SPECTROSCOPIC STUD’uES OF SONIC SOLVATION
EN NO‘NAQUEOUS MEDIA

Dissertation for the Degree of Ph. D.
MiCHEGAN STATE UNIVERSETY
MARK S. GREENBERG
1974

   

is”!
.

  
   

L I13 2? A
,3 Michigan 5;- .
13% University a

m
. . . \ . pm ‘ a
'. .' Y? r
I :_ w
. I, I - ‘ ' I!
. _ , | .
t . g
5 - “ h
This“! to certify that the
- thesis entitled "
'l

SPECTROSCOPIC STUDIES OF IONIC SOLVATION
IN NONAQUEOUS MEDIA

 

' O

    
    
       
 

3:; Mark az‘éfien . “ ' ’1

. _ ,i A

.. .V - {h - . presempd ‘ ‘5 .

. "h t \ ( > I“ ' l
a;

has been accepted » ar-‘f

 

3 a of the requilém , k.
' egre” 7
i

I: .

79"

/

   

'.

  

‘7 professor

   
 

 

e. r" 4-.-
q .‘ ' _ ‘ I
t . ‘—
V 33' 1' u"
0 ‘ - . ' LI .’ .
. .
5 . - . ‘3
fl 1‘
r
a I.
I I

 

 

 

 

 

 

ABSTRACT
SPECTROSCOPIC STUDIES OF IONIC SOLVATION IN NONAQUEOUS MEDIA
By

Mark S. Greenberg

Sodium-23 NMR measurements have been performed on several
sodium salts in eighteen nonaqueous solvents as a function of salt
concentration. The chemical shifts for solutions of sodium
tetraphenylborate and perchlorate exhibited little or no concentration
dependence, whereas chemical shifts of corresponding solutions of
sodium thiocyanate, bromide and iodide showed marked concentration
dependence. The static nature of the 23Na shifts in the former
case is proposed to be indicative of either free Na+ ion or solvent
separated ion pairs. The 23Na shifts in the latter case are proposed
to be indicative of contact ion pair formation.

A plot of the infinite dilution 23Na chemical shifts in these
solvents against the Gutmann donor numbers of these solvents yielded
a straight line. Hence, the magnitude and direction of the 23Na
chemical shift reflects the relative donicity of these solvents.

A quantitative evaluation of ion pair formation constants suggests
that the formation of contact ion pairs is strongly influenced

not only by solvent dielectric constant but also by the "donicity"

(solvating ability) of the solvent. The interpretations obtained

@\

Mark S. Grecnberg
from the 23Na chemical shift correlate well with the data obtained
from electrical conductance measurements.

Competitive solvation studies for the Na+ ion in binary solvent
mixtures of nitromethane, acetonitrile, hexamethylphosphoramide,
dimethylsulfoxide, pyridine and tetramethylurea were monitored by
23Na NMR. Generally, these studies reflected the relative donicity
of each solvent in a given solvent pair where the solvent of higher
donicity was preferentially contained in the inner solvation shell
of the Na+ ion. However, in binary solvent mixtures of pyridine
with dimethylsulfoxide and tetramethylurea, the donicity of pyridine
seemed to be repressed. Infrared studies of the former mixture
suggest that strong solvent-solvent interactions disrupt the
associative structure of DMSO molecules resulting in enhanced donicity
for DMSO in these mixtures. Because tetramethylurea exhibits
structure similar to that of DMSO, the same solvent—solvent
interactions are proposed to enhance its donicity.

The slight upfield shift of the 23Na resonance with increasing
concentration for solutions of sodium perchlorate was reexamined by

23 5

Na and 3 Cl NMR, Raman and infrared spectroscopy. Linear upfield

shifts of the 23Na resonance were noted for solutions of NaClO4 in

methanol, ethanol, water and formic acid whereas non-linear upfield

shifts were observed in the other solvents studied. In pyridine,

acetonitrile and tetrahydrofuran, a new Raman band at N 470 cm-1

in addition to the other C10 - bands at 456, 626, and 935 cm"1 was

4

observed. With increasing concentration, the 470 cm-1 band increased

Mark S. Greenberg

in intensity relative to the 456 cm-1. Hence, these bands were proposed

to be indicative of bound and free C104“, respectively. The line-

width of the 35Cl resonance was identical within experimental error

35

(30 :_5 Hz) for all solutions of NaClO as was the Cl chemical

4
shift (-lO40 1.5 ppm). These data suggest that contact ion pair
formation of NaClO4 does occur in the solvents studied; moreover,
the interaction is weak. In the hydroxylic solvents, the energy

of interaction is proposed to be less than kT, since the 23Na Shift

varies linearly with concentration.

SPECTROSCOPIC STUDIES OF IONIC SOLVATION

IN NONAQUEOUS MEDIA

By

Mark Sf Greenberg

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

1974

ACKNOWLEDGEMENTS

The author wishes to thank Professor Alexander I. Popov for
his guidance, encouragement, friendship and counseling throughout
this study.

He also wishes to thank Professor Joseph A. Caruso of the
University of Cincinnati for his unique introduction to research,
his friendship and many good times.

Gratitude is also extended to the Department of Chemistry,
Michigan State University, the National Science Foundation and the
National Institutes of Health of the Department of Health, Education,
and Welfare for financial aid.

Appreciation is extended to Dr. Richard Bodner for numerous
enlightening discussions and experimental assistance during the
early part of this work. The present members of the "group",

Paul Gertenbach, Duke DeWitte, Yves Cahen, Robert Baum, DavidDeBrosse
and John Thompson, should be acknowledged for their gifts of friendship
and enthusiasm, and an incessant willingness to lunch at McDonalds.
Special thanks to Mr. Patrick Kelly for interfacing the Raman
instrument and providing program PEAKSBF and to Wayne DeWitte

and Robert Baum for their expenditure of time in proofreading

this thesis.

Professor Stanley R. Grouch, Crouchl, Crunch, Couchy, etc. is
acknowledged not only for his helpful suggestions as second reader,
but for his outstanding sense of humor. The author would also like
to thank Professor Robert Hammer and the students of Honors
Chemistry, 1970-1971, for providing a most unusual and rewarding
teaching experience and some important friendships.

Special thanks are given to Messrs. Eric T. Roach, A. Wayne
Burkhardt and Frank Bennis, without whose cooperation the NMR
investigations would have been much more difficult.

Deep appreciation is extended to my wife, Roni, for her love,
patience and encouragement throughout the years of graduate study.

To her and to our families, I dedicate this thesis.

ii

TABLE OF CONTENTS

Chapter
I. HISTORICAL
NUCLEAR MAGNETIC RESONANCE .
INFRARED SPECTROSCOPY
RAMAN SPECTROSCOPY .
CONCLUSIONS
II. EXPERIMENTAL PART
SALTS
SOLVENTS .
SAMPLE PREPARATION .
INSTRUMENTAL MEASUREMENTS
Nuclear Magnetic Resonance
Infrared Spectra.
Laser Raman Spectra .

Conductance Measurements
Data Handling . . . . . .

III. A SODIUM-23 NUCLEAR MAGNETIC RESONANCE AND ELECTRICAL
CONDUCTANCE STUDY OF CONTACT ION PAIRS IN

NONAQUEOUS SOLVENTS
INTRODUCTION .

RESULTS AND DISCUSSION . . .

IV. STUDIES OF PREFERENTIAL SOLVATION OF THE SODIUM ION

IN MIXED SOLVENTS BY SODIUM-23 NMR
INTRODUCTION . . .
RESULTS AND DISCUSSION .

CONCLUSIONS . . . . . . . . .

iii

Page

. 12

. 15

. l6

. 16

17

. 20

. 20

22
22
23

. 23

. 24

. 25

51

O 52

78

Table of Contents (Continued)

v. A SPECTROSCOPIC STUDY OF CONCENTRATED SOLUTIONS OF
SODIUM PERCHLORATE - THE NATURE OF THE UPFIELD
CHEMICAL SHIFT

INTRODUCTION .
RESULTS AND DISCUSSION .
CONCLUSIONS
v1. APPENDICES
I. DESCRIPTION OF COMPUTER PROGRAM KINFIT AND
SUBROUTINE EQN FOR THE CALCULATION OF ION
PAIR FORMATION CONSTANTS BY THE NMR
TECHNIQUE
11. DESCRIPTION OF COMPUTER PROGRAM KINFIT
AND SUBROUTINE EQN FOR THE RESOLUTION
OF OVERLAPPING RAMAN AND INFRARED BANDS
III. DESCRIPTION OF COMPUTER PROGRAM PEAKSBF
FOR THE RESOLUTION OF OVERLAPPING RAMAN
AND INFRARED BANDS .

IV. DESCRIPTION OF COMPUTER PROGRAM SHEDLOV
FOR EVALUATION OF CONDUCTANCE DATA .

VII. LITERATURE CITED

iv

80

81

. 107

. 108

. 112

. 115

. 123

. 126

LIST OF TABLES

Table Page
1. Salvation Band Frequencies of Alkali Metal Ions
in Nonaqueous Media . . . . . . . . . . . . . . . . . . 9
2. Key Solvent Properties . . . . . . . . . . . . . . . . 18
23

3. Na Chemical Shifts vs. 3.0 M_Aqueous NaCl . . . . . . 27

4. Ian Pair Formation Constants for Various Sodium
Salts by 23Na NMR . . . . . . . . . . . . . . . . . . . 50

5. Variation of the Sodium-23 Resonance as a Function of
Solvent Composition for Binary Solvent Mixtures . . . . 53

6. Isosolvation Points for Salvation of the Sodium‘
Ian in Binary Solvent Mixtures. . . . . . . . . . . . . 63

7. Variation in the Frequency of the Sodium Ian Salvation
Band for 0.50 M NaBPh4 Solutions in DMSO- Pyridine
Solvent Mixtures . . . . . . . . . . . . . . . . . . . 66

8. Variation of the S- O Stretching Frequency of DMSO for
DMSO- Pyridine Solvent Mixtures . . . . . . . . . . 69

9. Covington Treatment of Preferential Salvation . . . . . 77
10. 23Na Chemical Shifts for Sodium Perchlorate Solutions
in Nonaqueous Media . . . . . . . . . . . . . . . . . . 82

11. Association Constants of Sodium Perchlorate from
Conductance Studies . . . . . . . . . . . . . . . . . . 88

12. Computer Analysis of Sodium Perchlorate Solutions in
Tetrahydrofuran and Acetonitrile by PEAKSBF and
KINFIT O O O O O O O O O O O O O O O O O O O O O I O O 96

13. KINFIT Analysis of Raman Spectra of Sodium Perchlorate
Solutions on Acetonitrile from 900- 950 cm‘1 . . . . . . 100

List of Tables (Continued)

14. 35Cl Chemical Shifts for Sodium Perchlorate
Solutions in Nonaqueous Media .
15. 35Cl Linewidth for Sodium Perchlorate Solutions

in Nonaqueous Media .

vi

. 103

. 104

Figure

10.

11.

12.

LIST OF FIGURES

Sodium-23 Chemical Shifts of Various Sodium Salts in
l,l,3,3-Tetramethylurea and N,N-Dimethylformamide .

Sodium-23 Chemical Shifts of Various Sodium Salts
in Sulfolane and Acetone . . . . . . . . . . . .

Sodium-23 Chemical Shifts of Various Sodium Salts
in Pyridine and l,l,3,3-Tetramethy1guanidine

Sodium-23 Chemical Shifts of Various Sodium Salts
in Tetrahydrofuran. . . . . . . . . . . . . . .

Conductance Curve far Sodium Iodide in
1,1,3,3-Tetramethy1guanidine. . . . . . . . . . . .

Conductance Curve for Sodium Iodide in Pyridine .

Sodium-23 Chemical Shifts of Various Sodium Salts in
Methanol and Ethanol . . . . . . . . . . . . . . .

Sodium-23 Chemical Shifts of Various Sodium Salts in
Propylene Carbonate and Dimethylsulfoxide . . . . . . .

Plot of Infinite Dilution Sodium-23 Chemical Shifts.
versus the Donor Number of the Solvent. . . . .-. .

Sodium-23 Chemical Shifts of Sodium Iodide in
Sulfolane, Formamide, Acetone and Dimethylsulfoxide . . .

Variation of the Chemical Shift of the Sodium-23
Resonance as a Function of Solvent Composition for
Binary Solvent Mixtures of Nitramethane with
Acetonitrile, Tetramethylurea, Dimethylsulfoxide,
Pyridine and Hexamethylphospharamide. . . . . . . .

Variation of the Chemical Shift of the Sodium-23
Resonance as a Function of Solvent Composition for
Binary Solvent Mixtures of Acetonitrile with
Tetramethylurea, Dimethylsulfoxide, Pyridine and
Hexamethylphospharamide . . . . . . . . . . . . . . . . .

vii

Page

. 26

. 33

. 34

. 35

. 37

. 38

. 39

. 41

. 43

. 58

59

List of Figures (Continued)

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Variation of the Chemical Shift of the Sodium-23

Resonance as a Function of Solvent Composition

for Binary Solvent Mixtures of Hexamethylphospharamide
with Tetramethylurea, Dimethylsulfoxide and

Pyridine . . . . . . . . . . . . . . . . . . . . . . . . 60

Variation of the Chemical Shift of the Sodium-23

Resonance as a Function of Solvent Composition

for Binary Solvent Mixtures of Tetramethylurea

with Dimethylsulfoxide and Pyridine . . . . . . . . . . 61

Variation of the Frequency of the Sodium Ian
Salvation Band for 0.50 M_NaBPh4 Solutions in
Dimethylsulfoxide-Pyridine Solvent Mixtures . . . . . . 67

Variation Of the 5-0 Stretching Frequency for
Dimethylsulfaxide—Pyridine Solvent Mixtures . . . . . . 7O

Covington Plat for the Binary Solvent System
Dimethylsulfaxide-Pyridine . . . . . . . . . . . . . . . 76

Sodium-23 Chemical Shifts of Sodium Perchlorate-
Solutians in Nonaqueous Media . . . . . . . . . . . . . 86

Vibrations of the C104- group as a Function
of Symmetry . . . . . . . . . . . . . . . . . . . . . . 91

Raman Spectra of 1.0 M_Sodium Perchlorate
Solutions in Tetrahydrofuran and Acetonitrile
fI‘Om 400-500 C111 0 o o O a o a o o o o o o o o o o o o 93

KINFIT Analysis of 1.0.M_Sadium Perchlorate
Solution in Tetrahydrofuran from 400-498 cm‘l. . . . . . 94

PEAKSBF Analysis of 1.0 M Sodium Perchlorate
Solution in Tetrahydrofuran from 430—498 cm'l. . . . . . 95

Raman Spectra of Sodium Perchlorate Solutions
in Acetonitrile from 900-950 cm-l. . . . . . . . . . . .101

viii

LIST OF NOMENCLATURE, ABBREVIATIONS AND SYMBOLS

Contact Ion Pairs. Pairs of ions, linked electrostatically, but

 

with no covalent bonding between them.

Solvent Shared Ion Pairs. Pairs of ions, linked electrostatically

 

by a single, oriented solvent molecule.

Solvent Separated Ian Pairs. Pairs of ions, linked electrostatically

 

but separated by more than one solvent molecule.

ACN: Acetonitrile

PC: Propylene Carbonate

THF: Tetrahydrofuran

DMF: Dimethylfarmamide

TMU: 1,1,3,3-Tetramethylurea
DMSO: Dimethylsulfoxide

HMPA: Hexamethylphospharamide

TMG: l,1,3,3-Tetramethy1guanidine

A = equivalent conductance of the solution

= equivalent conductance at infinite dilution

S = the Onsager slope = OAO + B

a = the relaxation effect = (8.205 X 10-5)/(DT)3/2
B = the electrophoretic effect = (82.43)/n(DT)1/2

D = solvent dielectric constant

ix

List of Nomenclature (Continued)

['11
ll

L.
ll

m
ll

absolute temperature
viscosity
concentration

Ele + E

ale + 02

degree of dissociation

2

association constant

mean activity coefficient

CHAPTER I

HISTORICAL

NUCLEAR MAGNETIC RESONANCE

 

One of the central problems in bath aqueous and nonaqueous
solutions of electrolytes is the characterization of the interaCtions
that exist between the constituent ions and among the ions and
the solvent. Chemical literature contains more than nine hundred
electrical conductance studies of electrolyte solutions since
classically this was one of the first approaches and still continues
to be one of the more important methods for characterization of
ionic equilibria in solutions. Other techniques which have been
employed to study complexatian and salvation in electrolyte solutions
include spectrophotometry in the visible and ultra-violet region,
distribution equilibria, ultrasonic relaxation techniques, potentio-
metry, vibrational spectroscopy and magnetic resonance (1-6).

Nuclear magnetic resonance (NMR) has become a powerful tool for
the investigation of electrolyte solutions. There have been many
studies of the chemical shift and relaxation times of protons of the
solvent molecules or in the solvated species (e.g., tetraalkylammonium
salts) (7-15). However, relatively wa studies of the magnetic
resonances of nuclei other than protons present in the ions themselves
have been reported, although the chemical shifts and line widths of
the nuclear resonances would yield information about both ion-ion
and ion—solvent interactions.

All the alkali metal and halide ions possess at least one

isotope with a magnetic nucleus, i.e., 7Li, 23Na, 39K, 87Rb,

133Cs, 19F, SSCI, 79'813r and 127I. Deverell and Richards (16)

2

studied the chemical shifts of 23Na, 39K, 87Rb and 133Cs nuclei in

aqueous solutions of alkali halides and nitrates as a function of
salt concentration. The concentration dependence of the chemical
shifts were attributed to interactions between the cations and
anions in solution. These same workers surveyed the chemical shifts
of 35Cl, 81Br and 1271 nuclei in aqueous solutions of alkali

halides and determined that direct cation-anion collisions were the
predominant cause of the chemical shifts of the halogen resonances
in solutions of potassium, rubidium and cesium halides. In lithium
and sodium halide solutions the cation-water-halide interaction was
of prime importance (17). WennerstOrm, gt, a}: (18) measured NMR

linewidths of 35C1, 793i, Slat and 127

I nuclei in aqueous solutions
of various substituted ammonium, phosphonium.and sulfonium salts.
The observed line broadening was attribUted to anion—solvent
interactions due to the structure making or breaking action of the
cation. Noting that 35Cl chemical shifts reflect the solvent
environment of the Cl' ion, Langfard and Stengle (19) studied
competitive salvation of Cl' ion in acetanitrile-water and dimethyl-

35C1 and 7Li line-

sulfoxide-water mixtures. Bryant (20) monitored
widths far concentrated LiCl solutions (1-15 M) and suggested that
changes in the linewidth were indicative of ion-pairing
phenomena.
Maciel, gt, 9;, (21) observed 7Li chemical shifts of dilute solutions

of lithium bromide and perchlorate in water and eleven organic solvents.

Recent studies by Akitt and Downs and work in this laboratory reveal

the frequency of the 7Li resonance to be quite sensitive to the
environment (22-24). Halliday, 333 213 (25) investigated 133Cs chemical
shifts of cesium halides, hydroxide and nitrate in water and seVeral
nonaqueous solvent. The shifts were found to vary non-linearly with
salt concentration. The degree of nan-linearity depends upon the
dielectric constant of the solvent. Hence, the shifts reflect cation-
anian collisions. Carrington, gt, 31. (26) observed linear shifts in
133Cs resonance as halide salts are added to a solution of CsCl and
postulated contact ion pair formation to be responsible for these shifts.
Bloor and Kidd (27) determined the 39K chemical shifts in a number of
aqueous electrolyte solutions as a function of concentration and observed
both upfield and dawnfield chemical shifts due to random cation-anion
collisions.

The sodium-23 nucleus is also well suited for such.nmr studies,
and therefore, it was the subject of investigation in this thesis.
The relative sensitivity of 0.1 at constant field with respect to the
proton indicates that measurements can be obtained with fairly dilute
solutions. The small natural linewidth of the sodium ion resonance
(~15 Hz) permits the use of high resolution NMR eqUipment. Finally,

24 esu cm2 renders this nucleus

the large quadrupole moment of 0.1 K 10'
a sensitive probe of the neighboring electronic environment.

The first detailed 23Na nmr investigation was performed by
Jardetzky and Wertz (28) who reported an Observed broadening of the
23Na resonance line as a fhnction of concentration and anion which they

attributed to an interaction of the 23Na nuclear quadrupole with an

4

electric field gradient caused by the formation of weak complexes of
sodium with the added anion. Rechnitz and Zamochnick (29) evaluated
the formation constant of sodium ion with several acid anions. They
plotted these formation constants vs. the Jardetzky-Wertz linewidths
and observed a linear relationship. Eisenstadt and Friedman (30)
determined relaxation times for aqueous solutions of sodium perchlorate
and chloride finding that the former has a marked effect on the 23Ma
relaxation time whereas the latter did not.

In a study similar to that Of Jardetzky and Wertz, Griffiths and
Socrates (31) found the linewidth of the 23Na resonance for several
inorganic salts to be prOportional to the concentration, indicative of
ion pair formation. However, they likewise did not observe any
variation in the 23Na chemical shifts in their systems. Richards and
Yorke (32) studied 79Br, 81Br and 23Na resonances for aqueous and
nonaqueous solutions of sodium, calcium and cesium bromide. The
results indicate that for sodium and calcium bromide solutions relaxation
arises from ion-solvent interactions whereas ion-ion interactions are
important for cesium bromide solutions.

In 1968, Bloor and Kidd (33) reported 23Na chemical shifts of
sodium iodide solutions in fourteen nonaqueous solvents as a fUnction
of concentration. The concentration dependence of the observed chemical
shifts was attributed to cation-anion interactions; moreover, the
observed range of the chemical shifts was related to changes in the
paramagnetic term of the general screening equation. A fair correlation
was found between the magnitude of the chemical shift extrapolated to

infinite dilution in a given solvent and the Lewis basicity of the

solvent. In order to extend these measurements to new solvents and also
to determine the influence of different anions on 2:I’Na chemical shifts,
Erlich, 333 21, (34) studied 23Na chemical shifts of sodium tetraphenyl-
borate, perchlorate, thiocyanate and iodide in several nitrogen and
oxygen donor nonaqueous solvents. In the case of the first two salts,
the chemical shifts were found to be independent of concentration, while
for the last two, varying degrees of concentration dependence were
noted. In this latter case, it was assumed that the chemical shifts
were influenced by the formation of contact ion pairs. A linear

23Na chemical shift

correlation was found between the magnitude of the
in a given solvent and the Gutmann donor number* of that solvent.
Templeman and Van Geet (35) measured 23Na chemical shifts and
relaxation times for aqueous Solutions of sodium tetraphenylborate,
perchlorate, hydroxide and chloride and postulated that short-lived
collisional ion pairs are responsible for the observed chemical shifts.
By monitoring the 23Na chemical shift as a function of added water,
Van Geet (36) determined the hydration number of sodium ion to be
between three and four. Canters (37) examined the interactions of glyme

ethers with the sodium ion as a function of temperature, anion,

solvent and viscosity.

 

*Gutmann's donor numbers (96) are the enthalpy of complex formation between
the given solvent and antimony pentachloride in 1,2-dichloraethane
solution

S + SbClS w S'SbCls

Gutmann used the term "donicity" when referring to the donor ability
of a solvent.

6

Since the Na+ ion is of biological significance, several studies
were pursued to further elucidate the magnitude of interaction of the
sodium ion with biologically important molecules. The interaction of
Na+ with biological tissues has been reported by Cope (38) and
Ratunno (39). The camplexing of ionophores which function as mobile
carriers of alkali cations in membranes, such as valinomycin, manensin,

23Na NMR (40). A good

nigericin and monactin, were studied with
correlation was observed between the 23Na resonance position of the 1:1
complex and the stability constant of the complex.

Sodium-23 NMR relaxation time measurements have been used to
investigate the formation of weak complexes of sodium ion with some
aminopalycarboxylic acids (41), soluble RNA (42), sodium-potassium
transport adenosinetriphophatase (43), some phosphate-containing compounds
of biological interest (44) and interactions with cysteine, aspartic
acid and citric acid in aqueous solution (45). The magnitude of the
change of relaxation time yields a qualitative indication of the
strength of the sodium ion interactions with the ligand. Andrasko and
Forsen (46) applied pulsed fourier transform 23Na NMR to the study of
the binding of sodium ion with simple carbohydrates.

Cerasa and Dye (47) employed 23Na NMR to study the exchange rate
of the sodium ion between two environments in solution; ethylenediamine
and a hexaoxadiamine macrobicyclic camplexing agent (cryptate). Below
the coalescence temperature of 50°C, two well defined resonance
absorptions were Observed, indicative of free and complexed sodium ion.

Shchari, gt: a}, (48) monitored complexation of sodium ion by dibenzo-

l8-crown-6 in dimethylfarmamide but were unable to observe two peaks

7

because the linewidth of the complexed Na+ ion was very broad, although

the line shape analysis indicated that exchange was slow.

INFRARED SPECTROSCOPY

 

While studying the far infrared spectra of tetrabutyl- and
tetrapentylammonium chlorides and bromides in benzene solution,

Evans and Lo (49) noted bands at 120 cm-1 and 80 cm.1 respectively,
which could not be attributed to a vibrational mode of the solute or
the solvent. Since the band position was dependent both on the mass
of the cation and anion, it was ascribed to the direct cation-anion
ion pair vibration and constituted the first report Of an ionic
vibration in solution.

At the same time, Edgell, 33: 31: (50) observed far infrared bands
of alkali metal tetracarbonylcobaltates and pentacarbonylmanganates in
tetrahydrofuran solutions. These bands likewise did not arise from
either the solute or the solvent. After extending these studies to
such solvents as dimethylsulfoxide, pyridine and piperidine, it was
concluded that the band position was a function of cation and
solvent. Hence, the observed band represented the alkali ion vibrating
in a cage (51). These vibrations were named "salvation bands".

POpov and ca-warkers extended these far infrared studies to a
wide range of nonaqueous media (52—60) and observed these salvation
bands to exhibit strong cation dependence and much weaker solvent
dependence. The bands, which follow Beer's Law within experimental
accuracy, are intense, but broad, with band widths equal to or greater

1

than 50 cm' . Isotopic substitution far the cations or solvent

indicated that band frequencies vary inversely with the change in

mass Of the cation or the solvent. Generally, no anion dependence is
noted in solvents of high polarity or donicity; however, in some
solvents of low polarity or donicity, such as tetrahydrofuran (50,51),
acetone (56) and propylene carbonate (60) some anion dependence is
noted. Such dependence is postulated to be indicative of contact ion
pairing in solution.

The frequencies Of the alkali metal ion salvation bands in several
solvents are presented in Table 1. Generally, the Li+ salvation band
occurs at N400 cm-l, Na+ and NH4+ at mZOO cm-l, K+ at mlSO cm-l, Rb+
at m120 cm-1 and Cs+ at mllO cm-l. These data are indicative of the
strong cation dependence Of the frequency Of these bands.

A unique electrolyte, sodium tetrabutylaluminate, which is soluble
in solvents of very low polarity, has been the subject of extensive
investigation. Tsatsas and Risen (61) reported two concentration

1 for this salt in

1

dependent far infrared bands at 195 cm-1 and 160 cm-
cyclahexane and tetrahydrofuran solutions in addition to a 202 cm-
Raman band in the farmer which is attributed to sodium ionic motion.

In addition to a 1H nmr study (8), Olander and Day investigated salvation

of sodium by tetrahydrofuran in the system THF-NaAlBut -cyclohexane

4

by monitoring the v stretching frequency of THF in the region

COC
900-1150 cn’1 (62,63).
Edgell's group (64) investigated the effects of different solvents
and cations on the infrared spectrum of metal tetracarbonylcobaltate
salt (MCo(C0)4) solutions. The 1890 cm“1 band (C-O stretch) was faund

to be quite symmetrical in dimethylfarmamide, dimethylsulfoxide, and

wet tetrahydrofuran, indicative of a symmetrical "solvent-surrounded"

mm

Rm

cm

em
mm.em
am

mm

mm
mm.mm.mm
Hm

Hm
mm.Hm

Hc.oc

Oucoaowom

wvH
OOH ova
mvH
NmH
MNH omH
OHH mNH mmH
omH
+mU +Dm +M +

OON

NON

oHN

CNN

NNN

vHN

emz

 

AHIEOV moflocoswouu mcmm coaum>fiom

owa

mad

VON

NON

VNN

CNN

OON

owH

me

ooN

omH

m2

ONv

omm

va

mHv

mom

oov

va

ONv

va

va

now

“A

ocfioflusa

vwo< proo<

ocopoo<
ocowflaouhxthachw>tH
Ocowfiaonpxthtfixcpoth
Ocowwfiohhxth
ocfixomasaasesefia
ObfixomH5meOHmHo
ooaxomfisoaseSoeAa
ocwkuxm

oceceeoafla
cc-ocflxomaanAEOoeaa

cmaawoavxsmuuoe

muco>fiom

«five: msoosvmcoz :fi mcoH Hope: flamxa< mo mofiocoscoam wcmm :ONpm>Hom .H CHQMH

10

00

mm

mm

mm

mm

mm

Nam

mad

va

va

CON

owH

owa

oov

mmm

mmm

Nom

vwm

0mm

apmcopnmu ecoHNQOHm
ocfiwfihxmonofisutN
ocflvfiexmaxzuoaflotv.m
ocflwflhxmaxsuoEthV.N
ocfivfihxmaxnuozum

onwwfiaxmfixnuoztv

AEODCMucouv H OHDMH

11

ion environment. However, in pyridine, piperidine, dry tetrahydrofuran
and dimethoxyethane the band is split into two components above and
below the 1890 cm—1 band. These data indicate increasing asymmetry
about the anion resulting from contact ion pairing. A temperature
study Of NaCO(C0)4 in tetrahydrofuran solutions enabled the complex
band envelope around 1890 cm-1 to be resolved into four components
indicative Of two anion environments assigned as contact and solvent
separated ion pairs (65). Recently Barbetta and Edgell (66) examined
the infrared spectrum of thallium tetracarbonylcobaltate in seven
solvents as a function of temperature. Only a single ion site was
found in dimethylfarmamide, dichloromethane and dimethylsulfoxide
solutions. Several kinds of sites were found in tetrahydrofuran,
acetonitrile and nitromethane solutions including free ions, solvent
separated ion pairs, contact ion pairs and triple ions.

Tsatsas and Risen (67) observed far infrared bands for lithium,
sodium and calcium ions in ethylene-methacrylate ionic capolymers.
These same workers also investigated far infrared bands of alkali
metals in cyclic polyether "crown" compounds in pyridine and dimethyl-
sulfoxide solutions (68).

The results presented in this thesis are applications of the
aforementioned studies. A more extensive historical discussion of
salvation studies by infrared spectroscopy can be found in the Ph.D.
theses of B. W. Maxey (69), J. L. Wuepper (70), P. R. Handy (71)

and M. K. Wong (72).

12

RAMAN SPECTROSCOPY

 

Recently, with the advent of laser excitation Raman spectroscopy,
the methods Of vibrational spectroscopy to probe ion-ion and ion-
salvent interactions are very much in the forefront in the studies of
electrolyte solutions. For axyanions such as perchlorate, nitrate and
sulfate, the splittings and intensities of the anion Raman lines have
been suggested as probes Of the disposition of solvent molecules
around the ions and indicators of the presence of contact or solvent
separated ion pairs.

Such studies were originally pursued in aqueous solutions of metal
oxyanion salts. Hester, 33, 31, (73) investigated aqueous solutions of
indium sulfate, nitrate and perchlorate and found in all solutions a
broad, polarized Raman band at 400 cm.1 which probably resulted from
hydrated In3+. In both the indium sulfate and nitrate solutions, three
new polarized Raman lines appeared which were accounted for in terms
of C3v and C2V symmetry for bound sulfate and nitrate anion respectively.
NO new bands due to bound perchlorate were observed. Hester and
Plane (74) examined Cu2+, Zn2+, Hg2+, Mg2+, In3+ and Ga3+ sulfate,
nitrate and perchlorate solutions and noted a new polarized line at
360-400 cm-1 in each solution. Because, for a given cation, this band
was independent of anion, it was suggested that the origins of these
lines are in some symmetric farms of vibrations within the hydration
sheaths of the solvated metal ions. These same workers studied
saturated aqueous metal ion-oxyanian solutions and concluded that
cation-nitrate complexes are much more common than the corresponding

sulfate complexes, while there is no evidence for the existence of

perchlorate complexes (75).

13

Several research groups have suggested that the magnitude of
splitting of the v2(E') band of D3h free nitrate into v4(Bl) and
v1(A1) lines characteristic of bound C2v nitrate may be used as a
measure of the strength of the cation-nitrate interaction (76-78).
Using this criterion, Hester and Plane (75) proposed the following

. . . . 4+
order of cation-nltrate Interaction strength: Th > In3+ > Cu2+ >

Hg2+ > Ce3+ > Ca2+ > Zn2+, A13+, Ag+ > Na+, K

+

, NH4+ (75).

While studying the Raman spectra of the nitrate ion in aqueous
solutions of cadmium and zinc nitrate, Irish, gt: a}: (79) proposed
spectroscopic criteria for solvent separated and contact ion pairs.
A splitting of the v3(E') band in the 1300-1500 cm.1 region and the

absence of splitting in the 720 cm.1 region is proposed to be indicative

3

A contact ion pair interaction features an additional loss of

of an interaction of the type M+(H20)nN0 (solvent separated ion pair).

degeneracy of the v4(E') mode resulting in two bands at 720 and
740 cm-1. After examining the 1050 cm-1 region of aqueous solutions
of calcium, cadmium, zinc, and silver nitrate in detail, Janz,

gt: a}, (80) further proposed that a nitrate band at 1036 cm"1 is
indicative of a contact ion pair whereas a band at 1041 cm'1 is
indicative of a solvent separated ion pair.

Hence, these detailed Raman studies of aqueous solutions of metal
nitrate salts resulted in several criteria to serve as guidelines for
the identification of ionic species in solution. In order to
evaluate these criteria in nonaqueous media, detailed Raman studies
of metal nitrates in acetonitrile were pursued, principally by Janz.

In addition, solvent bands could also be monitored to gain further

insight as to the disposition of solvent in these solutions.

14

Balasubrahmanyam and Janz (81) examined silver nitrate solutions
in acetonitrile and noted the data to be in accord with spectroscopic
criteria for contact ion pairing. In the N03” stretching frequency
region, bands at 1036 and 1041 cm—1 are observed, indicative of free
and bound nitrate respectively. The relative concentration dependence
of these bands permits calculation of a, the fraction of free nitrate
ions, as a function of concentration. The band contours of the CEN
and C-C stretching frequencies of the solvent are each resolved into
two components, indicative of salvation of the metal ion. This study
confirms the conclusions of an earlier infrared investigation of
complexes of AgNOS-CHSCN and AgNOS-ZCHSCN by the same workers (82).
Addison, et: 31, (83) examined infrared and Raman spectra of solutions
of zinc, cadmium and mercury(II) nitrates in acetonitrile and found
the nitrate spectra to be consistent with a strong perturbation of
nitrate ions by solvated cations. Only in the case of mercury(II)
was a metal-oxygen vibration frequency observed.

Vibrational studies of electrolytes in liquid ammonia have
recently received much attention. Gardiner, 33, 313 (84) examined
lithium and ammonium nitrates in liquid ammonia and noted large
changes in the N-H stretching region, 3000-3500 cm-l, which were
correlated with the metal ion salvation. Both solvent separated and
contact ion pairing were observed for lithium nitrate solutions, but
not for the ammonium nitrate solution. In addition, two bands at
561 and 361 cm"1 are assigned to cation—solvent, Li-N, stretching
vibrations. At the same time, Roberts, 33, 31, (85) studied the

effect of sodium perchlorate and iodide on the hydrogen bonded

structure of liquid ammonia. In addition to the expected changes

IS

in the N—H stretching region, a broad, weak, polarized band at
m200 cm.1 was observed in the sodium perchlorate spectrum. Two bands
at 325 and 450 cm.1 were observed for the sodium iodide solution.
In both cases, these anion dependent bands could not be attributed to
salute or solvent; hence, it was proposed that solvent separated ion
pairs were responsible. Plowman and Lagowski (86) examined Raman
spectra of ammonia solutions of some alkaline earth and alkali metal
perchlorates and nitrates. In all cases, the perchlorate bands were
unperturbed whereas the nitrate bands were split. Again, low frequency
bands assigned to the symmetric stretching mode of the solvated cation
2+ 2+

were observed for Li+, Na+, Mg , Ca , Sr2+, and Ba2+ at 241, 194,

328, 266, 243 and 215 cm.1 respectively.

CONCLUSIONS

 

It thus appears that spectroscopic techniques such as nuclear
magnetic resonance, infrared and Raman vibrational spectroscopy may
be used to characterize ion—ion and ion-solvent interactions in
solutions. The aim of such investigations would be to comprehend more
fully the effect of the solvent on these interactions and then to relate

these effects to physicochemical properties of the solvent.

GMWERII

EXPERIMENTAL PART

SALTS

 

Sodium tetraphenylborate (J. T. Baker), thiocyanate (Mallinkrodt),
perchlorate (G. F. Smith), iodide and bromide (Matheson, Coleman and
Bell) were of reagent grade and were used without further purification
except for drying. The first two salts were dried under vacuum at 60°C
for 72 hours, and the other salts were dried at 110°C for 72 hours.
After drying, all salts were stored in a vacuum dessicator charged

with granulated barium oxide.

SOLVENTS

Nitramethane (Aldrich) was fractionally distilled over granulated
barium oxide followed by drying over freshly activated Linde 5A
molecular sieves for 24 hours. Acetonitrile was fractionally distilled
over calcium hydride after refluxing for 48 hours. Sulfolane (Shell)
was purified by fractional freezing six times followed by fractional
distillation over sodium hydroxide pellets. Propylene carbonate
(Aldrich) was dried over Linde 4A molecular sieves followed by vacuum
distillation at 40 torr. Acetone (Fisher), ethyl acetate (J. T. Baker)
and tetrahydrofuran (Matheson, Coleman and Bell) were fractionally
distilled over calcium sulfate (Drierite), phosphorous pentoxide and
calcium hydride, respectively. Formamide (Fisher) was purified by
six repeated fractional freezings. Methanol (Baker) was fractionally
distilled over calcium sulfate. Dimethylformamide (Fisher) was dried
over Linde 4A molecular sieves followed by vacuum distillation over
phosphorous pentoxide. Tetramethylurea (Aldrich) and tetramethylguanidine
(Eastman) were purified by refluxing over granulated barium oxide

followed by fractional distillation under vacuum. Dimethylsulfoxide

16

17

was dried over Linde 4A molecular sieves and vacuum distilled. Ethanol
(commercially available, 200 proof) was fractionally distilled over
calcium hydride. Pyridine (Fisher) was refluxed over granulated barium
oxide and fractionally distilled. Hexamethylphospharamide (Aldrich)
was vacuum distilled over granulated barium oxide at 20 torr. Acetic
acid (Baker) and formic acid (Baker) were each purified by six repeated
fractional freezings. Purified solvents were stored over Linde 4A
molecular sieves in brown bottles with the tops covered with cellophane
wrap. Important solvent properties and solvent abbreviations to be

used in this thesis are listed in Table 2.

SAMPLE PREPARATION

 

For the NMR concentration studies, stock solutions of the sodium
salts, generally 0.50 M, were prepared by weighing out the desired
amount of salt into a 5 ml volumetric flask and diluting to the mark
with solvent. The remaining solutions were prepared by appropriate
dilutions of the stock solution.

The mixed solvent solutions were prepared by taring a snap cap
vial, adding the desired volume of solvent A, weighing, adding the
desired amount of solvent B and weighing again. Knowing these weights,
the number of moles of each solvent and resultant mole fraction were
calculated. Approximately 0.1171 g of sodium tetraphenylborate was
weighed into a 1 ml volumetric flask and the desired solvent or solvent
mixture was then added up to the mark (No.50 M_Na+).

Samples for the infrared and Raman studies were prepared in 1116

Same manner.

18

«m.Hm

w.mN

m.wN

0.0N

«m.mN

an.vN

o.oN

H.NH

o.NH

H.mH

w.vH

H.vH

N.N

 

Hopesz

Hocoo m.:cm5uso

 

mm.e~ mam.o
wo.oe omo.o
oo.m~ fine 0
Ha.cm oom.o
k.mm omm.o
m.aoa 2mm.o
mm.a aam.o
No.c omm.o
k.om oce.o
o.mc oac.o
o.me weo.o
w.wm amm.o
s.mm Hmm.o
Beachcou OH x a-

uwspoofiofla Nuflflfinflpmoomom oflouoEDHo>

Hocmcum

AOmZQV ovwxomHSmecuoEMQ
mazev moasfixnuoemauoe
flazev oeAEaenomHEESosaa
Hocmguoz

oowEmEpom

AmIHV Caedmopwxnmhuoh
oompoomaxcum

o:ouoo<

numv oumconumo oonAQOHm
ocmfiomfism

Azuav oHAeoAcoeoo<

QCGSHGEOHHHZ

pco>How

mofluaomoom uco>aom xox .N manme

19

mm

coeoaeoea.

Howe:
wflo< ofiswom

Cflo< 0fluo0<

Auzkv ocfleficeomflxguoEmoooH
m<mzzv oofiEmsozmmozmezuomeo:

acacuesa

flooscsocoov N oases

20

For the conductance studies, the stock solutions were prepared by

weighing solvent into a flask containing previously weighed solute.

INSTRUMENTAL MEASUREMENTS

 

Nuclear Magnetic Resonance

 

Sodium-23 nuclear magnetic resonance measurements were made on a
highly modified NMRS—MP-IOOO spectrometer operating at 60 MHz at a
field of 53 k0. The time-sharing method of Baker, gt: 31: (87) with
frequency sweep was used. Several crossed coil 5 mm probes were used.
Spectra were time averaged for a few scans on a Nicolet 1083 computer.

One modification of the MP-lOOO was the replacement of the supplied
function generator for frequency sweep with the voltage ramp output of
the Nicolet 1083, which drives the analog frequency sweep of a General
Radio 1164-A frequency synthesizer. The ramp from the 1083 is
proportional to the digital word corresponding to the memory address
being loaded; hence, the frequency of any given point on the display
may be determined with a digital counter by stopping the sweep at that
point. The frequencies of all peaks reported were measured in this
manner.

The Wilmad 506-PP, 5 mm 0D polished nmr sample tube was fitted with
a Wilmad precision coaxial 520-2 nmr tube for the reference solution.
The reference for the 23Na measurements was 3.0 M_aqueous sodium
chloride solution; however, when the chemical shifts were so small that
the sample was masked by the reference, a secondary reference of 2.5 M
sodium perchlorate in methanol was used. In this latter case, the
shifts were corrected so as to apply to the 3.0 M aqueous sodium chloride

reference solution. A positive shift from the reference is upfield.

21

The Chemical shifts reported are corrected for differences in bulk
diamagnetic susceptibility between sample and reference according to the

relationship of Live and Chan for high-field spectrometers (88).

= 6 _ 4w (xref _ Xsample)

corr abs '3— v v (1)
For low field spectrometers, such as the Varian DA-60, the correction

for the observed chemical shift is given by:

_ ___ ref _ sample
Scorr _ 6abs + 3 (Xv Xv

) (2)

For each solution, the contribution of the salt to the susceptibility

of the solution is assumed to be negligible - a reasonable assumption

as shown by Templeman and Van Geet (<0.05 ppm) (35). Hence the corrections
simply reflect differences between the aqueous reference and the organic
sample. For the mixed solvents, the volume diamagnetic susceptibility

of a given mixture may be calculated by Wiedemann's Law:

Xcalc = V7—7£N7_ - X3 + I7—€?77— - X5 - (3)
V AB AB
calc . . . .
where X = calculated volume susceptibility of the mlxture
VA’VB = volumes of solvents A and B, respectively
X3,X5 = volume susceptibility of pure solvents A and B, respectively

Chlorine-35 nmr measurements were performed on the Varian DA-60
Spectrometer operating at 4.30 MHz at an applied field of 10.39 kG
using the 14N Varian probe. The chart paper was calibrated by the
sideband technique (111) where an audiofrequency of 600 Hz was super-

imposed on the radiofrequency field. The samples to be analyzed were

22

placed in a standard 15 mm test tube with a caaxially mounted 8 mm
nmr tube, which contained the saturated sodium chloride reference
solution. A negative value for the Chemical Shift is dawnfield. For
the linewidth studies, signals were recorded in the dispersion mode

and the linewidth in Hz measured as the peak to peak distance.

Infrared Spectra

 

Far infrared (fir) measurements (600-50 cm-l) were performed on a
Digilab FTS—16 Fourier transform spectrometer. The theory and operation
cf this instrument have been described by P. Handy (71). Most of the
spectra were obtained at a nominal resolution of either 2 or 4 cm-l.
A standard demountable Barnes liquid cell with polyethylene windows and
a 0.1 mm path length was used. 0

Near infrared measurements in the 4000-600 cm.1 spectral region were
Obtained on the Perkin Elmer Model 225 Spectrophotometer. The solution
samples were held between potassium bromide salt plates in a standard

demountable Barnes cell holder. The cell pathlength was varied from

0.1-0.015 mm by the use of teflon spacers.

Laser Raman Spectra

 

Raman spectra were obtained on the Spox Ramalog 4 Laser-Roman system
equipped with the Spectra-Physics Model 164 Argon-Ion Laser. The
5145 A line was employed for excitation and data were obtained in the
PUIse counting mode with a nominal resolution of 2-4 cm-l. The
i“Strument may be interfaced with the Digital Equipment Corp. PLP-S/E
18b mini-computer to record data and transfer to cards for computer

analysis, an option that was employed in this thesis work. Samples were

injected into 1.6-1.8 X 90 mm melting point capillary tubes and sealed.

23

Conductance Measurements

 

The conductance bridge has been previously described (90) and was
operated at 1000 Hz. The cells, constructed similarly to those of
Daggett, Bair and Kraus (91), were steamed for at least one hour
followed by oven drying at 110°C. Calibration with standard aqueous
potassium chloride solutions gave cell constants of 0.0826 :_0.0001,
1.442 :_0.001 and 3.916 :_0.001 cm-l. All measurements were made at
25.00 :_0.02°. A Sargent S-84805 thermostatic bath assembly filled with
light mineral oil was used for temperature control. Specific conductances
of the pure solvents were found to be negligible with respect to the
conductances of the solutions even at the lowest concentrations.

All solution transfers to the cell were made by means of a weight
buret. To maximize accuracy over the entire concentration range

5-5 x 10'1

(NIO- m), the final solutions were prepared by one of the
following methods. Method A involved pragressive addition of stock
solution to the cell which contained a previously weighed amount of
solvent. Conversely, method B involved progressive additions of the
solvent to the weighed stock solution in the cell. In both methods,

the cell contents were thoroughly mixed, temperature equilibrated and

resistance measurements obtained.

Data Handling

 

Extensive use of the CDC-6500 computer was made to evaluate data.
Pragram KINFIT (98) was employed to evaluate ion pair formation
constants and complexation constants. Program SHEDLOV was used to
evaluate the conductance data and program PEAKSBF and KINFIT were
employed to fit and resolve Raman spectral data. The application of

these programs is described in the appendices.

CHAPTER III

A SODIUM-23 NUCLEAR MAGNETIC RESONANCE AND
ELECTRICAL CONDUCTANCE STUDY OF CONTACT ION

PAIRS IN NONAQUEOUS SOLVENTS

INTRODUCTION

 

A large majority of the chemical reacrions used industrially and
analytically as well as those in biological systems, occur in solution.
Yet, until recently, the role of the solvent in these reactions has
been largely ignored, with the solvent considered as an inert diluting
medium. However, the choice of solvent may affect the equilibrium and
the rate and mechanism of a reaction; hence, we should strive to
develOp some generalizations concerning the choice of an Optimum
solvent for a given reaction.

Such generalizations require a sound knowledge of solute-solute,
solute-solvent and solvent-solvent interactions. A prime reason for
a lack of this knowledge lies in the dynamic complexity of solution
structures and a lack of adequate experimental techniques to study‘
them.

In recent years, spectroscopic techniques such as far infrared,
near infrared, Raman, proton and non-proton nuclear magnetic resonance
techniques have made it possible to get a new perspective on the
nature and structure of species in electrolyte solutions. Alkali metal
NMR, particularly sodium-23 NMR, has proven a sensitive probe of the
immediate chemical environment of alkali metal ions (16,27-48).

The purpose of this study is to elucidate the role of the solvent
in determining the species present in nonaqueous solutions of sodium
salts by use of sodium-23 NMR as the primary probe. By carefully
extending the earlier fundamental studies of Deverell and Richards
(16), Bloor and Kidd (27) and Erlich and Popov (34), it is hoped to
identify a key solvent property or combination of properties which may

serve as a useful guide in predicting electrolyte solution structure.

24

25

RESULTS AND DISCUSSION

 

The chemical shift of the sodium—23 resonance with respect to
3.0 M_aqueous sodium chloride was determined for sodium tetraphenyl-
borate, perchlorate, thiocyanate, iodide and bromide over the concentration
range 0.01-0.50 M_in various solvents. The data are presented in
Table 3. With few exceptions, the 23Na Chemical shifts for solutions
of the latter three salts exhibit marked concentration dependence,
whereas concentration independence is noted for solutions of the
tetraphenylborate and perchlorate salts. The results obtained in
l,l,3,3-tetramethylurea (TMU) and N,N-dimethylformamide (DMF), shown
in Figure 1, illustrate this behavior.

The dielectric constants (D) of DMF and TMU are 36.7 and 26.0,
respectively. Since the 23Na chemical shifts reflect changes in the
inner salvation sphere of the cation and since these shifts seem to
converge at infinite dilution to a value indicative of free sodium
ion solvated by TMU or DMF, it seems reasonable to assume that the
dawnfield shifts occuring with increasing concentration far sodium
thiocyanate, bromide and iodide, are due to the formation of Contact
ion pairs. Conversely, the concentration independence of the chemical
shifts for sodium tetraphenylborate and perchlorate is indicative of
either free solvated cations or solvent separated ion pairs. Hence,
chemical shift measurements can differentiate between contact ion
pairs and free solvated ions and/or solvent separated ion pairs, but
apparently not between free solvated ions and solvent separated ion

pairs.

26

 

 

 

 

 

 

 

 

 

 

TETRAMETHYLUREA
D _J, n .9 c_i°________4) CHC24
3 .1 E; e a W BPh4
2 _—
1 ... SCN
0 — I
1 ..
E DIMETHYLFORMAMIDE
‘3' o .. 004
2'1 3 " BPh4
—‘ I
.. *0 SCN
:3 _—
2 -t
1 -1
Br
() .-
l I l I l l
0.0 0.1 0.2 0.3 0.4 0.5
CONCENTRATION (M) 1

Figure l. Sodium-23 Chemical Shifts of Various Sodium Salts in
l,l,3,3-Tetramethylurea and N,N-Dimethylformamide

27

0m.m
mm.0t
mv.mt
mN.Hn
mv.0
0N.0

00.v
mn.m
mN.v
0H.n
05.5

mv.0

H0.0

m0.o
mw.mt
mm.mn
mN.Ht
mm.0
m~.0
m0.N
m0.v
00.m
0m.v
0H.n
0N.w
mm.w
0v.0
00.0
mH.n

m0.0

mw.m
m0.mt
mv.mt
0m.~t
0v.0
mH.0
mm.N
m0.v
mn.m
mN.v
mH.n
0N.w
mN.w
mm.0
00.0
0N.h

0H.0

00.0
05.0:
0m.mt
mN.Ht
mm.0
mN.0
m0.N
m0.v
mn.m
0m.v
m~.n
0m.w
0m.w
0N.0
mw.w
mw.o

seamaz
om.o

mmv cofiumaucoocou

oflo< aflenoa

oa.m om.c er< oaooo<
o~.oH- oo.oH- mo.oH- ocaowcasmaseHosaepoe
om.m- mm.m- oc.m- ocaeoeoeanoeaaseHoEAAom
om.H- mN.H- m~.H- acacaesa
mc.o ow.o oo.~ Hocoeem
om.o oe.o me.o oeaxoMHsoaxeeoEAa
cm.m mw.m ow.~ noesHAEDoEaeooe
mm.e mm.e oo.m oefieaeeomHAEeo5aa
ma.m ow.m ow.m Hosanna:
mm.s H H ocanenoa
mH.R om.k me.a caesmoecaesneoa
on.w mk.m om.w ooaoooaHAESm
om.w 0m.w mm.w acouoo<
0N.0 mv.m mm.m oumconamu ocoaxmoam
mw.w ma.w ma.w ocaaomHsm
mH.E mH.a ma.“ oaaeoacoooo<
QCMSHQEOHHNZ

om.o oe.o om.o oeo>aom
ausz msooss< m o.m a> macaem Haoaeoeu oz .m oases

MN

28

00.0
00.0:
00.0:
0v.0:
00.0
00.0

00.0

0~.v

00.0

00.0

0v.5

00.0
00.0
00.0:
00.0:
00.0:
0H.H
00.0
00.0
00.0
00.0
0N.v
00.0
00.0
05.0
00.0
00.0
00.5

00.0
0H.0
00.5:
0v.0:
00.0:
00.0
00.0
0H.0
00.v
00.0
00.v
00.0
0N.0
05.0
00.0
00.0
0v.5

00.0
0N.0
00.5:
0v.0:
0H.0
00.H
00.0
00.0
00.0
00.v
0N.v
00.0
00.0
00.0
00.0
0N.0
0v.5

ofiumz

00.0
0N.0
0~.5:
0v.0:
0v.0
05.H
0N.0
0N.0
0H.0
0H.v
00.?
00.0
0v.0
05.0
00.0
0N.0
00.5

00.0
0N.0
00.5:
0v.0:
0v.0
00.H
v0.0
0N.0
00.0
0H.v
0H.v
00.0
00.0
05.0
00.0
0H.0
00.5

00.0
00.0
00.5:
0v.0:
05.0
0H.N
0v.0
00.0
00.0
0N.v
00.0
00.0
05.0
00.0
00.0
0N.0
05.5

aflo< oafihom

ufio< uapoo<
ocfivwemawfixnuoamnpoe
ovflsmuocmmozmaxguoamxo:
ocwvwnxm

Hocmnom
ooaxomHsmHREDosAa
mannaxcuoEMHuoh
owMEmEHomenuoafia
Hocmnuoz

owMEHBHOm
:MHSMOHvxnmnuoh
oumuoomfixnum

o:ouoo<

oomconhmu oaoaxmonm
ocmaomHsm
mafinuflcouoo<

QCNr—HOEOHHMZ

Aconcaocouv m oases

30

00.0
0N.00:
00.0:
0v.5:
00.0
00.0
00.N
00.v
00.0
0N.v
00.0:
0

00.0
00.0

0v.0
00.0
00.00:
00.0:
0N.0:
00.0
00.0
00.0
00.v
00.0
0N.v
0N.0:

05.0
00.0
00.0
00.0

0v.0
0v.0
00.00:
00.0:
0N.0:
00.0:
00.0
00.0
05.0
00.0
0N.0
0v.0:

0N.v
00.0
00.0
00.0

om.w
ce.e
when

HI

00.0:

00.0
00.0
00.0
00.0
00.v
0v.0
00.0
00.5

00.N
0N.0
00.5
00.0

sz

oc.w
mm.c
oo.aa-
oc.m-
o2.a-
ca.o-
mo.o
oe.o
mm.e
m.m

00.0

00.N
00.5
00.5
00.v

on.w
mm.c
mm.ma-
mc.m-
om.a-
mm.o-
ea.o
mo.o
mm.e
om.m
ma.e

00.0
00.5
00.0
00.0

00.0
0v.0
00.00:
05.0:
00.0:
0N.0:
00.0
00.0:

0N.0
00.v

00.0
00.0
00.0
0v.0

00o< O0Euom

00o< owuoo<
0:000cm50050uoemuu05
000amnocmmocm0xapoemxoz
0:000me

00cmzum
ocaxomHSoHAEHoeAa
moas0xnuosmuuob
o005m500005000600
00cm0uoz

0005msnom
amasmonwxcmnuoh
oumHoOM0xnum

onouoo<

opmconumu oco0xmoum
o:M0ow0:0
o00Hp0couoo<

ocmnposoau0z

mooscfiecouv m oHoee

31

00.0:
0N.v
00.0
00.0

00.0:
00.0
0N.v
00.0

00.0:
00.N
0N.0
00.0

00.0:
00.0
00.0
0v.0

Hmmz

00.0:
0v.0
00.v
00.0

0N.0:
00.0
00.0
00.0

00.0:
00.0
00.v
00.0

ocaxomHsAerDoeAa
000EmEa00050u0600
0000:8890"—

accesses

fleescaocouv m oaooh

32

Figure 2 displays the data obtained in sulfolane (D = 44.0) and
acetone (0 = 20.7). As in the previous case, the thiocyanate and the
iodide salts exhibit marked concentration dependence attributable
to contact ion pair formation whereas the perchlorate and tetraphenyl-
borate salts are relatively concentration independent. Moreover, the
shifts seem to be converging toward a single point at infinite
dilution characteristic of free sodium ion solvated by acetone or
sulfolane respectively.

It is interesting to note that in formamide (D = 109.5)

(Table 1) the chemical shifts of all salts used do not display
concentration dependence, and the positions of the 23Na resonance for
each salt are almost identical (~4.20 ppm). These data suggest that
over the concentration range studied (0.01-0.50 M) contact ion

pairs do not exist in these solutions. Classically, such behavior

is to be expected in solvents of high dielectric constant.

However, the 23Na chemical shifts in solvents of very low
dielectric constant such as tetramethylguanidine (TMG) (D = 11.0),
pyridine (D = 12.3) and tetrahydrofuran (THF) (D = 7.6) also show
little or no concentration dependence from 0.01-0.50 M_as shown in
Figures 3 and 4. It should be noted, however, that in these
solvents the chemical shifts do not converge at lower concentrations
as they do in solvents of medium and high dielectric constant. In
solvents of low dielectric constant, the concentration of free ions
is vanishingly small in the 0.01-0.50 M_range, and the predominant
species must be contact ion pairs. It should be noted that in the
chemical shift vs concentration plots, the lines do seem to be curving

upward in the 0.01-0.l M_concentration range where one would expect

33

 

 

 

 

 

 

10 __ SULFOLANE

9 .—

E} .—

7 0

6 -—J

5 _J

9 __ ACETONE
E 8 _ c s v——1——-——o-——-° 0%,,
G.
d 7 —'i

 

 

 

 

I I l l l l
0.0 0.1 0.2 0.3 0.4 0.5

CONCENTRATION (M)

Figure 2. Sodium-23 Chemical Shifts of Various Sodium Salts in
Sulfolane and Acetone

34

 

 

 

 

 

 

 

 

 

 

 

1 _ PYRIDINE CIO
W e 4" 4
'1 -!:— 3“ u ”if “—U 9T* 49 BPh4
-3 '- v—r : 40 v 4 SCN
-5 .0.
.37 .—
2;- e
a. -9 - ° I
0.
<1 -6 __ TETRAMETHYLGUANIDINE
K n o C —-d CIO4
-{3 .0
oo\tt\<>
. 0 a 9 e SCN
'10 -‘\ ‘U‘ 0 J3 BPh4
-1;Z .—
-14 -°\°\°\ ° . .
e I
I I l I I T
0.0 0.1 0.2 0.3 0.4 0.5

CONCENTRATION (M)

Figure 3. Sodium-23 Chemical Shifts of Various Sodium Salts in
Pyridine and 1,1,3,3-Tetramethy1guanidine

35

 

 

 

 

10 '—
8 _ 9-9 Jr 0 9 00;) 00,,
W 9* 17—4 BPh4
6 ..
4 _
2;. _ 0—-0 SCN
(1. 2
0.
fl 0 __,
-2 .1
-4 .—
-6 —\9\0
I
‘3 I I I T l
0.0 0.1 0.2 0.3 0.4 0. 5
CONCENTRATION (M)

Figure 4. Sodium-23 Chemical Shifts of Various Sodium Salts in
Tetrahydrofuran

36

ion pair dissociation to become observable. The present detection
limit of 0.01 M solutions prevents investigations at lower
concentrations.

The conductance of Nal solutions in TMG and pyridine was determined

5-5 X 10‘1 M concentration range. The data are shown

in the l x 10‘

in Figures 5 and 6, respectively. It is seen that ion pair dissociation

is negligible in solutions with concentrations above 10.2 M, The

conductance data were analyzed by the Fuoss-Shedlovsky technique (92).

For TMG, A0 = 51.6 and the ion pair dissociation constant, Kd =

6.2 x 10‘5. In pyridine, A0 = 73.4 and Kd = 3.91 x 10‘4. The latter ;

I

value agrees with Kd = 3.7 X 10-4 as determined by Burgess and Kraus

(93). No literature data seem to be available for the conductance of

sodium iodide in TMG. A literature search for conductance studies of

the other salts in these three solvents revealed only the following

data for sodium tetraphenylborate in THF; A0 = 88.5, Kd= 8.52 X 10"5

(94,95). Hence, the conductance data indicate that for the salts

studied in these low dielectric solvents, ion pairing phenomena

prevail over the concentration range examined in the NMR study.
Results obtained in hydrogen bonding solvents such as methanol

(0 = 32.7) and ethanol (D = 24.5) show one significant difference as

compared to the other solvents studied thus far. As shown in

Figure 7, the 23Na resonance observed for NaClO4 in both Solvents and

NaBPh4 in ethanol shifts upfield with increasing concentration, while

in the case of NaI and NaSCN the cOrrespoNding shifts are dawnfield.

Upfield shifts have previously been reported for aqueous alkali metal

nitrate solutions (16). With increasing concentration, the alkali

metal resonances of the nitrate solutions shifted upfield while those

37

50'-

40—?

 

30"

-91r-c-t-—$

Atom/Is"CMZEOUIM")

 

 

 

20'—
10—
' 3’ 313 953 ::: e 3 3 =$-o a a a
0‘ l I l l I l I

00 01 a2 03 a4 05 06 D7
{CU2 ~

Figure 5. Conductance Curve for Sodium Iodide in l,l,3,3-Tetramethyl-
guanidine

38

70-1

60-

0"
O
l
w—cfl“

A (OH MS'ICMzEQUIV.'1)
00 «b
o O
l :

200

10"“

 

 

T l l l l I I l
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

CI/ 2

Figure 6. Conductance Curve for Sodium Iodide in Pyridine

39

METHANOL 004
‘4 .7:y—<T——'Gr-—————LL——-—"'£)——'————‘>—“————‘>
A o-——o 09:14

L)

Q U G U
W SCN
. Br,I

ETHANOL 004

BPh4

'0 SCN

 

 

 

 

 

'1 I l l l i
0.0 0.1 0.2 0.3 0.4 0.5

CONCENTRATION (M)

Sodium—23 Chemical Shifts Of Various Sodium Salts in

Figure 7.
Methanol and Ethanol.

40

in alkali halide solutions shifted dawnfield. Similar behavior has
recently been observed by Van Geet, again, in aqueous solution (35).
While both upfield and dawnfield shifts with increasing salt
concentration are due to increasing cation-anion interaction, it has
been postulated that upfield shifts occur when a water molecule in the
cation salvation shell is replaced by an anion, which results in a
decreased electron density around the alkali cation. Although it is
possible that the same explanation applies to the upfield shifts
observed in methanol and ethanol solutions, a detailed NMR and
vibrational study of systems exhibiting upfield shifts was pursued.
This study is presented in Chapter V of this thesis.

'To this point, we have been examining contact ion pair formation
in several nonaqueous solvents and rationalizing the data with respect
to the dielectric constant of the solvent. Consider the results
obtained in propylene carbonate (PC) (D = 65.0) and dimethylsulfoxide
(DMSO) (0 = 46.7), as shown on Figure 8. In PC, contact ion pairing
occurs in iodide and thiocyanate solutions, whereas perchlorate and
tetraphenylborate solutions are composed of free solvated iohs and/or
solvent separated ion pairs. In DMSO, a solvent of’lggg£_dielectric
constant, not only is there no contact ion pairing in the perchlorate
and the tetraphenylborate solutions, but there is no contact ion
pairing in the iodide or the thiocyanate solutions as well. It
appears that, to a large degree, ion pair formation is closely related
to the bulk solvent dielectric constant; however, these studies show
that other factor(s) may be important. After defining a quantity

which describes solvent donicity on a molecular basis (as opposed to

41

10 __1 PROPYLENE CARBONATE

 

 

G O'——") C|O4
_ 0 o
C} -—— BPh4

   
 
  

 

 

 

 

 

 

 

9 -—I
8 .—
'7 .—
5? I
g] 6 _4 ° 0 0a SCN
‘I — DIMETHYLSULFOXIDE
BPh4
O _ .1--. ‘— I"! ‘)‘——() I
SCN
-1 __
Br
-;Z._a
I I l l I N
0.0 0.1 0.2 0.3 0.4 0.5

CONCENTRATION (M)

Figure 8. Sodium-23 Chemical Shifts of Various Sodium Salts in
Propylene Carbonate and Dimethylsulfoxide

42

bulk preperties) we will investigate the interplay of solvent
dielectric constant and donicity on contact ion pairing phenomena.

The frequency of the 23Na resonance of sodium ion in solution is
affected primarily by the nearest neighbors of the ion, hence if
the effects of cation—anion interactions were removed from the system
under investigation, the observed resonance would indicate solely
the interaction of the ion with the solvent. Hence, the variation of
the 23Na chemical shift with solvent may be explained by the ability
of the solvent to alter the electronic environment about the sodium-23
nucleus, which should be related to the donor ability of the solvent.
An empirical approach to this donation or camplexing ability of the
solvent is given by Gutmann's donor numbers (96). These are simply
the enthalpies of complex formation between the given solvent and
antimony pentachloride in 1,2-dichloroethane solution. As shown by
Gutmann, the donor numbers may be quite useful in predicting the
behavior of nonaqueous systems.

It is reasonable to expect, therefore, that there might be a
relationship between the donor numbers and the relative chemical shifts
for the solvated sodium ion provided that the chemical shift is
unperturbed by cation—anion interactions. A plot of the infinite
dilution 23Na chemical shift in a given solvent versus the donor
number of that solvent, shown in Figure 9, reveals that this is indeed
the case. The linearity of the plot shows that the relative 23Na
chemical shifts yield useful information on the solvating abilities

of the solvents at least vis a vis the sodium ion.

43

A

53
I

K

APPM

 

 

'4 I | l l
0 10 20 30 40

DONOR NUMBER

Figure 9. Plot of Infinite Dilution Sodium-23 Chemical Shifts versus
the Donor Number of the Solvent: (1) Nitramethane,
(2) Acetonitrile, (3) Sulfolane, (4) Propylene Carbonate,
(S) Acetone, (6) Bthylacetate, (7) Tetrahydrofuran,
(8) Dimethylformamide, (9) Tetramethylurea, (10) Dimethyl-
sulfoxide, (11) Water, (12) Pyridine, (l3) Hexamethyl-
phospharamide, (14) Formamide, (15) Methanol and
(16) Ethanol

44

The plot of infinite dilution chemical shifts as a function of
donor numbers was fitted by linear least squares analysis to yield

the following equation:

chemical shift = (-0.5074)(DN) + 16.6105 (3)

The donor numbers of methanol, ethanol and formamide have not been
determined experimentally but from the above equation they can be
predicted to be 25, 32 and 24 respectively, again provided that the
chemical shifts are unperturbed by ion-ion interactions.

It should be noted that, while the salvation process involves
electrostatic ion-dipole interactions, donor numbers represent the
enthalpy of formation of a covalent bond between a given solvent and
antimony pentachloride. A priori it seems, therefore, that a
parallelism between the solvent's ability to solvate alkali metal
cations and its tendency to form covalently bonded complexes with
SbCl5 should not be expected. As Gutmann points out, however, the
fact that such parallelism exists suggests that some covalent
interaction is involved in the salvation of sodium ion by donor
solvents and that the extent of covalency in the salvation bonds
increases with the solvent donicity (97). On the other hand, this
parallelism may suggest that some electrostatic interaction is
involved in the solvent-antimony pentachloride bonding.

The fact that donor properties of solvents can strongly influence
ion pair dissociation has been shown by Gutmann (97). The data

reported in this thesis suggest that both solvent dielectric constant

and donicity influence ion pair dissociation, as summarized in Figure 10.

45

 

10's
9 ..
3 "'I
SULFOLANE
7 " o
6 ..
E 5 ._
0..
a. FORMAMIDE
<1

 

 

03—00

0
(I
C
O

 

 

I I I I I
0.0 0.1 0.2 0.3 0.4 0.5

CONCENTRATION (M)

Figure 10. Sodium-23 Chemical Shifts of Sodium Iodide in Sulfolane,
Formamide, Acetone and Dimethylsulfoxide

46

For example, in formamide solutions (D = 109.5, DN = 24) the formation
of contact ion pairs in the 0.01-0.50 M_concentration range would not
be expected due to the high dielectric constant and medium donor
number of the solvent. The data seem to confirm this conclusion.
The comparison of dimethylsulfbxide (D = 46.7, DN = 29.8), sulfolane
(D = 44.0, DN = 14.8) and acetone (D = 20.7, DN = 17) shows the
importance of the donicity of the solvent in contact ion pair
formation. While dimethylsulfoxide and sulfolane have nearly
identical dielectric constants, the former is a much better solvating
agent, hence contact ion pair fOrmation is expected to be more probable
in the latter solvent. Indeed, the 23Na chemical shift fbr NaI
shows no concentration dependence in dimethylsulfoxide but a marked
concentration dependence in sulfolane. 0n the other hand, acetone
and sulfolane have nearly the same donicity, but the latter has
a much higher dielectric constant. Qualitative indications from the
data reveal the tendency toward ion pair formation in acetone is much
higher than in sulfolane. Additional evidence has been obtained in
propylene carbonate solutions (D = 65.0, DN = 15.1). As shown in
Figure 8, the 23Na chemical shifts for sodium iodide solutions are
stronger concentration dependent, which indicates that despite the
high dielectric constant, the formation of contact ion pairs occurs
to a considerable extent.

While to this point only qualitative indications of ion pair
formation have been obtained from 23Na chemical shifts, it seems
reasonable to assume that these data can be used for the determination

of ion pair formation constants. Since only one signal is observed

47

for the 23Na resonance in solution, it may be assumed that exchange
is rapid compared to the NMR time scale; hence, a population weighted

average shift is observed and is described by equation (4):

sobs = XF6F + XCGC (4)

bs is the observed chemical shift in ppm, 6F is the chemical

shift characteristic of free solvated sodium ion in a given solvent

where 5
0

in ppm, 6C is the chemical shift characteristic of the contact ion

pair in that solvent in ppm, and X and XC are the fractions of free

F

and complexed sodium ion respectively. Now,

sobs = XFGF + (1 - XF)6c (5)
Rearranging,
Gobs = XF(<SF - 6C) + 5c (6)
By definition,
XF = 2?; (7)
T

where C? is the analytical concentration of free Na+ and C? is the

. . + . . .
total analytical concentration of Na ; substituting (6) into (7),

we obtain equation (8)

5 =

obs (6F ‘ 6c) + 6c (8)

Ola
H3713

The equilibrium constant for the reaction

M+A=M,A (9)

48

may be written as

 

 

+ ..
K = —-——————-[M+’ A_] (10)
[M HA ]
Again, by definition
+ - M M
[M , A ] — CT - CF (11)
and
P
[M“] = [A‘] = (2:3 (12) '
Substitution of (11) and (12) into (10) yields
Cde ' CF
K = T?“ (13)
(CF)
Rearranging and setting the equation equal to zero, we obtain
M2 M M
CFK+CF-CT=O (14)
The solution to this quadratic is written as,
M 1/2
M -1 :_(1 + 4KCT)
CF = (15)
2K
Substituting (15) into (8), we obtain
M 1/2
-1 1-(1 + 4KCT)
Gobs = [ M ] (5F - 6c) + 6c (16)
ZKCT

Inspection of equation (16) reveals that the observed chemical

shift is expressed in terms of the known C? and SF and the unknowns

6c and K. Since the expression C¥ is obtained by solving a quadratic

49

equation, two solutions are obtained. However, only the positive (+)
value in equation (15) was found to yield a reasonable fit. The
procedure is to input the experimental éobs and known Cg parameters

holding 6 constant and vary the K and 6c until the calculated chemical

F
shifts correspond to the observed chemical shifts within a preset error
limit. This procedure requires that good initial values for 6c and

K be read in for the KINFIT program to converge readily.

Ion pair formation constants for sodium thiocyanate, iodide and
bromide in several solvents are presented in Table 4. It should be
noted that the K value is a concentration K and not a thermodynamic K
where activity corrections are applied. Also, because so few
experimental points were used, these K's represent little more than
the order of magnitude of ion pair formation. The numbers in

parentheses next to the nmr K's are the K's obtained from conductance

data with the appropriate reference denoted.

SO

.osam .m .m paw ocfipm .m

 

a «go. a a
.naomfiv mmee Ha .eflnw among .2 .m ecm Hofimmso .4 .m flow .mmmmfiv 6H

 

.moomfiv 5mm“ .ww ..oom smumpma .mcmge

 

Amy manomfiv ”OHH .mw ..Em;u .n .cmu .ofime .m .2 new News .5 .0 new

 

8?
ecu moe< .a .o any “fluamfiv Hema .mm ..oom .aogu .Hos< .w .0msgau

ofla.4v om.o + mm.o
eflmnv mw.m.u H.4H
Hm.N.H om.e
em.o.n Hq.¢
w.om.n e.mm
Haz
mzz «2

mm

Ha.o + ma.m
o.om.n
Umofiv mo.e.u N.oH me.e.n
nflmfi.ov mo.H.H wm.a nnmfl.oV Na.o.n
aflN.Hmv mm.v.n
w.NH.n
o.oH.u
gmmz zummz
Ind
fla-zv x

 

.wm ..Eo;u .msna .n .muaom .u .a

.< .w cam smegmm .n .m Amy

n.mm

m.NH

Hm.v

m.om

n.0H

0.0m

ocmfiomflsm
mnuxomfismflsgpmaflo
ocouoo<

Hocmnuoz
oufisaeuomflscumefia
nonnaxcuoamnuoh
Hocmnum

opmconhmu ocoaxmonm

pco>aom

xn madam Esflpom msoflum> How mucmumcou :ofiumahom yawn :oH .v manna

CHAPTER IV

STUDIES OF PREFERENTIAL SOLVATION OF THE

SODIUM ION IN MIXED SOLVENTS BY SODIUM-23 NMR

INTRODUCTION

 

In the preceding chapter, we saw that NMR spectroscopy may
serve as a valuable tool for the elucidation of the structure of
electrolyte solutions. The variation of the chemical shift as a
function of solvent, anion of the sodium salt and concentration allows
one to differentiate contact ion pairs versus free solvated ions or
solvent separated ion pairs. More importantly, the magnitude and
direction of the 23Na chemical shift varies linearly with the donicity
of the solvent while the resonance shifts further downfield as
solvent donicity increases.

These studies were extended to binary solvent mixtures where the
variation of the 23Na resonance as a function of solvent composition
was monitored and interpreted in terms of preferential solvation.

It is further hoped to rationalize preferential solvation as a
manifestation of solvent donicity.

When an electrolyte is dissolved in a binary solvent mixture, the
first solvation shell of the cation need not have the same composition
as the bulk solvent composition; it may preferentially contain one
solvent over the other. If the variation of the cation resonance is
monitored as a function of mole fraction of one solvent in a binary
mixture, generally, a smooth transition of the resonance from its value
in one pure solvent to the other is observed where the rate of
transition is dependent on the relative solvating abilities of the
two solvents in the given mixture. At some point in the study, the
chemical shift has progressed one half the way from one limiting

value to the other. This point has been defined as the equisolvation

51

52

or isosolvation point, where there is equal competition of each of
the solvent components for the cation. In this chapter, the
isosolvation point will be employed as a measure of preferential
solvation.

Frankel, etc 31: (99) examined preferential solvation of Co3+
ion (Co(acac)3) in chloroform-carbon tetrachloride mixtures by 59Co
NMR. Likewise, preferential solvation of Cr3+ ion (Cr(acac)3)
in the same solvent system was examined by observing the effect of the
paramagnetic Cr3+ ion on the transverse relaxation time (T2) of
the solvent nuclei (100). These same workers also studied 35Cl
chemical shifts in dimethylsulfoxide-water and acetonitrile-water
mixtures, but were not able to draw any conclusions concerning
preferential solvation (19). Bloor and Kidd (33) examined 23Na

resonance of sodium iodide solutions in acetonitrile-water mixtures

and concluded that Na+ ion was preferentially solvated by water.

RESULTS AND DISCUSSION

 

The chemical shift of the sodium-23 nucleus as a function of
solvent composition in binary solvent mixtures of nitromethane with
dimethylsulfoxide, pyridine, tetramethylurea, hexamethylphosphoramide
and acetonitrile; acetonitrile with pyridine, dimethylsulfoxide,
tetramethylurea and hexamethylphosphoramide; tetramethylurea with
pyridine and dimethylsulfoxide; hexamethylphosphoramide with
tetramethylurea, dimethylsulfoxide and pyridine, and dimethylsulfoxide
with pyridine using sodium tetraphenylborate as the solute are

presented in Table 5 and illustrated in Figures 11-14.

53

mN.m I 000.H

m~.m I 005.0 00.H I 000.H mN.n 000.H
00.m I w0m.0 0N.H I 0mw.0 0v.h N00.0
0m.m I m0¢.0 0m.0 I 000.0 mw.n H00.0
mm.H I m0m.0 00.0 Nov.0 00.0 H0m.0
m©.0 I 0m~.0 mw.H 00m.0 0N.0 0Hm.0
mm.~ m0~.0 0N.v 00H.0 m.0H v0N.0
0N.m mmo.0 mm.m HmH.0 0.HH ovH.0
mn.m 000.0 0m.n 000.0 N.NH 000.0
v.0H HN0.0 0.0a Nm0.0 o.mH wv0.0
m.mH 000.0 n.va 000.0 N.mH 000.0
200< <02: 02 Emm< mzHQHm>m m: Emmq qumHHZOHmu< 02
mz<szzomHqu<mzr m2<IPmZOmHqumzHOHm>0 mz<EHmZOMHququmHHZOHmU<

mthuxflz uco>Hom

thcfim How cofiuwmomsou uco>Hom mo coHuUCSm < mm mocmcomom MNIESflvom map mo :ofiumfihm> .m oHan

54

mo.HI

m5.NI

00.NI

0N.0I

00.0I

mm.mI

0m.mI

00.0I

00.0I

00.0I

mm.mI

200d

000.H

500.0

000.0

H05.0

«00.0

H00.0

00v.0

00N.0

mom.0

00H.0

000.0

mzHQHm>0 02

<0EIHMZHQH0>Q

00.N

0N.0

00.0I

mm.HI

05.HI

mv.NI

05.NI

0H.mI

0v.mI

mv.mu

mm.mI

200<

000.H

v00.0

505.0

005.0

m0m.0

00v.0

00v.0

v0m.0

v0H.0

500.0

000.0

02% m:

0N.0

mm.HI

00.NI

0v.mI

00.NI

00.mI

mH.mI

0~.mI

mm.mI

m©.mI

mm.mI

200<

<mzmnomza

000.H

000.0

000.0

000.0

000.0

N00.0

N0¢.0

V00.0

NON.0

000.0

000.0

0020 02

fleozcflpcoov m magma

55

0N.0

0v.0

0v.0

05.0

00.0

0H.H

00.H

00.H

00.H

0H.m

00.N

Zamq

000.H

H00.0

H00.0

005.0

000.0

00m.0

00v.0

00N.0

00N.0

00H.0

000.0

0020 02

mo.HI
ma.0I
00.0
mv.H
00.a
00.N
mH.m
0H.N
mN.m
00.N
00.N

Zamq

DEHHmZHQHm>0

000.~

N00.0

N55.0

N05.0

N00.0

000.0

v0v.0

50N.0

00N.0

mHH.0

000.0

mZHQHm>0 02

0N.N

mN.N

00.N

00.N

00.0

00.0

00.0

mo.v

00.0

05.0

Emm<

000.H

«00.0

m0m.0

000.0

H00.0

00H.0

0vH.0

000.0

0v0.0

000.0

32% 02

qumHHZOHmu<uszH

memscupcouv m magma

S6

00.0I

00.0I

mfi.mI

00.NI

00.NI

0N.NI

00.HI

m©.HI

vm.m

05.0

200<

000.H

N05.0

N0m.0

000.0

00N.0

00N.0

0vH.0

000.0

0v0.0

000.0

<02$ m:

qumHHZOHmu<H<mzm

mo.HI
mm.HI
mv.0I
00.0
0N.0
00.0
m0.H
0m.N
0v.m
m5.v
05.0

2004

000.H

H00.0

005.0

H05.0

H00.0

00v.0

NOV.0

000.0

HON.0

HOH.0

000.0

mzHon>m

mAHmHHZOHmu<HmzHQHm>m

02

0N.0
00.0
05.0
00.0
00.H
00.H
00.H
00.N
0N.0
05.0

200<

000.~

0H0.0

000.0

000.0

00N.0

va.0

m0~.0

00H.0

vHH.0

000.0

0020 m:

maHmBHZOHm0<Homza

hemsafiucouv m manna

57

00.0

00.0

00.0

00.0

00.0

0v.0

ma.0

0H.0I

mm.0I

00.0I

mv.HI

E004

000.H

.5v0.0
.005.0

.mom.o

000.0

000.0

00m.0

00H.0

000.0

0m0.0

000.0

0020 02

mzHOH0>0uomEQ

0N.0

00.H

05.0

0H.H

00.H

00.?

00.0

00.5

v.HH

0.0M

2004

000.H
005.0
000.0
N0¢.0
00N.0
NOH.0
V50.0
mmo.0
mm0.0

000.0

0020 02

mz<EHMZOmHHZHOmEQ

00.N
00.N
00.N
00.N
05.0
00.5
00.0
0.HH
N.NH
0.vH

200d

000.“

N05.0

000.0

000.0

~0N.0

500.0

050.0

0v0.0

0N0.0

000.0

32% 02

mZ<ZHmZOMHHZHDEH

memscflpcoov m magma

16 58

 

 

 

 

 

 

12
If .
A
8—
2
CL
0.
Q
4...
fl
0_ a
I
Q
4

 

I F l l 7
00 0.2 0.4 0.6 0.8 1.0
MOLE FRACTION SOLVENT

FlgUTG 11. Variation of the Chemical Shift of the Sodium-23 Resonance
as a Function of Solvent Composition for Binary Solvent
Mixtures of Nitramethane with Acetonitrile (1;),
Tetramethylurea ((D), Dimethylsulfbxide ([3), Pyridine
([3) and HexamethylphosPhoramide (())

 

59

 

 

 

 

 

 

l I I I I

0.0 0. 2 0.4 0.6 0.8 1.0

Figure 12.

MOLE FRACTION SOLVENT

Variation of the Chemical Shift of the Sodium-23 Resonance
as a Function of Solvent Composition fer Binary Solvent
Mixtures of Acetonitrile with Tetramethylurea ([3),
Dimethylsulfoxide (0), Pyridine (A) and
Hexamethylphosphoramide (())

Figure 13.

60

 

Q
it
0
'I
o ‘
"" I
9 -e
O U A
ll
-- 0
'9 A
\ a ll

 

I I I I I
0.0 0.2 0.4 0.6 0.8 1.0

MOLE FRACTION SOLVENT

Variation of the Chemical Shift of the Sodium-23
Resonance as a Function of Solvent Composition for
Binary Solvent Mixtures of Hexamethylphosphoramide with

Tetramethylurea (¢)), Dimethylsulfoxide (L1)
and Pyridine (1;).

61

APPM

 

1.5 9—

 

 

2-0“ I 'l I I I

0.0 0.2 0.4 0.6 0.8 1.0
MOLE FRACTION SOLVENT

Figure 14. Variation of the Chemical Shift of the Sodium-23
Resonance as a Function of Solvent Composition for
Binary Solvent Mixtures of Tetramethylurea with
Dimethylsulfoxide (£1) and Pyridine ((D)

62

We note a smooth transition of the 23Na resonance as a function
of solvent composition as we proceed from one pure solvent to the
other. The fact that in these plots we note curves and not straight
lines indicates preferential solvation by one of the solvents. The
isosolvation points for these solvent systems are summarized in
Table 6.

It has been shown that nitromethane is a solvent of weak donicity
as reflected by its donor number of 2.7. As a result, it has been
employed as an "inert" medium to study the complexation of Na+ and
Li+ with biologically active molecules (23,101). The data in Tables
5 and 6 and Figure 11 seem to justify using nitromethane as an
"inert" solvent. Binary solvent mixtures of nitromethane with
dimethylsulfoxide, pyridine, tetramethylurea, hexamethylphosphoramide
and acetonitrile exhibit isosolvation points of 0.05, 0.12, 0.06,
0.05 and 0.15 mole fraction of the latter respectively (Figure 11).
These data imply that Na+ is, to a large degree, preferentially
solvated by these solvents relative to nitromethane. Hence, compared
to nitromethane, the relative order of donor ability is
HMPA 3 DMSO 3 TMU > Py > ACN >>> CHSNO2 which is to be expected as
shown by the studies of donicity in Chapter III. The only exception
seems to be pyridine. With a donor number of 33.0, pyridine should
be a stronger donor than DMSO (DN = 29.8) and TMU (DN = 28.9) but
weaker than HMPA (DN = 38.0). Finally, it is interesting to note
that HMPA, DMSO and TMU seem to exhibit the same relative donor
abilities in nitromethane although their respective donor numbers are

widely differing. Hence the donicity of these solvents in

nitromethane is "leveled".

 

63

Table 6. Isosolvation Points for Solvation of the Sodium Ion in

Binary Solvent Mixtures

Binary Solvent System

DMSO:Nitromethane 0.05
PyridinezNitromethane 0.12
TMUzNitromethane 0.06
HMPAzNitromethane 0.05
Acetonitrile:Nitromethane 0.15
TMU:HMPA 0.23
DMSO:HMPA 0.15
PyridinezHMPA 0.10
Pyridine:Acetonitrile 0.29
DMSO:Acetonitrile 0.10
TMU:Acetonitrile 0.11
HMPAzAcetonitrile 0.06
Pyridine:TMU 0.16
DMSO:TMU 0.39

DMSO:Pyridine 0.10

MP

MF

MP

MP

MP

MF

MP

MP

ME

ME

Isosolvation Point

DMSO
Pyridine

TMU

HMPA
Acetonitrile
HMPA

HMPA

HMPA
Pyridine

DMSO

TMU

HMPA

MP

MP

MF

TMU

DMSO

DMSO

< h'mV-V‘ '1-_:w

I._.-

64

Acetonitrile is a solvent of medium donicity (DN = 15.0) and as
such should be more competitive with the preceeding solvent series
for Na+ ion than was nitromethane. Binary solvent mixtures of
acetonitrile with pyridine, dimethylsulfoxide, tetramethylurea,
hexamethylphosphoramide and nitromethane exhibit isosolvation points
of 0.29, 0.10, 0.11, 0.06 and 0.85 mole fraction of the latter
respectively, as shown in Figure 12. Again, the strong donor solvents
HMPA, DMSO and TMU are "leveled” but we now seen an even greater
differentiation of pyridine from these solvents. In acetonitrile, the
relative order of donor ability is HMPA 3 DMSO 3 TMU > Py > ACN >>
CH3N02.

The "repression” of the donicity of pyridine in binary mixed
solvents was surprising. The data seem to indicate that DMSO and
TMU are better solvating agents for the sodium ion than pyridine,
which is contrary to the 23Na NMR chemical shift data and to the
solvents' donor numbers.

To investigate further this unexpected donor repression of
pyridine relative to other high donor solvents, solvent mixtures of
pyridine with hexamethylphosphoramide, tetramethylurea and
dimethylsulfoxide were examined. The 23Na nmr data revealed
isosolvation points of 0.10, 0.16 and 0.10 mole fraction of the
latter indicative of strong preferential solvation by that solvent.
Recalling the donor numbers of 38.0, 29.8, 28.9, and 33.0 for HMPA,

DMSO, TMU, and pyridine respectively, we would expect small

preferential solvation of the Na+ ion by HMPA with respect to pyridine,

but very little or no preferential solvation of the Na+ ion by DMSO

 

L

65

or TMU relative to pyridine since these three solvents have similar
donor numbers. Instead, strong preferential solvation of the Na+
ion is noted for DMSO and TMU versus pyridine.

It was decided to examine the mixed solvent system DMSO-pyridine
in greater detail using vibrational spectroscopy to gain further
insight into either the enhanced donicity of DMSO or the repressed
donicity of pyridine in their mixtures.

It has been shown in previous publications that the vibration of
the sodium ion in the DMSO solvation shell is observed at 200 cm-1
(53), while in the pyridine solvation shell the corresponding
frequency is 180 cm-1 (58). The position of the solvation band in
DMSO-pyridine mixtures was determined and the results are summarized
in Table 7 and are shown in Figure 15. It is seen that even the
addition of small amounts of DMSO to pyridine results in a nearly
complete shift of the solvation band frequency to that corresponding
to pure DMSO. In fact, the rate of change of the band frequency with
composition strongly parallels the corresponding change in the 23Na
chemical shift. These results seem to confirm strong preferential
solvation by DMSO contrary to "donicity" values.

A possible explanation of this apparent anomaly may be in the
different structures of the two solvents. As evidenced by its Trouton
constant of 29.6 cal deg-1 mole-l, DMSO is a highly associated
liquid with a chainlike structure consisting of aligned S-O dipoles
(102). Even in dilute benzene (102) and carbon tetrachloride
solutions (103) the molecules are said to be associated into a cyclic

dimer. It seems reasonable, therefore, to assume that when a sodium

salt is introduced into neat DMSO, a considerable amount of energy

Table 7.

66

Variation in the Frequency of the Sodium Ion Solvation

Band for 0.50 M_NaBPh

Mixtur

MOLE FRACTION

O

0

0.

85

DMSO

.000

.050

099

.151

.198

.300

.400

.501

.598

.703

.800

.946

.000

4

Solutions in DMSO-Pyridine Solvent

FREQUENCY
(cm-1)

178 :.2
184
188
191
195
198
200
201
199
199
201
200

201

 

I'm
x—an- .

 

II c..-

67

I. monsuxfiz ocflvwnx0IonfixomHamfixnuosfio a“
e200mz : 00.0 mo teem nowum>aom gem samvom one mo 50co30090 059 we cowumwnm> .mH onswfim

OmED 20.....04m—n. MACS—
O.— 0.0 0.0 5.0 ed 0.0 .vd md Nd .5 Cd
_ _ . PI PI — _ — — — _ — myFP

meowusaow

I 00—

I
In
0
._

r
O
92

(kwm ACNE-10038:!

I we
* v I com

I mow

 

“—
—-°_

 

68

must be expended to break up the structure of the solvent before
solvation of the cation can occur. 0n the other hand, pyridine is a
relatively unstructured polar liquid with a dipole moment of 2.2 D.
The introduction of even small amounts of pyridine into neat DMSO may
result in the break up of the polymeric structure of the latter
solvent via a dipole interaction.

To investigate this possibility, the S-O stretching frequency of
DMSO was monitored in neat DMSO and in DMSO-pyridine mixtures. The
variation of the S-O stretching frequency with solvent composition
would be indicative of DMSO-pyridine interactions causing structural
changes in the medium. The data are presented in Table 8 and
Figure 16. It is seen that the addition of small amounts of pyridine
to DMSO results in a sharp increase in the vS-O frequency to N 0.05
mole fraction pyridine (0.95 mole fraction DMSO). In the region

0.10-0.80 mole fraction pyridine, the v frequency gradually

S-O
increases. Above 0.80 mole fraction pyridine, the increase in the
VS-O frequency becomes steep again. These data then do reveal a
strong solvent-solvent interaction causing extensive disruption of
DMSO structure by pyridine. A recent study of the Brillouin
scattering by pyridine-DMSO mixtures confirms the conclusions reached
by examining nmr and vibrational data (112).

Although no specific vibrational study of the TMU-pyridine
system was undertaken here, it should be noted that the enhanced
donicity of TMU relative to pyridine may also be rationalized in

terms of structure breaking by pyridine. As Luttringhaus (104)

points out, TMU is structured by strong dipole association as is

69

Table 8. Variation of the S-O Stretching Frequency of DMSO for

DMSO-Pyridine Solvent Mixtures

MOLE FRACTION FREQUENCY
DMSO ' (cm-1)
0.050 1068 1 1
0.100 1064
0.150 1063
0.198 1063
0.300 1062
0.400 1061
0.500 1061
0.598 1059
0.700 1057
0.800 1057
0.946 . 1056

1.000 1049

 

70

moHSHxfiz onwvflax0Iowfixowfismchuosmo new xocosvoum mcwgouonum 0-0 0:9 mo :ofiumfiua> .oH 093000

Oman—0 20:049....— MAO—2
0.5 0.0 0.0 5pc 9.0 0.0 Jo 0.0 N._0 5 0.0
_ _. _ _ _

 

 

 

32
I So.

14‘

H

1..

I 89 m

a

N

3

I 82 M

3

w.

I mac. 1
r 22

71

DMSO. Beguin and Gunthard (105) showed the C=0 stretching frequency

to be strongly solvent dependent as is the S-O stretching frequency

in DMSO. Hence, a strong TMU-pyridine dipole-dipole interaction may
account for the enhanced donicity of TMU by causing extensive disruption
of TMU structure by pyridine.

Finally, we wished to differentiate the three solvents that were
effectively leveled in nitromethane and acetonitrile.« For binary .
solvent mixtures of hexamethylphosphoramide with tetramethylurea and
dimethylsulfoxide, the isosolvation points were 0.23 and 0.15 mole
fraction HMPA as shown in Figure 13. An isosolvation point of 0.39
mole fraction DMSO was noted for the dimethylsulfoxide-tetramethylurea
system shown in Figure 14. These data imply that the relative order j
of solvating ability is HMPA > DMSO 3 TMU where DMSO is just slightly
stronger than TMU.

To this point preferential solvation has been discussed in a
qualitative sense as reflected by the isosolvation point. Covington,

33, El: (106,107), points out that if we make a few reasonable

physical assumptions, we may develop a quantitative model for
competitive solvation. In the above papers, he presents a thermodynamic
derivation of preferential solvation that allows the calculation of
equilibrium constants and the changes in free energy as the solvation
shell of an ion X is progressively changed from n molecules of solvent
W to n molecules of solvent P. The remainder of this chapter will

be devoted to a discussion of the physical assumptions underlying
Covington's treatment followed by an application to the data presented

in this chapter.

 

72

Let us assume that initially ion X is solvated by four molecules
of the solvent W. As the second solvent P is introduced into the
system, there is a step—wise replacement of W by P in the inner

solvation shell. The series of equations can be expressed as,

K K

1 2
XW4 + pP + (w-4)W-——-—+ XW3P + (p-1)P + (w-3)W'-——+
K3 K4
-——+ ———+
XWZP2 + (p-2)P + (w-2)W'+-- XWP3 + (p-3)P + (w-1)W'+-————
XP4 + (p-4)P + wW , (17)

Considering the first step as an example, we can write

o__ =0 _o
AG1 — RTJLnK1 “stp “XW4

+ u; - u; (18)

At this point, we assume that the chemical potential, u°, is made up

of a number of contributions, for example

0 0 int 0 hyd 0 int 0 elec ochem
p = u + u = u + u + u p (19)
XP4 X XP4 X XP4 XP4 .
Oint o o I o o hyd o
where “X is the contribution from the bare ion, ”XP is the
4

contribution from the solvation shell and bulk solvent, which is then

°elec
XP

4
. . ochem . . .
contribution,uXP , dependent only on the compOSition of the solvation
4

split up into an electrostatic contribution, u and a chemical

shell.

If we insert equation (19) into equation (18), we obtain

0 _ 0 Chem 0 Chem 0 o
AG1 ‘ "stp ' uxw4 + “w + “P (20)

73

o elec

where the 0 terms cancel if we assume that the radii of the

two solvated ions, XW and XW P are equal in the Born sense. This

4 3

assumption requires that the solvent system be isodielectric.

However, the treatment may be extended to non-isodielectric systems

if the electronic term can be calculated from the Born treatment.
Another assumption concerns the nature of the intrinsic chemical

shift of each species. If 6? is the shift in the resonance of X

from pure W to pure P, then Covington assumes that wa = 0;
4

XW3P = 1/4 6P; 6XW2P2 = 1/2 6P; 6 = 3/4 6?; oxp4 = 0?.

the intrinsic shifts of the various solvated species are assumed to

6 Hence,

XWP3

be proportional to the amount of P which they contain. Now, there
seems to be no physical evidence for this assumption. Recent 13C
NMR studies for a series of halomethanes (CH4_nXn where X = Cl,
Br, I) reveal that as the protons are replaced by chlorine, the
chemical shift varies proportionally, however, this is not the case
when protons are replaced by bromine or iodine (110). Lauterbur
(109) observed that in some cases additive shift changes are produced
by interchanging ha10gens in these polyhalomethanes. For example,
successive replacement of chlorine atoms by bromine atoms in carbon
tetrachloride causes linear variations in the 13C chemical shifts.
The intrinsic shift assumption implies that each P molecule
will interact equally and independently with X. That is, once a P
molecule replaces a W, its interaction with X will be unchanged as
the next P molecule solvates X, etc. It should be pointed out that

the intrinsic shift assumption must be made so that the number of

unknowns in the final equation is reduced to one. Finally, it is

74

interesting to note that it is this assumption that underlies the
isosolvation point, hence, the discussion contained herein also
applies to this concept.

The next assumption requires that the individual equilibrium

constants, K1, are related solely by statistical requirements, that is,

1/4 /4.

K' = K (21)

1
‘ (K1K2K3K4)

K = 4K', K2 = 3/2 K', K

1 = 2/3 K', K

3 4 = 1/4 K' (22)

This implies that the equilibrium constants are solvent independent.
If there indeed is some degree of preferential solvation, then these
equilibrium constants should be solvent dependent.
The final equation in this treatment allows calculation of Klln
as fellows,
%=%“(1*—‘1'T
P Kl/n.(_P_

aw)

) (23)

By plotting 1/6 vs Xw/XP, we obtain 1/6P from the intercept and Kl/n

from the slope. Hence, we must assume that solutions are ideal. Then,
we may accordingly substitute mole fractions for activities to facilitate
the calculations. However, our solutions are not ideal at all. The
solvation of the sodium ion, which is essentially an ion-dipole
interaction, contributes to non-ideality. To expect no solvent-solvent
interaction is unrealistic, as we have shown marked solvent interaction

in dimethylsulfoxide-pyridine mixtures earlier in this chapter.

75

Finally, Covington concludes that since the plot of equation (23)

o elec

is linear as predicted, the assumptions that u terms cancel in

equation (20) and that assuming the individual equilibrium conStants
are solvent independent are acceptable.

We treated our data according to equation (23) and calculated the

l/n

geometric equilibrium constant, K , and the corresponding free

0

P.S.
Table 9. It is interesting to note that a plot of 1/6 vs XB/XA did

energy of preferential solvation, AG /n, which are shown in
in fact yield straight lines in 2220,9353» except for the solvent
system acetonitrile-pyridine, which displayed a small degree of
curvature. The lines were treated by a linear least squares procedure
from which we obtained the values of Kl/n. It is particularly
interesting to note that such a plot for the data in DMSO-pyridine
mixtures, where there exists marked solvent-solvent interactions,
yields a straight line as shown in Figure 17.

It must be recalled that this treatment assumes that the solvation
number, n, is the same in both solvents in the given binary mixture.

We noted a small amount of curvature in the plot for acetonitrile-
pyridine, which may be a result of n not being the same on this
case. Langford (19,108) noted curvature in similar plots, which he
ascribed to non-equal solvation numbers.

If the values for Kl/n are meaningful (they are certainly as
valid as the isosolvation point model in that both embody the same
assumptions), we may again comment on relative donor strengths. We
may note the same trends in nitromethane and acetonitrile as earlier
mentioned, except that in the former, TMU appears to be much more

competitive than either DMSO or HMPA. Another interesting observation

76

1.0 ‘—

0.9 -

0.8 -

\ 0.7-1

0.6 -

0.5 —

 

 

0-4 I I I I I
0 .- 2 4 6 8 10

XPYRIDINE/XDMSO

Figure 17. Covington Plot for the Binary Solvent System
Dimethylsulfoxide-Pyridine

77

Table 9. Covington Treatment of Preferential Solvation

Binary Solvent System Kl/n
DMSO:Nitromethane 14.4
PyridinezNitromethane 7.44
TMU:Nitromethane 33.8
HMPAzNitromethane 13.2
Acetonitrile:Nitromethane 4.48
TMU:HMPA 0.0669
DMSO:HMPA 0.166
PyridinezHMPA 0.1245
PyridinezAcetonitrile 2.93
DMSO:Acetonitrile 12.7
TMU:Acetonitrile 7.01
HMPAzAcetonitrile 62.9
PyridinezTMU 0.297
DMSO:TMU 0.92

DMSO:Pyridine 17.2

o
"AGP.s./"

(kJ mole-1)
6.61
4.97
8.72
6.39
3.72

-6.70
-4.45
-5.16
2.66
6.30
4.82
10.3
-3.01
-0.206

7.04

78

is that Kl/

n for TMU-DMSO mixtures seems to favor TMU as the stronger
donor whereas the location of the isosolvation point seems to favor
DMSO. Finally, it is interesting to note that the chemical shift

vs mole fraction plot for this system is almost linear, which

indicates little preferential solvation, and that Covington's

065 S /n is close to zero, indicative of no preferential solvation.

CONCLUSIONS

 

The location of the isosolvation point for a given binary solvent
system gives a qualitative measure of the solvation abilities of the
solvents in the given mixture. In order to extend the comparison to
variations in isosolvation values to a series of binary mixtures, more
knowledge is needed on the structures of the individual solvents and
the changes in this structure on formation of liquid mixtures. These
results emphasize the danger of extrapolating the behavior and
properties of pure solvents to their properties in solvent mixtures.

Covington's quantitative approach appears successful despite the
number of physical assumptions, some of which are not reasonable
(intrinsic shifts, solution ideality, etc.). This observation has
two implications. First, as Covington states, "The treatment . . .
although not completely rigorous, assists, we believe in understanding
the present approach to preferential solvation . . .". 0r second,
this treatment is not sensitive enough; that is, the physical
assumptions are not as crucial as they appear. Since the NMR
technique reflects the influence of the first solvation sphere on the
ion of interest, perhaps the non-ideality of the bulk solvent mixture

outside this sphere has small influence on the ion. Moreover, perhaps

79

the ion-dipole solvation phenomenon does not cause significant

deviations from ideal behavior.

CHAPTER V

A SPECTROSCOPIC STUDY OF CONCENTRATED SOLUTIONS OF
SODIUM PERCHLORATE - THE NATURE OF THE

UPFIELD CHEMICAL SHIFT

INTRODUCTION

 

Generally we noted that the 23Na resonance in solutions of sodium
perchlorate and tetraphenylborate exhibited no concentration dependence
over the concentration range 0.01-0.50 M5 whereas, the corresponding '
shifts for the thiocyanate, bromide and iodide salts exhibited marked
concentration dependence, shifting downfield with increasing concentration.
In the latter case, these downfield shifts were attributed to contact
ion pairing, while in the former case, the static shift was proposed
to be indicative of free ions or solvent separated ion pairs. These
assignments seem reasonable as it would be easier for the smaller
bromide, iodide and thiocyanate anions to be part of a solvation
sphere as opposed to the larger, bulky perchlorate and tetraphenylborate
anions.

It was pointed out before that small upfield chemical shifts of
the 23Na resonance were observed for sodium tetraphenylborate and
perchlorate solutions in methanol and ethanol from 0.01-0.50 M,
Reexamining the data and figures in Chapter III, subtle upfield shifts
of the 23Na resonance are noted for sodium perchlorate solutions in
tetramethylurea, acetone, tetrahydrofuran and pyridine to name a
few. The problems that confront us are: (1) What type of species
is responsible for the upfield shift? (2) Are the occurence and
magnitude of the upfield shift related to some solvent property
or is it characteristic of the electrolyte? (3) How do these upfield
shifts vary with concentration over a wider range (0.01 Mfsaturation)?
To answer these questions, detailed 23Na, 1H and 3501 NMR, Raman and
infrared measurements were made of sodium perchlorate solutions

in several nonaqueous media and are the subject of this chapter.

80

81

RESULTS AND DISCUSSION

 

The chemical shifts of the 23Na resonance for sodium perchlorate
solutions in several solvents are presented in Table 10 and diSplayed
in Figure 18. With two exceptions, the range of the upfield chemical
shifts in a given solvent is large (:_1 ppm), indicative of significant
changes in the environment of the sodium ion. In formic acid,
water, methanol and ethanol the resonance shifts upfield linearly
with concentration, whereas, non-linear shifts are noted in the other
solvents.

Upfield shifts of alkali metal ion resonance with increasing
concentration have been noted for aqueous alkali metal nitrate
solutions by Deverell and Richards (16), for aqueous solutions of
sodium perchlorate by Van Geet (35), and for aqueous solutions of
potassium dichromate, nitrate, chromate and sulfate by Bloor and
Kidd (27). Since the chemical shift results from overlap of the
outer electron shells of the ions, these authors conclude that when
water in the first solvation sphere is replaced by some oxyanion, the
oxyanion overlaps less effectively with the cation orbitals than
water, resulting in decreased electron density about the cation
nucleus and hence, an upfield shift in the cation resonance.

An alternative explanation is that collisional ion pairs, if
formed, remain solvent separated, and the alkali metal ion is
directly surrounded only by water molecules. The shift then
reflects a change in the "bond" between the cation and water as the
salt concentration increases. This long range shift would be

analogous to proton chemical shifts in organic compounds. Since

82

00.0
0v.0
00.0
0H.0
0H.0
00.0
00.0
05.0
05.0
00.0
0v.0
00.0

flamov owfirm

00.N
00.H
00.H
0v.~
0N.~
00.H
00.0
00.0
0v.0
0N.0
0H.0
00.0
A20 0:00
mZOHmu<

wave: msoosvmcoz :« meowusfiom oumHoHcoho0 550000 How mumflsm Hmofiaecu «2

00.0
0v.0
0N.0
00.0
00.5
00.5
0v.5
0N.5

00.0

fleoov umfigm

mqmthZOHmu<

00.N
m5.H
00.H
0N.H
00.H
05.0
00.0
0N.0

0H.0

0:0 ocou

mN.m
o~.m
o~.m
oo.m
OH.m
oo.m
mw.w
ma.w
oo.m
mm.w
05000 om

firm

00.”
00.0
00.0
05.0
00.0
00.0
00.0
00.0
0N.0
0H.0

fizv ocou

z<030000>m<MHmh

mm

.oH.oHnme

83

00.0 00.H 0N.0 00.N

00.0 00.0 00.0 00.N 0.00 00.0
00.0 00.0 00.0 00.N 0.00 00.N
00.0 0N.0 00.0 00.0 00.0 .00.~
00.0 00.0 00.0 00.0 00.0 00.0
00.0 00.0 00.0 00.0 00.0 00.0
00.0 00.0 00.0 00.0 00.0 00.0
00.0 00.0 00.0 00.0 00.0 00.0
00.0 00.0 00.0 00.0 00.0 00.0
00.0 00.0 00.0 00.0 00.0 00.0
00.0 00.0 00.0 00.0 00.0 00.0

00000 00000 020 0000 00000 00000 020 0:00 00000 00000 020 0000

00Hz<200000200200 002<0002 000< 002000

00000000000 00 ofiame

84

o~.m
ow.~
om.N
om.m
om.H
mo.~
mm.~
mo.~
ow.o
om.o

09.0

AE000 00000

mmk<z

oo.m
om.v
oo.v
om.m
oo.m
om.m
00.N
om.H
oo.H
om.o

mm.o

020 0000

o~.m
ow.m
mo.~
mm.m
mv.~
mm.m
mH.N
om.~

on.~

fle000 000:0

AOZ<ZHm

ow.o
05.0
oc.o
om.o
ov.o
om.o
om.o
OH.o

mo.o

mzv 0:00

om.m
ov.m
oo.m
ov.m
ov.m
om.m
mm.m
om.m
om.m

om.m

00000 00000
<05: 050035...

00.0
00.0
0000
00.0
00.0
00.0
00.0
00.0
00.0

o~.o

~20 0:00

A0000000000 00 00000

85

00.0 00.0
00.0 00.0
00.0 00.0
00.0 00.0
00.0 00.0
00.0 00.0
00.0 00.0
00.0 00.0
00.0 00.0
00.0 00.0

00.0

00000 00000 020 0:00

000x0000000200200

m.oH

m.o~

H.oH

H.0H

H.oH

mn.m
om.m
mw.m
mn.m
oo.m

oo.m

00000 00000

oo.~
om.o
om.o
05.0
oo.o
om.o
ov.o
om.o
o~.o
o~.o

mo.o

020 0:00

mh<zomm<u mzmqrmomm

00000000000 00 00000

6')
I

APPM
01
l

 

 

 

 

Figure 18.

I l I l
1 2 3 4
CONCENTRATION (M)

Sodium-23 Chemical Shifts of Sodium Perchlorate Solutions
in Nonaqueous Media: (a) Propylene Carbonate, (b) Formic
Acid, (c) Tetrahydrofuran, (d) Acetone, (e) Acetonitrile,
(f) Methanol, (g) Dimethylformamide, (h) Tetramethylurea,
(i) Ethanol, (j) Dimethylsulfoxide and (k) Water

87

such an effect would have to be transmitted through at least three
bonds, it should be fairly small. Hence, we may tentatively
attribute the upfield shift to a direct cation-anion interaction of
some kind.

If cation-anion interactions (contact ion pairs) are responsible
for the shift, the linearity of the upfield chemical shift with salt
concentration in the four solvents noted is surprising when compared
to the corresponding downfield shifts of the 23Na resonance for the
thiocyanate, bromide and iodide salts in the same solvents. It is
interesting to note that despite a high dielectric constant of 78.6,
the data suggest that a rather significant degree of cation-anion
interactions occur in water. The solubility of NaClO4 in water
exceeds the mole fraction of 0.25; hence, for every two ions in
solution, less than three molecules of water are available (35).
Therefore, a least some of the Na+ must be in direct contact with
ClO4-. Alternatively, we may suppose that such contact ion pairs
only form at high concentrations. However, if this is so, the shift
would not vary linearly with concentration. It is concluded then that
at least some contact ion pairs are present at all concentrations.

In order to obtain an overview of the associative behavior of
NaClO4, a literature search of conductance studies of this salt in
nonaqueous media and water was initiated. The results of this search
are presented in Table 11 where the solvent, the ion pair association
constant (Ka) and the reference are listed. The data indicate that
even in water, the C104' ion is associated with the Na+ ion into
either a contact or a solvent separated ion pair. The lack of

association noted in DMSO, TMU and DMF may be attributed to a weakness

88

Table 11. Association Constants of Sodium Perchlorate From Conductance

Studies
SOLVENT Ka REF
Water 0.20 :_0.14* 113
Methanol 18.70 i 0.04* 114
Acetonitrile 20.77 i 0.50* 115
36.0+ 116
10 :_1+ 117
70.92+ 121
Dimethylsulfoxide 0+ 118,122
Sulfolane 9.05 :_0.04* 115
Tetramethylurea 0+ 120
Dimethylformamide o+ 119,122

+Eva1uated using the Fuoss equation (123)

*Evaluated using the Fuoss, Hsia, D'Aprano equation (113,124)

89

in the linearized form of the Fuoss conductance equation for associated

electrolytes,
1/2 2
A = A0 - S(Cy) + ECleg Cy + JCY - KaCyf A (24)

and unassociated electrolytes,

A = A0 - 5(0)”2 + EClog c + JC (25)

where the symbols have their usual meanings (123). These equations
were obtained by retaining only terms linear in concentration in the
integration of the differential equation which describes the
relaxation field. Recently, Fuoss and Hsia (124) reexamined the
problem retaining concentration terms of order C3/2 and proposed a

new equation for associated electrolytes,

3,273/2

A = A0 - SCCy)1/2 + ECylog Cy + my + J20 - KaCysz (26)

D'Aprano (113) has shown that previous conductance data evaluated

by the Fuoss equation for unassociated species, when reevaluated by
the Fuoss, Hsia, and D'Aprano equation predicted associative behavior.
The important idea is that the association of alkali perchlorates

in water should ngt_be unexpected.

While the analysis of precise conductance data is a well
established criterion for identifying ionic association in electrolytes,
it seems that laser Raman spectroscopy may be applied to explore
the interfacing of criteria advanced independently from Spectroscopy
and electrical conductance relative to ionic association in solutions

of electrolytes. The C104- ion has its nine vibrational degrees of

90

freedom distributed into four normal modes of vibration: v1(A),

the totally symmetric stretch; 02(E), the symmetric bend; v3(F2), the
asymmetric stretch; and v4(F2), the asymmetric bend. By studying

the infrared spectra of some hydrated transition metal perchlorates,
Hathaway and Underhill (134) noted splittings of these bands as the
perchlorate ion changes from free ion (Td) to monodentate perchlorate
(C3v) or bidentate perchlorate (sz) as shown in Figure 19 which
lists the assignment of each mode, the motion of the ion corresponding
to that assignment, the approximate wavenumber of that mode and its
infrared (IR) and/or Raman (R) activity. It should be noted that
these assignments are based on a ”partial covalent" interaction of
C10 - ion with the metal cation. Contact ion pairing is a purely

4

electrostatic interaction and as such, may not be strong enough to
perturb significantly the tetrahedral symmetry of the C104- ion as
does the covalent interaction. Therefore, when observing the Raman
spectra of NaClO4 solutions, a lack of splittings or intensity
differences does not obviate cation-anion interactions; rather, it
implies that either the interaction is too weak to alter the Td
symmetry to C3V or C2V or that the number of such interactions is
too small to contribute to the spectrum. 0n the other hand, the
appearance of new bands or distortion of existing bands may be taken
as definite evidence for ion pairing phenomena.

The laser Raman spectrum of NaClO4 in water, tetramethylguanidine,
pyridine, methanol, tetrahydrofuran, acetonitrile, ethanol,

dimethylsulfoxide, formic acid, acetic acid, dimethylformamide,

propylene carbonate and tetramethylurea was monitored from

91

000
«:00000
00.00000
m?
00.0000
m?
000

O00
meuxoou
00.00000

up

owe
neon Exam

00.00000

0vo
vamp 5x0
00.0000<

neon 500
00.000<

chm

AHpoEEXm wo cowuocsm m 00 @5000 uVOHU 0:0 mo macaumhn0> .mH ohsmfim

0000 0000 0000 000 000 >00
econ same 000 350a 000 5x0 :000000 0:00 500 up» 500 . \\\p//r
00.00000 00.00000 00.0000< 000~< 00.0000< 00.0000< -0000111 \\\2

w> o: 0: n: v: N: o

0000 000 000 0
000 3000 econ 350 ppm 500 b
00.00000 0000 0000 -0000

m? N? #9
>0

0:00 5000 000 Ex» «:0xoou pm 0

00.0000 00.000< 00.0000 00.0000< 0000-0-2
V? H? 09 N?

Nana awed omv own

92

100 cm-1 to 1500 cm.1 from the exciting line. Generally, only the
01 (935 cm-l), 02 (460 cm-1) and v4 (626 cm-1) bands were observed,
as the 03 (1110 cm'l) band is very weak in the Raman and was
observed only at high concentrations (e.g., :_l fl_in water). These
three bands appeared in all solvents studied. The only new band
readily observed was m 470 cm”1 in acetonitrile, tetrahydrofuran and
pyridine as shown in Figure 20 for 1.0 M_NaClO solutions for

4
the first two solvents.
The splittings observed in THF and ACN in the 400-500 cm-1
region were further examined as a function of concentration and the
data treated with a curve fitting computer routine. The equation

employed to resolve the overlapping bands shown in Figure 20, is a

Lorentzian-Gaussian product described by Irish, gt: El: (125):
I = Io{exp[(-'J-Uo)2/2o21}{1 + (3-3012/021'1

where I is the arbitrary intensity at frequency 3; 3; is the position
of the line center with maximum intensity, Io; and o is the variance,
which gives the halfwidth when multiplied by the factor 1.46. This
equation was employed to fit the data by using either a non-linear
weighted least squares analysis, KINFIT (98) or a sequential simplex
method, PEAKSBF (126). The results of this data treatment are

listed in Table 12 and a sample of the computer output is shown

in Figures 21 and 22. Presently, PEAKSBF is limited to resolving
only two overlapping bands, whereas, KINFIT may resolve two, three

or four overlapping bands.

93

TETRAHYDROFURAN

 

ACETONITRILE

INTENSITY

 

 

[ I 1 I i
500 475 450 425 400
A FREQUENCY (CM'I)

Figure 20. Raman Spectra of 1.0 H_Sodium Perchlorate Solutions in
Tetrahydrofuran and Acetonitrile from 400-500 cm'1

94

H

-000 0000.000 0.000. 00.000000000000000. :0 000.002.0000 050000 m 0.0 00 000.0002 0.00200. .00 0.000000

30¢ I ccc 7¢0=ucac>2¢ahwb 2— whdaCJICnuo !:hccv ac.- eeocoooooooetootoooooooo 07w

Fm----m---'m----m--"W'---W"--V'--'m'---V'-'-v----U---'m--'-w-lv'-r'-'-U---IU--'-VII'l U U I I-"--'-I-"-—

mu nocc . .muuuuw
xuuooc . ccsnccc guarcc *
uxncc . . Ctr x ywxuuu ccc
xxcc urn x "vary :0 u
u xx! . C" .C( x Ctr ~
m. "x . x run (0:20; 0
— Ox . u u >0 f
0 n w
— Cx
0 0.. xu _
V x u
— Cx C u
— x 0
— Cx C _
0 x x .0
U C u
_ c 0
0 x x 0
0 c 0
0 0 0
V C X} H
0 x 0
0 0
0 c C
0 x
0 .. . 0
0 c 0
0 c 0
0 x . 0
V C x u
0 c 0
p xC —
0 —
p C , xxulflfl —
U x CCuCC n O u
0 . 0
xx x 0
0 x 0 ~
0 m C —
w -
0 w
0 ux c 0
0 x 0
0 0 _
V C u
0 u x 0
0 C u —
— xC . 0
0 - 0 . 0
Nm---'U'-'-w-"'m--'-m--"W----w--'-m'---v----ui---v----m-'--U'--'v-'I'm'-'-T--'-m--"u'|"T"-'U--"h

- > than: >3 x drama wzcv w:» 20 mm: 020cc cahagzuacc c:c 04020:020axu (a u2<wx u
.cuppcaa m0 o0muu 201: cum: w0 u > 2:. 020cc ou»«0:u0<u.« 0:00: 0 .0.0cc .cazuz0auoxu 2c uzcmz 0

00.0000. 0 +2wzmauz0. - .o u zcwpca ~10 04 w:0¢>.noou~nn. u-ac» at» he m24<> .6000: cc ._.~.x m0 0qu00au>
ocoucoo. u pzuxuaoz0.noou¢oe. n rte—a 01» >4 w:0¢>.ncoucce. u 000. u:» 04 u:.<> .0..oc:x.x ac 00.0.x v0 vvcocq

 

 

 

I
~...I. u. u . up u u no .IIIIII . O
. III II’.,.I I b t I D b 0 I p p n ..... IooIIIIIIII
II II. III-90¢ . III’..I00
I I II. 3!. . Io IIIIIIIO.
I III..I.!0.¢. . It...l..lcu
I.....II..II. . . II .o.)!.to.

95

I I It!"

I I .2396.-

I I II. III—Ion

I II II. ’0...

I II II. 3000‘.

.I II 93.33,.

I I I II. ’2’.

I I I II.IIII.III

I I I I2;

I I I I3};-

I I I 0331’...

I I I 3.113.

O I I II.IIIIIII.

' I IT'IIIIII

. I I223...

I I I II.IIII¢II.

I I I II.IIIIIII.

I I I III-Iago.

I O I OOIIIOI‘II
O I I II. IIIIoII.
.. 0.50.1...

II II.§IIII.

I I IIIIIIIII.

I I II..IIIIII.

I I II.III¢I¢I.

I I 2.3.1:.

I I IIIIIIIII.

I I I II.IIII¢II.
I I I II.IIIIIII.
I I I II.IIIIQII.

I I I II.IIIIIII.

I I I IIIIIIIII.

I I . . 0......03.

I . I I 2......3.
I I I 3.335..
I I 2.8.310.

I I II.‘IIIII.

I I 3.80.3...

I I 2.3.3:.

I I 3.333..

I I IIIIISII.

II 3.33:...

II II.IIII;I.

II II.IIIIIII.

I II.IIIIIII.

I II..IIIIII.

I I I II.III..II.
I I . I II.IIIIIOI.

I O I II. .IIIIIII.

II I I333...-

I I. I 3.3-3...

I I I II.IIII~.I.

C I I II.IIII.III I

I I I II.IIIIIII.
I I I 07330:.
I I I ITIIIIII.

I I I II.IIIIIIIo

I I I 3.83.11

8 II II..I,III

I II I130.

I II II..IIIII.

I I IICII‘II.

I I III!:I.

C I IIoiII

 

 

II..IIIoOQo
I I II..I-I.In.

I I. .3083.
I .- I00IIIOI
I I.IIIIIIOQ

II. ‘1‘. .

09800.08
III."
OIaIIIolI‘
IooIIIIIQII
II.IIIIIIII
II.‘I¢IIQI
II.‘I'II!I
3.18.0.
III-III.
II...IIIIO.
II.IIIIIIII
II.II.IIII.
II.IIIIIII.
II.IIIIIII.
0.0.0IIIIII
II.IIIIIII.
II.I.IIIIIO
IILIIIIII.
II. IIIIIII.
QC. I..IIIII
II. 30.1:-
‘0II..I'.OI
II. IIIIIIII
.II3IOFIOI
IIIDIIOIOII
II.:I0 III.
II.IIIIIIII
II. 87....
II.III.IIO.
II..III.IIoo
II.II~IIII0
OI. 'IIIIII
000:0I0flflo
II.III.000.
II. IIIIIII.
I233...-
II.II..III.
II. IIIIIII.
II.III.III.
GI. IIIIIIII
II.IIIIIIII
II. IIIIIII.
00.50000!
II.IIIIIcI.
II. III. II.
II. IIIIIII.
II. 0.0:...
II.II¢.9IC.
II. III. I.
II.I.I.II¢0
II.’I1¢I
II. III-(I I.
II. 0301‘
II. .101-Io
IOIIII’IOI
I III-III o.
I I. III: to

in

ate

1Perchlor

PEAKSBF Analysis of 1.0 M deium
Tetrahydrofuran from 430-498 cm

Figure 22.

tal data

I

experimen

x:

’

calculated'data

O = eXperimental data; X = calculated data

0

ABOVE
BELOW

96

Table 12. Computer Analysis of NaClO Solutions in Tetrahydrofuran

4
and Acetonitrile Using PEAKSBF and KINFIT

Conc (M) Position (cm-1) Intensity 61/2(cm-1) 1470/1456
ACETONITRILE/PEAKSBF

1.00 459 213 31 ‘ 0.40
476 85.6 19

1.50 456 214 24 0.60
472 128 22

2.00 455 142 20 0.80
468 114 29

TETRAHYDROFURAN/PEAKSBF

1.00 453 307 22 0.75
470 231 22

O 80 453 272 20 O 83
470 226 22

0.60 454 157 19 0.74
471 116 19

ACETONITRILE/KINFIT

1.00 457 167 18 0.36
476 59.3 22

1.50 456 217 16 0.59
472 127 15

2.00 455 159 16 0.62

470 98.2 15

Table 12 (Continued)

1.0

0.80

0.60

453

470

453

470

453

470

97

TETRAHYDROFURAN/KINFIT

299
218
267
212
157

106

16

18

14

18

13

16

0.73

0.79

0.68

98

The data in Table 12 suggest that the m 456 tm'1 band is

characteristic of free 0104- ion whereas the W 470 cm-1 band is

assigned to ion paired C104 ion. These assignments are based On

the fact that the ratio of intensities, 1470/1456’ increases with
increasing concentration dramatically in acetonitrile and very subtly
in tetrahydrofuran. Relating this to the 23Na NMR data, in
acetonitrile, the chemical shift varies upfield m 1.5 ppm, indicative
of dramatic changes in the chemical environment of the Na+ ion due to
a cation-anion interaction. Hence, we would expect the Raman bands
characteristic of free perchlorate ion and complexed perchlorate ion Q
to also change dramatically. In acetonitrile, this ratio, 1470/1456’

varies from 0.40 to 0.80, reflecting an increase in the relative

amount of cation—anion interactions. 0n the other hand, in tetrahydro-

furan the chemical shift moves upfield only by W 0.15 ppm over the

concentration range 0.60-1.00 M, These data imply very little change

in the Na+ ion environment; equilibrium between free sodium ion and

complex sodium ion is almost established. The static nature of the

1470/1456 ratio confirms this hypothesis. In addition, Figure 20
displays a new, weak, polarized band at m 412 cm'1 in the tetrahydrofuran
solution. A corresponding band is not observed in the acetonitrile

solution, perhaps due to a strong, broad solvent band at N 380 cm-1.

1 and 935 cm'1 bands appeared symmetrical,

with no new bands present in the region 100-1500 cm-l.

In both solvents, the 626 cm-

Acetonitrile also affords the Opportunity to monitor the
disposition of solvent molecules in solution as well. The C-C
. . . . -1
symmetric stretching mode for acetonitrile occurs at 919 cm .

Janz (80,81,127) has noted a new band at t 928 cm.1 for AgNO3

99

solutions in acetonitrile which he assigns to complexed solvent.

The Raman spectrum of NaClO solutions in acetonitrile as a function

4
of salt concentration is shown in Table 13 and Figure 23.
With increasing concentration, the ratio of complexed to free

acetonitrile, I I919, increases as would be expected. The half-

927/
width (VI/2) of the 934 cm"1 perchlorate band seems to decrease with
decreasing concentration which may be indicative of complexed
perchlorate ion going to free ion.
As mentioned earlier, both the 456 and 470 cm.1 bands appear in
a 0.5 flNaClO4 solution in pyridine. From the 23Na NMR data, it was
concluded that contact ion pairs are present in this solution; these
Raman data lend support to that conclusion. More interesting was i
the marked broadening of the 934 cm-1 perchlorate band in this
solution, which may also be indicative of a cation-anion interaction.
An inspection of Figure 19 suggests that in addition to
splitting upon interaction, the new 460 cm.1 band(s) should also
become infrared active provided the complexation interaction is
strong enough to perturb the Td symmetry of the C104- ion. The
far infrared region from 100 cm.1 to 600 cm.1 was scanned fbr
NaClO4 solutions in ethanol, dimethylsulfoxide, propylene carbonate,
dimethylformamide, pyridine, acetone, acetonitrile, tetrahydrofuran,
tetramethylurea and tetramethylguanidine. No new bands were noted
in any of these solvents except for the sodium ion solvation band
at N 200 cm-1.
Therefore, the vibrational studies suggest that the Na+-C104'

interaction is not capable of producing strong changes in the

C104- ion spectrum as did the covalent metal perchlorate salts

100

Table 13. KINFIT Analysis of Raman Spectrum of Sodium Perchlorate

Solutions in Acetonitrile from 900-950-cm‘1
Conc (M) Position(cm-l) Intensity v (cm—1) I /I
1/2 927 919

920 92.6 8

2.00 927 116 16 1.25
934 209 10
919 147 7

1.50 927 99.2 18 0.67
934 150 7
919 213 7

1.00 927 110 18 0.52

934 153 S

101

H

 

 

 

 

 

 

 

 

 

 

-Eo ommncom Eoum oflwuuwcouoo< cw mcowusHom oumuofigonom ssfipom mo mnuoomm cmemm .mm enamfim
.729 55:85. a

a...» new as see e5 93

_ _ _ - _ _ 01/. . _-
\ M
I.
a j a 3
N
Ah
I.
a .A

a a
flow 2.3 23

102

studied by Hathaway and Underhill. In these electrolyte solutions,

only the 460 cm.1 Raman band is strongly affected by cation-anion

interactions, while some line broadening is noted in the 935 cm-1

Raman band. These changes were quite apparent in NaClO4 solutions

in acetonitrile, tetrahydrofuran and pyridine. However, a reexamination

1

of the 460 cm- Raman band in acetic acid, ethanol, water and fermic 1

acid shows marked asymmetry on the high wavenumber side which may
be indicative of a weak band at m 470 cm-1. However, befbre these
bands are studied in detail, we must be assured that the noted

tailing is not an instrumental artifact.

In addition to these vibrational studies, 35Cl NMR was employed

to monitor the disposition of the C104 ion in solution. Complexation

of the C104- ion should effect the quadrupolar relaxation mechanism

of the 35C1 nucleus and result in a broadening of the resonance
signal; in addition, such complexation may alter the electronic
environment of the 35Cl nucleus and result in chemical shifts.
Tables 14 and 15 diSplay the 35Cl linewidths and chemical shifts for

NaClO4 solutions in water and several nonaqueous solvents. In all

solvents, the 35C1 linewidths seem to be the same within experimental

error (m 30 :_5 Hz). Similar solutions of LiClO4 exhibit marked

1/2 values ranging from
35

30 Hz to 170 Hz in a given solvent (128). The C1 chemical shifts

linewidth changes with concentration, with 9

all seem to be identical within experimental error (N -1041 :_5 ppm)

except for concentrated solutions in water. The relative insensitivity

35

of C1 linewidth and chemical shifts for sodium perchlorate solutions

suggests the cation-anion interaction does not significantly perturb

103

m.“ mNoH- o.H

mQHxOm43m4>EHmZHQ

m.“ emofi- o.H

fleaav emaem flag eceu
z<mzaomoam<aeme

mflwoz macadamcoz :fl mcoflusfiom oumHoHsopom esflvom How muwfigm HmoflEogu Ho

m.” mega- o.H

mon<2m0aa>zam2Ho

m.“ eeofi- o.H
mzoemo<

m.“ meoa- o.H

neaav emagm flzv ecou

qumHHZOHmu<

m.“ emofi- o.~
aHu< qumoa

m.“ Nmofi- o.~
85.5%

m.“ meofi- o.H

m.“ mica- o.m

m.” asofi- o.m

magma peacm fizv ecou

mme<z

.eH magma

mm

104

mm m.o

em o.H

mm m.H
m.“ 5N o.m

mSHmHHzoemu<

mm m.o

mm o.H
m.“ mm m.H

fiazv gueazecag “ES 0:00
maazazxoaq>zemzHo

muco>flom msooscmcoz :fi wcofipsaom oumpofinohom esfiwom pom Lppfl3ocflq do

away nuwflzmcflq

mm m.o
m.“ me, o.H
z<maaoma>m<meme
mm o.H
om o.~
em o.m
mm o.e
m.“ mm o.m
mme<z
Hm o.H
em o.N
m.“ on o.m

QHu< UHZmOm

mzv ocou

em m.o

om o.H

mm m.H

m.“ mm o.~
mzoemu<

em m.o

mm o.H

mm m.H

om mum

m.“ em m.~
SE59

em om.o

mm oo.~

m.“ me om.H

hazy eeeazecea 52v ecou
monoaqzmq>zemzHo
.mH eaeee

mm

105

the 35C1 nucleus. In contrast, marked changes in 35C1 NMR linewidths and

C104' Raman and infrared bands have been observed for LiClO4 solutions in
nonaqueous media (128,129,130). Hence, relative to the Li+-C104

interaction, the Na+-C104- interaction is significantly weaker as
shown by 35C1 NMR, Raman and infrared measurements.

Altogether, the data suggest that in NaClO4 solutions, weak
cation-anion interactions exist; the 23Na NMR data strongly suggest
that this cation-anion interaction is contact ion pairing. The problem
is to rationalize contact ion pair formation in a high donor, high 1
dielectric constant solvent like water even at low salt concentrations 5
(~0.1 fl) and to explain the linear upfield shift in the 23Na resonance
with increasing NaClO4 concentration in water, ethanol, methanol, and i

fbrmic acid. This problem can be resolved by considering an interaction

between the C104' anion and the solvent molecules or a lack of such

interaction.

The chemical shifts of solvent protons are generally altered by
the addition of electrolytes. Among factors known to contribute to such
changes are polarization of solvent molecules by the ions and
modifications in the solvent structure produced by the ions.
The polarization effect usually leads to deshielding of the solvent
[arotons and hence a low field shift, unless magnetic anisotr0py
caffects are present, which may cause a high field shift. The structural
effect results in a low field shift if the solute is a net structure
lulker, and a high field shift if it is a net structure breaker.
Since the C104' ion is known to be a powerful structure breaker, we

“mtlld expect a low field shift if the polarization effect dominates

“K1 a high field shift if the structural effect dominates.

106

While studying the 1H resonance of the hydroxylic solvents
water, methanol and ethanol, Krumgal'z, gt, a}; (131) noted high field
shifts of the OH protons when NaClO4 was the solute. In another study,
they found the protons in acetone and acetonitrile to shift to lower
field with the addition of NaClO4 (132). In the former case, the data
were rationalized in terms of strong structure breaking, whereas the

polarization effect was dominant in the latter case. Coetzee and

Sharpe (130) noted that the C104

ion did not effect the C-H stretching
frequency of CHSCN at all, whereas other anions caused significant

changes. They concluded that C10 ' does not interact with acetonitrile,

4

confirming the 1H NMR data of Krumgal'z.
Hence, by invoking the concept of structure breaking, the upfield

shift problem may be rationalized. The upfield nature of the shift

4

. + .
cannot donate as much electron den51ty to the Na ion as could the

itself is the result of a ctaion-anion complex where the C10 ion

4- complex may

exist in water because the C104‘ will not be hindered by solvent

. . . . +
solvent. Despite the high dielectric constant, a Na -C10

structure about the cation in its quest for the Na+ ion. In this
respect, it is interesting to note that the linear upfield shift only
occurs in the hydroxylic, structured solvents water, methanol, ethanol
and formic acid. These shifts may imply that when contact ion pairs

are formed in these solvents, it is by collision. Also, if the

energy of interaction is less than kT,the probability of their

occurence would vary linearly with concentration. In contrast, the
energy of the classical ion pair formed in acetonitrile, tetrahydrofuran,

etc., would be greater than kT and result in a non-linear dependence

107

on concentration. The fact that only a slight upfield shift in the
23Na resonance is noted in tetramethylurea and dimethylformamide
implies that contact ion pair formation is not favorable in these

high donor solvents.

CONCLUSION

 

Although further investigations are in order, the use of spectro-
scopic techniques to examine the upfield shift problem has at least
displayed that cation-anion interactions are responsible for the
observed shift. Moreover, speculations as to the types of species
present in solution were advanced. This investigation emphasizes the
idea that to fully understand the role of solvent in a chemical
process, a sound knowledge of solute-solute, solute-solvent and solvent-

solvent interactions in that solvent is required.

APPENDICES

APPENDIX I

DESCRIPTION OF COMPUTER PROGRAM KINFIT AND SUBROUTINE EQN
FOR THE CALCULATION OF ION PAIR FORMATION CONSTANTS

BY THE NMR TECHNIQUE

DESCRIPTION OF COMPUTER PROGRAM KINFIT AND SUBROUTINE EQN
FOR THE CALCULATION OF ION PAIR FORMATION CONSTANTS

BY THE NMR TECHNIQUE

The numerical calculations for the ion pair formation constants
were performed on the Control Data 6500 digital computer using the
KINFIT program (98) by manipulating SUBROUTINE EQN. This appendix
lists a typical deck for KINFIT analysis and describes the construction
of this deck.

Recall equation 16 which expresses the observed chemical shift,

6 , in terms of the total salt concentration, CM the chemical

obs T’

shifts characteristic of free and complexed metal ion, 6F and 6C

respectively, and the ion pair formation constant, K.

-1 :_(1 + 4KC¥)1/2
M ] (6F - 6C) + 6C (16)

ZKCT

 

aobs = [

_ M
dobs’ ”(1) - 6C’ T

we obtain the FORTRAN equation for use in SUBROUTINE EQN for each

By letting S U(2) = K, XX(1) = C and CONST(1) = 6

F

data point.

8 = ((-1,0+SQRT(1.0+(4.0*U(2)*XX(1))))/(2o0*U(2)*XX(1)))

*(CONST(l)-U(l))+U(l) . (28)

108

109

Note that we assume a constant value for 5F and that 6C and K are
unknown. In order to fit the calculated shift (the right hand side
of equation 27) to the observed shift, the program may vary the
values of 6C and K. Hence, the number of unknowns, NOUNK, equals
two as does the number of variables, NOVAR.

The first data card contains the number of experimental points
in columns l-5(FIS), the maximum number of iterations allowed in
columns 10-15 (FIS), the number of constants in columns 36-40 (PIS)
and the maximum value of (A parameter/parameter) for convergence to
be assumed (0.0001 works well) in columns 41-50 (F10.6). The second
data card contains any title the user desires. The third data card
gives the value of CONST(1), SF, in columns 1-10 (F10.6) in ppm.

The fourth data card contains the initial guesses for U(l), 6C,
and U(2), K, in columns 1-10 and 11-20 (F10.6), respectively. The
fifth through N data cards contain XX(1), concentration in columns
l-10 (F10.6), the relative variance of XX(1) in columns 11-20
(F10.6), XX(2), the chemical shift at XX(1) in columns 21-30 (F10.6),
the relative variance of XX(2) in columns 31-40 (F10.6) followed by
the same parameters for the next data point. Hence, each card may
contain two data points. The relative variance is error in each
measurement normalized to each other (e.g., if the concentration
error is :_0.04 M_and the shift error is :_0.1 ppm, the relative
variance of the concentration is 0.04/0.04=1 and the relative
variance of the shift is 0.1/0.04:25). If there are an odd number

of data points, the last data card may contain that one point with

columns 41-80 blank. If no further data is to be analyzed, the next

110

card after the last data point(s) should be a 6789 card. If more
data set is to be analyzed, the next card after the last data
point(s) is the number of points, number of iterations . . . etc.
for the next data set.

Generally the most common error in using this program is an
error MODE44in an address of the SQRTE subroutine. Usually, this
implies that the initial guess for K is quite inaccurate. Before
using this program, the user should see reference (98) or the
materials of CEM 883, Chemical Kinetics, to become familiar with the

mode of operation and further application of program KINFIT.

33°C

111

ItpMfi P1665“ 12117) 1, ”nqu
DOPOVcCN45000.1201.1c£J.n..q,.

FTN.
ATTACH1K1KF1T¢K1NFIT1

. LOAD(K1NFII)
L60.
1 0000000000000003101900

snanourvvf ~09

suastolzen unsun-M5.uanuuaun§?z'

FGNVOM VWII' .[Tu”v.1ThW"o1J“T IADQRINCP. N09T NOVAQQVOHNKQXoUoITNAKQFON ‘52;

luTYoIrCT-IofVOQ‘KlqoI‘QoFOQQITYPOIIoUXTYPQH;lIofOPQFOQFUOPQZLOTnOEFQN
Ztnvol_.xsr.7.1r.t.2.111.v.nv.vrct.anT.cn~.q
nlurnqlow x14.1uu). ”(771. utx(4-1001. xx<a). #001100). rncloow. Frau I289
1”(100)- 31?29911- VFCT("o81)- 7L11301. 10(20). EIGVAL1201¢ XST11OEON 1286
70). Y110). DYIIUI. FON<T1151
. GO If.) (903040501). ITVP F0" 1295
l CO'TIMHF
1T0 0°F=60
JTADF=61 5”" ‘102
wPITE (JTMDF-fil EQN 1103
5 rnovar(e n:r:w«1n\v[om nF FORMATInN CONQTANT RY NA-23 NMR'1
mnukk:7
NOVA”: 2 FON 1319
pFTURM EQN 170
ON 1321

2 cnwtlnur

9: (1- “1.OOSORT11.0.(h. )0H(?)9XR(1))))/(?. 0°U121'XX111)1°(CON57111-U

11111111)
DFSID=Q-XX1?1
DFTH

1 CONIIMUF

4 CONTINHF

Q Cnutluufi
RFTURH

END
0000000000000000000000

1.0001
gngtun THIOCYAVATE IN TFTPAMEIHYLUQEA
0.60 12.00
.50 .0 0.73 79.0 0.00
.30 {.0 Ooqq PROD 0.22
.10 .0 1.hS 25.0 0.0
.01 1.0 ?o§S 29.0

rm—H
.0.
COO

133%

raw 1134
ten 1342
gnu 1168
.ow 341

NM»
IND“
no.

90‘!

APPENDIX II

DESCRIPTION OF COMPUTER PROGRAM KINFIT AND SUBROUTINE EQN

FOR THE RESOLUTION OF OVERLAPPING RAMAN AND INFRARED BANDS

DESCRIPTION OF COMPUTER PROGRAM KINFIT AND SUBROUTINE EQN

FOR THE RESOLUTION OF OVERLAPPING RAMAN AND INFRARED BANDS

One program employed to resolve overlapping peaks was program
KINFIT by manipulating SUBROUTINE EQN. This appendix lists a
typical deck for KINFIT analyses and describes the construction of
this deck.

Recall equation 27 which describes the lineshape of a Raman band

as a Lorentzian-Gaussian product:
I = Io{exp[(-3L3;)2/202]}{1+(3;;;)2/02}-1 (27)

where I is the arbitrary intensity at frequency 3} 3; is the position
of the line center with maximum intensity, Io; and o is the variance,
which gives the halfwidth when multiplied by 1.46. Letting U(l) = 10’

U(Z) = d, U(3) = 3;, and XX(1) = 3; we may rewrite the right hand

side of equation (27) into two equations to be summed as follows:

B = U(l)*EXP(-0.S*((XX(l)-U(3))**2)/(U(2)**2)) (29)
c = 1/(1+(lXX(1)-U(3))**2)/(U(2)**2)) (so)
A = B*C w (31)

Equations 29-31 describe one peak; to resolve two overlapping peaks,

we need to take a linear sum of two sets of equations similar to

112

113

28-30; and to resolve three overlapping peaks. A linear sum of
three sets is required. The deck listed here is an example of the
last case. Hence NOUNK=9 (three per peak) and NOVAR=2 (if two peaks
were to be resolved, NOUNK=6, one peak, NOUNK=3).

As before the first data card contains the number of data points
in columns l~5 (F15), the maximum number of iteractions allowed in
columns ll-lS (FIS) and the maximum value of A parameter/parameter
allowed for convergence to be assumed (0.0001 works well) in columns
41—50 (F10.6). The second data card displays any title the user
desires. The third data card contains initial guesses at the maximum
peak intensity, 10, in columns l-lO (F10.6), the variance, 0 (halfwidth/
1.46), in columns ll-20 (F10.6), the position of the peak, 36, in
columns 21-30 (F10.6) for one peak. This pattern is repeated
on the same data card for two or three peaks; however, if three
peaks are to be resolved, the guess at 3; for the third peak must
appear on the next data card as shown on the deck listing. The
fourth through N data cards display the wavenumber, XX(1) in columns
1-10 (F10.6), the relative variance of XX(1) in columns 11-20 (F10.6),
the intensity, I, at XX(1) in columns 21-30 (F10.6) and the relative
variance of the intensity in columns 31-40 (F10.6). This order is
repeated for the next data point yielding two data points per card.
If there are an odd number of data points, the last data card may
contain that one point with columns 41-80 blank. If no further data
are to be analyzed follow the last data card with a 6789 card. To
analyze more data sets, follow the last data card with the number of

points, number of iterations . . . etc. for the next data set.

114

 

 

2 ?3 96 5 23 901 “281 0000000000000000000000000
2 RR 88 9 00 122 3666 0000000000000000000000000
E 2? 72 a 31 ‘33 3‘13 0000000000000000000000000
0 11 11 1 11 111 1111 0000000000000000000000000
N 0.0000000000000000.one...
A NN NN N NN NNN NNNN 0000000000000000000000000
H 00 00 0 00 000 0000 22????2??2???227?722???7?
M FE FF E EF FFE EEFE .
H OE F0 0
0 x0 1 7
H An .1 cocoa...0.000000000000000
0 MT 1T 0006378253390307963278062
E To 05 1126069676870061762
H 1L 0x 111111123211
I 02 1
a U0 10
M .9 01 O
S ‘9 F0 9
H 9U 2 0
H KF 9‘ 3
No 1L 0000000000000000000000000
0 00 0A 0000000000000000000000000
E OF 0V 0000000000000000000000000
Z N0 15 0000000000000000000000000
1 Op (1 000.000.000.000.000000000
D 90 DE M llllll 1011011111110d10h1
1 AF 0 o
S V0 F0 6
3 01 1 2
U N1 90 9 .
5 0X )2 1 1 1 0 onooooooooooooooooooooooo
TOT6( 1 1 1 S 1357913579135791397913979
p.510 2 2 2 9 0000011111222223333366666
OPNXT C O O 9999999999999999999999999
NYOX 9 O O .
OTC O 1 \I 1 1
9X0.) 2 5 8 00
CDT1O I\ l! l‘ 00
NoSOO U U U 09.
1XC01 1 111111 0 8
XXNI‘ ’1/1/1 00000000000000000000000000
QQQQL 9 121212 E 00000000000000000000000000
ppTh? 7a 2029 .L 00000000000000.00.0000000000
AYC‘ R OGGGOO 1 00000000000000000000000000
LTEX' E 010161 R 0.000...00.000000000000000
01VT1 Z 121518 T 00000000000000000000000000
TooWl) Y 111111 10 22222222222222222222222222
WSY 26 L 3U6U9U . NS
91p“ 9 01 A ((111111 07
V oEo101 N U/U/U/ T1
0. FoYflET A .1.1.1 E cocoa...oooooooooooooooooo
DZ 09.?(9p 121212 C 01016753712695932892660900
HG AAJCTNY E 1a1910 A 17769967670757696111
PR TIJUCOT V (01619 11111122211
9 JoJ FCI Q X1X1X1 N
0 9n 9 9V ”II x1x1x11 I
00 Flu) 00 C (116196 e
10 pg 90 9)) ((1111. E9
2 0 AEL0101. T aUerUF 11
C 1 0 TR.1110 11 9.5.9.1 A9
2J T 0N19T0715 6F 01.1.10 9 00000000000000000000000000
7. o 1 00 0V0“ vi 9 ON 01010111 0 00000000000000000000000000
10 F 06:10 91004 E1 .1-1 .152 .L 00000000000000000000000000
30 N 0 NQTXZ 9 9K (X1X1X91 H 00000000000000000000000000
22 1 0EU10 (’3 A pXPXPXDx C .0000.00.00.00...coco-coon.
11 K 0N0 OTNP) O 10 x1x1x11x 0R la. 11111 0110011de1 . HHuw»1 . 111111.“
0 o OIKTSO 02E01J FE1E1E16. EEE SE
0 T 0T 3X1o110661192 UOooooolA UUU P
60 I1 0UNF0§11 N2: T=:VN)11111C:NNNNN o
60 FT OOOTLNOYOIEEEAKRR111617C9DRIIIR M3
65 k1 OPNQAFO TTPPTNKAUT1/1/1/EIUTTTU U 0.00000000.000000000000000
66 1F ORHXVMlo NAAIQUVTNU1U1U11STNNNTD 1 02668026680266802668026680
1M KN 0U01611190TT900050=::::=:EEOOOFN D 000001111122?2233333666665
2C (1 OSCHIDUOGCIJNFNNPCHCDEFGAQPCCCRE 0 99999999999999999999999999
O HK 0 12 12 5
CV C1 0 .1 6 2 365 .1
N0.AD.0 5H6.
PPNTAOO 033
[OTTOGO 063
[PFALLO 219

APPENDIX III

DESCRIPTION OF COMPUTER PROGRAM PEAKSBF FOR THE RESOLUTION

OF OVERLAPPING RAMAN OR INFRARED BANDS

DESCRIPTION OF COMPUTER PROGRAM PEAKSBF FOR THE RESOLUTION
OF OVERLAPPING RAMAN OR INFRARED BANDS

A second program used to resolve overlapping peaks was program
PEAKSBF, written by Mr. Patrick Kelly of this Department. This
appendix lists this program, which is on APLIB, and a typical deck to
access PEAKSBF.

As did the KINFIT program, PEAKSBF uses the Lorentzian-Gaussian
lineshape equation (27) to resolve the data. The first data card
contains any title the user desires. The second data cards contain
the initial wavenumber in columns 1-6 (F6.0), the final wavenumber in
columns 7-12 (F6.0) and the interval between data points in columns'
13-18 (F6.0). The third data card through N data cards contain
the input data. For a given pair of data points, the data card
contains the wavenumber in columns 1-10 (F10,4) and the corresponding
intensity in columns 21-30 (F10.4) for the first point and the
wavenumber and intensity of the second point in columns 41-50 and
61-70, respectively. In the deck listing, note that the relative
variances of the wavenumber and intensity appear in columns 11-20,
31-40, 51-60, and 71-80. PEAKSBF does not read these numbers; they
are present only to make each deck of data cards in proper format for
PEAKSBF and KINFIT analysis. The next card’contains the number of

peaks to be resolved in column 1 (F11) and the number of unknowns

115

116

per peak in column 2 (F11). Presently, the program can resolve
only two peaks or fit one peak, hence, column 1 will contain either
the number 2 or 1. There are always three unknowns per peak, so
column 2 will always contain the number 3. The next data card
contains the following initial guesses; the intensity of peak 1 in
columns 1-6 (F6.0), the position (cm-1) of peak 1 in columns 7-12
(F6.0), the variance (halfwidth/1.46) of peak 1 in columns 13-18
(F6.0), the intensity of peak 2 in columns 19-24 (F6.0), the position
of peak 2 in columns 25-30 (F6.0), and the variance of peak 2 in
columns 31—36 (F6.0). If fitting only one peak, leave columns
19-80 blank. The next data card contains the initial wavenumber in
columns 1-6 (F6.0) and the final wavenumber in columns 7-12 (F6.0).
The final data card contains the initial wavenumber in columns 1-6
(F16) and the final wavenumber in columns 7-12 (F16). Integer
format requires these values to be right justified. If no further
data is to be analyzed, the next card should be a 6789. To analyze
more data, the next card should be the title card for the next set
of data. -

To more fully understand the operation of this program and the

sequential simplex method, the user is referred to reference 126.

117

2

v

00000000000000000000000000000000000

 

 

Z 00000000000000000000000000000000000
E 00000000000000000000000000000000000
. 00000000000000000000000500000000000
" O....0OOOOOOOOOOOOOOOOOOOOOO..00...
‘ 00000000000000000000000000000000000
W Z?222222222227222222222222272222222
m
N 000.00.000.00...0000000000000000000
O 612722847134320818833315696335981‘“
E 11133456037012096666530976653211
H lllzzzallllllll
9
9
H
S
w
00000000000000000000000000000000000
D 00000000000000000000000000000000000
E 00000000000000000000000000000000000
7 0O000000000000000000000000000000000
ll OOOOOOOOOOOOOOOOOOOOOIOOOOOOO0.0...
D 1W1 hill 4 llJlelllllh.hndeWHlWl
I
S
W
S 0 oooooooooooooooocoo-cocoa...-one...
0 [357°135791357913579135791397913979
5 33313464449555566666777778888899999
. “4446644646“66444444446444444444444
o
3
4
000000000000000000000000000000000000
E 000000000000000000000000000000000000
L 000““.000000 00000000000 00000000000000
) 1 000000000000000000000000000000000000
ul- R .0.00.000.00.00.00.00.00.000.0......
: T 000000000000000000000000000000000000
m m 222222222222222222222222222222222222
o
9 0 0
V V T l
0.0 E OOOOCOOOOOOOOO0.000.00.00.00.0......
p?p C 031821037264109608092161161770412830
"GO A 1111149981691010766764203790122? o
POP 11122221111111! 0
c: N 7
on I 6
001
100 E
2F 0T 0
CH 0A . 2
2J5 0R 000000000000000000000000000000000000 7
70K 00 000000000000000000000000000000000000 1
10A 0L 000000000000000000000000000000000000
306 CH 000000000000000000000000000000000000
22p 0c coco...00000003000000.0000...0.0.000
110 ORnllLHllLJHHWLLTWWWnhntdnhhhnnnlnndnnd 0
OF GE. 0
09 opl l
60g 0 0
60K 0H 0
655 000000000000000000.coo-0000000000000... 0&4
6&5 01002“680246802463025680246802465024630 50
1MPo0”0333334“64495595666667777788888999990 50
2CMWWM5444444¢4644¢4444464b4446464b44446445 650
9
ccho 3
NOAKOHQ 00“
PPTAOOO 00
[OtEOOJ 303
IPAPOEQ 22“

.~.

118

punnqau DFAVC(I‘PHT. nuanT. TADF an : INPUT)
nyucwcy~u anxis!‘ly.-?\?(|3.).arCC([un)
nrnrmclww ,.(,.

PITCH"'“2 in“:

I_‘~
DIWJHRIVW EWL‘f' l)
DIMCHQIWW 55 ".I.
Qtufimqyny H( g)

(
1!
Pluc‘clnv 0(4)
nz'wg,<:n~a x(u)

9899 If:

- V(o\. ’(Q). w(w)o Htl).xX(H)

KO = n
DF\F(An.‘17OI) (3‘. “H"
p»,)!’ T H] 1h (”I‘D-"2‘7

Q79? Fn.)'JI\T{:\} ’)\

9596 FnD”Q[(!‘?iofi!7‘
DF'M". 1.51. (I. ‘:K
{Ftror

oaen ODIN!

1 rnqua
w =

A‘s-4
Ida-s

:RINU(I”H)oAQ(IMH)oCCCC‘INMDOIWN=19UL3
"(IV”)ohHII“N)oCFCC(IWN)oIdN=IOWL’

v0 o'.o-o 7‘

¢
11
3
U
2 - ..
D
>-Q #4
«a

.g
)
"Mllflfi

D
.71
>
1')

point
9 FnouA
[\(\ 53
no 7 to = I
sa¢7u1.tn»
CI‘NT {I 0“-
CoHTImwF
no 9 11 = Iota:
10! = In 0 l
sA(xnl.rn) = .HI 4 satrnl.ln) o sAtInl.IQ)

«'4
QafiJ—c—o—d—nb to a'vJE
~CODJDDJ

\

Dfl

ark", ‘00 FF. a";
pugwr in. r:. in
10 rnnuar l?V*.‘l
9th" [0%. lk‘. In?
DO’KY 10“. ‘FF. "an
106 FOQVAT (81%|
7a = O.
In? : [11 o i
102 = l
(‘0 In [AI = l-II)n
HH = 0.0
AM : (Ff - WI)/WV o l.
as no 15 In = I¢¥.:nc.l
H = Flflfif (II)
no I] I“ 1 0. I”!
”(10‘ = m‘('”lol‘!
l] COHTIHNF
IF (Inn - /) av.¢'.l?
]? IF (1%“ - 7i Flo"i.l~
11 no 10 h!
120 FORMAT (IN/o “Iowa ”an: [1 In THIQ pvtnfv)

119

5 AIB/A1419(H(4)°EXP(AI§/dl6)1/(10417/

,
’
2
U
I
‘
V:
I. o
l\ H o 0 \l \l
/ H I. 7 AU 0 I.
\l 9. «I ‘- 1: II. uh.
\l I. 0 9 on“ Q 0 ) 1|
9 D 9 A 1.. 1R nac. Q1. 1 I
n a I. 0? 7. «I? :40. A. )0 o X
o 9. a DA. Co. o co nos). 4 :1 A l.
1 o l a/ 9a Al 0 7% 00. c o! 1 A
DA 63% $1. 0) .44 ? )? 41“.? 0 PA n .I. o
01971¢147.\rll ll, 9 9 Q- R 0 011.0 7 R Cc ell ll
11* Quill». 14‘ "A {wet I) II?) on.“ 0 ”U '19.. vlln '11 ll
\l/\I 01/1 .0 llx o o a? O 9 11).”) A I 0 9n. OI! Cu
(19 /(fiQ/F 1A RI. cl 1.! =1 ? I. In ( .l. 1|
0((1UI\(1) nl 1 H O :11) :Kllko C1n.‘ \. 2V.“ 2A.! 2‘ 2x0
.wofi.uuh) . Hl) V2 V4 23118 1 = S x
U... .(U( _(1 . o o .9 13c s) a. nr 1 QKK‘O K .111 HUI. X 2 o 1 1:2
IOIO‘GH 0‘ .l E21410F 10 .qu/an .11?r|P.flfll-..U'I?CP. WU.” n n. C. ..rll M... FW:F. .J. WP
.?(1.2H\(D\ ADMIN . .1”. (I (OH. 0'2 (OH. 1. o 7.......(fi.”h7?111~1011:1‘l:.fllUK .I. n U1 ”'11....”
I. 0(AM,)QA M1V 1V1nvnv DAVIW. onnoasvolu GK 0 NH I. 1.1M. 1“ .l H
3: == = = == K 12.27....1 I.” .l 71...“ .10 011.1“ '10.... ======01Q§|111= .12 I?..l.|.:3W:I
3": =rln‘l\ TI: ._?l\==7l=2((::TIT=1I|I\(= u—TYI(= .1...P~«/l\n.1.vll.a :3H211H' «1.! fl
‘73h567n 9 NJ! 7.... V N N 21./4.56.5 7.. II. “1.1... 71 1V cl”? (3 N

IIIIIIIIIIFHHunt‘Fprnnlfiin—PWInlwnrpr 6?.nnqin—rFFq110nFIIWMucuvnwwnl‘nFID 1000(“N0I0U0‘U nu
AAAAAAAAAAVIHACII'YCKRDIRKCRDIIRKCF“0'31!“KCFIKIIIIIIO‘nleerl'lnv-vlcnxrclhvXclr.
l

a .I. 56 4730. 0 .1 11.567 2 345 0 .108 30,0 . .I- 2 3
l 0 11 all}. 2 2 2222 0 000 0 07“ 27.3 3 3 3
0 3 333 3 31
II.

3%

39
96
36
37

39

40
41

81
4?

41

4.1.
as

97
46

(.7
310
99

49

'11

m1:
V—L 3

120

i
x rm . '2 .-. «unau- vnwan

"21.-m. .Ls .. (~th - vuuan
9
0

falls 2 ['-
COMT [Ml [F
101 = K1
DO 16 I“ = 19109
§°(lnloTC) = XX(I“)
CGHYIHHF

IF (lfi-h.) 17.17.53
H” =
AM :

O

1
F - WI)/MK o l.
O

J'.

1
) 10041.“?

“I END-dd
I
o OI

.100
7(1w1)

3
3
4
D
3.~ x a Jxxcno

DO

9 Jul
ll

_fia
A—Hhflb"

DZ-‘d ~

179179h6

~
II N".
d
«E
O
U

0- XI.
—‘
933‘

= 1.100
= HIIWQ)

A
kw

r—e:<~
u

,
_
a—a.
-u/
7'1
~

A-d "AA II-d ll Ilnbll H A-IIII-O‘D II
',~ V—

“o’ 37OQQ.Q?

(I,
I”: : 90‘

Hl'flfil-QQJ ~3.a1.aa

"R—vadaxufixD-filxx~fihmnD—ww—fl—un3-o
~Ofl3bfiflbJfiLflDfiflflSDDDODflDODDJ)

a(7u5)
hut"
H [MHF
7“ . 10
HQ - 41) J?-07.&7
Kg - a) 16010.4“
0.
O 40
KM“) 11°o110.1w9
q = Kl
= K?
«on — «on . 1
Q1 : 94
7" - 0.
an m an
Ml =w1)
vnw = 0
TT = 11 . .
On Q) ["2 : 2.10M
A316 = AqS(-1-’II-\I‘(I‘9H
I" (mm - .rm ) shew?"
'7" = K
7F (TY-“00.) 10049.“?
rrw‘HH'
no :9 v0? = 1.309

121

I
polar R1. (0'. 64(1‘“"7‘
S3 FOP-”LT (l,:/.-§"1‘\I...'4C‘f,f.. -‘..‘;.T?: (>.L~1,).H)
otwr/v : Ka(l~' ’1

£34 (2M1 hill?
033‘ T 3v»

55 FOWUQT I'M/o//0|¥.’V WNWEW'MzYkfi'H"*V rOQDDIMATE¢tho”CALCULATF3
1v (QODHYNQTF*-4!9*UFCIWHAI*9/’
Kfih = «ma 0 l

H = WY
5‘! = I.
(In T” ‘(0
sq nn(’(:\y:) : .11
r» = AH(AH) - AI ‘

no[»r g7. u.A”(P”\.\!.'§
97 anunr «la/.?17.A.~<-Fl>.fiollYoFl’-6~l?x9El?-5’
IF (U - wJ) 89.6w.=0
SQ H = U o NK
A“. = «N ’ ‘0
an 10 1a
50 pork} «0

60 Fn'H’AT‘l“!ol“(.*1*).?fl(§+¢.+[u)./)
HRH : §J(1)
ARF = «0(1)
no 65 I = 2.‘w
IF (Ann-AH(I)) 87.69.51
6’ ARO = AR([)
63 IF (ARF-AH(‘)) bqohboAh
64 59F = AH(I)
Cnutxnur
“\I 3 III!
no 67 I = '0 4L
1” = (AQ(I)/(AHU-5WF\3000,o1,§
1"! = (ANC(I)/(AHn-1\HE))uqq.o1_q
D0 F” J = 1.10]
TD!PT(J‘ = 1H
6H leTIMUF
IF (IV-IN! 70.71.70
70 IDLCT(I”) = lfli .
I°|07(IN) = 1'“)
PRIhT 7?. AJ.IDLnT
72 FOQVAT (1H/~/. El?.6o 1x. 101a1)
- no 10 71
71 ID[01(IH) = 1H:
OQIKT 7 o AJ~I°I7T
o wx

3
lax.%IG.P1(fl‘o¢+‘fi)./)
. (aun-qur)/CQ.
ou[\1 u 1. a“:
801 Fn)”AT“H/.Q‘Hr UAOHF HF THF Y CHOQUINATE AT THE ROTTOW = “061205,
ought 204. D}
804 FnQVAT (IN/.vruF WiIH' 1F THV r CnnDUIMAIE AT THE Tap = «. 612.6!
hVK : (AQW-AJ?)/0).
DDVPT Q91. AV“
321 pnovny (IN/.OTHF ugLHF 0? EACH IMCQFMFNT = “9E1?.6)
(.0 = r
A; = {
I; ([51 - 2) QQ.0|.Q9
q? puHTr Q1
93 rnovnr (1H1.//.1y.ev Pvnnw'“1Tt*-‘Vo-“Y - CHQV‘ 1* ~7X-°Y - CUPVE
1p».7x.ev - FUQVF 1%./)
an to q”
91 DJIKT 90
94 rnqwnT (1H1.//.}x.uv rnnanxnnTFo.Ax.ov - CHDVF 19.7x.6Y - CURVE 2*
o/,
Ian TO PO
9n Data? OR
9% FnDVATl(lHl-//ol<.*¥ CWnnWIUFTE*q6X-9Y - CHPVE 1’0/)

89 I“! =
R! = 0.
HI = n.
RX 2 n.

122

n” F8 {:1 = 1.1;0.1)z
IF? = {fl o
IF." :: Vk'l 0
fi'; = -',\"-, .IL}‘.“ -
AVA : 70(:(v.‘,/,. ‘4w9
5‘ = l”(I*l)-~‘-("" n1a))
IV (IFI - 1 ) ‘ ..I~ 1.34;.
183 Q! = a!
G0 1'0 09
194 IF (Ir1-1-1nu; Hq.dk,fifi
49 0I : \I
an 1n H)
R". H’ (IF'l-l-Wu-‘T‘WH fun-17,43
R7 n“ = A!
82 (‘rWTIr-‘ut‘
1F (JhQ-l) H*o719.llé
RR 1‘ (Ian-9) 7fi1.7w;.7q1
701 DJYAT 704. n7. n!
70» FUUVAT (lH/.f12,~,g(,rlp.g)
GU TO 70%
70? nwle 79%. av. 3:. n)
70:; FWDVQT '1"/0*-‘9.‘.‘~X.F1r’.fi.fii.fi|?.h)

QJFfllzl7F3loq[#uv
ADFA82ADF&20H’§.<
AQEA3zbnfifii+fiK4uw
Gn T0 70”

701 ODIN! 706. n/.n[. 9'. AK
70“ FORMAT (lH/G’fil)ohoi‘xor[aofififixo'El206Ofi‘QF-lao6)
70R IF (07—NJ) 707.703.7QQ
707 02 = 07 0 wK
an TO 90
709 polnr an
07 : 1}
728 JflD 2 Jan . 1
GO TO R0
710 I“ = (AR(JAD)/lndn-nqc‘suqo..l.q
V” = (HI/(ARD-M'C'))uo<)..1.t,
IO = (RI/(AHH-ckryyouu..1.g
ID = (PK/(AHn-AHF;)§QQ.,1.5
Dfl 711J= ‘Qljl
- IDICY(J) = 1H
1 PONTIMHF
2 IF (IN-1n: 712.7)“,71:
h IULCT(ID) = [Ht
Y°IPT(IOJ = I”!
Ioln1(1”) : 14:

GO To 770
720 DQINT 79.07.13Lor
an In 7?h

71% I? (IN-tn) 724.717.71;

717 YCLPT(ip) z gu.
YDICT(IM) 2 !H1
19L"Y(lu) = 14:

G0 70 77C
716 IF (I"-13) 714.714.71‘3
713 IDLCT(Y“)

= 1H1
YOLPT(10) : [HY
IDIPTtI") = 14:
fin T0 790
719 IDlOT(IM) = lax
VQICT(FO) = 1”!
IOIPT(ID) : 14R
I”LC1(VM) = 1%”
N) IO 7 30
725 IF ()7-MJ) ~07.771.791

907 07 3.07 0 JK

DUV’I 9?]. .‘.L(
9 DQIKT ‘9“). .‘.~"‘\]. -'."C‘I\f). r.u."\1
O? FqD‘JAT( '4/Q/lol,.\‘a)i-\;( I 3 Q. (.r 9 h. XoGL’ A v = o
IQ'ZJ’K)FAX '11: “01Xorl?.fi) l l ' In E K a 'lx'EIZO‘NIO‘
Gn Tu Qqu
6] C’VITIMHF
FHn

APPENDIX IV

DESCRIPTION OF COMPUTER PROGRAM SHEDLOV FOR EVALUATION

OF CONDUCTANCE DATA

DESCRIPTION OF COMPUTER PROGRAM SHEDLOV FOR EVALUATION

OF CONDUCTANCE DATA

Program SHEDLOV was used to evaluate conductance data using the
CDC-6500 computer. Since this program is listed in the Ph.D. thesis
of J. A. Caruso, this appendix lists and describes the construction
of the data deck.

The first data card contains a title of the user's choice in
columns 1-56 (F7A8), a value for PER, which deletes those data
points where the solvent conductance is greater than PER of the
experimental conductance, in columns 62-69 (F8.2) and a value for
CALCL in columns 70-72 (F13). If CALCL=0, the complete program

/2

is used; if CALCL=1, then only Ao', C, A and C1 are calculated.
The second data card contains ETA, the viscosity of the solvent in
poise in columns l-lO (F10.0); DIELEC, the dielectric constant of

the solvent in columns 11-20 (F10.0); TEMP, the absolute temperature
of the experiment in columns 21-30 (F10.0); and ZERO, an initial
guess at A0 in columns 31-40 (F10.0). The third data card designates
control of the input data, IAM, in columns 1-5 (F15). For
experimental data, IAM=O whereas for literature data, IAM=1. The
number of data points, N, is in columns 6-10 (F15) and an integer
which controls the extent of program execution, IAC, is in columns

11-15 (PIS). IAC=1 for weak acid and bases, whereas, IAC=O for

electrolytes.

123

124

The fourth data card contains an integer which designates the
method of experimentation, OPT, in columns l-S (F15). If solvent
is added to a stock solution, OPT=1; OPT=O if stock solution is
added to solvent. The fifth data card contains the density of
the solvent, RHO, in columns 1-10 (F10.8); the conductance cell
constant, KONST, in columns 11-20 (F10.8); the conductance of pure
solvent, LSOLV, on columns 21-30 (FlO.8); the molecular weight of
the solute, M2, in columns 31-40 (F10.8); the weight of original
stock solution if OPT=l, WOSS, or the weight of original solvent if
OPT=O, SOLV, in columns 41-50 (F10.8); the value of (grams of solute)/
(grams of stock solution), RATIOl, in columns 51-60 (F10.8); and
the value of (grams of solvent/grams of stock solution), RATIOZ
in columns 61-70 (F10.8). The sixth through N data cards contain
for each data point the resistance of solution R(I), in columns
l-lO (F10.5); the weight of added (not total) solvent if 0PT=1,
ADSU(I) or the weight of added (not total) stock solution if 0PT=2,
WTSS(I) in columns 11-20 (F10.5); and an integer signifying the end
of a given set of data, IJ, in columns 21-22 (F12) on the last data
point card of that given set. If more data are to be analyzed,
the next card should be the title card of the next data set. When
all set(s) of data are run, any integer, IM, in columns 57-61
(FIS) should appear on the last data point card of the final set

along with IJ.

125

0.10

CONDUCTAKCF OF ROOIUM [GUIDE IN PYQIDINE

.000160209.999859791

l
S
7
l
o
h
7..
9
R.
O
o 9
0 a
7 l
R.
.
F
a
6 0
1|. 0
O l
8
q 10
2 7.1..
0
0
o
6 6,456
5 29217257796520
0 Q: COQQQCR,3449?H.O
3 9 QDROQAOR97OR§0
o 30.5300759918 o o .0
? nanosecoooooalQo
I. 00112222314579.1230
0
0
O

LITERATURE C ITED

 

(1)
(2)
(3)

(4)

(5)
(6)
(7)
(8)
(9)
(10)

(11)
(12)

(13)

(14)

(15)

(16)

LITERATURE CITED

T. R. Griffiths and M. C. R. Symons, M01. Phys., g, 90 (1960).

A. K. Covington and M. J. Tait, Electrochim. Acta, 12, 123 (1967).

 

K. H. Wong, G. Konizer and J. Smid, J. Amer. Chem. Soc., 92,
666 (1970).

 

S. Kopolow, Z. Machacek, V. Takaki and J. Smid, J. Macromol.
Sci.—Chem., AZ, 1015 (1973).

 

H. Wang and P. Hemmes, J. Amer. Chem. Soc., 2a, 5115 (1973).

 

H. K. Frensdorff, J. Amer. Chem. Soc., fig, 600 (1971).

 

B. W. Maxey and A. I. Popov, J. Amer. Chem. Soc., 90, 4470 (1968).

 

 

E. Schaschel and M. C. Day, J. Amer. Chem. Soc., kg, 503 (1968).

V. I. Chizhik, Strukt. Rol. Vody. Zhivm. Orgizme, k, 126 (1966).

 

B. P. Fabrikand, S. S. Goldberg, R. Leifer and S. G. Angen,
M01. Phys., Z, 425 (1963).

J. N. Shoolery and B. J. Adler, J. Chem. Phys., 2;, 805 (1955).

 

D. N. Nicholls and M. Swarc, J. Phys. Chem., 71, 2727 (1967).

 

 

a) Z. Luz and S. Meiboom, J. Chem. Phys., 40, 1066 (1964).
b) Ibid., an, 1058 (1964).

A. Fratiello, R. E. Lee, D. P. Miller and V. M. Nishida,
Mol. Phys., 12, 551 (1966).

R. P. Taylor and I. D. Kuntz, Jr., J. Amer. Chem. Soc., 22,
4815 (1970).

 

C. Deverell and R. E. Richards, Mol. Phys., 10, 551 (1966).

126

 

127

(17) C. Deverell and R. E. Richards, Mol. Phys., 16, 421 (1969).

(18) H. Wennerstorm, B. Lindman and S. Forsen, J. Phys. Chem., 75,
2936 (1971).

 

(19) C. H. Langford and T. R. Stengle, J. Amer. Chem. Soc., 91,
4014 (1969).

 

(20) R. G. Bryant, J. Phys. Chem., 73, 1153 (1969).

 

(21) G. E. Maciel, J. K. Hancock, L. F. Lafferty, P. A. Mueller
and W. K. Musker, Inorg. Chem., g: 554 (1966).

 

(22) J. W. Akitt and A. J. Downs, in "The Alkali Metals Symposium",
The Chemical Society, London, 1967, p. 199.

(23) E. T. Roach, P. R. Handy and A. I. Popov, Inorg; Nucl. Chem.
Letters, 2» 359 (1973).

 

(24) Y. M. Cahen, P. R. Handy, E. T. Roach and A. I. Popov,to be
published.

(25) J. D. Halliday, R. E. Richards, F. R. Sharp and R. R. Sharp,
Proc. Roy,Soc. Lond. A, 313, 45 (1969).

 

(26) A. Carrington, F. Dravnicks and M. C. R. Symons, Mol. Phys.,
3, 174 (1960).

(27) E. G. Bloor and R. G. Kidd, Can. J. Chem., 52, 3926 (1972).

 

(28) o. Jardetzky and J. E. Wertz, J. Amer. Chem. Soc., 8g, 318
(1960).

 

(29) G. A. Rechnitz and S. B. Zamochnick, J. Amer. Chem. Soc.,
88' 2953 (1964).

 

(30) M. Eisenstadt and H. L. Friedman, J. Chem. Phys., 44, 1407
(1966).

 

 

(31) V. S. Griffiths and G. Socrates, J. Mol. Spectrosc., 27,
358 (1968).

(32) R. E. Richards and B. A. Yorke, Mol. Phys., 6, 289 (1963).

 

(33) E. G. Bloor and R. G. Kidd, Can. J. Chem., 46, 3425 (1968).

 

(34) a) R. H. Erlich, E. T. Roach and A. I. Popov, J. Amer. Chem.
§2£:’ 9%, 4989 (1970).

 

b) R. H. Erlich and A. I. POpov, ibid., 23, S620 (1971).

c) M. Herlem and A. I. Popov, ibid., 94, 1431 (1972).

 

(35)

(36)
(37)
(38)

(39)

(40)

(41)

(42)

(43)

(44)
(45)

(46)

(47)

(48)

(49)

(50)

(51)

(52)
(53)

(54)

128

G. J. Templeman and A. L. Van Geet, J. Amer. Chem. Soc., 94,
5578 (1972).

 

A. L. Van Geet, J. Amer. Chem. Soc., 94, 5583 (1972).

 

G. W. Canters, J. Amer. Chem. Soc., 94, 5230 (1972).

 

F. W. Cope, J. Gen. Physiol., 50, 1353 (1967).

 

C. A. Rotunno, V. Kowalewski and M. Cereijido, Biochim. Biophys.
Acta, 11%, 170 (1967).

 

 

D. H. Haynes, B. C. Pressman and A. Kowalsky, Biochemistry,
(Q, 852 (1971).

 

T. L. James and J. H. Noggle, J. Amer. Chem. Soc., 21, 3429
(1969).

 

T. L. James and J. H. Noggle, Proc. Nat. Acad. Sci. (U.S.A.)
8%, 644 (1969).

 

F. Ostroy, T. L. James, J. H. Noggle and L. E. Hokin, Fed.
Proc., 31, 1207 (1972).

T. L. James and J. H. Noggle, Anal. Biochem., 49, 2081 (1972).

 

T. L. James and J. H. Noggle, Bioinorg. Chem., 2, 69 (1972).

 

J. Andrasko and S. Forsen, Biochem. Biophys. Res. Commun., 52,
233 (1971).

 

J. M. Ceraso and J. L. Dye, J. Amer. Chem. Soc., 95, 4432 (1973).

 

E. Shchori, J. Jagur-Gradzinski, Z. Luz and M. Shporer,
J. Amer. Chem. Soc., 93, 7133 (1971).

 

J. C. Evans and G.Y—S. Lo, J. Phys. Chem., 69, 3223 (1965).

 

W. F. Edgell, A. T. Watts, J. Lyford, IV and W. M. Risen, Jr.,
J. Amer. Chem. Soc., 88, 1815 (1966).

 

W. F. Edgell, A. Lyford, IV, R. Wright, W. Risen, Jr. and
A. Watts, J. Amer. Chem. Soc., 92, 2240 (1970).

 

B. W. Maxey and A. I. Popov, J. Amer. Chem. Soc., 89, 2230 (1967).

 

B. W. Maxey and A. I. Popov, J. Amer. Chem. Soc., 91, 20 (1969).

 

 

J. L. Wuepper and A. I. Popov, J. Amer. Chem. Soc., 21, 4253 (1969).

(55)

(56)

(57)

(58)
(59)

(60)

(61)

(62)

(63)

(64)

(65)

(66)

(67)

(63)

(69)

(70)

(71)

(72)

129

J. L. Wuepper and A. I. Popov, J. Amer. Chem. Soc., 9%, 1493
(1970).

 

M. K. Wong, W. J. McKinney and A. I. Popov, J. Phys. Chem.,
ZR, 56 (1971).

 

M. K. Wong and A. I. Popov, J. Inorg. Nucl. Chem., 33, 1203
(1970).

 

W. J. McKinney and A. I. Popov, J. Phys. Chem., 74, 535 (1970).

 

P. R. Handy and A. I. Popov, Spectrochim. Acta, 28A, 1545 (1972).
W

 

M. S. Greenberg, D. M. Wied and A. I. Popov, Spectrochim. Acta,
29%, 1927 (1973).

 

A. T. Tsatsas and W. M. Risen, Jr., J. Amer. Chem. Soc., 9%,
1789 (1970).

 

 

J. A. Olander and M. C. Day, J. Amer. Chem. Soc., 93, 3584
(1971).

E. G. Hohn, J. A. Olander and M. C. Day, J. Phys. Chem., 73,
3880 (1969).

 

W. F. Edgell, J. Lyford, IV, A.Barbetta and C. I. Jose, J. Amer.
Chem. Soc., 93, 6403 (1971).

W. F. Edgell and J. Lyford, IV, J. Amer. Chem. Soc., 23, 6407
(1971).

 

W. R. Robinson, A. Barbetta and W. F. Edgell, to be published.

 

A. T. Tsatsas and W. M. Risen, Jr., Chem. Phys. Lett., 7, 354
(1970).

A. T. Tsatsas, R. W. Stearns and W. M. Risen, Jr., J. Amer.
Chem. Soc., 94, 5247 (1972).

B. W. Maxey, Ph.D. Thesis, Michigan State University, East
Lansing, Michigan, 1968.

J. L. Wuepper, Ph.D. Thesis, Michigan State University, East
Lansing, Michigan, 1969.

P. R. Handy, Ph.D. Thesis, Michigan State University, East
Lansing, Michigan, 1972.

M. K. Wong, Ph.D. Thesis, Michigan State University, East
Lansing, Michigan, 1971.

 

130

(73) R. E. Hester and R. A. Plane, J. Chem. Phys., 38, 249 (1963).

 

 

(74) R. E. Hester and R. A. Plane, Inorg. Chem., 3, 768 (1964).

(75) R. E. Hester and R. A. Plane, Inorg. Chem., 3, 769 (1964).

 

(76) B. M. Gatehouse, S. E. Livingstone and R. S. Nyholm, J. Chem.
Soc., 4222 (1957). "“““

(77) J. I. Bullock and F. W. Parrett, Chem. Commun., 157 (1969).

 

(78) A. B. P. Lever, E. Mantovani and B. S. Ramaswamy, Can. J. Chem.,
49, 1957 (1971).

 

(79) D. E. Irish, A. R. Davis and R. A. Plane, J. Chem. Phys., 50,
2262 (1969). -

 

(80) G. J. Janz, K. Balasubrahmanyam and B. G. Oliver, J. Chem. Phys.,
51, 5723 (1969).

 

 

(81) K. Balasubrahmanyam and G. J. Janz, J. Amer. Chem. Soc., 22,
4189 (1970).

(82) G. J. Janz, M. J. Tait and J. Meier, J. Phys. Chem., 71,
963 (1967).

 

(83) C. C. Addison, D. W. Amos and D. Sutton, J. Chem. Soc. (A),
2285 (1968).

 

(84) D. J. Gardiner, R. E. Hester_and W. E. L. Grossman, J. Chem.
Phys., 59, 175 (1973).

(85) J. H. Roberts, A. T. Lemley and J. J. Lagowski, Spectr. Lett.,
5, 27 (1972).

 

 

(86) K. R. Plowman and J. J. Lagowski, J. Phys. Chem., 18, 143 (1974).

(87) E. B. Baker, L. W. Burd and G. N. Root, Rev. Sci. Inst., 36,
1495 (1965).

 

(88) 0. H. Live and s. I. Chan, Anal. Chem., 42, 791 (1970).

 

(89) M. J. 0. Low, J. Chem. Ed., 47, A163 (1970).

 

(90) H. B. Thompson and M. T. Rogers, Rev. Sci. Instrum., 27, 1079
(1956).

 

(91) H. M. Daggett, E. J. Bair and C. A. Kraus, J. Amer. Chem. Soc.,
75, 799 (1951).

 

 

(92)

(93)

(94)

(95)

(96)

(97)

(98)

(99)

(100)

(101)

(102)

(103)

(104)

(105)

(106)

(107)

(108)

(109)

131

 

T. Shedlovsky, J. Franklin Inst., 22%, 739 (1938).

D. A. Burgess and C. A. Kraus, J. Amer. Chem. Soc., 70,
706 (1948).

 

D. N. Bhattacharyya, C. L. Lee, J. Smid and M. Szwarc,
J. Phys. Chem., 69, 608 (1965).

 

C. Carjaval, K. J. Tolle, J. Smid and M. Szwarc, J. Amer.
Chem. Soc., 87, 5548 (1965).

a) V. Gutmann and E. Wychera, Inorg. Nucl. Chem. Lett., 2,
257 (1966).

 

b) V. Gutmann, "Coordination Chemistry in Nonaqueous Solvents",
Springer—Verlag, Vienna, 1968.

U. Mayer and V. Gutmann, Struct. Bonding (Berlin), 12, 113
(1972).

 

J. L. Dye and v. A. Nicely, J. Chem. Ed., $8, 443 (1971).

 

L. S. Frankel, T. R. Stengle and C. H. Langford, Chem. Commun.,
393 (1965).

 

Ibid., J. Phys. Chem., 74, 1376 (1970).

 

R. L. Bodner, M. S. Greenberg and A. I. Popov, Spectr. Lett.,
5, 489 (1972).

 

J. J. Lindberg, J. Kenttamaa and A. Nissema, Suomen Kemistilehti,
34%, 98 (1961).

 

R. Figueroa, E. Roig and H. Szmant, Spectrochim. Acta, 22,
587 (1966).

 

A. Luttringhaus and H. W. Dirksen, Angew. Chem. Intern. Ed.
Engl., 3, 260 (1964).

 

C. Beguin and H. Gunthard, Helv. Chim. Acta, 42, 2262 (1959).

 

A. K. Covington, T. H. Lilley, K. E. Neuman and G. A. Porthouse,
J. C. S. Faraday I, 69, 963 (1973).

 

A. K. Covington, K. E. Neuman and T. H. Lilley, J. C. S.
Faraday I, 69, 973 (1973).

 

L. S. Frankel, C. H. Langford and T. R. Stengle, J. Phys. Chem.,
74, 1376 (1970).

 

P. C. Lauterbur, Ann. N. Y. Acad. Sci., ZQ’ 841 (1958).

 

(110)

(111)

(112)

(113)
(114)
(115)
(116)

(117)

(118)

(119)

(120)

(121)

(122)

(123)

(124)

(125)

(126)
(127)
(128)
(129)

(130)

132

W. M. Litchman and D. M. Grant, J. Amer. Chem. Soc., 29,
1400 (1968).

 

R. H. Erlich, Ph.D. Thesis, Michigan State University, East
Lansing, Michigan, 1971.

J B. Kinsinger, M. M. Tannahill, M. S. Greenberg and
A. I. Popov, J. Phys. Chem., 17, 2444 (1973).

 

A. D'Aprano, J. Phys. Chem., ZS, 3290 (1971).

 

A. D'Aprano, J. Phys. Chem., 76, 2920 (1972).

 

B. F. Prine and J. E. Prue, Trans. Faraday Soc., 6%, 1257 (1966).

 

S. Minc and L. Werblan, Roczniki Chemii, 4Q, 1537 (1966).

 

R. L. Kay, B. J. Hales and G. P. Cunningham, J. Phys. Chem.,
71, 3925 (1967).

 

P. G. Sears, G. L. Lester and L. R. Dawson, J. Phys. Chem.,
60, 1433 (1956).

 

D. P. Ames and P. G. Sears, J. Phys. Chem., 59, 16 (1955).

 

B. J. Barker and J. A. Caruso, J. Amer. Chem. Soc., 93,
1341 (1971).

 

S. Mine and L. Werblan, Electrochimica Acta, Z, 257 (1962).

 

J. E. Prue and P. J. Sherrington, Trans. Faraday Soc., 57,
1795 (1961).

 

R. M. Fuoss, J. Amer. Chem. Soc., 81, 2659 (1969).

 

K. L. Hsia and R. M. Fuoss, Proc. Nat. Acad. Sci., U.S.,
57, 1550 (1967).

 

A. R. Davis, D. E. Irish, R. B. Roden and A. J. Weerheim,
Appl. Spec., 26, 384 (1972).

 

S. N. Deming and S. L. Morgan, Anal. Chem., 45, 278A (1973).

 

G. J. Janz, J. Electroanal. Chem., 29, 107 (1971).

 

R. Baum, private communication.
Y. Cahen and A. I. Popov, to be published.

J. F. Coetzee and W. R. Sharpe, J. Soln. Chem., 1, 77 (1972).

 

(131)

(132)

(133)

(134)

133

B. S. Krumgarz, K. P. Mishchenko, D. G. Traber and Y. P.
Tseretell, J. Structural Chem., USSR (Eng. Trans.), 13, 373
(1972).

 

B. S. KrumgaPz, Y. I. Gerzhberg, gt: 31:, Zh. Fiz. Khim.,
44, 227 (1970).

 

J. A. Caruso, Ph.D. Thesis, Michigan State University, East
Lansing, Michigan, 1967.

B. J. Hathaway and A. E. Underhill, J. Chem. Soc., 3091 (1961).

 

 

   

WWITIWIWW1111111111111111mm!“
3 1293 030617793