Electroanalytical Chemistry: New Research

ELECTROANALYTICAL CHEMISTRY: NEW RESEARCH No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

ELECTROANALYTICAL CHEMISTRY: NEW RESEARCH

GRAHAM M. SMITHE Editor

Nova Science Publishers, Inc. New York

Copyright © 2008 by Nova Science Publishers, Inc.

All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS.

Library of Congress Cataloging-in-Publication Data Electroanalytical chemistry : new research / Graham M. Smithe (editor). p. cm. ISBN 978-1-60741-857-3 (E-Book) 1. Electrochemical analysis--Research. I. Smithe, Graham M. QD115.E5112 2008 543'.4--dc22 2007052758

Published by Nova Science Publishers, Inc.

New York

CONTENTS Preface

vii

Expert Commentary: Electrodes Based on Metallophthalocyanines Integrated with Carbon Nanotubes: Potential Hybrids for Enhancing Electron Transport Kenneth I. Ozoemena Chapter 1

Chapter 2

Chapter 3

Chapter 4

Index

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies Peter Englezos, John Ripmeester and Robin Susilo

1

9

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate. On the Concept of Time of Wetness and Its Interaction with Contaminants Deposition F. Corvo, T. Pérez, Y. Martin, J. Reyes, L.R. Dzib, J.A. González and A. Castañeda

61

The Application of D-Statistics Based Tests of Randomness, Independence, and Trend to Electrochemical Observations Thomas Z. Fahidy

93

Self-Assembly Assisted Polypolymerization (SAAP): A Novel Approach to Prepare Multiblock Copolymers with a Controllable Chain Sequence and Block Length Liangzhi Hong, Fangming Zhu, Guangzhao Zhang, To Ngai and Chi Wu

109

123

PREFACE Electrochemistry can be broadly defined as the study of charge-transfer phenomena. As such, the field of electrochemistry includes a wide range of different chemical and physical phenomena. These areas include (but are not limited to): battery chemistry, photosynthesis, ion-selective electrodes, coulometry, and many biochemical processes. Although wide ranging, electrochemistry has found many practical applications in analytical measurements. The field of electroanalytical chemistry is the field of electrochemistry that utilizes the relationship between chemical phenomena which involve charge transfer (e.g. redox reactions, ion separation, etc.) and the electrical properties that accompany these phenomena for some analytical determination. This new book presents the latest research in this field. Expert Commentary - Carbon nanotubes (CNTs), notably single-walled (SWCNTs) or multi-walled (MWCNTs), have continued to attract immense research interests in electroanalytical chemistry because of their ability to exhibit unusual but excellent electrical conductivity and mechanical properties. They also possess modifiable sidewalls and openends, making them suitable for use as new materials for constructing efficient electrocatalysts and electrochemical sensors. Transition metal metallophthalocyanine (MPc) complexes have also proved themselves as powerful redox-active materials for modifying electrode surfaces for use as sensors. My group has been engaged in the rational design and integration of CNTs with certain transition MPc complexes, and exploring their potential applications in the fabrication of electrochemical responsive CNT-MPc based electrodes. This short commentary gives insights into the recent developments in this emerging research area as well as future trends. Chapter 1 - Clathrate or gas hydrates are non-stoichiometric crystalline materials formed by the inclusion of certain molecules into a framework of hydrogen-bonded water molecules under suitable temperature and pressure conditions. The resulting host-guest networks consist of cavities formed by water molecules enclosing the guest molecules. Typical guests include light hydrocarbon gases, carbon dioxide, hydrogen sulfide, hydrogen and nitrogen. The basic cavity formed by water molecules through hydrogen bonding is the pentagonal dodecahedron (512). This cavity is common to the three best known hydrate structures (I, II and H). A unit cell of structure I hydrate has 46 water molecules forming two dodecahedral (512) and six tetrakaidecahedral (51262) cavities. A unit cell of structure II has 136 water molecules forming 16 (512) and eight hexakaidecahedral (51264) cavities. The Structure H hydrate unit cell has 34 water molecules, three 512 cavities, two different dodecahedral cavities which have threesquare faces, six-pentagonal faces and three-hexagonal faces (435663) and a larger

viii

Graham M. Smithe

iscosahedral cavity with twelve pentagonal faces and six hexagonal faces (51268). It should be noted that unlike structures I and II which may form with a single guest species, structure H requires the presence of a small guest (like methane) and a large molecule guest substance (LMGS) like neohexane at ordinary pressures. Gas hydrates were first reported at the beginning of the 19th century, and until the 1930s they remained a scientific curiosity. At that time it was realized that hydrates were more likely to be the causative agent in blocking pipelines than ice. Today, gas hydrate control continues to be a problem in the oil and gas industry. In the 1960’s it was realized that natural gas hydrates are present in the geo-sphere with worldwide reserves estimated at 10,000 to 40,000 trillion cubic meters (TCM). Considerable efforts are underway to refine global estimates and to develop technology and exploit this resource. On the other hand these hydrates may decompose as a result of global warming or seafloor instability and release the methane gas. There is speculation that a runaway greenhouse effect could result, with some evidence for changes of such magnitude in the global paleoclimate (15,000 and 55 million years ago). Application of clathrate hydrate crystallization offers the possibility of the development of innovative technologies for natural gas storage and transportation, hydrogen storage and gas separation with applications for carbon dioxide capture from flue gas (CO2/N2/O2) or fuel gas (CO2/H2) mixtures. Clathrate hydrates have become an important research area spanning a variety of disciplines. It is of continuing great interest to chemical engineers who have played a key role in past developments. This report discusses areas where chemical engineers can advance the knowledge frontier. Chapter 2 - Atmospheric corrosion is the most extended type of corrosion in the World. Over the years, several papers have been published in this subject; however, most of the research has been made in non-tropical countries and under outdoor conditions. Results of outdoor and indoor corrosion rate and corrosion aggressivity in tropical corrosion test stations of Cuba and Mexico are reported. Time of wetness (TOW), considered as the time during which the corrosion process occurs, is an important parameter to study the atmospheric corrosion of metals. According to ISO-9223 standard, TOW is approximately the time when relative humidity exceeds 80% and temperature is higher than 0oC. No upper limit for temperature is established. In tropical climates, when temperature reaches values over 25oC, evaporation of water plays an important role and the possibility to establish an upper limit respecting temperature should be analyzed. The concept of TOW assumes the presence on the metallic surface of a water layer; however, there are recent reports about the formation of water microdrops during the initial periods of atmospheric corrosion, showing that the idea of the presence of thin uniform water layers is not completely in agreement with the real situation in some cases (particularly indoor exposures). Most of the research carried out to study the initial stages of atmospheric corrosion have been made on a clean surface without corrosion products; however, the metal is very often covered by thin or thick corrosion products after a given exposure time and these products usually act as retarders of the corrosion process. In the Cuban Isle, the influence of chloride ions is very significant in determining the corrosion rate. In the coastal territory of the Mexican Gulf, particularly at Campeche, the deposition of Chloride ions is lower. No previous reports have been made about the interaction between chloride deposition rate and rain. The influence of rain seems to be

Preface

ix

important in determining the acceleration rate of chloride ions on metals due to its washing effect. To consider the influence of the interaction chloride deposition rate–rain regime could be useful to improve the prognosis of corrosion aggressivity. The predominant wind direction corresponding to geographic sites result in an important parameter for chloride deposition and their influence on surface wetness. The calculation of Time of Wetness established in ISO 9223 should be revised based on new results obtained in outdoor and indoor conditions in tropical humid marine climate. Some proposals are made to improve the estimation of TOW, taking into account changes in its nature depending on outdoor or indoor exposure, linear relationship between time and TOW, the effect of rain, and the role of contaminants and air temperature. Chapter 3 - Randomness, independence and trend (upward, or downward) are fundamental concepts in a statistical analysis of observations. Distribution-free observations, or observations with unknown probability distributions, require specific nonparametric techniques, such as tests based on Spearman’s D – type statistics (i.e. D, D*, D**, Dk ) whose application to various electrochemical data sets is herein described. The numerical illustrations include surface phenomena, technology, production time-horizons, corrosion inhibition and standard cell characteristics. The subject matter also demonstrates cross fertilization of two major disciplines. Chapter 4 - Block copolymers have attracted much attention because of their novel properties and various promising potential applications. However, it is still difficult, if not impossible, to prepare multiblock copolymers with a controllable chain sequence and block length even though a variety of synthetic methods, such as anionic and controlled free radical living polymerization have been advanced. In recent years, we have proposed and developed a novel method of using the self-assembly of A-B-A triblock copolymers in a solvent which is selectively good for the two A-blocks. Such self-assembly concentrates and exposes the active groups attached on the two A-block ends so that they can be coupled together to form a long multiblock copolymer chain with its sequence and block length controlled by the initial triblock copolymer. In this review, we first illustrate how the SAAP concept was developed and exemplified in some real copolymer systems. Furthermore, we compare the coupling efficiency with and without the self-assembly, and demonstrate that SAAP provides an elegant way to prepare long multiblock copolymers.

In: Electroanalytical Chemistry: New Research Editor: G. M. Smithe

ISBN: 978-1-60456-347-4 © 2008 Nova Science Publishers, Inc.

Expert Commentary

ELECTRODES BASED ON METALLOPHTHALOCYANINES INTEGRATED WITH CARBON NANOTUBES: POTENTIAL HYBRIDS FOR ENHANCING ELECTRON TRANSPORT Kenneth I. Ozoemena* Department of Chemistry, University of Pretoria, Pretoria 0002, South Africa.

Abstract Carbon nanotubes (CNTs), notably single-walled (SWCNTs) or multi-walled (MWCNTs), have continued to attract immense research interests in electroanalytical chemistry because of their ability to exhibit unusual but excellent electrical conductivity and mechanical properties. They also possess modifiable sidewalls and open-ends, making them suitable for use as new materials for constructing efficient electrocatalysts and electrochemical sensors. Transition metal metallophthalocyanine (MPc) complexes have also proved themselves as powerful redox-active materials for modifying electrode surfaces for use as sensors. My group has been engaged in the rational design and integration of CNTs with certain transition MPc complexes, and exploring their potential applications in the fabrication of electrochemical responsive CNT-MPc based electrodes. This short commentary gives insights into the recent developments in this emerging research area as well as future trends.

1. Introduction Phthalocyanine complexes are organic macrocycles with 18 π-electrons, structurally resembling the naturally-occuring porphyrins complexes [1-3]. Electrodes modified with transition metal (notably Fe, Co, Mn, Ni) phthalocyanine (MPc, Fig.1) complexes have continued to generate immense research interests because of their well-established electrocatalytic properties [3-6]. *

E-mail address: [email protected], Tel.: +27-12-420-2515; Fax: +27-12-420-4687 (Corresponding author)

2

Kenneth I. Ozoemena

R

R

Peripheral position N N

N M

N N

Non-peripheral position N

N N

R

R

Figure 1. Molecular structures of metallophthalocyanine (MPc) complex. MPc complexes may be substituted at the peripheral or non-peripheral positions as shown. M = metal ion, and R = substituent containing terminal functional groups such as –NH2 and –OH.

Carbon nanotubes (CNTs), notably single-walled (SWCNTs) or multi-walled (MWCNTs), have been receiving some attention amongst electroanalysts as electrode modifiers that are capable of enhancing electrochemical response [7-10]. Interestingly, both CNT and MPc contain π-electrons, thus facilitating their integration via π-π integration. Recent reports have indicated that MPc integrated with CNTs enhance electrochemical responses of MPc-based electrodes toward the detection of certain important analytes such as the hydrolysis products of V-type nerve agents [11-13], herbicide asulam [14] and mercaptoethanol and nitric oxide [15]. It may be predicted that one of the future implications of these reports is that many previous works on MPc-modified electrodes are likely to be revisited by many researchers on MPc-based electrodes or surfaces for sensing, catalysis, fuel cell, and many other potential applications. Thus, there is an urgent need to start building knowledge in this emerging field of CNT-MPc based electrodes. This short commentary gives insights into the recent developments, specifically touching on the main fabrication strategies (self-assembly, electrodecoration and drop-coating) currently being explored for integrating these two redox-active species onto electrode surfaces, their impacts on the heterogeneous electron transport properties, and future trends.

2. Electrode Modification Strategies CNTs are insoluble in any solvent, thus prior to use they are treated in strong acids to introduce oxygen-containing moieties (mainly –COOH group) to make them soluble [16]. The following strategies are currently employed.

Electrodes Based on Metallophthalocyanines Integrated with Carbon Nanotubes

3

2.1. Abrasive Adhesion (or Drop-Coating) of CNTs Preceding MPc Coating In this method, about 5 µL of the CNT DMF solution (1mg / 1ml DMF) is first placed onto the electrode and the solvent allowed to dry off in air or at mild oven temperature (ca. 80 o C). The same process is repeated with DMF solution of MPc solution (1 mM). The morphology of the films depends on the concentration of the casting solution, rate of solvent evaporation, nature of the solvent and roughness of the electrode surface. Important substrates for this are the pyrolytic graphites (highly oriented, basal or edge planes) because of the inherent ability of these electrodes to interact with CNTs via π-π interactions [7,17,18]. Adhesion of CNT onto glassy carbon electrode (GCE) is difficult and fraught with problems such as irreproducibility [19]. Abrasive immobilization is preferred for the MWCNT than the more expensive SWCNT. The drop-coating is possible when the CNT is pretreated in harsh acid conditions to introduce COOH, OH functionalities.

2.2. CNT Coating Preceding Electro-Decoration This method involves electrochemical deposition of the MPc onto CNT-modified electrode surface by repetitive cycling in a concentrated MPc solution (1 mM) within a specific potential window. The first cyclic voltammetric scan is usually similar to subsequent scans, indicating the formation of monomeric species only. Ozoemena et al [11] found that on certain occasions, as reported recently [11] during the electro-deposition of CoTAPc onto a basal plane pyrolytic graphite electrode (BPPGE) pre-modified with SWCNT, both cathodic and anodic waves may decrease continually and then stabilizes at a certain scan (a process known as ‘electrochemical adsorption’ or simply called ‘electrosorption’). Another form of electro-decoration is electropolymerization of the MPc complex, especially the MTAPc complex such as the NiTAPc onto CNT-modified electrodes [13]. The beauty of this technique is that the thickness and morphology of the resulting MPc polymeric film may be easily controlled by manipulating the deposition voltage, the number of cycle scans and concentration of the MPc solution.

2.3. Self-Assembly (Chemisorption) Process Self-assembly is a foremost strategy for forming highly stable, well-ordered ultra thin ‘self-assembled monolayer’ (SAM) films of redox-active species onto coinage metal surfaces [20,21]. The use of SAMs for sensing and catalytic purposes is well documented. It involves strong and irreversible chemisorption of the MPc complexes onto coinage metal based electrodes, notably gold. Gold surfaces are preferred for forming thiol-derivatised SAMs because of the well established specific and strong interaction of sulfur atoms with gold. From past experiences in the formation of MPc-based SAMs [22-31], we (my research group and collaborators) have now begun to exploit the SAM strategy in the integration of CNTs with transition MPc complexes [32,33] (Figure 1). Amino-substituted metallophthalocyanine complexes (MPc, Figure 2) can be covalently linked to acid-treated CNTs via the formation of amide bond (if R = amino group [32]) or ester bond (if R = hydroxyl group [33]).

4

Kenneth I. Ozoemena R

R N

Substituted MPc moiety

N

N

N

M

N

N

N N

R HO

R

O OH O

O

O

Self-assembled carboxylated SWCNT

O

S

O O NH

O NH S

NH

NH

Self-assembled cysteamine S

S

Figure 2. Carton showing the integration of a substituted metallophthalocyanine (MPc) complex onto a self-assembled single-walled carbon nanotube. The substitutent R could either be terminal hydroxyl (– OH) group forming an ester bond, or amino (–NH2) group forming an amide bond.

3. Impact of CNTs on Heterogeneous Electron Transport We have employed electrochemical impedance spectroscopy using common redox probe, [Fe(CN)6]4-/[Fe(CN)6]3- to interrogate the influence of CNTs on the electron transport behaviour of the MPc complexes [11-14,32,33]. Whatever the strategy used in forming the CNT-MPc based electrode, the apparent electron transfer rate constants (kapp) is highest for the MPc-CNT constructs (Bare-CNT-MPc) compared to either the bare electrode or the electrode modified with either the CNT (Bare-CNT) or MPc (Bare-MPc). Reports so far have also established that these MPc-CNT constructs (Bare-CNT-MPc) improve the electrochemical response of analytes, following the trend: Bare electrode < Bare-CNT < Bare-MPc < Bare-CNT-MPc. However, it is important to caution at this juncture that this trend may not be generalized for every analyte. This again calls for continued need for intense research in this field. The electron transfer mechanism for the CNT-MPc modified electrode may be represented as shown in Figure 3, where the immobilized MPc and CNT act as electrocatalyst and electron conducting species, respectively.

Electrodes Based on Metallophthalocyanines Integrated with Carbon Nanotubes

Analyte (Re d uc e d )

An aly te (Oxidiz e d )

So lu e t y l An a

MPc (o xidiz e d )

5

MPc (re d uc e d )

t io n

r Hy b c P -M CNT

id

Ele ctro de

eFigure 3. Proposed schematic representation of oxidative electrocatalysis at an electrode modified with CNT-MPc hybrid. In this case, the surface-confined MPc and CNT are hypothesized to act as electrocatalyst and electron conducting species, respectively.

The electron transport occurring in CNT-MPc SAM based electrodes is thought to proceed via four main steps: (i) electron-tunneling from SWCNT to gold, (ii) electron transport occurring within SWCNTs, (iii) electron-tunneling from the MPc ring and/or central metal, and (iv) heterogeneous electron transfer between the central metal of the MPc and analyte. The ‘cutting’ process of harsh acid-treatment adopted for introducing carboxylic moiety onto SWCNTs may generate local traps for charge transport, however, the high conductivity of the immobilized SWCNTs is sufficient enough to render SWCNTs as efficient conductive nanowires rather than charge traps [34,35].

4. Future Trends and Conclusion Research on the use of CNT-MPc based electrode in electroanalytical chemistry is still in its infancy. Without doubt, there is an enormous potential for using CNT-MPc-based electrodes for applications in areas such as environmental, industrial, food, pharmaceutical, clinical, and biomedical fields. Few studies have only been attempted with MPc complexes with Co, Fe and Ni as the central metals, meaning that there are many open doors for research on these and many other MPc complexes as redox mediators for the development of electrochemical sensors. Given the many advantages of electrochemical techniques (especially sensitivity to redox-active analytes, and amenability to automation,

6

Kenneth I. Ozoemena

miniaturization and remote operation) over other analytical techniques such as the spectroscopic and chromatographic methods, I envision that (i) important research works on screen-printed electrodes for one-shot analysis, ultramicroelectrodes, lab-on-chips, microband arrays, electrochemical sensors integrated with scanning electron microscopy and atomic force microscopy, etc, are likely to dominate the development of electrochemical sensors in the future; (ii) the prevalence of new types of diseases such as the drug-resistance tuberculosis, and HIV and AIDS will continue to heighten the need for rapid and sensitive onsite or point-of-care analysis; and (iii) the use of CNT-MPc composite (paste-based) electrodes, especially with the less expensive MWCNTs than the more expensive SWCNT. The main advantage of paste-based electrodes lies in their ease of regenerating the electrode surface. For example, in situations where the electrode surface becomes irrevocably contaminated or fouled new surface could easily be regenerated by polishing on clean aluminum paper.

References [1] Phthalocyanines: Properties and Applications, Lever, A.P.B.; Leznoff, C.C., Ed.; VCH Publishers: New York, 1989, 1993, 1996, Vol.1–4. [2] McKeown, N.B. Phthalocyanine Materials: Synthesis, Structure and Function, Cambridge University Press: Cambridge, 1998. [3] The Porphyrin Handbook, Kadish, K.M.; Smith K.M.; Guilard, R., Eds.; Academic Press: Boston, 1999, Vol.1-10; and 2003, Vol. 11–20. [4] Ozoemena, K.I.; Nyokong, T. In Encyclopedia of Sensors, Grimes, C. A.; Dickey, E.C. Pishko, M.V., Eds.; American Scientific Publishers: California, 2006, Vol.3, Chapter E, pp.157 – 200 [5] Nyokong, T. Coord. Chem. Rev. 2007, 251, 1707 [6] Vasudevan, P.; Phougat, N. Shukla, A.K. Appl. Organomet. Chem. 1996, 10, 591. [7] Banks, C.E.; Moore, R.R.; Davies, T.J.; Compton, R.G. Chem. Commun. 2004, 1804 [8] Jurkschat, K.; Xiaobo, J.; Crossley, A.; Compton, R.G. Analyst 2007,132, 21. [9] Banks, C.E.; Crossley, A.; Salter, C.; Wilkins, S.J.; Compton, R.G. Angew. Chem., Intl. Ed. 2006, 45, 2533. [10] Valcárcel, M.; Cárdenas, S.; Simonet, B.M. Anal. Chem. 2007, 79, 4788. [11] Ozoemena, K.I.; Pillay, J.; Nyokong, T. Electrochem. Commun. 2006, 8, 1391. [12] Pillay, J.; Ozoemena, K.I. Electrochim. Acta 2007, 52, 3630. [13] Pillay, J.; Ozoemena, K.I. Chem. Phys. Lett. 2007, 441, 72-77. [14] Siswana, M.; Ozoemena, K.I.; Nyokong, T. Electrochim. Acta 2006, 52, 114 [15] Silva, J. F.; Griveau, S.; Richard, C.; Zagal, J.H.; Bedioui, F. Electrochem. Commun. 2007, 9, 1629 [16] Liu, J.; Rinzler, A.G.; Dai, H.; Hanfer, J.H.; Bradley, R.K.; Boul, P.J.; Lu, A.; Iverson, T.; Shelimov, K.; Huffman, C.B.; Macias, F.R.; Shon, Y.S.; Lee, T.R.; Colbert, D.T.; Smalley, R.E. Science 1998, 280, 1253. [17] Moore, R.R.; Banks, C.E.; Compton, R.G. Anal. Chem. 2004, 76, 2677. [18] Wildgoose, G.G.; Leventis, H.G.; Streeter, I.; Lawrence, N.S.; Wilkins, S.J.; Jiang, L.; Jones, T.G.J.; Compton, R.G. ChemPhysChem. 2004, 5, 669. [19] Salimi, A.; Hallaj, R. Talanta 2005, 66, 967

Electrodes Based on Metallophthalocyanines Integrated with Carbon Nanotubes

7

[20] Finklea, H.O. Electrochemistry of Organised Monolayers of Thiols and Related molecules on Electrodes. In Electroanalytical Chemistry, Bard, A.J.; Rubinstein, I., Eds.; Marcel Dekker: New York, 1996; Vol. 19, pp.109-335. [21] Finklea, H.O. Self-assembled monolayers on Electrodes. In Encyclopedia of Analytical Chemistry: Applications, Theory and Instrumentations, Meyers, R.A. Ed.; John Wiley & Sons: Chichester, 2000; Vol. 11, pp.10090-101000 [22] Ozoemena, K.; Westbroek, P.; Nyokong, T. Electrochem. Commun. 2001, 3, 529. [23] Ozoemena, K.I.; Zhao, Z.X.; Nyokong, T. Electrochem. Commun. 2005, 7, 679. [24] Ozoemena, K.; Nyokong, T. Electrochim. Acta, 2002, 47, 4035. [25] Ozoemena, K.; Westbroek, P.; Nyokong, T. Electroanalysis, 2003, 15, 1762. [26] Ozoemena, K.; Nyokong, T. Talanta 2005, 67, 162. [27] Ozoemena, K.; Nyokong, T. J. Electroanal. Chem. 2005, 579, 283. [28] Ozoemena, K.I.; Nyokong, T. Electrochim. Acta 2006, 51, 5131 [29] Mashazi, P.N.; Ozoemena, K.I.; Maree, D.M.; Nyokong, T. Electrochim. Acta 2006, 51, 3489. [30] Mashazi, P.N.; Ozoemena, K.I.; Nyokong, T. Electrochim. Acta 2006, 52, 177. [31] Agboola, B.; Westbroek, P.; Ozoemena, K.I.; Nyokong, T. Electrochem. Commun. 2006, 9, 310 [32] Ozoemena, K.I.; Nyokong, T.; Nkosi, D.; Chambrier, I.; Cook, M. J. Electrochim. Acta 2007, 52, 4132 [33] Ozoemena, K.I; Nkosi, D. Electrochim. Acta 2008, 53, 2782. [34] Sheeney-Haj-Ichia, L.; Basnar, B.; Willner, I. Angew. Chem. Int. Ed. 2005, 44 78. [35] P. Diao, Z. Liu, J. Phys. Chem. B. 2005, 109, 20906.

In: Electroanalytical Chemistry: New Research Editor: G. M. Smithe

ISBN: 978-1-60456-347-4 © 2008 Nova Science Publishers, Inc.

Chapter 1

CLATHRATE HYDRATE CRYSTALLIZATION FOR CLEAN ENERGY AND ENVIRONMENTAL TECHNOLOGIES Peter Englezos1, John Ripmeester2 and Robin Susilo1,2 1.

Department of Chemical & Biological Engineering, University of British Columbia, Vancouver, BC, Canada 2. Steacie Institute for Molecular Sciences, National Research Council Canada, Ottawa, ON, Canada

Abstract Clathrate or gas hydrates are non-stoichiometric crystalline materials formed by the inclusion of certain molecules into a framework of hydrogen-bonded water molecules under suitable temperature and pressure conditions. The resulting host-guest networks consist of cavities formed by water molecules enclosing the guest molecules. Typical guests include light hydrocarbon gases, carbon dioxide, hydrogen sulfide, hydrogen and nitrogen. The basic cavity formed by water molecules through hydrogen bonding is the pentagonal dodecahedron (512). This cavity is common to the three best known hydrate structures (I, II and H). A unit cell of structure I hydrate has 46 water molecules forming two dodecahedral (512) and six tetrakaidecahedral (51262) cavities. A unit cell of structure II has 136 water molecules forming 16 (512) and eight hexakaidecahedral (51264) cavities. The Structure H hydrate unit cell has 34 water molecules, three 512 cavities, two different dodecahedral cavities which have threesquare faces, six-pentagonal faces and three-hexagonal faces (435663) and a larger iscosahedral cavity with twelve pentagonal faces and six hexagonal faces (51268). It should be noted that unlike structures I and II which may form with a single guest species, structure H requires the presence of a small guest (like methane) and a large molecule guest substance (LMGS) like neohexane at ordinary pressures. Gas hydrates were first reported at the beginning of the 19th century, and until the 1930s they remained a scientific curiosity. At that time it was realized that hydrates were more likely to be the causative agent in blocking pipelines than ice. Today, gas hydrate control continues to be a problem in the oil and gas industry. In the 1960’s it was realized that natural gas hydrates are present in the geo-sphere with worldwide reserves estimated at 10,000 to 40,000 trillion cubic meters (TCM). Considerable efforts are underway to refine global estimates and to develop technology and exploit this resource. On the other hand these hydrates may

10

Peter Englezos, John Ripmeester and Robin Susilo decompose as a result of global warming or seafloor instability and release the methane gas. There is speculation that a runaway greenhouse effect could result, with some evidence for changes of such magnitude in the global paleoclimate (15,000 and 55 million years ago). Application of clathrate hydrate crystallization offers the possibility of the development of innovative technologies for natural gas storage and transportation, hydrogen storage and gas separation with applications for carbon dioxide capture from flue gas (CO2/N2/O2) or fuel gas (CO2/H2) mixtures. Clathrate hydrates have become an important research area spanning a variety of disciplines. It is of continuing great interest to chemical engineers who have played a key role in past developments. This report discusses areas where chemical engineers can advance the knowledge frontier.

Introduction It is well known that when sufficient amounts of water and a hydrate-forming substance are brought into contact under appropriate temperature and pressure conditions a crystalline solid known as a gas or clathrate hydrate forms (van der Waals and Platteeuw, 1959; Davidson, 1973; Englezos, 1993; Sloan, 1998, Ripmeester, 2000; Koh, 2002; Sloan, 2003a; 2003b; 2004a; 2004b; 2005; Englezos and Lee, 2005; Chatti et al. 2005; Bishnoi and Clark, 2006). Gas hydrates were reported as early as 1810 (Davy, 1811). Following Sir Humphrey Davy’s report of aqueous solutions of chlorine that remained solid at temperature above 0°C (Davy, 1811) and Faraday’s confirmation in 1823, gas hydrates were a steady object of scientific curiosity for more than 100 years. Research efforts became more focused from the mid 1930’s on due to the suggestion that the unwanted solid material in gas transmission pipelines was in fact gas hydrate rather than ice, frequently forming plugs at temperatures above the icepoint (Hammerschmidt, 1934). Since then, extensive experimental and computational studies have been carried out in order to identify the equilibrium formation conditions for various hydrate forming systems. This research was driven by the need to establish methods to prevent the occurrence of hydrate of hydrate plugs in oil and gas pipelines. Hydrate research on flow assurance is still carried out especially with oil and gas exploration into deeper water and remote offshore areas. Despite the initial negative impression on gas hydrates, recently it has been realized that gas hydrates possess important roles towards energy and environmental issues as well, mainly due to its potential for gas holding capacity and separation purposes. Methane, and other natural gas components, trapped in ice-like lattices known as ‘gas hydrates’ or ‘clathrate hydrates’ have been found to exist naturally in the earth, especially offshore on the continental margins and under the permafrost in the Arctic (Makogon et al., 1972; 1987; Suess et al., 1999; Kvenvolden 1988; 1999; 2000; Reeburgh, 2003; Buffet and Archer, 2004; Klauda and Sandler, 2005). The exact quantity of natural gas in hydrate form is not known accurately, but is significant and considered to be a huge potential unconventional energy source for the future, with the latest estimates some 5 – 20 % of the global carbon budget. Unfortunately technologies for gas production do not exist as yet due to complications involved. Besides the economic aspect that has to be taken into consideration, the impacts on the environment, ecology, and geological stability are also important. The decomposition of natural hydrate may destabilize and alter the earth’s geological features (Glasby, 2003; Sultan et al., 2004a; 2004b) causing geohazards such as landslides, earthquakes, and even tsunamis if the mass movements are under water. On the other hand, current global warming may initiate the decomposition of natural hydrates (Hatzikiriakos and

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 11 Englezos, 1993; Kvenvolden, 2002). The first production test of methane hydrate was conducted in the Mackenzie Delta in the Canadian Arctic. Two test wells were drilled where samples cores were collected to understand its characteristic and evaluate the possibility for successful gas production (Bybee, 2004). Thermal and pressure stimulations were tested to measure both input conditions and reservoir responses that enabled the calibration and refinement of reservoir simulation models (Collet, 2005). Moreover, hydrates were also suggested to exist in extraterrestrial space especially on the outer planets and satellites, eg. Mars, Saturn, Uranus and Neptune (Delsemme and Swings, 1952; Delsemme and Wenger, 1970; Lunine and Stevenson, 1985; Koh, 2002; Osegovic and Max, 2005; Tobie et al., 2006; Machida et al., 2006; Hand et al., 2006). Titan, the largest moon of the Saturn is believed to have 100 km thickness of high pressure methane hydrate within its ice mantle (Loveday et al., 2001). Gas hydrate also offers opportunities to develop innovative technologies for gas storage and separation (Gudmundsson et al., 1994; 2000; Mori, Y. H., 2003; Englezos and Lee, 2005) as well as cool energy storage (Inaba, 2000; Tanasawa and Takao, 2002). The applications include storing and transporting natural gas (Khokkar et al., 1998; Seo and Lee, 2003; Mori, 2003; Thomas and Dawe, 2003; Lee et al., 2005b; Tsuji et al., 2005a; 2005b; 2004; Javanmardi et al., 2005; Abdalla and Abdullatef, 2005), hydrogen storage for the hydrogen economy of the future (Mao et al., 2002; Patchkovskii and Tse, 2003; Lee et al., 2005a; Schuth, 2005 Strobel et al., 2006; Hester et al., 2006; Hu and Ruckenstein, 2006), sequestration of carbon dioxide with in situ methane hydrate decomposition (Lee et al. 2003; Park et al., 2006a; House et al., 2006), separation of carbon dioxide from the flue gas (Seo et al., 2005a; Yoon et al., 2006; Park et al., 2006b), seawater desalination (Parker, 1942; Barduhn et al., 1962; Javanmardi and Moshfeghian, 2003; Rautenbach and Pennings, 1973) and refrigeration (Tomlison, 1982; Ternes, 1984; Mori and Mori, 1989a,b; Bi et al., 2006). For hydrate technology to be industrially applicable, several challenges need to be addressed such as: high gas storage capacity, efficient gas separation, fast phase transformations, and mild pressure-temperature conditions during processing, storage, and transportation. Research on gas hydrates involves engineers and scientists. Chemical engineers in particular can make significant contributions towards the commercialization of gas hydratebased technologies. Therefore, the objective of this chapter is to review developments and highlight areas of opportunity for the involvement of chemical engineers. The structure of the chapter is as follows. First clathrate hydrates are described with emphasis on their structures and the fundamentals of thermodynamic and kinetic properties. Subsequently, the clean energy technologies under development are discussed. Here, chemical engineers can play a crucial role in the successful scale-up of the various existing concepts. These include natural gas storage and transport, hydrogen storage, and carbon dioxide capture and sequestration. Flow assurance in the hydrocarbon production and transportation industry is the only industrial-scale area where continuous research activities both at fundamental and practical levels are ongoing. However the potentials of hydrates for other practical applications have not been exploited mostly because the technology is relatively new and not established. Finally the recovery of natural gas from natural deposits is briefly discussed and the interested reader is directed to a number of recent references.

12

Peter Englezos, John Ripmeester and Robin Susilo

Clathrate Hydrates Gas hydrates are true inclusion compounds and hence also are known as clathrates (Davidson, 1973). They are non-stoichiometric crystalline solids with the physical appearance of ice, and consist of water molecules that serve as the host material and guest molecules trapped in the host. Natural gas components such as: methane, ethane, propane, carbon dioxide, and hydrogen sulfide are typical guest molecules, although any hydrophobic molecule that fits the host cavities can be the guest. The water molecules are connected through hydrogen bonds forming cages/cavities that completely encage individual guest molecules. The crystal lattice consisting of empty cages is not stable thermodynamically, and there is a minimum guest content required to give a stable lattice. Hence, guest molecules must have the correct sizes and geometry in order to fit into the different cages. Normally there is a single molecule per cage except in the case of high pressure hydrates of atoms or small molecules (N2, H2, CH4, rare gases). There are no specific or directional interactions between the host and guest molecules with the weak van der Waals forces providing the key interaction. Thus the guest molecule is free to rotate and translate inside the cage. The crystal lattices have limited flexibility to accommodate guests of different size so that the crystal lattice parameters have well defined limits. Further discussion of structures, kinetics and thermodynamic properties are discussed in the next section.

Structures Hydrate cages are arranged and packed in different configurations, with three distinct crystal structures commonly encountered today: cubic structure I (sI) (Müller and vStackelberg, 1952; McMullan and Jeffrey, 1965), cubic structure II (sII) (vStackelberg and Müller, 1951; Mak and McMullan, 1965), and hexagonal structure H (sH) (Ripmeester et al., 1987). The cubic structures have two types of cavities (small/S and large/L) but the hexagonal Table 1. Hydrate structures and cage properties Structure Crystal system Space group Lattice parameters Number of cage Cage identification Ideal unit cell formula Cage radius

Coordination number

I Cubic Pm3n

II Cubic Fd3m

a = 12A

a = 17.3A

2

2

Small/S (512) Large/L (51262)

Small/S (512) Large/L (51264)

2S.6L.46H2O

16S.8L.136H2O

rS = 3.95A rL = 4.33A

rS = 3.91A rL = 4.73A

S = 20 L = 24

S = 20 L = 28

H Hexagonal P6/mmm a = 12.2 A c = 10A 3 Small/S (512) Medium/M (435663) Large/L (51268) 3S.2M.1L.34H2O rS = 3.91A rM = 4.06A rL = 5.71A S = 20 M = 20 L = 36

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 13 one has three cavities (small/S, medium/M and large/L). Experience has shown that the large cavities in all common hydrate structures usually are filled completely by a guest molecule. For the cubic structures, the small cavity may be empty and this depends on molecular size, so that a single guest species is enough to stabilize sI and sII hydrates. However a small “help guest” is required to maintain sH hydrate so that two guest species are required. Generally there is only one guest molecule in a cage. Few exceptions were reported at high pressures where smaller molecules like nitrogen (Kuhs et al., 1997), hydrogen (Mao et al., 2002), rare gases (Loveday et al., 2003a; Alavi et al., 2005; 2006b), and methane (Loveday et al., 2001; 2003a) may fill and stabilize the large cage of sII and sH with multiple guest occupancies of a cage. Small (S)

Medium (M)

Large (L)

-

512

Structure I (sI)

51262

-

512

Structure/Formula

2S.6L.46H2O

Structure II (sII)

51264

16S.8L.136H2O

Structure H (sH)

512

435663

51268

3S.2M.1L.34H2O

Figure 1. Hydrate cages.

The small cage is a polyhedron made of twelve-pentagonal faces (pentagonal dodecahedron/512). It is the basic cage that is commonly present in all hydrate structures. However, since lattice stability is derived from the filling of the largest cage in the system, their presence is key, however, they cannot be packed together to fill three-dimensional space. For the cubic structures the basic cages are a polyhedron made of twelve-pentagonal faces with two-hexagonal faces (51262) for sI hydrate and four-hexagonal faces (51264) for sII hydrate. The pentagonal dodecahedra then provide the space-filling building blocks. A unit cell of sI hydrates consist of 2-small cages, 6-large cages and 46 water molecules whereas sII hydrates consist of 16-small cages, 8-large cages and 136 water molecules. The hexagonal sH hydrate has 3-small cages, 2-medium cages, 1-large cage and 34 water molecules. The large cavity has twelve-pentagonal faces with eight-hexagonal faces (51268). Six-hexagonal faces

14

Peter Englezos, John Ripmeester and Robin Susilo

are located on the equatorial plane and linked to another large cage via the medium cages (435663). The other two-hexagonal faces are located at the polar plane and connected to another large cage. The small cage (512) fills the space remaining in the unit cell. The cages are shown in Figure 1 and their properties are summarized in Table 1. sI and sH hydrate both can be derived from the stacking of layers of pentagonal dodecahedra. The space between layers stacked in different ways is taken up by the other cages in the crystal lattice. The stable crystal structure formed depends on the guest molecule(s) present. Guest molecules that fill the void space of the cages efficiently are generally preferred. As expected, see Table 1, the size of all of the dodecahedral cages in the hydrate structures is quite similar although the cage symmetry is different in the three structures. However the large cage size increases on going from sI, to sII and sH. Hence the stable hydrate structure formed is governed by the size of the largest molecule(s) in a particular hydrate. Guest molecules with a van der Waals diameter less than ~9.7A have been reported as hydrate formers. Structure I hydrate is generally formed by smaller molecules (4.2A
Thermodynamics and Kinetics of Gas Hydrates The thermodynamics of gas hydrates is a mature subject (Sloan, 2004b). However, new data continue to appear, mainly for mixtures required in new applications, or for more severe conditions of temperature and pressure. A list with experimental data up until the late 1990s is available (Sloan, 1998). From a computational standpoint the main interest is in calculating the incipient hydrate formation pressure for a given hydrate-forming mixture at a particular

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 15 temperature. Modeling of hydrate phase equilibria has been based on the statistical thermodynamics model of van der Waals and Platteeuw (van der Waals and Platteeuw, 1959; Parrish and Prausnitz, 1972) and various computer methods have been developed. Hammerschmidt developed the first method used in the industry for predicting the inhibiting effect of methanol (Hammerschmidt, 1934). The method is empirical and the reliability of the calculations is variable (Ng and Robinson, 1985). Anderson and Prausnitz presented a thermodynamics-based method for calculating the inhibiting effects of methanol (Anderson and Prausnitz, 1986). They used the van der Waals-Platteeuw model for the solid hydrate phase, Redlich-Kwong equation of state for the vapour phase and the UNIQUAC model for the liquid phase. Henry’s constants were used for calculating the fugacities of components in their supercritical state in the liquid phase. Furthermore, empirical correlations were used for calculating the molar volumes, partial molar volumes at infinite dilution and the fugacity of the hypothetical liquid phase below the ice-point temperature. Robinson and Ng presented a computer program for the calculation of the depression of hydrate formation temperatures due to methanol (Robinson and Ng, 1986). A computational method based on the Trebble-Bishnoi equation of state was presented by Englezos et al. for calculating the depression effects of methanol and the amounts of methanol required (Englezos et al., 1991; Trebble and Bishnoi, 1988). All equilibrium hydrate prediction methods use either an equation of state for all fluid phases or an equation of state for the vapor phase and activity coefficient models for the liquid (Englezos, 1993). The use of a molecular-based equation of state (SAFT) to compute the inhibiting effect of methanol, glycols and glycerol was also presented (Li and Englezos, 2006; Li et al., 2007). More recently Sloan’s group at the Colorado School of Mines (CSM) provided software to predict the hydrate phase equilibria for mixtures and in the presence of thermodynamics inhibitors (Ballard and Sloan, 2002; Jager et al., 2003; 2005; Ballard and Sloan, 2004a; 2004b). Flash calculations in which one is also interested in determining the amount of hydrate formed have also appeared (Bishnoi et al., 1989). It is difficult to determine the quantity of hydrate or hydrate phase fraction experimentally, and this is an area in need of future attention. Kinetics is concerned with the rate at which the phase transformation occurs and the identification of the factors affecting it. Gas hydrate formation, being a crystallization process is characterized by nucleation followed by crystal growth and agglomeration (Bishnoi and Natarajan, 1996; Englezos, 1996). The rate of nucleation e.g. number of hydrate crystal nuclei formed per unit time per unit volume is extremely difficult to measure and to date there is no data available. Most gas hydrate kinetic studies have focused on the growth phase and involve measuring the rate of uptake of the hydrate forming substance (Englezos et al., 1987a; Englezos et al., 1987b). This is a macroscopic measurement which may also be coupled with measurement of the particle size distribution (Clarke and Bishnoi, 2004; Clarke and Bishnoi, 2005). Bishnoi’s laboratory at the University of Calgary pioneered these kinetic studies. They also develop a mechanistic hydrate kinetics model (Englezos et al., 1987a; 1987b) that was later improved (Dholabai et al., 1993). Another macroscopic-type measurement coming online is based on differential scanning calorimetry (Le Parlouer et al., 2004). Figure 2 shows the typical methane uptake profile (Lee et al., 2005b). An initial increase in methane consumption is observed due to the gas dissolution in water. Once the liquid is saturated with the gas, no more gas consumption is seen. The hydrate nuclei start to form at this point until it reaches a critical size as indicated by the arrow and an increase in the temperature profile due to heat released during the hydrate formation (exothermic). The curve

16

Peter Englezos, John Ripmeester and Robin Susilo

representing methane moles consumed versus time is approximately linear after the nucleation but the slope of the curve gradually decreases at longer experimental times after a certain level of conversion to hydrate. The decrease was suggested as being due to limited contact between the water and gas. 277.5 277.0

Moles consumed Temperature

0.15

276.5 0.10 276.0 0.05

275.5

Temperature (K)

CH4 moles consumed

0.20

275.0

0.00 ntb (0.0071mol, 482.7 min)

0

100

200

300

400

500

600

700

274.5 800

Time (min)

Figure 2. Rate of methane consumption for sI hydrate formation of CH4-H2O system at 275.50 K and 3.96 MPa. (Reprinted from Energy and Fuels (Lee et al., 2005b), Copy right (2005) with permission from American Chemical Society).

The macroscopic approach does not provide any information on the solid phase properties (Sloan, 2003b). Recently a more sophisticated in-situ measurements using Raman spectroscopy (Park et al., 2006a), NMR spectroscopy and imaging (Moudrakovski et al., 2001; 2004; Gao et al., 2005; Seo et al., 2005a; Park et al., 2006a; Chen et al., 2006; Susilo et al., 2006), X-Ray diffraction (Uchida et al., 2003; Takeya et al., 2002), neutron diffraction (Kuhs et al., 2006; Hester et al., 2006) and digital scanning calorimetry (Dalmazzone et al., 2006) have been employed to follow hydrate conversion. Molecular dynamics simulation is also being exploited to understand the kinetics of hydrate formation (Nada, 2006; Vatamanu and Kusalik, 2006). In order to gain more insight into hydrate kinetics, hydrate crystal morphology has also been studied in order to elucidate the mechanism of hydrate nucleation, migration, and growth at the surface of water droplets and at the interface between water and gas or nonaqueous liquid (Lee et al., 2005c; 2006b; Ohmura et al., 2004; 2005b; 2005d; Servio and Englezos, 2003a; 2003b; Ito et al., 2003; Kobayashi et al., 2001; Uchida et al., 1999b). A hydrate film growth model can be derived from the macroscopic observation (Mochizuki and Mori, 2006; Mori, 2001; Mochizuki, 2003). In general, hydrate preferentially nucleates at an interface where the guest concentration is at a maximum. However hydrate nucleation from the bulk liquid phase was also observed and this occurs only when the liquid phase is supersaturated or where there is local sub-cooling. The nucleation and crystal growth rates generally increase with driving force, given by pressure increase or temperature decrease from the equilibrium values. After nucleation, hydrates tend to grow into the adjacent volumes. If the nucleation occurs at an interface, hydrates will grow along the interface, forming a film with a certain thickness. If the nucleation occurs in the bulk phase, the crystal

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 17 will grow and then float towards the interface if the density of the hydrate is lower (e.g. methane hydrate) than that of water. Subsequent hydrate growth is limited by mass transfer of water or guest molecules across the hydrate film. It has been suggested that the gas may diffuse and grow towards the water phase and water may also migrate out through the hydrate due to capillary action. Hairy or needle-like crystal morphologies were observed as going from the interface towards the gas phase. Columnar or dendritic crystals were observed as growing from the interface towards the liquid water phase. Crystal morphology has been suggested to depend on the degree of sub-cooling (Knight and Rider, 2002; Lee et al., 2006b). The presence of inhibitors changes the crystal morphology (Lee and Englezos, 2006; Zeng et al., 2003, Larsen et al., 1998). Kinetics through gas uptake measurements and macroscopic techniques, in general, is in reality the “average kinetics” over the whole sample. The observation of gradual conversion in bulk samples only arises as a result of averaging over many local environments (Moudrakovski et al., 2004), as locally the conversion to hydrate is quite an inhomogeneous process. This significant finding, based on microscopic techniques such as NMR microimaging, poses the question how microscopic measurements of kinetics are related to kinetics obtained from macroscopic techniques. Susilo et al. (2006) employed Nuclear Magnetic Resonance (NMR) spectroscopy and imaging (MRI) to monitor the kinetics of structure I and H methane hydrate growth and found that the results agree with the results from gas uptake measurements obtained by Lee et al. (2005a). At least, both methods give the same answers when the local techniques are averaged over the sample inhomogeneities. Kinetic studies involve measurement of induction times for crystallization and the determination of the rate of hydrate crystal growth, and the latter is usually defined operationally. Thus, one may use the gas uptake rate to describe kinetics or the rate at which an interface involving a hydrate phase advances. The question that arises is how an intrinsic rate can be distinguished from the relevant transport processes. The difficulty in this regard as well as the variety of methods to study the dynamics of hydrates gives rise to a variety of kinetic studies on gas hydrates. In addition, in spite of kinetic models that have appeared and aided the mechanistic understanding of hydrate growth it is still difficult to come up with a predictive model that will allow prediction of the onset of hydrate formation. Thus, there is a need to study particular hydrate forming systems and hydrate vessel configurations. This is an area for future growth and it is ripe for new approaches. In this regards, a new computer simulation approach based on cellular automata and Monte Carlo methods appeared recently (Kvamme et al., 2004; Svandal et al., 2006; Buanes et al., 2006). It is also important to mention the nucleation time during hydrate formation. Nucleation by far is known as a stochastic phenomenon hence it is unpredictable meaning the induction data cannot be reproducible from system to system even though each system has the same thermal history (Lingelem et al., 1994). The time required to nucleate hydrate seed may further delay the productivity of hydrate formation. It is suggested that heterogeneous nucleation occurs in most hydrate systems due to lower driving force (higher temperature or lower pressure) required than homogeneous nucleation (Zeng et al., 2006a,2006b). Hydrate nucleation time is also affected by the history of water sample but the growth is not. Liquid water and also ice that has experienced hydrate formation tends to have a shorter nucleation time (Moudrakovski et al., 2001; Buchanan et al., 2005; Lee et al., 2005c). This is also referred as memory effect. It is still unclear and debatable the reasoning behind the “memory effect”. It was suggested that there is residual hydrate structures and/or dissolved gas in

18

Peter Englezos, John Ripmeester and Robin Susilo

memory solution. Neutron diffraction study did not detect any residual hydrate structures in the memory solution as previously suggested (Buchanan et al., 2005). The structure of water changes only when hydrate crystallites are present in the solution (Thompson et al., 2006). Perhaps the dissolved gas in liquid water after hydrate decomposition alters the water molecule orientation due to hydrophobic interaction that leads to faster hydrate reformation. The statistical study on induction time indeed show that the “memory effect” tends to disappear when the samples are exposed to higher temperature or longer decomposition time (Ohmura et al., 2003b). Nevertheless, the utilization of memory solution is beneficial to speed up the nucleation time. This is an important aspect for industrial practice. The water from dissociated hydrate can be sent back to the hydrate production plant where it is reused to speed up hydrate nucleation. Anti-freeze protein (AFP) is the only inhibitor that that has shown to have the ability to successfully eliminate the memory effect (Zeng et al., 2006a; 2006b). a combination of a commercial inhibitor and polyethylene oxide was found to substantially suppress the memory effect (Lee and Englezos, 2006).

Clean Energy Technologies Based on Hydrates There are more than 100 molecules identified as hydrate formers (Ripmeester and Davidson, 1977; Ripmeester and Ratcliffe, 1990). Most of them are mainly of scientific interest but only a fraction has received attention because of possible practical significance. Table 2 gives a list of hydrate formers that are either relevant to humanity because of issues such as climate change, energy, and nature in general, or that may serve some future applications. The corresponding hydrate stability condition and heat of hydrate dissociation at 0°C in equilibrium with liquid water (L) and gas phase (V) are also given. Generally, the Table 2. Structural, stability properties and heat of hydrate decomposition at 0°C of most technologically relevant hydrate formers Hydrate former Hydrogen Nitrogen Oxygen Methane Hydrogen sulfide Carbon dioxide Ethane Tetrahydrofuran Propane i-butane Methyl tert-butyl ether Neohexane 1,1-dimethylcyclohexane Methylcyclohexane *

Molecule size [A] 2.72 4.10 4.20 4.36 4.58 5.12 5.50 5.90 6.28 6.50 7.60 7.99 8.40 8.59

Structure

P equilibria [MPa]

II [152] II [106] II [275] I I I I II II II H H H H

>100 [151] 16.10 [311] 12.02 [310] 2.65 [95] 0.10 [235] 1.22 [314] 0.50 [45] 0.1; 277K [70] 0.16 [174] 0.11 [231] 1.52 [84;*] 1.13 [166,173,200,*] 0.88 [45;*] 1.29 [173,177,200;*]

Phase equilibria - methane occupies the 512 and 435663 cages Regressed using Clausius-Clayperon equation from the H-L-V phase line

**

ΔH H-L-V [kJ/mol] NA 49.54 [41*] 49.16 [41*] 54.19 [71] 63.10 [321] 65.22 [98] 71.80 [71] 99.50 [295] 129.2 [71] 133.2 [71] 76.6 [**] 76.8 [**] 81.8 [**] 79.5 [**]

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 19 molecule size is correlated with hydrate structures with exception from those of the smaller ones. Hydrate decomposes below the equilibrium pressure at a given temperature. Higher pressure is required at higher temperature. The decomposition process is an endothermic reaction, which absorb heat that can be measured by calorimetric method or regressed from the corresponding phase equilibrium data (Yoon et al., 2003; Sloan and Fleyfel; 1992). Hydrogen is the simplest molecule and it has been suggested to be the starting material for the creation of planets (Stevenson, 1999). Hydrogen is also a suitable fuel for fuel cells. Nitrogen and oxygen are the essential components of air which have been found trapped within the polar ice sheet in the form of bubbles as well as hydrates. The explanation of how air hydrate forms in the deep glaciers such as in Antarctica and the hydrate composition may help explain some of the paleoclimate records. (Uchida et al., 1994). Most of the other hydrate formers listed in Table 2 are typical components of natural gas that are commonly found in gas reservoirs or natural hydrate samples. Tetrahydrofuran is a water soluble guest that is known to have a strong hydrate stabilizing effect. It is particularly attractive for hydrogen storage where a tremendous pressure reduction can be achieved by forming mixed hydrogen and THF hydrate (Florusse et al., 2004; Lee et al., 2005a; Alavi et al., 2006a). The large molecules that occupy the large cage of sH hydrate are also called LMGS (Ohmura et al., 2002; Tsuji et al., 2004). LMGS may be found as condensates in gas transmission pipelines and in naturally occurring sH hydrate (Lu et al., 2007). The LMGS in condensate are undesirable from the point of view of flow assurance because methane and LMGS may form hydrate and plug the pipelines. On the other hand, the LMGS may be beneficial for gas storage and transport applications as methane can be stored under milder conditions. The maximum pressure reduction for methane storage in sH hydrate was reported for 1,1dimethylcyclohexane (Hara et al., 2005). The following section discusses the opportunities for gas hydrate applications to clean energy and the environment. The motivation, current knowledge, opportunities, challenges, research efforts in our laboratories and directions towards commercializing the technology are discussed.

Natural Gas Transport and Storage Transporting and storing natural gas from the gas field to the energy users is one of the promising industrial applications for hydrate technology. This is supported by the fact that there are more than 50% natural gas field worldwide considered as stranded and abandoned gas fields (Ivanhoe, 1993). It is profitable to recover the gas from the gas fields where the gas quantities are not sufficient and/or located in remote areas. This is simply because it is too costly to transport the gas via pipelines or in liquefied form (LNG). Hence there is a need for cost-effective technology in terms of capital and operational costs. Compressed natural gas (CNG) in ships and gas to solid (GTS) technology via gas hydrates are being considered as promising candidates to compete with the established pipeline or LNG technology. However there are many factors that have to be taken into consideration before selecting which technology to be implemented. Natural gas composition varies from field to field. The main component of purified natural gas generally is methane (~90 %+) along with other light hydrocarbons. Hence, hydrate formation conditions and the hydrates obtained may vary according to gas composition. Obviously, each component interacts differently with water and hence there will

20

Peter Englezos, John Ripmeester and Robin Susilo

be fractionation of the components between the hydrate and the hydrocarbon phase. The formation of the thermodynamically favored phases should result in the lowest equilibrium pressure. If the natural gas contains only methane, the stable hydrate structure is sI. However the most stable hydrate structure is more likely to be sII due to the presence of larger components that are sII hydrate former such as propane, i-butane (Østergaard et al., 2001), and various combinations of ethane and methane also give sII hydrate (Hester and Sloan, 2005; Takeya et al., 2003; Subramanian et al., 2000a; 2000b). The formation of sH hydrate is possible if the gas contains condensates. A mixture of hydrate structures may also be present. Methane gas is the preferred model system representing natural gas in most hydrate study because it is the cleanest fuel among hydrocarbons. The simplified system provides a sound basis for further development and avoids the complexities that may be encountered when working with gas mixtures (Schicks et al., 2006) although the ideal methane system may not approximate real life. Produced natural gas also contains unwanted compounds such as hydrogen sulfide, carbon dioxide, and heavier hydrocarbon. Thus one has to treat the gas first to meet the safety regulations before converting it into hydrate. The fractionation of produced natural gas could also be performed with hydrate technology. Hydrate fractionation allows the gas phase to be methane-rich due to the preference of the heavier hydrocarbon components as well as CO2 and H2S for the hydrate phase. This enables the trapping of methane in hydrate after fractionation (Uchida et al., 2004). Obviously hydrate processes have to be coupled to other processes such as membrane technology to further purify the gas phase.

Methane Hydrate Properties Methane is one of the most studied guest molecules associated with gas hydrate. This is simply because methane is found in huge quantities in natural gas, coalbed methane, and natural hydrates, and it is the cleanest fuel among hydrocarbons. Under moderate pressure conditions, methane and water form a stable sI hydrate with the large and most small cages occupied by methane if the hydrate is prepared near equilibrium conditions. Hydrate phase equilibrium data, dating back to 1946, has been reported by many researchers (Deaton and Frost, 1946). Hydrate phase equilibrium data at temperature below the icepoint is also available (Makogon and Sloan, 1994). The phase diagram can be predicted satisfactorily by using the van der Waals-Platteew model (Van der Waals and Platteeuw, 1959). The model was improved by Sloan’s group at the Colorado School of mines to better predict the hydrate phase equilibria for mixtures and in the presence of thermodynamics inhibitors (Ballard and Sloan, 2002; Jager et al., 2003; 2005; Ballard and Sloan, 2004a; 2004b). The hydration number of methane sI hydrate has been measured by various techniques such as calorimetry (Handa, 1986a,b), Raman (Uchida et al., 1999a; 2003; Sum et al., 1997) and NMR spectroscopy (Collins et al, 1990; Ripmeester et al., 1988) with the value reported quite consistent at around ~6.0 ±0.2 irrespective whether it is a synthesized or natural hydrate formed at different conditions. Methane cage occupancies are also given in the literature obtained from Raman (Sum et al., 2000), NMR (Ripmeester et al., 1988), and single crystal X-Ray diffraction (Udachin et al., 2002; Kirchner et al., 2004). They are generally in good agreement with the predictions of the statistical thermodynamic model, which suggests that all large cages are fully occupied with methane and approximately 90% of the small cages are occupied when hydrates are prepared near Hydrate-Gas-Liquid water (HG-L) equilibrium

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 21 conditions. The measured and calculated hydration numbers decrease slightly with increasing pressure with higher methane cage occupancy (Sum et al., 1997; Klapproth et al., 2003; Circone et al., 2005). The reported hydration number at very high pressure (620 MPa) is 5.67 (Ogienko et al., 2006), which correspond to full methane occupancy of both cages. The lattice constant of sI methane hydrate for synthetic and natural hydrate is practically the same at 11.85A (113K) which corresponds to hydrate density of 0.95 g/cm3 (Takeya et al., 2006). The lattice constant increases with temperature due to thermal expansion (Shpakov et al., 1998) but decreases with pressure due to compressibility (Klapproth et al., 2003). The reported lattice constant at 273K is 11.99A and refers to a hydrate density of 0.92 g/cm3 (Takeya et al., 2006). Consequently the methane content in hydrates can be calculated based on the hydration number and density, which is approximately ~160-180 v/v depending on the cage occupancy and temperature.

Gas Storage Potential in Hydrate Gas storage capacity in hydrate is one of the most important factors and depends strongly on the hydrate structure and its ability to accommodate the guest molecules. Each hydrate structure has its characteristic cage sizes and cage packing configuration. Methane is a small molecule with a van der Waals diameter of 4.36A and can fit into all of the hydrate cages. However there is a preference regarding the cages over which methane is distributed as driven by thermodynamic and kinetic factors. Methane and water generally form a stable sI hydrate under moderate pressures. The formation of stable pure methane sII or sH hydrates occurs only at extremely high pressure (25 MPa or above) (Chou et al., 2000). In the presence of another guest molecule such as double hydrates are formed that are stable eg propane for sII (Kini et al., 2004) and methylcyclohexane for sH (Mehta and Sloan, 1993). A sII methane hydrate was also reported as a kinetic product (meta-stable) (Schicks and Ripmeester, 2002). Methane itself does not stabilize the sII or sH lattices at moderate pressures because of inefficient interactions between the host and guest, and hence the presence of a larger molecule is required. The methane content of all hydrate structures is summarized in Table 3. The number for LNG and the corresponding hydrate stability conditions are also given. It is again important to note that not all hydrate cages are fully occupied by methane under three phase equilibrium conditions. The numbers for full methane occupancy and the realizable methane content are also given. For sII and sH hydrates most or all of the large cages are filled with a larger guest molecule. Table 3. Methane gas content in hydrates and stability condition Hydrate structure Structure I Structure II Structure H LNG

Methane storage in hydrate [m3/m3] Maximum Realistic 180 170 172 < 110 168-280 140 600

Hydrate stability condition Temperature -20°C -20°C -20°C -160°C

Pressure [MPa] 1.2 < 1.0 0.5 ~ 0.7 0.1

22

Peter Englezos, John Ripmeester and Robin Susilo

It is clear from Table 3 that storing methane in LNG has the maximum energy density. However this requires costly cryogenic conditions during storage and transport. On the other hand, storing methane in hydrates has approximately 3-4 times lower energy density than LNG but the conditions are much milder. Methane hydrates, when palletized, can be stored in a freezer at -20°C and atmospheric condition for about 2 weeks without significant hydrate decomposition. Some gas hydrates show “self-preservation” properties, allowing storage well below the equilibrium pressure at freezer temperatures, (Stern, 2001a; 2001b; Takeya et al., 2001; Kuhs et al., 2004; Shimada et al., 2005), allowing storage and transport methane from the gas field to the energy users. It has been suggested that the endothermic decomposition of hydrate can be delayed when palletized and stored adiabatically as heat input is minimized. The larger the hydrate pellet, the longer hydrate decomposition can be delayed (Takeya et al., 2005). The maximum possible and realizable methane content is obtained with sI hydrate but this hydrate structure is the least stable of the methane-containing hydrates. Hence it is essential to look at the overall performance in selecting the appropriate system that offers the optimum in terms of gas content, stability condition, and the kinetics during formation and decomposition. The addition of a second molecule to form sII and sH hydrate reduces the equilibrium pressure. However the methane content is also reduced. The pressure reduction depends on the type of guest molecule chosen and the concentration. For a double hydrate of sII a much lower pressure is required however, the methane content decreases significantly, too, which is unfavorable. In case of sH hydrate, the methane content reduction is not as large with an appreciably reduction in pressure by a factor of two. Accordingly, there are many recent efforts attempting to store methane in sH hydrate. Tuning gas hydrate properties to find the compromise between storage capacity and pressure will be discussed at a later section. There is additional gas storage potential in sH hydrate because the large cage might be occupied by more than one methane molecules. Unfortunately this occurs only under extremely high pressures and hence is not or practical interest. Currently the study of high pressure methane hydrates of sH hydrate are relevant only to study the behavior of water and gas molecules in the outer solar system (icy satellites/planetary studies). However further studies are required to get the complete understanding on the feasibility of methane storage in sH hydrate with multiple guest occupancy.

Hydrate Synthesis The hydrate phase equilibria and gas storage capacity are constrained by the thermodynamic conditions (temperature, pressure), crystallographic properties (structure, cage size and shape) and guest molecule properties (size, shape). Since these properties are often relatively well defined and fixed by specific external requirements, the optimization of kinetics becomes an important topic from a practical viewpoint. It is indispensable to have good recipes for hydrate synthesis. The rate of hydrate formation and the yield or conversion level of hydrate are important from the economic point of view. The role of mass and heat transfer is significant in order to ensure the progression of the reaction towards optimum conversion. The contact between water and gas is also an important aspect for controlling the hydrate conversion.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 23 Hydrate formed from liquid water generally requires efficient mixing due to the poor methane solubility in water. A rapid hydrate formation rate is essential for hydrate technology to become useful to industry. Mixing can be achieved by using a mechanical mixer, a magnetic bar coupled to a rotating magnet (Vysniauskas and Bishnoi, 1983; Lee et al., 2005b), a static mixer (Tajima et al., 2004; 2005a; 2005b; 2006), a spray of water droplets into a gas atmosphere (Tsuji et al., 2004; 2005), or bubbling the gas into the liquid water phase (Ma et al., 2002; Takahashi et al., 2003). The performance of several contacting modes is currently being investigated. Mori’s laboratory at Keio University (Japan) has proposed the formation of hydrate by spraying water droplets into a gas atmosphere (Ohmura et al., 2002; Fukumoto et al., 2001). From the hydrate formation rates measured for different systems it was acknowledged that hydrate conversion was limited, especially after a certain amount of solid crystals had been produced. The crystals have to be removed from the water circulation line to prevent plugging the transmission lines. It is also challenging to avoid clogging the nozzle. The other contacting mode suggested was the injection of micro-bubbles into bulk water (Takahashi et al., 2003). The micro-bubble has a very long lifetime in the liquid which allows sufficient time for hydrate formation. Static mixers have also been proposed recently in a commercial hydrate plant to enhance the hydrate formation rate by increasing the contact of gas and liquid (D’Aquino, 2003). Shell and tube heat exchangers are employed to remove the heat generated by hydrate formation immediately. This technology was initially proposed by Yamasaki’s group at AIST Tsukuba, Japan that claimed that much less energy consumption is needed (Tajima et al., 2004; 2005a; 2005b; 2006). It has been reported that the exposure of water on a cold metal surface may help to improve the kinetics (Matsuda et al., 2006; Li et al., 2005) although hydrate conversion is limited when the surface becomes covered by hydrates. All previously mentioned contacting modes work well mostly at the early stages of hydrate formation. However complete water to hydrate conversion is unlikely for most methods. There are always possibilities that the water molecules are occluded between hydrate chunks and which limit the contact between gas and water, and hence the extent of conversion. High power consumption is also required to create new interfaces between water and gas and this increases the operational cost (except for the static mixer). Hence additional stages may be required to convert un-reacted water or to separate the hydrate from the slurry. There are also several alternatives to hydrate synthesis that do not involve mixing. The first one uses suitable surfactants to enhance the contact between water and gas and forms porous hydrates (Zhong and Rogers, 2000; Sun et al., 2003; Link et al., 2003; Lin et al., 2004; Wanatabe et al., 2005; Gayet et al. 2005). The addition of a small amount of sodium dodecyl sulfate (SDS) anionic surfactant was reported to increase hydrate formation rates and water to hydrate conversion by preventing the agglomeration of hydrate crystals (Sun et al., 2003; Link et al., 2003; Lin et al., 2004; Wanatabe et al., 2005; Gayet et al. 2005). Hydrate growth along the wall of the reaction vessel forms open pore structures that transport the water via capillary action towards the hydrate crystallization front and this keeps the water-gas interface free of hydrate. It was reported that 97% of water can be converted into hydrate (Link et al., 2003; Gayet et al. 2005). SDS also enhances the decomposition kinetics (Lin et al., 2004). Rapid hydrate formation towards full conversion was successfully reported by confining water in porous media (silica gel) where ~90% of water could be converted into hydrate within an hour when exposed to a carbon dioxide and hydrogen gas (Seo et al., 2005a; Lee et al., 2005a). Unfortunately such kinetics study has never been reported for other systems

24

Peter Englezos, John Ripmeester and Robin Susilo

including methane hydrate. This contacting mode seems promising although the storage density will be less because of the presence of silica gel. The most recently reported contacting mode proposed is fluidization, giving better mass and heat transfer, but this has not been tested experimentally (Servio et al., 2004). It seems that dispersing water in silica gels or by adding SDS surfactant are the best approaches to synthesize hydrate from liquid water considering the fast kinetics achieved, high water to hydrate conversion and lack of energy requirements for mixing. Despite the study on liquid water, hydrate can also be grown from ice. Early hydrate formation kinetics study from ice was reportedly immeasurable due to limited growth (Hwang et al., 1990). This is likely due to the formation of hydrate film that covers the ice surface and exhibits significant mass transfer resistance. Thus a means of mixing is required to expose the surface contact between ice and gas. Measurable rates were observed only when the ice was melted (Hwang et al., 1990). This thermal ramping above the icepoint has been reported to successfully produce pure methane hydrate in large quantities (Stern et al., 1998). Although this is commonly employed in laboratory practice only, the thermal ramping procedure seems attractive from the economical point of view to be applied in a larger scale because agitation is not required. Kinetics studies on hydrate formation from ice have been reported recently with in-situ measurements using diffraction and spectroscopy techniques (Susilo et al., 2006; Kuhs et al., 2006; Staykova et al., 2003). Field emission-scanning electron microscopy (FESEM) was also employed to examine the grain structure, and texture on ice/hydrate (Stern et al., 2004). The formation of meso to macro porous on the surface of hydrate covered ice particles was observed (Klapproth et al., 2003; Staykova et al., 2003). Generally, hydrate growth from ice can be described into 2 stages (Susilo et al., 2006; Kuhs et al., 2006; Staykova et al., 2003). The growth in the first stage is the formation of hydrate film covering the ice surface which occurs relatively fast due to direct contact between ice surface and the gas phase. This occurs until approximately 10% of ice is converted into hydrate (Wang et al., 2002). Recently the presence of gas film was also observed (Susilo et al., 2006). The gas concentration at the bulk and on the ice surface is not equal. Mixing in the gas phase is required to remove the gas film resistance (Lee et al., 2005c). The second growth stage can be divided into two steps. The first step is mass transport from the bulk gas phase across the hydrate film, which is a slow process and considered as the rate limiting step. The second step is the interfacial reaction at the hydrateice interface. This hydrate reaction kinetics is the same as the well-known gas-solid reaction with the product (ash) cover the solid interface which can be modeled with the shrinking core model (Levenspiel, 1999). A more sophisticated model was developed and improved (Salamantin et al., 1998; Staykova et al., 2003; Genov et al., 2004; Kuhs et al., 2006). However the phenomena on temperature ramping cannot be described by the shrinking core model. The melting of ice covered by the hydrate layer melts at the usual melting point (Moudrakovski et al., 1999). Ice melting causes the pressure inside the hydrate shell to drop as a result of different density between ice and water. Hence the hydrate on the inside shell may decompose that leads to the breakage of the shell and more contact between unreacted water and gas. The broken pieces from the hydrate shell may in turn act as the hydrate seeds that further speed up the kinetics (Susilo et al., 2006).

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 25

Hydrate Storage and Decomposition Hydrate dissociates at temperature and pressure condition below the phase boundary. Hence recovery of methane gas from hydrates can be achieved by increasing the temperature or lowering the pressure. Hydrate decomposes faster at pressure and temperature conditions further away from the three-phase line. The gas release rate is usually fast at the beginning and decreases monotonically with time. However, methane hydrate is known to have anomalies or also called as “self-preservation” effect at temperature between 240-268K (Stern et al., 2001a; Stern et al., 2003; Kuhs et al., 2004). Surprisingly, this behavior is not observed in methane-ethane in sII hydrate (Stern et al., 2003). The study on sH methane hydrates is not available yet. This is rather interesting because apparently ice shielding for gas diffusion is not the real explanation of the “self preservation” mechanism. Hence it remains a puzzle why the decomposition of sI methane hydrate is delayed and others are not. Methane sI hydrate at 268K can be stored for more than two weeks at atmospheric condition and over a month at pressure above 1MPa (Circone et al., 2004a). Rapid hydrate decomposition at temperature above the icepoint is reported, that is governed by the intrinsic dissociation reaction (Sun and Chen, 2006). This is advantageous because methane gas from the gas fields can be stored and transported to energy importing countries in form of hydrate pellets. The hydrates can be stored in a freezer at atmospheric condition for about 2 weeks or a slightly higher pressure (10-20 atm) for longer storage time. A specialized-design hydrate carrier ship was proposed (Ota et al., 2002). It is also of practical interests to control hydrate decomposition. The early studies were intended to understand the mechanism of hydrate decomposition when plugging in pipeline was encountered. The hydrate crystals generally decompose by de-pressurization (Kelkar et al., 1998, Peters et al., 2000; Hong et al., 2006). Thermal stimulation has also been considered in order to provide strategies for methane recovery from the natural hydrate by thermal stimulation (Selim and Sloan, 1989; Ji et al., 2001; Hong et al., 2003; Hong and PooladiDarvish, 2005). Hydrate decomposition studies may also be applicable to gas storage in hydrates. The first hydrate decomposition kinetics model was reported from Bishnoi’s laboratory that incorporates the intrinsic kinetics (Kim et al., 1987). Another rigorous model was also reported by using Fourier’s law in one dimension with moving boundary problem due to thermal conduction (Ullerich et al., 1987). The model assumes that the solid ice forms right after hydrate decomposes that eventually melt into gas and water when heat is applied on the wall. Hence there are two boundaries (ice-hydrate and ice-water) that move inward away from the heat source. The extended model that incorporates the heat transfer effect and intrinsic kinetics was also presented (Jamalludin et al., 1989). A model for hydrate dissociation in porous media by thermal stimulation was presented (Selim and Sloan, 1989). The model was improved over time by taking into account the porosity of the hydrate plug and extended the dimension from Cartesian to cylindrical coordinates (Kelkar et al., 1998, Peters et al., 2000). Most models predict the time required to completely dissociate the solid phase and the decomposition rates at different pressure, temperature, or porosity. Simulation results were generally in good agreement with the experimental data. The simulation approach using molecular dynamics was also reported (English et al., 2005).The study on hydrate decomposition in silica gel meso-porous (Aladko et al., 2006a) and mixed hydrate with methane, ethane, and propane mixture (Kawamura et al., 2003; 2006) were also reported.

26

Peter Englezos, John Ripmeester and Robin Susilo

Tuning Clathrate Hydrates Methane in sI hydrate requires a certain pressure and temperature condition for long term storage. Although the “self-preservation” effect does sound attractive, it may not be the perfect solution because the hydrate is essentially in meta-stable condition. A more stable hydrate is preferable especially from the technical, safety, and economical point of view. Hence the challenge is to find an additive which can help stabilize the hydrate at lower pressures and/or higher temperatures. A more stable hydrate can be achieved by introducing a second guest molecule to form sII or sH hydrate. The obvious drawback is that the addition of a second molecule implies a reduction in methane content in hydrates due to less accessible cages for methane. The question is then, is it worthy to have a stable hydrate with a lesser amount of methane content?

Methane in sII hydrate There are several sII hydrate formers which can be added to methane and significantly lower the hydrate equilibrium pressure at a given temperature. They include water soluble guests (de Deugd et al., 2001; Seo et al., 2001): tetrahydrofuran (THF), tetrahydropyran (THP), acetone, propylene oxide, 1,4-dioxolane, 1,3-dioxolane and water insoluble guests (Tohidi et al., 1996; 1997; Sloan, 1998; Deaton and Frost, 1946): propane, cyclopentane, cyclohexane, neopentane. Hydrate phase diagram for methane, ethane, and propane mixtures are available (Ballard and Sloan, 2001). The pressure reductions depend on the type and concentration of the larger molecule. Large pressure reduction was reported when THF or cyclopentane is employed as the second guest. Larger pressure depression is achievable at higher concentration up to the stoichiometric (5.6%mol). Unfortunately the addition of THF reduces the methane occupancy as predicted from thermodynamics simulations (Mooijer-van den Heuvel et al., 2000; de Deugd et al., 2001). The large guest occupies the large cage fully with methane restricted to the small cage only. The measured methane occupancy was also reported from NMR measurements (Seo and Lee, 2003). It was found that methane occupies only 37% of the small cages whereas the large cage is fully occupied by THF when synthesized at 4MPa and 292.15K. This corresponds to methane volume density of 51 (v/v). Hence there is a trade-off between operating pressure and occupancy in hydrate. The most innovative approach is to tune the methane content with varying the large guest concentrations. This was successfully achieved by using THF (Seo et al., 2005b) and t-butyl amine (Kim et al., 2005) as the large guest. Methane is capable to enter the large cage of sII hydrate when the large guest concentration is lowered below the stoichiometric. Surprisingly methane occupancy diminishes below the so called the “critical guest concentration (CGC)” which is specific to the large guest molecule (Kim et al., 2006). The critical concentration for THF and t-butyl amine is 0.2 % and 1%. The methane content in hydrates is at the maximum at CGC but decreases below CGC. Unfortunately the hydrate phase equilibrium is also shifted to a higher pressure when the large guest concentration is reduced below stoichiometric (Seo et al., 2005b). The other drawback identified was the slow kinetics to synthesize such hydrate. Nevertheless, there is sufficient evidence that the gas content in hydrate can be adjusted although the hydrate stability has to be compromised. Hydrate kinetics studies of methane mixed hydrate in sII hydrate are limited. This is mostly due to complication when working with mixture where the gas composition changes over time due to preferential guest dissolution in water and preferential encapsulation in the

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 27 hydrate cages. Therefore there may be a change in hydrate composition or even the structure as well over time. In a NMR kinetics study of methane-propane hydrate propane was found to fully occupy the large cage with methane in the small cage only (Kini e al., 2004). The large cages filling rate was twice faster than the small cages although the number of large cages is only half of the small cages. Hence it was hypothesized that the methane occupancy was small. A two-step hydrate formation was also reported from the pressure profile and gas chromatography measurements (Uchida et al., 2004). It was found that sII hydrates was formed in the first step but transform to sI hydrate in the second step. Kinetics study with water spraying mode for methane (90%), ethane (7%), and propane (3%) mixture was also reported (Tsuji et al., 2005a; 2005b). The gas uptake rates and the gas phase composition were measured during the hydrate formation. It was found that the gas consumption rate for methane in sI hydrate and gas mixture in sII hydrate are not noticeably different. The composition of ethane and propane in the gas phase was found to decrease over time until it reaches a steady state. Consequently the gas phase becomes methane-rich. These results further suggest that methane content in sII hydrate is low hence storing methane in sII hydrate seems not too promising. Tuning sII hydrate has been successfully demonstrated to increase the methane content but at the expense of slower kinetics as well.

Methane in sH Hydrate There are at least 26 large molecules known to form sH hydrate (Ripmeester and Ratcliffe, 1990; Lee et al., 2006a). Hydrate phase equilibria at different temperature and pressure conditions are widely available in literature for most sH hydrate formers with methane (Mehta and Sloan, 1993; Hutz and Englezos, 1996; Sun et al., 2002; Ohmura et al., 2003a; 2003c; 2005a; 2005c; Makino et al., 2004; Nakamura et al., 2003; Hara et al., 2005; Ohmura et al., 2005a; 2005c). A hydrate phase equilibrium prediction method is also available (Mehta and Sloan, 1996; Chen et al., 2003; Ma et al., 2005). Morphology studies of sH hydrate systems were reported (Servio and Englezos, 2003b; Ohmura et al., 2005b). The lattice constants of sH hydrate are dependent on the size of the LMGS (Takeya et al., 2006). Single crystals of sH hydrate with adamantane (Kirchner et al., 2004) and methylcyclohexane (Udachin et al., 2002) have been analyzed by X-Ray diffraction. The occupancy of the large cage was determined to be full with the methane occupancy at around 80%. The solid-state NMR analysis also confirms the occupancy values (Seo and Lee, 2003; Susilo et al., 2007a). The actual gas content in sH hydrate was reported recently (Susilo et al., 2007a). It was found that the methane content in sH hydrate is approximately 20% to 40% less than in sI hydrates depending on the LMGS employed and pressure during the synthesis. This confirms the prediction by using the thermodynamics model (Mooijer van Heuvel et al., 2000). Because most sH hydrate former are poorly miscible in water, the presence of LMGS may act as additional mass transfer barrier resulting a slow kinetics. Hence an efficient way to maintain the contact among methane, water, and LMGS is important in synthesizing sH hydrate. Mori’s laboratory suggested the water spraying technique (Ohmura et al., 2002; Tsuji et al., 2004). Six LMGSs were tested: methylcyclohexane (MCH), 2,2-dimethylbutane (neohexane/NH), tert-butyl methyl ether (TBME), 3-methyl-1-butanol (isoamyl alcohol), 3,3dimethyl-2-butanone (pinacolone), and 2-methylcyclohexanone (2MCH). Interestingly, the rates of sH hydrate formation may exceed that of sI, especially for the system with TBME, 2MCH and pinacolone. However TBME is recommended due to economics (Tsuji et al., 2004). It was found that the hydrate formation rates were dependent on the type of LMGS and

28

Peter Englezos, John Ripmeester and Robin Susilo

hence it was suggested that the formation rates are correlated with the solubility of LMGS in water. A systematic effort is underway in our laboratories on structure H hydrates. Three large guest molecules (LMGS) were chosen: tert-butyl methyl ether (TBME), neohexane (NH), and methylcyclohexane (MCH). Liquid-liquid equilibria (LLE) of water-LMGS and vapor-liquidliquid equilibria (VLLE) with methane were reported (Susilo et al., 2005). LMGS solubility in water was found to increase with pressure but decrease with temperature as shown in Figure 3. Methane solubility in water and LMGS also increase with pressure. This explains why hydrates generally form at higher pressure and lower temperature. It is important to note that TBME is two orders of magnitude more soluble in water than MCH or NH. Hence one may expect hydrate to form faster with TBME as the LMGS followed by NH and MCH. NH

MCH

TBME 6000

35

4000

25 20

3000

15

2000

10 1000

5 0

0 0

500 NH

NH and MCH in water / ppm

TBME in water / ppm

5000

30

1000

1500

Pressure / kPa MCH

2000 TBME

10

14000

9 8

12000 10000

7 6

8000

5 4

6000

3

4000

2

TBME in water / ppm

NH and MCH in water / ppm

40

2000

1 0

0 275.5

283.15

298.15

Temperature / K

Figure 3. Solubility data of neohexane (NH), methylcyclohexane (MCH), and tert-butyl methyl ether (MTBE) in water. The upper figure is the vapor-liquid-liquid equilibria at 275.5K and the lower figure is the liquid-liquid equilibria at atmospheric pressure. (Reprinted from Fluid Phase Equilibria (Susilo et al., 2005), Copy right (2005) with permission from Elsevier).

A kinetic study in a well-stirred semi-batch reactor was conducted to determine the rate of methane as shown in Figure 4 (Lee et al., 2005b). As seen, the system with TBME has the shortest nucleation time and fastest hydrate growth rate followed by NH and MCH. This trend

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 29 agrees with that of the solubility measurements. The measured rates are also in good agreement with the measurements by using the water spraying contacting mode (Tsuji et al., 2004). The rate of sI hydrate formation for comparison is also indicated. As seen, the nucleation without LMGS is almost instantaneous however the rate is slower than TBME. The observed rates are generally linear in the first 30 minutes however they tend to slow down as soon as sufficient hydrate slurry is present in the suspension. This occurs when ~10% of water has been converted into hydrate where the magnetic induced mixing cannot perform its job well. Hence better contacting mode may be required that implies higher power consumption required due to an increase in viscosity of the suspension. This is not encouraging because of the low conversion. 0.20

CH4 moles consumed

Driving force =1.00 MPa at 275.50 K CH4-H2O-TBME

0.15

O -H 2 CH 4

O-MCH CH 4-H 2

0.10 O-NH CH 4-H 2

0.05

0.00 0

50

100

150

1000

1500

2000

Time (min) Figure 4. Methane uptake profile in sI and sH hydrate at 275.5K and 1MPa pressure above their corresponding equilibrium pressure. (Reprinted from Energy and Fuels (Lee et al., 2005b), Copy right (2005) with permission from American Chemical Society).

It is also indicated that there are two hydrate growth stages for the MCH system. By employing NMR spectroscopy and imaging the methane partition and evolution in the gas, liquid, and hydrate phase and ice was monitored over time. In addition, hydrate conversion, methane diffusivity in LMGS and LMGS contact with ice were studied (Susilo et al., 2006). It is noted at this point that inadequate mixing and high pressures requirements would be the limitation with NMR. Hence the use of magic angle spinning (MAS) technique is impossible which makes it difficult to resolve the spectra. It is also challenging to obtain acceptable signal to noise ratio without acquiring the spectra longer than the actual kinetics. Hence deuterated methane-d4 was used in our study to eliminate signal interference from the proton of ice and LMGS. Besides, hydrate has to be synthesized from ice powder to eliminate the long nucleation time of liquid water and better contact between LMGS and ice. Hence the mass transfer resistance is due to the dissolution of methane in the liquid LMGS film covering the ice powder. Methane in the dissolved liquid or gas phase and solid phase could

30

Peter Englezos, John Ripmeester and Robin Susilo

be distinguished due to the difference in molecular motions. The motion of methane molecules in the fluid phase show a sharp isotropic lorentzian NMR line-shape but in the solid phase is restricted in the hydrate cage hence a broad Gaussian line-shape is observed. The normalized intensity of methane diffusion in LMGS liquid and growth in hydrate phase is shown in Figure 5. It was found that methane growth in hydrate phase increases from MCH
1.00

Intensity

0.80

0.60

0.40 Ice-CD4 Ice-TBME-CD4 Ice-NH-CD4 Ice-MCH-CD4

0.20

0.00 0

200

400

600

800

1000

1200

1000

1200

Time [min]

1.00 Ice-TBME-CD4 Ice-CD4 Ice-NH-CD4 Ice-MCH-CD4

Intensity

0.80

0.60

0.40

0.20

0.00 0

200

400

600

800

Time [min]

Figure 5. Normalized intensity evolution of methane in the gas + dissolved in LMGS liquid phase (left) and hydrate phase (right) at 253K and ~4.5MPa. (Reprinted from J. Phys. Chem. B (Susilo et al., 2006), Copy right (2006) with permission from American Chemical Society).

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 31 All intensities were calibrated so that the amount of methane in the hydrate phase can be quantified. Hence the ice to hydrate conversion can be calculated assuming methane occupies both small and medium cage by 80% (Udachin et al., 2002). The corresponding ice to hydrate conversion is given in Figure 6. At 253K and ~4.5MPa, the maximum conversion achieved is ~35% for TBME that is still far from completion. The conversions for the systems with NH, MCH and sI methane hydrate are even lower than 20%. Generally ice gets converted into hydrate quickly during the first 5 hours before the rates slow down significantly. This is due to the formation of hydrate film on ice surface that add resistances to the mass transfer. Fortunately, the temperature ramping procedure that is commonly used in synthesizing sI methane hydrate was also applicable in sH hydrate. Rapid hydrate formation was observed when the temperature was ramped above the icepoint (Figure 6). The total conversion obtained after 40 hours of reaction time was still far below completion unfortunately. This is due to limited space in the NMR cell. The cell was fully loaded with ice particles to acquire sufficient signal intensity and the formation of hydrate eventually requires volume expansion due to methane inclusion. Thus there is not anymore space available for the unreacted ice to get converted into hydrate.

Hydrate Conversion [%]

60 Ice-TBME-CD4 Ice-NH-CD4 Ice-CD4 Ice-MCH-CD4

50 40 30 20 10 0 0

300

600

900

1200

1500

1800

2100

2400

Time [min]

Figure 6. Ice to hydrate conversion rate, assuming both 512 and 435663 are 80% occupied with methane, T = 253K for the first 20 hours and T = 274K afterward. (Reprinted from J. Phys. Chem. B (Susilo et al., 2006), Copy right (2006) with permission from American Chemical Society).

Figure 7 shows the MCH contact with ice at different ice packing densities. Generally LMGS may fill the free space between ice particles upon pressurization. However if the ice is highly packed, The LMGS contact with the ice particles is poor. Thus some ice particles are isolated and this limits hydrate conversion. The bulk measurement in an external pressure vessel shows that complete hydrate conversion can be achieved with a thermal ramping procedure when the ice is loosely packed (Susilo et al., 2007b). Surprisingly, a second hydrate growth stage for the system with NH at 274K was observed. This is similar to the one observed from the kinetic data mentioned earlier but for the MCH system. This is attributed to rupture when the occluded water is transformed into hydrate. The rapid conversion from

32

Peter Englezos, John Ripmeester and Robin Susilo

melting ice into hydrate creates internal pressure due to volume expansion of hydrate formation that eventually provides new access for the water to get converted into hydrate.

Figure 7. Proton density image of Ice+MCH with different degree of ice packing: the left images correspond to highly packed ice and the right images correspond to not so highly packed ice. (Reprinted from J. Phys. Chem. B (Susilo et al., 2006), Copy right (2006) with permission from American Chemical Society).

The trend of hydrate formation rates changes between TBME and NH or MCH system when the unreacted ice is melted (Susilo et al., 2007b). The TBME system shows a much slower phase transformation towards full conversion at temperatures above the icepoint whereas the other hydrophobic guests MCH and NH show a rapid conversion. It was suggested that the high solubility of TBME in water lowers the water activity resulting the slow kinetics. Hence the TBME acts as hydrate inhibitor the same way as other thermodynamics inhibitor except TBME is also a hydrate guest molecule. The most recent work from our group reported that methane occupancy in sH hydrate is dependent on the LMGS type. The methane occupancy was higher for MCH and NH system than TBME when synthesized at the same operating condition (Susilo et al., 2007a). However it is unsure if the lower methane occupancy was due to differences in hydrate growth rates (kinetics) or thermodynamics (stability). Further studies are required to elucidate this and it is currently being investigated. The contact between LMGS and ice was monitored through micro-imaging NMR experiments. The polar guest (TBME) wets the ice surface whereas the hydrophobic guest (NH and MCH) does not wet ice well as shown in Figure 8. The poor wetting between hydrophobic molecule and ice limits the contact and hence the hydrate growth/conversion. The fact that hydrate growth rates vary among different LMGS systems can be explained by the difference observed from methane diffusivity and wetting between LMGS and ice. This clarifies why the kinetics of the NH system is slower than in the TBME system although the methane diffusivity in LMGS is similar. The system with MCH has the slowest kinetics (nucleation time and hydrate growth rate) due to slow methane diffusivity and poor contact between MCH and ice powder. It is clear that the contact between water and guest molecules is very important in hydrate kinetics. The presence of LMGS does not necessarily limit the mass transfer of methane from the bulk gas phase to the water interface because of high

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 33 methane solubility in LMGS. However methane diffusivity has to be considered when no mixing is applied. The wetting and solubility of LMGS with water play an important role in sH kinetics.

Figure 8. Wetting and diffusion of LMGS (MCH and TBME) between ice particles at atmospheric condition observed by 1H micro-imaging NMR. (Reprinted from J. Phys. Chem. B (Susilo et al., 2006), Copy right (2006) with permission from American Chemical Society).

Multiple Occupancy in large cage. The challenge for methane storage in sH hydrate is to increase the storage capacity by putting multiple methane molecules into the large cage. This can be done at extremely high pressures although there is a disagreement about the number of methane molecules in the large cage. The solid phase analysis obtained by X-Ray and neutron diffraction suggested there are 5 methane molecules in the large cage (Loveday et al., 2003a) but only 2-3 methane molecules reported from Raman spectroscopy (Kumazaki et al., 2004). Molecular dynamics (MD) simulations are currently employed to verify the multiple guest occupancy. Such comparison between MD simulation and experimental measurements was presented for rare gases (Ne, Ar, Kr, Xe) in sH hydrate at high pressures (Alavi et al., 2006b; Ogienko et al., 2006; Manakov et al., 2004). It still remains to be seen if sH hydrate tuning can also be performed in the same way as sII hydrates. Recently it was reported that methane can replace the methylcyclohexane in sH hydrate by pressurizing the hydrate with methane up to 110 bar (Yeon et al., 2006).

Commercialization of Hydrate Technology for Gas Storage Application Hydrate or gas to solids technology (GTS) is expected to be positioned between the LNG and CNG technologies in terms of gas storage density, temperature and pressure condition. The motivation though is the substantial reduction in the capital and operating cost. Hence GTS technology may be utilized to recover the stranded gas and even compete with the current established LNG and CNG technologies for larger gas fields. Japan leads the way in the effort to commercialize GTS. Mitsui Engineering and Shipbuilding Corp., Inc (MES) has completed a hydrate production pilot plant which is able to produce 600 kg of hydrate per

34

Peter Englezos, John Ripmeester and Robin Susilo

day. The most recent interest came from an Indian oil and gas company who has reached an agreement with Aker Kvaerner (AK) for storing and transporting natural gas via hydrates. AK is a Norwegian company who collaborates with Norwegian University of Science and Technology (NTNU) and MES to develop, establish, and commercialize the GTS technology.

2. CO2 Capture from Treated Flue Gas (CO2/N2/O2) and Fuel Gas (CO2/H2) Gas Mixtures Global climate change and ocean acidification as a result of increased emissions of carbon dioxide are now serious global environmental issues. Consequently, capture of carbon dioxide from existing power plants along with improving energy efficiency in energy utilization is now a necessary action plan to stabilize and reduce carbon emissions in the atmosphere. There are two general approaches to CO2 capture. Post-combustion capture refers to separating carbon dioxide from a flue gas (Klara and Srivastava, 2002). Precombustion capture refers to capturing carbon prior to combustion. This would be a suitable approach for new power plants in which the fossil fuel is gasified and then through the shift reaction is converted into a stream of H2 and CO2 (Barchas and Davis, 1992). The CO2 can then be removed for disposal. The resultant stream of H2 could be used in fuel cells and not just in a gas turbine. Processes for the removal of a component from a multi-component gaseous stream include cryogenic fractionation, selective adsorption by solid adsorbents, gas absorption, membrane separation etc. (Barchas and Davis, 1992; Kikkinides et al., 1993). It has been estimated that, the cost of separation and disposal of CO2 from existing coal fired, air blown boilers would increase the cost of electricity by about 75% (Hendriks and Turkenburg, 1990). The cost of separation of CO2 alone reduces the power generation efficiency from 38 to 26% (Chakma et al., 1995). Aaron and Tsouris (2005) have reviewed the processes for CO2 separation from flue gases. Liquid absorption using amines was considered the most promising current method while some other methods are promising but too new for comparison. The use of gas hydrates was considered on of the novel methods under investigation for separating CO2 from flue gas (mixture of CO2, N2 and O2) or synthesis gas (mixture of CO2 and H2). Treated flue gas contains CO2, N2, O2. Since N2 and O2 form hydrate crystals at approximately the same conditions the treated flue gas is considered a CO2/N2 mixture. Thus in post-combustion capture from power plants the task is to separate CO2 from a CO2/N2 mixture in which the CO2 molar content is approximately 15-20 %. Pre-combustion capture involves separation of CO2 from a mixture with H2 in which the CO2 molar concentration is approximately 40 % (Klara and Srivastava, 2002; Kang and Lee, 2000; Englezos and Lee, 2005). CO2 capture from flue gas mixtures. Figure 9 below shows a conceptual process based on gas hydrate data from Linga et al. (2006; 2007a; 2007b). Following a one-stage hydrate formation / decomposition process for the CO2/N2 mixture, a CO2-rich gas is obtained which contains 57 % CO2 at 10 MPa. A second hydrate foration/decomposition stage is used with the CO2-rich gas that results in a new CO2-rich stream containing about 83 % CO2. A hybrid process is advocated where the lean CO2 streams are passing through a gas separation process

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 35 like a membrane one. The process is completed with a third gas hydrate process stage that may result in 98-99 % CO2 stream. N2

Membrane separation

17 % CO2 83 % N2

Hydrate Process -stage 1

CO2

H2 O

Hydrate Process -stage 2

Hydrate Process -stage 3 98-99 % CO2

H2 O

for Disposal

H2 O

Figure 9. A hybrid process for CO2 recovery from fuel gas.

CO2 capture from fuel gas mixtures. Figure 10 shows a conceptual hydrate-based process for the separation of CO2 from a fuel gas. Only two hydrate stages are needed in this case to recover a 98-99 mol % of CO2 (Linga et al., 2007b; Kumar et al., 2006).

H2 Membrane separation

40 % CO2 60 % H2

Hydrate Processstage 1

CO2

Hydrate Processstage 2

H2O

H2O

98-99 % CO2

Figure 10. A hybrid hydrate-membrane process for CO2 recovery from fuel gas.

From an engineering standpoint the challenges are in some regards similar to those faced in the natural gas storage area and discussed above. One has to find efficient ways of contacting gas and water in order to scale up the process effectively. The above conceptual

36

Peter Englezos, John Ripmeester and Robin Susilo

processes are based on a laboratory stirred vessel operating in a semi-batch mode. Andersson and Haines (2005) have used an industrial scale stirred vessel hydrate rector to produce hydrates for storage in the ocean. An alternative approach is hydrate formation in a packed bed with silica gel (Seo et al., 2005). Park et al. (2006b) reported phase equilibrium data which led them to assume that a three-stage hydrate formation/decomposition process would result a stream with more than 96 mol % CO2. Finally, the use of additives is required to lower the operating conditions and reduce compression costs. This should be done in a way that does not compromise the separation factor. Use of tetrahydrofuran in the hydrate formation process has been suggested in the literature in order to reduce the hydrate formation pressure for the CO2/N2 system (Kang and Lee, 2000; Seo et al., 2005). A suitable additive is also required to lower the hydrate forming conditions for the CO2/H2 system. One such additive might be propane (Kumar et al., 2006).

3. Hydrogen Storage Hydrogen is a light gas that is easily combustible. It is a clean fuel with high energy density per weight but low on volume. It is the preferred fuel for fuel-cell but hydrogen storage is always the barrier for the development of fuel-cell technology (Ross, 2006; Chalk and Miller, 2006). A safe way to store hydrogen in a reasonable energy density is always challenging. Large energy density can be obtained when hydrogen gas is compressed or liquefied. However both processes require extreme conditions that are not practical and safe. Hydrate is one of the alternative technologies that are currently being explored for storing hydrogen. This was realized since hydrogen was discovered to form sII hydrates back in 2002 (Mao et al., 2002). Two hydrogen molecules were reportedly able to fill the small cages and four molecules in the large cages. Hence a total hydrogen content of 5 wt% is achievable in hydrate. Hydrogen occupancies were also determined as a function of pressure and temperature using neutron diffraction (Lokshin et al., 2004). It was found that the small cage is singly occupied by hydrogen but the large cage can be occupied by two to four molecules depending on the temperature. This was also supported by simulation work using the molecular dynamics (Alavi et al., 2005) and lattice dynamics (Inerbaev et al., 2006). The drawback is that pure hydrogen hydrates require high pressure condition that is not practicable. Therefore addition of a second molecule that is able to reduce the formation pressure significantly is needed. Recently it was found that introducing tetrahydrofuran (THF) as a second guest molecule may lower the required pressures significantly (Florusse et al., 2004). This is a real breakthrough that increases the hopes for hydrogen storage via gas hydrates. Obviously lower hydrogen content in hydrate is expected. A hydrate tuning strategy to increase the hydrate content in hydrate was also reported (Lee et al., 2005a). Lowering the THF concentration allows pressure reduction while maintaining reasonably high hydrogen content in hydrate up to 4 wt%. Surprisingly this tuning procedure is not repeatable (Strobel et al., 2006; Hester et al., 2006). It was reported that the maximum hydrogen content is about 1wt% only with all THF occupy the large cage and only a single hydrogen molecule in the small cage at moderate pressure (less than 60MPa). The molecular dynamics study confirms the stabilizing effect of THF (Alavi et al., 2006a). The small cage is more favorable when there is only a single hydrogen molecule inside. The unit cell volume increases significantly when a second

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 37 hydrogen molecule is added. Hence the hydrogen content is practically limited to ~1wt%. It is unsure if the double occupancy of the small cage is observed as a meta-stable condition where the hydrogen may escape or diffuse to another empty small cage or out from the hydrate cage. Further studies are needed to clarify this discrepancy. The requirement for hydrogen storage in USA is currently 6wt% although in certain countries a lower storage density may be acceptable. It is well-known that the maximum hydrogen storage capacity in sII hydrate is only 5wt%, which is a bit short from 6wt%. This occurs only at a high pressure that is unrealistic for practical applications. A more realistic pressure limits the hydrogen content around 1wt% only. Hence it is clear that storing hydrogen in hydrate still have a long shot to go. The other options available are to employ other sII guest molecules or to form sH hydrate. So far there is no publication on hydrogen storage using other help guest than THF. The success finding of THF should open up the possibility of other guest molecule to work the same way as THF or even better. Storing hydrogen in sH hydrate is also attractive since the large cage of sH hydrate may allow much more hydrogen molecule inside. However up to date there is no published work available yet on this. Hence there are still many opportunities to explore the potential of hydrate for hydrogen storage. The other drawback for the hydrate tuning that is worthwhile to mention is the slow reaction time. Significant improvement in the kinetics was accomplished by dispersing the water in a confined space using silica beads (Lee et al., 2005a). Full hydrate conversion can be achieved within an hour with the proposed methodology where the normal reaction time takes a week or more. However more studies are required to verify this.

4. Flow Assurance in Hydrocarbon Pipelines Currently, the only area where gas hydrates are practiced on an industrial scale is the area of “flow assurance” in oil and gas pipelines (Sloan, 1998; 2005). This subject was reviewed recently by Kelland (2006). The oil and gas industry has relied mostly on the use of methanol and glycols to avoid plugging of transportation and processing facilities (Dholabhai et al. 1992). Use of methanol and glycol is costly because of the high treatment amounts required. This has motivated the search for methods based on the injection of polymer-based chemicals at low dosages in the water phase (Fu, 2002; Lovell and Pakulski, 2003, Kelland, 2006).These inhibiting chemicals do not prevent the formation of hydrates since the operating conditions are still within the hydrate formation region but interfere with hydrate nucleation, growth and agglomeration of hydrate particles (Huo et al., 2001). Thus, they are subdivided into so called kinetic inhibitors (KI) and anti-agglomerates (AA). The action of LDHI is analogous to that of glycoproteins in the blood of Antarctic fish that enable them to exist at low temperatures (Zeng et al. 2003; Marshall et al. 2004). Research in this field is ongoing aiming to understand the mechanism of action of kinetic inhibitors. Lee and Englezos (2005) showed that inclusion of polyethylene oxide (PEO) to a kinetic inhibitor solution was found to enhance by an order of magnitude the performance of the hydrate inhibitor. Binding of inhibitor molecules to the surface of hydrate crystals was considered to be the key aspect of the mechanism of kinetic inhibition (Anderson et al., 2005). Kinetic inhibitors exhibit unusual effects on hydrate formation with implications to processing (Lee and Englezos, 2006). Gas hydrate formation experiments were conducted

38

Peter Englezos, John Ripmeester and Robin Susilo

using water droplets or water contained in cylindrical glass columns. It was found that the water droplet containing the strong commercial inhibitor was found to collapse prior to nucleation and spread out on a Teflon surface. Subsequently, hydrate was formed as a layer on the surface. Catastrophic growth and spreading of the hydrate crystals was also observed during hydrate formation in the glass columns in the presence of the kinetic inhibitor. Finally, when polyethylene oxide (PEO) was added into the kinetic inhibitor solution the memory effect on the induction time decreased dramatically (“suppression of memory”). Antifreeze protein from winter flounder was also found to impact the memory effect. In fact Zeng et al. (2006a) reported that the memory effect was eliminated. The knowledge of the mechanism of action is required in order to synthesize more effective inhibitors. Effectiveness refers to the ability to act increased degrees of undercooling. Environmental considerations have also motivated the search for biologically derived inhibitors instead of synthetic ones (Kelland, 2006; Zeng et al. 2006b)

5. Recovery of Methane from In-situ Methane Hydrate with Carbon Dioxide Injection The idea here is to capture carbon dioxide from power plants and then inject it into natural gas hydrate reservoirs assumed to contain primarily methane hydrate. Thus one achieves the simultaneous sequestration of carbon dioxide with the production of natural gas. Lee et al. (2003) presented laboratory data that showed the replacement of methane molecules by CO2. Yoon et al. (2004) and Ota et al. (2005) confirmed these laboratory findings. Park et al. (2006a) used a CO2/N2 mixture containing 20 mol % carbon dioxide (flue gas) instead of pure CO2 and noticed that the methane recovery increased from 64 to 85 %. A similar idea for sequestering captured CO2 is to use it as cushion gas for natural gas storage in reservoirs (Oldenburg, 2003).

6. Recovery of Energy from Gas Hydrates. The potential of hydrates to become a factor in the energy supply system is discussed in the recent work by Max et al. (2006). Moridis discusses the efforts to classify hydrate reservoirs according to their production potential and uses a reservoir simulation package to establish the reservoir potential (Moridis, 2003; 2004; Moridis et al. 2005). Poladi-Darvish (2004) also uses reservoir simulation to assess production of natural gas from hydrates in the earth. It is noteworthy that the general believed was that the naturally occurring hydrate in the earth is methane hydrate. It is now known from the analysis of natural samples that hydrates in the earth are not simple (Lu et al., 2007). The most ambitious program to assess the economic viability of drilling for the production of gas form hydrates is the International Mallik program. This effort is presented in a comprehensive report by Dallimore and Collett (2005). Max et al. (2006) discuss the issue of gas hydrates as an energy resource extensively in a recent publication.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 39

7. Relationship of Hydrates with Climate Change. The relationship between the naturally occurring gas hydrates and current global warming has become a topic of interest not only among earth and ocean scientists but engineers as well (Kvenvolden, 1993; 1999; 2000, Englezos, 1993). Although the physical processes involved in the phenomenon known as greenhouse effect are well established the resulting global temperature rise can not be easily determined (Schneider, 1990; Taylor, 1991). Hatzikiriakos and Englezos (1993) assessed the possibility of a "runaway" greenhouse effect (RGE) due to the decomposition of in situ methane hydrates due to current global warming. It was found that under a catastrophic scenario (annual temperature rise of 0.08°C), the temperature at the top of a typical methane hydrate zone will begin to rise within the next 100 years. The results also indicated that sub-oceanic hydrates would remain stable within the next 1,000 years. Nisbet (1992) proposed that decomposition of sub-oceanic and continental hydrates could have triggered a RGE and thus contributed to the rapid warming at the end of the last glaciations 13,500 years ago. Moreover, there is evidence that the temperature in the deep ocean increased by about 6oC 55.5 millions of years ago (Zachos et al., 1993, Dickens et al., 1997, Simpson, 2000). This phenomenon is known as latest Paleocene thermal maximum (Kerr, 1997). Characterization of carbonate and organic matter deposited during the latest Paleocene thermal maximum revealed that a large influx of 12C occurred that resulted in 0.25 % drop in the 13C values. The relative compositional imbalance is known as a carbon isotope excursion and lasted approximately 2 x 105 years (Suess et al., 1999, Dickens et al., 1997). It was suggested that submarine seismicity, volcanism or simple gravitational slumping induced catastrophic slope failure on continental margins containing methane hydrate reservoirs (Bains et al., 1999). Based on mass balance calculations it was hypothesized that the source of the lighter carbon was methane hydrate that decomposed (Dickens et al., 1995). Katz et al. (1999) suggested that the Blake Nose east of the Florida coast is the site where most of the methane release took place.

8. Other Applications. The fact that solutes are excluded from the hydrate lattice motivated work on using hydrates for the desalination of seawater and the concentration of aqueous solutions in general. Most of that work occurred in the 1960s and 70s and was reviewed by Englezos (1993). More recent work dealt with water recovery from pulp mill effluents (Ngan and Englezos, 1996); and concentration of liquid foods (Purwanto et al. 2001). The motivation for such work is potential energy savings compared with evaporation and in the case of food applications the preservation of the aroma volatile compounds. Other idea involving hydrates is the photo-catalysis of methane hydrate (Taylor, 2005) and as a medium for cool energy storage. This is particularly attractive to store the cool energy during the off-peak period and discharge it when needed. The hydrate process can be implemented in a closed-loop vaporcompression refrigerator where the hydrate crystallizer may replace a conventional evaporator (Mori and Mori, 1989). The refrigerant in the evaporator can be exposed to water where the energy released by the evaporation of the refrigerant is absorbed by the water to form hydrate. Hence the refrigerant which is a volatile organic compound may be stored in hydrate while it

40

Peter Englezos, John Ripmeester and Robin Susilo

is not in use. The hydrate can be decomposed to cool off the air due to the endothermic reaction of hydrate decomposition.

Concluding Remarks Clathrate of gas hydrates are a research topic of interest to a variety of disciplines in addition to chemical engineering. In most cases the issues are complex and require the collaborative effort of scientists and engineers. In this work we have described the various hydrate topics of interest to chemical engineering research. These areas include the development of clean energy technologies such as natural gas transport and storage, capture of carbon dioxide from flue gases, capture of carbon dioxide from fuel gases and recovery of hydrogen, hydrogen storage. Natural gas storage and transport have reached pilot plant stage in Japan while the CO2 capture technologies and hydrogen storage are at the laboratory stage. The recovery of energy from hydrates in the earth and the relationship of hydrates in the earth with global climate change are briefly mentioned since they are covered extensively elsewhere. Flow assurance in hydrocarbon pipelines is currently the only area where hydrates are involved in an industrial scale. The shift there is towards finding kinetic inhibitors derived from biological sources rather than using synthetic macromolecules. The motivation is environmental considerations. Other applications such as aqueous solution concentration and cool energy storage are also mentioned. The emphasis is on exposing the reader to the state of the art on the various topics and highlighting research needs.

References [1] [2]

[3]

[4]

[5] [6] [7]

Aaron, D. & Tsouris, C. (2005). Separation of CO2 from Flue Gas: A Review. Separation Science and Technology, 40 (1-3), 321-348. Abdalla, B.K. & Abdullatef, N.A. (2005). Simulation and economic evaluation of natural gas hydrates [NGH] as an alternative to liquefied natural gas [LNG]. Catalysis today, 106, 256-258. Aladko, E.Y. Dyadin, A.Y. Fenelonov, V.B. Larionov, E.G. Manakov, A.Y. Mel’gunov, M.S. & Zhurko, F.V. (2006a). Formation and decomposition of ethane, propane, and carbon dioxide hydrates in silica gel mesopores under high pressure. J. Phys. Chem. B, 110 (39), 19717-19725. Aladko, E.Y. Ancharov, A.I. Goryainov, S.V. Kurnosov, A.V. Larionov, E.G. Likhacheva, A.Y. Manakov, A.Y. Potemkin, V.A. Sheromov, M.A. Teplykh, A.E. Voronin, V.I. & Zhurko, F.V. (2006b). New Type of Phase Transformation in Gas Hydrate Forming System at High Pressures. J. Phys. Chem. B, 110 (42), 2137121376. Alavi, S. Ripmeester, J.A. & Klug, D.D. (2005). Molecular-dynamics study of structure II hydrogen clathrates. J. Chem. Phys., 123, 024507. Alavi, S. Ripmeester, J.A. & Klug, D.D. (2006a). Molecular-dynamics simulations of binary structure II hydrogen and THF clathrates. J. Chem. Phys., 124 (1), 014704. Alavi, S. Ripmeester, J.A. & Klug, D.D. (2006b). Stability of rare gas structure H clathrate hydrates. J. Chem. Phys., 125 (10), 104501.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 41 [8] Anderson, F. E. & Prausnitz, J. M. (1986). Inhibition of Gas Hydrates by Methanol. AIChE J., 32, 1321-1333. [9] Anderson, B.J. Tester, J. W. Borghi, G. P. Trout, B.L. (2005). Properties of Inhibitors of Methane Hydrate Formation via Molecular Dynamics Simulations. J Am. Chem. Soc., 127, 17852-17862. [10] Andersson, V. & Haines, M. (2005). Gas Hydrates for Deep Ocean Storage of CO2Novel technology for Utilizing Hydrates for Transport of CO2. Greenhouse Gas Control Technologies, Volume II, in Wilson, M., T. Morris, J. Gale, K. Thambimuthu (Eds). Amsterdam: Elsevier, 1487-1492, [11] Bains, S., Corfield, R.M. & Norris, R.D. (1999). Mechanisms of climate warming at the end of the Paleocene. Science, 285, 724. [12] Ballard, A.L. & Sloan, E.D., Jr. (2001). Hydrate phase diagrams for methane + ethane + propane mixtures. Chem. Eng. Sci., 56, 6883–6895. [13] Ballard, A.L. & Sloan, E.D., Jr. (2002). The next generation of hydrate prediction I. Hydrate standard states and incorporation of spectroscopy. Fluid phase equilibria, 194-197, 371-383. [14] Ballard, L. & Sloan, E.D., Jr. (2004a). The next generation of hydrate prediction IV. A comparison of available hydrate prediction programs. Fluid phase equilibria, 216 (2), 257-270. [15] Ballard, L. & Sloan, E.D., Jr. (2004b). The next generation of hydrate prediction - Part III. Gibbs energy minimization formalism. Fluid phase equilibria, 218 (1), 15-31. [16] Barchas, R. & Davis, R. (1992). The Kerr-McGee/ABB Lummus Crest technology for the recovery of CO2 from stack gases. Energy Conversion and Management, 33 (5-8), 333-340. [17] Barduhn A.J., Towlson, H.E., Hu Y.C. (1962) The Properties of Some New Gas Hydrates and Their Use in Demineralizing SeaWater. AIChE Journal, 22, 176-183. [18] Bi, Y., Guo, T., Zhu, T., Zhang, L., & Chen, L. (2006). Influences of additives on the gas hydrate cool storage process in a new gas hydrate cool storage system. Energy Convers. Manage., 47 (18-19), 2974-2982. [19] Bishnoi, P.R. Gupta, A.K. Englezos P. & Kalogerakis, N.E. (1989). Multiphase Equilibrium Flash Calculations in Systems Containing Gas Hydrates. Fluid Phase Equilibria, 53, 97-104. [20] Bishnoi, P.R. &, Natarajan, V. (1996). Formation and decomposition of gas hydrates. Fluid Phase Equilibria, 117 (1-2), 168-177. [21] Bishnoi, P.R. & Clarke, M.A. (2006). Natural Gas Hydrates. In: Lee, S. editor. Encyclopedia of Chemical Processing. Taylor and Francis, 1849-1863. [22] Buanes, T. Kvamme, B. & Svandal, A. (2006). Computer simulation of CO2 hydrate growth. J. Crystal Growth, 287, 491-494. [23] Buchanan, P. Soper, A.K. Thompson, H. Westacott, R.E. Creek, J.L. Hobson, G. & Koh, C.A. (2005). Search for memory effects in methane hydrate: Structure of water before hydrate formation and after hydrate decomposition. J. Chem. Phys. 123 (16), Art. No. 164507. [24] Buffet, B. & Archer, D. (2004). Global inventory of methane clathrate: sensitivity to changes in the deep ocean. Earth and Planetary Science Letters, 227, 185– 199. [25] Bybee, K. (2004). Overview of the mallik gas-hydrate production research well. J. Petroleum Technology, 56 (4), 53-54.

42

Peter Englezos, John Ripmeester and Robin Susilo

[26] Chakma, A. Mehrotra, A.K. & Nielsen, B. (1995). Comparison of chemical solvents for mitigating CO2 emissions from coal-fired power plants. Heat Recovery Systems & Chp., 15 (2), 231-240. [27] Chalk, S.G. & Miller, J.F. (2006). Key challenges and recent progress in batteries, fuel cells, and hydrogen storage for clean energy systems. J. Power Resources, 159 (1), 73-80. [28] Chatti, I. Delahaye, A. Fournaison, L. & Petitet, J.P. (2005). Benefits and Drawbacks of Clathrate Hydrates: a Review of their areas of interest. Energy Conversion and Management, 46, 1333-1343. [29] Chazallon, B. & Kuhs, W. F. (2002). In situ structural properties of N-2-, O-2-, and airclathrates by neutron diffraction. J. Chem. Phys., 117 (1), 308-320. [30] Chen, G.-J. Sun, C.-Y., Guo, T.-M. (2003). Modelling of the formation conditions of structure-H hydrates. Fluid Phase Equilibria, 204 (1), 107-117. [31] Chen, A.B. Zhang, W.P. Lan, X.J. Zheng, H. Liu, M.X. Han, X.W. Bao, X.H. (2006). In situ high-pressure NMR observations on the formation and dissociation of methane hydrate. Chinese Chemical Letters, 17 (6), 791-794. [32] Chou, I.-M. Sharma, A. Burrus, R.C. Shu, J. Mao, H. Hemley, R.J. Goncharov, A.F. Stern, L.A. Kirby, S.H. (2000). Transformation in Methane Hydrate. PNAS, 97 (25), 13484-13487. [33] Circone, S. Stern, L.A. & Kirby, S.H. (2004a). The effect of elevated methane pressure on methane hydrate dissociation. American Mineralogist, 89, 1192–1201. [34] Circone, S. Kirby, S.H. & Stern, L.A. (2005). Direct measurement of methane hydrate composition along the hydrate equilibrium boundary. J. Phys. Chem. B, 109, 94689475. [35] Clarke, M.A. & Bishnoi, P.R. (2004). Determination of the intrinsic rate constant and activation energy of CO2 gas hydrate decomposition using in-situ particle size analysis. Chem. Eng. Sci., 59 (14), 2983-2993. [36] Clarke, M.A. & Bishnoi, P.R. (2005). Determination of the intrinsic kinetics of CO2 gas hydrate formation using in situ particle size analysis. Chem. Eng. Sci., 60 (3), 695709. [37] Collett, T.S. (2005). Results at Mallik highlight progress in gas hydrate energy resource research and development. Petrophysics, 46 (3), 237-243. [38] Collins, M.J. Ratcliffe, C.I. Ripmeester, J.A. (1990). Nuclear magnetic resonance studies of guest species in clathrate hydrates: line-shape anisotropies, chemical shifts, and the determination of cage occupancy ratios and hydration numbers. J. Phys. Chem., 94 (1), 157-162. [39] D’Aquino, R. (2003). A wake-up call in water management. CEP magazine 99 (3), 10-17. [40] Dallimore, S.R. Collett, T.S. (Eds.). (2005) Scientific Results from the Mallik 2002 Gas Hydrate Production Research Well Program, Mackenzie Delta, Northwest Territories, Canada, Ottawa, Canada, Geological Survey of Canada Bulletin 585, including CD. [41] Dalmazzone, D. Hamed, N. Dalmazzone, C. & Rousseau, S. (2006). Application of high pressure DSC to the kinetics of formation of methane hydrates in water-in-oil emulsion. J. Thermal Analysis & Calorimetry, 85 (2), 361-368.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 43 [42] Davidson, D.W. Gas Hydrates. In: Frank, F. Eds. Water: A Comprehensive Treatise. New York, Plenum Press, 1973, Vol. 2, Chapter 3, 115. [43] Davy, H. (1811). The Bakerian Lecture. On Some of the Combinations of Oxymuriatic Gas and Oxygene, and on the chemical Relations of these Principle, to inflammable Bodies. Phil. Trans. Roy. Soc. London, 101, Part I, 1-35. [44] de Deugd, R.M. Jager, M.D. & de Swaan Arons, J. (2001). Mixed Hydrates of Methane and Water-Soluble Hydrocarbons Modeling of Empirical Results. AIChE J., 47(3), 693-704. [45] Deaton, W.M. Frost, E.M. Jr. (1946). Gas hydrates and their relation to the operation of natural-gas pipe lines. U.S. Bureau of Mines Monograph, 8, 101. [46] Delsemme, A.H., Swings, P. (1952) Gas hydrates in cometary nuclei and interstellar grains, Ann. Astrophys, 15, 1. [47] Delsemme, A.H., Wenger, A. (1970) Physico-chemical phenomena in comets-I: Experimental study of snows in a cometary environment, Planet Space. Sci., 18, 709-715. [48] Dholabhai, P.D. Kalogerakis, N. & Bishnoi, P.R. (1993). Kinetics of methane hydrate formation in aqueous-electrolyte solutions. Can. J. Chem. Eng., 71 (1), 68-74. [49] Dholabhai, P.D. Kalogerakis, N. & Bishnoi PR. (1992). Inhibition of gas hydrate deposition in offshore pipelines: effect of methanol. Preprint of Paper # CIM 92-83, Annual Meeting of The Petroleum Society of CIM, Calgary, Alberta, June 7-10. [50] Dickens, G. R. Paull, C. K. Wallace, P. & the ODP Leg 164 Scientific Party. (1997). Direct measurement of in situ methane quantities in a large gas-hydrate reservoir. Nature, 385, 426. [51] Dickens, G. R. O’Neil, J. R. Rea, D. K. & Owen, R. M. (1995). Dissociation of oceanic methane hydrate as a cause for the carbon isotope excursion at the end of the Paleocene. Paleoceanography, 10, 965. [52] Englezos, P. Kalogerakis, N. Dholabhai, P.D. & Bishnoi, P.R. (1987a). Kinetics of formation of methane and ethane gas hydrates. Chem. Eng. Sci., 42 (11), 2647-2658. [53] Englezos, P. Kalogerakis, N. Dholabhai, P.D. & Bishnoi, P.R. (1987b). Kinetics of gas hydrate formation from mixtures of methane and ethane gas. Chem. Eng. Sci,. 42 (11), 2659-2666. [54] Englezos, P. Huang, Z. & Bishnoi, P.R. (1991). Prediction of Natural Gas Hydrate Formation Conditions in the Presence of Methanol Using the Trebble-Bishnoi Equation of State. The Journal of Canadian Petroleum Technology, 30, 148-155. [55] Englezos, P. (1993). Clathrate Hydrates, Ind. Eng. Chem. Res., 32, 1251-1274. [56] Englezos, P. (1996). Nucleation and growth of gas hydrate crystals in relation to ''kinetic inhibition''. Revue Del Institut Francais Du Petrole, 51 (6), 789-795. [57] Englezos, P. & Lee, J. D. (2005). Gas Hydrates: A Cleaner Source of Energy and Opportunity for Innovative Technologies. Korean J. Chem. Eng., 22, 671-681. [58] English, N.J. Johnson, J.K. & Taylor, C.E. (2005). Molecular-dynamics simulations of methane hydrate dissociation. J. Chem. Phys., 123 (24), Art. No. 244503. [59] Florusse, L. J. Peters, C. J. Schoonman, J. Hester, K. C. Koh, C. A. Dec, S. F. Marsh, K.N. & Sloan, E.D. (2004). Stable Low-Pressure Hydrogen Clusters Stored in a Binary Clathrate Hydrate. Science, 306, 469. [60] Fu, B. Neff, S. Mathur, A. & Bakeev, K. (2002). Application of low-dosage hydrate inhibitors in deepwater operations. SPE Production & Facilities, 17(3), 133-137.

44

Peter Englezos, John Ripmeester and Robin Susilo

[61] Fukumoto, K. Tobe, J. Ohmura, R. & Mori, (2002). Y.H. Hydrate formation using water spraying in a hydrophobic gas: A preliminary study. AIChE J., 47, 1899-1904. [62] Gao, S. House, W. & Chapman, W.G. (2005). NMR/MRI Study of Clathrate Hydrate Mechanisms. J. Phys. Chem. B, 109, 19090-19093. [63] Gayet, P. Dicharry, C. Marion, G. Graciaa, A. Lachaise, J. & Nesterov, A. (2005). Experimental determination of methane hydrate dissociation curve up to 55 MPa by using a small amount of surfactant as hydrate promoter. Chem. Eng. Sci., 60 (21), 5751-5758. [64] Genov, G. Kuhs, W.F. Staykova, D.K. Goreshnik, E. & Salamatin, A.N. (2004). Experimental studies on the formation of porous gas hydrates. American Mineralogist, 89, 1228-1239. [65] Glasby, G.P. (2003). Potential impact on climate of the exploitation of methane hydrate deposits offshore. Marine &Petroleum Geology 20 (2), 163-175. [66] Gudmundsson, J.S., Parlaktuna, M., Khokhar, A.A. (1994). Storing Natural Gas as Frozen-Hydrate. SPE Prod. Facilities 9 (1), 69-73. [67] Gudmundsson, J.S., Andersson, V., Levik, O.I., Mork, M. (2000). Hydrate technology for capturing stranded gas. Ann. N.Y. Acad. Sci., 912, 403-410. [68] Hammerschmidt, E.G. (1934). Formation of Gas Hydrates in Natural Gas Transmission Lines. Ind. Eng. Chem., 26, 851-855. [69] Hand, K.P. Chyba, C.F. Carlson, R.W. & Cooper, J.F. (2006). Clathrate Hydrates of Oxidants in the Ice Shell of Europa. Astrobiology, 6(3), 463-482. [70] Handa, Y.P. (1985). Heat capacities in the range 95 to 260 K and enthalpies of fusion for structure-II clathrate hydrates of some cyclic ethers. J. Chem. Thermodynamics, 17 (3), 201-208. [71] Handa, Y.P. (1986a). Compositions, enthalpies of dissociation, and heat capacities in the range 85 to 270 K for clathrate hydrates of methane, ethane, and propane, and enthalpy of dissociation of isobutane hydrate, as determined by a heat-flow calorimeter. J. Chem. Thermodynamics, 18 (10), 915-921. [72] Handa, Y.P. (1986b). Calorimetric determinations of the compositions, enthalpies of dissociation, and heat capacities in the range 85 to 270 K for clathrate hydrates of xenon and krypton. J. Chem. Thermodynamics, 18 (9), 891-902. [73] Hara, T. Hashimoto, S. Sugahara, T. & Ohgaki, K., (2005). Large pressure depression of methane hydrate by adding 1,1-dimethylcyclohexane. Chem. Eng. Sci., 60, 31173119. [74] Hatzikiriakos, S. G. & Englezos, P. (1993). The relationship between global warming and methane gas hydrates in the earth. Chem. Eng. Sci., 48 (23), 3963-3969. [75] Hendriks, C.B. K. & Turkenburg, W. (1990). Climate and Energy. in, (Dordrec, The Netherlands), Kluwer Academic Press, 125-188. [76] Hester, K.C. & Sloan, E.D. (2005). sII Structural Transitions from Binary Mixtures of Simple sI Formers. Int. J. Thermophysics, 26 (1), 95-106. [77] Hester, K.C. Strobel, T.A. Sloan, E.D. Koh, C.A. Huq, A. & Schultz, A.J. (2006). Molecular hydrogen occupancy in binary THF-H-2 clathrate hydrates by high resolution neutron diffraction. J. Phys. Chem., B 110 (29), 14024-14027. [78] Hong, H. Pooladi-Darvish, M. & Bishnoi, P.R. (2003). Analytical modelling of gas production from hydrates in porous media. J. Canadian petroleum technology, 42 (11), 45-56.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 45 [79]

[80] [81] [82] [83] [84] [85]

[86] [87]

[88]

[89] [90]

[91]

[92]

[93]

[94]

[95] [96]

Hong, H. & Pooladi-Darvish, M. (2005). Simulation of depressurization for gas production from gas hydrate reservoirs. J. Canadian petroleum technology, 44 (11), 39-46. Hong, D.N. Gruy, F. & Herri, J.-M. (2006). Experimental data and approximate estimation for dissociation time of hydrate plugs. Chem. Eng. Sci., 61 (6), 1846-1853. House, K.Z. Schrag, D.P. Harvey, C.F. & Lackner, K.S. (2006). Permanent carbon dioxide storage in deep-sea sediments. PNAS, 103 (33), 12291-12295. Hu, Y.H. & Ruckenstein, E. (2006). Clathrate Hydrogen Hydrate - A Promising Material for Hydrogen Storage. Angew. Chem. Int. Ed., 45 (13), 2011-2013. Huo, Z., Lamar, M., Sannigrahi, B., Knauss, D.M., Sloan, E.D. (2001). Hydrate plug prevention by anti-agglomeration. Chem Eng Sci., 56, 4979-4991. Hutz, U. & Englezos, P. (1996). Measurement of structure H hydrate phase Equilibrium and the effect of electrolytes. Fluid Phase Equilibria, 117 (1-2), 178-185. Hwang, M.J. Wright, D.A. Kapur, A. & Holder, G.D. (1990). An experimental study of crystallization and crystal growth of methane hydrates from melting ice. J. Incl. Phenom.Mol. Recog. Chem., 8, 103-116. Inaba, H. (2000). New challenge in advanced thermal energy transportation using functionally thermal fluids. Int. J. Therm. Sci., 39 (9-11), 991-1003. Inerbaev, T.M. Belosludov, V.R. Belosludov, R.V. Sluiter, M. & Kawazoe, Y. (2006). Dynamics and equation of state of hydrogen clathrate hydrate as a function of cage occupation. Computational Materials Science, 36 (1-2), 229-233. Ito, Y. Kamakura, R. Obi, S. & Mori, Y.H. (2003). Microscopic observations of clathrate-hydrate films formed at liquid/liquid interfaces. II. Film thickness in steadywater flow. Chem. Eng. Sci., 58 (1), 107-114. Ivanhoe, L.F. (1993). Global oil, gas fields, sizes tallied, analyzed. Oil and Gas Journal, 91 (7), 87-91. Jager, M.D. Ballard, A.L. & Sloan, E.D. Jr. (2003). The next generation of hydrate prediction - II. Dedicated aqueous phase fugacity model for hydrate prediction. Fluid Phase Equilibria, 211, 85-107. Jager, M.D. Ballard, A.L. & Sloan, E.D. Jr. (2005). Comparison between experimental data and aqueous-phase fugacity model for hydrate prediction. Fluid phase equilibria, 232 (1-2), 25-36. Jamalludin, A.K.M. Kalogerakis, N. & Bishnoi, P.R. (1989). Modelling decomposition of a synthetic core of methane gas hydrate by coupling intrinsic kinetics with heat transfer rates. Can. J. Chem. Eng., 67, 948-955. Javanmardi, J. & Moshfeghian, M. (2003). Energy consumption and economic evaluation of water desalination by hydrate phenomenon. App. Therm. Eng., 23 (7), 845-857. Javanmardi, J. Nasrifar, K. Najibi, S.H. & Moshfeghian, M. (2005). Economic evaluation of natural gas hydrate as an alternative for natural gas transportation. App. Therm. Eng., 25 (11-12), 1708-1723. Jhaveri, J. & Robinson, D.B. (1965). Hydrates in methane-nitrogen system. Can. J. Chem. Eng., 43 (2), 75. Ji, C. Ahmadi, G. & Smith, D.H. (2001). Natural gas production from hydrate decomposition by depressurization. Chem. Eng. Sci., 56 (20), 5801-5814.

46 [97]

[98]

[99]

[100]

[101]

[102] [103] [104] [105] [106]

[107] [108]

[109]

[110] [111] [112] [113]

[114]

Peter Englezos, John Ripmeester and Robin Susilo Kang, S.P. & Lee, H. (2000). Recovery of CO2 from flue gas using gas hydrate: Thermodynamic verification through phase equilibrium measurements. Environ. Sci. Technol., 34(20), 4397-4400. Kang, S.-P. & Lee, H. (2001). Enthalpies of dissociation of clathrate hydrates of carbon dioxide, nitrogen, (carbon dioxide + nitrogen), and (carbon dioxide + nitrogen + tetrahydrofuran). J. Chem. Thermodynamics, 33 (5), 513-521. Katz, M. E. Pak, D. K. Dickens, G. R. & Miller, K. G. (1999). The source and fate of massive carbon input during the latest paleocene thermal maximum. Science, 286, 1531. Kawamura, T. Ohga, K. Higuchi, K. Yoon, J.H. Yamamoto, Y. Komai, T. & Haneda, H. (2003). Dissociation Behavior of Pellet-Shaped Methane-Ethane Mixed Gas Hydrate Samples. Energy and Fuels, 17, 614-618. Kawamura, T. Sakamoto, Y. Ohtake, M. Yamamoto, Y. Komai, T. Haneda, H. Yoon, J.-H. & Ohga, K. (2006). Dissociation behavior of pellet shaped mixed gas hydrate samples that contain propane as a guest. Energy Convers. Manage., 47 (15-16), 24912498. Kelkar, S.K. Selim, M.S. & Sloan, E.D. (1998). Hydrate dissociation rates in pipelines. Fluid Phase Equilibria, 150, 371-382. Kelland, M.A. (2006). History of the Development of Low Dosage Hydrate Inhibitors. Energy and Fuels, 20(3), 825-847. Kerr, R. (1997). Did a blast of sea-floor gas hydrates usher in a new age? Nature, 275,1267. Khokkar, A.A. Gudmundsson, J.S. & Sloan, E.D. (1998). Gas storage in structure H hydrates. Fluid Phase Equilibria, 150-151, 383-392. Kikkinides, E.S. Yang, R.T. & Cho, S.H. (1993). Concentration and recovery of carbon dioxide from flue gas by pressure swing adsorption. Industrial & Engineering Chemistry Research, 32 (11), 2714-2720. Kim, H.C. Bishnoi, P.R. Heidemann, R.A. & Rizvi, S.S.H. (1987). Kinetics of methane hydrate decomposition. Chem. Eng. Sci., 42, 1645-1653. Kim, D.-Y. Lee, J.W. Seo, Y.-T. Ripmeester, J.A. & Lee. H. (2005). Structural Transition and Tuning of tert-Butylamine Hydrate. Angew. Chem. Int. Ed., 44 (47), 7749-7752. Kim, D.-Y. Park, J. Lee, J.-W. Ripmeester, J.A. & Lee, H. (2006). Critical guest concentration and complete tuning pattern appearing in the binary clathrate hydrates. J. Am. Chem. Soc., 128 (48), 15360-15361. Kini, R.A. Dec, S.F. & Sloan, E.D. Jr. (2004). Methane + Propane Structure II Hydrate Formation Kinetics. J. Phys. Chem. A, 108, 9550-9556. Kirchner, M.T. Boese, R. Billups, W.E. & Norman, L.R. (2004). Gas Hydrate SingleCrystal Structure Analyses. J. Am. Chem. Soc., 126, 9407-9412. Klapproth, A. Goreshnik, E. Staykova, D. Klein, H. & Kuhs, W.F. (2003). Structural studies of gas hydrates. Can. J. Phys., 81, 53-518. Klara, S.M. & Srivastava, R.D. (2002). U.S. DOE Integrated Collaborative Technology Development Program for CO2 Separation and Capture. Environmental Progress, 21(4), 247-253. Klauda, J., B. & Sandler, S.I. (2005). Global distribution of methane hydrate in ocean sediment. Energy & Fuels, 19, 459-470.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 47 [115] Knight, C.A. &Rider, K. (2002). Free-growth forms of tetrahydrofuran clathrate hydrate crystals from the melt: plates and needles from a fast-growing vicinal cubic crystal. Philosophical Magazine 82 (8), 1609-1632. [116] Kobayashi, I. Ito, Y. & Mori, Y.H. (2001). Microscopic observations of clathratehydrate films formed at liquid/liquid interfaces. I. Morphology of hydrate films. Chem. Eng. Sci., 56 (14), 4331-4338. [117] Koh, C.A. (2002). Towards a fundamental understanding of natural gas hydrates. Chem. Soc. Rev., 31 (3), 157-167. [118] Kuhs, W.F. Chazallon, B. Radaelli, P.G. & Pauer, F. (1997). Cage Occupancy and Compressibility of Deuterated N2-Clathrate Hydrate by Neutron Diffraction. J. Incl. Phen. Microcyclic Chem., 29, 65-77. [119] Kuhs, W.F. Genov, G. Staykova, D.K. & Hansen. T. (2004). Ice perfection and onset of anomalous preservation of gas hydrates. Phys. Chem. Chem. Phys. 6 (21), 49174920. [120] Kuhs, W.F. Staykova, D.K. & Salamantin, A.N. (2006). Formation of methane hydrate from polydisperse ice powders. J. Phys. Chem. B, 110 (26), 13283-13295. [121] Kumar, R. Wu, H-J. & Englezos, P. (2006). Incipient Hydrate Phase Equilibrium for Gas mixtures Containing Hydrogen, Carbon Dioxide & Propane. Fluid Phase Equilibria, 244, 167-171. [122] Kumazaki, T. Kito, Y. Sasaki, S. Kume, T. & Shimizu, H. (2004). Single-crystal growth of the high-pressure phase II of methane hydrate and its Raman scattering study. Chem. Phys. Letter, 388 (1-3), 18-22 [123] Kvamme, B. Graue, A. Aspenes, E. Kuznetsova, T. Granasy, L. Toth, G. Pusztai, T. & Tegze, G. (2004). Kinetics of solid hydrate formation by carbon dioxide: Phase field theory of hydrate nucleation and magnetic resonance imaging. Phys. Chem. Chem. Phys., 6, 2327-2334. [124] Kvenvolden, K.A. (1988). Methane hydrate – a major reservoir of carbon in the shallow geosphere. Chemical Geology, 71 (1-3), 41-51. [125] Kvenvolden, K. A. (1993). Gas hydrates-geological perspective and global change. Reviews in Geophysics, 31(2), 173. [126] Kvenvolden, K. A. (1999). Potential Effects of Gas Hydrate on Human Welfare. Proceedings of National Academy of Sciences U.S.A., 96(7), 3420. [127] Kvenvolden, K. A. (2000). Gas Hydrate and Humans. Annals of the New York Academy of Sciences, 912, 17. [128] Kvenvolden, K.A. (2002). Methane hydrate in the global organic carbon cycle. Terra Nova, 14 (5), 302-306. [129] Larsen, R., Knight, C.A., Sloan, E.D., Jr. (1998). Clathrate hydrate growth and inhibition. Fluid Phase Equilibria, 150, 353-360. [130] Le Parlouer, P. Dalmazzone, C. Herzhaft, B. Rousseau, L. & Mathonat, C. (2004). Characterisation of gas hydrates formation using a new high pressure MICRO-DSC. J. Thermal Analysis & Calorimetry, 78, 165–172. [131] Lee, H. Seo, Y. Seo, Y-T. Moudrakovski, I.L. & Ripmeester, J. A. (2003) Recovering Methane from Solid Methane Hydrate with Carbon Dioxide. Angew. Chem. Int. Ed., 42, 5048-5051.

48

Peter Englezos, John Ripmeester and Robin Susilo

[132] Lee, H. Lee, J. Kim, D.Y. Park, J. Seo, Y.-T. Zeng, H. Moudrakovski, I.L. Ratcliffe, C.I. & Ripmeester, J.A. (2005a). Tuning clathrate hydrates for hydrogen storage. Nature, 434, 743-746. [133] Lee, J.-D. Susilo, R. & Englezos, P. (2005b). Kinetics of Structure H Gas Hydrate. Energy and Fuels, 19, 1008-1015. [134] Lee, J.D. Susilo, R. Englezos, P. (2005c). Methane-ethane and methane-propane hydrate formation and decomposition on water droplets. Chem. Eng. Sci., 60, 42034212. [135] Lee, J.D. & Englezos, P. (2005) Enhancement of the Performance of Gas Hydrate Kinetic Inhibitors with Polyethylene Oxide. Chem Eng Sci, 60, 5323-5330. [136] Lee, J.D. & Englezos, P. (2006). Unusual kinetic inhibitor effects on gas hydrate formation. Chem. Eng. Sci. 61 (5), 1368-1376. [137] Lee, J. Lu, H. Moudrakovski, I.L. Ratcliffe, C.I. & Ripmeester, J.A. (2006a). npentane and n-hexane as coguests in structure-H hydrates in mixtures of 2,2dimethylbutane and methane. Angew. Chem. Int. Ed., 45 (15), 2456-2459. [138] Lee, J.D. Song, M. Susilo, R. & Englezos, P. (2006b). Dynamics of Methane-Propane Clathrate Hydrate Crystal Growth from Liquid Water with or without the Presence of n-Heptane. Crystal Growth & Design, 6 (6), 1428-1439. [139] Lekvam, K. & Ruoff, P. (1993). A reaction kinetic mechanism for methane hydrate formation in liquid water. J. Am. Chem. Soc., 115 (19), 8565-8569. [140] Lekvam, K. & Bishnoi, P.R. (1997). Dissolution of methane in water at low temperatures and intermediate pressures. Fluid Phase Equilibria, 131, 297-309. [141] Levenspiel, O. (1999). Chemical Reaction Engineering, 3rd ed., New York, NY: Wiley & Sons. [142] Li, J. Liang, D. Guo, K. & Wang, R. (2005). The influence of additives and metal rods on the nucleation and growth of gas hydrates. J. Coll. & Int. Sci., 283 (1), 223230. [143] Li, X.-S. & Englezos, P. (2006). Prediction of Gas Hydrate Formation Conditions in the Presence of Methanol, Glycerol, Ethylene glycol and Triethylene glycol with the SAFT Equation of State. Ind Eng Chem Res., 45 (6), 2231-2137. [144] Li, X.-S. Wu, H.-J. Li, L.-G. Feng, Z.-P. Tang, L.-G. & Fan, S.-H. (2007). Hydrate dissociation conditions for gas mixtures containing carbon dioxide, hydrogen, hydrogen sulfide, nitrogen, and hydrocarbons using SAFT. J Chem Thermodynamics. 39, 419-425. [145] Lin, W., Chen, G.-J., Sun, C.-Y., Guo, X.-Q., Wu, Z.-K., Liang, M.-Y., Chen, L.-T., Yang, L.-Y., Effect of surfactant on the formation and dissociation kinetic behavior of methane hydrate, Chem. Eng. Sci. 59: 4449-4455 (2004). [146] Linga, P. Kumar, R. & Englezos, P. (2006). Capture of Carbon Dioxide from Conventional Power Plants or from Integrated Gasification Plants through Gas Hydrate Formation/Dissociation. Journal of Energy & Climate Change, 1 (2), 75-82. [147] Linga, P. Kumar, R. & Englezos P. (2007a). Gas Hydrate formation from Hydrogen/carbon dioxide and nitrogen/carbon dioxide gas mixtures. accepted for publication in Chem. Eng. Sci., April 26, 2007 [148] Linga, P. Kumar, R. & Englezos P. (2007b). The clathrate hydrate process for post and pre combustion capture of carbon dioxide. J Hazardous Materials. revised submitted May 10, 2007

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 49 [149] Lingelem, M. N. Majeed, A. I. & Stange, E. (1994). Industrial experience in evaluation of hydrate formation, inhibition, and dissociation in pipeline design and operation. Ann. N.Y. Acad. Sci., 715, 75. [150] Link, D. D. Ladner, E. P. Elsen, H. A. & Taylor, C. E. (2003). Formation and dissociation studies for optimizing the uptake of methane by methane hydrates. Fluid Phase Equilibria, 211, 1–10. [151] Lokshin, K.A. Zhao, Y. He, D. Mao, W.L. Mao, H.-K. Hemley, R.J. Lobanov, M.V. & Greenblatt, M. (2004). Structure and dynamics of hydrogen molecules in the novel clathrate hydrate by high pressure neutron diffraction. Phys. Rev. Lett., 93, Art. No. 125503. [152] Lokshin, K.A. & Zhao, Y. (2006). Fast synthesis method and phase diagram of hydrogen clathrate hydrate. App. Phys. Lett., 88, 131909. [153] Loveday, J.S. Nelmes, R.J. Guthrie, M. Belmote, S.A. Allan, D.R. Klug, D.D. Tse, J.S. & Handa, Y.P. (2001). Stable methane hydrate above 2 GPa and the source of Titan's atmospheric methane. Nature, 410 (6829), 661-663. [154] Loveday, J.S. Nelmes, R.J. Klug, D.D. Tse, J.S. & Desgreniers, S. (2003a). Structural systematics in the clathrate hydrates under pressure. Can. J. Phys., 81 (1-2), 539-544. [155] Lovell, D. & Pakulski, M. (2003). Two low-dosage hydrate inhibitors. Journal of Petroleum Technology, 55(4), 65-68. [156] Lu, H. Seo, Y-T. Lee, J.-W. Moudrakovski, I. Ripmeester, J. A. Chapman, N. R. Coffin, R.B. Gardner, G. & Pohlman, (2007). J. Complex Gas Hydrate from the Cascadia Margin. Nature, 445, 303-306. [157] Lunine, J.I., Stephenson, D.J. (1985). Thermodynamics of Clathrate Hydrate at Low and High Pressures with Application to the Outer Solar System. Astrophys. J. Suppl. Ser. 58, 493-531. [158] Ma, C.F. Chen, G.J. & Guo, T.M. (2002). Kinetics of hydrate formation using gas bubble suspended in water. Science in China - Series B - Chem., 45 (2), 208-215. [159] Ma, Q.L. Chen, G.J. Guo, T.M. Zhang, K. Zuo, J.Y. Zhang, D. Ng, H.J. (2005). Prediction of structure-H gas hydrate formation conditions for reservoir fluids. Chinese Journal of Chemical Engineering, 13 (4), 484-490. [160] Machida, S. Hirai, H. Kawamura, T. Yamamoto, Y. & Yagi, T. (2006). A new highpressure structure of methane hydrate surviving to 86 GPa and its implications for the interiors of giant icy planets. Physics of the Earth and Planetary Interiors, 155 (1-2), 170-176. [161] Mak, T.C.W. & McMullan, R.K. (1965). Polyhedral Clathrate Hydrates. X. Structure of the Double Hydrate of Tetrahydrofuran and Hydrogen Sulfide. J. Chem. Phys., 43 (8), 2732-2737. [162] Makino, T. Nakamura, T. Sugahara, T. Ohgaki, K. (2004). Thermodynamic stability of structure-H hydrates of methylcyclopentane and cyclooctane helped by methane. Fluid Phase Equilibria, 218, 235–238. [163] Makogon, Y. F. Trebin, F. A. Trofimuk, A. A. Tsarev, V. P. & Chersky, N.V. (1972). Detection of a Pool of Natural Gas in a Solid (Hydrate Gas) State. Doklady of the Academy of Sciences of the USSR-Earth Science Sections, 196, 197. [164] Makogon, Y. (1987). Gaseous hydrate – the frozen energy. Recherche, 18 (192), 1192-1200.

50

Peter Englezos, John Ripmeester and Robin Susilo

[165] Makogon, T.Y. & Sloan, E.D. Jr. (1994). Phase Equilibrium for Methane Hydrate from 190 to 262 K. J. Chem. Eng. Data, 39 (2), 351-353. [166] Makogon, T.Y. Mehta, A.P. & Sloan, E.D. Jr. (1996). Structure H and Structure I Hydrate Equilibrium Data for 2,2-Dimethylbutane with Methane and Xenon. J. Chem. Eng. Data, 41, 315-318. [167] Manakov, A.Y. Voronin, V.I. Kurnosov, A.V. Teplykh, A.E. Komarov, A.Y. & Dyadin, A.Y. (2004). Structural Investigations of Argon Hydrates at Pressures up to 10 kbar. J. Inclusion Phenom. Mol. Recognit. Chem., 48 (1-2), 11-18. [168] Mao, W.L. Mao, H. Goncharov, A.F. Struzhkin, V.V. Guo, Q. Hu, J. Shu, J. Hemley, R.J. Somayazulu, M. & Zhao, Y. (2002). Hydrogen clusters in clathrate hydrate. Science, 297 (5590), 2247-2249. [169] Marshall, C.B. Fletcher, G.L. & Davies, P.L. (2004). Hyperactive antifreeze protein in a fish. Nature, 429, 153. [170] Matsuda, S. Tsuda, H. & Mori, Y.H. (2006). Hydrate Formation Using Water Spraying onto a Cooled Solid Surface in a Guest Gas. AIChE J., 52 (8), 2978-2987. [171] Max, M.D. Johnson, A. H. Dillon, W. P. (2006). Economic Geology of Natural Gas Hydrates. The Netherlands: Springer-Verlag. [172] McMullan, R.K. & Jeffrey, G.A. (1965). Polyhedral Clathrate Hydrates. IX. Structure of Ethylene Oxide Hydrate. J. Chem. Phys., 43 (8), 2725-2732. [173] Mehta, A.P. & Sloan, E.D. Jr. (1993). Structure H Hydrate Phase Equilibria of Methane + Liquid Hydrocarbon Mixtures. J. Chem. Eng. Data, 38, 580-582. [174] Miller, B. & Strong, E.R. (1946). Am. Gas Assn. Monthly, 29(2), 63. [175] Mochizuki, T. (2003). Clathrate hydrate formation at liquid/liquid interface under shear water flow. J. Crystal Growth, 249, 372–380. [176] Mochizuki, T. & Mori, Y.H. (2006). Clathrate-hydrate film growth along water/hydrate-former phase boundaries - numerical heat-transfer study. J. Crystal Growth, 290 (2), 642-652. [177] Mooijer van Heuvel, M.M. Peters. C.J. & de Swaan Arons, J. (2000). Influence of water-insoluble organic components on the gas hydrate equilibrium conditions of methane. Fluid Phase Equilibria, 172, 73-91. [178] Mori, T. & Mori, Y.H. (1989a). Characterization of gas hydrate formation in directcontact cool storage process. Int. J. Refrigeration, 12 (5), 259-265. [179] Mori, T. & Mori, Y.H. (1989b). Formation of gas hydrate with CFC alternative R134a, AIChE J., 35 (7), 1227-1228. [180] Mori, Y.H. (2001). Estimating the thickness of hydrate films from their lateral growth rates: application of a simplified heat transfer model. J. Crystal Growth, 223, 206212. [181] Mori, Y.H. (2003). Recent Advances in Hydrate-based Technologies for Natural gas Storage-A Review. J. Chem. Ind. Eng. (China), 54, Suppl., 1−17. [182] Moridis, G. J. (2003). Numerical studies of gas production from methane hydrates. paper SPE 75691, SPE Journal, 8(4), 359. [183] Moridis, G. J. (2004). Numerical studies of gas production from class 2 and class 3 hydrate accumulations at the Mallik site, Mackenzie Delta, Canada. SPE Reservoir Evaluation & Engineering, 7(3), 175.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 51 [184] Moridis, G.J. Kowalsky, M. B. & Pruess, K. (2005). Depressurization-Induced Gas Production from Class 1 Hydrate Deposits. SPE 97266, SPE Annual Technical Confrence, Dallas, TX, 9-12 October. [185] Moudrakovski I.L. Ratcliffe, C.I. McLaurin, G.E. Simard, B. & Ripmeester, J.A. (1999). Hydrate Layers on Ice Particles and Superheated Ice: a 1H NMR Microimaging Study. J. Phys. Chem. A, 103 (26), 4969-4972. [186] Moudrakovski, I.L. Sanchez, A.A. Ratcliffe, C.I. & Ripmeester, J.A. (2001). Nucleation and Growth of Hydrates on Ice Surfaces: New Insights from 129Xe NMR Experiments with Hyperpolarized Xenon. J. Phys. Chem. B, 105, 12338-12347. [187] Moudrakovski, I.L. McLaurin, G.E. Ratcliffe, C.I. & Ripmeester, J.A. (2004). Methane and Carbon Dioxide Hydrate Formation in Water Droplets: Spatially Resolved Measurements from Magnetic Resonance Microimaging. J. Phys. Chem. B, 108, 17591-17595. [188] Müller, H.R. & Stackelberg, M.V. (1952). Zur struktur der gashydrate. Naturwiss, 39, 20–21. [189] Nada, H. (2006). Growth mechanism of a gas clathrate hydrate from a dilute aqueous gas solution: A molecular dynamics simulation of a three-phase system. J. Phys. Chem. B, 110 (33), 16526-16534. [190] Nakamura, T. Makino, T. Sugahara, T. & Ohgaki, K. (2003). Stability boundaries of gas hydrates helped by methane—structure-H hydrates of methylcyclohexane and cis1,2-dimethylcyclohexane. Chem. Eng. Sci., 58, 269-273. [191] Ng, H.-J. & Robinson, D.B. (1985). Hydrate Formation in Systems Containing Methane, Ethane, Propane, Carbon Dioxide or Hydrogen Sulfide in the Presence of Methanol. Fluid Phase Equilibria, 21, 145-155. [192] Ngan, Y. T. & Englezos, P. (1996). Concentration of Mechanical Pulp Mill Effluents and NaCl Solutions through Propane Hydrate Formation. Ind. Eng. Chem. Res., 35(6), 1894. [193] Nisbet, E. G. (1992). Sources of atmospheric CH4 in early postglacial time. Journal of Geophysical Research, 97(D12), 12859. [194] Ogienko, A.G. Kursonov, A.V. Manakov, A.Y. Larionov, E.G. Ancharov, A.I. Sheromov, M.A. & Nesterov, A.N. (2006). Gas hydrates of argon and methane synthesized at high pressures: Composition, thermal expansion, and self-preservation. J. Phys. Chem. B, 110 (6), 2840-2846. [195] Ohmura, R. Kashiwazaki, S. Shiota, S. Tsuji, H. & Mori, Y.H. (2002). Structure-I and Structure-H Hydrate Formation Using Water Spraying. Energy & Fuels, 16, 11411147. [196] Ohmura, R. Uchida, T. Takeya, S. Nagao, J. Minagawa, H. Ebinuma, T. & Narita, H. (2003a). Clathrate hydrate formation in (methane + water + methylcyclohexanone) systems: the first phase equilibrium data. J. Chem. Thermodynamics, 35, 2045-2054. [197] Ohmura, R. Ogawa, M. Yasuoka, K. & Mori, Y.H. (2003b). Statistical study of clathrate-hydrate nucleation in a water/hydrochlorofluorocarbon system: Search for the nature of the "memory effect". J. Phys. Chem. B, 107 (22), 5289-5293. [198] Ohmura, R. Uchida, T. Takeya, S. Nagao, J. Minagawa, H. Ebinuma, T. & Narita, H. (2003c). Phase Equilibrium for Structure-H Hydrates Formed with Methane and

52

[199]

[200]

[201]

[202]

[203]

[204] [205] [206]

[207]

[208]

[209]

[210]

[211] [212] [213] [214]

Peter Englezos, John Ripmeester and Robin Susilo either Pinacolone (3,3-Dimethyl-2-butanone) or Pinacolyl Alcohol (3,3-Dimethyl-2butanol). J. Chem. Eng. Data, 48, 1337-1340. Ohmura, R. Shimada, W. Uchida, T. Mori, Y.H. Takeya, S. Nagao, S. Minagawa, H. Ebinuma, T. & Narita, H. (2004). Clathrate hydrate crystal growth in liquid water saturated with a hydrate-forming substance: variations in crystal morphology. Philosophical Magazine, 84 (1), 1-16. Ohmura, R. Matsuda, S. Uchida, T. Ebinuma, T. & Narita, H. (2005a). Phase Equilibrium for Structure-H Hydrates at Temperatures below the Freezing Point of Water. J. Chem. Eng. Data, 50, 993-996. Ohmura, R. Matsuda, S. Uchida, T. Ebinuma, T. & Narita, H. (2005b). Formation and Growth of Structure-H Hydrate Crystals on a Water Droplet. Crystal Growth & Design, 5 (5), 1821-1824. Ohmura, R. Matsuda, S. Takeya, S. Ebinuma, T. & Narita, H. (2005c). Phase equilibrium for structure-H hydrates formed with methane and methyl-substituted cyclic ether. Int. J. Thermophysics, 26 (5), 1515-1523. Ohmura, R. Matsuda, S. Itoh, S. Ebinuma, T. & Narita, H. (2005d). Clathrate Hydrate Crystal Growth in Liquid Water Saturated with a Guest Substance: Observations in a Methane + Water System. Crystal Growth & Design, 5(3), 953-957. Oldenburg, C.M. (2003). Carbon Dioxide as Cushion Gas for natural Gas Storage. Energy & Fuels, 17, 240-246. Osegovic, J.P. & Max, M.D. (2005). Compound clathrate hydrate on Titan's surface. J. Geophysical Research 110, E08004. Østergaard, K.K. Tohidi, B. Danesh, A. Burgass, R.W. & Todd, A.C. (2001). Equilibrium data and thermodynamic modelling of isopentane and 2,2dimethylpentane hydrates. Fluid Phase Equilibria, 169, 101–115. Ota, M. Morohashi, K. Abe, Y. Watanabe, M. Smith, R.L. Jr. & Inomata, H. (2005). Replacement of CH4 in the Hydrate by Use of Liquid CO2. Energy Conversion Management, 46, 1680-1691. Ota, S. Uetani, H. & Kawano, H. (2002). Use of hydrate pellets for transportation of natural gas – III – safety measures and conceptual design of natural gas hydrate pellet carrier. Proc. Fourth Int. Con. Gas Hydrate, Yokohama, Japan, Vol. I: 991-996. Park, Y. Kim, D.Y. Lee, J.-W. Huh, D.-G. Park, K.-P. Lee, J. & Lee, H. (2006a). Sequestering carbon dioxide into complex structures of naturally occurring gas hydrates. PNAS, 103 (34), 12690-12694. Park, J. Seo, Y.T. Lee, J. & Lee, H. (2006b). Spectroscopic analysis of carbon dioxide and nitrogen mixed gas hydrates in silica gel for CO2 separation. Catalysis today, 115 (1-4), 279-282. Parker, A. (1942) Potable Water from Sea-water. Nature, 149, 184-186. Parrish, W.R. Prausnitz, J.M. (1972). Dissociation Pressures of Gas Hydrates Formed by Gas Mixtures. Ind. Eng. Chem. Pro. Des. Develop. 11, 26-34. Patchkovskii, S. & Tse, J.S. (2003). Thermodynamic stability of hydrogen clathrates. PNAS, 100, 14645-14650. Peters, D. Selim, M.S. & Sloan, E.D. Jr. (2000). Hydrate dissociation in pipelines by two-sided depressurization - Experiment and model. Annals. NY Acad. Sci., 912, 304313.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 53 [215] Pooladi-Darvish, M. (2004). Gas Production from Hydrate Reservoirs and its Modelling. Distinguished author series, Journal of Petroleum Technology (JPT), 65. [216] Purwanto, Y.A. Oshita, S. Seo Y. & Kawagoe, Y. (2001). Concentration of Liquid Foods by the Use of Gas Hydrate. J. of Food Engineering, 47, 133-138. [217] Rautenbach, R. & Pennings, P. (1973). Development and optimization of a hydrate process for desalination of seawater. Chemie Ingenieur Technik, 45 (5), 259-264. [218] Reeburgh, W.S. (2003). Global Methane Biogeochemistry, Treatise on Geochemistry. Vol 4, Editor: Ralph F. Keeling. Executive Editors: Heinrich D. Holland and Karl K. Turekian. pp. 347. ISBN 0-08-043751-6. Elsevier, 65-89. [219] Ripmeester, J.A. & Davidson, D.W. (1977). Some new clathrate hydrates. Mol. Crys. Liq. Crys., 43, 189-195. [220] Ripmeester, J.A. Tse, J.S. Ratcliffe, C.I. & Powell, B.M. (1987). A new clathrate hydrate structure. Nature, 325, 135-136. [221] Ripmeester, J.A. & Ratcliffe, C.I. (1988). Low-temperature cross-polarization/magic angle spinning carbon-13 NMR of number solid methane hydrates: structure, cage occupancy, and hydration. J. Phys. Chem., 92, 337-339. [222] Ripmeester, J.A. & Ratcliffe, C.I. (1990). Xenon-129 NMR studies of clathrate hydrates: new guests for structure II and structure H. J. Phys. Chem., 94 (25), 87738776. [223] Ripmeester, J.A. & Ratcliffe, C.I. Klug, D.D. Tse, J.S. (1994). Molecular perspectives on structure and dynamics in clathrate hydrates. Ann. N.Y. Acad. Sci., 715, 161-176. [224] Ripmeester, J.A. & Ratcliffe, C.I. (1999). On the contributions of NMR spectroscopy to clathrate science. J. Structural Chemistry, 40(5), 654-662. [225] Ripmeester, J.A. (2000). Hydrate Research – From Correlations to a Knowledgebased Discipline. Ann. N.Y. Acad. Sci., 912, 1-16. [226] Robinson, D.B. & Ng, H.-J. (1986). Hydrate Formation and Inhibition in Gas or Gas Condensate Streams. Journal of Canadian Petroleum Technology, 25, 26-30. [227] Ross, D.K. (2006). Hydrogen storage: The major technological barrier to the development of hydrogen fuel cell cars. Vacuum, 80 (10), 1084-1089. [228] Salamatin, A.N. Hondoh, T. Uchida, T. & Lipenkov, V.Y. (1998). Post-nucleation conversion of an air bubble to clathrate air-hydrate crystal in ice. J. Cryst. Growth, 193, 197-218. [229] Schicks, J.M. & Ripmeester, J.A. (2002). The Coexistence of Two Different Methane Hydrate Phases under Moderate Pressure and Temperature Conditions: Kinetic versus Thermodynamic Products. Angew. Chem. Int. Ed., 43, 3310–3313. [230] Schicks, J.M. Naumann, R. Erzinger, J. Hester, K.C. Koh, C.A. & Sloan, E.D. Jr. (2006). Phase transition in mixed gas hydrates: Experimental observations versus calculated data. J. Phys. Chem. B, 110, 11468-11474. [231] Schneider, G.R. & Farrar, J. (1968). U.S. Dept. of Interior, Rsch. Dev. Rpt No 292, p.37, January. [232] Schneider, S. H. (1990). The global warming debate: science or politics? Environ. Sci. Technol., 24, 432. [233] Schuth, F. (2005). Technology - Hydrogen and hydrates. Nature, 434 (7034), 712713. [234] Selim, M.S. & Sloan, E.D. (1989). Heat and mass transfer during dissociation of hydrate in porous media. AIChE J., 35, 1049-1052.

54

Peter Englezos, John Ripmeester and Robin Susilo

[235] Selleck, F.T. Carmichael, L.T. & Sage, B.H. (1952). Phase Behavior in the Hydrogen Sulfide-Water System. Ind. Eng. Chem., 44, 2219-2226. [236] Seo, Y.-T. Kang, S.-P. & Lee, H. (2001). Experimental determination and thermodynamic modeling of methane and nitrogen hydrates in the presence of THF, propylene oxide, 1,4-dioxane and acetone. Fluid Phase Equilibria, 189, 99–110. [237] Seo, Y.T. & Lee, H. (2003). 13C NMR Analysis and Gas Uptake Measurements of Pure and Mixed Gas Hydrates: Development of Natural Gas Transport and Storage Method using Gas Hydrate. Korean J. Chem. Eng., 20(6), 1085-1091. [238] Seo, Y.T. Moudrakovski, I.L. Ripmeester, J.A. Lee, J.-W. & Lee, H. (2005a). Efficient Recovery of CO2 from Flue Gas by Clathrate Hydrate Formation in Porous Silica Gels. Environ. Sci. Technol., 39, 2315-2319. [239] Seo, Y.-T. Kim, D.Y. Lee, H. Lee, J. Moudrakovski, I.L. & Ripmeester, J.A. (2005b). Molecular behavior and guest distribution in the cages of CH4 + THF double hydrate. Proc. Fifth Int. Con. Gas Hydrates, Vol. II, 568-572, Trondheim, Norway. [240] Servio, P. & Englezos, P. (2003a). Morphology of methane and carbon dioxide hydrates formed from water droplets. AIChE J., 49 (1), 269-276. [241] Servio, P. & Englezos, P. (2003b). Morphology study of structure H hydrate formation from water droplets. Crystal Growth & Design, 3 (1), 61-66. [242] Servio, P. Macchi, A. & Luiz, D.S. (2004). Formation of gas hydrates by fluidizedbed granulation. U.S. Pat. Appl. Publ.. [243] Shimada, W. Takeya, S. Kamata, Y. Uchida, T. Nagao, J. Ebinuma, T. & Narita, H. (2005). Texture Change of Ice on Anomalously Preserved Methane Clathrate Hydrate. J. Phys. Chem. B, 109, 5802-5807. [244] Shpakov, V.P. Tse, J.S. Tulk, C.A. Kvamme, B. & Belosludov, V.R. (1998). Elastic moduli calculation and instability in structure I methane clathrate hydrate. Chemical Physics Letter, 282, 107–114. [245] Simpson, S. (2000). Methane Fever. Scientific American, 2, 24. [246] Sloan, E.D. Jr. & Fleyfel, F. (1992). Hydrate dissociation enthalpy and guest size. Fluid Phase Equilibria, 76, 123-140. [247] Sloan, E.D. Jr. (1998). Clathrate Hydrates of Natural Gasses, 2nd edition, New York, NY: Marcel Dekker. [248] Sloan, E.D. Jr. (2003a). Fundamental principles and applications of natural gas hydrates. Nature, 426, 353-359. [249] Sloan, E.D. Jr. (2003b). Clathrate hydrate measurements: microscopic, mesoscopic, and macroscopic. J. Chem. Thermodynamics, 35, 41–53. [250] Sloan, E.D. (2004a). Introductory overview: Hydrate knowledge development. American Mineralogist, 89, 1155–1161. [251] Sloan, E.D. (2004b). Thermodynamics of Natural Gas Hydrates, in Chemical Thermodynamics for Industry, Fletcher, T.M., (editor), RSC, chapter 6. [252] Sloan, E.D. (2005). A changing hydrate paradigm—from apprehension to avoidance to risk management. Fluid Phase Equilibria, 228, 67-74. [253] Stackelberg, M.V. & Müller, H.R. (1951). Zur struktur der gashydrate. Naturwiss, 38, 456. [254] Staykova, D.K. Kuhs, W.F. Salamatin, A.N. & Hansen, T. (2003). Formation of Porous Gas Hydrates from Ice Powders: Diffraction Experiments and Multistage Model. J. Phys. Chem. B, 107 (37), 10299-10311.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 55 [255] Stern, L.A. Circone, S. Kirby, S.H. & Durham, D.B. (2003). Temperature, pressure, and compositional effects on anomalous or "self" preservation of gas hydrates. Can. J. Phys., 81 (1-2), 271-283. [256] Stern, L.A. Kirby, S.H. & Durham, W.B. (1998). Polycrystalline Methane Hydrate: Synthesis from Superheated Ice, and Low-Temperature Mechanical Properties. Energy and Fuels, 12 (2), 201-211. [257] Stern, L.A. Circone, S. Kirby, S. H. & Durham, W.B. (2001a). Anomalous Preservation of Pure Methane Hydrate at 1 atm. J. Phys. Chem. B, 105, 1756-1762. [258] Stern, L.A. Circone, S. Kirby, S.H. & Durham, W.B. (2001b). Preservation of methane hydrate at 1 Atm. Energy and Fuels, 15 (2), 499-501. [259] Stern, L.A. Kirby, S.H. Circone, S. & Durham, W.B. (2004). Scanning Electron Microscopy investigations of laboratory-grown gas clathrate hydrates. American Mineralogist, 89, 1162–1175. [260] Stevenson, D.J. (1999). Life-sustaining planets in interstellar space. Nature, 400(6939), 32. [261] Strobel, T.A. Taylor, C.J. Hester, K.C. Dec, S.F. Koh, C.A. Miller, K.T. &Sloan, E.D. Jr. (2006). Molecular hydrogen storage in binary THF-H-2 clathrate hydrates. J. Phys. Chem. B, 110(34), 17121-17125. [262] Subramanian, S. Kini, R.A. Dec, S.F. & Sloan, E.D. Jr. (2000a). Evidence of structure II hydrate formation from methane-ethane mixtures. Chem. Eng. Sci., 55, 1981-1999. [263] Subramanian, S. Ballard, A.L. Kini, R.A. Dec, S.F. & Sloan, E.D. Jr. (2000b). Structural transitions in methane-ethane gas hydrates, Part I: upper transition point and applications. Chem. Eng. Sci., 55, 5763-5771. [264] Suess, E. Bohrmann, G. Greinert, J. & Lausch, E. (1999). Flammable Ice. Scientific American, 11, 76. [265] Sultan, N. Cochonat, P. Foucher, J.-P. & Mienert, J. (2004a). Effect of gas hydrates melting on seafloor slope instability. Marine Geology, 213 (1-4), 379-401. [266] Sultan, N. Cochonat, P. Canals, M. Cattaneo, A. Dennielou, B. Haflidason, H. Laberg, J.S. Long, D. Mienert, J. Trincardi, F. Urgeles, R. Vorren, T.O. & Wilson, C. (2004b). Triggering mechanisms of slope instability processes and sediment failures on continental margins: a geotechnical approach. Marine Geology, 213 (1-4), 291-321. [267] Sum, A.K. Burrus, R.C. & Sloan, E.D. Jr. (1997). Measurement of clathrate hydrates via Raman spectroscopy. J. Phys. Chem. B, 101 (38), 7371-7377. [268] Sun, Z.-G. Fan, S.-S. Guo, K.-H. Shi, L. & Wang, R.-Z. (2002). Equilibrium hydrate formation conditions for methylcyclohexane with methane and a ternary gas mixture. Fluid Phase Equilibria, 198, 293-298. [269] Sun, Z. Wang, R. Ma, R. Guo, K. & Fan, S. (2003). Effect of surfactants and liquid hydrocarbons on gas hydrate formation rate and storage capacity. Int. J. Energy Res., 27, 747-756. [270] Sun, C.-Y. & Chen, G.-J. (2006). Methane hydrate dissociation above O degrees C and below O degrees C. Fluid Phase Equilibria, 242 (2), 123-128. [271] Susilo, R. Lee, J.D. & Englezos, P. (2005). Liquid–liquid equilibrium data of water with neohexane, methylcyclohexane, tert-butyl methyl ether, n-heptane and vapor– liquid–liquid equilibrium with methane. Fluid Phase Equilibria, 231, 20-26.

56

Peter Englezos, John Ripmeester and Robin Susilo

[272] Susilo, R. Moudrakovski, I.L. Ripmeester, J.A. & Englezos, P. (2006). Hydrate kinetics study in the presence of non-aqueous liquid by NMR spectroscopy and imaging. J. Phys. Chem. B, 110 (51), 25803-25909. [273] Susilo, R. Ripmeester, J.A. & Englezos, P. (2007a). Characterization of gas hydrates with PXRD, DSC, NMR and Raman Spectroscopy, accepted for publication in Chem. Eng. Sci.. [274] Susilo, R. Ripmeester, J.A. & Englezos, P. (2007b). Methane uptake and conversion rate into structure H hydrate, submitted for publication to AIChE J. [275] Svandal, A. Kvamme, B. Granasy, L. Pusztai, T. Buanes, T. & Hove, J. (2006). The phase-field theory applied to CO2 and CH4 hydrate. J. Crystal Growth 287, 486-490. [276] Tajima, H. Yamasaki, A. & Kiyono, F. (2004). Continuous formation of CO2 hydrate via a kenics-type static mixer. Energy and Fuels, 18 (5), 1451-1456. [277] Tajima, H. Yamasaki, A. & Kiyono, F. (2005a). Effects of mixing functions of static mixers on the formation of CO2 hydrate from the two-phase flow of liquid CO2 and water. Energy and Fuels, 19 (6), 2364-2370. [278] Tajima, H. Yamasaki, A. & Kiyono, F. (2005b). Process design of a new injection method of liquid CO2 at the intermediate depths in the ocean using a static mixer. Fuel Processing Technology, 86 (14-15), 1667-1678. [279] Tajima, H. Yamasaki, A. Kiyono, F. & Teng, H. (2006). Size distribution of CO2 drops in a static mixer for ocean disposal. AIChE J., 52 (8), 2991-2996. [280] Takahashi, M. Kawamura, T. Yamamoto, Y. Ohnari, H. Himuro, S. & Shakutsui, H. (2003). Effect of Shrinking Microbubble on Gas Hydrate Formation. J. Phys. Chem. B, 107 (10), 2171-2173. [281] Takeya, S. Shimada, W. Kamata, Y. Ebinuma, T. Uchida, T. Nagao, J. & Narita, H. (2001). In Situ X-ray Diffraction Measurements of the Self-Preservation Effect of CH4 Hydrate. J. Phys. Chem. A, 105, 9756-9759. [282] Takeya, S. Hondoh, T. & Uchida, T. (2002). Insitu observations of CO2 hydrate by X-Ray diffraction. Ann. NY Acad. Sci., 912, 973-982. [283] Takeya, S. Kamata, Y. Uchida, T. Nagao, J. Ebinuma, T. Narita, H. Hori, A. & Hondoh, T. (2003). Coexistence of structure I and II hydrates formed from a mixture of methane and ethane gases. Can. J. Phys., 81, 479–484. [284] Takeya, S. Uchida, T. Nagao, J. Ohmura, R. Shimada, W. Kamata, Y. Ebinuma, T. & Narita, H. (2005). Particle size effect ofCH4 hydrate for self-preservation. Chem. Eng. Sci., 60, 1383-1387. [285] Takeya, S. Kida, M. Minami, H. Sakagami, H. Hachikubo, A. Takahashi, N. Shoji, H. Soloviev, V. Wallmann, K. Biebow, N. Obzhirov, A. Salomatin, A. & Poort, J. (2006). Structure and thermal expansion of natural gas clathrate hydrates. Chem. Eng. Sci., 61, 2670-2674. [286] Tanasawa, I. & Takao, S. (2002). Clathrate hydrate slurry of tetra-n-butylammonium bromide as a cold storage material. Proc. 4th Int. Conf. Gas Hydrates, 963-967, Yokohama, Japan. [287] Taylor, F. W. (1991). The greenhouse effect and climate change. Reports on Progress in Physics, 54, 881. [288] Taylor, C.E. (2005). Photocatalytic conversion of methane contained in methane hydrates. Topics in Catalysis, 32 (3-4), 179-184.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 57 [289] Ternes, M.P. (1984). Characterization of refrigerant - 12 gas hydrate formation for heat pump cool storage applications. ASME energy sources technology conference, 84-AES-12, American Society of Mechanical Engineers, New York, USA. [290] Thomas, S. & Dawe, R.A. (2003). Review of ways to transport natural gas energy from countries which do not need the gas for domestic use. Energy, 28 (14), 14611477. [291] Thompson, H. Soper. A.K. Buchanan, P. Aldiwan, N. Creek, J.L. & Koh, C.A. (2006). Methane hydrate formation and decomposition: Structural studies via neutron diffraction and empirical potential structure refinement. J. Chem. Phys., 124 (16), Art. No. 164508. [292] Tobie, G. Lunine, J.I. & Sotin, C. (2006). Episodic outgassing as the origin of atmospheric methane on Titan. Nature, 440 (7080), 61-64. [293] Tohidi, B. Danesh, A. Todd, A.C. Burgass, R.W. Østergaard, K.K. (1996). Equilibrium data and thermodynamic modelling of cyclohexane gas hydrates. Chem. Eng. Sci. 51(1), 159-163. [294] Tohidi, B. Danesh, A. Burgass, R.W. Todd, A.C. Østergaard, K.K. (1997). Equilibrium data and thermodynamic modelling of cyclopentane and neopentane hydrates. Fluid Phase Equilibria, 138, 241-250. [295] Tombari, E. Presto, S. Salvetti, G. & Johari, G.P. (2006). Heat capacity of tetrahydrofuran clathrate hydrate and of its components, and the clathrate formation from supercooled melt. J. Chem. Phys., 124 (15), 154507. [296] Tomlison, J.J. (1982). Heat pump cool storage in a clathrate of Freon. Proc. 17th Intersoc Energy Conversion Eng. Conf., Vol.4, 2060-2064, New York, USA. [297] Trebble, M. A. & Bishnoi P. R. (1988). Extension of the Trebble-Bishnoi Equation of State to Fluid Mixtures. Fluid Phase Equilib., 40, 1-21. [298] Tse, J.S. Handa, Y.P. Ratcliffe, C.I. & Powell, B.M. (1986). Structure of oxygen clathrate hydrate by neutron powder diffraction. J. Incl. Phen. Microcyclic Chem., 4, 235-240. [299] Tsuji, H. Ohmura, R. & Mori, Y.H. (2004). Forming Structure-H Hydrates Using Water Spraying in Methane Gas: Effects of Chemical Species of Large-Molecule Guest Substances. Energy and Fuels, 18(2), 418-424. [300] Tsuji, H. Kobayashi, T. Ohmura, R. & Mori, Y.H. (2005a). Hydrate Formation by Water Spraying in a Methane + Ethane + Propane Gas Mixture: An Attempt at Promoting Hydrate Formation Utilizing LMGS for sH Hydrates. Energy and Fuel, 19, 869-876. [301] Tsuji, H. Kobayashi, T. Okano, Y. Ohmura, R. Yasuoka, K. & Mori, Y.H. (2005b). Thermodynamic Simulations of Isobaric Hydrate-Forming Operations: Formulation of Computational Scheme and Its Application to Hydrate Formation from a Methane + Ethane + Propane Mixture. Energy and Fuel, 19, 1587-1597. [302] Tulk, C.A. Ripmeester, J.A. & Klug, D.D. (2000). The application of Raman spectroscopy to the study of gas hydrates. Ann. N.Y. Acad. Sci., 912, 859-872. [303] Uchida, T. Hondoh, T. Mae, S. Lipenkov, V.Y. & Duval, P. (1994). Air-hydrate crystals in deep ice-core samples from Vostok Station Antartica. J. Glaciology, 40 (134), 79-88.

58

Peter Englezos, John Ripmeester and Robin Susilo

[304] Uchida, T. Hirano, T. Ebinuma, T. Narita, H. Gohara, K. Mae, S. & Matsumoto, S. (1999a). Raman spectroscopic determination of hydration number of methane hydrates. AIChE J., 45 (12), 2641-2645. [305] Uchida, T. Ebinuma, T. Kawabata, J. & Narita, H. (1999b). Microscopic observations of formation processes of clathrate-hydrate films at an interface between water and carbon dioxide. J. Crystal Growth, 204 (3), 348-356. [306] Uchida, T. Takeya, S. Wilson, L.D. Tulk, C.A. Ripmeester, J.A. Nagao, J. Ebinuma, T. & Narita, H. (2003). Measurements of physical properties of gas hydrates and in situ observations of formation and decomposition processes via Raman spectroscopy and X-ray diffraction. Can. J. Phys., 81, 351–357. [307] Uchida, T. Moriwaki, M. Takeya, S. Ikeda, I.Y. Ohmura, R. Nagao, J. Minagawa, H. Ebinuma, T. Narita, H. Gohara, K. & Mae, S. (2004). Two-Step Formation of Methane–Propane Mixed Gas Hydrates in a Batch-Type Reactor. AIChE J. 50(2), 518-523. [308] Udachin, K.A. Ratcliffe, C.I. Ripmeester, J.A. (2002). Single Crystal Diffraction Studies of Structure I, II and H Hydrates: Structure, Cage Occupancy and Composition. J. Supra. Chem. 2, 405-408. [309] Ullerich, J.W. Selim, M.S. & Sloan, E.D. (1987). Theory and measurement of hydrate dissociation. AIChE J., 33, 747-752. [310] Van Cleef, A. & Diepen, G.A.M. (1960). Gas hydrates of nitrogen and oxygen. I. Recueil Trav. Chim., 79, 582-586. [311] Van Cleef, A. & Diepen, G.A.M. (1965) Gas hydrates of nitrogen and oxygen. II. Recueil Trav. Chim., 84, 1085-1093. [312] Van der Waals, J.H. & Platteeuw, J.C. (1959). Clathrate solutions. Adv. Chem. Phys., 2, 1-57. [313] Vatamanu, J. Kusalik, P.G. (2006). Molecular insights into the heterogeneous crystal growth of sI methane hydrate. J. Phys. Chem. B, 110 (32), 15896-15904. [314] Vlahakis, J.G., Chen, H.-S., Suwandi, M.S., Barduhn, A.J., The growth rate of ice crystals: properties of carbon dioxide hydrate, a review properties of 51 gas hydrates, Syracuse U. Research and Development report 830, prepared for U.S. Department of the Interior, November (1972). [315] Vysniauskas, A. & Bishnoi, P.R. (1983). A kinetics study of methane hydrate formation. Chem. Eng. Sci., 38 (7), 1061-1072. [316] Wanatabe, K. Imai, S. & Mori Y.H. (2005). Surfactant effects on hydrate formation in an unstirred gas/liquid system: An experimental study using HFC-32 and sodium dodecyl sulfate. Chem. Eng. Sci., 60 (17), 4846-4857. [317] Wang, X. Schultz, A. Halpern, Y. (2002). Kinetics of methane hydrate formation from polycrystalline deuterated ice. J. Phys. Chem. B, 106 (32), 7304-7309. [318] West, O.R. Tsouris, C. Lee, S. McCallum, S.D. & Liang, L. (2003). Negatively Buoyant CO2-Hydrate Composite for Ocean Carbon Sequestration. AIChE J., 49(1), 283. [319] Wilson, L. D. Tulk, C. A. Ripmeester, J. A. (2002). Instrumental Techniques for the Investigation of Methane Hydrates: Cross-calibrating NMR and Raman Spectroscopic Data. Proc. 4th Int. Conf. Gas Hydrates, 2, 614-618.

Clathrate Hydrate Crystallization for Clean Energy and Environmental Technologies 59 [320] Yeon, S.-H. Seol, J. & Lee, H. (2006). Structure Transition and Swapping Pattern of Clathrate Hydrates Driven by External Guest Molecules. J. Am. Chem. Soc., 128(38), 12388-12389. [321] Yoon, J.-H. Yamamoto, Y. Komai, T. Haneda, H. (2003). Rigorous Approach to the Prediction of the Heat of Dissociation of Gas Hydrates. Ind. Eng. Chem. Res., 42, 1111-1114. [322] Yoon, J.-H. Kawamura, T. Yamamoto, Y. & Komai, T. (2004). Transformation of Methane Hydrate to Carbon Dioxide Hydrate: In Situ Raman Spectroscopic Observations. J. Phys. Chem. A, 108, 5057-5059. [323] Yoon, J.-H. Kawamura, T. Ohtake, M. Takeya, S. Komai, T. Yamamoto, Y. Emi, H. Kohara, M. Tanaka, S. Takano, O. & Uchida, K. (2006). Highly selective encaging of carbon dioxide molecules in the mixed carbon dioxide and nitrogen hydrate at low temperatures. J. Phys. Chem. B, 110 (35), 17595-17599. [324] Zachos, J. C. Lohmann, K. C. Walker, J. C. G. & Wise, S. W. (1993). Abrupt climate change and transient climates during the paleocene: a marine perspective. Journal of Geology, 101, 191. [325] Zeng, H. Wilson, L.D. Walker, V.K. & Ripmeester, J.A. (2003). The inhibition of tetrahydrofuran clathrate-hydrate formation with antifreeze protein. Can. J. Phys., 81 (1-2), 17-24. [326] Zeng, H. Wilson, L.D. Walker, V.K. & Ripmeester, J.A. (2006a). Effect of antifreeze proteins on the nucleation, growth, and the memory effect during tetrahydrofuran clathrate hydrate formation. J. Am. Chem. Soc., 128 (9), 2844-2850. [327] Zeng, H. Moudrakovski, I.L. Ripmeester, J.A. & Walker, V.K. (2006b). Effect of antifreeze protein on nucleation, growth and memory of gas hydrates. AIChE J., 52 (9), 3304-3309. [328] Zhong, Y. & Rogers, R.E. (2000). Surfactant effects on gas hydrate formation. Chem. Eng. Sci., 55, 4175-4187.

In: Electroanalytical Chemistry: New Research Editor: G. M. Smithe

ISBN: 978-1-60456-347-4 © 2008 Nova Science Publishers, Inc.

Chapter 2

CORROSION RESEARCH FRONTIERS. ATMOSPHERIC CORROSION IN TROPICAL CLIMATE. ON THE CONCEPT OF TIME OF WETNESS AND ITS INTERACTION WITH CONTAMINANTS DEPOSITION F. Corvo, T. Pérez*, Y. Martin**, J. Reyes*, L.R. Dzib* J.A. González* and A. Castañeda** Instituto de Ciencia y Tecnología de Materiales, Universidad de la Habana, Cuba. *Centro de Investigaciones de Corrosión (CICORR), Universidad Autónoma de Campeche, Campeche, México. **Centro Nacional de Investigaciones Científicas (CNIC), La Habana, Cuba

Abstract Atmospheric corrosion is the most extended type of corrosion in the World. Over the years, several papers have been published in this subject; however, most of the research has been made in non-tropical countries and under outdoor conditions. Results of outdoor and indoor corrosion rate and corrosion aggressivity in tropical corrosion test stations of Cuba and Mexico are reported. Time of wetness (TOW), considered as the time during which the corrosion process occurs, is an important parameter to study the atmospheric corrosion of metals. According to ISO-9223 standard, TOW is approximately the time when relative humidity exceeds 80% and temperature is higher than 0oC. No upper limit for temperature is established. In tropical climates, when temperature reaches values over 25oC, evaporation of water plays an important role and the possibility to establish an upper limit respecting temperature should be analyzed. The concept of TOW assumes the presence on the metallic surface of a water layer; however, there are recent reports about the formation of water microdrops during the initial periods of atmospheric corrosion, showing that the idea of the presence of thin uniform water layers is not completely in agreement with the real situation in some cases (particularly indoor exposures).

62

F. Corvo, T. Pérez, Y. Martin et al. Most of the research carried out to study the initial stages of atmospheric corrosion have been made on a clean surface without corrosion products; however, the metal is very often covered by thin or thick corrosion products after a given exposure time and these products usually act as retarders of the corrosion process. In the Cuban Isle, the influence of chloride ions is very significant in determining the corrosion rate. In the coastal territory of the Mexican Gulf, particularly at Campeche, the deposition of Chloride ions is lower. No previous reports have been made about the interaction between chloride deposition rate and rain. The influence of rain seems to be important in determining the acceleration rate of chloride ions on metals due to its washing effect. To consider the influence of the interaction chloride deposition rate–rain regime could be useful to improve the prognosis of corrosion aggressivity. The predominant wind direction corresponding to geographic sites result in an important parameter for chloride deposition and their influence on surface wetness. The calculation of Time of Wetness established in ISO 9223 should be revised based on new results obtained in outdoor and indoor conditions in tropical humid marine climate. Some proposals are made to improve the estimation of TOW, taking into account changes in its nature depending on outdoor or indoor exposure, linear relationship between time and TOW, the effect of rain, and the role of contaminants and air temperature.

Introduction Atmospheric corrosion is the most extended type of corrosion in the World. Over the years, several papers have been published in this subject; however, most of the research has been made in non-tropical countries and under outdoor conditions. The tropical climate is typical of equatorial and tropical regions and is characterized by permanently high temperatures and relative humidity with considerable precipitation, at least during part of the year. A high corrosion rate of metals is usually reported for this climate. Results of outdoor and indoor corrosion rate and corrosion aggressivity in tropical corrosion test stations of Cuba and Mexico are reported. The results mainly concern to natural atmospheric corrosion tests obtained in the western side of the Isle of Cuba and in the Campeche State located at the Yucatán Peninsula in Mexico. The two regions are located in the tropical climate and receive the influences of the waters of the Atlantic Ocean and the Mexican Gulf. Data processed in this paper correspond to atmospheric corrosion tests carried out during a long period of time, about the last 20 years in Cuba and the last 10 years at Campeche up to the present.

Climate of Cuba and the Yucatán Peninsula The humid tropical climate of Cuba and the Yucatán Peninsula (México) is characterized by an average air temperature always higher than 15oC, frequently high relative humidity, a summer or wet season (may to october) with frequent and heavy precipitations and a winter of dry season (november to april) with lower precipitations. In the case of these two regions there is a natural source of airborne salinity: the waters of the Atlantic Ocean, the Gulf of Mexico and the Caribbean Sea. Airborne salinity plays an important role in determining corrosion aggressivity in Cuba [1-4] and in the Yucatán Peninsula [2, 5-6]. Other anthropogenic contaminants can be present also in this region, particularly sulfur compounds coming from the oil production and manufacture industries and

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

63

transportation. Depending on the contaminants sources existing in different regions, the kind of contaminants may change, but in general, the main types of contaminants are Chlorides and Sulfur compounds Air temperature in the Yucatán Peninsula reaches higher values than in Cuba, the top average air temperature in Cuba is 26oC in the eastern shoreline [7]. In the Yucatán Peninsula part of the territory can be classified as Tropical very warm with an average temperature over 26oC [8]. Absolute maximum temperature in Cuba is of 38,6oC, meanwhile a temperature over 40oC is frequently reported in sites of the Yucatán Peninsula, particularly in Campeche. About ¼ of the Cuban territory is composed by mountains and the rest is mainly terrain flatness. The Yucatán Peninsula has no mountains. The lower temperatures and higher precipitations rates in Cuba are localized in the mountain regions. Orography is a transformation factor of local wind regime. Average relative humidity in the west side of Cuba is frequently higher than in Campeche, the average is about 80% for Cuba and about 76% for Campeche. Daily maximum relative humidity values are over 90% for the west part of Cuba and over 88% for Campeche. In spite of this, the classification of TOW according to ISO-9223[9] is the same for the two territories: τ4, corresponding to “Outdoor atmospheres at all climates, excepting dry and cold climates”. The influence of air temperature may change depending on the conditions. The effect of increasing temperature appears to be an increase in corrosion rate under conditions of permanent surface wetting, such as those obtained during precipitation. Under conditions of varying surface wetting, however, the corrosion rate increases with temperature up to a certain maximum value, and thereafter decreases [10]. ISO 9223 standard established that for purposes of aggressivity determinations the use of temperature-humidity complex data is the recommended methodology, because corrosion rate depends on this complex and not only on temperature or humidity independently. Taking into account this report, a decrease in corrosion rate should be expected in tropical climate for an increase in average air temperature.

Time of Wetness (TOW) and ISO 9223 Definition. Taking into account the electrochemical nature of the atmospheric corrosion process it is absolutely necessary to use the concept of Time of Wetness (TOW). It is a concept commonly used in atmospheric corrosion of metallic materials and refers to the time when the metal is sufficiently wet for corrosion reaction to occur, that is, when an electrolyte is present in the metallic surface. Under the particular characteristics of atmospheric corrosion there are time periods where corrosion could not occur due to the absence of an electrolyte in the metallic surface. The lowest outdoor TOW values are observed in the desert regions, as also in the Antarctic and Arctic regions. Atmospheric corrosion rates of metals at these climatic conditions are also very low and in the case of cold regions, the increase of temperature leads to the increase of TOW and corrosion rate [11]. In principle, TOW is a parameter that depends upon both the climatic conditions and in the characteristics of the metallic surface. The definition of TOW presented on ISO standard 9223 is the following: “The period during which a metallic surface is covered by adsorptive and/or liquid films of electrolyte that are capable of causing atmospheric corrosion”.. In addition, the new document ISO WD/9223

64

F. Corvo, T. Pérez, Y. Martin et al.

[11] defines: The wetting of surfaces is caused by many factors, for example, dew, rainfall, melting snow and a high humidity level. The length of time when the relative humidity is greater than 80% at a temperature greater than 00C is used to estimate the calculated time of wetness (τ) of corroding surfaces. Information on calculated time of wetness is helpful for informative atmosphere corrosivity estimation. In this standard, TOW is “estimated” based on the characteristics of the temperature humidity complex, independently of the pollutant level and the nature of the metal or alloy. It is not the actual time of wetness, but it is an estimation based only on climatic factors (temperature and relative humidity) and independent of the nature and characteristics of the metallic surface. Some results obtained in the tropical regions of the Gulf of Mexico suggest that an upper limit in temperature for the definition of TOW could be established [3-4,12-14]. TOW as defined in ISO 9223 does not cover all the aspects of climate. According to the ISO definition, in the case of tropical climate where air temperature never reach 0oC, TOW is estimated as the time when relative humidity is over 80%. The diminition of the electrolyte layer at relative humidity over 80% when temperature is over 25oC has been reported for Santiago de las Vegas rural station in Cuba [12] using TOW sensors based in a circuit print covered with a gold layer. By other way, metal temperature and TOW were registered in a rural station at 30 km of the seashore located in Merida, Yucatan Peninsula, using copper/gold sensors [14] according to ASTM Practice G8489 (1999). It was reported the influence of the nature of the metal and their corrosion products, the orientation of the sample (skyward or groundward), the direction of the winds and rain precipitation on the measured TOW values. None of this factor is taken into consideration by ISO 9223. A quite different temperature was determined for the metal surface respecting air temperature. It is perhaps an explanation to the very possible increase in evaporation of the surface electrolyte when air temperature is over 25oC. Water adsorption on silver surfaces exposed into a ventilated shed in an urban-rural site of Cuba was studied [13] using quartz resonators covered with a silver layer. It was determined that in these indoor conditions water adsorption significantly diminishes when air temperature increases over 25oC at relative humidity ranges of 80-90% and 90-100%. All these results confirm the idea that an upper limit of temperature should be established for the estimation of time of wetness. The presence of water does not only create conditions for the existence of an electrolyte, but it acts as a solvent for the dissolution of contaminants [10]. Oxygen plays an important role as oxidant element in the atmospheric corrosion process. The thickness of the water layer determines the oxygen diffusion toward the metallic surface and also the diffusion of the reaction products to the outside interface limited by the atmosphere. Another aspect of ISO definition is that “a metallic surface is covered by adsorptive and/or liquid films of electrolyte”. According to new results, the presence of adsorptive or liquid films of electrolyte perhaps could be not in the entire metallic surface, but in places where there is formed a central anodic drop due to the existence of hygroscopic particles or substances surrounded by microdrops where the cathodic process takes place. This phenomenon is particularly possible in indoor conditions [15-18]. Using ISOCORRAG, MICAT and Russian data, Tidblad et al [19] showed that the inclusion of temperature among the environmental parameters improves considerably the usefulness of the dose-response functions and should be adapted in the revision of ISO 9223 standard. It is reported an increase in corrosion rate with average air temperature in the range of -15 to 30oC.

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

65

In the eighties of the last century the adsorption of water layers on surfaces was studied [20]. It was demonstrated that the change in thickness of these layers depends on the physicochemical properties of water in these thin water layers. It is reported that on iron surfaces, the number of adsorbed water layers is about 15 at RH 55% and 90 at 100%. Similar values are obtained for Copper and Zinc; however significant differences are reported for Platinum, gold, aluminum and silver. These monolayers have been calculated only in presence of water (without oxygen) where the corrosion process is very slow and, consequently, in conditions far from the reality. Almost all tests carried out to study the starting process of atmospheric corrosion have been performed in a surface without corrosion products; however, in real conditions, the metal is covered with corrosion products after a given time and these products begin to play its role as retarders of the corrosion process in almost all cases. Corrosion products acts as a barrier for oxygen and contaminants diffusion, the free area for the occurrence of the corrosion is lower; however, the formation of the surface electrolyte is enhanced. Only in very polluted areas the corrosion products accelerate the corrosion process. Water adsorption isoterms were determined to corrosion products formed in Cuban natural atmospheres[21]. Sorption properties of corrosion products (taking into account their salt content-usually hygroscopics) determine the possibilities of surface adsorption and the possibility of development of corrosion process For a single type of climate could be more practical and easy the use of time instead of TOW, because it is expected the same category of TOW according to ISO 9223. In other words, it makes no sense to work using TOW-ISO in tropical climate, because in general it is in the same category at all places, so its influence is the same. Using data for different climates or wide regions a significant change in TOW-ISO could take place, including changes in wet/dry cycles. In these conditions it should be recommendable to use TOW according to ISO definition.

Time and TOW-ISO. Outdoor TOW-ISO Corrosion rate is a function of time of wetness, considered as the time during which corrosion occurs, but in general it should not be a linear function because corrosion rate changes with time. There are different factors influencing, for example, the protective properties of the corrosion products, the increase or decrease of the acceleration caused by contaminants, increase or decrease of the thickness and conductivity of the electrolyte layer, i.e. If the definition of TOW established by ISO is used (TOW-ISO), a linear relationship between time and TOW is obtained, in spite of the different possible changes in corrosion rate caused by changes in the nature of TOW. It has to be remarked that it is not the same effect on corrosion rate caused by a heavy rain than dew, fog or water adsorption, so for the same interval of relative humidity (80-100%), notable changes in nature of TOW-ISO and consequently in corrosion rate could take place. If a linear regression is fitted between time (sum of time in months) and TOW-ISO (sum of TOW corresponding to every month up to one year), a perfect linear relationship is obtained (see tables I, II, III, IV and V). The results show that a perfect linear relation exists

66

F. Corvo, T. Pérez, Y. Martin et al.

between time and TOW-ISO. It means that both variables are equivalent, so for purposes of long term prognosis, work carried out time give the same result than using TOW-ISO. It does not mean that the actual time of wetness should be the same as estimated according to ISO, because as it has been pointed out above, the ISO definition does not take into account the nature of the different components of TOW regarding climate (rain, dew, fog, water adsorption) and the nature of the metal and the corrosion products. Table I. Simple regression of TOW-ISO data and time for Campeche PCGM coastal test station in a period of 2 years. Equation t = a + b TOW-ISO, r = correlation coefficient, r2 = percentage of variation explained by the independent variable, P = Statistical Probability Year 2004 2005

a -230.42 -300.15

b 234.39 258.77

r 0.979 0.974

r2 95.90 94.88

P <0.0000 <0.0000

Campeche PCGM

Table II. Simple regression of TOW-ISO data and time for Viriato coastal test station in a period of 10 years. Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 = percentage of variation explained by the independent variable, P = Statistical Probability Year 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

a -487.20 -14.83 -160.20 89.30 -167.38 -237.26 -91.18 -70.60 -428.05 83.97 -101.30

b 406.48 423.15 365.59 299.48 395.20 375.55 348.37 380.73 413.16 252.75 390.60

r 0.987 0.997 0.998 0.996 0.998 0.996 0.996 0.999 0.994 0.997 0.999

r2 97.49 99.38 99.58 99.11 99.52 99.16 99.29 99.71 98.73 99.489 99.73

P <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000

Table III. Simple regression of TOW-ISO data and time for Holguin coastal station (located in the eastern side of Cuba) in a period of 3 years. Year 1987 1988 1989

a 126.42 27.13 54.44

b 469.84 435.98 403.36

r 0.999 0.999 0.999

r2 99.79 99.88 99.96

P <0.0000 <0.0000 <0.0000

Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 = percentage of variation explained by the independent variable, P = Statistical Probability

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

67

Table IV. Simple regression of TOW-ISO data and time for Cojimar coastal station in a period of 10 years. Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 = percentage of variation explained by the independent variable, P = Statistical Probability Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

a -127.68 -172.08 -54.0 -51.32 -289.58 -164.67 -51.44 -75.36 -286.80 -159.21

b 389.13 367.73 342.23 381.34 389.58 376.75 369.95 344.79 352.04 371.14

r 0.998 0.996 0.998 0.999 0.993 0.997 0.999 0.997 0.989 0.995

r2 99.52 99.14 99.53 99.78 98.60 99.41 99.83 99.40 97.89 99.02

P <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000

Table V. Simple regression of TOW-ISO data and time for Quivican rural station in a period of 8 years. Equation t = a + b TOW-ISO. , r = correlation coefficient, r2 = percentage of variation explained by the independent variable, P = Statistical Probability Year 1990 1991 1992 1993 1994 1995 1996 1997

a -89.37 -121.50 -66.07 -33.53 -75.86 -105.04 -62.50 -71.86

b 405.78 421.15 440.34 443.97 448.22 458.38 444.04 457.37

r 0.997 0.997 0.998 0.999 0.999 0.999 0.999 0.999

r2 99.34 99.34 99.59 99.85 99.73 99.72 99.53 99.78

P <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000 <0.0000

TOW-ISO Under Heat Trap Conditions A simple regression was made between time and TOW-ISO for the different test stations under heat trap conditions for a two year exposure time. The results were the following: . Heat trap conditions[22] refer to the exposure of samples inside metallic boxes. In those conditions of partially closed metallic boxes, sun irradiation causes a significant increase in temperature and the maximum values of air temperature are obtained under these conditions. Some electro-electronic items are used under these conditions. The results of the regression are the following: Quivican rural test station, Cuba: t= -0.005 + 0.0026 TOW-ISO r= 0.999 r2 = 99.94 P <0.0000 Campeche PCGM coastal test station, Mexico: t = 0.177 + 0.0032 OW-ISO r= 0.999 r2 = 99.79 P <0.0000

68

F. Corvo, T. Pérez, Y. Martin et al. CNIC urban rural test station, Cuba: t = 0.0011 + 0.003 TOW-ISO r = 0.999 r2 = 99.84 P < 0.0000 Cojimar coastal test station, Cuba: t = 0.097 + 0.0035 TOW-ISO r = 0.9997 r2 = 99.9383 P = 0.0000

The results show that a perfect relation exists between time and TOW-ISO, so both variables are equivalent and can be used for purposes of long term prognosis. There are some small differences between the metallic boxes used in Cuba and Campeche. On figures 1 and 2 it can be observed that Cuban metallic boxes are relatively more closed than the one used at Campeche PCGM corrosion station.

Figure 1. Metallic box exposed at Cojimar coastal station in the north shore of the City of Havana. Similar metallic boxes are exposed at CNIC and Quivican Cuban stations. Samples and contaminant detector are placed inside the metallic boxes.

Figure 2. Metallic box exposed at Campeche PCGM coastal station at 300 m of the shoreline in the City of Campeche (front view). Samples and contaminant detectors are placed inside the metallic boxes. As it can be observed, this metallic box has a wider window than those used in Cuban stations.

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

69

TOW-ISO Under Heat Trap Conditions at Long Term Exposure. It is very well known the equation: K=aτb Where: a = constant b = coefficient that indicates the protective properties of the corrosion products layer

τ = Actual time of wetness (effective time during which corrosion process occurs) It is very wide used in studying the atmospheric corrosion process and its development with time. As smaller will be coefficient b, it is supposed that the protective properties of the corrosion products layer are higher. On Table VI it is shown a comparison between data fitness using TOW calculated according to ISO 9223 and time. It can be observed that there are not significant differences in using TOW or time because r2 is almost the same in both cases. It could be explained supposing a direct relationship between time and TOW, that is, TOW increases with time. If, in general, there should be differences in TOW for the different climatic seasons, these are not significant for the total time. Table VI. Statistical fitness of data of corrosion of Copper, Steel, Nickel and Tin inside a metallic box (heat trap conditions) respecting time and TOW according to ISO definition. Station Cojimar (coastal)

CNIC (urban)

Quivicán (rural)

Campeche PCGM (coastal)

Metal Copper Steel Tin Nickel Copper Steel Tin Nickel Copper Steel Tin Nickel Copper Steel Tin Nickel

n.w.l.= No weight loss detected

K=aτb a b 0.0009 0.847 0.0038 1.185 0.0014 0.817 n.w.l. n.w.l. 1.1270 1.239 1.7930 1.596 0.0044 0.766 9.7180 1.216 0.0017 0.986 0.0206 1.087 0.0135 0.705 7.2561 0.953 0.00003 1.528 0.0080 1.137 0.0022 0.914 0.0000002 1.851

2

r 48.93 96.27 88.09 n.w.l. 90.85 94.39 85.15 97.54 98.31 97.83 70.76 94.82 93.68 93.33 88.23 88.25

a 0.138 4.284 0.177 n.w.l. 0.205 2.823 0.467 0.015 0.453 9.412 0.728 0.158 0.158 4.141 0.354 0.006

K= a t b b 0.805 1.455 0.795 n.w.l. 1.179 1.532 0.716 1.163 0.976 1.083 0.697 0.945 1.628 1.209 0.975 1.974

r2 49.76 97.11 98.41 n.w.l. 91.76 95.79 82.98 95.98 97.81 98.68 70.30 94.67 90.68 90.84 85.62 85.55

70

F. Corvo, T. Pérez, Y. Martin et al.

Indoor Humidity Under indoor conditions, in the same way than outdoors, it is necessary the presence of surface humidity for corrosion to occur due to the electrochemical nature of the atmospheric corrosion process; however, in indoor conditions there are no precipitations and the presence of surface water depends mainly on water content in the air and changes in temperature on the surface, as well as the presence of hygroscopic substances on the metallic surface. Indoor temperature and relative humidity appreciably depends on the ventilation level, on the use or not of air conditioner or heating systems and of the thermal isolation. Leygraf reports[23] that indoor relative humidity ranges from 15 to 85% with an average value of 50%. In Cuba [24] average values significantly different to 50% have been found, because inside a closed storehouse (no windows) during one year an average value of 89,5% is reported. In a storehouse having windows the average value is of 79,5%. In a storehouse having a dehumidifier system the annual average relative humidity diminishes to 74,5%, still significantly high respecting Leygraf report. In two storehouses without climatic control it is reached a 100% of relative humidity, while in the storehouse submitted to dehumidification the relative humidity reaches 85%. Changes in indoor relative humidity are lower than those occurring outdoors, without the strong change caused by the difference between night and day. In this way, the most probable way the metallic surface can be humid is through water adsorption, although it can be observed that a significant relative humidity is reported indoors respecting Europe, so a higher corrosion should be determined. The higher temperature of the Cuban climate causes also the existence of higher water content and increases the possibilities of water condensation. Under indoor conditions, wind rate is lower than outdoors [25]. In this way, outdoor conditions promotes a fast surface drying due to the influence of air and sun radiation, and also a more frequent surface wetness thicker water layers due to rainfall, factors almost nonexistent in indoor conditions. Another factor nonexistent in indoor conditions is the washing effect on the metallic surface by precipitations. It is important to remark that indoors, TOW-ISO is higher in storehouses without ventilation than outdoors in tropical humid climates. ISO 9223 classifies TOW for no ventilated storehouses as τ5 (more than 5500 h/a), whereas for outdoor conditions it is τ4 (2500-5500 h/a). Taking into account that indoors corrosion rate is significantly lower respecting outdoors, it can be perfectly understood that the development of corrosion under adsorbed water is very low respecting outdoors. In these conditions, the magnitude of corrosion will be determined by the possibilities of acceleration that could cause contaminants, being the presence of water on the surface a requirement for the occurrence of corrosion, but not for determining the rate of the corrosion reaction. When the wind velocity is lower, drying process could be larger and could be a significant part of the dry/wetness cycle. It also depends on the temperature existing on the metallic surface. On ISO-9223 there are not considered changes in the dry/wet cycles. The deposition of contaminants in the surface usually causes an acceleration of the corrosion process; however, there should not be excluded the possibility that a given contaminant could diminish corrosion rate, as it could be the case of ammonia and its influence on steel (due to its alkaline properties, it could induce the passivation of steel). An important difference between outdoor and indoor conditions is that in the last one there is no washing of contaminants by precipitations. Contaminants deposition is lower, but

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

71

it increases with time. In these conditions, the presence of adsorbed water is necessary for corrosion to occur, but the rate of the process depends on the acceleration cause by contaminants. The calculation of TOW-ISO should not be as important as outdoors, because adsorbed water could be present all time. It has been reported a significant decrease in adsorbed water as temperature increases in Cuban climate [13]. In indoor conditions, the role of airborne salinity significantly diminishes, because the deposition rate is significantly lower than outdoors [24]. Stratmann reports [26-27] polarization curves on metallic surfaces covered by electrolyte layers as thin as 2 micrometers without using Luggin capillary, but using a Kelvin probe as reference (it is not necessary a contact to the surface). In this way there is avoided the effect in the polarization curve caused by the presence of a capillary having aqueous solution inside. It was possible to study the drying process, observing that at the beginning oxygen reduction increases due to a fast transport of this element through the thin electrolyte layer. Corrosion rate was determined by the oxygen consumption in a closed volume. It was shown that corrosion rate of pure steel could be divided into three parts: the increase in the oxygen diffusion current at the beginning, passivity of the surface in a second period and the diminishing in the oxygen reduction at the end of the drying process. Recent reports about the microdroplets formation in the starting periods of atmospheric corrosion [15-18] show that the idea of a thin uniform water layers is not completely in accordance with the reality. It has been observed that when a water drop is on the metallic surface, formed in the place where a salt deposit existed before, microdroplets are formed around this central drop. The cathodic process takes place in these surrounding microdroplets, meanwhile the anodic process takes place in the central drop. This idea is not consistent with the proposal of an uniform water layer on the surface and it is very probable that this situation could be obtained under indoor conditions. It has been determined that microdrops (about 1 micron diameter) clusters are formed around a central drop. An important influence of air relative humidity is reported on microdrops formation. There is a critical value of relative humidity for the formation of microdroplets. Under this value no microdroplets are formed. This value could be considered as the critical relative humidity. This situation is very similar to the process of indoor atmospheric corrosion: presence of humid air, deposition of hygroscopic contaminants in the surface, formation of microdrops. Water is necessary for corrosion reaction to occur, but the reaction rate depends on the deposition rate and nature of contaminants. In ISO 11844-3[28] it is established that the combination of different parameters is what determines the corrosivity of the atmosphere. Under indoor conditions, corrosion process depends on a more complex number of parameters than outdoors; however in the same way than outdoors two types of parameters are proposed: • •

Air temperature and humidity Air contaminants (gas and particulates)

The effect of these two groups of parameters is linked, because contaminants need a given level of humidity to act on the corrosion process. A combination of contaminants could have different effects than the sum of the individual effects. In ISO 11844-1 [29] it is explained that the impact of temperature and relative humidity can not be expressed

72

F. Corvo, T. Pérez, Y. Martin et al.

according to ISO-9223. It could be explained taking into account that the nature of TOW is very different to outdoors conditions. Recent reports [30-31] on the use of atmospheric corrosion sensors based on changes in electrical resistance showed that when there were no contaminants[29] , in tests of 100-110 h., corrosion rate was zero or insignificant. These sensors can determine changes in metal thickness lower than one nanometer. However, in the presence of 0.08 ppm of SO2 or 20 μg/cm2 of NaCl in the system, changes in thickness where always detected over 75% of relative humidity. Corrosion rate was determined at temperatures of 20, 30 and 40oC and the Arrhenius equation was used to calculate the activation energy of the reactions. This method is very similar to the natural conditions.

TOW at Different Exposure Conditions TOW depends mainly on the meteorological parameters, the nature of the metal, the properties of the metallic surface and of the corrosion products layer formed. The most important climatic factors on the corrosion process are relative humidity, sunshine hours, temperature of the air and the metal surface, wind velocity and duration and frequency of the rain, dew and fog. Dew or condensation of humidity is considered an important cause of the corrosion of metals. Its formation depends on the relative humidity and on the changes of temperature. Dew does not wash the metallic surface so the concentration of pollutants is relatively high and could be more aggressive than rain. Rain gives rise to the formation of a thick layer of water and also adds corrosive agents such as H+ and SO42-;however it can wash away the contaminants as well. It will depend on the intensity and duration of the rainfall. The results of a wide evaluation carried out on different exposure conditions in Cuba have already been published [3-4].The following Cuban corrosion test stations and conditions were considered: − − −

Coastal test station under outdoor, sheltered and ventilated shed conditions Urban-industrial test station under outdoor and sheltered conditions Rural test station under outdoor, sheltered, ventilated shed and closed space conditions

As was expected, the higher corrosivity corresponds to the coastal station. It was also noted that the higher the corrosivity of the atmosphere, the higher the difference between the outdoor corrosion rate and the indoor corrosion rate. The atmospheric corrosion rate of metals depends mainly on TOW and pollutants; however, if the differences in the corrosion process between outdoor and indoor conditions are taken into account, the influence of direct precipitation such as rain is very important for outdoor and negligible for indoor conditions. The acceleration effect of pollutants could change depending on wetness conditions of the surface, so the influence of the rain time and quantity should be very important in determining changes in corrosion rate. A model was proposed which consider the influence of the relationship between rain quantity/time and the interaction between pollutants at different TOW. TOW-ISO was divided into two parts, when air temperature was lower than 25oC and when air temperature

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

73

was higher than 25oC (up to 35oC). In this way three different times of wetness are considered: time due to rain (included independently of TOW-ISO), time including rain, dew and fog (TOW-ISO at air temperature <25oC), and time when evaporation of the electrolyte layer prevails (air temperature >25oC)-in some cases this last variable should also include rarely time of rain. This model was developed based on previous results indicating that when air temperature is over 25oC a significant diminution of the surface water layer takes place, possible due to a significant increase in evaporation of this water layer. By considering that the average corrosion rate is influenced principally by the deposition rates of chloride and SO2 and adding the effect of cleansing of the metallic surface caused by rain the following empirical model was proposed: K = a + b [Cl-] TOW-ISO5-25 + c [Cl ] TOW-ISO25-35 + d [SO2]TOW-ISO5-25 + [SO2] TOW-ISO25-35 + f [Cl ] Train + g [SO2] T rain + h mm/Train where K = Weight loss in g/m2 for the exposure period. [Cl ] = deposition rate of chloride ions in mg/m2d [SO2] = deposition rate of sulfur compounds in mg/m2d TOW-ISO5-25 = Time of wetness estimated according to ISO-9223 for air temperatures between 5 and 25oC. Lower temperatures are not considered because they are almost nonexistent in Cuban climate. TOW-ISO25-35 = Time of wetness estimated according to ISO-9223 for air temperatures between 25 and 35oC. Train = Time of rainfall (h) for the exposure period. This time is also included in TOWISO, but its effect is significantly different. mm = Amount of rain in mm for the period of exposure. Data of corrosion rate of carbon steel, copper, zinc and aluminum together with different TOW and contaminants were statistically processed (stepwise regression) and the following results were obtained: The variable [SO2] T rain in all cases was not statistically significant. Metal: carbon steel Outdoor conditions: K = 17.74 + [2.47 [Cl ] + 0.071 [SO2] ] TOW-ISO5-25 – 1.5 [Cl ] TOW-ISO25-35 r = 0.99 r2 = 99.0 Indoor conditions: K = 31.55 + [2.02 [Cl-] + 0.13 [SO2]] TOW-ISO5-25 – 1.06 [Cl-] TOW-ISO 25-35 R = 0.99 r2 = 96.0 Outdoor and Indoor conditions:

74

F. Corvo, T. Pérez, Y. Martin et al.

K = 21.29 + [0.62 Train + 1.59 TOW-ISO 5-25 – 1.11 TOW-ISO TOW-ISO25-35 r = 0.98 r2 = 95.0

25-35]

[Cl-] + 0.09 [SO2]

Metal: Copper Outdoor conditions: K = 6.13 + 0.97 [Cl-] Train + 0.29 [SO2] ] TOW-ISO5-25 + 0.24 mm/Train r = 0.98 r2 = 95.0

Indoor conditions: K = 0.88 [Cl-] TOW-ISO5-25 + 0.17 [SO2] TOW-ISO 5-25 r 0.99 r2 = 97.0 Outdoor and Indoor conditions: K = [1.03 [Cl-] + 0.24 [SO2] ]TOW-ISO 5-25 – 0.46 [Cl-] T rain + 0.19 mm/Train r = 0.98 r2 = 94.0 Metal: Zinc Outdoor conditions:

K = 12.22 + 0.98 [Cl-] Train r = 0.98 r2 = 96.0

Indoor: K = [1.98 [Cl-] + 0.13 [SO2] ]TOW-ISO5-25 – 1.09 [Cl-] TOW-ISO 25-35 r 0.99 r2 = 99.0 Outdoor and indoor: K = 6.28 + [1.86TOW-ISO 5-25 – 0.96 TOW-ISO25-35] + 0.12 [Cl-] Train + 0.06 mm/Train r = 0.98 r2 = 95.0 Metal: Aluminum Outdoor conditions:

K = 0.46 + 0.96 [Cl-] Train r = 0.96 r2 = 90.0

Indoor:

K = 0.80 [Cl-] TOW-ISO5-25 + 0.37 [SO2] TOW-ISO5-25 r 0.99 r2 = 96.0 Outdoor and indoor: K = 0.55 + [1.81[Cl-] + 0.18 [SO2] ]TOW-ISO 5-25 –[ 0.24 Train + 0.89 TOW-ISO25-35] [Cl ] - 0.22 mm/Train r = 0.96 r2 = 89.0 The suggestion of dividing the time of wetness into three different contribution parts was made in order to get a more quantitative approach step to study the atmospheric corrosion process. The influence of time and quantity of rain is very important for characterizing differences between indoor and outdoor corrosion. It can be observed that in all cases the variable TOW 25-35 is affected by a negative sign, indicating a diminishing of corrosion rate

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

75

when this variables increases, what is according to the fact that when temperature is over 25oC the electrolyte layer significantly diminishes instead of existing a high relative humidity. The cleaning effect of rain is important in the corrosion process. In many cases this variable is significant, as well as the inclusion of time of rain as an independent variable in explaining the influence of different parameters in atmospheric corrosion rate of basic metals.

Comparison of Air Temperature, Relative Humidity, TOW-ISO, Corrosion of Steel and Deposition of Contaminants at Different Exposure Conditions. A comparison of the behavior of air temperature for different exposure conditions in Cuba using data corresponding to one year is presented on figure 3. It is only an approximated behavior because there was not used the same year for all exposure conditions. Average air temperature is around 25oC for all conditions, slightly higher in the storehouses in urban and rural sites. No data of average minimum and maximum temperatures were available for storehouses. It can be observed that higher temperatures are reported for conditions of heat trap and lowers for storehouses. The wider interval of temperature is also obtained at these conditions, probably due to the emission of heat of the painted galvanized steel used in the construction of the metallic boxes. The lower interval between maximum and minimum temperature is obtained in the case of storehouse. It is due to the fact that these are wide and close spaces with concrete walls where air temperature changes are relatively low. The lower temperature interval is reported for the storehouse with air conditioner. In outdoor, sheltered conditions and ventilated sheds there is reported a second wide interval of temperature. The lower and higher temperature is reported for the outdoor rural station, because it is relatively far from the seashore and its influence is lower. The behavior of relative humidity is shown of figure 4. It is interesting to note that maximum relative humidity is obtained in almost all exposure conditions, excepting the storehouse having air conditioner and rural ventilated shed, this last with a very small difference. It means that regarding relative humidity the possibilities of condensation of moisture are significantly higher if the required conditions are available. Average relative humidity is around 80%, higher in the storehouse having no windows and lower in the ventilated shed located in the coastal site. It is very well known that in a closed place relative humidity increases. No data of average maximum and minimum relative humidity for storehouses were available. The lower average in relative humidity between storehouses is reported for the storehouse having air conditioner as it should be expected. These two figures refer to the temperature-relative humidity complex, but there is an important difference in the environment between outdoor, sheltered, ventilated shed and indoor corrosion. Precipitation is only possible in outdoor conditions, although condensation of moisture is possible in sheltered conditions and less probable in ventilated shed. In storehouses and heat trap conditions there are no possibilities of precipitation and condensation of moisture is more difficult. A calculation of time of wetness according to ISO 9223, considered as the time when relative humidity is over 80% and air temperature over 0oC is presented on figure 5.

76

F. Corvo, T. Pérez, Y. Martin et al.

It can be noted that the lower TOW is obtained for the storehouse having air conditioner and the higher for the storehouse located in the coastal site and the metallic box located in the rural site. According to these results, the higher possibilities of corrosion should be in these last conditions if we only consider the definition of TOW and the values obtained. However, it is very well known that the higher corrosion rate is usually obtained in outdoor or sheltered conditions. This figure shows that the concept of TOW established on ISO 9223 could not be used simply and many considerations have to be made. It should be limited to outdoor conditions and some considerations have to be made respecting the nature and characteristics of different meteorological parameters such as rain, dew, fog, i.e.. In other words, the concept of time of wetness (established in ISO 9223) shows limitations, because the nature and the changes occurred in the electrolyte formed on the metallic surface have clear differences between exposure conditions. There is a significant difference between outdoor time of wetness and sheltered or indoor TOW. Outdoor and indoor air temperature

60

50 Air conditioned storehouse

)

40 (

Average T Ave.min.T 30 p

Ave.max T Min T Max T

20

10

no windows storehouse opened windows storehose

0 Rural

Coastal

coastal

Urban

Rural

Outdoor and

Heat trap

sheltered

conditions

Coastal

Figure 3. Changes in average, average maximum and minimum, maximum and minimum air temperature depending on exposure condition and type of atmosphere for test stations located in the western side of the Cuban Isle.

Actually, the general concept of TOW (time during which corrosion reaction occurs) is, undoubtedly, correct and generally applicable, taking int account the electrochemical nature of the atmospheric corrosion process, but the estimation made by ISO 9223 has to be limited to specific conditions, mainly outdoors and taking into consideration the differences in the corrosion reaction as a function of the way the water is present on the surface, because, for example, rain forms an electrolyte in the metallic surface different to dew, and the formation of the electrolyte during rain also causes a diminishing rate of deposition of contaminants in the surface depending on the extent of rain, How can this complicated process be

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

77

represented?. Obviously, not by the simple concept established by ISO 9223. It should be also considered the nature of the metal and the corrosion products. Outdoor and indoor relative humidity

100

90

no windows st orehouse

80

70

opened w indow s storehouse

60

Average RH Ave.min.RH

50

Ave.max.RH Max RH air condit ioned st orehouse

40

Min RH

30

20

10

0 Rural

Coast al

coast al

Urban

Out door and

Rural

Coast al

Heat t rap

shelt ered

condit ions Ex posur e c ondi t i ons

Figure 4. Changes in average, average maximum and minimum, maximum and minimum air relative humidity depending on exposure condition and type of atmosphere for test stations located in the western side of the Cuban Isle.

Average annual corrosion rate of steel reported for test stations depending on exposure conditions and type of exposure (see figure 6) is significantly different to the behavior of time of wetness for the same conditions. It is perfectly explained based on the influence of contaminants. Chloride deposition rate causes a significant acceleration of corrosion rate, which is the reason why the maximum annual corrosion lost is reported for outdoor conditions in the coastal zone, following sheltered and ventilated shed in the same coastal zone. Urban industrial atmospheres (outdoor, sheltered and ventilated shed) are the second one in weight lost and rural stations the less aggressive atmosphere. Regarding indoor corrosion, the lower weight lost is determined in the storehouses and heat trap conditions increase respecting the last one.

78

F. Corvo, T. Pérez, Y. Martin et al. Outdoor/ indoor time of wetness 8000 No direct wat er precipit at ions Closed st orehouse

7000 Rain, dew, condensat ion, spray 6000

5000

4000

t 80-100

Windows St orehouse

3000

2000

Air condit ioned st orehouse

1000

0 Rur a l

Rur a l

Out door a nd

Ve nt i l a t e d she ds

Ur ba n

Ur ba n

she l t e r e d Ex posur e c ondi t i ons

Figure 5. Time of wetness at different exposure conditions and types of atmospheres calculated according to ISO 9223 for test stations of the western side of the Cuban Isle. Annual corrosion rate of carbon steel at different types of exposures and atmospheres

4000

Corrosion rate (g/m2a)

3500

3000

2500 Rural Urban-industrial

2000

Coastal 1500

1000

500

0 Outdoor

Sheltered

Ventilated sheds

Storehouses

Heat trap conditions

Type of exposure

Figure 6.Average corrosion rate (g/m2a) depending on exposure condition and type of atmosphere for test stations located in the western part of the Cuban Isle

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

79

As it can be observed, in the storehouse where there is determined the maximum time of wetness, it is reported the lower corrosion rate. It is explained based in the fact that in this case TOW consists mainly in air humidity without the contribution of any precipitation or condensation (no rain, no dew, no significant condensation). It is a demonstration that the estimation of time of wetness according to ISO 9223 presents limitations. Changes in steel corrosion rate for different exposure conditions and types of atmosphere are better explained based on average deposition rate of Chlorides, as can be observed on figure 5. It can be note that Chloride deposition is higher in coastal stations, as it should be expected, and it is in agreement with a higher weight lost for these stations. The lower values of Chloride deposition are reported for the storehouses, where coincidently, there are reported the lower weight losses. Under heat trap conditions there is a given increase in Chloride deposition, that is why corrosion rate of steel is higher than in storehouses.

Average Chloride deposition rate at different exposure conditions

1000

Deposition rate (mg/m2d)

100

Rural 10

Urban-industrial Coastal

1

0,1 Outdoor

V.Shed

Storehouse

Heat-trap

Type of exposure

Figure 7. Annual average Chloride deposition rate for test station located at the western side of the Cuban Isle depending on exposure conditions and type of atmosphere.

On figure 7 there is represented Average Chloride Deposition rate determined in Cuba for different types of atmosphere and exposure conditions. The results are very similar to those presented on figure 6, particularly respecting outdoor and ventilated shed conditions. Chloride deposition is determined under shelter, that is why it is assumed the same value for ourdoor and sheltered conditions; however, it is very well known, the significant influence of precipitations upon outdoor corrosion and its negligible effect under sheltered conditions. In

80

F. Corvo, T. Pérez, Y. Martin et al.

storehouses and heat trap conditions the influence of Chloride deposition significantly diminishes; however, under heat trap conditions the higher corrosion of steel is reported for the coastal test station.

Average sulphur com pounds deposition rate at different exposure conditions 50 45

Deposition rate (mg/m2d)

40 35 30 Rural 25

Urban-industrial Coastal

20 15 10 5 0 Outdoor

V. Shed

Storehouse

Heat-trap

Type of exposure

Figure 8. Annual average sulfur compound deposition rate for test stations located in the western side of the Isle of Cuba depending on exposure conditions and type of atmosphere.

The average deposition of sulfur compounds found at the Cuban western side test station is presented on figure 8. As can be expected, the higher sulfur compound deposition rate corresponded to the urban industrial station in outdoor conditions. There is no report for ventilated shed. Part of the sulfur compound deposition for coastal stations should not be attributed to gaseous SO2 or SO3, because it is well known that airborne salinity contains sulphate ions and the procedure used for determining sulfur compounds deposition is sensible to sulphate ions (see ISO 9225 [31]). A direct relationship between average corrosion rate and sulfur compound deposition is not observed as in the case of Chloride deposition. It confirms the importance of Chloride deposition in determining corrosion aggressivity in Cuba.

TOW-ISO and Rain Chloride ion is one of the most important natural pollutants influencing corrosion; particularly in the tropical humid conditions of Cuba and Campeche in Mexico; however, its role may change depending on climate. A very humid climate can cause a fast leaching of the chloride ions and decreases its effect on the acceleration of corrosion rate. A different acceleration rate has been reported for chloride ions between Eastern and Western Caribbean

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

81

[2]. In a given climate the influence of rain could change the acceleration caused by chloride ion on metal corrosion. Differences between rainy and dry periods could be important in determining such acceleration. The electrochemical mechanism of corrosion by chloride ions does not change, but the time during which chloride ions acts on the metallic surface and their concentration may appreciable change, depending on factors such as climatic conditions and nature of corrosion products. A research about the role of the rainfall characteristics on the acceleration rate caused by chloride ions between two stations having noticeable differences in rain regime was already published[32]. The influence of chloride ions at Medellin and Havana corrosion stations was studied by submitting samples of steel and copper to salt spray during exposure to the open atmosphere. Up to 12 months of exposure, the acceleration on the deterioration caused by chloride ions was notably up to 12 months of exposure, being higher at Havana station for steel and copper. The corrosivity category for steel and copper (C3) in natural conditions in Havana increases to C5 for copper and over C5 for steel when salt spray is applied; however at Medellin, copper corrosion increases to C4 for copper and C5 for steel under the application of salt spray. The acceleration rate caused by the addition of a salt spray under natural conditions causes a higher acceleration of corrosion rate at Havana. For copper, the difference does not appreciably change with time; but it does for steel. A more aggressive action of chloride ions is observed in the case of steel. The remarkable difference in the acceleration caused by chloride deposition rate in Havana and Medellin could be due to the considerable difference in the rain regime between both sites. Other characteristic of the environment could also have influence, but rain should play an important role. Samples were submitted to the same artificial chloride deposition rate at Havana and Medellin stations (supposing that differences concerning natural values of chloride deposition rate are not very significant). Samples were prepared from the same origin material, so the differences on the corrosion behavior can be associated only to climatic characteristics. In the case of Medellin the washing effect of rain should be higher than in Havana because the rain amount is more than four times than the reported for Havana during the exposure period. It explains why a lower acceleration rate of chloride ions is obtained at the former station. Based on this fact, a model for studying the influence of the washing or cleaning effect of rain in the determination of the acceleration rate of chloride ions is proposed. The bi-logarithm equation for atmospheric corrosion establishes that: K=atb where K = mass loss; a and b = constants; t = time of exposure. In the presence of a given value of chloride deposition rate, an acceleration of corrosion takes place; this acceleration means that corrosion increases with time. At the same time, the acceleration of corrosion caused by chlorides depends on the washing or cleaning effect of rain. Under this condition the following model is proposed: K = a t b[Cl]c (W/D)d

82

F. Corvo, T. Pérez, Y. Martin et al.

where K = mass loss; a, b, c and d = constants; [Cl] = chloride deposition rate; W = rainfall (mm); D = rainy days; t = time of exposure. It has been considered that the washing or cleaning effect of rain could be represented by the ratio W/D (amount of rain/frequency of rain). This washing effect could affect the influence of chloride deposition rate on corrosion. A regression analysis including steel mass loss data for Havana and Medellin showed the following results: K = 18.8516 t1.0947 [Cl]0.4313 (W/D)–0.2447 r2 = 98.11 where K = mass loss (g/m2); t = time of exposure (months); [Cl] = chloride deposition rate (mg/m2 d); w = rainfall amount (mm); D = number of rainy days. A regression analysis including copper mass loss data for Havana and Medellin showed the following results: K =1.2466 t 0.7858 [Cl]0.2616(W/D)–0.2379 r2 = 98.08 A good data fit is also obtained for copper. It confirms that the complex chloride deposition rate–rain regime is important for determining mass loss of copper and steel. In this model TOW-ISO has not been used and a good fitness has been obtained. It confirms that the acceleration rate caused by chloride ions on atmospheric corrosion of steel and copper depends on the characteristics of rain regime. For a place having high amount and time of rain, a lower acceleration on corrosion rate should be expected for a given chloride deposition rate The washing and cleaning effect of rain is not included and is far from the concept of TOW established on ISO 9223 standard.

Corrosion in Tropical Coastal Atmospheres. Role of TOW-ISO Airborne salinity can be determined using different methods. In corrosion research the standard method (Wet Candle method) is established in ISO-9225: 1992 [33]; however, it is not the only method traditionally used. In the case of Cuba it has been widely used the method named as dry plate method, consisting in the employment of a dry cotton fabric of known area exposed under a shed. The amount of chloride deposition on the gauze is determined analytically at the end of the exposure period (two months) and the deposition rate is calculated. A report [34] about the simultaneous comparison between values obtained using Wet Candle and Dry plate methods at different corrosion stations in Cuba showed that there is not a good correlation for the rural station having lower values of salinity; however, a good correlation was obtained for stations having higher values of salinity. The following regression equation was obtained: [Cl-] W.C. = -54.5 + 1.6 [Cl-] D.P.

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

83

r = 0.98 n=14 P<0.005 where: [Cl-] W.C = Chloride deposition rate determined using Wet Candle method [Cl-] D.P = Chloride deposition rate determined using Dry Plate method Chloride deposition rate was determined using Wet Candle and Dry Plate methods in the corrosion stations Santiago de las Vegas (rural-urban), Casa Blanca (industrial-marine-urban), Via Blanca (industrial-urban-marine) and Cojimar (marine). On figure 9 it can be noted a significant difference between Steel corrosion rate and Chloride deposition between the north and the south shores. It is caused because, in general, trade winds in the north shore come from the Ocean and in the south shore from the earth. In addition, cold fronts always come from the north. The territory is flat, presenting only small hills, so Chloride deposition reaches almost all the territory. It can be seen that even in places located at 15-30 km from the north seashore a significant Chloride deposition rate is determined (3.4-8.5 mg/m2d-classified as S1 according to ISO-9223). Sulfur compounds deposition has also some relation with Chloride deposition. The higher values correspond to coastal and industrial stations. It means that, taking into account that the determination using alkaline surfaces is sensible to different sulfur compounds, it includes, in addition to sulfur oxides, sulfates coming in airborne salinity. Another possible sulfur compound determined could be H2S 600

4000

500 Corrosion rate (g/m2)

3000

Steel corrosion rate Chloride deposition SO2 deposition rate

2500

400

300

2000 1500

200 1000 100

Chloride and Sulphur compound depositiion rate (mg/m2d)

3500

500 0

0 0.01

North shore

0.15

2.0

2

2.0

3.0

3

6

15

18.0

30

30

30.0

48.0

South shore

Distance to the north sea shore (km)

Figure 9. Changes in annual Corrosion rate of mild steel flat samples, annual average Chloride deposition rate determined by the dry plate method and annual average Sulfur compounds deposition rate determined by alkaline surfaces method in the west side of the Cuban Isle.

Changes in TOW-ISO are negligible between all the test stations of the Cuban western side. All values are classified as τ4 according to ISO 9223. In these conditions main changes depend on contaminants deposition. Due to the hygroscopic properties of the salts deposited

84

F. Corvo, T. Pérez, Y. Martin et al.

in the surface there is no doubt that corrosion should occur at a relative humidity lower than 80%, so in this case the use of TOW-ISO has no sense. Deposition rate of Chloride at Campeche, Veracruz, Puerto Progreso and Puerto Morelos 500

Chloride deposition (mg/m2d)

450

Campeche Veracruz

400

Puerto Morelos - Puerto Progreso

350 300 250 200 150 100 50 0 -10

0

10

20

30

40

50

60

70

80

distance to the shoreline (km)

Figure 10. Annual average of deposition rate of Chlorides for different sites of the Yucatán Peninsula: Puerto Progreso and Puerto Morelos (Yucatán and Quintana Roo States), Campeche and Veracruz States. Steel corrosion rate at Campeche, Veracruz, Puerto Progreso and Puerto Morelos (flat specimens) 3500 Campeche

3000 2

Corrosion rate (g/m )

Veracruz Puerto Morelos - Puerto Progreso

2500 2000 1500 1000 500 0 -10

0

10

20

30

40

50

60

70

80

distance to the shoreline (km)

Figure 11. Annual average corrosion rate of carbon steel at Campeche, Veracruz and Puerto Morelos as function of distance to the shoreline.

On figure 10 the annual average Chloride deposition rate determined at three zones of the Yucatán Peninsula is presented: Puerto Progreso and Puerto Morelos in the mouth of the Gulf of Mexico and Campeche and Veracruz inside this Gulf as function of the distance to the shoreline. As can be seen, Chloride deposition is very similar at Puerto Progreso, Puerto Morelos and Veracruz; however, the values of Chloride deposition are significantly lower for Campeche. It could be explained based on the fact that predominant winds in Campeche most of the year are from earth to the Gulf of Mexico. Another factor is that sea Campeche usually

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate

85

has no wave movement. It is a similar situation to the south shore of the Cuban Isle where a lower Chloride deposition is determined. Changes of corrosion rate of steel as function of distance to the shoreline are lower respecting Veracruz. It is in perfect agreement with the influence of Chloride deposition. In the case of Puerto Morelos and Puerto Progreso it is also reported a significantly higher corrosion rate respecting Campeche. Changes in corrosion rate depending on the climatic season in coastal atmospheres. Chloride ion is one of the most important natural pollutants influencing corrosion; however, its role may change depending on climate. A very humid climate can cause a fast leaching of the chloride ions and diminish their effect on the acceleration of corrosion rate. A different acceleration rate has been reported for chloride ions between Eastern and Western Caribbean [2] It has been reported that the acceleration rate caused by chloride ions on atmospheric corrosion of steel and copper depends on the characteristics of rain regime. For a place having high amount and time of rain, a lower acceleration on corrosion rate should be expected for a given chloride deposition rate[32]. It is very well known that in tropical climate there are two main seasons, rainy season and dry season. Under this conditions, the acceleration caused by chlorides should be higher in the dry season (winter period) and lower in the rainy season. As an example, on Table VII presents statistical parameters calculated for corrosion rate of steel at Viriato coastal stations for periods of six months corresponding to the wet season (may to october) and dry season (november to april). All steel samples were exposed for a six months period corresponding, starting on may or on november. Data correspond to the period may/1987 to November/1991. As can be seen, a remarkable difference is determined for samples exposed in winter and in summer six months period at Viriato coastal station. It shows that a marked difference in aggressivity exists between climatic seasons in tropical climate. It is confirmed by the ratio Summer/Winter. It means that a higher acceleration caused by Chloride ions take place in the season having lower amount of rain. Changes in TOW-ISO are not remarkable. A slightly higher TOW-ISO is reported for summer period, the time when corrosion rate is lower. It seems that it has no sense to use TOW-ISO in studying atmospheric corrosion in tropical coastal atmospheres, because due to the hygroscopic nature of coastal salts, corrosion rate could take place very often at RH lower than 80%. On Table VII presents data for four test sites in Campeche (having lower deposition rates than in Havana). Chloride deposition is higher in winter periods for Campeche CRIP and Campeche TEC, the two sites nearer to the shoreline and with S1 ISO classification. In the case of Campeche SMN and Campeche PCGM (classification So according to ISO)the tendency is to a higher Chloride deposition in winter, but there are values of average deposition higher during the summer period.

Table VII. Average corrosion rate of steel, standard deviation, average Chloride deposition rate, standard deviation and aerage TOWISO (h) for six months exposure periods at Viriato coastal station, Cuba Climatic season

Average corrosion rate of steel (g/m2)

Standard deviation

Summer(May/oOct) Winter(Nov/Apr) Summer/Winter

569.4 2111.5 2.53

167.6 892.8 --

Average Chloride deposition rate (mg/m2d) 225.0 486.3 4.34

Standard deviation

TOW-ISO (h)

107.7 174.2 --

2112 2000 1.06

Table VIII. Average Chloride deposition rate for differents test stations of Campeche Test station SMN PCGM CRIP TEC

Average Chloride deposition rate (mg/ m2d) Summer/04 Summer/05 Winter/03-04 Winter/04-05 16,16 26,31 18,82 26,05 26,40 18,02 27,42 25,24 47,97 74,95 157,24 114,99 45,77 42,86 77,47 67,11

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate…

87

TOW and Corrosion Products of Steel. Chloride ions move through the corrosion products layer and arrive at metal surfaces producing a notable acceleration in corrosion rate. A catalyser role has been suggested for anions in the corrosion process. Chloride ions are very soluble, that is why they are easily removed from metallic surfaces due to liquid precipitation. Usually, a low concentration of chloride ions is determined in corrosion products on outdoor exposures. A relation between chloride concentration present in corrosion products and corrosion rate has been reported. Rust samples scraped off from exposed AISI 1019 steel surfaces were ground to fine size in a morter with a pestle. A fraction of the samples was subjected to a salts extraction process using distilled water to determine the concentration of most common ions by chemical analysis and to study changes in adsorption isotherm after elimination of hygroscopic salts. Samples were obtained from AISI 1019 steel coupons, exposed for up to six months at coastal and rural locations. The water adsorption isotherms were determined by the liquid-volumetric method[21] The rust samples were put in an atmosphere with certain amount of water up to obtaining an established equilibrium, the pressure is measured and the water adsorbed may be calculated. The experimental isotherms were fitted to the BET and Langmuir isotherms, selecting the best fit. The values of the monolayer capacity, the specific surface and the net heats of adsorption were calculated. The adsorption isotherms for metallic surfaces are reported in the literature; however, an important part of the atmospheric corrosion process takes place under rust layers, which play a decisive role in the long-term course of corrosion because of its sorption capacity for water. The influence of the chloride and sulfate anions has a real effect only when the corrosion products layer is already formed. Thus, the adsorption isotherms of the steel corrosion products formed in different atmospheres were determined. It can be seen that chloride, sulfate and bicarbonate concentrations are higher in the sample from Viriato station, because here the atmosphere is more contaminated. Nevertheless, the relation between the chloride concentration in the corrosion products from Viriato and Quivican is (24.64/15 = 1.65) far smaller than the Cl-DR relation (495.68/3.9 = 127.1) obtained in these two atmospheres. This is because the corrosion products are washed away from the corroding surface during rain and chlorides are easily eliminated from the metallic surface. It confirms the role of rain in washing and cleaning contaminants in the corrosion products layer. Chloride collectors are protected from liquid precipitations. There should be periods of very high chloride concentration with rapids changes due to liquid precipitations. In the case of sulfate, the ratio Viriato/Quivican is very similar to the atmosphere (17.28/3.84 =4.5 and 34.55/10.5 = 3.3, respectively). It seems that liquid precipitations do not appreciably affect sulfate concentration in corrosion products. The value of steel corrosion rate ratio for Viriato and Quivican stations at six months of exposure (1154.67/108.37 = 10.65) is between Cl-DR and chloride concentration in corrosion products ratio and is higher than sulfate. There should be rapid changes in chloride concentrations due to liquid precipitations and arrival of aerosol particles. Changes in sulfate concentration should be lower than for chloride. The corrosion rate in coastal zones should depend mainly on a sum of time of wetness at different chloride and sulfate concentrations.

88

F. Corvo, T. Pérez, Y. Martin et al.

Bicarbonate ions concentration at Viriato is higher as the value of pH increases. Thus, the corrosion process and the corrosion products transformation occur to a higher pH and this can lead to changes in the corrosion mechanism and in the nature of corrosion products formed. Sodium and potassium concentration (Na+ should be the main constituent) is higher at Viriato according to a very high anionic concentration and also a high conductivity is determined. Water adsorption should be caused by salts content but, perhaps, the presence of high levels of contaminants can determine the formation of a corrosion product having a more “hygroscopic” nature. To make clear this possibility, adsorption isotherm of corrosion products from Viriato corrosion station after elimination of salts were determined . The values of adsorption are lower after salt elimination, but higher than those corresponding to Quivican. It could be due to a more “hygroscopic” nature of corrosion products formed in coastal stations. The superficial characteristics of atmospheric corrosion products of steel depend on the type of atmosphere where the sample has been exposed. The way of adsorption of the corrosion products obtained in the coastal atmosphere is polymolecular due to a higher content of salts. This makes easier the presence of water in the metal-corrosion products interface and determines a high corrosion rate. The adsorption of water of a corrosion product formed in a rural zone obeys a Langmuir isotherm, i.e. a monomolecular adsorption takes place. It causes a lower corrosion rate. Chloride ions are easily washed away from the metallic surface by precipitation, while sulfate ions concentration remains more constant. It means that corrosion takes place under very variable concentrations of chloride ions and in the presence of less variable concentration of sulfate ions. Changes in the amount of chloride ions concentration should depend on the particular rain regime of the place. A notable difference has been found between chloride deposition rate determined by collectors protected from liquid precipitations and chloride content in corrosion products. The determination of adsorption isotherms offers the possibility to study the atmospheric corrosion process when the metal is covered by a corrosion products layer.

Conclusion TOW considered as the time during which corrosion occurs is an important parameter in atmospheric corrosion of metals. It defines the possibility for atmospheric corrosion to occur based on its electrochemical nature, but corrosion rate will depend mainly in the acceleration caused by contaminants deposition and other factors. • •

•

TOW according to ISO definition shows several limitations in tropical climate: It is a variable perfectly linked to time. It is not necessary to calculate TOW-ISO when the development of corrosion rate on time is studied because a linear relationship exists between time and TOW-ISO. Its nature changes from outdoor to indoor conditions. Higher TOW-ISO values are reported for indoor conditions where corrosion rate is significantly lower. It is mainly due to the absence of precipitations.

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate… •

• •

89

It has no sense to calculate TOW-ISO for coastal tropical atmospheres, because in those conditions corrosion process occurs at relative humidity lower than 80%. It has been determined that water adsorption by corrosion products is polymolecular in these conditions. As analogy, in highly polluted atmospheres, corrosion process should proceed at RH lower than 80%, so it has no sense to use TOW-ISO. TOW-ISO does not take into account the washing and cleaning effect of rain, a very important aspect in atmospheric corrosion outdoors. An upper limit for air temperature has not been included in TOW-ISO definition. Several results show that a significant change in TOW-ISO takes place at air temperature higher than 25oC and RH >80%.

In order to improve the estimation of TOW-ISO it is proposed: • To establish an upper limit of temperature for TOW-ISO calculation, dividing TOWISO in two categories: Air temperature 0>25< and air temperature >25 • To include time and amount of rain as an additional variable, taking into account the washing and cleaning effect of rain. • To limit the use of TOW-ISO to outdoor and not highly contaminated environments. • Always take into account the influence of contaminant deposition and its interaction with TOW. • The concept of an adsorptive or liquid film of electrolyte on the metallic surface should include the possibility of the existence of localized presence of adsorptive or liquid films on the metallic surface, particularly in indoor conditions.

Acknowledgments The authors appreciate the contribution of M. Sc. Miguel Ramón Sosa Baz (Corrosion Research Center, University of Campeche, Campeche, Mexico) and Tec. Eva González Mellor and Julia Pérez Acosta (National Center for Scientific Research (CNIC), Havana City, Cuba)

References [1] [2] [3] [4] [5] [6]

F. Corvo, N. Betancourt, A. Mendoza, Corrosion Science, vol. 37, No. 12, p. 18891901, 1995. F. Corvo, C. Haces, N. Betancourt, L. Maldonado, L. Veleva, M. Echevarria, O.T. de Rincón, A. Rincón, Corrosion Science, vol. 39, No. 5, p. 823-833, 1997. R. Mendoza, F. Corvo, Corrosion Science, vol. 41, 1999, p.75-86. A.R. Mendoza, F. Corvo, Corrosion Science, vol. 42, 2000, p. 1123-1147 L. Veleva, L. Maldonado, British Corrosion Journal, vol. 33, No.1, p. 53-57,1998. D. C. Cook, A.C. Van Orden, J. Reyes, S.J.Oh, R. Balasubramanian, J.J. Carpio, H.E. Townsend, Atmospheric Corrosion in Marine Environments along the Gulf of México, Marine Corrosion in Tropical Environments, ASTM STP 1399, S.W. Dean, G. Hernández Duque-Delgadillo and J. Bushman, Eds. American Society for Testing and Materials, West Conshohocken, PA, 2000.

90 [7] [8]

[9] [10] [11] [12] [13]

[14]

[15] [16] [17] [18]

[19]

[20] [21] [22] [23] [24] [25] [26]

F. Corvo, T. Pérez, Y. Martin et al. M. Sosa, A. del Sol, Revista de Climatología, Cuba, vol. 2, 2004, pp. 1-4. L. Mariaca, J. Genesca, J. Uruchurtu, L. Salvador. Corrosividad Atmosférica (MICAT-México). Programa Iberoamericano de Ciencia y Tecnología para el Desarrollo (CYTED), 1999. Editorial Plaza y Valdés S.A. de C.V., México. ISO 9223:1992. Corrosion of metals and alloys. Corrosivity of atmospheres. Classification. Leygraf, T.E. Graedel, Electrochemical Society Series, Wiley Intersciences, 2000 ISO/WD 9223. Corrosion of metals and alloys. Corrosivity of atmospheres. Classification R. Pascual, F. Corvo, Rev. Iberoamericana de Corrosión y Protección, mayo-junio, 1980. F. Corvo, J. Rocha. Characterization of environments and their influences on Copper at different atmospheres and exposure conditions in tropical climate, International Corrosion Congress (ICC), Granada, Spain, 2002. L. Veleva, M. A. Alpuches-Aviles, Outdoor atmospheric corrosion, ASTM, STP 1421, H.E. Townsend, Ed. American Society for Testing and Materials International, West Conshochoken, PA, 2002. Tooru Tsuru, Ken-Ichiro Tamiya, Atsushi Nishikata, Electrochimica Acta 49 (2004) 2709–2715. Li Bian, Yongji Weng , Xiangyi Li, Electrochemistry Communications, 7 (2005), 1033-1038. Jibiao Zhang, Jia Wang, Yanhua Wang, Electrochemistry Communications, 7 (2005) 443–448. Wang Jia, Zhang Jibiao. The effect of micro-droplets formation caused by the deliquescence of the deposited salt particle on atmospheric corrosion of metals. Proceedings16th International Corrosion Congress, Beijing, China, September 19-24, 2005. J. Tidblad , A. A. Mikhailov and V. Kucera, “Application of a Model for prediction of atmospheric corrosion in tropical environments”, Marine Corrosion in Tropical Environments, ASTM STP 1399, S.W. Dean, G. Hernández Duque-Delgadillo, and J. B. Bushman, Eds., American Society for Testing and Materials, West Conshohocken, PA, 2000. Iu. N. Mikhailovskii, Atmospheric corrosion of metals and protection methods (in russian). Editorial “Metallurgiia” , Moscow, 1989. F. Corvo, A. R. Mendoza, M. Autie, N. Betancourt, Corrosion Science, vol. 39, No. 4, pp. 815-820, 1997. José Rocha Andrade da Silva Editor. Productos electro-electrónicos en ambientes tropicales, Campinas, Sao Paulo, Sitta Gráfica 2003. Leygraf. Indoor atmospheric corrosion. Proceedings 15th International Corrosion Congreso, Granada, Spain, September 22-27, 2002. F. Corvo, A. D. Torrens, N. Betancourt, J. Perez, E. Gonzalez, Corrosion Science, vol. 49, No. 2, February 2007, , pp. 418-435. S. Cole, R. Holgate, P. Kao and W. Ganther, Corrosion Science, vol. 37, No. 3, 1995, pp. 455-465. M. Stratmann, H. Streckel, Corrosion Science, vol. 30, No. 6/7, 1990, pp. 697-714.

Corrosion Research Frontiers. Atmospheric Corrosion in Tropical Climate… [27] [28] [29] [30] [31] [32] [33] [34]

91

M. Stratmann, H. Streckel, K. T. Tim, S. Crockett, Corrosion Science, vol. 30, No. 6/7, 1990, pp. 715-734. ISO 11844-3.-Corrosion of metals and alloys. Classification of corrosivity of indoor atmospheres. Measurement of environmental parameters affecting indoor corrosivity. ISO 11844-1.-Corrosion of metals and alloys. Classification of corrosivity of indoor atmospheres. Determination and estimation of indoor corrosivity. J.P. Cai, S.B. Lyon, Corrosion Science, vol. 47, (2005) 2956-2973. Seon Yeob Li, Young-Geun Kim, Sungwon Jung, Hong-Seok Song, Seong-Min Lee, Sensors and Actuators B, vol. 120, 10 January 2007, pp. 368-377. F. Corvo, J. Minotas, J. Delgado, C. Arroyave, Corrosion Science, vol. 47, 2005 p. 883-892. ISO-9225: 1992. Corrosion of metals and alloys. Corrosivity of atmospheres. Measurement of pollution. F. Corvo, R. Marbot, Revista CENIC (Ciencias Químicas), vol. 15, No.1, 1984

In: Electroanalytical Chemistry: New Research Editor: G. M. Smithe

ISBN: 978-1-60456-347-4 © 2008 Nova Science Publishers, Inc.

Chapter 3

THE APPLICATION OF D-STATISTICS BASED TESTS OF RANDOMNESS, INDEPENDENCE AND TREND TO ELECTROCHEMICAL OBSERVATIONS Thomas Z. Fahidy Department of Chemical Engineering, University of Waterloo, Waterloo Ontario, Canada, N2L 3G1

Abstract Randomness, independence and trend (upward, or downward) are fundamental concepts in a statistical analysis of observations. Distribution-free observations, or observations with unknown probability distributions, require specific nonparametric techniques, such as tests based on Spearman’s D – type statistics (i.e. D, D*, D**, Dk ) whose application to various electrochemical data sets is herein described. The numerical illustrations include surface phenomena, technology, production time-horizons, corrosion inhibition and standard cell characteristics. The subject matter also demonstrates cross fertilization of two major disciplines.

List of Symbols A. B A); B* b D

ordinal form of an observation set; A* same with rank ties ordinal form of an observation test (to be tested for association/independence vis-à-vis set same with rank ties total number of blocks in two-factor analysis Spearman’s D-statistic, defined by Eq.(2); D* the same with rank ties, defined by Eq.(8)

D** d Ecorr

D-statistic extended to ordinal forms of two observation sets, defined by Eq.(11) numerical value of D corrosion potential

94

Thomas Z. Fahidy

e EH EMF FM FN g H i,j icorr k N NPS PH RBE Ri

total number of tied ranks in rank array A expectation of a random parameter related to hypothesis H electromotive force Friedman’s statistic, defined by Eq.(14), without rank ties cumulative normal (Gaussian) distribution function total number of tied ranks in rank array B hypothesis of lack of association, dependence or trend; HA alternative hypothesis array element index corrosion current density total number of treatments in two-factor analysis size of an observation set nonparametric statistics probability of a statement or event related to hypothesis H randomized block experiment (in conventional ANOVA)

rs S2Bj b SEC Ti VH χα2(ν)

Spearman’s rank correlation coefficient sample variance of the adjusted ranks within blocks j = 1,…, specific energy consumption i-th element in an observation set; Ti* same with rank ties; ti its numerical value variance of a random parameter related to hypothesis H chi-square statistic, critical at probability α, with degree of freedom ν

sum of the i-th treatment rank in two-factor analysis; Ri mean rank in the i-th treatment

I. Introduction Nonparametric statistics (NPS) differs primarily from its traditional, distribution-based counterpart by dealing with data of unknown probability distributions. Its principal attractiveness lies, in fact, in not requiring the knowledge of a probability distribution. NPS is especially inviting when the assumption of normal distribution of small data sets is hazardous (if at all admissible), even if NPS-based calculations are more time consuming than in traditional statistics. The steadily growing importance of NPS has been amply demonstrated by numerous textbooks and monographs published within the last few decades, e.g. [1-7]. Exploration of the scope of NPS in electrochemical science and engineering has so far been rather limited. The estimation of confidence intervals of population mean and median, permutation-based approaches and elementary explorations of trends and association involving metal deposition, corrosion inhibition, transition time in electrolytic metal deposition processes, current efficiency, etc.[8] provides a general framework for basic applications. Two-by-two contingency tables [9], and the analysis of variance via the NPS approach [10] illustrate two specific areas of potential interest to electrochemical process analysts. The current subject matter deals with randomness, independence and trend concerning small sets of electrochemical observations or measurements, whose probability distribution is unknown, and the assumption of an even approximately normal distribution would be statistically unsound. Under such circumstances the theoretical null-distribution related to the hypothesis H of lack of randomness, independence and lack of trend, has to be established from the data themselves on the basis of equal probability of all possible data arrangements.

The Application of D-Statistics Based Tests of Randomness, Independence…

95

In a conventional sense, rejection of hypothesis H is significant at the 5%, and highly significant at the 1% level, i.e. at a 95% and 99% level of confidence, respectively. The required tests are applied to ranks, or differences in ranks, instead of numerical observations; this is a fundamental characteristic of nonparametric versus classical statistics. Electrochemical applications are grouped with respect to each test, and are presented in combination with pertinent theory for immediate illustration. Only a modicum of understanding of critical normal-, and chi-square tables, and analysis of variance techniques, is required as a necessary background.

II.The Hypothesis of Randomness [11], and Testing Independence Against Trend [12] II.1 Elementary Analysis Based on the D- and the D*- Statistic The first illustration is fashioned after Problem 29 [13] of iron corrosion in deoxygenated aqueous acid. It is assumed that four identically shaped iron-containing specimens labeled A, B, C, D are chosen randomly, and inserted in the acid carrying a corrosion inhibitor. The percentage reduction in the specific rate loss is 93.3(A), 97.3(B), 96.7(C) and 90.0(D). As shown in Table 1, the hypothesis H of no treatment effect (i.e. the hypothesis of all rankings being equally possible) can be stated as Table 1. Establishment of the null distribution of ranks and the D-statistic in testing the hypothesis of upward trend in the rate-loss reduction data in Section II. d 0 2 4 6 8 10

P[D = d] 1/4! 3/4! 1/4! 4/4! 2/4! 2/4!

P[D ≤ d] 1/4! = 0.0147 4/4! = 0.1667 5/4! = 0.2083 9/4! = 0.3750 11/4!=0.4583 13/4!=0.5417

d 12 14 16 18 20

P[D = d] 2/4! 4/4! 1/4! 3/4! 1/4!

P[D ≤ d] 15/4!=0.6250 19/4!=0.7917 20/4!=0.8333 23/4!=0.9584 24/4! = 1

Ranking Sample rank, i 1 2 3 4 Observation rank, Ti 2 4 3 1 Assignment possibilities First positions: 1 2 3 4 Second positions: 2 3 4 1 3 4 1 2 4 1 2 3 Third positions: 34 24 23 34 14 13 24 14 12 23 13 12 Final assignments: 1234;1324;1423;2134;2314;2413;3124;3214;3412;4123;4213;4312 1243;1342;1432;2143;2341;2431;3142;3241;3421;4132;4231;4321 Total number of equally possible positions: N = 4! = 24 Null distribution of ranks: 1/4! Significance test D = (2-1)2 + (4-2)2 + (3-3)2 + (1-4)2 = 14; Table N [4;p.433]: P[D ≤ 14;N = 4] = 0.7919 Normal approximation: P[D ≤ 14;N = 4] ≈ FN [(14-10)/5.773] = FN(0.69) = 0.7549 (with continuity correction: FN[(15-10)/5.773] = FN(0.87) = 0.8078) Spearman D-statistic probabilities (N = 4)

96

Thomas Z. Fahidy i Ti

1 3 2 1 1

2 1 3 3 4

3 2 1 2 2

4 4 4 2 3

Illustration for d = 6 In all four cases the sum of the squared differences (Ti – i)2 is 6. Hence, P[D = 6] = 4(1/4!) = 1/6

PH (T1 = t1 ; T2 = t 2 ;..., TN = t N ) =

1 1 = N ! 4!

(1)

in terms of the probability that the ranks Ti take on any particular set of values ti ,i = 1,…,N, respectively. This is also the desired null distribution of the ranks, expressing formally the hypothesis of randomness. The alternative (or counter) hypothesis HA of upward trend can be tested via the numerical value of the Spearman D-statistic N

d = ∑ (t i − i ) 2

(2)

i =1

by determining the cumulative probability P[D ≤ d]. As shown in Table N [4,p.433], there exists an approximately 79% probability that the hypothesis of randomness can be sustained. The normal distribution-based approximation, correct in a strict sense for large observation sets:

P[ D ≤ d ; N ] ≅ FN (

d − EH VH

)

(3)

with expectation

EH ≡

N3 − N = 10 6

(4)

and variance

( N − 1) N 2 ( N + 1) 2 VH ≡ ≈ 33.33 36

(5)

yields a somewhat lower probability of about 75%. If a continuity correction is applied, i.e. (d + 1/2) is used in Eq.(3) instead of d, the probability is about 78%.The alternative hypothesis of upward trend cannot be supported. The next illustration portrays the high level of cautiousness carried by the testing-againsttrend approach. The case in point is the production of caustic soda in seven major industrial

The Application of D-Statistics Based Tests of Randomness, Independence…

97

countries in the years 1970, 1975 and 1979 [14], with ranking distributions assembled in Table 2. Since in the presence of ties, the modified D-statistic D* related to the Ti* ranks: Table 2. D-statistic analysis of the ranking distribution of caustic soda production in seven major industrial countries, based on tonne per annum data [14] in brackets Country USA West Germany USSR Japan France Canada Italy

1 (1970) 2 (9.2) 1 (1.7) 1 (1.8) 1 (2.6) 1.5 (1.1) 2 (0.9) 1.5 (1.0)

2 (1975) 1 (8.4) 2 (2.5) 2 (2.4) 2.5 (2.9) 1.5 (1.1) 1 (0.8) 1.5 (1.0)

3 (1979) 3 (11.1) 3 (3.4) 3 (3.0) 2.5 (2.9) 3 (1.4) 3 (1.1) 3 (1.1)

Case A: USA Canada D = (2-1)2 + (1-2)2 + (3-3)2 = 2 P[D ≤ 2; N = 3] = 3/3! = 0.5 Case B: West Germany USSR D = (1-1)2 + (2-2)2 + (3-3)2 = 0 P[D ≤ 0; N = 3] = 1/3! = 0.167 Case C: France Italy D = (1.5-1)2 + (1.5-2)2 + (3-3)2 = 0.5 Case D: Japan D = (1-1)2 + (2.5-2)2 + (2.5-3)2 = 0.5 EH = (33-3)/6 – (23-2)/2 = 3.5 VH = [(2)(32)(42)/36][1-(23-2)/(33-3)] = 6 P[D ≤ 0.5] ≈ FN[(1.0-3.5)/√6] = FN(-1.02) = 0.1539 N

D * = ∑ (Ti* − i ) 2

(6)

i =1

can also have a fractional value {computed as d* = ∑(ti* - i)2 }, Table N [4, p.233] provides probabilities for the occurrence of integer values of D* flanking the computed value. The appropriately modified normal approximation with parameters

N3 − N 1 g EH (D ) ≡ − ∑ (d 3j − d j ) 6 12 j =1 *

(7)

and g

( N − 1) N ( N + 1) VH ( D * ) = [1 − 36 2

2

∑ (d j =1

3 j

−dj)

N3 − N

]

(8)

provides the probability estimate via Eq.(3). The obvious upward trend in Case B is indicated to be not significant at an about 17% probability of D = 0. At N = 4 the probability is significant, whereas at N > 4, it is highly significant (0.0417 at N = 4; 0.0083 at N = 5; 0.0014 at N =6, etc.), with gradually decreasing significance probabilities. As shown in Section III.3, a much more realistic result is obtained by appropriate two-factor analysis, employing Friedman’s FM-statistic.

98

Thomas Z. Fahidy

II.2. Analysis based on the Dk – Statistic The approach applies to repeated observations pertinent to each treatment, or to observations in two-dimensional arrays, if one factor (which could serve as the block in a two-factor analysis) is considered as a replicate entity. The idea is illustrated in the case of permeation fluxes through a membrane [15] with pertinent calculations given in Table 3. If the effect of the k = 6 salts is considered as a replicate (random) phenomenon, and the observed and calculated values (N=2) are ranked according to the flux observations, the probability computed by an appropriate modification of Eq.(3) as Table 3. Dk – analysis of the ranking distribution of observed and theoretical permeation fluxes. The bracketed numbers are reported flux values [15] Salt LiCl NaCl KCl MgCl2 CaCl2

1: observed 2 (37.7) 2 (6.60) 1 (3.28) 1 (5.61) 1 (3.16)

2: theoretical 1 (31.5) 1 (6.42) 2 (3.53) 2 (5.70) 2 (3.19)

Dk (2-1)2 + (1-2)2 = 2 (2-1)2 + (1-2)2 = 2 (1-1)2 + (2-2)2 = 0 (1-1)2 + (2-2)2 = 0 (1-1)2 + (2-2)2 = 0

∑Dk = 4; E(∑Dk) = 5(23-2)/6 = 5; V(Dk) = 5(1)(22)(32)/36 = 5 P[∑Dk ≤ 4] ≈ FN[(4.5-5)/√5] = FN(-0.22) = 0.4129

P[ D ≤ ∑ Dk ] ≅ FN (

∑D

k

− E H ( ∑ Dk )

V H ( ∑ Dk )

(9)

with parameters

E H (∑ Dk ) ≡ k

N3 − N 6

(10)

and

( N − 1) N 2 ( N + 1) 2 V ( ∑ Dk ) ≡ k 36

(11)

is not significant. If the salt effect is a-priori assumed to be non-random, this case is an obvious candidate for two-factor analysis. So is the similar case of the corrosion [16] of Al2.5Mg alloy in 0.5 mol dm-3 NaCl solution, in the presence of inhibitor sinapic acid(cinamic acid, 3,5-dimethoxy-4-hydroxy-sinapine), portrayed in Table 4. The hypothesis of no upward trend in corrosion potential (i.e. no significant rpm effect) cannot be rejected, if the column entries are considered simply as random replicates. By contrast, upward trend in corrosion current is highly significant. An identical result can be obtained for inhibition efficiency, not discussed here.

The Application of D-Statistics Based Tests of Randomness, Independence…

99

Table 4. Dk – based analysis of the ranking distribution of the corrosion potential (Ecorr) and corrosion current density (icorr) of an Al – 2.5Mg alloy. The brackets contain mV and μA cm-2 values, respectively [16] (a)Corrosion potential Inhibitor conc’n mol dm-3 0 1 10 50 100 500

Rotation speed, rpm 1 0 2 150 3 2000 1 (0.771) 3 (0.776) 2 (0.773) 1 (0.767) 2 (0.768) 3 (0.771) 2 (0.770) 1 (0.768) 3 (0.773) 1 (0.771) 2 (0.775) 3 (0.776) 3 (0.800) 1 (0.772) 2 (0.773) 3 (0.820) 2 (0.806) 1 (0.774)

Dk (&) 2 0 2 0 6 8

(&) computation pattern as in Table 3 ∑Dk = 18; EH(∑Dk) = 6(33-3)/6 = 24; VH(∑Dk) = 6[(2)(32)(42)/36] = 48 P[∑Dk ≤ 18] ≈ FN[(18.5-24)/√48] = FN(-0.79) = 0.2148

(b) Corrosion current density Inhibitor conc’n mol dm-3 0 1 10 50 100 500

Rotation speed, rpm 1 0 2 150 3 2000 1 (1.44) 2 (2.28) 3 (2.84) 1 (0.86) 2 (1.63) 3 (2.32) 1 (0.72) 2 (1.40) 3 (2.10) 1 (0.62) 2 (1.22) 3 (1.93) 1 (0.51) 2 (1.02) 3 (1.68) 1 (0.38) 2 (0.79) 3 (1.40)

Dk (&) 0 0 0 0 0 0

(&) computation pattern as in Table 3 ∑Dk = 0; EH(∑Dk) = 24; VH(∑Dk) = 48; P[∑Dk = 0] ≈ FN[(0-24)/√48] = 2.7x10-4

II.3. Analysis Based on the D**-Statistic The D-statistic concept can be readily extended to the two-factor case, where (Ai ; Bi) are corresponding ranks, i.e. if A and B are the ordinal form of observation sets (the A,B notation is used instead of Lehman’s (R,S) notation in order to avoid confusion between similar symbols). If rank positions are tied, they are replaced by their mid-rank, yielding modified distributions A* and B*. By a straightforward extension of Eq.(6), the modified D-statistic is computed as [17] N

D ** ≡ ∑ ( Ai* − Bi* ) 2

(12)

i =1

If N is sufficiently large, the usual normal approximation applies with parameters

N3 − N 1 e 1 g 3 EH (D ) ≡ − ∑ (d j − d j ) − ∑ ( f j3 − f j ) 6 12 j =1 12 j =1 **

(13)

100

Thomas Z. Fahidy

and e

( N − 1) N ( N + 1) VH ( D ) ≡ [1 − 36 2

**

2

∑ (d j =1

3 j

g

−dj)

N3 − N

][1 −

∑( f j =1

3 j

− fj)

N − N3

]

(14)

where dj and fj are the number of mid-ranks in set A and B in increasing order, their total numbers being e and g, respectively. The first electrochemical application of the D**-statistic deals with the lack- ofassociation (i.e. independence) hypothesis concerning current efficiency and current load in diaphragm-type industrial scale chlor-alkali cells [18]. Table 5 demonstrates that the two factors are independent with the understanding that the current efficiency/current load relationship may indirectly be influenced by other technical variables, e.g. cell potential, and impurities. Table 5. Testing the independence of current efficiency and current load via eight different diaphragm-type chlor-alkali cells [18] Cell type DS-31 DS-45 D8-85 H-2A H-4 V-1144 B-40 DS-7

Current efficiency (%) Observation Rank 96.5 5 96.5 5 96.5 5 96.4 3 96.6 7 97 8 95 1 96 2

Current load (kA) Observation Rank 40 1.5 80 4 150 7.5 80 4 150 7.5 80 4 40 1.5 75 6

D** = (1.5-5)2 + (4-5)2 + (7.5-5)2 + (4-3)2 + (7.5-7)2 + (4-8)2 + (1.5-1)2 + (6-2)2 = 53 e = 1; d1 = 3 (i.e. 5;5;5); 1 – (33-3)/(83-8) = 0.9523 g = 3; f1 = 2 (i.e.1.5;1.5); f2 = 3(i.e. 4;4;4); f3 = 2(i.e.7.5;7.5) 1 – [(23-2) + (33-3) + (23-2)]/(83-8) = 0.9286 EH(D**) = (83-8)/6 – [(33-3) + (23-2) + (33-3) + (23-2)]/12 = 79 VH(D**) = [(7)(82)(92)/36](0.9523)(0.9286) = 891.353 P[D** ≤ 53] ≈ FN[(53.5-79)/√891.353] = FN(-0.85) = 0.1977 Table N [4;p.433]: P[D ≤ 52;N = 8] = 0.1799; P[D ≤ 54;N = 8] = 0.1947

Analysis of the impedance parameters/potential sweep rate observations [19], summarized in Table 6, demonstrates significant dependence only in the case of the oxide film resistance where a negative trend is evident. The probability, 1 – P[D** ≤ 38.5] = 1 0.9744 = 0.0256 is significant by the normal approximation, whereas Table N {4;p.433] indicates simply that the probability is less than 0.5250, a quantitatively “puny” result.

The Application of D-Statistics Based Tests of Randomness, Independence…

101

Table 6. Testing the independence of impedance parameters and potential sweep rate in surface oxide formation [19]. The observed values are in brackets Sweep rate (mV s-1)

Electrolyte Resistance (Ω cm2)

Surface/species Parameter (106Ω-1sncm-2)

Constant.phase/ element exponent

1 (10) 2 (20) 3 (30) 4 (40) 5 (50) D** EH(D**) VH(D**) z (#) FN(z) Table N

1 (26.4) 4.5 (26.9) 4.5 (26.9) 2 (26.5) 3 (26.8) 16.5 19.5 95 - 0.308 0.3783 0.392-0.475

1 (5.36) 2 (5.37) 4 (5.40) 5 (5.41) 3 (5.39) 6 19.5 100 - 1.30 0.0968 0.117

4 (0.95) 1.5 (0.94) 4 (0.95) 4 (0.95) 1.5 (0.94) 22.5 19.5 75 0.346 0.6368 > 0.525

Oxide filmResist ance (kΩ cm2) 5 (320) 4 (316) 2.5 (315) 2.5 (315) 1 (311) 38.5 19.5 95 1.949 0.9744 > 0.525

Ox.film thickness (nm) 4.5 (1.98) 4.5 (1.98) 2.5 (1.97) 1 (1.96) 2.5 (1.97) 34 19.5 90 1.581 0.9429 > 0.525

(#) z ≡ [(D** - EH(D**)]/√VH(D**)

Proof for an intuitively correct inference is provided by the treatment of electromotive force versus CdSO4 concentration data for a standard cell [20] at 25 0C, shown in Table 7. Dependence is highly significant, as expected. If the dimensionless potential is defined alternatively as (EMF – EMF0 )/EMF0 , D** = (1-10)2 + (2-9)2 +…+(10-1)2 = 330 yields 1 – FN(3) = 0.0013, and the same conclusion is reached. Table 7.Rank distribution analysis of potential versus cadmium sulfate content in a standard cell at 25 0C [20]. The bracketed numbers are observed values; EMF* = 1.018399 V. Cadmium sulfate content (%) 1 (41.84) 2 (42.39) 3 (42.63) 4 (42.77) 5 (42.90) 6 (42.94) 7 (43.06) 8 (43.12) 9 (43.22) 10 (43.35)

104 (EMF* - EMF)/EMF* 1 (-28.682) 2 (-16.045) 3 (-12.657) 4 ( -9.122) 5 ( -7.207) 6 ( -6.451) 7 ( -4.851) 8 ( -3.142) 9 ( -2.121) 10 (0)

D** = (1-1)2 + (2-2)2 = 0; EH(D**) = (103 – 10)/6 = 165; VH(D**) = [(9)(102)(112)]/36 = 3025 P[D = 0;N =10] ≈ FN[(0 – 165)/√3025] = FN(-3) = 0.0013 Table N [4;p.433]: P[D = 0;N=10] = 0.0000

102

Thomas Z. Fahidy

Table 8.Rank distribution analysis of the SEC versus current density, cell potential, and degree of efficiency in industrial alkaline bipolar water electrolyzers at 80 0C electrolyte temperature and 1 bar pressure [21]. The bracketed numbers are observed values. Electrolyzer type ABB De Nora Hydrotechnique Norsk Hydro

Current density (kA) 2 (2) 1 (1.5) 3 (2.5) 4 (3)

D** EH(D**) VH(D**) Qualitative inspection z& FN(z) Table N [4;p.433] &

Cell potential (V) 4 (2.1) 2.5 (1.9) 2.5 (1.9) 1 (1.8)

SEC/current density 13.5 59.5 30 Slightly upward - 1.533 0.063 N/A

Degree of efficiency (%) 1 (72) 3 (80) 2 (77) 4 (82)

SEC/cell potential 0 59 27 Upward - 11.355 3.26x10-7 P[D = 0] = 0.0417

SEC (kWh m-3) 4 (4.9) 2.5 (4.6) 2.5 (4.6) 1 (4.3) SEC/degree of effn’cy 18.5 59.5 30 Downward - 7.49 1 – FN(-z) = 1.31x10-7 P[D ≥ 10] < 0.5417

z ≡ [D** - EH(D**)]/√VH(D**)

Table 8 depicts tests against trend involving current density, cell potential, degree of efficiency, and specific energy consumption, the latter considered as the trend variable. Critical D** data lead to conservative predictions, but the normal approximation indicates a (i) strongly upward SEC/cell potential, and a strongly downward SEC/degree of efficiency trend, and (ii) a not significant SEC/current density trend, although FN = 0.06 is only slightly larger than the significant 0.05 value. On account of symmetry, the distributions in Table N [4, p.433] are symmetric around (N3-N)/6, and FN(z) = 1 – FN(-z), hence the testing of downward trend is essentially the “mirror image” of the upward trend test. Other entries in [21] have been excluded due to their different operation protocols (cell type, pressure) in order to avoid “dilution” due to too many (possibly interacting) variables.

III. Discussion III.1 Comparison with Nonparametric Analysis of Variance A close inspection of Table 4 raises the question whether inhibitor concentration should be considered as a block factor, preferably requiring a nonparametric randomized complete block experiment approach. Employing for the corrosion potential the Friedman statistic, e.g. [22-24]:

FM =

k 12 Ri2 − 3b(k + 1) ∑ bk (k + 1) i =1

(15)

The Application of D-Statistics Based Tests of Randomness, Independence…

103

in the absence of rank ties, and in view of b = 6, k = 3, R1 = R2 = 11, and R3 = 14, the computed FM = 1 is well below even the 10% critical value 5.333 [25], as well as the approximate critical χ20..5(2) = 1.386 (the Friedman statistic has an approximately chi-square distribution). The blocking effect, therefore, is insignificant. As for the corrosion current, since R1 = 6; R2 = 12; R3 = 18, the computed FM = 12 is highly significant, since it is also the 0.1% critical value; alternatively, χ2(2) = 12 falls between the 0.5% critical value 10.60 and the 0.1% critical value 13.82. Similarly, in the case of the membrane permeation flux (Table 3), with b = 5, k = 2, R1 = 7, R2 = 8, FM = 0.2 being well below even the 25% critical χ2(1) = 1.32, the blocking effect is found to be insignificant.

III.2. Comparison with Spearman’s Rank Correlation Coefficient Apparently the oldest nonparametric measure of the strength of association between two factors [26], the Spearman rank correlation coefficient

rs = 1 −

6D N3 − N

(16)

yields rs = 1 when D = 0 (fully upward trend, regardless of the scatter of data!), and it approaches rs = -1 for large values of D, i.e. when the downward trend is very strong. Although rs may be regarded as an “…equivalent test statistic to D…” [27], and it yields qualitatively identical results with D – type tests, as demonstrated in Table 9, it does not provide a quantitative measure of H-hypothesis significance. Certain significances tests related to rs are described briefly in the Appendix. Table 9.Comparison of (i) D-statistic based decisions on the hypothesis of independence 9 Chlor-alkali cells Water electrolyzers SEC/current density SEC/cell potential SEC/degree of efficiency Standard cell

D – statistic Not significant Not significant Highly significant Highly significant

Rank correlation coefficient rs = 0.369 ;weak rs = - 0.35;weak rs = 1; very strong rs = - 0.85;strong

Highly significant

rs = -1;very strong

(i.e. lack of association) and (ii) decisions based on Spearman’s rank correlation coefficient

If the two-factor cases considered here were known to originate (at least approximately) from a normal population, the ‘standard” randomized block experiment approach would be admissible for testing the significance of the block effect. A detailed discussion of this technique, widely documented in the statistical textbook literature, is omitted. Table 10 indicates the possibility of drawing qualitatively identical inferences from nonparametric and conventional analysis of variance, even if only one of the two is correct, in principle.

104

Thomas Z. Fahidy

Table 10.Comparison of two-factor analysis via Friedman’s statistic and conventional randomized block experiment in (parametric) ANOVA Case Corrosion of Al2.5Mg alloy Corrosion potential Corrosion current Inhibition efficiency Membrane permeation flux Caustic soda production

Friedman’s test FM = 1; 50% critical FM =1.389: NS FM = 12; 0.1% critical FM =12.00: HS FM = 10; 0.1% critical FM =10.00; HS FM = 0.2; 25% critical χ2 =1.32: NS FM’ = 9.92; 1% critical FM =8.857: HS

RBE* P-value: 0.38 for treatments; 0.07 for blocks: NS P-value: 2x10-9 for treatments; 2x10-7 for blocks: HS P-value: 2x10-9 for treatments; 2x10-9 for blocks: HS P-value: 0.39 for treatments; 3x10-4 for blocks: HS& P-value: 2x10-9 for treatments; 0.03 for blocks: HS

NS: not significant; HS highly significant P-value: the probability of an observation-based statistic being larger than the computed value & the block effect is highly significant FM’: modified Friedman statistic for ties in ranks, defined as

FM ' =

b2 b

∑S j =1

*

k

∑ (R

2 i =1 Bj

i

−

k +1 2 ) 2

from ANOVA tables in EXCEL (Windows XP Professional), for randomized block experiments. Blocks are from top down: sinapic acid; salt types; major producers.

III.4. Fallacious use of Tests for Trend A major fallacy is made when observations obeying a known physical law are subjected to trend-oriented tests, but without allowing for a specific behaviour predicted by the law in certain sub-domains of the observation set. This can be seen in Table 11 where a partial set of classical cathode polarization data has been reconstructed from a current versus total polarization graph [28]. If all data pairs were equally treated, rank distribution analysis would lead to an erroneous conclusion, inasmuch as the (admittedly short) limiting-current plateau for cupric ion discharge, albeit included in the data, would be ignored. Along this plateau, the independence of current from polarization potential follows directly from the theory of natural convection at a flat plate, with ample empirical support from electrochemical mass transport experiments. Table 11.Analysis of association based on a partial set of cathode polarization data from experiments with an aqueous 0.732 mol dm-3 CuSO4; 1.484 mol dm-3 H2SO4 solution, reconstructed from Fig.6 [27] Total cathodic polarization Observation (mV)& Rank 100 1 150 2 200 3 250 4

Cathode current density Observation ( mA cm2 )& Rank 6.9 1 16.1 2 34.3 3 40.6 4

The Application of D-Statistics Based Tests of Randomness, Independence…

105

Table 11. Continued Total cathodic polarization Observation (mV)& Rank 300 5 350 6 400 7 # 483 8 # 648 9 # 700 10 730 11

Cathode current density Observation ( mA cm2 )& Rank 50.0 5 53.0 6 54.3 7 55.2 8 56.6 9.5 56.5 9.5 58.7 11

&

experimental points read from an appropriate enlargement of the mA cm-2/mV graph # pertaining to the limiting current plateau Rank distribution analysis: D** = (9 - 9.5)2 + (10 - 9.5)2 = 0.5; EH(D**) = (113 – 11)/6 – (23 – 2)/12 = 219.5 VH(D**) = (10)(112)(122)[1 – (23 – 2)/(113 – 11) = 4818.22 FN[(1.0 – 219.5)/√4818.22] = FN(-3.15) = 0.0008 Table N [4, p.433]: P[D = 0; N = 11] = 0.0000 P[D ≤ 1; N = 11] = 0.0000 Conclusion: highly significant association Conclusion drawn from physical considerations: The steadily rising current density reaches a limiting current plateau in the 483 – 700 mV polarization range, prior to a further increase at higher polarization causing proton discharge at the cathode.

More generally, testing for a trend is a self-defeating exercise, if the observation sets are not connected by causality. This cardinal rule, applying equally to parametric and nonparametric tests, has been amply emphasized in the literature.

III.5. Alternative Approaches to Testing Randomness, Independence, Trends and Association While D-statistic based analyses are arguably straightforward, other alternatives, based e.g. on contingency tables [29,30] are also useful means of analysis, but beyond the scope of the current presentation. Their potential usefulness to electrochemical systems, not repeated here, has recently been documented [8].

IV. Conclusion The preceding applications furnish a small, albeit representative sample of a nonparametric treatment of electrochemical observations when their probabilistic properties are unknown, or if no specific a-priori probability distribution can be associated with them. D-statistic based techniques have much to offer to the electrochemical process analyst, but a full exploration of this useful tool remains a subject of future research.

106

Thomas Z. Fahidy

Acknowledgments Facilities for this work were provided by the University of Waterloo and the Natural Sciences and Engineering Research Council of Canada (NSERC).

Appendix A Brief Summary of Some Significance Tests Related to Spearman’s Rank Correlation Coefficient The Spearman-Hotelling-Pabst test may be used for determining if a D-type statistic in Eq.(15) is significant, by comparing the computed D, D*, etc. to lower quantiles of the statistic, tabulated e.g. in [31]. The diaphragm-type chlor-alkali cells in Section II.3 with rs = 0.369 (Table 9) and D** = 53 (Table 5) are taken for illustration. The current efficiency observations suggesting an upward trend with current load, acceptance of the counterhypothesis of significant positive correlation requires that D** = 53 be less than the 5% lower quantile. The latter being 32 at N = 8, the hypothesis H of no positive correlation cannot be rejected (it cannot be rejected even at a 10% level with lower quantile 42). In an alternative approach, critical values of Spearman’s rank correlation coefficient, tabulated as a function of observation size and significance probability [32], are employed. The 5% critical value is 0.738 when N = 8, hence rs = 0.369 is not significant (not even at the 10% level, with critical value 0.643). If the observations suggest a negative correlation, the computed D-type statistic is to be larger than the critical upper quantile, computed as N(N2 – 1)/3 minus the critical lower quantile, for rejecting H: insignificant negative correlation. In Case A of water electrolysis (Table 9) with N = 4, the 5% upper quantile is 4(42 – 1)/3 – 2 = 18, and D** being 13.5, H cannot be rejected. In Case C, D** = 18.5 exceeds 18 slightly, hence negative association may be considered significant at the 5% level, although not strongly. The critical rs – tabulation does not carry entries for N < 5, but they are expected to be near unity.

References [1] [2] [3] [4] [5] [6] [7]

Kraft, C. H., van Eden, C; A Nonparametric Introduction to Statistics; Macmillan: New York, NY,1968. Mosteller, F., Rourke, R.; Sturdy Statistics; Addison-Wesley:Reading, MA, 1973. Hollander, M., Wolfe, D.; Nonparametric Statistical Methods; Wiley: New York, NY, 1973. Lehman, E.; Nonparametrics: Statistical Methods Based on Ranks; Holden Day: San Francisco, CA, 1975. Daniel, W.; Applied Nonparametric Statistics; Houghton, Miffin: Boston, MA, 1978. Conover, W. J.; Practical Nonparametric Statistics; J. Wiley and Sons: New York, NY, 3rd edn., 1999. Higgins, J. J.; An Introduction to Modern Nonparametric Statistics; Duxbury, Thomson Brooks/Cole: Belmont, CA, 2004.

The Application of D-Statistics Based Tests of Randomness, Independence… [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

[22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]

107

Fahidy, T. Z. in New Developments in Electrodeposition and Pitting Research, A. El Nemr, Transworld Research Network, Trivandrum, India, in press. Fahidy, T. Z.; On The Potential Utility of 2x2 Contingency Tables in Electrochemical Engineering, J. Appl. Electrochem., in press. Fahidy, T. Z. in Electroanalytical Chemistry Research Developments, P. N. Jiang, Nova Science Publishers: Hauppage, NY, in press. Lehman, E.; loc. cit., p. 287. Lehman, E.; loc. cit., p. 290. Pletcher, D.; A First Course in Electrode Processes, ECC, Alresford Press: UK, 1991, pp. 246; 266-267. Hine, F.; Electrode Processes and Electrochemical Engineering, Plenum: New York, NY, 1985, Table 7.1, p.140. Oguzie, E.E., Onuchukwu, A. I., Ekpe, U. J.; J. Appl. Electrochem. 2007, 37, 10471053. Vrsavolić, L.; Kliškić, M.;Radošević, J.; Gudić, S.; J. Appl. Electrochem. 2007, 37, 325 – 330. Lehman, E.; loc. cit., p. 391. Hine, F.; loc. cit., Table 8.3, p.176. Hasenay, D., Šeruga, M.; J. Appl. Electrochem. 2007, 37, 1001-1008. Hamer, W.J. in The Primary Battery, G. W. Meise, N. C. Cahoon; J. Wiley and Sons: New York, NY, 1971, p. 454. Sandstede, G., Wurster, R. in Modern Aspects of Electrochemistry, R. E. White, J. O’M. Bockris, B. E. Conway, Plenum Press: New York, NY, Vol. 27., 1995, Table 13, pp. 440- 441. Higgins, J. J.; loc. cit., p. 132. Higgins, J. J.; loc. cit., p. 262. Devore, J. L.; Probability and Statistics for Engineering and the Sciences, Thomson Brooks/Cole: Belmont, CA, 6th edn., 2004, p. 691. Lindley, D. V.; Scott, W. F.; New Cambridge Statistical Tables, Cambridge Univ. Press: Cambridge, UK, 2nd edn., 1984, Table 24, p. 71. Spearman, C.; Am. J. Psychol. 1904, 15, 72-101. Lehman, E. L.; loc. cit., p. 300. Wilke, C. R., Tobias, C. W., Eisenberg, M.; Chem. Eng. Progr. 1953, 49, 663-674. Higgins, J. J.; loc. cit., p. 300. Higgins, J. J.; loc. cit., p. 303. Powell, F. C.; Statistical Tables for the Social, Biological and Physical Sciences, Cambridge Univ. Press: Cambridge, UK,1982, p.80. Handbook of Tables for Probability and Statistics, W. H. Beyer, CRC Press: Boca Raton, FL, 1st edn. 1966, p. 186; 2nd edn., 1968, p.446.

In: Electroanalytical Chemistry: New Research Editor: G. M. Smithe

ISBN: 978-1-60456-347-4 © 2008 Nova Science Publishers, Inc.

Chapter 4

SELF-ASSEMBLY ASSISTED POLYPOLYMERIZATION (SAAP): A NOVEL APPROACH TO PREPARE MULTIBLOCK COPOLYMERS WITH A CONTROLLABLE CHAIN SEQUENCE AND BLOCK LENGTH Liangzhi Hong1, Fangming Zhu3, Guangzhao Zhang2, To Ngai*,1 and Chi Wu1,2 1

Department of Chemistry, The Chinese University of Hong Kong Shatin, N. T., Hong Kong. 2 The Hefei National Laboratory of Physical Science at Microscale Department of Chemical Physics, University of Science and Technology of China Hefei, Anhui, China. 3 Institute of Polymer Science, School of Chemistry and Chemical Engineering Sun Yat-Sen University, Guangzhou 510275, P. R. China.

Abstract Block copolymers have attracted much attention because of their novel properties and various promising potential applications. However, it is still difficult, if not impossible, to prepare multiblock copolymers with a controllable chain sequence and block length even though a variety of synthetic methods, such as anionic and controlled free radical living polymerization have been advanced. In recent years, we have proposed and developed a novel method of using the self-assembly of A-B-A triblock copolymers in a solvent which is selectively good for the two A-blocks. Such self-assembly concentrates and exposes the active groups attached on the two A-block ends so that they can be coupled together to form a long multiblock copolymer chain with its sequence and block length controlled by the initial triblock copolymer. In this review, we first illustrate how the SAAP concept was developed and exemplified in some real copolymer systems. Furthermore, we compare the coupling efficiency with and without the self-assembly, and demonstrate that SAAP provides an elegant way to prepare long multiblock copolymers.

110

Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.

Introduction In the last two decades, block copolymers have attracted much attention in both polymer chemistry and physics because of their synthetic challenges and rich phase diagrams in bulk as well as in solution.[1-13] The preparation of well-defined and complex architectures of block copolymers requires a chain-growth polymerization mechanism that can be conducted in the absence of undesired transfer and termination steps.[14] Living polymerization, especially living anionic polymerization (LAP), is the most powerful tool for the preparation of block copolymers with designed block lengths and architectures. Chemical modification after polymerization[15-17] is one of three general strategies to overcome the primary limitation of using LAP in block copolymer synthesis; namely, the restriction to a minority of all interesting or otherwise useful monomers. Protection of the functional groups in monomers is the second approach to extend the application of LAP in the preparation of block copolymers.[18] The third strategy, combined alternative living or controlled polymerization protocols with LAP, has been successfully demonstrated.[19-25] Recently, due to the development of controlled living free-radical polymerization, using one block as a macroinitiator for another monomer has also been developed.[22, 24, 25] For preparation of multicomponent block copolymers, it is difficult, if not impossible, to sequentially polymerize more than five blocks by a single mechanism.[26] A combination of different mechanisms thereby has provided an alternative route and become more attractive.[14] In principle, one could sequentially add different types of monomers into a living anionic polymerization system to prepare long and multiblock copolymer chains. However, in reality, each addition of new monomer will inevitably terminate some of the living chain ends because of introducing the impurities,[27] and result in a broad distribution of the final block copolymer chains. Practically, it is not feasible to fractionate these multiblock chains because their physical properties, such as solubility and viscosity, are similar. On the other hand, each sequentially added monomer must be reactive to initiate the following monomers so that the chain can propagate again.[28] However, in most cases, a living A block can initiate comonomer B, but a living B block may not initiate comonomer A. This is the crucial reason why the sequential addition method can only be used to make copolymers with a few blocks, typically diblock or triblock copolymer chains.[26, 29-31] Alternatively, one could prepare multiblock copolymers by directly coupling different blocks with two reactive groups at the end, like the step-growth polymerization. Preparation of polyurethane is a typical example. However, it should be addressed that such a coupling reaction is extremely ineffective, especially when long polymer blocks (Mw > 104 g/mol) are used because most of the reactive ends are wrapped and hidden inside the long coiled blocks in solution. Moreover, the concentration of the active end is low because of the long chains used in the coupling reaction. Therefore, it still remains a challenge in polymer chemistry to synthesize long multi-block heteropolymer chains with an ordered sequence and controllable block lengths. Despite the difficulties, the significance of producing such long multiblock heteropolymers could not be overlooked because they will not only provide unique systems for the study of polymer physics, but also could lead to a new type of polymeric materials. The key to a successful preparation of long multiblock copolymer chains is how to effectively couple the active ends together. It has been well known in polymer physics that the self-

Self-Assembly Assisted Polypolymerization (SAAP)

111

assembly of block copolymer chains in a selective solvent can form different core–shell micelle-like structures.[27, 32-34] Using such self-assembly, one should be able to force the reactive ends to stick out and concentrate on the periphery of the micelle-like structure if the reactive functional groups are attached to the ends of the soluble blocks of copolymer chains. A combination of the self-assembly and the condensation polymerization (coupling) has enabled us to propose a new method, the self-assembly assisted polypolymerization (SAAP), for the preparation of long multiblock copolymers with an ordered chain sequence and controllable block lengths. Figure 1 schematically illustrates the principle of SAAP.[35]

Figure 1. Schematic of self-assembly assisted polypolymerization (SAAP) of triblock copolymers in a selective solvent for the synthesis of long multiblock copolymers with a controllable chain sequence and block length.[35]

In the past several years, we have used the SAAP method to prepare different long multiblock copolymers. Generalities about the preparation and characterization of different end-functionalized triblock copolymers are first outlined. Then, the micellization of triblock copolymer as well as coupling efficiency with and without self-assembly method are discussed.

1. End-Functionalization of Block Copolymer with Oxalyl Chloride To demonstrate this novel SAAP method, we first prepared triblock copolymer, ClOCOC-(-PMMA-b-PS-b-PMMA-)-COCOCl, with two functionalized ends via sequential addition anionic polymerization and termination with an excess amount of oxalyl chloride (ClCOCOCl).[35, 36] The weight-average molar masses (Mw) of the PMMA and PS blocks were 8.75 x 102 and 2.10 x 104 g/mol, respectively. Laser light-scattering data, as shown in Figure 2, reveals that it is soluble in a solvent mixture of methyl acetate and acetonitrile (10/1, v/v) at 45 °C.[35] As the temperature decreases, they self-assemble into a core-shell micellelike structure with a collapsed PS core and a swollen PMMA shell at the room temperature.

112

Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.

The peak in f(Rh) located at 3-4 nm represents individual triblock copolymer chains. At 29 °C, an additional peak appears indicating the self-assembly of the triblock copolymer chains. Pentanediol (HO(CH2)5OH) was added as the linking agent to couple each two functional ends of the triblock copolymer chains in the presence of pyridine. The resultant multiblock heteropolymer chains have a structure like (PMMA-b-PS-b-PMMA-c-)n, where “c” denotes the linking agent, pentanediol. The structure can also be written as (PMMA-b-PS)n, in which the PMMA block is twice longer than that in the initial triblock PMMA-b-PS-b-PMMA copolymer chain because each two PMMA blocks are connected together in the resultant multiblock copolymer. o

T = 45 C

f(Rh)

3.0 o

T = 29 C

2.0 1.0 0.0 0 10

1

2

10

10

Rh / nm

i

2

Figure 2. Typical hydrodynamic radius distributions (f(Rh)) of individual triblock PMMA-b-PS-bPMMA copolymer chain end-capped with oxalyl chloride in a solvent mixture of methyl acetate and acetonitrile (10/1, v/v) at 45 °C and the aggregates formed via the self-assembly of the triblock copolymer chains at 29 °C, where the triblock copolymer concentration is 1 x 10-4 g / mL.[35]

W(M)

1.2

0.8

0.4

0.0 3 10

4

10

5

10

6

10

7

10

M / (g/mol) Figure 3. Weight distributions of molar mass of triblock PMMA-b-PS-b-PMMA copolymer (dashed line) and long multi-block (PS-b-PMMA)n heteropolymer (solid line) measured by size exclusion chromatography.[35]

Figure 3 shows typical size exclusion chromatographs (SEC) of PMMA-b-PS-b-PMMA and multiblock copolymer (PMMA-b-PS)n.[35] The polydispersity index (Mw/Mn) of the two

Self-Assembly Assisted Polypolymerization (SAAP)

113

peaks is ~1.25. The weight-average molar mass (Mw) of the resultant multiblock copolymer is 4.72 x 105 g/mol, indicating that on average, more than 20 triblock copolymer chains are coupled together. According to a normal definition of block copolymers, this is a 40-block copolymer. The results have been confirmed by static laser light scattering. For comparison, a control experiment was performed at 45 °C at which no self-assembly of the triblock PMMAb-PS-b-PMMA chains was detected by LLS. We are not able to detect any long multi-block heteropolymer chains by both SEC and LLS after a week, indicating that the self-assembly plays a crucial role for the preparation of long multiblock copolymers.

V (a.u.)

2. End-Functionalization of Block Copolymer with Photosensitive Groups

14

1 hr UV in n-heptane. 1 hr UV in THF. without UV.

SI42

1hr UV in n-heptane. without UV.

SI44

16

18

20

22

Retention Volume / mL Figure 4. SEC profiles of SI42 and SI44 before and after 1 h UV irradiation in n-heptane as well as of SI42 after 1 h UV irradiation in THF without the self-assembly.[46]

Photodimerization of coumarin and its derivatives have been widely investigated and used.[37-45] It is known that under a UV irradiation with light wavelength longer than 300 nm, coumarin and its derivatives can undergo [2+2] dimerization to form four types of dimmers depending on the reaction conditions. Importantly, such formed photodimers can be

114

Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.

reversed when irradiated by a shorter UV light (λ < 254 nm). One might therefore expect that by terminating of each anionic end with such a photoreactive molecule, like 7chlorodimethylsilanoxy-4-methylcoumarin, we can couple two triblock copolymer chains via the [2 + 2] photodimerization through the coumarin groups.[46] Photo-coupling has exhibited special advantages over the conventional coupling reaction. For example, the curing time is fast and the rate of coupling is easily controlled. Additionally, it can avoid the problem of linking agent addition. It is helpful to note that both the addition of insufficient or excessive amount of the linking agent reduce the coupling efficiency. In order to test our hypothesis about the photo-induced coupling reaction, we started with a model precursor diblock polystyrene-b-polyisoprene copolymer which can self-assemble in n-heptane, a poor solvent for the PS block. The self-assembly assisted photodimerization of such diblock chains successfully results in novel polystyrene-b-polyisoprene-c-polyisoprene-b-polystyrene (PS-bPI-c-PI-b-PS) “triblock” copolymers (B-A-A-B). Figure 4 shows typical SEC profiles of two PS-b-PI diblock copolymers before and after the photodimerization reaction.[46] The appearance of a peak after the UV irradiation at a short retention time with a doubled molar mass indicates the formation of PS-b-PI-c-PI-b-PS copolymer chains. In contrast, the irradiation of SI 42 in THF without the self-assembly only leads to a slight increase of the molar mass. It confirms that the self-assembly helps to concentrate and expose the reactive ends of precursor diblock copolymers on the periphery of the core-shell micelles, which greatly increases the coupling efficiency of the photodimmerization reaction between two coumarin end groups.

3. End-Functionalization of Block Copolymer with 1,4Dibromobutane After demonstrating the SAAP concept in the previous sections, we moved to increase the coupling yield by modifying the two anionic ends of triblock A-B-A chains with bromobutyl groups. In comparison with the functional groups mentioned above, the bromobutyl groups are insensitive to water, carbon dioxide and oxygen so that we can exchange the good reaction solvent for both the blocks with another selective solvent for the A block, but without worrying about the deactivation of the chain ends. It is helpful to note that reactions of living anionic polymer with haloalkane have attracted much attention because such reaction can lead to the preparation of narrowly distributed end-functionalized polymer chains with a controllable chain length.[47-51] However, analogous reactions of organolithium with haloalkanes can only proceed to a certain extent with some competitive side reactions, such as β-elimination, metal-halogen interchange and dimerization via the single electron transfer reaction.[49, 50, 52-54] In order to simplify the problem and have a better understanding of the reaction mechanism, we designed a model reaction of difunctional polystyryllithium that was prepared by initiating styrene with lithium naphthalenide and then reacted with 1-bromopentane in THF at -78 ℃.[55] The reaction excludes the effect of the Wurtz-type coupling reaction, otherwise, those side reactions, such as metal-halogen interchange or single electron transfer, would lead to dimers and multi-mers in the resultant product. Figure 5 shows the SEC result of the product from the reaction of difunctional polystyryllithium and 1-bromopentane. There

Self-Assembly Assisted Polypolymerization (SAAP)

115

is no dimer and multi-mer in the final product, indicating that difunctional polystyryllithium predominantly undergoes the nucleophilic substitution reaction with 1-bromopentane, where the side reactions are not significant in THF. Many reports have shown that solvents also play a pronounceable effect on the reactions. For reactions carried out in a mixture of a hydrocarbon solvent and polar aprotic solvent, such as THF, there has nearly no side reactions.[47] In contrast, the reaction in a hydrocarbon solvent can often lead to dimerterminated living polymers.[51] The Wurtz-type coupling is likely the reaction pathway, leading to the formation of dimeric products in the reaction of difunctional polystyryllithium with dihaloalkanes, but it can be suppressed by the addition of an excess amount of dihaloalkanes. Our result shows that anionic polymerization provides a feasible route to prepare well-defined polymer chains with halogen end groups. Difunctional polystyryllithium terminated by 1-bromopentane. 4 Mn = 4.25 x 10 g/mol

V (a.u.)

Mw/Mn = 1.02

Difunctional polystyryllithium terminated by methanol as reference. 4 Mn = 4.22 x 10 g/mol Mw/Mn = 1.02

10

15

20

25

Retention Volume / mL

30 Fi

5

Figure 5. SEC profiles of difunctional polystyryllithium terminated by 1-bromopentane and methanol, respectively.

Narrowly distributed PI-b-PS-b-PI triblock copolymer chains with both of their ends capped with bromobutyl groups were prepared by sequential addition of living anionic polymerization and terminated by excess of 1,4-dibromobutane (PS block: Mw = 3.5 × 103 g/mol; PI blocks: Mw = 3.1 × 103 g/mol; Mw/Mn = 1.12; The degree of end-functionalization was 92% characterized by 1HNMR). Figure 6 shows the SEC profile of such prepared triblock copolymer chains. The small but a detectable amount (~5%) of PI-b-PS-b-PI dimers, PI-b-PS-b-PI-c-PI-b-PS-b-PI, is presumably formed via the Wurtz-type coupling reaction.

116

Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.

Before the coupling reaction, the self-assembly of PI-b-PS-b-PI triblock copolymer chains in n-hexane was investigated by LLS. Figure 7 shows typical hydrodynamic radius distributions (f(Rh)) of individual PI-b-PS-b-PI triblock chains in THF, a good solvent for both the PI and PS blocks, and the core-shell micelles formed via the self-assembly of the triblock copolymer chains in n-hexane, a solvent selectively good for the PI block. The shifting of the peak from 3.6 nm to 22.8 nm clearly reveals the self-assembly of individual triblock copolymer chains and the formation of narrowly distributed micelles. The coexistence of two well-separated peaks indicates that there is an equilibrium between unimers and micelles. 3

V (a.u.)

Unimer. Coupled in n-hexane. Coupled in THF.

5.9 x 10 g/mol

4

1.3 x 10 g/mol 4

5.7 x 10 g/mol

16

20

24

28

Retention Volume / mL Figure 6. SEC profiles of PI-b-PS-b-PI triblock copolymer chains end-capped with butyl bromide group (SI44) before and after the coupling reaction in n-hexane, with the self-assembly as well as in THF without the self-assembly.

It should be emphasized once more that the self-assembly concentrates and exposes the functional ends of precursor A-B-A triblock copolymer chains on the periphery of each resultant core-shell micelle. Figure 6 shows that in the formation of −(−PI-b-PS-b-PI−)n− multiblock copolymer chains, the efficiency of the coupling reaction of 1,4-dilithio-1,1,4,4tetraphenylbutane (DD2-) and PI-b-PS-b-PI triblock copolymer chain terminated with the bromobutyl group is significantly increased by the self-assembly, after we compare the SEC profiles of the PI-b-PS-b-PI triblock copolymer chains before and after the coupling reaction. With the self-assembly in n-hexane, the coupling results in a mixture of multi-block copolymer chains, dimers and the precursor triblock chains. The peak value (5.7 × 104 g/mol) suggested that on average, 10 triblock chains were linked together to form 21-block copolymer chains with a defined sequence and controlled block lengths. This is reasonable because the degree of the end-functionalization is ~90%, which means that on average we can only link ~10 chains together. In order to make longer multiblock copolymer chains, we will have to increase the efficiency of the end capping. In contrast, when the coupling reaction was performed in THF without the self-assembly, the SEC profile contains only two peaks, namely, a “dimer” peak with Mw ≈ 1.2 × 104 and precursor triblock copolymer peak. It shows that the coupling reaction stops after two triblock copolymer chains are linked together. This is because for longer copolymer chains (Mw > 104 g/mol), the chain ends are likely wrapped and hidden inside the chains coiled in a good

Self-Assembly Assisted Polypolymerization (SAAP)

117

solvent. Therefore, the self-assembly is the key step in the preparation of long multiblock chains in SAAP. In practice, we will have to find a better selective solvent to shift the unimer ↔ micelle equilibrium towards the micelle formation and a better method to functionalize the ends of the precursor triblock copolymer chains.

f(Rh)

12 in n-hexane. in THF.

8

4

0

0

10

1

10

2

10

Rh / nm Figure 7. Typical hydrodynamic radius distributions (f(Rh)) of individual PI-b-PS-b-PI triblock copolymer chains end-capped with butyl bromide group (SI44) in THF and their self-assembled coreshell micelle in n-hexane, where C = 1.0 × 10-2 g/mL and T = 25.0 °C.

4. End-Functionalization of Block Copolymer with Carbon Dioxide It has been known that living anionic polymerization can be terminated by the addition of some suitable additives to form active ends.[50, 56-58] The carboxylation of the carbanionic group at the end of a living polymer chain with CO2 is a classic example of such endfunctionalization.[59-63] Such a reaction can quantitatively proceed in THF at -78 oC,[60, 63] which avoids the Wurtz-type coupling side reaction in the capping of the living ends with dibromoalkanes.[55] Furthermore, the formation of a amide linkage between the carboxyl acid and amine groups has been extensively studied[64, 65] and widely used in polymer synthesis.[66-72] Again, we first synthesized living PI-b-PS-b-PI triblock copolymer chains and capped their ends with the carboxylic acid group. Using the SAAP method, we successfully coupled such triblock copolymer chains together to form long (PS-b-PI)n multiblock copolymer chains. The coupling efficiency with and without the self-assembly is also compared.[73] The carboxylic acid groups on the end of PI-b-PS-b-PI triblock copolymer chains was coupled with 1,6-hexamethylenediamine (HMDA) in the presence of 1,3-dicyclohexylcarbodiimide (DCC). Figure 8 shows that the SAAP method results in the formation of dimer chains and the coupling efficiency is close to 50%. On the other hand, there is no coupling reaction in THF. This is because the carboxylic acid end groups wrapped inside the polymer coil are not able to collide with the amine groups.

118

Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.

V (a.u.)

Triblock copolymer precursor. Coupled in hexane by SAAP. Coupled in THF.

18

20

22

24

26

28

Retention Volume / mL Figure 8. SEC profiles of the triblock copolymer precursor ended with carboxyl acid group coupled by SAAP with 1,6-hexamethylenediamine (HMDA) catalyzed by 1,3-dicyclohexylcarbodiimide (DCC) as well as coupled directly in THF as a reference. 4

Mw1 = 5.96 x 10 g/mol 6

Peak 1

V (a.u.)

Mw2 = 1.76 x 10 g/mol Peak 2

18

20

22

24

Retention Volume / mL Figure 9. SEC profiles of the triblock copolymer precursor ended with acyl chloride group coupled by SAAP with 1,6-hexamethylenediamine (HMDA) but without 1,3-dicyclohexylcarbodiimide (DCC).

In the above coupling reaction, an excess amount of DCC was added to increase the reactivity of the carboxylic acid group. The carboxylic acid groups have to be activated by the DCC first then form the amide linkage with the HMDA. In order to increase the coupling efficiency, we convert the carboxylic acid group to acyl chloride to increase its reactivity. In this way, the α,ω-diacyl chloride terminated PI-PS-PI triblock copolymer chains can be coupled by the SAAP method with HMDA without DCC. The SEC result of the coupling product of the α,ω-diacyl chloride terminated PI-b-PS-b-PI triblock copolymer chains, as shown in Figure 9, clearly reveals that ~30 triblock copolymer chains are coupled together to form long multiblock copolymer chains, PI-b-(-PS-b-PI-c-PI-b-)29-PS-b-PI, where “c” represents the linkage unit. If we treat each PI-c-PI as one PI block, the multiblock copolymer chains can be regarded as a 60-block copolymer. To the best of our knowledge, the maximum reported number of blocks of multiblock copolymers was 11 up to now.[29, 30] Our results

Self-Assembly Assisted Polypolymerization (SAAP)

119

clearly demonstrate the advantage of SAAP in the preparation of long multiblock copolymer with a controllable consequence and different block lengths.

Conclusion We have developed a novel self-assembly assisted polypolymerization (SAAP) method for the preparation of long multiblock copolymers with a controllable sequence and different block lengths. The self-assembly of precursor triblock A-B-A copolymer chains in a solvent selectively good for the A block concentrates and exposes the functional groups at the end of the A block on the periphery of each self-assembled core-shell micelle. The SAAP method enables us to synthesize long multiblock copolymer chains with its sequence and block length well controlled by the initial living polymerization. Using this novel method, we have successfully prepared 60-block copolymer chains, (PS-b-PI)30. It opens a door for further studies of effects of comonomer composition and distribution on the chain backbone on its solution and bulk properties. The success of this novel approach clearly shows how synthetic chemistry is combined with polymer physics.[74] The problem-driven synthesis will be critically important in the study of polymer physics.

Acknowledgment The financial support of the Research Grants Council of the Hong Kong Special Administration (CUHK4267/00P, CUHK 4209/99P, 2160122, 2060255), NNSFC 29974027, the special funds for Major State Basic Research Projects (G1999064800), and the CAS Bai Ren Project is gratefully acknowledged. Dr. F. M. Zhu wishes to thank Lingnan Foundation of Zhongshan University and the Postdoctoral Fellowship of the Chinese University of Hong Kong.

References [1] Discher, B. M.; Won, Y. Y.; Ege, D. S.; Lee, J. C. M.; Bates, F. S.; Discher, D. E.; Hammer, D. A. Science 1999, 284, 1143. [2] Discher, D. E.; Eisenberg, A. Science 2002, 297, 967. [3] Dubertret, B.; Skourides, P.; Norris, D. J.; Noireaux, V.; Brivanlou, A. H.; Libchaber, A. Science 2002, 298, 1759. [4] Jain, S.; Bates, F. S. Science 2003, 300, 460. [5] Jenekhe, S. A.; Chen, X. L. Science 1999, 283, 372. [6] Jeong, B.; Bae, Y. H.; Lee, D. S.; Kim, S. W. Nature 1997, 388, 860. [7] Li, Z. B.; Kesselman, E.; Talmon, Y.; Hillmyer, M. A.; Lodge, T. P. Science 2004, 306, 98. [8] Russell, T. P.; Karis, T. E.; Gallot, Y.; Mayes, A. M. Nature 1994, 368, 729. [9] Savic, R.; Luo, L. B.; Eisenberg, A.; Maysinger, D. Science 2003, 300, 615. [10] Seul, M.; Andelman, D. Science 1995, 267, 476.

120

Liangzhi Hong, Fangming Zhu, Guangzhao Zhang et al.

[11] Stoykovich, M. P.; Muller, M.; Kim, S. O.; Solak, H. H.; Edwards, E. W.; de Pablo, J. J.; Nealey, P. F. Science 2005, 308, 1442. [12] Zhang, L. F.; Eisenberg, A. Science 1995, 268, 1728. [13] Zhang, L. F.; Yu, K.; Eisenberg, A. Science 1996, 272, 1777. [14] Lodge, T. P. Macromol. Chem. Phys. 2003, 204, 265. [15] Ren, Y.; Lodge, T. P.; Hillmyer, M. A. Macromolecules 2000, 33, 866. [16] Ren, Y.; Lodge, T. P.; Hillmyer, M. A. Macromolecules 2001, 34, 4780. [17] Hillmyer, M. A.; Schmuhl, N. W.; Lodge, T. P. Macromol. Symp. 2004, 215, 51. [18] Mori, H.; Hirao, A.; Nakahama, S. Macromolecules 1994, 27, 35. [19] Lu, Z. J.; Huang, X. Y.; Huang, J. L.; Pan, G. Q. Macromol. Rapid Commun. 1998, 19, 527. [20] Ramakrishnan, A.; Dhamodharan, R. J. Macromol. Sci.-Pure Appl. Chem. 2000, 37, 621. [21] Lu, G. Q.; Jia, Z. F.; Yi, W.; Huang, J. L. J. Polym. Sci. Polym. Chem. 2002, 40, 4404. [22] Bernaerts, K. V.; Du Prez, F. E. Prog. Polym. Sci. 2006, 31, 671. [23] Miura, Y.; Yamaoka, K.; Mannan, M. A. Polymer 2006, 47, 510. [24] Yagci, Y.; Tasdelen, M. A. Prog. Polym. Sci. 2006, 31, 1133. [25] Hadjichristidis, N.; Iatrou, H.; Pitsikalis, M.; Mays, J. Prog. Polym. Sci. 2006, 31, 1068. [26] Ekizoglou, N.; Hadjichristidis, N. J. Polym. Sci. Polym. Chem. 2002, 40, 2166. [27] Hadjichristidis, N.; Pispas, S.; Floudas, G. A. Block Copolymers: Synthetic Strategies, Physical Properties, and Applications John Wiley & Sons, Inc.: Hoboken, N.J., 2003. [28] Hsieh, H. L.; Quirk, R. P. Anionic Polymerization: Principles and Practical Applications. Marcel Dekker: New York, 1996. [29] Nagata, Y.; Masuda, J.; Noro, A.; Cho, D. Y.; Takano, A.; Matsushita, Y. Macromolecules 2005, 38, 10220. [30] Wu, L. F.; Cochran, E. W.; Lodge, T. P.; Bates, F. S. Macromolecules 2004, 37, 3360. [31] Wu, L. F.; Lodge, T. P.; Bates, F. S. Macromolecules 2006, 39, 294. [32] Hamley, I. W. The Physics of Block Copolymers. Oxford University Press: New York, 1998. [33] Fredrickson, G. H.; Bates, F. S. Annu. Rev. Mater. Sci. 1996, 26, 501. [34] Riess, G. Prog. Polym. Sci. 2003, 28, 1107. [35] Wu, C.; Xie, Z. W.; Zhang, G. Z.; Zi, G. F.; Tu, Y. F.; Yang, Y. L.; Cai, P.; Nie, T. Chem. Commun. 2002, 2898. [36] Wu, C.; Xie, Z. W.; Zhang, G. Z.; Zi, G. F. Chin. J. Polym. Sci. 2001, 19, 451. [37] Kuznetsova, N. A.; Kaliya, O. L. Russ. Chem. Rev. 1992, 61, 1243. [38] Morrison, H.; Curtis, H.; McDowell, T. J. Am. Chem. Soc. 1966, 88, 5415. [39] Hoffman, R.; Wells, P.; Morrison, H. J. Org. Chem. 1971, 36, 102. [40] Muthuramu, K.; Ramamurthy, V. J. Org. Chem. 1982, 47, 3976. [41] Chen, Y.; Geh, J. L. Polymer 1996, 37, 4473. [42] Chen, Y.; Geh, J. L. Polymer 1996, 37, 4481. [43] Ngai, T.; Wu, C. Macromolecules 2003, 36, 848. [44] Jiang, J. Q.; Qi, B.; Lepage, M.; Zhao, Y. Macromolecules 2007, 40, 790. [45] Li, W. J.; Lynch, V.; Thompson, H.; Fox, M. A. J. Am. Chem. Soc. 1997, 119, 7211. [46] Zhu, F. M.; Ngai, T.; Xie, Z. W.; Wu, C. Macromolecules 2003, 36, 7405. [47] Asami, R.; Takaki, M.; Hanahata, H. Macromolecules 1983, 16, 628. [48] Hirao, A.; Hayashi, M.; Nakahama, S. Macromolecules 1996, 29, 3353.

Self-Assembly Assisted Polypolymerization (SAAP) [49] [50] [51] [52] [53] [54] [55] [56] [57]

121

Hirao, A.; Tohoyama, M.; Nakahama, S. Macromolecules 1997, 30, 3484. Hirao, A.; Hayashi, M. Acta Polym. 1999, 50, 219. Quirk, R. P.; Han, K.; Lee, Y. Polym. Int. 1999, 48, 99. Bailey, W. F.; Patricia, J. J. J. Organomet. Chem. 1988, 352, 1. Ashby, E. C.; Sun, X. J.; Duff, J. L. J. Org. Chem. 1994, 59, 1270. Tillman, E. S.; Hogen-Esch, T. E. J. Polym. Sci. Polym. Chem. 2002, 40, 1081. Zhu, F. M.; Hong, L.; Li, X.; Wu, C. Manuscript in Preparation. Szwarc, M. Nature 1956, 178, 1168. Morton, M. Anionic Polymerization: Principles and Practice. Academic Press: New YorK, 1983. [58] Jagur-Grodzinski, J. J. Polym. Sci. Polym. Chem. 2002, 40, 2116. [59] Quirk, R. P.; Yin, J.; Fetters, L. J. Macromolecules 1989, 22, 85. [60] Quirk, R. P.; Yin, J.; Fetters, L. J.; Kastrup, R. V. Macromolecules 1992, 25, 2262. [61] Pan, J.; Chen, M. F.; Warner, W.; He, M. Q.; Dalton, L.; Hogen-Esch, T. E. Macromolecules 2000, 33, 4673. [62] Choi, S. B.; Han, C. D. Macromolecules 2003, 36, 6220. [63] Ji, H. N.; Nonidez, W. K.; Advincula, R. C.; Smith, G. D.; Kilbey, S. M.; Dadmun, M. D.; Mays, J. W. Macromolecules 2005, 38, 9950. [64] Sheehan, J. C.; Hess, G. P. J. Am. Chem. Soc. 1955, 77, 1067. [65] Tang, P. Org. Synth. 2005, 81, 262. [66] Kubo, M.; Hayashi, T.; Kobayashi, H.; Tsuboi, K.; Itoh, T. Macromolecules 1997, 30, 2805. [67] Pantazis, D.; Schulz, D. N.; Hadjichristidis, N. J. Polym. Sci. Polym. Chem. 2002, 40, 1476. [68] Huang, H. Y.; Kowalewski, T.; Remsen, E. E.; Gertzmann, R.; Wooley, K. L. J. Am. Chem. Soc. 1997, 119, 11653. [69] Huang, H. Y.; Remsen, E. E.; Wooley, K. L. Chem. Commun. 1998, 1415. [70] Zhang, Q.; Remsen, E. E.; Wooley, K. L. J. Am. Chem. Soc. 2000, 122, 3642. [71] Ma, Q. G.; Wooley, K. L. J. Polym. Sci. Polym. Chem. 2000, 38, 4805. [72] Liu, G. J.; Yan, X. H.; Li, Z.; Zhou, J. Y.; Duncan, S. J. Am. Chem. Soc. 2003, 125, 14039. [73] Hong, L.; Zhu, F. M.; Li, J.; Ngai, T.; Wu, C. Manuscript in Preparation. [74] Wu, C. Chem. Eng. News 2001, 79, 197.

INDEX A AA, 37 absorption, 34 access, 32 acetate, 111, 112 acetone, 26, 54 acetonitrile, 111, 112 acid, 3, 5, 95, 98, 104, 117, 118 acidification, 34 activation, 42, 72 activation energy, 42, 72 adamantane, 27 additives, 36, 41, 48, 117 adsorption, 3, 34, 46, 64, 65, 66, 70, 87, 88, 89 adsorption isotherms, 87, 88 aerosol, 87 age, 46 agent, 9, 72, 112, 114 aggregates, 112 AIDS, 6 air, 3, 19, 34, 40, 42, 53, 62, 63, 64, 67, 70, 71, 72, 73, 75, 76, 77, 79, 89 Alberta, 43 alcohol, 27 alkali, 100, 103, 106 alkaline, 70, 83, 102 alloys, 90, 91 alternative, 36, 40, 45, 50, 94, 96, 106, 110 alternative hypothesis, 94, 96 alternatives, 23, 105 alters, 18 aluminum, 6, 65, 73, 74 amide, 3, 4, 117, 118 amine, 26, 117 amines, 34 amino, 3, 4 ammonia, 70

Amsterdam, 41 analysis of variance, 94, 95, 103 analysts, 94 analytical techniques, 6 anionic surfactant, 23 anions, 87 anomalous, 47, 55 ANOVA, 94, 104 Antarctic, 37, 63 anthropogenic, 62 application, 50, 57, 81, 93, 100, 110 aprotic, 115 aqueous solution, 10, 39, 40, 71 Arctic, 10, 63 argon, 51 Arrhenius equation, 72 artificial, 81 ash, 24 ASTM, 64, 89, 90 Atlantic Ocean, 62 atmosphere, 23, 34, 64, 71, 72, 76, 77, 78, 79, 80, 81, 87, 88 atmospheric pressure, 28 atomic force microscopy, 6 atoms, 3, 12 attention, 2, 15, 18, 109, 110, 114 attractiveness, 94 automata, 17 automation, 5 averaging, 17 avoidance, 54

B barrier, 27, 36, 53, 65 batteries, 42 behavior, 22, 25, 46, 48, 54, 75, 77, 81 Beijing, 90

124

Index

bicarbonate, 87 biological, 40 biologically, 38 biomedical, 5 bipolar, 102 blocks, 93, 94, 104, 109, 110, 111, 114, 115, 118 blood, 37 boilers, 34 Boston, 6, 106 British, 9, 89 bubble, 19, 23, 49, 53 building blocks, 13 butane, 14, 18, 20 butyl ether, 18

C cadmium, 101 calibration, 11, 14 California, 6 calorimetric method, 19 calorimetry, 20 Canada, 9, 42, 50, 93, 97, 106 candidates, 19 capacity, 10, 11, 21, 22, 33, 37, 55, 57, 87 capillary, 17, 23, 71 capital, 19, 33 carbon, 3, 4, 9, 10, 11, 12, 14, 20, 23, 34, 38, 39, 40, 43, 45, 46, 47, 48, 52, 53, 54, 58, 59, 73, 84, 114 Carbon, 1, 2, 18, 38, 47, 48, 51, 52, 58, 59, 117 carbon atoms, 14 carbon dioxide, 9, 10, 11, 12, 20, 23, 34, 38, 40, 45, 46, 47, 48, 52, 54, 58, 59, 114 Carbon nanotubes (CNTs), 1, 2 carboxyl, 117, 118 carboxylic, 5, 117, 118 Caribbean, 62, 80, 85 carrier, 25, 52 CAS, 119 casting, 3 catalysis, 2, 39 catalytic, 3 cathode, 104, 105 cathode polarization, 104 cathodic process, 64, 71 causality, 105 cavities, 9, 12, 13 CD, 42 cell, 9, 12, 13, 31, 36, 93, 100, 101, 102, 103 CGC, 26 CH4, 12, 16, 52, 54, 56 chemical, 10, 11, 14, 40, 42, 43, 87 chemical engineering, 40

chemicals, 37 chemisorption, 3 chemistry,1, 5, 110, 119 China, 49, 50, 90, 109 Chinese, 42, 49, 109, 119 chloride, 62, 73, 80, 81, 82, 85, 87, 88, 111, 112, 118 chlorine, 10 chromatography, 112 circulation, 23 cis, 51 classical, 95, 104 classification, 63, 85 clean energy, 11, 19, 40, 42 cleaning, 75, 81, 82, 87, 89 climate change, 18, 34, 56, 59 climate warming, 41 climatic factors, 64, 72 clinical, 5 closed-loop, 39 clusters, 50, 71 CNTs, 1, 2, 3, 4 CO2, 10, 20, 34, 35, 36, 38, 40, 41, 42, 46, 52, 54, 56, 58, 117 coal, 34, 42 coastal zone, 77, 87 coil, 117 Colorado, 15, 20 combustion, 34, 48 commercial, 18, 23, 38 complications, 10 components, 10, 12, 15, 19, 20, 50, 57, 66 composite, 6 composition, 19, 26, 42, 119 compounds, 12, 20, 39, 62, 73, 80, 83 compressibility, 21 compression, 36, 39 computation, 99 computer, 15, 17 Computer simulation, 41 concentrates, 109, 116, 119 concentration, 3, 16, 22, 24, 26, 34, 36, 39, 40, 46, 72, 81, 87, 88, 101, 102, 110, 112 concrete, 75 condensation, 70, 72, 75, 79, 111 conductivity, 5, 65, 88 confidence, 94, 95 configuration, 21 confusion, 99 Congress, 90 construction, 75 consumption, 15, 16, 23, 27, 29, 45 contaminants, 62, 64, 65, 70, 71, 72, 73, 76, 77, 83, 87, 88

125

Index contingency, 94, 105 continuing, 10 continuity, 95, 96 control, 9, 25, 70, 113 controlled, 3, 109, 110, 114, 116, 119 convection, 104 conversion, 16, 17, 22, 23, 29, 31, 32, 37, 53, 56 conversion rate, 31, 56 cooling, 16 copolymer, 109, 110, 111, 112, 114, 115, 116, 117, 118, 119 copolymers, 109, 110, 111, 113, 114, 118, 119 copper, 64, 73, 81, 82, 85 core-shell, 111, 114, 116, 117, 119 correlation, 66, 67, 82, 94, 103, 106 correlation coefficient, 66, 67, 94, 103, 106 corrosion, 61, 62, 63, 64, 65, 66, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 93, 94, 95, 98, 99, 102, 103 corrosivity, 64, 71, 72, 81, 91 cost-effective, 19 costs, 19, 36 cotton, 82 coupling 45, 109, 110, 111, 114, 115, 116, 117, 118 covering, 24, 29 CRC, 107 critical value, 71, 103, 106 cross fertilization, 93 cryogenic, 22, 34 crystal, 12, 14, 15, 16, 17, 20, 45, 47, 52, 53, 58 crystal lattice, 12, 14 crystalline, 9, 10, 12 crystallites, 18 crystallization, 10, 15, 17, 23, 45 crystallographic, 22 crystals, 17, 23, 25, 27, 34, 37, 43, 47, 57, 58 Cuba, 61, 62, 63, 64, 66, 67, 68, 70, 72, 75, 79, 80, 82, 86, 89, 90 curing, 114 curiosity, 9, 10 cycles, 65, 70 cycling, 3 cyclohexane, 26, 57

D Dallas, 51 data set, 93, 94 dating, 20 decisions, 103 decomposition, 10, 11, 18, 19, 22, 23, 25, 34, 36, 39, 40, 41, 42, 45, 46, 48, 57, 58 deep-sea, 45

definition, 63, 64, 65, 66, 69, 76, 88, 89, 113 degree, 17, 32, 94, 102, 103, 115, 116 Delta, 11, 42, 50 density, 17, 21, 22, 24, 26, 32, 33, 36, 37, 94, 99, 102, 103, 104, 105 Department of the Interior, 58 deposition, 3, 43, 62, 70, 71, 73, 76, 77, 79, 80, 81, 82, 83, 84, 85, 86, 88, 89, 94 deposition rate, 62, 71, 73, 77, 79, 80, 81, 82, 83, 84, 85, 86, 88 deposits, 11, 44 depression, 15, 26, 44 derivatives, 113 desalination, 11, 39, 45, 53 desert, 63 detection, 2 deviation, 86 dew, 64, 65, 66, 72, 73, 76, 79 diaphragm, 100, 106 differential scanning calorimetry, 15 diffraction, 14, 16, 18, 24, 33, 36, 42, 44, 49, 57 diffusion, 30, 33, 64, 65, 71 diffusivity, 29, 32 dimer, 115, 116, 117 dimeric, 115 dimerization, 113, 114 diseases, 6 dissociation, 18, 25, 42, 43, 44, 45, 46, 48, 49, 52, 53, 54, 55, 58 distilled water, 87 distribution, 15, 46, 54, 56, 94, 95, 96, 97, 98, 99, 101, 102, 103, 104, 105, 110, 119 distribution function, 94 dosage, 43, 49 droplets formation, 90 dry, 3, 62, 63, 65, 70, 81, 82, 83, 85 DS-3, 100 DSC, 42, 47, 56 duration, 72

E earth, 10, 38, 39, 40, 44, 83, 84 Earth Science, 49 ecology, 10 economic, 10, 22, 38, 40, 45 economy, 11 Eden, 106 effluents, 39 electrical, 1, 72 electricity, 34 Electroanalysis, 5, 7 electrocatalyst, 4, 5

126

Index

electrocatalytic, 1 electrochemical, 1, 2, 3, 4, 5, 63, 70, 76, 81, 88, 93, 94, 100, 104, 105 electrochemical deposition, 3 electrochemical impedance, 4 electrodes, 1, 2, 3, 5 electrolysis, 106 electrolyte, 43, 63, 64, 65, 71, 73, 75, 76, 89, 102 electromotive force, 94, 101 electron, 2, 4, 5, 114 electronic, 67 electrons, 1, 2 EMF, 94, 101 emission, 24, 75 empirical potential, 57 employment, 82 encapsulation, 26 endothermic, 19, 22, 40 energy, 10, 11, 18, 19, 22, 23, 24, 25, 34, 36, 38, 39, 40, 42, 49, 57, 94, 102 energy consumption, 23, 94, 102 energy density, 22, 36 energy efficiency, 34 energy supply, 38 engineering, 35, 40, 94 English, 25, 43 Enhancement, 48 enlargement, 105 environment, 10, 19, 43, 75, 81 environmental, 5, 10, 34, 40, 64, 91 equilibrium, 10, 15, 16, 18, 19, 20, 21, 22, 26, 27, 29, 36, 42, 46, 50, 51, 52, 55, 87, 116, 117 ester, 3, 4 ethane, 11, 12, 20, 25, 26, 27, 40, 41, 43, 44, 48, 55, 56 ethers, 44 Europe, 70 evaporation, 3, 39, 61, 64, 73 evidence, 10, 26, 39 evolution, 29, 30 exclusion, 112 exercise, 105 exothermic, 15 exploitation, 44 exposure, 23, 62, 67, 72, 73, 75, 76, 77, 78, 79, 80, 81, 82, 86, 87, 90 extraction, 87 extraction process, 87

F fabric, 82 fabrication, 1, 2

factor analysis, 93, 94, 97, 98, 104 failure, 39 February, 90 field theory, 47, 56 film, 3, 16, 24, 29, 31, 50, 100, 101 financial support, 119 fish, 37, 50 fitness, 69, 82 flatness, 63 flexibility, 12 float, 17 flow, 10, 19, 37, 44, 45, 50, 56 flue gas, 10, 11, 34, 38, 40, 46 fluid, 15, 30 fluidization, 24 food, 5, 39 fossil, 34 fossil fuel, 34 Fourier, 25 Fox, 120 fractionation, 20, 34 France, 97 free radical, 109, 110 freedom, 94 fuel, 2, 10, 19, 20, 34, 35, 36, 40, 42, 53 functionalization, 115, 116, 117 funds, 119 fusion, 44

G gas, 9, 10, 11, 12, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 29, 30, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 71 gas chromatograph, 27 gas diffusion, 25 gas exploration, 10 gas phase, 17, 18, 20, 24, 27, 29, 32 gas separation, 10, 11, 34 gas turbine, 34 gases, 9, 12, 13, 33, 40, 56 Gaussian, 30, 94 gel, 23, 25, 36, 40, 52 Germany, 97 Gibbs energy, 41 glaciations, 39 glaciers, 19 glass, 38 global climate change, 40 global warming, 10, 39, 44, 53 glycerol, 15 glycol, 37, 48

127

Index glycoproteins, 37 gold, 3, 5, 64, 65 grain, 24, 43 graph, 104, 105 greenhouse, 10, 39, 41, 56 groups, 2, 71, 109, 110, 111, 114, 115, 117, 118, 119 growth, 15, 16, 17, 23, 24, 28, 29, 31, 32, 37, 38, 41, 43, 47, 48, 50, 58, 59, 110 growth rate, 28, 30, 32, 50, 58 Guangzhou, 109 Gulf of Mexico, 62, 64, 84

H H2, 10, 12, 34, 36, 104 H-2A, 100 Haj, 7 halogen, 114 heat, 15, 18, 19, 22, 23, 24, 25, 44, 45, 50, 57, 67, 69, 75, 77, 79, 80 heat transfer, 22, 24, 25, 45, 50 heating, 70 heptane, 55, 113, 114 herbicide, 2 heterogeneous, 2, 5, 17, 58 heteropolymers, 110 hexane, 14, 48, 116, 117 high pressure, 11, 12, 13, 21, 22, 29, 33, 36, 37, 40, 42, 47, 49, 51 high resolution, 44 high temperature, 62 HIV, 6 Holland, 53 homogeneous, 17 Hong Kong, 109, 119 host, 9, 12, 21 House, 11, 44, 45 humanity, 18 humidity, 61, 62, 63, 64, 65, 70, 71, 72, 75, 77, 79, 84, 89 hybrid, 5, 34, 35 hydration, 20, 42, 53, 58 hydro, 14, 19, 20, 48, 55 hydrocarbon, 9, 11, 20, 40, 115 hydrocarbons, 14, 19, 20, 48, 55 hydrodynamic, 112, 116, 117 hydrogen, 9, 10, 11, 12, 13, 19, 20, 23, 36, 37, 40, 42, 44, 45, 48, 49, 52, 53, 55 hydrogen bonds, 12 hydrogen gas, 23, 36 hydrogen sulfide, 9, 12, 20, 48 hydrolysis, 2 hydrophobic, 12, 18, 32, 44

hydroxyl, 3, 4 hypothesis, 94, 95, 96, 98, 100, 103, 106, 114

I ice, 9, 10, 12, 15, 17, 24, 25, 29, 31, 32, 33, 45, 47, 53, 57, 58 identification, 12, 15 images, 32 imaging, 16, 17, 29, 32, 33, 56 immobilization, 3 impurities, 100, 110 in situ, 6, 11, 39, 42, 43, 58 inclusion, 9, 12, 31, 37, 64, 75 independence, 93, 94, 100, 101, 103, 104 independent variable, 66, 67, 75 India, 107 Indian, 34 induction time, 17, 18, 38 industrial, 5, 11, 18, 19, 36, 37, 40, 72, 77, 80, 83, 96, 97, 100, 102 industrial application, 19 industry, 9, 11, 15, 23, 37 infancy, 5 inferences, 103 infinite, 15 inhibition, 37, 43, 47, 49, 59, 93, 94, 98 inhibitor, 18, 32, 37, 48, 95, 98, 102 inhibitors, 15, 17, 20, 37, 38, 40, 43, 49 inhomogeneities, 17 injection, 23, 37, 56 insight, 16 inspection, 102 instability, 10, 54, 55 integration, 1, 2, 3, 4 intensity, 14, 30, 31, 72 interaction, 12, 18, 62, 72, 89 interactions, 3, 12, 21 interface, 16, 17, 23, 24, 32, 50, 58, 64, 88 interference, 29 interval, 65, 75 intrinsic, 17, 25, 42, 45 Investigations, 50 ions, 62, 73, 80, 81, 82, 85, 87, 88 iron, 65, 95 irradiation, 67, 113, 114 ISO, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 78, 79, 80, 82, 83, 85, 86, 88, 89, 90, 91 isobutane, 44 isolation, 70 isopentane, 52 isotherms, 87

128

Index

isotope, 39, 43 isotropic, 30 Italy, 97

M

J January, 53, 91 Japan, 23, 33, 40, 52, 56, 97 Jung, 91

K kinetic model, 17 kinetic studies, 15, 17 kinetics, 12, 15, 16, 17, 22, 23, 24, 25, 26, 27, 29, 32, 37, 42, 45, 56, 58 Korean, 43, 54 krypton, 44

L Langmuir, 87, 88 laser, 113 lattice, 12, 13, 14, 21, 27, 36, 39 lattices, 10, 12, 21 law, 25, 104 leaching, 80, 85 lead, 88, 102, 104, 110, 114 lifetime, 23 light scattering, 113 limitation, 29, 110 limitations, 76, 79, 88 linear, 16, 29, 62, 65, 88 linear function, 65 linear regression, 65 linkage, 117, 118 liquefied natural gas, 40 liquid film, 63, 64, 89 liquid interfaces, 45, 47 liquid phase, 15, 16, 30 liquid water, 17, 18, 23, 24, 29, 48, 52 literature, 20, 27, 36, 87, 103, 105 lithium, 114 LNG, 19, 21, 22, 33, 40 London, 43 long period, 62 long-term, 87 low temperatures, 37, 48, 59

macromolecules, 40 magnet, 23 magnetic, 23, 29, 47 magnetic resonance imaging, 47 management, 42 mantle, 11 Mars, 11 MAS, 29 mass loss, 81, 82 mass transfer, 17, 24, 27, 29, 31, 32, 53 measurement, 15, 17, 31, 42, 43, 58 measures, 52 mechanical, 1, 23 median, 94 mediators, 5 melt, 25, 47, 57 melting, 24, 32, 45, 55, 64 memory, 17, 38, 41, 51, 59 MES, 33 mesoscopic, 54 metals, 5, 61, 62, 63, 72, 75, 88, 90, 91 meteorological, 72, 76 methane, 9, 10, 11, 12, 13, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 38, 39, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58 methanol, 15, 37, 43, 115 methylcyclohexane, 21, 27, 28, 33, 51, 55 Mexico, 61, 62, 67, 80, 84, 89 micelles, 114, 116, 117 migration, 16 mines, 20 miniaturization, 6 minority, 110 mirror, 102 mixing, 23, 24, 29, 33, 56 model system, 20 modeling, 54 models, 11, 15, 25 moieties, 2 moisture, 75 molar volume, 15 molecular dynamics, 25, 36, 51 molecules, 7, 9, 12, 13, 14, 17, 18, 19, 20, 21, 22, 23, 27, 28, 30, 32, 33, 36, 37, 38, 49, 59 monolayer, 3, 87 monolayers, 7, 65 monomer, 110 monomeric, 3 Monte Carlo method, 17 morphology, 3, 16, 52

129

Index Moscow, 90 motion, 30 motivation, 19, 33, 39, 40 mountains, 63 mouth, 84 movement, 85 MRI, 17, 44 MTBE, 28

N NA, 18 Na+, 88 NaCl, 51, 72, 98 nanometer, 72 nanotube, 4 nanowires, 5 National Academy of Sciences, 47 National Research Council, 9 natural, 9, 10, 11, 19, 20, 25, 34, 35, 38, 40, 43, 45, 47, 52, 54, 56, 57, 62, 65, 72, 80, 81, 85, 104 natural gas, 9, 10, 11, 19, 20, 34, 35, 38, 40, 45, 47, 52, 54, 56, 57 needles, 47 nerve, 2 nerve agents, 2 Netherlands, 44, 50 New York, 6, 7, 43, 47, 48, 54, 57, 106, 107, 120 next generation, 41, 45 Ni, 1, 5 Nielsen, 42 nitric oxide, 2 nitrogen, 9, 13, 45, 46, 48, 52, 54, 58, 59 NMR, 14, 16, 17, 20, 26, 27, 29, 31, 32, 33, 42, 44, 51, 53, 54, 56, 58 noise, 29 nonparametric, 93, 94, 95, 102, 103, 105 non-random, 98 normal, 37, 94, 95, 96, 97, 99, 100, 102, 103, 113 NPS, 94 NS, 104 Nuclear magnetic resonance, 42 nucleation, 15, 16, 17, 28, 29, 32, 37, 38, 47, 48, 51, 53, 59 nuclei, 15, 43

O observations, 42, 45, 47, 53, 56, 58, 93, 94, 98, 100, 104, 105, 106 offshore, 10, 43, 44 oil, 9, 10, 34, 37, 42, 45, 62

oil production, 62 one dimension, 25 online, 15 optimization, 22, 53 organic, 1, 39, 47, 50 orientation, 18, 64 Ottawa, 9, 42 oxidative, 5 oxide, 18, 26, 37, 54, 100, 101 oxygen, 2, 19, 57, 58, 64, 65, 71, 114 oxygen consumption, 71

P palletized, 22 paper, 6, 43, 50, 62 parameter, 61, 62, 63, 88, 94 particles, 24, 31, 33, 37, 64, 87 partition, 29 passivation, 70 pentane, 14, 48 performance, 22, 23, 37 permafrost, 10 permeation, 98, 103, 104 Petroleum, 41, 43, 44, 49, 53 pH, 88 pharmaceutical, 5 phase boundaries, 50 phase diagram, 20, 26, 41, 49, 110 phase transformation, 11, 15, 32 Photocatalytic, 56 physical properties, 58, 110 physico-chemical properties, 65 physics, 110, 119 pipelines, 9, 10, 19, 37, 40, 43, 46, 52 planetary, 22 planets, 11, 19, 49, 55 plants, 34 play, 11, 33, 65, 81, 87, 115 PMMA, 111, 112 polar ice, 19 polarization, 53, 71, 104, 105 politics, 53 pollutants, 64, 72, 80, 85 pollution, 91 polycrystalline, 58 polydispersity, 112 polyethylene, 18, 37 polyisoprene, 114 polymer, 37, 110, 114, 115, 117, 119 polymerization, 109, 110, 111, 115, 117, 119 polymerization mechanism, 110 polymers, 115

130

Index

polystyrene, 114 polyurethane, 110 poor, 23, 31, 32, 114 population, 94, 103 pore, 23 porosity, 25 porous media, 23, 25, 44, 53 porphyrins, 1 positive correlation, 106 potassium, 88 potential energy, 39 powder, 14, 29, 32, 57 power, 23, 29, 34, 38, 42 power generation, 34 power plants, 34, 38, 42 precipitation, 62, 63, 64, 72, 75, 79, 87, 88 prediction, 15, 17, 27, 41, 45, 90 predictive model, 17 preference, 20, 21 preparation, 110, 111, 113, 114, 117, 119 pressure, 9, 10, 11, 14, 16, 17, 19, 20, 22, 24, 25, 26, 27, 28, 29, 31, 33, 36, 37, 42, 44, 46, 47, 49, 55, 87, 102 Pretoria, 1 prevention, 45 probability, 93, 94, 96, 97, 98, 100, 104, 105, 106 probability distribution, 93, 94, 105 probe, 4, 71 producers, 104 production, 10, 11, 18, 33, 38, 41, 44, 45, 50, 93, 96, 97, 104 productivity, 17 prognosis, 62, 66, 68 program, 15, 38 promoter, 44 propane, 12, 20, 21, 25, 26, 27, 36, 40, 41, 44, 46, 48 propylene, 26, 54 protection, 90 protein, 18, 38, 50, 59 protocols, 102, 110 pulp, 39 P-value, 104 pyrolytic graphite, 3

rain, 62, 64, 65, 66, 72, 73, 74, 75, 76, 79, 81, 82, 85, 87, 88, 89 rainfall, 64, 70, 72, 73, 81, 82 Raman, 14, 16, 20, 33, 47, 55, 56, 57, 58, 59 random, 94, 98 range, 44, 64, 105 reaction mechanism, 114 reaction rate, 71 reaction time, 31, 37 reactive groups, 110 reactivity, 118 reality, 17, 65, 71, 110 reasoning, 17 recovery, 11, 25, 35, 38, 39, 40, 41, 46 redox, 1, 2, 3, 4, 5 reduction, 19, 22, 26, 33, 36, 71, 95 refrigerant, 39, 57 refrigeration, 11 regression, 66, 67, 73, 82 regulations, 20 rejection, 95 relationship, 39, 40, 44, 62, 65, 69, 72, 80, 88, 100 reliability, 15 research, 1, 3, 4, 5, 10, 11, 19, 40, 41, 42, 61, 62, 81, 82, 105 research and development, 42 Research and Development, 58 researchers, 2, 20 reservation, 55 reserves, 9 reservoir, 11, 38, 43, 47, 49 reservoirs, 19, 38, 39, 45 resistance, 6, 24, 29, 100 retention, 114 risk, 54 rods, 48 room temperature, 111 roughness, 3 runaway, 10, 39 rural, 64, 67, 68, 69, 75, 76, 77, 82, 83, 87, 88 Russian, 64 rust, 87

S Q quartz, 64

R radiation, 70 radius, 12, 112, 116, 117

safety, 20, 26, 52 SAFT, 15, 48 salinity, 62, 71, 80, 82, 83 salt, 65, 71, 81, 88, 90, 98, 104 salts, 83, 85, 87, 88, 98 sample, 17, 64, 87, 88, 94, 105 sample variance, 94 Sao Paulo, 90

131

Index savings, 39 scanning calorimetry, 16 scanning electron microscopy, 6, 24 scatter, 103 scattering, 111 science, 53, 94 scientific, 9, 10, 18 scientists, 11, 39, 40 SDS, 23 search, 37, 38 seawater, 11, 39, 53 SEC, 94, 102, 103, 112, 113, 114, 115, 116, 118 sediment, 46, 55 seeds, 17, 24 selecting, 19, 22, 87 Self, 3, 7, 56, 109 self-assembly, 2, 109, 111, 112, 113, 114, 116, 117, 119 SEM, 24 sensing, 2, 3 sensitivity, 5, 41 sensors, 1, 5, 64, 72 separation, 10, 11, 34, 35, 36, 52 series, 53 shape, 14, 22, 30, 42 shear, 50 Shell, 23, 44 shelter, 79 shores, 83 sign, 74 signals, 14 silica, 23, 25, 36, 37, 40, 52 silver, 64, 65 simulation, 11, 16, 17, 25, 33, 36, 38, 51 simulations, 26, 33, 40, 43 sites, 62, 63, 75, 81, 84, 85 SO2, 72, 73, 74, 80 sodium, 23, 58 sodium dodecyl sulfate (SDS), 23 software, 15 solar, 22 solid phase, 14, 16, 25, 29, 33 solid-state, 27 solubility, 23, 28, 29, 32, 33, 110 solutions, 43, 58 solvent, 2, 3, 64, 109, 111, 112, 114, 115, 116, 117, 119 solvents, 42, 115 sorption, 87 Spearman rank correlation coefficient, 103 species, 2, 3, 4, 5, 9, 13, 42, 101 specific surface, 87 spectra, 14, 29

spectroscopy, 4, 14, 16, 17, 20, 24, 29, 41, 53, 56 speculation, 10 speed, 18, 24, 99 stability, 10, 13, 18, 21, 22, 26, 32, 49, 52 stabilize, 13, 21, 26, 34 stack gas, 41 stages, 23, 24, 29, 35, 62 standard deviation, 86 statistical analysis, 93 statistics, 93, 94, 95 steady state, 27 steel, 70, 71, 73, 75, 77, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88 stochastic, 17 storage, 10, 11, 19, 21, 22, 24, 25, 26, 33, 35, 36, 37, 38, 39, 40, 41, 42, 45, 46, 48, 50, 53, 55, 56, 57 strategies, 2, 25, 110 strategy use, 4 streams, 34 strength, 103 strong interaction, 3 styrene, 114 substances, 64, 70 substitution, 115 substrates, 3 sulfate, 58, 87, 88, 101 Sulfide, 49, 51, 54 sulfur, 3, 62, 73, 80, 83 sulphate, 80 summer, 62, 85 Sun, 23, 25, 27, 42, 48, 55, 109, 121 supercritical, 15 suppression, 38 surface water, 70, 73 surfactant, 24, 44, 48 surfactants, 23, 55 surviving, 49 symbols, 99 symmetry, 14, 102 synthesis, 22, 23, 27, 34, 49, 110, 111, 119 synthetic, 21, 38, 40, 45, 109, 110, 119 systematic, 28, 49 systems, 10, 17, 23, 27, 30, 31, 32, 42, 51, 70, 105, 109, 110

T technological, 53 technology, 9, 11, 19, 20, 23, 33, 36, 41, 44, 45, 57, 93 Teflon, 38

132

Index

temperature, 3, 9, 10, 11, 14, 15, 16, 17, 19, 20, 22, 24, 25, 26, 27, 28, 31, 33, 36, 39, 53, 61, 62, 63, 64, 67, 70, 71, 72, 75, 76, 89, 102, 111 territory, 62, 63, 83 test statistic, 103 tetrahydrofuran, 26, 36, 46, 47, 57, 59 textbooks, 94 theoretical, 94, 98 theory, 95, 104 thermal, 17, 21, 24, 25, 31, 39, 45, 46, 51, 56, 70 thermal conduction, 25 thermal energy, 45 thermal expansion, 21, 51, 56 thermodynamic, 11, 12, 20, 21, 22, 52, 54, 57 thermodynamics, 14, 15, 20, 26, 27, 32 Thomson, 106, 107 three-dimensional, 13 Ti, 94, 95, 96, 97 time, 9, 15, 16, 17, 23, 25, 26, 28, 29, 32, 37, 45, 51, 61, 62, 63, 64, 65, 66, 67, 68, 69, 71, 72, 73, 74, 75, 76, 77, 79, 81, 82, 85, 87, 88, 89, 93, 94, 114 time consuming, 94 time periods, 63 Titan, 11, 49, 52, 57 trade, 26, 83 trade-off, 26 transfer, 4, 5, 50, 110, 114 transformation, 63, 88 transition, 1, 3, 53, 55, 94 transmission, 10, 19, 23 transport, 2, 4, 5, 11, 17, 19, 22, 23, 24, 40, 57, 71, 104 transportation, 10, 11, 37, 45, 52, 63 traps, 5 trend, 4, 28, 30, 32, 93, 94, 95, 96, 97, 98, 100, 102, 103, 104, 105, 106 tsunamis, 10 tuberculosis, 6 tunneling, 5 two-dimensional, 98

U UK, 107 uniform, 61, 71 UNIQUAC, 15 urban, 64, 68, 69, 75, 80, 83 users, 19, 22 USSR, 49, 97 UV, 113, 114 UV irradiation, 113, 114

V values, 16, 27, 39, 61, 63, 64, 65, 67, 70, 76, 79, 81, 82, 83, 84, 85, 87, 88, 96, 97, 98, 99, 101, 102, 103 van der Waals, 10, 12, 14, 15, 20, 21, 58 vapor, 15, 28, 39, 55 variable, 15, 73, 74, 75, 88, 89, 102 variables, 66, 68, 75, 100, 102 variance, 94, 96 variation, 66, 67 velocity, 70, 72 ventilation, 70 viscosity, 29, 110 voltammetric, 3

W water, 9, 10, 12, 13, 15, 16, 17, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 31, 32, 35, 37, 38, 39, 41, 42, 44, 45, 48, 49, 50, 51, 52, 54, 55, 56, 58, 61, 64, 65, 66, 70, 71, 72, 73, 76, 87, 88, 89, 102, 106, 114 weight loss, 69, 79 wells, 11 wet, 32, 62, 63, 65, 70, 85 wetting, 32, 63, 64 wind, 62, 63, 70, 72 windows, 70, 75 winter, 38, 62, 85

X xenon, 44 X-ray, 56, 58 X-Ray diffraction, 14, 16, 20, 27, 56, 58

Y yield, 22, 114

Z zinc, 73 Zinc, 65, 74