SPECTROSCOPIC AND THERMOELECTRIC VAPOR PHASE OSMOMETRIC STUDIES, OF ALKALI AND AMMONIUM ION, SOLVATION. IN NONAQUEOUS SOLVENTS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY MING KEONG WONG 1971 This is to certify that the thesis entitled SPECTROSCOPIC AND THERNOELECTRIC VAPOR PHASE OSHOMETRIC STUDIES OF ALKALI AND AM‘ONIUN ION SOLVATION IN NONAQUEOUS SOLVENTS presented by M1m: Keong Wong has been accepted towards fulfillment of the requirements for Ph. D. degreein Chemistry flew/I 47 WW Major professor Date July 29, 1971 0-7639 ' “E"? & suns I mm mm mc ING BY LIBRWY BINDERS ammonium mum] .._' .__ .-'—_‘-—a-——‘-——-“‘ __—w—-B . -.. .wl.i.'4o.. ,.. . 1h . lbs. .o.\.:.- o -l: (l. $1.... «with gin» :i.1!lorl1t r . INT. J. . --¢A- !_-h“‘ . \r'_ R ‘: ' b..s.L"VU ‘ P'v " x I“ ' Iliad“ The techni £;e::r3520pies 2&2: 201': ined :LIS in aceto In acetor ”2'53 'u’hich CE "6" Or so alien- IEdEpen .2: the halid- ABSTRACT SPECTROSCOPIC AND THERMOELECTRIC VAPOR PHASE OSMOMETRIC STUDIES OF ALKALI AND AMMONIUM ION SOLVATION IN NONAQUEOUS SOLVENTS By Ming Keong Wong The techniques of infrared, Raman, and nuclear magnetic resonance spectrosc0pies as well as thermoelectric vapor phase osmometry have been combined to study the solutions of alkali metal and ammonium salts in acetone, acetic acid, tetrahydrofuran and mixed solvent systems . In acetone, solutions of a given cation show new far infrared bands which can be ascribed to an alkali ion vibrating in a solvent cage. For solutions of lithium salts, the far infrared bands are anion-independent for the polyatomic anions, but anion-dependent for the halides, e.g., the bands occur at 423, 412, and 409 em’1 for the iodide, bromide and chloride, reSpectively. The anions chloride and bromide are shown to have a stronger affinity for the cation than the acetone molecule. The anion-dependent far infrared bands may arise from the vibration of solvated contact ion pair or the vibration of the cation in a cage formed by the anion and near neighbor solvent molecules. The anion-independent far infrared bands arise from the cation vibrating in a solvent cage. In an attempt to determine the lithium ion coordination number in acetone, mole ratio studies were carried out in nitromethane. 0‘; ¢;"?tlbn - n'h :5 La": per-“ .. ; : :Jm 35 a ‘ m5 indicat It acetic ‘4’:qu salts salts used 51:: :i the nature aLeali and th .azis 'iere 0': 13511 i335, 31.1-3“, ace 3 is a pool- lil. ihieh 't ifial ions, Them. “hall 2613 tat in SO: 52517935. uglier in Studied. 1 EUbefits Ming Keong Wong The absorption intensity of the far infrared band, the shifting of the Raman perchlorate band and the nmr proton shift of acetone were followed as a function of the acetone/lithium mole ratio. The results indicate that the lithium ion is solvated by four acetone molecules. In acetic acid, new far infrared bands are also observed for lithium salts and some of the metal fluorides. All of the lithium salts used showed a band at 390 cm-1 whose frequency was independent of the nature of the anion. A band was observed at 280 cm-1 for the alkali and the tetramethylammonium fluorides. No cation-dependent bands were observed at lower frequencies (< 150 cm-1) for the heavy alkali ions. It is concluded that due to its hydrogen-bonding ability, acetic acid solvates the fluoride ion. On the other hand, it is a poor solvating agent for the cations and only the lithium ion, which has a stronger tendency for solvation than other alkali metal ions, shows the solvation band in acetic acid. Thermoelectric vapor phase osmometric measurements of the alkali metal salts in acetone and tetrahydrofuran solutions indicate that in some cases aggregates higher than ion pair exist in these solvents. The degrees of aggregation of the electrolytes are higher in tetrahydrofuran solutions. Among the electrolytes studied, lithium chloride is the most highly aggregated in both solvents. Nmr proton chemical shift studies of solvent mixture systems of acetone, acetic acid, dimethyl sulfoxide and nitromethane indicate thAt dimethyl sulfoxide is the strongest donor solvent, while nitromethane is the weakest. The donor strengths of acetone and acetic acid are approximately equal. P.'~'-An 5“" \II ' ' “_'.o‘|"1' VU‘ "ID '4 SPECTROSCOPIC AND THERMOELECTRIC VAPOR PHASE OSMOMETRIC STUDIES OF ALKALI AND AMMONIUM ION SOLVATION IN NONAQUEOUS SOLVENTS By Ming Keong Wong A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1971 Tee aut': Friessx A1. averages: Appreci I. Seiinney 2 :vercsnin Egezial than :eaiing or t . 9 j~3T'Eative Acknow] Ewing Unix fife this 5‘ 368p a. Ziierstal‘ldi East, but 11 ring Yea“, 53013:”). Gr a hot: 11 ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to Professor Alexander I. Popov for his patient guidance and encouragement throughout this research. Appreciation is extended to Dr. Frank M. D'ltri and Dr. William J. McKinney for their valuable assistance, academic and otherwise, in overcoming the many hurdles that the author encountered. Special thanks are extended to Mr. Paul R. Handy for his critical reading of this manuscript and for numerous enlightening and informative discussions. Acknowledgment is given to the Singapore Government and Nanyang University for granting the author leave of absence which made this study possible. Deep appreciation is extended to my wife, Irene, for her love, understanding and encouragement during the course of this study. Last, but not least, special mention is extended by my daughter, Ting Yean, who always so cheerfully saw "daddy" off to work in the laboratory regardless of whether it was a cold, snowy winter night, or a hot, muggy summer evening. ii ..-~ —A ‘u-uc éfivmzfifl.‘ . .. "'~ 1' " EC u:. U a. L- .4 II. HISTC ACKNOWLEDGMENTS LIST OF TABLES LIST OF FIGURES LIST OF Chapter I. II. III. IV. TABLE OF CONTENTS APPENDICES . . . . . INTRODUCTION . . . HISTORICAL SECTION Spectrosc0pic Studies of Ionic Solvation Acetone . . . Acetic Acid . Tetrahydrofuran . Dimethyl Sulfoxide EXPERIMENTAL SECTION Reagents . Analyses . . . . Preparation of Salt Solutions Instrumental Measurements THERMOELECTRIC VAPOR PRESSURE OSMOMETRY . RESULTS AND DISCUSSION (A) (B) (C) Acetone . . Far Infrared and Raman Spectra . Studies of Acetone Vibrations Effects of Anion . . . . Nuclear Magnetic Resonance Mole Ratio Studies. Vapor Pressure Osmometric Measurements Acetic Acid Dimethyl Sulfoxide . iii Page ii 11 13 l7 18 18 21 22 22 25 30 30 30 36 46 51 56 84 9O \_I II. Cw III. (“C (D) Tetrahydrofuran . . . . . . . . . . . . . 91 Far Infrared Studies . . . . . . . . . . 91 Vapor Pressure Osmometric Measurements . 94 (E) Mixed Solvents . . . . . . . . . . . . . 120 Solvent I. 2.0 M Acetone in Acetic Acid . . . . . . . . . . . . . . . . 120 Solvent II. 2.0 M Acetic Acid in Acetone . . . . . . . . . . . . . . . 121 Solvent III. 2.0 M Dimethyl Sulfoxide in Acetone . . . . . . . . . . . . . 122 Solvent IV. 2.0 M Acetone in Dimethyl Sulfoxide . . . . . . . . . . . . . . 122 Solvent V. 2.0 M Dimethyl Sulfoxide in Acetic Acid . . . . . . . . . . . . Solvent VI. 1.2 M Dimethyl Sulfoxide and 1.2 M Acetone in Nitromethane . . 124 . 123 VI. SUMMARY AND CONCLUSION . . . . . . . . . . . . . 143 VII. SUGGESTIONS FOR FUTURE STUDY . . . . . . . . . . 146 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . 148 APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . 155 iv . Dissceia Soiuticn '. Differe: \I '__4 t‘. a Measure: Dissoci; Acid 50 Frequen Tetrahy Dissocj Tetrah§ Absorp Aceton The Sh vaPOr in Ace Data I BiPhet The A Salts SOIVe ACeti Tetr; in I Data ‘iph Table 10. ll. 12. 13. 14. LIST OF TABLES Dissociation Constants of Electrolytes in Acetone salutions O O O O O I I O O I O 0 Different Kd Values of LiBr and KI from Conductance Measurements . . . . . . . . . . . . . Dissociation Constants of Electrolytes in Acetic Acid Solutions . . . . . . . . . Frequencies of Alkali Ion Vibrations in Tetrahydrofuran Solutions . . . . . . . . . . . Dissociation Constants of Electrolytes in Tetrahydrofuran Solutions . . . . . . . . . . . . . Absorption Bands (cm-1) of Alkali Metal Salts in Acetones . . . . . . . . . . . . The Shifting of the 424 cm_1 and 369 cm‘1 Bands Vapor Phase Osmometry Calibration Data — Benzil in Acetone . . . . . . . . . . Data of Vapor Phase Osmometric Measurements - Biphenyl, Lithium and Sodium Salts . . . . . . . The Apparent Molecular Weights of Alkali Metal Salts in Acetone . . . . . . . . . . . . . . . . . Solvation Bands of the Alkali Metal Salts in Acetic Acid Solutions (cm-l) . . . . Frequencies (cm-1) of Alkali Ion Vibrations in Tetrahydrofurans . . . . . . . . . . Vapor Phase Osmometry Calibration Data - Benzil in Tetrahydrofuran . . . . . . . . . . . Data of Vapor Phase Osmometric Measurements - Biphenyl, Lithium and Sodium Salts . . . . . . Page 12 15 16 31 50 59 6O 76 85 92 96 97 D. (heaical Protons 1 Chemical Protons i Chaiical Sulfoxic Chemical Acetone fiance Acetic . CREE-.1 C a Sulfoxi Effect Shifts Solubi Variou 15. 16. 17. 18. 19. 20. 21. 22. 23. The Apparent Molecular Weights of Alkali Metal Salts . 103 in Tetrahydrofuran . . . . . . . . . . . . . . . . Chemical Shifts (Hz) of Acetone and Acetic Acid Protons in Solvent I . Chemical Shifts (Hz) of Acetone and Acetic Acid Protons in Solvent II . . . . . . . . . . . . Chemical Shifts (Hz) of Acetone and Dimethyl Sulfoxide Protons in Solvent III . Chemical Shifts (Hz) of Dimethyl Sulfoxide and Acetone Protons in Solvent IV . . . . . . Chemical Shifts (Hz) of Dimethyl Sulfoxide and Acetic Acid Protons in Solvent V . . . . . . . Chemical Shifts (Hz) of Protons of Dimethyl Sulfoxide, Acetone and Nitromethane in Solvent VI. Effect of Sodium Tetraphenylborate on the Chemical Shifts (Hz) of Acetone and Nitromethane Protons. Solubilities and Maximum Concentrations (M) of Various Salts in the Solvents Used in this Study . vi . 125 . 128 . 131 . 134 . 137 . 140 . 157 158 Eye 1. Abscrban' aittcmet Z. Spectra methane 3' Spectra methane 1' SPECtr; met-mac. 5' 5?€Ctr metha; 6' Spec t1 nitro' I. The 7 Ch em: aCetc acet SOIV P—J LA! LIST OF FIGURES Figure l. 10. ll. 12. 13. Absorbance of 425 cm-1 band XE: concentration of LiClOQ. Concentration of acetone = 1.5 M. Solvent: nitromethane . . . . . . . . . Spectra of solutions of acetone + LiC104 in nitro- methane. The 425 cm"1 and 390 cm"1 bands. Spectra of solutions of acetone + LiClOA in nitro— methane. The 528 cm"1 band. . . . . . Spectra of solutions of acetone + LiClO4 in nitro- methane. The 1224 cm'1 acetone band . . . . . . Spectra of solutions of acetone + LiClO4 in nitro- methane. The 1712 cm'1 acetone band Spectra of solutions of acetone + 6LiC104 in nitromethane . . . . . The 789 cm.1 acetone and 935 cm-1 C104” Raman bands. Chemical shift of the methyl protons of acetone vs. acetone/LiClO mole ratio. Concentration of acetone = 1.5 M. Concentration of LiC104 varied. Solvent: nitromethane. . . . . . . . . . . Chemical shift of the methyl protons of acetone XE: acetone/LiClO mole ratio. Concentration of LiClO4 = 0.5 M. Concentration of acetone varied. Solvent: nitromethane . . . . . . The AR/Cm vs. Cm calibration plot. Benzil in acetone. . . . . . . . . . . . . . . . The AR/Cg/l gs. Cg/l plots for (A) LiCl. (B) LiSCN. (C) LIN 3 . . . . . . . . . . . . . . . . . . . . The AR/C vs. C 8/1 plots for (A) Biphenyl. (B) L1B§h4.— (C) gNaBPhAH . . . . . . . . . The AR/Cg/l vs. Cg/l plots for (A) NaSCN. (B) Lic104. (C) LiBr . . . . . . . . . . . vii Page 34 38 4O 42 44 47 52 54 57 66 68 7O 72 $1.39 LRI’CE (C) Nal. ' "" “.T“ A. the arga. COECEHIIE E. The appa conceotr The apps concent: 880 The e111 in acet 17. The :‘R/ tetrahy 'Ifi :.. The ;R C. Li} 21. The LR C. Li B. x; 23' The L C-N 24, The - c C0nc< 23' The . COnQ % I u . “Le COUC 27' Chen Pro )9 cf, Che Drc 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. The AR/Cg/l XE: cg/l plots for (A) NaClO4. (B) LiI. (C) NaI 0 o o o o o o o o o o o o o o o o o a o o 74 The apparent molecular weights (Mapp) vs, molar concentration plots . . . 77 The apparent molecular weights (Mapp).1§' molar concentration plots , , , , , , . , 79 The apparent molecular weights (Mapp) XE: molar concentration plots.. . . . . . . . . 81 The effect of water on the solvation band of LiClOA in acetic acid solution . . . . . . . 87 The AR/Cm‘zg. Cm calibration plot. Benzil in tetrahydrofuran . . . . . . . . . . 104 The AR/Cg/l vs. cg/l plots. A. LiCl. B. LiSCN C. L1N03 o o o o o o o o o o . 106 The AR/CEII XE: Cg/l plots. A. NaSCN. B. LiBr. C. LiCl 4. . . . . . . . . 108 The AR/C /1 vs, C /1 plots. A. Biphenyl. B. NaBP 4. C. EiBPh4 . . . . . . . 110 The AR/Cg/l vs, Cg/l plots. A. NaClO4. B. Lil. C0 N81. 0 o o o o o o o o o o o o o o o o o 112 The apparent molecular weights (M ) vs. molar app -—— concentration plots . . . . . . . . . . 114 The apparent molecular weights (Mapp) 3§3molar concentration plots . . . . . . . . . 116 The apparent molecular weights (Mapp) vs, molar concentration plots . . . . . . . . . . . . . . . . 118 Chemical shifts of the acetone and acetic acid protons vs, Acetone/LiC104 mole ratio in Solvent I. 126 Chemical shifts of the acetone and acetic acid protons vs, HOAc/LiClO4 mole ratio in Solvent II. . 129 Chemical shifts of the dimethyl sulfoxide and acetone protons XE: DMSO/LiClO4 mole ratio in Solvent III . . . . . . . . . . . . . . . . . 132 Chemical shifts of the dimethyl sulfoxide and acetone protons gs, Acetone/LiC104 mole ratio in SCI-vent IV 0 o o o o o 0 o o o o o o o o o I 135 Chemical shifts of the dimethyl sulfoxide and acetic acid protons gs, DMSO/LiClO4 mole ratio in Solvent V. . . . . . . . . . . . . . . . . . . . . 138 viii Chezical sulfoxid - 1P1 LILIO‘I‘ D 32. Chemical shifts of the nitromethane, dimethyl sulfoxide and acetone protons vs, (DMSO + Acetone)/ LiClOA mole ratio in Solvent VI. . . . . . . . . . . 141 ix -. I.. .A. . J. x J.bfl.. LIST OF APPENDICES Appendix Page 1. Miscellaneous Observations . . . . . . . . . . . 155 II. Solubilities and Maximum Concentrations of Solutions Used in this Study . . . . . . . . . . 158 The 11:, sautlcns o. ' " ‘iel l: [2L5 Le. ' . in t1 :15 in .a . ZYC-és OI S; equiLLnria solvent pr: ‘5 Q‘llte 1; Eden u... I. INTRODUCTION The importance of ion—solvent and ion-ion interactions in solutions of electrolytes is exemplified by hundreds of publications in this field. Yet at this time these interactions are understood only in rather crude qualitative terms. The knowledge of the various types of species that are present in electrolyte solution, the equilibria between these species as well as the influence of the solvent prOperties on the nature of these species and equilibria is quite limited. When an electrolyte is dissolved in a solvent, four general tYpes of interactions occur: ion-ion, ion-dipole, dipole-dipole, and hydrogen-bonding. These interactions result in the formation of many different types of solvated species. The identification and the characterization of these solvated Species present a major challenge in solution chemistry. The difficulty lies in different- iating among solvated ions, solvated ion pairs and higher aggregates. If the species is a solvated ion, characterization of the Sc>lvation sphere is difficult, e.g., depending on the investigation method, vastly different solvation numbers for the same system haVe been reported in the literature. If the solvated species is an ion pair, discerning the nature of the ion pair itself 18 another major problem. Several different types of ion pairs may be present. For the present purpose, three distinct Species will be considered: (1) contact ion pair, in which an anion and 1 ' are ind “N ‘0. $.eVu rift one solve :i jv'et retain :f Lhasa three fines has bee fears, and to. The detec 131‘ pair in 5 7.5.2:); y 531. utils'll’ely. 31225. in an Tariflus elm 2 cation are in direct contact; (2) solvent—shared ion pair, in which one solvent molecule separates the two ions; (3) solvent- separated ion pair, in which each ion is solvated individually and yet retains an attraction for the other. The characterization of these three different ion pairs by various measurement tech— niques has been pursued by this research group for a number of years, and the present work is part of this larger study. The detection and characterization of aggregates higher than ion pair in solution is also a major problem, eSpecially for low polarity solvents, in which higher aggregates are known to exist extensively. For this purpose, vapor phase osmometry is intro— duced in an attempt to study the degree of aggregation of the various electrolytes in the solvents used in this study. Extensive experimental data for various electrolyte solution systems are needed in order to elucidate the nature of the various species present, and thus be able to postulate their behavior and the equilibria between them. In previous studies conducted in this laboratory, the solvents dialkylsulfoxides (1) and pyrrolidones (2) were examined. These solvents are highly polar and have good donor properties. It is well known that the nature of the sol- vents play an important role in determining the properties of the Electrolyte solutions. However, not much is known on the influence 0f Solvent on the various solvated species and their equilibria. Thus it is of interest to examine the electrolyte solution system in solvents with divergent prOperties. The work presented in this thesis is an extension of the aforementioned studies in solvents with medium and low polarities: acetone. acetic acid, and 3 tetrahydrofuran. A qualitative comparison of the donor strengths of some of the solvents studied thus far is also performed. ice-solvent extensively sto actions are stf ‘3. Qualitative .a. the far 1' iiiues can be II . HISTORICAL SECTION Spectroscopic Studies of Ionic Solvation Ion-solvent interactions in electrolyte solutions have been extensively studied for many decades. Yet until now these inter- actions are still vaguely understood and can be discussed only 111 qualitative terms. During the past decade it became evident tfllat the far infrared, Raman and nuclear magnetic resonance tech- rniques can be very useful in the elucidation of the structure of eilectrolyte solutions (1-27). A more extensive historical dis- cnlssion of solvation studies can be found in the Ph. D. theses of Brian W. Maxey (1) and John L. Wuepper (2) of Michigan State University. In the far infrared measurements of alkali metal and ammonium salts, a band was observed which could not be assigned to the sol- vent or the solute and whose frequency was dependent on the mass 0f tflne cation. In polar solvents such as dimethyl sulfoxide (l) and SL-methyl-Z-pyrrolidone (2) or in solvents with strong donor abilclty, such as pyridine (8), the frequency of the band, with very :few exceptions, was independent of the nature or the mass of the anion. In a nonpolar solvent, such as tetrahydrofuran (9. 10) . the band frequencies were anion dependent. It has been postulélted that in the first case the bands were due to the vi— bration Of the cation in the solvent cage, while in the latter, 4 22':st were or 323;: or the I :ii'Ie Salt. l: are also obsen Inst Kazan s astralytes is :affecting 0- :;ftared solva tactive, indi a2: the solve: :2' sodium tet reaently repc In tecem 25-25 if): the x). x v uh Do tnes' 5 the bands were probably due to either the vibration of an unsolvated ion pair or the cation vibrating in a cage which contains the anion of the salt. Ion pair vibrations of tetraalkylammonium salts were also observed in benzene solutions (11) . Most Raman SpectroscOpic studies in solutions have been carried out on the aqueous electrolyte systems (14—19). A Raman study of electrolytes in methanol showed that the anions are more influential in affecting O-H bonds than are cations (20). Most of the far infrared solvation bands reported thus far were shown to be Raman inactive, indicative of electrostatic bonding between the cation and the solvent. A Raman band at 202 cm-1, ascribed to ion motion of sodium tetrabutylaluminate (NaAlBu4) in cyclohexane has been recently reported (13). In recent years, proton magnetic resonance has been widely used for the study of aqueous and nonaqueous electrolyte solutions. A more extensive discussion of such studies can be found in the Ph. D. thesis of John L. Wuepper (2). Most of these studies involve the measurement of the proton chemical shifts of the solvent. A few Studies have been reported on 7L1 and 23Na chemical shifts in aClueous and nonaqueous solutions of electrolytes (21-25) . Acetone Many Acetone is one of the most common nonaqueous solvents. °rSanic and inorganic compounds are readily soluble in acetone. The solvent has wide liquid rance (—95.4° to 56.2°C) , a moderate dielectric constant of 20.76 at 25° (28), and a Trouton constant of 231-5, indicating that it is a relatively unassociated liquid. The solvent has intermediate solvating power, and according to team's scheme A 39 acetonEIte 3.5 [5 its use in 1,5 sslid diacetc :ia:e:one at 3 Electrical c girlamonium s 1:: pair dissoc 5:0: such meas- sesra‘. studie one the var? ‘13: pair diss iziide in ace aperimental NEIOd 0f da :‘ne salts. The dis EI‘Pectecl f: itctro 1m; ‘1) II n 5 ries 0i 6 Gutmann's scheme (29), it has a donor number of 17. The acetonate of sodium iodide, NaI'BCH3COCH3, is well known due to its use in a method for the purification of acetone (30). The solid diacetonate of lithium bromide has been prepared by Bell, gt_a_1_. (31). It was found to decompose into the unsolvated salt and acetone at 35.5°C. Electrical conductance studies of alkali, ammonium and tetra- alkylannnonium salts in acetone have been plentiful (32-45) . The ion pair dissociation constants for the various salts obtained from such measurements are tabulated in Table 1. In cases where several studies on the same salt have been reported, the agreement among the various values is rather poor. Examples of reported ion pair dissociation constants of lithium bromide and potassium iodide in acetone are shown in Table 2. The discrepancies, besides experimental error, can be attributed to the difference in the method of data treatment, and to the purity of the solvent and the salts. In such cases, the most recent values are listed. The dissociation constants show that in acetone, as would be expected for a solvent with a dielectric constant of 20.76, dissolved electrolytes are mostly undissociated. Adams and Laidler (36) in a Series of papers on mass transport of tetraalkylammonium salts in acetone, have rationalized the results with the assumption of at least two species of ion pairs, "solvent-separated" and "solvent- 8hared" ion pairs. The presence of contact ion pairs was postulated in solvents of lower dielectric constant. Beronius, _e_t_:__a_]_.. (47) studied the ion pair reactivity of lithium bromide in acetone by kinetically measuring the exchange 82 Of Br between the salt and butyl bromide. Their study showed '. 1, 9155 u~a' l .‘U‘ 0 n J' O O .4 . “A ‘.'f'"‘. “Vuofld I 'Q‘ ”I. .'. p-tcmeoes \f-' .K. I'.:.‘ .k. A r310. I 33.“- I! 7 Table l. Dissociation Constants of Electrolytes in Acetone Solutions Compound Temp (°C) Kd X 104 Ref LiCl 25 0.033 33 LiBr 25 2.14 32 Lil 25 69.1 33 LiC104 25 2.0 46 LiPi 25 12.2 37, 40 Li p-toluenesulfonate 25 0.096 33 NaI 25 62.5 34 NaPi 25 14.7 37, 40 NaC104 25 5.44 46 NH4Pi 25 11.1 38 NH4I 25 1.59 46 [(1 25 55.7 33 KPi 25 34.3 37 KCNS 25 34.0 39 KC104 25 1.41 46 Me4NF 25 8.77 37, 40 Me4NPi 25 149.3 38, 40 MeaNI 26 . 6 34 .0 36 Me4NFB (C6H5) 3 25 69 . 3 37 EtANCl 25 27.0 40, 41 Et4NPi 25 222.2 37, 40 Et4NI 26.6 79.0 36 n-Pr4NI 26 . 6 80 .0 36 n—PrANPi 25 370.4 40, 41 n—Bu4NC1 25 23 . 3 35 n-Bu4N8r 25 37.9 37, 40 n-Bu4NI 25 69.9 37. 40 n—Bu4NC104 25 125.0 37, 40 n-Bu4NNO3 25 69.9 37, 40 n-Bu4NPi 25 588.2 37, 40 n-BU4NFB(C6H5) 3 25 19 7 . 0 37 Am4NBr 25 45.5 38, 4O l'l-Bu41*l-p-toluenesul fonate 25 24 . 6 33 Me =- methyl Pr = prOpyl Bu = butyl Am = amyl P1 = picrate KI 8 Table 2. Different K Values of LiBr and K1 from Conductance d Measurements Compound Kd X 104 Investigators Ref LiBr 2.14 Nilsson, Wikander and Beronius 32 2.19 Savedoff 33 4.49 Dippy, Jenkins and Page 43 2.56 Pistoia, Polcaro and Schiavo 44 15.7 Singh and Mishra 45 KI 55.7 Savedoff 33 91.3 Dippy and Hughes 42 80.2 Reynolds and Kraus 37 93.0 Walden, Ulich and Busch 41 186.0 Dippy, Jenkins and Page 43 119.3 Hartley and Hughes 42 297.0 Bauer 42 :2: the lit'niu 12:23:: ion pa: 2.33 x 10“, :::c’u:tan:e me here are 3536151 88115 Base Studies €35 nearly a; nirared and 3"‘1115 o'ota :tt-eeo foe suits in t"! “he 33‘; Ear a 9 that the lithium bromide ion pairs are unreactive species. Assuming contact ion pairs, they calculated the ion pair dissociation constant as 2.33 X 10-4, in close agreement with the value obtained from conductance measurement. There are numerous reports in the literature on the influence of metal salts on the vibrational spectrum of acetone (48-58). Most of these studies were carried out in the 4000 to 400 cm.1 spectral range and nearly all measurements were made in pure acetone as solvent. The primary emphasis in these studies has been on the changes in the infrared and Raman Spectra of acetone upon addition of the salts. The results obtained were interpreted in terms of a complex formation between the cation and the carbonyl group of acetone. The frequency shifts in the acetone bands were found to be independent of anions. The appearance of a band in the 420—430 cm-1 region of lithium perchlorate- acetone solutions was observed by Pullin and Pollock (48) and was attributed to a higher frequency component of the 380 cm‘1 acetone band. Driessen and Groeneveld (56), in their study of solid acetone complexes of alkaline earth and transition metals, also recorded a band at about 420 cm_1 which is cation dependent. They assigned this band to the shifting of the 380 cm'1 acetone band. Yamada (50) also studied the effect of lithium and sodium perchlorates on the n - n* transition of acetone in the ultraviolet Spectral region, and found that the absorption band for this transition is shifted to shorter wavelength upon addition of the two salts. The effect is more pronounced in the case of the lithium perchlorate. These results are interpreted in terms of a charge transfer complex formation between the alkali metal ion and the acetone molecule. Assuming this type of interaction, the author calculated the percent n'leent C131 1 at :225 co"; ’3; the cor-1;". :zar at high in can be e the amber o :easuring t'r. azetone solt lithium PEIC r3389 Studie illutmn the Vapor 1;): rather Spar Rights, H a themflect Vere I'Ound talculated Vapor PIESS I'm conCiuc EffeCts 0f icdide in a calculate d co nCentrati 10 covalent character of the oxygen—lithium and oxygen-sodium bonds to be 132 and 8% reSpectively. In their infrared study of silver and lithium perchlorates in acetone solutions, Pullin and Pollock (48) reported that at lithium perchlorate mole fraction of about 0.3, the uncomplexed acetone bands at 1225 cm"1 and 530 cm"1 disappeared completely and were replaced by the complexed acetone bands at 1239 cm"1 and 540 cm‘l. They prOposed that at high concentration the association of acetone with the metal ion can be expressed by M+(acetone)2. In less concentrated solutions the number of acetone molecules in the complex may be higher. By Ineasuring the ultrasonic velocity in the lithium perchlorate in acetone solutions, Fogg (59) determined the solvation number of lithium perchlorate as varying from 1.2 to 3.4 in the concentration range studied (0.34 to 0.014 M), and indicated that at infinite dilution the solvation number would be about 4. Vapor pressure osmometric studies of metal salts in acetone are rather Sparse. Peska, gt_al, (60) determined the apparent molecular ‘weights, Mapp’ of sodium perchlorate and potassium thiocyanate by tilermoelectric method. The apparent molecular weights for both salts were found to be strongly concentration dependent. The authors by using the osmotic coefficients from calculated Mr from Ma eal PP vapor pressure osmometric measurement and the degrees of association from conductance measurement. Sokolov and Lindberg (61) studied the effects of concentration and temperature on the Mapp of sodium iodide in acetone solutions by osmometric measurements. They also calculated the dissociation constants of the salt at various concentrations and temperatures. Both Mapp and Kd were found to be concentration and temperature dependent. kgov (63): iégits his aha; :e distinctim seems solver. references cit Acetic aci :5 6.2, and a I: is a highly ;I:;er:ies, or 1: the medium ”any stuc; 1: acetic aci Pics: of them SpectreI'vhotor: ti5315513“ On ism in Refe if alkali met Edd solutior Fair diSSocia several Stu d; 5€Sirab1e. From Tab .:r elec “01 11 Acetic Acid Popov (62), in his review chapter on anhydrous acetic acid, Imagine his chapter by stating, "Acetic acid shares with ammonia tflne distinction of being one of the two most investigated non- aaqueous solvents." The statement is amply verified by over 400 references cited in the review chapter. Acetic acid is a protic solvent with a low dielectric constant of 6.2, and a reasonably wide liquid range (from 16.6° to 117.7°C). It is a highly associated liquid, and because of its poor donor properties, only a limited number of inorganic compounds are soluble in the medium (62). Many studies on ionic dissociation of inorganic electrolytes in: acetic acid solutions have been reported in the literature. Most of them are conductance studies. Some potentiometric and spectrOphotometric measurements have also been reported. A detailed discussion on the various studies and the results obtained can be found in Reference (62). The ion pair dissociation constants 0f alkali metal, ammonium and tetraalkylammonium salts in acetic acid solutions are given in Table 3. Again, as in the case of ion pair dissociation constants of electrolytes in acetone, where several studies have been reported, the agreement is less than desirable. IFTom Table 3, it is seen that the ion pair dissociation constants for electrolytes in acetic acid is generally smaller by two to three orders of magnitude than those in acetone. Thus it is to be exPected that the salts will exist in acetic acid solutions pri- marily as ion pairs or higher ionic aggregates. 12 Table 3. Dissociation Constants of Electrolytes in Acetic Acid Solutions CompOund Temp (°C) Kd X 107 Method Ref LiCl 25 0 .83 Potentiometric 67 LiBr 30 7 . 2 Conductance 66 Li (HCOO) 30 0 . 87 Conductance 66 LiOAc 30 6 .0 Conductance 66 LiClO4 25 49 Potentiometric 7O NaBr 30 l . 3 Conductance 66 Na (HCOO) 30 0 . 65 Conductance 66 NaOAc 3O 2 .1 Conductance 66 NaC104 25 33 Potentiometric 67 KCl 25 l . 3 Potentiometric 67 KB}: 30 1 . 1 Conductance 66 K(HCOO) 30 1 . 1 Conductance 66 KOAc 30 3 . 6 Conductance 66 NH40Ac 25 1.00 SpectroscOpic 68 RbOAc 25 l . 28 Spectrosc0pic 68 CsOAc 25 l .66 Spectrosc0pic 68 Et: 4NPi 25 16 . 3 Conduc tance 69 OAc - acetate P1 " picrate Smefcus 37358811 re; 526513395 t acii mlecul‘ he found in I :egc-rted the :agzesium. m. acid as liga' salts (with and acetic a acid molecul studies show :etal ion vi :mplexes a] Site. Cryosco; 2'31}‘Iserized tohounds o; rshined (6 e.g., LIOAQ Tetrahy to 66°C) . If 7.39 at Slightly hi i". “v lutiOns h l3 Numerous spectral studies dealing with the acetate ion as ligand have been reported in the literature. However, relatively very few studies have been done on inorganic complexes containing acetic acid molecules as ligands. A detailed review on the subject can be found in Reference (62). Recently Groeneveld, 953$. (63) reported the preparation of some complexes of the divalent cations magnesium, manganese, cobalt, nickel, c0pper and zinc with acetic acid as ligands. The compounds were prepared from the hydrated salts (with tetrafluoroborate, perchlorate and nitrate anions) and acetic anhydride. The metal ions are coordinated to six acetic acid molecules forming hexacoordinated complex cations. Spectral Studies showed that the acetic acid molecules are bound to the metal ion via the carbonyl oxygen. Earlier studies (62) of other complexes also indicate that the carbonyl oxygen is the coordinating Site. Cryoscopic studies (64,65) showed that lithium salts are pelymerized in acetic acid solutions. Isolations of some addition compounds of alkali metal salts with acetic acid have also been reported (62). The compounds have either 1:1 or 1:2 stoichiometry, e . g . , LiOAc °HOAc , KOAc -HOAc , LiCl ° ZHOAc and Lil - ZHOAc . Tetrahydrofuran Tetrahydrofuran is a solvent with moderate liquid range (—65° to 66°C). It is relatively non-polar, with a dielectric constant 013 7.39 at 25° (71). The donor number of the solvent is 20 (29), 8lightly higher than that of acetone. Several SpectroscOpic solvation studies in tetrahydrofuran solutions have been reported in recent. years (9,10,12,13,26,27,72,76). :::ntr85t t 5:5:xiie (1) 22315305“! :resence of i frequencies a Day and c 3:131:25. 1 res::ance (2'. cf 1:1 and l: tizns cf the intession i: There 81‘ finance Stu (71:74) dete 3": a Series mm the dat Sflis are t} 5110mm Sa] than the SO( 531VatiOn oi Stants ob ta‘ m°ride (7' Edgell, Scintions 0 CO a 14 In contrast to the alkali ion solvation bands observed in dimethyl sulfoxide (l) and pyrrolidones (2), the alkali ion vibrations in tetrahydrofuran solutions are strongly anion dependent, indicating presence of ion pairs or higher ionic aggregates. The reported frequencies are tabulated in Table 4. Day and co-workers studied the complexation between sodium tetrabutylaluminate (NaAlBuq) and tetrahydrofuran in cyclohexane solutions. In their more ratio studies, using both proton magnetic resonance (27) and infrared (76) techniques, they showed the formation of 1:1 and 1:4 complexes, depending on the respective concentra- tions of the two interacting species. The over-all equilibrium expression is: Na'THF+, AlBua‘ + 3 THF 2 Na'4THF+, AlBu4' There are only a few reports in the literature (71-75) on con- ductance studies in tetrahydrofuran solutions. Szwarc, §_t_a_l_. (71,74) determined the conductances and the dissociation constants Of a series of tetraphenylborates in tetrahydrofuran solutions. From the data obtained, they concluded that the lithium and sodium salts are the most dissociated, while the cesium and tetraalkyl— ammonium salts, in Spite of their bulkiness, are less dissociated than the sodium or lithium salts, indicating weaker or lack of solvation of the bulkier cations. The ion pair dissociation con— Stants obtained by Szwarc, 3531., together with that of lithium chloride (75), are listed in Table 5. Edgell, _e_t__a_l_. (72) measured the conductances of tetrahydrofuran solutions of sodium tetracarbonylcobaltate (NaCo(CO)4) at various °°“°—entrations. In the equivalent conductance E- square root of 22mm .4“ quJJ A ”:3 nude.L 7.;Co(CC‘, 15 Table 4. Frequencies of Alkali Ion Vibrations in Tetrahydrofuran Solutions Compound v(cm-l) Ref LiCl 387 10 LiBr 378 10 Lil 373 10 LiNO3 407 10 LiBPh4 412 10 LiCo(C0)4 413 10 NaI 184 10 NaBPh4 198 10 NaCo(C0)4 192 10 NaAlBu4 195 13 KAlBu4 150 13 KCo(CO) 4 142 10 ’— fipflfllyu \“JJV Mn 16 Table 5. Dissociation Constants of Electrolytes in Tetrahydrofuran Solutions Compound Temp (°C) Kd X 105 Ref LiCl 20 0.000012 75 LiBPh4 25 7.96 74 NaBPh4 25 8.52 74 KBPh4 25 3.22 74 CsBPh4 25 0.187 74 BuANPhé 25 4.32 74 Bu(isoamyl)3NPh4 25 6.04 74 Ph - phenyl Bu 8 butyl .«g‘ y- .- ...4- ct. LT“ I: (n (_) (I .o : l ‘7 .a. ] FAF'!‘ cy5“ b ‘- (f) L) *1 rt. 'U (J 3', :3: 17 concentration plot, a deep minimum was evident at low concentration {/5 x 0.1 M) and there was an inflection point at higher concentration (/C R 0.3 M). They postulated the formation of ion pairs in dilute tetrahydrofuran solutions, and the formation of ionic clusters, e.g., triplets and quadruplets, in the more concentrated solutions. Dimethyl Sulfoxide Dimethyl sulfoxide has a relatively high dielectric constant of 46.6 (77), and a broad liquid range of l8.4-189°C (78). It is a good donor solvent, with a donor number of 30 (29). A detailed discussion of solvation studies in dimethyl sulfoxide can be found in Reference (1). In dimethyl sulfoxide, the frequencies of the far infrared solvation bands are independent of anions. Bands due to the cations lithium, ammonium, sodium, potassium, rubidium, and cesium 1 9 occur at 430 cm-1, 214 cm-1, 200 cm-1, 154 cm_l, 125 cm_l, and 110 cm— respectively. The solvation numbers, as determined by proton magnetic resonance mole ratio studies, are 2 for lithium ion (1) and 6 for sodium ion (2). Conductance studies (79-85) in dimethyl sulfoxide solutions reported in the literature indicate that, in dilute solutions, with few exceptions, electrolytes are generally completely dissociated. sti‘IEd 0‘ :istille: calm. OI ’ w '9/7 JV“ ll .4... ‘ ' a” ”I5 “‘2‘“ C III. EXPERIMENTAL SECTION Reagents Acetone: Matheson Coleman & Bell reagent grade acetone was stored over calcium sulfate several days and then fractionally distilled from fresh calcium sulfate through a one meter packed column. The middle fraction boiling at 54°C at 737 mm (lit. (31): 55-1°/760 mm) was collected. The water content was approximately 0.022. Acetone-d6: Diaprep Inc., acetone-d6 with a minimum isotOpic _¥ purity of 99.5% was used without further treatment. Acetic Acid: Fisher reagent grade 99.7% acetic acid was Purified by addition of approximately 17. (by volume) of acetic anhydride, the mixture was refluxed for several hours and then fractionally distilled through a one meter packed column. The purified acetic acid melted at 16.6 : 0.2°C (lit. (62): m.p. = 16.63S°C). The water content was approximately 0.01%. Acetic Acid-d4: Diaprep Inc., acetic acid-d with a minimum 4 isotopic purity of 99.57. was used without further purification. Dimethyl Sulfoxide: Baker reagent grade dimethyl sulfoxide Was Purified by refluxing under high vacuum (4 mm or less) for Several hours over barium oxide, and then fractionally distilled through a 30 cm packed column. Throughout the vacuum distillation, t he temperature was maintained at approximately 45°C. The middle 18 ISZIEHE 1 l9 fraction was collected and subjected to fractional crystallization three times. The purified product melted at 18.3 j; 0.2°C. The best literature value of 18.45° (78) can be achieved only after The water very elaborate and tedious purification procedures. content was approximately 0.017.. Nitromethane: Matheson, Coleman and Bell practical grade nitromethane was purified by percolation through a 20-cm column of Dowex 50W-X8 cation-exchange resin (86) . The eluent was refluxed over anhydrous calcium Sulfate for several hours and then fractionally distilled through a one meter packed column. The middle fraction boiling at 99.5°C at 738 mm was collected (lit. (87): b-P- a 101.25°). The water content was approximately 0.1%. 760 Etrahydrofuran: Matheson, Coleman and Bell reagent grade tatratlydrofuran, with water content of approximately 0.027., was used withou t further pur ificat ion . T£.=—t:rahydrofuran-d8: Norell Chemical Company, Inc., tetrahydrofuran- d8 With a minimum isotOpic purity of 997. was used without further treatment. Eighium Thiocyanate: Lithium thiocyanate from City Chemical corporation, N.Y., could not be dried by simple heating. The salt dissolved in its water of hydration when the temperature was raised to about 40°c, Anhydrous lithium thiocyanate was prepared (88) by firSt dissolving the salt in anhydrous ether, then adding about 2 values of petroleum ether. If sufficient water were present in the compound, two immiscible liquid phases would form. The aqueous Phase at the bottom was removed by decantation. An etherate precipitate, LiSCN°(C21-l5) 20, crystallized upon cooling and stirring. The e'iherate crystals were filtered with minimum exposure to moist ;:v' r38 ether f.” (11 day) ‘ ' b . 1‘ ’Ali'A ., , 3. Li: a?" :58 undel gang point ‘ “my conten‘ as found to 0 resulted in h)’ tater content Tze salt was d atetczate by c :tystals were :35 then repee inc-spherus per 'v'as then raist Lithium iodid lithium T 5? metathesis in tatrahydrc “3*T3Ptteny lbc :‘i ~ 1 . “Petated o ‘59 residue 20 air. The ether in the etherate was removed under high vacuum at 30°C ('bl day). Then the temperature in the vacuum oven was raised gradually. Lithium thiocyanate was finally dried over phosphorus pentoxide under vacuum at 90°C for 2 days. The dried product has a melting point range of 279—282°C (Lit. (88): m.p. = 281°), and a water content of 0.47.. Lithium Iodide: Anhydrous lithium iodide from K & K Laboratories was found to contain over 3% water. Heating the salt in vacuum resulted in hydrolysis and decomposition (89). However, the water content can be reduced by recrystallization from acetone (90). The salt was dissolved in purified acetone and precipitated as the acetonate by cooling the solution in a dry ice bath. The acetonate crystals were filtered with minimum exposure to air. The procedure was then repeated. The acetonate crystals were then heated over Phosphorus pentoxide under vacuum at 35°C for 1 day. The temperature W38 then raised gradually until final drying at 85°C for 3 days. Lithium iodide thus obtained had a water content of 0.67.. Lithium Tetraphenylborate: Lithium tetraphenylborate was prepared by metathesis of excess lithium chloride and sodium tetraphenylborate in tetrahydrofuran solution (74) . Lithium chloride and sodium tetraphenylborate in the mole ratio of approximately 2:1 were dissolved separately in tetrahydrofuran. After mixing the solutions, the less soluble sodium chloride was filtered out, and the tetrahydrofuran evaporated off by blowing a stream of nitrogen over the solution. The Irasidue was dissolved in 1,2-dichloroethane and the solution was filtered to remove excess lithium chloride. Lithium tetraphenylborate was then precipitated by the addition of cyclohexane. After f 11traItion, the procedure was repeated twice. The salt was dried '.fl mu! " * ..e. 3... ..vl ufiflIQ ,..¢...-I a! A¢¢1 'n A I .- 47" A: . . Jr . p. _— " I'- n n n q a'WT' Mitts. Ernie a “was“: . ‘l \"‘-.. "-c‘; . . n4 m he 'u. to use [‘0 r_- fl! Earl 3:1" ‘t I 21 over phosphorus pentoxide under vacuum at 60°C for 3 days. Flame emission analysis showed the sodium content was about 0.05% and the lithium content 2.0 i 0.2% (theoretical: 2.15%). The water content was approximately 0.4%. Other Alkali Metal Salts: All other alkali metal salts, with the exceptions of sodium thiocyanate, sodium tetraphenylborate, ammonium thiocyanate, potassium thiocyanate, were reagent grade chemicals and were used after drying at 180-200°C for 2-3 days. Sodium thiocyanate, sodium tetraphenylborate, ammonium thiocyanate and potassium thiocyanate were vacuum dried at 60°C for 3 days prior °Li salts has been described previously (91). to use. The preparation of Tetraalkylammonium Salts: Tetra-n-propylammonium bromide, tetra-n—butylammonium bromide, tetra-n-butylammonium iodide from Eastman Kodak, tetramethylammonium fluoride from Aldrich Chemical Co. Inc., and tetra-n—butylammonium perchlorate from K & K Laboratories, all were vacuum dried at 60°C for 3 days prior to use. Benzil : Eastman Kodak reagent grade benzil was recrystallized twice from absolute ethanol and then vacuum dried at 50°C for 2 days. The purified product melted at 96 :_0.5°C (92). Biphenyl: Eastman Kodak reagent grade biphenyl was used without further treatment. Analyses All quantitative determinations of water were carried out by the Karl Fischer procedure (93). Flame emission analyses of sodium and lithium were performed by using a Beckman Model B Spectrophotometer with flame photometry attachment. .I I. . I. . i. . .: . “Pb .~Lv .Av. HmN an ”In ‘N . f f ‘ .‘o .u o v . w“ .w. ..l .I‘ I.» t up. . .3 . . 22 The isotOpic purity of all deuterated solvents was checked by mass Spectrometry, a service provided by this Laboratory. Preparation of Salt Solutions All the salt solutions were prepared on a molar basis. In mole ratio and mixed solvents studies, several stock solutions of various concentrations of one solvent in another were first prepared, and solutions of intermediate or lower concentrations were made by dilution. In vapor pressure osmometry studies, salt solutions of the highest concentrations were first prepared, and subsequent series of solutions were prepared by dilution. Since most of the solvents and salts employed in these studies were hygroscopic, care was taken to prepare them in as nearly an anhydrous condition as possible. During preparation or transfer, exposure of the solvents and solutions to the air was minimized by making all transfers with a syringe or pipet. Measurements were made immediately after preparation of solutions. Solutions were prepared at room temperature of about 24°C. Instrumental MeaSurements Far Infrared Measurements: Most of the far infrared spectra (from 600-80 cm-l) were obtained with the Perkin Elmer 301 Far Infrared Spectrometer. Some spectra were obtained with the Digilab FTS-l6 Interferometer. The Perkin Elmer 301 instrument was operated in the double beam mode, and single beam mode was used in the operation of the Digilab FTS-16 Interferometer. Nitrogen was used to purge the instruments of water vapor prior and throughout the In c: {$111192 .- e: previ b '9 o. v. "I A .C .< \ .a 1.- Ta . s .1 .. r ”a s .. m R \ I o A. M ,0 t P Q .1 . u M»: E u. m e t t}. C U m a a c u .c c e u e we. a z 1 I QC RI... . . NM :4“. Na.“ 9.. .1. «In 5; HM My». 11. at HQ 3 S Qt . a. ..V. {We .ml. find «L n :- pH... t was 7.. r 23 measurements. Polyethylene cells with 0.1 or 0.2 mm pathlengths, purchased from Barnes Engineering Co., were used. The frequency scale of the Perkin Elmer 301 Spectrometer was calibrated with water vapor. Detailed calibration procedure has been previously described (2). In cases where duplicate measurements were made from the two instruments, the solvation band frequencies always agreed to :3 cm-l, which is within experimental uncertainty. Infrared Measurements: All infrared Spectra in the 4000-600 cm-l region were obtained on a Perkin Elmer Model 225 Spectrometer. Nitrogen was used to purge the instrument of water vapor throughout the measurements. Standard demountable liquid cells, with KBr windows and teflon Spacers giving pathlengths from 0.025 to 0.1 mm, all obtained from Barnes Engineering Co., were used. Filling of the liquid cells was done by using small volume syringes to insure uniformity and minimize sample exposure to the atmosphere. Raman Measurements: Raman spectra were obtained using a Spectra-Physics Model 700 Raman Spectrometer, with a 40 mw O 6328 A He-Ne laser. The frequency scale was calibrated with carbon tetrachloride Raman lines. Nuclear Magnetic Resonance Measurements: Varian Associates A56/60D Spectrometer was used to obtain all nuclear magnetic resonance spectra. Tetramethylsilane was used as an internal standard for all samples. Vapor PreSSure Osmometric Measurements: Vapor pressure osmometry studies were performed using a Mechrolab Inc., High Temperature Vapor Pressure Osmometer Model 302. The method outlined in the Operational n“ y . .— .k kl! 1" 24 manual was used (94). The sample chamber was maintained at a constant temperature of 37°C, and at least 4 hours were allowed for it to equilibrate and saturate with solvent vapor prior to measure- ments. Benzil was used as calibration standard for all the solvents. The concentrations ranged from 0.02 to 0.2 M for benzil, and 0.02 to 0.16 for other salts. Biphenyl was used as an additional check on the calibration constants obtained from benzil. In all the measurements, the solvent and sample dr0ps at the thermister beads were kept at approximately the same size. The resistance AR values were recorded 2 minutes (accurate to within 5 seconds) after the sample dr0pS were deposited to the thermister bead. Approximately 15-20 minutes prewarm periods were allowed for each solution. The solvent in the solvent cup (used to saturate the chamber) was replaced after every two or three series of runs. At least two readings of AR were taken for each solution and the averaged value was used for calculations. an ”I 21 :“ H ubo' . fi'fi‘ u.u., .IU 5L IV. THERMOELECTRIC VAPOR PRESSURE OSMOMETRY Thermometric vapor pressure osmometry was first introduced by Hill (95). The determination of the molecular weight of a substance by the method of vapor pressure osmometry involves the following steps. An enclosed chamber is thermostated and saturated with solvent vapor. A drOp of the solution of the substance studied is placed on one of the two matched thermistor beads and a drop of solvent on the other. Since the vapor pressure of the solution drOp is lower than that of the solvent, solvent vapor will condense on it, the heat of condensation will raise its temperature, and a steady state is soon reached. The steady state temperature difference between the two drops is proportional to the vapor pressure difference. The change in resistance, AR, generated by the temperature difference AT, is measured by a Wheatstone bridge. The value AR is used to calculate the molecular weight of the unknown substance by methods described below. When this technique was first introduced, thermOpiles were used to measure the temperature difference, and the method was used exclusively for aqueous systems. Subsequently, the Speed and the sensitivity of the measurements were greatly improved by using thermistors, and the method was adapted to organic solvents (96-100). The theory and the limitations of the method has been discussed by several authors (92,100-106). Presently, matched thermistors are almost exclusively used for measuring the temperature difference, 25 26 and besides the determination of molecular weights, an increasing use of vapor pressure osmometry for the determination of osmotic and activity coefficients of salts in nonaqueous solvents has been reported (101,107-111). There are also several reports on the use of vapor pressure osmometry for the study of association or aggregations of salts in nonaqueous solvents (112—115). In this study, the Mechrolab Vapor Pressure Osmometer Model 302 was used (see Chapter II). It is stated in the Mechrolab Instruction Manual (94) that the sample and solvent drOp size variations have no appreciable effect on the value of molecular weight determined, except in the case of solvents with high surface tensions. The statement has been disputed by Meeks and Goldfarb (116), who studied the quantitative effects of the drop size and the time variations on the value of the measured molecular weight. They found that the vapor pressure osmometer readings AR (and hence, the molecular weight) are strongly dependent on drOp size and time variations, and prOposed methods for correcting both of these effects. The molecular weights as determined by vapor pressure osmometry are the so-called number-average molecular weight. The term "number— average" indicates that every molecule present, regardless of its size and mass, gives rise to the same response. In order for this to be achieved, the system must behave ideally, i.e., the force of interaction (solute-solute, solute-solvent, solvent-solvent) should be the same for all molecules present in solution. In general, near ideal conditions are reached in the limit at infinite dilution and the response will be prOportional to the number of molecules present. p‘- v'. n~¢ but f n tn I'M- '5‘ ‘. i L] 27 However, departures from the ideal behavior of the systems can still result from association or dissociation of the solute or solvent. There are two steps involved in the determination of molecular weight of the substance studied: (1) calibration of the instrument, and (ii) measurement of the unknown. For each solvent, a calibration constant K for the instrument used has to be found. The calibration constant is determined by use of a standard substance of known molecular weight which gives near ideal solution at low concentrations. Several methods have been proposed for the determination of the calibration constant and the molecular weights. Only the two methods used in this study will be discussed. (I) The limiting slopemethod (117) (or "constant K" method by the Mechrolab Manual): For small temperature changes observed in this work, the results can be represented by the equation _ 2 3 AR - a0 + alC + a2C + a3C . . . (l) where AR is the osmometer dekastat reading observed at a solution concentration C. The coefficient a0 represents the zero point displacement and normally should be zero. The coefficient a1 can be expressed as K/Mn where K is the calibration constant and Mn is the number-average molecular weight. Thus this is the coefficient of interest. The term a C3 is normally small enough to be negligible. 3 The term a2C2 is usually large and cannot be neglected. However, the coefficient a2 is not required for calculating apparatus constants or molecular weights. It is required to evaluate the data and in error considerations (117). Hence for general cases, ch‘ p nuv i.» g... '5 .5. 55. '0 28 equation (1) is reduced to 2 AR - alC + aZC . . (2) Equation (2) can be rewritten as AR/C = a1 + aZC . . (3) Therefore, a plot of AR/C vs. C should give a straight line with intercept a and slope a In cases where scattering or nonlinearity 1 2' of points at low concentrations are encountered, the higher concentration points are emphasized. Thus (AR/C)C=O = a = K/Mn 1 If molar concentration is used for the calibration plot, then (AR/C)C=O = Km in units of ohms liters/mole. For a molecular weight determination, similar AR/C vs. C plots are obtained. However, C is expressed in g/liter, so that the molecular weight is calculated from the expression, Mn = Km % (AR/C)C=O (C in unit g/l) (II) The "variable K" method: In this method, the calibration graph is obtained by plotting AR/Cm vs. AR. The AR value of each concentration of sample is entered into the calibration plot to find AR/Cm, from which a series of apparent Cm values can be calculated. The reciprocal apparent molecular weight of each solution is obtained by dividing each Cm value by weight concentration of the solution. A plot of the reciprocal of apparent molecular weight (l/Mapp) vs. weight concentration is extrapolated to zero concentration to obtain the l/Mapp value at zero concentration. Reciprocal of (1/Map is the number-average molecular weight. p)C=0 29 In the present study, no attempt is made to extrapolate to zero concentration and find the number-average molecular weight by method (II). Instead the apparent molecular weight for each concentration is calculated and then plotted vs. the actual concentration (from weight of solute used in preparing the solution) so as to elucidate the magnitude of the concentration dependence of the apparent molecular weight of each electrolyte. V. RESULTS AND DISCUSSION (A) Acetone Far Infrared and Raman Spectra Preliminary investigations in the 600-80 cm‘l infrared region indicated that some spectral windows are available in acetone. The solvent has strong fundamental bands at 528 cm’l, 390 cm'l, and a weak band at 495 cm-1. It has a narrow spectral window between 480 and 410 cm-1, and is fairly tranSparent in the 360 to 150 cm“1 spectral region. Solutions of common lithium and sodium salts were studied in the far infrared region and new, broad absorption bands were observed. The new bands could not be ascribed to the solvent or to the anions of the salts. The exact positions of the band maxima were obtained by subtracting out the solvent absorption. The frequencies of the new bands are cation dependent, e.g., 425 cm‘1 for lithium perchlorate and 195 cm"1 for sodium perchlorate in acetone solutions. Thus the new bands were assigned to the cation (or salt) - acetone vibrations. A summary of the data obtained is given in Table 6. It is seen from the data that for lithium salts with polyatomic anions as well as for the iodide, the band frequencies do not depend on the mass or the nature of the anion. There is, however, a decrease in the frequency of the observed band for the bromide (412 cm_1) and the chloride (409 cm'l), with the chloride having the lowest band frequency. Anion dependence of the bands was also observed in Spectra 30 . .3 U. 5‘ ‘1 \_d b V. «4‘ Q“ n\ r Y.- . L i . t-J 31 Table 6. Absorption Bands (cm'l) of Alkali Metal Salts in Acetones Salt Acetone Acetone-d6 L1c104 425 :_3 390 :_4 LiSCN 425 390 LiBPhA 426 390 L1No3 420 387 LiCl 409 376 LiBr 412 378 Lil 423 389 NHASCN 212 :,4 NaSCN 196 192 :_s NaC104 195 191 NaBPh4 196 190 NaI 192 186 KSCN 154 6L1c1o4 436 :.3 61.11103 434 °LiC1 412 °LiBr 420 6 Lil 436 U. 9‘. .‘u AI» ". 1‘ .s o . . :1 II. n: ‘5 32 of lithium salts in acetone-d6 solutions, where lithium salts with polyatomic anions, as well as the iodide, show bands at about 390 cm-1, while the bromide and chloride salts give bands at 378 cm.1 and 376 cm'l, reSpectively. The large shift of about 35 cm"1 with deuterated solvent is unexpected since a Hook's law calculation, based on a pseudo-diatomic situation in which the acetone and acetone-d6 molecules are considered as point masses in the same manner as the lithium atoms, predicts a shift of only two cm-l. A similar Hook's law calculation of the effect of substitution of 6L1 salts in place of the naturally occuring 7Li salts predicts a shift of 30 cm'l. Experimentally the shift is 1 except in the case of the lithium found to be approximately 10 cm- chloride salts where the shift is only 3 cm—l. No valid explanation can be given for this discrepancy at this time. It seems reasonable to assume that the vibrating Species is complex, involving the cation, the solvent molecules and in some cases (especially the bromide and chloride) the anions. The anion dependence of the lithium bromide and chloride band frequencies is similar to that reported by Edgell, g£_al, (10), for lithium salts in tetrahydrofuran solutions. However, since the anion dependence of the solvation band does not follow the mass dependence which would be expected if an ion pair vibration (such as was observed by Evans and Lo (11) for some tetraalkylammonium salts in benzene) was being detected, it is quite possible that the anion is acting as a perturbing influence on the cation-solvent vibration. The very small shift (about 3 cm'l) observed for lithium 6 chloride with isotopic substitution of Li may indicate the existence of a contact ion pair - solvent vibration. 8‘ r at 15: i 4 sciatica. Weak band the USual band at a] 33 l was also observed for all the A shoulder at about 365 cm— 1ithium salts in acetone solutions. The shoulder, as will be shown below, is a band which is anion dependent in some cases. It is also dependent on the mass of the cation. In general at least 0.1 M solutions are required to yield observable Spectra. Among the common potassium and ammonium salts only the thiocyanates were found to be sufficiently soluble. A band at 154 :;4 cm"1 was observed for potassium thiocyanate in acetone solution. Ammonium thiocyanate gave a band at 212 :_4 cm—1 and a weak band at about 350 cm-1. The band at 212 cm”1 has the shape of the usual solvent-cation vibrational band. The origin of the weak band at about 350 cm”1 is uncertain. The solubilities of the more common rubidium and cesium salts (MCI, MBr, MI, MNO3, MC104, MSCN, MBPh4) in acetone were too low (less than 0.1 M) to allow meaningful infrared Spectral measurements. A concentration range of between 0.1 and 1.5 M was used to study the lithium and sodium salts. The intensities of the observed cation-solvent bands were proportional to the concentrations in this range. High absorption and high baseline at concentrations higher than about 1.5 M limited further comparison. The intensity of the 425 cm.1 band was also measured as a function of lithium perchlorate concentration in nitromethane solutions which were 1.5 M in acetone. The absorbance at peak maximum of 425 cm"1 band was plotted gs, molar concentration of lithium perchlorate (Figure 1). A straight line was obtained for solutions which were §_0.4 M lithium perchlorate. At higher concentrations, while the plot was still linear, the slope was considerably different. At the same time, while the frequency of the band is not changed 34 Figure 1. Absorbance of 425 cm-1 band XE: concentration of LiC104. Concentration of acetone = 1.5 M. Solvent: nitromethane. A. Without subtraction of baseline. B. With the baseline subtracted. 35 r, '1 LOI- '1 O ’0 O .8 a ’0 0 0.6- 3 g, o '9 9. .n < ’0 Ar 0; 0 C; .2' 0 0 .14 .s 1.2 AA [JC:“DZ Figure l. pa! and .vv 4.. .6 36 appreciably, the band begins to broaden. The broadening may indicate that a new absorption band is formed whose frequency does not differ appreciably from that of the original band. The data, therefore, indicate that below 4:1 mole ratio of acetone/Li+, a new absorbing species is formed. The change in absorbance and the band broadening may be due to a change in the solvation number of the lithium ion or to a replacement of either one or more acetone molecules in the solvation shell by the perchlorate ion (see below). Raman Spectra in the 3000—150 cm‘l spectral region were obtained of the lithium nitrate, bromide, and perchlorate and of sodium perchlorate solutions in acetone and in acetone—nitromethane mixtures in the 0.6-3.0 M concentration range. the 425 cm"1 Li+-acetone and 195 cm.1 Na+-acetone bands were not observed. It seems that the bands are Raman-inactive and, therefore, the cation-solvent bonding is essentially electrostatic as postulated by Edgell. gt_al. (9,10). Studies of Acetone Vibrations There are many reports in the literature discussing the influence of metal salts on the vibrational Spectrum of acetone (48-58). These studies were carried out in the 4000-400 cm.1 range and all measurements were made in pure acetone as solvent. The shifting or splitting of acetone fundamental bands were ascribed to interaction or complex formation between the cation and the carbonyl oxygen of the solvent molecule. However, in pure acetone, the fundamental vibration bands are rather intense, and the observation of the shifting or Splitting of those bands relative to salt concentrations is difficult, if not impossible. ... -' Jo-rvh t a" irfl‘u ‘\¢~- In~ D '7 out... ‘h-na ......‘_ . ‘- ‘A-‘ (1" - p ‘ b. a“ or ' iJ (I7 37 In this study the behavior of the acetone vibrations as a function of acetone/Li+ mole ratio was observed. It was necessary, therefore, to use an inert solvent as diluent. The difficulties of selecting inert solvents for such studies have been discussed by Wuepper (2). In the present case it was found possible to use nitromethane as the diluent. While it cannot be Stated that nitromethane is devoid of all solvating ability, its relative inertness is well known (29) and it is quite unlikely that it can compete significantly with acetone in the solvation of lithium ions. A series of nitromethane solutions was prepared which were 1.5 M in acetone and contained varying amounts of lithium perchlorate. The changes in the frequencies of four Strong acetone vibrational bands at 390, 528, 1224, and 1712 cm.1 were followed as a function of acetone/LiClO4 mole ratio. The 390 (C-C-C deformation), 528 (C-C-O bending) and 1224 cm—1 (C-C asymmetric stretch) bands were Split, with new bands appearing at 369, 539 and 1239 cm'l. These new bands can be attributed to vibrations of the complexed acetone. The relative intensities of the various bands as a function of acetone/LiClO4 mole ratios are shown in Figures 2,3, and 4. No splitting was observed for the 1712 cm.1 C=O stretch, but the band was shifted progressively to lower wave number with decreasing acetone/LiC104 mole ratio, as shown in Figure 5. Similar shifting and Splitting of the acetone fundamental frequencies were observed with lithium iodide, indicating that the lithium ion is the Species that interacts with acetone. 1 The intensity of the 390 cm- acetone band decreases gradually with decreasing acetone/LiClO4 mole ratio, and a new band at 38 Figure 2. Spectra of solutions of acetone + LiClO4 in nitromethane. The 425 cm"1 and 390 cm.1 bands. A. Blank = 1.5 M acetone B. Acetone/LiClO4 ratio = 8:1 C. Acetone/LiClO4 ratio = 4:1 D. Acetone/LiClO ratio = 2:1 4 39 A 70" I \ ,’ I, \\ ’B / \/"\ I, / \\ ,,’ / / 50*- / r-~ // /’ “\ / 440 400 360 c m ‘1 N Figure Figure 3. Spectra of solutions of acetone + LiClO 40 The 528 cm-1 band. Blank = 1.5 M acetone Acetone/LiClO4 ratio = 8:1 Acetone/LiClO4 ratio = 4:1 = 2:1 Acetone/LiClO4 ratio 4 in nitromethane. 41 80* 60- :hate. 20- 550 530 510 0 m"‘ 570 Figure 3. 42 Figure 4. Spectra of solutions of acetone + LiClO4 in nitromethane. The 1224 cm.1 acetone band. A. Blank = 1.5 M acetone B. Acetone/LiClO4 ratio = 10:1 C. Acetone/LiClO4 ratio = 4:1 D. Acetone/LiC104 ratio = 1.7:1 43 80' 60' 4o- 20” 1230 1210 c m-‘ 1250 Figure 4. 44 Figure 5. Spectra of solutions of acetone + LiClOA in nitromethane. The 1712 cm"1 acetone band. A. Blank = 1.5 M acetone B. Acetone/LiC104 ratio 10:1 II b C. Acetone/LiClO4 ratio :1 45 I..- III ’ lllll I '''''' I '-'- I 0’ I 7" I l." I '0 I " I UUUUU I ’ I! 'l I I I I I I \\ ‘\\ \B“ “ \II \ \‘tu I\ \‘\ \ \‘ \ ‘5 \ \ ‘ I“ 5‘ \ ‘ \ ‘ \ \ \ “\ \ ‘\ \‘I \ \ “ \ “ \ \ \ \ \ \ \ ‘\\ \.\ 1690 1710 1730 80‘ 60- 4o- 2m» h on c m—‘ Figure 5 . :e Splitti bands were 321::‘1 at was observ: iecreased. 3513+3cm Figure 6. The a for the 11 5331613 Stu if the So] has been I Exact i( It» be for} an intern. hams fOr he “an, that of l tetraphen the catio lt’u‘er fre‘ and then 1 Era - Rude 3. 46 369 :_3 cm.1 appears. This new band, in contrast to the 1239 and 539 crn-l complexed acetone bands, was observed to be dependent on the mass of the cation. When 6LiClO4 was substituted for 7LiClO4, the splittings and shiftings of the 1224, 528 and 1712 cm'1 acetone bands were identical with those observed with 7LiClO4. However, the 390 cm-1 acetone band was shifted gradually (no apparent Splitting was observed) to lower frequency as the acetone/Li+ mole ratio was decreased. The band appeared to reach a limiting value of 380 :_3 cm-1 at acetone/6LiClO4 mole ratio of 4:1, as shown in Figure 6. Effects of Anion The anion dependence of some of the solvation bands, notably for the lithium halides, seems to indicate that at least in the system studied, the halides (bromide and chloride) may form a part of the solvation sphere. From electrical conductance Studies, it has been postulated (33) that lithium chloride forms unsolvated contact ion pairs, while with lithium iodide the ion pair was said to be formed from fully solvated ions. Lithium bromide represents an intermediate case. The observed frequencies of the solvation bands for the lithium halides are consistent with the above explanation. The frequency of the lithium iodide solvation band is identical with that of lithium perchlorate, lithium thiocyanate, and lithium tetraphenylborate, indicating that the vibration must be that of the cations in the solvent cage. The band shifts to progressively lower frequencies as the iodide ion is replaced first by the bromide and then by the chloride ions. Thus, if our model is correct, the bromide and chloride ions can successfully compete with the acetone 47 Figure 6. Spectra of solutions of acetone + 6LiClO4 in nitromethane. A. Blank = 1.5 M acetone 10:1 B. Acetone/6LiClO4 ratio C. Acetone/6LiC104 ratio = 6:1 D. Acetone/6LiClO4 ratio = 4:1 E. Acetone/6Lic1o4 ratio = 2:1 48 A so ~ sot .— a& 40- 2o- 4§o 410 370 c m-‘ Figure 6. 49 molecules for the position in the solvent cage even when the concentration of acetone is relatively high as compared with that of the two halide ions. Recent kinetic study (47) of isotOpic exchange of 82Br between lithium bromide and butyl bromide in acetone shows that the exchange rate is very small, and that the lithium bromide ion pairs are unreactive Species, indicating that the bromide ion is very strongly bound to the lithium cation. The fact that the chloride and bromide ions have a strong affinity for the cation is further demonstrated by the following studies. The position and the intensity of the solvation band in 0.5 M lithium perchlorate in acetone solutions were measured as a function of the concentration of added salts, tetrabutylammonium perchlorate, tetrabutylammonium iodide, and tetrabutylammonium bromide. The solvation and the 369 cm"1 bands remained practically unchanged upon the addition of 0.4 M tetrabutylammonium perchlorate or 0.4 M tetrabutylammonium iodide. However, upon the addition of tetrabutylammonium bromide, the band shifted progressively from 424 to 412 om’l, and the 369 em-l band also shifted to 358 em‘l. The data are given in Table 7. Similar results were obtained with lithium iodide as the solute. Acetone solutions of the various tetrabutylammonium salts did not show any new bands in the measured Spectral region. It seems from the above data that the 369 cm'1 band is also associated with the cation-solvent vibration. Symmetrical polyatomic anions, such as the perchlorate ion, have little tendency to form contact ion pairs in the presence of sufficient amounts of a solvating solvent. The Raman band of the perchlorate 1 ion at 935 cm- remains unchanged when the acetone/LiClo4 mole ratio 50 Table 7. The Shifting of the 424 em‘1 and 369 om-l Bands Conc. LiClO Conc. Bu NBr 4 4 (M) (M) Frequency 0.54 - 424 cm—1 i 3 369 om'l :_4 0.53 0.12 420 368 0.53 0.23 418 368 0.53 0.37 415 364 0.52 0.55 413 362 0.53 0.87 412 358 51 is 3_4. Below this value, however, the band progressively broadens and shifts to higher frequency, and at mole ratio of approximately 1:1, the frequency is at 945 cm'l. These data offer additional evidence that the solvation number of the lithium ion is 4, and that if the system does not contain enough acetone to maintain the salvation shell, the deficiency is made up by the perchlorate ion which then forms a contact ion pair with the cation. The results are shown in Figure 7. It is seen from Figure 7 that parallel with the change in the perchlorate band, there is a change in the appearance of the 789 cm"1 (C-C symmetric stretch) Raman band of acetone. As the concentration of lithium is increased, a new band appears at 803 cm‘l, representing vibration of the acetone molecule bound to the lithium ion. Nuclear Magnetic Resonance Mole Ratio Studies In order to further establish the stoichiometry of the solvated species, mole ratio studies on the acetone-Li+ system were carried out using proton magnetic resonance Spectra (27). Again nitromethane was used as the inert solvent, and the position of the methyl proton resonance of acetone was followed as a function of the acetone/Li+ ratio. Four studies were performed. In the first study, the acetone concentration was kept constant at 1.5 M and the concentration of lithium perchlorate was varied. The result (Figure 8) shows that there is a definite break in the curve at acetone to salt ratio of 4.3. The position of the methyl proton resonance of nitromethane remains practically unchanged down to about 1:1 mole ratio. In the second study, the lithium perchlorate concentration was kept constant at 0.5 M and the concentration of acetone was varied. Again a break 52 Figure 7. The 789 cm—1 acetone and the 935 cm'1 C104- Raman bands. A. B. 3.5 M acetone in nitromethane Acetone/LiClO4 ratio Acetone/LiClO4 ratio Acetone/LiClO4 ratio Acetone/LiC104 ratio Acetone/LiClO4 ratio 6:1 4:1 3:1 2:1 ml:l Band intensity semi—quantitative 0 . i5 7 8] e _ r m m 0C F to 8 ]O F F. D C B A 54 Figure 8. Chemical shift of the methyl protons of acetone X§° acetone/ LiClO4 mole ratio. Concentration of acetone = 1.5 M. Concentration of LiClO4 varied. Solvent: nitromethane. A. Nitromethane. B. Acetone. 55 262b () C) 1L58 _1 I\Iv I4C) cps 1136 T 132L D D P j l 1:28 2 4 6 8 1() . Acetone M'R' UC|04 Figure 8. 56 at acetone/LiClOA mole ratio of 4.4:1 was observed, as shown in Figure 9. Two other studies were carried out using lithium iodide instead of lithium perchlorate. Similar breaks at acetone/Lil mole ratio of 4.4:1 and 4.5:1 were obtained. The results indicate that lithium ion is solvated by four molecules of acetone. It is possible that the deviation from the integral value of 4 are due to the solvation of the anions. Vapor Pressure Osmometric Measurements The apparent molecular weights of seven lithium and four sodium salts were measured in acetone. Method I as discussed in Chapter IV was used to calculate the molecular weights. Generally, with the exceptions of the tetraphenylborate salts, the AR/C vs, C (concentration) plots yielded straight lines in the 0.03 to 0.14 M concentration range. For tetraphenylborate, the linear concentration range was 0.02 to 0.10 M. The method of plotting AR/C XE: AR (Method II) was used to calculate the apparent molecular weight at each concentration of the salt solutions. The data are given in Tables 8 and 9. The AR/C gs, CM calibration curve is shown in Figure 10. The calibration curve of AR/C vs, AR is similar to Figure 10, except that AR is used instead of C. The AR/C vs, C plots for biphenyl, the seven lithium and the four sodium salts are given in Figures 11 to 14. A summary of the results obtained is given in Table 10. The nonlinearity of the AR/C 22' C plots at concentrations lower than about 0.03 M for most of the salts renders the molecular weights obtained from extrapolation to infinite dilution subject to sizable error (as can be seen from the apparent molecular weight at each concentration 33. concentration plots, Figures 15 to 17). However, 57 Figure 9. Chemical shift of the methyl protons of acetone vs. acetone/LiClO4 mole ratio. Concentration of LiClO4 = 0.5 M. Concentration of acetone varied. Solvent: nitromethane. A. Nitromethane. B. Acetone. 58 O O (3 On 262- O p ' ' AV 260r 4:- 138+ :71 o. . 0 134+ 130- 2 4 (T 8 10 Acetone M. R. “Ti—c.254— Figure 9. 59 Table 8. Vapor Phase Osmometry Calibration Data - Benzil in Acetone Run Conc (M) AR AR/CIn I 0.1773 68.93 388.8 0.1460 57.95 396.9 0.1147 46.00 401.0 0.08344 33.93 406.6 0.06258 25.47 407.0 0.04172 17.36 416.1 0.02186 9.14 438.2 0.01043 4.59 440.1 11 0.1870 72.78 389.2 0.1603 62.52 390.0 0.1389 55.12 396.8 0.1176 46.74 397.6 0.06412 25.92 404.2 0.04275 17.57 411.0 0.03206 13.39 417.7 0.02137 9.20 430.5 0.01069 4.72 441.5 111 0.1911 76.02 397.9 0.1804 71.27 395.2 0.1486 58.98 396.9 0.1274 51.30 402.8 0.1061 42.75 402.8 0.09022 36.88 408.8 0.07430 30.33 408.2 0.05838 24.23 415.0 0.04777 19.81 414.7 0.03715 15.67 421.8 0.02654 11.58 436.3 0 2 .01592 7.04 442. 2518 9. 1..1CCFLF\ COCO... OOOOOPUO 60 Table 9. Data of Vapor Phase Osmometric Measurements - Biphenyl, Lithium and Sodium Salts Run Conc (M) Conc(g/l) AR AR/C M 8/1 app Biphenyl I 0.1751 27.00 68.84 2.550 153.3 0.1401 21.60 53.75 2.488 159.9 0.1138 17.55 45.55 2.595 154.9 0.08754 13.50 34.35 2.544 159.9 0.06128 9.450 24.83 2.628 156.8 0.04377 6.750 17.64 2.613 158.8 0.01751 2.700 7.68 2.844 153.3 LiNO3 I 0.1300 8.960 41.63 4.646 87.0 0.1056 7.280 35.36 4.857 83.8 0.08123 5.600 28.34 5.061 81.0 0.06498 4.480 23.71 5.292 77.9 0.04874 3.360 28.36 5.464 75.9 0.03249 2.240 13.40 5.982 70.6 11 0.1304 8.992 41.10 4.571 88.4 0.1060 7.306 34.96 4.785 85.1 0.08152 5.620 28.91 5.144 79.7 0.06522 4.496 23.62 5.254 78.5 0.04891 3.372 18.53 5.494 75.5 0.03261 2.248 13.06 5.810 72.7 0.01630 1.124 7.00 6.228 70.3 (C 38189 Cont . Q" ”in mm hm hm hm am 0 O C ( ’\ 61 Table 9. (Continued) Run Conc (M) Conc (g/l) AR AR/Cg/l M aPP LiSCN I 0.1492 9.700 56.77 5.853 67.8 0.1193 7.759 45.76 5.898 68.2 0.09697 6.305 38.13 6.048 67.1 0.07459 4.850 29.44 6.070 67.5 0.05221 3.395 21.38 6.297 65.7 0.03730 2.425 15.48 6.384 65.5 0.02238 1.455 9.67 6.646 64.8 II 0.1354 8.806 52.45 5.956 67.0 0.1078 7.009 41.82 5.966 67.7 0.09560 6.216 37.91 6.099 66.6 0.07967 5.180 31.64 6.108 66.9 0.06374 4.144 26.11 6.301 65.2 0.04780 3.108 20.08 6.461 64.1 0.03187 2.072 14.01 6.762 62.5 LiC1O4 1 0.1614 17.168 78.26 4.558 84.7 0.1311 13.949 63.98 4.587 85.7 0.1008 10.730 51.47 4.797 83.2 0.08068 8.584 41.20 4.800 84.2 0.05650 6.009 30.14 5.016 81.6 0.03530 3.756 20.27 5.397 76.7 0.02020 2.146 12.26 5.713 74.3 II 0.1573 16.740 77.55 4.633 83.5 0.1337 14.229 66.28 4.658 84.1 0.08653 9.207 45.53 4.945 81.3 0.06293 6.696 33.56 5.012 81.4 0.04720 5.022 26.10 5.197 79.1 0.03147 3.348 18.56 5.544 74.9 0.01573 1.674 10.08 6.022 71.4 .‘Le 9. CI C(((tl vi *1 62 Table 9. (Continued) Run Conc (M) Conc (g/l) AR. AR/Cg/l Mapp LiCl I 0.1030 4.366 27.77 6.361 64.5 0.08240 3.493 23.20 6.642 62.2 0.06700 2.840 19.16 6.747 61.5 0.05150 2.183 15.14 6.935 60.1 0.03610 1.530 11.23 7.340 58.2 0.02060 0.873 6.93 7.938 55.2 11 0.1033 4.380 27.98 6.388 64.2 0.07750 3.285 21.80 6.640 62.3 0.05683 2.409 16.76 6.957 59.8 0.04133 1.752 12.82 7.317 57.7 0.03100 1.314 10.02 7.626 56.4 LiBr I 0.1425 12.375 53.74 4.343 91.7 0.1173 10.191 45.51 4.466 90.0 0.1006 8.735 39.45 4.516 89.7 0.08382 7.280 34.38 4.723 86.2 0.16705 5.823 27.77 4.769 86.1 0.05029 4.368 21.86 5.005 82.6 0.03353 2.912 15.00 5.151 81.3 0.01676 1.456 8.20 5.632 77.2 11 0.1349 11.712 52.04 4.443 89.8 0.1111 9.646 44.26 4.587 87.8 0.09519 8.267 38.43 4.649 87.2 0.07933 6.890 32.88 4.772 85.5 0.06346 5.512 27.05 4.908 83.7 0.04760 4.134 21.02 5.084 81.3 0.03173 2.756 14.72 5.341 78.5 0.01587 1.378 7.82 5.674 76.8 189. u u a q A 000000.0. 00.00.0.0.0.0. 63 Table 9. (Continued) Run Conc (M) Conc (g/l) AR AR/Cg/l Mapp ii}. 1 0.1449 19.390 66.10 3.409 115.0 0.1177 15.760 55.64 3.530 112.4 0.09959 13.330 46.95 3.522 113.8 0.07243 9.695 36.95 3.811 106.5 0.05432 7.271 28.40 3.906 105.0 0.03622 4.848 20.64 4.257 97.2 0.018109 2.424 11.37 4.691 90.9 II 0.1375 18.400 62.81 3.414 115.4 0.1203 16.099 56.42 3.505 113.3 0.1031 13.800 49.57 3.592 111.4 0.08591 11.500 42.76 3.718 108.5 0.06873 9.200 35.90 3.902 104.2 0.05155 6.901 27.44 3.976 103.2 0.03436 4.599 19.18 4.171 99.5 0.01718 2.300 10.71 4.657 92.0 LiBPh4 I 0.1043 34.016 64.51 1.896 207.5 0.07821 25.512 48.75 1.911 209.6 0.06518 21.262 41.38 1.946 207.6 0.05214 17.008 34.54 2.031 200.6 0.03911 12.758 25.95 2.034 202.2 0.02607 8.504 18.21 2.141 193.8 0.01304 4.254 9.48 2.228 193.4 11 0.1147 37.415 69.50 1.857 210.2 0.09447 30.816 58.26 1.891 209.5 0.07422 24.211 45.97 1.899 211.6 0.05978 19.500 38.63 1.981 204.4 0.03985 12.999 26.84 2.065 199.1 0.02699 8.804 18.50 2.101 197.5 0.01993 6.501 14.00 2.154 195.5 0.01350 4.404 9.59 2 178 197.8 Table 9. (Continued) 64 Run Conc (M) Conc(g/l) AR AR/Cg/1 Mapp NaBPh4 I 0.1220 41.76 80.42 1.926 200.1 0.1005 34.39 65.61 1.908 205.4 0.07895 27.02 52.16 1.930 206.7 0.06460 22.11 42.04 1.901 212.4 0.05024 17.194 33.56 1.952 209.0 0.03589 12.283 24.01 1.956 210.8 0.02153 7.370 15.06 2.043 205.1 0.01436 4.913 10.19 2.074 207.0 11 0.08615 29.48 56.03 1.901 208.9 0.06627 22.68 44.62 1.967 204.5 0.04639 15.877 31.59 1.990 205.6 0.03314 11.342 22.36 1.972 209.5 0.01325 4.535 9.71 2.142 201.3 NaClO4 I 0.01505 18.429 71.64 3.887 100.2 0.1239 15.172 61.02 4.022 98.1 0.1062 13.004 52.96 4.072 97.8 0.08853 10.840 45.18 4.168 96.4 0.07082 8.672 36.45 4.203 96.7 0.05312 6.504 28.08 4.317 95.1 0.03541 4.336 19.50 4.497 92.2 0.01771 2.168 10.77 4.968 86.2 II 0.1659 20.32 78.52 3.864 99.9 0.1481 18.130 70.57 3.892 100.2 0.1219 14.930 60.54 4.055 97.4 0.1045 12.797 52.40 4.095 97.5 0.08709 10.664 44.35 4.159 96.8 0.06968 8.532 36.35 4.260 95.4 0.05226 6.399 28.05 4.384 93.6 0.04006 4.906 21.96 4.476 92.4 0.02177 2.666 13.22 4.959 85.2 00000000 0000000. 65 Table 9. (Continued) Run Conc (M) Conc (g/l) AR AR/Cg Mapp ya; I 0.1292 19.375 63.06 3.255 121.0 0.1016 15.224 51.21 3.365 119.0 0.08308 12.455 42.18 3.387 119.2 0.06462 9.688 33.57 3.465 117.7 0.04616 6.920 25.77 3.724 110.5 0.02770 4.153 16.54 3.983 104.7 II 0.1491 22.36 71.25 3.186 122.4 0.1228 18.410 58.54 3.180 124.6 0.1017 15.290 51.14 3.345 119.3 0.07017 10.520 36.32 3.452 117.7 0.04386 6.575 24.26 3.690 111.7 0.01754 2.630 10.97 4.171 102.5 NaSCN 1 0.1555 12.608 58.10 4.608 85.9 0.1372 11.124 52.22 4.694 85.0 0.1189 9.640 47.11 4.887 82.0 0.09146 7.415 37.12 5.006 81.1 0.06402 5.191 27.57 5.311 77.3 0.04573 3.708 20.68 5.577 74.2 0.02744 2.225 13.27 5.964 70.8 0.01463 1.186 7.75 6.532 66.7 11 0.1613 13.078 60.98 4.663 84.6 0.1371 11.116 52.38 4.712 84.7 0.1129 9.156 44.55 4.867 82.7 0.08873 7.194 35.35 4.914 82.8 0.06453 5.232 27.43 5.243 78.4 0.04033 3.270 18.32 5.602 74.1 0.01613 1.308 8.15 6.231 69.8 66 Figure 10. The AR/Cm vs. Cm calibration plot. Benzil in acetone. 67 o... «a . 2: u may. vmv. .oH ooamae hone. 68 Figure 11. The AR/C vs. C plots for (A) LiCl. (B) LiSCN. g/1‘—— 8/1 (C) LiN03. 5. 2'- Figure 11. CW“) ms? 12 70 Figure 12. The AR/Cg/l XE: Cg/l plots for (A) Biphenyl. (B) LiBPhq. (C) NaBPha. 71 CD mm” C) C) :33 u 0 CD 0.. 4 100 .NH ooowae ..V.N \\ fiV’ no.) 2|‘7 AV.N l Fi ure 1 . h (C) LiBr. Cg/ 72 1 plots for (A) NaSCN. 5E 4J 5.7 (B) LiC104. ‘ AR CQ/I 73 C(9/l) Figure 13. 74 Fi r 1 . gu e 4 The AR/Cg/l vs, Cg/l plots for (A) NaClOa. (B) Lil. (C) NaI. 4.5 75 o— m— :3: .qa wwswwm O _1_\\4 06 nd mi table 10- LI 6 u ’_A re 76 Table 10. The Apparent Molecular Weights of Alkali Metal Salts in Acetone Calibration Constant Km = 424 M01. Wt. M01. Wt. Linear Salt OAR/Cg/l)c=0 (apparent) (real) Mapp/Mreal Conc Range* LiCl 7.60 55.8 42.4 1.32 0.03 -- 0.10 LiBr 5.40 78.5 86.9 0.90 0.03 -- 0.14 LiI 4.46 95.1 133.8 0.77 0.03 -- 0.14 LiNO3 6.00 70.7 68.9 1.03 0.04 -- 0.13 LiSCN 6.80 52.4 65.0 0.96 0.03 -- 0.13 LiClO4 5.30 82.3 106.4 0.77 0.04 -- 0.14 LiBPha 2.24 189.3 326.2 0.58 0.01 -- 0.09 NaI 3.70 114.6 149.9 0.76 0.04 -- 0.15 NaSCN 5.65 75.0 81.1 0.92 0.04 -- 0.15 NaClOa 4.59 92.2 122.4 0.75 0.04 -- 0.15 NaBPh4 2.06 205.8 342.2 0.60 0.01 -- 0.10 Biphenyl 2.68 158.2 154.2 1.03 0.02 -- 0.14 *In the AR/C vs, C (unit g/l) plots. 77 Figure 15. The apparent molecular weights (Mapp) vs. molar concentration plots. (A) LiN03. (b) Lil. (C) LiBr. (D) LiCl. 78 Al 80. / 6(N' ;. 110' O 3 o O E O c o. 9 MW 70" W o 5()P .04 .08 f2 .7 C (M) Figure 15. 79 Figure 16. The apparent molecular weights (Mapp) vs, molar concentration plots. (A) LiBPhé. (B) Biphenyl. (C) LiClOa. (D) LiSCN. Mapp Figure 16 . 80 250 200? 150’ 90* 70' 50 . 1 . 1 1 i C (M) 81 Figure 17. The apparent molecular weights (Mapp) XE: molar concentration plots. (A) NaBPhA. (B) NaI. (C) NaClOA. (D) NaSCN. _e_riELfio 0 no A. - A 1901- A, 0 8 120 0 Q O. O E 100 (DC 1 .0 GD 6* 80* o 60- .054 .08 1‘2 1'6 C (M) Figure 17. 83 in the 0.03 to 0.14 M concentration range, a qualitative comparison of the apparent molecular weights of the various salts may be of significance. From the apparent molecular weights it is seen that among the lithium salts, the tetraphenylborate is the most dissociated, the iodide and the perchlorate appear to be less dissociated, and no appreciable dissociation is observed for the nitrate, thiocyanate and the bromide. Apparently, higher aggregates (ionic or molecular) are present in appreciable amount for lithium chloride. Sodium iodide, sodium perchlorate and sodium tetraphenylborate are also appreciably dissociated, with greatest dissociation being observed for the tetraphenylborate. Some association is observed, however, in the sodium thiocyanate solutions. From the greater dissociation of lithium tetraphenylborate, lithium iodide, lithium perchlorate and the sodium salts, we would expect that the above electrolytes exist in solutions mainly as solvent-separated ion pair. The molecular weights for lithium nitrate, lithium thiocyanate and lithium bromide indicate no appreciable dissociation in the concentration range studied. Even in_ these cases, solvent separated ion pairs cannot be ruled out for? the three salts, eSpecially for the polyatomic anion salts. From the apparent molecular weight, it is reasonable to assume that, in addition to ion pairs, appreciable higher ionic or molecular aggregates exist for lithium chloride in acetone solution. Thus the far infrared band for the lithium chloride in acetone may arise from vibrations of contact ion pair, or solvated contact ion The fact that pair, or other solvated species of higher aggregates. the chloride and the bromide ions do form strong bonds with lithium 84 ion have been shown by the low ion pair dissociation constants from conductance (33) and kinetic isotOpic exchange studies (47). Thus the results from vapor pressure osmometric measurements agree with the trends observed from the various studies. With the exceptions of biphenyl and sodium tetraphenylborate, the apparent molecular weights of all the salts studied show some degree of concentration dependence. The apparent molecular weight at each salt concentration vs. molar concentration plots for the eleven salts studied are shown in Figures 15 to 17. Besides biphenyl and sodium tetraphenylborate, the salts lithium tetraphenylborate and lithium thiocyanate also show very small degrees of concentration dependence of their apparent molecular weights. (B) Acetic Acid Preliminary investigations indicated that acetic acid is quite transparent in the 430-250 cm—1 spectral region as well as below 150 cm-1. It is Opaque between 250 and 150 cm-1. The lithium Salts are quite soluble in acetic acid. Some sodium salts are also filightly soluble. Among the common salts of other alkali metals CHLly the fluorides have sufficient solubilities for spectral measure- IDEIIts. The frequencies of the far infrared solvation bands of lithium Séhlts, together with those of potassium fluoride, rubidium fluoride, Cesium fluoride and tetramethylammonium fluoride are shown in Table 1].. The values are given for the averages of at least three runs. 131E! concentration range studied was 0.1 to 1.5 M. The intensities of the observed bands were proportional to the salt concentrations. 85 Table 11. Solvation Bands of the Alkali Metal Salts in Acetic Acid Solutions (cm-l) Salt CHBCOOH CD3C00D LiCl 390 :_4 364 :.5 LiBr 391 367 Lil 392 368 LiNO3 390 366 LiSCN 391 366 LiClO4 390 366 LiBPh4 389 362 6 . L1C1 405 i 5 6LiBr 408 6Lil 410 6 LiNO3 407 6 L10104 405 KF‘ 281 :.10 RbF 283 CsF 281 (CH3)4NF 278 86 It is interesting to note that all the lithium salts show a band at 390 :.4 cm"1 irrespective of the nature of the anion. In acetone solutions the band frequencies of the lithium salts with polyatomic anions were also found to be independent of the anion but the halides showed bands at slightly different frequencies, indicating the possible formation of contact ion pairs. It seems, therefore, that in acetic acid solutions contact ion pairs are not formed by lithium salts to any appreciable extent although, of course, due to the low dielectric constant of the solvents the electrolytes exist largely in the form of solvent—separated ion pairs. The low frequency of the lithium - acetic acid solvation band, as compared with other solvents (e.g., 429 cm.1 in dimethyl sulfoxide (1), 425 cm-1 in acetone, etc.) seem to reflect lower energy of interaction between acetic acid and the lithium ion. The fact that the frequency of this band changed upon isotopic substitution (6Li and CD COOD) substantiates the assumption that 3 the band is due to the cation-solvent vibration. The shifts with isotopic substitution are similar to those Observed with acetone. Application of the same Hook's law 1 Cétlculation to acetic acid predicts a shift of 28 cm- with the Stflastitution of 6L1 salts in place of the 7Li salts. Experimentally the shift is found to be approximately 15 om’l. Similarly, Hook's lLawv calculation of the effect of substitution of deuterated solvent Pradicts a shift of less than 2 cm-l. Experimentally a large shift 015 about 24 cm"1 is observed. The effect of water on the far-infrared solvation bands was e"amined by the gradual addition of water to solutions of lithium perchlorate. The results are shown in Figure 18. It is seen that Figure 18. 87 The effect of water on the solvation band of LiClO4 in acetic acid solution. A. Pure acetic acid. B. 0.7 M LiClO4 in acetic acid. Water 0.06% (with no addition of water). C. 0.6 M LiClOQ in acetic acid. Water D. 0.6 M LiClO in acetic acid. Water 4 E. 0.6 M LiClOA in acetic acid. Water content content content content The same cell was used to obtain B, C, D and E. was obtained from a different cell. 88 530 4305—130 620 340 80F 50- 404 10 czlrr“ Figure 18. 89 the position of the solvation band does not change upon the addition of small amounts of water. Even at 2% water, the band can still be vaguely distinguished. However, due to water absorption, the baseline increases with increasing water content. The solvation band practically disappeared at water content above about 2.5%. Since most of the measurements were done with 0.5 M solution of salts and the combined water content of salt and solvent at this concentration was less than 0.1%, the effect of this small amount of water would be negligible. The solvation bands of sodium and ammonium ions usually occur at %200 cm—1. Since the solvent is opaque in this region these bands could not be observed. The solvation bands of rubidium and cesium ions occur below 150 cm-1. The spectra of the respective fluorides were measured in this region but the solvation bands were not found, indicating that the solvent-cation interaction is quite weak, at least by comparison with that found in other solvents studied thus far. On the other hand, a broad band was observed at W280 cm"1 for bOth fluorides as well as for solutions of potassium and tetramethylammonium fluorides. Thus, unlike the solvation bands Sttndied thus far, the 280 cm_1 band must be due to fluoride anion - SOJJvent vibration. Acetic acid is a strongly hydrogen—bonding Solsvent and since the fluoride ion is particularly susceptible to thia formation of hydrogen-bond, it is not surprising that this interaction is manifested by the anion solvation band. The other halide anions must also undergo some hydrogen-bonding with the solvent, althOugh to a lesser extent than the fluoride ion and, therefore, It Seems reasonable to assume that the lithium halides in acetic aci<1 Solutions do not form strong contact ion pairs. 90 A one molar hydrogen fluoride solution in acetic acid was prepared by mixing acetic anhydride with a stoichiometric amount of aqueous 48% hydrofluoric acid. The Spectrum of the solution was similar to that of pure acetic acid. The 280 cm‘1 fluoride anion - acetic acid solvation band observed for other fluorides was conSPicuously absent, and no new band was detected in the 600—250 cm"1 spectral region. Since hydrogen fluoride forms very Strong hydrogen-bonds within itself and thus is highly associated, it is likely that the fluoride anion in hydrogen fluoride does not interact appreciably with the acetic acid hydroxyl proton, and thus the fluoride anion solvation band can not be observed. Raman Spectra in the 3000-150 cm“l spectral region were also ob tained of the lithium perchlorate, bromide, nitrate and of I‘llbidium fluoride and cesium fluoride solutions in acetic acid in the 0.6-2.0 M concentration range. The 390 cm"1 lithium - acetic acid and 280 cm-‘1 fluoride - acetic acid solvation bands were not obServed. Thus the bands are Raman-inactive and, therefore, the SOlvent — ion interactions must be essentially electrostatic. Attempts to measure the apparent molecular weights of the electrolytes in acetic acid by vapor pressure osmometry were Unsuccessful. The thermistors of the osmometer could not be balanced with pure solvent, due to excessive drifting. The low Vapor pressure of the solvent at 37°C may produce too small a temperature difference to allow meaningful measurements. (CL Dimethyl Sulfoxide Attempts to run vapor pressure osmometry measurements using dimchYl sulfoxide as solvent were unsuccessful. As in the case of 91 acetic acid, the thermistors of the osmometer could not be properly balanced with the pure solvent. The vapor Pressure of dimethyl sulfoxide at 37°C is probably too low to give meaningful readings. (D) Tetrahydrofuran Far Infrared Studies Tetrahydrofuran is a relatively good solvent for far infrared solvation studies. It has a good spectral window with only one absorption band at 283 cm’1 in the 600 to 100 cm-1 spectral range. The band is shifted to 235 cm-1 in tetrahydrofuran-d8. Most of the common lithium salts and some of the sodium salts are quite soluble in the solvent. The far infrared Spectra of six lithium salts and one sodium salt in tetrahydrofuran solutions were measured. In all cases, a broad band, which has been assigned to alkali ion vibration, appeared in the Spectrum of each salt solution. The frequency for the lithium perchlorate band, which has not been determined previously, was 404 1.3 cm-l. The frequencies of the solvation bands for other salts, with the exception of lithium iodide, agree with those of Edgell, ££.§l-’ to within 2 cm.1 (see Chapter II). The spectra Of several lithium salts in tetrahydrofuran-d8 solutions were also ‘measured. The data are given in Table 12. From Table 12, it is seen that the substitution of deuterated Solvent has practically no effect on the frequency of the lithium chloride band in tetrahydrofuran solution. The shifting of the lithium bromide band frequency is also within the experimental uncertainty. The band frequency of the lithium perchlorate 92 Table 12. Frequencies (cm—l) of Alkali Ion Vibrations in Tetrahydrofurans Salt Tetrahydrofuran Tetrahydrofuran-d8 LiCl 389 :.3 389 :_3 LiBr 380 377 Lil 381 LiClO4 404 396 LiNO3 408 LiBPh4 412 NaI 182 solution shifted from 404 cm.1 in tetrahydrofuran to 396 cm"1 in 1, which is beyond the experimental tetrahydrofuran—d8, a shift of 8 cm" uncertainty and shows that the solvent participates in the vibration. The observed constancy of the lithium chloride band frequency upon Substitution of the deuterated solvent may be fortuitous; however, it may also indicate that the solvent does not participate in the vibration of lithium chloride in tetrahydrofuran solution. Ferraro, §£_al, (118, 119) have studied the effect of pressure on the vibration of ionic crystals at long wavelengths. They find that internal modes of vibration of polyatomic ions (e.g., N03”, 3042-) are relatively insensitive to pressure -- the change being about 0.0-0.2 cm-l/kbar of applied pressure. 0n the other hand, large shifts in frequency were observed for external (lattice) modes, e.g., 0.7—1.9 cm-l/kbar. Based on these findings, Edgell, 93 leg 31. (10), studied the variation with pressure of the vibrational frequency of lithium chloride in tetrahydrofuran solution. They obtained the average value of 1.5 cm'l/kbar of applied pressure and concluded that the vibration is characteristic of a lattice mode in an ionic solid which means that solvent molecules are also involved in the vibration. In the present study, the non-shifting of the frequency of lithium chloride band upon substitution of the deuterated tetra- hydrofuran as well as the high degree of association of lithium chloride in the solvent, as indicated by vapor pressure osmometric measurements (see below) indicate that the possibility of the vibrational frequency at 389 cm”1 arising from the dimer vibration can not be ruled out. The halides of alkali metals, especially the lithium halides, are known to be polymerized in vapor phase, in matrices and in solvents with low polarities. The occurrence of dimer and trimer molecules in addition to the monomers has been well established. For some of the halides (notably the lighter halides) the dimer to monomer ratio is as high as 3:1 (120-123). The infrared spectra of the lithium salts, both in gaseous phase (124, 125) and in matrices of argon, krypton, xenon and nitrogen (126—128) have been investigated and the vibrational frequencies of the monomers and dimers assigned (124,129). The structure of lithium chloride has also been investigated by electron diffraction (130), which confirmed the theoretical predicted configuration of a planar diamond shaped structure. 94 There are three vibrational frequencies of the dimers, and with the exception of those of lithium fluoride (which occur at higher frequencies), all the frequencies lie within the 500 to 150 cm-1 spectral range. The frequencies are found to be matrix dependent, e.g., in nitrogen matrix, at 20°K, the two higher dimer vibrational frequencies of lithium chloride are at 425 cm‘1 (b3u) and 325 cm‘1 (qu)’ The presence of lithium bromide tetramers in ether has also been reported (131). In the light of the findings described above, and from the data obtained in the present study, it is not unreasonable to ascribe the 389 cm"1 lithium chloride band in tetrahydrofuran solution to that of the dimer vibration, while the vibrations of lithium iodide and other lithium salts with polyatomic anions may involve the solvent molecules and the anions. Lithium bromide may represent the intermediate case. The dimer vibration is that of the lattice mode, and hence is sensitive to pressure, as was found by Edgell, £1331- (10). Vapor Pressure Osmometric Measurements The apparent molecular weights of seven lithium and four sodium Salts in tetrahydrofuran solutions were measured. The procedure used is the same as that described in the vapor pressure osmometric measurements in acetone solutions. The data obtained are given in Tables 13 and 14. The AR/C vs. Cm calibration curve for the solvent iS given in Figure 19. The AR/C XE: C plots for the various salts are given in Figures 20 to 23. A summary of the results obtained is given in Table 15. From the apparent molecular weights obtained it is seen that none 0f the salts studied show any degree of dissociation in 95 tetrahydrofuran solutions. Association is appreciable for all the salts, especially at higher concentrations. Among the salts studied lithium chloride showed the greatest extent of association. In fact, the apparent molecular weight of lithium chloride approaches that of a dimer. The infrared band frequencies of alkali metal salts in tetrahydrofuran solutions are all anion dependent, which indicates that the anions are either still bound to the cations or are present in the near neighbor environment, and together with the solvent molecules, form a cage in which the cations vibrate. Lithium chloride may represent a possible exception in that in view of the strong association of the salt, the solvent, with its low polarity, is unable to break the bond of the dimers. Thus the vibrational band may be due to the dimers, and not the ion pair vibration or vibrations involving solvent molecules. From the apparent molecular weight at each salt concentration Kg. concentration plots, as shown in Figures 24 to 26, it is seen that the apparent molecular weights of all the salts are concentration dependent, though the degree of dependence varies. There is very Small concentration dependence of the apparent molecular weights of the lithium and sodium tetraphenylborate salts in acetone solutions. However, in tetrahydrofuran solutions, the apparent molecular weights of the two salts are highly concentration dependent. The contrasting data may reflect the difference in the dissociating Powers of the two solvents. 96 Table 13. Vapor Phase Osmometry Calibration Data — Benzil in Tetrahydrofuran Run Conc (M) AR AR/Cm 1 0.1776 72.12 406.0 0.1110 46.47 418.5 0.07772 32.65 420.1 0.05551 23.31 419.9 0.02924 13.02 445.3 11 0.2018 82.55 409.1 0.1816 75.02 413.1 0.1413 57.87 409.7 0.1211 50.41 416.4 0.1009 41.96 415.9 0.07305 31.00 424.4 0.05549 23.63 425.8 0.04036 17.34 429.6 0.03027 13.04 430.8 0.01514 6.81 449.8 111 0.1989 81.31 408.8 0.1889 76.88 406.9 0.1790 73.23 409.1 0.1691 69.03 408.3 0.1492 61.36 411.2 0.1293 53.36 412.8 0.1094 45.53 416.3 0.09447 39.28 415.8 0.07955 33.12 416.3 0.06464 27.17 420.3 0.04972 21.10 424.4 0.03480 15.00 431.0 0.01989 8.90 447.5 97 Table 14. Data of Vapor Phase Osmometric Measurements - Biphenyl, Lithium and Sodium Salts Run Conc (M) Conc (g/l) AR AR/Cg/l Mapp Biphenyl 1 0.1836 28.321 74.90 2.645 154.1 0.1561 24.072 62.22 2.585 159.0 0.1280 19.739 51.14 2.591 159.8 0.09902 15.270 39.70 2.600 160.6 0.07346 11.328 30.15 2.662 157.8 0.05510 8.497 22.89 2.694 156.8 0.03673 5.664 15.37 2.714 156.2 0.01836 2.832 7.81 2.758 154.5 LlNO3 1 0.1532 10.560 36.45 3.452 121.2 0.1379 9.504 32.92 3.464 121.1 0.1148 7.917 28.82 3.640 115.5 0.09957 6.864 25.34 3.692 114.2 0.08425 5.805 22.12 3.809 110.9 0.06893 4.752 19.08 4.015 105.5 0.05361 3.696 15.31 4.142 102.5 0.03830 2.640 11.69 4.428 96.1 0.02298 1.584 7.72 4.874 87.5 11 0.1545 10.650 36.80 3.455 121.1 0.1390 9.585 33.35 3.479 120.5 0.1236 8.520 30.47 3.576 117.4 0.1081 7.455 27.50 3.689 114.1 0.09269 6.390 24.43 3.823 110.4 0.07724 5.325 21.05 3.953 107.0 0.06179 4.260 17.56 4.122 102.8 0.04634 3.195 14.00 4.382 97.0 0.03090 2.130 10.12 4.751 89.6 0.01545 1.065 5.72 5.371 79.5 98 Table 14. (Continued) Run Conc (M) Conc (g/l) aR AR/Cg/1 Mapp LiSCN 1 0.1489 9.682 49.32 5.094 81.4 0.1314 8.542 44.10 5.163 80.6 0.1051 6.834 36.40 5.326 78.6 0.08759 5.695 31.40 5.514 76.2 0.07007 4.556 25.68 5.637 74.8 0.05255 3.417 20.20 5.912 71.6 0.03504 2.278 13.67 6.001 70.8 0.01752 1.139 7.20 6.321 67.6 11 0.1572 0.220 51.45 5.034 82.2 0.1415 9.201 46.56 5.060 82.1 0.1179 7.667 39.94 5.209 80.2 0.09434 6.134 33.43 5.450 76.9 0.06289 4.089 23.36 5.713 73.9 0.04717 3.067 17.92 5.843 72.6 0.03145 2.045 12.51 6.117 69.5 LiClo4 1 0.1381 14.688 46.42 3.160 131.5 0.1251 13.306 42.30 3.179 131.1 0.1056 11.233 37.37 3.327 125.7 0.08121 8.640 29.82 3.451 121.8 0.06496 6.912 24.87 3.598 117.3 0.03248 3.456 12.95 3.747 113.4 0.01624 1.728 6.76 3.912 109.2 11 0.1570 16.700 53.31 3.192 129.7 0.1334 14.196 45.98 3.239 128.4 0.1177 12.525 41.62 3.323 125.5 0.09418 10.021 34.49 3.442 121.7 0.07848 8.350 29.56 3.540 118.8 0.06671 7.098 25.78 3.632 116.1 0.05494 5.846 21.65 3.703 114.2 0.03924 4.175 15.68 3.756 113.0 0.02747 2.923 11.27 3.856 110.4 99 Table 14. (Continued) Run Conc (M) Conc (g/l) AR AR/Cg/1 Mapp L1C1 I 0.1613 6.840 32.74 4.787 87.6 0.1372 5.814 27.68 4.761 88.4 0.1210 5.130 25.00 4.873 86.6 0.1049 4.446 21.71 4.883 86.6 0.08068 3.420 17.28 5.053 83.9 0.06454 2.736 13.82 5.051 84.1 0.04841 2.052 10.83 5.278 80.7 0.03631 1.539 8.30 5.393 79.1 0.02017 0.855 4.90 5.731 74.6 11 0.1503 6.370 30.53 4.793 87.6 0.1277 5.415 26.16 4.831 87.2 0.1127 4.777 23.39 4.896 86.2 0.08265 3.503 17.68 5.047 84.0 0.07514 3.185 16.25 5.102 83.1 0.06011 2.548 13.22 5.188 81.9 0.04508 1.911 10.21 5.343 79.7 0.03005 1.274 7.12 5.589 76.4 LiBr I 0.1534 13.320 41.85 3.142 132.7 0.1304 11.322 35.73 3.156 132.7 0.1150 9.990 32.20 3.223 130.2 0.09969 8.658 28.74 3.320 126.8 0.07669 6.661 23.13 3.473 121.5 0.06135 5.328 18.71 3.512 120.6 0.04601 3.996 14.49 3.626 117.1 0.03451 2.997 11.51 3.831 110.8 0.01917 1.665 6.91 4.150 102.9 11 0.1506 13.080 41.16 3.147 132.5 0.1355 11.772 37.56 3.191 131.0 0.1054 9.156 30.29 3.308 127.0 0.09036 7.848 26.07 3.322 126.9 0.07154 6.213 21.72 3.496 120.9 0.05648 4.905 17.94 3.658 115.9 0.04142 3.597 13.61 3.784 112.3 0.03012 2.616 11.16 4.266 99.0 100 Table 14. (Continued) Run Conc(M) Conc (g/l) aR AR/Cg/l Mapp £31 I 0.1640 21.950 51.27 2.336 177.2 0.1476 19.755 47.05 2.382 174.5 0.1312 17.560 43.20 2.460 169.3 0.1148 15.366 37.81 2.461 169.9 0.09839 13.171 32.86 2.495 168.1 0.08199 10.975 28.32 2.580 163.1 0.06559 8.780 23.14 2.636 160.1 0.04919 6.585 17.97 2.729 155.4 0.03280 4.391 12.31 2.804 151.8 0.01640 2.195 6.36 2.898 147.3 11 0.1559 20.860 49.50 2.373 174.8 0.1403 18.777 44.40 2.365 175.9 0.1247 16.690 41.73 2.500 166.8 0.1013 13.560 34.65 2.555 164.0 0.07793 10.432 27.30 2.617 160.9 0.06234 8.345 22.97 2.753 153.4 0.04675 6.259 17.57 2.807 151.0 0.03896 5.215 14.78 2.834 149.9 0.02727 3.650 10.90 2.986 142.6 LiBPh4 I 0.1025 33.448 31.41 0.939 447.2 0.08716 28.432 26.61 0.936 450.0 0.07178 23.415 22.45 0.959 440.6 0.06152 20.068 19.63 0.978 432.6 0.05127 16.724 16.76 1.002 423.1 0.04102 13.381 14.06 1.051 404.5 0.03076 10.034 11.06 1.102 386.2 0.02051 6.690 7.82 1.168 365.2 0.01025 3.344 4.46 1.334 320.6 11 0.09660 31.510 30.50 0.970 434.1 0.08211 26.785 26.22 0.979 430.6 0.07245 23.634 23.54 0.996 423.7 0.05796 18.907 19.43 1.028 411.9 0.04589 14.968 16.27 1.087 390.2 0.03623 11.818 13.22 1.119 379.9 0.02657 8.666 10.05 1.160 367.4 0.01691 5.516 6.93 1.256 339.9 101 Table 14. (Continued) Run Conc (M) Conc (g/l) AR [AR/Cg/1 Mapp NaBPh4 1 0.1031 35.251 34.06 0.966 433.6 0.08836 30.240 30.00 0.992 423.8 0.07363 25.199 24.86 0.987 427.8 0.05890 20.158 20.22 1.003 421.8 0.04418 15.120 16.18 1.070 396.5 0.02945 10.079 11.45 1.136 374.8 0.01473 5.040 6.25 1.240 344.3 11 0.1130 38.680 36.40 0.941 444.7 0.09183 31.428 30.25 0.963 436.6 0.07064 24.176 23.80 0.984 428.7 0.05298 18.132 18.55 1.023 414.0 0.03885 13.296 14.30 1.076 395.0 0.02119 7.251 8.71 1.201 355.1 NaClO4 I 0.1402 17.172 39.97 2.328 179.4 0.1169 14.310 34.88 2.437 171.2 0.09349 11.448 28.89 2.524 166.6 0.06233 7.632 20.38 2.670 160.7 0.04675 5.725 16.21 2.831 158.4 0.03116 3.816 11.55 3.027 149.9 0.01558 1.908 6.29 3.297 140.6 11 0.1538 18.830 42.90 2.278 182.9 0.1307 16.005 37.50 2.343 178.4 0.1077 13.182 31.91 2.421 173.5 0.08458 10.357 26.41 2.550 165.2 0.06920 8.474 22.40 2.643 159.8 0.05382 6.590 17.93 2.721 155.8 0.03845 4.708 13.47 2.861 148.6 0.02307 2.825 8.97 3.175 134.2 102 Table 14. (Continued) Run Conc (M) Conc (g/l) AR AR/Cg/l Mapp E8}. 1 0.1535 23.010 42.00 1.825 228.3 0.1264 18.948 35.85 1.892 221.3 0.1083 16.242 31.71 1.952 215.1 0.09028 13.535 26.82 1.982 212.4 0.07222 10.827 22.14 2.045 206.6 0.05417 8.121 17.46 2.150 197.2 0.03611 5.414 12.26 2.265 187.8 0.01806 2.707 6.87 2.538 168.2 11 0.1638 24.560 44.52 1.813 229.5 0.1393 20.876 38.43 1.841 227.1 0.1147 17.191 32.75 1.905 220.2 0.09830 14.737 28.82 1.956 215.1 0.08192 12.281 24.34 1.982 212.9 0.06554 9.826 20.32 2.068 204.5 0.04915 7.369 16.25 2.205 192.4 0.02458 3.685 8.74 2.372 179.8 313st I 0.1682 13.640 41.96 3.076 135.6 0.1262 10.230 32.20 3.148 133.4 0.1009 8.184 26.29 3.212 131.2 0.08412 6.820 22.55 3.306 127.7 0.06729 5.456 18.77 3.440 123.1 0.05047 4.092 14.68 3.587 118.4 0.03365 2.728 10.64 3.900 109.2 0.01682 1.364 5.73 4.201 101.7 11 0.1639 13.290 40.37 3.038 137.4 0.1475 11.961 36.16 3.023 138.4 0.1311 10.632 33.14 3.117 134.5 0.1147 9.303 29.23 3.142 133.8 0.09015 7.309 23.86 3.265 129.3 0.07376 5.981 19.90 3.327 127.1 0.05737 4.652 16.06 3.452 122.9 0.04098 3.323 12.31 3.705 114.8 0.02459 1.994 7.95 3.987 107.0 103 Table 15. The Apparent Molecular Weights of Alkali Metal Salts in Tetrahydrofuran Calibration Constant Km = 429 Salt (AR/Cg/1)C=O M01. Wt. M01. Wt. 2322 Linear Conc. (apparent) (real) Mreal Range (M)* L1C1 5.54 77.4 42.4 1.83 0.04 - 0.14 LiBr 4.03 106.5 86.9 1.23 0.03 - 0.13 LiI 2.96 144.9 133.8 1.08 0.02 - 0.16 LiNO3 4.59 93.5 68.9 1.36 0.05 - 0.14 LiSCN 6.18 69.4 65.0 1.07 0.02 — 0.14 L1C104 4.01 107.0 106.4 1.01 0.02 - 0.14 LiBPhA 1.27 337.8 326.2 1.04 0.02 - 0.09 NaI 2.30 186.5 149.9 1.24 0.03 - 0.16 NaSCN 3.74 114.7 81.1 1.41 0.05 - 0.15 Na0104 3.00 143.0 122.4 1.17 0.04 — 0.15 NaBPh4 1.14 376.3 342.2 1.10 0.02 - 0.11 Biphenyl 2.75 156.0 154.2 1.01 0.02 - 0.16 *In the AR/Cg/l XE: C plots. 104 Figure 19. The AR/CIn XE- Cm calibration plot. Benzil in tetra- hydrofuran. 105 ON. op. NL. A<C>P‘Fi c>uwa>c>m SYIT LIB l WINES 56 mumummmlumnmmumum