J . Phys. Chem. 1988, 92, 4172-4771
4112
TABLE IV: R E / I Calculated from p E / I = w2 + '/@,where @ and w Are Fitted Parameters (See Table- 11)
amlied electric field, kV/cm
UEII
10.00 11.25 12.50 13.75 15.0
11 814 18 148 26 506 38 075 53 628
The increase in w with temperature (Table I) is in qualitative agreement with the decrease in the aggregate effective length with temperature, pointed out by Fonseca and Barbosa.z As absolute measurements were not done for the constants in eq 35-38, we were not able to obtain the microscopic related quantities p , I , and$ To have these parameters, we must obtain those constants, preferably from separated wax components and under dilution in nonpolar medium. While no definite conclusions were drawn from these results with respect to the main chemical constituents responsible for the polar behavior of the wax, they are quite convincing about the existence of large polar structures or aggregates in it. Possibly, the aggregation mechanism, which is not yet understood, resembles
that of formation of micelles in a polar-nonpolar liquid mixture. Although the chemical components involved were not clearly identified, independent measurements such as optical disymmetry could help to provide direct data on the aggregates sizes and shapes. The study of the wax principal fractions might be particularly elucidating with respect to the polarization mechanism of the wax in the liquid phase. The possibility of a separation of these polar structures is quite appealing, as technical applications for a strengthened electret would certainly appear. The basic raw material, being extracted from a natural Brazilian palm quite abundant in the northeast of Brazil, is practically inexhaustible if not misguided by human actions. Acknowledgment. We thank Dr. Oscar N. Mesquita for some discussions. This research was partially supported by the Brazilian agencies FINEP (Financiadora de Estudos Projetos), CNPq (Conselho Nacional de Desenvolvimento Cientifico e TecnolBgico), an! CAPES (Coordena@o de Aperfeiqoamento de Pessoal de Nivel Superior). This work has been carried out by C.H.M. in partial fulfillment of the requirements for a Mestre em Fisica degree.
e
Electrical Conductivity Behavior of Poly(ethy1ene oxide) in Aqueous Electrolyte Solutions F. Bordi, C. Cametti,* Dipartimento di Fisica, Universitd di Roma "La Sapienza", Rome, Italy, and "Gruppo Nazionale di Struttura della Materia", Rome, Italy
and A. Di Biasio Dipartimento di Matematica e Fisica, Universitd di Camerino, Camerino. Italy, and "Gruppo Nazionale di Struttura della Materia", Rome, Italy (Received: October 23, 1987)
The electrical conductivity of poly(ethy1ene oxide) of different molecular weights in 0.1 M KCl electrolyte solution has been measured in the temperature interval from 0 to 55 "C. The data have been interpreted on the basis of the Looyenga equation for heterogeneous systems, found to apply with good approximation also for other transport properties. The hydration number deduced from these measurements has been estimated to be lower than that reported elsewhere. Its dependence on temperature, however, supports the occurrence of a cloud point temperature which agrees with the values observed in similar systems. These findings are consistent with the measured activation enthalpy values, which are very similar to that of the pure solvent, indicating that poly(ethy1ene oxide)-water interactions give rise to a structure that would form a rather open hydrogen-bonded network.
Introduction Poly(ethy1ene oxide) (PEO), an uncharged hydrophilic polymer, is a very interesting model system to study the role of water in polymer solutions owing to its simple molecular structure, the absence of any charged groups, and its complete miscibility with water in all proportions for a very wide range of degrees of polymerization. This unusual property is probably due to the fact that PEO chains may be modeled in a hexagonal water structure' with hydrogen bonds between water and polar groups of the polymer, forming highly linked networks. This behavior is of great significance in terms of an enhanced water structureZaround the chain to further elucidate questions (1) Blandamer, M. J.; Fox, M. F.;Powell, E.; Stafford, J. W. Mukromol. Chem. 1969, 124, 222.
0022-3654/88/2092-4772$01 .50/0
involved in specific dipolar or hydrogen bond interactions, particularly with regard to various transport properties. In aqueous solutions, the polymer influences through hydrophilic and hydrophobic interactions the structure of the solvent, water, reflected in changes of the electrical conductivity. This effect can be used to study the polymer-water interactions involving changes in the polymer hydration or change in the water structure. Recently, different transport properties of PEO in water have been reported, including solvent ~ e l f - d i f f u s i o n thermal , ~ ~ ~ cond ~ c t i v i t yelectrical ,~ conductivity at audio frequencies,' and di(2) Kjellander, R.; Florin, E. J . Chem. Soc., Furuduy Trans. 1 1981, 77, 2053. (3) Foster, K. R.; Cheever, E.; Leonard, J. B.; Blum, F. D. Biophys. J . 1984, 45, 915. (4) Blum, F. D.; Pickup, S.; Foster, K. R. J . Colloid Interface Sci. 1986, 113, 336.
0 1988 American Chemical Society
Electrical Conductivity Behavior of PEO
The Journal of Physical Chemistry, Vol. 92, No. 16, 1988 4773
0 a1 0.2 0.3 0.4 0.5 9 Figure 1. Conductivity of PEO in 0.1 M KCI electrolyte solution as a function of the fractional volume @ for various molecular weights. The temperature is T = 25.0 f 0.1 OC. This work: ( 0 )M , = 6 X lo3;(m) M, = 15 X 10); (A)M , = 2 X los; (v)M , = 5 X lo6. From ref 3: (0)M,.= 200; (0) M , = 400; (A) M , = 600; (v)M , = 3400; (0) M , = 14000. The full and dotted lines represent the calculated values according to eq 1-3 for the = 1.2@ (see text). particular case up = 0: (a) eq 1 with f = 2; (b) eq 2; (c) eq 3; (d) eq 3 with an excluded volume electric relaxation measurements of dilute solutions with aqueous5 and nonaqueous6 solvents. Deviations from the predicted values from the usual heterogeneous mixture equations have been generally reported and interpreted as caused by a reduction in the translational or orientational motion of the solvent due to interactions with the polymer, although, as pointed out by Blum et a1.: no chemically specific effects from the above measurements can be resolved. In the present work, the electrical conductivity of poly(ethy1ene oxide)-water-KC1 systems has been measured over a wide range of the polymer molecular weight, from 6 X lo3 to 5 X lo6 daltons, at various temperatures, from 0 to 55 OC. This study extends toward highest molecular weights the results of previous investigations3 providing further support to structural models involving specific spatial interactions in the PEO-water coupling. The observed values of the activation enthalpy indicate a slight dependence on the polymer concentration, suggesting appreciable cooperative effects with an increasing solvent organization between polymer chains. The amount of the hydration water associated with the polymer is estimated from a simple mixture theory of heterogeneous systems proposed by Looyenga, yielding hydration values considerably lower than those obtained by Foster et al.,I3 who used the Maxwell and Hanai mixture equations. On the other hand, some discrepancies found by these authors between different transport properties have been removed. The hydration number estimated from conductivity measurements differs from that obtained from thermodynamic measurements' corresponding to two water molecules per ethylene oxide group. It is noteworthy, however, that the dependence of hydration on temperature determined from the Looyenga equation gives a cloud-point temperature very close to that experimentally determined by visual observation for the same system. This fact indicates that the hydration shell around the PEO chains exerts a marked influence on the phase-separation behavior, ( 5 ) (a) Kaatze, U. Ber. Bunsen-Ges, Phys. Chem. 1978, 82,690. Kaatze, U. Prog. Colloid Polymn. Sci. 1978,65, 214. (b) Kaatze, U.; Gbttmann, 0.; Podbielski, R.; Pottel, R.; Terveer, U. J . Phys. Chem. 1978, 82, 112. ( 6 ) Mashimo, S.; Yagihara, S. Macromolecules 1984, 17, 630. (7) Molyneux, P. In Water, a Comprehensive Treatise; Franks, F., Ed.; Plenum: New York, 1975; Vol. 4, p 617.
supporting the hypothesis that PEO in aqueous solution is surrounded by an extended region with enhanced water structure.
Experimental Section Two samples of poly(ethy1ene oxide) (PEO), viscosity-average molecular weights of which are 2 X lo5 and 5 X lo6, respectively, were provided by Aldrich Chemie, and were used without further purification. Samples of PEO with molecular weights of 6 X lo3 and 15 X lo3 were obtained courtesy of Dr. G. Paradossi, University of Naples, Italy. The samples were prepared by dissolving an appropriate amount of polymer in 0.1 M KCI electrolyte solution. All the samples were then kept standing at 20 OC for several days to ensure homogeneity, before the first set of conductivity measurements was made. The fractional volume of the polymer was varied from 0.01 to 0.5, depending on the molecular weight. The weight concentrations were converted to volume concentrations by assuming a partial specific volume of 0.885 ~ m ~ / g . ~ Electrical conductivity measurements were performed by using an impedance analyzer, Hewlett-Packard Model 41 92A, in the frequency range 1 kHz to 10 MHz at the temperature of 25.0 f 0.1 OC. The conductivity cell, consisting of a cylindrical guide excited far beyond its cutoff frequency, has been calibrated with standard liquids of known conductivity and dielectric constant, according to the procedure proposed by B ~ t t o m l e y . ~ The measured conductivity was approximately constant in the whole frequency range, ruling out artifacts due to electrode polarization. High molecular weights are known to be sensitive to degradation,1° even if the reaction mechanism is still not known precisely. However, at the ionic strength employed and with minimization of the time in the heating and cooling cycles, reproducible results were obtained over long period of time, ensuring that degradation does not take place. To obtain a better picture of the water structure effects on the PEO-water systems, we carried out the conductivity measurements (8) Galin, M.; Mathis, A. Macromolecules 1981, 14, 677. (9) Bottomley, P. A. J . Phys. E Sci. Instrum. 1978, 1 1 , 413. (10) Bailey, F. E., Jr.; Koleske, J. V. Poly(ethy1ene oxide) Academic: New York, 1976.
4714
The Journal of Physical Chemistry, Vol. 92, No. 16, 1988
Bordi et al.
of samples of higher molecular weights in the temperature range 0 to 55 “C within 0.1 OC. The maximum temperature investigated was maintained well below the cloud-point temperature,” at which the sample starts to get cloudy and macroscopic phase separation occurs.
similar to that introduced by Bruggeman. Finally a more general equation, from which the above two formulas can be derived with appropriate approximations, has been proposed by L00yenga.l~ This equation, written in terms of conductivity, reads
Results and Discussion The electrical conductivity u/u,, normalized to that of aqueous phase u, of the PEO solutions of different molecular weights as a function of the fractional volume CP is shown in Figure 1. The values for samples of the smallest molecular weights, from 0.2 X lo3 to 14 X lo3, are taken from ref 3 and shown in Figure 1 for a complete vision. As can be seen, all normalized conductivity values fall approximately on the same curve, providing no evidence for a possible dependence of conductivity on the molecular weight of the polymer. The slight dependence of the conductivity on the molecular weight observed by Foster et aL3,and interpreted as due to changes in the polymer flexibility is confined to very small molecular weight polymers, below approximately lo3, where end effects are probably relevant. In fact, the contribution of end-group effects disappears as the polymer chain increases and the polymer itself takes up a more complex configuration. For example, no molecular weight dependence was observed by Breen et a1.’* in 4H nuclear magnetic relaxation rates for PEO aqueous solutions with polymerization degree larger than 100. The flexibility of the chain is also suggested by the short relaxation time (about lo-” s at T = 20 “ C ) derived by Davies et al.I3 from dielectric measurements of PEO of various molecular weights in benzene solutions. This indication is also strengthened by the low activation enthalpy (2.5 kcal/mol) observed for these systems. For large polymers as those investigated in this work, the absence of any dependence on the molecular weight, in addition to the marked flexibility of the PEO chain (compare a value of the persistence length of about 0.5-0.6 nm (ref 12 and references quoted therein) with an average bond length of 0.146 nm), suggests that the polymer in aqueous solution assumes a globular configuration. The conductivity measurements have been analyzed on the basis of appropriate heterogeneous system equations relating the bulk properties of the suspended and aqueous phases to those of the whole solution. Three different mixture equations will be considered. The most widely used conductivity mixture equation (Maxwell-Wagner equation) is given byI4
(3)
where um and up are the conductivity of the aqueous phase and the polymer, respectively, and CP is the fractional volume of the dry polymer. The configuration of the polymer in solution is taken into account by the shape factorfranging from 2 for spherical particles to 1.5 for randomly oriented rodlike particles. For concentrated suspensions of spherical particles, BruggemanlS derived the following experession:
(-)(
= 1 - cf,
This equation also results from a theory of the interfacial polarization proposed by HanaiI6 based on an integration method (1 1) Florin, E.; Kjellander, R.; Eriksson, J. C. J . Chem. SOC.,Faraday Trans. 1984, 80, 2889. (12) Breen, J.; van Duijn, D.; de Bleijsen, J.; Leyte, J. C. Ber. Bunsen-Ges Phys. Chem. 1986, 90, 1 1 12. (13) Davies, M.; Williams, G.; Loveluck, G. D. Z . Elekrrochem. 1960,64, 575.
(14) Fricke, H. Phys. Reu. 1924, 24, 575. (15) Bruggemann, D. A. G. Ann. Phys. 1935, 24, 636.
As pointed out by this author, because the derivation of this expression does not involve the shape of the dispersed phase, it is in principle applicable to any kind of particles, especially when the “internal conductivity” of the inclusions is not exactly defined. A comparison of the conductivity values calculated from eq 1-3 as a function of the fractional volume with experimental results is shown in Figure 1 in the particular case up = 0. As can be seen, a very good agreement over the whole concentration range is obtained by using the Looyenga equation (curve c) in comparison with the Maxwell (curve a) and Hanai (curve b) equations. These results make it possible to attempt an adequate interpretation of the conductivity data on the basis of eq 3. The apparent hydration of the PEO molecules deduced from conductivity measurements can be evaluated by considering in eq 3 an excluded volume CPerf = a@that includes the effective fractional volume CP of the dry polymer and the volume of the hydration water. From a nonlinear regression fit of the experimental data to eq 3, for all the solutions investigated, the parameter a is approximately independent of molecular weight and takes the average value of 1.2, which corresponds to about 0.6 water molecules per ethylene oxide group. This result must be compared with the apparent hydration reported by various authors3p4deduced from the transport measurements and with that obtained from thermodynamic measurement~.~ Foster et al.3 have found from electric conductivity and from water self-diffusion coefficient measurements an excluded volume of 2.5 or 2.3 times that of the dry polymer, employing the Maxwell or Hanai equation, respectively. These values correspond to 3.3 and 2.8 water molecules per ethylene oxide group. Moreover, Kaatze et aL5 obtained a hydration number of 5.4 from dielectric relaxation measurements in a wide frequency range from 1 to 35 GHz. From permittivity data at high frequencies, assuming a real value of 4 for the dielectric constant of the hydration water, Foster et aL3 found hydration values of 1.8 and 1.6 according to eq 1 or 2, respectively. On the other hand, from the capability of each ethylene group to form two hydrogen bonds, a hydration number of 2 is expected.18 Our results do indicate hydration water content somewhat lower than the above results, but it is not clear at present how much of this discrepancy results from the lack of an appropriate theory or from peculiar motional properties of hydration water. It must be noted, however, that hydration depends considerably on the experimental technique used to probe the system, beside the theoretical model employed. However, the use of the Looyenga equation instead of the Maxwell or Hanai equations seems also supported by the following considerations. Foster et aL3 have pointed out that some transport properties of PEO-water systems such as water self-diffusion and ionic conductivity deviate markedly from the expected values of the Maxwell or Hanai equations, and, on these deviations, these authors have evaluated the hydration properties of this polymer, while thermal conductivity agrees well enough with the calculated values, using the known transport properties of pure liquid water. A possible mechanism based on changes in translational and orientational motion of solvent has been proposed. On the contrary, the different behavior of the water self-diffusion coefficient and ionic conductivity on the one side and (16) Hanai, T. Kolloid Z.1960, 171, 23. (17) Looyenga, H . Physica, 1965, 31, 401. (18) Maxfield, J.; Shepherd, J. W. Polymer 1975, 16, 505
Electrical Conductivity Behavior of PEO
The Journal of Physical Chemistry, Vol. 92, No. 16, 1988 4775
Figure 2. Normalized water self-diffusion coefficient of PEO solutions as a function of the fractional volume @ for different polymer molecular weights: (0)M, = 200; (0) M, = 400;(A)M, = 600;(V)M, = 3400; ( 0 )M, = 14000 (data from ref 3, table I). The full lines represent the calculated = 1.29. values according to eq 1-3 written for the self-diffusion coefficient: (a) eq 1 withf= 2; (b) eq 2; (c) eq 3; (d) eq 3 with
M
0.9
0.4
I
I
8
&I
I
I
I
I
I
I
02
w3
w
0.5
08
0.7
I
I
0.9
o8 9 Figure 3. Thermal conductivity, normalized to that of pure solvent, of PEO solutions as a function of the fractional volume @ for different polymer
molecular weights; symbols are as in Figure 2 (data from ref. 3, figure 1). The full lines represent the calculated values according to eq 1-3 written = 1.2@. The thermal conductivity of the polymer is taken as 0.31 for thermal conductivity: (a) eq 1 withy= 2; (b) eq 2; (c) eq 3; (d) eq 3 with times that of water. thermal conductivity on the other side is due to the choice of the mixture equation employed to describe the transport properties studied in that work. In fact the Looyenga equation is able to give a unique description of all these properties. To strengthen this statement, we have reexamined the data reported by Foster et al.3 on the basis of this equation. Figures 2 and 3 show the normalized water self-diffusion coefficient and the thermal conductivity (data taken from ref 3) compared with the values calculated from eq 1-3.
As can be seen, a very good agreement is obtained for all these transport properties with a unique value of the excluded volume %ff.
The low amount of hydration water deduced from conductivity measurements might be related to the activation enthalpy of this process, which should be very similar to that of the simple electrolyte solution. Accordingly, we decided to make a more comprehensive study of the conductivity properties over a range of temperatures from 0 to 55 O C so as to confirm the above assumptions.
(I
1
m-']
[Q-l
I
30 -
0.21 I
3x)
320
3.30
3.40
3 60
3.50
T-'x 103[K-'] Figure 4. Arrhenius plot of conductivity as a function of inverse of temperature for PEO-water systems of various molecular weights, at two polymer contents. Open symbols, molecular weight 2 X lo5: (0)4.0% (wt/wt); (0) 7.5% (wt/wt); (A) 10% (wt/wt); (v)15% (wt/wt); ( 0 )20% (wt/wt). Full symbols, molecular weight 5 X lo6: (0) 4.0% (wt/wt); (D) 7.5% (wt/wt); (A) 10%(wt/wt). The conductivity of 0.1 M KCI electrolyte solution
(asterisk) is also shown for comparison. TABLE I: Activation Enthalpy of PEO-Water Systems at Various Concentrations Accordine to Ea 4' H , kcal/mol concn, 9% (wtlwt) ~,=2x105 ~ , = 5 x i 0 6 4.0 3.38 f 0.18 3.39 f 0.18
3.55 f 0.18 3.68 f 0.20 4.01 f 0.20 4.35 f 0.23
7.5 10 15
20
3.56 f 0.20 3.67 f 0.18
"The aqueous solution 0.1 M KC1 yields an activation enthalpy of 3.17 f 0.22 kcal/mol. The quoted uncertainties represent a 95% confidence interval. 3.0
1
0
0 10
0.05
035
0.20
cp
Figure 5. Enthalpy of activation AH as a function of the fractional volume 0 for PEO of different molecular weights in 0.1 M KC1 electrolyte solutions: ( 0 )M , = 5 X lo6; (0) M , = 2 X lo5. The value corresponding to the aqueous phase (D) is also reported.
The electrical conductivity of PEO-water solutions increases exponentially with temperature according to the usual relation:I9 u(T)
=
uo exp(-AH/RT]
(4)
S is the activation enthalpy for the process, R is the gas constant, and uo is the extrapolated conductivity at infinite temperature. When the logarithm of the conductivity was plotted against 1/T, straight lines were obtained. The natural logarithm of the conductivity versus the reciprocal temperature for various solutions with different PEO contents a t two different molecular weights is shown in Figure 4. The upper solid line represents the conductivity behavior of 0.1 M KCl aqueous solution. The dependence of the activation enthalpy as a function of fractional volume of the polymer is shown in Figure 5 . As can be seen,the slope of the various curves are approximately the same, indicating that the major contribution to the electrical conductivity is originated from the aqueous phase, whereas the polymer contribution is confined to a little correction to this value. (19) Nandy, P.; Data, R.; 115, 277.
In fact, in the limit of our experimental accuracy, some statistically significant differences can be appreciated. Linear regression analysis of the experimental data for PEO samples of M , = 2 X lo5 and 5 X lo6 according to eq 4,leads to activation enthalpy values listed in Table I. A test of hypothesis on regression coefficients based on the t-test was employedZoto ascertain if observed differences might be significant. It consists in comparing each slope b, with each other b, by using a variable: t = (b" - bIll)/(Sb,2 + Sb,2)
which follows a t-distribution function with 2 ( N - 2 ) degrees of freedom. Here, the quantity Sbzis defined as
SbZ
1
=
m
(m
- 2 ) x k ( X k - x)' 1
The other symbols have their usual meanings. The confidence interval estimated for the difference of the regression coefficients indicates that these parameters are obtained from two different populations, and, therefore, it must be concluded that the activation enthalpy depends on the polymer concentration. Since the differences in the regression, although significant, are small, to provide further evidence to this analysis, a F-test of whether one regression coefficient might be used for all the observations has been performed. Also in this case, our
Bhowmik, B. B. J. Colloid Interface Sci. 1987, (20) Ostle, B. Statistic in Research; Iowa State College Press: Iowa, 1954.
The Journal of Physical Chemistry, Vol. 92, No. 16, 1988 4777
Electrical Conductivity Behavior of PEO
m
10
0
Figure 6. Ratio a of the apparent volume
30
40
5m
6.0
70
80
TPC]
to the effective volume 3 of the dry polymer as a function of temperature, deduced from the Looyenga
equation. TABLE II: Apparent Hydration of PEO ( M , = 2 X lo5) Measured by the Parameter a in Eq 5 as a Function of Temperature and the Number N of Water Molecules w r Ethvlene Oxide GrouD"
T, O C 0 5 10 15 20
(Y
=
1.480 f 0.005 1.427 f 0.005 1.370 f 0.005 1.330 f 0.005 1.282 f 0.005
N
T, OC
1.04 0.93 0.80 0.72 0.61
25 30 35 40 45
LY
@.a/@
1.254 f 0.005 1.218 f 0.004 1.182 f 0.005 1.148 f 0.005 1.123 f 0.004
N 0.55 0.47 0.39 0.32 0.27
The quoted uncertainties represent the standard error. hypotheses are confirmed with a confidence level higher than 0.99. On the contrary, no dependence on the molecular weight appears. This may reflect a structural change in the water at the polymer-bulk medium interface, whose percentage relevance depends on the polymer concentration. Nevertheless, the activation enthalpies result is very similar to that of the pure solvent, making it evident that the polymer-water interactions influence a moderate fraction of the total water. The dependence of the excluded volume and hence the change of the hydration on temperature was obtained by fitting for each temperature the experimental data to the equation ./a,
= (1 - a@)3
(5)
which is the Looyenga equation in the limit up = 0. The values of the parameter a in the case of PEO with molecular weight 2 X lo5 are reported in Table 11. The hydration number, calculated from stoichiometric relationships, is also shown. As can be seen, the hydration number decreases with increasing temperature, indicating a reduction in the polymer-solvent interactions. The quantity a takes into account the polymer hydration. When the hydration reduces to zero, a phase separation should occur.
To complete this analysis and to provide further evidence in using the Looyenga equation, we have calculated the extrapolation of the curve a versus T toward temperatures higher than those investigated by fitting the data to a quadratic form in T . This procedure yields a temperature of 85 OC. Figure 6 shows the values of a as a function of T and the extrapolation to a = 1 . In the limit of the above assumption this valve agrees surprisingly well with the cloud-point temperature of 91.92 OC reported by Florin et al." for PEO of molecular weight of 4 X lo6 in 0.1 M KC1 electrolyte solution. It must be noted that the dependence of the cloud-point temperature on the molecular weight has been proved to be small" and its dependence on polymer concentration2' for concentration higher than 1 wt % is contained within 2-4%. Consequently this value can be properly assumed as reliable for the sample investigated. This excellent agreement gives further support to use of the Looyenga equation. Addition of an electrolyte to an aqueous solution of nonionic polymer such as PEO polymer does not cause marked changes in the polymer conformation. As pointed out by Breen et al.,'* on the basis of nuclear magnetic relaxation measurements, ion-PEO complexes are highly improbable and the ion effects are confined to a change in the water structure around the polymer segments, resulting in a decrease of the cloud-point temperature.22 Since different ions affect differently the structure of water, one would expect that their influence on the hydration shell of PEO is probably different. More systematic studies are needed to give a general description of the effects of addition of simple electrolytes on aqueous nonionic solutions. This work is in progress. Registry No. PEO, 25322-68-3; KC1, 7447-40-7. (21) Ataman, M. Colloids Polym. Sci. 1987, 265, 19.
(22)
Ataman, M.; Boucher, E. A. J . Polym. Sci. Polym. Phys. Ed. 1982,
20, 1585.
