Hansen Solubility Parameters: A User's Handbook, Second Edition

7248_C000.fm Page i Thursday, May 24, 2007 1:40 PM

HANSEN SOLUBILITY PARAMETERS A User’s Handbook Second Edition

7248_C000.fm Page ii Thursday, May 24, 2007 1:40 PM

7248_C000.fm Page iii Thursday, May 24, 2007 1:40 PM

HANSEN SOLUBILITY PARAMETERS A User’s Handbook Second Edition

Charles M. Hansen

7248_C000.fm Page iv Thursday, May 24, 2007 1:40 PM

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2007 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 0-8493-7248-8 (Hardcover) International Standard Book Number-13: 978-0-8493-7248-3 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http:// www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Hansen solubility parameters : a user’s handbook. -- 2nd ed. / edited by Charles Hansen. p. cm. Rev. ed. of: Hansen solubility parameters / Charles M. Hansen. c2000. Includes bibliographical references and index. ISBN 0-8493-7248-8 (alk. paper) 1. Solution (Chemistry) 2. Polymers--Solubility. 3. Thin films. I. Hansen, Charles M. II. Hansen, Charles M. Hansen solubility parameters. QD543.H258 2007 547’.70454--dc22 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

2006051083

7248_C000.fm Page v Thursday, May 24, 2007 1:40 PM

Contributors Dr. John Durkee Consultant in Critical and Metal Cleaning Hunt, Texas U.S.A. Dr. techn. Charles M. Hansen Consultant Hoersholm, Denmark Prof. Georgios M. Kontogeorgis Technical University of Denmark Department of Chemical Engineering Lyngby, Denmark Prof. Costas P anayiotou Department of Chemical Engineering University of Thessaloniki Thessaloniki, Greece Tim S. P oulsen Sr. Research Scientist Molecular Pathology Glostrup, Denmark

Dr. rer. nat. Hanno Priebe Sr. Research Scientist Chemical Development – Process Research GE Healthcare Amersham Health AS Oslo, Norway Per Redelius Research Manager Nynas Bitumen Product Technology Nynashamn, Sweden Prof. Laurie L. Williams Department of Ph ysics & Engineering Fort Lewis College Durango, Colorado U.S.A.

7248_C000.fm Page vi Thursday, May 24, 2007 1:40 PM

7248_C000.fm Page vii Thursday, May 24, 2007 1:40 PM

Preface to the First Edition My w ork with solv ents started in Denmark in 1962 when I w as a graduate student. The major results of this w ork were the realization that polymer film formation by sol ent evaporation took place in two distinct phases and the de velopment of what has come to be called Hansen solubility (or cohesion) parameters, abbre viated in the follo wing by HSP. The first phase of film formati by solvent evaporation is controlled by surf ace phenomena such as solv ent vapor pressure, wind velocity, heat transfer, etc., and the second phase is controlled by concentration-dependent diffusion of solvent molecules from within the film to the air sur ace. It is not controlled by the binding of solvent molecules to polymer molecules by h ydrogen bonding as w as pre viously thought. My solubility parameter work was actually started to define a finities between sol ent and polymer to help predict the de gree of this binding which w as thought to control solv ent retention. This was clearly a futile endeavor as there was absolutely no correlation. The solvents with smaller and more linear molecular structure dif fused out of the films more quickly than those with la ger and more branched molecular structure. HSP were de veloped in the process, ho wever. HSP ha ve been used widely since 1967 to accomplish correlations and to mak e systematic comparisons which one would not have thought possible earlier. The effects of hydrogen bonding, for e xample, are accounted for quantitati vely. Man y of these correlations are discussed later , including polymer solubility , swelling, and permeation; surf ace wetting and de wetting; solubility of inorganic salts; and biological applications including w ood, cholesterol, etc. The experimental limits on this seemingly universal ability to predict molecular affinities are apparently g verned by the limits represented by energies of the liquid test solvents themselves. There had/has to be a more satisfactory explanation of this uni versality than just “semiempirical” correlations. I decided to try to collect my e xperience for the purpose of a reference book, both for myself and for others. At the same time, a search of the major theories of polymer solution thermodynamics was undertak en to e xplore how the approaches compared. A k ey element in this w as to e xplain why the correlations all seemed to fit with an apparently “un versal” 4 (or 0.25 depending on which reference is used). This is described in more detail in Chapter 2 (Equation 2.5 and Equation 2.6). My present view is that the “4” is the result of the v alidity of the geometric mean rule to describe not only dispersion interactions but also permanent dipole–permanent dipole and hydrogen bonding (electron interchange) interactions in mixtures of unlik e molecules. The Hildebrand approach uses this and w as the basis of my earliest approach. The Prigogine corresponding states theory yields the “4” in the appropriate manner when the geometric mean rule is adopted (Chapter 2, Equation 2.11). Any other kind of averaging gives the wrong result. Considering these f acts and the massive amount of data that has been correlated using the “4” in the follo wing, it appears pro ven beyond a reasonable doubt that the geometric mean assumption is v alid not only for dispersion-type interactions (or perhaps more correctly in the present conte xt those interactions typical of aliphatic hydrocarbons) but also for permanent dipole–permanent dipole and h ydrogen bonding as well. For those who wish to try to understand the Prigogine theory , I recommend starting with an article by Donald P atterson.1 This article e xplains the corresponding states/free v olume theory of Prigogine and co workers in a much simpler form than in the original source. P atterson2 has also reviewed in understandable language the progression of developments in polymer solution thermodynamics from the Flory–Huggins theory, through that of Prigogine and coworkers, to the so-called “New Flory Theory.”3 Patterson also has been so kind as to aid me in the representations of the earlier theories as the y are presented here (especially Chapter 2). All of the pre vious theories and their extensions also can be found in a more recent book. 4 For this reason, these more classical

7248_C000.fm Page viii Thursday, May 24, 2007 1:40 PM

theories are not treated extensively as such in this book.The striking aspect about all of this previous work is that no one has dared to enter into the topic of h ydrogen bonding. The present quantitative treatment of permanent dipole–permanent dipole interactions and h ydrogen bonding is central to the results reported in e very chapter in this book. An attempt to relate this back to the pre vious theories is gi ven briefly here and more xtensively in Chapter 2. This attempt has been directed through Patterson,1 which may be called the Prigogine–Patterson approach, rather than through the Flory theory, as the relations with the former are more ob vious. I strongly recommend that studies be undertak en to confirm the usefulness of the “structura parameters” in the Prigogine theory (or the Flory theory). It is recognized that the effects of solvent molecular size, segment size, and polymer molecular size (and shapes) are not fully accounted for at the present time. There is hope that this can be done with structural parameters. The material presented here corresponds to my knowledge and experience at the time of writing, with all due respect to confidentiality agreements, etc I am greatly indebted to man y colleagues and supporters who ha ve understood that at times one can be so preoccupied and lost in deep thought that the present just seems not to e xist. Charles M. Hansen October 19, 1998

REFERENCES 1. Patterson, D., Role of Free Volume Changes in Polymer Solution Thermodynamics, J. Polym. Sci. Part C, 16, 3379–3389, 1968. 2. Patterson, D., Free Volume and Polymer Solubility . A Qualitati ve View, Macromolecules, 2(6), 672–677, 1969. 3. Flory, P. J., Thermodynamics of Polymer Solutions , Discussions of the Faraday Society, 49, 7–29, 1970. 4. Lipatov, Y. S. and Nestero v, A. E., Polymer Thermodynamics Library, Vol. 1, Thermodynamics of Polymer Blends, Technomic Publishing Co., Inc., Lancaster , PA, 1997.

7248_C000.fm Page ix Thursday, May 24, 2007 1:40 PM

Preface to the Second Edition When the question about a second edition of this handbook w as posed, I w as not in doubt that several additional authors were necessary to meet the demands it w ould require. The writings of the fi e contributors that were chosen speak for themselv es. There is theoretical impact in Chapter 3 (Costas Panayiotou) and in Chapter 4 (Georgios M. Kontogeorgis). Chapter 3 introduces statistical thermodynamics to confirm the d vision of cohesi ve ener gy into three parts enabling separate calculation of each. Chapter 4 describes ho w the Hansen solubility parameters (HSP) fit into othe theories of polymer solutions. The practical applications and understanding pro vided in Chapter 9 (Per Redelius) related to asphalt, bitumen, and crude oil should accelerate new thinking in this area and emphasize that simple e xplanations of seemingly comple x phenomena are usually the right ones. The thermodynamic treatment of carbon dioxide gi ven in Chapter 10 (Laurie L. Williams) is a model for similar w ork with other g ases and emphatically confirms the usefulness of Hanse solubility parameters for predicting the solubility beha vior of g ases in liquids and therefore also in polymers. Chapter 11 (John Durkee) goes through the process of demonstrating ho w “designer” solvents can be used in cleaning operations to replace, or partly replace, ozone-depleting solv ents, in spite of the problem of their HSP not being suf ficiently close to the HSP of the soils that are to be remved. I have added two chapters because of apparent need. Chapter 14 discusses environmental stress cracking (ESC). ESC is a major cause of unexpected and sometimes catastrophic failure of plastics. The recent impro ved understanding pro vided by HSP seemed appropriate for inclusion in this context. Chapter 16 discusses absorption and dif fusion in polymers. Many of the HSP correlations presented in this handbook cannot stand on HSP alone but must include consideration of absorption and diffusion of chemicals in polymers. These effects are often disguised by use of a molecular volume, as molecular size/volume correlates reasonably well with diffusion coefficients, especiall at low concentrations. Polymer surface layers are often significantly diferent from the bulk polymer. Surface mobility of polymer chain se gments plays an important role in surf ace dewetting, ESC, and resistance and/or delays to the absorption of chemicals. This chapter tries to unify the ef fects of a v erifiable sur ace resistance and v erifiable concentration-dependent di fusion coef ficients Solutions to the diffusion equation simultaneously considering these two effects explain the “anomalies” of absorption and also correctly model desorption phenomena, including the drying of a lacquer film from start to finis Each of the chapters in the first edition has been r viewed and added to where this w as felt appropriate without increasing the number of pages unduly . There is still a lack of significan activity in the biological area, in controlled release applications, and in other areas discussed in Chapter 18, such as nanotechnology . The relative affinity of molecules or s gments of molecules for each other can be predicted and in many cases controlled in self-assembly with the understanding provided by HSP. Chapter 15 treating biological materials has been e xpanded more than the others included in the first edition. This was done with the help of Tim Svenstrup Poulsen. Perhaps the most surprising of the additions in Chapter 15 is a HSP correlation for the (nonco valent) solvent interactions with DNA. The δD;δP;δH values of 19.0;20.0;11.0 for DN A, all in MP a1/2, clearly sho w that h ydrogen bonding interactions (H) contrib ute much less to the nonco valent interactions that determine the structure of the DN A than the dispersion (D) and dipolar interactions (P). Only about 14% of the cohesion energy involved in what is commonly called “h ydrogen bonding” derives from hydrogen bonding.

7248_C000.fm Page x Thursday, May 24, 2007 1:40 PM

Table Appendix A.1 is greatly e xpanded both in number and in information. The latter is due to the generous help of Hanno Priebe, the e xtent of which is clearly evident for those familiar with the first edition. There are close to 1200 entries in this table vs. the approximately 860 in the firs edition. However, please be advised that most of these are calculated and not e xperimental values as indicated in the comments to the table. Table Appendix A.2 is not greatly expanded. There have been too man y restrictions on what may be published to allo w any major e xpansion of this table. The majority of my work as a consultant has usually in volved agreements that prohibit or se verely limit publication of results paid for by pri vate sources. I have also included Appendix A.3 with the original solubility data on which the di vision of the ener gy was based. I ha ve regularly found this more specific data of considerable interest Once more resources and timing ha ve not been conduci ve to do a complete literature search to provide additional explanations of phenomena that should have had Hansen solubility parameters included in their interpretation. In view of the large expansion in the number of pages over the firs edition it is hoped that the principles, both theoretical and practical, are well illuminated. F or those who still lack information in a gi ven situation I can suggest a search using the k ey words “Hansen solubility parameters” followed by additional k ey words as required. This is true both for Internet searches as well as for searches in the more traditional literature. It has been satisfying to see ho w much can be interpreted with v ery simple observ ations and calculations. If it cannot be done simply , then rethink. I want to once more thank those who have contributed to this second edition. Let us hope others will take up the ef fort and relate their findings for the benefit of al Charles M. Hansen

7248_C000.fm Page xi Thursday, May 24, 2007 1:40 PM

The Author Charles M. Hansen consults on the topics co vered by this book. He works from his home in Hoersholm, 22 kilometers north of Copenhagen, Denmark. He received a BChE from the University of Louisville and an MS de gree from the Uni versity of Wisconsin. After being awarded the Dr. techn. de gree from the Technical University of Denmark in 1967, he held leading positions with PPG Industries in Pittsburgh, and as director of the Scandina vian P aint and Printing Ink Research Institute in Hoersholm, Denmark. Dr . Hansen dealt with polymers at FORCE Technology, Broendby, Denmark, for the 17 years prior to the start of the current state of semi-retirement. Dr. Hansen is perhaps best known for his extension of the Hildebrand solubility parameter to what are no w called Hansen solubility parameters. These have been found mutually confirming with the I. Prigogine corresponding states theor of polymer solutions and can be used to directly calculate the Flory–Huggins interaction coefficient The statistical thermodynamics approach de veloped by Costas P anayiotou and co workers, which is reported in Chapter 3 of this second edition, also confirms the viability of the d vision of the cohesion energy into separate parts, and allo ws their independent calculation. Dr. Hansen has published widely in the fields of polymer solubilit , diffusion and permeation in polymers and films, sur ace science, and coatings science. He is currently vice president of the Danish Society for Polymer Technology, having recently completed a 5-year period as president. He frequently re views papers for leading journals, and is on the editorial board of Progress in Organic Coatings, as well as being a member of the DanishAcademy of Technical Sciences (ATV).

7248_C000.fm Page xii Thursday, May 24, 2007 1:40 PM

7248_C000.fm Page xiii Thursday, May 24, 2007 1:40 PM

Key to Symbols Note: The symbols used in Chapters 3 and 16 are so numerous and dif ferent that the y have been placed in these chapters, respecti vely. A12 D D DM ED EP EH ΔEv G G ΔGM ΔGMnoncomb H ΔHv ΔHM KH L P P P Q P* R Ra RA RM Ro RED S ΔSM T T Tb Tc Tr V

Energy difference defined by Chapter 2, Equation 2.1 Diffusion coefficient in Chapter1 Dispersion cohesion (solubility) parameter — in tables and computer printouts Dipole moment — debyes Dispersion cohesion ener gy Polar cohesion ener gy Hydrogen bonding cohesion ener gy Energy of vaporization (=) cohesion ener gy Number of “good” solv ents in a correlation, used in tables of correlations Gibbs Energy in Chapter 4 Molar free ener gy of mixing Noncombinatorial molar free ener gy of mixing Hydrogen bonding cohesion (solubility) parameter — in tables and computer printouts Molar heat of v aporization Molar heat of mixing Henry’s law constant in Equation 10.5 Ostwald coefficient in Equation 10. Permeation coefficient in Chapter 1 Polar cohesion (solubility) parameter — in tables and computer printouts Pressure in Chapter 10 Solvent quality number Total pressure, atm. (Chapter 13, Figures 13.4 and 13.5) Gas constant (1.987 cal/mol K) Distance in Hansen space, see Chapter 1, Equation 1.9 or Chapter 2, Equation 2.5 Distance in Hansen space, see Chapter 2, Equation 2.7 Maximum distance in Hansen space allo wing solubility (or other “good” interaction) Radius of interaction sphere in Hansen space Relative energy difference (Chapter 1, Equation 1.10) Solubility coefficient in Chapter 1 Molar entropy of mixing Absolute temperature “Total” number of solv ents used in a correlation as gi ven in tables (Normal) boiling point, de grees K Critical temperature, de grees K Reduced temperature, Chapter 1, Equation 1.12 Molar volume, cm 3/gram molecular weight

7248_C000.fm Page xiv Thursday, May 24, 2007 1:40 PM

V Vf V* VW VM a ai bi b c c ci f fi fi0 i k k n n nD p pi pis r r ts x y H Ω ΩI ∞ Σ ΔT α α β β δD δH δP δt δ ε ε γ γ

Total volume in Chapter 4 Free volume (Equation 4.2) Hard core or close pack ed volume in Equation 4.2 van der Waals volume Volume of mixture Constant in van der Waals equation of state (Chapter 4) Activity coefficient of the “i”th component in Appendix 10.A.1 Coefficients in Equations 10.17 and 10.1 Constant in van der Waals equation of state (Chapter 4) Dispersion cohesion ener gy density from Chapter 1, Figure 1.2 or Figure 1.3 Concentration in Chapter 8, Equation 8.4 Coefficients (state constants) in Equations 10.17 and 10.1 Fractional solubility parameters, defined by Chapter 5, Equations 5.1 to 5. Fugacity of the “i”th component in Appendix 10.A.1 Fugacity at standard state in Appendix 10.A.1 Component “i” in a mixture Constant in Equation 6.1 Constant in Equations 10.21–10.23 Coefficient in Equation 10.1 Coefficient in Equaitons 10.21, 10.22, and 10.2 Index of refraction in Equation 10.25 Partial pressure (of carbon dioxide) in Chapter 10 Partial pressure of the “i”th component in Appendix 10.A.1 Saturation pressure of the “i”th component in Appendix 10.A.1 Number of se gments in a gi ven molecule, Chapter 2 Ratio of polymer v olume to solv ent volume (Chapter 4) Sedimentation time, see Chapter 7, Equation 7.1 Mole fraction in liquid phase (Chapter 13, Figures 13.4 and 13.5, and Chapter 10) Mole fraction in vapor phase (Chapter 13, Figures 13.4 and 13.5, and Chapter 10) Ratio of cohesive energy densities; Chapter 2, Equation 2.6 Bunsen coefficient (Equation 10.6 Infinite dilution act vity coefficien Summation Lydersen critical temperature group contrib ution Thermal expansion coefficien Constant in Equation 4.15 Constant in Chapter 2, Equation 2.1 Compressibility in Chapter 10 Dispersion cohesion (solubility) parameter Hydrogen bonding cohesion (solubility) parameter Polar cohesion (solubility) parameter Total (Hildebrand) cohesion (solubility) parameter Prigogine normalized interaction parameter , Chapter 2, Equation 2.8 Cohesive energy for a polymer se gment or solv ent in Chapter 2 Dielectric constant in Equation 10.25 Surface free ener gy of a liquid in air or its o wn vapor Activity coefficient in Chapter

7248_C000.fm Page xv Thursday, May 24, 2007 1:40 PM

η ηs ηo [η] [η]N ϕi μ ν Θ Θa Θr ρ ρ ρ ρp ρs σ χ χ12 χc χlit χs 1 2 D P H d p h

Viscosity of solv ent, Chapter 7, Equation 7.1 Viscosity of solution Viscosity of solv ent Intrinsic viscosity, see Chapter 8, Equation 8.4 Normalized intrinsic viscosity Volume fraction of component “i” Dipole moment Interaction parameter, see Chapter 2, Equation 2.11 Contact angle between liquid and surf ace Advancing contact angle Receding contact angle Prigogine parameter for dif ferences is size in polymer se gments and solv ent, Chapter 2, Equation 2.10 Density in Chapter 7, Equation 7.1 Density in Chapter 10 Particle density in Chapter 7, Equation 7.1 Solvent density in Chapter 7, Equation 7.1 Prigogine segmental distance parameter, Chapter 2, Equation 2.10 Polymer–liquid interaction parameter (Flory–Huggins), Chapter 2 Interaction parameter — “Ne w Flory Theory” Critical polymer–liquid interaction parameter , Chapter 2 Representative χ value from general literature Entropy component of χ (Subscript) indicates a solv ent (Subscript) indicates a polymer (or second material in contact with a solv ent) (Subscript) dispersion component (Subscript) polar component (Subscript) hydrogen bonding component (Subscript) dispersion component (Subscript) polar component (Subscript) hydrogen bonding component

7248_C000.fm Page xvi Thursday, May 24, 2007 1:40 PM

7248_C000.fm Page xvii Thursday, May 24, 2007 1:40 PM

Table of Contents Chapter 1

Solubility Parameters — An Introduction ...................................................................1

Abstract ..............................................................................................................................................1 Introduction ........................................................................................................................................1 Hildebrand Parameters and Basic Polymer Solution Thermodynamics ...........................................2 Hansen Solubility P arameters ............................................................................................................4 Methods and Problems in the Determination of P artial Solubility P arameters ...............................6 Calculation of the Dispersion Solubility P arameter δD ...................................................................13 Calculation of the Polar Solubility P arameter δP ............................................................................16 Calculation of the Hydrogen Bonding Solubility P arameter δH .....................................................17 Supplementary Calculations and Procedures ..................................................................................17 Temperature Dependence .......................................................................................................18 Some Special Effects Temperature Changes .........................................................................19 Effects of Solv ent Molecular Size .........................................................................................19 Computer Programs ................................................................................................................20 Hansen Solubility P arameters for Water .........................................................................................21 Conclusion........................................................................................................................................22 References ........................................................................................................................................24 Chapter 2

Theory — The Prigogine Corresponding States Theory, χ12 Interaction Parameter, and Hansen Solubility P arameters ...........................................................27

Abstract ............................................................................................................................................27 Introduction ......................................................................................................................................27 Hansen Solubility P arameters (HSP) ...............................................................................................28 Resemblance between Predictions of Hansen Solubility P arameters and Corresponding States Theories...............................................................................................30 The χ12 Parameter and Hansen Solubility P arameters.....................................................................32 Comparison of Calculated and Experimental χ12 Parameters .........................................................34 Polybutadiene .........................................................................................................................35 Polyisobutylene.......................................................................................................................36 Polystyrene .............................................................................................................................38 Polyvinylacetate......................................................................................................................39 Polyacrylonitrile .....................................................................................................................39 General Discussion ..........................................................................................................................39 Postscript ..........................................................................................................................................40 Conclusion........................................................................................................................................41 References ........................................................................................................................................42 Chapter 3

Statistical Thermodynamic Calculations of the Hydrogen Bonding, Dipolar, and Dispersion Solubility P arameters..........................................................45

Key words ........................................................................................................................................45 Abstract ............................................................................................................................................45 Introduction ......................................................................................................................................45

7248_C000.fm Page xviii Thursday, May 24, 2007 1:40 PM

Theory ..............................................................................................................................................46 The Equation-of-State Framework.........................................................................................46 The Contribution from Dipolar F orces ..................................................................................50 Applications .....................................................................................................................................52 Discussion and Conclusions ............................................................................................................59 Acknowledgments ............................................................................................................................62 List of Symbols Special to this Chapter ..........................................................................................63 References ........................................................................................................................................64 Appendix 3.I: The Acid Dimerization .............................................................................................65 Appendix 3.II: An Alternative Form of the Polar Term..................................................................66 Appendix 3.III: A Group-Contribution Method for the Prediction of δ and δD.............................66 Chapter 4

The Hansen Solubility P arameters (HSP) in Thermodynamic Models for Polymer Solutions ......................................................................................................75

Abstract ............................................................................................................................................75 Group Contribution Methods for Estimating Properties of Polymers ............................................76 The Group-Contribution Principle and Some Applications (Density, Solubility Parameters) ................................................................................................76 GC Free-Volume-Based Models for Polymers (Entropic-FV , Unifac-FV) ...........................77 The Free-Volume Concept .........................................................................................77 The UNIFAC-FV Model ............................................................................................77 The Entropic Model ...................................................................................................78 The Flory–Huggins Model and the Re gular Solution Theory ..............................................80 Rules of Thumb and Solv ent Selection Using the Flory–Huggins Model and Solubility Parameters ..................................................................................81 Activity Coefficients Models Using the HS ..................................................................................82 Flory–Huggins Models Using Hildebrand and Hansen Solubility P arameters (HSP) .........82 The FH/Hansen Model vs. the GC Methods .............................................................84 Applications............................................................................................................................85 Solvent Selection for P aints (Activity Coefficients at Infinite Dilutio ..................85 Mixed Solvent–Polymer Phase Equilibria .................................................................88 Conclusions and Future Challenges ................................................................................................90 List of Abbreviations........................................................................................................................91 Symbols in this Chapter ...................................................................................................................92 Appendix 4.I: An Expression of the Flory–Huggins Model for Multicomponent Mixtures .........92 References ........................................................................................................................................93 Chapter 5

Methods of Characterization — Polymers ................................................................95

Abstract ............................................................................................................................................95 Introduction ......................................................................................................................................95 Calculation of Polymer HSP ...........................................................................................................97 Solubility — Examples ....................................................................................................................98 Swelling — Examples ...................................................................................................................106 Melting Point Determinations — Ef fect of Temperature..............................................................106 Environmental Stress Cracking ......................................................................................................107 Intrinsic Viscosity Measurements ..................................................................................................107 Other Measurement Techniques ....................................................................................................109 Conclusion......................................................................................................................................109 References ......................................................................................................................................110

7248_C000.fm Page xix Thursday, May 24, 2007 1:40 PM

Chapter 6

Methods of Characterization — Surf aces................................................................113

Abstract ..........................................................................................................................................113 Introduction ....................................................................................................................................113 Hansen Solubility P arameter Correlations with Surf ace Tension (Surface Free Ener gy)............113 Method to Evaluate the Cohesion Ener gy Parameters for Surf aces.............................................114 A Critical View of the Critical Surf ace Tensions..........................................................................116 A Critical View of the Wetting Tension ........................................................................................117 Additional Hansen Solubility P arameter Surface Characterizations and Comparisons ...............118 Self-Stratifying Coatings................................................................................................................120 Maximizing Physical Adhesion .....................................................................................................122 Conclusion......................................................................................................................................122 References ......................................................................................................................................122 Chapter 7

Methods of Characterization for Pigments, Fillers, and Fibers ..............................125

Abstract ..........................................................................................................................................125 Introduction ....................................................................................................................................125 Methods to Characterize Pigment, Filler , and Fiber Surf aces ......................................................126 Discussion — Pigments, Fillers, and Fibers .................................................................................127 Hansen Solubility P arameter Correlation of Zeta Potential for Blanc Fix e.................................131 Carbon Fiber Surf ace Characterization .........................................................................................131 Controlled Adsorption (Self-Assembly) ........................................................................................132 Conclusion......................................................................................................................................134 References ......................................................................................................................................134 Chapter 8

Applications — Coatings and Other Filled Polymer Systems ................................137

Abstract ..........................................................................................................................................137 Introduction ....................................................................................................................................137 Solvents ..........................................................................................................................................137 Techniques for Data Treatment......................................................................................................142 Solvents and Surf ace Phenomena in Coatings (Self-Assembly) ..................................................144 Polymer Compatibility ...................................................................................................................145 Hansen Solubility P arameter Principles Applied to Understanding Other Filled Polymer Systems ..................................................................................................................147 Conclusion......................................................................................................................................147 References ......................................................................................................................................148 Chapter 9

Hansen Solubility P arameters of Asphalt, Bitumen, and Crude Oils .....................151

Abstract ..........................................................................................................................................151 Symbols Special to Chapter 9 .......................................................................................................151 Introduction ....................................................................................................................................151 Models of Bitumen ........................................................................................................................152 Asphaltenes ....................................................................................................................................154 Molecular Weight .................................................................................................................154 Polarity..................................................................................................................................155 Solubility Parameters of Bitumen ..................................................................................................155 Testing of Bitumen Solubility ........................................................................................................156 Hildebrand Solubility P arameters ..................................................................................................156 Hansen Solubility P arameters (HSP) .............................................................................................158

7248_C000.fm Page xx Thursday, May 24, 2007 1:40 PM

The Solubility Sphere ....................................................................................................................159 Computer Program for Calculation and Plotting of the Hansen 3D Pseudosphere .....................161 Components of Bitumen ................................................................................................................164 Bitumen and Polymers ...................................................................................................................166 Crude Oil ........................................................................................................................................169 Turbidimetric Titrations .................................................................................................................170 BISOM Test ...................................................................................................................................170 Conclusion......................................................................................................................................173 References ......................................................................................................................................174 Chapter 10 Determination of Hansen Solubility P arameter Values for Carbon Dioxide ..........177 Abstract ..........................................................................................................................................177 Introduction ....................................................................................................................................177 Methodology ..................................................................................................................................178 One-Component Hildebrand P arameter as a Function of Temperature and Pressure ..................187 Three-Component (Hansen) Solubility P arameters — Pure CO 2 .................................................189 Temperature and Pressure Ef fects on HSPs: δd.............................................................................190 Temperature and Pressure Ef fects on HSPs: δp.............................................................................191 Temperature and Pressure Ef fects on HSPs: δh.............................................................................191 Conclusion......................................................................................................................................196 Acknowledgments ..........................................................................................................................196 Chapter 10 Addendum ...................................................................................................................196 Symbols Special to this Chapter ....................................................................................................197 References ......................................................................................................................................197 Appendix 10.A.1: Ideal Solubility of Gases in Liquids and Published CO 2 Solubility Data .....199 Ideal Solubility of Gases in Liquids ..............................................................................................199 References ......................................................................................................................................201 Chapter 11 Use of Hansen Solubility P arameters to Identify Cleaning Applications for “Designer” Solvents .................................................................................................203 Abstract ..........................................................................................................................................203 Introduction ....................................................................................................................................203 A Variety of Solv ents.....................................................................................................................204 Pathology of Soils ..........................................................................................................................204 HSP of Multiple-Component Soils ................................................................................................204 Method for Calculating HSP of Composites (Soils or Solv ents) .................................................205 More Realistic View about Evaluating HSP of Composite Soils .................................................206 Method for Choice of Suitable Solv ents .......................................................................................206 Reference Soils for Comparison ....................................................................................................208 Identification of Designer Sol ents ...............................................................................................208 An Open Question — Answered ...................................................................................................208 Limiting R A Value for Expected Good Cleaning Performance ....................................................210 Application of HSP Methodology to Cleaning Operations ..........................................................212 Analysis of Capability of Designer Solv ents ................................................................................213 Conclusions ....................................................................................................................................215 Notes ..............................................................................................................................................227

7248_C000.fm Page xxi Thursday, May 24, 2007 1:40 PM

Chapter 12 Applications — Chemical Resistance ......................................................................231 Abstract ..........................................................................................................................................231 Introduction ....................................................................................................................................231 Chemical Resistance — Acceptable-or-Not Data .........................................................................232 Effects of Solv ent Molecular Size .................................................................................................232 Chemical Resistance — Examples ................................................................................................233 Tank Coatings .......................................................................................................................233 PET Film Coating ................................................................................................................234 Acceptable or Not — Plastics ..............................................................................................234 Tensile Strength ....................................................................................................................237 Special Effects with Water.............................................................................................................238 Conclusion......................................................................................................................................239 References ......................................................................................................................................240 Chapter 13 Applications — Barrier Polymers ............................................................................243 Abstract ..........................................................................................................................................243 Introduction ....................................................................................................................................243 Concentration-Dependent Diffusion ..............................................................................................244 Solubility Parameter Correlations Based on Permeation Phenomena ..........................................245 Solubility Parameter Correlations of Breakthrough Times .................................................245 Solubility Parameter Correlation of Permeation Rates .......................................................248 Solubility Parameter Correlation of Polymer Swelling ................................................................250 Solubility Parameter Correlation of Permeation Coef ficients for Gase ......................................251 Laminates..............................................................................................................................253 General Considerations ..................................................................................................................255 Conclusion......................................................................................................................................256 References ......................................................................................................................................257 Chapter 14 Applications — Environmental Stress Cracking in Polymers ................................259 Abstract ..........................................................................................................................................259 Introduction ....................................................................................................................................259 ESC Interpreted Using HSP ..........................................................................................................260 ESC with Nonabsorbing Stress Cracking Initiators ......................................................................263 Discussion ......................................................................................................................................264 Conclusion......................................................................................................................................267 References ......................................................................................................................................267 Chapter 15 Hansen Solubility P arameters — Biological Materials ...........................................269 Abstract ..........................................................................................................................................269 Introduction ....................................................................................................................................270 Hydrophobic Bonding and Hydrophilic Bonding (Self-Association) ...........................................271 DNA ..............................................................................................................................................273 Cholesterol .....................................................................................................................................275 Lard ................................................................................................................................................277

7248_C000.fm Page xxii Thursday, May 24, 2007 1:40 PM

Human Skin....................................................................................................................................277 Proteins — Blood Serum and Zein ...............................................................................................279 Chlorophyll and Lignin ..................................................................................................................279 Wood Chemicals and Polymers .....................................................................................................279 Urea ..............................................................................................................................................283 Water ..............................................................................................................................................289 Surface Mobility ............................................................................................................................290 Chiral Rotation, Hydrogen Bonding, and Nanoengineering .........................................................290 Conclusion......................................................................................................................................291 References ......................................................................................................................................291 Chapter 16 Absorption and Diffusion in Polymers ....................................................................293 Abstract ..........................................................................................................................................293 List of Symbols Used in This Chapter ..........................................................................................293 Introduction ....................................................................................................................................294 Steady State Permeation ................................................................................................................296 The Diffusion Equation..................................................................................................................296 Constant Diffusion Coefficient ...........................................................................................296 Concentration Dependent Diffusion Coefficient ................................................................297 Surface Resistance .........................................................................................................................298 Mathematical Background....................................................................................................298 Surface Resistance in Absorption Experiments ...................................................................300 Surface Resistance in Permeation Experiments ..................................................................301 Surface Resistance — A Discussion ....................................................................................302 Side Effects ....................................................................................................................................304 Measuring Diffusion Coefficients with Sur ace Resistance and Concentration Dependence.......................................................................................304 Film Formation by Solv ent Evaporation .......................................................................................305 Anomalous Diffusion (Case II, Super Case II) .............................................................................306 General Comments .........................................................................................................................308 Conclusion......................................................................................................................................308 References ......................................................................................................................................309 Chapter 17 Applications — Safety and En vironment ................................................................311 Abstract ..........................................................................................................................................311 Introduction ....................................................................................................................................311 Substitution.....................................................................................................................................311 Alternative Systems .......................................................................................................................312 Solvent Formulation and Personal Protection for Least Risk .......................................................313 The Danish Mal System — The Fan.............................................................................................313 Selection of Chemical Protecti ve Clothing ...................................................................................315 Uptake of Contents by a Plastic Container ...................................................................................315 Skin Penetration .............................................................................................................................316 Transport Phenomena.....................................................................................................................316 Conclusion......................................................................................................................................317 References ......................................................................................................................................318

7248_C000.fm Page xxiii Thursday, May 24, 2007 1:40 PM

Chapter 18 The Future ................................................................................................................321 Abstract ..........................................................................................................................................321 Introduction ....................................................................................................................................321 Hansen Solubility P arameter Data and Data Quality ....................................................................324 Group Contribution Methods .........................................................................................................328 Polymers as Points — Solv ents as Spheres ..................................................................................328 Characterizing Surfaces .................................................................................................................330 Materials and Processes Suggested for Further Attention ............................................................332 Surface Active Agents ..........................................................................................................332 Surface Mobility (Self-Assembly) .......................................................................................333 Water.....................................................................................................................................334 Gases.....................................................................................................................................336 Organic Salts ........................................................................................................................337 Inorganic Salts ......................................................................................................................337 Organometallic Compounds .................................................................................................338 Aromas and Fragrances ........................................................................................................338 Absorption of Chemicals in Plastics ....................................................................................339 Chemical Resistance.............................................................................................................339 Controlled Release................................................................................................................339 Nanotechnology....................................................................................................................340 Theoretical Problems Awaiting Future Resolution ........................................................................341 Polymer Solubility ................................................................................................................341 Surface Phenomena ..............................................................................................................342 Conclusion......................................................................................................................................342 References ......................................................................................................................................342 Appendix A: Comments to Table A.1 ...........................................................................................345 References ......................................................................................................................................346 Table A.1 ........................................................................................................................................347 Appendix A: Comments to Table A.2 ...........................................................................................485 References ......................................................................................................................................490 List of Trade Names and Suppliers ...............................................................................................491 Table A.2 ........................................................................................................................................493 Appendix A: Comments to Table A.3 ...........................................................................................507 Table A.3 ........................................................................................................................................508 Index...............................................................................................................................................511

7248_C000.fm Page xxiv Thursday, May 24, 2007 1:40 PM

7248_C001.fm Page 1 Monday, April 23, 2007 1:56 PM

Parameters — 1 Solubility An Introduction Charles M. Hansen ABSTRACT Solubility parameters have found their greatest use in the coatings industry to aid in the selection of solv ents. They are used in other industries, ho wever, to predict compatibility of polymers, chemical resistance, and permeation rates, and even to characterize the surfaces of pigments, fibers and fillers. Liquids with similar solubility parameters will be miscible, and polymers will dissol e in solvents whose solubility parameters are not too dif ferent from their o wn. The basic principle has been “like dissolves like.” More recently, this has been modified to “li e seeks like,” as many surface characterizations ha ve also been made, and surf aces do not (usually) dissolv e. Solubility parameters help put numbers into this simple qualitati ve idea. This chapter describes the tools commonly used in Hansen solubility parameter (HSP) studies. These include liquids used as energy probes and computer programs to process data. The goal is to arri ve at the HSP for interesting materials either by calculation or , if necessary , by e xperiment and preferably with agreement between the tw o.

INTRODUCTION The solubility parameter has been used for man y years to select solv ents for coatings materials. A lack of total success has stimulated further research. The skill with which solvents can be optimally selected with respect to cost, solv ency, workplace environment, external environment, evaporation rate, flash point, etc., has impr ved over the years as a result of a series of impro vements in the solubility parameter concept and widespread use of computer techniques. Most commercial suppliers of solv ents ha ve computer programs to help with solv ent selection. One can no w easily predict how to dissolve a given polymer in a mixture of two solvents, neither of which can dissolve the polymer by itself. Unfortunately, this book cannot include discussion of all the significant e forts leading to our present knowledge of the solubility parameters.An attempt is made to outline developments, provide some background for a basic understanding, and gi ve examples of uses in practice. The key factor is to determine those af finities that the important components in a system h ve for each other. For many products this means e valuating or estimating the relati ve af finities of sol ents, polymers, additives, pigment surf aces, filler sur aces, fiber sur aces, and substrates. It is note worthy that the concepts presented here ha ve de veloped to ward not just predicting solubility that requires high affinity between sol ent and solute, but for predicting affinities betwee different polymers, leading to compatibility , and af finities to sur aces to impro ve dispersion and adhesion. In these applications the solubility parameter has become a tool, using well-define liquids as energy probes, to measure the similarity, or lack of the same, of key components. Materials with widely different chemical structures may be v ery close in affinities. Only those materials tha interact differently with dif ferent solvents can be characterized in this manner . It can be e xpected that many inorganic materials, such as fillers, will not interact di ferently with these energy probes

1

7248_C001.fm Page 2 Monday, April 23, 2007 1:56 PM

2

Hansen Solubility Parameters: A User’s Handbook

as their energies are very much higher. An adsorbed layer of w ater on the high-energy surface can also play an important role. Re gardless of these concerns, it has been possible to characterize pigments, both or ganic and inor ganic, as well as fillers li e barium sulf ate, zinc oxide, etc., and also inorganic fibers (see Chapter 7). Changing the sur ace energies by various treatments can lead to a surf ace that can be characterized more readily and often interacts more strongly with gi ven organic solvents. When the same solvents that dissolve a polymeric binder are those which interact most strongly with a surf ace, it can be e xpected that the binder and the surf ace have high affinit for each other. Solubility parameters are sometimes called cohesion energy parameters as the y are deri ved from the energy required to convert a liquid to a gas. The energy of vaporization is a direct measure of the total (cohesi ve) energy holding the liquid’ s molecules together. All types of bonds holding the liquid together are brok en by e vaporation, and this has led to the concepts described in more detail later . The term cohesion energy parameter is more appropriately used when referring to surface phenomena.

HILDEBRAND PARAMETERS AND BASIC POLYMER SOLUTION THERMODYNAMICS The term solubility parameter w as first used by Hildebrand and Scott 1,2 The earlier w ork of Scatchard and others w as contributory to this de velopment. The Hildebrand solubility parameter is defined as the square root of the cohes ve energy density: δ = (E/V)1/2

(1.1)

Where V is the molar volume of the pure solvent, and E is its (measurable) energy of vaporization (see Equation 1.15). The numerical v alue of the solubility parameter in MP a1/2 is 2.0455 times larger than that in (cal/cm 3)1/2. The solubility parameter is an important quantity for predicting solubility relations, as can be seen from the follo wing brief introduction. Thermodynamics requires that the free ener gy of mixing must be zero or ne gative for the solution process to occur spontaneously . The free ener gy change for the solution process is gi ven by the relation: ΔGM = ΔHM – ΔTSM

(1.2)

where ΔGM is the free energy of mixing, ΔHM is the heat of mixing, T is the absolute temperature, and ΔSM is the entrop y change in the mixing process. Equation 1.3 gives the heat of mixing as proposed by Hildebrand and Scott: ΔHM = ϕ1ϕ2VM(δ1 – δ2)2

(1.3)

The φ1 and φ2 are volume fractions of solvent and polymer, and VM is the volume of the mixture. Equation 1.3 is not correct, and it has often been cited as a shortcoming of this theory in that only positive heats of mixing are allo wed. It has been sho wn by Patterson, Delmas, and co workers that ΔGMnoncomb is given by the right-hand side of Equation 1.3 and not ΔGM. This is discussed more in Chapter 2. The correct relation is 3–8: ΔGMnoncomb = ϕ1ϕ2VM(δ1 – δ2)2

(1.4)

The noncombinatorial free energy of solution, ΔGMnoncomb, includes all free energy effects other than the combinatorial entropy of solution that results by simply mixing the components. Equation

7248_C001.fm Page 3 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

3

1.4 is consistent with the Prigogine corresponding states theory (CST) of polymer solutions (see Chapter 2) and can be dif ferentiated to gi ve e xpressions3,4 predicting both positi ve and ne gative heats of mixing. Therefore, both positi ve and ne gative heats of mixing can be e xpected from theoretical considerations and ha ve been measured accordingly . It has been clearly sho wn that solubility parameters can be used to predict both positi ve and ne gative heats of mixing. Pre vious objections to the ef fect that only positi ve values are allowed in this theory are incorrect. This discussion clearly demonstrates wh y the solubility parameter should be considered as a free energy parameter. This is more in agreement with the use of the solubility parameter plots to follow. These use solubility parameters as ax es and have experimentally determined boundaries of solubility defined by the act that the free ener gy of mixing is zero. The combinatorial entrop y enters as a constant f actor in the plots of solubility in dif ferent solv ents, for e xample, as the concentrations are usually constant for a gi ven study. It is important to note that the solubility parameter , or rather the dif ference in solubility parameters for the solv ent–solute combination, is important in determining the solubility of the system. It is clear that a match in solubility parameters leads to a zero change in noncombinatorial free energy, and the positi ve entropy change (the combinatorial entrop y change), found on simple mixing to result in a disordered mixture compared to the pure components, will ensure that a solution is possible from a thermodynamic point of vie w. The maximum dif ference in solubility parameters that can be tolerated where the solution still occurs is found by setting the noncombinatorial free ener gy change equal to the combinatorial entrop y change: ΔGMnoncomb = TΔSMcomb

(1.5)

This equation clearly sho ws that an alternate vie w of the solubility situation at the limit of solubility is that it is the entrop y change that dictates ho w closely the solubility parameters must match each other for the solution to occur . It will be seen in Chapter 2 that solv ents with smaller molecular v olumes will be thermodynamically better than lar ger ones ha ving identical solubility parameters. A practical aspect of this effect is that solv ents with relati vely low molecular v olumes, such as methanol and acetone, can dissolve a polymer at larger solubility parameter differences than might be expected from comparisons with other solv ents with larger molecular volumes. An average solvent molecular volume is usually taken as about 100 cc/mol. The converse is also true. Lar ger molecular species may not dissolve, even though solubility parameter considerations might predict the y would. This can be a difficulty in predicting the beh vior of plasticizers solely based on data for lower molecular weight solvents. These effects are also discussed elsewhere in this book, particularly in Chapter 2, Chapter 12, Chapter 13, and Chapter 16. A shortcoming of the earlier solubility parameter w ork is that the approach w as limited to regular solutions, as defined by Hildebrand and Scott2 and does not account for association between molecules, such as those that polar and h ydrogen-bonding interactions w ould require. The latter problem seems to ha ve been largely solved with the use of multicomponent solubility parameters; however, the lack of accurac y with which the solubility parameters can be assigned will al ways remain a problem. Using the dif ference between two large numbers to calculate a relati vely small heat of mixing, for e xample, will al ways be problematic. A more detailed description of the theory presented by Hildebrand, and the succession of research reports that have attempted to improve on it, can be found in Barton’s extensive handbook.9 The slightly older , e xcellent contrib ution of Gardon and Teas10 is also a good source of related information, particularly for coatings and adhesion phenomena. The approach of Burrell, 11 who divided solv ents into h ydrogen bonding classes, has found numerous practical applications; the approach of Blanks and Prausnitz 12 divided the solubility parameter into tw o components, “nonpolar” and “polar.” Both are w orthy of mention, ho wever, in that the first has found wide use an the second greatly influenced the author s earlier activities. The Prausnitz article, in particular, was

7248_C001.fm Page 4 Monday, April 23, 2007 1:56 PM

4

Hansen Solubility Parameters: A User’s Handbook

farsighted in that a corresponding states procedure was introduced to calculate the dispersion energy contribution to the cohesi ve energy. This is discussed in more detail in Chapter 2. It can be seen from Equation 1.2 that the entrop y change is beneficial to mixing. When multiplied by the temperature, this will w ork in the direction of promoting a more ne gative free energy of mixing. This is the usual case, although there are e xceptions. Increasing temperature does not always lead to impro ved solubility relations. Indeed, this w as the basis of the pioneering work of Patterson and coworkers,3–8 to show that subsequent increases in temperature can predictably lead to insolubility. Their work was done in essentially nonpolar systems. Increasing temper ature can also lead to a nonsolv ent becoming a solvent and, subsequently, a nonsolvent again with still further increase in temperature. Polymer solubility parameters do not change much with temperature, but those of a liquid frequently decrease rapidly with temperature.This situation allows a nonsolvent, with a solubility parameter that is initially too high, to pass through a soluble condition to once more become a nonsolv ent as the temperature increases. These are usually “boundary” solvents on solubility parameter plots. The entropy changes associated with polymer solutions will be smaller than those associated with liquid–liquid miscibility, for example, as the “monomers” are already bound into the config uration dictated by the polymer the y mak e up. They are no longer free in the sense of a liquid solvent and cannot mix freely to contrib ute to a lar ger entropy change. This is one reason polymer–polymer miscibility is dif ficult to achi ve. The free ener gy criterion dictates that polymer solubility parameters match e xtremely well for mutual compatibility , as there is little to be g ained from the entrop y contribution when progressi vely larger molecules are in volved. However, polymer–polymer miscibility can be promoted by the introduction of suitable copolymers or comonomers that interact specifically within the system. Further discussion of these phenomena is b yond the scope of the present discussion; ho wever, see Chapter 5.

HANSEN SOLUBILITY PARAMETERS A solubility parameter approach proposed by the author for predicting polymer solubility has been in wide use. The basis of these so-called HSPs is that the total ener gy of v aporization of a liquid consists of se veral individual parts. 13–17 These arise from (atomic) dispersion forces, (molecular) permanent dipole–permanent dipole forces, and (molecular) hydrogen bonding (electron exchange). Needless to say, without the work of Hildebrand and Scott 1,2 and others not specifically reference here, such as Scatchard, this postulate could ne ver have been made. The total cohesive energy, E, can be measured by e vaporating the liquid, i.e., breaking all the cohesi ve bonds. Thus the total cohesive energy is considered as being identical to the ener gy of v aporization. It should also be noted that these cohesive energies arise from interactions of a given solvent molecule with another of its o wn kind. The basis of the approach is, therefore, v ery simple, and it is surprising that so many different applications ha ve been possible since 1967 when the idea w as first published. A rather large number of applications are discussed in this book. Others are found in the w orks of Barton.9 A lucid discussion by Barton 18 enumerates typical situations where problems occur when using solubility parameters. These appear most often where the en vironment causes the solv ent molecules to interact, with or within themselv es, dif ferently from the w ay the y do in situations where they make up their own environment, i.e., as pure liquids. Several cases are discussed where appropriate in the follo wing chapters. Materials with similar HSP ha ve high af finity for each othe . The extent of the similarity in a given situation determines the e xtent of the interaction. The same cannot be said of the total or Hildebrand solubility parameter.1,2 Ethanol and nitromethane, for e xample, have similar total solubility parameters (26.1 vs. 25.1 MPa1/2, respectively), but their affinities are quite di ferent. Ethanol is water soluble, whereas nitromethane is not. Indeed, mixtures of nitroparaf fins and alcohols wer demonstrated in man y cases to pro vide syner gistic mixtures of tw o nonsolv ents that dissolv ed

7248_C001.fm Page 5 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

5

polymers.13 This could ne ver ha ve been predicted by Hildebrand parameters, whereas the HSP concept readily confirms the reason for this e fect. There are three major types of interactions in common or ganic materials. The most general are the nonpolar interactions. These are derived from atomic forces and have also been called dispersion interactions in the literature. As molecules are b uilt up from atoms, all molecules contain those types of attractive forces. For the saturated aliphatic hydrocarbons, for example, these are essentially the only cohesi ve interactions, and the ener gy of v aporization is assumed to be the same as the dispersion cohesive energy, E D. Finding the dispersion cohesi ve energy as the cohesion ener gy of the homomorph, or h ydrocarbon counterpart, is the starting point for calculating the three Hansen parameters for a gi ven liquid. As discussed in more detail later , this is based on a corresponding states calculation. The permanent dipole–permanent dipole interactions cause a second type of cohesion ener gy, the polar cohesive energy, E P. These are inherently molecular interactions and are found in most molecules to one e xtent or another. The dipole moment is the primary parameter used to calculate these interactions. A molecule can be mainly polar in character without being w ater soluble, hence there is a misuse of the term polar in the general literature. The polar solubility parameters referred to here are well-defined, xperimentally verified, and can be estimated from molecular parameter as described later. As noted previously, the most polar of the solv ents include those with relatively high total solubility parameters that are not particularly w ater soluble, such as nitroparaf fins propylene carbonate, and tri-n-b utyl phosphate. Induced dipoles ha ve not been treated specificall in this approach b ut are recognized as a potentially important f actor, particularly for solv ents with zero dipole moments (see the Calculation of the Polar Solubility P arameter section). The third major cohesi ve ener gy source is h ydrogen bonding, E H. This can be called more generally an electron exchange parameter. Hydrogen bonding is a molecular interaction and resembles the polar interactions in this respect. The basis of this type of cohesi ve energy is attraction among molecules because of the h ydrogen bonds. In this perhaps o versimplified approach, th hydrogen bonding parameter has been used to more or less collect the ener gies from interactions not included in the other two parameters. Alcohols, glycols, carboxylic acids, and other hydrophilic materials have high-hydrogen-bonding parameters. Other researchers ha ve divided this parameter into separate parts — for e xample, acid and base cohesion parameters — to allo w both positi ve and negative heats of mixing. These approaches will not be dealt with here b ut are described in Barton’s handbook 9 and else where.19–21 The most e xtensive di vision of the cohesi ve ener gy has been done by Karger et al.,22 who developed a system with fi e parameters: dispersion, orientation, induction, proton donor, and proton acceptor. As a single parameter, the Hansen hydrogen bonding parameter has serv ed remarkably well in the e xperience of the author and k eeps the number of parameters to a le vel that allows ready practical usage. It is clear that there are other sources of cohesion ener gy arising in various types of molecules from, for e xample, induced dipoles, metallic bonds, electrostatic interactions, or whate ver type of separate energy can be defined. The author stopped with the three major types found in or ganic molecules. It has been recognized that additional parameters could be assigned to separate ener gy types. F or e xample, the description of or ganometallic compounds could be an intriguing study . This would presumably parallel similar characterizations of surf ace-active materials, where each part of the molecule requires separate characterization for completeness. The Hansen parameters have mainly been used in connection with solubility relations, mostly , but not e xclusively, in the coatings and related industries. Solubility and swelling have been used to confirm the solubility parameter assignments of ma y of the liquids. Group contrib ution methods and suitable equations based on molecular properties were then derived from these. They make possible estimates of the three parameters for additional liquids. The goal of a prediction is to determine the similarity or dif ference of the cohesion energy parameters. The strength of a particular type of hydrogen bond or any other bond is important only to the e xtent that it influences the cohes ve energy density.

7248_C001.fm Page 6 Monday, April 23, 2007 1:56 PM

6

Hansen Solubility Parameters: A User’s Handbook

HSPs do have direct applications in other scientific disciplines, such as sur ace science, where they have been used to characterize the wettability of various surfaces and the adsorption properties of pigment surfaces,10,14,16,23–26 and have even led to systematic surface treatment of inorganic fiber so that the y could be readily incorporated into polymers of lo w-solubility parameters such as polypropylene27 (see also Chapter 7). Man y widely dif ferent applications ha ve been discussed by Barton9 and Gardon.28 Surface characterizations have not been given the attention deserved in terms of a unified similarity-of-ene gy approach. The author can certify that thinking in terms of similarity of energy, whether surface or cohesive energies, can lead to rapid decisions and plans of action in critical situations that lack data. In other w ords, the e veryday industrial crisis situation often can be reduced in scope by appropriate systematic approaches based on similarity of ener gy. The success of the HSPs for surf ace applications are not surprising in vie w of the similarity of predictions offered by these, and the Prigogine corresponding states theory of polymer solutions discussed in Chapter 2. Flory also emphasized that it is the surf ace of molecules that interact to produce solutions,29 so the interactions of molecules residing in surf aces should clearly be included in an y general approach to interactions among molecules. Surf ace mobility and surf ace rotation are important f actors in en vironmental stress cracking (Chapter 14), certain biological phenomena (Chapter 15), the wetting of surfaces, and in other important phenomena relating to nanotechnology (Chapter 18). The basic equation go verning the assignment of Hansen parameters is that the total cohesion energy, E, must be the sum of the indi vidual energies that mak e it up. E = ED + EP + EH

(1.6)

Dividing this by the molar v olume gi ves the square of the total (or Hildebrand) solubility parameter as the sum of the squares of the Hansen D, P , and H components. E/V = E D/V + E P/V + E H/V

(1.7)

δ2 = δ2D + δ2P + δ2H

(1.8)

To sum up this section, it is emphasized that HSPs quantitati vely account for the cohesion energy (density). Up to this point of time, an e xperimental latent heat of v aporization has been considered a more reliable method to arrive at a cohesion energy rather than using molecular orbital calculations or other calculations based on potential functions. Indeed, the goal of such e xtensive calculations for polar and hydrogen bonding molecules should be to accurately arrive at the energy of vaporization. The statistical thermodynamics approach of Panayiotou and coworkers reported in Chapter 3 may ha ve changed this. An alternative method of calculating the three parameters has been presented, b ut full evaluation of this ne w information has not been possible as yet.

METHODS AND PROBLEMS IN THE DETERMINATION OF PARTIAL SOLUBILITY PARAMETERS The best method to calculate individual HSPs depends to a great extent on what data are available. The author originally adopted an essentially e xperimental procedure and established v alues for 90 liquids based on solubility data for 32 polymers. 13 This procedure in volved calculation of the nonpolar parameter according to the procedure outlined by Blanks and Prausnitz.12 This calculation procedure is still in use and is considered the most reliable and consistent one for this parameter . It is outlined in the follo wing section. The division of the remaining cohesi ve energy between the polar and h ydrogen bonding interactions w as initially done by trial and error to fit xperimental polymer solubility data. A key to parameter assignments in this initial trial-and-error approach was

7248_C001.fm Page 7 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

7

that mixtures of two nonsolvents could be systematically and synergistically (but predictably) found to dissolve given polymers. This meant that these had parameters placing them on opposite sides of the solubility re gion, a spheroid. Using a lar ge number of such predictably syner gistic systems as a basis, reasonably accurate di visions into the three ener gy types were possible. Using the experimentally established, approximate, δP and δH parameters, Hansen and Skaarup15 found that the Böttcher equation (Equation 10.25) could be used to calculate the polar parameter quite well, and this led to a revision of the earlier values to those now accepted for the same liquids. These values were also consistent with the e xperimental solubility data for 32 polymers a vailable at that time and with Equation 1.6. Furthermore, Skaarup de veloped the equation for the solubility parameter “distance,” Ra, between tw o materials based on their respecti ve partial solubility parameter components: (Ra)2 = 4( δD2 – δD1)2 + ( δP2 – δP1)2 + ( δH2 – δH1)2

(1.9)

This equation was developed from plots of experimental data where the constant “4” was found convenient and correctly represented the solubility data as a sphere encompassing the good solvents (see Chapter 5). When the scale for the dispersion parameter is doubled, in comparison with the other two parameters essentially spherical, rather than spheroidal, re gions of solubility are found. This greatly aids two-dimensional plotting and visualization. There are, of course, boundary regions where deviations can occur. These are most frequently found to involve the larger molecular species as being less ef fective solv ents compared to their smaller counterparts that define the solubilit sphere. Likewise, smaller molecular species such as acetone, methanol, nitromethane, and others often appear as outliers, in that they dissolve a polymer even though they have solubility parameters placing them at a distance greater than the xe perimentally-determined radius of the solubility sphere, Ro. This dependence on molar v olume is inherent in the theory de veloped by Hildebrand, Scott, and Scatchard discussed pre viously. Smaller molar v olume f avors lo wer ΔGM, as discussed in Chapter 2. This in turn promotes solubility . Such smaller -molecular-volume species that dissolv e “better” than predicted by comparisons, based on solubility parameters alone, should not necessarily be considered outliers. The molar v olume is frequently and successfully used as a fourth parameter to describe the effects of molecular size. F or e xample, these are especially important in correlating dif fusional phenomena with HSP (see Chapter 12, Chapter 13, and Chapter 16). The author has preferred to retain the three, well-defined partial-solubility parameters with a fourth, separate, molar olume parameter, rather than multiplying the solubility parameters by the molar v olume raised to some power to redefine them The reason for the e xperimentally determined constant 4 in Equation 1.9 will be discussed in more detail in Chapter 2. It will be noted here, however, that the constant 4 is theoretically predicted by the Prigogine corresponding states theory of polymer solutions when the geometric mean is used to estimate the interaction in mixtures of dissimilar molecules. 30 The constant 4 differentiates between atomic and molecular intereactions. This is exceptionally strong evidence that dispersion, permanent dipole–permanent dipole, and h ydrogen bonding interactions all follo w the geometric mean rule. P atterson and co workers ha ve been especially instrumental in relating the Prigogine theory to solubility parameters and to the Flory–Huggins theory of polymer solutions. 3–8 The HSP approach of dividing the cohesive energy into parts derived from different types of cohesive forces has been confirmed both by xperimental studies, as well as the Prigogine theory . The use of the geometric mean is basic to this agreement between the HSP approach and that of Prigogine (see Chapter 2). The approach of optimizing solubility data to spheres is still very much in use. Plotting regions of solubility based on experimental solubility data, or computer-optimizing boundaries of solubility by locating the maximum dif ference in solubility parameters allo wed by Equation 1.9 are both used. The total free energy of mixing, ΔGM, is equal to zero on the boundary. It should be recognized

7248_C001.fm Page 8 Monday, April 23, 2007 1:56 PM

8

Hansen Solubility Parameters: A User’s Handbook

that using the solubility parameters relating to ΔGMnoncomb in Equation 1.4 dif fers from this by the combinatorial entropy of mixing. Another promising approach to arri ve at the HSP for materials based on e xperimental data is to use multivariable analysis of one type or another, as discussed in Chapter 5.This type of approach has not been attempted by the author , but it clearly has adv antages in some cases. The author’s preferred approach of locating the polymer HSP as the center of a sphere has a problem in that it is, in reality, the poor solv ents or nonsolvents located near the boundary of the sphere that fix th boundary (and center) rather than the best solv ents in the middle. This may present problems for smaller sets of data, b ut it is an adv antage when e xtrapolating into re gions of HSP higher than those of an y liquid that can be used in testing. This is discussed in more detail in Chapter 5 and the definition of the limited s gment of the boundary of the HSP sphere deri vable from such correlations is based on Equation 1.9. Equation 1.9 is readily used on a computer (or on a hand calculator), and supplementary relations allow easier scanning of lar ge sets for data. It is ob vious that solubility, or high af finit , requires that Ra be less than Ro. The ratio Ra/Ro has been called the RED number, reflecting th relative energy difference. RED = Ra/Ro

(1.10)

A RED number of 0 is found for no ener gy difference, RED numbers less than 1.0 indicate high affinity; RED equal to or close to 1.0 is a boundary condition; and progress vely higher RED numbers indicate progressively lower affinities. Scanning a computer output for RED numbers les than 1.0, for example, rapidly allows location of the most interesting liquids for a given application. Parenthetically, it should be noted that the ratio of Ra to Ro is really a ratio of quantities having the same units as the solubility parameter . The ratio (Ra/Ro) 2 = (RED) 2 is a ratio of cohesion energies. The latter quantity is important for relating the HSP approach to that of Huggins and Flory, as discussed in Chapter 2. The revised set of parameters for the 90 original solv ents was the basis for group contrib ution procedures developed (most notably) by van Krevelen,31 Beerbower,32 and Hansen and Beerbower,17 who also used Fedors’ w ork.33 These v arious de velopments ha ve been summarized by Barton, 9 although Beerbo wer’s latest v alues ha ve only appeared in the National Aeronautics and Space Administration (NASA) document. 32 Table 1.1 is an e xpanded table of Beerbower group contributions, which w as distributed among those who were in contact with Beerbo wer in the late 1970s. The majority of the data in this table, as well as Table 1.2, ha ve also appeared in Reference 34. Beerbower also de veloped a simple equation for the polar parameter ,17 which in volved only the dipole moment and the square root of the molar v olume. This is also gi ven later (Equation 1.13) and has been found quite reliable by Koenhen and Smolders.35 This equation has been found reliable by the author as well, giving results generally consistent with Equation 1.6 to Equation 1.8, which, again, is the basis of the whole approach. K oenhen and Smolders also give correlation coefficient for other calculation procedures to arri ve at the indi vidual Hansen parameters. The group contrib utions in Table 1.1 ha ve been used e xtensively to arri ve at the collection of HSP data in Appendix Table A.1. Most of the chemicals of primary interest for which full data are available are presumably already in this table. The trend has been to calculate HSP for lar ger and still larger molecules. Many of these have multiple groups, and it becomes more and more difficul to make decisions as to ho w to treat them best. At times the HSP for the lar ger molecules can be estimated from the HSP of lar ger segments that mak e them up. Rather than e xpanding Table 1.1 with additional data, e xcept as noted briefly late , the usual practice has been to locate chemicals with similar groups and to use their HSP v alues in a group contrib ution-type calculation. The procedure has de veloped to the point where its principle features can be identified in th following table. If a boiling point is a vailable, the procedures for calculating δD have been used. If a boiling point is not a vailable, the similarity with related molecules has been used. If a dipole

7248_C001.fm Page 9 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

9

moment is available, the procedures gi ven here were used in preference to group contrib utions. If necessary, group contributions can be derived from similar molecules to the one in question, when dipole moments are a vailable for these, and not for the molecule in question. There is often a change in the group contrib ution as a function of molecular size. This is the main reason for the lack of e xpansion of Table 1.1. It is thought best that the uncertainty be clear to the user . F or example, it has been found that the group contrib ution for the polar component of aliphatic esters should not be less than 300 cal/mol as gi ven in Table 1.1. This is necessary to pre vent δP for materials like plasticizers from being clearly too low, based on their compatibility with, for example, polyvinyl chloride. Sulfur containing compounds have also been somewhat difficult in this respect with major changes in estimated group contributions depending on molecular weight of the chemical in question. Group contrib utions for sulfur , amides, and other groups not found in Table 1.1 can be easily deri ved from the data on similar compounds reported in Appendix Table A.1. The same is true of the δH component. The problem with this procedure is ob vious: any error or distortion of value for a class of compounds is perpetuated. This has been recognized and dealt with to the extent possible, but there are limits to what can be done with limited data.The scope of this situation has been be yond the resources a vailable for its fully satisf actory resolution. The extensive list of group contributions at the end of Chapter 3 pro vides what may be a partial replacement and/or a supplement for Table 1.1. This requires some e xperience with the techniques in volved. A sizable number of materials ha ve been assigned HSPs using the procedures described here. Many of these have not been published. Exxon Chemical Corporation 36,37 has indicated a computer program with data for o ver 500 solv ents and plasticizers, 450 resins and polymers, and 500 pesticides. The author’s files contain the three parameters for about 1200 chemicals (See Appendix Table A.1), although several of them appear with two sets of possible values awaiting experimental confirmation. In some cases, this is due to questionable p ysical data, for example, for latent heats of vaporization, or wide variations in reported dipole moments. Another reason is that some liquids are chameleonic, 38 as defined by H y, in that the y adopt configurations depending on their e vironment. Ho y38 cites the formation of c yclic structures for glycol ethers with (nominally) linear structure. The formation of h ydrogen-bonded tetramers of alcohols in a fluoropolymer has als been pointed out. 39 The term compound formation can be found in the older literature, particularly where mixtures with w ater were in volved and structured species were postulated to e xplain phenomena based on specific interactions among the components of the mixtures. Barton has discusse some of the situations where cohesion parameters need a more careful use, and points out that Hildebrand or Hansen parameters must be used with particular caution where the e xtent of donor–acceptor interactions, especially , h ydrogen bonding within a compound, is v ery dif ferent from that between compounds. 18 Amines, for e xample, are kno wn to associate with each other . Pure component data cannot be e xpected to predict the beha vior in such cases. Still another reason for dif ficulties is the la ge variation of dipole moments reported for the same liquid. The dipole moment for some liquids depends on their environment, as discussed later. A given solvent can be listed with dif ferent values in files to eep these phenomena in mind. Large data sources greatly enhance a search for similar materials and the locating of ne w solvents, as an e xample, for a polymer for which there are limited data. Unfortunately , different authors have used different group contribution techniques, and there is a proliferation of dif ferent “Hansen” parameters for the same chemicals in the literature. This would seem to be an unfortunate situation, b ut may ultimately pro vide benefits. In particula , partial solubility parameter v alues found in Ho y’s e xtensive tables 9,40 are not compatible with the customary Hansen parameters reported here. Hoy has provided an excellent source of total solubility parameters. He independently arrived at the same type of di vision of cohesion ener gies as Hansen, although the methods of calculation were quite dif ferent. Many solvent suppliers ha ve also presented tables of solv ent properties and/or use computer techniques with these tables in their technical service. Partial solubility parameters not taken directly from earlier well-documented sources should be used with caution. The Hoy dispersion parameter,

3.8 10.8 (23.2) 18.0 28.5 10.0

24.0 52.0 81.9 30.0 62.0 97.2 31.5 66.6

16.1 –1.0 –19.2 28.5 13.5 –5.5 — 16 16 18.0

Same Same Same Same Same Same Same 71.4 — Same 22.0 48.0 78.0 28.0 60.0 73.9 34.0 70.0 109.2 35.5 74.6 123.0 Same Same (31.4) Same Same Same

Aliphatic Aromatic 1,125 1,180 820 350 850 ± 100 875 ± 100 800 ± 100 — — — 0 0 0 1,400 ± 100 3,650 ± 160 4,750 ± 300 c 1,950 ± 300 c 4,300 ± 300 c 5,800 ± 400 c 2,350 ± 250 c 5,500 ± 300 c ? 0 —e 950 ± 300 —f 3,350 ± 300 1,770 ± 450

Alkane Same Same Same Same ? ? ? — 250 250 0 0 0 ? ? ? 1,500 ± 175 ? ? 2,200 ± 250 c ? ? 0 2,350 ± 400 ? ? 3,550 ± 250 1,370 ± 500

Cyclo Same Same Same Same ? ? ? 7,530 — 250 0 0 0 1,300 ± 100 3,100 ± 175 c ? 1,650 ± 140 3,500 ± 300 c ? 2,000 ± 250 c 4,200 ± 300 c ? 0 2,800 ± 325 550 ± 275 —f 3,600 ± 400 1,870 ± 600

Aromatic

London Parameter, ΔVδD2 (cal/mol)

0 0 0 0 25 ± 10 18 ± 5 60 ± 10 — 0 0 1,000 ± 150 700 ± 250 c ? 1,250 ± 100 800 ± 150 300 ± 100 1,250 ± 100 800 ± 250 c 350 ± 150 c 1,250 ± 100 800 ± 250 c ? 500 ± 150 (15, 000 ± 7%)/V 2,100 ± 200 (56,000 ± 12%)/V 500 ± 150 700 ± 200

Alkane 0 0 0 0 ? ? ? — 0 0 ? ? ? 1,450 ± 100 ? ? 1,700 ± 150 ? ? 1,350 ± 100 ? ? 600 ± 150 1,000 ± 300 3,000 ± 500 ? 300 ± 50 1,100 ± 300

Cyclo

Aromatic 0 0 0 0 ? ? ? 50 ± 25 — 0 700 ± 100 500 ± 250 c ? 800 ± 100 400 ± 150 c ? 800 ± 100 400 ± 150 c ? 575 ± 100 400 ± 150 c ? 450 ± 150 950 ± 300 2,750 ± 200 (338,000 ± 10%)/V 750 ± 350 800 ± 150

Polar Parameter, ΔVδP2 (cal/mol

0 0 0 0 180 ± 75 180 ± 75 180 ± 75 — 0 0 0 0 0 100 ± 20 c 165 ± 10 c 350 ± 250 c 500 ± 100 825 ± 200 c 1,500 ± 300 c 1,000 ± 200 c 1,650 ± 250 c ? 450 ± 25 800 ± 250 d 1,000 ± 200 1,250 ± 150 2,750 ± 250 4,650 ± 400

Aliphatic 0 0 0 0 ? ? ? 50 ± 50 c — 0 0 0 0 Same 180 ± 10 c ? 500 ± 100 800 ± 250 c ? 1,000 ± 200 c 1,800 ± 250 c ? 1,200 ± 100 400 ± 125 c 750 ± 150 475 ± 100 c 2,250 ± 250 c 4,650 ± 500

Aromatic

Electronic Transfer Parameter, ΔVδH2 (cal/mol)

1,125 1,180 820 350 1,030 1,030 1,030 — 250 250 1,000 1,700 1,650 2,760 4,600 5,400 3,700 5,900 7,650 4,550 8,000 11,700 800 4,150 (4,050) 4,300 6,600 7,120

Aliphatic

Same Same Same Same Same Same Same 7630 — 250 800 b 1,360b 1,315b 2,200b 3,670b 4,300b 2,960b 4,700b 6,100b 3,600b 6,400b 9,350b (1,650 ± 150) Same Same Same Same Same

Aromatic

Total Parametera ΔVδ2 (cal/mol)

10

CH3 33.5 CH2< –CH< >C< CH2 = olefi –CH = olefi >C = olefi PhenylC-5 ring (saturated) C-6 ring –F F2 twin f 40.0 F3 triplet f 66.0 –Cl Cl2 twin f Cl3 triplet f –Br Br2 twin f Br3 triplet f –I I2 twin e I3 triplet e 111.0 –O– ether >CO ketone –CHO –COO-ester –COOH –OH

Functional Group

Molar Volume,a ΔV (cm3/mol)

TABLE 1.1 Group Contributions to Partial Solubility Parameters

7248_C001.fm Page 10 Monday, April 23, 2007 1:56 PM

Hansen Solubility Parameters: A User’s Handbook

Same Same 32.0 Same Same Same Same

26.0

24.0 24.0 19.2 4.5 (6.7) 28.0

1,600 ± 850 c 3,000 ± 600 1,050 ± 300 1,150 ± 225 ? —e

0 ? ? 1,050 ± 450 c ? ? ?

? 0 2,550 ± 125 150 ± 150 c ? ? ?

? 4,000 ± 800 c 3,600 ± 600 600 ± 200 100 ± 50 ? (81,000 ± 10%)/V

1,500 ± 100 ? ? 600 ± 350 c ? ? ?

? 3,750 ± 300c 1,750 ± 100 800 ± 200 ? ? ?

? 500 ± 200 d 400 ± 50 d 1,350 ± 200 750 ± 200 2,700 ± 550 c 3,000 ± 500

9,000 ± 600 400 ± 125 c 350 ± 50 c 2,250 ± 200 d ? ? ?

9,300 ± 600 4,150 7,000 3,000 2,000 (5,850) (7,000)

10,440 Same (4,400) Same Same Same Same

Same

Source: From Hansen, C. M., Paint Testing Manual, Manual 17, K oleske, J. V., Ed., American Society for Testing and Materials, Philadelphia, 1995, 388. Cop yright ASTM. Reprinted with permission.

b

a

Data from Fedors, R.F ., A method for estimating both the solubility parameters and molar v olumes of liquids, Polym. Eng. Sci., 14(2), 147–154, 472, 1974. With permission. These values apply to halogens attached directly to the ring and also to halogens attached to aliphatic double-bonded C atoms. c Based on v ery limited data. Limits sho wn are roughly 95% confidence; in ma y cases, values are for information only and not to be used for computation. d Includes unpublished infrared data. e Use formula in ΔVδ 2 column to calculate, with V for total compound. P f Twin and triplet v alues apply to halogens on the same C atom, e xcept that ΔVδ 2 also includes those on adjacent C atoms. P

(OH)2 twin or adjacent –CN –NO2 –NH2 amine >NH2 amine –NH2 amide PO4 ester

7248_C001.fm Page 11 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction 11

7248_C001.fm Page 12 Monday, April 23, 2007 1:56 PM

12

Hansen Solubility Parameters: A User’s Handbook

TABLE 1.2 Lydersen Group Constants Group

Aliphatic, ΔT

Cyclic, ΔT

ΔPT

Aliphatic, ΔP

Cyclic, ΔP

CH3 CH2 >CH– >C< CH2 CH– C< CH aromatic CH aromatic

0.020 0.020 0.012 0.000 0.018 0.018 0.000 — —

— 0.013 0.012 –0.007 — 0.011 0.011 — —

0.0226 0.0200 0.0131 0.0040 0.0192 0.0184 0.0129 0.0178 0.0149

0.227 0.227 0.210 0.210 0.198 0.198 0.198 — —

— 0.184 0.192 0.154 — 0.154 0.154 — —

–O– >O epoxide –COO– >CO –CHO –CO2O

0.021 — 0.047 0.040 0.048 —

0.014 — — 0.033 — —

0.0175 0.0267 0.0497 0.0400 0.0445 0.0863

0.16 — 0.47 0.29 0.33 —

0.12 — — 0.02 — —

–OH→ –H→ –OH primary –OH secondary –OH tertiary –OH phenolic

— — 0.082 — — 0.035

— — — — — —

0.0343 –0.0077 0.0493 0.0440 0.0593 0.0060

0.06 — — — — –0.02

— — — — —

–NH2 –NH– >N– –CN

0.031 0.031 0.014 0.060

— 0.024 0.007 —

0.0345 0.0274 0.0093 0.0539

0.095 0.135 0.17 0.36

— 0.09 0.13 —

–NCO HCON< –CONH– –CON< –CONH2 –OCONH–

— — — — — —

— — — — — —

0.0539 0.0546 0.0843 0.0729 0.0897 0.0938

— — — — — —

— — — — — —

–S– –SH

0.015 0.015

0.008 —

0.0318 —

0.27 —

0.24 —

–Cl 1° –Cl 2° Cl1 twin Cl aromatic

0.017 — — —

— — — —

0.0311 0.0317 0.0521 0.0245

0.320 — — —

— — — —

–Br –Br aromatic

0.010 —

— —

0.0392 0.0313

0.50 —

— —

–F –I

0.018 0.012

— —

0.006 —

0.224 0.83

— —

7248_C001.fm Page 13 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

13

TABLE 1.2 (CONTINUED) Lydersen Group Constants Group

Aliphatic, ΔT

Cyclic, ΔT

ΔPT

Aliphatic, ΔP

Cyclic, ΔP

Conjugation cis double bond trans double bond

— — —

— — —

0.0035 –0.0010 –0.0020

— — —

— — —

4 Member ring 5 Member ring 6 Member ring 7 Member ring

— — — —

— — — —

0.0118 0.003 –0.0035 0.0069

— — — —

— — — —

Ortho Meta Para

— — —

— — —

0.0015 0.0010 0.0060

— — —

— — —

Bicycloheptyl Tricyclodecane

— —

— —

0.0034 0.0095

— —

— —

Source: Hansen, C.M., Solubility parameters, in Paint Testing Manual, Manual 17, Koleske, J.V., Ed., American Society for Testing and Materials, Philadelphia, P A, 1995, 383–404. Reprinted with permission.

in particular, is consistently lo wer than that found by Hansen. Ho y subtracts estimated v alues of the polar and h ydrogen-bonding energies from the total ener gy to find the dispersion ene gy. This allows for more calculational error and underestimates the dispersion energy, as the Hoy procedure does not appear to fully separate the polar and h ydrogen-bonding ener gies. The v an Kre velen dispersion parameters appear to be too lo w. The author has not attempted these calculations, being completely dedicated to the full procedure based on corresponding states described here, but values estimated independently using the v an Krevelen dispersion parameters are clearly lo w. A comparison with related compounds or the similarity principle gi ves better results than those found from the van Krevelen dispersion group contrib utions. In the follo wing, calculation procedures and e xperience are presented according to the procedures most reliable for the e xperimental and/or ph ysical data a vailable for a gi ven liquid.

CALCULATION OF THE DISPERSION SOLUBILITY PARAMETER δD The δD parameter is calculated according to the procedures outlined by Blanks and Prausnitz. 12 Figure 1.1 to Figure 1.3 can be used to find this paramete , depending on whether the molecule of interest is aliphatic, cycloaliphatic, or aromatic. These figures h ve been inspired by Barton,9 who converted earlier data to Standard International (SI) units. All three of these figures h ve been straight-line e xtrapolated into a higher range of molar v olumes than that reported by Barton. Energies found with these extrapolations have also provided consistent results. As noted earlier, the solubility parameters in SI units (MP a1/2) are 2.0455 times lar ger than (ca1/cc) 1/2 in the older cgs centimeter gram second (cgs) system, which still finds xtensive use in the U.S., for example. The figure for the aliphatic liquids g ves the dispersion cohesive energy, ED, whereas the other two figures directly report the dispersion cohes ve energy density, c. The latter is much simpler to use, as one need only tak e the square root of the v alue found from the figure to find the respec ve partial solubility parameter. Barton also presented a similar figure for the aliphatic sol ents, but it

7248_C001.fm Page 14 Monday, April 23, 2007 1:56 PM

14

Hansen Solubility Parameters: A User’s Handbook

Δ ED kJ/mol 70

60 = Tr

50

0 0.4 5 0.4 0 0.5 5 0.5 0.60 0.65

40

0.70

30

20

10

0 50

100

150

200

250

V, cm3/mol

FIGURE 1.1 Energy of v aporization for straight chain h ydrocarbons as a function of molar v olume and reduced temperature. (From Hansen, C. M., Paint Testing Manual, Manual 17, K oleske, J. V., Ed., American Society for Testing and Materials, Philadelphia, 1995, 389. Cop yright ASTM. Reprinted with permission.)

400

350 c, MPa

Tr = 0.40 0.45

300

0.50 0.55 0.60 0.65 0.70

250

50

60

70

80

90

100

110

120

130

3

V, cm /mol FIGURE 1.2 Cohesive energy density for c ycloalkanes as a function of molar v olume and reduced temper ature. (From Hansen, C. M., Paint Testing Manual, Manual 17, K oleske, J. V., Ed., American Society for Testing and Materials, Philadelphia, 1995, 389. Cop yright ASTM. Reprinted with permission.)

7248_C001.fm Page 15 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

15

400

c, MPa

Tr = 0.40 350 0.45 0.50

300

0.55 0.60 0.65 0.70

250

80

90

100

110

120 130 V, cm3/mol

140

150

160

170

FIGURE 1.3 Cohesive energy density for aromatic hydrocarbons as a function of molar v olume and reduced temperature. (From Hansen, C. M., Paint Testing Manual, Manual 17, K oleske, J. V., Ed., American Society for Testing and Materials, Philadelphia, 1995, 389. Cop yright ASTM. Reprinted with permission.)

is inconsistent with the ener gy figure and in erro . Its use is not recommended. When substituted cycloaliphatics or substituted aromatics are considered, simultaneous consideration of the tw o separate parts of the molecule is required. The dispersion ener gies are e valuated for each of the molecules in volved, and a weighted a verage is tak en for the molecule of interest based on the number of significant atoms. or example, hexyl benzene w ould be the arithmetic a verage of the dispersion energies for an aliphatic and an aromatic liquid, each with the gi ven molar v olume of hexyl benzene. Liquids such as chlorobenzene, toluene, and ring compounds with alk yl substitutions that have only tw o or three carbon atoms ha ve been considered only as c yclic compounds. Such weighting has been found necessary to satisfy Equation 1.6. The critical temperature, Tc, is required to use the dispersion ener gy figures. If the critica temperature cannot be found, it must be estimated. A table of the L ydersen group contributions,41 ΔT , as given by Hoy40 for calculation of the critical temperature is included as Table 1.2. In some cases, the desired groups may not be in the table, which requires some educated guessing. The end result does not appear too sensiti ve to these situations. The normal boiling temperature, Tb, is also required in this calculation. This is not always available and must be estimated by similarity, group contribution, or some other technique. The Lydersen group contrib ution method in volves the use of Equation 1.11 and Equation 1.12 as follo ws: Tb/Tc = 0.567 + ΣΔT – ( ΣΔT)2

(1.11)

Tr = T/Tc

(1.12)

and

where T has been tak en as 298.15 K. The dispersion parameter is based on atomic forces. The size of the atom is important. It has been found that corrections are required for atoms significantly lager than carbon, such as chlorine, sulfur, bromine, etc., b ut not for oxygen or nitrogen that ha ve a similar size. The carbon atom in hydrocarbons is the basis of the dispersion parameter in its present form. These corrections are applied by first finding the dispersion cohe ve energy from the appropriate figure. This requires multiplication by the molar v olume for the c yclic compounds using data from the figures here, a these figures g ve the cohesive energy densities. The dispersion cohesive energy is then increased by adding on the correction factor. This correction factor for chlorine, bromine, and sulfur has been

7248_C001.fm Page 16 Monday, April 23, 2007 1:56 PM

16

Hansen Solubility Parameters: A User’s Handbook

taken as 1650 J/mol for each of these atoms in the molecule. Di viding by the molar v olume and then taking the square root gi ves the (lar ge atom corrected) dispersion solubility parameter . The need for these corrections has been confirmed ma y times, both for interpretation of experimental data and to allo w Equation 1.6 to Equation 1.8 to balance. Research is definitel needed in this area. The impact of these corrections is, of course, lar ger for the smaller molecular species. Taking square roots of the lar ger numbers in volved with the lar ger molecular species reduces the errors in volved in these cases, as the corrections are relati vely small. It can be seen from the dispersion parameters of the c yclic compounds that the ring has an effect similar to increasing the ef fective size of the interacting species. The dispersion energies for cycloaliphatic compounds are lar ger than their aliphatic counterparts, and the y are higher for aromatic compounds than their corresponding c ycloaliphatics. Similar effects also appear with the ester group. This group appears to act as if it were, in ef fect, an entity that is lar ger than the corresponding compound containing only carbon (i.e., its homomorph), and it has a higher disper sion solubility parameter without an y special need for corrections. The careful evaluation of the dispersion cohesi ve energy may not ha ve a major impact on the value of the dispersion solubility parameter, because square roots of rather large numbers are taken. Larger problems arise because of Equation 1.6. Ener gy assigned to the dispersion portion cannot be reused when finding the other partial parameters using Equation 1.6 (or Equation 1.8). This is one reason group contrib utions are recommended in some cases, as discussed later .

CALCULATION OF THE POLAR SOLUBILITY PARAMETER δP The earliest assignments of a “polar” solubility parameter were gi ven by Blanks and Prausnitz. 12 These parameters were, in f act, the combined polar and h ydrogen bonding parameters as used by Hansen, and the y cannot be considered polar in the current conte xt. The first Hansen pola parameters13 were reassigned new values by Hansen and Skaarup according to the Böttcher equation (Equation 10.25).15 This equation requires the molar volume, the dipole moment (DM), the refractive index, and the dielectric constant. These are not available for many compounds, and the calculation used is more dif ficult than the much simpler equation d veloped by Hansen and Beerbo wer17: δP = 37.4(DM)/V1/2

(1.13)

The constant 37.4 gi ves this parameter in SI units. Equation 1.13 has been consistently used by the author over the past years, particularly in view of its reported reliability.35 This reported reliability appears to be correct. The molar volume must be known or estimated in one w ay or another. This leaves only the dipole moment to be found or estimated. Standard reference works have tables of dipole moments, with the most extensive listing still being McClellan.42 Other data sources also have the same, as well as other relevant parameters, and data such as latent heats and critical temperatures. The Design Institute for Ph ysical Property Research (DIPPR) 43 database has been found useful for man y compounds of reasonably common usage, but many interesting compounds are not included in the DIPPR. When no dipole moment is a vailable, similarity with other compounds, group contrib utions, or e xperimental data can be used to estimate the polar solubility parameter . It must be noted that the f act of zero dipole moment in symmetrical molecules is not basis enough to assign a zero polar solubility parameter . An outstanding e xample of v ariations of this kind can be found with carbon disulfide. The reported dipole moments are mostly 0 for g as phase measurements, supplemented by 0.08 in hexane, 0.4 in carbon tetrachloride, 0.49 in chlorobenzene, and 1.21 in nitrobenzene. There is a clear increase with increasing solubility parameter of themedia. The latter and the highest v alue has been found e xperimentally most fitting for correlating perme ation through a fluoropolymer film used for chemical protec ve clothing. 44 Many fluoropolymer

7248_C001.fm Page 17 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

17

have considerable polarity. The lower dipole moments seem to fit in other instances. Diet yl ether has also presented problems as an outlier in terms of dissolving or not and permeating rapidly or not. Here, the reported dipole moments 42 vary from 0.74 to 2.0, with a preferred v alue of 1.17 and 1.79 in chloroform. Choosing a gi ven value seems rather arbitrary . The chameleonic c yclic forms of the linear glycol ethers would also seem to provide a basis for altered dipole moments in various media.38 When Equation 1.13 cannot be used, the polar solubility parameter has been found using the Beerbower table of group contrib utions, by similarity to related compounds and/or by subtraction of the dispersion and h ydrogen bonding cohesi ve ener gies from the total cohesi ve ener gy. The question in each case is, “Which data are a vailable and judged most reliable?” Ne w group contributions can also be developed from related compounds whose dipole moments are available. These new polar group contrib utions then become supplementary to the Beerbo wer table. For large molecules, especially those with long h ydrocarbon chains, the accurate calculation of the relatively small polar (and h ydrogen bonding) contributions present special difficulties. The latent heats are not generally a vailable with suf ficient accura y to allo w subtraction of tw o large numbers from each other to find a ery small one. In such cases, the similarity and group contrib ution methods are thought best. Unfortunately , latent heats found in a widely used handbook 45 are not clearly reported as to the reference temperature. There is an indication that these are 25°C data, but checking indicated man y of the data to be identical with boiling point data reported else where in the literature. Subsequent editions of this handbook 46 have a completely different section for the latent heat of e vaporation. Again, even moderate v ariations in reported heats of v aporization can cause severe problems in calculating the polar (or h ydrogen bonding) parameter when Equation 1.6 or Equation 1.8 are strictly adhered to.

CALCULATION OF THE HYDROGEN BONDING SOLUBILITY PARAMETER δH In the earliest w ork, the hydrogen bonding parameter w as almost always found by subtracting the polar and dispersion ener gies of v aporization from the total ener gy of v aporization. This is still widely used where the required data are a vailable and reliable. At this stage, ho wever, the group contribution techniques are considered reasonably reliable for most of the required calculations and, in f act, more reliable than estimating se veral other parameters to ultimately arri ve at the subtraction step just mentioned. Therefore, in the absence of reliable latent heat and dipole moment data, group contributions are judged to be the best alternative. Similarity to related compounds can also be used, of course, and the result of such a procedure should be essentially the same as for using group contrib utions. The above paragraph is not changed from the first edition of this handbook.This is to emphasize the importance of the w ork of P anayiotou and co workers reported in Chapter 3. It no w appears possible not only to calculate the h ydrogen bonding parameter independently, but also to arri ve at all three parameters by statistical mechanics. This is clearly a major step forw ard. Whether or not one understands all of the equations and methodology of Chapter 3, the procedure in itself confirm the need for (at least) three cohesion ener gy parameters, and similar results are found by the approach of the first paragraph as well as with statistical thermodynamics

SUPPLEMENTARY CALCULATIONS AND PROCEDURES The procedures listed pre viously are those most frequently used by the author in calculating the three partial solubility parameters for liquids when some data are a vailable. There are a number of other calculations and procedures that are also helpful. Latent heat data at 25°C ha ve been found consistently from those at another temperature, using the relation gi ven by Fishtine. 47

7248_C001.fm Page 18 Monday, April 23, 2007 1:56 PM

18

Hansen Solubility Parameters: A User’s Handbook

ΔHv(T1)/ΔH(T2) = [(1 – Tr1)/(1 – Tr2)]0.38 (1.14) This is done e ven if the melting point of the compound being considered is higher than 25°C. The result is consistent with all the other parameters, and to date no problems with particularly faulty predictions ha ve been noted. It appears as if the predictions are not significantly in erro when e xperimental data are a vailable for checking. When the latent heat at the boiling point is given in cal/mol, Equation 1.14 is used to estimate the latent heat at 25°C. RT equal to 592 cal/mol is then subtracted from this according to Equation 1.15, to find the total cohesion ene gy, E, in cgs units at this temperature: E = ΔEv = ΔHV – RT

(1.15)

where R is the g as constant and T is the absolute temperature. A computer program has been de veloped by the author to assign HSP to solv ents, based on experimental data alone. This has been used in se veral cases where the parameters for the gi ven liquids were desired with a high degree of accuracy. The procedure is to enter solvent quality, good or bad, into the program for a reasonably large number of polymers where the solubility parameters, and appropriate radius of interaction for the polymers are kno wn. The program then locates that set of δD, δP , and δH parameters for the solv ent that best satisfies the requirements of a locatio within the spheres of the appropriate polymers, that ha ve good solv ent quality, and outside the appropriate spheres where the solv ent quality is bad. An additional aid in estimating HSP for many compounds is that these parameters can be found by interpolation or e xtrapolation, especially for homologous series. The first member may no necessarily be a straight-line extrapolation, but comparisons with related compounds should always be made where possible to confirm assignments. Plotting the parameters for homologous serie among the esters, nitroparaf fins, etones, alcohols, and glycol ethers has aided in finding th parameters for related compounds.

TEMPERATURE DEPENDENCE Only very limited attempts have been made to calculate solubility parameters at a higher temperature prior to the second edition of this handbook. The inclusion of Chapter 3 and Chapter 10 in this handbook helps by pro viding a more accurate treatment of temperature dependence when the situation warrants it. Solubility parameter correlations of phenomena at higher temperatures ha ve generally been found satisfactory when the established 25°C parameters ha ve been used. Recalculation to higher temperatures is possible b ut has not generally been found necessary . In this direct but approximate approach, it is assumed that the parameters all demonstrate the same temperature dependence, which, of course, is not the case. It might be noted in this connection that the ydrogenh bonding parameter, in particular, is the most sensitive to temperature. As the temperature increases, more and more h ydrogen bonds are progressi vely brok en or weak ened, and this parameter will decrease more rapidly than the others. The g as-phase dipole moment is not temperature dependent, although the v olume of a flui does change with the temperature, which will also change its cohesi ve energy density. The change of the δD, δP , and δH parameters for liquids with temperature, T, can be estimated by the following equations, where α is the coef ficient of thermal xpansion17: dδD/dT = –1.25 αδD

(1.16)

dδP/dT = –0.5 αδP

(1.17)

dδH/dT = – δH(1.22 × 10 –3 + 0.5 α)

(1.18)

7248_C001.fm Page 19 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

19

Higher temperature means a general increase in rate of solubility/dif fusion/permeation, as well as larger solubility parameter spheres. δD, δP , and δH decrease with increased temperature, as can be seen by a comparison of Equation 1.16, through Equation 1.18. This means that alcohols, phenols, glycols, and glycol ethers become better solv ents for polymers of lower solubility parameters as the temperature increases. Thus, increasing the temperature can cause a nonsolv ent to become a good solv ent, a f act that is often noted in practice. As mentioned earlier , it is possible that a boundary solv ent can be a good solv ent at a gi ven temperature, b ut turn bad with either an increase or a decrease in temperature. These phenomena are discussed in great detail by P atterson and coworkers.3,4 They can be explained either by the change in solubility parameter with temper ature or more completely by the Prigogine CST . The effects of temperature changes on solubility relations are most ob vious with systems ha ving a high h ydrogen-bonding character. Examples are given in the ne xt section for some special situations in volving water and methanol.

SOME SPECIAL EFFECTS TEMPERATURE CHANGES Water (and methanol) uptak e in most polymers increases with increasing temperature. This is because the solubility parameters of the w ater and the polymer are closer at higher temperatures. The δH parameter of w ater (and methanol) f alls with increasing temperature, whereas that of most polymers remains reasonably constant.Water is also well known as an exceptionally good plasticizer because of its small molecular size. The presence of dissolv ed water not only softens (reduces the glass transition temperature) a polymer as such, b ut it also means dif fusion rates of other species will be increased. The presence of water in a film can also influence the upt e of other materials, such as in solubility parameter studies or resistance testing, with h ydrophilic materials being more prone to enter the film This can cause blistering on rapid cooling as discussed in Chapter 12 and in Reference 48 (see Chapter 8 and Chapter 12). Figure 8.3 sho ws how rapid cooling from a w ater-saturated state at higher temperature can lead to blistering. Figure 12.3 and Figure 12.4 sho w how this effect can be measured e xperimentally with an increase in w ater content abo ve the equilibrium v alue when temperature cycling is encountered. This leads to premature f ailure of polymeric products used in such environments. A related problem has been encountered with methanol. It w as intended to follo w the rate of uptake of methanol in an epoxy coating at room temperature by weighing coated-metal panels periodically on an analytical balance. Blistering was encountered in the coating near the air surface shortly after the experiment began. The methanol that had been absorbed into the coating near the surface became insoluble as the temperature of the coating near the surf ace was lowered by the evaporation of e xcess methanol during the handling and weighing of the panels. This is a rather extreme case, and, as mentioned earlier , use of the HSP (determined at 25°C) at ele vated temperatures can most often be done without too much trouble from a practical point of vie w. One should be aware that the changes in the δH parameter would be larger than those in the other parameters, and this effect would be most significant for those liquids with la ger δH values.

EFFECTS

OF

SOLVENT MOLECULAR SIZE

The size of both solv ent and solute molecules is important for solubility , permeation, dif fusion, and chemical resistance phenomena. Smaller molecules tend to be more readily soluble than lar ger ones. As stated previously, the Hildebrand solubility parameter theory also points to smaller molar volume solvents as being better than those with lar ger molar volumes, even though they may have identical solubility parameters. 1,2 This fact of e xpected improved solvency for smaller molecules is also known from the Flory–Huggins theory of polymer solutions.29 Solvents with smaller molecular size ha ve also been repeatedly noted as being superior to those with lar ger molecular size, when highly crystalline polymers or solids are being tested for solubility . So it is not surprising

7248_C001.fm Page 20 Monday, April 23, 2007 1:56 PM

20

Hansen Solubility Parameters: A User’s Handbook

that solvent molecular size can be an important fourth parameter in solubility and, in some cases, in chemical resistance. Specific xamples are given in Chapter 5 and Chapter 12. The size and shape of the solvent molecule are also very important for kinetic phenomena such as diffusion, permeation, and attainment of equilibrium. Smaller and more linear molecules diffuse more rapidly than lar ger and more b ulky ones. The dif fusion coef ficient may be so l w that equilibrium is not attained for hundreds of years at room temperature. This was demonstrated in common solvent exposures of rigid polymers like polyphenylene sulfide (PPS) with thicknesses o several millimeters. 49 Likewise, the second stage in the tw o-stage drying process in polymer fil formation by solvent evaporation can last for man y years. 16,50 Polymer samples used for solubility parameter or other testing may well retain solvent or monomer for many years, and this may affect the evaluations. Attempts to include the molecular v olume in ne w composite solubility and size parameters have not been particularly successful. 20,21 This may be because the size ef fect is most often not caused due to the thermodynamic considerations on which the solubility parameters are based, b ut rather through a kinetic effect of diffusion rates or other free volume considerations. The similarities in the HSP approach and the Prigogine theory, discussed in Chapter 2, indicate a remarkably close, if not identical, relation between the Prigogine ρ (segment size parameter) and the δD parameter, suggesting that molecular size dif ferences are at least partially accounted for in the δD parameter. The Prigogine theory also has a parameter to describe “structural effects,” including size of polymer molecules, but this has not been e xplored in relation to the present discussion. The increase of δD with increasing molecular size among the aliphatic hydrocarbons, the higher δD values for the larger units represented by cycloaliphatic and aromatic rings, and the need for corrections for larger atoms discussed earlier tend to support the molecular size dif ferences. Sorting output data according to the molecular volume of the test solvents in a computer analysis helps to discover whether solvent molecular size is indeed an additional significant actor in a given correlation or testing program.

COMPUTER PROGRAMS The author has used two computer programs extensively in his own studies and in collecting material for this book. These are called SPHERE and SPHERE1. They are very similar, the only difference being that SPHERE optimizes the polymer (or other material, of course) parameters based on all the data, whereas SPHERE1 considers data for only those solvents considered as “good.” It neglects the nonsolv ent data. SPHERE1 has been most useful in correlations with pigments, fillers, an fibers, as described in Chapter 7 The data input is by solv ent number follo wed by an indication of the quality of interaction with that solv ent. A “1” indicates a “good” solv ent, whereas a “0” is used for a “bad” solv ent. What is considered good or bad v aries according to the level of interaction being studied. This can be solution or not, a gi ven percentage of swelling or uptak e, breakthrough time being less than a given interval, permeation coefficients higher than a gven value, long-time suspension of a pigment, etc. The program systematically e valuates the input data using a quality-of-fit function called th desirability function.51 This suggestion w as made by a reputed statistician man y years ago as the most appropriate statistical treatment for this type of problem. It has been in use since the late 1960s. The function has the form: DATA FIT = (A 1 * A2 *...A n)1/n

(1.19)

where n is the number of solv ents for which there is e xperimental data in the correlation. The DATA FIT approaches 1.0 as the fit impr ves during an optimization and reaches 1.0 when all the good solvents are included within the sphere and all the bad ones are outside the sphere.

7248_C001.fm Page 21 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

21

Ai = e –(ERROR DISTANCE)

(1.20)

The Ai quotient for a given good solvent within the sphere and a bad solvent outside the sphere will be 1.0. The error distance is the distance of the solv ent in error to the sphere boundary . It could either denote being good and outside the sphere or being bad and inside the sphere. Ro is the radius of the sphere, and Ra is the distance from a gi ven solvent point to the center of the sphere. F or a good solv ent outside the sphere, an error enters the D ATA FIT according to: Ai = e +(Ro – Ra)

(1.21)

Such errors are often found for solv ents having low molecular v olumes. For a bad solv ent inside the sphere, the contrib ution to the D ATA FIT is Ai = e +(Ra – Ro)

(1.22)

Such errors can sometimes be found for lar ger molecular species such as plasticizers. This is not unexpected for the reasons mentioned earlier . Solvents with large and/or small molecules that give the “errors” can sometimes be (temporarily) disregarded by generating a ne w correlation; this gi ves an excellent DATA FIT for an abbre viated range of molecular v olumes. There is a special printout with the solv ents arranged in order of molecular volume that helps to analyze such situations.The computer printouts all include a column for the RED number . The program assumes a starting point, based on an a verage of each of the HSP for the good solvents only. The program then e valuates eight points at the corners of a cube, with the current best values as center. Different radii are evaluated at each of these points in the optimization process. When better fits are found among the eight points, the point with the best fit is t en as a ne w center, and eight points around it are e valuated in a similar manner. This continues until the DATA FIT cannot be impro ved upon. The length of the edge of the cube is then reduced in size to fine tune the fit. The initial length of the cube is 1 unit, which is reduced to 0.3 unit, and finally to 0. unit in the last optimization step. Experimental data for the solv ents are entered with solv ent number (comma) and a “1” for a good solvent, or a “0” for a bad one. Errors in the correlations are indicated with an “*” in the SOLUB column where the e xperimental input data are indicated. As stated abo ve, systematic errors can sometimes be seen in the molar volume printout. This may suggest a new analysis of the data. Nonsystematic errors may be real, such as for reactions or some e xtraneous effect not predictable by the solubility parameter . They may also be bad data, and rechecking data indicated with an “*” in the output has become a routine practice. The output of this program is for the least radius allo wing the maximum DATA FIT. An example is found in Table 5.4. Results from the SPHERE program reported in this book generally include the HSP , given as D (δD), P (δP), and H (δH), respectively, and Ro for the correlation in question, as well as the DATA FIT, the number of good solv ents (G), and the total solv ents (T) in the correlation. This latter information has not always been recorded and may be lacking for some correlations, especially the older ones.

HANSEN SOLUBILITY PARAMETERS FOR WATER Water is such an important material that a special section is dedicated to its HSP at this point. The behavior of w ater often depends on its local en vironment, which mak es general predictions v ery difficult. Water is still so unpredictable that its use as a test solv ent in solubility parameter studies

7248_C001.fm Page 22 Monday, April 23, 2007 1:56 PM

22

Hansen Solubility Parameters: A User’s Handbook

TABLE 1.3 HSP Correlations Related to Water Correlation

δD

δP

δH

Ro

FIT

G/T

Water — Single molecule Water — >1% soluble in a Water — Total miscibility 1 a Water — Total miscibility “1 b”

15.5 15.1 18.1 18.1

16.0 20.4 17.1 12.9

42.3 16.5 16.9 15.5

— 18.1 13.0 13.9

— 0.856 0.880 1.000

— 88/167 47/166 47/47

a b

Based on SPHERE program. Based on SPHERE1 Program.

is not recommended. This is true of w ater as a pure liquid or in mixtures. Table 1.3 includes data from various HSP analyses of the behavior of water. The first set of data is der ved from the energy of vaporization of water at 25°C. The second set of data is based on a correlation of the solubility of various solvents in w ater, where “good” solv ents are soluble to more than 1% in w ater. “Bad” solvents dissolve to a lesser e xtent. The third set of data is for a correlation of total miscibility of the given solvents with water. The second and third entries in Table 1.3 are based on the SPHERE program where both good and bad solv ents af fect the D ATA FIT and hence the result of the optimization. The last entry in Table 1.3 is for an analysis using the SPHERE1 program. The HSP data are for the minimum sphere that encompasses only the good solv ents. The bad solv ents are simply not considered in the data processing. This type of comparison usually results in some of the parameters being lo wer than when all the data are included. A frequent problem is that a considerable portion of the HSP spheres, such as in the case for w ater, covers such high ener gies that no liquid can be found. The cohesion energy is so high as to require solids. The constant 4 in the correlations (Equation 1.9) is still used for these correlations, primarily based on successes at lower levels of cohesion ener gies, but this is also supported by the comparison with the Prigogine cst of polymer solutions, discussed at some length in Chapter 2. The HSP for w ater as a single molecule, based on the latent heat at 25°C is sometimes used in connection with mixtures with water to estimate average HSP. More recently, it has been found in a study involving water, ethanol, and 1,2-propanediol that the HSP for water indicated by the total water solubility correlation could be used to explain the behavior of the mixtures involved. The averaged values are very questionable as water can associate, and w ater has a v ery small molar v olume as a single molecule. It almost appears to ha ve a dual character . The data for the 1% correlation, 52 as well as for the total w ater miscibility, suggest that about six w ater molecules associate into units. Traditionally, solvents are considered as points. This is practical and almost necessary from an experimental point of vie w as most solv ents are so miscible as to not allo w an y e xperimental characterization in terms of a solubility sphere. An exception to this is the data for w ater reported in Table 1.3. The HSP reported here are the center points of HSP spheres where the good solv ents are either those that are completely miscible or those that are miscible to only 1% or more, as discussed previously. It should also be mentioned that amines were a major source of outliers in these correlations. No solids were included. Their use to predict solubility relations for amines and for solids must therefore be done with caution.

CONCLUSION This chapter has been dedicated to describing the tools with which dif ferent HSP characterizations can be made and some of the pitf alls that may be encountered in the process. The justification fo

7248_C001.fm Page 23 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

Chlorinated

23

Alcohols

Esters

Glycols

Ethers

Amides

Amines

Phenols

Methyl ethyl ketone

Chloroform

Cyclohexanone

Dichloromethane

THF

Nitriles

Toluene

Dimethylsulfoxide

Isooctane Nitromethane

Bold type indicates relatively high δD 24

20

N

Nitriles A S

δP, Polar Parameter

16

Amides F

Alcohols

B

12

M

A Ketones

E

M E

8

C

Chlorinated

Glycols

Phenols

E T

P C

B Esters

4

P

B

O

Ethers E P

Tol

Amines

0 0

I

4

8

12

16

20

24

28

δH, Hydrogen Bonding Parameter

FIGURE 1.4 δp vs. δH plot showing the location of various common solvents. The glycols are ethylene glycol and propylene glycol. The alcohols include methanol (M), ethanol (E), 1-b utanol (B), and 1-octanol (O). The amides include dimeth yl formamide (F) and dimeth yl acetamide (A). The nitriles are acetonitrile (A) and butyronitrile (B). The esters are ethyl acetate (E) and n-butyl acetate (B). The amines are ethyl amine (E) and propyl amine (P). The phenols are phenol ( P) and m-cresol ( C). The ethers are symbolized by dieth yl ether.

the tools is further confirmed in Chapter 2 and Chapter 3, and their use is demonstrated in all th subsequent chapters. Figure 1.4 is included to sho w where many common solvents are located on a δp vs. δH plot.

7248_C001.fm Page 24 Monday, April 23, 2007 1:56 PM

24

Hansen Solubility Parameters: A User’s Handbook

REFERENCES 1. Hildebrand, J. and Scott, R.L., The Solubility of Nonelectrolytes, 3rd ed., Reinhold, Ne w York, 1950. 2. Hildebrand, J. and Scott, R.L., Regular Solutions, Prentice-Hall, Englewood Cliffs, NJ, 1962. 3. Patterson, D. and Delmas, G., Ne w aspects of polymer solution thermodynamics, Off. Dig. Fed. Soc. Paint Technol., 34(450), 677–692, 1962. 4. Delmas, D., Patterson, D., and Somcynsky, T., Thermodynamics of polyisobutylene-n-alkane systems, J. Polym. Sci., 57, 79–98, 1962. 5. Bhattacharyya, S.N., Patterson, D., and Somcynsky, T., The principle of corresponding states and the excess functions of n-alkane mixtures, Physica, 30, 1276–1292, 1964. 6. Patterson, D., Role of free v olume changes in polymer solution thermodynamics, J. Polym. Sci. Part C, 16, 3379–3389, 1968. 7. Patterson, D.D., Introduction to thermodynamics of polymer solubility , J. Paint Technol., 41(536), 489–493, 1969. 8. Biros, J., Zeman, L., and Patterson, D., Prediction of the C parameter by the solubility parameter and corresponding states theories, Macromolecules, 4(1), 30–35, 1971. 9. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press, Boca Raton, FL, 1983; 2nd ed., 1991. 10. Gardon, J.L. and Teas, J.P., Solubility parameters, in Treatise on Coatings, Vol. 2, Characterization of Coatings: Physical Techniques, Part II, Myers, R.R. and Long, J.S., Eds., Marcel Dekk er, Ne w York, 1976, chap. 8. 11. Burrell, H., Solubility parameters for film formers, Off. Dig. Fed. Soc. Paint Technol., 27(369), 726–758, 1972; Burrell, H., A solvent formulating chart, Off. Dig. Fed. Soc. Paint Technol., 29(394), 1159–1173, 1957; Burrell, H., The use of the solubility parameter concept in the United States, VI Federation d’Associations de Techniciens des Industries des Peintures, Vernis, Emaux et Encres d’Imprimerie de l’Europe Continentale, Congress Book, 21–30, 1962. 12. Blanks, R.F . and Prausnitz, J.M., Thermodynamics of polymer solubility in polar and nonpolar systems, Ind. Eng. Chem. Fundam., 3(1), 1–8, 1964. 13. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities I, J. Paint Technol., 39(505), 104–117, 1967. 14. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities II, J. Paint Technol., 39(511), 505–510, 1967. 15. Hansen, C.M. and Skaarup, K., The three dimensional solubility parameter — key to paint component affinities III, J. Paint Technol., 39(511), 511–514, 1967. 16. Hansen, C.M., The Three Dimensional Solubility P arameter and Solvent Diffusion Coefficient, Doc toral dissertation, Danish Technical Press, Copenhagen, 1967. 17. Hansen, C.M. and Beerbo wer, A., Solubility parameters, in Kirk-Othmer Encyclopedia of Chemical Technology, Suppl. Vol., 2nd ed., Standen, A., Ed., Interscience, Ne w York, 1971, pp. 889–910. 18. Barton, A.F.M., Applications of solubility parameters and other cohesion ener gy parameters, Polym. Sci. Technol. Pure Appl. Chem., 57(7), 905–912, 1985. 19. Sørensen, P ., Application of the acid/base concept describing the interaction between pigments, binders, and solv ents, J. Paint Technol., 47(602), 31–39, 1975. 20. Van Dyk, J.W ., P aper presented at the F ourth Chemical Congress of America, Ne w York, August 25–30, 1991. 21. Anonymous [Note: This was, in fact, Van Dyk, J.W., but this does not appear on the b ulletin], Using Dimethyl Sulfoxide (DMSO) in Industrial F ormulations, Bulletin No. 102, Gaylord Chemical Corp., Slidell, LA, 1992. 22. Karger, B.L., Sn yder, L.R., and Eon, C., Expanded solubility parameter treatment for classificatio and use of chromatographic solv ents and adsorbents, Anal. Chem., 50(14), 2126–2136, 1978. 23. Hansen, C.M. and Wallström, E., On the use of cohesion parameters to characterize surfaces, J. Adhes., 15, 275–286, 1983. 24. Hansen, C.M., Characterization of surfaces by spreading liquids, J. Paint Technol., 42(550), 660–664, 1970. 25. Hansen, C.M., Surface dewetting and coatings performance, J. Paint Technol., 44(570), 57–60, 1972.

7248_C001.fm Page 25 Monday, April 23, 2007 1:56 PM

Solubility Parameters — An Introduction

25

26. Hansen, C.M. and Pierce, P.E., Surface effects in coatings processes, Ind. Eng. Chem. Prod. Res. Dev., 13(4), 218–225, 1974. 27. Hennissen, L., Systematic Modification of Filler/Fiber Sur aces to Achieve Maximum Compatibility with Matrix Polymers, Lecture for the Danish Society for PolymerTechnology, Copenhagen, February 10, 1996. 28. Gardon, J.L., Critical review of concepts common to cohesive energy density, surface tension, tensile strength, heat of mixing, interf acial tension and b utt joint strength , J. Colloid Interface Sci., 59(3), 582–596, 1977. 29. Flory, P.J., Principles of Polymer Chemistry, Cornell University Press, New York, 1953. 30. Prigogine, I. (with the collaboration of Bellemans, A. and Mathot, A.), The Molecular Theory of Solutions, North-Holland, Amsterdam, 1957, chap. 16, 17. 31. van Krevelen, D.W. and Hoftyzer, P.J., Properties of Polymers: Their Estimation and Correlation with Chemical Structure, 2nd ed., Else vier, Amsterdam, 1976. 32. Beerbower, A., Environmental Capability of Liquids, in Interdisciplinary Approach to Liquid Lubricant Technology, NASA Publication SP-318, 1973, 365–431. 33. Fedors, R.F., A method for estimating both the solubility parameters and molar v olumes of liquids, Polym. Eng. Sci., 14(2), 147–154, 472, 1974. 34. Hansen, C.M., Solubility parameters, inPaint Testing Manual, Manual 17, Koleske, J.V., Ed., American Society for Testing and Materials, Philadelphia, P A, 1995, pp. 383–404. 35. Koenhen, D.N. and Smolders, C.A., The determination of solubility parameters of solv ents and polymers by means of correlation with other physical quantities, J. Appl. Polym. Sci., 19, 1163–1179, 1975. 36. Anonymous, Brochure: Co-Act — A Dynamic Program for Solv ent Selection, Exxon Chemical International Inc., 1989. 37. Dante, M.F., Bittar , A.D., and Caillault, J.J., Program calculates solv ent properties and solubility parameters, Mod. Paint Coat., 79(9), 46–51, 1989. 38. Hoy, K.L., New values of the solubility parameters from vapor pressure data, J. Paint Technol., 42(541), 76–118, 1970. 39. Myers, M.M. and Abu-Isa, I.A., Elastomer solv ent interactions III — ef fects of methanol mixtures on fluorocarbon elastomers, J. Appl. Polym. Sci., 32, 3515–3539, 1986. 40. Hoy, K.L., Tables of Solubility Parameters, Union Carbide Corp., Research and De velopment Dept., South Charleston, WV, 1985; 1st ed. 1969. 41. Reid, R.C. and Sherw ood, T.K., Properties of Gases and Liquids, McGra w-Hill, Ne w York, 1958 (Lydersen Method — see also Reference 31). 42. McClellan, A.L., Tables of Experimental Dipole Moments, W.H. Freeman, San Francisco, 1963. 43. Tables of Physical and Thermodynamic Properties of Pure Compounds, American Institute of Chemical Engineers Design Institute for Physical Property Research, Project 801, Data Compilation, Danner, R.P. and Daubert, T.E., Project Supervisors, DIPPR Data Compilation Project, Department of Chemical Engineering, Pennsylvania State University, University Park. 44. Hansen, C.M., Selection of Chemicals for Permeation Testing Based on Ne w Solubility P arameter Models for Challenge 5100 and Challenge 5200, under contract DTCG50-89-P-0333 for the U.S. Coast Guard, June 1989, Danish Isotope Centre, Copenhagen. 45. Weast, R.C., (Editor-in-Chief), CRC Handbook of Chemistry and Physics, 65th ed., CRC Press, Boca Raton, FL, 1988–1989, pp. C-672–C-683. 46. Majer, V., Enthalpy of v aporization of or ganic compounds, in Handbook of Chemistry and Physics, 72nd ed., Lide, D.R., (Editor -in-Chief), CRC Press, Boca Raton, FL, 1991–1992, pp. 6-100–6-107. 47. Fishtine, S.H., Reliable latent heats of v aporization, Ind. Eng. Chem., 55(4), 20–28, 1963; Ind. Eng. Chem., 55(5), 55–60; Ind. Eng. Chem., 55(6), 47–56. 48. Hansen, C.M., Ne w developments in corrosion and blister formation in coatings, Prog. Org. Coat., 26, 113–120, 1995. 49. Hansen, C.M., Solvent Resistance of Polymer Composites — Glass Fiber Reinforced Polyphenylene Sulfide, Centre for Polymer Composites (Denmark), Danish Technological Institute, Taastrup, 1993, 1–62, ISBN 87-7756-286-0. 50. Hansen, C.M., A mathematical description of film drying by sol ent evaporation, J. Oil Color Chem. Assoc., 51(1), 27–43, 1968.

7248_C001.fm Page 26 Monday, April 23, 2007 1:56 PM

26

Hansen Solubility Parameters: A User’s Handbook 51. Harrington, E.C., Jr., The desirability function, Ind. Qual. Control, 21(10), 494–498, April 1965. 52. Hansen, C.M. and Andersen, B.H., The affinities of o ganic solvents in biological systems, Am. Ind. Hyg. Assoc. J., 49(6), 301–308, 1988.

7248_C002.fm Page 27 Wednesday, May 9, 2007 8:25 AM

— The Prigogine 2 Theory Corresponding States Theory, χ 12 Interaction Parameter, and Hansen Solubility Parameters Charles M. Hansen ABSTRACT Patterson has sho wn that the χ12 interaction parameter can be estimated from the corresponding states theory (CST) of Prigogine. Correlations using Hansen solubility parameters (HSP) confir the usage of the term cohesive energy difference proposed in the Prigogine CST . Therefore, the HSP approach can be expected to be useful to predict the Flory interaction coefficient, χ12. Equations for this purpose are presented and discussed based on comparisons of calculated and e xperimental values for fi e polymers. There is agreement in man y cases, especially for essentially nonpolar systems, b ut full understanding of the interrelationship has not yet been achie ved. The lack of accounting for permanent dipole–permanent dipole and h ydrogen bonding (electron interchange) in the “New Flory” theory leading to the coefficient χ12” is thought to be largely responsible for this. It does appear , however, that the constant “4” (or 0.25) in the HSP correlations and the 0.25 in the leading term of the Prigogine theory ha ve identical functions. They modify the specifi interactions described by the Prigogine δ and also the polar and h ydrogen bonding HSP ( δP and δH). This could imply that the Prigogine ρ attempts to describe what the δD parameter describes, that is, the nondirectional (nonpolar) atomic interactions. Neither the Flory nor the Prigogine approaches can lead to the type of predictions possible with the HSP approach. The many correlations and other predictions contained in this book w ould not be possible with these theories, as the y do not separate the polar and h ydrogen bonding effects independently. The Prigogine theory must be used with the geometric mean to estimate the inter action between different species. The Hildebrand and HSP approaches inherently use the geometric mean. This implies that the geometric mean is capable of describing not only dispersion interactions but also those due to permanent dipoles and h ydrogen bonding.

INTRODUCTION The Flory–Huggins “chi” parameter, χ, has been used for man y years in connection with polymer solution behavior,1,2 but now the χ12 parameter derived from the New Flory theory is being currently accepted for general use instead of the older χ. It w ould be desirable to relate the widely used HSP3–10 more directly to χ12. This w ould allo w estimates of χ12 for systems where the HSP are known, b ut χ12 is not. The re verse is not possible as a single χ12 parameter cannot be used to divide the cohesion ener gy into contrib utions from dispersion (nonpolar) forces, permanent 27

7248_C002.fm Page 28 Wednesday, May 9, 2007 8:25 AM

28

Hansen Solubility Parameters: A User’s Handbook

dipole–permanent dipole forces, and h ydrogen bonding (electron interchange), which is the basis of the HSP. Reliable χ12 values for numerous solvents and the same polymer can be used to determine the HSP for the polymer, in the same manner as solv ency or swelling data being used for a similar purpose. In principle, the weighting schemes (described in Chapter 5) to a verage the solv ent parameters for obtaining the polymer HSP can also be used with the χ12 parameter, just as they are used with weight g ain or intrinsic viscosity . Patterson15 and coworkers17 have shown how to predict the χ12 parameter using corresponding states theories (CST),2,11–17 as well as using the Hildebrand solubility parameter in (strictly) nonpolar systems. They use the symbol ν2 instead of χ12 for the same quantity . The Hildebrand solubility parameter is the square root of the cohesi ve ener gy density (ced). 18,19 Also, it has been sho wn recently that HSP and the Prigogine/P atterson CST are mutually confirming and g ve similar predictions.20,21 This is discussed in more detail later . The customary equation to calculate χ12 from the Hildebrand solubility parameters for a nonpolar solvent and a nonpolar polymer is: χ12 = [V( δ1 – δ2)2]/RT + β (2.1) where V is the molar volume of the solvent, δ is the Hildebrand solubility parameter for the solvent (1) and polymer (2), R is the g as constant, and T is the absolute temperature. The empirical constant β has been discussed as being necessary for polymer systems, 22 as a correction to the Flory combinatorial entrop y. β, although combinatorial in origin, w as attached to χ12 in order to preserv e the Flory form of the chemical potential e xpression. β has a generally accepted average value near 0.34. Biros et al. 17 state that this v alue of β presents difficulties as a explanation of an error in the Flory combinatorial entrop y approximation. These authors state that β should be interpreted as aligning the χ12 values from the solubility parameters with those found from CST. The CST predict χ to be about 0.3 units larger than that found when using the Hildebrand solubility parameter. β is not required for essentially nonpolar systems when HSP are used in a relation similar to Equation 2.1, as sho wn later. The Hildebrand parameters are applicable to re gular solutions, which, in the current conte xt, implies strictly nonpolar systems. Hildebrand solubility parameters ha ve been shown to reflect th 12,13 (see also Chapter 1, noncombinatorial free ener gy directly via the first term in Equation 2. Equation 1.3 and Equation 1.4). It w as previously thought that the heat of mixing w as given by the Hildebrand theory as φ1φ2VM(δ1 – δ2)2, where VM is the v olume of the mixture and the ϕs are volume fractions of the solvent and polymer, respectively. This is not true. The heat of mixing must be found by dif ferentiating this relation as sho wn by Delmas and co workers.12,13 The w ork of Delmas, Patterson, and co workers has sho wn that predictions with the nonpolar Hildebrand solubility parameter and the Prigogine CST are in e xcellent agreement with each other with re gard to heats of mixing in essentially nonpolar systems. Both positi ve and ne gative heats of mixing are allowed, predicted, and found. The argument that solubility parameters are inadequate, as the y do not allow for negative heats of mixing, is not valid. These studies also show that subsequent increases in temperature lead to improved solvency when a solvent has higher solubility parameters than the polymer. When the solv ent has lo wer solubility parameters than the polymer , an increase in temperature leads to poorer solv ency. Precipitation can e ven occur with increasing temperature. This temperature is called the lower critical solution temperature. (See also the discussion in Chapter 1.)

HANSEN SOLUBILITY PARAMETERS (HSP) It has been shown that the total energy of vaporization can be divided into at least three parts.6 These parts come from the nonpolar/dispersion (atomic) forces, ED; the permanent dipole–permanent dipole

7248_C002.fm Page 29 Wednesday, May 9, 2007 8:25 AM

Theory

29

(molecular) forces, E P; and h ydrogen bonding (molecular) forces, E H. The latter is more generally called the electron exchange energy. E = ED + E P + E H

(2.2)

E/V = E D/V + E P/V + E H/V

(2.3)

δ2 = δD2 + δP2 + δH2 (2.4) δD, δP , and δH are the HSP for the dispersion, polar, and hydrogen bonding interactions, respectively. δ is the Hildebrand solubility parameter , (E/V) 1/2. It might be noted that the v alue of a solubility parameter in MP a1/2 is 2.0455 times lar ger than in the often used (cal/cm 3)1/2 units. As described in Chapter 1, a corresponding states calculation using h ydrocarbons as reference is used to find the part of the cohes ve energy of a liquid that is attributable to dispersion (nonpolar) forces. Subtracting this nonpolar contrib ution from the total cohesion ener gy then gi ves the sum of the permanent dipole–permanent dipole and h ydrogen bonding (electron interchange) contrib utions to the total cohesion ener gy. These can then be separated by calculation and/or e xperiment into the polar and h ydrogen bonding parameters. HSP also include v olume ef fects, as the y are based on cohesion energy density. Volume effects are also basic to the Prigogine CST. As described in Chapter 3, P anayiotou and co workers ha ve used a statistical thermodynamics approach to calculate all three parameters, thus giving support to the approach used in Equation 2.2 to Equation 2.4. HSPs have been applied to the study of polymer solubility and swelling, biological materials, barrier properties of polymers, surf aces,4,20,23–26 etc. and ha ve been described in greater detail elsewhere7,8,10 (see also the follo wing chapters). The three parameters described in Equation 2.4 are fundamental energy parameters that can be calculated from the mutual interactions of identical molecules in a pure liquid. The quantities required are E, V, the dipole moment (and perhaps the refractive index and the dielectric constant), and generalized corresponding states correlations for hydrocarbons, (E D). Group contrib ution methods and simpler calculational procedures ha ve also been established.10 These procedures were described in Chapter 1. The calculated values for a large number of the liquids ha ve been confirmed xperimentally by solubility tests. The usual equation used in HSP correlations is: (Ra)2 = 4( δD2 – δD1)2 + ( δP2 – δP1)2 + ( δH2 – δH1)2

(2.5)

Ra, in this equation, is a modified di ference between the HSP for a solv ent (1) and polymer (2). Ra must not e xceed Ro, the radius of interaction of a HSP solubility sphere, if solubility is to be maintained. Both Ra and Ro ha ve the same units as solubility parameters. These correlations have been v ery convenient for practical use, for e xample, in solv ent selection. The constant “4” has been found empirically useful to convert spheroidal plots of solubility into spherical ones using δD and either of the other parameters (see Chapter 5). It has been used with success in well o ver 1000 HSP correlations with a computer program that optimizes a solubility sphere according to Equation 2.5, where all the good solv ents are within the sphere and the bad ones are outside. This program was described in Chapter 1. This experimental procedure is still thought to be the best way to arri ve at the HSP for polymers; the polymer HSPs being gi ven by the coordinates for the center of the sphere. The reliability of the spherical characterizations and the need to di vide the total cohesion ener gy (E) into at least three parts has been confirmed by systematically locatin nondissolving solvents that are syner gistic and dissolv e a given polymer when mix ed.3 They only need to be located on opposite sides of the sphere of solubility for the gi ven polymer.

7248_C002.fm Page 30 Wednesday, May 9, 2007 8:25 AM

30

Hansen Solubility Parameters: A User’s Handbook

For present purposes of comparison, Equation 2.5 must be normalized by 4R M2 to mak e its predictions consistent with the quantities commonly used in the literature, in connection with the CSTs and χ12: Η = (RA) 2/RM2 = [( δD2 – δD1)2 + ( δP2 – δP1)2/4 + ( δH2 – δH1)2/4]/RM2

(2.6)

RA = Ra/2; R M = Ro/2

(2.7)

RM is the maximum solubility parameter dif ference that still allo ws the polymer to dissolv e based on Equation 2.6. R M is the radius of a HSP sphere (spheroid) based on Equation 2.7. The HSP difference between solv ent and polymer , RA, must be less than R M for solution to occur . It can be seen that the quantities RA/R M and H = (RA/R M)2 will be 1.0 on the boundary surf ace of a sphere describing polymer solubility . RA/R M is a ratio of solubility parameters, whereas H is a ratio of cohesion ener gy densities. H is zero when the solubility parameters for the solv ent and polymer match and subsequently increases to lar ger values as the differences between solvent and polymer increase.

RESEMBLANCE BETWEEN PREDICTIONS OF HANSEN SOLUBILITY PARAMETERS AND CORRESPONDING STATES THEORIES Patterson and co workers 15,17 have explained the Prigogine theory in concise form and simplifie some of the most important aspects. A key parameter is the Prigogine δ. This describes normalized cohesive energy differences between polymer se gments and solv ents. The Prigogine δ parameter is defined a δ = ( ε2 – ε1)/ε1

2.8)

where ε is the cohesi ve energy for a polymer se gment (2) or for a solv ent (1). For the present discussion, it is adv antageous to define the Prigogine δ using cohesive energy densities as follows: δ = [(ced 2)1/2 – (ced 1)1/2]2/(ced1) = ( δ1 – δ2)2/δ12

(2.9)

The numerator is the difference in cohesion energy densities between solvent and polymer, and this is normalized by the ced of the solv ent. As indicated abo ve, cohesion ener gies (HSP) for solvents can be calculated whereas those for polymers currently require e xperimental data on solubility or other rele vant testing procedures. The Prigogine ρ accounts for differences in the size of the solv ent and polymer segments. The segmental distance parameter is s. ρ is defined a ρ = ( σ2 – σ1)/σ1

(2.10)

Another k ey parameter in the Prigogine/P atterson CST is ν2, which is in f act equal to χ12.15 “ν ” is approximated by 2

ν2 app = ( δ2/4 + 9 ρ2) app = ( δ/2 – 3 ρ/2)2

(2.11)

“ν2” includes effects from differences in segmental energy (the δ effect) and in segmental size (the ρ effect). The geometric mean rule [ ε12 = ( ε1ε2)1/2] was used to arri ve at this result, just as it w as used to arrive at the equations ha ving differences in solubility parameters. P atterson states that the

7248_C002.fm Page 31 Wednesday, May 9, 2007 8:25 AM

Theory

31

coefficient in front of ρ is uncertain15 and, furthermore,27 that “ρ was a (misguided) attempt to take into account se gment-size differences.” Patterson27 has recently helped the author to clarify some points about the relations among the theories. I feel a fe w quotes from these communications, in addition to the one just cited, are in order at this point: In my opinion, the Flory theory w as a v ery usable and particularly successful case of the Prigogine theory (which w as in f act difficult to use). An additional point that Flory made much of b ut was only touched by Prigogine, is that surf ace/volume fraction of the polymers are v ery different from those of solvents, i.e., the polymer is very bulky. These are interesting differences between the Flory theory and that of Prigogine. But, again in my opinion, Flory always presented his theory as something absolutely different from that of Prigogine’s, using different symbols, different names for concepts, etc. In reaction against this I ha ve al ways lik ed to call the whole thing the Prigogine–Flory theory. Ho wever, since about 1970, I ha ve done v ery little with polymer solutions and, hence, when using the term Prigogine–Flory theory, I have used it with respect to mixtures of small molecules and not polymer solutions. I think that the w ork by a lot of people has established the utility of the Prigogine–Flory theory , or if you like, the Flory theory , for small-molecule mixtures. Also: Very specificall , the Prigogine parameters delta and rho are no w out of f ashion, and the y have been lumped together in the χ12 parameter. P articularly, the rho parameter does not ha ve nearly as much importance as Prigogine thought, and Flory completely discarded it. I think, too, the origin of the parameter beta in the solubility parameter approach w ould, in the Prigogine–Flory approach, be ascribed to free-volume differences, which must inevitably exist between any polymer and an y solvent and which gi ve a contrib ution to the chi parameter ….

This demonstrates that there is not complete agreement among those who ha ve concerned themselves with these theories. The following is an attempt to unify all of these thoughts. The ideas have not been fully tested as of yet, b ut the implications appear v ery clear to the author , at least. The discussion concentrates on the ν2 parameter, being lo yal to the P atterson article 15 (sometimes referred to as the Prigogine–Patterson theory) where this part of the book got its start. More specificall , ν2 accounts for se gmental ener gy dif ferences, and dif ferences in size of solvent and polymer se gments for breakage of solv ent–solvent (1–1) bonds and polymer–polymer (2–2) bonds to allo w formation of solv ent–polymer (1–2) bonds. In nonpolar systems, the Prigogine δ is small (perhaps zero in this conte xt), and the quantity ν2 depends on se gmental-size differences only. The Prigogine δ parameter becomes important in systems with specific interactions, i.e., those with polar and ydrogen bonding. Differences in ced arising from these sources in such systems are modified by a actor of 0.25 according to Equation 2.11. If we no w consider Equation 2.6, it can be seen that each of the three terms in this equation is in the form of a Prigogineδ as given by Equation 2.9.These terms describe normalized differences in the respective types of cohesive energy in corresponding states terminology. The cohesive energy differences in Equation 2.6 are normalized by R M2, the ced of the worst possible good solvent, i.e., a solv ent located on the boundary of a Hansen solubility sphere rather than the ced of a gi ven solvent under consideration. In strictly nonpolar systems, the polar and h ydrogen bonding terms in Equation 2.6 are zero, and the interaction is described by the dif ference in δD. One is led to the conclusion that the firs term in Equation 2.6 relates directly to the second term (the ρ ef fect) in Equation 2.11. In the future, this relationship could be explored in more detail with the hope of experimental verificatio of the coefficient in front of ρ.

7248_C002.fm Page 32 Wednesday, May 9, 2007 8:25 AM

32

Hansen Solubility Parameters: A User’s Handbook

If we no w consider a system where δD1 is equal to δD2, the polymer–solv ent interaction will be either polar or hydrogen bonding (or both) in nature, i.e., there will only be specific interactions Such differences in δP and δH will be modified by 0.25 in Equation 2.6. It is not worthy that the same f actor, 0.25, modifies the Prigogine δ term in Equation 2.11, i.e., when there are specifi interactions between solv ent and polymer. The same 0.25 is present for a similar purpose both in Equation 2.6 (HSP) and in Equation 2.11 (CST) where the geometric mean was used. The geometric mean appears to be applicable to all types of ener gies discussed here.

THE χ12 PARAMETER AND HANSEN SOLUBILITY PARAMETERS Patterson and coworkers have shown that χ12 can be calculated, using ν2 as a symbol for the same quantity.17 Therefore, according to the pre vious discussion, it is e xpected that Equation 2.6 can be used in a similar w ay to predict χ12. There remains a general belief that χ12 can be calculated by Equation 2.1 using Hildebrand solubility parameters and a value of 0.34 for β. The change required to progress from the nonpolar Hildebrand solubility parameter to include polar and h ydrogen bonding ef fects with the HSP , in calculating χ12, is to replace the Hildebrand solubility parameter dif ference of Equation 2.1 by a corresponding HSP term, i.e., A1,2 A1,2 = [( δD2 – δD1)2 + 0.25( δP2 – δP1)2 + 0.25( δH2 – δH1)2]

(2.12)

and χ12 is estimated from: χ12 = VA1,2/RT

(2.13)

The empirical f actor β (0.34) in Equation 2.1 w as found from studies on almost nonpolar systems using the Hildebrand solubility parameters. This is an a verage correction to these calculations because of the ne glect of some relati vely small b ut significant alues of ( δP1 – δP2) and/or (δH1 – δH2). β is not required in Equation 2.13. This same assumption w as made by Zellers and coworkers in their approach to correlate the swelling and permeation of elastomers used in chemical protective clothing. 28–31 An estimate for χ12 can also be found by noting that the total χ12 parameter in common solutions of polymers ha ving high molecular weight is required to be close to 0.5 at the point of mar ginal solution or precipitation. 1 This boundary v alue is called the critical chi parameter, χc. In HSP terminology, this is a boundary solv ent with a placement directly on the sphere of solubility , and the quality is indicated by H in Equation 2.6 being equal to 1.0. This allows a simple estimate for χ12 for higher-molecular-weight polymers by the relation: χ12 = χc(RED)2 = Η/2 (2.14) This last equation assumes an a verage V, just lik e the HSP correlations ha ve done up to the point. It has been noted man y times that liquids with lo wer V are often better solvents than liquids having essentially identical HSP b ut with lar ger V. This is seen with liquids lik e methanol (V = 40.7 cc/mol) and acetone (V = 74.0 cc/mol), which are sometimes good solv ents, even when H is greater than 1.0 and for liquids lik e the phthalate and other plasticizers (V > 150 cc/mol), which are not good solvents in spite of H being less than 1.0.An explanation for this is found by comparing Equation 2.13 with Equation 2.14. Equation 2.15 can be deri ved from this comparison. The dependency of χc on polymer molecular size is also included in Equation 2.15, as this is a partial explanation for some of the results discussed later .1

7248_C002.fm Page 33 Wednesday, May 9, 2007 8:25 AM

Theory

33

RM2 = (Ro/2) 2 = {0.5(1 + 1/r 1/2)RT/V} (2.15) “r” is the ratio of the polymer size to that of the solvent and is usually considered as the approximate degree of polymerization, assuming the size of the solv ent molecule as being close to that of the polymer segment size. For a solvent with V = 100 cc/mol, this added term leads to a correction of 1.1 for a polymer molecular weight of 10,000 and to a correction of 1.03 for a polymer molecular weight of 100,000. The correction term is e xpressed more generally by the relation in Equation 2.16. Correction = {0.5(r1–1/2 + r2–1/2)2} (2.16) Here, r1 and r2 are the number of statistical segments in the molecules in question. For mixtures of low molecular weight, where r 1 is approximately equal to r 2 and both are approximately equal to 1, the correction amounts to a f actor of about 2; when one of the molecules is a high polymer , the correction amounts to 0.5 (as discussed earlier); when both the molecules ha ve v ery high molecular weights, the correction approaches 0, meaning compatible mixtures of polymers are very difficult to achi ve. A study of the kind discussed in the follo wing for smaller molecules w ould seem appropriate to help clarify some of the questions raised. Table 2.1 gives the expected RM and Ro values based on Equation 2.15 for polymers of molecular weight high enough to ne glect the effect of r. It is usually assumed that V for the average solvent is near 100 cc/mol. Table 2.1 indicates that all polymers of reasonably high molecular weight will be insoluble in solv ents with V greater than about 100 cc/mol for a corresponding RA that is greater than an R M of 3.5 MPa1/2 (Ra greater than 7.0 MPa1/2). This is not generally the case as man y values of R M have been reported that are much higher than this 10 (see also Appendix, Table A.2); meaning the y are more easily dissolv ed than Equation 2.15 indicates. Values for R M greater than 5 MP a1/2 are common, with some polymer R M values being considerably larger, although these are generally for lower-molecular-weight materials. This immediately points to potential problems in directly calculating χ12 from HSP data when R M is significantly la ger than about 3.5 MP a1/2. Some improvement in the estimates of χ12 using Equation 2.14 is possible by including V in a correction term. Equation 2.14 has been retained for present purposes of comparison, ho wever, because of its simplicity . The column for χ12, estimated from Equation 2.14 in Table 2.3 to Table 2.7, is placed adjacent to that of the solv ent molar volume to allow an easy mental multiplication by V/100 if desired.

TABLE 2.1 Expected Solubility Parameter Differences for Marginal Solubility as a Function of the Molecular Volume of the Solvent V (cc/mol) 50 100 200

RM (MPa1/2) 5.0 3.5 7.0 2.5

Ro (MPa1/2) 10.0 5.0

7248_C002.fm Page 34 Wednesday, May 9, 2007 8:25 AM

34

Hansen Solubility Parameters: A User’s Handbook

TABLE 2.2 Hansen Solubility Parameter Data for Polymers Selected for the Comparisons Given in Table 2.2 to Table 2.6 Polymers

δD

δP

δH

Ro

Polybutadiene Polyisobutylene Polystyrene Polyvinylacetate Polyacrylonitrile

17.5 16.9 21.3 20.9 21.7

2.3 2.5 5.8 11.3 14.1

3.4 4.0 4.3 9.7 9.1

6.5 7.2 12.7 13.7 10.9

Note: Units are MPa1/2. Values in these units are 2.0455 times lar ger than in (cal/cm 3)1/2.

TABLE 2.3 Comparison of Experimental, Indicative Chi Parameter Data, χlit, with Calculations Based on HSP for Polybutadiene, Buna Hüls CB 10 cis-Polybutadiene Raw Elastomer, Chemische Werke Hüls Solvent

V

Benzene Toluene Xylene Pentane n-Hexane n-Heptane n-Octane Chloroform Carbon tetrachloride Methanol Water

89.4 106.8 123.3 116.2 131.6 147.4 163.5 80.7 97.1 40.7 18.0

χ12 (Equation 2.14) 0.12 0.04 0.02 0.62 0.52 0.43 0.39 0.07 0.16 5.68 20.3

χ12 (Equation 2.13)

χlit

0.10 0.04 0.08 0.62 0.58 0.54 0.54 0.05 0.13 1.97 3.1

0.4 0.3 0.4 0.7 0.6 0.5 0.6 0.15 0.3 3.3 3.5

Source: Solubility data from Hansen, C.M., J. Paint Technol., 39(505), 104–117, 1967; Hansen, C.M., The Three Dimensional Solubility P arameter and Solv ent Diffusion Coefficient, Their Importance in Surface Coating Formulation, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967.

COMPARISON OF CALCULATED AND EXPERIMENTAL χ12 PARAMETERS The predictions of χ12 using Equation 2.13 and Equation 2.14 ha ve been compared with χ12 parameter data in standard references.32,33 The polymers used for the comparisons are listed inTable 2.2 with their HSP data. The calculated and indicati ve e xperimental values for χ12 for the gi ven solvent–polymer systems are reported in Table 2.3 to Table 2.7. The polymers were chosen because

7248_C002.fm Page 35 Wednesday, May 9, 2007 8:25 AM

Theory

35

TABLE 2.4 Comparison of Experimental, Indicative Chi Parameter Data, χlit, with Calculations Based on HSP for Polyisobutylene, Lutonal® I60, BASF Solvent

V

χ12 (Equation 2.14)

χ12 (Equation 2.13)

χlit

Benzene Toluene Decalin Cyclohexane Pentane n-Hexane n-Octane n-Nonane Chloroform Carbon tetrachloride Methylene dichloride

89.4 106.8 156.9 108.7 116.2 131.6 163.5 179.7 80.7 97.1 63.9

0.19 0.10 0.35 0.20 0.44 0.37 0.29 0.27 0.06 0.20 0.25

0.17 0.11 0.43 0.23 0.53 0.49 0.50 0.51 0.05 0.21 0.17

0.5 0.5 0.4 0.45 0.5 0.5 0.5 0.3+ 1.0 0.5 0.6

Source: Solubility parameter data from Hansen, C.M., Solubility parameters, in Paint Testing Manual, Manual 17, Koleske, J.V., Ed., American Society for Testing and Materials, Philadelphia, 1995, pp. 383–404.

of similarities/differences in the HSP data, as well as the a vailability of suf ficient data on the χ12 parameter. This is not a complete e valuation of Equation 2.13 and Equation 2.14, b ut it points out clearly that there are factors which are not completely understood. Table 2.8 provides these solvents along with their HSP. A casual inspection of the measured and calculated χ12 values in Table 2.3 through Table 2.7 would gi ve the impression that there are significant discrepancies between these alues which require further e xplanation. The calculated and literature v alues for χ12 agree in some cases and differ significantly in others. Some possible reasons for this are discussed in the foll wing sections.

POLYBUTADIENE The calculated and experimental χ12 values for polybutadiene are given in Table 2.3. The first thre entries are for aromatic solv ents. The solubility parameter predictions indicate that these are exceptionally good solvents, whereas the χ12 values indicate that they are moderately good. Before one adds on a constant v alue of about 0.3 to bring an agreement, it should be noted that the calculated and e xperimental χ12 values for the aliphatic solv ents are in good agreement. The solubility parameters for the higher-molecular-weight homologs are closer to those of the polymer, but the size ef fect reduces solv ent quality . Agreement for the aliphatic solv ents is considered excellent. It should be noted that Ro is v ery near the ideal v alue for such calculations according to Table 2.1. Chloroform and carbon tetrachloride are predicted by HSP to be very good solvents, especially confirmed by the χlit for chloroform. HSP considerations indicate that chloroform and the aromatic solvents are near neighbors with similar HSP and might ha ve similar qualities. This is not borne out by the χlit values for the aromatics, which are suspected as being too high for presently unknown reasons. The calculated and literature v alues for methanol and w ater are dif ferent enough to w arrant a comment. HSP considerations indicate that the dif ference in beha vior between these tw o liquids should be sizable, which the χlit v alues do not indicate. A problem of some significance in a y

7248_C002.fm Page 36 Wednesday, May 9, 2007 8:25 AM

36

Hansen Solubility Parameters: A User’s Handbook

TABLE 2.5 Comparison of Experimental, Indicative Chi Parameter Data, χlit, with Calculations Based on HSP for Polystyrene, Polystyrene LG, BASF Solvent

V

χ12 (Equation 2.14)

χ12 (Equation 2.13)

χlit

Benzene Toluene Xylene Ethyl benzene Styrene Tetralin Decalin (cis) Cyclohexane Methyl cyclohexane n-Hexane n-Heptane n-Octane Acetone Methyl ethyl ketone Methyl isobutyl ketone Cyclohexanone Ethyl acetate n-Butyl acetate sec-Butyl acetate

89.4 106.8 123.3 123.1 115.6 136.0 156.9 108.7 128.3 131.6 147.1 163.5 74.0 90.1 125.8 104.0 98.5 132.5 133.6

0.23 0.21 0.25 0.26 0.16 0.09 0.24 0.41 0.49 0.67 0.61 0.58 0.51 0.38 0.45 0.15 0.40 0.40 0.54

0.66 0.73 0.99 1.05 0.61 0.38 1.22 1.44 2.03 2.87 2.92 3.08 1.22 1.12 1.83 0.52 1.29 1.72 2.35

0.40–0.44 0.40–0.44 0.4 0.45 0.35 0.4 0.5 0.50–1.0 0.5 0.8 0.8 0.9 0.6 0.49 0.5 0.5 0.5 0.5 0.4

Source: Solubility data from Hansen, C.M., J. Paint Technol., 39(505), 104–117, 1967; Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Their Importance in Surface Coating Formulation, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967.

study of solvents at low concentrations in polymers is that the smaller amounts of solv ent relative to the polymer can lead to preferential association of the solv ent with those local re gions/segments/groups in the polymer which ha ve similar ener gies (HSP). These local re gions may not necessarily reflect the same a finities as the polymer as a whole, such as are reflected by the solubl or-not approach commonly used in HSP e valuations. These local association ef fects can influenc results on swelling studies in both good and bad solv ents, for e xample. Other types of studies carried out at lo w-solvent concentrations can also be influenced by these s gregation/association phenomena. An extension of this type of situation can be cited in the tendencies of w ater to associate with itself, as well as with local re gions within polymers. This has made simple predictions of its behavior impossible. A detailed discussion of this is beyond the scope of this chapter. It is suggested, however, that the potential dif ferences observed here between HSP predictions and observ ed χlit may be derived from such phenomena.

POLYISOBUTYLENE The calculated and e xperimental χ12 values for polyisob utylene are gi ven in Table 2.4. There are some similarities with polybutadiene both chemically and in the Ro value of 7.2 MPa1/2 being near the ideal for a polymer of v ery high molecular weight. The cyclic and aromatic solv ents are again better as judged by HSP than the χlit values indicate, whereas the estimates for the aliphatic solvents are in e xcellent agreement with Equation 2.13, in particular . Again, HSP finds chloroform

7248_C002.fm Page 37 Wednesday, May 9, 2007 8:25 AM

Theory

37

TABLE 2.6 Comparison of Experimental, Indicative Chi Parameter Data, χlit, with Calculations Based on HSP for Polyvinylacetate, Mowilith® 50, Farbwerke Hoechst Solvent

V

χ12 (Equation 2.14)

χ12 (Equation 2.13)

χlit

Benzene Toluene Decalin (cis) Tetralin Cyclohexane Methyl cyclohexane n-Nonane n-Decane Acetone Methyl ethyl k etone Methyl isobutyl ketone Ethyl acetate Dimethyl phthalate Dioxane Chloroform Chlorobenzene n-Propanol

89.4 106.8 156.9 136.0 108.7 128.3 179.7 195.9 74.0 90.1 125.8 98.5 163.0 85.7 80.7 102.1 75.2

0.56 0.51 0.64 0.37 0.76 0.49 0.88 0.88 0.33 0.33 0.45 0.39 0.12 0.29 0.32 0.33 0.47

1.91 2.05 3.80 1.92 3.13 2.03 6.00 6.54 0.92 1.11 1.83 1.46 0.58 0.95 0.99 1.27 1.34

0.3–0.5 0.5 2.7 1.3 2.4 0.5 3.3 3.4 0.3–0.46 0.4–0.44 0.5 0.4 0.4 0.4 0.4 0.5 1.2–1.6

Source: Solubility data from Hansen, C.M., J. Paint Technol., 39(505), 104–117, 1967; Hansen, C.M., The Three Dimensional Solubility P arameter and Solv ent Dif fusion Coefficient, Their Importance in Surf ace Coating F ormulation, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967.

TABLE 2.7 Comparison of Experimental, Indicative Chi Parameter Data, χlit, with Calculations Based on HSP for Polyacrylonitrile Solvent

V

χ12 (Equation 2.14)

χ12 (Equation 2.13)

χlit

Ethylene carbonate γ-Butyrolactone Ethanol Water N,N-Dimethyl formamide N,N-Dimethyl acetamide Dimethyl sulfoxide Tetramethylene sulfoxide

66.0 76.8 58.5 18.0 77.0 92.5 71.3 90.0

0.40 0.16 1.15 5.3 0.33 0.44 0.21 0.25

0.63 0.30 1.61 2.3 0.61 0.97 0.36 0.53

0.4 0.36–0.40 4.0 2.0 0.2–0.3 0.4 0.3–0.4 0.3

Source: Solubility data from Brandrup, J. and Immer gut, E.H., Eds., Polymer Handbook, 3rd ed., Wiley-Interscience, New York, 1989. (a) Gundert, F . and Wolf, B.A., Polymer solvent interaction parameters, pp. VII/173–182. (b) Fuchs, O., Solvents and non-solvents for polymers, pp. VII/379–407.

7248_C002.fm Page 38 Wednesday, May 9, 2007 8:25 AM

38

Hansen Solubility Parameters: A User’s Handbook

TABLE 2.8 Hansen Solubility Parameters for the Solvent Included in Table 2.3 to Table 2.7 Solvent

δD

δP

δH

Solvent

δD

δP

δH

Benzene Toluene Xylene Ethyl benzene Styrene Decalin (cis) Tetralin Cyclohexane Methyl cyclohexane n-Pentane n-Hexane n-Heptane n-Octane n-Nonane Acetone Methyl ethyl k etone Methyl isobutyl ketone Cyclohexanone Ethyl acetate

18.4 18.0 17.6 17.8 18.6 18.0 19.6 16.8 16.0 15.6 14.9 15.3 15.5 15.7 15.5 16.0 15.3 17.8 15.8

0.0 1.4 1.0 0.6 1.0 0.0 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 10.4 9.0 6.1 6.3 5.3

2.0 2.0 3.1 1.4 4.1 0.0 2.9 0.2 1.0 0.0 0.0 0.0 0.0 0.0 7.0 5.1 4.1 5.1 7.2

n-Butyl acetate sec-Butyl acetate Dimethyl phthalate 1,4-Dioxane Chloroform Chlorobenzene Carbon tetrachloride Methylene dichloride Methanol Ethanol n-Propanol Ethylene carbonate γ-Butyrolactone N,N-dimethyl formamide N,N-dimethyl acetamide Dimethyl sulfoxide Tetramethylene sulfoxide Water

15.8 15.0 18.6 19.0 17.8 19.0 17.8 18.2 15.1 15.8 16.0 19.4 19.0 17.4 16.8 18.4 18.2 15.5

3.7 3.7 10.8 1.8 3.1 4.3 0.0a 6.3 12.3 8.8 6.8 21.7 16.6 13.7 11.5 16.4 11.0 16.0

6.3 7.6 4.9 7.4 5.7 2.0 0.6 6.1 22.3 19.4 17.4 5.1 7.4 11.3 10.2 10.2 9.1 42.3

Note: Units are MP a1/2. The value 0.0 is v alid in nonpolar media and deri ves from a zero dipole moment; a progressi vely higher value in increasingly polar media is required because of induced dipoles10 (see also Chapter 1).

a

methylene dichloride, and carbon tetrachloride as being v ery good, in agreement with solubilityor-not experiments, whereas the χlit values indicate these are not good or at best, marginal in quality. The results of Equation 2.13 for the aliphatic h ydrocarbons are particularly in good agreement with χlit.

POLYSTYRENE The calculated and e xperimental χ12 values for polystyrene are gi ven in Table 2.5. The Ro v alue of 12.7 MPa1/2 is now much higher than the ideal value indicated in Table 2.1. The polymer molecular weight is thought to be reasonably high, b ut is unkno wn. As a consequence of the Ro v alue, practically all the χ12 values calculated by Equation 2.13 are too high. One is tempted to di vide by a factor of 2 or 3, b ut there is no consistent pattern. Equation 2.14 includes the boundary v alue of χc equal to 0.5, so the results are more in agreement with χlit . HSP predicts that the aromatic and cyclic solvents are somewhat better than that e xpected from χlit. The agreement would be better if the χ12 values obtained from Equation 2.14 for these were increased by a factor 2. The values found by Equation 2.14 for the aliphatic h ydrocarbons are also lo wer than χlit, b ut are qualitati vely in agreement. The χ12 values for k etones and esters, using Equation 2.14, are in generally good agreement with the literature v alues. An exception of some note is the well-kno wn good solv ent cyclohexanone that is predicted as a much better solvent by HSP than the χlit value would indicate. There is considerably more differentiation in predictions of solvent quality found by Equation 2.14 than values indicated by χlit.

7248_C002.fm Page 39 Wednesday, May 9, 2007 8:25 AM

Theory

39

POLYVINYLACETATE The calculated and experimental χ12 values for polyvinylacetate are given in Table 2.6. The molecular weight for this polymer is reported as being 260,000. It can initially be noted that Ro is 13.7 MPa1/2, which ag ain means χ12 values found from Equation 2.13 will be higher than those found in the literature. This difference varies considerably, but a factor of 2 to 4 is generally required to give reasonable agreement. Equation 2.13 is certainly not generally acceptable as an instrument to predict χlit. Equation 2.14 gi ves reasonably good approximations to χlit as long as the solv ents are good enough to dissolv e the polymer; ho wever, there are some major disagreements. Tetralin dissolves the polymer, but has χlit equal to 1.3. n-Propanol is an error from the HSP prediction of it being a mar ginal solv ent, whereas it is a nonsolv ent. Alcohols of higher and lo wer molecular weight do have a significant e fect on this polymer, however, and the azeotropic mixture of ethanol and water actually dissolv es it. 3,6 When dealing with nonsolv ents, the HSP predictions of χ12 are generally lower than the data found in the literature. Once ag ain, a f actor of 3 to 4 is required to bring the values into agreement.

POLYACRYLONITRILE The calculated and experimental χ12 values for polyacrylonitrile are given in Table 2.7. This polymer has high polar and h ydrogen-bonding parameters and Ro equal to 10.9 MP a1/2 which, once more, is somewhat above the ideal range. The agreement with Equation 2.14 is reasonably good for the good solvents. The nonsolvents are not in good agreement. Equation 2.13 agrees surprisingly well with the best solv ents, γ-butyrolactone and dimeth yl sulfoxide, b ut the agreement is not uniform when all the solv ents are considered.

GENERAL DISCUSSION It should be noted in general that χ12 can either increase or decrease with concentration of the polymer. Barton 32 presents data to e xamine the potential magnitude of this ef fect. The correlations given in Table 2.2 are based on whether or not the polymer dissolv es at a concentration of 10%, with the exception of the data for polyacrylonitrile where no polymer concentration is indicated in the original solubility data. 33 The HSP data for correlations of the type gi ven in Table 2.2 can also be expected to change for higher polymer concentration and molecular weight. R M is e xpected to decrease only slightly for mar ginally higher polymer molecular weight, while considering a reasonably high molecular weight, and R M is e xpected to decrease some what for higher polymer concentration, although this can vary, especially for lower-molecular-weight species. An interesting fact to keep in mind is that a polymer with molecular weights (in millions) will only swell in e ven the best solvents. The present evaluations are at the same polymer concentration unless otherwise noted. No significant corrections of the type included in Equation 2.16 are required, as the polyme molecular weight is v ery high in all cases. Therefore, corrections of this type are not responsible for the differences in the calculated and observ ed χ12 parameters. A point of some concern is that ne gative values for χ12 are found in the literature, but these are not allowed in either the CST or HSP approaches. There is no obvious general explanation for this situation. A negative χ12 implies a solvent of a quality superior to anything that normal polymer–liquid interactions could pro vide. Normal here also includes the specific interactions attributable to permanent dipole–permanent dipole and h ydrogen bonding interactions, as discussed earlier . A closer review of this situation is desired. No systems with ne gative χ12 are included in Table 2.2 to Table 2.7. An additional problem of some concern is that, in general, there is considerable scatter in the χ12 parameter data from different sources. Clearing up this situation is f ar beyond the scope of this book. However, one cannot help but wonder why, and the seeming discrepancies do not contrib ute

7248_C002.fm Page 40 Wednesday, May 9, 2007 8:25 AM

40

Hansen Solubility Parameters: A User’s Handbook

to blind confidence in a y of the reported χ12 values. Indicative χ12 values are used here. One can also find ariations in HSP values for polymers from different sources,32 so there are also problems in determining which v alues are best in this approach. The χ12 parameter does not specificall account for permanent dipole–permanent dipole or h ydrogen bonding interactions, which must be considered a major source of potential dif ferences. There has been some discussion as to whether the coef ficient 4 in Equation 2.5 (correspondin to a coefficient of 0.25 in Equation 2.12) should be a di ferent number. Barton cites a case where a coefficient of 0.2 (rather than 0.25) in Equation 2.12 as determined. 34 Skaarup has mentioned a case of 5 (rather than 4) as a v alue for the coef ficient in Equation 2.5, which, of course, g ves 0.2 in Equation 2.12. 35 The author has also e xplored situations in volving water, where the D ATA FIT was equal for either a constant 4 or 5 in Equation 2.5. Zellers and co workers used this coefficien as an adjustable parameter for individual solvents in their studies. 28–31 One significant actor in this discussion is that solv ents with higher solubility parameters generally ha ve lower molecular v olumes. This means they will be better than e xpected by average comparisons of behavior. This fact tends to lead one to stretch the spheres a little more in the δP and δH directions to encompass these good solvents that would otherwise lie outside of the spheroids. This would lead to a number that is slightly higher than the 4 in Equation 2.5, and it is the author’ s feeling that this could be the case whenever a complete understanding of the ef fect of solvent molecular volume, and other size effects is accomplished. The use of the coef ficient 4 is also confirmed in Chapter The Prigogine theory contains structural parameters that have not been explored in this context. There are also structural parameters in the New Flory theory. It is possible that the use of structural parameters will allo w better understanding, enhance the possibility of impro ved calculations, and reduce the need for e xperimental studies. The experimentally-determined radius for the solubility spheres automatically tak es these f actors into account, b ut reliable calculation of the radius of interaction has not been possible as yet.

POSTSCRIPT The author has al ways e xperienced consistenc y in the quality of the predictions using the HSP . Care is required to generate the necessary data, and there should al ways be a ree valuation of experimental data based on an initial correlation. The solv ent parameters ha ve been used with success for many years in industrial practice to predict solv ent quality using computer techniques by most, if not all, major solvent suppliers. Mixing rules have been established for even complicated solvent blends. These are usually based on summing up simple volume fraction times the solubility parameter values. (An e valuation of the quality of the χ12 values in the literature could be made with precipitation e xperiments for mixtures to see whether a mixing rule gi ves consistent results for these as well.) The solvent δD, δP, and δH values that were established with e xtensive calculations ha ve been supported by tens of thousands of experimental data points based on solubility, permeation, surface wetting, etc. 10 It has become quite clear that the HSP for the solv ents are not precise enough for sophisticated calculations, but they certainly represent a good and satisf actory means for practical applications. The HSP for the solv ents relati ve to each other are correct for the majority of the common solvents. The “nearest neighbors” to a good solv ent are clearly e xpected to be of nearly comparable quality unless they are in a boundary re gion of the HSP solubility sphere. The solvent quality indicated by the ratio Ra/Ro (the RED number) has been particularly satisfactory. This ratio was defined years ago as a ratio of solubility parameters, as plotting and interpretation of data use solubility parameters. Use of the ratio of cohesion ener gy densities is also possible, of course, as this is indeed closer to an energy difference number and would agree more with the Flory approach as seen in Equation 2.6 and Equation 2.14, as H is really nothing other than (RED) 2.

7248_C002.fm Page 41 Wednesday, May 9, 2007 8:25 AM

Theory

41

As a result of ha ving reviewed all of this here, the author senses that the HSP approach is a practical e xtension in complete agreement with the Prigogine–Flory theory when the geometric mean is used, at least as f ar as the major f actors discussed earlier are concerned. The comparisons presented previously confirm some relation, ut the single χ12 parameter may ha ve been o versimplified, such that the more complete HSP approach cannot be immediately recognized. The ability of HSP to describe molecular affinities among so ma y different materials listed in this book speaks for the general application of both the Prigogine and the HSP treatments. The Prigogine treatment is acknowledged as difficult to use in practice. This is not true of the HSP approach.

CONCLUSION The Prigogine/Patterson CST and the HSP approach (which also in volves a corresponding states calculation) are shown to have very close resemblance. Both can be used to estimate the Flory χ12. Two equations involving HSP are given for this purpose. Reasonably good predictions are possible under f avorable circumstances. F avorable circumstances in volve a system with an essentially nonpolar polymer whose Ro v alue is not too dif ferent from 7.0 MP a1/2. χ12values for the better solvents are calculated by HSP at lo wer le vels than those found in the literature. χ12 values for nonsolvents are also generally calculated by HSP as being significantly l wer than the reported literature values. The most f avorable circumstances are, of course, not al ways present, and some problems still e xist and need to be solv ed before these calculations can be used with confidenc to estimate χ12 values for any solvent–polymer system. The HSP values for the polymers used for the present comparisons are based on solubility-or-not type experiments which reflect the propertie of the polymer as a whole. These may not completely correspond to the type of e valuation often used to find χ12, as less-than-dissolving amounts of solv ent may be used, and the solv ent may associate with given segments or groups in the polymer and not reflect the beh vior of the polymer as a whole (see also the discussion in Chapter 5). An empirical f actor, β, equal to about 0.34 appears in man y sources in the literature in connection with calculation of χ12 using Hildebrand solubility parameters. β disappears when HSP are used for this purpose, b ut the resulting equation has not been studied enough to allo w general use of HSP to calculate χ12 parameters. Studies on the ef fect of molecular size, se gmental size, and polymer size are still required. It is suggested that the structural f actors discussed by Prigogine be tried in this respect. 11 Use of the geometric mean in conjunction with the Prigogine theory brings the HSP and Prigogine approaches into agreement. The massive amount of experimental data presented in this book strongly supports the use of the geometric mean. As a curiosity, it might be noted that the use of the geometric mean (Lorenz–Berthelot mixtures) generated an ellipsoidal miscibility plot essentially identical to those given in Chapter 5, Figure 5.1 and Figure 5.2. 36 This approach w as not continued because it w as stated that “the boundary of this ellipse is of little practical importance as there are no known cases of immiscibility in mixtures kno wn to conform to the Lorenz–Berthelot equations. ” As stated in the Pref ace to this book, it has not been its purpose to recite the de velopments of polymer solution thermodynamics in a historical manner with full e xplanations of each theory or modifications thereof. The references cited in the Pref ace do this already . Chapter 3 and Chapter 4 have been added to this edition of this handbook to gi ve broader co verage in this respect. This chapter has attempted to sho w relations between the classical theories of polymer solution ther modynamics and the HSP approach, which includes a quantitati ve accounting of both permanent dipole–permanent dipole and h ydrogen bonding interactions as an inte gral part. The relation between the Prigogine–P atterson theory and HSP w as the most ob vious.

7248_C002.fm Page 42 Wednesday, May 9, 2007 8:25 AM

42

Hansen Solubility Parameters: A User’s Handbook

REFERENCES 1. Flory, P.J., Principles of Polymer Chemistry, Cornell University Press, New York, 1953. 2. Eichinger, B.E. and Flory , P.J., Thermodynamics of polymer solutions, Trans. Faraday Soc., 64(1), 2035–2052, 1968; Trans. Faraday Soc. (2), 2053–2060; Trans. Faraday Soc. (3), 2061–2065; Trans. Faraday Soc. (4), 2066–2072. 3. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities I Solvents, plasticizers, polymers, and resins, J. Paint Technol., 39(505), 104–117, 1967. 4. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities II Dyes, emulsifiers, mutual solubility and compatibilit , and pigments, J. Paint Technol., 39(511), 505–510, 1967. 5. Hansen, C.M. and Skaarup, K., The three dimensional solubility parameter — key to paint component affinities III. Independent calculation of the parameter components J. Paint Technol., 39(511), 511–514, 1967. 6. Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Their Importance in Surf ace Coating F ormulation, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967. 7. Hansen, C.M., The universality of the solubility parameter , Ind. Eng. Chem. Prod. Res. Dev., 8(1), 2–11, 1969. 8. Hansen, C.M. and Beerbo wer, A., Solubility parameters, in Kirk-Othmer Encyclopedia of Chemical Technology, Suppl. Vol., 2nd ed., Standen, A., Ed., Interscience, Ne w York, 1971, pp. 889–910. 9. Hansen, C.M., 25 Years with solubility parameters (25 År med Opløselighedsparametrene, in Danish), Dan. Kemi, 73(8), 18–22, 1992. 10. Hansen, C.M., Solubility parameters, inPaint Testing Manual, Manual 17, Koleske, J.V., Ed., American Society for Testing and Materials, Philadelphia, 1995, pp. 383–404. 11. Prigogine, I. (with the collaboration of Bellemans, A. and Mathot, A.), The Molecular Theory of Solutions, North-Holland, Amsterdam, 1957, chap. 16, 17. 12. Patterson, D. and Delmas, G., Ne w aspects of polymer solution thermodynamics, Off. Dig. Fed. Soc. Paint Technol., 34(450), 677–692, 1962. 13. Delmas, D., Patterson, D., and Somcynsky, T., Thermodynamics of polyisobutylene-n-alkane systems, J. Polym. Sci., 57, 79–98, 1962. 14. Bhattacharyya, S.N., Patterson, D., and Somcynsky, T., The principle of corresponding states and the excess functions of n-alkane mixtures, Physica, 30, 1276–1292, 1964. 15. Patterson, D., Role of free v olume changes in polymer solution thermodynamics , J. Polym. Sci. Part C, 16, 3379–3389, 1968. 16. Patterson, D.D., Introduction to thermodynamics of polymer solubility , J. Paint Technol., 41(536), 489–493, 1969. 17. Biros, J., Zeman, L., and P atterson, D., Prediction of the χ parameter by the solubility parameter and corresponding states theories, Macromolecules, 4(1), 30–35, 1971. 18. Hildebrand, J. and Scott, R.L., The Solubility of Nonelectrolytes, 3rd ed., Reinhold, Ne w York, 1950. 19. Hildebrand, J. and Scott, R.L., Regular Solutions, Prentice-Hall, Englewood Cliffs, NJ, 1962. 20. Hansen, C.M., Cohesion parameters for surf aces, pigments, and fillers ( ohæsionsparametre for Overflade , Pigmenter, og Fyldstoffer, in Danish), Färg och Lack Scand., 43(1), 5–10, 1997. 21. Hansen, C.M., Polymer solubility — prigogine theory and Hansen Solubility parameter theory mutually confirmed (Polymeropløselighed — Prigogine Teori og Hansen Opløselighedsparameterteori Gensidigt Bekræftet, in Danish), Dan. Kemi, 78(9), 4–6, 1997. 22. Hildebrand, J. and Scott, R.L., The Solubility of Nonelectrolytes, 3rd ed., Reinhold, Ne w York, 1950, chap. 20. 23. Hansen, C.M., Characterization of surfaces by spreading liquids, J. Paint Technol., 42(550), 660–664, 1970. 24. Hansen, C.M., Surface dewetting and coatings performance, J. Paint Technol., 44(570), 57–60, 1972. 25. Hansen, C.M. and Pierce, P .E., Surface effects in coatings processes, XII Federation d’Associations de Techniciens des Industries des Peintures, Vernis, Emaux et Encres d’Imprimerie de l’Europe Continentale, Congress Book, Verlag Chemie, Weinheim/Bergstrasse, 1974, 91–99; Ind. Eng. Chem., Prod. Res. Dev., 13(4), 218–225, 1974.

7248_C002.fm Page 43 Wednesday, May 9, 2007 8:25 AM

Theory

43

26. Hansen, C.M. and Wallström, E., On the use of cohesion parameters to characterize surfaces, J. Adhes., 15(3/4), 275–286, 1983. 27. Patterson, D., personal communication, 1997. 28. Zellers, E.T., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. I. Modeling the solubility of organic solvents in Viton® gloves, J. Appl. Polym. Sci., 50, 513–530, 1993. 29. Zellers, E.T . and Zhang G.-Z., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. II. Modeling diffusion coefficients, breakthrough times, and steady-state perme ation rates of or ganic solvents in Viton® gloves, J. Appl. Polym. Sci., 50, 531–540, 1993. 30. Zellers, E.T., Anna, D.H., Sulewski, R., and Wei, X., Critical analysis of the graphical determination of Hansen’s solubility parameters for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2069–2080, 1996. 31. Zellers, E.T., Anna, D.H., Sule wski, R., and Wei, X., Impro ved methods for the determination of Hansen’s solubility parameters and the estimation of solv ent uptake for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2081–2096, 1996. 32. Barton, A.F.M., Handbook of Polymer-Liquid Interaction Parameters and Solubility Parameters, CRC Press, Boca Raton, FL, 1990. 33. Brandrup, J. and Immer gut, E.H., Eds., Polymer Handbook, 3rd ed., Wiley-Interscience, New York, 1989. (a) Gundert, F . and Wolf, B.A., Polymer -solvent interaction parameters, pp. VII/173–182. (b) Fuchs, O., Solv ents and non-solv ents for polymers, pp. VII/379–407. 34. Barton, A.F.M., Applications of solubility parameters and other cohesion ener gy parameters, Polym. Sci. Technol. Pure Appl. Chem., 57(7), 905–912, 1985. 35. Skaarup, K., pri vate communication, 1997. 36. Rowlinson, J.S., Liquids and Liquid Mixtures, Butterw orths Scientific Publications, London, 1959 chap. 9.

7248_C002.fm Page 44 Wednesday, May 9, 2007 8:25 AM

7248_book.fm Page 45 Tuesday, April 24, 2007 9:19 AM

Thermodynamic 3 Statistical Calculations of the Hydrogen Bonding, Dipolar, and Dispersion Solubility Parameters Costas Panayiotou KEY WORDS Statistical thermodynamics, cohesive energy density, Hansen solubility parameters, solvent screening.

ABSTRACT The main objecti ve of this chapter is the presentation of an equation-of-state frame work for the calculation of the h ydrogen-bonding component of the solubility parameter as well as the other partial solubility parameters. A ne w statistical thermodynamic approach has been de veloped for the estimation of these partial components o ver a broad range of temperature and pressure. K ey to this approach is the development of explicit expressions for the contribution of hydrogen bonding, dispersion, and dipolar interactions to the potential ener gy of the fluid. The approach is applicable to ordinary solvents, supercritical fluids, and high polymers. Information on various thermodynamic properties of fluids is used in order to estimate the three solubility parameter components. Extensive tables with the k ey parameters are presented. When information on h ydrogen bonding interaction is a vailable from other sources, the proposed method is essentially a predicti ve method for the hydrogen bonding component of the solubility parameter. On the other hand, available information on these separate components is exploited for extracting information on the thermodynamic behavior of the fluids o ver an e xtended range of e xternal conditions.

INTRODUCTION The conceptual simplicity of the solubility parameter , δ, makes it most attracti ve in industry and 1 it remains today one of the k in academia as well. Originally introduced by Hildebrand, ey parameters for selecting solv ents in industry , characterizing surf aces, predicting solubility and degree of rubber swelling, polymer compatibility , chemical resistance, permeation rates, and for numerous other applications. There is also much interest in utilizing solubility parameters for rationally designing new processes, such as the supercritical fluid, the coating, and the drug delivery processes.2–8 Of course, the use of solubility parameter, or the closely related cohesive density is not always successful and this lack of total success stimulates continuing research.The central principle behind the use of δ is the historic alchemist maxim, similia similibus solvuntur (“lik e dissolv es lik e”), 45

7248_book.fm Page 46 Tuesday, April 24, 2007 9:19 AM

46

Hansen Solubility Parameters: A User’s Handbook

probably the oldest rule of solubility. This rule can, indeed, be a good guide in the study of solubility , as long as it enables definition of the degree of likeness in the given system with sufficient precision. This need for precision in the definition of likeness led to the division of δ into its partial components or Hansen solubility parameters 5 δd, δp, and δhb, for the dispersion, the polar , and the h ydrogen bonding contributions, respectively. Thus, liquids with similar δd, δp, and δhb are very likely to be miscible. The bulk of the developments in solubility parameter reside on the principle of “similarity matching” of properties. As it is recognized, ho wever, that a more appropriate principle w ould be the “complementary matching” of properties, 9 the h ydrogen bonding component, δhb, is further subdivided into an acidic component, δa, and a basic component, δb, in order to account for the Lewis acid and Le wis-base character of the substance. 8–10 Over the years, the partial solubility parameters were determined for an enormous number of substances and led to critical compilations a vailable in the open literature. 3–5 These compilations are most v aluable sources of information for the nature of the substances and their interactions with other substances. Starting from the original definition of cohesi ve energy density and solubility parameter , we have already proposed a systematic approach for estimating the solubility parameter o ver an extended range of temperature and pressure. 11–12 In this w ork, it became clear that the h ydrogen bonding contribution could be calculated rather accurately from the h ydrogen bonding part of the potential energy E and the v olume V of the system as obtained, for e xample, from the lattice-fluid hydrogen bonding (LFHB) equation-of-state model. 13 The model, however, could not separate the dispersion and the polar components of the solubility parameters. The proposition w as made to calculate δd from the solubility parameter of the corresponding homomorph hydrocarbon. Although this proposal could be v alid for some classes of fluids, it could not be generalized. In a recent publication,14 we have proposed a group contribution method for the estimation of the total solubility parameter of a large variety of substances. The very same method could be used for the estimation of δd, as sho wn later. Knowledge of the h ydrogen bonding and the dispersion components of the solubility parameter could lead to an estimation of the polar component as well. This chapter is, ho wever, heavily based on a more recent publication 15 in which the pre vious approach11–12 w as e xtended in an ef fort to account for all three components of the solubility parameter. This w as done by adopting the more recent and more accurate NRHB (nonrandom hydrogen-bonding) equation-of-state frame work,16 which w as modified in order to e xplicitly account for dipole–dipole interactions and, thus, e xplicitly calculate the polar component, δp.

THEORY THE EQUATION-OF-STATE FRAMEWORK Let us consider a system of N molecules of a fluid at temperature T, external pressure P, and of volume V, which are assumed to be arranged on a quasi-lattice that has a coordination number z, number of sites Nr , N0 and that denotes empty sites. Each molecule is assumed to be di vided in r segments of se gmental volumes v*, and to ha ve zq = zrs external contacts, s being its surf ace-tovolume ratio, a geometric characteristic of the molecule. The total number ( Nr) of lattice sites is given by: Nr = rN + N0

(3.1)

Following previous practice,15,16 one may write for the configurational partition function of the fluid in the N,P,T ensemble and in its maximum term approximation: Q ( N , T , P ) = QR QNR Qhb = Ω R Ω NR Ω hb exp

−E p − Ed − E hb − PV exp exp exp kT kT kT kT

(3.2)

7248_book.fm Page 47 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

47

Ed, Ep, and Ehb in Equation 3.2 are the dispersion, polar , and h ydrogen bonding components, respectively, of the potential ener gy of the system. The detailed rationale behind the form of the combinatorial term, ΩR, its correction factors for nonrandom distribution of free volume, ΩNR, and for the h ydrogen bonding, Ωhb, can be found in the pre vious work.16 Here, the final equations are simply reproduced, namely: N r ! N rl N ⎛ N q ! ⎞ ΩR = ω N 0 ! N ! ⎜⎝ N r ! ⎟⎠

z

2

N

(3.3)

where l=

(

) (

)

z r −q − r −1 2

(3.4)

whereas the total number of intermolecular contacts in the system is gi ven by: (3.5)

zNq = zqN + zN0

In Equation 3.3, ω is a characteristic quantity for each fluid that takes into account the flexibility and the symmetry of the molecule, and this quantity cancels out in all applications of interest here. In the following, we will need the site fractions f0 and f for the empty sites and the molecular segments, respectively. The relation is gi ven by: f0 =

N 0 N r − rN = = 1− f Nr Nr

(3.6)

For the second f actor, ΩR, we may use v arious expressions available in the open literature. The most classical is Guggenheim’ s quasi-chemical e xpression17:

16

2

⎡⎛ N 0 ⎞ ⎤ N ! N ! ⎢ ⎜ r 0 ⎟ !⎥ ⎢⎣⎝ 2 ⎠ ⎥⎦ = 2 ⎡⎛ N r 0 ⎞ ⎤ N rr ! N 00 ! ⎢⎜ ⎟ !⎥ ⎢⎣⎝ 2 ⎠ ⎥⎦ 0 rr

Ω NR

0 00

(3.7)

where N rr is the number of e xternal contacts between the se gments belonging to molecules; N 00 is the number of contacts between the empty sites; and N r0 is the number of contacts between a molecular segment and an empty site. The superscript 0 refers to the case of randomly distrib uted empty sites. In this random case, we ha ve: N rr0 =

1 qN z zqN = qN θr 2 N 0 + qN 2

(3.8)

7248_book.fm Page 48 Tuesday, April 24, 2007 9:19 AM

48

Hansen Solubility Parameters: A User’s Handbook

0 N 00 =

1 N z N 0 z 0 = N 0 θ0 2 Nq 2

(3.9)

and N r00 = zqN

N0 qN = zN 0 = zqN θ0 = zN 0 θr Nq Nq

(3.10)

where θr = 1 − θ0 =

q r q r + v −1

(3.11)

and the reduced v olume is defined as: v=

V rNvv ∗ 1 = = ∗ ρ V rNv ∗

(3.12)

ρ˜ being the reduced density . The corresponding number of interse gmental contacts (N ij) in the nonrandom case are gi ven by the following equations: N rr = N rr0 Γ rr =

z qN θr Γ rr 2

0 N 00 = N 00 Γ 00

(3.13)

N r 0 = N r00 Γ r 0 The nonrandom f actors, Γ, in these equations are equal to unity in the random case. numbers must satisfy the follo wing material balance equations: 16–18 2 N 00 + N 0 r = zN 0

These

(3.14)

2 N rr + N 0 r = zqN By combining these equations, we obtain: θ0 Γ 00 + θr Γ r 0 = 1

(3.15)

θr Γ rr + θ0 Γ r 0 = 1 These two equations along with the quasi-chemical condition:

16–18

7248_book.fm Page 49 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

49

⎛ 2 * ε* / z ⎞ 4 N rr N 00 Γ rr Γ 00 = = exp ⎜⎝ RT ⎟⎠ = A N r20 Γ r20

(3.16)

form a system of three equations from which one may obtain the f actor Γ. The reduced density needed in Equations 3.15, is obtained from the equation of state (cf. Equation 3.22 sho wn later). The intersegmental interaction energy, ε* = z ε/2, in Equation 3.16 is related to the scaling temper ature, T*, and scaling pressure, P*, of the fluid by: ε∗ = RT ∗ = P ∗ v ∗

(3.17)

whereas the reduced temperature and pressure are defined as: T=

T P , P= * * T P

(3.18)

Combining Equation 3.15 and Equation 3.16, one can obtain a quadratic equation for Γr0, with the physically meaningful solution: Γr0 =

2

(

)

1 + ⎡⎣1 − 4θ0 θr 1 − A ⎤⎦

1

(3.19) 2

Most general e xpressions for Ωhb may be found in the original w ork.13 In the case of a fluid with d proton donors and a proton acceptors forming NH hydrogen bonds, one has:

Ω hb =

⎡⎣ N !⎤⎦

(

2

NH ! ⎡ N − NH ⎣

⎛ ρ −S ⎞ exp H ⎟ 2 ⎜ R ⎠ ⎝ rN !⎤ ⎦

NH

)

(3.20)

where SH is the entrop y change upon h ydrogen bond formation and the NH is given by: B+d +a− N νH = H = rN

(B + d + a)

2

− 4ad

2r

(3.21)

With these definitions, the equation of state of the fluid is gi ven by: ⎡ ⎤ ⎛l ⎞ z ⎛ q ⎞ z P + T ⎢ ln 1 − ρ − ρ ⎜ − ν H ⎟ − ln ⎜ 1 − ρ + ρ⎟ + ln Γ 00 ⎥ = 0 r ⎠ 2 ⎝r ⎠ 2 ⎝ ⎣ ⎦

(

)

(3.22)

and the chemical potential by: μ dp μ H μ = + RT RT RT

(3.23)

7248_book.fm Page 50 Tuesday, April 24, 2007 9:19 AM

50

Hansen Solubility Parameters: A User’s Handbook

where μ dp RT

= ln

⎡ q P 1 z q ⎤ zq ⎡⎣ ln Γ rr ⎤⎦ − + r − ρl + ln ρ − q ln ⎢1 − ρ + ρ⎥ + ωr 2 r ⎦ 2 T T ⎣

(3.24)

is the chemical potential for the dispersion and polar interactions, and μH d a = r ν H − d ln − a ln RT d − νH a − νH

(3.25)

is the chemical potential for the h ydrogen bonding interactions. The heat of v aporization is given by: ⎡⎛ ⎛ ⎞ ⎤ ⎞ q q HV = rN ε∗ ⎢⎜ Pv − θr Γ rr ⎟ − ⎜ Pv − θr Γ rr ⎟ ⎥ + E H ⎡⎢ N H ⎣ r r ⎠ liq ⎥ ⎠ vap ⎝ ⎢⎣⎝ ⎦

( )

liq

( )

− NH

vap

⎤ ⎥⎦

(3.26)

Equation 3.15, Equation 3.19, and Equation 3.22 are coupled equations and must be solv ed simultaneously for the reduced density and the nonrandom f actors. In the above formalism, the contributions from dispersion and polar forces are lumped into one contribution. An attempt will be made in the ne xt section to separate them.

THE CONTRIBUTION

FROM

DIPOLAR FORCES

In an initial attempt, the contrib ution of dipole–dipole interactions w as approximated by the multipolar u-expansion of Twu and Gubbins19 by keeping the leading term of the point dipole–point dipole interaction and the P ade approximations, 20–21 as well as by using the perturbation model of Nezbeda and Pavlicek.22–24 In an o verwhelming majority of cases, this procedure led to underestimations of δp that often f all in the range of one to tw o orders of magnitude belo w the e xpected value. Thus, a drastically dif ferent approach w as adopted that preserv es the simplicity of the formalism of the pre vious paragraph. In the previous paragraph, as in NRHB 16, it was assumed that only first neighbor segment–segment interaction contacts contrib ute to the potential ener gy (E) of the system and, thus, we may write for the non-h ydrogen-bonding part (dispersion and polar): − E dp = N rr ε dp = Γ rr qN θr ε∗dp

(3.27)

Obviously, in the absence of dipolar interactions we should ha ve εdp = ε and εdp* = ε*, that is, only the contrib ution from dispersion forces. We may then write quite generally:

(

)

ε∗dp = ε∗ ⎡1 + f m, T , ρ, r,... ⎤ ⎣ ⎦

(3.28)

The crux of the problem is the e xplicit form of the function f in Equation 3.28. This function should be zero in the absence of dipole–dipole interactions or when the dipole moment, m, of the fluid is zero. As, however, in Equation 3.27 we count se gmental interactions, the function f might be approximated by:

7248_book.fm Page 51 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

51

b

⎛m⎞ f m, T , ρ, r,... = ⎜ ⎟ g m, T , ρ, r,... ⎝r⎠

(

)

(

)

(3.29)

Hansen5, summarizing years of e xperience, has observ ed that δp is directly proportional to m, which implies that b in Equation 3.29 could be set equal to 2. His additional observ ation that δp is in versely proportional to V1/2 could be reconciled with Equation 3.27 and Equation 3.29 by writing: 2

⎛m⎞ c f m, T , ρ, r,... = ⎜ ⎟ ⎝ r ⎠ Γ rr θr

(

)

(3.30)

c being a constant. The second attempt for the estimation of δp w as made by using Equation 3.30 along with Equation 3.27 and Equation 3.28. This approach with the constant c replaced by the e xpression: c = πs 4

(3.31)

could, indeed, lead to a simultaneously satisf actory estimation of δp and a satisf actory description of the thermodynamic beha vior of the fluids. Ho wever, it turned out that the follo wing simpler form of Equation 3.30: 2

⎛m⎞ f = ⎜ ⎟ πs 2 ⎝r⎠

(3.32)

led to better results and pro vided both a satisfactory description of the thermodynamic behavior of fluids over a broad range of e xternal conditions and a satisf actory estimation of δp and the other partial solubility parameters for the o verwhelming majority of fluids. One additional reason for adopting Equation 3.32 is that the only change that should be made in the formalism of the previous 2 ⎡ ⎤ ⎛m⎞ paragraph is to replace ε* by ε∗ ⎢1 + π ⎜ ⎟ s 2 ⎥ . ⎝r⎠ ⎢ ⎥ ⎣ ⎦ Thus, the final form of the potential ener gy that w as adopted is:

2 ⎡ ⎛m⎞ 2⎤ ⎢ − E = Γ rr qN θr ε 1 + π ⎜ ⎟ s ⎥ − N H E H ⎝r⎠ ⎢ ⎥ ⎣ ⎦ ∗

(3.33)

On the basis of the abo ve equation, the partial solubility parameters are gi ven by:

δd =

Γ rr qN θr ε∗ V

(3.34)

7248_book.fm Page 52 Tuesday, April 24, 2007 9:19 AM

52

Hansen Solubility Parameters: A User’s Handbook

δp =

⎡ ⎛ m ⎞2 ⎤ Γ rr qN θr ε∗ ⎢ π ⎜ ⎟ s 2 ⎥ ⎢ ⎝r⎠ ⎥ ⎣ ⎦ V

δ hb =

− N H EH V

(3.35)

(3.36)

where, the total v olume, V, of the system is gi ven by: V = rNvv ∗ + N H VH

(3.37)

VH being the v olume change upon h ydrogen bond formation. An alternati ve e xpression for the factor f in Equation 3.32, which tak es explicitly into consideration its dependence on temperature, is given in Appendix 3.II.

APPLICATIONS In this section, we will apply the model presented in the pre vious two sections in a multitude of cases. As a first step, we will describe the v apor pressure, the orthobaric densities, and the heat of vaporization of fluids by determining their three scaling constants through a least-squares fit. 16,25–26 These constants are reported in Table 3.1a for a number of common fluids. The critical compilation (Design Institute for Ph ysical Property Research [DIPPR]) 25 was used as a source for the thermodynamic data and the dipole moments of the studied fluids. As in NRHB, 16 the geometric constant s of each fluid w as obtained through the widely used group contrib ution calculation scheme of UNIFAC.27–28 Having determined the scaling constants through the previously mentioned procedure, this approach can estimate (essentially predict) the dispersion and the polar components of the solubility parameter o ver a broad range of temperature and pressure. In the case of hydrogen-bonded fluids, the energy, entropy, and volume change upon hydrogen bond formation are also needed. F or simplicity, the v olume change, VH, was set equal to zero for all fluids. In addition, for lack of reliable information pertinent to h ydrogen bonding o ver an extended range of temperature and/or pressure, the entrop y change, SH, w as set equal to –26.5 JK1mol1, as for alkanols. 13,16 This is a rather crude approximation b ut it permits a more direct comparison of the strength of the v arious types of h ydrogen bonds through the mere comparison of the energy change, EH. Essentially, the energy change, EH, which is the only adjustable parameter for the description of h ydrogen bonding in our approach, w as adjusted through the e xperimental value5 of δhb. As this is only one datum, it does not suf fice to reliably determine SH as well. The parameter EH is also reported in Table 3.1a. In the overwhelming majority of cases, the number of hydrogen bonds in the system is obtained through Equation 3.21. This equation, ho wever, cannot be used in the case of carboxylic acids where the main mode of h ydrogen bonding is dimerization. The case of carboxylic acids is treated separately, as shown in Appendix 3.I. In a similar manner , the scaling constants for high polymers were obtained by correlating the available extensive experimental pressure-volume-temperature (PVT) data 29 with the equation of state, Equation 3.22, and are reported in Table 3.1b. Table 3.2a and Table 3.2b compare the e xperimental5 solubility parameters with those estimated/predicted by our approach for a number of common fluids. As observed and in vie w of the

7248_book.fm Page 53 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

53

TABLE 3.1A Characteristic Constants of Pure Fluidsa Fluid

ε* = RT*/J.mol1

ν* = ε*P*–1/cm3 .mol–1

ν*sp = ρ*–1/ cm3 .g–1

m/ Debye

–EH/ J.mol–1

s = q/r s

Propane n-Butane n-Pentane n-Hexane n-Heptane n-Octane n-Nonane n-Decane n-Undecane n-Dodecane 3-Methyl pentane 2,4-Dimethyl hexane 2,2,4-Trimethyl pentane Cyclohexane

3319 3800 4295 4557 4734 4870 4999 5111 5212 5265 4514. 4821 4794 5171

Nonpolar Fluids 9.121 10.748 13.15 13.57 14.00 14.45 14.925 15.317 15.577 15.637 13.750 15.916 17.142 13.040

1.426 1.392 1.373 1.317 1.290 1.283 1.278 1.266 1.259 1.252 1.313 1.301 1.297 1.205

0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0

0.903 0.881 0.867 0.857 0.850 0.844 0.839 0.836 0.833 0.830 0.856 0.843 0.857 0.800

Benzene Toluene Styrene

4986 5247 5656 5639 5403 5845 5664 3197 5295 3888 4406 3992 4383 4226 3840 4460 1811 4827 4650 3859 5304 2632 3267 3787 3740 4130 4376 4683 4977 5294 6199 4824

0.370 0.360 0.130 0.330c

2720 3585 6421 6444 5100 5080 5120 8250 6274 11133 8705 6390 7384 6780 7547 6570 5860 8565 5116 4412 4275 25100 25100 24771b 25100 25100 25100 25100 25100 23000 22940 21130

0.753 0.757 0.760 0.760 0.759 0.720 0.720 0.908 0.769 0.896 0.869 0.894 0.903 0.818 0.888 1.030 0.909 0.840 0.881 0.800 0.746 0.941 0.903 0.881 0.881 0.867 0.857 0.850 0.839 0.833 0.757 0.806

o-Xylene Tetralin Acetone Acetophenone Ethyl acetate n-Butyl acetate Vinyl acetate Methyl methacrylate -Caprolactone Diethyl ether 1,4-Dioxane Carbon dioxide Chloroform Dichloromethane Vinyl chloride Chlorobenzene Methanol Ethanol 1-Propanol 1-Butanol 1-Pentanol 1-Hexanol 1-Octanol 1-Decanol Phenol Ethylene glycol

Polar/Hydrogen-Bonded Fluids 9.551 1.079 10.922 1.098 11.971 1.075 11.960 1.075 11.445 1.087 11.016 1.004 10.431 0.992 9.113 1.134 10.350 0.952 13.264 1.018 13.050 1.049 10.548 0.961 12.823 0.978 14.757 0.941 11.575 1.182 10.830 0.870 7.050 0.739 10.423 0.619 9.786 0.736 9.880 0.960 10.418 0.873 10.737 1.140 10.820 1.126 11.400 1.124 11.560 1.122 12.140 1.131 12.700 1.131 14.652 1.156 14.25 1.141 14.698 1.146 12.220 0.928 11.635 0.904

0.630 0.219 1.050c 2.887 3.028 1.780 1.841 1.790 1.670 4.437 1.151 0.400 2.320c 1.010 1.439 1.451 1.690 1.700 1.690 1.680 1.680 1.660 1.650 1.650 1.650 1.619 1.451 2.308

7248_book.fm Page 54 Tuesday, April 24, 2007 9:19 AM

54

Hansen Solubility Parameters: A User’s Handbook

TABLE 3.1A (CONTINUED) Characteristic Constants of Pure Fluidsa Fluid 1,2-Propylene glycol Glycerol Diethylamine n-Butylamine Tetrahydrofuran Formamide Dimethylformamide Acrylonitrile Dimethylsulfoxide Acetic acid Acrylic acid Propionic acid Butyric acid Methacrylic acid Octanoic acid Oleic acid Stearic acid Ammonia Water a b c

ε* = RT*/J.mol1

ν* = ε*P*–1/cm3 .mol–1

2325 2751 4377 4417 4025 4193 4610 3634 3065 4175 4703 4840 4676 4787 5028 5525 5166 6400 5610 1234 2676 4222

13.650 16.61 12.411 12.430 12.474 11.432 5.959 12.146 8.110 9.160 6.75 8.95 7.963 9.088 9.752 10.052 10.014 19.413 12.322 6.000 8.709 10.527

ν*sp = ρ*–1/ cm3 .g–1

m/ Debye

–EH/ J.mol–1

s = q/r s

0.909 0.798 1.265 1.229 1.036 1.025 0.899 0.762 1.175 0.908 0.896 0.903 0.939 0.963 0.935 0.938 1.036 1.108 1.041 1.341 0.992 0.991

3.627 4.197 0.920 1.391 1.631 1.360c 3.717 3.807 3.867 3.957 1.739 1.460 1.751 1.649 1.649 0.650c

22450 22520 12268 12325 8700 8770 16238 16120 9152 11038 20882 31600 24700 28880 21638 21553 26369 21090 19200 12277 17493 18100

0.866 0.767 0.861 0.874 0.925 0.925 0.869 0.855 0.887 0.855 0.910 0.876 0.902 0.888 0.922 0.922 0.850 0.823 0.824 1.039 0.861 0.861

1.700 1.739 1.670 2.750 1.850 0.970c

SH was set equal to – 26.5 JK–1mol–1 in all cases. Adjusted to fit δhb. Adjusted to fit δp.

TABLE 3.1B Characteristic Constants of Pure Fluids/Polymersa Fluid (Polymer)

ε* = RT*/J.mol1

ν* = ε*P*–1/cm3 mol1

ν*sp = ρ*–1/cm3 .g–1

Polyethylene-lin. Polypropylene Polystyrene Poly(vinyl chloride) Polyacrylonitrile Poly(methyl methacrylate) Polycarbonate (bisphenol A) Poly(ε-caprolactone) Poly(vinyl acetate) Nylon 66

5401 5993 6335 4876 5862 5294 5840 4817 5607 5029

13.169 13.348 11.692 8.472 9.614 11.569 11.915 10.976 13.121 9.169

1.130–0.000039P 1.147–0.000164P 0.710–0.000087P 0.676–0.000017P 0.839–0.000049P 0.821–0.000083P 0.806–0.000020P 0.870–0.000016P 0.810–0.000070P 0.898–0.000037P

a b

m/ Debye 0.080 0 1.298 1.451 1.810 2.600 4.300 1.750 0.450 10.00

Pressure, P, in column 4 is in MPa. Numbers in parentheses are proton-donors and proton-acceptors, respecti vely, per repeat unit.

–EH/ J.mol–1

s = q/r s

1350(4,2)b 400(6,2) 3600(8,1) 4400(3,1) 8700(3,1) 4800(8,2) 9550(6,3) 7800(10,2) 3700(6,2) 68000(2,2)

0.800 0.799 0.667 0.780 0.887 0.843 0.728 0.818 0.825 0.783

7248_book.fm Page 55 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

55

TABLE 3.2A Total and Partial Solubility Parameters (in MPa1/2) of Pure Fluids δ Total

δ HB

5

5

Fluid

Exp

Propane n-Butane n-Pentane n-Hexane n-Heptane n-Octane n-Nonane n-Decane n-Undecane n-Dodecane 3-Methyl pentane 2,4-Dimethyl hexane 2,2,4-Trimethyl pentane Cyclohexane

13.10 14.10 14.40 14.90 15.20 15.40 15.60 15.70 15.80 15.90 14.67 14.65 14.08 16.76

Benzene Toluene Styrene o-Xylene Tetralin Acetone Acetophenone Ethyl acetate n-Butyl acetate Vinyl acetate Methyl methacrylate ε-Caprolactone Diethyl ether 1,4-Dioxane Carbon dioxide Chloroform Dichloromethane Vinyl chloride Chlorobenzene Methanol Ethanol 1-Propanol 1-Butanol 1-Pentanol 1-Octanol 1-Decanol Phenol Ethylene glycol 1,2-Propylene glycol

Calc

Exp

Nonpolar Fluids 12.85 0 14.10 0 14.34 0 14.58 0 15.12 0 15.21 0 15.25 0 15.30 0 15.40 0 15.52 0 14.65 0 14.39 0 13.90 0 16.30 0

δP Exp 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Polar/Hydrogen-Bonded Fluids 18.41 18.27 2.05 2.05 18.32 17.78 2.00 2.00 19.07 18.29 4.10 4.04 19.07 18.30 4.10 4.05 18.20 18.01 3.10 3.10 19.80 18.80 2.90 2.90 19.80 19.07 2.90 2.90 19.95 20.04 6.95 7.00 21.73 20.48 3.68 3.68 18.48 18.31 9.20 9.33 17.59 17.43 6.30 6.30 18.58 18.33 5.90 5.90 17.92 17.77 5.40 5.40 21.41 21.28 7.40 7.40 15.66 15.73 5.11 5.12 20.47 20.08 7.36 7.36 14.56 11.64 4.10 4.09 18.94 19.18 5.73 5.74 20.79 19.92 4.09 4.09 17.77 16.00 2.40 2.40 19.61 18.91 2.05 2.05 29.61 29.89 22.30 24.08 26.13 26.08 19.43 19.98 24.45 24.19 17.40 17.41 24.45 24.17 17.40 17.58 23.35 22.90 15.80 15.80 21.65 21.95 13.91 14.52 20.87 20.27 11.86 11.94 20.32 19.37 10.00 10.03 24.63 24.69 14.90 14.95 33.70 33.64 25.77 25.74 29.52 29.19 23.32 23.73

5

Calc

0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0

1.02 1.40 1.00 1.00 1.00 2.00 2.00 10.43 8.59 5.85 3.70 7.20 6.50 15.0 2.86 1.84 6.90 3.07 7.36 6.50 4.30 12.27 8.80 6.80 6.80 5.70 4.50 3.27 2.60 5.90 11.05 9.41

1.01 0.92 0.33 1.00 1.48 0.43 2.00 10.14 7.04 6.07 4.70 5.90 5.60 13.22 3.48 1.91 6.86 3.79 5.63 5.07 4.33 11.34 8.24 6.79 6.79 5.72 4.99 3.75 3.15 5.16 12.23 12.56

7248_book.fm Page 56 Tuesday, April 24, 2007 9:19 AM

56

Hansen Solubility Parameters: A User’s Handbook

TABLE 3.2A (CONTINUED) Total and Partial Solubility Parameters (in MPa1/2) of Pure Fluids Total Fluid Glycerol Diethylamine n-Butylamine Tetrahydrofuran Formamide Dimethylformamide Acrylonitrile Dimethylsulfoxide Acetic acid Acrylic acid Propionic acid Butyric acid Methacrylic acid Octanoic acid Oleic acid Stearic acid Ammonia Water

HB

5

Calc

Exp

34.12 16.61 18.28 19.46 19.46 36.65 23.95 21.59 26.75 21.35 24.01 19.95 20.2 21.00 21.00 22.24 17.38 19.04 24.63 47.82

34.34 16.80 18.48 18.93 19.26 38.18 23.62 22.29 26.20 27.58 25.90 25.44 23.96 23.85 23.67 21.61 16.81 18.87 26.52 48.68

29.25 6.10 8.00 8.00 8.00 19.00 11.25 6.80 10.20 13.52 14.90 12.40 10.60 10.20 10.20 8.20 5.50 5.50 17.80 42.32

Exp

P

5

Exp 29.18 6.13 8.08 8.00 8.00 18.80 11.16 6.80 10.28 12.09 14.90 12.19 12.21 10.25 10.22 8.66 4.95 4.30 17.80 42.17

5

Calc

12.07 2.30 4.50 5.70 5.70 26.20 13.70 12.80 16.40 7.98 6.40 5.30 4.10 2.80 2.80 3.30 3.10 3.30 15.70 16.00

14.31 2.89 4.75 6.98 5.70 21.41 13.99 13.16 14.75 8.22 6.31 6.81 5.51 6.87 2.80 3.35 2.49 1.86 15.70 16.00

TABLE 3.2B Total and Partial Solubility Parameters (in MPa1/2) of Common Polymers Total

HB

Fluid

Exp

5

5

Calc

Exp

Polyethylene-lin. Polypropylene Polystyrene Poly(vinyl chloride) Polyacrylonitrile Poly(methyl methacrylate) Polycarbonate (bisphenol A) Poly(ε-caprolactone) Poly(vinyl acetate) Nylon 66

16.26 18.10 19.26 19.55 27.43 21.52 20.25 20.20 18.18 30.87

17.37 18.15 19.26 21.93 28.37 22.35 20.43 20.33 18.42 33.09

2.80 1.00 2.90 3.40 9.10 5.10 6.90 8.40 4.00 24.00

P 5

Calc

Exp

2.80 1.00 2.90 3.42 9.10 5.10 6.90 8.40 4.00 23.90

0.80 0.0 4.50 7.80 14.10 10.50 5.90 5.00 2.20 11.00

Calc 0.80 0.0 4.50 8.08 14.08 10.44 5.90 5.00 2.20 11.00

approximations made, the o verall picture is rather satisf actory. We are not a ware of an y similar predictive approach in the literature in order to mak e the analogous comparison. There are a number of comments that should be made re garding Table 3.1a and Table 3.1b. First, the scaling constants for the nonpolar substances are identical to those reported pre viously.16 Thus, their calculated solubility parameters, reported in Table 3.2a, are essentially predictions of

7248_book.fm Page 57 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

57

the model. Second, the series of 1-alkanols is another case where the calculated solubility parameters are, essentially, predictions of the model, as the h ydrogen bonding parameters are the same as in the original NRHB model. 16 Of course, the scaling constants ha ve been changed as the interaction energy is no w split into its dispersion and polar components, b ut the ne w parameter, the dipole moment, m, is not an adjustable parameter . Third, in all other cases of polar substances there is always an e xperimental hydrogen bonding contrib ution5 even in cases where there is no ob vious proton-donor and proton-acceptor pair. There are cases where e ven the dipoles themselv es are not obvious and, apparently, the polar component δp refers to quadrupole or higher -order interactions. A typical example is the case of carbon dioxide, where neither protons nor dipoles are present. In such cases, the donor–acceptor interaction w as replaced by the acid–base or the electrophilic–nucleophilic (carbon–oxygen) interaction, and a fictitious v alue of m w as adjusted on the basis of the corresponding experimental5 value for δp. In a similar manner, in aromatic hydrocarbons, all hydrogens were considered as equi valent proton donors and (the π-electrons of) the aromatic ring as the proton acceptor. Fourth, in the case of polymers, the picture is somewhat more complex. The repeating unit was considered the basis for the calculations, and the reported numbers in Table 3.1b of proton donors and acceptors refer to this basis. Thus, the total number of proton donors and acceptors (for the polymer) are the reported numbers in Table 3.1b multiplied by the de gree of polymerization of each polymer . In this case, the dipole moment of the polymer w as ag ain adjusted on the basis of the e xperimental5 δp. Fifth, for some fluids there are tw o entries in Tables 3.1a and 3.1b. In these cases, the estimated δp on the basis of the literature value25 for m was largely deviating from the experimental one, and thus, in the second entry the v alue of m was adjusted on the basis of the e xperimental5 δp. Figure 3.1 sho ws the calculated components of the solubility parameter of w ater o ver an extended range of saturation temperatures. As was expected, the main contrib ution to δ of w ater, especially at lo w temperatures, is h ydrogen bonding. This type of diagram is most useful for designing applications involving subcritical or supercritical w ater. An analogous diagram for ammonia is sho wn in Figure 3.III.1 of Appendix 3.III. In this case, the contribution of the polar component is as important as that of the hydrogen bonding component, and it appears to override the hydrogen bonding component at the supercritical region. In the same figure, one may compare the temperature dependence of these components as estimated by the alternative approach of Appendix 3.III. 50 δ 40

δ / MPa½

δhb 30

20

δd δp

10

0 250

300

350

400

450 T/K

FIGURE 3.1 Fractional solubility parameters for w ater.

500

550

600

650

7248_book.fm Page 58 Tuesday, April 24, 2007 9:19 AM

58

Hansen Solubility Parameters: A User’s Handbook

17.6

δd, alkanol

17.2

16.8

16.4

16.0

15.6 13.0

13.5

14.0

14.5

15.5 δalkane

15.5

16.0

FIGURE 3.2A Comparison of the experimental dispersion component of solubility parameters of 1-alkanols 5 with the solubility parameters of the corresponding n-alkane homomorphs.

16.4

δd, alkanol

16.0

15.6

15.2

14.8

14.4 12.5

13.0

13.5

14.0

14.5

15.0

15.5

δalkane FIGURE 3.2B Comparison of the calculated dispersion component of solubility parameters of 1-alkanols with the solubility parameters of the corresponding n-alkane homomorphs. The correlation line is: δd,alkanol = 6.86 + 0.60 δalkane .

It is often proposed, as in our previous approach,11–12 to estimate the dispersion component, δd, in polar fluids from the total solubility parameter of the corresponding homomorph h ydrocarbon. This approach, ho wever, is not al ways successful as sho wn in Figure 3.2A, where the dispersion component, δd, of alkanols is compared with the total solubility parameter of the corresponding homomorph hydrocarbon. As shown, the experimental5 data not only f all away from the diagonal; they do not e ven f all on a straight line. As a consequence, the route for the estimation of the solubility parameter components through the homomorph concept is not al ways a safe w ay. In

7248_book.fm Page 59 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

59

contrast, when we compare these tw o solubility parameters as calculated by the present approach, the data seem to f all at least on a straight line, as sho wn in Figure 3.2B. Ho wever, care must be exercised once ag ain, as in this figure, the slope of the straight line is much lo wer than unity. The splitting of the potential energy into its dispersion, polar, and hydrogen bonding components in Equation 3.33, which led to the e xplicit forms of Equation 3.34 through Equation 3.36, is most useful for an additional reason: If the partial solubility parameters and the molar volume are known (e.g., compilations by Hansen, 5 Barton,3 van Krevelen,30 etc.), and the normal boiling point (or the vapor pressure at some other temperature) are either kno wn or can be estimated with reasonable accuracy, then one may use this information and the abo ve formalism to estimate the scaling constants and the h ydrogen bonding ener gy of the fluid through a least squares fit. Of course, the dipole moment is also needed and, if not kno wn, it may be estimated by v arious rather widely available ab initio or semiempirical quantum mechanics calculations. We have applied the above procedure to acetophenone by using Hansen’ s compilation5 for the partial solubility parameters and molar v olume, and DIPPR 25 for the normal boiling point and the dipole moment. The obtained scaling constants are: ε* = 5267 J/mol, v * = 10.177 cm 3/mol, v sp* = 0.9495 cm3/g, and EH = 6293 J/mol, rather close to the corresponding parameters reported in Table 3.1A. These scaling constants can no w be used for the estimation of the basic thermodynamic properties of the fluid at an y temperature and pressure. The e xperimental25 and the calculated (essentially, predicted) vapor pressures and saturated liquid densities for acetophenone are compared in Figure 3.3A and Figure 3.3B, respecti vely. As observed, this procedure leads to a reasonably accurate estimation of the thermodynamic properties of fluids. A most useful concept that quantifies the similarity of tw o substances 1 and 2, especially the similarity of a polymer , 2, and a potential solv ent, 1, for it, is the solubility parameter distance, Ra, defined by: 5

(

Ra = ⎡ 4 δ d 2 − δ d 1 ⎣⎢

) + (δ 2

p2

− δ p1

) + (δ 2

hb 2

)

2 − δ hb1 ⎤ ⎥⎦

(3.38)

The idea is: the smaller the Ra, the better is the solv ent for the polymer. A sphere with radius Ro encompasses the good solv ents for this polymer . A refined discussion on Ra and the related quantities Ro and RED = Ra/Ro is pro vided by Hansen. 5,31 The partial solubility parameters for (Bisphenol A) Polycarbonate as functions of temperature, as calculated using the scaling constants in Table 3.1B, are shown in Figure 3.4. It can be seen that all three components are nonnegligible and there is a cross-over in the polar and hydrogen bonding components for this polymer . The distances ( Ra) of this polymer with three common solv ents are compared in Figure 3.5. According to the calculations, chloroform is the best of the three solv ents for this polymer , follo wed by tetrah ydrofuran (THF). Heptane is the w orst and, essentially , a nonsolvent for the polymer , and all these findings agree with e xperiment. The distances ( Ra) for polyprop ylene with three solv ents: tetrah ydrofuran, chloroform, and tetralin are similarly compared in Figure 3.6. As shown, the best solvent for polypropylene appears to be tetralin, which is ag ain corroborated by the e xperiment. This type of figure is most useful not only for the mere selection of the solv ent, but also for the selection of the e xternal conditions (especially, temperature) for the dissolution of the polymer or an y other solute.

DISCUSSION AND CONCLUSIONS A new approach has been presented for the estimation of the partial solubility parameters of pure substances. The capacity of the approach appears satisf actory for both the estimation of the partial solubility parameters and the description of the thermodynamic beha vior of fluids o ver a broad

7248_book.fm Page 60 Tuesday, April 24, 2007 9:19 AM

60

Hansen Solubility Parameters: A User’s Handbook

2.5

2.0

P / MPa

1.5

1.0

0.5

0.0 250

300

350

400

450

500 T/K

550

600

650

700

FIGURE 3.3A Experimental (symbols) 25 and predicted (line) v apor pressures for acetophenone.

1.1

ρ / g.cm3

1.0

0.9

0.8

0.7

0.6 250

300

350

400

450

500 T/K

550

600

650

700

FIGURE 3.3B Experimental (symbols) 25 and predicted (line) liquid densities for acetophenone.

range of temperature and pressure. The author and his co workers are not a ware of an y similar integral approach in the literature in order to mak e comparison. The equation-of-state approach for the estimation of the partial solubility parameters, which is presented in this w ork, has a number of features. First, it is in principle, applicable to an y fluid regardless of its size and shape. Second, it permits the estimation of the partial solubility parameters over an extended range of temperature and pressure. Third, it may utilize a vailable information on

7248_book.fm Page 61 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

61

20

16

d

δ / MPa½

12

8 p

4

hb

280

320

360

400 T/K

440

480

330

340

FIGURE 3.4 Partial solubility parameters for (bisphenol A) polycarbonate.

10

n-Heptane

Ra / MPa½

8

6

4

2

0 290

THF¹

THF²

Chloroform

300

310

320 T/K

FIGURE 3.5 The estimated solubility parameter distance, Ra, of (bisphenolA) polycarbonate (MW = 100000) with three common solv ents, as a function of temperature. The lines mark ed with THF1 and THF2 were obtained by using the first and second set of scaling constants in Table 3.1A, respectively.

7248_book.fm Page 62 Tuesday, April 24, 2007 9:19 AM

62

Hansen Solubility Parameters: A User’s Handbook

10

THF

Ra / MPa½

8

Chloroform 6

4

Tetralin

2 290

300

310

320

330

340

T/K FIGURE 3.6 The estimated solubility parameter distance, Ra, of (linear) Polyprop ylene with three common solvents, as a function of temperature. The lines for tetralin and tetrahydrofuran (THF) were obtained by using the second set of scaling constants in Table 3.1A.

the partial solubility parameters for the estimation of the scaling constants of substances for which there are no a vailable e xtensive e xperimental data on v apor pressures, heats of v aporization, orthobaric densities, etc. Fourth, it may act as useful guide for the selection of appropriate solv ents and/or dissolution conditions. Of course, the statistical thermodynamic model, on which the abo ve approach resides, can be used for a detailed description of the phase diagrams of pairs (mixtures) of fluids when the scaling constants and the binary interaction parameters are a vailable. However, when a f ast screening of potential solvents is needed, the approach of this w ork is sufficient and most v aluable. A no vel element in our approach is the w ay the potential ener gy is split into its dispersion, polar, and h ydrogen bonding components. The calculation of the polar component, in particular , is rather oversimplified and there is much room for improvement if one wishes to use more involved expressions for the function f in Equation 3.30. The alternative approach in Appendix 3.II is one example. Significant progress could be made if e xperimental information on the partial solubility parameters as functions of temperature and pressure were a vailable. One further possibility is the use of a group contrib ution method for the estimation of the dispersion component in much the same way as suggested in, 14 as sho wn in Appendix 3.III. As shown, δd can be estimated with an average absolute error of 0.40, less than half the corresponding error for the estimation of total with the same method. 14 Such a group contrib ution method f ails dramatically, ho wever, for the hydrogen bonding component, which enhances further the usefulness of the approach reported in the present work. Once the estimations of total δ, δhb, and δd are available, the polar component is obtained by a simple subtraction.

ACKNOWLEDGMENTS The contribution of E. Stef anis and I. Tsivintzelis in the preparation of tables and figures of this chapter is gratefully ackno wledged. The contribution of C. M. Hansen through his v aluable comments is also gratefully ackno wledged.

7248_book.fm Page 63 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

LIST OF SYMBOL SPECIAL TO THIS CHAPTER E G H K L N N Nr No Nij P R R m S s T Q(N, P, T) V* V X Y z zq

Potential energy Gibbs free ener gy Enthalpy Boltzmann’s constant Staverman’s parameter Number of e xperimental points Number of molecules Total number of lattice sites Number of empty lattice sites Number of contacts of type i-j Pressure Gas constant Number of se gments per molecule Dipole moment Entropy Surface to v olume fraction Temperature Configurational partition function of fluid in the N, P , T ensemble, see Equation 3.2, p. 46 Average segmental volume Volume ole fraction in liquid phase ole fraction in v apor phase Lattice coordination number Average number of e xternal contacts per molecule

GREEK LETTERS ν* Γ Δ E Θ Θ M P φ Ω Ω

Segmental volume Non random f actor Solubility parameter Interaction energy Surface (contact) fraction Hole-free surface (contact) fraction Chemical potential Density Segment fraction Geometric–flexibility parameter Combinatorial term (In Equation 3.7 and Equation 3.20)

63

7248_book.fm Page 64 Tuesday, April 24, 2007 9:19 AM

64

Hansen Solubility Parameters: A User’s Handbook

SUPERSCRIPT ~ * L V

Reduced quantity Scaling constant Liquid phase Vapor phase

SUBSCRIPT d, hb, p Dm H O R

Dispersion, hydrogen bonding, and polar component, respecti vely Quantity pertinent to dimer Hydrogen bonding quantity Property pertinent to holes Property pertinent to molecular se gments

REFERENCES 1. Hildebrand, J. and Scott, R.L., Regular Solutions, Prentice-Hall, Englewood Cliffs, NJ, 1962. 2. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities I., J. Paint Technol., 39(505), 104–117, 1967. 3. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press, Boca Raton, FL, 1983. 4. Barton, A.F.M., Applications of solubility parameters and other cohesion parameters, Polym. Sci. Technol. Pure Appl. Chem., 57(7), 905–912, 1985. 5. Hansen, C.M., Hansen Solubility Parameters: A User’s Handbook, CRC Press, Boca Raton, FL, 1999. 6. Hansen, C.M., Aspects of solubility , surf aces, and dif fusion in polymers, Prog. Org. Coat., 51(1), 55–66, 2004. 7. Tehrani, J., Am. Lab., 40hh-40mm, February 1993. 8. Bustamante, P., Peña, M.A., and Barra, J., Int. J. Pharm., 174, 141–150, 1998. 9. Jensen, W.B., in Surface and Colloid Science in Computer Technology, Mittal, K.L., Ed., Plenum Press, New York, 1987, pp. 27–59. 10. Karger, B.L., Sn yder, L.R., and Eon, C., J. Chromatogr., 125, 71–88, 1976. 11. Panayiotou, C., Fluid Phase Equilibria, 131, 21–35, 1997. 12. Panayiotou, C., Fluid Phase Equilibria, 236, 267, 2005. 13. Panayiotou, C. and Sanchez, I.C., J. Phys. Chem., 95, 10090–10097, 1991. 14. Stefanis, E., Constantinou, L., and P anayiotou, C., Ind. Eng. Chem. Res., 43, 6253–6360, 2004. 15. Stefanis, E., Tsivintzelis, I., and P anayiotou, C., Fluid Phase Equilibria, 240, 144–154, 2006. 16. Panayiotou, C., P antoula, M., Stef anis, E., Tsivintzelis, I., and Economou, I., Ind. Eng. Chem. Res., 43, 6592–6606, 2004. 17. Guggenheim, E.A., Mixtures, Clarendon Press, Oxford, 1952. 18. Panayiotou, C. and Vera, J.H., Polym. J., 14, 681–694, 1982. 19. Twu, C.H. and Gubbins, K.E., Chem. Eng. Sci., 33, 863–878, 1978. 20. Kraska, T. and Gubbins, K.E., Ind. Eng. Chem. Res., 35, 4727–4737, 1996. 21. Stell, G., Rasaiah, J.C., and Narang, H., Mol. Phys., 27, 1393–1414, 1974. 22. Nezbeda, I. and P avlíek, J., Fluid Phase Equilibria, 116, 530–536, 1996. 23. Nezbeda, I. and Weingerl, U., Mol. Phys., 99, 1595–1606, 2001. 24. Karakatsani, E., Sp yriouni, T., and Economou, I., AIChE J., 51(2005), 2328–2342. 25. Daubert, T.E. and Danner , R.P., Eds., Data Compilation Tables of Properties of Pure Compounds, AIChE Symp. Ser. No. 203, American Institute of Chemical Engineers, Ne w York, 1985. 26. Perry, R. and Green, D., Ed., Chemical Engineers’ Handbook, CD, McGra w Hill, New York, 1999. 27. Fredenslund, A., Jones, R.L., and Prausnitz, J.M., AIChE J., 21, 1086–1099, 1975.

7248_book.fm Page 65 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

65

28. Fredenslund, A., in Models for Thermodynamic and Phase Equilibria Calculations, Sandler, S., Ed., Marcel Dekker, New York, 1994. 29. Zoller, P. and Walsh, D., PVT Data for Polymers, Technomic Publ. Co., Lancaster , Basel, 1995. 30. van Krevelen, D.W., Properties of Polymers, Elsevier, Amsterdam, 2nd ed., 1976. 31. Hansen, C.M., Fifty years with solubility parameters — past and future, Prog. Org. Coat., 51(1), 77–84, 2004.

APPENDIX 3.I: THE ACID DIMERIZATION Following the LFHB practice, 13 one may derive the formalism for the acid dimerization in a rather straightforward manner . F or simplicity , we will consider dimerization only , as dimers are the overwhelming majority of the association species in h ydrogen-bonded acids. Let Ndm be the number of dimers in the system. One can select these dimerized molecules out of the N acid molecules in, N! ! N − 2Ndm !

(2 N ) ( dm

(3.I.1)

)

ways, and form the Ndm dimers in N! N! 2 N dm − 1 2 N dm − 3 ...1 = ! N − 2 N dm ! N − 2 N dm ! N dm ! 2 N dm

(2 N ) ( dm

)

(

)(

)

(

)

(3.I.2)

ways. If Gdm = E dm + PVdm − TS dm

(3.I.3)

is the free energy change upon formation of one dimer, the hydrogen bonding factor in the partition function becomes: QH =

(

⎛ ρ ⎞ N! N dm ⎜ ⎝ rN ⎟⎠ N − 2 N dm ! N dm ! 2

N dm

)

⎛ N G ⎞ exp ⎜ − dm dm ⎟ RT ⎠ ⎝

(3.I.4)

The equilibrium number of dimers per mol of segments of acid, νdm, is obtained from the above equation through the usual free ener gy minimization condition, or 2+ ν dm =

1 − K dm

1 4 + 2 K dm K dm 4r

(3.I.5)

where, K dm =

⎛ −Gdm ⎞ ρ exp ⎜ r ⎝ RT ⎟⎠

(3.I.6)

In this case of dimerization, the h ydrogen bonding contrib ution to the chemical potential is: μH 1 = r ν dm − ln RT 1 − 2 r ν dm

(3.I.7)

7248_book.fm Page 66 Tuesday, April 24, 2007 9:19 AM

66

Hansen Solubility Parameters: A User’s Handbook

APPENDIX 3.II: AN ALTERNATIVE FORM OF THE POLAR TERM The factor f in Equation 3.32 for the dipolar interactions implies that the dipolar forces depend on temperature and/or v olume in the same manner as do the dispersion forces. This is a rather gross simplification as one w ould expect the dipolar forces to be in versely proportional to temperature or to a function of temperature. 19–24 This has been e xplored by adopting the follo wing alternative form for the f actor f: 2

f =

2 ⎛ m ⎞ s2 π ⎜⎝ r ⎟⎠ T

(3.II.1)

With this expression, the formalism of the main te xt remains the same e xcept for the equation for the potential ener gy, which now becomes: 2 ⎡ 4 ⎛ m ⎞ s2 ⎤ ⎥ − N H EH − E = Γ rr qN θr ε∗ ⎢1 + ⎜ ⎟ ⎢ π⎝ r ⎠ T ⎥ ⎣ ⎦

(3.II.2)

This change will change, of course, the scaling constants of the fluids. These ne w scaling constants are reported in Table 3.III.1 for some representati ve fluids. These constants describe the key thermodynamic properties of the fluids in a similar, almost identical manner to the one obtained by the corresponding scaling constants of the main text. In addition, the predicted partial solubility parameters by the tw o sets of the scaling constants are compared in Table 3.II.2. As observ ed, Equation 3.II.1 and Equation 3.II.2 do not lead to any clear improvement in this respect either. The essential difference is the dependence of δp on temperature, which is no w given by:

δp =

⎡ 4 ⎛ m ⎞ 2 s2 ⎤ ⎥ Γ rr qN θr ε∗ ⎢ ⎜ ⎟ ⎢π ⎝ r ⎠ T ⎥ ⎣ ⎦ V

(3.II.3)

As sho wn in Figure 3.II.1, the tw o alternati ve approaches for the estimation of the polar component lead to differences not only in δp but also in δd and to the total . The hydrogen bonding component appears, ho wever, intact. This is important, as δhb may be used in approaches lik e the one reported in Appendix 3.III.

APPENDIX 3.III: A GROUP-CONTRIBUTION METHOD FOR THE PREDICTION OF δ AND δD The details of the group contrib ution method may be found in the original w ork.14 Two kinds of functional groups are used: First-order (UNIF AC groups) and second-order groups that are based on the conjug ation theory. The basic equation that gives the value of each property according to the molecular structure is: f(p) =

∑n F i

i

i

+

∑m j

j

Sj

(3.III.1)

7248_book.fm Page 67 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

67

TABLE 3.II.1 Characteristic Constants of Pure Fluidsa Fluid

ε* = RT*/J.mol–1

Benzene Toluene Tetralin Acetone Acetophenone Ethyl acetate n-Butyl acetate Methyl methacrylate Diethyl ether 1,4-Dioxane Chloroform Dichloromethane Chlorobenzene Methanol Ethanol 1-Butanol 1-Octanol Phenol Ethylene glycol 1,2-Propylene glycol Glycerol Diethylamine n-Butylamine Tetrahydrofuran Acetic acid Butyric acid Ammonia Water

5041 5236 5858 4207 5875 4444 4741 5046 4040 4435 4859 5062 5517 3798 4042 4656 5280 6868 5681 4386 2595 4471 4927 4491 5110 4981 2506 2070

a b

ν* = ε*P*–1/cm3 .mol–1

ν*sp = ρ*1/cm3.g–1

9.526 10.684 11.034 10.709 10.899 14.170 13.484 17.060 12.293 11.216 10.026 10.704 10.420 13.710 14.040 14.031 15.250 14.805 20.048 12.800 30.000 12.232 14.685 11.261 7.019 7.776 7.500 23.862

1.078 1.093 1.000 1.171 0.958 1.033 1.054 1.019 1.192 0.865 0.617 0.747 0.873 1.178 1.155 1.145 1.145 0.940 0.961 0.942 0.798 1.262 1.257 1.029 0.902 0.962 1.399 0.997

–EH/J.mol–1 3724 3590 5055 7970 6192 10915 8674 5383 6630 6501 8610 4992 4269 25100 25100 25100 25100 23300 21775 22400 24600 12240 11980 8810 23735 23822 11940 18198

SH was set equal to 26.5 JK –1mol–1 in all cases. Adjusted to fit δhb.

where Fi is the contribution of the first-order group of type i that appears n i times in the compound and, Sj, is the contribution of the second-order group of typej that appears mj times in the compound. f(p) is a single equation of the property , p, and is selected after a thorough study of the ph ysicochemical and thermodynamic beha vior of the property . The determination of the contrib utions is done by a two-step regression analysis for the F is and the S js, respectively. A least-square analysis has been carried out to estimate the first-order and second-order group contrib utions for the solubility parameters. In Table III.1, the first-order group contrib utions for total solubility parameter , δ, and the dispersion partial solubility parameter (Hansen), δd, at 25°C are presented. Table III.2 sho ws the second-order group contrib utions for the same properties. The selected equations for the estimation of each property are the follo wing: Total solubility parameter, δ, at 25°C ((kJ/m 3)(1/2)): δ=(

∑n F i

i

i

+

∑m j

j

S j + 75954.1)

0.383837

− 56.14

(3.III.2)

7248_book.fm Page 68 Tuesday, April 24, 2007 9:19 AM

68

Hansen Solubility Parameters: A User’s Handbook

TABLE 3.II.2 Total and Partial Solubility Parameters (in MPa1/2) of Pure Fluids Polar/Hydrogen-Bonded Fluids Total

HB

P

Fluid

Expa

Calc

Expa

Calc

Expa

Calc

Benzene Toluene Tetralin Acetone Acetophenone Ethyl acetate n-Butyl acetate Diethyl ether 1,4-Dioxane Chloroform Dichloromethane Methanol Ethanol 1-Butanol 1-Octanol Phenol Ethylene glycol 1,2-Propylene glycol Glycerol Diethylamine n-Butylamine Tetrahydrofuran Acetic acid Butyric acid Ammonia Water

18.41 18.32 19.80 19.95 21.73 18.48 17.59 15.66 20.47 18.94 20.79 29.61 26.50 23.35 20.87 24.63 33.70 29.52 34.12 16.61 18.31 19.46 21.35 20.20 27.40 47.82

18.41(18.27)a 17.95(17.78) 18.85(18.80) 21.29(20.04) 21.20(20.48) 18.66(18.31) 17.72(17.43) 15.71(15.73) 19.67(20.08) 19.46(19.18) 20.02(19.92) 30.86(29.89) 26.26(26.08) 22.92(22.90) 20.30(20.27) 24.59(24.69) 33.97(33.64) 31.90(29.19) 34.14(34.34) 16.97(16.80) 18.32(18.48) 20.05(18.93) 27.70(27.58) 24.74(23.96) 28.73(26.52) 47.80(48.68)

2.05 2.00 2.90 6.95 3.68 9.20 6.30 5.11 7.36 5.73 4.09 22.30 19.43 15.80 11.86 14.90 25.77 23.32 29.25 6.10 8.00 8.00 13.52 10.60 17.80 42.82

2.05 2.00 2.90 6.95 3.68 9.20 6.30 5.11 7.36 5.73 4.09 24.15(24.08) 20.08(19.98) 15.84(15.80) 11.97(11.94) 14.90 25.93 23.44 30.91 6.10 8.00 8.00 13.52 10.60(12.21) 17.80 43.15

1.02 1.40 2.0 10.43 8.59 5.32 3.70 2.86 1.84 3.07 7.36 12.27 8.80 5.70 3.27 5.90 11.05 9.41 12.07 2.30 4.50 5.70 7.98 4.10 15.70 16.00

0.93(1.01) 0.84(0.92) 1.16(0.41) 10.44(10.14) 7.46(7.04) 5.70(6.07) 4.39(4.70) 2.98(3.48) 1.84(1.91) 3.34(3.79) 5.56(5.63) 12.11(11.34) 8.49(8.24) 5.70(5.72) 3.27(3.75) 6.31(5.16) 16.01(12.2) 13.93(12.56) 12.10(14.31) 2.49(2.89) 4.86(4.75) 6.04(6.98) 7.98(8.22) 4.69(5.51) 16.10(15.70) 18.70(16.00)

a

Values in parenthesis from Table 3.2A.

Dispersion partial solubility parameter , δd, at 25°C ((kJ/m 3)(1/2)): δd =

∑n F i

i

i

+

∑m

j

S j + 17.3231

(3.III.3)

j

The quantity mjSj is considered to be zero for compounds that do not have second-order groups. Table 3.III.3 illustrates the statistics concerning the o verall improvement in the estimation of solubility parameters that has been achie ved after the introduction of second-order groups in the regression. As observed, the method is rather quite satisf actory.

Standard Deviation =

∑ (X

est

− X exp )2

N

Average Absolute Error = AAE =

1 N

∑X

,

est

− X exp , and

7248_book.fm Page 69 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

69

30

30 (A) 25

(B) 25

δT

20 δ / MPa½

δ / MPa½

20 δhb

15

10

δT

δp

δhb 15

δd

δp 10

δd

5 290 300 310 320 330 340 350 360 370 380 T/K

5 290 300 310 320 330 340 350 360 370 380 T/K

FIGURE 3.II.1 Fractional solubility parameters for ammonia. A: with the scaling constants from Table 3.1A. B: with the constants from Table 3.II.1.

Average Absolute Percent Error = AAPE =

1 N

∑

Xest − X expp X exp

100% ,

where N is the number of data points, X est is the estimated v alue of the property , and X exp the experimental value. It is w orth pointing out that the solubility parameters of comple x structures that occur in aromatic or multiring compounds of biochemical interest can no w be predicted by only using their molecular structure and without an y other data kno wn. Syntactic isomers can be distinguished, whereas stereoisomers cannot. The estimation of one of the other Hansen solubility parameters, such as δhb, as described in the main te xt, could lead to the estimation of δp as well.

Example of Prediction of the Hansen Partial Solubility Parameter, δd, for 1-Butanol First-Order Groups

Occurrences, ni

Contributions, Fi

niFi

1 3 1 — —

–0.97135 –0.02686 –0.34621 — —

–0.97135 –0.08058 –0.34621 –1.39814 17.3231

–CH3 –CH2 –OH ΣniFi Universal constant, C

No second-order groups are in volved. δd =

∑n F i

i

i

+ 17.3231 = 15.92496 (kJ/m 3 )(1/2)

7248_book.fm Page 70 Tuesday, April 24, 2007 9:19 AM

70

Hansen Solubility Parameters: A User’s Handbook

TABLE 3.III.1 First-Order Group Contributions to (Total) δ and δd at 25°C First-Order Groups -CH3 -CH2 -CH< >C< CH2 = CH-CH = CHCH2 = C< -CH = C< >C = C< CH2 = C = CHCHECCEC ACH AC ACCH3 ACCH2CH3CO CH2CO CHO COOH CH3COO CH2COO HCOO COO OH ACOH CH3O CH2O CHO CH2O (CYCLIC) CH2NH2 CHNH2 CH3NH CH2NH CHNH CH3N CH2N ACNH2 CONH2 CONHCH3 CON(CH3) 2 C 5 H 4N C 5 H 3N CH2SH CH2S I BR CH2CL CHCL

Contributions to δ –2308.6 –277.1 –355.5 –176.2 –2766.2 –381.9 –980.2 1887.1 1601.8 –3745.0 –975.5 2169.3 –6.4 684.3 –221.8 1023.4 3269.1 7274.2 5398.2 9477.8 1865.1 5194.2 1716.0 3671.8 12228.9 8456.1 –480.8 –206.7 1229.1 3733.9 3650.7 560.4 8616.2 4183.8 3381.8 2166.5 –2662.6 9228.4 14930.1 27386.9 12770.8 4686.3 6574.7 2191.2 3585.2 3183.8 2163.8 1923.3 426.3

Contributions to δd –0.97135 –0.02686 0.64501 1.26857 –1.05853 0.00476 –0.48289 0.53723 0.35922 –1.65178 0.23203 –0.20284 0.11050 0.84464 0.21737 0.69325 –0.35506 0.65267 –0.40303 –0.29100 –0.54006 0.29130 na 0.20386 –0.34621 0.52883 –0.58280 0.03098 0.88334 0.27531 –0.58277 0.01116 na 0.81162 na di-Isoprop 0.87693 1.46810 1.69868 –0.06889 na 0.44822 na 2-Meth na 2,6-Dimeth 1.27971 1.05949 0.77971 0.57169 0.26226 0.44622

Sample Group Assignment (occurrences) Propane (2) Butane (2) Isobutane (1) Neopentane (1) Propylene (1) cis-2-Butene (1) Isobutene (1) 2-Methyl-2-butene (1) 2,3-Dimethyl-2-butene (1) 1,2-Butadiene (1) Propyne (1) 2-Butyne (1) Benzene (6) Naphthalene (2) Toluene (1) m-Ethyltoluene (1) Methyl ethyl ketone (1) Cyclopentanone (1) 1-Butanal (1) Vinyl acid (1) Ethyl acetate (1) Methyl propionate (1) n-Propyl formate (1) Ethyl acrylate (1) Isopropanol (1) Phenol (1) Methyl ethyl ether (1) Ethyl vinyl ether (1) Diisopropyl ether (1) 1,4-Dioxane (2) 1-Amino-2-propanol (1) Isopropylamine (1) n-Methylaniline (1) di-n-Propylamine (1) ylamine (1) Trimethylamine (1) Triethylamine (1) Aniline (1) 2-Methacrylamide (1) n-Methylacetamide (1) N,N-Dimethylacetamide (1) ylpyridine (1) ylpyridine (1) n-Butyl mercaptan (1) Diethyl sulfide (1) Isopropyl iodide (1) 2-Bromopropane (1) n-Butyl chloride (1) Isopropyl chloride (1)

7248_book.fm Page 71 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

71

TABLE 3.III.1 (CONTINUED) First-Order Group Contributions to (Total) δ and δd at 25°C First-Order Groups CCL CHCL2 CCL2 CCL3 ACCL ACF CL-(C=C) CF3 CH2NO2 CHNO2 ACNO2 CH2CN CF2 CF C 4 H 3S F (except as abo ve) CH2 = C = C< CH = C = CHCHCO O (except as abo ve) Cl (except as abo ve) NH2 (except as abo ve) >C = N-CH = NNH (except as abo ve) N = NCN (except as abo ve) NO2 (except as abo ve) O = C = NCHSH CSH SH (except as abo ve) S (except as abo ve) SO2 >C = S >P>C = 0 (e xcept as abo ve) N (except as abo ve)

Contributions to δ –1415.6 1164.0 na 0.36532 –1208.7 1332.2 –701.5 –473.5 –5199.5 10030.7 12706.7 6303.5 9359.8 –3464.4 na 0.17069 4722.7 –2965.3 –2326.1 –795.6 7805.8 2467.6 636.3 –841.5 3380.7 5026.4 3459.4 –7339.6 10253.0 1655.1 2694.6 1234.8 2230.2 na 4770.2 14215.0 26271.8 –1643.4 na na

Contributions to δd

Sample Group Assignment (occurrences)

2.75755 1.17971

t-Butyl chloride (1) 1,1-Dichloropropane (1) 2,2-dichloropropane (1) na Benzotrichloride (1) 0.84750 m-Dichlorobenzene (2) 0.11704 Fluorobenzene (1) 0.22893 2,3-Dichloropropene (1) –0.22931 Perfluorohexane (2) na 1-Nitropropane (1) na 2-Nitropropane (1) 1.41953 Nitrobenzene (1) –0.33919 n-Butyronitrile (1) –0.97290 Perfluoromethylcyclohexane (5) Perfluoromethylcyclohexane (1) na 2-Meth ylthiophene (1) –0.70693 2-Fluoropropane (1) –0.28043 3-Methyl-1,2-butadiene (1) na 2,3-Pentadiene (1) na Diisopropyl ketone (1) 0.04716 Divinyl ether (1) 0.22562 Hexachlorocyclopentadiene (2) na Melamine (3) –0.30737 2,4,6-Trimethylpyridine (1) 0.96719 Isoquinoline (1) na Dibenzopyrrole (1) na p-Aminoazobenzene (1) 0.08615 cis-Crotonitrile (1) na Nitroglycerine (3) –0.13065 n-Butyl isocyanate (1) na Cyclohexyl mercaptan (1) na tert-Butyl mercaptan (1) 1.04271 2-Mercaptobenzothiazole (1) 1.48988 Thiophene (1) 1.55021 Sulfolene (1) 0.77470 N-Methylthiopyrrolidone (1) na T riphenylphosphine (1) –0.43429 Anthraquinone (2) 1.54378 Triphenylamine (1)

Note: na = not a vailable.

Thus, estimated δd =15.925 MP a1/2, experimental5 δd =16.00 MP a1/2 Percentage error = (16.00-15.925)/16.00 = 0.47% According to Table 3.II.2, this equation-of-state approach estimates = 22.92 MP a1/2 and δhb = 15.80 MPa1/2. These data combined with the group contribution result for δd give: δp = 4.70 MPa1/2. The experimental value5 is 5.72 MP a1/2.

7248_book.fm Page 72 Tuesday, April 24, 2007 9:19 AM

72

Hansen Solubility Parameters: A User’s Handbook

TABLE 3.III.2 Second-Order Group Contributions to (Total) δ and δd at 25°C Second-Order Groups

Contributions, δ

Contributions, δd

Sample Group Assignment (occurrences)

(CH3)2-CH(CH3)3-C-CH(CH3)-CH(CH3)-CH(CH3)-C(CH3)2-C(CH3)2-C(CH3)2ring of 5 carbons ring of 6 carbons -C = C - C = CCH3-C = -CH2-C = >C{H or C}-C = string in c yclic >CHCHO CH3(CO)CH2C(cyclic) = O ACCOOH >C{H or C}-COOH CH3(CO)OC{H or C}< (CO)O(CO) ACHO >CHOH >C < OH -C(OH)C(OH)-C(OH)C(N) C(in cyclic)-OH C-O-C = C AC-O-C >N{H or C}(in c yclic) -S-(in cyclic) ACBr ACI (C = C)-Br CH3(CO)CH< ring of 3 carbons ring of 4 carbons ring of 7 carbons ACCOO AC(ACHm)2AC(ACHn)2 Ocyclic-Ccyclic = O AC-O-AC CHn-O-OH CHm-O-O-CHn NcycHm-Ccyc = O Ocyc-CcycHm = Nc yc -O-CHm-O-CHnAC-NH-AC C(= O)-C-C(= O)

142.1 592.3 1581.2 2678.4 5677.6 –2637.7 –524.2 –426.8 11.9 –762.7 –1257.2 626.1 –1634.4 142.0 –3745.0 –3076.5 511.1 134.4 –2875.9 3315.0 –359.5 –23.4 5020.6 3306.4 4022.7 –228.5 2493.0 –492.7 2389.4 337.4 1267.1 na –437.1 –9764.5 –3673.4 –1486.4 –83.5 –69.8 9215.6 –4646.5 2002.5 –2029.1 11489.1 –8721.6 –620.3 2.8 –3668.9

0.04604 –0.07377 na na na –0.66808 0.38742 –0.13554 –0.07853 –0.32357 –0.27979 –0.19450 na –0.04509 –0.29806 –0.22930 na –0.52196 –0.27069 0.37724 0.11231 –0.06801 na –0.08088 –0.08764 0.20629 0.25679 0.22183 0.48916 0.12341 0.00000 –0.40589 na 0.02003 na na –0.18466 –0.37514 0.24676 –0.56461 na na 0.29563 na 0.08394 na –0.48615

Isobutane (1) Neopentane (1) 2,3-Dimethylbutane (1) 2,2,3-Trimethylbutane (1) 2,2,3,3-Tetramethylpentane (1) Cyclopentane (1) Cyclohexane (1) 1,3-Butadiene (1) Isobutene (2) 1-Butene (1) 3-Methyl-1-butene (1) Ethylcyclohexane (1) 2-Methylpropanal (1) Methyl ethyl ketone (1) Cyclopentanone (1) Benzoic acid (1) Isobutyric acid (1) Isopropyl acetate (1) Acetic anhydride (1) Benzaldehyde (1) 2-Propanol (1) Tert-Butanol (1) 1,2-Propanediol (1) 1-Amino-2-propanol (1) Cyclohexanol (1) Ethyl vinyl ether (1) Methyl phenyl ether (1) Cyclopentimine (1) Tetrahydrothiophene (1) Bromobenzene (1) Iodobenzene (1) 2-bromo-propene (1) Methyl isopropyl ketone (1) Cyclopropane (1) Cyclobutane (1) Cycloheptane (1) Methyl benzoate (1) Naphthalene (1) Diketene (1) Diphenyl ether (1) Ethylbenzene hydroperoxide (1) di-t-Butyl peroxide (1) 2-Pyrrolidone (1) Oxazole (1) Methylal (1) Dibenzopyrrole (1) 2,4-Pentanedione (1)

Note: na = not a vailable.

7248_book.fm Page 73 Tuesday, April 24, 2007 9:19 AM

Statistical Thermodynamic Calculations

73

TABLE 3.III.3 Comparison of the First- and Second-Order Approximations Standard Standard Data Deviation Deviation AAE AAE AAPE (%) AAPE (%) Property Points First-Order Second-Order First-Order Second-Order First-Order Second-Order δ 1017 δd

344

1.47 0.61

1.31 0.58

1.00 0.44

0.90 0.41

5.15 2.62

4.67 2.42

7248_book.fm Page 74 Tuesday, April 24, 2007 9:19 AM

7248_C004.fm Page 75 Thursday, May 10, 2007 12:51 PM

Hansen Solubility 4 The Parameters (HSP) in Thermodynamic Models for Polymer Solutions Georgios M. Kontogeorgis ABSTRACT Polymer thermodynamics plays an important role in a lar ge number of processes and in the design of many different polymer-based products. Examples include: 1. The removal of unreacted monomers, colorants, by-products, toxic compounds, and other additives after polymerization 2. The selection of mix ed solv ents in the paints and coatings industry to ward designing environmentally-friendly paints (w ater-based, fewer VOC) 3. The control of emissions from paints as well as the swelling of the film in the presenc of water 4. The recycling of polymers based on ph ysicochemical methods like selective dissolution 5. The compatibility of polymer blends including those with no vel structures (star -like, dendrimers), permeabilities of g ases in fl xible polymeric pipes used in the North Sea and other major oil and g as producing areas for transporting of h ydrocarbons on the seabed and from the seabed to the surf ace 6. Compatibility of plasticizers in PVC 7. In the biotechnology , aqueous tw o-phase systems based on polymers for separating proteins This is only a short list, and man y more applications of polymer thermodynamics e xist. In several of these cases it is not suf ficient to empl y only the Hansen solubility parameters (HSP), as much more detailed calculations may be needed, including solv ent activities, for e xample, for solvent emission assessment or even full phase diagrams and at both lo w (e.g., biotechnology) and high pressures (e.g., polyolefin industr , gas permeabilities in polymers). Polymers form highly nonideal liquid solutions with lo w-molecular weight chemicals and liquid–liquid phase separation (LLE) is the rule rather than the e xception in polymer -solvent mixtures. Moreo ver, such LLE may tak e v arious forms; UCST , LCST , closed loop, etc., and temperature, polymer molecular weight, and polydispersity have great effects. Free-volume effects, special structures and crystallinity cause additional comple xities. It is rather tempting to combine the e xtensive use/tables available for the HSP with a rigorous thermodynamic approach and compare the performance of this method to more adv anced approaches. This is the purpose of this chapter . As several of the literature approaches used for

75

7248_C004.fm Page 76 Thursday, May 10, 2007 12:51 PM

76

Hansen Solubility Parameters: A User’s Handbook

comparison purposes are based on the group-contrib ution principle, a short introduction to this principle is pro vided first

GROUP CONTRIBUTION METHODS FOR ESTIMATING PROPERTIES OF POLYMERS THE GROUP-CONTRIBUTION PRINCIPLE SOLUBILITY PARAMETERS)

AND

SOME APPLICATIONS (DENSITY,

Many properties of pure polymers and polymer solutions can be estimated with group contributions (GC), e.g., density, the solubility parameter (Hildebrand and HSP), the melting and glass transition temperatures, activity coefficients, and the sur ace tension. The GC method is based on the assumption that the properties of molecules can be estimated using additive rules from the values of the corresponding groups they are composed of. For example, n-hexane (CH 3-(CH2)4-CH3) can be considered to ha ve two CH 3 and four CH 2 groups. Similarly , butanone has one CH3, one CH2, and one CH3CO group. If the group values are known for a specifi property F, then the total v alue of the property for the whole molecule is often e xpressed by a general additive rule of the form: F=

∑n F i

i

(4.1)

i

or similar additive equations. In Equation 4.1, ni is the number of groups of type i and Fi is the corresponding group v alue. In some cases, Fi values are also functions of temperature for temperature-dependent properties such as the v olume and the v apor pressure. F or several properties, the general GC equation has a more complicated form than that indicated by Equation 4.1. The GC methodology has been applied to man y properties and for both lo w molecular weight compounds and polymers. Activity coefficients h ve also been predicted with group contrib utions, e.g., the UNIFAC model by Fredenslund et al. 1 Van Krevelen2 gives an overview of the application of group contrib ution methods to se veral properties of pure polymers, including also mechanical and environmentally-related properties. Van Krevelen provides extensive GC tables for the Hildebrand and Hansen solubility parameters as well. An alternative GC method for the polymer (and solvent) density has been de veloped by Elbro et al. 3 (GCVOL) and was recently extended to cover several group families.4,5 A list of GCVOL parameters is provided elsewhere.6 GCVOL can predict satisfactorily the density of solvents, oligomers, and polymers, including copolymers, often within 2%.7 The great advantage of the group contribution method is its simplicity: Even though there may be thousands of different molecules (and mixtures), the corresponding number of groups is significantly smaller (no more than 100 or so). Thus, instead of kno wing the parameter v alues of a specific property for thousands of molecules, it is su ficient to kn w the group parameters for a much smaller number of groups. Two limitations of the approach should be k ept in mind: 1. The GC methodology is a v ery useful technique leading to good results in man y cases. However, it is an approximation, based often on a some what unjustified d vision of the molecule into groups. F or some properties, such as for density , GC methods perform much better than for others, e.g., melting point. Specific molecules are assigned a separate groups (e.g., methanol) because further di vision is not possible if good results are to be obtained. Problems can also be e xpected for multifunctional groups and where more than one polar groups are close to each other (e.g., in alcohols and acids with more

7248_C004.fm Page 77 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

77

than one OH and COOH groups; for h ydroxyl-acids or for alkanolamines). Ho wever, despite these problems, the GC principle is e xtensively used for property calculations for specific molecules and also, in the r verse w ay, for selecting suitable compounds having a required set of properties. The latter technique is called computer aided product design. 2. The exact definition of groups may be di ferent from method to method. In some cases, even two different methods for the same property may emplo y different definitions fo the groups. For example, the division of butadiene rubber according to the van Krevelen and GCVOL methods for density is different. In other cases, and for the same GC method, a particular molecule can be di vided into groups in tw o different ways that may yield different results. The latter problem will be further discussed later .

GC FREE-VOLUME-BASED MODELS

FOR

POLYMERS (ENTROPIC-FV, UNIFAC-FV)

The Free-Volume Concept The classical Flory–Huggins model (See Equation 4.9 later) pro vides a first approximation fo polymer solutions. Both the combinatorial and the energetic terms need improvement via inclusion of free-volume effects and nonrandom local-composition terms such as those of the UNIQ UAC, NRTL, and UNIFAC models. The concept of free-volume (FV) is rather loose b ut still very important. Elbro 8 showed, using a simple definition for the FV (Equation 4.2), that the FV percentages of sol ents (40–50%) are greater than those of polymers (30–40%), with the e xception of w ater and polydimeth yl siloxane (PDMS). Many mathematical expressions have been proposed for the FV. One of the simplest and most successful equations is: V f = V − V * = V − Vw

(4.2)

originally proposed by Bondi 9 and later adopted by Elbro et al. 10 and others 11 in the Entropic-FV model. In Equation 4.2, free-volume is simply the “empty” volume available to the molecule when the molecules’ own (hard-core or closed-packed V*) volume is subtracted. But what is actually the hard-core volume? This also is rather difficult to determine. Elbr 8,10 suggested using V* = Vw , i.e., equal to the v an der Waals volume (V w), which is obtained from the group increments of Bondi and is tabulated for almost all existing groups in the UNIFAC tables. Other investigators interpreted the hard-core volume somewhat differently; most agree today that V* > Vw due to the closed packed structure of molecules. For example, the closed-packed cubic structure suggests that V* = 1.35V w, and Bondi mentions that for many organic molecules it can be expected that the ratio V*/Vw should be between 1.28 and 1.43. The UNIFAC-FV Model The original UNIF AC model does not account for the free-v olume differences between solv ents and polymers; consequently , it highly underestimates the solv ent activities in polymer solutions. Empirical modified UNI AC v ersions (de veloped in L yngby and Dortmund) with e xponential segment fractions are also inadequate for polymer solutions; the y systematically o verestimate the solvent activities. Various modifications — xtensions of the classical UNIF AC approach to polymers — ha ve been proposed. All of these approaches include the FV effects, which are neglected in the UNIFAC combinatorial term, and most of them emplo y the energetic (residual) term of UNIFAC. The most well-known is the UNIF AC-FV model by Oishi and Prausnitz 12:

7248_C004.fm Page 78 Thursday, May 10, 2007 12:51 PM

78

Hansen Solubility Parameters: A User’s Handbook

fv ln γ i = ln γ comb + ln γ res i i + ln γ i

(4.3)

The combinatorial (comb) and residual (res) terms are tak en from original UNIF AC.1 All the ener getic parameters in the residual term are the same as in original UNIF AC, i.e., estimated based on vapor-liquid equilibria data for low molecular weight compounds. No parameter reestimation is performed. An additional term is added for the free-volume effects. The FV term used in UNIFAC-FV has a theoretical origin, and it is based on the Flory equation of state:

ln γ

fv i

( (

) ⎤⎥ − c ⎡⎢⎛ v ) ⎥⎥⎦ ⎢⎣⎜⎝ v

⎡ v1i / 3 − 1 ⎢ = 3ci ln ⎢ 1/ 3 ⎢ vm − 1 ⎣

i

i

m

−1 ⎞⎛ 1 ⎞ ⎤ − 1⎟ ⎜ 1 − 1/ 3 ⎟ ⎥ ⎠⎝ v1 ⎠ ⎥⎦

(4.4)

where the reduced v olumes are defined as vi =

Vi bVi,w

x1V1 + x2V2 vm = b x1V1,w + x2V2,w

(

(4.5)

)

In Equation 4.5, the volumes Vi and the van der Waals volumes Vi,W are all expressed in cm3/mol. In the UNIFAC-FV model as suggested by Oishi and Prausnitz 12 the parameters ci (3ci is the number of e xternal de grees of freedom) and b are set to constant v alues for all polymers and solvents ( ci = 1.1 and b = 1.28). The performance of UNIF AC-FV is rather satisf actory, as sho wn by man y in vestigators, for a lar ge v ariety of polymer -solvent systems. Some researchers ha ve suggested that, in some cases, better agreement is obtained when these parameters ( ci and b) are fitted to xperimental data.13 The UNIFAC-FV model was originally developed for solvent activities in polymers and does not give satisfactory results for polymer activities; thus, it has not been applied to polymer-solvent LLE. The Entropic Model A similar to UNIF AC-FV b ut some what simpler approach, which can be readily e xtended to multicomponent systems and liquid-liquid equilibria, is the so-called Entropic-FV model proposed by Elbro et al. 10 and Kontogeorgis et al. 11: ln γ i = ln γ icomb− fv + ln γ ires ln γ icomb− fv = ln ϕ ifv =

ϕ ifv ϕ fv +1− i xi xi

xiVi , fv

∑x V j

=

j . fv

j

ln γ ires → UNIFAC

xi (Vi − Vwi )

∑ x (V − V j

j

j

wj

(4.6)

)

7248_C004.fm Page 79 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

79

As can been seen from Equation 4.6, the free-v olume definition g ven by Equation 4.2 is employed. The combinatorial term of Equation 4.6 is v ery similar to that of Flory–Huggins. However, instead of v olume or se gment fractions, free-v olume fractions are used. In this w ay, combinatorial and free-volume effects are combined into a single expression. The combinatorial–FV expression of the Entropic-FV model is deri ved from statistical mechanics, using a suitable form of the generalized v an der Waals partition function. The residual term of Entropic-FV is taken by the so-called new or linear UNIFAC model, which uses a linear temperature dependent parameter table: 14

(

amn = amn,1 + amn,2 T − To

)

(4.7)

This parameter table has been developed using the combinatorial term of the original UNIFAC model (which, as mentioned, does not account for free-v olume effects). As with UNIFAC-FV, no parameter reestimation has been performed. The same group parameters are used in linear-UNIFAC and Entropic-FV. Both UNIFAC-FV and Entropic-FV require as input the v olumes of solvents and polymers (at the temperatures of interest). If not available, these can be estimated with the GC methods mentioned previously, e.g., GCV OL. Activity coef ficient calculations with UNI AC-FV and Entropic-FV , especially the former, are rather sensiti ve to the density v alues employed. As already mentioned, the Entropic-FV model has been derived from the van der Waals partition RT - – ----afunction. The similarity of the model with the v an der Waals equation of state P = -----------V – b V2 becomes apparent if the latter is written (when the classical Van der Waals one fluid mixing an classical combining rules are used) as an acti vity coefficient model ln γ i = ln γ icomb− fv + ln γ ires ⎛ ϕ fv 2 ϕ fv ⎞ ⎛ V ⎞ = ⎜ ln i + 1 − i ⎟ + ⎜ i ( δ i − δ j ) ϕ 2j ⎟ ⎠ xi ⎠ ⎝ RT ⎝ xi ϕ ifv =

(4.8)

xi (Vi − bi )

∑ x (V − b ) j

j

j

j

δi =

ai Vi

ϕi is the v olume fraction as defined later in Equation 4.10 The first term in Equation 4.8 is the same as in Entropic-FV with Vw = b, whereas the latter term is a re gular solution theory-type term. Various efforts in impro ving Entropic-FV ha ve been published, focusing especially on its combinatorial-FV term; they are reviewed elsewhere.6,15 For example, Kouskoumvekaki et al.16 suggested using Equation 4.6 with V* = 1.2V w, which yields better results for athermal polymer solutions, compared to the assumption V* = Vw adopted in the original Entropic-FV model. This is in agreement with what is stated pre viously, i.e., a co volume being higher than the v an der Waals volume. Entropic-FV has been extensively applied to various types of phase equilibria (VLE, LLE, and SLE) of polymer–solv ent and polymer–polymer (blends) systems as well as solutions including dendrimers, mixed solvents, copolymers, and paint-related polymers. It is considered one of the state-of-the-art models in the field

7248_C004.fm Page 80 Thursday, May 10, 2007 12:51 PM

80

Hansen Solubility Parameters: A User’s Handbook

THE FLORY–HUGGINS MODEL

AND THE

REGULAR SOLUTION THEORY

The Flory–Huggins (FH) model is the most well-kno wn approach for polymer solutions and can, for binary systems, be e xpressed under some assumptions as follo ws for the solv ent acti vity coefficient ln γ 1 = ln

ϕ1 ϕ + 1 − 1 + χ12 ϕ 22 x1 x1

(4.9)

ϕ ⎛ 1⎞ = ln 1 + ⎜ 1 − ⎟ ϕ 2 + χ12 ϕ 22 x1 ⎝ r⎠

The most important assumption in deri ving Equation 4.9 is that the Flory–Huggins interaction parameter ( χ12) is independent of composition (see Appendix 4.A.1). The parameter r is the ratio of the polymer volume to the solvent volume V2/V1 (approximately equal to the de gree of polymerization), and the v olume fraction is defined as xiVi xiVi + x jV j

ϕi =

(4.10)

The first term of Equation 4.9 is due to combinatorial e fects and is deri ved from the lattice theory, whereas the second rather empirical (v an Laar -type) ener getic term includes the only adjustable parameter of the model, the so-called FH interaction parameter χ12. The FH theory can be extended to multicomponent systems (see Appendix 4.A.1) but (at least) one χ12-value is required per binary . Moreover, unfortunately, the FH parameter is typically not a constant and should be estimated from e xperimental data. Usually it v aries with both temperature and concentration, which renders the FH model useful basically for correlating e xperimental data. Accurate representation of miscibility curv es with the FH model is possible using rather comple x equations for the temperature and the concentration-dependence of the FH-parameters: ΔG = RT

∑ x ln ϕ + g i

12

i

ϕ1ϕ 2

i

⎞ ⎛ c g12 = B(ϕ)C (T ) = a + bϕ 2 ⎜ 1 + + dT + e ln T ⎟ T ⎝ ⎠

(

)

(4.11)

Although we could use equations like Equation 4.11, it should be mentioned that the adjustable parameters of such equations (a,b,c,d,e) ha ve no apparent ph ysical significance; th y cannot be generalized and are specific for each polyme -solvent system. F or practical applications, it often suffices to use Equation 4.9 with a composition-independent Flory–Huggins interaction paramete . Even in this w ay, the FH model cannot be used for predictions unless a predicti ve scheme for the FH parameter is a vailable. Such a predicti ve scheme can be based on a solubility parameter , either the Hildebrand or the Hansen. Due to the similarity of the v an Laar term with the re gular solution theory (see Equation 4.8), we can relate the FH parameter with the solubility parameters. This is an approximate approach, but in some cases a reasonable v alue of the FH parameter can be obtained, using the follo wing equation: χ12 = χs + χ h = 0.35 +

(

V1 δ1 − δ 2 RT

)

2

(4.12)

7248_C004.fm Page 81 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

81

Equation 4.12, without the empirical 0.35 term, is deri ved from the re gular solution theory (compare Equation 4.8 and Equation 4.9). The constant 0.35 is added for correcting for the deficiencies of the FH combinatorial and residual terms. These deficiencies become vident when comparing experimental data for athermal polymer and other asymmetric solutions to the results obtained with the FH model. A consistent underestimation of the acti vity coef ficient data i observed, which is often attributed to the inability of the FH model to account for the free-v olume differences between polymers and solv ents or between compounds dif fering significantly in siz such as n-alkanes with v ery dif ferent chain lengths. The term, which contains the 0.35 f actor, corrects in an empirical w ay for these free-v olume ef fects. Ho wever, and although satisf actory results are obtained in some cases, we cannot generally recommend using Equation 4.12 for estimating the FH parameter. Moreover, for many nonpolar systems with compounds having similar solubility parameters, the empirical f actor 0.35 should be dropped. Rules of Thumb and Solvent Selection Using the Flory–Huggins Model and Solubility Parameters The FH model and the solubility parameters of fer v arious alternati ve approaches for solv ent selection for polymers, and these rules of thumb are summarized here. A chemical (1) will be good solvent for a specific polymer (2), or in other ords, the tw o compounds will be miscible if one (or more) of the follo wing “rules of thumb” are v alid: 1. Using Hildebrand solubility parameters. If the polymer and the solvent have “similar polar and hydrogen bonding degrees:” ⎛ cal ⎞ ≤ 1.8 ⎜ 3 ⎟ ⎝ cm ⎠

δ1 − δ 2

1/ 2

(4.13)

2. Using Hansen solubility parameters (HSP). If the polymer and the solvent have very different polar and hydrogen bonding degrees:

(

4 δd1 − δd 2

) + (δ 2

p1

− δ p2

) + (δ 2

h1

− δh2

)

2

≤R

(4.14)

where R is the Hansen solubility parameter sphere radius. 3. χ12 ≤ 0.5 (the lo wer the Flory–Huggins parameter v alue, the greater the miscibility or , in other words, the better a solvent is a specific chemical).Values much above 0.5 indicate nonsolvency. 4. Ω1∞ ≤ 6 (the lower the infinite dilution act vity coefficient of the sol ent, the greater the solvency of a chemical).Values of the infinite dilution actvity coefficient ab ve 8 indicate nonsolvency.17 In the intermediate re gion (between 6 and 8), it is dif ficult to conclude i the specific chemical is a sol ent or a nonsolv ent. This latter rule of thumb requires some further e xplanations. The weight-based acti vity coefficient at infinite dilution is defined γ i∞ = lim xi →0 γ i ⎛xγ ⎞ M Ωi∞ = lim wi→0 ⎜ i i ⎟ = γ 1∞ 2 M1 ⎝ wi ⎠

7248_C004.fm Page 82 Thursday, May 10, 2007 12:51 PM

82

Hansen Solubility Parameters: A User’s Handbook

(The latter part of the equation is v alid for a binary solv ent(1)-polymer(2) mixture.) Extensive collections of experimental Ω1∞ data are available;18 otherwise, these can be estimated by thermodynamic models lik e the ones mentioned abo ve (UNIF AC-FV, Entropic-FV , and Flory–Huggins). Thermodynamic models often perform better for this type of calculation rather than for predicting full LLE phase diagrams. However, the results depend not only on the accuracy of the model b ut also on the reliability of the rule of thumb, which in turns depends on the assumptions of the Flory–Huggins approach. A thermodynamically more correct method is to calculate the acti vity–concentration diagram with a thermodynamic model lik e Entropic-FV or UNIFAC-FV; the maximum indicates phase split, whereas a monotonic increase of acti vity with concentration indicates a single liquid phase (homogeneous solution).

ACTIVITY COEFFICIENTS MODELS USING THE HSP FLORY–HUGGINS MODELS USING HILDEBRAND PARAMETERS (HSP)

AND

HANSEN SOLUBILITY

Several of the problems of the Flory–Huggins model are associated with the dif ficulties in predictin the FH interaction parameter and the f act that this parameter depends on both temperature and concentration. Recently, Lindvig et al. 19 proposed an e xtension of the Flory–Huggins equation using the Hansen solubility parameters for estimating activity coefficients of compl x polymer solutions. The expression for the solv ent activity coefficient in a binary sol ent-polymer solution is: ln γ 1 = ln

ϕ1 ϕ + 1 − 1 + χ12 ϕ 22 x1 x1

2 V 2 2 χ12 = α 1 ⎡( δ d1 − δ d 2 ) + 0.25 ( δ p1 − δ p 2 ) + 0.25 ( δ h1 − δ h 2 ) ⎤ ⎦ RT ⎣

(4.15)

This model is hereafter abbre viated as FH/Ha(nsen) or FH/HSP . Lindvig et al. 19 ha ve tested three dif ferent combinatorial e xpressions, i.e., dif ferent w ays of expressing the composition fraction ϕi: 1. Based on v olume fractions (Equation 4.10) 2. FV fractions (Equation 4.6) 3. Segment fractions, the latter being defined via Equation 4.10 volumes used instead of v olumes.

ut with v an der Waals

The universal parameter α has been fitted in each case to a la ge number of polymer -solvent VLE data. In total, 358 data points ha ve been considered for solutions containing acrylates and acetates (PBMA, PMMA, PEMA, PVAc). A minimum does exist for the different types of solutions, as can be seen in Figure 4.1. In particular , the minima for the nonpolar and h ydrogen bonding solvents are v ery close, whereas there is little sensiti vity to the parameter for polar solv ents. This means that a uni versal α value can be established. These are shown for the v arious combinatorial terms in Table 4.1. In all cases, the results are better when the optimum v alue is used than when α is set equal to one or when the term with the Hansen solubility parameters is ignored ( α = 0). The best results are obtained when the v olume-based combinatorial term is used (Equation 4.10) together with α = 0.6 (see also Figure 4.1). Table 4.2 pro vides results for se veral polymer solutions with the FH/Hansen model (FHHa, Equation 4.15) using the volume-based combinatorial with both the optimum parameter and α = 1.

7248_C004.fm Page 83 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

Combinatorial part based on volume fractions

100

PBMA PEMA PMA PMMA PVAc Average

90 80 Average absolute percentage deviation

83

70 60 50 40 30 20 10 0

0

0.1

0.2

0.3

0.4 0.5 0.6 0.7 Value of the correction constant

0.8

0.9

1

FIGURE 4.1 Influence of the α-parameter on the performance of Equation 4.15 for all polymer solutions in the database when the Flory–Huggins part of the model is based on v olume fractions. (From Lindvig, Th., Michelsen, M.L., and Kontogeorgis, G.M., Fluid Phase Equilibria, 203, 247, 2002. Reprinted with permission.)

Results are also sho wn with three dif ferent group-contribution models, the pre viously described Entropic-FV (EFV) and UNIF AC-FV (UFV) acti vity coef ficients and an equation of state, th GC–Flory (GCFl) model by Bogdanic and Fredenslund. 20 The presentation of the results is organized into three categories according to the nature of the solv ents (nonpolar, polar, and hydrogen bonding). Finally, Table 4.3 pro vides an o verall comparison of the FH/Hansen model using all three choices for the combinatorial term and the three GC models mentioned abo ve. Results are sho wn for all systems considered in the database for the estimation of the α-parameter as well as tw o commercial epoxy resins for which acti vity coefficient data are vailable. It can be concluded that: 1. The α-parameter is higher when v olume and especially se gment fractions are used in the combinatorial term. This may be e xpected as entropic ef fects are not accounted for and compensation is required by a higher parameter value. It seems that the HSP account not only for ener getic effects but also for some residual-free v olume contributions. 2. For the FH/Hansen model: in all cases better results are obtained when the optimum αparameter is used compared to α = 1. Moreover, FH/Hansen performs in all cases better than FV/Hildebrand which is the best possible model implementing the Hildebrand parameters.21 The a verage de viations with FV/Hildebrand are: 36% (nonpolar), 24% (polar), and 48% (h ydrogen bonding). The FV/Hildebrand model is gi ven by Equation 4.8 with solubility parameters being the total Hildebrand parameters.

7248_C004.fm Page 84 Thursday, May 10, 2007 12:51 PM

84

Hansen Solubility Parameters: A User’s Handbook

TABLE 4.1 Optimum Values of the α-Parameter with the FH/HSP Model, Equation 4.1 (From Lindvig, Th., Michelsen, M.L., and Kontogeorgis, G.M., Fluid Phase Equilibria, 203, 247, 2002. Reprinted with permission.) Fraction

Non-Polar

Polar

H.B.

Total

Volume

αopt % AAD (αopt) % AAD (α = 0) % AAD (α = 1)

0.55 20 37 40

1.00 23 31 23

0.60 25 54 53

0.60 22 41 40

Segment

αopt % AAD (αopt) % AAD (α = 0) % AAD (α = 1)

0.85 19 47 22

1.00 34 48 34

0.75 28 63 40

0.80 25 51 29

Free volume

αopt % AAD (αopt) % AAD (α = 0) % AAD (α = 1)

0.25 28 31 76

0.05 20 20 33

0.40 22 40 87

0.30 26 31 71

Note: The a verage absolute de viations (AAD) are pro vided using these optimum values as well as when the α-parameter is equal to zero or one. Results are sho wn for all systems a vailable in the database (denoted as “total”) and for the dif ferent types of solv ents. H.B. indicates h ydrogen bonding solvents.

3. The FH/Hansen model is as accurate as the other group-contrib ution models for the systems used in the database for estimating the α-parameter. It is particularly better than the GC models for h ydrogen bonding solv ents. 4. The FH/Hansen model is more accurate than the other models, especially the well-known UNIFAC-FV, for the tw o epoxy resins. 5. The de viations are within the reported e xperimental uncertainty for infinite dilutio activity coefficients, which is typically between 10–20% The FH/Hansen Model vs. the GC Methods There are similarities and dif ferences between the FH/Hansen and the GC methods. The most important similarities are that the y both need as input the density of polymer and solv ents (though not GC–Flory, which is an equation of state) and that the y are formulated as acti vity coefficien models, thus their application is limited to lo w pressures. An advantage of the FH/Hansen model compared to the GC methods is that the xact e knowledge of the structure of the polymers is not needed. The only information required is the HSP and the densities, which are a vailable for many polymers, solvents, and other chemicals, or can be readily estimated. Moreover, the FH/Hansen method does not suffer from the often problematic assignment of groups in the GC methods. As an example, we can mention that three dif ferent definitions h ve been proposed for the “acetate” group of, for e xample, PBMA: CCOO, 21 COOCH 2,12 and COO. 22 Use of different main groups in models lik e EFV or UFV will ha ve different results, and it is not always apparent beforehand which group should be chosen. A difficulty with the FH/Hansen model is that arious values of HSP are sometimes reported for the same polymers in the literature. 19

7248_C004.fm Page 85 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

85

TABLE 4.2 Average Absolute Percentage Deviations (% AAD) between Calculated and Experimental Activity Coefficients for Various Groups of Solvents at Infinite Dilution in Polymers often Used in Paints and Coatings Applications Non-Polar Solvents Polymer

Nsys

NDP

EFV

UFV

GCFl

α=1 FHHa

α = 0.6 FHHa

PBMA PEMA PMA PMMA PVAc Total

11 4 5 5 8 33

60 5 17 17 103 202

22 47 45 40 20 31

22 49 46 25 19 29

11 36 28 26 17 20

65 86 18 19 53 51

18 44 24 38 14 24

Polar Solvents Polymer

Nsys

NDP

EFV

UFV

GCFl

α=1 FHHa

α = 0.6 FHHa

PBMA PEMA PMA PMMA PVAc Total

4 2 2 3 4 15

12 6 6 10 30 64

9 10 35 25 36 23

13 11 36 25 25 21

11 11 25 23 24 19

28 22 37 25 17 25

35 27 22 35 17 25

Hydrogen Bonding Solvents Polymer

Nsys

NDP

EFV

UFV

GCFl

α=1 FHHa

α = 0.6 FHHa

PBMA PEMA PMA PMMA PVAc Total

6 3 2 3 2 16

40 10 4 21 17 92

65 46 16 108 15 57

87 47 34 117 63 26

24 21 9 12 15 18

65 86 18 19 53 53

18 44 24 38 14 25

Note: Nsys and NDP are, respectively, the number of systems and datapoints.

APPLICATIONS Solvent Selection for Paints (Activity Coefficients at Infinite Dilution) Lindvig et al. 19,21 have performed an e valuation of various models for solv ent selection for paints. These included Entropic-FV, UNIFAC-FV, GC-Flory as well as the the FH/Hansen model presented previously and the classical Hansen method (Equation 4.14). The results of these e valuations are summarized in Table 4.4a, whereas selected results are sho wn in Table 4.4b. For the three group-contrib ution models, the solv ent selection is based on the rule of thumb: Ω1∞ ≤ 6 : good solv ent Ω1∞ ≥ 8 : poor solv ent (nonsolvent)

7248_C004.fm Page 86 Thursday, May 10, 2007 12:51 PM

86

Hansen Solubility Parameters: A User’s Handbook

TABLE 4.3 Average Absolute Percentage Deviations (% AAD) between Experimental and Calculated Activity Coefficients for PaintRelated Polymer Solutions Using the Flory-Huggins/Hansen Method and Three Group Contribution Models Model

%AAD (systems in database)

%AAD Araldit 488

%AAD Eponol-55

FH/Hansen volume fractions FH/Hansen segment fractions FH/Hansen FV fractions Entropic-FV UNIFAC-FV GC-Flory

22 25 26 35 39 18

31 — — 34 119 29

28 — — 30 62 37

Note: The second column represents the systems used for optimization of the uni versal parameter (solutions containing acrylates and acetates). The last tw o columns sho w predictions for tw o epoxy resins. The density of the epoxies is estimated using the GCVOL method. Source: Adapted from Lindvig, Th. et al., Fluid Phase Equilibria, 203, 247, 2002.

TABLE 4.4A Validity of the Solubility Answers Obtained from Five Methods for Solvent Screening in Various Polymer-Solvent Systems Model

Correct Answers

Incorrect Answers

No Answer

No Calculation

FH/HSP (Equation 4.16) Original Hansen (Equation 4.14) FV/Hildebrand (Equation 4.17) Entropic-FV UNIFAC-FV GC-Flory

102 99 86 91 78 72

20 23 23 19 21 26

— — 13 19 17 14

7 7 7 0 13 17

Source: Adapted from Lindvig, Th. et al., Fluid Phase Equilibria, 203, 247, 2002; Lindvig, Th, et al., Thermodynamics of paint-related systems with engineering models, AIChE J., 47(11), 2573, 2001.

No answer is obtained when the infinite dilution act vity coefficient alue is between 6 and 8. For the FH/Hansen and FV/Hildebrand models, the solvent selection is based on whether the value of the FH parameter is belo w or abo ve 0.5 as e xplained pre viously. All the FH/Hansen results shown in this section are with the “best” combination, i.e., using the v olume fraction-based combinatorial and α = 0.6. The FH/Hansen and FV/Hildebrand models are summarized as follo ws:

7248_C004.fm Page 87 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

87

TABLE 4.4B Prediction of the Solubility for Characteristic Polymer-Solvent Systems Using Various Rules of Thumb and Models for Solvent Selection System

Experiment

PBMA/nC10 PBMA/xylene PBMA/CHCl3 PBMA/acetone PBMA/ethyl acetate PBMA/ethanol PMMA/acetone PMMA/ethyl acetate PMMA/butanol PEMA/MEK PEMA/diethyl ether PEMA/nitropropane PVAc/hexane PVAc/methanol PVAc/ethanol PVAc/nitromethane PVAc/THF

NS S S S S NS S S NS S S NS NS S NS S S

EFV

UFV

FH/HSP

6.5 ( – ) 2.3 (S) 1.9 (S) 0.2 (NS) 6.7 ( – ) 29.2 (NS) 10.0 (NS) 6.6 (– ) 26.8 (NS) 8.1 (NS) 5.8 (S) 4.5 (S) 38.7 (NS) 18.9 (NS) 15.2 (NS) 3.9 (S) 8.4 (NS)

6.1 ( – ) 3.6 (S) 9.1 (NS) 14.1 (NS) 6.7 ( – ) 31.3 (NS) 16.5 (NS) 8.4 (NS) 14.4 (NS) 11.7 (NS) 7.6 ( – ) 1.4 (S) 38.6 (NS) 19.4 (NS) 38.9 (NS) 3.8 (S) 5.6 (S)

1.20 (NS) 0.41 (S) 0.14 (S) 0.2 (S) 0.27 (S) 1.01 (NS) 0.18 (S) 0.36 (S) 0.67 (NS) 0.09 (S) 0.57 (NS) 0.11 (S) 1.09 (NS) 0.71 (NS) 0.63 (NS) 0.43 (S) 0.05 (S)

Note: The values in the table for EFV and UFV indicate infinite dilution act vity coefficient while those of FH/HSP are the Flory-Huggins parameters estimated based on Equation 4.16. S = good solv ent, NS = non solv ent, - = no answer according to the rule of thumb . Source: Adapted from Lindvig, Th. et al., Fluid Phase Equilibria, 203, 247, 2002.

FH–Hansen (FH/HSP) ln γ 1 = ln

ϕ1 ϕ + 1 − 1 + χ12 ϕ 22 x1 x1

χ12 = 0.6

2 V1 ⎡ (δ d1 − δ d 2 )2 + 0.25 (δ p1 − δ p 2 ) + 0.25 (δ h1 − δ h 2 )2 ⎤⎦ RT ⎣

ϕi =

(4.16)

xiVi xiVi + x jV j

FV–Hildebrand ln γ 1 = ln ϕ ifv =

xiVi , fv

∑ j

χ12 =

ϕ ifv ϕ fv + 1 − i + χ12 ϕ 22 xi xi

x jV j . fv

=

xi (Vi − Vwi )

∑

V1 ( δ1 − δ 2 ) 2 RT

j

x j (V j − Vwj )

(4.17)

7248_C004.fm Page 88 Thursday, May 10, 2007 12:51 PM

88

Hansen Solubility Parameters: A User’s Handbook

The basic conclusions are: 1. The FH/Hansen method is, on a verage, as good as the original method of Hansen and the three GC methods 2. The FH/Hansen method is better than the “best” combination based on the Hildebrand parameters, i.e., the FV/Hildebrand method 3. For most polymers (PBMA, PMMA, PEMA) the GC methods have problems for ketonecontaining solutions, where only the FH/Hansen method performs well 4. Most models ha ve problems with nitrocompounds and PEMA 5. The behavior of the v arious models for PVAC is rather mix ed and peculiar , and unlik e the other polymers, each model has dif ferent strengths and weaknesses. Mixed Solvent–Polymer Phase Equilibria Lindvig et al. 23 extended the applicability of v arious models to mix ed solvent-polymer VLE and typical results are presented in Table 4.6 divided according to the experimental source. Only a few experimental data are a vailable for such multicomponent systems, and the accurac y of the data may in some cases be doubtful. Two fundamentally different modelling approaches ha ve been tested: 1. Purely predicti ve GC models for which calculations can be made without re gressing parameters from binary data for the systems considered (EFV , UFV, FH/Ha, GC-Flory, and GCLF). These models are the f astest and simplest tools for thermodynamic calculations. All calculations are based on e xisting parameters (typically group-based ones). 2. Molecular models using binary molecular interaction parameters estimated from e xperimental data for the corresponding binary systems. This is a more time-consuming approach than the former one b ut is still a predicti ve one for ternary mixtures in the sense that no multicomponent data are used for the re gression of the model parameters. All interaction parameters ha ve been fitted to binary xperimental data as e xplained by Lindvig et al.23 All correlative models contain a single binary parameter except EFV/UNIQUAC, which has tw o. In addition to this distinction (purely predicti ve and molecular models), the models tested for multicomponent systems ha ve similarities and dif ferences, and their v arious characteristics are summarized in Table 4.5.

TABLE 4.5 Presentation of the Characteristics of the Various Models Tested for Mixed Solvent–Polymer Phase Equilibria Model

Correlative

SAFT EFV/UQ FH Pa-Ve EFV/UN UFV GCFl GCLF FH/Ha

X X X X

Fully Predictive

Activity Coefficient Model

Equation of State

Group-Contribution

X X X X X X X X

X X X X

Partially X X X X

X X X Partially

7248_C004.fm Page 89 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

89

TABLE 4.6 Average Absolute Logarithmic Percentage Deviations between Experimental and Predicted Equilibrium Pressures and Average Absolute Deviation (X100) between Calculated and Experimental Vapor-Phase Compositions (Mole Fractions) for Various Ternary Polymer-Mixed Solvent Systems Sys. No.

Variable

SAFT

EFV/U

FH

Pa-Ve

EFV/U

UFV

GCFI

GCLF

FHHa

1

P y

13 28

16 30

15 29

16 29

16 32

36 32

16 28

15 27

17 21

2

P y

7 14

4 13

12 14

7 14

8 8

14 8

149 6

14 23

7 13

3

P y

– –

– –

– –

– –

30 17

37 18

16 8

22 5

50 5

4

P y

– –

– –

– –

– –

72 4

57 5

52 4

32 4

20 4

Note: 1. PMMA-butanone-toluene at 308 K. 2. PS-benzene-toluene at 308 K. 3. PMMA-b utanone-acetone at 308 K. 4. PMMA-benzene-toluene at 308 K. Source of experimental data: Liu et al. (2002), Fluid Phase Equilibria, 2002, 194–197: 1067–1075]. Sys. No.

Variable

SAFT

EFV/U

FH

Pa-Ve

EFV/U

UFV

GCFI

GCLF

FHHa

1

P y

11 4

6 3

6 3

6 3

2 3

1 3

21 3

8 3

2 3

2

P y

4 5

2 2

14 5

8 1

2 2

2 2

– –

2 4

11 5

3

P y

– –

– –

– –

– –

3 3

3 3

12 2

5 4

18 4

4

P y

– –

– –

– –

– –

4 18

4 18

13 19

9 19

5 15

Note: 1. PS-toluene-ethylbenzene at 303 K. 2. PS-toluene-cyclohexane at 303 K. 3. PVAc-acetone-ethyl acetate at 303 K. 4. PV Ac-acetone-methanol at 303 K. Source of experimental data: Katayama et al. (1971) [Kag aku K ogaku, 1971, 35: 1012]; Matsumara and Katayama (1974), Kagaku Kogaku, 1974, 38: 388. Sys. No. 1

Variable

SAFT

EFV/U

FH

Pa-Ve

EFV/U

UFV

GCFI

GCLF

FHHa

P y

14 17

16 17

11 16

4 11

17 17

19 13

93 2

5 18

52 14

Source of experimental data: Tanbonliong and Prausnitz (1997), Polymer, 38: 5775; PS-chloroform-carbon tetrachloride at 323.15 K. Adapted from Lindvig et al. 23

Besides the acti vity coefficient models mentioned pr viously (Entropic-FV, UNIFAC-FV, and FH/Hansen) and the GC-Flory equation of state, three adv anced equations of state are considered: the GCLF by Lee and Danner ,24 SAFT by Chapman and co workers,25 and P anayiotou-Vera.26 The basic conclusions are:

7248_C004.fm Page 90 Thursday, May 10, 2007 12:51 PM

90

Hansen Solubility Parameters: A User’s Handbook

1. The FH/Hansen is as successful as the GC methods and other more comple x models, both the correlative and the predicti ve. 2. The models perform sometimes dif ferently for isolated systems, b ut the o verall differences are minor , and we belie ve that the y fall within the e xperimental uncertainties of the data considered. In many cases the results with the various models are closer to each other than to the e xperimental data. 3. For 9 out of the 13 systems considered, good predictions of both the equilibrium pressures and the vapor phase compositions are obtained by all models. 4. The correlative models do not of fer any improvements over the group-contribution models. This is a surprising result as it is e xpected that the molecular information in a model should lead to a better representation of phase equilibria for multicomponent mixtures. It may be that some experimental data for such multicomponent mixtures are of lo w accuracy. 5. This investigation does not point to a “clear winner” among the models, and more data are required for further in vestigations. However, this preliminary study pro vides confi dence for use of the FH/Hansen approach.

CONCLUSIONS AND FUTURE CHALLENGES Many successful calculations in polymer thermodynamics can be carried out using simple groupcontribution methods based on UNIFAC, which contain corrections for the FV effects. Models like Entropic-FV and UNIF AC-FV can be used for such calculations and are sho wn to satisf actorily predict the solvent activities and vapor–liquid equilibria for binary and ternary polymer solutions. Such methods require accurate v alues of the densities and, moreo ver, are based on the a vailability of group parameters in the UNIF AC tables. An alternative equally successful approach is offered by the combination of the Flory–Huggins model with the Hansen solubility parameters (HSP). The FH/HSP (Equation 4.16) includes a single universal parameter that has been re gressed to e xperimental data for man y polymer solv ent solutions. The combinatorial term that gi ves the best results is based on v olume fractions. The FH/HSP model is sho wn to be as successful as the state of the art GC models (EntropicFV, UNIFAC-FV, and GC-Flory) in predicting infinite dilution act vity coefficients including com plex epoxy polymers, solv ent selection for paints, and VLE for mix ed solvent–polymer systems. Moreover, FH/HSP is as successful for mix ed solvent–polymer phase equilibria as comple x, theoretically-based equations of state lik e SAFT and the Group Contrib ution Lattice Fluid. This chapter has been limited to v apor–liquid equlibria (both at finite concentrations and infini dilution), mixed solvents, and solv ent selection. The methods can be, in principle, e xtended to polymer–solvent LLE, and this has been indeed done, for e xample, for Entropic-FV, GC-Flory, GCLF, and SAFT. The predictive group contribution methods are less successful for the prediction of liquid–liquid equilibria as their parameters are based on VLE. Some results are summarized in the recent literature.6,15 Better results are obtained when a molecular local composition model is used as, for example, in the Entropic-FV/UNIQUAC model as shown by Pappa et al.27 As such models include the liquid volume as input parameter, successful results have been reported even for high pressure LLE. However, solvent selection can be based on the use of infinite dilution act vity coefficient for which these models are quite successful. So far most acti vity coefficient models, including FH/HS , have been applied mainly to or ganic polymer solutions of rather simple structures andVLE/solvent selection studies, although some complex paint-related polymers ha ve been considered as well. Thorlaksen et al. 28 have recently combined the Entropic-FV term with Hildebrand's re gular solution theory and de veloped a model for estimating g as solubilities in elastomers. A similar approach can be adopted for the FH/HSP model presented here. Future developments can include complex structures such as dendrimers (where EFV and UFV already ha ve been applied 29); star -like, h yperbranched polymers as well as v arious copolymers;

7248_C004.fm Page 91 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

91

liquid–liquid and solid–liquid equilibria, including the ef fect of crystallinity; and cross-linking, inorganic polymers and polyelectrolytes.

LIST OF ABBREVIATIONS % AAD Comb Comb-FV CST Exper./exp. EFV EFV/UN EFV/UQ FH FH/HSP FV GC GC-F(lory) GCLF GCVOL HB HSP LCST LLE MW NS Pa-Ve PBMA PDMS PEMA PMMA Pred. PVAC Res S SAFT SLE UCST U-FV UNIFAC UQ vdW vdW1f VLE VOC

Average percentage absolute de viation Combinatorial Combinatorial-free volume Critical solution temperature Experimental Entropic-FV (Same as EFV) Entropic-FV (using UNIF AC for the residual term) Entropic-FV using UNIQ UAC as the residual term Flory–Huggins (model/equation/interaction parameter) The FH model using the Hansen solubility parameters, Equation 4.16 Free-volume Group contribution (method/principle) Group contribution Flory equation of state Group contribution lattice flui Group contribution volume (method for estimating the density) Hydrogen bonding Hansen solubility parameters Lower critical solution temperature Liquid–liquid equilibria Molecular weight Nonsolvent/nonsoluble Panayiotou–Vera equation of state Polybutyl methacrylate Polydimethylsiloxane Polyethyl methacrylate Polymethyl methacrylate Predicted Polyvinyl acetate Residual Solvent/soluble Statistical associating fluid theor Solid–liquid equilibria Upper critical solution temperature UNIFAC-FV Universal functional activity coefficient UNIQUAC Van der Waals equation of state Van der Waals one fluid (mixing rules Vapor–liquid equilibria Volatile organic content

7248_C004.fm Page 92 Thursday, May 10, 2007 12:51 PM

92

Hansen Solubility Parameters: A User’s Handbook

SYMBOLS IN THIS CHAPTER G R R T a amn b ni r x v a b c F Fi V Vf V* VW δ φ γ χ12 ΩI∞

Gibbs energy Radius of Hansen solubility sphere Gas constant (in connection with T) Temperature Constant in van der Waals equation of state Coefficient defined in Equation 4 Constant in van der Waals equation of state Number of given groups of type “i” in molecule Ratio of polymer v olume to solv ent volume Mole fraction Reduced volumes in Equation 4.4 and Equation 4.5 Constant in Equation 4.15 Constant in Equation 4.5 Degree of freedom in Equation 4.4 Property in Equation 4.1 Group value for given property in Equation 4.1 Total volume Free volume (Equation 4.2) Hard core or close pack ed volume in Equation 4.2 van der Waals volume Solubility parameter (as in rest of handbook) Volume fraction (see also Equation 4.10) Activity coefficient in Chapter Flory–Huggins interaction parameter Infinite dilution act vity coefficien

APPENDIX 4.I: AN EXPRESSION OF THE FLORY-HUGGINS MODEL FOR MULTICOMPONENT MIXTURES The Flory–Huggins model w as originally de veloped as a model for the entrop y of mixing for mixtures containing molecules of dif ferent size, b ut it w as soon modified also to account fo energetic interactions. The model can be formulated in terms of the excess Gibbs energy as follows (Lindvig et al. 23): G E = G E,comb + G E,res G E,comb = RT G E,res = RT

N

∑ n ln φx

i

i

i

i =1 N

N

i =1

j =1

∑∑φ φ a i

χij = 2 aij νi

j ij

7248_C004.fm Page 93 Thursday, May 10, 2007 12:51 PM

The Hansen Solubility Parameters (HSP) in Thermodynamic Models for Polymer Solutions

93

Using basic thermodynamics, the follo wing expression for the acti vity coefficient is obtained ln γ i = ln γ comb + ln γ res i i where the combinatorial term is gi ven by: ln γ icomb = ln

ϕi ϕ +1− i xi xi

and the residual term is: NC

= 2 vi ln γ res i

∑ j =1

NC

ϕ j aij − vi

NC

∑∑ϕ ϕ a j

k

jk

j =1 k =1

The abo ve formulation of the FH model is slightly dif ferent from the con ventionally used formulation using the Flory–Huggins interaction parameter (χ12), although there is an interrelationship based on the simple equation sho wn above. For a binary mixture, the multicomponent equation reduces to the traditional FH residual term: ln γ 1res = χ12 ϕ 22

REFERENCES 1. Fredenslund, Aa., Jones, R.L., and Prausnitz, J.M., Group contrib ution estimation of acti vity coeffi cients in nonideal liquid mixtures, AIChE J., 25(1), 1086–1098, 1975. 2. Van Krevelen, D.W., Properties of polymers, Their correlation with chemical structure; their numerical estimation and prediction from additive group contributions, Elsevier, 1990. 3. Elbro, H.S., Fredenslund, Aa., and Rasmussen, P ., Group contrib ution method for the prediction of liquid densities as a function of temperature for solv ents, oligomers, and polymers , Ind. Eng. Chem. Res., 30, 2576, 1991. 4. Tsibanogiannis, I.N., Kalospiros, N.S., and Tassios, D.P., Extension of the GCV OL method and application to some comple x compounds, Ind. Eng. Chem. Res., 33, 1641, 1994. 5. Ihmels, E.C. and Gmehling, J., Extension and re vision of the group contribution method GCVOL for the prediction of pure compound liquid densities , Ind. Eng. Chem. Res., 42(2), 408–412, 2003. 6. Kontogeorgis, G.M., Thermodynamics of polymer solutions, in Handbook of Surface and Colloid Chemistry, 2nd ed., Birdi, K.S., Ed., CRC Press, Boca Raton, FL, 2003, chap.16. 7. Bogdanic, G. and Fredenslund, Aa., Prediction of VLE for mixtures with co-polymers, Ind. Eng. Chem. Res., 34, 324, 1965. 8. Elbro, H.S., Phase Equilibria of Polymer Solutions — with Special Emphasis on Free Volumes, Ph.D thesis, Department of Chemical Engineering, Technical University of Denmark, 1992. 9. Bondi, A., Physical Properties of Molecular Crystals, Liquids and Glasses, John Wiley & Sons, New York, 1968. 10. Elbro, H.S., Fredenslund, Aa., and Rasmussen, P., A new simple equation for the prediction of solvent activities in polymer solutions, Macromolecules, 23, 4707, 1990. 11. Kontogeorgis, G.M., Fredenslund, Aa., and Tassios, D.P., Simple acti vity coefficient model for th prediction of solv ent activities in polymer solutions, Ind. Eng. Chem. Res., 32, 362, 1993. 12. Oishi, T. and Prausnitz, M., Estimation of solv ent acti vities in polymer solutions using a groupcontribution method, Ind. Eng. Chem. Process Des. Dev., 17(3), 333, 1978. 13. Fried, J.R., Jiang, J.S., and Yeh, E., Comput. Polym. Sci., 2, 95, 1992.

7248_C004.fm Page 94 Thursday, May 10, 2007 12:51 PM

94

Hansen Solubility Parameters: A User’s Handbook 14. Hansen, H.K., Coto, B., and K uhlmann, B., UNIF AC with lineary temperature-dependent groupinteraction parameters, IVC-SEP Internal Report 9212, 1992. 15. Kontogeorgis, G.M., Models for polymer solutions, in Computer Aided Property Estimation for Process and Product Design, Kontogeorgis, G.M. and Gaani, R., Eds., Else vier, 2004, chap. 7. 16. Kouskoumvekaki, I., Michelsen, M.L., and Kontogeorgis, G.M., Fluid Phase Equilibria, 202(2), 325, 2002. 17. Holten-Andersen, J. and Eng, K., Activity coef ficients in polymer solutions, Progress in Organic Coatings, 16, 77, 1988. 18. High, M.S. and Danner , R.P., Polymer Solution Handbook; DIPPR 881 Project, Design Institute for Physical Property Data, 1992. 19. Lindvig, Th., Michelsen, M.L., and K ontogeorgis, G.M., Fluid Phase Equilibria, 203, 247, 2002. 20. Bogdanic, G. and Fredenslund, Aa., Ind. Eng. Chem. Res., 34, 324, 1995. 21. Lindvig, Th., Michelsen, M.L., and K ontogeorgis, G.M., Thermodynamics of paint-related systems with engineering models, AIChE J., 47(11), 2573, 2001. 22. Lee, B.C. and Danner , R.P., Prediction of infinite dilution act vity coefficients in polymer solutions comparison of prediction models, Fluid Phase Equilibria, 128, 97, 1997. 23. Lindvig, Th., Economou, I.G., Danner, R.P., Michelsen, M.L., and K ontogeorgis, G.M., Modeling of multicomponent v apor-liquid equilibria for polymer -solvent systems, Fluid Phase Equilibria, 220, 11–20, 2004. 24. Lee, B.C. and Danner , R.P., Prediction of polymer -solvent phase equilibria by a modified group contribution EoS, AIChE, 42, 837, 1996. 25. Chapman, W.G. et al., Ne w reference equation of state for associating fluids, Ind. Eng. Chem. Res., 29, 1709–1721, 1990. 26. Panayiotou, C. and Vera, J.H., An improved lattice-fluid equation of state for pure component poly meric fluids, Polym. Eng. Sci., 22, 345, 1982. 27. Pappa, G.D., Voutsas, E.C., and Tassios, D.P., Liquid-liquid phase equilibrium in polymer -solvent systems: correlation and prediction of the polymer molecular weight and the pressure effect, Ind. Eng. Chem. Res., 40(21), 4654, 2001. 28. Thorlaksen, P., Abildskov, J., and K ontogeorgis, G.M., Prediction of g as solubilities in elastomeric polymers for the design of thermopane windo ws, Fluid Phase Equilibria, 211, 17, 2003. 29. Kouskoumvekaki, I., Giesen, R., Michelsen, M.L., and K ontogeorgis, G.M., Free-v olume acti vity coefficient models for dendrimer solutions, Ind. Eng. Chem. Res., 41, 4848, 2002.

7248_C005.fm Page 95 Wednesday, May 23, 2007 10:51 AM

of Characterization 5 Methods — Polymers Charles M. Hansen ABSTRACT The simplest e xperimental method to determine the Hansen solubility parameters (HSP) for a polymer is to e valuate whether or not it dissolv es in selected solv ents. Those solvents dissolving the polymer will ha ve HSP closer to those of the polymer than those solv ents that do not. A computer program or graphical method can then be used to find the HSP for the polyme . Other types of evaluations can also lead to polymer HSP. These include swelling, melting point reduction, surface attack, chemical resistance, barrier properties, viscosity measurements, and an y other measurement reflecting di ferences in polymer af finities among the sol ents. Polymer HSP can be higher than the HSP of an y of the test solv ents. This means that some of the methods suggested in the literature to interpret data, i.e., those which use a verages of solv ent HSP to arrive at the polymer HSP , must be used with care.

INTRODUCTION Experience has sho wn that if it is at all possible, an e xperimental evaluation of the beha vior of a polymer in contact with a series of selected liquids is the best way to arrive at its HSP. Experimental data can be generated and treated in v arious ways to arrive at the v alues of interest. Examples are included in the follo wing. The author’s usual approach to generate data in solubility parameter studies is to contact a polymer of interest with 40 to 45 well-chosen liquids. One may then observ e or measure a number of different phenomena including full solution at a given concentration, degree of swelling by visual observation or by measurement of weight change, v olume change, clarity, surface attack, etc. The object of the studies is to determine differences in affinity of the polymer for the di ferent solvents. These differences are then traditionally used to di vide the solv ents into tw o groups, one which is considered “good” and the other which is considered “bad. ” Such data can be entered into the SPHERE program as discussed in Chapter 1. Whenever possible, the author uses a set of solv ents as described belo w, often supplemented by selected solv ents depending on the purpose of the investigation. Supplementary test solv ents are usually in the boundary re gions as it is these that determine the parameters of the sphere. Adding more good solvents well within the sphere or more bad solvents well outside of it will not change an ything but the data fit The goal of the e xperimental work is to arrive at a set of data sho wing differences in behavior among the test solvents. These data are then processed to arrive at the four parameters characteristic of HSP correlations, three describing the nonpolar , polar, and h ydrogen-bonding interactions for the liquids and the fourth, Ro, a radius of interaction for the type of interaction described. The author has most often used computer techniques to e valuate the data to find the polyme HSP. In earlier w ork simple plots were used. A simple plot of δP vs. δH is also helpful in man y practical situations to get guidance as discussed in Chapter 8. The approximate determination of polymer HSP can be done with three plots of experimental data using the HSP parameters pairwise. Figure 5.1 to Figure 5.3 demonstrate how this was attempted initially.1 The spheroids in the figure including the δD parameter gave problems. Hansen and Skaarup 2 simply used a scaling f actor of 2 (the coefficient “4” in Chapter 1, Equation 1.9) to produce spheres in all three plots. As Ro must 95

7248_C005.fm Page 96 Wednesday, May 23, 2007 10:51 AM

96

Hansen Solubility Parameters: A User’s Handbook

12

10

8 δp R

6 (P,H) 4

2

2

4

6

8 δh

10

12

14

FIGURE 5.1 Two-dimensional plot of δP vs. δH for the solubility of polymeth yl methacrylate (Polymer B in Table 5.2). The circle is the projection of a sphere on the gi ven coordinates. Units are (cal/cm 3)1/2. (From Hansen, C.M., Färg och Lack, 17(4), 71, 1971. With permission.)

be the same in all of these plots, a single compass setting is tried for a set of δD, δP , and δH to see how well the separation into good and bad solv ents is accomplished. Calculations for points in doubt can be made using Chapter 1, Equation 1.9. Plots with the modified δD axis are gi ven for the solubility of polystyrene3 shown in Figure 5.4 to Figure 5.6. These are the original figures fro this thesis, and the numbers refer to a table of solv ents found there. An idea of the accurac y of the graphical approach can be found in T able 5.1, where comparisons are made between the “hand” method and results of the SPHERE program. T able 5.2 contains a listing of the polymers included in Table 5.1. Specific solubility data are g ven for these polymers in 88 solv ents in Appendix A.3. Teas4 has developed a triangular plotting technique which helps visualization of three parameters on a plain sheet of paper. Examples are found in Reference 5 to Reference 7 and in Chapter 8. The triangular plotting technique uses parameters for the solv ents, which, in f act, are modified HS parameters. The individual Hansen parameters are normalized by the sum of the three parameters. This gives three fractional parameters defined by Equation 5.1 to Equation 5.3 fd = 100 δD/(δD + δP + δH)

(5.1)

fP = 100 δP/(δD + δP + δH)

(5.2)

fh = 100 δH/(δD + δP + δH)

(5.3)

The sum of these three fractional parameters is 1.0. This allows the use of the special triangular technique. Some accuracy is lost, and there is no theoretical justification for this plotting technique

7248_C005.fm Page 97 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

97

12

10

8 δp 6

4

2

2

4

6

8

10

12

14

δh FIGURE 5.2 Two-dimensional plot of δH vs. δD for the solubility of polymeth yl methacrylate (Polymer B in Table 5.2). Expansion of the δD scale by a f actor of 2 w ould yield a circle (a sphere in projection). Units are (cal/cm3)1/2. (From Hansen, C.M., Färg och Lack, 17(4), 71, 1971. With permission.)

but one does get all three parameters onto a tw o-dimensional plot. This plotting technique is often used by those who conserv e old paintings, because it w as described in a standard reference book very shortly after it w as developed.7 Figure 8.4 sho ws how such a plot can be used in finding suitable solvent when dealing with such an older oil painting. HSP for the polymers and film formers discussed in the foll wing examples are given in Table 5.3. These data are based on solubility determinations unless otherwise noted. Barton 6,8 has also provided solubility parameters for many polymers. Values for a number of acrylic, epoxy, and other polymers potentially useful in self-stratifying coatings ha ve been reported by Benjamin et al. 9 (see Chapter 8). Rasmussen and Wahlström10 pro vide additional HSP data in relation to the use of replenishable natural products (oils) in connection with solv ents. The data processing techniques and data accumulated by Zellers and co workers11–14 on elastomers used in chemical protecti ve clothing are also useful. Zellers et al. also point out man y of the problems encountered with these characterizations. Such problems are also discussed belo w. There are other sources of HSP for polymers in the literature, but a full review of these and their uses is beyond the scope of this book.

CALCULATION OF POLYMER HSP Calculation of the HSP for polymers is also possible. The results are not yet fully satisf actory, but there is hope for the future. One of the more significant e forts in this has been made by Utracki and coworkers.15,16 They assumed the δD parameter for polymers did not dif fer too much between polymers and interpreted e valuations of polymer–polymer compatibility using calculated v alues for δP and δH. A word of caution is advisable here and that is that the constant “4” in Equation 1.9 is v ery often if not most often significant, and should not be replaced with a “1 ” either. Group

7248_C005.fm Page 98 Wednesday, May 23, 2007 10:51 AM

98

Hansen Solubility Parameters: A User’s Handbook

12

10

8 δp 6

4

2

2

4

6

8 δd

10

12

14

FIGURE 5.3 Two-dimensional plot of δP vs. δD for the solubility of polymeth yl methacrylate (Polymer B in Table 5.2). Expansion of the δD scale by a f actor of 2 w ould yield a circle (a sphere in projection). Units are (cal/cm3)1/2. (From Hansen, C.M., Färg och Lack, 17(4), 72, 1971. With permission.)

calculations were used. This is probably the best calculation approach currently a vailable, b ut improvements are thought possible. See Chapter 3. The group contributions given in Chapter 1 can be used for this purpose, although the estimated dispersion parameters are thought to be too lo w. It is suggested that HSP for polymers determined by these calculations not be mix ed with e xperimentally determined HSP until confirmation of agreement is found. It can be presumed that th errors involved in either process will cancel internally , but these may not necessarily be the same for the calculated results as for the e xperimental ones. The author has never been particularly successful in calculating the same values as were found experimentally, although a serious ef fort to use weighting and similar f actors, as discussed in the following, has never been tried.

SOLUBILITY — EXAMPLES The most direct method to determine the three HSP for polymers or other soluble materials is to evaluate their solubility or de gree of swelling/uptak e in a series of well-defined sol ents. The solvents should ha ve different HSP chosen for systematic e xploration of the three parameters at all levels. As indicated earlier, a starting point could be the series of liquids used by the author for many years. These are essentially those included in Table 5.4. Sometimes boundaries are define better by inclusion of additional test solv ents. A computer analysis quickly gives a choice of many of these, as solv ents with RED numbers (Chapter 1, Equation 1.10) near 1.0 are located near the sphere boundary. It is actually the boundary which is used to define the center point of the spher using Chapter 1, Equation 1.9. Some changes are also possible to remove or replace solvents which are now considered too hazardous, although good laboratory practice should allo w use of the ones

7248_C005.fm Page 99 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

99

16

14

SOLUBLE NOT SOLUBLE STRONG INTERACTION

12

δp

10 28

8

45

6

43

38A

69

29 RA

4

2

32 23

0 δd 0

2

21 19 19A 15A 18A 18 42 15 67A 89 25

4

6

8 δh

10

12

14

16

FIGURE 5.4 Two-dimensional plot δP vs. δH of solubility data for polystyrene (Polymer G in Units are (cal/cm 3)1/2.

Table 5.2).

indicated. The HSP generally in use for liquids ha ve all been e valuated/calculated at 25°C. These same values can also be used to correlate ph ysical phenomena related to solubility at other test temperatures with some care, as noted in the follo wing. Several examples of HSP correlations based on solubility are found in Table 5.3. The entry for polyethersulfone (PES) found in Table 5.3 w as determined from data included in the computer output reported in Table 5.4. The solubility of PES w as evaluated in 41 dif ferent solvents. It w as found that fi e of them actually dissolv ed the polymer . The input data to the SPHERE program described in Chapter 1 are included in Table 5.4 in the SOLUB column. A “1” means a good solvent and a “0” means a bad solv ent. A 1* means that a good solv ent lies outside the sphere, where it should not, and a 0* means a bad solv ent lies inside the sphere, which means it is an outlier. Each of these error situations reduces the data fit. D, , H, and R for the solubility of PES are given at the top. In addition, there is an indication of the data fit, which is 0.999 here. A perfect fit is 1.000. A data fit slightly less than 1.0 is actually preferred, as the computer program has the optimized the data to a single set of v alues that are so close to being correct as the y can be within experimental error. An unknown number of sets of the parameters can give a data fit of 1.0 when ver this result is found. Perfect fits are rather easily obtained with small sets of data, and the boundarie are rather poorly defined, which means the center is also poorly defined. One can continue testi with additional solv ents located in the boundary re gions of the established sphere as stated pre viously. These can be found easily by listing the solvents in order of their RED numbers and choosing

7248_C005.fm Page 100 Wednesday, May 23, 2007 10:51 AM

100

Hansen Solubility Parameters: A User’s Handbook

16

14 SOLUBLE NOT SOLUBLE STRONG INTERACTION

12

δh

10

8 15

9

6

67A21 19A 69 15B 35 25 29

4

45 42 47A 28

23

RA

2

56

32 45

0 δp

5

6

7

26A

8 63 48 9

10

11

12

δd

FIGURE 5.5 Two-dimensional plot δH vs. δD of solubility data for polystyrene (Polymer G in Table 5.2). Expansion of the δD scale by a factor of 2 has given a spherical representation according to Chapter 1, Equation 1.9. Units are (cal/cm 3)1/2.

those with v alues not too dif ferent from 1.0. The RED number is gi ven for each solv ent in the RED column. A quality number, Q = 1 – RED, is also conceptually useful. Finally, there is a column in Table 5.4 indicating the molar volume, V, of the solvents in cc/mol. There was no need to analyze the influence of this parameter in the present case A second example of this type of approach is given in Table 5.3. Data on good and bad solvents17 for polyacrylonitrile (PAN) have been used as input to the computer program. There are 13 solvents indicated as good, and 23 indicated as bad. These test solv ents do not dif fer as widely from each other as the test series suggested earlier , but the data are still useful in finding the HSP for thi polymer. These are reported in Table 5.3. The data fit of 0.931 is good for this kind of data. H ving found the HSP for a polymer in this manner, one can then search a database for additional solvents for the polymer in question. This was done for the HSP database with o ver 800 solvent entries in Table A.1 of the first edition of this handbook. A significantly la ge number of the 123 additional solvents found to ha ve RED numbers less than 1.0 can be e xpected to dissolv e this polymer , but such an e xtensive experimental study w as not undertak en to confirm the predictions A special problem that can be encountered is when only a few solvents with very high solubility parameters dissolve a polymer. An example is polyvinyl alcohol with true solvents being 1-propanol and ethanol in a data set with 56 solv ents.6 The entry in Table 5.3 places a big question mark o ver the solubility parameters, as well as with the radius 4.0 and the perfect fit of the data.The computer analysis quickly encompasses the tw o good solv ents in the data set within a small sphere as the y

7248_C005.fm Page 101 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

101

16

14 SOLUBLE NOT SOLUBLE STRONG INTERACTION

12

δp

10 27

8

40 16 38A

44 26B

28

45 43

6

69 29

66

21 58

4

RA 8

67A

2

9

56

32 8

23 25

0 δh

5

6

7

26A

8

9

10

11

12

δd

FIGURE 5.6 Two-dimensional plot δP vs. δD of solubility data for polystyrene (Polymer G in Table 5.2). Expansion of the δD scale by a factor of 2 has given a spherical representation according to Chapter 1, Equation 1.9. Units are (cal/cm 3)1/2.

have reasonably similar parameters. Based on reasonable similarity with other solubility correlations for w ater-soluble polymers, one anticipates spheres with a radius much lar ger than the distance between these solv ents. This result is not good and should not be used. Another e xample of determining HSP for a polymer with v ery high solubility parameters is Dextran C (British Drug Houses). Only 5 out of 50 solv ents were found to dissolv e Dextran C. 18 In this case, there w as enough spread in the solubility parameters of the test solv ents such that the spherical model correlation (Chapter 1, Equation 1.9) forced the program to find a radius of 17. MPa1/2. This appears to be a reasonable number for this situation. The problem can be made clearer by noting the dissolving solv ents with their RED numbers in parentheses. These were dimeth yl sulfoxide (1.000), ethanolamine (0.880), eth ylene glycol (0.880), formamide (0.915), and glycerol (0.991). Some dissolving liquids had RED equal to 1.0 or higher and included dieth ylene glycol (1.000), propylene glycol (1.053), and 1,3-butanediol (1.054). These helped to define the boundar of the Hansen solubility sphere. Note that the HSP for the polymer are in a re gion higher than that defined by the alues of test liquids. Any technique using an average of the HSP for the test solvents will inherently underestimate the solute HSP in such a situation. The solubility data for the polymer Dextran C led to the HSP data reported in Table 5.3 when the SPHERE program used a starting point based on a verages of the HSP v alues for the good solv ents. When the starting point w as 25 MPa1/2 for D, P, H, and Ro, respecti vely, a perfect data fit as found for D, P, H, and Ro equal to 26, 26, 26, and 24, all in MP a1/2. When the starting point w as for D, P, H, and Ro equal to 30, 30,

7248_C005.fm Page 102 Wednesday, May 23, 2007 10:51 AM

102

Hansen Solubility Parameters: A User’s Handbook

TABLE 5.1 Calculated vs. Trial-and-Error Solubility Parameter Data for Various Polymersa Computed Handtrials A B C D E F G H I J K L M N O P Q

(First Values) (Second Values) δD δP δH

Ro

FIT

8.60 9.2 9.11 9.2 9.95 8.5 9.98 8.5 9.79 9.4 9.09 8.5 10.40 8.6 10.23 9.3 10.17 9.5 7.53 7.0 9.90 9.3 9.08 9.5 11.37 9.0 9.65 9.4 10.62 8.9 8.58 8.5 9.87

5.20 5.3 4.20 4.0 6.20 4.7 6.70 5.3 5.70 5.0 5.20 4.8 6.20 3.5 6.70 4.9 6.20 4.7 5.60 5.5 5.20 4.2 3.70 4.5 9.70 6.4 6.20 5.5 7.70 4.5 5.20 5.0 5.70

0.960 0.923 0.945 0.923 0.853 0.829 0.974 0.957 0.930 0.929 0.948 0.871 0.955 0.915 0.891 0.855 0.924 0.909 0.933 0.918 0.949 0.933 0.921 0.896 0.978 0.923 0.897 0.867 1.000 0.952 0.942 0.940 0.942

4.72 5.3 5.14 5.0 5.88 5.5 1.68 2.5 2.84 3.2 2.13 4.3 2.81 3.0 5.51 5.0 4.05 4.0 7.20 7.0 3.09 3.7 6.22 6.0 3.20 4.0 5.68 5.3 0.46 3.0 4.58 4.7 6.43

1.94 2.1 3.67 4.2 5.61 5.5 2.23 3.0 5.34 5.1 6.37 5.5 2.10 2.0 4.72 4.0 7.31 6.4 4.32 4.3 2.64 2.1 5.38 6.0 4.08 5.5 7.13 7.4 4.17 3.8 7.00 6.5 6.39

Computed Handtrials R S T U V X Y Z A B C D E F G L

(First Values) (Second Values) δD δP 9.04 9.2 10.53 8.8 8.58 8.7 9.10 9.3 8.10 8.5 7.10 7.8 8.57 8.8 8.52 8.2 9.60 8.7 9.95 9.5 8.05 8.5 10.34 9.2 8.58 8.5 8.91 9.4 9.49 8.8 9.86 10.8 9.3

4.50 4.5 7.30 7.0 1.64 1.8 4.29 4.5 0.69 1.5 1.23 1.0 1.10 2.5 –0.94 0.8 2.31 2.5 4.17 4.0 0.18 1.0 6.63 5.8 0.58 1.5 3.68 4.5 2.68 2.7 7.14 7.0 6.2

δH

Ro

FIT

2.40 2.6 6.00 5.5 1.32 1.8 2.04 2.0 –0.40 1.5 2.28 3.6 1.67 1.2 7.28 5.7 3.80 3.5 5.20 5.5 1.39 2.0 6.26 4.2 1.76 1.8 4.08 3.5 2.82 2.7 7.35 8.8 4.7

5.20 5.0 8.20 6.0 3.20 3.5 4.70 4.7 4.70 3.4 6.20 4.0 3.20 3.8 4.70 2.9 5.20 4.2 7.20 7.0 4.20 3.4 6.70 5.0 3.20 2.6 1.70 3.2 4.70 4.0 5.70 7.1 4.2

0.985 0.972 0.910 0.879 0.974 0.965 0.969 0.950 0.974 0.964 0.921 0.881 0.950 0.914 0.971 0.954 0.942 0.951 0.980 0.976 0.966 0.960 0.964 0.868 0.968 0.956 0.992 0.895 0.961 0.963 0.970 0.936 0.892

Note: Units are (cal/cm 3)1/2. a

See Table 5.2 for polymer types.

Source: From Hansen, C.M., Färg och Lack, 17(4), 73, 1971. With permission.

30, and 30, all in MP a1/2, a perfect correlation w as found to D, P , H, and Ro equal to 30, 28, 28, and 32, all in MPa1/2. These data show that extrapolations into regions where there are no data can be problematic. It is thought that the data gi ven in Table 5.3 for Dectran C are the most representative, because of the data fit being slightly less than 1.0 g ving a better definition of a boundar . The properties of good solv ents alone cannot al ways lead to a good estimate of the solubility parameters for these polymers, and the radii of spheres using only a fe w solvents with high solubility parameters will be very uncertain. One can sometimes find better results by correlating d grees of swelling or uptak e, rather than correlate on solubility or not. The work of Zellers and co workers

7248_C005.fm Page 103 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

103

TABLE 5.2 List of Polymers and Resins Studied A B C D E F G H I J K L M N O P Q R S T U V X Y Z A B C D E F G L

Lucite® 2042-poly (eth yl methacrylate), E. I. du Pont de Nemours & Co., Inc. Poly (methyl methacrylate), Rohm and Haas Co. Epikote® 1001-epoxy, Shell Chemical Co. Plexal P65-66% oil length alk yd, Polyplex. Pentalyn® 830-alcohol soluble rosin resin, Hercules Incorporated. Butvar® B76-poly (vin yl butyral), Shawinigan Resins Co. Polystyrene LG, Badische Anilin- und Soda F abrik. Mowilith® 50-poly (vin yl acetate), F arbwerke Hoechst. Plastopal H-urea formaldeh yde resin, Badische Anilin- und Soda F abrik. H Sec. Nitrocellulose-H 23, A. Hagedorn and Co. Parlon® P10-chlorinated poly (prop ylene), Hercules Incorporated. Cellulose acetate, Cellidora A-Bayer AG. Super Beckacite ® 1001-Pure Phenolic Resin, Reichhold Chemicals Co. Phenodur 373U-phenol-resol resin, Chemische Werke Albert. Cellolyn 102-modified pentaerythritol ester of rosin, Hercules Incorporated Pentalyn 255-alcohol soluble resin, Hercules Incorporated. Suprasec F5100-blocked isocyanate (phenol), Imperial Chemical Ind. Ltd. Plexal C34-34% coconut oil-phthalic anh ydride alkyd, Polyplex. Desmophen 850, Polyester -Farbenfabriken Bayer AG. Polysar 5630 — styrene-b utadiene (SBR) ra w elastomer, Polymer Corp. Hycar® 1052-acrylonitrile-butadiene raw elastomer, B. F. Goodrich Chemical Corp. Carifl x IR 305-isoprene ra w elastomer, Shell Chemical Co. Lutanol IC/123-poly (isob utylene), Badische Anilin- und Soda F abrik. Buna Huls CB 10-cis poly b utadiene raw elastomer, Chemische Werke Huels. Versamid® 930-polyamide, General Mills, Inc. Ester gum BL, Hercules Incorporated. Cymel® 300-hexamethoxy melamine, American Cyanamid Co. Piccolyte® S100-terpene resin, Pennsylv ania Industrial Chemical Corp. Durez® 14383-furfuryl alcohol resin, Hook er Chemical Co. Piccopale® 110-petroleum h ydrocarbon resin, Pennsylv ania Industrial Chemical Corp. Vipla KR-poly (vin yl chloride), K = 50, Montecatini. Piccoumarone 450L-cumarone-indene resin, Pennsylv ania Industrial Chemical Corp. Milled wood lignin — special sample from Prof. A. Björkman.

TABLE 5.3 Hansen Solubility Parameter Correlations for Selected Materials Material

δD

δP

δH

Ro

FIT

G/T

PES solubility PAN solubility PP swelling Polyvinyl alcohol ? (see te xt) Hexamethylphosphoramide PVDC melting temperature 110°C PVDC melting temperature 130°C Dextran C solubility

19.6 21.7 18.0 17.0 18.5 17.6 20.4 24.3

10.8 14.1 3.0 9.0 8.6 9.1 10.0 19.9

9.2 9.1 3.0 18.0 11.3 7.8 10.2 22.5

6.2 10.9 8.0 4.0 — 3.9 7.6 17.4

0.999 0.931 1.00 1.00 — 0.992 0.826 0.999

5/41 13/36 13/21 2/56 — 6/24 13/24 5/50

Note: Units are (cal/cm 3)1/2.

7248_C005.fm Page 104 Wednesday, May 23, 2007 10:51 AM

104

Hansen Solubility Parameters: A User’s Handbook

TABLE 5.4A Calculated Solubility SPHERE for PES Solubility D = 19.6 P = 10.8 H = 9.2 RAD = 6.2 FIT = 0.999 NO = 41 Solvent

D

P

H

SOLUB

RED

V

Acetone Acetophenone Benzene 1-Butanol Butyl acetate γ-Butyrolactone Carbon tetrachloride Chlorobenzene Chloroform Cyclohexanol Diacetone alcohol o-Dichlorobenzene Diethylene glycol Diethyl ether Dimethyl formamide Dimethyl sulfoxide 1,4-Dioxane Ethanol Ethanolamine Ethyl acetate Ethylene dichloride Ethylene glycol Ethylene glycol monob utyl ether Ethylene glycol monoethyl ether Ethylene glycol monomethyl ether Formamide Hexane Isophorone Methanol Methylene dichloride Methyl ethyl k etone Methyl isobutyl ketone Methyl-2-pyrrolidone Nitroethane Nitromethane 2-Nitropropane Propylene carbonate Propylene glycol Tetrahydrofuran Toluene Trichloroethylene

15.5 19.6 18.4 16.0 15.8 19.0 17.8 19.0 17.8 17.4 15.8 19.2 16.6 14.5 17.4 18.4 19.0 15.8 17.0 15.8 19.0 17.0 16.0 16.2 16.2 17.2 14.9 16.6 15.1 18.2 16.0 15.3 18.0 16.0 15.8 16.2 20.0 16.8 16.8 18.0 18.0

10.4 8.6 0.0 5.7 3.7 16.6 0.0 4.3 3.1 4.1 8.2 6.3 12.0 2.9 13.7 16.4 1.8 8.8 15.5 5.3 7.4 11.0 5.1 9.2 9.2 26.2 0.0 8.2 12.3 6.3 9.0 6.1 12.3 15.5 18.8 12.1 18.0 9.4 5.7 1.4 3.1

7.0 3.7 2.0 15.8 6.3 7.4 0.6 2.0 5.7 13.5 10.8 3.3 20.7 5.1 11.3 10.2 7.4 19.4 21.2 7.2 4.1 26.0 12.3 14.3 16.4 19.0 0.0 7.4 22.3 6.1 5.1 4.1 7.2 4.5 5.1 4.1 4.1 23.3 8.0 2.0 5.3

0 1 0 0 0 1 0 0 0 0 0 0 0 0 1 0*a 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0

1.371 0.955 2.129 1.777 1.741 0.998 2.301 1.576 1.483 1.467 1.321 1.204 2.101 2.183 0.915 0.996 1.493 2.077 2.241 1.547 1.007 2.837 1.563 1.395 1.618 3.044 2.745 1.094 2.575 0.990 1.368 1.782 0.655 1.580 1.899 1.387 1.429 2.457 1.237 1.978 1.485

74.0 117.4 89.4 91.5 132.5 76.8 97.1 102.1 80.7 106.0 124.2 112.8 94.9 104.8 77.0 71.3 85.7 58.5 59.8 98.5 79.4 55.8 131.6 97.8 79.1 39.8 131.6 150.5 40.7 63.9 90.1 125.8 96.5 71.5 54.3 86.9 85.0 73.6 81.7 106.8 90.2

Note: Units are MP a1/2. a

Outlier (a bad solv ent lying inside sphere).

7248_C005.fm Page 105 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

105

ALTERNATE TABLE 5.4B Calculated Solubility SPHERE for PES Solubility (Listed in RED Order) D = 19.6 P = 10.8 H = 9.2 RAD = 6.2 FIT = 0.999 NO = 41 Solvent

D

P

H

SOLUB

RED

V

Methyl-2-pyrrolidone Dimethyl formamide Acetophenone Methylene dichloride Dimethyl sulfoxide γ-Butyrolactone Ethylene dichloride Isophorone o-Dichlorobenzene Tetrahydrofuran Diacetone alcohol Methyl ethyl k etone Acetone 2-Nitropropane Ethylene glycol monoethyl ether Propylene carbonate Cyclohexanol Chloroform Trichloroethylene 1,4-Dioxane Ethyl acetate Ethylene glycol monob utyl ether Chlorobenzene Nitroethane Ethylene glycol monomethyl ether Butyl acetate 1-Butanol Methyl isobutyl ketone Nitromethane Toluene Ethanol Diethylene glycol Benzene Diethyl ether Ethanolamine Carbon tetrachloride Propylene glycol Methanol Hexane Ethylene glycol Formamide

18.0 17.4 19.6 18.2 18.4 19.0 19.0 16.6 19.2 16.8 15.8 16.0 15.5 16.2 16.2 20.0 17.4 17.8 18.0 19.0 15.8 16.0 19.0 16.0 16.2 15.8 16.0 15.3 15.8 18.0 15.8 16.6 18.4 14.5 17.0 17.8 16.8 15.1 14.9 17.0 17.2

12.3 13.7 8.6 6.3 16.4 16.6 7.4 8.2 6.3 5.7 8.2 9.0 10.4 12.1 9.2 18.0 4.1 3.1 3.1 1.8 5.3 5.1 4.3 15.5 9.2 3.7 5.7 6.1 18.8 1.4 8.8 12.0 0.0 2.9 15.5 0.0 9.4 12.3 0.0 11.0 26.2

7.2 11.3 3.7 6.1 10.2 7.4 4.1 7.4 3.3 8.0 10.8 5.1 7.0 4.1 14.3 4.1 13.5 5.7 5.3 7.4 7.2 12.3 2.0 4.5 16.4 6.3 15.8 4.1 5.1 2.0 19.4 20.7 2.0 5.1 21.2 0.6 23.3 22.3 0.0 26.0 19.0

1 1 1 1 0*a 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0.655 0.915 0.955 0.990 0.996 0.998 1.007 1.094 1.204 1.237 1.321 1.368 1.371 1.387 1.395 1.429 1.467 1.483 1.485 1.493 1.547 1.563 1.576 1.580 1.618 1.741 1.777 1.782 1.899 1.978 2.077 2.101 2.129 2.183 2.241 2.301 2.457 2.575 2.745 2.837 3.044

96.5 77.0 117.4 63.9 71.3 76.8 79.4 150.5 112.8 81.7 124.2 90.1 74.0 86.9 97.8 85.0 106.0 80.7 90.2 85.7 98.5 131.6 102.1 71.5 79.1 132.5 91.5 125.8 54.3 106.8 58.5 94.9 89.4 104.8 59.8 97.1 73.6 40.7 131.6 55.8 39.8

Note: Units are MP a1/2. a

Outlier (a bad solv ent lying inside sphere).

7248_C005.fm Page 106 Wednesday, May 23, 2007 10:51 AM

106

Hansen Solubility Parameters: A User’s Handbook

reports extensive studies of this type. 11–14 It should be noted, ho wever, that the HSP-sphere parameters usually v ary some from correlation to correlation based on the same data when dif ferent criteria are used for good and bad solv ents. This is because the absorbed solv ent tends to locate in regions with similar solubility parameters, and there are local variations in HSP within most, if not all, polymers. This is particularly true of polymers which are not homopolymers. This situation relates to self-assembly . Solvents or se gments of molecules with similar HSP will tend to reside near each other . An e xample of this is w ater residing at local h ydrophilic sites, such as alcohol groups, in polymers. Utilization of the HSP af finity between molecules or s gments of molecules is a viable w ay to control self-assembly . See also Chapter 18.

SWELLING — EXAMPLES The correlation for swelling of polypropylene reported in Table 5.3 is based on solvent uptake data reported by Lieberman and Barbe at 22°C. 19 The limit of 0.5% w as arbitrarily set to dif ferentiate good solv ents from bad ones. As mentioned earlier , e xperience has sho wn that a dif ferent limit usually gives different parameters. It should be noted that swelling data reflect the properties o the regions in the polymer where the solv ent has chosen to reside because of ener getic similarity (self-assembly). The principle is not necessarily “lik e dissolves like,” but rather “lik e seeks lik e.” If the solv ent is homogeneously distrib uted in the polymer , the solubility parameters found will reflect the properties of the whole polyme . Crystalline re gions will not contain solv ent. If the solvent collects locally in re gions with chemical groups dif ferent from the b ulk of the polymer , then the HSP so deri ved will reflect at least partially the p ysical nature of these chemical groups. The parameters reported in Table 5.3 seem appropriate for what is expected in terms of low polarity and low hydrogen-bonding properties for a polyprop ylene-type polymer. An example of a characterization using swelling data that did not result in a good correlation is that for Viton® (The Du Pont Compan y, Wilmington, DE). This problem has been discussed by Zellers and Zhang 11,12 and is also discussed in Chapter 13. If one tries to force-fit data where ther are several different comonomers into a single HSP sphere, the result is usually reflected in a poo correlation coef ficient. Figure 13.3 sh ws that impro vements can be made by using a separate sphere for each comonomer. One reason for the poor correlation of swelling beha vior is that Viton is not a homopolymer , and also contains a cross-linking chemical. The dif ferent se gments ha ve different af finities. Indeed, there are s veral qualities of Viton, each of which has significantl differing chemical resistance. Swelling of Viton has also been treated by Ev ans and Hardy 20 in connection with predictions related to chemical protecti ve clothing, and by Nielsen and Hansen, 21 who presented curv es of swelling as a function of the RED number .

MELTING POINT DETERMINATIONS — EFFECT OF TEMPERATURE Partly crystalline polymers that are placed in dif ferent liquids will ha ve melting points which are lowered to a de gree depending on the solv ent quality of the indi vidual liquids. The melting points of polyvin ylidine chloride (PVDC) ha ve been measured in dif ferent solv ents.22 These data ha ve been analyzed by e valuating solubility parameter re gions based on those solv ents which dissolv e the polymer at 110°C and above and also at 130°C and above. As expected, there are more solvents which dissolve the semicrystalline polymer at the higher temperature. The results for these correlations are included in Table 5.3. The main reasons for the somewhat lower data fit at 130°C includ two nondissolving solv ents within the solubility parameter sphere. These are dimeth yl phthalate, where the lar ge molecular size is a f actor, and benzyl alcohol, where temperature ef fects can be larger than e xpected compared with the other solv ents as discussed later and in Chapter 1. The solubility parameters for PVDC at this temperature, based on tab ulated solvent values at 25°C, are not affected significantly by this type of situation. A single room temperature solv ent for PVDC

7248_C005.fm Page 107 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

107

is reported by Wessling.22 This is hexamethylphosphoramide and its solubility parameters are also reported in Table 5.3 for comparison.The change in the values of the individual solubility parameters with temperature is discussed in Chapter 1 (Equation 1.16 to Equation 1.18). Chapter 3 also treats the temperature dependence of the HSP. See also Chapter 10 where the HSP of specifically carbo dioxide are treated in depth as a function of temperature and pressure.

ENVIRONMENTAL STRESS CRACKING Environmental stress cracking (ESC) is unfortunately a v ery frequent mode of f ailure for plastics. For this reason a whole chapter is de voted to the topic (see Chapter 14). It has been possible to correlate HSP with ESC phenomena, and this can also provide an estimate of the HSP for the given polymers. Care is advised since the stress/strain le vel is important, as is the molecular size and shape of the chemicals in volved. Several collections of ESC data in the older literature 23–25 should not be forgotten in these days of “Internet and only Internet.” Such collections have particular value as it is considered impossible to get a commercially a vailable polymer without some additi ves. These can also affect ESC behavior. These older data were the basis of the ESC correlations gi ven in Chapter 14.

INTRINSIC VISCOSITY MEASUREMENTS One of the more promising methods to e valuate polymer HSP for limited data is that using the intrinsic viscosity. Van Dyk et al. found a correlation with the intrinsic viscosity of an acrylic polymer (polyethyl methacrylate) in v arious solvents and the polymer HSP 26 (see the discussion on polymer compatibility in Chapter 8). Segarceanu and Leca 27 have devised a method to calculate the polymer HSP from data on its intrinsic viscosity in different solvents. The intrinsic viscosities will be higher in the better solvents because of greater interaction and greater polymer chain e xtension. The intrinsic viscosity gi ves an indication of the solv ent quality. It has been used earlier to calculate the Flory–Huggins chi parameter, for e xample.28 In the ne w technique, the intrinsic viscosities are normalized by the intrinsic viscosity of that solvent giving the highest v alue. These normalized data (numbers are 1.0 or less) are then used in a weighted a veraging technique to arri ve at the center of the Hansen sphere. δD2 = Σ(δDi × [η]i)/Σ[η]i

(5.4)

δP2 = Σ(δPi × [η]i)/Σ[η]i

(5.5)

δH2 = Σ(δHi × [η]i]/Σ[η]i

(5.6)

The subscript 2 is for the polymer , and the respecti ve solv ents are indicated by an “i. ” The intrinsic viscosity in the i-th solv ent is given by [ η]i. Those solvents with the greatest weighting factor have higher intrinsic viscosities and are closest to the geometric center of the sphere. Those solv ents which do not dissolv e the polymer were assumed to have a zero weighting factor. The HSP for a polyesterimide were reported as an example. HSP values were assigned both by the “classical” e valuation and with this ne wer approach. These data are included as the first entries in Table 5.5. This is a v ery promising method of arri ving at the polymer HSP with limited data. There are se veral aspects of this w ork which deserv e comment. It w as demonstrated earlier that many polymers ha ve higher solubility parameters than an y of the solv ents which are or can be used to test them. The present method only allows for polymer HSP within the range attainable

7248_C005.fm Page 108 Wednesday, May 23, 2007 10:51 AM

108

Hansen Solubility Parameters: A User’s Handbook

TABLE 5.5 HSP Data for the Same Polyesterimide Polymer Based on Data Given in Reference 27 Correlation

δD

δP

δD

Ro

FIT

Classicala1 Newa1 HSP SPHERE a HSP SPHERE b Classicala Classicalb Newa Newb

17.4 18.0 20.0 19.0

12.3 11.1 11.0 11.0

8.6 8.8 10.0 9.0

4.1 8.6 8.3 7.0

— — 1.000 1.000 0.426 0.447 0.506 0.364

Note: Units are MP a1/2. Indicates use of the solubility parameters for the solvents given in Reference 27. b Indicates use of the solv ent HSP data in the author’ s files a

by the test solv ents. The method will lead to v alues that are too lo w in some cases, including the example with the polyesterimide used as an e xample in Segarceanu and Leca. 27 It is not surprising that the polymer HSP are often higher than solv ent HSP, as the y are in a ph ysical state between that of a liquid and a solid. When the cohesion energy becomes too high, a material is a solid rather than a liquid. Lo w molecular weight solids frequently ha ve HSP some what higher than the HSP of liquids. Man y examples can be gi ven, including urea, eth ylene carbonate, etc. When the data (as soluble or not) for the 11 solvents were processed by the SPHERE computer program, the parameters found were those given by the third set of HSP inTable 5.5. The agreement with the “new” method is acceptable, even though none of the test solv ents have δd as high as that of the polymer . Further inspection sho wed that the solubility parameters used in the study were not in agreement with those published in the latest reference to Hansen listed by Se garceanu and Leca.27 It also appears that the radius of the HSP sphere for the classical determination is in error , being far too low. To further clarify the situation, se veral runs with the SPHERE program were done with the parameters listed in this book, as well as with those listed in the article being discussed. In both cases the data fit is not good for the HSP reported by S garceanu and Leca. 27 In the classical case, the data fit is only 0.426 (1.0 is perfect), and four of the f e good solvents are located outside of the sphere. Only N-methyl-2-pyrrolidone is inside. In the ne w case, the data fit is not much bette , being 0.506. Here, four of the fi e bad solvents are inside the sphere with only one being outside. It has been possible to estimate the polymer parameters within acceptable v ariation, but the radius of the sphere has not been accounted for in a satisf actory manner. Further inspection of the data suggests that morpholine, the solv ent with the highest [ η] that was used to normalize the data, is not as good as might ha ve been e xpected from the intrinsic viscosity data. This can be seen in Table 5.6. The reason for this is unkno wn, but experience has shown that amines often are seen to react with v arious materials in a manner which does not allo w their inclusion in correlations of the type discussed here. To conclude this section, it is noted that a similar weighting technique w as used by Zellers et al.13,14 where the weighted measurements were solv ent uptake by elastomers customarily used to make chemical protecti ve clothing. The same precautions must be tak en in analyzing this type of

7248_C005.fm Page 109 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

109

TABLE 5.6 Calculated Solubility SPHERE for Polyesterimide (Listed in RED Order) D = 19.0 P = 11.0 H = 9.0 RAD = 7.0 FIT = 1.000 NO = 11 Solvent

[η]aN

D

P

H

SOLUB

RED

V

Methyl-2-pyrrolidone Dimethyl formamide Dimethyl sulfoxide γ-Butyrolactone Morpholine Cyclohexanone Diacetone alcohol Acetone Diethylene glycol monomethyl ether Ethylene glycol monoethyl ether Ethylene glycol monoethyl ether acetate

0.970 0.947 0.182 0.689 1.000 0.718 — — — — —

18.0 17.4 18.4 19.0 18.8 17.8 15.8 15.5 16.2 16.2 15.9

12.3 13.7 16.4 16.6 4.9 6.3 8.2 10.4 7.8 9.2 4.7

7.2 11.3 10.2 7.4 9.2 5.1 10.8 7.0 12.6 14.3 10.6

1 1 1 1 1 1 0 0 0 0 0

0.427 0.682 0.809 0.832 0.874 0.937 1.031 1.044 1.055 1.131 1.283

96.5 77.0 71.3 76.8 87.1 104.0 124.2 74.0 118.0 97.8 136.1

Note: Units are MP a1/2. a

Normalized intrinsic viscosity data from Reference 27.

measurement, but as the polymers studied were reasonably nonpolar, some of the solvents had HSP which were higher than those of the polymers studied. Zellers et al. 14 and Athey29 also describe multiple variable statistical analysis techniques to find the HSP of a g ven polymer. Barton’s work6 contains man y literature sources of intrinsic viscosity studies using the solubility parameter for interpretation.

OTHER MEASUREMENT TECHNIQUES There are man y other techniques to dif ferentiate between the beha vior of dif ferent solv ents in contact with a polymer . Man y of these are discussed in the follo wing chapters and will not be treated here. These include permeation measurements, chemical resistance determinations of avrious kinds including ESC, and surface attack, etc. Some of the techniques can be very useful, depending on the polymer in volved. Others will present problems because of the probable influence of othe factors such as solvent molar volume and length of time before attainment of equilibrium. Se veral of these phenomena can be correlated with HSP, but the techniques used in the measurements will present problems in using the data for direct HSP characterization of polymers because other effects are also important.

CONCLUSION HSP for polymers can be e valuated experimentally by correlations of data where a suitably lar ge number of well-chosen solvents are brought into contact with the polymer . The observed behavior which can be correlated includes true solubility, swelling, weight gain, dimensional change, degree of surf ace attack, reduction of melting point, permeation rate, breakthrough time, and tensile strength reduction. Correlations for simple e valuations of chemical resistance of the suitable-or not type and ESC are also possible. In each case, the molecular size of the liquids used can affect the result and should be considered in some w ay. The use of w ater as a test liquid is not recommended for these purposes.

7248_C005.fm Page 110 Wednesday, May 23, 2007 10:51 AM

110

Hansen Solubility Parameters: A User’s Handbook

REFERENCES 1. Hansen, C.M., Solubility in the coatings industry , Färg och Lack, 17(4), 69–77, 1971. 2. Hansen, C.M. and Skaarup, K., The three dimensional solubility parameter — key to paint component affinities III. Independent calculation of the parameter components, J. Paint Technol., 39(511), 511–514, 1967. 3. Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Their Importance in Surf ace Coating F ormulation, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967. 4. Teas, J.P., Graphic analysis of resin solubilities, J. Paint Technol., 40(516), 19–25, 1968. 5. Gardon, J.L. and Teas, J.P., Solubility parameters, in Treatise on Coatings, Vol. 2, Characterization of Coatings: Physical Techniques, P art II, Myers, R.R. and Long, J.S., Eds., Marcel Dekk er, Ne w York, 1976, chap. 8. 6. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press, Boca Raton, FL, 1983; 2nd ed., 1991. 7. Torraca, G., Solubility and Solvents for Conservation Problems, 2nd ed., International Centre for the Study of the Preserv ation and the Restoration of Cultural Property (ICCR OM), Rome, 1978. 8. Barton, A.F.M., Handbook of Polymer-Liquid Interaction Parameters and Solubility Parameters, CRC Press, Boca Raton, FL, 1990. 9. Benjamin, S., Carr , C., and Wallbridge, D.J., Self-stratifying coatings for metallic substrates, Prog. Org. Coat., 28(3), 197–207, 1996. 10. Rasmussen, D. and Wahlström, E., HSP — solubility parameters: a tool for de velopment of ne w products — modelling of the solubility of binders in pure and used solv ents, Surf. Coat. Int., 77(8), 323–333, 1994. 11. Zellers, E.T., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. I. Modeling the solubility of organic solvents in Viton® gloves, J. Appl. Polym. Sci., 50, 513–530, 1993. 12. Zellers, E.T . and Zhang, G.-Z., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. II. Modeling diffusion coefficients, breakthrough times, and steady-state perme ation rates of or ganic solvents in Viton® gloves, J. Appl. Polym. Sci., 50, 531–540, 1993. 13. Zellers, E.T., Anna, D.H., Sulewski, R., and Wei, X., Critical analysis of the graphical determination of Hansen’s solubility parameters for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2069–2080, 1996. 14. Zellers, E.T., Anna, D.H., Sule wski, R., and Wei, X., Impro ved methods for the determination of Hansen’s solubility parameters and the estimation of solv ent uptake for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2081–2096, 1996. 15. Luciani, A., Champagne, M.F., and Utracki, L.A., Interfacial tension in polymer blends. Part 1: Theory, Polym. Networks Blends, 6(1), 41–50, 1996. 16. Luciani, A., Champagne, M.F ., and Utracki, L.A., Interf acial tension in polymer blends. P art 2: Measurements, Polym. Networks Blends, 6(2), 51–62, 1996. 17. Fuchs, O., Solv ents and non-solv ents for polymers, in Polymer Handbook, 3rd ed., Brandrup, J. and Immergut, E.H., Eds., Wiley-Interscience, New York, 1989, p. VII/385. 18. Hansen, C.M., The universality of the solubility parameter , Ind. Eng. Chem. Prod. Res. Dev., 8(1), 2–11, 1969. 19. Lieberman, R.B. and Barbe, P .C., Polypropylene polymers, in Encyclopedia of Polymer Science and Engineering, 2nd ed., Vol. 13, Mark, H.F ., Bikales, N.M., Ov erberger, C.G., Menges, G., and Kroschwitz, J.I., Eds., Wiley-Interscience, New York, 1988, pp. 482–483. 20. Evans, K.M. and Hardy , J.K., Predicting solubility and permeation properties of or ganic solvents in Viton glove material using Hansen’s solubility parameters, J. Appl. Polym. Sci., 93, 2688–2698, 2004. 21. Nielsen, T.B. and Hansen, C.M., Elastomer swelling and Hansen solubility parameters, Polym. Testing, 24, 1054–1061, 2005. 22. Wessling, R.A., The solubility of poly(vinylidine chloride), J. Appl. Polym. Sci., 14, 1531–1545, 1970. 23. Wyzgoski, M.G., The role of solubility in stress cracking of n ylon 6,6, in Macromolecular Solutions, Seymour, R.B. and Stahl, G.A., Eds., Per gamon Press, New York, 1982, pp. 41–60. 24. Mai, Y.-W., Environmental stress cracking of glassy polymers and solubility parameters, J. Mater. Sci., 21, 904–916, 1986.

7248_C005.fm Page 111 Wednesday, May 23, 2007 10:51 AM

Methods of Characterization — Polymers

111

25. White, S.A., Weissman, S.R., and Kambour , R.P., Resistance of a polyetherimide to en vironmental stress crazing and cracking, J. Appl. Polym. Sci., 27, 2675–2682, 1982. 26. Van Dyk, J.W., Frisch, H.L., and Wu, D.T., Solubility, solvency, and solubility parameters, Ind. Eng. Chem. Prod. Res. Dev., 24(3), 473–478, 1985. 27. Segarceanu, O. and Leca, M., Improved method to calculate Hansen solubility parameters of a polymer , Prog. Org. Coat., 31(4), 307–310, 1997. 28. Kok, C.M. and Rudin, A., Prediction of Flory-Huggins interaction parameters from intrinsic viscosities, J. Appl. Polym. Sci., 27, 353–362, 1982. 29. Athey, R.D., Testing coatings: 6. Solubility parameter determination, Eur. Coat. J., 5, 367–372, 1993.

7248_C005.fm Page 112 Wednesday, May 23, 2007 10:51 AM

7248_C006.fm Page 113 Wednesday, May 23, 2007 11:00 AM

of Characterization 6 Methods — Surfaces Charles M. Hansen ABSTRACT Relations between cohesion parameters and surface energy parameters and their practical significanc are discussed. Cohesion parameters (solubility parameters) can be used with full theoretical justifi cation to characterize man y surf aces, including substrates, coatings, plastics, pigment and fille surfaces, etc., in addition to the binder or polymer used in a gi ven product. Important molecular relations between a binder in a coating or adhesi ve and its surroundings then become ob vious. Use of cohesion parameters, i.e., Hansen solubility parameters in a total characterization of surface energy, clearly shows how the single point concepts of the (Zisman) critical surf ace tension and the wetting tension fit into a la ger energy concept. A complete match of surface energies of two surfaces requires that exactly the same liquids (in a lar ger number of well-chosen test liquids) spontaneously spread on both surfaces. The dewetting behavior (wetting tension test) of the liquids must also be the same, in that the same liquids should not retract when applied to the surf aces as films

INTRODUCTION Interfacial free ener gy and adhesion properties result from intermolecular forces. It has been recognized for man y years that molecules interact by (molecular) surf ace to (molecular) surf ace contacts to enable solutions to be formed. 1 As molecular surf ace-to-surface contacts control both solution phenomena and surface phenomena, it is not surprising that various correlations of cohesion parameters and surface phenomena can be found. This idea has been well e xplored and dealt with elsewhere.2 The various treatments and correlations in the literature will not be explicitly dealt with here, other than those directly related to Hansen solubility parameters (HSP). In this chapter , solubility parameters are called cohesion (energy) parameters and refer more specifically to HS . Solubility as such does not necessarily enter into the ener getics of interf acial phenomena, b ut the energy characteristics of surf aces can still be correlated with HSP . This chapter will emphasize methods of surface characterization using HSP. The orientation of adsorbed molecules is a significant added e fect that must also be considered in man y cases. The “like dissolves like” concept is e xtended and applied as “lik e seeks lik e” (self-assembly).

HANSEN SOLUBILITY PARAMETER CORRELATIONS WITH SURFACE TENSION (SURFACE FREE ENERGY) Skaarup w as the first to establish a correlation between liquid sur ace tension and HSP . This correlation with surface tension had been long lost in an internal report to members of the Danish Paint and Printing Ink Research Laboratory in 1967, as well as in an abstract for a presentation to the Nordic Chemical Congress in 1968. 3,4 γ = 0.0688V 1/3[δD2 + k( δP2 + δH2)]

(6.1) 113

7248_C006.fm Page 114 Wednesday, May 23, 2007 11:00 AM

114

Hansen Solubility Parameters: A User’s Handbook

γ is the surface tension, and k is a constant depending on the liquids in volved. This k was reported as 0.8 for se veral homologous series, 0.265 for normal alcohols, and 10.3 for n-alk yl benzenes. Beerbower independently published essentially the same type of correlation in 1971. 5,6 With the exception of aliphatic alcohols and alkali halides, Beerbo wer found γ = 0.0715V 1/3[δD2 + 0.632( δP2 + δH2)]

(6.2)

where γ is the surf ace tension. The constant w as actually found to be 0.7147 in the empirical correlation. The units for the cohesion parameters are (cal/cm 3)1/2, and those of the surf ace tension are dyn/cm in both Equation 6.1 and Equation 6.2. Ho wever, v alues in dyn/cm are numerically equal to those in mN/m. The constant w as separately deri ved as being equal to 0.7152 by a mathematical analysis in which the number of nearest neighbors lost in surf ace formation w as considered, assuming that the molecules tend to occup y the corners of re gular octahedra. The correlations presented by K oenhen and Smolders 7 are also rele vant to estimating surf ace tension from HSP. The author has never explored them in detail, however, so they are not discussed here. It is interesting to note thatδP and δH have the same coefficient in the surace tension correlations. They also ha ve the same coef ficient when solubility is correlated (see Chapter 1, Equation 1.9 o Chapter 2, Equation 2.6). The reason for this is the molecular orientation in the specific interaction derived from permanent dipole–permanent dipole and h ydrogen bonding (electron interchange) interactions. The dispersion or London forces arise because of electrons rotating around a positi ve atomic nucleus. This causes local dipoles and attraction among atoms.This is a completely different type of interaction and requires a different coefficient in the correlations. It is this diference between atomic and molecular interactions that is basic to the entire discussion of similarity between HSP and the Prigogine corresponding states theory in Chapter 2. The finding that the polar and ydrogen bonding (electron interchange) ef fects require the same coef ficient for both ulk and surf ace correlations suggests that the net effects of the (often mentioned) unsymmetrical nature of hydrogen bonding are no dif ferent from the net ef fects occurring with permanent dipole–permanent dipole interactions. The lack of specific consideration that ydrogen bonding is an unsymmetrical inter action led Erbil 8 to state that HSP has limited theoretical justification, for xample. The previous discussion and the contents of Chapter 1 and Chapter 2 clearly indicate that the author is not in full agreement with this viewpoint. In fact, it appears that this book presents massi ve experimental evidence, related both to b ulk and surf ace phenomena, which sho ws that the geometric mean is valid for estimating interactions between dissimilar liquids. This includes dispersion, permanent dipole–permanent dipole, and h ydrogen bonding (electron interchange) interactions.

METHOD TO EVALUATE THE COHESION ENERGY PARAMETERS FOR SURFACES One can determine the cohesion parameters for surf aces by observing whether or not spontaneous spreading is found for a series of widely dif ferent liquids. The liquids used in standard solubility parameter determinations are suggested for this type of surf ace characterization. It is strongly suggested that none of the liquids be a mixture, as this introduces an additional f actor into the evaluations. The liquids in the series often used by the author are indicated in Chapter 5, Table 5.4 or Chapter 7, Table 7.2. Droplets of each of the liquids are applied to the surf ace and one simply observes what happens. If a droplet remains as a droplet, there is an adv ancing contact angle and the cohesion ener gy/surface energy of the liquid is (significantly) higher than that of the sur ace. The contact angle need not necessarily be measured in this simplified procedure, h wever. Contact angles have generally been found to increase for greater differences in cohesion parameters between the surface and liquid 9 (see also Figure 6.5). If spontaneous spreading is found, there is presumed

7248_C006.fm Page 115 Wednesday, May 23, 2007 11:00 AM

Methods of Characterization — Surfaces

115

15 18 *

Wetting Tension – ADV = O

C

19 *

10 24 *

δP

B

20 *

25

A 26

5 27 *

12 *

21 * 14 15 *

22 23

7

17 *

11 *

13 *

3 *

4

2 *

1 *

42

6 5 *

8 16 9

5

10

δH

15

20

FIGURE 6.1 HSP surface characterization of an epoxy surf ace showing regions of spontaneous spreading of applied droplets (A), lack of dewetting of applied films (B), and d wetting of applied films (C). Note that thi characterization may not be valid for all epoxy surfaces. Units are MPa1/2. (From Hansen, C.M. and Wallström, E., J. Adhes., 15[3/4], 281, 1983. With permission.)

to be some “similarity” in the energy properties of the liquid and the surface. The apparent similarity may be misleading. As discussed in greater detail later , the f act of spontaneous spreading for a given liquid does not mean that its HSP are identical with those of the surf ace being tested. If a given liquid does not spontaneously spread, it can be spread mechanically as a film and be obser ed to see whether it retracts. This can be done according to ASTM D 2578-84 or ISO 8296:1987 (E). This test determines whether or not there is a receding contact angle under the gi ven conditions. Figure 6.1 sho ws a complete ener gy description for an epoxy polymer surf ace 10,11 based on the testing procedure described previously. The Hansen polar and hydrogen bonding parameters δP and δH are used to report the data. Further explanation of these parameters themselves can be found in Chapter 1. The circular lines can be considered as portraying portions of HSP spheres, b ut the third Hansen parameter, δD, has not been specifically accounted for in the t o-dimensional figure Figure 6.1 shows two curves that are concave toward the origin. The lower of these divides the test liquids into tw o groups based on spontaneous spreading or not. Belo w the line one finds tha liquids applied as droplets will spontaneously spread. Liquids that are found in the re gion above the upper curve will retract when applied as films. A test method to determine this is found in the ASTM and ISO standards gi ven previously, for e xample, except that one uses a lar ge number of pure liquids instead of the liquid mixtures suggested in the standards. Receding contact angles will generally increase as one progresses to liquids with still higher HSP. Intermediate between the two curves in Figure 6.1 is a re gion where liquids applied as droplets will remain as droplets, whereas liquids applied as films will remain as films This region deserves more attention in future research. The energy properties of these liquids are not as close to those of the surf ace as are the ener gy properties of the liquids that spontaneously spread. Spontaneous spreading is more related to adhesion since such liquids w ant to cover the surface spontaneously. The wetting tension test uses

7248_C006.fm Page 116 Wednesday, May 23, 2007 11:00 AM

116

Hansen Solubility Parameters: A User’s Handbook

an e xternal force to spread the liquids, after which the y may continue to remain as a film. The mobility of the surf ace layer(s) will play a role in the wetting tension test. Hydrophilic se gments can (perhaps) rotate toward a water droplet at some rate, for example, and increase the hydrophilic nature of the surf ace accordingly. This is discussed more in Chapter 18. As mentioned earlier , there is still a problem in simplifying these results for easier use and improved understanding. Hexane, for example, does not dissolv e an epoxy polymer, but in Figure 6.1 it is almost in the middle of the region describing spontaneous spreading of the liquids. Hexane will not contrib ute to a “bite” into an epoxy coating for impro ving intercoat adhesion with a subsequent coating. He xane is within the re gion of spontaneous spreading because it has a lo wer surface energy (surf ace tension) than the epoxy surf ace. Nature reduces the free ener gy level of the surface by requiring he xane to cover the epoxy coating. The result of this is that the center of the normal HSP sphere for describing spontaneous spreading can be assigned sizable ne gative values.11 This is both impractical and impossible. A better method of handling this situation is still desired, and until it is found, one must presumably refer to simple plots or other simple comparisons rather than to refined computer techniques, which are more desirable in most cases. In the mean time, interest will still be focused onto the usual test method(s) for determining surf ace tensions based on the Zisman critical surf ace tension plots (lack of adv ancing contact angle) or by using the ASTM procedure for wetting tension (lack of receding contact angle). The following discussion relates these to the HSP-type characterizations discussed earlier . Additional surface characterization plots for spontaneous spreading and wetting tension using HSP are included in Figure 6.2 for a plasticized polyvin yl chloride (PVC) and in Figure 6.3 for a polyethylene (PE).

A CRITICAL VIEW OF THE CRITICAL SURFACE TENSIONS 12,13 The Zisman critical surf ace tension is determined by measuring the e xtent that af finity is lackin (contact angles) for a surf ace using pure liquids or liquid mixtures in a series. The surface tension of each of the liquids is kno wn. One can then plot cosine of the contact angle vs. liquid surf ace tension and e xtrapolate to the limit where the contact angle is no longer present (see Figure 6.4). Liquids with higher surface tensions than this critical v alue allow measurement of a contact angle, whereas liquids with lo wer surface tensions than the critical v alue will spontaneously spread. The fact that the liquid with a surf ace tension just under the critical v alue spontaneously spreads is often tak en as an indication of high af finit . This is dif ficult to understand and appears to be misunderstanding. The limiting critical surf ace tension 12,13 has v ery little to do with the “best” solvent for the surface. It is more appropriately compared with a v ery poor solvent which can only marginally dissolv e a polymer , for e xample. This is similar to the condition for a RED number equal to 1.0 discussed in Chapter 1 and Chapter 2. Measuring the critical surf ace tension has been and still will be a useful technique to better understand surf aces, but it should be done with the following in mind. Who would determine the solubility parameter for a polymer by the follo wing method? One makes up a series of liquids with dif ferent, known solubility parameters. The polymer dissolves in some of them, and the de gree of swelling of the polymer in question is measured in those liquids which do not dissolve it fully. One subsequently determines the solubility parameter of the polymer by e xtrapolating the de gree of swelling to infinit , which corresponds to total solution. This extrapolation can be done by plotting 1/(de gree of swelling) vs. solv ent composition (solubility parameter). One now focuses attention upon that liquid which (by e xtrapolation) just dissolves the polymer. One assumes that there is no better solv ent than this one and, consequently , assigns the polymer solubility parameters corresponding to those of this boundary solvent. This is exactly what one does when the critical surf ace tension is measured. This method should clearly ne ver be used to determine solubility parameters for polymers. At the same time, it sheds some light onto the true meaning of the critical surf ace tension.

7248_C006.fm Page 117 Wednesday, May 23, 2007 11:00 AM

Methods of Characterization — Surfaces

117

15

18 *

19 *

10 24 * δP

20 *

Wetting Tension

25 * 26 *

12 *

21 *

5

14 15 * *

22 * 27 *

17 *

11 *

23

13 *

3 *

4 *

1 *

C B

6 5 7 *

8

2 *

A –

ADV = O

16 9 10 5

10

δH

15

20

FIGURE 6.2 HSP surface characterization of spontaneous spreading of applied droplets and wetting tension for applied films for plasticized polyvi yl chloride (PVC). Note that these characterizations may not be v alid for all PVC surfaces. Units are MPa1/2 (From Hansen, C.M. and Wallström, E., J. Adhes., 15[3/4], 280, 1983. With permission.)

If we now consider the re gion for spontaneous spreading in Figure 6.1 to Figure 6.3, it can be seen that the critical surf ace tension is a point on its boundary . In practice, one finds di ferent critical surface tensions for the same surf ace depending on which liquids (or liquid mixtures) are used. This is explained by the fact that the cohesion parameter regions of the type shown in Figure 6.1 to Figure 6.3 are not symmetrical around the zero axis. The individual liquid series used to determine the critical surf ace tension will intersect the cohesion parameter spontaneous spreading boundary at dif ferent points. The corresponding total surf ace tension will v ary from intersection to intersection as mentioned earlier . Hansen and Wallström11 compared the critical surf ace tension plotting technique with one where a dif ference in HSP w as used instead of liquid surf ace tension. One arrives at the same general conclusions from both types of plotting techniques.This comparison is made in Figure 6.4 and Figure 6.5.

A CRITICAL VIEW OF THE WETTING TENSION A region larger than that for spontaneous spreading will be found on a δP vs. δH plot when one plots data for those liquids that remain as films (do not break up or contract) when th y are applied as films. This type of experiment measures the wetting tension. Mixtures of formamide and ethylene glycol monoethyl ether are usually used in practice for these measurements according to ASTM

7248_C006.fm Page 118 Wednesday, May 23, 2007 11:00 AM

118

Hansen Solubility Parameters: A User’s Handbook

15

18 *

19 *

10 δP

24 *

20 *

25 * 26 *

B

5

11 *

C 21 * 14 15 * *

22 * 27

12 *

23

13 *

Wetting Tension

3 *

4

2 *

1 *

– ADV = O

6 5

A

7 *

8 16 9 10 5

10

δH

15

20

FIGURE 6.3 HSP surface characterization of spontaneous spreading of applied droplets and wetting tension for applied films for a polyet ylene (PE) surf ace. Note that these characterizations may not be v alid for all PE surfaces. Units are MP a1/2. (From Hansen, C.M. and Wallström, E., J. Adhes., 15[3/4], 279, 1983. With permission.)

D 2578-84 or ISO 8296:1987 (E). One can also use the same liquids suggested earlier for cohesion parameter determinations and mak e a plot lik e that in Figure 6.1. If tw o different surfaces are to have the same wetting tension beha vior, their plots must be the same. The results of the ASTM test are usually stated in terms of the surf ace tension of the liquid or liquid mixture which just stays intact as a film for 2 sec. This simple single point determination corresponds to determining a single point on the boundary of the HSP plot describing wetting tension for all liquids. A single point determination may not al ways be sufficient information an certainly neglects the complete picture possible from HSP considerations. Comments identical in principle to those included in the earlier section, “A Critical View of the Critical Surface Tensions,” on measurement of the critical surf ace tension are also v alid here. It is hoped the reader no w better understands the total ener gy conte xt of the simple ASTM wetting tension measurements.

ADDITIONAL HANSEN SOLUBILITY PARAMETER SURFACE CHARACTERIZATIONS AND COMPARISONS Beerbower14 has reported man y other correlations of surf ace phenomena with HSP . Examples include the work of adhesion on mercury; frictional properties of untreated and treated polyethylene

7248_C006.fm Page 119 Wednesday, May 23, 2007 11:00 AM

Methods of Characterization — Surfaces

119

ADV REC

7

1.0

0.8

3

COS θ

2

0.6 1

0.4

0.2

20

30

40

γ mN.m-1

50

60

FIGURE 6.4 Zisman critical surface tension plot of cosine of the static advancing and receding contact angles vs. liquid surface tension for low density polyethylene. The same data are used in Figure 6.5. (From Hansen, C.M. and Wallström, E., J. Adhes., 15[3/4], 282, 1983. With permission.)

1.0

ADV. REC.

7

0.8

3

COS θ

2

0.6 1

0.4

0.2

16

18

20

22

RA

24

26

28

FIGURE 6.5 Critical HSP plot of cosine of the static adv ancing and receding contact angles vs. the HSP difference as defined by Chapter 1, Equation 1.9 for l w density polyeth ylene. The same data are used in Figure 6.4. (From Hansen, C.M. and Wallström, E., J. Adhes., 15[3/4], 282, 1983. With permission.)

7248_C006.fm Page 120 Wednesday, May 23, 2007 11:00 AM

120

Hansen Solubility Parameters: A User’s Handbook

for 2 and 5 min, respectively, with H2S2O7; the Joffé effect — effect of liquid immersion on fracture strength of soda-lime glass; and the Rehbinder ef fect — crushing strength of Al2O3 granules under various liquids. Beerbo wer has also brought cohesion parameters into the discussion of wear and boundary lubrication. 14 It appears that these f actors should still ha ve some consideration, e ven though recent progress and understanding in the area are much more adv anced.15 Additional surface characterizations using HSP are reported in Chapter 7. These include characterizations of the surf aces of pigments, fillers, and fibers. Both ganic and inor ganic materials have been characterized. The test method used is to determine sedimentation rates for the materials of interest in the same lar ge number of solv ents traditionally used in HSP studies. Adsorption of given liquids onto the particle or fiber sur ace slows the sedimentation rate, and indeed some (fine particles with rather high densities suspend for years in organic liquids with rather modest densities. A significant ad antage in this testing method is that he xane, for e xample, is not able to retard sedimentation where it may spontaneously spread, as discussed abo ve. Hexane is not an isolated example of this beha vior. The characterizations using standard HSP procedures indicate it is truly high affinity for the sur ace, which is important in these characterizations and not just spontaneous spreading. The reason for this may be the e xtent (or depth) of the adsorption layer , as well as whether the adsorption occurs at specific sites, or both. Results may be a fected when molecules in a surface can orient differently from their original state upon contact with a liquid, for e xample, with water (see the discussion in Chapter 18). An indirect correlation between HSP and the phenomena discussed above, spontaneous spreading and de wetting, has been established through measurements of en vironmental stress cracking (ESC).16 As discussed in Chapter 14, ESC correlates with the strain and the HSP and molecular size and shape of the cracking agent. The polymers polycarbonate (PC), c yclic olefinic copolyme (COC), and acrylonitrile/b utadiene/styrene (ABS) terpolymer could be described in terms of the regions A, B, and C as sho wn in Figure 6.1 to Figure 6.3. A large number of test liquids in each category were used to e valuate the critical strain required for ESC. It w as found that in e very case tested, category A liquids gave ESC. All category B liquids also g ave ESC, but the critical strains were somewhat higher on an a verage. Category C liquids could also gi ve ESC in some cases. nhexane was a category C liquid for some of these polymers in spite of its lo w surface tension. The HSP differences outweighed the expected spreading based on surface tension differences. Although these observ ations should not replace testing, a simple test of applying a droplet of liquid and possibly spreading it, if it does not do so itself, is a rapid w ay to assess a potential problem. Before lea ving this section, it is appropriate to mention that thinking of the type described above has led to a Nordtest Method, NT POL Y 176, “Spreading Surf ace Tension by the Applied Droplet Method.” This method is based on visual observation of droplets of known surface tension after they are applied to a test surface. The test surface may be a polymer, metal, or other material. The spreading surface tension is found to within ±1 mN/m by locating two liquids in a series where one of them spreads spontaneously and the other with a slightly higher surf ace tension does not. The preferred set of liquids is made with ethanol and w ater with a dif ference of 2 mN/m between them. Surf aces of man y different geometries (from 4 μm diameter wire to ships being painted), states of contamination (from clean for internal medical use to contamination with oil, silicones, pressure sensitive adhesive, etc.), and orientation (ceiling in a tunnel, inside pipes, etc.) ha ve been tested with remarkable success using this simple test. The usual procedure is to assign a v alue to a clean(ed) surface and then compare test surfaces, wherever they may be, against this to determine the presence of contamination.

SELF-STRATIFYING COATINGS A newer development in the coatings industry is to apply a single coat of paint which separates by itself into a primer and topcoat. A special issue of the journal Progress in Organic Coatings was devoted to this type of coating.17 Misev has also discussed formulation of this type of product using

7248_C006.fm Page 121 Wednesday, May 23, 2007 11:00 AM

Methods of Characterization — Surfaces

121

HSP concepts.18 The separation of the binders into primer and topcoat must occur while the coating is still liquid enough to allo w the necessary transport processes to occur . The solv ent must just dissolve the binders such that the y become incompatible when it be gins to e vaporate. The binder with the lowest energy (surface tension/cohesion parameters) will naturally migrate toward the low energy air interf ace and, therefore, this determines which of the binders mak es up the topcoat. There are a number of other f actors which are important for the process, including polymer molecular weight, rate of solv ent e vaporation, etc., b ut these will not be discussed here. This discussion is included because it once more demonstrates ho w cohesion parameters are coupled with surface energy and also to interf acial energy. The interface between the topcoat and primer is formed from an otherwise homogeneous system. The previous considerations lead to the e xpectation that the magnitude of the interf acial surface tension between tw o incompatible polymers is closely related to the dif ference in their cohesion parameters. Without going into greater detail, it is widely known among those who work with partially compatible polymers that this is indeed the case.19,20 See also Chapter 9 where partial compatibility in bitumen (asphalt) is discussed. Figure 6.6 shows the principles involved for selecting the solv ent which can make these work. The polymer with HSP nearest the origin will be the topcoat, as it has the lo wer (surface or cohesion) energy of the two. A solvent is required which dissolves both polymers, so it will be located in the common region to the spheres portrayed. Mutual solubility of tw o polymers is promoted when the solvent favors the polymer which is most dif ficult to dissol e.21 This is usually the one with the higher molecular weight. It is clear that selection of the optimum solv ent for this process of designed generation of an interf ace is aided by systematic use of HSP . This is a prime e xample of selfassembly where proper formulation can be aided by the concepts discussed abo ve.

POLAR PARAMETER

PRIMER

PARAMETERS REQUIRED FOR COMMON SOLVENT TOP COAT (LOWEST ENERGY)

HYDROGEN BONDING SOLUBILITY PARAMETER

FIGURE 6.6 Sketch of HSP principles used to formulate a self-stratifying coating from an initially homogeneous solution (see discussion in te xt). (From Birdi, K.S., Ed., Handbook of Surface and Colloid Chemistry, CRC Press, Boca Raton, FL, 1997, p. 324.)

7248_C006.fm Page 122 Wednesday, May 23, 2007 11:00 AM

122

Hansen Solubility Parameters: A User’s Handbook

MAXIMIZING PHYSICAL ADHESION If one wishes to maximize ph ysical adhesion, the ph ysical similarity (same HSP) of the tw o interfaces being joined must be as close as possible. The previous discussion suggests that physical similarity can be obtained when tw o criteria are met. The first criterion is that xactly the same liquids spontaneously spread on each of the surf aces to be joined. The second criterion is that exactly the same liquids maintain films when spread (ASTM method for wetting tension) on eac of the surf aces to be joined. Any dif ferences in this spontaneous spreading or wetting tension behavior can be interpreted as being a dif ference in ph ysical similarity . The dif ferences in the behavior of liquid droplets or films that are obser ed may suggest which steps can be tak en to minimize differences, if this is required. Should one add aliphatic segments to reduce the polar and hydrogen bonding contrib utions? Should alcohol and/or acid groups be incorporated to increase the hydrogen bonding in the system? This type of approach can be used to establish guidelines for action relative to each of the HSP parameters. Aromatic character and halogens other than fluorin characteristically increase δD; nitro and phosphate groups characteristically increaseδP; and alcohol, acid, and primary amine groups characteristically increase δH. Reference can be made to the table of group contrib utions in Chapter 1 (T able 1.1) for more precise comparisons. The discussion of forming good anchors on pigments and other surf aces found in Chapter 8 is also rele vant to the present discussion, as such anchors can also be used to enhance adhesion.

CONCLUSION Greater insight into the mak eup of a product is possible when one not only kno ws the cohesion parameters, i.e., HSP, for polymers and solvents it contains, but also the HSP for the various surfaces which these encounter . The surf aces of substrates, pigments, fillers, plastics, fibers, and oth materials can also be characterized by HSP (see Chapter 5 and Chapter 7). This allo ws mutual interactions to be inferred by comparisons of which materials are similar and which materials are different in terms of their HSP . Similar materials in this conte xt have similar HSP re gardless of differences in composition. The critical surf ace tension and wetting tension are single point determinations. Cohesion parameters allow a more complete characterization of surfaces than do these single point measurements and, at the same time, allo w insight as to ho w the single point measurements fit into th overall ener gy picture for the product. Guidelines for systematically changing the af finities o surfaces can also be obtained from HSP concepts. Both the spontaneous spreading re gion and the wetting tension re gion on HSP plots for tw o different surfaces must be identical if the y are to ha ve identical o verall surface characteristics.

REFERENCES 1. Flory, P.J., Principles of Polymer Chemistry, Cornell University Press, New York, 1953. 2. Barton, A.F.M., Applications of solubility parameters and other cohesion ener gy parameters, Polym. Sci. Technol. Pure Appl. Chem., 57(7), 905–912, 1985. 3. Skaarup, K. and Hansen, C.M., The Three-Dimensional Solubility Parameter and Its Use (Det Tredimensionale Opløselighedsparametersystem og dets Anvendelse), Rapport No. 54 (TM 2-67), Lakog Farveindustriens Forskningslaboratorium, København, 1967 (in Danish). 4. Skaarup, K., Surf ace Tension and 3-D Solubility P arameters (Ov erfladespænding og 3-D Opløse lighedsparametre), Nordiske Kemikermøde, København, 1968 (in Danish). 5. Beerbower, A., Surface free energy: a new relationship to bulk energies, J. Colloid Interface Sci., 35, 126–132, 1971. 6. Hansen, C.M. and Beerbo wer, A., Solubility parameters, in Kirk-Othmer Encyclopedia of Chemical Technology, 2nd ed., Suppl. Vol., Standen, A., Ed., Interscience, Ne w York, 1971, pp. 889–910.

7248_C006.fm Page 123 Wednesday, May 23, 2007 11:00 AM

Methods of Characterization — Surfaces

123

7. Koenhen, D.N. and Smolders, C.A., The determination of solubility parameters of solv ents and polymers by means of correlation with other physical quantities, J. Appl. Polym. Sci., 19, 1163–1179, 1975. 8. Erbil, H.Y., Surface tension of polymers, in Handbook of Surface and Colloid Chemistry, Birdi, K.S., Ed., CRC Press, Boca Raton, FL, 1997, pp. 265–312. 9. Hansen, C.M., Characterization of liquids by spreading liquids, J. Paint Technol., 42(550), 660–664, 1970. 10. Hansen, C.M. and Pierce, P .E., Surface effects in coatings processes, XII Federation d’Associations de Techniciens des Industries des Peintures, Vernis, Emaux et Encres d’Imprimerie de l’Europe Continentale, Congress Book, Verlag Chemie, Weinheim/Bergstrasse, 91–99, 1974; Ind. Eng. Chem. Prod. Res. Dev., 13(4), 218–225, 1974. 11. Hansen, C.M. and Wallström, E., On the use of cohesion parameters to characterize surfaces, J. Adhes., 15(3/4), 275–286, 1983. 12. Zisman, W.A., Relation of the equilibrium contact angle to liquid and solid constitution, in Contact Angle, Wettability and Adhesion, Advances in Chemistry Series No. 43, Gould, R.F ., Ed., American Chemical Society, Washington, D.C., 1964, chap. 1. 13. Zisman, W.A., Surface energetics of wetting, spreading, and adhesion, J. Paint Technol., 44(564), 41, 1972. 14. Beerbower, A., Boundary Lubrication — Scientific and Technical Applications Forecast, AD747336, Office of the Chief of Research and D velopment, Department of the Army, Washington, D.C., 1972. 15. Krim, J., Friction at the Atomic Scale, Scientific American, 275(4), October 1996, pp. 48–56. 16. Nielsen, T.B. and Hansen, C.M., Surf ace wetting and the prediction of en vironmental stress cracking (ESC) in polymers, Polym. Degradation Stability, 89, 513–516, 2005. 17. Special issue de voted to self-stratifying coatings, Prog. Org. Coat., 28(3), July 1996. 18. Misev, T.A., Thermodynamic analysis of phase separation in self-stratifying coatings — solubility parameters approach, J. Coat. Technol., 63(795), 23–28, 1991. 19. Luciani, A., Champagne, M.F., and Utracki, L.A., Interfacial tension in polymer blends. Part 1: Theory, Polym. Networks Blends, 6(1), 41–50, 1996. 20. Luciani, A., Champagne, M.F ., and Utracki, L.A., Interf acial tension in polymer blends. P art 2: Measurements, Polym. Networks Blends, 6(2), 51–62, 1996. 21. Hansen, C.M., On application of the three dimensional solubility parameter to the prediction of mutual solubility and compatibility, Färg och Lack, 13(6), 132–138, 1967.

7248_C006.fm Page 124 Wednesday, May 23, 2007 11:00 AM

7248_C007.fm Page 125 Wednesday, April 4, 2007 1:00 PM

of Characterization 7 Methods for Pigments, Fillers, and Fibers Charles M. Hansen ABSTRACT Cohesion parameters for pigments, fillers, and fibers can often be valuated by observation of the suspension and/or sedimentation beha vior of particulate matter in dif ferent liquids. These characterizations are based on relatively stronger adsorption by some of the liquids compared with others. Those liquids with stronger interaction can suspend finer fractions of solids indefinitely or reta sedimentation, compared with the other liquids. Data should be interpreted by accounting for differences in the densities and viscosities of the test liquids, such that a relative sedimentation rate can be used for comparisons. The absolute sedimentation rates are generally not of primary interest. Data from such evaluations can be computer-processed to assign Hansen cohesion parameters (HSP) to the material in question. Cohesion parameter data are gi ven for some ne wer pigments, fillers and a carbon fiber to demonstrate the principles

INTRODUCTION The possibilities offered by cohesion parameter characterization of pigments, fillers, and fibers ve not been generally recognized, judging from the relati vely small number of publications appearing on the topic. Pigments and a few fillers were characterized in some of the author s first publication dealing with the solubility parameter .1,2 These were gi ven δD, δP, and δH parameters (HSP) and a characteristic radius of interaction e xactly analogous to the polymer characterizations discussed in Chapter 2 and Chapter 5. These data together with some more recent pigment characterizations are included in Table 7.1, Table 7.2A, and Table 7.2B. Shareef et al. 3 have also characterized pigment surfaces, including metal oxides. Gardon and Teas4 clearly sho wed the dif ferences between zinc oxides treated and untreated with or ganic phosphate using a cohesion parameter characterization. Inorganic fibers h ve also been characterized. 5 All of these characterizations ag ain confirm th universality possible with these parameters. They reflect molecule–molecule interactions whethe at surfaces or in b ulk. In the future, more systematic selection of dispersion aids should be possible, as these can also be described with the same energy parameters. Hansen and Beerbower have touched on this topic.6 Each se gment of such molecules requires its o wn HSP. The discussion in Chapter 15 for the interactions within cell w alls in wood demonstrates how this could be done. It has been sho wn by calculation that hemicelluloses act lik e surface-active agents, with some se gments seeking lo werenergy lignin regions and some segments (those with alcohol groups) orienting to ward the higherenergy cellulose.

125

7248_C007.fm Page 126 Wednesday, April 4, 2007 1:00 PM

126

Hansen Solubility Parameters: A User’s Handbook

TABLE 7.1 HSP Correlations for Older Inorganic Pigments1,2 and Metal Oxides3 Material

δD

δP

δH

Ro

Kronos® RN57 TiO2a Aluminum pulver lack 80 a Red iron oxide a Synthetic red iron oxide b Synthetic yellow iron oxide b

24.1 19.0 20.7 16.1 17.3 16.1 16.9 16.2

14.9 6.1 12.3 8.6 6.0 8.6 7.8 10.8

19.4 7.2 14.3 15.0 14.5 15.0 10.6 12.7

17.2 4.9 11.5 11.3 12.5 11.3 13.2 9.8

Zinc oxide Note: Units are MP a1/2.

From Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Doctoral disse tation, Danish Technical Press, Copenhagen, 1967. With permission. b From Shareef, K.M.A. et al., J. Coat. Technol., 58(733), 35–44, February 1986. With permission. a

METHODS TO CHARACTERIZE PIGMENT, FILLER, AND FIBER SURFACES The cohesion parameter (HSP) approach to characterizing surf aces gained impetus by experiments where the suspension of fine particles in pigment p wders was used to characterize 25 organic and inorganic pigment surfaces.1,2 Small amounts of the pigments are shak en in test tubes with a gi ven volume of liquid (10 ml) of each of the test solv ents, and then sedimentation or lack of the same is observed. When the solid has a lo wer density than the test liquid, it will float. Rates of floati have also been noted, b ut the term sedimentation will be retained here for both sedimentation and floating. The amounts of solid sample added to the liquids can v ary depending on the sample in question, and some initial experimentation is usually advisable. If the pigment or filler particle siz is large — say o ver 5 µm — the surf ace effects become less significant compared with a sampl where the particle size is only 0.01 µm. Problems arise when the pigments are soluble enough to color the liquid such that sedimentation cannot be e valuated. The larger particle size pigments and fillers may sediment ery rapidly. Sedimentation rates have still been used successfully in some of these cases. The sedimentation rate is most easily e xpressed as the time at which the amount of particles, at a gi ven point in the test tubes, has f allen to some small amount, perhaps zero. Observ ations can be made visually , or perhaps instrumentally , in a direction perpendicular to the incidence of a laser light. A visual observation is required in an y event, as some samples seem to coat out rapidly on glass surf aces. Some pigments ha ve portions that suspend for years in spite of lar ge-density dif ferences and relatively low-solvent viscosity. Satisfactory results from this type of measurement require some experience regarding what to look for . This can vary from sample to sample. A characterization is less certain when there are only 4 or 5 good liquids out of the perhaps 40 to 45 tested, although this depends some what on which liquids are in volved. “Good” in this context means suspension of particulates is prolonged significantl , compared with the other test solvents, after compensating for differences in density and viscosity. A corrected relative sedimentation time (RST) can be found by modifying the sedimentation time, t s

7248_C007.fm Page 127 Wednesday, April 4, 2007 1:00 PM

Methods of Characterization for Pigments, Fillers, and Fibers

127

TABLE 7.2A List of Pigments Studied. HSP Results are Given in Table 7.2B Pigment

Description

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

TiO2, Kronos RN 57, Titan Co. A/S., Frederikstad, Norw ay. Phthalocyanine Blue, B6, E. I. du Pont de Nemours and Co. (1949). Isolbonared Nr. 7522, C. I. Pigment Red 48 (C.I. 15865) (MnSalt), Køge Chemical Works, Køge, Denmark. Peerless Carbon Black Isol Fast Yellow IO GX 2505, C.I. Pigment Yellow 3, Køge Chemical Works, Køge, Denmark. Refl x Blau TBK Ext. (No C.I. Inde x-pigment mixture), F arbwerke Hoechst, Frankfurt (M), German y. Isol Ruby BKS 7520, C.I. Pigment Red 57 (C.I. 15850) (Ca Salt), Køge Chemical Works, Køge, Denmark. Hansagelb 10 G, C.I. Pigment Yellow 3 (C.I. 11710), F arbwerke Hoechst, Frankfurt (M), German y. Fanalrosa G Supra Pulv er, Pigment Red 81 (C.I. 45160), B ASF, Ludwigshafen, German y. Heliogenblau B Pulv er, C.I. Pigment Blue 15 (C.I. 74160), B ASF, Ludwigshafen, German y. Heliogengrün GN, C.I. Pigment Green 7, (C.I. 74260), B ASF, Ludwigshafen, German y. Permanentgelb H 10 G, C.I. PigmentYellow 81, (No C.I. index), Farbwerke Hoechst, Frankfurt (M), Germany. Permanent Bordeaux FRR, C.I. Pigment Red 12 (C.I. 12385), F arbwerke Hoechst, Frankfurt (M), Germany. Permanent Violet RL Supra, C.I. Pigment Violet 23, (C.I. 12505), F arbwerke Hoechst, Frankfurt (M), Germany. Isol Benzidine Yellow G 2537, C.I. Pigment Yellow 12 (C.I. 21090), Køge Chemical Works, Køge, Denmark. Brillfast Sky Blue 3862, C.I. Pigment Blue 3 (C.I. 42140), J. W. and T. A. Smith Ltd., London. Permanent Orange G, C.I. Pigment Orange 13 (C.I. 21110), F arbwerke Hoechst, Frankfurt (M), German y. Permanent Red, FGR Extra Pulver, C.I. Pigment Red 112, (C.I. 12370). Farbwerke Hoechst, Frankfurt (M), Germany. Isol Fast Red 2G 2516, C.I. Pigment Orange 5, (C.I. 12075), Køge Chemical Works, Køge, Denmark. Monolite Fast Blue 3 RS, Po wder, C.I. Vat Blue 4 (C.I. 69801), Imperial Chemical Industries. Heliogenblau LG, Pulv er, C.I. Pigment Blue 16 (C.I. 74100), B ASF., Ludwigshafen, German y. Red Iron Oxide. Carbon Black, Printe x V (5519-1), De gussa, Frankfurt (M), German y. Aluminum Pulver Lack 80, Eckart-W erke, 851 Fürth/Bayern, German y. Isol Benzidene Yellow GA-PR, 9500, C.I. Pigment Yellow 12, Køge Chemical Works, Køge, Denmark.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

Source: From Hansen, C.M., The Three Dimensional Solubility P arameter and Solv ent Diffusion Coefficient, Doctora dissertation, Danish Technical Press, Copenhagen, 1967. With permission.

RST = t s(ρp – ρs)/η

(7.1)

ρp and ρs are densities of particle and test liquid, respecti vely, and η is the liquid viscosity . A prolonged RST implies greater adsorption of the gi ven solvent onto the surf ace in question. Characterizations based on these techniques tend to place emphasis on the nature of the surf aces for the smaller -particle-size fractions. An example of a data sheet used for such studies is included in Table 7.3.

DISCUSSION — PIGMENTS, FILLERS, AND FIBERS It can be reasoned that a pigment, fille , or fiber is most beneficial when the pigment su ace and the binder in question ha ve the same cohesion parameters. There are apparently no publications indicating a systematic modification of pigment sur aces to achieve a given set of cohesion parameters. The characterizations for some or ganic pigments are gi ven in Table 7.4. These data indicate that their respecti ve surf aces are essentially identical. An e xception is the first one in the tabl where the analysis is based on only three good solv ents that were able to e xtend sedimentation significantly relat ve to the other solv ents tested.

7248_C007.fm Page 128 Wednesday, April 4, 2007 1:00 PM

128

Hansen Solubility Parameters: A User’s Handbook

TABLE 7.2B Characteristic Parameters for Various Pigments Given in Table 7.2A Pigment

δt

δD

δP

δH

Ro

Comments

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

16.8 10.5 10.0 13.6 11.9 13.2 10.5 10.5 13.0 12.0 12.0 8.8 13.2 11.5 10.2 13.3 11.5 11.2 14.2 15.2 13.5 13.7 13.1 10.4 9.1

11.8 9.3 8.7 10.3 10.2 10.8 9.6 9.1 9.8 10.8 10.0 8.4 10.7 9.6 9.3 9.5 9.7 10.0 10.9 10.8 10.7 10.1 10.3 9.3 9.0

7.3 3.1 3.5 6.0 4.8 3.8 3.0 4.0 7.0 3.5 4.8 1.5 4.8 5.2 3.0 7.2 3.9 3.5 5.6 6.5 5.0 6.0 6.0 3.0 2.7

9.5 3.7 3.5 6.6 3.8 6.6 3.2 3.3 5.0 4.0 4.5 2.3 6.1 3.6 2.9 6.0 4.7 3.5 7.1 8.5 6.5 7.0 5.5 3.5 2.3

8.4 2.3 2.5 6.0 4.4 7.0 3.9 3.3 5.2 5.2 4.8 2.2 5.2 4.4 3.9 5.1 4.5 5.0 7.0 7.0 6.0 5.6 5.5 2.4 2.5

Suspension Few suspending solv ents Few suspending solv ents Suspension Color only Mixed color and suspension Suspension Color only Color only Suspension Primarily suspension Suspension Color only Mixed color and suspension Mixed color and suspension Suspension Color only Color only Primarily color Suspension Suspension Suspension Suspension Suspension Suspension

Note: Units are (cal/cm3)1/2. Source: From Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967. With permission.

These results suggest that pigment manufacturers have essentially arrived at the same result — a surf ace ener gy compatible with a wide v ariety of currently used binders. The solv ents most frequently appearing as good for adsorption onto these surfaces include several chlorinated solvents, toluene, and tetrahydrofuran. As these solvents dissolve the most commonly used binders, one can conclude that the common binders will adsorb readily onto these pigment surf aces. This will give a good result, provided the solvent is not so good for the binder that it can remo ve the binder from the pigment surf ace. Schröder7 (BASF) confirms that the optimum polymer adsorption will be found when the binde and pigment surface have the same HSP. He indicates that the solv ent should be v ery poor for the pigment and located on the boundary re gion for the binder . He prefers the pigment to ha ve HSP values placing it intermediate between the solv ent and binder . This is suggested for conditions where the solv ent has higher HSP than the pigment, as well as for conditions where the solv ent has lower HSP than the pigment. This situation, with the solv ent and binder on opposite sides of the pigment, means the composite v ehicle has parameters v ery closely matching those of the pigment.

7248_C007.fm Page 129 Wednesday, April 4, 2007 1:00 PM

Methods of Characterization for Pigments, Fillers, and Fibers

129

TABLE 7.3 Sedimentation Study Date Sample: Density of sample (Dp):

Reference No.:

Solvent

D(Ds) 20°C

Acetone Acetophenone Benzene 1-Butanol Butyl acetate Butyrolactone Carbon tetrachloride Chlorobenzene Chloroform Cyclohexane Cyclohexanol Diacetone alcohol o-Dichlorobenzene Diethylene glycol Diethyl ether Dimethyl formamide DMSO 1,4-Dioxane Dipropylene glycol Ethanol Ethanolamine Ethyl acetate Ethylene dichloride Ethylene glycol Ethyelene glycol monob utyl ether Ethyelene glycol monoeth yl ether Ethyelene glycol monometh yl ether Formamide Hexane Isophorone Methanol Methylene dichloride Methyl isobutyl ketone Methyl-2-pyrrolidone Nitrobenzene Nitroethane Nitromethane 2-Nitropropane Propylene carbonate Propylene glycol Tetrahydrofurane Toluene Trichloroethylene

0.79 1.03 0.88 0.81 0.87 1.29 1.59 1.10 1.48 0.78 0.95 0.94 1.31 1.12 0.72 0.95 1.10 1.04 1.03 0.82 0.91 0.89 1.25 1.12 0.90 0.93 0.96 1.13 0.66 0.92 0.79 1.33 0.96 1.03 1.21 1.05 1.13 0.99 1.20 1.04 0.89 0.87 1.47

Sedimentation Time (min)

Viscosity 20°C

No.

0.35 1.90 0.65 4.00 0.74 1.92 0.99 0.80 0.37 1.00 68.00 3.20 1.27 35.70 0.23 0.82 1.98 1.31 107.0 1.22 24.10 0.44 0.84 20.90 2.90 2.05 1.72 3.30 0.33 2.60 0.59 0.43 0.59 1.80 2.03 0.55 0.62 0.75 2.80 56.00 0.55 0.59 0.58

4 6 13 28 30 37 40 41 44 47 48 56 61 75 82 90 94 96 98 104 105 106 120 121 123 124 126 131 140 148 153 162 167 172 177 178 179 181 204 205 222 225 229

D p- D s

From

To

Relative Sedimentation Time (RST) From

To

7248_C007.fm Page 130 Wednesday, April 4, 2007 1:00 PM

130

Hansen Solubility Parameters: A User’s Handbook

TABLE 7.4 HSP Correlations for Selected Materials Material

δD

δP

δH

Ro

Fit

G/T

Organic Pigments Paliotol® Gelb L1820 B ASF Heliogen ® Blau 6930L B ASF Socco Rosso L3855 B ASF Perm Rubin F6B Hoechst Perm Gelb GRL02 Hoechst Perm Lackrot LC Hoechst

18.9 18.0 17.3 16.7 16.7 19.0

3.5 4.0 5.7 3.7 2.5 5.0

10.5 4.0 2.7 3.1 3.7 5.0

5.4 4.0 4.1 4.8 4.5 4.0

0.99 1.00 0.99 0.88 0.95 1.00

3/35 5/34 4/34 6/33 5/37 7/28

Inorganic Pigments, Fillers, etc. Cabot Hochdisperse a 16.7 Cabot Hochdisperse 19.3 Zeta Potential Blanc Fix eb 26.5

9.3 9.5 19.1

11.5 10.3 14.5

11.7 12.7 20.4

— 0.79 0.95

23/23 23/31 5/19

Note: A perfect data fit of 1.0 means that there most probably are other sets o the same parameters that will ha ve a data fit of 1.0 and also define a sphere th surrounds all the good solv ents. A data fit of 0.99+ is preferred to define t optimum sphere for this reason. G/T is the number of good (G) liquids and the total (T) number of liquids in a correlation. Units are MP a1/2. Data analysis which only considers the good solv ents to define the least spher possible. See discussion of the SPHERE1 program in Chapter 1. b Data from Winkler, J., Eur. Coat. J., 1–2/97, 38–42, 1997. With permission. a

There is also a relation between ho w clearly a pigment can be characterized by sedimentation measurements and its zeta potential. Low zeta potential means sedimentation is rapid in all solvents, and this type of characterization becomes difficult or perhaps impossible. The zeta potential reflect the intensity (percentage co verage and number of layers) of the surf ace energy characteristics. It does not clearly indicate specific a finity relations of a g ven binder for the pigment surf ace, as a result of a given surface treatment, for example. This is given by HSP. To obtain a complete picture of the ener getics of the surf ace, one needs an intensity f actor, i.e., the zeta potential, as well as a qualitative factor, i.e., the cohesion ener gy parameters. The latter are generally lacking. One can suspect that some pigments ha ve such high-intensity zeta potential — at some cost — that e ven though the cohesion parameters match poorly with a gi ven binder , a system can still function satisfactorily. An HSP correlation for the zeta potential of blanc fi e is gi ven in Table 7.4 using data from Winkler.8 This is discussed further in the follo wing section. Acid–base theories ha ve been popular .9–11 The author has not found it necessary to resort to this type of approach in an y acti vity, although man y ha ve clearly found them beneficial. Mor research is needed to fully understand the successes of the acid–base as well as the HSP approaches. It would seem that the HSP approach allows predictive ability that is not possible with an acid–base approach. However, the current problem is the lack of data. Organic pigments normally have a good organic substrate on which to base an or ganic surface modification. The characterizations may reflect both a sur ace treatment and the surface of the base particles, depending on ho w the test liquids interact with these. It should not be too dif ficult t modify an organic surface to an alternative organic surface with satisfactory properties, if desired. It is conceptually and, in practice, more dif ficult to modify an ino ganic surf ace to mak e it compatible with or ganic systems. This requires a significant change in sur ace energy from high to much lower and, presumably, also requires a greater degree of coverage to mask the base inorganic surface. The producers of inor ganic pigments and fillers must either g ve their products suitable

7248_C007.fm Page 131 Wednesday, April 4, 2007 1:00 PM

Methods of Characterization for Pigments, Fillers, and Fibers

131

surfaces, probably after much ef fort, or else one needs one or more superef fective additives to be able to achieve a good and stable dispersion. It helps to incorporate gi ven (high-cohesion energy) groups in a grinding resin, such as acid, alcohol, amine, etc. The relati vely high local cohesion parameters in the binder that are associated with these groups w ould indicate a high af finity fo the high-cohesion-energy surface of the inor ganic material. At the same time, these local re gions of adsorbed polymer se gments are not particularly soluble or are insoluble in the cheaper h ydrocarbon solvents — for e xample, those that ha ve much lower cohesion parameters. This provides a good, stable anchor on the pigment surf ace. The solvent will not dissolve that polymer or polymer segment away from the surface. Binders with high-acid numbers are frequently used, with success, in printing inks for the same reason. This is discussed in more detail in the follo wing section. It is felt that those who understand the use of cohesion parameters are able to more systematically modify surfaces of inorganic materials to optimize or improve their compatibility with organic polymers and binders. This has been done for inorganic Rockwool® fibers that are to be incorporate into polypropylene.5 It must be presumed that this type of systematic procedure can guide surf ace treatment of other inor ganic materials in a more directed w ay toward a desired goal. Data fits h ve been generally lo wer for characterizations of particulate solid surf aces, such as fillers, than for characterizations of polymers based on solubilit . When testing is finished, th polymeric macromolecule is no longer a solid in the good solv ents, whereas the particulate fille remains a solid. A macromolecule has v arious possibilities for contortions and the positioning of significant act ve groups in solution (or when swollen), giving a large number of possible (dynamic) structures that can be formed with the solv ent. A rigid solid surf ace does not ha ve this potential for adopting ener getically desirable positions for its acti ve groups. The adjustments for optimum local association must be made by the solv ent molecules alone in the sedimentation testing. There are many solvents that do not retard sedimentation significantl , whereas the predictions based on the behavior of other solv ents that do significantly retard sedimentation indicate that this shoul be the case. A contribution to the formation of the ener getically desirable geometrical structures is not possible from the mo vement of rigid solid surf aces. Therefore, some solvents may not be able to retard sedimentation because the y cannot adopt the geometrical positioning required to do this without the help of a mobile substrate. This lack of e xpected performance may be also partly due to solvent size, the location of the active groups, or combinations of these. These phenomena appear to be a significant area for future research

HANSEN SOLUBILITY PARAMETER CORRELATION OF ZETA POTENTIAL FOR BLANC FIXE Winkler8 has reported zeta potentials measured for 1% v/v blanc fi e with 0.34% moisture content. There were 19 liquids included in this careful study. These liquids could easily be divided into two groups. There were 5 systems with zeta potentials greater than about 10 mV and 14 systems with zeta potentials less than about 5 mV . Table 7.3 includes the results of the correlation of these data with cohesion parameters. The only major “error” was for hexamethylphosphoramide, with a RED of 0.951 and a zeta potential of 1.9 mV . This correlation supports the contention that cohesion parameters are significant for characterization of pigment, fill , and fiber sur aces. This is a good correlation and supports the views presented earlier. According to Winkler, there was no correlation with the acceptor or donor numbers (acid–base).

CARBON FIBER SURFACE CHARACTERIZATION Hansen solubility parameters ha ve been assigned to a carbon fiber sur ace, Panex 33 from Zoltek. After considerable refinement of the xperimental technique, it w as determined that tw o separate sets of HSP are required to describe the fiber sur ace. One of these sets has δD;δP;δH equal to

7248_C007.fm Page 132 Wednesday, April 4, 2007 1:00 PM

132

Hansen Solubility Parameters: A User’s Handbook

TABLE 7.5 HSP Correlations for Various “Carbon” Materials Material

δD

δP

δH

Ro

Comments

Carbon fiber (high r gion)b Carbon fiber (l w region)b Carbon black a Carbon black 1 Carbon black 1 HT b Carbon black 2 HT b Carbon black 3 HT b Carbon black 4 HT b Petroleum coke Coal tar pitch C60 fullerene d

13.4 17.2 21.1 17.9 21.5 20.5 20.5 18.9 16.4 18.7 19.7

17.8 3.4 12.3 8.1 7.3 8.7 8.7 11.1 4.0 7.5 2.9

14.2 2.0 11.3 8.9 13.3 12.7 12.5 8.5 10.0 8.9 2.7

10.3 3.9 16.6 6.2 11.4 9.2 9.1 5.6 10.7 5.8 3.9

Sedimentation rate 3mm fiber Sedimentation rate 3mm fiber Suspension Sedimentation (slow) Sedimentation (rapid) Sedimentation (rapid) Sedimentation (rapid) Sedimentation (rapid) Sedimentation (slow) Solubility Solubilityc

Note: Units are MP a1/2. a b c d

Printex V (5519-1), De gussa HT indicates a special heat treatment w as performed prior to testing. Log mole fraction solubility greater than –3. Data from Hansen, C.M. and Smith, A.L., Carbon, 42(8–9), 1591–1597, 2004. With permission.

13.4;17.8;14.2, all in MPa1/2. This corresponds to a highly polar surface with a significant ydrogen bonding component as well. The second set is characteristic of a hydrocarbon material with δD;δP;δH equal to 17.2;3.4;2.0, ag ain in MP a1/2. The h ydrocarbon-like surface can be the backbone of the polyacrylonitrile (PAN) precursor for the fibe. Two separate sets of HSP assignments are confirme by x-ray photoelectron spectroscop y (XPS) analysis. Two separate regions are found to coe xist on the carbon fiber sur ace. The hydrogen bonding and polar contrib utions arise from the both bound and unbound (sizing/finish agent) chemical functionalities mainly in the form of ydroxyl, ether, carbonyl, carboxyl, amide, and nitrile groups. The carbonaceous backbone of the carbon fibe primarily accounts for the nonpolar region. This work was done as a part of the Framework program Interface Design of Composite Materials with the support of the Danish Research Agency, Ministry of Science, Technology and Innovation (STVF). The HSP characterizations were done at FORCE Technology, Broendby, Denmark, whereas the analyses were done at the Risø National Laboratory, Roskilde, Denmark. Table 7.4 contains the HSP data used to construct Figure 7.1. These data confirm that “carbon can be man y things with widely dif ferent surface energies. The origins of the material, as well as the method in which it has been handled or treated, can completely dominate the nature of the surface of the gi ven materials and their solubility , if this is possible.

CONTROLLED ADSORPTION (SELF-ASSEMBLY) Significant tasks for formulators are to control the sur ace and interf acial ener gies of products, especially if they are water reducible. This is required to allow substrate wetting, to maintain stable dispersions, and to pro vide/ensure adequate and durable adhesion to gi ven substrates. Guidelines for courses of action are frequently a vailable when cohesion ener gy parameters are referred to. Some guides are discussed in the follo wing. It is a well-known fact that a small percentage of acid groups (or alcohol groups) on a polymer chain will promote adhesion and adsorption to man y surfaces. The cohesion ener gy parameter of an isolated acid group is high. One can consider the cohesion ener gy properties of formic acid ( δD;

7248_C007.fm Page 133 Wednesday, April 4, 2007 1:00 PM

Methods of Characterization for Pigments, Fillers, and Fibers

133

20

Carbon fiber, high

18

16

14

“Carbon Black”

12 δp 10

Carbon Black 1

8

Coal tar pitch 6

4

Carbon fiber, low

Petroleum Coke

Fullerene 2

0 0

2

4

6

8 δH

10

12

14

16

FIGURE 7.1 Characterization of a carbon fiber and comparisons of this with other carbon materials. Unit are MPa1/2. The work on which this figure is based as supported by the Framework program Interface Design of Composite Materials (STVF fund No. 26-03-0160). Reproduced with permission.

δP; δH = 14.3; 11.9; 16.6) as an isolated part of a polymer chain.The polar cohesion energy parameter of an acid group is not so high. It w ould seem logical to systematically use acid groups for adsorption to high-energy surfaces and to mak e certain that the cohesion ener gy parameters for the solv ent and bulk of the product are much lo wer, such that isolated acid groups w ould not be dissolv ed. This would provide an anchor that the product will not be able to remo ve. This type of adsorption may be called hydrophilic bonding. If, on the other hand, the solvent were too good for the anchor,

7248_C007.fm Page 134 Wednesday, April 4, 2007 1:00 PM

134

Hansen Solubility Parameters: A User’s Handbook

it could be presumed that e ven an acid group may be too readily dissolv ed off the surf ace or at least tak e part in a dynamic equilibrium of adsorption and desorption. Absorbed/adsorbed w ater can sometimes interfere with such anchors at high-ener gy surfaces. The reverse of this thinking is systematically used by those designing associati ve thickeners and also by nature, such as the h ydrophobic bonding in proteins. Certain se gments of gi ven molecules have such low cohesive energy parameters that they are no longer soluble in the media, which is usually aqueous, and the y either seek out their o wn kind (associate) or perhaps adsorb on or penetrate into a low-energy surface where cohesion energy parameters more suitably match. The positive effects of associative thickeners can be counteracted by the presence of solv ents preferentially locating where the hydrophobic bonding is to occur. The hydrophobic bonds lose strength or may even dissolve away. A challenge to the creati ve mind is to deri ve ne w uses for high-ener gy groups that are not particularly water soluble or sensitive. The division of the cohesion ener gy into at least three parts allows these considerations to be made in a reasonably quantitati ve manner. One can choose nitro groups or perhaps groups containing phosphorus as e xamples of species characterized by highpolar-cohesion-energy parameters and lo w or moderate h ydrogen bonding parameters. The total cohesion ener gy parameters for ethanol and nitromethane are v ery close: 26.1 and 25.1 MP a1/2, respectively. Ethanol is soluble in water, nitromethane is not. Ethanol has a relatively high-hydrogen bonding parameter (19.4 MP a1/2) compared with nitromethane (5.1 MP a1/2). This mak es all the difference. Would not the nitro group be a suitable anchor analogous to the pre vious discussion concerning acid groups? Also, it w ould not be h ydrophilic with the inherent problems of w ater sensitivity associated with high-hydrogen bonding parameters. Several of the pigments reported in Table 7.4 did indeed ha ve moderate af finity for the nitropara fins, for xample, b ut the y were included in the lesser interacting group by the arbitrary di vision into good and bad groups.

CONCLUSION Many pigments and fillers h ve now been characterized by Hansen cohesion parameters (HSP). Many e xamples are gi ven. A method based on relati ve sedimentation time and/or suspension is described for doing this. This method has generally allowed useful characterizations, although some experience is helpful. For example, the data are often scattered and not nearly of the quality usually found when observing polymer solution beha vior. This scatter of data for untreated surf aces in particular may cause some to disre gard the method; hopefully, they can develop a better one. The obvious advantages of ha ving solvents, plasticizers, polymers, pigments, fillers, fibers, etc., ch acterized with the same ener gy parameters should pro vide incentive for impro ving on the present state, both in terms of numbers of characterizations as well as improved methodology. One assumes that maximum physical adsorption is accompanied by closely matching HSP . Local adsorption by so-called active groups (alcohol, acid, amine) ha ving the required match, can gi ve anchors on a surface that may no longer be soluble in the continuous media, and therefore will remain in place as required.

REFERENCES 1. Hansen, C.M., The Three Dimensional Solubility P arameter and Solvent Diffusion Coefficient, Doc toral dissertation, Danish Technical Press, Copenhagen, 1967. 2. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities II. J. Paint Technol., 39(511), 505–510, 1967. 3. Shareef, K.M.A., Yaseen, M., Mahmood Ali, M., and Reddy, P.J., J. Coat. Technol., 58(733), 35–44, February 1986.

7248_C007.fm Page 135 Wednesday, April 4, 2007 1:00 PM

Methods of Characterization for Pigments, Fillers, and Fibers

135

4. Gardon, J.L. and Teas, J.P., Solubility parameters, in Treatise on Coatings, Vol. 2, Characterization of Coatings: Physical Techniques, P art II, Myers, R.R. and Long, J.S., Eds., Marcel Dekk er, Ne w York, 1976, chap. 8. 5. Hennissen, L., Systematic Modification of Filler/Fibre Sur aces to Achieve Maximum Compatibility with Matrix Polymers, Lecture for the Danish Society for PolymerTechnology, Copenhagen, February 10, 1996. 6. Hansen, C.M. and Beerbo wer, A., Solubility parameters, in Kirk-Othmer Encyclopedia of Chemical Technology, Suppl. Vol., 2nd ed., Standen, A., Ed., Interscience, Ne w York, 1971, pp. 889–910. 7. Schröder, J., Colloid chemistry aids to formulating inks and paints,Eur. Coat. J., 5/98, 334–340, 1998. 8. Winkler, J., Zeta potential of pigments and fillers, Eur. Coat. J., 1–2/97, 38–42, 1997. 9. Vinther, A., Application of the concepts solubility parameter and pigment char ge, Chim. Peint. (England), 34(10), 363–372, 1971. 10. Soerensen, P., Application of the acid/base concept describing the interaction between pigments, binders, and solv ents, J. Paint Technol., 47(602), 31–39, 1975. 11. Soerensen, P., Cohesion parameters used to formulate coatings (K ohaesionsparametre an vendt til formulering af f arver og lak), Färg och Lack Scand., 34(4), 81–93, 1988 (in Danish). 12. Hansen, C.M. and Smith, A.L., Using Hansen solubility parameters to correlate solubility of C 60 fullerene in or ganic solvents and in polymers, Carbon, 42(8–9), 1591–1597, 2004.

7248_C007.fm Page 136 Wednesday, April 4, 2007 1:00 PM

7248_C008.fm Page 137 Wednesday, May 23, 2007 11:05 AM

— Coatings and 8 Applications Other Filled Polymer Systems Charles M. Hansen ABSTRACT Hansen solubility parameters (HSP) are widely used in the coatings industry to help find optimu solvents and solvent combinations. They also aid in substitution to less hazardous formulations in various other types of products such as cleaners, printing inks, adhesi ves, etc. The discussion in this chapter includes the ph ysical chemical reasons wh y solv ents function as the y do in man y practical cases. The behavior of solvents in connection with surf aces of various kinds and the use of HSP to understand and control surf ace phenomena is especially emphasized. Products where HSP concepts can be used in a manner similar to coatings include other (filled) polymer system of various types such as adhesi ves, printing inks, che wing gum, etc. There are man y examples of controlled self-assembly.

INTRODUCTION There are man y applications documented in the literature where HSP ha ve aided in the selection of solvents, understanding and controlling processes, and, in general, of fering guidance where affinitie among materials are of prime importance 1–5 (see also the follo wing chapters and e xamples below). This chapter emphasizes coatings applications and discusses the practical application of HSP to solv ent selection. Computer techniques are helpful, b ut not necessary. The same principles useful for under standing the behavior of coatings are useful in understanding beha vior in a lar ger number of related products, including adhesives, printing inks, and chewing gum, to mention a few. These contain widely different materials, both liquid and solid, which can be characterized by HSP.This allows their relative affinities to be established. Pr vious chapters ha ve discussed ho w to assign HSP to solv ents, plasticizers, polymers, and resins, as well as to the surf aces of substrates, pigments, fillers, and fiber Various additives such as resins, surf actants, fl vors, aromas, scents, drugs, etc., can also be characterized by HSP to infer ho w they behave in seemingly comple x systems.

SOLVENTS In order to find the optimum sol ent for a polymer , one must ha ve or estimate its HSP . Matching the HSP of an already e xisting solvent or combination of solv ents can be done, b ut this procedure does not necessarily optimize the ne w situation. The optimum depends on what is desired of the system. A solvent with the highest possible affinity for the polymer is both xpensive and probably not necessary and will rarely be optimum. In more recent years, optimization increasingly includes considerations of worker safety and the e xternal environment. Volatile organic compounds (VOC) are to be reduced to the greatest extent possible. Chapter 11 is devoted to replacing ozone-depleting chemicals in cleaning operations. 137

7248_C008.fm Page 138 Wednesday, May 23, 2007 11:05 AM

138

Hansen Solubility Parameters: A User’s Handbook

Whereas hand calculations and plotting of data are still quite useful and at times more rapid than computer processing, it is becoming almost mandatory that computers be used. To this end, most solvent suppliers and many large users of solvents have computer programs to predict solution behavior as well as e vaporation phenomena. In spite of these pressures to let the computer do the thinking, an experienced formulator can often arri ve at a near -optimum result without recourse to paper or to computers. A major f actor in this almost immediate o verview is the decrease in the number of solvents useful in coatings. By putting this together with other necessary considerations such as flash point, proper vaporation rate, cost, odor, availability, etc., the experienced formulator who knows the HSP for the relati vely few solv ents possible in a gi ven situation will be able to select a near -optimum combination by a process of e xclusion and simple mental arithmetic. This does not mean the use of HSP is on the w ay out. The real benefit of this concept is in interpretin more complicated behavior, such as affinities of polymers with polymers and polymers with suraces as described in the follo wing. Much more w ork needs to be done in these areas, b ut the following gives an indication of what might be e xpected. As indicated previously, computer techniques can be v ery useful but are not always necessary, and simple two-dimensional plots using δP and δH can often be used by those with limited xeperience with these techniques to solve practical problems. The nonpolar cohesion parameter, δD, cannot be neglected in every case, but, for example, when comparing noncyclic solvents in practical situations, it has been found that their dispersion parameters will be rather close regardless of structure. Cyclic solvents, and those containing atoms significantly la ger than carbon, such as chlorine, bromine, metals, etc., will ha ve higher dispersion parameters. The total solubility parameter for aliphatic hydrocarbon solvents is identical with their dispersion parameter and increases only slightly with increased chain length. This same trend is expected for oligomers of a polymer as molecular weight increases. Re gardless of the means of processing data, the follo wing e xamples are intended to illustrate principles on which to base a systematic course of action. Most coatings applications involve solvents reasonably well within the solubility limit which is indicated by the boundary of a solubility plot such as that sho wn in Figure 8.1. 1 A maximum of cheaper hydrocarbon solvent is also desired and can frequently be used to arri ve at such a situation for common polymers used in coatings. Some safety mar gin in terms of e xtra solvency is advised because of temperature changes, potential variations in production, etc. These can lead to a situation where solvent quality changes in an adv erse manner. Balance of solv ent quality on e vaporation of mixed solvents is also necessary. Here again, computer approaches are possible, and calculations of solvent quality can be made at all stages of evaporation. It is usually good practice to include a small or moderate amount of slo wly evaporating solvent of good quality and lo w water sensitivity to take care of this situation. These have frequently been slo wly evaporating ketones and esters. An oxygenated solv ent which is frequently added to h ydrocarbon solvents and has been cost effective in increasing the v ery important h ydrogen-bonding parameter has been n-butanol (or sometimes 2-butanol). The mixture of equal parts xylene and n-butanol has been widely used in conjunction with man y polymers such as epoxies, b ut a third solv ent, such as a k etone, ester, or glycol ether, is often included in small amounts to increase the polar parameter/solv ency of the mixture. Neither xylene nor n-butanol satisfactorily dissolves an epoxy of higher molecular weight by itself. These are located in boundary regions of the solubility region for epoxies, but on opposite sides of the characteristic Hansen spheres (see Figure 8.2). 1 Glycol ethers also can be added to hydrocarbon solvents with adv antage, and the polar and h ydrogen-bonding parameters are higher than if n-butanol had been added to the same concentration. There are man y possibilities, and a solubility parameter approach is particularly valuable in quickly limiting the number of candidates. The addition of glycol ethers or other w ater-soluble solv ents can ha ve adv erse ef fects, such as increased water sensitivity and poorer corrosion resistance of the final film, as some soent retention must be anticipated, and the least v olatile solvent is enriched and left behind. Relative costs for impro ving solvency from a h ydrocarbon base solv ent can be estimated by n-butanol, for the relation ( δP2 + δH2)1/2/cost. This relation has generally pointed to the use of

7248_C008.fm Page 139 Wednesday, May 23, 2007 11:05 AM

Applications — Coatings and Other Filled Polymer Systems

139

12

10

8

6 δp MEK •

4

• E GMEE •B • E GMEEA

2 ARH 0

ALH 0

2

4

6

8

10

12

δH FIGURE 8.1 Sketch showing location of typical solvents relative to the HSP of a binder. Aliphatic hydrocarbons (ALH) and aromatic h ydrocarbons (ARH) do not al ways dissolve well enough so other solv ents must be added to bring the mix ed solvent composition into the re gion of solubility for the binder . Ketones (MEK, methyl ethyl ketone), alcohols (B, n-butanol), or other solv ents such as glycol ethers and their acetates (here ethylene glycol monoeth yl ether and eth ylene glycol monoeth yl ether acetate) can be used to do this. The expected solvent improvement at least cost is discussed in the te xt as the quantity ( δP2 + δH2)1/2/cost. Units in the figure are in (cal/c 3)1/2. The choice of solv ent today w ould involve glycol ethers based on prop ylene glycol as discussed in Chapter 18. (From Hansen, C.M., Färg och Lack, 17(4), 69–77, 1971. With permission.)

example, as a cost-ef ficient sol ent to increase the h ydrogen-bonding parameter in particular . Solvents can be ranked in this manner to arrive at the least cost solutions to given solvent selection problems. Coalescing solvents in water-reducible coatings are often (but not always) those with somewhat higher hydrogen-bonding parameters than the polymer , which also means the y are w ater soluble or have considerable w ater solubility. The distribution between the w ater phase and the dispersed polymer phase depends on the relative affinities for ater and the polymer. Solvents which are not particularly w ater soluble will preferentially be found in the polymer phase. Such coalescing solvents may be preferred for applications to porous substrates, making certain they are where they are needed when the y are needed. Otherwise, w ater-soluble coalescing solv ents w ould tend to follow the aqueous phase, penetrating the substrate f aster than the polymer particles, which also get filtered out and th y are not therefore a vailable to do their job in the film when the ater evaporates. When w ater e vaporates, the solv ent must dissolv e to some e xtent in the polymer to promote coalescence. Of course, this affinity of the coalescence sol ent for the polymer is a function of its HSP relati ve to those of the polymer .

7248_C008.fm Page 140 Wednesday, May 23, 2007 11:05 AM

140

Hansen Solubility Parameters: A User’s Handbook

12

10

8 δp 6 MEK •

4

MC •

0

•B THF •

2

2

4

6

8 δH

10

12

14

16

FIGURE 8.2 Sketch sho wing formulation principles using tw o relati vely poor solv ents in combination to arrive at a good solv ent. Xylene (X) can be mix ed with n-butanol (B) to arri ve at a mixture which can be improved by additions of tetrah ydrofuran (THF), meth ylene chloride (MC), or meth yl eth yl ketone (MEK) among others. These three very volatile solvents have often been used in analytical work, paint removers, etc., because the y dissolv e all of the typical coatings binders sho wn in the figure. Labeling requirements h ve dictated other choices in more recent years. Units are (cal/cm3)1/2. (From Hansen, C.M., Färg och Lack, 17(4), 69–77, 1971. With permission.)

Amines are frequently added to w ater-reducible coatings to neutralize acid groups b uilt into polymers, thus providing a water-solubilizing amine salt. Amine in excess of that required for total neutralization of the acid groups acts lik e a solv ent. Such amine salts ha ve been characterized separately to demonstrate that the y have higher solubility parameters than either (acetic) acid or organic bases. 6 These salts are h ydrophilic and ha ve very little af finity for the polymers used i coatings, which means the y are to be found in a stabilizing role in the interf ace in the aqueous phase while still being attached to the polymer . Electrostatic repulsion contrib utes to stability as well, and the dispersed solubilized polymer can be visualized in terms of a porcupine with raised quills. Surface-active agents, whether nonionic or ionic, are also to be found where the af finities o the respective parts of their molecules dictate their placement; lik e seeks like. The hydrophilic end with a high h ydrogen-bonding parameter will seek the aqueous phase, and the h ydrophobic end will seek out an environment where energy differences are lowest (self-assembly). It might be noted here that some solv ents have surfactant-like properties as well. Eth ylene glycol monob utyl ether, in particular, has been shown to be a good coupling agent, as well as contributing to lowered surface tension.7 The hydrophobic end of such molecules may reside within the polymer if HSP relations dictate this. Otherwise, if the HSP differences are too great, the hydrophobic portion may be forced to remain in the interf acial region, not being accepted by the aqueous phase either . Increases in temperature especially lead to lo wer hydrogen-bonding parameters (see Chapter 1, Chapter 3, and Chapter 10). F or this reason, solv ents with high h ydrogen-bonding parameters, such as glycols, glycol ethers, and alcohols, become better solv ents for most polymers at higher

7248_C008.fm Page 141 Wednesday, May 23, 2007 11:05 AM

Applications — Coatings and Other Filled Polymer Systems

% H2O

141

T1 T2

T3

T1 > T2 > T3

Time FIGURE 8.3 Sketch of water uptake in a polymer as a function of temperature. Higher temperature leads to more rapid uptake and to higher equilibrium levels. Quenching to a lower temperature (arrow) leads to excess water in the film and possibly to ater blisters and delamination (see te xt for further discussion). (Reprinted from Hansen, C.M., Ne w developments in corrosion and blister formation in coatings, Prog. Org. Coat., 26, 113–120, 1995. With permission from Else vier Science.)

temperatures. This can markedly affect hot-room stability in water-reducible coatings, for example, as more of the solv ent will partition to the polymer phase, which swells, becomes more fluid, an has altered af finities for stabilizing sur ace-active agents. These may dissolv e too readily in the swollen, dispersed polymer. When carefully controlled, these temperature ef fects are an advantage in water-reducible, oven-cured coatings, leading to higher film int grity, as poor solv ents at room temperature become good solv ents in the o ven after the w ater has evaporated. A very special destructi ve effect of w ater is caused by the reduction of its h ydrogen-bonding parameter with increases in temperature. The solubility of w ater in most polymers is higher at a higher temperature than it is at a lo wered temperature because the HSP for the polymer and w ater match better at the higher temperature. It has been documented in man y cases that a rapid quench from hot water to cold water can cause blisters in coatings. 8 Previously dissolved water within the film n w becomes in excess of that soluble in the film. This can be seen in Figure 8.3 where w ater uptake curves are sho wn for three temperatures. The amount and rate of uptak e is higher for the higher temperatures. Rapid cooling to belo w the solubility limit at a lo wer temperature means the system is supersaturated. Excess w ater freed by this mechanism has been called SWEA T (soluble water e xuded at lo wered temperatures). If the SWEA T w ater cannot rapidly dif fuse out of the coating, it will appear as a separate phase, perhaps first as clusters, ut ultimately at h ydrophilic sites or at a substrate. The coating fails by blistering or delamination. This special effect has been noted by the author in coatings (alkyd, polyester, and epoxy), in rigid plastics such as poly(phen ylene sulfide) and poly(ether sulfone), and ven in EPDM rubber. Examples of measurements of this type are sho wn in Chapter 12, Figure 12.3 and Figure 12.4 for an EPDM rubber g asket and for a poly(ether sulfone) tensile bar . This effect is not restricted to w ater; it has also been seen for an epoxy coating that w as repeatedly remo ved from room temperature methanol to measure weight gain. The cooling due to the methanol evaporation was sufficient to produce methanol blisters nea the air surface of the coating because of e xcess amounts of methanol over that soluble at the lower temperature resulting from the methanol e vaporation. The use of supercritical g ases as solv ents has become possible in recent years. The solubility parameters for carbon dioxide have been reported 9 earlier and in the first edition of this handbook based on the room temperature solubility of the g as in different liquids. These are now revised as discussed in Chapter 10 to δD, δP, and δH equal to 15.7, 6.3, and 5.7. HSP v alues are reported as a

7248_C008.fm Page 142 Wednesday, May 23, 2007 11:05 AM

142

Hansen Solubility Parameters: A User’s Handbook

function of temperature and pressure. This same type of analysis can be used to e valuate the temperature and pressure ef fects for other g ases. See also Chapter 3. These parameters are found using the solvents that dissolve more than the theoretical amounts of carbon dioxide that are reported in Table 10.2. The use of such g ases is considered an adv antage for the environment, but their use has been limited to relati vely smaller items because of the size of pressure equipment. Solvent technology has also been used in a wide v ariety of other products and processes as listed by Barton. 2 One can mention the formulation of solv ent cleaners based on v egetable oils as an additional example.10 Such “green” products ha ve found increasing use, as ha ve those solvents with low volatility, low VOC, and low labeling requirements.

TECHNIQUES FOR DATA TREATMENT As mentioned earlier, a simple approach to man y practical problems is to mak e a two-dimensional plot of polar vs. h ydrogen-bonding parameters with a circle (or estimated circle) for the polymer in question. The circle should encompass the good solv ents. One can then plot points for potential solvents and quickly arri ve at a starting composition for an e xperiment. Subsequently, this can be adjusted if necessary. A linear mixing rule based on the volume (or weight) fractions of the solvent components is usually satisf actory. Plasticizers should be included in the calculations. They will be very slow to dissolve rigid polymers, in particular, and are, of course, nonvolatile for all practical purposes. A special plotting technique for solv ent selection de veloped by Teas11 is used frequently by those who restore old paintings. The art in volved in this stage of the conserv ation process is to remove the old varnish without attacking the underlying original masterpiece. HSP principles ha ve been used since the late 1960s for selecting solv ents and solv ent blends for this purpose. 12 The triangular plotting technique uses parameters for the solv ents, which, in f act, are modified HS parameters. The individual Hansen parameters are normalized by the sum of the three parameters. This gives three fractional parameters defined by Equation 8.1 to Equation 8.3 fd = 100 δD/(δD + δP + δH)

(8.1)

fP = 100 δP/(δD + δP + δH)

(8.2)

fh = 100 δH/(δD + δP + δH)

(8.3)

The sum of these three fractional parameters is 100 in the form the equations are written. This allows use of the special triangular technique. Some accurac y is lost, and there is no theoretical justification for this plotting technique, ut one does get all three parameters onto a two-dimensional plot with enough accuracy that its use has survi ved for this type of application (at least). The Teas plot in Figure 8.4 includes an estimate of the solubility/strong attack of older , dried oil paint. A varnish which could be considered for use is P araloid® B72, a copolymer of eth yl methacrylate and methyl methacrylate from Rohm and Haas. There is a re gion in the lo wer, right-hand part of this plot where the v arnish is soluble and the dried oil is not. The varnish remover should be in this region. Mixtures of h ydrocarbon solvent and ethanol are located in this re gion and could be considered. HSP correlations for materials of interest in restoration of older paintings are included in Table 8.1. A helpful simplifying relation to use in solvent selection calculations using solubility parameters is that the resultant v alues for mixtures can be estimated from v olume fraction a verages for each solubility parameter component. Solv ent quality can be adjusted by the RED number concept, which is discussed in Chapter 1 (Equation 1.10), or graphically as described abo ve.

7248_C008.fm Page 143 Wednesday, May 23, 2007 11:05 AM

Applications — Coatings and Other Filled Polymer Systems

0

143

100

fh

fp

DRIED OIL + PAR B72 E PAR B72

100 0

CH 0 100

fd

FIGURE 8.4 Teas plot for a typical painting conserv ation situation where a v arnish is to be remo ved or applied without attacking the underlying original oil painting. Solvents indicated are cyclohexane (C), heptane (H), and ethanol (E) (see te xt for further discussion).

TABLE 8.1 HSP Correlations for Materials of Interest in the Conservation of Older Paintings MATERIAL

δD

δP

δH

Ro

FIT

G/T

Paraloid® 22 solubility Dammar gum de waxed Dried oil (estimate)

17.6 18.4 16.0

7.4 4.2 6.0

5.6 7.8 7.0

9.4 8.3 5.0

1.000 0.915 1.000

17/26 30/56 9/22

A computer search with the SPHERE computer program (Chapter 1) for “nearest neighbors” for a gi ven single solv ent has been used man y times to locate alternates for a wide v ariety of product types including coatings of v arious descriptions, cleaners, etc. A similar application is to predict which other solv ents will probably be aggressi ve to a chemically resistant coating where very limited data have indicated a single solvent or two are somewhat aggressive. A nearest neighbor search involves calculation of the quantity Ra (Chapter 1, Equation 1.9) for a whole database, for example, and then arranging the printout in RED number order (Chapter 1, Equation 1.10). The potentially most aggressi ve liquids are at the top of the list. Solv ents with RED less than 1.0 are “good” and therefore easily recognized. Sorting out these possibilities considering toxicity , evaporation rate, cost, etc. leads to the most promising candidates for the substitution.

7248_C008.fm Page 144 Wednesday, May 23, 2007 11:05 AM

144

Hansen Solubility Parameters: A User’s Handbook

SOLVENTS AND SURFACE PHENOMENA IN COATINGS (SELF-ASSEMBLY) Chapter 6 and Chapter 7 ha ve been de voted to the characterization of surf aces for substrates, pigments, fillers, and the li e. This means the interplay between solvent, polymer, and surfaces can be inferred by their relati ve affinities. These depend on their HSP relati ve to each other , and the RED number concept can be quite useful. As stated pre viously, the desired solv ent quality in man y coatings is just slightly better than that of a mar ginal solvent. This means RED numbers just under 1.0 relati ve to the polymer will be sought. One reason for the desired mar ginal solv ent quality is that this will ensure that the polymer adsorbed onto pigment surf aces during pigment dispersion has little reason to dissolv e away from that surf ace. The dispersion stabilizing polymer should remain on the pigment surf ace where it is desired. If this polymer is dissolved away, the result is most likely pigment flocculation which leads to color change, undesired settling, and perhaps e ven rheological dif ficulties. The solvent in this case should have a RED number for the pigment surf ace greater than 1.0, or at least reasonably high, to aid in the planned af finity approach to pigment dispersion stabilit . Of course, the polymer, or some portion of it, and the pigment surface should have high affinity for each othe . A sketch of the optimum relations in coatings is gi ven in Figure 8.5 where the mar ginal solvent is number 1. Solvent 2 would probably be too e xpensive and, in addition, will probably dissolv e the polymer too well. Schröder13 (B ASF) confirms that the optimum polymer adsorption will be found when th binder and pigment surf ace have the same HSP. He indicates that the solv ent should be v ery poor for the pigment and located in the boundary re gion for the binder. He prefers the pigment to ha ve HSP values placing it intermediate between the solvent and binder. This is suggested for conditions where the solv ent has higher HSP than the pigment, as well as for conditions where the solv ent has lower HSP than the pigment. This situation, with the solv ent and binder on opposite sides of the pigment, means the composite v ehicle has parameters v ery closely matching those of the pigment. A v ery similar type of result w as found by Skaarup, 14 who especially emphasized that optimum color strength w as found for solv ents marginal in quality for the binder and poor for the pigment in question. Pigment

3

δp 2 1

Polymer δh FIGURE 8.5 Sketch showing influence of sol ent quality on e xpected pigment dispersion stability (see te xt and Figure 1.1 for discussion). (Reprinted from Hansen, C.M., Paint and Coating Testing Manual, 14th ed. of the Gardner -Sward Handbook, 1995, p. 400. With permission. Cop yright ASTM.)

7248_C008.fm Page 145 Wednesday, May 23, 2007 11:05 AM

Applications — Coatings and Other Filled Polymer Systems

145

In special applications, an extended polymer chain configuration is desirable, ut a solid anchor to the pigment surf ace is also desired. This means a better -than-marginal solvent for the polymer is desired. A good anchor has high af finity for the pigment sur ace and mar ginal affinity for th solvent. Solvent 3 (Figure 8.5) w ould adsorb onto the pigment surf ace preferentially, and pigment dispersion stability would be poor. An extension of this thinking may be required for pigment pastes and other very highly filled products. In these cases, there is little dispersing ehicle relative to the pigment, and the solv ent must be considered as being part of the dispersing v ehicle. In such cases the solvent may ha ve high af finity for the pigment sur ace as well as for dispersing polymer . An ideal situation here is where all the ingredients ha ve the same HSP.

POLYMER COMPATIBILITY In some cases, closer -than-usual matches between solv ent and polymer solubility parameters are required. This is true when tw o polymers are mix ed and one of them precipitates. This is most likely the polymer with the lar ger molecular weight, and it must be dissolv ed even better. Lower RED numbers with respect to this polymer are desired, while still maintaining af finity for th other polymer. Miscible blends of tw o polymers ha ve been systematically found using a solv ent mixture composed exclusively of nonsolvents.15 This is demonstrated schematically in Figure 8.6, where it can be seen that different percentage blends of solvents 1 and 2 will have different relative affinities for the polymers. No other alternat ve theory of polymer solution thermodynamics can duplicate this predictive ability. Polymer miscibility is enhanced by lar ger overlapping solubility regions for the polymers as sk etched in Figure 8.7. Polymers A and B should be compatible, whereas polymer C would not. Such a systematic analysis allows modification of a g ven polymer to provide more o verlap and enhanced compatibility . The advantages of a copolymer containing the monomers of A or B and C should also be e vident. Such a copolymer will essentially couple the system together.

1

B 1+2

δp A 2

δh FIGURE 8.6 Sketch showing how two otherwise immiscible polymers can be brought into a homogeneous solution by the use of mix ed nonsolvents. (Reprinted from Hansen, C.M., Paint and Coating Testing Manual, 14th ed. of the Gardner -Sward Handbook, 1995, p. 400. With permission. Cop yright ASTM.)

7248_C008.fm Page 146 Wednesday, May 23, 2007 11:05 AM

146

Hansen Solubility Parameters: A User’s Handbook

C δp

A B

δh FIGURE 8.7 Sketch describing e xpected polymer miscibility relations (see te xt for discussion). (Reprinted from Hansen, C.M., Paint and Coating Testing Manual, 14th ed. of the Gardner-Sward Handbook, 1995, p. 401. With permission. Cop yright ASTM.)

Van Dyk et al. 16 ha ve correlated the inherent viscosity of polymer solutions with HSP inherent (intrinsic) viscosity used in this study , [ η], is given by Equation 8.4. [η] = ( ηs/η0)/c

. The (8.4)

ηs is the solution viscosity , η0 is the solv ent viscosity, and c is the solution concentration. The concentration used was about 0.5 g/dl. This is an expression reflecting polymer chain xtension in solution, with higher v alues reflecting greater chain entanglements because of greater polyme extension. This is interesting in that the solubility parameter is a thermodynamic consideration, whereas the viscosity is a kinetic phenomenon. Higher [ η] were found for solv ents with HSP nearest those of the polymer . As stated above additional uses of HSP (and the total solubility parameter) in solv ent technology can be found in Barton,2 but these are too numerous to include here. However, a couple of examples relating to guided polymer compatibility are w orthy of special mention. These are the formulation of asymmetric membranes for separations, 17,18 where polymer solutions — ha ving gi ven HSP relations — and at least one solv ent soluble in w ater are used. The solution is immersed in w ater, the solvent quality becomes bad, and a controlled porous membrane results. Another example of controlled phase relations during a dynamic process is found in the formulation of self-stratifying coatings. This is discussed in Chapter 6 in terms of the creation of interf aces and therefore interfacial surface tension. The HSP principles involved in this type of coating can be seen in Figure 8.8. The solvent must dissolv e both the topcoat and primer and allo w the lo wer surface tension topcoat to migrate to the surf ace during film formation. ormulation principles have been discussed in detail elsewhere.19,20 Before concluding this section, some of the recent work on miscible polymer blends should also be noted. 21,22 This work used group contribution estimates of the δP and δH parameters only in an effort to correlate interfacial tension between polymers, assuming that the δD parameters would not be too dif ferent. Although this is a good starting point to pro ve the procedure has possibilities, further differentiation between the polymers and improved group contribution methods may offer even more impro vement.

7248_C008.fm Page 147 Wednesday, May 23, 2007 11:05 AM

Applications — Coatings and Other Filled Polymer Systems

147

POLAR PARAMETER

PRIMER

PARAMETERS REQUIRED FOR COMMON SOLVENT TOP COAT (LOWEST ENERGY)

HYDROGEN BONDING SOLUBILITY PARAMETER

FIGURE 8.8 Sketch illustrating the principles of solv ent selection for self-stratifying coatings. (From Birdi, K.S., Ed., Handbook of Surface and Colloid Chemistry, CRC Press, Boca Raton, FL, 1997, p. 324. With permission.)

HANSEN SOLUBILITY PARAMETER PRINCIPLES APPLIED TO UNDERSTANDING OTHER FILLED POLYMER SYSTEMS Recent characterizations of inor ganic fillers and fibe 23 have confirmed that HSP concepts can b applied to engineered fibe -filled systems such as those based on polypro ylene. The behavior of chewing gum can also be analyzed in terms of solubility parameter principles.24 In addition to rheological beha vior, appearance, and other performance considerations, a desired product characteristic is that the release of the taste components should be controlled. Greater differences in solubility parameters between fl voring agents and w ax-free gum bases lead to enhanced fl vor release. Similarity of HSP can lead to stopping the desired release too soon. Perhaps the most important practical w ork dealing with solubility parameters and the stability of pigment dispersions is that attributable to Stephen.25 He concludes that all the (solid) ingredients in a paint formulation should ha ve the same ener gy characteristics. If the y do not, there will be a driving force for this to occur. This can lead to problems. One can just as well make the formulation stable from the start, and then e verything will remain stable just where it is because there are no driving forces for anything to move around. Although this sounds expensive, obvious, and perhaps too simple, the truth of the matter is well documented in v ery practical terms.

CONCLUSION Many practical uses of the solubility parameter concept ha ve been described in detail, including optimizing solvent selection, improving polymer compatibility, and enhancing pigment dispersion. When all of the materials in volved in a gi ven product and application can be characterized with the same af finity (solubility/cohesion) parameters, the possibility xists to predict interactions

7248_C008.fm Page 148 Wednesday, May 23, 2007 11:05 AM

148

Hansen Solubility Parameters: A User’s Handbook

among them. This is true e ven in complicated situations, such as the formulation of v arious types of filled systems including coatings, printing inks, adhes ves, and other filled polymer system including chewing gum.

REFERENCES 1. Hansen, C.M., Solubility in the coatings industry , Färg och Lack, 17(4), 69–77, 1971. 2. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press, Boca Raton, FL, 1983; 2nd ed., 1991. 3. Gardon, J.L. and Teas, J.P., Solubility parameters, in Treatise on Coatings, Vol. 2, Characterization of Coatings: Physical Techniques, P art II, Myers, R.R. and Long, J.S., Eds., Marcel Dekk er, Ne w York, 1976, chap. 8. 4. Beerbower, A., Boundary Lubrication — Scientific and Technical Applications Forecast, AD747336, Office of the Chief of Research and D velopment, Department of the Army, Washington, D.C., 1972. 5. Hansen, C.M. and Beerbo wer, A., Solubility parameters, in Kirk-Othmer Encyclopedia of Chemical Technology, Suppl. Vol., 2nd ed., Standen, A., Ed., Interscience, Ne w York, 1971, pp. 889–910. 6. Hansen, C.M., Some aspects of acid/base interactions (Einige Aspekte der Säure/Base-W echselwirkung, in German), Farbe und Lack, 83(7), 595–598, 1977. 7. Hansen, C.M., Solv ents in w ater-borne coatings, Ind. Eng. Chem. Prod. Res. Dev., 16(3), 266–268, 1977. 8. Hansen, C.M., Ne w developments in corrosion and blister formation in coatings, Prog. Org. Coat., 26, 113–120, 1995. 9. Hansen, C.M., 25 years with solubility parameters (25 År med Opløselighedsparametrene, in Danish), Dan. Kemi, 73(8), 18–22, 1992. 10. Rasmussen, D. and Wahlström, E., HSP-solubility parameters: a tool for development of new products — modelling of the solubility of binders in pure and used solv ents, Surf. Coat. Int., 77(8), 323–333, 1994. 11. Teas, J.P., Graphic analysis of resin solubilities, J. Paint Technol., 40(516), 19–25, 1968. 12. Torraca, G., Solubility and Solvents for Conservation Problems, 2nd ed., International Centre for the Study of the Preservation and the Restoration of Cultural Property (ICCR OM), Rome, 1978. (13, Via Di San Michelle, 00153 Rome) 13. Schröder, J., Colloid chemistry aids to formulating inks and paints,Eur. Coat. J., 5/98, 334–340, 1998. 14. Skaarup, K., The three dimensional solubility parameter and its use. II. Pigmented systems, skandinavisk tidskrift for Fårg och Lack, 14(2), 28–42, 1968; 14(3), 45–56, 1968. 15. Hansen, C.M., On application of the three dimensional solubility parameter to the prediction of mutual solubility and compatibility, Färg och Lack, 13(6), 132–138, 1967. 16. Van Dyk, J.W., Frisch, H.L., and Wu, D.T., Solubility, solvency, solubility parameters, Ind. Eng. Chem. Prod. Res. Dev., 24(3), 473–478, 1985. 17. Klein, E. and Smith, J.K., Assymetric membrane formation, Ind. Eng. Chem. Prod. Res. Dev., 11(2), 207–210, 1972. 18. Chawla, A.S. and Chang, T.M.S., Use of solubility parameters for the preparation of hemodialysis membranes, J. Appl. Polym. Sci., 19, 1723–1730, 1975. 19. Misev, T.A., Thermodynamic analysis of phase separation in self-stratifying coatings — solubility parameters approach, J. Coat. Technol., 63(795), 23–28, 1991. 20. Special issue de voted to self-stratifying coatings, Prog. Org. Coat., 28(3), July 1996. 21. Luciani, A., Champagne, M.F., and Utracki, L.A., Interfacial tension in polymer blends. Part 1: Theory, Polym. Networks Blends, 6(1), 41–50, 1996. 22. Luciani, A., Champagne, M.F ., and Utracki, L.A., Interf acial tension in polymer blends. P art 2: Measurements, Polym. Networks Blends, 6(2), 51–62, 1996. 23. Hennissen, L., Systematic Modification of Filler/Fibre Sur aces to Achieve Maximum Compatibility with Matrix Polymers, Lecture for the Danish Society for PolymerTechnology, Copenhagen, February 10, 1996.

7248_C008.fm Page 149 Wednesday, May 23, 2007 11:05 AM

Applications — Coatings and Other Filled Polymer Systems

149

24. Song, J.H. and Reed, M.A., Petroleum Wax-Free Chewing Gums Ha ving Improved Flavor Release, U.S. Patent No. 5,286,501, February 15, 1994, assigned to Wm. Wrigley Jr. Company, Chicago, IL. 25. Stephen, H.G., Parameters controlling colour acceptance in late x paints, J. Oil Colour Chem. Assoc., 69(3), 53–61, 1986.

7248_C008.fm Page 150 Wednesday, May 23, 2007 11:05 AM

7248_book.fm Page 151 Tuesday, April 24, 2007 9:19 AM

Solubility Parameters 9 Hansen of Asphalt, Bitumen, and Crude Oils Per Redelius ABSTRACT Hansen solubility parameters (HSP) are sho wn to be a useful ne w tool for understanding compatibility relations among bitumens and crude oils. Bitumen and crude oils are comple x mixtures of hydrocarbons which are k ept in solution mainly by their mutual solubility . They are not colloidal dispersions as previously thought. Although the solubility of the hydrocarbons is mainly determined by the dispersive interactions, it is not possible to mak e correct estimates of their stability without also taking polar interactions and h ydrogen-bonding interactions into consideration. HSP ha ve proven their ability to gi ve a good estimate of the stability of bitumen and/or crude oil ha ving different origins in relation to solv ents and polymers. Relations between the HSP of dif ferent materials is visualized using 3D plots showing the HSP as ellipsoids. A more precise determination of the extension of the ellipsoids can be found by turbidimetric titrations with three different titrants, each representing a direction in the HSP space, respecti vely. It is no w possible with the help of simple laboratory e xperiments to predict the consequences of dif ferent courses of action, thus eliminating expensive trial and error testing.

SYMBOLS SPECIAL TO CHAPTER 9 C P FR pa po

Amount of bitumen/total amount of solv ent and titrant Stability index given by Equation 9.3 Volume of solv ent/total volume of solv ent plus titrant Defined by Equation 9. Defined by Equation 9.

INTRODUCTION Even if most of us are not familiar with bitumen, we all know the “black” roads on which we drive every day. The majority of road surf aces are black because the binding agent used to manuf acture the surfacing is bitumen, which is mix ed with crushed rock aggre gate. Road surf aces can also be grey to white in color, in which case an alternative binder has been used: Portland cement concrete. Bitumen is a semisolid material that can be produced from certain crude oils by distillation. It can also be found in nature as “natural asphalt.” It consists of a mixture of hydrocarbons of different molecular sizes containing small amounts of heteroatoms such as sulfur , nitrogen, and oxygen, as well as traces of metals like vanadium and nickel. Bitumen behaves as a viscoelastic thermoplastic solid at ambient temperature and turns into a viscous liquid at high temperature. It presents unique adhesive and waterproofing properties, which ma e it ideal in the manuf acture of asphalt for road 151

7248_book.fm Page 152 Tuesday, April 24, 2007 9:19 AM

152

Hansen Solubility Parameters: A User’s Handbook

construction and to use in a wide range of industrial application, from waterproofing in constructio to sound dampening in the automoti ve industry. The term bitumen is not completely unambiguous as it has been gi ven different meanings in different parts of our world. In Europe the term is defined as ab ve, whereas in Canada, for example, it is used for hea vy crude oils. In the U.S. the term asphalt is used instead of bitumen. Sometimes bitumen is confused with tar, which is a product of completely dif ferent origin. Tar is produced by dry distillation of coal or w ood. The most common process for production of bitumen is by distillation under acuum v of properly selected crude oils. There are ho wever just a limited number of crude oils which permit direct distillation to proper bitumen grades suitable for production of road asphalt. Although the reserves of such crude oils are v ery large worldwide, they are not primarily produced as the y contain too small amounts of fuel, which is the most important and profitable product for refiner The functional properties of bitumen are usually related to its use as binder in asphalt for roads. Thus, the most common properties are related to the rheology of bitumen. As the road construction area is very conservative, and bitumen has been used for about 100 years, most tests are empirical and have been used for a long time. Two of the most common tests are penetration at 25 °C and softening point Ring&Ball. The penetration gives a measure of the stiffness of the bitumen at most common service temperatures of a road, whereas the Ring&Ball gi ves the stif fness close to the highest expected temperature in practice. In Europe bitumens are graded according to their penetration at 25 °C — for e xample, 50/70, where the tw o numbers gi ve the highest and lo west limit for the particular grade. It is also common, particularly in the U.S., to use viscosity gradation based on viscosity at 60 °C. Bitumen is, ho wever, a viscoelastic material with a comple x rheology and can thus not be completely described by simple penetration testing and softening point. The development of modern and reliable rheometers — for e xample, the dynamic shear rheometer (DSR) — has made it possible to describe the full rheology of bitumen. During the last 20 years we ha ve seen an increase in the use of polymer modified bitume (PMB) with impro ved properties. The main reason for modification of bitumen is to impr ve the rheological properties, particularly to mak e the binder less sensiti ve to temperatures. It is desired to ha ve a reasonable stif fness of the binder e ven at the highest surf ace temperatures a road can reach on a hot summer day , as well as being reasonably fl xible at the lo west temperatures on a cold winter day. Another reason for modification with polymers is to increase durabilit . This will be improved if a proper polymer is selected. A large number of dif ferent polymers ha ve been tested as modifiers for bitumen. In the end just a fe w of them ha ve reached lar ger commercial use. The main restriction for the choice of polymer is the e xpected improvement of the rheological properties in comparison with the price of the polymer . But e ven more important is the compatibility or the solubility of the polymer in the bitumen. Until now, there have been very few tools for prediction of compatibility between the polymer and bitumen, so the de velopment of new PMB has to a lar ge extent been done on a “trial and error” basis. The better understanding of the true nature and the solubility properties of bitumen provided by Hansen solubility parameters (HSP) has given a new tool for understanding of polymer compatibility with bitumen as discussed in the follo wing.

MODELS OF BITUMEN Crude oils ha ve been found in man y places around the w orld. Although the true origin of crude oils is still under discussion, most scientists agree that the y have been formed by de gradation and transformation of ancient or ganisms. The properties of crude oil v ary depending on age and conditions during formation. Some crude oils are liquids with lo w viscosity, whereas others are semisolid materials that ha ve a viscosity making them impossible to handle at room temperature. The low viscosity crude oils contain lar ge amounts of fuel b ut very little bitumen, if an y, and the high viscosity crude oils contain v ery little fuel b ut large amounts of bitumen.

7248_book.fm Page 153 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

153

From a chemical point of vie w crude oil is an e xtremely complex mixture of h ydrocarbons. Usually small amounts of heteroatom lik e nitrogen, oxygen, and sulfur , as well as trace amounts of metals like vanadium and nickel, are found, although the content v aries depending on type and origin of the crude oil. The smallest molecules are the gaseous methane, ethane, and propane. These are dissolved in the liquid h ydrocarbons. The heaviest molecules ha ve molecular weights higher than 1000 and are thus h ydrocarbons with 70 carbon atoms or more. The separation of crude oils into different fractions is done in refineries by distillation, with the diferent fractions being collected based on their boiling points. The lo w-boiling fractions consist of g asoline and g as oils. The constituents in these fractions have been characterized by modern analytical techniques until almost every single component has been identified. The heavier fractions (heavy gas oil), and particularly the residue after distillation, ha ve escaped such detailed characterization. Most residual oils are further upgraded by dif ferent refining processes to fuels. Bitumen may be produced only after proper distillation process of a selected crude oil using v acuum. Although the residual oil and bitumen have been e xtensively analyzed with modern equipment, most of the understanding is in terms of a verages of dif ferent chemical functional groups or structures. From these data tentati ve structures of the molecules ha ve been suggested. 1 In fact hardly any one single molecule from the complex mixture has been chemically analyzed. There are se veral reason wh y this has been a superior challenge: • • • • •

The number of dif ferent molecules is v ery large. There is no major population of identical molecules. The material is black and viscous. The range of molecules of dif ferent polarities and sizes is continuous. The boiling point is higher than approximately 450°C, making the molecules afirly large.

The most common approach for chemical characterization of bitumen in volves a separation into generic fractions based on chromatographic principles.The most common separation procedure is called SARA analysis (saturates, aromatics, resins, and asphaltenes). It consists of two principally different steps: first, creation and precipitation of a solid fraction by dilution of the bitumen wit n-heptane, and then a separation of the soluble fraction with respect to polarity . The precipitated fraction is called asphaltenes and is defined as the fraction of bitumen that is insoluble inn-heptane. The n-heptane soluble fraction is named maltenes and is further separated by polarity into three more fractions. These fractions have been given names like “resins,” “aromatics,” and “saturates.” The most common and widespread h ypothesis about the structure of bitumen, which is found in most books and papers on bitumen chemistry , teaches that bitumen is a colloidal dispersion of asphaltenes in maltenes. The dispersion is assumed to be stabilized by the resins. The first one t introduce this concept w as Nelensteyn (1924). 2 The model was later refined by Pfei fer and Saal. 3 Although the model might be attracti ve for mechanical engineering, it is more dif ficult to accep for an organic chemist, particularly since colloidal dispersions of h ydrocarbons in other h ydrocarbons are rare, e xcept in the case of polymers. A number of questions are immediately raised: “Do the asphaltenes have enough different chemistry to permit dispersion rather than dissolution?” and “If it is a colloidal dispersion, what is the mechanism for its stabilization?” Other models that question the e xistence of micelles ha ve also been proposed. Examples of models are the continuous thermodynamic model by P ark and Mansoori 4 and Buduszynski et al., 5 and the micro structural model as a result from the SHRP development program in the U.S.6 Recent research has shown that the asphaltenes do not form micelles b ut are soluble in the maltenes, and thus no micelles can exist in the bitumen.7,8 These models describe bitumen as a solution of organic material of different polarity and dif ferent molecular weight ha ving a kind of mutual solubility in each other. When a solvent such as n-heptane is added to the system, the balance is disturbed. P art of the system precipitates. The precipitation beha vior of asphaltenes is what could be predicted from regular solution theory and could be described as flocculation. In spite of the solubility model

7248_book.fm Page 154 Tuesday, April 24, 2007 9:19 AM

154

Hansen Solubility Parameters: A User’s Handbook

being a more precise description of the true nature of bitumen, it has recei acceptance in the research on bitumen and crude oils.

ved surprisingly lo w

ASPHALTENES During production, transport, and refining of certain crude oils there are sometimes problems wit the formation of precipitates and deposits. The deposits have been claimed to be asphaltenes, and therefore there is considerable interest in them and mone y spent to sa ve, if the formation of precipitates could be controlled. Thus, e xtensive research has been performed to in vestigate the chemistry of asphaltenes9,10 as well as mechanisms of formation of the precipitates. There is a clear definition of the term asphaltenes11,12 as the material that precipitates on dilution of bitumen or crude oil with n-heptane. Most of the characterization w ork has been conducted on precipitated asphaltenes, and very little attention has been gi ven to asphaltenes in their natural en vironment in the bitumen. Much confusion has come from the misuse of the term asphaltenes to mean all kinds of precipitates from bitumen, suggesting that the insolubles inn-heptane could represent precipitates in general. This assumption might have been correct if the asphaltenes were a colloidal fraction in bitumen, but this it is not the case. As will be pro ven later in this chapter the cause of formation of precipitates is more related to general solubility rather than just solubility in n-heptane. The mechanism of precipitate formation is certainly not only an academic matter b ut is of major importance for the whole oil industry as precipitates may cause blocking and fouling of equipment used in crude oil production as well as in transport and refining. It is orth discussing some of the more common statements about the chemistry of asphaltenes and to compare them with e xperimental facts.

MOLECULAR WEIGHT A general statement about the molecular weight of asphaltenes w ould be that they are high molecular weight material. The true molecular weight of asphaltenes has been under discussion for man y years. Investigations using vapor phase osmometry (VPO) on precipitated asphaltenes dissolv ed in different solvents have shown molecular weights from 1000 up to 10000, depending on the source of asphaltenes. The apparent molecular weight is strongly dependent on the solv ent. This indicates that the asphaltenes associate in solution. 1 Other attempts to determine molecular weight using field ionization mass spectrometry (FIMS) r veal an apparent molecular weight of 700–1000.These results also vary depending on crude oil source. 5 It is ob vious that the VPO overestimates the true molecular weight due to interactions between the molecules, and FIMS lik ely gives a more correct value, although there might be a risk that some de gradation has tak en place in the ion source. Recent studies with fluorescence depolarization techniques h ve confirmed the FIMS results 13 It may be speculated that lar ge size molecules are less soluble in n-heptane, and thus asphaltenes should consist mainly of high molecular weight material. A high dependency of molecular weights on solubility is well kno wn from polymers. There are, ho wever, se veral h ydrocarbons of lo wer molecular weight that are not soluble inn-heptane (for example, coronene or dibenz(a,h)anthracene, where the very high aromatic content leads to v ery high dispersion parameters compared with the relatively low dispersion parameter for n-heptane in the HSP concept), and similar molecules may be part of the asphaltenes fraction. It is thus reasonable to assume that the lowest molecular weight in the asphaltenes is equal to the smallest molecule with a boiling point at the cut-point of the bitumen. This varies with different crude oils b ut may be estimated as being 500°C. This roughly corresponds to h ydrocarbons with 35 carbon atom, less for polyc yclic aromatics and more for nalkanes. The conclusion is that the asphaltenes fraction lik ely consists of a range of molecules of different molecular weight, which might range from as lo w as 300 up to more than 1000.

7248_book.fm Page 155 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

155

POLARITY

Asphaltenes are claimed to be a “highly polar” fraction in bitumen, in contrast to the more nonpolar oils (maltenes). This statement is based on the f act that asphaltenes are insoluble in n-heptane, a nonpolar solvent. The asphaltenes are, ho wever, easily soluble in relati vely nonpolar solvents like benzene, toluene, and dichloromethane, whereas the y are insoluble in polar solv ents lik e w ater, glycerine, and methanol. It is thus more correct to state that the asphaltenes are not polar in a chemical sense, but they might be considered as more polar than the other hydrocarbons in bitumen and crude oil. As nitrogen and oxygen are the only atoms in asphaltenes that could contrib ute significantly to a permanent polarit , an estimation of the relati ve polarity can be made by considering the amounts of nitrogen and oxygen atoms compared to the amount of carbon atoms. Elemental analyses have revealed that the total amount of oxygen and nitrogen in the asphaltenes is usually lo wer than 4%. 14 This is not more than about one to three nitrogen and oxygen atoms per asphaltene molecule assuming a molecular weight of about 1000. This is not enough to mak e them particularly polar . The apparent polarity might, ho wever, be increased by the content of polyaromatic compounds in some asphaltenes. These are polarizable and thus may act as polar molecules in contact with other polar molecules. In spite of this, the asphaltenes remain mainly nonpolar, and the claims that they are highly polar have without any doubts been misleading in the attempts to understand the role of the asphaltenes in bitumen and crude oils.

SOLUBILITY PARAMETERS OF BITUMEN The first attempts to determine the solubility parameters of bitumens were made using the Hilde brand solubility parameter concept. 15–20 The focus in these investigations was to study the onset of precipitation of asphaltenes and their solubility properties. In these in vestigations traditional systems using ratios between a good solv ent and a poor solv ent are used. The choice of good solv ent was usually toluene and the poor solv ent was usually n-heptane, but sometimes other n-alkanes were used. This approach gives reasonably good results, as long as it is in accordance with the definitio of asphaltenes. As bitumen and crude oil mainly consist of h ydrocarbons, the simple Hildebrand solubility parameters were belie ved to gi ve a good prediction of solubility properties. When the solubility properties of bitumen are e xtended to more v aried types of solv ents than aromatic and aliphatic hydrocarbons, the good agreement with the Hildebrand solubility parameter is to some extent lost. 21 The authors of Reference 22, for e xample, found that all good solv ents for bitumen fall between = 15 MP a1/2 and = 23 MP a1/2, but not all solv ents in this range were good solv ents. This sho ws that the Hildebrand solubility parameters are not appropriate for bitumen, probably because there are other interactions between the molecules that are not tak en into consideration. The authors of this paper and others23 found that using two-dimensional solubility parameters gives a better description of the solubility properties, b ut the best estimation w as given by the Hansen three- dimensional solubility parameter.24,25 There are still some de viations. This indicates that the prediction could be slightly impro ved if more than three types of interactions are used, b ut this will make the model unnecessarily complicated. Determination of solubility parameters of bitumen and crude oil is rather complicated as these consist of a very complex mixture of hydrocarbons. In fact, it is not completely evident that solubility parameters should be applicable for such mixtures, and particularly not if the assumed colloidal model would be correct. Use of common methods based on ph ysical and chemical parameters, which easily can estimate the solubility parameter of pure compounds, cannot be applied to such complicated mixtures as bitumen. The best approach is probably to mak e solubility tests of the material in a lar ge number of solvents with known solubility parameters and then try to find the best verage of the good solvents. Ev en this seemingly simple approach is rather complicated, ho wever, when applied to bitumen. The first complication comes from the act that bitumen is very black, and it is rather difficul to see with the e ye whether the solution is clear or not. Another complication is that se veral solvents

7248_book.fm Page 156 Tuesday, April 24, 2007 9:19 AM

156

Hansen Solubility Parameters: A User’s Handbook

may partly dissolve the bitumen, leaving a small precipitate or residue. The third complication is that one has to tak e the mutual solubility between the bitumen molecules into consideration. The effect of the mutual solubility is that a higher concentration of bitumen results in better solubility , which is contradictory to normal solubility theory that teaches that a saturation le vel for the solute is reached. In this case solubility becomes better for higher concentrations.

TESTING OF BITUMEN SOLUBILITY Solubility testing may be used for calculation of the solubility parameters of bitumen by the method given in the follo wing. In the testing of the solubility we find that most sol ents give a kind of partial solubility with more or less residue. As bitumen is v ery black, it is sometimes dif ficult t notice small traces of precipitate. In uncertain cases a drop of the solution can be placed on a filte paper. If a black dot appears at the spot of the drop, the solution contains precipitate, b ut if the staining of the filter paper is a uniform darkish br wn, it does not contain an y precipitate. As it is so difficult to estimate true solubilit , it is sometimes better to gi ve a grading of the solubility in several steps, although the final calculation requires only “soluble” or “not soluble ” In an e xperiment using 15 different bitumens, the solubility was determined in 6 different grading levels, ranging from completely soluble to completely insoluble. 26 Each le vel of solubility w as designated as a solubility grade according to the follo wing rules: 1. 2. 3. 4. 5. 6.

Totally dissolved: no residue by filter paper test Almost totally dissolv ed: light residue w as noticed by filter paper test Partly dissolved: large residue w as noticed in dark bro wn liquid. Slightly dissolved: large residue w as noticed in red-bro wn liquid. Very slightly dissolv ed: mainly residue in bro wnish liquid. Not dissolved: colorless or almost colorless liquid.

The bitumens were selected to co ver a wide v ariation of dif ferent properties. Some samples were taken from the mark et, and some were made e xperimentally for this purpose. It is known that the solubility of bitumen is concentration dependent.Thus, a fi ed concentration was used in all experiments to get comparable data. In all experiments, 0.5 g bitumen was dissolved in 5 ml solv ent. In most cases the samples were left to dissolv e for at least 24 h and sometimes for up to 48 h.

HILDEBRAND SOLUBILITY PARAMETERS Solubility data for 15 different bitumens are given in Table 9.1. All solvents with no visible residue (grade 1) were considered as “good solv ents,” and all others were considered “poor solv ents.” A bar diagram of the solv ents for bitumen No. 1 in relation to the Hildebrand solubility parameter is given in Figure 9.1. It is e vident that the majority of the “good solv ents” can be found in a range between = 17.8 MP a1/2 and = 25.8 MP a1/2, but it is also ob vious that several “poor solv ents” are found in the same range. The range of solubility parameters is slightly higher than claimed in Reference 21, which is probably due to a slightly different selection of solvents and bitumen types. The results confirm the earlier findings that the Hildebrand solubility parameter is of little or value to predict solubility properties or compatibility between solv ents or other materials with bitumen. One may speculate that the reason could be the presence of other kinds of interactions in bitumen such as, for e xample, polar interactions, h ydrogen bonding, or π-interactions between the molecules. If these interactions are of significant importance, it xplains the poor correlation with the Hildebrand solubility parameter, and also indicates that a better correlation may be achie ved when more interactions are tak en into consideration.

7248_book.fm Page 157 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

157

TABLE 9.1 Solubility Test of 15 Different Bitumens in 42 Different Solvents Solvent HSP No. and Name

1

2

3

4

5

Bitumen Sample – Code No. 6 7 8 9 10 11

12

13

14

15

7 – Acetone 11 – Acetophenone 46 – Aniline 52 – Benzene 92 – 1-Butanol 102 – n-Butyrolactone 115 – y-Butyrolactone 122 – Carbon tetrachioride 148 – Chloro benzene 182 – Cyciohe xanol 209 – Diacetone alcohol 234 – Dichlorobenzene 255 – Dieth ylether 263 – Dieth ylene glycol 297 – Dimeth ylformamide 303 – Dimeth ylsulfoxide 306 – 1,4-Dioxan 325 – Ethanol 326 – Ethanolamine 328 – Eth yl acetate 367 – 1,2-Dichloroethane 368 – Eth ylene glycol 375 – Eth ylene glycol b utyl ether 376 – Eth ylene glycol eth yl ether 380 – Eth ylene glycol meth yl ether 397 – F ormamide 417 – n-He xane 438 – Isophorone 456 – Methanol 481 – Meth ylethyl ketone 491 – Meth ylisobutyl ketone 521 – N-Meth yl-2-pyrrolidone 524 – Meth ylene chloride 532 – Nitroethane 534 – Nitromethane 536 – 2-Nitropropane 584 – Prop ylene carbonate 585 – Prop ylene glycol 617 – Tetrahydrofuran 637 – Toluene 649 – Trichloroethylene 698 – Xylene

4 1 4 1 5 2 5 1 1 3 5 1 3 6 5 5 3 5 6 4 1 6 3 5 5 6 3 1 6 3 2 3 1 5 5 4 6 6 1 1 1 1

4 2 4 1 5 2 5 1 1 4 5 1 3 6 4 5 3 6 6 4 1 6 3 5 5 6 3 1 6 3 1 3 1 4 5 4 5 6 1 1 1 1

4 2 4 1 5 3 5 1 1 5 5 1 3 5 4 5 3 5 5 4 1 6 4 5 5 6 3 1 6 4 3 3 1 4 5 4 6 6 1 1 1 1

4 3 4 1 5 2 5 1 1 3 5 1 3 6 4 5 3 5 5 4 1 6 3 5 5 6 3 1 6 3 2 3 1 4 5 4 5 6 1 1 1 1

4 3 4 1 5 2 5 1 1 3 5 1 3 6 5 5 3 6 6 4 1 6 3 5 5 6 3 1 6 3 2 3 1 5 5 4 6 6 1 1 1 1

4 1 4 1 5 2 6 1 1 3 5 1 2 6 4 5 3 6 6 3 1 6 3 5 5 6 3 1 6 3 2 3 1 5 5 5 6 6 1 1 1 1

Solubility Grade 4 4 4 4 1 1 2 2 4 5 4 4 1 1 1 1 5 5 5 5 2 3 3 2 5 5 5 5 1 1 1 1 1 1 1 1 3 4 3 3 5 5 5 5 1 1 1 1 2 3 3 2 6 6 6 6 4 5 4 4 5 5 5 5 2 3 3 3 6 6 6 6 6 6 6 6 3 3 4 4 1 1 1 1 6 6 6 6 3 3 3 3 5 5 5 5 5 5 5 5 6 6 6 6 3 3 3 3 1 1 1 1 6 6 6 6 3 3 3 3 2 3 2 2 3 4 3 3 1 1 1 1 4 5 4 5 5 5 5 5 4 4 4 4 6 6 6 6 6 6 6 6 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

4 2 4 1 5 2 5 1 1 3 5 1 2 6 4 5 3 5 5 4 1 6 3 4 5 6 3 1 6 3 2 3 1 4 4 3 5 6 1 1 1 1

4 2 4 1 5 2 5 1 1 3 5 1 2 6 4 5 3 6 6 3 1 6 3 5 5 6 3 1 6 3 2 3 1 4 5 4 5 6 1 1 1 1

4 1 5 1 5 2 5 1 1 5 5 1 3 6 4 5 3 6 6 4 1 6 3 5 6 6 3 1 6 3 2 3 1 4 5 4 6 6 1 1 1 1

4 2 4 1 5 2 5 1 1 5 5 1 3 6 4 5 3 6 6 4 1 6 3 5 5 6 3 1 6 3 3 3 1 5 5 4 6 6 1 1 1 1

Note: The solubility is graded from 1 (completely soluble) to 6 (completely insoluble).

4 2 4 1 5 2 5 1 1 3 5 1 2 6 4 5 3 6 6 4 1 6 3 5 5 6 3 1 6 3 2 3 1 4 5 4 6 6 1 1 1 1

7248_book.fm Page 158 Tuesday, April 24, 2007 9:19 AM

158

Hansen Solubility Parameters: A User’s Handbook

FORMAMIDE ETHYLENGLYCOL ETHANOLAMINE PROPYLENGLYCOL METHANOL DIETHYLENGLYCOL PROPYLENCARBONAT DIMETHYLSULFOXIDE ETHANOL γ-BUTYROLACTONE DICHLORBENZENE NITROMETHANE DIMETHYLFORMAMIDE GLYCOLMETHYLETHER GLYCOLETHYLETHER METHYLPYRROLIDONE NITROETHANE ANILINE CYCLOHEXANOL 1-BUTANOL CHLORBENZEN ACETOPHENONE GLYCOLBUTYLETHER 1,2-DICHLOROETHANE 2-NITROPROPANE 1,4-DIOXAN METHYLENECHLORIDE ISOPHORONE ACETONE TETRAHYDROFURAN METHYLETHYLKETONE TRICHLORETHYLENE BENZENE TOLUENE ETHYL ACETATE o-XYLENE C-TETRACHLORIDE n-BUTYLACETATE DIACETONALCOHOL METHYLBUTYLKETONE DIETHYLETHER n-HEXANE

0.0

5.0

10.0

15.0

20.0

25.0

Hildebrand solubility parameter

30.0

35.0 MPa0.5

FIGURE 9.1 Solubility of bitumen No 1 (T able 9.1) in dif ferent solv ents of kno wn Hildebrand solubility parameter. White bars = poor solv ents, gray bars = good solv ents.

HANSEN SOLUBILITY PARAMETERS (HSP) The data set for bitumen No.1 in Table 9.1 was used for testing whether HSP gi ves a better model for bitumen solubility than Hildebrand solubility parameters. HSP consists of three components, each gi ving a quantitati ve v alue for the dispersion (D), polar (P), and h ydrogen bonding (H)

7248_book.fm Page 159 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

159

insoluble ▼ soluble

Bitumen 25 ctions polar intera

20 15 10 5

di sp

er

▼▼ ▼ 0 ▼ ▼ ▼ ▼ 15 ▼

ei

siv

er nt

20

ac ns

tio

25 0

20 15 ng 10 nd i o b 5 en rog hyd

25

FIGURE 9.2 Plot of the solv ents in Table 9.1, bitumen No.1, in a 3D, x-y-z plot, where each axis is one of the Hansen solubility parameters.

interactions, respectively. The suitability of HSP may be illustrated by using a three-dimensional (3D) diagram where each axis constitutes one of the interactions. All solv ents with a solubility grade 1 were considered as “good solv ents” and all other grades as “poor solv ents.” The result is illustrated in Figure 9.2 where all “good solv ents” are f alling within a certain region separated from the “poor solv ents.” This confirms that the solubility properties of bitume can be reasonably well predicted by HSP . Although the “good solv ents” are found in a re gion of relatively high dispersion interaction and relati vely low polar and hydrogen bonding interaction, it seems like the latter two types of interactions are still of fundamental importance for understanding the properties of bitumen. Ev en if we cannot completely rule out the possibility that there e xist other types of interactions, we may, however, conclude that the HSP estimate is good enough, and particularly for understanding the true nature of bitumen. It can be assumed that the same situation is valid also for crude oils, which indicates that the use of HSP w ould be a valuable tool, also, for crude oil production, transport, and processing.

THE SOLUBILITY SPHERE Chapter 1 includes a discussion of a computer program called SPHERE for calculation of the best estimated HSP as well as the radius of the best fitted pseudo sphere, which includes the “goo solvents” and e xcludes the “poor solv ents,” based on a set of solubility data. The program w as applied on the data in Table 9.1 for calculation of the best estimate for HSP for 15 bitumens. The program permits only tw o le vels of solubility , “good solv ents” and “poor solv ents,” ho wever, whereas the solubility in Table 9.1 w as determined in 6 grades. F or comparison, the HSP were calculated using tw o dif ferent criteria for “good solv ents.” In the first calculation only the bes solvents (grade 1) were selected as “good solv ents” and then in a second calculation the tw o best grades (1 and 2) were tak en as “good solv ents.” All other solv ents were considered as “poor solvents.” The results are listed in Table 9.2. It is ob vious that the calculated HSP for the dif ferent bitumens become slightly dif ferent, depending on the choice of solubility grade for the “good solvents.” Although the dif ferent bitumens are selected to represent a range of products produced

7248_book.fm Page 160 Tuesday, April 24, 2007 9:19 AM

160

Hansen Solubility Parameters: A User’s Handbook

TABLE 9.2 Calculated HSP for 15 Bitumens Using Two Levels of Solubility as the “Good Solvents” Grade 1 = “Good Solvents” Bitumen

D

P

H

RAD

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Average

18.0 17.6 17.9 17.9 17.9 18.0 18.0 18.0 17.9 17.9 17.9 17.9 17.9 18.0 17.9 17.91

4.8 5.0 4.5 4.5 4.5 4.8 4.8 4.8 4.5 4.5 4.5 4.5 4.5 4.8 4.5 4.63

3.2 2.8 3.3 3.3 3.3 3.2 3.2 3.2 3.3 3.3 3.3 3.3 3.3 3.2 3.3 3.23

5.5 5.5 5.3 5.3 5.3 5.5 5.5 5.5 5.5 5.5 5.5 5.5 5.5 5.5 5.5 5.46

Grades 1 and 2 = “Good Solvents” D 17.9 17.9 18.0 17.5 17.5 17.4 17.9 18.0 18.1 17.4 17.4 17.4 17.4 17.9 18.1 17.72

P

H

5.1 5.1 4.8 4.7 4.7 4.0 3.3 4.8 5.5 4.0 4.0 4.0 4.0 5.1 5.3 4.56

3.1 3.1 3.2 2.7 2.7 2.0 2.5 3.2 2.9 2.0 2.0 2.0 2.0 3.1 3.1 2.64

RAD 5.8 5.8 5.5 5.7 5.7 6.6 7.3 5.5 6.0 6.6 6.6 6.6 6.6 5.8 5.9 6.13

from different crude oils as well as different process conditions, the difference in HSP is surprisingly small. The average HSP for bitumen based on calculations using “grade 1” as “good solv ents” are D = 17.9 MP a0.5, P = 4.6 MP a0.5, and H = 3.2 MP a0.5. The “sphere” radius (RAD) is 5.5 in the same units. The small variation in HSP between the dif ferent bitumens is a result of the small v ariation in solubility as seen in the data from Table 9.1, with only a fe w solvents giving different solubility for different types of bitumen. If solv ents giving a small residue (solubility grade 2) are accepted as “good solvents,” one still gets a very similar average HSP, but the variation between the different binders becomes more e vident. The main general trend is a small shift to ward lower h ydrogenbonding interactions and a lar ger radius of the solubility sphere. The larger radius is an e xpected consequence when more solv ents are accepted as “good solv ents.” The decrease in h ydrogen bonding is more dif ficult to xplain, b ut it might indicate that the “sphere” is not completely symmetrical. In applications where bitumen is used — for e xample road building and w ater proofing — i is well kno wn that bitumen produced by dif ferent methods and from dif ferent crude oils ha ve different performance. Although the 15 bitumens listed in Table 9.1 are primarily intended for use in the water-proofing industr , they are selected and manufactured to cover a wide variety of crude sources as well as dif ferent types of manuf acturing processes. Laboratory e xperiments, and fiel experience for some of the samples, sho w that there is a lar ge v ariation in performance of the bitumens. One example is the compatibility with polymers, such as styrene/butadiene/styrene (SBS), which varies to a lar ge extent. It is e xpected that some of these dif ferences should be reflected i the different chemical compositions of the bitumens and that these same dif ferences should also be reflected in the HS . The results given in Table 9.2 show, however, that there are only very small differences, particularly when calculated with only the best solv ents as “good solv ents.” If “grade 2” is also accepted as “good solv ent,” the v ariation between the binders becomes more e vident, but a comparison with known composition and performance still does not allow a simple correlation.

7248_book.fm Page 161 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

161

This lack of correlation is without an y doubt disappointing. We may however speculate that it is mainly due to a lack of precision. The solvents in Table 9.1 are selected to co ver a large area in the 3D solubility space, whereas most bitumens are mixtures of hydrocarbons where the differences in chemical properties are relatively small. Obviously it is necessary to have much better precision than the solubility testing as shown in Table 9.2. The better precision may be achieved in two ways. The first impr vement is to use a better selection of solv ents for the solubility testing. Solv ents that have HSP close to the border of solubility are preferred to better define the borde . Another approach is to perform turbidimetric titrations to estimate the e xact HSP at the precipitation point calculated from the ratio of a “good solv ent” and a “poor solv ent” at precipitation. This approach is further discussed as BISOM titrations below. An improved selection of solvents should focus on solvents with RED v alues around 1, as these are close to the boundary . The RED (relative energy difference) concept is discussed in Chapter 1. As much v ariation as possible with respect to the dispersion, polar, and h ydrogen bonding interactions is desired. This requires, of course, that an approximate HSP of the material is already available. And finall , nontoxic and inexpensive solvents are preferred. A suggested set of solv ents, optimized for determination of HSP of bitumens and similar materials is presented in Table 9.3. These solvents have RED between 0 and 2 related to the estimated HSP of bitumen as presented above. When using this set of solvents for a Venezuelan binder, the HSP is D = 18.6 MP a0.5, P = 3.0 MP a0.5, H = 3.4 MP a0.5, and the radius of the sphere is 6.3 in the same units. This set of numbers is dif ferent from the pre viously estimated v alues in Table 9.2. A comparison can be made with binder No. 9 (D = 17.9 MP a0.5, P = 4.5 MP a0.5, H = 3.3 MP a0.5, and a radius of 5.5 MP a0.5), which is similar to the binder used to obtain the data reported in Table 9.3. If the HSP of other types of materials than bitumen are going to be measured, also in the petroleum area, it is suggested that other sets of solv ents may be needed to get the best precision. Examples are light crude oils, distillates, base oils, petroleum w axes, etc.

COMPUTER PROGRAM FOR CALCULATION AND PLOTTING OF THE HANSEN 3D PSEUDOSPHERE The SPHERE program described in Chapter 1 has gi ven very good approximations of the HSP as well as the diameter of the (solubility) sphere for a lar ge number of materials. In the SPHERE program, a f actor 4 is used as a multiplier for the dif ference in the dispersion interactions of the species concerned. This means that the “sphere, ” with the three dif ferent types of interactions as coordinates, is in f act an ellipsoid (spheroid). A disadvantage with the SPHERE program is the lack of a tool for plotting the ellipsoid in a diagram that w ould be beneficial for illustration purposes Thus, an improved program which permits 3D plotting of the ellipsoid w as developed. During the development it was discussed that although the factor 4 has been proven to be a good approximation for most materials there might be complex mixtures which could give a better fit with xperimental data if other f actors were used. The new program has the follo wing features: • • • • •

Permits plotting of the HSP solubility ellipsoid in a 3D diagram. Permits plotting of up to three ellipsoids representing dif ferent materials in the same 3D diagram. The input data should be based on “poor solv ents,” “good solv ents,” and “borderline solvents.” There should be an option to mak e other types of fitting than the SPHERE program t the available data. Negative values of HSP interaction coef ficients are not ta en into consideration.

7248_book.fm Page 162 Tuesday, April 24, 2007 9:19 AM

162

Hansen Solubility Parameters: A User’s Handbook

TABLE 9.3 Solvents Used for Determination of the Solubility of Bitumen with Their HSP in MPa0.5 HSP No. 56 93 717 1060 118 955 156 182 183 184 188 194 1019 791 269 1084 889 328 333 353 345 758 412 419 440 1063 450 464 472 481 500 502 1029 524 531 546 1051 704 617 618 885 637 953

Solvent

D

P

H

Solubility

Benzophenone 2-Butanol 2-Butyl octanol Butyraldehyde Caprolactone (epsilon) 1-Chloro pentane Chloroform Cyclohexanol Cyclohexanone Cyclohexylamine Cyclopentanone cis-Decahydronaphthalene 1.4-Dichlorobutane 1.1-Diethoxy ethanol (acetal) Ethylene glycol monoethyl ether acetate Diisopropylamine 1.2-Dimethoxybenzene Ethyl acetate Ethyl benzene Ethyl lactate 2-Ethyl-hexanol Ethylene glycol dib utyl ether Hexadecane Hexyl acetate Isopropyl acetate Laurylalcohol Mesityl oxide Methyl acetate Methyl benzoate Methyl ethyl k etone 1-Methyl naphthalene Methyl oleate 3-Methyl-2-butanol Methylene dichloride Nitrobenzene Oleyl alcohol Pyrrolidine Salicylaldehyde Tetrahydrofuran Tetrahydronaphthalene 1.2.3.5-Tetramethylbenzene Toluene 2-Toluidine

19.6 15.8 16.1 15.6 19.7 16.0 17.8 17.4 17.8 17.2 17.9 18.8 18.3 15.2 16.2 14.8 19.2 15.8 17.8 16.0 15.9 15.7 16.3 15.8 14.9 17.2 16.4 15.5 17.0 16.0 20.6 14.5 15.6 18.2 20.0 14.3 17.9 19.4 16.8 19.6 18.6 18.0 19.4

8.6 5.7 3.6 10.1 15.0 6.9 3.1 4.1 6.3 3.1 11.9 0 7.7 5.4 5.1 1.7 4.4 5.3 0.6 7.6 3.3 4.5 0 2.9 4.5 3.8 6.1 7.2 8.2 9.0 0.8 3.9 5.2 6.3 8.6 2.6 6.5 10.7 5.7 2.0 0.5 1.4 5.8

5.7 14.5 9.3 6.2 7.4 1.9 5.7 13.5 5.1 6.5 5.2 0 2.8 5.3 9.2 3.5 9.4 7.2 1.4 12.5 11.8 4.2 0 5.9 8.2 9.3 6.1 7.6 4.7 5.1 4.7 3.7 13.4 6.1 4,1 8.0 7.4 14.7 8.0 2.9 0.5 2.0 9.4

1 0 0 0 0 1 1 0 1 1 0 1 1 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 1 0 0 1 0 0 1 0 1 1 1 1 0

7248_book.fm Page 163 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

163

TABLE 9.3 (CONTINUED) Solvents Used for Determination of the Solubility of Bitumen with Their HSP in MPa0.5 HSP No. 648 653 667 670 698

Solvent 1.1.2-Trichloroethane Tricresyl phosphate 1.2.4-Trimethylbenzene 18.0 2.2.4-Trimethylpentane o-Xylene

D

P

H

Solubility

18.2 19.0

5.3 12.3 1.0 0 1.0

6.8 4.5 1.0 0 3.1

1 0 1 0 1

14.1 17.8

Note: Good solv ents are indicated with a “1” and poor solv ents are indicated with a “0. ” This set of solv ents better defines the boundary r gion as discussed in the te xt.

The computer program hsp3D was developed on a MATLAB platform.27 The program permits 6 different kinds of fit to create a 3D bod , based on a lar ge set of solubility data. In each case all good solvents are included and all poor solv ents are e xcluded. 1. Convex hull fit which could be described as the points for the good sol ents being wrapped with a fl xible membrane. This fit ma es use of only the good solv ents. 2. The Hansen fi is the same type of fit as in the SPHERE program using Equation 1.9 The search algorithm is ho wever slightly dif ferent, so the results compared to the SPHERE program might be slightly dif ferent. 3. Axis-aligned ellipsoid fit, which is similar to the Hansen fit a ve, b ut with v ariable coefficients for the three a es (the three types of interactions). In the normal Hansen fi a factor 4 is used for transformation of the dispersion interactions, in the axis-aligned fit this actor as well as the factors for the other two axes are adjusted to optimize the fit 4. Rotated ellipsoid fit, which is similar to the Axis-aligned ellipsoid above but allows the body to rotate and tilt to obtain a better fit. In all of the fits a ve it is assumed that the axis of the ellipsoid is aligned along the three ax es. In the rotated ellipsoid the program can tilt the ax es to impro ve the fitting, and at the same time also optimize the transfo mation factors for the ax es. 5. Rotated ellipsoid with convex hull center and volume. The body for this fit has the sam center coordinates and v olume as the con vex hull b ut attempts to align with the “good solvents” to minimize distance to its surf ace. 6. Minimum enclosing ellipsoid is the body with the smallest v olume that encloses all the “good solvents.” The features of the impro ved computer program hsp3D were further e xamined using the solubility data from Table 9.3. The results from the dif ferent available fits were compared in 3 diagrams with three different fits in each (Figure 9.3 and Figure 9.4). From Figure 9.3 it is vident that there is a very small difference between the resulting ellipsoids using different fitting algorithms Transformation or tilting of the axis did not give any major improvement compared to the SPHERE data. This indicates that the f actor 4 in Equation 1.9 seems also to be v alid for such complicated mixtures as bitumen. In Figure 9.4 we see a comparison between the con vex hulls, which probably is the best figure to describe the solubility properties, as it is the truest body constructed withou approximations. This might be the first choice if di ferent materials are going to be compared. Another way of comparing the quality of the fit using the di ferent algorithms is to compare some indicators like volume, number of outliers, and fitting coe ficient.

7248_book.fm Page 164 Tuesday, April 24, 2007 9:19 AM

164

Hansen Solubility Parameters: A User’s Handbook

Venezuelan bitumen soluble Venezuelan bitumen insoluble

14

δH H-bonding

12 10 8 6 4 2 0 15 10

20 18

5 16 δP Polar

0

δD Dispersive

FIGURE 9.3 3D solubility body of bitumen using computer program hsp3D. Hansen fit, axis-aligned ellipsoid, and rotated ellipsoid are compared

The ellipsoids according to

It can be seen in Table 9.4 that the HSP for the particular Venezuelan bitumen, and most likely also for other bitumens, is more or less independent of the fitting method. This sho ws that the approximation with an ellipsoid is rather rob ust. The best solubility body is the one ha ving the smallest volume, the least number of outliers, and the highest fitting coe ficient. The Hansen sphere and the axis aligned ellipsoid gi ve almost the same results. The rotated ellipsoid gi ves a smaller volume b ut at the e xpense of more outliers and less good fitting. The most e xtreme case is the ellipsoid with the same center point (HSP) and the same v olume as the con vex hull, which gi ves the smallest volume, most outliers, and less good fitting.This is, of course, a result of the algorithm. If a body with multiple corners is transferred to an ellipsoid with the same v olume, most of the corners mathematically will f all outside the ellipsoid. The f act that the coordinates are dif ferent indicates that the con vex hull is sk ewing for this material compared to the Hansen Sphere. This might, however, also be due to an uneven selection of solvents rather than properties in the material.

COMPONENTS OF BITUMEN Bitumen is a very complex mixture of different hydrocarbons but yet with very similar properties. It is almost impossible to isolate chemically uniform fractions; instead, bitumen is usually di vided into fractions that are defined by the selection of the separation method. Perhaps the most commo separation of bitumen is the precipitation of asphaltenes from the maltenes. As stated abo ve, the definition of asphaltenes is the material that precipitates upon dilution of bitumen (or oil) with nxtraction of n-heptane soluble heptane.11,12 The fractionation could also be considered as an e molecules from the bitumen, lea ving a residue named “asphaltenes. ” The asphaltene-free fraction from bitumen is called “maltenes. ” In almost all of the literature about bitumen and asphalt it is

7248_book.fm Page 165 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

165

Venezuelan bitumen soluble Venezuelan bitumen insoluble

14 12 δH H-bonding

10 8 6 4 2 0 15 10 20 5

δP Polar

18 16

0

δD Dispersive

FIGURE 9.4 Plots of fit of bitumen solubility data using hsp3D sh wing the Convex hull model, the ellipsoid with the same center and v olume, and also the minimum enclosing ellipsoid.

TABLE 9.4 Precision Indicators for Fitting the Data in Table 9.3 to Ellipsoids Type of Fitting

D

P

H

Volume

Outliers

Fit Coeficient

Hansen Sphere Axis aligned ellipsoid Rotated ellipsoid Ellipsoid: convex hull c and v Minimum enclosing ellipsoid

18.4 18.3 18.4 18.0 18.4

3.9 3.9 4.1 4.4 4.1

3.6 3.5 3.6 4.1 3.7

399 399 242 99 371

3 3 5 10 6

0.980 0.987 0.939 0.798 0.983

Note: Outliers = number of “good solv ents” with RED > 1 + number of “poor solv ents” with RED < 1.

erroneously claimed that the asphaltenes are dispersed in the maltenes as a colloidal dispersion. That this is not correct can easily be pro ven by solubility testing and plotting of the solubility ellipsoids using the hsp3D program. Asphaltenes isolated by the standard method ASTM D6560 12 have been tested for solubility in the set of solvents listed in Table 9.3. The isolated maltene fraction is also tested for solubility in the same set of solvents. The solubility ellipsoids for the two materials are plotted using the hsp3D program (Figure 9.5). Figure 9.5 confirms that there is no verlap of the HSP for n-heptane and the ellipsoid for asphaltene, and it can be considered that the y are so f ar apart that the asphaltenes are not soluble in n-heptane. This agrees with the definition of asphaltenes. It is also vident that the HSP of the

7248_book.fm Page 166 Tuesday, April 24, 2007 9:19 AM

166

Hansen Solubility Parameters: A User’s Handbook

14 12

δH H-bonding

10

Asfaltenes

8 6

Maltenes 2 0 15 20

10 18

n-heptane δP Polar

16 0 δD Dispersive

FIGURE 9.5 Solubility ellipsoids for asphaltenes and maltenes compared with n-heptane.

maltenes is dif ferent from the HSP of n-heptane. Thus, there is no reason to belie ve that the asphaltenes will appear in the same state in maltenes as inn-heptane. The fact that they are insoluble in n-heptane is no e vidence that the y are insoluble in the maltenes. In f act, there is such a lar ge overlap between the solubility ellipsoids of the maltenes and the asphaltenes that the y are quite likely to be soluble in each other . This strongly suggests that the asphaltenes are not dispersed in the maltenes as a colloidal dispersion b ut are more lik ely dissolved. It might be ar gued that some of the asphaltenes molecules with e xtreme HSP might not be soluble in the maltenes, and thus could still be dispersed rather than dissolved. This is, however, less likely as long as the continuum in the asphaltenes and the maltenes is k ept intact. In some e xperiments the asphaltenes ha ve been further fractionated into “soluble” asphaltenes and “insoluble” asphaltenes. 28 If a fraction of the “insoluble” asphaltenes is mixed with the maltenes the y might be insoluble. The reason is that the continuum has been brok en and w ould probably not reflect the conditions in the original sample In fact, removal of fractions from either the maltenes or the asphaltenes will create a risk for phase separation. This is also the reason wh y one should be v ery careful in making an y claims or predictions of bitumen properties based on the properties of fractions.

BITUMEN AND POLYMERS It is a v ery common practice to impro ve bitumen properties by adding dif ferent additi ves. The reason is to impro ve the lo w temperature properties by making the bitumen softer at v ery lo w temperatures (<20°C) and at the same time mak e the bitumen more stif f at higher temperatures (+60°C). The temperatures are representative of the highest and lowest temperatures on the surface of an asphalt road during winter and summer , although in reality the maximum and minimum temperatures v ary considerably with geographical location. The most common and well kno wn modification is the addition of di ferent types of polymers to bitumen. Probably all possible types

7248_book.fm Page 167 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

167

of polymers ha ve been tested in bitumen — for e xample, plastomers, elastomers, tw o component curing systems, and e ven rec ycled rubber and plastics. The requirements on such products are, however, v ery strict, so in practice v ery fe w polymers ha ve found a wider use as modifiers fo bitumen. One of the most important requirements is the “compatibility” between the bitumen and the polymer. In this case, the meaning of “compatibility” is the stability ag ainst phase separation. Another important factor is the cost efficien y, which means that a good improvement of the bitumen properties is achieved with fairly small levels of polymeric additives. In the road building industry where the volumes are large and the price constraints are strong, the maximum level of modificatio is typically below 5%. In the roofing industry higher l vels are generally accepted as product quality is more important than price. The Hansen solubility parameter concept provides a good tool for selection of suitable polymers, based on predictions of compatibility between dif ferent polymers and bitumen. If the HSP of a particular polymer is not known, it can easily be determined with a simple solubility test as described above. To illustrate the usefulness we may compare tw o types of polymers with kno wn HSP with the HSP of bitumen. To mak e it simple we selected tw o polymers, polyethersulfone (PES) and polyethylensulfide, for which solubility data are presented in Chapter 5 and Chapter 18, respectvely. Neither of these polymers is a common modifier for bitumen. The solubility ellipsoids of the tw o polymers compared to the HSP sphere of Venezuelan bitumen are illustrated in Figure 9.6. It is evident that PES is not soluble in bitumen, as the solubility ellipsoid is almost completely outside the ellipsoid of bitumen. In case PES is mix ed with bitumen it will be dispersed rather than dissolv ed.

20

δH H-bonding

15 Polyether sulphone

10

5

Bitumen

0 Polyethylene sulphide

25 20

20

15 10

18 5

16 0

δP Polar

δD Dispersive

FIGURE 9.6 Solubility ellipsoid for bitumen compared with solubility ellipsoid for polyether sulfone and polyethylene sulfide

7248_book.fm Page 168 Tuesday, April 24, 2007 9:19 AM

168

Hansen Solubility Parameters: A User’s Handbook

H-bonding

10 SBS 5 Bitumen

0 15

10 Polar 20

5 18 16 0

14

Dispersion

FIGURE 9.7 Solubility ellipsoid of bitumen compared to a solubility ellipsoid of SBS.

The effect will be an increase in stif fness at temperatures where PES will act as a solid fille . In case of polyeth ylenesulphide the solubility ellipsoid is inside the ellipsoid of bitumen, and thus polyethylenesulphide is e xpected to be completely soluble (compatible). As the polymer is completely soluble we do expect the effect to be related to the concentration of the modifie. No problem with storage stability is foreseen. None of the polymers discussed abo ve have been frequently used for modification of bitumen The polymer most commonly used for bitumen is styrene b utadiene block copolymer (SBS) or similar polymers based on styrene and b utadiene. This polymer gi ves a good modification e fect at f airly lo w concentration (3–5%). The major adv antages are increased stif fness at f airly high temperatures (60°C) and improved fl xibility at low temperature. The higher stiffness will decrease the risk for rutting (permanent deformation). This risk is highest on hot, sunn y summer days. The HSP of SBS was determined with a solubility test as abo ve and the solubility ellipsoid was plotted together with bitumen in Figure 9.7. It is e vident that there is a considerable o verlap between the SBS and the bitumen.This implies partial solubility. In reality the situation is even more complicated as SBS consists of tw o dif ferent kinds of polymer se gments based on b utadiene and styrene, respectively. Each of these segments has different HSP. SBS belongs to the group of thermoplastic elastomers. These become plastic-lik e and can be processed at higher temperatures, at the same time having rubber-like properties at room temperature because of physical crosslinking caused by polystyrene and polyb utadiene being mutually insoluble. The polystyrene blocks ha ve a glass transition temperature of approximately 100°C, and therefore SBS is airly f workable at temperatures above 100°C b ut is still rubber -like at lo wer temperatures. It has been proposed that the same mechanism is also giving a good effect in bitumen, with the polystyrene being soluble/compatible in bitumen at the mixing temperature (180°C) b ut becoming less soluble or insoluble at lo wer temperatures. The effect is the same physical crosslinking as in pure SBS. Figure 9.7 supports this

7248_book.fm Page 169 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

169

δH Hydrogen bonding

10 8 6 4 2 0 15 10 δP Polar

5 0

12

14

16

18

20

22

24

δD Dispersive

FIGURE 9.8 Comparison between a hea vy Venezuelan crude oil (Laguna) and a light crude oil (Leadon) from the North Sea.

picture as the part of the SBS ellipsoid located outside bitumen presumably represent the polystyrene, although this has not been v erified by xperiments.

CRUDE OIL Crude oil is found almost all o ver the world with large reserves in the Middle East, Russia, China, North America, Venezuela, and the North Sea, just to mention a fe w examples. It is produced by drilling wells in the ground or under the sea. Crude oil is pumped up to the surf ace where it is transported by pipeline or ships to refineries for further processing into desired products. The crude oils are v ery dif ferent, depending on origin. Some crude oils are v ery light and contain a lar ge percentage of the most desired products, g asoline and diesel fuel, whereas other crude oils are heavy and bitumen-rich. The heaviest of the crude oils have such a high viscosity that they can not be pumped at normal ambient temperature b ut always have to be handled at higher temperature. Only a fe w selected crude oils can be used for production of high quality bitumen suitable for making asphalt for roads. Under certain conditions of storage and transport of crude oils there are sometimes problems with the formation of precipitates and/or deposits. These might decrease the capacity of pipelines by formation of solid contaminants in the crude oil. These deposits are sometimes blamed on asphaltenes and sometimes on w axes. The exact nature of these precipitates and the mechanism of their formation are not fully understood and is thus the subject for intense research. There are lar ge economic benefits to be ained if the problem with deposits could be decreased. The use of HSP to study the precipitates in comparison with the solubility parameters of the crude oils is a good tool for better understanding of the precipitation mechanism. To have the complete picture it is also necessary to understand ho w temperature and pressure influence th HSP of different molecules in the crude oil. The difference between tw o crude oils, hea vy Venezuelan Laguna and medium Leadon from the North Sea, may be illustrated by 3D plots of the solubility ellipsoids of each crude oil calculated with the hsp3D program (Figure 9.8). It is sho wn that Leadon is co vering a lar ger space than Laguna and is thus expected to have better solubility properties. This is probably an effect of Leadon being a lighter crude which contains more lo w viscosity oils. These are better solv ents than the

7248_book.fm Page 170 Tuesday, April 24, 2007 9:19 AM

170

Hansen Solubility Parameters: A User’s Handbook

higher molecular weight components of the Laguna. It is also seen that the solubility ellipsoid for the Laguna is located completely inside the solubility ellipsoid for the Leadon. This means that the Laguna crude should be completely soluble in the Leadon crude, and no problems with the formation of precipitates are to be expected by dilution of Laguna crude oil with Leadon crude oil.

TURBIDIMETRIC TITRATIONS Although the determination and visualization of solubility parameters for bitumen and other mineraloil-derived materials have proven to be very illustrative, there is still a desire for better precision. This has partly been met, as discussed abo ve, by a better selection of test solv ents and by better methods to optimize the ellipsoid. There are very obvious differences when it comes to practical applications, and particularly with modification with SBS polymers, among bitumens h ving v ery similar solubility ellipsoids (Table 9.2). Bitumens that are seemingly very similar with respect to solubility give still very different properties after mixing with SBS. One of the most important properties is the separation stability. Most mixtures of bitumen and SBS sho w a tendency for separation if the y are stored at high temperature for a long time. A typical separation test is made at 180°C for 3 d. The separation is usually seen as the polymer floating to the surace, but sometimes, particularly at concentrations of SBS between 10 and 15%, a phase separation can tak e place, also in the bitumen. This is seen as a hard precipitate at the bottom of the bitumen tank. The separation tendency can be o vercome by a proper selection of bitumen, alternatively by selection of a suitable polymer . The selection of components is mainly done on a “trial and error” basis, although there are some empirical rules. HSP may be an e xcellent tool for selection of suitable combinations of bitumen and polymer , but better precision is required than can be obtained with simple solubility testing with pure liquids. Impro ved estimation of the best solubility ellipsoid is required for optimal use of the HSP concept.

BISOM TEST The procedure discussed in the follo wing has been developed at Nynas Bitumen based on turbidimetric titrations to precisely determine the boundary of the surf ace of solubility. The procedure is called BISOM, an acron ym for BItumen Solubility Model. The principle of the test method is visualized in Figure 9.9. The HSP ellipsoid of the bitumen is constructed using the “poor solvents” illustrated as solid triangles in the figure and the “good sol ents” being illustrated with open triangles. Three nonsolv ents ha ve been selected. These ha ve HSP placing them just outside the surface of the solubility ellipsoid. They may be seen as the black triangles in the center of the circles in Figure 9.9. In the case of BISOM titration the selected “poor solvents” are 2,2,4-trimethyl pentane (isooctane), 2-b utanone (methyl ethyl ketone), and 2-eth yl hexanol. As bitumen is a high viscosity liquid or a semisolid material, it has to be diluted to decrease the viscosity to permit proper stirring during the titrations. For this purpose, a solvent with a solubility parameter as close to the center of the ellipsoid as possible has to be selected. For the BISOM titration we have selected toluene (or in some cases xylene) as the good solv ent. The titration is illustrated as arro ws, going from the HSP of the good solv ent toluene to an y of the three poor solv ents. The titration can be considered as a dilution of the bitumen with a mixture of a good solv ent and a poor solv ent. The HSP of the mixture is proportional to the concentration of each solv ent. A precipitate will appear when the HSP of the mixture has a v alue placing it on the surf ace of the solubility ellipsoid. An important ef fect to tak e into consideration in the turbidimetric titration of bitumen is the concentration effect. This comes from the f act that bitumen is k ept homogeneous by the mutual solubility of all its dif ferent molecules. The ef fect is seen as a higher concentration of bitumen giving better solubility . This situation is contradictory to what is usually kno wn for solubility of pure substances. To understand this phenomenon we must consider that the first sign of turbidit

7248_book.fm Page 171 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

171

Laguna B180 soluble Laguna B180 insoluble

14

δH H-bonding

12

2-ethyl-1-hexanol

10 8 6 4

MEK

2 0 15

Toluene 20

10

Iso-octane

18

5

16 0

δP Polar

δD Dispersive

FIGURE 9.9 Summary of the BISOM titration with the “good solvent” and the three “poor solvents” titrants.

is interpreted as “insolubility ,” but more precisely, it is the insolubility of the molecule(s) that are least soluble in the particular titrant/solv ent mixture that has been used. The concentration effect was first described by Hithaus 29 He developed a kind of turbidimetric test for what he called “peptization of asphaltenes.” In this titration only one good solvent, toluene, and one poor solv ent, n-heptane, were used. To account for the concentration ef fect, Heithaus performed the titration at several different concentrations of bitumen. The details of the calculations can be found in Reference 29, b ut Figure 9.10 gi ves an illustration of the principle. Heithaus Titration 1 FR = Vs/(VS + VT)

0.8 0.6 VT = volume titrant

0.4

VS = volume solvent 0.2

-0.1

0

-0.2

0

0.1

0.2

0.3

0.4

0.5

C(g/mL) = Wa/(VS + VT)

FIGURE 9.10 Illustration of the Heithaus titration of a Venezuelan bitumen.

7248_book.fm Page 172 Tuesday, April 24, 2007 9:19 AM

172

Hansen Solubility Parameters: A User’s Handbook

In Figure 9.10 the dilution ratio FR = volume solvent/total volume solvent and titrant is plotted against C = amount of bitumen/total amount of solvent and titrant. At the start of the titration FR = 1 as no titrant (V T = 0) has been added, whereas at the same time C is equal to the concentration of bitumen in the solvent. During the addition of titrant both FR and C become smaller and smaller. At infinite dilution FR = 0 and C = 0, ut before this a precipitate has been formed, pro vided that the titrant has been properly selected. The point where the first sign of precipitate is noticed i marked with a black dot in the diagram. F or each experiment the titration is repeated several times using different concentrations of bitumen. In Figure 9.10 the titration has been repeated four times, illustrated by the four titration arro ws showing the decrease of the FR and the C v alue during the titration, and four black dots indicating the first sign of precipitation.A straight line is fitted throug the four points using the least R-squared method. The equation for the line is used for extrapolation to find the intercepts for FR when C = 0 and for C when FR = 0. The meaning of the FR v alue at C = 0 is the ratio of solv ent to titrant where there is solubility for all concentrations of bitumen. The solvent to titrant ratio can be used to calculate the HSP at the precipitation point for infinit dilution of the bitumen. The fact that a higher concentration of bitumen requires more titrant to reach the precipitation point confirms that the solubility of bitumen increases as the concentratio increases. This effect is also visible with v ery dilute solutions. The meaning of the intercept for FR = 0 is the lo west concentration of bitumen that is needed to gi ve full stability in pure titrant. It could also be expressed as the maximum titrant which can be added to bitumen without causing precipitation. In practice the precipitation point at the BISOM titration can be determined by dif ferent methods. There are at least tw o commercial instruments which can be used for BISOM titrations although they are both originally developed for automatic Hethaus titration. The testing procedures used by the instruments are described in tw o ASTM standards. One of the instruments measures the transmission of light through a cuvette with a short beam length30 and the other instrument uses variation in the intensity of a reflected beam of light (attenuated total reflectance principle TR]).31 Both instruments can be equipped with more than one titrant for BISOM titration. There is a modified ersion of the ATR instrument 31 which is suitable for BISOM titrations. The computer program hsp3D also has the possibility to handle HSP of solv ents or solv ent mixtures considered as being e xactly on the surf ace of the ellipsoid. This is the case with HSP calculated from the ratio of good and poor solv ents at the precipitation point. Thus, the BISOM data may be combined with solubility data for better precision. The BISOM data may easily be recalculated from FR max to the FR v alue for the same concentration used in the solubility testing by using the equation for the line in Figure 9.10. The results from a BISOM titration may be reported in dif ferent ways. The most simple is to give the FRmax for the intercept C = 0 and Cmin for the intercept FR = 0. Heithaus29 proposed further calculations of f actors: pa = 1 – FR max

(9.1)

⎡⎛ 1 ⎞ ⎤ p 0 = FR max ⎢⎜ ⎟ + 1⎥ ⎣⎝ C min ⎠ ⎦

(9.2)

P=

p0 1 − pa

(9.3)

where pa is considered to be related to the solubility of the molecules in bitumen with the lo west solubility, p0 is related to the solubility po wer of the bitumen, and finall , P is a balanced stability index describing the internal stability of the bitumen. A higher number indicates a higher stability .

7248_book.fm Page 173 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

173

A high internal stability can be seen as a high allo wance for blending with additi ves, polymers, solvents, or other types of bitumen or crude oil. The interpretation of the BISOM titration is a determination of the internal stability of the bitumen or crude oil, rather than determination of HSP , although it is based on the principles of HSP. The BISOM could also be seen as an identification of those molecules that are the leas soluble in the bitumen or crude oil, and ho w close to insolubility the y are. To ha ve a complete picture it is necessary to ha ve several titrants with dif ferent HSP as the traditional determinations of internal stability using only n-heptane as precipitant will usually gi ve an incomplete picture. It is not e xpected that the material precipitated by addition of n-heptane would be the same as that precipitated by the addition of 2-eth yl-1-hexanol or 2-b utanone. The use of three titrants permits calculation of the HSP at three precipitation points, each of which could be considered to be on the surface of the solubility ellipsoid. It is, however, not possible to estimate the solubility ellipsoid from only three points on the surf ace, particularly since the e xact center point is not kno wn. If the HSP calculated from turbidimetric titrations are going to be compared to the HSP ellipsoid of the same material, the concentration at the solubility testing has to be tak en into consideration. The precipitation point at a certain concentration can easily be estimated from data such as are reported in Figure 9.10.

CONCLUSION • • • • • • • • • • • •

The Hansen solubility parameters of a comple x material such as bitumen or crude oil can be estimated by determination of its solubility in a lar ge number of solv ents with known HSP. To have the best precision, the test solvents should be selected with respect to the material to be tested. It is preferred to ha ve solvents near or at the borderline of solubility . The factor 4 in Equation 1.9 as a multiplier for the difference in the dispersion parameters for the materials being considered seems to be v alid, also for comple x mixtures lik e bitumen. Comparison of HSP for bitumen, asphaltenes, and maltenes confirms that asphaltene are soluble in maltenes, and bitumen should thus not be considered as a colloidal dispersion as is frequently claimed. The best estimated HSP v alues for Venezuelan bitumen are D = 18.6 MP a0.5, P = 3.0 MPa0,5, H = 3.4 MP a0.5. A computer program hsp3D can estimate the best fit to the solubility data and from thi calculate the HSP. The program can also estimate the best coef ficients for the ellipsoid model to illustrat the extension of solubility re gions in a 3D diagram. Up to three dif ferent materials can be compared in the same 3D plot. The program is not limited to bitumen and crude oils, b ut could equally well be used for other types of materials. A procedure for turbidimetric titrations has been de veloped to further impro ve the precision to determine the surf ace of the HSP solubility ellipsoid. This procedure is called a BISOM titration. BISOM titration is well suited for measurement of the internal stability of comple x hydrocarbon mixtures lik e bitumen or crude oils. BISOM titration is also a determination of the least soluble molecules in bitumen or crude oils.

7248_book.fm Page 174 Tuesday, April 24, 2007 9:19 AM

174

Hansen Solubility Parameters: A User’s Handbook

REFERENCES 1. Speight, J.G., The Chemistry and Technology of Petroleum, 2nd ed., Marcel Dekker, New York, 1991. 2. Nellensteyn, F.J., Relation of the micelle to the medium in asphalt, Inst. Pet. Technol., 14, 134–138, 1928. 3. Pfeiffer, J.P. and Saal, R.N.J., Asphaltic bitumen as colloidal system, Phys. Chem., 44 139–149, 1940. 4. Park, S.J. and Mansoori, G.A ., Aggregation and deposition of hea vy organics in petroleum crudes, Energy Sources, 10, 109–125, 1988. 5. Boduszynski, M.M., McKay , J.F., and Lathham, D.R., Asphaltenes where are you? Proc. Assoc of Asphalt Paving Technologists, 49, 1980, pp. 124–143. 6. Petersen, J.C., Robertson, R.E., Branthaver, J.F., Harnsberger, P.M., Duvall, J.J., Kim, S.S., Anderson, D.A., Christiansen, D.W., and Bahia, H.U., Binder Characterization and Ev aluation, Vol. 1: Physical Characterization, Strategic Highway Research Program SHRP-A-367, 1994. 7. Redelius, P .G., Bitumen solubility model using Hansen solubility parameter , Energy Fuels, 18, 1087–1092, 2004. 8. Sirota, E.B., Understanding the Physical Structure of Asphaltenes to Optimize Bitumen Manufacture, 3rd Euroasphalt and Eurobitume Congress Vienna, Paper 097, 930–939, 2004. 9. Yen, T.F. and Chilingarian, G.V., Asphaltenes and Asphalts, 1 in Developments in Petroleum Science 40 A, Elsevier Science BV, Amsterdam 1994. 10. Sheu, E.Y. and Mullins, O.C., Asphaltenes, Fundamentals and Applications, Plenum Press, New York, 1995. 11. Test Method for n-heptane Insolubles. Annual book of ASTM Standards Section 04.03, ASTM D327997, 2001. 12. Test Method for Determination ofAsphaltenes (Heptane insolubles) in Crude Petroleum and Petroleum Products, Annual book of ASTM Standards Section 05.04, ASTM D6560-00. 13. Buch, L., Groenzin, H., Buenrostro-Gonzales, E., Andersen, S.I., Lira-Galena, C., Mullins, O.C., Molecular size of asphaltene fractions obtained from residuum h ydrotreatment, Fuel, 82, 1075–1084, 2003. 14. Boduszynski, M.M., Asphaltenes in petroleum asphalts: composition and formation, Am. Chem. Soc. Meet., Div. Pet. Chem., Washington, D.C., September 9–14, 1979. 15. Mitchell, D.L. and Speight, J.G., The solubility of asphaltenes in h ydrocarbon solv ents, Fuel, 52, 149–152, 1973. 16. Hirschberg, A., deJong, L.N.J., Schipper , B.A., Meijer , J.G., Influence of temperature and pressur on asphaltene flocculation, Soc. Pet. Eng. J., 283–293, June 1984. 17. Hildebrand, J.H. and Scott, R.L., The Solubility of Nonelectrolytes, 3rd ed., Do ver, New York, 1964. 18. Laux, H. and Rahimian, I., Colloid-disperse Crude Oil Systems; Phase Behaviour and Stability, Erdöl und Kohle — Erdgas — Petrochemie vereinigt mit Brennstoff-Chemie, 47(11), 430–435, 1994. 19. Hirschberg, A. and Hermans, L., Asphaltene phase beha viour: a molecular thermodynamic model , Proc. Characterisation of Heavy Crude Oils and Petroleum Residues, Lyon, June 25–27, 1984. 20. Burke, N.E., Hobbs, R.E., Kashou, S.F ., Measurement and modelling of asphaltene precipitating, J. Pet. Technol., November, 1440–1446, 1990. 21. Rahimian, I. and Zenk e, G., Zum Verhalten organicher Lösemittel ge genüber Bitumen, Bitumen 1, 2–8, 1986 (in German). 22. Neumann, H.J., Rahimian, I., Zenk e, G., Einfluss der Löslich eitseigenschaften von Asphaltene auf die Rückstandsverarbeitung. Erdöl und Kohle –Erdgas, 42(7/8), 278–286, 1989 (in German). 23. Hagen, A.H., Jones, R., Hofner , R.M., Randolf, B.B., Johnson, M.P ., Characterisation of asphalt by solubility profiles, J. Assoc. Asphalt Paving Technol., 53, 119–137, 1984. 24. Redelius, P.G., Solubility parameters and bitumen , Fuel, 79, 27–35, 2000. 25. Hansen, C.M., Skaarup, K., The three dimensional parameter — k ey to paint component af finities III. Independent calculation of the parameter components, J. Paint Technol., 39(511), 511–514, 1967. 26. Warnez, M. and Redelius, P .G., unpublished results from Icopal – Nynas joint research project: Correlation between Bitumen Characteristics and Fire Properties of Roofing Membranes, 2001–2003 27. Turner, F. and Redelius, P ., The program hsp3D is a vailable as share w are in Europe from: Nynas Petroleum, S-149 82 Nynäshamn, Sweden, www .nynas.com or in USA from Western Research Institute, 265 North 9th Street, Laramie, Wyoming 82070–3380, www.westernresearch.org.

7248_book.fm Page 175 Tuesday, April 24, 2007 9:19 AM

Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils

175

28. Östlund, J.-A., Wattana, P., Nydén, M., and Fogler, H.S., Characterisation of fractionated asphaltenes by UV-VIS and NMR self-dif fusion spectroscopy, J. Colloid Interface Sci., 271(2), 372–380, 2004. 29. Heithaus J.J., Measurement and significance of asphatene pepetization, J. Inst. Pet., 48(458), 45–53, 1962. 30. Standard Test Method for Automated Heithaus Titrimetry, Annual book of ASTM Standards Section 04.03, ASTM D6703-01. 31. Standard Test Method for Determining Stability and Compatibility of Hea vy Fuel Oils and Crude Oils by Hea vy Fuel Oil Stability Analyzer (Optical Detection), Annual book of ASTM Standards Section 05.04, ASTM D7112-05a

7248_book.fm Page 176 Tuesday, April 24, 2007 9:19 AM

7248_book.fm Page 177 Tuesday, April 24, 2007 9:19 AM

of Hansen 10 Determination Solubility Parameter Values for Carbon Dioxide Laurie L. Williams ABSTRACT Reference values of the Hansen solubility parameters (HSP) for carbon dioxide (CO 2) have been determined. The methodology adopted for this is based on room temperature solubility of the g as in different liquids of known HSP values. CO2 gas solubility data, at 25°C and a CO2 partial pressure of 1 atmosphere, in 101 liquid solv ents have been g athered and e valuated, yielding the v alues: δd = 15.6 MP a1/2, δp = 5.2 MP a1/2, and δh = 5.8 MP a1/2. A methodology for e xtending this reference set of HSP v alues to an y temperature and pressure has been de veloped utilizing an equation of state (EOS) of the form P = f(ρ, T).

INTRODUCTION The HSP concept is widely used for selecting suitable solvents for organic compounds. In addition, HSPs ha ve been applied to biological materials, barrier properties of polymers, as well as the characterization of surf aces, pigments, fillers, and fibers. 1 The basis of the HSP approach is the assumption that the total cohesi ve ener gy ( E) of a pure compound is made up of the additi ve contributions from nonpolar (dispersion) interactions (Ed), polar (dipole–dipole and dipole–induced dipole) interactions ( Ep), and hydrogen bonding or other specific association interactions including Lewis acid–base interactions ( Eh): E = Ed + EP + Eh

(10.1)

Dividing each contrib ution by the molar v olume, E Ed E p Eh = + + V V V V

(10.2)

gives the square of the total solubility parameter as the sum of the squares of the Hansen dispersion (δd), polar ( δp), and h ydrogen bonding ( δh) solubility parameters, so that δ t2 = δ 2d + δ 2p + δ 2h

(10.3)

where

177

7248_book.fm Page 178 Tuesday, April 24, 2007 9:19 AM

178

Hansen Solubility Parameters: A User’s Handbook

δ 2d =

Ed ; V

δ 2p =

Ep V

;

and δ 2h =

Eh V

(10.4)

The determination of HSPs for compounds that are gases at ambient conditions is usually based on room temperature solubility of the g as in a range of dif ferent liquids of kno wn δd, δp, and δh. Those liquids that show the highest solubility for the gas are assumed to have HSPs closer to those of the g as than those liquids that ha ve lo wer solubilities for the g as. In the follo wing section, published data of CO 2 gas solubility at 25°C and a CO 2 partial pressure of 1 atmosphere in a lar ge number of liquid solv ents are e valuated. From this data, a set of HSP v alues at a single reference temperature and pressure is determined.

METHODOLOGY Literature values of CO 2 solubility in v arious liquid solv ents at 25°C ha ve been collected and are summarized in Table 10.1. All the solubility data sho wn in Table 10.1 w as either e xperimentally determined at a CO 2 partial pressure of 0.1 MP a, or has been corrected to pCO2 MPa using Henry’s law,2 xCO2 =

pCO2 KH

(10.5)

where KH is the Henry’ s law coefficient, pCO2 is the partial pressure of CO 2, and xCO2 is the mole fraction of dissolv ed CO 2. In addition to correcting CO 2 partial pressure to 0.1 MP a, it has been necessary to correct the reported v alues of CO 2 gas solubility to a common set of units. Generally , solubilities ha ve been reported as mole fraction of dissolv ed CO 2, xCO2, or as one of the tw o dimensionless quantities; the Bunsen coefficient or the Ostwald coefficient. The Bunsen coefficient, Ω, is defined as the v olume of g as, reduced to 0°C and 0.1 MP a, dissolv ed per unit v olume of solvent at a system temperature, T, under a gas pressure of 1 atmosphere. The Ostwald coefficient, L, is defined as the ratio of the v olume of g as absorbed to the v olume of the absorbing liquid, all measured at the same temperature.3 If the solubility is small and the gas phase is ideal, the Ostwald coefficient is independent of total pressure and these tw o coefficients are simply related by 4: L=

T Ω 273

(10.6)

The mole fraction of dissolv ed CO 2 can then be calculated using

xCO2

⎡⎛ RT ⎞ ⎤ = ⎢⎜ ⎟ + 1⎥ ⎢⎣⎝ LpCO2 V10 ⎠ ⎥⎦

2

−1

(10.7)

where R is the g as constant, T is the absolute temperature, and V10 the molar v olume of the pure solvent. The Hansen dispersion, polar , and h ydrogen bonding parameters in Table 10.1 are from Hansen’s solublity parameter handbook,5 and the total solubility parameter value is calculated from Equation 10.3, noted earlier . Using the collected v alues of xCO2 at 25 °C and pCO2 = .1 MPa, HSP v alues were calculated based on a simple weighted a verage using the data set in Table 10.1, hereafter called data set #1. Where multiple CO 2 gas solubility data were a vailable for an indi vidual solvent, an average value for that solv ent was used.

7248_book.fm Page 179 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

179

TABLE 10.1 CO2 Solubility in Various Solvents at T = 25ºC and a CO2 Partial Pressure of 1 Atmosphere

Formula C2H4O2 C2H4O2 C4H6O3 C3H6O C3H6O C3H6O C3H6O C3H6O C3H6O C7H14O2 C7H14O2 C7H14O2 C5H11Br C5H11Cl C6H7N C6H7N C6H7N C6H7N C7H6O C7H6O C 6H6 C 6H6 C7H7Cl C6H5Br C4H10O C4H10O C4H10O C4H10O C4H10O C22H42O2 C4H8O2 CS2 CS2 CCl4 CCl4 CCl4 C6H5Cl C6H5Cl CHCl3 CHCl3 CHCl3 C9H12 C7H14 C7H12O C6H12 C6H12 C6H12 C6H12O

Solvent Acetic acid Acetic acid Acetic anhydride Acetone Acetone Acetone Acetone Acetone Acetone Amyl acetate Amyl acetate Amyl acetate Amyl bromide Amyl chloride Aniline 0.00493 Aniline 0.00486 Aniline 0.00487 Aniline 0.00487 Benzaldehyde Benzaldehyde Benzene Benzene 0.00880 Benzyl chloride Bromobenzene Butanol Butanol Butanol 2-Butanol t-Butanol 0.00725 Butyl oleate Butyric acid Carbon disulfide Carbon disulfide Carbon tetrachloride Carbon tetrachloride Carbon tetrachloride Chlorobenzene Chlorobenzene Chloroform 0.01121 Chloroform Chloroform 0.01126 Methyl ethyl benzene Cycloheptane Cycloheptanone Cyclohexane Cyclohexane Cyclohexane Cyclohexanol

Mole Fraction CO2 0.01089 0.01122 0.01977 0.01876 0.01903 0.01896 0.01896 0.02259 0.02092 0.02800 0.02452 0.02612 0.01235 0.01424

0.01140 0.01171 0.00962 0.00912 0.00789 0.00887 0.00734 0.00718 0.00660 0.02790 0.01297 0.00215 0.00328 0.01100 0.01059 0.00904 0.00938 0.00981 0.01277 0.01008 0.00721 0.01588 0.00771 0.00760 0.00774 0.00471

Actual/ Ideal

δd (MPa)1/2

0.4757 0.4899 0.8635 0.8193 0.8309 0.8279 0.8279 0.9865 0.9136 1.2227 1.0708 1.1407 0.5394 0.6220 0.2151 0.2121 0.2129 0.2129 0.4979 0.5114 0.4201 0.3843 0.3983 0.3444 0.3872 0.3207 0.3135 0.2882 0.3166 1.2183 0.5665 0.0940 0.1432 0.4803 0.4624 0.3948 0.4095 0.4283 0.4897 0.5576 0.4918 0.4401 0.3148 0.6934 0.3367 0.3319 0.3380 0.2057

14.5 14.5 16.0 15.5 15.5 15.5 15.5 15.5 15.5 15.8 15.8 15.8 20.3 15.5 19.4 19.4 19.4 19.4 19.4 19.4 18.4 18.4 18.8 20.5 16.0 16.0 16.0 15.8 15.2 14.7 14.9 20.5 20.5 17.8 17.8 17.8 19.0 19.0 17.8 17.8 17.8 16.1 17.2 17.2 16.8 16.8 16.8 17.4

δp (MPa)1/2 8.0 8.0 11.7 10.4 10.4 10.4 10.4 10.4 10.4 3.3 3.3 3.3 4.8 5.0 5.1 5.1 5.1 5.1 7.4 7.4 0.0 0.0 7.1 5.5 5.7 5.7 5.7 5.7 5.1 3.4 4.1 0.0 0.0 0.0 0.0 0.0 4.3 4.3 3.1 3.1 3.1 7.0 0.0 10.6 0.0 0.0 0.0 4.1

δh (MPa)1/2 13.5 13.5 10.2 7.0 7.0 7.0 7.0 7.0 7.0 6.1 6.1 6.1 2.8 1.3 10.2 10.2 10.2 10.2 5.3 5.3 2.0 2.0 2.6 4.1 15.8 15.8 15.8 14.5 14.7 3.4 10.6 0.6 0.6 0.6 0.6 0.6 2.0 2.0 5.7 5.7 5.7 0.0 0.0 4.8 0.2 0.2 0.2 13.5

δt (MPa)1/2

Ref.

21.37 21.37 22.29 19.94 19.94 19.94 19.94 19.94 19.94 17.26 17.26 17.26 21.05 16.34 22.50 22.50 22.50 22.50 21.43 21.43 18.51 18.51 20.26 21.62 23.20 23.20 23.20 22.19 21.75 15.47 18.74 20.51 20.51 17.81 17.81 17.81 19.58 19.58 18.95 18.95 18.95 17.56 17.20 20.77 16.80 16.80 16.80 22.40

29 30 29 29 30 31 32 33 34 35 29 30 29 29 29 30 33 3 30 29 36 29 29 29 32 37 38 39 40 38 29 29 41 35 36 29 29 34 29 34 30 29 42 43 44 2 42 45

7248_book.fm Page 180 Tuesday, April 24, 2007 9:19 AM

180

Hansen Solubility Parameters: A User’s Handbook

TABLE 10.1 (CONTINUED) CO2 Solubility in Various Solvents at T = 25ºC and a CO2 Partial Pressure of 1 Atmosphere

Formula C6H10O C8H16 C8H16 C5H10 C5H8O C10H22 C10H22 C10H22O C2H4Br2 C2H4Br2 C2H4Br2 C2H4Br2 C3H6Cl2O CH2Cl2 C8H14O C2H6OS C2H6OS C2H6OS C3H7NO C4H8O2 C12H26 C12H26 C12H26 C12H26O C2H6O C2H6O C2H6O C4H8O2 C8H10 C2H4Cl2 C2H6O2 C10H12O2 C5H9NO2 C3H8O3 C7H16 C7H16 C7H16 C7H16 C7H16O C16H34 C16H34 C6H14 C6H14 C6H14O C5H5I C5H12O C4H10O C4H10O

Solvent Cyclohexanone Cyclooctane Cyclooctane Cyclopentane Cyclopentanone Decane Decane Decanol 1,2-Dibromoethane 0.00812 1,2-Dibromoethane 0.00804 1,2-Dibromoethane 0.00797 1,2-Dibromoethane 1,3-dichloro-2-propanol Dichloromethane Dimethyl cyclohexanone Dimethyl sulfoxide Dimethyl sulfoxide Dimethyl sulfoxide Dimethyl formamide 1,4-Dioxane Dodecane Dodecane Dodecane Dodecanol 0.01811 Ethanol Ethanol Ethanol Ethyl acetate Ethyl benzene Ethylene chloride Ethylene glycol Eugenol N-formyl morpholine Glycerol (glycerin) Heptane Heptane Heptane Heptane Heptanol Hexadecane Hexadecane Hexane Hexane Hexanol Iodobenzene Isoamyl alcohol Isobutanol Isobutanol

Mole Fraction CO2 0.01600 0.00688 0.00686 0.00491 0.01641 0.01204 0.01250 0.00973

0.00838 0.00746 0.01250 0.01680 0.00908 0.00907 0.00945 0.01610 0.02272 0.01428 0.01290 0.01191 0.00642 0.00642 0.00725 0.02300 0.01006 0.01133 0.00220 0.01023 0.01475 0.00009 0.01328 0.01190 0.01194 0.01202 0.01258 0.01161 0.01420 0.01318 0.01190 0.01174 0.00592 0.00809 0.00696 0.00690

Actual/ Ideal

δd (MPa)1/2

δp (MPa)1/2

0.6987 0.3004 0.2996 0.2146 0.7166 0.5256 0.5459 0.4249 0.3546 0.3509 0.3481 0.3659 0.3258 0.5459 0.7336 0.3965 0.3962 0.4125 0.7031 0.9921 0.6235 0.5633 0.5202 0.7908 0.2804 0.2805 0.3166 1.0044 0.4393 0.4946 0.0961 0.4469 0.6441 0.0039 0.5801 0.5197 0.5212 0.5248 0.5495 0.5069 0.6201 0.5754 0.5197 0.5124 0.2585 0.3530 0.3041 0.3014

17.8 17.5 17.5 16.4 17.9 15.7 15.7 17.5 17.8 17.8 17.8 17.8 17.5 18.2 15.2 18.4 18.4 18.4 17.4 19.0 16.0 16.0 16.0 15.5 15.8 15.8 15.8 15.8 17.8 19.0 17.0 15.1 16.6 17.4 15.3 15.3 15.3 15.3 15.1 16.3 16.3 14.9 14.9 14.1 19.5 15.8 15.1 15.1

6.3 0.0 0.0 0.0 11.9 0.0 0.0 2.6 6.4 6.40 6.4 6.4 9.9 6.3 8.8 16.4 16.4 16.4 13.7 1.8 0.0 0.0 0.0 6.5 8.8 8.8 8.8 5.3 0.6 7.4 11.0 8.8 11.7 12.1 0.0 0.0 0.0 0.0 8.0 0.0 0.0 0.0 0.0 8.6 6.0 5.2 5.7 5.7

δh (MPa)1/2 5.1 0.0 0.0 1.8 5.2 0.0 0.0 10.0 7.0 7.0 7.0 7.0 14.6 6.1 3.3 10.2 10.2 10.2 11.3 7.4 0.0 0.0 0.0 10.8 19.4 19.4 19.4 7.2 1.4 4.1 26.2 9.8 10.0 29.3 0.0 0.0 0.0 0.0 13.0 0.0 0.0 0.0 0.0 12.7 6.1 13.3 15.9 15.9

δt (MPa)1/2

Ref.

19.56 17.50 17.50 16.50 22.11 15.70 15.70 20.32 20.17 20.17 20.17 20.17 24.85 20.20 17.87 26.68 26.68 26.68 24.86 20.47 16.00 16.00 16.00 19.98 26.52 26.52 26.52 18.15 17.87 20.80 33.11 20.04 22.64 36.16 15.30 15.30 15.30 15.30 21.47 16.30 16.30 14.90 14.90 20.83 21.29 21.30 22.66 22.66

46 42 47 42 48 32 49 50 34 29 30 45 29 38 51 44 52 33 32 38 53 49 32 32 30 31 38 38 38 29 38 29 54 29 53 49 36 52 32 32 49 32 49 32 29 30 2 29

7248_book.fm Page 181 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

181

TABLE 10.1 (CONTINUED) CO2 Solubility in Various Solvents at T = 25ºC and a CO2 Partial Pressure of 1 Atmosphere

Formula

Solvent

C6H12O2 Isobutyl acetate C4H9Cl Isobutyl chloride C3H8O Isopropanol CH4O Methanol CH4O Methanol CH4O Methanol CH4O Methanol C3H6O2 Methyl acetate C3H6O2 Methyl acetate C7H14 Methyl cyclo hexane C7H12O 2-Methylcyclohexanone C4H8O Methyl ethyl ketone C11H10 1-Methyl naphthalene C19H36O2 Methyl oleate C6H5NO2 Nitrobenzene C6H5NO2 Nitrobenzene C5H9NO N-methyl-2-pyrrolidone C9H20 Nonane C9H20O Nonanol C8H18 Octane C8H18 Octane Octanol C8H18O C18H34O2 Oleic acid C15H32 Pentadecane C5H12 Pentane C5H12O Pentanol 0.00806 C7F16 Perfluroheptane C8H7N Phenyl acetonitrile C3H8O Propanol C3H8O Propanol C3H8O Propanol C3H6O2 Propionic acid Propionitrile C3H5N C5H10O2 Propyl acetate C3H6Br2 Propylene bromide C4H6O3 Propylene carbonate C4H6O3 Propylene carbonate C4H6O3 Propylene carbonate C5H5N Pyridine C5H5N Pyridine C5H5N Pyridine C9H7N Quinoline C4H8O2S Sulfolane 0.00799 C14H30 Tetradecane C14H30 Tetradecane C4H8O Tetrahydrofuran C10H12 Tetrahydronaphthalene C7H9N m-Toluidine

Mole Fraction CO2 0.02500 0.01410 0.00654 0.00641 0.00564 0.00563 0.00635 0.02089 0.02253 0.00927 0.01660 0.02444 0.00674 0.02690 0.01020 0.00997 0.01590 0.01231 0.01481 0.01254 0.01210 0.00930 0.01570 0.01167 0.01385 0.02088 0.00957 0.00782 0.00680 0.00759 0.01234 0.01677 0.02429 0.00977 0.01074 0.01162 0.01210 0.01193 0.01169 0.01198 0.00912 0.01171 0.01360 0.02700 0.00752 0.00635

Actual/ Ideal

δd (MPa)1/2

1.0919 0.6158 0.2856 0.2797 0.2463 0.2460 0.2774 0.9123 0.9837 0.4046 0.7249 1.0672 0.2943 1.1747 0.4455 0.4355 0.6943 0.5376 0.6465 0.5474 0.5284 0.4062 0.6856 0.5094 0.6048 0.3519 0.9118 0.4180 0.3414 0.2969 0.3316 0.5390 0.7323 1.0607 0.4266 0.4688 0.5073 0.5284 0.5211 0.5104 0.5231 0.3983 0.3489 0.5115 0.5939 1.1790 0.3285 0.2771

15.1 14.7 15.8 15.1 15.1 15.1 15.1 15.5 15.5 16.0 17.6 16.0 20.6 14.5 20.0 20.0 18.0 15.7 15.3 15.5 15.5 17.0 16.2 16.8 14.5 15.9 12.0 19.5 16.0 16.0 16.0 14.7 15.3 15.3 17.4 20.0 20.0 20.0 19.0 19.0 19.0 19.4 18.4 16.2 16.2 16.8 19.6 19.3

δp (MPa)1/2 3.7 5.3 6.1 12.3 12.3 12.3 12.3 7.2 7.2 0.0 6.3 9.0 0.8 3.9 8.6 8.6 12.3 0.0 7.3 0.0 0.0 3.3 3.1 0.0 0.0 4.5 0.0 12.3 6.8 6.8 6.8 5.3 14.3 4.3 7.5 18.0 18.0 18.0 8.8 8.8 8.8 7.0 16.6 0.0 0.0 5.7 2.0 3.8

δh (MPa)1/2 6.3 0.9 16.4 22.3 22.3 22.3 22.3 7.6 7.6 1.0 4.7 5.1 4.7 3.7 4.1 4.1 7.2 0.0 12.0 0.0 0.0 11.9 5.5 0.0 0.0 13.9 0.0 3.8 17.4 17.4 17.4 12.4 5.5 7.6 2.9 4.1 4.1 4.1 5.9 5.9 5.9 7.6 7.4 0.0 0.0 8.0 2.9 10.1

δt (MPa)1/2

Ref.

16.77 15.65 23.58 29.61 29.61 29.61 29.61 18.70 18.70 16.03 19.28 19.05 21.14 15.46 22.15 22.15 22.96 15.70 20.77 15.50 15.50 21.01 17.39 16.80 14.50 21.59 12.00 23.37 24.60 24.60 24.60 19.95 21.65 17.62 19.17 27.22 27.22 27.22 21.75 21.75 21.75 21.98 25.86 16.20 16.20 19.46 19.91 22.11

29 29 38 32 30 31 29 29 34 55 56 33 38 38 29 34 38 32 32 32 49 50 38 32 32 29 41 34 57 58 29 29 34 34 29 32 33 38 29 30 34 38 32 32 49 38 38 29

7248_book.fm Page 182 Tuesday, April 24, 2007 9:19 AM

182

Hansen Solubility Parameters: A User’s Handbook

TABLE 10.1 (CONTINUED) CO2 Solubility in Various Solvents at T = 25ºC and a CO2 Partial Pressure of 1 Atmosphere

Formula

Solvent

C7H9N C 7H 8 C 7H 8 C 7H 8 C 7H 8 C12H27O4P C2Cl3F3 C13H28 C6H15N C8H18 C11H24 C11H24O H2O H2O H2O H2O C8H10 C8H10 C8H10 C8H10

o-Toluidine Toluene 0.00994 Toluene 0.01010 Toluene 0.01042 Toluene 0.01039 Tributyl phosphate Trichlorotrifluoroethane Tridecane Triethylamine 2,2,4-Trimethylpentane Undecane Undecanol 0.01708 Water Water Water Water o-Xylene m-Xylene 0.01042 m-Xylene 0.01063 p-Xylene

Note:

ideal CO 2

x

Mole Fraction CO2 0.00605

0.03550 0.01823 0.01175 0.02321 0.01387 0.01148 0.00070 0.00059 0.00060 0.00061 0.00994

0.01087

Actual/ Ideal

δd (MPa)1/2

0.2641 0.4340 0.4411 0.4550 0.4536 1.5502 0.7961 0.5132 1.0134 0.6057 0.5014 0.7457 0.0306 0.0258 0.0261 0.0267 0.4339 0.4552 0.4642 0.4744

19.4 18.0 18.0 18.0 18.0 16.3 14.7 16.4 17.8 14.1 16.0 15.4 15.5 15.5 15.5 15.5 17.8 17.4 17.4 17.4

δp (MPa)1/2 4.2 1.4 1.4 1.4 1.4 6.3 1.6 0.0 0.4 0.0 0.0 6.7 16.0 16.0 16.0 16.0 1.0 1.0 1.0 1.0

δh (MPa)1/2 10.7 2.0 2.0 2.0 2.0 4.3 0.0 0.0 1.0 0.0 0.0 11.2 42.3 42.3 42.3 42.3 3.1 3.1 3.1 3.1

δt (MPa)1/2

Ref.

22.55 18.16 18.16 18.16 18.16 18.00 14.79 16.40 17.83 14.10 16.00 20.19 47.81 47.81 47.81 47.81 18.10 17.70 17.70 17.70

29 29 55 52 34 38 59 32 33 59 32 32 35 29 30 31 60 29 60 60

= 0.0229.

n

∑xδ

i di

δ

CO 2 d

=

i=1 n

∑x

’

(10.8)

’

(10.9)

.

(10.10)

i

i=1

n

∑xδ i

δ

CO 2 p

=

pi

i=1 n

∑x

i

i=1

n

∑xδ

i hi

δ

CO 2 h

=

i=1 n

∑x i=1

i

7248_book.fm Page 183 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

183

TABLE 10.2 Solvents Showing Greater Than Ideal Solubility for CO2 at 25°C Solvent Tributyl phosphate (C12H27O4P) Amyl acetate (C7H14O2) Butyl oleate (C22H42O2) Tetrahydrofuran (C4H8O) Methyl oleate (C19H36O2) Isobutyl acetate (C6H12O2) Methyl ethyl ketone (C4H8O) Propyl acetate (C5H10O2) Ethyl acetate (C4H8O2) Methyl acetate (C3H6O2) Note: PCO2 = 1,

ideal xCO 2

Exptl xCO 2

δd (MPa)1/2

δp (MPa)1/2

δh (MPa)1/2

0.03550

16.3

6.3

4.3

0.02800

15.8

3.3

6.1

0.02790

14.7

3.4

3.4

0.02700

16.8

5.7

8.0

0.02690

14.5

3.9

3.7

0.02500

15.1

3.7

6.3

0.02444

16.0

9.0

5.1

0.02429

15.3

4.3

7.6

0.02300

15.8

5.3

7.2

0.02253

15.5

7.2

7.6

= 0.0229.

A second HSP evaluation was undertaken using a subset of data set # 1. The subset was chosen to consist of solv ents where the measured CO 2 solubility w as greater than the ideal solubility at ideal 25°C and pCO2 = .1 MPa, xCO 2 = 0.0229. (The calculation of this ideal solubility v alue is given in the Appendix 10.A.1). This data subset, hereafter called data set #2, is comprised of the 10 solvents shown in Table 10.2. These two evaluations resulted in the follo wing HSP values for CO 2 at 25°C: Data set #1: δd = 16.4 MP a1/2 δp = 5.5 MP a1/2 δh = 5.8 MP a1/2 Data set #2: δd = 15.6 MP a1/2 δp = 5.2 MP a1/2 δh = 5.8 MP a1/2

7248_book.fm Page 184 Tuesday, April 24, 2007 9:19 AM

184

Hansen Solubility Parameters: A User’s Handbook

20 15 10

δh (M Pa)½

25

5

5

10

δd (M Pa)½

15

20

25 0

5

10

15

20

0

δp (M Pa)½

FIGURE 10.1 Interaction radius, where R o incorporates all good solv ents and e xcludes all bad solv ents.

The HSP values derived from the two data sets g ave identical results in terms of the h ydrogen bonding parameter, δh, and similar v alues for δp and δd. To assist in the determination of a final set of HSP v alues, a second approach, kno wn as the solubility sphere,5–7 was also used to evaluate the published solubility data and resulting HSP values for data set #1 and data set #2.The solubility sphere approach is essentially a trial and error method, whereby all the good solv ents are included within a sphere in δd, δp, and δh space, whereas simultaneously excluding all the bad solv ents. The criterion of good v ersus bad is arbitrary and is defined based on the particular interaction being e valuated, such as de gree of polymer swelling, dissolution, barrier breakthrough time, permeation coefficients higher than a given value, suspension time of a pigment, etc. In this e valuation, we are concerned with optimizing solv ents for their CO 2 solubility. Based upon the selected criteria, two-dimensional plots are produced for δd vs. δp, δd vs. δh, and δp vs. δh, and the three circle radii are adjusted until a single optimized radius for all three plots is found. This solubility sphere approach is essentially a trial and error method whereby all the good solvents are included within a sphere inδd, δp, and δh space while simultaneously excluding the bad ones. The resulting radius for the three plots of δd vs. δp, δd vs. δh, and δp vs. δh, is then used to plot a sphere in a three-dimensional plot of δd vs. δp vs. δh. The radius of this sphere is known as the interaction radius, Ro, and is considered a fourth parameter in HSP v alue determinations. Figure 10.1 is a schematic representation of the solubility sphere approach. The adv antage of the solubility sphere approach is that once an interaction radius has been determined, solvents that ha ve not yet been e xperimentally tested for the desired interaction can be quickly screened and, therefore, should be considered for further study . This solubility sphere evaluation is aided by an equation developed by Skaarup for determining the straight-line distance, Ra, in a plot of δd vs. δp vs. δh between tw o materials based on their respecti ve HSP values,5

( Ra )

2

(

= 4 δd 2 − δd1

) + (δ 2

p2

− δ p1

) + (δ 2

h2

− δ h1

)

2

(10.11)

where δd2, δp2, and δh2 are associated with a gi ven solvent and δd1, δp1, and δh1 with the center of the optimized solubility sphere. This equation w as developed from plots of e xperimental data where the leading constant of 4 in the leading right-hand term w as found to correctly represent the solubility data as a sphere encompassing the good solv ents. An extended discussion of the validity of this coefficient is found in Chapter 2. Further confirmation is found in Chapter 9. For the present e valuation, a f avorable interaction is defined as CO 2 solubility greater than ideal at 25°C and pCO2 = 0.1 MPa, whereas an unfavorable interaction is one where the CO2 solubility is less than ideal at these same conditions. It is clear that for cases where CO 2 solubility is greater than ideal, and therefore where the (attracti ve) CO 2–solvent interactions are greater than

7248_book.fm Page 185 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

185

solvent–solvent interactions, Ra should be less than Ro. A convenient index for relati ve goodness of a solvent is the ratio Ra/Ro, which has been called the relative energy difference (RED) number5 (10.12)

RED = Ra Ro

For an indi vidual solv ent, a v alue of RED < 1 indicates f avorable CO 2–solvent interactions, whereas a RED ≈ 1 represents a boundary condition between good and bad. Progressi vely higher values of RED indicate progressively more unfavorable interactions. Computing Ra from Equation 10.11 and the RED from Equation 10.12 allo ws for easy scanning of lar ge data sets, such as the 101 solvents listed in data set #1. The solubility spheres optimized for data set #1 and data set #2 with the two HSP center points are sho wn in Figure 10.2a, Figure 10.2b, and Figure 10.2c. As can be seen from these plots, an interaction radius Ro = 4.0 for data set #2 incorporates the good solvents, whereas for data set #1, an interaction radius Ro = 4.7 is necessary to incorporate the good solvents. In addition, the solubility sphere analysis for data set #1 results in the inclusion of 7 bad solv ents (2-meth ylcyclohexanone, c yclohexanone, oleic acid, dichloromethane, trichloromethane, propylene bromide, and 1,2-dibromoethane), whereas the sphere analysis generated for data set #2 results in the inclusion of only 1 bad solv ent (oleic acid). In terms of the solubility sphere technique, occurrences of good solvents falling outside of the sphere radius, and bad solvents falling inside the sphere radius can be vie wed as an indication of the goodness of the fit. 5

CO2 Solubility in Liquid Solvents, PCO2 = 1 atm., T=25°C 1 Data Set #1

2 Data Set #2 20 18 16

10 8 6 4

δP (MPa)

12

½

14

Alcohols Alkanes/Nonpolar compounds Halogenated compounds Simple acids/Carboxylic acids Esters Ketones Amides Cyclic Alkenes Nitro compounds Aldehyde Amines Nitriles Anhydrides Sulfoxide compounds

2 0 12

13

14

15

16 17 18 ½ δd (MPa) (a)

19

20

21

22

FIGURE 10.2 Two-dimensional plots of CO 2 in organic solvents. (a) δp vs. δd, (b) δh vs. δp, and (c) δh vs. δd.

7248_book.fm Page 186 Tuesday, April 24, 2007 9:19 AM

186

Hansen Solubility Parameters: A User’s Handbook

CO2 Solubility in Liquid Solvents, PCO2 = 1 atm., T=25°C 1 Data Set #1

2 Data Set #2

22 Alcohols Alkanes/Nonpolar compounds

20 18

Halogenated compounds Simple acids/Carboxylic acids Esters Ketones

16

δh (MPa)½

14

Amides

12

Cyclic Alkenes Nitro compounds Aldehyde Amines

10 8 6

Nitriles Anhydrides Sulfoxide compounds

4 2 0

2

4

6

8

10 δp (MPa)

12

14

16

18

20

½

(b) FIGURE 10.2 (Continued)

From the refinement of the two sets of CO2 HSP values, using the solubility sphere methodology, the HSP values from data set #2 were selected as the optimum reference values for CO2 at T=25°C; δ d = 15.6 MPa1/ 2 δ p = 5.2 MPa1/ 2 δ h = 5.8 MPa1/ 2 This determination is further supported based on problems noted by Hansen, 5 who observ ed that the approach of using all solv ents to establish the center of a solubility sphere can result in this sphere boundary (and center) being determined by the poor solv ents or nonsolv ents, rather than the best solv ents in the middle. A comparison of these CO 2 values can be made with the lar ge database of HSP v alues found in Hansen Solubility Parameters: A User’s Handbook.5 Similar v alues are reported for liquid solvents such as diprop yl k etone: δd = 15.8 MP a1/2, δp = 5.7 MP a1/2, and δh = 4.9 MP a1/2; 1,3dimethoxybutane: δd = 15.6 MP a1/2, δp = 5.5 MP a1/2, and δh = 5.2 MP a1/2; and eth yl acetate: δd = 15.8 MPa1/2, δp = 5.3 MPa1/2, and δh = 7.2 MPa1/2. It should be noted, however, that the CO2 optimum HSP reference v alues correspond to a reference temperature of 25°C and a reference pressure of 90.5 MPa8, that is, a higher operating pressure than found in common industrial applications. A methodology for e xtending this reference set of HSP v alues to an y temperature and pressure has been developed and is discussed in the follo wing sections.

7248_book.fm Page 187 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

187

CO2 Solubility in Liquid Solvents, PCO2 = 1 atm., T=25°C 1 Data Set #1

2 Data Set #2

20 18

Alcohols Alkanes/Nonpolar compounds Halogenated compounds

16

Simple acids/Carboxylic acids Esters Ketones Amides

δh (MPa)½

14 12 10 8

Cyclic Alkenes Nitro compounds Aldehyde

6

Amines

4

Nitriles Anhydrides Sulfoxide compounds

2 0

13

14

15

16

17

δd (MPa)

18

19

20

21

22

½

(c) FIGURE 10.2 (Continued)

ONE-COMPONENT HILDEBRAND PARAMETER AS A FUNCTION OF TEMPERATURE AND PRESSURE Hildebrand’s solubility parameter theory was derived from an approximation of the internal pressure of a fluid. This was later termed the cohesi ve energy density (ced), based on w ork conducted in 1928,9 1929, 10 1932, 11 and 1950 12 where the tw o terms, internal pressure and cohesive energy density were found to be related by the quantity , n, as shown in Equation 10.13: ⎛ ∂E ⎞ n ΔE ⎜⎝ ∂V ⎟⎠ = V T

(10.13)

where (ΔE/V) is defined by Hildebrand as ced, and (∂E/∂V)T is the internal pressure . Hildebrand and coworkers found that for nonpolar/nonassociating liquids, where intermolecular interactions are weak, n is not f ar from unity 9–11 and the equality of ced and internal pressure is a good approximation. This same work also demonstrates that n is near unity for nonpolar liquids and also for polar liquids where the dipole moment is less than 2 D , and where specific interactions (particularly hydrogen bonding) are largely absent. 1 D is equal to 1.0 × 10-18 (ESO) or 3.336 × 10–30 Cm. Whereas no direct e valuation of the v alue of n has been found in the literature for carbon dioxide (CO 2), a comparison of the v alues found by Hildebrand and others 13–16 strongly suggests that the value of n for CO 2 is e xpected to be near unity . As a result, the internal pressure and ced are approximately equal.

7248_book.fm Page 188 Tuesday, April 24, 2007 9:19 AM

188

Hansen Solubility Parameters: A User’s Handbook

Accordingly, Hildebrand's solubility parameter, defined as the square root of the ced,12 can also be approximated by the square root of the internal pressure for nonpolar/nonassociating fluids. 1/ 2

1/ 2

⎛ ΔE ⎞ ⎛ ∂E ⎞ ≈⎜ δ=⎜ ⎝ V ⎟⎠ T ⎝ ∂V ⎟⎠ T

(10.14)

And the internal pressure can be calculated fromthe thermodynamic equation of state, Equation 10.15. ⎛ ∂E ⎞ ⎛ ∂P ⎞ ⎜⎝ ∂V ⎟⎠ = T ⎜⎝ ∂T ⎟⎠ − P T V

(10.15)

⎛ ∂E ⎞ ⎛ ∂P ⎞ δ2 ≈ ⎜ =T⎜ −P ⎟ ⎝ ∂V ⎠ T ⎝ ∂T ⎟⎠ V

(10.16)

so that

Total (one-component) solubility parameters can therefore be calculated using an EOS of the form, P = f (ρ,T). This approach is used in this w ork to calculate the total solubility parameter for pure CO 2, using the empirical EOS of Huang et al. 17 ⎡ ⎢1 + b ρ' + b ρ' 2 + b ρ' 3 + b ρ' 4 + b ρ' 5 + b ρ' 2 exp p −c21ρ' 2 + b8 ρ' 4 exp −c21ρ' 2 2 3 4 5 6 7 ⎢ ⎢ 2 2 2 Δρ P = ρ R T ⎢⎢ +c22 ρ' 2 exp ⎡ −c27 ΔT ⎤ + c23 ' exp ⎡ −c25 Δρ − c27 ΔT ⎤ ⎢⎣ ⎦⎥ ⎣⎢ ⎦⎥ ρ ⎢ ⎢ 2 2 ⎢ +c24 Δρ exp ⎡ −c26 Δρ − c27 ΔT ⎤ ' ⎢ ⎥⎦ ⎣ ρ ⎣⎢

(

( )

( )

)

( )

(

( )

( )

⎤

)⎥⎥

⎥ ⎥ (10.17) ⎥ ⎥ ⎥ ⎥ ⎥⎦

where T ′ = T Tc ;

ΔT = 1 − T ′;

ρ′ = ρ ρc ;

Δρ = 1 − 1 / ρ′

(10.18)

and, ⎛ c c c c c ⎞ b2 = ⎜ c1 + 2' + 3' 2 + 4' 3 + 5' 4 + 6' 5 ⎟ ; ⎝ T T T T T ⎠

⎛c ⎞ b6 = ⎜ 14' ⎟ ⎝T ⎠

⎛ c c ⎞ b3 = ⎜ c 7 + 8' + 9' 2 ⎟ ; ⎝ T T ⎠

⎛c c c ⎞ b7 = ⎜ 15' 3 + 16' 4 + 17' 5 ⎟ ⎝T T T ⎠

⎛ c ⎞ b4 = ⎜ c10 + 11' ⎟ ; ⎝ T ⎠

⎛c c c ⎞ b8 = ⎜ 18' 3 + 19' 4 + 20' 5 ⎟ ⎝T T T ⎠

⎛ c ⎞ b5 = ⎜ c12 + 13' ⎟ ⎝ T ⎠ The state constants (c i) are as defined in the Huang reference.

(10.19)

7248_book.fm Page 189 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

189

25 25 δ (MPa)½

15

15 10

SUPERCRITICAL FLUID

10

5

5

δ (MPa)½

20 LIQUID

20

0

0 GAS

0

25

0

30

T (K

)

0 35

0

40

0 0

10

20

30

40

50

)

Pa

M P(

45

FIGURE 10.3 Total (one-component) solubility parameter of pure CO 2 calculated using Equation 10.16 and Equation 10.17.

It should be noted that there are a wide range of a vailable EOSs for carbon dioxide and a comparison of Huang’s and others can be found in a re view by Span and Wagner.18 These equations and the appropriate deri vatives are then used to calculate CO 2 solubility parameters over the temperature and pressure range for which the EOS is stated to be v alid (220 K ≤ T ≤ 420 K, and 0.1 MP a ≤ P ≤ 60 MPa). Figure 10.3 is a plot of the resulting one-component solubility parameters. Other notable w orks include Allada’s19 proposed generalized solubility parameter (that uses analytical equations of state (EOSs), such as Lee-Kelser or modified Redlich-Kwong for evaluation), the modified solubility parameter proposed by Ikushima et al. 20 (where the solubility parameter is expressed in terms of reduced parameters), and the EOS model proposed by P anayiotou.21 This later work utilizes the lattice fluid theory and a lattice fluid h ydrogen bonding model to e valuate solubility parameters and tw o separate components: ph ysical (or van der Waals) and chemical (or specific, e.g., h ydrogen bonding).

THREE-COMPONENT (HANSEN) SOLUBILITY PARAMETERS — PURE CO2 Extending the HSP methodology to supercritical fluids would significantly enhance the understanding of their solv ent properties; ho wever, no such studies appear to ha ve been done. The pressure volume temperature (PVT) EOS that calculates the total (Hildebrand) CO 2 solubility parameter value (Equation 10.16) w as used to determine the combination of pressure and molar v olume ⎛ ⎛ ∂P ⎞ ⎞ corresponding to T = 25°C and δt = ⎜ T ⎜ − P⎟ ⎟ ⎝ ⎝ ∂T ⎠ V ⎠

1/ 2

= 17.4 MP a1/2 that g ave:

P = 91.7 MPa VCO2 = 39.13 cm 3 mole

(10.20)

7248_book.fm Page 190 Tuesday, April 24, 2007 9:19 AM

190

Hansen Solubility Parameters: A User’s Handbook

HSP values at other pressures and temperatures will be based on this set of HSP v alues, using pressure and temperature inte gral functions, which will be deri ved subsequently. Both temperature and pressure will influence total solubility parameters. Ho wever, other than Giddings’ extension of the one-component (Hildebrand) solubility parameter model to supercritical fluids,22 there appears to be no published reports on methods to calculate total solubility parameters as a function of pressure and only limited reports on the calculation of solubility parameters as a function of temperature.5,7,23 Generally, an increase in pressure at constant temperature will increase the total solubility parameter through an increase in the solv ent density. Similarly, an increase in temperature at constant pressure will decrease the total solubility parameter . Both of these trends can be seen in Figure 10.3, where the total CO 2 solubility parameter was calculated using Equation 10.16 and Equation 10.17. The temperature and pressure dependence of individual HSPs as a function of temperature and pressure has apparently not been e valuated for an y liquid, g as, or supercritical fluid. A suggested approach for this calculation is outlined as follo ws where the temperature deri vatives, originally derived by Hansen and Beerbo wer,24 are v erified. Pressure deri vatives, not found in an y literature search, are derived in a manner parallel to the temperature deri vatives. In addition, inte gral forms are developed.

TEMPERATURE AND PRESSURE EFFECTS ON HSPS: δd Hildebrand12 in his 1950 w ork considered the ef fect of temperature on solubility parameters by recalling the e xpression for the dependence of E on the v olume: E=−

k Vn

(10.21)

where k is a constant dependent upon the nature of the particular liquid, and n is about 1.5 for normal (nonassociating or van der Waals) liquids. Substituting Equation 10.21 into Hansen’s definition for the dispersion solubility parameter, k1/2 δ d = − ( n+1) 2 V

(10.22)

allows one to calculate the change in δd produced by a change in v olume by differentiating Equation 10.22. ⎛ ∂δ d ⎞ ⎛ n + 1⎞ = k 1/ 2 ⎜ ⎜⎝ ∂V ⎟⎠ ⎝ 2 ⎟⎠ T ,P

⎡ − ( n + 1) ⎤ ⎢V 2 ⎥ ⎡V −1 ⎤ ⎦ ⎥⎣ ⎢ ⎣ ⎦

(10.23)

⎛ n + 1⎞ ⎛ 1 ⎞ = −δ d ⎜ ⎝ 2 ⎟⎠ ⎜⎝ V ⎟⎠ and, ∂δ d ∂V = −1.25 δd V

(10.24)

7248_book.fm Page 191 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

191

Equation 10.24 can no w be dif ferentiated for either a change in temperature or pressure, or integrated. Results of both deri vations are shown in Table 10.4 and Table 10.5.

TEMPERATURE AND PRESSURE EFFECTS ON HSPS: δp The first v alues of δp were assigned by Hansen and Skaarup using the Böttcher equation, sho wn here as Equation 10.25, δ 2P =

⎡ cal ⎤ 12108 ε − 1 nD2 + 2 μ 2 ⎢ 3 ⎥ V 2 2 ε + nD2 ⎣ cm ⎦

(

)

(10.25)

A simplified equation w as later developed by Hansen and Beerbo wer,24 δp =

37.4μ ⎡ MPa1/ 2 ⎤⎦ V 1/ 2 ⎣

(10.26)

where μ is the dipole moment (in Debyes). This equation is utilized for determining the change in δp with respect to either temperature at constant pressure or with respect to pressure at constant temper ature. ⎛ ∂δ p ⎞ ⎛ 1 ⎞ −3 2 ( 37.4μ ) ⎜⎝ ∂V ⎟⎠ = − ⎜⎝ 2 ⎟⎠ V T ,P δp 1 ⎛ 37.4 μ ⎞ =− =− ⎜ 1/2 ⎟ ⎠ 2V ⎝ V 2V

(10.27)

and ∂δ p δp

=−

∂V . 2V

(10.28)

Equation 10.28 can no w be dif ferentiated for either a change in temperature or pressure, or integrated. Results of both deri vations are shown in Table 10.4 and Table 10.5.

TEMPERATURE AND PRESSURE EFFECTS ON HSPs: δh In Hansen’s early work, the hydrogen bonding parameter w as almost always found by subtracting the polar and dispersion energies of vaporization from the total energy of vaporization. This is still widely used where the required data are available and reliable. Hansen, 5 however, noting that there is no rigorous w ay of arri ving at v alues of the temperature dependence of the h ydrogen bonding solubility parameter , de veloped an empirical approach for the determination of the temperature dependence of δh, which in volves e xperimental heats of v aporization data for h ydrogen-bonded substances, which, in turn, are tak en from Bondi. 25 From Equation 10.4, the h ydrogen bonding solubility parameter , δh, is defined as: δ 2h =

Eh V

(10.29)

7248_book.fm Page 192 Tuesday, April 24, 2007 9:19 AM

192

Hansen Solubility Parameters: A User’s Handbook

so that E h = V δ 2h

(10.30)

where Eh is the hydrogen bonding contribution to the total cohesive energy. Differentiating Equation 10.30 with respect to temperature at constant pressure, ⎛ ∂E h ⎞ ⎛ ∂δ h ⎞ 2 ⎛ ∂V ⎞ ⎜⎝ ∂T ⎟⎠ = V 2 δ h ⎜⎝ ∂T ⎟⎠ + δ h ⎜⎝ ∂T ⎟⎠ P P P

( )

⎛ ∂δ ⎞ ⎛ ∂E ⎞ ⎛ ∂V ⎞ 2V δ h ⎜ h ⎟ = ⎜ h ⎟ − δ 2h ⎜ ⎝ ∂T ⎠ P ⎝ ∂T ⎠ P ⎝ ∂T ⎟⎠ P

⎛ ∂δ h ⎞ ⎜⎝ ∂T ⎟⎠ = P

α,

(10.31)

⎛ ∂E h ⎞ 2 ⎛ ∂V ⎞ ⎜⎝ ∂T ⎟⎠ − δ h ⎜⎝ ∂T ⎟⎠ P P 2V δ h

Simplifying, rearranging terms, and substituting in the isobaric coefficient of thermal expansion,

⎛ ⎛ ∂E h ⎞ ⎞ ⎜ ⎜⎝ ∂T ⎟⎠ ⎟ ⎛ ∂δ h ⎞ α⎟ P ⎜ = δ − h ⎜⎝ ∂T ⎟⎠ ⎜ 2 Eh 2⎟ P ⎜ ⎟ ⎝ ⎠

(10.32)

Bondi,25 through e xploratory calculations, has sho wn that the dif ference between the heat of vaporization of a h ydroxylic compound (a compound displaying strong h ydrogen bonding) and that of its h ydrocarbon (or other nonpolar) homomorph constitutes a good measure of h ydrogen bond strength. This work also discusses the decrease in the heat of formation of the hydrogen bond with increasing temperature. Reference curves of (dEh/dT) were constructed23 for various functional groups and are sho wn in Table 10.3 along with e xperimentally derived values of Eh.25 Averaging the rate of change of the h ydrogen bond heat of v aporization with temperature (dEh/dT), and di viding by the a verage e xcess heats of v aporization (heat of v aporization of the hydrogen bonding compound minus the heat of v aporization of its nonpolar homomorph) results in the following form of Equation 10.32, ⎛ 2.64 × 10 −3 α ⎞ ⎛ ∂δ h ⎞ = − δ + ⎟ h ⎜ ⎜⎝ ∂T ⎟⎠ 2 2⎠ ⎝ P ⎛ α⎞ = −δ h ⎜ 1.32 × 10 −3 + ⎟ 2⎠ ⎝

(10.33)

7248_book.fm Page 193 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

193

TABLE 10.3 Experimentally Determined Values of Eh and (dEh /dT) Functional Group

Hydrogen-Bond Parameter, Eh (cal/mole)

–OH (aliphatic) –NH2 (aliphatic) –CN (aliphatic) –COOH (aliphatic)

4650 1350 500 2750

dEh/dT (cal/mole·K)

± 400 ± 200 ± 200 ± 250

–10 –4.5 –7.0 –2.9

The change in the δh with respect to pressure at constant temperature is obtained by utilizing the relationship: ∂δ h ∂δ h ∂T = ⋅ ∂P ∂T ∂P

(10.34)

∂T β =− ∂P α

(10.35)

⎛ 1.32 × 10 −3 β β ⎞ ⎛ ∂δ h ⎞ = δ + ⎟ h ⎜ ⎜⎝ ∂P ⎟⎠ 2⎠ α ⎝ T

(10.36)

where,

so that

Equation 10.36 can be rearranged to a form that can also be easily inte grated, Table 10.5. The derivative forms are summarized in Table 10.4 and the inte grated forms in Table 10.5. The total solubility parameter, incremented for small changes in temperature and pressure, can be calculated from equations (deri vative form) in Table 10.4,

TABLE 10.4 Equations (Derivative Form) for the Temperature and Pressure Effects on HSP Temperature Increment

δd

δp

δh

Pressure Increment

⎛ ∂δ d ⎞ ⎜ ∂T ⎟ = −1.25δ d α ⎝ ⎠P

⎛ ∂δ d ⎞ ⎜ ∂P ⎟ = 1.25δ d β ⎝ ⎠T

⎛ ∂δ p ⎞ ⎛ α⎞ ⎜ ⎟ = −δ p ⎜ ⎟ ⎝ 2⎠ ⎝ ∂T ⎠ P

⎛ ∂δ p ⎞ ⎛ β⎞ ⎜ ⎟ = δp ⎜ ⎟ ⎝ 2⎠ ⎝ ∂P ⎠ T

⎛ ⎛ ∂δ h ⎞ α⎞ −3 ⎜⎝ ∂T ⎟⎠ = −δ h ⎜⎝ 1.32 × 10 + 2 ⎟⎠ P

⎛ 1.32 × 10 −3 β β ⎞ ⎛ ∂δ h ⎞ = δ + ⎟ h ⎜ ⎜⎝ ∂P ⎟⎠ 2⎠ α ⎝ T

7248_book.fm Page 194 Tuesday, April 24, 2007 9:19 AM

194

Hansen Solubility Parameters: A User’s Handbook

TABLE 10.5 Equations (Integrated Form) for the Temperature and Pressure Effects on HSP δd

δ dref δd

δp

δ pref δp

δh

δ href δh

⎛ Vref ⎞ =⎜ ⎟ ⎝ V ⎠ ⎛ Vref ⎞ =⎜ ⎟ ⎝ V ⎠

−1.25

−0.5

.5 ⎡ ⎛ Vref ⎞ ⎤ −3 ⎢ ⎥ = exp −1.32 × 10 Tref − T − ln ⎜ ⎟ ⎢ ⎝ V ⎠ ⎥ ⎣ ⎦

(

)

2 ⎡ ⎤ ⎡ ⎤ ⎛ ∂δ p ⎞ ⎛ ∂δ p ⎞ ⎛ ∂δ d ⎞ ⎛ ∂δ d ⎞ 2 ⎢δ p + ⎥ + δ = ⎢δ d + ⎜ + + Δ T Δ P Δ T Δ P ⎥ ⎜ ⎟ ⎜ ⎟ ⎜⎝ ∂P ⎟⎠ ⎢ ⎥ ∂ ∂ T P ⎝ ∂T ⎟⎠ P ⎝ ⎠ ⎝ ⎠ ⎢⎣ ⎥ T ⎦ P T ⎣ ⎦

⎡ ⎛ ∂δ ⎞ ⎛ ∂δ ⎞ + ⎢δ h + ⎜ h ⎟ ΔT + ⎜ h ⎟ ΔP T ∂ ⎝ ⎠P ⎝ ∂P ⎠ T ⎢ ⎣

⎤ ⎥ ⎥ ⎦

2

(10.37)

2

or from the equations (inte grated form) in Table 10.5 2 2 ⎡ ⎤ ⎡ ⎡ ⎤ ⎤ ⎢ ⎥ ⎢ ⎢ ⎥ ⎥ ⎢ ⎥ ⎢ δ pref ⎥ ⎢ δ dref ⎥ δ href 2 ⎢ ⎥ δ =⎢ + +⎢ −1.25 ⎥ −0.5 ⎥ 0.5 ⎞ ⎥ ⎢ ⎛ ⎢ ⎛ Vref ⎞ ⎢ ⎛ Vref ⎞ ⎥ ⎥ ⎢ exp ⎜ −1.32 × 10 −3 T − T − ln ⎛ Vref ⎞ ⎟ ⎥ ⎢⎜ ⎢⎜ ⎥ ⎥ ⎟ ⎟ ref ⎜ V ⎟ ⎢ V V ⎠ ⎠ ⎜⎝ ⎝ ⎠ ⎟⎠ ⎥ ⎢⎣ ⎝ ⎢⎣ ⎝ ⎥⎦ ⎥⎦ ⎢⎣ ⎥⎦

(

2

)

(10.38)

where the reference v alues are as determined earlier; δdref = 15.6 MP a1/2, δpref = 5.2 MP a1/2, δhref = 5.8 MPa1/2, Vref = 39.13 cm 3/mole, and Tref =298.15 K. CO2 HSP values calculated with the equations in Table 10.4, as a function of temperature and pressure, are shown in the CO 2 HSP surf ace diagrams illustrated in Figure 10.4. From this w ork though, CO 2 HSP v alues at a temperature of 25°C and a pressure of 200 bar , δd = 12.2 MPa1/2, δp = 4.7 MPa1/2, and δh = 5.2 MPa1/2 can also be compared to liquid solv ent HSP values. Comparible liquid solvents include chlorodifluoromethane: δd = 12.3 MPa1/2, δp = 6.3 MPa1/2, and δh = 5.7 MP a1/2; isopropyl ether: δd = 13.7 MP a1/2, δp = 3.9 MP a1/2, and δh = 2.3 MP a1/2; and vinyl trifluoroacetate: δd = 13.9 MP a1/2, δp = 4.3 MP a1/2, and δh = 7.6 MP a1/2. It is also interesting to note that the total solubility parameter v alue of he xane, a solv ent that CO 2 is often closely compared to 26–28 (δt = 14.9 MP a1/2) is very near the total solubility parameter of CO 2 (at 25°C and 200 bar), δt = 14.0 MPa1/2. Yet, when the two are compared in terms of HSP values, little similarity is noted; he xane: δd = 14.9 MP a1/2, δp = 0.0 MP a1/2, and δh = 0.0 MP a1/2.

7248_book.fm Page 195 Tuesday, April 24, 2007 9:19 AM

20 17.5 15 12.5 10 7.5 5 2.5 0 2

50

30

0

0 35

T (K

)

40

0 45

10

00

20

30

40

20 17.5 15 12.5 10 7.5 5 2.5 0

50

195

δd (MPa)½

δd (MPa)½

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

)

Pa

M P(

(a) CO2 dispersion parameter as a function of temperature and pressure.

5

5

4

4

3

3

2

2

1

1

δp (MPa)½

δp (MPa)½

6 6

0

0 25

0 3

00 3

T (K

50

40

)

0

0 45

0

10

20

30

40

50 )

Pa

M P(

7 6 5 4 3 2 1 0

7 6 5 4 3 2 1 0 2

50

30

T (K

0

)

3

50

40

0

0 45

0

10

20

30

40

δh (MPa)½

δh (MPa)½

(b) CO2 polar parameter as a function of temperature and pressure 45K)

50 )

Pa

M P(

(c) CO2 hydrogen bonding parameter as a function of temperature and pressure

FIGURE 10.4 HSP values for CO 2 as a function of T and P. (a) CO 2 dispersion parameter as a function of temperature and pressure. (b) CO2 polar parameter as a function of temperature and pressure. (c) CO2 hydrogen bonding parameter as a function of temperature and pressure.

7248_book.fm Page 196 Tuesday, April 24, 2007 9:19 AM

196

Hansen Solubility Parameters: A User’s Handbook

CONCLUSION A set of Hansen solubility parameters at T = 25°C ha ve been determined for CO 2, based on the room temperature solubility in different liquid solvents of known HSP values: δd = 15.6 MPa1/2, δp = 5.2 MP a1/2, and δh = 5.8 MP a1/2. Further, this set of HSPs were refined using the RED af finity number and Hansen solubility plots to correlate the solubility of carbon dioxide in the solv ents identified in Table 10.1. It is important to note that the solubility parameter, or rather the difference in solubility parameters, for a given solvent–solute combination has been foremost in determining the mutual solubility of the system.5 An analogy to “lik e dissolv es lik e” is appropriate. Therefore, the accurac y to which the solubility parameters for a binary pair can be kno wn will be v aluable in predicting the system’ s behavior. This work introduces a theoretical methodology for generating solubility parameter v alues, both one-component Hildebrand and three-component Hansen parameters, for a pure supercritical fluid, using CO 2 as an e xample. The ability to e xpress molecular interactions, in terms of HSPs, for a pure fluid solv ent in a w ay that unites the liquid, g as, and supercritical phases, represents an advancement in the understanding of the role of solv ents in both e xisting and new applications.

ACKNOWLEDGMENTS Special thanks to James Rubin of Los Alamos National Laboratory for his excellent assistance and collaboration in the de velopment of this w ork.

CHAPTER 10 ADDENDUM Research has been the k ey to man y progressi ve de velopments. Significant steps to ward impro ved understanding have been made with limited resources, and these result in still other impro vements in the same direction when other resources are applied. The HSP for carbon dioxide reported in the chapter ( δd, δp, and δh equal to 15.6 MP a1/2, 5.2 MP a1/2, and 5.8 MP a1/2) were found by a plotting technique with a radius for the solubility sphere of 4.0 MP a1/2. The data used were collected from a wide variety of sources, and the criterion for a good solvent was solubility in excess of the theoretical. A computer analysis of the same data has no w shown that it is possible to describe the solubility of carbon dioxide with a slightly dif ferent solubility sphere. The δd, δp, and δh values 15.7 MP a1/2, 6.3 MPa1/2, and 5.7 MP a1/2 were found to gi ve a perfect data fit of 1.000 (v ersus 0.981) with a radius of only 3.3 MPa1/2. Both correlations emphatically sho w the ability of this procedure, whether by hand or by computer , to correlate g as/liquid solubility data for a wide v ariety of chemically dif ferent solvents. It has not been possible, nor has it been deemed necessary , to re vise the contents of this chapter using these slightly dif ferent numbers. This note is only to indicate wh y a slightly dif ferent set of HSP is reported else where in this handbook (T able A-1 and Chapter 13, Table 13.4).

7248_book.fm Page 197 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

197

SYMBOLS SPECIAL TO THIS CHAPTER KH L P ai fi fio k pi pis Ω β ρ

Henry’s law constant (Equation 10.5) Ostwald coefficient in Equation 10.7 Pressure Activity coefficient of the “i”th component in Appendix 10.A.1, Equation 10.A.1 Fugacity of the “i”th component in Appendix 10.A.1, Equation 10.A.1 Fugacity at standard state in Appendix 10.A.1, Equation 10.A.1 Constant in Equation 10.21–Equation 10.23 Partial pressure of the “i”th component in Equation 10.A.2 Saturation pressure of the “i”th component in Equation 10.A.2 Bunsen coefficient in Equation 10.6 Compressibility Density

REFERENCES 1. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities. I. Solvents, plasticizers, polymers, and resins, J. Paint Technol., Vol. 39, No. 505, 104–117, 1967. 2. Wilhelm, E. and Battino, R., Thermodynamic functions of the solubilities of gases in liquids at 25°C, Chem. Rev., Vol. 73, No. 1, 1–9, 1973. 3. Battino, R. and Cle ver, H.L., The solubility of g ases in liquids, Solutions and Solubilities, Dack, M.R.J., Ed., John Wiley & Sons, Ne w York, 1965, p. 379, chap. 7. 4. Reid, R.C., Prausnitz, J.M., and Poling, B.E., The Properties of Gases and Liquids, 4th ed., McGraw Hill, New York, 1987, p. 334. 5. Hansen, C.M., Hansen Solubility Parameters: A User’s Handbook, CRC Press, Boca Raton, FL, 1999. 6. Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Copenhagen Danish Technical Press, Denmark, 1967, pp. 33–38. 7. Barton, A.F.M., CRC Handbook of Solubility Parameters and Other Cohesion Parameters, 2nd ed., CRC Press, Boca Raton, FL, 1991. 8. Williams, L.L., Rubin, J.B., and Edw ards, H.W., Calculation of Hansen solubility parameter v alues for a range of pressure and temperature conditions, including the supercritical fluid re gion, Ind. Eng. Chem. Res., Vol. 43, 4967–4972, 2004. 9. Westwater, W., Frantz, H.W., and Hildebrand, J.H., The internal pressure of pure and mix ed liquids, Phys. Rev., Vol. 31, 135–144, 1928. 10. Hildebrand, J.H., Intermolecular forces in liquids, Phys. Rev., Vol. 34, 984–993, 1929. 11. Hildebrand, J.H. and Carter, J.M., A study of v an der Waals forces between tetrahalide molecules, J. Am. Chem. Soc., Vol. 54, 3592–3603, 1932. 12. Hildebrand, J.H. and Scott, R.L., The Solubility of Nonelectrolytes, 3rd ed., Reinhold, New York, 1950. 13. Dack, M.R.J., Solvent structure. The use of internal pressure and cohesi ve energy density to examine contributions to solv ent-solvent interactions, Aust. J. Chem., Vol. 28, 1643–1648, 1975. 14. Renuncio, J.A.R., Breedveld, G.J.F., and Prausnitz, J.M., Internal pressures and solubility parameters for carbon disulfide, benzene, and c yclohexane, J. Phys. Chem., Vol. 81, No. 4, 324–327, 1977. 15. Allen, G., Gee, G., and Wilson, G.J., Intermolecular forces and chain fle xibilities in polymers: I. Internal presssures and cohesive energy densities of simple liquids, Polymer, Vol. 1, No. 4, 456–476, 1960. 16. MacDonald, D.D. and Hyne, J.B., The thermal pressure and energy-volume coefficients of the methyl alcohol-water and t-b utyl alcohol-water systems, Can. J. Chem., Vol. 49, 2636–2642, 1971. 17. Huang, F., Li, M., Lee, L., and Starling, K., An accurate equation of state for carbon dioxide, J. Chem. Eng. Jpn., Vol. 18, No. 6, 490–496, 1985.

7248_book.fm Page 198 Tuesday, April 24, 2007 9:19 AM

198

Hansen Solubility Parameters: A User’s Handbook

18. Span, R. and Wagner, W., A new equation of state for carbon dioxide co vering the fluid re gion from the triple-point temperature to 1100 K at pressures up to 800 MP a, J. Phys. Chem. Ref. Data, Vol. 25 No. 6, 1509, 1996. 19. Allada, S.R., Solubility parameters of supercritical fluids, Ind. Eng. Chem. Process Des. Dev., Vol. 23, 344, 1984. 20. Ikushima, Y., Goto, T., and Arai, M., Modified solubility parameter as an inde x to correlate the solubility in supercritical fluids, Bull. Chem. Soc. Jpn., Vol. 60, 4145, 1987. 21. Panayiotou, C., Solubility parameter revisited: an equation-of-state approach for its estimation, Fluid Phase Equilibria, Vol. 131, 21, 1997. 22. Giddings, C.J., Myers, M.N., McLaren, L., and K eller, R.A., High Pressure Gas Chromatograph y of Nonvolatile Species, Science, Vol. 162, 1968, pp. 67–73. 23. Bondi, A., Physical Properties of Molecular Crystals, Liquids, and Glasses, John Wiley & Sons, New York, 1968. 24. Hansen, C. and Beerbo wer, A., Solubility parameters, Kirk-Othmer Encyclopedia of Chemical Technology, Suppl. Vol., Interscience, New York, 1971. 25. Bondi, A. and Simkin, D.J., Heats of v aporization of h ydrogen bonded substances, AIChE J., Vol. 3 No. 4, 473, 1957. 26. O’Neill, M.L., Cao, Q., F ang, M., Johnston, K.P., Wilkinson, S.P., Smith, C.D., K erschner, J.L., and Jureller, S.H., Solubility of homopolymers and copolymers in carbon dioxide, Ind. Eng. Chem. Res., Vol. 37, 3067–3079, 1998. 27 Fedotov, A.N., Sinevich, E.A., and Simonov, A.P., Intermolecular interactions of supercritical carbon dioxide with polymers of dif ferent types, Russ. J. Phys. Chem., Vol. 71, No. 11, 1900–1903, 1997. 28. Hyatt, J.A., Liquid and supercritical carbon dioxide as or ganic solv ents, J. Org. Chem., Vol. 49, 5097–5101, 1984. 29. Just, G., Z. Phys. Chem., Vol. 37, p. 342, 1901. 30. Kunerth, W., Solubility of CO 2 and N 2O in certain solv ents, Phys. Rev., Vol. 19, 512–524, 1922. 31. De Ligny, C.L. and v an der Veen, N.G., On the applicability of Pierotti’ s theory for the solubiliy of gases in liquids, J. Solution Chem., Vol. 4, No. 10, 841–851, 1975. 32. Pirig, Y.N., Polyuzhin, I.V., and Makitra, R.G., Carbon dioxide solubility , Russ. J. Appl. Chem., Vol. 66, No. 4, P art 2, 691–695, 1993. 33. Podvigaylova, I.G., Zaynalov, B.K., Kruglikov, A.A., Radzhabov, D.T., Shagidanov, E.N., Shestakova, T.G., K orbutova, Z.V ., and Mel’nik ova, L.I., Solubility of CO2 in Organic Solvents, The Soviet Chemical Industry, No. 5, pp. 19–21, 1970. 34. Gjaldbaek, J.C. and Anderson, E.K., The solubility of carbon dioxide, oxygen, carbon monoxide, and nitrogen in polar solv ents, Acta Chem. Scand., Vol. 8, 1398–1413, 954. 35. Gerrard, W., Solubility of Gases and Liquids: A Graphic Approach, Plenum Press, Ne w York, 1976, p. 73. 36. Gjaldbaek, J.C., The solubility of carbon dioxide in perfluoro-n-heptane, normal heptane, c yclohexane, carbon tetrachloride, benzene, carbon disulphide, and aqueous solution in aerosol,Acta Chem. Scand., Vol. 7, 537–544, 1953. 37. Pardo, J., Lopez, M.C., Santafe, J., Ro yo, F.M., and Urieta, J.S., Solubility of g ases in b utanols. I., Fluid Phase Equilibria, Vol. 109, 29–37, 1995. 38. Fogg, P.G.T., Ed., Carbon Dioxide in Nonaqueous Solvents, Solubility Data Series, Vol. 50 1992 International Union of Pure and Applied Chemistry (IUPAC), Research Triangle Park, NC. 39. Pardo, J., Lopez, M.C., Mayoral, J.A., Ro yo, F.M., and Urieta, J.S., Solubility of g ases in b utanols. III., Fluid Phase Equilibria, Vol. 134, 133–140, 1997. 40. Pardo, J., Mainar , A.M., Lopez, M.C., Ro yo, F.M., and Urieta, J.S., Solubility of g ases in b utanols. IV., Fluid Phase Equilibria, Vol. 155, 127–137, 1999. 41. Kobatake, Y. and Hildebrand, J.H., Solubility and entrop y of solution of He, N 2, A, O2, CH 4, C 2H6, CO2 and SF6 in various solvents; regularity of gas solubilities, J. Phys. Chem., Vol. 65, 331–334, 1961. 42. Gironi, F. and La vecchia, R., A simple method for determining the solubility of g ases in liquids: application to CO 2-cycloparaffin systems, Fluid Phase Equilibria, Vol. 87, 153–161, 1993. 43. Gallardo, M.A., Lopez, M.C., Urieta, J.S., and Losa, C.G., Solubility of 15 non-polar g ases in cycloheptanone, Fluid Phase Equilibria, Vol. 58, 159–172, 1990.

7248_book.fm Page 199 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

199

44. Dymond, J.H., The solubility of a series of g ases in c yclohexane and dimeth ylsulfoxide, J. Phys. Chem., Vol. 71, 1829–1831, 1967. 45. Begley, J.W., Maget, H.J.R., and Williams, B., Solubility of carbon dioxide in c yclohexanol, 1,2dibromoethane, a mixture of 1-chloro-2-bromopropane and 2-chloro-1-bromopropane, and mineral oil, J. Chem. Eng. Data, Vol. 10, No. 1, 4–8, 1965. 46. Gallardo, M.A., Melendo, J.M., Urieta, J.S., and Losa, C.G., Solubility of non-polar g ases in c yclohexanone between 273.15 and 303.15 K at 101.32 kP a partial pressure of g as, Can. J. Chem., Vol. 65, 2198–2202, 1987. 47. Wilcock, R.J., Battino, R., and Wilhelm, E., The solubility of g ases in liquids 10., J. Chem. Thermodynamics, Vol. 9, 111–115, 1977. 48. Gallardo, M.A., Melendo, J.M., Urieta, J.S., and Losa, C.G., Solubility of He, Ne, Ar, Kr, Xe, H 2, D2, N 2, O 2, CH 4, C 2H4, C 2H6, CF 4, SF 6, and CO 2 in c yclopentanone from 273.15 and 303.15 K at 101.32 kPa partial pressure of g as, Fluid Phase Equilibria, Vol. 50, 223–233, 1989. 49. King, M.B. and Al-Najjar, H., The solubilities of carbon dioxide, h ydrogen sulphide and propane in some normal alkane solv ents I, Chem. Eng. Sci., Vol. 32, 241–1246, 1977. 50. Wilcock, R.J., Battino, R., Danforth, W.F., and Wilhelm, E., J. Chem. Thermodynamics, Vol. 10, 817–822, 1978. 51. Gallardo, M.A., Melendo, J.M., Urieta, J.S., and Losa, C.G., Solubility of nonpolar g ases in 2,6dimethylcyclohexanone, Can. J. Chem., Vol. 68, 435–439, 1990. 52. De Ligny, C.L. and van der Veen, N.G., A test of Pierotti’s theory for the solubility of gases in liquids, by means of literature data on solubility and entrop y of solution, Chem. Eng. Sci., Vol. 27, 391–401, 1972. 53. Hayduk, W., Walter, E.B., and Simpson, P ., Solubility of propane and carbon dioxide in heptane, dodecane, and he xadecane, J. Chem. Eng. Data, Vol. 17, No. 1, 59–61, 1972. 54. Jou, F.-Y., Deshmukh, R.D., Otto, F .D., and Mather , A.E., Solubility of H 2S, CO 2 and CH 4 in NFormyl Morpholine, J. Chem. Soc., Faraday Trans. I, Vol. 85, No. 9, 2675–2682, 1989. 55. Field, L.R., Wilhelm, E., and Battino, R., The solubility of g ases in liquids. 6. , J. Chem. Thermodynamics, Vol. 6, 237–243, 1974. 56. Gallardo, M.A., Lopez, M.C., Urieta, J.S., and Losa, C.G., Solubility of nonpolar g ases in 2-methylcyclohexanone between 273.15 and 303.15 K at 101.32 kP a partial pressure of g as, Can. J. Chem., Vol. 67, 809–811, 1989. 57. Makranczy, J., Rusz, L., and Balog-Me gyery, K., Hung. J. Ind. Chem., Vol. 7, No. 1, 41–46, 1979. 58. Tokunaga, J., J. Chem. Eng. Data, Vol. 20, 41–46, 1975. 59. Hiraoka, H. and Hildebrand, J.H., The solubility and entrop y solution of certain g ases in (C4F9)3N, CCl2F CClF2, and 2,24-(CH3)3C5H9, J. Phys. Chem., Vol. 68, No. 1, 213–214, 1964. 60. Byrne, J.E., Battino, R., and Wilhelm, The solubility of gases in liquids. 8., J. Chem. Thermodynamics, Vol. 7, 515–522, 1975.

APPENDIX 10.A.1: IDEAL SOLUBILITY OF GASES IN LIQUIDS AND PUBLISHED CO2 SOLUBILITY DATA. Published CO2 gas solubility data in 103 liquid solvents were gathered from the available literature and is presented in Table 10.1. Two of these solv ents, triethylamine and 1,4-dioxane were subsequently deleted from the data set based on their kno wn tendency to chemically react with CO 2.1

IDEAL SOLUBILITY OF GASES IN LIQUIDS Solutions that come close to approximating ideal solutions are those which are very dilute, or those where the molecular species are so nearly alik e that a gi ven molecule is subject to the same intermolecular forces (both attracti ve and repulsi ve) in the mixture as in its o wn pure phase. (In very dilute solutions, the intermolecular forces on a solute molecule may be quite dif ferent than in the pure solute phase, but the solute molecules are far enough apart that solute–solute interactions do not manifest themselves.) The concept of an ideal solution is often an appropriate approximation

7248_book.fm Page 200 Tuesday, April 24, 2007 9:19 AM

200

Hansen Solubility Parameters: A User’s Handbook

TABLE 10.A.1 Ideal Carbon Dioxide Solubility Calculated Using Raoult’s Law T (°C)

s PCO2 (atm)

0 15 20 25 30

34.40 50.19 56.60 63.50 71.12

ideal =1/P s xCO2 CO2

0.0291 0.0199 0.0177 0.0157 0.0141

Note: pCO2 = 1 atm.

for g ases dissolv ed in liquids, as at modest pressures; most g ases are only sparingly soluble in typical liquids. 2 Thermodynamically, an ideal solution is defined as one in which the activity, a, equals the mole fraction, xi, over the entire composition range and over a nonzero range of temperature and pressure.3 xi = ai =

fi fi o

(10.A.1)

The acti vity of a substance gi ves an indication of ho w acti ve a substance is relati ve to its standard state, as it provides a measure of the difference in chemical potential at the state of interest and that at the standard state. 4 The term fugacity, f, w as introduced by Le wis5 as a measure of thermodynamic escaping tendency and is equal to the effective gas pressure corrected for deviations from ideality. In Equation 10.A.1, fi is the fug acity of component i at partial pressure pi, and fio is the fugacity at the saturation pressure of i, Pis, at the solution temperature. Equation 10.A.1 is an empirical rule suggested by Le wis and Randall 5 that assumes imperfect g as mixtures to beha ve as ideal mixtures. When de viations from the ideal g as la w are small, generally at lo w pressures, the ef fect of pressure on the fug acity of component i is ne gligible, and the fug acity terms in Equation 10.A.1 approach the partial pressure and saturation pressure of i, respectively. In this situation, therefore, the ratio of the partial pressure and saturation pressure can now be used to express the mole fraction, xi . xi =

pi Pi s

(10.A.2)

Equation 10.A.2 is kno wn as Raoult’s law, and the mole fraction, as calculated from Raoult’ s law, is referred to as the ideal g as solubility. The ideal solubility calculated from Equation 10.A.2 usually gi ves correct order of magnitude results pro vided that Pis is not lar ge and the solution temperature is well belo w the critical temperature of the solv ent and not e xcessively abo ve the critical temperature of the g aseous solute. 4 Table 10.A.1 e valuates the ideal solubility of CO 2, calculated using Equation 10.A.2, for the temperature range 0°C to 30°C. ideal From Table 10.A.1, Raoult’ s la w predicts an ideal CO 2 solubility of xCO 2 = 0.0157 at T = 25°C, and this v alue has been used in se veral of the published CO 2 solubility studies. 2,4,60

7248_book.fm Page 201 Tuesday, April 24, 2007 9:19 AM

Determination of Hansen Solubility Parameter Values for Carbon Dioxide

201

TABLE 10.A.2 Ideal Carbon Dioxide Solubility Calculated Using Raoult’s Law and Fugacities T (°C)

s PCO2 (atm)

fCO2 PCO2 = 1 atm

0 15 20 25 30

34.4 50.19 56.6 63.5 71.12

.9928 .9941 .9944 .9947 .9950

o fCO2 at s PCO2

26.51 36.04 39.6 43.3 47.17

ideal xCO 2

ideal xCO 2

s = 1/PCO2

o = fCO2/fCO2

0.0291 0.0199 0.0177 0.0157 0.0141

0.0375 0.0276 0.0251 0.0229 0.0211

% Diff. 28 39 42 46 49

Note: pCO2 = 1 atmosphere.

It has been noted by Prausnitz et al.4 that the simplest way to reduce Equation 10.A.1 to a more useful form is to rewrite it in the manner suggested by Raoult’ s law, Equation 10.A.2. In doing so, however, they caution that se veral assumptions are made, and errors in the use of this estimation technique for the solubility of g ases in liquids can be significant, especially , when the saturation pressure of the g as is high. Therefore, in cases where the saturation pressure of the g as is above 1 atmosphere, it is necessary to consider the error in using pi /pis instead of fi /fi.0 Table 10.A.2 gi ves the saturation pressures of CO 2 for the temperature range 0°C to 30°C, as well as the fug acities and calculated ideal solubilities using both methodologies. From Table 10.A.2 it appears that the assumption of Raoult’s law for the determination of ideal CO2 solubility in liquids results in significant error . The ideal CO 2 solubility at 25°C and 1 ideal atmosphere partial pressure, as calculated from CO 2 fugacities,7 is xCO 2 = 0.0229 compared with ideal a Raoult’s law prediction of xCO 2 = 0.0157. This value was also used by Gjaldbaek et al. 8,9 in their work comparing e xperimental and calculated CO 2 gas solubilities.

REFERENCES 1. Charles M. Hansen, communication. 2. Reid, R.C., Prausnitz, J.M., and Poling, B.E., The Properties of Gases and Liquids, 4th ed., McGraw Hill, New York, 1987. 3. Hildebrand, J.H., Prausnitz, J.M., and Scott, R.L., Regular and Related Solutions, Van Nostrand Reinhold Company, New York, 1970. 4. Prausnitz, J.M., Lichtenthaler , R.N., de Azevedo, E.G., Molecular Thermodynamics of Fluid-Phase Equilibria, 2nd ed., Prentice-Hall, Engle wood Cliffs, New Jersey, 1986. 5. Lewis, G.N. and Randall, M.,Thermodynamics and the Free Energy of Chemical Substances, McGrawHill, New York, 1923. 6. Pirig, Y.N., Polyuzhin, I.V., and Makitra, R.G., Carbon dioxide solubility , Russ. J. Appl. Chem., Vol. 66, No. 4, P art 2, 691–695, 1993. 7. King, M.B., Phase Equilibrium in Mixtures, Pergamon Press Ltd., Oxford, U.K., 1969. 8. Gjaldbaek, J.C., The solubility of carbon dioxide in perfluoro-n-heptane, normal heptane, c yclohexane, carbon tetrachloride, benzene, carbon disulphide and aqueous solution of aerosol, Acta Chem. Scand., Vol. 7, 537–544, 1953. 9. Gjaldbaek, J.C. and Anderson, E.K., The solubility of carbon dioxide, oxygen, carbon monoxide and nitrogen in polar solv ents, Acta Chem. Scand., Vol. 8 1398–1413, 1954.

7248_book.fm Page 202 Tuesday, April 24, 2007 9:19 AM

7248_C011.fm Page 203 Thursday, May 10, 2007 8:00 AM

of Hansen Solubility 11 Use Parameters to Identify Cleaning Applications for “Designer” Solvents John Durkee ABSTRACT Choosing the most ef fective solv ents for cleaning requires characterization of both solv ents and soils. Hansen solubility parameters (HSP) can be used whether or not single or multiple components are present in either the soil composite or the solv ent blend. These HSP characterizations of solv ents and soils can be compared as v alues in HSP space just as similar values of solvents and polymers are compared to determine if a solv ent can dissolve a polymer. A previously developed value of R A for cleaning operations allo ws prediction of useful cleaning performance by solv ent on soil. In this chapter , six representati ve soils and 13 base solv ents, so-called “designer” solv ents, were selected as feedstocks for formulation of binary azeotropes, using a literature database. The selected solvent azeotropes were predicted to clean fi e or six representati ve soils. In this chapter, it will be sho wn that the base azeotropic feedstock must display significant polar an hydrogen bonding intermolecular forces if the azeotrope is to clean a soil composite that also contains significant polar and ydrogen bonding intermolecular forces.

INTRODUCTION Since the early 1990s, more than a dozen ne w solvents have been commercialized for cleaning 1, such as vapor degreasing, and other applications. The sobriquet “designer” has been applied to them as their structure w as designed to pro vide certain benefits. Chief , these benefits are a relat vely nonhazardous e xposure to humans and compliance with nearly all en vironmental re gulations (including the Montreal Protocol) of most nations. As designer goods vs. commodities, the y are high priced, which limits applications. Another limitation is their character as solv ents. Given the relative inertness of these designer solvents to both humans and the en vironment, it is not surprising to find them to be inert to ma y soils. The purpose of this chapter is to illustrate ho w to use HSPs to aid in identifying v aluable applications for these and other solv ents in cleaning applications. The approach tak en is the traditional one used in cleaning work: matching the solvent to the soil. 2 HSP values of the solvents will be those matched to soil materials. 3

203

7248_C011.fm Page 204 Thursday, May 10, 2007 8:00 AM

204

Hansen Solubility Parameters: A User’s Handbook

A VARIETY OF SOLVENTS There are almost too man y types of cleaning solv ents. It can be hard to discriminate among them. Simple chemicals such as he xane, more comple x chemicals such as N-methyl-2-pyrrolidone, flammable chemicals such as acetone, carcinogenic chemicals such as benzene, ozone-depletin chemicals such as carbon tetrachloride, and lo w-cost common chemicals such as w ater ha ve all been used as cleaning solv ents. One of the strengths of solv ent-cleaning technology is the v ariety of solvents that can be used in v arious solvent cleaning processes. Yet, no cleaning solv ent is perfect. All ha ve fl ws, both general and specific. Sol ents with general fl ws4 may pose environmental, safety, or health hazards. They may be physically unsuited to the application because of mismatched ph ysical properties such as surf ace tension, density , or viscosity; have excessive or limited v olatility; or simply be too e xpensive. However, the specific f w that is usually f atal to an application is the solv ent’s not being compatible with the soil materials. Pre vention of a mismatch of intermolecular forces between a cleaning solvent and a soil (an incompatibility) is a role uniquely fulfilled by HS .5 A soil material is probably soluble in a solv ent (or solv ent blend) if the Hansen parameters for the solv ent lie within the solubility sphere for the soil. 6

PATHOLOGY OF SOILS Those managing cleaning w ork must know at least as much about soils as the y do about cleaning agents, if their cleaning w ork is to meet their requirements. After all, the soil is the enemy to be conquered. Soils are multicomponent mixtures. They are composed of occasionally unkno wn ingredients prepared with unkno wn and v ariable proportions, which are not homogenized. They may include “tramp” materials or contain une xpected components that are not al ways present or recognized at best, often inadv ertently produced by users. Soil can also be the outcome of incomplete cleaning work. Cleaning of multiple soil components usually produces multiple cleaning results. The better matched cleaning agents and processes are to a soil, the better the result, while poor matches will still be apparent. A good e xample is h ydrocarbon-based w ax, which must normally be melted in order to be removed from surfaces. Microcrystalline wax, used as a binder in polishing compounds, has a sharp melting point because the range of molecular weights of its component paraf fins is narr w. Wax used in box-coating operations has a mixture of molecular weights and a broad melting point. If the boiling point of the cleaning solv ent only matches the a verage melting point of both w ax mixtures, the higher melting fractions of the coating wax probably will not be well-cleaned because they are not heated to a high enough temperature. In other words, the soil least matched to the cleaning solvent is likely to be the soil least cleaned. (See “More Realistic View about Evaluating HSP of Composite Soils. ”)

HSP OF MULTIPLE-COMPONENT SOILS How then is a user to match a single (or azeotropic) cleaning solv ent with a multicomponent soil? The answer combines observ ation along with conventional solubility theory and is seen in the following list: •

First, one observes7 if the soil is a single- or multiple-phase mixture. If the soil material is a single phase (presumably a solution and not an emulsion), the components must have similar intermolecular forces, and a useful solvent can probably be found from their study. If the mixture contains multiple phases, it is lik ely that a cleaning process must

7248_C011.fm Page 205 Thursday, May 10, 2007 8:00 AM

Polar HSP (soil or solvent)

Use of Hansen Solubility Parameters

205

#2 Soil Component #1 HSP Parameters for Common Soil or Solvent

H2 Bonding HSP (soil or solvent) FIGURE 11.1

•

•

be chosen that is not based on solvency but on application of mechanical force (aqueous cleaning, for e xample) or the chosen cleaning process must ha ve multiple steps. Second, one estimates the HSP v alues for a composite of soil materials. Information about the individual soil components is used here. This usually involves some compromise. For example, if the soil contains multiple components, HSP values for the two that represent the greatest volume fraction should be used. This is illustrated in Figure 11.1. 8 If that means other components are not represented in estimation of HSP v alues, then it is assumed that the observ ation of the soil form as a single phase w ould mean that the unrepresented components must be similar to those represented. Whereas it is perfectly possible (see next two sections) to compute the composite HSP using the entire composition, the recipe for that composition is only seldom kno wn.9 Third, one chooses a solv ent whose HSP v alues are similar . This, too, in volves some compromise. The choice of a solvent (or blend) similar to a soil component (or composite) is accomplished in a similar manner as solv ent for a polymer 10 is chosen (see section on “Method for Choice of Suitable Solv ents”). In other w ords, the required compromise solvent should be no further than the outer boundary of the lar gest circle (see discussion following).

METHOD FOR CALCULATING HSP OF COMPOSITES (SOILS OR SOLVENTS) Solubility parameters11 of mixtures are linear. That is, each of the three HSPs (disperse, polar , and hydrogen bonding) of a solv ent mixture is a linear function of composition. In this case, the composition v alue to be used in calculating solubility parameters for solv ent mixtures is the volume fraction (φ) for each component. 12 For a binary (tw o-solvent) mixture, the equation for all four13 solubility parameters is Equation 11.1. 14 This equation is correct for more than two components where the HSP 14 values are known.15,16 δ blend ≡ ⎡⎣ϕ comp1 × σ com1 ⎤⎦ + ⎡⎣ϕ comp 2 × σ com 2 ⎤⎦

(11.1)

Traditionally, without specific data, it is normally assumed that there is no olume change upon mixing of solv ents. That is:

7248_C011.fm Page 206 Thursday, May 10, 2007 8:00 AM

206

Hansen Solubility Parameters: A User’s Handbook

TABLE 11.1 Calculation of HSP for Mixtures via Equation 11.1 Component 1 2

Solvent

Mol Density δDispersion δPolar δHydrogen-Bonding Wt. g/cc MPa1/2 MPa1/2 MPa1/2

Cyclohexane 84.2 Isopropanol 60.1

0.779 0.789

16.8 15.8

0.0 6.1

0.2 16.4

Wt. Vol. δHydrogen% % ϕ δDispersion δPolar Bonding 67.0 67.2 33.0 133.8

16.5

⎛ Wt . Fraction ⎞ ⎜ ⎟ ⎝ Density ⎠1 (Vol. Fraction)1 = ⎛ Wt . Fraction ⎞ ⎛ Wt . Fraction ⎞ ⎟ ⎜ ⎟ +⎜ ⎝ Density ⎠1 ⎝ Density ⎠2

2.0

5.5

(11.2)

An azeotropic mixture of cyclohexane and isopropanol is used as an example of this calculation. The data needed and results are gi ven in Table 11.1. Note, for this e xample only, that the v olume and weight 17 concentrations are essentially the same. This is because the indi vidual solvent density values are essentially the same. Also note that the value of adding isopropanol (2-propanol) to cyclohexane is to add polar and, especially, hydrogen bonding intermolecular forces to the blend. Cyclohexane has essentially none. In this w ay, the solv ency character of the azeotropic blend is v ery different from that of the neat cyclohexane. Equation 11.1 is useful for both solv ents and soils. It is also useful for azeotropes, near azeotropes, or nonazeotropic blends.

MORE REALISTIC VIEW ABOUT EVALUATING HSP OF COMPOSITE SOILS Judgment is also a factor required for evaluation of the HSP of composite soils. In some cases, the volumetric-proportional HSP v alues (disperse, polar , and h ydrogen bonding) of the soil may not represent the actual nature of the cleaning problem. Calculated v alues of HSP for a composite soil (as are shown in Figure 11.1) may not lead to the right choice of cleaning solv ents. This happens when a single component in the composite soil limits cleaning performance vs. the well-mixed composite being the limit of cleaning performance. 18 Unless the soil is a single component, one component of the soil composite will al ways be less soluble with the cleaning solvent in comparison with the composite of soil components as a whole. And other components of the composites will also ha ve dif ferent rates of solubility . The least compatible component may well be the one that should define the sol ency of the cleaning solvent. This is illustrated, with component number 3, in Figure 11.2. In other words, cleaning (where all soil components must be equally well removed) is a situation where the v olumetric average may not represent the true situation. 19

METHOD FOR CHOICE OF SUITABLE SOLVENTS RA is the distance in HSP space between the soil and the solv ent. That distance should be as small as possible (R A ≈ 0). 20,21 The equation to be used for solv ent selection, 22 which defines A, is:

7248_C011.fm Page 207 Thursday, May 10, 2007 8:00 AM

Polar HSP (soil or solvent)

Use of Hansen Solubility Parameters

207

#2 Soil Component #1

#3

HSP Parameters for Common Soil or Solvent

H2 Bonding HSP (soil or solvent) FIGURE 11.2

RA2 Solution = ( δ Polar for Solvent – δ Polar for Soil )

2

+ ( δ H 2 – Bonding for Solvent – δ H 2 – Bonding for Soil )

2

(11.3)

2 + ⎡ 4 × ( δ Disperse for Solvent – δ Disperse for Soil ) ⎤ ⎣ ⎦

The preceding equation comprises the follo wing data: • • •

The basic data for Equation 11.3 is the three pre viously estimated HSP values for either a soil composite (via Equation 11.1) or HSP v alues for a component e xpected to be limiting. The independent variable in the equation is the choice of solv ent. The dependent variable in the equation shows how well the soil is dissolved, that is, how closely R A approaches zero.

Polar HSP (soil or solvent)

The individual values for HSPs for the soil are either those computed from Equation 11.1 or that for the single soil component judged to limit cleaning performance as sho wn in Figure 11.2. HSP data and a spreadsheet can aid in making this comple x choice because they allow evaluation of solution performance for man y solvents (or blends) ag ainst all components of the soil. Graphically, the matching of solvent (or blend of solvents) to soil (of one of more components) is described in Reference 3 as Figure 4.6 (reproduced here as Figure 11.3 23). If the distance between the loci

Soil or Solvent #1 HSP Parameters for Common Soil or Solvent Soil or Solvent #2

H2 Bonding HSP (soil or solvent) FIGURE 11.3

7248_C011.fm Page 208 Thursday, May 10, 2007 8:00 AM

208

Hansen Solubility Parameters: A User’s Handbook

of HSP values for solvent and soil is too great (RA is too large), there will be little graphical overlap between the tw o spheres, and the proposed solv ent cleaning process w ould be e xpected to be an unsatisfactory choice. Where there is HSP data for soil components and solv ents, an optimum choice of solv ents (or solvent blends) can be made (see section on “Identification of ‘Designer Solv ents.”). No parts need be found, wet, cleaned, weighed, or e xamined.

REFERENCE SOILS FOR COMPARISON To e valuate the suitability and limitations of designer solv ents and their azeotropes in solv ent cleaning operations, one must be able to position them relati ve to common soils. This situation is common to that found in mark eting. The question is: Against which soils should these solv ents (and blends) be positioned (e valuated)? The approach taken here is to choose six different single-component organic soils for reference. They are identified and described in Table 11.2. Each represents a soil type frequently encountered in industrial operations, and their collective HSP values cover a broad range of HSP space. The HSP values of each soil are plotted as in Figure 11.4. These reference soils serve both a specific and a general purpose. Specific azeotropes of design solvents will be proposed to clean each soil. But in general, the method of de veloping successful solvent cleaning application is illustrated by matching solv ents to soil components.

IDENTIFICATION OF DESIGNER SOLVENTS Choice of the cleaning solvents of the designer type is somewhat arbitrary. Four families of solvents are considered: Hydrofluoroethers (HFEs): HFE-7100, HFE-7200, and HFE-8200 24 are molecules uniquely containing both fluorine and oxygen, as well as three or f e hydrogen atoms. Hydrocarbon ether solvents partially based on silicon (not on carbon): OS-10, OS-20, and OS-30 are molecules also containing ether linkages, b ut the y are based on a silicon to oxygen bond rather than a carbon to oxygen bond. The h ydrocarbon meth yl group is plentiful. Solvents more useful as refrigerants which form azeotropes: These are HFC-245f a (based on partially fluorinated ethane), HFC-365mfc (based on partially fluorinated propane), a (HFC-4310mee), based on partially fluorinated propane, which does find applications a nonazeotropic cleaning solvent. In addition, the fully fluorinated sol ent based on butane (PFC-506025) is also included for comparison. Solvents with harmful environmental impacts but useable in de veloping countries: HCFC141b, HCFC-225 ca/cb, and CFC-113. All three ha ve ozone depletion potentials (ODP) of significance, which has led to a ban on their production in industrialized countries pe the Montreal Protocol. Ho wever, their production and use in unde veloped countries is permitted for se veral years be yond the publication date for this book. Their structural formulas, physical, and solubility properties are gi ven in Table 11.3. The two-dimensional graphs26 of the latter are illustrated in Figure 11.5. Numerically, they total a “baker’s dozen.”

AN OPEN QUESTION — ANSWERED It is interesting to note the o verlap of Figure 11.4 and Figure 11.5. This overlap would represent the similarity (see Endnote 8) of these designer solv ents for use in cleaning typical industrial soils. There is observ ed to be little o verlap.

7248_C011.fm Page 209 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

209

Table 11.2 Reference Soil Materials Soil

#

ASTM Fuel “A” 1

Butyl Stearate

27

Castor Oil

Ethyl Cinnamate

2

3

4

5

δ Dispersion δ Polar δ Hydrogen-Bonding Molecular Model Image MPa½ MPa½ MPa½

14.3

12.6

13.6

16.0

13.5

0

6.3

6.0

10.8

3.5

0

6

15.9

13.9

Industrial Use

CH3

Isooctane (2,2,4-trimethylpentane, or isobutyltrimethylpentane) is a component of gasoline (Octane Rating 100) and representative of the general class of hydrocarbon chemicals.

6.1

CH3, Ester

10.5

CH3, Acid Double Bond Hydroxyl

7.5

Aromatic, Ester, Double Bond

3.7

CH3, Esters (3), Double Bonds (6)

Linseed Oil

Tricresyl Phosphate

Contained Functional Groups

13.5

Phosphate, Aromatic

Representative of a class of soils as fatty acid esters, which are used as used as raw material of emulsifiers or oiling agents for foods; spin finishes and textiles; lubricants for plastics, paint and ink additives; surfactants and base materials for perfumery; solvents or cosolvents; and oil carriers in agricultural operations. Castor oil is a vegetable oil. It and its derivatives have applications in the manufacturing of soaps, lubricants, hydraulic and brake fluids, paints, dyes, coatings, inks, plastics, waxes and polishes, pharmaceuticals and perfumes. (Ethyl Phenyl Acrylate) is a natural chemical and a common fragrance and flavoring component. It is the ester primarily responsible for the smell of cinnamon.

(Triglyceride Ester of mainly Linoleic Acid28), a natural vegetable oil, is the most important drying oil of the oil painting industry.

This material is found in common metalworking fluids: anti-wear agents or lubricants. It is also used as a plasticizer (PVC and alkyd resins), as a detergent, and as flame retardant.

Other than cleaning h ydrocarbon-based soils that do not ha ve substantial polarity or h ydrogen bonding character, these designer solv ents are no more functionally useful than a simple solv ent such as heptane (which costs perhaps one twentieth as much per kg). Questions asked during the last decade by those who practice cleaning science are:What should these designer solv ents be used for?, wh y should manuf acturers char ge e xorbitant prices?, and “how can we g ain the value of the United States v olatile organic compound (USVOC) exemption and negligible ODP promised by their manuf acturers, when these designer solvents appear useless to remove typical industrial soils? ”

7248_C011.fm Page 210 Thursday, May 10, 2007 8:00 AM

210

Hansen Solubility Parameters: A User’s Handbook

HSP OF REPRESENTATIVE SOILS

HSP OF SINGLE COMPONENT SOLVENTS

Chosen to Cover a Broad Range of Value

13 ‘Designer’ Solvents 14 Hydrogen Bonding HSP, MPaˆ (½)

Hydrogen Bonding HSP, MPaˆ (½)

14 12 10 8 6 4 2 0

12 10 8 6 4 2 0

0

2

FIGURE 11.4

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

0

2

4 6 8 Polar HSP, MPaˆ (½)

10

12

14

FIGURE 11.5

The approach to answer these questions without commercial bias will be tw o fold: 1. To extend the power of designer solvents by using them not in neat form (as a comparison of Figure 11.4 and Figure 11.5 sho ws that approach to be fruitless) b ut in combination with secondary components in binary azeotropes identified by the manuacturers of these solvents. In other words, the type of cleaning29 solvent evaluated in this study is a binary azeotrope with one component being a designer solv ent and the other being a more common (and lo w-priced) chemical. Naturally , this formulation of solv ent blends (reduces) purchase price and blends (dilutes) en vironmental benefits 2. To use e xisting solvent data of ph ysical and chemical properties as well as HSP v alues to match one of the six common industrial soils to a binary azeotrope one of whose components as a designer solv ent. In all cases, the cleaning process contemplated is the con ventional vapor degreasing.

LIMITING RA VALUE FOR EXPECTED GOOD CLEANING PERFORMANCE A good cleaning performance that is attained by limiting R • • •

A

requires the follo wing:

A method for calculation of HSP values of blends that are identified (See Equation 11.1) Target soils to be identified (See Figure 11.4 for HSP alues); and A library of azeotropic formulations identified (See Reference 16), including their blen HSP v alues calculated through literature component v alues from Reference 15 and Equation 11.1. The questions remain as to what standard of cleaning performance speaks to how well the soil is dissolv ed and how closely does R A approach zero?

The answer is in a study also reported in Reference 16. It is that R A is related to the amount of soil not remo ved (uncleaned) from the parts being cleaned. Cleaning trials30 have been and are being conducted by the Surface Cleaning Laboratory within the Toxic Use Reduction Institute (TURI), located at the Uni versity of Massachusetts–Lowell. The work is aimed at pro viding practical, useful, and en vironmentally sound solutions to b usinesses within the New England re gion of the U.S.

7248_C011.fm Page 211 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

211

Table 11.3 Properties of “Designer” Solvents Solvent

Structure

Specific Mol. Gravity, Boiling δ Dispersion δ Polar δ Hydrogen-Bonding δ Overall Wt. g/cc Point, °C Values in Mp½

Cleaning Solvents Designed for Minimal Environmental Impact

HFE-7100

250.1 1.520

60.0

13.7

2.2

1.0

13.9

HFE-7200

264.1 1.430

76.1

13.3

2.0

1.0

13.5

HFE-8200

250.1 1.43

76

13.0

4.0

1.0

13.6

HFC-43-10 mee

252.1 1.580

54.4

11.6

0.0

0.0

11.6

OS-10 (OS-10 Hexamethyldisiloxane)

162.3 0.764

100.6

12.4

0.0

0.0

12.4

OS-20 (OS-20 Octamethyldisiloxane)

236.5 0.820

152.8

11.7

0.0

0.0

11.7

OS-30 (Decamethyltrisiloxane)

310.7 0.854

193.9

12.2

0.0

0.0

12.2

Solvents More Useful as Refrigerants, But Which Form Azeotropes

HFC-245 fa

134.0 1.320

15.3

15.7

0.0

0.0

15.7

HFC-365 mfc

148.1 1.270

40.2

16.4

0.0

0.0

16.4

PFC-5060

288.0 1.680

55

12.1

0.0

0.0

12.1

Solvents with Harmful Environmental Impacts But Usable in Developing Countries

HCFC-141B

116.9 1.230

32.2

15.7

4.0

1.0

16.2

HCFC-225 ca / cb

202.9 1.550

53.9

14.1

3.2

1.0

14.5

CFC-113

187.4 1.560

47.8

14.7

1.6

0.0

14.8

7248_C011.fm Page 212 Thursday, May 10, 2007 8:00 AM

212

Hansen Solubility Parameters: A User’s Handbook

A group of more than 50 tests with various solvents (and blends) in vapor degreasing operations were examined by this author .16 It w as found that the “bright line ” between 31 acceptable and not acceptable cleaning performance is represented by an R A value of about 8. This means that: If the R A calculated for soil components and a single solv ent or azeotrope is less than 8, the solvent cleaning operation has a good chance to be successful. If the R A calculated exceeds 8, the solv ent cleaning operation does not ha ve a good chance to be successful. Another solvent or blend, or a process dif ferent from v apor degreasing should be considered. The limiting value of 8 will be used for R A in this study .

APPLICATION OF HSP METHODOLOGY TO CLEANING OPERATIONS Figure 11.6 and Figure 11.7, in which the cleanliness performance data are e xhibited, should be considered as the tw o-dimensional spherical solubility plots in Reference 3. Data in Figure 11.7 are similar to that in Figure 11.6. Only the ranges ha ve been changed. These solution cleaning results are consistent with and similar to the teachings of Reference 3, which are about solution of polymers or other materials within solv ents (or the re verse). The results are enlisted as sho wn in the follo wing: 1. Quality of cleaning is related to the distance in HSP space (R A) between solvent and the soil least compatible with the solv ent. This means less soil is left on parts in actual cleaning tests when R A is smallest. The basis for Figure 11.6 and Figure 11.7 mak es 100% ▲ ▲

% Soiled

80%

▲

60% 40%

▲▲ ▲ ▲ ▲ ▲

▲

▲

▲ ▲ ▲ ▲

20% ▲ ▲▲ ▲ ▲ ▲ ▲▲ ▲ ▲ ▲▲ ▲ ▲ ▲ ▲▲

▲▲▲▲ ▲

0% 0

5

10

▲

15 MAX RA

20

25

30

% Soiled

FIGURE 11.6 10% 9% 8% 7% 6% 5% 4% 3% 2% 1% 0%

▲ ▲

▲ ▲

▲

▲

▲

▲ ▲ ▲

▲ ▲ ▲ ▲

▲▲ ▲ ▲ ▲ ▲ ▲▲ ▲

▲

▲ ▲ ▲ ▲

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 MAX RA

FIGURE 11.7

7248_C011.fm Page 213 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

213

physical sense; the material not cleaned (percentage soiled) by a solv ent is related to the soil least soluble. 2. There is a limit (breakpoint, “bright line,” demarcation, or separation) between one type of behavior (acceptable cleanliness) and the opposite (unacceptable cleanliness). There are v ery fe w or ne gligible anomalies (situations where HSP v alues suggest that there should be poor compatibility , and there is good compatibility , as well as the re verse). There are no results of poor cleaning when R A is 8 or belo w. 3. The value of R A (8), chosen to dif ferentiate acceptable and unacceptable cleanliness, is similar to the interaction radius between polymers and solv ents typical for selected correlations given in Appendix Table A.2 of Reference 3 and Appendix Table A.2 of this Handbook. The average value in Appendix Table A.2 is about 10. Results of calculation of the HSP values for each binary32 azeotrope blend based on the solvents in Table 11.3 and in Table 11.4 33 through Table 11.15 are gi ven and illustrated in Figure 11.8 through Figure 11.20.

ANALYSIS OF CAPABILITY OF DESIGNER SOLVENTS The limited utility 34 of neat designer solv ents has been e xperienced in the decade since the y were commercialized and is characterized via HSP methodology in Figure 11.5. As viewed within the HSP space, the v alue of the components from which these azeotropes are formed is to add h ydrogen bonding and polar intermolecular forces to designer solv ents that generally do not display those forces. This is illustrated schematically in Figure 11.21, and is essentially the approach described in the section on “An Open Question — Answered,” to find verlap between Figure 11.4 and Figure 11.5. Each azeotropic blend noted abo ve35 is a fluid with a single boiling point, ut with dif ferent physical properties, safety , health, en vironmental impacts, and economic consequences than are manifested by the 13 chosen designer solv ents. The proper metric 36 for discrimination among these azeotropes is the ef fective completion of the proposed cleaning work. That metric is the magnitude of the mismatch between intermolecular forces of the solvent (single or azeotropic blend) and composite soil. Quantization of this mismatch is produced for specific soils by the parameter A. As described in the section on limiting R A value for good cleaning performance (at least 95% remo val of all soils) is around 8. The capability to clean v arious soils with azeotropes of designer solv ents is gi ven in Table 11.16. The information is sorted to meet the needs of users by gi ving the various designer solvents that can be blended to clean the stated soil. Users can use HSP to dif ferentiate one solv ent or solv ent blend from another based on the values of the soil found in the specific application. Information in the preceding tables can be use to screen azeotropes formulated from designer solvents and reduce the needed burden of feasibility testing. Reference 3 or Reference 15, as well as the appendices in this book, can be used as sources of HSP v alues when the six soils chosen to be representati ve in this study are not representati ve of those in the current application. This analysis, using HSP methodology, clearly shows the superiority of the fluoroether structur (HFE) as a feedstock for azeotropes. F our to fi e units (MP a1/2) of polar HSP v alue can be added through addition of secondary solv ents to form constant-boiling 37 azeotropes. HSP methodology further allo ws recognition of the limits of azeotropic blends based on designer solvents. No composition is e xpected to be capable of cleaning soil number 6 (tricresyl phosphate). This is because the amount of polar and hydrogen bonding intermolecular forces needed to produce a compatible solution cannot be obtained by adding a second solvent rich in those forces to designer solv ents that are nearly de void of them.

7248_C011.fm Page 214 Thursday, May 10, 2007 8:00 AM

214

Hansen Solubility Parameters: A User’s Handbook

TABLE 11.4 Azeotropes with HFE-7100 Primary Azeo Component HFE-7100

Type of Solvent

Wt (%) HFE-7100 100%

BP (ºC) 60.0

SpG (g/cc) 1.520

Isobutyl alcohol

Azeotrope

99.0%

58.0

1.507

2-Butanol

Azeotrope

98.4%

58.0

1.499

1-Propanol (nPA)

Azeotrope

97.9%

56.0

1.492

t-Butyl alcohol

Azeotrope

94.0%

56.0

1.438

1,2-Dichloropropane

Azeotrope

94.8%

59.0

1.496

2-Propanol (IPA) Azeotrope 1-Chlorobutane Azeotrope

95.1% 87.1%

54.0 57.0

1.453 1.390

HCFC-225 ca/cb

28.5%

53.0

1.541

Ethanol Azeotrope

93.0%

52.0

1.428

Methanol Azeotrope

93.0%

46.0

1.427

Azeotrope

65.7%

55.0

1.430

Azeotrope Azeotrope Azeotrope

25.8% 64.0% 44.1%

54.0 31.2 40.0

1.376 1.232 1.361

Azeotrope Azeotrope Azeotrope

39.1% 31.4% 22.0%

52.6 50.2 35.0

1.100 1.103 0.950

1,2-Dichloroethylene (CIS) n-Propyl bromide Ethyl formate 1,2-D ichloroethylene (TRANS) Methyl acetate Methyl formate 2-Chloropropane Near

Azeotrope

HSP Reference WO 96/36689, U.S.P. 6,426,327, U.S.P. 6,008,179 WO 96/36689, U.S.P. 6,426,327, U.S.P. 6,008,179 WO 96/36689, U.S.P. 6,426,327, U.S.P. 6,008,179 WO 96/36689, U.S.P. 6,426,327 WO 96/36689, U.S.P. 6,426,327 U.S.P. 6,426,327 WO 96/36689, U.S.P. 6,426,327 WO 96/36689, U.S.P. 6,426,327 WO 96/36689, U.S.P. 6,426,327 WO 96/36689, U.S.P. 6,426,327 WO 96/36689, U.S.P.6,426,327 U.S.P. 6,689,734 U.S.P. 6,753,304 U.S.P. 6,426,327 U.S.P. 6,753,304 U.S.P. 6,753,304 U.S.P. 6,646,020, U.S.P. 6,426,327

δD 13.7

δP 2.2

δH 1.0

Soils Cleaned RA < 7.93

13.7

2.3

1.3

1, 2, 5

13.8

2.3

1.4

1, 2, 5

13.8

2.4

1.6

1, 2, 5

13.9

2.5

2.5

1, 2, 5

13.9

2.5

1.1

1, 2, 5

13.9 14.2

2.6 2.9

2.4 1.2

1, 2, 5 1, 2, 5

14.0

2.9

1.0

1, 2, 5

14.0

3.0

3.3

1, 2, 3, 5

13.9

3.5

3.7

1, 2, 3, 5

15.0

4.4

1.8

1, 2, 5

15.5 14.6 15.7

5.0 5.2 5.7

3.5 4.6 2.3

1, 2, 4, 5 1, 2, 3, 4, 5 1, 2, 4, 5

15.0 14.9 15.3

5.8 7.0 7.5

5.7 8.1 1.9

2, 3, 4, 5 2, 3, 4, 5 2, 4, 5

Figure 11.8 through Figure 11.20 are cluttered, rich in information, and capable of ef communication. At a single glance, one can estimate:

fective

In Figure 11.8, that no azeotrope of HFE-7100 can clean soils based on castor oil or tricresyl phosphate. In Figure 11.18, that HCFC-141b is a lik ely candidate to be a feedstock from which one or more azeotropes can be formulated to clean soils based on castor oil. In Figure 11.15 and Figure 11.16, that HFC-245f a and HFC-365mfc, respecti vely, are not likely candidates from which azeotropes can be formulated to clean an y but single hydrocarbon soils. In Figure 11.12, that OS-10 is a v ery useful feedstock from which one or more azeotropes can be formulated to clean a v ariety of soils.

7248_C011.fm Page 215 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

215

TABLE 11.5 Azeotropes with HFE-7200 Primary Azeo Component HFE-7200

Type Solvent

wt % HFE-7200 100%

BP, °C 76.1

SpG, g/cc 1.430

Ethanol

Azeotrope

88.0%

62.0

1.417

1,2-Dichloropropane

Azeotrope

87.0%

73.0

1.388

1-Chlorobutane

Near Azeo

87.1%

69.0

1.293

1-Butyl Alcohol

Azeotrope

84.0%

67.0

1.262

n-Propyl Bromide

Azeotrope

56.0%

63.0

1.385

Methanol

Azeotrope

84.0%

53.0

1.266

1,2-Dichloroethylene (TRANS)

Azeotrope

68.0%

48.0

1.307

HSP Reference WO 96/36688 & USP 6,288,018 WO 96/36688 & USP 6,288,018 WO 96/36688 & USP 6,288,018 WO 96/36688 & USP 6,288,018 WO 96/36688 & USP 6,288,018 WO 96/36688 & USP 6,288,018 USP 6,699,829

δD 13.3

δP 2.0

δH 1.0

Soils Cleaned RA < 7.93

13.8

2.8

1.6

1, 2, 5

13.9

2.8

1.3

1, 2,

13.8

2.8

4.0

1, 2, 3

13.8

2.8

4.5

1,. 2, 3

14.5

3.7

2.5

1, 2,

13.8

4.6

6.5

2, 3, 4

15.9

6.2

2.6

1, 2, 4

It w ould be dif ficult to imagine h w another technical approach could allo w ef ficient iden tification of sol ent capability for industrial cleaning from a table of azeotropes in a handbook.

CONCLUSIONS The developments described in this chapter are original and support the follo wing conclusions: 1. Soils are chemicals. They can be described by the component chemicals of which the y are comprised. As chemicals, their intermolecular forces can be characterized by HSPs. And as a mixture of chemical components, their o verall HSP v alue can be computed through a con ventional volumetric blend rule. This outcome allo ws soils to be characterized in a quantitati ve manner . In this chapter , six soil materials representati ve of common industrial soils were so characterized. 2. The same approach can be follo wed for a solv ent and multiple component solv ent blends with the same outcome — that is, characterization in a quantitative manner. In this chapter, thirteen “designer” solvents were so characterized. These solvents are notable for being similar to simple h ydrocarbons with lo w-polar and h ydrogen bonding forces and being considered to foster limited concerns about en vironmental, safety, or health hazards. 3. In a pre vious publication (Reference 16), it has been sho wn that the ef fectiveness of solvent cleaning systems (v apor degreasers) can be predicted by the similarity of inter molecular forces between soil composites and solv ent blends. Similarity means the distance in HSP space between the soil and the solv ent materials. The quantitative term is R A, and good cleaning is observ ed when R A is about 8 or less. 4. In the same pre vious publication (Reference 16), tw o-component azeotropes of these designer solvents were identified. 5. In this chapter, the blend HSP v alues for solvent azeotropes were compared to those of soil composites with the standard for comparison being that the R A value between them be less than 8. These comparisons were tab ulated and plotted.

7248_C011.fm Page 216 Thursday, May 10, 2007 8:00 AM

216

Hansen Solubility Parameters: A User’s Handbook

TABLE 11.6 Azeotropes with HFE-8200 Primary Azeo Component HFE-8200

Type of Solvent

Wt (%) HFE-8200 100%

BP (ºC) 76.0

SpG (g/cc) 1.430

HCFC 225 ca/cb

Azeotrope

30.6%

53.0

1.511

Isobutyl alcohol

Azeotrope

99.0%

58.0

1.419

98.8%

58.0

1.417

2-Butanol Azeotrope t-Butyl alcohol

Azeotrope

94.0%

56.0

1.362

1-Propanol (nPA)

Azeotrope

97.4%

56.0

1.402

1,2-Dichloropropane Azeotrope

94.5%

59.0

1.412

2-Propanol (IPA)

93.3%

55.0

1.356

1-Chlorobutane Azeotrope

87.8%

57.0

1.329

Ethanol Azeotrope

95.2%

52.0

1.376

Azeotrope

Ethyl formate

Azeotrope

79.9%

31.3

1.287

n-Propyl bromide

Azeotrope

25.8%

54.0

1.356

89.6%

46.0

1.319

Azeotrope

65.7%

55.0

1.376

Azeotrope

44.3%

52.7

1.104

Azeotrope

50.0% 40.0

Azeotrope

40.1%

50.4

1.122

22.0%

35.0

0.942

Methanol Azeotrope 1,2-Dichloroethylene (CIS) Methyl acetate

1,2-Dichloroethylene (TRANS) Methyl formate

2-Chloropropane Near

Azeotrope

1.338

HSP Reference U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,753,304, U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,753,304, U.S.P. 6,426,327 U.S.P. 6,426,327 U.S.P. 6,753,304, U.S.P. 6,426,327 U.S.P. 6,646,020, U.S.P. 6,426,327

δD 13.0

δP 4.0

δH 1.0

Soils Cleaned RA < 7.93

13.7

3.5

1.0

1, 2, 5

13.0

4.0

1.3

1, 2, 5

13.1

4.0

1.3

1, 2, 5

13.2

4.1

2.4

1, 2, 5

13.1

4.1

1.7

1, 2, 5

13.3

4.2

1.1

1, 2, 5

13.3

4.2

2.8

1, 2, 3, 5

13.6

4.3

1.2

1, 2, 5

13.2

4.4

2.5

1, 2, 5

13.7

5.2

3.1

1, 2, 3, 5

15.3

5.4

3.4

1, 2, 3, 4, 5

13.4

5.4

4.7

1, 2, 3, 5

14.5

5.5

1.8

1, 2, 5

14.6

6.1

5.3

2, 3, 4, 5

15.1

6.1

2.2

1, 2, 4, 5

14.6

7.0

7.3

2, 3, 4, 5

15.1

7.8

1.9

2, 4, 5

7248_C011.fm Page 217 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

217

TABLE 11.7 Azeotropes with HFC-4310mee Primary Azeo Component HFC-4310mee

Type of Solvent

Wt (%) HFC-4310mee 100%

BP (ºC) 54.4

SpG (g/cc) 1.580

Azeotrope Azeotrope Azeotrope Azeotrope

97.4% 97.1% 95.3% 23.0%

45.5 43.4 39.9 52.0

1.540 1.535 1.509 1.515

1,1-Dichloroethane Azeotrope

73.0%

43.0

1.445

Acetone Azeotrope

83.9%

57.6

1.361

Azeotrope 67.9%

42.3

1.471

Azeotrope 61.7%

39.0

1.443

2-Propanol (IPA) Ethanol Methanol n-Propyl bromide

1,2-Dichloroethylene (CIS) 1,2-Dichloroethylene (TRANS)

HSP Reference Vertrel XP Vertrel XE Vertrel XM U.S.P. 6,689,734 U.S.P. 5196,137 U.S.P. 5824,634 U.S.P. 5196,137 Vertrel MCA

δD 11.6

δP 0.0

δH 0.0

Soils Cleaned RA < 7.93

11.8 11.8 11.9 12.8

0.3 0.5 1.1 1.5

0.8 1.1 2.0 1.1

1, 5 1, 2, 5 1, 2, 5 1, 2, 5

13.2

2.6

1.0

1, 2, 5

12.7

2.9

1.9

1, 2, 5

13.6

2.9

1.2

1, 2, 5

14.0

3.5

1.4

1, 2, 5

TABLE 11.8 Azeotropes with OS-10 Primary Azeo Component OS-10 Hexamethyld isiloxane

Type of Solvent

Wt (%) OS-10 Hexamethyld isiloxane 100%

BP (ºC) 100.6

SpG (g/cc) 0.764

HSP

Sec-butyl acetate

Azeotrope

96.0%

100.5

0.768

2-Pentanol

Azeotrope

87.0%

97.8

0.770

Propylene glycol methyl ether n-Propyl acetate

Azeotrope

82.0%

95.7

0.788

Azeotrope

61.0%

96.7

0.808

2-Propanol (IPA)

Azeotrope

54.3%

76.4

0.774

Ethanol

Azeotrope

63.8%

71.4

0.773

Methanol

Azeotrope

59.1%

58.7

0.775

Reference U.S.P. 5834,416 U.S.P. 5478,493 U.S.P. 5478,493 U.S.P. 5834,416 U.S.P. 5773,403 U.S.P. 5773,403 U.S.P. 5773,403

δD 12.4

δP 0.0

δH 0.0

Soils Cleaned RA < 7.93

12.5

0.1

0.3

1, 5

12.8

0.8

1.6

1, 2, 5

12.9

1.0

2.4

1, 2, 5

13.4

1.5

2.7

1, 2, 5

13.9

2.7

7.4

1, 2, 3, 5

13.6

3.1

6.9

1, 2, 3, 5

13.5

4.9

8.9

2, 3, 4, 5

6. The comparison showed that expected cleaning performance of designer solvents can be enhanced by combining it with other solv ents into binary azeotropes. That enhancement is a g ain of about four or fi e units (MP a1/2) of total (Hildebrand) solubility parameter , which allows prediction of ef fective cleaning of se veral common industrial soils. Ho wever, soils that display high levels of polar and hydrogen bonding forces, such as tricresyl phosphate, are not e xpected to be remo ved in a solv ent cleaning process emplo ying azeotropes of these designer solv ents.

7248_C011.fm Page 218 Thursday, May 10, 2007 8:00 AM

218

Hansen Solubility Parameters: A User’s Handbook

TABLE 11.9 Azeotropes with OS-20 Primary Azeo Component OS-20 Octamethyltrisiloxane

Type of Solvent

Wt (%) OS-20 Octamethyltrisiloxane 100%

HSP BP (ºC) 152.8

SpG (g/cc) 0.820

Propylene glycol n-butyl ether

Azeotrope 89.0%

151.8

0.826

Propylene glycol n-propyl ether

Azeotrope 60.0%

141.8

0.844

Ethyl lactate

Azeotrope

139.4

0.888

63.0%

Reference U.S.P. 5454,970, 5628,833 U.S.P. 5516, U.S.P. 5628,833 U.S.P. 5454, U.S.P. 5628,833

δD 11.7

δP 0.0

δH 0.0

Soils Cleaned RA < 7.93

12.1

0.4

1.3

1, 2, 5

13.2

1.9

5.1

1, 2, 3, 5

13.1

2.4

4.0

1, 2, 3, 5

TABLE 11.10 Azeotropes with OS-30 Primary Azeo Component OS-30 Decamethyltrisiloxane Dipropylene glycol n-propyl ether Dipropylene glycol methyl ether acetate Dipropylene glycol methyl ether (DPM, DPGME) Propylene glycol monobutyl ether (PGBE, PnB)

Type of Solvent

Wt (%) OS-30 DecamethylBP SpG trisiloxane (ºC) (g/cc) 100% 193.9 0.854

Azeotrope 91.0%

186.7

Azeotrope 89.0%

193.8

Azeotrope 61.0%

170.4

Azeotrope 15.0%

170.6

HSP Soils Cleaned RA < 7.93

δD 12.2

δP 0.0

δH 0.0

0.859 U.S.P. 5824,632 0.866 U.S.P. 5824,632 0.889 U.S.P. 5824,632

12.4

0.4

1.0 1, 2, 5

12.5

0.5

0.8 1, 2, 5

13.4

2.1

4.1 1, 2, 3, 5

0.874 U.S.P. “5824,632

14.8

3.8

7.8 2, 3, 4, 5

Reference

The methodology used in this chapter can be easily emplo yed with other soil composites and solvent blends. This is a crucial point because additional solv ents other than the designer solv ents can be used as a feedstock to formulate azeotropes. Such additional solv ents will not be as limited with re gards to polar and h ydrogen bonding intermolecular forces as are the designer solv ents chosen for this study . It appears possible 16 to identify an azeotropic solv ent pair that is capable of being expected to clean an y proposed soil composite.

7248_C011.fm Page 219 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

219

TABLE 11.11 Azeotropes with HFC-245fa Primary Azeo Component HFC-245fa

Type of Solvent

Wt (%) HFC-245fa 100%

BP (ºC) 15.3

SpG (g/cc) 1.320

14.8

1.322

Trichloroethylene (TCE) HCFC-123 Near

Near-Azeotrope 98.1% -Azeotrope

98.0%

15.0

1.323

Methylene chloride 2-Propanol (IPA)

Near-Azeotrope

98.3%

15.2

1.320

Near-Azeotrope

96.8%

14.5

1.291

1-Propanol (nPA)

N ear Azeo

96.8%

14.4

1.293

Methanol Near

-Azeotrope

97.0%

14.2

1.294

Ethanol Near

-Azeotrope

95.8%

14.4

1.283

1,2Near-Azeotrope Dichloroethylene (TRANS) Water Near -Azeotrope

37.5%

—

1.295

83.0%

7.0

1.251

Wt (%) HFC-365mfc

BP, (ºC)

SpG, (g/cc)

Ethanol Azeotrope

98.4%

39.2

1.258

Water Azeotrope

98.0%

38.0

1.263

Methanol Azeotrope

96.2%

—

1.241

HCF2OCF2OCF2H

60.1%

32.6

1.450

46.0%

33.0

1.009

HSP Reference U.S.P. 6,100,229 U.S.P. 6,362,153 U.S.P. 6,100,229 U.S.P. 5683,974 U.S.P. 5683,974 U.S.P. 5683,974 U.S.P. 5683,974 U.S.P. 5851,977 U.S.P. 6,514,928

δD 15.7

δP 0.0

δH 0.0

Soils Cleaned RA < 7.93

15.7

0.1

0.1

1, 5

15.7

0.1

0.0

1, 5

15.7

0.1

0.1

1, 5

15.7

0.3

0.9

1, 5

15.7

0.4

0.9

1, 5

15.7

0.6

1.1

1, 5

15.7

0.6

1.3

1, 5

16.2

3.1

1.2

1, 5

15.7

3.4

0.9

1, 5

TABLE 11.12 Azeotropes with HFC-365mfc Primary Azeo Component HFC-365mfc

HSP Type of Solvent

NearAzeotrope 2-Chloropropane Azeotrope

Reference U.S.P. 5445757 U.S.P. 6,365566 U.S.P. 6,743,765 U.S.P. 6,255273 U.S.P. 6,646,020

Soils Cleaned RA < 7.93

δD 16.4

δP 0.0

δH 0.0

16.4

0.2

0.5

16.4

0.4

0. 1,

16.3

0.7

1.3

1, 5

13.0

3.8

2.5

1, 2, 5

16.1

5.3

1.3

1, 5

1, 5 5

7248_C011.fm Page 220 Thursday, May 10, 2007 8:00 AM

220

Hansen Solubility Parameters: A User’s Handbook

TABLE 11.13 Azeotropes with PFC-5060 Primary Azeo Component Perfluorohexane (PFC 5060) HCFC-141b HCFC-123

HSP

Type of Solvent

wt % Perfluorohexane (PFC 5060) 100%

BP, °C 56.0

SpG, g/cc 1.680

Azeotrope Azeotrope

50.0% 50.0%

26.0 26.5

1.420 1.561

Reference

δD 12.1

δP 0.0

δH 0.0

Soils Cleaned RA < 7.93

USP 5560,861 USP 5560,861

14.2 13.7

2.3 2.9

0.6 0.5

1, 2, 5 1, 2, 5

TABLE 11.14A Azeotropes with HCFC-141b HSP Primary Azeo Component HCFC-141b

Type of Solvent

wt % HCFC-141b 100%

BP, °C 32.1

SpG, g/cc 1.230

Reference

δD 15.7

δP 4.0

δH 1.0

Soils Cleaned RA < 7.93

Azeotrope

50.0%

20.0

1.414

USP 5494,601

13.8

2.3

0.6

1, 2, 5

Azeotrope

50.0%

26.0

1.420

USP 5560,861

14.2

2.3

0.6

1, 2, 5

Perfluoropentane (PFC 5050) Perfluoroh xane (PFC 5050) Cyclopentane 1,2-Dichloroethylene (TRANS) 2-Chloropropane Ethanol Methanol

Near Azeo Near Azeo

98.5% 98.1%

32.2 32.2

1.218 1.230

USP 5085798 USP 5126,067

15.7 15.7

3.9 4.1

1.0 1.0

1, 5 1, 5

Near Azeo Near Azeo Blend

98.5% 98.0% 95.6%

32.2 31.9 30.0

1.222 1.216 1.201

15.7 15.7 15.7

4.1 4.1 4.6

1.0 1.6 2.4

1, 5 1, 5 1, 2, 5

HCFC-123 Methylene Chloride

Near Azeo Blend

50.0% 70.0%

31.5 31.0

1.334 1.257

15.4 16.4

4.6 4.7

1.0 2.5

1, 2, 5 1, 4, 5

2-Propanol (IPA)

Near Azeo

50.0%

31.5

0.959

USP 5085797 USP 4,836,947 PROMOSOL 141b MS USP 5194,169 PROMOSOL 141b MC USP 5318,716

15.8

5.3

10.4

2, 3, 4, 5

TABLE 11.14B Azeotropes with HCFC-225 ca/cb Wt (%) HCFC-225 ca/cb 100%

BP (ºC) 54.0

SpG (g/cc) 1.550

HFE-7100 Azeotrope

28.5%

53.0

1.541

HFE-8200 Azeotrope

30.6%

53.0

1.511

Primary Azeo Component HCFC-225 ca/cb

Type of Solvent

HSP

Reference WO 96/36689, U.S.P. 6,426,327 U.S.P. 6,426,327

δD 14.1

δP 3.2

δH 1.0

Soils Cleaned RA < 7.93

14.0

2.9

1.0

1, 2, 5

13.7

3.5

1.0

1, 2, 5

7248_C011.fm Page 221 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

221

TABLE 11.15 Azeotropes with CFC-113 Primary Azeo Component 1,1,2-Trichlorotrifluoroethane (CFC-113)

Wt (%) 1,1,2-Trichlorotrifluoroethane BP (CFC 113) (ºC) 100% 47.6

Type of Solvent

HSP SpG (g/cc) 1.560

Nitromethane Azeotrope

97.1%

46.8

1.544

Methylene chloride Azeotrope

85.8%

37.0

1.522

Acetone Azeotrope

87.5%

45.0

1.391

Reference U.S.P. 3,573,213 U.S.P. 2,999,817 U.S.P. 2,999,815

δD 14.7

0.2 1, 5

15.3 2.4

1.0 1, 5

14.9

1.5 1, 2, 5

17 Azeotropes with HFE-7100 Hydrogen Bonding HSP, MPaˆ (½)

TRICRESYL PHOS

12 CASTOR OIL

10

Methyl Formate ETHYL CINNAMATE

8 6 4

BUTYL STEARATE Methyl Acetate Ethyl Formate LINSEED OIL Ethanol

2

ASTM 0FUEL “A” 0 2

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

FIGURE 11.8

HSP OF BINARY AZEOTROPES

Hydrogen Bonding HSP, MPa^(1/2)

7 Azeotropes with HFE-7200 14

CASTOR OIL

10 8 6

ETHYL CINNAMATE

Methanol BUTYL STEARATE

t-Butyl Alcohol 1-Chlorobutane LINSEED OIL ?????????????? ?????????????? 2 Ethanol 1,2-Dichloropropane 4

ASTM 0 FUEL “A” 0 2

FIGURE 11.9

TRICRESYL PHOS

12

4

6 8 10 Polar HSP, MPa^(1/2)

12

14

Soils δH Cleaned 0.0 RA < 7.93

14.7 2.3

HSP OF BINARY AZEOTROPES 14

δP 1.6

3.5

7248_C011.fm Page 222 Thursday, May 10, 2007 8:00 AM

222

Hansen Solubility Parameters: A User’s Handbook

HSP OF BINARY AZEOTROPES 17 Azeotropes with HFE-8200 Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8 6 4

ETHYL CINNAMATE Methyl Formate BUTYL STEARATE Methyl Acetate Methanol LINSEED OIL

2

ASTM 0FUEL “A” 0 2

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

FIGURE 11.10

FIGURE 11.11 HSP OF BINARY AZEOTROPES 7 Azeotropes with OS-10 Hexamethyldisiloxane Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 10 8

CASTOR OIL 2-Propyl (PA) Ethanol ETHYL CINNAMATE BUTYL STEARATE

6 4

LINSEED OIL

2

ASTM 0FUEL “A” 0 2

FIGURE 11.12

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

7248_C011.fm Page 223 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

223

HSP OF BINARY AZEOTROPES 3 Azeotropes with OS-20 Octamethyltrisiloxane Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8 6 4

ETHYL CINNAMATE BUTYL STEARATE Propylene Glycol n-Propyl Ether LINSEED OIL

2

ASTM 0FUEL “A” 0 2

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

FIGURE 11.13 HSP OF BINARY AZEOTROPES 4 Azeotropes with OS-30 Decamethyltrisiloxane Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8 6 4

ETHYL CINNAMATE BUTYL STEARATE Propylene Glycol n-Propyl Ether LINSEED OIL

2

ASTM 0FUEL “A” 0 2

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

FIGURE 11.14 HSP OF BINARY AZEOTROPES 9 Azeotropes with HFC-245fa Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8

ETHYL CINNAMATE BUTYL STEARATE

6 4

LINSEED OIL

2

ASTM 0FUEL “A” 0 2

FIGURE 11.15

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

7248_C011.fm Page 224 Thursday, May 10, 2007 8:00 AM

224

Hansen Solubility Parameters: A User’s Handbook

HSP OF BINARY AZEOTROPES 5 Azeotropes with HFC-365mfc Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8

ETHYL CINNAMATE BUTYL STEARATE

6 4 2

LINSEED OIL Methanol

ASTM 0FUEL “A” 0 2

2-Chloropropane 4 6 8 10 Polar HSP, MPaˆ (½)

12

14

FIGURE 11.16

HSP OF BINARY AZEOTROPES 2 Azeotropes with Perfluorohexane (PFC) Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8

ETHYL CINNAMATE BUTYL STEARATE

6 4

LINSEED OIL

2

ASTM 0FUEL “A” 0 2

FIGURE 11.17

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

7248_C011.fm Page 225 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

225

HSP OF BINARY AZEOTROPES 10 Azeotropes with HCFC-141b Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8

ETHYL CINNAMATE BUTYL STEARATE

6 4

LINSEED OIL

2

ASTM 0FUEL “A” 0 2

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

FIGURE 11.18

HSP OF BINARY AZEOTROPES 2 Azeotropes with HCFC-225 ca/cb Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8

ETHYL CINNAMATE BUTYL STEARATE

6 4

LINSEED OIL

2

ASTM 0FUEL “A” 0 2

FIGURE 11.19

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

7248_C011.fm Page 226 Thursday, May 10, 2007 8:00 AM

226

Hansen Solubility Parameters: A User’s Handbook

HSP OF BINARY AZEOTROPES 3 Azeotropes with 1,1,2-Trichlorotrifluoroethame Hydrogen Bonding HSP, MPaˆ (½)

14

TRICRESYL PHOS

12 CASTOR OIL

10 8

ETHYL CINNAMATE BUTYL STEARATE

6 4

LINSEED OIL

2

ASTM 0FUEL “A” 0 2

4 6 8 10 Polar HSP, MPaˆ (½)

12

14

Soils

Soils

iti

on

of

Co

m po

ne

nt s

Soils Soils Azeotropes

Ad d

Polar HSP (soil or solvent)

FIGURE 11.20

“Designer” Solvents

H2 Bonding HSP (soil or solvent) FIGURE 11.21

TABLE 11.16 Capability of Azeotropes of Designer Solvents Soil Number

Soil Name

1

ASTM fuel “A”

2

Butyl stearate

3 4 5

Castor oil Ethyl cinnamate Linseed oil

6T

ricresyl phosphate

Designer Solvents from which Azeotropes Can Be Blended to Clean the Stated Soil (RA < 8) HFE-7100, HFE-7200, HFE-8200, OS-10, OS-20, OS-30, HFC-245f a, HFC365mfc, PFC-5060, HCFC-141b, HCFC-225 ca/cb, CFC-113 HFE-7100, HFE-7200, HFE-8200, OS-10, OS-20, OS-30, HFC-365mfc, PFC5060, HCFC-141b, HCFC-225 ca/cb, CFC-113 HFE-7100, HFE-7200, HFE-8200, OS-10, OS-20, OS-30, HCFC-141b HFE-7100, HFE-7200, HFE-8200, OS-10, OS-30, HCFC-141b HFE-7100, HFE-7200, HFE-8200, OS-10, OS-20, OS-30, HFC-245f a, HFC365mfc, PFC-5060, HCFC-141b, HCFC-225 ca/cb, CFC-113 None

7248_C011.fm Page 227 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

227

NOTES 1. In Chapter 11, cleaning solv ents are assumed to be used in a v apor degreasing process. Subboiling (cold cleaning) is not the process for which use of the solvents mentioned in this chapter is envisioned. 2. Durkee, J.B., Management of Industrial Cleaning Technology and Processes, 2006 by Else vier Science, Oxford and Amsterdam. 3. Hansen, C.M., Hansen Solubility Parameters — A User’s Handbook, CRC Press, Boca Raton, FL, 1999, Equation 1.9 and Equation 1.10. 4. Durkee, J.B., What W ould the ‘Perfect’ Cleaning Solv ent Be?, Cleaning Times column in Metal Finishing Magazine, November 2005, p. 61. 5. See Chapter 1, Chapter 2, and Chapter 8 of Reference 3. 6. Note that thorough solutioning of a soil by a cleaning solv ent is no more a sufficient condition for a successful cleaning application than is thorough solutioning of a polymer in a coating carrier solv ent, a sufficient condition for a successful coating application. H wever, thorough solutioning is usually a necessary condition for either type of application to be successful. 7. An efficient approach to making this obser ation is to carefully apply some of the liquid soil composite to a glass slide. When the coated slide is illuminated from belo w, regions that are multiple phases can usually be delineated. One must be cognizant of Heisenber g’s uncertainty principle in preparing to make this observ ation, as the soil sample must be equi valent to the actual residue. 8. Solubility behavior of the tw o solvents is represented in Figure 11.1 as circles. The center point of the circle for each solvent is the locus of their polar and hydrogen bonding HSP values. Because these two solvents are different but miscible, there is a combined ef fect. Equation 11.1 is used to calculate that effect for any blend of solvents. That is the center point for the combined circle. The radius (that will be identified as R ) of each circle (sol ent) represents their sphere of similarity in two dimensions. This is a re gion or zone in HSP space within which HSP v alues are suf ficiently similar to those a the center point. Similarity has a practical meaning. It is that solv ents are miscible, and soils are soluble in solvents. In two dimensions, the region is visualized in Figure 11.1 as a circle. Figure 11.1 is an idealized representation.There are two aspects of this idealized representation of solvent character as graphical areas: (1) solvents or soils or polymers are compatible (lik ely miscible) when their areas overlap (2) whereas the center point HSP v alues can be computed from molecular characteristic (properties), the e xtent of graphical area from the center point must be measured. The latter will be shown in the section on “An Open Question — Answered.” 9. If the soil is a commercial product, its material safety data sheet (MSDS) will lik ely contain infor mation that either is the recipe or from which a suitable estimate of the recipe can be inferred. 10. See Equation 1.9 in Chapter 1 in Reference 3, and Equation 11.3 in “Method for Calculating HSP of Composites (Soils or Solv ents)” in this chapter . 11. Refers to the total or Hildebrand solubility parameter . 12. Reference 3, Chapter 1. 13. Disperse, polar, hydrogen bonding, or total. 14. In Equation 11.1, φ is the v olume fraction of component 1, and δ is an y solubility parameter . It is understood that φcomp 1 + φcomp 2 = 1. Volume fraction is easy to compute because solv ents are stored in pails or drums and used by v olume, although the y are sold by weight. The capacity of a v apor degreaser sump is gi ven in g allons or liters. 15. The appendix to Reference 3 is an excellent source of HSP values for solvents (Appendix Table A.1), polymers, and some soils (Appendix Table A.2). Another useful reference is Barton, A.F.M., CRC Handbook of Solubility Parameters and Other Cohesion Parameters, 2nd ed., CRC Press, Boca Raton, FL, 1991. Data found here are one feedstock for implementation of Equation 11.1. 16. The other feedstock is composition information about azeotropic solv ent composition. This can be found in Durkee, J.B., Management of Industrial Cleaning Technology and Processes, 2006 by Elsevier Science, Oxford and Amsterdam. 17. The reason for inclusion of molecular weight v alues in the basic data is because the mixture concentrations are often given in molar concentration, though that is not the case in this xample. e Composition is normally gi ven on a weight basis in this chapter , because it is that basis by which solv ents are normally sold.

7248_C011.fm Page 228 Thursday, May 10, 2007 8:00 AM

228

Hansen Solubility Parameters: A User’s Handbook

18. In this w ay, cleaning is unlik e construction of coatings. In a coating, all components are generally compatible with the carrier solv ent. In cleaning, where there can be destruction of coatings, all components must be manageable. In solv ent cleaning, soil components can be partially soluble, swollen, insoluble, or soluble. Generally , a cleaning process can be designed to manage remo val of soil components from surf aces without re gard to their relationship to the solv ent. In cleaning, management involves hydrodynamic (mechanical) force, heat, and solv ency. 19. Said in a colloquial manner , “A chain is only as strong as its weak est link.” 20. Hansen, C.M. and Skaarup, K., The three dimensional solubility parameter — key to paint component affinities, J. Paint Technol., Vol. 305, No. 511, 511–514, 1967. 21. Finding the number 4 in Equation 11.3 may be surprising. There is considerable discussion in the literature about the need for it. It has been found empirically useful o ver several decades of practical experience. There are two justifications for it: (1) it co verts three-dimensional spheroidal plots with dispersion solubility parameters to spherical ones, and (2) theoretical considerations associated with the Prigogine theory of corresponding states. See Chapter 5 in this book. 22. Unfortunately, selection of a suitable cleaning solv ent based on similarity of intermolecular forces is but a necessary first step in d velopment of cleaning process applications. Certain selection criteria are listed as follo ws: • A solv ent that has a strong af finity for a soil ut a lo w holding capacity for it (mass solubility) would be a poor choice. • A solvent selected to do cleaning w ork must be able to dissolv e the soil to the e xtent desired in the time allotted under the pre vailing conditions. Cleaning is a transport process. • Also, a poor choice is a solv ent that only gradually penetrates and swells the soil and allo ws it to be removed by rinse fluids • Finally, a solvent that efficiently dissol es an adequate mass of soil in an ef ficient time ut only at a temperature above its boiling point is nearly useless. Pressurized contacting equipment is expensive. Yet, without successful completion of a screening evaluation via use of HSP, development of cleaning process applications becomes either resource intensi ve or based on hearsay e vidence. 23. Note that Figure 11.1 is a generalized presentation. Soils can be matched to solv ents for the purpose of solvent selection; solv ents can be matched to other solv ents for the purpose of producing a blend with certain ph ysical properties, or soils can be matched to other soils for the purpose of e valuating if solution-based or force-based (aqueous technology) should be used, depending upon whether the soils are e xpected to combine in a single or multiple phases. 24. HFE-8200 is actually a component of the product sold as HFE-7100.The latter is mixture of Perfluoro n-butyl methyl ether (35%) and perfluoroiso utyl methyl ether (65%). HFE-8200 is the pure perflu oroisobutyl methyl ether. United States Pharmacopeia (USP) 6,426,327 teaches that azeotropic behavior of mixtures of the tw o isomers and another component is relati vely independent of distribution of isomers in the HFE material. Consequently, in this analysis, HFE-7100 and HFE-8200 are considered equivalent in terms of the boiling point and composition of azeotropes. Ho wever, the HSP properties of HFE-7100 and HFE-8200 are slightly different. This difference has been retained in the calculations of solubility beha vior of the azeotropes. 25. Owing to its inertness as a solv ent, high specific gr vity, and lo w surf ace tension, it finds use i cleaning work as a rinsing agent. Yet, because of its inertness, PFC-5060 is a designer solv ent that on emission migrates without de gradation to the Earth ’s stratospheric layer and acts as a global warming agent. Hence, its use is limited “by design.” 26. The disperse HSP v alue is momentarily ne glected in the interests of graphical clarity when a tw odimensional graph is used. These values are relatively similar. 27. The structure shown represents the most common (89%) component (ricinolenic acid, 12-hydroxyoleic acid) of castor oil. Castor oil is also referred to as the triglyceride of this acid. 28. The structure sho wn, linoleic acid, is the most common (52%) acidic component of linseed oil. The other major acidic component of the triglyceride is linolenic acid that has one fe wer double bond.

7248_C011.fm Page 229 Thursday, May 10, 2007 8:00 AM

Use of Hansen Solubility Parameters

229

29. In the main, these azeotropes are claimed in U.S. patents whose grant date is during the 1990s. The patent assignee is most often the manuf acturer of the designer solv ent. There is no published table of these binary azeotropes e xcept that found in Reference 15, which lists around 450 of them. They were identified in 2004 during an xtensive search of the world patent literature. Readers are reminded that the normal interpretation of U.S. patent law is that when a third component is added to a claimed binary composition, the ternary mixture is still co vered by the U.S. patent. In other w ords, if these binary azeotropes are not commercial products, users cannot formulate them without permission from the patent holder. 30. The parts used are those from the b usiness operation. The soils are those identified by the usiness as being of concern and are applied as pure components to pre viously cleaned parts (or coupons). The method of assaying cleanliness is gra vimetric. All information is publicly a vailable. See http://www.cleanersolutions.org. Most w ork was done around 2004. There are se veral uncertainties about use of these 53 data points: (1) a v ariety of v apor degreasing processes, test conditions, and test e xposure times were used, (2) soil components were identified b chemical abstracts service (CAS) number from the manufacturer’s MSDS that is not always a complete listing of ingredients, and (3) whereas the soil components were identified, their concentration in th applied soil mix w as not. Yet, there w as considerable le vel of replicates a vailable through the use of multiple runs where the same solvent was used under a dif ferent product identification. Hence, one can say with 95% conf dence that if the R A value is less than 8, then the w orst level of cleanliness of the least compatible soil component w ould be 94%. The expected value of cleanliness w ould be around 97%. 31. Normally, the limit of acceptability re garding cleaning performance is set by the requirements of the next processing or use step. F or these laboratory trials, no such information w as available. 32. Both ternary and quaternary azeotropes containing designer solvents have been identified and patented Some find applications in aerosol-p wered cleaning products. Here, there is little concern about composition change in use, as the entire v olume applied evaporated quickly after application. But in vapor degreasing operations, which is the venue of this chapter, the concern about composition change of ternary and quaternary (vs. binary) azeotropes with use cannot be ne glected. 33. Azeotropes in Table 11.4 through Table 11.14 are sorted by increasing v alue of polar HSP with that of the base designer solv ent given in the initial ro w. Boiling points are also gi ven, but not the ro ws that are not sorted by that parameter . In this study , all azeotropes noted are minimum-boiling azeotropes. As all of these designer solv ents have commercial v alue as unique compounds, the azeotropes the y form with other solv ents are claimed in patents. U.S. patent numbers are gi ven as references where available. 34. No denigration of designer solvents is intended in this work. Rather, these solvents represent the result of significant and successful chemical research into identification and man acture of chemical structures, which alle viate man y of the safety , health, and en vironmental concerns about solv ents commonly used in cleaning (and other) w ork in the late 20 th century. Elimination of those concerns was roundly and justifiably applauded by users. In a sense, that research w as too successful. Seemingly, as a class of product, designer solv ents were those whose use raised fe wer concerns, b ut whose value in use w as more limited. Binary azeotropes based on designer solv ents are an attempt to span the g ap between absence of concern and absence of v alue in use. 35. Only blends defined by measured p ysical property and boiling point data are included. Theoretical or predicted azeotropes are not included because the y cannot be purchased as blends for cleaning work. A good source of information about predicted azeotropes is Harding, S.T ., Locating All Heterogeneous and Reactive Azeotropes, distrib uted by the American Institute of Chemical Engineers, (January 1, 1998), ASIN: B0006R4J04. 36. A secondary standard is to fulfill the co ventional requirements upon which solv ent cleaning w ork (vapor degreasing) is based. See Reference 2 for details. 37. This phrase refers to the identifying characteristic of an azeotrope, that the composition of the v apor and the composition of the boiling liquid are identical. In other w ords, as the mixture boils, the composition (like that of a single component solv ent) remains constant.

7248_C011.fm Page 230 Thursday, May 10, 2007 8:00 AM

7248_C012.fm Page 231 Wednesday, May 23, 2007 11:14 AM

— Chemical 12 Applications Resistance Charles M. Hansen ABSTRACT Hansen solubility parameters (HSP) can correlate dif ferences in ph ysical beha vior observ ed in chemical resistance testing of polymers and polymer -containing systems when a suf ficiently la ge number of dif ferent organic solvents are included in a study . These correlations can then be used to predict the chemical attack e xpected in systems that have not yet been tested. Examples of HSP correlations included here are for solubility , de gree of surf ace attack, tensile strength reduction, and simple e valuations of chemical resistance of the suitable-for -use or not type. En vironmental stress cracking is discussed in more detail in Chapter 14. In each case, the molecular size of the liquids used can affect the result and should be considered in some way. Chapter 16 treats absorption and diffusion in polymers with this in mind. A common problem is that tests with larger molecular weight liquids have not reached equilibrium absorption within the timeframe of the e xposure. HSP correlations are presented for chemical resistance studies of epoxy and zinc silicate tank coatings, PET, POM, PA6/66, PUR, PPS, PEI, Neoprene ®, etc.

INTRODUCTION HSP are widely used in the coatings industry to select solv ents for dissolving polymers and binders. This has been discussed in Chapter 8 and also in References 1 through 12, as well as else where. The fact of solution is in itself clearly one simple form of chemical attack of the polymers the y dissolve. This means that chemical resistance for some polymers can be partly inferred from HSP correlations of their solubility and/or swelling. HSP correlations of this type ha ve been discussed in Chapter 5. An example is that if a chemical does not dissolve an epoxy component or the curing agent, then it is quite unlikely that it will attack a fully crosslink ed epoxy coating or glue. The HSP correlations of surf ace phenomena, which have been discussed in Chapters 6 and 7 and else where,13–18 can also provide some insight into chemical resistance. Liquids not wetting a surf ace are not as likely to attack it as those that do wet it, although there are no guarantees. A relation between spontaneous spreading and de wetting of liquids and en vironmental stress cracking has been found. 19 This is discussed in more detail in Chapter 14. Liquids that spontaneously spread were found to induce en vironmental stress cracking at lower critical strains than those liquids that do not. Some surf ace studies may in volve evaluation of a more direct form of chemical attack, such as the attack/whitening of PET coated with “amorphous” PET to improve weldability. This example is discussed in more detail later. Whatever is being correlated, the general considerations of the HSP characterizations discussed in earlier chapters are the same for the HSP correlations of chemical resistance reported here. However, there are certain additional pitfalls to be aware of when correlating chemical resistance. These include (lack of) attainment of equilibrium, the effects of molecular size of the test chemical, dif ference in local se gments of polymers (e ven in homopolymers), and acid/base reactions. Once a reliable HSP characterization of chemical resistance is a vailable, it can be used to calculate the behavior of other systems that have not been tested. Obtaining a good HSP correlation of chemical resistance that allo ws reliable predictions depends v ery much on careful treatment of the available data or generation of data with such a correlation in mind. Unfortunately , very few 231

7248_C012.fm Page 232 Wednesday, May 23, 2007 11:14 AM

232

Hansen Solubility Parameters: A User’s Handbook

studies of chemical resistance have been designed with the purpose of generating HSP correlations. Also, it must be clear that the chemical attack discussed in this chapter does not include true chemical reactions leading to covalent bonding or destruction, such as with acids and bases, whether they are or ganic or inor ganic. Chemical reactions forming ne w compounds are often found with amines and or ganic acids. These reactions often lead to discoloration in systematic solubility parameter testing with one or more of the amines used as test solv ents. Discolored systems should simply be neglected in HSP correlations of physical (reversible) solubility. The products of reactions of well-defined o ganic bases with well-defined o ganic acids ha ve actually been studied systematically from a solubility parameter point of vie w.20 Some of the results of this study are discussed in Chapters 15 and 18.

CHEMICAL RESISTANCE — ACCEPTABLE-OR-NOT DATA Additional sources of data for solubility parameter characterizations include chemical resistance tables reported by ra w material suppliers 21–26 or collected in books 27–29 and other sources such as those supplied on electronic media by the Plastics Design Library .30 Although these data are certainly valuable in themselv es, it has been found that gi ven data sets are not al ways as reliable/consistent/coherent as could be desired for solubility parameter correlations. Attainment of equilibrium may not have been achieved, and this ef fect is rarely confirmed or ven considered. Solvents with low dif fusion coef ficients will appear to be less aggress ve than the y might become at longer exposure times or at higher concentrations. As discussed earlier , true chemical attack with acids and bases must sometimes be sorted out. Likewise, the data are often limited in number and scope, and the chemical reagents have not been chosen for the purpose of solubility parameter correlations. Nevertheless, with the use of due caution, it is often possible to find xcellent solubility parameter correlations using chemical resistance data of the acceptable-or -not type, particularly when the list of agents is long. Additional precautions with re gard to data of the acceptable-or -not type include whether a molecular size ef fect is present as discussed in the follo wing. Also, it can be assumed that if a chemical attacks a polymer at, say , 20°C, then it will also attack it at, say , 70°C. It can then be included as data in a 70°C correlation, e ven though it may not ha ve been tested at that temperature (and in principle the HSP are only v alid at room temperature).

EFFECTS OF SOLVENT MOLECULAR SIZE It has been emphasized in Chapters 1, 2, and 5 that the size of solv ent molecules is important for polymer solubility. HSP correlations ha ve confirmed that this e fect is even more important when chemical resistance is being considered. Smaller molecules are e xpected to be better, that is, more aggressive from a thermodynamic point of view than larger ones, all else being equal. This is known from the theories of polymer solubility discussed in Chapters 1 through 4, and also from the discussion of barrier polymers and dif fusion found in Chapter 13 and 16, respecti vely. So it is not surprising that solv ent molecular size can be an important fourth parameter in correlations of chemical resistance. An appropriate w ay to check this is sorting output data from a computer (or other) HSP optimization according to the molecular v olume of the test solv ents. What appear to be errors in the correlation may become systematically arranged. It can easily be seen, if the top of the list includes the type of “error” where the smaller molecular species are “better” thanxpected e by comparison with all the other solv ents. This may tak e the form of une xpectedly dissolving, being more aggressive than expected, penetrating more rapidly, or reducing mechanical properties more severely. Larger molecular species which are “poorer” than expected by comparison with the data for the other solvents are often seen at the bottom of the list. One can focus upon the molecular size range of greatest interest in such cases and repeat the correlation, ne glecting those species which are outside of this range of molecular v olumes. The correlation is then strictly v alid only

7248_C012.fm Page 233 Wednesday, May 23, 2007 11:14 AM

Applications — Chemical Resistance

233

for the size range specified. Some indication of the beh vior of the solvents with V larger than the upper limit is possible if their RED numbers are greater than 1.0. These would not be e xpected to attack under an y circumstances. Lik ewise, the solv ents with V less than the lo wer limit can be expected to attack if their RED numbers are less than 1.0. Lar ger numbers of solv ents are needed in the study if this is to be done with an y benefit As stated abo ve, the size and shape of solv ent molecules are v ery important for kinetic phenomena such as dif fusion, permeation, and attainment of equilibrium. Chapters 13 and 16, respectively, discuss correlations of HSP and elaborate on dif fusion phenomena in more detail. However, it will be repeated here that smaller and more linear molecules dif fuse more rapidly than larger and more bulky ones. The diffusion coefficient may be so l w that equilibrium is not attained for hundreds of years at room temperature in common solv ent e xposures of rigid polymers lik e polyphenylene sulfide (PPS) with thicknesses of s veral millimeters. 31 Such ef fects lead to comparisons where some systems may ha ve reached an equilibrium uptak e, whereas others ha ve not. Likewise, the second stage in the tw o-stage drying process in polymer film formation by sol ent evaporation can last for man y years. 4,31 See Chapter 16. Polymer samples used for solubility parameter or chemical resistance testing may contain retained solvent or monomer for many years, and this may also af fect the e valuations. However, the ef fects of w ater can be e xtremely rapid as discussed in the follo wing.32 Attempts to include the molecular v olume into a ne w composite solubility parameter and size parameter have not been particularly successful. 33,34 This may be because the size ef fects are not necessarily caused through the thermodynamic considerations on which the solubility parameters are based, b ut rather through a kinetic ef fect of diffusion rate.

CHEMICAL RESISTANCE — EXAMPLES Chemical resistance means dif ferent things in dif ferent contexts. Various examples of HSP correlations of chemical resistance are included in the follo wing. The HSP data for the correlations discussed are included in Table 12.1. Experimental data are always preferred over predicted behavior based on a correlation. However, a good HSP correlation can be used to find ma y chemicals that will clearly attack or that will clearly not attack. There are also situations where the attack is mild, and whether or not satisfactory results are found with a product depends on its use. Data in chemical resistance tables are often of the type +, +/–, –, or satisf actory/questionable/unsatisfactory, recommended/not recommended (R/NR), or something similar . The liquids which attack are clearly good solv ents for the material in question and will be located within the HSP spheres with RED numbers being successi vely lower for more severe attack, all other things being equal. The correlations can include the solvents with mild attack (+/, questionable) either in the attacking (NR) group or in the nonattacking (R) group. They can also be neglected, not knowing which group to put them into. The treatment used in the individual correlations presented here is indicated in the follo wing. Unless otherwise specified the results are for room temperature.

TANK COATINGS Chemical resistance is important for tank coatings used in the transport of b ulk chemicals. The data in Table 12.1 include two older HSP correlations for chemical resistance for two types of tank coatings supplied by Hempel’ s Marine Paints. These are for a tw o-component epoxy type and for a zinc silicate type. The data and correlations are about 20 years old. They are included here for purposes of demonstration. Impro vements in chemical resistance are kno wn to ha ve been implemented in a ne wer epoxy tank coating, b ut no HSP correlation has been made. A HSP correlation of the solubility of a lower molecular weight epoxy, Epon® 1001 (Shell Chemical Corp.), is included for comparison. The numbers are not too dif ferent from those of the HSP correlation for chemical

7248_C012.fm Page 234 Wednesday, May 23, 2007 11:14 AM

234

Hansen Solubility Parameters: A User’s Handbook

TABLE 12.1 Hansen Solubility Parameter Correlations for Selected Materials Material

δD

δP

δH

Ro

FIT

G/T

Epoxy tank coat (tw o component) Epoxy solubility (Epon 1001) Zinc silicate coating PET-amorphous coating PET-CR (+/– R b) PUR-CR (+/– R b) POMC/POMH (+/ Rb) (+/– R b) POMC (+/– NR b) PA6/PA66 (+/– NR b) Halar 300 ECTFE 23°C Halar 300 ECTFE 50°C Halar 300 ECTFE 100°C Halar 300 ECTFE 120/149°C Neoprene-CR (+/– R b) PPS tensile strength <60% 93°C PEI ULTEM 1000 600 psi PEI ULTEM 1000 1200 psi PEI ULTEM 1000 2500 psi PEI ULTEM 1000 solubility PES mechanical properties PES solubility

18.4 18.1 23.5 17.0 18.2 18.1 17.1 17.9 18.9 16.8 18.1 18.1 18.3 18.1 18.7 17.3 17.0 17.4 19.6 17.1 19.6

9.4 11.4 17.5 11.0 6.4 9.3 3.1 5.9 7.9 8.4 7.5 7.9 8.7 4.3 5.3 5.3 6.0 4.6 7.6 9.9 10.8

10.1 9.0 16.8 4.0 6.6 4.5 10.7 8.3 11.7 7.8 8.5 7.9 7.9 6.7 3.7 4.7 4.0 9.0 9.0 6.3 9.2

7.0 9.1 15.6 9.0 5.0 9.7 5.2 6.6 8.7 2.7 5.2 6.7 7.5 8.9 6.7 3.3 4.0 7.2 6.0 6.3 6.2

—a —a —a 1.000 0.865 0.981 0.955 0.609 0.950 0.993 0.700 0.710 0.800 0.937 0.991 1.000 1.000 0.967 0.952 0.931 0.999

—a —a —a 7/11 7/34 16/26 2/28 11/28 9/31 2/102 18/92 49/91 48/74 30/48 9/16 3/20 4/20 9/20 8/45 6/25 5/41

Note: The symbols G for good solv ents and T for total solv ents are maintained, with the understanding that G solv ents are within the HSP correlation spheres and are not recommended for use. a b

Data not a vailable. See text: NR — not resistant, R — resistant.

resistance for the tw o component epoxy tank coating. These three correlations ha ve been reported earlier.8 The fact of a successful HSP correlation for a completely inor ganic type of coating lik e zinc silicate is surprising. This is still another demonstration of the uni versality of the applications possible with the HSP concept. Although the data fit numbers were not recorded at the time, th two chemical resistance correlations reported here were clearly considered reliable.

PET FILM COATING Another e xample of chemical resistance, or lack of the same, is the attack of the amorphous, modified PET coating on PET films to imp ve their weldability. This correlation is based on only 11 well-chosen data points b ut clearly sho ws that attack for man y chemicals can be e xpected. Among those chemicals not attacking are h ydrocarbons, glycols, and glycol ethers and higher alcohols which ha ve a reasonably high h ydrogen bonding character.

ACCEPTABLE

OR

NOT — PLASTICS

Several examples of HSP correlations of data reported in the form acceptable-or -not are included in Table 12.1. The data for these are all found in Reference 21. Other data sources for these are also a vailable. HSP correlations of this type are included for PET , PUR, POMH, POMC, and

7248_C012.fm Page 235 Wednesday, May 23, 2007 11:14 AM

Applications — Chemical Resistance

235

PA6/PA66 using data from Reference 21. The first three correlations of this type consider reporte evaluations of minor attack which will require further e valuation (+/–) as if these systems were suitable for use (resistant). The last two correlations consider this condition as not suitable for use (not resistant). This is to demonstrate that dif ferences are found, depending on ho w the data are considered and that outliers are often found when correlating this type of data. It is for this reason that more e xtensive tables of HSP correlations of chemical resistance are not reported here. Too much space is required to try to e xplain wh y gi ven results are outliers. Ho wever, some reasons have been given earlier and others are in the follo wing. The HSP v alues for PET based on chemical resistance are some what different from those of the amorphous PET coating, which is readily attack ed by far more solvents. The compositions are also different. The POM correlations ha ve typical problems in that the correlation considering all minor attack as ne gligible is based on only tw o severely attacking solvents among the 28 solv ents tested. When the solv ents demonstrating minor attack are considered as being in the attacking group, the data fit sh ws that there are man y outliers. A systematic analysis of wh y this is found will not be attempted for reasons of space, e ven if all the outliers could be e xplained in one w ay or another. The HSP found for a polymer in this type of correlation may not be representati ve of that polymer in all aspects of its beha vior. There is a question as to co verage over the whole range of solvent HSP possible by the test solvents. Additional solvents may be required to make an improved correlation based on the improved coverage possible. There is also a question as to which segments of the polymer may be subject to attack by which solv ents. Block copolymers may demonstrate two separate (o verlapping) correlations that cannot be reasonably force fitted into a single HS correlation. Viton® is an e xample of this. The most se vere attack or swelling may occur in one region or another of the polymer or maybe e ven on a third component, such as a cross-linking segment. Viton® is discussed in Chapter 5 and Chapter 13 in more detail. In spite of these pitf alls, it is strongly suggested that those generating this type of resistance data should try HSP correlations to evaluate the consistency of the data before reporting it. Outliers can be reconsidered; whether or not equilibrium has been attained can be inferred, and the probable effects of solv ent molecular size may become apparent. The effects of temperature on the chemical resistance of poly(eth ylene co-chlorotrifluoroeth ylene) (ECTFE) Halar ® 300 can be seen in Figure 12.1. The data on which these correlations are based are of the recommended-or -not type and were found in Reference 26. This figure ha affectionately been dubbed a “b ullseye,” as there appears to be symmetry about a central point, although this is not strictly true as the HSP data confirm. The radius of the chemical resistance spheres increases with increasing temperature, as e xpected, as more solv ents then become more severe in their attack. The HSP data for these correlations are also included in Table 12.1. The data fits are not particularly good at the higher temperatures To complete this section, a correlation of chemical resistance data for Neoprene ® rubber (Du Pont)22 is included. Solv ents in the intermediate cate gory, i.e., that of a questionable-for -use recommendation, are considered as being in the nonattacking group for this correlation. Previously it w as indicated wh y HSP correlations of this type lead most often to guidance rather than to a firm recommendation. There are man y pitfalls to be a ware of both in generating such correlations as well as in using them, b ut their usefulness becomes clearer with some e xperience. A suitable goal for a future project is to determine the ef fective HSP for various media frequently encountered in resistance lists, such as mustard, certain detergents and oils, etc. This could perhaps be done by composition in some cases. In other cases, one could see whether beha vior paralleled that of a kno wn chemical. A third method is to determine these parameters by recognizing a similarity to all materials attack ed and a dif ference from those not attack ed. This approach has been used to assign HSP to some liquids when calculations were uncertain. A computer program was de veloped similar to the SPHERE program as described in Chapter 1, b ut w orking in the

7248_C012.fm Page 236 Wednesday, May 23, 2007 11:14 AM

236

Hansen Solubility Parameters: A User’s Handbook

SOLVENT RESISTANCE OF HALAR 300 AT 4 DIFFERENT TEMPERATURES

14 13 12 11 HANSEN POLAR PARAMETER, δH

120° = 149° 10 100° 9

23°

120°✚ 23° ● 100°▲ 50° ✳

8

50°

7 6 5 4 3 2 1 0 0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

HANSEN HYDROGEN BONDING PARAMETER, δH

TEMP.

D

P

H

R

FIT

NO(NR)

23°

16.8

8.4

7.8

2.7

0.993

102 (2)

50°

18.1

7.5

8.5

5.2

0.70

92 (18)

100°

18.1

7.9

7.9

6.7

0.71

91 (49)

120/149°

18.3

8.7

7.9

7.5

0.80

74 (48)

FIGURE 12.1 Chemical resistance of Halar® 300 ECTFE at various temperatures. Liquids within the spheres (circles) are not recommended at the gi ven temperatures. HSP data gi ven in Table 12.1.

7248_C012.fm Page 237 Wednesday, May 23, 2007 11:14 AM

Applications — Chemical Resistance

237

opposite manner. The solubility of a number of polymers was evaluated in the solvent. The solvent parameters were then systematically v aried by the program to reduce the collecti ve error, that is, to locate the best possible set of HSP for the solv ent. In general, the data fits for this procedur were comparable to those found for polymer solubility using the SPHERE program. In other words, not all the predictions based on the HSP thus assigned to the solv ent agreed with the e xperimental data, but the errors were small.

TENSILE STRENGTH The long-term exposure of polymers or polymer composites to solv ents normally leads to changes in mechanical properties. One of the more direct techniques to measure such effects is to determine the tensile strength. The tensile strength reduction for glass fiber reinforced polyphe ylene sulfid (PPS) after e xposure to a number of solv ents at 93°C for 12 months has been reported. 23 A HSP correlation of these data using the “good” solv ents as those which reduce tensile strength under these conditions to less than 60% of the initial v alue is found in Table 12.1 and Figure 12.2. More extensive correlations for PPS are found in Reference 31. Additional HSP tensile strength correlations have been generated for polyetherimide, ULTEM® 1000, using data reported by General Electric.24 It is clear that the chemical resistance is dependent on the stress level. Higher stress levels lead to more severe attack by a larger number of chemicals. The solvents considered as being those which attack led to cracking during the study , which lasted 336 h. Some led to earlier cracking than others, which could be treated in a separate correlation, but this has not been done.A more rapid attack is expected from the better solvents with the smallest size and shape. The correlations all ha ve high data fits.

δP

δD

PPS

δH

5

FIGURE 12.2 HSP correlation of the tensile strength reduction of Ryton ® PPS. Within the sphere are liquids which reduce the tensile strength to less than 60% of the original v alue after e xposure for 1 year at 93°C. (From FORCE Institute, Solv ent Resistance of Polymer Composites, Glass Fibre Reinforced Polyether Sulphone (PES), 1st ed., Center for Polymer Composites, 1994, 31. With permission.)

7248_C012.fm Page 238 Wednesday, May 23, 2007 11:14 AM

238

Hansen Solubility Parameters: A User’s Handbook

The next entry in Table 12.1 is for true solubility of UL TEM 1000. These data were generated in a standard solubility parameter study. This correlation is not directly comparable with the previous ones for ULTEM 1000 as the number and range of solubility parameters included in the test solv ents are different. The parameters for the polymer found in this correlation are those e xpected to reflec its (thermodynamic) affinities most closel . The previous study24 did not include a sufficient numbe of solvents having widely different HSP to gi ve a true total picture of the UL TEM 1000. A final HSP correlation of the suitable-fo -use or not type is presented in Table 12.1. This is for polyethersulfone (PES) based on mechanical properties after e xposure to v arious liquids. The HSP correlation for the recommendation from the supplier 25 for glass fiber reinforced PES (Ultra son® E, BASF) can be compared with the HSP correlation for simple solubility of the polymer in standard test liquids, also gi ven in Table 12.1. There is a dif ference, but it is not lar ge.

SPECIAL EFFECTS WITH WATER As stated else where in this book (Chapter 1 and Chapter 15, in particular), the seemingly unpredictable behavior of water has often led to its being an outlier in HSP correlations. F or this reason, it is suggested that data for w ater used as a test solv ent not be included in HSP correlations. Water can be a v ery aggressi ve chemical. Water uptak e in most polymers increases with increased temperature. This is because the solubility parameters of the water and polymer are closer at higher temperatures. The very high δH parameter for w ater decreases more rapidly with increasing temperature than the δH parameter for most polymers. This has been discussed in Chapter 1 and Chapter 8, but it is repeated here with e xamples for those interested in chemical resistance. Water is an e xceptionally good plasticizer because of its small molecular size. The presence of water not only softens (reduces the glass transition temperature) a polymer as such, b ut it also means diffusion rates of other species will be increased. Therefore, the presence of w ater in a fil can influence the upta e of other materials, with h ydrophilic materials, in particular , being more prone to enter the film The increase of w ater uptak e with increased temperature can cause special problems with blistering if the temperature of a w ater-saturated polymer falls rapidly to a lower temperature. The previously soluble water can no longer be truly dissolved. Some of the water already in the polymer is now in e xcess and suddenly appears as small clusters or droplets of freed liquid w ater within the polymer itself (see Chapter 8, Figure 8.3). These droplets can quickly collect into blisters, especially if there are hydrophilic sites in the polymer or at an interface to which water will rapidly diffuse. This special type of failure has been discussed in more detail elsewhere32 (see also Chapter 1 and Chapter 8). The phase separated w ater has been called SWEA T (soluble w ater e xuded at lowered temperatures). The author has observ ed this phenomenon as a mechanism of f ailure for epoxies, polyesters, alkyds, polyethersulfone (PES), polyphenylene sulfide (PPS), and ven EPDM rubber. This mechanism can be confirmed xperimentally by cycling samples continually e xposed to water between two relevant temperatures using a quench from the higher one to the lo wer one. One follows weight g ain by rapidly weighing samples that are surf ace dry. Typical results for the SWEAT phenomena for EPDM are seen in Figure 12.3 and for PPS in Figure 12.4. Control samples that are not c ycled reach equilibrium and stay there, whereas the c ycled samples suddenly be gin to gain weight well beyond the equilibrium value. The extra weight is phase-separated water within the samples. This has been discussed in detail in Reference 35 for PPS and PES. A related problem can be encountered in chemical resistant coatings for tanks that ha ve been in contact with methanol. If a coated tank has been used to store methanol, and perhaps hot methanol in particular, the coating is more than lik ely saturated with methanol. It may tak e several days of exposure to fresh air (to reduce the amount of methanol to acceptable le vels) before subsequent direct contact with w ater or sea water can be tolerated. If there is too much methanol retained in the coating, the w ater diffusing into the coating will associate with the methanol. The increasing

7248_C012.fm Page 239 Wednesday, May 23, 2007 11:14 AM

Applications — Chemical Resistance

239

WATER ABSORPTION, CYCLING 120°-15° EPDM-GASKET 5.00 4.50

WATER UPTAKE, W/W%

4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00

0

1

2

3 4 EXPOSURE TIME, √DAYS

5

6

7

FIGURE 12.3 A rapid quench to a lo wer temperature can free w ater already dissolv ed in a polymer in the form of SWEAT. SWEAT can lead to blistering, cracking, and delamination. The data in the figure are weigh gain for EPDM with c ycling in w ater between 120 and 15°C. Water in e xcess of the equilibrium v alue at longer cycling times is SWEAT.

water content in the mixture of methanol and w ater will ultimately cause the solubility parameters of the mixture to be suf ficiently high so that it becomes incompatible with the coating. Blister form and total delamination can occur . These blisters are often near the substrate, as this is where the retained methanol will be found at highest concentrations.

CONCLUSION HSP can correlate dif ferences in ph ysical beha vior observ ed in chemical resistance testing of polymers and polymer containing systems when a suf ficiently la ge number of dif ferent organic solvents are included in a study. HSP correlations including systematic consideration of the solvent molar volume (or other suitable size parameter[s]) should be an inherent part of all future studies of chemical resistance. These correlations aid in the determination of whether equilibrium has been attained, as well as pro vide insight into the beha vior expected from untested solv ents whose HSP are stored in a solv ent database or can be calculated. Examples include HSP correlations of true solubility and swelling, de gree of surf ace attack, tensile strength reduction, and correlations for simple e valuations of chemical resistance of the suitable-or-not type. It is reemphasized that, in each case, the molecular size of the liquids vealuated will affect the result, and this should be considered in some w ay. Data for true acidic or basic chemical attack must not be included in HSP correlations, as HSP correlations reflect p ysical attack and not chemical attack. It is strongly suggested that data for water not be included in these correlations as well. Its behavior is too unpredictable compared with other test liquids, and if it is included as an outlier , this f act will force a correlation with less predictive ability than had it been ne glected.

7248_C012.fm Page 240 Wednesday, May 23, 2007 11:14 AM

240

Hansen Solubility Parameters: A User’s Handbook

WATER ABSORPTION VS.TEMPERATURE RYTON R4 XT-HV, plate thickness 1 and 2 mm 2 90 C, 1 mm ■

1.5

90 C, 1 mm

.5

▲ ▲ ▲ ▲ ▲ ■ ▲ ▼

▲ ▲ ▲

■

■ ■

▲ ▲

90 C, 2 mm ▲ ▲ ■▲ ▲

■ ■

50 C, 1 mm

■

✕

■

▼ ▼ ✕ ▲ ▲ ▲ ▼ ▲ ▲ ▲ ✕ ✛ ▼ ■ ✛ ▼ ▼ ▼ ▼ ▼ ▼ ▼

▲

0

▲ ▲ ▲ ▲ ▲ ▲

▲▲ ▲ ▲ ▲ ▲▲ ▲

✕ ✛

▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲

WEIGHT GAIN (%)

■

1

✕ ✛

✕ ✛

✕

23 C, 1 mm

✛

✛

▼

▼

-.5

▼ ▼

140 C, 2 mm ▲▲ ▼▼

-1 0

10

20 30 40 50 60 70 SQUARE ROOT OF TIME / THICKNESS SQRT (HRS)/mm

80

FIGURE 12.4 A rapid quench to a lo wer temperature can free w ater already dissolv ed in a polymer in the form of SWEAT. SWEAT can lead to blistering, cracking, and delamination. The data in the figure are weigh gain for PPS with cycling in water between 90 and 23°C. Water in excess of the equilibrium value is SWEAT. (From Hansen, C.M., The Resistance of PPS, PES and PA Polymer Composites to Temperature Cycling During Water Exposure, Center for Polymer Composites (Denmark), Danish Technological Institute, Taastrup, 1994, 11. With permission.)

REFERENCES 1. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities I Solvents, plasticizers, polymers, and resins, J. Paint Technol., 39(505), 104–117, 1967. 2. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities II Dyes, emulsifiers, mutual solubility and compatibilit , and pigments , J. Paint Technol., 39(511), 505–510, 1967. 3. Hansen, C.M. and Skaarup, K., The three dimensional solubility parameter — key to paint component affinities III. Independent calculation of the parameter component , J. Paint Technol., 39(511), 511–514, 1967. 4. Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Their Importance in Surf ace Coating F ormulation, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967. 5. Hansen, C.M., The universality of the solubility parameter , Ind. Eng. Chem. Prod. Res. Dev., 8(1), 2–11, 1969. 6. Hansen, C.M. and Beerbo wer, A., Solubility parameters, in Kirk-Othmer Encyclopedia of Chemical Technology, Suppl. Vol., 2nd ed., Standen, A., Ed., Interscience, Ne w York, 1971, pp. 889–910. 7. Hansen, C.M., 25 years with solubility parameters (25 År med Opløselighedsparametrene, in Danish), Dan. Kemi, 73(8), 18–22, 1992. 8. Hansen, C.M., Solubility parameters, inPaint Testing Manual, Manual 17, Koleske, J.V., Ed., American Society for Testing and Materials, Philadelphia, P A, 1995, pp. 383–404.

7248_C012.fm Page 241 Wednesday, May 23, 2007 11:14 AM

Applications — Chemical Resistance

241

9. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press, Boca Raton, FL, 1983. 10. Barton, A.F.M., Handbook of Polymer-Liquid Interaction Parameters and Solubility Parameters, CRC Press, Boca Raton, FL, 1990. 11. Anonymous, Brochure: Co-Act — A Dynamic Program for Solv ent Selection, Exxon Chemical International Inc., 1989. 12. Dante, M.F., Bittar , A.D., and Caillault, J.J., Program calculates solv ent properties and solubility parameters, Mod. Paint Coat., 79(9), 46–51, 1989. 13. Hansen, C.M., Characterization of surfaces by spreading liquids, J. Paint Technol., 42(550), 660–664, 1970. 14. Hansen, C.M., Surface dewetting and coatings performance, J. Paint Technol., 44(570), 57–60, 1972. 15. Hansen, C.M. and Pierce, P.E., Surface effects in coatings processes, Ind. Eng. Chem. Prod. Res. Dev., 13(4), 218–225, 1974. 16. Hansen, C.M. and Wallström, E., On the use of cohesion parameters to characterize surfaces, J. Adhes., 15(3/4), 275–286, 1983. 17. Shareef, K.M.A., Yaseen, M., and Reddy , O.J., Suspension interaction of pigments in solv ents. Characterization of pigments surfaces in terms of three-dimensional solubility parameters of solvents, J. Coat. Technol., 58(733), 35–44, February 1986. 18. Beerbower, A., Surface free energy: a new relationship to bulk energies, J. Colloid Interface Sci., 35, 126–132, 1971. 19. Nielsen, T.B. and Hansen, C.M., Surf ace wetting and the prediction of en vironmental stress cracking (ESC) in polymers, Polym. Degradation Stability, 89, 513–516, 2005. 20. Hansen, C.M., Some aspects of acid/base interactions (Einige Aspekte der Säure/Base-W echselwirkung, in German), Farbe und Lack, 7, 595–598, 1977. 21. Anonymous, Plastguide, SCS Dukadan AS, Randers, Denmark, 1990. 22. Anonymous, Fluid Resistance of Viton®, Du Pont Compan y, Polymer Products Department, Elastomers Division, Wilmington, DE, 1989. 23. Anonymous, RYTON® PPS Polyphen ylene Sulfide Resins — Corrosion Resistance Guide, Philip Petroleum Co., U.S. 24. Anonymous, Ultem ® Resin Design Guide, GE Plastics, Pittsfield, MA, 1989 25. Anonymous, Verhalten v on Ultrason ® ge gen Chemikalien — B ASF Technische Information TI-KTE/TH-01 d 82132, October 1991. 26. Anonymous, Expanded List — Chemical Resistance of Halar ® Fluoropolymer, Ausimont, USA, Inc. 27. Carlowitz, B., Thermoplastic Plastics (in German), Thermoplastische Kunststoffe, Zechner und Hüthig, Speyer am Rhein, 1980. 28. Anonymous, Chemical Resistance Data Sheets Volume 1 Plastics; Volume 2 Rubbers, new ed., Rapra Technology Limited, Sha wbury, Shrewsbury, Shropshire, 1993. 29. Anonymous, Chemical Resistance of Plastics and Elastomers used in Pipeline Construction, Geor ge Fischer +GF+, 1992. 30. Plastics Design Library, Chemical resistance data, William Andrew, Inc., Norwich, NY. 31. Hansen, C.M., Solvent Resistance of Polymer Composites — Glass Fibre Reinforced Polyphen ylene Sulfide, Centre for Polymer Composites (Denmark), Danish Technological Institute, Taastrup, 1993, pp. 1–62. 32. Hansen, C.M., Ne w developments in corrosion and blister formation in coatings, Prog. Org. Coat., 26, 113–120, 1995. 33. Van Dyk, J.W ., P aper presented at the F ourth Chemical Congress of America, Ne w York, August 25–30, 1991. 34. Anonymous [Note: This was, in fact, Van Dyk, J.W., but this does not appear on the b ulletin], Using Dimethyl Sulfoxide (DMSO) in Industrial F ormulations, Bulletin No. 102, Gaylord Chemical Corp., Slidell, LA, 1992. 35. Hansen, C.M., The Resistance of PPS, PES and P A Polymer Composites to Temperature Cycling During Water Exposure, Centre for Polymer Composites (Denmark), Danish Technological Institute, Taastrup, 1994.

7248_C012.fm Page 242 Wednesday, May 23, 2007 11:14 AM

7248_C013.fm Page 243 Wednesday, May 23, 2007 11:18 AM

— Barrier 13 Applications Polymers Charles M. Hansen ABSTRACT The permeation coef ficient, P, of a liquid or a g as through a polymer is gi ven by the product of the diffusion coefficient, D, and the solubility coef ficient, S: P = DS. S correlates with the Hansen solubility parameters (HSP). At lo w permeant concentrations D is a constant. Ho wever, as the permeant concentration increases, its plasticizing ef fect on the polymer becomes significant, an the dif fusion coef ficient increases mar edly. This ef fect can be v ery significant. The successful correlations of permeation phenomena with HSP are thought to be largely a result of this exceptional dependence of D on the dissolved permeant. As the amount of permeant being dissolv ed increases with closer matches of the HSP for permeant and barrier polymer , the end result is that both S and D, and therefore P, are functions of the HSP match. HSP correlations are gi ven for breakthrough times in chemical protecti ve clothing, permeation rates through barrier polymers, and barrier polymer swelling. Both liquids and g ases are treated. Absorption and dif fusion in polymers is treated extensively in Chapter 16.

INTRODUCTION The permeation of a liquid or a g as through a polymer can be described by the relation P = DS

(13.1)

P, the permeation coef ficient, is the product of the di fusion coefficient, D, and the solubility coefficient, S. The diffusion coefficient indicates h w fast the permeant molecules can move through the polymer. The solubility coefficient indicates h w much of the permeant can be dissolv ed in the polymer. The amount dissolv ed in the polymer determines the concentration gradient o ver a film and the concentration gradient is the dri ving force for mass transport. When solubility is higher , the concentration gradient is correspondingly higher, and, assuming the same diffusion coefficient mass transport will be proportionately higher . S will be lo wer when the HSP of the barrier fil and a solv ent are v ery different. A significant actor af fecting D is the molecular size and shape of the permeant molecules. Larger molecular size and more comple x and bulky molecular shape are major f actors that lead to lower diffusion coefficients. The diffusion coefficient for oxygen in polyvi yl chloride (PVC) is well over a million times greater than that of n-hexane (at low concentrations) in the same polymer.1 This difference in diffusion coefficients is a result of di ferences in molecular size. Likewise, it has been found that the rate of dif fusion at the same concentration is about the same for dif ferent solvents with approximately the same size and shape, even though they may have different solubility parameters (b ut not so dif ferent that both are able to dissolv e in the polymer at the le vel of comparison).2–4 The polymers in these studies were a copolymer of 87% vin yl chloride and 13% vinyl acetate, polyvin yl acetate, and polymeth yl methacrylate.

243

7248_C013.fm Page 244 Wednesday, May 23, 2007 11:18 AM

244

Hansen Solubility Parameters: A User’s Handbook

CONCENTRATION-DEPENDENT DIFFUSION Low molecular weight liquids are plasticizers for polymers if they can be dissolved in them. Water, for e xample, can significantly soften ma y polymers e ven though it is dissolv ed to only a fe w percent. The low molecular weight materials can greatly reduce the glass transition temperature of their mixtures with a polymer as the y have considerably more free v olume associated with them than the polymers themselv es. This e xtra free v olume allo ws easier polymer se gmental motion. The diffusion of the smaller species (and other species) becomes f aster as their local concentration and plasticizing effect become greater. The solvent diffusion coefficient data in Figure 13.1 were first presented in Reference 3. S also Chapter 16. This figure sh ws diffusion coefficients for s veral solvents in polyvin yl acetate (PVAc) at 25°C. The diffusion coefficient for ater shown in the figure as found by absorption and desorption e xperiments in thin films where a correction for the sur ace resistance w as also required.5 See Chapter 16. It can be seen in this figure that for moderate sol ent concentrations in this rigid polymer, the local dif fusion coefficient increases by a actor of about 10 for an increase in solvent concentration of about 3 to 4 v ol%. As this behavior is general for solvents in polymers, a rule of thumb indicates that the local dif fusion coef ficient for sol ents in rigid polymers can increase by a f actor of about one million when about 20 v ol% solvent is present compared with the solvent-free state. 3–7 This rule of thumb assumes that the polymer beha ves as a rigid polymer over the concentration range being considered. This difference corresponds to the speed of a snail in the w oods compared with a modern jet airliner .

7

-LOG DIFFUSION COEFFICIENT

CM2/SEC

A 8 B C

9

D 10 11 12 13 14 15

0

8 16 24 VOLUME FRACTION PENETRANT

32

FIGURE 13.1 Diffusion coefficients in polyvi yl acetate at 25°C for methanol (A), eth ylene glycol monomethyl ether (B), chlorobenzene (C), and c yclohexanone (D). Original data are in Reference 3. The data point for w ater (*) is included for comparison. (From Hansen, C.M., Permeability of polymers, Pharmaceutical and Medical Packaging 98, 1998, 7.12 With permission.)

7248_C013.fm Page 245 Wednesday, May 23, 2007 11:18 AM

Applications — Barrier Polymers

245

Concentration-dependent diffusion coefficients are also found for elastomers. Here, the rule o thumb is that the dif fusion coefficient increases by a actor of about 10 for an increase in solv ent concentration of about 15 vol%.7 This shows that liquid contact with chemical protecti ve clothing, for example, leads to concentration-dependent dif fusion coefficients because the amount ta en up at the contact surf ace on liquid contact is v ery often more than 15%. Concentration-dependent diffusion has been discussed at length by Crank. 8 It is also discussed here because it is a major factor in the success of HSP correlations of permeation phenomena. The Crank-Nicholsen finite di ference treatment for concentration-dependent dif fusion8 was extended by Hansen3 and used to describe film formation by sol ent evaporation,4 to explore what is termed anomalous diffusion,5 to develop an easy method to e valuate data leading to concentration-dependent diffusion coefficients 6 and to account for the effects of concentration-dependent diffusion and surface boundary resistance simultaneously .5–7 Klopfer 9 de veloped analytical solutions in volving concentration-dependent dif fusion for man y situations found in practical b uilding applications, particularly with respect to transport of w ater in b uilding materials. Concentration-dependent diffusion can be handled properly without great dif ficulty for most situations of practical interest Neglect of this ef fect can lead to errors, the significance of which will increase with increasin amounts of the dissolv ed materials. In addition to demonstrating concentration dependence, the diffusion coefficient data for P Ac in Figure 13.1 also sho w the well-established relations that those solv ents with lar ger and more complicated chemical structures are those with lo wer diffusion coefficients. Water has one “significant” atom, methanol has tw o, and eth ylene glycol monometh yl ether (EGMME) has fi e. The diffusion coefficient for ater in PVAc at lo w concentration, Do, is 10,000 times lar ger than that for the latter. An example of how to estimate diffusion coefficients in P Ac for other liquids, such as meth ylene chloride, is as follo ws. The diffusion coefficients in P Ac for meth ylene chloride, with three significant atoms, can be xpected to be some what lower than those for methanol, b ut much higher than those for EGMME. Planar chlorobenzene dif fuses more rapidly than nonplanar cyclohexanone, even though the number of significant atoms is the same. Another type of comparison which is possible is to state that the dif fusion coefficients for toluene are xpected to be close to those for chlorobenzene because of a similarity in molecular size and shape. This was confirme by solvent retention studies where toluene and chlorobenzene were retained in identical amounts in a film of VYHH® (87 wt% vinyl chloride, 13 wt% vinyl acetate, Union Carbide). Toluene, which does not dissolv e this polymer , w as introduced by placing a completely dry polymer film in closed container o ver toluene vapors. Diffusion can be e xpected to be slo wer in more rigid polymers, i.e., those with higher glass transition temperatures, unless the rigidity is such as to allo w decided holes of suitable size to enable quite rapid dif fusion of much smaller molecules. These considerations lead to the best combination of properties for a barrier polymer as being one with a high glass transition temperature and with HSP far removed from those of the permeant. If, in practice, this leads to water sensitivity, an alternate strate gy, such as a laminated system, may be required.

SOLUBILITY PARAMETER CORRELATIONS BASED ON PERMEATION PHENOMENA SOLUBILITY PARAMETER CORRELATIONS

OF

BREAKTHROUGH TIMES

Extensive permeation studies and collections of permeation data are a vailable within the chemical protective clothing industry.10,11 Such data can also be used to establish correlations with HSP . A list of HSP for barrier polymers used in chemical protecti ve clothing has been published 12 based on data by Forsberg and Olsson.10 Some of these correlations have been improved in most instances by correlating the more e xtensive data of F orsberg and Keith.11 The definition of a “good” sol ent which was used for these correlations w as that the breakthrough time w as less than some selected

7248_C013.fm Page 246 Wednesday, May 23, 2007 11:18 AM

246

Hansen Solubility Parameters: A User’s Handbook

TABLE 13.1 HSP Correlations of Breakthrough Times for Barrier Polymers Typically Used in Chemical Protective Clothing. Units are MPa1/2 Type

δD

Neoprene® 16.0 Neoprene 19.0 Butyl 17.0 Viton® 15.6 Nitrile 19.8 Challenge 5100 ® 16.6 PE 16.5

δP

δH

Ro

V Limits

FIT

No.

8.8 8.0 1.5 9.6 13.3 7.0 2.7

4.0 0.0 0.0 7.8 2.2 3.8 6.1

10.1 13.2 7.3 7.1 13.6 2.3 7.9

None 71.0/172 71.0/175.8 72.6/148.9 84.3/177.2 None None

0.574 1.000 0.902 0.896 0.907 0.925 0.969

66 50 86 77 58 116 32

Note: “Good” solvents in these correlations have breakthrough times of less than 1 h.

value, either 20 min, 1 h, or 4 h. Table 13.1 includes some of these impro ved HSP correlations based on a 1-h breakthrough time for commercial film thickness HSP alone cannot always correlate barrier properties unless comparisons are limited to solv ent molecules with approximately the same size (and shape). This, of course, means that the dif fusion coefficients at the reasonably l w concentrations e xpected in better barrier polymers do not v ary too greatly from each other. In many cases, satisfactory correlations could only be found when the differences in HSP between the permeant and the barrier polymer were combined with a size (and shape) parameter(s). The molecular v olume, V, w as found to be a reasonably successful single parameter for this purpose. Printing the correlation data arranged in increasing order of permeant V clearly sho wed whether the molecular size w as important. With regard to the protecti ve ability of the dif ferent garments, it w as found that, in general, and as e xpected, the solv ents with lar ger, more complicated structures required much longer times for breakthrough for a gi ven protective membrane type than comparison with other solv ents would indicate. Outliers were usually these larger molecular species and permeants with smaller or more linear structure, where dif fusion is much more rapid than expected in average comparisons. This size effect is in agreement with what has been known about solvent retention in coatings 2–4,12,13 and what has been discussed pre viously. An e xcellent e xample of this type of impro ved correlation is included in Table 13.1 for the breakthrough times of less than 1 h for neoprene rubber used in chemical protecti ve clothing. The first correlation for this material listed in Table 13.1 gives a very poor data fit (0.574). There were 46 out of 66 liquids which had breakthrough times shorter than 1 h. It is clear from closer analysis of the details of the correlation that the outliers are methanol, carbon disulfide, and al yl alcohol with shorter breakthrough times than predicted and the phthalate plasticizers which ha ve longer breakthrough times than predicted. A perfect fit is found when the molecular olume range of the permeants included in the correlation is abbreviated to between 71 and 172 cc/mol. This eliminates these “outliers.” This correlation is based on 39 liquids with breakthrough times of less than 1 h out a total of 50. The HSP correlations for 1-h breakthrough times for other barrier polymers discussed in Table 13.1 give polymer HSP in the range of those e xpected from their composition. This includes butyl rubber, Viton® (The Du Pont Compan y), nitrile rubber , Challenge ® 5100 (Chemical F abrics Corporation, Merrimack, NH), and polyeth ylene (PE). The thicknesses of all of the films discusse here are those commonly used in chemical protecti ve clothing. HSP correlations of the swelling of Viton are discussed here as well as in Chapter 5 and by Zellers, 14,15 Evans and Hardy,16 and by Nielsen and Hansen. 17

7248_C013.fm Page 247 Wednesday, May 23, 2007 11:18 AM

Applications — Barrier Polymers

247

D, P, H, R = 16.6, 5.4, 4.0, 3.8 FIT = 0.997 FOR 68 < MV < 98

150

BREAKTHROUGH TIME “=” “NO=”

■ BTC ● BCN

NONE

< 3 HR

■

●

> 3 HR

■

●

Evaluation uncertain

x

● MIK

MOLAR VOLUME

BCL

100

MSO ■ ● STY ● CHA ● EAC ■ TCE EET MMA ■ ●● NTB ● TCR ●CHK DEN●● CRB ● ETA● TTE ■ ATC DOUB– EVE ■ ● ACA MAM BNZ ■ LE CRPVAM ■TCL■●MEK● BTR ■ ● ANL● ● DOX MAN ■ THF● PRA BOND ■ALC ● FFA EPC ● PYR ■ MVK ■ CRF●● VOC ■ EDC● ● ACI ALN ● DMF ACB● ● ■ ALM ● ACE POX●■ ACN ■ DCM● ARL MIC● CBB● ● AAC ● AAD ● MBR ALL

50

0.0

1.0

2.0 3.0 RED NUMBER

■ ALA ATN ● ● NME

4.0

5.0

FIGURE 13.2 Graphical method to present HSP correlations. The data are plotted using permeant molar volume vs. RED number. HSP correlation for breakthough times of less than 3 h in Challenge 5100. Symbols used are explained in Table 13.2. (Reprinted from Hansen, C.M. et al.,The Performance of Protective Clothing: Fourth Volume, ASTM STP 1133, McBriarty, J.P. and Henry , N.W., Eds., American Society for Testing and Materials, Philadelphia, 1992, 906. With permission. Cop yright ASTM.)

The RED number (Chapters 1 and 2) is a k ey parameter to judge solv ent quality. This is given in HSP correlations using Chapter 1, Equation 1.10 as Ra (Chapter 1, Equation 1.9), the dif ference in HSP between a solv ent and a polymer , divided by the radius of the correlating sphere, Ro. The radius of the sphere is actually determined as the dif ference in HSP of the “w orst” good solvent(s) and the HSP for the polymer (which is the center point of the sphere). RED numbers near zero indicate v ery good solv ents (rapid breakthrough). RED numbers increase as the solv ent quality decreases. F or RED numbers greater than 1.0, the solv ent quality is considered “bad, ” although swelling may still occur . Figure 13.2 sho ws one w ay to graphically use RED numbers to present data from HSP correlations of permeation phenomena. In this correlation, “good” permeants ha ve breakthrough times of less than 3 h. The data are plotted using V vs. solv ent–polymer af finit , i.e., the RED glass. The number.18 The barrier material, Challenge 5100, is a fluoropolymer supported by fib abbreviations for the permeants in Figure 13.2 are e xplained in Table 13.2. This correlation shows that molecules with molar volumes greater than about 100 cc/mol will not have breakthrough times of less than 3 h, re gardless of RED number. Molecules with molar v olumes greater than about 75 cc/mol require a terminal double bond and lo wer RED numbers to breakthrough under these conditions. Molecules with still lo wer molar v olumes appear to come through with only a slight dependence on the RED number . The effect with the terminal double bonds clearly indicates the preferential direction of motion for this type of molecule. The molecules in effect worm their way through the barrier polymer .

7248_C013.fm Page 248 Wednesday, May 23, 2007 11:18 AM

248

Hansen Solubility Parameters: A User’s Handbook

TABLE 13.2 List of Symbols Used in Figure 13.2 Symbol AAC AAD ACA ACB ACE ACI ACN ALA ALC ALM ALN ANL ARL ATC ATN BCM BCN BNZ BTC BTR BUT BZN CAC CBB CBT CCF CHA CHK CLA CLB CRB CRF CRP DCM DEN DMF DOX DSO EAC

Compound

Symbol

Acetic acid Acetaldehyde Acetic anhydride Acetyl bromide Acetyl chloride Acetone Acrylonitrile Allyl alcohol Allyl chloride Allyl amine Allyl cyanide Aniline Acrolein Allyl isothiocyanate Acetonitrile Bromochloromethane Butyl acetate Benzene Butyl acrylate Butyraldehyde Butane Benzonitrile Chloroacetylchloride Carbon disulfid Carbon tetrachloride Dichloromonofluoromethane (Freon 21 Cyclohexylamine Cyclohexanone Chloroacetone 1-Chlorobutane Chlorobenzene Chloroform Chloroprene Dichloromethane Diethylamine Dimethyl formamide 1,4-Dioxane Dimethyl sulfoxide Ethyl acrylate

EBR EDC EET EIM EPC ESH ETA EVE EVK F12 FFA HXA MAL MAM MAN MAT MBR MEK MES MIC MIK MMA MSO MVK NEE NME NTB POX PRA PYR STY TCE TCL TCR THF TOL TTE VAM VDC

Compound Ethyl bromide Ethylene dichloride Diethyl ether Ethyleneimine Epichlorohydrin Ethanethiol Ethyl acetate Ethyl vinyl ether Ethyl vinyl ketone Dichlorodifluoromethane (Freon 12 Furfural Hexane Methanol Methyl acrylate Methacrylonitrile Methyl acetate Methyl bromide Methyl ethyl ketone Methyl sulfid Methyl isocyanate Methyl isobutyl ketone Methyl methacrylate Mesityl oxide Methyl vinyl ketone Nitroethane Nitromethane Nitrobenzene Propylene oxide Propylamine Pyridine Styrene 1,1,2,2,-Tetrachloroethylene Trichloroethylene 1,1,1-Trichloroethane Tetrahydrofuran Toluene Tetrachloroethylene Vinyl acetate 1,1-Dichloroethylene

Source: Reprinted from Hansen, C.M. et al., The Performance of Protective Clothing: Fourth Volume, ASTM STP 1133, McBriarty, J.P. and Henry , N.W., Eds., American Society for Testing and Materials, Philadelphia, 1992, 903. With permission. Cop yright ASTM.

SOLUBILITY PARAMETER CORRELATION

OF

PERMEATION RATES

Permeation rates for dif ferent permeants in a polymer can also be correlated to find HSP for th polymer. This is done by di viding a data set into tw o groups. The “good” solv ents will ha ve permeation rates greater than an arbitrarily selected v alue, and the “bad” solv ents will ha ve

7248_C013.fm Page 249 Wednesday, May 23, 2007 11:18 AM

Applications — Barrier Polymers

249

TABLE 13.3 HSP Correlations Related to Barrier Polymers Material

δD

δP

δH

Ro

FIT

G/T

LDPE permeation coefficient 21.1° Permeation viable skin a PP swelling b ACLAR® 22 >5 wt% swelling ACLAR® 22 2%<swelling<5% Psoriasis scales swelling (Chap. 14) Viton® swell c >10 wt% 20°C EVOH sol/swell d Polyvinyl chloride swell e Cellophane — >25% swell PETP chemical resistance (+/– OK) f PA6/PA66 chemical resistance (+/– OK) g PA6/PA66 chemical resistance (+/– bad) g

16.3 17.6 18.0 14.7 18.0 24.6 13.1 20.5 18.7 16.1 18.2 17.0 18.9

5.9 12.5 3.0 3.9 1.0 11.9 13.7 10.5 9.7 18.5 6.4 3.4 7.9

4.1 11.0 3.0 6.7 2.0 12.9 3.9 12.3 7.7 14.5 6.6 10.6 11.7

8.2 5.0 8.0 6.8 4.0 19.0 14.7 7.3 6.4 9.3 5.0 5.1 8.0

1.000 1.000 1.000 1.000 1.000 0.927 0.742 0.925 1.000 0.955 0.865 0.984 0.950

26/47 4/13 13/21 6/26 4/21 35/50 36/57 5/24 13/47 4/22 7/34 2/31 9/31

Note: Data fit and the number of good liquids (G) and total number of liquids (T) in th correlations are also indicated. Units are MP a1/2. This correlation is discussed in more detail in Chapter 15 and is based on limited data. 20 This correlation is based on data by Lieberman and Barbe 22 and is discussed in more detail in Chapter 5. c This correlation is discussed in Chapter 5. Swelling data is from Reference 23. d Ethylene vin yl alcohol copolymer (EV OH), four liquids dissolv ed and one (morpholine) swelled very strongly. e Visual observation of v ery strong swelling and/or solubility . f Polyethylene terephthalate (PETP) chemical resistance based on rather uncertain data24 (see discussion in Chapter 12). Recommendation of uncertain-for -use is considered as acceptable for use. Attacking solvents are within correlating HSP sphere. g Polyamide 6/66 chemical resistance based on rather uncertain data 24 (see discussion in Chapter 12). Recommendation of uncertain-for -use is used as indicated. Attacking solvents are within the correlating HSP sphere. a

b

Source: From Hansen, C.M., Permeability of polymers, Pharmaceutical and Medical Packaging 98, 7.6, 1998. With permission.

permeation rates lo wer than this v alue. Such a correlation based on permeation coef ficients fo various liquids in PE is included in Table 13.3. The permeation coef ficient data, (g x mm)/( 2 x d), are reported by P auly19 for lo w density polyeth ylene (LDPE). “Good” solv ents are arbitrarily considered as those which ha ve permeation coef ficients in these units which are greater than 1. at 21.1°C. The parameters reported correlate the data well b ut are some what different from those which might be expected for a polyolefin. Reasons for this are not vident but may include additives in the polymer , local oxidation, or some other local v ariation in the composition of the polymer . It should be remembered that permeation occurs in the amorphous re gions only. This is why high density PE is a better barrier polymer than lo w density PE; the higher densities are attrib utable to a higher percentage of crystallinity . A problem of some concern is the permeation through b uried water pipes by chemicals or oil products which somehow reach them, either by general pollution or by g asoline or oil spills. One clearly expects more extensive permeation by chemicals that have HSP not too different from those of the polymer from which the pipe is made, all other things being equal. These pipes are often made from polyolefins

7248_C013.fm Page 250 Wednesday, May 23, 2007 11:18 AM

250

Hansen Solubility Parameters: A User’s Handbook

A HSP correlation has been possible in a v ery special case of polymer permeability where the barrier polymer is viable human skin. 20 This is discussed in more detail in Chapter 15. Human skin is a polymeric barrier with se veral functions, one of which is to help k eep undesirable chemicals out of the body . Some chemicals readily permeate this boundary , and this f act has been used to establish a tentative HSP correlation for the permeation rate of viable human skin. This correlation 21 which is also also has a relation to the HSP correlation for the swelling of psoriasis scales, discussed in Chapter 15.

SOLUBILITY PARAMETER CORRELATION OF POLYMER SWELLING Solubility is a major f actor in the equation P = DS, so correlations of solv ent uptake in polymers are important to understand their barrier properties. The correlation for swelling of polyprop ylene reported in Table 13.3 is based on solv ent uptak e data reported by Lieberman and Barbe. 22 The limit of 0.5% weight g ain was arbitrarily set to dif ferentiate “good” solv ents from “bad” ones. A different limit might gi ve different parameters. The HSP found in a gi ven correlation of swelling depends on which polymer se gments the smaller amounts of permeant prefer to associate with. The predictive ability of the correlation will depend on the number of test liquids used in the study and their gi ven HSP v alues. How different are the HSP of the test liquids? What are their v alues compared with the predictions desired? The parameters reported in Table 13.3 for polyprop ylene seem to accurately reflect what is xpected in terms of lo w polarity and lo w h ydrogen bonding properties for this type of polymer . As stated pre viously, a problem of some significance in a y study of solv ents at lo w concentrations in polymers is that the smaller amounts of solv ent relative to the polymer can lead to preferential association of solv ent with those local re gions/segments/groups in the polymer that have energies (HSP) most similar to their own. Like seeks like. These local regions may not necessarily reflect the same a finities as the polymer as a whole, such as are indicate by the totally soluble-or -not approach most commonly used in HSP e valuations. These local association effects can influence results on swelling studies at l w solvent uptakes in both good and bad solv ents, for e xample. Copolymers, such as Viton, are particularly susceptible to this problem. Swelling data for Viton23 are correlated by the HSP v alues included in Table 13.3. A poor data fit can be anticipated when a single HSP sphere is used to describe what should b represented by (at least) tw o overlapping HSP spheres (see also Figure 13.3). Zellers 14,15 also had difficulty correlating the swelling of Viton. Other types of studies carried out at low solvent concentrations can also be influenced by this s gregation/association phenomena. An extension of this type of situation can be cited in the tendencies of w ater to associate with itself as well as with local hydrophilic regions within polymers. The amount of water taken up at equilibrium is not reflected by an verall HSP correlation of polymer solubility or swelling. As little as 1% of hydrophilic additive can ef fectively destroy the w ater barrier properties of a polymer film but this small amount cannot be measured in swelling or solubility studies leading to HSP correlations. This f act, among other things, has made simple predictions of the beha vior of water very difficult Correlations of polymer solubility and swelling ha ve led to se veral of the HSP data sets reported in Table 13.3 (see also the data reported in Chapter 5).The HSP correlations for chemical resistance based on data of the acceptable-for -use or recommended/not recommended type are not as reliable as those usually found for solubility and swelling where a suitably lar ge number of liquids are used in the testing. The reasons for this are discussed in depth in Chapter 12. The data used in the chemical resistance correlations reported in Table 8.3 were tak en from Reference 24.

7248_C013.fm Page 251 Wednesday, May 23, 2007 11:18 AM

Applications — Barrier Polymers

251

ACLAR®22 Bimodal Uptake HSP Correlations Uptake

δD

δP

δH

R

5 - 10 % 2-5% > 10 %

14.7 3.9 6.7 6.8 18.0 1.0 2.0 4.0 Trichloroethylene within both

FIT 1.000 1.000

● ■ ✕

12

δP, Polar Parameter

10

8

6

Ro = 4

6.8

(14.7)

2

(18.0)

Ro = 4.0

0 0

2

4

6

8

10

12

14

δH, Hydrogen Bonding Parameter FIGURE 13.3 Bimodal HSP correlation(s) for uptak e of liquids in ACLAR® 22. Trichloroethylene uptake is the lar gest among the test solv ents because it is the only solv ent found within both re gions. It has RED numbers of 0.999 for the >5% correlation and 0.978 for the correlation of uptak e between 2 and 5%. As the data fit is 1.0 for both correlations, other sets of parameters can also gve data fits of 1.0. H wever, the numbers are approximately correct. Units are MP a1/2.

SOLUBILITY PARAMETER CORRELATION OF PERMEATION COEFFICIENTS FOR GASES Gases can also be assigned δD, δP, and δH parameters. For strictly nonpolar g ases, the values of δP and δH will be zero, b ut other g ases, such as carbon dioxide, h ydrogen sulfide, etc., will h ve significant alues for all three parameters. Table 13.4 gi ves the δD, δP, and δH parameters for a number of gases. It is not surprising that there are HSP correlations of permeation coef ficients fo gases in dif ferent polymers as a function of their solubility parameter dif ferences. One such correlation using the total solubility parameter has been given by König and Schuch,25 who showed

7248_C013.fm Page 252 Wednesday, May 23, 2007 11:18 AM

252

Hansen Solubility Parameters: A User’s Handbook

TABLE 13.4 HSP for Common Gases of Interest in Permeation Phenomena Gas

δD

δP

δH

Water Ammonia Chlorine Sulfur dioxide Carbon dioxide a Carbon monoxide Ethane Ethylene Helium Hydrogen Hydrogen sulfid Methane Nitrogen oxide Nitrogen Nitrous oxide Oxygen Acetylene

15.5 13.7 17.3 15.8 15.7 11.5 15.6 15.0 1.0 5.1 17.0 14.0 11.5 11.9 12.0 14.7 14.4

16.0 15.7 10.0 8.4 6.3 4.9 0 2.7 0 0 6.0 0 20.0 0 17.0 0 4.2

42.3 17.8 0.0 10.0 5.7 0 0 2.7 0 0 10.2 0 0 0 0 0 11.9

Note: Units are MP a1/2. Values changed from 1st Edition. See Chapter 10 Addendum. a

that the better barrier polymers for oxygen, i.e., those with lo w oxygen permeation coef ficients were those whose solubility parameters were most dif ferent from the solubility parameters of oxygen. The better barrier polymers for oxygen include polyacrylonitrile and polyvin yl alcohol, whereas the poorer barriers include polyolefins and polytetrafluoroe ylene (PTFE). The amounts of gases dissolved at low pressures are usually low, and constant diffusion coefficients are xpected. This may not be true at higher pressures where solubility parameters for the g ases increase more rapidly than those of the polymers and polymers can absorb them to a greaterxtent. e See Chapter 10. An example of ho w HSP principles can be applied to interpreting the beha vior of g as barrier films can be found in the performance of poly(chlorotrifluoroe ylene). The data on which the example is based are tak en from the commercial literature supplied by Allied Signal concerning their barrier films under the tradename of ACLAR®.26 These films are xcellent barriers for w ater and oxygen, and various laminating possibilities e xist, including polyethylene, polyvinyl chloride, and polyeth ylene terephthalate. The barrier properties of films made from this material are no nearly as good for carbon dioxide as the y are for nitrogen or oxygen. A contributing factor in this is that the HSP of the polymer is some what different from the HSP of oxygen and nitrogen, b ut close to the HSP of carbon dioxide. A HSP correlation for the swelling of ACLAR 22 to greater than 5 wt% is included in Table 13.3. The RED numbers for w ater, nitrogen, oxygen, and carbon dioxide based on this correlation are 5.5, 1.4, 1.1, and 0.48, respecti vely. Nitrogen has slo wer permeation than oxygen, and both are much slower than carbon dioxide, in general agreement with this ranking. One might ha ve expected the permeation rate of carbon dioxide to be lo wer than that of nitrogen and oxygen as it is a lar ger molecule, but the enhanced solubility of carbon dioxide in the polymer o verrides this e xpectation.

7248_C013.fm Page 253 Wednesday, May 23, 2007 11:18 AM

Applications — Barrier Polymers

253

Figure 13.3 sho ws that there are tw o distinct spherical characterizations possible for ACLAR 22. The first of these, as discussed earlie , is for liquid uptak e to greater than 5% by weight. The second of these is for uptake between 2 and 5% by weight. This second correlation is also reported in Table 13.3. There is one liquid in the data which is common to both of the HSP re gions pictured in Figure 13.3. This is trichloroeth ylene (which w as assumed to be a good solv ent in both of the perfect correlations in spite of being absorbed to o ver 10% by weight). Ev en though trichloroethylene has high RED numbers in both correlations, this solv ent is absorbed more than an y of the other solvents tested because of this property . The primary uptak e region has HSP that might be expected from a fluoropolyme , whereas the secondary HSP region is what might be expected from a chlorinated species. Such secondary re gions can potentially allo w higher permeation rates and greater absorption of unpredictable materials based on a single HSP correlation. Searching a database of solv ents, plasticizers, aromatic compounds, etc., w ould sho w which of these could behave in an une xpected manner. Sometimes an indirect approach allo ws prediction of the uptak e of a g as in a polymer . This involves determining the uptake of the gas in a liquid ha ving solubility parameters that are similar to those of the polymer . This approach e xpands the usefulness of g as–liquid equilibrium data. Correlations of g as–liquid solubility with the solubility parameter are included in Figure 13.4 for the equilibium values for water27 and in Figure 13.5 for the equilibrium v alues for nitrogen. 28 The quantity P*y/x is gi ven by the total pressure, P*, the mole fraction in the g as phase, y , and the mole fraction in the liquid phase, x. The abbre viations used in Figure 13.5 are e xplained in Table 13.5. The solubility parameters for g ases not found in Table 13.4 may be found in standard references.29–31 HSP for gases can be calculated using the procedures outlined in Chapter 1 with the special figure for ases included in Chapter 18 (Figure 18.2). There can be problems of dividing the total cohesive energy into three parts. Sulfur trioxide is a good e xample of how one cannot come further in dividing the energy of vaporization into components without experimental data. The techniques of Chapter 3 ha ve not been e xplored in this conte xt ho wever. The total (Hildebrand) solubility parameter , δt, indicated from the total ener gy of v aporization is 31.3 MPa1/2. δD found by the usual techniques is 15.6 MP a1/2. This leaves a residual corresponding to a solubility parameter v alue of 27.2 MP a1/2 to be distrib uted between permanent dipole and hydrogen bonding (electron transfer) ef fects. There is no dipole moment, and neither is there a hydrogen atom. This clearly requires e xperimental data to resolv e the distribution of the ener gy of vaporization into components for all of these ef fects, even if there are supporting estimates from the techniques of Chapter 3. It might be noted that the scale in Figure 13.4 for the uptak e of w ater in v arious liquids is exponential with data covering almost fi e decades in concentration. The phenomena correlated in this figure confirm the xpectation that nonpolar polymers, with solubility parameters f ar different from those of w ater, will be good barrier polymers for w ater because of low water solubility. This is generally true, of course. As mentioned earlier, such polymers include the polyolefins as well a chlorinated and fluorinated polymers. These comments and generalities are not necessarily v alid for polymers containing additives. Depending on the nature of the additive and the amounts present, some of these can totally change the barrier performance of the base polymer .

LAMINATES Laminated barrier polymer systems are designed to mak e the best use of the properties of each of the individual layers, as well as to optimize cost with performance. The most common type has a polyolefin on the xterior surfaces to protect the inner barrier polymer from w ater. These interior barrier polymers often ha ve relatively high solubility parameters, such as eth ylene vinyl alcohol copolymers (EVOH), polyamides (P A), or polyeth ylene terephthalate (PET). If the inner barrier polymer takes up water, it will be plasticized, and its barrier properties will be reduced.A polyolefi

7248_C013.fm Page 254 Wednesday, May 23, 2007 11:18 AM

254

Hansen Solubility Parameters: A User’s Handbook

D P H 1 % IN WATER 15.1 20.4 16.5 H from Perry: “CHEMICAL ENGINEERS’ HANDBOOK” (1963). 106

H6

105

N2  CO

H2S  NO

C2H4

104

H = P+ y/x

H2

O

 2 CH4  C2H6



 N2O  CO2

 C2H2

103

 C12

102  SO2

101

 NM3

100 0

10

20

30

40

DISTANCE TO “1% IN WATER” MPa½ FIGURE 13.4 HSP correlation of gas–water equilibrium data using water HSP values found with a correlation using a limit of >1% liquid soluble in water as a “good” solvent,21 and gas–water equilibrium data from Perry et al. 27 P* is the total pressure, y is the mole fraction in the g as phase, and x is the mole fraction in the liquid phase. (From Hansen, C.M., Dan. Kemi, 73(8), 21, 1992. With permission.)

laminated to such a potentially w ater-sensitive barrier film can significantly delay this upt e and loss of barrier properties and maintain reasonable costs. Depending on the performance desired, various combinations of laminates can be systematically designed using HSP considerations as one of the design parameters.

7248_C013.fm Page 255 Wednesday, May 23, 2007 11:18 AM

Applications — Barrier Polymers

255

5000 NITROGEN

D = 11.9, P = 0.0 H = 0.0, RAD. = 1.0

4000

MTA

3000

t

P y⁄x

ETA

2000 BEN ETY

MTC HEP

1000

PEN

ISA

DEC HEX

PTF PFM PFH

0 0

1

RED ⁄ 10

2

3

FIGURE 13.5 HSP correlation of nitrogen–liquid equilibrium data at temperatures near 25°C and lo w pressure. P* is the total pressure, y is the mole fraction in the g as phase, and x is the mole fraction in the liquid phase. (From Hansen, C.M., Dan. Kemi, 73(8), 20, 1992. With permission.)

GENERAL CONSIDERATIONS HSP correlations have been possible for many phenomena where differences in behavior on contact with different solvents have been studied. The HSP correlations are preferably based on systems in thermodynamic equilibrium, although the correlations presented pre viously on breakthrough times are an e xception to this. These correlations were possible because of the e xceptional dependence of the permeation phenomena on the amount of permeant being dissolv ed. There is a strong dependence of the dif fusion coefficient of permeants in polymers on thei size and shape. This can clearly affect HSP correlations of permeation coefficients, as t o permeants with identical HSP will ha ve different D if their sizes and shapes are significantly di ferent. This

7248_C013.fm Page 256 Wednesday, May 23, 2007 11:18 AM

256

Hansen Solubility Parameters: A User’s Handbook

TABLE 13.5 Key to Symbols Used in Figure 13.5 Symbol PFH PTF PFM PEN HEX HEP MYC ISA ETY BEN ETA MTA

Compound Perfluoroheptan Perfluorotri utylamine Perfluoromet ylcyclohexane Pentane Hexane Heptane Methyl cyclohexane Isobutyl acetate Ethyl acetate Benzene Ethanol Methanol

differential in diffusion rate based on solv ent size and shape can also gi ve apparent errors in HSP correlations of polymers for their chemical resistance, for e xample, where not enough e xposure time has been allowed for attainment of equilibrium. This is clearly a problem in the determination of equilibrium de gree of swelling and lo w amounts of uptak e. This problem has also been found, for example, for exposures of thick samples (3 to 4 mm) of rigid polymers used for tensile testing after solvent exposure for given times. Crystalline polymers also have a tendency to be more readily soluble in solvents with lower V, all other parameters being equal, b ut this is explained by thermodynamic considerations rather than a relati vely faster diffusion process. In all of these cases, the majority of the outliers in the correlations are the test liquids with higher V. The time required for attainment of equilibrium with the lar ger diffusing molecules can be so long as to be prohibiti ve for their reasonable inclusion in HSP correlations. It is suggested that dif fusion rates be carefully considered when liquids with v ery high V are outliers in HSP correlations. Ne glecting such data for the sak e of a correlation can be justified, ut this is a w arning that the dif fusion process for these liquids has not yet achie ved equilibrium and that the ef fects of such liquids can be e xpected to be more se vere at still longer times than those used in the study . HSP correlations can be used in this way to find those xposure liquids which have not reached equilibrium at the exposure time chosen for evaluations.

CONCLUSION Successful HSP correlations for the permeation and solubility behavior of selected barrier polymers have been presented to demonstrate the use of simple principles to arri ve at optimum barrier systems, as well as to determine reliable HSP for the polymers studied. Selection of polymer–permeant combinations with widely different solubility parameters will ensure low solubility of the permeant in the polymer. This reduces the concentration gradient and pre vents significant self-plasticizatio of the polymer . The self-plasticization leads to concentration-dependent dif fusion coefficients, a effect which becomes more significant with increasing amounts of permeant being dissol ed, i.e., a closer HSP match. See Chapter 16 for more discussion of dif fusion in polymers. HSP correlations have been presented for breakthrough times in chemical protecti ve clothing, permeation rates in barrier polymers, and swelling of v arious types of polymers. Both g ases and liquids are treated.

7248_C013.fm Page 257 Wednesday, May 23, 2007 11:18 AM

Applications — Barrier Polymers

257

REFERENCES 1. Rogers, C.E., Permeation of gases and vapours in polymers, in Polymer Permeability, Comyn, J., Ed., Elsevier Applied Science, London, 1985, pp. 11–73. 2. Hansen, C.M., Some aspects of the retention of solvents in high polymer films, Färg och Lack, 10(7), 169–186, 1964. 3. Hansen, C.M., The Three Dimensional Solubility P arameter and Solvent Diffusion Coefficient, Doc toral dissertation, Danish Technical Press, Copenhagen, 1967. 4. Hansen, C.M., A mathematical description of film drying by sol ent evaporation, J. Oil Colour Chem. Assoc., 51(1), 27–43, 1968. 5. Hansen, C.M., Dif fusion in polymers, Polym. Eng. Sci., 20(4), 252–258, 1980. 6. Hansen, C.M., The measurement of concentration-dependent diffusion coefficients — the xponential case, Ind. Eng. Chem. Fundam., 6(4), 609–614, 1967. 7. Hansen, C.M., Dif fusion coef ficient measurements by sol ent absorption in concentrated polymer solutions, J. Appl. Polym. Sci., 26, 3311–3315, 1981. 8. Crank, J., The Mathematics of Diffusion, Oxford University Press, Oxford, 1956. 9. Klopfer, H., Water Transport by Diffusion in Solid Materials (Wassertransport durch Diffusion in Feststoffen, in German), Bauv erlag GMBH, Wiesbaden, 1974. 10. Forsberg, K. and Olsson, K.G., Guidelines for Selecting Chemical Protective Gloves, (Riktlinjer för val av kemiskyddshandskar, in Swedish), Förening Teknisk Företagshälso vård (FTF), Stockholm, 1985. 11. Forsberg, K. and Kieth, L.H., Chemical Protective Clothing Performance Index, 4th ed., Instant Reference Sources, Austin, TX, 1991. 12. Hansen, C.M. and Hansen, K.M., Solubility parameter prediction of the barrier properties of chemical protective clothing, Performance of Protective Clothing: Second Symposium. ASTM STP 989, Mansdorf, S.Z., Sager, R., and Nielsen,A.P., Eds., American Society for Testing and Materials, Philadelphia, PA, 1988, pp. 197–208. 13. Hansen, C.M., The free v olume interpretation of plasticizing ef fectiveness and dif fusion in high polymers, Off. Dig., 37(480), 57–77, 1965. 14. Zellers, E.T., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. I. Modelling the solubility of organic solvents in Viton gloves, J. Appl. Polym. Sci., 50, 513–530, 1993. 15. Zellers, E.T., Three-dimensional solubility parameters and chemical protective clothing. II. Modelling diffusion coefficients, breakthrough times, and steady-state permeation rates of o ganic solv ents in Viton gloves, J. Appl. Polym. Sci., 50, 531–540, 1993. 16. Evans, K.M. and Hardy , J.K., Predicting solubility and permeation properties of or ganic solvents in Viton Glove material using Hansen’s solubility parameters, J. Appl. Polym. Sci., 93, 2688–2698, 2004. 17. Nielsen, T.B. and Hansen, C.M., Elastomer swelling and Hansen solubility parameters, Polym. Testing, 24, 1054–1061, 2005. 18. Hansen, C.M., Billing, C.B., Jr., and Bentz, A.P., Selection and use of molecular parameters to predict permeation through fluoropolyme -based protective clothing materials, in The Performance of Protective Clothing: Fourth Volume, ASTM STP 1133, McBriarty, J.P. and Henry, N.W., Eds., American Society for Testing and Materials, Philadelphia, P A, 1992, pp. 894–907. 19. Pauly, S., Permeability and diffusion data, in Polymer Handbook, 3rd ed., Branderup, J. and Immergut, E.H., Eds., Wiley-Interscience, New York, 1989, pp. VI/445–446. 20. Ursin, C., Hansen, C.M., Van Dyk, J.W., Jensen, P.O., Christensen, I.J., and Ebbehoej, J., Permeability of commercial solv ents through living human skin, Am. Ind. Hyg. Assoc. J., 56, 651–660, 1995. 21. Hansen, C.M. and Andersen, B.H., The affinities of o ganic solvents in biological systems, Am. Ind. Hyg. Assoc. J., 49(6), 301–308, 1988. 22. Lieberman, R.B. and Barbe, P .C., Polypropylene polymers, in Encyclopedia of Polymer Science and Engineering, 2nd ed., Vol. 13, Mark, H.F ., Bikales, N.M., Ov erberger, C.G., Menges, G., and Kroschwitz, J.I., Eds., Wiley-Interscience, New York, 1988, pp. 482–483. 23. Anonymous, Fluid Resistance of Viton®, Du Pont Compan y, Polymer Products Department, Elastomers Division, Wilmington, DE, 1989. 24. Anonymous, Plastguide, SCS Dukadan AS, Randers, Denmark, 1990.

7248_C013.fm Page 258 Wednesday, May 23, 2007 11:18 AM

258

Hansen Solubility Parameters: A User’s Handbook

25. König, U. and Schuch, H., Structure and permeability of polymers (K onstitution und Permeabilität von Kunststoffen, in German), Kunststoffe, 67(1), 27–31, 1977. 26. Anonymous, ACLAR® Barrier Films, AlliedSignal — Advanced Materials, Allied Signal Inc., Mor ristown, NJ. 27. Perry, J.H., Chilton, C.H., and Kirkpatrick, S.D., Eds., Chemical Engineers’ Handbook, 4th ed., McGraw-Hill, New York, 1963, Section 14, pp. 2–7. 28. Hansen, C.M., 25 years with solubility parameters (25 År med Opløselighedsparametrene, in Danish), Dan. Kemi, 73(8), 18–22, 1992. 29. Hildebrand, J. and Scott, R.L., The Solubility of Nonelectrolytes, 3rd ed., Reinhold, Ne w York, 1950. 30. Hildebrand, J. and Scott, R.L., Regular Solutions, Prentice-Hall Inc., Engle wood Cliffs, NJ, 1962. 31. Barton, A.F.M., Handbook of Polymer-Liquid Interaction Parameters and Solubility Parameters, CRC Press, Boca Raton, FL, 1990.

7248_C014.fm Page 259 Wednesday, May 23, 2007 11:24 AM

— Environmental 14 Applications Stress Cracking in Polymers Charles M. Hansen ABSTRACT Hansen solubility parameters (HSP) can be used to help predict which chemicals can cause environmental stress cracking (ESC) in polymers. ESC requires tensile stress and correlates when the RED number (relati ve energy difference in the polymer -solvent interaction) found from HSP considerations is plotted vs. a molecular size parameter , the molar v olume, V. There are three distinct regions on this plot. There is a re gion at lo w RED including those challenge liquids that dissolve the polymer or are v ery aggressive, and ESC is not found as such. There is a re gion at high RED where the absorption is zero or not great enough to matter or else the absorption rate is slow enough to allo w relaxation of the polymer in preference to ESC. ESC can occur in an intermediate re gion where there is some absorption of challenge liquid, although e xamples are given where ESC takes place without measurable absorption for good matches in HSP at relatively high stress/strain. The ESC re gion on these plots increases in size with increased tensile stress and/or increased critical strain.

INTRODUCTION The pre vious chapter dealt with the chemical resistance of polymers and briefly touched o environmental stress cracking (ESC) as one aspect of chemical attack. This chapter e xpands the previous discussion of this special type of ph ysical chemical attack on polymers. This form of failure represents at least 25% of all f ailures in plastics and therefore deserv es special attention. 1 A failure by ESC can appear almost immediately , after minutes, after hours, or e ven after years, and often occurs without prior w arning. There is a considerable literature on ESC. An excellent general source of the ESC literature is Wright’s encompassing book on f ailures in polymers in general.1 The present chapter is in man y ways a supplement to this e xcellent work. It is no w clear that the polymer and the chemical that initiates the stress cracking must ha ve similar or reasonably similar HSP . The ESC initiator need not dissolv e the polymer as such. It has generally been assumed that some similarity in HSP is required such that some absorption occurs. More recently it has been recognized that measurable absorption is not al ways necessary for ESC to occur in some polymers. It is theorized that some physical movement, such as rotation of the polymer chain segments at the contact surface, can initiate the cracking process. In the latter case the similarity of the HSP of the challenge liquid may be to the HSP of an entity in the polymer chain that otherwise might be oriented away from the polymer surf ace. Under given conditions of strain it may prefer the environment of the ESC initiator once there is contact with the polymer . An understanding of this phenomenon seems to be e volving, but it is not complete as discussed belo w. As stated above, the generally accepted mechanism for ESC has been that some absorption of the active chemical weakens the polymer structure locally , such that the tensile stress increases in adjacent regions. The tensile stress must be suf ficient to locally pull polymer chains or s gments of chains from the bulk. As the structure weakens locally, there is added stress in adjacent re gions. This may be suf ficient to cause a craze or crack, ut in man y cases further absorption or further chemical penetration seems necessary to repeat the same process until finally the stress become 259

7248_C014.fm Page 260 Wednesday, May 23, 2007 11:24 AM

260

Hansen Solubility Parameters: A User’s Handbook

too high and a craze or crack occurs. The number of chemicals gi ving ESC in a gi ven polymer increases with increases in the le vel of stress/strain. The critical strain is that minimum v alue of strain at which the given challenge chemical will cause crazing/cracking in the given polymer. Any strain le vels abo ve this will result in ESC on contact with the gi ven chemical. In some cases absorption of liquid permits stress relaxation, and e xpected crazing and cracking does not occur . This is presumably for a more or less massi ve uptake but not enough to completely dissolv e the polymer. On the other hand, the polymer is plasticized and/or weak ened by the absorption of these chemicals, and there is potentially a dif ferent type of problem to cope with. If a chemical reaction is in volved, such as with acids and bases, then the phenomenon is more properly called stress corrosion cracking (SCC). Wright indicates that about 90% of ESC f ailures involve amorphous thermoplastic polymers with the remaining being found with partly crystalline thermoplastic polymers. Lustiger 2 lucidly describes the mechanism of ESC in partly crystalline polymers with polyeth ylene as an e xample. There are three types of polymer chains to consider in this case. Those with free ends (cilia) extending into the amorphous phase, those with chains e xtending into the amorphous phase b ut that loop back into the same lamella (loose loops), and tie-molecules or chains that e xtend from one lamella and anchor into an adjacent one. It is such tie-molecules that gi ve the ultimate resistance to ESC. One must either break the tie molecules or pull them out of one of the lamella. In a sense these act lik e strong ph ysical cross-links. Similar considerations are v alid for amorphous thermoplastic polymers. Higher molecular weight enhances polymer chain entanglements, and thus reduces the tendency for ESC when solv ent is absorbed. There are balances in properties and processing ability that are required for both amorphous thermoplastic polymers as well as partly crystalline polymers in order to achie ve maximum ESC resistance and still maintain other desired beha vior. It is beyond the scope of this chapter to discuss models for the fracture mechanics mechanisms of ESC. Emphasis is placed on the consequences of similarity of HSP for the polymer and challenge chemical.

ESC INTERPRETED USING HSP Wright1 has discussed ESC data in terms of HSP for polycarbonate (PC), polyvinylchloride (PVC), polymethyl methacrylate (PMMA), and polystyrene (PS). ESC data are discussed for man y other polymers including polyeth ylene (PE), polyamides (P A), polyether ether k etone (PEEK), polytetrafluoroet ylene (PTFE), and styrene-acrylonitrile copolymer (SAN). There are man y practical examples of how not to do things. Barton3 has discussed in detail the theory and application of the cohesion (solubility) parameters including HSP for man y systems, as witnessed by the 739 pages in his book and hundreds of references. En vironment-induced de gradation is also discussed with v arious plots for poly(2,6dimethyl-1,4-phenylene oxide), PS, PMMA, PVC, and polysulfone. The latter three are found originally in Vincent and Raha4. Wyzgoski5 constructed several plots to help interpret ESC data for nylon (PA) 6,6 using HSP. Hansen and Just 6 studied ESC in the COC-type polymer called Topas® 6013 from Ticona. Figure 14.1, sho ws two essentially concentric HSP spheres resulting from this work. The inner sphere encompasses those solv ents that dissolv e the polymer , whereas the outer sphere encompasses these as well as those gi ving ESC for the gi ven samples. Additional HSP correlations for ESC in PET , PCTG, and PC were reported for ESC (critical strain) data from Moskala and Jones 7. HSP correlations for ESC in PEI using data from 8 were also reported. All of these HSP correlations are included in Table 14.1. The HSP correlations for ESC reported in Table 14.1 must be considered with care. In the firs place there are other factors than HSP that are important, including the size and shape of the gi ven challenge molecule in addition to the state of stress/strain, which is not al ways well defined.Another point to be remembered is that the RED number (Equation 1.10) refers to the correlation in question. The same solv ent polymer pair will ha ve dif ferent RED numbers for a HSP correlation of true

7248_C014.fm Page 261 Wednesday, May 23, 2007 11:24 AM

Applications — Environmental Stress Cracking in Polymers

Dispersion Solubility Parameter, δD

261

40

30

20 15

15 10

10 Polar 5 Solubility Parameter, δP

10 Hydrogen Bonding Solubility Parameter, δH

5 0

0

FIGURE 14.1 Three-dimensional HSP plot sho wing those solv ents that dissolv e the COC polymer , Topas 6013, Ticona, in the shaded re gion, and those that induce ESC in the clear shell. The two HSP spheres are almost concentric as can be seen from the data in Table 14.1. (Reprinted with permission from Hansen, C.M. and Just, L., Ind. Eng. Chem. Res., 40(1), 21–25, 2001. Cop yright 2001 American Chemical Society.)

TABLE 14.1 HSP Correlations for ESC in Polymers Polymer

δD

δP

δH

R0

FIT

G/Ta

Topas 6013 Solubility Topas 6013 Sol. + cracks PC critical strain <0.6% b PVC crit.str. <0.6% c PET crit. str. <0.6% PCTG crit. str. <0.55% PCTG crit. str. <0.50% PC crit. str. <0.31% PEI Ultem ® 1000 600 psi PEI Ultem 1000 1200 psi PEI Ultem 1000 2500 psi PEI Ultem 1000 solubility

18.0 17.3 21.5 16.0 21.3 18.3 18.8 18.0 17.3 17.0 17.4 19.6

3.0 3.1 9.5 10.0 4.5 9.3 8.8 9.0 5.3 6.0 4.6 7.6

2.0 2.1 5.1 5.0 12.3 11.3 10.8 6.0 4.7 4.0 9.0 9.0

5.0 6.4 12.9 10.7 13.9 8.0 7.9 10.0 3.3 4.0 7.2 6.0

1.000 0.974 0.857 1.000 1.000 1.000 0.989 1.000 1.000 1.000 0.967 0.952

8/43 15/43 18/47 5/16 12/19 8/19 6/19 9/18 3/20 4/20 9/20 8/45

Note: Units are MP a1/2. All correlations from Reference 6, e xcept as noted. See also Chapter 12; Ultem is a re gistered trademark of the General Electric Company. a G/T is the number of “good” or attacking solv ents relative to the total number of solvent data points used in this correlation. b Data from Reference 10 through Reference 14. c Data from Mai, Y.-W., J. Mater. Sci., 21, 904–916, 1986. With permission.

7248_C014.fm Page 262 Wednesday, May 23, 2007 11:24 AM

262

Hansen Solubility Parameters: A User’s Handbook

solubility than it will have for a correlation for ESC where the “good” solvents have critical strains less than, say , 0.6%. HSP correlations can be made with man y dif ferent types of data (barrier properties, wetting behavior, swelling, solubility, sedimentation rates, etc.) as discussed else where in this book. The term RED number is intimately connected with a gi ven HSP correlation. Hansen9 used data from v arious sources 6,10–13 to construct a ne w type of plot to correlate ESC data. This plot uses the RED number, Equation 1.10, and the molar volume, V, of the test chemicals. Figure 14.2 is such a plot for a lar ge number of liquids in contact with small, injection-molded cylinders of a COC-type polymer calledTopas® 6013 from Ticona.6,9 There are three distinct regions on this plot. There is a re gion at lo w RED including those challenge liquids that dissolv e the polymer or are very aggressive, and ESC is not found as such. There is a region at high RED where the absorption is zero or not great enough to matter , or else the absorption rate is slo w enough to allow relaxation of the polymer in preference to ESC. ESC can occur in an intermediate re gion where there is some absorption of challenge liquid, although e xamples are gi ven below for other polymers where ESC tak es place without measurable absorption for good matches in HSP . The ESC region on these plots increases in size with increased tensile stress and/or higher v alues of strain. Closer study of Figure 14.2 sho ws that there are se veral liquids that do not gi ve ESC whereas this would normally be e xpected from their position on this plot. From Table 14.2, it can be seen that these include meth yl isobutyl ketone, acetophenone, and nitrobenzene. These are surrounded on the plot by liquids that do result in ESC including n-hexane, n-butyl acetate, ethyl acetate, and diethyl ether. The RED numbers and molecular v olumes for all se ven of these liquids are comparable. The explanation lies in the fact that the four liquids giving ESC have measurable absorption, whereas the three not gi ving ESC apparently do not absorb at all under the test conditions. It has been determined that 1,4-dioxane, meth yl isob utyl k etone, acetophenone, and phen yl acetate do not absorb into this polymer at room temperature.14 The methyl “side-group” on the methyl isobutyl ketone, and the benzene rings in acetophenone, nitrobenzene, and phen yl acetate, are e vidently sufficient to pr vide the steric hindrance that prevents absorption. Methyl isobutyl ketone does give ESC at higher stress le vels. All of the test liquids causing ESC f ailure in the immersed COC c ylinders had measurable surface resistances retarding absorption, b ut they did absorb.14 Surface resistance phenomena may play an important part in the ESC process itself. Delayed absorption can also potentially lead to postponing a catastrophic ESC f ailure be yond normal testing times. Chapter 16 in this book is therefore dedicated to absorption and dif fusion in polymers, and especially emphasizes the frequently overlooked surface resistance. This surface resistance is thought to originate primarily from the rate at which adsorbing molecules can locate a hole in the polymer surf ace large enough to accommodate them. Larger and more structurally complicated molecules have much more difficult finding such a suitable hole, so the sur ace transport coef ficient is i versely proportional to the molecular cross-section. Molecules that are too lar ge can simply not enter the polymer . Once an adsorbed molecule locates in a suitable hole, the rate of motion into the b ulk is dependent on the local diffusion coefficient. Therefore the surface transport coefficient is directly proportional to th diffusion coefficient. See the discussion in Chapter 16 for more details and xamples. Figure 14.3 and Figure 14.4 use the same parameters, RED vs. V, for correlating ESC in the polymers PC and PVC, respecti vely. The data used to construct these were tak en from references 10–13, that is, from the older literature. It can be seen that additional liquids will gi ve ESC as the critical stress levels become higher. Before proceeding to perhaps still more complicated theories and situations in the next section, one should be reminded that there are lar ge v ariations in the critical strain e ven for the he xane isomers, which all have very similar total (Hildebrand) or Hansen dispersion solubility parameters. For PC, these isomers ha ve critical strains that increase from 0.85% to 1.68% as branching increases.12 These data clearly reinforce the need for consideration of shape.

7248_C014.fm Page 263 Wednesday, May 23, 2007 11:24 AM

Applications — Environmental Stress Cracking in Polymers

TOPAS 6013 solubility

●

3.0 2.8

δD

δP

δH

RO

18.0

3.0

2.0

5.0

Soluble

C ESC  No ESC DMF

D Severe deformation

NEE

2.6 No problem at stress level

2.4 HSP, Red number for true solubility

263

CHK DAA NMP

2.2 ACI

2.0 1.8 ESC possible

1.6

2NP THF

1.4

D

1.2

MCHL

C

1.0

EET ETA C C NTB

ISOPH HXA

C

ACETOPH MIK C BCN

DOX EDC

C

●●

0.8

●

Solubility

● ●

0.6

● ●

0.4

●

0.2 0

10

20

30

40

50

60

70 80 90 100 110 120 130 140 150 160 170 Size parameter, V

FIGURE 14.2 Plot of ESC and solubility data for injection-molded cylinders of a COC polymer (Topas 6013, Ticona) using RED number vs. molar v olume. There is a re gion of solubility (RED less than 1.0), an intermediate region for ESC, and a re gion at higher RED where ESC is not found. The ESC solvents all have linear molecular structure. Data from Reference 6. Symbols are e xplained in Table 14.2. (Reproduced from Hansen, C.M., Polym. Degradation Stability, 77, 43–53, 2002. With permission from Else vier Science.)

ESC WITH NONABSORBING STRESS CRACKING INITIATORS Recent research has shown that even nonabsorbing chemicals can induce ESC, in some cases v ery rapidly.15–17 It is presumed that there are man y additional cases of this kind, but they have not been reported as such. One possible e xplanation for this is HSP-induced motion of the polymer chain segments in the surf ace. This phenomenon is called surface mobility here and has also been discussed in Chapter 15 and Chapter 18. An e xample of surf ace mobility is contact with w ater converting surf aces lik e those of peat moss from h ydrophobic to h ydrophilic. An applied w ater droplet initially beads up but soon soaks readily in. The surface again becomes hydrophobic when the water is gone. This might be considered as Nature’ s valve to conserv e water within a system. There are other e xamples given in Chapter 18. Rotation or other form of polymer chain se gment

7248_C014.fm Page 264 Wednesday, May 23, 2007 11:24 AM

264

Hansen Solubility Parameters: A User’s Handbook

TABLE 14.2 Symbols Used in Figure 14.2 Symbol

Compound

ACI ACETOPH BCN CHK DAA DMF DOX EDC EET ETA HXA MCI MIK 2NP NEE NMP NTB THF

Acetone Acetophenone Butyl acetate Cyclohexanone Diacetone alcohol Dimethyl formamide 1,2-Dioxane Ethylene dichloride Diethyl ether Ethyl acetate n-Hexane Methylene dichloride Methyl isobutyl ketone 2-Nitropropane Nitroethane N-methyl-2-pyrrolidone Nitrobenzene Tetrahydrofuran

motion may also be suf ficient to start the cracking process. All of the ESC cases kno wn to the author where absorption has not been measurable have had relatively high stress/strain imposed on the samples. There have been many other undocumented cases of ESC where one would not suspect absorption of the ESC initiator . These include contact with chemicals in hair spray , deodorants, hand creams, b utter, and v arious kinds of oils including essential oils. In all cases, it is thought that the af finity of the act ve chemical must be high or moderately high, such that e ven though it cannot absorb, it is still capable of inducing motion in the polymer chains at the surf ace. A related study of surf ace phenomena by Nielsen and Hansen 18 explored whether ESC could be predicted by the wetting beha vior of the challenge chemicals on PC, COC, and ABS type polymers. A large number of challenge chemicals were divided into three groups. Group A included those that spontaneously spread when applied as a droplet. Group B included those that w ould not spontaneously spread, but which would not retract either, when they were applied as films. Grou C liquids did not spontaneously spread and retracted when they were applied as films. It as found that all A and B liquids gave ESC, although the B types had higher critical strains in general. Some type C liquids g ave ESC and others did not. Retraction of an applied film is not an indication tha ESC will not occur . Contamination of the test surf ace may also lead to retraction of an applied film, for xample, where this w ould not occur otherwise. Care must be tak en, and this test is only an indication of a potential problem. ESC w as found for PC polymers with polydimeth ylsiloxanes having molecular weights of 340 and belo w but not for those with 410 and abo ve. Exactly wh y this happens is not known, but surface entry or lack of same (see Chapter 16) and surf ace mobility of polymer molecules may both be in volved.

DISCUSSION HSP correlations of ESC phenomena ha ve been presented for a number of common polymers. As stated above, the mechanisms for crazing and crack initiation and gro wth have not been discussed in detail here. The basis of the failures is the pulling of polymer chains from each other in all cases,

7248_C014.fm Page 265 Wednesday, May 23, 2007 11:24 AM

Applications — Environmental Stress Cracking in Polymers

265

CALCULATED ESC FOR PC AT CRITICAL STRAIN = 0.6 2.5

▲

2 ●

Glycerol 0.77, 1.99

●

● ● 1.5

●  

RED NO.

 ● 2-Propanol 1.02  ◆  ◆ ▲



● 





●

●

 

▲ ●  ⴐ ⴐ Cyclohexanol 0.98, 1.48

◆●  ◆ ● ⴐ ⴐ Dimethylformamide 1.55 ⴐⴐ 

◆▲ ◆ ◆

1

0.5

●

◆

ⴐ ◆

0

0

50

100

150

200

250

300

350

MOL VOLUME - CC/MOL  0-0.2ⴐ0.2-0.4 ◆ 0.4-0.6 ▲ 0.6-0.8  0.8-1 ●>1 FIGURE 14.3 Plot of ESC data for polycarbonate (PC) using data from Mai 10 and others 11–13 to establish a HSP correlation based on critical strain since true solubility data w as lacking. The limit for critical strain w as chosen as 0.6%. Liquids with critical strains belo w this value should ha ve RED less than 1.0, just as liquids with critical strains abo ve this should ha ve RED larger than 1.0. Specific data and a discussion r garding the correlation are included in the te xt. Higher critical strains lead to lar ger ESC regions on this plot. The critical strains are indicated in the caption. (Reproduced from Hansen, C.M., Polym. Degradation Stability, 77, 43–53, 2002. With permission from Else vier Science.)

however. A more detailed discussion of the phenomena in volved is considered be yond the scope of this chapter . HSP has been used as a correlating parameter , necessarily with other parameters such as molecular size and shape of the challenge chemical and the stress/strain condition of the samples, to pro vide improved predictability, and to ask ne w questions re garding ESC. Figure 14.1 shows that the ESC solvents have RED numbers slightly larger than those required for solubility. The data for this figure are found in 6 Figure 14.2 clearly sho ws that the molecular shape of the challenge molecules is important in addition to V. ESC may occur for a challenge chemical with a linear molecular structure, b ut not

7248_C014.fm Page 266 Wednesday, May 23, 2007 11:24 AM

266

Hansen Solubility Parameters: A User’s Handbook

HSP CORRELATION FOR ESC IN PVC AT CRITICAL STRAIN = 0.6 2.5 ●

2

●

▲

1.5 PENTANE

RED NO.

▲

CYCLOHEXANE

METHYL CYCLOHEXANE HEPTANE

▲ ▲ ▲



1 ⴐ



ⴐ

● ●

0.5

●

●

 ▲

ETHYL BENZENE

●

NITROMETHANE

●

ⴐ

●

0 0

20

40

60

80

100

120

140

160

MOL VOLUME - CC/MOL ● DISSOLVES  0-0.2 Ⳮ 0.2-0.4 ◆ 0.4-0.6 ▲ 0.6-0.8  0.8-1.0 ● >1 FIGURE 14.4 Plot of ESC data for polyvin ylchloride (PVC) using data from Mai 10 to establish a HSP correlation based on critical strain. ESC data were used for the correlation since specific solubility data o these samples were not available. The limit for critical strain was chosen as 0.6%. Solvents with critical strains below this v alue should ha ve RED less than 1.0, just as solv ents with critical strains abo ve this should ha ve RED larger than 1.0. Higher critical strains lead to lar ger ESC re gions on this plot. The critical strains are indicated in the caption. (Reproduced from Hansen, C. M., Polym. Degradation Stability, 77, 43–53, 2002. With permission from Else vier Science.)

for one with the same RED and V, but with a c yclic or more branched structure. The absorption rate of the b ulkier molecules is too slo w or the y may not e ven be able to absorb altogether . The samples providing the data for this figure were small ylinders, so it w as impossible to determine critical strains for the gi ven liquids with these samples. Figure 14.3 clearly shows larger ESC regions for higher critical strain limits for polycarbonate. More solvents logically gi ve ESC as the strain increases. The data used in this figure were foun in.10–14

7248_C014.fm Page 267 Wednesday, May 23, 2007 11:24 AM

Applications — Environmental Stress Cracking in Polymers

267

Figure 14.4 shows the ESC behavior of PVC using data tak en from the literature. 10 This figur confirms the ability to use this type of plot to correlate and predict ESC in polymers It is not unusual for suppliers of polymers to change compositions without indicating what has been done and to continue to use the general indication of polymer type, such as PC. Thus, the PC data in Figure 14.3 and the PVC data in Figure 14.4 may not be v alid for all polymers of these nominal types. This is a significant problem in further research and d veloping fundamental understanding. Researchers either must w ork with commercial materials the y do not ha ve full knowledge of, or else tak e the path of making test materials to be certain of what is being dealt with. In the latter case, the information may not apply directly in practice because the practical materials may have added components. This dilemma apparently has no fully satisfactory solution.

CONCLUSION HSP have been used to de velop correlations of ESC. It w ould appear that testing is still required to ascertain what beha vior is to be e xpected, b ut these correlations clearly indicate where the problems will be greatest. It is usually the une xpected events that cause the catastrophic f ailures, but designs or changes in designs that lead to increased tensile stress ha ve also been the causes of ESC in practice. Care is especially required whene ver the HSP of the polymer gets too close to the HSP of potential challenge chemicals. The required af finity between act ve chemical and polymer allows correlations of ESC with HSP , most often with the simultaneous need to consider the size and shape of the acti ve chemical and the stress/strain condition of the polymer .

REFERENCES 1. Wright, D., Failure of Plastics and Rubber Products, Rapra Technology Limited, Shawbury, Shrewsbury, Shropshire, U.K., 2001. 2. Lustiger, A., Understanding En vironmental Stress Cracking in Polyeth ylene, Medical Plastics and Biomaterials Magazine, MPB Article Inde x (originally published July 1996), http://www .devicelink.com/mpb/archive/96/07/001.html. 3. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, 2nd ed., CRC Press, Boca Raton, FL, 1991. 4. Vincent, P.I. and Raha, S., Influence of ydrogen bonding on crazing and cracking of amorphous thermoplastics, Polymer, 13, 283–287, 1972. 5. Wyzgoski, M.G., The role of solubility in stress cracking on n ylon 6,6, in Macromolecular Solutions, Seymour, R.B. and Stahl, G.A., Eds., Per gamon Press, New York, 1982, pp. 41–60. 6. Hansen, C.M. and Just, L., Prediction of en vironmental stress cracking in plastics with Hansen solubility parameters, Ind. Eng. Chem. Res., 40(1), 21–25, 2001. 7. Moskala, E.J. and Jones, M., Evaluating Environmental Stress Cracking of Medical Plastics, Medical Plastics and Biomaterials Magazine, May, 1998, pp. 34–45. 8. Anonymous, Ultem Resin Design Guide, GE Plastics, Pittsfield, MA, 1989 9. Hansen, C.M., On Predicting En vironmental Stress Cracking in Polymers, Polym. Degradation Stability, 77, 43–53, 2002. 10. Mai, Y.-W., Environmental stress cracking of glassy polymers and solubility parameters, J. Mater. Sci., 21, 904–916, 1986. 11. Kambour, R.P., Gruner, C.L., and Romagosa, E.E., Bisphenol-A polycarbonate immersed in or ganic media. Swelling and response to stress, Macromolecules, 7, 248–253, 1974. 12. Jacques, C.H.M. and Wyzgoski, M.G., Prediction of en vironmental stress cracking of polycarbonate from solubility considerations, J. Appl. Polym. Sci., 23, 1153–1166, 1979. 13. Henry, L.F., Prediction and evaluation of the susceptibilities of glassy thermoplastics to environmental stress cracking, Polym. Eng. Sci., 14(March), 167–176, 1974. 14. Nielsen, T.B. and Hansen, C.M., Significance of sur ace resistance in absorption by polymers, Ind. Eng. Chem. Res., 44(11), 3959–3965, 2005.

7248_C014.fm Page 268 Wednesday, May 23, 2007 11:24 AM

268

Hansen Solubility Parameters: A User’s Handbook

15. Al-Saidi, L.F., Mortensen, K., and Almdal, K., En vironmental stress cracking resistance. Beha viour of polycarbonate in different chemicals by determination of the time-dependence of stress at constant strains, Polym. Degradation Stability, 82, 451–461, 2003. 16. Hansen, C.M., Environmental stress cracking of PTFE in k erosene, Polym. Degradation Stability, 77, 511–513, 2002. 17. Kjellander, C.K., Publications being prepared. 18. Nielsen, T.B. and Hansen, C.M., Surf ace wetting and the prediction of en vironmental stress cracking (ESC) in polymers, Polym. Degradation Stability, 89, 513–516, 2005.

7248_C015.fm Page 269 Wednesday, May 9, 2007 9:37 AM

Solubility 15 Hansen Parameters — Biological Materials Charles M. Hansen and Tim S. Poulsen ABSTRACT The Hansen solubility parameters (HSP) of man y biological materials can be found from correlations of ho w the y interact with well-defined liquids. The three HSP parameters, δD, δP , and δH quantitatively account for the cohesion ener gy density arising from atomic, dispersion type inter actions (D), molecular, dipolar interactions (P), and molecular, hydrogen bonding interactions (H). Examples of HSP correlations included in this chapter are DN A, cholesterol, chloroph yll, w ood chemicals, polypeptides (proteins), human skin, nicotine, lard, and urea. The often-quoted “lik e dissolves like” has been expanded to “like seeks like” (self-association) to discuss the implications of these correlations. The ability of HSP to correlate surf ace phenomena has made this change mandatory. Biological materials such as proteins and DN A have well defined structures in a g ven environment. DNA adopts double helices, whereas proteins consist of a combined shape of the secondary, tertiary, and in some cases quaternary structure that together determine the conformation of the protein. The ultrastructure of wood is another example of Nature’s way of establishing order in complex systems. The proper function of a protein requires that certain functional groups are at precise locations within its tertiary and/or in some cases quaternary structure. The conformation of proteins and DN A can be influenced, and in ma y cases controlled, by solv ent quality. The solvent quality in a gi ven environment is e xpected to determine whether a protein is dissolv ed or not, and also to control the w ay it adsorbs onto other materials or interacts with itself. Controlled changes in solv ent quality can lead to controlled changes in conformation. Solv ents can change not only the ability of noncovalent interactions such as van der Waals, hydrogen bonding, and ionic bonding, but also induce chiral rotation. The key to the importance of nonco valent interaction is that such interactions can continually be brok en and reformed under physiological conditions. The portion of the molecule with ener gy properties most similar to the surrounding liquid will be oriented toward the liquid (“lik e seeks lik e”). The term hydrogen bonding is generally used to describe the noncovalent interactions in DNA, proteins, and other biological molecules, implying that this is the dominating interaction. The HSP correlation based on solv ent interactions with DN A resulted in δD;δP;δH v alues equal to 19.0;20.0;11.0. These numbers clearly sho w that h ydrogen bonding pro vides by f ar the smallest contribution of the three types of interaction, representing only about 14% of the cohesi ve energy involved (using Chapter 1, Equations 1.6–1.8). The term hydrogen bonding must be considered as an insufficient description of the interactions that determine the structure in such molecules

269

7248_C015.fm Page 270 Wednesday, May 9, 2007 9:37 AM

270

Hansen Solubility Parameters: A User’s Handbook

INTRODUCTION HSP have been used to characterize man y biological materials. 1–7 Most of the materials discussed in these references are also included in the present discussion, b ut many more can be added by experiment or calculation. There are man y simple e xperimental methods to determine the HSP for biological materials. These involve contacting a material of interest with a series of well-chosen liquids. The f act of solubility, differences in degree of equilibrium swelling, rapid permeation or not, significant surace adsorption or not, or other measurable quantity significantly influenced by ysical affinity relation can be observ ed and used to find the HSP for a material being studied. These methods have been discussed in more detail in earlier chapters. The basis of the division of the cohesive energy density into three parts accounting for the atomic dispersion (D), molecular dipolar (P), and molecular hydrogen bonding (H) interactions, respecti vely, is given in detail in Chapter 1. The HSP for simpler compounds can be calculated according to the methods gi ven in Chapter 1. HSP v alues for nicotine, skatole, w ood chemicals, etc., that are discussed in this chapter were calculated using these methods. Figure 15.1 sho ws a typical HSP sphere correlating e xperimental solubility data for lignin. 1 The good solv ents are located within the sphere which is based on Chapter 1, Equation 1.9. Again, as stated in pre vious chapters, this equation is in agreement with the Prigogine corresponding states theory of polymer solutions as discussed in Chapter 2. The statistical thermodynamics approach presented in Chapter 3 also shows agreement with the concepts to which this book is dedicated. Furthermore, this equation has also been sho wn to be correct for such comple x materials as asphalt and bitumen, as described in Chapter 9, and carbon dioxide solubility in solv ents, as described in Chapter 10. A HSP correlation can, of course, be used to predict the beha vior of solv ents not included in the experimental work. It is con venient to print the solv ent database in order from best solv ent to worst solvent to aid in finding alternat ves. This is a quantitati ve application of the generally used statement “Lik e Dissolv es Lik e.” In the follo wing discussion, this concept is e xpanded to “Lik e Seeks Lik e” (self-association). This implies that se gments of molecules seek re gions of similar HSP if this is possible. This may result in solutions or in selecti ve orientation of se gments of molecules in more complicated systems. Table 15.1 contains HSP data for se veral biologically interesting materials. These are discussed in the following in more detail with an indication of ho w such data may be used. The data included in this table are the δD, δP , and δH parameters; the radius of interaction for the HSP correlation, Ro; if appropriate, the data fit (where a fit of 1.000 is perfect as discussed in Chapter 1). G is number “good” solvents and the total number of solvents in a given correlation is T. The units for the solubility parameters and Ro are MP a1/2. Plots of the kind gi ven in Figure 15.1 for lignin are sometimes used to interpret relations among dif ferent materials. RED numbers indicate solv ent quality with lo wer values being indicati ve of better solv ents (see Chapter 1, Equation 1.10). The correlations reported here are a result of data processing with the SPHERE program described in Chapter 1. The output is often arranged with the best solv ent (lowest RED number) at the top of the list. A most interesting and important class of molecules are called amphipathic. These exhibit both hydrophilic and h ydrophobic properties simultaneously . An example from biology is the amphipathic molecules (lipids) that form the basis of the biological membrane bilayers that surround cells. Such amphipathic molecules ha ve a head group that is strongly h ydrophilic, coupled to a hydrophobic tail – usually h ydrocarbon in nature. When one attempts to dissolv e these molecules in water, they form special structures. These may be monolayers on the w ater surface, with only the head groups immersed. Alternatively, if the mixture is vigorously stirred, micelles (spherical structures stabilized by a single layer of molecules at the w ater interface) or bilayer v esicles may form. Another example is amino acid side chains. These are by nature not only dif ferent in size and shape, but also in the char ge they carry, their general affinity for ater (hydrophilicity) and/or their general a version to w ater (hydrophobicity). The native conformation of proteins is a strong

7248_C015.fm Page 271 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

271

Hansen Solubility Parameters

Lignin

δD

δP

δH

RO

21.9

14.1

16.9

13.7

40

Dispersion Solubility Parameter, δD

30

Lignin

20 15 Polar Solubility 10 Parameter, δP

15 10

10

5

5 0

0

Hydrogen Bonding Solubility Parameter, δH

FIGURE 15.1 HSP correlation showing the solubility of lignin. Good solv ents are located within the sphere. Units are MP a1/2. (From Hansen, C.M. and Björkman, A., Holzforschung, 52, 339, 1998. With permission.)

function of the interactions that tak e place within and between polypeptide chains. This is also highly dependent on the interaction that tak es place with w ater, as proteins e xist in an aqueous environment. These general concepts of hydrophilic and hydrophobic entities can also be quantifie using HSP.

HYDROPHOBIC BONDING AND HYDROPHILIC BONDING (SELF-ASSOCIATION) The concept of “lik e seeks lik e” offers a general e xplanation of h ydrophobic bonding. An aliphatic hydrocarbon chain on a protein, for e xample, is not soluble in w ater and ultimately finds anothe aliphatic h ydrocarbon chain with which to associate. This same type of process leads to micelle formation when the solubility limit of surface active agents is exceeded. Hydrophobic bonding is found when the HSP for the associating se gments are too lo w to allow solubility in the continuous phase. When it is immersed in w ater a polypeptide chain will not stay in an elong ated form. It will instead fold up into secondary structures according to the polarity of the side chains it contains and the rotation of peptide backbone bond angles that are lar gely determined by Van der Waals radii of side chains. This can be called hydrophilic bonding. Hydrophilic bonding is formed when the HSP for the associating se gments are too high to allo w solubility in the continuous phase. If the continuous phase is a h ydrocarbon liquid, the associating se gments may be characterized by high δH, for e xample, because of the presence of an alcohol, acid, or amide group.

7248_C015.fm Page 272 Wednesday, May 9, 2007 9:37 AM

272

Hansen Solubility Parameters: A User’s Handbook

TABLE 15.1 Hansen Solubility Parameter Correlations for Biologically Interesting Materials, MPa1/2 Material

δD

δP

δH

Ro

FIT

G/T

DNA Cholesterol solubility Lard 37°C solubility Lard 23°C solubility Olive oil solubility Psoriasis scales swelling Human skin — permeation Nicotine — calculation Skatole — calculation Chlorophyll — solubility Sinapyl alcohol calculation Coniferyl alcohol calculation p-Coumaryl alcohol calculation Lignin — solubility Dextran C (= amorphous cellulose) See Chapter 5 Sucrose solubility N-methyl-morpholine-N-oxide calculation Blood serum — swelling Zein — solubility Urea — solubility Water — >1% soluble in Water — totally miscible Water — single molecule

19.0 20.4 15.9 17.7 15.9 24.6 17.6 18.8 20.0 20.2 19.2 19.0 19.1 21.9 24.3 23.4 19.0 25.5 22.4 22.9 15.1 18.1 15.5

20.0 2.8 1.2 2.7 1.2 11.9 12.5 7.8 7.1 15.6 7.3 7.0 7.0 14.1 19.9 18.4 16.1 10.3 9.8 14.9 20.4 17.1 16.0

11.0 9.4 5.4 4.4 5.4 12.9 11.0 6.4 6.2 18.2 16.1 16.3 17.3 16.9 22.5 20.8 10.2 22.1 19.4 21.3 16.5 16.9 42.3

11.0 12.6 12.0 8.0 12.0 19.0 5.0 — — 11.1 — — — 13.7 17.4 16.0

1.000 1.000 1.000 1.000 1.000 0.927 1.000 — — 0.864 — — — 0.990 0.999 0.981

6/12 25/41 29/50 21/50 29/50 35/50 4/13 — — 7/35 — — — 16/82 5/50 6/50

17.8 11.9 16.2 18.1 13.0 —

0.980 0.964 0.984 0.856 0.880 —

4/51 4/50 14/50 88/167 47/166 —

Note: The units for the solubility parameters and Ro are MP a1/2. G/T represents the number of good liquids (G) and the total number of liquids (T) in the correlation.

Figure 15.2 demonstrates ho w hydrophilic bonding between v ersamid polymer blocks reacted into an alkyd (polyester) polymer gives a thixotropic alkyd paint with its special nondrip properties. Agitation of the paint is enough to break the h ydrophilic bonds allowing easy spreading, b ut they reform quickly ag ain after application. The most common secondary structures are alpha helices and beta sheets that are stabilized by local inter-residue interactions mediated by h ydrogen bonds. An alpha helix can tak e the form of an amphipathic helix with a polar and a nonpolar side. This plays a crucial role in helix–helix interactions and in the interaction of small peptides that ha ve a helical conformation with membranes, air–w ater interf aces, and self-assembly processes. Beta sheets are alternati ve secondary structure to the alpha-helix in proteins. Lik e alpha-helices, beta-sheet backbones are stabilized by hydrogen bonds between tw o beta sheets, but the bonds occur between neighboring strands. If the beta−strand contains alternating polar and non-polar residues it forms an amphipathic beta sheet. This distrib ution of h ydrophilic and h ydrophobic residues has been observ ed in the membrane protein porin that forms a beta-barrel structure. Here the nonpolar residues stick into the h ydrophobic part of the lipid membrane and the h ydrophilic residues form part of the channel interior responsible for the passage of small molecules across the membrane. Hydrophobic bonding is a major ef fect that dri ves proper protein folding. Hydrophobic sidechains are oriented to minimize the ener gy lost by the intrusion of amino acids into the w ater solvent, which disrupts lattices of w ater molecules. Hydrophobic bonding forms an interior ,

7248_C015.fm Page 273 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

273

A Mineral spirits B A δP

B

B regions are not soluble and “Precipitate” together

B

A

Polymer B (versamid)

Butanol

M.S. Polymer A (alkyd) δH FIGURE 15.2 HSP relations for establishing thixotrop y in an alk yd-type paint. The solid circle represents the solubility of the alkyd (A) and the dotted circle that of the Versamid (B). The Versamid segments associate because they are not soluble in mineral spirits. Addition of n-butanol destroys the thixotropic effect, since the solvent then becomes too good. Similar relations e xist for the true solution of some proteins by additions of urea to w ater. This denatures them, by ef fectively dissolving them in a solv ent mixture that is better than water itself.

hydrophobic protein core, where most hydrophobic sidechains can closely associate and are shielded from interactions with solv ent water. Formation of “h ydrogen bonds” within proteins is based on the lack of solv ency in the continuous media, w ater, because the HSP of these se gments is too high. Additions of urea, as discussed later in more detail, increase the HSP of the continuous media to such an extent that it can now dissolve the “hydrogen bonded” segments. The protein is denatured, which in f act means that these se gments are dissolv ed in a good solv ent. Additions of salts can also improve solvency for a gi ven material or se gments of materials. Additions of salts can also reduce solvency. These phenomena must also ha ve their e xplanation in the “lik e seeks/dissolves like” phenomena, but more research is required to quantify them. Such mechanisms of controlling solvent quality can be e xpected to be used by Nature in man y biological systems to control adsorption and/or transport of v arious types of materials as in self-association.

DNA The double helix structure of DN A suggested by Watson and Crick is stabilized by h ydrogen bonding between bases on opposite strands when the bases are paired in one particular w ay (A+T or G+C). In the Watson–Crick model the base pairs are stack ed on one another with their planes nearly perpendicular to the helix axis where the hydrophilic phosphate–deoxyribose backbones are on the outside, in contact with the aqueous environment. This complementary base pairing (hybridization) is central to all processes involving nucleic acids. In cells it occurs in, e.g., DNA replication,

7248_C015.fm Page 274 Wednesday, May 9, 2007 9:37 AM

274

Hansen Solubility Parameters: A User’s Handbook

TABLE 15.2 Hansen Solubility Parameter Correlation for DNA Solvent

δD

δP

δH

RED

V

Dimethyl sulfoxide 2,6-Dichloro-7-methyl purine Coumarin Purine Caffeine Formamide Pyrimidine Phenol Urea Cyclohexanol Methyl riboside Adonitol

18.4 20.5 20.0 20.5 19.5 17.2 20.5 18.0 20.9 17.4 17.0 18.0

16.4 11.7 12.5 11.7 10.1 26.2 9.4 5.9 18.7 4.1 12.0 12.0

10.2 14.2 6.7 14.2 13.0 19.0 11.3 14.9 26.4 13.5 32.8 36.0

0.353 0.651 0.807 0.853 0.923 0.977 1.002 1.342 1.447 1.492 2.142 2.393

71.3 162.4 156.3 100.0 157.9 39.8 78.8 87.5 45.8 106.0 117.2 95.1

DNA

D = 19.0 P = 20.0 H = 11.0 R

o

= 11.0 FIT = 1.000 NO = 12

Note: Units of D, P, H and Ro are MP a1/2. V is in cc/mole. The order in the table is from e xpected best at the top to e xpected worst at the bottom.

transcription, rRNA, and tRNA structure, b ut it is also used in laboratories in RN A and DNA gel blots, PCR, sequencing, genotyping, microarrays, in situ hybridization, etc. DNA melts (denatures) at 90-100 °C in 0.1-0.2 M Na+. This may lead to deterioration of morphology. Fortunately, organic solvents reduce the thermal stability of double-stranded polynucleotides, so that h ybridization can be performed at lo wer temperatures in the presence of formamide, for e xample. Formamide is often used in connection with DN A.8 For in situ h ybridization this implies that microscopic preparations must be hybridized at 65–75° for prolonged periods. The melting temperature, Tm, is found when a population of particular DN A sequences is at a midpoint between fully double-stranded and single-strand. F ormamide reduces the Tm of DN A-DNA and DNA-RNA duplexes in a linear f ashion by about 0.65 °C for each v olume percent of the solv ent that is present. Other common solv ents can also reduce Tm, including dimeth yl sulfoxide. An article in the older literature 9 reports aspects of the interaction of dif ferent low molecular weight materials with DNA. The summary of this article states that the order of increasing acti vity was found to be: adonitol, meth yl riboside (both negligible) < cyclohexanol < phenol, p yrimidine, uridine < c ytidine, thymidine < purine, adenosine, inosine, deoxyguanosine < caf feine, coumarin, 2,6-dichloro-7-methylpurine. Urea w as inef fective with poly A and only slightly ef fective with DNA. At a concentration of 0.3M, purine lo wered the Tm of DNA by about 9 °C. The HSP for se veral of these ha ving reasonably simple structures were estimated by the methods of Chapter 1. These HSP data were di vided into tw o arbitrary groups of “good” and “bad” with a dividing line between purine as good and pyrimidine as bad. The compounds intermediate in the above list were structurally too complicated to allow a reliable calculation. Formamide and dimethyl sulfoxide were also considered as “good” and added to the data for the correlation reported in Table 15.2. The encouraging correlation reported in Table 15.2 ranks the gi ven solvents in approximately the same order as that gi ven in Reference 9. All the solv ents from p yrimidine and lo wer were considered as being “bad” and all those abo ve this were considered as being “good. ” Even urea, where performance may be af fected significantly by the presence of ater, seems to be placed correctly. Formamide is not at the top of the list, b ut is the preferred solv ent of use today in man y cases. The effectiveness of formamide is primarily because of its low molecular volume, but it will also be a good solv ent for phosphate salts, which may also contrib ute some ef fect. Dimeth yl

7248_C015.fm Page 275 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

275

sulfoxide will also be a reasonably good solvent for phosphate salts. Low molecular volume is very conducive to dissolving polymers with structure or crystallinity , as the small molecules can reach the critical sites more readily than lar ger ones. Smaller molecules are also predicted to be thermodynamically better, all else being equal. The radius is arbitrary and depends on the criterion used for good and bad. If p yrimidine had been considered as being good, then the D, P , and H could be maintained with a slightly larger Ro, and the data fit ould still be 1.000. There are many different D, P, and H, combinations possible when the data fit is 1.000, ut the present correlation, in spite of the v ery few solvents, is still considered reasonably reliable because of the essentially correct ranking. Other supporting e vidence that the correlation is reasonable can be found in the estimated HSP for adenine and th ymine. These can be considered as single rele vant portions of DNA. The HSP are δD;δP;δH equal to 20.0;16.0;14.9 for adenine and 19.0;20.5;13.0 for th ymine. Both of these are reasonably close to HSP equal to 19.0;20.0;11.0, the estimated v alues for DNA based on its interaction with a number of solv ents as reported in Table 15.2. All units are MP a1/2. The δH value for DN A is only 11.0 MP a1/2 compared with δD equal to 19.0, and δP equal to 20.0. This clearly sho ws that the h ydrogen-bonding interactions are f ar less important than the other two types of interaction. The cohesive energy derived from hydrogen bonding is about 14% of the total using Chapter 1, Equations 1.6 to 1.8.

CHOLESTEROL Cholesterol has been characterized with HSP based on its solubility in a lar ge number of solvents. δD, δP, and δH and Ro for cholesterol solubility were found as 20.4;2.8;9.4 and 12.6, all v alues having units of MPa1/2. The test method involved placing 0.5 g of cholesterol in test tubes together with 5 ml of each of 41 dif ferent solvents. The temperature was 23°C. Total solution or not at this concentration w as e valuated visually . The 25 “good” solv ents dissolv ed the entire amount of cholesterol added. These data were analyzed by the SPHERE computer program described in Chapter 1 to find the HSP for cholesterol. This has also been reported in Reference 10. Figure 15.3 shows this HSP correlation for cholesterol. This figure also includes s veral solv ents that are discussed in the follo wing. The data fit of 1.0 indicates that there are other sets of parameters for spheres which can b expected to give a perfect separation of the good solv ents from the bad ones by a “spherical” HSP correlation. Continued testing with additional test solv ents located in the boundary re gion of the sphere is possible to define it more precisel . This was not w arranted under the present circumstances, but is recommended if more e xtensive use of these data is planned. A general confirmation of the HSP correlation for cholesterol as done by studying mixtures of nonsolvents. Many mixtures of tw o nonsolvents which dissolv e polymers when admix ed have been reported in the literature. 1 Such synergistic mixtures can be predictably found when the y are pairwise on opposite sides of an HSP sphere. The 50:50 vol mixtures of n-hexane with 2-nitropropane and n-hexane with ethanol predictably dissolv ed cholesterol at 0.5 g/5 ml. During the course of this study , it also became ob vious that the solubility of cholesterol in hydrocarbons was limited and quite temperature dependent, being considerably higher at slightly elevated temperatures. This behavior in hydrocarbon solvents relates to the interactions of cholesterol in the hydrocarbon (hydrophobic) portions of lipid layers.The limited solubility in hydrocarbon media and v ery low solubility in w ater favors a location at an aqueous interf ace with the alcohol group of the cholesterol molecule oriented toward the high energy aqueous phase, where it is more compatible, and the h ydrocarbon portions oriented into the lipid layer . Changes to ward lo wer temperature will tend to force more cholesterol out of a h ydrocarbon matrix. The δH parameter of alcohol solv ents decreases relati vely more rapidly with increasing temperature than for solv ents where the δH parameter is lo w (or zero), such as with the h ydrocarbon solvents. This brings the HSP of the alcohol solvent closer to the HSP of the hydrocarbon solvents, and miscibility improves markedly as temperature increases.

7248_C015.fm Page 276 Wednesday, May 9, 2007 9:37 AM

276

Hansen Solubility Parameters: A User’s Handbook

MATERIAL

δD

δP

δH

CHOLESTEROL

20.4

2.8

9.4

2 – NITROPROPANE

16.2

12.1

4.1

HEXANE

14.9

0.0

0.0

ETHANOL

15.8

8.8

19.4

RO 12.6

DISSOLVING MIXTURES OF NON–SOLVENTS

2–NITRO– PROPANE

15

10 ETHANOL δP 5

(9.4, 2.8, 20.4) 0 20

15

10

5

0

HEXANE

δH

FIGURE 15.3 HSP sphere correlating the solubility of cholesterol. Nonsolvents which synergistically interact to become impro ved solvents when mix ed are indicated. These can predictably be found by selecting pairs located on opposite sides of the HSP solubility parameter sphere. Units are MP a1/2.

One can also surmise what might happen when ethanol or other or ganic solvent is present in the body. Or ganic solv ents with HSP resembling those of the lipid layer may be found due to occupational e xposure or for other reasons, such as drinking alcohol-containing be verages. The presence of ethanol or other organic solvent in the lipid layer allows greater cholesterol miscibility in its hydrocarbon portions. The reason for this is the synergistic effect of ethanol and hydrocarbon segments described earlier. The simple e xperiments described pre viously indicate that the cholesterol uptak e in h ydrocarbon portions of a lipid layer will be greatly enhanced when ethanol is present. This, of course, preferentially remo ves some of the cholesterol from the blood stream. The solubility of cholesterol in an essentially nonsolv ent such as w ater can be enhanced by additions of a solv ent improver such as ethanol. The average HSP for these mixtures are closer to those of cholesterol itself. Therefore, those persons with alcohol in their blood can anticipate a

7248_C015.fm Page 277 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

277

slightly higher solubility of cholesterol in their blood because the continuous phase has solubility parameters closer to those of cholesterol. This effect and that discussed earlier should help to reduce cholesterol levels in the blood and blood v essels of those who ingest small to moderate amounts of alcohol on a re gular basis.

LARD Experimental data and HSP correlations for the solubility of refined lard at 23°C and 37°C h ve been reported. 2 The criterion for a good solv ent is that it totally dissolv es the sample at the gi ven temperature. The concentrations chosen were 10%. The results of the correlations are gi ven in Table 15.1. The refined lard is a semisolid with a melting point of 42°C The composition of refined lard is ery similar to that of human depot f at, so the conclusions drawn for the solubility of lard will also be generally v alid for depot f at. Olive oil is a con venient material to use at room temperature to study the beha vior of depot f at (lard), as the same solv ents that dissolve it at room temperature also dissolv e lard at 37°C. This is reported in Table 15.1. The best room temperature solv ents for lard include trichloroeth ylene, styrene, toluene, and methyl methacrylate. Octyl alcohol does not ha ve a strong af finity for lard at room temperatur with a RED number (see Chapters 1 and 2) of 0.96. The good solvents reflect the crystalline natur of the lard, as toluene, for e xample, is an e xcellent swelling solvent for partly crystalline polyethylene. Esters are among the best solv ents for lard at 37°C, reflecting the presence of the este groups in the lard, which is v ery nearly a liquid at this temperature.

HUMAN SKIN A first attempt to characterize human skin with HSP as made by visually evaluating the swelling of psoriasis scales immersed for a prolonged time in dif ferent solvents.2 Uptake could clearly be seen by dimensional changes and a mark ed enhancement of clarity . It w as anticipated that the solubility parameter correlation for the psoriasis scales (k eratin) w ould to some e xtent reflec permeation in human skin b ut that other f actors, such as the presence of w ater and lipids, for example, would also be important. The data fit for this correlation (0.927) indicates that a reasonabl reliable correlation for swelling of the psoriasis scales (k eratin) has been found. Ho wever, the δD parameter is thought to be too high. Permeation data generated in an e xtensive study allowed placement of the tested solv ents into groups according to actual permeation rates through viable human skin. 4 Figure 15.4 graphically shows the HSP correlation that resulted. There are too fe w data to establish a reliable correlation, but a sphere with center at δD, δP , and δH of 17.6, 12.5, and 11.0, which has a radius of 5.0, encompasses the parameters for the four solvents with the highest permeation rates while excluding the others. The units for these parameters are MP a1/2. n-octyl acetate has a near zero permeation rate. This correlation cannot be considered precise because of insuf ficient data, and there are, i fact, numerous spheres with some what similar b ut different combinations of the parameters that also can accomplish this. Ne vertheless, there is a good guideline for future w ork, whether it be an expanded correlation or formulation of products designed for a prescribed compatibility with human skin. Calculations for skatole and nicotine predict that moderate rates of skin permeation can also be expected for these. It might be noted that the four solv ents with high permeation rates also ha ve very high affinit for psoriasis scales according to the correlation pre viously noted. Lik ewise, the c yclic solv ents propylene carbonate, gamma-butyrolactone, and sulfolane have, or are predicted to have, high affinit for psoriasis scales, b ut the y are placed in the lo w permeation rate group for actual permeation through viable human skin. These all have high δP and low δH. n-butyl acetate and toluene are also in this group. This reflects the compl xity of actual skin permeation and the importance of using viable skin for testing. The cyclic nature of the solv ents, however, is also e xpected to slow the rate

7248_C015.fm Page 278 Wednesday, May 9, 2007 9:37 AM

278

Hansen Solubility Parameters: A User’s Handbook

SOLUBILITY PARAMETER PLOT FOR SKIN PERMEATION RATE δD

δP

δH

MV

DMSO DMF DMAC NMP

18.4 17.4 16.8 18.0

16.4 16.7 11.5 12.3

10.2 11.3 10.2 7.2

71.3 77.0 92.5 96.5

MCL MEK ETH

18.2 16.0 15.8

6.3 9.0 8.8

6.1 5.1 19.4

63.9 90.1 58.5

BAC PPC TOL BTA SUL

15.8 20.0 18.0 19.0 18.4

3.7 18.0 1.4 16.6 16.6

6.3 4.1 2.0 7.4 7.4

132.5 85.0 106.8 76.8 95.3

OAC

15.8

2.9

5.1

196.0

l HIGH

l MODERATE

ⴛ LOW

à “0”

CIRCLE: δ P = 12.5, δ H = 11.0, RO = 5.0

20 HANSEN POLAR SOLUBILITY PARAMETER, δ H

PERMEATION RATE

PPCⴛ

BTA ⴛ ⴛ SUL

15

lDMSO lDMF

lNMP 10

l

MEK l

5

TOLⴛ

lDMAC

l

ETH

MCL

BACⴛ Ã OAC

0 0

5 10 15 20 HANSEN HYDROGEN BONDING PARAMETER, δ H

FIGURE 15.4 Permeation rates of selected solv ents through viable human skin sho w a correlation with the HSP4 although the data are not e xtensive. Units are MP a1/2.

7248_C015.fm Page 279 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

279

of permeation relati ve to linear solv ents of comparable af finit . Factors affecting permeation ha ve been discussed at length in Chapter 13. Of course, the presence of w ater and/or other skin components can also have an effect on the permeation rate. Finally , the swelling of the psoriasis scales in volved equilibrium swelling of the individual systems, whereas the permeation rate studies did not have this uniformity. Concentration gradients are required for permeation to occur .

PROTEINS — BLOOD SERUM AND ZEIN HSP correlations for the swelling of blood serum and for the solubility of zein, a protein deri ved from corn, are included in Table 15.1. The data used in these correlations are found in Reference 3. Solvents with the lo west RED numbers in the correlation for the solubility of zein are listed in Table 15.3. The HSP parameters for blood serum and zein are not too dif ferent. The blood serum data are based on visual observ ation of swelling, while the zein data are for visual observ ation of true solution. It is note worthy that there are only four good solv ents in the data set reported in Table 15.3, and that the HSP parameters for the proteins are much higher than for any liquid which can be used in such testing. These HSP parameters are found by a form of e xtrapolation, where all of the good solv ents are located in the boundary re gion of the respecti ve spheres. The values are very much dependent on the mathematical model which includes the coef ficient “4” (see Chapte 1 and Chapter 2). The saturated solution of urea and w ater is also a (predictably) good solv ent in that it swells blood serum and dissolv es zein, b ut it w as not included as a data point in the correlations as such. Mixtures of solv ents, water, and mixtures of solv ents with w ater have been avoided as test solvents to the extent possible because of too many interactions, which are apparently not always predictable by these simple considerations. The general prediction that additions of urea to water will impro ve solvency of proteins is discussed belo w.

CHLOROPHYLL AND LIGNIN5 The results of HSP correlations of solubility for lignin and chloroph yll are gi ven in Table 15.1. More specific information on the lignin correlation is found in Table 15.4A and Table 15.4B. It can be seen that these indeed ha ve high affinity/p ysical resemblance to each other , with the HSP values not being too different. A major difference is that chlorophyll is soluble in ethanol, whereas lignin is not. This indicates a higher h ydrophilicity, of course, and gi ves a higher δH parameter to chlorophyll compared with lignin. It can be presumed that the HSP for these materials are the result of natural selection by nature for optimum compatibility relations with immediate surroundings and function. A discussion of this is beyond the scope of this work, but this point has been studied in more detail for the relations among w ood chemicals and w ood polymers as outlined in the ne xt section. Here, the HSP for lignin have a demonstrated clear importance with re gard to compatibility relations.

WOOD CHEMICALS AND POLYMERS The results of HSP calculations and correlations for se veral w ood chemicals and polymers are given in Table 15.1. These results are part of a study considering the ultrastructure of w ood from a solubility parameter point of vie w.6 The study is based on the principle of “lik e seeks lik e” and leads to a proposed configuration of the ultrastructure. The HSP for amorphous cellulose are presumed to be similar to those of De xtran (Dextran C, British Drug Houses). The crystallinity in cellulose will require that good solv ents have higher af finity/HSP than most of those dissolvin Dextran, however. N-methyl-morpholine-N-oxide is an e xample. The HSP for De xtran are higher than those of sucrose (which v alues are similar to the other sug ars as well). It is common for polymers to ha ve higher HSP than the monomers from which the y are made. It is also common

7248_C015.fm Page 280 Wednesday, May 9, 2007 9:37 AM

280

Hansen Solubility Parameters: A User’s Handbook

TABLE 15.3 Calculated Solubility Sphere for Zein Solvent

δD

δP

δH

1,3-Benzenediol Benzyl alcohol Diethanolamine Phenol o-Methoxyphenol Furfuryl alcohol Hexamethylphosphoramide 3-Chloro-1-propanol 1,3-Butanediol Propylene glycol Diethylene glycol Ethylenediamine m-Cresol Aniline Dipropylene glycol 1,1,2,2-Tetrabromoethane Ethanolamine Succinic anhydride 2-Pyrolidone Allyl alcohol Ethylene glycol Ethylene glycol monometh yl ether Cyclohexanol Diethylenetriamine Benzoic acid Triethyleneglycol 1,1,2,2-Tetrachloroethane Ethanol 1-Propanol Morpholine Ethylene glycol monoeth yl ether Dimethylformamide Propylene glycol monophen yl ether Quinoline Hexylene glycol Dimethyl sulfone Dimethyl sulfoxide Ethylene cyanohydrin 1-Butanol 2-Propanol Ethylene dibromide Tetramethylurea Glycerol Diethylene glycol monometh yl ether Diethylene glycol monoeth yl ether N,N-Dimethylacetamide Bromoform

18.0 18.4 17.2 18.0 18.0 17.4 18.5 17.5 16.6 16.8 16.6 16.6 18.0 19.4 16.5 22.6 17.0 18.6 19.4 16.2 17.0 16.2 17.4 16.7 18.2 16.0 18.8 15.8 16.0 18.8 16.2 17.4 17.4 19.4 15.7 19.0 18.4 17.2 16.0 15.8 19.2 16.7 17.4 16.2 16.1 16.8 21.4

8.4 6.3 10.8 5.9 8.2 7.6 8.6 5.7 10.0 9.4 12.0 8.8 5.1 5.1 10.6 5.1 15.5 19.2 17.4 10.8 11.0 9.2 4.1 13.3 6.9 12.5 5.1 8.8 6.8 4.9 9.2 13.7 5.3 7.0 8.4 19.4 16.4 18.8 5.7 6.1 3.5 8.2 12.1 7.8 9.2 11.5 4.1

21.0 13.7 21.2 14.9 13.3 15.1 11.3 14.7 21.5 23.3 20.7 17.0 12.9 10.2 17.7 8.2 21.2 16.6 11.3 16.8 26.0 16.4 13.5 14.3 9.8 18.6 9.4 19.4 17.4 9.2 14.3 11.3 11.5 7.6 17.8 12.3 10.2 17.6 15.8 16.4 8.6 11.0 29.3 12.6 12.2 10.2 6.1

SOLUB

0* 0* 1 1* 0 0 0

1* 1* 0

0

0 0

0 0

RED

V

0.761 0.876 0.891 0.893 0.910 0.933 0.950 0.976 0.991 0.997 0.998 0.999 1.001 1.004 1.004 1.021 1.037 1.043 1.061 1.068 1.068 1.073 1.087 1.090 1.099 1.101 1.108 1.112 1.117 1.127 1.128 1.130 1.136 1.137 1.140 1.155 1.165 1.166 1.169 1.179 1.180 1.198 1.198 1.200 1.221 1.226 1.228

87.5 103.6 95.9 87.5 109.5 86.5 175.7 84.2 89.9 73.6 94.9 67.3 104.7 91.5 130.9 116.8 59.8 66.8 76.4 68.4 55.8 79.1 106.0 108.0 100.0 114.0 105.2 58.5 75.2 87.1 97.8 77.0 143.2 118.0 123.0 75.0 71.3 68.3 91.5 76.8 87.0 120.4 73.3 118.0 130.9 92.5 87.5

7248_C015.fm Page 281 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

281

TABLE 15.3 (CONTINUED) Calculated Solubility Sphere for Zein Solvent 2-Butanol 1-Octanol Ethyl lactate Methyl salicylate Zein

δD

δP

δH

15.8 17.0 16.0 16.0

5.7 3.3 7.6 8.0

14.5 11.9 12.5 12.3

SOLUB

RED

V

1.232 1.233 1.236 1.239

92.0 157.7 115.0 129.0

D = 22.4 P = 9.8 H = 19.4 RAD. = 11.9 FIT = 0.964 NO = 50

Note: Units are MPa1/2. This table contains the first entries in a much la ger database to show which solvents are most likely to affect proteins. The SOLUB column indicates good solv ents with a 1, bad solv ents with a 0, and untested solvents with a blank. The “*” points out those solv ents that do not conform e xactly with the correlation.

that the solubility of crystalline polymers requires good solvents to have higher HSP than otherwise expected and that smaller molecular v olume is an adv antage. The relati vely high HSP for cellulose, which also includes a lar ge number of –OH groups, provides a proper ener getic environment for the backbones of hemicelluloses, as well as those of their side groups which contain –OH groups. The hemicellulose side groups with acetyl and ether linkages can be e xpected to orient toward the lower HSP lignin. Neither lignin nor hemicelluloses are compatible with cellulose in the usual sense, b ut the hemicelluloses can form oriented config urations in connection with cellulose and with lignin. The monomers for lignin, sinap yl alcohol, coniferyl alcohol, and p-coumaryl alcohol all have HSP which are on the boundary of the solubility sphere for solubility of De xtran (amorphous cellulose), so their af finities indicate th y will seek the lower HSP domain of the lignin. Hemicelluloses act lik e surfactants, with some side groups favoring the cellulose environment and others favoring the lignin environment. If one considers the HSP for higher k etones, esters, and ethers in Table 15.4, it can be seen that none of these simple liquids will dissolve lignin. This indicates that the acetyl- and ether -containing side groups on the hemicelluloses may not penetrate lignin as such but prefer to remain on its surface, probably findin a local (interf ace) site with closest possible HSP . A sketch of these predicted relations is found in Figure 15.5. This is a clear e xample of self-association in nature. In addition to those pre viously mentioned, one can deduce which chemicals are most prone to penetrate directly through w ood. These will dissolv e lignin. Included are chlorinated phenols and other w ood impre gnation materials. It is kno wn that pentachlorophenol, for e xample, readily diffuses into and through w ood specimens. Still another question is ho w wood transports its o wn chemicals at various stages of the life of a tree. The same principles are valid. A preferred pathway is where HSP are similar . This can be made possible by molecular rotation and orientation. This can perhaps change with time and local en vironment. Other types of predictions are possible from comparisons of the HSP correlations inTable 15.1. For example, it can be determined that all the solv ents dissolving lignin are also predicted to swell psoriasis scales. This generality then suggests special care is in order when handling w oodimpregnating chemicals. The protective clothing chosen should ha ve HSP quite dif ferent from the HSP of the chemical in volved, as discussed in Chapter 13. An important ef fect that may ha ve been o verlooked in the solubility of w ood and w ood components is that there are acid groups present in hemicelluloses, for e xample, and these can be neutralized by bases. This gives an organic salt with high HSP.11 (See also Chapter 18.) Such a salt is hydrophilic and will collect w ater. This may lead to phase separation, and some destruction of ultrastructure is possible. This is an ef fect which is kno wn to ha ve caused blistering in coatings.

7248_C015.fm Page 282 Wednesday, May 9, 2007 9:37 AM

282

Hansen Solubility Parameters: A User’s Handbook

TABLE 15.4A Calculated Solubility Sphere for Lignin Solubility Solvent

δD

δP

δH

SOLUB

RED

V

Acetic acid Acetic anhydride Acetone Acetonitrile Acetophenone Aniline Benzaldehyde Benzene 1-Bromonaphthalene 1,3-Butanediol 1-Butanol Butyl acetate Butyl lactate Butyric acid gamma-Butyrolactone Butyronitrile Carbon disulfid Carbon tetrachloride Chlorobenzene 1-Chlorobutane Chloroform m-Cresol Cyclohexane Cyclohexanol Cyclohexanone Cyclohexylchloride Diacetone alcohol o-Dichlorobenzene 2,2-Dichlorodiethyl ether Diethylamine Diethylene glycol Diethylene glycol monob utyl ether Diethylene glycol monometh yl ether Diethyl ether Diethyl sulfid Di(isobutyl) ketone Dimethylformamide Dimethyl sulfoxide 1,4-Dioxane Dipropylamine Dipropylene glycol Ethanol Ethanolamine Ethyl acetate Ethylbenzene 2-Ethyl-1-butanol Ethylene glycol Ethylene glycol monob utyl ether Ethylene glycol monoeth yl ether Ethylene glycol monoeth yl ether acetate

14.5 16.0 15.5 15.3 19.6 19.4 19.4 18.4 20.3 16.6 16.0 15.8 15.8 14.9 19.0 15.3 20.5 17.8 19.0 16.2 17.8 18.0 16.8 17.4 17.8 17.3 15.8 19.2 18.8 14.9 16.6 16.0 16.2 14.5 16.8 16.0 17.4 18.4 19.0 15.3 16.5 15.8 17.0 15.8 17.8 15.8 17.0 16.0 16.2 15.9

8.0 11.7 10.4 18.0 8.6 5.1 7.4 0.0 3.1 10.0 5.7 3.7 6.5 4.1 16.6 12.4 0.0 0.0 4.3 5.5 3.1 5.1 0.0 4.1 6.3 5.5 8.2 6.3 9.0 2.3 12.0 7.0 7.8 2.9 3.1 3.7 13.7 16.4 1.8 1.4 10.6 8.8 15.5 5.3 0.6 4.3 11.0 5.1 9.2 4.7

13.5 10.2 7.0 6.1 3.7 10.2 5.3 2.0 4.1 21.5 15.8 6.3 10.2 10.6 7.4 5.1 0.6 0.6 2.0 2.0 5.7 12.9 0.2 13.5 5.1 2.0 10.8 3.3 5.7 6.1 20.7 10.6 12.6 5.1 2.0 4.1 11.3 10.2 7.4 4.1 17.7 19.4 21.2 7.2 1.4 13.5 26.0 12.3 14.3 10.6

0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 1* 0 0 0 1 1 0 0 1 1 1 0 0 0 1* 0 1 0

1.195 1.006 1.212 1.277 1.096 0.897 1.044 1.582 1.254 0.895 1.060 1.403 1.158 1.337 0.833 1.298 1.586 1.683 1.369 1.506 1.293 0.917 1.761 1.013 1.193 1.424 1.085 1.210 1.006 1.552 0.836 1.105 1.001 1.605 1.543 1.480 0.774 0.727 1.211 1.631 0.831 0.988 0.788 1.306 1.615 1.169 1.002 1.134 0.925 1.204

57.1 94.5 74.0 52.6 117.4 91.5 101.5 89.4 140.0 89.9 91.5 132.5 149.0 110.0 76.8 87.3 60.0 97.1 102.1 104.5 80.7 104.7 108.7 106.0 104.0 118.6 124.2 112.8 117.2 103.2 94.9 170.6 118.0 104.8 107.4 177.1 77.0 71.3 85.7 136.9 130.9 58.5 59.8 98.5 123.1 123.2 55.8 131.6 97.8 136.1

7248_C015.fm Page 283 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

283

TABLE 15.4A (CONTINUED) Calculated Solubility Sphere for Lignin Solubility Solvent Ethylene glycol monometh yl ether Furan Glycerol Hexane Isoamyl acetate Isobutyl isobutyrate Isooctyl alcohol Isophorone Mesityl oxide Methanol Methylal Methyl ethyl ketone Methyl isoamyl k etone Methyl isobutyl carbinol Methyl isobutyl ketone Morpholine Nitrobenzene Nitroethane Nitromethane 2-Nitropropane 1-Pentanol 1-Propanol Propylene carbonate Propylene glycol Pyridine Styrene Tetrahydrofuran Tetrahydronaphthalene Toluene 1,1,1-Trichloroethane Trichloroethylene Xylene Lignin

δD

δP

δH

SOLUB

RED

V

16.2 17.8 17.4 14.9 15.3 15.1 14.4 16.6 16.4 15.1 15.0 16.0 16.0 15.4 15.3 18.8 20.0 16.0 15.8 16.2 15.9 16.0 20.0 16.8 19.0 18.6 16.8 19.6 18.0 16.8 18.0 17.6

9.2 1.8 12.1 0.0 3.1 2.9 7.3 8.2 6.1 12.3 1.8 9.0 5.7 3.3 6.1 4.9 8.6 15.5 18.8 12.1 4.5 6.8 18.0 9.4 8.8 1.0 5.7 2.0 1.4 4.3 3.1 1.0

16.4 5.3 29.3 0.0 7.0 5.9 12.9 7.4 6.1 22.3 8.6 5.1 4.1 12.3 4.1 9.2 4.1 4.5 5.1 4.1 13.9 17.4 4.1 23.3 5.9 4.1 8.0 2.9 2.0 2.0 5.3 3.1

1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0* 1 0 0 0 0 0 0 0

0.906 1.372 1.128 1.904 1.447 1.516 1.237 1.125 1.268 1.076 1.479 1.274 1.411 1.279 1.464 0.986 1.054 1.254 1.286 1.260 1.143 1.117 1.513 0.944 0.987 1.421 1.163 1.392 1.538 1.500 1.298 1.524

79.1 72.5 73.3 131.6 148.8 163.0 156.6 150.5 115.6 40.7 169.4 90.1 142.8 127.2 125.8 87.1 102.7 71.5 54.3 86.9 108.6 75.2 85.0 73.6 80.9 115.6 81.7 136.0 106.8 99.3 90.2 123.3

D = 21.9 P = 14.1 H = 16.9 R

o

= 13.7 FIT = 0.990 NO = 82

UREA Data for the HSP correlation for urea solubility in or ganic solvents are given in Table 15.1. All of the parameters are rather high, which is characteristic of a lo w molecular weight solid. The data fit is ery good. Perhaps the most interesting thing about this correlation is that it clearly sho ws that additions of urea to w ater will improve solubility for a variety of materials including proteins. This is the reason for the improved solubility discussed previously in connection with the destruction of h ydrophilic bonding in proteins. The saturated solution of urea and w ater is also the best physically acting solvent for whole, dried blood that the author could locate in a pre vious (unpublished) study. The f act of high HSP for urea/w ater mixtures has led to its use in man y v aried types of products.7 The saturated solution of urea in water has found particular successes in the following examples.

7248_C015.fm Page 284 Wednesday, May 9, 2007 9:37 AM

284

Hansen Solubility Parameters: A User’s Handbook

TABLE 15.4B Calculated Solubility Sphere for Lignin Solubility Solvent 2-Pyrolidone Succinic anhydride Dimethyl sulfone Dimethyl sulfoxide Hexamethylphosphoramide o-Methoxyphenol 1,3-Butanediol Ethylene cyanohydrin Dimethyl formamide Diethylenetriamine Ethanolamine Diethanolamine Benzyl alcohol Furfuryl alcohol Dipropylene glycol gamma-Butyrolactone Diethylene glycol Phenol Ethylenediamine Allyl alcohol Triethyleneglycol 1,3-Butanediol Aniline 3-Chloro-1-propanol Ethylene glycol monometh yl ether N,N-Dimethyl acetamide Trimethylphosphate Benzoic acid m-Cresol Methyl-2-pyrrolidone 1,1,2,2-Tetrabromoethane Ethylene glycol monoeth yl ether Quinoline Propylene glycol Triethylphosphate 1,1,2,2-Tetrachloroethane Tetramethylurea Diethylene glycol monoeth yl ether Morpholine Pyridine Ethanol Furfural Hexylene glycol Propylene glycol monophen yl ether Diethylene glycol monometh yl ether Ethylene glycol 2,2-Dichlorodiethyl ether Acetic anhydride Tricresyl phosphate Cyclohexanol

δD

δP

δH

19.4 18.6 19.0 18.4 18.5 18.0 18.0 17.2 17.4 16.7 17.0 17.2 18.4 17.4 16.5 19.0 16.6 18.0 16.6 16.2 16.0 16.6 19.4 17.5 16.2 16.8 16.7 18.2 18.0 18.0 22.6 16.2 19.4 16.8 16.7 18.8 16.7 16.1 18.8 19.0 15.8 18.6 15.7 17.4 16.2 17.0 18.8 16.0 19.0 17.4

17.4 19.2 19.4 16.4 8.6 8.2 8.4 18.8 13.7 13.3 15.5 10.8 6.3 7.6 10.6 16.6 12.0 5.9 8.8 10.8 12.5 10.0 5.1 5.7 9.2 11.5 15.9 6.9 5.1 12.3 5.1 9.2 7.0 9.4 11.4 5.1 8.2 9.2 4.9 8.8 8.8 14.9 8.4 5.3 7.8 11.0 9.0 11.7 12.3 4.1

11.3 16.6 12.3 10.2 11.3 13.3 21.0 17.6 11.3 14.3 21.2 21.2 13.7 15.1 17.7 7.4 20.7 14.9 17.0 16.8 18.6 21.5 10.2 14.7 16.4 10.2 10.2 9.8 12.9 7.2 8.2 14.3 7.6 23.3 9.2 9.4 11.0 12.2 9.2 5.9 19.4 5.1 17.8 11.5 12.6 26.0 5.7 10.2 4.5 13.5

SOLUB

1

1 1

1 1 1

1 1 1

1

1 0*

1 1 1

1* 1 0 0 0

RED

V

0.599 0.609 0.665 0.727 0.758 0.761 0.766 0.769 0.774 0.785 0.788 0.792 0.800 0.821 0.831 0.833 0.836 0.839 0.865 0.866 0.878 0.895 0.897 0.902 0.906 0.911 0.913 0.915 0.917 0.918 0.919 0.925 0.929 0.944 0.965 0.968 0.973 0.981 0.986 0.987 0.988 0.989 0.998 1.000 1.001 1.002 1.006 1.006 1.008 1.013

76.4 66.8 75.0 71.3 175.7 109.5 87.5 68.3 77.0 108.0 59.8 95.9 103.6 86.5 130.9 76.8 94.9 87.5 67.3 68.4 114.0 89.9 91.5 84.2 79.1 92.5 115.8 100.0 104.7 96.5 116.8 97.8 118.0 73.6 171.0 105.2 120.4 130.9 87.1 80.9 58.5 83.2 123.0 143.2 118.0 55.8 117.2 94.5 316.0 106.0

7248_C015.fm Page 285 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

285

TABLE 15.4B (CONTINUED) Calculated Solubility Sphere for Lignin Solubility Solvent 1-Propanol Propylene carbonate Triethyanolamine Nonyl phenoxy ethanol Methyl salicylate Dimethyl phthalate Ethyl lactate Benzaldehyde Trifluoroacetic aci Di-(2-Chloro-isopropyl) ether Nitrobenzene Ethylene dibromide 1-Butanol 2-Propanol Methanol Bromoform Diacetone alcohol Ethylene carbonate Epichlorohydrin 2-Butanol Acetophenone Diethylene glycol monob utyl ether Methylene dichloride Benzyl butyl phthalate Acrylonitrile Formic acid Isophorone 1-Octanol Glycerol Formamide Ethylene glycol monob utyl ether Ethylene dichloride 1-Pentanol 1-Nitropropane Bromobenzene Ethylene glycol monometh yl ether acetate Ethyl cinnamate Propylene glycol monometh yl ether Diethyl phthalate Diethyl sulfate Butyl lactate Diethylene glycol he xyl ether Propylene glycol monoeth yl ether Propylamine Tetrahydrofuran 2-Octanol 2-Ethyl-1-butanol Isobutyl alcohol 2-Methyl-1-propanol 1-Decanol

δD

δP

δH

SOLUB

RED

V

16.0 20.0 17.3 16.7 16.0 18.6 16.0 19.4 15.6 19.0 20.0 19.2 16.0 15.8 15.1 21.4 15.8 19.4 19.0 15.8 19.6 16.0 18.2 19.0 16.4 14.3 16.6 17.0 17.4 17.2 16.0 19.0 15.9 16.6 20.5 15.9 18.4 15.6 17.6 15.7 15.8 16.0 15.7 16.9 16.8 16.1 15.8 15.1 15.1 17.5

6.8 18.0 22.4 10.2 8.0 10.8 7.6 7.4 9.9 8.2 8.6 3.5 5.7 6.1 12.3 4.1 8.2 21.7 10.2 5.7 8.6 7.0 6.3 11.2 17.4 11.9 8.2 3.3 12.1 26.2 5.1 7.4 4.5 12.3 5.5 5.5 8.2 6.3 9.6 14.7 6.5 6.0 6.5 4.9 5.7 4.9 4.3 5.7 5.7 2.6

17.4 4.1 23.3 8.4 12.3 4.9 12.5 5.3 11.6 5.1 4.1 8.6 15.8 16.4 22.3 6.1 10.8 5.1 3.7 14.5 3.7 10.6 6.1 3.1 6.8 16.6 7.4 11.9 29.3 19.0 12.3 4.1 13.9 5.5 4.1 11.6 4.1 11.6 4.5 7.1 10.2 10.0 10.5 8.6 8.0 11.0 13.5 15.9 15.9 10.0

0 0

1.013 1.015 1.018 1.021 1.026 1.028 1.034 1.044 1.044 1.052 1.054 1.059 1.060 1.066 1.076 1.077 1.085 1.088 1.090 1.095 1.096 1.105 1.112 1.113 1.116 1.121 1.125 1.125 1.128 1.129 1.134 1.136 1.143 1.144 1.144 1.145 1.149 1.149 1.149 1.154 1.158 1.160 1.160 1.162 1.163 1.163 1.169 1.169 1.169 1.171

75.2 85.0 133.2 275.0 129.0 163.0 115.0 101.5 74.2 146.0 102.7 87.0 91.5 76.8 40.7 87.5 124.2 66.0 79.9 92.0 117.4 170.6 63.9 306.0 67.1 37.8 150.5 157.7 73.3 39.8 131.6 79.4 108.6 88.4 105.3 121.6 166.8 93.8 198.0 131.5 149.0 204.3 115.6 83.0 81.7 159.1 123.2 92.8 92.8 191.8

0

0 0 0 0

0 0

0 0 0 0

0

0 0

7248_C015.fm Page 286 Wednesday, May 9, 2007 9:37 AM

286

Hansen Solubility Parameters: A User’s Handbook

TABLE 15.4B (CONTINUED) Calculated Solubility Sphere for Lignin Solubility Solvent

δD

δP

δH

Propylene glycol monoprop yl ether Dibutyl phthalate Dipropylene glycol meth yl ether Cyclohexanone Acetic acid Ethyl formate Trichlorobiphenyl Anisole Ethylene glycol monoeth yl ether acetate o-Dichlorobenzene 1,4-Dioxane Acetone Nonyl phenol Acetaldehyde 1,1-Dimethyl hydrazine bis-(m-Phenoxyphenyl) ether 2,4-Pentanedione Ethyl chloroformate Dibenzyl ether 2-Ethyl hexanol Isooctyl alcohol Tetrachloroethylene 2-(Diethylamino) ethanol Benzyl chloride Benzonitrile Ethyl bromide Nitroethane 1-Bromonaphthalene Naphthalene 2-Nitropropane Methyl acetate 2,2,4-Trimethyl 1,3-pentanediol monoisobutyrate Methylene diiodide Butylamine Mesityl oxide 1,1-Dichloroethylene Propionitrile Methyl ethyl ketone Acetonitrile Methyl isobutyl carbinol Ethanethiol Methyl methacrylate Nitromethane Chloroform Diethylene glycol b utyl ether acetate Butyronitrile Trichloroethylene Cyclohexylamine Methyl acrylate Ethyl acetate

15.8 17.8 15.5 17.8 14.5 15.5 19.2 17.8 15.9 19.2 19.0 15.5 16.5 14.7 15.3 19.6 17.1 15.5 17.3 15.9 14.4 19.0 14.9 18.8 17.4 16.5 16.0 20.3 19.2 16.2 15.5 15.1 17.8 16.2 16.4 17.0 15.3 16.0 15.3 15.4 15.7 17.5 15.8 17.8 16.0 15.3 18.0 17.2 15.3 15.8

7.0 8.6 5.7 6.3 8.0 8.4 5.3 4.1 4.7 6.3 1.8 10.4 4.1 8.0 5.9 3.1 9.0 10.0 3.7 3.3 7.3 6.5 5.8 7.1 9.0 8.0 15.5 3.1 2.0 12.1 7.2 6.1 3.9 4.5 6.1 6.8 14.3 9.0 18.0 3.3 6.5 5.5 18.8 3.1 4.1 12.4 3.1 3.1 9.3 5.3

9.2 4.1 11.2 5.1 13.5 8.4 4.1 6.7 10.6 3.3 7.4 7.0 9.2 11.3 11.0 5.1 4.1 6.7 7.3 11.8 12.9 2.9 12.0 2.6 3.3 5.1 4.5 4.1 5.9 4.1 7.6 9.8 5.5 8.0 6.1 4.5 5.5 5.1 6.1 12.3 7.1 4.3 5.1 5.7 8.2 5.1 5.3 6.5 5.9 7.2

SOLUB

0 0

0 0 0 0

0

0 0 0

0

0 0 0

0 0 0 0

0

RED

V

1.174 1.180 1.192 1.193 1.195 1.196 1.200 1.202 1.204 1.210 1.211 1.212 1.212 1.212 1.213 1.224 1.226 1.232 1.232 1.236 1.237 1.237 1.241 1.247 1.247 1.250 1.254 1.254 1.257 1.260 1.260 1.263 1.267 1.267 1.268 1.271 1.273 1.274 1.277 1.279 1.280 1.286 1.286 1.293 1.295 1.298 1.298 1.301 1.302 1.306

130.3 266.0 157.4 104.0 57.1 80.2 187.0 119.1 136.1 112.8 85.7 74.0 231.0 57.1 76.0 373.0 103.1 95.6 192.7 156.6 156.6 101.1 133.2 115.0 102.6 76.9 71.5 140.0 111.5 86.9 79.7 227.4 80.5 99.0 115.6 79.0 70.9 90.1 52.6 127.2 74.3 106.5 54.3 80.7 208.2 87.3 90.2 113.8 113.8 98.5

7248_C015.fm Page 287 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

287

TABLE 15.4B (CONTINUED) Calculated Solubility Sphere for Lignin Solubility Solvent Propylene glycol monoisob utyl ether Propylene glycol monob utyl ether Di(2-Methoxyethyl) ether Ethylene glycol b utyl ether acetate 1-Methyl naphthalene Bromochloromethane Butyric acid Diethyl ketone Ethyl acrylate Tributyl phosphate Diethyl carbonate Chlorobenzene Furan Dioctyl phthalate Di-iso-butyl carbinol Methacrylonitrile Tetrahydronaphthalene Butyl acrylate Butyl acetate Stearic acid Methyl isoamyl k etone Ethyl butyl ketone Octanoic acid Styrene Cyclohexylchloride Amyl acetate Butyraldehyde sec-Butyl acetate Ethyl amyl k etone Isoamyl acetate Biphenyl Dichloromonoflouromethan Propyl chloride Methyl butyl ketone Methyl isobutyl ketone Methyl amyl acetate Isobutyl acetate Methyl chloride Methylal Di(isobutyl) ketone Ethyl chloride Tridecyl alcohol 1,1,1-Trichloroethane 1,1-Dichlorethane 1-Chlorobutane o-Xylene Isobutyl isobutyrate Xylene Oleyl alcohol Toluene

δD

δP

δH

15.1 15.3 15.7 15.3 20.6 17.3 14.9 15.8 15.5 16.3 16.6 19.0 17.8 16.6 14.9 15.3 19.6 15.6 15.8 16.3 16.0 16.2 15.1 18.6 17.3 15.8 14.7 15.0 16.2 15.3 21.4 15.8 16.0 15.3 15.3 15.2 15.1 15.3 15.0 16.0 15.7 14.3 16.8 16.5 16.2 17.8 15.1 17.6 14.3 18.0

4.7 4.5 6.1 4.5 0.8 5.7 4.1 7.6 7.1 6.3 3.1 4.3 1.8 7.0 3.1 10.8 2.0 6.2 3.7 3.3 5.7 5.0 3.3 1.0 5.5 3.3 5.3 3.7 4.5 3.1 1.0 3.1 7.8 6.1 6.1 3.1 3.7 6.1 1.8 3.7 6.1 3.1 4.3 8.2 5.5 1.0 2.9 1.0 2.6 1.4

9.8 9.2 6.5 8.8 4.7 3.5 10.6 4.7 5.5 4.3 6.1 2.0 5.3 3.1 10.8 3.6 2.9 4.9 6.3 5.5 4.1 4.1 8.2 4.1 2.0 6.1 7.0 7.6 4.1 7.0 2.0 5.7 2.0 4.1 4.1 6.8 6.3 3.9 8.6 4.1 2.9 9.0 2.0 0.4 2.0 3.1 5.9 3.1 8.0 2.0

SOLUB

0

0 0

0 0 0

0 0

0

0

0 0

0 0 0 0 0

RED

V

1.313 1.317 1.318 1.330 1.331 1.336 1.337 1.346 1.351 1.356 1.366 1.369 1.372 1.372 1.374 1.389 1.392 1.395 1.403 1.408 1.411 1.417 1.418 1.421 1.424 1.427 1.428 1.432 1.434 1.447 1.450 1.451 1.462 1.464 1.464 1.465 1.470 1.473 1.479 1.480 1.485 1.486 1.500 1.502 1.506 1.512 1.516 1.524 1.535 1.538

132.2 132.0 142.0 171.2 138.8 65.0 110.0 106.4 108.8 345.0 121.0 102.1 72.5 377.0 177.8 83.9 136.0 143.8 132.5 326.0 142.8 139.0 159.0 115.6 118.6 148.0 88.5 133.6 156.0 148.8 154.1 75.4 88.1 123.6 125.8 167.4 133.5 55.4 169.4 177.1 70.0 242.0 99.3 84.8 104.5 121.2 163.0 123.3 316.0 106.8

7248_C015.fm Page 288 Wednesday, May 9, 2007 9:37 AM

288

Hansen Solubility Parameters: A User’s Handbook

TABLE 15.4B (CONTINUED) Calculated Solubility Sphere for Lignin Solubility Solvent Diethyl sulfid Diethyl amine Benzene Naphtha.high-flas Carbon disulfid Oleic acid Diethyl ether Triethylene glycol monoole yl ether Ethylbenzene Methyl oleate Dipropylamine Dibutyl stearate Triethylamine Trimethylbenzene Isopropyl palmitate Dibutyl sebacate cis-Decahydronaphthalene para-Diethyl benzene Mesitylene Carbon tetrachloride trans-Decahydronaphthalene Chlorodiflouromethan Cyclohexane Methyl cyclohexane Eicosane Trichlorofluoromethan Hexadecane Dodecane Mineral spirits Decane Nonane Octane 1,1,2-Trichlorotrifluoroethan Heptane Hexane Pentane Tetraethylorthosilicate Butane 2,2,2,4-Trimethylpentane Isopentane 1,2-Dichlorotetrafluoroethan Dichlorodifluoromethan Water Perfluoro(dimet ylcyclohexane) Perfluoromet ylcyclohexane Perfluoroheptan Bromotrifluoromethan Lignin

δD

δP

δH

SOLUB

RED

V

16.8 14.9 18.4 17.9 20.5 14.3 14.5 13.3 17.8 14.5 5.3 14.5 17.8 17.8 14.3 13.9 18.8 18.0 18.0 17.8 18.0 12.3 16.8 16.0 16.5 15.3 16.3 16.0 15.8 15.7 15.7 15.5 14.7 15.3 14.9 14.5 13.9 14.1 14.1 13.7 12.6 12.3 15.5 12.4 12.4 12.0 9.6

3.1 2.3 0.0 0.7 0.0 3.1 2.9 3.1 0.6 3.9 1.4 3.7 0.4 0.4 3.9 4.5 0.0 0.0 0.0 0.0 0.0 6.3 0.0 0.0 0.0 2.0 0.0 0.0 0.1 0.0 0.0 0.0 1.6 0.0 0.0 0.0 0.4 0.0 0.0 0.0 1.8 2.0 16.0 0.0 0.0 0.0 2.4

2.0 6.1 2.0 1.8 0.6 5.5 5.1 8.4 1.4 3.7 4.1 3.5 1.0 1.0 3.7 4.1 0.0 0.6 0.6 0.6 0.0 5.7 0.2 1.0 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.0 42.3 0.0 0.0 0.0 0.0

0 0 0

1.543 1.552 1.582 1.585 1.586 1.603 1.605 1.614 1.615 1.628 1.631 1.643 1.645 1.645 1.647 1.652 1.669 1.673 1.673 1.683 1.704 1.719 1.761 1.774 1.790 1.797 1.803 1.823 1.823 1.844 1.844 1.858 1.860 1.873 1.904 1.936 1.944 1.969 1.969 2.003 2.042 2.065 2.081 2.122 2.122 2.161 2.340

107.4 103.2 89.4 181.8 60.0 320.0 104.8 418.5 123.1 340.0 136.9 382.0 138.6 133.6 330.0 339.0 156.9 156.9 139.8 97.1 156.9 72.9 108.7 128.3 359.8 92.8 294.1 228.6 125.0 195.9 179.7 163.5 119.2 147.4 131.6 116.2 224.0 101.4 166.1 117.4 117.6 92.3 18.0 217.4 196.0 227.3 97.0

0 0 0 0

0

0

0

D = 21.9 P = 14.1 H = 16.9 RAD. = 13.7 FIT = 0.990 NO = 82

7248_C015.fm Page 289 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

289

--M1β4M1β4M1β4G1β4M1β4G1β4M1β4M1β4G– 2 332 3 6 Ac Ac β Ac Ac α (LIGN) 1 (LIGN) 1 M Ga (CELL) (CELL) FIGURE 15.5 Expected generalized sk etch of the configuration of cellulose, hemicelluloses, and lignin i wood cell w alls. See te xt or Reference 6 for further details. The sketch is for glucomannan. M is mannose monomer; G is glucose monomer; Ga is galactose monomer; Ac is an acetyl group; (LIGN) is a region similar in HSP to lignin (or acetal etc.); (CELL) is a re gion similar in HSP to cellulose, being an y of cellulose, hemicellulose backbone, or hemicellulose side chain with an alcohol group (M, Ga).

1. Lithographic stones were previously conditioned to make them more receptive to ink by application of this liquid to change wetting beha vior. 2. The saturated solution of urea and w ater, which swells and softens w ood, has been used to give wood fl xibility so that it can easily be formed. 3. It has been used by Eskimos to soften seal skins by swelling and softening them. A similar application in Me xico involves curing leather . This application probably originated in prehistoric times. 4. It has been used to impro ve the fl w of house paints on cold days or when no other source of liquid has been available (such as on a scaffold), as it is a good solvent miscible in many paints. 5. It is reported to ha ve been used to set hair , as it also softens and swells it. 6. It w as used in the early manuf acture of gunpo wder as a dispersion medium during grinding because of impro ved wetting for the po wder. 7. Amazonian Indians used this liquid to coagulate late x prior to sale and shipment. This was practiced particularly during World War II. Other unspecified and undocumented uses include those possible because the liquid has th ability to soften human skin, thus allo wing easier transport of medicinal chemicals into the body . Urea itself has HSP v ery close to those of sug ar and proteins. As all of these are biocompatible materials, it is clear that the incorporation of significant numbers of urea groups in, for xample, polyurethane polymers or other products, can greatly enhance biocompatibility .

WATER Water has been discussed in detail in Chapter 1. Briefly stated, one can use the HSP for ater or a correlation for w ater solubility to get a general e xplanation for observ ed phenomena. Accurate calculation of the HSP for solv ent–water mixtures cannot be expected because of the irregularities of w ater associating with itself, the solv ent, and a potential solute. Lindenfors 12 described the association of two molecules of water with one molecule of dimethyl sulfoxide, a solvent frequently mentioned in connection with biological systems. A simplistic approach based on the ratio of δH for w ater as a single molecule vs. that in the correlation(s) for w ater solubility suggests that (42.3/16.5)2 or about six water molecules are linked by hydrogen bonding into some type of entity. Various structures for assemblies of w ater molecules ha ve been discussed in the literature. The clusters with six water molecules are among the more probable ones.13 The data on water solubility used in the HSP correlations are reported by Wallström and Sv enningsen.14

7248_C015.fm Page 290 Wednesday, May 9, 2007 9:37 AM

290

Hansen Solubility Parameters: A User’s Handbook

SURFACE MOBILITY Surface mobility allows given segments of molecules to orient at surfaces in a direction where their HSP match more closely . The surfaces of h ydrophobic polymers (peat moss) can become h ydrophilic when contacted with w ater. One can speculate as to wh y this occurs. One possibility is that this phenomenon conserves water within the structure. Whenever water is present on an otherwise hydrophobic surf ace, it can become h ydrophilic if the surf ace molecules can rotate or mo ve hydrophilic entities toward the water. This allows the water to spontaneously spread and potentially enter the structure if there are suitable passages. When this is accomplished, and contact with water ceases, the surface dries and becomes h ydrophobic once more. The molecules rotate with a lo wer energy moiety toward the air. This hydrophobic surface helps prevent evaporation of water, as water is not particularly soluble in it, and the h ydrophilic segments oriented to ward the interior of the structure will help bind the w ater where it is. The basis of the orientation ef fects described earlier for hemicelluloses is another e xample of orientation to ward re gions where HSP matches better . These phenomena are also discussed in Chapter 18. It is also appropriate to repeat that solv ent quality has a great deal to do with pigment dispersion stability , in that the adsorbed stabilizing polymer should remain on the pigment surf ace. A solvent which is too good can remo ve it. This is discussed in detail in Chapter 5. The implication of these e xamples is that solv ent quality is v ery important for the orientation of molecules at interf aces. A change in solv ent quality can easily lead to a change in the configu ration of molecules at surfaces. It is not surprising that Nature has used this to advantage in various ways.

CHIRAL ROTATION, HYDROGEN BONDING, AND NANOENGINEERING It has been found that anthracene units appended to a single scre w-sense helical polyguanidine changed orientation when the temperature was increased beyond 38.5°C.15 The configuration foun above 38.5°C was the same as that found in tetrah ydrofurane. At temperatures lower than this, the orientation of the appended anthracene was that found in toluene.A mixture of tetrahydrofurane/toluene equal to 90/10 v ol% approximated the conditions at the critical temperature. F or those who have diligently read this handbook, it w ould appear ob vious that it is the cohesi ve energy density just above or just belo w the critical temperature that controls the structure. More specifically it i the set of HSP values that do this, as these reflect the mix of sources of the cohes ve energy density according to Equation 1.6 to Equation 1.8. There is also massi ve e vidence sho wing that the interactions can be interpreted as the dif ference in HSP using Equation 1.9. It is well kno wn that solubility limits can be passed by lo wering the temperature in some cases and by increasing it in other cases. When the cohesi ve energy density of the solv ent is higher than that of the polymer , solvency increases with increased temperature. When the cohesive energy density of the solvent is lower than that of the polymer, solvency decreases with increases in temperature. This is discussed in Chapter 2 and has been thoroughly treated by P atterson.16,17 In the present case the HSP of toluene are comparable to those of anthracene whereas tetrah ydrofuran has much higher v alues. δD, δP, and δH equal to 18.7, 4.1, and 3.3 for anthracene have been reported by a multiple regression technique based on its solubility in a lar ge number of solv ents.18 The corresponding v alues are 18.0, 1.4, and 2.0 for toluene and 16.8, 5.7, and 8.0 for tetrah ydrofurane. Thus increasing the temperature will increase the solv ency in tetrah ydrofurane to the point where it becomes able to cause the appended anthracene to adopt the same orientation as it has in toluene. As toluene has lower HSP than anthracene, the solv ency will decrease with increases in temperature. Extending this w ay of thinking to the problem of mo ving very large biological molecules — while they are being assembled, for example — is presumably one of controlling the local solubility. When the molecule is locally able to reside in the surrounding fluid, it can m ve much more readily

7248_C015.fm Page 291 Wednesday, May 9, 2007 9:37 AM

Hansen Solubility Parameters — Biological Materials

291

than when it is not. An insoluble molecule or molecular se gment will adsorb at a location where the ener gies (HSP) match, and where the geometry is also accommodating. This is most often called hydrogen bonding, b ut it must be all three (or more) types of cohesi ve ener gy that are collectively acti ve. The molecule or molecular se gment can be remo ved ag ain when it and the surrounding liquid ha ve a f avorable energy relation.

CONCLUSION Many materials of biological significance h ve been assigned HSP based on their interaction with a large number of solv ents whose HSP are kno wn. A correlation for solv ent effects on DN A has ranked the extent of these effects for different solvents in essentially the same order as that reported in an older study .9 This correlation for DN A can presumably be impro ved by additional data, b ut still reflects the magnitudes of the types of ene gies that are in volved in forming/destro ying the double helices. The δD;δP;δH found for DN A are 19.0;20.0;11.0, all in MP a1/2. This clearly sho ws that hydrogen bonding is by far the smallest of the energies involved in the noncovalent interactions that determine the DN A structure. A HSP correlation has been used to find predictably syne gistic solvent mixtures where tw o nonsolvents dissolve cholesterol when mix ed. The ethanol/aliphatic h ydrocarbon synergistic mixture is discussed as being of particular interest to the f ate of cholesterol in lipid layers. The HSP of chlorophyll and lignin are quite similar , indicating they will be compatible with v ery much the same kind of surroundings. The physical interrelationships for wood chemicals and wood polymers (lignin, hemicelluloses, and cellulose) are discussed. The side chains on hemicelluloses which contain alcohol groups and the hemicellulose backbone will be most compatible with cellulose and will orient to ward this. The hemicellulose side chains without alcohol groups (acetal, acid) are closer in HSP to lignin and will orient in this direction. The acetal side chains actually ha ve lower HSP than will dissolve lignin, for which reason they are expected to lie on the surface of the lignin or perhaps penetrate slightly into the lignin at v ery special local points where the HSP match is better than the a verage values seen o ver the lignin molecule as a whole. Molecular design of molecules or structures that change conformation with slight changes in the cohesive energy characteristics of given continuous media seems possible using HSP concepts. The changes are caused by preferred orientation of se gments of one conformation to ward the continuous phase, where its HSP match better , thus reducing the free ener gy of the system. If the cohesive energy characteristics of the continuous media change in a direction that no longer f avors this orientation, the molecule will change configuration to one where the free ene gy is lower. The attraction of the se gments not oriented to ward the continuous phase to neighboring molecules is commonly called hydrogen bonding in proteins and similar materials. This attraction is caused collectively by all the types of ener gy involved through the pre vailing difference in HSP and is a result of insolubility (rejection by) the continuous media. Geometrical considerations are clearly also a major f actor in addition to the cohesi ve energy density focused upon here. HSP analyses of relative affinities can be applied to a la ge number of other biological materials and may pro vide insights into relationships which are not readily ob vious or cannot be studied otherwise. The best situation is where the materials in question can be tested directly , otherwise the calculation procedures described in Chapter 1 can be used with some loss of reliability in the predictions.

REFERENCES 1. Hansen, C.M., The three dimensional solubility parameter — k ey to paint component af finities I Solvents, plasticizers, polymers, and resins, J. Paint Technol., 39(505), 104–117, 1967.

7248_C015.fm Page 292 Wednesday, May 9, 2007 9:37 AM

292

Hansen Solubility Parameters: A User’s Handbook

2. Hansen, C.M. and Andersen, B.H., The affinities of o ganic solvents in biological systems, Am. Ind. Hyg. Assoc. J., 49(6), 301–308, 1988. 3. Hansen, C.M., The universality of the solubility parameter , Ind. Eng. Chem. Prod. Res. Dev., 8(1), 2–11, 1969. 4. Ursin, C., Hansen, C.M., Van Dyk, J.W., Jensen, P.O., Christensen, I.J., and Ebbehoej, J., Permeability of commercial solv ents through living human skin, Am. Ind. Hyg. Assoc. J., 56, 651–660, 1995. 5. Hansen, C.M., 25 years with the solubility parameter (25 År med Opløselighedsparametrene, in Danish), Dan. Kemi, 73(8), 18–22, 1992. 6. Hansen, C.M. and Björkman, A., The ultrastructure of w ood from a solubility parameter point of view, Holzforschung, 52(4), 335–344, 1998. 7. Hansen, C.M., Solv ents for coatings, Chem. Technol., 2(9), 547–553, 1972. 8. Blake, R.D. and Delcourt, S.G., Thermodynamic effects of formamide on DN A stability, Nucl. Acid Res., 24, 2095–2103, 1996. 9. Ts’o, P.O.P., Helmkamp, G.K., and Sander, C., Interaction of nucleosides and related compounds with nucleic acids as indicated by the change of helix-coil transition temperature, Proc. Natl. Acad. Sci. U S A, 48, 686–698, 1962. 10. Hansen, C.M., Cohesion energy parameters applied to surface phenomena, Handbook of Surface and Colloid Chemistry, Birdi, K.S., Ed., CRC Press, Boca Raton, FL, 1997, pp. 313–332. 11. Hansen, C.M., Some aspects of acid/base interactions (Einige Aspekte der Säure/Base-W echselwirkung, in German), Farbe und Lack, 7, 595–598, 1977. 12. Lindenfors, S., Solubility of cellulose ethers (Löslichkeit der Celluloseeäther, in German), Das Papier, 21, 65–69, 1967. 13. Gregory, J.K., Clary, D.C., Liu, K., Bro wn, M.G., and Saykally , R.J., The Water Dipole Moment in Water Clusters, Science, 275, 1997, pp. 814–817. 14. Wallström, E. and Svenningsen, I., Handbook of Solvent Properties, Report T1-84, Scandinavian Paint and Printing Ink Research Institute, Hoersholm, Denmark, 1984. 15. Tang, H.-Z., No vak, B.M., He, J., and Pola varapu, P.L., A thermal and solv ocontrollable cylindrical nanoshutter based on a single scre w-sense helical polyguanidine, Angew. Chem. Int. Ed., 44, 7298–7301, 2005. 16. Patterson, D. and Delmas, G., Ne w aspects of polymer solution thermodynamics, Off. Dig. Fed. Soc. Paint Technol., 34(450), 677–692, 1962. 17. Delmas, D., Patterson, D., and Somcynsky, T., Thermodynamics of polyisobutylene-n-alkane systems, J. Polym. Sci., 57, 79–98, 1962. 18. Wu, P.L., Beerbower, A., and Martin, A., Extended Hansen approach: calculating partial solubility parameters of solid solutes, J. Pharm. Sci., 71(11), 1285–1287 (1982).

7248_C016.fm Page 293 Wednesday, May 23, 2007 11:36 AM

and Diffusion in 16 Absorption Polymers Charles M. Hansen ABSTRACT Predicting whether or not a gi ven chemical will attack a gi ven polymer is important. Hansen solubility parameters (HSP) ha ve been used for this purpose as discussed else where in this book. Consideration of the absorption and dif fusion of the chemical in the polymer is often required in addition to HSP in order to make reliable predictions, however. This has been discussed in particular in Chapter 12 through Chapter 14, where chemical resistance, barrier properties, and environmental stress cracking are treated in detail. Chemicals with smaller and more linear molecules absorb and diffuse more readily than those with lar ger and more b ulky structures. Surf ace resistance to absorption is sometimes so dominating that absorption does not occur in some cases, e ven though this might be e xpected based on simple HSP considerations. This chapter e xamines surface resistances in connection with absorption and dif fusion in polymers in order to help impro ve understanding of these f actors and to emphasize the necessity of simultaneous consideration of surf ace resistance when absorption rates and dif fusion within the bulk of the polymer itself are of interest. Methods to measure surf ace resistance and concentration-dependent dif fusion coef ficients ar discussed. Solving the diffusion equation with simultaneous consideration of surface resistance and with a concentration dependent dif fusion correctly models absorption, desorption, film formatio by solvent evaporation, and v arious forms of so-called anomalous dif fusion such as “time-dependent,” Case II, and Super Case II. Surf ace phenomena such as surf ace resistance to absorption deserve far more attention than has been gi ven in the past.

LIST OF SYMBOLS USED IN THIS CHAPTER (Please note that these are dif ferent from those used in the other chapters.) A B C CA Cs D D0 D1 D2 Dapp Dav Dmax Dv F FM FB Fa

Minimum cross-sectional area of molecule in Equation 16.20 Ratio of diffusion resistance to surf ace resistance. See Equation 16.13 Dimensionless concentration. See Equation 16.5 Concentration at break in curv e in Figure 16.1 Dimensionless surface concentration Diffusion coefficient. Preferred units are c 2/s Diffusion coefficient at zero concentration or l west concentration in an e xperiment Diffusion coefficient on xposed side of fil Diffusion coefficient on xit side of fil Apparent diffusion coefficien Average diffusion coefficien Maximum diffusion coefficient in an xperiment Increase in the dif fusion coefficient ver zero conditions for a gi ven situation Mass flu Correction factor for concentration dependent dif fusion in Equation 16.11 Correction factor for surf ace resistance in Equation 16.11 Correction factor for concentration dependent dif fusion in absorption e xperiments 293

7248_C016.fm Page 294 Wednesday, May 23, 2007 11:36 AM

294

Fd L Mt M∞ P Papp P∞ R Rd Ri Rs T X Vf Vt V2 b c cs c0 c1 c2 c∞ h hav k k2 l t t1/2 w x Δp ΦR

Hansen Solubility Parameters: A User’s Handbook

Correction factor for concentration dependent dif fusion in desorption e xperiments Film thickness Absorbed mass at time t Mass at equilibrium conditions Permeation coefficien Apparent permeation coefficien True permeation coefficien Radius of c ylindrical sample in Equation 16.24 Resistance to mass transport from dif fusion Resistances to permeation from sources 1, 2, 3, etc., in Equation 16.17 to 16.19 Resistance to mass transport by surf ace resistance(s) Dimensionless time given by Equation 16.3 Dimensionless distance Volume fraction of dif fusing material Constant in Figure 16.4 used to find local di fusion coefficients at Vf greater than 0.20 Diffusion coefficients at Vf equal 0.20 relative to diffusion coefficient at zero concentra tion (Figure 16.1 and Figure 16.4) Thickness of c ylindrical sample in Equation 16.24 Concentration of diffusing material Surface concentration Initial concentration Concentration on e xposed side of sample Concentration on e xit side of sample Concentration at equilibrium conditions Surface mass transfer coef ficien Average surface mass transfer coef ficien Constant in Equation 16.6 Constant in Equation 16.20 Length of sample in Equation 16.23 Time Time required to absorb (or desorb) one half of the equilibrium amount Width of sample in Equation 16.23 Distance Pressure difference of diffusing material across membrane Local concentration is Figure 16.6

INTRODUCTION Chapters 12 through 14 ha ve dealt with chemical resistance, en vironmental stress cracking, and barrier polymers, respectively. Absorption and diffusion of the chemicals into polymers are impor tant in each of these. F actors af fecting absorption and dif fusion in polymers are therefore of considerable importance and must frequently be included along with the Hansen solubility parameters (HSP) to mak e correlations and predictions for these phenomena. This chapter emphasizes the importance of surface resistance for absorption in polymers as this has been largely overlooked in the literature. The focus of much of the rele vant literature has been on anomalous dif fusion, but even in this context the influence of sur ace resistance has largely been neglected. Surface resistance can delay or pre vent the absorption of solv ents that should absorb readily based on simple HSP considerations. This could lead to a f alse sense of security based on short time testing only . Absorption requires some de gree of solubility . Therefore, equilibrium absorption can be expected to correlate with HSP. In studies of the absorption of a chemical into a polymer, its surface concentration is usually assumed to reach the equilibrium v alue immediately. As discussed in the

7248_C016.fm Page 295 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

295

following, this is not al ways true. Whatever the rate at which the surf ace concentration increases, absorption will proceed according to the la ws of dif fusion, Fick’s First La w, Equation 16.1, and Fick’s Second La w, Equation 16.2. The latter is often called the dif fusion equation. Its deri vation can be found in Crank. 1 For the sak e of simplicity, these equations are gi ven here for dif fusion in one dimension (x) only . For a constant dif fusion coefficient F = - D 0(∂c/∂x)

(16.1)

∂c/∂t = ∂/∂x (D 0∂c/∂x)

(16.2)

The diffusion equation is deri ved in a v ery general w ay and also accounts for concentration dependent diffusion coefficients when ver this is encountered. F is the mass flux, 0 is the (constant) diffusion coefficient or the di fusion coefficient at the l west concentration if there is concentration dependence (see below), c is the local concentration, x is the distance in the x dimension, and t is the time. The solutions to the diffusion equation given in this chapter use the dry film thickness a reference. This is because it is f ar simpler to k eep track of what is going on instead of continually adjusting local film thickness as a function of the amount of sol ent present. When a dif fusing solvent is present, for example, then the actual local thickness should be increased proportionately according to its local v olume fraction. The use of dimensionless v ariables to solv e these equations mak es the solutions more useful by making them applicable to all v alues of the combined v ariables. For this purpose the follo wing are defined Dimensionless time: T = D 0t/L2

(16.3)

X = x/L

(16.4)

C = (c – c 0)/(c∞ – c 0)

(16.5)

Dimensionless distance:

Dimensionless concentration:

L is the thickness of the plane sheet being considered. c 0 is the initial uniform concentration in the film. The (local) dimensionless concentrations rise from 0 to 1.0 in an absorption experiment where equilibrium absorption, c ∞, is finally obtained. For the sak e of completeness the predicted and e xperimentally confirmed xponential dependence of the dif fusion coefficient on concentration, D(c), will be introduced at this early stage. A more detailed discussion of its significance is g ven in a special section belo w. D(c) = D 0ekc = D 0Dv

(16.6)

“k” is a constant that is v alid up to a gi ven concentration as discussed belo w. D v is the increase in the diffusion coefficient ver that at zero conditions for a gi ven concentration. At the maximum concentration this becomes D max, and for a constant dif fusion coefficient it is 1.0 Using these variables, Equation 16.2 can be re written in dimensionless form as Equation 16.7. The dimensionless diffusion equation for an e xponential diffusion coefficient is ∂C/∂T = ∂/∂X (D v∂C/∂X)

(16.7)

7248_C016.fm Page 296 Wednesday, May 23, 2007 11:36 AM

296

Hansen Solubility Parameters: A User’s Handbook

The earliest edition of Crank’s monumental work1 has the advantage of many graphical solutions to the dimensionless dif fusion equation being presented with plots that can be used with a high degree of accurac y. These plots include numerous reference lines not found in later editions.

STEADY STATE PERMEATION In the absence of significant sur ace resistances, solving Equation 16.1 (or Equation 16.7) for steady state permeation in one direction only for a constant dif fusion coefficient g ves Equation 16.8. F = –D 0(c1 – c 2)L

(16.8)

The surface concentration on the exposed side, c1, is usually assumed to be c∞, and c2 is usually assumed to be zero. Initially , c 1 will be less than c ∞ when there is a significant sur ace resistance on the e xposed side, as discussed in the follo wing. Likewise, c 2 will be greater than c 0 if there is a significant sur ace resistance on the lo w concentration side of the film. The preferred units for these quantities are F in g/(cm 2×s), L in cm, c in g/cm 3, and D 0 in cm 2/s. It can be seen from Equation 16.8 that the surface concentration on the exposed side determines the concentration gradient over the film, assuming 2 is zero. This is a situation that seems to prevail in general, b ut exceptions are discussed belo w, and there will be significant sur ace resistance in presumably all cases as film thickness approaches zero. The surface concentrations, c 1 (or c s), will be higher for closer matches in HSP between challenge chemicals and polymers. Therefore, it is not surprising that HSP correlations can be made when permeation rates for a lar ge number of chemicals have been measured for a gi ven polymer, as reported in Chapter 13. This is particularly true when the molecules involved are all relatively small. When the molecular sizes of the challenge chemicals become too lar ge and/or their shape becomes suf ficiently complicated with side group and cyclic structures, simple HSP correlations are no longer possible. The diffusion coefficient i affected by these size and shape f actors, and the HSP can no longer be used as a single correlating parameter. In addition, surf ace resistances can also become v ery significant, as discussed bel w and in Chapter 14. The molecular v olume, V, of the challenge chemical has been used with some success to account for size ef fects, b ut this does not directly account for dif ferences in shape. Examples of correlations for dif fusion through chemical protecti ve clothing, for e xample, demonstrated that molecular size had to be tak en into account 2 (see also Figure 13.2 and the discussion in Chapter 13). The most reliable HSP correlations in these cases do not immediately consider the solvents with smaller molecules that may permeate f aster than e xpected by comparison with all the others. Likewise, improved understanding and correlations are obtained by initially ne glecting the solvents with larger molecules that do not permeate as fast as expected, in spite of close matches in HSP with those solvents that do permeate rapidly. Once a reliable HSP correlation is established without these ob vious outliers, their beha vior is better understood. Predictions then often become possible for other v ery small and/or v ery large molecular species as well.

THE DIFFUSION EQUATION CONSTANT DIFFUSION COEFFICIENTS Relevant solutions to Equation 16.2 or Equation 16.7 are used to measure dif fusion coefficients The diffusion equation must be solved with two boundary conditions and an initial condition. This discussion will consider a plane film xposed to absorbing chemical on tw o sides. The initial condition is chosen as a uniform concentration within a film, so that C is 0 for all X, r gardless of whether the initial concentration is zero or not. Dif fusion is also usually assumed to tak e place in one direction only, but side effects can become important in thicker samples as discussed below. The chemical concentration at the e xposed surf ace(s) is assumed to immediately rise to the

7248_C016.fm Page 297 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

297

equilibrium value. The second boundary condition is found at the middle of the free film xposed on two sides where there is no transfer in either direction. In other words Equation 16.1 is set equal to zero at this point. For a constant dif fusion coefficient, solving the di fusion equation with these conditions gi ves an initial straight-line absorption curv e as a function of the square root of time. This is generally called Fickian diffusion. The straight line passes through the origin, and the time required to absorb one-half of the equilibrium amount, t 1/2, can be used in Equation 16.9 to find 0. D0 = 0.049 L 2/t1/2

(16.9)

This equation is based on T having a value of 0.049 when half of the equilibrium amount has been absorbed. There is an identical result for desorption e xperiments where the time required for half of a uniformly absorbed material to lea ve the film also requires the same T value.

CONCENTRATION DEPENDENT DIFFUSION COEFFICIENTS Solutions for the diffusion equation have also been generated for concentration dependent diffusion coefficients for the same boundary and initial conditions as described ab ve. In this case it can be shown that Equation 16.7 reduces to Equation 16.10. 3–5 ∂Dv/∂T = D v(∂2Dv/∂X2)

(16.10)

This equation has been solv ed numerically man y times for dif ferent values of D max for both absorption and desorption with a uniform initial concentration. 3–5 Dmax is the ratio of the maximum diffusion coefficient found at the maximum concentration encountered in an xperiment to D0. The half-times calculated for both absorption and desorption were con verted to F a or F d, respectively, their ratio to the v alue 0.049. These values are given in Table 16.1. F a and F d are the F M for use in Equation 16.11. The correction f actors for desorption e xperiments were also found for the time required for only one fourth of the material to lea ve the film in desorption xperiments. This was necessary because of the extremely long experimental times (months) required for even this amount to leave, and also holds true for very thin films. These quarter-time correction factors for desorption were used to generate the results reported in Figure 16.1 and are discussed in the follo wing.6 The results reported in this figure confirm that a and F d are correct and useful. The same dif fusion coefficients were found by both absorption and quarte -time desorption measurements. The correction factors for the absorption measurements, Fa, were close to 2.0, whereas those for the quartertime desorption measurements, F d, were in the range of 40 to 144. D (c ) = FM × FB

0.049 × L2 tE

(16.11)

The factor F B in Equation 16.11 is a related correction accounting for an y surface resistance as discussed belo w. F B will al ways be greater than 1 as a surf ace resistance slo ws the transport process and leads to an apparent dif fusion coefficient that is too l w. The exponential increase in dif fusion coefficient with concentration is xpected based on free The main volume theory.7 It is be yond the scope of this chapter to include this theory in detail. feature of diffusion in polymers is that the macromolecular chains are barriers to transport. F actors that either promote mobility of the chain se gments or increase the distance between them will enhance the movement of smaller molecules. The transport occurs as very small movements rather than larger jumps. An increase in concentration of plasticizing smaller molecules leads to an increase in the free volume of the system, as the smaller molecules ha ve more free volume associated with

7248_C016.fm Page 298 Wednesday, May 23, 2007 11:36 AM

298

Hansen Solubility Parameters: A User’s Handbook

TABLE 16.1 Correction Factors for the Measurement of Concentration Dependent Diffusion Coefficients for Use in Equation 16.11 Desorption Dmax

(Fd)1/2

(Fd)1/4

100 2 5 101 102 103 104 105 106 107 108

1.00 1.56 2.70 4.00 13.40 43.30 138.7 443 1,370 4,300 13,670

1.00 1.55 2.61 3.84 10.20 23.10 47.40 89.0 160.5 290 506

Absorption (Fa)1/2 1.00 1.30 1.70 2.01 3.30 4.85 6.14 7.63 8.97 10.60 12.10

Note: (Fd)1/2 is for desorption half-times, (Fd)1/4 is for desorption quarter -times, and (Fa)1/2 is for absorption half-times.

them than do polymers. Thus diffusion coefficients increase as sol ent concentration increases. As can be seen in Figure 16.1 and Figure 13.1, the diffusion coefficient for common sol ents increases by a f actor of about 10 for an increase of solv ent concentration 0.03 v olume fraction at lo wer solvent concentrations. This corresponds to slightly more than doubling the dif fusion coefficien for each added 0.01 v olume fraction of solv ent. Concentration dependence in measuring dif fusion coef ficients by absorption can also b accounted for by the method of integrals given by Crank.1 Treatment of the experimental data given by Crank with the correction f actors given in Table 16.1 leads to e xactly the same result for the exponential dif fusion coef ficients of chloroform in polystyrene as as found by the method of integrals. Both procedures require iterations as the true v alue of D max is not kno wn initially and must be estimated to find a (and a ne w D max) until convergence is obtained.

SURFACE RESISTANCE MATHEMATICAL BACKGROUND The diffusion equation must be solv ed with the appropriate boundary conditions at the surf aces when significant sur ace resistances are encountered. For a film xposed on two sides, the absorption process then can be modeled with the follo wing boundary condition at the surf aces: F = h (c∞ − cs )

(16.12)

The surface mass transfer coefficient, h, has preferred units of cm/s. The surface concentration at any given time is c s. One estimate of h can be obtained by plotting the weight g ain against time. The limiting slope at time approaching zero is used to find . As c s is zero at time equal to zero, h can be estimated from this initial flux d vided by c ∞.

7248_C016.fm Page 299 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

299

- LOG diffusion coefficient at 20 °C, cm²/sec

6 Isotope

8 Absorption Fa = 1.8

F = Fa x FB = 1.2 x 250 = 300

F = Fa x FB = 1.3 x 1.25 = 1.63

Fd = 144 10

Desorption (to vacuum) 12 Fd = 40

14 0.1

0.2

0.3

0.4

0.5

0.6

Vf FIGURE 16.1 Diffusion coef ficients for chlorobenzene in polyvi ylacetate measured by absorption halftimes, desorption quarter-times, and isotope experiments as a function of the volume fraction of chlorobenzene, Vf.6 The lower curve is based on dry polymer content. The upper curve is based on total thickness. Corrections for surface resistance, F B, are also required for Vf above about 0.2 v olume fraction. In an e xtreme case for absorption with Vf near 0.5, the correction for surface resistance was a factor of 250.21 (Reprinted from Hansen, C.M., Prog. Org. Coat., 51(1), 55–66, 2004. With permission from Else vier.)

It is useful to rewrite this boundary condition in dimensionless terms using the quantity B. This is the ratio of dif fusion resistance, R d, to that of surf ace resistance R s. Thus, B = R d/Rs = (L/D 0)/(1/h) = hL/D 0

(16.13)

Large B is indicati ve of a dif fusion-controlled process. F or B = 1, R d = R s, and for lo wer B surface resistance dominates. Surf ace resistance becomes increasingly important as the film thick ness decreases. The dimensionless boundary condition corresponding to Equation 16.12 for use with Equation 16.7, considering the x direction only is ∂C/∂X = B(1 – C s)

(16.14)

For an e xponential concentration dependence of the dif fusion coefficient, this boundary con dition can be used with Equation 16.10 as 3,5 ∂Dv/∂X = BlnD v

(16.15)

7248_C016.fm Page 300 Wednesday, May 23, 2007 11:36 AM

300

Hansen Solubility Parameters: A User’s Handbook

TABLE 16.2 Correction Factors, FB, for Use with a Constant Diffusion Coefficient in Equation 16.11 B

1/B

FB

∞ 10 2 1 0.5 0.1

0 0.1 0.5 1 2 10

1.0 1.45 3.14 4.95 6.8 37.5

Note: This means that F equal to 1.0 in this case.

SURFACE RESISTANCE

IN

M

is

ABSORPTION EXPERIMENTS

Solutions to the diffusion equation have been presented for v arious surface resistances (surface conditions) by Crank for a constant diffusion coefficient 1 FB has been evaluated from these graphical results as the ratio of the half-time for a gi ven B value to 0.049. These results are reported in Table 16.2. Equation 16.16 can be used for B less than about 0.5. At higher B v alues it is not e xact. FB = (3.75/B) +1

(16.16)

As stated abo ve, when surf ace resistance can be ne glected, a plot of the (relati ve) uptake vs. the square root of time initially is a straight line that passes through the origin. When surf ace resistance becomes important, the delayed uptak e can be seen as a form of time-lag phenomena, with a clear “S” shape. It is also interesting to note at this point that solutions to the dif fusion equation with the boundary condition of an e xponential increase of the surf ace concentration with time give absorption curves with exactly the same “S” shapes. Numerical solutions to the diffusion equation of the type described in Reference 3 and Reference 5 confirm that significant su ace resistance leads to an e xponential increase in the surf ace concentration. The factors leading to and controlling the pre vailing surface concentrations are of major interest and not necessarily the f act that these increase in an e xponential manner with time. Figure 16.2 sho ws S-shaped curv es for absorption in the COC polymer , Topas® 6013 from Ticona. Surface resistance is significant in all three cases sh wn, as can be seen by the S-shaped curves that do not pass through the origin.8 The apparent B values for these cases are 10 for ethylene dichloride, 0.5 for dieth yl ether , and 20 for prop yl amine. Solv ent absorption w as follo wed in injection-molded samples for 13 solv ents in c yclic olefinic copolymer (COC), 4 sol ents in tw o different grades of polycarbonate (PC), and 2 solvents in the terpolymer acrylonitrile/butadiene/styrene (ABS). It was discovered that a surface resistance to absorption was significant in 19 of thes 23 cases. Approximate surf ace mass transfer coef ficients and approximate di fusion coef ficient were determined where possible. There is no significant sur ace resistance to absorption in those cases where the absorbing molecules are smaller and linear such as for tetrahydrofurane, n-hexane, and 1,3-dioxolane in the COC polymer , and b utyric acid in PC (Le xan® 104R, General Electric). When the challenge molecules are too large or bulky, no absorption occurs. Such systems included 1,4-dioxane, methyl isobutyl ketone, acetophenone, and phenyl acetate in the COC polymer. There

7248_C016.fm Page 301 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

301

Ethylenedichloride in COC

Diethylether in COC

Propylamine in COC

Weight change [mg/g]

300 250 200 150 100 50 0 0

25

50

75

100 125 Time [sqrt(min)]

150

200

200

FIGURE 16.2 Absorption of ethylene dichloride, diethyl ether, and n-propylamine in a COC polymer, Topas® 6013, Ticona, at 23 °C. (Reprinted with permission from Nielsen, T.B. and Hansen, C.M., Ind. Eng. Chem. Research, 44(11), 3959–3965, 2005. Cop yright 2005 American Chemical Society.)

was no weighable absorption, e ven though HSP w ould predict this. The molecules can simply not get through the surf ace layer , and its resistance is therefore ef fectively infinitely la ge. Other examples of lack of absorption are oleic acid that simply does not absorb in the PC polymers or ABS. Between these extremes are situations where surface resistance clearly affects the absorption process. Surface resistance becomes significant when molecules can be transported way from the surface into the bulk of the polymer faster than they can be adsorbed/absorbed just at/in the surface.

SURFACE RESISTANCE

IN

PERMEATION EXPERIMENTS

As stated above, Equation 16.13 clearly shows that the importance of surface resistance will increase as film thickness decreases, and vice ersa. This can be used in permeation measurements to fin the true permeation coef ficient as well as the sum of all other resistances to the transport process One measures apparent permeation coefficients app, for different film thicknesses and xtrapolates the inverse of the apparent transport coef ficient ersus the in verse of the film thickness to zero corresponding to infinite film thickness This is portrayed in Figure 16.3 for the permeation of water through an acrylic coating. 9 If a single measurement had been made at a film thickness o 40 microns, which is a normal film thickness, the apparent permeation coe ficient ould have been one-half that of the true permeation coef ficient The data in this type of figure can be interpreted using the foll wing set of equations: F = Δp/(L/Papp) = Δp/(L/P∞ + R 1 + R 2 + R 3 ….)

(16.17)

L/Papp = L/P ∞ + R 1 + R 2 + R 3 ….

(16.18)

1/Papp = 1/P ∞ + (R 1 + R 2 + R 3 ….)/L

(16.19)

The extrapolation to 1/L equal to zero gives the inverse of the true permeation coefficient, 1/ ∞, and the slope gi ves the sum of the resistances. Δp is the o verall vapor pressure dif ference in the system, and the gi ven R represent dif ferent sources of resistance. Other extrapolations are possible to g ain more information, depending on the situation. 10 One can find the apor diffusion coefficient (resistance), for xample, by varying the amount of liquid

7248_C016.fm Page 302 Wednesday, May 23, 2007 11:36 AM

302

Hansen Solubility Parameters: A User’s Handbook

1 x 10-12 Papp 20

15

10

P∝

5

05

10

15

20

25

1 x 10-3 L

FIGURE 16.3 A plot of the in verse of the apparent permeability coef ficient ( app in kg P a1 s1 m1) vs. the inverse of the film thickness (L in m) for an acrylic coating. Extrapolation to 1/L to zero g ves the inverse of the true permeability coef ficient. (Reprinted from Huldén, M. and Hansen, C.M., Prog. Org. Coat., 13(3/4), 171–194, 1985. With permission from Else vier.)

in a cup-type e xperiment. It has been found that these surf ace and vapor diffusion effects must be accounted for in cup-type and related e xperiments with thinner polymer films and porous types o materials. These include thinner paint films, ood in the fiber direction, and pape , for example.9–11 This type of e xperiment using paper as a film separated resistances for the permeation of th paper by water as well as for dif fusion of water in air, (heat transfer) evaporation of the water, and an estimate of the surf ace resistances on the tw o sides of the paper .10

SURFACE RESISTANCE — A DISCUSSION In the literature surface resistance has alternatively been called skin layer effect, surface condition, interfacial resistance, or boundary layer resistance, and leads to what is often called an induction time and “time-dependent” diffusion. In addition to the discussion and literature cited above, surface effects ha ve been noted in man y studies of dif fusion in polymers. The following also deal with surface effects in the absorption of solv ents into polymers. A skin layer was found on the surface of injection molded polypropylene.12 NMR microimaging was used to confirm that carbon tetrachloride absorption as retarded by this layer . The rapid surface cooling in the injection molding process gi ves a surf ace different from the interior , where the cooling rate (and orientation) is dif ferent. This effect is presumably found with man y injection molded polymers. McDonald et al. also arri ved at the conclusion that surf ace flux limited di fusion of solv ent into polymer could e xplain observ ed beha vior such as Case II and transitions between Fickian

7248_C016.fm Page 303 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

303

diffusion and Case II dif fusion.13 Toluene diffusion in polystyrene is discussed in detail. Case II diffusion involves a linear absorption curv e when time is used rather than the customary square root of time. This is discussed in more detail belo w. Shankar studied interf acial resistance in absorption e xperiments.14 The conclusions were that non-Fickian characteristics can result from slo w transfer to the surf ace layer, and that some of the observed features of anomalous sorption can be e xplained with the help of this model. Surf ace resistance was found to be particularly important in thinner sheets. Systematic variations in absorption phenomena with film thickness are a clear indication of a significant su ace resistance. When there is a surf ace resistance, the surf ace concentration only slo wly rises to the equilibrium v alue. Surface resistance in the meth yl iodide-cellulose acetate system produced the characteristic Sshaped absorption curv e. Characteristic S-shaped absorption curves were found in the methylene chloride-PEEK system by Gryson et al. 15 The initial rate of absorption w as found to be strongly dependent on the surf ace condition but the equilibrium v alues were not. Surface resistance w as shown to af fect permeation through polymer films by Kim and Kam mermeyer who measured actual concentration profiles in Nylon-6, cellulose acetate, and polyeth ylene membranes. 16 Water permeation through Nylon-6 films as studied for dif ferent film thick ness. It w as clearly sho wn that the surf ace concentration of w ater did not reach the equilibrium value at the equilibrium permeation rate unless the film thickness as greater than 0.05 cm at 35°C. Similar studies sho wed that there were also significant sur ace resistances for dioxane, benzene, and n-hexane in polyeth ylene, and for w ater in cellulose acetate. Hwang and Kammerme yer showed significant sur ace resistance by studying permeation as a function of film thickness for ater in acetyl cellulose acetate and Nylon-6, h ydrogen through stainless steel, and p-dioxane through Nylon-6 and polyeth ylene.17 Skaarup18 performed man y permeation e xperiments on Polyamide 6 (P A 6) (B ASF Ultramid B4) and polyvin ylacetate (Hoechst, Mo wilith 50) where surf ace resistance w as significant. O particular interest are the studies on P A 6 where the permeation of eth yl laurate, 2,4-dimeth yl-2pentanol, benzyl alcohol, n-butanol, ethyl acetate, and n-pentane was studied. Skaarup arri ved at the following general relation for the permeation of these liquids through P A hav = k 2(Dav/A1/2)

(16.20)

hav (in cm/s) is the average surface mass transfer coefficient, av (in cm 2/s) is the average diffusion coefficient in the film (see Equation 16.21), an A (in cm 2) is the (minimum) cross-sectional area of the molecule in question. k 2 is a constant with the v alue approximately 4.5(10) -6. This equation indicates that a molecule in the surf ace of a polymer will proceed inw ard in direct proportion to the diffusion coefficient. It li ewise indicates that the greater the cross-sectional area of the molecule, the more difficulty it will h ve to reside at the surface in a condition where it can take a small jump into the b ulk. h on the e xposed side of the films as approximately 1/3 of that on the e xit side. The reason for this is thought to be that the orientation of molecules leaving the film is directe more toward the e xit surface, whereas the orientation of molecules approaching the entry side is more random. A molecule landing “sideways” and hitting a polymer segment will be rejected. The right orientation at the right place allo ws adsorption/absorption. The significance of the sur ace resistances can be demonstrated by the film thickness at which the resistance from di fusion within the film is equal to the sum of the t o surface resistances. This was about 180 microns or higher for the linear aliphatic molecules, and increased for more complicated structures. Dav can be found from Equation 16.21 as the log arithmic mean where D 1 is the dif fusion coefficient on the xposed side and D 2 is the dif fusion coefficient on the xit side. Dav = (D 1/D2 – 1)/ln(D 1/D2)

(16.21)

7248_C016.fm Page 304 Wednesday, May 23, 2007 11:36 AM

304

Hansen Solubility Parameters: A User’s Handbook

This equation shows how much the average diffusion coefficient di fers from that at essentially zero concentration corresponding to D 2. If the ratio D 1/D2 is 10 then D av is 3.9 times lar ger than D1. This implies that concentration dependent diffusion can lead to significant errors when difusion and permeation coefficients at ery low concentrations are needed. D 1/D2 equal to 10 implies that the concentration dif ference is only about 0.03 v olume fraction across the membrane when rigid polymers are involved. For elastomers this same ratio implies a concentration dif ference near 0.15 volume fraction, as judged from the dif fusion coefficient data reported in Figure 16.1 at concen trations above about 0.20 volume fraction where the system is above its glass transition temperature. These same considerations are v alid for permeation coef ficients since P = DS

(16.22)

P is the permeation coef ficient measured at the g ven steady state concentration dif ference, (c1 – c2), and S is the solubility coefficient. If 1 is 10 times greater than D2 then P is approximately 3.9 times lar ger than w ould ha ve been measured for an e xtremely lo w concentration dif ference across the membrane.

SIDE EFFECTS Corrections for absorption or desorption from the sides of thick er samples should be applied to measured diffusion coefficients if there is significant d fusion through the sides of thicker samples. Equation 16.23 for dif fusion in plane samples of isotropic media can be used for this purpose. 19 D0 = D app/(1 + L/l + L/w) 2

(16.23)

Dapp is the apparent dif fusion coefficient found from initial slope measurement 1 or Equation 16.9 if there is a linear absorption curv e as f ar as t 1/2 when the square root of time is used. The length and width of the film are l and , respectively, with L being retained as the thickness. It can be seen that D 0 and D app are equal to the e xtent that the second and third terms in Equation 16.18 are not significant. Co ventional tensile bars, for example, have significant side e fects when they are used to measure diffusion coefficients. The uptake is more rapid than otherwise anticipated, so the apparent diffusion coefficients must be reduced accordingl . Tensile bars that are 4 mm thick and 10 mm wide will require corrections that are at least as great as a f actor of 1.96. A square sample that is 10 mm on each side and 1 mm thick requires a correction for end ef fects equal to 1.44. For c ylindrical geometry a deri vation similar to that used to find Equation 16.23 results i Equation 16.24. D0 = D app/(1 + b/R) 2

(16.24)

R is the radius of the c ylinder and b is its thickness. These equations are based on the initial uptake (or loss) to e valuate D app. Here ag ain, an easy procedure is to e xtrapolate the initial slope on a plot of uptak e versus the square root of time to a fict ve t 1/2 found when the relati ve uptake is equal to 0.5. This value for t 1/2 can be used in Equation 16.9 to find app.

MEASURING DIFFUSION COEFFICIENTS CONCENTRATION DEPENDENCE

WITH

SURFACE RESISTANCE

AND

It was necessary to measure diffusion coefficients ver as wide a range as possible in order to solve the dif fusion equation to simulate film formation by sol ent e vaporation.3,20 This w as done by comparison of e xperimental results with solutions to the dif fusion equation accounting for both

7248_C016.fm Page 305 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

305

concentration dependence and surf ace resistance at the same time. These techniques are reported elsewhere.3–5,21 The technique used w as to interpret the e xperiments initially assuming that the diffusion coefficient as a constant at the concentration assigned to the e xperiment. Corrections to this estimate (multiplying f actors) were then made by comparison with suitable solutions to the diffusion equation using the data in Table 16.1 in Equation 16.11. The correction f actors, F a, at low concentrations were close to 1.5–2.0 (corresponding to increases of from 50 to 100%) for the usual step-wise absorption e xperiments from one concentration to a slightly higher one. The correction factors, Fd, were well above 100 for desorption measurements from an equilibrium state near 15–20%v ol solv ent to v acuum. The results in Figure 16.1 sho w that the same dif fusion coefficients were found in both cases. Absorption e xperiments only were rele vant abo ve about 20%vol because there is a marked change in diffusion behavior at concentrations above and below this value (for the system chlorobenzene and polyvin ylacetate). When concentrations reach higher than about 20%v ol, the correction f actors for surf ace resistance, F B, increase rapidly . They can become as high as 100 or more. 21 The total dif fusion coef ficient cur e from 0 to 100% solv ent (chlorobenzene) was completed with a self-dif fusion coefficient and s veral isotope experiments.3 The isotope experiments are for very high solvent concentrations that correspond to liquid lacquer formulations. The solvent-in-polymer diffusion coefficient for chlorobenzene in polyvi ylacetate increases more than 9 decades as the solvent concentration increases from essentially zero to 100%, where the self-diffusion coefficient is 1.65(10 5 cm 2/s. These coordinated absorption and desorption experiments were necessary to determine the true diffusion coefficients at all concentrations so that the solutions to the di fusion equation for fil drying by solvent evaporation were correct.3,20 It is also necessary to consider concentration dependence simultaneously with surf ace resistance in man y cases of practical importance. Surf ace resistance has been present in many studies reported in the literature and clearly affects many results involving diffusion in polymers, b ut only rarely has there been mention of this f act. This situation naturally led to solving the dif fusion equation with rele vant values for dif fusion coefficients an surface resistances for additional situations of interest. The overall result was a simple explanation for the various types of so-called anomalous diffusion. It is the balance between the (concentrationdependent) diffusion resistance and the surf ace resistance that determines whether the dif fusion is "Fickian" or whether it is presumed to be anomalous. This is discussed in the ne xt sections.

FILM FORMATION BY SOLVENT EVAPORATION The process of film formation by sol ent evaporation has been fully described using the dif fusion equation with a significant sur ace resistance and local dif fusion coefficients found from the dat in Figure 16.1. Loss of solv ent was followed experimentally for about 2 years from an initial Vf of 0.75 until it w as totally lost. 3,20 Surface resistance was significant at concentrations ab ve about 0.2 v olume fraction solv ent in the system that w as studied most e xtensively (chlorobenzene in polyvinylacetate). This was also confirmed by the calculations since the sur ace concentration fell to essentially zero when this amount was reached, and the continued loss of solvent was controlled by internal diffusion to the surface. In this situation h is af fected by factors such as vapor pressure and latent heat of the solv ent, heat transfer to the surf ace, and air v elocity past the surf ace. The local diffusion coefficient changed from 5(10 –7 cm 2/s initially to 1(10) –14 cm 2/s at zero concentration. The experimental and calculated curv es are reported in Figure 16.4. The process of solv ent loss tak es place in tw o distinct stages. Surf ace resistance controls the first stage, whereas the second stage is controlled by the rate at which sol ent molecules can diffuse to the film/air sur ace in order to e vaporate. The quantity V2 in Figure 16.4 is gi ven as 10 6. As can be seen in Figure 16.1, where dif fusion coefficients are reported for the same system, this is the xtent of the v ariation of the dif fusion e. Vt is a coefficient from 0 concentration up to A, the concentration at the break in the curv fictitious di fusion coef ficient to calculate the di fusion coef ficients at concentrations ab ve C A.

7248_C016.fm Page 306 Wednesday, May 23, 2007 11:36 AM

306

Hansen Solubility Parameters: A User’s Handbook

Volume Solvent / Volume Polymer

101

V2= 106 Vt= 1010 CA= 0.2 B as indicated Exptl. 165 microns

10

5

10-1

B=10 Exptl. B=106 7 22 microns B=10 ~MO –CA –CA CS= O CS= O For B=107 For B=106

CS= O For B=105 Calculated Experimental Effect of water - a steeper slope One day L=30 microns

-2

10

10-7

10-6

10-5

10-4 Dot Dimensionsless T, (L)2

10-3

10-2

FIGURE 16.4 Calculated and experimental drying curves for the evaporation of chlorobenzene from polyvinyl acetate.3,20 See the discussion in the accompan ying text.

Diffusion coefficients were calculated for each local site in the films The iterative procedure at each successi ve time interv al w as that described in Crank 1 as the Crank–Nicolson method. The films were d vided into a suf ficient number of finite d ference elements to assure correct results. Film drying in a climatized room w as f aster than film drying in a acuum apparatus, where diffusion coef ficients were measured. Absorbed w ater plasticizes the film. The calculated and measured desorption curves under vacuum coincided at long times (several months). Vacuum does not hasten release of solv ent by dif fusion at longer times. It only mak es certain that the surf ace concentration is zero, which it will be in almost all cases an yway.

ANOMALOUS DIFFUSION (CASE II, SUPER CASE II) For absorption with a constant dif fusion coefficient one normally finds that the initial weight ain is linear with the square root of time as stated abo ve. This is called Fickian or normal dif fusion. In Case II diffusion, the weight increases linearly with time. This is largely a result of concentration dependent diffusion coefficients with a la ge jump in concentration. This is true e ven when there is v ery little surf ace resistance of significance 5 Super Case II emer ges as the surf ace resistance becomes more significant relat ve to the dif fusion resistance. 5 Uptake is linear with time early in the process, b ut at some longer time the rate of g ain increases mark edly. See Figure 16.5. This change occurs when the dif fusing material reaches the middle of a sheet e xposed on both sides, for example. This can be seen in Figure 16.6 where concentration gradients ha ve been calculated for Super Case II behavior. Simultaneous diffusion and surface resistance combine in a special way to produce this beha vior. The chronology of the e vents is that surf ace concentration increases as time goes on. There is an adv ancing front into the film that is som what more pronounced than that sho wn in Figure 16.6, because the distances in this figure are based on dry film thicknes When the concentration at the center of the film starts to increase ab ve zero, the rate of absorption also begins to increase. The concentration profiles ultimately become flat at moderate elapsed tim as diffusion within the film is n w rapid compared to the surface resistance. The diffusion resistance

7248_C016.fm Page 307 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

307

1.0 B:

0.8 Mt / M∝

109 108

0.6

107

0.4 0.2 106 0.0 0.0

0.2

0.4

0.6 0.8 6 T x 10

1.0

1.2

1.4

FIGURE 16.5 Solvent uptake curves for v arious values of B and concentration dependent dif fusion coeffi cients equal to those reported for chlorobenzene in polyvinylacetate (See Figure 16.1). B is the ratio of diffusion resistance to surf ace resistance, including concentration dependence. (Reproduced from Hansen, C.M., Diffusion in Polymers, Polym. Eng. Sci., 20(4), 252–258, 1980. With permission from the American Chemical Society.)

1.0 0.869 0.8

ØR

0.6 0.4 0.2 0.0 0.0

0.037

0.125

0.098

0.250

0.146

0.375

0.216

X

0.500

0.319

0.625

0.386

0.750

0.467

0.875

0.562

1.0

FIGURE 16.6 Concentration gradients for conditions corresponding to Super Case II type (anomalous) diffusion. Values on the curv es are relati ve weight g ains. ∅R is the local concentration. (Reproduced from Hansen, C.M., Polym. Eng. Sci., 20(4), 252–258, 1980.With permission from theAmerican Chemical Society.)

becomes less and less as the concentration in the middle of the film increases, and the rate of uptae increases. There is still another ef fect at very long times where the surf ace resistance begins to be more important ag ain relative to the dif fusion resistance. The rate of weight g ain decreases just before the equilibrium v alue is attained. The concentration gradients in Figure 16.6 are for the curv e in Figure 16.5 for B = 10 7. This is a typical Super Case II type curv e with initial absorption linear with time follo wed by a sudden increase in the absorption rate as the solv ent reaches the center of a film xposed on tw o sides. The final l veling off at very long times is also characteristic of Super Case II. The absolute value of B depends on the de gree of concentration dependence o ver the selected concentration interv al

7248_C016.fm Page 308 Wednesday, May 23, 2007 11:36 AM

308

Hansen Solubility Parameters: A User’s Handbook

in the given case. See Reference 5 for greater detail. The parameters chosen to calculate this curv e are realistic.

GENERAL COMMENTS Discussion of the v arious other e xplanations for anomalous dif fusion is be yond the scope of this chapter. These discussions are focused on time-dependent replication of the observ ed dif fusion phenomena, such as an advancing front,22 and on explanations based on stress relaxation phenomena at the head of this adv ancing front. 23,24 Whereas these may ha ve aspects of v alidity, they can only be convincing when the v erifiable sur ace resistances and the v erifiable xponential concentration dependence of the dif fusion coefficient are also ta en into account. It w ould appear that enhanced stress relaxation, like the increase of dif fusion coefficient with sol ent concentration, is dependent on the increase in free v olume brought locally by the solv ent molecules themselv es. The measurement of surf ace resistances and concentration-dependent dif fusion coefficients did not requir use of an y mathematical tools or e xplanations other than the dif fusion equation solv ed with appropriate initial and boundary conditions. There were adv ancing fronts in volved both in the mathematics and in the samples in the e xperiments. The simple approach of solving the dif fusion equation with a v erifiable ( xponential) concentration-dependent dif fusion coefficient, along wit appropriate and verifiable parameters for the boundary conditions, xplains and replicates both the absorption and the desorption of solvents in polymers over an extremely large concentration range. This is true with and without significant sur ace resistance. The methodology and concepts presented in this chapter on dif fusion in polymers combined with the methodology and concepts in the rest of this handbook should pro vide insight into man y situations of industrial and theoretical interest. These include controlled release of drugs, absorption and transport through polymeric packaging of v arious kinds, impro ved prediction of the beha vior of chemical protecti ve clothing, absorption into coatings, release of absorbed chemicals from plastics, release of sterilization g as, etc.

CONCLUSION Many organic liquids are aggressive with respect to many types of polymers. Predicting aggressive behavior is important, and the absorption of the or ganic liquids into the polymers is a k ey factor in this respect. This absorption can be strongly af fected by both surf ace phenomena and the rate of diffusion within the bulk of the polymer. This chapter has emphasized the importance of surface resistance in this process. Additional material on dif fusion in polymers including simple mathematical descriptions of film drying by sol ent evaporation and so-called anomalous dif fusion have been included to present several aspects of the importance of surf ace resistance to contribute to its better understanding. There is no surf ace resistance to absorption where the absorbing molecules are small enough and/or linear. When the challenge molecules are too lar ge or b ulky, no absorption occurs, e ven though Hansen solubility parameter considerations lead one to predict that this is clearly e xpected. The molecules can simply not get through the surface layer, and its resistance is therefore effectively infinitely la ge. Between these e xtremes are situations where surf ace resistance necessarily af fects the absorption process. Surf ace resistance becomes significant when molecules can be transporte away from the surf ace into the bulk of the polymer f aster than they can be adsorbed/absorbed just at/in the surf ace. The simple approach of solving the dif fusion equation with appropriate and v erifiable param eters for the boundary conditions e xplains and replicates both the absorption and desorption of solvents in polymers o ver an e xtremely large concentration range. It is not yet possible to predict

7248_C016.fm Page 309 Wednesday, May 23, 2007 11:36 AM

Absorption and Diffusion in Polymers

309

which systems will gi ve significant sur ace resistance to absorption. Experimental studies are still required, but some similarity in HSP must be present for absorption to occur in an y event.

REFERENCES 1. Crank, J., The Mathematics of Diffusion, Oxford University Press, Oxford, 1956. 2. Hansen, C.M. and Hansen, K.M., Solubility parameter prediction of the barrier properties of chemical protective clothing, Performance of Protective Clothing: Second Symposium, ASTM STP 989, Mansdorf, S.Z., Sager, R., and Nielsen,A.P., Eds., American Society for Testing and Materials, Philadelphia, PA, 1988, pp. 197–208. 3. Hansen, C.M., The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient, Their Importance in Surface Coating Formulation, Doctoral dissertation, Danish Technical Press, Copenhagen, 1967. 4. Hansen, C.M., The measurement of concentration-dependent diffusion coefficients — the xponential case, Ind. Eng. Chem. Fundam., 6(4), 609–614, 1967. 5. Hansen, C.M., Dif fusion in polymers, Polym. Eng. Sci., 20(4), 252–258, 1980. 6. Hansen, C.M., Aspects of solubility , surf aces, and dif fusion in polymers, Prog. Org. Coat., 51(1), 55–66, 2004. 7. Korsmeyer, R.W., Von Meerwall, E., and Peppas, N.A., Solute and penetrant dif fusion in swellable polymers. II. Verification of theoretical models, J. Polym. Sci.: Polym. Phys. Ed., 24, 409–434, 1986. 8. Nielsen, T.B. and Hansen, C.M., Significance of sur ace resistance in absorption by polymers, Ind. Eng. Chem. Res., 44(11), 3959–3965, 2005. 9. Huldén, M. and Hansen, C.M., Water permeation in coatings, Prog. Org. Coat., 13(3/4), 171–194, 1985. 10. Nilsson, E. and Hansen, C.M., Evaporation and vapor diffusion resistance in permeation measurements by the cup method, J. Coat. Technol., 53(680), 61–64, 1981. 11. Hansen, C.M., Potential errors in water/water vapor permeation measurements using the cup method, Färg och Lack, 39(3), 57–60, 1993. 12. Abbott, R.J., Chudek, J.A., Hunter , G., and Squires, L., Skin layer ef fects on the dif fusion of carbon tetrachloride into injection moulded polyprop ylene studied by 1H NMR microimaging, J. Mater. Sci. Lett., 15, 1108–1110, 1996. 13. McDonald, P.J., Godward, J., Sackin, R., and Sear, R.P., Surface flux limited di fusion of solvent into polymer, Macromolecules, 34, 1048–1057, 2001. 14. Shankar, V., Influence of inter acial resistance on kinetics of sorption, Polymer, 22, 748–752, 1981. 15. Grayson, M.A., P ao, P.S., and Wolf, C.J., Transport of meth ylene chloride in poly(aryl-ether -etherketone) (PEEK), J. Polym. Sci.: Part B: Polym. Phys., 25, 935–945, 1987. 16. Kim, N.K. and Kammerme yer, K., Actual concentration profiles in membrane permeation, Sep. Sci., 5(6), 679–697, 1970. 17. Hwang, S.T. and Kammerme yer, K., Ef fect of thickness on permeability , in Permeability of Plastic Films and Coatings, Hopfenberg, H.B. Ed., Plenum, Ne w York, 1974, pp. 197–205. 18. Skaarup, K., Abstract, Lecture at Nordic Polymer Days, 1988; Grænseflademodstand ved Diffusionsprocesser I Polymerer, Danmarks Ingeniør Akadami, K emiafdelingen (in Danish) 1981, Surf ace Resistance in Dif fusion Processes in Polymers (in English). 19. Marom, G., The role of w ater transport in composite materials, in Polymer Permeability, Comyn, J., Ed., Elsevier, London, 1985, chap. 9. 20. Hansen, C.M., A mathematical description of film drying by sol ent evaporation, J. Oil Colour Chem. Assn., 51(1), 27–43, 1968. 21. Hansen, C.M., Dif fusion coef ficient measurements by sol ent absorption in concentrated polymer solutions, J. Appl. Polym. Sci., 26, 3311–3315, 1981. 22. Windle, A.H., Case II sorption, in Polymer Permeability, Comyn, J., Ed., Else vier, London, 1985, chap. 3. 23. Petropoulos, J.H., Interpretation of anomalous sorption kinetics in polymer-penetrant systems in terms of a time-dependent solubility coef ficient, J. Poly. Sci. Polym. Phys. Ed., 22, 1885–1900, 1984.

7248_C016.fm Page 310 Wednesday, May 23, 2007 11:36 AM

310

Hansen Solubility Parameters: A User’s Handbook

24. Sanopoulou, M., and Petropolous, J.H., Systematic analysis and model interpretation of micromolecular non-fickian sorption kinetics in polymer films Macromolecules, 34, 1400–1410, 2001.

7248_C017.fm Page 311 Wednesday, May 23, 2007 11:40 AM

— Safety and 17 Applications Environment Charles M. Hansen ABSTRACT Hansen solubility parameters (HSP) can be used to gain insight into many safety and environmental issues. These include substitution to more desirable materials, products, and processes, where a listing of solv ents having HSP similarity to the one(s) to be substituted pro vides an o verview of the potential choices for impro vement. Selection of suitable chemical protecti ve clothing can be improved by considering HSP correlations of breakthrough time. Ev aluating risks for inadv ertent chemical uptak e in plastic can be helped by HSP correlations of chemical resistance and/or permeation phenomena. Similarity of HSP suggests which chemicals are most lik ely to be rapidly absorbed into given plastic types. These same approaches can be used to e valuate the potential for uptake of chemicals through human skin.

INTRODUCTION Many organic materials are potential safety hazards. They can also be harmful to the en vironment. Unfortunately, it is often a matter of e xperience before the risks are unco vered because of damage being done. Thus, over the years, there ha ve been a series of substitutions with or without the aid of HSP to eliminate or to at least reduce such problems. An example is the lack of emphasis on the use of some eth ylene glycol ethers as solv ents because of their teratogenic ef fects, whereas the y were used in massi ve quantities earlier . The problem of replacing ozone-depleting chemicals is a case involving the e xternal environment. This is discussed a great deal in Chapter 11. Other lar ge-scale substitutions can also be cited where HSP can aid, b ut a list of this type is not the purpose of this chapter. The emphasis here is on the use of sound formulating principles to reduce the potential hazard in terms of reformulation or substitution. When a satisfactory substitution cannot be found, personal protection of one type or another may be required. Here, ag ain, HSP can help. Evaluating other forms of en vironmental risks can be aided by using HSP . An e xample is the occasional misuse of plastic containers normally used for soft drinks to store chemicals such as herbicides and pesticides. These are lik ely to dif fuse into the plastic container w all itself, making customary washing insufficient. HSP can indicate which chemicals can do this, thus pr viding information on the means to impro ve handling of the problem. This type of information can be generated for any polymer where HSP correlations of chemical resistance, weight g ain, etc., can be generated. All of these situations are discussed in more detail in the follo wing.

SUBSTITUTION Substitution involves the replacement of a potentially dangerous process or chemical with a ne w process or chemical ha ving less hazardous properties. The hazards can be judged using accepted approaches — for e xample, labeling requirements, toxicology assessments, biode gradability, and physical properties for the chemical or products. The v olatility of products is also a significan 311

7248_C017.fm Page 312 Wednesday, May 23, 2007 11:40 AM

312

Hansen Solubility Parameters: A User’s Handbook

factor with lo wer volatility being preferred due to reduced w orkplace concentrations and reduced replacement requirements for cleaners and the like which often recirculate in nearly closed systems. On the other hand, the problem should not just be transferred from the air e xhaust system to the sewer. The use of technologies in volving w ater or mechanical methods, such as mechanical joints rather than the use of solv ents, are preferred. Examples of preferred coatings technologies are the use of po wders which fl w at higher temperatures or polymerization by radiation, both of which use solvent-free base products to pro vide the coating. Other product types which may be tar geted include cutting fluids, cleaners of arious types, adhesives, sealers, and fillers In general, one primarily wishes to substitute for • • • • •

Carcinogens or suspected carcinogens Substances with risk phrases for being very toxic, toxic, allergenic, carcinogenic, teratogenic, mutagenic, or causing cumulati ve or irreversible effects Substances with moderate or serious aquatic toxicity Nonbiodegradable substances Substances with high predicted aquatic ef fects — for e xample, chemicals which prefer entially distribute to a nonaqueous phase to a v ery high de gree

Efforts should be made to de velop products with the lo west possible hazard. Those who ha ve read the earlier chapters in this book will immediately recognize HSP as a tool to aid in the substitution and systematic formulation for reduced safety and en vironmental risks. A key element in this is a listing of solv ents where those most resembling the candidate for substitution are at the top of the list. The program described in Chapter 1 can do this by entering the HSP for the solv ent to be replaced and requesting a listing with those solv ents most similar to it (i.e., the lo west RED numbers as defined in Chapter 1, Equations 1.10) at the top of the list. One must then sort throug these potential replacement candidates using other information to arri ve at a better alternati ve. It is clear that much more data than HSP are required to mak e the desired substitutions. However, a further discussion of this is be yond the scope of this chapter , which emphasizes HSP only . The currently used HSP techniques and correlations can aid in some aspects of substitution, and it is anticipated that future correlations will help in this endeavor. Many cases of substitutions in practice have been listed by Goldschmidt, 1 Olsen, 2 Soerensen and Petersen, 3 and Filskov et al. 4 A long list of references for the Danish experience with occupational risks and solutions is given by Soerensen and Petersen. 3

ALTERNATIVE SYSTEMS Alternative systems with less solv ent or no solv ent ha ve been focused on by the coatings and printing ink industries for man y years. Examples of such systems are coatings with higher solids, radiation-curable inks and coatings, po wder coatings, electrodeposition coatings, and other w aterreducible products. It might appear that solvent technology and use of HSP will not be as important as it has been in the past. This is not the case, ho wever, as demonstrated in earlier chapters. F or example, HSP principles can be used to aid in impro ved stability and adhesion, to predict polymer/filler interactions, to impr ve barrier polymers, and to aid in understanding some biological phenomena. The use of solv ents in alternati ve coatings systems has been the topic of se veral pre vious publications by the author .5–8 Some general principles of solv ent selection have been discussed in Chapter 8 and earlier ,9 as well as else where more recently.10

7248_C017.fm Page 313 Wednesday, May 23, 2007 11:40 AM

Applications — Safety and Environment

313

SOLVENT FORMULATION AND PERSONAL PROTECTION FOR LEAST RISK Solubility parameter principles have been used in formulating alternative, low VOC (volatile organic compound) products. A number of the general formulation principles can be briefly stated for th sake of completeness. These include the follo wing: 1. Solvents with lower viscosity most often lead to polymer solutions with lo wer viscosity. Such a change allo ws the use of higher solids at the original viscosity . However, these may evaporate more rapidly and can be e xpected to ha ve a lower flash point 2. Solvents with linear and smaller structures diffuse more rapidly than those with branched and larger structures. Inclusion of slo wer evaporating, more linear solv ents can hasten the through-drying of a coating. 3. Two (or more) mix ed solvents with lower labeling requirements may be able to replace a single solv ent. HSP can be used in this type of endea vor. 4. The surf ace tension of w ater-reducible coatings can often be significantly reduced b relatively small additions of ethanol or other alcohol-type solv ent. These can, of course, also be used in conjunction with other surf ace active materials. Materials with least potential risk are to be used in the Nordic countries where ver possible. The risk must be indicated by the seller/producer in terms of a labeling code. The risk can then be assessed by users or , perhaps more specificall , by primarily professional users. Such labeling is required on paints, printing inks, cleaners, or for an y product containing significant amounts o solvent or hazardous chemical. The labeling code dictates the personal protection required for the product, depending on the way it is used. Spraying product in a smaller room with limited entilation v requires much more protection than applying paint with a brush outdoors. Tables ha ve been published which gi ve the protection required (glo ves, dust mask, fresh air mask, body suit, etc.) for a gi ven set of application conditions for a wide v ariety of products from paints and printing inks through cleaners. 11 A key element in these tables is the labeling code de veloped in Denmark according to the MAL (in Danish: Maleteknisk e Arbejshygieniske Luftbehov) system. For present purposes, this is translated as the F AN (fresh air number). Higher MAL/F AN dictate that more extensive personal protection is required.

THE DANISH MAL SYSTEM — THE FAN12 As indicated pre viously, the quality of the w orking environment must be considered in all cases where organic solvents are being used. The Danish MAL system or other labeling system can be systematically used for this purpose. The Danish MAL reflects the cubic meters of fresh air require for ventilation of 1 l of product to belo w the threshold limit value (TLV). This number is modifie by a constant, depending on the e vaporation rate (or v apor pressure). Higher e vaporation rates imply greater hazard, so the multiplier is lar ger. The concept behind the MAL system can be better understood in English by translating the MAL number as the FAN. Other numbers in addition to the TLV/GV/OEL (occupational exposure limit) and FAN have been generated to help e valuate risks by inclusion of e vaporation rate/vapor pressure considerations. The vapor pressure di vided by the TLV is called the vapor hazard ratio (VHR) and the actual calculated vapor composition (using activity coefficients) d vided by the TLV has been called SUBFAC. A Danish publication comparing se veral of these is a vailable.4 To demonstrate the principle, the simple MAL = F AN has been tab ulated in Table 17.1 for se veral solvents.

7248_C017.fm Page 314 Wednesday, May 23, 2007 11:40 AM

314

Hansen Solubility Parameters: A User’s Handbook

TABLE 17.1 Fresh Air Numbers (FAN/MAL) for Selected Solvents from the Danish MAL Labeling Systema FAN/MAL

Solvent

1400 1100 880 110 88 78 74 58 54 48 46 29 28 26 25 24 23

Chloroform Tetrachloromethane Benzene Dichloromethane Trichloroethylene n-Hexane Toluene C9 Aromatics Methanol Methyl ethyl ketone Xylene 2-Propanol Propylene glycol monomethyl ether 1,1,1-Trichloroethane C>9 Aromatics Tetrahydrofurane Acetone

FAN/MAL 20 19 17 15 14 14 13 13 13 12 7 6 5 4b 0 0

Solvent n-Propanol Propylene glycol monomethyl ether acetate Propyl acetate Propylene glycol monoprop yl ether Mineral spirits/white spirit Butyl acetates Ethyl acetate Cyclohexane Benzin/petroleum ether (as heptane) Heptane Ethanol Propylene glycol monob utyl ether Dipropylene glycol monomethyl ether DBE (dibasic esters) Ethylene glycol Propylene glycol

These numbers are de veloped primarily with re gard to health hazards from v apors. The second number in the FAN code is added for hazard for skin contact, e ye contact, respiratory system contact, and/or ingestion. In addition to these, the European Union requires use of Xi, Xn, C, T, etc. Se veral of the solv ents require such labeling as well. One must also consider R (risk) and S (protecti ve measure) labeling requirements. b Estimated from composition of the mix ed solvent.15 a

Each product containing solv ent is assigned a tw o-digit number to place it into a potential hazard category. This number is a summation of the hazards possible for the components which are considered potentially hazardous. The first number relates to the potential hazard from th vapors and will v ary from 00 through 0, 1, 2, 3, 4, 5, to 6 as the potential hazard increases. The second number varies similarly and relates to the potential hazard from direct contact with the skin, eyes, breathing system, and by ingestion. This second number will not be less than 1 if or ganic solvents are included in the product in significant amounts.The following is a list of several solvents which are considered less desirable, based on this second number being 3 (or higher for higher concentrations in some cases): Toluene and Xylene at >10%, all common eth ylene glycol based ethers and their acetates (including dieth ylene glycol monob utyl ether , for e xample), terpenes, monomers at rather low concentrations, amines at moderate concentrations, and the most common chlorinated solvents. A “3” in this cate gory places the protection required in a significantly highe category with requirements for glo ves as a minimum and frequently fresh air masks as well. As indicated, a tw o-digit MAL code defines which safety precautions are required for each o a lar ge number of processing operations and conditions, including interior and e xterior painting and gluing, whether or not lar ge surfaces are involved, the quality of v entilation provided, surface preparation, painting of ships, larger construction sites, each of the printing processes, and industrial coating (spray box es, cabinets, etc.). 11 The protective measure required may be a f ace guard, e ye protection, a dust mask, a g as filter mask, a combination filter mask, a fresh air supplied mask, a body suit, in order of increasing requirements.

7248_C017.fm Page 315 Wednesday, May 23, 2007 11:40 AM

Applications — Safety and Environment

315

Examples of complete labeling of products and solv ents are beyond the scope of this chapter . The purpose of the discussion is to suggest that possible substituting solv ents can be listed, such as in Table 17.1, in an attempt to find a substitute with a l wer labeling requirement. Systematic consideration of labeling requirements is becoming a significant parameter i commercial applications of solv ents and products containing solv ents. This is happening all o ver the world, both with regard to worker safety as well as to the external environment. Such a procedure has been used to arri ve at optimum commercially useful solv ent compositions with the lo west possible risk for w orkers in the serigraphic printing industry as described in Danish patents DK 153797B (1989) and DK 160883 (1991) which correspond to European patents EP 0 205 505 B1 and EP 0 270 654 B 13,14 The preferred compositions reduce the MAL number to a minimum and also consider lo west possible internationally required labels as a requirement. The lo w label requirements of DBE (dibasic esters) ha ve been emphasized in comparison with other solv ents.15 Systematic solvent selection procedures ha ve also been strongly suggested for use in the selection of solvents for restoring older paintings. 16 This is discussed in Chapter 5.

SELECTION OF CHEMICAL PROTECTIVE CLOTHING HSP correlations for barrier properties of some types of chemical protecti ve clothing are gi ven in Chapter 13, Table 13.1. These correlations are based on data presented by F orsberg and Kieth. 17 Other e xamples of HSP correlations of barrier properties of protecti ve clothing are discussed in Chapter 13. Earlier publications also include HSP correlations of barrier properties of chemical protective clothing. 18–21 The procedure for using these correlations requires kno wledge of the HSP of the chemicals involved. These may be found in a suitable table or can be calculated according to the procedures outlined in Chapter 1. One then e valuates the RED number for the situation of interest. The RED number is discussed in Chapter 1 (Equation 1.10). If this number is less than 1, the system in not expected to be suitable for use. If the RED number is close to 1.0, there may be some doubt about the recommendation. RED numbers significantly greater than 1.0 can be considered for use. As discussed in Chapter 13, the molecular size of the chemical in volved is important in these e valuations. The major use of such correlations is to e valuate potential barrier types for chemicals where test results are not available. One can usually divide the results into groups of clearly not acceptable, questionable, and w orthy of further consideration. There ha ve been recent attempts to impro ve on the direct correlation of breakthrough times and permeation rates with HSP by trying to estimate the solubility and dif fusion coef ficient separately using HSP.22–25 These efforts have been discussed in Chapter 2 and Chapter 13.

UPTAKE OF CONTENTS BY A PLASTIC CONTAINER Plastic containers ha ve become increasingly popular in recent years. They have many advantages (which will not be discussed here), but there is also one disadvantage that HSP can shed more light on. This is the f act that plastic materials are able to absorb v arious liquids to some e xtent. The extent of absorption clearly depends on the HSP of the plastic used in the container compared with the HSP of the liquid which is in contact with it. Containers in contact with food ha ve been tested well for suitability for this purpose, including barrier properties relati ve to the contents. This is not the point of the present discussion. A problem e xists with the inadv ertent storage of hazardous liquids in the plastic container prior to its e xpected recycling as a container for a food or be verage. Many types of liquids can be temporarily stored in such containers. Whereas the earlier glass or metal containers could not absorb potentially dangerous materials, a plastic container can do this. A simple w ashing operation cannot be e xpected to remo ve all of the absorbed material. Washing

7248_C017.fm Page 316 Wednesday, May 23, 2007 11:40 AM

316

Hansen Solubility Parameters: A User’s Handbook

only removes what is on the surface or what can diffuse to the surface during the washing process, which presumably tak es place at some higher temperature. HSP concepts can focus attention on the types of chemicals that can absorb into a gi ven type of plastic container. This is useful information in terms of what analyses should be performed prior to recycling. The principles discussed here can possibly contrib ute in other w ays to impro ve the recycling process based on the increased le vel of kno wledge. There may be other w ays to reduce the problem.

SKIN PENETRATION Human skin is a complicated system. Nevertheless, it has been possible to characterize some aspects of the beha vior of human skin by HSP . The HSP found in a correlation of permeation rates of liquids in contact with viable skin26 are similar to those found for the swelling of psoriasis scales.27,28 This has been discussed in Chapter 15 in more detail, b ut also relates to w orker safety. The HSP for these correlations are included in Chapter 13, Table 13.1 and Chapter 15, Table 15.1. A skin penetration w arning has been attached to man y liquids taken up in the lists of limiting values for workplaces which are published in dif ferent countries. It was found earlier based on the HSP correlation with the swelling of psoriasis scales that this practice could be misleading, as HSP predicted man y liquids without this w arning also swelled psoriasis scales (k eratin) and could therefore be e xpected to penetrate the skin. 27 The lack of a skin penetration w arning for these liquids is partly attributable to the fact that this warning is based on experience. The bad experience giving the warning includes a combination of all effects, most notably the combination of dose and toxicity, rather than the potential dose ef fect only which is indicated by similarity of HSP . Earlier discussions also led to the impression that those in volved in this area did not consider the swelling of psoriasis skin as having relevance to the permeation of li ving skin. The finding that comparabl δP and δH are found from correlating the permeation rates of solv ents through living skin is a ne w input into this discussion. It is recognized that the δD parameter is dif ferent, but reasons for this are not clear. An improved HSP correlation of the permeation rates of solv ents through living skin based on a lar ger number of solv ents than the 13 included in the w ork of Ursin et al. 26 is perhaps required to give improved predictions in marginal cases, i.e., those near the boundaries of the HSP sphere describing the situation. The size and shape of the penetrating liquid molecules must also be considered. Predictions of the barrier properties of viable human skin should recei ve more attention. In addition, there is some discussion of the use of HSP in this respect in Chapter 15.

TRANSPORT PHENOMENA Many chemicals ha ve been the subject for concern in the past for v arious environmental reasons. Among these is the presence in artic re gions of chemicals that do not readily break do wn. Some chlorinated materials, such as pentachlorophenol, ha ve the ability to penetrate skin and w ood, and to be transported by animals, birds, or aquatic species after the y have taken them up. The HSP of given chemicals can gi ve a clue as to whether or not the y can follo w the same pathw ays in the environment. An example is given in Table 17.2 where the HSP for tetrabromobisphenolA (TBBPA) are reported along with the similarity of these with other rele vant materials. TBBPA has a distance from Pentachlorophenol (PCP) of only 3.5 units. This is very close and means that where PCP is soluble, TBBPA will also be soluble. TBBPA has a distance from the center of the spherical HSP correlation for Lignin solubility of only 6.8 units. Dividing this by the radius of the sphere to find the RED number (relat ve energy difference) shows that it is well within the solubility re gion with a RED of 0.5. TBBPA is readily soluble in lignin (w ood).

7248_C017.fm Page 317 Wednesday, May 23, 2007 11:40 AM

Applications — Safety and Environment

317

TABLE 17.2 The Affinities of Tetrabromobisphenol A (TBBPA) for Selected Biological Materials Material

δD

δP

δH

Ro

Dist. to TBBPA

TBBPA Pentachlorophenol Lignin solubility correlation Rapid skin penetration correlation Swelling of psoriasis scales correlation Depot fat (37 °C) total solubility Blood serum

20.2 21.5 21.9 17.6 24.6 15.9 25.5

9.1 6.9 14.1 12.5 11.9 1.2 10.3

13.8 12.8 16.9 11.0 12.9 5.4 22.1

— — 13.7 5.0 19.0 12.0 17.8

0.0 3.5 6.8 (RED = 0.50) 6.5 (RED = 1.36) 9.3 (RED = 0.49) 14.3 (RED) = 1.20 13.1 (RED = 0.73)

Note: Units are MP a1/2. Source: Reprinted with permission from Hansen, C.M., Conference Proceedings, Pharmaceutical and Medical Packaging 2001, Sk ov, H.R., Ed., He xagon Holding, Copenhagen, 2001, pp. 20.1–20.10.

TBBPA has a distance from the center of the spherical correlation for rapid skin permeation of 6.5. This is just outside the region for rapid permeation, but means that permeation will certainly take place at a moderate rate. The size and shape of the molecule will dictate the rate of permeation, not the solubility relations. TBBPA will readily absorb into the outer layer of the skin (k eratin) here described by swelling of psoriasis scales. The rate of absorption is dictated by the size and shape of the molecule, and not by the solubility relations. These data confirm that TBBPA is readily soluble where pentachlorophenol is soluble. If the chemical stability is comparable or better , then it can be presumed to appear at the same places if it gets into the en vironment. The same is true of related brominated compounds in general. Both pentachlorophenol and TBBPA can readily penetrate w ood and w ood products. They will also be readily taken up at lower concentrations without delay by human skin as shown by the correlations of rapid skin penetration and the swelling of psoriasis scales. Similar analyses can be done with other chemicals. HSP are a vailable in Appendix Table A.1 for man y phthalate plasticizers, mono 2-ethylhexyl phthalate, bisphenol A, N-methyl-2-pyrrolidone, and v arious glycol ethers based on ethylene oxide. In principle an y compound of interest can be assigned HSP for use in predicting behavior in connection with a lar ge number of en vironmental subjects. The correlations of the solubility of depot f at at 37 °C, the solubility of a protein (human blood serum), and swelling of k eratin (psoriasis scales swelling) indicate that TBBPA prefers to reside in the nonf atty tissue, b ut that it will ha ve some solubility in the f atty tissues as well. The collection of chemicals in f atty tissue is another topic that could be e xplored. A simple rule of thumb is that lack of w ater solubility will encourage collection in the f atty tissues, but this could be given more precision with HSP. The major problem with this collection is that the central nervous system is almost completely f atty in nature, and an e xcess of foreign materials can shortcircuit, misdirect, or stop messages, leading to memory problems, etc.

CONCLUSION In conclusion, it can be noted that HSP pro vides a tool to aid in substitution and in systematic formulation of less hazardous products and processes. One can also use HSP to more rapidly arrive at an optimum choice of chemical protecti ve clothing. HSP pro vides other insights with re gard to

7248_C017.fm Page 318 Wednesday, May 23, 2007 11:40 AM

318

Hansen Solubility Parameters: A User’s Handbook

uptake of undesirable chemicals in the human skin, in packaging materials, and perhaps e ven in a wide variety of other materials such as those found in nature.

REFERENCES 1. Goldschmidt, G., An analytical approach for reducing workplace health hazards through substitution, Am. Ind. Hyg. Assoc. J., 54, 36–43, January 1993. 2. Olsen, E., Substitution: a method to fulfill the orking environment law for air quality (Substitution: En Metode til at Overholde Arbejdsmiljoelovens Krav til Luftkvaliteten, in Danish), Dan. Kemi, 67(5), 146–153, 1986. 3. Soerensen, F. and Petersen, H.J.S., Substitution of hazardous chemicals and the Danish e xperience, Occup. Hyg., 1, 261–278, 1995. 4. Filskov, P., Goldschmidt, G., Hansen, M.K., Höglund, L., Johansen, T., Pedersen, C.L., and Wibroe, L., Substitution in Practice — Experience from BST (Substitution i Praksis — Erf aringer fra BST), Arbejsmiljøfondet, in Danish), Copenhagen, 1989. 5. Hansen, C.M., Solv ents in w ater-borne coatings, Ind. Eng. Chem. Prod. Res. Dev., 16(3), 266–268, 1977. 6. Hansen, C.M., Organic solvents in high solids and w ater-reducible coatings, Prog. Org. Coat., 10(3), 331–352, 1982. 7. Holten-Andersen, J. and Hansen, C.M., Solv ent and w ater e vaporation from coatings, Prog. Org. Coat., 11(3), 219–240, 1983. 8. Saarnak, A. and Hansen, C.M., Ev aporation from high solids coatings (A vdunstningen Från LFFårger), Färg och Lack, in Swedish, 30(5), 100–105, 1984. 9. Hansen, C.M., Solv ents for coatings, Chem. Technol., 2(9), 547–553, 1972. 10. Wu, D.T., F ormulating solv ents to remo ve hazardous air pollutants, Polym. Paint Colour J., 185, 20–23, December 1995. 11. Anonymous, Directive on Work with Products Having Codes (Bekendtgørelse om arbejde med kodenummererede produkter, Arbejdstilsynets Bekendtgørelse nr.302 af 13. maj 1993, in Danish), Danish Directorate for Labor Inspection. 12. Anonymous, Directi ve on Determination of Codes ( Bekendtgørelse om fastsættelse af kodenumre, Arbejdstilsynets bekendtgørelse nr. 301 af 13 maj 1993, in Danish), Danish Directorate for Labor Inspection. 13. Madsen, C.H. and Hansen, C.M., EP 0 205 505 B1, 1988. Assigned to CPS K emi Aps (Now a part of the Autotype/MacDermid Concern). 14. Madsen, C.H. and Hansen, C.M., EP 0 270 654 B1, 1991. Assigned to CPS K emi Aps (Now a part of the Autotype/MacDermid Concern). 15. Altnau, G., Risik opotentiale v on Lösemitteln systematisch be werten, Farbe+Lack, 103(9), 34–37, 1997; Systematic Ev aluation of Risk Potentials of Solv ents, Eur. Coat. J., 6/98, 454–457, 1998. 16. Hansen, C.M., Conserv ation and Solubility P arameters, Nordic Conserv ation Congress Preprints, Copenhagen, 1994, pp. 1–13. 17. Forsberg, K. and Kieth, L.H., Chemical Protective Clothing Performance Index, 4th ed., Instant Reference Sources, Inc., Austin, TX, 1991. 18. Hansen, C.M., The systematic choice of material in personal protection equipment for organic solvents in safety and health aspects of or ganic solvents, Progress in Chemical and Biological Research 220, Riihimäki, V. and Ulfv arson, U., Eds., Alan R. Liss, Ne w York, 1986, pp. 297–302. 19. Hansen, C.M. and Hansen, K.M., Which glo ves should I put on? (Hvilk e Handsk er Skal Je g Tage På?, in Danish), Färg och Lack, 33(3), 45–49, 1987. 20. Hansen, C.M. and Hansen, K.M., Solubility parameter prediction of the barrier properties of chemical protective clothing, Performance of Protective Clothing: Second Symposium, ASTM STP 989, Mansdorf, S.Z., Sager, R., and Nielsen,A.P., Eds., American Society for Testing and Materials, Philadelphia, PA, 1988, pp. 197–208.

7248_C017.fm Page 319 Wednesday, May 23, 2007 11:40 AM

Applications — Safety and Environment

319

21. Hansen, C.M., Billing, C.B., and Bentz, A.P., Selection and use of molecular parameters to predict permeation through fluoropolyme -based protective clothing materials, The Performance of Protective Clothing; Fourth Volume, ASTM STP 1133, McBriarty, J.P. and Henry, N.W., Eds., American Society for Testing and Materials, Philadelphia, P A, 1992, pp. 894–907. 22. Zellers, E.T., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. I. Modeling the solubility of organic solvents in Viton® gloves, J. Appl. Polym. Sci., 50, 513–530, 1993. 23. Zellers, E.T . and Zhang, G.-Z., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. II. Modeling diffusion coefficients, breakthrough times, and steady-state perme ation rates of or ganic solvents in Viton® gloves, J. Appl. Polym. Sci., 50, 531–540, 1993. 24. Zellers, E.T., Anna, D.H., Sulewski, R., and Wei, X., Critical analysis of the graphical determination of Hansen’s solubility parameters for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2069–2080, 1996. 25. Zellers, E.T., Anna, D.H., Sule wski, R., and Wei, X., Impro ved methods for the determination of Hansen’s solubility parameters and the estimation of solv ent uptake for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2081–2096, 1996. 26. Ursin, C., Hansen, C.M., Van Dyk, J.W., Jensen, P.O., Christensen, I.J., and Ebbehoej, J., Permeability of commercial solv ents through living human skin, Am. Ind. Hyg. Assoc. J., 56, 651–660, 1995. 27. Hansen, C.M., The Absorption of Liquids into the Skin, Report No. T 3-82, Scandina vian Paint and Printing Ink Research Institute, Hoersholm, Denmark, 1982. 28. Hansen, C.M. and Andersen, B.H., The affinities of o ganic solvents in biological systems, Am. Ind. Hyg. Assoc. J., 49(6), 301–308, 1988.

7248_C017.fm Page 320 Wednesday, May 23, 2007 11:40 AM

7248_C018.fm Page 321 Wednesday, May 23, 2007 11:46 AM

18 The Future Charles M. Hansen

ABSTRACT Hansen solubility parameters (HSP) help to quantify the statements “lik e dissolves like” and “lik e seeks like.” These parameters have found use in many fields of research and practice, primarily becaus their unique predictive capabilities are based on sound theoretical principles. HSP ha ve extended the original Hildebrand single solubility parameter approach by quantitati vely taking into account the molecular permanent dipole–permanent dipole and molecular h ydrogen bonding (electron inter change) interactions. HSP and the Prigogine corresponding states theory of polymer solutions are mutually confirming with r gard to treatment of specific interactions, as sh wn in Chapter 2. This is important, as it confirms that the HSP correlations must continue to include a constant not too diferent from the currently used “4” (or 0.25). This is necessary to dif ferentiate between the atomic ( δD) and the molecular (specific) interactions δP and δH). Neglecting this differentiation will lead to misinter pretations. The geometric mean a verage for the interaction of unlik e molecules is inherently used in the Hildebrand approach and in the HSP approach as well. This same mean must be used in the Prigogine corresponding states theory if agreement is to be found with the HSP correlations presented in this book. As the agreement is general, the conclusion must be that the geometric mean can be used to average not only dispersion interactions b ut also those attrib utable to permanent dipoles and to hydrogen bonding. These findings h ve been supported more recently by the statistical thermodynamics approach of P anayiotou and co workers summarized in Chapter 3. This approach allo ws independent calculation of each of the three parameters. Based on the large number of current uses of HSP, one can easily suppose that there are man y more practical uses which remain to be discovered and developed. One need not necessarily extend its theoretical scope to accomplish this. The existing data can be used in a strictly empirical manner if so desired. Ho wever, a glimpse has been gi ven of a v ery general ener getic approach to systematically predict and control molecular interactions among man y materials of widely dif ferent composition. The general predictions possible for these ph ysical interactions ha ve been demonstrated for both b ulk phenomena (solubility , swelling, compatibility) and surf ace phenomena (adsorption, dewetting, spontaneous spreading). In the future, the theory should be e xplored and used with this general applicability in mind. Problems and situations clearly needing further attention are discussed in the follo wing.

INTRODUCTION There are many matters related to HSP which still need clarification and xpansion. Some limitations are clear, but others are not so clear . As this book is primarily directed to ward the practitioner, the following discussions will start with more practical topics. The first matter of concern is the vailability of data. This handbook attempts to help impro ve this situation by publishing HSP for a lar ger number of liquids, about 1200, primarily in the Appendix, Table A.1. This handbook also contains ne w HSP correlations not present in the firs edition. Many of these are gi ven as examples in the te xt, and others are included in the Appendix, Table A.2. Other sources are discussed belo w. The second matter of concern is ho w reliable the HSP data are and ho w accurately the correlations can predict the beha vior of untested systems. Qualitati ve indications of this for the 321

7248_C018.fm Page 322 Wednesday, May 23, 2007 11:46 AM

322

Hansen Solubility Parameters: A User’s Handbook

data generated by the author are gi ven in the rele vant tables. In general correlation coef ficient approach 1.0. This indicates perhaps only a fe w minor outliers in the correlations, as can be seen from those included in this handbook. There are very rarely major outliers, and these usually ha ve another explanation for their behavior, such as very large molecular size, very small molecular size, reactions, or the lik e. The experience reported in Chapter 4, Table 4.4A, for the reliability of the “original Hansen” approach does not correspond to this e xperience. Normally there are perhaps 5 or 6 boundary solvents that are not predicted correctly out of about 100 test solvents in a correlation of experimental data. This was the case for the correlations presented in Table 5.1, rather than the ratio of 99 correct answers with 23 incorrect answers indicated in Table 4.4A. The reason for this discrepancy is not kno wn, but if group contrib ution or other estimates are in volved, especially for polymers, then the number of “errors” will increase. A third point which is sometimes irritating is that the scope of the characterizations possible is limited to the cohesi ve energy spectrum of the test liquids. A situation is often met where only a few solvents having high solubility parameters dissolv e a polymer which has still higher HSP . Similarly, only a fe w solv ents may interact intimately with a surf ace which has v ery high HSP . These surfaces are clearly wet because of the lo wer surface tension of all of the liquids, b ut only a few with high HSP prolong suspension of finer particles, for xample. The energy characteristics of such surf aces are apparently higher than those of an y liquids which can be used to study them by these techniques. Very high cohesi ve energies lead to the formation of solids, so there are no pure liquids which can be used to test the v ery high ener gy materials. Ne w thinking and ne w techniques are required to accurately characterize such high energy materials. A full understanding of the beha vior of w ater, or ganometallic materials, and salt solutions might be helpful in these situations (see the following corresponding sections). The current practice is to extrapolate into the region of v ery high HSP using Chapter 1, Equation 1.9 which includes the constant “4. ” It is assumed that this constant is still v alid, e ven for these v ery high ener gy characterizations. The given good solv ents are often in the boundary re gion of the HSP spherical characterizations. The solubility of De xtran C (British Drug Houses) 1 is an e xample of this as sho wn in Table 18.1 and Table 18.2. See Chapter 5 and Chapter 7 for further discussion of this problem which is present for both polymers and particulate matter. In a sense, the problem is similar to measuring the surface tension of a surface which has such a high v alue that even water spontaneously spreads on it. One can only conclude that its surf ace tension is greater than that of w ater. In the present case, there is a model to e xtrapolate HSP to higher v alues than can be measured directly . Another concern related to reliable HSP v alues is based on the f act that most chemicals in the intermediate molecular weight range, such as that characteristic of plasticizers, are soluble in almost all of the test liquids, e xcept for , for e xample, glycerin, w ater, and he xane. It is impossible to establish the three HSP based on such data. One generally has to rely on group contrib ution methods or other calculations or comparisons, and there will be some uncertainty in volved with this. Once the necessary reliable HSP data are a vailable, decisions and ideas are needed on ho w the data should be used. It is here that the existing theory and future extensions of it are most important. In man y cases, engineering approximations leading to a systematic course of action ha ve been possible using data which is currently a vailable. One can often arri ve at a prediction for e xpected behavior using the “lik e seeks lik e” principle, e ven though accurate numbers and an appropriate detailed theory may be lacking. It is hoped that this book will aid in the generation of still more HSP data ha ving a uniformly high quality , such that the interactions among still more materials can be predicted. Logical applications for HSP will be found in self-assembling systems and in what is called nanotechnology , for e xample. One e xample is the self-stratifying paints discussed in Chapter 8. Another is the ultrastructure of cell w alls in w ood discussed in Chapter 15.

7248_C018.fm Page 323 Wednesday, May 23, 2007 11:46 AM

The Future

323

TABLE 18.1 Calculated Solubility Sphere for Dextran C Solubility The Solvents with Their Parameters Solvent

δD

δP

δH

SOLUB

RED

V

Acetone Acetophenone Aniline Benzaldehyde Benzene 1,3-Butanediol 1-Butanol Butyl acetate gamma-Butyrolactone Carbon disulfid Carbon tetrachloride Chlorobenzene Chloroform m-Cresol Cyclohexanol Cyclohexanone Diacetone alcohol o-Dichlorobenzene 2,2-Dichlorodiethyl ether Diethylene glycol Diethyl ether Dimethyl formamide Dimethyl sulfoxide 1,4-Dioxane Dipropylene glycol Ethanol Ethanolamine Ethyl acetate Ethylene dichloride Ethylene glycol Ethylene glycol monob utyl ether Ethylene glycol monoethyl ether Ethylene glycol monomethyl ether Formamide Glycerol Hexane Isophorone Methanol Methylene dichloride Methyl isobutyl carbinol Methyl isobutyl ketone Nitrobenzene Nitromethane 2-Nitropropane Propylene carbonate Propylene glycol

15.5 19.6 19.4 19.4 18.4 16.6 16.0 15.8 19.0 20.5 17.8 19.0 17.8 18.0 17.4 17.8 15.8 19.2 18.8 16.6 14.5 17.4 18.4 19.0 16.5 15.8 17.0 15.8 19.0 17.0 16.0 16.2 16.2 17.2 17.4 14.9 16.6 15.1 18.2 15.4 15.3 20.0 15.8 16.2 20.0 16.8

10.4 8.6 5.1 7.4 0.0 10.0 5.7 3.7 16.6 0.0 0.0 4.3 3.1 5.1 4.1 6.3 8.2 6.3 9.0 12.0 2.9 13.7 16.4 1.8 10.6 8.8 15.5 5.3 7.4 11.0 5.1 9.2 9.2 26.2 12.1 0.0 8.2 12.3 6.3 3.3 6.1 8.6 18.8 12.1 18.0 9.4

7.0 3.7 10.2 5.3 2.0 21.5 15.8 6.3 7.4 0.6 0.6 2.0 5.7 12.9 13.5 5.1 10.8 3.3 5.7 20.7 5.1 11.3 10.2 7.4 17.7 19.4 21.2 7.2 4.1 26.0 12.3 14.3 16.4 19.0 29.3 0.0 7.4 22.3 6.1 12.3 4.1 4.1 5.1 4.1 4.1 23.3

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1* 0 0 0 1 0 0 1* 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0

1.454 1.371 1.241 1.346 1.776 1.054 1.313 1.640 1.077 1.756 1.858 1.601 1.556 1.246 1.312 1.473 1.363 1.474 1.313 1.000 1.795 1.082 1.000 1.485 1.080 1.180 0.880 1.559 1.416 1.003 1.406 1.211 1.170 0.915 0.991 2.037 1.410 1.144 1.411 1.517 1.679 1.336 1.399 1.479 1.172 1.053

74.0 117.4 91.5 101.5 89.4 89.9 91.5 132.5 76.8 60.0 97.1 102.1 80.7 104.7 106.0 104.0 124.2 112.8 117.2 94.9 104.8 77.0 71.3 85.7 130.9 58.5 59.8 98.5 79.4 55.8 131.6 97.8 79.1 39.8 73.3 131.6 150.5 40.7 63.9 127.2 125.8 102.7 54.3 86.9 85.0 73.6

7248_C018.fm Page 324 Wednesday, May 23, 2007 11:46 AM

324

Hansen Solubility Parameters: A User’s Handbook

TABLE 18.1 (CONTINUED) Calculated Solubility Sphere for Dextran C Solubility Solvent Tetrahydrofuran Tetrahydronaphthalene Toluene Trichloroethylene

δD 16.8 19.6 18.0 18.0

δP

δH

5.7 2.0 1.4 3.1

8.0 2.9 2.0 5.3

SOLUB

RED

V

0 0 0 0

1.450 1.618 1.744 1.560

81.7 136.0 106.8 90.2

Note: δD = 24.3 δP = 19.9 δH = 22.5 Ro = 17.4 FIT = 0.999 NO = 50.

HANSEN SOLUBILITY PARAMETER DATA AND DATA QUALITY The author and others including most solv ent suppliers and some paint companies (at least) ha ve databases including HSP data for solv ents and HSP correlations for polymer solubility etc. Tables of HSP data for man y materials are also included in standard reference w orks.2–5 There is still a tendency to regard the contents of such databases as proprietary information for the benefit of th owner and/or his/her customers. Exxon, for e xample, has indicated a computer program based on HSP where data for over 500 solvents and plasticizers, 450 resins, and 500 pesticides are included.6,7 The use of these parameters is becoming so commonplace that, in man y studies, the δD, δP, and δH parameters often appear without an y specific reference to where th y came from or what the y actually represent. The solvent listing in the Appendix, Table A.1 includes the pre viously published set of some 240 solvents which ha ve appeared earlier in se veral sources. 2,4,5,8,9 Some of the v alues have been revised o ver the years. The materials gi ven in dark type ha ve had some de gree of e xperimental verification. All the others are based on calculations only . The methods described in Chapter 1 were used, although in man y cases data w as lacking to such an e xtent that group contrib utions were used. In some cases data for whole, smaller molecules whose HSP are kno wn can be used to derive group contributions for estimating the HSP of lar ger molecules wherein they appear as a part. There are many additions to the original set of data. The calculated values have been checked against performance data reported in the literature where this has been possible. An example is the solubility data reported for poly(vinylidene chloride) (PVDC).10 Appendix, Table A.1 also includes HSP for a number of lo w molecular weight solids. Lo w molecular weight solids with relati vely low melting points ha ve been treated as if the y were liquids for e xtrapolation of latent heats to 25°C. This seems to be satisf actory, and it is consistent with the treatment of high boiling liquids. See Chapter 1 for details of the calculations. The first edition of this handbook contained ver 800 HSP values for chemicals. This has been e xpanded to about 1200 v alues in the second edition. HSP correlations in addition to those given in connection with examples in the text are included in Appendix, Table A.2. Only data judged (reasonably) reliable are reported. There are limitations on the accuracy of the HSP data deri ved from Burrell’s solvent range studies reported in standard reference works,2,11,12 but many correlations based on these data are included for reference anyway. The solvent range chosen for the studies does not completely fill out the possibilities selection o different liquids w ould have allowed. The problem of estimating a sphere based on limited data which do not experimentally define the whole sphere becomes more acute. This problem is greatest for polymers with high HSP , as not only is there a lack of possible data, b ut much of the v olume of the HSP sphere is located where there are no liquids. The cohesion energies are so high that no liquids are possible and only solids are present. An example of a good HSP correlation from the solvent range studies is that of polyeth ylene sulfide. This polymer has relati vely low HSP, and the solvents in the test series pro vide nonsolvents at higher HSP than those of the polymer to locate the boundaries with suf ficient accura y. This is sho wn in Tables 18.3 and 18.4. A comparison of

7248_C018.fm Page 325 Wednesday, May 23, 2007 11:46 AM

The Future

325

TABLE 18.2 Calculated Solubility Sphere for Dextran C Solubility The Solvents with Their Parameters Solvent

δD

δP

δH

Succinic anhydride Triethanolamine Dimethyl sulfone Ethylene cyanohydrin 2-Pyrolidone Ethanolamine Formamide Diethanolamine 1,3-Butanediol Glycerol Dimethyl sulfoxide Diethylene glycol Ethylene glycol Propylene glycol 1,3-Butanediol Diethylenetriamine Triethyleneglycol gamma-Butyrolactone Dipropylene glycol Dimethyl formamide Allyl alcohol o-Methoxyphenol Hexamethylphosphoramide Ethylenediamine Methanol Furfuryl alcohol Trimethylphosphate Benzyl alcohol Ethylene carbonate Phenol Ethylene glycol monomethyl ether Propylene carbonate Ethanol 1,1,2,2-Tetrabromoethane Ethylene glycol monoethyl ether N,N-Dimethyl acetamide 3-Chloro-1-propanol Hexylene glycol Methyl-2-pyrrolidone Furfural Aniline m-Cresol 1-Propanol Benzoic acid Triethylphosphate Quinoline Diethylene glycol monoethyl ether Acetic anhydride

18.6 17.3 19.0 17.2 19.4 17.0 17.2 17.2 18.0 17.4 18.4 16.6 17.0 16.8 16.6 16.7 16.0 19.0 16.5 17.4 16.2 18.0 18.5 16.6 15.1 17.4 16.7 18.4 19.4 18.0 16.2 20.0 15.8 22.6 16.2 16.8 17.5 15.7 18.0 18.6 19.4 18.0 16.0 18.2 16.7 19.4 16.1 16.0

19.2 22.4 19.4 18.8 17.4 15.5 26.2 10.8 8.4 12.1 16.4 12.0 11.0 9.4 10.0 13.3 12.5 16.6 10.6 13.7 10.8 8.2 8.6 8.8 12.3 7.6 15.9 6.3 21.7 5.9 9.2 18.0 8.8 5.1 9.2 11.5 5.7 8.4 12.3 14.9 5.1 5.1 6.8 6.9 11.4 7.0 9.2 11.7

16.6 23.3 12.3 17.6 11.3 21.2 19.0 21.2 21.0 29.3 10.2 20.7 26.0 23.3 21.5 14.3 18.6 7.4 17.7 11.3 16.8 13.3 11.3 17.0 22.3 15.1 10.2 13.7 5.1 14.9 16.4 4.1 19.4 8.2 14.3 10.2 14.7 17.8 7.2 5.1 10.2 12.9 17.4 9.8 9.2 7.6 12.2 10.2

SOLUB

1 1

1 1 0 1* 0 0

0 0 0

0

0 0 0 0

0 0

RED 0.739 0.819 0.846 0.866 0.867 0.880 0.915 0.972 0.984 0.991 1.000 1.000 1.003 1.053 1.054 1.063 1.068 1.077 1.080 1.082 1.117 1.121 1.132 1.136 1.144 1.144 1.147 1.152 1.152 1.167 1.170 1.172 1.180 1.199 1.211 1.215 1.216 1.219 1.220 1.230 1.241 1.246 1.250 1.258 1.259 1.265 1.272 1.277

V 66.8 133.2 75.0 68.3 76.4 59.8 39.8 95.9 87.5 73.3 71.3 94.9 55.8 73.6 89.9 108.0 114.0 76.8 130.9 77.0 68.4 109.5 175.7 67.3 40.7 86.5 115.8 103.6 66.0 87.5 79.1 85.0 58.5 116.8 97.8 92.5 84.2 123.0 96.5 83.2 91.5 104.7 75.2 100.0 171.0 118.0 130.9 94.5

7248_C018.fm Page 326 Wednesday, May 23, 2007 11:46 AM

326

Hansen Solubility Parameters: A User’s Handbook

TABLE 18.2 (CONTINUED) Calculated Solubility Sphere for Dextran C Solubility Solvent Tricresyl phosphate Formic acid Tetramethylurea

δD

δP

δH

19.0 14.3 16.7

12.3 11.9 8.2

4.5 16.6 11.0

SOLUB

RED 1.277 1.284 1.285

V 316.0 37.8 120.4

Note: δD = 24.3 δP = 19.9 δH = 22.5 Ro = 17.4 FIT = 0.999 NO = 50.

the solvents included in Table 18.3 with those in Table 18.1 sho ws which ones are lacking in the high HSP range. An example of a poor correlation using solv ent range data is that of the solubility of polyvinyl alcohol. Only tw o of the solv ents, ethanol and 2-propanol, dissolv e it. This leads to a correlation with the following data: δD;δP;δH;Ro equal to 17.0;9.0;18.0;4.0 in MPa1/2 with a perfect fit for t o good solv ents out of 56 in the set of data. The use of these data is not recommended. Ro is clearly too small by comparison with Ro found in HSP correlations for solubility for other water-soluble polymers. One of the problems with some of the reported correlations in the Appendix, Table A.2 is that the data on which the y are based were not generated for this purpose. There are shortcomings in terms of lack of full co verage of the HSP space as well as in the total number of liquids for which there are data. Note that a standard set of test solv ents such as that used in Table 18.1 tak es full coverage into account. Ho wever, some of these liquids must be handled with care for reasons of toxicity. Data for chemical resistance, permeation, and other phenomena related to solubility which can be correlated with HSP are practically ne ver accumulated with an HSP correlation in mind. This does not pre vent use of such data as demonstrated else where in this book, b ut it does place some limitations on the reliability of the predictions obtainable from the correlations. A qualitative indication of the reliability of the correlations is gi ven for this reason. Reliable HSP data for man y polymers of practical importance are not a vailable at this time. It would seem advisable for ra w material suppliers to determine the HSP for their rele vant products in a reliable manner and to publish these data on their product data sheets or else where. Including them in a possible future edition of this book may also be a possibility . For the sak e of completeness, a couple of w arnings are appropriate before proceeding to the next section. As noted in Chapter 1, the three partial solubility parameters tabulated by Hoy13,14 are not compatible with those of the author . As discussed in the ne xt section, the group contrib ution procedure presented by van Krevelen and Hoftyzer 15 does not give satisfactory agreement with the procedures given in Chapter 1. Finally , water (or its mixtures) should not be included currently in any HSP correlations without a v ery careful analysis of the results. The small molecular v olume, exceptionally high δH parameter, and tendency to self-associate depending on the local environment all lead to the lik ely result that w ater will be an outlier for the correlation. This results in HSP values which are less reliable, and ha ve lo wer predicti ve ability than had w ater been ne glected. Mixtures of or ganic solvents with w ater are still more problematic when used as test liquids (see Figure 18.1 and the follo wing discussion). A goal of future w ork should be to be able to account for the beha vior of w ater in a reliable manner , such that it can be included in studies leading to HSP correlations. The HSP v alues for w ater found from the correlation for total w ater solubility reported in Chapter 1 (T able 1.3) appear promising for some applications where the HSP v alues for water as a single molecule are clearly not applicable.

7248_C018.fm Page 327 Wednesday, May 23, 2007 11:46 AM

The Future

327

TABLE 18.3 Calculated Solubility Sphere for Polyethylenesulfide The Solvents with Their Parameters Solvent

δD

δP

δH

SOLUB

RED

V

Acetic acid Acetone Acetonitrile Aniline Benzene 1-Butanol sec-Butyl acetate Butyraldehyde Carbon tetrachloride Chlorobenzene p-Chlorotoluene m-Cresol Cyclohexane Cyclopentanone 1,2-Dichloro ethylene (cis) o-Dichlorobenzene 2,2-Dichlorodiethyl ether Dichlorodifluoromethane (Freon 12 Dichloromonofluoromethan Diethyl amine Diethyl ether Diethylene glycol Di-isobutyl ketone N,N-Dimethyl acetamide Dimethyl formamide 1,4-Dioxane Ethanol Ethyl acetate 2-Ethyl hexanol Ethylene carbonate Ethylene glycol Ethylene glycol monob utyl ether Ethylene glycol monoethyl ether Furfural Furfuryl alcohol Glycerol Isoamyl acetate Isoamyl alcohol Isopropyl acetate Methanol Methyl acetate Methyl ethyl k etone Methyl n-amyl k etone Nitroethane Nitromethane Octane

14.5 15.5 15.3 19.4 18.4 16.0 15.0 14.7 17.8 19.0 19.1 18.0 16.8 17.9 17.0 19.2 18.8 12.3 15.8 14.9 14.5 16.6 16.0 16.8 17.4 19.0 15.8 15.8 15.9 19.4 17.0 16.0 16.2 18.6 17.4 17.4 15.3 15.8 14.9 15.1 15.5 16.0 16.2 16.0 15.8 15.5

8.0 10.4 18.0 5.1 0.0 5.7 3.7 5.3 0.0 4.3 6.2 5.1 0.0 11.9 8.0 6.3 9.0 2.0 3.1 2.3 2.9 12.0 3.7 11.5 13.7 1.8 8.8 5.3 3.3 21.7 11.0 5.1 9.2 14.9 7.6 12.1 3.1 5.2 4.5 12.3 7.2 9.0 5.7 15.5 18.8 0.0

13.5 7.0 6.1 10.2 2.0 15.8 7.6 7.0 0.6 2.0 2.6 12.9 0.2 5.2 3.2 3.3 5.7 0.0 5.7 6.1 5.1 20.7 4.1 10.2 11.3 7.4 19.4 7.2 11.8 5.1 26.0 12.3 14.3 5.1 15.1 29.3 7.0 13.3 8.2 22.3 7.6 5.1 4.1 4.5 5.1 0.0

0 0 0 0 1 0 0 0 0 1 0* 0 0 0 1* 1 0 0 0 0 0 0 0* 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

3.352 2.285 3.793 2.125 0.973 3.462 1.898 1.947 1.006 0.600 0.869 2.631 1.155 2.107 1.123 0.954 1.605 2.771 1.308 1.744 1.772 4.970 0.993 2.752 3.286 1.480 4.476 1.604 2.521 4.491 6.077 2.634 3.325 2.825 3.286 6.916 1.699 2.898 2.043 5.483 1.919 1.697 1.019 3.038 3.852 1.551

57.1 74.0 52.6 91.5 89.4 91.5 133.6 88.5 97.1 102.1 118.3 104.7 108.7 89.1 75.5 112.8 117.2 92.3 75.4 103.2 104.8 94.9 177.1 92.5 77.0 85.7 58.5 98.5 156.6 66.0 55.8 131.6 97.8 83.2 86.5 73.3 148.8 109.4 117.1 40.7 79.7 90.1 139.8 71.5 54.3 163.5

7248_C018.fm Page 328 Wednesday, May 23, 2007 11:46 AM

328

Hansen Solubility Parameters: A User’s Handbook

TABLE 18.3 (CONTINUED) Calculated Solubility Sphere for Polyethylenesulfide Solvent 1-Octanol Pentane 1-Pentanol 2-Propanol Propionitrile Propylene carbonate Styrene t-Butyl alcohol Tetrahydronaphthalene Xylene

δD

δP

δH

SOLUB

RED

V

17.0 14.5 15.9 15.8 15.3 20.0 18.6 15.2 19.6 17.6

3.3 0.0 4.5 6.1 14.3 18.0 1.0 5.1 2.0 1.0

11.9 0.0 13.9 16.4 5.5 4.1 4.1 14.7 2.9 3.1

0 0 0 0 0 0 1 0 1 1

2.401 1.933 3.005 3.642 2.948 3.655 0.913 3.317 0.996 0.724

157.7 116.2 108.6 76.8 70.9 85.0 115.6 95.8 136.0 123.3

Note: δD = 17.8; δP = 3.8; δH = 2.2; Ro = 4.1; FIT = 0.981; NO = 56.

GROUP CONTRIBUTION METHODS Suggested calculation procedures to arri ve at the HSP for solv ents are gi ven in Chapter 1. The group contribution methods need e xpansion with new groups. New group contributions should be checked for reliability of the predictions in some w ay, which is not al ways possible within the timeframe of most projects. The group contrib ution v alues consistently used by the author are reported in Chapter 1. Values added o ver the years are appended to the original table which w as attributable to Beerbo wer.4,17,18 Barton has also collected man y tables of group contrib utions for various purposes. 2 As stated pre viously, the group contrib utions tabulated by v an Kevelen15 have not been found reliable. The δD parameter, in particular , is not predicted well. The author chose not to use these at an early date, although man y other authors ha ve chosen to do so. The use of various predictive methods which arri ve at dif ferent results has al ways been a problem. K oenhen and Smolders 19 evaluated various equations for predicting HSP . Methods for reliable a priori calculation of the HSP for polymers are not a vailable. This is a serious shortcoming. The author has tried several times to calculate the HSP for individual polymers using the same group contributions suggested for the liquids, and almost e very time has ultimately resorted to experiment. Calculation of the radius of interaction is a particular problem in this respect. This is definitely an area requiring attention. Chapter 2 discusses some of the actors which must be taken into account when calculating the radius of interaction. If one consistently uses the same method of estimating HSP, it can be assumed that some of the inherent errors will not affect relative evaluations. Utracki and co workers20 estimated HSP for a number of polymers assuming their δD parameters were not different and group contributions for the δP and δH parameters. This is discussed in Chapter 5.

POLYMERS AS POINTS — SOLVENTS AS SPHERES One way to possibly improve predicting the behavior of polymers is to consider them as points (or more accurately, spheres, with v ery small radii of interaction that depend on molecular weight) rather than as spheres with lar ge radii, as is presently done. A given solvent is assigned a rather large radius of interaction. This radius is larger for smaller molar volume in this inverted approach. This idea was presented many years ago,8,21 but it has never been fully explored. The first indication were that there seemed to be no real benefit in terms of impr ved reliability of predictions for polymer solubility in organic solvents, which was of primary interest, so there was no need to start

7248_C018.fm Page 329 Wednesday, May 23, 2007 11:46 AM

The Future

329

TABLE 18.4 Calculated Solubility Sphere for Polyethylenesulfide Solvent

δD

δP

Chlorobenzene Xylene p-Chlorotoluene Styrene o-Dichlorobenzene Benzene Di-isobutyl ketone Tetrahydronaphthalene Carbon tetrachloride Methyl n-amyl k etone 1,2-Dichloro ethylene (cis) Cyclohexane Dichloromonofluoromethan 1,4-Dioxane Octane Ethyl acetate 2,2-Dichlorodiethyl ether Methyl ethyl k etone Isoamyl acetate Diethyl amine Diethyl ether sec-Butyl acetate Methyl acetate Pentane Butyraldehyde Isopropyl acetate Cyclopentanone Aniline Acetone 1-Octanol 2-Ethyl hexanol m-Cresol Ethylene glycol monob utyl ether N,N-Dimethyl acetamide Dichlorodifluoromethane (Freon 12 Furfural Isoamyl alcohol Propionitrile 1-Pentanol Nitroethane Dimethyl formamide Furfuryl alcohol t-Butyl alcohol Ethylene glycol monoethyl ether Acetic acid 1-Butanol 2-Propanol Propylene carbonate

19.0 17.6 19.1 18.6 19.2 18.4 16.0 19.6 17.8 16.2 17.0 16.8 15.8 19.0 15.5 15.8 18.8 16.0 15.3 14.9 14.5 15.0 15.5 14.5 14.7 14.9 17.9 19.4 15.5 17.0 15.9 18.0 16.0 16.8 12.3 18.6 15.8 15.3 15.9 16.0 17.4 17.4 15.2 16.2 14.5 16.0 15.8 20.0

4.3 1.0 6.2 1.0 6.3 0.0 3.7 2.0 0.0 5.7 8.0 0.0 3.1 1.8 0.0 5.3 9.0 9.0 3.1 2.3 2.9 3.7 7.2 0.0 5.3 4.5 11.9 5.1 10.4 3.3 3.3 5.1 5.1 11.5 2.0 14.9 5.2 14.3 4.5 15.5 13.7 7.6 5.1 9.2 8.0 5.7 6.1 18.0

δH 2.0 3.1 2.6 4.1 3.3 2.0 4.1 2.9 0.6 4.1 3.2 0.2 5.7 7.4 0.0 7.2 5.7 5.1 7.0 6.1 5.1 7.6 7.6 0.0 7.0 8.2 5.2 10.2 7.0 11.9 11.8 12.9 12.3 10.2 0.0 5.1 13.3 5.5 13.9 4.5 11.3 15.1 14.7 14.3 13.5 15.8 16.4 4.1

SOLUB

RED

1 1 0* 1 1 1 0* 1 0 0 1* 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0.600 0.724 0.869 0.913 0.954 0.973 0.993 0.996 1.006 1.019 1.123 1.155 1.308 1.480 1.551 1.604 1.605 1.697 1.699 1.744 1.772 1.898 1.919 1.933 1.947 2.043 2.107 2.125 2.285 2.401 2.521 2.631 2.634 2.752 2.771 2.825 2.898 2.948 3.005 3.038 3.286 3.286 3.317 3.325 3.352 3.462 3.642 3.655

V 102.1 123.3 118.3 115.6 112.8 89.4 177.1 136.0 97.1 139.8 75.5 108.7 75.4 85.7 163.5 98.5 117.2 90.1 148.8 103.2 104.8 133.6 79.7 116.2 88.5 117.1 89.1 91.5 74.0 157.7 156.6 104.7 131.6 92.5 92.3 83.2 109.4 70.9 108.6 71.5 77.0 86.5 95.8 97.8 57.1 91.5 76.8 85.0

7248_C018.fm Page 330 Wednesday, May 23, 2007 11:46 AM

330

Hansen Solubility Parameters: A User’s Handbook

TABLE 18.4 (CONTINUED) Calculated Solubility Sphere for Polyethylenesulfide Solvent Acetonitrile Nitromethane Ethanol Ethylene carbonate Diethylene glycol Methanol Ethylene glycol Glycerol

δD

δP

15.3 15.8 15.8 19.4 16.6 15.1 17.0 17.4

18.0 18.8 8.8 21.7 12.0 12.3 11.0 12.1

δH 6.1 5.1 19.4 5.1 20.7 22.3 26.0 29.3

SOLUB

RED

V

0 0 0 0 0 0 0 0

3.793 3.852 4.476 4.491 4.970 5.483 6.077 6.916

52.6 54.3 58.5 66.0 94.9 40.7 55.8 73.3

Note: δD = 17.8; δP = 3.8; δH = 2.2; Ro = 4.1; FIT = 0.981; NO = 56.

all over again with this in verted system. On the other hand, there may be adv antages in terms of more reliable prediction of polymer–polymer miscibility, for example. This was not explored. The requirement of polymer miscibility will be that the respecti ve points (v ery small spheres) for the polymers must be very close to each other; comparing distances between small spheres is relatively easy. This type of comparison is sometimes dif ficult to ma e in the present approach where the degree of overlapping of rather lar ge spheres is used to estimate polymer–polymer miscibility . No fi ed rules of thumb have been established to estimate how much overlap is required for miscibility. However, guidelines for impro ving polymer–polymer miscibility are easily found in the present approach. These include selection of an improved solvent, reduction of polymer molecular weight, and modification of a polymer s HSP in a desired direction based on the HSP group contrib utions of its repeating unit or comonomers, for e xample. Traditionally, solvents are considered as points. This is practical and almost necessary from an experimental point of vie w as most solv ents are so miscible as to not allo w an y e xperimental characterization in terms of a solubility sphere. An exception to this is the data for w ater reported in Table 1.3. The HSP reported here are the center points of HSP spheres where the good solv ents are either those that are completely miscible or those that are miscible to only 1% or more.

CHARACTERIZING SURFACES The characterization of surfaces with HSP, or perhaps more correctly cohesion parameters (ha ving exactly the same numerical v alues), is still in its inf ancy. This possibility was demonstrated many years ago, ho wever.22 As sho wn in Chapters 6 and 7, this type of approach can lead to a ne w understanding of surface phenomena, which in turn allows systematic study and design of surfaces for desired beha vior. Data on surf ace characterizations, in addition to that in Chapter 6 and Chapter 7, are not provided here. This is primarily because such data are lacking b ut also because surf ace cohesion parameters may not be reflected by nominal ulk composition. The same basic pigment or fille , for example, can have widely different surface cohesion parameters, depending on how it has been surface treated. Neither has the ef fect of adsorbed w ater been clarified. Li ewise, a surf ace characteristic for a polyvinyl chloride or a polyethylene cannot be expected to be valid for all polymers normally said to be of these compositions. There may also be additi ves which ha ve dif ferent compositions and which may ha ve migrated to the surf aces. It appears that the relati ve simplicity of the surf ace characterizations discussed in this book would lead to their wider use. One current problem is that blindly entering wetting or spontaneous spreading data into the usual computer routine for finding the HSP alues often leads to ne gative

7248_C018.fm Page 331 Wednesday, May 23, 2007 11:46 AM

The Future

331

40

δh/MPa½

30

20

10

0

0

10

δp/MPa½

20

FIGURE 18.1 HSP plot of characterization of Rhodamin FB (C.I. BasicViolet 10) showing potential problems with incorporation of w ater mixtures as test solv ents (see te xt for discussion). (From Riedel, G., Farbe und Lack, 82(4), 281–287, 1976. With permission.)

7248_C018.fm Page 332 Wednesday, May 23, 2007 11:46 AM

332

Hansen Solubility Parameters: A User’s Handbook

numbers for one or more of them. This was discussed in Chapter 6. Currently , the best approach is to compare plots or e ven to just compare tab ulated data for the test solv ents to determine where two surfaces differ in af finities. Guides for action can also be found by simple comparison of th HSP of those solv ents which sho w a dif ference in beha vior. A more systematic approach for the use of cohesion parameters to describe surf ace phenomena w ould be desirable.

MATERIALS AND PROCESSES SUGGESTED FOR FURTHER ATTENTION Examples of the use of HSP for man y types of materials and phenomena ha ve been presented in earlier chapters. Some special types of materials are singled out here as w orthy of still more attention in the near future. These include surf ace active agents, w ater, gases, organic and inor ganic salts, organometallic materials, and aromatic (fragrances) materials. The uptake of potentially dangerous chemicals in recyclable packaging also needs attention. An additional area of interest may be found in that many commonly used reaction solvents have similar HSP. These include dimethyl sulfoxide, dimethyl formamide, dimeth yl acetamide, and sulfolane, for e xample. It seems unlik ely that this is a coincidence. It could be that the solubility of an acti vated species ha ving high polarity ( δP) and moderate hydrogen bonding (δH) is determining the reaction rate(s). Still another area of major interest is the systematic formulation of filled systems using HS . This is also still in its inf ancy. Pigments and fillers need to be characterized. S veral of these applications are discussed in more detail below. Surface active materials have remained essentially untouched in terms of HSP , although Beerbower started on this man y years ago. 8,17,23

SURFACE ACTIVE AGENTS Surface active agents have not been systematically characterized by HSP yet, although Beerbo wer has developed some aspects of a theory for gi ven situations. 8,17,23 The statement “lik e seeks lik e” indicates that surface active agents should be extensively treated in terms of HSP. Each end of such molecules will require its o wn set of HSP , as demonstrated by the e xample of lithium stearate, discussed later in the Or ganometallic Compounds section (Figure 18.3). An example to help illustrate the type of prediction possible is to try to answer the question of where the h ydrophobic end of a gi ven surfactant will tend to preferentially reside. An aliphatic end group w ould have lower affinity for polystyrene, for xample, than an aromatic one. Octane will not dissolv e polystyrene, whereas toluene will. This is reflected by their cohesion ene gy parameters. This same reasoning applies to other polymers. A surfactant with a fluorinated end wil not dissolv e in man y polymers where a h ydrocarbon end will. The cohesion ener gy parameters characteristic of fluorocarbons are too l w. Although these e xamples are ob vious to those skilled in the science of surf aces, they point to the possibility of quantifying af finities of sur ace active materials in terms of the cohesion energy parameters of their respective end groups. Those familiar with cohesion energy parameters can already discern dif ferences that may impro ve the chances of success. The data in Chapter 11 confirm that di ferences in HSP are critical if soils are to be removed effectively. HSP for surfactants can be assigned by experiment or by the methods described in the ne xt paragraph. Each surfactant must be assigned three sets of HSP . The first is for the ydrophobic end, the second is for the hydrophilic end, and the third is for the molecule as a whole. Figure 18.3 confirm the need for the first t o characterizations, and the third one is required for predictions when the whole molecule is soluble in the system. Even in a completely dissolved condition, one anticipates some degree of orientation of the ends of the surf actant molecule to ward regions of similar HSP . The HSP for the h ydrophobic end can be estimated by the methods discussed in Chapter 1, with group contributions, or by simple comparison with similar (usually) h ydrocarbon molecules of the same size. Barton has collected group contrib utions that can also be used in connection with the

7248_C018.fm Page 333 Wednesday, May 23, 2007 11:46 AM

The Future

333

different ends of surf ace acti ve agents. 2 The HSP of the h ydrophilic end can be estimated by comparison with existing HSP of, for e xample, sucrose or other w ater soluble entity, organic salts (see below), inorganic salts (see belo w), polyeth ylene oxide or polyeth ylene glycol, or whate ver resembles this end best. Experiments are preferred, of course, and the estimation for given inorganic salts is still uncertain. When a surfactant denatures a protein, there will be some similarity between the HSP of the protein or urea, if this also denatures the protein (see Chapter 15), and the HSP of the hydrophilic end of the surf actant. If enough data of this kind can be generated, the HSP of the surfactant can be e xperimentally confirmed. The estimation for the molecule as a whole in volves combining the two sets of HSP for the ends. It is thought that this is best done by a veraging using estimated molecular volumes for the respective ends relative to the molecular volume of the whole molecule. In closing this section, it can also be repeated that thermodynamic surf ace and interf acial phenomena correlate with HSP . This has been amply sho wn in Chapter 6 and Chapter 7. The kinetics of the situation may also be important. Chapter 16 discusses adsorption and absorption in polymer surfaces where there is a surf ace resistance. There will also be some form of interf acial or boundary layer resistance influencing the adsorption and absorption of sur actants into soils, for example. If the HSP do not match sufficiently well, adsorption and absorption will presumably no occur as readily as when the HSP do match to within some required limit for the desired ef fect, as shown in Chapter 11. The size and shape of the adsorbing/absorbing entity is also presumed to be important from a kinetic point of vie w, as demonstrated by the e xamples in Chapter 16.

SURFACE MOBILITY (SELF-ASSEMBLY) The rule of thumb that “lik e seeks lik e” can be v ery useful in understanding the structure of complicated systems. That this type of consideration can lead to useful results can be seen in the way that the beha vior of w ood polymers and the ultrastructure of cell w alls in w ood was treated 24 Hemicelluloses appear to in Chapter 13 and in much more detail by Hansen and Björkman. function much like surfactants with the backbone and those side chains containing hydroxyl groups favoring placement to ward cellulose (or their o wn kind). Hemicellulose side chains containing acetyl, acid, or ether groups are expected to favor orientation toward lignin regions. In this example, it is interfacial mobility that is in focus, and it can be expected that the orientations may be changed with the transport or presence of other materials such as w ater through a given local environment. These predictions and inferences appear to agree with what is e xpected or has been established by independent measurement, b ut it is too early to say that confirmation has been obtained indepen dently. The treatment of dif ferent segments of block copolymers as separate entities is a related endeavor where more quantitative predictions of compatibility should be possible. It is kno wn that additions of a block of polymer C to both polymer A and polymer B impro ves their chances of compatibility (at some molecular level). The association of blocks of polymers is also the basis for the thermoplastic elastomers (TPE). These are made with a wide v ariety of dif ferent immiscible (hard and soft) blocks where the phase separation is critical to performance. Typical e xamples include the styrene/butadiene/styrene block copolymer (SBS), the polyether/polyamide block copolymer (PEBA), and polyurethanes combined with polyethers or polyesters (TPU). Some types are also vulcanized to impro ve properties. The rotation of some hydrophobic materials to become more hydrophilic when in contact with water is still another example of like seeking like. Peat moss is an example. A drop of water initially pearls on the surf ace but shortly thereafter disappears into the interior in a spontaneous manner . The peat moss has become hydrophilic (but returns to the hydrophobic state on drying again). This phenomena was actually employed to develop an electrodeposition coating for an evaporator where film-wetting by ater was required for good evaporation efficien y.25 After several hours of exposure of a fresh coating to w ater, the static contact angles with w ater disappeared and a coherent w ater film as obtained.

7248_C018.fm Page 334 Wednesday, May 23, 2007 11:46 AM

334

Hansen Solubility Parameters: A User’s Handbook

Many surface phenomena can be understood from the preferences of gi ven segments or materials to seek out regions of similar HSP. Some inferences may be possible from the studies performed on compatible (or nearly compatible) polymers. The HSP data leading to formulation of selfstratifying coatings also pro vides useful information 26 (see also Chapter 6). Systematic studies of these effects are badly needed. One such study 27 confirmed that the rotation ability (mobility) o aging polymer surfaces could be followed by measuring the (static) receding contact angle of water. Aging can be expected to lead to increased oxygenation and perhaps also to a decrease in a verage molecular weight. These ef fects both contrib ute to the tendenc y/ability for oxygenated species attached to an otherwise more h ydrophobic polymer to rotate into an applied w ater droplet. When the (static) receding contact angle for w ater w as measured, it fell with e xposure time/aging at shorter times, whereas the (static) adv ancing contact remained constant. At longer exposure times, when the surface was oxygenated to a greater extent, the advancing contact angle also started to fall. Surface mobility also has an important role in biological processes, as described in Chapter 15. The orientation of molecules to allo w given segments to locate in re gions of similar HSP is presumed to be a general phenomenon. Hydrophilic bonding (usually referred to in the present context as intermolecular h ydrogen bonding) is responsible for the configurations of proteins i water. The proteins that can be dissolved in mixtures of water and urea or given salts, for example, are no longer “h ydrogen bonded” in the con ventional usage of the term, as the y are no w truly dissolved by an ef fectively good solv ent that can also dissolv e these se gments/bonding sites. The usual solvent, water, does not ha ve the correct set of HSP to truly dissolv e these se gments of the protein molecules. The urea additions correct for this deficien y, and the protein is said to be denatured in the process. The concept of hydrophilic (hyperphilic?) bonding, which is the opposite of hydrophobic bonding, is discussed in more detail with examples in Chapter 15. These phenomena also point to the use of urea or urea groups to impro ve biocompatibility. Many of the concepts discussed here are directly applicable for self-assembling systems and to procedures and products within the concepts of nanotechnology .

WATER The current treatment of the HSP for w ater discussed in Chapter 1 and Chapter 15 needs confi mation and/or modification. As noted earlier on se veral occasions, w ater is v ery special because of its lo w molecular v olume, its v ery high δH parameter for a liquid, and its tendenc y to selfassociate or to associate with other materials forming special structures. The HSP correlations for the solubility of solvents in water presented in Chapter 1, Table 1.3 have not been tested extensively as yet, b ut do seem promising. They are clearly useful to mak e predictions for the solubility of untested solv ents in w ater, but whether or not these HSP data for w ater can be used in a lar ger context remains to be determined. General beha vior can be predicted, b ut can specific beh vior be predicted? More research is needed in this area, b ut, in the meantime, w ater can be considered as having (at least) duality. Sometimes it acts like a single molecule, and sometimes it acts as a cluster of about six molecules (according to the HSP comparison, at least). There may also be other possibilities. The use of the HSP for w ater found from the correlation of total w ater solubility appears to be the most promising set of v alues to work with at the present time. This is especially true for w ater in lower energy systems. It is not yet advisable to include water in a standard set of test liquids for xeperimental evaluation of the HSP for polymers or other materials because of its tendenc y to be an outlier . This means a challenge still exists to understand ho w to be able to incorporate w ater into a standard set of HSP test liquids without al ways being concerned about special interpretations for w ater, and only for water. An example of how this can lead to oddities is discussed in the follo wing. A characterization problem caused by nonideal mixtures with w ater is the interpretation of HSP correlations for materials such as the dye Rhodamin FB (C.I. Basic Violet 10). 2,16 Use of mixtures of solv ent and w ater as test solv ents led to a v ery nonspherical (noncircular) cohesion

7248_C018.fm Page 335 Wednesday, May 23, 2007 11:46 AM

The Future

335

TABLE 18.5 Calculated Solubility Sphere for Rhodamin FB Solvent

δD

δP

δH

Solubility

RED

V

Acetone Benzene 1-Butanol Butyl acetate gamma-Butyrolactone Cyclohexane Diacetone alcohol Diethylene glycol Diethylene glycol monomethyl ether Diethylenetriamine Dimethyl formamide Dimethyl sulfoxide Dipropylene glycol Ethanol Ethylene glycol Ethylene glycol monoethyl ether 2-Ethyl hexanol Glycerol Methanol Methylisoamyl ketone Methyl-2-pyrrolidone 1-Pentanol 1-Propanol Propylene glycol Tetrahydrofuran Toluene Trichloroethylene Water

15.5 18.4 16.0 15.8 19.0 16.8 15.8 16.6 16.2 16.7 17.4 18.4 16.5 15.8 17.0 16.2 15.9 17.4 15.1 16.0 18.0 15.9 16.0 16.8 16.8 18.0 18.0 15.5

10.4 0.0 5.7 3.7 16.6 0.0 8.2 12.0 7.8 13.3 13.7 16.4 10.6 8.8 11.0 9.2 3.3 12.1 12.3 5.7 12.3 4.5 6.8 9.4 5.7 1.4 3.1 16.0

7.0 2.0 15.8 6.3 7.4 0.2 10.8 20.7 12.6 14.3 11.3 10.2 17.7 19.4 26.0 16.4 11.8 29.3 22.3 4.1 7.2 13.9 17.4 23.3 8.0 2.0 5.3 42.3

0 0 0* 0 0* 0 1* 1 1 1 1 1 1 1 1 1 0 1 1 0 1* 0 0* 1 0 0 0 0

1.125 1.991 0.999 1.517 0.988 2.076 1.001 0.486 0.934 0.487 0.677 0.741 0.570 0.732 0.815 0.707 1.294 0.996 0.589 1.530 1.042 1.138 0.889 0.772 1.295 1.902 1.615 1.965

74.0 89.4 91.5 132.5 76.8 108.7 124.2 94.9 118.0 108.0 77.0 71.3 130.9 58.5 55.8 79.1 156.6 73.3 40.7 142.8 96.5 108.6 75.2 73.6 81.7 106.8 90.2 18.0

Note: δD = 16.7; δP = 17.5; δH = 18.5; Ro = 12.2; FIT = 0.930; NO = 28. Source: Riedel, G., Farbe and Lack, 82(4), 281–287, 1976; Birdi, K.-S., Ed., Handbook of Surface and Colloid Chemistry, CRC Press, Boca Raton, FL, 1997, chap. 10.

energy parameter plot (see Figure 18.1). The irregular plot can presumably still be used as such, but the characterization of the dye in question is not useful in relation to prediction of interactions with other materials. The plot has given several individuals the impression that there are significan problems with the HSP approach when it is applied to this kind of material. This is not true. A computer analysis based on the pure solvent data given by Riedel16 confirms that a good “spherical characterization of Rhodamin FB is possible using the same data otherwise used in Figure 18.1. 5 The HSP data for this correlation are gi ven in Table 18.5. The data fit as 0.93 for 28 data points. Figure 18.1 clearly shows that this HSP sphere co vers more space than the data, with a significan portion in the high ener gy re gion where there are no liquids. Chapter 1, Equation 1.9 (with the constant 4) w as used in this correlation, as it has been used in all the other HSP correlations in this book. The HSP correlations for water-soluble polymers and other high energy materials involved similar extrapolations into domains where there are no liquids. This procedure may be subject to revision at some future point in time, but for the present it seems to be the only procedure possible to maintain consistenc y in the HSP procedures de veloped. It should be remembered that man y

7248_C018.fm Page 336 Wednesday, May 23, 2007 11:46 AM

336

Hansen Solubility Parameters: A User’s Handbook

25

ENERGY OF VAPORIZATION FOR “NON POLAR” LIQUIDS BUTANE C4H10 (.642) ●

20

●

ISOBUTANE (.639)

TR =0.5

ΔET KJ ∕ MOLE

TR =0.6

15 Xe (.569) ✳

N

E BL

O

G

ES AS

● C3H8

(.624)

TR =0.7 ●C

2H6 (.6025)

10 Kr (.578) ✳ O2 (.583) Ar (.577) ✳

5

Ne (.607) ✳

10

●

CH4 (.586)

N2 (.613)

H2 (.613) He (.803)

●

0 0

●

20

✳ 30

40 50 60 MOLAR VOLUME CC/MOLE

70

80

90

100

FIGURE 18.2 Cohesion energy for various low molecular weight materials as a function of molecular volume and reduced temperature (gi ven by curv es or in parentheses). (See te xt for discussion.)

(most) liquids with high HSP (w ater, methanol, glycols) also ha ve lo w molecular v olumes (V). This makes them “better” solvents than expected by comparison with all the other solvents (whose average V is closer to 100 cc/mol). This f act might gi ve the impression that the constant 4 in Chapter 1, Equation 1.9 should be increased. This is discussed more in Chapter 2. A unifying concept and procedure for the use of w ater in all testing is needed. The HSP considerations discussed in this book pro vide help toward reaching this goal.

GASES HSP can also be used to impro ve understanding of the solubility beha vior of g ases. Solubility parameters are usually deri ved from data at the normal boiling points. HSP deri ved from these numbers seem to be in good qualitati ve agreement with e xpectations (even at 25°C), and in man y cases quantitative agreement with physical behavior has also been found. Some examples are given by Barton. 2 Solubility parameter correlations for oxygen 28 (Chapter 13) and nitrogen 29 (Chapter 13) have been used as e xamples in this book. The δP and δH parameters for these tw o gases are zero. HSP for man y gases where this is not true are reported in Chapter 13, Table 13.4. A specifi example where this is not the case is carbon dioxide. Carbon dioxide is e xtensively discussed in Chapter 10 where the HSP are calculated for lar ge v ariations in pressure and temperature. This same procedure is applicable to other g ases. Chapter 3 also treats a method to calculate the three partial solubility parameters for g ases. In the process of calculating the HSP for g ases, it was found necessary to e xtrapolate the data in Chapter 1, Figure 1.1 to lo wer molar v olumes. Figure 18.2 is deri ved from this. This figure i worthy of some consideration from a theoretical point of vie w. The basis of the HSP is a corresponding states calculation for E D as the ener gy of v aporization of a corresponding h ydrocarbon solvent (same V and structure) at the same reduced temperature.This is, of course, 298.15 K divided

7248_C018.fm Page 337 Wednesday, May 23, 2007 11:46 AM

The Future

337

by the solv ent’s critical temperature. The reduced temperature at the boiling point is indicated in parentheses in Figure 18.2. Questions can be raised as to wh y the noble g ases dif fer from the hydrocarbon solv ents and whether the h ydrocarbon solv ents were the best choice as reference materials. Also, why is oxygen among the noble g ases in cohesion behavior rather than rated with the other g ases? At the time of choice for a reference for the dispersion bonding ener gies, there were ample data on latent heats for the yhdrocarbons and the aliphatic hydrocarbons were considered as having δP and δH values equal to zero. This may not quite be true, b ut the corrections w ould be minor, and the necessary data for a re vised reference are lacking. It appears that the currently chartered course of using h ydrocarbon solv ents as a basis will be maintained. Some additional considerations may be found in further study of the relation of HSP to the corresponding states theory of Prigogine and co workers as discussed in Chapter 2. The beha vior of the h ydrocarbon solvents appears to be included within the Prigogine parameter for dif ferences in size ( ρ).

ORGANIC SALTS The HSP of se veral or ganic salts ha ve been compared with the HSP for the or ganic acids and organic bases from which the y were made. 30 The result w as that the or ganic salts al ways had considerably higher HSP than either of the components making them up. As examples, the salts made from formic acid and acetic acid combined indi vidually with dimeth yl ethanolamine had δD;δP;δH equal to 17.2;21.5;22.5 and 16.8;19.8;19.8, respecti vely, whereas the δD;δP;δH are 14.3;11.9;16.6 for formic acid, 14.5;8.0;13.5 for acetic acid, and 16.1;9.2;15.3 for dimeth yl ethanolamine. All of these values are in MPa1/2. This general relationship was also found for other salts formed by combinations of or ganic acids with a v ariety of amines. These values are reported in Appendix Table A.1 in dark type, as there are e xperimental data to justify the numbers. The HSP for the salts are generally close to those mentioned earlier . These values are high enough to mak e the salt entities insoluble in most polymers. Their affinities for ater will be v ery high, ho wever, both because of high HSP and also because of the char ges associated with the salt groups. There was about 10% shrinkage in v olume compared with the original v olumes of the acids and bases. In some cases, the cohesion ener gy of the salts is high enough to mak e them solids rather than liquids. This study sho wed that or ganic salts can indeed be characterized by HSP . More w ork is necessary, however, with other types of salts. In particular , the acid groups found in nature, such as in hemicelluloses, deserv e more attention (see Chapter 15 and Reference 24).

INORGANIC SALTS The solubility of magnesium nitrate [Mg(NO 3)2·6H2O] was evaluated in a standard set of solv ents1 and later correlated more precisely with HSP. The HSP derived from this are δD;δP;δH, and Ro equal to 19.5;22.1;21.9, and 13.2, respecti vely, all in MP a1/2. Nitrates are kno wn to be among the most soluble of salts. Somewhat less soluble than the nitrates are chlorides. These are only partly soluble in a few organic liquids with very high HSP. Group contributions to the HSP from the nitrate group are e xpected to result in lo wer HSP, and, in particular , lower δD for the nitrate portion of a salt than would be e xpected from the group contrib utions from a chloride. This would lead to greater solubility of the nitrates in or ganic solvents, which is indeed the case. The δD parameter seems to be qualitatively capable of describing the beha vior of metals to some e xtent. It may be possible to arrive at an approximate description of inor ganic salt solubility in or ganic media (perhaps w ater, too) using HSP or some modification/ xtension thereof. The salting in and salting out of v arious polymers can perhaps pro vide clues to assign HSP in this connection. Finally , it should be noted that an e xcellent HSP correlation of the chemical resistance of an inor ganic zinc silicate coating is reported in Chapter 12, Table 12.1.

7248_C018.fm Page 338 Wednesday, May 23, 2007 11:46 AM

338

Hansen Solubility Parameters: A User’s Handbook

11 cc ABSORBED ? > .8 δH δP

10 ?

9

2

8

1 3.1 1 3.7

?

2.7 8.7

7

δH

6

HANSEN PLOT - LITHIUM STEARATE

5 4 ??

3

?

?

?

?

?

1

2 ?

1 1

2

3

4

5 δP

6

7

8

9

FIGURE 18.3 HSP plot for solvent uptake by lithium stearate. 31,32 Units for δP and δH are (cal/cm 3)1/2. (From Alan Beerbower, personal communication.)

ORGANOMETALLIC COMPOUNDS No systematic studies of the HSP of or ganometallic compounds have been made. An exception is perhaps that shown in Figure 18.3 where Beerbower31 used data from Panzer32 to show that lithium stearate does indeed have two distinct regions of solvent uptake and that a HSP plot can show why. This example shows that one can calculate HSP values where the relevant data can be found in the literature and then test these with rele vant e xperiments. Group contrib utions w ould be v aluable. Metallic bonds differ in nature from those usually discussed in connection with organic compounds. A suspicion is that, at least in practice, the cohesion ener gy derived from the “metallic” bonding in organometallic compounds can be coupled with the dispersion parameter.There is also a question, for example, of whether metal atoms in the center of more complicated molecules are ef fectively shielded from an y (surface) contact with a solv ent. Surface contacts are clearly important, b ut it appears that the nature of the central atom also has an ef fect. Finally , it might be noted that Hildebrand and Scott presented a chapter on the solubility parameters of metals. 33 Unfortunately, we do not often deal with pure metals in this conte xt, but rather metal oxides, for which no HSP work has been reported, as least not to the author’ s knowledge.

AROMAS

AND

FRAGRANCES

Aromas and fragrances are important in connection with packaging materials, foodstuffs, cosmetics, chewing gum, etc. A recent report 34 discussed HSP in connection with fragrances and aromas. It is clear that HSP e xist for these materials, b ut very little work has been published in the area. One of the examples included in Reference 34 was the development of an artificial nose based on coate oscillating sensors, which oscillate more slo wly when the y g ain weight. Matching HSP for the coating and material to be detected leads to increased weight g ain and increased sensitivity. Other

7248_C018.fm Page 339 Wednesday, May 23, 2007 11:46 AM

The Future

339

examples where HSP could be systematically used include: counteracting undesirable odors using fragrances that ha ve reasonably similar HSP; absorption of odors into plastics, coatings, sealants, etc.; development of packaging with designed HSP to either function as a barrier or as a sink; and an estimate of where a gi ven aromatic material is lik ely to reside. The key to interpretation is, as usual, that similarity of HSP means higher affinit . It is thought that the masking effect is the result of adsorption of the fragrance on the regions of the sensory system where HSP are similar to those of the undesired odor . There may not be a complete match re garding steric adsorption, b ut the sensitive areas are covered anyway, and this barrier prevents the odor from arriving at the given sites.

ABSORPTION

OF

CHEMICALS

IN

PLASTICS

HSP correlations e xist for chemical resistance, permeation phenomena, and uptak e of solv ents in many polymers. The recycling of polymeric containers has a potential problem in that the polymers used are able to readily absorb those chemicals whose HSP are not too dif ferent from their o wn. Once a chemical has absorbed into a polymer , and particularly if it is a rigid polymer with a relatively high glass transition temperature, it can be v ery difficult to get it out a ain. A relatively slow diffusion process is required to do this. See Chapter 16. It is suggested that an e xtensive HSP analysis be done for those polymers where potential misuse or contamination of containers prior to rec ycling is a possibility . This can point out which chemicals are most lik ely to present the greatest problems.

CHEMICAL RESISTANCE Chemical resistance studies ha ve generally been performed with too fe w liquids and without the necessary spread of HSP to allow the data to be correlated with confidence. In addition, attainmen of equilibrium is not usually confirmed.These shortcomings mean that HSP correlations of chemical resistance must be done with great care. This has been discussed in Chapter 12 in more detail. An additional activity, which should be done for practical reasons, is to assign effective HSP to various test materials such as mustard, k etchup, and other gi ven products that often appear in tests of chemical resistance. Such data will allow greater use of the correlations since guidelines for potential improvements can be obtained.

CONTROLLED RELEASE HSP considerations can provide an extra formulating parameter for the controlled release of drugs. When the HSP relations between a drug and its surroundings are known, predictions of its behavior can be made. When there is a good match in the HSP v alues, the drug will be more soluble with the ability to mo ve at some rate within a polymer matrix. On the other hand it can be surrounded by a matrix with similar HSP , and this may slo w release more than desired. A poor match in HSP may leave holes and expose the drug for more rapid release. When the match is poor, the drug will not be able to permeate through its polymeric surroundings, b ut it can, of course, pass through open passageways. Drugs will also tend to adsorb at surf aces where there is a good HSP match. It has been sho wn (personal communication, Andreas Gryczk e, De gussa Pharma Polymers, Darmstadt) that calculations of HSP for a number of drugs and for a series of EUDRA GIT® polymers provided a correlation confirming that the drugs are soluble in those polymers where the HS match closely enough. When the drug and polymer HSP were within 10 MPa1/2 at the concentration studied (20% wgt.), what appeared to be true solubility w as found. This was evidenced by x-ray diffraction and DSC measurement that confirmed the drugs were embedded completely amorphou (solid solution). A closer match would presumably be required to do this for higher concentrations of the drugs. Table 18.6 contains HSP for some materials whose solubility in avrious media may have interest. It is possible to calculate HSP for essentially all drugs, although h ydrochlorides and other salts

7248_C018.fm Page 340 Wednesday, May 23, 2007 11:46 AM

340

Hansen Solubility Parameters: A User’s Handbook

TABLE 18.6 Calculated HSP for Various Materials Material

δD

δP

δH

V

Adrenaline 4-Aminopyridine Ascorbic Acid Caffeine Cycloheximide DDT Dopamine Ecstacy Meclofenoxate (base only) Norephedrin 2-Oxopyrrolidinacetamid Quinine Saccharin Serotonin Spermidin

20.5 20.4 18.0 19.5 18.3 20.0 18.2 18.0 16.0 18.0 17.5 19.0 21.0 18.0 16.7

8.7 16.1 12.6 10.1 11.0 5.5 10.3 5.1 6.2 10.7 15.6 4.6 13.9 8.2 11.2

19.9 12.9 27.6 13.0 13.8 3.1 19.5 6.1 9.0 24.1 11.2 11.0 8.8 14.4 12.0

154.5 87.1 106.7 157.9 171.0 268.8 180.0 202.9 198.3 141.9 116.2 310.7 206.8 144.4 155.6

Note: Units are MP a1/2. Source: Hansen, C.M., Conference Proceedings, Pharmaceutical and Medical Packaging 2001, Skov, H. R., Ed., Hexagon Holding, Copenhagen, 2001, pp. 20.1–20.10.

should be measured rather than calculated only , as there is v ery little e xperimental data on which to base an estimate with group contrib utions. There are correlations of HSP with skin permeation presented in Chapter 15 that should allo w an estimate of which drugs can enter via this route. Simple calculations confirm that dopamine nicotine, skatole, and nitroglycerine, for e xample, have high affinity for skin

NANOTECHNOLOGY Controlling the orientation of molecules can be a k ey for switches and the lik e. An example has been given in Chapter 15 where anthracene units appended to a polymer molecule adopted one orientation in toluene at room temperature, but changed to a new orientation at temperatures above 38.5° in toluene. The configuration at the higher temperatures as also that adopted in tetrah ydrofurane. This change of orientation is the result of a poorer solv ent at the higher temperatures compared with a good one at room temperature. More details and additional e xamples are found in Chapter 15. The following link reports what follo ws for studies of or ganically modified nanoclays (o ganoclay) in nanocomposites: http://www.hwi.buffalo.edu/ACA/ACA03/abstracts/text/W0383.html Quote: W0383 Evaluating Or ganoclays for Nanocomposites by Small Angle Scattering . Derek L. Ho 1,2 and Charles J. Glinka 1, 1Center for Neutron Research, National Institute of Standards and Technology,

7248_C018.fm Page 341 Wednesday, May 23, 2007 11:46 AM

The Future

341

Gaithersburg, MD 20899, 2Dept. of Materials and Nuclear Engineering, Univ. of Maryland, College Park, MD 20742. Understanding the interaction between or ganically modified clay (o ganoclay) platelets and or ganic solvent molecules as well as the corresponding structure of or ganoclays in a suspension is a critical step toward tailoring and characterizing nanocomposites formed by or ganoclays in a polymer matrix. Recently, nanocomposites composed of clays and polymers ha ve been found to ha ve impro ved mechanical properties as well as enhanced thermal stability . The improved properties are related to the degree of dispersal and e xfoliation of the clay platelets in the polymer matrix. In order to understand and optimize potential processing conditions, or ganoclays were dispersed in a number of or ganic solvents covering a range of solubility parameters and characterized using small-angle neutron scattering and wide-angle x-ray scattering techniques. With Hansen’s solubility parameters, δo2 = δd2 + δp2 + δh2, the correlation between the de gree of e xfoliation of or ganoclays and the solv ent in which the clay platelets are dispersed/mix ed has been analyzed. It has been found that the dispersion force of the solvent, reflected by δd, is the principal factor determining whether the clay platelets remain suspended in the solv ent while the polar ( δp) and h ydrogen-bonding ( δh) forces af fect primarily the tactoid formation/structure of the suspended platelets. The or ganically modified clays studied in this ork precipitated in an y solv ent with molecules with moderately strong h ydrogen-bonding groups. The correlation found has been used to correctly identify a solv ent, trichloroeth ylene, which completely exfoliates the or ganoclays studied in this w ork.

These and the many other examples in this handbook related to self-assembly (see abo ve), surface adsorption, molecular orientation, and af finity among molecules and molecular s gments should provide ample evidence that molecular guidance in nanotechnology endea vors can be found in the HSP concept.

THEORETICAL PROBLEMS AWAITING FUTURE RESOLUTION POLYMER SOLUBILITY The Flory chi parameter has been used to describe polymer–solvent interactions for many years.35,36 If this single parameter is to be complete in this function, it must include both the atomic/dispersion interactions as well as the specific interactions reflected by th δP and δH parameters. Attempts to calculate chi using HSP are reported in Chapter 2. More understanding is required before chi can be calculated with reasonable accuracy, but intensified e forts seem warranted. Zellers and coworkers have recently made an attempt to use this theory in conjunction with HSP studies. 37–40 A major problem is the reliability of the chi parameter (and also HSP) v alues in the literature (see Chapter 2 for more details). Chapter 4 is a length y discussion of the use of HSP in thermodynamic models for polymer solutions. The author’s current view as expressed in Chapter 2 is that the Flory approach is a special case of the more general Prigogine corresponding states theory . This is in agreement with the vie w of Patterson discussed in Chapter 2. Furthermore, the v ery general applications of the HSP approach demonstrated in this book and else where, and the apparent agreement in the treatment of specifi interactions by both the HSP and Prigogine treatments, leads to confidence in the HSP approach The geometric mean must be used in the Prigogine theory to arri ve at this similarity of treatment. The statistical thermodynamic approach of P anayiotou and co workers thoroughly discussed in Chapter 3 provides convincing evidence that the division of the cohesion energy density into three parts (at least) is the correct procedure to understand af finities among ma y types of materials. Finally, Chapter 4 deals with theories of vapor–liquid equilibrium in particular and how these relate to HSP.

7248_C018.fm Page 342 Wednesday, May 23, 2007 11:46 AM

342

Hansen Solubility Parameters: A User’s Handbook

SURFACE PHENOMENA Surfaces can clearly be characterized by HSP as demonstrated in Chapter 6 and Chapter 7. The work of Beerbo wer contained in Reference 17 and Reference 23 has also sho wn applications of HSP to such varied phenomena as the work of adhesion of liquids on mercury, friction of polyethylene untreated and treated with sulfuric acid, the Rehbinder ef fect — the crushing of aluminum oxide (Al 2O3) under v arious liquids, and the Jof fé ef fect — ef fect of liquid immersion on the fracture strength of soda-lime glass. Here ag ain, the successful use of HSP to such applications might not have been anticipated had it been considered as a parameter for use in bulk systems only. A formalized unifying theory linking HSP to both b ulk and surf ace phenomena is still lacking. Presently, the best that can be said is that the generality “lik e dissolves like” can be quantified i many cases. The extension of this, “lik e seeks lik e,” also seems to ha ve been demonstrated. It is the surfaces of molecules which interact with each other (also in b ulk and solution phenomena), so it is not surprising that cohesion parameters can be applied with success to both solubility and surface phenomena. Much more research needs to be done with these relations. A good starting point is the Handbook of Surface and Colloid Chemistry, edited by K.S. Birdi. 41 If we consider chromatographic techniques as depending primarily on surf ace phenomena, mention should also be made of the e xtension of the three-parameter HSP approach to a fi e-parameter approach by Karger and coworkers.42 HSP characterizations of surf actants are also badly needed.

CONCLUSION HSP have been sho wn useful in solv ent selection; predicting polymer–polymer miscibility; char acterizing the surfaces of polymers, fillers, and fibers; correlating permeation phenomena; chara terizing organic salts and inor ganic salts; g as solubility; etc. No other parameter can be assigned to such a range of materials spanning from g ases and liquids, o ver surf aces, to inor ganic salts. These results and the close relation with the Prigogine corresponding states theory of polymer solutions, and more recently to the w ork of P anayiotou summarized in Chapter 3, indicate that a still more general theory exists. This theory should quantify the adage “like seeks like,” i.e., include surface phenomena as well as b ulk phenomena. Specific areas needing more theoretical ork related to HSP in the near future include better understanding of usage for predictions of beha vior for w ater, gases, organic salts, inor ganic salts, and organometallic compounds. Water remains special because of its lo w molecular v olume and high δH. Most materials having HSP in the range of the customary test liquids can be studied using HSP with reasonable success. This is not fully the case for g ases, many of which have much lower HSP than the well-studied liquids, and salts, many of which apparently have HSP very much higher than any of the liquids. Extensions of practical applications related to chemical resistance and the uptake of potentially dangerous materials in polymers are also required. Finally there is a great deal to be done in the areas of controlled drug release, impro ved understanding of some biological processes, and last b ut not least, the systematic use of HSP in nanotechnology to control the assembly and orientation of molecules or se gments of molecules.

REFERENCES 1. Hansen, C.M., The universality of the solubility parameter , Ind. Eng. Chem. Prod. Res. Dev., 8(1), 2–11, 1969. 2. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press, Boca Raton, FL, 1983; 2nd ed., 1991. 3. Barton, A.F.M., Handbook of Polymer-Liquid Interaction Parameters and Solubility Parameters, CRC Press, Boca Raton, FL, 1990.

7248_C018.fm Page 343 Wednesday, May 23, 2007 11:46 AM

The Future

343

4. Hansen, C.M., Solubility parameters, inPaint Testing Manual, Manual 17, Koleske, J.V., Ed., American Society for Testing and Materials, Philadelphia, P A, 1995, pp. 383–404. 5. Hansen, C.M., Cohesion energy parameters applied to surface phenomena, CRC Handbook on Surface Science, Birdi, K.S., Ed., CRC Press, Boca Raton, FL, 1997, chap. 10. 6. Anonymous, Brochure: Co-Act — A Dynamic Program for Solv ent Selection, Exxon Chemical International Inc., 1989. 7. Dante, M.F., Bittar , A.D., and Caillault, J.J., Program calculates solv ent properties and solubility parameters, Mod. Paint Coat., 79(9), 46–51, 1989. 8. Hansen, C.M. and Beerbo wer, A., Solubility parameters, in Kirk-Othmer Encyclopedia of Chemical Technology, Suppl. Vol., 2nd ed., Standen, A., Ed., Interscience, Ne w York, 1971, pp. 889–910. 9. Hansen, C.M. and Andersen, B.H., The affinities of o ganic solv ents in biological systems, J. Am. Ind. Hyg. Assoc., 49(6), 301–308, 1988. 10. Wessling, R.A., The solubility of poly(vinylidene chloride), J. Appl. Polym. Sci., 14, 1531–1545, 1970. 11. Grulke, E.A., Solubility parameter values, in Polymer Handbook, 3rd ed., Brandrup, J. and Immergut, E.H., Eds., John Wiley-Interscience, New York, 1989, pp. VII/519–559. 12. Burrell, H., Solubility of polymers, in Kirk Othmer Encycopedia of Polymer Science and Technology, Vol. 12, 2nd ed., John Wiley & Sons, Ne w York, 1970, pp. 618–626. 13. Hoy, K.L., New values of the solubility parameters from vapor pressure data, J. Paint Technol., 42(541), 76–118, 1970. 14. Hoy, K.L., Tables of Solubility Parameters, Union Carbide Corp., Research and De velopment Dept., South Charleston, WV, 1985; 1st ed. 1969. 15. van Krevelen, D.W. and Hoftyzer, P.J., Properties of Polymers: Their Estimation and Correlation with Chemical Structure, 2nd ed., Else vier, Amsterdam, 1976. 16. Riedel, G., The solubility of colorants in organic solvents (Löslichkeit von Farbstoffen in organischen Lösungsmitteln, in German), Farbe und Lack, 82(4), 281–287, 1976. 17. Beerbower, A., Environmental capability of liquids, in Interdisciplinary Approach to Liquid Lubricant Technology, NASA Publication SP-318, 1973, 365–431. 18. Fedors, R.F., A method for estimating both the solubility parameters and molar v olumes of liquids, Polym. Eng. Sci., 14(2), 147–154, 472, 1974. 19. Koenhen, D.N. and Smolders, C.A., The determination of solubility parameters of solv ents and polymers by means of correlation with other physical quantities, J. Appl. Polym. Sci., 19, 1163–1179, 1975. 20. Utracki, L., Personal communication, 1995. 21. Kähler, T. and Knudsen, S.L., Student Report, Technical University of Denmark, 1967. 22. Hansen, C.M., Characterization of surfaces by spreading liquids, J. Paint Technol., 42(550), 660–664, 1970. 23. Beerbower, A., Boundary Lubrication — Scientific and Technical Applications Forecast, AD747336, Office of the Chief of Research and D velopment, Department of the Army, Washington, D.C., 1972. 24. Hansen, C.M. and Björkman, A., The ultrastructure of w ood from a solubility parameter point of view, Holzforschung, 52(4), 335–344, 1998. 25. Hansen, C.M. and Knudtson, J., Spreading and wetting of an electrodeposition coating (Ausbreitung und Benetzung bei einem Elektrotauchlack, in German), Farbe und Lack, 2, 115–118, 1973. 26. Progress in Organic Coatings, Special issue de voted to Self-Stratifying Coatings, 28(3), July 1996. 27. Thulstrup, D., Characterization of surf aces by static contact angle measurements (lecture in Danish: Karakterisering af Overflader ed Statisk Kontaktvinkelmåling), Surface Properties and Modificatio of Plastics (Conference Notes), Ingeniørforening i Danmark, No vember 13, 1997, Copenhagen. See also DSM Materialenyt 2:97, Dansk Selskab for Materialeproe vning og -forskning, Ingeniørforening in Danmark, Copenhagen, 1997, 63–69. 28. König, U. and Schuch, H., Molecular composition and permeability of plastics, (K onstitution und Permeabilität von Kunststoffen, in German), Kunststoffe, 67(1), 27–31, 1977. 29. Hansen, C.M., 25 years with solubility parameters (25 År med Opløselighedsparametrene, in Danish), Dan. Kemi, 73(8), 18–22, 1992. 30. Hansen, C.M., Some aspects of acid/base interactions (Einige Aspekte der Säure/Base-W echselwirkung, in German), Farbe und Lack, 83(7), 595–598, 1977.

7248_C018.fm Page 344 Wednesday, May 23, 2007 11:46 AM

344

Hansen Solubility Parameters: A User’s Handbook

31. Beerbower, A., Personal communication, Predicting Wetting in the Oil-Brine-Rock System, Unedited manuscript for presentation at the National AIChE Meeting, Tulsa, OK, March 10–13, 1974. 32. Panzer, J., Components of solid surf ace free energy from wetting measurements, J. Colloid Interface Sci., 44(1), 142–161, July 1973. 33. Hildebrand, J. and Scott, R.L., The Solubility of Nonelectrolytes, 3rd ed., Reinhold, Ne w York, 1950. 34. Hansen, C.M., Solubility parameters for aromas and scents (Aromastof fers Opløselighedsparametre, in Danish), Plus Process, 11(9), 16–17, 1997. 35. Flory, P.J., Principles of Polymer Chemistry, Cornell University Press, New York, 1953. 36. Eichinger, B.E. and Flory , P.J., Thermodynamics of polymer solutions, Trans. Faraday Soc., 64(1), 2035–2052, 1968; 64(2), 2053–2060, 1968; 64(3), 2061–2065, 1968; 64(4), 2066–2072, 1968. 37. Zellers, E.T., Three-dimensional solubility parameters and chemical protecti ve clothing permeation. I. Modelling the solubility of organic solvents in Viton® gloves, J. Appl. Polym. Sci., 50, 513–530, 1993. 38. Zellers, E.T., Three-dimensional solubility parameters and chemical protective clothing. II. Modelling diffusion coefficients, breakthrough times, and steady-state permeation rates of o ganic solv ents in Viton® gloves, J. Appl. Polym. Sci., 50, 531–540, 1993. 39. Zellers, E.T., Anna, D.H., Sulewski, R., and Wei, X., Critical analysis of the graphical determination of Hansen’s solubility parameters for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2069–2080, 1996. 40. Zellers, E.T., Anna, D.H., Sule wski, R., and Wei, X., Impro ved methods for the determination of Hansen’s solubility parameters and the estimation of solv ent uptake for lightly crosslinked polymers, J. Appl. Polym. Sci., 62, 2081–2096, 1996. 41. Birdi, K.S., Handbook of Surface and Colloid Chemistry, CRC Press, Boca Raton, FL, 1997. 42. Karger, B.L., Sn yder, L.R., and Eon, C., Expanded solubility parameter treatment for classificatio and use of chromatographic solv ents and adsorbents, Anal. Chem., 50(14), 2126–2136, 1978.

7248_A001.fm Page 345 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1 Charles M. Hansen (with the help of Hanno Priebe) COMMENTS TO TABLE A.1 Table A-1 is greatly e xpanded compared with the first edition of this handbook. There are more chemicals, and the information for each of them is much more comprehensi ve. There are two lists for chemical names. The first list foll ws the general nomenclature that is used in the first edition This list follo ws usage that is commonly found in industrial conte xts. The second list of names and also the structures ha ve been compiled by Dr . Hanno Priebe, GE Healthcare. Most of these chemical names originate from AutoNom 2000 (Cop yright © 1988-2006, Beilstein-Institut zur Förderung der Chemischen Wissenschaften licensed to Beilstein GmbH and MDL Information Systems GmbH. All rights reserv ed). These are ob viously more correct in man y cases. In Autonom the Beilstein ring system naming convention was chosen. In cases where Autonom failed, the chemical names originate from ACD/Name freeware, version 8.0 (Advanced Chemistry Development, Inc., Toronto ON, Canada, www.acdlabs.com, 2005). In cases where the ACD/Name freeware failed, the name field as left empty. Most of the HSP entries in Table A-1 are calculated v alues. Those entries ha ving some experimental confirmation are the original 90 from Reference 1 to Reference 3, together with som from industrial experience at PPG Industries. These are indicated in dark type. The usual procedure for the calculations w as to use the procedure gi ven in Chapter 1. The figures were used to fin δD from boiling point data. δP was estimated either with a dipole moment, where this could be found, or by group contrib utions. There are special problems where the dipole moment is zero or close to zero for symmetrical molecules. In some cases tw o sets of data are given. δH was almost always found by group contrib utions or by similarity to related compounds. It must be said that none of the parameters are precise. Even use of experimental data does not necessarily confirm the accura y of the values. This is particularly true where a gi ven HSP parameter is relati vely high. Whether or not given solutes dissolv e is not suf ficient in most cases to narr w the potential error(s) in such cases. Where latent heats are kno wn, Equations 1.6–1.8 do limit potential errors, and e xperience has shown that for all practical purposes, the v alues assigned are most useful. Notwithstanding, several parameters ha ve been changed since the first edition appeared. These are indicated with a “*” in the first column of names. H wever, no changes in the older , published figures h ve been made for this reason, as this w as judged unnecessary. Many of the solv ents for which e xperimental data were used to help assign the HSP were commercial in origin. This means that purity of the samples w as that specified in the g ven case, but was usually very high, and that a mixture of isomers was present if isomers were possible. This is particularly true of the glycol ethers based on prop ylene oxide. F or these chemicals and others where isomers are possible, the chemical names and structures indicated in Table A-1 are indicative of the type of compound present. It w ould be impractical to present all the possible isomers. The HSP for the or ganic salts formed from equimolar mixtures of or ganic acids and amines are those reported in Reference 4. The assignments were made on the basis of which polymers or other solutes in a test series dissolv ed in the gi ven salt. This was compared with ho w these same solutes behaved in a series of liquids with kno wn HSP. When the same pattern of dissolving or not w as found for both a salt and a known liquid, it was assumed that their HSP were the same (or at least very close). The units used are MP a1/2 for the HSP and cc/mole for the molar v olume.

345

7248_A001.fm Page 346 Wednesday, May 23, 2007 12:31 PM

346

Hansen Solubility Parameters: A User’s Handbook

REFERENCES 1. Hansen, C.M., The Three Dimensional Solubility P arameter and Solvent Diffusion Coefficient, Doc toral Dissertation, Danish Technical Press, Copenhagen, 1967. 2. Hansen, C.M., The Three Dimensional Solubility P arameter — Key to Paint Component Affinities I — Solvents, Plasticizers, Polymers, and Resins, J. Paint Techn., 39(505), 104-117, 1967. 3. Hansen, C.M., and Skaarup, K., The Three Dimensional Solubility P arameter — Key to Paint Component Affinities III. — Independent Calculation of the arameter Components, J. Paint Techn. 39(511), 511-514, 1967. 4. Hansen, C.M., Some Aspects of Acid/Base Interactions (in German) (Einige Aspekte der Säure/BaseWechselwirkung), Farbe und Lack, 83(7), 595-598, 1977.

7248_A001.fm Page 347 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

347

TABLE A.1 No. 1

Solvent Name Acetaldehyde*

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetaldehyde

14.7

12.5

7.9

56.6

Acetaldehyde oxime

16.3

4.0

20.2

61.2

Acetamide

17.3

18.7

22.4

60.8

N-Phenyl-acetamide

20.6

13.3

12.4

110.9

Acetic acid

14.5

8.0

13.5

57.1

Acetic anhydride

16.0

11.7

10.2

94.5

Propan-2-one

15.5

10.4

7.0

74.0

2-Hydroxy-2-methylpropionitrile

16.6

12.2

15.5

94.0

O H3C 2

Acetaldoxime

H3C 3

N

OH

Acetamide

O H3C 4

NH2

Acetanilide

O CH3

N 5

Acetic Acid

O H3C OH 6

Acetic Anhydride

O

O

O

H3C 7

Acetone

CH3

O H3C 8

CH3

Acetonecyanhydrin N HO H3C

CH3

7248_A001.fm Page 348 Wednesday, May 23, 2007 12:31 PM

348

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 9

Autonom/ ACD Name

Solvent Name Acetonemethyloxime

Propan-2-one O-methyl- 14.7 oxime

CH3 N

4.6

4.6

96.7

O

Acetonitrile

Acetonitrile

15.3

18.0

6.1

52.6

1-Phenyl-ethanone

19.6

8.6

3.7

117.4

Propan-2-one oxime

16.3

3.7

10.9

80.2

Acetic acid b uta-1,3dienyl ester

16.1

4.4

8.3

118.4

1-Acetyl-azepan-2-one

18.9

8.7

4.8

155.0

1-Morpholin-4-ylethanone

18.3

5.3

7.8

115.6

1-Piperazin-1-ylethanone

18.5

10.0

6.5

125.8

N

H3C 11

Hydrogen Molar Bonding Volume

CH3

H3C

10

Dispersion Polarity

Acetophenone O CH3

12

Acetoxime

OH

N

CH3

H3C 765

1-Acetoxy-1,3-butadiene

O

H3C

CH2

O 13

N-Acetyl Caprolactam O

O N

14

CH3

N-Acetyl Morpholine CH3

O

N O

15

N-Acetyl Piperidine CH3

O N N

7248_A001.fm Page 349 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

349

TABLE A.1 (CONTINUED) No. 16

Solvent Name N-Acetyl Pyrrolidone O

O N

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Acetyl-pyrrolidin-2one

17.8

13.1

8.3

127.0

2-Acetoxy-benzoic acid

19.0

6.6

9.3

133.5

1-Thiophen-2-ylethanone

19.1

12.2

9.3

108.0

3-Acetoxy-3ethoxycarbonylpentanedioic acid diethyl ester

16.6

3.5

8.6

279.9

Pentane-2,4-dione

16.1

11.2

6.2

103.0

Acetyl bromide

16.7

10.6

5.2

74.0

Acetyl chloride

16.2

11.2

5.8

71.4

Ethyne

14.4

4.2

11.9

42.1

CH3

1116 Acetyl Salicylic Acid O

Autonom/ ACD Name

OH O

O

CH3

785

2-Acetyl Thiophene O S

CH3

1219 Acetyl Triethyl Citrate O O H3C

O

O O

CH3 O

CH3

O O

CH3

17

Acetylacetone

O

O

CH3

H3C 18

Acetylbromide

O Br

H3C 19

Acetylchloride

O Cl

H3C 20

Acetylene (Ethyne)

HC

CH

7248_A001.fm Page 350 Wednesday, May 23, 2007 12:31 PM

350

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 21

Solvent Name Acetylfluorid

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetyl fluorid

14.7

14.0

5.7

62.0

Acridine

21.7

5.9

2.0

163.0

Propenal

15.0

7.2

7.8

66.7

Acrylamide

15.8

12.1

12.8

63.4

Acrylic acid

17.7

6.4

14.9

68.5

Acrylonitrile

16.0

12.8

6.8

67.1

Acryloyl chloride

16.2

11.6

5.4

81.3

Hexanedioic acid

17.1

9.0

14.6

107.5

4-((R)-1-Hydroxy-2methylamino-ethyl)benzene-1,2-diol

20.5

8.7

19.9

154.5

O H3C 860

F

Acridine

N 22

Acrolein

O

H2C 23

Acrylamide

O

H2C

NH2 24

Acrylic Acid

O

H2C OH 25

Acrylonitrile

H2C 26

N

Acrylylchloride

O

H2C Cl

1152 Adipic Acid (1,6-Hexanedioic) O OH

HO O

1195 Adrenaline HO HO

N OH

CH3

7248_A001.fm Page 351 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

351

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1088 alfa-Chloro Acetophenone O

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Chloro-1-phenylethanone

20.0

9.6

5.7

116.8

Acetic acid allyl ester

15.7

4.5

8.0

108.5

Pent-4-enoic acid

16.7

4.7

11.3

102.1

3-Oxo-butyric acid allyl ester

15.9

6.9

8.6

137.8

Pent-4-enenitrile

16.3

11.2

5.0

98.5

Prop-2-en-1-ol

16.2

10.8

16.8

68.4

Allylamine

15.5

5.7

10.6

74.9

3-Bromo-propene

16.5

7.3

4.9

86.5

3-Chloro-propene

17.0

6.2

2.3

82.3

But-3-enenitrile

16.0

14.3

5.6

80.5

Cl

27

Allyl Acetate

CH3

O

H2C

O 28

Allyl Acetic Acid

O

OH

H2C 769

Allyl Acetoacetate

O

O

H3C 29

O

Allyl Acetonitrile (4-Pentenenitrile)

CH2

N 30

Allyl Alcohol

H2C 31

Br

Allyl Chloride

H2C 34

NH2

Allyl Bromide (3-Bromoprene)

H2C 33

OH

Allyl Amine

H2C 32

CH2

Cl

Allyl Cyanide

N H2C

7248_A001.fm Page 352 Wednesday, May 23, 2007 12:31 PM

352

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 35

Solvent Name Allyl Ethylether

O

H2C 36

Dispersion Polarity

Hydrogen Molar Bonding Volume

3-Ethoxy-propene

15.0

4.8

5.1

112.6

3-Fluoro-propene

14.9

6.4

1.0

71.5

Formic acid allyl ester

15.7

5.4

8.8

91.0

3-Iodo-propene

17.4

6.4

3.3

90.8

3-Isocyano-propene

16.1

13.0

5.4

84.5

3-Isopropoxy-propene

15.0

4.1

7.1

129.0

3-Isothiocyanatopropene

17.0

11.3

8.5

98.0

Prop-2-ene-1-thiol

16.4

6.2

7.9

80.2

2-Methyl-acrylic acid allyl ester

15.2

4.1

7.5

134.8

3-Methoxy-propene

15.0

4.3

5.9

93.7

CH3

Allyl Fluoride

F

H2C 37

Autonom/ ACD Name

Allyl Formate

O

CH2

O 38

Allyl Iodide

I

H2C 39

Allyl Isocyanide

H2C

+

N 967

C

Allyl Isopropyl Ether

O

H2C

CH3

CH3 40

Allyl Isothiocyanate

N

H2C 42

S

Allyl Mercaptan

SH

H2C

1018 Allyl Methacrylate CH3 O

H2C

CH2

O

43

Allyl Methylether

H2C

O

CH3

7248_A001.fm Page 353 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

353

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1181 2-Amino-2-Methyl-1-Propanol/Acetic Acid O O

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetate 2-hydroxy-1,1dimethyl-ethylammonium

17.2

22.5

23.3

1-Amino-cyclopropanecarboxylic acid

17.0

6.3

13.6

93.7

Pyridin-2-ylamine

20.4

8.1

12.2

94.1

Pyridin-4-ylamine

20.4

16.1

12.9

87.1

Ammonia

13.7

15.7

17.8

20.8

1-Methyl-2-phenylethylamine

17.5

4.3

6.3

148.1

Acetic acid pentyl ester

15.8

3.3

6.1

148.0

1-Methoxy-4-((E)propenyl)-benzene

19.0

4.3

8.7

150.0

+

NH3

OH

1206 1-Aminocyclopropane Carboxylic Acid

O H2N

OH

1090 2-Amino Pyridine

N

NH2

1205 4-Amino Pyridine NH2

N

44

Ammonia

NH3 1179 Amphetamine

CH3 NH2 45

Amyl Acetate

H3C

O

CH3

O 699

Anethole (Trans)

H3C

CH3 O

7248_A001.fm Page 354 Wednesday, May 23, 2007 12:31 PM

354

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 46

Solvent Name Aniline

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Phenylamine

19.4

5.1

10.2

91.5

2-Methoxybenzaldehyde

19.4

11.9

8.3

120.8

2-Methoxy-phenylamine 19.5

5.4

11.4

112.2

4-Methoxy-phenylamine 19.9

6.5

11.3

113.3

Methoxy-benzene

17.8

4.1

6.7

119.1

Anthraquinone

20.3

7.6

4.8

145.6

(R)-5-((S)-1,218.0 Dihydroxy-ethyl)-3,4dihydroxy-5H-furan-2one

12.6

27.6

106.7

NH2 1113 Anisaldehyde (2-Methoxy Benzaldehyde) O O

949

CH3

Anisidine (2-Methoxyaniline) NH2 O

47

CH3

p-Anisidine (Methoxy Aniline) CH3 O NH2

48

Anisole

H3C

O

1099 9,10-Anthraquinone O

O

1163 Ascorbic Acid OH O HO HO

O OH

7248_A001.fm Page 355 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

355

TABLE A.1 (CONTINUED) No. 49

Solvent Name Azidoethane

H3C 50

N

N

N

807

Benzal Chloride

Hydrogen Molar Bonding Volume

Azido-ethane

15.9

8.9

12.9

79.0

3-Azido-propene

16.8

7.7

13.4

83.0

N

Dichloromethyl-benzene 19.9

6.6

2.4

134.2

Benzaldehyde

19.4

7.4

5.3

101.5

Benzamide

21.2

14.7

11.2

90.3

2.0

89.4

N

Cl

Cl

51

Dispersion Polarity

N

3-Azidopropene

H2C

Autonom/ ACD Name

Benzaldehyde O

1055 Benzamide O NH2

52

Benzene

Benzene

18.4

0

53

1,3-Benzenediol

Benzene-1,3-diol

18.0

8.4

21.0

87.5

1H-Benzoimidazole

20.6

14.9

11.0

102.7

OH

OH

894

Benzimidazole

N N

7248_A001.fm Page 356 Wednesday, May 23, 2007 12:31 PM

356

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 893

Solvent Name Benzisoxazole

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Benzo[d]isoxazole

20.6

11.5

8.8

100.7

Benzofuran

18.7

5.1

5.7

110.2

Benzoic acid

18.2

6.9

9.8

113.1

2-Hydroxy-1,2diphenyl-ethanone

19.9

9.7

10.7

187.7

Benzonitrile

17.4

9.0

3.3

102.6

Diphenyl-methanone

19.6

8.6

5.7

164.2

Benzothiazole

20.6

5.2

8.4

108.5

1H-Benzotriazole

18.7

15.6

12.4

96.2

N O 892

2,3-Benzofuran (Cumaron)

O 54

Benzoic Acid

OH O

891

Benzoin O

OH

55

Benzonitrile N

56

Benzophenone O

962

Benzothiazole

N S 1053 1,2,3-Benzotriazole

N N N

7248_A001.fm Page 357 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

357

TABLE A.1 (CONTINUED) No. 866

Benzotrichloride Cl

57

Autonom/ ACD Name

Solvent Name

Cl

Cl

Benzoyl Chloride

Dispersion Polarity

Hydrogen Molar Bonding Volume

Trichloromethylbenzene

20.2

6.6

3.2

142.1

Benzoyl chloride

20.7

8.2

4.5

116.0

Acetic acid benzyl ester

18.3

5.7

6.0

142.8

Phenyl-methanol

18.4

6.3

13.7

103.6

Benzylamine

19.2

4.6

11.7

109.2

Benzoic acid benzyl ester

20.0

5.1

5.2

191.2

Terephthalic acid 1benzyl ester 4-b utyl ester

19.0

11.2

3.1

306.0

Chloromethyl-benzene

18.8

7.1

2.6

115.0

Cl O

903

Benzyl Acetate O O

58

CH3

Benzyl Alcohol

OH

971

Benzyl Amine

NH2

1149 Benzyl Benzoate O O

59

60

Benzyl Butyl Phthalate O

O

O

O

Benzyl Chloride

Cl

CH3

7248_A001.fm Page 358 Wednesday, May 23, 2007 12:31 PM

358

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 61

Solvent Name Benzyl Methacrylate CH3

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-acrylic acid benzyl ester

16.8

4.1

4.1

167.8

1-Benzyl-pyrrolidin-2one

18.2

6.1

5.6

160.0

Ethoxymethyl-benzene

18.4

3.8

3.8

144.2

O

H2C

O

62

N-Benzyl Pyrrolidone O N

887

Benzylethyl Ether

CH3

O

867

Bicyclohexyl

Bicyclohexyl

18.6

0

0

188.5

868

Biphenyl

Biphenyl

19.7

1.0

2.0

155.1

Dicarbonimidic diamide 20.0

14.6

18.8

70.3

17.0

10.2

6.9

41.8

16.6

2.5

0

85.3

1066 Biuret

O N

H2N 64

O NH2

Borine Carbonyl +

H3B

O

1172 Boron Trichloride

Cl

B

Cl

Cl 1170 Bromine (P From Group Cont.)

Bromine

18.2

14.9

0

51.5

1213 Bromine (P From Dipole Moment)

Bromine

18.2

2.1

0

51.5

7248_A001.fm Page 359 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

359

TABLE A.1 (CONTINUED) No. 65

Autonom/ ACD Name

Solvent Name 2-Bromo Allyl Alcohol

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Bromo-prop-2-en-1-ol 17.1

9.9

16.2

84.5

2-Bromo-propene

16.1

6.0

4.9

88.8

(Z)-1-Bromo-propene

16.3

6.4

5.0

84.7

4-Bromo-but-1-ene

16.5

6.0

4.5

102.0

4-Bromo-buta-1,2-diene

17.0

6.5

4.7

93.3

4-Bromo-1-nitro-2trifluoromet ylbenzene

20.0

6.0

4.9

150.1

1-Bromo-4-ethoxybenzene

19.5

7.7

5.3

140.0

Bromo-ethyne

15.7

9.9

5.6

67.7

1-Bromo-2-methoxybenzene

19.8

8.4

6.7

124.5

Br OH

H2C 66

2-Bromo Propene

Br CH2

H3C 67

1-Bromo Propene (CIS)

H3C Br 68

4-Bromo-1-Butene

H2C 69

Br

4-Bromo-1,2-Butadiene

Br

H2C 854

5-Bromo-2-Nitrobenzotrifluorid F

F

F

O +

N

O

Br

815

1-Bromo-4-Ethoxy Benzene

H3C

O Br

70

Bromoacetylene

Br

HC 800

o-Bromoanisole O Br

CH3

7248_A001.fm Page 360 Wednesday, May 23, 2007 12:31 PM

360

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 71

Solvent Name Bromobenzene

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Bromo-benzene

20.5

5.5

4.1

105.3

4-Bromo-benzonitrile

20.4

9.3

5.8

113.8

4-Bromo-benzoyl chloride

20.2

6.5

5.5

137.2

2-Bromo-butane

16.3

7.7

4.4

109.5

1-Bromo-2-chlorobenzene

20.3

7.7

4.6

116.9

Bromo-chloro-methane

17.3

5.7

3.5

65.0

Bromo-ethene

16.1

6.3

2.3

71.6

Tribromo-methane

21.4

4.1

6.1

87.5

Bromo-methoxymethane

16.9

8.5

7.0

81.6

Br 806

p-Bromobenzonitrile N

Br

805

p-Bromobenzoyl Chloride Br Cl O

729

2-Bromobutane

Br CH3

H3C 802

o-Bromochlorobenzene Cl Br

72

Bromochloromethane

Br 73

Cl

Bromoethylene

Br

H2C 74

Bromoform

Br

Br Br 748

Bromomethyl Methyl Ether

Br

O

CH3

7248_A001.fm Page 361 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

361

TABLE A.1 (CONTINUED) No. 75

Solvent Name 1-Bromonaphtalene

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Bromo-naphthalene

20.3

3.1

4.1

140.0

1-Bromo-4-nitrobenzene

20.9

9.9

5.9

103.7

2-Bromo-buta-1,3-diene

16.9

6.4

4.7

95.2

1-Bromo-propane

16.4

7.9

4.8

90.9

2-Bromo-propane

16.1

8.3

4.7

93.6

3-Bromo-propyne

18.1

6.5

5.3

75.3

1-Bromo-2-vinylbenzene

19.5

5.2

5.3

130.0

2-Bromo-thiophene

20.1

5.2

4.6

96.8

1-Bromo-2-methylbenzene

19.3

5.0

4.2

119.0

Br

803

p-Bromonitrobenzene Br O

+

N

O

76

Bromoprene

Br CH2

H2C 1124 1-Bromopropane

Br

H3C 1125 2-Bromopropane

Br CH3

H3C 77

3-Bromopropyne

HC Br 710

o-Bromostyrene Br CH2

78

2-Bromothiophene

Br

S 713

o-Bromotoluene CH3 Br

7248_A001.fm Page 362 Wednesday, May 23, 2007 12:31 PM

362

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 714

Solvent Name p-Bromotoluene

Autonom/ ACD Name 1-Bromo-4-methylbenzene

Dispersion Polarity

Hydrogen Molar Bonding Volume

19.3

6.8

4.1

18.3

0.8

0

99.2

Bromo-trichloromethane

18.3

8.1

6.0

99.2

Bromotrichloro Methane (P from Group Bromo-trichlorocont.) methane

18.3

8.1

0

99.2

Bromo-trifluoro methane

9.6

2.4

0

97.0

Buta-1,2-diene

14.7

1.7

6.2

82.3

Buta-1,3-diene

14.8

2.8

5.6

83.2

1-Chloro-buta-1,3-diene

15.6

8.5

2.0

92.2

Buta-2,3-dien-1-ol

16.2

6.6

16.8

76.5

CH3

122.9

Br

743

Bromotrichloro Methane (P from Dipole Bromo-trichloroMoment) methane

Cl

Cl Cl 744

Bromotrichloro Methane (P and H from Group cont.)

Cl

Cl Cl 745

Cl

Br

Bromotrifluoromethane (Freon 1381

F

F F

80

Br

Cl

Cl

79

Br

Br

1,2-Butadiene

H3C 81

1,3-Butadiene

H2C 82

CH2

CH2

1,3-Butadiene-1-Chloro

Cl 83

CH2

2,3-Butadiene-1-ol

OH H2C

7248_A001.fm Page 363 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

363

TABLE A.1 (CONTINUED) No. 84

Autonom/ ACD Name

Solvent Name 1,3-Butadiene-1,2-Di-Chloro

Cl

Butadiene-4-Cyano

H2C 86

Hydrogen Molar Bonding Volume

1,2-Dichloro-buta-1,3diene

17.0

10.7

2.8

102.5

Penta-2,4-dienenitrile

16.2

11.7

5.2

93.7

[2,2']Bioxiranyl

18.3

14.4

6.2

77.8

Butane-2,3-dione

15.7

5.1

6.8

87.8

Acrylic acid 416.8 acryloyloxy-butyl ester

9.1

4.2

194.4

Butane

14.1

0

0

101.4

Butane-1,3-diol

16.6

10.0

21.5

89.9

Butane-1,4-diol

16.6

15.3

21.7

88.9

Butane-1-thiol

16.3

5.3

4.5

107.8

Butan-1-ol

16.0

5.7

15.8

91.5

CH2

Cl 85

Dispersion Polarity

N

Butadienedioxide

O O 87

Butadione O CH3

H3C O

88

1,4-Butandiol Diacrylate O O

H2C

O

CH2

O

89

Butane

CH3

H3C 90

1,3-Butanediol

OH H3C 730

1,4-Butanediol

HO 91

OH

1-Butanethiol

H3C 92

OH

SH

1-Butanol

H3C

OH

7248_A001.fm Page 364 Wednesday, May 23, 2007 12:31 PM

364

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 92

Autonom/ ACD Name

Solvent Name 2-Butanol

Dispersion Polarity

Hydrogen Molar Bonding Volume

Butan-2-ol

15.8

5.7

14.5

92.0

But-1-ene

13.2

1.3

3.9

94.3

(Z)-But-2-ene

14.7

1.3

4.1

90.3

(E)-But-2-ene

14.6

0

2.9

92.9

But-3-enenitrile

16.2

14.3

5.6

80.5

4-Methoxy-but-1-ene

15.1

5.3

5.2

111.9

(Z)-1-Methoxy-but-2ene

15.2

3.4

5.0

118.8

(E)-1-Methoxy-but-2ene

15.3

4.4

4.3

110.4

1-Butoxy-3-ethoxypropan-2-ol

15.5

6.5

10.2

185.9

CH3

H3C OH 94

1-Butene

CH3

H2C 95

2-Butene (cis)

H3C CH3 96

2-Butene (trans)

CH3

H3C 97

3-Butenenitrile

N H2C 98

1-Butenyl Methyl Ether

O

H2C 99

CH3

2-Butenyl Methyl Ether (cis)

CH3 O 100

2-Butenyl Methyl Ether (trans)

O

H3C 101

CH3

CH3

Butoxy Ethoxy Pr opanol Commercial H3C

O

O

OH

CH3

7248_A001.fm Page 365 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

365

TABLE A.1 (CONTINUED) No. 728

Autonom/ ACD Name

Solvent Name 3-Butoxybutanol

Dispersion Polarity

Hydrogen Molar Bonding Volume

3-Butoxy-butan-1-ol

15.9

5.5

10.6

166.3

Acetic acid b utyl ester

15.8

3.7

6.3

132.5

Acetic acid sec-b utyl ester

15.0

3.7

7.6

133.6

Acetic acid tert-b utyl ester

15.4

6.2

6.2

134.1

3-Oxo-butyric acid butyl 16.6 ester

5.8

7.3

164.3

Acrylic acid b utyl ester

15.6

6.2

4.9

143.8

2-Methyl-propan-2-ol

15.2

5.1

14.7

95.8

Butylamine

16.2

4.5

8.0

99.0

CH3

O OH

H3C

102

n-Butyl Acetate

O

103

CH3

O

H3C

Sec-Butyl Acetate

CH3

O

963

tert-Butyl Acetate

CH3

O

CH3 CH3

O

H3C 768

CH3

O

H3C

n-Butyl Aceto Acetate

O

O

104

CH3

O

H3C n-Butyl Acrylate

CH3

O

H2C O 611

t-Butyl Alcohol

H3C H3C 105

CH3 OH

n-Butyl Amine

H3C

NH2

7248_A001.fm Page 366 Wednesday, May 23, 2007 12:31 PM

366

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1182 N-Butyl Amine/Acetic Acid

O

+

NH3

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetate butylammonium;

16.0

20.3

18.4

Benzoic acid b utyl ester

18.3

5.6

5.5

178.0

Butyric acid b utyl ester

15.6

2.9

5.6

166.7

Butyl-cyclohexane

16.2

0

0.6

176.7

Butyl-cyclopentane

16.4

0

1.0

162.0

Formic acid b utyl ester

15.7

6.5

9.2

114.8

1-Isopropenyloxybutane

14.8

5.3

5.0

145.7

2-Hydroxy-propionic acid butyl ester

15.8

6.5

10.2

149.0

O 718

Butyl Benzoate* O CH3

O

774

n-Butyl Butyrate

O

722

CH3

O

H3C

n-Butyl Cyclohexane

CH3

723

n-Butyl Cyclopentane H3C

927

Butyl Formate

O

H3C 106

O

Butyl Isopropenyl Ether

CH2 O

H3C 107

CH3

Butyl Lactate OH

O

H3C O

CH3

7248_A001.fm Page 367 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

367

TABLE A.1 (CONTINUED) No. 108

Autonom/ ACD Name

Solvent Name n-Butyl Methacrylate CH3

O

H2C

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-acrylic acid butyl ester

15.6

6.4

6.6

159.4

2-Butyl-octan-1-ol

16.1

3.6

9.3

224.2

Propionic acid b utyl ester

15.7

5.5

5.9

149.7

1-Butyl-pyrrolidin-2-one 17.5

9.9

5.8

148.0

2-Hydroxy-benzoic acid butyl ester

17.9

4.8

11.7

181.7

Octadecanoic acid butyl ester

14.5

3.7

3.5

382.0

1-Butyl-3-methylbenzene

17.4

0.1

1.0

173.7

6-Methyl-cyclohex-3enecarboxylic acid butyl ester

16.1

2.5

5.7

209.1

CH3

O

717

2-Butyl Octanol OH

H3C

CH3

1006 n-Butyl Propionate

CH3

O

H3C O 109

N-Butyl Pyrrolidone O

CH3

N

764

n-Butyl Salicylate O CH3

O OH

223

Butyl Stearate O O

H3C H3C

724

3-n-Butyl Toluene CH3

CH3

747

Butyl-6-Methyl-3-Cyclohexene Carbolyate O H3C

O

CH3

7248_A001.fm Page 368 Wednesday, May 23, 2007 12:31 PM

368

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 725

Solvent Name n-Butylbenzene

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Butyl-benzene

17.4

0.1

1.1

157.0

4,5-Dimethyl[1,3]dioxolan-2-one

18.0

16.8

3.1

105.5

2-Ethyl-oxirane

16.3

6.2

5.9

87.5

1-Butyl-2-methylbenzene

17.6

0.1

1.0

171.3

1-Butyl-4-methylbenzene

17.4

0.1

1.0

174.2

But-2-ynedinitrile

15.2

16.2

8.0

78.4

Butyraldehyde

15.6

10.1

6.2

90.5

Butyramide

16.9

13.7

12.3

98.4

Butyric acid

14.9

4.1

10.6

110.0

CH3

110

2,3-Butylene Carbonate O

O H3C

111

O CH3

Butyleneoxide

H3C 719

O

o-n-Butyltoluene H3C

CH3

720

p-n-Butyltoluene CH3

CH3

112

2-Butynedinitrile

N

N

1060 Butyraldehyde*

H3C

O

1072 n-Butyramide

O NH2

H3C 114

Butyric Acid

O H3C

OH

7248_A001.fm Page 369 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

369

TABLE A.1 (CONTINUED) No. 767

Butyric Anhydride

Hydrogen Molar Bonding Volume

Butanoic anhydride

16.0

6.3

7.7

164.4

Dihydro-furan-2-one

19.0

16.6

7.4

76.8

Butyronitrile

15.3

12.4

5.1

87.3

Butyryl chloride

16.8

9.4

4.8

103.6

1,3,7-Trimethyl-3,7dihydro-purine-2,6dione

19.5

10.1

13.0

157.9

Oxepan-2-one

19.7

15.0

7.4

110.8

9H-Carbazole

21.7

6.4

6.2

152.0

Methanedione

15.7

6.3

5.7

38.0

CH3

O

H3C

Dispersion Polarity

O

O

115

Autonom/ ACD Name

Solvent Name

Gamma-Butyrolactone O O

116

Butyronitrile

H3C N 117

Butyrylchloride

O Cl

H3C 1200 Caffeine O H3C

CH3 N

N

N

N

O

CH3

118

Caprolactone (Epsilon) O O

869

Carbazole (Diphenylenimine)

N

119

Carbon Dioxide* Ro = 3.3

O O

7248_A001.fm Page 370 Wednesday, May 23, 2007 12:31 PM

370

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 120

Autonom/ ACD Name

Solvent Name Carbon Disulfid P for 0 Dipole Moment

Dispersion Polarity

Hydrogen Molar Bonding Volume

Methanedithione

20.5

0

0.6

60.0

Methanedithione

19.9

5.8

0.6

60.0

Tetrachloro-methane

17.8

0

0.6

97.1

Tetrachloro-methane

16.1

8.3

0

97.1

2-Oxo-malononitrile

15.0

6.3

8.0

71.2

Thioxo-methanone

17.4

3.7

0

51.0

15.9

4.6

12.0

436.0

15.1

3.7

8.1

298.7

S S 121

Carbon Disulfid P for Group Contrib ution

S S 122

Carbon tetrachloride P for 0 Dipole Moment

Cl

Cl

Cl

Cl 862

Carbon Tetrachloride P by Group Cont.

Cl

Cl

Cl

Cl 124

Carbonyl Cyanide

O N 123

N

Carbonyl Sulfid

S O 1224 Castor Oil

976

Cetyl Alcohol (1-Hexadecanol) H3C

Hexadecan-1-ol OH

7248_A001.fm Page 371 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

371

TABLE A.1 (CONTINUED) No. 125

Solvent Name Chloral

Cl Cl

Autonom/ ACD Name

Dispersion Polarity

Trichloro-acetaldehyde

17.2

7.4

Hydrogen Molar Bonding Volume 7.6

97.5

O Cl

126

Chlorine

Chlorine

17.3

10.0

0

46.0

128

Chloro Acetaldehyde

Chloro-acetaldehyde

16.2

16.1

9.0

60.4

Chloro-acetic acid

17.7

10.4

12.3

68.6

1-(4-Chloro-phenyl)ethanone

19.6

7.6

4.0

129.7

3-Chloro-prop-2-en-1-ol 17.2

10.3

16.5

78.6

2-Chloro-prop-2-en-1-ol 17.1

10.2

16.4

79.6

3-Chloro-furan-2,5dione

20.4

17.3

11.5

89.5

Acrylic acid chloromethyl ester

15.9

7.3

8.5

101.4

1-Chloro-pentane

16.0

6.9

1.9

120.8

Cl 129

O

Chloro Acetic Acid

O

Cl 797

OH

p-Chloro Acetophenone O CH3 Cl

130

3-Chloro Allyl Alcohol

OH

Cl 131

2-Chloro Allyl Alcohol

Cl OH

H2C 983

Chloro Maleic Anhydride

O

O

O Cl

132

1-Chloro Methyl Acrylate

O

H2C

Cl

O 955

1-Chloro Pentane

H3C

Cl

7248_A001.fm Page 372 Wednesday, May 23, 2007 12:31 PM

372

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 133

Solvent Name 2-Chloro Propene (Isopropen yl Chloride)

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Chloro-propene

15.5

6.7

2.2

84.9

1-Chloro-1-ethoxyethene

16.8

6.5

5.7

104.5

1-Chloro-1-fluoro ethene

16.0

6.9

4.0

67.6

1-Chloro-1-nitro-ethane

16.8

13.5

4.7

85.1

1-Chloro-1-nitropropane

16.8

13.0

4.5

102.2

3-Chloro-propan-1-ol

17.5

5.7

14.7

84.2

4-Chloro-buta-1,2-diene

16.6

8.0

6.7

89.5

1-Chloro-but-2-ene

16.2

7.7

2.0

97.4

1-Chloro-2-ethoxybenzene

19.2

8.1

4.4

138.7

Cl CH3

H2C 135

1-Chloro Vinyl Ethyl Ether

Cl H2C 127

O

CH3

1-Chloro-1-Fluoro Ethylene

Cl

CH2 F

136

1-Chloro-1-Nitroethane

Cl

CH3 +

N

O

O

1083 1-Chloro-1-Nitropropane

O

+

Cl

N

CH3

O 137

3-Chloro-1-Propanol

OH

Cl 138

4-Chloro-1,2-Butadiene

H2C Cl 139

1-Chloro-2-Butene

H3C 817

Cl

1-Chloro-2-Ethoxy Benzene Cl O

CH3

7248_A001.fm Page 373 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

373

TABLE A.1 (CONTINUED) No.

Solvent Name

1164 1-Chloro-2-Ethyl Benzene Cl

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Chloro-2-ethylbenzene

18.9

4.9

2.2

133.0

2-Chloro-2-methylpropane

15.6

7.6

2.0

110.0

3-Chloro-2-methylpropene

16.2

5.6

2.0

98.8

1-Chloro-2-methylpropene

16.1

7.1

4.2

95.6

1-Chloro-2-methyl-3nitro-benzene

20.3

9.6

3.8

132.0

4-Chloro-1-methyl-2nitro-benzene

19.9

11.8

3.8

132.0

1-Chloro-4-ethoxybenzene

19.3

6.3

4.4

139.2

2-Chloro-4-methylphenylamine

19.7

7.4

8.9

123.0

CH3

970

2-Chloro-2-Methyl Propane

CH3 H3C 140

Cl

CH3

3-Chloro-2-Methyl Propene CH3 H2C Cl

141

1-Chloro-2-Methyl Propene CH3 CH3 Cl

813

6-Chloro-2-Nitrotoluene CH3

819

O +

N

Cl

O

4-Chloro-2-Nitrotoluene CH3

O +

N

O

Cl

816

1-Chloro-4-Ethoxy Benzene Cl

O CH3

950

2-Chloro-4-Methylaniline NH2 Cl

CH3

7248_A001.fm Page 374 Wednesday, May 23, 2007 12:31 PM

374

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 977

Solvent Name 2-Chloro-5-Methyl Phenol OH

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Chloro-5-methylphenol

19.0

7.7

13.1

117.4

2-Chloro-acetamide

17.6

12.4

12.8

93.5

1-Chloro-propan-2-one

16.8

9.6

5.5

80.5

Chloro-acetonitrile

17.4

13.6

2.0

63.3

Chloro-acetyl chloride

17.5

9.2

5.5

79.5

Chloro-ethyne

16.2

2.1

2.5

63.7

Acetic acid 1-acetoxy-2- 16.7 chloro-allyl ester

7.3

8.8

160.2

3-Chloro-phenylamine

20.6

9.9

9.8

105.0

1-Chloro-4-methoxybenzene

19.6

7.8

6.7

122.5

Cl H3C

1067 2-Chloroacetamide

O Cl 142

NH2

Chloroacetone O H3C Cl

143

Chloroacetonitrile

N Cl 144

Chloroacetylchloride O

Cl Cl

145

Chloroacetylene

HC 766

Cl

2-Chloroallylidene Diacetate O

H3C Cl O

H2C

H3C

O

O

1012 m-Chloroaniline NH2

Cl

146

4-Chloroanisole

O

Cl

CH3

7248_A001.fm Page 375 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

375

TABLE A.1 (CONTINUED) No. 147

Solvent Name 4-Chlorobenzaldehyde

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

4-Chloro-benzaldehyde

19.9

7.2

5.6

113.4

Chloro-benzene

19.0

4.3

2.0

102.1

4-Chloro-benzonitrile

19.5

8.0

4.1

125.1

1-Chloro-4trichloromethylbenzene

20.3

5.5

3.5

153.8

4-Chloro-benzoyl chloride

19.9

6.7

5.1

127.1

(4-Chloro-phenyl)methanol

19.6

7.5

13.0

117.7

1-Chloro-3chloromethyl-benzene

19.9

9.3

2.6

117.7

1-Bromo-2-chloroethene

17.2

6.6

2.3

78.7

O

Cl 148

Chlorobenzene

Cl

811

4-Chlorobenzonitrile N

Cl

853

4-Chlorobenzotrichloride Cl

Cl

Cl

Cl

812

p-Chlorobenzoyl Chloride O Cl Cl

149

4-Chlorobenzyl Alcohol

OH Cl 150

3-Chlorobenzylchloride

Cl

151

Cl

1,2-Chlorobromoethylene

Cl

Br

7248_A001.fm Page 376 Wednesday, May 23, 2007 12:31 PM

376

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 152

Solvent Name 1-Chlorobutane

H3C 750

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Chloro-butane

16.2

5.5

2.0

104.5

2-Chloro-butane

15.8

7.6

2.0

106.8

2-Chloro-cyclohexanone 18.5

13.0

5.1

113.9

Cl

2-Chlorobutane

CH3

H3C

Cl 780

2-Chlorocyclohexanone O Cl

153

Chlorocyclopropane

Chloro-cyclopropane

17.6

7.2

2.2

84.9

Chloro-difluoro-methan

12.3

6.3

5.7

72.9

[Chloro(methyl)amino] methane

16.0

7.8

7.9

87.4

Acetic acid 2-chloroethyl ester

16.7

9.6

8.8

107.5

(2-Chloro-ethyl)benzene

19.3

6.3

2.2

131.5

1-Chloro-2-ethoxyethane

16.3

7.9

4.6

109.5

Cl

154

Chlorodifluoromethane (Freon 22

F

Cl F

155

N-Chlorodimethylamine

Cl H3C 772

N

CH3

2-Chloroethyl Acetate

O H3C

Cl

O

1165 2-Chloroethyl Benzene

Cl

756

2-Chloroethyl Ethyl Ether

H3C

O

Cl

7248_A001.fm Page 377 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

377

TABLE A.1 (CONTINUED) No. 834

2-Chloroethyl Ethyl Sulfid

Cl

S

H3C 808

Autonom/ ACD Name

Solvent Name

o-Chlorofluorobenzen F

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Chloro-2ethylsulfanyl-ethane

17.2

5.0

6.1

116.9

1-Chloro-2-fluoro benzene

19.4

8.7

2.0

104.9

Trichloro-methane

17.8

3.1

5.7

80.7

1-Chloro-hexane

16.1

6.2

1.7

138.1

Chloro-chloromethoxymethane

17.2

4.9

6.6

86.6

Chlorochloromethylsulfanylmethane

16.6

6.4

2.0

95.0

1-Chloro-naphthalene

19.9

4.9

2.5

136.2

1-Chloro-4-nitrobenzene

20.4

9.6

4.2

103.7

Chloro-nitro-methane

17.4

13.5

5.5

65.1

Cl

156

Chloroform

Cl

Cl Cl 959

1-Chlorohexane

Cl

H3C 157

Bis(Chloromethyl) Ether

Cl 158

O

Chloromethylsulfid

Cl 870

Cl

S

Cl

1-Chloronaphthalene Cl

809

p-Chloronitrobenzene O

+

N

O

Cl

159

Chloronitromethane

O

Cl

+

N

O

7248_A001.fm Page 378 Wednesday, May 23, 2007 12:31 PM

378

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 160

Solvent Name 2-Chlorophenol

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Chloro-phenol

20.3

5.5

13.9

102.3

Trichloro-nitro-methane

17.6

6.8

7.0

101.7

2-Chloro-buta-1,3-diene

16.1

5.4

2.1

93.2

2-Chloro-propenal

17.1

12.9

8.1

75.5

1-Chloro-propene

15.3

6.9

2.2

82.3

2-Chloro-acrylic acid

19.1

9.4

12.4

86.6

3-Chloropropionaldehyde

17.0

13.3

8.2

73.0

3-Chloro-propionitrile

17.3

15.9

6.1

77.4

3-Chloro-propyne

16.7

7.4

2.3

72.4

OH Cl

1081 Chloropicrin (Trichloronitromethane)

O

Cl

+

Cl

N

Cl

O 161

Chlorprene

Cl CH2

H2C 162

2-Chloropropenal

Cl O

H2C 163

1-Chloropropene

Cl

H3C 164

2-Chloropropenoic Acid Cl OH

H2C O

165

3-Chloropropionaldehyde

Cl 166

O

3-Chloropropionitrile

Cl N 167

3-Chloropropyne

HC Cl

7248_A001.fm Page 379 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

379

TABLE A.1 (CONTINUED) No. 711

Solvent Name p-Chlorostyrene CH2

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Chloro-4-vinylbenzene

18.7

4.3

3.9

128.3

1-Chloro-2-vinylbenzene

18.7

4.7

3.9

126.8

2-Chloro-thiophene

19.2

6.2

8.0

92.2

4-Chloro-benzenethiol

20.8

8.6

10.6

100.0

2-Chloro-benzenethiol

20.2

7.0

10.0

113.4

1-Chloro-4-methylbenzene

19.1

6.2

2.6

118.3

1-Chloro-1,2,2-trifluoro ethene

15.3

6.3

0

(E)-3-Phenyl-propenal

19.4

12.4

Cl

712

o-Chlorostyrene CH2 Cl

989

2-Chlorothiophene

Cl

S 168

4-Chlorothiophenol SH

Cl

787

o-Chlorothiophenol SH Cl

814

p-Chlorotoluene CH3

Cl

134

Chlorotrifluoroet ylene (CTFE)

Cl

F

F

F

1111 trans-Cinnamaldehyde O

6.2

75.6

125.9

7248_A001.fm Page 380 Wednesday, May 23, 2007 12:31 PM

380

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Solvent Name

1110 cis-Cinnamic Acid

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

19.1

3.9

10.6

115.4

(E)-3-Phenyl-prop-2-en- 19.1 1-ol

6.0

13.0

129.1

1,3,3-Trimethyl-2-oxabicyclo[2.2.2]octane

16.7

4.6

3.4

167.5

3-Methyl-furan-2,5dione

19.2

17.0

11.2

89.9

4-((E)-3-Hydroxypropenyl)-2-methoxyphenol

19.0

7.0

16.3

171.6

Chromen-2-one

20.0

12.5

6.7

156.3

4-((E)-3-Hydroxypropenyl)-phenol

19.1

7.0

17.3

136.5

3-Methyl-phenol

18.0

5.1

12.9

104.7

(Z)-3-Phenyl-acrylic acid

OH O

1107 Cinnamyl Alcohol OH

1135 Cineol (Eucalyptol) H3C O H3C

CH3

H

982

Citraconic Anhydride O

O

O CH3

1102 Coniferyl Alcohol OH

H3C

O HO

974

Coumarin

O 1103 p-Coumaryl Alcohol OH

O

HO

169

m-Cresol CH3

OH

7248_A001.fm Page 381 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

381

TABLE A.1 (CONTINUED) No. 170

Solvent Name Crotonaldehyde

Dispersion Polarity

Hydrogen Molar Bonding Volume

(E)-But-2-enal

16.2

14.9

7.4

82.5

(E)-But-2-enoic acid

16.8

8.7

12.0

84.6

5H-Furan-2-one

19.0

19.8

9.6

76.4

(E)-But-2-enenitrile

16.4

18.8

5.5

81.4

Cyanamide

15.5

27.6

16.8

32.8

Oxalonitrile

15.1

11.8

0

54.6

Cyanic bromide

18.3

15.2

0

52.6

Cyanic chloride

15.6

14.5

0

51.8

Cyclobutanone

18.3

11.4

5.2

73.4

Cyclodecanone

16.8

8.0

4.1

161.0

O

H3C 171

Autonom/ ACD Name

Crotonic Acid

O OH

H3C 172

Crotonlactone O O

173

Trans-Crotononitrile

N H3C 174

Cyanamid (Carbamonitrile)

N 175

NH2

Cyanogen

N 176

Cyanogen Bromide

N 177

Br

Cyanogen Chloride

N 178

N

Cl

Cyclobutanone

O

179

Cyclodecanone

O

7248_A001.fm Page 382 Wednesday, May 23, 2007 12:31 PM

382

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 180

Autonom/ ACD Name

Solvent Name Cycloheptanone

Dispersion Polarity

Hydrogen Molar Bonding Volume

Cycloheptanone

17.2

10.6

4.8

118.2

Cyclohexane

16.8

0

0.2

108.7

Cyclohexane-1,2dicarboxylic acid bis(7-methyl-octyl) ester

16.4

2.2

5.0

422.4

Cyclohexane-1,2-diol

17.4

9.8

18.3

112.8

Cyclohexane-1,2-dione

18.6

10.3

8.0

103.8

Cyclohexanol

17.4

4.1

13.5

106.0

Cyclohexanone

17.8

6.3

5.1

104.0

Cyclohexene

17.2

1.0

5.0

101.9

O

181

Cyclohexane

1214 Cyclohexane-1,2-Dicarboxylic Acid Di-(Isononyl) Ether H3C

CH3 H3C

CH3

O O O

O

1093 1,2-Cyclohexanediol

OH OH 1094 1,2-Cyclohexanedione

O O 182

Cyclohexanol

OH

183

Cyclohexanone

O

871

Cyclohexene

7248_A001.fm Page 383 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

383

TABLE A.1 (CONTINUED) No.

Solvent Name

1199 Cycloheximide H3C CH3

O H

OH H

O

N

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

4-[(R)-2-((1S,3S,5S)3,5-Dimethyl-2-oxocyclohexyl)-2hydroxy-ethyl]piperidine-2,6-dione

18.3

11.0

13.8

171.0

O

872

Cyclohexyl Benzene

Cyclohexyl-benzene

18.7

0

1.0

169.9

969

N-Cyclohexyl-2-Pyrrolidone

1-Cyclohexylpyrrolidin-2-one

18.2

6.8

6.5

163.0

Cyclohexylamine

17.2

3.1

6.5

113.8

Chloro-cyclohexane

17.3

5.5

2.0

118.6

Cyclooctanone

17.0

9.6

4.5

131.7

N

184

O

Cyclohexylamine

NH2

185

Cyclohexylchloride

Cl

186

Cyclooctanone O

920

Cyclopentadiene

Cyclopenta-1,3-diene

17.2

1.9

6.1

82.1

187

Cyclopentane

Cyclopentane

16.4

0

1.8

94.9

7248_A001.fm Page 384 Wednesday, May 23, 2007 12:31 PM

384

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 188

Solvent Name Cyclopentanone

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Cyclopentanone

17.9

11.9

5.2

89.1

O

189

Cyclopentene

Cyclopentene

16.7

3.8

1.7

89.0

735

2-Cyclopentenyl Alcohol

Cyclopent-2-enol

18.1

7.6

15.6

86.2

OH 190

Cyclopropene

Cyclopropene

17.2

2.4

2.0

50.0

191

Cyclopropane

Cyclopropane

17.6

0

0

58.3

192

Cyclopropylmethylketone

1-Cyclopropyl-ethanone 17.0

11.1

4.6

93.6

Cyclopropanecarbonitrile

18.6

16.2

5.7

75.4

(1R,4R)-1,7,717.8 Trimethylbicyclo[2.2.1]heptan-2one

9.4

4.7

153.7

CH3 O 193

Cyclopropylnitrile N

1223 d-Camphor H3C

CH3 H

CH3

O

7248_A001.fm Page 385 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

385

TABLE A.1 (CONTINUED) No. 873

Autonom/ ACD Name

Solvent Name DDT Cl Cl

Cl

Cl

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Chloro-4-[2,2,2trichloro-1-(4chlorophenyl) ethyl]benzene

20.0

5.5

3.1

268.8

Cl

194

Cis-Decahydronaphthalene

Decahydro-naphthalene

18.8

0

0

156.9

195

Trans-Decahydronaphthalene

Decahydro-naphthalene

18.0

0

0

156.9

196

Decane

Decane

15.7

0

0

195.9

Decan-1-ol

16.0

4.7

10.0

191.8

Decan-2-ol

15.8

3.9

10.0

192.8

Dec-1-ene

15.8

1.0

2.2

190.6

1-Chloro-2-(2-chloro17.1 ethoxymethoxy)-ethane

10.2

7.1

141.1

CH3

H3C

197

1-Decanol* OH

H3C

734

2-Decanol CH3

H3C

OH

726

1-Decene CH2

H3C

759

Di-(2-Chloroethoxy) Methane

O

Cl 773

O

Cl

Dibutyl Fumarate O H3C

16.7

3.0

6.7

232.7

1-Methyl-4-[(4methylphenyl)sulfi yl] benzene

20.3

11.4

3.1

209.0

CH3

O

O

(E)-But-2-enedioic acid dibutyl ester O

199

Di-p-Tolyl Sulfoxide CH3

H3C S O

7248_A001.fm Page 386 Wednesday, May 23, 2007 12:31 PM

386

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 204

Di-(2-Chloro-Isopropyl) Ether

CH3

CH3

Cl 235

Autonom/ ACD Name

Solvent Name

Di-(2-Chloroethyl) Ether

Cl

O

1139 Di-(2-Ethyl Hexyl)azelate O

O H3C

O

O

CH3

O H3C

O

O

O

CH3

N

Di-2-Ethyl Hexyl Ether CH3

O

1-Chloro-2-(2-chloroethoxy)-ethane

18.8

9.0

5.7

117.2

Nonanedioic acid bis(2-ethyl-hexyl) ester

16.7

1.4

4.8

449.9

Decanedioic acid bis(2-ethyl-hexyl) ester

16.8

1.0

4.7

468.7

1-Methoxy-2(2-methoxy-ethoxy)ethane

15.7

6.1

6.5

142.0

Bis-(2-ethyl-hexyl)amine

15.6

0.8

3.2

301.5

3-(2-Ethylhexyloxymethyl)heptane

15.9

2.6

2.5

300.5

2,6-Dimethyl-heptan4-ol

14.9

3.1

10.8

177.8

2,6-Dimethyl-heptan4-one

16.0

3.7

4.1

177.1

2-Methyl-1-(2-methylpropane-1-sulfi yl)propane

16.3

10.5

6.1

195.0

CH3

Di-Isobutyl Carbinol

CH3

OH

CH3 CH3

H3C Di-Isobutyl Ketone

CH3

O

CH3 CH3

H3C Di-Isobutyl Sulfoxide

CH3 H3C

146.0

CH3

H3C

845

CH3

Di-2-Ethyl Hexyl Amine

H3C

203

O

O

H3C

208

5.1

CH3

Di-(2-Methoxyethyl) Ether

H3C

736

8.2

O

H3C

205

19.0

CH3

1138 Di-(2-Ethyl Hexyl)sebacate

H3C

1-Chloro-2-(2-chloro-1methyl-ethoxy)propane

CH3

H3C

202

Hydrogen Molar Bonding Volume

Cl

O

Cl

Dispersion Polarity

O S

CH3 CH3

7248_A001.fm Page 387 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

387

TABLE A.1 (CONTINUED) No. 279

Autonom/ ACD Name

Solvent Name Di-Isodecyl Phthalate O

H3C

CH3

H3C

CH3

Dispersion Polarity

Hydrogen Molar Bonding Volume

Phthalic acid bis-(8methyl-nonyl) ester

16.6

6.2

2.6

464.2

Phthalic acid bis-(5methyl-hexyl) ester

16.8

7.2

3.4

364.9

Hexanedioic acid bis-(7- 16.2 methyl-octyl) ester

1.8

4.9

433.7

Phthalic acid bis-(7methyl-octyl) ester

16.6

6.6

2.9

432.4

Methyl-phosphonic acid diisopropyl ester

16.4

10.0

5.7

184.4

Phosphorofluoridic acid diisopropyl ester

15.7

10.2

5.9

174.5

2-(Propane-2-sulfi yl)propane

17.0

11.5

7.4

159.9

1-Butoxy-butane

15.2

3.4

4.2

170.3

1-(Butane-1-sulfi yl)butane

16.4

10.5

6.1

195.0

O O O

280

Di-Isoheptyl Phthalate CH3

O O

CH3

O

CH3 CH3

O

281

Di-Isononyl Adipate O

H3C

O CH3

282

O

CH3

O

Di-Isononyl Phthalate CH3

O

CH3

O O

CH3 CH3

O

840

CH3

Di-Isopropyl Methyl Phosphonate CH3 O

H3C O H3C

842

CH3

P

CH3

O

Di-Isopropyl Phosphonofluoridat CH3 O

H3C

O P H3C

198

CH3

F

O

Di-Isopropyl Sulfoxide CH3 H3C

CH3

S

CH3

O

219

Di-n-Butyl Ether

H3C 200

O

CH3

Di-n-Butyl Sulfoxide

H3C

S O

CH3

7248_A001.fm Page 388 Wednesday, May 23, 2007 12:31 PM

388

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 957

Di-n-Pentyl Ether O

H3C

956

O

Hydrogen Molar Bonding Volume

1-Pentyloxy-pentane

15.6

3.1

3.0

203.2

1-Propoxy-propane

15.1

4.2

3.7

137.7

1-(Propane-1-sulfi yl)propane

17.0

13.0

7.4

159.9

4-Hydroxy-4-methylpentan-2-one

15.8

8.2

10.8

124.2

Diallyl-amine

15.6

4.5

6.7

124.1

3-Allyloxy-propene

15.3

4.4

5.3

118.8

3-(1-Allyloxy-ethoxy)propene

15.4

5.1

4.8

163.3

Diazomethane

14.7

6.1

11.3

78.1

Pentanedioic acid dimethyl ester

16.2

4.7

8.4

159.0

CH3

Di-n-Propyl Sulfoxide

H3C

Dispersion Polarity

CH3

Di-n-Propyl Ether

H3C 201

Autonom/ ACD Name

Solvent Name

CH3

S O

209

Diacetone Alcohol

H3C H3C

CH3

OH 210

O

Diallyl Amine

N

H2C 211

Diallyl Ether

O

H2C 749

CH2

CH2

1,1-Diallyloxyethane

O

H2C

O

CH2

CH3 212

Diazomethane

H2C

+

N

N

1009 Dibasic Esters (dupont) Mix of Dimethyl Esters of Succinic, Glutaric, and Adipic Acids

O H3C

O

O O

CH3

7248_A001.fm Page 389 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

389

TABLE A.1 (CONTINUED) No. 213

Autonom/ ACD Name

Solvent Name Dibenzyl Ether*

Dispersion Polarity

Hydrogen Molar Bonding Volume

[(Benzyloxy)methyl]benzene

19.6

3.4

5.2

197.4

Decanedioic acid dibenzyl ester

17.8

2.2

5.5

362.1

1,1-Dibromo-ethene

15.1

4.8

7.0

85.3

Dibromo-methane

17.8

6.4

7.0

69.8

1,2-Dibromo-benzene

20.7

6.5

5.3

120.0

1,1-Dibromo-ethane

18.5

8.4

8.8

91.4

1,2-Dibromo-ethene

18.0

4.9

3.0

82.7

2,3-Dibromo-buta-1,3diene

17.7

11.8

6.4

103.4

Dibutyl-amine

15.0

3.0

4.3

170.0

O

1148 Dibenzyl Sebacate

O O

O O

214

1,1-Dibromo Ethylene

Br

Br

CH2 215

Dibromo Methane

Br 804

Br

o-Dibromobenzene Br Br

216

1,1-Dibromoethane

Br

Br

CH3 217

1,2-Dibromoethylene

Br 218

Br

2,3-Dibromoprene Br

CH2

H2C Br

1015 Dibutyl Amine

H3C

N

CH3

7248_A001.fm Page 390 Wednesday, May 23, 2007 12:31 PM

390

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 220

Autonom/ ACD Name

Solvent Name N,N-Dibutyl Formamide

H3C

N

Dispersion Polarity

Hydrogen Molar Bonding Volume

N,N-Dibutyl-formamide 15.5

8.9

6.2

182.0

Nonan-5-one

16.0

7.7

4.4

173.4

(Z)-But-2-enedioic acid dibutyl ester

16.5

6.1

7.2

230.4

Phthalic acid dib utyl ester

17.8

8.6

4.1

266.0

Decanedioic acid dibutyl 16.7 ester

4.5

4.1

339.0

1,2-Dichloro-4trifluoromet ylbenzene

20.0

4.7

2.4

145.5

1,3-Dichloro-propane

18.6

8.2

3.0

95.1

1,1-Dichloro-1-nitroethane

16.8

12.1

4.3

125.2

CH3

O 1068 Dibutyl Ketone

CH3

H3C O 1146 Dibutyl Maleate O O

CH3

O

CH3

O

221

Dibutyl Phthalate O O

CH3

O

CH3

O

222

Dibutyl Sebacate* O O

O

H3C

CH3

O

700

3,4-Dichloro α,α,α-Trifluorotoluen F

F

F

Cl Cl

874

1,3-Dichloropropane

Cl

Cl

1082 1,1-Dichloro-1-Nitroethane

O

+

N O

Cl

Cl CH3

7248_A001.fm Page 391 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

391

TABLE A.1 (CONTINUED) No. 229

Solvent Name 2,3-Dichloro-1,3-Butadiene Cl

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2,3-Dichloro-buta-1,3diene

17.1

2.3

2.8

104.0

1,3-Dichloro-but-2-ene

16.9

7.8

2.7

107.7

1,4-Dichloro-but-2-ene

17.8

7.6

2.0

97.8

1,3-Dichloro-2-fluoro benzene

19.4

9.1

2.7

117.1

1,3-Dichloro-propan-2ol

17.5

9.9

14.6

95.5

1,3-Dichloro-2-fluoro-4 nitro-benzene

19.9

7.2

4.2

140.0

1,5-Dichloro-2-nitro-4trifluoromet ylbenzene

19.9

7.6

3.7

162.0

Dichloro-acetaldehyde

16.7

9.1

7.5

94.0

CH2

H2C Cl

206

1,3-Dichloro-2-butene

Cl Cl

H3C 207

1,4-Dichloro-2-Butene

Cl

Cl 825

1,3-Dichloro-2-Fluorobenzene Cl F Cl

739

1,3-Dichloro-2-Propanol

OH Cl 824

Cl

2,4-Dichloro-3-Fluoronitrobenzene O

+

N

O Cl F

Cl

828

2,4-Dichloro-5-Nitrobenzotrifluorid F

Cl

F

Cl

F

+

N

O

O

230

Dichloroacetaldehyde

Cl

O Cl

7248_A001.fm Page 392 Wednesday, May 23, 2007 12:31 PM

392

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 945

Solvent Name Dichloroacetic Acid

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Dichloro-acetic acid

18.2

8.1

12.2

82.5

1,1-Dichloro-propan-2one

17.1

7.6

5.4

97.3

Dichloro-acetonitrile

17.4

9.4

6.4

80.3

2,4-Dichlorophenylamine

20.9

6.2

10.0

103.4

1,3-Dichloro-2methoxy-benzene

19.8

8.4

6.5

137.1

2,4-Dichlorobenzaldehyde

19.7

8.8

5.4

135.7

1,2-Dichloro-benzene

19.2

6.3

3.3

112.8

1,4-Dichloro-benzene

19.7

5.6

2.7

118.6

O Cl

OH Cl

231

1,1-Dichloroacetone O Cl

CH3 Cl

232

Dichloroacetonitrile

Cl N Cl 1000 2,4-Dichloroaniline NH2 Cl

Cl

233

2,6-Dichloroanisole O

CH3 Cl

Cl

846

2,4-Dichlorobenzaldehyde O Cl

Cl

234

o-Dichlorobenzene Cl Cl

715

p-Dichlorobenzene Cl

Cl

7248_A001.fm Page 393 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

393

TABLE A.1 (CONTINUED) No. 716

Autonom/ ACD Name

Solvent Name m-Dichlorobenzene

Dispersion Polarity

Hydrogen Molar Bonding Volume

1,3-Dichloro-benzene

19.7

5.1

2.7

114.8

1,4-Dichloro-2trifluoromet ylbenzene

20.0

4.7

2.4

145.5

1,4-Dichloro-butane

18.3

7.7

2.8

109.5

Dichloro-difluoro methane

12.3

2.0

0

92.3

1,1-Dichloro-ethane

16.5

7.8

3.0

84.2

Bis-chloromethyl-amine 16.8

7.6

8.0

90.8

1,1-Dichloro-ethene

16.4

5.2

2.4

79.9

(Z)-1,2-Dichloro-ethene

17.0

8.0

3.2

75.5

Bis-(2-chloro-ethyl)amine

16.8

7.6

7.7

98.3

Cl

Cl

826

2,5-Dichlorobenzotrifluorid F

F

F Cl

Cl

1019 1,4-Dichlorobutane

Cl

Cl 236

Dichlorodifluoromethane (Freon 12

Cl

F

Cl

F 237

1,1-Dichloroethane

Cl

Cl CH3

225

N,N-Dichloromethyl Amine

Cl 239

N

Cl

1,1-Dichloroethylene

Cl Cl 224

1,2-Dichloroethylene (cis)

Cl 238

CH2

Cl

N,N-Dichloroethyl Amine

Cl

N

Cl

7248_A001.fm Page 394 Wednesday, May 23, 2007 12:31 PM

394

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 241

Autonom/ ACD Name

Solvent Name Dichloromethyl Methyl Ether

Cl

O

H3C

Dispersion Polarity

Hydrogen Molar Bonding Volume

Dichloro-methoxymethane

17.1

12.9

6.5

90.5

Dichloro-fluoro-methan

15.8

3.1

5.7

75.4

1,2-Dichloro-3-nitrobenzene

19.7

12.6

4.4

132.5

1,2-Dichloro-4-nitrobenzene

20.1

7.2

4.1

130.6

2,4-Dichloro-1-nitrobenzene

20.4

8.7

4.2

133.4

1,5-Dichloro-pentane

19.0

7.8

1.5

127.5

2,5-Dichloro-phenol

20.0

6.3

12.1

119.0

2,6-Dichloro-phenol

20.1

7.5

10.9

119.0

Cl 242

Dichloromonofluoromethane (Freon 21

Cl

Cl F 243

2,3-Dichloronitrobenzene O

+

N

O Cl Cl

801

3,4-Dichloronitrobenzene O

+

N

O

Cl Cl

822

2,4-Dichloronitrobenzene O

+

N

O Cl

Cl

833

1,5-Dichloropentane

Cl

Cl 244

2,5-Dichlorophenol OH Cl Cl

852

2,6-Dichlorophenol OH Cl

Cl

7248_A001.fm Page 395 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

395

TABLE A.1 (CONTINUED) No.

Solvent Name

1159 2,4-Dichlorophenol

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2,4-Dichloro-phenol

20.0

7.2

13.1

119.0

(3,4-Dichloro-phenyl)acetonitrile

20.5

10.8

4.4

148.8

1,1-Dichloro-propane

16.1

7.8

3.5

98.3

2,3-Dichloro-propan1-ol

17.5

9.2

14.6

95.2

2,3-Dichloro-propene

16.2

7.8

3.0

91.6

1,1-Dichloro-propene

16.9

6.7

2.9

93.5

(Z)-1,2-Dichloropropene

17.0

8.5

2.9

93.9

1,2-Dichloro-1,1,2,2tetrafluoro-ethan

12.6

1.8

0

OH Cl

Cl

848

3,4-Dichlorophenyl Acetonitrile N

Cl Cl

245

1,1-Dichloropropane

Cl

H3C Cl 738

2,3-Dichloropropanol

Cl OH

Cl 226

2,3-Dichloropropene

Cl Cl 246

CH2

1,1-Dichloropropene

Cl

H3C

Cl 227

1,2-Dichloropropene (cis)

Cl Cl 247

CH3

1,2-Dichlorotetrafluoroethane (Freon 114)

Cl

Cl

F

F F

F

117.6

7248_A001.fm Page 396 Wednesday, May 23, 2007 12:31 PM

396

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 248

Autonom/ ACD Name

Solvent Name 3,4-Dichlorotoluene CH3

Dispersion Polarity

Hydrogen Molar Bonding Volume

1,2-Dichloro-4-methylbenzene

19.8

9.8

2.5

128.7

1,2-Dichloro-1-ethoxyethene

16.9

10.5

6.0

117.8

2-(2-Hydroxyethylamino)-ethanol

17.2

10.8

21.2

95.9

1,1-Diethoxy-butane

15.4

4.9

4.6

177.6

[(Ethoxydithio)oxy]etha ne

15.1

8.3

7.4

141.5

1,1-Diethoxy-ethane

15.0

3.4

4.0

143.9

1,1-Diethoxy-ethane

15.2

5.4

5.3

142.2

2,5-Diethoxytetrahydro-furan

16.6

6.4

7.3

165.7

N,N-Diethyl-acetamide

16.4

11.3

7.5

124.5

Cl Cl

228

1,2-Dichlorovinyl Ethyl Ether

Cl

CH3

O

Cl 249

Diethanolamine

N

HO 777

OH

1,1-Diethoxy Butane

CH3

O

H3C 792

O

Diethoxy Disulfid

H3C 250

CH3

S

O

S

O

CH3

1,1-Diethoxy Ethane

CH3

791

CH3

O

O

H3C

1,1-Diethoxy Ethane (Acetal)

CH3 O

H3C

O

CH3

1045 2,5-Diethoxy Tetrahydrofuran

251

O

O

H3C

N,N-Diethyl Acetamide O H3C

N

CH3 CH3

O

CH3

7248_A001.fm Page 397 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

397

TABLE A.1 (CONTINUED) No. 252

Autonom/ ACD Name

Solvent Name Diethyl Amine

H3C

Dispersion Polarity

Hydrogen Molar Bonding Volume

Diethyl-amine

14.9

2.3

6.1

103.2

Acetate diethylammonium;

16.0

20.3

18.4

1,4-Diethyl-benzene

18.0

0

0.6

156.9

1,2-Diethyl-benzene

17.7

0.1

1.0

153.5

Carbonic acid dieth yl ester

15.1

6.3

3.5

121.0

Acetate diethyl-(2hydroxy-ethyl)ammonium;

16.0

20.3

18.4

Ethoxy-ethane

14.5

2.9

5.1

104.8

N,N-Diethyl-formamide

16.4

11.4

9.2

111.4

CH3

N

1183 Diethyl Amine/Acetic Acid O

+

NH2

O

253

p-Diethyl Benzene CH3

H3C

721

1,2-Diethyl Benzene CH3 CH3

254

Diethyl Carbonate*

O O

H3C

O

CH3

1184 Diethyl Ethanolamine/Acetic Acid H + N

O

OH

255

Diethyl Ether

H3C 256

O

O

CH3

N,N-Diethyl Formamide O N

CH3

CH3

7248_A001.fm Page 398 Wednesday, May 23, 2007 12:31 PM

398

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 257

Autonom/ ACD Name

Solvent Name Diethyl Ketone

H3C

Dispersion Polarity

Hydrogen Molar Bonding Volume

Pentan-3-one

15.8

7.6

4.7

106.4

Malonic acid dieth yl ester

16.1

7.7

8.3

152.5

Oxalic acid diethyl ester

16.2

8.0

8.8

136.6

Phthalic acid dieth yl ester

17.6

9.6

4.5

198.0

Sulfuric acid dieth yl ester

15.7

14.7

7.1

131.5

Ethylsulfanyl-ethane

16.8

3.1

2.0

107.4

2-Diethylamino-ethanol

14.9

5.8

12.0

133.2

Ethyldisulfanyl-ethane

16.7

6.7

5.7

123.1

2-(2-Hydroxy-ethoxy)ethanol

16.6

12.0

20.7

94.9

CH3 O

979

Diethyl Malonate

O

O

CH3

O

O

H3C

1143 Diethyl Oxalate O O

H3C

O

CH3

O

258

Diethyl Phthalate O O

CH3

O

CH3

O

259

Diethyl Sulfate

O

H3C 260

S O

CH3

O

Diethyl Sulfid

H3C 261

O

CH3

S

2-(Diethylamino) Ethanol

H3C

OH

N

H3C 262

Diethyldisulfid

H3C 263

S

S

CH3

Diethylene Glycol Commercial

HO

O

OH

7248_A001.fm Page 399 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

399

TABLE A.1 (CONTINUED) No. 264

Autonom/ ACD Name

Solvent Name

Dispersion Polarity

Diethylene Glycol Butyl Ether Acetate Acetic acid 2-(2-b utoxy- 16.0 Commercial ethoxy)-ethyl ester H3C

O

O

O

Hydrogen Molar Bonding Volume

4.1

8.2

208.2

1-[2-(2-Butoxy-ethoxy)- 15.8 ethoxy]-butane

4.7

4.4

248.1

1-Ethoxy-2-(2-ethoxyethoxy)-ethane

15.8

5.9

5.6

179.8

[2-(2-Vinyloxy-ethoxy)ethoxy]-ethene

16.0

7.3

7.9

164.1

2-(2-Hexyloxy-ethoxy)ethanol

16.0

6.0

10.0

204.3

2-[2-(2-Methoxyethoxy)-ethoxy]-2methyl-propane

16.0

7.2

7.2

193.9

2-(2-Butoxy-ethoxy)ethanol

16.0

7.0

10.6

170.6

2-(2-Ethoxy-ethoxy)ethanol

16.1

9.2

12.2

130.9

Acetic acid 2-(2-ethoxy- 16.2 ethoxy)-ethyl ester

5.1

9.2

175.5

2-(2-Methoxy-ethoxy)ethanol

16.2

7.8

12.6

118.0

2-(2-Propoxy-ethoxy)ethanol

16.0

7.2

11.3

153.9

CH3 O

757

Diethylene Glycol Dib utyl Ether H3C

762

O

O

O

CH3

O

O

CH2

Diethylene Glycol He xyl Ether H3C

266

O

O

Diethylene Glycol Di vinyl Ether H2C

265

O

CH3

Diethylene Glycol Dieth yl Ether H3C

755

O

O

OH

O

Diethylene Glycol Meth yl t-Butyl Ether commerical H3C

O

O

O

CH3 CH3

CH3

267

Diethylene Glycol Monob utyl Ether Commercial H3C

268

O

Diethylene Glycol Monoeth yl Ether Commerical

H3C 269

OH

O

O

OH

O

Diethylene Glycol Monoeth yl Ether Acetate Commerical H3C

O

O

O

CH3 O

270

Diethylene Glycol Monomethyl Ether Commerical

H3C 271

O

OH

O

Diethylene Glycol Monoprop yl Ether H3C

O

O

OH

7248_A001.fm Page 400 Wednesday, May 23, 2007 12:31 PM

400

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 272

Diethylenetriamine Commercial

H2N 810

Autonom/ ACD Name

Solvent Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

N*1*-(2-Amino-ethyl)ethane-1,2-diamine

16.7

13.3

14.3

108.0

1,2-Difluoro-benzen

18.0

9.0

1.0

98.5

2,6-Difluoro benzonitrile

18.8

11.2

3.2

111.6

3,5-Difluoro benzonitrile

18.8

11.2

3.2

111.6

1,1-Difluoro-ethan

14.9

10.2

3.0

69.5

1,1-Difluoro-ethen

15.0

6.8

3.6

58.2

2,4-Difluoro-1-nitro benzene

19.4

11.0

3.7

109.6

1-Hexyloxy-hexane

16.0

3.0

2.8

235.8

Phthalic acid dihe xyl ester

17.0

7.6

3.6

332.3

NH2

N

o-Difluorobenzen F F

829

2,6-Difluorobenzonitril N F

F

830

3,5-Difluorobenzonitril N

F

273

F

1,1-Difluoroethan

F

F

CH3 274

1,1-Difluoroet ylene

F

F

CH2 823

2,4-Difluoronitrobenzen O

+

N

O F

F

1069 Dihexyl Ether

275

CH3

O

H3C

Dihexyl Phthalate O

O

O

CH3

O

CH3

7248_A001.fm Page 401 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

401

TABLE A.1 (CONTINUED) No. 276

Dispersion Polarity

Hydrogen Molar Bonding Volume

17.3

6.3

10.7

49.6

3,4-Dihydro-2H-pyran

17.5

5.5

5.7

91.2

Benzene-1,4-diol

21.0

10.2

27.2

82.7

Benzene-1,2-diol

20.0

11.3

21.8

81.9

Hexanedioic acid diisobutyl ester

16.7

2.5

6.2

269.6

Diisopropyl-amine

14.8

1.7

3.5

141.1

4-Methylene-oxetan-2one

16.2

15.1

8.1

75.9

1,3-Dimethoxy-butane

15.6

5.5

5.2

140.0

Dihydrogen Disulfid

HS 277

Autonom/ ACD Name

Solvent Name

SH

Dihydropyran

O 917

1,4-Dihydroxybenzene (1,4Benzenediol) OH

OH

904

1,2-Dihydroxybenzene (Catechol)

OH OH 1140 Diisobutyl Adipate O H3C

CH3 O

O CH3

1084 Diisopropylamine

H3C

CH3

N CH3

283

CH3

Diketene O O

H2C

763

1,3-Dimethoxy Butane

O

H3C

CH3

O

CH3

O

CH3

7248_A001.fm Page 402 Wednesday, May 23, 2007 12:31 PM

402

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 284

Autonom/ ACD Name

Solvent Name 1,1-Dimethoxy Ethane

O

H3C

O

Dispersion Polarity

Hydrogen Molar Bonding Volume

1,1-Dimethoxy-ethane

15.1

4.9

4.9

106.7

2,6-Dimethoxy-phenol

19.3

7.6

13.7

136.4

1,2-Dimethoxy-benzene

19.2

4.4

9.4

127.7

2,5-Dimethoxytetrahydro-furan

16.8

7.2

8.2

129.6

N,N-Dimethylacetamide

16.8

11.5

10.2

92.5

But-2-yne

15.1

3.4

7.6

78.9

Dimethyl-amine

15.3

4.8

11.2

66.2

Dimethyl-amine; compound with dimethyl-amine

15.3

4.8

7.9

132.4

N,N-Dimethylbutyramide

16.4

10.6

7.4

127.8

CH3

CH3 1108 2,6-Dimethoxy Phenol OH H3C

889

O

O

1,2-Dimethoxybenzene

O

CH3 CH3

O 988

2,5-Dimethoxytetrahydrofuran

H3C 285

CH3

O

O

O

N,N-Dimethyl Acetamide O N

H3C

CH3

CH3

CH3

286

Dimethyl Acetylene

H3C 287

Dimethyl Amine

H3C 288

CH3

N

CH3

Dimethyl Amine-Dimer

H N

H N 289

N,N-Dimethyl Butyramide O

H3C

N

CH3

CH3

7248_A001.fm Page 403 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

403

TABLE A.1 (CONTINUED) No. 290

Autonom/ ACD Name

Solvent Name Dimethyl Carbonate

O

H3C

O

Dispersion Polarity

Hydrogen Molar Bonding Volume

Carbonic acid dimeth yl ester

15.5

3.9

9.7

84.2

1,2-Dimethoxy-ethane

15.4

6.0

6.0

104.5

1-Methoxy-2-(2methoxy-ethoxy)ethane

15.8

6.1

9.2

142.0

Butane-2,3-dione

15.7

5.3

11.7

88.2

Methyldisulfanylmethane

17.3

7.8

6.5

88.6

2-Dimethylaminoethanol

16.1

9.2

15.3

101.1

Acetate (2-hydroxyethyl)-dimethylammonium;

16.8

19.8

19.8

Formate (2-hydroxyethyl)-dimethylammonium;

17.2

21.5

22.5

2-Methyl-acrylate (2hydroxy-ethyl)dimethyl-ammonium;

17.2

18.8

17.6

CH3

O 291

Dimethyl Cellosolve

O

H3C 292

CH3

Dimethyl Diethylene Glycol

O

H3C 293

O

O

O

CH3

Dimethyl Diketone O CH3

H3C O

294

Dimethyl Disulfid

H3C 295

S

S

CH3

Dimethyl Ethanolamine

CH3 N

HO

CH3

1189 Dimethyl Ethanolamine/Acetic Acid H + N

O O

OH

1188 Dimethyl Ethanolamine/Formic Acid H + N

O O

OH

1185 Dimethyl Ethanolamine/Methacrylic Acid O

H + N OH

O

7248_A001.fm Page 404 Wednesday, May 23, 2007 12:31 PM

404

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1186 Dimethyl Ethanolamine/p-Toluene Sulfonic Acid

H + N

O OH

S

O OH

297

Toluene-4-sulfonate (2hydroxy-ethyl)dimethyl-ammonium;

17.2

21.5

22.5

Mercapto-acetate (2hydroxy-ethyl)dimethyl-ammonium;

17.2

21.5

22.5

Methoxymethane

15.2

6.1

5.7

63.2

N,N-Dimethylformamide

17.4

13.7

11.3

77.0

N,N-Dimethylhydrazine

15.3

5.9

11.0

76.0

Acetate (2-hydroxypropyl)-dimethylammonium;

16.6

19.4

18.0

2-Methyl-propen-1-one

15.2

7.4

4.8

87.6

(Z)-But-2-enedioic acid dimethyl ester

16.3

8.3

9.8

125.8

O

H + N

HS

O

Dimethyl Ether

O

H3C

Hydrogen Molar Bonding Volume

O

1187 Dimethyl Ethanolamine/Thioglycolic Acid

296

Dispersion Polarity

CH3

Dimethyl Formamide O CH3

N

CH3

298

1,1-Dimethyl Hydrazine

NH2 N

H3C

CH3

1190 Dimethyl Isopropanol Amine/Acetic Acid H + N

O OH

299

Dimethyl Ketene

O

O

H3C

CH3

1145 Dimethyl Maleate O O O O

CH3 CH3

7248_A001.fm Page 405 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

405

TABLE A.1 (CONTINUED) No. 839

Dimethyl Methyl Phosphonate H3C

Dispersion Polarity

Hydrogen Molar Bonding Volume

Methyl-phosphonic acid dimethyl ester

16.7

13.1

7.5

106.9

2,6-Dimethyl-phenol

19.1

4.9

12.9

116.3

3,4-Dimethyl-phenol

19.2

6.0

13.4

121.0

Phthalic acid dimeth yl ester

18.6

10.8

4.9

163.0

2,5-Dimethyl-1Hpyrrole

18.3

7.6

6.8

101.8

Decanedioic acid dimethyl ester

16.6

2.9

6.7

233.3

Sulfuric acid dimethyl ester

17.7

17.0

9.7

94.7

Methylsulfanyl-methane 16.1

6.4

7.4

73.2

O

O

P

H3C

849

Autonom/ ACD Name

Solvent Name

CH3

O

2,6-Dimethyl Phenol OH CH3

H3C

850

3,4-Dimethyl Phenol OH

CH3 CH3

300

Dimethyl Phthalate O CH3

O O

CH3

O

782

2,5-Dimethyl Pyrrole

H3C

CH3

N

1144 Dimethyl Sebacate O H3C

O

O

CH3

O

1080 Dimethyl Sulfate

O

H3C O

S

CH3

O

O 301

Dimethyl Sulfid

H3C

S

CH3

7248_A001.fm Page 406 Wednesday, May 23, 2007 12:31 PM

406

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 302

Solvent Name Dimethyl Sulfone

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Methylsulfonyl-methane 19.0

19.4

12.3

75.0

Methylsulfi yl-methane

18.4

16.4

10.2

71.3

2,3-Dimethyl-but-1-ene

14.9

1.2

2.8

125.2

2,2-Dimethyl-propan1-ol

15.6

6.5

13.5

107.5

2,4-Dimethylphenylamine

19.2

5.2

8.7

123.7

3,5-Dinitro-phenol

19.5

12.9

14.4

108.2

1,2-Dinitro-benzene

20.6

22.7

5.4

107.4

3,4-Dinitro-phenol

19.5

12.8

14.3

110.1

O S

H3C 303

CH3

O

Dimethyl Sulfoxide

O S

H3C 304

CH3

2,3-Dimethyl-1-Butene CH3 CH3

H2C

CH3

1027 2,2-Dimethyl-1-Propanol (Neopentyl alc,)

CH3 OH

H3C 952

CH3

2,4-Dimethylaniline NH2 CH3

CH3

942

3,5-Dinitrophenol O +

N

HO

O

+

O

940

N

O

1,2-Dinitrobenzene O +

N

+

N

O O

O

941

3,4-Dinitrophenol O

HO

+

N

+

N

O

O

O

7248_A001.fm Page 407 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

407

TABLE A.1 (CONTINUED) Dispersion Polarity

Hydrogen Molar Bonding Volume

20.0

13.1

4.9

137.9

Hexanedioic acid dioctyl 16.7 ester

2.0

5.1

400.0

Phthalic acid dioctyl ester

16.6

7.0

3.1

377.0

[1,4]Dioxane

19.0

1.8

7.4

85.7

[1,3]Dioxane

18.1

6.6

9.3

69.9

4-Isopropenyl-1-methyl- 17.2 cyclohexene

1.8

4.3

162.9

(Phenylethynyl)benzene

20.1

0

3.2

184.3

1167 Diphenyl Acetylene P from Group Cont. (Phenylethynyl)benzene

20.1

2.0

3.2

184.3

No. 938

Autonom/ ACD Name

Solvent Name 2,4-Dinitrotoluene CH3

1-Methyl-2,4-dinitrobenzene

O +

N

O

+

O

N

O

1142 Dioctyl Adipate Commercial O H3C

O

O

CH3

O

305

Dioctyl Phthalate Commercial O O

CH3

O

CH3

O

306

1,4-Dioxane

O O 307

1,3-Dioxolane

O O 1197 Dipentene CH3

H3C

CH2

1166 Diphenyl Acetylene P from 0 Dipole Moment

7248_A001.fm Page 408 Wednesday, May 23, 2007 12:31 PM

408

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 778

Autonom/ ACD Name

Solvent Name Diphenyl Ether*

Diphenyl ether

Dispersion Polarity

Hydrogen Molar Bonding Volume

19.5

3.4

5.8

160.4

(Phenylsulfonyl)benzene 21.1

14.4

3.4

174.3

O 779

Diphenyl Sulfone O S O

875

Diphenylamine

Diphenyl-amine

20.0

3.3

5.9

145.9

N 876

Diphenylmethane

Benzylbenzene

19.5

1.0

1.0

168.2

308

Dipropyl Amine

Dipropyl-amine

15.3

1.4

4.1

136.9

Heptan-4-one

15.8

5.7

4.9

140.8

3-(3-Hydroxy-propoxy)- 16.5 propan-1-ol

10.6

17.7

130.9

3-(3-Methoxy-propoxy)- 15.5 propan-1-ol

5.7

11.2

157.4

15.7

6.5

10.0

211.2

N

H3C 737

CH3

Dipropyl Ketone*

O CH3

H3C 309

Dipropylene Glycol Commercial

310

OH

O

HO

Dipropylene Glycol Meth yl Ether Commercial H3C

O

O

OH

1007 Dipropylene Glycol Mono n-Butyl Ether 1-(2-Butoxy-1-methylethoxy)-propan-2-ol CH 3

H3C

O OH

O

CH3

7248_A001.fm Page 409 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

409

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1008 Dipropylene Glycol Mono n-Prop yl Ether

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-(1-Methyl-2-propoxyethoxy)-propan-2-ol

15.6

6.1

11.0

185.6

Acetic acid 3-(3methoxy-propoxy)propyl ester

16.3

4.9

8.0

195.7

Methyldisulfanylmethane

17.3

7.8

6.5

88.6

Phthalic acid ditridec yl ester

16.6

5.4

1.9

558.3

1,4-Divinyl-benzene

18.6

1.0

7.0

142.8

Vinylsulfanyl-ethene

16.5

4.6

5.6

93.6

Dodecane

16.0

0

0

228.6

Dodecan-1-ol

16.0

4.0

9.3

224.5

4-(2-Amino-ethyl)benzene-1,2-diol

18.2

10.3

19.5

183.0

CH3 H3C

O

O

CH3

OH

311

Dipropylene Glycol Monomethyl Ether Acetate O H3C

312

CH3

2,3-Dithiabutane

S

H3C 313

O

O

O

S

CH3

Ditridecyl Phthalate O CH3

O O

CH3

O

314

p-Divinyl Benzene CH2

H2C

315

Divinyl Sulfid

H2C 316

S

CH2

Dodecane CH3

H3C

1075 Dodecanol H3C

OH

1180 Dopamine

HO HO

NH2

7248_A001.fm Page 410 Wednesday, May 23, 2007 12:31 PM

410

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1177 Ecstasy

CH3

O O 317

H3C

Hydrogen Molar Bonding Volume

(2-Benzo[1,3]dioxol-5yl-1-methyl-ethyl)methyl-amine

18.0

5.1

6.1

202.9

Icosane

16.5

0

0

359.8

2-Chloromethyl-oxirane

18.9

7.6

6.6

78.4

2-Ethyl-oxirane

15.6

8.0

4.6

87.9

2-Vinyl-oxirane

16.6

7.7

7.4

80.7

2-Methylene-oxirane

16.5

8.6

6.7

70.0

Azepan-2-one

19.4

13.8

3.9

110.7

Ethane-1,2-dithiol

17.9

7.2

8.7

83.9

Ethanesulfonyl chloride

17.7

14.9

6.8

94.7

Ethanethiol

15.7

6.5

7.1

74.3

N

Eicosane CH3

H3C

318

Dispersion Polarity

Epichlorohydrin

O Cl 987

1,2-Epoxy Butane

O 320

CH3

3,4-Epoxy-1-Butene

H2C 319

O

1,2-Epoxy-2-Propene

O 321

CH2

Epsilon-Caprolactam O N

322

1,2-Ethane Dithiol

HS 323

SH

Ethanesulfonylchloride

O S H3C 324

Cl

O

Ethanethiol (Ethyl Mercaptan)

H3C

SH

7248_A001.fm Page 411 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

411

TABLE A.1 (CONTINUED) No. 325

Ethanol

Dispersion Polarity

Hydrogen Molar Bonding Volume

Ethanol

15.8

8.8

19.4

58.5

2-Amino-ethanol

17.0

15.5

21.2

59.8

Acetate 2-hydroxyethyl-ammonium;

17.2

20.3

18.4

1-(4-Ethoxy-phenyl)ethanone

18.8

10.3

6.4

162.6

1,3-Diethoxy-propan-2ol

15.9

5.7

11.7

156.0

3-Ethoxypropionaldehyde

16.0

8.8

7.4

112.1

Propionic acid 2-ethoxy- 16.2 ethyl ester

3.3

8.8

155.5

Acetic acid eth yl ester

15.8

5.3

7.2

98.5

3-Oxo-butyric acid ethyl 16.5 ester

10.8

8.3

125.6

OH

H3C 326

Autonom/ ACD Name

Solvent Name

Ethanolamine

OH NH2 1191 Ethanolamine/Acetic Acid

O

+

NH3

O

OH 794

4-Ethoxy Acetophenone O

CH3

O CH3

776

1-Ethoxy Ethoxy-2-Propanol

OH H3C 761

3-Ethoxy Propionaldehyde

O

O

H3C 327

CH3

O

O

Ethoxyethyl Propionate

O

H3C

O

CH3

O 328

Ethyl Acetate

O

H3C

CH3

O

1016 Ethyl Aceto Acetate (Keto)

O H3C

O O

CH3

7248_A001.fm Page 412 Wednesday, May 23, 2007 12:31 PM

412

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 329

Autonom/ ACD Name

Solvent Name Ethyl Acetylene

Dispersion Polarity

Hydrogen Molar Bonding Volume

But-1-yne

15.1

3.4

5.0

81.5

Acrylic acid eth yl ester

15.5

7.1

5.5

108.8

Ethylamine

15.0

5.6

10.7

65.6

Octan-3-one

16.2

4.5

4.1

156.0

Ethyl-benzene

17.8

0.6

1.4

123.1

Benzoic acid eth yl ester

17.9

6.2

6.0

144.3

Bromo-ethane

16.5

8.4

2.3

74.6

Heptan-3-one

16.2

5.0

4.1

139.0

Butyric acid eth yl ester

15.5

5.6

5.0

132.9

HC CH3 330

Ethyl Acrylate

CH3

O

H2C O 331

Ethyl Amine

NH2

H3C 332

Ethyl Amyl Ketone

CH3

H3C O 333

Ethyl Benzene CH3

954

Ethyl Benzoate O CH3

O

334

Ethyl Bromide

H3C 335

Br

Ethyl Butyl K etone

CH3

H3C

O 928

Ethyl Butyrate

O H3C

O

CH3

7248_A001.fm Page 413 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

413

TABLE A.1 (CONTINUED) No. 929

Autonom/ ACD Name

Solvent Name Ethyl Caproate

H3C

O

Dispersion Polarity

Hydrogen Molar Bonding Volume

Hexanoic acid ethyl ester 15.5

3.2

5.9

149.6

Carbamic acid ethyl ester 16.8

10.1

13.0

91.2

Isocyano-ethane

15.6

15.2

5.8

74.4

Chloro-ethane

15.7

6.1

2.9

70.0

Ethyl chloridocarbonate

15.5

10.0

6.7

95.6

(E)-3-Phenyl-acrylic acid ethyl ester

18.4

8.2

4.1

166.8

2-Ethyl-but-2-enal

16.1

8.0

5.5

115.2

Cyano-acetic acid ethyl ester

16.7

7.9

8.3

107.1

2-Cyano-acrylic acid ethyl ester

15.2

10.3

9.0

117.1

CH3 O

336

Ethyl Carbamate

CH3

O

O 337

NH2

Ethyl Carbylamine +

H3C 338

N

Ethyl Chloride

H3C 339

C

Cl

Ethyl Chloroformate

O Cl 340

O

CH3

Ethyl Cinnamate O O

341

CH3

2-Ethyl Croton Aldehyde

CH3

O

CH3 1042 Ethyl Cyanoacetate

CH3

O N 342

O

Ethyl Cyanoacrylate N

O O CH2

CH3

7248_A001.fm Page 414 Wednesday, May 23, 2007 12:31 PM

414

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 343

Autonom/ ACD Name

Solvent Name Ethyl Ethynylether

Dispersion Polarity

Hydrogen Molar Bonding Volume

Ethoxy-ethyne

15.4

7.9

5.9

87.6

N-Ethyl-formamide

17.2

10.0

14.0

76.5

Formic acid eth yl ester

15.5

8.4

8.4

80.2

Acetic acid 2-eth ylhexyl ester

15.8

2.9

5.1

196.0

Acrylic acid 2-eth ylhexyl ester

14.8

4.7

3.4

208.2

17.3

6.2

6.8

265.0

Ethyl hypochlorite

15.7

8.6

6.5

79.5

Iodo-ethane

17.3

7.9

7.2

81.2

Isocyanato-ethane

15.4

12.0

2.5

78.7

O

HC

CH3 966

N-Ethyl Formamide

O CH3

N 344

Ethyl Formate

O CH3

O 346

2-Ethyl Hexyl Acetate O H3C

CH3

O H3C

347

2-Ethyl Hexyl Acrylate O

H2C

CH3

O H3C

1153 Mono 2-Ethyl-Hexyl Phthalate (MEHP) Phthalic acid mono-(2ethyl-hexyl) ester OH O O O CH3 CH3

348

Ethyl Hypochlorite

H3C 349

Cl

Ethyl Iodide

H3C 350

O

I

Ethyl Isocyanate

O H3C

N

7248_A001.fm Page 415 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

415

TABLE A.1 (CONTINUED) No. 351

Solvent Name Ethyl Isopropenyl Ether

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Ethoxy-propene

14.8

3.5

5.1

113.3

Isothiocyanato-ethane

17.2

14.7

9.0

87.1

2-Hydroxy-propionic acid ethyl ester

16.0

7.6

12.5

115.0

2-Methyl-acrylic acid ethyl ester

15.8

7.2

7.5

125.8

Methylsulfanyl-ethane

17.1

4.8

2.5

91.2

4-Ethyl-morpholine

17.7

5.0

6.6

116.5

(Z)-Octadec-9-enoic acid ethyl ester

14.5

3.8

3.7

357.3

4-Ethyl-phenol

19.2

5.3

12.8

118.6

CH3

352

CH2

O

H3C

Ethyl Isothiocyanate

S H3C 353

N

Ethyl Lactate O HO

CH3

O CH3

354

Ethyl Methacrylate O H2C

O

CH3

CH3

355

Ethyl Methyl Sulfid

H3C

S

CH3

1085 4-Ethyl Morpholine O N H3C

1117 Ethyl Oleate CH3

H3C

O O

1071 4-Ethyl Phenol OH

H3C

7248_A001.fm Page 416 Wednesday, May 23, 2007 12:31 PM

416

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Solvent Name

1010 Ethyl Propionate

Dispersion Polarity

Propionic acid ethyl ester 15.5

Hydrogen Molar Bonding Volume

6.1

4.9

115.5

CH3

O

H3C

Autonom/ ACD Name

O 356

Ethyl Thiocyanate

Thiocyanato-ethane

15.4

13.4

5.4

87.1

2-Ethoxy-but-1-ene

15.3

4.2

4.6

125.6

Ethoxy-ethene

14.9

4.9

5.6

94.9

Pent-1-en-3-one

15.8

11.3

4.5

99.3

2-Ethyl-butan-1-ol

15.8

4.3

13.5

123.2

3-Methylene-pentane

14.9

1.7

3.5

123.1

1-Ethoxy-propyne

15.9

6.4

5.4

101.6

3-Methylene-pent-1-ene

15.3

1.6

3.8

115.2

N H3C 357

S

1-Ethyl Vinyl Ethyl Ether

CH3 H3C 358

Ethyl Vinylether

H3C 359

CH2

O

CH2

O

Ethyl Vinylketone

CH2

H3C O 360

2-Ethyl-1-Butanol

OH

H3C H3C 727

2-Ethyl-1-Butene

CH2

H3C H3C 361

Ethyl-1-Propynylether

CH3

H3C 362

O

2-Ethyl-1,3-Butadiene

CH3 H2C

CH2

7248_A001.fm Page 417 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

417

TABLE A.1 (CONTINUED) No. 961

Solvent Name N-Ethyl-2-Pyrrolidone

Autonom/ ACD Name

Dispersion Polarity

1-Ethyl-pyrrolidin-2-one 18.0

Hydrogen Molar Bonding Volume

12.0

7.0

113.9

O

N H3C

345

2-Ethyl-hexanol

2-Ethyl-hexan-1-ol

15.9

3.3

11.8

156.6

Ethene

15.0

2.0

3.8

63.0

[1,3]Dioxolan-2-one

19.4

21.7

5.1

66.0

2-Chloro-ethanol

16.9

8.8

17.2

67.3

3-Hydroxy-propionitrile

17.2

18.8

17.6

68.3

1,2-Dibromo-ethane

19.2

3.5

8.6

87.0

1,2-Dichloro-ethane

19.0

7.4

4.1

79.4

OH

H3C H3C 1058 Ethylene

H2C 363

CH2

Ethylene Carbonate O O

364

O

Ethylene Chlorohydrin

OH Cl 365

Ethylene Cyanohydrin N

OH

366

Ethylene Dibromide

Br Br 367

Ethylene Dichloride

Cl Cl

7248_A001.fm Page 418 Wednesday, May 23, 2007 12:31 PM

418

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 937

Autonom/ ACD Name

Solvent Name Ethylene Dinitrate

Dispersion Polarity

Hydrogen Molar Bonding Volume

1,2-Bis-nitrooxy-ethane

16.2

13.0

4.5

101.7

Ethane-1,2-diol

17.0

11.0

26.0

55.8

Acetic acid 2-b utoxyethyl ester

15.3

4.5

8.8

171.2

1-(2-Ethoxy-ethoxy)butane

15.3

4.9

4.6

175.5

1-(2-Methoxy-ethoxy)butane

15.5

5.2

4.9

157.2

2-(2-tert-Butoxyethoxy)-2-methylpropane

14.7

4.1

8.2

210.0

Acetic acid 2-acetoxyethyl ester

16.2

4.7

9.8

132.8

1-(2-Butoxy-ethoxy)butane

15.7

4.5

4.2

209.5

1,2-Diethoxy-ethane

15.4

5.4

5.2

141.6

1,2-Dimethoxy-ethane

15.4

6.3

6.0

103.9

O +

O

N

O

O

+

N

O

O

368

Ethylene Glycol

OH OH 369

Ethylene Glycol Butyl Ether Acetate O

O

H3C

752

Ethylene Glycol Butyl Eth yl Ether H3C

753

O

O

O

CH3

Ethylene Glycol Di-t-Butyl Ether CH3

HC H3C3

O

O

CH3

371

CH3

O

Ethylene Glycol Butyl Meth yl Ether

H3C 370

CH3

O

CH CH3 3

Ethylene Glycol Diacetate O H3C

O

CH3

O

O

758

Ethylene Glycol Dib utyl Ether O

H3C

760

CH3

Ethylene Glycol Dieth yl Ether

H3C 775

O

O

O

Ethylene Glycol Dimeth yl Ether

H3C

O

O

CH3

CH3

7248_A001.fm Page 419 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

419

TABLE A.1 (CONTINUED) No. 372

Autonom/ ACD Name

Solvent Name Ethylene Glycol Meth yl t-Butyl Ether*

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-(2-Methoxy-ethoxy)2-methyl-propane

15.3

5.1

8.2

157.4

2-(2-Ethyl-hexyloxy)ethanol

16.0

4.1

10.5

194.7

2-Benzyloxy-ethanol

17.8

5.9

12.2

143.0

Acrylic acid 2-ethoxyethyl ester

15.9

5.1

2-Hexyloxy-ethanol

16.0

5.0

11.4

164.5

2-Propoxy-ethanol

16.1

8.7

13.5

114.1

2-tert-Butoxy-ethanol

15.3

6.1

10.8

131.0

2-Butoxy-ethanol

16.0

5.1

12.3

131.6

CH3

H3C 373

O

O

CH3 CH3

Ethylene Glycol Mono-2-Eth yl Hexyl Ether* HO

CH3

O

CH3

731

Ethylene Glycol Monobenzyl Ether O

771

OH

Ethylene Glycol Monoeth yl Ether Acrylate

9.3

147.7

O H3C

O

CH2

O

1056 Ethylene Glycol Mono n-He xyl Ether HO

CH3

O

1004 Ethylene Glycol Mono n-Prop yl Ether

HO 374

O

CH3

Ethylene Glycol Mono-t-Butyl Ether Commercial

CH3 HO

375

O

CH CH3 3

Ethylene Glycol Monob utyl Ether Commerical

HO

O

CH3

7248_A001.fm Page 420 Wednesday, May 23, 2007 12:31 PM

420

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 376

Ethylene Glycol Monoeth yl Ether Commercial

HO 377

Autonom/ ACD Name

Solvent Name

O

Hydrogen Molar Bonding Volume

2-Ethoxy-ethanol

16.2

9.2

14.3

97.8

Acetic acid 2-ethoxyethyl ester

15.9

4.7

10.6

136.1

2-Isobutoxy-ethanol

15.2

4.9

9.6

132.5

2-Isopropoxy-ethanol

16.0

8.2

13.1

115.8

2-Methoxy-ethanol

16.2

9.2

16.4

79.1

Acetic acid 2-methoxyethyl ester

15.9

5.5

11.6

121.6

2-Phenoxy-ethanol

17.0

7.2

12.3

125.3

[1,3,2]Dioxathiolane 2oxide

20.0

15.9

5.1

75.1

CH3

Ethylene Glycol Monoeth yl Ether Acetate Commercial

H3C

Dispersion Polarity

O

O

CH3

O 378

Ethylene Glycol Monoisob utyl Ether Commercial

HO

CH3

O

CH3 379

Ethylene Glycol Monoisoprop yl Ether

CH3 HO 380

O

Ethylene Glycol Monometh yl Ether Commercial

HO 381

CH3

O

CH3

Ethylene Glycol Monometh yl Ether Acetate

O

H3C

O

CH3

O 924

Ethylene Glycol Monophen yl Ether

O

HO

789

Ethylene Glycol Sulfit O O

S

O

7248_A001.fm Page 421 Wednesday, May 23, 2007 12:31 PM

Appendix A: Table A.1

421

TABLE A.1 (CONTINUED) No. 382

Ethylene Methyl Sulfonate

O

O O

S

O S

O

O

Dispersion Polarity

Hydrogen Molar Bonding Volume

Methanesulfonic acid 2methanesulfonyloxyethyl ester

16.9

9.3

9.6

97.9

Oxirane

15.6

10.0

11.0

49.9

Thiirane

19.3

9.0

6.5

59.5

Ethane-1,2-diamine

16.6

8.8

17.0

67.3

Aziridine

18.6

9.8

7.7

51.8

Ethyl-hexyl-amine

15.7

4.2

6.1

163.3

Pent-3-en-2-one

16.2

12.1

4.5

99.0

Methoxy-ethyne

15.8

8.1

6.5

70.1

1-Ethynyloxy-propane

15.4

3.8

5.4

104.1

CH3

CH3 383

Autonom/ ACD Name

Solvent Name

Ethylene Oxide

O 384

Ethylene Sulfid

S 385

Ethylenediamine

NH2 NH2 386

Ethyleneimine

N 946

2-Ethylhexylamine

387

CH3

N

H3C

Ethylidene Acetone

O CH3

H3C 389

Ethynyl Methyl Ether

HC

O CH3

388

Ethynyl Propyl Ether

HC

O CH3

7248_A001.fm Page 422 Wednesday, May 23, 2007 12:31 PM

422

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Solvent Name

1109 Eugenol OH

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

4-Allyl-2-methoxyphenol

19.0

7.5

13.0

154.0

7,7-Dimethyl-2methylenebicyclo[2.2.1]heptane

16.9

1.5

3.1

157.3

(E)-3-(4-Hydroxy-3methoxy-phenyl)acrylic acid

19.0

6.6

15.1

155.5

CH3

O H2C

898

Fenchene(Alfa) CH3

H3C

CH2

1104 Ferulic Acid OH O O

H3C

HO

877

Fluorene

9H-Fluorene

20.0

1.7

1.7

138.2

390

1-Fluoro Acrylic Acid

2-Fluoro-acrylic acid

16.0

8.7

13.0

90.0

2-Fluoro-acrylonitrile

14.1

15.4

5.7

88.8

1,4-Difluoro-2-nitro benzene

19.4

7.2

3.7

108.4

1-Fluoro-4-methoxybenzene

18.7

7.3

6.7

113.2

F O

H2C OH

391

1-Fluoro Acrylonitrile

F

H2C 827

N

4-Fluoro-3-Nitrobenzofluorid F

+

N

F

799

O

O

p-Fluoroanisole O

F

CH3

7248_A001A.fm Page 423 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

423

TABLE A.1 (CONTINUED) No. 392

Solvent Name Fluorobenzene

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Fluoro-benzene

18.7

6.1

2.0

94.7

Fluoro-ethene

15.2

7.0

1.0

57.5

Fluoromethane

13.4

10.6

9.5

40.7

2-Fluoro-buta-1,3-diene

14.2

5.8

1.0

85.5

1-(4-Fluoro-phenyl)butan-1-one

19.6

7.1

3.5

138.8

Formaldehyde

12.8

14.4

15.4

36.8

Formamide

17.2

26.2

19.0

39.8

Formic acid

14.3

11.9

16.6

37.8

Formyl fluorid

15.0

10.1

8.6

56.5

Azepane-1-carbaldehyde 18.5

10.4

7.6

127.0

F

393

Fluoroethylene

H2C 394

Fluoromethane

H3C 395

F

F

Fluoroprene

F CH2

H2C 820

4-Fluoropropiophenone CH3 O

F

396

Formaldehyde

H2C 397

O

Formamide

O 398

NH2

Formic Acid

OH

O

741

Formyl Fluoride

O 400

F

N-Formyl Hexamethylene Imine O N

7248_A001A.fm Page 424 Wednesday, May 23, 2007 12:38 PM

424

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 401

Autonom/ ACD Name

Solvent Name N-Formyl Piperidine O

Dispersion Polarity

Hydrogen Molar Bonding Volume

Piperidine-1carbaldehyde

18.7

10.6

7.8

111.5

(E)-But-2-enedinitrile

16.7

13.6

7.8

83.0

Furan

17.8

1.8

5.3

72.5

Furan-2-carbaldehyde

18.6

14.9

5.1

83.2

Furan-2-yl-methanol

17.4

7.6

15.1

86.5

Furan-2-carbonitrile

18.4

15.0

8.2

87.5

Propane-1,2,3-triol

17.4

12.1

29.3

73.3

Acetic acid 3-acetoxy-2- 16.4 hydroxy-propyl ester

8.9

14.2

149.4

N

402

Fumaronitrile

N N 403

Furan

O 404

Furfural

O 405

O

Furfuryl Alcohol

O 994

OH

2-Furonitrile

N

O 406

Glycerol OH

OH

OH

1222 Glycerol Diacetate (Isomer Mix) CH3

O H3C

O

O OH

O

7248_A001A.fm Page 425 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

425

TABLE A.1 (CONTINUED) No. 990

Autonom/ ACD Name

Solvent Name Glycidaldehyde*

O

Dispersion Polarity

Hydrogen Molar Bonding Volume

Oxirane-2-carbaldehyde

17.5

13.4

9.8

63.2

Oxiranyl-methanol

18.2

9.0

17.9

66.5

2-Methyl-acrylic acid oxiranylmethyl ester

16.3

8.5

5.7

136.4

Ethanedial

15.0

17.0

13.3

50.9

Heptane

15.3

0

0

147.4

Heptan-1-ol

16.0

5.3

11.7

141.4

Heptan-2-ol

15.7

5.4

11.7

142.9

Heptan-3-ol

15.9

5.4

11.7

141.2

Hept-1-ene

15.0

1.1

2.6

141.9

Acetic acid heptyl ester

15.8

2.9

5.5

181.1

O

1087 Glycidol

O

OH 407

Glycidyl Methacrylate CH3 O

H2C

O

O

408

Glyoxal (Ethandial)

O 409

O

Heptane

H3C 931

CH3

1-Heptanol

H3C

OH

1044 2-Heptanol

OH CH3

H3C 1013 3-Heptanol

CH3

H3C

OH 410

1-Heptene

CH2

H3C 411

n-Heptyl Acetate H3C

CH3

O

O

7248_A001A.fm Page 426 Wednesday, May 23, 2007 12:38 PM

426

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 706

Hexachloroacetone P Based on 0 Dipole Moment Cl

Hydrogen Molar Bonding Volume

1,1,1,3,3,3-Hexachloropropan-2-one

18.3

0

0

151.9

1,1,1,3,3,3-Hexachloropropan-2-one

18.3

6.6

6.4

151.9

1,2,3,4,5,6-Hexachlorobenzene

21.9

2.1

0

139.3

1,1,1,2,2,2-Hexachloroethane

22.0

4.7

0

113.2

Hexadecane

16.3

0

0

294.1

1,1,2,3,4,4-Hexafluoro buta-1,3-diene

13.8

0

0

104.3

1,1,1,3,3,3-Hexafluoro propan-2-ol

17.2

4.5

14.7

105.3

1,2,3,4,5,6-Hexafluoro benzene

16.9

0

0

115.8

Cl Cl Cl

Hexachloroacetone P based on group contrib utions Cl

Dispersion Polarity

O

Cl Cl

707

Autonom/ ACD Name

Solvent Name

O

Cl

Cl Cl Cl

Cl

1128 Hexachlorobenzene (Lindane) Cl Cl

Cl Cl

Cl Cl

1121 Hexachloroethane

Cl

Cl

Cl

Cl

Cl

Cl 412

Hexadecane CH3

H3C

413

Hexafluoro 1,3-Butadien F

F F

F F

F

414

Hexafluo o Isopropanol F

OH

F

F F

F

F

1119 Hexafluorobenzen F F

F

F

F F

7248_A001A.fm Page 427 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

427

TABLE A.1 (CONTINUED) No. 415

Autonom/ ACD Name

Solvent Name Hexafluo ohexanol

1,2,3,4,5,6-Hexafluoro hexan-1-ol

F

F

F F

Dispersion Polarity

Hydrogen Molar Bonding Volume

15.1

4.4

9.9

123.3

1,2,3,4,5,6-Hexamethyl- 19.2 benzene

1.6

0

152.7

19.4

11.6

13.8

105.3

18.5

8.6

11.3

175.7

Hexanal

15.8

8.5

5.4

120.2

n-Hexane

14.9

0

0

131.6

Hexanoic acid

15.0

4.1

9.4

125.9

Hexan-1-ol

15.9

5.8

12.5

124.9

Hex-1-ene

14.7

1.1

0

126.1

OH

F

F

1123 Hexamethyl Benzene CH3 CH3

H3C

CH3

H3C CH3

948

Hexamethylenetetramine

1,3,5,7-Tetraazatricyclo[3.3.1.1*3,7*] decane

N N 416

N

N

Hexamethylphosphoramide H3C O

H3C

CH3 N CH3 P

N

N

CH3 CH3

1097 Hexanal

O

H3C 417

n-Hexane

H3C

CH3

1022 Hexanoic Acid

O

H3C 930

1-Hexanol

H3C 418

OH

OH

1-Hexene*

H3C

CH2

7248_A001A.fm Page 428 Wednesday, May 23, 2007 12:38 PM

428

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 419

Autonom/ ACD Name

Solvent Name Hexyl Acetate

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetic acid he xyl ester

15.8

2.9

5.9

165.0

Hexane-1,6-diol

15.7

8.4

17.8

123.0

Acetic acid 6-acetoxyhexyl ester

15.3

4.5

7.2

204.3

Hydrazine

14.2

8.3

8.9

32.1

Nitrilomethane

12.3

17.6

9.0

39.3

Hydrogen peroxide

15.5

12.2

42.7

23.2

17.9

6.0

10.2

36.1

4-Hydroxybenzaldehyde

19.4

15.2

14.5

108.2

(E)-3-(4-Hydroxyphenyl)-acrylic acid

19.1

6.7

15.9

128.3

Tetrahydro-furan-3-ol

18.9

9.4

16.3

80.0

O

H3C 420

O

CH3

Hexylene Glycol

OH

HO 421

Hexylene Glycol Diacetate O CH3

O

O

H3C

O

422

Hydrazine

H2N 423

NH2

Hydrogen Cyanide

HC

N

1173 Hydrogen Peroxide

HO 424

OH

Hydrogen Sulfid

SH2 905

4-Hydroxy Benzaldehyde O

OH

1105 4-Hydroxy Cinnamic Acid OH O

HO

991

3-Hydroxy Tetrahydrofuran

O

OH

7248_A001A.fm Page 429 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

429

TABLE A.1 (CONTINUED) No. 425

Autonom/ ACD Name

Solvent Name Hydroxyethyl Acrylate

O

H2C

Dispersion Polarity

Acrylic acid 2-h ydroxy- 16.0 ethyl ester

Hydrogen Molar Bonding Volume

13.2

13.4

114.9

OH

O 960

N-(2-Hydroxyethyl)-2-Pyrrolidone

1-(2-Hydroxy-ethyl)pyrrolidin-2-one

18.0

9.2

15.7

113.4

1078 Indene

1H-Indene

18.7

2.6

9.0

116.5

831

1H-Indole

19.8

7.5

6.5

110.1

4-Iodo-buta-1,2-diene

17.4

6.3

6.2

105.1

Iodo-benzene

19.5

6.0

6.1

114.4

Triiodo-methane

20.2

3.6

10.6

98.2

2-Iodo-buta-1,3-diene

17.2

2.5

6.2

104.2

O

N

OH

Indole

N 426

4-Iodo-1,2-Butadiene

I H2C 427

Iodobenzene* I

1126 Iodoform

I

I I

428

Iodoprene

I H2C

CH2

7248_A001A.fm Page 430 Wednesday, May 23, 2007 12:38 PM

430

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 429

Isoamyl Acetate

CH3

O

H3C 754

Autonom/ ACD Name

Solvent Name

O

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetic acid 3-meth ylbutyl ester

15.3

3.1

7.0

148.8

3-Methyl-butan-1-ol

15.8

5.2

13.3

109.4

Propionic acid 3-methyl- 15.7 butyl ester

5.2

5.6

165.7

Acetic acid isobutyl ester 15.1

3.7

6.3

133.5

Acrylic acid isob utyl ester

15.5

6.2

5.0

145.0

2-Methyl-propan-1-ol

15.1

5.7

15.9

92.8

Formic acid isob utyl ester

15.5

6.5

6.7

117.4

Isobutyric acid isob utyl ester

15.1

2.9

5.9

163.0

CH3

Isoamyl Alcohol (3-Methyl-1-Butanol)

CH3

H3C 925

OH

Isoamyl Propionate

H3C

O

CH3 O

CH3 430

Isobutyl Acetate CH3 CH3

O

H3C

O

770

Isobutyl Acrylate CH3 O

H3C

CH2 O

431

Isobutyl Alcohol CH3 H3C OH

1035 Isobutyl Formate O CH3

O

CH3

432

Isobutyl Isobutyrate CH3

CH3 H3C

O

CH3

O

7248_A001A.fm Page 431 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

431

TABLE A.1 (CONTINUED) No. 786

Isobutyl Sulfoxide

O

CH3

S

H3C 433

Autonom/ ACD Name

Solvent Name

CH3

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-1-(2-methylpropane-1-sulfi yl)propane

16.3

10.5

6.1

195.0

Isobutene

14.5

2.0

1.5

89.4

2,2-Dimethyl-oxirane

16.1

4.8

5.8

90.0

Isobutyric acid

16.5

5.4

11.1

93.4

Imino-methanone

15.8

10.5

13.6

37.7

6-Methyl-heptan-1-ol

14.4

7.3

12.9

156.6

2-Methyl-butane

13.7

0

0

117.4

3,5,5-Trimethylcyclohex-2-enone

16.6

8.2

7.4

150.5

CH3

Isobutylene

CH2 CH3

H3C 434

Isobutyleneoxide

O CH3

H3C

1039 Isobutyric Acid O H3C

OH CH3

435

Isocyanic Acid

O HN 436

Isooctyl Alcohol

CH3 OH

H3C 437

Isopentane

CH3 CH3

H3C 438

Isophorone O CH3 CH3 H3C

7248_A001A.fm Page 432 Wednesday, May 23, 2007 12:38 PM

432

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 439

Autonom/ ACD Name

Solvent Name Isoprene (2-Methyl-1,3-Butadiene)

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-buta-1,3-diene 14.7

1.4

4.1

100.9

Acetic acid isoprop yl ester

14.9

4.5

8.2

117.1

Isopropylamine

14.8

4.4

6.6

86.8

Isopropyl-benzene

18.1

1.2

1.2

139.1

2-Chloro-propane

15.9

8.3

2.1

91.7

2-Isopropoxy-propane

13.7

3.9

2.3

140.9

Hexadecanoic acid isopropyl ester

14.3

3.9

3.7

330.0

3-Methyl-butyraldehyde 14.7

9.5

5.0

106.0

3-Methyl-butyric acid

4.1

10.7

111.0

CH3 CH2

H2C 440

Isopropyl Acetate

CH3

H3C 441

O CH3

O

Isopropyl Amine (2-Propan Amine)

CH3 NH2

H3C 878

Isopropyl Benzene (Cumene) CH3

H3C

442

Isopropyl Chloride (2-Chloro Propane)

Cl CH3

H3C 443

Isopropyl Ether

CH3

Isopropyl Palmitate H3C

CH3

O CH3

445

CH3

O

H3C 444

CH3

O

Isovaleraldehyde

CH3

O

H3C 1040 Isovaleric Acid

CH3 H3C

O OH

16.4

7248_A001A.fm Page 433 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

433

TABLE A.1 (CONTINUED) No. 446

Isoxazole

Ketene

H2C 709

Dispersion Polarity

Hydrogen Molar Bonding Volume

Isoxazole

18.8

13.4

11.2

64.1

Ethenone

15.4

7.3

5.8

53.0

2-Hydroxy-propionic acid

17.0

8.3

28.4

72.1

2-Methyl-acrylic acid dodecyl ester

14.4

2.2

5.1

293.1

Furan-2,5-dione

20.2

18.1

12.6

66.3

Malononitrile

17.7

18.4

6.7

55.5

(4-Chloro-phenoxy)acetic acid 2dimethylamino-ethyl ester

16.0

6.2

9.0

198.3

3,6-Dimethyl-4,5,6,7tetrahydro-benzofuran

17.2

4.6

5.6

158.0

(1R,2S,5R)-2-Isopropyl- 16.6 5-methyl-cyclohexanol

4.7

10.6

175.6

N

O 447

Autonom/ ACD Name

Solvent Name

O

Lactic Acid (DL) O CH3

HO OH

448

Lauryl Methacrylate CH3 O

H2C

CH3

O

980

Maleic Anhydride

O

449

O

O

Malononitrile

N

N

1208 Meclofenoxate (Base Only) O O

CH3 O

N

CH3

Cl

1129 Menthofuran CH3 O

CH3

1132 L-Menthol H3C

CH3

H3C HO

7248_A001A.fm Page 434 Wednesday, May 23, 2007 12:38 PM

434

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Solvent Name

1130 L-Menthone O H3C

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

(2S,5R)-2-Isopropyl-5methyl-cyclohexanone

17.0

8.1

4.4

172.3

Acetic acid (1R,2S,5R)2-isopropyl-5-methylcyclohexyl ester

16.8

4.7

4.9

215.8

4-Methyl-pent-3-en-2one

16.4

6.1

6.1

115.6

Mesitylene

18.0

0

0.6

139.8

2-Methyl-propenal

15.7

11.1

7.4

83.3

2-Methyl-acrylamide

15.8

11.0

11.6

76.0

2-Methyl-acrylic acid

15.8

2.8

10.2

84.8

2-Methyl-acrylonitrile

15.8

15.1

5.4

83.9

CH3

H3C

1131 L-Menthyl Acetate H3C O

CH3

O CH3 CH3

450

Mesityl Oxide

O

CH3

CH3

H3C 451

Mesitylene CH3

CH3

H3C

452

Methacrylaldehyde

CH3

O

H2C 453

Methacrylamide CH3 NH2

H2C O

454

Methacrylic Acid CH3 OH

H2C O

455

Methacrylonitrile

CH3 H2C

N

7248_A001A.fm Page 435 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

435

TABLE A.1 (CONTINUED) No. 456

Solvent Name Methanol

4-Methoxy Acetophenone CH3

O

H3C

781

Dispersion Polarity

Hydrogen Molar Bonding Volume

Methanol

15.1

12.3

22.3

40.7

1-(4-Methoxy-phenyl)ethanone

18.9

11.2

7.0

137.8

4-Methoxy-benzonitrile

19.4

16.7

5.4

121.0

3-Methoxy-butan-1-ol

15.3

5.4

13.6

113.2

Acetic acid 3-methoxybutyl ester

15.3

4.1

8.1

153.9

2-Methoxy-tetrahydropyran

17.2

6.6

6.0

105.1

1-Methoxy-buta-1,3diene

15.5

8.3

5.4

101.8

2-Methoxy[1,3]dioxolane

17.8

8.4

7.7

95.3

OH

H3C 793

Autonom/ ACD Name

O

4-Methoxy Benzonitrile N

H3C

O

1003 3-Methoxy Butanol

O

CH3

OH

H3C

1001 3-Methoxy Butyl Acetate

2-Methoxy Tetrahydropyrane

O 457

O

CH3

1-Methoxy-1,3-Butadiene

H3C 992

CH3

O

H3C 996

CH3

O

O

O

CH2

2-Methoxy-1,3-Dioxolane O H3C

O O

7248_A001A.fm Page 436 Wednesday, May 23, 2007 12:38 PM

436

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Solvent Name

1021 1-Methoxy-2-Nitrobenzene CH3 O O

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Methoxy-2-nitrobenzene

19.6

16.3

5.5

123.5

4-Methoxy-4-methylpentan-2-one

15.3

6.0

5.9

143.5

2-Methoxy-phenol

18.0

8.2

13.3

109.5

3-Methoxy-propionitrile 16.6

14.4

7.8

91.1

Acetate 3-methoxypropyl-ammonium;

17.2

22.5

23.5

(Z)-But-2-enoic acid

16.8

5.2

12.4

83.9

1-Methoxy-propene

15.0

4.3

5.7

93.5

N-Methyl-acetamide

16.9

18.7

13.9

76.9

+

N

458

Methoxyhexanone (Pentoxone) O

CH3 H3C

H3C

459

O

CH3

O

o-Methoxyphenol (Guaiacol) OH O

460

CH3

3-Methoxypropionitrile

H3C

N

O

1192 3-Methoxypropyl Amine/Acetic Acid O

+

NH3

O O

461

2-Methyl (cis) Acrylic Acid

OH CH3 463

Methyl 1-Propenyl Ether

H3C 958

O

O

CH3

N-Methyl Acetamide

O H3C

N

CH3

7248_A001A.fm Page 437 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

437

TABLE A.1 (CONTINUED) No. 464

Autonom/ ACD Name

Solvent Name Methyl Acetate

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetic acid meth yl ester 15.5

7.2

7.6

79.7

3-Oxo-butyric acid methyl ester

16.4

8.6

8.9

108.3

Propyne

15.1

3.8

9.2

59.6

Acrylic acid methyl ester 15.3

6.7

9.4

90.3

But-2-en-1-ol

16.0

6.0

15.5

84.4

2-Methyl-but-3enenitrile

16.4

11.3

5.1

97.7

Methylamine

13.0

7.3

17.3

44.4

Acetic acid 1,3dimethyl-butyl ester

15.2

3.1

6.8

167.4

Benzoic acid meth yl ester

18.9

8.2

4.7

124.9

O O

H3C

CH3

1037 Methyl Aceto Acetate

O

O

O

H3C 465

Methyl Acetylene

HC 467

CH3

CH3

Methyl Acrylate

O

H2C

CH3

O 468

3-Methyl Allyl Alcohol

H3C 469

OH

Methyl Allyl Cyanide

N H2C CH3 470

Methyl Amine

H3C 471

NH2

Methyl Amyl Acetate

H3C

O CH3

CH3 472

Methyl Benzoate* O O

CH3

O CH3

7248_A001A.fm Page 438 Wednesday, May 23, 2007 12:38 PM

438

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 473

Solvent Name Methyl Bromide

Dispersion Polarity

Hydrogen Molar Bonding Volume

Bromomethane

17.0

8.8

2.6

56.8

Hexan-2-one

15.3

6.1

4.1

123.6

Chloromethane

15.3

6.1

3.9

55.4

Methyl chloridocarbonate

16.3

9.5

8.5

77.3

Cyano-acetic acid methyl ester

16.8

14.8

9.1

88.3

Methyl-cyclohexane

16.0

0

1.0

128.3

1-Methyl-cyclohexanol

17.1

6.4

12.5

123.4

3-Methyl-cyclohexanol

17.2

6.4

12.5

124.8

4-Methyl-cyclohexanol

17.2

6.3

12.5

125.5

Br

H3C 474

Autonom/ ACD Name

Methyl Butyl K etone

O CH3

H3C 475

Methyl Chloride

Cl

H3C 476

Methyl Chloroformate

Cl

O

CH3

O 1041 Methyl Cyanoacetate

O N 477

CH3

O

Methyl Cyclohexane CH3

1030 1-Methyl Cyclohexanol

H3C

OH

1031 3-Methyl Cyclohexanol (mix) OH

CH3

1032 4-Methyl Cyclohexanol (mix) OH

CH3

7248_A001A.fm Page 439 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

439

TABLE A.1 (CONTINUED) No.

Solvent Name

1033 2-Methyl Cyclohexanol (mix)

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-cyclohexanol

17.1

6.5

12.5

124.1

3-Methylcyclohexanone

17.7

6.3

4.7

122.5

2-Methylcyclohexanone

17.6

6.3

4.7

121.3

Methoxy-ethane

14.7

4.9

6.2

84.1

Butan-2-one

16.0

9.0

5.1

90.1

Butan-2-one oxime

14.7

4.9

7.8

94.8

N-Methyl-formamide

17.4

18.8

15.9

59.1

Formic acid methyl ester 15.3

8.4

10.2

62.2

OH CH3

478

3-Methyl Cyclohexanone O

CH3

479

2-Methyl Cyclohexanone O CH3

480

Methyl Ethyl Ether

481

CH3

O

H3C

Methyl Ethyl Ketone

O CH3

H3C 482

Methyl Ethyl Ketoxime

N

OH

CH3

H3C 965

N-Methyl Formamide

O N 483

CH3

Methyl Formate

O

O

CH3

7248_A001A.fm Page 440 Wednesday, May 23, 2007 12:38 PM

440

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 995

Solvent Name 2-Methylfuran

Methyl Furoate

O O 484

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-furan

17.3

2.8

7.4

89.7

Furan-2-carboxylic acid methyl ester

17.4

6.9

9.7

107.0

2-Oxo-propionaldehyde

15.5

16.1

9.7

68.9

Methyl-hydrazine

16.2

8.7

14.8

52.7

Methyl-hydroperoxide

15.0

15.0

30.0

24.1

1-Methyl-1H-imidazole

19.7

15.6

11.2

79.5

Iodomethane

17.5

7.7

5.3

62.3

5-Methyl-hexan-2-one

16.0

5.7

4.1

142.8

4-Methyl-pentan-2-ol

15.4

3.3

12.3

127.2

CH3

O 998

Autonom/ ACD Name

CH3

O

Methyl Glyoxal O CH3 O

485

Methyl Hydrazine

N

H3C 486

Methyl Hydroperoxide

O

H3C 487

NH2

OH

1-Methyl Imidazole N N CH3

488

Methyl Iodide

I

H3C 489

Methyl Isoamyl K etone O CH3

H3C

CH3

490

Methyl Isobutyl Carbinol

HO

CH3 CH3

CH3

7248_A001A.fm Page 441 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

441

TABLE A.1 (CONTINUED) No. 491

Methyl Isobutyl Ketone

O

Hydrogen Molar Bonding Volume

4-Methyl-pentan-2-one

15.3

6.1

4.1

125.8

Isocyanatomethane

15.6

7.3

2.5

61.8

3-Methyl-but-3-en-2one

15.9

12.1

4.5

99.3

Isothiocyanatomethane

17.3

16.2

10.1

68.4

3-Methyl-isoxazole

19.4

14.8

11.8

57.7

Methanethiol

16.6

7.7

8.6

54.1

2-Methyl-acrylic acid methyl ester

15.8

6.5

5.4

106.1

4-Methyl-morpholine 4oxide

19.0

16.1

10.2

97.6

Heptan-2-one

16.2

5.7

4.1

139.8

CH3

Methyl Isocyanate

N

H3C 493

Dispersion Polarity

CH3 CH3

492

Autonom/ ACD Name

Solvent Name

O

Methyl Isopropenyl Ketone O

CH2

H3C

CH3

494

Methyl Isothiocyanate

N

H3C 495

S

3-Methyl Isoxazole

CH3 N

O 496

Methyl Mercaptan

SH

H3C 497

Methyl Methacrylate CH3 O

H2C

CH3

O

1106 N-Methyl Morpholine-N-Oxide O +

N H3C

498

O

Methyl n-Amyl K etone

O H3C

CH3

7248_A001A.fm Page 442 Wednesday, May 23, 2007 12:38 PM

442

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 499

Solvent Name Methyl n-Propyl Ketone

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Pentan-2-one

16.0

7.6

4.7

106.7

1-Methyl-naphthalene

20.6

0.8

4.7

138.8

Nitrooxymethane

15.8

14.0

4.8

63.8

(Z)-Octadec-9-enoic acid methyl ester

14.5

3.9

3.7

340.0

Methylsulfanyl-benzene

19.6

4.8

4.7

117.4

Methanesulfonylbenzene

20.0

16.9

7.8

124.6

Methylphosphonic difluorid

14.0

14.0

8.4

73.9

Propionic acid meth yl ester

15.5

6.5

7.7

96.8

O H3C

CH3

500

1-Methyl Naphthalene CH3

501

Methyl Nitrate

O

H3C

+

N

O

O 502

Methyl Oleate H3C

CH3 O O

879

Methyl Phenyl Sulfid

S

CH3

1076 Methyl Phenyl Sulfone CH3 O

841

O

S

Methyl Phosphonic Difluorid

F H3C P F 503

O

Methyl Propionate

O

H3C O

CH3

7248_A001A.fm Page 443 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

443

TABLE A.1 (CONTINUED) No. 902

Solvent Name 2-Methyl Pyrazine

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-pyrazine

18.3

12.3

10.5

91.4

1-Methyl-pyrrolidine

17.0

2.8

6.9

104.0

2-Hydroxy-benzoic acid methyl ester

18.1

8.0

13.9

129.6

Methyl-silane

15.5

3.3

0

71.0

1-Methyl-2-vinylbenzene

18.5

2.4

2.4

130.0

3-Methyl-tetrahydrothiophene 1,1-dioxide

19.4

17.4

5.3

112.7

2-Methyl-tetrahydrofuran

16.9

5.0

4.3

100.2

Thiocyanatomethane

17.3

15.0

6.0

68.5

Thiocyanatomethane

17.4

15.0

8.7

68.5

N

H3C

N 986

N-Methyl Pyrrolidine CH3 N

504

Methyl Salicylate O

OH

O

740

Methyl Silane

SiH3

H3C 886

CH3

alfa-Methyl Styrene CH3 CH2

506

Methyl Sulfolane

O

S

CH3

O 997

2-Methyl Tetrahydrofuran

O

507

Methyl Thiocyanate

H3C 968

CH3

S N

Methyl Thiocyanate

N

S

CH3

7248_A001A.fm Page 444 Wednesday, May 23, 2007 12:38 PM

444

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 508

Solvent Name Methyl Vinyl Ether

H3C 509

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Methoxy-ethene

14.9

5.3

6.3

75.2

But-3-en-2-one

15.6

12.5

5.0

81.2

2-Methoxy-propene

14.8

4.2

5.6

96.1

Methylsulfanyl-ethene

16.4

4.9

6.0

82.1

Methylsulfonyl-ethene

16.8

19.6

4.8

87.6

2-Methyl-butan-1-ol

16.0

5.1

14.3

109.5

3-Methyl-but-1-ene

14.0

1.4

3.8

112.9

2-Methyl-but-1-ene

14.2

1.8

2.3

108.7

2-Chloro-but-2-enal

17.1

10.6

7.3

91.7

CH2

O

Methyl Vinyl Ketone

O CH2

H3C 510

1-Methyl Vinyl Methyl Ether

CH3 O

H2C 511

Methyl Vinyl Sulfid

H3C 512

CH3

CH2

S

Methyl Vinyl Sulfone

O H3C 462

CH2

S

O

2-Methyl-1-Butanol

H3C

OH CH3

513

3-Methyl-1-Butene

CH3 CH2

H3C 514

2-Methyl-1-Butene

CH2

H3C

CH3 515

2-Methyl-1-Chloro Acrolein

Cl

H3C

O

7248_A001A.fm Page 445 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

445

TABLE A.1 (CONTINUED) No.

Solvent Name

1028 2-Methyl-1-Pentanol

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Methyl-pentan-1-ol

15.8

4.9

13.5

124.0

2-Methyl-propan-1-ol

15.1

5.7

15.9

92.8

1-Methoxy-propyne

15.7

6.3

5.9

85.4

3-Methyl-buta-1,2-diene 15.1

2.5

4.5

99.7

2-Methyl-[1,3]dioxolane 17.3

4.8

5.8

89.8

2-Methyl-butan-2-ol

15.3

6.1

13.3

109.6

3-Methyl-butan-2-ol

15.6

5.2

13.4

107.7

2-Methyl-but-2-ene

14.3

2.0

3.9

106.7

CH3 OH

H3C 516

2-Methyl-1-Propanol

CH3 OH

H3C 517

Methyl-1-Propynyl Ether

H3C

O CH3

518

3-Methyl-1,2-Butadiene

CH3 H2C CH3 519

2-Methyl-1,3-Dioxolane O

O CH3

732

2-Methyl-2-Butanol

OH H3C

CH3

CH3

1029 3-Methyl-2-Butanol CH3

CH3

H3C OH

520

2-Methyl-2-Butene CH3 H3C CH3

7248_A001A.fm Page 446 Wednesday, May 23, 2007 12:38 PM

446

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 521

Solvent Name N-Methyl-2-Pyrrolidone O

1002 3-Methyl-3-Methoxy Butyl Acetate O

O

972

CH3

CH3

O

S

Hydrogen Molar Bonding Volume

1-Methyl-pyrrolidin-2one

18.0

12.3

7.2

96.5

Acetic acid 3-methoxy3-methyl-butyl ester

15.3

3.8

7.7

168.6

Toluene-4-sulfonic acid methyl ester

19.6

15.3

3.8

152.6

4-Methyl-benzoic acid methyl ester

19.0

6.5

3.8

140.4

2-Methoxy-2-methylpropane

14.8

4.3

5.0

119.8

Dimethoxy-methane

15.0

1.8

8.6

169.4

1-Isocyanato-4-(4isocyanatobenzyl) benzene

19.5

4.1

1.7

212.1

Methyl-phenyl-amine

19.5

6.0

11.5

108.4

CH3 CH3

O

Methyl-4-Toluenesulfonate

H3C

Dispersion Polarity

CH3

N

H3C

Autonom/ ACD Name

O

O

796

Methyl-p-Toluate CH3 O CH3 O

522

Methyl-t-Butyl Ether

H3C

H3C O

CH3

CH3 523

Methylal (Dimethoxymethane)

H3C

O

O

CH3

1217 4,4'-Methylenebis(Phenylisocyanate)

951

O

O

N

N

N-Methylaniline N

CH3

7248_A001A.fm Page 447 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

447

TABLE A.1 (CONTINUED) No. 524

Solvent Name Methylene Dichloride

Methylene Diiodide

I 888

Dispersion Polarity

Hydrogen Molar Bonding Volume

Dichloro-methane

18.2

6.3

6.1

63.9

Diiodo-methane

17.8

3.9

5.5

80.5

Benzo[1,3]dioxole

19.0

6.7

5.9

114.8

Morpholine

18.8

4.9

9.2

87.1

Acetate morpholin-4ium;

17.2

20.3

18.4

Tetramethyl-thiourea

17.3

6.0

10.5

132.2

17.9

0.7

1.8

181.8

Cl

Cl 525

Autonom/ ACD Name

I

1,2-Methylenedioxybenzene

O O 526

Morpholine

O N 1193 Morpholine/Acetic Acid O O +

N H2

844

O

N,N,N,N-Tetramethylthiourea S

H3C

CH3

N

N

CH3

CH3

529

Naphtha.High-Flash

530

Naphthalene

Naphthalene

19.2

2.0

5.9

111.5

896

1-Naphthol

Naphthalen-1-ol

19.7

6.3

12.3

131.7

OH

7248_A001A.fm Page 448 Wednesday, May 23, 2007 12:38 PM

448

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Solvent Name

1198 Nicotine H3C

N

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

3-((S)-1-Methylpyrrolidin-2-yl)pyridine

18.5

7.8

6.5

160.7

1-Methyl-4-nitrobenzene

20.1

9.6

3.9

98.5

2-Nitro-propan-1-ol

16.1

13.7

15.4

88.8

3-Nitro-phenylamine

21.2

18.7

10.3

96.1

Nitro-benzene

20.0

8.6

4.1

102.7

1-Chloro-4-nitrobenzene

20.0

8.8

3.9

121.2

Nitro-ethane

16.0

15.5

4.5

71.5

Nitro-ethene

16.3

16.6

5.0

59.9

N

784

p-Nitro Toluene CH3

+

O

N

O

1092 2-Nitro-1-Propanol O

O

+

N

OH

H3C

939

3-Nitroaniline NH2

N* O

531

O

Nitrobenzene

+

O

973

N

O

4-Nitrochlorobenzene Cl

+

O

532

N

O

Nitroethane

H3C +

O 533

N

O

Nitroethylene

H2C +

O

N

O

7248_A001A.fm Page 449 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

449

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1112 Nitroglycerin (Glyceryl Trinitrate) O

O

Hydrogen Molar Bonding Volume

1,2,3-Tris-nitrooxypropane

16.2

17.8

5.9

142.5

Nitromethane

15.8

18.8

5.1

54.3

4-Nitro-phenol

20.4

20.9

15.1

93.1

1-Nitro-propane

16.6

12.3

5.5

88.4

2-Nitro-propane

16.2

12.1

4.1

86.9

Nitroso-benzene

20.0

12.7

4.0

89.3

2-Nitro-thiophene

19.7

16.2

8.2

94.6

Nonane

15.7

0

Nonane-1,9-diol

15.7

7.0

+

+

O

Dispersion Polarity

N

O

O O

N

O

O

+

N

O

534

Nitromethane

CH3 +

N

O 936

O

4-Nitrophenol OH

+

O

535

N

O

1-Nitropropane CH3

+

N

O

536

O

2-Nitropropane

CH3

H3C +

O

N

O

1073 Nitrosobenzene

N

984

2-Nitrothiophene

S 537

O

+

N

O

O

Nonane

H3C

179.7

15.1

170.5

CH3

1095 1,9-Nonanediol HO

0

OH

7248_A001A.fm Page 450 Wednesday, May 23, 2007 12:38 PM

450

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 932

Autonom/ ACD Name

Solvent Name 1-Nonanol H3C

Hydrogen Molar Bonding Volume

Nonan-1-ol

16.0

4.8

10.6

174.4

Non-1-ene

15.4

1.0

2.2

170.5

4-Nonyl-phenol

16.5

4.1

9.2

231.0

2-(4-Nonyl-phenoxy)ethanol

16.7

10.2

8.4

275.0

2-Methylamino-1phenyl-propan-1-ol

18.0

10.7

24.1

141.9

Octadecane

16.4

0

0

326.9

Octane

15.5

0

0

163.5

Octanoic acid

15.1

3.3

8.2

159.0

Octan-1-ol

16.0

5.0

11.9

157.7

Octan-2-ol

16.1

4.9

11.0

159.1

Oct-1-ene

15.3

1.0

2.4

158.0

OH

1070 1-Nonene

CH2

H3C 538

Dispersion Polarity

Nonyl Phenol HO CH3

539

Nonyl Phenoxy Ethanol

1209 Norephedrin OH CH3 H3C

923

N

Octadecane CH3

H3C

540

Octane

CH3

H3C 541

Octanoic Acid

O OH

H3C 542

1-Octanol*

OH

H3C 543

2-Octanol

CH3

H3C OH 544

1-Octene

HH33CC

CH3

CH2

7248_A001A.fm Page 451 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

451

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1065 Octyl Acetate

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetic acid octyl ester

15.8

2.9

5.1

196.0

(Z)-Octadec-9-enoic acid

16.0

2.8

6.2

317.0

(Z)-Octadec-9-en-1-ol

14.3

2.6

8.0

316.0

Oxalic acid

17.0

14.3

22.0

47.4

Oxalyl dichloride

16.1

3.8

7.5

85.8

Oxetane

18.0

9.1

5.4

65.0

2-(2-Oxo-pyrrolidin-1yl)-acetamide

17.5

15.6

11.2

116.6

19.8

4.2

0

24.0

17.8

10.5

13.9

151.2

O

545

CH3

O

H3C

Oleic Acid* O HO

CH3

546

Oleyl Alcohol HO

CH3

1156 Oxalic Acid O OH

HO O

547

Oxalylchloride O Cl

Cl O

1026 Oxetane (Trimethylene Oxide)

O 1210 2-Oxopyrrolidinacetamid

O N

O NH2

1169 Ozone +

O

O

O

1211 Paracetamol

N-(4-Hydroxy-phenyl)acetamide

HO

CH3 N

O

7248_A001A.fm Page 452 Wednesday, May 23, 2007 12:38 PM

452

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 978

Solvent Name Paraldehyde

2,4,6-Trimethyl[1,3,5]trioxane

CH3 O

Autonom/ ACD Name

Hydrogen Molar Bonding Volume

16.6

7.5

7.3

132.4

19.5

8.8

5.9

205.7

1,1,1,2,2-Pentachloroethane

18.2

3.2

2.4

120.9

1,1,2,2,3-Pentachlorocyclopropane

18.5

10.5

3.7

128.5

2,3,4,5,6-Pentachlorophenol

21.5

6.9

12.8

134.7

(E)-Penta-1,3-diene

14.7

2.5

5.0

101.7

1-Pentafluorophe ylethanone

19.3

8.1

5.4

185.5

Tetrahydro-thiopyran

18.5

6.3

8.9

103.6

Pentanal

15.7

9.4

5.8

106.4

O O

H3C

CH3

1114 Paraquat 1,1'-Dimethyl-4,4'Cl

H3C

Dispersion Polarity

bipyridinium dichloride

Cl

+

+

N

N

CH3

1161 Pentachloro Ethane

Cl

Cl

Cl

Cl

Cl 818

Pentachlorocyclopropane Cl Cl

Cl

Cl

Cl

788

Pentachlorophenol OH Cl

Cl Cl

Cl Cl

548

1,3-Pentadiene (Trans)

CH3

H2C 821

Pentafluorobenzophenon F

O

F

CH3 F

F F

549

Pentamethylene Sulfid

S 1061 Pentanal (Valeraldehyde)

H3C

O

7248_A001A.fm Page 453 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

453

TABLE A.1 (CONTINUED) No. 550

Pentane

H3C 551

Autonom/ ACD Name

Solvent Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Pentane

14.5

0

0

116.2

Pentane-2,4-dione

17.1

9.0

4.1

103.1

Pentanoic acid

15.0

4.1

10.3

109.2

Pentan-1-ol

15.9

5.9

13.9

108.6

Pentan-2-ol

15.6

6.4

13.3

109.6

Pent-4-enal

15.5

8.1

6.8

98.7

Pent-1-ene

13.9

1.4

3.8

110.4

Propionic acid pentyl ester

15.8

5.2

5.7

165.3

1,1,2,2,3,4,4,5,5,6Decafluoro-3,6-bis trifluoromet ylcyclohexane

12.4

0

0

217.4

CH3

2,4-Pentanedione

O

O

CH3

H3C 1023 Pentanoic Acid

O H3C 552

OH

1-Pentanol*

OH

H3C 733

2-Pentanol

OH CH3

H3C 553

4-Pentenal

O

H2C 554

1-Pentene

CH2

H3C 1005 n-Pentyl Propionate

O

H3C

CH3

O 556

Perfluoro Dimet ylcyclohexane F F

F

F F F

F

F

FF F F

F

F F

F

7248_A001A.fm Page 454 Wednesday, May 23, 2007 12:38 PM

454

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 555

557

Perfluoro Et ylene (Tetrafluoro Ethylene)

F

F

F

F

Perfluoroheptan F F F

558

F

F

F

F

F

F

F

F F FF

F F

F

F

Dispersion Polarity

Hydrogen Molar Bonding Volume

1,1,2,2-Tetrafluoro ethene

15.1

0

0

65.8

1,1,1,2,2,3,3,4,4,5,5,6,6, 7,7,7-Hexadecafluoro heptane

12.0

0

0

227.3

1,1,2,2,3,3,4,4,5,5,6Undecafluoro-6 trifluoromet ylcyclohexane

12.4

0

0

196.0

Phenanthrene-9,10dione

20.3

17.1

4.8

148.2

Ethoxy-benzene

18.4

4.5

4.0

127.3

Phenol

18.0

5.9

14.9

87.5

4-[1-(4Hydroxyphenyl)-1methylethyl]phenol

19.2

5.9

13.8

207.5

4-[(4Hydroxyphenyl)sulfon yl]phenol

20.0

14.6

16.3

183.2

F

F F F F

9,10-Phenanthrenequinone O

O

Phenetole (Ethyl Phenyl Ether)

H3C

559

F

F

F

F

880

F

Perfluoromet ylcyclohexane F F

975

Autonom/ ACD Name

Solvent Name

O

Phenol OH

1100 Bisphenol A OH

H3C

OH CH3

1155 Bisphenol-S OH

O

S O

OH

7248_A001A.fm Page 455 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

455

TABLE A.1 (CONTINUED) No. 560

2-Phenoxy Ethanol

Bis-(m-Phenoxyphenyl) Ether

Hydrogen Molar Bonding Volume

2-Phenoxy-ethanol

17.8

5.7

14.3

124.7

1-Phenoxy-3-(3phenoxyphenoxy)benz ene

19.6

3.1

5.1

373.0

Acetic acid phen yl ester 19.8

5.2

6.4

126.9

O

O

O

881

Dispersion Polarity

OH

O 561

Autonom/ ACD Name

Solvent Name

Phenyl Acetate

O H3C 851

O

Phenyl Acetonitrile

Phenyl-acetonitrile

19.5

12.3

3.8

114.9

Ethynyl-benzene

18.8

2.8

4.0

109.1

2-Phenyl-ethanol

19.0

5.8

12.8

120.0

Phenyl-hydrazine

20.4

6.5

14.0

98.5

Carbonyl dichloride

16.4

5.3

5.3

71.7

N 1168 Phenyl Acetylene CH

1074 2-Phenyl Ethanol

OH

1047 Phenylhydrazine

N

562

Phosgene

Cl

Cl O

NH2

7248_A001A.fm Page 456 Wednesday, May 23, 2007 12:38 PM

456

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Solvent Name

1216 Phosphoric Acid

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Phosphoric acid

14.7

18.6

26.8

52.8

Phosphoric trichloride

18.1

9.3

0

93.2

Phosphorous trichloride

18.4

3.6

0

87.3

Isobenzofuran-1,3-dione 20.6

20.1

10.1

96.8

19.2

7.0

6.0

130.0

15.6

3.0

9.8

155.0

(1S,5S)-2,6,6-Trimethyl- 16.9 bicyclo[3.1.1]hept-2ene

1.8

3.1

159.5

Piperazine

5.6

8.0

97.9

O HO

P

OH

OH 1171 Phosphorous Oxychloride (Phosphoryl Trichloride) O P

Cl

Cl

Cl

563

Phosphorus Trichloride

Cl

Cl

P Cl

564

Phthalic Anhydride O O O

1089 Picric Acid (2,4,6-Trinitrophenol) O +

O

Picric acid

O

OH

+

N

N

O

+

O

N

O

565

Pine Oil

900

2-Pinene (dl) H3C CH3 H

H

CH3

1052 Piperazine

N N

18.1

7248_A001A.fm Page 457 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

457

TABLE A.1 (CONTINUED) No.

Solvent Name

1050 Piperidine

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Piperidine

17.6

4.5

8.9

98.9

Propa-1,2-diene

15.3

3.0

6.8

60.1

Propane

13.4

0

0

89.5

[1,2]Oxathiolane 2,2dioxide

18.4

16.0

9.0

87.7

Propane-1,3-diol

16.8

13.5

23.2

72.5

Propane-2-thiol

16.3

6.8

6.5

94.1

Propane-1-thiol

16.1

5.8

5.7

90.5

Propan-1-ol

16.0

6.8

17.4

75.2

Propan-2-ol

15.8

6.1

16.4

76.8

Acetic acid prop-2-yn yl ester

16.3

5.2

8.3

98.8

N 566

1,2-Propadiene (Allene)

CH2

H2C 1064 Propane

CH3

H3C 943

1,3-Propane Sultone

O O

O S

1038 1,3-Propanediol (Trimethyleneglycol)

OH

HO 567

2-Propanethiol

SH CH3

H3C 568

1-Propanethiol

SH

H3C 569

1-Propanol

OH

H3C 570

2-Propanol

OH H3C

CH3

1036 Propargyl Acetate

O

HC O

CH3

7248_A001A.fm Page 458 Wednesday, May 23, 2007 12:38 PM

458

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 571

Autonom/ ACD Name

Solvent Name Propargylaldehyde

Dispersion Polarity

Hydrogen Molar Bonding Volume

Propynal

16.2

11.9

8.7

60.0

Oxetan-2-one

19.7

18.2

10.3

65.5

Propionaldehyde

15.3

11.1

6.9

73.4

Oxirane-2-carbaldehyde

17.5

13.4

9.8

63.2

Propionamide

16.7

9.8

11.5

78.9

Propionic acid

14.7

5.3

12.4

75.0

Propanoic anhydride

16.2

10.0

8.7

129.4

Propionitrile

15.3

14.3

5.5

70.9

Propionyl chloride

16.1

10.3

5.3

86.9

HC O 572

beta-Propiolactone

O O 1062 Propionaldehyde*

O

H3C 574

Propionaldehyde-2,3-Epoxy*

O O 575

Propionamide

NH2

H3C

O 576

Propionic Acid

OH

H3C

O 1043 Propionic Anhydride

O

H3C

O

O 577

Propionitrile

H3C 578

CH3

N

Propionylchloride

Cl

H3C O

7248_A001A.fm Page 459 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

459

TABLE A.1 (CONTINUED) No. 579

Solvent Name n-Propyl Acetate

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetic acid prop yl ester

15.3

4.3

7.6

115.3

Propylamine

16.9

4.9

8.6

83.0

1-Chloro-propane

16.0

7.8

2.0

88.1

Formic acid prop yl ester 15.5

7.1

8.6

97.9

2-Methyl-acrylic acid propyl ester

15.5

6.3

6.6

158.8

1-Nitrooxy-propane

15.8

11.3

4.1

99.7

Propene

15.1

1.6

1.5

68.8

4-Methyl-[1,3]dioxolan- 20.0 2-one

18.0

4.1

85.0

9.8

15.3

85.4

CH3

O

H3C

Autonom/ ACD Name

O 580

Propyl Amine

NH2

H3C 581

Propyl Chloride

Cl

H3C 926

Propyl Formate

O

H3C 582

O

Propyl Methacrylate CH3 O

H2C

CH3

O

935

n-Propyl Nitrate

O +

N

O 583

Propylene

CH3

H2C 584

CH3

O

Propylene Carbonate O O

O CH3

746

Propylene Chlorohydrin

OH Cl

CH3

1-Chloro-propan-2-ol

16.8

7248_A001A.fm Page 460 Wednesday, May 23, 2007 12:38 PM

460

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 585

Autonom/ ACD Name

Solvent Name Propylene Glycol

Dispersion Polarity

Hydrogen Molar Bonding Volume

Propane-1,2-diol

16.8

9.4

23.3

73.6

1-tert-Butoxy-propan2-ol

15.3

6.1

10.8

151.6

1-Butoxy-propan-2-ol

15.3

4.5

9.2

132.0

1-Ethoxy-propan-2-ol

15.7

6.5

10.5

115.6

Acetic acid 2-ethoxy-1methyl-ethyl ester

15.6

4.3

9.0

155.1

1-Isobutoxy-propan-2-ol 15.1

4.7

9.8

132.2

1-Isopropoxy-propan-2ol

15.5

6.1

11.0

134.6

1-Methoxy-propan-2-ol

15.6

6.3

11.6

93.8

OH H3C OH

586

Propylene Glycol Mono-t-Butyl Ether CH3

CH3

O

HO

CH3

H3C

587

Propylene Glycol Monob utyl Ether

HO

O

CH3

CH3 588

Propylene Glycol Monoeth yl Ether

HO

O

CH3

CH3 589

Propylene Glycol Monoeth yl Ether Acetate

H3C

O O

590

O CH3

Propylene Glycol Monoisob utyl Ether

HO

CH3

O CH3

857

CH3

CH3

Propylene Glycol Monoisoprop yl Ether CH3 HO

O

CH3

CH3

591

Propylene Glycol Monometh yl Ether

HO

O CH3

CH3

7248_A001A.fm Page 461 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

461

TABLE A.1 (CONTINUED) No. 592

Propylene Glycol Monometh yl Ether Acetate

H3C

O

O

O 593

Autonom/ ACD Name

Solvent Name

Hydrogen Molar Bonding Volume

Acetic acid 2-methoxy1-methyl-ethyl ester

15.6

5.6

9.8

137.1

1-Phenoxy-propan-2-ol

17.4

5.3

11.5

143.2

1-Propoxy-propan-2-ol

15.8

7.0

9.2

130.3

2-Methyl-aziridine

18.1

8.4

6.5

71.2

2-Methyl-oxirane

15.2

8.6

6.7

67.6

Prop-2-yn-1-ol

16.1

8.8

19.1

Propynenitrile

15.5

17.0

6.3

62.5

(R)-2-Isopropylidene-5methyl-cyclohexanone

17.5

8.9

5.5

162.9

1H-Pyrazole

20.2

10.4

12.4

68.1

CH3

CH3

Propylene Glycol Monophen yl Ether

HO

Dispersion Polarity

O CH3

594

Propylene Glycol Monoprop yl Ether

HO

CH3

O CH3

1086 Propylene Imine (2-Methyl Aziridine)

N

H3C 595

Propylene Oxide

O CH3 1091 2-Propyn-1-ol

57.7

HC OH 596

Propynonitrile

N

HC 1133 Pulegone O H3C

CH3

H3C

985

Pyrazole

N

N

7248_A001A.fm Page 462 Wednesday, May 23, 2007 12:38 PM

462

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 597

Solvent Name Pyridazine

Dispersion Polarity

Hydrogen Molar Bonding Volume

Pyridazine

20.2

17.4

11.7

72.6

Pyridine

19.0

8.8

5.9

80.9

Benzene-1,2,3-triol

20.7

10.7

21.1

87.0

1H-Pyrrole

19.2

7.4

6.7

69.2

Pyrrolidine

17.9

6.5

7.4

83.5

Pyrrolidin-2-one

19.4

17.4

11.3

76.4

2-Oxo-propionitrile

15.9

18.9

8.0

70.9

(R)-(6-Methoxyquinolin-4-yl)((2S,4S,5R)-5-vinyl-1aza-bicyclo[2.2.2]oct2-yl)-methanol

19.0

6.6

11.0

310.7

N

N 598

Autonom/ ACD Name

Pyridine

N 901

Pyrogallol (1,2,3-Trihydroxybenzene) OH OH OH

600

Pyrrole

N 1051 Pyrrolidine

N

599

2-Pyrrolidone O N

601

Pyruvonitrile

O N H3C 1202 Quinine CH2 H N

HO H3C

H

O N

7248_A001A.fm Page 463 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

463

TABLE A.1 (CONTINUED) No. 602

Autonom/ ACD Name

Solvent Name Quinoline

Dispersion Polarity

Hydrogen Molar Bonding Volume

Quinoline

19.8

5.6

5.7

118.0

1,1-Dioxo-1,2-dihydro1lambda*6*benzo[d]isothiazol-3one

21.0

13.9

8.8

206.8

2-Hydroxybenzaldehyde

19.4

10.7

14.7

104.6

2-Hydroxy-benzoic acid 19.4

10.1

17.4

95.7

N 1196 Saccharin O N S

O

O

704

Salicylaldehyde O

OH

1204 Salicylic Acid OH

OH

O

1225 Sebacic Acid

Decanedioic acid

17.1

7.1

11.7

167.6

3-(2-Amino-ethyl)-1Hindol-5-ol

18.0

8.2

14.4

144.4

4-((E)-3-Hydroxypropenyl)-2,6dimethoxy-phenol

19.2

7.3

16.1

194.6

3-Methyl-1H-indole

20.0

7.1

6.2

122.6

O O

HO OH

1178 Serotonin NH2 HO N

1101 Sinapyl Alcohol OH

H3C

O HO O

832

CH3

Skatole CH3

N

7248_A001A.fm Page 464 Wednesday, May 23, 2007 12:38 PM

464

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1207 Spermidin H2 N

603

NH2

N

Stearic Acid

Dispersion Polarity

Hydrogen Molar Bonding Volume

N*1*-(3-Aminopropyl)-butane-1,4diamine

16.7

11.2

12.0

155.6

Octadecanoic acid

16.3

3.3

5.5

326.0

Vinyl-benzene

18.6

1.0

4.1

115.6

2-Phenyl-oxirane

19.4

5.8

6.6

114.2

Succinaldehyde

16.8

9.8

10.5

81.2

Dihydro-furan-2,5-dione 18.6

19.2

16.6

66.8

Succinonitrile

17.9

16.2

7.9

81.2

4-Aminobenzenesulfonamide

20.0

19.5

10.7

159.5

Tetrahydro-thiophene 1,1-dioxide

20.3

18.2

10.9

95.7

OH

H3C O

604

Styrene

CH2

1014 Styrene Oxide (Phen yl Oxirane) O

605

Succinaldehyde (Butanedial)

O 606

O

Succinic Anhydride

O

O

607

O

Succinonitrile

N N 1077 Sulfanilamide

NH2 O

S

NH2

O 608

Sulfolane

O

S

O

7248_A001A.fm Page 465 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

465

TABLE A.1 (CONTINUED) No. 609

Solvent Name Sulfur Dicyanide

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Cyanic thiocyanate

18.1

13.5

0

60.0

Oxosulfane oxide

15.8

8.4

10.0

44.0

Sulfuryl dichloride

17.6

7.2

0

81.0

Terephthalic acid

18.8

10.7

166.0

2-tert-Butyl-4-methylphenol

17.3

3.7

10.5

177.6

2,6-Dibromo-4-[1-(3,5dibromo-4hydroxyphenyl)-1methylethyl]phenol

20.2

9.1

13.8

249.5

1,1,2,2-Tetrabromoethane

22.6

5.1

8.2

116.8

etrachlorobenzene

21.2

10.7

3.4

116.2

S N

N 610

Sulfur Dioxide

S

O

O

1098 Sulfuryl Chloride

O Cl

S

Cl

O 897

Terephthalic Acid

6.1

OH

O

HO

O

1079 2-tert-Butyl-4-Methyl Phenol OH

CH3 CH3 CH3

CH3

1127 Tetrabromo Bisphenol A OH Br

Br

Br H3C

OH CH3

612

708

Br

1,1,2,2-Tetrabromoethane

Br

Br

Br

Br

1,2,4,5-Tetrachlorobenzene 1,2,4,5-T Cl Cl Cl Cl

7248_A001A.fm Page 466 Wednesday, May 23, 2007 12:38 PM

466

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 783

2,2,6,6-Tetrachlorocyclohexanone

882

O

Cl

Cl

614

Autonom/ ACD Name

Solvent Name

Cl

Dispersion Polarity

Hydrogen Molar Bonding Volume

2,2,6,6-Tetrachlorocyclohexanone

19.5

14.0

6.3

138.8

1,1,2,2-Tetrachloroethane

18.8

5.1

5.3

105.2

1,1,1,2-Tetrachloroethane

18.0

4.4

4.2

105.0

1,1,2,2-Tetrachloroethene

18.3

5.7

0

101.2

1,1,2,2-Tetrachloropropane

17.9

6.7

3.3

123.7

Tetradecane

16.2

0

0

261.3

Tetradec-1-ene

16.1

0.5

1.9

253.4

13.9

4.3

0.6

224.0

16.8

5.7

8.0

81.7

Cl

1,1,2,2-Tetrachloroethane

Cl

Cl

Cl

Cl

1,1,1,2-Tetrachloroethane

Cl Cl

Cl

Cl 615

613

Tetrachloroethylene

Cl

Cl

Cl

Cl

1,1,2,2-Tetrachloropropane Cl Cl Cl

Cl

922

CH3

n-Tetradecane H3C

CH3

1011 1-Tetradecene CH2

H3C

616

Tetraethylorthosilicate CH3 O O H3C

Si O

CH3

O CH3

617

Tetrahydrofuran

O

Tetrahydro-furan

7248_A001A.fm Page 467 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

467

TABLE A.1 (CONTINUED) No. 999

Solvent Name 2,5-Tetrahydrofuran Dimethanol

HO

OH

O

1096 Tetrahydrofurfuryl Alcohol

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

17.9

8.5

18.9

114.5

(Tetrahydro-furan-2-yl)- 17.8 methanol

8.2

10.2

97.0

(5-Hydroxymethyltetrahydro-furan-2-yl)methanol

OH O 618

Tetrahydronaphthalene

1,2,3,4-Tetrahydronaphthalene

19.6

2.0

2.9

136.0

619

Tetrahydropyran

Tetrahydro-pyran

16.4

6.3

6.0

97.8

Tetrahydro-thiopyran

18.5

6.3

8.9

103.6

1,1,3,3-Tetramethoxypropane

15.0

7.1

6.8

164.7

1,2,3,4-Tetramethylbenzene

18.8

0.5

0.5

148.3

1,2,3,5-Tetramethylbenzene

18.6

0.5

0.5

150.8

O

620

Tetrahydrothiapyran

S

895

1,1,3,3-Tetramethoxypropane

H3C

O

H3C 884

O O

O

CH3

CH3

1,2,3,4-Tetramethylbenzene CH3 CH3 CH3 CH3

885

1,2,3,5-Tetramethylbenzene CH3

CH3 H3C

CH3

7248_A001A.fm Page 468 Wednesday, May 23, 2007 12:38 PM

468

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 622

Solvent Name Tetramethylene Sulfid

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Tetrahydro-thiophene

18.6

6.7

9.1

88.3

Tetrahydro-thiophene 1,1-dioxide

20.3

18.2

10.9

95.7

Tetrahydro-thiophene 1oxide

18.2

11.0

9.1

90.0

Tetramethyl-urea

16.7

8.2

11.0

120.4

Tetranitro-methane

15.5

9.9

7.5

119.7

[2,2']Bi[[1,3]dithiolylide 21.0 ne]

8.2

8.4

204.0

Methylsulfanyl-ethane

16.2

5.9

5.3

90.4

Thiirane

19.3

9.1

5.0

58.0

Thiazole

20.5

18.8

10.8

70.9

S

790

Tetramethylene Sulfone (Sulfolane)

O

623

O

S

Tetramethylene Sulfoxide O S

624

Tetramethylurea

H3C

CH3

CH3

N

N

CH3

O

933

Tetranitromethane O

O

+

N

+

890

625

N +

N

O

O

O

2,2',5,5'-Tetrathiafulvalen

S

S

S

S

2-Thiabutane

S

H3C 626

O

+

N O

O

CH3

Thiacyclopropane

S 627

Thiazole*

N S

7248_A001A.fm Page 469 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

469

TABLE A.1 (CONTINUED) No. 628

Autonom/ ACD Name

Solvent Name Thioacetamide

Dispersion Polarity

Hydrogen Molar Bonding Volume

Thioacetamide

17.5

20.6

20.2

75.0

Thioacetic acid

17.0

6.7

8.9

71.5

Dihydro-thiophen-2-one 19.0

6.9

6.2

86.6

Thiocyanic acid

16.8

8.9

10.9

51.7

2-(2-Hydroxyethylsulfanyl)-ethanol

17.3

8.8

19.8

103.6

Mercapto-acetic acid

16.0

8.6

14.8

69.5

Thionyl dichloride

16.9

6.2

5.9

79.0

Thiophene

18.9

2.4

7.8

79.0

Benzenethiol

20.0

4.5

10.3

102.4

S NH2

H3C 629

Thioacetic Acid

O SH

H3C 630

Gamma-Thiobutyrolactone O S

631

Thiocyanic Acid

N

SH

1034 Thiodiethylenglycol

HO

S

OH

1017 Thioglycolic Acid (Mercapto Acetic Acid)

O SH

HO 632

Thionyl Chloride

Cl

S

Cl

O 633

Thiophene

S 703

Thiophenol SH

7248_A001A.fm Page 470 Wednesday, May 23, 2007 12:38 PM

470

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 634

Solvent Name Thiourea

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

Thiourea

20.0

21.7

14.8

72.8

[1,4]Oxathiane

19.0

6.6

7.7

93.5

2-Isopropyl-5-methylphenol

19.0

4.5

10.8

166.9

(E)-2-Methyl-but-2-enal 16.2

12.9

6.8

96.6

NH2

H2N S 635

1,4-Thioxane

S O 1203 Thymol CH3

OH CH3

H3C

636

Tigaldehyde

H3C

O CH3

637

Toluene

Toluene

18.0

1.4

2.0

106.8

o-Tolylamine

19.4

5.8

9.4

107.8

2,4-Diisocyanato-1methyl-benzene

19.3

7.9

6.1

143.5

1147 Triacetin (1,2,3-Propanetriol Triacetate) Acetic acid 2-acetoxy-1- 16.5 acetoxymethyl-ethyl CH O ester O

4.5

9.1

188.2

CH3

953

2-Toluidine NH2 CH3

638

Tolylene Diisocyanate O CH3 N

N O

3

O O H3C

O O H3C

7248_A001A.fm Page 471 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

471

TABLE A.1 (CONTINUED) No.

Solvent Name

1137 Tri-n-Butyl Acetyl Citrate CH3

CH3

O O

O

O O

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

3-Acetoxy-3butoxycarbonylpentanedioic acid dibutyl ester

16.7

2.5

7.4

384.3

Tri-n-butyl borate

16.7

1.8

4.6

269.7

3-Butoxycarbonyl-3hydroxy-pentanedioic acid dibutyl ester

16.6

3.8

10.1

345.5

Phosphoric acid tributyl ester

16.3

6.3

4.3

274.0

1H-[1,2,3]Triazole

20.7

8.8

15.0

58.2

2,4,6-Tribromo-phenol

20.6

10.2

14.1

129.7

1,1,2-Tribromo-ethene

18.3

9.4

8.0

97.8

3,3,3-Trichloro-propene

17.7

15.5

3.4

106.2

O CH3

O O

H3 C

1020 Tri-n-Butyl Borate CH3

O H3C

B

O

CH3

O

1220 Tri-n-Butyl Citrate CH3

CH3

O O

O

O O

O

OH

H3C

641

Tri Butyl Phosphate* CH3

O P

O

O

O

H3C

CH3

639

1,2,3-Triazole

N N

N

1118 2,4,6-Tribromo Phenol OH Br

Br

Br

640

Tribromo Ethylene*

Br Br 642

Br

3,3,3-Trichloro Propene

Cl

H2C Cl

Cl

7248_A001A.fm Page 472 Wednesday, May 23, 2007 12:38 PM

472

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 643

Solvent Name 1,1,2-Trichloro Propene

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

1,1,2-Trichloro-propene

17.7

15.7

3.4

104.8

1,2,3-Trichloro-propene

17.8

15.7

3.4

105.0

Trichloro-methyl-silane

16.5

6.6

3.5

117.4

Trichloro-acetic acid

18.3

5.8

11.4

100.2

Trichloro-acetonitrile

16.4

7.4

6.1

100.0

1,3,5-Trichloro-2methoxy-benzene

21.0

3.9

7.0

129.0

1,2,4-Trichloro-benzene

20.2

6.0

3.2

125.5

1,1,1-Trichloro-ethane

16.8

4.3

2.0

99.3

Cl Cl

CH3 Cl

644

1,2,3-Trichloro Propene

Cl Cl 981

Cl

Trichloro-Methyl-Silane

Cl H3C

Si Cl Cl

861

Trichloroacetic Acid O Cl

OH

Cl

645

Cl

Trichloroacetonitrile

Cl

Cl

N

Cl 1115 2,4,6-Trichloroanisole O

CH3 Cl

Cl

Cl

701

1,2,4-Trichlorobenzene* Cl Cl

Cl

647

1,1,1-Trichloroethane

Cl H3C

Cl Cl

7248_A001A.fm Page 473 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

473

TABLE A.1 (CONTINUED) No. 648

Solvent Name 1,1,2-Trichloroethane

Autonom/ ACD Name

Dispersion Polarity

Hydrogen Molar Bonding Volume

1,1,2-Trichloro-ethane

18.2

5.3

6.8

92.9

1,1,2-Trichloro-ethene

18.0

3.1

5.3

90.2

Trichloro-fluoro methane

15.3

2.0

0

92.8

2,4,6-Trichloro-phenol

20.3

5.1

10.8

132.5

1,2,3-Trichloro-propane

17.8

12.3

3.4

106.1

14.2

3.6

3.8

101.1

2,4,5-Trichlorobenzenethiol

21.0

4.5

9.1

145.0

1,1,2-Trichloro-1,2,2trifluoro-ethan

14.7

1.6

0

119.2

Cl Cl 649

Cl

Trichloroethylene

Cl Cl 650

Cl

Trichlorofluoromethane (Freon 11

F

Cl

Cl

Cl 702

2,4,6-Trichlorophenol OH Cl

Cl

Cl

883

1,2,3-Trichloropropane Cl

Cl

Cl

1174 Trichlorosilane

Cl

Si

Cl

Cl 651

2,4,5-Trichlorothiophenol SH Cl Cl Cl

652

1,1,2-Trichlorotrifluoroethane (Freon 113)

Cl

F

F

Cl

F

Cl

7248_A001A.fm Page 474 Wednesday, May 23, 2007 12:38 PM

474

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1160 Triclosan OH

Cl O

Hydrogen Molar Bonding Volume

5-Chloro-2-(2,4dichloro-phenoxy)phenol

20.0

7.7

10.0

263.0

Phosphoric acid m-tolyl ester o-tolyl ester p-tolyl ester

19.0

12.3

4.5

316.0

(1S,2S,4R,6R)-1,7,7Trimethyltricyclo[2.2.1.0*2,6*] heptane

16.9

0

0

161.4

Tridecan-1-ol

16.2

3.1

9.0

242.0

2-[Bis-(2-hydroxyethyl)-amino]-ethanol

17.3

22.4

23.3

133.2

Acetate tris-(2-hydroxy- 17.2 ethyl)-ammonium;

20.3

18.4

Cl

Cl

653

Dispersion Polarity

Tricresyl Phosphate O O

P

O

CH3

O

H3C

899

CH3

Tricyclene CH3

H3C

CH3

H

H H

654

Tridecyl Alcohol* OH

H3C

655

Triethanolamine OH

N

HO

OH

1194 Triethanolamine/Acetic Acid OH

HO H + N

O O

OH

1218 Triethyl Citrate O H3C

OH

O

O O

CH3

3-Ethoxycarbonyl-3hydroxy-pentanedioic acid diethyl ester

16.5

4.9

12.0

243.0

Triethyl-amine

17.8

0.4

1.0

138.6

2-[2-(2-Hydroxyethoxy)-ethoxy]ethanol

16.0

12.5

18.6

114.0

O O

CH3

656

Triethylamine

CH3 H3C 657

N

CH3

Triethylene Glycol

HO

O

O

OH

7248_A001A.fm Page 475 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

475

TABLE A.1 (CONTINUED) No. 856

Triethylene Glycol Monometh yl Ether O

HO

658

Autonom/ ACD Name

Solvent Name

O

O

CH3

Triethylene Glycol Monoole yl Ether* HO

O

O

O

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-[2-(2-Methoxyethoxy)-ethoxy]ethanol

16.2

7.6

12.5

160.0

2-(2-{2-[((Z)-Octadec9-enyl)oxy]-ethoxy}ethoxy)-ethanol

16.0

3.1

8.4

418.5

Phosphoric acid trieth yl ester

16.7

11.4

9.2

171.0

2,2,2-Trifluoro-ethano

15.4

8.3

16.4

72.3

1,2,3-Trifluoro-4-nitro benzene

19.5

7.7

3.5

122.0

Trifluoromethyl-benzen

17.5

8.8

0

122.9

Trifluoro-acetic acid

15.6

9.9

11.6

74.2

1,1,1-Trifluoro-ethan

14.6

10.7

0

64.6

CH3

659

Triethylphosphate

O

H3C

O

P

H3C 944

O

CH3

O

2,2,2-Trifluoro Ethano

F F

F 795

OH

2,3,4-Trifluoro Nitrobenzen O

+

N

O F F

F

1158 alpha,alpha,alpha Trifluoro oluene F

660

F

Trifluoroacetic Aci

F 661

F

F

O

F

OH

1,1,1-Trifluoroethan

F H3C

F F

7248_A001A.fm Page 476 Wednesday, May 23, 2007 12:38 PM

476

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 662

Autonom/ ACD Name

Solvent Name Trifluoromethane (Freon 23

F

Dispersion Polarity

Hydrogen Molar Bonding Volume

Trifluoro-methan

14.4

8.9

6.5

46.1

1-(4-Trifluoromet ylphenyl)-ethanone

18.8

6.1

3.5

151.7

1,3-Bis-trifluoromethyl benzene

17.0

6.8

0

155.2

Benzene-1,2,4tricarboxylic acid tris(7-methyl-octyl) ester

16.6

5.7

2.2

602.9

Benzene-1,2,4tricarboxylic acid tris(6-methyl-heptyl) ester

16.6

6.0

2.5

553.1

Trimethyl-amine

14.6

3.4

1.8

90.3

Isobutyric acid 3hydroxy-2,2,4trimethyl-pentyl ester

15.1

6.1

9.8

227.4

1,2,4-Trimethyl-benzene 18.0

1.0

1.0

137.3

F

F 847

4-(Trifluoromet yl) Acetophenone CH3

O

F

F

F

1157 1,3-Bis(Trifluoromethyl)Benzen F

F

F

F F

663

F

Triisononyl Trimellilate CH3 H3C

O O

CH3

O

O

CH3

O

O

H3C CH3

664

Triisooctyl Trimellitate H3C

O

CH3

O

O

H3C

665

CH3 O

O

CH3

CH3

O

Trimethyl Amine

CH3 H3C 666

N

CH3

2,2,4-Trimethyl-1,3-Pentanediol Monoisobutyrate CH3

O

H3C OH

667

CH3

Trimethylbenzene* CH3 CH3

CH3

CH3

CH3

CH3 O

7248_A001A.fm Page 477 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

477

TABLE A.1 (CONTINUED) No. 669

Autonom/ ACD Name

Solvent Name Trimethylenesulfid

Dispersion Polarity

Hydrogen Molar Bonding Volume

Thietane

18.8

7.8

9.4

72.8

2,2,4-Trimethyl-pentane

14.1

0

0

166.1

Phosphoric acid trimethyl ester

16.7

15.9

10.2

115.8

Trinitro-methane

15.5

10.3

7.3

94.6

2-Methyl-1,3,5-trinitrobenzene

19.5

3.7

4.5

137.6

2-Methyl-1,3,5-trinitrobenzene

19.5

10.0

4.5

137.6

Phosphoric acid trioctyl ester

16.2

5.9

4.2

469.8

[1,3,5]Trioxane

18.7

9.2

8.6

77.0

S 670

2,2,4-Trimethylpentane

CH3 H3C 671

CH3

CH3

Trimethylphosphate O

H3C O

CH3

P

O

O

H3C

934

CH3

Trinitomethane O

O

+

+

N

O

N

O

+

N

O

918

O

Trinitrotoluene (TNT) P by Dipole Moment O

O

CH3

+

+

N

N

O

O

+

O

919

N

O

Trinitrotoluene (TNT) P by Group Cont. O

O

CH3

+

O

+

N

N

O

+

O

N

O

1151 Trioctylphosphate H3C

O O P

O

O

H3C

855

CH3

1,3,5-Trioxane

O O

O

7248_A001A.fm Page 478 Wednesday, May 23, 2007 12:38 PM

478

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No.

Autonom/ ACD Name

Solvent Name

1150 Triphenyl Phosphate

Dispersion Polarity

Hydrogen Molar Bonding Volume

Phosphoric acid triphenyl ester

20.1

6.4

6.8

271.9

1-[2-(2-Methoxy-1methyl-ethoxy)-1methyl-ethoxy]propan-2-ol

15.3

5.5

10.4

214.0

Undecane

16.0

0

0

212.7

Urea

20.9

18.7

26.4

45.8

Pentanenitrile

15.3

11.0

4.8

103.8

4-Hydroxy-3-methoxybenzaldehyde

19.4

9.8

11.2

144.1

(2-Chloro-ethoxy)ethene

16.3

6.7

5.8

101.7

3-Vinyloxymethylheptane

15.6

3.4

4.2

194.2

(2-Methoxy-ethoxy)ethene

15.9

6.7

6.8

96.2

O O P

O

O

672

Tripropylene Glycol Monometh yl Ether* CH3 HO

O

O CH3

CH3

O CH3

1059 Undecane H3C

860

CH3

Urea (Ro = 19.4)

H2N

NH2

O 673

Valeronitrile H3C N

1162 Vanillin (4-Hydroxy-3-Methoxy Benzaldehyde) OH

CH3 O

O

674

Vinyl 2-Chloro Eth yl Ether

H2C 908

O

Cl

Vinyl 2-Ethyl-Hexyl Ether

H2C

CH3

O

H3C 675

Vinyl 2-Methoxy Eth yl Ether

H2C

O

O

CH3

7248_A001A.fm Page 479 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

479

TABLE A.1 (CONTINUED) No. 676

Autonom/ ACD Name

Solvent Name Vinyl Acetate

Dispersion Polarity

Hydrogen Molar Bonding Volume

Acetic acid vin yl ester

16.0

7.2

5.9

92.6

But-3-enoic acid

16.8

5.2

12.3

85.3

But-1-en-3-yne

15.1

1.7

12.0

74.3

3-Vinyloxy-propene

14.9

6.5

5.3

105.4

Vinylamine

15.7

7.2

11.8

51.8

Bromo-ethene

15.9

6.3

5.4

71.6

1-[2-(2-Vinyloxy16.0 ethoxy)-ethoxy]-butane

5.0

6.0

206.2

1-Vinyloxy-butane

15.2

4.1

5.1

129.4

1-Vinylsulfanyl-butane

16.0

5.0

5.4

136.7

Butyric acid vin yl ester

15.6

3.9

6.9

126.5

O

H3C 677

CH2

O

Vinyl Acetic Acid

CH2

O OH

678

Vinyl Acetylene

HC CH2 679

Vinyl Allyl Ether

H2C 742

Vinyl Amine

H2C 680

O

CH3

S

Vinyl Butyrate

O H3C

CH2

CH3

O

Vinyl Butyl Sulfid

H2C 682

O

O

Vinyl Butyl Ether

H2C 681

Br

Vinyl Butyl Carbitol H3C

964

NH2

Vinyl Bromide

H2C 914

CH2

O

O

CH2

7248_A001A.fm Page 480 Wednesday, May 23, 2007 12:38 PM

480

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 906

Vinyl Carbitol

O

HO 683

Vinyl Crotonate

O

H3C

O

O

O

1024 Vinyl Ethyl Ether

H2C 686

687

O

7.5

13.8

129.2

Chloro-ethene

16.0

6.5

2.4

68.7

(E)-But-2-enoic acid vinyl ester

15.9

5.0

9.0

118.8

Vinyloxy-ethene

14.8

4.2

5.8

90.7

[2-(2-Ethoxy-ethoxy)ethoxy]-ethene

15.9

6.0

6.6

171.9

S

CH2

Ethoxy-ethene

14.5

4.9

6.0

95.0

Ethylsulfanyl-ethene

16.4

5.8

6.3

101.3

Formic acid vin yl ester

15.3

6.5

9.7

74.7

Iodo-ethene

17.1

5.5

7.3

75.6

2-Methyl-1-vinyloxypropane

15.1

4.1

5.1

131.3

CH3

Vinyl Ethyl Sulfid

H2C

16.3

CH2

O

Vinyl Ethyl Carbitol H3C

2-(2-Vinyloxy-ethoxy)ethanol

CH2

Vinyl Ether

H2C 907

Hydrogen Molar Bonding Volume

Cl

O

685

Dispersion Polarity

CH2

O

Vinyl Chloride

H2C 684

Autonom/ ACD Name

Solvent Name

CH3

Vinyl Formate

O

CH2

O 688

Vinyl Iodide (Iodoethene)

H2C 909

I

Vinyl Isobutyl Ether

H2C

O

CH3 CH3

7248_A001A.fm Page 481 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

481

TABLE A.1 (CONTINUED) No. 910

Autonom/ ACD Name

Solvent Name Vinyl Isopropyl Ether

Dispersion Polarity

Hydrogen Molar Bonding Volume

2-Vinyloxy-propane

14.7

3.5

5.2

115.3

(2-Methoxy-ethoxy)ethene

15.6

6.2

6.8

114.7

Propionic acid vin yl ester

15.6

8.0

4.7

110.1

1-Vinyloxy-propane

14.9

3.5

5.2

113.0

4-Vinyl-pyridine

18.1

7.2

6.8

107.3

1-Vinyl-pyrrolidin-2-one 16.4

9.3

5.9

106.9

1-Ethoxy-4vinylsulfanyl-butane

16.5

5.0

5.8

174.0

(2-Ethoxyethylsulfanyl)-ethene

16.4

7.0

6.0

139.7

Vinyl-silane

15.5

2.6

4.0

89.4

CH3 H2C 911

CH3

Vinyl Methyl Cellosolve

H3C 689

O

O

O

CH2

Vinyl Propionate

CH2

O

H3C O 690

Vinyl Propyl Ether

H2C 843

CH3

O

4-Vinyl Pyridine CH2

N

691

Vinyl Pyrrolidone O N

913

Vinyl S-Butyl Mercapto Butyl Ether H2C

912

S

O

CH3

Vinyl S-Ethyl Mercapto Eth yl Ether H2C

692

CH2

S

Vinyl Silane

H2C

SiH3

O

CH3

7248_A001A.fm Page 482 Wednesday, May 23, 2007 12:38 PM

482

Hansen Solubility Parameters: A User’s Handbook

TABLE A.1 (CONTINUED) No. 916

Autonom/ ACD Name

Solvent Name 2-Vinyl Toluene CH3

Dispersion Polarity

Hydrogen Molar Bonding Volume

1-Methyl-2-vinylbenzene

18.6

1.0

3.8

131.8

Trifluoro-acetic acid vinyl ester

13.9

4.3

7.6

116.4

Trimethyl-vinyl-silane

14.5

1.0

2.5

145.3

2-Ethyl-hexanoic acid vinyl ester

15.6

2.7

5.8

196.0

[1,3]Dioxol-2-one

17.3

18.1

9.6

86.0

Water

15.5

16.0

42.3

18.0

Water

15.1

20.4

16.5

18.0

Water

18.1

17.1

16.9

18.0

p-Xylene

17.6

1.0

3.1

123.3

CH2

693

Vinyl Trifluoro Acetate O F

694

CH2

O

F

F

Vinyl Trimethyl Silane

CH3 H2C 915

Si CH3

CH3

Vinyl-2-Ethyl Hexanoate O H3C

O

CH2

H3C

695

Vinylenecarbonate

O O O 696

Water

OH2 859

Water - 1% in Ro = 18.1

OH2 858

Water - Complete Miscibility Ro = 13.0

OH2 697

Xylene CH3

CH3

7248_A001A.fm Page 483 Wednesday, May 23, 2007 12:38 PM

Appendix A: Table A.1

483

TABLE A.1 (CONTINUED) No. 698

Solvent Name o-Xylene

Autonom/ ACD Name o-Xylene

CH3 CH3

Dispersion Polarity

Hydrogen Molar Bonding Volume

17.8

3.1

1.0

121.2

7248_A001A.fm Page 484 Wednesday, May 23, 2007 12:38 PM

7248_A002.fm Page 485 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2 COMMENTS TO TABLE A.2 Documentation for the sources of data used for the HSP correlations is gi ven in the following. The quality of the entries is sometimes less than desired because of the data being too fe w, too limited in scope and range of HSP , or for other reasons discussed in the te xt, such as the influence o molecular weight (molecular v olume) of the test solv ents in the given study. All entries have been included (with some apologies) as the y have some value in terms of estimating, ho wever.

POLYMERS 1–109 These polymers are listed in Reference 1 with suppliers. This report from the Scandina vian Paint and Printing Ink Research Institute updates an earlier one from 1982. The institute no longer exists. See also Reference 2.

POLYMER 110 This is an intermediate v alue for the permeation of chemicals through Challenge ® materials [3]. See also Table 13.1 and Figure 13.2. Impro ved v alues are found belo w in 141 and 142. This correlation was based on fe w data to help locate additional solv ents for testing. Results from tests with these then resulted in the correlations belo w.

POLYMERS 111–112 These are correlations of true solubilities for the DO W epoxy No volacs 438 and 444.

POLYMERS 113–114 These are correlations of the chemical resistance of coatings based on inor ganic zinc silicate and a two component epoxy produced by Hempel’ s Marine P aints. Data tak en from resistance tables.

POLYMER 115 The data are solubilities determined for PVDF with the correlation being pre viously published in [4].

POLYMER 116 Data for coal tar pitch generated for the solubility of the solids not dissolv ed in some cases where the solution w as darkened with only partial solution.

POLYMERS 117–140 Permeation correlations for chemical protecti ve clothing described in detail in Reference 5. See also Chapter 13, Table 13.1.

485

7248_A002.fm Page 486 Wednesday, May 23, 2007 12:53 PM

486

Hansen Solubility Parameters: A User’s Handbook

POLYMERS 141–142 Final permeation correlations for Challenge ® 5100 and 5200 materials. Data from Reference 3 where there is considerable discussion. See also Chapter 13, Table 13.1, and Figure 13.2.

POLYMERS 143–144 These correlations are based on which solv ents dissolve PVDC at ele vated temperatures and use data from Wessling [6]. These were additionally used to check ne w calculations for solubility parameters of the solv ents where these were lacking.

POLYMERS 145–148 These chemical resistance data for PES (ICI-V ictrex®) and PPS (Philips-Ryton ®) were based on supplier data sheets and are reported in Reference 7.

Polymers 149–160 These correlations for man y common plastics types are based on the resistance tables reported in the PLASTGUIDE (1989) published by the Danish compan y Dukadan, which no longer e xists. A single correlation for the solubility of P A 6,6 is based on its solubility only with data from Reference 8.

POLYMER 161 Beerbower treated several sets of data and made correlations of swelling and solubility (and other phenomena). This one is for polyvin yl silane.

POLYMERS 162–163 These correlations for swelling of cellophane and solubility of eth ylene vinyl alcohol copolymer are based on data generated at NIF (Scandina vian Paint and Printing Ink Research Institute).

POLYMERS 164–167 These are supplementary breakthrough time correlations for Sarane x®, Safety 4 ® 4H, and polyvinylalcohol protective gloves. See also Reference 5 and Chapter 13. Elimination of plasticizer data for the 4H glo ves improved predictability for lo wer molecular weight materials.

POLYMERS 168–181 These correlations for common polymer types are based on data in resistance tables in the Modern Plastics Encyclopedia in the 1984/1985 issue [9]. Such data are not al ways suf ficiently encom passing to allow good correlations.

POLYMER 182 Correlation based on high temperature solv ents for ECTFE.

POLYMER 183 Data for this correlation of solubility of polyacrylonitrile were tak en from the Polymer Handbook [10], Table of solv ents and nonsolv ents, p. VII/385-VII/386. See also Chapter 5, Table 5.3.

7248_A002.fm Page 487 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

487

POLYMERS 184–186 Data for this correlation are the tendency of Polyethylene imide (PEI) (GE Ultem®) to environmental stress crack (ESC) at dif ferent stress/strain le vels. These data were generated by General Electric as published in the Modern Plastics Encyclopedia 1984/1985 [9].

POLYMERS 187–224 The Handbook of Solubility Parameters and Other Cohesion Parameters [11] as well as thePolymer Handbook [12] included so-called “solvent range” data. Solvents were divided into groups of poor, moderate, and strong h ydrogen bonding, and man y experiments were run. The correlations sho w that not all the data were well tak en, b ut a reasonable indication is possible. The full Hansen solubility parameter system is not covered very well by this limited solubility data. These polymers are included in Reference 11, Table 1, on page 280. Heating samples to speed up the solution process was also done. This can easily lead to errors.

POLYMERS 225–346 These entries have the same problem as those in 187–224 in that the data are sometimes questionable and not suf ficient enough to do what has been done, i.e. co vert solv ent range data to Hansen solubility parameter spheres. These entries cover the acrylics, polyesters, polystyrenes, vin yls, and miscellaneous categories. Some cate gories are not yet included. Data on page 281-289 (T able 2) in Reference 11.

Polymer 347 These values for VYHH® (Union Carbide) were tak en from Reference 1.

POLYMER 348 This questionable correlation for PVF includes only one solv ent as being good [13].

POLYMER 349 Data on PES true solubility tak en by author. See Chapter 5 and Table 5.4.

POLYMERS 350–358 These entries are not all polymers but mostly biological materials with the source of data being [14].

POLYMER 359 The solubility of cholesterol, data collected by the author . See Chapter 15.

POLYMER 360 Solubility data generated by high school students as part of project. Included in Reference 4. Source of chlorophyll was crushed lea ves.

Polymer 361 Correlation on strength of paper immersed in dif ferent solvents reported in Reference 4. Data w as taken from Reference 15.

7248_A002.fm Page 488 Wednesday, May 23, 2007 12:53 PM

488

Hansen Solubility Parameters: A User’s Handbook

POLYMER 362 Solubility of ULTRASON® PES has been reported by BASF in their product data. These data were combined with supplementary solubility data for this correlation. Also reported in Reference 16. See Chapter 5.

POLYMERS 363–364 Chemical resistance of B AREX® 210 from data in BP Chemicals datasheet. Styrene is an outlier in the first, whereas its rem val from consideration gives a perfect fit and presumably a more usefu correlation.

POLYMERS 365–367 These data were generated in connection with a lecture to the Nordic Conserv ation Congress in Copenhagen [17]. All give perfect fits, partly because of too f w data, b ut the correlations can be useful. Paraloid B72 and Dammar are used as protecti ve lacquers.

POLYMERS 368–369 These correlations di vide the permeation coef ficients g ven in Reference 18 into >80 and >0.8, respectively. The units are (g x mm)/(m 2 x d). The fits are good. See Chapter 13.

POLYMERS 370–371 These are correlations of e xperimental solubility data for the Rhône-Poulenc reacti ve isocyanates Tolonate® HDT (which g ave the same result as Tolonate® HDT-LV) and Tolonate® HDB (which gave the same results as Tolonate® HDB-LV). The fits were perfect and the numbers reasonable The data could not include alcohol or amine solv ents because of reactions.

POLYMERS 372–389 The data correlated for these 18 rubbers are from a RAPRA database [19]. The information used was satisfactory or unsatisf actory, all other information such as limited suitability w as neglected. No precise weight g ain or other information is a vailable, just the general suitability or not. The values in parentheses are (data fit/number of sol ents). ACM acrylate rubbers (.981/55) ECO epichlorohydrin rubbers (.988/37) CSM chlorosulphonated polyethylene rubber (.906/53) E ebonite (.722/41) EPM ethylene-propylene copolymer (.987/47) EPDM ethylene-propylene terpolymer (.968/51) FQ fluorosilicone rubber (.844/40 FKM hexafluoroprop.-vi ylidine fluoride copol. ( iton) (.769/50) NR natural rubber (1.000/59) NBR nitrile rubber (.990/65) FFKM Kalrez ® (Du Pont) (too resistant to correlate) CR polychloroprene (.877/54) AU polyester polyurethane (.959/63) EU polyether polyurethane (.959/63) T polysulphide rubber (.799/48) Q silicone (.748/53)

7248_A002.fm Page 489 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

SBR TFP

489

styrene butadiene rubber (.942/54) tetrafluoroet ylene-propylene copolymer (.744/26)

POLYMERS 390–412 These correlations use data from the RAPRA collection of data on chemical resistance for plastics [20]. Approach same as for RAPRA rubber data just abo ve.

POLYMERS 413–450 These data are from the collected report of the EC project on self-stratifying coatings reported in a full issue of Progress in Organic Coatings. The specific reference is Reference 21.The evaluations were made at dif ferent concentrations in man y cases. Some alk yds were omitted here.

POLYMERS 451–452 These data are for strong swelling of tw o different film samples of brominated utyl rubber.

POLYMER 453 The correlation is based on strong swelling of a film of polyisoprene

POLYMERS 454–458 These correlations are based on chemical resistance data from Reference 22.

POLYMER 459 Correlation based on solubility of Eth ylene Vinylacetate adhesive EVA 4055.

POLYMER 460 Correlation based on solubility of Topas® 6013 from Ticona GmbH (Hoechst AG).

POLYMER 461 Correlation based on solubility of CZ

®

Resin from the West Company.

POLYMER 462 An older correlation for the solubility of Kauri Gum, used in the Kauri-Butanol test, w with a data fit of 0.95 for the standard sol ents.

as made

POLYMER 463 The data for the solubility of polyvin ylpyrrolidone used in this correlation are found in Reference 23. The data fit as 0.992, b ut as with man y w ater soluble polymers, there is a considerable extrapolation into the “unkno wn” where there are no liquids.

ENTRY 464 The data fit for the correlation of solubility of palm oil with the standard set of sol ents was 0.992.

7248_A002.fm Page 490 Wednesday, May 23, 2007 12:53 PM

490

Hansen Solubility Parameters: A User’s Handbook

ENTRY 465 This is a correlation of the solubility of a fungicide and algaecide called Bethoxazin using solubility data in 19 liquids from Reference 24. The data fit as 0.976.

ENTRY 466 This is a correlation for the solubility of carbon-60 at a gi ven small level as reported in Reference 25; 15 of the 87 liquids were considered as “good” gi ving a data fit of 0.972

REFERENCES 1. Saarnak, A., Hansen C.M., and Wallström E., Solubility Parameters — Characterization of Paints and Polymers, Report from Scandina vian Paint and Printing Ink Research Institute, January 1990, Hoer sholm, Denmark 2. Hansen, C.M., Solubility Parameters, in Paint Testing Manual, Manual 17, J.V. Koleske, Ed., American Society for Testing and Materials, Philadelphia, 1995, pp. 38–404. 3. Hansen, C.M., Billing, C.B., and Bentz, A.P., Selection and Use of Molecular P arameters to Predict Permeation Through Fluoropolymer-Based Protective Clothing Materials, The Performance of Protective Clothing; Fourth Volume, ASTM STP 1133, J.P. McBriarty and N.W. Henry, Eds., American Society for Testing and Materials, Philadelphia, 1992, pp. 894–907. 4. Hansen, C.M., 25 Years with Solubility Parameters (in Danish: 25 År med Opløselighedsparametrene), Dansk Kemi, 73(8), 18–22, 1992. 5. Hansen, C.M. and Hansen, K.M., Solubility Parameter Prediction of the Barrier Properties of Chemical Protective Clothing, Performance of Protective Clothing: Second Symposium. ASTM STP 989, S.Z. Mansdorf, R. Sager, and A.P. Nielsen, Eds., American Society for Testing and Materials, Philadelphia, 1988, pp. 197–208. 6. Wessling, R.A., The Solubility of Poly(vinylidene Chloride), Journal of Applied Polymer Science, 14, 1531–1545, 1970. 7. Hansen, C.M., Solubility Parameters for Polyphenylene Sulfide (PPS) and Polyether Sulphone (PES), Centre for Polymer Composites (Denmark), Danish Technological Institute, Taastrup, 1991, 89 pages. ISBN 87-7756-139-2 8. Wyzgoski, M.G., The Role of Solubility in Stress Cracking of Nylon 6,6,in Macromolecular Solutions — Solvent Property Relationships in Polymers, R.B.Se ymour and G.A.Stahl, Eds. Per gamon, New York, 1982, pp. 41–60. 9 Anonymous, Modern Plastics Enc yclopedia 1984/1985, McGra w-Hill, New York, pp. 482–455. 10. Fuchs, O., Tables of Solv ents and Non-solv ents, Polymer Handbook, 3 rd Ed., J. Branderup and E.H. Immergut, Eds., Wiley, New York, 1989, pp. VII/379-VII/407. 11. Barton, A.F.M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press Inc., Boca Raton, FL. 1983, pp. 280-289. 12. Grulke, E.A., Table 3.4, Solubility Parameter Ranges of Commercial Polymers , Polymer Handbook, 3rd Ed., J. Branderup and E.H. Immer gut, Eds., Wiley, New York, 1989, pp. VII/544–VII/550. 13. Fuchs, O., Tables of Solvents and Non-solvents, Polymer Handbook, 3rd. Ed., J. Branderup and E.H. Immergut, Eds., Wiley, New York, 1989, p. VII/385. 14. Hansen, C.M. and Andersen, B.H., The Affinities of O ganic Solvents in Biological Systems, American Industrial Hygiene Association Journal, 49(6), 301–308, 1988. 15. Robertson, A.A., Cellulose-Liquid Interactions, Pulp and Paper Magazine of Canada, 65(4), T-171T-178, 1964. 16. Hansen, C.M., Solvent Resistance of Polymer Composites — Glass Fibre Reinforced Polyether Sulfone (PES), Centre for Polymer Composites (Denmark), Danish Technological Institute, Taastrup, 1994. 17. Hansen, C.M., Conservation and Solubility Parameters, Nordic Conserv ation Congress Preprints, Copenhagen, 1994, pp. 1–13. 18. Pauly, S., Permeability and Dif fusion Data, Polymer Handbook, 3rd. Ed., J. Branderup and E.H. Immergut, Eds., Wiley, New York, 1989, pp. VI/435–VI/449.

7248_A002.fm Page 491 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

491

19. Anonymous, Chemical Resistance Data Sheets, Volume 2. Rubbers, Ne w Edition — 1993, Rapra Technology, Shawbury, Shrewsbury, Shropshire, 1993. 20. Anonymous, Chemical Resistance Data Sheets, Volume 1. Plastics, Ne w Edition — 1993, Rapra Technology, Shawbury, Shrewsbury, Shropshire, 1993. 21. Benjamin, S., Carr, C., and Walbridge, D.J., Self-stratifying Coatings for Metallic Substrates, Progress in Organic Coatings, 28, 197-207, 1996. 22. Anonymous, Engineering Guide to Du Pont Elastomers, The Du Pont Company, Switzerland, 1987. 23. Hansen, C.M., The Universality of the Solubility P arameter, Ind. Eng. Chem. Prod. Res. Dev., 8(1), 2–11, 1969. 24. Bosselaers, J., Blancquaert, P ., Gors, J., He ylen, I., Lauw aerts, A., Nys, J., Van der Flaas, M., and Valcke Janssen, A., A New Fungicide and Algaecide, Färg och Lack Scandinavia, 49(1), 5–11 2003. 25. Hansen, C.M., and Smith, A.L., Using Hansen Solubility P arameters to Correlate Solubility of C60 Fullerene in Or ganic Solvents and in Polymers, Carbon, 42(8-9), 1591–1597, 2004.

LIST OF TRADE NAMES AND SUPPLIERS PAINTS AND BINDERS: Bayer (D): Cellit, Desmophen, Desmolac, Pergut, Cellidora, Desmodur, Baysilon, Alkydal Hercules (US): Piccolyte, Cellolyn, Pentalyn, Ester Gum, Parlon Ciba-Geigy (CH): Araldite Shell (D): Epikote, Carifl x Union Carbide (US): Vinylite, Phenoxy Hoechst (D): Macrynal, Phenodur, Alpex, Mowithal, Alfthalat, Mowilith Reichhold (CH): Super Beckasite, Uformite Polymer Corp. (CAN): Polysar Goodrich (US): Hycar Hüls (D): Vilit, Vesturit, Buna Hüls, Lutonal, Larofl x, Plastopal, Polystren Monsanto (US): Modafl w, Multifl w, Butvar Montecatini Edison (I): Vipla ICI (GB): Cereclor, Allopren, Suprasec Du Pont (US): Lucite Hagedorn (D): 1/2 sec. Nitrocellulose H 23 Röhm (D): Plexigum Rohm and Haas (U.S.): Paraloid Dynamit Nobel (D): Dynapol SOAB (S): Soamin BIP Chemicals (GB): Beetle Dyno Cyanamid (N): Dynomin DSM Resins (NL): Uracron Wacker (D): Wacker Dow Chemical (CH): Ethocel Cray Valley (GB): Versamid W. Biesterfeld (D): Chlorparaffi Synres (NL): Synresin American Cyanamide (US): Cymel Polyplex (DK): Plexal Pennsylvania Industrial Chemical Corp. (US): Piccopale, Piccoumarone OTHERS: Chemical Fabrics Corporation: Challenge

7248_A002.fm Page 492 Wednesday, May 23, 2007 12:53 PM

492

Hansen Solubility Parameters: A User’s Handbook

Chevron Phillips: Ryton ICI (Victrex plc): Victrex Saranex: Dow Safety 4, 4H: North General Electric: Ultem BASF: Ultrason BP Chemicals: Barex Rhône-Poulenc: Tolonate Ticona (Celanese): Topas West Company (DAIKYO): CZ Resin The capital letters in parenthesis are the international symbols for the respecti ve countries: D US CH CAN I GB S N NL DK

Germany United States of America Switzerland Canada Italy Great Britian Sweden Norway Netherlands Denmark

7248_A002.fm Page 493 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

493

TABLE A.2 Hansen Solubility Parameters for Selected Correlations Number

Polar

Hydrogen Bonding

Interaction Radius

12.00

6.70

10.20

18.20

12.40

10.80

7.40

17.90 20.10

4.30 6.90

3.90 5.90

5.90 9.90

14.00 23.10 20.00 17.40 21.00 19.30 23.40

7.40 14.60 10.32 10.50 11.10 9.37 7.20

9.40 5.00 10.11 9.00 13.40 10.95 14.80

13.70 20.50 10.02 7.90 11.70 8.26 14.90

Epoxy Curing Agents 23.80 20.30 24.90 26.90

5.30 6.60 3.10 2.40

16.20 14.10 18.70 18.50

16.10 9.60 20.30 24.00

17.70 19.10 21.54 16.00 20.60 19.40 17.90 18.70 19.90

10.60 12.20 14.94 13.10 7.80 7.40 9.60 9.60 8.10

11.60 9.90 12.28 9.20 11.60 6.00 5.90 9.90 6.00

9.50 8.00 16.78 11.40 13.10 9.80 8.20 8.20 9.80

23.26 19.74

6.55 11.62

8.35 14.59

19.85 12.69

Hydrocarbon Resins 16.47 17.55 19.42

0.37 1.19 5.48

2.84 3.60 5.77

8.59 6.55 9.62

Polymer

Dispersion Cellulose Acetobutyrate 16.60

1

CELLIT BP-300

2

CELLIDORA A

3 4

ETHOCEL HE10 ETHOCEL STD 20

5 6 7 8 9 10 11

ARALDITE DY O25 EPIKOTE 828 EPIKOTE 1001 EPIKOTE 1004 EPIKOTE 1007 EPIKOTE 1009 PKHH

12 13 14 15

VERSAMID 100 VERSAMID 115 VERSAMID 125 VERSAMID 140

16 17 18 19 20 21 22 23 24

DESMOPHEN 651 DESMOPHEN 800 DESMOPHEN 850 DESMOPHEN 1100 DESMOPHEN 1150 DESMOPHEN 1200 DESMOPHEN 1700 DESMOLAC 4200 MACRYNAL SM 510N

25 26

SUPER BECKACITE 1001 PHENODUR 373U

27 28 29

PLIOLYTE S-100 PICCOPALE 110 PICCORONE 450L

30

POLYSAR 5630

Styrene-Butadiene (SBR) 17.55

3.35

2.70

6.55

31

HYCAR 1052

Acrylonitrile-Butadiene 18.62

8.78

4.17

9.62

Cellulose Acetate

Ethyl Cellulose

Epoxy

Polyurethane

Phenolic Resins

7248_A002.fm Page 494 Wednesday, May 23, 2007 12:53 PM

494

Hansen Solubility Parameters: A User’s Handbook

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Polar

Hydrogen Bonding

Interaction Radius

17.53

2.25

3.42

6.55

16.57

1.41

–0.82

9.62

14.20 16.90 17.40

2.50 2.50 4.30

4.60 4.00 8.40

12.40 7.20 7.40

20.17 16.10

14.61 3.70

15.04 7.90

11.66 8.90

18.40

6.60

8.00

3.00

Dispersion Polybutadiene

32

BUNA HULS B10

33

CARIFLEX IR 305

34 35 36

LUTONAL IC/1203 LUTANAL I60 POLYVINYLBUTYL ETHER

37 38

LIGNIN MODAFLOW

39

VIPLA KR (PVC)

40 41

CERECLOR 70 CHLOROPAR 40

20.00 17.00

8.30 7.60

6.80 7.90

9.80 11.90

42 43

PERGUT S 5 ALLOPREN R10

Chlorinated Rubber 17.40 17.40

9.50 4.30

3.80 3.90

10.00 6.10

44

PARLON P 10

Chlorinated Polypropylene 20.26

6.32

5.40

10.64

45 46

HYPALON 20 HYPALON 30

Chlorosulfonated PE 18.10 18.20

3.40 4.70

4.90 2.00

3.60 5.00

47

ALPEX

19.90

0.00

0.00

9.40

48

1/2-sec.-NITRO CELLULOSE H 23

15.41

14.73

8.84

11.46

49 50 51 52

CELLOLYN 102 PENTALYN 255 PENTALYN 830 ESTER GUM BL

21.73 17.55 20.03 19.64

0.94 9.37 5.81 4.73

8.53 14.32 10.93 7.77

15.75 10.64 11.66 10.64

53 54 55

VERSAMID 930 VERSAMID 961 VERSAMID 965

17.43 18.90 20.15

–1.92 9.60 6.04

14.89 11.10 12.90

9.62 6.20 9.20

Polyisoprene

Polyisobutylene

Special

Polyvinylchloride

Chloroparaffin

Cyclized Rubber

Nitrocellulose

Rosin Derivatives

Polyamide

7248_A002.fm Page 495 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

495

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

17.50 17.60 19.70

11.30 10.00 12.90

5.90 3.70 12.80

8.50 9.30 11.40

18.60 20.20 18.60

12.90 11.20 4.36

10.30 13.30 13.03

8.30 11.20 10.64

17.60 16.20 18.60 18.40 18.64

9.66 6.80 10.80 9.40 10.52

3.97 5.70 4.10 6.50 7.51

10.64 9.10 11.50 10.70 8.59

20.93

11.27

9.66

13.71

22.28

5.75

4.30

12.68

Vinyl Chloride Copolymers 18.40 20.00 20.00 18.40 17.10 16.50 17.70 17.60 17.60 17.40 18.10

8.40 8.30 8.30 7.60 10.40 10.90 11.10 11.10 11.10 10.20 10.30

5.80 6.70 6.70 6.70 6.50 6.40 6.90 6.80 6.40 5.90 4.20

9.00 9.40 9.40 6.80 7.50 7.70 8.70 8.80 8.60 7.80 8.30

Binders in Solution: Alkyds and Polyesters 18.60 10.00 23.00 2.20 22.90 15.20 22.60 13.80 20.50 9.30 19.20 5.30 23.60 1.00 20.60 4.60 17.30 4.20 22.60 13.10

5.00 4.20 7.60 8.10 9.10 6.30 7.60 5.50 7.90 5.80

10.40 16.90 18.10 17.10 12.40 11.90 19.00 12.60 9.30 16.80

Polymer Isocyanate

56 57 58

DESMODUR L DESMODUR N SUPRASEC F-5100

59 60 61

MOWITAL B 30 H MOWITAL B 60 H BUTVAR B 76

62 63 64 65 66

LUCITE 2042 PEMA LUCITE 2044 PMMA PLEXIGUM MB319 PLEXIGUM M527 PMMA

67

MOWILITH 50 PVAC

68

POLYSTYRENE LG

69 70 71 72 73 74 75 76 77 78 79

LAROFLEX MP 45 VILIT MB 30 VILIT MC 31 VILIT MC 39 VINYLITE VAGD VINYLITE VAGH VINYLITE VMCA VINYLITE VMCC VINYLITE VMCH VINYLITE VYHH VINYLITE VYLF

80 81 82 83 84 85 86 87 88 89

ALFTALAT AC 366 ALFTALAT AM 756 ALFTALAT AN 896 ALFTALAT AN 950 ALFTALAT AT 316 ALFTALAT AT 576 ALKYDAL F 261 HS ALKYDAL F 41 DUROFTAL T 354 DYNAPOL L 812

Polyvinylbutyral

Polyacrylate

Polyvinylacetate

Polystyrene

7248_A002.fm Page 496 Wednesday, May 23, 2007 12:53 PM

496

Hansen Solubility Parameters: A User’s Handbook

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

20.00 18.50 18.00 18.80 17.70

6.20 9.21 11.60 12.00 13.00

7.00 4.91 8.50 6.00 7.60

9.50 10.64 9.00 11.50 11.50

20.70 22.20 19.35 25.50 18.80 19.90 15.90 22.10 20.81 22.70

6.10 –0.40 12.83 15.20 14.00 15.80 8.10 5.00 8.29 2.80

12.70 10.10 12.87 9.50 12.30 13.40 6.50 11.30 14.96 5.40

14.80 18.40 9.82 22.20 10.50 11.70 10.60 15.50 12.69 16.20

19.20 19.20 19.60 18.40

7.70 9.60 9.10 9.80

5.70 9.30 6.80 10.00

10.60 12.20 12.20 12.40

90 91 92 93 94

DYNAPOL L 850 PLEXAL C-34 SOALKYD 1935-EGAX VESTURIT BL 908 VESTURIT BL 915

95 96 97 98 99 100 101 102 103 104

BE 370 BEETLE 681 CYMEL 300 CYMEL 325 DYNOMIN MM 9 DYNOMIN UM 15 SOAMIN M 60 SYNRESIN A 560 PLASTOPAL H UFORMITE MX-61

105 106 107 108

URACRON 15 PARALOID P 400 PARALOID P 410 PARALOID EXPER. RES. QR 954

109 110

BAYSILON UD 125 TEFLON (SL2-)

19.40 17.10

9.90 8.10

10.10 1.30

6.90 4.70

111 112 113 114 115 116

Special Data DOW EPOXY NOVOLAC 438 DOW EPOXY NOVOLAC 444 ZINK SILICATE - CHEMICAL RES. 2-COMP EPOXY CHEMICAL RES. POLYVINYLIDINE FLUORIDE SOL. COAL TAR PITCH SOL.

20.30 19.50 23.50 18.40 17.00 18.70

15.40 11.60 17.50 9.40 12.10 7.50

5.30 9.30 16.80 10.10 10.20 8.90

15.10 10.00 15.60 7.00 4.10 5.80

Amino Resins

Acrylate Resins

Silicone Resins

117 118 119 120 121 122 123 124 125 126 127 128

Breakthrough Time (Bt) Correlations for Common Types of Chemical Protective Films at Practical Film Thickness NITRILE 20 MIN 17.50 7.30 6.50 5.10 NITRILE 1 HR 16.60 9.10 4.40 10.00 NITRILE 4 HR 19.00 12.60 3.80 13.30 BUTYL 20 MIN 16.50 1.00 5.10 5.00 BUTYL 1 HR 15.80 –2.10 4.00 8.20 BUTYL 4 HR (2) 17.60 2.10 2.10 7.00 NATURAL RUBBER 20 MIN 14.50 7.30 4.50 11.00 NATURAL RUBBER 1 HR 15.60 3.40 9.10 14.00 NATURAL RUBBER 4 HR 19.40 13.20 7.70 19.00 PVC 20 MIN 16.10 7.10 5.90 9.30 PVC 1 HR 14.90 11.10 3.80 13.20 PVC 4 HR 24.40 4.90 9.90 22.70

7248_A002.fm Page 497 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

497

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

11.20 15.30 17.20 16.90 17.10 24.10 10.90 16.50 13.60 17.60 19.00 14.60 16.60 16.60

129 130 131 132 133 134 135 136 137 138 139 140 141 142

POLYVINYLALCOHOL 20 MIN POLYVINYLALCOHOL 1 HR POLYVINYLALCOHOL 4 HR POLYETHYLENE 20 MIN POLYETHYLENE 1 HR POLYETHYLENE 4 HR VITON 20 MIN VITON 1 HR VITON 4 HR NEOPRENE 20 MIN NEOPRENE 1 HR NEOPRENE 4 HR CH 5100 3 HR CH 5200 3 HR

12.40 13.20 13.60 3.30 3.10 14.90 14.50 8.10 15.40 2.50 8.00 13.90 5.40 6.00

13.00 13.50 15.40 4.10 5.20 0.30 3.10 8.30 8.60 5.90 0.00 2.30 4.00 4.80

12.10 8.80 10.90 8.10 8.20 24.30 14.10 6.60 14.40 6.20 13.20 15.90 3.80 3.70

143 144

High Temperature Solubility of PVDC PVDC (110C) SOLUBILITY 17.60 9.10 PVDC (130C) SOLUBILITY 20.40 10.00

7.80 10.20

3.90 7.60

145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161

Chemical Resistance of High Performance and Other Polymers PES L C=1 18.70 10.50 7.60 PES L B + C =1 17.70 9.70 6.40 PPS CR 93°C 18.80 4.80 6.80 PPS TS60%12MO 18.70 5.30 3.70 PA6 CR 17.00 3.40 10.60 PA66 SOL 17.40 9.80 14.60 PA11 CR 17.00 4.40 10.60 POMH+POMC CR 17.10 3.10 10.70 PETP CR 18.20 6.40 6.60 PTFE L80 CR 16.20 1.80 3.40 PMMA CR 16.00 5.00 12.00 PE? CR Q UESTIONABLE VALUES 16.80 5.40 2.40 PPO CR 17.90 3.10 8.50 PUR CR 18.10 9.30 4.50 ABS CR 16.30 2.70 7.10 PSU CR 16.00 6.00 6.60 VINYL SILANE 16.40 3.70 4.50

9.10 9.30 2.80 6.70 5.10 5.10 5.10 5.20 5.00 3.90 13.00 4.70 8.60 9.70 7.80 9.00 10.00

162 163 164 165 166 167

Correlations for Some Barrier-Type Polymers CELLOPHAN SW 16.10 18.50 EVOH SOL 20.50 10.50 SARANEX 4HR 17.70 18.30 4H 35°C 19.40 13.40 4H 35°C no plasticizer included 20.50 11.30 POLYVINYLALCOHOL 15.00 17.20

14.50 12.30 0.70 18.00 10.30 17.80

9.30 7.30 18.40 8.60 6.70 10.20

168 169 170

Chemical Resistance Data - Modern Plastics Encylopedia ACETAL CELANESE 21.10 9.30 ACETALHOMO-DUO 19.00 5.00 CELLULOSE ACETATE 16.90 16.30

5.90 8.00 3.70

11.40 5.00 13.70

7248_A002.fm Page 498 Wednesday, May 23, 2007 12:53 PM

498

Hansen Solubility Parameters: A User’s Handbook

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

17.20 9.80 14.10 19.00 19.00 19.90 16.70 21.60 18.00

13.80 13.60 2.70 4.00 6.00 3.90 7.70 5.20 3.00 0.00 0.00

2.80 11.40 5.50 3.00 8.00 5.10 –0.50 18.80 4.00 2.00 1.00

12.60 15.20 6.60 4.00 5.00 3.80 8.10 15.40 6.00 2.00 6.00

7.30

1.70

5.10

14.10

171 172 173 174 175 176 177 178 179 180 181

CELL. ACET. BUTYRATE CELL. ACET. PROPIONATE PCTFE FEP FURAN FURF ALC PFA(?) PHENOLIC PETG HDPE 18.00 PP

182

Poly(Ethylene/Chlorotrifluoroethylene) PECTFE SOL AT HIGH TEMP. 19.50

183

PAN

9.10

10.90

184 185 186

PEI - Polyethylene Imide - Environmental Stress Cracking (ESC) PEI 1200PSI 17.20 6.40 5.20 PEI 2400PSI 17.40 4.60 9.00 PEI 600PSI 17.30 5.30 4.70

3.60 7.20 3.30

187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212

Based on Solvent Range Solubility Data - Not too Reliable ESTER GUM 16.90 4.50 ALKYD 45 SOYA 17.50 2.30 SILICONE DC-1107? 19.60 3.40 PVETHYLETHER? 15.10 3.10 PBUTYLACRYLATE 16.20 9.00 PBMA? 15.90 5.50 SILICONE DC 23? 16.40 0.00 PE 16.00 0.80 GILSONITE 17.10 2.10 PVINYLBUTYLETHER 17.40 3.40 NAT RUBBER 16.00 4.00 HYP 20 CHLOROSULFONATED PE 17.40 3.20 ETHCEL N22? 22.70 0.50 CHLORINATED RUBBER 17.90 6.30 DAMMAR GUM 18.40 4.20 VERSAMID 100? 18.80 3.00 PS 18.50 4.50 PVAC 17.60 2.20 PVC 17.60 7.80 PHENOLICS 19.80 7.20 BUNA N B UTADIENE/ACRYLONITRILE 17.80 3.20 PMMA 18.10 10.50 PEO 4000 ? HEATED SAMPLES 21.50 10.90 POLYETHYLENESULFIDE (GOOD) 17.80 3.80 PC 18.10 5.90 PLIOLITE P1230 18.10 4.70

9.20 10.00 9.80 12.90 10.10 8.50 5.50 3.20 4.90 8.40 1.30 4.80 20.10 7.60 8.30 7.80 5.30 4.10 8.20 12.80 3.70 9.50 15.90 4.10 5.50 3.90

18.00

Solubility of Polyacrylonitirile 21.70

6.50 7.70 10.80 11.90 3.00 5.90 7.80 2.80 3.90 7.80 6.00 4.00 16.50 5.10 7.80 9.20 2.90 4.00 3.40 10.80 3.40 5.10 13.10 2.20 6.90 3.70

7248_A002.fm Page 499 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

499

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

18.00 17.30

6.20 6.10 3.70 7.60 13.00 7.80 15.10 7.60 11.90 10.30 18.00 24.00

5.00 7.80 2.70 4.80 10.50 7.10 10.70 3.80 8.80 9.90 4.00 3.00

213 214 215 216 217 218 219 220 221 222 223 224

MYLAR PET VCVA COPOLY PUR 17.90 SAN VINSOL ROSIN EPON 1001 SHELLAC POLYMETHACRYLONITRILE CELLULOSE ACETATE CELLULOSE NITRATE PVOH (NOT GOOD, SEE CHAP. 5) NYLON 66

16.60 17.40 17.00 19.70 17.20 18.30 16.90 17.00 16.00

6.20 8.70 6.90 9.80 10.00 9.60 10.10 14.40 16.50 13.50 9.00 11.00

225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251

Acrylics - Solvent Range ACRYLOID B-44 19.40 ACRYLOID B-66 18.00 ACRYLOID B-72 19.20 ACRYLOID B-82 19.10 R+H PBA 16.00 R+H PiMBA 20.70 R+H PNBMA 16.00 R+H PEMA 19.00 R+H PMAA 25.60 R+H PMMA 19.10 BMA/AN 80/20 17.50 ISOB MALANH/CYCLOL 75/25 16.80 MAA/EA/ST 15/38/47 17.60 MAA/MA/VA 15/27.5/57.5 28.50 MAA/MA/VA 15/17.5/67.5 25.50 MMA/CYCLOL 58/42 18.70 MMA/EA 50/50 17.50 MMA/EA 25/75 19.00 MMA/EA/AGE 40/40/20 17.60 MMA/EA/AA 15.90 MMA/EA/AN 55/30/15? 16.70 MMA/EA/AN 40/40/20 20.40 MMA/EA/BAMA 40/40/20 17.90 MMA/EA/CYCLOL 17.60 MMA/EA/MAM 40/40/20 19.00 MMA/EA/MAM 45/45/10? 19.50 MMA/EA/BVBE 40/40/20 17.80

11.20 9.00 11.20 9.10 8.00 4.10 6.20 9.00 11.20 11.30 9.90 –0.40 5.20 15.70 15.70 9.90 9.90 9.00 9.80 15.90 10.90 13.20 8.50 9.80 9.00 11.10 10.00

4.40 3.00 1.80 3.30 8.00 10.70 6.60 8.00 19.60 4.10 4.10 7.20 7.00 18.10 18.10 8.70 4.10 15.00 5.60 11.50 8.50 11.00 11.70 6.40 15.00 8.70 6.60

10.50 9.00 11.00 9.00 12.00 11.50 9.50 11.00 20.30 10.30 9.50 8.50 4.50 21.50 21.50 8.80 9.50 12.00 9.70 11.10 8.50 12.30 12.90 9.80 12.00 11.20 9.80

252 253 254 255 256 257

ACID DEGMP CARB DEG PTH CRYPLEX 1473-5 DEG ISOPH DEG PHTH DPG PHTH

Polyesters - Solvent Range 15.30 19.40 19.20 19.20 21.00 20.10

13.30 13.40 9.40 17.20 15.20 11.50

14.90 11.60 5.60 14.60 13.20 6.70

15.60 11.10 8.90 11.80 13.70 11.60

7248_A002.fm Page 500 Wednesday, May 23, 2007 12:53 PM

500

Hansen Solubility Parameters: A User’s Handbook

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

17.80 17.80 14.90 21.30 17.00 18.70 19.40 19.00 19.00 18.80 18.10 17.30

258 259 260 261 262 263 264 265 266 267 268 269

DOW ADIP TEREP DOW X-2635 MALEATE VITEL PE LINEAR VITEL PE101-X HYD BIS A FUM ISPH HYD BIS A PG FUM ISPH PENTA BENZ MAL SOL MYLAR 49001 SOL MYLAR 49002 TEG EG MAL TEREP TEG MALEATE VAREZ 123

10.40 5.60 10.10 6.30 4.40 8.90 12.20 5.00 5.00 11.40 13.90 10.90

6.80 6.80 2.90 4.70 6.20 5.50 10.20 4.00 5.00 9.20 12.10 11.90

9.30 4.50 6.10 7.30 5.00 8.40 10.80 5.00 5.00 10.20 9.70 10.70

270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296

Styrene Polymers And Copolymers - Solvent Range AMOCO 18-290 19.30 3.70 BUTON 100 B UTAD-STY 17.00 4.00 BUTON 300 17.30 3.70 KOPPERS KTPL-A 19.30 3.70 RUBBER MOD PS 20.00 5.00 STY MAL ANH 23.40 13.80 LYTRON 820 21.10 13.10 MARBON 9200 19.00 4.00 PARAPOL S-50 17.90 3.90 PARAPOL S-60 17.90 3.90 PICCOFLEX 120 17.40 7.80 SHELL POLYALDEHYDE EX 39 19.60 10.00 SHELL POLYALDEHYDE EX 40 19.60 10.00 SHELL X-450 19.30 9.50 SMA 1430A 18.80 11.40 SAN 85/15 19.10 9.50 STY/BUTENOL 85/15 17.40 7.80 STY/CYCLOL 82/18 18.20 5.60 STY/2EHA/AA 81/11/8 17.70 4.90 STY/MAA 90/10 18.70 6.30 STY/MA 85/15 18.00 9.00 STY/HALF ESTER MA 60/40 18.90 10.90 STY/PROP HALF E MA 57/43 18.00 9.80 STY/VBE 85/15 17.40 7.80 STYRON 44OM-27 MOD PS 20.00 5.00 STYRON 475M-27 20.00 5.00 STYRON 480-27 20.00 6.00

7.90 3.00 7.30 7.90 1.00 15.20 14.50 4.00 4.90 4.90 3.80 3.60 3.60 11.10 16.40 3.10 3.80 7.20 5.90 7.30 3.00 10.70 8.40 3.80 1.00 1.00 4.00

7.80 7.30 7.00 7.80 7.00 16.50 14.40 6.00 3.90 3.90 7.70 9.40 9.40 11.10 14.10 8.70 7.70 5.70 5.90 6.70 9.00 9.70 10.10 7.70 7.00 7.00 5.30

297 298 299 300 301 302

ACRYLOID K120N DODA 6225 DODA 3457 ELVAX 250 ELVAX 150 ELVAX EOD 3602-1

3.80 1.00 1.00 1.00 0.70 2.70

9.50 3.00 3.00 3.00 6.00 5.40

Vinyl Resins - Solvent Range 17.60 19.00 19.00 19.00 18.70 17.70

10.00 2.00 2.00 2.00 2.30 3.30

7248_A002.fm Page 501 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

501

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

17.40 17.90 17.40 19.50 19.00

3.80 2.50 3.80 11.10 15.00 8.30 12.00 14.20 14.20 8.00 0.80 0.80 9.60 7.70 13.00 11.00 11.10 13.80 0.80 7.70 6.80 7.70 4.60 5.00 4.60 15.00 15.00 7.50

7.70 9.00 7.70 8.00 12.00 8.60 14.00 14.00 14.00 10.00 23.70 23.70 7.30 5.30 12.60 15.00 9.60 12.30 11.70 20.40 7.10 8.50 8.40 10.00 8.40 12.00 12.00 5.90

19.90 6.00 3.00 13.20 5.00 3.00 9.00 12.10 7.40 14.00 2.00 4.40 3.20 13.20 8.40 16.40

17.00 3.00 18.10 17.10 7.00 18.10 9.00 8.50 1.00 11.90 5.00 4.10 5.10 12.20 10.60 20.90

303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330

EXON 470 PVC EXON 471 EXON 473 GEON 121 POLYCYCLOLa PVBE 16.70 PVEE FORMVAR 7/70E PVFORMAL FORMVAR 15/95E PVIBE SARAN F-120 VCL2/AN? SARAN F-220 ? SINCLAIR 3840A VA/EHA/MA 63/33/4 VA/EHA/CYC/MAA/76/12/8/4 VA/EA/CY 70/20/10 VBE/AN/MAA 46/27/27 VBE/MA/MAC46/27/27 VDC/AA 75/25 ? VINYLITE AYAA PVAC VINYLITE VAGH VINYLITE VMCH VINYLITE VXCC VINYLITE VYHH VINYLITE VYLF VINYLITE XYHL PVB UTYRAL VINYLITE XYSG PVB UTYRAL VYSET 69

16.00 22.20 22.20 16.00 28.80 28.80 18.40 17.70 21.20 20.00 18.90 19.40 20.40 22.90 17.00 18.30 18.00 19.00 18.00 19.00 19.00 17.90

7.80 8.70 7.80 6.70 9.00 3.70 4.00 12.60 12.60 1.00 16.80 16.80 4.00 6.30 12.40 12.00 11.70 13.00 11.00 18.30 7.80 9.70 9.40 11.00 9.40 9.00 9.00 3.50

331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346

Miscellaneous - Solvent Range ACRYLAMIDE MONOMER 16.90 BAKELITE SULFONE P-47 20.00 BECKOLIN 27 MODIF OIL 11.40 PEO 4000 ? SAMPLES HEA TED 22.20 CHLORINATED RUBBER 18.00 CONOCO H-35 HYDROCARBON M 11.40 DAMMAR GUM DEWAXED 19.00 EPOCRYL E-11 ? 17.30 ESTANE X-7 ?? DIO XANE ONLY 19.00 HEXADECYL MONOESTER TRIME 19.00 HYDR SPERM OIL WX135 20.00 HYPALON 20 CHL SULF PE 17.80 HYPALON 30 17.80 KETONE RESIN S588 18.00 SANTOLITE MHP ARYLSULFONA 18.40 pTOLSULFONAMIDE-FORMALDEH 24.60

18.10 3.00 0.00 11.20 6.00 0.00 2.00 12.90 1.80 11.60 4.00 3.20 3.40 10.80 12.00 18.60

7248_A002.fm Page 502 Wednesday, May 23, 2007 12:53 PM

502

Hansen Solubility Parameters: A User’s Handbook

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

347 348 349

Polymer Solubility Data from Various Sources VYHH-NIF REPT 17.40 9.90 PVF? (DMF ONLY GOOD SOLVENT) 17.40 13.70 PES SOL 19.60 10.80

6.70 11.30 9.20

7.50 2.00 6.20

350 351 352 353 354 355 356 357 358 359 360 361

Biologically Interesting Systems LARD 37C 15.90 LARD 23C 17.69 1%IN WATER -AMINES 15.07 1%IN WATER +AMINES 14.96 BLOOD SERUM 23.20 SUCROSE 21.67 UREA 20.90 PSORIASIS SCALES 24.64 LIGNIN 20.61 CHOLESTEROL 20.40 CHLOROPHYLL 20.20 CELLULOSE-PAPER STRENGTH 25.40

1.16 2.66 20.44 18.33 22.73 26.26 18.70 11.94 13.88 2.80 15.60 18.60

5.41 4.36 16.50 15.15 30.60 29.62 26.40 12.92 15.25 9.40 18.20 24.80

12.03 7.98 18.12 16.22 20.48 20.44 19.42 19.04 11.83 12.60 11.10 21.70

362

PSU ULTRASON S

19.70

8.30

8.30

8.00

363 364

BAREX 210 CR BAREX 210 CR-STYRENE

20.10 17.70

9.10 8.90

12.70 10.90

10.90 6.40

365 366 367

Polymers of Interest for Conservation of Paintings PARALOID B72 17.60 7.40 ESTIMATE DRIED OIL 16.00 6.00 DAMMAR DEWAXED 19.00 2.00

5.60 7.00 9.00

9.40 5.00 9.00

368 369

LDPE PERM>80 LDPE PERM<0.8

Permeation of LDPE by Organic Liquids 16.50 4.50 15.30 5.30

0.50 2.50

6.00 10.10

370 371

Tolonate Solubility TOLONATE HDT (RH-POULENC) 19.00 TOLONATE HDB (RH-POULENC) 19.00

11.00 11.00

3.00 2.00

12.00 11.30

372 373 374 375 376 377 378 379 380 381 382

Chemical Resistance of Elastomers R ACM 16.80 R BUTYL 18.00 R ECO 21.30 R CSM 28.00 R EBONITE (DATA FIT 0.722) 18.70 R ETHYLENE/PROPYLENE 16.60 R EPDM 18.60 R FQ FL/SI 15.90 R FKM (VIT ON) (0.76 DATA FIT) 11.60 R NR NAT RUB 20.80 R NBR 19.80

11.80 0.00 8.10 14.00 6.10 0.00 –3.40 20.10 23.00 1.80 17.80

11.60 3.00 6.10 3.40 2.70 5.20 4.40 6.90 5.00 3.60 3.20

17.00 9.00 12.00 28.30 6.60 9.10 10.70 16.80 21.60 14.00 19.00

Polysulfone PSU

Barex

7248_A002.fm Page 503 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

503

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

24.60 17.90 17.90 25.30 13.80 17.20 16.60

8.60 13.30 13.30 17.30 5.00 6.00 6.80

6.40 10.70 10.70 6.70 1.20 4.60 0.60

20.40 17.10 17.10 23.60 14.30 9.80 7.90

8.60 7.10 6.50 12.20 9.00 8.10 8.80 5.70 5.60 10.90 4.30 6.30 19.50 16.70 1.40 8.90 11.20 5.60 10.80 12.30 11.30 17.40 17.40

383 384 385 386 387 388 389

R CR CHLOROPRENE R AU ESTER PU R PEU ETHER PU R T SULPHIDE R Q SILICONE (0.748 D ATA FIT) R SBR R TFP TETFLPROP (0.744 DATA FIT)

390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412

Chemical Resistance of Plastics R ABS 17.60 R CELLULOSE ACETATE 14.90 R CHLORINATED PVC 17.50 R DIALLYLPHTHALATE 22.20 R POM ACETAL 17.20 R PA12 18.50 R PA66 18.20 R POLYAMIDEIMIDE 18.50 R POLYBUTYLENETEREPH 18.00 R POLYCARBONATE 19.10 R HDPE/LDPE 17.50 R PET 19.10 R POLYIMIDES 24.30 R PMMA 19.30 R TPX 18.80 R POLYPHENYLENEOXIDE 16.90 R POLYSULPHONE 19.80 R POLYPROPYLENE 17.20 R EPOXY COLD CURING 16.80 R EPOXY HOT CURING 18.30 R HET RESIN 17.50 R ISOPHTHALIC 19.80 R TEREPHTALIC 19.80

6.40 11.10 5.50 8.60 9.80 9.10 10.80 8.70 8.40 5.10 8.30 9.10 22.90 4.70 6.40 2.70 6.20 –0.40 8.80 9.70 8.30 4.20 4.20

10.90 12.40 6.30 15.80 5.30 6.30 5.20 4.20 4.50 12.10 3.90 4.80 21.60 17.40 7.90 11.70 11.30 4.50 8.20 7.30 8.60 18.00 18.00

413 414 415 416 417 418 419 420 421 422 423 424

Polymers at Different Test Concentrations - (Conc) Epoxy Polymers EPIKOTE 828 (60%) 16.60 14.00 2.80 EPIKOTE 828 (30%) 16.30 16.40 1.90 EPIKOTE 1001 (60%) 15.80 16.30 3.30 EPIKOTE 1001 (40%) 16.30 13.10 6.30 EPIKOTE 1001 (20%) 19.80 13.60 8.90 EPIKOTE 1001 (10%) 18.10 11.40 9.00 EPIKOTE 1004 (60%) 17.70 10.10 7.60 EPIKOTE 1004 (30%) 18.50 9.30 8.00 EPIKOTE 1007 (30%) 18.60 10.60 8.10 EPIKOTE 1009 (60%) 17.00 9.60 8.50 EPIKOTE 1009 (30%) 19.80 10.60 10.30 EPIKOTE 1009 (10%) 19.00 9.10 10.70

14.90 16.70 16.40 10.90 12.00 9.10 9.80 9.60 8.80 7.60 9.70 8.00

425 426

PIBMA (10%) PIBMA (30%)

Acrylics 17.00 17.10

4.60 5.90

7.60 0.70

9.50 7.30

7248_A002.fm Page 504 Wednesday, May 23, 2007 12:53 PM

504

Hansen Solubility Parameters: A User’s Handbook

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Dispersion

Polar

Hydrogen Bonding

Interaction Radius

17.80 17.20 20.60 18.10 17.60 17.50 16.50 16.90 17.80 21.20 16.30 16.30

10.40 7.20 3.50 5.70 10.10 5.50 8.70 7.80 6.40 1.40 10.60 10.60

2.90 3.50 7.20 0.00 5.80 3.80 5.00 0.50 4.70 10.70 7.40 7.40

9.60 4.80 12.80 8.50 9.40 4.50 10.40 7.30 10.70 12.30 12.90 12.90

427 428 429 430 431 432 433 434 435 436 437 438

PMMA (10%) PMMA (30%) PBMA (10%) PBMA (30%) PMMA (10%) PMMA (30%) PEMA (10%) PEMA (30%) CRODA AC500 THERMOSET CRODA AC500 THERMOSET CRODA AC550 THERMOSET CRODA AC550 THERMOSET

439 440 441 442

LUMFLON LF200 (10%) LUMFLON LF200 (30%) LUMFLON LF916 (10%) LUMFLON LF916 (30%)

Fluorinated Polyethers 18.50 20.10 17.50 18.10

5.40 4.40 6.80 3.90

6.90 3.20 10.50 8.30

9.90 8.50 12.50 8.80

443 444 445 446

PLASTOKYD S27 (30%) PLASTOKYD SC140 (30%) PLASTOKYD SC400 (30%) PLASTOKYD AC4X (30%)

Acrylic Modified Alkyd 20.10 25.20 23.70 23.90

5.70 9.20 0.50 7.80

5.30 3.70 10.30 8.80

20.00 20.00 20.00 19.90

447 448 449

ALLOPRENE R10 (10%) ALLOPRENE R10 (30%) ALLOPRENE R10 (60%)

Chlorinated Rubber 19.50 17.90 19.60

9.20 5.60 6.50

6.90 6.70 5.80

7.50 5.80 9.10

450

HYPALON 20 (30%)

Chlorosulfonated Polyethylene 20.30

3.20

0.70

11.30

451

POLYISOPRENE SW

Polyisoprene Swelling 17.00

4.00

4.00

7.30

452 453

BROMOBUTYL RUBBER S BROMOBUTYL RUBBER L

Bromobutyl Rubber Swelling 17.60 17.00

1.70 3.40

2.00 2.00

6.00 6.00

454 455 456 457 458

Supplemental Chemical Resistance Corrlations NEOPRENE CR 18.10 4.30 HYTREL +/- OK 24.20 14.60 HYTREL +/- NOT OK 26.40 18.80 HYPALON +/- OK 18.40 3.60 HYPALON +/- NOT OK 18.40 5.60

6.70 13.20 7.40 6.40 6.00

8.90 18.80 26.30 9.00 9.40

459

EVA 4055 SOL

3.70

4.70

(10%) (30%) (10%) (30%)

Ethylene Vinylacetate (EVA) Solubility 17.70

3.50

7248_A002.fm Page 505 Wednesday, May 23, 2007 12:53 PM

Appendix A: Table A.2

505

TABLE A.2 (CONTINUED) Hansen Solubility Parameters for Selected Correlations Number

Polymer

Polar

Hydrogen Bonding

Interaction Radius

18.00 18.00

3.00 1.00

2.00 3.00

5.00 4.00

18.7 21.4 17.7 22.4 19.7

8.1 11.6 3.5 7.6 2.9

13.0 21.6 3.7 10.8 2.7

8.2 17.3 4.7 13.9 3.9

Dispersion COC Solubility

460 461

TOPAS 6013 SOL CZ RESIN SOL

462 463 464 465 466

KAURI GUM POLYVINYLPYRROLIDONE (PVP) PALM OIL BETHOXAZIN CARBON-60

Miscellaneous

7248_A002.fm Page 506 Wednesday, May 23, 2007 12:53 PM

7248_A003.fm Page 507 Wednesday, May 23, 2007 12:57 PM

Appendix A: Table A.3 COMMENTS TO TABLE A.3 The data in Appendix A.3 is for the solubility of the polymers listed in Table 5.2. 0.5 g of polymer was placed in a test tube with 5 ml of the gi ven solvent (except for the Milled Wood Lignin). The evaluation of solubility w as made on a scale as follo ws: 1. 2. 3. 4. 5. 6.

Soluble Almost soluble Strongly swollen, slight solubility Swollen Little swelling No visible effect

Note:* indicates a reaction. Data reproduced from Hansen, C.M., “The Three Dimensional Solubility P arameter and Solv ent Dif fusion Coef ficient ” doctoral dissertation, Danish Technical Press, Copenhagen, 1967.

507

1 1 1 1 1 1 1 1 1 5 5 1 1 1 1 1 4 1 1 1 1 3 6 5 1 1 1 1 1

A 6 5 1 6 1 4 1 5 6 6 6 1 1 6 1 1 5 5 1 4 1 3 6 6 1 1 3 1 3

B 2 1 1 1 1 1 1 1 2 5 3 1 1 1 1 1 3 4 2 2 1 1 6 6 1 1 4 1 4

C 4 3 1 6 1 1 1 1 1 5 5 1 1 1 6 1 1 1 1 1 1 1 1 1 1 1 1 1 1

D 2 2 1 5 1 1 1 1 1 4 1 1 1 1 1 3 5 2 1 1 1 1 6 1 1 1 1 5 1

E 1 5 1 5 1 1 1 1 1 6 1 1 1 1 3 2 5 5 1 4 1 1 5 1 1 1 4 3 5

F 6 6 3 4 1 1 1 1 1 6 6 1 5 4 1 1 1 1 1 1 1 5 1 6 1 1 1 1 3

G 1 1 1 1 1 1 1 1 1 6 5 1 1 1 1 1 4 1 1 1 1 1 4 1 1 1 5 1 5

H 1 1 6 6 1 1 1 5 6 5 1 6 1 1 3 5 5 6 1 3 1 1 6 1 1 1 5 5 6

I 1 1 1 1 1 3 1 6 6 6 6 1 1 6 1 1 6 6 6 6 6 5 6 6 1 * 6 5 1

J 6 6 3 6 1 1 1 1 1 6 6 1 1 6 3 3 1 1 1 1 1 6 6 6 1 * 1 1 1

K 3 1 1 4 1 3 4 6 6 6 6 6 6 6 1 4 6 6 6 6 3 4 6 3 3 3 6 5 6

L 1 1 1 1 1 1 1 1 1 6 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

M 4 1 1 5 1 1 1 6 6 1 1 6 1 6 1 3 6 6 6 6 6 1 6 6 1 1 6 6 6

N 4 5 3 5 1 1 1 1 1 6 1 1 1 1 5 3 1 1 1 1 1 1 1 1 1 1 1 1 1

O 1 1 1 6 1 1 1 6 6 1 1 1 1 1 1 4 5 5 5 3 3 1 5 1 1 5 6 6 2

P 4 4 1 5 1 1 1 6 6 4 6 6 3 6 1 5 4 6 6 6 3 1 6 5 1 6 6 5 6

Q 1 1 1 2 1 1 1 1 1 6 4 1 1 1 1 1 3 1 1 1 1 1 5 2 1 1 1 1 1

R 1 1 1 1 1 1 1 2 4 1 3 1 1 1 1 1 3 5 1 3 1 1 5 1 1 * 4 1 1

S 6 6 6 6 1 6 2 1 1 5 6 1 6 4 6 6 1 1 1 1 1 5 1 5 1 1 1 5 1

T 6 2 1 5 1 1 1 1 1 5 6 1 1 5 1 1 4 5 1 1 1 1 6 5 1 1 1 1 1

U 6 6 6 6 5 6 5 1 1 6 6 1 6 4 6 5 1 1 1 1 1 5 1 5 1 1 1 6 1

V 6 6 6 6 5 6 5 1 5 6 1 1 5 1 6 6 1 1 1 1 1 1 1 1 1 1 1 6 1

X 6 6 6 6 2 6 4 1 1 6 5 1 6 4 6 6 1 1 1 1 1 5 1 5 1 1 1 5 1

Y 3 6 6 6 6 5 3 5 6 6 1 6 3 1 6 6 4 4 5 5 1 1 6 1 6 1 6 6 6

Z 4 5 1 4 1 1 1 1 1 6 3 1 1 1 5 1 1 1 1 1 1 1 2 1 1 1 1 1 1

A 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 1 1 1 1 1 1

B 6 5 6 6 5 6 5 1 1 6 6 1 6 1 6 6 1 1 1 1 1 6 1 5 1 1 1 4 1

C 1 1 1 4 1 1 1 4 5 6 4 2 1 3 1 2 6 4 3 3 1 1 6 1 1 1 3 1 6

D 6 5 6 6 5 6 4 1 1 6 6 2 6 4 6 6 1 1 1 1 1 6 1 6 1 1 1 6 1

E

6 6 5 6 2 3 1 5 4 6 6 5 6 6 4 6 6 6 4 6 5 6 6 6 1 6 5 5 6

F

5 5 2 5 1 1 1 1 1 6 6 1 6 1 2 1 1 1 1 1 1 1 2 5 1 5 1 1 1

G

4 5 5 6 4 1 3 5 4 6 6 6 1 6 1 5 4 5 5 6 6 1 6 6 5 * 6 6 6

L

508

Solvent Acetic Acid Acetic Anhydride Acetone Acetonitrile Acetophenone Aniline Benzaldehyde Benzene 1-Bromonaphthalene 1,3-Butanediol 1-Butanol n-Butyl Acetate n-Butyl Lactate Butyric Acid Gamma-Butyrolactone Butyronitrile Carbon Disulf de Carbon Tetrachloride Chlorobenzene 1-Chlorobutane Chloroform m-Cresol Cyclohexane Cyclohexanol Cyclohexanone Cyclohexylamine Cyclohexylchloride Di-(2-Chloroethyl)ether Di-isobutyl Ketone

Polymer

TABLE A.3 Solubility Data for the Original 33 Polymers and 88 Solvents

7248_A003.fm Page 508 Wednesday, May 23, 2007 12:57 PM

Hansen Solubility Parameters: A User’s Handbook

Diacetone Alcohol o-Dichlorobenzene Diethyl Amine Diethyl Ether Diethylene Glycol Diethylene Glycol Monob utyl Ether Diethylene Glycol Monomethyl Ether Dimethyl Formamide 1,4-Dioxane Dipropyl Amine Dipropylene Glycol Ethanol 96% Ethanol 99.9% Ethanolamine Ethyl Acetate Ethyl Benzene Ethyl Lactate 2-Ethyl-1-Butanol 2-Ethyl-1-Hexanol 1,2-Dichloroethane Diethyl Sulfid Dimethyl Sulfoxide Ethylene Glycol Ethylene Glycol Monob utyl Ether Ethylene Glycol Monoethyl Ether Ethylene Glycol Monoethyl Ether Acetate Ethylene Glycol Monomethyl Ether Formic Acid 90% Furan Glycerol n-Hexane Isoamyl Acetate Isobutyl Isobutyrate Isophorone Mesityl Oxide Methanol

1 1 1 3 6 3 1 1 1 1 6 4 5 6 1 1 1 3 4 1 1 4 6 1 1 1 4 1 1 4 6 1 1 1 1 5

5 1 6 6 6 3 1 1 1 6 6 6 6 6 1 5 1 2 2 1 4 1 6 5 5 1 3 6 4 6 5 1 5 1 1 6

1 2 1 3 1 1 1 1 1 1 1 4 3 1 1 4 1 5 5 1 4 1 6 1 1 1 1 5 1 6 6 1 5 1 1 4

1 1 1 1 5 1 1 1 1 1 6 6 5 6 1 1 1 3 1 1 1 3 4 1 1 1 1 3 1 5 1 1 1 1 1 5

1 1 1 4 3 1 1 1 1 1 1 4 1 2 1 2 1 1 1 1 1 1 6 1 1 1 1 6 1 6 6 1 5 1 1 4

1 1 1 5 5 1 1 1 1 1 5 1 1 6 1 3 1 1 1 1 5 2 6 1 1 1 1 1 1 6 6 1 5 1 1 5

5 1 1 3 6 4 6 1 1 1 6 6 6 6 1 1 6 6 6 1 1 6 6 6 5 1 6 6 1 6 3 1 1 1 1 6

1 1 5 4 6 1 1 1 1 5 5 1 4 6 1 1 1 5 5 1 3 1 6 2 1 1 1 1 1 6 6 1 5 1 1 1

1 2 4 6 4 1 1 1 6 6 1 1 1 5 6 4 1 1 1 1 5 1 6 1 1 5 1 1 3 6 6 5 5 1 5 3

1 6 5 6 1 1 1 1 6 5 1 6 3 * 1 6 1 6 6 6 6 1 6 1 1 1 1 6 6 6 6 1 1 1 1 1

3 1 5 6 6 1 4 1 1 5 6 6 6 6 1 1 5 6 6 1 1 1 6 2 5 1 6 6 3 6 6 1 1 1 1 6

4 6 6 6 6 6 4 1 1 6 6 6 6 4 5 6 1 6 6 5 6 1 6 6 6 4 4 1 6 6 6 6 6 6 6 6

1 1 1 1 4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 6 1 1 1 1 5 1 6 1 1 1 1 1 2

1 6 1 5 1 1 1 1 1 1 1 1 1 1 1 6 1 6 6 6 6 1 1 1 1 1 1 1 6 6 6 6 6 1 1 1

1 1 1 1 6 1 1 1 1 1 5 5 5 5 1 1 1 1 1 1 1 4 6 1 1 1 2 5 1 6 5 1 1 1 1 4

1 4 5 3 1 1 1 1 1 5 1 1 1 1 1 6 1 1 1 4 5 1 6 1 1 1 1 6 5 6 5 5 6 1 1 1

1 6 5 6 1 1 1 1 1 5 1 5 5 5 1 6 1 6 6 5 6 1 6 1 1 1 1 2 5 6 6 6 6 1 1 5

1 1 1 5 5 1 1 1 1 1 5 5 4 6 1 1 1 3 4 1 1 1 5 1 1 1 1 5 1 4 5 1 1 1 1 4

1 1 1 3 1 1 1 1 1 2 1 1 1 1 1 3 1 3 4 1 4 1 2 1 1 1 1 1 1 3 6 1 3 1 1 1

6 1 1 1 6 6 6 6 1 1 6 6 6 6 3 1 6 6 6 1 1 6 6 6 6 5 6 6 1 6 3 1 1 1 1 6

1 1 1 6 6 4 3 1 1 4 6 6 6 6 1 1 4 6 6 1 1 5 4 4 4 1 5 6 1 6 6 1 1 1 1 5

6 1 1 1 6 5 6 5 4 1 6 6 6 6 5 1 6 6 5 1 1 6 6 6 6 6 6 6 1 5 1 1 1 1 1 6

6 1 1 1 6 5 6 5 5 1 6 6 6 6 1 1 6 1 1 1 1 6 6 1 6 6 6 6 1 6 1 1 1 1 1 6

6 1 1 1 6 6 6 6 1 1 6 6 6 5 5 1 6 6 5 1 1 6 5 6 6 6 6 6 1 6 1 1 1 1 1 6

6 6 5 6 6 6 6 6 6 6 6 5 6 6 6 6 6 1 1 6 6 6 6 4 6 6 6 3 6 6 6 6 6 6 6 5

3 1 1 1 6 1 5 4 1 1 6 5 4 6 1 1 3 1 1 1 1 6 6 1 1 1 4 6 1 6 4 1 1 1 1 6

1 1 1 1 1 1 1 1 1 1 1 1 1 6 1 1 1 1 1 1 1 1 6 1 1 1 1 1 1 6 6 1 1 1 1 1

6 1 1 1 6 6 6 6 4 1 6 6 6 6 4 1 6 5 1 1 1 6 6 6 6 6 6 6 1 6 1 1 1 1 1 6

1 3 2 4 1 1 1 1 1 3 1 5 5 1 1 6 1 4 4 1 4 1 6 1 1 1 1 6 4 6 6 4 5 6 1 5

6 1 1 1 6 6 6 6 5 1 6 6 6 6 4 1 6 6 5 2 1 6 6 6 6 6 6 6 1 6 2 1 1 1 4 6

6 4 6 6 6 6 6 4 3 6 6 6 6 6 6 6 6 6 6 3 6 2 6 6 6 5 6 6 6 6 6 5 6 6 2 6

5 1 1 1 6 1 4 1 1 1 6 6 6 6 1 1 5 6 6 1 1 6 6 1 2 1 5 6 1 6 5 1 1 1 1 6

1 4 5 6 1 1 1 1 2 6 1 5 6 1 6 6 3 6 6 6 6 1 1 6 1 6 1 1 6 6 6 6 6 6 4 5

7248_A003.fm Page 509 Wednesday, May 23, 2007 12:57 PM

Appendix A: Table A.3 509

A 1 1 4 1 1 1 1 1 1 1 1 4 5 3 6 1 1 1 1 1 1 1 1

Polymer

Methyl Ethyl K etone Methyl Isoamyl K etone Methyl Isobutyl Carbinol Methyl Isobutyl Ketone Methylal Methylene Dichloride Morpholine Nitrobenzene Nitroethane Nitromethane 2-Nitropropane 1-Pentanol 1-Propanol Propylene Carbonate Propylene Glycol Pyridine Styrene Tetrahydrofuran Tetrahydronaphthaline Toluene 1,1,1-Trichloroethane Trichloroethylene Xylene

1 1 5 1 1 1 1 1 1 2 2 5 3 1 5 1 2 1 4 2 4 2 3

C 1 1 1 1 1 1 1 1 1 5 1 3 5 5 6 1 1 1 1 1 1 1 1

D 1 1 1 1 1 1 1 1 5 6 6 1 1 6 4 1 1 1 1 1 1 1 5

E 1 1 1 1 1 1 1 1 5 5 5 1 1 6 6 1 1 1 2 2 4 1 3

F 1 1 6 1 1 1 1 1 2 5 1 6 6 6 6 1 1 1 1 1 1 1 1

G 1 1 5 1 1 1 1 1 1 1 1 2 4 1 6 1 1 1 5 1 1 1 4

H 6 6 1 6 6 1 1 1 6 5 6 1 1 6 1 1 3 1 5 2 6 1 5

I 1 1 6 1 1 6 1 1 1 1 1 6 6 1 6 1 6 1 6 6 6 6 5

J 1 1 6 1 5 1 1 1 5 6 3 6 6 6 6 1 1 1 1 1 1 1 1

K 4 6 6 6 5 5 1 4 5 1 5 6 6 3 6 1 6 1 6 6 6 6 6

L 1 1 1 1 1 1 1 1 1 2 1 1 1 1 6 1 1 1 1 1 1 1 1

M 1 1 6 1 3 6 1 5 6 6 6 1 1 1 1 1 6 1 6 6 6 6 6

N 1 1 1 1 1 1 1 1 3 6 4 1 2 6 6 1 1 1 1 1 1 1 1

O 1 1 1 1 1 5 1 3 6 6 6 1 1 6 4 1 4 1 4 6 3 4 6

P 1 6 6 4 5 3 1 2 5 5 4 6 6 1 6 1 6 1 6 6 6 5 6

Q 1 1 4 1 1 1 1 1 1 2 1 5 4 1 5 1 1 1 1 1 1 1 1

R 1 1 4 1 1 1 1 1 1 1 1 6 2 1 1 1 1 1 2 3 3 1 3

S 4 1 6 1 5 1 4 1 6 6 6 6 6 6 6 1 1 1 1 1 1 1 1

T 1 1 5 1 1 1 1 1 1 1 1 6 6 4 6 1 1 1 1 1 1 1 1

U 4 1 6 1 4 1 5 1 6 6 6 6 6 6 6 5 1 1 1 1 1 1 1

V 1 1 1 1 1 1 4 6 6 6 6 1 1 6 6 4 1 1 1 1 1 1 1

X 4 1 6 4 3 1 5 1 6 6 5 6 5 6 6 1 1 1 1 1 1 1 1

Y 6 6 1 6 6 2 6 6 6 6 6 1 1 6 6 5 6 1 5 6 6 2 6

Z 1 1 1 1 1 1 1 1 3 4 1 1 5 6 6 1 1 1 1 1 1 1 1

A 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

B 5 1 5 1 5 1 6 5 6 6 6 6 6 6 6 5 1 1 1 1 1 1 1

C 1 4 4 4 2 1 1 1 3 4 3 4 4 1 5 1 4 1 5 5 4 4 5

D 5 2 6 4 3 1 6 6 6 6 6 6 6 6 6 5 1 1 1 1 1 1 1

E

3 5 6 4 6 2 1 2 6 6 5 6 6 6 6 2 5 1 6 5 6 5 6

F

1 1 6 1 1 1 1 1 2 5 1 6 6 6 6 1 1 1 1 1 1 1 1

G

6 6 6 6 6 6 1 4 6 6 6 6 6 6 3 1 5 3 4 5 6 5 6

L

510

1 1 5 1 5 1 1 1 1 6 1 6 6 1 6 1 2 1 3 2 5 4 4

B

TABLE A.3 (CONTINUED) Solubility Data for the Original 33 Polymers and 88 Solvents

7248_A003.fm Page 510 Wednesday, May 23, 2007 12:57 PM

Hansen Solubility Parameters: A User’s Handbook

7248_Index.fm Page 511 Wednesday, May 9, 2007 10:49 AM

Index A Absorption concentration-dependent, 298 equations, 294–296 film formation by sol ent evaporation and, 305–306 plastics, 315–316, 339 side effects, 304–305 surface resistance and, 300–301 Acetophenone, 59, 262 Acid-base theories and HSP of pigments, 130–131 Acids amines and, 140 dimerization, 65 hydrophilic bonding, 133–134 ACLAR®, 251, 252–253 Acrylonitrile/butadiene/styrene (ABS) terpolymer, 120, 300 Active agents, surf ace, 332–333 Activity coefficient model Flory-Huggins models, 82–84 future challenges, 90–91 at infinite dilution, 85–8 mixed solvent-polymer phase equilibria, 88–90 Adenine, 275 Adhesion maximization, ph ysical, 122, 342 Adsorption, controlled, 132–134 Affinity and Hansen solubility parameters, 4– Alcohol in the bloodstream, 275–276 Alpha-helices, 272 Alternative systems, solv ent, 312 Amines, 140 Amino acid side chains and w ater, 270–271 Ammonia, 57 Amphipathic molecules, 270–271 Anomalous diffusion, 306–308 Aromas and fragrances, 338–339 Asphalt, 152 Asphaltenes, 153, 164–166 definition, 15 molecular weight of, 154 polarity, 155 Availability of HSP data, 321–322 Azeotropes, 212, 213–215, 216–226

B Barrier polymers breakthrough times, 245–247 concentration-dependent diffusion, 244–245 human skin as, 250, 316 laminated, 253–254

RED numbers, 247, 253 solubility parameter correlation of permeation coefficients for ases, 251–254 solubility parameter correlation of permeation rates, 248–250 solubility parameter correlation of polymer swelling, 250 solubility parameter correlations based on permeation phenomena, 245–250 Benzene, 204 Beta sheets, 272 Binders coating, 138, 139, 144 pigment, 128–131 Biological materials blood serum and zein, 279, 280–281 chiral rotation, hydrogen bonding and nanoengineering, 290–291 chlorophyll, 279 cholesterol, 275–277 DNA, 273–275 HSP characterization of, 270–271 human skin, 250, 277–279 hydrophobic bonding and h ydrophilic bonding of, 271–273 lard, 277 lignin, 279, 282–283, 284–288, 316–317 surface mobility, 290 urea, 274, 283, 289 water, 21–22, 289 wood chemicals and polymers, 279, 281 BISOM test, 170–173 Bitumens asphaltenes in, 153, 154–155 BISOM test, 170–173 calculation and plotting of Hansen 3D pseudosphere, 161–164 components of, 164–166 Hildebrand solubility parameters, 156, 158 HSP, 158–164 hydrocarbons in, 155–156, 164–166 models of, 152–154 polymer modified, 15 polymers and, 166–169 production process, 151–152 RED numbers, 160 solubility parameters of, 155–156 solubility SPHERE, 159–164 solvents used for determination of solubility of, 161–164 temperature effect on, 168 testing solubility of, 156, 157 turbidimetric titrations, 170

511

7248_Index.fm Page 512 Wednesday, May 9, 2007 10:49 AM

512 uses, 151, 160 Blanc fi e, 131 Blistering of coatings, 141, 281 Blood serum, 279, 280–281 Boundaries of solubility, 7–8 Breakthrough times, HSP correlations of, 245–247 Butadiene block copolymer (SBS), 168 Butyl rubber, 246

C Calculations versus experimental χ12 parameters, 34–39 polymer HSP, 97–98 Carbon dioxide, 57, 252 Hildebrand parameters of, 187–190 HSP of, 336–337 ideal solubility of g ases in liquids and, 199–201 pressure effects on solubility parameters of, 187, 191–196 solubility data in v arious solvents, 178–186, 187 solubility in liquid solv ents, 185–186 solubility parameters, 141–142 temperature effects on solubility parameters of, 190–191 three-component solubility parameters, 189–190 Carbon disulfide, 16–1 Carbon fiber sur ace characterization, 131–132, 133 Carbon tetrachloride, 35–36, 204 Carcinogens, 312 Castor oil, 214 Cellulose, 281, 289 acetate, 303 Challenge® 5100, 246 Chemical protective clothing, 315 Chemical resistance acceptable-or-not data on, 232 effects of solv ent molecular size on, 232–233 HSP characterization of, 231–232, 339 plastics, 234–237 special effects with w ater, 238–239 tank coatings, 233–234 tensile strength and, 237–238 Chiral rotation and HSP, 290–291 Chlorobenzene, 245, 305 Chloroform, 35–36 Chlorophyll, 279 Cholesterol, 275–277 Cleaning solvents application of HSP methodology to, 212–213 azeotropes, 212, 213–215, 216–226 calculating HSP of composites and, 205–206 HSP of multiple-component soils and, 204–205 identification of designe , 208 method for choice of suitable, 206–208 pathology of soils and, 204 performance, 210–212 reference soils, 208, 209–210 variety of chemicals used in, 204

Hansen Solubility Parameters: A User’s Handbook Coatings amines added to, 140 binders, 138, 139, 144 blistering in, 141, 281 chemical resistance and, 231–239, 240 gases as solv ents in, 141–142 for hazardous materials, 312 PET film, 234, 23 polymer compatibility in, 145–146 RED numbers, 142–143, 144 self-stratifying, 120–121 solvents, 137–142 and surface phenomena in, 144–145 SPHERE and, 143 surface mobility, 333–334 tank, 233–234 water-reducible, 139–140 Cohesion energy parameters, 2, 5, 13–18, 113 Cohesive energy density, carbon dioxide, 187 Cohesive energy difference, 27 polymer segments and solv ents, 30–32 Composite soils, 204–206 Compound formation, 9 Concentration-dependent diffusion, 244–245, 297–298, 304–305 Constant diffusion coefficients, 296–29 Controlled adsorption, 132–134 Controlled release of drugs, 339–340 Cooling, rapid, 19 Corresponding states theory (CST), 3, 6 HSP and, 30–32 predicting the χ12 parameter using, 28 Critical chi parameter, 32 Critical surface tensions, 116–117 Crude oils BISOM test, 170–173 HSP of molecules in, 169–170 origins and structure, 152–153 uses, 151–152 Cyclic olefinic copolymer (COC), 120, 300–30 Cyclohexanone, 185

D Danish MAL system, 313–315 Data, HSP availability, 321–322 quality, 324, 326 Dendrimers, 90–91 Designer solvents. See also Cleaning solv ents analysis of capability of, 213–215 application of HSP methodology to, 212–213 azeotropes, 212, 213–215, 216–226 cleaning performance of, 210–212 defined, 20 identification of, 20 variety of chemicals used in, 204 Desirability function, 20 Dextran C, 101–102, 322, 323–324, 325–326 Dibasic esters (DBE), 315

7248_Index.fm Page 513 Wednesday, May 9, 2007 10:49 AM

Index 1,2-dibromoethane, 185 Dichloromethane, 185 Diethyl ether, 17, 262 Diffusion anomalous, 306–308 concentration-dependent, 244–245, 297–298, 304–305 constant coefficients, 296–29 equations, 294–296 film formation by sol ent evaporation and, 305–306 side effects, 304–305 steady state permeation, 296 surface effects and, 302–304 surface resistance and, 298–305 time-dependent, 308 vapor, 301–302 Dilution, infinite, 85–8 Dimerization, acid, 65 Dimethyl sulfoxide, 289 1,4-dioxane, 262 Dipolar interactions, 50–52, 66 Dipole moments, 8–9 zero, 16–17 Dispersion solubility parameters, 5, 13–16 hydrogen bonding, 52–59 Distance, solubility parameter, 7 DNA, 273–275 Dynamic shear rheometer (DSR), 152

E Elastomers, 245, 333 Electron exchange parameters, 5, 29 Electrostatic repulsion, 140 Energy density, cohesive, 13–16 Gibbs, 92–93 polar cohesive, 5 relative difference, 8 surface free, 113–114 of vaporization, 13–16 Entropic group contrib ution model, 78–79, 83 Environmental impact of solv ents, 208 Environmental stress cracking (ESC), 107, 120 incidence of, 259–260 interpreted using HSP, 260–262 mechanism for, 259–260, 264–267 with nonabsorbing stress cracking initiators, 263–264 RED number and, 260–262 EPDM rubber, 141 Epon® 1001, 233 Epoxies, 138, 141, 233–234 Equation-of-state framework, 46–50, 60–62 carbon dioxide, 188–189 Ethanol, 4–5, 276 Ethyl acetate, 262 Ethylene glycol monometh yl ether (EGMME), 245 Ethylene vinyl alcohol copolymers (EV OH), 253–254 Evaporation, solvent, 305–306

513

F Fickian diffusion, 306–308 Field ionization mass spectrometry (FIMS), 154 Fillers and fibers, 147–14 Film formation by solv ent evaporation, 305–306 Films, PET, 234, 235 First-order and second-order groups in solubility parameter prediction, 66–73 Floating rates of particles in pigments, 126–127 Flocculation, 153 Flory-Huggins model, 27–28, 77, 341 versus the GC methods, 84–90 for multicomponent mixtures, 92–93 regular solution theory and, 80–82 using Hildebrand and HSP, 82–84 Fluoropolymers, 16–17 Formamide, 274 Fragrances and aromas, 338–339 Free-volume-based models for polymers, 77–79 solvent selection for paints, 85–88 Fresh Air Numbers (FAN/MAL) system, 313–315

G Gases characterization problems, 336–337 HSP of, 336–337 ideal solubility in liquids, 199–201 permeation through polymers, 243 solubility parameter correlation of permeation coefficients fo , 251–254 as solvents, 141–142 Gibbs energy, 92–93 Glycol ethers, 138, 140–141 Group-contribution method applications, 76–77 entropic model, 78–79, 83 Flory-Huggins/Hansen model v ersus, 84–90 Flory-Huggins model and re gular solution theory, 80–82 free-volume-based models for polymers, 77 reliability of calculation procedures for , 328 UNIFAC model, 76–78, 83 Guggenheim expression, 47

H Halar®, 235, 236 Hansen 3D pseudosphere, 161–164 Hansen solubility parameters (HSP) acid-base theories and, 130–131 aromas and fragrances, 338–339 biological materials, 269–291 bitumen, 158–164 blood serum and zein, 279, 280–281 carbon dioxide, 177–201 for chemical protecti ve clothing, 315

7248_Index.fm Page 514 Wednesday, May 9, 2007 10:49 AM

514 chemical resistance characterization using, 231–239, 240, 339 chiral rotation and, 290–291 chlorophyll, 279 cholesterol, 275–277 coatings, 145–146 composite soils, 204–206 controlled release drugs, 339–340 correlation of permeation coef ficients for ases, 251–254 correlation of permeation rates, 248–250 correlation of polymer swelling, 250 correlation of zeta potential for blanc fi e, 131 correlations for chemical resistance, 234–236 correlations of breakthrough times, 245–247 correlations with surf ace tension, 113–114 corresponding states theories and, 30–32 data availability, 321–322 quality, 324, 326 defined, 1, 4– Dextran C, 322, 323–324, 325–326 electron exchange energy and, 29 environmental stress cracking (ESC) and, 260–267 fillers and fibers, 147–1 first-order and second-order groups contri ution to, 66–73 Flory-Huggins model and, 81–82 gases, 336–337 GC methods v ersus the FH/, 84–90 human skin, 277–279, 316 hydrogen bonding and, 290–291 inorganic salts, 337 lard, 277 lignin, 279, 282–283, 284–288 methodology applied to cleaning operations, 212–213 methods and problems in determination of, 6–13 mixtures, 205–206 molecules in crude oils, 169–170 multiple-component soils, 204–206 nanoengineering and, 290–291, 340–341 organic compounds, 177–178 organic salts, 337 organometallic compounds, 338 pigments, 128–131 fillers and fibers, 125–1 plastics, 315–316, 339 polymers, 29–30, 95–109 solvents, 29–30, 40–41, 137–143 special effects with w ater, 238–239 surface characterization, 113–122, 131–132, 330–332 liquid adhesion, 342 mobility and, 263–264, 290 tensile strength and, 237–238 three-component, 189–190 urea, 283, 289 water, 21–22, 289, 326, 334–336 wood chemicals and polymers, 279, 281 χ12 parameter and, 32–39

Hansen Solubility Parameters: A User’s Handbook Hazardous materials chemical protective clothing for, 315 Danish MAL system, 313–315 skin penetration of, 316 solvent formulation and personal protection for least risk from, 313 substitutions for, 311–312 transport phenomena, 316–317 uptake by plastic containers, 315–316 Heat of v aporization, 50 Hemicelluloses, 281, 289 Henry's law coefficient, 17 Hexane, 116, 120, 204 Hildebrand parameters, 2–4 of bitumen, 155, 156, 158 of carbon dioxide, 187–190 solvent selection for paints, 85–88 χ12 coefficient and, 28, 3 Hoy dispersion parameter, 9, 13 Human skin HSP of, 277–279, 316 penetration of hazardous materials, 316 as polymeric barrier, 250 Hydrocarbons in bitumen, 155–156, 164–166 carbon dioxide solubility parameters and, 192–194, 336–337 coatings applications, 138–139 and corresponding states theories, 29 in crude oils, 153 ether solvents partially based on silicon, 208 Hydrochlorides, 339–340 Hydrofluoroethers (HFEs), 20 Hydrogen bonding, 5, 7 acid dimerization and, 65 coatings, 138 cohesive energy differences and, 31–32 defined, 26 dipole moments and, 9 DNA, 273–275 HSP and, 290–291 lattice-fluid, 4 nonrandom, 46, 50–59 polymer surfaces, 115 self-stratifying coatings, 120–121 solubility parameters, 17, 52–59 techniques for data treatment, 142–143 temperature increases and, 140–141 water temperature and, 141 Hydrophilic bonding, 133–134, 271–273, 333–334 Hydrophobic bonding, 271–273, 332

I Inorganic salts, 337 Interaction radius, 184 Internal pressure, 187 Intrinsic viscosity of polymers, 107–109

7248_Index.fm Page 515 Wednesday, May 9, 2007 10:49 AM

Index

515

J

N

Joffé effect, 342

Nanoengineering, 290–291, 340–341 n-butanol, 138, 140 n-butyl acetate, 262 Neoprene®, 235 New Flory theory, 27–28 n-heptane, 153, 154, 165–166 n-hexane, 243, 262 Nitrates, 337 Nitrile rubber, 246 Nitromethane, 4–5 N-methyl-morpholine-N-oxide, 279 N-methyl-2-pyrrolidone, 204 Nonabsorbing stress cracking initiators, 263–264 Nonbiodegradable substances, 312 Nonpolar interactions in or ganic materials, 5 Nonrandom hydrogen bonding, 46, 50–59 Normal diffusion, 306–308 Nylon-6, 303

K Keratin, 277, 317 Ketones, 138, 140

L Laminates, 253–254 Lard, 277 Lattice-fluid ydrogen bonding (LFHB), 46 Lignin, 279, 282–283, 284–288, 316–317 Liquid-liquid phase separation (LLE), 75 Liquids absorption by plastic containers, 315–316 adhesion phenomena, 342 cohesion energy parameters and surf ace characterization of, 114–116 environmental stress cracking (ESC) and, 262 first-order and second-order groups in, 66–7 ideal solubility of g ases in, 199–201 solubility parameters for, 17–21, 32–33, 52–59 solvents, CO 2 solubility in, 185–186 spontaneous spreading, 115–118 surface tension, 116–117 Lithium, 338 Lorenz-Berthelot mixtures, 41 Low density polyeth ylene (LDPE), 249 Lower critical solution temperature, 28 Lydersen group constants, 12–13t, 15

O Octyl alcohol, 277 Oleic acid, 185 Olive oil, 277 One-component Hildebrand parameter as a function of temperature and pressure, 187–189 Organic compounds, HSP and solv ents for, 177–178 Organic salts, 337 Organoclays, 340–341 Organometallic compounds, 338 Ostwald coefficient, 17 Ozone depletion potentials (ODP), 208

M

P

MAL system, Danish, 313–315 Maltenes, 153, 164–165 MATLAB platform, 163 Melting points of polymers, 106–107 Metallic bonding, 338 Methanol, 19, 141, 238–239 2-methylcyclohexanone, 185 Methylene chloride, 140 Methyl ethyl ketone, 140 Methyl iodide-cellulose acetate, 303 Methyl isobutyl ketone, 262 Methyl methacrylate, 277 Mixtures, solubility parameters of, 205–206, 283, 289 Mobility, surface, 263–264, 290, 333–334 Molar volume and solubility parameters, 7 Molecular size and solubility parameters, 19–20, 32–33, 232–233, 243 Morpholine, 108 Multicomponent mixtures, Flory-Huggins model for , 92–93 Multiple-component soils, 204–206 Multivariable analysis and solubility parameters, 8

Paints hydrophilic bonding in, 272 older, dried, 142–143 self-stratifying coatings, 120–121 solvent selection for, 85–88 Partial solubility parameters determination, 6–13 dipolar interactions and, 50–52 for liquids, 17–21, 52–59 Particle suspensions, 126–127 Pathology of soils, 204 P-dioxane, 303 Peat moss, 333 Pentachlorophenol, 316 Permeation of liquid or g as through polymers, 243 rates, 248–250 size and shape and dependence of diffusion coefficient, 255–256 solubility parameter correlation of permeation coefficients for ases, 251–254 solubility parameter correlations based on, 245–250 steady state, 296

7248_Index.fm Page 516 Wednesday, May 9, 2007 10:49 AM

516 surface resistance in, 301–302 through human skin, 277–279, 316 PET film coating, 234, 23 Phase equilibria, mix ed solvent-polymer, 88–90 Phenyl acetate, 262 Physical adhesion maximization, 122 Pigments, fillers, and fibe binders, 128–131 carbon fibe , 131–132, 133 cohesion parameter characterization study , 125 controlled adsorption, 132–134 manufacturing, 128 organic versus inorganic, 130–131 sedimentation rates, 126–127, 129 solvents, 128–131, 144–145 surface characterization methods, 126–127 Plasticizers, 21, 142, 244 Plastics absorption of chemicals in, 315–316, 339 bitumen in, 167 chemical resistance, 234–237 containers uptake, 315–316 environmental stress cracking and, 107 HSP correlations acceptable or not, 234–237 SWEAT and, 141 Plastomers, 167 Points, polymers as, 328, 330 Polar bonding and cohesi ve energy acid groups, 132–134 differences, 31–32 Polarity of asphaltenes, 155 Polar solubility parameters, 5, 16–17 Polyacrylonitrile (PAN), 39, 100, 132, 252 Polyamides (PA), 253–254, 260 Polybutadiene, 35–36 Polycarbonate (PC), 120, 260, 300 Poly(chlorotrifluoroet ylene), 252 Polydimethyl siloxane (PDMS), 77 Polydimethylsiloxanes, 264 Polyesterimide, 108–109 Polyesters (TPU), 333 Polyether ether k etone (PEEK), 260 Polyether/polyamide block copolymer (PEB A), 333 Polyether sulfone, 141 Polyethersulfone (PES), 167–168, 238 Poly(ethylene co-chlorotrifluoroet ylene) (ECTFE), 235, 236 Polyethylene (PE), 118, 246, 260, 303 low density, 249 Polyethylenesulfide, 167–168, 327–328, 329–330 Polyethylene terephthalate (PET), 252, 253–254 Polyisobutylene, 36–38 Polymer modified bitumen (PMB), 15 Polymers barrier concentration-dependent diffusion and, 244–245 human skin as, 250, 316 laminated, 253–254 RED numbers, 247, 253 solubility parameter correlation of permeation coefficients for ases, 251–254

Hansen Solubility Parameters: A User’s Handbook solubility parameter correlation of permeation rates, 248–250 solubility parameter correlations based on permeation, 245–250 swelling, 250 bitumen and, 166–169 calculated versus experimental χ12 parameters, 34–39 coatings, 139–140 cohesive energy differences between solv ents and, 30–32 compatibility in coatings, 145–146 concentration and χ12, 39 concentration-dependent diffusion of, 244–245 dendrimer, 90–91 environmental stress cracking, 107 experimental evaluation of beha vior of, 95–97 Flory-Huggins models using Hildebrand and HSP , 82–84 group contribution applications, 76–77 entropic model, 78–79 Flory-Huggins model and re gular solution theory, 80–82 free-volume-based models for polymers, 77 UNIFAC-FV model, 77–78 high temperature solv ents, 140–141 HSP for, 29–30 calculation of, 97–98 solubility and, 98–106 SPHERE program and, 95–96, 99–106, 108–109 two-dimensional plot of, 96–97 intrinsic viscosity of, 107–109 melting point determinations, 106–107 permeation of liquid or g as through, 243 as points, 328, 330 RED number, 98–100 rigid, 245 solubility, 341 solution thermodynamics, 2–4, 19, 75–76 -solvent mixtures phase equilibria, 88–90 SPHERE program and, 95–96, 99–106, 108–109 swelling of, 106 wood, 279, 281 Polymethyl methacrylate (PMMA), 243, 260 Polyolefins, 252, 253–25 Polyphenylene sulfide (PPS), 20, 141, 23 Polypropylene, 59, 106, 302 Polystyrene (PS), 38, 260 Polytetrafluoroet ylene (PTFE), 252, 260 Polyurethanes, 333 Polyvinylacetate, 39 Polyvinyl acetate, 243 Polyvinylacetate, 305 Polyvinyl alcohol, 252 Polyvinyl chloride (PVC), 116–117, 243 Polyvinylchloride (PVC), 260 Polyvinyl chloride (PVC), 264–267 Polyvinylidine chloride (PVDC), 106–107 Pressure effects on carbon dioxide solubility parameters, 187, 191–196 Pressure-volume-temperature (PVT) data, 52–59

7248_Index.fm Page 517 Wednesday, May 9, 2007 10:49 AM

Index Prigogine corresponding states theory (CST), 3, 6 HSP and, 30–32 predicting the χ12 parameter using, 28 Prigogine-Patterson theory, 31, 40–41 Propylene bromide, 185 Protective clothing, chemical, 315 Proteins in blood serum, 279, 280–281 Pyrimidine, 275

Q Quasi-chemical condition, 47–49

R Rapid cooling, 19 Rates, permeation, 248–250 RED (relative energy difference) numbers, 8, 21, 40 barrier polymers, 247, 253 bitumen, 160 carbon dioxide, 185 coatings, 142–143, 144 environmental stress cracking and, 260–262 polymer solubility and, 98–100 solvent quality and, 142 transport phenomena and, 316 Reference soils, 208, 209–210 Refrigerants, 208 Regular solution theory and Flory-Huggins model, 80–82 Rehbinder effect, 120, 342 Relative energy difference, 8 Resins, 153 Resistance, surface in absorption e xperiments, 300–301 mathematical background, 298–299 measuring diffusion coefficients with, 304–30 in permeation e xperiments, 301–302 skin layer ef fect, 302–303 Rhodamin FB, 331, 334–335 Rigid polymers, 245 RNA, 274

S Safety and health issues. See Hazardous materials Salts inorganic, 337 organic, 337 SARA analysis, 153 Sedimentation rates, 126–127, 129, 130 Self-assembly (controlled adsorption), 132–134, 341 Self-stratifying coatings, 120–121 Side effects, 304–305 Silicon, hydrocarbon ether solv ent partially based on, 208 Similia similibus solvuntur, 45–46 Skin, human. See Human skin Skin layer ef fects, 302–303 Soils composite, 204–206, 213–215

517 HSP of multiple-component, 204–205 method for choice of suitable solv ents for, 206–208 mixtures with solv ents, 205–206 pathology of, 204 reference, 208, 209–210 Solubility parameters. See also Hansen solubility parameters (HSP); Hildebrand parameters of bitumen, 155–161 boundaries and, 7–8 carbon dioxide, 141–142 dipolar interactions and, 50–52 dispersion, 13–16 distance equation, 7 first-order and second-order groups contri ution to, 66–73 hydrogen bonding, 17 introduction to, 1–2 for liquids, 17–21 partial, 6–13, 17–21 polar, 5, 16–17 polymers, 341 of solvent-solute combinations, 3–4 study of, 45–46 surface science, 6 temperature dependence and, 18–19 for water, 21–22 Solubility sphere, 184 Solvents absorption, 300–301 alternatives, 312 azeotropes, 212, 213–215, 216–226 biological materials as, 274–275 carbon dioxide solubility in v arious, 178–186, 187 cleaning analysis of capability of, 213–215 application of HSP methodology to, 212–213 calculating HSP of mixtures of soil and, 205–206 HSP of multiple-component soils and, 204–205 method for choice of suitable, 206–208 pathology of soils and, 204 performance, 210–212 reference soils, 208, 209–210 variety of chemicals used in, 204 coating applications, 137–148 cohesive energy differences between polymers and, 30–32 diffusion coefficients, 244–24 environmental impact of, 208 evaporation film formation, 305–30 formulation and personal protection for least risk, 313 gases as, 141–142 for hazardous materials, 312 high temperature, 140–141 HSP for, 29–30, 32–33, 40–41, 137–142 for human skin, 277–279 hydrocarbon base, 138–139, 208 for lard, 277 mixtures with soils, 205–206 molecular size and solubility parameters, 19–20, 232–233 permeation rates, 248–250

7248_Index.fm Page 518 Wednesday, May 9, 2007 10:49 AM

518 pigments adsorption and, 128–131, 144–145 polymer compatibility in coatings, 145–146 -polymer mixtures phase equilibria, 88–90 selection for paints, 85–88 solubility of bitumens in, 156, 157–158 solute-combination solubility parameters, 3–4 as spheres, 328, 330 and surface phenomena in coatings, 144–145 titration tests, 170–173 used as refrigerants, 208 used for determination of solubility of bitumens, 161–164 vegetable oils in, 142 viscosity, 145–146 SPHERE and SPHERE1 analysis, 20–21 bitumen solubility and, 159–164 cholesterol, 275–276 coatings and, 143 Hansen 3D pseudo-, 161–164 polymers and, 95–96, 99–106, 108–109 Spheres, solvents as, 328, 330 Spontaneous spreading of liquid droplets, 115–118, 120 Steady state permeation, 296 Stress corrosion cracking (SCC), 260 Styrene, 277 Styrene-acrylonitrile copolymer (SAN), 260 Styrene/butadiene/styrene block copolymer (SBS), 333 Substitutions, 311–312 Sulfur trioxide, 253 Super Case II beha vior, 306–308 Surfaces active agents, 332–333 -active agents in coatings, 140 carbon fibe , 131–132 characterizations and comparisons using HSP,118–120, 131–132, 330–332 cohesion energy parameters e valuation for, 114–116 film formation by sol ent evaporation and, 305–306 free energy, 113–114 HSP for, 113–122 liquid adhesion phenomena, 342 mobility, 263–264, 290, 333–334 phenomena in coatings, 144–145 physical adhesion maximization, 122 pigment, fillers, and fib , 126–131 resistance, 298–305 self-stratifying coatings, 120–121 tension correlations with HSP, 113–114 tensions, critical, 116–117 wetting tension, 115–116, 117–118 Surfactants, 332–333 SWEAT (soluble water exuded at lower temperatures), 141, 238–240 Swelling polypropylene, 106 solubility parameter correlation of polymer , 250

Hansen Solubility Parameters: A User’s Handbook

T Tank coatings, 233–234 Tar, 152 Temperature bitumen fl xibility and, 168 chiral rotation and, 290 dependence and solubility parameter calcuation, 18–19 effect on chemical resistance, 235 effects on solubility parameters of carbon dioxide, 190–196 hydrocarbon solvents and, 337 hydrogen-bonding parameters in coatings and, 140–141 melting point of polymers and, 106–107 one-component Hildebrand parameter as a function of pressure and, 187–189 SWEAT phenomena, 141, 238–240 water absorption v ersus, 238–240 Tensile strength, 237–238 Tetrabromobisphenol A (TBBPA), 316–317 Tetrahydrofurane, 290 Tetrahydrofuran (THF), 59, 140 Thermoplastic elastomers (TPE), 333 Three-component solubility parameters, 189–190 Thymine, 275 Titration BISOM test, 170–173 turbidimetric, 170 Toluene, 245, 277, 290, 314 Toxic substances, 312 Transport phenomena, 316–317 Trichloroethylene, 253, 277 Trichloromethane, 185 Tricresyl phosphate, 214 Turbidimetric titrations, 170

U ULTEM® 1000, 237–238 UNIFAC model, 76–78, 83 Urea, 274, 283, 289

V Van der Waals volume, 77, 79 Van Krevelen dispersion parameters, 13 Vapor degreasing operations, 210–212 Vapor diffusion, 301–302 Vapor phase osmometry (VPO), 154 Vegetable oils in solv ent cleaners, 142 Vinyl acetate, 243 Vinyl chloride, 243 Viscosity crude oil, 169–170 intrinsic, 107–109 solvent, 145–146 V iton®, 106, 246, 250 Volatile organic compounds (VOC), 137, 142

7248_Index.fm Page 519 Wednesday, May 9, 2007 10:49 AM

Index

W Water absorption versus temperature, 238–240 amino acid side chains and, 270–271 characterization problems with, 334–336 chemical resistance and HSP correlations with, 238–239 as a cleaning solv ent, 204 HSP for, 21–22, 289, 326 hydrogen-bonding parameter and temperature of, 141 polymer softening using, 244–245 -reducible coatings, 139–140 surface mobility, 263–264, 290 urea mixtures, 283, 289 Wetting tension, 115–116, 117–118 Wood chemicals and polymers, 279, 281, 289

X χ12 parameters

519 comparison of calculated and e xperimental, 34–39 defined, 27–2 HSP and, 32–39 polyacrylonitrile, 39 polybutadiene, 35–36 polyisobutylene, 36–38 polymer concentration and, 39 polystyrene, 38 polyvinylacetate, 39 scatter, 39–40 X-ray photoelectron spectroscop y (XPS) analysis, 132 Xylene, 138, 140, 314

Z Zein, 279, 280–281 Zero dipole moments, 16–17 Zeta potential, 130, 131

7248_Index.fm Page 520 Wednesday, May 9, 2007 10:49 AM