Complex Solubilities of the Silver Halides
429
Complex Solubilities of the Silver Halides in Propionitrile and Propylene Carbonate Mixtures with Tetrahydrothiophene Mark Salomon Power Sources Technical Area, U.S. Army Eiectronics Technologyand Devices Laboratory, Fort Monmouth, New Jersey 07703 (Received July 3 1, 1974) Publication costs assisted by the U.S. Army Electronics Technologyand Devices Laboratory
The complex solubilities of the silver halides (AgCl, AgBr, and AgI) have been determined in propylene carbonate and propionitrile mixtures with tetrahydrothiophene (THT). The results are compared with previous data for mixtures with SOg. Molecular orbital calculations were carried out to determine the nature of anion solvation in these systems. The results indicate that in SO2 containing solvents, C1- ion solvation is stabilized by coordination to the sulfur by considerable charge transfer to both the S and 0 atoms. In the THT-Cl- complex, charge transfer is not as complete and the S-C1 c bond contains a much larger s-orbital contribution thereby resulting in a much weaker bond.
Introduction In a previous paper1 the complex solubilities of the silver halides were studied in mixtures of propionitrile (PN) and S02. It was found that the presence of SO2 in P N stabilizes both the halide ion and Ag+. In the present work the complex solubilities of the silver halides are determined in mixtures propylene carbonate (PC) and propionitrile with tetrahydrothiophene (THT). A major difference between T H T and SO2 is that in the former, the sulfur atom possesses a net negative charge in the furan type ring whereas in the latter, the sulfur is in a positive oxidation state and carries a net positive charge. The complex equilibria studied are AgX = Ag” + x’ Ks 0 (1) Ag+ + 2 x ’ = AgX2- pz (2) Ag’ + 3X- = AgX3- p3 (3)
-
where X is C1, Br, or I; KSois the solubility product and @2 and @3are complex stability constants. Solid AgX can dissolve in the presence of excess halide according to
AgX +
X- = AgXz-
Ks2
(4)
Ks3
(5)
and AgX - + 2 x ‘ = AgXS2-
Experimental Section The stability and solubility constants were determined from potentiometric titration data as described previously. 1,2 The electrochemical titration cell can be represented bY Ag(indicator electrode) 1 halide solution 1 AgCIO,, Ag (refer ence) (8) The electrolyte solutions were mixtures of a tetrapropylammonium halide (TPAX) with tetrapropylammonium perchlorate (TPAP) and the total ionic strength was kept constant at either 0.1 or 0.01 M (molar). The AgC104 solu-
tion in the reference electrode compartment was kept at the same ionic strength as in the halide solution compartment and the two were separated by a salt bridge of the type described previously.lS2 The emf of cell 8 is given by
E = E’
+
( 2 . 3 R T / F ) log [Ag’]
(9)
In eq 9, [Ag+] is the silver ion concentration (molar units) and E’ is a formal potential in that it contains unknown contributions from the liquid junction and activity coefficients. These latter two quantities are assumed to be constant because of the constancy of the ionic strength. All chemicals and solvents with the exception of PC and T H T were purified as described previously.1,2PC was distilled at 71’ at 0.4 Torr and T H T was distilled at atmospheric pressure and collected at 119 f 1’. In both cases only the middle third of each distillate was saved for use. All purified salts and solvents were kept in an argon filled drybox in which the water and oxygen content was less than 1ppm. The potentiometric titrations were carried out at 25 f 0.05”. Further experimental details can be found elsewhere.lP2 Several rough determinations of the total solubilities of the silver halides in pure T H T were performed by weighing the residue of a given volume of the saturated solution. The solubilities were obtained by approaching equilibrium from the “undersaturated” solution. The solvent was evaporated with a heat lamp and the residue weighed directly. No attempt was made to analyze for solvate formation. Two determinations were made for AgCl and one each for AgBr and AgI.
Results The experimental quantities3 used to calculate the vari, and E (the total silous equilibrium constants are C A ~Cx, ver and halide concentrations and emf of cell 8, respectively). The [Ag+] required for the calculations are obtained from eq 9 having previously determined E’ by titrating AgClO4 into TPAP s01ution.~For the two most dilute THT-PC solutions ([THT] = 0.09 and 0.56 M ) , the Nernst slope was obeyed to within f0.6 mV (standard deviation) and for the PC-1.58 M T H T and PN-0.10 M T H T solutions, the standard deviation from the Nernst slope was 3 The Journal of Physical Chemistry, Vol. 79, No. 5, 1975
Mark Salornon
430
TABLE I: Equilibrium Constants at 25"
PC-O.09 M THT"
pc-0.56
MTHT~
PC-1.56 itz T H T ~ PN-O.10 ;LI THTb
AgCl AgBr A gI AgCl AgBr '%I AgCl AgCl AgBr AgI
12.07 12.50 13.73 9.40 10.02 10.93 8.30
* 0.02
13.083 i 0.003 13.22 i 0.04 14.49 f 0.05
15.01 15.29
f i.
0.03 0.05
1.01 f 0.02 0.72 f 0.05 0.77 i 0.07
2.94 2.79
i i.
0.03 0.05
f
0.02 0.05
i i i
0.02 0.03 0.10
10.538 i 0,002 11.06 f 0.01 12.33 f 0.08
12.13 11.97
i
0.01 0.11
1.14 1.04 1.40
i i. i
0.02 0.04 0.13
2.74 1.95
f i.
0.02 0.12
1.00
i
0.36
2.44
f
0.35
f
* 0.35
10.7 * 0.1 10.8 i 0.2 11.1 i 0.5
9.30 11.8 13.0 13.8
i i i i
0.07
* 0.04
10.74
0.2 0.1 0.1
14.3 15.0 15.8
*
i. i
+
0.1 0.1 0.1
1.1 i. 0.3 2.2 i 0.3 2.7 i 0.5
3.6 f 0.3 4.2 i 0.3 4.7 i 0.5
T HT
0.65 f 0.2 AgCl AgBr -0.5 -0.3 A gI AgCl so2 (OOY -3.4 AgBr -7.6 -6.3 AgI PC"d 19.87 AgCl 20.87 23.39 1.00 3.52 AgBr 21.2 22 20.5 1.5 0.7 22.8 21.8 1.0 AgI PNb9' 14.29 AgCl 15.94 16.71 2.42 1.66 AgBr 16.24 17.63 1.29 14.95 2.68 17.25 1.18 16.08 .~ A gI a Ionic strength = 0.01 M . b Ionic strength = 0.10 M . c Data from ref 4. Data from ref 5 . e Data from ref 6. f Data from ref 1.
TABLE 11: Free Energies of Transfer of Single Ions from Water to Various Aprotic Solvents
___-
AG,(ion), kcal/mol
-
Solvent
Ag'
C1'
Br-
I-
AgC12-
AgBr2-
A@,-
-
PC PC-O.09 M PC-0.56 izI PN PN-O.10 M PN-0.95 M PN-3.30 M
5.3 8.6 6.0 2.7 1.5 -0.8 -7.1 THTa -5.7 8.8 6 .O 2.6 1.4 -1.4 -5 .O THT' -9.5 9.0 6.4 2.6 1.4 -1.4 -5.9 -2.0 9.5 7.0 3.5 1.2 -1.2 4.7 THT' -8.9 10.1 6.2 2.2 2.6 -2.6 -8.1 SO:, -2.8 5.1 3.2 0.9 1.2 -1.2 4.5 SO2 -3.0 3.8 1.9 -0.2 1.2 -1.2 -3.6 DMSO -7.6 8.7 5.6 1.6 0.6 -2.6 -6.6 C H3N02 5.1 7.8 5.1 1.4 -1.4 a From present data uncorrected for activity effects in solution. Remaining data are corrected for activity effects and were discussed in ref 1.All data refer t o the molar scale at 25" except the data for nitromethane. b Data for nitromethane at molar scale, are from ref 19. ZOO,
mV. The equilibrium constants in eq 1-7 were calculated by an iterative non-linear least-squares method.1,2 The results are given in Table I with the corresponding standard deviations. Since only one run was made for K,o for AgBr and AgI in pure THT, no standard deviation is given. Included for comparison in Table I are the data for pure PC, PN, and SOz. No attempt was made to correct the equilibrium constants for activity coefficients and the possibility of ion pairing. By analogy to other system^^,^,*,^ the latter effect is small and correcting for activity coefficients in 0.1 M solutions typically decreases log KSo by about 1 log unit and increases log 6 2 by about 1 log unit: log Ks2 is independent of concentration effects (but not ion pairing). In Table I1 values for the free energy of transfer of single ions, AGt(ion), are given for the transfer of the ions in reactions 1-3 from water to the aprotic solvent. The values for The Journai of Physicai Chemistry, Voi. 79, No. 5, 1975
the present solvent systems are based on the assumption1 that AGt(AgC12-) = -AGt(AgBrz-) and from which it follows that
2AG,(Ag+) M AGt(Ag+,AgC1,')
+
AG,(Ag+,AgBr,-)
(10) Discussion The effect of T H T additions to PN and PC is to greatly reduce the @'s and increase KSo. This can undoubtably be attributed to the increased stability of Agf in solution due to d n back bonding between the 4d electrons of Ag+ and low-lying d-orbitals of the sulfur atom as proposed by Chat@ and discussed in a previous paper in this series.l This effect is also reflected in the large negative values of AGt(Ag+) given in Table 11. What does seem surprising at
Complex Solubilities of the Silver Halides
431
TABLE 111: Gross Bond Populations for Neutral Speciesa Molecule
Orbital
so2
S
1.822 2.362 1.331 0.485
0
THT
S
DMSO
c2, 3
-
c4,5
S
-
0
CH3
--
1.778 1.020 0.988 2.000 1.929 4.594 3.030 3.863 2.988 2.980 0.483 0.486 0.484 4.522 0.019 -0.008 0.032 4.242 -0.124 4 a Basis functions are 39, 3p, 3d for sulfur and 2s, 2p for oxygen and carbon. DMSO results are from ref 15. q is the net charge on each atom are the carbons at the apex in C-S-C and C4,5 are the carbons at the base of the pentagonal THT molecule. in electron units. S
P d
1.811 4.431
TABLE IV: Gross Bond Populations for Charged Speciesa Charged species
0 - w
so,-C1' Orbital
(free) C1-
c1
2 .ooo 6.000 0 .ooo -1 .ooo
1.914 5.191 0.190 -0.295
0
S
1.852 1.816 2.136 4.601 d 1.882 4 0.130 -0.417 a See footnote for Table III. Basis functions for C1 are 3s, 3p, 3d. S
P
first is the fact that Ks2 is virtually unaffected by the T H T additions whereas SO2 additions to P N resulted in negative values for log KS2.lIt thus appears that the stability of the anion also plays a major role in these equilibria. For example, an increase in the stability of X- would increase the solubility product KSo (eq 1) and decrease the complex stability constant Ksz (eq 2) whereas as increase in the Age ion stability would increase KSo and leave Ks2 virtually unchanged. Clearly the later effect predominates for the present systems, but a combination of effects appears to be operable in the PN-SO2 systems. There is spectroscopicg and kineticlo evidence that halides form charge-transfer complexes with SO2 in organic solvents and water. In our previous paper,l the possibility of forming complexes of the type SOZ-Cl- was neglected. Below molecular orbital calculations are presented which demonstrate the importance of these charge-transfer type complexes in SO2 solutions and their relative unimportance in T H T mixed-solvents. The molecular orbital calculations were performed by the CNDOI.2 method.11J2 The program is essentially that given by Pople and Beveridge13 but was modified to run on the Burroughs B5700: the original parameterizations were retained. The reactions studied by this method are and
c1- + so*= co,s-c1,-
AEl
(11)
where aE is the energy difference (the stabilization energy) between the charged complex and the free ion and solvent molecule. In Table I11 the gross bond populations are given for various sulfur containing molecules. The results for dimethyl sulfoxide (DMSO) are from ref 15. In SO2 the bond angle was taken as 119.54' and S-0 distance was optimized and found to equal 1.51 A. For THT, no structural data
c1
S
1.go9 5.179 0.277 -0.364
1.838 3.353 1.098 -0.289
c2,3 1.045 3.003 -0.049
C4, j 0.996 3.005 0.033
exist and dimensions were assumed based on structurally similar mole~ules:1~ bond lengths were taken as C-S = 1.80 A, C-C = 1.54 A, and C-H = 1.09 A; bond angles are C-SC = 98O, S-C-C = 115O, and C-C-C = 106'. For coordination between the sulfur atom and a C1- ion, the CND0/2 calculations resulted in the gross bond populations given in Table IV. In these calculations the optimum S-C1 distance was found to equal 1.90 8. The stabilization energies, hE1 and hE2, are abnormally large and are -230.8 and -204.6 kcal/mol, respectively. These large values for hE are not particularly alarming since they are characteristic of the theory and are highly dependent upon the parameterization.16~17The CNDOI2 method does however successfully predict differences (and trends) in stabilities and in bond popu1ations.l6-l8 On this basis the S-Cl bond in SO2 is (neglecting zero-point and kinetic energies) some 26 kcal/mol stronger than in THT. In both cases there is charge transfer from the halide to the solvent molecule. In SO2 this effect is quite large amounting to some 71%; charge is transferred mainly to the u and a orbitals on S and the u and p a orbitals on 0. Some back bonding into the d orbitals of C1 occurs. The major difference between SO2 and T H T mixtures lies in the greater amount of s-orbital contribution to the S-C1 u bond in the latter system and in the increased amount of d a back bonding. In THT-Cl- the excess negative charge on the sulfur is dissipated from the a orbitals to the antibonding orbitals of the methylene groups (all the methylene hydrogens acquire a net negative charge). Because of this unfavorable configuration, the net charge transfer from C1 to the S atom in T H T is not as great as in SO2 and the net bonding in the former complex is therefore weaker. Calculations were not performed to determine if the C1- ion might bond to other parts of the THT molecule and this possibility cannot be ruled out. This possibility is however extraneous to the present conclusions. The Journal of Physical Chemistry, Vol. 79, No. 5, 1975
Mark Salomon
432
The role of the solvent in regards to the behavior of Ag+ is quite diverse. In general the Ag+ ion is seen to gain energy when transferred to solvents which can participate in dn back bonding. From Table I1 it is seen that AGt(Ag+) is positive for transfer to PC and CHsN02 despite the fact that, according to CNDOIB calculation^,^^^^^ the net charge on the oxygens is more negative than in water. In DMSO, as in S02, there are two possible binding sites for Ag+ and based on the gross bond population (Table 111) which show the sulfur to possess a net positive charge, it might be anticipated that Ag+ would coordinate with the negatively charged oxygen. This apparently is not the case. Spectroscopic studies in DMSO show two distinct coordination sites for cations which appear to follow the "hard-soft" classification of Pearson20 and the earlier observations of Ahrland, et aL8 For the typically hard Lewis acids Li+, Na+, etc., the oxygen atom in DMSO is the site of coordination.21 For the intermediate acids Cu(II), Co(II), Ni(II), and Mn(II), the oxygen atom is again the site of coordination whereas the soft acids Pd(I1) and Pt(I1) apparently bind to the sulfur atom.Z2 Since Ag+ is a typical soft Lewis acid and since AG,(Ag+) for the transfer from water to DMSO is negative, it is logical to conclude that Ag+, and probably all the soft acids, bind to the sulfur atom in DMSO. Carrying this analogy to the PN-SO2 system, it appears that both Ag+ and C1- solvation are more stable than in the pure solvent (PN) and that both ions coordinate to the sulfur atom.
(105 X 148 mm, 24X reduction, negatives) containing all of the supplementary material for the papers in this issue may be obtained from the Journals Department, American Chemical Society, 1155 16th St., N.W., Washington, D. C. 20036. Remit check or money order for $4.00 for photocopy or $2.00 for microfiche, referring to code number JPC-75429.
Acknowledgments. The author is grateful to Leon Leskowitz for his help in modifying the CNDOI2 program. Grateful acknowledgment is also due Dr. B. K. Stevenson for performing the experiments in PN-0.1 M THT.
(1970). (17)K. G. Breitschwerdt and H. Kistenmacher, Chem. Phys. Lett., 14, 288 (1972). (18)J. A. Pople and M. S.Gordon, J. Amer. Chem. Soc., 89, 4253 (1967). (19)J. Bardoz-Lambling and J.-C. Bardin, C. R. Acad. Sci., Ser. C, 266, 95 (1968). (20)R. G. Pearson, J. Amer. Chem. Soc., 65, 3533 (1963). (21)B. W. Maxey and A. I. Popov, J. Amer. Chem. Soc., 91, 20 (1969). (22)(a) F. A. Cotton, R. Francis, and W. D. Horricks, J. Phys. Chem., 64, 1534 (1960);(b) F. A. Cotton and R. Francis, J. Amer. Chem. Soc., 82, 2986 (1960);(c) B. F. Johnson and R. A. Walton, Spectrochim. Acta, 22, 1853 (1966). (23)H. L. Yeager, J. D. Fedyk, and R. J. Parker, J. Phys. Chem., 77, 2407 (1973).
Supplementary Material Available. Listings of concentrations, cell emf's, and formal potentials for each solvent system will appear following these pages in the microfilm edition of this volume of the journal. Photocopies of the supplementary material from this paper only or microfiche
The Journal of Physical Chemistry, Voi. 79, No. 5, 1975
References and Notes (1) M. Salomon and B. K. Stevenson, J. Phys. Chem., 77, 3002 (1973). (2)M. Salomon, J. Phys. Chem., 78, 1817 (1974). (3)See paragraph at end of text regarding supplementary material. (4) J. N. Butler, Anal. Chem., 38, 1799 (1967). (5)J. Courtot-Copez and M. L'Her, Bull. Soc. Chim. Fr., 675 (1969). (6) P. J. Eiving and J. M. Markowitz, J. Chem. Educ., 37, 75 (1960):note; there is a slight difference in Ksa(AgCI) quoted in thls reference and in the earlier work of Jander (see ref 7). (7)G. Jander, Naturwissenschaften, 26, 779 (1936). (8) S.Ahrland, J. Chatt, and N. R. Davies, Quart. Rev., Chem. Soc., 12, 265
(1958). (9)(a) E. R. Lippincott and F. E. Welch, Spectrochim. Acta, 17, 123 (1961); (b) N. N. Lichtin, Progr. Phys. Org. Chem., 1, 75 (1963):(c)E. J. Woodhouse and T. H. Norris, J. horg. Chem., 10, 614 (1971). (IO) 0. F. Zeck and D. W. Carlyle, J. lnorg. Chem., 13, 34(1974). (1 1) J. A. Popie and G. A. Segal, J. Chem. Phys., 43, SI36 (1965);44, 3289 (1966). (12)D. P. Santryand G. A. Segal, J. Chem. Phys., 47, 158 (1967). (13)J. A. Pople and D. L. Beveridge, "Approximate Molecular Orbital Theory," McGraw-Hill, New York, N. Y., 1970. (14) (a) "Tables of Interatomic Distances," Chem. Sbc., Spec. Pub/., No. 11, (1958);(b) P. J. Wheatley, "Handbook of Molecular Dimensions," Academic Press, New York, N. Y., 1972. (15)V. I. Baranovskii, Yu. N. Kukushkln, N. S.Panina, and A. I. Panin, Ross. J. lnorg. Chem., 18(6), 844 (1973). (16)H. Lischka, Th. Plesser, and P. Schuster, Chem. Phys. Lett., 6, 263