Influence of the Chain Composition on the Thermodynamic Properties

Department of Colloid Chemistry, Eo¨tvo¨s University, P.O. Box 32,. Budapest 112, Hungary H-1518. Received August 31, 2004. In Final Form: October 2...
0 downloads 0 Views 133KB Size
Download PDF
Langmuir 2005, 21, 761-766

761

Influence of the Chain Composition on the Thermodynamic Properties of Binary and Ternary Polymer Solutions Katalin F. Csa´ki, Miklo´s Nagy, and Ferenc Csempesz* Department of Colloid Chemistry, Eo¨ tvo¨ s University, P.O. Box 32, Budapest 112, Hungary H-1518 Received August 31, 2004. In Final Form: October 26, 2004 Thermodynamic properties of binary and ternary polymer solutions (one or two uncharged polymers in one solvent) were studied. Poly(vinyl pyrrolidone) (PVP), fully hydrolyzed poly(vinyl alcohol) (PVA) homopolymers, and water-soluble poly(vinyl alcohol-co-vinyl acetal, -vinyl propional, and -vinyl butiral) copolymers with various acetal content and chain structure, respectively, were used in the experiments. The hydrophilic/hydrophobic character of the PVA-based macromolecules and their compatibility with the PVP homopolymer were systematically regulated by changing the chemical structure of the copolymers (acetal content and/or length of side chains). The water activities in binary and ternary solutions of the chemically different polymers were determined by a gel-deswelling method developed here for ternary solutions. On the basis of the Flory-Huggins theory, the relevant solvent-segment and segment-segment pair interaction parameters (χ) have been calculated. The χ12 segment-solvent interaction parameters proved to be sensitive indicators for changes in the chemical structure of the copolymers. With increase of either the acetal content or the length of side chains in the copolymer, χ12 approached the value characteristic of a theta condition. No significant differences could be revealed in the segment-segment interaction parameters obtained for the PVP-copolymer mixtures with various acetal derivatives, when the χ12 and χ13 interaction parameters determined in binary solutions were used in the calculations for χ23. Determination of the parameters χ1,23 as suggested by Panayiotou, however, showed that increasing the acetal content or the length of the hydrophobic side chains in the copolymer resulted in a reduction in the interaction between the PVA “acetals” and PVP molecules.

1. Introduction Water-soluble polymers are extensively used in a variety of practical applications such as in flocculation or stabilization of dispersions, in foaming, emulsification, and surface modification, and in many other fields. Using copolymers of various chemical structure may provide particular advantages, since varying the composition of a copolymer enables producing macromolecular substances with specific properties required for several technologies. Namely, changing the chemical structure of a copolymer may bring about considerable alterations in their thermodynamic properties, solubility, and compatibility with other polymers and, also, in the interfacial behavior1 of the polymeric compounds. The thermodynamic characteristics of macromolecular solutions are closely related to the magnitude of relevant (solvent-segment and segment-segment) interaction energies. The quality of solvent is, therefore, a key factor that substantially affects the chain dimensions in solution and, also, how a dissolved polymer acts in an application. As Shiomi et al.2 showed the thermodynamic interactions between a random copolymer and the solvent are affected even by the interactions between the different segments constituting the random copolymer. The classical mean-field theory derived by Flory3 and Huggins4 is still widely used to represent the thermody* Corresponding author. Tel.: +36-1-372-2544. Fax: +36-1-2090602. E-mail address: [email protected] (F. Csempesz). (1) Csempesz, F.; Csa´ki, K.; Kova´cs, P.; Nagy, M. Colloids Surf., A 1995, 101, 113-121. (2) Shiomi, T.; Tohyama, M.; Endo, M.; Sato, T.; Imai, K. J. Polym. Sci., B 1996, 34 (15) 2599-2606. (3) Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953.

namic properties of polymer solutions. In a binary solution, the quality of solvent is characterized through the interaction parameter (χ12). A high positive value of χ12 indicates a repulsive interaction between polymer segments and solvent molecules and causes a positive enthalpy of mixing which tends to inhibit the dissolution process. The value of χ12 slightly depends on the molar mass, the temperature, and the composition of the solution. For a ternary system the polymer-polymer interaction parameter (χ23) is used to describe the interaction between the two polymeric components of solution. Whereas the change of entropy is extremely small in the event of admixing polymers, generally two polymers are compatible if their χ23 parameter is negative or close to zero. In general, the values of the polymer-polymer interaction parameters obtained from measurements performed in ternary systems depend significantly on the solvent used.5 Furthermore, the two polymer-solvent interaction parameters are also dependent on the composition of the solution.6,7 Accordingly, the χ23 parameter, calculated from the experimental χ12 and χ13 parameters determined in binary solutions, can be regarded as estimated values. The recent studies in this field are mainly focused on investigating the miscibility of polymer blends. Few attentions are devoted to the solution properties of multicomponent polymer systems, but the results of thermodynamic investigations may also have relevance in an adequate characterization of the adsorption behavior of chemically different polymers. (4) Huggins, M. J. Phys. Chem. 1942, 46, 151. (5) Prolongo, M. G.; Masegosa, R. M.; Horta, A. Macromolecules 1989, 22, 4346-4351. (6) Pinder, D. N. Macromolecules 1997, 30, 226-235. (7) Riedl, B.; Prud′homme. R. E. Polym. Eng. Sci. 1984, 24, 12911299.

10.1021/la047827h CCC: $30.25 © 2005 American Chemical Society Published on Web 12/16/2004

762

Langmuir, Vol. 21, No. 2, 2005

Csa´ ki et al.

In previous papers8 we reported on thermodynamic studies carried out in binary and ternary solutions of homopolymers. The aim of this study is to reveal how the alterations in the chemical structure of the dissolved polymers are reflected in the relevant pair interaction parameters and in the compatibility of chemically different polymers in solution. The present work therefore involves thermodynamic investigations on binary solutions of copolymers with tailored chemical structure and also on their ternary solutions formed with a homopolymer. By changing the chemical structure of poly(vinyl alcohol-covinyl acetal, -vinyl propional, and -vinyl butiral) copolymers (“acetal” content or length of side chains), respectively, the hydrophilic/hydrophobic character of the macromolecules and their compatibility with poly(vinyl pyrrolidone) samples were systematically regulated. To characterize the thermodynamic properties of the polymer solutions, the thermodynamic activity of the solvent has been determined by a gel-deswelling method, and the relevant χ12, χ13, and χ23 Flory-Huggins pair interaction parameters have been calculated. Another theoretical approach to solvent-segment interactions based on the determination of mixed χ1,23 interaction parameters in ternary solutions has also been tested, and the results are compared.9 Methods of Calculation The well-known Flory-Huggins expression for the chemical potential of the solvent in a moderately concentrated polymer solution is as follows

∆µ1 ) RT[ln(1 - v2) + (1 - 1/x)v2 + χ12v22]

(1)

where v1 and v2 are the volume fractions of the solvent and polymer, respectively, and x is the ratio of the number average molar volume of a polydisperse polymer and the molar volume of the solvent, which can be obtained from molar mass and density data. From the solvent activity data determined at various polymer volume fractions, the χ12 interaction parameter can be calculated. For a ternary solution, where a solvent (component 1) and two polymeric components (2 and 3) are present, the chemical potential of solvent is

∆µ1 ) RT[ln v1 + (1 - v1) - v2(x1/x2) - v3(x1/x3) + (χ12v2 + χ13v3)(v2 + v3) - χ23(x1/x2)v2v3] (2) If the volume fractions (v1, v2, and v3) and the corresponding solvent-segment interaction parameters (χ12 and χ13) for both polymers are known, from the solvent chemical potentials determined in ternary solutions, the polymer-polymer interaction parameter (χ23) can be calculated

-χ23 ) [(∆µ1/RT) - ln v1 - (1 - v1) + v2(x1/x2) + v3(x1/x3)](x2/x1)v2-1v3-1 [(χ12v2 + χ13v3)(v2 + v3)](x2/x1)v2-1v3-1 (3) The approach developed by Panayiotou9 involves determination of a mixed χ1,23 interaction parameter in ternary solvent/polymer/polymer mixtures, as proposed by Eliassi.10 χ1,23 can be regarded as a mean parameter characterizing the solvent-segment interactions in solution of chemically different macromolecules. The mixed (8) Csempesz, F.; Csa´ki, K. F.; Nagy, M. Colloids Surf., in press. (9) Panayiotou, C.; Vera, J. H. Polym. J. 1984, 16, 89-102. (10) Eliassi, A.; Modarress, H. Eur. Polym. J. 2001, 37, 1487-1492.

χ1,23 parameter can be calculated from the solvent activity data as follows

(

ln a1 ) ln v1 + 1 -

)

x1 v + χ1,23v232 x23 23

(4)

It can be related to the individual solvent-segment interaction parameters

χ1,23 ) [(χ12v2 + χ13v3)(1 - v1) - χ23′v2v3]/v232 (5) where 3

v23 )

v i ) 1 - v1 ∑ i)2 3

x23 )

yixi ∑ i)2

χ23′ )

x1 χ x2 23

yi denotes the mole fraction of component i at zero solvent concentration. The above-defined “mixed” parameter offers the advantage that it can be computed directly from the measured solvent activity data. χ1,23 can be regarded as a mean characteristic of the solvent-segment interactions in a ternary solution. For the calculation of this “mixed” parameter, knowledge of χ12 and χ13 is not necessary. 2. Experimental Section Materials. Water-soluble uncharged polymers, such as hydrolyzed11 poly(vinyl alcohol) (PVA), poly(vinyl alcohol-co-vinyl acetal, -vinyl propional, and -vinyl butiral) copolymers, and poly(vinyl pyrrolidone) (PVP), were used. All polymers were fractionated samples, prepared from commercial products of Poval 420 poly(vinyl alcohol) (Kuraray Co., Japan) and Fluka K-90 poly(vinyl pyrrolidone), respectively. The hydrophilic/hydrophobic character of the macromolecules and their compatibility with the homopolymer were systematically regulated by changing the chemical structure of the copolymers. Poly(vinyl alcohol)-based copolymers with various acetal, propional or butiral contents and chain structure were prepared by inducing a chemical reaction between the OH groups of the hydrolyzed PVA molecules and acetaldehyde, propionaldehyde, and butyraldehyde, respectively, at pH 2, using hydrochloric acid as catalyst.12,13 The reactions were carried out in water with acetaldehyde, in 20% (V/V) water/ethanol mixture with propionaldehyde, and in 41% (V/V) water/ethanol mixture with butyraldehyde. Acetal Formation. To check the kinetics of acetal formation, the aldehyde conversion was continuously followed by measuring accurately the concentration of the nonreacted aldehyde in the presence of polymer. The concentration of the “free” aldehyde in the solution was measured spectrophotometrically by using a modified Schiff reagent:14 0.68 g of NaHSO3 was dissolved in distilled water and then 13 cm3 of 1 g‚dm-3 fuchsine solution and 0.63 cm3 of HCl was added. The solution was used after 1 day. Determination of the Aldehyde Concentration in Polymer Solutions. A 1 cm3 portion of reagent was added to 3 cm3 of aldehyde solution, and 10 min after the mixing of the two solutions, the absorbance of the mixture was measured at a (11) Nagy, M.; Gyo¨rgyi-Edele´nyi, J. Acta Chim. Hung. 1989, 126, 507-517. (12) Flory, P. J. J. Am. Chem. Soc. 1938, 61, 1518. (13) Solomons, T. W. G.; Fryhle, C. B. Organic Chemistry; John Wiley & Sons. Inc.: New York, 1998. (14) Naumann; et al. Anal. Chem. 1960, 32, 307.

Thermodynamic Properties of Polymer Mixtures

Langmuir, Vol. 21, No. 2, 2005 763

Figure 1. Kinetics of acetal formation: Change of the acetaldehyde content (mol %) with the conversion time (t) during the reaction of the aldehyde with PVA (pH ) 2.0, T ) 298 K).

Figure 2. Water activity curves for PVAL/7.0, PVPr/7.0, and PVBu/7.0 copolymer samples in binary polymer solutions. T ) 298 K.

wavelength of λ ) 560 nm. For calculating the aldehyde concentrations, calibration plots were used. Due to the weak solubility of the propionaldehyde and butyraldehyde in water, the propional and butiral copolymers have been prepared in an alcohol/water mixture. In the alcoholic medium the sensitivity of the aldehyde determination has also enlarged. In an alcohol/water mixture of 45% (v/v), aldehyde concentration as low as 10-5 mol/dm3 could be accurately determined. (A 20-fold increase in this mixture could be achieved compared to that obtained in pure water.) The following acetal copolymers were prepared and used: poly(vinyl alcohol/7.0 mol % vinyl acetal) copolymer (PVAL/7.0); poly(vinyl alcohol/9.5 mol % vinyl acetal) copolymer (PVAL/9.5); poly(vinyl alcohol/13.9 mol % vinyl acetal) copolymer (PVAL/ 13.9); poly(vinyl alcohol/22.6 mol % vinyl acetal) copolymer (PVAL/22.6); poly(vinyl alcohol/7.0 mol % vinyl propional) copolymer (PVPr/7.0); poly(vinyl alcohol/9.5 mol % vinyl propional) copolymer (PVPr/9.5); poly(vinyl alcohol/7.0 mol % vinyl butyral) copolymer (PVBu/7.0). The time dependence of acetal formation showed that the time taken to reach equilibrium was approximately 3 days, and the highest conversion was 93-97%. Figure 1 illustrates a typical kinetic curve for a poly(vinyl alcohol-co-vinyl acetal) copolymer with 10 mol % acetal content. In the figure, the relative amount (mol %) of the residual acetaldehyde is plotted as a function of the conversion time (h). (The kinetic curve represents experimental data of four independent measurements.) Solvent Activity Measurements. Gel deswelling proved to be a useful technique for thermodynamic investigations in binary polymer solutions at wide concentration range of the macromolecular solutes.15-17 We have successfully applied this method for thermodynamic studies of mixtures of homopolymers.8 To characterize the thermodynamic properties of binary and ternary solutions of the uncharged homopolymers and copolymers used in the present investigations, the thermodynamic activity of solvent (a1) has also been determined by a gel-deswelling method. For the measurements of solvent activity, the polymer samples were intensively dialyzed. The solvent activities in the range of polymer volume fractions of 1 × 10-2 < v2 < 1.0 × 10-1 were measured at 298 ( 0.1 K. Poly(vinyl alcohol-co-vinyl acetate) gels with 8 mol % acetate content were used for these investigations. Before activity measurements the gels, placed in acetylated (Visking) cellulose membrane (Medicell Int. England), were calibrated by poly(vinyl pyrrolidone) solutions of known water activity.18 The detectable limit with adequate accuracy of solvent activity was ln a1 ∼ 1 × 10-6, and the lowest measurable value for ln a1 was ∼2 × 10-6. The time taken to reach equilibrium in these measurements was ∼10 days.

Table 1. The Number Average Molecular Mass of the Polymers

(15) Nagy, M.; Horkay, F. Acta Chim. Acad. Sci. Hung. 1980, 104, 49-61. (16) Nagy, M. Phys Chem Chem Phys. 2000, 2, 2613-2622. (17) Boyer, R. F. J. Chem. Phys., 1945, 13 (9), 363-372. (18) Vink, H. Eur. Polym. J. 1971, 7, 1411-1419.

Mn PVA PVAL/7.0 PVAL/9.5 PVAL/22.6

74 070 75 600 76 150 79 020

Mn PVPr/7.0 PVPr/9.5 PVBu/7.0 PVP

76 430 77 270 77 260 117 000

The number average molecular mass of the polymers used in these investigations is given in Table 1.

3. Results and Discussion Individual Polymer Solutions. To characterize the thermodynamic properties of the solutions of individual polymers, the thermodynamic activity of the solvent has been determined at a wide range of polymer volume fractions (1 × 10-2 < v2 < 1.0 × 10-1), and the relevant Flory-Huggins pair interaction parameters (χ12, χ13) have been calculated. Typical solvent activity curves for poly(vinyl alcohol/ 7.0 mol % vinyl acetal), poly(vinyl alcohol/7.0 mol % vinyl propional), and poly(vinyl alcohol/7.0 mol % vinyl butiral) copolymers are shown in Figure 2. In the figure, the water activities in logarithmic scale (ln a1) are plotted as a function of the polymer volume fraction (v2). It can be seen that the dissolved polymers reduce the solvent chemical potential to different degrees depending on the chemical structure of the macromolecules. In other words, changing the chemical structure of the copolymers is well reflected in their solvent activities. Both Figures 2 and 3 illustrate the effect of side chains on the thermodynamic properties of the polymer solutions. Increasing either the length of the hydrophobic side chains or the acetal content of the copolymer results in less reduction in the activity of the solvent. In accordance with theoretical considerations, it means that the water becomes a poorer solvent for the more hydrophobic copolymers. The solvent activity vs polymer volume fraction curves can be well described by a third degree polynomial for each homopolymer and copolymer. From the water activity vs polymer volume fraction functions the relevant solventsegment pair interaction parameters (χ12) have been calculated for both the homopolymers and the copolymers by means of eq 1. Figures 4 and 5 illustrate the volume fraction dependence of χ12 parameters for two homopolymers and poly(vinyl alcohol-co-vinyl acetal) copolymers with various acetal content and chain structure, respectively. For each binary solution, the χ12-v2 function can

764

Langmuir, Vol. 21, No. 2, 2005

Figure 3. Water activity curves for PVA, PVPr/7.0, and PVPr/ 9.5 in binary polymer solutions. T ) 298 K.

Figure 4. Solvent-segment interaction parameters for the PVP, PVA, PVAL/7.0, PVAL/9.5, and PVAL/22.6 in aqueous binary solutions: T ) 298 K, sd ) 0.0015.

Figure 5. Solvent-segment interaction parameters for the PVAL/7.0, PVPr/7.0, and PVBu/7.0 in aqueous binary solutions: T ) 298 K, sd ) 0.0015.

be well fitted by a straight line of small positive slope. These results confirm that the segment-solvent pair interaction parameters are responsive indicators to changes in the chemical structure of the copolymers. With increase in either the acetal content or the length of the hydrophobic side chains in the copolymer, χ12 approaches the value characteristic of a theta condition.

Csa´ ki et al.

Figure 6. Water activity of the PVPr/7.0 and the PVP in binary solutions and of the 1:1(w/w) PVPr/7.0-PVP mixture in ternary solutions: T ) 298 K.

Figure 7. Segment-segment interaction parameters for the 1:1 (w/w) PVA-PVP, PVAL/7.0-PVP, PVAL/9.5-PVP, PVAL/ 22.6-PVP, PVPr/7.0-PVP, and PVBu/7.0-PVP mixtures in ternary solutions: T ) 298 K, sd ) 0.02.

Polymer Mixtures. Similar solvent activity curves were obtained for the aqueous mixtures of PVP with several PVA-“acetals” at 1:1 w/w ratio of the homopolymer and the copolymer, as those determined for the solutions of the individual homopolymers. (See Figure 6, where the water activity-polymer volume fraction curves for the binary solution of the PVP K-90 sample and of the PVPr/ 7.0 copolymer, as well as, for the ternary solutions of these two polymers {1:1 (w/w)} are shown together.) On the theoretical basis of the Flory-Huggins approximation for ternary systems (eq 3), the segmentsegment (χ23) interaction parameters were calculated for different polymer mixtures by using the appropriate χ12 and χ13 solvent-segment interaction parameters. The relevant segment-segment interaction parameters at various polymer volume fractions are shown for some polymer pairs in Figure 7. It was found that for most of the mixtures, the values of the χ23 segment-segment interaction parameters are close to zero or slightly negative in a wide concentration range. For compatible polymer pairs, these results are in accordance with theoretical considerations. It must be noted, however, that in the 1:1 (w/w) mixtures of the PVBu/ 7.0-PVP and the PVAL/22.6- PVP phase separation was observed even at low polymer concentrations. In Figure 7, arrows indicate the concentration of this incipient phase separation. No significant differences in the χ12 segment-

Thermodynamic Properties of Polymer Mixtures

Figure 8. Solvent-segment interaction parameters for the PVAL/7.0 (χ12) and the PVP ( χ13) in aqueous binary solutions and the χ1,23 interaction parameters for the 1:1 (w/w) PVAL/ 7.0-PVP mixture in ternary solutions: T ) 298 K.

Figure 9. Solvent-segment interaction parameters for the PVAL/9.5 (χ12) and the PVP (χ13) in aqueous binary solutions and the χ1,23 interaction parameters for the 1:1 (w/w) PVAL/ 9.5-PVP mixture in ternary solutions: T ) 298 K.

solvent interaction parameters obtained for these polymer pairs could be observed. It is worthwhile to point here that the accuracy of the determination of the χ23 parameters at low polymer concentrations considerably diminishes. With increase of the polymer concentration, the calculated χ23 parameters show a slight decrease. This contradiction is probably due to the fact that in the calculations for χ23 the χ12 and χ13 parameters determined in binary polymer solutions were used. For this reason, in dilute polymer mixtures these parameters can be adequate indicators of the solventsegment interactions. At higher concentrations, however, the solution of one polymer is presumably a “worse solvent” for the other polymer than the pure solvent. In more concentrated polymer mixtures, therefore, the used χ12 and χ13 parameters may really be regarded only as estimated characteristics of the individual solventsegment interactions. Also, this is probably one reason why no significant differences could be detected in the values of the χ23 parameters for the different mixtures of PVP and PVA-acetal copolymers. Some of the shortcomings owing to the approximation used in the calculations for χ23 can be eliminated by determining the χ1,23 interaction parameters for two polymer-component mixtures on the theoretical basis suggested by Panayiotou. In principle, the latter param-

Langmuir, Vol. 21, No. 2, 2005 765

Figure 10. Solvent-segment interaction parameters for the PVAL/22.6 (χ12) and the PVP (χ13) in aqueous binary solutions and the χ1,23 interaction parameters for the 1:1 (w/w) PVAL/ 22.6-PVP mixture in ternary solutions: T ) 298 K.

Figure 11. Solvent-segment interaction parameters for the PVPr/7.0 (χ12) and the PVP (χ13) in aqueous binary solutions and the χ1,23 interaction parameters for the 1:1 (w/w) PVPr/ 7.0-PVP mixture in ternary solutions: T ) 298 K.

eters do not provide information directly on the segmental interactions between chemically different polymers but χ1,23 may be related to χ23 (eq 5). Nevertheless, χ1,23 can be simply calculated from exact solvent activity data (eq 4), i.e., knowledge of χ12 and χ13 interaction parameters are not definitely required. It is quite evident from eq 5 that χ1,23 increases when either χ12 or χ13 increases, and it decreases when χ23 increases. A larger value of χ1,23 indicates a stronger attraction between the segments of chemically different polymers in the mixture. In Figures 8-10, the χ1,23 vs v2 functions for PVPPVAL/7.0, PVP-PVAL/9.5, and PVP-PVAL/22.6 mixtures are shown. The χ12 and χ13 parameters obtained for the corresponding polymer components and “hypothetic χ1,23 values” are also shown in the figures. The “hypothetic” values of χ1,23 parameters were calculated according to eq 5, assuming that there are no interactions between the chemically different polymer components in the mixture. For simulating this situation, χ23 was taken to be zero (χ23 ) 0). Figures 8-10, adequately demonstrate that increasing the acetal content of the copolymer results in a decrease in χ1,23 compared to its hypothetical value. The 1:1 (w/w)

766

Langmuir, Vol. 21, No. 2, 2005

Csa´ ki et al.

when the hydrophilic character of a copolymer is reduced by increasing the length of its side chains. In Figures 11 and 12, similar χ1,23 vs polymer volume fraction curves can be seen for the mixtures of PVPr/7 and PVBu/7.0 with PVP. These results demonstrate that for miscible polymer pairs at higher polymer concentrations the values of the χ1,23 parameter significantly increase. This effect can probably be attributed to the increase of the individual χ12 and χ13 parameters due to the presence of the second polymer in the mixture.

Figure 12. Solvent-segment interaction parameters for the PVBu/7.0 (χ12) and the PVP (χ13) in aqueous binary solutions and the χ1,23 interaction parameters for the 1:1 (w/w) PVBu/ 7.0-PVP mixture in ternary solutions: T ) 298 K.

mixture of PVP and PVAL/22.6 seems to correspond to the conditions where χ23 is nearly zero, i.e., this mixture exhibits the “hypothetic χ1,23” (Figure 10). According to eq 5 three quantities contribute to the value of this complex interaction parameter. In principle, either χ12 and χ13 may decrease in a mixture compared with their values in individual (binary) polymer solutions or χ23 may become larger when the acetal content of the copolymer increases. The effect that one polymer enlarges the solubility of another one without promoting attractive interaction between them can be probably excluded. Hence it is very unlikely that the values of the solvent-segment parameters (χ12 or χ13) considerably diminish in the mixture. The incipient phase separation occurring in homopolymer/copolymer/water ternary mixtures with copolymers of adequately high acetal content (e. g. PVAL/ 22.6-PVP-water mixture) confirms these premises. Hereupon, it can be concluded from these results that with increase of the acetal content of the copolymer, the attraction between the copolymer and PVP in their mixture becomes weaker. Similar conditions may arise

4. Conclusions The relevant segment-solvent pair interaction parameters (χ12) have been determined for homopolymers (PVP, PVA) and copolymers (PVA-“acetals”) from the water activity vs polymer volume fraction functions. For each binary solution, the χ12-v2 function can be described by a straight line with a small positive slope. It was shown that changes in the chemical structure of the copolymers are well reflected in the value of the segment-solvent pair interaction parameters. Increasing either the acetal content or the length of side chains in the copolymer, χ12 approaches the value characteristic of theta condition. On the theoretical basis of the Flory-Huggins approximation for ternary systems, the segment-segment (χ23) interaction parameters have been calculated for different polymer mixtures. It was found that for compatible polymer pairs, the values of the χ23 segment-segment interaction parameters are close to zero or slightly negative in a wide concentration range. No significant differences could be revealed in the χ23 segment-segment interaction parameters obtained for the PVP-copolymer mixtures with various acetal derivatives. However the χ1,23 parameters calculated for these mixtures show that increasing the acetal content or the length of the hydrophobic side chains in the copolymer results in a reduction in the interaction between the “acetal” copolymers and the PVP molecules. Acknowledgment. The Hungarian Science Foundation under Grant OTKA T034929 is gratefully acknowledged. LA047827H