Advances in Crystal Growth Inhibition Technologies

Advances in Crystal Growth Inhibition Technologies

This page intentionally left blank.

Advances in Crystal Growth Inhibition Technologies Edited by

Zahid Amjad The B.F Goodrich Company Brecksville, Ohio

Kluwer Academic Publishers

New York, Boston, Dordrecht, London, Moscow

eBook ISBN: Print ISBN:

0-306-46924-3 0-306-46499-3

©2002 Kluwer Academic Publishers New York, Boston, Dordrecht, London, Moscow

All rights reserved

No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher

Created in the United States of America

Visit Kluwer Online at: and Kluwer's eBookstore at:

http://www.kluweronline.com http://www.ebooks.kluweronline.com

PREFACE This volume documents the proceedings of the symposium entitled “Advances in Crystal Growth Inhibition Technologies” sponsored by the Division of Colloid and Surface Chemistry of the American Chemical Society. The symposium was held in New Orleans on August 22 – 26, 1999 at the 218th American Chemical Society National Meeting. A total of 27 papers were presented by a wide spectrum of scientists. There was a strong attendance by representatives from government, academia, and industrial research centers. The objective of the symposium was to present and discuss recent developments in crystal growth and inhibition processes. The precipitation and dissolution of sparingly soluble salts from aqueous solution have major consequences in a wide range of technologically important processes such as mineral processing, petroleum recovery, geothermal, pulp and paper industry, production of potable water, and industrial water treatment. A significant part is also played in environmental processes such as wastewater treatment, agriculture, and the damage to building materials by climatic conditions and airborne pollution. In the last two decades a large variety of crystal growth inhibitors have been developed and successfully applied to prevent the precipitation of scale forming salts on equipment surfaces. The build-up of these salts on heat exchangers and membrane surfaces can lead to loss in production, increased corrosion rates, increased pumping costs, etc. The crystallization of sparingly soluble salts is also of primary importance in biological systems. Dental calculus, or tartar, consists primarily of salts of calcium, phosphate, and carbonate. Calcium oxalates are the main components of pathological deposits in the urinary tract. The medical community is seeing a renewed interest in the regulation of these species. To enhance the performance of industrial and biological systems, crystal growth inhibitors are generally incorporated into the formulations. The commonly used inhibitors include polyphosphates, organophosphonates, polycarboxylates, and phosphonocarboxylates. The role of inhibitor, especially polymeric (i.e., polyacrylates, polymaleates, copolymers of acrylic, maleic, and sulfonic acids) is not only to prevent the precipitation of sparingly soluble salts from aqueous solutions but also to disperse the

v

suspended matter such as clay, silt, corrosion products, etc. Therefore, a fundamental knowledge of the solution properties of inhibitors (polymeric and non-polymeric) is of vital importance in developing high performance formulations for achieving optimum system efficiency. This volume is organized into four sections which essentially maintain the same structure as the Symposium. The six papers which comprise Section 1 deal with precipitation and inhibition of calcium phosphates, calcium pyrophosphates, and calcium phosphonates. The inhibitors evaluated include metal ions (i.e., Cu, Co, Zn, etc.), poly(acrylate), acrylic acid-based copolymers, phosphocitrate, and alkylphoshates. In Section 2, the precipitation and inhibition of calcium carbonate are presented in two papers, while the remaining two papers are concerned with the precipitation of calcium carbonate in sea water and mixed scale formation at elevated temperature. Section 3 is comprised of four papers which address the precipitation of salts that are important in petroleum and sugar refining processes, also addressing the very important aspect of inhibition properties of polymeric and non-polymeric materials. The four papers of Section 4 are concerned with the treatment of industrial water systems, complexation of proteins with polymers, and crystal-growth restriction through clathrate hydrate formation. In this book the academic researchers and technologists will find important information on the interaction of polymeric and non-polymeric inhibitors with a variety of scale forming crystals such as calcium phosphates, calcium carbonate, calcium oxalates, barium sulfate, calcium pyrophosphates, and calcium phosphonates. Moreover, the book delivers information to plant managers and formulators who would like to broaden and deepen their knowledge about processes involved in precipitation of sparingly soluble salts and learn more about the inhibitory aspects of various commercial available materials. Furthermore, experienced researchers will get fruitful and inspiring ideas due to the easily accessible information about overlapping research areas, which will promote discoveries of new inhibitors (synthetic and/or natural) for the currently unmet challenges. Zahid Amjad Cleveland, Ohio

vi

ACKNOWLEDGMENTS I wish to thank all the participants and the contributors to this book and to the symposium on which it is based. Financial support of the national and international scientists is greatly appreciated. Special thanks are extended to the American Chemical Society Division of Colloid and Surface Chemistry, ACS Corporation Associates, Avlon Industries, BetzDearborn, Colgate-Palmolive Company, RNA Corporation, and the BFGoodrich Company. Their generous assistance contributed substantially to the success of the symposium. I express my appreciation to Drs. Peter G. Koutsoukos and Michael M. Reddy for chairing various sessions. Thanks are also extended to Jennifer Thomas-Wohlever for her efficient and organized handling of the considerable correspondence associated with both the symposium and the book. I am thankful to Dr. Victoria F. Haynes for her encouragement and support and to the management of the BFGoodrich Company for permitting me to organize this symposium and to edit this volume. I also acknowledge the support of my family and Susan Safren, the acquisition editor at Kluwer Academic/Plenum Publishers who kept me on task in a firm but patient manner.

vii

This page intentionally left blank.

CONTENTS 1. Influence of Foreign Metal Ions on Crystal Morphology and Transformation of Brushite (CaHPO4 . 2 H 2O .) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. E. L. Madsen and J. B. Pedersen

1

2. Inhibition of Calcium Phosphate Precipitation by Polymers in the Presence of Iron (III). The Influence of Chelating Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Z. Amjad, J. F. Zibrida, and J. A. Thomas-Wohlever

15

3. Interaction of Sodium Polyacrylate with Octacalcium Phosphate . . . . . . . . . . . . A. Bigi, E. Boanini, G. Cojazzi, G. Falini, S. Panzavolta, and N. Roveri 4. Effect of Alkyl Phosphates on the Formation and Crystallization of Calcium Phosphates in an Aqueous Phase. ........................................... S. Shimabayashi, K. Furukawa, and T. Hino 5. Calcium Pyrophosphate Crystal Salt Forms and the Influence of Phosphocitrate J. D. Sallis, A. Wierzbicki, and H. S. Cheung 6. Performance of Anionic Polymers and Precipitation Inhibitors for Calcium Phosphonates: The Influence of Cationic Polyelectrolytes. . . . . . . . . . . . . . . . Z. Amjad, R. W. Zuhl, and J. A. Thomas-Wohlever 7. Crystal Growth of Calcium Carbonate in Seawater. The Effect of Temperature and of the Presence of Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Kladi, P. G. Klepetsanis, T. Ostvold, C. G. Kontoyiannis, and P. G. Koutsoukos

29

43 57

71

85

8. Calcite Crystal Growth Rate Inhibition by Aquatic Humic Substances. . . . . . . . . 107 M. M. Reddy and A. R. Hoch 9. The Inhibition of Calcium Carbonate Formation in Aqueous Supersaturated Solutions. Spontaneous Precipitation and Seeded Crystal Growth. . . . . . . . . . . . . . . 123 P. G. Klepetsanis, A. Kladi, T. Ostvold, C. G. Kontoyiannis, P. G. Koutsoukos, Z. Amjad, and M. M. Reddy

ix

10. Calcium Carbonate and Calcium Phosphate Scale Formation and Inhibition at Elevated Temperature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P. G. Klepetsanis, P. G. Koutsoukos, and Z. Amjad

139

11. Effect of Temperature on Barium Sulfate Scale Inhibition of Diethylene Triamine Penta (Methylene PhosphonicAcid). . . . . . . . . . . . . . . . . . . . . . M. Yuan

151

12. The Role of Calcium Phosphino-Polycarboxylate Complexation in Inhibiting BaSO4 Precipitation from Brine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. J. Xiao, A. T. Kan, and M. B. Tomson

165

13. Inorganic Mineral Scale Control in Sugar Evaporators Using Scale Inhibitors. . J. S. Gill

187

14. Control of Crystallization Processes by Diblock Copolymers. .. . . . . . . . . . . . 0. Dogan, E. Akyol, S. Baris, and M. Oner

197

15. An Algicide Using Anionic Polymers to Sequester and Stabilize Copper in an Oxidizing Aqueous Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. P. Garris

207

16. Rational Development of New Cooling Water Chemical Treatment Programs for Scale and Microbial Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . K. D. Demadis, B. Yang, P. R. Young, D. L. Kouznetsov, and D. G. Kelley

215

17. Complexation of Proteins with Polymers in Aqueous Solutions . . . . . . . . . . . . B. Wang and E. Kokufuta

235

18. Crystal-Growth Restriction through Clathrate Hydrate Formation: Applications to Nanoparticle Synthesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. Irvin Jr., S. Li, B. Simmons, V. John, G. McPherson and C.J. O' Conner

255

19. About the Editor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

267

20. Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

269

x

INFLUENCE OF FOREIGN METAL IONS ON CRYSTAL MORPHOLOGY AND TRANSFORMATION OF BRUSHITE (CaHPO4 .2H 2O)

H. E. Lundager Madsen and Jo Bech Pedersen Chemistry Department, Royal Veterinary and Agricultural University, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark

ABSTRACT Brushite, forming tabular crystals, has been precipitated at 25°C in the presence of each of 14 different di- and trivalent metal ions. The effect of additives was pH-dependent, which could be related to the presence or absence of amorphous precipitate. Zn favoured aggregates, and the transition metals with the exceptions of Mn(II), Co(II) and Cu(II) favoured irregular growth. Zn inhibited lateral growth, as did Cd and Cr(III) at low and Cu(II) at high pH. Several of these ions have a marked effect on the solvent-mediated phase transformation of brushite to more basic calcium phosphates even at concentrations below 10 µM. Cu(II) and Zn are strong inhibitors, whereas Pb(II) is a moderate promotor. This is explained by the formation ofnuclei with apatite/pyromorphite structure, i.e. (Ca,Pb)5 OH(PO4 )3 .

INTRODUCTION Calcium hydrogen phosphate dihydrate, CaHPO4 .2H2 O, known as the rare mineral brushite, is a metastable compound which is normally observed as the primary crystalline product when calcium phosphate is precipitated at low pH and low temperature.1,2 This is in good accordance with Ostwald's well-known "Stufenregel" (rule of stages) for precipitation. For a sparingly soluble salt it easily forms large, well-developed crystals well suited for morphological studies.3 Its theoretical growth habit was determined on the basis of a PBC (periodic bond chain) analysis by Heijnen and Hartman4 and agrees well with observations. It belongs to the crystal class m (monoclinic domatic class), i.e. it has a mirror plane as the sole symmetry element. The space group, according to the revised setting of Heijnen and Hartman, isAa. The metastability of brushite manifests itself in the solvent-mediated transformation

Advances in Crytal Growth Ihibition Technologies Edited by Amjad. Kluwer Academic/Plenum Publishers. New York. 2000

1

(hydrolysis) in neutral solution to more basic calcium phosphate, notably octacalcium phosphate (OCP): 10 CaHPO 4 .2H2 O

Ca8 H2 (PO4 )6 .5H2 O + 2 Ca2+ + 4 H2 PO4 . + 15 H2 O

OCP is metastable itself, but its transformation to the stable end member hydroxyapatite (HAP: Ca5 OH(PO4 )3 ) is much slower than the above process of formation. The transformation of brushite is initiated by heterogeneous nucleation of OCP or, maybe, HAP, on brushite crystals,5,6 a process which seems to be strongly favored by crystal defects of the substrate.6 The liberation of dihydrogen phosphate ions shown in the above reaction scheme causes a decrease of pH of the solution, which makes pH measurement a convenient method of recording the progress of the reaction. Interest in the influence of foreign metal ions on calcium phosphate crystallization has, with few exceptions, been focused on biological mineralization and, therefore, on ions abundant in biological systems. Bigi et al.7 studied the influence of Mg2+, Sr2+ and Pb2+ on the transformation of brushite and OCP to HAP and found significant inhibition only by Mg2+. The effect of the latter on the transformation of brushite was also the subject of investigations by Rowles,8 who found that whitlockite (Ca9MgH(PO4)7) rather than OCP and HAP is formed at [Mg2+] > 10-4 M. Fleish, Bisaz and Russel9 have demonstrated that Pb2+ promotes the precipitation of calcium phosphate under physiological conditions, i.e. ionic strength 0.16, pH 7.4 and temperature 37°C. Finally, Cu2+ has been shown to be a potent inhibitor of the transformation of brushite to OCP.10 The object of the present investigation is a systematic screening of a number of di- and trivalent metal ions for their effects on the reaction crystallization of brushite at 25°C and the transformation of brushite to more basic calcium phosphate at 37°C. In crystallization of brushite a concentration of foreign metal of 1 % of the calcium concentration was chosen; we regard this level as the upper limit for a study of inhibition in the proper sense. All experiments were carried out at two different pH values to study the influence of amorphous calcium phosphate (ACP) as a transitory precipitate. In the hydrolysis experiments two levels of foreign metal ion concentration were chosen, viz. 5 and 10 µM, because this was the range of transition from weak to strong inhibition found for Cu.10

EXPERIMENTAL Apparatus pH was measured with a Metrohm pH meter model 704 or, when higher accuracy was required, with a Radiometer pH/ion meter model PHM 240. The instruments were calibrated using ISO standard buffers: potassium hydrogen phthalate, 0.05 m, pH = 4.008, and potassium dihydrogen phosphate + sodium hydrogen phosphate, both 0.025 m, pH = 6.865, both pH values valid at 25°C. Crystalline products were characterized by optical microscopy, either in situ, through the bottom of the flask, with a Zeiss Axiovert 25C inverted microscope, or ex situ using a Zeiss Jenapol polarization microscope equipped with suitable accessories for measurements, including a tilting compensator of the Ehringhaus type. Further characterization was made by x-ray powder diffraction, using a Philips diffractometer and Co Kα-radiation (λ=179.02 pm).

Materials All solutions were prepared from analytical reagent grade chemicals. The water used as

2

solvent was demineralized water further purified by passing through a filter with activated carbon and a Silhorko Silex-1 mixed-bed ion-exchange column. The conductivity of the effluent never exceeded 0.1 µS/cm. Brushite for hydrolysis experiments was prepared by the method of Tovborg Jensen and Rathlev11 using a 2-liter stirred crystallizer and a 4-channel peristaltic pump to add reagents.

Procedures Brushite was precipitated at 25°C by fast mixing of two 25-ml aliquots of solutions,13,12 one being 0.1 M calcium nitrate, the other a mixture of potassium dihydrogen phosphate and potassium hydrogen phosphate with a total phosphate concentration of 0.1 M. Whenever a foreign metal was present, it was added as the nitrate at a concentration of 1 mM to the calcium nitrate solution prior to mixing. Initial precipitates were observed with the inverted microscope, after which the flasks were left in the thermostat without agitation to the next day, where samples were withdrawn for microscopy. A drop of the sample with crystal suspension was placed on a microscope slide and the aqueous phase removed with a drop of alcohol, making use of the Marangoni effect. When the alcohol had evaporated, the crystals were immersed in cedarwood oil (nD = 1.515) and covered with a cover glass. The crystals were studied in the microscope between crossed polars to determine size distribution and crystal habits. Thirty crystals were selected at random, the largest and smallest dimension ofeach measured with the aid of a scale in the eyepiece, and the thickness estimated from the interference color, making use of the known birefringence in a section perpendicular to the b -axis, viz. nβ - nα= 0.00526. l3 Because only the size distribution was sought, high precision was not attempted, Morphologically the crystals were classified as regular single crystals, symmetric crystals, irregular, twins, or aggregates.3 For the hydrolysis experiments 0.5 g of brushite was placed in a 250-ml Pyrex glass bottle with screw cap. 250 ml 0.1 mM phosphoric acid with foreign metal as nitrate was added. The bottles were rotated end-over-end in a thermostat at 37°C. pH was measured at regular time intervals, until it became constant. The contents of each bottle were then filtered, the solid phase washed with alcohol and dried and finally examined by microscopy and x-ray diffraction. Initial supersaturations were calculated from solution compositions with the aid of a previously described computer program14,15 using literature values of solubility products and other equilibrium constants.16,17,18,19

RESULTS In one series of brushite precipitations the phosphate mixture was made up from 23 ml KH2PO4 and 2 ml K2HPO4, in the other 20 ml KH2PO4 and 5 ml K2HPO4 were used. These series will be denoted I and II, respectively. We define the supersaturation (saturation ratio) β(b) with respect to brushite as

where Ksp is the solubility product. Similarly, we define the ion product I3 of ACP as

since it has been shown that it is the value of this expression which determines whether an amorphous precipitate is formed or not.20 We find the values in Table 1, neglecting the influence of foreign metals on supersaturation. Table 1. Calculated values for supersaturations, etc. in brushite precipitations. β(b) initial

log I3 initial

pHatsaturation

Series I

31.2

-1.14

3.973

Series II

75.3

0.54

4.075 2

Since the limiting value of log I3 for formation of ACP at 25°C is -0.85, there should be an amorphous precipitate in the experiments of series II and so was actually the case. The following day all the precipitate was crystalline.

pH

t/ min Figure 1. Decrease of pH during crystallization of brushite in the presence of different foreign metal ions. Upper: main group and closed-shell (d10) ions. Lower: transition metal ions.

4

Figure 2. Mean dimensions (upper row) and morphologies (lower row) of brushite crystals. Left: series I. Right: series II. Note that the ordinate on the upper diagrams is logarithmic.

5

foreign metal ion Figure 2. (Continued).

The course of crystallization of brushite in experiments of series I was followed during the first hour by pH measurements. Figure 1 shows the results. The formation of ACP in series II means that such measurements will not provide the same information about brushite crystallization in this case. Figure 2 shows the average dimensions and the crystal morphologies from the two series of experiments, and Figures 3 and 4 show typical crystals from experiments in series I and II, respectively. The examples shown are those where the morphological effects of the additives are strongest according to Figure 2. Regular crystals are of triangular outline and rather small, but thicker than most other crystals (Figure 4d). Their habit agrees well with the theoretical one of Heijnen and Hartman.3,4 Most other crystals are elongated. Symmetric crystals have a pair of long and a pair of short parallel edges, but the latter are often distinctly curved (Figure 4c). They are believed to be polysynthetic twins with (010) as the contact plane and the b axis as the twin operator.3 The irregular crystals often show parallel overgrowth. A Student's t-test was carried out on the size distributions of crystals to see whether the differences between mean dimensions of crystals from experiments with additives and from blanks were statistically significant. The cases where the difference was significant on either the 5 % (significant) or the 0.5 % level (highly significant) are indicated in Table 2. Additives not mentioned in the table caused no statistically significant changes of crystal dimensions. Finally, Figure 5 shows the results of hydrolysis experiments with 10 µM additive. The first pH measurement was made 1 hour after start, at which time it was expected that the solution was saturated with respect to brushite. The measured value of about 7.5 for the systems with intermediate or strong inhibition confirms this assumption, since it agrees with the calculated value. The total duration of the majority of experiments was about 200 hours, i.e. 7-8 days. A few experiments with particularly strong inhibitors were continued for about one further week, at which time pH had stabilized. The final pH of most of the systems, about 5.5. agrees with the value calculated on the assumption of complete conversion into OCP, the solution being saturated with respect to this compound. The results obtained with additive concentration of 5µm were similar, but the effect was less pronounced, as expected.

6

Figure 3. Photomicrographs of brushite crystals from experiments of series I Additives a) Blank, b) Pb, c) Zn d) Fe (II)

7

Figure 4. Photomicrographs of brushite crystals from experiments of series II. Additives: a) Blank, b) Mg, c) Cr (III), d) Mn (II).

8

Table 2. Significance of changes of brushite crystal dimensions with additives, compared to blanks Mg Series 1

Ba

A1

Pb

Length

Zn

Cd

Cr

⇓

↓

⇓

⇓

⇓

Breadth ⇓

Thickness

Series 2

↓

⇓

↓

⇓

⇓

↓

↓

Length Breadth Thickness

↓

⇑

↓

⇓

Mn

Fe2+

Fe3+

Ni

↓

↓

Cu

↓ ⇓

⇓

↓

⇓

⇓ ↓ ⇑

↓

⇓

↓

⇓

Meaning of symbols: Increase significant on 5% level, ⇑: on 0 5% level ↓: Decrease significant on 5% level, ⇓: on 0 5% level No symbol Change not significant

t/h Figure 5. Decrease of pH during hydrolysis of brushite at 37 C in the presence of different foreign metal ions at a concentration of 10 µM. Upper: main group and closed-shell ions. Lower: transition metal ions.

9

The results of phase identifications of products of hydrolysis by x-ray diffraction are shown in Table 3. The most abundant phase, as judged from line intensities, is mentioned first in each case. Essentially the same phase composition was found for both levels of foreign metal concentration. Table 3. Phase compositions of products of hydrolysis of brushite at 37 C in the presence of different foreign metal ions. Additive

Phases

Additive

Phases

Additive

Phases

blank

HAP,OCP

Pb

HAP,OCP

Fe(II)

B, OCP

Mg

HAP,OCP

Zn

OCP,B

Fe(III)

B, HAP, OCP

Sr

OCP,HAP

Cd

HAP,OCP

Co(II)

OCP, HAP

Ba

HAP,OCP

Cr(III)

HAP,OCP

Ni

HAP, OCP

AI

B,OCP,HAP

Mn(II)

OCP,HAP

Cu(II)

OCP, B

DISCUSSION Since all the metal ions studied in this work form sparingly soluble phosphates, they have a high affinity to phosphate ions. We should therefore expect a measurable effect of most of them on brushite crystallization as well as on phase transformation of brushite. This is, in fact, clearly verified by the results, as is evident from the figures and Tables 2 and 3.

Brushite Crystallization In the experiments of series I, Zn2+ is the strongest inhibitor of the main group and closedshell ions, followed by CD2+ and A13+. The latter shows its effect mainly on the rate of crystallization (Figure 1), not to any significant extent on crystal dimensions or morphology (Figure 2). On the other hand, very characteristic aggregates are formed in the presence of Zn2+ (Figure 3c). Pb2+ is a strong promotor (Figure 1), favoring regular crystals (Figures 2 and 3b). This extends the validity of the findings of Fleisch et al.9 Of the transition metal ions Cr3+ and Mn2+ are the strongest inhibitors, though without any marked effect on morphology (Figure 2). Ni2+ is a moderate promotor, but in contrast to Pb2+ it increases the proportion of irregular crystals (Figure 2). The two iron ions influence kinetics in a way distinctly different from the effects of other ions (Figure 1). It is well known from numerous examples in the literature that Cr3+ inhibits crystal growth of many phosphates, even those with high solubility like KH2PO4. The formation of amorphous precipitate in the experiments of series II may have three consequences: I. The supersaturation with respect to brushite is lowered which, in particular, reduces nucleation rate and leads to larger and/or more regular brushite crystals. The latter effect comes from the fact that growth of crystals is fed by dissolution of amorphous precipitate. 2. The individual particles of the amorphous precipitate serve as heteronuclei for brushite, which reduces nucleation of aggregates. 3. Foreign metal ions coprecipitate with the amorphous precipitate, whereby their effect on' brushite crystallization is reduced. The importance of these factors is evident from Figure 2. In most of the systems the size of crystals and the proportion of regular crystals have increased in comparison with the

10

products of experiments of series I. Mg2+, Cr3+, Fe2+ and Ni2+ tend to suppress regular crystal growth, but Cr3+ nevertheless causes a highly significant increase in average crystal thickness (Table 2). On the other hand, both Zn2+ and Cu2+ cause highly significant decrease in crystal dimensions without any noticeable morphological effect. It is remarkable that neither Mg2+ nor Cu2+ show any significant effect in experiments of series I except a slight influence ofthe latter on kinetics (Figure 1).

Brushite Hydrolysis When it comes to hydrolysis of brushite, most of the foreign metal ions are inhibitors, the only exceptions being Mg2+, which has no significant effect, and Pb2+, which is a promotor like in the case of brushite crystallization (Figure 5). In both cases the final solid phases are HAP and OCP like in the case ofthe blank (Table 3). The behavior ofthe Fe3+ system is remarkable. Kinetically it follows the blank very closely with the same end value of pH, but the final phases are unchanged brushite, HAP and OCP, which discloses the ion as a rather strong inhibitor. Of the main group and closed-shell ions Sr2+ and Zn2+ are the strongest inhibitors, followed by Al3+. This conforms with the compositions of final phases, where OCP dominates over HAP, the latter being absent with Zn2+; brushite is still present with Al3+ and Zn2+. Of the transition metal ions Cu2+ is by far the strongest inhibitor, followed by Co2+ and Mn2+. As judged from Figure 5, Fe2+ is only a weak inhibitor, but the phase compositions show that both iron ions are, in fact, comparable to Zn2+ and Cu2+ in that respect. Most of the inhibitors cause a distinct induction period, during which nothing seems to happen, after which pH drops almost to the end value in relatively short time. The strong inhibitors, Zn2+ and Cu2+, lead to a three-stage process: 1) the induction period, 2) a relatively long period where pH decreases slowly, and 3) a relatively short period where pH drops to the final level. The latter stage is not shown in Figure 5, but was, in fact, observed in the present and has been demonstrated in previous work.10 The middle stage is observed only at Cu2+ concentrations above ~5µM. Our findings are somewhat at variance with those of Bigi et al.7 since we found no significant inhibition by Mg2+, but by Sr2+ instead. The level of [Mg2–] in our study is far below that at which Rowles observed the formation of whitlockite instead of OCP.8 Again, the promotive effect of Pb2+ agrees with observations of Fleisch, et al.9

Mechanisms of Inhibition and Promotion Inhibition of crystal growth is caused by adsorption of the inhibitor, and so is most probably inhibition of nucleation, the inhibitor being adsorbed to minute nuclei and preventing their further development. Promotion, on the other hand, involves most likely the nucleation stage only, this stage being facilitated by the possibility of formation of a compound less soluble than that formed in the absence of additive. Thus, the promotive effect of Pb2+ on both crystallization and hydrolysis of brushite may be explained by the nucleation of hydroxypyromorphite (Pb5OH(PO4)3), which is isomorphous with HAP, being much less soluble, but showing a higher rate of crystallization.21 Mixed nuclei with both calcium and lead may be formed as well. The increased proportion of regular brushite crystals in the presence of Pb2+ in series I may have the same cause as the general trend in this direction in series II, viz. lowering of supersaturation by initial precipitation of another phase, in this case hydroxypyromorphite. Reduction of average crystal size in comparison with the blank may be caused by promotion of nucleation, inhibition of crystal growth, or both. Growth inhibition is usually, at least in the case of brushite in the present study, accompanied by irregular growth, because adsorption is not uniform, or growth rate fluctuates in space and time. If the degree of adsorption is different for different crystal faces, then habit modification results. Good examples

11

are the experiments with Zn2+ in series I and with Cr3+ in series II, where it is obvious that inhibition is stronger on the lateral than on the (010) faces. The cases of Zn2+ and Cu2+ merit particular interest. The high proportion of aggregates formed in the presence of the former ion, especially in series I, indicates that nuclei with a structure widely different from those of pure calcium phosphates have been formed. These nuclei start growing into brushite crystals in different directions more or less simultaneously, according to the goodness of fit of the two structures, that of brushite and that of the nucleus The fact that Cu2+ shows significant inhibition only in series 11, i.e. at high pH, isjust an example among many of the peculiar properties of this ion. Whenever most of the divalent metal ions in the first part of the periodic system form salts with similar crystal structures (sulfates, carbonates, phosphates, etc.), Cu2+ is always an exception. In a number of experiments, notably in series II, two distinct crystal populations are observed: large irregular and small regular crystals. The former are formed first, and their crystallization would normally lower supersaturation to a level below the critical one for nucleation. However, they also consume the inhibitor, and the second population develops from residual, not completely inhibited nuclei in an almost pure solution at low supersaturation. Inhibition ofheterogeneous nucleation of OCP or HAP on brushite crystals may take place either by inhibiting growth of the nuclei or by poisoning the substrate, thereby hindering the initial stage of the process. Either mechanism may account for the majority of observations in the present study. With the former, the induction period should be the time it takes for the nucleus to grow to such a size, well above the critical one, that the chemical potential of its growth units is well below that of growth units in the critical nucleus.22 This time increases in the presence of a growth inhibitor. The latter mechanism has been used to explain inhibition by Cu2+, which gradually disappears from solution by adsorption and subsequent solid-state diffusion into the brushite crystals.10 This is the only mechanism which can explain the existence of an intermediate, slow stage of the crystallization process, and it probably applies to the case of Zn2+ as well. These two ions are also those which most strongly inhibit crystal growth of brushite in the higher pH range, a fact which further supports the second mechanism. Results for Al3+ and the two iron ions are more difficult to interpret due to their relatively complex aqueous chemistry. These cases will need additional investigations. Finally, the appearance of both OCP and HAP in most of the systems is not unexpected and is not necessarily the result of further hydrolysis of OCP to HAP. It has been shown by high-resolution transmission electron microscopy that apparently pure, single OCP crystals as judged by both optical microscopy and x-ray diffraction often contain regions with HAP structure.23,24

ACKNOWLEDGEMENTS This work has been supported by grants from the Carlsberg Foundation and the Danish National Scientific Research Council. We wish to thank Miss Stina Lykkegaard for valuable technical assistance.

REFERENCES 1. H. E. Lundager Madsen and G. Thorvardarson, Precipitation of calcium phosphate from moderately acid solution, J. Crystal Growth 66:369 (1984). 2. F. Abbona, H. E. Lundager Madsen and R. Boistelle, The initial phases of calcium and magnesium phosphates precipitated from solutions of high to medium concentrations, J. Crystal Growth 74:581 (1986). 3. F. Abbona, F. Christensson, M. Franchini Angela and H. E. Lundager Madsen, Crystal habit and growth

12

conditions of brushite, CaHPO4. 2H2O, J. Crystal Growth 131:331 (1993). 4. W. M. M. Heijnen and P. Hartman, Structural morphology of gypsum (CaSO4. 2H2O), brushite (CaHPO4. 2H2O) and pharmacolite (CaHAsO4. 2H2O), J. Crystal Growth 108:290 (1991). 5. H. E. Lundager Madsen, Heterogeneous nucleation of calcium phosphates I. Kinetics of nucleation of tetracalcium monohydrogen phosphate on crystals of calcium monohydrogen phosphate dihydrate, Acta Chem. Scand. 24:1677 (1970). 6. H. E. Lundager Madsen, Calcium phosphate crystallization. IV. Kinetics of heterogeneous nucleation of tetracalcium monohydrogen phosphate on brushite crystals, Acta Chem. Scand. A 37:25 (1983). 7. A. Bigi, M. Gazzano, A. Ripamonti and N. Roveri, Effect of foreign ions on the conversion of brushite and octacalcium phosphate into hydroxyapatite, J. Inorg. Biochem. 32:251 (1988). 8. S. L. Rowles, The precipitation of whitlockite from aqueous solution, Bull. SOC. Chim. Fr. (special issue): 1797 (1968). 9. H. Fleisch, S. Bisaz and R. Russell, The activating effect of lead on the precipitation of calcium phosphate, Proc. SOC. Exp. Biol. Med. 118:882 (1965). 10. H. E. Lundager Madsen, Heterogeneous nucleation of calcium phosphates. II. Inhibition by cupric ions, Acta Chem. Scand. A 29:277 (1975). 11. A. Tovborg Jensen and J. Rathlev, Calcium hydrogen orthophosphate 2-hydrate and calcium hydrogen orthophosphate, Inorg. Syn. 4: 19 (1953). 12. H. E. Lundager Madsen and F. Christensson, Precipitation of calcium phosphate at 40°C from neutral solution, J. Crystal Growth 114:613 (1991). 13. H. E. Lundager Madsen, unpublished results. 14. H. E. Lundager Madsen, Aspects physicochimiques de la lithiase urinaire, Néphrologie 5: 151 (1984). 15. Y. Berland, M. Olmer, M. Grandvuillemin, H. E. Lundager Madsen and R. Boistelle, In vitro and clinical study of oxalate influence on calcium oxalate crystal formation, J. Crystal Growth 87:494 (1988). 16. R. M. Smith and A. E. Martell. Critical Stability Constants, Vol. 4: Inorganic Complexes, Plenum, New York (1976). 17. E. C. Moreno, W. E. Brown and G. Osborn, Solubility of dicalcium phosphate dihydrate in aqueous systems, Soil Sci. SOC. Am. Proc. 24:94 (1960); Stability of dicalcium phosphate dihydrate in aqueous solutions and solubility of octocalcium phosphate, ibid. 24:99 (1960). 18. E. C. Moreno, T. M. Gregory and W. E. Brown, Solubility of CaHPO4. 2H2O and formation of ion pairs in the system Ca(OH)2 - H3PO4 - H2O at 37.5°C, J. Res. Natl. Bur. Stand. A 70:545 (1966). 19. P. R. Patel, T. M. Gregory and W. E. Brown, Solubility of CaHPO4. 2H2O in the quatenary system Ca(OH)2 - H3PO4 - NaCl - H2O at 25°C, J. Res. Natl. Bur. Stand. A 78:675 (1974). 20. H. E. Lundager Madsen, I. Lopez-Valero, V. Lopez-Acevedo and R. Boistelle, The formation product of amorphous tricalcium phosphate at 37°C, J. Crystal Growth 75:429 (1986). 21. M. Ayati and H. E. Lundager Madsen, Crystallization of some heavy-metal phosphates alone and in the presence of calcium ion, J. Crystal Growth, in press. 22. H. E. Lundager Madsen, Theory of long induction periods, J. Crystal Growth 80:371 (1987). 23. E. I. Suvorova, F. Christensson, H. E. Lundager Madsen and A. A. Chernov, Terrestrial and spacegrow HAP and OCP crystals: effect of growth conditions on perfection and morphology, J. Crystal Growth 186:262 (1998). 24. E. I. Suvorova and H. E. Lundager Madsen, Observation by HRTEM the hydroxyapatite-octacalcium phosphate interface in crystals grown from aqueous solutions, J. Crystal Growth 198/199:677 (1999).

13

This page intentionally left blank.

INHIBITION OF CALCIUM PHOSPHATE PRECIPITATION BY POLYMERS IN THE PRESENCE OF IRON (III). THE INFLUENCE OF CHELATING AGENTS

Zahid Amjad, John F. Zibrida, Jennifer A. Thomas-Wohlever Performance Materials The BFGoodrich Company Brecksville, OH 44 14 1

ABSTRACT The formation and adherence of alkaline earth metal phosphates and phosphonates is a problem in many heat exchanger applications where the equipment is fed with waters containing high levels of calcium, phosphate, and phosphonate. Historically, polymeric and non-polymeric additives have been used to prevent the formation and deposition of scale forming salts. In the present study the effect of various polymeric and non-polymeric additives as Fe (III) chelants have been evaluated for their effect on the performance of calcium phosphate inhibiting polymers. It has been observed that polymer composition and polymer dosage have a significant impact on the performance of calcium phosphate inhibiting polymers especially in the presence of low concentration of Fe (III). The data on the effect of various polymeric and non-polymeric additives as chelants for Fe (III) show that additives containing hydroxyl and/or carboxyl groups exhibit excellent chelating ability.

INTRODUCTION The occurrence of various types of mineral scales in industrial water systems such as desalination (thermal and membrane-based), boiler, recirculating and once-through, etc. is very common due to the presence of various dissolved and suspended impurities in feed streams. Scale deposits such as calcium sulfate and calcium carbonate form readily on flow surfaces such as reverse osmosis membranes, heat exchangers, and other process equipment handling supersaturated waters. When deposited on equipment surfaces, the mineral scales coupled with corrosion products, microbiological mass, and suspended matter (i.e., clay, silt, organic debris, etc.) result in reduced heat transfer efficiency, material damage, blocked pipes, and increased maintenance cost. In desalination processes

Advances in Crystal Growth Inhibition Technologies Edited by Amjad Kluwer Academic/Plenum Publishers. New York. 2000

15

the consequences of unwanted deposits on reverse osmosis membrane surfaces range from reduced quality and quantity of produced water and increased cleaning cost to pre-mature membrane failure. Polymeric and non-polymeric agents are often used to inhibit the corrosion of metal surfaces and the growth of mineral crystals. For many years, chromate-based corrosion inhibitors were considered the performance standard in the cooling water treatment industry. Unfortunately, chromates are toxic and due to environmental regulations their use has declined considerably in recent years. Accordingly, there has been a trend away from chromate-based corrosion inhibitors. Among the more environmentally acceptable mild steel corrosion inhibitors which have been displacing chromates since the 1960s are silicates, molybdates, phosphate/polyphosphates, phosphonates, and zinc salts. From the perspective of their versatility, cost, and performance, phosphates have become the new performance standard. However, phosphate-based cooling water treatment programs require more careful control of system parameters than do chromate programs, mainly because they rely upon relatively high phosphate levels for corrosion protection. These high phosphate levels, especially in cooling systems operating at alkaline pH, can lead to calcium phosphate precipitation and fouling of heat exchangers. In many cases cooling systems operating under high cycles of concentration due to water conservation efforts dramatically increase the potential for calcium phosphate scale formation, even with the relatively low phosphate concentration present in the makeup water. It has been suggested that the presence of polymer in the water treatment formulation serves a dual role. It controls the thickness of the calcium phosphate film on the metal surface and also prevents precipitation of the calcium phosphate in the recirculating water.1 Over the years, different polymers have been developed and are extensively used as calcium phosphate scale inhibitors. Studies have shown that polymer performance strongly depends upon polymer composition (type and amount of comonomers), molecular weight, and ionic charge. Typically, an effective scale inhibitor contains at least one of the following functional groups: carboxylic acid, -COOH; acrylamide, -CONH2; sulfonic acid, -SO3H; ester, -COOR; phosphonic acid, -PO3H2, etc. The molecular weight of an effective polymeric calcium phosphate inhibitor is generally below 15,000. It is generally believed that these polymeric inhibitors operate by adsorption onto mineral scale surfaces at growth sites, thereby preventing both further crystal growth and deposition on heat exchanger surfaces.1 The influence of both dissolved and suspended impurities present in the make-up water on the performance of calcium phosphate inhibiting polymers has been the subject of extensive research. Results of a recent study3 have shown that the presence of clay in calcium phosphate supersaturated solutions markedly reduces the performance of calcium phosphate inhibiting polymer. Presumably, this is due to a reduction in polymer concentration in solution as a result of adsorption onto clay particles. It is generally agreed that the presence of soluble metal ions (i.e., aluminum, iron, manganese, zinc, etc.) in the water can lead to heat exchanger fouling in cooling water system.4 It is also well documented that the presence of low levels (0.5 to 5 ppm) of iron (III), aluminum, and cationic polymer due to the “carryover” of coagulants/flocculants in the feed stream has a marked antagonistic effect on the performance of calcium phosphate inhibiting polymers. Thus, the presence of impurities in the water may require additional amounts of polymer for inhibiting scale formation and growth.5,6 Natural organic compounds (i.e., humic substances) are commonly encountered in surface and groundwater used for water supply. Humic substances are mixtures of natural products with a complex structure and chemistry. Isolation and characterization of humic substances from surface and groundwater is an ongoing area of research interest. Recently, the influence of humic substances on crystal growth of sparingly soluble salts has attracted

16

the attention of several researchers. Freche et al.7 and Amjad et al.8 in their studies on the influence of humic acid as crystal growth inhibitor for dicalcium phosphate dihydrate and hydroxyapatite reported that humic acid reduced crystal growth rates. A Langmuiradsorption process at surface growth sites was proposed to explain the inhibitory effect of humic substances. Effects of polyelectrolytes (i.e., poly(acrylic acid), poly(aspartic acid), fulvic acid, tannic acid, etc.) on calcium carbonate precipitation has also been reported. X-ray diffraction studies of calcium carbonate precipitated in the presence of polyelectrolytes reveal that whereas poly(acrylic acid) and poly(aspartic acid) favor the formation of vaterite, fulvic acid and tannic acid influence the formation of calcite. The polymorph formed in the absence of polyelectrolytes is vaterite.9 In the present work, we have studied the effect of acrylic acid-based co- and terpolymers on the precipitation of calcium phosphate in the absence and presence of low levels of iron (III). The influence of functional group i.e., carboxyl, sulfonic, substituted acrylamide, and aromatic, on the inhibitory power of polymer was tested. For the quantitative assessment of the effectiveness of chelating agents precipitation experiments in the presence of ter-polymer and iron (III) were done by the pH-stat method, which gives the most reliable and reproducible results. The chelants tested include: A. Hydroxy/carboxylic acids i.e., fulvic acid, tannic acid, gluconic acid, glucose, salicylic acid, benzene hexacarboxylic acid, tricarballylic acid, B. Phosphonates i.e. 2phosphonobutane 1,2,4-tricarboxylic acid, aminotris(methylene phosphonic acid), hydroxyethylidine 1,1 -diphosphonic acid, and C. polymers i.e., poly(acrylic acid), poly(maleic acid), poly(acryalmide), poly(2-acrylamido 2-methylpropane sulfonic acid) and acrylic acid containing co- and terpolymers. It was hoped that the study of such a variety of chelating agents would not only quantify the effectiveness of these chelating agents, but also throw some light on the mechanism of inhibition of calcium phosphate scale formation in aqueous systems.

EXPERIMENTAL The chemicals used to prepare the solutions were Fisher Scientific ACS reagent grade. Stock solutions of known concentrations of calcium chloride and disodium phosphate, hydrochloric acid, sodium hydroxide, chelating agents and various polymers were prepared and used to makeup test solutions. The polymeric inhibitors were selected from commercial materials. All inhibitor solutions were prepared on dry weight basis. The desired concentrations were obtained by dilution, using double de-ionized and distilled water. A Metrohm Brinkmann pH-stat unit equipped with a combination electrode was used to maintain the experimental solution pH. The pH electrode was calibrated before each experiment with standard buffers. The test solution temperature was maintained to 50 ± 1°C by passing controlled water through the outer jacket of the jacketed reaction cell. The subsaturated calcium phosphate solution was made by the slow addition of the phosphate stock solution to the desired water. The polymer, chelating agent, and other ionic constituents were added and the solution pH was adjusted if necessary, after a 30minute equilibration period. The calcium stock solution was added, making up the total volume to 600 mL. Experiments involving the effect of soluble iron a known amount of ferric chloride was added to the test solution following temperature stabilization but prior to pH adjustment and before the introduction of the calcium solution. For all experiments the reaction period was fixed at 20 hours. During each experiment, the test solution pH was maintained (within ± 0.01 pH units) at the desired level using the pH-stat unit. Unless specified, the standard test conditions for the calcium phosphate inhibition were 140 mg/L, [Ca], 9.0 mg/L phosphate, pH 8.50, 50°C, and 20 hr. The pH-stat apparatus used is illustrated in Figure 1.

17

Figure 1. pH-stat apparatus used in calcium phosphate scale inhibition studies.

The reaction progression was determined by spectrophotometric analysis of filtered (0.45 micron) aliquots of the test solution for the phosphate ion.5 The efficacy of the polymer was calculated as percent inhibition as illustrated in Equation 1.

[PO4]s

I=

---

[PO4]I --where: I [PO4]S [PO4]C [PO4]I

[PO4]c

-----------------------------------

= = = =

[PO4]c

x 100%

(1)

Inhibition, % Phosphate concentration in the presence of polymer at t =20 hr Phosphate concentration in the absence of polymer at t = 20 hr Phosphate concentration at the beginning of the experiment.

RESULTS AND DISCUSSION The use of polymers in water treatment formulations has increased significantly over the last several decades. The role of polymer in these formulations is to inhibit the precipitation of scale forming salts and also to prevent the deposition of unwanted materials on the equipment surfaces. It is generally agreed that these polymers operate by adsorption onto the mineral surface at growth sites, thereby preventing further crystal formation and deposition. Most commercially available polymers are acrylic acid or maleic acid based copolymers that have proven to be effective under most typical cooling water conditions. However, the performance of these copolymers has been limited by the recent trend toward operating cooling systems under more severe conditions to increase process efficacy, safety, and water conservation. Additionally, results of recent studies have shown that the efficacy of these copolymers is adversely impacted by the presence of low levels of impurities commonly encountered in feed streams.2,3,5,6 The following section presents comparative performance data on several acrylic acid-based homo-, co-, and terpolymers. The polymers used in this investigation are summarized in Table 1. Chelating agents and phosphonates used in further studies are summarized in Tables 2 and 3, respectively.

18

Table 1. Polymeric inhibitors of calcium phosphate scale used in this study. Acronym

Polymer

Structure

Poly-A

Poly(acrylic acid)

Poly-B

Poly(acrylic acid:diacetone acrylamide)

Poly-c

Poly(acrylic acid:2-acrylamido2-methylpropane sulfonic acid: sulfonated styrene)

Table 2. Chelating agents evaluated in this study. Acronym

Polymer

BHCA

Benzene Hexacarboxylic Acid

GLA

Gluconic Acid

GL

Glucose

SA

Salicylic Acid

CA

Citric Acid

FA TA

Fulvic Acid Tannic Acid

Structure

phenolic, carboxyl groups phenolic, carboxyl groups

Table 3. Phosphonates evaluated in this study. Acronym

Phosphonate

AMP

Arninotris(methylene phosphonic acid)

HEDP

Hydroxyethylidine 1,1 -diphosphonic acid

PBTC

2-phosphonobutane 1,2,4-tricarboxylic acid

Structure

Effect of Polymer Dosage The effect of varying concentrations of Poly-A, poly(acrylic acid), Poly-B (acrylic acid co-polymer) and Poly-C (acrylic acid ter-polymer) on calcium phosphate precipitation is illustrated in Figure 2. The results indicate that Poly-A exhibits poor performance as calcium phosphate inhibitor. For example, % Inhibition value obtained in the presence of 10 ppm (parts per million) of Poly-A is ~ 8%. As shown, an increase in the polymer concentration by a factor of two and one half (from 10 to 25 ppm) results in a five fold increase in the polymer's ability to inhibit calcium phosphate. Figure 2 also shows the 19

calcium phosphate inhibition data on co- and ter-polymer. It is apparent that acrylic acid homopolymer is not particularly effective as a calcium phosphate inhibitor. However, when acrylic acid is copolymerized with diacetone acrylamide, Poly-B, the performance is dramatically enhanced. The effectiveness of Poly-C (a terpolymer of acrylic acid, 2acrylamido-2-methylpropane sulfonic acid, and sulfonated styrene) as a calcium phosphate inhibitor is surprisingly better than Poly-B, suggesting that both the ionic charge and the bulk of the comonomers exert a positive influence in improving the overall inhibitory activity of the polymer. From the data shown in Figure 2, the ranking of the polymers (in terms of effectiveness) is: terpolymer > copolymer > homopolymer.

Figure 2. Percent inhibition of calcium phosphate scale by varying levels of Poly-A, Poly-B, or Poly-C (140 mg/L Ca, 9 mg/L PO4, pH 8.5, 50ºC, 20 hours).

Figure 3. Percent inhibition of calcium phosphate scale by varying levels of Poly-A, Poly-B, and Poly-C in the presence of 3 ppm Fe (140 mg/L Ca, 9 mg/L PO4, pH 8.5, 50°C, 20 hours).

Effect of Iron(III) on Polymer Performance The influence of trace levels of metal ions (i.e., copper, zinc, iron, manganese, etc.) on the precipitation of scale forming salts has been the subject of numerous investigation.3,10,11 It has been documented that polymer performance as calcium 20

phosphate inhibitor is reatly affected by the presence of iron(III), aluminum, and cationic polymeric flocculant.12 In these cases polymer must not only prevent calcium phosphate scale formation but must also act as a very good dispersant for iron oxide particles formed as a result of iron hydrolysis and/or corrosion products. Deposition of these unwanted materials must be minimized in order to prevent under-deposit corrosion and increased pitting.13 In order to evaluate the impact of soluble iron (III) a series of precipitation experiments were conducted under standard test conditions. Figure 3 shows the effect of 3 ppm iron (III) on the performance homo-, co-, and ter-polymers. By comparing the inhibition data versus dosages for Poly-A, Poly-B, and Poly-C, it is evident that presence of 3 ppm of iron (III) has an antagonistic effect on the inhibitory power of these polymers. The data presented in Figure 3 also reveal that additional amounts (i.e, 25 % for co- vs. 15 % for ter-polymer) of polymers are needed to successfully inhibit the calcium phosphate precipitation. The results presented in Figure 3 clearly show that compared to homo- and co-polymers, ter-polymer exhibits superior performance in inhibiting calcium phosphate precipitation, especially in the presence of iron (III).

Effect of Chelating Agents As mentioned previously, cooling water is generally contaminated with various forms of oxidized iron due to corrosion of steel equipment and/or its introduction with the feed water.4,14 Various inorganic, organic, and mixed organic acids are used for the removal of deposits in general cleaning operations in industry (heat exchanger, boiler, reverse osmosis and untrafiltration membranes), in homes (laundry machines, dishwashers), and in laboratories. The most commonly used acids and their salts are: hydrochloric acid, phosphoric acid, sulfuric acid, organic acids (i.e., oxalic, citric, lactic, malic, sulfamic, gluconic), and acid salts such as sodium and ammonium citrates. In order to study the effectiveness of chelants on calcium phosphate inhibiting polymers in the presence of Fe(III), the concentration of ter-polymer chosen was 10 ppm. As illustrated in Figure 4, the % inhibition value obtained in the presence of 10 ppm of ter-polymer and 3 ppm Fe(III) is 75%. The results presented below were collected under these standard conditions and with varying concentration of chelants.

Humic Compounds. Tannins are a class of compounds that have been investigated for use in on-line removal of iron deposits. These polymers are attractive because they contain hydroxyaromatic substituents which are available for binding to ferric ions. To study the effect of humic compounds a number of calcium phosphate precipitation experiments were carried out in the presence of humic compounds and iron(III). The inhibition data presented in Figure 4 clearly shows that compared to Poly-C, both fulvic and tannic acids are poor calcium phosphate inhibitors when iron is not present. It is interesting to note that whereas it takes ~10 ppm of Poly-C to achieve > 90% inhibition, the inhibition values obtained in the presence of 10 ppm fulvic and tannic acids are < 10%. It is clear from this data that phenolic and carboxyl groups present in humic compounds do not contribute significantly to the inhibitory power of fulvic and tannic acids. In view of the above results indicating that humic compounds exhibit poor inhibitory activity for calcium phosphate precipitation, additional experiments were performed to evaluate the iron(III) chelation ability of these compounds. Results of calcium phosphate inhibition experiments carried out in the presence of 10 ppm Poly-C, 3 ppm iron (III), and varying concentration of humic compounds are illustrated in Figure 4. It can be seen that compared to tannic acid, fulvic acid is a better chelating agent for iron (III). It is worth

21

noting that it took 5 ppm FA to yield >96 % inhibition compared to 10 ppm TA to achieve similar results. It should be noted that similar superior performance of FA has been reported in earlier seeded growth studies involving the evaluation of humic compounds as inhibitors for calcium phosphates and calcium carbonate.8,15

Figure 4. Percent inhibition of calcium phosphate scale formation by Poly-C and humic compounds, with and without 3 ppm Fe (10 ppm Poly-C, 140 mg/L Ca, 9 mg/L PO4, pH 8.5, 50°C, 20 hours)..

Hydroxy/Carboxylic Acids. Several studies have been reported pertaining to the influence of di- and tri-carboxylic acids on the crystal growth of scale-forming salts. Meyer and Sellinger16 in their study on the effect of citrate ion on calcium phosphate phase transitions have shown that citrate ions appear to have only a minimal effect on the rate of amorphous to crystalline transformation, through a small, but measurable, stabilizing effect on the amorphous phase at higher concentrations (>20mM). Nancollas et al.17 investigated the influence of tricarboxylic acids on calcium phosphate precipitation using the seeded growth method. Results of this study reveal that citric acid is a better inhibitor than tricarballylic acid in inhibiting the crystal growth of calcium phosphate acids on the precipitation of amorphous calcium phosphate. Recently, Amjad investigated the influence of a variety of hydroxyl and carboxyl groups containing acids on the rate of crystal growth of calcium fluoride. Comparison of the effect of citric, isocitric acid, and tricarballylic acid suggest that the hydroxyl group in the molecular backbone plays a key role in the effectiveness of these tricarboxylate ions as inhibitors.l8 Figure 5 presents inhibition data on the influence of various acids containing hydroxyl and carboxyl groups. It can be seen that compared to Poly-C, all the chelating agents evaluated show poor to moderate activity for calcium phosphate inhibition. Among the non-polymeric chelants (i.e., citric acid, CA, benzene hexacarboxylic acid, BHCA, salicyclic acid, SA, gluconic acid, GLA), BHCA shows the best performance as a calcium phosphate inhibitor. The data presented in Figure 5 suggest that both the number of hydroxyl and/or carboxyl groups present in the chelant influence the inhibitory power of the chelants. As illustrated in Figure 5, glucose, which is devoid of carboxyl groups compared to gluconic acid, is an ineffective calcium phosphate inhibitor. It is worth noting 22

that similar trend in chelants performance was observed in the seeded growth of calcium fluoride and calcium phosphates. Figure 6 presents results on the performance of Poly-C in the presence of 5 ppm chelating agents and 3 ppm of Fe(III). It can be seen that both citric acid and benzene hexacarboxylic acid, compared to gluconic acid and salicyclic acid, appears to have a positive effect on the prevention of precipitation of calcium phosphate in the presence of iron (III).

Figure 5. Percent Inhibition of -OH and -COOH containing acids vs. Poly-C(10 ppm inhibitor, 140 mg/L Ca, 9 mg/L PO4, pH 8.5, 50°C, 20 hours).

Figure 6. Effect of chelating agents containing -OH and -COOHon percent inhibition of calcium phosphate scale by Poly-C(10 ppm Poly-C, 140 mg/L Ca, 9 mg/L PO4, pH 8.5, 50°C, 20 hours).

Phosphonates The influence of trace amounts of polyphosphates and phosphonates on the precipitation of calcium phosphate has been the subject of numerous investigations. Results of previous studies have shown that hydroyethylidine 1,1 -diphosphonic acid, HEDP, is an effective inhibitor when added to barium sulfate, calcium fluoride, calcium carbonate, and calcium oxalate solutions. 19-21 However, with more soluble calcium sulfate dihydrate, this inhibitor appeared to have only a slight effect on the growth kinetics. Varsanik 22 reported that among various phosphonates studied in laboratory and pilot plant 23

experiments, HEDP was the most effective inhibitor for limiting calcium phosphate scale formation. However, system pH had a dramatic effect upon inhibitor effectiveness. At neutral pH values, the performance of aminotri(methylene phosphonic acid), AMP, was comparable to that HEDP. Amjad23 in another also reported that HEDP performed better than AMP as calcium phosphate inhibitor. Certain phosphonates have been used to remove iron oxide deposits and leave a surface which is passivated against further corrosion. To evaluate the effectiveness of phosphonates as iron chelating agents, a series of calcium phosphate precipitation experiments were carried out in the presence of chelating agents. Results of these precipitation experiments are shown in Figure 7. It can be seen that chelants performance as calcium phosphate inhibitor depends upon the chelant concentration present in calcium phosphate supersaturated solution. For example, in the presence of 5 ppm phosphonates the percent calcium phosphate obtained is <10% suggesting that under these experimental conditions phosphonates are ineffective calcium phosphate inhibitors.

Figure 7. Effectiveness of polyphosphonates as calcium phosphate inhibitors vs. Poly-C (140 mg/L Ca, 9 mg/L PO4, pH 8.5, 50°C, 20 hours).

Figure 8 presents results on calcium phosphate inhibition by Poly-C in the presence of 3 ppm Fe(III) and 2 ppm of phosphonates. It is evident that even at low concentrations (i.e., 2 ppm), phosphonates have a marked effect on the performance of Poly-C. It is interesting to note that under the experimental conditions employed HEDP, compared to AMP and PBTC, appears to exhibit the most adverse impact on the performance of Poly-C. The marked antagonistic effect by HEDP may be due to the poor tolerance of HEDP to calcium ions and/or co-precipitation of calcium phosphate and calcium-HEDP salts. A study is currently underway to further understand the role of phosphonate in the precipitation and inhibition of calcium phosphate by polymers. Homo- Co-, and Terpolymers. In addition to organic acids, homopolymers and copolymers of acrylic acid have been tried in the removal of deposits from fouled heat exchanger and membrane surfaces. Lipinski and Chang in their study on the evaluation of a variety of polymers as scale removal agents from heat exchanger surfaces reported that acrylic acid containing polymers were effective in dissolving calcium carbonate and calcium phosphate scales. Recently, Kreh et al.14 reported that among the additives tested, the polymers containing sulfonate and carboxylate functionality showed superior performance for on-line prevention of rust scale by maintaining soluble ferric ions and dispersing any precipitated iron oxide.

24

Figure 8. Percent inhibition by Poly-C in the presence of 3 ppm Fe (III) and 2 ppm phosphonates(l0 ppm Poly-C, 140 mg/L Ca, 9 mg/L PO4, pH 8.5, 50ºC 20 hours).

Figure 9 presents calcium phosphate inhibition data by several polymers later evaluated as iron chelators. It can be seen that the with the exception of Poly-C all polymers evaluated at 10 ppm polymer concentration show poor performance as calcium phosphate inhibitors. For example, % inhibition value obtained for Poly-C is >90% compared to <10 % obtained for Poly-A. As illustrated in Figure 9, % inhibition values for homopolymers P-SA and P-AM at 10 ppm are <5%. The copolymer Poly-B gives intermediate inhibition data. In order to study the effect of polymers as chelating agents for Fe (III) a series of experiments were carried out in the presence of 10 ppm of Poly-C, 3 ppm Fe(III) and varying concentration of homo- , co-, and terpolymers. The data presented in Figure 10 on homo- and co-polymers of acrylic acid show that compared to homo acrylic acid polymer (Poly-A), acrylic acid-based copolymer, Poly-B is a better iron chelating agent. For example, to achieve >95% inhibition it requires 1.5, 2.0, and 2.5 ppm respectively of ter-, co-, and homo- polymers. Under similar experimental conditions to achieve > 95% inhibition value, it required 5 ppm of Poly-AM and Poly-SA. Clearly, terpolymer exhibits superior performance in inhibiting the precipitation of calcium phosphate and in overcoming the antagonistic effect of iron (III), under the experimental conditions tested.

Figure 9. Calcium phosphate inhibition by homo-, co-, and terpolymers.

25

Figure 10. Calcium phosphate inhibition by Poly-C in the presence of 3 ppm Fe(III) and varying concentrations of homo-, co-, and terpolymers.

CONCLUSIONS In this paper we have demonstrated that low levels of both polymeric and non-polymeric additives can significantly inhibit the precipitation of calcium phosphate from aqueous solution. The results of the present study also suggest that the inhibition of calcium phosphate precipitation strongly depends upon the concentration of polymer and polymer composition. It has also been observed that low levels of Fe (III) adversely affects the performance of calcium phosphate inhibiting polymers. In addition, it has also been shown that antagonistic effect exhibited by Fe (III) can be overcome by the addition of homo-, co-, and ter-polymers as well as by non-polymeric chelants. The ranking in terms of effectiveness as various additives evaluated as iron chelants is:

• Polymeric additives: ter-polymer > co-polymer > homo-polymer • Non-polymeric additives: benzene hexacarboxylic acid > hydroxy carboxylic acid > phosphonates

REFERENCES 1. L. Perez, Mechanism of Calcium Phosphate Scale Formation and Inhibition in Cooling Systems, Chapt. 17 in Calcium Phosphates in Biological and Industrial Systems, Amjad, Z., Ed.,Kluwer Academic Publishers, Boston, MA (1998). 2. Z. Amjad, Development of Calcium Phosphate Inhibiting Polymers for Cooling Water Applications, Chap. 16, in Calcium Phosphates in Biological and Industrial Systems, Amjad, Z. Ed., Kluwer Academic Publishers, Boston, MA (1998). 3. Z. Amjad, D. Butala, and J. Pugh, The Influence of Recirculating Water Impurities on the Prerformance of Calcium Phosphate Inhibiting Polymers, Paper NACE/99, No. 99 1 18, NACE International, Houston, TX (1999). 4. Z. Amjad, Controlling Metal Ions Fouling in Industrial Water Systems, ULTRAPURE WATER, accepted for publication (1999). 5. 2. Amjad, J. Pugh, J. Zibrida, and R. Zuhl. Polymer Performance in Cooling Water: The Influence of Process Variables, Material Performance 36:32 (1997). 6. J.E. Hoots, K.P. Fivizzani, and R.W. Cloud, Influence of Mechanistic Studies on the Development of New Cooling Water Programs, CORROSION/89, Paper No. 175, National Association of Corrosion Engineers, Houston, TX (1989). 7. M. Freche, N. Rouquet, P.G. Koutsoukos, and J.L. Lacout, Effect of Humic Compounds on the Crystal Growth of Dicalcium Phosphate Dihydrate, Agrochimica, 36:500 (1992).

26

8. Z. Amjad, M. Reddy, Influence of Humic Compounds on the Crystal Growth of Hydroxyapatite, Chap. No. 7: in Water Soluble Polymers: Solution Properties and Applications, Amjad, Z. Ed., Plenum Publishing Corporation, New York, New York (1998). 9. Z. Amjad, Precipitation of Calcium Carbonate in Aqueous Systems: The Influence of Natural and Synthetic Polyelectrolytes, Tenside Surfactant Detergents, accepted for publication (1999). 10. S.K. Hamdona, R.B. Nesim, and S.M. Hamza, Spontaneous Precipitation of Calcium Sulfate Dihydrate in the Presence of Some Metal Ions, Desalination 94:69 (1993). 11. P.G. Koutsoukos, P. G., Influence of Metal Ions on the Crystal Growth of Calcium Phosphates, Chap. 7 in Calcium Phosphates in Biological and Industrial Systems, Amjad, Z., Ed., Kluwer Academic Publishers, Boston, MA (1998). 12. Z. Amjad, Calcium Phosphate Inhibition by Polymeric Inhibitors: The Influence of Aluminum (III), Iron (III), and Iron (II) Ions, J. Phos. Res. , accepted for publication (1999). 13. L. Perez, Novel Calcium Phosphate Scale Inhibitor, Chap. 12 in Water Soluble Polymers: Solution Properties and Applications, Amjad, Z., Ed., Plenum Publishing Corporation, New York, New York (1998). 14. R. Kreh, W. Henry, J. Richardson, and V. Kuhn, The Use of Chelants and Dispersants for Prevention and Removal of Rust Scale, Chap. 13, in Mineral Scale Formation and Inhibition, Plenum Publishing Corporation, New York, New York (1995). 15. Z. Amjad, J. Pugh, and M. Reddy, Kinetic Inhibition of Calcium Carbonate Crystal Growth in the Presence of Natural and Synthetic Organic Inhibitors, Chapt. 11 in Water Soluble Polymers: Solution Properties and Applications, Plenum Publishing Corporation, New York, New York, 1998 16. J.L. Meyer, and A.H. Selinger, Effect of Citrate on Calcium Phosphate Phase Transformation, Miner. Electrolyte Metab. 3:207 (1980). 17. G.H. Nancollas, and M.B. Tomson, The Precipitation of Biological Minerals, Faraday Discussion Chem. Soc., 61:176 (1976). 18. Z. Amjad, Performance of Inhibitors in Calcium Fluoride Crystal Growth Inhibition, Langmuir 9:597 (1993). 19. P. Dubin, The Effect of Organophosphorus Compounds and Polymers on Calcium Carbonate Crystal Morphology, Paper No. 222, CORROSION/80, National Association of Corrosion Engineers, Houston, TX (1980). 20. S.M. Hamza, and G.H. Nancollas, Kinetics of Dissolution of Magnesium Fluoride in Aqueous Solution, Langmuir 1:573 (1985). 21. G.L. Gardner, Effect of Pyrophosphate and Phosphonate Anions on the Crystal Growth of Calcium Oxalates Hydrates, J. Phys. Chem. 82:864 (1978). 22. Z. Amjad, Calcium Sulfate Dihydrate (Gyspsum) Scale Formation on Heat Exchanger. Surfaces. The Influence of Scale Inhibitors, J. Colloid Interface Sci . 123:523 (1988). 23. R.G. Varsanik, Calcium Phospate in Industrial Water Systems. Muter. Perform, 14:16 (197.5). 24. Amjad, Z., Effect of Precipitation Inhibitors on Calcium Phosphate Scale Formation, Can. J. Chem. 67, 850, 1989

27

This page intentionally left blank.

INTERACTION OF SODIUM POLYACRYLATE WITH OCTACALCIUM PHOSPHATE

A. Bigi1 , E. Boanini2, G. Cojazzi3, G. Falini1, S. Panzavolta1, N. Roveri1 1

Department of Chemistry "Ciamician" Research Unity of Bologna CSFM (CNR), c/o Dept. Of Chemistry "Ciamician" Bologna University, 40126 Bologna, Italy

2 C.I.R.C.M.S.B., 3

INTRODUCTION Octacalcium phosphate, Ca8H2(PO4)6 • 5H2O (OCP), often occurs as a transient intermediate in the precipitation of the thermodynamically more stable hydroxyapatite (HA)1, and has been suggested to be involved in the mineralization of biological tissues, such as bone and dentine.1-3 The inorganic phase in these calcified tissues is a poor crystalline carbonated apatite, which, however, could be just the final phase of a process where OCP acts as a precursor phase. HA crystallizes in the hexagonal space group P63/m. Biogenic apatite crystals from bone, dentine and calcifying enamel exhibit plate or ribbon-like shapes and are elongated along the c-axis.4-6 The plate-like habit, which is not consistent with apatite symmetry, is one of the reason why OCP is thought to be involved in the first stages of mineralization of bone and tooth tissues.7 In fact, OCP crystallizes as {100} blades of triclinic pinacoidal symmetry, elongated along the c-axis and bordered by the forms {0l0}, {001} and {011}.4 Although the triclinic structure of OCP, space group P1, does not exhibit any sixfold symmetry relation, its similarity with HA structure is remarkable. The unit cell consists of a layer practically identical to the unit cell of HA and a layer with more widely spaced PO43- and Ca2+ ions and H2O molecules interdispersed. Apatitic layers, about 1.1 nm thick alternate with hydrated layers, about 0.8 nm thick, parallel to the (100) face.8 Figure 1 shows the projections along the (100) plane of the structures of OCP and HA, together with a tridimensional scheme of the layered OCP structure. OCP hydrolyzes in water to HA. It has been suggested that the transformation of OCP into HA takes place through in situ hydrolysis of OCP, and/or dissolution of OCP followed by HA precipitation.2,8-10 The involvement of the exchange

Advances in Crystal Growth Inhibition Technologies Edited by Amjad Kluwer Academic/Plenum Publishers New Y ork 2000

29

Figure 1. Projections along the (100) face of the structures of OCP and HA, together with a tridimensional scheme showing the alternating “apatitic” and “hydrated” layers parallel to the (100) face of OCP. The dark spheres indicate the Ca atoms, the white spheres show the oxigens atoms of water molecules in OCP, as well as of OH group in HA, and the shadowed tetrahedra indicate the phosphate groups.

of ions and water molecules between solid and solution phases,8 as well as the great influence of pH on the transition rate,9,10 suggests that the dissolution/reprecipitation process is the most probable. Fluoride, carbonate, citrate, as well as several mono and divalent cations, have been found to influence the conversion of OCP into HA.11-16 Furthermore, the same ions, when present during the synthesis of OCP, can affect its growth and morphology: the presence of magnesium, carbonate , and fluoride, for instance, suppresses the growth of OCP along the c-axis under most conditions.l7-19 OCP structure can incorporate a large variety of inorganic and organic species.1,20 Thus, the presence of impurities which stabilize or destabilize OCP21,22 could affect not only the relative rate of crystallization and hydrolysis, but also the composition and structure of the calcium phosphate finally formed.13,23 Ionic incorporation during OCP transformation into HA could account also for the nonstoichiometry and poor crystallinity of biological apatites.3 OCP has been found in dental calculi24,25 as well as in other pathological calcifications.26 Furthermore, high resolution electron microscopy images of calcifying dentine revealed the presence of OCP in the central part and of HA at the extremities of the same crystal.6 It must be remembered that the process of mineralization of biological tissues takes place in an environment rich in acidic macromolecules, which can affect crystallization and growth of the mineral phase.27 Because of the presence of many charged groups, such as carboxylate, phosphate and sulphate groups, the acidic macromolecules can interact with the charged crystal surfaces. The interaction can be specific, with the protein adopting a conformation and exhibiting charge distribution and repeating

30

distances matching some structural motif of the mineral crystal, or even non-specific, due to multiple electrostatic interaction.28 The results of in vitro experiments carried out on OCP crystallization suggest a specific interaction of the highly phosphorylated acidic protein, phosphophoryn, with the (010) face, whereas carboxylate rich proteins seem to interact preferentially with the hydrated layer of the (100) face.28 It has been verified that some synthetic polyelectrolytes can affect the nucleation and growth of calcium phosphates playing a role similar to that played in vivo by biological macromolecules.29 In particular it has been verified that the presence of polyacrylic acid, as well as its sodium salt, inhibits the synthesis of hydroxyapatite.29,30 The presence of the carboxylate-rich polyelectrolyte affects also the morphology of the crystals, suggesting that its significant adsorption on HA crystals takes place through interaction with the phosphate sites on the (100) faces.31 Furthermore, polyacrylic acid, as well as its sodium salt, is easily and irreversibly adsorbed on HA through a process which has been ascribed to electrostatic interactions and/or hydrogen bonding forces.32

OCTACALCIUM PHOSPHATE Octacalcium phosphate was synthesized in aqueous medium by reaction of calcium acetate with sodium phosphate under stirring at 60°C, starting pH 5.21,33 Due to the similarity between the structure of OCP and HA, the positions of most of the powder Xray diffraction reflections present in the pattern reported in fig. 2 are coincident with those characteristic of HA. However, the strong peak at about 4.7º, due to the (100) reflection of OCP, as well as the relative intensity distribution of the reflections, allows to identify the synthetic product as octacalcium phosphate, with cell parameters , a = 19.62(5) Å b = 9.62(9) Å c = 6.83 Å α = 89.4(8)º, β = 92.6(8)º, γ =107.4(8)º, very close to those reported for OCP single crystal.34 At the scanning electron microscope, OCP appears as spherules made up of long blades originating from a common center (fig. 3a,b). Next to these aggregates, which are characteristic of OCPI3 and exhibit a mean diameter of about 100 µm, couples of spheres with diameter of 100-200 µm are present. The spheres seem to have an inner hollow core and thick walls constituted of short blades arranged almost orthogonally to the surface of the sphere (fig. 3c,d). OCP hydrolysis in water for 48 hours at 60°C, starting pH 7.4, leads to a complete transformation into HA, which exhibits lattice constants a=9.4 17(7)Å and c=6.87(1)Å.

Figure 2. Powder X-ray diffraction pattern of OCP.

31

Figure 3. SEM images of OCP. Spherules constituted of long blades originating from a common center (a,b), as well as couples of spheres, whose surfaces are made up of shorter blades (c,d), are present. Bars = (a) 50µm; (b) 20µm; (c) 100µm; (d) 25µm. Reproduced from ref. 33, with permission from Elsevier Science.

The transmission electron microscopy investigation carried out on OCP before storage in water reveals a plate-like morphology of the crystals, which exhibit straight edges and mean dimensions of 2 x 8 µm (fig. 4a). HA crystals obtained by OCP hydrolysis appear as long platelets with straight edges and quite great mean dimensions (about 70-80 x 700-800 nm), as it can be appreciated in fig. 4b.

Figure 4. TEM image of (a) OCP, bar = 2 µm; (b) HA obtained by OCP hydrolysis in water, bar = 0.2 µm.

EFFECT OF SODIUM POLYACRYLATE ON THE SYNTHESIS OF OCP Low molecular weight sodium polyacrylate (MW 2100) inhibits the synthesis of OCP from aqueous solution: OCP can be obtained as unique crystalline phase up to a NaPA concentration of 2 µM, while the synthesis is completely inhibited at concentrations higher than 5µm.33 The values of the lattice parameters do not show any significant variation as a function of NaPA concentration in solution, as reported in Table 1.

32

Table 1. Lattice constants of OCP synthesized in presence of different NaPA concentrations. Standard deviations in the last significant figure are given in parentheses. [NaPa] (uM) -0.375 0.5 1 2

a-axis(Å)

b-axis(Å)

19.62(5) 19.62(9) 19.70(4) 19.63(1) 19.61(6)

9.62(9) 9.61(9) 9.61(6) 9.57(7) 9.63(9)

c-axis(Å)

a (°)

B (°)

Y (°)

6.83(6) 6.83(6) 6.82(7) 6.87(5) 6.84(6)

89.4(8) 89.5(9) 89.6(9) 88.9(9) 89.5(9)

92.6(8) 92.4(9) 92.3(9) 92.1(8) 92.2(9)

107.4(8) 107.5(8) 107(1) 107.9(5) 107.4(8)

However, the coherence lengths of the perfect crystalline domains reduce on increasing the polymer concentration in solution, as results from the line broadening of the (100), (010) and (002) reflections. The Dhkl values reported in Table 2, which were calculated from the width at half maximum intensity (B½) using the Scherrer equation,35 indicate that the reduction is almost isotropic along the three different direction and suggest that the interaction with OCP does not take place along specific directions. In Table 2. Coherence lengths (Dhkl) evaluated from the width at half maximum intensity of the (100), (010) and (002) reflections of OCP as a function of NaPA concentration in solution.a [NaPA] (µM)

D100(nm)

D010(nm)

D002(nm)

-34±1 48±1 0.375 28.1±0.9 42±2 0.5 29.8o0.8 44±1 1 23.5±0.8 3 1.0±0.4 2 18.3±0.5 25.0±0.4 aReproduced from ref. 33, with permission from Elsevier Science.

72± 1 54 ±2 62 ± 2 34 ± 2 34± 1

agreement, OCP crystals synthesized in presence of NaPA display very irregular habits and significantly reduced dimensions with respect to those obtained in absence of the polyelectrolyte, as shown by the TEM image reported in fig. 5. The crystals do not contain significant amounts of the polyelectrolyte, as indicated by the infrared adsorption spectra, which do not display any bands due to the presence of polyacrylate. Thus, the inhibition of crystal growth cannot be ascribed to polyacrylate adsorption on crystal surfaces and different mechanisms, such as the formation of chelates between carboxylate and calcium ions, for instance, must be suggested. The main effect of the polyelectrolyte on the morphology of the crystal aggregates seen at the scanning electron microscopic investigation is a reduction of the length of the blades which constitute the spherules, as well as the hollow spheres. The comparison of the images in fig. 6 a,b with those in fig. 3 c,d indicates that the thickness of the sphere walls appears reduced when the synthesis is carried out in presence of NaPA. At high polyelectrolyte concentration, 2 µM, SEM images display very large single hollow spheres, whose diameters reach the dimension of 1 mm, and whose walls are constituted of very short blades (fig. 6c,d).

33

Figure 5. TEM image of OCP synthetized in 1µM NaPA solution. Bar = 400 nm

Figure 6. SEM images of OCP synthesized in presence of 1 µM NaPA (a,b) and in presence of 2 µM NaPA (c,d). The OCP blades on the surface of the double spheres appear shortened with respect to those obtained in absence of NaPA (fig. 3c,d). The reduction in length is even more evident for the blades on the surface of the big hollow spherical aggregates obtained in presence of 2 µM NaPA. Bars= (a) 200µm; (b) 20µm; (c) 500µm; (d) 25µm. Reproduced from ref. 33, with permission from Elsevier Science.

34

EFFECT OF SODIUM POLYACRYLATE ON THE HYDROLYSIS OF OCP OCP hydrolysis is inhibited by the presence of NaPA. A polyelectrolyte concentration of l0-2 mM is sufficent to reduce the extent of OCP hydrolysis into HA, which is completely inhibited at higher polymer concentrations. The crystalline phases present in the products obtained after OCP storage in aqueous solutions at different polyelectrolyte concentrations are reported in Table 3, together with the values of final pH of the solutions. HA obtained by OCP hydrolysis in l0-3mM NaPA solution exhibits lattice parameters, coherence length of the crystalline domains, as well as morphology of the needle-like structures seen at the scanning electron microscopy, very close to Table 3. Crystalline phases present in the powder X-ray diffraction patterns of the products of hydrolysis of OCP in aqueous solutions at different NaPA concentrations. The values of final pH are also reported.a 48 hours [NaPA](mM) 24 hours -OCP + HA HA 10-3 OCP – HA HA 10-2 OCP OCP + HA l0-1 OCP OCP 1 OCP OCP 5 OCP OCP aReproduced from ref. 33, with permission from Elsevier Science.

PH 5.6 5.9 6.4 7.0 7.4 7.4

those of the apatitic phase prepared from OCP hydrolysis in water. However, the results of TEM investigation reveal that NaPA affects the shape of the HA crystals which appear quite irregular and significantly shorter (fig. 7) than those obtained by hydrolysis in water (fig. 4a). SEM images of OCP stored in solutions at greater NaPA concentrations (fig. Sa) show spherules, as well as spheres, whose presence suggest that OCP, which was grounded before storage in NaPA solution, can rebuild crystal aggregates morphologically similar to those of the starting material. SEM micrographs of OCP stored in l0-2 mM NaPA, which is partially hydrolyzed to HA, as results from X-ray diffraction analysis, display further SEM images where numerous isolated long blades, as well as very small needle-like structures, are present (fig. 8b).

Figure 7. TEM image of HA obtained by OCP hydrolysis in 10-3mM NaPA solution. Bar= 80 nm.

35

Figure 8. (a) SEM image of OCP after storage in 5mM NaPA solution for 48h. Bar = 30 µm. (b) SEM image of partially hydrolyzed OCP after storage in 10-2mM NaPA solution for 48h. Bar= 10µm.

At the TEM investigation, OCP submitted to storage in NaPA solutions appears as long plate-like crystals (fig. 9a), morphologically similar to, but even longer than those of the starting material.36 TEM images of partially hydrolyzed OCP show also numerous aggregates of much smaller and thinner crystals, as well as isolated small crystals (fig 9b).

Figure 9. (a) TEM image of OCP after storage in 1mM NaPA solution for 48h. Bar= 4 µm. (b) TEM image of partially hydrolyzed OCP after storage in 10-2mM NaPA solution for 48h. Bar= lµm.

After storage in solutions, OCP contains significant amounts of the polyelectrolyte. In fact, IR absorption spectra display not only the bands characteristic of phosphate and monohydrogen phosphate, but also several absorption bands at 2920, 1720, 1650, 1560, 1465 and 1410 cm-1 due to polyacrylate,37,38 as illustrated in fig. 10. The quantitative evaluation of the polyelectrolyte content was performed by means of thermogravimetric analysis, from the weight loss between 250 and 500ºC, due to the polyacrylate combustion.33,37 The relative amount of polyacrylate associated to the inorganic phase increases on increasing NaPA concentration in solution up to about 22 wt%.33 The polyelectrolyte adsorption on OCP induces an anisotropic variation of the coherence lengths of the crystals, as results from the scheme reported in fig. 11, which shows the variation of the ratios among the coherence lengths along the three crystallographic directions on increasing polyacrylate adsorption on OCP. The reduction

36

Figure 10. Infrared absorption spectrum of OCP after storage in 1mM NaPA solution for 48h.

Figure 11. Scheme of the relative coherence lengths of the crystalline domains along the three crystallographic directions of OCP as a function of the polyacrylate content.

37

Figure 12. Electron diffraction patterns of (a) an OCP crystal, obtained normal to the [110] axis; and (b) polycrystalline HA.

of crystal perfection along the direction parallel to the c-axis direction and the simultaneous increase of the crystals mean length along the same direction could be ascribed to crystal aggregation along the c-axis direction due to polymer bridging. The absence of stirring during OCP storage in solution could favour the adsorption of polyacrylate on the hydrated layer of the OCP (100) face, that is the layer most likely exposed to solution, and allow crystal aggregation in spite of the short chain length (6.0 nm)32 of the low molecular weight NaPA. The preferential adsorption of the polyelectrolyte on the (1 00) face would fit well with the similar behaviour exhibited by carboxylate rich proteins, which was attributed to carboxylate complexation of the calcium ions separated by water molecules and exposed on the hydrated layer.28 The long plate-like crystals seen in the TEM images of the products obtained after OCP storage in NaPA solutions exhibit electron diffraction patterns characteristic of OCP single crystals:36 they are generally normal to the [110] axis and indicate that the crystallographic c-axis of OCP coincides with the long axes of the crystals (fig.12a). Similar ED patterns have been recorded from the much smaller crystals seen in the TEM images of partially hydrolyzed OCP (fig. 9b). However, quite often these crystals give ED patterns characteristic of polycrystalline HA, as that reported in fig. 12b. This means that, while the big plate-like crystals can be recognized on the basis of their TEM images as OCP, the habit of the small platelet-like crystals is not sufficient to their identification as OCP or HA. Furthermore, the closeness of their shape and that of HA crystals obtained by OCP hydrolysis in water suggests that OCP transformation into HA proceeds by in situ hydrolysis, rather than by OCP dissolution and successive HA reprecipitation.2,13

CONCLUSIONS NaPA inhibits nucleation and growth of OCP, as well as its hydrolysis into HA. The first of the two phenomena is clearly due to a nonspecific interaction, as indicated by the lack of any significative adsorption of the polyelectrolyte on OCP crystals and by the isotropic reduction of the coherence lengths of the perfect crystalline domains on increasing polymer concentration in solution. On the other hand, the inhibition of OCP

38

hydrolysis into HA takes place through a significant polyelectrolyte adsorption on OCP crystals. We verified that, in similar conditions, the adsorption of NaPA on HA crystals reaches a maximum relative amount of 2 wt%, in agreement with previous studies about NaPA adsorption on HA.32 The polymer preference for OCP structure is confirmed by the comparison among the different crystal habits of HA crystals obtained by hydrolysis at different NaPA concentration.36 The crystals obtained in l0-2 mM NaPA, that is when just a partial hydrolysis occurs, exhibit a morphology quite similar to that of HA crystals obtained by hydrolysis in water, whereas those obtained in 10-3 mM NaPA are clearly affected by the polymer. This finding suggests that when the polyelectrolyte can choose, as in partially hydrolyzed material, it is preferentially adsorbed on OCP rather than on HA crystals. On the basis of the structure of OCP and HA, it must be supposed that the preference of the polyelectrolyte for OCP is due to the presence of the hydrated layer, where it is adsorbed on the (100) face causing an anisotropic variation of the coherence lengths. The morphology of HA crystals obtained by OCP hydrolysis is very similar to that of the plate-like crystals of bone and dentine, and supports the hypothesis of an OCP involvement in the biomineralization process of these tissues.4,6,39 The adsorption on the hydrated layer of the (100) face prevents OCP hydrolysis into HA, in agreement with the hypothesis that hydrolysis takes place along this face, as suggested to explain the contemporary presence of OCP and HA in the same crystal of calcifying dentine.6 Although these results cannot be extrapolated to in vivo interaction between acidic proteins and calcium phosphates, they point out the important role that the presence of acidic macromolecules can play on the relative stability of octacalcium phosphate and hydroxyapatite, as well as on the shape, dimensions and crystallinity of the final crystals.

ACKNOWLEDGEMENTS This research was carried out with the financial support of MURST and the University of Bologna (Funds for Selected Research Topics).

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

J.C. Elliott, Structure and Chemistry of the Apatites and Other Calcium Orthophosphates, Elsevier Sci., The Neitherlands (1994). G. Graham and P.W. Brown, Reactions of octacalcium phosphate to form hydroxyapatite, J. Crystal Growth 165:106 (1996). G.H. Nancollas, Biological Mineralization and Demineralization, Dahlem Konferenzen, Springer-Verlag, Berlin (1 982). R.A. Terpstra and P. Bennema, Crystal morphology of octacalcium phosphate: theory and observation, J. Cryst. Growth 82:416 (1987) M.U. Nylen, E.D. Evans and K.A. Omnel, Crystal growth in rat enamel, J. Cell. Biol. 18: 109 (1963). P. Bodier-Houllé, P. Steuer, JC. Voegel and F.J.C. Cuisinier, First experimental evidence for human dentine crystal formation involving conversion of octacalcium phosphate to hydroxyapatite, Acta Cryst. D54: 1377 (1998). D.G.A. Nelson and J.C. Barry, High resolution electron microscopy of nonstoichiometric apatite crystals, Anat. Rec. 224:265 (1989). W.E. Brown, J.P. Smith, J.R. Lehr and A.W. Frazier,Octacalcium phosphate and hydroxyapatite, Nature 196:1048 (1962). B.B. Tomazic, M.S. Tung, T.M. Gregory and W.E Brown, Mechanism of hydrolysis of octacalcium phosphate, Scanning Microsc. 3: 119 (1989).

39

10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31. 32. 33. 34.

40

J. Zhang and G.H. Nancollas, Kinetics and mechanisms of octacalcium phosphate dissolution at 37ºC, J. Phys. Chem. 96:5478 (1992). M. lijima, H. Tohda, H. Suzuki, T. Yanagisawa and Y. Moriwaki, Effect of F- on apatite octacalcium phosphate intergrowth and crystal morphology in a model system of tooth enamel formation, Calcif. Tissue Int. 50:357 (1992). A. Bigi, M. Gazzano, A. Ripamonti and N. Roveri, Thermal conversion of octacalcium phosphate into hydroxyapatite, J. Inorg. Biochem. 32:25 1 (1988). R.Z. LeGeros, G. Daculsi, I. Orly, T. Abergas and W. Torres, Solution-mediated transformation of octacalcium phosphate (OCP) to apatite, Scanning Electron Microsc. 3: 129 (1989). B.B. Tomazic, I. Mayer and W.E. Brown, Ion incorporation into octacalcium phosphate hydrolyzates, J. Cryst. Growth 108:670 (1991). M. Iijima, H. Kamemizu, N. Wakamatsu, T. Goto, Y. Doi and Y. Moriwaki, Effects of C032- ion on the formation of octacalcium phosphate at pH 7.4 and 37ºC, J. Crystal Growth 135:229 (1994). M. Iijima, K. Iijima, I. Moriwaki and Y. Kuboki, Oriented growth of octacalcium phosphate crystals on type I collagen fibrils under physiological conditions, J. Crystal Growth 140:91 (1994). Y. Moriwaki, Y. Doi, T. Kani, T. Aoba, J. Takahashi, M. Okazaki, Synthesis of enamel-like apatite at physiological temperature and pH using ion-selective membranes, in Mechanisms of Tooth Enamel Formation, S. Suga, ed., Quintessence Publishing Co, Tokyo (1983). M. Iijima, H. Tohda and Y. Moriwaki, Growth and lamellar mixed crystals of octacalcium phosphate and apatite in a model system of enamel formation, J. Crystal Growth 116:3 19 (1992). V.K. Sharma, M. Johnsson, J.D. Sallis and G.H. Nancollas, Influence of citrate and phosphocitrate on the crystallization of octacalcium phosphate, Langmuir 8:676 (1 992). M. Markovic, B.O. Fowler and W.E. Brown, Octacalcium phosphate carboxylates. 5. Incorporation of excess succinate and ammonium ions in the octacalcium phosphate succinate structure, in Hydroxyapatite and Related Materials, P.W. Brown and B. Constantz, eds., CRC Press, Boca Raton (1994). R.Z. LeGeros, R. Kijkovska and J.P. LeGeros, Formation and transformation of octacalcium phosphate, OCP: a preliminary report, Scanning Electron Microscopy 4 :1771 (1984). M.H. Salimi, J.C. Heughebeart and J.H. Nancollas, Crystal growth of calcium phosphates in the presence of magnesium ions, Langmuir 1: 119 (1985). I.Y. Pieters, E.A.P. De Maeyer and R.M.H. Verbeeck, Stoichiometry of K+ and CO32- containing apatites prepared by the hydrolysis of octacalcium phosphate, Inorg. Chem. 35:579 1(1996). R.Z. Le Geros, Variations in the crystalline components of human dental calculus: I. Crystallographic and spectroscopic methods of analysis, J. Dent. Res. 53:45 (1974). D.W. Holcomb and R.A. Young, Thermal decomposition of human tooth enamel, Calcif Tissue Int. 31:189(1980) M.S. Tung and W.E. Brown, The role of octacalcium phosphate in subcutaneous heterotopic calcification, Calcif Tissue Int. 37:329 (1985). H.A. Lowenstam and S. Weiner, On Biomineralization, Oxford University Press, Oxford (1989). L. Addadi, J. Moradian-Oldak, H. Füredi-Milhofer, S. Weiner and A. Veis, Stereochemical aspects of crystal regulation in calcium phosphate- associated mineralized tissues, in Chemistry and biology ofmineralized tissues, H. Slavkin and P. Price, eds., Elsevier Science Publ., The Netherlands (1 992). S.I. Stupp and G.W. Ciegler, Organoapatites: materials for artificial bone. 1. Synthesis and microstructure, J. Biomed. Mater. Res. 26: 169 (1992). E. Bertoni, A. Bigi, G. Cojazzi, M. Gandolfi, S. Panzavolta, N. Roveri, Nanocrystals of magnesium and fluoride substituted hydroxyapatite, J. Inorg. Biochem. 72:29 (1 998). E. Bertoni, A. Bigi, G. Falini, S. Panzavolta and N. Roveri, Hydroxyapatite/polyacrylic acid composite nanocrystals, J. Mater. Chem., 9:779 (1999). D.N. Misra, Adsorption of polyacrylic acids and their sodium salts on hydroxyapatite: effect of their relative molar mass, J Colloid Interface Sci., 18 1 :289 (1996) A. Bigi, E. Boanini, M. Borghi, G. Cojaui, S. Panzavolta, N. Roveri, Synthesis and hydrolysis of octacalcium phosphate: effect of sodium polyacrylate, J. Inorg. Biochem. 75: 145 (1999) M. Mathew, W.E. Brown, L.W. Schroeder and B. Dickens, Crystal structure of octacalcium bis(hydrogenphosphate) tetrakis(phosphate)pentahydrate, Ca8(HPO4)2(PO4)4·5H2O, J. Cryst. Spectrosc. Res. 18:235 (1988).

35. 36. 37. 38. 39.

L. E. Alexander, X-ray diffraction methods in polymer science, Wiley- Interscience, New York (1969). A. Bigi, E. Boanini, G. Falini, S. Panzavolta, N. Roveri, Effect of sodium polyacrylate on the hydrolysis of octacalcium phosphate, submitted for publication Q. Liu, J.R. de Wijn and C.A. van Blitterswjik, Nano-apatite/polymer composites: mechanical and physicochemical characteristics, Biomaterials 18: 1253 (1997). D. Belton and S.I. Stupp, Adsorption of ionizable polymers on ionic surfaces: poly(acrylic acid), Macromolecules 16:1143 (1983). F.J.G. Cuisinier, P. Steuer, A. Brisson and J.C. Voegel, High resolution electron microscopy study of crystal growth mechanisms in chicken bone composites, J. Cyst. Growth 156:443 (1995).

41

This page intentionally left blank.

EFFECT OF ALKYL PHOSPHATES ON THE FORMATION AND CRYSTALLIZATION OF CALCIUM PHOSPHATES IN AN AQUEOUS PHASE

Saburo Shimabayashi,* Keiko Furukawa, and Tomoaki Hino The University of Tokushima, Faculty of Pharmaceutical Sciences Sho-machi 1-78-1, Tokushima, Tokushima 770-8505, Japan

INTRODUCTION Hydroxyapatite(HAP) directly crystallizes and grows in an aqueous phase when the degree of supersaturation is low. On the other hand, it is formed via amorphous calcium phosphate(ACP), octacalcium phosphate pentahydrate (OCPP), or dicalcium phosphate dihydrate(DCPD) when the degree of supersaturation is high. HAP is formed after Ostwald ripening, so called, in its mother solution.1,2 It is known that the transformation from these precursors to HAP and its crystal growth after the induction time are altered by various additives. Taking these facts into consideration, many researchers are devoting themselves until now to the studies for inhibiting the mineral scale formation in industrial areas such as water treatment.3 Understanding the mechanism of formation of sparingly soluble calcium phosphates is also of primary importance in connection with biological mineralization, i.e., growth and maturation of hard tissues (i.e., mammalian bones and teeth), pathological calculi, and so on. In the previous papers4,5 it was shown that phosphorylated serine (PSer) and polyvinylalcohol (Phos. PVA), which are soluble in water, retard the transformation and inhibited the crystallization of ACP to HAP. These effects were explained in terms of adsorption of their ester phosphate groups on the active growth sites of the nuclei/embryos through isomorphous substitution with inorganic phosphate ion(Pi) on the surface of HAP and through electrostatic attractive force toward the surface calcium ion(Ca2+ ). As for the adsorption of Phos. PVA, the significance of the thick polymer layer, which is negatively charged, was emphasized. Some of the phosphate groups along the adsorbed polymer chain participated in the adsorption in contact with the HAP surface as train segments, while others remained on the polymer tails or loops protruding from the surface. Negative charges of the ester phosphate groups in the thick adsorption

Advances in Crystal Growth Inhibition Technologies Edited by Amjad Kluwer Academic/Plenum Publishers New York 2000

43

layer contributed to repelling the Pi approaching to the growth site, resulting in effective inhibition of the crystal growth. Thus, the effect of Phos.PVA was more pronounced than that of PSer which was adsorbed separately on the HAP surface. Insoluble phosphorylated cellulose(Phos. Cell), however, was a promoter of the crystallization from ACP to HAP.4,6,7 This effect was explained as follows. The initial trigger is capturing Ca2+ by its ester phosphate groups, onto which Pi is bound, resulting in formation of a complex of Pi-Ca-Phos.Cell. This complex grows in its extent, and the clusters of ---Pi-Ca-Pi-Ca--- are formed on the cellulose surface. The Phos.Cell acts as the crystal nucleus or embryo for calcium phosphates. In other words, the ester phosphate group of Phos. Cell is an important core for the heterogeneous nucleation of calcium phosphates. With these clusters acting as nuclei, the transformation/formation of HAP from ACP is also accelerated Thus, the induction time from ACP to HAP decreases with the amount of Phos. Cell added. The effect of Phos. Cell is quite a contrast to that of Phos.PVA and PSer. It seems that the insoluble additive behaves as a promoter (i.e., crystal nucleus) while the soluble acts as an inhibitor (i.e., crystal poison) for the crystallization. In the present paper, the effects of monoalky1 phosphate(carbon number: 12, 14, and 16) on the formation and crystallization of calcium phosphates were discussed. Hereafter these monoalkyl phosphates will be described as alkyl phosphates. Although the acid form was almost insoluble in water, it was easily soluble in the presence of arginine, forming the arginine salt, where the mixing ratio of arginine to alkyl phosphate was kept constant at 3 mole/mole. Following the results mentioned above, we expected that the insoluble acid might be a crystallization promoter while the soluble salt an inhibitor. It was found, however, that both acted as promoters. The soluble arginine salt formed an insoluble calcium salt in a mother solution which contained both Pi and Ca2+, resulting in formation of the crystal seed for calcium phosphates/HAP. Thus, some of physicochemical properties of soluble alkyl phosphates are quite different from those of soluble Phos.PVA and PSer in the fact that the alkyl phosphate precipitate in the presence of Ca2+ whereas Phos.PVA and PSer did not, although all of these are a monoester of an alcohol and inorganic phosphoric acid. Primary particles of HAP after ripening were aggregated to their secondary particles, of which size increased with a concentration of the acid form while decreased with that of the arginine salt form. The effect of them on the formation of secondary particles was reversed, depending on whether they are an acid form or salt one, or water-insoluble or soluble. That is, alkyl phosphate is regulating the size of secondary particles of HAP, depending on the species of the alkyl phosphate.

EXPERIMENTAL Materials HAP(Ca10(PO4)6(OH)2) was obtained from Nakarai Chemicals, Ltd.(Kyoto). Its specific surface area was estimated as 56.0 m2/g by N2 gas adsorption, while its molar ratio of Ca/Pi in the bulk was determined as 1.67 after chemical analysis. This sample is the same as that used in the previous study.8,9 Alkyl phosphates (i.e., monoalkyl phosphates, carbon number 12, 14, and 16) are a gift from KAO, Inc.(Tokyo), which were used without further purification. These samples were of white and fine powder.

44

The acid form is almost water-insoluble but soluble in the presence of arginine, as mentioned above. Chemical structures are shown in Figure 1.

Figure 1. Chemical structures of alkyl phosphoric acid, alkyl phosphate ion, and arginine

Methods Calcium ion activity was determined at 35 ºC using an Orion ion-sensitive electrode connected to an Orion expandable ion analyzer (model EA 940). Prior to the measurement on the sample solution, the calcium electrode was calibrated with an aqueous solution of CaC12 in the presence of 0.9% NaC1, taking its activity coefficient into consideration. Because the ionic strength was determined mainly by the added NaCl (0.9% = 154 mM), an activity coefficient for Ca2+ is almost constant irrespective of its concentration. Therefore, the measured activity is easily converted to the concentration of free Ca2+ (shown as [Ca2+] in Figure 2). The suspension/solution pH was measured by means of an Orion pH-electrode also connected to the model EA 940. The values of pH and calcium ion activity were traced for about 70 minutes beginning just prior to adding K2HPO4 (2.50 mM) into an aqueous solution of 0.9% NaCl containing 1.25 mM CaCl2 and a given amount of alkyl phosphate in order to determine the induction time, Ttrans , from ACP to HAP in the presence or absence of an alkyl phosphate. The induction time, Ttrans , was determined from the intersection of the tangents drawn to the time courses of pH and [Ca2+] , as shown in Figure 2. The induction periods, thus obtained from these two curves, were in fair agreement with each other. Surface tension of an aqueous solution was measured by means of a Du Nouy tensiometer (Shimadzu, Kyoto). Zeta potential of the precipitate particles was obtained by a Laser Zee™ (model 501, Pen Kem, Inc.).9 Mean diameter of the secondary particles of HAP was determined by a Coulter counter (type TAII, Coulter Electronics, Inc.). These data were obtained in an aqueous solution of 0.9% NaCl at room temperature( 18 – 24 °C). An aqueous solution of 0.9% NaCl as a solvent and solutions of CaC12 and K2HPO4 were filtered through a membrane filter (Supor 200, pore size 0.2 micrometer, Gelman Sciences Inc., Ann Arbor) immediately before supplying for the experiments.

45

time/min.

[Ca2+]

Figure 2. Time courses of and pH at 35°C. Initial concentrations of CaC12 and K2HPO4 are 1.25 mM and 2.50 mM, respectively, in an aqueous solution of 0.9% NaCl in the absence of an alkyl phosphate, ACP is dominant in the area of (1) while HAP in the area of (3). At the period of (2), ACP is transformed to HAP via OCP. The time at the intersections of the tangents is the induction time, Ttrans , for this system, which will hereafter be referred to as T0 because this is the reference time for estimating the effect of an alkyl phosphate on Ttrans .

RESULTS AND DISCUSSION Effect of Alkyl Phosphate on the Induction Time and Surface Tension The induction time,Ttrans , at 35 ºC was determined after mixing 1.25 mM CaC12 with 2.50 mM K2HPO4 in an aqueous solution of 0.9% NaCl containing a given concentration of an alkyl phosphate. Relationship between the concentration and the induction time relative to that in the absence of alkyl phosphate,Ttrans /T0 , is shown in Figure 3. The Ttrans was reduced with a concentration of an alkyl phosphate irrespective of its form (i.e., acid or arginine salt form, or chain length of the hydrocarbon). Closed symbols in Figure 3 show Ttrans/T0 in the presence of arginine, where a mixing ratio of arginine to alkyl phosphate was kept constant at 3 mole/mole. The Ttrans steeply decreased at ca. 2x10-2 mM for lauryl phosphate (circle) and at ca. 6 xl0-3 mM for myristy1 phosphate (triangle) after retaining a plateau in a low concentration. These results suggest that the transformation and crystallization of ACP to HAP is accelerated at a concentration higher than some characteristic one. In order to identify what the trigger is, surface tension of the solution was measured in the presence and absence of Ca2+, because these compounds are surface active agents. Figure 4 shows surface tension of the solution as a function of a concentration of lauryl phosphate (A) or myristy1 phosphate (B) in the presence and absence of Ca2+. Surface tension continuously and gradually decreased with a concentration of alkyl phosphate and attained a minimum at 10-11 mM for lauryl phosphate (closed circle) and at 2.0-2.1 mM for myristyl phosphate (closed triangle) in the absence of Ca2+.

46

These concentrations are known as the critical micellization concentrations (CMC) for these compounds. On the other hand, in the presence of 1.25 mM Ca2+ (open symbols), surface tension steeply decreased after remaining a plateau and attained an apparent CMC at 12-13 mM for lauryl phosphate (A) and at ca. 5.0 mM for myristyl phosphate (B).

Figure 3. Relationship between Ttrans /T0 and concentration of an alkyl phosphate added. Open symbols are data in the absence of arginine, where an alkyl phosphate is in the acid form. Closed symbols are those in the presence of arginine, where an alkyl phosphate is ionized as the arginine salt form. Open and closed circle: lauryl phosphate, Open and closed triangle: myristyl phosphate, open cross: cetyl phosphate. T0 = 35 min.

Figure 4. Surface tension as a function of concentration of alkyl phosphate at a room temperature. [Arginine]/[alkylphosphate] = 3 mole/mole. [NaC1] = 0.9 % K2HPO4 was not added. Closed symbols: without CaC12, Open symbols: in the presence of 1.25 mM CaC12. Surface tension was determined after filtration of the solution when the precipitate of calcium alkyl phosphate came out. (A) for lauryl phosphate, (B) for myristyl phosphate. Solution pH increased with a concentration of arginine from 6.8(solvent) to 9.0(300 mM arginine).

47

Comparing these data, it is obvious that open symbols deviates upward from closed symbols at ca. 1 x10-2 mM for lauryl phosphate(A) or at ca. 5 xl0-3 mM for myristyl phosphate (B). This deviation or loss of surface activity is owed mainly to the consumption of the surface active ion after its binding to Ca2+ in a mother solution. Binding of Ca2+ and consequent precipitate formation are also responsible for higher CMC in the presence than in the absence of Ca2+. It seems that micelles are formed from the alkyl phosphate ions which are free from Ca2+. The concentrations at which the steep decrease began in Figure 3 (ca. 2x10-2 mM and ca. 6 xl0-3 mM) are in fair agreement with those at which the deviations occurred in Figure 4 (ca. 1 x10-2 mM and ca. 5 xl0-3 mM). On the other hand, the CMC’s, mentioned above, are far different from these concentrations. These results mean that the precipitate of calcium alkyl phosphate filled the role of crystal nucleus/embryo for the calcium phosphates/HAP (i.e., mechanism of heterogeneous nucleation). Returning to Figure 3, open symbols show the data for alkyl phosphoric acids which are almost insoluble in water. The Ttrans steeply decreased at 3 - 6 xI0-3 mM after the stable period at a low concentration. This behavior is quite similar to that for soluble alkyl phosphate, mentioned above (closed symbols). It is interesting that one is water-insoluble while the other soluble, but both of them reduced the Ttrans in a similar manner. The suspension pH of the acid decreased with its concentration. This is owed to the dissociation of H+ from the head groups of alkyl phosphoric acid protruding to the surface of the particles, and to the ion-exchange between H+ on the surface and Ca2+ in the mother solution, which occur concurrently. Thus, calcium phosphate layer is formed on the surface of particles, resulting in formation of the crystal nucleus/embryo for calcium phosphates/HAP. This mechanism is also of heterogeneous nucleation.

Heterogeneous Nucleation by Alkyl Phosphates Figure 5 shows schematic illustrations for heterogeneous nucleation, where (A) is for soluble alkyl phosphates (arginine salt) and (B) for insoluble alkyl phosphates (acid form). Alkyl phosphate ion captures Ca2+, to which Pi also approaches, resulting in formation of the cluster of ---Ca-Pi-Ca-Pi--- . Some of alkyl phosphate groups may aggregate like a micelle or pre-micelle due to hydrophobic interaction between the hydrocarbon chains during and after the cluster formation.

Figure 5. Schematic illustration of heterogeneous nucleation. (A) for soluble alkyl phosphate in the presence of arginine. (B) for insoluble alkyl phosphate in the absence of arginine. Hydrogen ion, sodium ion, chloride ion, and water molecule are not shown here. Shaded circle is a head group of an alkyl phosphate. Dotted line is for coordination of Pi to Ca2+.

48

Figure 6. Time course of mean diameter of the secondary particles in the presence of lauryl phosphate and arginine. [Arginine]/[lauryl phosphate] =3 mole/ mole. Initial concentrations of CaC12 and K2HPO4 are 1.25 and 2.50 mM. [NaC1] = 0.9 %. [lauryl phosphate] /mM = 0(closed circle), 0.005(open triangle), 0.0 10 (open cross), 0.030(open diamond), and 0.050(half closed diamond).

The insoluble particle of the acid captures Ca2+ on the surface after the ionexchange, to which Pi is bound also, resulting in formation of the surface cluster or layer of -Ca-Pi-Ca-Pi-. The calcium ion may or may not penetrate into the inside of the particle, which is not necessarily a primary particle but may sometimes be a secondary one after the mutual aggregation. Because the size of the cluster in (B) is larger than that in (A), the former is more effective on the transformation and crystallization than the latter.

Formation of Secondary Particles in the Presence of Alkyl Phosphate The time course of the mean diameter of secondary particles of calcium phosphates was obtained by means of a Coulter counter at a given concentration of alkyl phosphate in the presence and absence of arginine. The mean diameter rapidly increased within 5 minutes immediately after mixing Ca2+ with Pi, during which primary particles of ACP grew and aggregated to the secondary particles. Figure 6 shows the formation of the secondary particles in the presence of arginine and lauryl phosphate. In the first plateau, where ACP is dominantly prevailing, the particle size increased with the addition of lauryl phosphate. This occurs because the size of the cluster or network structure of ---Ca---Pi---Ca---PI--- (Figure 5(A)) becomes large with a concentration of lauryl phosphate which was mixed with CaC12 prior to the addition of K2HPO4. The particle size steeply increased again 35 – 60 minutes after addition, and attained the second plateau, where the dominant species in the precipitate is HAP because the time of Ttrans (Figure 3) has already elapsed. The transition time from the first plateau (ACP) to the second (HAP) should be closely related to the time of Ttrans. The transition (Figure 6) seems to occur certainly earlier in the presence than in the absence of lauryl phosphate. However, an apparent relationship between the transition time and the Ttrans or a concentration of lauryl phosphate was not obtained. The reproducibility of the transition time was rather subtle and difficult in this work. 49

Figure 7. Time course of mean diameter of the secondary particles in the presence of lauryl phosphoric acid without arginine. Initial concentrations of CaCl2 and K2HPO4 are 1.25 and 2.50 mM. [NaC1] = 0.9%. [lauryl phosphoric acid] /mM = 0 (closed circle), 0.003 (open triangle), 0.007 (open cross), and 0.010 (half closed diamond).

Figure 8. Effect of concentration of alkyl phosphate or arginine on a mean diameter of secondary particles of HAP at 100 minutes after the preparation. Open symbols are showing the data in the absence of arginine, where alkyl phosphates are in the acid form. Closed ones are showing those in the presence of arginine, where alkyl phosphates are ionized as an arginine salt. Mixing ratio of [arginine]/[alkyl phosphate] = 3 mole/mole. Open and closed circle: lauryl phosphate, open and closed triangle: myristyl phosphate, and open cross: cetyl phosphate. The X mark shows the effect of arginine added alone. A concentration of this arginine is given on the abscissa as [arginine]/3 to compare its effect with that of the arginine salt of an alkyl phosphate. Initial concentrations of CaC12 and K2HPO4 are 1.25 and 2.50 mM. [NaC1] = 0.9%.

50

The mean diameter of HAP at the second plateau was larger than that of ACP. This is because the latter was rapidly formed at a high degree of supersaturation while the former was gradually formed on the seed after the dissolution and recrystallization of ACP consuming OH-, Ca2+, and Pi in the mother solution. This is a mechanism of Ostwald ripening.10 On the other hand, the particle size decreased with a concentration of lauryl phosphate added. This fact is also explained in terms of the Ostwald ripening. During the ripening, some of the lauryl phosphate ion were desorbed from the old cluster/ACP and adsorbed again to the surface of HAP newly formed through electrostatic attractive force toward Ca2+ and through isomorphous substitution with Pi on the surface. The hydrocarbon chain of a lauryl phosphate ion after the adsorption is protruding to the solution.' Another lauryl phosphate ion in the mother solution may be bound to the protruding tail through hydrophobic interaction, forming the surface dimer. Arginine distributes around the particle as a counter ion for the adsorbed alkyl phosphate anion and Pi on the surface. The HAP particles are, thus, charged and the degree of dispersion increases with the amount of adsorption and/or concentration of lauryl phosphate. The effect of lauryl phosphate on the size at 100 minutes is shown in Figure 8. Figure 7 shows time course of the mean diameter in the presence of waterinsoluble lauryl phosphoric acid. The particle size increased to the second plateau (HAP phase) after remaining the first plateau (ACP phase) for a certain period. This tendency is quite similarly to that observed in Figure 6. However, the particle size in both the plateaus increased with a concentration of lauryl phosphoric acid added, because the size of the nucleus/seed (Figure 5(B)) increases with the concentration. The effect of lauryl phosphoric acid on the size at 100 minutes is shown also in Figure 8 with a open circle. Vertical distance between closed circle (concentration of lauryl phosphoric acid = 0 mM) and open cross (0.007 mM) or half closed diamond (0.010 mM) in the second plateau is larger than that in the first plateau. This fact is showing that the HAP particles are more aggregated in size than the ACP particles after the Ostwald ripening. Lauryl phosphoric acid is contributing to the aggregation in contrast with lauryl phosphate ion (Figure 6). The size at low concentration of lauryl phosphoric acid (open triangle, 0.003 mM) was almost the same as that in the absence of it (closed circle). This is because the amount of lauryl phosphoric acid was too small to affect the size or was dissolved and disappeared from the solution. The effect after the dissolution of 0.003 mM lauryl phosphoric acid is rather similar to that of 0.005 mM lauryl phosphate ion, shown in Figure 6. Figure 8 summarizes the effects of alkyl phosphates and arginine on a mean diameter of the particles at 100 minutes after the mixing of Ca2+ with Pi, where HAP is the prevailing species. The mean diameter increased with a concentration of the acid powder, where the effect of lauryl phosphoric acid was the most remarkable. The reason why the effect was weak and/or the size levelled off after a small increase in the cases of myristyl and cetyl phosphoric acids is not clear, although it might be closely related to the solubility, surface and crystal structure of the seed particles. However, the most important factor may be the size of the seed. The mean diameter of the seed was 6.5, 3.2, and 2.5 micrometers for lauryl, myristyl, and cetyl phosphoric acids, respectively. With increase in the size, the degree of the aggregation may have been increased. On the other hand, the mean diameter decreased with a concentration of the arginine salt form, where the effect of lauryl phosphate was slightly weaker than that of myristyl phosphate. Unfortunately, cetyl phosphate was not studied here because it was not solubilized even in the presence of arginine owing mainly to a high Kraft point.

51

Figure 9. Zeta potential of HAP particles as a function of concentration of alkyl phosphate added. The HAP was ripened for 120 minutes at 35 °C after mixing 1.25 mM CaC12 with 2.50 mM K2HPO4 in an aqueous solution of 0.9% NaCl in the presence or absence of arginine. Closed symbols show the data in the presence of arginine, where [arginine]/[alkyl phosphate]=3 mole/ mole. Open symbols are those experiments perfomed in the absence of arginine, where an alkyl phosphate is in an acid form. The solution pH was 6.96 in the absence of both alkyl phosphate and arginine. It increased to 7.04 with a concentration of the arginine salt, and decreased to 6.80 with that of the acid. Open and closed circle: lauryl phosphate. Open and closed triangle: myristyl phosphate.

Arginine alone did not exhibit any effects on the mean diameter (x in Figure 8) or on the Ttrans (data not shown) up to 0.3 mM. It could, therefore, be concluded that the anionic form of the arginine salt of alkyl phosphate is mainly responsible for the size decreasing, Arginine was just a solubilizing agent for the alkyl phosphoric acids, and it behaved as a counter ion for the alkyl phosphate anion. A probable solubilizing mechanism is proposed as follows. When arginine, a strong base, attacks the acid in crystal, the crystal structure gradually collapses during the neutralization and ionization due to the bulkiness of the arginine as a counter ion, resulting in dissolution of the acid.

Zeta Potential of the HAP Particles Zeta potential of the HAP particles9 was measured after ripening for 120 minutes at 35 ºC. The results are shown in Figure 9. The pH of a mother solution slightly changed with added alkyl phosphate (0 - 0.1 mM) from a reference value of 6.96 in the absence of both alkyl phosphoric acid and arginine. It decreased to 6.80 alkyl phosphoric acid added in the absence of arginine, while it increased up to 7.04 with a concentration of alkyl phosphate added in the presence of arginine, where the mixing ratio of [arginine]/[alkyl phosphate] was 3 mole/mole. The decrease is due to the dissociation of H+ from an alkyl phosphoric acid, while the increase due to the excess amount of arginine. As the Pi was given more than Ca2+ (the mixing ratio of Pi/Ca2+ = 2 mole/ mole, as mentioned in above in the experimental section), the excess Pi was adsorbed on the surface of the HAP particles, and, therefore, zeta potential of the HAP was negative in the absence of both an alkyl phosphate and arginine, as shown on the ordinate. However, it gradually increased with a concentration of the acid added. This tendency

52

was the same in both lauryl phosphoric acid (open circle) and myristyl phosphoric acid (open triangle). The increase in zeta potential, i.e., the decrease in negative charge of the particle, was caused by the adsorption of H+ , which was dissociated from the seed acid, onto the HAP particles. On the other hand, zeta potential steeply increased and reversed to a positive value with a concentration of an alkyl phosphate in the presence of arginine. This phenomenon was observed both with lauryl phosphate (closed circle) and with myristyl phosphate (closed triangle). During the period of transformation and crystallization from ACP to HAP, the calcium ion from the calcium alkyl phosphate, which was formed prior to ACP/HAP, was also used in some degree to form the HAP, resulting in release of an alkyl phosphate ion into a mother solution. This ion is adsorbed again to the surface of the newly formed HAP particles. Some of the alkyl phosphate ions may form dimers on the surface by virtue of hydrophobic interaction between the tails and capture arginine as a counter ion. Arginine may play a role of counter ion also for an inorganic phosphate ion (Pi) exposed on the surface of the particle. The surface charge or zeta potential, thus, increases with a concentration of alkyl phosphate in the presence of arginine, although the pH (or concentration of OH-) in the mother solution increases.

Model for Crystal Growth and Aggregation of the HAP Particles Summarizing the facts mentioned above, the schematic illustrations for the crystal growth and aggregation are given in Figure 10. In the presence of arginine (A), calcium alkyl phosphate behaves as a crystal seed for the HAP, which is remained in part in the core of the particles. However, large portion of the alkyl phosphate ions, which are released from the seed, are re-adsorbed on the surface of the particles by electrostatic attractive force toward the surface Ca2+ and by isomorphous substitution with the surface Pi. Some of these adsorbed alkyl phosphate ions probably form the dimer on the surface. Arginine is bound also to the surface as a counter ion for the dimer and surface Pi. Therefore, the surface is hydrophilic but contaminated by the

(B) in the absence of arginine

Primary particle

Secondary particle

Figure 10. Schematic illustration for the crystal growth and aggregation. (A) in the presence of arginine, and (B) in the absence of arginine. The left-hand side is for the growth of the primary particles, and the right-hand side for that of secondary particles. See the text.

53

accumulated adsorbates, resulting in the fact that the HAP primary particles are rather inhibited from coagulating/aggregating. Needless to say, the electric charges also contribute toward the dispersion. On the other hand, in the absence of arginine, HAP crystals are formed on the surface of the seed, the core of which is an alkyl phosphoric acid, as shown in Figures 5(B) and 10(B). Because this acid is sparingly soluble in water, the adsorption amount of alkyl phosphate ion or alkyl phosphoric acid on the surface of newly formed HAP particles is also low. Hydrogen ion dissociated from an alkyl phosphoric acid is bound to the Pi localized on the surface, as mentioned before. These effects are, however, assumed to be small in magnitude according to the fact that the change in zeta potential with a concentration of an alkyl phosphoric acid is small, compared to that in the presence of arginine (Figure 9). In this way, the particle retains a negative charge as high as that for the bare particles, shown on the ordinate in Figure 9. This result is in contrast with that in the presence of both an alkyl phosphate ion and arginine. Because the adsorption amount of alkyl phosphate ion is low, the tail is protruding into the solution but not forming the dimer on the surface. A large portion of the surface of the particle is naked with respect to the adsorbate species. After the mutual collision, these particles are easily aggregated to form secondary particles by matching the arrangement of the ions on the surface contacting to that of the neighboring particles with one another with the assistance of interparticle van der Waals force and interparticle hydrophobic interaction between the protruding tails in spite of high zeta potential. Opposite charges on the neighboring particles are partly arrayed in such a manner as (+)---(-) and (-)---(+) along the interface, resulting in operation of the electrostatic attractive force between the neighboring particles. The van der Waals attractive force is rather strong because the seed is large in size. As for the particles formed in the presence of arginine ((A) in Figure l0), the aggregation was rather inhibited despite low zeta potential, as mentioned above, because of small particles (i.e., interparticle van der Waals force is weak) and modification of the surface of the particles (i.e., much of adsorbates is interspaced between the particles). It is quite interesting that arginine alone did not affect both the Ttrans and the mean diameter, although the particle is negatively charged whereas arginine positively charged, where the pKa values of the acid dissociation constants for arginine are 2.17 for alpha-COOH, 9.04 for alpha-NH3+, and 12.48 for guanidium NH2+. The affinity of arginine for the bare particles of calcium phosphates seems weak. On the other hand, arginine solubilized an alkyl phosphoric acid in an aqueous solution of 0.9% NaCl and it behaved as a counter ion for an alkyl phosphate in an aqueous phase as well as on the surface. Alkyl phosphate regulated the transformation and dispersion/aggregation of the particles in the presence or absence of arginine. The adsorption of alkyl phosphate together with arginine played an important role in the regulation. It is known that treatment of HAP particles with alkyl phosphates in acetone-water (2 : 1) modifies the surface properties after the adsorption.11,12 It might be important and interesting to compare the physicochemical properties of the samples treated in an organic solvent with those prepared in an aqueous phase in the presence or absence of arginine. This is now under investigation.

CONCLUSIONS The following conclusions may be drawn from the results presented here. 1. HAP was formed via ACP in the presence of alkyl phosphate of either form, the acid form (barely soluble in water) and the arginine salt form (easily soluble in water). 54

2. Transformation of ACP to HAP was accelerated with a concentration of alkyl phosphate irrespective of its form. The acid form was more effective in the transformation and crystallization than the arginine salt form. 3. Particles of the insoluble acid adsorbed Ca2+ after the ion-exchange with H+, while alkyl phosphate ion of the soluble arginine salt captured Ca2+ to form the precipitate. Both resulted in formation of the ---Ca--Pi--Ca--Pi--- clusters or embryos (i.e., heterogeneous nucleation). 4. Mean diameter of the secondary particles of HAP increased with a concentration of the acid (i.e., aggregation), whereas it decreased with that of the arginine salt(i.e., dispersion). Size of the secondary particle was, thus, regulated. 5. Negative charge or zeta potential of the particles was higher in the presence of the acid form than in the presence of the arginine salt form. The mechanism for aggregation/dispersion was discussed, taking the particle size and the adsorption amounts into consideration. 6. Arginine alone did not show any remarkable effects on the formation and dispersion of the particles up to 0.3 mM. It filled the role of solubilizing agent and/or a counter ion for an alkyl phosphoric acid.

REFERENCES 1. G.H. Nancollas and A. Zieba, Constant composition kinetics studies of the simultaneous crystal growth of some alkaline earth carbonate and phosphates, in: Mineral Scale Formation and Inhibition, Z. Amjad, ed., Plenum Press, New York(1995). 2. J.C.Elliott, Structure and Chemistry of the Apatite and Other Calcium Orthophosphates, Elsevier, Amsterdam( 1994). 3. Z.Amjad,Mineral Scale Formation and Inhibition, Plenum Press, New York (1995). 4. S. Shimabayashi, N. Hashimoto, H. Kawamura, and T. Uno, Formation of hydroxyapatite in the presence of phosphorylated and sulfated polymers in an aqueous phase, in: Mineral Scale Formation and Inhibition, Z. Amjad,ed., Plenum Press, New York( 1995). 5 . S. Shimabayashi, Y .Tanizawa, Formation of hydroxyapatite in the presence of phosphorylated polyvinylalcohol as a simplified model compound for mineralization regulator phosphoproteins, Chem. Pharm. Bull., 38 : 1810 (1990). 6. S. Shimabayashi, N. Hashimoto, and T. Uno, Effect of phosphorylated cellulose and bovine serum albumin on crystallization of hydroxyapatite, Phosphorus Research Bulletin, 3: 7(1993). 7. S. Shimabayashi and T. Uno, Crystal growth of calcium phosphates in the presence of polymeric inhibitors, in: Calcium Phosphates in Biological and Industrial Systems, Z.Amjad, ed., Kluwer Academic, Boston(1998). 8. S. Shimabayashi, S. Nishine, T. Uno, and T. Hino, Adsorption of hydroxypropylcellulose on hydroxyapatite via formation of surface complex with sodium dodecylsulfate, in: Water Soluble Polymers, Solution Properties and Applications, Z.Amjad, ed., Plenum Press, New York (1998). 9. S. Shimabayashi, K. Kishimoto, and T. Hino, Competitive and cooperative adsorption of bovine serum albumin and sodium chondroitin-6-sulfate on hydroxyapatite in an aqueous phase, Phosphorus Research Bulletin, 9: 23 (1999). I O . W. E. Brown, M. Mathew, and M. S. Tung, Crystal chemistry of octacalcium phosphate, in: Inorganic Biological Crystal Growth, Part 2, B.R. Pamplin, ed., Pergamon Press, Oxford(l981). 11. H. Tanaka, A.Yasukawa, K.Kandori, and T.Ishikawa, Surface modification of calcium hydroxyapatite with hexyl and decyl phosphate, Colloids Surfaces (A), 125: 53 (1997). 12. H. Tanaka, A.Yasukawa, K.Kandori, and T.Ishikawa, Modification of calcium hydroxyapatite using alkyl phosphates, Langmuir, 13: 821 (1997).

55

This page intentionally left blank.

CALCIUM PYROPHOSPHATE CRYSTAL SALT FORMS AND THE INFLUENCE OF PHOSPHOCITRATE

John D. Sallis,a Andrzej Wierzbicki,b and Herman S. Cheungc a

Dept. Biochemistry, University of Tasmania. Hobart, Australia 7001 Dept. Chemistry, University SouthAlabama. Mobile, AL 36688 c Dept. Medicine, University Miami. VA Medical Center. Miami, FL 33135 b

INTRODUCTION TO CALCIUM PYROPHOSPHATES Controlling the development and crystallization of calcium pyrophosphates offers an interesting research challenge as many diverse morphological forms can be generated both in nature and synthetically. Brown and colleagues1 in pioneer work described the preparation and characterization of a variety of mixed salts of calcium pyrophosphate types for the ultimate purpose of identifying calcium pyrophosphates in fertilizers. Calcium ammonium pyrophosphates are of value in fertilizers to provide a leaching source of nitrogen, phosphorus and calcium elements for plant growth. Industrially, calcium pyrophosphates also have been incorporated into many other useful products such as abrasives, ceramics and dentifrices. An important discovery from a biological standpoint was the recognition early in 1962 by McCarty et al2 that calcium pyrophosphate crystals in addition to basic calcium phosphates are present in specific human tissues and fluid during a state of “pseudogout”. Such crystals can cause acute attacks of inflammatory arthritis but more generally, the crystals are associated in the osteoarthritic joint producing a chronic, non-inflammatory disease. Interrelationships of crystals and osteoarthritis have recently been reviewed.3 Crystals present in articular connective tissue, particularly hyaline cartilage, amplify severe cartilage degeneration. Synovial fluid and synovium also participate in forming new crystals and combating shed crystals, adding to the complex cycle of events associated with the joint environment. With an increasing geriatric population and greater recognized prevalence of arthropathies, control over the initiation and crystallization of calcium pyrophosphates and also calcium phosphates, is creating a much greater research focus to seek an understanding of the underlying events taking place. A major unanswered question is just how the crystals form in vivo. Of all the types of calcium pyrophosphates which have been isolated or synthesized so far, it is only the calcium pyrophosphate dihydrate, monoclinic [CPPD(M)] and calcium pyrophosphate dihydrate, triclinic

Advanced s in Crystal Growth Inhibition Technologies Edited by Amjad Kluwer Academic/Plenum Publishers New York 2000

57

[CPPD(T)] forms that are ever recognized in tissue.4 Conceptually, it might be thought that provided only the two ionic species were present together in water in appropriate concentrations, then the respective dimorphs would eventually result. Unfortunately, this does not occur implying that a much more complex milieu is needed. Pyrophosphate is an ubiquitous anion being the product of numerous biological reactions, but concentrations in tissues are very variable and often short-lived due to the activity of pyrophosphatases hydrolysing the molecule. Cartilage itself spontaneously elaborates extracellular pyrophosphate. Calcium ion concentrations on the other hand tend to be maintained at more constant levels in fluids and cells but again, if transient intracellular fluctuations do occur, they can trigger many different responses. It is, for example, well known that many agonists upon binding to cell membrane receptors are capable of causing transduction of a signal to stimulate release of calcium from stores or the opening of calcium channels to raise intracellular calcium levels.5 These observations collectively, suggest that it would be difficult to sustain a correct ratio of ions for product formation and given that most biologic media will contain sodium, potassium , magnesium, chloride and other ions, it is surprising that the pure CPPD dihydrates form rather than a mixed salt. It needs also to be recognized that at best, calcium pyrophosphates are only sparingly soluble but phase changes can occur in an aqueous environment.6 Despite the apparent formidable barriers imposed, the fact that there is a selected specificity of tissue involved perhaps offers a key to biomediated crystal formation. Matrix vesicles with possible enzymes and other proteins serving as a nidus, are recognized to be capable of accumulating calcium pyrophosphate crystals.7 Numerous investigators have reported on the chemistry, preparation and characterization of a variety of calcium pyrophosphate crystalites. Aside from the studies of Brown and colleagues1 who prepared and characterized 25 calcium pyrophosphates which included the ammonium and potassium mixed salts together with pure hydrated calcium pyrophosphates, extensive chemical characterization studies also have been pursued in Pritzker’s laboratory on dihydrate, tetrahydrate calcium pyrophosphates and the sodium and magnesium mixed salts.8 The growth specifically of CPPD(T) under synthetic aqueous conditions also has been reported.9,10,11 Not withstanding all of these attempts, the initiation and growth of the recognized natural in vivo dimorphs is difficult to establish under physiological neutral pH conditions. The predominant CPPD(T) type crystal normally requires acid conditions (
58

PHOSPHOCITRATE CHARACTERISTICS AND MOLECULAR MODELING STUDIES Phosphocitrate (PC; Figure 1) is a powerful inhibitor of the nucleation, crystallization and aggregation of a number of calcium and magnesium salts. Initially shown to inhibit strongly the development of calcium phosphates such as hydroxyapatite,15,16,17 it has since been demonstrated to exert equally powerful inhibitory activity toward calcium oxalates18,19 both COD (calcium oxalate dihydrate) and COM (calcium oxalate monohydrate), calcium sulfate and calcium carbonate,20 and more recently calcium pyrophosphate formation.21 In addition, the crystallization of struvite, a mixed salt of magnesium ammonium phosphate present in a urinary stone form is also prevented.22,23

Figure 1. Phosphocitrate structure,

Contact with calcium salt crystals (including CPPD(T)) can cause lysis of membranes.10 Aside from crystal growth inhibition then, PC also has the ability to protect membranes from lysing as a result of crystal contact.24 In respect to basic calcium phosphate and CPPD, in vitro cell culture studies reveal that in addition, there is stimulated intracellular mitogenesis and the synthesis and secretion of proteases and prostaglandin,25 in fact, the expression of many of the commonly observed responses observed in a clinical arthritic condition.26 Of relevance then, PC inhibits these responses when present in incubation media bathing cells in culture. In addition, in vitro studies have demonstrated that both ATP and non-ATP dependent calcium pyrophosphate deposition (type unknown) occurring in matrix vesicles and articular cartilage also is blocked by PC21 Insights into the mechanism of action of PC suggest that it’s inhibitory activity is in large part due to the ionization at neutral to slightly alkaline pH, of hydroxyl and carboxylic groups associated with the molecule. Dissociation values at a pH for PC of less than 8 are pK1 <2.0 (POH); pK2 3.67 (β-COOH); pK3 5.15 (α-COOH) ; pK 4 7.69 (P-OH). Ionic charge is not the only determinant though as the stereochemical relationship between groups and the molecular framework as a whole plays a significant role.27 Thus PC as a water soluble, highly negatively charged small molecule has excellent affinity for surface binding to calcium crystallites.28,29 An additional mode of action (yet to be proven) is that PC may exert an influence by binding to membrane components, thus limiting crystal contact and any subsequent signalling of a mitogenactivated intracellular protein kinase cascade pathway. From previous studies employing molecular modeling techniques, the nature of a PC molecule interacting with crystals of calcium oxalate19 and magnesium ammonium phosphate23 has been revealed. The molecular model of PC used was generated using parameters of computational geometry optimization and charge distributions (Figure 2a). The modeled

59

structure was found to be virtually identical with X-ray analysis of synthetic crystalline PC (Figure 2b). In respect to calcium oxalate, PC binds to the (-101) planes of a COM crystal as these loci have calcium ions in perfect position to interact with PC , stabilising the faces and slowing growth in a perpendicular direction. Distortion of crystal morphology was readily verified in COM growth experiments.19 For the more complex struvite crystal where three ions, ammonium, magnesium and phosphate plus water molecules, are present in the unit cell lattice, again the morphology of growth in the presence of PC could be matched with computer models, indicating both surface orientation of PC and penetration to displace ions within the lattice.23

Figure 2. The molecular PC ion structure and crystalline monosodium PC.

Of interest then, was our recent study exploring the theoretical molecular interaction of PC with a CPPD(T).30 In consideration of the influence that PC might have to block CPPD(T) formation, it was necessary also to model this specific crystal. Mandell31,32 previously had described a structural representation from X ray data of CPPD crystals produced either in vitro or sourced in vivo. With this information then, a computer model of the CPPD unit cell was generated as shown in Figure 3.

Figure 3. Computer model of the unit cell lattice for CPPD . Expressed faces are (0l0), (100) and (001) with number of formula cell units =2.

60

Computer modeling of the morphology of a CPPD(T) crystal reveals two possible crystal habits (Figure 4). A crystal morphology with (l00), (010) and (0-11) faces expressed is shown in Figure 4a, a morphology previously observed.32 An alternate morphology for CPPD involves expression of (l00), (010) and (001) faces and the model of this form is indicated in Figure 4b. It also has been noted experimentally.33

Figure 4. Computed morphologies of a CPPD(T) crystal.

When PC interacts with the CPPD crystal, there are a number of possibilities. Binding is strongest at the (010) surface with the negatively charged groups favorably placed to interact with the calcium ions on the surface and also at an (011) face. In the latter position, there is a double row of calcium ions sandwiched between rows of pyrophosphate. The best binding energies then of PC to CPPD (-2848 Kcal/mol and –2961 kcal/mol) are calculated to be with PC binding to the (010) and (011) calcium rich faces respectively. Molecular modeling of the binding of PC to (010) and (011) faces of CPPD is visualized in Figures 5 & 6. Strong binding of PC to these planes, mainly due to the presence on homoplanes of calcium ions parallel to these planes, leads to a very strong and stereospecific interaction between the PC and CPPD crystals. This PC interaction is going to be reflected in the CPPD crystal morphology.

Figure 5. Hypothetical binding of a PC molecule to the (010) face of a CPPD.

61

Figure 6. Hypotneticai binding of a PC molecule to the (011) face of a CPPD.

Binding of PC to (010) for example, should result in crystal morphology similar to that shown in Figure 4a, except that the (010) face will become enlarged, relative to other faces, and the crystal as the dose of inhibitor increases will grow much slower in the direction perpendicular to (010) faces. This in turn will result in thin, plate-like crystals with its shortest dimension aligned in the direction perpendicular to the (010) face. In summary, strong evidence exists that PC has the potential to not only prevent the initiation of crystal formation but in addition, the compound has the capacity to nullify continuing growth of preformed crystals which might be present. The postulated mechanism of its action which has emerged from molecular modeling data is of a strong binding affinity that PC possesses for selected crystal faces. The corollary then is that this molecule has the essential elements to control many deleterious events that follow unwanted biomineralization in humans and other species.

GENERATION OF SOME CALCIUM PYROPHOSPHATES AND THE INFLUENCE OF PHOSPHOCITRATE Given all of the above information, an attempt therefore was made to experimentally produce and study the morphology of some CPPD crystals grown in the presence of PC.

Figure 7. An electron micrograph of calcium dihydrogen pyrophosphate crystals (bar = 50µm).

62

CPPD(T) crystals were first generated at acid pH to provide ‘reference’ material. The precursor, CaH2P2O7, was prepared as previously described1 by heating orthophosphoric acid to 21 0°C and slowly adding anhydrous Ca(H2PO4)2 until a heavy precipitate was present. The crystals were filtered hot and air dried (Figure 7). Chemical analysis confirming product was made following acid hydrolysis. Ca++ was estimated by capillary electrophoresis;34 PO4 by spectroscopy of the acid-molybdate chromophore.35 Thereafter, CPPD(T) was generated by a scaled-down version of that described by Burt and Jackson.9 The addition of CaH2P2O7 to an acidic solution (
Figure 8. An electron micrograph of synthetic CPPD(T) crystals (bar = 250µm).

Developing pure CPPD(T) crystals at neutral or slightly more alkaline pH imposes problems. For technical reasons outlined earlier, the trial of a silica-gel system did not prove useful for examining PC influence. The alternate approach was to use pure solution chemistry at the pH sought but as this necessitates introducing alkali to neutralize the phosphoric acid, it leads to the inevitable generation of mixed salt crystals. Nevertheless, it was considered of value to determine what crystal phase changes might occur during growth in the presence of PC. Growth medium at pH 7.5 was prepared with final concentrations of sodium pyrophosphate (80mM) and calcium chloride (2mM). PC (<1–6 mM) was added as the neutral sodium salt and

63

vials were incubated at 30°C with gentle agitation for up to 3 days or more. Crystal formation was monitored daily by light microscopy and analysis for chemical composition verified crystals of CaNa2P2O7.4H2O.

Figure 9. Phosphocitrate influence on the crystal habit of the sodium salt of calcium pyrophosphate. Crystals grown in the absence of PC are in the left panel (Figure 9a; bar = 200µm) while those grown in the presence of 6mM PC are in the right panel (Figure 9b; bar = 150µm).

Large crystals with parallel sides and sharp, clear features were formed within 24h in the controls (Figure 9a). With continued incubation, these crystals became heavily overlayered. No crystals were evident in a 24h period when PC was present but were slowly forming around 60h. Incubation conditions were changed to allow the start of normal crystal formation in a period of 6h before adding PC, thus presenting PC to growing faces rather than preventing nucleation. Limiting PC concentrations (2–6mM) were used to retard, rather than totally inhibit growth so that any face changes occurring would be identified. With increasing PC concentration (up to 10mM), there was a correspondingly lesser number and smaller-sized, thinner crystals, reflecting in part perhaps the shorter time period available for any growth. The morphology of the crystal indicated that some faces were enlarged with respect to the other faces (Figure 9b). In some instances there was a serrated appearance at edges suggesting growth was taking place at defect sites on the crystal. Total elimination of crystal appearance for a seven day period occurred if concentration of PC was increased to 10mM. Simulated modeling of these changes has not yet been possible as to date, x-ray co-ordinates are not available. Another mixed salt was crystallized and examined when lithium ions were substituted for sodium ions, the rationale being that this smaller ion species might more closely represent the pure CPPD(T) type. Pyrophosphoric acid was carefully neutralized with LiOH to reach a pH of 7.5. Calcium chloride was mixed with lithium pyrophosphate to give final concentrations of 2mM and 80mM respectively. PC (as the free acid as opposed to the sodium salt) was included in a range of concentrations (2-10mM). The addition of this form of PC did not alter the final PH. Growth of these crystals was much slower, occurring over 1-2 weeks and their morphology was different from any of the other crystals examined. The control crystals were smaller (as compared to a sodium salt type) and in the earliest stage of growth appeared as thin hexagons with one longer axis. As growth proceeded, there was considerable thickening, giving the typical appearance displayed in Figure l0a. When grown in the presence of 4mM PC, it was evident that strong binding of PC had occurred (Figure l0b). 64

Figure 10. Changed morphology of a lithium salt of calcium pyrophosphate in the presence of phosphocitrate. An electron micrograph of crystals grown in the absence of PC is in the left panel (Figure l0a; bar = 100µm); a single crystal grown in the presence of PC in the right panel (Figure l0b; bar = 40µm).

Again total inhibition of crystal formation was possible when higher concentrations of PC were present. These data indicate that even when an atom as small as lithium is present in the lattice, PC still has the ability to bind and disrupt the formation of the normal unit lattice, thus negating crystallization. Lithium concentration present in the crystal structure was estimated by capillary electrophoresis ion analysis34 and ionic composition of the crystals suggested CaLi2P2O7.2H2O. The unavailability of single crystal x-ray data for this compound curtails any model building but our experimental data suggest that strong interaction with PC has occurred to limit growth at several faces of the crystal. Development of two other calcium pyrophosphate crystal types was also examined. In respect to potassium salts which are known to exist,1 potassium hydroxide was used to neutralize pyrophosphoric acid before adding calcium to the medium. However, no crystals were observed to develop under the growth conditions and time limits imposed.

Figure 11. Electron micrographs of calcium pyrophosphate crystalline forms grown in the presence of magnesium and influenced by PC. The image in the left panel (Figure 1 la; bar =l00µm ) represents crystals forming in the absence of PC while the image in the right panel (Figure 11b; bar = 35µm) is of crystals grown in the presence of PC.

65

Instead, crystallization of a calcium pyrophosphate salt in the presence of magnesium was examined as detailed studies on the growth of calcium magnesium pyrophosphates have been described by P-Cheng and Pritzker.36 They report however, that the type of crystal polymorph that develops is strongly influenced by prevailing concentrations of the ionic species in solution together with pH, time of growth and other factors. This certainly proved to be the case under the standard comparative conditions that were employed in the present study. The inclusion of magnesium in a sodium pyrophosphate medium did result in a crystal form being produced, but the overall result was unexpected. Only one type was observed when concentrations of calcium and magnesium were equivalent. Control crystals (Figure 1 la) were very fine needle-like forms in sphere or feather formations whereas in the presence of 5mM PC (Figure 1lb), a changed morphology with blunted ends was observed. This particular CPPD salt form was initially thought perhaps to be of a calcium magnesium pyrophosphate crystal phase designated CMPP2 by Cheng and Pritzker.36 However, an elemental analysis using energy dispersive spectroscopy (Figure 12) revealed a complete absence of the magnesium element indicating exclusion of this atom during formation. The composition of this crystallite is currently being reviewed.

Figure 12. EDS spectroscopy profile of calcium pyrophosphate crystals grown in the presence of magnesium.

General Comments While it is evident that PC can bind to calcium pyrophosphate crystals irrespective of any additional ions present in their structure, ongoing investigations are still required to assert that in the specific cellular environment of hyaline cartilage, the compound will relieve or eliminate CPPD crystal-induced arthropathy. Certainly, there are no other proven compounds which have the capability of preventing initiation, crystallization and growth of calcium pyrophosphates at a neutral pH. Magnesium treatment for example, previously has been trialed with a view to making crystals dissolve more readily37 but secondary responses can lead to a worsening of the condition. More recently it has been suggested that alkaline phosphatase activity38 (serving in the role of pyrophosphatase) and working synergistically with polyamines to assist crystal dissolution could be a biological control on the extent of CPPD build up. It is known that responses are directed specifically to the end faces (010) of CPPD crystals.

66

In respect then to PC, there are no negative aspects associated with PC to raise doubts that it couldn’t be developed into a useful therapeutic agent. Our biological assessments from using this compound would indicate that small animals show no undesirable secondary responses39 nor is it considered toxic as it has been identified as a natural compound in mammalian mitochondria.40,41 With calcium pyrophosphate generated in the extracellular matrix of articular cartilage and shed into the surrounding aqueous environment it would be important for management that PC should reach its target. Our in vitro and in vivo evidence to date supports the contention that PC can penetrate through cartilage.21 PC also of course is more than capable of preventing the damage that any shed crystals can promote. This is because of its known ability to protect against membranolysis, stop crystal-induced mitogenesis and prevent secretion of destructive enzymes such as collagenase, stromolysin and proteases. The potential for PC to be a useful future therapeutic agent in crystal associated arthropathies is encouraging from all of the evidence accrued to date, but testing of PC in clinical rheumatoid patients is still some distance away. Although two PC-prodrugs have been investigated which target the kidney and urinary tract,42 a general oral delivery presentation of the compound would be a valuable future goal.

ACKNOWLEDGEMENTS The ESEM services of Dr. David Steele, University of Tasmania are gratefully acknowledged.

REFERENCES 1. E.H. Brown, J.R. Lehr, J.P. Smith, and A.W. Frazier. Preparation and characterization of some calcium pyrophosphates. J Agricultural and Food Chemistry. 11 :214 (1963). 2. D.J. McCarty, N.N. Kohn, and J.S. Faires. The significance of calcium pyrophosphate crystals in the synovial fluid of arthritis patients; the “pseudogout” syndrome. Ann Intern Med. 56: 711 (1962). 3. L.M. Ryan and H.S. Cheung. The role of crystals in osteoarthritis, in: Osteoarthritis, Rheumatic Diseases of America. K.D.Brandt, ed; WB Saunders Co. Philadelphia. 25:257 (1999). 4. K.P.H. Pritzker. Calcium pyrophosphate crystal arthropathy: a biomineralization disorder. Human Pathology 17:543 (1986). 5. R.A.Terkeltaub. Pathogenesis and treatment of crystal-inducedinflammation, in: Arthritis and Allied Conditions. A Textbook of Rheumatology. W.J. Koopman, ed; Williams and Wilkins. Baltimore. MD 2085 (1997). 6. W.E. Brown and T.M. Gregory. Calcium pyrophosphate crystal chemistry. Arthritis and Rheum. 19:446 ( 1976). 7. L.M. Ryan, I.V. Kurup, B.A. Derfus and V.M. Kushnaryov. ATP-induced chondrocalcinosis. Arthritis Rheum. 35: I520 (1992). 8. K.P.H. Pritzker. Calcium pyrophosphate crystal formation and dissolution, in: Calcium Phosphates in Biological and lndustrial systems. Z. Amjad , ed; Chpt. 12: 277 Kluwer Acad. Publ.Dordrecht. Netherlands (1998). 9. D.J. McCarty, D.W. Palmer, and P.B. Halverson. Clearance of calcium pyrophosphate dihydrate crystals in vivo. I. Studies using 169yb labelled triclinic crystals. Arthritis Rheum. 22:718 (1979). 10. H.M. Burt and J.K. Jackson. Characterization and membranolytic effects of triclinic calcium pyrophosphate crystals. J Rheum. 14:968 (1987). 11. P-T. Cheng, K.P.H. Pritzker, M.E. Adams, S.C. Nyburg, and S.A. Omar. Calcium pyrophosphate crystal formation in aqueous solutions. JRheum. 7: 609 (1980). 12. N.S. Mandel and G.S. Mandel. Calcium pyrophosphate crystal deposition in model systems in: Rheum Disease Clinics of North America 14: 321 (1988). 13. N. Mandel, and G. Mandel. A model for human calcium pyrophosphate crystal deposition disease. Crystallization kinetics in a gelatin matrix model. Scanning Electron Microscopy IV: 1779 (1984). 14. K.P.H. Pritzker, P-T. Cheng, M.E. Adams, and S.C. Nyburg. Calcium pyrophosphate dihydrate crystal formation in model hydrogels. J of Rheumatology. 5:469 (1978)

67

15. G. Williams and J.D. Sallis. Structure-activity relationship of inhibitors of hydroxyapatite formation. Biochem J. 184: 181 (1979). 16. W.P. Tew, C. Mahle, J. Benavides, J.E.Howard, and A.L. Lehninger. Synthesis and characterization of phosphocitrate, a potent inhibitor of hydroxyapatite crystal growth. Biochemistty.l9:1983 (1980). 17. G. Williams and J.D. Sallis. Structural factors influencing the ability of compounds to inhibit hydroxyapatite formation. Calc Tissue Int. 34: 169 (1982). 18. J.D. Sallis, N.F.G. Parry, J.D. Meehan, H. Kamperman, and M.E. Anderson. Controlling influence of phosphocitrate in vitro and in vivo on calcium oxalate crystal formation and growth. Scanning Microscopy 9: 127 (1995). 19. A. Wierzbicki, C.S. Sikes, J.D. Sallis, J.D. Madura, E.D. Stevens, and K.L. Martin. Scanning electron microscopy and molecular modeling of inhibition of calcium oxalate monohydrate crystal growth by citrate and phosphocitrate. Calc Tissue Int. 56:297 (1995). 20. J.D. Sallis, W. Juckes, and M.E. Anderson. Phosphocitrate: Potential to influence deposition of scaling salts and corrosion. In: Mineral Scale Formation and Inhibition (Z. Amjad, ed.). Plenum Press, NY. Chapt.8: 87 (1996). 21. H.S. Cheung, I.V. Kurup, J.D. Sallis, and L.M. Ryan. Inhibition of calcium pyrophosphate dihydrate crystal formation in articular cartilage vesicles and cartilage by phosphocitrate. J Biol Chem. 271:28082 (1996). 22. J.D. Sallis, R.Thomson, B. Rees, and R. Shankar. Reduction of infection stones in rats by combination antibiotic and phosphocitrate therapy. J Urol. 140: 1063 (1989). 23. A. Wierzbicki, J.D. Sallis, E.D. Stevens, M. Smith, and C.S. Sikes. Crystal growth and molecular modeling studies of inhibition of struvite by phosphocitrate. Calc Tissue Int. 61:216 (1997). 24. J.D. Sallis, R. Shankar. B. Rees, and R. Thomson. Protection of crystal-induced polymorphonuclear leukocyte membranolysis by phosphocitrate. Biochem Med Metab Biol. 4 1 :56 (1 989). 25. D. Nair, R.P. Misra, J.D. Sallis, and H.S. Cheung. Phosphocitrate inhibits a basic calcium phosphate and calcium pyrophosphate dihydrate crystal-induced mitogen-activated protein kinase cascade signal transduction pathway. J Biol Chem. 272: 18920 (1 997). 26. L.M. Ryan and D.J. McCarty. Calcium pyrophosphate crystal deposition disease, pseudogout, and articular chondrocalcinosis, in: Arthritis and Allied Condition. A Textbook of Rheumatology. W.J. Koopman, ed; Williams and Wilkins, Baltimore, Md. 2103 (1997). 27. J.D. Sallis. Structure/performance relationships of phosphorous and carboxyl containing additives as calcium phosphate crystal growth inhibitors, in: Calcium Phosphates in Biological and Industrial Systems. Z. Amjad, ed; Kluwer Acad. Publ. Dordrecht, Netherlands. 173 (1998). 28. M. Johnsson, C.F. Richardson, J.D. Sallis, and G.H. Nancollas. Adsorption and mineralization effects of citrate and phosphocitrate on hydroxyapatite. Calc Tissue Int 49: 134 (1991). 29. V.K. Sharma, M. Johnsson, J.D. Sallis, and G.H. Nancollas. Influence of citrate and phosphocitrate on the crystallization of octacalcium phosphate. Langmuir. 8:676 (1992). 30. A. Wierzbicki and H.S. Cheung. Molecular modeling of inhibition of crystals of calcium pyrophosphate by phosphocitrate. Theochem. 454:287 (1998). 31. N.S. Mandel. The crystal structure of calcium pyrophosphate dihydrate. Acta Ctystallographica. B3 1:1730 (1973). 32. G.S. Mandel, K.M. Renne, A.M. Kolbach, A.M. Kaplan, J.D. Miller, and N.S. Mandel. Calcium pyrophosphate crystal deposition disease: preparation and characterization of crystals. J of Crystal Growth. 87: 453 (1988). 33. T. Shinozaki, Y. Xu, T.F. Cruz, and K.P.H. Pritzker. Calcium pyrophosphate dihydrate (CPPD) crystal dissolution by alkaline phosphatase: interaction of alkaline phosphatase on CPPD crystals. J Rheum. 22: 117 (1995). 34. K.D. Altria, M. Wallberg, and D. Westerlund. Separation of a range of cations by nonaqueous capillary electrophoresis using indirect and direct detection. J Chromatogr. B 714:99 (1998). 35. C.H. Fiske and J. Subbarow. Colorimetric detection ofphosphorus. J Biol Chem. 66:375 (1925). 36. P-T. Cheng and K.P.H. Pritzker. The effect of calcium and magnesium ions on calcium pyrophosphate crystal formation in aqueous solutions. J Rheum. 8: 772 (I981). 37. M. Doherty and P.A. Dieppe. Double blind, placebo controlled trial of magnesium carbonate in chronic calcium pyrophosphate arthropathy. Ann Rheum. Dis. 42 (suppl): 106 (1983). 38. T. Shinozaki and K.P.H. Pritzker. Polyamines enhance calcium pyrophosphate dihydrate crystal dissolution. J Rheumatol 22: 1907 (1995). 39. H.E. Krug, M.L. Mahowald, P.B. Halverson, J.D. Sallis, and H.S. Cheung. Phosphocitrate prevents disease progression in murine progressive ankylosis. Arthritis and Rheumatism. 36: 1603 (1993). 40. A.L. Lehninger. The possible role of mitochondria and phosphocitrate in biological calcification, in: Biomineralization and Biological Metal Accumulation. P. Westbroek and E. W. de Jong, eds., D. Reidel Publishing Co., NY. 107 (1983).

68

41. G. Williams and J.D. Sallis. The sources of phosphocitrate and its role as an inhibitor of calcium phosphate and calcium oxalate crystallization, in: Urolithiasis. L.H. Smith and W.R. Finlayson, ed., Plenum Press. NY. 569 (1981). 42. J.D. Meehan and J.D. Sallis. Kidney selective prodrugs of phosphocitrate: Synthesis and in vivo hydrolysis of γ-glutamyl derivatives of phosphocitrate and their in vivo influence on nephrocalcinosis. Pharm Sci. 1: 289 (1995).

69

This page intentionally left blank.

PERFORMANCE OF ANIONIC POLYMERS AS PRECIPITATION INHIBITORS FOR CALCIUM PHOSPHONATES: THE INFLUENCE OF CATIONIC POLYELECTROLYTES

Zahid Amjad, Robert W. Zuhl, and Jennifer A. Thomas-Wohlever Performance Materials The BFGoodrich Company Brecksville, Ohio 44141

ABSTRACT This paper addresses the precipitation of calcium salts of phosphonates (i.e., hydroxyethylidine 1,1 -diphosphonic acid, HEDP; aminotri(methylene phosphonic acid), AMP; 2-phosphono-butane 1,2,4-tricarboxylic acid, PBTC) at pH 8.50 and 50°C. The aim of the study was to investigate the role of cationic polyelectrolytes (i.e., and poly(acrylamide:2poly(diallyldimethyl ammonium chloride), Poly-E (acrolyloxy)ethyltrimethyl ammonium chloride), Poly-F) on the performance of several anionic calcium phosphonate inhibiting polymers. The anionic polymers studied include: acrylic acid/maleic acid-based homo-, co-, and terpolymers; fulvic acid (FA) and tannic acid (TA). The inhibiting data on several anionic polymers indicate that the performance of polymers depends on the type of the functional groups present in the polymer. It has been found that addition of low concentrations (from 0.25 to 1 .0 parts per million) of PolyE and Poly-F adversely impacts the performance of calcium phosphonate inhibiting polymers. The results also indicate that under similar experimental conditions Poly-F exhibits a stronger antagonistic affect than Poly-E.

INTRODUCTION Organophosphorus compounds form a broad family of chemicals which find applications in a variety of industrial processes including crude oil production, pigment dispersion, electroplating, paper and pulp slurries, scale removal, and industrial water treatment.1 In biological systems, phosphonates have been shown to decrease tartar buildup on tooth enamel and assist in preventing kidney stones.2 For their in vivo application, phosphonates are best known for their ability to block bone resorption, treating afflictions

Advances in Crystal Growth Inhibition Technologies Edited by Amjad Kluwer Academic/Plenum Publishers. New York. 2000

71

such as Paget’s disease, hypercalcemia of tumor origin and in other bone resorptive process disorders such as osteoporosis.3 These organophosphonates differ structurally from polyphosphates in that they have a P-C bond rather than a P-O bond. The structural differences account for their superior stability under extremes of pH and temperature. In water treatment processes, phosphonates are used for a variety of reasons. Most important, they inhibit the formation of scale forming salts and control steel corrosion. Scale formation is prevented in a “substoichiometric” manner as phosphonates adsorb onto the crystal growth sites of sub-microscopic crystallites, interfering with crystal growth, and altering the morphology of those crystals that do grow. This process can delay or prevent crystal growth. Phosphonates are also used to sequester metal ions such as iron, aluminum, copper, manganese, and zinc. Additionally, phosphonates exhibit some dispersancy activity towards suspending matter such as clay, silt, calcium carbonate, etc. As corrosion inhibitors phosphonates may generally be described as cathodic inhibitors. They function by reacting with calcium and other polyvalent metal ions, particularly those of the corrosion products. The metal-phosphonate salts form a protective nonconductive layer on the metal surfaces. This barrier separates the metal from the bulk water and prevents diffusion of oxygen to the metal surface, thereby preventing corrosion. Although there are many phosphonates available, three of the most commonly used phosphonates in water treatment formulations are aminotris(methylene phosphonic acid), AMP; hydroxyethylidine 1,1 -diphosphonic acid, HEDP; and 2-phosphonobutane 1,2,4– tricarboxylic acid, PBTC. However, under certain pH, concentration, and temperature conditions, HEDP has been shown to precipitate in the presence of calcium. The precipitation of calcium-HEDP salt not only creates fouling of heat exchanger and reverse osmosis membrane surfaces, it also decreases the solution concentration of HEDP to such an extent that severe calcium carbonate scaling can occur. The structures of phosphonates investigated in the present study are shown in Table 1. Table 1. Phosphonate structures

Major progress has been made toward the understanding of phosphonate solution and precipitation chemistry and how it relates to fouling problems in industrial water systems. Ashcraft4 studied the influence of various factors such as pH, temperature, calcium, and phosphonate concentrations on the performance of HEDP, AMP, and PBTC in the presence of soluble iron. Also studied was the effect of these factors on the loss of phosphonate due to adsorption onto Fe (III) sludge. These evaluations showed PBTC to be superior to AMP

72

and HEDP under conditions of high hardness or pH, presumably due to the greater solubility of the Ca-PBTC salt under harsh conditions. Hogue,5 in his investigation of the precipitation of phosphonates as a function of calcium ion concentration and pH, reported that under cooling water conditions, phosphonates could form precipitates, presumably due to the formation of an insoluble calcium phosphonate complex. Furthermore, AMP and HEDP were shown to exhibit about the same susceptibility to precipitation. Inhibition of calcium phosphonate precipitation by polymeric inhibitors has recently been the subject of intensive research. Smyk and Hoot6 in their investigation on the influence of various polymers on the precipitation of calcium phosphonate showed that acrylic-acid based terpolymers perform better than homo-polymers of acrylic acid and copolymers of acrylic acid and hydroxylpropyl acrylate. Boffardi and Schweitzer7 recognized that the relatively poor calcium tolerance of phosphonates could be overcome by the addition of a copolymer of acrylic acid and 2-acrylamido -2-methylpropane sulfonic acid. The influence of various factors such as calcium hardness, pH, and temperature on the precipitation of Ca-HEDP in the presence of various polymeric inhibitors has recently been reported.8 Results of this study reveal that inhibitor performance decreases with increasing pH (7.5 to 9.0), temperature (30° to 70°C), and calcium hardness (75 to 300 mg/L). Furthermore, inhibition data also suggests that polymer performance strongly depends upon the composition, molecular weight, and ionic charge of the polymer.9 The type and extent of impurities such as iron, manganese, aluminum, suspended matter, etc., present in recirculating cooling water has been reported to exhibit a marked antagonistic effect on the performance of polymers used in cooling water treatment formulations.10 It has also been shown that the effectiveness of surface water treatments in reducing suspended solids is dependent upon the proper selection and feed rate of coagulants or flocculents, pH, mixing, and residence time, etc. Chemicals commonly used in a coagulating or flocculating capacity include alum, ferric chloride, and diallyldimethyl ammonium chloride, DAC. These chemicals are known to "carry over" and have been reported to decrease the performance of calcium phosphate inhibiting polymers.11 Thus, the presence of these chemicals in the water may require additional amounts of polymer for inhibiting scale formation and growth in the system. The present work is a quantitative study of the effects of various commercial scale control polymers on the precipitation of calcium phosphonates. This study also reports the impact of cationic polymeric flocculants on the performance of calcium phosphonate precipitation inhibitors. In addition, for comparative analysis, precipitation experiments have also been made in the presence of fulvic acid and tannic acid for their influence in inhibiting the precipitation of calcium phosphonates.

EXPERIMENTAL Analytical reagent grade chemicals and grade A glassware were used. Calcium stock solutions were prepared from calcium chloride dihydrate and were standardized as described previously.9 Double deionized distilled water was used in this study. The phosphonates used in this study are commercial materials and were standardized by spectrophotometric method.9 The inhibitors were selected from a range of experimental and commercial polymers. The polymer concentrations given in this report are on a dry polymer basis. All polymers were tested under equal precipitation conditions, i.e., at the same supersaturation, temperature, and pH of the solution. A Brinkmann pH meter (model 632, Brinkmann Instruments, Westbury, New York) equipped with a combination electrode was used to measure pH. The pH was calibrated before each set of measurements against standard buffer solutions.8

73

Supersaturated solutions of calcium phosphonate were prepared by adding a known volume of phosphonate (10 to 20 mL) stock solution to a glass bottle (250mL) thermostated at constant temperature, containing a known volume of water and sodium bicarbonate solution. Following temperature equilibration, the calcium stock solution was added in such an amount that the final concentration would be 150 mg/L. Polymer solutions were added after the addition of phosphonate, but before the addition of calcium chloride, as a dilute solution in water. The pH of the calcium phosphonate supersaturated solutions was adjusted to the required value using dilute solutions of sodium hydroxide and lor hydrochloric acid. Precipitation in calcium phosphonate supersaturated solutions was monitored by analyzing aliquots of the filtered (0.22 micron) solution for phosphonate concentration using the standard colorimetric method.9 Polymer efficacy as a calcium phosphonate inhibitor was calculated using the following equation: Inhibition (%I) where: [Phosphonate]sample = [Phosphonate]control = [Phosphonate]initial =

[Phosphonate]sample – [Phosphonate]control

= x 100 (1)

[Phosphonate]initial – [Phosphonate]control concentration of phosphonate in the filtrate in the presence of inhibitor at 20 hours concentration of phosphonate in the filtrate in the absence of inhibitor at 20 hours concentration of phosphonate in the solution at time = 0 hour

Unless noted otherwise, all precipitation experiments were carried out under similar experimental conditions i.e., Ca 150 mg/L, 15 mg/L phosphonate, pH 8.50, 50°C, 22 hr.

RESULTS AND DISCUSSION Currently, many cooling water systems are being operated under high cycles of concentration due to water conservation and waste water discharge regulations. However, these practices increase scaling tendencies in recirculating water and often require effective scale inhibitors to prevent the formation and deposition of unwanted materials on heat exchanger and other equipment surfaces. In order to deal with these challenging problems, an effective cooling water treatment must control scale, corrosion, particulate matter, and microbiological growth. In the past two decades many water soluble polymers with different compositions and molecular weights have been developed and studied as precipitation inhibitors for a variety of alkaline earth metal salts of carbonate, sulfate, phosphate, etc.10-12 These polymers are mainly homo-, co- or ter-polymers having acrylic, maleic, or methacrylic acid as the initial monomer in conjunction with other monomers containing acrylamide, ester, sulfonic, phosphono, phosphino, and hydroxyl groups. In addition to acting as precipitation inhibitors some of the polymers act as a dispersant for particles suspended in solution and also as a metal ion stabilizer for iron, aluminum, manganese, and zinc.15,16 The ability of these polymers to act as a precipitation inhibitor, dispersant and/or metal ion stabilizer depends on the polymer composition, ionic charge, and the molecular weight. Table 2 shows the structures of polymers evaluated in the present study.

74

Table 2. Polyelectrolytes evaluated in this study Polymer Structure Poly(acrylic acid)

Poly H-1

Poly(methacrylic acid)

Poly H-2

Poly(maleic acid)

Poly H-3

Poly(2-acrylamido-2methylpropane sulfonic acid)

Poly H-4

Poly(acrylamide)

Poly H-5

Poly(acrylic acid: hydroxypropyl acrylate)

PolyC-1

Poly(acrylic acid: 2acry lamido-2-methylpropane sulfonic acid)

PolyC-2

Poly(acrylic acid: vinyl acetate)

PolyC-3

Poly(acrylic acid: 2acry lamido-2-methylpropane sulfonic acid: sulfonated styrene)

Poly T-1

Poly(acrylic acid: 2acrylamido-2-methylpropane sulfonic acid: substituted acrylamide)

Poly T-2

75

Calcium Phosphonates Precipitation and Inhibition For industrial water treatment, phosphonates have been shown to be broadly effective in controlling scale and mild steel corrosion. A loss of scale and corrosion control can occur if calcium phosphate salts precipitate. Although phosphonates are generally considered to be tolerant to hardness ions, however, under harsh conditions phosphonates can precipitate as calcium or magnesium salts or coprecipitate with other salts and deposit on heat exchanger surfaces. Thus, for better scale and corrosion control it is important to prevent the precipitation of calcium phosphonates in the recirculating waters. Screening tests for Ca-AMP, Ca-HEDP, and Ca-PBTC precipitation were carried out under standard test conditions. Figure 1 shows the plots of residual phosphonate as a function of time. As illustrated in Figure 1 within 1 hr >90% HEDP precipitated out as calcium salt compared to ~15%AMP and <5% PBTC. It is interesting to note that further increase in reaction time from 1 hr to 20 hours does not significantly affect the residual phosphonate concentration in the calcium phosphonate supersaturated solution. Based on the precipitation data presented in Figure 1, the ranking of phosphonates in terms of decreasing calcium salt precipitation (or calcium ion tolerance) follows: PBTC>>AMP>HEDP.

Figure 1. Residual phosphonate concentration in the absence of polymer as a function of time.

Performance of Polymers. During the last decade there has been an increasing interest in the application of ter-polymers as scale and deposit control agents, primarily because they are highly efficient in preventing the nucleation and crystallization of many scale forming salts. The main advantage in using ter-polymers lies in their excellent thermal stability compared with other co-polymers. Figure 2 presents results on the performance of Poly T-1 as a calcium phosphonate inhibitor. It can be seen that under similar experimental conditions performance of Poly T1 as calcium phosphonate inhibitor depends upon the type of phosphonate present in the supersaturated solution. For example, in the presence of 5 ppm Poly T-1, % inhibition values obtained for Ca-HEDP and Ca-AMP salts are 4% and 77% compared to 92% obtained for Ca-PBTC salt. The data presented in Figure 3 clearly indicate that Poly T-1 concentration has to be increased by a factor of two (i.e., 10 ppm) to achieve > 90% inhibition.

76

Figure 2. Calcium phosphonate inhibition by Poly T-1.

Figure 3 profiles calcium-HEDP inhibition (%I) versus polymer concentration for two inhibitors namely homo poly(acrylic acid) , Poly H- 1 and acrylic acid-based terpolymer, Poly T- 1 are illustrated. The inhibition data suggest that polymer concentration strongly affects the ability of a polymer to inhibit the precipitation of Ca-HEDP salt. For example, in the presence of 5 ppm dosage, Poly-T1 shows poor inhibitory activity (~5% I). However, at 7.5 ppm, polymer shows marked improvement (29 %I) and at 10 ppm polymer concentration, the Ca-HEDP precipitation is essentially inhibited (>90% I) for at least 20 hour. For performance comparison, inhibition data on Poly-H1 is also shown in Figure 3. The inhibition data on these polymers clearly indicate that under similar precipitation condition, a ter-polymer outperforms poly(acrylic acid) thus suggesting that carboxyl group alone is not sufficient in controlling the precipitation of Ca-HEDP. It is worth noting that where as for Ca-HEDP and calcium phosphate, ter-polymer is an effective inhibitor, for other sparingly soluble salts such as CaSO4•2H2O, and CaCO3, homo-polymer of acrylic acid has been reported to offer the best performance.12,15

Polymer (ppm) Figure 3. Calcium-HEDP inhibition vs. polymer concentration for Poly H-1 and Poly T-I,

77

Figure 4 presents performance data on several commercial polymers used in preventing the deposition of unwanted materials on heat exchanger surfaces. As can be seen from Table 2 the polymers contain a variety of monomers of different functionality. From the data presented in Figure 4, the polymers evaluated fall into the following three categories: A. Polymers that exhibit poor performance are those homopolymers which contain carboxyl group such as poly(acrylic acid), Poly-H 1 ; poly(methacrylic acid); PolyH2 and poly(maleic acid), Poly-H3; sulfonic group (i.e, Poly H-4) or acrylamide group (i.e., Poly H-5). It is worth noting that polymers which are devoid of – COOH group (i.e, Poly H4 and Poly H-5) are relatively ineffective Ca-HEDP inhibitors compared to those that do contain – COOH groups. For example, under similar experimental conditions inhibition values obtained for – COOH containing polymers are -10 to 18% compared to ~ 2% obtained for other homopolymers. B. Polymers that show mediocre performance (% inhibition 40-70%) contain at least one and/or two comonomers of different functional groups (i.e., -COOR, -CONH2, -SO3H, etc.). Results presented in Figure 4 clearly demonstrate that, compared to homopolymers, the copolymers (i.e, Poly C-1, Poly C-2, and Poly C-3) show better performance as Ca-HEDP inhibitors. C. Polymers in this category show superior performance in terms of preventing Ca-HEDP precipitation. Among the terpolymers tested it appears that the terpolymer which is more ionic (i.e., Poly-T1 vs. Poly-T2) in nature exhibits better performance in terms of inhibiting the precipitation of calcium phosphonate. Based on the inhibition data presented in Figure 4 the polymers can be ranked (in terms of decreasing effectiveness) as: Terpolymers > Copolymers > Homopolymers

Figure 4. Inhibition of calcium phosphonate precipitation by polymers of varying composition at 10 ppm.

Performance of Humic Compounds. Humic and fulvic acids, commonly found in the natural environment, are polymers whose molecular weights range from several hundred to several thousand. They mostly contain carboxylic and phenolic acid functionalities, and can behave as negatively charged colloids or anionic polyelectrolytes in

78

surface waters. Nystrom et al.16 in their study on the filtration of process waters containing humic acid by membrane-based filtration procedures reported that humic acid forms a gellike layer on the filter and blocks the pores of the filter. Freche et al.17 in a seeded growth study showed that low levels of fulvic acid markedly inhibit dicalcium phosphate dihydrate crystal growth. Similar results were also obtained in studies involving the effect of fulvic acid on the crystal growth of hydroxyapatite and calcium carbonate. A Langmuiradsorption process at surface growth sites was proposed to explain the inhibitory effect of humic substances. Effects of polyelectrolytes i.e., fulvic acid, tannic acid, poly(aspartic acid), and poly(acrylic acid) on calcium carbonate precipitation has also been recently investigated. X-ray diffraction studies of calcium carbonate precipitated in the presence of polyelectrolytes reveal that whereas poly(acrylic acid) and poly(aspartic acid) favor the formation of vaterite, fulvic acid and tannic acid influence the formation of calcite. The polymorph formed, however, in the absence of polyelectrolytes is vaterite.18 In order to compare the performance of humic substances as inhibitors for Ca-HEDP with synthetic polymers, a series of experiments were performed under similar conditions. Results are presented in Figure 5. Also shown, for comparison, are the inhibition data on synthetic polymers. It can be seen that fulvic acid and tannic compared to Poly-T1 show poor inhibitory activity. For example, % inhibition values obtained for Poly-T1 at 10 ppm concentration are 91% compared to ~5% obtained for fulvic acid and tannic acid, respectively. It should be noted that increasing the concentration of humic acid by a factor of two (i.e., from 10 to 20 ppm) does not show any significant improvement on the inhibitory power of the humic compounds. The poor inhibitory activity shown by humic compounds is consistent with the performance seen for Poly H-1 and Poly H-2 suggesting that polymers containing only carboxyl and/or hydroxyl groups show poor affinity for CaHEDP nuclei. Another possible factor contributing to the poor performance of humic compounds is their relatively high molecular weight. It is well documented that an effective precipitation inhibitor normally has a molecular weight of ~ 10,000. Studies are currently underway to further develop inhibitor structure - performance relationships. Effect of Cationic Polyelectrolytes For decades cationic polyelectrolytes have been used as flocculant/coagulant for isolation and separation of colloidal particles from various water streams. Commonly used polyelectrolytes include aluminum and iron containing

Polymer (ppm) Figure 5. Performance of humic substances as calcium phosphonate inhibitors versus Poly T-1 and Poly H-1.

79

compounds such as alum, ferric chloride, ferric sulfate, etc. These polyelectrolytes neutralize the charge of the colloidal particles in the water, and hydrolyze to form insoluble precipitates that, in turn, entrap additional particles. In most cases these large particle (flocs) are removed via settling in a clarifier and are recollected as sludge. Occasionally, clarifier upsets cause the metal ion-containing particles to carryover. In such instances, aluminum or iron-based deposits can form on heat exchanger and reverse osmosis membrane surfaces. Although aluminum- and iron-based compounds can exhibit positive effects in terms of clarifying the water, the optimum performance of these compounds is very sensitive to water pH and alkalinity. To overcome the shortfall of inorganic flocculants, synthetic organic polyelectrolytes (linear, branched, and lightly crosslinked) have been introduced in the past several years. Among the commercial cationic polyelectrolytes, homopolymer of diallyldimethyl ammonium chloride (Poly E) has been the preferred choice due to its high performance and reasonable cost. Recently, from a water conservation perspective, there has been an increasing trend in operating cooling water systems at high cycles of concentration resulting in greater driving force for the precipitation of scale forming salts. To overcome this increased scaling potential a high performing polymer is required in the water treatment formulation in order to protect the heat exchanger from fouling. Another measure of the effectiveness of a CaHEDP inhibitor is its efficacy in the presence of process water impurities such as cationic polyelectrolytes. The influence of varying concentrations of cationic polymer Poly-E on the performance of calcium phosphonate inhibiting polymers was investigated at fixed calcium phosphonate supersaturation. Figure 6 shows calcium phosphonate inhibition at varying concentrations of Poly T-1 in the presence of set levels of Poly-E (0 to 1 ppm in solution). It can be seen from Figure 6 that Poly-E exerts an antagonistic influence on the performance of Poly T-1 . For example, % inhibition obtained in the presence of 0.25 and 0.50 ppm Poly-E at 12.5 ppm Poly T-1 are 90% and 81%, respectively compared to 96% obtained in the absence of Poly-E. It should be noted that increasing the Poly-E from 0.50 to 1.0 ppm results in ~50% further reduction in inhibition value. In other words, in order to overcome this marked negative influence by cationic polymer, additional amounts (depending upon the level of Poly-E present in the make-up water) of Poly T-1 would be required to ensure complete inhibition of calcium phosphonate precipitation.

Polymer (ppm) Figure 6. Effect of varying levels of cationic polymer, Poly E, on calcium phosphonate inhibition by Poly T- 1.

80

The effect of Poly-E on the Ca-HEDP inhibiting performance of several commercial polymers containing varying functional groups was also investigated. Figure 7 presents comparative performance data on these polymers. It is evident from the inhibition data that polymer performance strongly depends on the nature of the comonomers and/or ionic charge of the functional groups present in the polymer. For example, polymers containing three monomers perform better than polymers containing two monomers. It is worth noting that based on the performance data of two terpolymers (Poly T-1 and Poly T-2) the ionic charge of the polymer also plays an important role in improving the inhibitory activity of the polymer. It is also important to note that similar antagonistic effect by cationic olymer has been reported on the performance of calcium phosphate inhibiting polymers.11 Thus, from a practical point of view, in cooling water applications where Poly E is used, the incorporation of a polymer that is more tolerant to Poly E into water treatment formulations is beneficial. In order to study the influence of the nature of cationic polymer on the performance of anionic polymers a series of calcium phosphonate precipitation experiments were carried

Figure 7. Effect of cationic polymer, Poly E, on percent inhibition of calcium phosphonate precipitation by various polymers.

Figure 8. Effect of cationic polymers Poly E and Poly F on the inhibition of calcium phosphonate by Poly T-1.

81

under similar experimental conditions (i.e., Poly T1 12.5 ppm, cationic polymer 0.5 ppm). Performance data presented in Figure 8 clearly show that compared to diallyldimethyl ammonium chloride polymer (Poly E), hydrolyzed acrylamide polymer (Poly F), exhibits a larger antagonistic effect. For example, inhibition values obtained in the presence of 0.5 ppm of Poly E and Poly F are 82 % and 62 % respectively, compared to obtained in the absence of cationic polymer. From the information presented in Figures 7 and 8, the performance of anionic polymers as calcium phosphonate inhibitors strongly depends upon the amount and the type of the cationic polymer present in the calcium phosphonate supersaturated solution. Thus, in view of the results discussed above the impact of several factors on the performance of polymer should be considered before a final selection of polymer is made to control the precipitation of calcium phosphate under stressed field conditions.

CONCLUSIONS The results presented in this paper support the following conclusions regarding the performance of calcium phosphonate inhibitors in the presence of cationic polyelectrolytes under test conditions commonly encountered in cooling water applications.

•

Phosphonates such as PBTC, AMP, and HEDP interact with calcium ions to form insoluble salts of varying solubility. Based on the precipitation data the order in terms increasing solubility of calcium phosphonate is: PBTC>> AMP > HEDP

•

Humic compounds suchas fulvic acid and tannic acid are ineffective calcium phosphonate inhibitors.

•

Homopolymers such as poly(acrylic acid), poly(maleic acid), poly(acrylamide) and poly(2-acrylamido 2-methyl propane sulfonic acid) exhibit poor performance as calcium phosphonate inhibitors.

•

Incorporation of bulky comonomers containing anionic and/or nonionic groups in the polymer significantly increase the inhibitory power of the polymer. Based in the inhibitory the ranking in terms of polymer effectiveness as calcium phosphonate inhibitor is: terpolymer > copolymer > homopolymer

•

Polymer composition is critical to the the performance of calcium phosphonate inhibiting polymer. An increase in co- or ter-polymer concentration results in a significant increase in the inhibition of calcium phosphonate precipitation. The ability of a polymer to perform at low concentration is an important selection criteria.

• The presence of low concentrations of cationic polyelectrolytes adversely impacts the the inhibitory performance of polymers. The incorporation of sulfonated styrene in a terpolymer greatly improves its tolerance to cationic polyelectrolyte.

REFERENCES 1. Cowan, J. C. and Weintritt, D. J., Water-Formed Scale Deposits, Gulf Publishing Company, Houston, TX, 1975 2. Amjad, Z., Calcium Phosphates in Biological and Industrial Systems, Kluwer Academic Publishers, Boston, MA, 1998

82

3. 4. 5.

6.

7.

8. 9. I0. 1I. 12. 13. 14. 15. 16. 17.

Ebrahimpour, A,, Ebetino, F. H., Sethuraman, G., and Nancollas, G. H., in Mineral Scale Formation and Inhibition, Amjad, Z., Ed., Chap. 24, Plenum Publishing Corporation, New York, NY, 1995 Ashcraft, R. H., Scale Inhibition Under Harsh Conditions by 2-Phosphonobutane-l,2,4Tricarboxylic Acid, CORROSION/85, Paper No. 123, National Association of Corrosion Engineers, Houston, TX, 1985 Hogue, R. R., Sensitivity of DequestR Phosphonates to Elevated Levels of Calcium, Report No. 8433, Monsanto Company, September 13, 1976 Smyk, E. E., Hoots, H. E., Fivizzani, K. F., Fulks, K. E., The Design and Application of Polymers in Cooling Water Programs, CORROSION/88, Paper No. 14, National Association of Corrosion Engineers, Houston, TX, 1988 Boffardi, B. P.,and Schweitzer, G. W., Advances in the Chemistry of Alkaline Cooling Water Treatment, CORROSION/85, Paper No. 132, National Association of Corrosion Engineers, Houston, TX, 1985 Masler, W. F. and Amjad, Z., Advances in the Control of Calcium Phosphonate with a Novel Polymeric Inhibitors, CORROSION/88, Paper No. 11, National Association of Corrosion Engineers, Houston, TX, 1988 Amjad, Z., Performance of Polymers as Precipitation Inhibitors for Calcium Phosphonate, Tenside Surfactants Detergents 34, 102, 1997 Amjad, Z., Butala, D., and Pugh, J., The Influence of Recirculating Water Impurities on the Perfromance of Calcium Phosphate Inhibiting Polymers, Paper , NACE International, Houston, TX, 1999 Amjad, Z., Pugh, J., Zibrida, J. and Zuhl, R., Polymer Performance in Cooling Water: The Influence of Process Variables, Materials Performance 36, 32, 1997 Amjad, Z., Hooley, J., Effect of Antiscalants on the Precipitation of Calcium Carbonate in Aqueous Solutions”, Tenside Surfactants Detergents, 3 1, 12, 1994 Amjad, Z., Controlling Metal Ions Fouling in Industrial Water Systems, accepted for publication, ULTRAPURE WATER 1999 Amjad, Z., Factors to Consider in Selecting a Dispersant for Industrial Water Systems, accepted for publication, ULTRAPURE WATER, 1999 Amjad, Z., Calcium Sulfate Dihydrate (Gypsum) Scale Formation on Heat Exchanger Surfaces: The Influence of Scale Inhibitors, J. Colloid and Surface Science, 123, 523, 1988 Nystrom, M., Ruohomaki, N, kaipia, L, Humic Acid as Fouling Agent in Filtration”, Desalination, 106,79, 1996 Freche, M., Rouquet, N., Koutsoukos, P. G., and Lacout, J. L., effect of Humic Compounds on the Crystal Growth of Dicalcium Phosphate Dihydrate, Agrochimica 36, 500, 1992

83

This page intentionally left blank.

CRYSTAL GROWTH OF CALCIUM CARBONATE IN SEAWATER. THE EFFECT OF TEMPERATURE AND OF THE PRESENCE OF INHIBITORS

Angeliki Kladi,1 Pavlos G. Klepetsanis,1,3 Terje Østvold,4 Kontoyiannis,1,3 and Petros G. Koutsoukos1,2

Christos G.

1

Institute of Chemical Engineering and High Temperature Chemical Processes, P.O. Box 1414 Patras, GR-26500, Greece 2 Department of Chemical Engineering, Patras, Greece 3 Department of Pharmacy University of Patras Patras, GR-26500, Greece 4 Institute of Inorganic Chemistry The Norwegian Institute of Technology N-7034 Trondheim, Norway

ABSTRACT The spontaneous precipitation of calcium carbonate from supersaturated seawater solutions has been investigated in the temperature range 25-80°C and under conditions of sustained supersaturation at pH = 8.5. The supersaturation values with respect to calcite varied from 10 to 100 and in all cases the phase precipitated at the initial stages was vaterite, which afterwards transformed into aragonite. This has been established by Scanning Electron Microscopy and powder X-Ray Diffraction analysis of the precipitates. The precipitation was largely homogeneous, as suggested by an average value of surface energy equal to 41.6 mJ/m2, which was calculated for 25, 35 and 50ºC. The induction times measured increased significantly with decreasing supersaturation and the rates of precipitation showed a parabolic dependence on supersaturation, pointing to a surface diffusion mechanism. An activation energy of 34.2 kJ/mol was calculated. At 80°C and at supersaturation values lower than 6 the precipitation of magnesium hydroxide made it impossible to obtain measurements. Amorphous magnesium hydroxide has been detected

Advances in Crystal Growth Inhibition Techologies Edited by Amjad Kluwer Academic/PlenumPublishers NewYork 2000

85

by SEM in the precipitates obtained. In the absence of magnesium ions from the synthetic seawater, the solutions were stable at 80°C at supersaturations lower than 17. The precipitates obtained in this case were analyzed by SEM and XRD and small vaterite and aragonite crystals transforming into calcite were detected. The dominance of calcite was due to the absence of magnesium, which favors the formation of aragonite. The inhibition of the calcium carbonate crystal growth in seawater has also been investigated at 80°C with the use of a polymer of polyacrylic acid with molecular weight equal to 2000. Its strong inhibitory effect (concentrations in the seawater solutions lower than 0.5 ppm) has been attributed to adsorption on active crystal growth sites.

INTRODUCTION The formation of calcium carbonate scale is one of the main problems encountered in the North Sea oil production wells, as seawater and oil well brines are often supersaturated with respect to CaCO3. Changes in temperature and pressure shift the aquatic equilibria and lead to deposition of scale, which adheres to the inner walls of tubing, pumps and other production equipment. This may lead to clogging of boreholes, well damage and expensive clean-up operations. 1,2,3 In most cases the deposits consist of calcite, which is the most thermodynamically stable crystalline polymorph of CaCO3.4 In order of decreasing thermodynamic stability, aragonite and vaterite are the two other polymorphs. To prevent or retard precipitation of calcium carbonate, oil companies often use inhibitors in order to alter nucleation and crystal growth patterns.5,6,7,8 However, a good knowledge of the calcium carbonate precipitation kinetics is necessary for this measure to be effective. The characteristics of the scale crystallites are also important for the determination of the properties of the solid deposit. In the present work spontaneous precipitation experiments of calcium carbonate in supersaturated seawater solutions have been performed in the temperature range 25-8O°C and at pH=8.5. Rates of precipitation at the initial stages of the reaction, rates of crystal growth and induction times are presented as a function of the solution supersaturation with respect to the calcite phase (SRcalcite). The SR values reported have been calculated by use of the Multiscale program supplied by STATOIL, Norway and they lie in the labile region of the calcium carbonate system. The calculation of rates of precipitation and crystal growth is based on the volume of titrant solutions added to maintain constant supersaturation during each experiment. The technique of constant supersaturation9,10 establishes a pseudo-steady state throughout the precipitation process during which the time zero conditions dominate, metastable phases are stabilized and any phase changes occurring can be easily observed. Moreover, it offers the advantage of accurate determination of the rate through monitoring of the rate of addition of reagents to replace the solution components removed from the working solution to the precipitate. In addition to the above experiments, the inhibition of CaCO3 formation in seeded growth experiments by a polymer of polyacrylic acid has been investigated at 80°C.

EXPERIMENTAL Spontaneous Precipitation Experiments The experiments were done in a 400 ml double-walled pyrex vessel thermostatted by water circulated from a constant temperature bath. The working solution was stirred by a

86

plastic propeller mounted on the top of the reactor and the stirring speed was adjusted to 350 rpm. The stock solutions were prepared from reagent grade chemicals (Merck) and triply distilled water and they were filtered through membrane filters (0.2µm, Millipore). The calcium chloride solution was standardized by atomic absorption analysis. The sodium bicarbonate and sodium carbonate solutions were prepared on the day of each experiment. The solution pH was measured by a combination pH electrode (Metrohm). The supersaturated solutions (total volume=250 ml) were prepared in the reactor by mixing 125 ml double concentration seawater (containing neither CaC12 nor NaHCO3), 25 ml NaHCO3 solution (solution B) and 100ml CaCl2 solution (solution C). The desired supersaturation value for each experiment was achieved by variation in the concentrations of solutions B and C. This was done by defining as Tca the total concentration of Ca2+ in the working solution and as THCO3 the total concentration of NaHCO3 in the working solution and calculating their values by selecting arbitrary values for the parameter V* from the equations:

Tca=CcaSW+

0.5V * 250 + 2V *

(1)

and THCO3=CHCO3SW+

0.5V* 250 + 2V *

(2)

where CcaSW and CHCO3SW are the concentrations of Ca2+ and HCO3-, respectively in the seawater. Apart from the experiments at 80ºC, for which lower values were used for the achievement of low SRcalcite values, these concentrations were equal to 0.0103 mo1/1 and 0.002334 mo1/1, respectively. The composition of the double concentration seawater is illustrated in Table 1. As will be discussed in the following section, the precipitation of Mg(OH)2 at 80°C and at supersaturations lower than about 6.0 made it impossible to obtain measurements, so experiments were also performed at 80°C without the use of MgC12 in the seawater. In these cases, no MgC12 was used in the titrant solutions, either. The ionic strength of the supersaturated solutions was adjusted by the addition of the appropriate volume of sodium chloride stock solution in solution C. After the mixing of the above solutions the pH was adjusted to 8.5 by the addition of VN ml NaOH 0.1N.

Table 1. Composition of double concentration seawater

87

The induction time, ô, is defined as the time between the pH adjustment and the onset of precipitation, which is signaled by a decrease in the pH of the solution. The induction times measured were reproducible within ± 1 5%.The first decrease in pH triggered the addition of titrants from two mechanically coupled burettes of an automatic titrator controlled by a personal computer unit equipped with the appropriate I/O interface and software. The titrant solutions compensate for the ions that take part in the formation of CaCO3, so that the solution supersaturation is kept constant. They also contain seawater, so that the seawater in the solution is not diluted after the addition of the titrants. Their composition is illustrated in Table 2. Table 2. Composition of titrant solutions

For each experiment, the concentrations CCaCI2, CNaHCO3, CNa2CO3 and CNaOH were calculated from the following equations (Volume of working solution = 250 ml, total volume of titrant solutions = 50 ml): CCaCI2 = 2 CCaSW + (n+2) (TCa-CCaSW)

(3)

CNaHCO 3 = 2 CNaHCO3 SW

(4)

+2 (THCO3-CNaHCO3SW)

CNa2CO3 = n (THCO3-CNaHCO3SW)

(5)

CNaOH = 4 E-5 VN

(6)

Preliminary experiments have pointed to the use of n =10 so that the process does not accelerate beyond control. However, at T = 80°C, higher values were used to ensure control of the reaction (n=20, 30, 40). The concentration of Na2CO3 corresponds to the stoichiometry of the precipitating CaCO3. The factor 2 corresponds to the dilution anticipated by the addition of the titrant solutions from the two burettes. During the precipitation process samples were withdrawn and filtered through membrane filters (0.2µm, Millipore). The filtrates were analyzed for calcium by Atomic Absorption Spectroscopy (Perkin Elmer Analyst 300), in order to confirm the constant

88

solution composition. In all experiments the analysis showed that throughout the precipitation process the calcium concentration remained constant to within 2%. The solid phase on the filter from the sample withdrawn after the first addition of titrants was analyzed by Scanning Electron Microscopy (SEM, JEOL JSM 5200), so that the first CaCO3 phase formed could be determined. The solid phase after the end of the experiment was also collected by filtration of the solution, dried and analyzed by SEM and Powder XRay Diffraction (XRD, Philips 1840/30). During the precipitation process the pH of the working solution and the added volume of titrants as a function of time were recorded and stored. The rates of precipitation were calculated from the traces of titrant volume added as a function of time, using curve-fitting software. More specifically, the rates of precipitation (expressed in mol CaCO3/min) at the initial stages of the precipitation process were calculated as follows: Rate ofprecipitation [mo1/min] = dV/dt [ml/s] * Ceff [mol/ml] * 60 s/min,

(7)

where Ceff is the effective concentration of the titrants given by: Ceff = n * (TCa — CCaSW)

(8)

and dV/dt is the slope of the V vs. time curve obtained for each experiment, where V is the volume of titrant solutions added. The slope is taken at the point that corresponds to 1% of the theoretically calculated mass mth of CaCO3 produced, according to the Ca2+ and HCO3concentrations of the particular solution studied. The calculation of the rates is illustrated in Figure 1, for a typical experiment. The rates measured were reproducible within 10%.

Time,s Figure 1. IIustration of the point at which the slope of the volume of titrants-time profile of a constant supersaturation experiment is measured, for the calculation of the precipitation rate.

89

Samples were also withdrawn from the working solution and analyzed for the number of crystals formed and their size ditribution in a laser particle counter (SPECTREX, ILIl000).11 The time at which the number of particles reached a constant value indicated the time at which the rate of crystal growth should be determined. At that point the process starts being a seeded crystal growth because calcium carbonate is formed on the particles already present in the supersaturated medium. At this stage the Vtitrant vs. time curve is linear and the rate of crystal growth is calculated from the slope. The calculation is described below for an experiment performed at 35°C and at SRcalcite=39.3. The laser particle counter gave the variation of total number of particles with time illustrated in Figure 2.

Figure 2. Use of the total counts vs. time curve for the determination of the time when crystal growth in seawater begins.

For the experiment corresponding to the example of Figure 2, crystal growth begins at t*=8000s, because at that time the number of CaCO3 particles stops increasing. The mass of CaCO3 produced at time t* is: mCaCO3(1) =ceff[mol/ml] V* [ml] *100g/mol

(9)

where ceff is the concentration of titrants and V* is the volume of titrants at time t*. The Vtitrant vs. time curve for the above experiment is shown in Figure 3. The slope dV/dt of the curve in Figure 3 between the volume V* and and the volume which corresponds to mCaCO3(2)=1. 1mCaCO3(1) is used for the determination of crystal growth according to Equation 10: Rate of crystal growth = dV/dt * Ceff [mol/ml] * Sg-1 [g/m2]/ mCaCO3(1) [g] [mol/min/m2] [ml/min] where Sg is the surface area of calcium carbonate (1 m2/g).

90

(10)

Seeded Growth Experiments in Seawater with the Use of a Polyacrylic Polymer as an Inhibitor The seeded growth experiments were done in the reactor described above at 80°C. No MgC12 was added to the seawater either in the titrants, to ensure that the pH decrease would be due only to precipitation of CaCO3 and not to Mg(OH)2, which is favored at this temperature. Before the use of the inhibitor, a blank experiment was performed with the same amount of seeds and the rate of crystal growth was recorded. The inhibitor used was a water soluble polyacrylic polymer with molecular weight (MW) = 2000. Its structure is described by the formula: CH2– COOH This was supplied in liquid form, it was diluted with triply distilled water and an appropriate volume was added to the solution B containing NaHCO3. The three solutions were afterwards mixed as described above. After the adjustment of pH at 8.5 with NaOH 0.1N, 20 mg calcite seeds (surface area = 0.3 m2/g) were added in the reactor and the experiment started immediately with the addition of titrants. The rates of crystal growth were calculated from the initial slope of the Vtitrant vs. time plots obtained. All experiments were done at supersaturation values low enough to ensure stability of the supersaturated solutions until the seed crystals were introduced. The SRcalcite of the working solutions were 8.44, 11.4 and 13.9. At these supersaturation values, at 80°C, pH = 8.5, no Mg2+ present and without the addition of seeds the seawater solutions are stable. The total concentration of calcium under these conditions was 2.8x10-3 – 3.4x10-3 mol/l.

Figure 3. Determination of rate of crystal growth (example for an experiment at 35°C and Srcalcite=39.3 in seawater).

91

RESULTS AND DISCUSSION Spontaneous CaCO3 Precipitation Experiments in Seawater Measurements of the time lapsed between the establishment of supersaturation and the onset of spontaneous precipitation provide a measure of the tendency of the solutions under study to promote the formation of calcium carbonate. Plots of the induction times as a function of the supersaturation are defined as the stability diagrams or formation curves of the solids and are useful for the definition of the metastability limits of the supersaturated solutions with respect to the mineral scale salt which may precipitate. Figure 4 shows the formation curves for calcium carbonate in seawater over a range of temperatures. As may be seen the induction times increased with decreasing supersaturation.

Figure 4. Stability diagrams for the precipitation of calcium carbonate in seawater at pH=8.5 and at T=25, 35, 50, and 80°c.

At 80°C and at very low supersaturation values it was not possible to obtain measurements because the pH decreased in spite of the fact that the solution was theoretically expected to be stable with respect to CaCO3. At these conditions the precipitation of Mg(OH)2 is favored.12 For this reason the experiments at 80°C were repeated in the absence of MgC12 from the seawater and from the titrant solutions. Figure 5 shows the formation curve for CaCO3 under these conditions and compares it with the previous formation curves. It was found that in the absence of Mg2+ the seawater solutions are stable for supersaturation values lower than 17, implying that at supersaturations lower than 17 there was no calcium carbonate in the precipitate. This was verified by XRD spectra and SEM micrographs of solid precipitates taken from precipitation experiments done in the presence and in the absence of magnesium and at various supersaturation values. ( Figures 6-8 & Pictures 1-5 ). Figure 6 shows the XRD spectrum of the solid precipitate obtained from an experiment done at 80°C in the presence of Mg2+ and at SRcalcite = 41. 92

The three distinct peaks at 2θ = 31.21, 44.97, and 56.07° coincide with the characteristic lines of NaC1, which is present in seawater. As shown, the remaining peaks are assigned to vaterite and aragonite crystals. The SEM micrograph (Figure 7) of the precipitate from the same experiment shows many small spherical crystals characteristic of vaterite.

SRcaIcite Figure 5. Effect of Mg2+ on induction times of supersaturated seawater solutions at 80°C.

Figure 6. XRD spectrum of the solid precipitate from an experiment done at 80°C and SRcalcite=41 in seawater in the presence of magnesium ions; comparison with the standard spectra of vaterite and aragonite phases.

93

Interesting vaterite structures have been observed on solid precipitates obtained under similar conditions (SRcalcite = 31.9). Figures 8 and 9 show lens-shaped and cross-shaped vaterite crystals in addition to spherical vaterite crystals, as observed by SEM. No aragonite crystals were found, probably because their relative amount was small.

Figure 7. T=80°C, seawater in the presence of magnesium ions SRcalcite=41. The precipitate consists of many vaterite crystals (SEM micrograph).

Figure 8. T=80°C, seawater in the presence ofmagnesium ions, SRcalcite=3 1.9. Vaterite crystals forming lens shaped structures (SEM).

94

Figure 9. T=80°C, seawater in the presence of magnesium ions, SRcalcite=3 1. Vaterite crystals forming crossshaped structures(SEM).

Figure 10. XRD spectrum of the solid precipitate obtained at 80°C and SRcalcite=2.99 in seawater in the presence of Mg2+.

No CaCO3 phase has been identified on the solid precipitates obtained at 80ºC in the presence of Mg2+ and at supersaturation values lower than 17. Figure 10 shows the XRD spectrum of a solid precipitate obtained at SRcalcite = 8.44. Only the characteristic peaks of NaC1 are distinguished. The SEM photograph of a precipitate obtained at SRcalcite=l6 (Figure 11) showed the formation of an amorphous mineral phase which was attributed to Mg(OH)2.

95

Figure 11. T=80°C, seawater in the presence of magnesium ions, SRcalcite=16. Amorphous Mg(OH)2 phase (SEM micrograph).

Figure 12 shows the XRD spectrum of the solid precipitate obtained from an experiment done at 80°C in the absence of Mg2+ and at SRcalcite=17.0. It is interesting to note that apart from NaC1, vaterite and aragonite, calcite was also detected. Calcite crystals were also observed by SEM (Figure 13 ). The dominance of calcite is due to the absence of Mg2+, which would stabilize the more soluble aragonite .13,14 As shown by SEM analysis, at all temperatures examined and in the presence of Mg2+ vaterite was initially formed, which afterwards transformed gradually into aragonite. In

Figure 12. XRD spectrum of the solid precipitate obtained from an experiment done at 80°C and SRcalcite =I7 in seawater without Mg2+.

96

Figure 13. T=80°C, seawater in the absence of magnesium ions, SRcalcite=17.0. Characteristic calcite crystals (SEM micrograph).

Figure 14. T=25°C, seawater, initial stages of the precipitation process. Characteristic vaterite crystals (SEM micrograph).

Figure 14, the SEM micrograph shows the characteristic spherullitic vaterite crystals. This sample was taken at 25°C at the initial stages of the precipitation process. It is important to note that the identification of vaterite, which is a transient phase, was made possible by the constant supersaturation approach. The pseudo-steady state established allows the stabilization of metastable phases and any phase changes occurring can be easily observed. At later stages the dominant phase was aragonite. As may be seen in Figure 15, vaterite crystals transform into aragonite upon contact with the mother supersaturated medium. Often the aragonite bundles retain the spherullitic morphology of the initially forming vaterite (Figure 16, T=35°C). The presence of aragonite and vaterite peaks in the XRD spectra of precipitates obtained at 25 and 50°C is shown in Figures 17 and 18.

97

Figure 15. T=25ºC seawater, later stages of the precipitation process. The transformation of vaterite crystals to aragonite crystals.

Figure 16. T=35ºC, seawater, later stages of the precipitation process. Aragonite bundles retaining the spherullitic morphology of the initially formed vaterite.

98

Figure 17. Powder XRD spectra of vaterite, aragonite, and precipitate obtained at 25ºC in seawater.

Figure 18. Powder XRD spectra of precipitate obtained at 50°C; comparison with vaterite and aragonite spectra.

99

According to the classical nucleation theory,15 the induction time is inversely proportional to the supersaturation with respect to the precipitating solid. The dependence of the measured induction time on SR is described by the equation: log τ = A + B γs3 / ((2.303kT)3log2SR)

(11)

where: A is constant,y is the surface energy of the growing phase, B = ßυm2 with ß a shape factor (ß =16π/3 for spheres) and υm the molecular volume of the growing phase υ -29 m3 for CaCO ), k is the Boltzmann constant and T is the temperature in K. As m=3.29x10 3 already discussed, the CaCO3 polymorph that precipitates at the initial stages in seawater and at all temperatures examined is vaterite, which afterwards transforms into aragonite. For this reason, the supersaturation values of the seawater solutions for this part of the analysis will be referred to with respect to vaterite. Plots of the logarithm of the induction time vs. the 1/log2(SR varerite) are presented in Figure 19 for 25, 35, and 50ºC. According to Equation 11, it is possible to calculate the surface energy γs of the precipitating vaterite for each temperature from the values the slopes of the straight lines obtained from the fit of kinetics data in eq. 11. The calculated value of the surface energy is an indirect parameter of a value difficult to be obtained otherwise for polycrystalline solids. However due to experimental uncertainties the calculated value may be used as a measure of the degree of homogeneity of a nucleation process. As may be seen in Figure 19, the surface energy of vaterite did not show any significant variation over the temperature range investigated. Its order of magnitude γ savg= 41.6 2 mJ/m suggests that the process is largely homogeneous. Values in the range 55-90 mJ/m2 have been reported in the literature for homogeneous processes.16,17 A value as low as 11 mJ/m2 has been calculated for heterogeneous nucleation of vaterite on cholesterol.18

Figure 19. Calculation of the surface energy γs of the growing vaterite phase for T = 25, 35, and 50°C in seawater.

100

The dependence of the rates of calcium carbonate precipitation as a function of the solution supersaturation at various temperatures is shown in Figure 20. As may be seen, a parabolic dependence of the rates of precipitation as a function of the supersaturation was found, suggesting that the rate determining step is surface diffusion.19 The same mechanism of precipitation was found for all temperatures examined. The rates measured were reproducible within 10%. In the absence of Mg2+ at 80ºC the rates of precipitation were significantly higher over the same range of supersaturation values, as illustrated in Figure 2 1. This may be due to an inhibitory effect of magnesium ions on the rate of crystallization. A second-order polynomial fit was applied for the precipitation rate curves presented in Figure 20. The apparent activation energy computed from the respective rate constants at each temperature was found to be equal to 34.2 kJ/mol (Figure 22). This value corroborates the evidence from the value of the apparent order that the crystallization is controlled by the surface reaction rather than the diffusion from the bulk solution to the surface of the crystals.20 The rates of crystal growth showed a second-order dependence on the solution supersaturation (Figure 23) thus following the same pattern as the rates of precipitation at the initial stages of the process. This parabolic dependence suggested a spiral growth mechanism.l9 Compounds which block the active growth centers of vaterite possibly by adsorption may therefore cause inhibition to CaCO3 precipitation under the present conditions.

Seeded-Growth Experiments in Seawater Inhibited by a Polyacrylic Acid Polymer XRD spectra of the seed crystals before and after the overgrowth of calcium carbonate confirmed that calcite was the CaCO3 phase which formed on the seed crystals in all experiments. SEM micrographs showed that the inhibitor used did not cause any morphological changes to the calcite crystals grown on the seed crystals.

Figure 20. Effect of temperature on CaCO3 precipitation kinetics in seawater.

101

Figure 21. Effect of magnesium ions on CaCO3 precipitation rate in seawater.

The addition of the above polymer at concentrations lower than 0.5ppm or 2.5x10-7 mol/l caused significant retardation to the rates of calcium carbonate crystal growth. At polymer concentrations of 0.5 ppm the precipitation was completely suppressed. The rate of crystal growth Ro measured in the absence of additive (blank experiment) was equal to 4.5x10-4 mol min-1 m-2 (T=80°C, SRcalcite=8.39). Table 3 summarizes the relative inhibition caused over a range of concentrations of the inhibitor tested with respect to their effectiveness in the artificial seawater solutions at 80°C, pH=8.5 and at SRcalcite=8.39. The relative inhibition Ri/Ro is the ratio of the rate of crystal growth Ri measured for a certain concentration of the additive over the rate measured for the blank experiment and the

Figure 22. CaCO3 precipitation in seawater. Calculation of activation energy of precipitation.

102

Figure 23. Kinetics of crystal growth for CaCO3 precipitation in seawater.

percent relative inhibition is defined by Equation 12: Percent

Relative

Inhibition

=

Ro – Ri Ro

Xl00

(12)

The relative inhibition Ri/Ro as a function of the concentration of the polyacrylic polymer is shown in Figure 24. The relative inhibition decreases sharply for inhibitor concentrations from 0 to 5x10-8 mol/l (0. 1ppm), while for higher concentrations a tendency to reach a plateau was found. The inhibition was also found to depend on the solution supersaturation. Figure 25 shows the dependence of the rate of crystal growth of calcite on

Concentration of inhibitor, mol/l Figure 24. Dependence of the relative inhibition on additive concentration in seawater (T=80°C, SRcalcite=8.39, pH=8.5).

103

the supersaturation for an inhibitor concentration equal to 0.lppm at 80°C and pH=8.5. A linear increase of the rate of crystal growth with supersaturation for the same additive concentration was found. Table 3. Crystal growth of calcite on calcite seed crystals in seawater in the presence of a polyacrylic acid polymer at 80°C, pH=8.5 and SRcalcite=8.39. Relative inhibition vs. additive concentration. Concentration, ppm

Relative inhibition, Ri/Ro

0 0.02 0.05 0.1 0.5

1 0.53 0.56 0.054 0

Percent relative inhibition, ((Ro-Ri)/Ro)*100 0 47.1 44.4 94.6 1

Crystal growth in the absence of additives is generally believed to proceed through a repetitive screw dislocation mechanism which continuously creates active sites.19,21 As may be seen in Figure 25, for the same inhibitor concentration in the working solution the rate of crystal growth increased with increasing supersaturation value. This result suggested that the more active sites in the working solution, the higher the concentration of the additive used is needed to achieve the same inhibition. If it is also taken into account that the polymer used was effective at substoichiometric proportions with respect to the total calcium concentration in solution (TCa=2.8x10-3 — 3.4×10-3 mol/l), it may be assumed that the additives adsorb at the active sites through their ionized functional groups, thus restricting further growth of the crystals. The same mechanism has been reported for other inhibitors in the literature.22 Assuming that adsorption takes place according to the

SRcalcite Figure 25. Dependence of the rate of crystal growth on the seawater solution supersaturation in the presence of 0.1ppm of the inhibitor polymer; T=80°C, seawater without Mg2+, pH=8.5.

104

simplified Langmuir adsorption model,23 and that the rates of crystal growth depend on the surface coverage, the relative inhibition as a function of additive concentration may be expressed by Equation 13 : 1 Ro = ——— 1 + —– Ro–Ri bC

(13)

where C is the additive concentration and b the affinity constant for the additive and the crystal surface. For the polymer used in the present study an affinity coefficient equal to 8.29x107 mol/l has been calculated. In a study of CaCO3 crystallization inhibition at 25°C by maleic acid polyelectrolytes (MW in the range 15,000-145,000) values of 6.71x1081 .66x1010 mol/l for the affinity constant have been reported. This parameter is considered to be the effectiveness of the inhibitors and depends on the active groups, their number in the molecule of the inhibitor compound and the hydrophobic character of the molecule, i.e. factors related directly to the adsorbate-adsorbent interactions.

CONCLUSIONS The crystallization of calcium carbonate in seawater is controlled by surface diffusion mechanism in the temperature range 25-80°C. Vaterite is the first CaCO3 polymorph formed, which at later stages transforms into aragonite. The process is largely homogeneous, as indicated by a value of surface energy equal to 41.6 mJ/m2. At 80°C the formation of Mg(OH)2 competes with the calcium carbonate formation resulting in low induction times at supersaturation values with respect to calcite close to zero. At the same temperature and in the absence of magnesium ions the induction times are substantially higher, as well as the rates of precipitation. The presence of a polyacrylic acid polymer at concentrations lower than 0.5 ppm inhibited strongly the crystal growth of calcite from seawater supersaturated solutions. An affinity coefficient equal to 8.29x107 mol/l has been calculated and the inhibitory effect has been attributed to adsorption of the polymers at the active growth sites.

REFERENCES 1. 2. 3. 4. 5. 6. 7.

T. Haarberg, I. Seim, D.B. Granbakken, and T .Østvold, Scale formation in reservoir and production equipment during oil recovery: An equilibrium model, SPE Production Engineering , February 1992:75. D. Granbakken, T. Haarberg, M. Rollheim, T .Østvold, P. Read, and T. Schmidt, Scale formation in reservoir and production equipment during oil recovery. III. A kinetic model for the precipitation/dissolution reactions, Acta Chem. Scand. 45:892, 1991, M. Abtahi, B. Kaasa, J.E. Vindstad, and T .Østvold, Calcium carbonate precipitation and pH variations in oil field waters. A comparison between experimental data and model calculations, Acta Chem. Scand. 50: 114 (1996). J.W. Morse, J. de KaneI, and H.Z. Craig Jr, A literature review of seawater with respect to calcium carbonate and its possible significance for scale formation on OTEC heat exchangers, Ocean Engineering, 6:267 (1979). L.M. Cenegy, D.F. Griffith, and G.W.M. Hobbs, Method of treating geothermal wells with acrylate/acrylamide scale inhibitor, U.S.Pat. 5,044,439,3, Sept. 1991. L. Iosifescu, G. Chitanu, and A. Carpov, Inhibition of calcium carbonate scale from geothermal water using anionic polyelectrolytes, Hidrotechn. (Buch.) 29: 143 (1984). M.P.C. Weijnen and G.M. van Rosmalen, The influence of various polyelectrolytes on the precipitation of Gypsum, Desalination 54:239 (1985).

105

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

10 6

O.J. Vetter, An evaluation of scale inhibitors, J. Pet. Techn. 997 (1972). T. Sabbides, E. Giannimaras, and P.G. Koutsoukos, The precipitation of calcium carbonate in artificial seawater at sustained supersaturation, Environment Technology 13:73 (1992). T. Sabbides and P.G. Koutsoukos, The crystallization of calcium carbonate in artificial seawater;Role ofthe substrate, J. Crystal Growth 133:13 (1993). I.X. Malollari, P.G. Klepetsanis, and P.G. Koutsoukos, Precipitation of strontium sulfate in aqueous solutions at 25°C, J. Crystal Growth 155:240 (1996). The NALCO Water Handbook, 2nd Ed., F.N.Kemmer , McGraw Hill Co. , N.York (1988). R.A. Berner, The role of magnesium in the crystal growth of calcite and aragonite in seawater, Geochim. Cosmochim. Acta, 39:489 (1975). Folk, The natural history of crystalline calcium carbonate: Effect of magnesium content and salinity, Journal of sedimentary Petrology, 44(1):40 (1974). J.W. Mullin, Crystallization, 3d ed., Butterworth-Heinemann, Oxford (1993). O.Sohnel and J.Garside, Precipitation, Butterworth-Heinemann, Oxford OX2 8DP Koutsoukos PG and Kontoyannis CG, The Spontaneous Precipitation of Calcium Carbonate in Aqueous Solutions J.Chem.Soc.Farad Trans.I, 80: 11 (1984). E.Dalas and P.G.Koutsoukos, The Precipitation of Vaterite on cholesterol. J.Colloid Interface Science, 127:273 (1989). E.A. Burton and L.M. Walter, The role of pH in inhibition of calcite and aragonite precipitation rates in seawater, Geochim. Cosmochim. Acta 54:797 (1990). S.T. Liu and G.H. Nancollas, The Kinetics of Crystal Growth of Calcium Sulfate Dihydrate J. Crystal Growth, 6:281 (1970). A.G. Xyla, E.K. Giannimaras, and P.G. Koutsoukos, The precipitation of calcium carbonate in aqueous solutions, Colloids and Surfaces 53:24 1 (1991). P.G. Klepetsanis and P.G. Koutsoukos, The inhibition of calcium carbonate formation by copolymers containing maleic acid, Water Soluble Polymers, p.117 (1998). C.W. Davies and G.H. Nancollas, The precipitation of silver chloride from aqueous solutions, Part III. Temperature coefficients of growth and solutions, Trans. Faraday Soc., 5 1 :8 18 (1955).

CALCITE CRYSTAL GROWTH RATE INHIBITION BY AQUATIC HUMIC SUBSTANCES

M. M. Reddy1 and A. R. Hoch1,2 1 U.S.

Geological Survey 3215 Marine St. Boulder, CO 80303 Current address: 2 Geology Department Lawrence University Appleton, WI 54912-0599

ABSTRACT Calcium carbonate crystallization is an important process in numerous terrestrial and aquatic ecosystems. Moreover, calcium carbonate scale formation on heat transfer surfaces is an important industrial problem. Humic substances, which are ubiquitous in natural waters, often reduce or inhibit calcite crystal growth and may, therefore, have application in scale prevention. This study quantifies and interprets calcite crystal growth rate reduction due to humic substance fractions isolated from surface water. We conducted calcite growth experiments at constant pH (8.5), temperature (25°C), supersaturation (Σ = 4.5), PCO2 (l0-3.5atm), and ionic strength (0.1 M KNO3) with various concentrations of natural organic acid isolates. Plant-derived aquatic hydrophobic acids from the Everglades were effective growth inhibitors. Organic acid molecular weight correlated with growth inhibition as did aromaticity, aliphatic content and heteroatom content; however, carboxyl content did not. Growth inhibition over a range of organic inhibitor concentrations for Everglades’ samples is consistent with a Langmuir adsorption mechanism blocking growth sites on the calcite surface.

INTRODUCTION Crystal growth on heat transfer surfaces, leading to scale formation and reduced heat transfer efficiency, is an important industrial problem. For example, calcium carbonate scale formation occurs in cooling water technology, desalination processes, and oil

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers New York 2000

107

production. Moreover, an important aspect of carbonate mineral formation in current water management strategies in South Florida and elsewhere is surface water storage in unsaturated carbonate aquifers. In this situation, carbonate minerals formed after injection will alter aquifer porosity and permeability. Calcium carbonate minerals are ubiquitous in the environment. Fresh water lakes and ocean waters are fre uently supersaturated with respect to calcite with no observed mineral precipitation.1,2,3 Where inorganic carbonate precipitation is thermodynamically favored, natural kinetic inhibitors may prevent carbonate mineral nucleation and crystal growth.2,4,5 Calcite crystal growth mechanisms have been examined extensively,6,7 and it is generally agreed that calcite grows by a surface reaction mechanism.8,9 Calcium carbonate growth is inhibited by common inor anic ions, such as magnesium ion,1,10 phosphate ion,9,11 and simple organic compounds.2,12,13,14,15 Nancollas and others16 discuss crystal growth inhibition mechanisms and show that slightly different compounds may have different effects on calcite growth rates depending on how these compounds interact with growth sites on calcite surfaces. Compounds that bind strongly to active growth sites are efficient growth inhibitors at low concentrations. Several classes of chemical additives, known as "threshold inhibitors", have been found to reduce or eliminate scale by inhibitin calcium carbonate formation at very low solution additive concentrations. Nancollas17 summarizing recognized aspects of threshold inhibition, notes that effective inhibitors are usually also good sequestering agents for lattice cations. However, not all good sequestering ligands are good threshold inhibitors because the inhibition mechanism is an adsorption process at growth sites on the crystal surface. Adsorption affinity is favored by low solubility of salts involving the lattice cations and the inhibiting anion. Model compound studies examining the molecular relationship to growth inhibition have yielded interesting results in precipitation reactions. Miura et al.18 reported that phosphates adsorbed to calcium carbonate, but pyrophosphate and tri hosphate were better inhibitors than orthophosphate and trimetaphosphate. Berner et al,14 concluded that aragonite growth can be inhibited by the presence of benzenecarboxyl groups. Amjad19 examined calcite growth inhibition in the presence of benzenepolycarboxylic acid and reported that more highly carboxylated benzenes are better inhibitors (at the same solution concentration) than less carboxylated benzenes. Strong calcite growth rate inhibitors are effective at concentrations less than one milligram per liter. Important functional groups for crystal growth rate inhibition contain phosphorus, nitrogen or carboxylate groups.11,15,16,17 However, threshold inhibitors may pose water quality problems because they contribute to nutrient loading. Perez and Zidovec20 stress environmental advantages of using nonphosphorus containing compounds in scale inhibition applications. Recent publications have shown natural organic acids markedly reduce calcium carbonate crystal growth rates.21,22,23 Humic substances, including fulvic acid, humic acid and humin, are biogenic, refractory organic matter ubiquitous in terrestrial and a uatic environment.24 Some of these substances are effective crystal growth inhibitors.14,15,21,22,25 In the past, recognition of humic substances as growth inhibitors has focused on comparing inhibition effects of model compounds with one or two poorly characterized or uncharacterized humic substance sample15,26 or comparing the effects of a few unrelated and less wellcharacterized soil humic extracts.25 The chemistry of aquatic humic substances may vary with regard to molecular weight, functional group content and acidity. It is the goal of this study to examine the effectiveness of well-characterized dissolved organic material (DOM) isolates from the Florida Everglades and Lake Fryxell, Antarctica, in inhibiting

108

calcite growth. An additional goal is to identify those chemical and structural features of these natural materials that may have application as threshold inhibitors.

MATERIALS AND METHODS Experimental details appear elsewhere27 and are briefly summarized here. We utilized a constant composition experimental apparatus28 in which solution pH and chemical composition were kept constant to fix calcite supersaturation (Fig. 1). Experiments were done at constant pH of 8.5, tem erature (25°C), supersaturation (Ω) of 4.5, solution carbon dioxide partial pressure of l0-3.5 atmosphere and ionic strength of 0.1 M. Solution supersaturation is regulated by the potentiometrically-controlled addition of solutions containing calcium and carbonate ions. This approach enables measurement of calcite growth rate inhibition with high precision and avoids complications due to changing solution supersaturation. The calcite growth reaction at a pH of 8.5 can be written as: Ca2+ + HCO3-→CaCO3 solid + H+

(1)

Figure 1. Diagram of constant composition experimental apparatus.

During calcite growth a pH-stat apparatus senses decreasing pH and causes the double burette system (Fig. 1) to respond by adding calcium chloride and sodium carbonate titrant solutions replacing calcium and carbonate ions lost to crystal growth, keeping pH and solution composition constant. Titrant solutions also contained electrolyte (potassium nitrate) to maintain constant ionic strength (Table 1). Constant composition was verified

109

by measuring total dissolved calcium ion concentration and alkalinity before and after all experiments. Solution pH and volume of titrant added over time were recorded on a dualpen chart recorder. All experiments were run for 100 minutes.

Table 1. — Inorganic Compositions of solutions used in all experiments Reagent CaC12 • 2H2O NaHCO3 Na2CO3 KNO3

A.calcium stock solution 0.004

Working Solutions (M) B.carbonate A+B reactor stock working solution solution 0.002 0.004 0.002 0.185

0.093

A.calcium titrant solution 0.053 0.038

Titrant Solutions (M) B.carbonate A+B net titrant composition of solution titrant added 0.026 0.004 0.002 0.049 0.024 0.046 0.042

Metastable working solutions were prepared by adding calcium ion containing solutions dropwise to carbonate solution in the thermostatted growth reaction cell. The pH was then adjusted to 8.5 using a few drops of 0.05 M KOH and metastability verified by monitoring constant pH for an hour prior to the experiments. Total working solution volumes were 400 milliliters and 100 milligrams of dry calcite seed were added to metastable solutions to initiate growth. Calcite seed was analyzed using a Siemens X-ray diffractometer to verify mineral purity. The specific surface area of the seed (SAseed) material was 0.256 ± 0.008 m2/g (determined by BET gas absorption measurement.27 Stock solutions containing organic acids were prepared in base working solutions. Organic acid containing solutions were refrigerated in dark glass bottles. Prior to each experiment, stock solutions were added to base working solutions to obtain the desired organic acid concentrations and then filtered through 0.1 µm Millipore cellulose nitrate filters. Organic acids used as growth inhibitors in this study were the hydrophobic organic acid fraction of DOM from three Florida Everglades’ sites, F1, U3 and 2BS.29,30 A fulvic acid sample from Lake Fryxell, Antarctica, was also used in the calcite growth inhibition experiments.31 The hydrophobic organic acids are 96% to 97% fulvic acid with the balance being humic acid. Weight averaged (Mw) and number averaged (Mn) molecular weights were determined by high pressure size exclusion chromatography (HPSEC), carboxyl acidity was determined by titration with a base to an endpoint of pH = 8.5; functional group abundances and aromaticity were estimated from 13C-NMR spectra.30,32 The slope of the line corresponding to titrant addition due to calcite growth (reciprocal minutes) is proportional to the calcite growth rate and is converted to crystal growth rates by the following equation: Rate ((mol/m2)/min) = slope (1/min) * Mtitrant (mol/l)/(massseed(g) * SAseed(m2/g))

(2)

Most experiments yielded straight lines for titrant addition over time except for experiments with higher organic acid isolate concentrations. These experiments initially exhibited higher rates, but quickly (< 5 min) slowed to a steady rate of growth as the organic acid isolate reached adsorption equilibrium with the calcite surface. For these experiments, rates were calculated using the slope starting at 10 minutes. Calcite growth rate inhibition by organic acids is expressed as reduced rates R/Ro for each experiment: R is the rate in the presence of inhibitor and Ro is the rate in the absence of inhibitor. Thus, a smaller reduced rate indicates greater growth inhibition.

110

RESULTS Dissolved calcium ion speciation calculations were performed for experiments in the presence and absence of organic acids. There are uncertainties concerning DOM complexation reactions with calcium ion in solution. Here we assume (as an upper limit) that all carboxyl groups of DOM form one-to-one complexes with calcium ion in solution. In this situation, complexes of calcium were small compared with the free calcium ion concentration. Ion pair formation did not significantly reduce calcite supersaturation. For each organic acid isolate studied a series of experiments were performed at concentrations of 0.2 to 5.0 mg/L DOM. Examples of calcite growth rate data for experiments using hydrophobic organic acids from site F1 are shown in Figure 2. For the organic acid isolate from site F1, measurable rate reduction was observed at the lowest experimental organic acid isolate concentrations and calcite growth was almost completely inhibited at concentrations of 5.0 mg/L.

Time (min) Figure 2. Crystal growth of calcite at constant supersaturation in the presence of varying concentrations of F1 hydrophobic organic acid (slope is proportional to rate).

The marked effect of organic acid isolates in reducing the calcite growth rate is illustrated in Figure 3 where reduced growth rate is plotted versus DOM concentration (expressed in units of milligrams of carbon per liter). Different organic acid isolates had different effects on calcite crystal growth rates. Hydrophobic acid from the F1 site was the strongest inhibitor at all concentrations, followed by U3 hydrophobic acid and 2BS hydrophobic acid. The isolate from site F1 reduced calcite growth rates by more than 50% at concentrations as low as 0.5 milligrams per litter. Fulvic acid from Lake Fryxell was the weakest growth inhibitor of all compounds studied. Reacted and unreacted crystals were examined. by scanning electron microscopy to visualize morphological changes that occurred during growth with and without added organic acids (Fig. 4). Unreacted crystals (Fig. 4A) were well-formed sharp-edged rhombohedra. Crystals that grew for 100 minutes in the absence of organic inhibitors

111

(Fig. 4B) increased their mass by about 25% and exhibited smooth planes of growth with step features on the face perimeters. Crystal morphology produced by growth in the presence of organic acid isolates is different than in the control experiments. For example, in the 0.5 mgL F1 isolate experiments the growth rate was reduced by more

Figure 3. Reduced calcite growth rates (R/Ro) at constant supersaturation at varying concentrations of DOM (milligrams of carbon per liter) isolates from Florida Everglades sites A) F1, B) U3 and C) 2BS and from Lake Fryxell, Antarctica, D) LF.

than 50% compared with control experiments; crystal mass increased by only 12% and crystal surfaces did not exhibit smooth planes of growth but planes in which growth had been interrupted, resulting in a broken or discontinuous appearance (Fig. 4C). Crystals reacted at higher organic acid isolate concentrations showed even smaller “islands” of growth or no growth at all.

112

Figure 4. SEM photomicrographs of calcite seed material before and after crystal growth experiments. A) unreacted calcite seed; B) Calcite seed after 100 minutes growth in an experiment with no added organic materials; C) Calcite seed after growing for 100 minutes in a solution containing 0.5 mg/L F1 hydrophobic acid; added mass was about 50% of that observed in control experiment (B).

113

DISCUSS ION Chin et al.33 observed that there is a direct relationship between DOM molecular weight and degree of aromaticity in aquatic fulvic acids. Weight average molecular weight (Mw) and number average molecular weight (Mn) for the Everglades’ organic acids vary by as much as 65% for the organic acid isolates. Perez and Zidovec20 reported that calcium carbonate growth inhibition in the presence of low-molecular weight polyacrylic acids (PAA) (2000 to 5000 daltons) is proportional to molecular weight. Although the molecular weights of the organic acids used here are slightly below this range (1100 to 1900 weight average molecular weight), higher molecular weight Everglades’ organicacid isolates produced greater rate reduction than the lighter organic acids at all concentrations (Fig. 5). These results (Fig. 5) support the influence of organic acid molecular weight in

Figure 5. Plots of all experimentally-determined reduced rates R/Ro versus: A) number-average molecular weights and B) weight-average molecular weight for DOM isolates used in experiments. Each symbol type represents a given concentration of organic acid. Sets of symbols at given molecular weights represent suite of experiments run in the presence of a given organic acid, noted on the figure.

114

the inhibition mechanism. However, Everglades’ organic acid isolates have greater functional group heterogeneity (Table 2) than the PAAs of Perez and Zidovec,20 and this may confound the molecular weight/inhibition relationship. At all concentrations less than 5 mg/L DOM the rate reduction for Lake Fryxell acid is approximately equal to that from site 2BS. This is an exception to the influence of molecular weight. It is noteworthy that the organic acid that is least effective as a growth rate inhibitor (Lake Fryxell, Table 2) has the lowest molecular weight. Table 2. Acidity and molecular weight data for organic acids Sample/date collected

Description

MOLECULAR WEIGHT* (Daltons) Mn Mw/Mn Mw

Everglades U3, 3/95

Hydrophobic Fraction

1747

1180

Everglades F 1,4196

Hydrophobic Fraction

1907

1238

ACIDITY ** COOH (meq/g)

OH (meq/g)

1.48

5.99

0.735

1.54

5.14

1.35

1.41 4.74 1.03 Everglades Hydrophobic 1519 1080 2BS, 4196 fraction Lake Fryxell Fulvic acid 1080 713 1.51 5.26 1.06 - 5.5m *Everglades’ high pressure size exclusion chromatography (HPSEC) molecular weight data from Aiken and Reddy (1997); Lake Fryxell HPSEC data is from Chin et al.33 ** All acidity data are from Ravichandran.30

Humic substances vary in chemical functionality and molecular structure.24 We have used 13C NMR (Table 3) to identify bonding environments of the 13C isotope in Everglades’ organic acid isolates.34 The 13C NMR data is used to estimate average carbon atoms in given bonding environments as follows: Ci (daltons) = M W. (daltons) *fC *fI

(3)

where Ci is the average total mass of carbon in functional group i in an organic acid isolate, M W. is the weight-averaged or number-averaged molecular weight of the organic acid isolate, fc is the fraction of carbon in the organic acid isolate (Table 4), and fi is the fraction of carbon in that functional group i (Table 3). Aliphatic I carbons (carbon-carbon single bond) are the most abundant functional groups (Table 3). Figure 6A shows the relationship between rate reduction and aliphatic I carbon content. Zullig and Morse36 found that fatty acids with longer carbon chains (which would be represented by aliphatic I carbon signal) adsorb to calcite surfaces. The aliphatic II region signal represents C-O, C-N and C-S heteroatom single bonds; most aliphatic II content is C-O bonds (Table 4). Aliphatic II moieties contribute to the hydrophilic nature of the organic acid isolates and are expected to preferentially bind to calcite surfaces. This affinity enhances organic acid crystal growth inhibition. Greater aliphatic II carbon content coincides with greater rate reduction (Fig. 6B). In summary, data presented in Figure 6 show that the only moiety which does not correlate with rate reduction is carboxyl (6D). All others exhibit rate reduction with increasing average

115

Table 3. Functional character of organic acids as determined by SampIe/date collected

Description

Everglades U3, 3/95;

Hydrophobic fraction

Everglades F1,4/96* Everglades 2BS, 4/96;

13CNMR — percent

13CNMR

analysis

C in functional groups (ppm chemical shift)

Aliphatic I(0-62) 42.3

Aliphatic II (62-90)

Acetal (90-110)

Aromatic (110-169)

Carboxyl (160-190)

Ketone (190-230)

15.0

1.4

19.0

14.4

4.1

Hydrophobic fraction

45.0

17.0

5.3

18.2

11.6

2.7

Hydrophobic fraction

48.5

14.9

4.1

15.4

15.4

1.8

4.4

15.2

19.6

0.0

Lake Fryxell Fulvic acid 46.4 14.5 - 5.5m * Everglades NMR data from Aiken and Reddy.33 ** Lake Fryxell NMR data from Aiken et al.31

Table 4. Chemical composition of organic acids as weight percent constituent; standard deviations are given in parentheses Sample

C (± 0.5)

H (± 0.05)

0 (± 0.6)

N (± 0.04)

S (± 0.05)

Ash (± 0.2)

F1*

52.2

4.64

39.9

1.53

1.73

5.8

U3*

54.7

4.79

37.5

1.88

1.15

4.6

2BS*

52.3

4.19

40.2

1.58

1.23

7.3

LF-5.5m**

54.9

5.5

34.9

3.3

1.2

2.3

* Everglades’ data from Aiken and Reddy.35 ** Lake Fryxell data from Aiken et al.31

116

Average carbon type/molecule (daltons) Average carbon type/molecule (daltons)

Average carbon type/molecule (daltons) Average carbon type/molecule (daltons) Figure 6. Plots of all experimentally-determined reduced rates R/Ro versus functional group: A) aliphatic I carbon (C-C single bond); B) aliphatic II carbon (primarily C-O single bond); C) aromatic carbon; and D) carboxyl carbon.

moiety carbon mass/molecule. The patterns of plots 6A, 6B and 6C are remarkably similar suggesting that aromaticity may not be central in determining DOM inhibition effectiveness. The most efficient growth inhibitors (F1 and U3) contained more aliphatic and aromatic carbon/total carbon than the weaker inhibitors (Fig. 6). Reynolds2 reported calcite precipitation rates decreased with increasing aromatic humic substance (polyphenol) concentrations in Lake Powell waters. Berner et al.14 studied the effects of four sets of model compounds (amino acids/proteins, fatty acids, aliphatic/polycarboxylic acids and aromatic acids) on aragonite precipitation. Of these four classes of compounds, only the aromatic acids were found to be strong inhibitors. Stereochemical reasons were invoked for the strong interaction of the more aromatic acids with the active growth sites on carbonate surfaces. Scanning Force Microscopy imaging demonstrates that calcite growth on seed crystals at moderate supersaturation is initiated by multiple surface nucleation events, followed by spiral growth.9 In inhibitor-free solutions, smaller growth spirals coalesce forming smooth mineral surfaces that grow in straight steps. In solutions containing 117

solutes that sorb to mineral surfaces, the smooth steps are blocked, producing a jagged, discontinuous appearance as shown in Figure 4C. If growth inhibition is caused by DOM adsorption at active growth sites on crystal surfaces, an adsorption isotherm should describe the rate reduction with changing inhibitor concentration. Here we consider Langmuir-type adsorption37 at active growth sites on the seed surfaces to cause growth inhibition as found for calcite growth rate inhibition by other ions.10,11,15 If we assume α is the fraction of sites blocked by organic inhibitors, the rate in the absence of inhibitor, Ro, is reduced to a slower rate, R, according to the relationship: R = Ro (1-α)

(4)

which may be rewritten as: α=

(R o -R)/R o

(5)

In a given system, DOM adsorption and desorption rates at the growth sites may be expressed as follows: adsorption rate = kads [DOM] (1-α)

(6)

desorption rate = kdes α

(7)

and, at DOM adsorption equilibrium: kads [DOM] (1-α )= kdes

α

(8)

Solving equation 8 for α, and substituting into equation 5 we obtain: R o /(R o -R) = (k ads /k des (1/[DOM]) + 1

(9)

which is the equation for a line with the slope defined by the adsorption/desorption rates and an intercept of 1. Using our reduced rate notation, a plot of 1/(1-R/Ro) versus the reciprocal of the organic acid isolate concentration reveals excellent linear relationships for experiments utilizing hydrophobic acids isolated from Everglades’ DOM (Fig. 7A-C). These data plots are consistent with adsorption of organic acid isolates at active growth sites; plots of the Everglades’ isolates are linear with an intercept near unity. Agreement of transformed experimental calcite growth rate data with that expected for the Langmuir adsorption model supports a mechanism of growth rate inhibition through attachment of inhibitor ions at growth sites on the crystal surface. Linearity of best fit lines is consistent with growth inhibition as a result of monolayer adsorption at active growth sites. Everglades’ DOM best-fit correlation coefficients are significant at the 99% confidence level. The Langmuir plot for Lake Fryxell DOM (Fig. 7D) is not significant at the 90% confidence level. Absence of strong inhibitors in the Lake Fryxell DOM sample may explain the larger experimental scatter in the dataset.

CONCLUSIONS Aquatic hydrophobic acids isolated from Florida Everglades’ DOM are effective calcite growth rate inhibitors. The most effective inhibitors reduced calcite growth rates by more than 50% at concentrations as low as 0.5 milligrams per liter. Molecular weight

118

1/[DOC]

1/[DOC]

Figure 7. Langmuir isotherm type plots for rate reduction in the presence of organic inhibitors at varying concentrations of DOM isolated from Florida Everglades sites A) F 1, B) U3 and C) 2BS and from Lake Fryxell, Antarctica, D) LF.

of Everglades’ DOM isolates correlated with growth inhibition. Carboxyl carbon content is less important in identifying an organic acid isolate as an effective inhibitor. The morphology of crystals grown in the presence of organic acids suggests that surface growth sites are blocked by adsorbing organic acid isolates, as demonstrated in other studies using atomic force microscopy. Calcite growth inhibition by Everglades’ hydrophobic organic acid isolates, over a range of concentrations, is consistent with a mechanism of Langmuir adsorption of organic ligands to calcite surfaces, blocking growth sites. Our results quantitatively show the significance and variability of the inhibitory effects of natural hydrophobic organic acids when considering kinetics of calcite growth and its control on natural water chemistry. In natural systems DOM concentrations may be an order of magnitude greater than those examined in this study, indicating the importance of time and/or localized areas of increased pH (such as in algal mats) for achieving significant inorganic calcite mineralization.

ACKNOWLEDGEMENTS This paper benefited from discussions with George Aiken, Greg Brown, Jerry Leenheer, and M. Ravichandran regarding the chemical character of humic substances. Review comments by Z. Amjad, R. Antweiler and G. Brown helped improve the manuscript. A. R. Hoch acknowledges the National Research Council associateship program and the USGS, Water Resources Division, National Research Program for 119

financial support of his work at the U.S. Geological Survey. Humic isolates supplied by George Aiken and co-workers is gratefully acknowledged. Assistance in revision of several figures by Micaela Beth Reddy is appreciated. Pat La Tour and Charmaine Gunther provided invaluable assistance in the preparation of the camera-ready manuscript. The use of trade names in this report is for identification purposes only and does not constitute endorsement by the USGS.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

120

R.A. Berner, The role of magnesium in the crystal growth of calcite and aragonite from sea water, Geochim. Cosmochim.Acta. 39:489(1975). R.C. Reynolds Jr., Polyphenol inhib ition of calcit e precipitation in Lake Powell, Limnology and Oceanography. 23:585( 1978). M.M. Reddy, Carbonate precipitation and Pyramid Lake, Nevada: Probable control by magnesium ion, in: Mineral Scale Formation and Inhibition, Z. Amjad, ed., Plenum Press, New York City( 1995). K. Chave and E. Seuss, Calcium carbonate supersaturation in seawater: effects of dissolved organic matter, Limnology and Oceanography, 15:633( 1970). D.L. Suarez, Calcite supersaturation and precipitation kinetics in the lower Colorado River, AllAmerican Canal and East Highline Canal: Water Resour.Res. 19:653( 1983). J.W. Morse, The kinetics of calcium carbonate dissolution and precipitation, in: Review in Mineralogy I I, Mineralogical Society of America, R.J. Reeder, ed., 227( 1983). R. Shiraki and S.L. Brantley, Kinetics of near-equilibrium calcite precipitation at 100°C: An evaluation of elementary reaction-based and affinity-based rate laws, Geochim.Cosmochim.Acta, 59: 1457( 1995). A.J. Gratz, P.E. Hillner, and P.K.Hansma, The dynamics of the growth of calcite crystals, Geochim.Cosmochim.Acta, 57:491(1993). P.M. Dove and M.F. Hochella Jr., Calcite precipitation mechanisms and inhibition by orthophosphate: In situ observations by Scanning Force Microscopy, Geochim.Cosmochim.Acta, 57:705( 1993). M.M. Reddy, Effect of magnesium ion on calcium carbonate nucleation and crystal growth in dilute aqueous solutions at 25° Celsius, in: Studies in Diageneses, F.A. Mumpton, ed., U.S. Geological Survey Bulletin 1578, Denver, Colorado, 169( 1985). M.M. Reddy, Crystallization of calcium carbonate in the presence of trace concentrations of phosphorus-containing anions I. Inhibition by phosphate and glycerophosphate ions at pH 8.8 and 25°C, Journal of Crystal Growth, 41:287(1977). Y. Kitano and D.W. Hood, The influence of organic material on the polymorphic crystallization of calcium carbonate, Geochim. Cosmochim. Acta, 29:29( 1965). A. Otsuki and R.G. Wetzel, Interaction of yellow organic acids with calcium carbonate in freshwater, Limnol.Oceanogr. 18:490(1973). R.A. Berner, J.T. Westrich, R. Graber, J. Smith, and C.S. Martens, Inhibition of aragonite precipitation from supersaturated seawater: a laboratory study and field study, Amer.J.of Sci. 278:816(1978). M.M. Reddy, Kinetic inhibition of calcium carbonate formation by wastewater constituents, in The Chemistry of Wastewater Technology, A.J. Rubin, ed., Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan(1978). G.H. Nancollas, K. Sawada, and E. Schuttringer, Mineralization reactions involving calcium carbonates and phosphates, in Biomineralization and Biological metal Accumulation, P. Westbroek and E.W. de Jong, eds., D. Reidel Publishing Company, (1996). G.H. Nancollas, The mechanism of growth and dissolution of sparingly soluble salts, Ceramic Transactions, 1A:8(1998). M. Miura, H. Naonó, S. and Otani, Effects of various condensed phosphates on the crystallization and crystal habit of calcium carbonates, Kogyo Kagaku Zasshi, 66:597(1963). Z. Amjad, Kinetic study of the seeded growth of calcium carbonate in the presence of benzenepolycarboxylic acids, Langmuir, 3:224( 1987). L.A. Perez and D.F. Zidovec, Scale control by using a new non-phosphorous, environmentally friendly scale inhibitor, in: Mineral Scale Formation and Inhibition, Z. Amjad, ed., Plenum, New York City( 1995).

21. Z. Amjad and M.M. Reddy, Influence of humic compounds on the crystal growth of hydroxyapatite, in: Water Soluble Polymers: Solution Properties and Applications, Z. Amjad, ed., Plenum, New York City(1998). 22. Z. Amjad, J. Pugh, and M.M. Reddy, Kinetic inhibition of calcium carbonate crystal growth in the presence of natural and synthetic organic inhibitors, in: Water Soluble Polymers: Solution Properties and Applications, Z. Amjad, ed., Plenum, New York City(1998). 23. I. Lebron and D.L. Suarez, Calcite nucleation and precipitation kinetics as affected by dissolved organic matter at 25°C and pH > 7.5, Geochim.Cosmochim.Acta, 60:2765(1996). 24. G.R. Aiken, D.M. McKnight, R.L Wershaw and P. McCarthy, An introduction to humic substances in soil, sediment and water, in: Humic Substances in Soil, Sediment and Water: Geochemistry, Isolation and Characterization, G.R. Aiken, D.M. McKnight, and R.L Wershaw, eds. Wiley Interscience( 1985). 25. W.P. Inskeep and P.R. Bloom, Kinetics of calcite precipitation in the presence of water soluble organic ligands, Soil Sci.Soc.Amer.J. 50: 1167(1986). 26. P.R. Grossl and W.P. Inskeep, Kinetics of octacalcium phosphate crystal growth in the presence of organic acids, Geochim.Cosmochim.Acta, 56: 1955(1992). 27. A.R. Hoch, M.M. Reddy, and G.R. Aiken, Inhibition of calcite growth by natural organic material from the Florida Everglades at pH - 8.5 and 25°C, Geochimica et Cosmochimica Acta, In Press( 1999). 28. M.B. Tomson and G.H. Nancollas, Mineralization kinetics: a constant composition approach, Sci. 200: 1059(1978). 29. M. Ravichandran, G.R. Aiken, M.M. Reddy, and J.N. Ryan, Enhanced dissolution of cinnabar (mercuric sulfide) by dissolved organic matter isolated from the Florida Everglades, Environ.Sci. Technol. 32:3305(1998). 30. M. Ravichandran, Role of speciation and complexation of dissolved organic matter with mercury in the Florida Everglades. Ph.D. Thesis, University of Colorado at Boulder( 1999). 31. G.R. Aiken, D.M. McKnight, K.A. Thorn, and E.M. Thurman, Geochemistry of aquatic humic substances in the Lake Fryxell basin, Antarctica, Biogeochem, 34: 157(1996). 32. M. Ravichandran, G.R. Aiken, J.N. Ryan, and M.M. Reddy, Inhibition of precipitation and aggregation of metacinnabar (mercuric sulfide) by dissolved organic matter isolated from the Florida Everglades, Environ.Sci. Technol. 33: 14 18( 1999). 33. Y.-P. Chin, G.R. Aiken, and E. O'Loughlin, Molecular weight, polydispersity and spectroscopic properties of aquatic humic substances, Environmental Science & Technology, 28: 1853(1994). 34. K.A. Thorn, D.W. Folan, and P. MacCarthy, Characterization of the international humic substances society standard and reference fulvic and humic acids by solution state carbon-13 (13C) and hydrogen- 1 (1H) nuclear magnetic resonance spectrometry, Water-Resources Inv.Rpt. 894196: 1( 1989). 35. G.R. Aiken, and M.M. Reddy, Dissolved organic carbon in the Everglades, Florida, US. Geological Survey Open-File Report, 97-385: 1(1997). 36. J.J. Zullig and J.W. Morse, Interactions of organic acids with carbonate mineral surfaces in seawater and related solutions, Geochim. Cosmochim.Acta, 52: 1667( 1988). 37. W. Stumm, W. Chemistry of the Solid- Water Interface: Processes at the Mineral- Water and ParticleWater Interface in Natural Systems, Wiley-Interscience, New York City( 1992).

121

This page intentionally left blank.

THE INHIBITION OF CALCIUM CARBONATE FORMATION IN AQUEOUS SUPERSATURATED SOLUTIONS, SPONTANEOUS PRECIPITATION AND SEEDED CRYSTAL GROWTH

Pavlos G. Klepetsanis, 1,3 Angeliki Kladi, 1 Terje Ostvold,4 Christos G. Kontoyiannis, 1,3 Petros G. Koutsoukos, 1,2 Zahid Amjad, 5 and Michael M. Reddy 6 1Institute

of Chemical Engineering and High Temperature Chemical Processes, P.O.Box 1414 Patras, GR-26500, GREECE 2Department of Chemical Engineering 3Department of Pharmacy University of Patras Patras, GR-26500, GREECE 4Institute of Inorganic Chemistry The Norwegian Institute of Technology N-7034 Trondheim, NORWAY 5The BFGoodrich Company 99 11 Brecksville Road Brecksville, Ohio 44141, USA 6U.S. Geological Survey 32 15 Marine Street Boulder, Colorado 80303, USA

ABSTRACT The influence of the presence of humic, fulvic, and polyacrylic acid on the nucleation and crystal growth of calcium carbonate in aqueous supersaturated solutions was investigated in batch reactors at 25°C and pH = 8.50. The nucleation of calcium carbonate was investigated by free drift methods, and the crystal growth was investigated with seeded crystal growth experiments at constant supersaturation. In all cases calcite was found to form exclusively, and the presence of all tested compounds at

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers. New York, 2000

123

concentrations between 0.1 – 1.0 ppm prolonged the induction time preceding the spontaneous formation of calcite. Humic acid at concentration up to 0.5 ppm inhibited the growth of calcite seeds up to 95% and polyacrylic acid at concentrations up to 0.1 ppm gave the same degree of inhibition. Polyacrylic acid was found to be stronger inhibitor. A concentration of 0.25 ppm of polyacrylic acid completely stopped crystal growth of calcite. Humic acid at concentration 1 .0 ppm completely stopped crystal growth of calcite seed crystals. The retardation was explained by the adsorption of the polyelectrolytes onto the active growth sites of the crystals. Application of a Langmuir-type adsorption model on the kinetics data obtained in the presence of the inhibitors tested yielded a higher affinity constant of polyacrylic acid for the calcite seed crystals.

INTRODUCTION The formation of deposits of alkaline earth insoluble salts is a serious problem in installations where untreated natural waters are used, such as in geothermal energy exploitation, in secondary oil production by waterflooding, cooling towers, production of tap water by reverse osmosis etc.1,2 The formation of these deposits reduces heat transfer and the internal diameter of pipes, increases the operating pressure of pumps, and enhances the probability of corrosion damage. In many cases, the removal of deposits leads to discontinuous operation of the installations, resulting in higher operation costs. Calcium carbonate is one of the most commonly encountered scale deposits and occurs in different crystalline forms : calcite, aragonite, vaterite, calcium carbonate monohydrate and calcium carbonate hexahydrate. Calcite, the thermodynamically most stable polymorph of calcium carbonate, forms armoring, hard mineral deposits. Precipitation and stabilization of calcium carbonate polymorphs depend on solution conditions, i.e. level of supersaturation, pH, temperature, pressure, and the concentration and chemical structure of additives.1,3 Numerous methods have been applied to inhibit calcium carbonate formation.4 The most promising approach is addition of water soluble polyelectrolytes.5-8 Watersoluble compounds (polyelectrolytes, phosphonates) prevent calcium carbonate formation even at very low concentrations (< 1 ppm) and are cost effective scale inhibitors. In several cases, the presence of polyelectrolyte and phosphonate crystal growth inhibitors may cause significant modifications of the crystal habit, reducing their ability to adhere on the surfaces. The main disadvantage of polyelectrolytes (polyacrylates, polymaleates and their copolymers) and phosphonates is their low biodegradability. As a result these compounds contribute to environmental pollution. Phosphonates decompose at high temperatures and release phosphate ions which may cause the formation of calcium phosphates scale. In the last decade there has been a need for “green inhibitors” (i.e. environmentally friendly) to control mineral scale formation. Humic and fulvic acids, commonly found in the natural environment, are polymeric molecules with molecular weights ranging fiom a few hundreds to several thousands. These acids predominantly contain phenolic and carboxylic acid functional groups and behave as negatively charged colloids or anionic polyelectrolytes in natural waters.9 Fulvic and humic acids inhibit hydroxyapatite (HAP) and calcium phosphate dihydrate (DCPD) formation.10,11 Presented here are the effects of two natural polyelectrolytes, humic and fulvic acids (Suwannee River) and one synthetic polyelectrolyte, polyacrylic acid, on the precipitation of calcium carbonate from labile and metastable supersaturated aqueous solutions.

124

EXPERIMENTAL Two types of experiments are reported in the present investigation: a) spontaneous precipitation experiments at variable supersaturation conditions, and b) seeded growth experiments at sustained supersaturation conditions. All experiments were done in a batch-type, magnetically-stirred glass reactor, thermostatted by water at constant temperature circulated through a water jacket. The temperature in all experiments was kept at 25.0 ± 0.1°C. The stock solutions were prepared from crystalline, reagent grade chemicals (Merck, pro analisi) dissolved in triply distilled, CO2-free water and were filtered through membrane filters (0.2 µm, Millipore). Calcium chloride and sodium chloride stock solutions were standardized by atomic absorption spectroscopy (Perkin Elmer, AAnalyst 300). The sodium bicarbonate and sodium carbonate solutions were prepared fresh for each experiment. Solids were dried at 105°C overnight followed by dissolution in triply distilled water. The fulvic acid and humic acid stock solutions were prepared by exact weighing the amounts of the respective solids. Polyacrylic acid stock solution was prepared with dilution of appropriate volume of its concentrated solution. The molecular weight of the polyacrylic acid used for the experiments was 2100 (the use of trade names in this report is for identification purposes only and does not constitute endorsement by the USGS). The calcite seed crystals were prepared by slow mixing of calcium chloride and sodium carbonate solutions at 70°C.12 The crystalline solid was aged for one week under continuous stirring, filtered, washed with saturated calcium carbonate solution, and dried. The final solid was characterized by physicochemical methods including powder x-ray diffraction (XRD, Philips PW 1840), scanning electron microscopy (SEM, JEOL JSM 5200) and by specific surface area measurements (GEMINI, Micromeritics). The powder x-ray diffraction pattern of the crystalline calcite preparation coincided with that of the respective reference material.13 The specific surface area of the seed crystals was determined by a multiple point method BET and was found equal to 0.30 m2/g. The total volume of supersaturated solutions was 500 ml. The supersaturated working solutions were prepared in the reaction vessel by careful and rapid mixing equal volumes (250 ml) of calcium chloride and sodium bicarbonate solutions under continuous stirring. In all experiments the total calcium was equal to the total carbonate concentration. The ionic strength of the supersaturated solutions was adjusted by the addition of appropriate volume of concentrated sodium chloride stock solution. For the crystal growth experiments in the presence of fulvic and humic acids, the appropriate volume of respective additive stock solution was added in the working solution. The solution pH was adjusted to 8.50 by the slow addition of 0.10 M standard sodium hydroxide solution (Merck, Titrisol). The solution pH was measured by a combination glass/Ag-AgC1 pair of electrodes (Ingold), standardized before and after each experiment with NBS buffer solutions (7.413 and 9.180 at 25°C).14 The working solutions were stirred by a magnetic stirrer with a Teflon coated stirring bar. A schematic representation of the constant composition experimental setup is shown in Figure 1. In all experiments the precipitation of calcium carbonate resulted in the decrease of the solution pH due to proton release concomitant with the solid formation according to the reaction: Ca2+ + xH2CO-3 + yHCO3 + zCO32- ⇔ CaCO3(s) + uH+

(1)

where x,y,z and u are stoichiometric coefficients.

125

Figure 1. Experimental apparatus for the investigation of spontaneous precipitation and seeded crystal growth kinetics of calcium carbonate.

In the case of spontaneous precipitation experiments, the supersaturated solutions were initially stable. Initial total calcium concentration in spontaneous precipitation experiments was equal to 5.0 mM. The decrease of pH was monitored by a pH-meter, connected to a personal computer. Throughout the course of the spontaneous precipitation experiments, the pH of the working solution as a function of time was recorded and the collected data were stored in the computer for further analysis. The end of precipitation was verified by pH-constancy following initial pH drop. Solutions were filtered and the collected solids were dried overnight at 70°C for further characterization. The precipitation rates were determined from the pH variation as a function of time using curve-fitting software. For seeded growth experiments, the working solutions were stable for periods exceeding 24h and were supersaturated with respect to all calcium carbonate polymorphs. The initial total calcium concentration in the seeded growth experiments was equal to 2.0 mM. One or two hours following pH adjustment in the working solutions, a precisely weighed amount of calcite crystals (ca. 20mg) was introduced in the supersaturated solutions and the precipitation started immediately. A pH drop as small as 0.005 pH units triggered the addition of titrant solutions from two mechanically coupled syringes of a computer controlled titrator through the appropriate software. Throughout the course of the seeded growth experiments, the pH of the working solution and the added volume of titrants as a function of time were recorded and stored in the computer for further analysis. The titrant solutions in the two burettes consisted of calcium chloride (titrant A) and a mixture of sodium carbonate, sodium bicarbonate, and an additive as appropriate (titrant B). The supersaturated working solution contained sufficient concentration of sodium chloride (inert electrolyte) so as to maintain the solution ionic strength during titrants addition. The titrant solutions were prepared so as to have the following composition : Titrant A : Titrant B :

126

(10C1 + 2C1) M CaC12 l0C1 (M) Na2CO3 + 2C1 (M) NaHCO3 + 2CIN (M)

where C1 and CIN are the total calcium and tested additive concentrations in the working solution respectively. The concentration of the additive included in titrant B was twice its concentration in the working solution. This accounted for dilution caused by titrant addition. The amounts of total calcium and total carbonate in the titrants solutions were calculated according to the stoichiometry (1:1) of the precipitating calcium carbonate. During the course of seeded growth experiments, samples were withdrawn and filtered through membrane filters (0.2 µm). The filtrates were analyzed for total calcium by atomic absorption spectroscopy in order to confirm the constancy of the solution composition. In all seeded growth experiments, analysis showed that the total calcium concentration remained constant to within ± 2%. Working solutions at the end of each experiment were filtered and the solids were dried for further analysis. In seeded growth experiments, the precipitation rates were determined from the traces of the titrant volume added as a function of time, using curve-fitting software. The rates of calcium carbonate formation were normalized for the seed crystal total surface area as follows: dV C t Rg = —– —– (mol.min-1.m-2) dt At

(2)

where dV/dt is the rate of titrant addition of concentration Ct (mol/l) (10C1) and At is the total surface area of the added calcite seed crystals.

RESULTS AND DISCUSSION Characterization of the Precipitated Solids In all experiments, both at constant and decreasing supersaturation conditions, calcite was the only calcium carbonate phase found. Calcite is the thermodynamically most stable polymorph of calcium carbonate and its formation was favored by the experimental conditions. The identification of solid precipitates was done by XRD and SEM analysis. The XRD spectrum of spontaneously precipitated solid in the presence of humic acid and the reference spectral lines of calcite from the JCPDS database13 are shown in Figure 2. There is a close correlation between the reflections of the spectra. Calcite was also formed under the same conditions with fulvic and polyacrylic acids. A typical scanning electron micrograph of calcite seed crystals grown in the presence of fulvic acid is shown in Figure 3. As may be seen, the presence of fulvic acid did not cause any appreciable morphological changes in the characteristic rhombohedral shape of the calcite crystals formed in the absence of additives. Results were the same for the growth of calcite crystals in the presence of humic and polyacrylic acids.

Spontaneous Precipitation Experiments The inhibition of the crystal growth rates relative to the absence of additives was expressed by the relative inhibition defined as : o -Ri x100 Relative Inhibition (%) = R ——— Ro

(3)

127

where R0 and Ri are the calcite crystal growth rates measured in the absence and in the presence of the additives respectively. The experimental conditions for the calcium carbonate spontaneous formation, the induction times, and the relative inhibition obtained in the presence of the tested compounds are summarized in Table 1.

Figure 2. Powder x-ray diffraction spectrum (a) for calcium carbonate formed in the presence of humic acid at decreasing supersaturation and (b) the reference spectrum of calcite from JCPDS database (Card NO. 05-0586).

Figure 3. Scanning electron micrograph of calcite seed crystals grown in the presence of fulvic acid at conditions of sustained supersaturation.

128

As may be seen in Table 1, the induction time for the spontaneous precipitation of calcite in the presence of humic, fulvic, and polyacrylic acids was found to depend strongly on their concentration. It should be noted that polyacrylic acid was tested at higher supersaturations than humic and fulvic acids. This was done because polyacrylic acid at a concentration of 0.1 ppm gave significantly higher induction time (11515 s) than humic (1066 s) and fulvic (809 s) acids at the same initial calcium concentration (5.0 mM). The dependence of induction time on the concentration of humic, fulvic, and polyacrylic acids is shown in Figure 4. The influence of the additive’s concentration on induction times was found to be stronger for polyacrylic acid than for humic and fulvic acids, although polyacrylic acid was tested at higher supersaturations. As may be seen in Figure 4, humic acid was more effective than fulvic acid at concentrations higher than 0.2 ppm. For concentrations exceeding 0.2 ppm, humic acid gave longer induction times than with fulvic acid at the same initial concentration. The difference in the behaviour in the presence of the three additives tested may be explained by the fact that polyacrylic acid adsorbs more strongly on the surface of the newly formed calcite seed nuclei and crystals at pH = 8.50 and blocks more effectively the active growth sites than humic and fulvic acids. The difference between humic and fulvic acids may also be due to the different adsorption behaviour exhibited by these acids. Adsorption of the additives on the crystals surface is influenced by the molecular charge density, the size and the type of functional groups of the molecule, the ratio of ionizable to non-ionizable groups, pH, and the ionic strength of aqueous media.14

Table 1. Spontaneous precipitation of calcium carbonate from aqueous solutions in the presence of humic, fulvic and polyacrylic (PAA) acids at 25°C and pH = 8.50. Initial conditions, induction times, and relative inhibition. Exp. #

Additive

Concentration / ppm 0 0.1 0.2 0.3 0.4 0.5 0.1 0.2 0.3 0.4 0.5 0.1 0 0.1 0.125 0.15 0.175 0.2 0.225 0.25

Induction Time % Relative /S Inhibition 1 Blank-1 176 2 Humic Acid 1066 53.2 3 Humic Acid 1723 81.2 4 Humic Acid 5940 87.5 5 Humic Acid 8164 92.7 6 Humic Acid 11355 95.3 7 Fulvic Acid 809 48.6 8 Fulvic Acid 2338 71.3 9 Fulvic Acid 2921 75.9 10 Fulvic Acid 3585 81.5 11 Fulvic Acid 5442 85.1 12 PAA 11515 102 13 Blank-2 14 PAA 1199 83.1 15 PAA 1631 87.1 16 PAA 2443 91.1 17 PAA 3857 93.7 18 PAA 9400 96.3 19 PAA 11651 98.3 20 PAA Complete 100 Inhibition Blank-1 : Total Calcium = Total Carbonate = 5.00 × l0-3 M, Total Sodium Nitrate = 5.00 x 10-2 M. Blank-2 : Total Calcium = Total Carbonate = 6.00 × l0-3 M, Total Sodium Nitrate = 6.00 × l0-2 M.

129

The concentrations of humic, fulvic, and polyacrylic acids in spontaneous precipitation experiments were significantly lower than the initial calcium concentration. The concentrations of the additives did not decrease to any significant extent the solution supersaturation with respect to calcite due to complex formation of the respective anions with the free calcium ions. This fact suggested that the inhibition of these compounds were caused by preferential adsorption at the active growth sites of the new nuclei and of the seed crystals. Similar behaviour has been reported for water soluble polymeric compounds8,5 and for organophosphorus compounds.16 The results of the relative inhibition in the presence of the humic, fulvic, and polyacrylic acids as a function of their concentration are shown in Figure 5. The percent inhibition of each additive tested was found to depend strongly on the respective additive concentration. As may be seen in Figure 5, polyacrylic acid gave significantly higher relative inhibition than humic and fulvic acids at the same concentration. Polyacrylic acid at a concentration of 0.25 ppm resulted in a complete stop of calcium carbonate precipitation while humic and fulvic acids at 0.30 ppm gave relative inhibition 87.5 and 75.9% respectively. Humic acid gave higher relative inhibition than fulvic acid for the same concentration at all tested concentrations. This trend was also observed at the induction time dependence on the additive concentration at the same conditions. For concentrations higher than 0.2 ppm, there is tendency for a linear increase of the additive effectiveness while at higher additive concentrations the tendency to reach a plateau of the relative inhibition is higher. It is assumed that the measured inhibition was due to the adsorption of additive molecules at the active growth sites of the crystals to fit the kinetics data obtained from rate measurements. A Langmuir type kinetic model is used, in which it is assumed that the reduction of the rates of crystal growth is proportional to the coverage of the surface of the seed crystals by the additive molecule.17 This model assumes monolayer coverage, lack of lateral interactions between adsorbate molecules, and energetically equivalent adsorption sites. The relationship between the rates of crystal growth in the presence and in the absence of additives with the additive concentration is given by the Equation (4): 1 R0 ——— =1+ ––– R0 -Ri bC

(4)

where C is the additive concentration and b is the affinity constant, being a measure of preferential affinity of the additive for the crystal surface. The above equation describes a linear relationship between the ratio R0/(R0-RI) and the inverse of additive concentration, 1/C. For concentrations higher than 0.2 ppm experimental data gave a satisfactory fit for the tested additives (Fig. 6). For humic and fulvic acids at concentrations lower than 0.2 ppm a second linear part was found. This was probably due to either the presence of active growth sites with different energy on the crystal surface or to the lack of validity of the Langmuir adsorption model over the entire concentration range. In contrast to humic and fulvic acids, polyacrylic acid did not give a second linear dependence. According to the model described by Eq. (4) the intercept is expected to be equal to unity. For the polyacrylic and humic acids the value of the intercept was lower than 1. This value suggests complete inhibition at concentrations below the concentration corresponding to the complete coverage of crystal surface with a monolayer of polyacrylic and humic acids molecules.18 The same behaviour has been reported for the crystal growth of several insoluble salts of the alkaline earth metals in the presence of

130

Additive Concentration / ppm Figure 4. Dependence of the induction time preceding the spontaneous precipitation of calcium carbonate in aqueous solutions in the presence of humic, fulvic, and polyacrylic acids at pH = 8.50, 25°C.

Additive Concentration / ppm Figure 5. Dependence of the ratio Ri/R0 on the humic, fulvic, and polyacrylic acids concentration for the spontaneous precipitation of calcium carbonate at pH = 8.50, 25°C.

131

1 /C Figure 6. Fit of the kinetic results for the spontaneous precipitation of calcium carbonate in the presence of humic, fulvic, and polyacrylic acids at pH = 8.50, 25°C.

organophosphorus compounds.2 For fulvic acid the intercept was found to be greater than 1. This suggested that complete inhibition is not possible and that the relative inhibition reaches a limiting value at certain additive concentration.

Seeded Growth Experiments The seeded growth experiments for the evaluation of the effectiveness of polyacrylic, humic, and fulvic acids as precipitation inhibitors were done with stable supersaturated solutions. The experimental conditions for the seeded growth experiments and the relative inhibition obtained are summarized in Table 2. In all seeded growth experiments, overgrowth of calcium carbonate on calcite seed crystals started immediately after the addition of seed crystals without any induction times at all concentrations of tested additives. The relative inhibition of calcium carbonate overgrowth on calcite seed crystals in the presence of humic, fulvic, and polyacrylic acids was found to depend strongly on their concentrations but to a different extent. The dependence of % relative inhibition on the tested additive concentration is shown in Figure 7. As may be seen in Figure 7, the polyacrylic acid gave higher relative inhibition than humic and fulvic acids for the same concentration. Also, the presence of polyacrylic acid resulted in the complete inhibition of calcium carbonate precipitation at concentrations exceeding 0.2 ppm. Similar results were obtained for humic acid which was found to be a more efficient inhibitor than fulvic acid. Humic acid completely inhibited calcium carbonate precipitation at concentrations higher than 0.8 ppm.

132

Table 2. Crystal growth of calcium carbonate on calcite seed crystals in the presence of humic, fulvic, and polyacrylic acid at 25°C and pH = 8.50. Initial conditions and relative inhibition. Exp. #

Additive

Concentration Relative / ppm Inhibition 1 Blank 0 2 Humic Acid 0.05 25.9 3 Humic Acid 0.1 56.4 4 Humic Acid 0.2 79.1 5 Humic Acid 0.3 89.3 6 Humic Acid 0.4 95.1 7 Humic Acid 0.5 96.9 8 Humic Acid 0.6 98.7 9 Humic Acid 0.8 99.1 10 Humic Acid 1.0 100 11 Fulvic Acid 0.1 12.5 12 Fulvic Acid 0.2 39.4 13 Fulvic Acid 0.3 59.3 14 Fulvic Acid 0.5 78.1 15 Fulvic Acid 1.0 85.9 16 Fulvic Acid 1.5 94.2 17 Fulvic Acid 2.0 94.4 18 PAA 0.05 90.0 19 PAA 0.1 97.5 20 PAA 0.15 96.9 21 PAA 0.2 98.5 22 PAA 0.25 100 Blank : Total Calcium = Total Carbonate = 2.00 x 10-3 M, Total Sodium Chloride = 2.00 x I0-2 M.

Additive Concentration / ppm Figure 7. Dependence of the ratio Ri/Ro on the humic, fulvic, and polyacrylic acids concentration for the calcium carbonate overgrowth on calcite seed crystals at pH = 8.50, 25°C.

133

Fulvic acid did not completely inhibit calcite crystal growth even at concentrations of 2 ppm, significantly higher than the concentration in which humic acid resulted in complete inhibition. Relative inhibition for fulvic acid reached a plateau. This was probably due to the stronger adsorption of polyacrylic and humic acid molecules on the active growth sites of the calcite seed crystals in comparison with fulvic acid at pH = 8.50. A linear dependence of the relative inhibition on the additive concentration was observed at concentrations higher than 0.4 ppm for humic acid and 0.5 ppm for fulvic acid. Thus, from the kinetic analysis according to the Langmuir model (Eq. (4)) a linear relationship between the ratio R0/(R0-Ri) and the inverse of additive concentration, C, is predicted for concentrations higher than 0.4 ppm for humic acid and 0.5 ppm for fulvic acid respectively. Polyacrylic acid gave linear dependence between the ratio R0/(R0-RI) and the inverse of additive concentration, 1/C, at all concentrations. The same response was observed for the tested compounds in the spontaneous precipitation experiments. The analysis of the kinetic data obtained from the overgrowth of calcium carbonate on calcite seed crystals is shown in Figure 8. At concentrations lower than 0.4 ppm for humic acid and 0.5 ppm for fulvic acid, a second linear dependence was observed. Again this is probably due to either the presence of active growth sites with different energy at the calcite surface or to changes in the mode of adsorption with increasing surface concentration. At higher surface coverage, the structure of the molecules adsorbed on the mineral surfaces may change. The polyacrylic acid molecule is linear and simpler in structure than humic and fulvic acid molecules. From the kinetic analysis (Eq. (4)), the value of the intercept for polyacrylic and humic acids was found to be lower than unity and for fulvic acid slightly higher than unity. This is in agreement with the variation of relative inhibition versus additive concentration (Figure 7). The values of the affinity constants for humic, fulvic, and polyacrylic acids, as calculated from the kinetics analysis according to, the Langmuir type model, are

1 /C Figure 8. Fit of the kinetic results for the overgrowth of calcium carbonate on calcite seed crystals in the presence of humic, fulvic and polyacrylic acids at pH = 8.50, 25°C.

134

summarized in Table 3. The affinity constant for polyacrylic acid was significantly higher than affinity constants for humic and fulvic acid. The affinity constant for humic acid was higher than the affinity constant for fulvic acid. Additives may be arranged in the following order with respect to their affinity constants for calcite crystal growth : Polyacrylic acid >> Humic Acid > Fulvic Acid Polyacrylic and humic acids, which caused complete inhibition at lower concentrations than fulvic acid, yielded higher values of the affinity constant. The affinity constant of polyacrylic acid was significantly higher than the affinity constants of the organophosphorus compounds, BPDMI and ETMPA, which were reported to effectively inhibit the precipitation of calcium carbonate. Polyacrylic acid is more effective than these organophosphorus compounds because it gave the same results but at lower concentration.19-21

Table 3. Affinity constants for the additive-calcite interface calculated from kinetics of crystal growth data according to the Langmuir model. Additive Affinity Constant Humic Acid 1.9 X 107 Fulvic Acid 4.6 X I06 PAA 1.6X 109 Mellitic Acid 2.0X 106 (19) Phosphate 5.9x 107 (20) BPDMI 1.6 X 107 (21) ETMPA 1.0 X 107 (21) BPDMI : 1,3-Bis[( I-phenyl- 1-dihydroxyphosphony1)-methy1]-2-imidazolidinone ETMPA : Ethylenediamine-tetra-bis-methylene phosphonic acid

The affinity constant for humic acid was of the same order of magnitude as the affinity constants for the organophosphorus compounds. BPDMI and ETMPA gave the same relative inhibition at almost the same concentration and completely inhibited the precipitation of calcium carbonate at the same low concentration at which humic acid was tested.21 In the contrast, the affinity constant of fulvic acid was lower (comparable to mellitic acid), and its inhibition efficiency was poorer than with polyacrylic and humic acids and BPDMI and ETMPA. The difference in effectiveness as precipitation inhibitors among polyacrylic, humic, and fulvic acids may be attributed to the characteristics of their molecules, such as the degree of ionization of their functional groups, their charge density at pH = 8.50, molecule size, stereochemical geometry, flexibility and to the ability for binding to more than one active growth sites by one molecule. The polyacrylic acid molecule is linear, fully ionizable at pH = 8.50, and more flexible for adsorption on crystal surface than humic and fulvic acids. The exact formulae for humic and fulvic acids are not known; thus it is very difficult to determine specific reasons for the better performance of humic acid as a precipitation inhibitor in comparison with fulvic acid.

135

CONCLUSIONS In the present work it has been shown that the presence of humic, fulvic, and polyacrylic acid at very low concentrations can strongly inhibit nucleation and crystal growth of calcium carbonate from labile and metastable supersaturated aqueous solutions. Humic, fulvic, and polyacrylic acids did not affect the nature of the calcium carbonate phase formed, which in all cases was calcite. Polyacrylic and humic acids gave higher relative inhibition than fulvic acid at the same conditions. Polyacrylic and humic acids were found to completely inhibit the crystallization of calcium carbonate at very low concentrations while fulvic acid failed to yield complete inhibition even at significantly higher concentrations. The higher inhibition efficiency of polyacrylic acid in comparison with humic and fulvic acids is due to the larger extent of adsorption of polyacrylic acid molecules on calcite surface in our experimental conditions, as indicated by the higher affinity constant of polyacrylic acid molecules for calcite surface. Also, humic acid was more effective as a precipitation inhibitor than fulvic acid. The difference in inhibition efficiency between polyacrylic, humic, and fulvic acids may be attributed to their molecular characteristics.

REFERENCES 1. J.C. Cowan and D.J. Weintritt, Water Formed Scale Deposits, Gulf Publishing, Houston, Texas, 60 (1976). 2. Z. Amjad, “Kinetic Study of the seeded growth of Calcium Carbonate in the presence of Benzenepolycarboxylic Acids”, Langmuir 3:224 (1987). 3. D. Wilson, ‘Influence of molecular weight on selection and application of Polymeric Scale Inhibitors’, Corrosion/94, Paper No. 48, NACE, Houston, Texas, (1994). 4. J. Glater, J.L. York and K.S. Campbell, “Scale Formation and Prevention”, in : Principles of Desalination, K.S. Spiegler and A.D.K. Laird. ed., Academic Press, 627 (1980). 5. M.P.C. Weijnen and G.M. van Rosmalen, ‘The influence of various polyelectrolytes on the precipitation of gypsum’ Desalination, 54:239 (1985). 6. O.J. Vetter, ‘An evaluation of scale inhibitors’, J.Pet.Techn., 997 (1972). 7. Z. Amjad, ‘Calcium Sulfate Dihydrate (Gypsum) scale formation on heat exchanger surfaces : The influence of Scale Inhibitors’, J. Colloid Interface Sci. 123:523 (1988). 8. A.M. Carrier and M.L. Standish, “Polymer mediated Crystal habit modification”, in : Mineral Scale Formation and Prevention. Z. Amjad, ed., Plenum New York, 63 (1995). 9. Mortvedt, P.M. Giordano and W. L. Lindsay, “Micronutrients in Agriculture”, American Society of Agronomy, Madison, Wiscosin, (1 972). 10. J.L. Lacout, P.G. Koutsoukos, N. Rouquet and M. Freche, ‘Effect of Humic compounds on the crystal growth of Dicalcium Phosphate Dihydrate’, Agrochimica, 36:500 (1992). 11. Z. Amjad and M.M. Reddy, ‘Influence of Humic compounds on the Crystal Growth of Hydroxyapatite’, in : Water Soluble Polymers : Solution, properties and applications, Z. Amjad, ed., Plenum Publishing Co., N.Y. 77 (1998). 12. M.M. Reddy and G.H. Nancollas, ‘The crystallization of Calcium Carbonate II. Calcite Growth mechanism’, J. Colloid Interface Sci. 37, 824 (1971). 13. JCPDS ASTM Card No 05-0586 14. M.P.C. Weijnen, ‘The influence of additives on the crystallization of Gypsum’, Ph.D. Thesis, University of Delft, (1 984). 15. A.E. Austin, J.F. Miller, D.A. Vaughan and J.F. Kircher, ‘Chemical additives for calcium sulfate scale control’, Desalination, 16:345 (1975). 16. C.W. Davies and G.H. Nancollas, ‘The precipitation of Silver Chloride from aqueous solutions. Part III. Temperature Coefficients of Growth and Solutions’, Trans.Faraday.Soc., 51 :818 (1955).

136

17. G.H. Nancollas and S.J. Zawacki, ‘Inhibitors of Crystallization and Dissolution’, Industrial Crystallization 84:5 1 (1 984). 18. W.H. Leung and G.H. Nancollas, ‘Nitrilotri(methylenephosphonic acid) adsorption on Barium Sulfate crystals and its influence on crystal growth’, J.Crystal Growth, 44:163 (1978). 19. E.K. Giannimaras and P.G. Koutsoukos, ‘The crystallization of Calcite in the presence of Orthophosphate’, J. Colloid Interface Sci. 116:423 (1987). 20. A.G. Xyla, J. Mikroyannidis, and P.G. Koutsoukos, ‘The inhibition of Calcium Carbonate precipitation in aqueous media by Organophosphorus compounds’, J. Colloid Interface Sci. 153:537 (1992).

137

This page intentionally left blank.

CALCIUM CARBONATE AND CALCIUM PHOSPHATE SCALE FORMATION AND INHIBITION AT ELEVATED TEMPERATURE

Pavlos G. Klepetsanis,

1,3

Petros G. Koutsoukos

1,2

and Zahid Amjad

4

1Institute

of Chemical Engineering and High Temperature Chemical Processes, P.O.Box 1414 Patras, GR-26500, GREECE 2Department of Chemical Engineering 3Department of Pharmacy University of Patras Patras, GR-26500, GREECE 4The BFGoodrich Company 9911 Brecksville Road, Breckville, Ohio 44141, USA

ABSTRACT In the present work, the effect of acrylic acid copolymers in the formation of calcium carbonate and calcium phosphate scale deposits on heated surfaces was studied. The experiments were done in a small heat exchanger and the effectiveness of acrylic acid copolymers was tested for 50 hours. In the calcium carbonate formation experiments, the presence of acrylic acid copolymers was found to stabilize thermodynamically unstable polymorphs of calcium carbonate and did not cause significant changes in morphology of the crystals formed. Ten ppm of polyacrylic acid, molecular weight ca. 2000, was found to be the most effective inhibitor for calcium carbonate yielding 95% inhibition efficiency. Polyacrylic acid effectiveness was found to decrease with increasing pH. The acrylic acid-sulfonic acid copolymer and acrylic acid-sulfonic acid-styrene sodium sulfonate tertpolymer were found to be less effective in comparison with polyacrylic acid at the same concentration (43 and 34% inhibition efficiency respectively) but their effectiveness was found to increase with increasing pH. In the calcium phosphate formation experiments hydroxyapatite, the thermodynamically most stable phase of calcium phosphate, was formed both in the presence and in the absence of acrylic acid copolymers. Polyacrylic acid at a concentration of 10 ppm was found to be the most

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers, New York, 2000

139

effective inhibitor for calcium phosphate with inhibition efficiency of 60%. The other copolymers of acrylic acid tested did not affect significantly the formation of calcium phosphate. The inhibition efficiency of PAA was found to decrease with increasing total calcium concentration in the solution.

INTRODUCTION The formation of mineral deposits of sparingly soluble calcium salts is a serious problem in many industrial operations and in particular in heat exchangers, where untreated natural waters are used1. Two of the most commonly encountered salts in the mineral deposits are calcium carbonate and calcium phosphate. These salts form hard and strongly adhering deposits on the metallic surfaces and their formation is favoured by the decrease of their solubility with increasing temperature. The formation of deposits on metallic surfaces of heat exchangers reduces heat transfer causing significant energy losses with the concomitant danger of overheating and unscheduled shutdown of operation. Moreover, scaling is often accompanied with corrosion which leads to damage of the metallic parts of heat exchangers. Calcium carbonate is found in three different crystalline polymorphs: calcite, aragonite and vaterite listed in order of increasing solubility. Calcium phosphate is found as hydroxyapatite (Ca5(PO4)3OH, HAP), tricalcium phosphate (Ca3(PO4)2, TCP), octacalcium phosphate (Ca8H2(PO4)6.5H2O, OCP), calcium phosphate anhydrous (CaHPO4, DCPA) and dicalcium phosphate dihydrate (CaHPO4.2H2O, DCPD) listed in order of increasing solubility. The formation of calcium phosphate phases depends strongly on supersaturation, ionic medium and pH of the aqueous phase. A large number of methods have been developed for the prevention of deposits formation2 including the decrease of the aqueous media pH and the addition of chelating agents or of water-soluble compounds (organophosphonates, polymers, etc.). The most promising method is the addition of water-soluble compounds at very low concentrations.2,3 These compounds give satisfactory results even at very low concentrations such as a few parts per million (ppm) or less, and affect either the nucleation or the crystal growth stages of the mineral formation. Organophosphonates are very effective inhibitors in the formation of many sparingly soluble salts but have two significant disadvantages in comparison with polymers: they decompose at elevated temperatures, releasing phosphate ions that may lead to the formation of insoluble calcium phosphate, in addition to forming calcium phosphonate salts at high concentrations. The efficiency of the polymeric compounds is strongly influenced by the nature of their functional groups, polymer composition, molecular weight, chemical composition and the pH of the aquatic medium as reported in earlier investigation.4,5 The introduction of various chemical groups in the structure of successful scale inhibitors, like polyacrylic acid, aims at increasing their thermal stability and dispersibility.6 The type of functional groups present influence significantly the effectiveness of the water soluble polymers as scale inhibitors, because of the incipient different interactions with the developing crystal surfaces. As a result of these interactions, morphological changes of the deposits are possible due to the different growth rates of the crystal faces. In addition, thermodynamically unstable phases may be favoured because of kinetic stabilization. In the present work, we have investigated the formation of calcium carbonate and calcium phosphate mineral deposits on heated metal surfaces in contact with untreated

140

tap water of well defined composition, in the presence of acrylic acid copolymers at various pH values and at different total calcium concentrations. The copolymer of acrylic acid with sulfonic acid (PAS), the tertpolymer of acrylic acid with sulfonic acid and sulfonated styrene (PSS), the poly(acrylic acid) (PAA) and 1 -hydroxyethylidene- 1,1diphosphonic acid (HEDP) were tested and compared.

EXPERIMENTAL All experiments were done at 60 ± 2°C in a small pilot heat exchanger. The heat exchanger consisted of an inner metal tube (SS304) and an outer plexiglass tube. The metal tube was heated internally by steam and the working solution was circulated with a pump in the space between the metal and the plexiglass tubes. A schematic representation of the pilot heat exchanger is shown in Figure 1.

Figure 1. Heat exchanger apparatus used for the evaluation of acrylic acid copolymers as crystallization inhibitors.

The working supersaturated solutions volume totalling 15L were prepared from tap water, filtered to remove suspended particles when needed. Supersaturation was adjusted to the desired levels as follows. For the calcium carbonate deposition calcium chloride from the respective stock solutions was used. For the calcium phosphate deposits in addition to calcium chloride, sodium dihydrogen phosphate, also from standard stock solutions was added. In experiments in which additives were tested, the additive was added from concentrated stock solutions to the supersaturated aqueous medium. The pH was adjusted with the addition of 1N NaOH or 1N HCI (Merck, Titrisol) solutions as needed. The initial pH in calcium carbonate formation experiments was 8.50 and in calcium phosphate formation experiments was 5.50. The working solutions during the pH adjustment and throughout the experiments were stirred with a small submersible water pump. Following pH adjustment, the working solution was circulated in the heat

141

exchanger and heating was turned on. The system reached thermal equilibrium at 60°C after approximately one hour. The temperature in the interior of the metal tube was maintained constant with a proportional temperature controller and was monitored. Each experiment lasted for 50 hours and during this time period samples were withdrawn from the working solution and were filtered through membrane filters (0.2 µm, Millipore). The filtrates were analyzed for calcium by atomic absorption spectroscopy (Perkin Elmer, AAnalyst 300) in calcium carbonate formation experiments and for phosphate by phosphovanadomolybdate method in calcium phosphate formation experiments. After the end of experiments, the solids formed on the outer surface of the metal tube were collected, dried overnight at 70°C and were further analyzed for the identification of the mineral phase by powder X-ray diffraction (XRD, Philips - PW 1840) and Scanning Electron Microscopy (SEM, JEOL - JSM 5300). The compounds tested in this work were water soluble acrylic acid copolymers kindly supplied by BF Goodrich, and HEDP by Monsanto. The names of acrylic acid copolymers and their chemical structures are presented in Table 1. Table 1. Polymers tested as calcium carbonate and calcium phosphate crystallization inhibitors.

RESULTS AND DISCUSSION Calcium Carbonate Formation Experiments In the experiments done in the absence of the tested polymers, a mixture of calcite and aragonite formed at pH = 7.5 and only aragonite was formed at pH = 9.0, as identified using XRD analysis. In the presence of polymers, only aragonite formed at all experimental conditions. Calcite is the thermodynamically most stable polymorph of calcium carbonate and its formation is favoured in our experimental conditions at pH = 7.5. The concentration of magnesium ions (24 ppm) is lower than the concentration necessary (54 ppm) for the stabilization of aragonite under the same condition.7 The presence of polymers was found to inhibit the transformation of aragonite to calcite at

142

pH = 7.5. Aragonite is thermodynamically less stable in comparison with calcite and may be considered as a precursor to calcite formation. The formation of transient phases has been reported for several sparingly soluble salts.8,9 In Figure 2, the XRD - spectrum of aragonite formed in the presence of polymers at pH = 9.0 and the spectral lines of aragonite from JCPDS Card file 41-147510 are shown. Scanning electron micrographs of solids formed in the presence and in the absence of acrylic acid copolymers did not show significant changes in the morphology of crystals, which maintained the rod-like prismatic shape exhibited in the absence of the additives.

Figure 2. Powder x-ray diffraction spectrum of aragonite formed in the presence of polyacrylic acid copolymers and reference lines for aragonite.

The inhibitor efficiency of the tested additive at a given concentration as a calcium carbonate inhibitor was calculated using the following equation:11

% Inhibition =

[Ca2+]e - [Ca2+]f [Ca2+]i - [Ca2+]f

*100

(1)

where [Ca2+]e, [Ca2+]i, and [Ca2+]f are the total calcium concentration in the presence of additive at a given time, the initial total calcium concentration and the final total calcium concentration, respectively, in the absence of additives. In Table 2, the initial experimental conditions, the percent inhibition for the tested polymers at the end of the experiments done and the respective percent decrease of precipitation rate are shown. The working solutions for calcium carbonate formation experiments were stable for at least 24 hours at room temperature. In all experiments the working solutions contained 24 ppm magnesium and 10 ppm chloride. In these experiments, the effect of HEDP on the formation of calcium carbonate polymorphs was investigated. HEDP is a very effective inhibitor for calcium carbonate system12 and was used as reference compound.

143

The variation of the percent inhibition efficiency versus time for the formation of calcium carbonate at pH = 7.5 and pH = 9.0 is shown in Figures 3 and 4 respectively.

Table 2. Initial experimental conditions, % Inhibition efficiency and % decrease of precipitation rate for the formation of calcium carbonate in heat exchanger in the presence of acrylic acid copolymers. Additive PAA PAS PSS HEDP PAA PAS PSS HEDP PAA PAS

Initial Experimental Conditions [Ca]T = 7mM [Add.] = l0ppm pH = 7.5 [Ca]T = 7mM [Add.] = l0ppm pH = 9.0 [ C a ]T = 10.5 m M [Add.] = 20ppm pH = 7.5

% Inhibition Efficiency 95 43 34 95 70 54 34 70 55

% Decrease Precipitation Rate 100 88 73 100 97 92 80 97 92

0

51

As shown in Figure 3 and in Table 2, the percent inhibition in the presence of PAA and HEDP at pH = 7.5 is significantly higher than the respective value obtained in the presence of PAS and PSS. The tested polymers and HEDP may be classified in the following order with respect to the inhibition efficiency after the lapse of test time (50 h): PAA ≈ HEDP >> PAS > PSS Also, the percent decrease of precipitation rate was found to follow the same order. PAA and HEDP inhibited the formation of calcium carbonate at pH = 7.5 almost completely

Figure 3. Variation of percent inhibition efficiency as a function of time for the formation of calcium carbonate in the presence of acrylic acid copolymers and HEDP at pH = 7.5. Initial total calcium = 280 ppm; polymers concentration = 10 ppm

144

(~100 %). PAS and PSS decreased significantly the precipitation rate (88 and 73 % respectively) but they failed to stop the deposit formation completely. Thus, the crystals formation was retarded to a large extent in the presence of PAS and PSS but the growth continued until the supersaturation decreased to a significantly lower value. The decrease in the inhibition efficiency of PAS and PSS in comparison with that of PAA may be ascribed to the presence of groups with different hydrophobic character and size in these polymers. These parameters may strongly affect the adsorption of the molecules on the charged crystal surface. The sulfonated styrene group and 2-acrylamido-2-methyl propane sulfonic acid group have larger size, lower charge density and they are more hydrophobic in comparison with the carboxylic groups, for which they substitute in the polymeric chain. It is possible therefore that PAA adsorbs to a larger extent onto calcite in comparison with PAS and PSS. As a result it exhibited a higher inhibition efficiency for the same concentration at the same experimental conditions.

Time / min Figure 4. Variation of percent inhibition efficiency as a function of time for the formation of calcium carbonate in the presence of acrylic acid copolymers and HEDP at pH = 9.0. Initial total calcium = 280 ppm; polymers concentration = 10 ppm

At pH = 9.0 the same behaviour of the tested polymers and HEDP was found, as may be seen in Figure 4. The inhibition efficiency and the percent decrease of the precipitation rates in the presence of PAA and HEDP were higher than those corresponding to PAS and PSS. It should be noted that the inhibition efficiency and the percent decrease of the precipitation rates in the presence of PAA and HEDP at pH = 9.0 were lower than the respective values at pH = 7.5. This was attributed to the higher supersaturation ratio of the experiments at pH = 9.0 (the initial total calcium concentration was the same for both pH values). The inhibition efficiency and the percent decrease of precipitation rates in the presence of PAS were found to increase at pH = 9.0 in comparison with the respective values at pH = 7.5. This increase may be ascribed to changes of the surface charge of the crystals formed.13 The acidic groups of acrylic acid copolymers were completely dissociated both at pH = 7.5 and 9.0 (the respective pK

145

values are about 3.5). Consequently the interactions between PAS molecules and the surface of the crystals formed and grown are expected to be higher resulting most probably to stronger adsorption of PAS molecules on the active growth sites of the crystals at this pH value. The inhibition efficiency obtained in the presence of PSS was found to remain constant as pH increased. This may be due to the high hydrophobic character and to the size of the sulfonated styrene group, which may hinder the adsorption of PSS molecules to the active crystal growth sites. 13 The copolymers PAA and PAS were tested at higher calcium concentration (total calcium = 420 ppm) and pH = 7.5. The variation of the percent inhibition with time in the presence of PAA and PSA is shown in Figure 5. As shown in Figure 5, PAS did not affect significantly the formation of calcium carbonate at this calcium concentration even at the highest concentration tested (20 ppm). PAA was also found to give satisfactory inhibition efficiency at these conditions but its inhibition efficiency was found lower (~50%) than the respective value (~95%) for total calcium concentration equal to 260 ppm at the same pH. This result was attributed to the higher supersaturation with respect to all calcium carbonate polymorphs. The decrease of PAA inhibition efficiency with increasing supersaturation was found in the experiments done at pH = 9.0 as well.

Calcium Phosphate Formation Experiments In the experiments done both in the absence and in the presence of the acrylic acid copolymers, HAP was formed exclusively. Examination of the XRD spectra showed that the mineral deposits consisted of HAP without the presence of any precursor phases. HAP, the thermodynamically most stable calcium phosphate phase was favoured in our experimental conditions at pH = 5.5. In Figure 6, the XRD - spectrum of HAP formed in the presence of polymers at pH = 5.5 and the spectral lines of HAP from JCPDS Card file 09-043214 are shown. Examination of the particulate deposits morphology from scanning electron micrographs of the minerals formed both in the presence and in the absence of polymers did not show significant changes in the morphology of crystals, which retained their microcrystalline prismatic shape (< 1 µm).

Time / min Figure 5. Variation of percent inhibition efficiency as a function of time for the formation of calcium carbonate in the presence of acrylic acid copolymers at pH = 7.5. Initial total calcium = 420 ppm; polymers concentration = 20 ppm

146

Figure 6. Powder x-ray diffraction spectrum of HAP formed in the presence of polyacrylic acid copolymers and reference lines for HAP.

Time / min Figure 7. Variation of percent inhibition efficiency as a function of time for the formation of calcium phosphate in the presence of acrylic acid copolymers at pH = 5.5. Total calcium = 120 ppm; polymers concentration = 10 ppm

147

The inhibition efficiency for a given concentration as a calcium phosphate inhibitor was calculated using the following equation:11 % Inhibition =

[PO43-] e - [PO3-4]f [PO3-4 ] i - [PO3-4 ]f

*100

(1)

\

where [ PO43 - ]e is the total phosphate concentration in the presence of additive at a given time, [PO43- ]i the initial total phosphate concentration and [PO3-4 ]f the final total phosphate concentration in the absence of additives. In Table 3, the initial experimental conditions, the percent inhibition for the tested polymers at the end of calcium phosphate formation experiments done and the respective percent decrease of precipitation rate are shown. The working solutions for calcium phosphate formation experiments were stable for at least 24 hours at room temperature. In all experiments the working solutions contained 24 ppm magnesium and 10 ppm chloride, as C1- ions. The variation of the percent inhibition efficiency with time for the formation of calcium phosphate at pH = 5.5 is shown in Figure 7. As shown, the percent inhibition for PAA at pH = 5.5 is significantly higher than the percent inhibition in the presence of PAS and PSS. The presence of PSS did not affect the formation of calcium phosphate and its percent inhibition was almost equal to zero. The tested polymers may be classified in the following order with respect to the inhibition efficiency past the lapse of the test time (50 hours) : PAA >> PSA > PSS Table 3. Initial experimental conditions, % Inhibition efficiency and % decrease of precipitation rate for the formation of calcium phosphate in heat exchanger in the presence of acrylic acid copolymers. Additive

Initial Exp. Conditions

% Inhibition Efficiency

% decrease Precipitation Rate

PAA PAS PSS PAA PAS PSS

[ PO 4] T = 3 m M pH = 5.5 [Add.] = l0ppm [PO4]T = 6.0 mM pH = 5.5 [Add.] = l0ppm

60 20 0 44 0 -

92 30 30 43 0 -

The decrease in the inhibition efficiency in the presence of PAS and PSS in comparison with that obtained in the presence of PAA may be ascribed to the presence of groups with different hydrophobic character and size in these polymers as described in the first section of this paper related with calcium carbonate formation. These parameters may strongly affect the adsorption of the molecules on the charged crystal surface. PSS and PAS molecules are larger in size, they have lower charge density and they are more hydrophobic in comparison with the carboxylic groups, for which they substitute in the polymeric chain. Thus, PAA seems to adsorb to a larger extent compared with PAS and PSS. As a result it exhibited higher inhibition efficiency for the same concentration under the same conditions.

148

The copolymers PAA and PAS were further tested at higher calcium concentration (total calcium = 260 ppm) and for the same pH. The percent inhibition in the presence of PAA was found to be lower in comparison with the percent inhibition for PAA in the experiments which were done in the at lower calcium concentration corresponding to lower supersaturations. The presence of PAS did not affect the formation of calcium phosphate at these conditions

CONCLUSIONS In the present work it has been shown that the presence of polyacrylic acid strongly inhibits the formation of calcium carbonate and calcium phosphate mineral deposits at 60°C. This may be explained by the adsorption of polyacrylic acid molecules on the active crystal growth sites. The modification of polyacrylic acid molecules with the introduction of sulfonic acid and sulfonated styrene groups in the polymeric chain was found to decrease its performance at the same experimental conditions. PAA was found to be the most effective inhibitor for both salts and its performance decreased with supersaturation and increasing pH. For calcium carbonate formation, the effectiveness of PAA was found similar to the effectiveness of HEDP which is a very good inhibitor for calcium carbonate. The other polyacrylic acid copolymers gave better results in calcium carbonate formation in comparison with the formation of calcium phosphate. The presence of acrylic acid copolymers was found to inhibit the transformation of aragonite to calcite at pH = 7.5 and did not affect the nature and morphology of calcium phosphate precipitates which in all cases consisted exclusively of HAP.

REFERENCES 1. J.C. Cowan and D.J. Weintritt, Water formed Scale Deposits, Gulf Publishing Co., Houston, TX, 60 (1976). 2. J. Glater, J.L. York and K.S. Campbell, “Scale Formation and Prevention”, In Principles of Desalination, K.S.Spiegler and A.D.K.Laird (Eds.), Academic Press, 627 (1980). 3. O.J. Vetter, “An evaluation of scale inhibitors”, J.Pet.Techn., 997 (1972). 4. D. Wilson, “Influence of Molecular Weight on Selection and Application of Polymeric Scale Inhibitors”, Corrosion/94, Paper No. 48, NACE, Houston, Texas, (1994). 5. Z. Amjad, “Influence of Calcium Fluoride Crystal Growth by Polyelectrolytes’’, Langmuir, 7:2405 (1991). 6. Z. Amjad, “Performance of polymers as precipitation inhibitors for calcium phosphonate”, Tenside Surf. Det., 34:2 (1997). 7. T.G. Sabbides and P.G. Koutsoukos, “The crystallization of Calcium Carbonate in artificial seawater; role of the substrate”, Journal of Crystal Growth, 133:13 (1993). 8. J.S. Manne, N. Biala, A.D. Smith and C.C. Gryte, ‘The effect of Anionic Polyelectrolytes on the Crystallization of Calcium Oxalate Hydrates’, Journal of Crystal Growth, 100:627 (1990). 9. G. Williams and J.D. Sallis, ‘Structural factors influencing the ability of compounds to inhibit hydroxyapatite formation’, Calcif. Tissue Int., 34: 169 (1982). 10. JCPDS ASTM Card No 41-1475 11. Z. Amjad, J. Pugh, J. Zibrida and B. Zuhl, ‘Polymer Performance in Cooling Water : The Influence of Process Variables’, Materials Performance, 36:32 (1997). 12. A.G. Xyla, E.K. Giannimaras and P.G. Koutsoukos, ‘The precipitation of Calcium Carbonate in aqueous solutions’, Colloids and Surfaces, 53:241 (1991). 13. M.P.C. Weijnen and G.M. van Rosmalen, “The influence of various polyelectrolytes on the precipitation of gypsum”, Desalination, 54:55 (1 985). 14. JCPDS ASTM Card No 09-0432

149

This page intentionally left blank.

EFFECT OF TEMPERATURE ON BARIUM SULFATE SCALE INHIBITION OF DIETHYLENE TRIAMINE PENTA (METHYLENE PHOSPHONIC ACID)

Mingdong Yuan Baker Petrolite Sugar Land, TX 77478

ABSTRACT Laboratory experiments have been carried out to investigate the effect of temperature on the performance of diethylenetriamine penta (methylene phosphonic acid), DETPMP, against barium sulfate scale formation. This paper presents the findings from this recent experimental study. By maintaining both supersaturation and amount of equilibrium precipitation of barium sulfate constant in the test brines, it is found that as temperature increases from 25°C to 60°C, barium sulfate inhibition efficiency of DETPMP decreases considerably. However, as temperature increases further from 60°C to 95°C, the performance of DETPMP remains almost constant. Such phenomena were observed with test brines of different pHs and in absence/presence of hydrogen carbonate. Additional experiments of DETPMP adsorption on barium sulfate reagent powder show that adsorption on barium sulfate continuously increases as temperature rises. This suggests that the adsorption level of DETPMP on barium sulfate surface may not be the key factor in the nucleation inhibition or crystal growth retardation of barium sulfate, when temperature is a variable. Also it is found that for a given supersaturation and a given amount of equilibrium precipitation, the rate of barium sulfate precipitation increases as temperature elevates. INTRODUCTION In an oilfield production system, temperature is one of the most important variables concerning scale formation in the produced waters. It is well known that temperature change has a significant impact on the solubility of barium sulfate,1,2 which in turn affects its scaling

Advances in Crystal Growth Inhibition Technologies Edited by Amjad. Kluwer Academic/Plenum Publishers New York. 2000

151

tendency. A previous study by the author et al. shows that, for a given brine, as temperature lowers from 95°C to 25°C, common oilfield scale inhibitors (including DETPMP) apparently exhibited much reduced inhibition efficiency against barium sulfate scale precipitation corresponding to the temperature reduction.3 However, it was not clear that whether the reduction in the inhibition efficiency was a result of decreased barium sulfate solubility (increased scaling tendency) alone, or whether the chemical inhibition function had also been impaired at lower temperatures. A mathematical equation developed based on the nucleation theory suggests that, as temperature decreases, barium sulfate scale inhibition efficiency of DETPMP as well as other scale inhibitors should actually improve.4 In order to experimentally prove whether and how temperature affects the chemical inhibition function against barium sulfate precipitation, this study attempts to focus on inhibitor performances at three different temperatures by maintaining supersaturation and amount of equilibrium precipitation of barium sulfate constant in the test brines. This was made possible by altering sulfate ion concentration (which was in great excess of barium ions in the brines) from one test temperature to another, while keeping other ion constituents in the brines intact. DETPMP was chosen as a model inhibitor in this study, as it is widely used for barium sulfate scale inhibition in oilfield produced waters. In order to validate the experimental results, comparative experiments were also conducted at different brine pHs (PH 5.5 vs. pH 7.0) and in the brines with or without hydrogen carbonate (HCO3-). Apart from using a typical oilfield brine system, two molal (2m) sodium chloride solutions containing barium sulfate were also constructed and used as test brines, in which barium sulfate solubility data at both 25°C and 95°C are available. Additional adsorption experiments of DETPMP on barium sulfate powder were performed at different temperatures to establish whether there was any correlation with the inhibitor inhibition function.

DESCRIPTION OF THE EXPERIMENTS The Test Brines The chemistry of the oilfield brine used as the model brine is shown in Table 1 below. It should be noted that sulfate ion concentration was modified for higher temperature tests so that the supersaturation and amount of equilibrium precipitation of barium sulfate in the brine remains constant at the three test temperatures. For inhibition tests, a test brine was prepared synthetically as equal volumes of anionic water, AW (containing half of the sodium chloride and all the other anions) and cationic water, CW (containing half of the sodium chloride and all the other cations). The pH of AW and CW was adjusted/buffered to a required value (PH 5.5 or pH 7.0) by acetic acidsodium acetate solution. For the experiments of DETPMP adsorption on barium sulfate reagent powder, barium, sulfate and hydrogen carbonate ions were excluded from the test brine. The brine was synthetically prepared and pH adjusted/buffered to 5.5 by acetic acidsodium acetate solution. The 2m sodium brine is shown Table 2. Again, sulfate ion concentration was modified for higher temperature tests so that the supersaturation and amount of precipitation of barium sulfate from the brine remained constant at the three test temperatures. Barium Sulfate Scale Precipitation and Inhibition For the following experiment, 50ml glass crimp top vials were used as test containers. Into one vial, 25ml anionic water (AW) was added and dosed with DETPMP to a pre152

Table 1. The oilfield brine. CONCENTRATION (mg/L)

ION

95°C tests 23°C tests 60°C tests 15370 5370 15370 338 338 338 749 749 749 663 663 663 57.5 57.5 57.5 26937 26435 25890 SO42365 1050 I750 364 HCO3364 364 According to ScaleSoft Pitzer prediction,5 the three tests brines at the coresponding test temperatures all have a BaSO4 saturation index of 2.34 and an amount of equilibrium precipitation of 97.4 mg/L BaSO4. Na+ K+ Mg2+ Ca2+ Ba2+ C1-

Table 2. The 2 molal NaCl brine. CONCENTRATION (mg/1000g H2O)

ION

Na+ C1+ Ba2+ SO42-

25°C tests

95°C tests

46000 71000 86 240

46000 7028 1 86 1214

Note: Both brines are ten times supersaturated based on the barium sulfate solubility data reported at the two temperatures.1 The amount of barium sulfate precipitation is 144.2 mg/L for both brines.

153

determined test concentration, Into another vial, 25ml cationic water (CW) was added. Both vials were sealed and then placed in a thermostatically controlled water bath to be equilibrated to the test temperature. After reaching the test temperature, the CW was quickly transferred into the AW test vial to mix with AW as the test brine. After mixing, the test vial was kept undisturbed at the test temperature in the water bath. At end of a 22 hour test, 1 0m1 of the supernatant test brine was sampled into 20ml of a stabilization solution. [Ba2+] in the sample was then analyzed by an ICP (interactively coupled plasma), and DETPMP was analyzed by a molybdenum blue based colorimetric method. The brine pH at end of each test was also measured. A blank test in which no inhibitor was included and a control test where AW was replaced with de-ionized water were also conducted along with the inhibited tests. Barium sulfate inhibition efficiency was calculated by: %BaSO4 Inhibition = 100 X

([Ba2+]lnhibited – [Ba2+]blank) ([Ba2+]control – [Ba2+]blank)

(1)

The DETPMP adsorption on the precipitated barium sulfate crystals was calculated by: % DETPMP Adsorbed = 100

X

([DETPMP] initial – [DETPMP] final) [DETPMP]

initial

(2)

The barium sulfate precipitation rate/kinetics experiments were carried out in the same manner, except that 2ml sample of the supernatant test brine was taken into 18ml of a stabilization solution at 0.5 hour, 1 hour, 2 hours, 3 hours and 22 hours respectively. The samples were analyzed by an ICP for [Ba2+]. Percentage of barium sulfate precipitated at each sampling time was calculated by: %BaSO4 Precipitated from Solution = 100 X

([Ba2+] control – [Ba2+]sample) ([Ba2+]control – [Ba2+]blank)

(3)

DETPMP Adsorption on Barium Sulfate Reagent Powder 50ml glass crimp top glass vials were used as test containers. Into each vial, 1.00 gram of barium sulfate reagent powder was added. 50ml of the test brine was then added and dosed with DETPMP to a pre-determined test concentration. The vial was then sealed and placed in a thermostatically controlled water bath to be equilibrated to the test temperature. At end of a 22 hour test, 5ml of the supernatant test brine was sampled into 5ml of a stabilization solution and analyzed by a molybdenum blue based colorimetric method for the DETPMP concentration. DETPMP adsorption on the barium sulfate powder was calculated by: DETPMP Adsorbed (mg/g-BaSO4) = 50/1000 ([DETPMP] initial – [DETPMP] final)

(4)

RESULTS AND DISCUSSION This study focuses on the temperature effect on DETPMP barium sulfate scale inhibition efficiency, and for which the main body of the experimental data was generated. Additional supporting data was also obtained to help understand/interpret the inhibition data, which

154

includes DETPMP adsorption on barium sulfate reagent powder, DETPMP adsorption on barium sulfate scale crystals in the test brines, and kinetic rate of barium sulfate precipitation.

DETPMP Inhibition of Barium Sulfate Scale Formation Tables 3 to 6 present the results of DETPMP barium sulfate scale inhibition in the oilfield brines (see Table 1) at different temperatures and/or altered brine chemistries, namely, at 23°C, 60°C and 90°C, in presence or absence of hydrogen carbonate, and at two brine pHs (5.5 and 7.0). The results are illustrated in Figures 1 to 5. It should be stressed that all the brines at their corresponding test temperatures have the same supersaturation (SI = 2.34) and amount of equilibrium barium sulfate precipitation (97.4 mg/L). Table 3 shows that the DETPMP inhibition efficiency of barium sulfate precipitation is markedly higher at 23°C than either that at 60°C or 95°C, indicating that DETPMP performs better at 23°C than at elevated temperatures. For example, at 20ppm concentration, the barium sulfate inhibition of DETPMP is 96% at 23°C, but only in the 50% range both at 60°C and at 95°C. However, between 60°C and 95°C, the DETPMP inhibition performance hardly changed. The results are graphically presented in Figure 1. Table 3. Barium sulfate inhibition efficiency of DEMTPMP vs. temperature. No HCO3- in the test; brine pH 5.5 DETPMP (ppm)

23°C

5 10 20

4.1 23.1 96.1

5 ppm

%BaSO4 Inhibition 60°C 3.1 14.5 50.3

95°C 2.9 14.6 56.5

10 ppm 20 ppm Concentration of DETPMP

Figure 1. Barium sulfate inhibition efficiency of DETPMP vs. temperature. No HCO3- in the test brines; brine pH 5.5.

155

The results obtained in absence of hydrogen carbonate in the brines was confirmed with the results in presence of hydrogen carbonate, as shown in Table 4 and Figure 2. Again, at 23°C DETPMP performs significantly better against barium sulfate precipitation than at elevated temperatures. As an example, at l0ppm, it shows more than 90% barium sulfate inhibition at 23°C, but less than 50% at either 60°C or 95°C. It should be noted that at 20ppm, because DETPMP achieved almost full (90%+) inhibition against barium sulfate precipitation at all the three temperatures, there is little room for differentiating its performance versus temperature. Interestingly, from 60°C to 95°C, there appears to be a small enhancement (up to 6%) in DETPMP inhibition efficiency, but it could be within the experimental error. Nonetheless, the results both in absence and presence of hydrogen carbonate in the brines suggest that, although DETPMP inhibition efficiency of barium sulfate clearly decreases as temperature increases from 23°C to 60°C, there is little change in DETPMP performance as temperature increases further to 95°C. In other words, the sensitivity of DETPMP to temperature may be restricted to only low and medium temperature ranges.

Table 4. Barium sulfate inhibition efficiency of DETPMP vs. temperature. With HCO3- in the test brines; brine pH5.5. DETPMP (ppm) 5 10 20

23°C 27.1 95.1 95.6

%BaSO4 Inhibition 60°C 21.6 43.9 93.0

95°C 23.1 49.7 95.7

Figure 2. Barium sulfate inhibition efficiency of DETPMP vs. temperature. With HCO3- in the test brines; brine pH 5.5.

156

Further experiments were conducted with brine pH7.0 at 23°C and 95°C. The results are shown in Table 5 as well as graphed in Figure 3, which confirm that DETPMP is more efficient at 23°C than at elevated temperatures.

Table 5. Barium sulfate inhibition efficiency of DETPMP vs. temperature. no HCO3- in the test brines; brine pH7.0. DETPMP (ppm) 5 10 20

23°C

%BaSO4 Inhibition

79.6 95.0 95.6

95°C 23.2 42.8 97.3

5 ppm 10 ppm 20 ppm Concentration of DETPMP Figure 3. Barium sulfate inhibition efficiency of DETPMP vs. temperature. No HCO3- in the test brines; brine pH 7.0.

Since the experiments on temperature effect have been carried out at both brine pH5.5 and brine pH 7.0, it was of some interest to compare inhibition performance at the two brine pHs. The results in Table 3 and Table 5 are therefore re-organized to appear in Table 6, and are illustrated in Figure 4 (23°C) and Figure 5 (95°C) respectively. The observed pH sensitivity of DETPMP is explained by the increased dissociation of the phosphonic acid groups, thus increased inhibitor activity, at a more neutral pH.6

157

Table 6. Barium sulfate inhibition efficiency of DETPMP vs. brine pH. No HCO3- in the test brines. DETPMP (ppm) 5 10 20

%BaSO4 Inhibition

23°C

95°C

pH5.5

pH7.0

pH5.5

pH7.0

4.1 23.7 96.1

79.6 95.0 95.6

2.9 14.6 56.5

23.2 42.8 97.3

5 ppm

10 ppm

Concentration

of

20 ppm DETPMP

Figure 4. Barium sulfate inhibition efficiency of DETPMP vs. brine pH. 23°C; no HCO3- in the test brines.

5 ppm

10 pm

Concentration

of

20 ppm DETPMP

Figure 5. Barium sulfate inhibition efficiency of DETPMP vs. brine pH. 95°C; no HCO3- in the test brines.

158

DETPMP Adsorption on Barium Sulfate Surface It is generally believed that barium sulfate scale inhibition is achieved by two main mechanisms: scale crystal nucleation inhibition and crystal growth retardation,7 although dispersion is also recognized.8 Phosphonic acids such as DETPMP inhibit scale formation mainly via crystal growth retardation by adsorbing on the active growth sites.7 In this study, experiments of DETPMP adsorption on barium sulfate reagent powder were performed at the three temperatures corresponding to the inhibition tests. The results are presented in Table 7 and plotted in Figure 6. It is found that DETPMP adsorption on barium sulfate continuously increases as temperature elevates from 23°C to 60°C and then to 95°C; the adsorption also increases as initial DETPMP concentration in a test brine increases from 5ppm to l0ppm and then to 20ppm. The enhanced adsorption at increased DETPMP concentrations correlates with increased barium sulfate inhibition efficiency at higher DETPMP concentrations (see Tables 3 to 5). However, it should be noted that as DETPMP adsorption on barium sulfate continuously increases from 23°C to 60°C then to 95°C, its barium sulfate inhibition eficiency actually decreases from 23°C to 60°C then levels out between 60°C and 95°C. This suggests that, although DETPMP adsorption on barium sulfate surface is a necessary part of inhibition process, the magnitude of the inhibitor adsorption may not necessarily determine its degree of barium sulfate inhibition. This is probably true especially when other factors such as temperature comes into play, which may also affect the precipitation kinetics of barium sulfate and the interactions between inhibitor and scale nuclei and/or scale crystals. This postulation is given support with the following experimental data appearing in Tables 8 and 9, as well as Figures 7 to 9.

Table 7. DETPMP adsorption on barium sulfate reagent powder. Effect of temperature; no HCO3- in the test brines. DETPMP* (ppm)

5 10

23°C 0.182 0.402

DETPMP Adsorption(mg/g) 60°C 0.196 0.427

20 0.646 * Concentrations are as initially dosed in the test brine.

0.782

95°C 0.211 0.441 0.898

Table 8 (as well as Figures 7 and 8) shows that, in the barium sulfate inhibition tests, the percentage of DETPMP lost to the scale crystals formed at end of 22 hour inhibition tests is lower at higher initial concentrations, but broadly increases with temperature increase. This temperature trend agrees with the results of DETPMP adsorption on barium sulfate reagent powder. It should be noted that, whereas DETPMP adsorption increases from 23°C to 60°C its inhibition eficiency against barium sulfate actually decreases (see Table 3 and Table 4). This again indicates that the level of DETPMP surface adsorption is not the determining factor in the inhibitor performance.

159

Initial Concentration of DETPMP (ppm) Figure 6. DETPMP adsorption on barium sulfate reagent powder. Effect of temperature; no HCO3- in the test brines.

Table 8. DETPMP adsorption on precipitated barium sulfate crystals. Effect of temperature. no HCO323°C 60°C

DETPMP (ppm) 5 10 20

89.0 68.2 0.0

0

90.5 14.3 12.2

5

10

%DETPMP Adsorbed 95°C 89.9 16.4 15.4

15

23°C

W/ HCO3-

60°C

95°C

85.3 56.9 0.0

86.3 60.5 2.5

84.0 46.8 1.2

20

25

Concentration of DETPMP (ppm) Figure 7. DETPMP adsorption on precipitated barium sulfate crystals. Effect of temperature; no HCO3- in the test brines.

160

0

5

10

15

20

25

Concentration of DETPMP (ppm) Figure 8. DETPMP adsorption on precipitated barium sulfate crystals. Effect of temperature; with HCO3in the test brines.

Precipitation Rate of Barium Sulfate One factor that may influence barium sulfate inhibition of DETPMP is kinetics of barium sulfate precipitation. Conventionally it is assumed that the degree of supersaturation of the scaling ions in a solution equates the driving force for scale formation. However, our latest experimental data reveals that, for a constant supersaturation, the kinetic rate of barium sulfate precipitation is higher at elevated temperatures as demonstrated in Table 9 and Figure 9. It shows that, after 30 minutes, 96% of supersaturated barium sulfate has precipitated out of solution at 95°C, only 89% at 23°C. Similarly, after only 2 hours, 100% of supersaturated barium sulfate has precipitated out of solution at 95°C, compared to less than 94% at 23°C. The faster rate of barium sulfate precipitation may have created an extra driving force for scale crystal nucleation and growth in addition to that by the supersaturation ratio, which in turn may create extra difficulties in inhibiting the scale formation. However, no explanation is readily available for why temperature effect on DETPMP inhibition function apparently is rather small between 95°C and 60°C, while it is significant from 60°C to 23°C

Table 9. Barium sulfate precipitation rate vs. temperature, no HCO3- in the test brines; no DETPMP in the test brines. Time (hour) 0.5 1.0 2.0 3.0

%BaSO4 23°C 88.9 90.5 93.5 94.6

Precipitated 95°C 95.8 98.4 100 100

161

0

0.5

1

1.5

2

2.5

3

3.5

Time (hour) Figure 9. Barium sulfate precipitation rate vs. temperature. No HCO3- in the test brines; no DETPMP in the test brines.

DETPMP (ppm) 2.5 5.0 7.5 10 25

%BaSO4 Inhibition 25°C 79.1 91.0 93.5 93.9 93.8

95°C N/A N/A N/A 0 0

DETPMP Inhibition of Barium Sulfate in 2m NaC1 Brine Additional experiments were carried out in a simple brine system, 2m NaC1 brine. The tests were conducted at 25°C and 95°C with the brines shown in Table 1, where sulfate ion concentration was increased from 25°C to 95°C so that the supersaturation and amount of equilibrium precipitation of barium sulfate remain the same between the two temperatures. The inhibition data is presented in Table 10 below. The difference in DETPMP inhibition efficiency at the two temperatures is more striking in the NaC1 brine. At 25°C, DETPMP was rather efficient against barium sulfate scaling, showing 79% inhibition at only 2.5ppm concentration. But at 95°C, DETPMP did not show any barium inhibition at concentrations as high as 25ppm.

CONCLUSIONS Laboratory experiments have been carried out to investigate the effect of temperature on the performance of diethylenetriamine penta (methylene phosphonic acid), DETPMP, against barium sulfate scale formation.

162

By maintaining supersaturation and amount of equilibrium precipitation of barium sulfate constant in the test brines, it is found that as temperature increases from 25°C to 60°C, barium sulfate inhibition efficiency of DETPMP decreases considerably. However, as temperature increases further from 60°C to 95°C, the performance of DETPMP remains almost constant. The previous observation that percentage inhibition of barium sulfate by DETPMP in a given brine decreases at lower temperatures3 is certainly the result of increased scaling tendency of barium sulfate in the brine at cooler temperatures. DETPMP adsorption on barium sulfate reagent powder at different temperatures indicates that the level of DETPMP adsorption on barium sulfate continuously increases as temperature rises, which could not explain why DETPMP inhibition efficiency decrease from 23°C to 60°C. Instead, it suggests that the level of inhibitor surface adsorption may not be the key factor in the nucleation inhibition or crystal growth retardation of barium sulfate. It is found that for a given supersaturation and a given amount of equilibrium precipitation, the rate of barium sulfate precipitation increases at elevated temperatures. This increased rate of precipitation at higher temperatures may have created additional driving force for barium sulfate scale formation, apart from that contributed by the degree of its supersaturation.

ACKNOWLEDGMENT The author would like to thank the management of Baker Petrolite for encouraging and permitting the publication of this paper.

REFERENCES 1. C.C. Templeton, Solubility of barium sulfate in sodium chloride solutions from 25°C to 95°C, J. Chem.&Eng. Data, 5:514 (1960). 2. C.W. Blount, Barite solubilities and thermodynamic quantities up to 300°C and 1400 bars, American Mineralogist, 62:942 (1977). 3. M.D. Yuan, M. Anderson, and E. Jamieson, Investigation and improvement of BaSO4 scale inhibition tests, Paper SPE 37304. The SPE International Symposium on Oilfield Chemistry (1997). 4. Rice University Brine Chemistry Consortium Semi-Annual Technical Meeting Report, Houston (Oct. 1998). 5. M. Tomson, ScaleSoft Pitzer verl.0, Rice University, Houston (1999). 6. D.W. Griffiths, S.D. Roberts, and S-T, Liu, Inhibition of calcium sulfate dihydrate crystal growth by phosphonic acids – influence of inhibitor structure and solution pH, Paper SPE 7862. The SPE International Symposium on Oilfield and Geothermal Chemistry (1 979). 7. L.S. Boak, G.M. Gordon, and K.S. Sorbie, The influence of divalent cations on the performance of BaSO4 scale inhibitor species, Paper SPE 50771. The SPE International Symposium on Oilfield Chemistry (1999). 8. M.D. Yuan, E. Jamieson, and P. Hammonds, Investigation of scaling and inhibition mechanisms and the influencing factors in static and dynamic inhibition tests, Paper 98067. NACE Corrosion/98 (1998).

163

This page intentionally left blank.

THE ROLE OF CALCIUM PHOSPHINO-POLYCARBOXYLATE COMPLEXATION IN INHIBITING BASO4 PRECIPITATION FROM BRINE

J. J. Xiao, A. T. Kan and M. B. Tomson Brine Chemistry Consortium Energy and Environmental Systems Institute Rice University MS 5 19 Houston, TX 77251-1892

ABSTRACT The kinetics of BaSO4 nucleation and its inhibition with phosphino-polycarboxylic acid (PPCA) has been studied. The induction period of BaSO4 nucleation under a wide range of ionic strength and temperature has been measured through a turbidity method. The presence of Ca2+ does not show a significant effect on the induction period of BaSO4 nucleation. The presence of PPCA increases the induction period of BaSO4 nucleation while the extent of the prolongation depends on the solution pH and Ca2+ concentration. Only the dissociated fraction and the metal complexed fraction of PPCA are effective to inhibit BaSO4 nucleation while the protonated fraction has no significant effect at any conditions. In a low supersaturated solution, the calcium complexed fraction of PPCA is more effective than the dissociated fraction. While in a highly supersaturated solution, the calcium complexed fraction of PPCA .is similarly effective as the dissociated fraction. Temperature has no significant effect on the efficiency of the calcium complexed fraction of PPCA and has only slight effect on that of the dissociated fraction of PPCA. The overall efficiency of PPCA to inhibit BaSO4 nucleation is proportional to the total concentration of Ca2+. Ca2+ enhances the overall efficiency of PPCA through decreasing the fraction of protonated PPCA and forming Ca-PPCA complexes. Ca2+ might play the role of a bridge connecting the crystal sites with exposed SO42- and PPCA functional groups so that the adsorption of PPCA on the active sites of BaSO4 crystal surfaces is enhanced.

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers, New York, 2000

165

INTRODUCTION Barium sulfate, BaSO4, is a troublesome scale found in oil/gas wells and geothermal wells. Once formed, BaSO4 is difficult to remove. Furthermore, it often enriches radium by coprecipitation and results in radiation problems.1-20 Therefore, the study of precipitation of BaSO4 from supersaturated solution is of both scientific and practical importance. Generally, phosphonates and polymers are good inhibitors of mineral nucleation and are widely used in industries. The efficiency of these inhibitors has been reported to be enhanced with divalent or multivalent metal ions in solution.14-15 Since Ca2+ is the predominant divalent metal ions in the most oil fields and PPCA is a common scale inhibitor, the object of this paper is to quantitatively evaluate the effect of Ca2+ on the efficiency of PPCA in inhibiting BaSO4 nucleation.

BasO4 Nucleation In this paper, the induction period (tind) of BaSO4 nucleation is defined as the time elapsing between the mixing of two solutions and the onset of an increase in turbidity reading, and is used as an index to evaluate BaSO4 nucleation kinetics and inhibitors' efficiency. The induction period for each nucleation condition can be determined from a turbidity-time curve of a supersaturated solution of barium sulfate, as Figure 1 shows. According to the classical theory of nucleation, this induction period can be related to the crystal/solution interfacial energy (σ), the heterogeneity (fh) of BaSO4 nucleation and the supersaturation index (SI) in solution as following:18 (1) where t0 is a constant, β is a geometric shape factor ( = 4A3/27V2 where A is the surface area and V is the crystal volume), Vm is the molar volume of the crystals (52.1 cm3/mol for Barite), NA= 6.02x1023 /mol, R = 8.314 J/mol/K, SI = log10({Ba2+} {SO42- }/Ksp) where { } means activity and Ksp is the solubility product of BaSO4 at temperature T. As equation 1 shows, the induction period of BaSO4 precipitation can be prolonged with inhibitors, such as PPCA, through modifying the geometric shape (β), interfacial energy (σ) etc. so that the mineral scale formation can be inhibited or retarded under given conditions.

PPCA Solution Chemistry PPCA is a polyacrylic acid (PAA) with imbedded phosphino- groups: (-CH2CHCOOH)-xP(OOH)(-CH2CHCOOH)-y. The solution chemistry of PPCA has been investigated by Xiao, et al21 under a wide range of ionic strength and temperature. With electrostatic theory, the acid/base equilibrium of PPCA can be expressed as:

HA with definition of:

166

H+ + A¯

(2)

KH =

[H+][A¯ ] [HA]

(3)

and the electrostatic relationship of: pKH = pKH,int + f • θu

(4)

where HA and A- are protonated and dissociated functional groups of PPCA, separately; pK H, int is the negative logarithm of K H, int, the intrinsic constant of PPCA proton dissociation reaction; f is a coefficient accounting for the electrostatic effect from the neighboring dissociated groups of the same PPCA molecule; and θu, is the dissociated fraction of PPCA functional groups. Both pKH, int and f have been obtained with acid/base titrations and can be expressed as:21

pKH,int = 4.798 – 0.954 • f = 2.757––1.056

+ 0.246 • I -

+ 0.220

187.8 T

(5)

—————

•I

(6)

where I and T are ionic strength (m) and temperature (K), respectively. Similarly, the calcium-PPCA complexation can be expressed as: Ca(A–A) ↔ Ca2+ + (A –A)2-

(7)

with definition of KCa =

[Ca2+][(A–A)2+] [Ca(A–A)]

(8)

and the electrostatic relationship of: pKCa = PKCa,int + 2f• θ u

(9)

where (A-A) is an arbitrary combination of two functional groups and KCa, int is the intrinsic constant of the dissociation reaction. All other symbols have the same physical meaning as in the acid/base equilibrium above. The value of pKCa, int was obtained by acid/base titration in the presence of Ca2+ under different conditions and can be expressed as:21 pKCa,int = 3.928 — 2.63 1 •

1099.4 + 0.738 • I – ——— T

(10)

In this study, the apparent efficiency of PPCA in inhibiting BaSO4 nucleation is separated into specific efficiency of three parts: protonated fraction, dissociated fraction

167

and calcium complexed fraction via equations.2-10 The objective is to assess the effect of Ca2+ on enhancing the efficiency of PPCA to inhibit BaSO4 nucleation.

0

20

40

60 80 Time(sec)

100

120

140

Figure 1. Determination of the induction period of BaSO4 nucleation. Initial condition: [Ba2+] = [SO42-] = 2.79 mm; [NaC1] = 1.0 m; pH ~ 6; T = 298.2 °K; I = 1.01; SI = 2.92; stirring at 350 rpm.

EXPERIMENTAL Chemicals BaCI2•2H2O (granular, AR®), Na2SO4 (granular, AR®), NaCl (crystal, reagent) and CaCI2•2H2O (granular, reagent) were from Fisher Scientifics Inc. PhosphinoPolycarboxylic acid (PPCA) is the active component of Bellasol®S29, used as received from FMC. All solutions were prepared with deionized water and filtered through 0.1µm microfilter.

Analytical Techniques PPCA was analyzed by solid phase separation-UV digestion-ascorbic acid/phosphomolybdic acid blue method. Samples containing PPCA were acidified by HCI to pH 2. A C18 cartridge (Waters Corporation) was conditioned with methanol, water and 0.01 N HCI in sequence. An acidified sample was then injected dropwise through the preconditioned C18 cartridge by a syringe, followed by about 5ml deionized water. The cartridge was then backward eluted dropwise with a 0.01 m borax solution (pH 9.18). This eluate was collected, neutralized to pH 5 and followed by an addition of potassium persulfate (0.1 gram c.a.) and UV radiation oxidation by the HACH Method. The UV digested solution was then cooled down to room temperature, and phosphorus was analyzed by the ascorbic acid method.22 The total PPCA concentration in the product Bellasol® S29 was determined by COD method.22 Ca2+ concentration was analyzed either by EDTA titration or ICP analysis.23 Ba2+ concentration was analyzed by ICP method.

168

Nucleation Procedures Two solutions were made for each nucleation experiment: one with crystallizing cations Ba2+ with/without Ca2+ in 1m NaCl solution; another with crystallizing anions SO42- with/without inhibitors in 1m NaCl solution. These two solutions were filtered through 0.1µm microfilter and preheated to the set temperature before they were used. The two solutions (10 ml each) were then rapidly added into a turbidity cell (30 ml) and mixed under continuous stirring at 350 rpm by a Teflon-coated magnetic stirring bar. The cell was kept at the set temperature by circulating water from a heated water bath through a stainless steel coil. The turbidity of the solution was tracked by a ratio/XR turbidimeter (HACH Company) and the data acquisition was accomplished via a digital multimeter (Radio Shack) connected with a PC.

RESULTS AND DISCUSSION All experimental data to be analyzed in this paper are listed in Table 1. The ranges of condition are: total PPCA concentration (Cp): 0 ~ 27 ppm; total Ca2+ concentration (CCa): 0 0.2 m; temperature (T): 298 ~ 363 K (25 ~ 90 °C); supersaturation index (SI): 1.0 ~ 4.0; ionic strength (I): ~ 1.0 m; and pH: 4.0 ~ 8.5. Data in Table 1 were sorted in order of Cp followed by T and SI. BaSO4 Nucleation in the Absence of Ca2+/PPCA For BaSO4 nucleation from 1.0 m NaCl solution at room temperature, the plot of log10(tind) versus Sr-2 produces two segments of straight line at low supersaturation and high supersaturation, respectively, as Figure 2 shows. The segment of straight line at high supersaturation (SI = 2.9 ~ 4.0) gives specific interfacial energy of σ 79 mJ/m2 from equation 1 with a fh of 0.4.8 This is comparative to 38-150mJ/m2 in literature8 at low ionic strength (IS<0.03m) and close to the value9 of 73 mJ/m2 in pure solution and 76mJ/m2 in 0.5m NaCl when [Ba2+] = [SO42-] > 1 mM.

0.0

0.1

0.2

0.3

sr2

0.4

0.5

0.6

0.7

Figure 2. Relationship of the induction period of BaSO4 nucleation with the supersaturation index – classical nucleation theory. Initial condition: [NaCI] = 1.0 m; I ~ 1.0 m; T = 298.2 °K; pH ~ 6; stirring at 350 rpm.

169

170

Table 1. Experiemntal data of BaSO4 nucleation under different conditions (Ref. to He at el18) Exptl #

Cp (ppm)

CCa (m)

T (K)

NaC1 (m)

BaSO4 (mm)

I (m)

pH

SI

log10(tind) Sec)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.003 0. 100 0.000 0.000 0.000 0.000 0.000 0.000 0.000

298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2

1.000 1.000 1.000 0.003 0.002 0.002 1.000 1.000 1.000 0.002 1.000 1.000 0.001 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

9.200 9.200 7.440 1.400 1.200 1.000 5.580 5.580 5.580 0.800 4.650 4.650 0.600 3.720 3.720 3.720 2.790 2.790 2.790 2.790 2.790 2.790 1.860 1.860 1.860 1.400 1.400 1.400 1.160

1.037 1.037 1.030 0.008 0.007 0.006 1.022 1.022 1.022 0.005 1.019 1.019 0.004 1.015 1.015 1.015 1.011 1.011 1.011 1.011 1.020 1.311 1.007 1.007 1.007 1.006 1.006 1.006 1.005

6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00

3.990 3.990 3.810 3.780 3.680 3.560 3.547 3.547 3.547 3.410 3.289 3.289 3.210 3.196 3.196 3.196 2.926 2.926 2.926 2.926 2.900 2.770 2.590 2.590 2.590 2.245 2.245 2.245 2.130

0.000 -0.020 0.230 0.110 0.270 0.430 0.530 0.477 0.661 0.760 0.720 0.732 1.120 0.860 0.954 0.845 1.030 1.027 1.279 1.301 1.278 1.301 1.279 1.643 1.301 1.520 1.778 1.523 1.778

θu

θH

0.880 0.880 0.880 0.632 0.630 0.627 0.879 0.879 0.879 0.624 0.879 0.879 0.621 0.879 0.879 0.000 0.878 0.878 0.878 0.878 0.880 0.960 0.878 0.878 0.878 0.878 0.878 0.878 0.878

0.120 0.120 0.120 0.368 0.370 0.373 0.121 0.121 0.121 0.376 0.121 0.121 0.379 0.121 0.121 1.000 0.122 0.122 0.122 0.122 0.120 0.040 0.122 0.122 0.122 0.122 0.122 0.122 0.122

θCa 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

171

Exptl #

Cp (ppm)

CCa (m)

T (K)

NaCl (m)

BaSO4 (mm)

I (m)

pH

30 31 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 60 61

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 308.2 318.2 318.2 323.2 323.2 323.2 323.2 323.2 323.2 323.2 323.2 323.2 323.2 323.2 323.2

1.000 0.500 1.000 0.000 0.010 2.000 3.000 6.000 1.000 1.000 0.100 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

1.160 0.660 0.940 0.120 0.170 1.340 1.600 1.830 0.930 0.930 0.220 0.750 0.700 0.700 0.700 0.470 0.470 1.060 2.790 1.160 12.800 11.520 10.240 8.960 7.680 6.400 5.120 3.840 3.200 2.560 1.920 1.280

1.005 0.503 1.004 0.001 0.011 2.005 3.006 6.007 1.004 1.004 0.101 1.003 1.003 1.003 1.003 1.002 1.002 1.004 1.011 1.005 1.051 1.046 1.041 1.036 1.031 1.026 1.020 1.015 1.013 1.010 1.008 1.005

6.10 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.29 7.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00

SI 2.130 2.010 2.010 2.000 2.000 2.000 2.000 2.000 1.993 1.993 1.990 1.790 1.750 1.746 1.746 1.291 1.290 2.000 2.640 1.850 3.990 3.910 3.800 3.690 3.560 3.400 3.200 2.950 2.800 2.600 2.250 2.000

log10(tind) (Sec) 1.954 1.980 2.280 2.060 2.180 1.980 1.940 2.180 2.267 2.079 2.080 2.480 2.530 2.559 2.633 3.343 3.340 1.800 0.602 1.820 -0.220 -0.050 0.000 0.080 0.200 0.260 0.510 0.630 0.690 0.720 1.090 1.430

θu 0.894 0.818 0.878 0.609 0.636 0.928 0.946 0.950 0.878 0.878 0.707 0.878 0.878 0.878 0.878 0.877 0.877 0.874 0.916 0.977 0.873 0.873 0.872 0.872 0.872 0.871 0.871 0.870 0.870 0.870 0.870 0.869

θH 0.106 0.182 0.122 0.391 0.364 0.072 0.054 0.050 0.122 0.122 0.293 0.122 0.122 0.122 0.122 0.123 0.123 0.126 0.084 0.023 0.127 0.127 0.128 0.128 0.128 0.129 0.129 0.130 0.130 0.130 0.130 0.131

θCa 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

172 Exptl #

Cp (ppm)

CCa (m)

T (K)

NaCl (m)

BaSO4 (mm)

I (m)

62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.160 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

323.2 323.2 323.2 323.2 323.2 323.2 323.2 333.2 339.2 339.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 343.2 353.2 356.2 356.2 363.2 363.2 363.2

1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1.000 1.000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000 1 .000

1.150 1.020 0.900 0.770 0.640 0.510 0.280 1.480 2.790 1.160 16.380 13.100 9.830 8.190 6.550 4.910 2.790 3.280 2.460 1.640 1.470 1.310 1.150 0.980 0.820 0.660 1.810 2.790 1.160 9.790 7.830 5.870

1.005 1.004 1.004 1.003 1.003 1.002 1.001 1.006 1.01 1 1.005 1.065 1.052 1.039 1.033 1.026 1.020 1.49 1 1.013 1.010 1.007 1.006 1.005 1.005 1.004 1.003 1.003 1.007 1.011 1.005 1.039 1.031 1.023

pH 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.67 7 .00 6.00 6.00 6.00 6 .00 6.00 6.00 6.50 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.70 7.00 6.00 6.00 6.00

SI 1.900 1.800 1.690 1.550 1.400 1.200 0.950 2.000 2.430 1.640 3.970 3.810 3.560 3.400 3.210 2.950 2.700 2.590 2.350 2.000 1.910 1.790 1.690 1.560 1.290 1.200 2.000 2.200 1.510 3.290 3.210 2.950

log10(tind) (Sec)

θu

1.650 1.790 2.070 2.220 2.610 3.080 3.860 1.240 0.477 1.820 -0.700 -0.460 -0.300 -0.190 0.210 0.460 0.699 0.560 0.790 1.030 1.320 1.490 1.830 2.200 2.650 3.070 0.900 0.201 1.845 -0.200 0.020 0.280

0.869 0.869 0.869 0.869 0.869 0.869 0.869 0.866 0.953 0.975 0.869 0.868 0.866 0.866 0.865 0.865 0.572 0.864 0.864 0.864 0.864 0.863 0.863 0.863 0.863 0.863 0.861 0.953 0.974 0.861 0.860 0.860

θH 0.131 0.131 0.131 0.131 0.131 0.131 0.131 0.134 0.047 0.025 0.131 0.132 0.134 0.134 0.135 0.135 0.005 0.136 0.136 0.136 0.136 0.137 0.137 0.137 0.137 0.137 0.139 0.047 0.026 0.139 0.140 0.140

θCa 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.423 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

173

Exptl #

Cp (ppm)

CCa (m)

T (K)

NaCl (m)

BaSO4 (mm)

I (m)

pH

SI

94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.540 0.540 1.000 1.078 1.078 1.080 1.080 1.080 1.080 1.080 1.080 1.080 1.080 1.080 1.080 1.080 2.500 2.684 3.586 5.000 7.157

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.100 0.160 0.200 0.000 0.000 0.000 0.020 0.050 0.100 0.000 0.100 0.000 0.100 0.000 0.100 0.160 0.100 0.000 0.160 0.100

363.2 363.2 363.2 363.2 363.2 363.2 363.2 363.2 363.2 363.2 363.2 298.2 298.2 343.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 318.2 318.2 339.2 339.2 356.2 356.2 343.2 298.2 298.2 343.2 298.2

1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

3.920 1.960 1.760 1.570 1.270 1.180 0.980 0.780 0.590 0.240 0.200 1.160 1.160 2.790 1.160 1.160 1.160 1.160 1.160 1.160 1.160 1.160 1.160 1.160 1.160 1.160 1.160 2.790 2.790 2.790 2.790 2.790

1.016 1.008 1.007 1.006 1.005 1.005 1.004 1.003 1.002 1.001 1.001 1.005 1.305 1.491 1.605 1.005 1.005 1.005 1.065 1.155 1.305 1.005 1.305 1.005 1.305 1.005 1.305 1.491 1.311 1.011 1.491 1.311

6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.26 6.61 6.50 6.21 4.75 7.80 6.31 6.27 6.14 6.11 6.21 6.24 6.98 6.90 7.26 6.04 6.50 6.44 5.00 6.50 6.48

2.600 2.000 1.910 1.810 1.690 1.560 1.290 1.190 0.950 0.790 0.600 2.134 1.977 2.700 1.888 2.134 2.134 2.130 2.099 2.049 1.977 1.850 1.760 1.640 1.550 1.510 1.430 2.700 2.770 2.926 2.700 2.770

log10(tind) (Sec) 0.490 0.730 0.940 1.050 1.190 1.530 1.770 2.230 2.700 3.220 3.680 2.243 2.602 0.699 3.954 2.176 2.845 2.633 3.301 3.602 3.699 3.176 3.431 3.176 3.298 3.176 3.301 1.079 1.776 1.477 1.279 2.697

θu 0.859 0.858 0.858 0.858 0.858 0.858 0.858 0.858 0.858 0.858 0.858 0.917 0.674 0.572 0.653 0.590 0.923 0.750 0.700 0.670 0.670 0.974 0.595 0.985 0.564 0.674 0.572 0.572 0.674 0.657 0.572 0.674

θH 0.141 0.142 0.142 0.142 0.142 0.142 0.142 0.142 0.142 0.142 0.142 0.083 0.007 0.005 0.013 0.410 0.077 0.030 0.028 0.022 0.022 0.026 0.003 0.015 0.018 0.007 0.005 0.005 0.011 0.343 0.005 0.010

θCa 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.319 0.423 0.334 0.000 0.000 0.220 0.273 0.308 0.308 0.000 0.403 0.000 0.419 0.319 0.423 0.423 0.316 0.000 0.423 0.316

174 Exptl #

Cp (ppm)

CCa (m)

T (K)

NaCl (m)

BaSO4 (mm)

I (m)

126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150

7.200 7.500 10.000 14.219 26.831 26.844 26.849 26.850 26.872 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 27.000 50.000

0.100 0.160 0.160 0.100 0.050 0.100 0.000 0.020 0.200 0.000 0.000 0.003 0.000 0.020 0.050 0.080 0.100 0.000 0.020 0.000 0.020 0.000 0.000 0.020 0.000

356.2 343.2 343.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 298.2 318.2 318.2 339.2 339.2 356.2 356.2 356.2 298.2

1.000 1.000 1.000 1.000 1,000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1,000 1.000 1.000 1.000 1.000 1.000 1.000

2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790 2.790

1.311 1.491 1.491 1.311 1.161 1.311 1.011 1.071 1.611 1.011 1.011 1.020 1.011 1.071 1.161 1.251 1.311 1.011 1.071 1.011 1.071 1.011 1.011 1.071 1.011

Note:

Cp - Total concentration of PPCA CCa - Total concentration of Ca2+

pH 6.56 6.50 6.50 6.28 6.10 6.08 4.00 5.86 7.82 5.76 8.41 6.15 5.82 5.86 5.76 5.84 5.75 6.21 6.04 6.67 6.74 6.86 7.77 6.93 3.90

SI 2.220 2.700 2.700 2.770 2.841 2.770 2.926 2.891 2.683 2.926 2.926 2.921 2.920 2.890 2.840 2.796 2.770 2.640 2.620 2.430 2.410 2.300 2.300 2.280 2.920

log10(tind) (Sec) 1.778 1.447 1.763 3.001 4.300 4.699 2.580 3.844 4.821 3.699 3.778 3.813 3.699 3.845 4.230 4.543 4.699 3.778 4.230 3.903 4.903 4.041 4.000 4.699 2.234

θu 0.567 0.572 0.572 0.673 0.700 0.671 0.387 0.738 0.657 0.834 0.999 0.843 0.846 0.738 0.691 0.674 0.664 0.905 0.710 0.953 0.684 0.966 0.995 0.658 0.360

θH 0.005 0.005 0.005 0.015 0.030 0.024 0.613 0.071 0.000 0.166 0.001 0.070 0.154 0.071 0.063 0.044 0.049 0.095 0.044 0.047 0.008 0.034 0.005 0.005 0.640

θCa 0.428 0.423 0.423 0.312 0.270 0.306 0.000 0.191 0.343 0.000 0.000 0.086 0.000 0.191 0.246 0.281 0.287 0.000 0.246 0.000 0.308 0.000 0.000 0.337 0.000

However, this segmented linear relationship becomes discontinuous and cannot cover the whole range of SI generally encountered so that their application is limited. For the practical purposes, a continuous fitting equation was derived over the SI range (1.2 to 3.6) at room temperature, as Figure 3 shows: Log 10 (tind 0) = 3.992- 6.224•log 10 (SI)

(11)

Figure 3. Relationship of the induction period of BaSO4 nucleation with the supersaturation index-continuous fitting equation. Initial condition: [NaC1] = 1.0 m; I ~ 1 .0 m; T = 298.2 °K; p ~ H 6; stirring at 350 rpm.

Similarly, the induction period of BaSO4 nucleation under different supersaturation index and temperature can be well expressed in the following fitting equation: Log10(tind0) = -3.030-6.192•log10(SI) +2120.6/T

(12)

BaSO4 Nucleation in the Presence of Ca2+/PPCA It was found that Ca2+ has no significant effect on the induction period of BaSO4 nucleation, especially at high supersaturation, even though the addition of Ca2+ causes a slight decrease in the supersaturation index (SI from 2.93 to 2.77 with Ca2+ concentration from 0 to 0.1 m, see Expt. # 20-22 in Table 1), as Figure 4 shows. In contrast, PPCA has strong effect on the induction period of BaSO4 nucleation at pH > 4. When Ca2+ was added into a solution of supersaturated Ba2+-SO42- solution in the presence of PPCA, it was observed that the induction period of BaSO4 nucleation increases. This prolongation of the induction period might result from the enhancement of Ca2+ on the overall inhibition efficiency of PPCA. However, it might also be a result of the reduction of the supersaturation index of BaSO4 since the ionic strength increases with the addition of calcium salts (see Figure 5 and Expt. 134-142 in Table 1). The quantitative analysis will be given below. It was observed that the prolongation of the induction period of BaSO4 nucleation is proportional to the concentration of PPCA, as well as pH, supersaturation index and temperature. Quantitatively, an overall fitting equation has been obtained from statistical analysis: Log10(tindPPCA) = log10(tind0) + (fu• θu+fCa •θCa} •Cp

(13)

175

Time (Sec) Figure 4. Effect of Ca on BaSO4 nucleation. Initial condition: [Ba2+] = [SO42 - ] = 2.79 mm; [NaC1] = 1.0 m; I = 1.01 m; SI = 2.92; pH ~ 6; T = 298.2 °K; stirring at 350 rpm. 2+

0.00

0.05

0.10

0.15

Cca (m)

0.20

0.25

Figure 5. The effect of Ca2+ on the efficiency of PPCA in inhibition of BaSO4 nucleation. Initial condition: [Ba2+] = [SO42- ] = 2.79 mm; C, = 27.0 ppm; [NaC1] = 1.0 m; I = 1.01 m; SI = 2.92; pH ˜ 6.5; T = 298.2 ºK; stirring at 350 rpm.

176

with fu = -0.365•SI0.917 + 322.7/T

(13a)

and f C a = 23.356 • SI - 5.596

(13b)

where log10(tind0) is the induction period of BaSO4 in the absence of Ca2+ and inhibitors (equation 12). θu and θCa are dissociated and calcium complexed fraction of PPCA in solution, respectively. Cp (mg/L) is the total PPCA concentration and T (K) is temperature. Figure 6 illustrates the measured induction period of BaSO4 nucleation (all 150 data in Table 1) with the predicted values from equation 13.

Measured log10(tind) Figure 6. Comparison between the predicted induction period from equation 13 with the measured induction period (data in table 1). Conditions: SI = 1.1 3.5; T = 298.2 ~ 363.2 °K; C, = 0 ~ 27 ppm; Cc a = 0 ~ 0.2 m; pH = 4 ~ 8; I = 1 ~ 1.6 m; stirring at 350 rpm.

Equation 13 shows the effect of each factor and species of PPCA on inhibition of BaSO4 nucleation. The efficiency of different species of PPCA on BaSO4 nucleation inhibition can be evaluated with the coefficient of each term (fu and fCa) in equation 13. Notice that equation 13 does not contain the protonated PPCA, meaning that the protonated PPCA is not significantly effective in inhibiting BaSO4 nucleation. This is consistent with the observations by other researchers9,10,14,16,18,19 that pH has a strong effect on the efficiency of PPCA. It was also observed in this study that an addition of two drops of 1 N HC1 into a well inhibited supersaturated BaSO4 solution (20 ml) resulted in an immediate precipitation (data is not shown), as expected. The pH dependence is implicitly contained in the species distribution in equation 13. The prolongation of the induction period (in logarithm) of BaSO4 nucleation is linearly proportional to the concentration of the dissociated and calcium complexed PPCA under a given condition. Supersaturation index and temperature have different effects on the efficiency of the dissociated and the calcium complexed PPCA. With an increase in supersaturation index, the efficiency of the dissociated PPCA decreases approximately in a linear manner while that of the calcium complexed PPCA decreases exponentially. Temperature has a slight effect on the efficiency of the dissociated PPCA, but no significant effect on that of the calcium complexed PPCA.

177

Figure 7. Relationship of the efficiency of dissociated and calcium complexed PPCA (fu & fca) with supersaturation index from equation 13. Assumption: T = 343.2 °K.

Figure 7 shows the effect of SI on the efficiency of PPCA species (fu and fCa) in inhibiting BaSO4 nucleation. The calcium complexed PPCA is much more effective than the dissociated PPCA when SI is lower than 2. At high SI, there is no significant difference between the efficiency of the dissociated PPCA and that of the calcium complexed PPCA. Furthermore, when SI increases, the efficiency of both species approaches zero. This is consistent with observations by Fernandez-Diaz, et al.9 Note that Ba2+-PPCA complexation has not been separated out. This should be reasonable when Ba2+ concentration is not high (low SI) or Ca2+/Ba2+ is very large. However, at higher SI (higher Ba2+ concentration) and in the absence of Ca2+, the Ba2+-PPCA complexation might be significant. It has been observed that new types of crystals form at high SI in the presence of PPCA (see Figure 8). However, this fiber type crystal was not found when both PPCA and Ca2+ (with higher concentration than Ba2+) present in solution. Although quite different from the regular BaSO4 morphologies, this fiber type crystal has been identified as BaSO4 with Electron Dispersive Spectrometric (EDS) analysis. This observation is consistent with that by Benton, et al.10 This new type crystal is proposed to be induced by the Ba2+-PPCA complexes. If it is true, this polymer-induced crystallization might be useful to produce artificial crystals with special morphologies.17

Figure 8. BaSO4 crystal morphology (SEM). Condition: [NaC1] = 1 .0 m; SI = 2.92; T = 298.2 °K; Cp = 27 ppm; CCa = 0 m; pH = 6.3; stirring at 350 rpm.

178

0

0.05

0.1 0.15 CCa (m)

0.2

0.25

Figure 9. Effect of Ca2+ on the overall efficiency of PPCA to inhibit BaSO4 nucleation from equation 13. Assumption: Cp = 1 ppm; I = 1.5 m; SI = 1.5; T = 298.2 °K.

Figure 9 shows the enhancement of Ca2+ on the overall efficiency of PPCA at two different pHs. At low pH, the Ca2+ has insignificant enhancing effect. This is because at pH 4 PPCA is mostly protonated, which is not effective in inhibiting BaSO4 nucleation. At pH 6, the dominant species are dissociated and calcium complexed PPCA. From the solution chemistry of PPCA and equation 13, the enhancement of Ca2+ on the efficiency of PPCA to inhibit BaSO4 nucleation might occur through: decreasing the fraction of protonated PPCA, which is ineffective as an BaSO4 nucleation inhibitor; increasing the fraction of calcium complexed PPCA, which is the most effective species in PPCA for BaSO4 nucleation; decreasing the fraction of the dissociated PPCA, of which efficiency is lower than the calcium complexed PPCA; and therefore, inhibiting/attenuating the Ba2+PPCA complexation, which might induce/accelerate BaSO4 nucleation. Figures 10-14 show BaSO4 crystal morphology under different conditions. In the absence of PPCA/Ca2+, BaSO4 solids are crystals although the morphology is temperature and supersaturation index dependent (Figure 10 and 11). In the presence of PPCA or PPCA/Ca2+, BaSO4 solids are not in any regular morphology. Different from other inhibitor molecule,11-13 PPCA seems to have no selective preference on inhibition of specific surface growth of BaSO4 crystals. This might be due to the linearity and flexibility of the PPCA backbone, which makes PPCA easily adjust to “match ” different lattices on different surfaces. Besides the all-surface modification (Figure 12), PPCA also tends to result in chained/agglomerated amorphous BaSO4 solids (Figure 13 and 14). As commonly interpreted, crystal growth is inhibited/modified by the adsorption of impurities on the active growth sites on crystal surfaces. For polyelectrolytes with negatively charged functional groups such as PPCA, the adsorption can only occur on the active sites with Ba2+ exposed (for BaSO4 nucleation). The addition of Ca2+ can change the surface charge distribution and more important, can function as a bridge connecting both PPCA functional groups and the sites with SO42- exposed on crystal surfaces, as Figure 15 illustrates. Thus, the adsorption of PPCA on BaSO4 surfaces can be enhanced. This mechanism can explain why Ca2+ can while Mg2+ cannot enhance the efficiency of PPCA on inhibition of BaSO4 nucleation, as observed in this lab and in other studies.14 Ca2+ can form both Ca2+-PPCA and CaSO40 complexes so as to be able to function as a bridge between PPCA functional groups and crystal sites with SO42- exposed. Mg2+ cannot form Mg2+-PPCA complexes or the complex is too weak so that Mg2+ cannot effectively

179

Figure 10. BaSO4 crystal morphology (SEM). Condition: [NaCl] = 1.0 m; SI = 2.30; T = 356.2 °K; pH 6.3; no PPCA and Ca2+; stirring at 350 rpm.

Figure 11. BaSO4 crystal morphology (SEM). Condition: [NaCI] = 1.0 m; SI = 2.92; T = 298.2 °K; pH 6.3; no PPCA and Ca2+; stirring at 350 rpm.

180

Figure 12. BaSO4 crystal morphology (SEM). Condition: [NaCl] = 1.0 m; SI = 2.28; T = 356.2 °K; pH 6.9; Cp = 27 ppm; CCa = 0.02 m; stirring at 350 rpm.

Figure 13. BaSO4 crystal morphology (SEM). Condition: [NaC1] = 1.0 m; SI = 1.43; T = 356.2 °K; pH 6.0; Cp, = 1.08 ppm; CCa = 0.1 m; stirring at 350 rpm.

181

Figure 14. BaSO4 crystal morphology (SEM). Condition: [NaC1] = 1.0 m; SI = 2.13; T = 298.2 °K; pH 6.3; Cp = 1.08 ppm; CCa = 0 m; stirring at 350 rpm.

Figure 15. Illustration of the role of Ca2+ in enhancing the efficiency of PPCA to inhibit BaSO4 nucleation.

182

"adhere" PPCA functional groups on crystal sites with SO42- exposed. On the other hand, Cu2+ and Zn2+ decrease the efficiency of PPCA on inhibition of BaSO4 nucleation, as observed in this lab (Table 2). The reason is that Cu2+ and Zn2+ can form strong Cu2+PPCA and Zn2+-PPCA complexes but not strong CuSO40 and ZnSO40 complexes, and the Cu2+-PPCA and Zn2+-PPCA might be multi-dentate coordination which changes the PPCA structure and reduces the efficiency of PPCA. Besides, CuPPCA and ZnPPCA precipitates have been observed to have much lower solubility than CaPPCA precipitates in this lab. Therefore, the apparent efficiency of PPCA is limited by the solubility of CuPPCA and ZnPPCA. This "bridging" mechanism can also explain why the efficiency of PPCA increases with an increase in the ratio of Ba2+/SO42- in the absence of Ca2+ when the overall supersaturated index (SI) keeps constant.20 This is because Ba2+ can form Ba2+-PPCA complexes and it also has strong affinity for SO42- This might also be the reason that PPCA can induce formation of fiber type BaSO4 crystals. According to this mechanism, a metal ion which has strong affinity to both SO42- and an inhibitor functional group but not too strong to precipitate with either of them can enhance the efficiency of the inhibitor on BaSO4 nucleation.

Table 2. Experimental data for BaSO4 nucleation - effect of Cu2+ and Zn2+ [Ba2+] = [SO42-] = 2.79 mm; [NaC1] = 1 .0 m; C, = 27 ppm; T = 298 K. Ions

Cu2+ Cu2+ Zn2+ Zn2+ Note:

CM (mm)

pH

I (m)

SI

l o g1 0 ( t i n d) (Sec.)

0.000 0.200 0.010 0.200 0.010

5.76 5.84 5.96 5.84 6.00

1.01 1 1.012 1.011 1.012 1.011

2.92 2.92 2.92 2.92 2.92

3.699 2.471 3.602 3.079 3.653

CM - Total concentration of metal ions

SUMMARY (1). Quantitative analysis of the effect of calcium ions on the efficiency of PPCA in inhibiting BaSO4 nucleation has been derived in this paper. (2). pH was observed to have strong effect on the efficiency of PPCA. Through PPCA solution chemistry, it was derived that the protonated PPCA has no significant effect on inhibiting BaSO4 nucleation. The effective species are the dissociated and calcium complexed fraction of PPCA. (3). Ca2+ can enhance the overall efficiency of PPCA through three possible ways: decrease the protonated fraction of PPCA so that increases the effective fraction; increase the calcium complexed fraction of PPCA, the most effective species; and decreases the

183

dissociated fraction of PPCA, which is not so effective as the calcium complexed fraction and tends to induce/accelerate BaSO4 nucleation at high SI. (4). The enhancement of Ca2+ on the efficiency of PPCA to inhibit BaSO4 nucleation is strongly influenced with the supersaturation index. Only at low supersaturation index, the enhancement is strong. (5). Temperature has slight effect on the efficiency of PPCA. (6). PPCA and Ca2+/PPCA have no specificity on inhibition of specific surface growth of BaSO4 crystals. All inhibited BaSO4 solids are chained or agglomerated amorphous solids.

ACKNOWLEDGEMENT The financial support of Rice University Brine Chemistry Consortium of companies: Aramco.; B. J. Services; Baker-Petrolyte; Champion Technologies, Inc.; Chevron Petroleum Technologies, Inc.; Conoco, Inc.; Texaco, Inc.; etc., to this research is greatly appreciated.

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

184

K. Wojciechowski, and W. Kibalczyc, Light scattering study of KH2PO4 and BaSO4 nucleation process, Journal of Crystal Growth, 76:379( 1986). E. N.Rizkalla, Kinetics of the crystallization of barium sulphate, J. Chem. Soc., Faraday Trans. 1(79): 1857(1983). K. Taguchi, J. Garside, and N. S. Tavare,, Nucleation and growth kinetics of barium sulphate in batch precipitation, Journal of Crystal Growth, 163:3 18( 1996). P. Hartman, and C. S. Strom, Structural morphology of crystals with the barite (BaSO4) structure: a revision and extension, Journal of crystal Growth, 97:502( 1989). N.L. Allan, A. L. Rohl, D.H. Gay, C.R. Catlow, R. J. Davey, and W.C. Mackrodt, Calculated bulk and surface properties of sulfates,. Faraday Discuss. 95:273( 1993). S. T.Liu, and G. H. Nancollas, Scanning electron microscopic and kinetic studies of the crystallization and dissolution of barium sulfate crystals, Journal of Crystal Growth, 33:1 1(1976). M. L. J. Leeuwen, O.S. L. Bruinsma, and G. M. Rosmalen, Three-zone approach for precipitation of barium sulfate. Journal of Crystal Growth, 166: 1004( 1996). S. He, J. E. Oddo, and M. B. Tomson, The nucleation kinetics of barium sulfate in NaCl solutions up to 6m and 90 °C, Journal of Colloid and Interface Science, 174:3 19 (1 995). L. Fernandez-Diaz, A. Putnis, and T. J. Cumberbatch, Barite nucleation kinetics and the effect of additives, Eur. J. Mineral., 2:495( 1990). W. J. Benton, I. R. Collins, I. M. Grimsey, G. M. Parkinson, and S. A. Rodger, Nucleation, growth and inhibition of barium sulfate-controlled modification with organic and inorganic additives. Faraday Discuss. 95:281(1993). S. N. Black, L. A. Bromley, D. Cottier, R. J. Davey, B. Dobbs, and J. E. Rout, Interactions at the organic/inorganic interface: binding motifs for phosphonates at the surface of barite crystals. J. Chem. Soc. Faraday Trans. 87(20):3409(1991). R. J. Davey, S. N. Black, L. A. Bromley, D. Cottier, B. Dobbs, and J. E. Rout, Molecular design based on recognition at inorganic surfaces. Nature, 353(10):549( 1991). A. L. Rohl, D. H. Gay, R. J. Davey, and C. R. Catlow, Interactons at the organic/inorganic interface: molecular modeling of the interaction between diphosphonates and the surfaces of barite crystals. J. Am. Chem. Soc. 118:642(1996).

14. L. S. Boak, G. M. Graham, and K. S. Sorbie, The influence of divalent cations on the performance of BaSO4 scaleinhibitor. SPE 50771, 1999. 15. Breen et al, Metal Ion Complexes for Use as Scale Inhibitors. US Patent 5,207,919. 16. W. H. Leung, and G. H. Nancollas, Nitrilotri(methylenephosphonic acid) adsorption on barium sulfate crystals and its influence on crystal growth, Journal of Crystal Growth, 44: 163( 1978). 17. B. R. Heywood, and S. Mann, Organic template-directed inorganic crystallization: oriented nucleation of BaSO4 under compressed Langmuir monolayers. J. Am. Chem. Soc., 114:4681(1992). 18. S. He, J. E. Oddo, and M. B. Tomson, The inhibition of gypsum and barite nucleation in NaCl brines at temperatures from 25 to 90°C. Applied Geochemistry, 9:561 (1994). 19. S. L. He, A. T. Kan, and M. B. Tomson, Mathematical inhibitor model for barium sulfate scale control. Langmuir, 12: 1901 (1996). 20. M. C. van der Leeden, and G. M. van Rosmalen, Inhibition of barium sulfate deposition by polycarboxylates of various molecular structures. SPE Production Engineering 5( 1):70, 17914 (Feb 1990). 21. J. Xiao, MS thesis, Rice University, 1997. 22. Water Analysis Handbook. Hach Company: Loveland CO, 1989. 23. Standard Methods for the Examination of Water and Wastewater, 1 8th ed.; Greenberg, A. E.; Clesceri, L. S.; Eaton, A. D.; Ed.; American Public Health Association: Washington, D. C., 1992.

185

This page intentionally left blank.

INORGANIC MINERAL SCALE CONTROL IN SUGAR EVAPORATORS USING SCALE INHIBITORS

Jasbir S. Gill Calgon Corporation POB 1346 Pittsburgh, PA 15230

INTRODUCTION Thermal efficiency of sugar evaporators is largely responsible for the overall efficiency and sometimes the capacity of the sugar mill. Although there are several factors which determine the overall thermal efficiency of an evaporator, heat exchanger surface cleanliness is the major limiting factor. Fouling can occur on either side of the heat exchanger tube, i.e., the juice side or the steam side, however, the deposit is mostly encountered on the juice side of the evaporative tubes. Deposit in beet sugar evaporators generally consist of multiple components' made up of materials brought in with the liquor (juice) or formed in place. The deposit composition varies with location in multi-effect evaporators due to changes in temperature, total solids, and pH. Deposit found in beet sugar evaporators can contain a variety of minerals such as calcium carbonate, calcium sulfate, calcium phosphate, calcium oxalate, and various silica and silicate compounds. The location of these minerals in the multi-effect evaporation system is determined by not only temperature and pH but also sugar contents of the effect. In the earlier effects, calcium carbonate and calcium sulfate are the most common minerals causing deposition due to high temperature. However, these minerals can also be found at a latter effect because of higher sugar loading. Solubility of most of these minerals is very much affected by the sugar loading. In a 55% sugar solution, calcium sulfate is four times less soluble than in thin juice. Silica/silicate and calcium oxalate deposits are generally found in the latter stages of the evaporators. Solubility of silica decreases with increasing sugar concentration and decreasing temperature. Silica is half as soluble in a 60 brix sugar solution compared to water at the same temperature. Thus the lower temperature and high sugar loading in the last effect makes it an ideal site for silica, calcium silicate, and

Advances in Crystal Growth Inhibition Technologies Edited by Amjad. Kluwer Academic/Plenum Publishers, New York. 2000

187

magnesium silicates deposits. A similar situation is true for calcium oxalate precipitation, as it is less soluble at lower temperature and high sugar loading. "his study is focused on the understanding and prevention of calcium oxalate deposit, since other scales such as calcium carbonate and calcium sulfate can be controlled using various scale inhibitors such as polyacrylic acid, polyaminopolyether tetra(methylene phosphonic acid), etc. The formation and dissolution of calcium oxalate hydrates as a common constituent of pathological deposits in the urinary tract of mammals have been the subject of numerous studies and publications.2-4 Precipitation of calcium oxalate occurs commonly in the processing equipment of many industries, such as paper-making and food and beverage processing. Beer-stones, hard deposits of calcium oxalate combined with proteinaceous and resinous material, are often encountered in ferments. Although there are many other areas where calcium oxalate poses deposition problems, the focus of this paper is on the fundamental aspects of calcium oxalate deposit control in sugar evaporators. However, this knowledge is widely applicable in other industries, as demonstrated in the case history describing the benefits of developing a proper treatment for the inhibition of calcium oxalate deposit in a tannin production process. In the beet sugar industry, calcium oxalate deposition in evaporators is a common occurrence. In spite of a significant reduction in the oxalic acid concentration during initial juice purification, thin juice of 12% dissolved solids contains approximately, 120460 parts per million (ppm) oxalic acid. During evaporation, additional oxalic acid accumulates from the invert degradation of sugar. The solubility of calcium oxalate in a 50% sucrose solution is one-tenth that in water at the same temperature. A combination of these factors, i.e., increased sugar contents, higher oxalic acid concentration, and lower temperature, creates ideal conditions for calcium oxalate deposition in the last effect evaporator. Thus, it is a complicated process to predict both the nature of a deposit and at which stage of a multi-effect evaporator a particular mineral will deposit. This depends on temperature, pH, total dissolved solids, the influence of total dissolved solids on the solubility of the minerals, and the kinetics of mineralization. In order to effectively treat these evaporators, a thorough understanding of all the factors is crucial to determine a proper inhibitor and apply it at a proper feed point.

EXPERIMENTAL The efficacies of several scale inhibitors were evaluated using stagnant flask tests. The test solutions representing different molar ratios of calcium to oxalate, concentrations, pH, and inhibitor concentrations were thermostated at different temperatures for various time intervals. At the end of the incubation time, the solutions (while still hot) were filtered through a 0.45µm membrane. The clear filtrate was analyzed for various ions to determine precipitation inhibition.

RESULTS AND DISCUSSION Calcium oxalate is precipitated in three distinct morphological phases with the same chemical composition but different hydration states: (1) a monoclinic monohydrate

188

CaC2O4 .H2O (COM), (2) a tetragonal dihydrate CaC2O4 .2H2O (COD), and (3) a triclinic trihydrate CaC2O4 .3H2O (COT). In the absence of additive, the following equilibria may be considered: H2C2O4 ⇔ H+ + HC2O4-

(1)

HC2O4- ⇔ H+ + C2O42-

(2)

Ca2+(aq) + HC2O4-(aq) ⇔ CaHC2O4+(aq)

(3)

Ca2+(aq) +C2O42-(aq) ⇔ CaC2O4(aq)

(4)

CaC2O4(aq) + C2O42- ⇔ Ca(C2O4)22-(aq)

(5)

Ca2+(aq) + C2O42-(aq) ⇔ CaC2O4.H2O(s)

(6)

Various equilibria constants can be obtained from Martell and Smith's book.5 Equilibrium (6) is directed at the formation of calcium oxalate monohydrate crystals. The driving force for the crystal growth process is calculated by the free energy expression: ∆ G = ½ RT In AP/ksp

(7)

where AP is the activity product of calcium oxalate ions in solution and ksp is the thermodynamic equilibrium solubility product for calcium oxalate monohydrate. The free ionic concentration of oxalate ion is controlled by equilibria (1) through (5) and that of calcium ion by equilibria (3), (4), and (5). Equilibria (1) and (2) are highly pH dependent, which largely determine the concentration of oxalate ions in solution. Figure 1 shows the driving force (in a simplified version by multiplying free calcium and free oxalate concentration) at various pH values for the same total calcium and oxalate concentration in water. As it can be seen, between pH 3.5 and 4.5 the driving force increases almost by three-fold. The concentration of free oxalate ions seems to level off at approximately pH 6. Such high dependence on pH makes it very difficult to inhibit the crystal growth of calcium oxalate. Although its most stable phase is COM, the stability of different hydrate phases varies according to the pH, temperature, presence of additives, induction time, and rate of crystal growth. Numerous studies6-9 on the kinetics of calcium oxalate precipitation have demonstrated the importance of kinetic factors in determining the phase formed during crystallization. Inhibition of pre-cursor phase(s), however, can delay (induction time) or inhibit the formation of COM. Tomazic and Nancollas10 found that the rate of calcium oxalate crystallization is of second order with respect to relative supersaturation for all three hydrates. They also found that different solid(seed) / solution ratios resulted in different ratios of COD, COT, and COM. As pointed out earlier the driving force for calcium oxalate precipitation is significantly increased by an increase in solution pH. Generally, it is quite a challenge to inhibit calcium oxalate in the pH range of 4-6 because of its high driving force and the

189

Figure 1. Effect of pH on the driving force for calcium oxalate precipitation.

190

inability of most anionic inhibitors to be fully functional. This is evident from the data shown in Table 1. With an increase in pH to above 4, the performance of polyphosphate (SHMP) is significantly reduced, while the effectiveness of polyacrylic acid (PAA) is improved only above pH 6. This observation results from a combination of several factors affected by the change in pH. In addition to increasing the driving force for calcium oxalate precipitation, increasing pH perhaps causes a deterioration of polyphosphate and increases ionization of polyacrylic acid. In such conditions, a combination of polyphosphate and polyacrylic acid is a better inhibitor than only one of the components, especially if the application has a changing, rather than constant, pH. In a water which contains 16 mg/L total calcium and 35.2 mg/L total oxalate, at pH 4.0, it takes 0.5 mg/L SHMP for complete inhibition. In the same water, PAA is not able to achieve 100% inhibition, even at 10 mg/L. However, a 1:1 mixture of SHMP and PAA requires 1.0 mg/L for complete inhibition (contains 0.5 mg/L SHMP), while at pH 7 for the same solution, either 1.5 mg/L of PAA or 3 mg/L of 1:1 mixture of SHMP and PAA gives compete inhibition. On the other hand, SHMP even at >10 mg/L does not provide 100% inhibition. AA/AMPS (acrylic acid:2-acrylamido-2-methyl propanesulfonic acid) copolymer is not as effective as PAA for precipitation inhibition and its dispersive properties were not investigated in this study. Table 1. Inhibitor required (mg/L) for 100% Calcium Oxalate inhibition at varying pH. Inhibitor (mgL) Inhibitor SHMP

3

4

5

0.1

0.5

8.0

pH 6

7

9-

10.0

>10

>10

STP

0.1

>5

>10

>10

>10

PAA*

0.5

>5

>5

3.5

1.5

SHMP/PAA

0.2

1 .0

10.0

6.0

3.0

SHMP/CP**

0.2

1 .0

10.0

7.5

4.0

1.2

9.5

2.5

*PAA = Polyacrylic acid (8-12K Mol. Wt.) **CP = Copolymer of AA and AMPS Experimental Conditions: Total Calcium = 16mg/L, Total Oxalate = 35.2 mg/L Temperature = 60 oC

The effect of temperature on calcium oxalate crystal growth inhibition was investigated by studying inhibition at two temperatures, 22 °C and 60 °C. The inhibitor performance improved with increased temperature as shown in Figure 2. This is perhaps due to the lower supersaturation of calcium oxalate as a result of increased equilibrium solubility of the mineral. This phenomenon is partly responsible for causing calcium oxalate deposition in the last effect with relatively cooler temperatures compared to the earlier effects in a multi-effect evaporator. The data presented in Table 2 show that there is a significant effect on inhibitor performance due to changes in the calcium and oxalate ionic product and calcium to oxalate ratios. As expected more inhibitor is required for complete inhibition for higher ionic product of calcium and oxalate ions than for the lower ionic product. However, calcium/oxalate ratios seem to have a rather significant effect on inhibitor requirements for calcium oxalate inhibition. A two-fold increase in calcium concentration for the same amount of oxalate lowers the calcium oxalate inhibition by 8%; however, a two-fold increase in oxalate concentration, for the same calcium level, lowers the inhibition by 30%. To achieve similar 30% reduction in inhibition the calcium concentration has to be increased six-fold to 96 mg/L. These concentrations are compared on equimolar bases. Tomson 191

SHMP

PAA

SHMP/PAA

Figure 2. Effect of temperature on calcium oxalate inhibition.

192

and coworkers11 observed a similar phenomenon for calcium carbonate inhibition; they explained it on the basis of cluster formation theory in which scale forming anions and anionic inhibitor compete for the same nucleating cluster for growth and inhibition, respectively. Alkaline earth metal, transition metal, and tervalent metal ions are known to interfere in the aqueous crystal growth process of sparingly soluble minerals. Inhibition of calcium carbonate and its morphological changes as a result of Mg2+ ion presence is well documented. A recent publication by Fernandez-Diaz and coworkers12 indicated that the presence of Mg2+ in the interstitial fluid inhibits nucleation of calcium carbonate, increasing the threshold supersaturation at which crystallization begins. The resulting Mg-calcite crystals show a range of morphologies depending on the Mg content and supersaturation at the point of crystallization producing a variety of crystal shapes, from a dumbbell-like form to a wheat-sheaf-like form. There have been a number of investigations on the role of Mg2+ in calcium oxalate crystal growth. Wunderlich13 concluded that Mg2+ may induce an increase in calcium oxalate solubility but more likely it broadens the Ostwald-Miers range.* As shown in Tables 3 and 4, a small quantity (2-4 mg/L) of Al3+ in solution alone or in combination with other crystal growth inhibitors

Table 2. Effect of Ionic product and Cationic/Anionic ratio on Calcium Oxalate Inhibition at 22o C, pH 6.0. %Inhibition with different doses (mg/L) of SHMP/PAA/CP TCalcium

TOxalate

(mg/L)

(mg/L)

5.0

10.0

20.0

16

35.2

87

100

-

16

70.4

57

100

-

32

17.6

100

32

35.2

79

100

-

32

70.4

60

100

-

32

176

0

0

64

35.2

65

100

64

70.4

40

70

92

64

141

0

0

35

96

35.2

59

100

96

70.4

21

53

76

96

141

0

0

16

128

35.2

35

87

100

128

70.4

11

31

74

192

70.4

0

35

69

192

35.2

22

67

89

17 -

-

*Ostwald-Miers (metastable) range is the space between the saturation curve and the spontaneous nucleation curve in a saturation -temperature diagram. Within this range nucleation does not take place but seed crystals can grow because of supersaturation. The breadth of the Ostwald-Miers range depends mainly on nucleation energy.

193

Table 3. Effect of Al3+ on Calcium Oxalate inhibition at various driving forces for Calcium Oxalate precipitation (22oC). % Calcium Oxalate Inhibition at different Al3+ levels (mg/L) TCalcium (mg/L)

16

TOxalate (mg/L)

pH

2

4

10

0

90

100

100

0

35.2

5.0

16

70.4

5.0

0

65

100

-

16

105.6

5.0

0

0

7

100

32

70.4

5.0

0

-

85

100

48

70.4

5.0

0

-

69

100

32

35.2

5.0

0

87

100

-

48

35.2

5.0

0

85

100

-

16

35.2

5.5

0

80

100

-

16

35.2

6.0

0

80

100

-

16

35.2

6.5

0

80

100

-

16

35.2

7.0

0

0

65

70

16

35.2

9.5

0

0

0

0

16

35.2

10.5

0

0

0

0

performs as an excellent calcium oxalate inhibitor. At pH 5 the performance of Al3+ is significantly better than any other inhibitor evaluated in this study. Above pH 6.5, however Al3+ becomes ineffective, perhaps due to increased hydrolysis. SHMP and pyrophosphate seem to have a significant negative impact on the ability of Al3+ to inhibit calcium oxalate precipitation. It appears that this negative effect is caused by the formation of Al(PO)4, No such negative effect is observed in the presence of AA/AMPS copolymer, since the polymer can effectively inhibit the formation of Al(PO)4 and thus maintains the concentration of aluminum ions in solution. Several other metallic ions like zinc, copper, manganese, and iron as well as, copper complexes of EDTA and citrate have been investigated14 for calcium oxalate inhibition. Zinc and Cucitrate were effective inhibitors for calcium oxalate precipitation. Langmuir adsorption treatment revealed that Zn2+ is effectively adsorbed on calcium oxalate crystals, thus blocking the active growth sites. Al3+ perhaps follows a similar mechanism, only more effectively than Zn2+ or Cu-citrate complex. The following case history depicts the success of this research work.

A CASE HISTORY A tannin manufacturing plant in the south west of France produces 25 tons per day of tannin extracted from chestnut trees. Extraction of tannin is carried out by digestion of wood chips under high temperature and pressure. Concentration of extracted liquid is carried out in evaporators and final purification occursin clarifiers before drying the concentrate to produce tannin powder.

194

Table 4. Comparison of different inhibitors for Calcium Oxalate threshold inhibition at pH 5.0. Effect of AI3+ on threshold inhibition. % Inhibition AI3+ (mg/L)

Inhibitor 0.0

2.0

4.0

SHMP

31

90

100

STP

10

PAA

43

83

100

SHMP/PAA

43

70

100

SHMP/PAA/CP

40

90

100

CP

33

85

100

(5.0 mg/L)

TCalcium =16mg/L Toxalate =35.2mg/L Temperature = 22 oC

The plant has been experiencing heavy calcium oxalate buildup in the evaporators’ tubes and channel heads; two to three inches of calcium oxalate were deposited in a matter of one week. Tannin is concentrated in the evaporators at a temperature of 85 oC and pH of 3.2. To prevent calcium oxalate buildup, the plant tried various chemical treatments without success. It was a standard practice to circulate trisodium phosphate twice a week and conduct a mechanical cleaning at least two times per month. The mechanical cleaning was carried out by hammering and drilling through three inches of deposit. We started our trial in September, 1992 on one train of evaporators by feeding,to the evaporator inlet, 80 ppm of a formulated product containing a mixture of sodium triphosphate and sodium hexametaphosphate (STP/SHMP). The performance was monitored by evaporator inspection. An evaporator was checked one week after startup and found to contain no deposits. This warranted a longer run for at least another two weeks without any inspection. Although the results were good, the treatment cost at 80 ppm was high. During the following month, the dosage was reduced to 60 ppm and started seeing some deposit build up after three days. Simultaneous feed of aluminum ions and STP/SHMP product was then tried an equal dosage ratio of aluminum to phosphate. With the addition of aluminum salt, the STP/SHMP product dosage was cut back to 50 ppm. Following a one month operation, the evaporators were opened and very little amount of extremely soft deposits were found. The deposits were easily wiped off. Results were even improved from the previous treatment with 80 ppm of the product alone. Since the introduction of the STP/SHMP/Aluminum treatment, the plant operated without any shutdown during the next six months. Mechanical cleaning has been eliminated. As a result of the trial success, the plant started treating the second train of evaporators at the beginning of June and the third and last train was treated after the turn around in September, 1993.

195

CONCLUSION Multiple effect evaporators consume about 70% of the total energy required in a typical sugar refinery. A small amount of scale buildup on evaporators tubes can significantly increase the energy consumption. The fouling consists of mineral scales, organic sludge, and corrosion products. Minerals commonly found as a part of the deposit include calcium carbonate, calcium sulfate, calcium oxalate, calcium phosphate, and silica/silicate compounds. During evaporation, changes in pH, sugar loading, and temperature alter the solubility of minerals, causing deposition at each stage. Laboratory data acquired using simple flask tests have been used to develop a multi-component product to inhibit most of the mineral precipitation and provide deposit control. At lower pH (<5.0) polyphosphate was found to very effective and at pH >5.0, polycarboxylates are effective in controlling calcium oxalate precipitation. Addition of trace amounts of aluminum ions to the water enhances the threshold inhibition with both polyphosphates and polycarboxylates. A careful system monitoring and a well-balanced formulated product fed at an appropriate feed-point can maintain the cleanliness of the evaporative tubes.

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

196

A. Browning, C. McDonough, R. Murtagh, Scale Control in Sugar Evaporators, presented before American Society of Sugar Beet Technologists, 24th General Meeting, Phoenix, AZ (1987). G.H. Nancollas, Urolithiasis Research 1976:5; latter published as a book by Plenum Publishing Corporation, New York. P.G. Koutsoukos , M.E. Sheehan, G.H. Nancollas, Investigative Urology. 18,5:358 (1981). W. Berg, A. Hesse, H.J. Shcneider, Urological Reasearch. 4:1976 (1976). M.L. Pulido, Sugary Azucar 13(1984). A.E. Martell, R.M. Smith, Critical Stability Constants, Plenum Pess, New York and London, 3: 92( 1979). G.H. Nancollas, G.L. Gardner, J. Crystal Growth 21:267 (1974). A.D. Randolph, G.W. Drach, J. Crystal Growth, 53:195 (1981). G.L. Gardner, J. Crystal Growth 30:158 (1975). B. Tomazic, G.H. Nancollas, J. Colloid Interface Science 75: 149 (1980). M.B. Tomson, et.al., Gas Research Institute Report #GRI-83/0048 (1983). L. Fernandez-Diaz, A. Putnis, M. Prieto, C.V. Putnis, J. Sediment Research, Section A, 66(3):482( 1996). W. Wunderlich, Urological Research 9:157 (1981). F. Grases, C. Genestar, A. Millan, J. Crystal Growth 94:507 (1989).

CONTROL OF CRYSTALLIZATION PROCESSES BY DIBLOCK COPOLYMERS

Özlem Dogan, Emel Akyol, Serkan Baris, Mualla Öner* Yildiz Technical University Department of Chemical Engineering Istanbul 80270, Turkey

INTRODUCTION The effect of polyelectrolytes on the precipitation of sparingly soluble salt has been known for several years. Studies related to the crystallization of calcium sulfate,] calcium oxalate2-3 and other alkaline earth carbonates4 in the presence of macroions have been published. Selected polyelectrolytes have been widely used to reduce the rate of crystal growth5 or to change the shape of the crystals6 and to improve the quality of crystalline products, powders, granular materials and in the stabilization of colloidal dispersions.7 On the other hand the notable examples of the roIe of crystal growth inhibition in biological processes are furnished by the action of certain peptides and glycopeptides in controlling the size and growth of cholesterol crystals, enamel, kidney stones and bone minerals.8 One question of general interest is that of how much structural detail is necessary in order that synthetic polymers may mimic some of the functions of proteins. Diblock copolymers are in principle substances that may serve for this purpose. In this work we examined the effect of poly(methacrylic acid)-blockpoly(butylmethacrylate) (i.e., MAA-b-BuMA) and poly(methacrylic acid)-block-poly(ethylene oxide) (i.e.,MAA-b-EO) copolymers on the crystallization of a calcium sulfate dihydrate. The growth of calcium sulfate dihydrate, or gypsum, is of considerable importance since it is frequently encountered both in nature and in industrial processes. Hence its precipitation is of particular interest not only for fundamental research concerning for biomineralization but also their importance in industrial applications. Large quantities of gypsum are produced as a byproduct in the production of phosphoric acid. The problem of scale formation on heat exchanger, reverse osmosis membrane surface and equipment surface is a persistent and ’an expensive problem in cooling water systems, boilers, secondary oil recovery utilizing water flooding techniques, desalination by evaporation and reverse osmosis method.9 * Corresponding author

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers, New York, 2000

197

EXPERIMENTAL PROCEDURE Group Transfer Polymerization was used to make block copolymers of acidic and hydrophobic methacrylates and acrylates of closely defined molecular weight, to be compared with random copolymers and homopolymers. MAA-b-EO block copolymers were kindly donated by MPI (Max-Planck Institute, Mainz, Germany). Crystal growth experiments were made in a water-jacketed Pyrex glass vessel of 1 L capacity. Supersaturated solutions for growth experiments were prepared by the slow mixing of equal volumes of calcium chloride and sodium sulfate solutions. In experiments where additives were used a similar procedure was followed and the freshly prepared additive solutions were normally added to the sulfate component. The complete experimentalprocedureswerereportedpreviously.10-11 The growth experiments were carried out by adding a known amount of the calcium or sulfate stock solution. After temperature equilibration the second component was added and conductivity, pH and temperature of the reaction solutions were monitored by computer during the crystallization. All experiments were performed at a temperature of 30 ± 0.1 ºC. The effect of polymers on the rate of precipitation of CaSO4 was determined by recording the decrease, as a function of time, in conductivity of a solution containing 8*10-2 molar CaC12 and Na2SO4 at 30°C. The effect of an additive can be quantified as the ratio of the rate of crystallization of the pure solution (ko) to the rate of crystallization in the presence of additive (k) at the same concentration and temperature. The induction period (tind ) was determined by monitoring the variations in the calcium concentrations accompanied by conductivity measurements. The time between the generation of a supersaturated state and the first observed change in calcium concentration or conductivity was defined as the induction period. The time periods were determined from the recordings of the time evolution of the conductivity of the solution, which is directly related to the volume of the precipitated calcium sulfate, and averaged from at least three separate experiments and only the average values of the induction period were reported. The reproducibility of this technique was about 4 to 5%. RESULTS AND DISCUSSION Table 1 and Table 2 summarize the acidic acrylate polymers used in this study and the effect of added polymer on the crystallization rate. The experiments have been carried out with methacrylate and acrylate block and random copolymers in order to see the differences in inhibitor performance. As shown in Table 1 all polymers at 1ppm concentration are reasonably good growth inhibitors but polyacrylate copolymers appear to be more effective than polymethacrylate copolymers at the same concentration, nearly same molecular weight range and same acid content. Upon adding polyacrylate copolymer concentrations of about 1 ppm in solution, crystallization was completely inhibited over the 5 hours period. The similar comparison was made using homopolymers of acrylate and methacrylate polymers as shown in Table 2. Four polyacrylic acids with different molecular weights ranging from 1200 to 240000 and three polymethacrylic acids having a molecular weight range of 8000 to 34000 have been investigated. At first inhibitor concentrations of 5 ppm were chosen, since polyelectrolyte dosages of several ppm are generally recommended. When this inhibitor concentration turned out to cause total blockage of the growth process at the established supersaturation, the concentration was lowered to 1 ppm. Even then it took several hours to obtain crystallization in solution. The homopolymer dosage was finally lowered to 0.05 ppm for determination of differences in inhibitor performance. The homopolymers showed appreciable differences in their influences on the length of the induction period. The crystallization reaction in the presence of 0.05 ppm of PAA is preceded by

198

an induction period of 360 minutes. In the presence of PMAA, the duration of the induction period is 278 minutes. The lesser retarding power of polymethacrylic acid and its copolymers compared with polyacrylic acid polymers may be attributed to hydrophobic methacrylate group. For an effective attraction and bonding, a flat configuration of the inhibitor upon crystal surface is desired. Hydrophobic methacrylate groups appeared to be detrimental to good inhibitor performance. This group is inclined to extend from the surface, resulting in a loopy configuration of the inhibitor molecule. They only increase the molecular weight of the polymer without adding extra bonding possibilities and they may even cause steric hindrance by methyl group interfering with the adsorption of the molecule to the crystal surface or by masking functional groups. Hence, the methyl group does prevent close lateral packing of chains on the surface, 830 pm vs. 440 pm for acrylic acid, thereby reducing the maximum negative charge density on the surface. A similar effect was seen previously for calcium oxalate crystallization.3 Table 1. Characterization of acrylate polymers and their effects on crytallization rate of calcium sulfate at 30°C and 1 ppm concentration. Polymer MAA-b-BuMA

MW 2300 13400 3400 18500 3000 10100 2000 6800 2700 7300 2800 11300 12100 5600 2200

MAA-co-BuMA

MAA-b-BuMA-b-MAA AA-b-BuA AA-co-BuA

Acid (%) 82 80 69 63 46 39 80 84 68 56 46 36 40 74 65

Induction Time (min) 156 356 135 343 110 152 122 188 96 127 87 133 142 >600 >600

ko/k 2.1 2.2 1.9 2.0 1.7 1.7 1.8 1.7 1.5 1.6 1.5 1.4 1.4 -----

Table 2. The growth retarding effect of PAA and PMAA homopolymers on CaSO4.2H2O crystallization. Polymer PAA

PMAA

MW 1200 5000 8000 240000 8000 20000 34000

1 ppm >800 >600 >400 355 >720 >370 >380

Induction Time (min) 0.1 ppm 0.05 ppm >600 >540 >500 44 8 >350 360 106 95 476 278 255 ----124

Figure 1 shows the effect of acid content of the polymer on the degree of inhibition, respectively, for block copolymers of molecular weight 2000-4000 and for random copolymers with similar molecular weights. The effectiveness of n-butylmethacrylate-methacrylic acid copolymers increases with increasing acid content of both block and random copolymers. Induction time increases from 152 min to 356 min for block copolymers of molecular weight

199

10000-15000 as the acid content of polymer increases from 39% to 80%. As shown in Table 1 and Table 2, PAA and PMAA are much more effective than equivalent copolymers. Several investigations have indicated that polymers which exhibit an inhibitory effect on crystallization kinetics of soluble salts are also highly substituted with carboxyl groups.12-13 Apparently the larger number of functional groups of negatively charged species increases the polar attraction between the adsorbate and the positive sites at the solution interface. It is also interesting to note that the crystallization reaction following the initial induction period commences with a rate close to that in the absence of inhibitor.

Acid (%) Figure 1. Effect of acid content of polymers on induction time.

The data for a typical crystal growth experiment in the presence of MAA-b-EO copolymers are shown in Table 3. It can be seen that the k0/k ratio and induction time increases with decreasing PEO block length. The crystallization reaction in the presence of 1 ppm of PEO is preceded by an induction period of 157 min and k0/k ratio of 3.8. In the presence of PMAA, the duration of induction period of >720 min. Induction time increases from 199 to 285 min and k0/k ratio increases from 6.1 to 34.6 as the PEO/PMAA ratio decreases from 10 to 1.4 for block copolymers (Fig.2). The similarity of response of MAA-b-BuMA and MAA-b-EO block copolymers, increase in inhibitor efficiency with MAA content, suggests that the degree of inhibition calcium sulfate crystallization by anionic polyelectrolytes is related to the maximum surface charge density due to adsorbed polymer. Previous studies have shown that block copolymers which consist of two different types of polymers, A and B are efficient dispersant.13 Polymer A has a strong affinity for the adsorbent and highly polar nature helps to anchor the polymer to the particle surface. A relatively nonpolar B is very compatible with the dispersion medium and has a little affinity for the surface, hence it is extended into the desorption medium and provides a steric barrier. The A segment is designed to adsorb on the surface of a solid, the B segment provides a steric stabilization of the particle. The A segment must be available for interaction with the particle surface; sufficient size to provide irreversible

200

adsorption. The B segment must be of sufficient size to provide steric stabilization. Extending this theory, one may assume specific interactions of the anchor block with crystal faces. The higher degree of polymerization of the anchor block results in much adsorbed polymer per unit surface area. Table 3. The growth retarding effect of 1 ppm MAA-b-EO copolymers on CaSO4.2H2O Crystallization. Polymer PMAA PEO MAA-b-EO

PEO : PMAA ratio (DPEO/DPMAA)

MW

Induction Time(min)

k0/k

n/a n/a 10 8 6 3.1 1.4

8000 8000 18000 17500 21600 22200 25400

>720 157 199 250 263 276 285

--3.8 6.1 8.28 9.61 10.81 34.6

DPEO/DPMAA Figure 2. Effect of PEO/PMAA ratio ofpolymers on induction time.

In contrast to the MAA-b-BuMA copolymers, the presence of MAA-b-EO copolymers resulted in a decrease to the rate of deposition of calcium sulfate dihydrate following the induction period. The much greater extent of inhibition by MAA-b-EO as compared to MAA-bBuMA is reflected by the ratios of k0/k which for acrylate copolymer is 2.2 and for MAA-b-EO copolymer, 34.6. Whether the increased activity of the MAA-b- EO copolymer relative to MAAb-BuMA is due to directly to the polymer structure is not certain. It may be that the slightly higher hydrophobicity and lower anionic charge density of the acrylate copolymers resulted in decrease in the degree ofassociation with the CaSO4 in the crystal lattice.

201

Molecular Weight of Polymers Figure 3. Effect of block copolymer weight on calcium sulfate dihydrate crystallization.

Molecular Weight of Polymers Figure 4. Effect of random copolymer weight on calcium sulfate dihydrate crystallization.

202

The Effect of Molecular Weight of Polyelectrolytes As shown in Figure 3 and 4, block and random copolymers are more effective at higher molecular weights – in the range of 2000-18000. Induction time values decrease 356 min to 156 min for 80% acid content block copolymers as the molecular weight decreased 13400 to 2300. Polyacrylic acids, which varied in molecular weight from 1200 to 240000 and polymethacrylic acids of varying molecular weight (MW= 8000-34000) were tested at concentrations of 0.1 ppm and 0.05 ppm. The tests showed that the effectiveness of these polymers decreases when the molecular weight of the polymers increased. As the molecular weight increases the homopolymers become progressively worse for inhibition crystallization probably as a result of the increased bridging effects, and decreased solution mobility associated with the higher molecular weights. Crystal growth experiments made in the presence of PMAA shows that the presence of trace amount (0.1 ppm) of polymer resulted in initial inhibition. The induction periods decrease from 360 min to 124 min with increasing molecular weight from 8000 to 34000. This can be attributed to the presence of kinetic barrier arising from the high electric field just above the initially adsorbed lower molecular mass polyelectrolyte. The molecular weight of the polyelectrolyte appears to be an important parameter. Polymer molecular weight influences the adsorption of the polyelectrolyte onto the growing crystal. For adsorption to occur a minimum number of binding possibilities have to occur. In general, polyelectrolyte inhibitors will not primarily adsorb along the steps but onto flat crystal terraces, thus reducing the attachment area for the crystal growth units and providing obstacles for the advancing growth steps that propagate along the terraces. A higher surface coverage might thus be needed for growth blockage in the case of larger molecules. This minimal coverage needed for blockage will also depend, as already mentioned, on how flatly the molecules are adsorbed at the crystal surface. It is reported that a drastic decrease in inhibitor effectiveness for polycarboxylates with a molecular weight below 1000 occurs when applied for scale inhibition in saline waters. They reported the same tendency with increasing molecular weight values above 3500.14 This is surprising since the number of cooperative bonding possibilities will increase with increasing chain length of the inhibitor molecule. For charged molecules preferential adsorption of the smaller molecules has sometimes been observed. It was found that for the adsorption on BaSO4 of polyacrylates with molecular weights varying from 17000 to 450000, low-molecular mass polymer was preferentially adsorbed and was not replaced by high molecular weight mass polymer even after a prolonged period.14 It was suggested that there should be an optimum molecular weight for a given additive which a maximum effects would be obtained.15 This optimum would occur as a result of the degree of adsorption increasing with increasing molecular weight whereas the rate of adsorption would be expected to decrease.

Effect of Polyelectrolytes on Crystal Morphology The influence of the additives on crystal habit may also be seen by scanning electron photomicrographs in Fig.5. Fig. 5-a shows the well known thin, elongated, fragile gypsum crystals precipitated from pure solutions, some probably broken due to the agitation employed. The noticeable effect of poly(n-butylmethacrylate-b-methacrylic acid) block copolymer on gypsum crystal habit and agglomeration is seen in Fig 5-b. The final precipitate consisted of the clumps of needle aggregates, together with a few small particles which had presumably escaped aggregation.The crystals are shorter and include some star like agglomerates which probably result from the relatively faster growth of agglomerated nuclei. The presence of polyacrylic acid polymer in solution results in an increase of agglomeration (Fig.5-c). The needles and plates disappear; crystals are deformed and covered with sponge-like mass of minute

203

Figure 5. Scanning electron micrographs of crystals grown (a) for 5 h from a solution with no polymer added (b) for 5 h from a solution containing methacrylic acid-butylmethacrylate block copolymer (c) for 8 h from a solution containing PAA (d) for 7 h from a solution containing PMAA (e) for 5 h from a solution containing methacrylic acidethylene oxide block copolymer (f) for 5 h from a solution containing PEO.

crystallites. PAA proved to stimulate an abundant agglomeration mechanism. It seems that coagulation occurs by a bridging mechanism in which some of the polymer molecules are simultaneously adsorbed onto more than one small crystal. Polymethacrylic acid homopolymer caused less agglomeration and the plate like crystals are increased (Fig.5-d). The presence of MAA-b-EO copolymer produced crystals exhibiting a morphology that differed from the control crystals (Fig. 5-e). The plate-like crystals are markedly increased. The average value of the length and width of the crystals grown in the presence of the MAA-b-EO copolymer were less than that of the control sample. The presence of PEO had no effect on crystal morphology or size (Fig. 5-f)

204

CONCLUSIONS All polymers tested in this study are effective as growth inhibitors in this experimental condition. The sequence in growth inhibiting effectiveness of polymethacrylates is measured to be homopolymer > block copolymer > random copolymer. The smooth increase in inhibitor efficiency with acid content of the block copolymers suggests that the adsorption of polyelectrolytes depend on the sign of the charge on the polyelectrolytes and on the solid surface. The larger number of negatively charged functional groups increases the attraction between the adsorbate and the positive sites at the precipitate solution interface. Higher molecular weight PAA and PMAA were less effective for inhibiting crystal growth. In the case of copolymers, an increase in effectiveness was found with increase in molecular weight in the range of 2000-1 8000. It appears, however, that the optimum molecular weight is a function of the particular polymer species. Agglomeration was found to occur with PAA homopolymer, suggesting that polymer adsorbed on the particle surfaces, linking them together.

ACKNOWLEDGMENT We appreciate the support of Yildiz Technical University Research Fund (Project No: 98-A07-01-01) and TUBITAK (Project No: TBAG-AY/29) for the accomplishment of this work.

REFERENCES 1. Z. Amjad, and J. Hooley, Influence of Polyelectrolytes on the Crystal Growth of Calcium Sulfate Dihydrate, J. Colloid and Interface Science 11 1 :496 (1986). 2. M. Öner, P. Calvert, Influence of Polymer Architecture on Crystal Growth Inhibition, ACS PMSE 69: 166 (1993). 3. M. Öner, P. Calvert, The Effect ofArchitecture ofAcrylic Polyelectrolytes on Inhibition of Oxalate Crystallization, Materials Science & Engineering C, Biomimetic Materials, Sensors and Systems C2:93 (1994). 4. D.R. Sexsmith, and E.Q. Petrey, The Use of Polyelectrolytes for Scale Control in Sea Water Evaporators, 13:87 (1993). 5. Z. Amjad, Kinetics of Crystal Growth of Calcium Sulfate Dihydrate. The Influence of Polymer Composition, Molecular Weight, and Solution pH, Can. J. Chem. 66:1529 (1987). 6. S.T. Liu, and G.H. Nancollas, The Crystal Growth and Dissolution of Barium Sulfate in the Presence ofAdditives, J. Colloid and Interface Science 52:582 (1975). 7. J. Rudolph, J. Patzsch, W.H. Meyer, and G. Wegner, The Interaction of Acrylic Diblock Copolymers with Aluminum Oxide Surfaces and Their Application for Ceramic Powder Processing, Acta Polymer, 44:230 (1993). 8. G.H. Nancollas, Biological Mineralization and Demineralization, Springer Verlag, Berlin, (1982). 9. G.M. van Rosmalen, P.J. Daudey, W.G.J. Marchee, An Analysis of Growth Experiments of Gypsum Crystals in Suspension, J.Crystal Growth 52:801 (1981). 10. M. Öner Ö. Dogan, G. her, The Influence of Polyelectrolyte Architecture on Inhibition of Calcium Sulfate Dihydrate Crystallization, 35 1, AIChE Annual Meeting, Los Angeles, (1997). 11. M. Öner Ö Dogan, G. Öner The Influence of Polyelectrolytes Architecture on Calcium Sulfate Dihydrate Growth Retardation, J. Crystal Growth, 186:427 (1998). 12. Z. Amjad, Performance ofInhibitors in Calcium Fluorite Crystal Growth Inhibition, Langmuir, 9:597 (1992). 13. E.R. McCartney, A.E. Alexander, The Effect of Additives Upon The Process of Crystallization, J. Colloid Interface Science 13:383 (1958). 14. M.C. van der Leeden, G.M. van Rosmalen, Effect of The Molecular Weight of Polyphosphinoacrylates on Their Performance in BaSO4 Growth Retardation, J. Crystal Growth 100: 109 (1990). 15. B.R. Smith, A.E. Alexander, The Effect of Additives on The Process of Crystallization II. Further Studies on Calcium Sulphate, J. Colloid and Interface Science 34:81 (1970).

205

This page intentionally left blank.

AN ALGICIDE USING ANIONIC POLYMERS TO SEQUESTER AND STABILIZE COPPER IN AN OXIDIZING AQUEOUS ENVIRONMENT

John P. Garris Technology & Development BioLab, Inc. Decatur, Ga 30030

INTRODUCTION Copper has been used for decades to kill algae in water systems such as swimming pools and spas.1 Unfortunately, copper can easily precipitate in alkaline or near-alkaline water conditions1,2 as well as in oxidizing conditions.3 In these environments, copper can precipitate as the insoluble oxide, hydroxide, and/or carbonate, salts. These precipitated salts reduce or eliminate the biocidal activity of the copper in solution. Additionally, these insoluble copper salts can precipitate directly onto swimming pool surfaces as an unsightly blue or black stain that is difficult to remove.3 In order to improve the efficacy of copper to act as an algicide in water, sequestering agents, such as alkanolamines, aminocarboxylic acids or citric acid, have been used to improve the stability of the copper in aqueous solution conditions.1,2,4,5 These simple, organic compounds contain amine, hydroxyl, and/or carboxyl functionalities that exhibit sequestration capacity for the divalent copper cation. However, problems remain because alkanolamines, aminocarboxylic acids, and citric acid may deteriorate in the presence of halogen oxidizers, bacteria, sunlight and heat that are present in the swimming pool environment. The net result in pools, spas, or other water systems is the loss of copper's biocidal activity as well as potential staining to swimming pool surfaces. Improving the stability of copper in solution would provide a great benefit to customers.

MATERIALS AND METHODS Copper Sources Copper was measured as elemental copper in each test system. Each copper source was applied on an equivalent weight basis as elemental copper. Commercial grade copper sulfate

Adv ances in Crystal Growth Inhibition Technologies Edited by Amjad. Kluwer Academic/Plenum Publishers. New York, 2000

207

pentahydrate (CuSO4) was used as a control of copper alone. Copper as a copperethanolamine complex was provided using a commercially available formula from BioLab, Inc. that contains 7.1% by weight elemental copper as a copper-triethanolamine complex and 2.5% by weight N-alkyl dimethyl benzyl ammonium chloride (Cu-TEA). A new formulation containing copper and a polyelectrolyte polymer (Cu-PA) was tested against these commercial copper sources. Polyelectrolyte polymers were identified as candidate compounds to prevent the precipitation of copper in swimming pools. Polyelectrolyte polymers were chosen since they are well known for binding transition metals such as iron and copper7,8 and are stable in high temperature and high halogen conditions.9 Some example monomers that can be made into suitable polymers or copolymers are shown in Figure 1 .9,10

Maleic Acid (MA)

Acrylic Acid (AA)

Styrene Sulfonic Acid (SS)

Methacrylic Acid (MAA)

Acrylamido-methyl Propane Sulfonic Acid (AMPS) Hydroxy-alkyl Propane Sulfonic Acid (HAPS)

Figure 1. Example monomeric units used to produce some water treatment polymers or copolymers.

Copper Persistence in the Absence of Chlorine Persistence of copper for a period is necessary in order to provide the proper dose over time to kill algae. Copper persistence was measured using a model swimming pool system. These model test systems consist of a 30 liter aquarium, a 1/100 horsepower pump, and a scaled down sand filter. Water from a swimming pool was introduced into each aquarium test

208

system. The water introduced to the test systems was balanced using the American National Standards Institute (ANSI) and National Spa and Pool Institute (NSPI) guidelines as shown in Table 1 with the exception of chlorine. Table 1. Relevant ANSI/NSPI standard water conditions for swimming pools. Water Parameter

Optimal Range

Calcium Hardness

200 - 400 ppm

Total Alkalinity

80 - 120 ppm

pH

7.4 - 7.6

Chlorine

* Free - 1-3 ppm Total - 1-3 ppm Combined - < 0.2 ppm

Bromine

* 3-5 ppm

Metals

0 ppm

* Chlorine and Bromine ranges are consistent with those set by the U S. Environmental Protection Agency (USEPA)

Copper was introduced to the test system at 1.25 mg/L as copper (ppm). Copper was initially measured 15 minutes after introduction of each test substance to insure proper mixing by the circulation of water through the pump and filter. The circulation system operated continuously during the test. Subsequent copper tests were performed at 24 and 48 hours. Copper was measured using a HACH DR/3000 spectrophotometer and CuVer®1 copper reagent powder pillows. These prepackaged reagents are based on the bicinchoninate method for copper measurement.

Chlorine Persistence & Copper Staining Using the model swimming pool system previously described, the halogen stability as well as the propensity of the copper compounds to stain swimming pool surfaces was measured. Using materials consistent with pool manufacturing, plaster plates were produced to simulate a pool’s surface. These plaster plates were cured for 2-3 days in a holding tank. The pH was measured regularly and hydrochloric acid was added to neutralize pH increases caused by the plaster curing process. A cured plaster plate was then placed in the bottom of the aquarium test system. The test system was filled with tap water. After allowing the water to circulate for several minutes, a sample was taken and tested for calcium hardness, total alkalinity, free chlorine, total chlorine, copper and pH. The test system was balanced per Table 1. Once the water was balanced, copper was introduced at 2.0 ppm as copper using copper sulfate pentahydrate, Cu-TEA, and Cu-PA. The tanks were allowed to circulate for 30 minutes. All of the pool water parameters were tested. The tanks were turned off for approximately five minutes to allow the water to stop moving. Halogen was added in the form of commercial grade dry calcium hypochlorite at a level of 8 to 10 ppm. Calcium hypochlorite is routinely added at this level to swimming pools as part of an algae treatment or as part of a maintenance program. The calcium hypochlorite was added directly to each tank onto the surface of the plaster plates. This scenario simulates a typical staining scenario in which calcium hypochlorite settles to the bottom of a pool. After 60 minutes a water analysis was performed and the pump was returned to normal operation.

209

The test system was monitored over the next three days with subsequent calcium hypochlorite additions at 24,48 and 72 hours as previously described. Any stains were identified visually. Acute Toxicity Toxicity tests were performed on the Cu-TEA formula and the Cu-PA formula under Good Laboratory Practices (GLP) Regulations11 as per USEPA published procedures. The standard toxicity tests performed are cited by reference Acute Oral Toxicity in Albino Rats,12 Acute Dermal Toxicity in Albino Rats13, Primary Eye Irritation in Albino Rabbits,14 and Primary Skin Irritation in Albino Rabbits.11 Classification of the formulas is based on USEPA Pesticide Label Review Manual.16 The scale is from 1 to 4 with 1 being most severe and 4 being least severe.

RESULTS & DISCUSSION Copper Persistence in the Absence of Chlorine Three replicate analyses were performed using the three copper containing materials. The results in Figure 2 show copper sulfate pentahydrate quickly looses activity in balanced pool water. The chelated forms of copper significantly increase the persistence of copper, thus allowing the copper to act as an algicide for longer periods of time. There does not seem to be much difference between Cu-TEA and Cu-PA.

Hours Figure 2. Persistence of copper is measured by the loss of copper in balanced swimming pool water as a function of time. (mean of 3 replicate samples with ± I std. dev. error bars)

The results show that copper sulfate pentahydrate is rapidly lost from solution compared with the complexed forms, Cu-TEA and Cu-PA. The lack of copper persistence in balanced pool water confirms the necessity of a complexing agent. The difference between Cu-PA and CuTEA is negligible.

210

Chlorine Persistence & Copper Staining Chlorine is utilized in the vast majority 90%) of swimming pools to kill bacteria and algae. When algae become resistent to chlorine or chlorine levels become low, an algicide may be used alone or in conjunction with chlorine. The algicide used should be stable in this oxidizing aqueous environment. Five replicate analyses were performed on the chelated copper compounds, Cu-TEA and CuPA. Figure 3 shows that a free chlorine level could be maintained in the test system containing Cu-PA; however, the test system containing Cu-TEA looses free chlorine during each addition of calcium hypochlorite. Additionally, blue stains were evident on the plaster plate in the CuTEA test system. No stains were evident on the Cu-PA plaster plate. Copper sulfate pentahydrate was tested as a control which upon addition of calcium hypochlorite provided a large black and blue stain on the plaster plate.

Hours Figure 3. Stability of free chlorine in the presence of Cu-PA and Cu-TEA formulas over a 73 hour period. (mean of 5 replicate samples with ± 1 std. dev. error bars)

At time zero the test system chlorine level is determined and 8 to 10 ppm of chlorine is added to the system. The measurement at 1 hour provides chlorine concentration after the 810 ppm chlorine addition has initially dissolved. Additional applications of calcium hypochlorite are made after chlorine determinations at 24, 48, and 72 hours. Additional measurements are made one hour, 25, 49, and 73 hours respectively, after the subsequesnt calcium hypochlorite additions.

211

The series of additions up to 72 hours adds approximately 24-30 ppm of free chlorine to the system. Of that total, as much as 71% is still measureable in the Cu-PA system. Less than 0.3% is measureable in the Cu-TEA system. Figure 3 shows that free chlorine is stable in the presence of Cu-PA and chlorine is not stable in the presence of Cu-TEA. The stair-step Cu-PA line suggests no substantial loss at each chlorine addition; whereas, Cu-TEA drops practically to zero after each chlorine addition. The lack of staining present on the Cu-PA plaster plate suggests chlorine stability while the blue staining that occurred on the Cu-TEA plaster plate is indicative of free copper precipitation.

Acute Toxicity Results from the acute toxicity testing are displayed in Table 2. Cu-PA has a lower acute oral toxicity towards albino rats (3571 mg/kg) than Cu-TEA (786 mg/kg). This difference does not change the classification; class 3 ranges from 500 to 5000 mg/kg. Cu-PA shows a substantially lower acute dermal toxicity towards albino rats (>5000 mg/kg) than Cu-TEA (> 277 & < 577 mg/kg). The difference in the dermal results place the materials in class 3 and class 2, respectively. The Draize dermal irritation score for Cu-PA was 0.0 out of 10.0. This score substantiates that Cu-PA has very little dermal irritation. The Draize dermal irritation score for Cu-TEA was 6.79 out of 10.0. This score suggests Cu-TEA is a strong irritant. The 0.0 score gives Cu-PA a class 4 rating while the 6.79 score gives Cu-TEA a class 2 rating. Because of the low Cu-PA dermal irritation score, a Primary Eye Irritation study was performed. The Draize ocular irritation score was 17.2 out of a 110 maximum score. This result gives the material a class 2 rating. A Primary Eye Irritation study was not performed on Cu-TEA since the high dermal irritation score would inflict severe ocular damage. The severity of the Dermal Irritation study supports a class 1 rating. Table 2. Results from acute toxicity testing. Toxicity Test

CU-PA

Acute Oral

3571 mg/kg

Acute Dermal

> 5000 mg/kg 0.0 Draize score (max score = 10)

Primary Dermal Irritation

*Class 3

Cu-TEA

*Class

786mg/kg

3

3

> 277 & < 577 mg/kg

2

4

6.79 Draize score (max score = 10)

2

Primary Eye 17.2 Draize 2 N/D based on result of Primary 1 Irritation (max score = 1 10) Dermal * USEPA Classification from EPA Label Review Manual, 1996, scale from 1 to 4; 1 being most severe and 4 being least severe.

The results show that Cu-PA has a superior toxicity profile compared to Cu-TEA. This toxicity profile lowers the risk of consumer exposure using the algicide.

CONCLUSION The copper ion is a well known and effective algicide, but has some performance issues related to known swimming pool water parameters. The various performance tests for copper

212

persistence, chlorine stability in the presence of copper, stain propensity, and acute toxicity demonstrate that the Cu-PA is an improvement over the Cu-TEA formula and copper sulfate alone, These desirable performance improvements should allow people to use Cu-PA in conjunction with chlorine as a swimming pool algae treatment with reduced risk of surface staining.

ACKNOWLEDGEMENTS I would like to thank my colleagues at BioLab, Inc. for their assistance in putting this paper together most especially Christopher Reed for his help with the Chlorine Persistence & Copper Staining work as well as Tom Lachocki, David Purdy and Kirk Mitchell for their guidance and help. I would also like to thank Tom Kern with WIL Research Lab for performing the toxicity studies on the Cu-PA formula. REFERENCES 1. B. Domogalla, Killing algae in water with copper alkanolamine salts. U. S. Patent No. 2,734,028. (1956). 2. R.K. Raman and B.C. Cook, Guidelines for applying copper sulfate as an algicide: Lake Loami field Study, Illinois Dept. of Energy and Natural Resources. (1988). 3. P.K. Mitchell, The Proper Management of Pool and Spa Water, Hydrotech Chemical Corp., Marietta, Ga (1988). 4. D.E. Seymour, G.A. Seymour and M.J. Jaber, Method of preparing a copper complex for use as an algaecide. U. S. Patent No. 4,324,578. (1982). 5. J.A Swader and W.Y. Chan, Citric acid enhancement of copper solubility and toxicity in bicarbonate solutions, Pest. Biochem. Phys. 5:405 (1975). 6. U. S. Patent No. 5,149,354 7. S.K. Chatterjee, F.H. Rajabi, B.V. Farahani, and N. Chatterjee, Investigations of polyelectrolyte-transition metal ion association in mixed solvents, Polymer Bull. 27: 179 (1991). 8. W.M. Hann and S.T. Robertson, Control of iron and silica with polymeric dispersants, International Water Conference Technical Paper Number IWC-90-29. (1 990). 9. I. Brase and J. Belcher, Laboratory development of novel multi-functional polymers for cooling water use, Cooling Tower Institute Technical Paper Number TP91-09. (1991). 10. R.W. Zuhl and Z. Amjad, The role of polymers in water treatment applications and criteria for comparing alternatives, Association of Water Technologies Technical Paper A WT-F-93. (1 993). 1 1. Code of Federal Regulations 40 Parts 160 and 792 (1993). 12. Environmental Protection Agency, Guidelines for registering Pesticides in the U.S. (Pesticides Assessment Guidelines, Subdivision F, Hazards Evaluation: Human and Domestic Animals, Section 81-1. 13. Environmental Protection Agency, Guidelines for registering Pesticides in the U.S. (Pesticides Assessment Guidelines, Subdivision F, Hazards Evaluation: Human and Domestic Animals, Section 81-2. 14. Environmental Protection Agency, Guidelines for registering Pesticides in the US. (Pesticides Assessment Guidelines, Subdivision F, Hazards Evaluation: Human and Domestic Animals, Section 81-4. 15. Environmental Protection Agency, Guidelines for registering Pesticides in the US. (Pesticides Assessment Guidelines, Subdivision F, Hazards Evaluation: Human and Domestic Animals, Section 81-5. 16. Environmental Protection Agency, Pesticide Label Review Manual (1996).

213

This page intentionally left blank.

RATIONAL DEVELOPMENT OF NEW COOLING WATER CHEMICAL TREATMENT PROGRAMS FOR SCALE AND MICROBIAL CONTROL

Kostas D. Demadis, Bo Yang, Paul R. Young, Dmitri L. Kouznetsov, and Douglas G. Kelley Nalco Chemical Company Global Water Research One Nalco Center Naperville, IL 60563-1 198

INTRODUCTION Crystal growth is part of a more extensive problem in supersaturated cooling waters known as scale formation. Scale usually refers to an intimate mixture of a multitude of sparingly soluble mineral salts. Depending on the pH, temperature and water chemistry, these salts include but are not limited to calcium carbonate, calcium phosphate, silica, silicate salts, calcium sulfate, corrosion products (iron (hydr)oxides) etc. Calcium carbonate is the predominant component of the hard and tenacious scale deposited from natural waters, especially in processes involving heat transfer. The main focus of this paper is on two major system stresses, high Ca2+ and oxidizing biocides, and under conditions common in Heating, Ventilation and Air Conditioning (HVAC) systems. The chief factor that promotes calcium carbonate scale formation is concentration of dissolved solids by repeated partial evaporation of the water.1,2 Therefore, a certain water that has no scale tendency will eventually become scale-forming when concentrated n times, where n = 2, 3, 4 and even higher. The number n is defined as cycles of concentration (COC's). Economic considerations to minimize water use require low blowdown and increased COC's. The severity of the scaling tendency of a certain water is related to the saturation index (SI), defined as follows: [Ca2+].[Ca32-] SI = KspCaCO (1) 3

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers. New York. 2000

215

[Ca2+] and [CO32-] are the activities of the corresponding free ions and KSp CaCO3 is the conditional solubility product for CaCO3. Cooling water represents the single largest portion of industrial water use. Hot processes, from air-conditioning to catalytic hydrocarbon cracking, have to be cooled. Water is the most abundant and cheapest cooling medium. The water is pumped across a metal heat exchanger in contact with the hot process. The water heats up upon cooling of the process and has to be cooled in order to be re-used for cooling purposes. This cooling of the water is usually achieved by evaporative cooling in a system known as a cooling tower. A schematic representation of a typical cooling tower set-up is shown in Figure 1.

Figure 1. Schematic drawing showing the basic components of a cooling tower.

Figure 2. The water treatment triangle model.

216

Evaporation causes an increase in dissolved solids concentration. Moreover, carbon dioxide is stripped away and the result is a pH increase. As a result, water becomes highly supersaturated with respect to mineral salts. At this highly metastable situation, alkaline earth salts (especially calcium carbonate) precipitate. This may lead to massive deposit of the scale onto critical surfaces. The metal surfaces (heat exchangers) are usually very critical pieces of equipment and great care has to be taken so scale deposits on them can be avoided. Further information on the operation of cooling towers from an engineering perspective can be found in numerous specialized reports.2-7 Calcium carbonate is an especially problematic scale because it possesses “inverse solubility”, i.e. it becomes less soluble at higher temperatures. Because of that, it tends to accumulate and deposit onto hot surfaces, where it is least desired. CaCO3 tends to form tenacious scale that strongly adheres to heat transfer surfaces. It can be cleaned by acid, but this involves a corrosive procedure and results in operations downtime. Mineral scale deposits are not the only issues that adversely affect the proper operation of a cooling tower and, by extension, of the hot process. Corrosion and microbiological growth are equally important factors that have to be controlled along with scale formation. These are the three most important challenges in cooling water treatment (Figure 2).Details on corrosion and bacterial growth are beyond the scope of this paper and will be subjects of future reports.

Figure 3. Schematic molecular structures of some representative calcium carbonate inhibitors. The symbol abbreviations are as follows: HEDP, 1 -hydroxyethylidene-1,1 -diphosphonic acid; PBTC, 2-phosphonobutane-l,2,4-tricarboxylic acid; HMDTMP, hexamethylene-N,N’,N’,N’-diamine tetramethylenephosphonic acid; DETMP, diethylene triaminepenta (methylenephosphonic acid); PAA, polyacrylic acid; PESA, polyepoxysuccinic acid; PAPEMP, polyaminopoly-ethylenephosphonic acid.

217

There are four basic approaches for calcium carbonate scale control: (a) use of acid; (b) use of sequestrants; (c) maintain low cycles of concentration; and (d) use of scale inhibitors. Use of acid is an effective way to prevent calcium carbonate formation by maintaining a low pH. However, it raises concerns due to its hazardous nature and the potential of accelerating metal corrosion. The use of sequestering agents is too expensive for recirculating water applications. Maintaining low cycles of concentration leads to rather costly water waste. Scale inhibitor use is the most widely accepted practice and involves the addition of minute amounts of organic compounds (usually at the ppm level) in the recirculating water. These compounds usually are phosphonates and carboxylates, and can be monomeric or polymeric. Representative schematic structures are shown in Figure 3. These molecules function at threshold levels, since the Ca:inhibitor concentration ratio is extremely high. It is believed that they perform “threshold inhibition” by a surface adsorption mechanism involving Langmuir adsorption. Adsorption onto the CaCO3 surface causes inhibition (or delay) of crystal growth at the very early stages. Several studies have been reported on this adsorption chemistry for CaCO3 and other scales. In water treatment programs, other components are often formulated with the scale inhibitors. For example, yellow metal corrosion inhibitors, dispersant polymers and tracers can be parts of the same treatment, whereas biocides for microbial control, such as chlorine or bromine, are fed separately.

Mechanism of Scale Formation As noted above, scale formation is a complex combination of a number of processes.8-11 Initial electrostatic interaction of dissolved anions and cations is the genesis of ion pairs that sequentially act as “building blocks”. They later assemble to form larger aggregates. Externally added molecules can interfere with thisassembly process and disrupt subsequent formation of larger particles. In the absence of such molecules, generation of these larger particles is inevitable. Supersaturation is the driving force for the enlargement and growth of these aggregates to a “critical size”. At that time, nucleation of a solid state particle occurs. Contrary to the smaller-size aggregates, these larger particles do not easily re-dissolve, and therefore serve as growth surfaces for crystal growth. The above process is called homogenous nucleation and refers to particle formation in the bulk of the solution. On the other hand, heterogenous nucleation relates to generation of nuclei onto a surface (e.g., a metal surface in contact with the supersaturated water). Following the nucleation event (either homogenous or heterogeneous), several deposit growth mechanisms can occur, which involve one or more of the following: (a) crystal growth, that is cation-anion incorporation (e.g., Ca2+ and CO32-) directly along certain crystallographic planes of a crystallite;, (b) particle agglomeration; (c) deposit on “hot” metal surfaces; (d) biofilm-induced scale growth.

EXPERIMENTAL SECTION Crystal Growth and Structure of a Ca-AMP Precipitate AMP (40% solution in water) and CaC12.2H2O are commercial samples. Single crystals of Ca-AMP were obtained in the following way. AMP (5.0 mL, 0.5 M solution in deionized water) was acidified with a 10% HC1 solution until the pH was 1.5.

218

Subsequently, a solution of CaC12·2H2O (5.0 mL, 0.5 M in deionized water) was added under vigorous stirring, while the pH was constantly monitored. The pH dropped and was re-adjusted to 1.5 by addition of 0.1 N NaOH solution. The beaker was covered and set aside. During the crystallization process, large (0.25 × 0.20 × 0.10 mm), transparent crystals (rectangular plates) formed and were isolated by filtration. They were washed with deionized water and air-dried. The crystals were stored in an air-tight container to minimize exposure to air and humidity. Several regularly shaped, colorless, transparent crystals were selected, and submitted for single crystal X-ray data collection. Data were collected at -100 "C using a cold stream of evaporated liquid N2 to minimize thermal motion. Standard crystallographic methods (direct methods) were used to initially locate the heavier atoms in the structure. The remaining non-hydrogen atoms were located in subsequent difference Fourier maps. Empirical absorption corrections were applied with SADABS. The ORTEP plotting program was used to computer-generate the structure.12 All atoms were refined anisotropically. Hydrogen atoms were located in the difference Fourier map and refined as well. All computations were performed by using the NRCVAX suite of programs.13 Atomic scattering factors were taken from a standard source14 and corrected for anomalous dispersion.

Turbidity Test (Ca Tolerance) Inhibitor tolerance towards calcium was measured by turbidity monitoring. The test set-up used a constant temperature water bath (130 ± 0.5 F, 54.4 °C), two precision syringe pumps (for the inhibitor and 0.1 N NaOH solution), and a Brinkman PC 700 Colorimeter with 420 nm filter (for transmittance measurements). Four hundred mL of a 1000 ppm solution of CaC12·2H2O (as CaCO3) in deionized water was placed in a 700 ml glass beaker. The beaker was covered with a rubber piece holding the turbidity and pH probes, and a thermometer was placed in the water bath. Appropriate time was allowed for temperature and pH to equilibrate. Then, a 2000 ppm (as actives) solution of inhibitor was pumped in at a rate of 0.25 ml/min. A constant pH of 9 ± 0.05 was maintained by addition of caustic and controlled by a PHCN-37 controller (from Omega Engineering). The transmittance response and maintained pH value were continuously recorded using a PC computer with a PCL-812 PG data acquisition board. When the concentration of the inhibitor in the beaker exceeded a certain limit, the solution became turbid, which was detected by a steep decrease in light transmittance. Data acquisition was continued for several minutes to precisely determine the inflection point. Calcium tolerance of an inhibitor was determined as the ratio of ppm inhibitor to ppm Ca2+ as CaCO3 in the solution at the inflection point adjusted to 1000 ppm Ca2+ (as CaCO3). Reproducibility of results was within 10% error.

CaCO3 Solubility Stress Tests pH Static Test. In this test, the desired amounts of CaCl2·2H2O and MgSO4·7H2O (from a 40,000 ppm stock solution) were first dissolved into a volumetric flask containing 60% of the final volume de-ionized water. Next, the inhibitor or inhibitor formulation was added to the flask. Another 30% of the final volume of de-ionized water was added into the flask and mixed well with the original solution. Then the desired amount of NaHCO3 (from a 40,000 ppm stock solution) was added into the solution. Finally, de-ionized water was added to the flask until the total final volume is reached. Depending on the test analysis

219

requirements, the final volume of the test solution was between 100 – 500 ml. The solution was then transferred to an Erlenmeyer flask. The flask was covered and put in a water bath set at the specified temperature. The solution inside the flask was under constant stirring with a magnetic stirring bar. pH 9.0 was maintained by addition 0.1 N NaOH via an autotitrator. After a chosen time period (2 –3 hr), the flask was removed from the water bath, and 0.45 µ filtered solution samples were taken immediately. Analyzing these samples yielded the soluble concentration of Ca2+. The remaining solution was covered and stored unstirred at room temperature. Twenty-four hours after the pH was first raised to 9.0, unfiltered solution samples were taken from just below the surface. The analytical results of these samples yielded the dispersed Ca2+ concentration. Aeration Test. In this method, the experimental procedure was similar to the one described above as "pH Static test", except that aeration was used to raise the solution pH. This was achieved with a constant flow of air through a sintered glass diffuser. The solution samples were taken and analyzed identically as in the pH Static test. It should be pointed out that the aeration test appears to be a better simulation of a commercial open recirculating cooling water system than the pH static test. Just as such systems, the solution pH depends on soluble M-alkalinity and other solution species. In addition, the amount of M-alkalinity added into each test solution will be well defined, in contrast to the pH static test where the amount of NaOH added depends on the performance of the inhibitor used. The worse an inhibitor performs, the more NaOH is added to the solution.

Pilot Cooling Tower (PCT) Tests Pilot cooling tower (PCT, Figure 1) tests were run as described by Nass & Reed and Fulks and Yeoman.15,16 Heat exchange tubes were 304 stainless steel, admiralty brass, copper, or carbon steel (mild steel). Most tests had two tubes of each metal, or three each SS and MS with no copper. Tubes were heated with electrical resistance heaters of 200 or 400 W power. Because the Heating, Ventilation and Air Conditioning (HVAC) systems usually use mild steel pipes to transport water, and use copper or brass as the prefered heat exchanger metallurgy, the mild steel tubes were not heated to better simulate this situation. Water flow was about 9 L/min, giving a velocity of 1m/s. The tubes’ surface temperatures are estimated to be about 5 - 8 °C higher than the water temperature. Parameters varied in this work were temperature, biocide, and treatment level of each component. Basin temperatures were controlled at either 27 or 32 ± 1 °C, with a ∆T of 2-3 °C. Formulations containing various amounts of scale inhibitors (AMP, PBTC, HEDP, PAA) and dispersant polymer were prepared. Many of these contained a tracer and/or a triazole corrosion inhibitor. Some formulations included a surfactant as a dispersant. A conductivity monitor controlled a blowdown pump, initiating blowdown whenever the conductivity exceeded the setpoint. After several days at a given conductivity with no scale observed on the tubes, the conductivity setpoint was increased by 100 - 300 µmhos. Daily chemical water analyses of the recirculating tower water were performed. Deposit formation on heated surfaces is monitored during each test visually and with a DATS fouling monitor manufactured by Bridger Scientific Co. This unit determines the resistance to heat transfer (HTR) across a precisely controlled heated surface. By measuring water temperature and heater block temperature, while controlling heat flux and water velocity, the DATS computer calculates a relative HTR every 30 seconds. When HTR increases, a deposit of some type has formed an insulating layer on the surface.

220

Commercial corrosion monitors, Corraters from Rohrback/Cosaco Co. provided continuous measurements of corrosion rates. These agreed well with visual and gravimetric data. Most formulations had a fluorescent tracer for better program control. For products with fluorescent tracers, an on-line fluorometer (TRASAR® technology) was used to measure and maintain product dosage.17 Other products were fed via calibrated syringe pumps activated whenever blowdown was turned on. These syringe pumps were set to deliver the calculated volume of product per liter of blowdown, given the known (usually measured independently) blowdown pump rate. A computer data acquisition system recorded the output from numerous on-line instruments. Conductivity, pH, basin and return temperatures, and outputs from fluorometer, Corrater and DATS were stored every 30 minutes. Oxidizing biocides were fed via a syringe pump at a constant rate, but the rate was occasionally increased or decreased in order to maintain the target residual. Most tests used either stabilized bromine at 0.2-0.8 ppm (as total Cl2) or stabilized chlorine at 1-5 ppm (as total C12). Three tests used bleach at 0.1-0.5 ppm total Cl2.

RESULTS AND DISCUSSION Effect of Ca2+ Stress Effect of High Ca2+ on Inhibitors. Calcium tolerance is defined as the ability of a certain chemical compound to remain soluble in the presence of calcium ions. Tolerance usually decreases as pH increases. This is because at higher pH’s the degree of deprotonation of inhibitors (usually phosphonates or carboxylates) is higher. Calcium tolerance becomes very critical as COC’s increase. An efficient inhibitor must interact strongly with Ca2+, but must be sufficiently soluble to remain in the system. A Cainhibitor precipitate is a scale by itself. Such precipitates have adverse effects on the whole treatment program because: (a) the system is depleted of inhibitor, resulting in poor scale control and higher chemical costs; (b) such Ca-inhibitor scales usually have the same inverse solubility features commonly observed with scales; (c) they impede heat transfer. Ca-inhibitor salt precipitation is well known for phosphonate as well as carboxylatebased inhibitors. Results obtained for AMP, PBTC, and HEDP phosphonates are shown in Figure 4. The above inhibitors can be rated according to their Ca tolerance in descending order (as ppm soluble inhibitor per 1000 ppm of Ca2+ as CaCO3): PBTC (1 85 ppm) > AMP (12 ppm) > HEDP (8 ppm). The presence of a polymeric dispersant has a profound effect on Ca-inhibitor salt precipitation. In one of the series, 20 ppm of Dispersant A was added to the calcium solution before HEDP addition started. As shown in Figure 5, the dispersant alleviated CaHEDP salt precipitation and effectively increased its solubility ~ 200 %. Finally, turbidity was noted at 19 ppm HEDP. Crystal and Molecular Structure of a Ca-AMP Complex. A better understanding of Ca-inhibitor interactions in the solid state may give insights to the inhibitory action of methylenephosphonate inhibitors. Thus, a Ca-AMP crystalline precipitate was studied by single crystal X-ray diffraction. Two representative views of different portions of the molecular structure, as well as salient metric features, are shown in Figure 6. There are no discrete molecular units of the Ca-AMP complex. Instead, the AMP methylenephosphonate “arms” participate in an intricate network of Ca-O bonds and hydrogen bonds.

221

The result is an intricate polymeric three-dimensional structure characterized by multiple bridging of the AMP molecules. There are four molecules of Ca-AMP per unit cell. The Ca:AMP molar ratio is 1:1 and is consistent with the absence of any other ions in the crystal lattice. The overall negative charge on the AMP is “–2” (“–3” from the phosphonate groups and “+1” from the protonated N). The crystal lattice of Ca-AMP is hydrated and contains overall 31/2 water molecules per one molecule of AMP. Water molecules of crystallization serve as “space fillers” and also participate in extensive hydrogen bonding superstructures, acting as bridges. Detailed crystallographic data on the Ca-AMP structure are available from the authors (KDD) upon request. Coordination Environments: Ca Atom. The coordination environment of the Ca atom is shown in Figure 6. Ca is pseudo-octahedral, surrounded by five phosphonate oxygens and a water molecule. Ca-O(P) bond lengths range from 2.2924(14) to 2.3356(14) Å The Ca-O(H2O) bond is 2.3693(17) Å long, somewhat longer than Ca-O(P) bonds. The Ca atom is situated in a slightly distorted octahedral environment, as judged by the O-Ca-O angles. Average Ca-O(P) bond lengths are comparable to similar bonds found in the literature.

ppm phosphonate per 1000 ppm Ca2+ as CaCO3 Figure 4. Precipitation of Ca-Inhibitor complexes as monitored by turbidity measurements. Results are presented as ppm phosphonate added to a 1000 ppm Ca2+ solution, at 130 °F and pH 9.

PO3 Groups. All three phosphonate groups in AMP are mono-deprotonated. This separates the P-O bonds into two groups: P-OH (protonated) and P-O ¯ (deprotonated). There is one P-OH group and two P-O groups per phosphonate moiety. The two P-O

222

ppm phosphonate per 1000 ppm Ca2+ as CaCO3 Figure 5. Effect of Dispersant A on the precipitation of Ca-HEDP as monitored by turbidity measurements. Results are presented as ppm phosphonate added to a 1000 ppm Ca2+ solution, at 130 °F and pH 9.

Figure 6. Molecular structure of Ca-AMP and important metric features.

223

groups share a single negative charge. Deprotonated P-O bond lengths are in the region 1.4931(15)-1.5102(14) Å On the other hand, protonated P-O bond lengths are slightly longer, in the region 1.5684( 15)-1.5801(14) Å It is worth noting that only the deprotonated P-O groups coordinate to the Ca atoms. P-C bond lengths fall in the normal range (1.8-1.9 Å) and are 1.8382(20) Å, 1.8347(20) Å and 1.8330(20) Å

Literature Cases and Comparisons. There is a plethora of crystal and molecular structures of mono- phosphinates18 and mono- phosphonates19,20 in the crystallographic literature. In addition, an extensive number of structure s of bis phosphonates21,22,23 exist, but only a limited number of reports on Ca2+ - bis -phosphonate complexes.24,25,26 Furthermore, ab initio studies on relevant organophosphorous compounds and their Ca complexes have been carried out.27 Substantial interest has been focused on 1,1- bis phosphonates (or gem- bis-phosphonates).28 On the other hand, two neighboring (geminal) phosphonate groups are not a necessary requirement for good scale inhibition. AMP and PBTC are both excellent CaCO3 inhibitors. A brief review of representative literature cases relevant to this study follows. Uchtman described the crystal structure of a Ca-HEDP complex.24 Ca is octacoordinate bound by two H2O molecules, two phosphonate oxygens, and the protonated hydroxyl group. The structure is polymeric. Ca-O(PO3) bond lengths are 2.352(4) and 2.420(3) Å, whereas the Ca-O(H) bond is much longer, at 2.608(3) Å In a similar Rb-HEDP complex Rb-O(PO3) bond lengths are 2.949(3) Å and 2.952(3) Å whereas the Rb-O(H) bond is much longer, at 3.078(3) Å.29 Nardelli et al. described the crystal structure of calcium dichloro-methylene diphosphonate.25 Phosphonate groups are monodeprotonated, and there is extensive hydrogen bonding. Ca is heptacoordinate and is bonded to 5 water molecules and two phosphonate O’s from the same chelating dichloro-methylene diphosphonate. Ca-O(P) bond lengths are 2.362(3) Å, and Ca-O(H20) bond lengths are in the region 2.418(4)2.428(4) Å. The Ca salt of N -(phosphonomethyl)glycine (glyphosate) was studied by Raymond et al.19 The structure is polymeric. The Ca is seven-coordinate with four oxygen atoms from three different glyphosate groups, one carboxylate oxygen from another glyphosate and two water oxygens. Each glyphosate is in turn bonded to four different Ca atoms through both the phosphonate and carboxylate groups. The N atom is protonated and therefore does not participate in any Ca binding. One phosphonate group from the glyphosate ligand acts as a bridge between two Ca centers and brings them to a distance of 3.93 Å apart. Ca-O(P) bond lengths vary. Ca-O(bridging) is 2.51 1(1) Å long, much longer than a “normal” Ca-O bond of 2.368(1) Å in the same molecule. Mathew et al. reported the crystal structure and spectroscopic properties of a Ca complex with a novel bis phosphonate, glutarylbis(phosphonate), G1BP.26 This is a phosphonate that has an acyl group next to the phosphonate group. The structure can be described in terms of a covalently pillared layer-type arrangement of neutral Ca-GlBP-Ca units along the b-axis. Each oxygen atom of the PO3 group is bonded to a different Ca ion, and each Ca in turn is, linked to three PO3 groups. Ca-O(P) bond lengths are ~ 2.38 Å Complex formation equilibria studies of amino phosphonates exist.30-33 More specifically, Sawada et al. reported studies on Ca-AMP complex formation.34 They suggested that AMP binds to alkaline earth metal ions in a tetradentate fashion (presumably at high pH’s), through three phosphoryl O’s (originating from three different PO3’s) and the central N atom. As the pH decreases, they propose that the tetradentate

224

AMP becomes tridentate because the central N is protonated, and gets “pushed-back,” away from the Ca center. The proposed structure was said to be consistent with the one proposed previously, which was based on infrared spectroscopy.35

Calcium Carbonate Scale Inhibition: Screening Methodology. Soluble and dispersed Ca, obtained by the pH Static test method as a function of PBTC concentration, are shown in Figure 7. The corresponding results obtained by the aeration method are shown in Figure 8. The test solutions contain 800 ppm Ca2+, 200 ppm Mg2+ and 800 ppm HCO3-. Tests were carried out at 110 °F. Results obtained from both methods show that in general, as the PBTC concentration increases beyond some threshold value ( >3 ppm based on the aeration method), the soluble and dispersed Ca concentrations increase as well. The increase is rapid up to 10 ppm PBTC, then begins to level off, finally reaching a plateau at 40 ppm PBTC.

PBTC (ppm actives) Figure 7. CaCO3 inhibition in the presence of various dosages of PBTC. Results were obtained with the pH static test. Conditions were: 800 Ca, 200 Mg, 800 NaHCO3 (all in ppm as CaCO3), pH 9, at 110 °F.

The results also show that the aeration method yielded a much higher soluble/dispersed Ca than the ones obtained by the pH static method. This can be explained by the potential presence of local high pH regions as the caustic is added to the Ca solution, which would lead to pronounced precipitation of CaCO3 in such regions. In addition, in the pH static test as CaCO3 precipitates more base is added to maintain the pH. This increases the alkalinity, forcing additional precipitation.

Inhibitor Performance. It is rather difficult to address all factors and variables affecting CaCO3 scale inhibition at the same time in a “beaker” test. Nevertheless, a simple screening test can provide useful performance and kinetic information on inhibition of CaCO3 scale. The aeration method was used to study the inhibiting properties of PBTC, HEDP, AMP, polyacrylate (MW = 5000), dispersant A (5 - 20 ppm actives) in solutions containing 600

225

PBTC (ppm actives) Figure 8. CaCO3 inhibition in the presence of various dosages of PBTC. Results were obtained with the aeration test. Conditions were: 800 Ca, 200 Mg, 800 NaHCO3 (all in ppm as CaCO3), at 110 °F.

(MW=5000)

Figure 9. CaCO3 inhibition in the presence of various dosages of inhibitors. Conditions were: 600 Ca, 200 Mg, 600 NaHCO3 (all in ppm as CaCO3), at 110 °F. The second and third bars (a and b) are duplicate trials.

226

ppm Ca2+, 200 ppm Mg2+ and 600 ppm HCO3-. These results are shown in Figure 9. Results obtained at 10 ppm inhibitor are duplicate tests. In general, the aeration test shows good reproducibility with error being ± 5 – 9 %. Based on the results obtained under the test conditions, CaCO3 scale inhibition performance has the following sequence at 5 ppm inhibitor actives. HEDP ~ AMP > PBTC > Polyacrylate >> Dispersant A ~ blank When the inhibitor concentration is increased to 10 ppm the performance ranking is PBTC > HEDP ~ AMP > Polyacrylate >> Dispersant A Increasing inhibitor concentration to 20 ppm seems to have little added effect.

Effect of Dispersant/Phosphonate Blends in CaCO3 Inhibition. The effect of Dispersant A on inhibitor performance was also studied with the aeration test. Inhibitors such as PBTC, HEDP, AMP, polyacrylate (MW = 5000) in concentrations 10 or 30 ppm actives were added to solutions containing 800 ppm Ca2+, 200 ppm Mg2+ and 800 ppm HCO3-. Based on soluble Ca measurements with and without 10 ppm Disperant A, the inhibition performance sequence established at lower Ca and M alkalinity stress remains largely the same (except that at 10 ppm polyacrylate performs better than 10 ppm AMP). It appears that the presence of 10 ppm Dispersant A improves the performance of 30 ppm AMP and HEDP noticeably under the test conditions. Little or no effect was observed on PBTC and polyacrylate. This suggests that the less soluble inhibitors benefit more from the dispersant. After 24 hours, the solutions were again tested (unfiltered) for Ca. These results give dispersed Ca and are presented in Figure 10. It is noticeable that Dispersant A has little effect on PBTC and polyacrylate. It again improves the performance of AMP and HEDP. We suggest that this is related to the ability of Dispersant A to disperse insoluble Ca-HEDP and Ca-AMP complexes that may form under the test conditions. The effect of dispersant polymers on inhibition performance by blends of phosphonates was also studied by the aeration method. The water used contained 800 ppm Ca2+, 200 ppm Mg2+, and 800 ppm HCO3- at 1 10 F. Results are shown in Figure 11. Dispersant A (15.5 ppm) seems to improve dispersed Ca in solutions containing 12 ppm PBTC and 8.4 ppm HEDP, over solutions containing Dispersant B. This effect is more pronounced compared to that of Dispersant C. It appears that different dispersant polymers impact CaCO3 scale inhibition performance in different ways. Dispersant A appears to be the most effective, whereas Dispersant C the least effective. Pilot Cooling Tower Tests and Data Manipulation. PCT tests are very useful probes for studying in detail all cooling water dynamics in a unified fashion. There are several ways to "run" a PCT. In the present study, conductivity is increased incrementally, eventually reaching the point where calcium begins to precipitate out. The conductivity or the maximum calcium/alkalinity at this point indicate the limits of the treatment. The limit is lower at higher temperature. The limit is higher at higher inhibitor levels. A total of 27 PCT tests were run. The variation of individual parameters was not varied according to a randomized factorial design. Fewer tests were run with polyacrylate, and different low molecular weight acrylate-based copolymers were used. More tests were run at the higher temperature. Several tests used identical formulations at different doses, contributing to a statistical correlation between the responses to the coefficients for the components in that formulation. Microbiological control was maintained with bleach, stabilized bleach, or stabilized bromine.

227

Figure 10. CaCO3 inhibition by PBTC, AMP and HEDP (30 ppm) in the absence and presence of Dispersant A (10 ppm). Conditions were: 800 Ca, 200 Mg, 800 NaHCO3 (all in ppm as CaCO3), at 110 °F. Dispersed Ca measurements were done after 24 h.

Some bench tests (eg., the pH static test described above) cannot differentiate between these inhibitors. They all reach a plateau of performance in the beaker tests. WaterCycleTM from French Creek Software38 was used to calculate the saturation ratios at the conditions of the conductivity limit. Qualitatively, it is clear that lower inhibitor doses have lower limits. It is of great interest to determine a more quantitative description. In particular, individual component contributions to the overall performance would allow easy optimization of product formulations. Two attempts to extract this information from the PCT data will be described. Both gave very similar results.

Phosphonate/Dispersant Blend Figure 11. Synergistic effects between various phosphonates and Dispersants . Blend 1 = 12 ppm PBTC + 8.4 ppm HEDP + 15.5 ppm Dispersant A. Blend 2 = 12 ppm PBTC + 8.4 ppm HEDP + 15.5 ppm Dispersant B. Blend 3 = 12 ppm PBTC + 8.4 ppm HEDP + 15.5 ppm Dispersant C. Conditions were: 800 Ca, 200 Mg, 800 NaHCO3 (all in ppm as CaCO3), at 110 °F. Dispersed Ca measurements were done after 24 h.

228

First, the maximum conductivity was determined (for each of 27 PCT tests) by graphically interpolating calcium and magnesium cycles as a function of conductivity. The “conductivity limit” is chosen to be where ∆Cycles (defined as MgCycles - CaCycles) first exceeds the value of 0.2. Figure 12 gives one example.

Conductivity Figure 12. Effect of conductivity increase on Ca and Mg Cycles in a PCT test. The conductivity limit in this test is ~ 1850 (Cy = Cycles).

There was a clear tendency for tests at higher temperatures to perform worse. This was as expected based on the known solubility of CaCO3. Using WaterCycleTM software, it was shown that the saturation ratio (SR) for this water increases almost exactly 1% per 1 degree F (0.556 °C). SR also is proportional to concentration cycles squared (or alternatively, conductivity squared) in the region of interest. A number, denoted as Cond’, proportional to SR but similar in magnitude to the conductivity was calculated as: Cond’=

(Conductivity) 2 × (T Factor) 2080

T Factor = 1 + 0.01 X (Tretum – 80)

(2) (3)

Where 2080 is the average conductivity limit, and Treturn, the temperature of the return water, is measured in degrees F. The resulting number for each test was fit (using JMP software) to a simple linear equation, where Cond’ is equal to: Cond0 + [a·(ppm AMP)] + [h·(ppm HEDP)] + [p·(ppm PBTC)] + [paa·(ppm PAA)] (4) The parameters a, h, p and are regression coefficients. In this treatment, for simplicity, paa refers to all the low molecular weigh acrylate-based copolymers. “Condo” was set at 1300, which is the approximate conductivity value where precipitation occurs without any inhibitors added to the system. The results were: a = 39 ± 13, h = 76 ± 21, p = 51 ± 14, paa= 41 ± 16, Adj. R2 = 0.60. For example, each ppm of AMP or PAA would be expected to raise the conductivity limit by about 40 µmhos. PBTC seems to be slightly better than AMP or polyacrylate, with

229

HEDP even better, but the differences are not much larger than experimental variability. The trend seems to be real, but the fit is less than ideal. Attempting a logarithmic or square root fit of the dose response improved the fit very little (Adj. R2 = 0.63). The ratios of parameters for each inhibitor were similar to the ratios for those calculated above. An alternate method was used to calculate the value of the “Ca × M alkalinity” for each day of the test. This was observed to level off at each conductivity set point. This leveling off, where ∆Cycles was small, was taken as indicating the maximum attainable concentration under the test conditions. The daily analyses for those days were used to calculate the SR at that point for each test. SR values were between 60 and 120. The points selected by this method coincided well with the point of the conductivity limits described above for low conductivity limits / low doses / low SR. For higher conductivity / SR ratios there was a tendency for the maximum to occur at slightly higher conductivity than the conductivity limits. This may introduce a small systematic difference between the two methods. A linear fit was poor (Adj. R2 = 0.45), but a logarithmic equation gave a good fit: log (SR-1) = (5) Const + n·log { [a·(ppm AMP)] + [h·(ppm HEDP)] + [p·(ppm PBTC)] + [paa·(ppm PAA)]} Where “Const” was set equal to the log of the SR where precipitation is observed with no treatment, and n = 0.25, a = 14.0 ± 4.3, h = 39.6 ± 10.6, p = 25.1 ± 4.4, paa = 9.8 ± 13.3, Adj. R2 = 0.78. Omitting the runs with polyacrylates (PAA) gave a similar fit, with n = 0.25, a =13.6 ± 4.5, h = 41.7 ± 12.1, p = 25.0 ± 4.6, Adj. R2 = 0.77. The value of n greatly affects the values of the other parameters. Allowing n to vary resulted in a value of 0.22, which seemed unrealistically low. Fixing n at 0.25 or 0.33 (fourth and cube roots of dosage) gave Adj. R2 values only slightly worse (0.75 and 0.73 respectively, compared to 0.78). For n = 0.5 (square root of dose), the Adj. R2 was still 0.71. The values of the fitted parameters changed with n, but the relative rankings did not. HEDP remained the best, according to this treatment of the data, followed by PBTC and AMP and PAA. Lack of fit for tests with blends of phosphonates was indistinguishable from lack of fit for tests with single phosphonates.36,37 Thus, the fit of the data was surprisingly insensitive to the exponent of the inhibitor dose. The data in Figure 8 is consistent with a square root or fourth root of dose. As a practical matter, the point is that the dose dependence of system performance is highly non-linear. Very high doses give only a little more inhibition than moderate doses. Money spent feeding excess product is largely wasted. However, slightly under-dosing the inhibitor can result in significantly higher deposit formation. This highlights the need for accurate control of product feed. l7

Inhibitor Stability Toward Oxidizing Biocides Although the primary criterion for potentially successful CaCO3 scale control is inhibition performance, secondary concerns are important as well. One of the most significant ones is inhibitor stability toward oxidizing biocides. One of the most common microbiological treatments is chlorine. Due to its strong oxidizing power, chlorine works effectively but, unfortunately, can degrade the scale inhibitors present in the water treatment program. As a result, chlorine-resistant technologies find widespread acceptance and popularity. Inhibitor degradation by oxidizing biocides is undesirable for several reasons: (a) the system is depleted from inhibitor leading to poor scale control (b) orthophosphate is one of

230

the by-products of decomposition of phosphorous-containing inhibitors and may lead to calcium phosphate scale (c) the oxidizing power of the biocide is “consumed” in decomposing the inhibitor and not to sterilize the system, which leads to (d) poor microbiological control. These problems can be minimized by use of halogen stabilizers or more halogen-resistant scale inhibitors, or an appropriate biocide-inhibitor combination. It is well known that oxidizers attack treatment chemicals.39-42 The severity and importance of this depends on system conditions, type of oxidizer, and type of treatment chemical. Benchtop testing in the present study showed that degradation of inhibitors depends on factors such as pH, [Ca2+], impurities, temperature, as well as the oxidizer itself. The two biocides most often used to control microbial growth in HVAC cooling systems are chlorine and bromine. Ozone, iodine, and non-oxidizing biocides are used much less frequently. Chlorine is less expensive per mole of oxidant, but bromine, at pH 8 - 9, is more effective for any given oxidant level. Hypobromous acid has a higher pKa than hypochlorous acid.40 This allows a higher percentage of the bromine biocide to be present in the more active acid form, while chlorine is predominantly in the less active ionized form. While phosphonates are, in general, more prone to oxidation than carboxylates, it is an oversimplification to say that all phosphonates are unstable in the presence of oxidizers. Figure 13 shows that PBTC is very resistant to oxidation. HEDP is quite resistant to chlorine, but readily attacked by bromine. AMP on the other hand, is readily attacked by chlorine, but to a lesser extent by bromine. The bromine results are in excellent agreement with the results of Vanderpool et al., who used High Performance Liquid Chromatography (HPLC) to study the decomposition of various inhibitors by bromine.43 This indicates that when chlorine is used as the biocide in a system, AMP is not a good choice for a scale inhibitor because it is rapidly degraded by chlorine. PBTC or a (less expensive) PBTC/HEDP blend would be a good choice. Indeed, such programs have been used

Figure 13. Effect of various oxidizing biocides on CaCO3 inhibitors. These data were obtained after one hour at 25 °C, pH 8.3, for solutions containing 5 ppm phosphonate and 5 ppm oxidizing biocide (as total Cl2) with the same makeup water as the PCTs.

231

successfully for many years. With bromine as the biocide, PBTC or a PBTC/AMP blend would be the best choice. In cooling towers a constant, low halogen residual is usually maintained. Therefore, it may seem that the “high-dosage” benchtop tests do not represent a practical situation. On the other hand, it is known that many operators “slug-feed” oxidizing biocides 1 - 3 times per week. In this case, high local concentrations of oxidizer may be experienced. In a related study (unpublished results), a low oxidizer residual (0.25 ppm) was maintained with an on-line chlorine analyzer in a small recirculating system. Degradation of HMDTMP (see Figure 3) was ~ 1 ppm/hour with chlorine, ~ 0.4 ppm/hour with bromine, and less than 0.03 ppm/hour for the stabilized halogens. In a PCT, phosphonate loss to oxidation can be estimated by calculating the difference between theoretical and measured organic phosphate. With PBTC and stabilized chlorine, there in no such degradation observed. With AMP or HEDP and stabilized bromine, the loss due to oxidation is discernable, but not large. An estimate of the relative reactivity of each phosphonate from the PCT data is in excellent agreement with benchtop data. It should be noted that the choice of oxidant does not affect either the program limit or the scaling rate. There were several additional experiments in which oxidizing biocides were added to PCT’s over a 10-20 minute period (slug-feed). Slugs of 2 - 4 ppm (as C12) had no detrimental effect on organophosphonate or PO43- (measured 16 - 40 hours after the slug). They also had no effect on loss of Ca ∆Cycles), and on deposition (visual or Heat Transfer Resistance). The only time such effects were observed was during an upset of over 20 ppm stabilized bromine (as total C12) for several days. HEDP decreased from 2.0 to 1.0 ppm, and PO43- increased from 0.2 to 1 .0 ppm during the same period. ∆Cycles increased from 0.6 to 1 .0, indicating increased CaCO3 precipitation. The principal role of oxidizing biocides in cooling systems is to control microbial growth. However, they may adversely affect and degrade the scale inhibitors present in the water treatment program. Proper combinations of oxidizing biocides and scale inhibitors have to be chosen to ensure both microbial and scale control. Stabilized bromine, for example, is an effective biocide and does not degrade phosphonates, eg., HEDP, to any appreciable extent under common cooling tower operating conditions.

CONCLUSIONS Cooling water accounts for most of the water used in industrial applications. Many cooling systems conserve water by operating under conditions of high mineral salt supersaturation, particularly CaCO3. Inhibitors, usually phosphonates or carboxylates, are required to prevent scale. However, simply measuring the ability of a molecule to inhibit calcite precipitation in a beaker is insufficient. Cooling water systems are complex and dynamic. They are affected by many factors not present in “beaker” tests. Individual systems are unique, and each is operated to achieve unique performance goals. Treatment programs must balance scale, corrosion, and microbial control for specific conditions. This paper described the effects of calcium and oxidizing biocides on scale inhibitors, as well as performance characteristics of those inhibitors. Effective cooling water treatment is a complex and challenging process. It is a multifaceted task that requires control of three major factors: scale, microbiological growth and corrosion. Monitoring and feed control of the cooling water treatment program is also very important for the success of the application. This eliminates over/under-feeding of treatment chemicals and ensures better control by the cooling tower operator.

232

The success of any water treatment program depends on its versatility, proper application and appropriate control of variables. Properly selecting and combining inhibitors and dispersants allows the water chemist to develop high-performing, costeffective formulations, which minimize both scale and corrosion at high cycles and with any biocide program. Inhibitors, either phosphonate or carboxylate, can be chosen for desired activity and sensitivity towards hardness and oxidizer.

ACKNOWLEDGMENTS The authors wish to thank R. V. Davis, M. A. Kamrath, B. S. Bedford, F.-Y. Lu, D. P. Pruss, J. C. Krcma, A. M. Shiely, N.M. Rao, and S. A. Borchardt for producing and sharing experimental results, D. A. Johnson for useful discussions, and P. S. White (University of North Carolina at Chapel Hill, Department of Chemistry) for collecting the single crystal X-ray data on Ca-AMP.

REFERENCES 1. Cowan, J.C.; Weintritt, D.J. Water-Formed Scale Deposits, Gulf Publishing Co. Houston, TX, 1976, p. 93. 2. Pilson, M.E.Q. An Introduction to the Chemistry of the Sea, Prentice Hall, Upper Saddle River, NJ, 1998, p. 123. 3. F.N. Kemmer, The Nalco Water Handbook, McGraw-Hill Company, New York (1988) 4. J. Katzel, Fundamentals of cooling towers, Plant Engineering April 27:32 (1989). 5. Power Special Report, Cooling towers, Power March:S-1 (1973). 6. E.C. Elliot, Cooling towers, Power December:S-1 (1985). 7. C.R. Branan, Rules of Thumb for Chemical Engineers, Gulf Publishing Co. Houston, TX, 1994, p. 127. 8. Sarig, S.; Ginio, O. A mechanism for retarded precipitation based on the time evolution of particle size and relative number density, J. Phys. Chem. 80:256 (1 976). 9. Koutsoukos, P.G.; Kontoyannis, C.G. Precipitation of calcium carbonate in aqueous solution, J. Chem. Soc. Faraday Trans.I 80: 1 18 1 (1984). 10. Spanos, N.; Koutsoukos, P.G. Kinetics of precipitation of calcium carbonate in alkaline pH at constant supersaturation. Spontaneous and seeded growth, J. Phys. Chem. B 102:6679 (1998). 11. Oddo, J.E.; Tomson, M.B., Why scale forms and how to predict it, SPE Production & Facilities February:47 (1994). 12. Johnson, C.K. ORTEP: A fortran thermal ellipsoid plot program; Technical Report ORNL-5138; Oak Ridge National Laboratory: Oak Ridge, T.N, 1976. 13. Gabe, E.J.; Le Page, Y.; Charland, J.-P.; Lee, F.L.; White, P.S., NRCVAX - an interactive program system for structure analysis, J. Appl. Crystallogr. 22:384 (1989). 14. International Tables for X-ray Crystallography; Kynoch Press: Bhigham, U.K., 1974; Vol. IV. 15. Reed, D.T.; Nass, R. Small-scale short-term methods of evaluating cooling tower treatments-are they worthwhile?, International Water Conference 1975, paper # 1. 16. Fulks, K.E.; Yeoman, A.M., Performance evaluation of non-metal cooling water treatments, Corrosion/83, Paper No. 279, National Association of Corrosion Engineers, Houston, TX, 1983. 17. Hale, E.R.; Hoots, J.E.; Nicolich, S.N., Tracers track down water problems, Power Engineering September:21 (1999). 18. Reis, A.H. Jr.; Peterson, S. W.; Dryan, M.E.; Gebert, E.; Mason, G. W.; Peppard, D.F., Sterically hindered extractants. 2. A neutron-difiaction study of the di-tert-butylphosphinic acid dimer showing strong asymmetric hydrogen bonding, Inorg. Chem. 15:2748 (1976). 19. Smith, P.H.; Raymond, K.N., Solid-state and solution chemistry of calcium N(phosphonomethyl)glycinate, Inorg. Chem. 27: 1056 (1988). 20. Langley, K.J.; Squattrito, P.J.; Adani, F.; Montoneri, E., Structures of fluorobenzylphosphonic acid isomers and their calcium salts, Inorg. Chim. Acta 253:77 (1996).

233

21. DeLaMatter, D.; McCullough, J.J.; Calvo, C., Crystal structure of methylenediphosphonic acid, J. Phys. Chem. 77: 1 146 (1973). 22. Gebert, E.; Reis, A.H. Jr.; Dryan, M.E.; Peterson, S.W.; Mason, G.W.; Peppard, D.F., Structural investigations of unsubstituted polymethylenediphosphonic acids 2. The molecular and crystal structure of propane-l,3-diphosphonic acid, J. Phys. Chem. 8 1 :47 1 (1977). 23. Uchtman, V.A.; Gloss, R.A., Structural investigations of calcium binding molecules. I. The crystal and molecular structures of HEDP.H2O, J. Phys. Chem. 76:1298 (1972). 24. Uchtman, V.A., Structural investigations of calcium binding molecules. II. The crystal and molecular structures of calcium-HEDP.2H2O; Implications for polynuclear complex formation, J. Phys. Chem. 76:1304 (1972). 25. Nardelli, M.; Pelizzi, G.; Staibano, G.; Zucchi, E., A structural study on metal binding of gemdiphosphonates, bonegrowth regulators, Inorg. Chim. Acta 80:259 (1983). 26. Mathew, M.; Fowler, B.O.; Breuer, E.; Golomb, G.; Alferiev, IS; Eidelman, N., Synthesis, characterization, and crystal structure of dicalcium glutarylbis(phosphonate) dihydrate: a covalently pillared layer structure with the potential for epitaxial growth on hydroxyapatite, Inorg. Chem. 37:5485 (1998). 27. Räsänen J.P.; Pohjala, E.; Nikander, H.; Pakkanen, T.A., Ab initio studies on organophosphorous compounds. 6. Interactions of dimethylphosphinic and dimethyl-phosphinothioic acid monoanions and methylenebisphosphonic acid dianion with calcium, J. Phys. Chem. A 101:5 196 (1997), and references therein. 28. Davis, R.V.; Carter, P. W.; Kamrath, M.A.; Johnson, D.A.; Reed, P.E. The use of modem methods in the development of calcium carbonate inhibitors for cooling water systems, in Mineral Scale Formation and Inhibition, Amjad, Z., ed., Plenum Press, New York (1995, p. 33. 29. Charpin, P.P.; Lance, M.; Nierlich, M.; Viper, D.; Lee, M.-R.; Silvestre, J.-P.; Dao, N.Q., Structure du Trihydrogéno hydroxy-1 -ethanedi(phosphonate)-1, 1 de rubidium dihydrate, Acta Crystallog. Sect. C Cryst. Struct. Comm. C44:990 (1988). 30. Vanderpool, D., New calcium carbonate scale inhibitors: understanding complexation constants as a tool for finding improved performance, International Water Conference, paper # 40, p. 383 (1997). 3 1. Carter, R.P.; Carrol, R.L.; Irani, R.R., Nitrilotris(methylenephosphonic acid), ethyliminodi(methylenephosphonic acid) and diethylaminomethylenephosphonic acid: acidity and calcium(II) and magnesium(II) complexing, Inorg. Chem. 6:939 (1967). 32. Deluchat, V.; Bollinger, J-C.; Serpaud, B.; Caullet, C., Divalent cations speciation with three phosphonate ligands in the pH-range of natural waters, Talanta 44:397 (1997). 33. Oddo, J.E.; Tomson, M.B., The solubility and stoichiometry of calciumdiethylenetriaminepenta(methylene phosphonate) at 70° C in brine solutions at 4.7 and 5.0 pH, Applied Geochem. 5:527 (1990). 34. Sawada, K.; Araki, T.; Suzuki, T., Complex formation of amino polyphosphonates. 1. Potentiometric and nuclear magnetic resonance studies of nitrilotris(methylenephosphonato) complexes of the alkaline-earthmetal ions, Inorg. Chem. 26:1199 (1987). 35. Kabachnik, M.I.; Medved, T.Y.; Dyatlova, N.M. Rudomino, M.V. Russ. Chem. Rev. 1974,43,733. 36. Matty, J.M.; Tomson, M.B., Effect of multiple precipitation inhibitors on calcium carbonate scale nucleation, Appl. Geochem. 3549 (1988). 37. Tomson, M.B., Effect of precipitation inhibitors on calcium carbonate scale formation, J. Cryst. Growth 62:106 (1983). 38. Ferguson, R.J.; Freedman, A.J.; Fowler, G.; Kulik, A.J.; Robson, J.; Weintritt, D.J., The practical application of ion association model saturation indices to commercial water treatment problem solving, in Mineral Scale Formation and Inhibition, Amjad, Z., ed., Plenum Press, New York (1995), p. 323. 39. Johnson, D.A.; Fulks, K.E.; Meier, D.A., Factors influencing the decomposition of HEDP by chlorine, Corrosion/86, Paper No. 403, National Association of Corrosion Engineers, Houston, TX, 1986. 40. Bartholomew, R.D., Bromine-based biocides for cooling water systems: a literature review, International Water Conference, paper # 74, p. 523 (1998). 41. Vaska, M.; Go, W., Microbial control. Evaluation of alternatives to gaseous chlorine for cooling water, Industrial Water Treatment March/April:39 (1993). 42. Brochures on “Dequest® Phosphonates” by Solutia. 43. Berg, D.; Vanderpool, D.; Rubin, D., Ion chromatographic analysis of organophosphonates in cooling water, International Water Conference, paper # 7, p. 56 (1987).

234

COMPLEXATION OF PROTEINS WITH POLYMERS IN AQUEOUS SOLUTIONS

Benlian Wang and Etsuo Kokufuta Institute of Applied Biochemistry University of Tsukuba Tsukuba, Ibaraki 305, Japan

INTRODUCTION Water-soluble polymers including polyelectrolytes have been known to form complexes with a variety of proteins in aqueous media. Of particular interest would be the way in which globular proteins bind to a flexible polymer chain, an understanding of which could provide a better explanation of the formation mechanism of polymer-protein complexes (PPCs). The present study has employed polymers of two sorts, i.e., potassium poly(vinyl alcohol) sulfate (KPVS) and poly(ethylene glycol) (PEG). The former is a typical polyanion, while the latter is a neutral water-soluble polymer frequently used in basic and technological studies. The experimental techniques used were turbidimetric titration, quasi-elastic light scattering (QELS), static light scattering (SLS), electrophoretic light scattering (ELS), dialysis method, and fluorescence spectroscopy. In addition, biochemical methods such as the measurement of enzymatic activity have also been employed in the appropriate cases. Studies on the PPC formation mechanism would be undertaken to characterize the structure of an “intrapolymer” complex in which several protein molecules bound to a chain of the polymer. Such a complex has been found to be formed in the PEG-protein system via hydrogen bonding between the polymer and protein. In the KPVS-protein system, however, the polyion charges should be neutralized by protein charges of the opposite sign; thus, the resulting intrapolymer complexes immediately associated with one another to yield neutral aggregates (AG1) of a characteristic size. Further addition of polyions caused an increase in the concentration of AG1 without change in its size. When the AG1 concentration exceeded a certain level, larger aggregates (AG2) form through association of AG1. We intend here to summarize our studies1-16 from which these conclusions have been drawn.

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers, New York, 2000

235

MATERIALS AND METHODS The following protein samples were obtained from Sigma Chemical Co. as 95-99 % pure lyophilized products: papain, human serum albumin (HSA), lysozyme (Lyz), ribonuclease (RNase), trypsin, and pepsin. PEG(1) was obtained from Nishio Industrial Co. and the others from Wako Pure Chemical Co. KPVS samples of various linear charge densities were prepared via the esterification of three PVAs which were commercial products (Kurare Co., Ltd.) with the degree of polymerization (DP) of 500, 1500, and 3500. Both polymers were characterized by studying the following quantities: molecular weight (MP); radius of gyration ( Rg); hydrodynamic radius ( Rh); the degree of esterification (De); and the change in refractive index with concentration, (dñ/dc)p, which is required for the analyses of SLS data for both polymer and complex. The results of characterization are shown in Table 1. The complexation was examined at 25°C by a combination of the following techniques: turbidimetric titration with a Hirama automatic recording titrator (model ART-3); QELS and SLS with a Brookhaven system (Holtsville, NY) equipped with a 256-channel digital auto-correlator (BI-2030 AT) and a 2 W Ar+ ion laser (Stabilite 2017, Spectra-Physics Lasers); ELS with a Coulter DELSA 400 apparatus Hialeah, Florida); fluorescence spectroscopy with a Hitachi model F-4500 spectrophotometer; refractive index measurements with an Otsuka electrophotometric differential refractometer (model DRM1021). We also employed the colloid titration17-19 for determining the esterification degree of KPVS. The standard titrant used was an aqueous solution containing 4 mM (based on the cationic groups) of 6-O-(2-hydroxyethyl-2- (trimethylamonio)-chitosan) iodide, which was commercially obtained from Wako Pure Chemical Co. Table 1. Properties of PEG and KPVS Polymer PEG( 1) PEG( 1') PEG(2) PEG(3) PEG(4) KPVS (5/7 8) KPVS( 15/13) KPVS( 15/33) KPVS( 15/53) KPVS( 15/76) KPVS( 15/92) KPVS(35/76)

Mp (104 g/mol) 13 20 29 62 428

Rn (nm) 55

6.1 8.7 12.1 14.9 18.1 18.7 46.6

12 15 18 20 21 22 36

Rg (nm)

De (–)

dñ/dc .1372

10 21

,1371 21 22 27 31 33 35 60

.778 .I30 .328 .530 ,760 ,922 .765

.1073 .1314 .1175 .1095 .0988 .0937 .0977

RESULTS AND DISCUSSION Formation of PPCs in the PEG System Early Studies with Pepsin. Kokufuta et al10 have attempted to study the complexation of PEG with porcine pepsin (pepsin A) as a typical acid gastric protease including a number of carboxyl groups, as a function of pH and PEG concentration. Neither

236

precipitation nor turbidity appeared in this system after the mixing of the protein and polymer at different ratios. The addition of PEG increased the reduced viscosity of the enzyme solution at pH 3 but not at pH 4.5. An increase in pH was observed by mixing both PEG and enzyme solutions which were previously adjusted to pH 3. Under conditions of pH 2.5 3 and 50°C, PEG contributed to an increase in the hydrolyzing activity of pepsin toward N -acetyl-L-phenylalanyl-3,5-diiodo-L-tyrosine. These results suggested that pepsin forms a water-soluble complex with PEG. Since the PPC formation occurred at pH 3 at which most of the COOH and phenolic OH groups are protonated, but not at pH 4.5 allowing partial dissociation of COOH groups, hydrogen bonds between the COOH and phenolic OH groups in the enzyme and the ether groups in PEG seemed to play an important role in the complexation. This was supported by examining the complexation by a combination of QELS and ELS.8

Complex Formation between HSA and PEG. In order to establish the generality of the above observations, we further studied another protein-polymer system using HSA and PEG.1 PPCs were then formed by mixing PEG with HSA in aqueous NaC1-containing systems at different ionic strengths (I) of 0.01 and 0.1 under a fixed pH over a range from 2 to 8. The mixing was controlled by the molar ratio (rm) of the ether groups in PEG to all of the HSA acidic groups. Figure 1 shows typical examples of QELS results. Two varieties of particles with different apparent Stokes diameters (ds) were observed in the HSA-PEG mixtures at rm = 1 and 10. The size of the small particles was very close to that of free HSA and independent of the following factors: rm, MP and I. Therefore, we may assign this particle to uncomplexed free HSA.

PEO(1); I = 0.01

PEO(4); I = 0.01

PEO(4); I = 0.1

Diameter (nm) Figure 1. Results of QELS for free HSA, free PEG, and HSA-PEG mixtures at pH 2 and atI = 0.01 and 0.1. Cpro = 0.1 mg/mL for both free HSA and HSA-PEG mixture. rm denotes the mixing ratio based on the functional groups of the protein and polymer.

237

Since the size of large particles varied depending on M p and I, the effects of both were studied in detail (Table 2). An increase in M p increased the size of large particles, whereas the particle size at I = 0.1 was always smaller than that at I = 0.01. A polymer chain with bound HSA would carry positive charges at pH 2 due to the protonated guanidyl, amino and imidazolyl groups. Repulsions among bound proteins are reduced by an increase in I due to the electrostatic screening effect. This leads to a contraction in the polymer chain. Therefore, these results indicate that the large particle is the HSA-PEG complex. Our samples were assumed to be multicomponent systems of PEG, HSA, and the complex of these two. An excess of uncomplexed PEG would remain in the systems at a high rm (e.g., rm = 10), although it did not appear in the QELS pattern. However, our previous studies2 have demonstrated that the contribution of free polymers (and also of free proteins) to the scattering signal assigned to the complex was essentially negligible. The same conclusion can be obtained here, because there was little change in the size with rm (see Table 2). Table 2. Apparent Stokes Diameters (ds) for HSA-PEG complexes formed at different conditions.

Polymer PEG(1) PEG(1) PEG(2) PEG(2) PEG(3) PEG(3) PEG(4) PEG(4) PEG(1) PEG(1) PEG(2) PEG(2) PEG(3) PEG(3) PEG(4) PEG(4) PEG(4) PEG(4) PEG(4) PEG(2)

Conditions of PPC formation rm 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 10 1 1 1 1

pH 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 5 8 2

ds (nm) I .01 .01 .01 .01 .01 .01 .01 .01 .1 .1 .1 .1 .1 .1 .1 .1 .01 .01 .01 .0 1

Urea (M) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1

69 68 82 83 89 91 127 128 55 52 70 69 71 73 108 108 127 314 90 7

Effect of pH on the Complexation between HSA and PEG. We have suggested from the pepsin-PEG system that hydrogen bonding of the ether groups in PEG with the carboxyl and phenolic OH groups in the protein plays a major role in the formation of a water-soluble complex; thus the complex was formed at pH 3 but not at pH 4.5. The complexation between HSA and PEG was therefore studied over a wide pH range including the isoelectric point (pI; 4.9±0.2) of HSA. A typical result is shown in Table 2 The QELS pattern at pH 3 was the same as that at pH 2 in Figure 1. However, at pH 5 (close to pl ), aggregation of free HSA was observed, and hence a HSA-PEG mixture (rm, = 1) included only very large particles (d s 300 nm). Although there was no aggregation at pH 8 at which a large part of carboxyl groups would be deprotonated, the particles with a 90 nm size still remained in the system. The PPC formation at pH 8 was unexpected from the results for the pepsin-PEG system. However, taking into account the

238

fact that pepsin has a lot of acidic groups (43 carboxyl and 16 phenolic OH) but a few of basic groups (2 amino, 1 imidazolyl, and 2 guanidyl), we become aware of the difference between pepsin and HSA with respect to the ratio of the acidic to the basic groups. In contrast to pepsin, HSA has 100 total basic groups, the number of which is about the same as that (116) of acidic groups. Thus, it is likely that the basic groups, most of which are deprotonated at pH 8, could form hydrogen bonds with PEG; for example,

Effect of Urea on the Complexation between HSA and PEG. The above results are indicative, yet not conclusive, of the formation of HSA-PEG complexes via hydrogen bonding. Thus, we attempted to study the effect of urea on the complexation using QELS and fluorescence spectroscopy. Urea has been frequently employed in the biochemical field as a tool to identify hydrogen bonds since it is generally believed that urea can break up intra- or intermolecular hydrogen bonding of proteins in aqueous systems. The urea concentration was adjusted to 1 M (mol/L) which is one eighth lower than the most commonly employed concentration (8 M) in the denaturation of proteins. Therefore, we did not observe a change in the size of free HSA either in the presence or the absence of urea. For the HSA-PEG mixture, however, 1 M urea inhibited the complex formation, even under the conditions at which the complex was formed (pH 2 and I = 0.01) (see Table 2). This could suggest that hydrogen bonding plays a role in the complexation between HSA and PEG, at least at pH 2. It has long been believed that urea disrupts the cluster structure of water molecules,20,21 i.e., “structure breaking effect.” Consequently, one may argue that urea weakens the hydrophobic hydration of a solute, leading to a reduction in the hydrophobic interaction (but not in hydrogen bonding) between PEG and HSA. We thus employed the fluorescence spectroscopy as a second tool for studying the complexation. The UV spectrum of an aqueous HSA solution showed an absorption band around 280 nm due to tryptophan and tyrosine residues; thus, the wavelength for the excitation in our fluorescence spectroscopic studies was fixed at 280 nm. An emission maximum (λem max) was then observed around 340 nm in the spectrum of free HSA at pH 8, the value of which was in fair agreement with that in the literature.22 The fluorescence around this λem max is mainly due to the tryptophan residues in the HSA molecule and on its surface; therefore, λem max is shifted upon changes in “micro-environment’’ around the tryptophan residues. The best known example of this is a blue shift of λem max upon a decrease in pH, the phenomena of which can be observed at pH
239

pH < 5 (i.e., pH-induced blue shift). (ii) A blue shift of λem max is also observed when increasing the concentration of free HSA at pH 5 (aggregation effect). (iii) Urea (1 M) has little affect on the pH curve of λem max for free HSA. (iv) In the presence of urea, there is little difference in the pH vs λem max curves between the free HSA and the HSA-PEG mixture. (v) In the absence of urea, however, PEG causes a marked blue shift over all the pH ranges studied, particularly at pH < 5 at which the λem max values are close to that (3 15 nm) observed under conditions where the free HSA molecules associate with one another to form aggregates with a 314 nm size (see Table 2). From the results (iv) and (v), it is clear that urea inhibits the complex formation. In addition, the result (iii) means that 1 M urea has no influence on the micro-environment around the tryptophan residues. Consequently, we may say that hydrogen bonding, but not hydrophobic interaction, plays an important part in the complexation between HSA and PEG.

Figure 2. pH Dependence of emission maximum λem max) for free HSA and HSA-PEG(3) mixture (Cpro = 0.01 mg/mL; rm = 1) in the presence and absence of 1 M urea: open circles, free HSA (Cpro= 0.01 mg/mL) in the absence of urea; closed circles, free HSA (Cpro= 0.01 mg/mL) in the presence of urea; open triangles, HSA-PEG mixture in the absence of urea; closed triangles, HSA-PEG mixture in the presence of urea. The change in λem max with free HSA concentration in the absence of urea at pH 5 (close to pI) was also shown: closed square, Cpro= 0.04 mg/mL; open square, Cpro = 0.1 mg/mL. Arrow shows an increase in Cpro from 0.01 to 0.1 mg/mL.

Structure of HSA-PEG Complex. We performed SLS experiments to study the — structure of the HSA-PEG complex. The weight-average molecular weight (Mx) and root mean square of gyration (Rg for the complex may be estimated by the following equation: (1) Here, Cx denotes the weight concentration of the complex, Rθ is the Rayleigh ratio, θ is the scattering angle, λ is the wave-length of the light in the medium, A2 is the second virial coefficient, and K is given as:

240

(2) where λo is the wavelength of the light in a vacuum, NA is Avogadro's number, ño is the refractive index of the medium, and (dñ/dc)x (in mL/g) is the _ change in refractive index with complex concentration. Therefore, we can obtain both M x and Rg from eq (1) if Cx and (dñ/dc)x are known. However, it is no easy matter to isolate an "intact"complex from a HSA-PEG mixture. We thus attempted to replace Cx with the concentration (Cp) of the polymer by taking into account the following relations between Cx and Cp as well as between Mx and M p: (3) (4) Here, β represents the mass ratio of bound protein to the polymer, and α is the number of polymer chains within one complex particle, i.e., the number of aggregated intrapolymer complexes. Moreover, we considered that (dñ/dc)x is given by: (5)

where (dñ/dc)P and (dñ/dc)pro are a change in refractive index with concentration for polymer and protein, respectively. Substitution of eqs (3) to (5) into eq (1) yields (6)

where

(7) (8)

As a result, we can analyze SLS data without using the concentration and refractive index increment of the complex because all the terms in eq (6), except β, are measurable quantities. In the actual analyses of SLS data, we ignore the particle interaction term, i.e., the second virial coefficient. Figure 3 shows a typical example of K' Cp/Rθ vs sin2(θ/2) plots for the HSA-PEG(4) system at pH 2. The plots at I = 0.1 provided a good straight line over all the ranges of (40 to 140 °), whereas at I = 0.01 deviations were observed at θ < 60 °, perhaps due to polyelectrolyte effects at low ionic strengths. Such deviations were also observed for the HSA-PEG(1) system; thus, we adopted only the data at I = 0.1 for the estimations of MX and R g. Assuming the formation of an intrapolymer HSA-PEG complex (i.e., α = 1), the intercept and slope of the straight line yield β and Rg, respectively. In addition, we can obtain by Mx the introduction of β into eq (4). These results are listed in Table 3. We made two assumptions to analyze SLS data by the above approaches; i.e., A 2 ~ 0 and α = 1. Therefore, the validity of such assumptions should be judged by another experimental method. For this purpose, we carried out the dialysis experiments through

241

Table 3. Results of SLS and Dialysis Experiments for Complexes of HSA with PEO(1) and PEO(4). β

Polymer PEG(1) PEG(4)

SLS 33.4 3.03

dialysis 33.2 32.0

(105 g/moI)

Rg (nm)

Rh (nm)

p=Rg/Rh (–)

4.92 18.6

77 148

28 54

2.8 2.7

which β can be determined. As can be seen from Table 3, there is a good agreement between the values of β from the SLS and the dialysis experiments. This evidently indicates that the assumption of α = 1 is correct ; in other words, M x corresponds to the molecular weight of an “intrapolymer” complex.

Figure 3. Results of SLS for HSA-PEG(4) complex at I = 0.01 and 0.1. The complex was obtained by dialyzing a HSA-PEG mixture (Cpro = 0.1 mg/mL; rm= 1; pH 2; and I = 0.01 or 0.1) against a large volume of the same solvent (HCl-NaC1 mixture; pH 2; I = 0.01 or 0.1) used for the preparation of the proteinpolymer mixture. The dialysis was completed when QELS showed that there was no detectable peak of the free protein.

Figure 4. Schematic illustration of an intrapolymer HSA-PEG complex in which several proteins bound to a polymer chain.

242

From Rg and the corresponding hydrodynamic radius ( Rh, half of ¯ds in Table 2), we obtain ρ = Rg/Rh for the intrapolymer HSA-PEG complex. Hydrodynamic theory23 shows that ρ changes from infinity to 0.775 when the polymer structure changes from a long rod to a sphere, with values from 1.3 to 1.5 for random coils. As can be seen from Table 3, ρ ranges from 2.8 to 2.7. This result suggests an extended structure of the HSA-PEG complex (see Figure 4). Such an extended structure appeared to be due to “steric” rather than “electrostatic” repulsion among the positively charged and polymer-bound proteins, since the ionic strength used here was considerably high. Moreover, a marked blue shift of λem max due to the complexation, suggestive of an inter-protein interaction within the complex, has been observed in Figure 2. In addition to the above, we should discuss the big difference in β between PEG( 1) and PEG(4) observed in Table 3. This result means that an MP-dependent increase in the number (n) of the HSA molecules bound per chain of the polymer is not constant over a wide MP range, but the increment of n with MP decreases with increasing MP (see the following section). Such MP effect is then explicable in terms of common knowledge that an intramolecular interaction among the segments assigned to a polymer chain in a good solvent becomes stronger as increasing its molecular weight. Thus, an increase in M P depresses the protein binding which should accompany an extension of the PEG chain due to the inter-protein interaction. Taking this into account, it is predictable that the binding of HSA to PEG with an enormous MP takes place only on the surface of the polymer coil. Under such a situation, we would not observe an extended structure of the complex as shown in Figure 4. In order to confirm this, it is necessary to employ a high molecular− weight PEG whose MP is one or more order of magnitude greater than that of PEG(4), but many difficulties would arise in handling such a high molecular-weight sample. Electrophoretic Behavior of HSA-PEG Complex. We employed the ELS technique for the examination of the complex charge. It was found that the electrophoretic mobility (U) decreases more or less with increasing M x in particular at I = 0.01, due to the effect of M x on n as stated above. However, a large difference in U between the complexed and free proteins was not observed, even when PEG(4) was used in the complexation. There was no influence of rm on U at I = 0.01 and 0.1. Seemingly, these results imply that during electrophoresis the complex consisting of PEG-bound proteins with positive charges behaves like a “free draining coil,” which is a generally accepted model in many theoretical and experimental studies on the electrophoretic behavior of polyelectrolyte ions.24-27 Nevertheless, one may argue that since an increase in Mx increases both the size of the complex and the number of bound proteins at the same time, it is possible to observe a constant U without the influence of Mx (i.e., a “non-free draining” behavior). Thus, we attempted to analyze the mobility data in more detail. Both SLS and dialysis experiments have demonstrated that the complexation between HSA and PEG yielded the intrapolymer complex. When such a particle behaves as a nonfree draining coil during electrophoresis, its mobility may be given by Henry's equation:28 U=

f(KR)(1+Kr) Ze 6πηR(1+ Kr + KR)

(9)

where f(KR) is Henry’s function, K is the Debye-Hückel parameter, R is the radius of a complex particle (R~Rh), r is the radius of a small ion, e is the electric charge, and ηis the viscosity of the medium. Moreover, Z is the charge of a complex particle and may be given by Z = nZ pro, where Z pro is the charge of a bound HSA molecule. Then, we may obtain n from in Table 3 using a relation of n= ß ( M p /M pro )

(10) 243

The charge of free HSA (Zpro) was then estimated to be +45 from the mobility (1.67 µm·cm/Vs) at I = 0.1 and pH 2 using eq (9). As a result, we can estimate the mobility for a non-free draining complex particle. The results are listed in Table 4, together with all the parameters used for the calculation. The calculated values were about one-fifth the observed ones. Even though some errors exist in our estimation, these results strongly suggest that the intrapolymer complex consisting of HSA and PEG did not behave as a non-free draining particle during electrophoresis. Table 4. Comparison of Calculated and Observed Electrophoretic Mobilities for Complexes of HSA with PEO( 1) and PEO(2) at pH 2 and = 0.1 Polymer PEG( 1) PEG(4)

R (nm) 28 54

κR

f(κR)

N

Z

28.6 56.2

.91 .95

7 21

3 15 45

Ucalc (µ·cm/Vs) .37 .31

Uobs (µ·cm/Vs) 1.61 1.62

It is significant to compare the values for PEG(1) and PEG(4) in Table 4. For this purpose, we attempted to estimate n for PEG(4) from β for PEG(1). Since MP= 428,000 for PEG(4) and M pro= 66,436, we obtain n ~ 215. This is almost 4 times that determined using the β values from the SLS and the dialysis experiments for PEG(4). The reason for this deviation has been discussed in the previous section.

Formation of PPCs in the KPVS System Early Studies with Several Proteins and Polyelectrolytes. In 1980s, our studies12-16 paid much attention to the PPC formation in the polyelectrolyte systems. The stoichiometry of the complexation was a topic of our research with regard to obtaining information about the way in which the charges on protein form “ion pairs” with the opposite charges of polyion. For this purpose, we employed a salt-free system at a pH level which forces to completion either the protonation of basic groups or the deprotonation of acidic groups bound to a protein. Under such experimental conditions, the formal charge of a protein should be equivalent to the number of acidic or basic groups per protein; therefore, we may calculate from the amino acid composition the number of ionic groups taking part in the complexation. Then it was possible to demonstrate that a number of proteins form PPCs with polyions through a “1:1 stoichiometric binding” between oppositely charged groups.7,12-16 The resulting PPCs were insoluble in water, and thereby the system underwent a phase separation during the course of the complexation. As shown in detail in the next section, however, there was an appreciable retention of biochemical function in the resultant PPCs. Protease Activity of Stoichiometric Complex between Trypsin and KPVS. Enzymes would be appropriate for studying the effects of complexation with polyelectrolytes on the ternary conformation of proteins, because one may expect to monitor slight variations in protein ternary structure due to the complexation through changes in enzyme activity. Figure 5 shows the pH-activity profiles of the native and complexed enzymes using N α-benzoyl-arginine-p-nitroanilide (BANA) as a low molecular weight substrate and casein as a high molecular weight substrate. The stoichiometric trypsin-KPVS( 15/92) complex is found to have an appreciable retention of enzymatic activity. This finding indicates that one imidazolyl group (histidine), which cooperates

244

with both COOH (aspartic acid) and OH (serine) in acylation-deacylation as an intermediate step in the enzyme catalytic action,29 is free of salt linkages with KPVS.

pH

pH

Figure 5. PH-activity curves of native (open circle) and KPVS(15/92)-complexed trypsin (closed circle) determined at 30 °C using a 2.66 mM BANA and 4 % casein solutions.

However, the complexation with KPVS results in two characteristic differences in the pH-activity curves of the native and complexed enzymes. First, we found from Figure 5 that in the acidic region the curve of the complexed trypsin shifts towards a higher pH range than the curve of the native enzyme. Goldstein et al.30 have observed a similar shifting in the pH-activity curves for two modified trypsin derivatives: acetyltrypsin obtained from acylation of the ε-amino groups of trypsin; and copolymer-bound, waterinsoluble trypsin in which the enzyme was covalently bound via itse amino groups to the COOH groups of a 1:1 copolymer of maleic acid and ethylene. The ε-substrate used in their study is Nα-benzoyl-L-arginine ethyl ester (BAEE). The magnitude of the shifting of the activity curve for copolymer-bound trypsin was reduced with increasing ionic strength, while the ionic strength did not alter the shifted curve for acetyltrypsin, in which the negative charges prevail over the positive charges because of the acylation of the amino groups. Goldstein et al. thus concluded that the electrostatic potential due to the negatively charged maleic acid copolymer affects the local concentrations of both H+ ions and positively charged BAEE molecules in the microenvironment of the copolymer-bound enzyme molecules. In the case of the complexed trypsin, as observed in acetyltrypsin, the ionic strength had little influence on the activity curve. It therefore seems most likely that the shifting of the activity curve for the trypsin-KPVS complex is due not to the negative charges of the KPVS component, but to the preponderance of negative charges arising from a 1:1 stoichiometric neutralization of the positive charges of the enzyme molecules through complexation with KPVS. Another important feature of Figure 5 is that in a narrow pH range from 8 to 9, a rapid increase in complexed enzyme activity was observed, whereas the native enzyme activity gradually decreased. In order to understand such abnormal activity on the part of complexed trypsin, the potentiometric titration curve with NaOH for the acidic groups (COOH and phenolic OH) remaining in the complex was examined. The results clearly indicated that the salt linkages between the imidazolyl and sulfate groups in the complex

245

are severed during the deprotonation of the charged imidazolyl groups caused by an increase in pH. Such a cleavage could cause a restoration of a complexation-induced and altered conformation of trypsin molecules, thereby resulting in a possible recovery of their enzymatic activity. A comparison of pH-activity profiles examined with low and high molecular weight substrates should help to clarify the ternary structure of trypsin in the complex, Haynes and Walsh31 have reported a large influence of substrate molecular size on the activity of a water-insoluble trypsin as an immobilized enzyme. The immobilized trypsin was then prepared by adsorbing the enzyme molecules as a monolayer onto colloidal silica, followed by crosslinking with glutaraldehyde. This preparation had about 80% retention of BAEE-hydrolyzing activity compared with native trypsin, whereas the activity towards casein as a high molecular weight substrate was approximately 17% that of the native enzyme. The observed low activity towards casein was interpreted in connection with a lower availability of active sites and a lower efficiency at each site, both of which could arise from a limited diffusion rate of the high molecular weight substrate due to a steric hindrance effect from the supporting matrix. A similar conclusion has also been drawn from enzymological studies of the copolymer-bound trypsin system with casein and hemoglobin.32 In contrast to these previous results, a large difference in the activities caused by the molecular weight of the substrate is not observed in the trypsin-KPVS system. Also, in the pH 8 to 9 range, where some of the salt linkages had been severed, the fall in the activity of the complexed trypsin towards casein is much smaller than that of the native enzyme. This appears to be evidence supporting the recovery of enzyme activity via the restoration of the complexation-induced conformation change of trypsin, although it showed up more clearly when using BANA. Both results reveal that the usual diffusion limitation of polymer substrates does not have a large influence on the complexed trypsin-catalyzed degradation of casein. Consequently, the salt linkages maintaining the structure of the complex are found to be very “loose” even if the enzyme was stoichiometrically complexed with KPVS. A similar result has been obtained from the study on protease activities of a stoichiometric complex of papain with KPVS(15/92).6,9 In addition, the active site of hemoglobin was little varied by the complexation with KPVS as well as poly(diallyldimethylammonium chloride) under conditions where stoichiometric PPCs were formed.16

Turbidimetric titration curves. In the complexation in salt-free polyelectrolyte systems, there appears a turbidity due to the resulting PPCs. Thus, we have theoretically dealt with the turbidimetric titration curve for the PPC formation process;2 that is, a change in the absorbance (A) due to the appearance of turbidity while a protein solution is titrated with a polyelectrolyte solution. Since the PPC formation takes place via a 1:1 stoichiometric binding between the opposite charges, the plots of A against titrant volume (Vt in mL) may be expressed as: A=H

Vt

(11)

where Vi denotes the initial volume of protein solution and the proportionality constant ( H ) is then written as: H= (dñ/dc)2X (12)

246

Thus, one may predict that the titration of proteins with polyions would cause a linear increase in A with Vt, when n remains unchanged during the titration. Indeed, the titration curves for many of the proteins were given by linear plots of A vs Vt, the slope of which was independent of Cpro Figure 6 shows typical titration curves of papain with KPVS. The absorbance increases linearly with an increase in Vt, and shows an abrupt increase when Vt reaches a certain volume (V t'), which can be viewed as the V t at the end point of the titration because (i) Vt’ is a linear function of Cpro and (ii) the relation between V t’ and Cpro is given by: (13) where Nc (in eq/g) denotes the number of cationic charges per unit mass of protein, and M° is the molecular weight per ionizable group (i,e., equivalent weight in g/eq) for KPVS. As a result, we may suggest that an electrically neutral PPC with a constant size is formed during the course of the titration.

Figure 6. Turbidimetric titration curves of HSA solutions (30 mL with different concentrations of KPVS. Cp = 0.415 mg/mL which is equivalent to 2.5X10-3 mol/L in molar concentration based on the ionizable groups.

Studies on Titration Process by QELS and ELS. First, the titration process was examined by QELS in order to investigate the size of the PPCs formed. The binding of proteins to the polyion was quite strong, as there was essentially no free polyion in our systems for V t' Vt. The contribution of free proteins to the scattering signal was essentially negligible; thus, only PPC particles were detectable by QELS. In addition, we were able to obtain reproducible data, even when the measurement was performed within at least one day after the sample preparation throughout the titration. Figure 7 shows the results for HSA as a typical example. There is little difference in the apparent distributions of ds for PPCs as well as in the average of ds (i.e., ds) until V t/V t' = 0.75. At V t/V t' = 1, however, very large particles of PPC (ds ~ 3000 nm) are observed. These results are in conformity with the previous interpretation of the turbidimetric titration curve: a linear

247

increase in A with V t at Vt < Vt’ is due to an increase in the PPC concentration at constant PPC size, followed by a very rapid increase in A at Vt = Vt’, associated with an increase in PPC size due to aggregation. We also employed the ELS technique for the examination of the titration processes. Figure 7 shows ELS spectral changes during the course of the titration of HSA with KPVS as a typical example. Free HSA molecules at pH 2 exhibited a peak at U ~ 2.30 µm·cm/Vs, while the addition of KPVS gave rise to a second peak at U = 0 µm·cm/Vs, with the peak due to free HSA disappearing completely at V t/V t’ = 1. This result clearly demonstrates that the HSA-KPVS complex is electrically neutral.

Figure 7. Changes in apparent distributions of Stokes diameter and electrophoretic mobility due to PPCs formed during the course of turbidimetric titration for a sample solution (30 mL; pH 2) containing 3 mg HSA with KPVS titrant (= 0.415 mg/mL). The amount of KPVS added was given by; = 1.81 mL.

Quantitative analyses of the titration curves. We may theoretically calculate the slope (Sc in mL-1) of titration curve and compare it with the slope (So) obtained experimentally, since all of the parameters in eqs (11) and (12) are known. As can be seen from Table 5, the values of Sc are two orders of magnitude smaller than those of So. The reason for this disagreement is, however, understandable considering that the calculation is made for the corresponding “intrapolymer” complexes. In other words, So would be assigned to an “multipolymer” aggregate (i.e., mentioned in Introduction) of intrapolymer complexes.

248

Table 5. Results of Turbidimetric Titration, QELS and SLS

KPVS(15/92) KPVS(1592) KPVS(1592) KPVS(1592) KPVS(1592)

Protein Mx (106g/mol) papain 126 HSA 946 Lyz 105 RNase 101 trypsin 151

KPVS(1592) KPVS(15/13) KPVS(15/33) KPVS(15/33) KPVS(15/53) KPVS(15/53) KPVS(15/76) KPVS(15/76) KPVS(15/78) KPVS(5/78) KPSV(3576) KPVS(35/76)

Lys 0236 669 RNase 0229 340 Lyz 0.482 7.14 RNase 0.466 320 Lyz 0.706 729 RNase 0.682 270 Lyz 0.954 733 RNase 0.921 190 0.324 7.46 Lyz RNAse 0312 0611 Lyz 246 721 RNase 238 614

Polymer

Mx (107g/mol) 736 428 667 664 784

α (-) 54 45 64 7 52 284 148 148 69 103 40 77 21 230 20 29 26

Rg (nm) 69 a 79 37 80

Rh (nm) 50 42 50 33 60

ρ (-) 138 126 158 109 133

102S ο 5.15 262 3.72 039 705

872 4.71 618 557 134

4.71 212 396 039 696

66 84 71 76 75 71 77 65 75 57 69 92

50 50 51 47 52 44 52 41 53 37 50 50

l32 168 139 168 144 161 148 159 142 154 I38 184

44 21 42 13

134 134 134 134

696 696 696 693

-

-

-

39 os 43 05 4.1 25

134 134 134 134 134 134

696 696 696 696 696 696

104S o 102(αSo)

Figure 8. Zimm plots for PPC consisting of HSA and KPVS at pH 2.0. Cx= 6.30X10-6 to 6.11X10-5 g/mL which were estimated using Equation (15); θ = 45 to 135°. The estimated second virial coefficient, 5.56X10-5 cm3mol/g2, is negligibly small.

249

For an intrapolymer complex, its molecular weight =

+nMpro

From this and eqs (4) and (10), we obtain α = system, C, in eq (1) can be given by

Cx=

1+(

may simply be written as: (14)

/

For PPCs in the polylectrolyte

(15)

Thus, we may obtain from SLS using eqs (1) and (5). The angular dependence of scattered light was determined as a function of Cx which was calculated from Vt using eq (15). Figure 8 shows a typical Zimm plot for HSA obtained at θ = 45 to 135° and at Cx = 6.30X10-6 to 6.1 lX10-5 g/mL. For all protein samples, the double extrapolations ofKCx / Rθ and also gave two straight lines with correlation vs. (sin2(θ/2) + kCx) plots to coefficients > 0.99, co-intersecting the Y axis, enabling us to determine Mx and Rg These results are listed in Table 5, together with α and It is found that the Mx values obtained are much larger than leading to ranging from 20 to 284 for all the complexes, except for RNase-KPVS(15/92) complex, for which α is much smaller than that observed for the other PPCs. To compare SLS results with turbidimetric titration curves, we introduce the term α into the right-hand side of eq (11). Thus, from α and n, we may calculate the expected turbidimetric slope αSc. This slope is compared to So, indicating that a multipolymer aggregate of intrapolymer PPCs forms during the course of the titration at V t < V t' From the results of QELS and SLS, we obtain ρ = Rg/Rh for aggregates of intrapolymer PPCs. As can be seen from Table 5, ρ ranges from 1.1 to 1.6 for many of PPCs, other than RNase complexes. This result may lead to a question as to why electrically neutral aggregates of intrapolymer PPCs do not form compact spherical structures. The binding of many protein molecules to an extended polymer chain may prevent the complex from adopting a fully collapsed configuration because of a steric repulsion among bound proteins. Lyz and RNase have the same number (19) of basic groups, while their distribution is almost homogeneous on Lyz but not on RNase. In addition, there is little difference in the molar mass between both proteins. Thus, effects of surface charge distribution of proteins in their complexation with polyions were studied using KPVS having a difference in the linear charge density and in the chain length (see Table 5). The slope of the linear plots of A vs Vt for Lyz was little dependent on MP and De. In the case of RNase, however, the slope increased with decreasing De and with increasing MP. From the results of QELS and SLS, it was found that the MP and De effects observed in the titration curve correspond to the changes of both size and mass of aggregated intrapolymer complexes formed during titration. Also found from Table 5 is that α decreases with increasing De in both protein systems, while increasing MP decreases α in the Lyz system but does not change it in the RNase system. These results suggest that both surface charge distribution of proteins and linear charge density of polyions strongly affect the formation of intrapolymer PPCs.

Trials to Detect Intrapolymer PPC by SLS and QELS. In order to clarify the process of PPC formation, it is necessary to establish whether an intrapolymer complex is ever produced during the course of titration. The formation of an intrapolymer PPC should be expected if we may add one polyion into an aqueous protein solution; however, such a way is impossible in experiments. In addition, many difficulties would arise in the

250

isolation of intrapolymer PPCs from their mixtures with free proteins. For these reasons, we cannot make use of intrapolymer PPCs to study their molecular properties. Nevertheless, we succeeded in observing the aggregation process of an intrapolymer PPC while adding very slight amounts of KPVS into aqueous protein solutions. Experimentally the aggregation process of intrapolymer PPCs was monitored at the beginnings of the titration. Both QELS and SLS were available for this purpose, although it was difficult to perform an accurate measurement of turbidity (A < 0.004). Figure 9 shows changes in α during the titration of HSA with KPVS(15/92) as well as of Lyz and RNase with KPVS(5/78). The values of α linearly increase against the normalized titrant volume (Vt/Vt’). Such a linear increase was also observed by plotting Rh and Rg against Vt/Vt’. These allow us to extrapolate each line to Vt/Vt’→ 0 and to study the molecular characteristics for a PPC arising from an infinitesimal quantity of KPVS. The values obtained by extrapolation were then distinguished by putting a dash for those in Table 5 and summarized in Table 6. It is found that in the HSA and RNase system α ~1, i.e. suggesting the formation of an intrapolymer PPC. Another important finding is that the value ofR g’ as well as R h’ is very close to that for the uncomplexed KPVS. Thus, it appears that the binding of RNase or HSA molecules to one KPVS ion little accompanies a conformational change of the polyion, since any contraction due to local collapse of polymer segments on the protein surface is compensated for by interprotein steric repulsion. In contrast to RNase and HSA, Mx' for Lyz is 15 times (3.24×105; for 7 the intrapolymer complex) and one-fifteenth (7.46×10 ; for AG1). Moreover, ρ for Lyz is not different from ρ forAG1. These clearly show that the intrapolymer PPC arising from Lyz immediately aggregates one another due to its high polarizability (see next section). From the results obtained here, it would be suggested that the complementarity of the spacing between charges on the protein to the uniform spacing between polymer charges is a primary factor for determining the stability of an intrapolymer PPC. HSA has one hundred positive charges, while only 19 positive charges bind to RNase. Nevertheless, the spacing between charges on RNase is not so different from that on HSA, because the basic groups on RNase are concentrated within narrow area of its surface (see the Protein Data Bank, Oct. 1996, Release No. 78, Brookhaven National Lab., Upton, NY, USA.). Thus, not only HSA but also RNase results in a stable intrapolymer complex with an infinitesimal quantity of KPVS, at least De > 0.7. In contrast to these proteins, Lyz has the basic groups which are scattered over its surface. For this reason, even if an intrapolymer complex results in the Lyz system, it is highly polarizable and immediately aggregates. Table 6. Molecular Characteristics for HSA, Lyz and RNase Complexes Obtained with Infinitesimal Quantity of KPVS. Polymer KPVS(15/92) KPVS(5/78) KPVS(5/78)

Protein HSA Lyz RNase

(-) 1 15 1

M x’ (105g · mol”) 9.5 49.1 3.15

R g’ (nm) 32 39 22

R h’ (nm) 13 27 11

ρ’ (-) 2.46 1.44 2.00

Mechanism of PPC formation. The processes of PPC formation inferred from our studies are schematically shown in Figure 10. Upon addition of KPVS to a protein solution, the protein molecules bind to a flexible polyion via Coulomb attraction, resulting in an intrapolymer complex, in which the polyion charges are balanced by protein charges

251

Figure 9. Changes in the degree of aggregation (α) for PPCs formed at the early stages of the titration of HSA (open triangles) with KPVS(15/92) as well as of Lyz (open circles) and RNase (closed circles) with KPVS(5/78).

of the opposite sign. The neutral intrapolymer complexes subsequently associate with one another to yield neutral aggregates (AG1) of a uniform size. Further addition of polyions causes an increase in the concentration of AG1 without change in its size. When the AG1 concentration exceeds a certain level (~Vt’),larger aggregates (AG2) form through association of AG1. The mechanism of formation needs to be considered and we also need to explain why the size of AG1 remains unaltered until the free proteins are nearly consumed. The answers to these questions should help to understand the mechanism of PPC formation. The spacing between charges on the proteins is not variable, as well as not enough to be complementary to the very small and uniform spacing between polymer charges. For this reason, and also from considerations of restrictions on polymer chain configurational entropy, we cannot imagine tight ion pairing in the PPC. Rather, we have to consider the proteins in the PPC as "loosely" bound, and the ion pairs as labile and prone to reconfiguration (for example, see Figure 5). Furthermore, there must be local regions of excess positive and negative charge. Thus, while the intrapolymer complex may have net electroneutrality, it is highly "polarizable". This polarizability promotes the formation of aggregates of intrapolymer complexes to form AG1.

protein Figure 10.

252

intrapolymer complex

AG1

Schematic illustration of processes of PPC formation inferred from the present study.

The amount of polarization for the intrapolymer PPC from Lyz should be larger than that from RNase, because the complementarity of the spacing between charges on the protein to the uniform spacing between polymer charges is poorer in the former than in the latter. This has been quite obvious from molecular models of Lyz and RNase. As a result, the intrapolymer PPC from Lyz would have a greater tendency to aggregate one another, compared with that from RNase. It is predictable that the polarizability of intrapolymer PPCs from the same protein should be changed by De, because the spacing between polymer charges is inversely proportional to the linear charge density. Therefore, De may cause a change in the aggregation of the intrapolymer PPC, the magnitude of which should be larger in the Lyz system than in the RNase system. We may predict from eq (14) that is increased with increasing De. On the other hand, the experiments showed that an increase in De leads to a decrease in α. Thus, it would be predictable that when the increase in and the decrease in a compensate for each other, the mass as well as the size of AG1 is little or not at all influenced by De. This is the case in the Lyz system, but not in the RNase system. Thus, we may consider that (i) there is a thermodynamically stable size of AG1 with a definite mass, and (ii) such a size would be attained by dissociation of a part of the bound proteins followed by their “reassociation” within the AG1 to minimize the amount of polarization during the aggregation of intrapolymer PPCs. Such dissociation and reassociation should be easier in the Lyz system than in the RNase system; thus, the former exhibited little change in the size as well as in the mass with the polyion charge density. As a result, it is reasonable to consider that the AG1 in the Lyz system prefers a random coil structure rather than an extended structure. This can be supported by the fact that the change in ρ with De for Lyz is within a range of 1.3 to 1.5. In contrast to De, MP has little or no influence on the aggregation when De remains constant since the chain length does not vary the spacing between polymer charges. Indeed, an increase in MP under a fixed De (~0.77) brought about little change of α in the RNase system. In addition, a linear increase in both size and mass of AG1 was observed, which is explicable in terms of the M P-dependent change in In the case of Lyz, however, MP led to a few changes in both mass and size, but to a marked change in α. Thus, we may say again that in the Lyz system in which the resulting intrapolymer PPC is highly polarizable, the increase in n and the decrease in α are compensate for one another, yielding AG1 with a thermodynamically stable size.

CONCLUSIONS AND TOPICS FOR FUTURE RESEARCH Formation mechanism of PPCs in aqueous non-ionic and ionic polymer systems has been studied at the molecular level by a combination of turbidimetric titration, QELS, SLS, ELS, dialysis method, and fluorescence spectroscopy. In addition, the measurements of enzymatic activity have also been employed in the appropriate cases. It has become apparent that an “intrapolymer” complex in which several protein molecules bound to a chain of the polymer is formed in the PEG-protein system via hydrogen bonding between the polymer and protein. In the KPVS-protein system, however, the polyion charges are neutralized by protein charges of the opposite sign; thus, the resulting intrapolymer complexes immediately associate with one another to yield neutral aggregates (AG1) of a characteristic size. Both surface charge distribution of proteins and linear charge density of polyions have a strong effect on the formation of intrapolymer complexes, and hence their aggregation.

253

In order to regulate the structure of intrapolymer complexes (and their aggregates), more systematic investigations will be of great importance. Polymer samples of varying hydrophobic character, hydrogen-bonding ability, as well as charge density, will provide a key for this purpose.

ACKNOWLEDGMENTS This research was supported by grants to E. K. from the Ministry of Education of Japan (#05044077and#08455434)

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

254

S. Azegami, A. Tsuboi, T. Izumi, M. Hirata, P.L. Dubin, B. Wang and E. Kokufuta, Langmuir 15:940 (1999). A. Tsuboi, T. Izumi, M. Hirata, J. Xia, P.L. Dubin and E. Kokufuta, Langmuir 12:6295 (1996). J. Xia, P.L. Dubin, L.S. Ahmed and E. Kokufuta, In: Macro-ion Characterization: From Dilute Solutions to Complex Fluids; K.S. Schmitz, ed., American Chemical Society, Washington, DC (1994), Chapter 17. L.S. Ahmed. J. Xia, P.L. Dubin andE. Kokufuta J. Macromol. Sci., Pure Appl. Chem. A31:17 (1994). T. Izumi, M. Hirata, E. Kokufuta, H.-J. Cha and C.W. Frank, J. Macromol. Sci., Pure Appl. Chem. A31:31 (1994). T. Izumi, M. Hirata, K. Takahashi and E. Kokufuta, J. Macromol. Sci., Pure Appl. Chem. A31:39 (1994). E. Kokufuta, In Macromolecular Complexes in Chemistry and Biology; P.L. Dubin, R. Davis, C. Thies, J. Bock and D. Schulz, eds.; Springer-Verlag, Heidelberg (1993) Chapter 18. J. Xia, P.L. Dubin and E. Kokufuta, Macromolecules 26:6688 (1993). E. Kokufuta, Prog. Polym. Sci. 17:647 (1992). E. Kokufuta and H. Nishimura, Polym. Bull. 26:277 (1991). E. Kokufuta and K. Takahashi, Polymer 31:1177(1990). E. Kokufuta, H. Shimizu and I. Nakamura, Macromolecules 15:1618 (1981). E. Kokufuta, N. Watanabe and I. Nakamura J. Appl. Polym. Sci. 26:2601(1981). E. Kokufuta, H. Shimizu and I. Nakamura, Macromolecules 14:1178 (1980). E. Kokufuta, H. Shimizu and I. Nakamura, Polym. Bull. 2: 157 (1980). E. Kokufuta, N. Watanabe and I. Nakamura, J. Appl. Polym. Sci. 26:2601(1981). H. Terayama, J. Polym. Sci. 8:243 (1952). E. Kokufuta and S. Iwai, Bull. Chem. Soc. Jpn. 50:3043 (1977). E. Kokufuta, Macromolecules 12:350 (1979). R.A. Kuharski and P.J. Rossky, J. Am. Chem. SOC. 106:5786 (1984). H. Tanaka, H. Touhara, K. Nakanishi and N. Watanabe, J. Chem. Phys. 80:5170 (1984). A. Teramoto, M. Watanabe, E. Iizuka and K. Abe, J. Macromol. Sci., Pure Appl. Chem. A3I:53 (1994). T. Konishi, T. Yoshizaki and H. Yamakawa,Macromolecules 24:5614 (1991). J.J. Hermans,J. Polym. Sci., 18:529 (1955). J.T.G. Overbeek andD. Stigter, Recl. Trav. Chim. 75:543 (1956). M. Nagasawa, I. Noda, T. Takahashi and N. Shimamoto, J. Phys. Chem. 76:2286 (1972). E. Kokufuta, Polymer 21:177 (1980). D.C. Henry, Proc. Roy. SOC. A133:106 (1931). K.J. Laidler and P.S. Bunting. The Chemical Kinetics of Enzyme Action. 2nd edn., Oxford University Press (1973), pp329-341. L. Goldstein, Y. Levin andE. Katchalski, Biochemistry 3:1913 (1964). R. Haynes and K.A. Walsh, Biochem. Biophys. Res. Commun. 36:235 (1969). Y. Levin, M. Pecht, L. Goldstein and E. Katchalski, Biochemisrry 3:1905 (1964).

CRYSTAL -GROWTH RESTRICTION THROUGH CLATHRATE HYDRATE FORMATION: APPLICATIONS TO NANOPARTICLE SYNTHESIS

G. Irvin Jr., S. Li, B. Simmons, V. John, G. McPherson1, C. J. O'Connor2 Department of Chemical Engineering and 1Department of Chemistry Tulane University, New Orleans, LA 70118 2 Department of Chemistry University of New Orleans, LA 70148

INTRODUCTION Nanophase materials are of significant interest due to the size quantization effects that result in novel electrooptical, magnetic, structural and chemical properties.1,2 Several preparative procedures have been proposed in the literature; these include mechanical attrition, laser ablation processes, and a variety of vapor and solution phase synthesis techniques. The techniques are based either on size reduction from the bulk scale (attrition), or various methods to quench growth after particle nucleation. For example, solution phase synthesis using water-in-oil microemulsions (reverse micelles) is a case where growth is limited by compartmentalizing the aqueous synthesis medium into water droplets therefore serving as microreactors for particle nucleation.3 Also, techniques using polymer technologies, polyelectrolyte stabilization, temperature effects, and capping/surface binding4 procedures are standard methods employed to inhibit crystal growth at various length scales. In this paper, we report a new procedure to quench particle growth in solution phase synthesis. We utilize a method to physically limit the growth medium (solvent). Interestingly, we use a co-nucleation route where the particle of interest is nucleated and its growth inhibited as a result of specifically crystallizing the water medium. The technique is based on the concept of clathrate hydrate formation, a naturally occurring phenomenon with interesting thermodynamic properties that we attempt to exploit for nanoparticle synthesis. Clathrate hydrates are crystalline inclusion compounds, formed from water (the host species) and small gas molecules (the guest species).5 They typically form in one of two structures ('just termed Structure I and Structure II), although a third structure (Structure H) was discovered a decade ago and has been the subject of intensive

Advances in Crystal Growth Inhibition Technologies Edited by Amjad, Kluwer Academic/Plenum Publishers, New York, 2000

255

study.6 A unit cell of Structure I hydrate is shown in Figure la illustrating the cage-like lattice formed by the hydrogen bonded water molecules. Gas molecules enclosed within the cavities stabilize the structure through van der Waals interactions with the host water molecules. Hydrate formation is thought to occur through the tendency of water molecules to retain hydrogen bonding in the presence of a non polar solute (the gas species). From a thermodynamic viewpoint, hydrates exhibit interesting univariant properties, the two-component (water+gas) system existing in three-phase (gas + liquid water/ice + hydrate) equilibria.7 For example, Figure 1 b describes the hydrate equilibria for ethylene (the gas used in this study), and the data in the figure indicates that hydrates of ethylene (the VHL1 curve) can form at pressures significantly lower than the vapor pressure of the gas (the VL2 curve).

Temperature (K)

(B) Figure 1. (a) Structure I hydrate. Unit cell 1.2 nm. (b) Ethylene hydrate equilibria. V - Vapor, H Hydrate, L1 - Liquid water, L2 - Liquid state of the hydrate forming gas species (C 2H4). The VHL1 data is from Deaton and Frost.21

256

The clathrate hydrates of small gas species (gas hydrates) have been studied primarily for their relevance to natural gas recovery, storage and transportation,5,8 and prior studies have focussed on their formation in bulk aqueous systems. Here, we attempt to exploit the thermodynamics of hydrate formation to modify inorganic particle formation. The concept behind the work depends on quenching of inorganic particle growth by the rapid formation of hydrates. In other words, if hydrates can be induced to form extremely rapidly, the aqueous phase becomes transformed to solid hydrate phase, thereby compartmentalizing the system volume and limiting ionic transport rates to the crystalline particle surface. Additionally, the rapid supersaturation of ionic precursors as the water phase is converted to the hydrate form should enhance the kinetics of particle formation. It is important to note that the hydrate crystal structure is well defined. It consists of water and the guest molecule and doesn't form with impurities such as a salt. Therefore, no concern is necessary regarding trapping salt in the hydrate crystal. The objective of the work therefore, was to initiate inorganic particle nucleation by contacting aqueous droplets of precursor solutions at hydrate formation conditions. It was our hypothesis that with appropriate design, hydrates can be induced to form extremely rapidly. The most important process variables for this experimental design are temperature, pressure, and salt concentration. By modifying the pressure and temperature we are able to control the rate of hydrate formation. Salt concentration can be an inhibitor to hydrate formation. The salt modifies the chemical potential of the water requiring a higher pressure to induce hydrate formation contrasted with pure water. We also can increase the probability for hydrate formation by increasing the surface area available for contact between the gas and water. That is, by injecting small water droplets into a gas saturated liquid hydrocarbon, the surface area of the water exposed to gas is significantly increased. Under these conditions, we expected hydrate formation to be rapid leading to a quench of inorganic particle growth after nucleation. In effect, the concept is analogous to extremely rapid ice formation. Yet, clathrate hydrates offer several advantages. Hydrate formation can be induced at temperatures above 273.15 K, with water activity being modified through pressure in addition to temperature. Further, recovery of the product is simple. The depressurization of the reaction chamber will initiate hydrate decomposition. A slurry is created that can be transferred to a collection vessel under system pressure Our initial experiments on ferrite synthesis indicate the feasibility of the approach, and the detailed methods and results are described next.

MATERIALS AND METHODS Materials Reagent grade ferrous sulfate, isooctane, methanol, and ammonium hydroxide were purchased from Aldrich-Sigma. Ethylene, the hydrate forming gas species was purchased from Matheson Gas Products (Analytical Grade).

Nanoparticle Synthesis Method The experimental set-up consists of a high pressure glass-windowed view cell (Jerguson Flat Gage, Clark-Reliance Corp.) submerged in a temperature controlled methanol/water bath typically maintained between 273 and 278 K. The view cell has a

257

volume of 120 ml and is equipped with inlet and outlet ports. Figure 2 illustrates a schematic of the experimental procedure and the processes occurring in the view cell. Initially, the cell is charged with ~60 ml isooctane and then contacted with a low molecular weight hydrocarbon gas such as ethylene which is the hydrate former. Typically, the initial charge also contains a very small amount of water (<0.1 ml) which serves as a seed for hydrate crystallization. The system is cooled to between 273.15 and 278.15 K and the pressure adjusted to be reasonably above the hydrate forming pressure (e.g. 2 MPa ) but below the vapor-liquid equilibrium line for the hydrate forming species (Figure 1b). At these conditions, the liquid hydrocarbon phase is saturated with the hydrate forming species. As reference data, at a pressure of 2 MPa and temperature of 274 K, ethylene dissolves in isooctane to a molefraction of 0.4, swelling the solution from a density of 0.7 x 10-3kg/m3 (pure isooctane) to about 0.57 x 10-3 kg/m3 (the calculations were done using the Peng-Robinson equation of state9). Additionally, since the thermodynamic conditions are appropriate for hydrate formation, the small amount of water introduced earlier as the seed for hydrate crystallization becomes converted to hydrates visualized as a trace amount of ice-like crystals at the bottom of the view cell. Hydrates, with an ice-like density of 0.91 x 10-3 kg/m3 have a higher density than the liquid hydrocarbon phase of isooctane saturated with ethylene, and therefore precipitate to the bottom of the view cell. The procedure then, is to inject aqueous precursor solutions of FeSO4 (1M) and NH4OH (5-15 M) through directly opposing nozzles using high pressure syringe pumps, into the liquid hydrocarbon phase. The opposing nozzle separation distance is less than 0.5 cm to facilitate coalescence between droplets. In repeated experiments, the injection flow rates

Figure 2. Schematic of the high pressure view cell, illustrating the system contents and the processes that occur.

were varied from 2-9 ml/min. The coalescence of the precursor-containing solutions leads to particle nucleation. Additionally, it is visually observed that hydrate formation is extremely rapid and virtually no aqueous liquid is observed as a bulk phase. In fact, it appears that hydrates are formed almost as soon as the aqueous droplets are ejected out of the nozzles, with the visual impression of ice crystals precipitating to the bottom of the view cell. Rapid hydrate formation is also facilitated by conditioning the precursor

258

solutions at temperatures between 273-278 K prior to pumping the solutions into the system. An interesting visual observation is the formation of a hydrate bridge from the view cell bottom to the confluence of the two nozzles. However, the nozzles do not become plugged with hydrates, presumably because the system pressure forces liquid hydrocarbon into the nozzles upon cessation of pumping thereby wetting the nozzles with an organic fluid. The coalescence of droplets containing FeSO4 and NH4OH leads to the nucleation of stable iron hydroxides with no further oxidation to ferrites in the oxygen free environment of the view cell. Visually, a dark color is observed mixed in with the hydrate mass at the bottom of the view cell, indicative of the presence of iron hydroxides. After a sufficient amount of iron hydroxides have been formed, the precursor pumping is ceased. The valve at the bottom of the view cell is opened thereby releasing the pressure at the system outlet. This results in hydrate dissociation to gas and liquid water. The system pressure is used to force the slurry of dissociating hydrates, iron hydroxide and hydrocarbon solution out of the view cell, through an 1/8 inch ID stainless steel tube, to a filtration apparatus. A 0.2 µm membrane filter is used to remove the liquid components (water from the melted hydrates, and isooctane) and recover the insoluble iron hydroxide particulates. A small amount of the particulates pass through the 0.2 µm filter but the filter cake rapidly builds up to retain subsequent particles smaller than 0.2 µm. It is also observed that the color of the retentate particulate mass changes from a very dark green to reddish brown. This is indicative of surface oxidation of the particle to the ferrite form. The retentate is collected, air dried for 12 hours at 3 18 K, and characterized through electron microscopy, x-ray diffraction and magnetization property measurements.

Characterization X-Ray Diffraction. A Scintag XDS-2000 equipped with a CuKα1 radiation source and a Si (Li) detector was used to record powder diffraction data. The X-ray tube was operated at 45 kV and 40 mA. A scan rate of 2°/min was employed to scan 2θ from 5° to 60°. Electron Microscopy. For TEM analysis, the particles were dispersed in isooctane by mild sonication. A drop of the dispersed solution was then placed on a carbon-coated Cu grid (100 mesh) and the grid was allowed to dry before insertion into the TEM chamber. A Phillips EM 410 electron microscope was operated at an acceleration voltage of 80 kV to determine particle size and morphology. Electron diffractograms were also recorded for the ferrite samples using the Phillips electron microscope equipped with an EDX unit. Magnetic Measurements. The magnetic properties of the ferrite nanophase materials were characterized using a Quantum Design, Inc. Model MPMS-5S superconducting quantum interference device (SQUID) magnetometer. Calibration has been described elsewhere in detail.10 Typically, two types of experiments were conducted: dc magnetic susceptibility as a function of temperature and magnetization as a function of field. Two different procedures were used for the dc magnetic susceptibility experiments. First, Zero-Field Cooling (ZFC) is utilized where the sample was slowly cooled in zero field to a temperature of 1.7 K. Upon reaching this temperature, the measuring field of 1.0 kG was switched on and the magnetization was measured as a function of temperature. Second, Field Cooling (FC) is utilized complementarily where

259

the field of 1.0 kG was turned on at a temperature well above the superparamagnetic blocking temperature before the sample was cooled to 1.7 K.

RESULTS AND DISCUSSION The initial particle nucleation in the pressurized view cell is done in an oxygen free environment, and the particles generated are expected to be iron hydroxides. Oxidation occurs when these particles are forced out of the cell under the system pressure and collected over a filtration set-up maintained at ambient conditions. Visually, the greenblack color of the hydroxides can be seen changing to a deep orange color upon exposure to air. No further oxidation was carried out, and the filtered particles were dried and imaged. Figures 3a and 3b illustrate the morphology of particles produced through the hydrate method in two separate experiments. The transmission electron micrograph of particles produced through the hydrate method illustrates extremely fine acicular particles interspersed with diffuse regions. The needle-shaped particles have a high aspect ratio with lengths approximately 20-30 nm and widths ~ 5 nm. We attribute the small crystallite size to the formation of the clathrate hydrates. The growth of hydrate crystals

Figure 3. Transmission electron micrographs of iron oxide/oxyhydroxide particles formed through the hydrate process. (a) and (b) show the results of two separate experiments illustrating some variability in experimental data. The inset to (a) is the electron difraction data.

260

inhibits the growth of the ferrites produced. Interestingly, ferrite crystal growth inhibition is achieved by the growth of hydrate crystals! The inset of Figure 3a is the electron diffractogram for that sample and shows a series of concentric rings characteristic of polycrystalline materials with nanoscale dimensions.11 The X-ray powder diffraction data of Figure 4 illustrates broad bands resulting from the ultrafine particle size. Particles of this length scale typically have very low intensity counts and broadening of peaks. The broad band centered around a 2θ of 36°C is indicative both of oxides of iron (maghemite, magnetite) and iron oxyhydroxides (goethite, lepidocrocite )12. Smaller bands are seen around 2θ values of 21°, 27°, 53°, 61° but are hard to ascribe definitively, although the results are suggestive of mixed oxide/oxyhydroxide phases. The computer generated structures indicate both these oxide and oxyhydroxide phases with confidence levels too low to discriminate between the relative forms.

Magnetic Properties Figure 5 illustrates magnetization behavior for the hydrate sample and the control sample. The magnetization vs field data at 300K for particles produced through the hydrate technique (Figure 5a) indicate that the data is virtually superimposable as the field is cycled between ±12 kG, with a coercivity, Hc=0. This lack of hysteresis is characteristic of superparamagnetic particles or some single domain-particles of very small dimensions. The particles have very small volumes, and thermal fluctuations are sufficient to overcome the anisotropy energy barrier, allowing the magnetization to spontaneously reverse direction. Additionally, the sample is only weakly magnetic and does not reach a saturation magnetization at high external field values. Figure 6 illustrates additional magnetic properties of the particles produced by the hydrate method. The magnetization data obtained at 2K (Figure 6a) indicates that the

2

Figure 4. X-ray difraction data. The data represents X-ray diffraction for particles formed by the hydrate route.

261

Figure 5. Magnetization vs. applied field data for particles formed by the hydrate process. The data were acquired at 300 °K.

sample exhibits hysteresis with a coercivity, Hc of 1430 G. The existence of Hc at low temperature and the lack of Hc at high temperature confirm the superparamagnetism. The temperature-dependendent dc magnetic susceptibility (Xdc) data is shown in Figure 6b. At high temperatures, the field-cooled (FC) and zero-field-cooled (ZFC) magnetization data exhibit the same trend. However, at low temperatures, they significantly diverge. The FC curve reaches a plateau, while the ZFC curve shows a decrease in magnetization. From the ZFC data, it is apparent that there is a maximum in the magnetization at ~ 37 K. Ultrafine magnetic particles are known to exhibit such a maximum in ZFC dc susceptibility plots as they tend to mimic spin glass materials.13,14,15 The temperature at which the maximum is observed is an estimation of the blocking temperature Tb, below which the material exhibits coercivity. The observations of Figure 6 are consistent with the behavior of ultrafine magnetic particles.16,17 The lack of saturation magnetization for the particles produced by the hydrate technique (Figures 5 and 6a) may constitute additional evidence of phase inhomogeneity. The quasi-linear response of the material may be indicative of an antiferromagnetic phase, and some of the iron oxyhydroxides (goethite/epidocrocite) are indeed antiferromagnetic. l8 The high temperature magnetic susceptibility data (T > 150 K) indicates Curie-Weiss paramagnetism with C = 63.69 emu.K/g,θ = -316.75 K. The large negative Weiss constant (θ)indicates that there is a substantial amount of antiferromagnetic exchange in this sample substantiating the ideal that the as synthesized material is of mixed phase. CONCLUSIONS The observation that the particles produced through the hydrate technique are significantly smaller than particles produced through direct mixing of reaction precursors constitutes evidence that the concepts of crystal growth quenching by conversion of the

262

Temperature (Kelvin) Figure 6. (a) Magnetization vs. applied field data at 2 °K for particles formed by the hydrate process. (b) dc magnetic susceptibility vs. temperature data. The blocking temperature is estimated at 37 °K.

263

aqueous medium to hydrates are valid. The hydrate process appears to be a simple method to produce nanoparticles in large quantities. The pressures and temperatures needed to form hydrates are very much within the limits of process equipment, the precursors are inexpensive, and the procedure involves simple pumping in of precursors at appropriate thermodynamic conditions. While it is particularly interesting that high aspect ratio nanoparticles of magnetic materials are formed through the hydrate process, there is some variability in morphology. Our continuing research seeks to address process conditions that affect this variability, whether initial precursor concentrations, particle recovery under pressure, etc. have an influence on morphology. Particle aggregation and the visually observed polydispersities are also of concern. Particles synthesized in reversed micelles, for example, have lower polydispersities (albeit with spherical morphologies).19 But it is equally feasible to form hydrates in the microaqueous core of reversed micelles20 and it would be of interest to determine if the growth-quenching concept can be extended to such systems. The weak magnetization of the particles is also of concern. But it must be noted that we have presented data of the as-synthesized materials, which may contain a significant fraction of the weakly magnetic oxyhydroxide phase. There is considerable scope to conduct controlled oxidation of the resulting materials to generate exclusively ferrites. The introduction of reducing agents (such as sodium borohydride) to directly form passivated iron is another potential route to nanomaterials with high magnetizations. The introduction of other materials (e.g. Co, Ba) to form doped ferrites, which can be subsequently annealed to high attain coercivity materials is yet another route to be investigated. It is reasonable to expect that all aspects of aqueous phase precipitation chemistry can be conducted in conjunction with the hydrate technique. The method could be extended to the synthesis of other ceramic materials with novel structural and/or electrooptical properties. The method is therefore simply proposed as an exciting new aspect of crystal growth inhibition to be evaluated extensively in order to determine its relative merits.

ACKNOWLEDGEMENT Funding from DARPA grant MDA972-97-1-0003 is gratefully acknowledged. REFERENCES 1. A. Henglein, Small-particle research: physicochemical properties of extremely small colloidal metal and semiconductor particles Chem. Rev., 89: 186 1 (1 989). 2. R.W. Siegel, Nanostruct. Mater. 24:79 (1993). 3. M. Pileni, Reverse micelles as microreactors J. Phys. Chem. 97:6961(1993). 4. A. D. J. Haymet, L. G. Ward, and M. M. Harding, Winter Flounder "Antifreeze proteins: synthesis and ice growth inhibition of analogues that probe the relative importance of hydrophobic and hydrogen-bonding interactions, J. Am. Chem. SOC., 121:941 (1999). 5. E.D. Sloan, Clathrate Hydrates of Natural Gases, Marcel Dekker, Inc.; New York (1997). 6. Ripmeester, J.A.; Tse, J.S.; Ratcliffe, (C.I.; Powell, B.M. Nature, 325:135 (1987). 7. J.H. van der Waals, J.C. Platteeuw, Clathrate solutions, Adv. Chem. Phys., 2:1 (1959). 8. E. Berecz, M. Balla-Achs, Gas Hydrates, Studies in Inorganic Chemistry 411, Elsevier Press; New York (1977, English Translation 1983). 9. D. Peng, D.B. Robinson, A new two-constant equation of state. Ind. Eng. Chem. Fundam., 15:59 (1976). 10. C. J. O. O'Connor, Magnetochemistry-Advances in theory and experimentation. Prog. Inorg. Chem. 29:203 (1982).

264

11. J. M. Cowley, Electron Diffraction Techniques Vol. I, Oxford Press (1992). 12. Powder diffraction files from the ICDD Database - International Center for Diffraction Data, Newton Square, PA. 13. J.H. Zhang, T.L. Birdwhistell, C.J.O. O’Connor, Magnetic and electrical properties of a new chromium telluride phase: CrTe2 Solid State Commun., 74:443 (1990). 14. C.J.O. O’Connor, In Research Frontiers in Magnetochemistry; C.J.O. O’Connor, Ed.; World Scientific Publishing, Inc.; London, (1993) p 109. 15. D. Fiorani, Comparison between fine particles and spin-glass properties in Studies of Magnetic Properties of Fine Particles and Their Relevance to Materials Science; J.L. Dormann, D. Fiorani, Eds.; Elsevier: Lausanne, (1992) p 135. 16. M.E. McHenry, S.A. Majetich, J.O. Artman, M. DeGraef, S.W. Staley, Superparamagnetism in carboncoated co particles produced by the kratschmer carbon arc process. Phys. Rev. B 49: 11358 (1994). 17. E.M. Brunsman, R. Sutton, S. Ekpatrick, K. Midelfort, J. Williams, P. Smith, M.E. McHenry, S. Majetich, J.O. Artman, M. DeGraef, S. W. Staley, Magnetic properties of carbon-coated, ferromagnetic nanoparticles produced by a carbon- arc method. J.Appl. Phys. 75:5882 (1994). 18. U. Schwertmann, R. M. Cornell, Iron Oxides in the Laboratoly: Preparation and Characterization. Germany (1991). 19. M.A. Lopez-Quintela, J. Rivas, Chemical reactions in microemulsions: a powerful method to obtain ultrafme particles J. Colloid Interface Sci., 158:446 (1993). 20. H. Nguyen, J.B. Phillips, V.T. John, Clathrate hydrate formation in reversed micellar solutions J. Phys. Chem., 93:8123 (1989). 21. W.M. Deaton, E.M. Frost, Oil and Gas Journal, 36:75 (1937).

265

This page intentionally left blank.

ABOUT THE EDITOR Zahid Amjad graduated with B.Sc. (Honours Chemistry) and M.Sc (Chemistry) degrees from Punjab University, Lahore, Pakistan, and obtained his Ph.D. degree in Physical Chemistry from Glasgow University, Scotland. Dr. Amjad was a lecturer at the Institute of Chemistry of Punjab University, and was Assistant Research Professor at the State University of New York at Buffalo, New York when he joined the BFGoodrich Company. Dr. Amjad has presented invited lectures to various national and international meetings, contributed to several books, and published numerous papers on the properties and behavior of water soluble polymers. His current major interests include the study of the interactions between polymers and different substrates in aqueous solution, and the applications of water soluble and water swellable polymers in personal care, pharmaceuticals, and industrial water systems.

267

This page intentionally left blank.

INDEX

Abrasives, calcium pyrophosphates and, 57 Acidic methacrylates, group transfer polymerization and, 198 Acids, calcium carbonate scale control, 218 ACP: see Amorphous calcium phosphate Acrylamide, scale inhibition and, 16 Acrylic acid, calcium phosphonate salts and, 71 Acrylic acid copolymers calcium carbonate scale deposits and, 139 calcium phosphate scale deposits and, 139 Adsorption isotherm, 118 Aeration test, 220 dispersant/phosphonate blend efficacy, 227 Agglomeration PAA and, 204 PAA homopolymers and, 205 Aggregates, neutral, 235 Aggregation, HAP particles, 53 Algae, chlorine absence and, 208 Algal mats, 119 Algicide copper as, 212 in oxidizing aqueous environment, 207-212 Alkaline earth metals, 15 aqueous crystal growth and, 193 Alkaline earth salts, deposit formation, 124 Alkanolamines, halogen oxidizers and, 207 Alkyl phosphates arginine and, 46 calcium phosphate crystallization and, 43.44 heterogeneous nucleation by, 48 induction time effects, 46 lauryl phosphate and, 46 particle size and, 5 1 secondary particle formation, 49 surface tension effects, 46 Alum, calcium carbonate precipitation and, 79 Aluminum brushite crystals and, 9 calcium carbonate precipitation and, 79 calcium oxalate inhibition and, 195 scale-forming salts and, 21 Aminocarboxylic acids, halogen oxidizers and, 207 Aminotris(methylene phosphonic acid), water treatment and, 19

Amorphous calcium phosphate crystallization to HAP, 43 hydroxyapatite and, 43 primary particles of, 49 Amorphous calcium phosphate crystallization, phosphorylated cellulose and, 44 Amorphous magnesium hydroxide, 85 Amorphous precipitate, 1, 10 AMP, calcium carbonate scale inhibition and, 225 Anionic polymers calcium phosphonates and, 71 calcium phosphonates inhibition and, 71 Aqueous crystal growth alkaline earth metals and, 193 tervalent metal ions and, 193 transition metal and, 193 Aqueous solutions algicide effects on copper in, 207 calcium carbonate crystal growth in, 123 copper sequestration/stabilization in, 207 oxidizing, 207 protein complexation with polymers in, 235 surface tension of, 45 Aragonite, 96 crystal formation, 97 precipitation of, 117 transformation to calcite, 142 Arginine, 52 alkyl phosphate and, 46 HAP particles and, 54 Articular cartilage, calcium pyrophosphates and, 58 Bacteria alkanolamines deterioration in, 207 aminocarboxylic acid deterioration in, 207 citric acid deterioration in, 207 Bacterial growth, cooling water treatment and, 217 Barium sulfate brushite crystals and, 9 crystal growth retardation and, 151 DETPMP adsorption on surface, 159 DETPMP inhibition of in brine, 162 electron dispersive spectrometric analysis, 178 HEDP and, 23 solubility of, 153

269

Barium sulfate (cont.) supersaturation index and, 166 temperature effects on, 151 Barium sulfate nucleation, 151 calcium ions effects on PPCA efficiency, 184 ionic strength and, 165 phosphino-polycarboxylic acid and, 165 PPCA and, 169, 175 PPCA inhibition efficiency, 179 PPCA inhibition of, 177 statistical analysis of, 175 temperature effects on, 165 turbidity and, 165 Barium sulfate precipitation calcium phosphino-polycarboxylate and, 165 PPCA and, 166 rate of, 161 Barium sulfate reagent powder, DETPMP adsorption on, 154 Barium sulfate scale DETPMP effects, 151 DETPMP inhibition of, 155 inhibition of, 152 precipitation of, 152 temperature inhibition effects, 155 Benzene hexacarboxylic acid, water treatment and, 19 Biocides calcium carbonate scale control and, 230 inhibitor stability and, 230 oxidizing power of, 231 Bleach, microbiological control and, 227 Boiler systems, 15 Bone mineralization, octacalcium phosphate and, 29 Bone resorption, organophosphorus compounds and, 71 Bromine microbial growth inhibition, 231 microbiological control and, 227 Brushite divalent metal ions and, I foreign metal ions influence on, 1-13 hydrolysis, 2, 6, 11 metastability of, 1-2 octacalcium phosphate and, 2 precipitation of, 3 reaction crystallization, 2 saturation ratio, 3 supersaturation, 3, 11 supersaturation values for, 4 transformation of, 1-13 trivalent metal ions and, 1 zinc and, 1 Brushite crystallization, 10-1 1 course of, 6 hydrolysis and, 1 1 Brushite crystals dimensions of, 5,9 distribution of, 6 inhibition of growth, 1 1 morphologies of, 5

270

Brushite crystals (cont.) morphology of, 1-13 nucleation of, 1 I photomicrographs of, 7-8 promotion of growth, 1 I Cadmium, brushite crystals and, 9 Calcite aragonite transformation to, 142 calcium carbonate and, 127 fulvic acid and, 127 inorganic, mineralization of, I19 polyacrylic acid and, 127 Calcite crystals/deposits aquatic humic substance inhibition of, 107 heat exchangers and, 140 humic substances and, 107 inorganic ion inhibition of, 108 magnesium ion inhibition of, 108 mechanisms of growth, 108 organic compound inhibition of, 108 phosphate ion inhibition of, 108 in seawater, 104 Calcite precipitation fulvic acid and, 129 humic acid and, 129 polyacrylic acid and, 129 Calcite seed crystals, 125 Calcium ammonium pyrophosphates, 57 Calcium atom, coordination environment of, 222 Calcium carbonate calcite and, 127 cooling water chemical treatment and, 215 growth in seawater, 85 HEDP and, 23 pH level and, 2 18 water systems and, 15 Calcium carbonate crystallization aquatic ecosystems and, 107 ecosystems and, 107 in supersaturated conditions, 123 Calcium carbonate formation, 86 cooling water systems and, 108, 217 desalination processes and, 108 inhibition of, 124 oil production and, 108 oil production wells, 86 oxidizing biocides and, 230 polyacrylic acid and, 149 temperature effects on, 139, 142 Calcium carbonate inhibition, 218 AMP efficacy in, 225 dispersant A efficacy in, 225 dispersant/phosphonate blend efficacy, 227 HEDP efficacy in, 225 inhibitor performance and, 225 PBTC efficacy in, 225 pH static test and, 225 polyacrylate efficacy in, 225 screening methodology for, 225 temperature effects on, 193

Calcium carbonate precipitation cationic polyelectrolytes and, 79 magnesium hydroxide and, 85 polyacrylic acid polymer effects on, 101 polyelectrolytes and, 79 seawater solutions and, 85 spherulitic vaterite crystals and, 97 Calcium dihydrogen pyrophosphates, 62 Calcium fluoride, HEDP and, 23 Calcium hydrogen phosphate dihydrate: see Brushite Calcium hypochlorite, 211 Calcium ions activity of, 45 PPCA efficiency and, 184 Calcium nitrate, 3 Calcium oxalate, HEDP and, 23 Calcium oxalate crystallization/deposits in evaporators, 187 growth retardation, 201 prevention in evaporators, 188 scale inhibitor efficacy, 188 temperature effects on, 191, 192 Calcium oxalate inhibition, 189 aluminum and, 195 cationic/anionic ratio and, 193 ionic product and, 193 pyrophosphate and, 194 SHMP and, 194 sodium hexametaphosphate and, 195 sodium triphosphate and, 195 Calcium oxalate monohydrate, 189 Calcium oxalate precipitation, 188 driving force for, 189 hydration states, 188 PAA for inhibition of, 191 pH effects on driving force for, 190 polyphosphate and, 190 Calcium phosphate crystallization/deposits alkyl phosphates and, 43, 44 foreign metal ions and, 2 monoalkyl phosphate effects, 44 orthophosphates and, 230-231 polyacrylic acid and, 149 temperature effects, 139 Calcium phosphate inhibition copolymers and, 24 homopolymers and, 24 poly(acrylic acid) and, 19 poly(acrylic acid copolymer) and, 19 poly(acrylic acid terpolymer), 19 terpolymers and, 24 Calcium phosphate scale formation, 146 Calcium phosphates cooling water chemical treatment and, 215 inhibition of precipitation, 15 precipitation of, 1, 15 scale inhibition, 16 secondary particles of, 49 Calcium phosphino-polycarboxylate, barium sulfate precipitation and, 165

Calcium phosphonates anionic polymers and, 71 corrosion and, 72 inhibition of, 76 precipitation chemistry, 72 precipitation of, 76 scale formation and, 76 steel corrosion and, 76 water treatment and, 72 Calcium pyrophosphates, 57 articular cartilage and, 58 crystallization of, 57 development of, 57 industrial uses of, 57 osteoarthritic joint disease and, 57 phosphocitrate and, 57 phosphocitrate influence on, 62 use in abrasives, 57 Calcium salt crystals mineral deposit formation, 140 phosphocitrate and, 59 Calcium silicate deposits, in sugar evaporators, 187188 Calcium sulfate cooling water chemical treatment and, 215 water systems and, 15 Calcium sulfate dihydrate crystallization block copolymer weight effects on, 202 copolymer weight effects on, 202 diblock copolymers control of, 197 Calcium sulfate precipitation, polymer effects on, 198 Carbonate, OCP conversion into HA, 30 Carboxylic acids scale inhibition and, 16 scale-forming salts and, 22 Catalytic hydrocarbon, cooling water and, 216 Cationic polymeric flocculant, scale-forming salts and, 21 Cedarwood oil, 3 Cellulose, phosphorylated, 44 Ceramics, calcium pyrophosphates and, 57 Chelating agents, 15 corrosion and, 2 I iron and, 15 polymers as, 25 Chlorine copper persistence in absence of, 208, 210 microbial growth inhibition, 231 Chlorine persistence, copper staining and, 209, 211 Chromate, corrosion inhibition and, 16 Chromium, brushite crystals and, 9 Citrate, OCP conversion into HA, 30 Citric acid halogen oxidizers and, 207 water treatment and, 19 Clathrate hydrates, 255 characterization of, 259 crystal-growth restriction through, 255 electron microscopy and, 259 gas hydrates and, 257

271

Clathrate hydrates (cont.) magnetic measurements, 259 x-ray diffraction and, 259 CMC: see Critical micellization concentration Cooling tower components of, 216 pilot: see Pilot cooling tower Cooling water systems aeration test, 220 bacterial growth and, 217 calcium carbonate scale formation and, 107 corrosion and, 217 corrosion products and, 215 corrosion rate monitoring, 221 effectiveness of treatment, 232 evaporation and, 217 iron oxides and, 215 microbiological growth and, 217 mineral scale deposits and, 217 model of, 216 pH static test, 220 saturation index and, 215 scale and microbial control, 215 scale inhibitors in, 218 Copolymers, 18 acrylic acid, 139 calcium phosphate inhibition and, 24 crystal growth in, 200 diblock, 197 group transfer polymerization and, 198 gypsum agglomeration and, 203 gypsum crystal habit and, 203 Copper acute toxicity of, 212 biocidal activity of, 207 brushite crystals and, 9 chlorine absence and, 208 scale-forming salts and, 20 sequestration in aqueous environment, 207 stabilization in aqueous environment, 207 toxicity tests, 210 Copper–ethanolamine complex, 208 Copper precipitation in oxidizing conditions, 207 polyelectrolyte polymer effects, 208 Copper staining, chlorine persistence and, 209, 211 Copper sulfate pentahydrate, 208 Copper-triethanolamine complex, 208 Corrosion, 15 calcium phosphonates and, 76 chromate-based inhibitors, 16 cooling water chemical treatment and, 215 inhibitors of, 16 monitoring rate of, 221 phosphonates and, 72 sugar evaporators and, 196 Critical micellization concentration, 47 Crude oil production, organophosphorus compounds and, 71 Crystal aggregates, polyelectrolytes and, 33

272

Crystal class m, 1 mirror plane, I monoclinic domatic class, 1 Crystal growth clathrate hydrate formation effects, 255 HAP particles, 53 inhibition of, 11, 33 promotion of, 11 Crystal morphology, polyelectrolytes effects on, 203 Crystallization of brushite, 10-11 calcium phosphate, 2 diblock copolymers control of, 197 foreign metal ions and, 2 Debye-Hückel parameter, 243 Dentine mineralization, octacalcium phosphate and, 29 Dentifrices, calcium pyrophosphates and, 57 Deposit, terpolymers and, 76 Desalination, 15 calcium carbonate scale formation and, 107 mineral scales and, 15 DETPMP: see Diethylenetriamine penta (methylene phosphonic acid) Diblock copolymers, crystallization control, 197 Dicalcium phosphate dihydrate, hydroxyapatite and, 43 Diethylenetriamine penta (methylene phosphonic acid) adsorption on barium sulfate reagent powder, 154 adsorption on barium sulfate surface, 159 barium sulfate inhibition in brine, 162 barium sulfate precipitation and, 161 barium sulfate scale formation and, 151, 155 temperature effects and, 151 Dispersant A, calcium carbonate scale inhibition and, 225 Dispersant/phosphonate blend aeration test and, 227 calcium carbonate scale inhibition and, 227 Electron dispersive spectrometric analysis, barium sulfate identification, 178 Electron microscopy, clathrate hydrates and, 259 Electrophoretic light scattering, 235 Electroplating, organophosphorus compounds and, 71 ELS: see Electrophoretic light scattering Equilibria, oxalate ion control and, 189 Ester, scale inhibition and, 16 Evaporation system, see also Sugar evaporators cooling water treatment and, 217 Ferric chloride, calcium carbonate precipitation and, 79 Ferric ions, rust scale prevention and, 24 Ferric sulfate, calcium carbonate precipitation and, 80 Fertilizers, calcium ammonium pyrophosphates and, 57

Fluoride, OCP conversion into HA, 30 Fulvic acid calcite and, 127 calcite precipitation and, 129 calcium carbonate crystal growth in, 123 calcium carbonate precipitation and, 79 calcium phosphonate salts and, 71 scale formation and, 78 seeded growth experiments with, 132 water treatment and, 19 Gas hydrates, clathrate hydrates and, 257 Geothermal energy exploitation, alkaline earth insoluble salt deposits and, 124 Gluconic acid, water treatment and, 19 Glucose, water treatment and, 19 Group transfer polymerization, 198 Gypsum, diblock copolymers control of, 197 Gypsum agglomeration, block copolymers and, 203 Gypsum crystal habit, block copolymer effects, 203 HA: see Hydroxyapatite Halogen oxidizers, 207 HAP: see Hydroxyapatite Heat exchange tubes, pilot cooling tower tests and, 220 Heat exchanger surfaces, polymers and, 78 Heat exchangers, 15 calcium salt formation in, 140 evaporator thermal efficiency and, 187 fouling of, 187 scale deposits in, 15 scale-forming salts in, 15 surface cleanliness, 187 Heat transfer calcium carbonate scale and, 215 mineral scales and, 15 Heat transfer efficiency, crystal growth and, 107 Heat transfer surfaces, calcium carbonate scale formation on, 107 HEDP: see Hydroxyethylidine I, 1-diphosphonic acid Henry’s function, 243 Homopolymers, 18, 20 agglomeration and, 205 calcium phosphate inhibition and, 24 HSA: see Human serum albumin HSA–PEG complex electrophoretic behavior of, 243 pH effects on, 238 structure of, 240 urea effects on, 239 Human serum albumin, polyethylene glycol and, 236 Humic acid affinity constant for, 135 calcite precipitation and, 129 calcium carbonate crystal growth in, 123 seeded growth experiments with, 132 Humic substances aquative, calcite crystal growth effects, 107 chemical functionality of, 115 molecular structure of, 115 scale formation and, 78

Humic substances (cont.) scale-forming salts and, 21 water supply and, 16 Hydrogen carbonate, barium sulfate scale formation and, 156 Hydrolysis of brushite, 11 brushite and, 6 x-ray diffraction and, 10 Hydrophobic methacrylates, group transfer polymerization and, 198 Hydroxy acids, scale-forming salts and, 22 Hydroxyapatite ACP crystallization to, 43 crystallization of, 43 dicalcium phosphate dihydrate and, 43 growth of, 43 nucleation on brushite, 12 OCP hydrolysis in water and, 31 OCP transformation to, 2 octacalcium phosphate and, 29 polyacrylic acid and, 3 1 supersaturation and, 43 Hydroxyapatite particles aggregation of, 53 arginine and, 54 growth of, 53 zeta potential of, 52 Hydroxyapatite precipitation, OCP hydrolysis and, 29 Hydroxyethylidine 1.1-diphosphonic acid calcium carbonate scale inhibition and, 225 water treatment and, 19 Hydroxyethyline 1,1-diphosphonic acid, scale-forming salts and, 23 Hypercalcemia, organophosphorus compounds and, 72 Induction time alkyl phosphate effects on, 46 barium sulfate nucleation, 177 nucleation theory and, 100 Industrial systems calcium salt formation in, 140 cooling water: see Cooling water systems Inhibitor stability, oxidizing biocides and, 230 Inorganic ion, calcite crystal growth and, 108 Iodine, microbial growth inhibition, 231 Iron brushite crystals and, 9 calcium carbonate precipitation and, 79 calcium phosphate precipitation and, 15 chelating agents and, 15 polymer performance and, 20 scale-forming salts and, 20 Iron chelants, calcium phosphate-inhibiting polymers and, 15 Iron oxides, cooling water chemical treatment and, 215 Kidney stones, organophosphorus compounds and, 71 KVPS: see Potassium poly(vinyl alcohol) sulfate

273

Laser ablation processes, 255 Lauryl phosphate, 48 alkyl phosphate and, 46 particle size and, 51 Lauryl phosphoric acid, water-insoluble, 5 I Lead, brushite crystals and, 9

Nucleation alkyl phosphates and, 48 barium sulfate, 169 brushite crystals and, 11 heterogeneous,48 phosphocitrate and, 59 scale formation and, 218 Nucleation theory, induction time and, 100

Magnesium, brushite crystals and, 9 Magnesium hydroxide, calcium carbonate precipitation and, 85 OCP: see Octacalcium phosphate Magnesium ion, calcite crystal growth and, 108 OCPP: see Octacalcium phosphate pentahydrate Octacalcium phosphate, 2 Magnesium silicate deposits, in sugar evaporators, I88 bone mineralization and, 29 Maleic acid, calcium phosphonate salts and, 71 calcium acetate synthetization of, 31 Manganese conversion into HA, 30 brushite crystals and, 9 dentine mineralization and, 29 scale-forming salts and, 20 hydrolysis in water, 31 hydrolysis of, 35 Marangoni effect, 3 Mechanical attrition, 255 hydroxyapatite and, 2, 29 Metal ions metastability of, 2 divalent, I nucleation of, 38 foreign nucleation on brushite, 12 brushite crystal morphology and transformation scanning electron microscopy of, 31 sodium polyacrylate and, 29-41 and, 1-13 brushite hydrolysis and, 11 sodium polyacrylate effects on growth, 38 supersaturation and, 4 triclinic structure of, 29 scale-forming salts and, 20 Octacalcium phosphate crystallization, acidic protein, tervalent, 193 31 trivalent, I Octacalcium phosphate pentahydrate, hydroxyapatite and, 43 Metal surfaces, corrosion of, I6 Methacrylates, group transfer polymerization and, Octacalcium phosphate synthesis, sodium 198 polyacrylate and, 32 Microbial growth Oil production biocides for control of, 231 alkaline earth insoluble salt deposits and, 124 bromine control of, 231 calcium carbonate scale formation and, 108 chlorine control of, 231 calcium carbonate scale formation in wells, 86 iodine control of, 23 I temperature and, 151 Organophosphorus compounds, 7 I ozone control of, 231 Microbiological growth affinity constant for, 135 bleach and, 227 hypercalcemia and, 72 bromine and, 227 industrial uses for, 71 cooling water treatment and, 217 osteoporosis and, 72 Microscopy Paget’s disease and, 72 use in biological systems, 71 optical, 2 scanning electron, 3 1 : see Scanning electron miOrthophosphate, calcium phosphate scale and, 230croscopy 231 transmission electron: see Transmission electron Osmosis,reverse,124 microscopy scale formation and, 197 Mineral crystallization Osteoarthritic joint disease, calcium pyrophosphates inhibition of growth, 16 and, 57 inorganic, 187 Osteoporosis, organophosphorus compounds and, 72 water systems and, 15 Ostwald’s Stufenregal, I Molybdates, corrosion inhibition and, 16 Ozone, microbial growth inhibition, 231 Monoalkyl phosphate, calcium phosphate crystallization and, 44 PAA, growth retarding effect, 199 Paget’s disease, organophosphorus compounds and, Monoclinic domatic class, I Myristyl phosphate, 46.48 72 Paper/pulp slurries, organophosphorus compounds Nanoparticle synthesis, 255 and, 71 NaPA: see Sodium polyacrylate Particles, secondary, formation, 49 Nickel, brushite crystals and, 9 PBC: see Periodic bond chain

274

PBTC: see 2-Phosphonobutane 1,2,4-tricarboxylic acid PC: see Phosphocitrate Pepsin, polymer-protein complex formation and, 236 Periodic bond chain, I pH, measurement of, 2 pH meter, 2 pH static test, 219 calcium carbonate scale inhibition and, 225 Phosphate ion, calcite crystal growth and, 108 Phosphates alkaline earth metal, 15 corrosion inhibition and, 16 Phosphino-polycarboxylic acid analysis of, 168 barium sulfate nucleation and, 165, 175 barium sulfate nucleation inhibition, 177, 179 barium sulfate precipitation and, 166 calcium-complexed, 177 temperature effects on, 184 2-Phosphobutane I ,2,4-tricarboxylic acid, water treatment and, 19 Phosphocitrate calcium pyrophosphates and, 57,62 calcium salt crystals and, 59 characteristics of, 59 molecular modeling studies, 59 Phosphonates, see also Calcium phosphonates alkaline earth metal, 15 corrosion inhibition and, 16 oxidation of, 231 scale-forming salts and, 23 Phosphonic acid, scale inhibition and, I6 2-Phosphonobutane 1,2,4-tricarboxylic acid, calcium carbonate scale inhibition and, 225 Phosphophoryn, OCP crystallization and, 31 Phosphoric acid, gypsum production and, 197 Pigment dispersion, organophosphorus compounds and, 71 Pilot cooling tower testing, 220, 227 Pipe blockage, mineral scales and, 15 Plant growth, calcium ammonium pyrophosphates and, 57 PMAA homopolymers, growth retarding effect, 199 Poly(acrylamide), 75 Polyacrylates calcium carbonate formation and, 124 calcium carbonate scale inhibition and, 225 Polyacrylic acid calcite and, 127 calcite precipitation and, 129 calcium carbonate crystal growth in, 123 calcium carbonate deposit formation and, 149 calcium phosphate deposit formation and, 149 hydroxyapatite and, 31 scale inhibition and, 188 seeded growth experiments with, 132 Poly(acrylic acid), 75 calcium carbonate precipitation and, 79 molecular weight of, 203 water treatment and, 19

Poly(acrylic acid polymer) calcite crystal grown in seawater and, 104 seawater and, 101 Poly(acrylic acid:diacetone acrylamide), water treatment and, 19 Polyaminopolyether tetra(methylene phosphonic acid), scale inhibition and, 188 Poly(aspartic acid), calcium carbonate precipitation and, 79 Polycrylic polymer, as inhibitor in seawater, 91 Polyelectrolyte polymers, copper precipitation and, 208 Polyelectrolytes, 75, 124 calcium carbonate precipitation and, 79 carboxylate-rich, 31 cationic calcium carbonate precipitation and, 79 calcium phosphonates and, 71 crystal aggregates and, 33 crystal morphology effects, 203 molecular weight of, 203 PPC formation and, 244 protein complexation and, 235 sodium polyacrylate and, 33 thermogravimetric analysis, 36 Poly(ethylene glycol), potassium poly(vinyl alcohol) sulfate and, 235 Polymaleates, calcium carbonate formation and, 124 Poly(maleic acid), 7.5 Polymeric agents, corrosion inhibition and, 16 Polymerization, group transfer, 198 Polymer-protein complexes, 235 formation in KVPS system, 243 formation in PEG system, 236 formation of, 235, 251 QELS detection of, 250 SLS detection of, 250 Polymers anionic: see Anionic polymers calcium phosphate precipitation by, 15 as chelating agents, 25 heat exchanger surfaces and, 78 iron(III) effects on, 20 molecular weight of, 202 protein complexation with, 235 water treatment and, 18 Polymethacrylates, growth inhibition effects, 205 Poly(methacrylic acid), 75 Polyphosphates calcium oxalate precipitation and, 190 corrosion inhibition and, I6 Polyvinylalcohol, 43 adsorption of, 43 Potassium dihydrogen phosphate, 3 Potassium hydrogen phosphate, 3 Potassium poly(vinyl alcohol) sulfate, 235 poly(ethylene glycol) and, 235 PPC: see Polymer-protein complexes PPCA: see Phosphino-polycarboxylic acid Pseudogout, calcium pyrophosphates and, 57 PVA: see Polyvinylalcohol Pyrophosphate, calcium oxalate inhibition and, 194

275

QELS: see Quasi-elastic light scattering Quasi-elastic light scattering, 235 intrapolymer PPC detection and, 250 Rayleigh ratio, 240 Recirculating water systems, 15 Reverse osmosis membranes, scale deposits in, 15 Rheumatoid joints, calcium pyrophosphates and, 67 Salicylic acid, water treatment and, 19 Salts scale-forming, 15 aluminum and, 21 carboxylic acids and, 22 cationic polymeric flocculant and, 21 hydroxy acids and, 22 metal ions effects on, 20 phosphonates and, 23 polymers and, 18 soluble, humic substances and, 16 Saturation index, scaling tendency of water, 215 Saturation ratio, of brushite, 3 Scale formation, see also specific types calcium carbonate, 86 cooling water chemical treatment and, 215 mechanism of, 218 nucleation and, 218 oil field production system and, 151 reverse osmosis and, 197 supersaturation and, 218 temperature effects on, 151 terpolymers and, 76 Scale inhibitors, see also specific types calcium carbonate scale control, 218 Scale removal, organophosphorus compounds and, 71 Scanning electron microscopy, OCP hydrolysis and, 35 Seawater CaCO3 precipitation in, 92 calcium carbonate growth in, 85 polyacrylic acid polymer effects, 101 polycrylic polymer as inhibitor, 91 Seeded growth experiments, 132 SEM: see Scanning electron microscopy Sequestrants, calcium carbonate scale control, 218 Serime, phosphorylated, 43 SHMP: see Sodium hexametaphosphate Silica cooling water chemical treatment and, 215 solubility of, 187 Silica deposits, in evaporators, 187 Silicate deposits corrosion inhibition and, 16 in evaporators, 187 Silicate salts, cooling water chemical treatment and, 215 SLS: see Static light scattering

276

Sodium hexametaphosphate, 195 Sodium polyacrylate OCP growth and, 38 OCP hydrolysis and, 35 OCP nucleation and, 38 OCP synthesis and, 32-34 octacalcium phosphate and, 29-41 synthesis and, 33 Sodium triphosphate, 195 Spectrophotometry, 18 Static light scattering, 235 intrapolymer PPC detection and, 250 STP: see Sodium triphosphate Stufenregal, 1 Sugar evaporators beet, deposits in, 187 calcium oxalate deposit prevention, 188 calcium oxalate deposits in, 187 calcium silicate deposits in, 188 corrosion of, 196 energy consumption and, 196 fouling of, 196 inorganic mineral scale control in, 187 magnesium silicate deposits in, 188 mineral scales in, 196 organic sludge, 196 silica deposits in, 187 silica solubility in, 187 silicate deposits in, 187 thermal efficiency of, 187 Sugar loading, 187 Sugar mill, evaporator thermal efficiency effects, 187 Sulfonic acid, scale inhibition and, 16 Supersaturation, 3,5 1 barium sulfate precipitation and, 166 brushite and, 11 calcium carbonate crystal growth in, 123 calculated values for, 4 foreign metal influence on, 4 PPCA inhibition of barium sulfate nucleation, I84 scale formation and, 218 seawater, 85 Surface tension alkyl phosphate effects, 46 aqueous solution and, 45 critical micellization concentration and, 47 Swimming pool chlorine absence, 210 copper precipitation in, 207 copper precipitation prevention, 208 standard water conditions for, 209 TA: see Tannic acid Tannic acid calcium carbonate precipitation and, 79 calcium phosphonate salts and, 71 water treatment and, 19 Tannins, calcium phosphate inhibition and, 21 TEM: see Transmission electron microscopy

Temperature barium sulfate nucleation and, 165 calcium carbonate inhibition and, 193 calcium carbonate scale formation and, 139 calcium oxalate crystal growth and, 191, 192 calcium phosphate scale formation and, 139 deposits in evaporation systems and, 187 DETPMP effects on barium sulfate scale formation and, 151 Tensiometer, 45 Terpolymers, 18 calcium phosphate inhibition and, 24 deposit formation and, 76 scale control and, 76 Tervalent metal ions, aqueous crystal growth and, 193 Tetradentate AMP, 224 Thermal efficiency, sugar evaporators and, I87 Thermogravimetric analysis, polyelectrolyte content, 36 Titration curves, quantitative analysis of, 248 Tooth enamel, organophosphorus compounds and, 71 Transition metal, aqueous crystal growth and, 193 Transmission electron microscopy, OCP hydrolysis and, 35 Triclinic pinacoidal symmetry, 29 Tridentate AMP, 224 Trypsin-KVPS system, stoichiometric complex, 244 Turbidimetric titration curves, 246 Turbidity, calcium inhibitor tolerance and, 219 Urea, HSA and PEG complexation and, 239

Vaterite crystals, 97 spherulitic, 97 Water, supersaturated, 15 Water conservation, phosphonates and, 74 Water supply, humic substances in, 16 Water systems cooling, 74 mineral scales in, 15 Water treatment alkaline earth insoluble salt deposits in, 124 aminotris(methylene phosphonic acid), 72 cooling: see Cooling water treatment corrosion inhibition and, 16 hydroxyethylidine 1,1diphosphonic acid, 72 industrial, calcium phosphonates and, 76 organophosphorus compounds and, 7 1 phosphonates and, 72 2-phosphonobutane 1,2,4-tricarboxylic acid, 72 polymers and, I8 surface, 73 X-ray diffraction, 3 clathrate hydrates and, 259 hydrolysis products and, 10 Yellow metal corrosion inhibition, 218 Zena potential, HAP particles and, 52 Zinc brushite and, I brushite crystals and, 9 scale-forming salts and, 20 Zinc salts, corrosion inhibition and, 16

277