Oct., 1958
METAL-POLYELFCTROLYTE COMPLEXES
there may be a charge interaction. This is not unexpected a t a pH so far from the isoelectric point. It \vas cited in the Results section that as rn, is increased above 40-50 pg. up to 100, af increases rapidly in such a fashion that cc remains fairly constant. This suggests that 40 or 50 pg. completely fills the mosaic and possibly a duplex layer is started which has a greatly decreased effect (ca. 5-fold) on P. However, if there were a sharp transition /3 should change by a corresponding factor. This is not observed even up to 150 fig. This idea of trypsin being increasingly excluded or forced from the mosaic upon increasing cs is consistent with the behavior of plots of the data as l a A us. P (rather than as In c us. P ) : above about 18 d./cm. (Ar 9 60 cm.2) all the F-C films show not only a large increase in compressibility but also equal pressures at equal areas independent of cs suggesting that posiibly material is continually farced out of the films. The relative sizes of measured and calculated values of &/af are consistent with a partition of the two types of trypsin between mosaic and the underlying phase. The partition coefficients would depend upon P22,T and the com-
1239
position of the support solution. d. Molecule? with Many Degrees of Unfolding.8-Many molecular species each with a different degree of unfolding co-exist in the plane of the monolayer. The exact distribution of molecular species will depend upon how the film is formed and manipulated. The compressibility of a given film will be a weighted average of the compressibilities of the various species of molecules. As surface concentration is increased the proportion of completely unfolded molecules decreases. Therefore, unless all molecular species have identical compressibilities not only should the a’s for F-A and F-C films be different but they should vary with the concentration of the F-C films. This model is similar to the mosaic; however, it is such more difficult to evaluate because it was less precisely defined and has an undetermined number of parameters. Nevertheless the same types of discussion, restrictions and alternatives should apply t o this model as to the mosaic. Acknowledgment.-The authors thank Dr. Henry Quastler for his helpful suggestions. (22) H. Bull, Methods in Medical Research, 3, 245-255 (1950).
METAL-POLYELECTROLYTE COMPLEXES. V. PREPARATION AND PROPERTIES OF A NEW POLYCHELATE-POLYVINYLACETONYL KETONE BY GUENTHER K. HOESCHELE,~ JULIAN B. ANDELMAN AND HARRY P. GREGOR Contribution from the Department of chemistry of the Polytechnic Institute of Brooklyn, New York Received March 8. lQ.58
Polyvinylacetonyl ketone was prepared by ionic poIymerization of vinyl methyl ketone and its simultaneous condensation with acetic anhydride. The acid dissociation constant and binding constants for metallic ions were determined in 4/1 dioxane-water solutions for the polymer and its monomeric analog, acetylacetone. A modified Bjerrum technique was used to calculate chelation constants. On the basis of the displacement of a roton from the chelate acid by the metal, one finds that the polymer is equivalent to its monomeric analog for Cu(II), a n f i s three orders of ma nitude stronger than its analog with UOz(I1). The polymer also demonstrated a slight preference for Nd(II1) over Pr(I1f).
Previous papers in this series2described the complexing properties of polyacrylic and polymethacrylic acids, the monomeric units of which form relatively weak complexes. This contribution describes the preparation and properties of a new polychelate, polyvinylacetonyl ketone (PAK) ; this material is the polymeric analog of acetylacetone (AA), a strong chelating agent, and one whose properties have been studied extensively and are summarized in several b0oks.~-5 Experimental
methyl vinyl ketone (10 g., 0.14 mole) was mixed with 51 g. of acetic anhydride and cooled to -70”. Dry boron trifluoride was passed slowly through the solution for 2 hours; 32.5 g. of the gas was absorbed. The solution became viscous and deep red, indicating that polymerization had taken place. It was then warmed to -15” and boron trifluoride bubbled through again for 2 hours, an additional 2.5 g. being absorbed. At this point the reaction mixture was added dropwise to a vigorously stirred solution of 50 g. of sodium acetate in a liter of water, The red-brown precipitate was filtered, washed, dried, then. redissolved in acetone and reprecipitated in water. The yield was 13 g. or 81% of the theoretical. Anal. Calcd. for PAK, CeH802 (112.1): C, 64.27;. H , Preparation of Polymer.-The polymer was prepared by 7.19. Found: C, 64.07; H, 6.91. Calcd. for polyvinyl(70.08): ~. . , C,. 68.56: . H,. 8.63. the ionic polymerization of vinyl methyl ketone and its methyl ketone,. CtHeO The polyvinylacetonyl ketone was readily soluble in acesimultaneous condensation with acetic anhydride. The ionic polymerization of vinyl methyl ketone was described tone, methyl acetate, dioxane, chloroform, pyridine, diby Schildknecht, Zoss and Grosser .6 Freshly distilled methylformamide and tetrahydrofurane. It was insoluble in ethanol and carbon disulfide, but soluble in a 0.002 M (1) Jackson Laboratory, E. I. du Pont de Nemours and Ca., Inc., basic solution of dioxane-water provided that the ratio Wilmington, Del. of the latter was greater than 1/1. (2) H. P. Gregor, with L. B. Luttinger and E. M. Loehl, THIS The above reaction can result in two end products
JOURNAL, 69, 34, 366, 559,
990 (1955).
(3) F. Hein, “Chemische Koordinations Lehre,” S. Hirael, Leipzig, 1950. (4) A. E. Martell and
M. Calvin, “Chemistry of the Metal Chelate
Compounds,” Prentice-Hall, Inc., New York, N . Y . , 1952. (5) J. C. Bailar, editor, “The Chemistry of t h e Codrdination Compounds,” Reinhold Puhl. Corp., New York, N. Y., 1956.
CHz=CH-
8
-CH3
+ (CH3-C0)20
BE -+
(13) C. E. Schildknecht, A. 0. Z O S S and F. Grosaer, Ind. 41, 2891 (1949).
Eng< Chew.,
G. K. HOESCHELE, J. B. ANDELMAN AND H. P. GREGOR
1240
1
r-CHa-CH-
r
Forms I and I1 could exist in the same chain, along with the methyl ketone unit. This reaction is analogous to one described by Hauser and Adams’ who treated methyl isopropyl ketone, the monomeric analog of polyvinylmethyl ketone, with acetic anhydride in the presence of boron trifluoride with the result
L bH- 3- (CHa-C0)20 I
CHa-
8
0 CHa
--CHz-&--bH
I
As was shown by Gregor, Luttinger and Loeb1,Z binding phenomena are essentially the same for soluble polychelates as for systems where a gel phase has formed. The polymer itself is soluble in dioxane. Fernelius, et al.,Ip studied chelation in dioxane-water mixtures using the glass electrode. They showed that the meter pH deviated significantly from the true H, especially for dioxane-water ratios greater than 4/1. &e solubility of PAK a t different pH levels and in the presence of metallic ions was examined using turbidity as the criterion. Most of the systems were insoluble in 1/1 dioxane-water solutions; solubility increased with dioxane content. It was decided that a 4/1 dioxane-water system was most suitable from the point of view of maximum polymer solubility, reasonably rapid ap roach to equilibrium and ease of pH interpretation. fiernelius, et a1.,12 purified dioxane by refluxing over metallic sodium for several hours followed by fractionation throough a 30-plate column, discarding the forerun up to 101 This technique was modified by the use of an atmosphere of pre-purified nitrogen throughout reflux and distillation. Purified dioxane was stored in the dark under nitrogen. It was found to maintain a negligible acid titer for a number of weeks. Dioxane available from different commercial sources was always found to have an appreciable acid content. A fiber tip saturated calomel electrode and a Beckman number 1190-80 glass electrode were used with a Beckman model G pH meter in all the determinations of hydrogen ion activity. Polyvinylacetonyl ketone (0.01 M) and acetylacetone (0.01 M), each in a 4/1 (v./v.) dioxane-water solution of 0.1 M sodium nitrate, were titrated in the presence and absence of several metal salts a t different concentrations by aqueous sodium hydroxide solutions a t room temperature. In several of the experiments nitric acid was added to the solution before the titrations were performed. A “stepwise” technique was used for PAK, with the solutions kept in screw-ca polyethylene bottles under nitrogen, such that the p l f was measured after 3 to 5 days of shaking, following each increment of added base. Rate experiments showed that equilibrium was reached within three days. The over-all reproducibility of each measurement was f 0.05 pH unit. Direct titrations with acetylacetone were performed without recourse to the stepwise procedure because no precipitate or colloid was observed below the pH of recipitation of the metal hydroxide. l!~ these titrations there is a competition between the chelating agent and the hydroxide ions for the metal ion. The data used for the calculations were, without exception, concerned with measurements in acidic solution where the hydrolysis of metal ions was a t a minimum; no data were used above the pH at which hydroxides precipitated. This point was determined by titrating 0.005 M dioxane-water solutions of the respective metal salts in 0.1 M sodium nitrate. Precipitation was observed at the following ~ H M (meter pH) values: nickel(I1) sulfate at ~ H M = 6.20; manganese(I1) nitrate at 6.50; copper(I1) nitrate a t 4.38; cobalt(I1) nitrate a t 5.50; uranyl nitrate a t 5.40; praseodymium(II1) nitrate at 5.83. Neodymium(II1) nitrate appeared to form a hydroxy complex a t 7.50. The functional groups I and I1 in PAK are distinguished by the fact that only form I has a titratable hydrogen and can act as a chelating agent; form I1 and methyl ketone mers cannot chelate metal ions. The base molecular weight of form I was determined by the conductometric titration of PAK in dioxane-water. The equivalent weight was 291, as compared to 112.1 calculated for the entire polymer in form I. The conductometric equivalence point coinoided with the approximate one obtained potentiometrically using a pH meter.
.
0 CH, CH3-
Vol. 62
BF3 -+ 0 CHsO and CHa-&-(!d%CH3 I
The carbon-hydrogen analysis of the polymer indicated that condensation with polyvinylmethyl ketone took place; it could not indicate relative amounts of the three forms. As will be seen from later characterization data, the amount of I, the only form important in chelation, can be determined. Product I1 would not be formed were the monomer methyl isopropenyl ketone. Polyvinylacetonyl ketone can also be prepared b condensation of acetic anhydride with polyvinyl methyl getone using boron trifluoride as catalyst. The monomeric analog of this reaction using acetone is described,8 with a yield. of 85%. This method of synthesis was also attempted, using polyvinylmethyl ketone prepared by peroxide-catalyzed polymerization of the monomer. About 40 g. of the polymer was dissolved in 300 g. of acetic anhydride, the mixture cooled to -15’ and boron trifluoride bubbled through it. After several hours a brown polymer was obtained with a yield indicating 80% conversion. However, this polymer was sli htly soluble only in acetone, acetylacetone or dimethylformarnide. Polyvinylmethyl ketone prepared by peroxide polymerization is usually branched, which may account for the limited solubility of the polymer prepared from this materia1.g Attempts also were made to prepare the monomer, vinyl acetonyl ketone, by treating acetoacetyl chloride (the preparation of which is described by Hurd and Kelsolo) with ethylene in the presence of aluminum chloride as catalyst, using the general procedure of McMahonlI who prepared ethyl vinyl ketone. However, under the conditions of this procedure i t was found that the monomer polymerized too readily to allow its isolation. Titrations.-Polyvinylacetonyl ketone is but slightly soluble in water; in the presence of metals it forms a precipitate. For example, the addition of 0.001 M copper(I1) to a 0.01 M (base moles of form I diketone) solution of PAK in water resulted in immediate precipitation; the pH of this system decreased slowly, reaching a constant value in about 30 days. With the addition of base this period of equilibration was even longer. From practical considerations water could not, therefore, be used as the solvent. Precipitation of the polymer does not invalidate the use of pH as an indication of the extent of chelation provided that equilibrium is attained when the pH is determined. (7) C. R . Hauser and J. T. Adama, J . Ana. Chem. Soc., 66, 345 (1944). (8) C. E. Denoon, Jr., Ors. Byntheses, 20, 6 (1940). (9) C. E. Schildknecht, “Vinyl and Related Polymers,” John Wiley and Sons, Ino., New York, N. Y., 1952, p. 090. (10) C. D. Hurd and C. D. Relso, J. A m . Chem. Soc., 6 2 , 1548 (1940).
(11) E. M. McMahon, J. N. Roper, Jr., W. P. Utermohlen. Jr., R. 13. Hasek, R . C. Harris and J. H. Brant, ibzd., 70, 2971 (1948).
Theoretical and Results Titration Curves.-Figure 1shows potentiometric titration data for AA and both conductometric and potentiometric data for PAK. The degree of neutralization CY is defined as the equivalents of base added per equivalent of chelate acid present initially, a negative CY occurring in a solution containing nitric acid initially. Figures 2, 3 and 4 (12) L. G . Van Uitert, W. C. Fernelius and B. E. Douglaa, ibid., 711, 457 (1953); see also L. G.Van Uitert, Dissertation, Pennsylvania State College, June, 1952.
Oct., 1958
METAL-POLYELECTROLYTE COMPLEXES
1241
T
ooTo .5
1.5
1.0
a. Fig. 1.-Titrations of 0.004 M solutions of chelate acids with sodium hydroxide in 4/1 dioxane-water mixtures 0.1 M in sodium nitrate: potentiometrically for polyvinylacetonyl ketone (0)and acetylacetone (0); conductometrically for polyvinylacetonyl ketone (A).
2
0
.4
.6
.8
1.0
a. Fig. 3.-Titrations of 0.004 M polyvinylacetonyl lcetone with sodium hydroxide in the absence of metal ions (A), with 0.003 M rasaeodymium(II1) nitrate (o), and 0.003 M neodymiumhI1) nitrate ( o ) , all in 4/1 dioxane-water and 0.1 M sodium nitrate.
4
PHW 3
2
1
-2
I 0
I
I
.2
.4
g.
Fig. 2.-Titrations of 0.004 M solutions of polyvinylacetonyl ketone with sodium hydroxide in the absence of metal ions (o), with 0.006 M nickel(I1) sulfate (A), 0.005 M copper(I1) nitrate (n), 0.006 M urany!(II) nitrate (a)] and 0,011 M uranyl(I1) nitrate (m), all in 4/1 dioxanewater and 0.1 M sodium nitrate.
represent the titrations of PAK and AA with typical concentrations of metal salts. In many cases titrations at more than one concentration of metal ion were required to determine a coordination constant. The titration curves were displaced to the right t o varying degrees by the different metal salts, indicating the extent of binding. Interpretation of Meter Readings.-For dioxane-water solutions, Fernelius, et al.,n defined an acidity function U H equal to the activity of hydrogen ions as indicated by the pH meter ( a M ) divided by the known hydrogen ion concentration (CH) from hydrochloric acid. For dioxane-water ratios greater than unity, U H is greater than unity and increases with this ratio. Glass electrode measurements in dioxane-water
I
-.2
0
.2
.4
a, Fig. 4.-Titrations of 0.1 M acetylacetone with sodium hydroxide in the absence of metal ions (o)]with 0.006 M nickel(I1) sulfate (V), 0.005 M copper(I1) nitrate (a), 0.006 M uranyl(I1) nitrate (m), 0.003 M uranyl(I1) nitrate (a), all in 4/1 dioxane-water and 0.1 M sodium nitrate.
systems differ from those in purely aqueous systems; the junction potentials will vary with the dioxane-water ratio, the ionic species present, and their distribution at the interfaces. If a constant excess of the same neutral salt is present in the same combination of solvents, it is reasonable that these junction potentials remain constant over a range of p H levels. That this was the case for measurements in 0.01 M sodium chloride in 3/1 dioxane-water was demonstrated by Fernelius. l2 Mean activity coefficient (7") data for nitric acid in dioxane-water systems are not available,
1242
G. K. HOESCHELE, J. B. ANDELMAN AND H. P. GREGOR
but are for hydrochloric acid.13 It may be assumed that mixed electrolytic systems show the same general behavior in mixed solvents as in water. Harned’s rule is log
Y1*
= log
Yl*(O)
- LY12/.42
where yl* refers t o component 1 of two electrolytes in a mixture and y*l(o) to that of component 1 in the absence of component 2, all mixtures being at the same total ionic strength; p2 is the ionic strength of component 2; a12is a constant unique for a mixture of two particular electrolytes. It has been shown14for mixtures of hydrochloric acid and sodium chloride in methanol-water of varying ratios up to 6001, methanol that Harned’s rule holds, and that the same a 1 2 values are obtained for the different solvent ratios. It is, therefore, not unreasonable to assume constancy of a 1 2 in dioxane-water for a given salt mixture. It has been shown16 that aHCl-NaCl is 0.043 in water. Using this value of a12and the value of -y* for 0.1 M hydrochloric acid in dioxane-water mixtures in the absence of sodium chloride, one can calculate y* for other concentrations of hydrochloric acid < 0.1 M when CHCI CNaCl = 0.1 M . For 0.1 M hydrochloric acid in 82% (by weight) dioxane a t 25”, Y*HCI = 0.0634.13 It readily can be shown that the term a12p2 in the Harned equation is so small in these dilute solutions that Y*HCl in hydrochloric acid-sodium chloride mixtures will nearly equal Y*HCl in 0.1 M acid. A series of hydrochloric acid-sodium chloride solutions of constant ionic strength (0.1) in 4/1 dioxane-water were prepared and their p H measured using the glass electrode. Using Y*HCl = 0.0634 and assuming that Y*HCl = Y+H, the activity of hydrogen ion in each of these solutions was calculated. The difference between pHa (that calculated from y* and CH) and ~ H (that M read by the pH meter) was used as the correction M 1.61 for 0.1 M acid, term. The pH, - ~ H was 1.68 for 0.01 M , and 1.67 for 0.001 M ; this constancy being in agreement with the results of Fernelius, the average value of 1.65 was used. For the nitric acid-sodium nitrate system, the same procedure would be preferable but cannot be employed because of lack of Y * H N O ~ data in dioxane-water systems. Accordingly, mixtures of nitric acid and sodium nitrate at M = 0.1 in 4/1 dioxane-water were prepared and the value of ~ H measured. M Assuming that the liquid junction potentials of the nitric acid-sodium nitrate mixtures are the same as those of the hydrochloric acid-sodium chloride mixtures and using the PHa-PHM. values of the latter system, the hydrogen ion activity coefficient Y*HN03 was calculated. The values obtained were: for 0.1 M nitric acid, Y * H N O ~ = 0.0672; 0.01 M , 0.0585; 0.001 M , 0.0533; 0.0001 M , 0.0424. While a trend does exist, an average value of p H N O 3 = 0.055 was calculated and employed a t all pH levels. A value of UH = 2.46 was calculated.
+
(13) H. S. Harned and B. B. Owen, “The Physical Chemistry of Electrolytic Solutions,” Reinhold Publ. Gorp., New York, N. Y., 1950. (14) G . Akerlof, J. W. Teare and H. E. Turck, J . Am. Chem. Soc., 59, 1916 (1937). (15) R . A. Robinson and R. H. Stokes, “Electrolyte Solutions,” \ Academic Press, Inc., New York, N . Y., 1955.
Vol. 62
It may be that an extrapolation to high pH values is not as valid in nitrate solutions as in chloride solutions. However, a t higher p H values the actual concentration of hydrogen ion is a minor term in the calculation of formation constants. The extrapolation from hydrochloric acid-sodium chloride data to nitric acid-sodium nitrate data is open to question because the nitrate ion would be expected t o behave rather differently a t phase boundaries. This possible error is, however, small relative t o the other experimental errors and approximative errors of the theoretical calculations. Calculation of Cob’rdination Constants.-The method used to calculate coordination constants is a modification of Bjerrum’s method and has been described in detail in the first papers of this series.2 For a chelation system with a maximum coordination number of two there are two successive formation constants ICl and IC2 such that
where M++ is the metal ion being coordinated and A- the ligand; MA+ and MA2, respectively, refer to the two complex species. For the cupric ion, Cu(II), binding to a P-diketone the following reaction can be written 2R-
8 1-CH=
-CHs
K
+ CU++I_ Cu(RCOCHCOCH3)z
or 2A-
+ CU++
K CuAz
For acetylacetone “R” is methyl, while for PAK it is the polymer chain. An extensive investigation of the coordination of various p-diketones with metal ions in dioxane-water solutions has been made.12 On coordination the charge of the diketone changes. Since changes in the charge of a polymeric chain influence binding reactions, the most meaningful comparisons and calculations are made when the charge on the chain is not altered as for 2HA
+ M++
B MA2
+ 2H+
Here the displacement constant is designated B. The relationship between B and K is readily seen to be B = KKa2 where Ka is the acid dissociation constant. For the successive steps in a displacement reaction one can by analogy write bl = klK,, bz = kzK., and B = blba
For a monobasic acid K, is well defined. For a polybasic acid it has been shown that
where Z is the ratio of charged to uncharged monomeric units along the polymer chain.2 For the titration of an acidic polymer in the absence of a chelating metal ion, the charge ( Y ) per monomeric unit is equal to a, the degree of neutralization. Thus in Fig. 1 the abscissa may be a! or Y . One
.
Oct., 1958
METAL-POLYELECTROLYTE COMPLEXES I
I
I ‘
I
1243 I
t
7.0
6.!
PK,. 6.C
5.!
Fig. 7.-Modified Bjerrum lot for 0.01 M acetylacetone uranyl(I1) nitrate ( 0 ) ; also with U.006 M (m) and 0.003 0.004 M polyvin lacetonyl ketone with 0.006 M ( 0 ) and 0.002 M uranyl(&) nitrate (0).
i 8
I
I
I .a
I
.3
Y.
Fig. 5.-pK. as a function of ne ative charge per monomeric unit ( Y ) for polyvin$acetonyl ketone.
\ Ti.
1.0
i
\
-
When A equals unity, klkz = 1/[A-I2. A plot of A VS. 1/A- or PA- leads to klk2 values. The slope of this plot indicates the separation between successive formation c o n ~ t a n t s . ~ By algebraic substitution one obtains the more useful relationship
Although bl and bz are not independent of the charge along the polymer chain, one may now make a modified Bjerrum plot of A vs. p( [HA]/[H+]). At@=1 From the relationships
.5
where [Mt] and [A,] are the total metal ion and weak acid concentrations, respectively, and -3
-2
Fig. 6.-Modiiied Bjerrum plot for 0.01 M acetylacetone with 0.005 M copper(I1) nitrate ( 0 )and 0.004 M polyvinylacetonyl ketone with 0.001 M ( o ) ,0.002 M (A), and 0.005 M copper(I1) nitrate (0).
may use this titration curve, making the appropriate conversion from PHMto a H , to calculate K, as a function of Y for PAK. This was done, the result being plotted in Fig. 5. Using Bjerrum’s concepts, one defines the degree of formation fi equal to the average number of ligands bound per metal ion present in any form suchthat .
[HA]
=
[At](1 - a) - [H+l
based upon electroneutrality and applying to all acids in the absence and presence of metal ions, one obtains For monomeric acids, a, [A,] and [Mt] are known, [ H f ] is determined from ~ H Mand , [A-] is calculated. The Bjerrum plot of rt vs. p ([HA]/ [Hf]) follows. For polymeric acids, however, one cannot similarly calculate [A-] since Y is not equal t o (Y in the presence of metal ions and, hence, K, is not known. Here one estimates [A-1, calculates a tentative value of Y , until by an iterative method the true value of [A-] is obtained from the general relationship of pK, and Y. This is then used t o calculate a and finally to make a Bjerrum plot.
1244
G. K. HOESCHELE, J. 13. ANDELMAN AND H. P. GREGOR
It should be noted that in this calculation the assumption is made that the intermediate species MA+ is not present. This is reasonable when one considers that the polymer is coiled and has a relatively high local negative charge concentration. Thus if a metal ion were to be found within this field, the MA+ species formed would react with an adjacent site to form MA2. Figures 6 and 7 are plots of rt vs. p ( [HA]/[ H f ] ) for Cu(I1) and UOz(II), respectively. Points calculated for experiments with varying metal concentration generally fall on the same line. For acetylacetone, pKa = 11.87 as calculated from its potentiometric titration. In the titration of PAK in the presence of UOZ(II), at fz = 1 the value of pKa = 5.3 and Y = 0.013; for PAK and Cu(II), pKa = 5.9and Y = 0.1. Formation and displacement constants are summarized in Table I. From the K f = klkz values it would appear that acetylacetone is a stronger complexing agent than PAK. However, the B values, which are the more correct criteria for binding when all the competing reactions are to be considered, indicate the same degree of binding of PAK and acetylacetone to Cu(II), while the polymer binds U02(II) more strongly than does the monomer. The binding of acetylacetone and PAK for Ni(I1) is relatively weak, as seen in Figs. 2 and 4; formation or displacement constants were not calculated. The titration curves in Fig. 3 show strong bindine: by PAK for Pr(II1) and-Nd(III), compirable t; that for U02(II).
Vol. 62
calculated AH = -8.7 kcal. mole-' and A S = 31 e.u. for the binding of two molecules of acetylacetone to Cu(I1); from their data we have calculated the comparable parameters for binding to U02(II) where AH = - 11.9 kcal. mole-1 and A S = 23.9 e.u. Thus, although the respective formation constants were 15.2 and 14.2, the intrinsic binding energies and entropy changes were quite different. Although the tetracoordinated Cu(I1) is known to be planar, little information is available concerning the stereochemistry of U02(II) coordination. There are indications" that the UOz(II) ion is linear in aqueous solution. If this configuration were maintained on coordination, there would be limitations on the possible configurations of the tetracoordinated ion, but the restriction would be generally in a plane as for Cu(I1). If integrity of the linear aonfiguration were not required, there would be less of a steric effect in polymeric binding with UOz(II) than with Cu(I1). This could account for the difference in binding constants of the polymer with these two ions, the monomeric unit binding being equal. Another consideration is the entropy effect due to the localization of ions and their solvation atmosphere on binding. In the tetravalent binding of Cu(I1) not all of its atmosphere would necessarily be displaced due to the open positions above and below the tetravalent plane. Thus there may be entropy changes due to the localization of water molecules as well as metal ions. For UOz(II), however, this effect would be much less due t o the occupation of these positions by oxygen atoms. TABLE I The exact nature of these steric and entropy effects and their relative importance cannot be evaluated DISPLACEMENT AND BINDINQ CONSTANTS Acid Metal ion Log B Log K pKa at this time. AA Cu(I1) -4.8 18.9 11.87" The displacement Constants, B, are reflections of all of the above described effects for both the PAK CU(I1) -4.8 7.0 5.9 AA UOdII) -5.0 18.7 11.87 metal binding and proton binding. Their importance is seen in the fact that these are competPAK UOdII) -1.8 8.8 5.3 a I n these and previous calculations,e [H+] appearing in ing reactions and that the amount of metal actually the expression for K , and [HA]/[H+]refers to hydrogen ion bound by the polymer will be determined by their activity. Because activity coefficient corrections are in- relative magnitude. From this practical consideradeterminate for polyions, these were not made for any anionic species, in order to provide a valid comparison be- tion we say that UOz(II)is bound more strongly by tween monomer and polymer. The appropriate correction three orders of magnitude, while the monomeric to [.4-]for acetylacetone yields a p K , value of 13.13. unit and the polymer bind Cu(I1) to about the same extent. Discussion The difference in the binding of PAK for Pr(II1) In 4/1 dioxane-water solution acetylacetone and Nd(II1) is not conclusive. It is worthy of furshowed approximately the same binding constant ther investigation from the point of view of sepfor Cu(I1) as for UOz(II). Izatt, Fernelius and aration of these species. BlocklG studied these binding phenomena in Acknowledgment.-This investigation was supaqueous solution at various temperatures. They ported in part by the National Science Foundation. (16) R. M. Izatt, W . C. Fernelius and B. P. Block, THISJOURNAL, (17)L. H. Jones and R. A. Pennernm, J . Chem. Phys., 21, 542 69, 235 (1955).
(1953).
.
