LIBRARY
Michigan State
University

 

 

 

PLACE ll RETURN BOXtonmavothhMomtlomywncord.

TO AVOID FINES Mum on or baton data duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU to An math. ActbrVEcpnl Opponmlty Inflation
Wanna-n1

FI'IR STUDIES OF WATER AND ARENE SORPTION
MECHANISMS ON 'I'MA- AND TMPA- MONTMORILLONITES

Jeffrey Jay Stevens

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

DOCI‘OR OF PHILOSOPHY

Department of Crap and Soil Science

1994

ABSTRACT

FTIR STUDIES OF WATER AND ARENE SORPTION
MECHANISMS ON TMA— AND 'I'MPA- MONTMORILLONITES

By

Jeffrey Jay Stevens

This research evaluated properties of normal- and reduced-charge Wyoming
rnontmor'illonites saturated with tetramethylarmnonium (TMA) and
trimethylphenylammonium (TMPA) ions. Infrared spectroscopy was used to determine
mechanisms of arcne and water sorption on these clays and other properties. For normal-
charge TMPA-montrnorillonite, infrared spectroscopy showed that the C-N bond axis of
TMPA was neither perpendicular nor parallel to the surface, yet X-ray data suggested that
the TMPA phenyl ring was perpendicular to the siloxane surface. Reduced-charge
montrnorillonite was a randomly interstratified mixture of 25% collapsed layers with no
adsorbed cations and 75% expanded layers that were propped open by TMPA’s methyl
groups, not the aromatic ring. Water vapor sorption isotherms showed water sorption was
greater for normal-charge clays than reduced-charge clays, though the N2 surface area of
the reduced-charge clays was larger. This suggested that water sorbed preferentially on
cations, not the siloxane surface. 'I'MA-saturated clays sorbed more water vapor than did
'I'MPA-samrated clays, indicating that the phenyl group of TMPA may sterically hinder

sorption of water vapor. The infrared spectra of TMA and TMPA cations saturating

normal- and reduced-charge montrnorillonite were perturbed by water vapor at 7.5% RH,
providing further evidence that water interacted preferentially with adsorbed TMA and
TMPA, not with uncharged siloxane surfaces, at low relative humidity. Arene sorption
perturbed vibrational frequencies of adsorbed TMA and TMPA differently than did water
sorption, allowing spectroscopic differentiation between water and arene interaction with
TMA and TMPA. The infrared data showed that benzene and ethylbenzene interact directly
with adsorbed TMA and TMPA ions on dry clay, but do not preclude sorption on the
siloxane surface. The cation vibrational frequencies of clay films exposed to both water
vapor and saturated benzene vapor remained at the frequencies characteristic of benzene
alone for all relative humidity treatments. The cation vibrational frequencies of films
exposed to ethylbenzene and water vapor shifted from those characteristic of ethylbenzene
vapor alone toward frequencies characteristic of water vapor alone as relative humidity
increased. Thus, water sorption drove ethylbenzene, but not benzene, from cation sites.
Water vapor sorption and higher clay charge density both inhibited sorption of larger arenes

more than smaller arenes.

ACKNOWIEGEMENTS

I would like to thank my advisor, Dr. Sharon Anderson for her encouragement,
training, advice, and hard work in obtaining financial assistance for my Doctoral program
and in editing the manuscript for this dissertation. I would also like to thank her for
motivating me when I fell short of my potential. Special thanks also go to my committee
members, Dr. Steve Boyd, Dr. Boyd Ellis, and Dr. Dave Long, for helping me during my
comprehensive exam, and in their suggestions for my dissertation. Thanks also go to Drs.
Max Mortland and George Leroi for their comments. I would also like to thank the Clay
Minerals Society and DOW for providing financial support for this research.

Thanks also go to my friends and fellow graduate students in the Crop and Soil
Sciences Department for their technical assistance and camaraderie, especially my office
mates Sven Boem and Inez Toro-Suarez.

Special thanks go to my wife and friend Donna for providing understanding, love
and support, and also for asking the question “So what’s this good for?” Finally, I would
like to thank the rest of my family (mom, dad, brother and sister) for their support and

understanding.

iv

TABLE OF CONTENTS

List of Tables ix
List of Figures xi
Chapter I. Introduction and Review of Literature 1
Introduction 1
Review of Literature 2
TMPA orientation 2

Water adsorption on clays 6
Competitive sorption of arenes and water 8

References 12

Chapter II. Orientation of the Phenyl Group of Trimethylphenylarnrnonium

on Normal- and Reduced-charge Wyoming Montrnorillonite 15
Abstract 15
Introduction 16
Materials and Methods 18

Preparation of reduced- and normal-charge TMPA montrnorillonite 18

Infrared dichroism 19
Preparation of methyl-deuterated TMPA 19
Clay film preparation 22
Spectral collection 23

X-ray diffraction and Fourier analysis 24

Results 26

Infrared dichroism 26

X-ray diffraction analysis 28
Glycerol-solvated Na-montmorillonite 28
TWA-montrnorillonite 30

Discussion 32
TMPA orientation on normal-charge montrnorillonite 32
TMPA orientation on reduced-charge montrnorillonite 36

Conclusions 39

References 41

Chapter III. An FI'IR study of water sorption on tetramethylammonium- and
trimethylphenylarnmonium-saturated normal- and reduced-charge Wyoming
montrnorillonite 4

Abstract 43
Introduction 44
Materials and Methods 46
Reduced- and normal-charge organoclay preparation 46
Water vapor sorption isotherms 48
Infrared spectroscopy ' 50
Clay film preparation 50

Water vapor sorption on clay films 50
Results and Discussion 53
Water sorption isotherms 53
Effect of sorbed water on cation vibrations 57
TMA-methyl vibrations 57
TMPA-methyl vibrations 63
TMPA ring vibrations 67
Conclusions 69

References

Chapter IV. An F'I'IR study of competitive water-arcne sorption on
tetramethylammonium- and trimethylphenylammonium—saturated normal-
and reduced-charge Wyoming montrnorillonite
Abstract
Introduction
Materials and Methods
Reduced— and normal-charged organoclay preparation
Infrared Spectroscopy
Clay film preparation
Competitive water and arcne vapor sorption on clay films
Results
TMA-montrnorillonite
Arene sorption on dry clay
Benzene sorption
Benzene-d6 sorption
Ethylbenzene sorption
Summary for arcne sorption on dry TMA-clay
Arene-water competition on hydrated TMA-montrnorillonite
Benzene-d6 sorption
Ethylbenzene sorption
Summary for arcne sorption on hydrated TMA-clay
Effect of layer charge on arcne sorption on TMA-clay
Effect of size on arcne sorption on TMA-clay
TMPA-montrnorillonite
Arene sorption on dry TMPA-clay

Benzene sorption

70

72

72
73
75
75
77
77
77
80
80
80
80
83
83
86
86
86
87
88
88
88
89
89
89

Ethylbenzene sorption
Summary for arcne sorption on dry TMA-clay
Arene-water eernpetition on hydrated TMA-montrnorillonite
Benzene sorption
Ethylbenzene sorption
Summary for arene sorption on hydrated TMA-clay
Effect of layer charge on arcne sorption on TMA-clay
Effect of solute size on arcne sorption on TMA-clay
Discussion
References
Appendix A. Supplementary FI'IR Dichroism and X-ray Diffraction Data
Materials and Methods
Clay film preparation

Spectral collection

Appendix B. Evaluation of ATR-FI'IR for Determination of Benzene
Sorption Mechanisms from Methanol on Clays

Materials and Methods
Clay film preparation
Spectral collection
Appendix C. Supplementary ATR-FI'IR sorption spectra
Appendix D. Supplementary mixed benzene-water sorption spectra

92
94
94
94
95
95
95
96
97
100
101
101
101
101

108

108
108
108
115
151

Chapter II

Table 1

Table 2

Table 3

Chapter 111

Table 1

Table 2

Table 3

LIST OF TABLES
Page

Selected physical and chemical properties of normal- and
reduced-charge TMPA-montrnorillonite 20

Terms used for calculating the Fourier transforms of X-ray
diffractograms of glycerol-solvated, normal- and reduced-
charge Na+-saturated montrnorillonite 24

Terms used for calculating the Former transforms of
X-ray diffractograms of normal- and reduced-charge
TMPA montrnorillonite 25

Selected physical and chemical properties of normal
(Norm)- and reduced (Red)—charge TMA— and TMPA-
montrnorillonite. 47

Solutions used to regulate the partial pressure of water

vapor in the desiccator and in the controlled-atmosphere

F'I'IR cell used to equilibrate TMA- and TMPA-saturated
smectites with water vapor. 52

Infrared band assignments for methyl symmetric and

asymmetric deformation vibrations of chloride, bromide,

and iodide salts of tetramethylammonium (TMA) in

pressed KBr pellets, and observed peak positions of

TMA-Cl 1)dispersed in KBr using diffuse reflectance

(DRIFT) and 2)dissolved in methanol using attenuated

total reflectance (ATR). 60

Table 4

ChapterIV

Table 1

Table 2

Table 3

Table 4

Table 5

Infrared band assignments for the v19, and vlgb C—C ring
stretch of methyl-demented (d9) TMPA-iodide and for
the v19... Vlgb, and methyl asymmetric and symmetric
deformation vibrations for TMPA-Br in l)KBr using
diffuse reflectance (DRIFT) and 2)in methanol using
attenuated total reflectance (ATR)

Selected physical and chemical properties of normal
(Norm)- and reduced (Red)—charge TMA- and TMPA-
montrnorillonite.

Solutions used to regulate the partial pressure of water
vapor in desiccator and in the controlled-atmosphere
FTIR cell.

Infrared band assignments for methyl symmetric and
asymmetric deformation vibrations of chloride, bromide,
and iodide salts of tetrarnethylammonium (TMA) in
pressed KBr pellets, and observed peak positions of
TMA-C1 1)dispersed in KBr using diffuse reflectance
(DRIFT) and 2)dissolved in methanol using attenuated
total reflectance (ATR).

Infrared band assignments for neat liquid benzene,
ethylbenzene, and benzene-d45-

Infrared band assignments for the v19. and vlgb C-C
ring stretch and the methyl asymmetric and symmetric
deformation vibrations for TMPA-Br in KBr using diffuse
reflectance (DRIFT) and in methanol using attenuated total
reflectance (ATR).

76

78

81

81

Chapter II

Figure 2-1

Figure 2-2

Figure 2-3

Figure 2-4

Figure 2-5

LIST OF FIGURES
Page

Selected infrared vibrations of TMPA adsorbed on normal-
charge and reduced-charge Wyoming montrnorillonite for

90' and 45' angles of incidence and for TMPA-dg iodide

pressed in a KBr pellet (bottom spectrum). Inset at right

shows the orientations of the symmetry axes of TMPA

and the band assignments for TMPA-dg vibrations along

these axes. 27

Fourier transforms of X-ray diffraction data for glycerol-
solvated Na—montmorillonites. Peaks due to combinations

of expanded (E) and collapsed (C) layers are labeled for
reduced-charge montrnorillonite. 29

Fourier transforms of X-ray diffraction data for TMPA-
montrnorillonites. Peaks due to combinations of expanded

(E) and collapsed (C) layers are labeled for reduced-charge
montrnorillonite. 31

Schematic representation of proposed TMPA orientation

on normal-charge montrnorillonite, with phenyl ring
perpendicular to the siloxane surface. Interlayer electron
probabilities from onedimensional Fourier analysis are

shown at the right. Both aromatic and methyl protons key

into the siloxane surface. 33

Schematic representation of TMPA orientation on expanded
layers of reduced-charge Wyoming montrnorillonite.

Layers are propped open by methyl groups, with some

methyl protons keying into siloxane surface. Phenyl ring

is not parallel to the surface, but exact orientation is not

known. 37

Chapter 111

Figure 3—1

Figure 3-2

Figure 3-3

Figure 3-4

Figure 3-5

DiagramoftheMcBain balanceusedtocollectwater

vapor sorption isotherms on TMA- and TMPA-saturawd
smectites. Water vapor pressure is controlled by releasing

water vapor from the sorbate reservoir into vacuum

Quantity ofwateron the clay is determined by measuring

the displacement of the quartz springs attached to the

sample holders. 49

Diagram of the controlled-atmosphere FI'IR cell used to
equilibrate TMA- and TWA-saunter! smectites with
water vapor and collect spectra of clay-water complexes 51

Example of the procedure used to normalize all spectra to

the same film thickness by normalizing the absorbance of

the lattice O-I-I stretch. A) Original spectrum.

B) Original spectrum with the lattice O-H Stretching

vibration removed. C) difference spectrum of A - B,

giving the lattice O-H stretching vibration only and the
multiplication factor necessary to bring the absorbance of

this band to 0.700A. D) Original spectrum (A) after

multiplying it by the factor calculated in (C). 54

Water vapor sorption isotherms on normal- (squares) and

reduced-(circles) charge clay saturated with TMA (left) and
TMPA (right). Graph also shows number of water

molecules sorbed per cation at different values of x/rn 55

Infrared spectra of reduwd-charge (top) and normal-

charge (bonom) TMA-saturated Wyoming montrnorillonite

at six different relative hurnidities. The left side

(3600 - 2800 cm'l) shows the changes in intensity of

the sorbed water O-H stretching vibration (3400 cm'l)

with changes in relative humidity. The right side

(1800 - 1500 cm4) shows the changes in intensity of

the sorbed water O-H bending (1630 cm'l) vibration

with changes in relative humidity. 58

Figure 3—6 Infrared spectra of reduced-charge (top) and normal-

Frgure 3-7

Figure 3-8

Figure 3-9

charge (bottom) TMPA-saturated Wyoming montrnorillonite
at six different relative hurnidities. The left side

(3600 - 2800 cm4) shows the changes in intensity of the
sorbed water 0-H stretching vibration (3400 cm'l) with
changes in relative humidity. The right side

(1800 - 1500 cm-1) shows the changes in intensity of

the sorbed water O—H bending (1630 cm4) vibration

with changes in relative humidity.

Effect of relative humidity on the center-of-mass
frequency of the methyl asymmetric (top) and methyl
symmetric (bottom) deformation vibrations of TMA
saturating normal-charge (left) and reduced-charge (right)
Wyoming montrnorillonite.

Infrared spectra of normal-charge (left) and reduced-

charge (right) TMA-saturated Wyoming montrnorillonite
showing the changes in frequency of the methyl asymmetric
and symmetric deformation vibrations with changes in _
relative humidity.

Effect of relative humidity on peak maximum frequencies
of the methyl symmetric (top), methyl asymmetric (middle)
deformations and low-frequency branch of the v19. ring
stretch, (bottom) of TMA satrnating normal-charge (left)
and reduced-charge (right) Wyoming montrnorillonite.

Figure 3-10 Infrared spectra of normal-charge (left) and reduced-

charge (right) TMPA-saturated Wyoming montrnorillonite
showing changes in frequency and intensity of the v19.
ring stretch and the methyl asymmetric and symmetric
deformation vibrations with changes in relative humidity.

Figure 3-11 Infrared spectra of normal-charge (left) and reduced-

charge (right) TMPA-saturated Wyoming montrnorillonite
showing changes in intensity of the methyl asymmetric
deformation vibration with changes in relative humidity.

59

61

62'

65

68

ChapterIV

Figure 4—1

Figure 42

Figure 4.3

Figure 4-4

Figure 4—5

Diagram of the controlled atmosphere FTIR cell used

to equilibrate arcne-saturated ’I'MA-and TMPA-saturated
clay films with water vapor and collect spectra of clay-
water ~arene complexes.

Effect of mixed benzene-water sorption on the
infrared spectrum of normal-charge (left) and
reduced-charge (right) TMA-saturated clay. The
clay spectra are drawn at the same scale, but stacked
for comparison. Band assignments for TMA are
given in Table 3 and benzene in Table 4. Peak
positions were estimated from peak maxima

Effect ofmixed deuterobenzene-water sorption on the
infrared spectrum of normal-charge (left) and reduced-
charge (right) TMA-saturated clay. The clay spectra

are drawn at the same scale, but stacked for comparison.
Band assignments for TMA are given in Table 3 and
benzene-d6 in Table 4. Peak positions were estimated
from peak center-of—mass

Effect of mixed ethylbenzene-water sorption on the
infrared spectrum of normal-charge (left) and reduced-
charge (right) TMA-saturated clay. The clay spectra
were drawn at the same scale, but stacked for
comparsion. Band assignments for TMA are given

in Table 3 and ethylbenzene in Table 4. Peak
positions were estimated from peak maxima.

Effect of mixed benzene-water sorption on the infrared
spectrum of normal-charge (left) and reduced-charge
(right) TMPA-saturated clay. The clay spectra are
drawn to the same scale, but stacked for comparison.
Band assignments for TMPA are given in Table 5.
Peak positions were estimated from peak maxima,

xiv

79

82

84

85

91

Figure 46 Effect of mixed ethylbenzene-water sorption on

Appendix A

Figure A1

Figure A2

Figure A3

the infraredspectrum of normal-charge (left) and

reduced-charge (right) TMPA-saturated clay. The
clayspectraaredrawntothesamescale,butstackedfor
comparison. Band assignments for TMPA are given

in Table 5. Peak positions were estimated from peak

maxima 93

Clay O-H stretching region of self-supporting

normal- and reduced-charge TMPA-d9—saturated
clayfilmsorientedat90°and45° withrespecttothe

IR beam. Normal-charge clay spectra were normalized

to a clay O-H stretch absorbance of 1.000 and the

reduced-charge clay spectra were normalized to a

clay O—H stretch absorbance of 0.800. 103

O-H stretching region of normal-charge and

reduced-charge TMPA-d9 montrnorillonite films

supported on AgCl disks. The overlaid spectra

show the films perpendicular to the IR beam and

tilted 30° from perpendicular. No normalization

factor was applied. 104

C-C ring-stretching region of the'I'MPA-d9-iodide

salt in a pressed KBr pellet (bottom spectra) and

normal-charge and reduced-charge TMPA-d9-

montrnorillonite films supported on AgCl disks.

The overlaid spectra show the film perpendicular

to and tilted 30° from perp.with respect to the IR

beam. 105

XV

Figure A4

FigureAS

Figure A6

Figure A7

Appendix B

Figure Bl

Figure BZ

Figure B3

Ring-C-H deformation region of the TMPA-d9
iodide salt in a pressed KBr pellet (bottom spectra)
and normal-charge and reduced-charge TMPA-d9-
montrnorillonite films supported on AgCl disks.
The overlaid spectra show the film perpendicular to
andtilted30°fromperpendicular. withrespectto
the IR beam.

C-C ring stretching region of normal-charge (top)
and reduced-charge (bottom) TMPA-montrnorillonite
films supported on AgCl disks.

X-ray diffractogrms of glycerol-solvated, Nat
saturated,normal-charge (left) and reduced-charge
(right) Wyornin g montrnorillonite. Inset numbers
indicate d(001)-d(0014).

X-ray diffractograms of normal- (left) and reduced-
(right) charge TMPA-montrnorillonite.

Effect of benzene and.methanol sorption on the infrared
spectra of reduced-charge hexadecyltrimethylamrnonium
(HDTMA) montrnorillonite. Spectra labeled A are clay
only in methanol, B 3% benzene in methanol, and C 3%
benzene in methanol on HDTMA-satmated clay.

Effect of benzene and methanol sorption on the infrared
spectra of reduced-charge tetrarnethylarrunonium (TMA)
montrnorillonite. Specua labeled A are clay only in
methanol, B 3% benzene in methanol, and C 3% benzene
in methanol on TMA-saturated clay.

Effect of benzene and methanol sorption on the infrared
spectra of reduced-charge trirnethylphenylammonium
(TMPA) montrnorillonite. Spectra labeled A are clay only
in methanol, B 3% benzene in methanol, and C 3% benzene
in rrrethanol on TMPA-saturated clay.

xvi

106

107

108

109

112

113

114

Figure B4

Figure B5

Appendix C
Figure C1
Figure C2
Figure C3
Figure C4
Figure C5
Figure C6

FrgureC7

Figure C8

Figure C9

Effect of benzene and methanol sorption on the infrared
spectra of reduced-charge Cs- montrnorillonite. Spectra
labeled A are clay only in methanol, B 3% benzene in
methanol, and C 3% benzene in methanol on Cs-
saturated clay.

Effect of benzene and methanol sorption on the infrared
spectra of reduced-charge Mg— montrnorillonite. Spectra
labeled A are clay only in methanol, B 3% benzene in
methanol, and C 3% benzene in methanol on Mg-
saturated clay.

ATR difference spectrum of 3% toluene in methanol.

ATR difference spectrum of 3% ethylbenzene in methanol.

ATR difference spectrum of 3% chlorobenzene in methanol.

ATR difference specu'um of 3% nitrobenzene in methanol.
ATR difference spectrum of 3% aniline in methanol.

ATR difference spectrum of 3% phenol in methanol.

ATR difference spectrum (methanol subtracted) of 3%
toluene in methanol on reduced-charge TMA-saturated
Wyoming montrnorillonite.

ATR difference spectrum (methanol subtracted) of 3%
toluene in methanol on reduced-charge TMPA-saturated
Wyoming montrnorillonite.

ATR difference spectrum (methanol subtracted) of 3%

toluene in methanol on reduced-charge HDTMA-
saturated Wyoming montrnorillonite.

xvii

115

116

117

118

119

120

121

122

123

124

125

Figure C10 ATR difference spectrum (methanol subtracted) of 3%
toluene in methanol on reduced-charge Cs+-saturated
Wyoming montrnorillonite.

Figure C11 A'I'R difference spectrum (methanol subtracted) of 3%
toluene in methanol on reduced-charge Mg2+-saturated
Wyoming montrnorillonite.

Figure C12 ATR difference spectrum (methanol subtracted) of 3%
ethylbenzene in methanol on reduced-charge TMA-
saturated Wyoming montrnorillonite.

Figure C13 ATR difference spectrum (methanol subuacted) of 3%
ethylbenzene in methanol on reduced-charge TMPA-
saturated Wyoming montrnorillonite.

Figure C14 ATR difference spectrum (methanol subtracted) of 3%
ethylbenzene in methanol on reduced-charge HDTMA-
saturated Wyoming montrnorillonite.

Figure C15 ATR difference spectrum (methanol subtracted) of 3%
ethylbenzene in methanol on reduced-charge Cs+-
saturated Wyoming montrnorillonite.

Figure C16 ATR difference spectrum (methanol subuacted) of 3%
ethylbenzene in methanol on reduced-charge Mg2+-
saturated Wyoming montrnorillonite.

Figure C17 ATR difference spectrum (methanol subtracted) of 3%
chlorobenzene in methanol on reduced-charge TMA-
saturated Wyoming montrnorillonite.

Figure C18 ATR difference spectrum (methanol subtracted) of 3%
chlorobenzene in methanol on reduced-charge TMPA-
saturated Wyoming montrnorillonite.

Figure C19 ATR difference spectrum (methanol subtracted) of 3%

chlorobenzene in methanol on reduced-charge I-IDTMA-

saturated Wyoming montrnorillonite.

xviii '

126

127

128

129

130

131

132

133

134

135

Figure C20 A'I'R difference specuum (methanol subtracted) of 3%
chlorobenzene in methanol on reduced-charge Cs+-
saturated Wyoming montrnorillonite.

Figure C21 ATR difference spectrum (methanol subtracted) of 3%
chlorobenzene in methanol on reduced-charge Mg“-
saturated Wyoming montrnorillonite.

Figure C22 ATR difference spectrum (methanol subtracted) of 3%
nitrobenzene in methanol on reduced-charge TMA-
saturated Wyoming montrnorillonite.

Figure C23 ATR difference spectrum (methanol subtracted) of 3%
nitrobenzene in methanol on reduced-charge TMPA-
saturated Wyoming montrnorillonite.

Figure C24 ATR difference spectrum (methanol subtracted) of 3%
nitrobenzene in methanol on reduced-charge I-ID'I'MA-
saturated Wyoming montrnorillonite.

Figure C25 ATR difference spectrum (methanol subtracted) of 3%
nitrobenzene in methanol on reduced-charge Cs+-
saturated Wyoming montrnorillonite.

Figure C26 ATR difference spectrum (methanol subtracted) of 3%
nitrobenzene in methanol on reduced-charge Mg2+-
saturated Wyoming montrnorillonite.

Figure C27 ATR difference spectrum (methanol subtracted) of 3%
aniline in methanol on reduced-charge TMA-
saturated Wyoming montrnorillonite.

Figure C28 ATR difference spectrum (methanol subtracted) of 3%
aniline in methanol on reduced-charge TMPA-
saturated Wyoming montrnorillonite.

Figure C29 ATR difference spectrum (methanol subtracted) of 3%
aniline in methanol on reduced-charge HDTMA-
saturated Wyoming monunorillonite.

136

137

138

139

140

141

142

143

144

145

Figure C30 ATR difference spectrum (methanol subtracted) of 3%
aniline in methanol on reduced-charge Cst-saturated
Wyoming montrnorillonite. 146

Figure C31 ATR difference spectrum (methanol subtracted) of 3%
aniline in methanol on reduced—charge M g2+-saturated
Wyoming montrnorillonite. 147

Figure C32 ATR difference spectrum (methanol subtracted) of 3%
phenol in methanol on reduwd-charge TMA-saturated
Wyoming montrnorillonite. 148

Figure C33 ATR difference spectrum (methanol subtracted) of 3%
phenol in methanol on reduced-charge TMPA-saturated
Wyoming montrnorillonite. 149

Figure C34 ATR difference spectrum (methanol subtracted) of 3%
phenol in methanol on reduced-charge I-ID'TMA-saturated
Wyoming montrnorillonite. 150

Figure C35 ATR difference spectrum (methanol subtracted) of 3%
phenol in methanol on reduced-charge Cs*-saturated
Wyoming montrnorillonite. 151

Figure C36 ATR difference specrrum (methanol subtracted) of 3%
phenol in methanol on reduced-charge Mg2+-saturated
Wyoming montrnorillonite. 152

Appendix D

Figure D1 Effect of water sorption on the intensity and position
of the OH out-of-plane deformation vibration of
sorbed benzene on normal-charge (left) and reduced-
charge (right) TMA-saturated clay. Spectra are drawn
to the same scale, but stacked for comparison. 153

XX

Figure D2 Effect of water sorption on the intensity and position
of the GB out-of-plane deformation vibration of
sorbed benzene on normal-charge (left) and reduced-
charge (right) TMPA-saturated clay. Spectra are drawn
to the same scale, but stacked for comparison. 154

Figure D3 Effect of vacuum treatment and exposure to saturawd
water vapor on the intensity of the OH out-of-plane
deformation vibration of sorbed benzene on normal-
charge (left) and reduced-charge (right) TMA-saturated
clay. Spectra shown are drawn at same scale, but
stacked for comparison. 155

Figure D4 Effect of vacuum treatment and exposure to saturated
water vapor on the intensity of the GB out-of-plane
deformation vibration of sorbed benzene on normal-
charge (left) and reduced-charge (right) TMPA-saturated
clay. Specua shown are drawn at same scale, but
stacked for comparison. 156

CHAPTER I

INTRODUCTION AND REVIEW OF LITERATURE

INTRODUCTION

The topic of this research is evaluation of properties of clay minerals saturated with
organic cations and mechanisms of arcne and water sorption on these clays by
spectroscopic methods. These materials have shown promise as sorbents for arenes and
other pollutant compounds (Boyd et. a1, 1991). Three major topics of research will be
studied: 1. Changes in cation orientation due to variations in clay charge density. 2.
Determining whether water preferentially sorbs on organic cation sites or on siloxane
surface sites of the clay mineral. 3. Evaluating competition between water and arcne
sorbates on sorption sites of the modified clay. Understanding interaction mechanisms of
arenes with modified clay rriineral surfaces will help us understand interaction mechanisms

between arenes and mineral surfaces in the environment.

REVIEW OF LITERATURE
I! [E E . .

Expanding 2:1 clay minerals saturated with organic cations like
trimethylphenylammonium (TMPA) have been proposed for remediating solvent waste
streams and contaminated soils (Boyd et al., 1991). Determining orientation of TMPA
phenyl groups in the interlamellar region of expanding 2:1 clay minerals may provide clues
to how solutes diffuse and sorb in the interlamellar space. Phenyl group orientation may
change depending on the charge density of the clay.

Jaynes and Boyd (1991) showed that sorption of nonpolar arenes from water onto
TMPA-saturated smectites increased as clay surface-charge density and the quantity of
adsorbed TMPA decreased As average separation between TMPA ions increased, the area
of accessible siloxane surface increased. Thus, arcne sorption increased as accessible
siloxane surface area increased. This suggested that aromatic compounds are adsorbed on
uncharged siloxane surfaces and that these surfaces are hydrophobic (J aynes and Boyd,
1991 ). ‘

An alternative to Jaynes and Boyd’s (1991) hypothesis is suggested by the carbon
content of their fully-charged Arizona montrnorillonite which showed there were 1050
mmols TMPA per kilogram of clay while the CEC was 1300 mmols kg‘l. Since they
added a 5- to 10- fold excess of TMPA, it may not be possible to fully saturate this clay
with TMPA because of the clay charge density and the cation size. Perhaps, there was
insufficient siloxane surface area to completely saturate the clay with TMPA cations. If it is
not possible to fully saturate this clay with TMPA, it might suggest that the sorption of
nonpolar arenes may also be restricted. Reducing the charge density might cause the

phenyl rings of the cations to reside farther apart without interaction with other TMPA

cations. This could permit greater sorption of arenes on the these clays regardless of
whether arenes rub on the siloxane surface or on the cation.

Determining TMPA cation arrangement in the interlayer of clays with different
charge densities would help determine the mechanism for the increase in arene sorption due
to charge reduction. Tighter cation packing of higher-charge clay cations might sterically
hinder diffusion of arenes into the interlayer and provide fewer available sorption sites for
arenes on normal-charge clays (Lee et al., 1989, 1990; Jaynes and Boyd, 1991). Less
competition for space on the siloxane surface of lower-charge smectites could allow T'MPA
phenyl groups to lie in a flatter orientation rather than being forced upright as might occur
on higher-charge smectites. A flat orientation may indicate greater spacing between
cations, which may favor diffusion and sorption in the interlayer.

The d(001) spacings observed by Jaynes and Boyd (1991) for normal-and reduced-
charge (~40% reduced) TMPA-smectite were 15.3 A and 14.4 A respectively. The shift to
14.4 A may be the result of random interstratification of different d(001) spacings
(Reynolds 1980). For example, in the higher—charge clay, phenyl groups may lie
perpendicular to the siloxane surface. Some phenyl groups in reduced-charge clay may be
oriented parallel to the siloxane surface in some of the interstratified layers, giving the
reduced spacing.

Further X-ray diffraction analysis can provide a clue to the orientation of MA
cations on expanding clays. Greene-Kelly (1955) collected X-ray powder photographs of
70 different montrnorillonite-aromatic complexes. Through Fourier uansforrns of
diffraction patterns and d-spacing measurements, Greene-Kelly (1955) proposed that 15 A
d(001) spacings of mononuclear aromatics were due to phenyl rings of the sorbates lying
perpendicular to the siloxane surface of the clay. Spacings of 12.5 A to 13 A were due to
all aromatic rings lying flat on the siloxane surface of the clay. Randomly interstratified 13

4
A and 15 A spacings due to flat and parallel orientations of the TMPA phenyl group could
give d(001) spacings between 14 and 15 A like those observed by Boyd et al. (1991).
Possible interstratified spacings for a reduced-charge smectite saturated with TMPA
(Jaynes and Boyd, 1991) might be:
1. 9.5 A for clay layers which completely collapsed with their charge reduction
treatment and were not re-expanded before adsorbing TMPA.
2. 15.3 A for clay layers where phenyl groups of the cations are oriented
perpendicular to the siloxane surface of the clay, similar to what might be the case in
normal-charge clay. I
3. All intermediate spacing of 12 A to 13 A where the change in charge density
allowed cation phenyl groups to shift to a parallel orientation with respect to the
siloxane surface.

The Fourier transform method was outlined by MacEwan (1956) for determining
components of the observed spacings of randomly interstratified mixtures of 2:1 clay
minerals. This method estimates the “..probability of finding a given layer to layer distance
in crystal space” (Reynolds (1980). For example, a randomly interstratified mica-
rnontmorillonite complex solvated with glycerol would likely give a single, broad X-ray
diffractograrn peak at 13 to 14 A. Yet, such a material would have interlayer spacings of 10
A due to mica layers and 17 A due to smectite layers in the clay mineral structure
(MacEwan 1956). Spacings due solely to mica would be visible on a Fourier transform in
10 A increments beginning at 10 A and continuing through the transform as far as it was
calculated Spacings due solely to the smectite would be visible in 17 A, increments
beginning at 17 A and continuing through the transform. The Fourier transform can also
give spacings due to mixtures of the two components. For example a peak at 27 A would
be observed for the combination of the first peak of mica and the first peak of smectite (10
A+17 A). 8

The Fourier transform can be used to verify various interlayer spacings of
interstratified unknowns and even estimate relative proportions of different interlayer
spacings (MacEwan, 1956). This method has also been used to determine orientation of
organic sorbates in the interlayer of smectite (Greene-Kelly, 1955) and vermiculite
(Serratosa, 1966) by estimating where regions of electron density due to sorbates in the
interlayer occm'.

Another tool used for determining orientation of adsorbed compounds on clays is
infrared spectroscopy. Vibrational transitions responsible for absorption of infrared
radiation occur along specific symmetry axes of molecules (Cotton, 1960). If sorbed
molecules assume a preferred orientation on an oriented clay mineral specimen, it is
possible that one of the vibrational symmetry axes of the sorbed molecules will be parallel
to incident infrared radiation. The absorbance of that vibration would be minimized and
tend to approach zero (Serratosa, 1966; Farmer and Mortland, 1966). If the clay mineral
specimen with the sorbed molecules is then tilted with respect to the infrared radiation of
the spectrometer, vibrations which were parallel to the incident radiation, now have a
greater cross section to absorb the infrared radiation, and the intensity observed for that
band will increase dramatically. This phenomenon, called infrared dichroism, has been
used by Farmer and Mortland (1966) and Serratosa (1966) to determine the orientation of
pyridine and pyridinium when adsorbed on clay minerals. Infrared band positions of
T'MPA can also be observed to determine if different interactions mm between the cation
and the clay (or other cations) due to differences in charge density. For example, if rt-rt
interactions occur between TMPA phenyl groups with high charge density, then a decrease
in charge density could cause band positions of T'MPA phenyl groups to change because of

less 1t-1t interaction.

W

Determining the preferred adsorption site for water on alkylammonium-saturated
clays is necessary for determining the effect of hydration on arcne sorption. There are two
major sites of water adsorption on expanding 2:1 clay minerals. Water molecules can form
coordination shells around the saturating cation or sorb on the siloxane surface (Sposito
and Prost, 1982). Most spectroscopic data for water sorption on clays has focused on
clays saturated with inorganic cations.

When clay is saturated with inorganic cations, the most important site for water
adsorption on clays is around cations saturating the clay mineral (Farmer and Russel, 1971;
Sposito and Prost, 1982; Sposito, 1984). Prost (1975) collected infrared spectra of water
and deuterated water vapor sorbed on montrnorillonite and hectorite films saturated with
different inorganic cations. Prost (1975) concluded that at low water contents, water
molecules were arranged in three-fold coordination around the saturating cation in a single
plane parallel to the siloxane surfaces. Other studies cited by Sposito and Prost (1982)
used nuclear magnetic resonance and electron spin resonance, in conjunction with X-ray
diffraction to verify that the structure of water around the inorganic saturating cations of
partially hydrated smectites was similar to that proposed by Prost (1975). Pinnavaia
(1980) described changes in water structure around saturating cations of clays as water
content changed from slightly hydrated to fully hydrated. When slightly hydrated, water
was arranged in planer coordination around the cation as described by Prost (1975). As
water content increased water formed a 3-layer coordination shell around the saturating
cation with water molecules above and below the cation as well as in the plane of the cation
parallel to the siloxane surface. With further increasing water content, the cation-water

coordination complex developed a tumbling motion suggesting that the siloxane surface

was no longer restricting its motion. This meant the cation-water complex was behaving
more like an ion in solution. This is important when considering the competitive sorption
of solutes and water on smectites. At low water contents, a solute can sorb on sites
unoccupied by hydrated cations. At high water contents, a solute must compete with a bulk
water-like phase occupying the clay interlayer.

Another possible site of water sorption on expanding 2:1 clay minerals is the
siloxane surface. Siloxane surface oxygens are weak, soft Lewis bases and could
potentially form hydrogen bonds with water (Sposito, 1984). Lewis basicity of the surface
oxygens increases with isomorphous substitution and is dependent on the location of the
substitution. When isomorphous substitution occurs in the tetrahedral layer (Sposito,
1984; Bleam, 1990), the negative charge is effectively distributed among three surface
oxygen atoms as opposed to ten surface oxygen atoms when substitution is in the
octahedral layer (Sposito and Prost, 1982). Lewis basicity, and thus hydrogen bonding, is
greatest when the substitution is in the tetrahedral layer. Cations balance the charge on
these sites, so any water in the vicinity of an isomorphous substitution will also be affected
by the cation.

The preferred site of water adsorption on clays saturated with quaternary
arrunonium cations has not been studied spectroscopically. Water vapor sorption isotherms
collected by Gast and Mortland (1971) showed that TMA- saturated montrnorillonite
consistently sorbed more water vapor than NI-lf- saturated montrnorillonite. Prost (1975)
collected water vapor sorption isotherms on K+-and Cs+- saturated hectorite which showed
similar amounts of water sorption to that observed by Gast and Mortland (1971) for NH;-
and TMA-saturated montrnorillonite. The shapes of these isotherms was also similar.

Since the primary site of water sorption on K"-and Cs+-saturated clays has been shown to

be coordinated around cations, this suggests that TMA cations are the primary sorption site,

rather than the siloxane surface. Sorption isotherms of arenes on TMA- and TMPA-
saturated montrnorillonites collected in the presence and absence of water have shown that
water inhibits arene sorption on these clays (Lee et al., 1990; Jaynes and Boyd, 1991).
The effect of charge density on the sorption of arenes from water provides indirect evidence
for sites of water sorption on clays. Lee et a1. (1990) stated that water sorption on either
T'MA cations or the siloxane surface might have caused the decrease in arene sorption from
aqueous solution compared with arene sorption on dry TMA-montrnorillonite. In addition,
' Lee et a1. (1990) noted greater sorption of arenes from water on low-charge TMA-clay than
higher-charge clay. In Jaynes and Boyd (1991) arcne sorption from water was compared
between reduced- and normal-charge Arizona montrnorillonite saturated with T'MPA to
determine whether aromatic hydrocarbons preferentially sorbed on the siloxane surface or
the cations. Sorption of arenes from water increased as the layer charge of the clay was
reduced, implying that arcne sorption was more favored than water adsorption on the
siloxane surface and that the siloxane surface was more hydrophobic than the cations. This
result was in agreement with the observations of Lee et a1. (1990) noting the effect of
charge on arcne sorption. These results suggested that the cations are the preferred
sorption site for water since the siloxane surface was more hydrophobic than the cations.
Although circumstantial evidence is rather strong that quaternary ammonium cation sites are
preferred for water sorption on TMA- and TMPA-saturated clay, spectroscopic evidence
will help to verify whether water sorbs preferentially on the siloxane surface or around
cations on these clays.

C .. 'f l

Sorption of arenes onto clays can be divided into two broad categories: simple
physical sorption and sorption involving strong chemical interactions of the 1t-electrons of

the aromatic ring with some substituent of the sorbent. An example of a strong interaction

9

was observed by Mortland and Pinnavaia (1971) where a red complex formed between
benzene and Cuztsaturated montrnorillonite when the clay was dry. Using infrared
spectroscopy, they determined that the red color formed because ctr2+ ions saturating the
clay withdrew rt-electrons from benzene. Similar complexes can form between Cu”-
saturated montrnorillonite and toluene, and ortho, meta, and para-xylene (Pinnavaia and
Mortland, 1971; Johnston et al., 1992). Clementz and Mortland (1972) also observed
formation of ligand complexes between Ag1+ saturated montrnorillonite and benzene,
toluene, and ortho-, meta-, and para-xylene. In all studies, physical sorption of nonpolar
arenes occurred on clays when ligand or charge-transfer complexes formed. Johnston et
a1. (1992) found spectroscopic evidence for physical sorption, but no ligand complexes
formed between p-xylene and (302+. saturated montrnorillonite. The Co2+ did not
withdraw it-electron density from the p-xylene as did copper. This result could be
predicted from the electronic structure of the ions. Cuz" tends to withdraw electron density
from the rt-cloud of benzene because it only needs one electron to fill its d orbital, while
Co2+ needs 3 electrons (Shriver, Atkins, and Langford, 1990).

The ligand and charge transfer complexes of benzene and alkylbenzenes on
smectites saturated with transition metal ions studied by Mortland and Pinnavaia (1971),
Pinnavaia and Mortland, ( 1971), Clementz and Mortland (1972) and Johnston et a1. ( 1992)
were destroyed by sorbed water. When clay films were exposed to water vapor, color
created by ligand complexes immediately disappeared Infrared spectroscopy revealed only
physical sorption of benzene or alkylbenzene on the clays. In all of these studies, the
cations on the clay surface were being hydrated by the water, destroying the complex
formed between the rt-cloud of the aromatic and the cation. The only sites remaining for

benzene and alkylbenzenes to sorb on these clays were on the siloxane surface.

10

The competitive sorption of arenes and water on clays saturated with alkali and
alkaline earth cations has been studied with vapor sorption isotherms. Rhue et a1. (1989)
collected ethylbenzene vapor sorption isotherms on a (Ca2+, Na1 +) saturated bentonite with
partial water vapor pressures ranging from 0.00 to 0.56. The amount of ethylbenzene
sorbed was the same at P/P0 of 0.00 as it was at 0.20. At 0.50, ethylbenzene sorption was
substantially reduced. The shape of the sorption isotherms also changed at higher relative
humidities. Isotherms at low relative humidities were S-shaped, while at higher relative
humidities isotherms were linear. The authors suggested that a change in the sorption
mechanism of ethylbenzene occurred as clay-water content increased, but the mechanism
could not be deduced from the isotherms alone. Rhue et al. (1989) theorized about
possible sorption mechanisms of the ethylbenzene on the hydrated surface of the bentonite,
but did not state whether the water was hydrating cations or forming H-bonds with the
siloxane surfaceRhue et al. (1989) felt that competition between water vapor and
ethylbenzene at higher water vapor pressures caused the decrease in ethylbenzene sorption
and changes in isotherm shape. Infrared spectroscopic data from Johnston et a1. (1992)
would suggest that Rhue et al. (1989) observed simple physical sorption of ethylbenzene
on the siloxane surface of the clay whichdecreased as water hydrating the cations saturating
the clay occupied more siloxane surface.

Clays saturated with quaternary ammonium cations may sorb arenes physically on
the siloxane surface and/or at the cation when dehydrated. Barrer and Perry (1961) used
X-ray diffraction to measure the d(001) spacings of a dry bentonite saturated with TMA
after sorbing benzene vapors at P/Po between 0.05 and 0.8. The spacing increased from
13:6 A to 14.6 A at P/Po = 0.05 and remained at that spacing for all other treatments.
Using estimated interlayer surface area, and the clay cation exchange capacity, Barret and

Perry (1961) concluded that the most likely orientation of the benzene in the interlayers was

11

tilted between T'MA cations. Perhaps the benzene-accessible siloxane surface area was
restricted by TMA cations forcing benzene to sorb on cation sites in a tilted orientation
between cations. Lee et a1. (1990) noted that sorption of vapor-phase benzene was greater
for dry T'MA- satlnated high-charge Arizona montrnorillonite than it was for lower-charge
TMA saturated Wyoming montrnorillonite at partial pressures of benzene greater than 0.6.
The Arizona clay may have had greater sorption because it had more cation sites.
Understanding the mechanism by which water reduces arcne sorption on quaternary
ammonium-saturated clays is important to understanding the sorptive properties of these
clays. Lee et a1. (1990) theorized that the decreased amount of aqueous benzene, toluene,
and o-xylene sorption on TMA-saturated high-charge Arizona montrnorillonite relative to
low-charge Wyoming montrnorillonite was because of a water-induced sieving effect in
high-charge clay. They stated that hydration of mineral surfaces or saturating cations
reduced accessibility of the interlayer to arenes. Lee et al. (1990) also observed that arcne
sorption on the clay from water decreased as the number of methyl groups on the arene
were increased. This was attributed to steric hinderance to diffusion of larger arenes into
the interlayer of clay. In further work, Jaynes and Boyd (1991) collected sorption
isotherms of arenes from water on TMPA clays with varied charge densities and found that
arcne sorption increased as charge density decreased. This finding indicated that the
siloxane surface of the clay was the probable sorption site for arenes and the cations
probably functioned as a pillars to hold the clay layers open. J aynes and Boyd (1991)
further concluded that “...a large. part of the siloxane surface in smectites has a hydrophobic
nature. A conclusion that can be drawn from this work is that since the siloxane surface of
the clay was hydrophobic, the TMPA cations were hydrophillic and were the probable site

for water sorption on these clays.

12

The basic mechanisms of arcne sorption and water inhibition of sorption on TMA-
and TMPA-saturated clay of different charge densities may be investigated using
specu'oscopic techniques. Clay films with different charge densities saturated with TMA
and TMPA can be prepared. These films can be treated with differen: combinations of
arene and water vapor and probed using infrared spectroscopy to determine the effect of
charge density, water sorption, and arcne sorption on the position of cation and arcne
vibrational frequencies. From this information, the preferred sorption sites for water and
arenes on TMA- and TMPA-saturated clay may be determined.

REFERENCES

Barrer, RM. and G.S. Perry. 1961. Sorption of mixtures and selectivity in
alkylarnmonium rnontrnorillonites. Part II. Tetramethylarrunonium bentonite. J.
Chem. Soc. 57:849-859.

Bleam, WF. 1990. The nature of cation substitution sites in phyllosilicates. Clay and V
Clay Minerals 38:527-536.

Boyd, S. A., W. F. Jaynes, and B. S. Ross. 1991. p. 181-372 (ed.) Organic
substances and sediments in water vol. 1. CRC Press, Boca Raton, FL

Clementz, D.M., and MM. Mortland 1972. Interlamellar metal complexes in layer
silicates III Silver(I)-Arene complexes in smectites. Clays and Clay Minerals
20: 181-187.

Cotton, F.A., 1960. The infrared spectra of transitional metal complexes. pp. 305-315. In
J. Lewis and R.G. Wilkins (ed.) Modern Coordination Chemistry. Interscience
Publishers Inc. New York.

Farmer, V. C., and M. M. Mortland. 1966. All Infrared Study of the Co-ordination of
Pyridine and Water to Exchangeable Cations in Montrnorillonite and Saponite. J.
Chem Soc. 344-351.

Farmer, V.C., and J .D. Russell. 1971. Interlayer complexes in layer silicates--The
structure of water in lamellar ionic solutions. Trans Farad. Soc. 67:2737-2749.

13

Gast, R.G., and M.M. Mortland. 1971. Self—diffusion of alkylarnmonium ions in
montrnorillonite. J. Coll. Int. Sci. 37:80-92.

Greene-Kelly, R. 1955. Sorption of Aromatic Organic Compounds by Montmorillonite
Pt. 1.—Orientation Studies. Trans. Faraday Soc. 51:412-424.

Jaynes, W.F., and S.A. Boyd. 1991. Hydrophobicity of siloxane surfaces in smectites as
revealed by aromatic hydrocarbon adsorption from water. Clays and Clay Minerals
39:428-436.

Johnston C.T., T. Tiupton, S.L. Trabue, C. Erickson, and DA. Stone. 1992. Vapor-
Phase Sorption of p-Xylene on Co- and Cu-Exchanged SAz-l Montmorillonite.
Env. Sci. Tech. 26:382-390.

Lee, J .-F., M.M. Mortland, S.A. Boyd and CT. Chiou. 1989. Shape-selective
adsorption of aromatic compounds from water by tetramethylammonium-smectite:
J. Chem Soc Faraday Trans. I 85: 2953-2962.

Lee, J.-F., M.M. Mortland, C.T. Chiou, D.E. Kile, and S.A. Boyd. 1990. Adsorption
of benzene, toluene, and xylene by two tetramethylammonium—smectites having
different charge densities. Clays Clay Minerals. 38: 1 13-120/

MacEwan, D.M.C., 1956. Fourier Transform Methods for Studying Scattering from
Lamellar Systems I. A Direct Method for Analysing Interstratified Mixtures.
Kolloid Zeitschn'ft 149:96-108.

Mortland, M.M. and T.J. Pinnavaia. 1971. Formation of copper (II) arcne complexes on
the interlamellar surfaces of montrnorillonite. Nature Physical Science 229:75-77.

Pinnavaia, T.J. 1980. Advanced Chemical methods for soil and Clay Minerals Research
Ch. 8. J.W Stucki and W.F. Banwart (Ed.) D. Reidel: Boston.

Pinnavaia, T.J., and M.M. Mortland. 1971. Interlamellar metal complexes on layer
silicates. I. Copper(ID-arene complexes on montrnorillonite. J. Phys Chem.
75:3957-3962.

Prost, R. 1975. Etude de l’hydration des argiles: Interactions can-mineral ct mécanisme de
la retention de L’eau II. Etude d’une smectite (Hectorite). Ann. Agron 26:463.

14

Reynolds, R.C. 1980. Interstratified Clay Minerals. in Mineralogical Society Monograph
no. 5. Crystal Structures of Clay minerals and their X-ray identification. G.W.
Brindley and G. Brown eds. Mineralogical Society 41 Queen’s Gate London SW7
5HR.

Rhue, R.D., K.D. Pennell, P.S.C. Rao, and W.H. Reve. 1989. Competitive adsorption
of alkylbenzene and water vapors on predominantly mineral surfaces.
Chemosphere 18:1971-1986.

Serratosa, J .M. 1966. Infrared Analysis of the orientation of Pyridine molecules in clay
minerals. Clays and Clay Minerals 14:385-391.

Shriver, D.F., P.W. Atkins, and CH. Langford. 1990. Inorganic Chemistry. W.H.
Freeman and Co. New York.

Sposito, G., and R. Prost. 1982. Structure of water adsorbed on smectites. Chemical
Reviews 82:553-573.

Sposito, G. 1984. The surface chemistry of soils. Oxford University Press. New York,
234pp.

CHAPTER II

Orientation of the Phenyl Group of Trimethylphenylammonium on
Normal- and Reduced-charge Wyoming montrnorillonite

ABSTRACT

The orientation of trirnethylphenylammonium (TMPA) cations adsorbed on
montrnorillonite affects the adsorbate-accessible siloxane surface and determines whether
the TMPA phenyl ring can interact with other aromatic adsorbates by rt—rt interactions. The
purpose of this study was to determine the orientation of TMPA ions in the interlayer of
normal-charge and reduced-charge Wyoming montrnorillonite. The orientation of TMPA's
phenyl group was investigated using infrared dichroism of selected aromatic ring
vibrations. X-ray diffraction and one-dimensional Fourier analysis were used to determine
interlayer spacings and the interlayer electron distribution of TMPA and to ascertain
whether reduced-charge Wyoming montrnorillonite is a randomly interstratified mixture of
layers with two different d-spacings. For normal-charge montrnorillonite, the infrared
results showed that the C-N bond axis is neither perpendicular nor parallel to the surface,
yet X-ray data suggested that the TMPA phenyl ring is perpendicular to the siloxane
surface. In this orientation, the average adsorbate-accessible sin-face area is 50 Azlcation,
which is consistent with N2 BET surface-area measurements. Reduced-charge
montrnorillonite is a randomly interstratified mixture of 25% collapsed layers with no
adsorbed cations and 75% expanded layers that are propped open by TMPA's methyl
groups, not the aromatic ring. The adsorbate-accessible surface area on expanded layers of

reduwd-charge montrnorillonite is about twice that on normal-charge TMPA-clay. When

15

16

the phenyl ring of TMPA is perpendicular to the surface, it should be possible for polar
compounds such as water to interact with the positively charged nitrogen atom while

aromatic compounds interact with the phenyl ring by 1M: interactions.

INTRODUCTION

Smectite and other expanding 2:1 clay minerals sorb aromatic hydrocarbons from
water when small quaternary alkylamrnonium cations such as trimethylphenylammonium
(TMPA) occupy the clay's cation exchange sites (Boyd et al., 1991 and references cited
therein). Information about the orientation and distribution of organocations within clay
interlayers will aid in developing suitable organoclay sorbents for a particular waste stream
because these properties affect sorption capacity and sorption mechanisms. For example,
the surface-charge density of a clay may affect the orientation of an adsorbed organic
cation's aromatic ring (Serratosa, 1966), which in turn may affect the interlayer surface
area accessible to uncharged organic sorbates such as aromatic pollutants. The orientation
of T'MPA phenyl rings may also determine whether TMPA's rt-electrons can interact with
other aromatic sorbates by 1M: interactions.

Infrared dichroism experiments can be used to determine the orientation of the
aromatic ring and the ON bond axis of TMPA. When an infrared beam is parallel to an
axis of molecular vibration, the absorption intensity of that vibration will be minimized. If
the sample is rotated so that the infrared beam is no longer parallel to the infrared vibration,
the absorption intensity will increase. Thus, any preferred orientation of molecules
adsorbed on siloxane surfaces will cause the intensity of one or more vibrational bands to

change when an oriented clay film is rotated in an infrared beam For example, infrared

dicluoism experiments have shown that uncharged polar aromatic molecules like pyridine

17

(Farmer and Mortland, 1966; Serratosa, 1966) and benzonitiile (Serratosa, 1968) lie
perpendicular to the siloxane surface of montrnorillonite, with the polar group midway
between opposing clay surfaces. This orientation allows the polar group to either solvate
adsorbed inorganic cations directly or to interact with water molecules in the hydration shell
of the inorganic cation. In contrast, when pyridinium ions satisfy a clay‘s negative charge,
infrared dicluoism experiments have shown that the pyridinium ring lies parallel to
montrnorillonite but perpendicular to vemiiculite surfaces (Serratosa, 1966). The
difference in orientation was attributed to the higher charge density of vermiculite: the
surface area per negative charge on Wyoming montrnorillonite can accommodate
pyridinium ions lying flat on the clay surface, whereas pyridinium can only satisfy the
structural charge of vermiculite in an upright position (Serratosa, 1966). In either case, the
orientation of the pyridinium ring minimizes charge separation between the NI-l+ group of
pyridinium and the structural charge of the clay. The orientation of adsorbed TMPA
cations should also minimize charge separation within steric constraints imposed by the size
of the TMPA ion and the clay's surface-charge density.

X-ray diffraction, particularly Fourier transform analysis of diffraction data, can
provide complementary information about clay structure and adsorbate orientation. Fourier
analysis of X-ray diffractograms can be used to verify the existence of random
interstratification, to determine d-spacings of individual components in an interstratified
mixture, and to estimate relative proportions of each component in an interstratified mixture
(MacEwan, 1956; Reynolds, 1980). Fourier transforms of X-ray diffraction data can also
be used to determine the orientation of organic adsorbates (e.g., Greene-Kelly, 1955;

Bradley et al., 1963: Johns and Sen Gupta, 1967). One-dimensional Fourier analysis of

nitrobenzene and pyridine complexes with montrnorillonite (Greene-Kelly, 1955) gave

18

interlayer electron density distributions that are completely consistent with infrared
dichroism results for polar aromatic compounds (Serratosa, 1966; 1968).

The objective of this research was to determine the orientation of TMPA cations on
normal-charge and reduced-charge Wyoming montrnorillonites. Specifically, infrared
dichroism experiments were used to test the hypothesis that TMPA phenyl groups are
perpendicular to the surface of normal-charge montrnorillonite and parallel to the surface of
reduced-charge montrnorillonite. Infrared dichroism will not be observed if the principal
vibration axes of TMPA are randomly oriented or if the preferred orientation is such that
none of the principal vibration axes is perpendicular to the surface. Additionally, little or
no dichroism will be observed if a clay comprises interstratified layers with different TMPA
orientations. One-dimensional Fourier analysis of X-ray diffraction data was used to test
the hypothesis that reduced-charge clay comprises a randomly interstratified mixture of

layers with different TMPA orientations and hence with different d-spacings.

MATERIALS AND METHODS

 

Wyoming montrnorillonite (SWy-l) was obtained from the Clay Minerals Society
Source Clays Repository at the University of Missouri-Columbia. The <2-llm fraction was
separated from the coarser material by sedimentation. Half of the <2-tlm clay was
saturated with Na"' and the other half with Li“ by shaking 4 g of clay with 200 ml of the
appropriate 0.1 M chloride solution, centrifuging the suspensions, and decanting the
supernatant solutions. This shaking-centrifugation-decantation process was repeated three
times. Excess salts were removed from the homoionic clays by dialysis until an AgNO3
test for chloride was negative. Part of the dialyzed Na+-saturated clay suspension

(“normal-charge” clay) was freeze-dried and set aside.

19

The layer charge of the remaining clay was reduwd by a method similarto that
described by Brindley and Ertem (1971). Briefly, a suspension composed of 50% Li"-
saturated and 50% Na+-saturated clay by mass was freeze-dried, then heated 18 h in quartz
crucibles at 250 °C to dehydrate adsorbed Li and promote Li migration into vacant
octahedral sites. The collapsed, redmd-charge clay was re-expanded by sonicating 2- g
subsamples of clay in 200 ml of a 70% methanol-water mixture for 20 min in an ice-water
bath (Greene-Kelly, 1953; Brindley and Ertem, 1971; Jaynes and Bigham, 1987). A I-Ieat
Systems, Ultrasonic Inc. model W-385 sonicator with a 7-mm diameter probe tip on a
setting of 7 (out of 10) was used.

To prepare TMPA-saturated montrnorillonites for the X-ray diffraction studies
described below, TMPA-Br was added to both normal- and reduced-charge
montrnorillonites in 70% methanol-water suspensions (10 g clay I") and stirred for 72 h.
The amount of MA added was ten times the CEC. After the suspensions were stirred,
clays were dialyzed until bromide-free and then were freeze—dried. Selected properties of
normal- and reduced-charge TMPA-clays are reported in Table 2-1.

I E 1 1 I .
' n h A

As will be discussed later, accurate assignment and identification of the v19. and
Vlgb infrared bands of TMPA is essential for using infrared dichroism to determine the
orientation of adsorbed TMPA. Methyl C-I-l deformation vibrations of TMPA overlap at
least partially with the v19, and Vlgb ring stretching vibrations, but the ring vibrations are
distinct in methyl-deuterated TMPA. Thus, methyl-deuterated TMPA (TMPA‘dg) was
prepared, and montrnorillonites saturated with TMPA-dg were used in the infrared

dichroism experiments described below.

20

Table 2-1. Selected physical and chemical properties of normal- and reduced-charge
TMPA- montrnorillonite.

 

d(001)1 N2 Surface Area2 Total C3 Adsorbed TMPA“ cr-zcs

 

Treatment (A) (m2 g‘l) (g kg‘l) (mmolkg'1) (mmol kg“)
Reduced 13.9 318:1:14 48.6103 449:1:8 39(330
Normal 14.9 252118 88. 111:1.0 814:1:1 1 870180

 

1 = Obtained from Fourier transform of XRD data.

2 = Avg 3: SD (N=2 or 3) from three-point BET N2 adsorption isotherms using a
Quantachrome Quantasorb Jr. surface area analyzer.

3 = Avg :1: SD (N=2) of total C determined by combustion at 900 °C using a Dohrmann
DC-190 high-temperature carbon analyzer.

4 = Avg :1: SD (N=2) calculated with the equation:

TotalgC mmolTMPA

 

5 = Avg :1: SD (N=3 or 4) Determined by Na+ saturation and ammonium displacement of
the parent normal- and reduced-charge clays. Statistically identical values were
obtained with Mg2+ saturation and Baz" extration.

21

Methyl-deuterated TMPA (TMPA-d9) was synthesized using a procedure modified
from that of Cope et a1. (1960). Aniline, NaHCO3, and CD31 in a mole ratio of 1:3:3 were
refluxed in methanol [10:1 methanol:aniline (v:v)] with constant stirring for 75 h.
Deuterated methyl iodide was added after 24 and 48 h to give a final CD31:aniline mole ratio
of 4.5:1. After the mixture was refluxed, it was evaporated to dryness in the reflux flask.
The residual solid was extracted three times with boiling chloroform. The volume of
chloroform for each extraction was equal to the volume of methanol used during reflux.
The hot chloroform extracts, which contained the deuterated TMPA-iodide, were decanted
from the solid material remaining in the flask and were filtered. Much of the deuterated
TMPA-iodide crystallized on the filter because the boiling chloroform cooled rapidly at
room temperature. Deuteration of the TMPA methyl groups was verified by obtaining the
FTIR spectrum of the white crystalline product in a pressed KBr pellet.

9131mm.

Oriented, self-supporting TMPA-dg-saturated clay films were prepared by
sedimentin g N a+-saturated normal-charge and reduced-charge montrnorillonite suspensions
onto glass slides. The Na-montmorillonite films were air-dried at room temperature and
then were reacted overnight at 60 ’C with 20 ml of TMPA-do in ethanol (5 mmol TMPA-d9
kg‘l solution) in covered Petri dishes. During this reaction, the clay films partially
separated from the glass slides. Next, the ethanol was allowed to evaporate, and the dry
films were removed from the slides and washed gently with 20 ml of ethanol to remove
excess salts. The ethanol wash solutions were removed with a Pasteur pipette, and the clay
was allowed to dry. The normal-charge film contained 4.6 mg clay cm'z, and the reduced-

charge film contained 2.2 mg clay cm'z.

 

 

Wanna.

Self-supporting TMPA-do montrnorillonite films were placed in the sample holder
of a Perkin-Elmer 1710 FTIR that was purged with high-purity N2. Single-beam
background (empty, purged sample compartment) and clay spectra were collected with a
DT’GS detector using 2 cm'1 resolution, no apodisation, and 100 scans. After spectra were
collected with the montrnorillonite films normal to the beam, a second spectrum ofeach
film was collected with the film tilted approximately 45‘. Spectra were stored on diskette
and converted from single-beam spectra to absorbance using the MS-DOS program
SpectraCalc (Galactic Software, Inc.). Baselines were leveled in spectral regions of
interest and zeroed. To correct for small differences (<10%) in absorbance that might be
caused by differences in the amount of clay probed by the infrared beam with the clay films
90' and 45' to the infrared beam, the spectra were normalized to make the absorbance of
the clay lattice O-I-I stretching band (3628 cm'l) equal at the two angles of incidence. To
determine whether the TMPA phenyl group adopts a preferred orientation with respect to
the siloxane surface in either the normal-charge or reduwd-charge montrnorillonite, the
normalized 90’ absorbance of each TMPA ring vibration was compared with the
normalized absorbance at 45’.

Fourier analysis of X-ray diffraction data for glycerol-solvated, normal-charge and
reduced-charged Na-montmorillonites was used to detemiine whether some layers in the
reduced-charge clay were collapsed and hence inaccessible to cations. Aqueous
suspensions of Na-montmorillonite suspensions were sedimented on X-ray slides, allowed
to dry overnight, then sprayed with a 10% aqueous glycerol solution. The slides then were
placed in a desiccator over CaClz to dry overnight. The slides were X-rayed from 3 to 90°

 

 

 

23
26 in 50-s, 0.05° steps with Cu-Ka radiation and a Ni filter at 10 mA and 25 kV using a
Phillips APD3270 diffractometer equipped with a monochrometer and theta-compensating
slit.

X-ray diffractograms were also collected for TMPA-saturated normal- and reduced-
charged montrnorillonite. Freeze-dried TMPA-saturated montrnorillonites were
resuspended in methanol, and the suspensions were sedimented onto glass slides. Slides
of normal-charge clay were X-rayed at 20 mA and 35 kV from 3 to 60° 26 in 4—s, 0.05°
steps and from 60 to 85° 26 in 10-s, 0.05° steps. Slides of reduced-charge montrnorillonite
were X-rayed in 50-s, 0.05° steps from 3 to 90° 26.

Fourier transforms of the X-ray diffraction peaks were calculated to determine the
amount of interstratified, collapsed material in the Na+-saturated clays and to determine
TMPA cation orientation in the interlayer of normal-charge and reduced-charge Wyoming .
montrnorillonites. The procedure outlined by MacEwan (1956) was programmed into a
computer spreadsheet. Form factors were taken from Figure 8-2 in MacEwan (1956).
Some form factors approached zero, resulting in amplitude factors that approached infinity.
As recommended by Reynolds (1980), X-ray peaks that caused the amplitude factor to
approach infinity were not used to calculate the Fourier transform. The Lorentz factor was
taken as a constant, since the diffractometer has a theta-compensating slit that causes the X-
ray beam to irradiate a constant volume of sample. The polarization factors were calculated
to account for use of the monochrometer as described by Azaroff (1955). The Fourier
transforms were calculated every 0.05 A from 0 to 45 A. Terms used for calculating the

Fourier transforms for Na+- and TMPA-saturated clays are shown in Tables 2 and 3

respectively.

 

 

24

Table 2-2. Terms used for calculating the Fourier transforms of X-
ray difl'iactograms of glycolated, normal- and reduced-charged,
Na+-montmorillonite.

Mil—WWIJQLZ_

 

 

 

 

Normal-Charge
1 19848 17.84 100 1058
2 2226 8.97 7 504
3 926 5.96 12 405
4 1947 4.48 10 689
5 2160 3.58 13 828
6 1655 2.99 32 808
7 265 2.56 10 353
8 428 2.23 1 474
9 441 1.99 6 498
10 156 1.79 4 303
11 67 1.69 1 204
12 376 1.45 7 514
13 213 1.38 12 403
14 97 1.29 10 283
15 64 1.20 3 237
Reduced-Charge
1 14339 17.84 60 490
2 10374 9.06 7 1226
3 1282 5.88 20 257
4 8131 4.54 3 1691
5 4336 3.57 18 512
6 6817 3.02 38 448
7 366 2.58 15 168
8 815 2.25 0.5 1387
9 1091 1.98 5 512
10 496 1.81 4 387
12 199 1.50 9 160
13 397 1.37 13 185
14 126 1.29 9 123
15 82 1.20 0.5 411

 

1 Estimated from Figure 8-2 of MacEwan (1956)
2 Fourier amplitude terms as calculated by MacEwan (1956)

 

 

25

Table 2-3. Terms used for calculating the Fourier transforms of X-
ray diffractograms of normal- and reduced-charged TMPA-
montrnorillonite.

WWI—Anz—

 

 

 

 

Normal-Charge
1 25619 14.84 38 823.29
3 2401 4.99 20 354.39
4 622 3.76 10 259.08
5 2685 3.01 38 281.01
6 75 2.54 0.5 416.42
8 335 1.89 4 317.82
9 76 1.67 0.4 477.48
10 140 1.49 8 142.69
11 130 1.37 15 98.39
12 15 1.25 6 51.36
13 42 1.15 0.4 297.56
Reduced-Charge
1 125400 13.57 42 267.48
2 893 6.72 8 119.64
3 8773 4.66 20 151.95
4 9982 3.35 22 150.45
5 212 2.75 37 13.27
7 1536 1.93 6 226.71
9 283 1.54 5 115.38
10 100 1.35 15 22.15
11 33 1.21 4 46.17

 

1 Estimated from Figure 8-2 of MacEwan (1956)
2 Fourier amplitude terms as calculated by MacEwan (1956)

 

26

RESULTS

Wain

To determine whether the TMPA phenyl ring is perpendicular or parallel to the
siloxane surface, the normalized absorbances of the v19” Vlgb, C-H out-of-plane, and ring
deformation bandsofTMPA-d9 withaclayfilmperpendiculartotheinfraredbeamcanbe
comparedtothenormalizedabsorbancesfortheclayfilmtilted45’ withrespecttothe
infrared beam. The vibration directions of the V“... vlgb, C-I-I out-of-plane, and ring
deformation vibrations of the TMPA phenyl ring are shown in the inset of Figure 2-1. If
the x-axis of the phenyl ring (Vlgb vibration) is perpendicular to the siloxane surface, the
absorbance of the Vlgb vibration should increase by as much as a factor of two (Serratosa,
1966) when a TMPA-dg-montmorillonite film perpendicular to the infrared beam is rotated
45’. The v19. absorbance should exhibit dichroism if the phenyl ring and z-axis (C-N
bond axis) are perpendicular to the infrared beam, whereas the GB out-of-plane and ring
deformation vibrations should be dichroic if the phenyl ring is parallel to the siloxane
surface.

For TMPA'dg adsorbed on normal-charge montrnorillonite, the normalized
absorbance of all of the TMPA ring vibrations increased slightly when the clay film was
rotated in the infrared beam. but none of the vibrations exhibited significant dichroism
(Figure 2-1a). The phenyl ring itself may be perpendicular to the siloxane surface even
though neither the x-axis nor z-axis is perpendicular, but infrared dichroism alone cannot
show the orientation of the phenyl ring when neither the x-, y-, nor z-axis is perpendicular.

For reduced-charge montrnorillonite, none of the TMPA ring vibrations showed
significant dichroism. Thus, none of the axes is perpendicular to the clay surface, and the
phenyl ring is not parallel to the siloxane surface. The original hypothesis that the phenyl

 

27

.823 805 wee—u Seep—e? ee.<e2h

Bu 8:08..»me page 05 new <m2h we mesa E0853 05 .«0 22.55.90 05 «30.7. Ewe 8 .35 482.00% 89.08

3:8 as. a e Becca See 3529 ecu as success as house A - - -v .3. as j .8 us 88:65.88
meme—o»? omafiéooeoou o5 owns—0-38.5: :0 Becomes ESE. Co 2285? Eng pace—0m .TN 0.5mm"—

 

     
 

  
    

 

...83 .353263
one 0mm 85 OR 09: omvp com _. omn—
lELlFLlFLLl—I‘TPELEFF. to-
e28. sodas: H
h
r o
Dmp> n
to
so as. _._-0 see. A
r Nd
$-80 mac 5 so; u
.1 no
$.80 8v: 8; .
.a.‘ Fae... h
.at‘.‘a .46
$-50 ~59 seam—520.0 osm ‘ W .
A 7&0 Not 26.30-50 1-0 > ..m o
m0m<=0tom030mc .m .
o6

 

 

 

 

 

 

 

 

$.50. bones—.033
8m 80 8a 289. 8: 8m. 8m.
.3-
eo_oo_ao-<a_2e ”
c 1‘ mo
1. .. to
__ 8 > .
.. .. ..N.O
... 9.0-0 ..m.o
.8 as. :6 n
“v.0
am; “mo
0052022252 .< h

 

 

 

eaueqlosqv

 

 

28

ring of TMPA is parallel to the surface of reduced-charge Wyoming montrnorillonite is
therefore false.
X- 1 E . l .

One-dimensional Fourier analysis of X-ray diffraction data for normal-charge,
glycerol-solvated Na-montmorillonite (Table 2-2) showed that this clay contains only
expanded layers with a d-spacing of about 18 A (Figure 2-2a). In contrast, the reduced-
charge montrnorillonite (Figure 2-2b) comprises a randomly interstratified mixture of 18-A
expanded layers and 9.25-A collapsed layers. Previous research (Clementz et al., 1974;
Clementz and Mortland, 1974) has shown that reduced-charge montrnorillonites with
greater than 50% charge reduction are interstratified mixtures of layers with two different
d-spacings. This conclusion was based on the irrationality of the X-ray diffraction peaks;
these authors did not calculate Fourier transforms to determine the primary spacings of the
collapsed and expanded layers (Clementz et al., 1974; Clementz and Mortland, 1974).
Clementz and Mortland (1974) proposed that collapsed layers occur because their structm'al
charge is reduced completely and no cations are adsorbed in these layers (which requires
that the collapsed layers have no tetrahedral charge).

Fourier transform peaks caused by combinations of expanded and collapsed layers
in glycerated reduced-charge Na-montmorillonite are labeled in Figure 2-2b. Based on the
amplitudes of the 9.25 A and 18 A peaks, the reduced-charge montrnorillonite contains
about 2/3 expanded layers and about 1/3 collapsed layers, assuming that the 18-A peak
from expanded layers does not contain a significant contribution at about 18.5 A fi‘om a
collapsed-collapsed layer sequence. For comparison, the proportion of expanded layers

also was estimated from the primary diffraction data (Table 2-2) using Mering's method

29

 

deco—€25.58 0w§=063=o2 .8.“ “cc—Ba. 2a «00.?—
AUV comes—e0 05 Amy Boga—x0 .«e $333500 2 030 “each dunno—Egg:—
éz 3828-3303» .8 8% 5:03am. apex we men—8mg com—now N-" 8:me

 

 

 

 

 

 

 

 

 

 

 

 

.3 8:220 2. 8:220
Mun-nor... mum—mp.”ppm-O—nnupr-nc—W-phmb-ppm men-pmV-pm-npm-nnw-n-nm :: -- pub-mbzm
Dost OOON-
ooew- ooe.-
o _ o m
a. .. v _ m
m 8 me we. 89 u
.v c. m 83 a
s m
n ooom w.
z c. .roooe m
m 0 n 88 m.
. r. m
0m . 88 a
. . 89. .u
e 8 8.: (
mm 3: n. 2.8
m a ooom Doom
mem<zodm0=amm .m . 352052252 .<

 

 

 

 

 

coco _. coco

 

30

(Reynolds, 1980): diffraction peaks at 9.06, 4.54, and 3.03 A are produced by a mixture
that comprises 65-75% expanded (18 A) layers and 25 to 35% collapsed (~93 A) layers

The Fourier transform for normal-charge TWA-montrnorillonite (Figure 2-3a)
shows that the primary interlayer distance is 15.0 A. the smaller peaks in the Fourier
transform are periodic with respect to the octahedral cations at o, 15, 30, and 45 A. The
peaks closest to the 0A and 15-A peaks for the octahedral cations are caused by planes of
0, OH, and Si atoms (Bradley, 1963). The peaks from 4.8 to 10.5 A and from 19.7 to
23.1 A can be attributed to interlayer electron density of TMPA. The interlayer peaks are
most periodic between 0 and 15 A and slightly less periodic between 15 and 30 A. At
distances greater than 30 A, resolution is limited by the poor crystallinity of
' montrnorillonite and the number of diffraction orders that can be observed (Brindley and
Hoffmann, 1962). The TMPA orientation consistent with the interlayer electron density
peaks will be described in the Discussion section.

The Fornier transform for the reduced-charge montrnorillonite (Figure 2-3b) is
dominated by a peak at 13.6 A from layers expanded by adsorbed TMPA. There also is a
smaller peak at 9.5 A from collapsed layers with no adsorbed TMPA ions, in addition to
peaks at greater distances that correspond to various combinations of expanded and
collapsed layers. Thus, reduced-charge TMPA-montrnorillonite consists of randomly
interstratified 13.6-A and 9.5-A collapsed layers. The 13.eA Fourier transform peak for
expanded layers is evidence that the layer charge in the expanded, TMPA-saturated layers is
partially reduced, allowing TMPA aromatic rings to assume a flatter orientation than in the
normal-charge montrnorillonite. If the layer charge of the expanded layers were not
reduced at all, there would be a 15-A peak in the Fourier transform, not a 13.6-A peak.

 

 

31

4000

 

00
O
O

1 .00
A. NonnaI-Charge 1 3TB

 

N
O
O

 

41 55
43.55

 

 

 

 

Electron Density (arbitrary)

 

 

 

 

.2000'IlijjjIIUITTUTIEYUV'VUUII‘IIII'U‘II'IUII‘IjT'U

0 5 1 0 1 5 20 25 30 35 40 45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1200

1 B. Reduced-Chargee
..... 0° 13.60 SE EEE

800
E n EC 2710 .3
g 600 0 ° cos 8
h '- 8 O O
s 400 1?, 3 9°C or 8 :3 5:, “e3

- In or - °
3;"? 200 8 ,2 8 °’ '1 ‘33 E 53, g 8. 8 0 ‘s‘r‘
: " oi to 8 89" w "" 9.’ v 8 8
. . O N

8 O N o v 0‘)
c '- no
2 -200 or g
:8 -400
I."

-600 U

-800

0 5 10 15 20 25 30 35 40 45

Distance (A)

Figure 2-3. Fourier transforms of X-ray diffraction data for TMPA-
montrnorillonites. Peaks due to combinations of expanded (E) and collapsed (C)
layers are labeled for reduced-charge montrnorillonite.

32

The proportions of expanded and collapsed layers in reduced-charge TMPA-
montrnorillonite cannot be estimated directly from the amplitudes of the 13.6-A and 9.5-A
peaks in the Fourier transform because the baseline for the 9.5-A peak cannot be
determined accurately. But the primary diffraction peaks at 4.46, 3.35, and 1.93 A (Table
2-2) and Mering's method indicate that reduced-charge TMPA-montrnorillonite comprises
70 to 80% TMPA-saturated, expanded layers, a result that is in good agreement with the
estimate for the glycerol-solvated Na-montmorillonite. Random interstratification in the
reduced-charge clay makes it impossible to interpret the Fourier u'ansform in terms of the
interlayer electron density of adsorbed TMPA.

DISCUSSION

 

Infrared dichroism experiments (Figure 2-1a) revealed that none of the symmetry
axes of TMPA is perpendicular to the siloxane sm'face of normalecharge montrnorillonite.
The 15-A d-spacing measured by X-ray diffraction, however, indicates that the aromatic
ring may be perpendicular to the surface. Other researchers have measured 143-A to
152-A d-spacings for montrnorillonite when polar aromatic compounds are adsorbed with
their aromatic ring and x-axis perpendicular to the siloxane surface (Greene-Kelly, 1955;
Serratosa, 1966). Therefore, a IS-A d—spacing supports the contention that the TMPA
phenyl ring is perpendicular to the siloxane surface. The x- and z-axes may be tilted in
order to minimize the distance between the negative charge in the clay and the positively

charged nitrogen atom of TMPA, as shown in Figure 2-4.

33

608.5,." 2.88:“ 05 95 .58. 22er 3508

v5 0388a Gen .53.. 05 .a :39..." oh 39:53 cornea Baommeofimvoeo Soc 333.38g
sauce—o bun—no.5 .ooatsm 0536:” 05 2 Esomccofiom met 3:23 55 sane—€955:—
owaaoénce: :o nous—Ste ESE. woodman mo catacomeqe omen—Bum .VN 033m

 

 

a....

EkHtitt‘tthSI‘S-01001-5“1‘11555'01'5“-5“0]-55'01!5-5-1011155555010555f

 

 

 

 

 

 

 

 

 

 

 

 

34

The lS-A d-spacing is smaller than the combined dimensions of the aromatic ring
(6.7 A in the x-direction) and the clay layer (9.4 A according to Brindley and Hoffmann,
1962). Consequently, a lS-A d-spacing requires "keying" of aromatic protons into the
siloxane surface when the TMPA phenyl ring is perpendicular to the surface of
montrnorillonite. The separation between adjacent protons in an aromatic ring is 2.5 A,
whereas the distance between adjacent siloxane ditrigonal cavities is 5.2 A Thus, one
aromatic proton may key into a ditrigonal cavity and another may fit into a closest-packing
arrangement with surface oxygen atoms. This type of keying is shown schematically for
TMPA in Figure 2-4.

The orientation of adsorbed TMPA must minimize the separation between the
negatively charged site in the clay and the positively charged nitrogen atom of TMPA.
With the phenyl ring roughly perpendicular to the siloxane surface, charge separation will
be minimized when one methyl proton from each of two methyl groups keys into the
siloxane surface (Figure 2-4). The interlayer electron probabilities from the one-
dimensional Fourier analysis are superimposed on Figure 2-4 to show that the proposed
orientation is consistent with the Fourier analysis results. The electron probability plot
shows forn' main regions of interlayer electron density that are nearly symmetric with
respect to the midpoint of the clay interlayer and to the upper and lower siloxane surfaces.
Keying of methyl and aromatic protons is required to give the Fourier transform peaks at
4.80 and 10.15 A (Figure 2-4), which are nearly coincident with the upper and lower
siloxane sm'faces. The peaks in the electron probability plot are symmetric with respect to
the siloxane surfaces and the midpoint of the interlayer, even though individual adsorbed
TMPA ions lack this symmetry, because some TMPA ions have their positively charged
nitrogen atom close to the bottom clay surface and some TMPA nitrogen atoms are closer to

the upper surface of each interlayer (Figure 2-4).

35

In the normal-charge montrnorillonite, the surface area per negative charge is about
151 A2, based on unit-cell dimensions and the structural formula of Wyoming
montrnorillonite. In the TMPA orientation depicted in Figure 2-4 with the aromatic ring
perpendicular to the surface, each interlayer TMPA ion occupies a total of 100 A2 on the
upper and lower siloxane surfaces, which includes 25 A2 on each surfaces for the edge area
of the aromatic ring and 25 A2 per surface for the projection of two methyl groups
(Pauling, 1960). Thus, TMPA ions occupy about 213 of the available siloxane surface area
per negative charge on normal-charge montrnorillonite, which leaves 1/3 is accessible to
other adsorbates. This molecular-level picture is consistent with macroscopic surface area
measurements, provided thath can form a bilayer in the interlayer of normal-charge
TMPA-montrnorillonite: the measured Nz-accessible surface area of 252 m2 g'1 (Table 2-
l) is about 1/3 the theoretical surface area of 770 m2 g'1 calculated from the unit-cell
dimensions and structural formula of Wyoming montrnorillonite (Gast, 1977). Thus,
surface area measurements support the X-ray Fourier analysis result that the aromatic ring
of TMPA is perpendicular to the siloxane surface of normal-charge montrnorillonite. In
this orientation, it is possible that neighboring TMPA phenyl rings may interact with one
another by 1M: interactions, although the experiments reported here provide no direct
evidence to either support or disprove this hypothesis.

If benzene were to adsorb perpendicular to the surface in order to interact with
TMPA phenyl rings by 1t-1t interactions, each benzene ring would require 50 A2 (25 A2 on
both the upper and lower siloxane surfaces of an interlayer), which is exactly the accessible
siloxane surface area on dry normal-charge TMPA-montrnorillonite. The accessible surface

would be even smaller in the presence of adsorbed water. Thus, no aromatic molecule
larger than benzene could adsorb on normal-charge TMPA-montrnorillonite if TMPA ions

were uniformly distributed on the surface. The fact that small amounts of alkylbenzenes

 

 

36

and naphthalene are sorbed from aqueous solution by high-charge TMPA-montrnorillonites
(Jaynes and Boyd, 1991) is evidence that TMPA ions are not uniformly distributed on the
surface.

The perpendicular orientation of TMPA on normal-charge montrnorillonite (Figure
2-4) will allow other aromatic compounds to sorb in the interlayer with their phenyl rings
perpendicular to the surface. Maximum sorption of aromatic compounds will occur when
their phenyl rings are perpendicular to the surface because the interlayer space of 5.4 A is
not large enough to accommodate two or more aromatic rings stacked parallel to the
surface, and sorption of a single ring parallel to the surface would be more sterically
resuicted and less energetically favorable than sorption with aromatic rings perpendicular to
the surface. Because sorption parallel to the surface is less favorable, it is unlikely that p-it
interactions between surface oxygen atoms and aromatic sorbates conuibute to arcne
sorption by normal-charge TMPA-montrnorillonite. A perpendicular orientation, however,
would allow aromatic sorbates to interact with the phenyl ring and with one another by rt-rt
interactions but at the same time permit polar sorbates such as water to interact with the

positively charged nitrogen atom of TMPA.

The 135A d-spacing for reduced-charge TMPA-montrnorillonite, which is not
large enough to accommodate the phenyl ring perpendicular to the surface, is the same as
that reported for tetramethylammonium (TMA)-montmorillonite (Bauer and Reay, 1957;
Barrer and Perry, 1961). Thus, a 13.6-A d-spacing suggests that TMPA's three methyl
groups prop open adjacent clay layers, with some methyl protons keying into the siloxane
surface, as shown schematically in Figure 2-5. Infrared dichroism showed that the phenyl

ring is not parallel to the surface of reduced-charge montrnorillonite. Therefore, the ring

37

.555.
8: mm ceasefire 888 :5 .833.“ 05 o. _o_§n 8: mm met 3:25 .835» usage—mm 95 9.38.
899:— 1388 2:8 53 3:8» 352: 3 some Began 2a 2&3 .ozceztoeneofi mew—5»?
owafileooaoe .«o flue?— 3335 no :ozscoto <2 we 5:32.889: oquoaom .m-~ 2:me

 

 

1

 

 

 

 

 

 

 

 

 

 

 

38

must be intermediate between parallel and perpendicular, or it may be free to rotate within
the interlayer space. One-dimensional Fourier analysis provided no information about the
orientation of interlayer TMPA cations because collapsed and TMPA-satmated (expanded)
layers are randomly interstratified in the reduced-charge montrnorillonite.

The surface area per negative charge on TMPA-containing expanded layers is 205
A2 per site when calculated with the following equation:

 

 

(814 mmolc)
kg NCM:____ (M) = (205 A2)
(1422221.) ,_ Ilayer RCM charge NCM —_charge “£854“
kg RCM 0.75 expanded layer RCM

where 449 and 814 mmolc kg'1 are the TMPA-exchange capacities of the reduced-charge
(RCM) and normal-charge montrnorillonites (NCM), 0.75 is the approximate proportion of
expanded layers in the reduced-charge clay, and 151 A2 is the area per structural charge in
normal-charge Wyoming montrnorillonite. Each TMPA cation would occupy 130 A2 if the
aromatic ring were parallel to the clay surface and 100 A2 if perpendicular. Because the
TMPA orientation must be intermediate between perpendicular and parallel, each TMPA ion
must occupy between 100 and 130 A2 on reduced-charge Wyoming montrnorillonite. This
leaves 75 to 105 A2 per cation available to other adsorbates, which is 1.5 to 2.1 times the
accessible area on normal-charge montrnorillonite. For comparison, the N2 BET surface
area calculated for expanded layers of reduced-charge TWA-montrnorillonite was 424 m2
g'1 (i.e., 318 m2 g'1 + 0.75 expanded layers), which is 1.68 times the Nz-accessible area

on normal-charge monmlorillonite.

39

Previous research has shown that benzene sorption causes TMA-montrnorillonite to
expand from 13.6 to 15.0 A so that the benzene ring can adsorb perpendicular to the
siloxane surface (Barter and Reay, 1957; Barrer and Perry, 1961), and preliminary
experiments in our laboratory confirm that TMPA montrnorillonite also expands when
benzene is sorbed Presumably the favorable energetics of n-rt interactions among
aromatic rings promote layer expansion and the upright orientation of the aromatic ring. In
the case of TMPA-montrnorillonite, the driving force for benzene to sorb perpendicular to
the surface would be even greater than for TMA montrnorillonite because the phenyl ring of
TMPA can interact with aromatic sorbates by n-rt interactions. As was discussed above,
the adsorbate-accessible surface area for the perpendicular orientation would be would be
about twice that on normal-charge TMPA montrnorillonite. The greater accessible surface
area of reduced-charge TWA-montrnorillonite should allow more sorption, particularly of
larger aromatic compounds (J aynes and Boyd, 1991), and should permit faster diffusion

into the interlayer.

CONCLUSIONS

In normal-charge montrnorillonite, Fourier transform X-ray analysis showed that
the phenyl ring of TMPA likely is perpendicular to the siloxane surface, though none of the
ring vibrations is perpendicular to the surface. The perpendicular phenyl ring could serve
as a “nucleating site” for aromatic sorbates to interact with TMPA and with one another by
1t—1t interactions. At the same time, sorbed water and other polar compounds would tend
to interact with the positively charged nitrogen atom of TMPA; water and aromatic sorbates
may interact with different parts of the TMPA ion yet compete for siloxane surface area.

The perpendicular orientation of TMPA maximizes the surface area available to other

4o

sorbates, yet on normal-charge montrnorillonite only 50 A2 per cation is not occupied by
TMPA ions, which is exactly the area required for benzene to sorb with its aromatic ring
perpendicular to the surface. The fact that small amounts of larger aromatic compounds
are sorbed by high-charge TMPA-montrnorillonites (J aynes and Boyd, 1991) is evidence
that TMPA ions are not uniformly distributed on the surface and that some of the interlayer
pore spaces between adjacent TMPA ions are sufficiently large to accommodate
alkylbenzenes and compounds such as naphthalene.

The reduced-charge TWA-montrnorillonite used in these experiments is a
randomly interstratified mixture composed of about 25% collapsed layers with no interlayer
cations and about 75% expanded layers with a 13.6-A d-spacing. The 13.6-A d-spacing
indicates first that layers are propped open by the methyl groups of TMPA, not the phenyl
ring, and second that the charge of expanded layers is reduced sufficiently that the aromatic
ring need not be perpendicular to the surface, as it is in normal-charge montrnorillonite.
The phenyl ring is neither parallel nor perpendicular to the surface of reduced-charge
TMPA-montrnorillonite, though preliminary experiments suggest that the layers may
expand in the presence of aromatic sorbates. This would allow the phenyl ring of TMPA to
rorate into a perpendicular orientation and would allow n—rt interactions between TMPA
and the aromatic sorbates, as well as among aromatic sorbates. Rotation or TMPA‘s
phenyl ring into a perpendicular orientation on reduced-charge montrnorillonite would not
only increase the d-spacing but also would increase both the arcne-accessible surface area
per cation and the interlayer pore dimensions to more than twice those in the normal-charge
montrnorillonite. Thus, macroscopic surface area measurements, in which the N2-
accessible surface area of reduced-charge montrnorillonite was only 1.25 times that of
normal—charge montrnorillonite, give an incomplete picture of interlayer accessibility to

aromatic sorbates. Measured surface areas do not reflect the fact that essentially all of the

41

accessible surface area on reduced-charge montrnorillonite resides on the 75% of the layers
that are expanded, and that the phenyl ring of TMPA likely rotates in the presence of
aromatic sorbates (but not N?) to create more adsorbate-accessible surface area on reduced-

charge montrnorillonite.

REFERENCES

Azaroff, L.V. (1955) Polarization correction for crystal-monochromatized X-radiation:
Acta Crystallogr. 8, 701-704.

Barrer, R.M., and Reay, J .S.S. (1957) Sorption and intercalation by methylammonium
montrnorillonites: Trans. Faraday Soc. 53, 1253-1261.

Barrer, R.M., and Perry, 6.8. (1961) Sorption of mixtures, and selectivity in
alkylammonium montrnorillonites. Part II. Tetramethylarrunonium
montrnorillonite: J. Chem. Soc. 850-858.

Boyd, S. A., Jaynes, W. F. and Ross, B. S. (1991) Immobilization of organic
contaminants by organo-clays: Application to soil restoration and hazardous waste
containment: in Organic substances and sediments in water, R.S. Baker, ed., vol.
1. CRC Press, Boca Raton, FL 181-372.

Bradley, W.F., Weiss, El, and Rowland, RA. (1963) A glycol-sodium vermiculite
complex: Clays & Clay Minerals 10, 117-122.

Brindley, G.W and Ertem, G. (1971) Preparation and salvation properties of some
variable charge montrnorillonites: Clays & Clay Minerals 19, 399-404.

Brindley, G.W., and Hoffmann, R.W. (1962) Orientation and packing of aliphatic chain
molecules on montrnorillonite: Clays & Clay Minerals 9, 546-556.

Clementz, D.M., and Mortland, M.M. (1974) Properties of reduced charge
montrnorillonites: Tetra-alkylammonium ion exchange forms: Clays & Clay
Minerals 22, 223-229.

Clementz, D.M., Mortland, M.M., and Pinnavaia, TJ. (1974) Properties of reduced
charge montrnorillonites: Hydrated Cu(II) ions as a spectroscopic probe: Clays &
Clay Minerals 22, 49-57.

42

Cope, A.C., Ciganer, E., Fleckenstein, LJ. and Meisinger, MAP. (1960) Tertiary
amines from methiodides and lithium alurrrinum hydride: J. Amer. Chem. Soc.
82, 4651-4655.

Farmer, V. C., and Mortland, M. M. (1966). An infrared study of the coordination of
pyridine and water to exchangeable cations in montrnorillonite and saponite: J.
Chem Soc. 344-351.

Gast, R.G. (1977) Surface and colloid chemistry. In Minerals in Soil Environments,
J.B. Dixon and SB. Weed, eds., American Society of Agronomy, Madison, WI,
27 -73.

Greene-Kelly, R. (1953) Theidentification of montrnorillonoids in clays: J. Soil Sci. 4,
233-237. '

Greene-Kelly, R. (1955) Sorption of aromatic organic compounds by montrnorillonite.
Pt. 1. Orientation studies: Trans. Faraday Soc. 51, 412-424.

Jaynes, W.J., and S.A. Boyd. (1991) Hydrophobicity of siloxane surfaces in smectites
as revealed by aromatic hydrocarbon adsorption from water: Clays & Clay
Minerals 39, 428-436.

Jaynes, WJ., and Bigham, J.M. (1987) Charge reduction, octahedral charge, and lithium
retention in heated, Li-saturated smectites: Clays & Clay Minerals 35, 440-448.

Johns, W.D., and Sen Gupta, PK. (1967) Vermiculite-alkylammonium complexes:
Amer. Mineral. 52, 1706-1724.

MacEwan, D.M.C. (1956) Fourier transform methods for studying scattering from
lamellar systems I. A dth method for analysing interstratified mixtures: K olloid
Zeitschrift 14, 96-108.

Pauling, L. (1960) The Nature of the Chemical Bond, Cornell University Press, Ithaca,
New York, 3rd ed., 644 p.

Reynolds, R.C. (1980) Interstratified Clay Minerals. kl Crystal Structures of Clay
Minerals and their X -ray Identification, G.W. Brindley and G. Brown eds.,
Mineralogical Society, London, 249-304.

Serratosa, J.M. (1966) Infrared analysis of the orientation of pyridine molecules in clay
minerals: Clays & Clay Minerals 14, 385-391.

Serratosa, J.M. (1968) Infrared study of benzonitrile (CgHs-Cm-montmorillonite
complexes: Amy. Mineral. 53, 1244-1251.

CHAPTER 111
An FTIR study of water sorption on tetramethylammonium- and
trimethylphenylammonium-saturated normal- and reduced-charge Wyoming
montrnorillonite

ABSTRACT

The presence of water inhibits sorption of solutes from water on clays saturated
with organic cations. The purpose of this study was to determine whether water sorbs
preferentially on the siloxane surface or on cation sites of montrnorillonite saturated with
tetramethylammonium (TMA) and trimethylphenylammonium (TMPA). Water vapor
sorption isotherms were collected on TMA- and TMPA-saturated, normal- and reduced-
charge montrnorillonite to determine the effect of cation type and clay charge density on
water sorption. Normal- and reduced-charge clay films saturated with T'MA and TMPA
were equilibrated at different water vapor pressures and probed by infrared spectroscopy to
determine if water vapor sorbs on the cations and support conclusions drawn from the
macroscopic data of the sorption isotherms. Water vapor sorption isotherms showed that
sorption was greater for normal-charge clays than reduced-charge clays though the N2
surface area of the reduced-charge clays was larger. This suggests that water sorbs
preferentially on cations, not the siloxane surface. TMA-saturated clays sorbed more water
vapor than did TMPA-saturated clays, indicating either that the phenyl group of TMPA may
Sterically hinder sorption of water vapor. The infrared Specua of both WA and TMPA
cations saturating normal- and reduced-charge montrnorillonite were perturbed by water
vapor at 7.5% RH, providing further evidence that water interacts preferentially with
adsorbed TMA and TMPA, not with uncharged siloxane surfaces, at low relative humidity.

43

INTRODUCTION

Clays saturated with quaternary alkylammonium cations have been proposed as
sorbents to remove organic pollutants from aqueous waste streams (Boyd, et al., 1991), an
application in which water competes with organic solutes for sorption sites on the clays
(Lee et al.,1990; Jaynes and Boyd, 1990, 1991). The sorption mechanism of water on
clays saturated with inorganic cations has been studied extensively and reviewed elsewhere
(Farmer and Russell, 1971; Sposito and Prost, 1982), but the sorption mechanism of water
on clays saturated with quaternary ammonium cations is not well understood. The object
of this study is to determine whether water sorbs preferentially at the cation or on the
siloxane surface of clays saturated with tetramethylammonium (TMA) and
trimethylphenylammonium (TMPA). This information is necessary to determine how
pollutant sorption is inhibited by water and may aid selection of organoclays appropriate for
waste cleanup applications.

The preferred site of water adsorption on clays saturated with quaternary
ammonium cations has not been studied spectroscopically. When clays are saturated with
inorganic cations, the most important site for water adsorption is around the cations
(Farmer and Russell, 1971; Sposito and Prost, 1982). Infrared spectroscopic data
indicates that at high water content, water may also form weak hydrogen bonds with
siloxane surface oxygen atoms, but only after exchangeable cations have been hydrated
(Sposito and Prost, 1982). It is probable that small alkylammonium ions like TMA and
TMPA also are hydrated before water sorbs or condenses on siloxane surfaces. Like NH4+
(Bernal and Fowler, 1933), TMA has a small but finite hydration energy (Nagano et al.,
1988), and water sorption isotherms on TMA-montrnorillonite are nearly identical to water

sorption isotherms for NHf-montmorillonite (Gast and Mortland, 1971).

45

Sorption isotherms of arenes on TMA-saunated montrnorillonites collected in the
presence and absence of water have shown that water inhibits arcne sorption on these clays
(Lee et al., 1990). The effect of charge density on the sorption of arenes from water
provides indirect evidence for sites of water sorption on clays (Jaynes and Boyd, 1991).
Lee et al. (1990) stated that water sorption on either TMA cations or the siloxane surface
might have caused the decrease in arene sorption from aqueous solution compared with
arcne sorption on dry TMA-montrnorillonite. In addition, Lee et al. (1990) noted greater
sorption of arenes from water on low-charge TMA-clay than higher-charge clay. In Jaynes
and Boyd (1991) arene sorption from water was compared between reduced- and normal-
charge Arizona montrnorillonite saturated with TMPA to determine whether aromatic
hydrocarbons preferentially sorbed on the siloxane surface or the cations. Sorption of
arenes from water increased as the layer charge of the clay was reduced, implying that
arcne sorption was more favored than water adsorption on the siloxane surface and that the
siloxane surface was more hydrophobic than the cations. This result was in agreement
with the observations of Lee et a1. (1990) noting the effect of charge on arcne sorption.
These results suggested that the cations are the preferred sorption site for water since the
siloxane surface was more hydrophobic than the cations. Preferential water sorption on
cations could be evaluated more directly by collecting water vapor sorption isotherms on
normal- and reduced-charge TMA- and TMPA-saturated clay. This would provide stronger
evidence that cation sites are preferred for water sorption on TMA- and TMPA-saturated
clay over the siloxane surface.

This paper describes a combination of macroscopic and spectroscopic experiments
used to test the hypothesis that water interacts preferentially with adsorbed TMA and
TMPA ions, not with the siloxane surface. Water vapor adsorption isotherms on normal-
and reduced-charge TMA- and TMPA-montrnorillonite will provide macrosc0pic

information about the site of water sorption. If water sorbs preferentially on adsorbed

46

cations, more water should be adsorbed on the normal-charge clay, which has a higher
cation concentration. If, however, water interacts preferentially with siloxane surfaces,
then water vapor sorption at low PIPo should be greater on reduced-charge than normal-
charge clay. In addition to the water sorption isotherms, infrared spectroscopy will be used
to provide molecular-level evidence to test the hypothesis that water preferentially hydrates
adsorbed TMA and MA, not the siloxane surface. Hydration of adsorbed WA and
MA should perturb the infrared spectra of these cations. Thus, displacement or splitting
of cation vibrational bands at low water vapor pressru'e would support the hypothesis that
water interacts preferentially with adsorbed WA and T'MPA.

MATERIALS AND METHODS

 

Wyoming montrnorillonite (SWy—l) was obtained from the Clay Minerals Society
Source Clays Repository at the University of Missouri-Columbia. The <2-um fraction was
separated from the coarser material by sedimentation and the charge was reduced for half
the clay as described in Chapter 2.

To prepare TMA- and TMPA-montrnorillonites, TMA-chloride and TMPA bromide
were added to normal-charge and reduced-charge montrnorillonites in 70% methanol-water
suspensions and stirred 72 h. Four times the CBC of TMA-chloride and ten times the CEC
of TMPA-bromide were added to the clays. The clays were dialyzed until chloride-free and
were freeze-dried. To improve the homogeneity of the clays, each clay (normal-or
reduced-charge, TMA- and TMPA-saturated) was mixed gently in a mortar aml pestle.

Selected properties of the clays are reported in Table 3-1.

 

47

Table 3-1. Selected physical and chemical properties of normal (Norm)- and reduced
(Red)-charge TMA- and TMPA-montrnorillonite.

 

d(001) N2 Surface Area2 Total C3 Adsorbed cation‘ CEC’

 

Treatment (A) (m2 g-1 ) (g kg‘l) (mmol kg‘l) (mmolkg'1)
Red TMA 13.6 289i30 27.7mm 474:14 390:30
NormTMA 13.6 - 2021-26 4o.7:1:0.3 847:1:6 870:1:80
RedTMPA 13.7l 318:1:14 48.6:l:O.9 449:1:8 390:30
Norm T'MPA 14.9l 252i28 88.1i1.0 814111 870:1:80

 

1 Obtained from Fourier transform of XRD data (Chapter 2).

2 Avg :1: SD (N=2 or 3) from three-point BET N2 adsorption isotherms using a
Quantachrome Quantasorb Jr. surface area analyzer.

3 Avg :1: SD (N=2) of total C determined by combustion at 900 °C using a Dohrmann DC-
190 high-temperature carbon analyzer.

4 Avg i SD (N=2) calculated with the equation:

Adsorbed cation = Tom] 8 C x mmol catron
kg clay g c

 

 

5 Avg :1: SD (N=3 or 4) Determined by Na+ saturation and ammonium displacement of the
parent normal - and reduced - charge clays (i.e. not treated with organic cations).
Statistically identical values were obtained with Mg2+ saturation and Ba2+ extration.

48

W

Figure 3-1 shows the device used to collect water vapor sorption isotherms
(McBain and Bakr, 1926). The precision of the mass measurements was i 0.25 mg.
Vapor pressure was accurate to :l: 0.01 torr. Between 200 and 500 mg of each of the four
freeze-dried, homogenized clays (dried at 80 °C and stored in a P205 desiccator) was
plawd in the sample holders before the balance was sealed. Vacuum was applied to the
balance and the sample chambers were heated to 80 °C for 72 h to remove any water which
may have sorbed on the clay while transferring it from the desiccator to the balance. Gast
and Mortland (1971) showed that the infrared spectrum of dry TMA-montrnorillonite did
not change when the clay was heated to 105 °C, a result that suggests that no
decomposition occurs when TMA- and TMPA-montrnorillonite is heated to 80 °C. After
the heat and vacuum treatment, the four clay samples were allowed to return to room
temperature and the sorption isotherms were collected. The clays were equilibrated with
water vapor at absolute pressures ranging from 0.90 to 17.25 torr. Each isotherm point
was collected by opening the sorbate reservoir to the vacuum or near- vacuum in the sample
chambers, then closing the reservoir when the pressure rose sufficiently. After one hour of
equilibration, the manometer reading, temperature, and spring displacements were noted.
For example, one point was collected by opening the reservoir until the pressure rose to
2.00 torr, then equilibrating for one hour, at which time the pressure had dropped to 1.47
torr. After the spring displacements were measured, the water pressure was increased to
3.20 torr for the next point. Because it was important to collect water sorption data for
each of the two charge treatments and two organic cations under identical conditions, it was
necessary to have each of the four clays in the sample chamber simultaneously.

Consequently, it was not possible to collect replicates of each data point because final

49

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Pressure ' f" Digital
Transduce . m Manometer
H -
orbate
Reservoir
/ C) __Quartzsl-
Vapor 39""9
Trap Telescope
/
I: W3:! 9
@ ,Dispiacement
Scale
Turbo>nolecular
Vacuum Pump -— — — -— '
Sample
Holder

Figtu'e 3-1. Diagram of the McBain balance used to collect water vapor sorption
isotherms on TMA-and TMPA-saturated smectites. Water vapor pressure is
controlled by releasing water vapor from the sorbate reservoir into vacuum.
Quantity of water on the clay is determined by measuring the displacement of the
quartz springs attached to the sample holders.

50

equilibrium water vapor pressures could not be accumme predicted from initial pressure,
and it was nearly impossible to exactly replicate the final pressure from one experiment to

the next.
10W

flaxfilmmnamimt

Seventy mg each of normal-and reduwd-charge TMA- and TMPA- clay were
resuspended in 10 ml of methanol by sonicating for three 10-min intervals in an ice bath
using a Heat Systems sonicator with 2.5-mm microprobe and a setting of 4.5. After the
third sonication, coarse material was allowed to settle from the suspensions, l-ml aliquots
of the clay suspension were pipetted onto 13-mm diameter Delrinm AgCl Disks (El.
duPont de Nemours"), and the methanol was allowed to evaporate. It was necessary to
repeat the pipetting several times to obtain a sufficiently thick film, defined as a film for
which the clay O-H stretching vibration at 3650 cm'1 was approximately 20%
transmittance. The amount of clay on the disks was 0.8 :l: 0.15 mg cm‘z.

WWW

TMA- and TMPA-saunated Wyoming montrnorillonite clay films were equilibrated
in a specially constructed IR cell (Figure 3-2) using the saturated salt solutions or
concentrated H2804 solutions shown in Table 3-2 to regulate relative humidity. During
this equilibration the O-H stretching region of sorbed water (3000 - 3600 cm“) was
monitored until no change occurred (usually 1.5 h) Spectra were collected using a Perkin-
Elmer 1600 FTIR specu'ophotometer with a DTGS detector, 2 cm1 resolution, no
apcdisation, and 50 to 75 scans. The background specu'um was of the empty cell with a
clean AgCl disk placed in the sample holder.

Single-beam spectra were stored on diskette, imported into the MS-DOS program
SpectraCach’ (Galactic Software Inc.) then transformed into absorbance units. Specua

51

 

Plastic magnets

  
 

   
  

 

 

A disk with
gccljdy film

 

1
l
lfl
l.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AgCl window

Saturated salt solution
for regulating humidity

Figure 3-2. Diagram of the controlled atmosphere FTIR cell used to
equilibrate TMA- and TMPA-saturated smectites with water vapor and
collect spectra of clay-water complexes.

52

Table 32. Solutions used to regulate the partial
pressure of water vapor in the desiccator and in the
controlled-atmosphere FTIR cell used to equilibrate
T'MA and TMPA saturated smectites with water

 

vapor‘.

211212 salt

0 P205

0.075 77% H2804

0.11 66.2% H2804

0.15 saturated LiCl

0.17 61.5% H2804
0.20 saturated K-acetate
0.25 56.5% H2804

0.33 sattuated CaC12
0.43 saturated K2C03
0.52 saturated NaI-ISO4
0.58 saturated NaBr
0.65 saturated Mg-acetate
0.69 34.5% H2804

0.72 saturated N114Cl
0.81 saturated (NI-14)2804
0.92 saturated KZHPO4

 

1 Weast, 1987.

53

were smoothed using an 11-point Savitsky-Golay procedure. The baseline was leveled
along six or seven baseline points in the spectrum, then zeroed. Variance in the absorbance
of the lattice O-H stretch suggested that the thickness of the clay film probed by the IR
beam was not always the same. The following procedure was used to normalize all spectra
to correct for differences in film thickness: A baseline-leveled, zeroed absorbance spectrum
between 3800 and 2800 cm‘1 (Frgln'e 3-3a) was displayed and the clay O-H stretching band
removed by extrapolating a straight line between 3720 cm‘1 and 3560 cm'1 in the spectrlun,
giving Figure 3-3b. This second spectrum was subtracted from the first, leaving the
spectrum of the clay O-H band only (Figure 3-3c). The peak absorbance of this band was
measured and the multiplication factor necessary to bring this peak to 0.700 A was
calculated. The original (Figure 3-3a) spectrum was multiplied by this factor to give the
normalized spectrum (Figure 3-3d).
RESULTS AND DISCUSSION

1!! . . 1

Figure 3-4 shows water vapor sorption isotherms on normal- and reduced-charge
TMA- and TMPA-saturated montrnorillonite. Isotherrn shapes were consistent with water
sorption isotherms collected by Gast and Mortland (1971) for TMA-saturated smectites,
which were type II by Brunauer’s (1945) classification, suggesting strong interaction
between water and the sorbent. The amount of water sorbed by TMA-montrnorillonite in
this study was greater than Gast and Mortland (1971) observed, possibly because they
used saturated salt solutions to control the partial pressure of water vapor in a system of
one atrn total pressure, whereas the present study was conducted at near-vacuum
conditions.

The water vapor sorption isotherms (Figure 3-4) show that TMA-saturated clays

sorb more water vapor than TMPA-clays. This result indicates that TMA-montrnorillonites

 

 

 

 

1.00
°~°° ‘ n—A = 0.7621 "

§ 0.60 - -

g d d
0.20 - -' -/\‘\‘\~
0.00 - -

 

 

 

 

 

 

fi I 1 I I ' I T I T I 1 l I l U 1 r
3800 3600 3400 3200 3000 2800 3800 3600 3400 3200 3000 2800

 

 

 

 

 

 

 

 

 

Wavenumbers (cnr‘) Wavenumbers (cm-1)
1.00
03°" ' A=O.7795
' —A=o.68437 ‘ ”—
§ 0.60- n -
a .
D Faeroe-91°—
; 0.407T 0.68437 -
3 - a
< 0.20- J k C - D
0.00- -
I l I j T

 

 

 

 

I I I U U l j —l I I I 1 1
3800 3600 3400 3200 3000 2800 3800 3600 3400 3200 3000 2800
Wavenumbers (cnr‘) Wavenumbers (901")

Figure 3-3. Example of the procedure used to normalize all spectra to the same film
thickness by normalizing the absorbance of the lattice O-H stretch. A) Original
spectrum. B) Original spectrum with the lattice O-H stretching vibration removed.
C) difference spectrum of A - B, giving the lattice O-H stretching vibration only and
the multiplication factor necessary to bring the absorbance of this band to 0.700A. D)
Original spectrum (A) after multiplying it by the factor calculated in (C).

55

 

 

 

 

 

 

P/Po

I Normal-charge El Normal-charge
O Reduced-charge o Reduced-charge

Figure 3-4. Water vapor sorption isotherms on normal-(squares) and reduced-(circles)
charge clay saturated with TMA (left) and TMPA (right). Graph also shows number of
water molecules sorbed per cation at different values of x/m.

56

are more hydrophillic than TMPA-montrnorillonites. This behavior could either be because
T'MA likely has a greater hydration energy than TMPA (assuming that water interacts with
the adsorbed cations) or because the phenyl groups of adsorbed TMPA sterically inhibit
sorption of water vapor by TMPA-saturated clays as described below.

The most notable feature of the sorption isotherms in Figure 3-4 is that within each
cation treatment, normal-charge clays sorb more water vapor than do reduced-charge clays,
even though reduced-charge clays have greater N2 surface area than do normal-charge clays
(Table 3-1). If water preferred to sorb on the siloxane surface, greater sorption would have
occurred on the reduced-charge clays. This is strong evidence that the cation, not the
siloxane surface is the preferred sorption site for water on clays saturated with TMA and
TMPA. This also supports the conclusion of Jaynes and Boyd (1991) that the siloxane
surface of TMPA-saturated clays is more hydrophobic than the cations.

When the amount of water vapor sorbed is calculated on the basis of water
molecules sorbed per cation (Figure 3-4), it appears that reduced-charge clays sorb at least
one more water molecule per cation than do normal-charge clays for mom values of P/Po.
On the TMA-saturated clay, this difference appears first at about P/Po=0.21, where there
are five water molecules per cation on the reduced-charge clay, but foru' on the normal-
charge clay. On the TMPA-saturated clay, at P/Po=0.21, there are three water molecules
per cation on the reduced-charge clay but two on the normal-charge clay. A possible
explanation for this behavior is that close proximity of adjacent cations on normal-charge
clays may sterically restrict the number of water molecules in cation hydration shells of
normal-charge clays. Steric restriction of hydration shell size may also explain why
TMPA-montrnorillonite has fewer molecules per cation at the same PIPo than does TMA-
rnontrnorillonite: the bulky TMPA phenyl group may also restrict the number of water
molecules in the TMPA hydration shell, especially on the normal-charge clay.

57

The trends in water vapor sorption shown in the sorption isotherms (Figure 3-4)
canalso beobserved in the infrared spectra ofadsorbed wateron TMA- (Figure 3-5) and
TMPA- (Figlne 3-6) saturated montrnorillonites. The sorbed water O-H stretching (3600
cm‘l) and deformation (1630 cm") vibrations increase in intensity with increasing relative
humidity (and amount of water sorbed). The IR spectra provide additional qualitative
evidence that more water is sorbw on TMA- than TWA-montrnorillonites, though the

effect of charge reduction is more difficult to discern. Because the molar absorptivities of
sorbed water O-H vibrations decrease with increasing water content (Johnston et al, 1992)
and may also depend on other factors such as layer charge, the intensities of the water O-H
bands are only shown to provide a link between the water sorption isotherms and the

infrared data described below.
Ell] E l 1 . 'l .

Infrared band assignments for methyl symmetric and asymmetric deformation
vibrations of TMA are given in Table 3-3. The infrared band positions of TMA (estimated
from the center-of mass, not peak maximum) on both normal- and reduced-charge clays are
shown as a function of relative humidity in Figure 3-7, with selected infrared spectra
shown in Figure 3-8. The increase in frequency of the methyl asymmetric deformation
(Figure 3-7, top) and the slight decrease in frequency of the methyl symmetric deformation
(Figure 3-7, bottom) with increases in relative humidity suggest that water vapor sorbs on
the cations at low humidity. The methyl symmetric deformation vibrations are less intense
(Figure 3-8) than the methyl asymmetric vibrations, which may have introduced some error
into their estimated center-of-mass peak positions and contributed to the greater scatter in
the methyl symmetric deformation data (Figure 3-7). The frequencies of both the methyl
symmetric and asymmetric deformation vibrations shift most between 0% and 20% RH,

58

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Reduced-charge
1.00
q
d 1 e
0.80-1 - m
. j 72% .
8 0'60“ :W
c J
S q " 33% 1630
§ 0.40- -l
. I 1
¢ ‘ 15% 629
020- -
IJ :W
- 0%
0.00- "W
'IfiT'I‘fi' lllTlllIlfil
3800 3600 3400 3200 3000 2800 1800 1700 1600 1500
Wavenumbers (cm-1) Wavenumbers (em-1)
Normal-Charge
1.00
‘ : 632
0.80- ..
q .3 72%
g 0.60 - .. 43% 630
g .1 .l 631
g 0.40" -, 33%
.n . ' 1630
< : 15%
020- 1M
3 0%
0.00- {A A g
' 1 T T ‘ i ' l ' l r r 1 r r r r r r_l'_l_‘
3800 3600 3400 3200 3000 2800 1800 1700 1600 1500
Wavenumbers (cm-1) Wavenumbers (cm-1)

Figure 3-5. Infrared spectra of reduced-charge (top) and normal-charge (bottom)
TMA- saturated Wyoming montrnorillonite at six different relative humidities. The

left side (3600 - 2800 cm'l) shows the increase in intensity of the sorbed water O-H
stretching vibration (3400 cm‘l) with changes in relative humidity. The right side
(1800 - 1500 cm4) shows the increase in intensity of the sorbed water O-H bending
( 1630 cm-1) vibration with changes in relative humidity.

59

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Reduced-Charge
1.00
' 3 1632
0.80- a
a 1 72%
8 .1 1630
c 0.60- 1 43%
g . 1‘ 1630
9, 0.40- . 33%
2 1 - 15% 103°
0'20“ 2' 75% 0
d 3 0%
0.00- . -M
l ' T ' l ' l I r r r r r r r r
3800 3600 3400 3200 3000 2800 1800 1700 1600 1500
Wavenumbers (cm°‘) Wavenumbers (6'0")
Normal-Charge
1.00 .
' 3 1632
0.80- -
. j 72%
8 .‘ 1630
3 05°" 1 43%
§ . 1‘ 1630
£0.40“ 1 33%
‘ - I 15% 1630
“2°" 1 7.5% 1 0
q 3 0%
0.00 "1.3 ‘2.“
l ' l I 1 r l r l r r r r r r
3800 3600 3400 3200 3000 2800 1800 1700 1600 1500

Wavenumbers (cm-1)

Wavenumbers (cm-1)

Figure 3-6. Infrared spectra of reduced-charge (top) and normal-charge (bottom)
TMPA-saturated Wyoming montrnorillonite at six different relative humidities.

The left side (3600 - 2800 cm'l) shows the changes in intensity of the sorbed water
O-H stretching vibration (3400 cm'l) with changes in relative humidity. The right
side (1800 - 1500 cm'l) shows the changes in intensity of the sorbed water O-H
bending (1630 cm'l) vibration with changes in relative humidity.

60

Table 3—3. Infrared band assignments for methyl symmetric and
asymmetric deformation vibrations of chloride, bromide, and
iodide salts of tetramethylammonium (TMA) in pressed KBr
pellets, and observed peak positions of of TMA-Cl 1)dispersed
in KBr using diffuse reflectance (DRIFT) and 2)dissolved in

methanol using attenuated total reflectance (ATR).

 

 

Methyl asym Methyl symm
defamation—.mm
frequency (cm'1)———
TMA-Cl (pressed)‘ 1490 1405, 1398
TMA-Br (pressed)' 1488 1405, 1397
TMA-1(pressed)‘ 1483 1403, 1395
TMA-Cl (DRIFDb 1488 1404, 1398
TMA-Cl Methanol (ATR)" 1491 1414

 

‘ Bottger and Geddes (1965)
b This study

Normal-Charge

 

.a
h
m
‘1

Wavenumber (cm‘l)
6 35
01 a:

1484

Methyl asymmetric def

 

 

61

Reduced-Charge

 

 

Methyl asymmetric def

0 1020304050 60708090100 0 1020304050 60708090100

1425
1424
{"1423
31422
~1421
£1420
31419
21416
51417
1416
1415

Relative humidity

 

Methyl symmetric def
0 10 20 30 40 50 60 70 80 90100 0 10 20 30 40 50 60 70 8090100

 

Relative humidity

 

Relative humidity

 

 

Methyl symmetric def

 

Relative humidity

Figure 3—7. Effect of relative humidity on the center-of-mass frequency of the methyl
asymmetric (top) and methyl symmetric (bottom) deformation vibrations of TMA
saturating normal-charge (left) and reduced-charge (right) Wyoming
montrnorillonite.

.chcwmmma 28.. 8a m 03$. 8m .3283 gun—8 E mow—3:0 .23 22353 cougar—80..

03088.3 93 0558152: 05 me 5:33.: 5 mow—35 9.36% EEO—EEEE «:35»?
3333-55 9&5 036-8652 Ba 26: $353.58 do 3.8% Babe .3 Beam

 

$.83 23:55.63
83 83 com. Sm. coop
_ . p — _ ii—lill _
99: £6 .1
i ii:
vdpfiw Ymmvp o\om.h .
2:3 983 $m. n
. . ill ll”.
:5 ~83 $8 .
ill/1
2:: 0.8! $9 w
2:: «.83 $~n .
T
«.83 I
1

 

 

 

 

 

 

 

 

0920.833:

 

 

 

 

 

$.53 23835.62,
82 . 8.3 . 3m. . 8m. 82
o .l
993 $0 n
. $m$ .
at: 1
.. 3: new: $9 n
3:: \ $8 1
Es: g 933 #3”
. «.33 1
at: , \\ $m\. 4
923 n
\ 3m: 1
1
‘\ 39: m
3.20.3882

eoueqrosqv

 

63

after which there is little change. This suggests that beyond 20% RH, water molecules
don’t interact directly with the cation, but either form a second hydration shell around the
cation or interact with the siloxane surface. It may also be possible that above 20% RH,
water continues to interact directly with cations, but does not cause any additional shift in
the methyl vibrations. However, at 20% and larger RH, there may already be at least 4
water molecules per cation on TMA-clay (Figure 3-4). Stearically, it seems unlikely that
more than four water molecules can be coordinated directly to each cation.

Infrared band assignments for some methyl vibrations of TMPA are given in Table
3-4. Infrared band positions on both normal- and reduced-charge clays for all relative
humidity treatments are plotted in Figure 3-9, and selected infrared spectra shown in Figure
3-10. Peak maxima, not peak center-of-masses, are plotted in Figure 3-9 because the
methyl asymmetric deformation vibration (~1490 cm’l) overlapped with the v19. and vlgb
ring stretch peaks, which made it impossible to estimate the center-of-mass for any of these
peaks. The methyl symmetric deformation vibration of TMPA on normal-charge clay
increased with increasing relative humidity, but no real trend is apparent in the reduced-
charge clay (Figure 3-9, top). The methyl asymmetric deformation frequency of adsorbed
TMPA definitely increased with increases in relative humidity on normal-charge clay
(Figure 3-9, bottom). On the reduced-charge clay, the methyl asymmetric band was only a
shoulder for many of the RH treatments, but appeared to exhibit the same trend as the
normal-charge clay (Figm'e 3—9, bottom). Since the peak maxima positions could only be
resolved to :i: 1 cm'1 on the TMPA spectra it was difficult to determine exactly at what
relative humidity the slope changed on the RH vs. peak position plots (Figure 3-9). On
normal-charge TMPA-montrnorillonite it appears that above 15 to 25% RH, additional

64

Table 3-4. Infrared band assignments for the V191. and v19}, C-C ring stretch of methyl-
deuterawd (d9) TMPA iodide and for the v19” vlgb, and methyl vibrations for TMPA-Br in
1)KBr using diffuse reflectance (DRIFT) and 2)in methanol using attenuawd total
reflectance (ATR).

v19a C-C vl9b C-C Methyl Asym Methyl Symm other

 

TMPA-l d9 1 1496 1459 -..... -..-.. .....
TMPA-Br (om 1500 1461 1475, 1431 1416 1451
TMPA-Br (methanol) 1493 1464 1475 1411 .....

 

‘ Chapter 2

  

65

Normal-Charge

 

Methyl symmetric dei

0 1020 3040 50 60 70 8090100
Relative humidity

 

 

 

Methyl Asymmetric def

 

0 10 20 3040 50 60 70 8090100

Relative humidity

Reduced-Charge

 

Methyl symmetric def

0 10 20 30 40 50 60 70 80 90100
Relative humidity

 

 

     

o = shoulder
Methyl Asymmetric def

 

0 10 20 304050 60708090100
Relative humidity

Figure 3-9. Effect of relative humidity on peak maximum frequencies of the methyl
symmetric (tap), and asymmetric (bottom) deformations of TMPA saturating normal-
charge (left) and reduced-charge (right) Wyoming montrnorillonite.

écoficwama e55 he.“ v 03¢. com 588::— 3:22 5 mowfino 55 wee—“En; 12.08
93 mew—Bob” met. 05 .«o .9225 23 35:62“ E mow—3:6 union.“ o=:o=:o§=oE mezzo»?
3223-....3... 25 69358962 65 26: 959.659. .6 8.8% 626.5 .2 -m .655

 

 

 

 

9.53 23555.63
83 8.: 8m. 89 82
p — — — _ — _

a: 6: 8: gal...
93 a: $3 H
o: l u

23 a: $21...
9.. .

33 mm: . u
23 $8 ”
#31. mag. o\omv n.
23 .
a: 8: 9
«we Fave u

 

 

 

09256830:

 

 

 

 

$.83 2353533
83 cm: comp own. 89
. -Iu I—p — — p _
93 w m at”.
A .
m 3 mm: «omfin.
we: “ il/ln
.9;
GFVw 6&va 1
oi— m 1
9: a
33 _’“ $8 H
33 {one ....
83 v— l/.:
0
co: emu n
came I owe. s

 

 

 

09.30-3832

eoueqrosqv

67

water sorption had no effect on the band position, whereas the same was true above 15%
RH on the reduced-charge TMPA clay (Figure 3-9, bottom).

Water sorption also affected the intensity of the methyl asymmetric deformation
vibration at ~149O cm'l. Assuming the v19. ring stretch vibration at 1496 cm’1 remains at
the same intensity with water sorption, the intensity of the methyl asymmetric deformation
vibration decreases with water sorption. On normal-charge clay the methyl asymmetric
deformation peak is resolved at all RH treatments, while on reduwd—charge clay, the
methyl asymmetric deformation vibration becomes a shoulder with water sorption.

Infrared band assignments for some ring stretching vibrations of TMPA are given
in Table 34. Peak maxima for the v19, and the Vlgb bands exhibited no trends with water
sorption (Figure 3-10), so neither band was plotted as a function of relative humidity. On
normal-charge clay, the v19. band remained at 1496 cm‘l, while on reduced charge clay,
the band appeared as a peak or shoulder at 1496 cm"or 1495 cm“. The low-intensity Vlgb
band is obscured by an unassigned methyl vibration at ==s1468cm'1 (this methyl band was
not present on methyl-deuterated TMPA-clay). The peak maximum of the Vlgb remained at
1462 cm‘1 for all relative humidity treatments on reduced-charge clay while on normal-
charge clay, this peak maximum shifted between 1462 cm'1 and 1464 cm'l, at least in part
because of shifts in the unidentified methyl band (Figure 3-10).

When absorbances of the methyl asymmetric deformation at 1490 cm'1 and the
V19. band at 1496 cm‘1 are compared between relative humidity treatments (Figure 3-11), it
appears that the v19. band remains at the same intensity while the methyl asymmetric

deformation band decreases in intensity with water sorption. This decrease in intensity is

68

$3355. 0332 5 mom—3:6 .23 .5595? 5:55.533.
03023 1505 05 be £525 5 women—.0 wfiBofi 352E655:— wEan? 3.888
355 5ng 935-6832 ea 83 6565.656: 56 658% Baas ._ 7m 652m

 

 

 

 

$.53 9.35.526; $.53 935.5325
Dog. 691. com F 0mm _. 009. 091 83. com _. 0mm _. coo _.
— — b _ — — — — _ _ _ — — b n n n _
f I
f .
925/ H v
. .. 4
:5 so 8.: .. .
omvm u T
. 225 1
n 1: $6 8: n
.. mow _. ..

 

 

 

 

 

 

3.20.0833: 0950.35.62

eoueqrosqv

69

most evident between 0% and 7.5% RH (Figme 3-10), with less distinct changes in
intensity with higher relative humidity.

The changes in intensity and position of the methyl asymmetric deformation
vibrations of TMPA on normal- and reduced-charge clay show that, as with the TMA-
montmorillonite, the infrared spectra of adsorbed TMPA are affected by water sorbed at
low relative humidity, strongly suggesting that the preferred sorption site for water on
TWA-saunted clays is on the adsorbed TMPA cations. On TMPA, it ftu'ther suggests
that water preferentially interacts with the trimethyl portion of TMPA cations, since the ring
stretching vibrations do not appear to be affected by water sorption. However, it is also
possible that the ring stretching vibrations are not perturbed because their positions are
more resistant to changes in the environment of the molecule.

CONCLUSIONS

Water vapor sorption isotherms showed more water sorption on the normal-charge
clays than on reduced-charge clays, even though the N2 surface area was greater for the
reduced-charge clays. This indicated that the organic cations, not the siloxane sm'face were
the preferred sorption sites for water vapor and that these cations are hydrophyllic. Water
vapor sorption isotherms also showed that TMA-saturated clays sorb more water vapor
than do TMPA-saturated clays, suggesting that TMA-saturated clays are more hydrophillic
than TMPA-saturated clays. The lower sorption of water by the TMPA-saunawd clays may
be due to steric hinderance by the TMPA phenyl group, the lower hydration energy of
TMPA or borh. The calculated number of water molecules sorbed per cation on the
reduced-charge clay was greater than on normal-charge clay, possibly because the closer
cation spacing on the normal-charge clay restricts the number of water molecules within
hydration shells around adjacent cations. The effect of water sorption on the infrared

spectra of adsorbed TMA and TMPA provide additional evidence that the preferred site of

70

water sorption on these clays at low water content was the organic cations, not the siloxane
sm'face.

The conclusion that the cations, and not the siloxane sm'face are the preferred site
for water sorption on 'I'MA- and TMPA-sauu'ated clays will aid the design of organoclays
for remediation of aqueous waste streams containing nonionic organic compounds. The
sorptivity of organoclays for nonionic organic compounds can be increased by decreasing
the number of cations on the clay, which will in turn reduce the amount of water competing

for sorption sites on the clay.

REFERENCES

Bemal, J.D., and Fowler, RH. (1933) A theory of water and ionic solution with
particular reference to hydrogen and hydroxyl ions. J. Chem. Phys. 1, 515-548.

Bottger , GL, and Geddes, AL. (1965). The infrared spectra of the crystalline
tetramethylammonium halides: Spectrochirn. Acta 21,1701-1708.

Boyd, 8. A., Jaynes, W. F. and Ross, B. S. (1991) Immobilization of organic
contaminants by organo—clays: Application to soil restoration and hazardous waste
containment: in Organic substances and sediments in water, R.S. Baker, ed., vol.
1. CRC Press, Boca Raton, FL 181-372.

Brindley, G.W, and Ertem, G. (1971) Preparation and solvation properties of some
variable charge montrnorillonites: Clays & Clay Minerals 19, 399-404.

Farmer, V.C., and Russell, JD. (1971). Interlayer complexes in layer silicates—The
structure of water in lamellar ionic solutions: J. Chem. Soc. Faraday Trans.. 67,
2737—2749.

Gast, R.G., and Mortland, M.M. (1971) Self-diffusion of alkylammonium ions in
montrnorillonite: Journal of Colloid and Interface Science. 37, 80-92.

Greene-Kelly, R. (1953) The identification of montrnorillonoids in clays: J. Soil Sci. 4,
233-237.

Jaynes, WJ., and Bigham, J.M. (1987) Charge reduction, octahedral charge, and lithium
retention in heated, Li-saturated smecdtes: Clays & Clay Minerals 35, 440-448.

71

Jaynes, W.J., and S.A. Boyd. (1991) Hydrophobicity of siloxane surfaces in smectites
as revealed by aromatic hydrocarbon adsorption from water: Clays & Clay
Minerals 39, 428-436.

Jaynes, W.F., and. Boyd, S.A (1990) Trimethyphenylammonium-smectite as an effective
adsorbent of water soluble aromatic hydrocarbons: J. Air and Waste Management.
Assoc. 40, 1649-1653.

Johnston, C.T., Sposito, G., and Erkckson, G. (1992) Vibrational probe studies of water
interactions with montrnorillonite: Clays &Clay Minerals 40, 722-730.

Lee, J.F., Mortland, M.M, Chiou, C.T, Kile, DE. and Boyd, S.A. (1990) Adsorption of
benzene, toluene, and xylene by two tetramethylammonium-smectites having
different charge densities: Clays & Clay Minerals 38, 113-120.

McBain, J.W., and Bakr, A.M. (1926) A new sorption balance: J. Am. Chem. Soc.
48, 690-695.

Nagano, Y., Sakiyama, M., Fujiwara, T., and Kondo, Y. (1988) Thermochemical study
of tetramethyl- and tetraethylammonium halides: Nonionic cohesive energies in the
crystals and hydration enthalpies of the cations. J. Phys. Chem. 92, 5823.

Sposito, (3., and R. Prost. 1982. Structure of water adsorbed on smectites: Chemical
Reviews 82, 553-573.

Weast, R.C. ed., (1986) CRC handbook of chemistry and physics 67th ed. CRC press,
Inc. Boca Raton, Florida. E-42.

CHAPTER IV

An FTIR study of competitive water-arcne sorption on
tetramethylammonium- and trimethylphenylammoniurn-saturated normal-
' and reduced-charge Wyoming montrnorillonite

ABSTRACT

Sorption of arenes by quaternary alkylammonium-saturated clays is enhanced by
reducing the charge density of the clay, but inhibited by water sorption and increasing
solute size. The objectives of this study are to determine: 1) whether arenes sorb on cation
sites of dry tetramethylammonium- (TMA) and trimethylphenylammonium- (TMPA)
saturated clays; 2) if water sorption inhibits arcne sorption on these sites; 3) if higher clay
charge density inhibits arene sorption on cation sites; 4) if water -arene competition and
high clay charge density inhibit arcne-cation interaction more for large arenes than for small
arenes. TMA- and TMPA-saturated clay films with different charge densities were treated
with different combinations of arcne and water vapor and probed using infrared
spectroscopy to determine the effect of charge density, water sorption, and arcne sorption
on the position of cation vibrational frequencies. Arene sorption perturbed vibrational
frequencies of adsorbed TMA and TMPA differently than did water sorption, allowing
specuoscopic differentiation between the water and arcne interaction with TMA and TMPA.
The infrared data show that benzene and ethylbenzene sorption occm's on cation sites on

dry TMA-and TMPA-sattn'ated clays. The closer cation spacing of higher charge TMA-

72

73

and TMPA-clays, apparently restricted arcne sorption from a larger proportion of cation
sites compared to reduced-charge clay. The cation vibrational frequencies of clay films
exposed towater vapor and saturated benzene vapor remained at the frequencies
characteristic of benzene alone for all relative humidity ueaunents. The cation vibrational
frequencies of films exposed to ethylbenzene and water vapor shifted from those
characteristic of ethylbenzene vapor alone toward frequencies characteristic of water vapor
alone as relative humidity increased. Thus, water sOrption drove ethylbenzene, but not
benzene, from cation sites. Water vapor sorption and higher clay charge density both

inhibited sorption of larger arenes more than smaller arenes.

INTRODUCTION

Clays saturated with quaternary alkylarrrrnonium cations have been proposed as
sorbents to remove organic pollutants from aqueous industrial waste streams (Boyd et al.,
1991). In this application, water and organic pollutants compete for sorption sites on the
clay. Possible pollutant sorption sites include the siloxane surface and the organic cations
(Lee et al., 1990; Jaynes and Boyd, 1990,1991).

Arene sorption on clays satmated with small organic cations such as .,
tetramethylammonium (TMA) and trimethylphenylammonium (TMPA) is affected by water
sorption, clay charge density and arene size. Vapor-phase arcne sorption on dry TMA-
montrnorillonite is much greater than is arcne sorption from aqueous solution (Lee et al.,
1990). The inhibition in aqueous sorption may be due to water hydrating the organic
cations on the clay (Chapter 3, Lee et al., 1990; Jaynes and Boyd, 1991). Lee et al.,
(1990) also observed that aqueous arcne sorption on TMA-montrnorillonite decreases with
increasing clay charge density and with increasing arene size. Jaynes and Boyd, 1991

observed decreased aqueous arene sorption with increased clay charge density on TMPA

74
clay, but they did not observe the arene size trend noted by Lee et al. (1990). With dry
vapor-phase sorption, Lee et al. (1990) observed similar clay charge effects where more
vapor phase sorption of toluene and o-xylene occurred on reduced-charge clay than on
normal-charge clay at all values of P/Po. At P/Po values less than 0.3, a similar reduction
in sorption occmred for benzene vapor on dry clay. Jaynes and Boyd (1991) concluded
that increases in arene sorption with decreasing charge density (and hence with increasing
uncharged siloxane surface area) indicated that in the presence of water, arenes sorb
preferentially onto uncharged siloxane surfaces. At P/Po values greater than 0.3, Lee et al.
(1990) observed more benzene sorption on the normal-charge clay than on reduced-charge
clay. This observation suggests that benzene may interact with cation sites at high P/Po as
a secondary sorption site.

The mechanism by which water inhibits arcne sorption is not well understood.
Spectroscopic studies can be conducted to deterrrrine whether arenes interact with TMA and
TMPA, with uncharged siloxane surfaces, or with both types of potential sorption sites.
Two hypotheses are suggested in the literature and Chapter 3 for the mechanism for arcne
sorption inhibition on TMA- and TMPA-clay. The first hypothesis is that water inhibits
arcne sorption on TMA- and TMPA-clay by hydrating the adsorbed cations and displacing
arenes that interact with the cations on dry TMA— and TMPA-monunorillonites. This
hypothesis is consistent with infrared spectroscopic data reported in Chapter 3 which
showed that water interacts with both TMA and TMPA cations even at PIP0 corresponding
to between one and two H20 molecules per cation, depending on cation type and layer
charge. The second theory (Lee et al., 1990) is that when water molecules hydrate cations
in clay interlayers, they sterically restrict the diffusion of solutes such as benzene, toluene,
and xylene into the interlayer region. On low-charge clay, hydrated cations are farther
apart, leaving more space in which arenes can diffuse and sorb. The latter theory (Lee et

al., 1990) is also consistent with published sorption data that show that the inhibitory effect

75

of water on arcne sorption increases (i.e., arene sorption decreases) as the number of arcne
methyl substituents increases (Lee et al, 1990). Another reason for the greater inhibitory
effect of water sorption of larger arenes, however, may be that water more readily displaces
methyl-substituted benzenes from cation sorption sites. The two hypothesized mechanisms
may act simultaneously to reduced arene sorption on TMA- and TMPA-clay.

Spectroscopic data are needed to test these hypotheses. This paper describes
research in which the positions of TMA and TMPA vibrational frequencies in the infrared
spectra of TMA- and TMPA-saturated normal-and reduced-charge montrnorillonite clay
films treated with different proportions of arene and water vapor were recorded and
interpreted to answer the following questions: 1. do arenes sorb on cation sites of dry
clays saturated with TMA and TMPA? 2. does water sorption cause arcne desorption from
cation sites? 3. does higher clay charge density inhibit arcne sorption on cation sites? 4. is
sorption of a larger arcne (ethylbenzene) on cation sites more inhibited by water sorption

and high charge density than is benzene sorption?

MATERIALS AND METHODS

- n h 1
Wyoming montrnorillonite (SWy-l) was obtained from the Clay Minerals Society
Source Clays Repository at the University of Missouri-Columbia. The <2-rrm fraction was
separated from the coarser material by sedimentation and the charge was reduced for half
the clay as described in Chapter 2. Normal- and reduced charge montrnorillonite satmated
with TMA- and TMPA-were prepared as described in Chapter 3. Selected properties of the
clays are reported in Table 4—1.

76

Table 4-1. Selected physical and chemical properties of normal (Norm)- and reduced
(Red)-charge TMA- and TMPA-montrnorillonite.

 

d(001) N2 Surface Area: Total c3 Adsorbedcation‘ crac5

 

Treatment (A) (m2 g'l) (g kg") (mmol kg“) (11111101 kg“)
Red TMA 13.6 289fl0 27.710.7 474114 390130
Norm TMA 13.6 202126 40.7103 84716 870180
Red TMPA 13.71 318114 48.6103 44918 390130
Norm TMPA 14.91 252128 88.111.0 81411 1 870180

 

1 Obtained from Fourier transform of XRD data (Chapter 2).

2 Avg :1: SD (N=2 or 3) from three-point BET N2 adsorption isotherms using a
Quantachrome Quantasorb Jr. surface area analyzer.

3 Avg 1 SD (N=2) of total C determined by combustion at 900 °C using a Dohrrnann DC-
190 high-temperatme carbon analyzer.

4 Avg :l: SD (N=2) calculated with the equation:
Total g C mmol cation

Adsorbed cation = kg clay x g C

 

5 Avg :l: SD (N=3 or 4) Determined by Na+ saturation and ammonium displacement of the
parent normal - and reduced - charge clays (i.e. not treated with organic cations).
Statistically identical values were obtained with Mgz“ saturation and Ba” extration.

77

W

W

Seventy mg each of normal-and reduced-charge TMA- and TMPA- clay were
resuspended in 10 ml of methanol by sonicating for three lO-min intervals in an ice bath
using a Heat Systems sonicator with 2.5-mm microprobe and a setting of 4.5. After the
third sonication, coarse material was allowed to settle from the suspensions, l-ml aliquots
of the clay suspension were pipetted onto 13-mm diameter Delrinn‘ AgCl Disks (E.I.
duPont de Nemours"), and the methanol was allowed to evaporate. It was necessary to
repeat the pipettin g several times to obtain a sufficiently thick film, defined as a film for
which the clay O-H stretching vibration at 3650 cm-1 was approximately 20%,

transmittance. The amount of clay on the disks was 0.8 i 0.15 mg cm‘z.

 

TMA- and TMPA-saturated Wyoming montrnorillonite clay films were equilibrated
overnight in a desiccator that contained a liquid reservoir of either benzene, benzene-dg, or
ethylbenzene, and either P205 or an aquous salt or H2804 solution. The free liquid
benzene, benzene-dg, and ethylbenzene provided an atmosphere that was saturated with
respect to that arcne, whereas the P205 and the aqueous salt and H2804 solutions
controlled the partial pressure of water (Table 4-2). After overnight equilibration in the
desiccator, the films were transferred rapidly from the desiccator into a gas-tight IR cell
(Figme 4-1) with the same salt or acid solution found in the equilibration desiccator. No
arenes were added in the IR cell. The films were equilibrated with water vapor in the IR
cell for 75 rrrin. Benzene and ethylbenzene sorption Specua were collected using a dry-air-
ptn'ged Perldn-Elmer P131600 FTIR spectrophotometer with a DTGS detector, 2 cm:1

resolution, no apodisation, and 50 to 75 scans. Benzene~dg sorption spectra were collected

78

Table 4—2. Solutions used to regulate the partial pressure of water vapor in desiccator and
in the controlled-atmosphere FTIR cell.

 

Ellie salt

0 P205

0.11 66.2% H2804
0.33 saturated CaClz
0.43 saturated K2003
0.58 saturated NaBr
0.72 saturated NH4C1

1.00 pme water

 

79

 

Plastic magnets

  
 

 

 

 

 

 

 

 

 

 

 

AgCl disk with
_‘ clay film
.4 ....................................... .fi.
nnnnn -—> —
> a
.J"

 

 

 

 

 

 

 

 

 

 

AgCl window

Saturated salt solution
for regulating humidity

Figure 4w]. Diagram of the controlled atmosphere FTIR cell used to
equilibrate arcne-saturated TMA-and TMPA-saturated clay films with
water vapor and collect spectra of clay-water -arene complexes

80

using an Nz-purged Nicollet lR/42 FTIR with a DTGS detector, 2 cm‘1 resolution,.l-Iapp-
Genzel apodisation, and 100 scans. A background spectrum of the empty cell with a clean
AgCl disk placed in the sample holder over the appropriate salt or acid solution was

collected for each arcne-water vapor treatment .

Single-beam spectra were stored on diskette and imported into the MS-DOS
program SpectraCach’ (Galactic Software Inc.) then transformed into absorbance units.
Specua were smoothed using an 11-point Savitsky-Golay procedure. The baseline was
leveled along six or seven baseline points in the spectrum, then zeroed. Variance in the
absorbance of the lattice O-H stretch suggested that the thickness of the clay film probed by
the IR beam may not always the same. The spectra were normalized for this variation to
make the intensities of all clay O~H stretching vibrations at 3650 cm'1 equal. (The details

of this proceedure were described in Chapter 3.

RESULTS

WW
Benzene sorption
Band assignments for TMA and benzene are given in Tables 3 and 4, respectively.
The 0% RH spectra in Figure 42 show the effect of benzene sorption on the infrared
spectrum of dry normal- and reduced-charge TMA-saturated Wyoming montrnorillonite.
The intense v19 ring stretching vibration of sorbed benzene (1478 cm") overlapped with

the weaker methyl asymmetric deformation vibration of TMA on both normal-charge

81

Table 4-3. Infrared band assignments for methyl symmetric and asymmetric deformation
vibrations of chloride, bromide, and iodide salts of tetramethylammonium (TMA) in
pressed KBr pellets, and observed peak positions of TMA-Cl 1)dispersed in KBr using
diffuse reflectance (DRIFT) and 2)dissolved in methanol using attenuated otal reflectance

(ATR).

 

 

Methyl Asym Methyl Symm
E I E H 1 fl . I E .
TMA-Cl (pressed) 1490 1405, 1398
TMA-Br (pressed) 1488 1405, 1397
TMAJ (pressed) 1483 1403, 1395
My
TMA-C1 DRIFT 1488 1404, 1398
TMA-Cl Methanol 1491 1414

 

Table 4-4. Infrared band assignments for neat
liquid benzenel, ethylbenzenez, and benzene-d3.

V19 CH3/CH2 bcnd
AERL. —quucncy (curl)—

Benzene 1478 ——
Ethylbenzene 1495(v19.) 1452
Benzene-dé 1333 —

 

 

1 Duinker and Mills, 1968
2 Green, 1962
3 Varsanyi, 1969

82

85:85 x3: 58.: 38558 803 «.528: gaun—

.v 03am. 5 2853 58 .m aid. 5 :03» 03 SE. 5.. 8:05:28.“ :55 58:85.8 :8 coo—08m
:5 .038 058 on. 5 :35: 2a 5.8% .920 2:. $20 355581253:th $559383: 58
30: 0985-555: :e 5.58% 33:5 05 cc 5558 533-0585: .855 5.8.1.5 .N-v Beam

 

 

 

 

 

 

 

 

  

 

 

83:3 Km .11 a}:

$.50. 5355.952, . :58 53535325
80.. 8.3 oo.m P 8.0 P on": 08.. com _. 8.1V P 0mm F on“? cm: :03
1 , 2353 =82
at... 0:033 .32 m an:
14 . as: .32 82:8 8 .1: .8
mp3 («.83 2353 o: .1: °er n
I! . 93 2353 am .1: exec
93 2553 .8 .1: .xb w.
. :3 33:0: 5m .11 $3

   

0:033 Km .11 $8

 

 

2.351518 119.58

  

m;

 

2853 .8 .1: exam: 2853 am .1: arms

on:

 

 

Ujfi'IIIII

mu? w

lll'UlUIIUleUUliijj

 

 

 

33506338: «92.0.3532

 

eoueqrosqv

83

(1485.4 cm'l) and reduced-charge (1484.5 cm") clay, but if the resulting peaks at 1480
cm‘1 on normal-charge clay and 1479 cm'1 on reduced-charge clay were due more to the
' methyl asymmetric deformation vibration of TMA than sorbed benzene, then benzene
sorption caused the methyl asymmetric deformation vibration to shift to lower frequency.
The methyl symmetric deformation vibration of TMA (=1418 cm'l) sorbed on norrnal- and
reduced-charge clay shifted to lower frequency in response to benzene sorption. These
shifts to lower frequency suggest that benzene interacts with TMA cations.
Benzene-d6 sorption

Infrared spectra of benzene-dfi sorption on TMA-montrnorillonites (Figure 4—3)
provided more information about possible interactions between benzene and adsorbed
TMA, because the v19 band of deuterobenzene (Table 44) did not overlap with any TMA
methyl vibrations. The 0% RH spectra in Figure 4-3 show the effect of benzene—dg
sorption on the infrared spectrum of norrnal- and reduced-charge TMA-saturated Wyoming
montrnorillonite. Benzene-dé sorption caused the center-of-mass of the methyl asymmetric
deformation vibration of TMA to shift from 1484.5 to 1482.2 cm'1 on the normal-charge
clay and from 1485.4 to 1481.9 cm‘1 on the reduced-charge clay. The methyl symmetric
deformation vibration also shifted to slightly lower frequency due to benzene-d6 sorption.
These shifts in the methyl deformation vibrations of TMA on normal- and reduced-charge
montrnorillonite suggest that benzene-d6 (and thus benzene) sorbs on cation sites of these
clays.

Ethylbenzene sorption

The 0% RH spectra in Figure 44 show the effects of ethylbenzene sorption on the

infrared spectrum of normal- and reduced-charge TMA-saturated clay films. The bands

84

635...?5058 Mao: 59: 003558 803 Segue: :3: .v 035. 5

00.05853 05 m 035. 5 5.3» Pa <3; he: 9:05:35.“ 055 585558 he: 09.05.." :5 .038
058 05 a 5350 2a 58% 2F :30 02558159..th ems—.9083».— 05 an: 0920
-355: :0 5.58% 0235 05 :e 50:98 8838585850500 5355 we 80mm .mé. 23E

$.53 23550225
89 83 89 com. 8.: 89

ONVP O: . 0
«.33. 0:0: I: ‘8

0.53 20: co .00 .1: $0

méovp
E0: mo .00 .75 $00

   

2.3 h. «03
89 23 8 a... .1: $8.

23 «.83
/

209008.751}.

 

 

09.050.002.00:

$.53 535.5963

009 cc: com F coop co: oomw

LLIrL-ppL—pppp—pP-phbnp.

 

2: m3: 2388.155.

 

. $531.38.... 1:... .
t: . ‘28:
89 . ES 8 a... .1: $8

2.3. ‘\

9mm:
/

20: on 0: .1: {one

 

 

ltjU'YUr

oouaqrosqv

jYUITjYII

 

 

0905040532

85

05505 0.00: 58.: 00:05:00 0:03 0:028: 0.00: .v 030,—.

5 0:00:00350 0:0 m 030.: :m :35 0:0 <32. :8 0505:3000 .500 580.8500 :8 000—0000 :5
.2000 0500 0:. .0 :35 000 0.000% >20 2F $20 00.5500-<2:. 9.th 090.20.000.60.— 0:0 30:
0901020500: 00 53.00% 000035 05 :0 5:38 0000305500350 0055 .8 0000.5 .04. 0.53.:

1.53 90:50:05.5

bhhnhbbbD-bbbb-IDID—bh-b

009 003. :09 89 cot oomw

 

0:000 .002

  

v.8fiw 5:08 8% $0

 

 

/\ 40f «0: MIN 1 .
:3. 1mm: 0: IE 2

E000 + IC $2.

 

 

  

:3 0:020 + I: axon:

 

 

  

20:0 + Im .xhn

 

20:0 0: .Im e\ewh

 

 

 

09010600300:

$.53 20:53:30;

Dunn-DDDI-bbbl—bbLI-bbbh

comp 003 009 000.. oo: 08..

 

 

 

 

 

 

 

 

 

 

 

 

 

  

t: 23. + 1: .000

 
  

mgr

0:000 02 u

000. 000. 000. n
0:00:30 0:11: .\.o .

0:; 0.003 .

7 11 it! n
F 2000 + 1: .00 r

(003 50:01:15 n

00.. 00: 230.1: .000 n.

00: BB. 8 .1: .00...

 

 

0905010532

eouaqrosqv

86

due to sorbed ethylbenzene (Table 4-4) at 1495 cm'1 and 1452 cm“1 were very small, and
the TMA methyl asymmetric deformation vibration can be distinguished clearly. With
ethylbenzene sorption, the peak maximum of the methyl asymmetric deformation vibration
of TMA (~1485 cm") shifted from 1484.5 to 1484 cm'1 on the normal-charge clay and
from 1485.4 to 1483 cm'1 on the reduced-charge clay. Peak maxima were used to
determine the position of the TMA methyl asymmetric deformation vibration because the
ethylbenzene shoulder at ~1495 cm'l skewed the calculation of the peak center-of-mass.
The TMA methyl symmetric vibration also shifted to lower frequency due to ethylbenzene
sorption. The shifts in the methyl vibrations of TMA on normal- and rcduwd-charge
montrnorillonite suggest that ethylbenzene interacts with TMA adsorbed on these clays.
Summaryfor arcne sorption on dry TMA -clay

Shifts in the methyl asymmetric and symmetric deformation vibrations of TMA-
montmorillonite due to benzene-dg and ethylbenzene sorption strongly suggest that these
arenes sorb on cation sites of dry TMA-saturated clay. Differences in the TMA peak shifts
of dry TMA-saturated clays due to arene size and clay charge density will be discussed in a

later section.

 

Benzene-d6 sorption

When benzene-d6 vapor and water vapor competed for sorption sites at 43% and
100% RH, the TMA methyl asymmetric deformation frequencies remained nearly the same
as for benzene-dg sorption on dry TMA-montrnorillonites (Figure 4-3). At 100% RH,
visual inspection of the clay films suggested that water containing benzene-ds had
condensed on the clay, yet the TMA methyl asymmetric defamation vibration remained at

1482 cm], characteristic of benzene-TMA interaction. In contrast, when water sorbed on

87
normal- and reduced-charge TMA-clay at 43% RH in the absence of benzene, the methyl
asymmetric deformation vibration was at ~1486 cm"1 (Figure 4-3). Thus, there was no
evidence that water prevented benzene-TMA interaction, or that water interacted with TMA
ions in the presence of benzene. The same result was obtained whether the clay was first
equilibrated with water vapor only, then exposed to the benzene-dg-water vapor
commbination, or by equilibrating with benzene-dis first, follwed by the combination. This
suggested that benzene-dg (and, by inference, benzene) had more affinity for TMA cation

sorption sites than water did.

Ethylbenzene sorption

As RH increased from 0% to 72% in the competitive ethylbenzene-water vapor
treatments, the TMA methyl asymmetric deformation increased from 1484 to 1485 cm‘1 on
the normal-charge clay and from 1483 to 1485 cm'1 on the reduced-charge clay (Figure 4-
4). When water vapor sorbed on TMA-montrnorillonite at 72% RH in the absence of
ethylbenzene, the TMA methyl asymmetric deformation vibration was 1486 cm'l. Thus,

I increases in the methyl asymmetric deformation frequency of adsorbed TMA with
increasing relative humidity (Figure 4-4) indicatd that water displaced ethylbenzene from
TMA sites on both normal- and reduced-charge clays.

Though water vapor sorption apparently drove ethylbenzene from TMA cation
sites, there is evidence that water and ethylbenzene also competed for sites other than
cations, possibly the siloxane surface. The water O-H deformation band (1630 cm'l) of
the 72% RH ethylbenzene-treated clays was less intense in the presence of ethylbenzene
than when ethylbenzene was absent (Figure 44), indicating that ethylbenzene vapor
sorption suppressed water vapor sorption. In the presence of water, which displaced
ethylbenzene from cation sites, ethylbenzene may be restricted to sorbing on the siloxane
surface of the clay. Ethylbenzene sorbed on the siloxane surface may in turn inhibit water

sorption on the siloxane surface, or it may sterically restrict the size of the hydration shell

88

around the TMA cations on the clay. This would cause the intensity of the OH bending
vibration of the sorbed water to be less intense when water and ethylbenzene sorb
competitively.
Summary for arcne sorption on hydrated TMA-clay

Water vapor, even at 100% RH, did not cause benzene.d6 to desorb from TMA
cations. In contrast, water vapor was apparently preferred over ethylbenzene on the cations
of TMA-montrnorillonite. Both benzene-dé and ethylbenzene inhibit water sorption on
TMA-montrnorillonite, as shown by the intensifies of the water 01-! deformation peak at
1630 cm".

1 n ' 1

.For dry TMA-montrnorillonite, arcne sorption caused larger shifts in the methyl
asymmetric deformation vibration of TMA on reduced—charge clay than on normal-charge
clay. Benzene-d6 sorption caused a 2.3 cm‘1 peak shift to lower frequency on nonnal-
charge clay, but a 3.5 cm'1 shift on reduced-charge clay. Ethylbenzene sorption on TMA-
clay caused a 0.5 cm'1 shift to lower frequency on normal-charge clay, but a 2.4 cm'1
downward shift on reduced-charge clay. These larger shifts for the reduced-charge clay
may have occurred because the TMA cations on the surface of the reduced-charge clay are
further apart than are the TMA cations on normal-charge clay. Close spacing of cations on
the surface of the normal-charge clay may sterically restrict the arenes from sorbing on
some cation sites, whereas most or all TMA cations may be accessible to arenes on
reduced-charge clay. Greater cation-arcne interaction on the reduced-charge clay may have
caused the larger bandshift.

Em E . . I! I E l

The methyl asymmetric deformation vibration of dry TMA-montrnorillonite shifted
more with benzene-dé sorption (Figure 4-3) than with ethylbenzene sorption (Figure 4—4).

89

Two possible reasons for this behavior are: l) The ethyl group of ethylbenzene may
sterically hinder ethylbenzene diffusion into interlayer regions. Consequently, benzene
may sorb on a larger proportion of cation sites on TMA-montrnorillonite than ethylbenzene
does, causing the greater bandshift for benzene; 2)benzene and ethylbenzene may sorb
equally on cation sites, but benzene-d6 may perturb the methyl asymmetric deformation
vibration of TMA more than ethylbenzene does because the two arenes may interact with
TMA by different mechanisms.

Both close cation spacing (i.e., high surface charge density) and increases in
relative humidity apparently inhibit sorption of ethylbenzene more than benzene-d6. First,
as mentioned in the previous section, larger frequency shifts for reduced-charge than
normal-charge clay due to arcne sorption suggest that arenes interact with a larger
proportion of TMA cations in the reduced-charge clay. This effect was much greater for
ethylbenzene than benzene-<16 (Figures 3 and 4); the methyl asymmetric bandshift for
ethylbenzene sorption on dry, reduced-charge clay was nearly five times greater than that
measured on normal-charge montrnorillonite. Second, water displaces ethylbenzene, but
not the smaller benzene-d6, from TMA cation sites. Thus, both higher clay charge density

and water sorption decrease sorption of larger arenes on TMA cation sites more than

sorption of smaller arenes.
I! [E E _ .1] .
n ' n W? la
Benzene sorption

Infrared band assignments for TMPA are given in Table 4-5. Figure 4-5 shows the
effect of benzene sorption on the infrared spectrum of TMPA-saturated normal- and
reduced-charge Wyoming montrnorillonite at 0% RH. Peak locations were determined

from peak maxima. The methyl asymmetric deformation vibration of TMPA shifts from

90

Table 4-5. Infrared band assignments for the V191. and Vlgb C-C ring stretch of methyl-
deuterated (d9) TMPA iodide and for the v19” Vlgb, and methyl vibrations for TMPA-Br in
1)KBr using diffuse reflectance (DRIFT) and 2)in methanol using attenuated total
reflectance (ATR).

vl9a C-C vl9b C-C Methyl Asym Methyl Symm other
ring . . .

TMPA-I d9 1 1496 1459 -..-.. ---.- -----
TMPA-Br (DRIFT) 1500 1461 1475, 1431 1416 1451
TMPA-Br (methanol) 1498 1464 1475 1411 .....

 

1 Chapter 2

91

050.05 0.00.. 50.... 00.05000

0003 0:00.000 0.00.. .0 0.00... 5 0:00:00 0:0 0 0.00... 5 :0...» 000 <52... .0. 0505:3000 0:00 58.000500
.0.. 000.00.... .00 .2000 0500 00. 0. 550.0 00 0.00000 0.0 00... 0.0.0 00.05.00-<n=>:. 00w... 0m0009000000.
0:0 800 09009—0500: .00 50.00000 00000.0. 00. :0 :0098 00.03.0552. 00.05 .0 80......“— .n.0 0.0m...“—

 

 

 

 

 

  

   

   

 

 

$.50. 2005.52.02. $.50. 0.0055303
S0... . £0... . 30. 00.0. 3.. 80. 80. 8.0. 8.0. 00.0. 3.. 30.
0:00 .002 . 0:00 .002
0.0. .T 0.0.
n 0.0. :0 08 28 oz
0.0. 000. :0 0.0 23 oz 1 000.
00 .. 000.
000 .
0... 10.00.053.00 ... 0.0. 00. :0. \oo 00.00
0.0 .
0.0. _ :0 0.... 500.00 .10“: 000. :0 0 0
:_ .1 0.0.00 0‘— :m .000 500 .00
0.0. 000. I... ...00 EB .00 .
000. . .. 0.0.
0.0. x r n. 00 _ 10.00.200.00
. 0.
o N: m
000. .00. :0 00. B. 0 . 0.000 .00.
an: hmvp ...I
n 0.0.
:0 0.0.. EB 02% 000 000. :0 .00. EB .

 

 

 

 

 

 

09. 00.0.00. .. .0. I002. , 09. 0.. 00.05.02

eouaqrosqv

92

1475 cm'1 for the bromide salt to ~1488 cm'l when sorbed on TMPA (Chapter 2).
Benzene sorption caused the methyl symmetric (~1412 cm'l) to shift to 1415 cm’1 on
normal-charge clay while no shift was observed on reduced-charge clay. On normal-
charge TMPA-montrnorillonite, the v19. ring stretch shifted 1 cm'1 to higher frequency,
whereas the methyl asymmetric deformation vibration decreased in frequency by 1 cm’l.
On reduced-charge clay, benzene sorption caused the v19.'to increase approximately 1
cm“, whereas the methyl asymmetric deformation vibration decreased 3 cm‘l. The shifts
in both the v19. ring stretch and the methyl asymmetric deformation vibration of TMPA
saturating normal- and reduced-charge clay indicated that benzene interacted with cation

sites of these clays, at both the aromatic ring and the methyl groups.
Ethylbenzene sorption

Figme 4.6 shows the effect of ethylbenzene sorption on the infrared spectrum of
TMPA-saturated normal- and reduced-charge Wyoming montrnorillonite at 0% RH.
Ethylbenzene sorption caused the methyl symmetric deformation vibration of TMPA on
normal-charge clay to shift up 1 cm— 1, while no shft occurred on the reduced-charge clay.
The TMPA v19, ring stretch cannot be distinguished from the V19. of ethylbenzene (1495
cm"), so the bandshift of the cation V19. cannot be determined. Clear shifts could be
observed in the methyl asymmetric defamation band of adsorbed TMPA. On normal-
charge clay, the methyl asymmetric deformation shifted down 2 cm‘l, while on reduced-
charge clay, the peak shifted down 3 cm'l. The shifts in the methyl asymmetric
deformation of TMPA saturating normal- and reduced-charge montrnorillonite indicated that
ethylbenzene interacted with cation sites of these clays at the methyl groups. Ring

interactions were not observable.

93

$.50. 0.0055052.

80. 000.

com. com. 8.... com.

bDbP-LDID—IPDb—IIbI-DDDD

 

 

0:0 .031.

mm:
000..

 

«m0. mm:

 

 

0. 0:30 .82

 

I: «8 0:00 oz

 

Omv _.

 

[1.3“ '3. I: ..\oo 0:000 .00

I _
gr 'L’ In. 0\0: neon: aw

.000. . 0.. 0
.5. 1:00 0.. .00

1.: .5

L2. 000. 10.0.0. 0:80.00

 

 

UiijfiUUIIIijTIT'UTTIIIIIIIUIITYIIY

 

 

00.000.002.00:

050.05 0.00.. 50.... 00.05000 500000.
0.00: .n 200... 5 «£2... 0:0 0 200 h 5 :35 0.0 505003.00 .0.. 0505:5000 0000 500000500 .0..
00200.0 .00 .2000 0500 00. 0. 550.0 0.0 0000.... 00.0 2:. 50.0 00.05.00-<:E c0»... 00.000.000.00. 0:0
92. 0900000500: .0 5.500% 00.0.05 00. :0 :00900 020305500350 0055 .0 .08....— .0-0 0.55.:

$.50. 0.0055302.

FDPh—bebb-PDI-bbhh-Dbb-

com. 80.. com. com. com. com.

 

 

 

0:000 .002

mm... mm: 0 .

   

I: $0 0:000 02

I: 0\oo 0:000 .8

3‘9
0

000.111

 

I... <0... Neon: Em

a... _.
'21 0.1:..000500000

0.0. - .
000 0.0. . 0..
000. a... I: 00. 80.00

_
000. 000 — 000.

0.0. 0.0.
000 I: 0.0. 0:30 02
000h—\ 000.

 

eoueqrosqv

 

j—T‘IU'VIIIIIUli'III'I'U'III'ij'I'

 

00.009.05.02

94

Summalyfor arcne sorption on dry T MFA-clay
Shifts in the methyl asymmetric defamation band of TWA-montmorillonite due to
benzene and ethylbenzene sorption strongly suggest that these arenes interact with the
methyl groups of TMPA on dry TMPA-saturated montrnorillonite. Shifts in the v19, of
TMPA with benzene sorption suggest that ring interactions, possibly by 1M: interactions

are possible as well.

 

Benzene somtion

When benzene and water sorbed competitively on the normal-charge clay at relative
humidities up to 72%, the methyl asymmetric deformation band stayed at 1487 cm'l, the
frequency characteristic of benzene sorption in the absence of water (i.e. noncompetitive
benzene sorption). When benzene and water sorbed competitively on the reduced-charge
clay at relative humidities up to 33%, the methyl asymmetric deformation band stayed at the
frequency characteristic of noncompetitive benzene sorption (1486 cm‘l), but increased to
1487 cm‘1 as RH was increased to 72%. The 1 cm'1 upward shift in the methyl
asymmeuic deformation of TMPA on the reduced—charge clay suggested that some
competition occm'ed between water and benzene with the higher relative humidity
‘ treatments, but not enough to cause the band to shift to the characteristic frequency of
noncompetitive water sorption (1492 cm'l). Therefore, water didn’t displace all of the
benzene that interacted with adsorbed TMPA.

The O-H bending frequencies of sorbed water at 72% RH on normal- and reduced-
charge TMPA-clays (1630 cm’l) were more intense in the absence of benzene than when

benzene was present (Figure 4-5), which indicated that benzene sorption inhibited water

95

sorption. The shifts in the TMPA methyl asymmetric deformation band clearly showed that
benzene inhibited water sorption on TMPA cations. In addition, benzene sorption on bath
the TMPA ions and on the uncharged siloxane sm'face of TMPA-montrnorillonite may have
inhibited water sorption on the siloxane surface and restricted the size of the hydration shell
around those cations from which water was able to displace benzene.
Ethylbenzene sorption

When ethylbenzene and water competed for sorption sites on TMPA-
montrnorillonite at relative humidities up to 72%, the methyl asymmetric deformation
increased from 1486 or 1487 cm'1 (normal- and reduced-charge, respectively) to 1490
cm'l, close to the frequency observed when water sorbed in the absence of ethylbenzene
(Figure 4-6). This result suggested that water displaced ethylbenzene from TMPA sites.

The lower intensity of the sorbed water O-H deformation vibration (1630 cm'l) of
ethylbenzene-water treated clays compared to water-only treated clays (Figure 4-6)
indicated that less water sorption occured when the two solutes competed, even at 72% RH
when ethylbenzene had been driven from the cation sites. A possible mechanism for this
decrease in water sorption is that ethylbenzene sorbed on siloxane surface sites inhibited
water sorption on the siloxane surface and may also have restricted the size of the hydration
shell around TMPA cations.

Summary for arene sorption on hydrated TMPA-clay

Water vapor was apparently preferred over ethylbenzene on the cations of TMPA-
montrnorillonite. Water vapor was not preferred over benzene on these sites, however.

Effecmflmcharmamzmfinmmax

When arenes sorbed on dry TMPA-montrnorillonite, larger shifts in the methyl
asymmetric deformation vibration of TMPA were observed for reduced-charge than
normal-charge TMPA-clay. Benzene sorption caused a 1 cm'1 peak shift to lower

96

frequency on normal-charge clay, but a 3 cm‘1 downward shift on reduced-charge clay.
Ethylbenzene sorption caused a 2 cm'1 peak shift on normal-charge clay, but a 3 cm'1
downward shift on reduced-charge clay. Thus, the effect of charge reduction on cation-
arene interaction for TMPA-montrnorillonite was similar to the effect for TMA-
montrnorillonite, and the explanation proposed above for the greater cation-arene interaction
in reduced-charge TMA-clay should be equally valid for the TWA-montrnorillonites.

E E: E l . . I! [E E I

There was no evidence that arene size affected arenes sorption on cation sites on dry
TMPA-clay. Benzene (Figure 4-5) and ethylbenzene (Figure 46) caused identical shifts in
the methyl asymmetric deformation vibration on dry, reduced-charge TMPA-
monmrorillonite. On normaLcharge TMPA-monumrillonite, ethylbenzene caused a 1 cm'1
greater band shift than did benzene. This latter result might suggest that less benzene than
ethylbenzene sorbed on TMPA sites, but this does not make sense sterically. It is more
likely that incomplete purging of water vapor from the spectrometer before the benzene
sorption spectra at 0% RH were collected may have caused the methyl asymmetric
deformation peak to apparently shift from 1486 cm’1 (its position with all other arene, 0%
RH TMPA-clay samples) to 1487 cm“, the position observed for benzene sorption on
reduced-charge TMPA-clay.

As was discussed above for TMA-montrnorillonite, increased relative humidity
appeared to inhibit ethylbenzene sorption more than benzene sorption on cation sites.
Shifts in the TMPA methyl asymmetric deformation showed that water drove ethylbenzene
from TMPA cation sites, but not smaller benzene molecules. In contrast to the data
reported for TMA-montrnorillonite, however, there was no clear evidence that higher
surface—charge density (i .e. tighter cation spacing) caused a greater decrease in cation-

arcne interaction with ethylbenzene than with benzene.

 

97

DISCUSSION

Spectroscopic data presented here showed that arene sorption from the vapor phase
occurs on cation sites on dry TMA- and TMPA-saturawd clays. This conclusion appears to
conflict with previous experiments (Jaynes and Boyd. 1991) showing that in the presence
of water, benzene and other arenes sorb primarily on uncharged siloxane surfaces, not on
cation sites of TMPA-saturated smectites. The spectroscopic data reported here, do not rule
out the siloxane surface as an important site of arcne sorption, particularly in the presence
of water because this specu'oscopic study was not designed to directly evaluate this type of
sorption. Clay films prepared for this study were too thick to observe the Si—O bands of
the clay, which might be perturbed by arcne sorption. Furthermore, the waterzarene ratio
was much higher in the work reported by Jaynes and Boyd (1991) than in the present
study. Since the preferred site of water sorption on these clays is on cation sites (Chapter
3), it is probable that the large excess of water during aqueous-phase arcne sorption
experiments (Jaynes and Boyd, 1991) caused complete arcne desorption from cation sites
and left the siloxane surface as the only remaining site for arcne sorption.

Benzene sorption caused greater infrared peak shifts in the methyl asymmetric
deformation vibration of TMA on dry normal- and reduced-charge clay than did
ethylbenzene sorption, implying that more benzene than ethylbenzene interacted with TMA
cations. However, another reason for this difference in peak shift could be that individual
benzene molecules perturb the methyl groups of TMA more than do individual
ethylbenzene molecules. On dry TMPA-saturated normal- and reduced-charge
montrnorillonite, benzene and ethylbenzene caused nearly identical shifts in the methyl
asymmetric deformation vibration. This result suggests that the same amount of benzene as

ethylbenzene may be sorbed on cation sites of dry TMPA-clay.

98

The band shifts observed for benzene and ethylbenzene sorption on dry, reduced-
charge TMA- and TMPA-saturated clays were greater than. those for normal-charge clay.
This suggested that the proportion of cation sites where arene sorption occurred was greater
on the reduced-charge clay than the normal-charge clay. This may be because tighter
pacldng of cations on the surface of the normal-charge clay sterically restricted arenes from
sorbing on some cation sites. On reduced-charge clay, proportionally more cation sites
may be accessible for arene sorption.

Water sorption drove ethylbenzene, but not benzene from cation sorption sites on
both TMA- and TMPA-clay. There are at least two explanations for this behavior. One
possibility is that the competitive sorption of arene and water vapor may depend more on
absolute vapor pressures (P) than their fraction of saturation vapor pressure (P/Po). The
room temperature saturation vapor pressure of benzene is 100 torr, ethylbenzene is 10 torr,
and water is 24 torr (W east, 1986). Water, with its lower absolute vapor pressure, may
not be able to compete with the higher vapor pressure of saturated benzene, though it
competed with the lower vapor pressure of saturated ethylbenzene. The second possibility,
discussed in the results section, is that the ethyl group of ethylbenzene sterically hindered
ethylbenzene from sorbing in the interlayer and interacting with cations, while smaller
benzene molecules had greater access to clay interlayers. Under this second possibility,
ethylbenzene may be excluded from many sorption sites at which benzene is capable of
sorbing, leaving water molecules better able to compete with those ethylbenzene molecules
that are able to interact with TMA or TMPA.

\ The combination of water sorption and high clay charge density probably inhibits
cation-arcne interaction more for larger arenes, like ethylbenzene, than for smaller arenes

like benzene. This is because higher layer charge alone probably reduces overall sorption

99

regardless of arcne size, and water sorption inhibits ethylbenzene sorption more than
benzene sorption.

In organoclay waste treatment design applications, for sorption of the pure arene
from vapor, it appears from these results that sorption of arenes may be improved by using
a lower charge density parent clay, which will allow more space between cations for the
arenes to diffuse through. This becomes more important when sorbing arenes with ring
substituents. A larger proportion of cations on the clay perturbed by arcne sorption
suggests that more cation sites are accessible for arene sorption on the lower charge clay.
Where more cation sites are accessible for arene sorption on the lower charge clay, it
implies there are also more siloxane sm'face sites available for sorption. When sorbin g
arenes from aqueous solution, it is likely that cation sites do not function as significant
sorption sites because they are hydrated (Chapter 3). Since water drove ethylbenzene from
cation sites, it is possible that, with the much higher ratio of waterzarene in solution, that
benzene could be driven from cation sites in solution. In aqueous systems, it is more
likely, the cations function as pillars, holding the clay layers open (J aynes and Boyd, 1991;
Lee et al., 1990) rather than as sorption sites.

100

REFERENCES

Boyd, S. A., Jaynes, W. F. and Ross, B. S. (1991) Immobilization of organic
contaminants by organo-clays: Application to soil restoration and hazardous waste
containment: in Organic substances and sedimenm in water, R.S. Baker, ed., vol. 1.
CRC Press, Boca Raton, FL 181-372

Brindley, G.W and Ertem, G. (1971) Preparation and solvation properties of some
variable charge montrnorillonites: Clays & Clay Minerals 19, 399-404.

Duinker, LG, and Mills, LM. (1968) Normal coordinates for the planar vibrations of
benzene: Spectrochim. Acta. 24A, 417-435.

Green, J.H.S. (1962) Vibrational Specua of benzene derivatives-III. Anisole,
ethylbenzene, phenetole methyl phenyl sulphide and ethyl phenyl sulphide:
Spectrochim. Acta. 18, 39-50.

Greene-Kelly, R. (1953) The identification of montrnorillonoids in clays: J. Soil Sci. 4,
233-237.

Jaynes, W.J., and Bigham, J.M. (1987) Charge reduction, octahedral charge, and lithium
retention in heated, Li-saturated smectites: Clays & Clay Minerals 35, 440-448.

Jaynes, W.F., and. Boyd, SA (1990) Trimethyphenylammonium-smectite as an effective
adsorbent of water soluble aromatic hydrocarbons: J. Air and Waste Management.
Assoc. 40, 1649-1653.

Jaynes, W.J., and Boyd, S.A. (1991) Hydrophobicity of siloxane surfaces in smectites
as revealed by aromatic hydrocarbon adsorption from water: Clays & Clay Minerals
39, 428-436.

Lee, J.F., Mortland, M.M., Chiou, C.T., Kile, DE. and Boyd, S.A. (1990) Adsorption
of benzene, toluene, and xylene by two teu'amethylammonium-smectites having
different charge densities: Clays & Clay Minerals 38, 113-120.

Vményi, G. (1969) Vibrational spectra of benzene derivatives. Academic Press, New
York and London. 430 pp.

Weast, R.C. ed., (1986) CRC handbook of chemistry and physics 67th ed. CRC press,
Inc. Boca Raton, Florida. D189-D212.

 

Appendix A

Supplementary FTIR Dichroism Data

 

101

MATERIALS AND METHODS

Wm

Methyl-deuterated TMPA-dg iodide was added to 10 mg of normal- or reduced-
charge clay (10 times the CBC) in 10 ml of a 70% methanol-water mixture and istirred for
25 hours. The clay suspension was washed free of excess salts by centrifuging, decanting
the supernatant, and re-suspending the clay in deionized water. Cenuifugation and
decanting was repeated three times. The TMPA-dg-clays were resuspended in 10 ml
methanol and disk-supported clay films prepared as described below below for
undeuterated TMPA-clays.

Seventy mg each of normal- and reduced-charge TMPA- clay were resuspended in
10 ml of methanol by sonicating for three lO-min intervals in an ice bath using a Heat
Systems sonicator with 2.5-mm microprobe and a setting of 4.5 (out of 10). After the third
sonication, coarse material was allowed to settle from the suspensions and l-ml aliquots of
the suspended clay were pipetted onto l3-mm diameter DelrinTM AgCl Disks (E.I. duPont
de Nemom‘s‘z’) and allowed to dry. It was necessary to repeat the pipetting several times to
obtain a sufficiently thick film, defined as a film for which the clay O-I-I stretching vibration
at 3650 cm‘1 was approximately 20% u'ansmittance. The amount of clay on the disks was
0.8 i 0.15 mg cm'z. These clay films will be referred to as “disk supported films”.

Maximum

Clay films satm'ated with TMPA-d9 plated on A gCl disks were stored in a

desiccator over P205 at least 24 h before infrared spectra were collected. The films were

102

removed from the desiccator and quickly placed in a sample holder in an Nz-purged
Nicollet IR/42 FI'IR. Transmittance spectra were collected with a DTGS detector using 4—
cm’1 resolution, Happ—Genzel apodisation, and 100 scans. Spectra were stored on diskette
as ASCII files and imported into the MS-DOS program SpectraCalc (Galactic Software
Inc.). then transformed into absorbance units using the spectrum of a clean Delrinm AgCl
disk as the background. Spectra were smoothed using an ll-point Savitsky-Golay
procedure. The baselines were leveled along seven selected baseline points in the
spectrum, then zeroed After collecting a spectrum of the clay films normal to the beam, a
second spectrum of each ueatrnent was collected with the clay film tilted approximately 30°.
The intensities of the v19, (A1 symmetry), and V19b(Bl symmetry) ring-stretching
vibrations and the OH out-of-plane deformation (B2 symmetry) vibration of the phenyl
group were compared between the two angles of incidence to determine whether the TMPA
phenyl group is either parallel or perpendicular to the siloxane surface in the clay interlayer.

Spectra of disk-supported clay films saturated with undeuterated TMPA were
collected on a Perkin-Elmer 1600 FTIR with no apodisation, 2 cm‘1 resolution and 50 to 75
scans with the film normal to the beam. Specua were stored on diskette, imported into
SpectraCalc" and analyzed as described above.
X- D . fifi .

Samples were prepared for X-ray diffraction and data collected as described in

Chapter 2.

Appendix A

Supplementary FTIR Diehroism and
X-Ray Diflraetlon Data

103

686 me 00:38»?
:80? 2.0 .30 a e. “cam—ago: 203 0:00am >30 eman0-oo0:o£ 05 one o8.— .«e 005883 :0009. :6
>20 a 9 30208.8: 203 0.500% >20 0305-38.82 .5009 5 05 2 8096.. 53 one Ea coo 3 3.590 ago

220 o3EB$-€-<mE 098:0..00052 ES -383: wcmteasméom he common 320.05 :6 .35 .3. 03mm

 

 

 

 

 

 

 

 

 

000520 9. .....
$.53 39:55.63 000500 om I .780. 8:83:96;
oomm ooom oomm oovm ooom ooom ooom ooom oowm oovm ooom ooom
p..——.—-—..-h——.- v.0- ——.———p~—.—.L.——..— 9.0.
i no .. t... flano
n 3 < .. n to
two two
1 I
[0.0 v 13
n . m n .
[co m rec
”no N. nmd
no.0 m ”0.0
m-.on<& I No I no
$34 no.0 «80% v.3
n00 95.0% nee
F F
092060030: 0920.38.02

 

  

 

 

 

 

eoueqrosqv

104

00:03 003 0900.“ 000002088: 02 .00—02000900 80c
com 00:0 05 8000 y: 05 o. 3306:0900 08—: 05 30:0 «000% 20:26 05. .930 .Um< :0 000253
“WEE EEO—2080008 00-593. 03200000000 0:0 0930.38.50 we comma 9.200000 :6 .~< 20mm

 

0032000900 I.
000600 on E...

 

.780. 9008:0033

88 88 8% 83 800 800

 

bbDb-bDIb-IDDD-DPBD—yrbbh

 

0905000000:

 

 

IIrTIYTUI'lUUUTrVII‘TIUIIIIUIU

 

u)
o.
C?

o
O.
o

I!)
0.
o

O
'-.
O

I!)
".
O

o
9!
o

in
9!
o

eaueqrosqv

$.53 2008:0020?
ooom ooom comm oovm coon

bbbb—bbDI-b-ID-hhhb-Dbh

oomm

 

090:0..0532

 

. mod.

to O I!) O

P. '- 0. O.

O O o O
eoueqrosqv

o
9!
o

UUU'liti‘II'U'lTTU‘lIT‘U‘jjj—t

 

 

ID
9!
o

105‘

.8000 E

0.0 0. 809.0. 5.3.9.0.. 80... com 00.... 000 o. .0.—00.000300 8.... 0.0 3000 8.00% 0.0.00.6 0E. .00—0.0
.0}. 00 00:00:00 08.: 0..00:t¢§008.00.<:2.~ 00.000-000.00. 000 0w..0._0-_08.80 000 .0002...
8000... 0:00 .5. 0000000 0 8 :00 00.00.-00.<:Eo_0 .0 00.00.. 00.00.000-00... 0-0 .3. 080.:

 

032000800 ll
000.000 on E...
$.80. 0.0080008? , - $.80. 0.0080008;

 

oomp o0: com. com. ooh. oomp oomp oov. oomp ooop ooh. com.

 

 

, 2305: l "8.0. 10-85 ll
.00. 000. .00 000.
I \\ , O
000.
000 .80. 00.0 . 03
v 0.0.
_ a. om:
00.: / 0..., 00 a \ 0r.
00“ 003 w 8:, 00.:
. U
0 r, 0. on 00F,
00.000.002.00: «6 00.009.08.02
, 00.0

 

 

 

 

mod-

mod

.6

5‘2
0'
aoueqrosqv

or
o

mud

106

.8000 :0 00. 0. .00000.

0.0.» 00.000.000.00 80.... com 00.... 000 0. 00.00.050.80 8.... 00. 3000 000000 0.0.8.6 00... 000.0
.00... 00 00.80000 08.... 0..00.....08.008-00.<0.2... 00.000-002.00. 000 00.009.08.80 000 8.00000
80.80. 8:00 .5. 00000.0 0 0. 000 00.00. .00-<02... 00. .8 00.00.. 0000.80.00 00-0-08: .04. 080.0

 

 

 

   

 

 

 

 

 

.0_00_0000.00 ll
000.000 00 E...
$80. 0808.80.63 $.80. 0808.80.62.
000 000 000 000 000 000 000 000 0.00 0“...- 0.00 0“... 0.00 000
_- 0.82. .0- 0.0-5.2. . .0-
000 n
o 10
00. 00 00. .
00v . 0 .- .0
00. m 00ml. ..
M 0.00 M mm 0
u 3. 000000200 000. .0 .
\ 000000200 000. .0 m 00.: -50 1-0 H
00.0 -30 1-0 0.0 . 00.0
0.0 .00
00.000.000.80: 00.009.08.02 n
0.0 . 0.0

eoueqrosqv

Absorbance

107

Normal—charge

Reduced-charge

M

1 I l I I I 1
1 800 1 700 1600 1500 1 400
Wavenumber (cm-1)

 

 

 

 

 

 

 

 

Figure A5. C-C ring stretching region of normal-charge (top)
and reduced-charge (bottom) TMPA-montrnorillonite films
supported on AgCl disks.

108

$880-28... 0.00.00. 0.00800 .00... .0....0...88.008 00.8003 000... 00000-000000.
000 0.0.. 00.009.05.00 .00.0.0.00-+0Z 00.038.080.20 .8 0800.00.00... 00.... .0... 0.00.0.

 

 

 

0.05-«000.000

wam-bcmbnbfimbeMDbfiwbb bmhbwDNh-bfihF-bb o
.1 ll ”L l 7.70 ...

7. 7. 9 . l. . c.
o 6.1... O mwmfi W .-
e n
x n
m. .- u
a 0 e n
o I
9 I
u.- n
.Qu I
00800-000000: v n

 

 

0.00....“ 000.000

 

 

 

00.000.08.02

 

 

7811

I I I

II'Ur‘

 

 

ooom.

oooo F

ooom P

oooom

ooomm

Ausuerul

109

.0..00.....08.008
002.. 00...... 0.0.5 00802 05 0.0.. -0000: .0 000.00.00.00 .0.... ...< 200.".

 

 

 

 

 

 

 

 

 

 

 

 

0.00.0.0 000. 000 0800.0 000. 000
0.0002000000000020. 0.0002000000000020
bth-Dbhb-Pbbb-b.bb—PDII-Db-D-th-hh-I-Dbbh o bh-bbhb—hhbh—bbbb-bbhb-PPPD—bhbb-bb-D-Dhbb o

m .- ... Mega-J m .
.- 00 000.9%.“ ..00000 n
.9: 9 x n ..0000
000000 .
. n
.0080 I .0000.
t u 1
000000 0 n
n 0. .0000.
000000..» n
.- 00000. ”.0800
“.0000... m
” wooomw
00.000.080.80 .00: 00000. 00.000.05.02 00...:
00000. 00000

 

 

Ausuarm

Appendix 8
Evaluation of ATR-FI'IR for Determination of Benzene

Sorption Mechanisms from Methanol on Clays

110

MATERIALS AND METHODS

W

Wyoming montrnorillonite (SWy-l) was obtained from the clay minerals repository
at the University of Missouri - Columbia. The d um fraction was separated from the
coarser material by sedimentation. Half of the 4 pm clay was saturated with Na“ and the
other half with Li+ by repeated saturation with the appropriate chloride salt and
ccntrifugation. Excess salts were removed by dialysis until testing negative for Cl‘ by the
AgNO3 test. Reduced-charge clay was prepared using a suspension composed of 50% Na+
saturated and 50% Li+ saturated clay, freeze-drying it and heating 18 horn in quartz
crucibles at 250°C. The collapsed, reduced-charge clay was re-expanded by suspending
the clay in a 70% rrrethanol-water mixture (10 g H) then sonicating 200 ml aliquots for 20
minutes (Green-Kelly, 1953; Brindley and Ertem, 1971; Jaynes and Bigham, 1987). A ‘
Heat Systems, Ultrasonic Inc. model #W-385 sonicator with a 7mm diameter probe tip on
a setting of 7 (out of 10) was used. Re-expansion was verified by X-ray diffraction and
cation exchange capacity measurements.

Organic cation salts were dissolved in methanol and added to the clay in 70%
methanol - water suspensions and agitated for 72 hours. Four times the CBC of
tetramethylammonium (TMA) chloride, ten times the CEC of trimethylphenylammonium
(TMPA) bromide and 1.5 times the CBC of hexadecyltrimethylammonium (HDTMA)
chloride were added to aliquots of this suspension. Separate, freeze-dried reduced - charge
clay was suspended in 200-250 ml solutions of 0.5 M CsCl or MgClz in 70% methanol
and sonicated as above. These were centrifuged and rinsed twice with 0.1M CsCl or
MgClz. All of the clays were dialized against deionized water until testing negative for
halides using AgNO3 then freeze-dried

 

Approximately 50 mg of each clay was measured into 25 ml Corex tubes and
shaken overnight in 25 mls of a 3% solution of benzene in methanol. After equilibration,
the clay was filteredfrom the suspension on a25 mmWhatmann glass fiberfilterand
placed on the surface of the ZnSe crystal of a Harrick single-reflection ATR cell The
sample compartment was scaled and a portion of the supemant solution was injected into
the cell to maintain equilibrium of the clay with the solution. The cell and sample were
placedinto aPerkin-Elmer 17lOFI'IRpurged with dryCOz-frec airwithaDTGS
detector, 2 cm'1 resolution, no apodisation and 500 scans. Spectral collections were
controlled by an IBM XT computer using Spectracalc" software and connected through the
serial port to the instrument. Single-beam spectra of each solute-solvent-sorbent-complex
were collected and ratioed against the single-beam spectrum of the empty cell to obtain
transmittance and absorbance spectra of the complex.

The absorbance spectrum of methanol injected into the ATR cell was collected and
subtracted from all clay-benzene-methanol spectra in an effort to isolate absorption bands
due to the sorbent or the benzene and remove interference due to the solvent. The positions
and intensities of these bands were observed and compared to literature values to determine

whether interpretations of the sorption mechanism of benzene on the clays could be made.

$0.0 0205.00-39: co 3:052: E 25003 8m 0 000 .3552: 5 2.00:3 00m m 42.052:
01.5 >03 20 < 3.2.0— 0buoam .o:=o=to§:o_e 22.0% 83080803508538.0000:
0900063002 me 058% @8805 05 0o coup—om 3:052: 0:0 28003 «o 88.5 .E Bani

112

:23 assuage?

can one cow own

bbbb-nnbh—hhb-hLb-h

 

 

J

 

/\

Now

 

 

 

eoueqrosqv

AVE": 3353:0203

com 099 Go: comp comp cot. comp 002 80a

bb-b-bbbhbbbbb-bbbb-Dhbb-hhbn-Pb

 

one F mv F

53

 

use p

 

 

eoueqrosqv

115

0003 520080000 5 3:053. 5 050:2.

fim U 5:0 3:052: 5 0:00:03 00m m 3:052: :m 25 >03 20 .0. 522.0— 0.00on 6033:2558
-5 00000958002 .3 0.08% 020.55 05 :o 5538 3:052: 5:0 33:2. .3 .8bm .vm 8:me

coo

 

 

 

:53 00:80:26.5
one can own 80
P P DL \- I b L b - P b I n - D b h b
<
w
m s
m
m
U
000 a
000
o

 

 

AVE": 3:80:30;

bib—bbPP-bbPP-FDFb-Dbbb-bbbb-bhbh

 

 

we:

use 0

 

 

009 83 89 009 co: comp comp ooom

eoueqrosqv

116

.03 5200200-»: :0 3:052: 5 0:00:09

93 0 5:0 3550:. 5 0:00:00 00m m 3:052: 5 3:0 003 20 < 50.2.0— 00000m 2305:2558
-02 000050500050: 3 0.000% 50.055 05 :0 350.30 3:052: 5:0 0:00:09 3 Seam .nm 02am

 

 

$.80. 00:50:03;
000 0.00 own 0.00 80
<
J m
000
000
o

 

 

 

eoueqrosqv

$.83 0350:0203

hD-bPhD-bIbD-DI-D-thP-D-hI-bbhb

 

E:

53.

 

 

 

009. cc: 08.. co? co: comp comp ooow

eoueqrosqv

Appendix C

Supplementary ATR-F118 sorption spectra

3:052: 5 0:039 00m 3 8.500% 00:000.:3 ~50. .6 0.52%

I b

coop

b

2283 20:50:03;

.P

ommp

h

omvp

b

owe?

camp

b

ooom

 

117

L69

 

 

 

 

 

SVOL

ZQOL

mOOAY

«cog?

POOAY

..
c:
‘2
c:

needy

TYTIIUITj‘T1i1UTT'T

I

I0

8.

O
eoueqrosqv

U

1

w 000.0
.u 000.0
.. :00

m0 .00

 

mpoxv

 

118

com .
I _ b

own

3:052: 5 0:00:05350 fin 3 8.500% 8:20.203 ”ER .NU 0.5mm..—

b

coop
. . .

$253 0:380:03;

oo.~ — 00.5 _. 0mm P can F

b b

 

 

 

999

LVL

 

00L

VLL

£06

996

 

 

1
r

1

t

i
t
I
I
1

T

IUY'II

 

 

0000
. . . 000.0-

01 mood-

.
.I 60.0-

Food

.0000
w 000.0

1
H 5006

wood

:06

m 5.0

m 56

eoueqrosqv

119

 

cow
.

L89

3:052: 5 0:00:23320 fin 3 8.500% 00:20.55 «.5. .n0 0...»...—

AF.<.:0. 0.355053

0mm 00.0 . oo.~ _. co... P 2% F 0mm.-

D b

£06

980
Vl-ll»

 

 

30L

 

 

0000
. . - 000.0-

f
.1 08.0-

“-.0000
”-.00..
_-. 000.0
.1 000.0
j
.u 000.0
.

.
”- 000.0
”-..0..
”-0.0..

I

. 0.0.0

 

 

eoueq-Iosqv

 

 

120
389

 

96L

3:05.05 5 0:35.353 fin 3 5.500% 00:20.05 «.5. .50 0.005

398

$2.50. 0.2.5.3052.
00.5 .

00.0 F oo.N P

 

 

 

 

 

098$

owe _.

 

 

L

can _.

0000 .
. 000.0-

.
.0

rmood

o 5.0

to
P
O.
o

UU'UTjTVU'UII

Home...

.0000

ii I

 

 

omod

eoueqrosqv

121

000
D - I

3:052: 5 05.30 fin 3 5.500% 00:20.53 .0.-5.. .no 200E

$450. 0.2.8.3303

owe 00.0 _. 00W .- oo.5 P owe P 2.6 w ooow

F + D b b -

 

 

 

 

969

891.

. o

moo-o

03.0

009 l
909i
u)
5.
O

o
N
0.
o

to
cu
0.
0

one...

mmod

TTIjIIU‘U'UIU'IT'U‘IUIIII‘VUY'UrUYTUWUI

 

 

 

 

O5o.o

eoueqrosqv

122

 

 

 

 

SLB

405508 E .0023 8m 00 8.500% 00000006 ER .00 0.5me

2.3500 0.0083263
08.. 0mm P 00.v _.

. b _ . . 0m0_. 0&2 000m

. . N000-

b b b

O

«00.0

 

 

@000

MB l
909 l
809

000.0

 

 

000.0

£69 l

090.0

N 5.0

30.0

05.0

88a:
8U”.
UII'V'IIUUU'IIUITUIIUUTIUUI'U‘U'UjleUU

 

 

 

 

05.0

eouaqmsqv

123

000

0.820.855... 0:59;.» 00.88%. <2...

00.20.000.600 :0 35.505 :. 0:022 00m .00 A3853... 3.8525 5.500% 00:30.00. #5. .8 05.0.".

000
. .

h

030 F

5

$6500 0:353:05;

00.N F

b

2va

000 F 0mg. 000m

b b L b b

 

 

 

P

P

b

 

 

 

E

 

 

00v P

 

 

aaueqjosqv

124

000

0.5250555 0550»? 00.850052.

0002900050.. :0 55.505 5 0:020. 00m 00 80.055... 65.505 5.500% 00:80.05 E< .00 050E
$6500 2005.50.63

0mm

00.0 P

0mm P

0&0. P 0J0 F 0&9 000m

 

 

 

 

 

D

 

000

 

vwm

 

333.5;

00v Fl

 

 

:3.

00¢ F

 

 

aouaqaosqv

125

000

D h

000
b - I

h

000 F
I .- D

b

0an F

0050:005505 0550»? 00.558-29.—
00350600002 :0 .05505 5 0:030. 00m 00 80.055... 35.505 5.500% 00:80.05 .E.< .00 0.50....
.550. 9.0050553

00.v F

000 F
I .— b

h

0&0 F

b

000w

 

 

 

 

 

 

 

FNOF

th F

 

 

aaueqlosqv

126

000

000
. . .

0&0F

2505.55.55 0550.»? 02550.30 00550
600.50. :0 .05505 5 0:059 00m ..0 0.0.0505 .055050 558% 00:80.05 #5. .08 055...—

$6500 0.350553

0an F

b

08F
L b D

b

00.0 F

b!

0&0 F

000m

 

 

 

 

N00

005

03

h

 

.5

 

N03

 

 

aouaqjosqv

127

000

000
. . .

00.0 F

b

0mm F

0.5055555 0550.»? 50.50.00.+002
00550500500. :0 .05505 5 0:050. 5m ..0 90.055... .05505. 5.500% 00:205.. «.5. .. .0 0.50....
$2.50. 2005:0053

0m¢F

000F

b

0&0 F

000w

 

 

b

I

 

 

 

F0:

 

 

000an osqv

128

dag—E08058 ugh—03$ :0.§00-<$F 0920
600.60.. :0 8550.: :m 0:022:28 fin .«0 8300.008 3550.5 8.5009. 00:20am. Eh .NFU Baum

$683 20:50:05.5
000 cm: oooF ommF omcF

. cow F can F ooow

b P

 

D D

km:
N FF

 

 

 

 

 

 

 

aauaqmsqv

129

.0._:0=..08.:0E «£50»? 00.0.0.2-..EZH 0920
600.60.. :0 .0550... :_ 0:00:3350 00m ..0 80.00.50... .0552... 8.500% 00:80.00. “5. .20 0...»...—

AF.<.:00 23:55.62.

 

 

 

 

 

000 own 0&0 F 00.N F 00.... F 00.0 F 0&0 F 000m
v00 F

N00 09. F\ —

FD. F

nah
i a.
0 F0
00v F
000
«N0

 

 

 

 

 

aouaqxosqv

130

000

000
D — D

. 2w...

00.N F

.0...:0.....0....:08 5.89.3 60.053-29. 005:0
600.60. :0 .05608 :. 0:082:58 fin .0 60.050... .0565. 8.500% 00:80.0... «.5. .30 0...»...

2280. 0.0080552.

0a.;

0mm F

0&0 F

D

000m

 

 

 

D

 

 

mow F
me F

 

000 F

#3. F

 

005 F

 

 

aauaqxosqv

131

000

0mm

00.0 F

D

00NF
. F

r

.0..:0.....0:..:08 3.89.3 60.55.-+m0 098.0
600.60. :0 .05608 :. 0:082:50 a...” .0 80.05:... .05608. 8.500% 00:80.0... ”:2. .20 0...»...
$2.80. 0.0085053

D

on“... F

0J0 F

b

can F

000w

 

 

 

h

 

 

00v F

 

 

aouaqsosqv

132

000

P

000
D - D

D

00.0 F

2280. 0.38.8053

00.N F

.0..:0....08.:08 »:.80.¢$ 60.5.8-3»... 0»5._0
600.60. :0 .05608 :. 0:052:50 .0» .0 80.0.50... .05608. 8.500% 00:80.0... a...» 0.0 0...»...

0m:

000F
. .

DL

000F
. .

D

000w

 

 

 

 

05m F

 

 

 

eauaqmsqv

133

.0..:0....08.:08 »:.80..3 3.58-<2... 0»5..0
600.60. :0 .0560... :. 058380.50 8m .0 60.0.56... .0562... 8.500% 00:80.0... M...< .20 0...»...

$280. 0.0080552.

00.N F

00% F

000 F on» F 000w

. p p — n p h -

 

~00

 

00..

N0»

«N0

 

 

5

 

 

00.. F

 

 

aoueqmsqv

134

000

0mm

00.0F

00.N F

.0..:0.....08.:08 8.80.3 60.883602... 0»5..0
600.68 :0 .05608 :. 808008020 .0» .0 80.0.56... .05608. 8.500% 0880...... y....< .».0 0...»...
$680. 0.0080552.

00... F

00.0 F 0an F 0000.

 

 

D

 

 

 

 

 

00v F

3.3.3}?

 

006 F

 

 

aouaqmsqv

135

000

000
. .

0&0F

0mm F

.0..:0....08.:08 8.89.3 3.5.8-29. 0»5..0
600.60. :0 .05608 :. 0588.80.50 .0» .0 30.0.56... .05608. 8.500% 0880...... MB... .30 0...»...

$680. 0.38.8052.

L

00.0 F

D

00.0 F

0mm F

000m

 

 

 

D

L

D

 

 

FBN F

 

 

 

«he F

 

 

aauaqxosqv

136

com

com
_ .

000 F
D h D

00W F

000050080008 @000»? vegamia 09000
600008 00 300508 0_ 00000300020 00m 00 80.00.0000 30050.5 00500.00 30800.50 yak .08 0.52"—
$6003 2000000053

00% F

00.0 F

D

oomF
D L D

ooow

 

Bo

 

D

man

D

«on

mum

 

 

 

vmv F

 

 

OOUBQJOSQV

137

com

own oo.o F

D

DL

oowF
L .

D

000050000008 N008»? 0205.00.00: 03000
6000000 00 00000008 00 0000000000000 8m .00 600000000 .00—000005 8.500% 0000000000 «.5. .08 0.0.me
$4.000 200050033

oo.v F

09.0 F

can F

D

ooow

 

 

D

ow

D

D

nah

me

mum

 

 

 

vmv F

 

 

aauaqaosqv

138

000

0000000000008 00:00.»? 030088-55. 000000
6000000 00 00000.08 .0 00000000000 03 .00 €200.53 0000.00.00 8.500% 000800.000 MHZ .30 00000"—

AFéEov 0000500053
00.NF on“:

own 0&0 F 00.0 F 0&0 F 000w

 

 

 

)5

 

 

 

 

 

 

000 F

 

meF

 

0mmF

 

 

 

aouaqmsqv

139

000

000
D b

P

D n

00.0 F

00W F

0.000.000.5008 M0080»? 00.0038-<02H 000000
6000000 00 3000008 0_ 00000000500 fin .00 60.00.0000 30000080 0050000. 000000.90 MHZ .30 00030

22.000 0.008000%?

00.0 F

00.0 F

D

0mm F

000m

 

 

 

 

 

 

Nov F

 

 

 

ommF

00¢ F

 

 

meF

 

 

aoueqlosqv

140

0mm

00.0 F

com F
D b! D

000000.255... 00.0.0»? 00.000.00-290 00.000
0.000000 00 0000508 0. 00000000000 00m .00 8200008 0000.00.00 8.500% 000000.000 «.5. .18 0.00.”.
$20.00 0.0005005?

D D

00.3

D D

000m

000 F 000 F
D — D D P D D

 

 

D

D

 

 

 

 

 

mFFF

.

 

 

 

09F

00F. F

 

E1 J

 

 

me F

 

 

aauaqmsqv

141

000

D

D

000
D — D

D

L

0&0 F

D

0mm F

D

.0000=_.0§000. 00.80»? 00.0.0.8.+00 00.000
600.00. 00 00000.0... 0. 0000000050 000 .0 60.00.0000. 00000.25 8.500% 0000.00.00 0.3. .08 0.00...

22.000 200050053
000 F

_L .

00.0 F

D

0mm F

D

0000

 

 

 

 

 

 

 

 

 

omm F

09. F

 

 

meF

 

 

eoueqxosqv

142

000

 

D

000
D - D

 

 

 

 

000000.803... 00.80.03 020.300.3002 00.000
0.00000. 00 8000.08 0. 00000000000 80 00 0.0.0000... 00005000 0050000 0000.00... 0.5. .000 0.00.".

$3.00. 0.0005003;

 

 

 

 

000 F

 

00.0 F

 

 

 

000 F

meF

D

0mm F

0000

 

 

aauaqjosqv

143

000
D - D

00.0 F

D

0000:...08008 00.80.03 020.300-0000...
00009000000. 00 3000.08 0. 000.00 000 .0 00200000.. $000.08. 80.00000 00020.0... 0.5.. .08 0.00.".

$280. 0.0080008;
ommF ova

00.0 F

00.0 F

D

0000

 

 

D

 

 

 

0NF 003

 

 

00v F

FomF

 

meF

 

v00 F

 

 

aouaqmsqv

144

.0..00=..08.008 00.80»? 020.800-...02...
00.000.000.60. 00 00:00.08 0. 000.00 00m ..0 8200.00... 3000.08. 80.00000 0000.0...0 ”F... .08 800...

 

$680. 0.008.830;
000 . . 000 . 000. 000. 000. 000. 000. 0000
00. 000.
0.
000. 000.

 

 

 

 

 

 

 

 

 

 

 

com F

 

 

eouaqmsqv

145

000
D - D

00.0 F

D

00.N F

.0..00=..08.008 00.8002. 00.80.00-<2...Q=
00.000-002.00. 00 8000.08 0. 000.00 00m .0 80.00.0000 8000.080 8.500% 0000.00.00 0.2. .08 0.00...
.580. 8008.82.02.

one F 00.0 F

0mm F

D

000w

 

 

 

D D

 

 

«on F

 

 

vow F
000 F

 

 

 

mow F
FomF

000 F

 

 

aouaqaosqv

146

.0..00:..08.008 00.80.? 00.0..800.+mU
00.000.002.00. 00 8000.08 0. 000.00 000 .0 60.00.0000 8000.08. 8.50000 0000.0...0 0...... .08 0.00.".

 

 

$480. 2008.80.62.
80 0.... 0."... £0. £3 0.... £0. 88
000 00
000. 000.
00...
.0 .8.
000 00.. 000 F

 

 

N...

 

 

 

aouaqmsqv

147

com
...

b

oo.N_.

P

b

0.20—Eggs. ”£8035 B.§8.+~m2
one 0- 0: =0 05 08 E 0:. Eu 0 00b 3. 05 08 E 00% 00:80 . 0.5m.

: 00 .000 _ 5 . ._. «yawn—0mg.»- Erccuwgv .5 $852 :U .m
0mm.

83

— b b

0mm.

b

b

ooom

 

 

b L

 

 

mm _.
0mm P

 

 

mm:

mm:

Pomp

3.6 F

mom _.

 

 

aoueqmsqv

148

com

owe

com:

b

oo_w F

b

0002—0000008 M0000»? 00.00000..<2.H
09000000000. 00 300508 E 30000 00m .«0 6080.003 30050.5 60.0009. 000000.90 MF< .90 0003.0
$6.03 2005:0053

2W0 F

oo.o P

can ..

ooom

 

 

h

h

b

 

 

mam P

 

 

map

3

 

now P

 

 

0mm P
hmv —

 

 

aauaqmsqv

149

000050000008 0000035 00.80.00-002
03000000000.— 00 300.000. 5 30000 fin .00 600000500 30050.5 6.500% 000800.00 ”ER .30 0.0.9.0

$6.00. 00005000203

 

 

 

 

 

 

 

 

80 £0 30. 8.0. 30. £0. 80. 080
000.
000.
an
000 . 000.
000 000.
km 0 0
Po mmv _.
000
000

 

 

 

 

 

aouaqmsqv

150

com

com
D - b

h

00.0 P

0002:8005... 00000.03 000E300-<2.—.e~
002908002 .5 .0552: c. .22... 00 00 .3800... .0552... 838% 8:20.00 0.2 .000 800.0

22.00. 00000500062.
0mm. 0&0. 0mm.

com P
D r b

ooow

 

 

b

P

 

 

 

 

KN P
00m P

mow _.

ham P

02‘ F

 

 

 

ooueqxosqv

151

com

own

6&3.

00.020.58.00... mafia»? 00.00.000.50
03000000000. 00 300.00... 0. .0000.— 00m .00 €088.03 300.0000 5.500% 000000000 «.5. .20 0.0%.".

$2.00. 2008000225

00W P

on“:

0mm F

onwm P

ooow

 

b h

0am

 

 

ow
mm

was

 

mam _.

 

mm.

500 _.

0mm P

 

 

aaueqxosqv

152

00.005.00.000... mac»? 00.0.0000.+0w2
09000000000. 00 3000.0... 0. .0000.— 00m .00 80.00.0000 3000.08. 05.08% 0000.800 Mb... .08 0.09....

 

 

$2.00. 0.0005005;
80 0mm cm... 00.0. 8.... £0. 30. 0000
00.0 .
30 .
00
um.
mm...

 

 

bah

 

 

 

 

 

aouaqjosqv

Appendix D

Supplementary mixed benzene-water sorptlon spectra

153

000000080 .0. 000.00% 0.0 6.000 00.00 0.0 0. 030.0 0.0 0.000%
$0.0 00.0.8025)... c0»... 00.000.002.00. 000 30.. 09000000000 00 0000000 000.00 .0 0000...?
000000800 0000.00-50 :6 0... .0 000.00.. 000 5.0080. 0.0 00 0000.00 .0003 .0 0000...". AD 0.0m...

.008. 2000500002. .200. 2000500003

000 0mm DON own com 80 00 com 0mm 00h 0mm. coo 0mm com

EON—.8 0: .IC o\oc .
0:05.00 0: .IC *6

 

 

mum 1‘ 2000ME02 who 1‘ 0000000“.002

 

 

 

 

 

 

 

 

>05.52... 09009000000: >295): 00.009.00.002

aaueqxosqv

154

000009000 .00 00000.0 .00 .2000 00.00 0.0 0. 030.0 0.0
0.000% .000 00.80.00-002 9005 00.000.002.00. 000 90: 00.000-000.00 00 0000000 000.00 .00 0000.00,
000000.800 0:0—0.00-000 2.0 0.0 .00 002000 000 00000000 00. 00 0000.00 .0003 .00 80.0.5 .NQ 080$
07.0.0. 0.00—0:00.63 “0.0.0. 20005030?
coo one can own com 000 com com one 85 ans can own com

10082302 100823
00.2.

   

3.000
08 :0 .00. ~03 oz

   

I: {can N000

"

000 Ewaoz

 

       

 

IE *0 Neon an
IE o\oo Ncwn 5w
$.8h
O

@985 IE 0:
BN8" . IE .3:

modes
2&0 . IE Q08

8. .5
3K" . IE <on

J 3. .5 IE Qawh

who” . IE o\oNh
v0.00 .

 

 

 

 

 

00.0-5.0: 00.000.800.80 00.0.5.2» 00.000.00.32

155

com cmo con own com omm com com omm con own 000 omm com

$000. 200050303

0000000 00 .IC .0qu

00005000
:0 $000

   
  

 

 

IE 006
000008 .3

000000.000

.0.. 03—0000 50 .00000 00.00 .0 030.0 0.0 03000 00000m .00—0 00.05.00-520. 0005 00.000
000000. 000 e000 0m.000._00000 00 0000000 000.00 00 0000.00» 0000000000 0020.00.50 2-0
05 .00 00000000 00. 00 .000> .0003 00.05.00 0. 0500000 000 0000000.. 0500.3 .00 00000.0 .mD 0503

9.08. 200050303

bub-Lb-bhbhbb-bbbb-DbbD-bbrb

 

  

0000000 00 .Im .80

  
 

02000000
I: 0.00.

 

 

 

 

 

00.0.45; 09000000000:

 

II *0 \
0000000 00w

 

 

00.0.03... 00.009.00.32

aauaqmsqv

156

$.80. 0.0050030;

80 cmo can own 000 one com com one com com com 000 com

0000.30.00
.0.. 00000.0 .00 .2000 00.00 .0 030.0 0.0 03000 000000 .00—0 00.05.000.02... 000... 00.000

-000000. 000 0.0.. 00.009.08.00 00 0000000 000.00 .0 0000.03 0000000000 0:0—0.00-0.0 3-0
00. .0 00000.0. 00. 00 .00? .20.? 020.300 0. 0.000000 000 .00—0.00.. E0000> .0 .00....m .09 300E

€00. 0.00.00.00.63

CbLthbb-Dbbbhb-bb-Db-b-hbbb

 

bbbb—bbDb-bbbb-PILbhbDII-bbbb

 

0000000 00 .10. $0

00.03000
:0 082

 

It o\oo
0000000 .00

 

0000000 00 .Im 08

 

 

 

 

 

00.0000.» 00.000-082.00

02.0.0000
:0 $8.
E::om>
:0 .00
0000000 .00 0.0 00
>0_0.<ns:. 00.009.00.002

oouaqmsqv

"IT'E'Tl'iifi‘lfliflifl"lfflliflmllflfiflfl'flT