we

r V :
fig”... ..
:3}.

.117)?» 1.

h} .. 5|. t.HW.n
3. .. HG 5...:

n .. 1f
1}!
.F. . . “Viagra.
. . . . ....|~ ,i 4I1
P???” i.»...
J 3.).39. I. t; ‘
gun

in.

hid

..
M1

.3 I

.. :.

. .i k...- l
0..

4!. f

.. fen
.1

 

 

y , ‘ 9" C. 1,
. {uh A mwwmm. 3&3...

3...: ‘

a ."yfig

mm

lllllllllllllllllllllllhllllllllllllllllllllllllllllllllll

3 1293 01567400

This is to certify that the

dissertation entitled

Sorption Characteristics, Mechanism(s), and
“v-1ace heterogeneity of Organo-Clays

presented by

Guangyao Sheng

has been accepted towards fulfillment
of the requirements for

Ph.D. . Soil Chemistry
degree In

 

 

 

Major prgf 0

Date W

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY

Michigan State
University

 

 

 

-‘~ "“ "‘T'du)
rues u RETURN aox to remove this mum yoUr more.

TO AVOID FINES Mum on Of Moro dd. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Afflnnativo Adlai/Equal Opponunity Intuition
WINS-9.!

Sorptlon Characteristics, Mechanism(s), and
Surface Heterogeneity of Organo-Clays

By

Guangyao Sheng

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Sciences

1996

ABSTRACT

Sorption Characteristics, Mechanism(s), and Surface
Heterogeneity of Organo-Clays

by

Guangyao Sheng

The sorptive characteristics of organoclays were compared in terms of magnitude
and mechanism to those of natural soil organic matter. It has been demonstrated that
multimechanisms control the sorption of aqueous phase neutral organic contaminants
(NOCs) by organoclays including solvation of the cationic ammonium centers, the alkyl
chains of HDTMA and the mineral surfaces, and solute partitioning. In contrast, solute
partitioning appears to be the singular mechanism for the sorption of aqueous phase
NOCs by natural soil organic matter. Organic contaminant sorption by the surfactant
derived organic matter of organoclays is substantially higher than by natural soil organic
matter due to the existence of multiple sorptive mechanisms for organoclays and the
greater solvency of their organic phases.

Sorption of NOCs by organophilic organoclays is either enhanced or depressed
by the presence of the second solute depending on the solubility parameters of both
solutes. Aliphatic solutes either increase or decrease the solvency of the surfactant
derived organic matter for NOCs. Aromatic molecules expand the swelling organoclays

by solvating the cationic centers and alkyl chains of surfactant cations and the mineral

surfaces. They also increase the solvency of the organic phase of organoclays for NOCs.
It appears that sorption of multiple organic contaminants in complex environmental
systems by organoclays generally produces a synergistic effect.

Organic cations in adsorptive organoclays (layer silicate clays modified with small
quaternary ammonium cations such as trimethylphenylammonium (TMPA)) take a
partially random distribution in the interlayers of organoclays. These cations function as
non-hydrated pillars and adsorption of organic molecules occurs predominantly on the
adjacent siloxane surface, which is hydrophobic in nature. In mixed Ca/TMPA-smectites,
the presence of TMPA and its partially random distribution result in far less water
associated with the clay. The interspersing of TMPA and Ca2+ ions prohibits the
formation of a network of large hydration shells around Ca” rendering the siloxane
surface available for the adsorption of NOCs. The partially random distribution also
creates an energetically (energy distribution) and structurally (fractal dimension)
heterogeneous surface in organoclays. The effect of such heterogeneity on adsorption of
organic molecules depends on the size of probe molecules. This heterogeneity is largely

generated by the microporosity of organoclays.

To my mother

iv

ACKNOWLEDGEMENTS

I would like to express my deep thanks to my major professor, Dr. Stephen A.
Boyd, for his guidance and for financial support during the course of my graduate study
at the Michigan State University. It has been a pleasure working with Dr. Boyd.

Thanks are given to Dr. Boyd G. Ellis, Dr. Sharon J. Anderson and Dr. Thomas
J. Pinnavaia for serving on my graduate committee and for their time, advice, discussion,
and suggestions.

Many thanks go to Dr. Max M. Mortland, Dr. Cary T. Chiou of US. Geological
Survey in Denver Federal Center and Dr. Shihe Xu of Dow Corning Company for their
research experience, and their valuable discussion and comments on my research and
manuscripts.

I would like to thank all people in Dr. Boyd's laboratory for their help, support,

and cooperation during my graduate study.

TABLE OF CONTENTS

List of Tables ................................................ viii
List of Figures ............................................... x
Chapter 1 Mechanism(s) Controlling Sorption of Neutral Organic

Contaminants by Surfactant Derived and Natural

Organic Matter .................................... 1
Abstract ......................................... 1
Introduction ...................................... 2
Experimental Section ............................... 4
Results and Discussion .............................. 7
References ....................................... 16

Chapter 2 Cosorption of Organic Contaminants from Water by

Hexadecyltrimethylammonium-Exchanged Clays ........... l9
Abstract ......................................... 19
Introduction ...................................... 20
Experimental Section ............................... 22
Results ......................................... 25
Discussion ....................................... 3 1
References ....................................... 39

Chapter 3 Surface Heterogeneity of Trimethylphenylammonium-

Smectite as Revealed by Adsorption of Aromatic

Hydrocarbons from Water ............................ 42
Abstract ......................................... 42
Introduction ...................................... 43
Background and Theory ............................. 46
Experimental Section ............................... 49
Results and Discussion .............................. 52
References ....................................... 68

Chapter 4 Function of Trimethylphenylammonium Cations in

Smectite for Adsorption of Aqueous Organic Compounds . . . . 72

Abstract ......................................... 72
Introduction ...................................... 73
Experimental Section ............................... 76
Results .......................................... 79
Discussion ....................................... 82
References ....................................... 90
Chapter 5 Conclusions and Significance .......................... 93

Sorption of Neutral Organic Contaminants by Organoclays . . . . 93
Application of Organoclays to Pollution Abatement ......... 101

References ....................................... 109

vii

Table 1.1

Table 1.2

Table 2.1

Table 2.2

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Table 4.1

Table 4.2

LIST OF TABLES

Properties of muck soil and HDTMA-treated clays ..........

Organic carbon normalized distribution coefficients of

organic compounds at a relative concentration of 0.4

and their solubility parameters (62) .....................

Properties of HDTMA organoclays .....................

Solubility parameters (6) of organic compounds ...........

Properties of TMPA-smectites ........................

Parameters Qm, q, v for adsorption of aromatic

hydrocarbons in water/TMPA-smectite systems ...........
The characteristics of adsorption energy distribution ........
Molecular lengths of adsorbates and fractal

dimensions of SAC-TMPAs ..........................
Properties of TMPA-smectites ........................
Monolayer adsorption capacities of organic compounds

on TMPA-smectites ................................

viii

7

ll

23

34

51

56

59

65

78

86

Figure 1.1

Figure 1. 2

Figure 1.3

Figure 1.4

Figure 2.1

LIST OF FIGURES

Sorption of aqueous phase organic solutes on a muck

soil and Hexadecyltrimethylammonium(HDTMA)-clays

normalized to organic carbon content ................... 9
Comparison of organic carbon normalized sorption isotherms

of SACH, SAzH, and hexane for chlorobenzene and
trichloroethylene ............................. 10
Sorption-dependent variation of the d-spacings of

SACH and SAzH .................................. 12
Sorption of trichloroethylene by a muck soil (left scale)

and SACH (right scale) in the absence and presence of
chlorobenzene as a competing solute ................... 15
Sorption of trichloroethylene by HDTMA-illite in the

absence and presence of carbon tetrachloride,

nitrobenzene or ethyl ether as a competing solute.

(A): Isotherms of trichloroethylene;

(B): Dependence of sorption distribution coefficient of

carbon tetrachloride and nitrobenzene on equilibrium

concentration of trichloroethylene ...................... 26

ix

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 3.1

Sorption of trichloroethylene by HDTMA-smectite in the
absence and presence of cosolute as a competing solute.
(A): Isotherms of trichloroethylene in the absence and
presence of carbon tetrachloride;

(B): Isotherms of trichloroethylene in the absence and
presence of nitrobenzene ............................. 28
Increase in sorption distribution coefficient of
trichloroethylene on HDTMA-smectite (K,,,) in the presence
of carbon tetrachloride or nitrobenzene (K) with different
concentrations normalized by respective amount of carbon
tetrachloride or nitrobenzene Sorbed (Q).

(A): Trichloroethylene - carbon tetrachloride system;

(B): Trichloroethylene - nitrobenzene system .............. 30
Sorption-dependent variation of basal spacing

for HDTMA-smectite ............................... 30
Comparison of HDTMA phase, hexane and soil organic

matter as partition media for trichloroethylene. (Data

normalized to organic carbon content. Data for hexane from

Sheng et al., 1996; linear line for soil organic matter

calculated by K0c from Garbarini & Lion, 1986.) ........... 33
Adsorption isotherms of benzene and alkylbenzenes by

mixed Ca/TMPA- and TMPA-smectites .................. 53

Figure 3.2

Figure 3.3

Figure 3 .4

Figure 3.5

Figure 3.6

Figure 4.1

Figure 4.2

Figure 4.3

Figure 5.1

Adsorption isotherms of benzene and n-propylbenzene

fitted to the Langmuir equation, the DA equation and

the DR Equation .................................. 55
Adsorption energy distributions of mixed Ca/TMPA-

and TMPA-smectites ................................ 57
Micropore size distributions of mixed Ca/TMPA- and
TMPA-smectites ................................... 61
Dependence of monolayer adsorption on molecular size ...... 64
Dependence of fractal dimension on mean pore size

of mixed Ca/TMPA- and TMPA-smectites ................ 67
Adsorption of TMPA cations by and release of Ca2+

from Ca2*-smectite(lefi), and corresponding basal

spacings (right) .................................... 80
Adsorption of organic compounds by TMPA-smectites

from aqueous solutions .............................. 81
Dependence of monolayer adsorption capacity on

organic carbon content on Ca/TMPA-smectites (square)

and reduced charge TMPA-smectites from

Jaynes and Boyd (1991) (circle) for benzene, ethylbenzene,
n-butylbenzene, and trichloroethylene ................... 87
Enhanced sorption of common groundwater contaminants

following treatment of soil with hexadecyltrimethyl-

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

ammonium. Sorption isotherms of benzene, toluene,

and ethylbenzene from water by untreated and treated

St. Clair (a-c) and Oshtemo (d-f) B, horizon soils are

shown. Reproduced from Lee et al. (1989a) .............. 94
Type III isotherms for trichloroethylene and

double-sigmoid isotherms for chlorobenzene sorption

by low-charge hexadecyltrimethylammonium(I-IDTMA)-

smectite (SACH) and high-charge HDTMA-

smectite (SAzI-I), compared to the linear isotherms for
trichloroethylene and chlorobenzene sorption by a

muck soil. Data from Sheng et al. (1996a) ................ 98
Possible interlayer arrangements of hexadecyltrimethyl-
ammoniumO-IDTMA) and corresponding interlayer

spacings (A). Reproduced from Jaynes and Boyd (1991a) . . . . 100
Simulated transport of 1,2,4-trichlorobenzene

through bentonite liners with and without

0.1% trimethylphenyl(TMPA)-bentonite added.

Reproduced from Gullick et a1. (1995) .................. 104
Benzene transport through landfill liners containing

organophilic bentonite: (A) 88% sand and 12% Na-bentonite;

(B) 88% sand, 8% Na-bentonite and 4% hexadecyltrimethyl-

ammonium-bentonite; and (C) 88% sand, 8% Na-bentonite

xii

Figure 5.6

and 4% benzyltrimethylammonium-bentonite. R is the

retardation factor, v is the linear velocity.

Reproduced from Smith and Jafi‘é (1994) ................. 106
Proposed in—situ modification of aquifer material

to create contaminant sorption zone, and coupled

sorption and biodegradation of organic contaminants

for groundwater cleanup ............................. 106

xiii

Chapter 1 Mechanism(s) Controlling Sorption of Neutral Organic
Contaminants by Surfactant Derived and Natural Organic Matter

Abstract

Layer silicate clays modified with cationic surfactants ("organoclays”) such as
hexadecyltrimethylammonium (HDTMA) are effective sorbents for neutral organic
contaminants (NOCs), and may be useful for in-situ and ex-situ remediation of contaminated
soils and waters. In this study, the sorptive characteristics of organoclays were compared in
terms of magnitude and mechanism to those of natural soil organic matter. The data presented
here support multiple mechanisms controlling the sorption of aqueous phase NOCs on
organoclays including solvation of the cationic ammonium centers, the alkyl chains of
HDTMA and the mineral surfaces, and solute partitioning. In contrast, solute partitioning
appears to be the singular mechanism for the sorption of aqueous phase NOCs by natural soil
organic matter. Organic contaminant sorption by the surfactant derived organic matter of
organoclays is substantially higher than by natural soil organic matter due to the existence of
multiple sorptive mechanisms for organoclays and the greater solvency of their organic
phases.

Key Words - mechanism, sorption, organoclay, solvation, partition.

Introduction

Earlier studies have demonstrated that replacing the native inorganic exchange cations
of clay minerals, soils and subsoils with organic cations of the form [(CH3,)3NR]+ results in
greatly enhanced abilities of the materials to remove organic contaminants from water (Boyd
et aI., 1988a,b,c,° Brixie et al., 1994; Jaynes et al., 1990, 1991a, b; Lee et al., 1989a, b, 1990;
Mortland et at, 1986; Wagner et a1., 1994). The impressive sorptive capabilities of surfactant
modified clays suggested their potential utility for treating contaminated waters, as extenders
for activated carbon, and as components of clay containment barriers, for example, in clay
slurry walls, hazardous waste landfills and petroleum tank farms (Boyd et al., 1991). These
studies also suggested that aquifer materials or subsoils could be modified in-situ via injection
of cationic surfactants to create sorptive zones that could intercept and immobilize advancing
contaminant plumes (Boyd et al, 1988a, 1991; Burris et al., 1992; Lee et al., 1989a).
Coupling contaminant immobilization with in-situ biodegradation would provide a
comprehensive restoration technology to permanently eliminate target contaminants (Burris
et a1., 1992; Nye etal., 1994).

Elucidating the mechanistic function of organoclays as sorbents for organic
contaminant molecules is important for understanding, predicting and maximizing their
sorptive capabilities, and for designing new families of organoclays. In the case where the
organic cations comprising the organoclay contain a large hydrophobic alkyl chain (e.g.,
HDTMA), the interlarnellar spaces of expandable clay minerals (e. g., smectite) are almost
fully occupied by the alkyl chains (Jaynes et al., 1991b). A conglomeration of the alkyl chains

on the clay surfaces and interlayer regions results in the formation of an organic phase that

3
controls the sorptive removal of organic contaminants from water.

In analogy to the mechanistic function of natural soil organic matter (Chiou et al.,
1979, 1983), the sorption of nonionic organic contaminants (N OCs) from water by
organoclays such as HDTMA-smectite was originally viewed as a partitioning process
involving the organic cation derived organic phase (Boyd et aI., 1988b; Jaynes et aI., 1991b).
The partition coeficients for the sorption of aqueous phase hydrocarbon contaminants (e.g.,
benzene, toluene, ethylbenzene) from water over a limited range of solute concentrations
were estimated using a simple linear relationship between the amount sorbed and the
equilibrium aqueous concentration to describe the sorption isotherms, in accordance with the
concept of solute partitioning (Chiou et al., 19 79, 1983). The general observation was that
sorption coefficients for contaminant sorption by organoclays were considerably higher (10
to 30 times) than those for soil organic matter (Boyd et aI., 1988a; Lee et aI., 1989a). This
was attributed to the nonpolar nature of the organic cation (e.g., HDTMA) derived phase as
compared to soil organic matter which contains an abundance of polar functional groups (e.g.,
- OH, -COOH).

In the present study, we have examined in greater detail the sorptive characteristics
of HDTMA-clays for aqueous phase NOCs in an attempt to define the operative sorptive
mechanism(s). Sorption was characterized over a wide range of solute concentrations
approaching their water solubilities. Changes in the d-spacing (dam) of HDTMA-clays due to
contaminant sorption were also quantitated using X-ray difli’action (XRD). The sorptive
characteristics of organoclays were compared in terms of magnitude and mechanism to those

of natural soil organic matter. The data presented here support multiple mechanisms

4
controlling the sorption of aqueous phase NOCs on organoclays including solvation of the
cationic ammonium centers, the alkyl chains of HDTMA and the mineral surfaces, and solute
partitioning. In contrast, solute partitioning appears to be the singular mechanism for the

sorption of aqueous phase NOCs by natural soil organic matter.

Experimental Section

The muck soil used in this study was collected fiom Michigan State University Muck
Farm (Lainsburg, MI). The clays used in this study were illite fiom Fithia, Illinois, low-charge
smectite (mostly low-charge montmorillonite, SAC) from Wyoming (American Colloid
Company, Chicago, IL), and high-charge smectite (high-charge montmorillonite, SAz-l) from
Arizona (the Source Clays Repository of the Clay Minerals Society). The <2 urn clay fractions
were obtained by wet sedimmtation and subsequently saturated with Ca2+ cations by washing
the clay repeatedly with CaCl2 solution (0.1 mol/l). HDTMA bromide (Aldrich Chemical Co.)
was dissolved in warm distilled water and used to prepare HDTMA organoclays. A portion
of clay suspensions containing 25 g of clays was treated by adding HDTMA bromide solution
(0.03 moi/l) in an amount just equal to the cation-exchange capacity of the clays (224
mmolclkg for illite, 900 mmoL/kg for SAC, and 1200 mmoleg for SAz-l). The mixtures
were agitated overnight on a magnetic stirrer at room temperature. The I-HD'I'MA-illite (1H),
HDTMA-SAC (SACH) and HDTMA-SAz-l (SAzH) suspensions were then washed with
distilled water repeatedly until free of bromide ions as indicated by AgN03. The HDTMA
organoclays were then stored in bottles for later use. Organic carbon contents of the HDTMA

organoclays were determined using Dohrrnann DC-l90 high temperature TOC analyzer

5
(Rosemount Analytical Inc., Santa Clara, CA).

Sorption of benzene, nitrobenzene, chlorobenzene, trichloroethylene, and carbon
tetrachloride (Aldrich Chemical Co.) on muck soil and HDTMA organoclays was conducted
using a batch equilibration technique. A weight of 1 g of muck soil, 0.8 g of IR, 0.2 g of
SACH or 0.2 g of SAzH and 25 ml of distilled water were placed in Corex glass centrifuge
tubes. Organic solutes were added neat using a Hamilton microliter syringe to produce a
range of initial and final concentrations. The tubes were capped immediately with teflon-
backed septa. The tubes were shaken for 24 hours on a rotator at room temperature.
Preliminary experiments showed that sorption reached equilibrium within 18 hours. After
equilibration, the tubes were centrifuged at 8000 rpm (RCF=4302 g) for 10 minutes to
separate liquid and solid phases. A volume of 5 nrl of supernatant was extracted with 10 nrl
of carbon disulfide in a glass vial.

Analysis of organic solutes in the carbon disulfide extracts was made by gas
chromatography. A Hewlett-Packard 5890A gas chromatograph fitted with a flame ionization
detector and a packed column (Alltech Co.) was used; N2 was the carrier gas. Peak areas
were recorded by a Hewlett-Packard 3392A integrator, and compared to external standards
to determine the concentrations of organic solutes. The recovery of organic solutes in blanks
not containing organoclays was about 90%. The amount of organic solutes sorbed was
calculated fi'om the difi‘erences between the amount of organic solutes added and that
remaining in the final equilibrium solutions. Sorption isotherms were made by plotting the
amount sorbed versus the equilibrium concentration in solution. All sorption isotherms were

normalized to organic carbon content of the corresponding sorbents.

6

Sorption of trichloroethylene in the presence of a cosolute was performed by adding
chlorobenzene into the glass centrifuge tubes containing either muck soil or SACH, distilled
water and trichloroethylene. A volume of 10 ul of chlorobenzene was added to each tube.
Immediately after addition of chlorobenzene, the experiments were continued following the
procedure described above.

Partition of chlorobenzene and trichloroethylene between hexane (Aldrich Chemical
Co.) and water was measured in glass centrifuge tubes containing 20 ml of water and 2 ml of
hexane. Difi‘erent amounts of chlorobenzene and trichloroethylene were added directly.
Following addition, the tubes were irmnediately sealed, and shaken vigorously for 10 nrin. The
aqueous phase was extracted with CS2. The hexane phase was diluted to a proper
concentration of chlorobenzene and trichloroethylene. The concentration of chlorobenzene
and trichloroethylene in both the hexane phase and CS2 extract was analyzed by gas
chromatography. Hexane-water partitioning was normalized to the organic carbon content
of hexane.

Basal spacings (dam) of HDTMA organoclays, and of SACH and SAzH with
chlorobenzene or trichloroethylene sorbed were determined by XRD analysis. HDTMA clay
suspensions were dropped on glass slides, and immediately covered by teflon tape to prevent
vapor loss. X-ray diffraction patterns were recorded using CuKa radiator and a Philips
APD3 720 automated X-ray difi‘ractometer using an APD3 521 goniometer fit with a 0-
compensating slit, a 0.2-mm receiving slit, and a diffracted-beam graphite monochromator,

fiom 2 to 10°26, in steps of 0.03°20, 1 sec/step.

Results and Discussion

Sorption isotherms of benzene, nitrobenzene, chlorobenzene, trichloroethylene, and
carbon tetrachloride fiom water onto IH, SACH and SAzH, and onto the muck soil are
shown in Figure 1.1. Table 1.1 shows some properties of the HDTMA-clays and muck soil.
The sorption isotherms of all five solutes tested were highly linear for the muck soil. In
contrast, NOC sorption by the organoclays resulted in isotherms with distinct curvature which
differed in shape and extent depending on the specific solute and sorbent. Isotherms of
trichloroethylene and carbon tetrachloride on all three organoclays are type III which is
convex to the abscissa commencing at the origin Sorption of these two solutes is much lower
on SACH than on SAzH and III. The isotherms of benzene, nitrobenzene and chlorobenzene
are more complex with a double-sigmoid shape. At low relative concentrations of these

aromatic solutes, the amount sorbed on SACH is much lower than on SAzH and 1H.

Table 1.1. Properties of muck soil and HDTMA-treated clays.

 

Sample Source Organic C Charge CEC d-spacing
(%) (mole/unit cell) (cmoL/kg) (A)

 

Muck Michigan State 49.0 - - -
Soil University Farm
Illite
IH Illinois 4.96 1.44 22.4 10.0
Smectites
SAzH Arizona 20.7 1.14 120 28.7
SACH American 16.5 0.64 90 21.6

Colloid Co.

 

8

The high linearity of isotherms on the muck soil is consistent with previous
observations and with the concept of solute partitioning into soil organic matter (Chiou et aI.,
19 79, 1983). Similarly, partitioning is apparently the dominant sorptive mechanism for
trichloroethylene and carbon tetrachloride sorption by HDTMA-clays. Here, the much higher
degree of sorption into the HDTMA phase, as compared to soil organic matter, substantially
changed the composition of the phase. This caused the solvency (i.e. the solubilization
capability) of HDTMA phase for these solutes to increase so that the isotherms became
convex to the abscissa, and hence are classified as type III. A similar phenomenon was
observed for the partitioning of trichloroethylene and chlorobenzene between hexane and
water (Figure 1.2).

The solubility parameter (Barton, 1991; Hildebrand et aI., 19 70) of the HDTMA
phase after sorption of solute molecules can be expressed as:

6 =(1-¢2)5m+¢252 (1.1)
where (b2 is the volume fi'action of the solute; 6m and 62 are the respective solubility
parameters of the pure HDTMA phase and the solute. The change in solubility parameter of

the HDTMA phase with solute content is:

d6
— = 6 - 6 1.2
d¢2 2 HDTMA ( )
The term {If— measures the ability of the solute to increase the solvency of the HDTMA
2

phase, and is directly related to the difference of the solubility parameters of the two mixing

phases. The solvency increase of the HDTMA phase is thus dependent on (41} XQXt) where
2

Q is the amount of solute sorbed and f the conversion factor to convert Q in mmng to the

volume fraction. The value of did?- for the solutes used in this study (Table 1.2) indicates that
2

 

 

 

 

o benzene v trichloroethylene
n nitrobenzene 0 carbon tetrachloride
A chlorobenzene
0.5 rl'IIIIVYT'ITIIIIIITTITII III'I—TTYIIIVT'IUIIIUIII- 35
0.4 E. [DUCK SOII IH " j 28
E )7 - " :
0.3 _- y/ . ., , 1 21
~ 1’ /‘ _- ' 3
§ 0.2 [- _/~ 1 14 5
'6 : ‘ ,/‘ . ‘ : g
20.1} J" , {7 a
V _.". :':‘( , 3 (on
E '* m . o g
8 - A
r. 3 3
c - 40
8 : c:
-j 20
I
5 1o
' o
1.0

 

Relative Concentration of Solute

Figure 1.1 Sorption of aqueous phase organic solutes on a muck soil and
hexadecyltrimethylammonium(HDTMA)-clays normalized to organic carbon
content

10

 

 
   

 

50 l. I I I I I I I I I I I I 1 T I I I I I l I I I I .
; o SACH :
a .. Chlorobenzene (9 SAzH .
E 40 :- 0 Hexane 0 -.
E : q SACH ." 2
E, - Trichloroethylene a SAzH q
E 30 f I Hexane 1
<3 20 — _‘
'E . I
3 . .
g 10 I ‘
< . '2
i q
__ : = 5 :

o f..- l l l

 

 

0.0 0.2 0.4 0.6 0.8 1.0
Relative Concentration of Solute

Figure 1.2 Comparison of organic carbon normalized sorption isotherms of
SACH, SAzH, and hexane for chlorobenzene and trichloroethylene

trichloroethylene more efi‘ectively increases the solubility parameter of the HDTMA phase
than carbon tetrachloride, resulting in a type HI isotherm despite the comparatively low

amount of trichloroethylene sorbed (small Q) on SACH.

11

Table 1.2. Organic carbon normalized distribution coefficients of organic compounds at a

relative concentration of 0.4 and their solubility parameters (6 Z).

 

 

 

Organic Distribution Coemcient (ng) a; 55.2 **
Compound IH SAzH SACH (calm/cm”)
benzene 1035.8 1369.7 675.8 9.1 1.1
nitrobenzene 970.6 2119.0 1276.6 10.9 2.9
chlorobenzene 3541.7 6678.6 4174.8 9.6 1.6
trichloroethylene 563 . 5 863 .6 133.8 9. 3 l .3
carbon tetrachloride 1764.8 1551.4 506.1 8.7 0.7

 

* Barton, ARM. Hcmcfiook of Solubility Pammeters and Other Cohesion Parameters.

CRC Press: Boca Raton, FL, 1991. 739pp.
" This term represents the change in the solubility parameter of the HDTMA phase with
solute content. 4), is the volume fi'action of the solute. 6HDTMA is not available and was

substituted for by 6cm,=8.0 calm/cm”.

A comparison of trichloroethylene and carbon tetrachloride distribution coefficients
(Table 1.2) on SACH, SAzH and III (external organic phase only) shows that both molecules
are intercalated in the interlarnellar region of SAzH while they are only partitioned into the
external HDTMA phase of SACH. This results in substantially lower organic carbon
normalized distribution coefficients for SACH, and is apparently due to the smaller d-spaeing
of SACH than SAzH (Table 1.1). XRD shows no increase in the d-spacing of SACH, but a
significant increase in the d-spacing of SAzH, accompanying sorption of trichloroethylene

(Figure 1.3), confirming its intercalation by SAzH. Intercalation of these solutes by SAzH is

12

 

  

 

    

   

 

 

36 L j I—I r r I I f‘ T j I I l I I I I l I T I I q
'1; 32 ~ -
.c ' o SACH I
g Chlorobenzene @ SAzH _
‘31; 28 f Trichloroethylene: 3231-7-
rs : I
m _ .
s 24 — -
U " q
.. A a A j

20 l l l l l l l l l l l l l l I l l l l l l l l l

 

 

0.0 0.2 0.4 0.6 0.8 1.0
Relative Concentration of Solute

Figure 1.3 Sorption-dependent variation of the d-spacings of SACH and SAzH

allowed by its initially greater d-spacing of 28.7A corresponding to a parafi'rn layer structure
of HDTMA (Xu et aI., 1995) as compared to a 21.6A d-spacing in SACH corresponding to
a flat-lying pseudotrimolecular layer arrangement of HDTMA.

We propose that multiple mechanisms are involved in the sorption of aromatic
compounds (viz., benzene, chlorobenzene, nitrobenzene) by HDTMA-clays. Aromatic
molecules, due to their planar shape and delocalized n-bonds, interact strongly with HDTMA,

causing a reorientation of the alkyl chains. These solute molecules solvate the cationic

13

ammonium centers and alkyl chains of HDTMA (Slabaugh et al., 1968; Solomon et aI.,
1983), causing a transition fiom a parallel to more vertical position relative to the silicate
sheets. Solute molecules are concomitantly adsorbed on the vacated mineral surfaces which
are hydrophobic in nature (Jaynes et aI., 1991a). An increase in the d-spacing accompanied
chlorobenzene but not trichloroethylene sorption by SACH (Figure 1.3). Solute partitioning
into the HDTMA phase is also an important concurrent mechanism. The combination of
solvation and partitioning produces substantially higher sorption of aromatic hydrocarbons
than of aliphatic hydrocarbons.

The d-spacing of SACH is initially too small to efi‘ectively intercalate aromatic
hydrocarbons, and hence they are sorbed to a similar extent as trichloroethylene and carbon
tetrachloride. This can be inferred from the low degree of sorption (as compared to SAzH)
at low relative solute concentrations (Figure 1.2). At a relative chlorobenzene concentration
of about 0.2, however, the chemical potential is apparently sufiicient to cause an interlayer
expansion, which can be observed directly with XRD (Figure 1.3). Similar isotherms were
observed for benzene and nitrobenzene, suggesting the same phenomena. Once the interlayers
are open, solvation and solute partitioning together produce high sorption of aromatic
molecules by SACH, similar to SAzH The greater d-spacing of SAzH as compared to SACH
provides immediate access of both aromatic and aliphatic solutes to the interlayer regions,
resulting in swelling of SAzH (Figure 1.3). Partitioning of trichloroethylene and carbon
tetrachloride increases the interlayer volume as well as the solvency of the HDTMA phase,
resulting in a type III isotherm. For the aromatic molecules, the combination of this

mechanism plus solvation of HDTMA and mineral surfaces leads to higher sorption and

14
double-sigmoid isotherms.

In contrast to organoclays, sorption of NOCs by natural soil organic matter is
characterized by highly linear sorption isotherms and similarity in the degree of uptake of
aromatic and aliphatic molecules of similar water solubilities (Figure 1.1). Also, the presence
of a cosolute caused differential effects on sorption by soil organic matter and organoclays.
Sorption of trichloroethylene in binary solute systems containing chlorobenzene is shown in
Figure 1.4. In the case of natural soil organic matter, the presence of chlorobenzene has
absolutely no effect on the sorption of trichloroethylene, consistent with the concept of simple
solute partitioning, and the comparatively low degree of chlorobenzene sorption. In contrast,
the addition of chlorobenzene causes substantially increased uptake of trichloroethylene by
SACH. Solvation of HDTMA by chlorobenzene separates the clay layers (as described
above), rendering the interlayers accessible for trichloroethylene sorption. Chlorobenzene
sorption by SACH also increases the solvency of the HDTMA phase for trichloroethylene,
resulting in greater uptake. Similar effects have been observed in cosolute systems containing
nitrobenzene (Sheng et al., 1996).

The sorption mechanisms proposed here for organoclays are consistent with the
partitioning results of chlorobenzene and trichloroethylene between hexane and water (Figure
1.2), the findingsof Slabaugh et. al. (1968; Solomon et al., 1983), and previous statistical
thermodynamic studies on the sorption of gases on polymers (Pyda et al., 1982, 1983, 1993).
The lack of solvation of hexane by chlorobenzene results in lower uptake of chlorobenzene
fi'om water by hexane than by HDTMA-clays where salvation of HDTMA by chlorobenzene

and partitioning both occur. Conversely, partitioning of trichloroethylene into hexane, which

15

 

0.5 h I I I I l I fiI r W I I I l I I I I l I I I I q 10
: Muck 0 without chlorobenzene :
’5 _ D with chlorobenzene r 3,
= 0.4 L . .. - 8 g
o - S ACH A without chlorobenzene -
E E V wrth chlorobenzene : a
V 0.3 — J 6 a,
E : r9 ' : '0‘
'5 I : §
g I: n) - : g
< 0.1 _- - 1 2 e
t _ - :
0.0 . ’1'? r r 1 L1 14 r 1 l r r r 1 1 r r r r d 0

 

 

 

0.0 0.2 0.4 0.6 0.8 1.0
Relative Concentration of Trichloroethylene

Figure 1.4 Sorption of trichloroethylene by a muck soil (left scale) and SACH
(right scale) in the adsence and presence of chlorobenzene as a competing
solute

is a free organic phase, is higher than the sorption of trichloroethylene by HDTMA-clays
(Figure 1.2) in which the HDTMA derived organic phase is somewhat restricted. Swelling of
I HDTMA-clays with a low d-spacing requires the solvation of HDTMA and the vacated
mineral surfaces (Slabaugh er al., 1968; Solomon et a1, 1983). Aromatic molecules (benzene,
nitrobenzene, chlorobenzene) fulfill these requirements, whereas aliphatic molecules

(trichloroethylene, carbon tetrachloride) are not able to adequately solvate HDTMA. It is

16
proposed that aromatic molecules solvating the alkyl chains of HDTMA interact with each
other indirectly, particularly at low concentrations, through the intra-chain conformational
subsystem of the HDTMA chains (Pyda et al., 1982, 1983, 1993), leading to a cooperative
effect. Solvation of HDTMA chains by aromatic molecules also weakens the inter-chain
interactions of HDTMA, resulting in a non-linear swelling efi‘ect. Both cooperative sorption
and non-linear swelling generate a sigmoid isotherm Combination of a sigmoid isotherm fiom
salvation and a type III isotherm from solute partitioning produces the double sigmoid-shaped

isotherms reported here.

References

Barton, A.F.M. 1991. Hancfliook of Solubility Parameters and Other Cohesion
Parameters. CRC Press. Boca Raton, FL. 739pp.

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. Organic
Substances and Sediments in Water, Vol 1. Baker R.S.(ed.) p181-200.

Boyd, S.A., Lee, J.-F., and Mortland, M.M. 1988a. Attenuating organic contaminant
mobility by soil modification. Nature 333:345-347.

Boyd, S.A., Mortland, M.M., and Chiou, C.T. 1988b. Sorption characteristics of
organic compounds on hexadecyltrimethylammonium-smectite. Soil Sci. Soc. Am. J.
52:652-657.

Boyd, S.A., Sun, S., Lee, J.-F., and Mortland, M.M. 1988c. Pentachlorophenol
sorption by organo-clays. Clays and Clay Miner. 36: 125-130.

Brixic, J.M., and Boyd, SA. 1994. Treatment of contaminated soils with organoclays to
reduce leachable pentachlorophenol. J. Environ. Qua]. 23: 1283- 1 290.

Burris, D.R., and Antworth, CF. 1992. In-situ modification of aquifer material by a
cationic surfactant to enhance retardation of organic contaminants. J. Contam. Hydrol.
10:325-337.

17

Chiou, C.T., Peters, L.J., and Freed, V.H. 1979. A physical concept of soil-water
equilibria for nonionic organic compounds. Science 206:831-832.

Chiou, C.T., Porter, P.E., and Schmedding, D.W. 1983. Partition equilibria of nonionic
organic compounds between soil organic matter and water. Environ Sci. T echno].
17:227-231.

Hildebrand, J.H., Pransnitz, J.M., and Scott, EL. 1970. Regular and Related
Solutions. Van Nostrand-Reinholdt. New York. 228pp.

Jaynes, W.F., and Boyd, SA. 1990. Trimethylphenylammonium-smectite as an efi‘ective
adsorbent of water soluble aromatic hydrocarbons. J. Air Waste Manag. Assoc. 40: 1649-
1653.

J aynes, W.F., and Boyd, S.A. 1991a. Hydrophobicity of siloxane surfaces in smectites as
revealed by aromatic hydrocarbon adsorption fiom water. Clays and Clay Miner. 39:428-
436.

Jaynes, W.F., and Boyd, S.A. 1991b. Clay mineral type and organic compound sorption
by hexadecyltrimethylarnmonium-exchanged clays. Soil Sci. Soc. Am. J. 55:43-48.

Lee, J.-F., Crum, J.R., and Boyd, S.A. 1989a. Enhanced retention of organic
contaminants by soils exchanged with organic cations. Environ. Sci. Technol. 2321365-
1372.

Lee, J.-F., Mortland, M.M., Chiou, C.T., and Boyd, S.A. 1989b. Shape-selective
adsorption of aromatic molecules from water by tetramethylammonium-smectite. J.
Chem. Soc. Faraday Trans. 1 85:2953-2962.

Lee, J.-F., Mortland, M.M., Chiou, C.T., Kile, D.E., and Boyd, S.A. 1990.
Adsorption of benzene, toluene, and xylene by two tetramethylarnmonium-smectites
having different charge densities. Clays ans Clay Miner. 38:113-120.

Mortland, M.M., Sun, S., and Boyd, S.A. 1986. Clay-organic complexes as adsorbents
for phenol and chlorophenols. Clays and Clay Miner. 34:581-585.

Nye, J.V., Guerin, W.F., and Boyd, S.A. 1994. Heterotrophic activity of
microorganisms in soils treated with quaternary ammonium compounds. Environ. Sci.
Techno]. 28:944-951.

Pyda, M., and Kunynski, M. 1982. Theory of sorption of gases on polymers. 1.
Conformational effects and the double-sigmoid shape. Chem. Phys. 67:7-16.

18

Pyda, M., and Kunynski, M. 1983. Theory of sorption of gases on polymers. II. Effects
of inter-chain secondary bonds. Chem. Phys. 79:219-224.

Pyda, M., and Lopes-Garcon, RI. 1993. Theory of sorption of gases on heterogeneous
solids-polymers sorbents. Langmuir 9:2676-2681.

Sheng, G., Xu, S., and Boyd, S.A. 1996. Cosorption of organic contaminants from water
by hexadecyltrimethylammonium-exchanged clays. Water Res. 30: 1483-1489.

Slabaugh, W.H., and Hiltner, RA. 1968. The swelling of alkylammonium
montrnorillonites. J. Phys. Chem. 72:4295-4298.

Solomon, D.E., and Hawthorne, D.G. 1983. Chemistry of Pigments and Fillers. John
Wiley & Sons. New York. p136-140.

Wagner, J., Chen, H., Brownawell, B.J., and Westall, J.C. 1994. Use of cationic
surfactants to modify soil surfaces to promote sorption and migration of hydrophobic
organic compounds. Environ. Sci. T echno]. 28:231-237.

Xu, S., and Boyd, S.A. 1995. Cationic surfactant adsorption by swelling and non-
swelling layer silicates. Langmuir 11:2508-2514.

Chapter 2 Cosorption of Organic Contaminants from Water by
Hexadecyltrimethylammonium-Exchanged Clays

Abstract

The characteristics of organoclays as sorbents for neutral organic contaminants
(N OCs) in aqueous systems containing multiple solutes were evaluated by measuring the
sorption of trichloroethylene on hexadecyltrimethylammonium (HDTMA) exchanged illite and
smectite in the absence and presence of carbon tetrachloride, nitrobenzene and ethyl ether.
Trichloroethylene partitioned into HDTMA-derived phases present on the external surfaces
of the clay particles manifesting type III isotherms. Carbon tetrachloride functioned to
increase the solvency of the external HDTMA phase, and hence enhanced sorption of
trichloroethylene. Nitrobenzene solvated HDTMA resulting in a more vertical orientation of
the C-16 alkyl chains and interlayer expansion. This rendered the interlamellar region of
HDTMA-smectite accessible for trichloroethylene sorption, and nitrobenzene sorption
simultaneously increased the solvency of HDTMA phase for trichloroethylene. Ethyl ether
sorption suppressed the uptake of trichloroethylene by decreasing the solvency of HDTMA
phase for trichloroethylene. Sorption of trichloroethylene also facilitated sorption of carbon
tetrachloride and nitrobenzene. The results suggest that multiple organic compounds generally
produce a synergistic effect on the uptake of NOCs from water by HDTMA-clays, and
support multiple mechanisms controlling the sorption process.

Key Words - Cosorption, organic compound, organoclay, mechanism, solvency, solvation.

19

20

Introduction

The errviromnent at the clay mineral surface is hydrophilic in nature due primarily to
the hydration of inorganic cations on the exchange sites. As a result, clays are ineffective
sorbents for neutral-organic compounds (N OCs) in the presence of water. Organoclays can
be formed by replacing natural inorganic cations on the surfaces of clay minerals with organic
cations via simple ion exchange reactions (Xu and Boyd 1995). The exchange reaction is
favored by van der Waals interaction between organic cations and the reduced solvent
shielding of the ions in the interlarnellar environment. Organoclay surfaces are hydrophobic
and efi‘ectively sorb a wide variety of organic compounds from water and air (Barrer and
Perry, 1961; Boyd et al., 1988a; Mortland et al., 1986). This is due to the lesser tendency
of organic cations to hydrate, and to their inherent organophilic nature.

Organoclays show several promising applications in pollution prevention and
environmental remediation including the treatment of waste efiluents, as extenders for
activated carbon and as components of clay barriers, for example in clay slurry walls,
hazardous waste landfills and petroleum tank farms (Boyd et al., 1988b, 1993; Jaynes and
Boyd, 1990, 1991a, 1991b; Lee et al., 1989a; McBride et al., 19 77). Recent studies have
also investigated the concept of in-m'tu soil modification using quaternary ammonium cations
such as HDTMA to enhance the sorptive capabilities of soils for aqueous phase organic
contaminants (Boyd et al., 1988c; Burris and Antworth, 1992; Lee et al., 1989b). These
studies demonstrated that HDTMA-derived organic matter in soils is 10-30 times more

effective than natural soil organic matter for sorbing organic contaminants fiom water, and

21

that soil modification greatly attenuates the mobility of organic contaminants in soils.
Coupling contaminant immobilization with in-situ biodegradation would provide a
comprehensive restoration technology to permanently eliminate target contaminants (Burris
tmd Antworth, 1992; Nye et al., 1994; Crocker et al., 1995).

The mechanisms controlling sorption of organic compounds on organoclays are
dependent on the type of organic cations forming organoclays (Boyd et al., 1993). We have
used quaternary ammonium cations of the general form [(CH3)3NR]*, where R is an aromatic
or alkyl hydrocarbon group, to replace inorganic cations on smectite clays, and evaluated their
sorptive properties for common organic contaminants (Boyd et al., 1988a, 1988b, 1993;
Jaynes and Boyd 1990, 1991a, 1991 b; Lee et al., 1989a; Mortland et al., 1986). When R
is a relatively small group (e. g., R=-CH3, -C6H,) the organic cations in the interlarnellar
region of smectite are isolated fi'om each other. Such organoclays have relatively high surface
areas and display characteristics of surface adsorbents. For example, adsorption of organic
compounds is often characterized by type I isotherms, strong competitive effects in
multisolute systems, and molecular shape selectivity (Boyd et al., 1988a; Jaynes and Boyd,
1990, 1991a; Lee et al., 1989a). Alternatively, when R is a large alkyl group (e. g.,
hexadecyltrimethylammonium (HDTMA), R=-C 16H33), the resultant organoclays have low
surface areas and act as partitioning media in the sorption of organic contaminants fiom water
(Boyd et al., 1988b; Jaynes and Boyd 1 991 b). Sheng et a]. (1996) recently proposed
multiple mechanisms controlling sorption of organic contaminants by HDTMA-smectites. In
addition to solute partitioning into the HDTMA-derived organic phase, solvation of HDTMA

and the mineral surfaces may occur, manifesting complex double sigmoid shaped isotherms.

22

In this paper, we investigate the sorption of trichloroethylene from water by HDTMA
organoclays in the absence and presence of carbon tetrachloride, nitrobenzene and ethyl ether
as cosolutes. The objective was to determine the sorptive characteristics of HDTMA
organoclays in the presence of multiple organic contaminants, and elucidate how the
interactions between sorberrt and sorbate influence sorption efi‘ectiveness of the organoclays.
Our results show that sorption of multiple organic compounds on HDTMA organoclays may
produce a synergistic or antagonistic effect, and supported multiple mechanisms involved in

the sorption process.

Experimental Section

The clays used in this study were illite from Fithia, Illinois and smectite (mostly low-
charge montmorillonite) fiom Wyoming (American Colloid Company, Chicago, IL). The
<2 um clay fi'actions were obtained by wet sedimentation and subsequently saturated with Ca2+
cations by washing the clay repeatedly with CaCl2 solution (0.1 moi/l). HDTMA bromide
(Aldrich Chemical Co.) was dissolved in warm distilled water and used to prepare HDTMA
organoclays. A portion of clay suspensions containing 25 g of either illite or smectite was
treated by adding HDTMA bromide solution (0.03 moi/l) in an amount just equal to the
cation-exchange capacity of the clays (224 mmoleg for illite and 900 mmoL/kg for smectite).
The mixtures were agitated overnight on a magnetic stirrer at room temperature. The
HDTMA-illite and HDTMA-smectite suspensions were then washed with distilled water
repeatedly until free of bromide ions as indicated by AgNO, These HDTMA organoclays

were subsequemly quick-frozen, freeze-dried, and stored in bottles for later use. Organic

23

carbon contents of these HDTMA organoclays were determined using Dohrrnann DC-190
high temperature TOC analyzer (Rosemount Analytical Inc., Santa Clara, CA). Some

properties of these HDTMA organoclays are given in Table 2.1.

Table 2.1 Properties of HDTMA organoclays.

 

 

HDTMA-clay illite smectite
Source Fithia, IL Wyoming Bentonite"
Charge 1.44" 0.64‘

(molJunit cell)

Cation-exchange

capacity 224 900
(mole/ks)
Organic carbon 4.96 16.5
(%)
d-Spacing 10. lb 21 .6”
(A)

 

a From American Colloid Co.

b measured.

0 Laird et a]. (1989).

Sorption of trichloroethylene (Aldrich Chemical Co.) on HDTMA organoclays was
conducted using a batch equilibration technique. A weight of 0.8 g of HDTMA-illite or 0.2
g of HDTMA-smectite and 25 ml of distilled water were placed in Corex glass centrifuge

tubes. Between 5 and 50 ul of trichloroethylene were delivered directly into the tubes as the

24
neat liquid using a Hamilton nricroliter syringe, and the tubes were capped immediately with
teflon-backed septa. The tubes were shaken for 24 hours on a rotator at room temperature.
Preliminary experiments showed that sorption reached equilibrium within 18 hours. After
equilibration, the tubes were centrifirged at 8000 rpm (RCF=43 02 g) for 10 nrinutes to
separate liquid and solid phases. A volume of 5 ml of supernatant was extracted with 10 ml
of carbon disulfide in a glass vial.

Analysis of trichloroethylene in the carbon disulfide extracts was made by gas
chromatography. A Hewlett-Packard 5890A gas chromatograph was used with flame
ionization detector and a packed column (Alltech Co.) with N2 as the carrier gas. Peak areas
were recorded by a Hewlett-Packard 3392A integrator, and compared to external standards
to determine the concentrations of trichloroethylene. The recovery of trichloroethylene in
blanks not containing organoclays was about 90%. The amount of trichloroethylene sorbed
on HDTMA organoclays was calculated from the difi‘erences between the amount of
trichloroethylene added and that remaining in the final equilibrium solutions. Sorption
isotherms were made by plotting the amount sorbed versus the equilibrium concentration in
solution.

Sorption of trichloroethylene in the presence of a cosolute was performed by adding
either carbon tetrachloride, nitrobenzene or ethyl ether (Aldrich Chemical Co.) into the tubes
containing HDTMA organoclays, distilled water and trichloroethylene. For carbon
tetrachloride, either 5 p] or 15 ul were added, corresponding to an initial relative
concentration (the ratio of the calculated initial aqueous concentration (C) to the solubility

(Cs) of carbon tetrachloride in water) of 0.40 or 1.20 , respectively. For nitrobenzene, either

25

10 pl or 50 ul were added, corresponding to an initial relative concentration of 0.25 or 1.25,
respectively. For ethyl etha, a volume of 50 ul was added, corresponding to an initial relative
concentration of 0.02. Immediately after addition of carbon tetrachloride, nitrobenzene or
ethyl ether, the experiments were continued following the procedure described above. The
aqueous phase equilibrium concentrations of carbon tetrachloride and nitrobenzene were
analyzed by gas chromatography. We were unable to determine the concentration of ethyl
ether by gas chromatography due to coelution of carbon disulfide and ethyl ether. The
amounts of carbon tetrachloride and nitrobenzene sorbed on HDTMA organoclays were again
calculated by difference between the amount added and that remaining in the equilibrium
solutions.

Basal spacings of HDTMA organoclays and of HDTMA-smectite with
trichloroethylene and carbon tetrachloride or nitrobenzene sorbed were determined by X-ray
diffraction analysis. HDTMA clay suspensions were dropped on glass slides, and immediately
covered by teflon tape to prevent vapor loss. X—ray diffraction patterns were recorded using
CuKa radiator and a Philips APD3 720 automated X-ray difii'actometer using an APD3 521
goniometer fit with a 0-compensating slit, a 0.2-mm receiving slit, and a difii'acted-beam

graphite monochromator, from 2 to 10°20, in steps of 0.03°20, l sec/step.

Results
The isotherms for sorption of trichloroethylene by HDTMA-illite were type 111 (Figure
2.1A). In the presence of either carbon tetrachloride or nitrobenzene, sorption of

trichloroethylene was enhanced while isotherms remained type III. The extent of sorption

26

 

.0
co

 

A I I I I I I I I I I I I I I I Ifi -‘
S I O trichloroethylene onIy .
g ; a col, added with initial CICs=1.20 ;
E 0.6 - A nitrobenzene added with initial CICs=1.25 -
V I O ethyl ether added with initial CIC -0.02 1
E I /’/ 2 :
o 0 4 .— " g 2‘
(D . ,l .
E I ' I
= 0.2 — . s , ‘ -
E : .- , = - (A) :
< " ”’53)," :
A 0.0 .. “tr-3'1-

g . n cc:4 with initial CICs=1.20 ..
E - A nitrobenzene with initial CICs=1.25 -
'0 0 2 ' I
iii:5 ' - .3. A +13 A 5% j
,5 0.1 - -
«s I- u:
,o - .
5 - 'll
0

 

 

O

O
h
p
p
h
—
h
l-
b
b
—
p
D
-
b
_
p
h
n
b

 

O
N
A
O)
on

Concentration of Trichloroethylene (mmol/l)

Figure 2.1 Sorption of trichloroethylene by HDTMA-illite in the absence and

presence of carbon tetrachloride, nitrobenzene or ethyl ether as a competing

solute.

(A): lsotherrns of trichloroethylene;

(8): Dependence of sorption distribution coefficient of carbon tetrachloride and
nitrobenzene on equilibrium concentration of trichloroethylene.

27

enhancement was significant; 63% in the presence of carbon tetrachloride and 70% in the
presence of nitrobenzene, at a trichloroethylene equilibrium concentration of 5 mmol/l.
Sorption of carbon tetrachloride and nitrobenzene occurred concomitantly with
trichloroethylene sorption. Although the amount of carbon tetrachloride and nitrobenzene
initially added into each tube was constant, the amount of carbon tetrachloride and
nitrobenzene sorbed on the organoclay, and hence their distribution coeflicients, increased
with the equilibrium concentration of trichloroethylene (Figure 2.1B).

Sorption of trichloroethylene by HDTMA-illite was suppressed in the presence of
ethyl ether as a cosolute while the isotherm remained type 111 (Figure 2.1A). The sorption of
trichloroethylene by HDTMA-illite was decreased by 26% at the trichloroethylene equilibrium
concentration of 5 mmol/l in the presence of ethyl ether having an initial relative concentration
equal to 0.02.

The isotherms for sorption of trichloroethylene by HDTMA-smectite were type HI
(Figure 2.2A). The presence of carbon tetrachloride enhanced the sorption of
trichloroethylene by HDTMA-smectite while the isotherms remained type III. The extent of
enhancement increased with increasing concentration of carbon tetrachloride; 48% and 245%
(calculated at a trichloroethylene equilibrium concentration of 5 mmol/l) in the presence of
carbon tetrachloride having initial relative concentrations (C/Cs) equal to 0.4 and 1.2,
respectively . As with HDTMA-illite, carbon tetrachloride was sorbed by HDTMA-smectite,
and increased with increasing sorption of trichloroethylene (data not shown).

The presence of nitrobenzene dramatically enhanced the sorption of trichloroethylene

by HDTMA-smectite while the isotherms remained type 111 (Figure 2.2B). The sorption of

28

 

A 2.0 I I I T l I I I I I I I I I I I I I I -
g 0 trichloroethylene only 3
E 1.5 D CCl, added with initial CICs=0.4 _
g A 0014 added with initial CICs=1.2 5
E 1.0 ‘ ., .:
8 0.5 ’ 5 e 9 -:
“’ - = = (A):
5 . I - : .. —-- : '2’: -
">'. 0 0 “:3.“ .
g _ O trichloroethylene only :
5 : D nitrobenzene added with initial CICs=0.25 1
E 4 f A nitrobenzene added with initial CICs=1.25 -:
E 2 f j
i * q
< '

 

 

   

 

0

Concentration of Trichloroethylene (mmolll)

Figure 2.2 Sorption of trichloroethylene by HDTMA-smectite in the absence
and presence of cosolute as a competing solute.

(A): lsothenns of trichloroethylene in the absence and presence of carbon
tetrachloride;

(B): Isotherms of trichloroethylene in the absence and presence of nitrobenzene.

29

trichloroethylene was increased by 181% and 563% (calculated at a trichloroethylene
equilibrium concentration of 5 mmol/l) in the presence of nitrobenzene having initial relative
concentrations of 0.25 and 1.25, respectively. Compared to carbon tetrachloride, nitrobenzene
was about twice as effective in enhancing the sorption of trichloroethylene on HDTMA-
smectite.

To quantify the sorption enhancement, the sorption distribution coeflicients of
trichloroethylene in the absence and presence of carbon tetrachloride or nitrobenzene were
calculated. The increase in the distribution coefficient (K-Ku), where K is the distribution
coefficient in the presence of carbon tetrachloride or nitrobenzene and Km the distribution
coefficient in the absence of carbon tetrachloride or nitrobenzene, divided by the amount of
carbon tetrachloride or nitrobenzene sorbed (Q) at the corresponding equilibrium
concentration of trichloroethylene, was plotted versus the equilibrium concentration of
trichloroethylene (Figure 2.3). Sorption enhancement of trichloroethylene by carbon
tetrachloride or nitrobenzene increased steadily with the equilibrium concentration of
trichloroethylene. The extent of trichloroethylene sorption enhancement per unit mass of
carbon tetrachloride sorbed increased when the initial relative concentration of carbon
tetrachloride increased fi'om 0.4 to 1.2 (Figure 2.3A); conversely, it decreased when the initial
relative concentration of nitrobenzene increased from 0.25 to 1.25 (Figure 2.38).

Sorption-dependent variation of HDTMA-smectite basal spacing is represented in
Figure 2.4. The basal spacing of HDTMA-smectite remained approximately 21 .6A with the
sorption of trichloroethylene. Addition of carbon tetrachloride also did not change the basal

spacing or the intensity of diffraction peak. In contrast, addition of nitrobenzene expanded

 

 

 

 

 

0.4 I. I I I I . I I I T l I I I I r I I I j u
3 n CCI4 added with initial CICs=0.4 1
O 3 :A CCl4 added with initial CICs=1.2 _:
E r: 3
2 0.2 :- j
g : U :
3 01 ' ‘
:5, ' t (A) :
a I I
g 0.O_:1‘:%Ji%#:{:#%:}:i::‘
O . D nitrobenzene added with initial Cle=0.2 .
3 0 6 IA nitrobenzene added with initial CICs=1.25 I
ac; E 3
0.4 :- 1
0.2 E- ‘ -§
_ (B) .
C I

0.0 r r r r r r r r r r r 1 I r r r r
O 2 4 6 8

Concentration of Trichloroethylene (mmolll)

Figure 2.3 Increase in sorption distribution coefficient of trichloroethylene on
HDTMA-smectite (Kw) in the presence of carbon tetrachloride or nitrobenzene
(K) with different concentrations normalized by respective amount of carbon
tetrachloride or nitrobenzene sorbed (Q).

(A): Trichloroethylene - carbon tetrachloride system;

(B): Trichloroethylene - nitrobenzene system.

31

 

 

30 T r I I I I I I I I T I I 1 I I I I I fi

0 trichloroeth lene only .

A u CCI4 add with initial CICs=1.20 I
’35 28 A nitrobenzene added with initial CICs=1.25 ".
a i
'6 26 j
as .
8 ..
— 24 1
3 .
g :
22 ‘-

 

 

 

N
0
co

0 2 4 6
Concentration of Trichloroethylene (mmolll)

Figure 2.4 Sorption-dependent variation of basal spacing for HDTMA-smectite.

HDTMA-smectite little if any at low trichloroethylene concentrations and then more sharply
at trichloroethylene concentrations > 4 mmol/l. It was also observed that the intensity of
difl‘raction peak decreased with increasing the sorption of trichloroethylene in the presence
of nitrobenzene. The diffraction peak disappeared at a trichloroethylene equilibrium

concentration of about 6 mmol/I.

Discussion

Recently, we demonstrated that sorption of aromatic molecules (viz. benzene,

32

chlorobenzene, nitrobenzene) by HDTMA-smectites produced complex double sigmoid
shaped isotherms. This was attributed to the occurrence of multiple mechanisms (Sheng etal.,
1996) including solvation of HDTMA, and solute partitioning into the HDTMA-derived
organic phases present in both the interlarnellar regions and on the external surfaces.
Apparently, aromatic compounds, due to their delocalized 1r -bond system, interact strongly
with the cationic ammonium centers and alkyl chains of HDTMA, causing a reorientation of
the alkyl chains to a more vertical position, which may allow solute interaction with the
revealed mineral surfaces. This process results in interlayer expansion of HDTMA-smectites,
making the interlarnellar regions accessible and hence promoting solute partitioning into the
expanded phase. In contrast, aliphatic compounds such as trichloroethylene and carbon
tetrachloride are apparently not able to adequately solvate HDTMA. For such compounds,
the singular sorptive mechanism is solute partitioning into the HDTMA-derived organic
phase.

The results obtained here further substantiate the mechanisms summarized above.
Sorption of trichloroethylene on HDTMA organoclays is due to partitioning into the organic
phase formed by HDTMA in organoclays. The type III isotherms observed, which are convex
to the abscissa commencing at the origin, are due to an increase in HDTMA phase solvency
for trichloroethylene as its content in the HDTMA phase increases (Chiou et al., 1992).
Similar type III isotherms are observed for the partitioning of trichloroethylene between water
and hexane (Figure 2.5).

The change in solvency of HDTMA-derived organic phase resulting from

trichloroethylene uptake can be described using the solubility parameters (Hildeber et aI.,

33

 

N
0
cl

I I I I I I I I l I I I T I I I If

0 HDTMA-smectite
D hexane
----- soil organic matter

A
(It
IIII'IIIrlIIIIlIIII

Amount Sorbed (mmollg)
a. 3

 

lllllllLllllllllllJ

   

 

 

O

O 2 4 6 8
Concentration of Trichloroethylene (mmolll)

Figure 2.5 Comparison of HDTMA phase, hexane and soil organic matter as
partition media for trichloroethylene. (Data normalized to organic carbon content.
Data for hexane from Sheng et al., 1996; Linear line for soil organic matter
calculated by Kw from Garbarini and Lion, 1986.)

l 9 70) of pure HDTMA (£5me and trichloroethylene (62) according to the following

equation:

6 = (1 - 43ng + (1)262 (2.1)
where (I); is the volume fraction of trichloroethylene (Barton, 1991; Hildebrand et al., 19 70).
The change in solubility parameter with trichloroethylene content in the HDTMA phase is:

d6

34

The value of 74:- is the measure of the ability of trichloroethylene to increase the solvency
2
of the HDTMA phase for trichloroethylene, and is directly related to the difference in the

solubility parameters of the two mixing phases. The solvency increase of the HDTMA phase

is thus dependent on ( 7‘?— xQ><t) where Q is the amount of trichloroethylene sorbed and f the
2

conversion factor to convert Q in mmng to the volume fraction. The value of j?— for
2

trichloroethylene (=13 calm/cm”) is greater than zero because 62 is greater than 6mm

(Table 2.2). Obviously, the value of(§£-><th) increases with increasing the sorption of
2

trichloroethylene, resulting in a type III isotherm (Figure 2.1, Figure 2.2).

Table 2.2 Solubility parameters (5) of organic compounds.

 

 

 

organic solubility parameter‘ d5 ( ab , 3‘ )**
compound (calm/cm”) d"2 34’: “a
trichloroethylene 9. 3 l .3
carbon tetrachloride 8.7 0.7
nitrobenzene 10.9 2.9
ethyl ether 7.7 -O.3
HDTMA 8.0 0.0

 

* Barton ARM. (1991).

* * This term represents the change in the solubility parameter of the HDTMA
phase with solute content. (b2 and d), are the volume fractions of the solute
and cosolute, respectively. cm is not available and was substituted for
by 60.11580 calm/cm”.

The efl‘ect of nitrobenzene and carbon tetrachloride on trichloroethylene sorption by

organoclays depends in part on the swelling properties of the clay. Because illite is a non-

35

swelling clay, interlayer expansion of HDTMA-illite due to HDTMA solvation by
nitrobenzene does not occur, and hence the effect of solvation on trichloroethylene sorption
is minimal. As a result, carbon tetrachloride and nitrobenzene both function primarily to
change the solvency of the HDTMA phase of HDTMA-illite for trichloroethylene. Here, the
solubility parameter of the HDTMA phase with trichloroethylene and either carbon
tetrachloride or nitrobenzene sorption follows the equation:
6 =0 - d), - (133)6W + (p262 + ¢363 (2.3)

where (I), is the volume fraction of either carbon tetrachloride or nitrobenzene, and 63 is the
solubility parameter of the corresponding compound. The change in the solubility parameter

of the HDTMA phase with cosorption of trichloroethylene and either carbon tetrachloride or

nitrobenzene is:
ab
— = 5 - 5 2.2
6% 2 HDTMA ( )
ab
— = 5 - 5 (2.4)
a¢3 3 HDTMA
Table 2.2 shows the values of g for trichloroethylene calculated by equation (2.2) and it.
2 3

for carbon tetrachloride and nitrobenzene calculated by equation (2.4). The value of %
being greater than zero indicates that sorption of carbon tetrachloride and nitrobenzene
increases the solubility parameter of the HDTMA phase towards trichloroethylene. The
increased sorption of trichloroethylene shown in Figure 2.1 is therefore the result of this
solvency increase of the HDTMA phase by either carbon tetrachloride or nitrobenzene
sorption and additional trichloroethylene sorption. Similarly, the sorption of trichloroethylene

also increases the solvency of the HDTMA phase for carbon tetrachloride and nitrobenzene

36

due to the fact that the solubility parameters of all three compounds are greater than that of
the HDTMA phase (Table 2.2). As a result, sorption of one compound produces a synergistic
effect on the sorption of the other, and hence the sorption coefficient of nitrobenzene and
carbon tetrachloride increases with the amount of trichloroethylene sorbed (Figure 2.13).
That trichloroethylene is sorbed primarily on external HDTMA phase of HDTMA-
smectite in the absence of cosolutes is supported by the fact that the organic carbon
normalized trichloroethylene uptake is much smaller for HDTMA-smectite than for hexane
(Sheng er al., 1996), yielding a log K“ value of 2.1 (at Ce/Cs of 0.4) which is similar to the
log K“, value of 2.0 reported for soil organic matter (Garbarini and Lion, 1986) (Figure 2.5).
Ifthe interlayer regions of HDTMA-smectite were fully accessible to trichloroethylene, the
trichloroethylene sorption isotherm and log K0° values should resemble those of hexane due
to the compositional similarity of the two phases as partition media, and furthermore, should
be substantially higher than soil organic matter which is a poorer partition phase for NOCs
due to its more polar nature. In addition, we previously observed that trichloroethylene
sorption into the interlayers of a higher charge HDTMA-smectite was accompanied by
interlayer expansion (Sheng er al., 1996). The basal spacing of the lower charge HDTMA-
smectite described here did not increase upon sorption of trichloroethylene, as reflected by
the constant basal spacing of ~21.6A (Figure 2.4). As in the case of HDTMA-illite, the type
III isotherms for trichloroethylene sorption remlted from the solvency increase of the external
HDTMA phase. Addition of carbon tetrachloride as a cosolute with trichloroethylene did not
increase the basal spacing of HDTMA-smectite indicating that it was not intercalated. This

is because carbon tetrachloride interacts only weakly with the HDTMA phase and the

37

molecular size of carbon tetrachloride is larger than that of trichloroethylene. Carbon
tetrachloride increases the uptake of trichloroethylene by increasing the solvency of the
external HDTMA phase of HDTMA-smectite, and this efl‘ect is more pronounced with higher
initial levels of carbon tetrachloride which manifest higher amounts of sorbed carbon
tetrachloride (positive 61:? Table 2.2). Because only the external HDTMA phases of both
HDTMA-illite and HDTMA-smectite are available for trichloroethylene sorption in the
absence and presence of carbon tetrachloride, the increase in sorption distribution coefficient
of trichloroethylene on both organoclays by carbon tetrachloride was within the same range
(0.06-0.30 (l/g)/(mmol/g)).

The addition of nitrobenzene as a cosolute with trichloroethylene causes an increase
in the basal spacing of HDTMA-smectite (Figure 2.4). Random interstratification of
nitrobenzene in the interlayers of HDTMA-smectite seems to occur, based on the diminution
of the original difl‘raction peak (at 21.6 A) and the eventual increase in basal spacing.
Apparently, nitrobenzene solvates both cationic ammonium centers and alkyl chains of
HDTMA, causing a more vertical reorientation of the alkyl chains and interlayer expansion.
Similar results have been obtained with the aromatic molecules such as benzene and
chlorobenzene (Sheng et al., 1996). Interlayer expansion by nitrobenzene renders the
interlarnellar space of HDTMA-smectite accessible to trichloroethylene. The nitrobenzene
activated HDTMA-smectite is thus able to intercalate trichloroethylene with swelling of the
organoclay and the formation of layer complexes in a bimolecular layered arrangement

(Solomon and Hawthorne, 1983), causing the intensity of the diffraction peak to decrease and

the apparent dom spacing to increase. The d001 spacing of HDTMA-smectite in a bimolecular

38
layered arrangement is not able to be determined because it is over the lower angle limit of
the X-ray difi'actometer used in this study. Nitrobenzene simultaneously functions to increase
the solvency of both external and interlarnellar HDTMA phases for trichloroethylene,
resulting in enhanced uptake.

In an attempt to separate the solvation and solvency effect of nitrobenzene on
trichloroethylene sorption by HDTMA-smectite, we have plotted the increase in
trichloroethylene sorption coefficient per unit mass of nitrobenzene sorbed (Figure 2.3B).
This plot reveals that the lower level of nitrobenzene addition is more efl‘ective than the higher
level, in direct contrast to the efl‘ect of carbon tetrachloride (Figure 2.3A) which only
influences solvency. This suggests that the major role of nitrobenzene when added at the
lower level is to solvate the HDTMA. This causes an expansion of the interlayers, rendering
the interlamellar regions available for sorption of trichloroethylene. Apparently, additional
nitrobenzene functions primarily to enhance the solvency of the HDTMA phase for
trichloroethylene sorption which continues to increase (Figure 2.28) since the interlayers are
already sufficiently expanded. As a result, the effectiveness of nitrobenzene on a unit mass
basis is diminished.

In accordance to the above discussion, sorption of ethyl ether, with a solubility
parameter of 7.7 calm/cm”z and :7:- of—0.3 (Table 2.2), should decrease the solvency of the
HDTMA phase for trichloroethylene. In fact, we observed experimentally a decrease in
trichloroethylene sorption by HDTMA-illite in the presence of ethyl ether as a cosolute
(Figure 2.1A).

Most of the common organic contaminants (>95 %) have higher solubility parameters

39
than that of the HDTMA phase. Sorption of such contaminants by organophilic organoclays
like HDTMA-smectite should produce a synergistic, rather than antagonistic (or competitive),
effect on contaminant removal from water. This represents an advantage of organoclays over
conventional adsorbents such as activated carbon or shale where sorption is depressed in the

presence of cosolutes due to competitive effects.

References

Barrer RM., and Perry 6.8. 1961. Sorption of mixtures, and selectivity in
alkylammonium montmorillonites. Part II. Tetramethylarnmonium montmorillonite. J.
Chem. Soc. 850-858.

Barton A.F.M. 1991. Hancbook of Solubility Parameters and Other Cohesion
Parameters. CRC Press, Boca Raton, FL.

Boyd S.A., Sun S., Lee J.- ., and Mortland M.M. 1988a. Pentachlorophenol sorption
by organo-clays. Clays and Clay Miner. 36: 125-130.

Boyd S.A., Mortland M.M., and Chiou C.T. 1988b. Sorption characteristics of organic
compounds on hexadecyltrimethylammonium-smectite. Soil Sci. Soc. Am. J. 52:652-657.

Boyd S.A., Lee J.- ., and Mortland M.M. 1988c. Attenuating organic contaminant
mobility by soil modification. Naturedondon) 333:345-347.

Boyd S.A., Jaynes W.F., and Ross B.S. 1993. Immobilization of organic contaminants
by organo-clays: application to soil restoration and hazardous waste containment. In
Organic Substances and Sediments in Water, Vol. 1., pp. 181-200. CRC Press, Boca
Raton, FL.

Burris D.R., and Antworth GP. 1992. In-situ modification of aquifer material by a
cationic surfactant to enhance retardation of organic contaminants. J. Contam. Hydro].
10:325-337.

Chiou C.T., Lee J.-F., and Boyd S.A. 1992. Comment on "the surface area of soil
organic matter”. Environ. Sci. T echnol. 26:404-406.

Crocker F.H., Guerin W.F., and Boyd S.A. 1995. Bioavailability of Naphthalene

4O

Sorbed to Cationic Surfactant-modified Smectite Clay. Environ. Sci. T echno]. 2922953-
2958.

Garbarini D.R., and Lion L.W. 1986. Influence of the nature of soil organics on the
sorption of toluene and trichloroethylene. Environ. Sci. Techno]. 20: 1263-1269.

Hildebrand J.II., Prausnitz J.M., and Scott, KL. 1970. Regular and Related
Solutions. Van Nostrant-Reinholdt, New York.

J aynes W.F., and Boyd S.A. 1990. Trimethylphenylammonium-smectite as an efl‘ective
adsorbent of water soluble aromatic hydrocarbons. J. Air Waste Manage. Assoc. 40:1649-
1653.

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

Jaynes W.F., and Boyd S.A. 1991b. Clay mineral type and organic compound sorption
by hexadecyltrimethylammonium-exchanged clays. Soil Sci. Soc. Am. J. 55 :43 ~48.

Laird I).A., Scott A.D., and Fenton T.E. 1989. Evaluation of the alkylammonium
method of determining layer charge. Clays and Clay Miner. 37:41-46.

Lee J.-F., Mortland M.M., Boyd S.A., and Chiou, C.T. 1989a. Shape-selective
adsorption of aromatic molecules from water by tetramethylammonium-smectite. J. Chem.
Soc., Faraday Trans. 1 85:2953-2962.

Lee J.-F., Crurn J.R., and Boyd S.A. 1989b. Enhanced retention of organic
contaminants by soils exchanged with organic cations. Environ. Sci. T echno]. 23 : 1365-
1372.

McBride M.B., Pinnavaia T.J., and Mortland M.M. 1977. Adsorption of aromatic
molecules by clays in aqueous suspension. In Fate of Pollutants in the Air and Water
Environments, Vol. 8, Pt. 1., pp. 145-154. Wiley, New York.

Mortland M.M., Sun S., and Boyd S.A. 1986. Clay-organic complexes as adsorbents
for phenol and chlorophenols. Clays and Clay Miner. 34:581-585.

Nye J.V., Guerin W.F., and Boyd S.A. 1994. Heterotrophic activity of microorganisms
in soils treated with quaternary ammonium compounds. Environ. Sci. Techno]. 28:944-
951.

Sheng G., Xu S., and Boyd S.A. 1996. Mechanism(s) controlling sorption of neutral

41

organic contaminants by surfactant derived and natural organic matter. Environ. Sci.
Technol. 30:1553-1557.

Solomon D.E., and Hawthorne D.G. 1983. Chemistry of Pigments and Fillers. Jonh
Wiley & Sons, New York.

Xu S., and Boyd S.A. 1995. Cationic surfactant adsorption by swelling and non-swelling
layer silicates. Langmuir 1 112508-2514.

Chapter 3 Surface Heterogeneity of Trimethylphenylammonium-
Smectite as Revealed by Adsorption of Aromatic Hydrocarbons
from Water

Abstract

Adsorption studies of aromatic hydrocarbons of various molecular sizes on
organoclays in aqueous solution were carried out for characterizing the surface heterogeneity
of organoclays. Benzene, toluene, p-xylene, ethyl-benzene and n-propylbenzene adsorption
by a smectite with five different exchange degrees of trimethylphenylammonium (TMPA)
cations for Ca” was measured. The Langmuir isotherm equation did not adequately describe
the experimental data, especially for small molecules, whereas the Dubinin-Radushkevich
(DR) equation combined with a gamma-type adsorption energy distribution function
described all experimental data well, suggesting the surface and structural heterogeneity of
TMPA-smectites. The calculated adsorption energy distributions indicated that the apparent
heterogeneity depends on the molecular size of adsorbates. Small adsorbate molecules such
as benzene explore a highly heterogeneous surface of TMPA-smectites while large molecules
such as n-propylbenzene detect a relatively homogeneous surface. The surface fractal
dimension was dependent on the extent of TMPA exchange for Ca”. When TMPA content
is less than 75 percent of the cation-exchange capacity of the smectite, the heterogeneity
decreases as TMPA content increases; it increases with TMPA content thereafier. These
results are related to the micropore size distributions of TMPA-smectites. The micropores in

TMPA-smectites are defined by the two serni-infinite aluminosilicate sheets and the interlayer

42

43

cations. The micropore size distributions and hence heterogeneity are created in part by the
inhomogeneity of the charge density of clay surfaces and the tendency for cation segregation
in these systems.

Key Words - adsorption, aromatic hydrocarbons, organoclays, heterogeneity, DR equation,

adsorption energy distribution, fractal dimension, micropore size distribution.

Introduction

Organoclays are formed when inorganic cations such as Ca2+ and Na“ in natural clay
minerals are replaced by organic cations. Such organoclays are efl‘ective sorbents for a variety
of aqueous organic compounds including many common groundwater contaminants (Boyd
et al., 1988a, b, 199]; Jaynes and Boyd 1991a; Sheng et al., 1996; Xu et al., 199_). When
relatively small organic cations, e.g. trimethylphenylammonium (TMPA) cations, are
employed, the organoclays behave with the adsorptive properties, resulting in Langmuir-type
adsorption isotherms (Jaynes and Boyd 1990, 1 991 b). Thus, the Langmuir adsorption
equation is often used to represent this type of experimental data.

Langmuir adsorption theory (Langmuir, 1918) has been extensively used in various
monolayer adsorption systems because of its simplicity of form and directness of derivation.
In most cases, however, difliculties arise in applying the Langmuir equation to describe
practical adsorption systems (Rudzinski and Wojciechowskr', 1993). One of the major
assumptions underlying the Langmuir adsorption equation is that the solid surface is
energetically homogeneous (Langmuir, 1918), which is often not true, especially for naturally

occuring solids such as clays. It is now well known that geometric and energetic

44
heterogeneity is a fundamental feature of many solid surfaces (Rudzinski and Wojciechowski,
1993). The main sources of heterogeneity are the complexity of crystallographical and
geometrical structure of solids and their complex chemical compositions.

Understanding the surface heterogeneity of adsorbents is important because it may
directly affect their adsorptive properties. Even though various modern techniques such as
electron microscopy, IR spectroscopy, electron spectroscopy, srnall-angle X—ray scattering,
NMR, etc. have been widely used to provide molecular scale information about the
heterogeneity of solid surfaces (Jaroniec and Brauer, 1986), macromolecular measurements
of adsorption and desorption on solid surfaces are still very useful for characterizing
heterogeneous adsorbents (Lee et al., 1989, 1990; Jones and Boyd 1991 b).

Adsorption energy distribution of solid surfaces, and pore-size distribution and fractal
dimension of fractally porous solids are among the most important quantities for
characterizing the heterogeneity of solids (Jaroniec and Madey, 1988). Adsorption energy
distribution, which characterizes the overall energetic heterogeneity of solid surfaces, has been
extensively studied on microporous activated carbons (Dubbin, 19 75; Jaroniec and Madey,
1989; Rudzinski and Wojciechmvsla‘, 1993). Pore-size distribution of porous solids is usually
calculated from the low-temperature nitrogen adsorption-desorption isotherm using the
Kelvin equation. Empirical relationships between pore size and characteristic adsorption
energy of activated carbons have been established (Stoeckli et al., 1989; Dubinin, 1988).
Fractal dimension depends on the microporosity of solids, and has receme been used to
characterize surface irregularity. Pfeifer et al. (1983) and Avnir et al. (1983, 1984) were the

first to introduce and apply Mandelbrot's non-Euclidean fractal geometry to characterize

45

adsorption systems. By using adsorbates of varying molecular cross section, they showed that
most solid surfaces are fractals, and that the fractal dimension measures the irregularity of the
surface accessible to molecules during adsorption processes, and is also useful to study the
porous structure of solids, the surface phase structure, and the adsorbent-adsorbent
interactions. Linear relationships between the monolayer adsorption and the molecular size
of adsorbates have been established to determine fractal dimension of a given solid surface
based on the adsorption experiments.

Most of the work on the heterogeneity of solid surfaces has been focused on activated
carbons and other inorganic adsorbents and catalysts (Jaroniec and Madey, 1988).
Comparatively little work on the heterogeneity of clay minerals (Barrow et al., 1993; Lagaly,
1979, 1981, 1982) and none on organoclays has been done. In this study, the adsorption of
toluene, p-xylene, and n-propylbenzene in aqueous solutions by Ca-smectite partially and firlly
exchanged with trimethylphenylammonium (TMPA) cations was measured. The resulting
isotherms were analyzed by several adsorption equations to characterize the heterogeneity of
such organo-smectites. In these systems, individual pores are defined by two semi-infinite
parallel silicate sheets which have permanent negative charge, and by adsorbed interlayer Ca”
and TMPA cations which neutralize this charge. The interlayer cations act as pillars to
separate silicate sheets, and to define the dimensions of the micropores. As such, the size and
structure of micropores in organo-srnectites are largely dependent on the nature of the
exchange cations (cg Ca2+ and TMPA), and the proportion of the cation-exchange capacity
(CEC) they occupy.

The objectives of this study were to evaluate the applicabilities of different adsorption

46
equations (the Langmuir equation, the DA equation and the DR equation) for describing the
uptake of aromatic hydrocarbons from water by mixed Ca/TMPA- and TMPA-smectites, and
to characterize these organo-smectites as heterogeneous adsorbents through an analysis of

adsorption energy distribution, fiactal dimension and micropore size distribution.

Background and Theory

Based on the Polanyi potential theory of adsorption (Goldman and Polanyi, 1928;
Polanyi and Welke, 1928), and extensive gas and vapor adsorption studies, Dubinin and
Astakhov (Dubinin, 19 75) developed a theory for the volume filling of micropores, expressed

semiempirically by the following equation (DA equation):

.4
0(A) = H—Y‘] (3.1)
exp 9E0
with A = anfl (3.1a)
P

where 6 is the degree of volume filling, expressed as the ratio of the amount adsorbed in the
micropores at temperature T and relative pressure p/po to the maximum adsorption amount.
R is the gas constant, A is the Polanyi adsorption potential, and E0 is the characteristic
adsorption energy of a reference adsorbate, and depends only on the adsorbate. The similarity
(affinity) coefficient [5 is a shitting factor, and is a measure of the similarities between the
characteristic vapor adsorption for the adsorbate being studied and that of a reference
adsorbate. [3 depends only on the adsorbate and can be defined as the ratio of the
characteristic adsorption energy of any adsorbate, E, to the characteristic adsorption energy

of a reference adsorbate, Eu, i.e., [Fl-31130 By convention, benzene is selected as the reference

47
adsorbate, as Dubinin (1975) suggested (i.e., Bo.rr.=1). The exponent n varies between 1 and
3, depending on the nature of adsorbent used. The value of n approaches 3 for adsorbents
with homogeneous micropores, whereas it approaches 1 for adsorbents with highly
heterogeneous micropores. When n=2, for average adsorbents, Equation 3.1 becomes:
6(A) = wi-(fi:fl (3.2)
This equation is referred to as Dubinin-Radushkevich equation (DR equation) (Dubinin,
1975). The DR equation is a good mathematical description of both vapor adsorption and
adsorption in dihrte liquid/solid systems (Jarmiec and Derylo, 1981). In liquid/solid systems,

the DR equation is written as
MA) = settLyt (3.3)
135 o
with A = RT 111% (3.3a)
where C and Co are the respective equilibrium liquid phase concentration and the liquid
solubility, and E0‘ is the characteristic adsorption energy of a reference adsorbate (C6116) in
such adsorption systems.
Choma et al. (1993) introduced an ”interfacial” similarity (affinity) coefficient p r to
relate the characteristic energies E(, and Eo’:
B*=-:—: (3.4)
They examined [3 "' value for benzene by using eight different activated carbons of difl‘erent
sources. The resulting B * value was 0.052:|:0.004 and did not depend on the heterogeneity
of the activated carbons used, illustrating the usefirlness of [i " as a measure of the similarity

between the characteristic adsorption curves of any adsorbate in the gas/ solid and liquid/solid

48
systems. Tints, the DR equation in the dilute liquid/solid adsorption systems can be rewritten

as:

 

9(A) = eel-t B ‘

)‘1 (3.5)
It should be emphasized that [3 defines the similaritie: between gas/solid adsorptions for an
adsorbate compound and a reference compound, while B‘, which is equal to 0.52 for
benzene, defines the similarities between liquid/solid and gas/solid adsorptions for an
adsorbate.

Recent studies (Stoecklr', 1989; Stoeckli et al., 1990; Innes et al., 1989) have shown
that the DR equation satisfactorily describes adsorption only in uniform and weakly
heterogeneous micropores. Stoeckli eta]. (Stoeckli, 19 77; Dubinin and Stoeckli, 1980)
suggested an integral equation combined with an adsorption energy distribution function to
describe adsorption in heterogeneous micropores. The DR equation is rewritten as:

6(A) = apt-fin (3.6)
where 10: “mp (3‘30 is the average adsorption potential. Letting 2.1/Z , adsorption in
heterogeneous micropores can be expressed as:

6(A) = jest-34 was» (3.7)
where 6(2) is the distribution function of z :orrelated to the characteristic adsorption energy
E0. Duan et a]. (198 7) assumed a Gaussian distribution for representing the adsorption
energy distribution function. However, because the Gaussian distribution has a non-zero value
at zero point argument and the resulting adsorption isotherm equation contains the error
function of the adsorption potential, Jaroniec and Madey (1989) proposed a gamma-type

distribution function for representing adsorption energy distribution firnction, which is

49
physically and mathematically meaningful and leads to a simple adsorption isotherm equation 5
. The following distribution function was obtained:

#2
6(2) = 2" zN exp(-qz’); q>0 and v>-1 (3.8)
MM)

 

with isms = 1 (3.8a)
where I‘(v/2) is the gamma function: q and v are parameters associated with the dispersion
a, of the distribution and mean E.

Integration of Equation 3.7 with the substitution of 6(2) leads to the following

adsorption isotherm equation:

 

6(A) = i ' l"2 (3.9)
,4: ma 2
Equation 3.9 does not contain the coeficient B’, and it can be used for adsorptions in both
gas/solid and liquid/solid systems.
Equations 3.8 and 3.9 are good mathematical representations of adsorptions in
gas/solid and dilute liquid/solid systems. The simplicity of these equations allows one to

properly characterize adsorption with only the parameters q, v, [i *, and B which can be

calculated fi'om experimental adsorption data.

Experimental Section

Smectite from a Wyoming bentonite (SAC) with a cation-exchange capacity (CEC)
of 90 01110ng was obtained from the American Colloid Company (Chicago, IL). The <2llm
clay fractions were obtained by wet sedimentation and subsequently saturated with Ca”

cations by washing the clay repeatedly with CaCl2 solution (0.1 mol/l).

50

Trimethylphenylammonium (TMPA) bromide (Aldrich Chemical Co.) was dissolved in
distilled water and used to prepare TMPA organo-smectites (SAC-TMPA). Clay suspensions
containing 25 g of Ca-clay were treated by adding different amounts of TMPA bromide
solution (0.03 mol/l) to form either mixed Ca/TMPA-clays or titlly saturated TMPA-clay
(SAC-TMPA). The amounts of TMPA added relative to the CEC were about 0.2, 0.4, 0.65,
0.75 and 1.0. The mixtures were agitated overnight on a magnetic stirrer at room
temperature. The organo-clay suspensions were then washed with distilled water repeatedly
until free of bromide ions as indicated by AgNO3, and stored in bottles for later use. Organic
carbon contents of clays were determined using Dohrrnann DC-l90 high temperature TOC
analyzer (Rosemount Analytical Inc, Santa Clara, CA), and were used to calculate the exact
percent of CEC occupied by TMPA cations. These organo-clays are referred to as SAC-
TMPA. 17, SAC-TMPA.38, SAC-TMPA.65, SAC-TMPA.75, and SAC-TMPAl.O,
respectively, in which the numerical value represents the fraction of the CEC occupied by
TMPA The properties of these SAC-TMPAs are given in Table 3.1.

Adsorption of benzene, toluene, ethylbenzene, p-xylene and n-propylbenzene by the
organoclays was conducted using a batch equilibration technique. A weight of 100 mg of
organoclay and 25 ml of distilled water were placed in Corex glass centrifuge tubes. Organic
solutes were added neat using a Hamilton microliter syringe to produce a range of initial and
final concentrations. The tubes were capped immediately with teflon-backed septa. The tubes
were shaken for 24 hours on a rotator at room temperature. Preliminary experiments showed
that sorption reached equilibrium within 18 hours. After equilibration, the tubes were

centrifuged at 8000 rpm (RCF=4302 g) for 10 minutes to separate liquid and solid phases.

51

Table 3.1 Properties of TMPA-smectites

 

 

sample % of CEC d001
clay name OC% occupied by TMPA (A)
1 SAC-MA17* 1.66 17 14.42, 18.89
2 SAC-TMPA.38 3.57 38 14.12, 18.97
3 SAC-TWA65 5.88 65 14.16
4 SAC-TWA. 75 6.75 75 14.95
5 SAC-MAI .0 8.84 100 14.38

 

" SAC stands for the smectite clay fi'om American Colloid Company;
TMPAXX and TMPA1.0 represent that SAC is treated with TMPA and
the degree of the CEC occupied by TMPA is .XX and 1.0, respectively.

A volume of 5 ml of supernatant was extracted with 10 ml of carbbn disulfide in a glass vial.

Analysis of organic solutes in the carbon disulfide extracts was made by gas
chromatography. A Hewlett-Packard 5890A gas chromatograph fitted with a flame ionization
detector and a packed column (Alltech Co.) was used; N2 was the carrier gas. Peak areas
were recorded by a Hewlett-Packard 3392A integrator, and compared to external standards
to determine the concentrations of organic solutes. The recovery of organic solutes in blanks
not containing organoclays was >90%. The amount of organic solutes sorbed was calculated
fiom the differences between the amount of organic solutes added and that remaining in the
final equilibrium solutions. Sorption isotherms were made by plotting the amount sorbed

versus the equilibrium concentration in solution.

52

Basal spacings (dam) of SAC-TMPA organoclays with organic solutes sorbed were
determined by XRD analysis. Organoclay suspensions were dropped on glass slides, and
immediately covered by teflon tape to prevent vapor loss. X-ray difl‘raction patterns were
recorded using CuKa radiator and a Philip APD3 720 automated X-ray difl‘ractometer using
an APD3521 goniometer fit with a 6-compensating slit, a 0.2-mm receiving slit, and a
difl‘racted-beam graphite monochromator, from 2 to 10°20, in steps of 0.03°26, 1 sec/step.

Pore size distributions of the organoclays were determined by QuantaChrome Co.
(Romeoville, IL). The organoclays were outgassed for 16 hrs at 70°C. N2 sorption
measurements were analyzed using the Brunauer-Emmett-Teller (BET) equation (Brunauer

et al., 1938) and the Horvath-Kawazoe (H-K) method (Horvath and Kawazoe, 1983).

Results and Discussion

The adsorption lsothenns of benzene, toluene, p—xylene, ethylbenzene and
propylbenzene are shown in Figure 3.1. All isotherms are type 1, indicating that the adsorption
occurs on the surfaces and in the micropores of SAC-TMPAs. Benzene shows the highest
uptake by the organoclays; progressively lower uptake occurs for larger molecules in the
order of benzene > toluene > p-xylene == ethylbenzene > n-propylbenzene. At Ce/Cs
(equilibrium concentration in water / aqueous solubility) =05, the sorbed concentration on
SAC-TMPA.65 is about 78 mg/g for benzene, 51 for toluene, 41 for p-xylene and
ethylbenzene, and 36 for n-propylbenzene. X-ray analysis showed that SAC-TMPAs do not
undergo interlayer expansion after adsorption of the test compounds over the experimental

concentration range used, indicating that the microstructures of SAC-TMPAs were not

53

 

 

 

 

 

-; o SAC-TMPA.17
_= a SAC-TMPA.38
_: a SAC-TMPA.65
v SAC-TMPA.75
- o SAC-TMPA1.0
"'rn o 300 600 9001200
E 60 : so :»
o: 45 C: 5 4° '5
E 30 i
3 3 / I, 20 1
“a v ' ; g
f 15 ' toluene 5 10 ‘:
C - I
3 o o
g o 100 200 300 400 0 40 so 120 160
<

 

 

55‘) IIII’ITPII r1 II'I II II I ‘8‘) IIIII'IIIWI'IIIII'IIIIFIIIII

11
l

   

20
10

0
0 40 80120160 01020304050

Equilibrium Concentration, Ce(mg L")

   

'7"
13
3
'o
‘5.
g
a
o

Lljlllllllllllllll

ethylbenzene

 

 

llllllllllllllllllllll

  

Figure 3.1 Adsorption isotherms of benzene and alkylbenzenes by mixed
Ca/TMPA- and TMPA-smectites

54
altered.

Choice of the adsorption 'notherm equation. All adsorption isotherms were fit to
the Langmuir isotherm equation and the Dubinin-Astakhov isotherm equation with n=2 and
n=3. Typical fitting results are represented in Figure 3.2. For small adsorbates such as
benzene, the Langmuir equation overestimates adsorption at low concentrations and
underestimates adsorption at high concentrations (Figure 3 .2). For larger molecules like n-
propylbenzene, the Langmuir equation gives a better fit, but it still deviates slightly from the
experimental data (Figure 3.2).

The DA equation with n=3 provides a good description of adsorption of gases and
aqueous compounds on activated carbons that contain uniform micropores (Stoeckli, 1989;
Stoeckli et al., 1990; Innes et al., 1989). When combined with the gamma-type adsorption
energy distribution firnction, it also describes adsorption on heterogeneous activated carbons.
However, it does not give a good fit to our experimental data in water/SAC-TMPA systems.
The DR equation, which is suitable for weakly heterogeneous activated carbons (Stoeckli,
1989; Imres et al., 1989; Kraehenbuehl et al., 1986), combined with the gamma-type
adsorption energy distribution, does provide an excellent fit of our experimental data over the
whole concentration range used (Figure 3.2).

The DR equation 'is a fundamental mathematical description of gas adsorption on
porous and microporous solids. It is also very useful for describing gas adsorption on
mesoporous, macroporous, and non-porous solids, although there is no theoretical basis for
the equation (Hobson, 1961; Cerofolinr', 1974). Our results suggest that the DR equation may

be extended to clay/organoclay systems, and in the following discussion it will be used to

55

 

PA 140
3 39 :
é . - so
G -l
- - n-propyl- -
E benzene I
- 20
o -t
(I) ..
'0 . .
< O expenmental ,
E —— Langmuir - 10
g ----- DR 2
e — - DA with n=3 -
< ' o

 

0 250 500 750 0 10 20 30 40
Equilibrium Concentration, Ce(mg L")

Figure 3.2 Adsorption isotherms of benzene and n-propylbenzene fitted to the
Langmuir equation, the DA equation and the DR equation
characterize the aqueous organoclay adsorption systems presented here.
Adsorption energy distribution function. Adsorption energy distribution firnctions
are dependent on the heterogeneity parameters q and v (Equation 3.8). In order to obtain q
and v for calculating the adsorption energy distribution functions for SAC-TMPAs, all
isotherms were fit to Equation 3.9. The resulting values of q and v, together with the

maximum adsorption amount Q,,,, are listed in Table 3.2.

56

Table 3 .2 Parameters Qm, v, q for adsorption of aromatic hydrocarbons in water/TMPA-

 

 

 

 

 

 

smectite systems
SAC-TMPA .17 .38 .65 .75 1.0
Qn(mg/g) 60.61 73.24 77.84 78.51 77.36
benzene v 1.095 1.588 1.482 1.589 1.508
q(kJ/mol)2 1.031x10‘ 2.594x10‘ 3.411x10‘ 4.238x10‘ 3.033x10‘
Qm(mg/g) 35.50 49.65 55.30 53.22 43.90
toluene v 1.328 1.219 1.979 1.598 2.466
q(kJ/mol)2 9.955x103 9.493x103 2.485x10‘ 2.028x10‘ 4.281x10‘
Q_(mg/g) 30.87 38.00 45.51 44.16 37.63
p-xylene v 1.219 2.109 2.170 2.356 2.977
q(lrJ/mol)2 8.368x103 3.224x10‘ 3.023x10‘ 4.144x10‘ 5.199x10‘
ethyl- Qn(mg/g) 32.72 38.62 44.98 44.64 38.48
benzene v 1.353 2.696 3.400 3.065 3.466
q(kJ/mol)2 8.520x103 4.545x10‘ 5.978x10‘ 6.140x10‘ 6.574x10‘
n-propyl- Qm(mg/g) 21.31 31.54 37.16 36.97 31.88
benzene v 33.42 65.39 86.98 56.33 75.79
q(kJ/mol)2 9.801x10’ 2.152x10‘ 2.848x10‘ 2.125x10‘ 2.430x10‘

 

The adsorption energy distribution firnctions versus the inverse value of the average

adsorption potential, 2, calculated from Equation 3.8, are shown in Figure 3.3. All are

asymmetrical gamma-type distributions with a broadening toward lower energies. To better

describe those adsorption energy distributions, the 2,, values (at which adsorption energy

57

 

2OOIUIT'rlrrlrIrr
‘
.

 

 

         

 

 

 

150 ,9 _ — SAC-TMPA.17
. 1 -— — SAC-TMPA.38
' ——- SAC-TMPA.65
---- SAC-TMPA.75
.:‘ ----------- SAC-TM PA1 .0
E
g 0 i I I I I I : I AI I I I l : 200
n I: :
5 :E' [A 5: 150
c" ul- .
.9 -_ p—xylene _:
c a
E i ‘ -‘ 50
c I
.g .
3 I I 0
.3 .I‘ :5 l E
o 180 ,A 1;- : 660
‘ ethylbenzene E n-propylbenzene E
120 1- 1 440
' L\_ qfil r r r r 1 r r r 1 q o

 

 

0 a
0.00 0.1 0.02 0.00 0.01 0.02 003
Energy Parameter, z(mol id")

Figure 3.3 Adsorption energy distributions of mixed CafTMPA- and
TM PA-smectites

58

distributions reach a maximum), the average I values, and the dispersion a, values were

calculated according to the following equations (Jaroniec and Madey, 1989):

 

z...= fi (3.10)
29
.. r61)
; = semis = 2 i (3.11)
0 111) J;

 

(3.12)

 

The results are listed in Table 3.3. For all SAC-TMPAs, the energy distributions of benzene
reach a maximum at higher adsorption energies compared to alkylbenzene studied, with an
average 2m value of 0.0028 mol/kJ. By comparison, zm values for toluene, p-xylene,
ethylbenzene and n-propylbenzene were all about 0.0040 mol/kJ. As indicated by 0,, the
energy distributions of toluene are broader than that of benzene, and then progressively
narrower for larger molecules in the sequence: benzene < toluene > p-xylene z ethylbenzene
> n-propylbenzene. For n-propylbenzene, the adsorption energy distributions are very sharp
and nearly symmetrical. For the adsorption of the same adsorbate on different SAC-TMPAs,
SAC-TMPA 17 gives the broadest distributions. The adsorption energy distributions for the
other four SAC-TMPAs are similar to each other, particularly in the case of larger molecules
such as ethylbenzene, but remarkably difl'erent from SAC-TMPA.17, the general trend in

width being SAC-TMPA.” z SAC-TMPALO > SAC-TMPA.65 z SAC-TMPA.75. In the

59

case of n-propylbenzene, they actually overlap, and only one curve is shown in Figure 3.3 for

n-propylbenzene.

Table 3.3

The characteristics of adsorption energy distribution

 

 

 

 

 

 

SAC-TMPA .17 .38 .65 .75 1.0 average
zm(mol/kJ) 0.0029 0.0034 0.0026 0.0026 0.0029 0.0028
benzene :(moVkJ) 0.0059 0.0048 0.0040 0.0037 0.0043
o,(moVkJ) 0.0043 0.0028 0.0024 0.0022 0.0026
zm(mol/kJ) 0.0041 0.0035 0.0044 0.0038 0.0041 0.0041
toluene ;(moVkJ) 0.0068 0.0066 0.0056 0.0054 0.0049
o,(mol/kJ) 0.0045 0.0045 0.0029 0.0032 0.0023
zm(mol/kJ) 0.0036 0.0041 0.0044 0.0040 0.0044 0.0040
p-xylene :(mol/kJ) 0.0071 0.0051 0.0054 0.0048 0.0049
oz(mol/kJ) 0.0048 0.0026 0.0027 0.0023 0.0021
ethyl- zm(mol/k.l) 0.0046 0.0043 0.0045 0.0041 0.0043 0.0044
benzene ;(mol/kJ) 0.0075 0.0050 0.0050 0.0046 0.0048
o,(moVkJ) 0.0048 0.0022 0.0020 0.0019 0.0019
n-propyl- zm(mol/kJ) 0.0041 0.0039 0.0039 0.0036 0.0039 0.0039
benzene I(mol/kJ) 0.0041 0.0039 0.0039 0.0036 0.0039
o,(mol/kJ) 0.0005 0.0005 0.0005 0.0005 0.0005

 

As a probe for the heterogeneity of adsorbents, adsorbate molecules only detect those

sites that interact with the probe molecules during the adsorption process. Therefore, the

60

adsorption experiments provide information concerning the "relative” heterogeneity. It is
apparent from the adsorption energy distributions summarized above that there is a steric
efl‘ect involved in adsorption. Benzene molecules, due to their small sizes, are adsorbed on
the sites of (relatively) high energies, whereas toluene and larger molecules can only be
adsorbed on sites of lower energies. Stoeckli eta]. (1989) have shown that the characteristic
adsorption energy and micropore size of activated carbons have an inverse relationship. A
similar relationship correlating the characteristic adsorption energy and micropore sizes may
exist in TMPA organo-smectites. This suggests the existence of fine micropores of high
adsorption energies in SAC-TMPAs which are accessible for benzene molecules but sterically
exclude toluene and larger molecules. Such molecules can only be adsorbed in larger
micropores, and on mesoporous, macroporous and flat surfaces which have the sites of lower
energies, as indicated by the shift of 2,, values for these molecules. These results also suggest
that sites of benzene adsorption have micropore diameters larger than the length of the
benzene molecules (5.6 A) but smaller than the length of toluene molecule (7.1 A). Molecular
planes of the adsorbates must be parallel to the basal plane of the clay as indicated previously
by Jaynes and Boyd (1 991 b). If adsorbed molecules were oriented vertically to the clay
surfaces, all five adsorbates used in this study would fill micropores since they have the same
widths of 4.8 A.

The above discussion is supported by the measured pore size distributions of SAC-
TMPAs shown in Figure 3.4. All pores are in the micropore region. A significant portion of
the micropores have pore diameters about 6.5 A, larger than the length of the benzene

molecule and smaller than the length of toluene.

61

100
- SAC-TMPA. 17 ~
50 - 1
0 r r r r
- LSAC—TMPAU: .38 ~
- l4 1 l L l J l l ‘L==l==l==rl=;

- SAC-TMPA.65 -

 

 

0'1
0
I

O

I- q

lMlMlllllJnaa

dV/dD (til/9A)
01
O
l

 

 

 

 

 

O r 1 r

- SAC-TMPA.75 -
5O " -*
0 1 1 l l 1 1 1 I I ‘-::t:-l'+1==l=

i SAC-TMPA1.0 -
50 - -
0 .

0 5 10 15 20

Pore Diameter D (A)

Figure 3.4 Micropore size distributions of mixed Ca/TMPA- and TMPA-
smectites

62

Although benzene molecules can fill the fine micropores, toluene, p-xylene,
ethylbenzene, and n-propylbenzene are adsorbed in larger micropores and on surfaces. The
smaller toluene molecules explore more sites of different energies on the clay surfaces than
the larger alkylbenzenes. As a result, toluene detects a more heterogeneous surface and gives
broader adsorption energy distributions. The adsorption energy distribution narrows and the
heterogeneity is concealed gradually as the adsorbate molecular size increases. In the case of
n-propylbenzene, the heterogeneity is concealed completely so that the adsorption isotherms
can be fit to Langmuir equation (Figure 3.2).

Cation segregation, especially at low TMPA content, may lead to a broadening of the
energy distribution profile. McBride and Mortland (19 73) observed the segregation of
different cations in a smectite when ethylammonium was partially exchanged by
tetrapropylamrnoniurn up to about 55 percent of the CEC. Faver and Lagaly (1991) reported
that large quaternary ammonium cations (e. g. hexadecyltrimethylammonium) are
preferentially adsorbed on low charge density clay surfaces while small organic cations such
as TMPA prefer to be adsorbed on surfaces of high charge density. We have also observed
that Ca2+ and TMPA cations are partially segregated in SAC-TMPAs, resulting in the
formation of CaZI-rich and TMPA-rich patches. This is due to self-interaction of TMPA
cations and the inhomogeneity of layer charge density (Sheng, 1996). This type of
topographical heterogeneity is especially significant for SAC-TMPA. l7, producing the broad
adsorption energy distributions. As the TMPA content increases in these clays, Ca” and
TMPA distribution becomes more random and hence the surfaces more homogeneous. This

is indicated by the progressively narrower adsorption energy distributions of SAC-TMPA.3 8,

63
SAC-TMPA65, and SAC-TMPA.75. Once the CEC is fully saturated by TMPA (in SAC-

TMPAl .0), the adsorption energy distribution is slightly broader. This may be related to the
formation of more fine micropores due to the high TMPA density on the surfaces.

Surface fractal dimension and micropore size distribution. Surface fractal
dimension has a promising application for characterizing heterogeneous adsorbents. For a
serial of spherical molecules, Pfeifer eta]. (1983) and Avnir et a]. (1983, 1984) derived the
following relationship which relates the monolayer amount adsorbed (Q,,) to the spherical
radius of the adsorbate molecule (r) and the surface fractal dimension (D) which is between
2 and 3:

Q. cc r’D (3.13)
However, the adsorbate molecules need not be spherical, provided they are adsorbed flat on
the adsorbent surface and belong to a homologous series for which the ratio of the square of
the linear extent (i.e., largest distance, r, between any two points on the probe) to the
effective cross sectional area is the same for all members (Pfeifer et al., 1983). For the
sorption data presented here, a plot of logarithm Qm versus logarithm r yields a straight line
(Figure 3.5), indicating that these requirements are satisfied. Values of fractal dimension (D)
are obtained fi'om the slopes of these plots and listed in Table 3.4 together with the lengths
of the aromatic molecules (r) which were used to estimate the fractal dimensions. Correlation
coeflicients (R2) in Table 3.4 verify the highly linear relationship between log Q“I and log r.
The high D value for SAC-TMPA 17 suggests a comparatively high degree of heterogeneity
for SAC-TMPA 17. This high heterogeneity can be attributed to the formation of TMPA-rich

patches in regions of higher charge density and the associated formation of numerous fine

 

.0
N
..
.
.
.
_
.
.
.
..
..
.

      

 

 

 

b I r1 I I I I I .
: 8’ 0 I
s r: q
0.0 — 8 a,» 5 -
’ ‘1’ at c: N ‘
Z 2 c a) c j
. w .9 22 3 .
-0.2 b C §~>~ ; -
3 0 5 a. :
E 1 2 do 9 2
8’ -0.4 r 8 2- 1
-0.6 :— o SAC-TMPA.17 -:
t n SAC-TMPA.38 :
- A SAC-TMPA.65 O .l
’0-8 :' v SAC-TMPA75 1
I 0 SAC-TMPA1.0 d
_1.0 - r #4 r 1 r r r r I 14 r r 1 r r r r
0.7 0.8 0.9 1.0 1.1

logr

Figure 3.5 Dependence of monolayer adsorption on molecular size

micropores due to the high TMPA density. The D value decreases from 2.550 for SAC-
TMPA 17 to 2.153 as the TMPA content increases to about 0.65CEC, indicating that TMPA-
srnectite becomes less heterogeneous. This can be attributed to the adsorption of TMPA on
the layers of lower charge density, and the resultant formation of larger pores which lowers
the fractal dimension. The D value increases again to 2.397 for SAC-“IMPALO as the TMPA
content increases to 1.0CEC due to the formation of additional micropores. The analysis of

fractal dimension results is consistent with the previous analysis of adsorption energy

65

distributions.

Table 3 .4 Molecular lengths of adsorbates and fractal dimensions of SAC-TMPAs

 

 

 

 

 

adsorbate benzene toluene p-xylene ethylbenzene n-propylbenzene
length (A) 5.6 7.1 8.4 8.4 9.6
SAC-TMPA .17 .38 .65 .75 1.0
D 2.550 2.360 2.153 2.164 2.397
R2 0.9704 0.9998 0.9989 0.9984 0.9842

 

These results are also consistent with the measured pore size distributions of SAC-
TMPAs shown in Figure 3.4. The volume fiaction of the portion of the micropores with pore
diameters of about 6.5 A increases (as the D value decreases) in the order: SAC-TMPA. 17
< SAC-TMPA.38 < SAC-TMPA.65, decreases thereafter (as the D value increases) in the
order: SAC-TMPA.65 > SAC-TMPA.75 > SAC-TMPALO.

Theoretically, as long as the relationship between the adsorption energy parameter 2
and pore size x is known, the adsorption energy distributions (G(z)) can be transformed into
micropore size distributions (J (x)) according to the following relationship (Jaroniec et al.,
1990):

J(x) = Gaxi) (3.14)

66
Stoeckli et al. (1989) have shown that, in activated carbons which are believed to have slit-
like micropores, the adsorption energy parameter (2) (and hence the characteristic adsorption
energy E0) is simply related to the size of the micropores (x). Although several complex
relationships between 2 and x have been proposed (Stoeckli et al., 1989; Dubinin, 1988), the
following simple inverse relationship has been successfully used on activated carbons (Stoeckli
et al., 1989; Dubinin, 1988):
rtEo = k (nm.kJ/mol) (3.15)

where k is a constant in the micropore region between 9 and 13 nm-kJ/mol for activated
carbons, and x is the size of the micropores (in nanometers), corresponding to the half-width
in the case of slit-like micropores in activated cm'bons. \Vrth the molecular probes of various
sizes, x is the half-width of the micropores accessible to probe molecules. A similar
relationship may be applicable to organoclays which also have slit-like micropores. However,
the constant k would likely be difi‘erent, and is unknown, which makes it impossible to directly
transform the adsorption energy distributions into micropore-size distributions in this study.

Based on the theory of adsorption on fi'actally porous solids, Avnir et al. (1983) have
developed the following theoretical proportionality relationship between pore-size distribution
J (x) and fractal dimension:

J(x) « x“) (3.16)

In the micropore region, J aroniec et al. (1990) derived the following equation:

J(x) = 7:29.33...” (3.17)
xn-x‘

where x“ is the upper limit of micropore size, and xm is the lower limit of micropore size.

Plotting fractal dimension against mean pore size, shown in Figure 3 .6, gives the following

67

 

2.8 I I I I I I 1 I I I I T I I

2.6

2.4

Fractal Dimension, D

2.2

lllllllllllllLllJl

 

 

TIIIIFIIIIIIUIIUIIU

p
p-
h
p
-
r-
l-
h
-
—
I-
y-
l-
_

 

2.0
0.45 0.48 0.51 0.54

Mean Pore Half-W1dth,x(nm)

Figure 3.6 Dependence of fractal dimension on mean pore size of mixed
CaITMPA- and TMPA-smectites

linear relationship:

D = 5.478 - 6.308 ; R’=1.000 (3.18)
This type of linear relationship was also obtained for activated carbons and consistent with
the theoretical analysis (Jaroniec et al., 1993). This relationship could be used to estimate the
fiactal dimension from experimental micropore size distribution of TMPA-smectites. In the

above relationship, the mean pore size is calculated via the following expression using

68

xm=1.0 nm and Xu,-50.1 nm:

‘-
f xJ(x)&

_. = t.-
x '- (3.19)

f J(x)d:
5‘

 

Our experimental results of micropore size distribution and the theoretical linear
relationship between D and i illustrate that D is simply correlated to the nricroporosity of
SAC-TMPAs and reflects the fact that the surface irregularities of SAC-TMPAs are largely

generated by their microporosity.

References

Avnir, D., D. Farin, and P. Pfeifer. 1983. Chemistry in noninteger dimensions between
two and three. 11. fi'actal surfaces of adsorbents. J. Chem. Phys. 79:3566-3571.

Avnir, D., D. Far-in, and P. Pfeifer. 1984. Molecular fractal surfaces. Nature 3082261-
263.

Barrow, N.J., G.W. Brummer, and R. Strauss. 1993. Effects of surface heterogeneity
on ion adsorption by metal oxides and by soils. Langmuir 9:2606-2611.

Boyd, S.A., M.M. Mortland, and CT. Chiou. 1988a. Sorption characteristics of
organic compounds on hexadecyltrimethylammonium-smectite. Soil Sci. Soc. Am. J.
52:652—657.

Boyd, S.A., W.F. Jaynes, and B.S. Ross. 1991. Immobilization of organic contaminants
by organo-clays: application to soil restoration and hazardous waste containment. In
Organic Substances and Sediments in Water, Vol. 1. Baker,R. S. ( ed). Boca Raton, FL:
CRC Press, pp181-200.

Boyd, S.A., J.-F. Lee, and M.M. Mortland. 1988b. Attenuating organic contaminant
mobility by soil modification. Nature 333:345-347.

Brunauer, S., P. Emmett, and E. Teller. 1938. Adsorption of gases in multimolecular
layers. J. Amer. Chem. Soc. 60:309-319.

69

Cerofolini, G.F. 1974. Localized adsorption on heterogeneous surfaces. Thin Solid Films
23: 129-152.

Choma, J., W. Burakiewicz—Mortka, M. Jaroniec, and RK. Gilpin. 1993. Studies of
the structural heterogeneity of microporous carbons using liquid/solid adsorption
isotherms. Langmuir 9:2555-2561.

Dubinin, M.M. 1975. Physical adsorption of gases and vapors in micropores. Prog. in
Surface and Membrane Sci. 9: 1-70.

Dubinin, M.M., and ILF. Stoeckli. 1980. Homogeneous and heterogeneous micropore
structures in carbonaceous adsorbents. J. Colloid Interface Sci. 75 :34-42.

Dubinin, M.M., and O. Kadlec. 1987. Adsorption properties and nricroporous
structures of carbonaceous adsorbents. Carbon 25:593-598.

Dubinin, M.M. 1988. On methods for estimating micropore parameters of carbon
adsorbents. Carbon 26:97-98.

Faver, 11., and G. Lagaly. 1991. Organo-bentonites with quaternary alkylammonium
ions. Clay Minerals 26:19-32.

Goldman, F., and M. Polanyi. 1928. Adsorption von dampfen an kohle und die
warmeausdehnung der benetzungsschicht. Z. Phys. Chem. 132:321-370.

Hobson, J.P. 1961. Physical adsorption of nitrogen on pyrex at very low pressures. J.
Chem. Phys. 34: 1850-1851.

Horvath, G., and K. Kawazoe. 1983. Method for the calculation of efi‘ective pore size
distribution in molecular sieve carbon. J.’ Chem. Eng. Jap., 16:470-475.

Innes, R.W., J. Fryer, and ILF. Stoeckli. 1989. On the correlation between micropore
distribution obtained from molecular probes and from high resolution electron microscopy.
Carbon 27:71-76.

Jaroniec, M., and A. Derylo. 1981. Application of Dubinin-Radushkevich-type equation
for describing bisolute adsorption from dilute aqueous solution on activated carbon. J.
Colloid Interface Sci. 84: 191-195.

Jaroniec, M., and P. Brauer. 1986. Surface Sci. Reports 6:65.

J aroniec, M., and R. Marley. 1988. Physical Adsorption on Heterogeneous Solids.
Amsterdam, The Netherlands: Elsevier Sci. Publishers B.V., 351pp.

70

Jaroniec, M., and R. Madey. 1989. A comprehensive theoretical description of physical
adsorption of vapors on heterogeneous microporous solids. J. Phys. Chem. 93:5225-5230.

Jaroniec, M., X. Lu, R. Marley, and D. Avnir. 1990. Thermodynamics of gas
adsorption on fi'actal surfaces of heterogeneous microporous solids. J. Chem. Phys.
92:7589-7595.

Jaroniec, M., X. Lu, and R. Marley. 1990. Evaluation of structural heterogeneities and
surface irregularities of microporous solids. Mater. Chem Phys. 26:87-97.

Jaroniec, M., RR. Gilpin, and J. Choma. 1993. Correlation between microporosity
and fractal dimension of active carbons. Carbon 31:325-331.

Jaynes, W.F., and S.A. Boyd. 1990. Trimethylphenylammonium-smectite as an effective
adsorbent of water soluble aromatic hydrocarbons. .1. Air Waste Manage. Assoc. 40: 1649-
1653.

Jaynes, W.F., and S.A. Boyd. 1991a. Clay mineral type and organic compound sorption
by hexadecyltrimethylammonium-exchanged clays. Soil Sci. Soc. Am. J. 55:43 -48.

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

Kraehenbuehl, F., H.F. Stoeckli, A. Addoun, P. EhrBurger, and J.B. Donnet. 1986.
The use of immersion calorimetry in the determination of micropore distribution of
carbons in the course of activation. Carbon 24:483-488.

Lagaly, G. 1979. The ”layer charge” of regular interstratifred 2:1 clay minerals. Clays
Clay Miner. 27: 1-10.

Lagaly, G. 1981. Characterization of clays by organic compounds. Clay Miner. 16: 1-21.

Lagaly, G. 1982. Layer charge heterogeneity in vernriculites. Clays Clay Miner. 30:215-
222.

Langmuir, I. 1918. The adsorption of gases on plane surfaces of glass, mica and
platinum. J. Am. Chem. Soc. 40:1361-1403.

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

71

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 tetrarnethylarnmonium-smectites having different
charge densities. Clays Clay Miner. 38:1 13-120.

McBride, M.B., and M.M. Mortland. 1973. Segregation and exchange properties of
alkylammonium ions in a smectite and vermiculite. Clay & Clay Miner. 21:323-329.

Pfeifer, P., and D. Avnir. 1983. Chemistry in noninteger dimensions between two and
three. I. fractal theory of heterogeneous surfaces. J. Chem. Phys. 79:3558-3565.

Polanyi, M., and K. Welke. 1928. Adsorption, adsorptionswarme und bindungscharakter
von schwefeldioxyd an kohle bei geringen belegungen. Z. Phys. Chem. 132:371-3 83.

Rudzinski, W., and B.W. Wojciechowski. 1993. Efi‘ects of surface heterogeneity in
adsorption and catalysis on solids. Langmuir 9: 1-2.

Sheng, G. 1996. Function of trimethylphenylammonium cations in smectite for adsorption
of aqueous organic compounds. Chapter 4. PhD Thesis. Michigan State University. East
Lansing, Michigan.

Sheng, G., S. Xu, and S.A. Boyd. 1996. Mechanism(s) controlling sorption of organic
contaminants by surfactant-derived and natural organic matter. Environ. Sci. Tech.
30: 1553-1557.

Stoeckli, [LE 1977 . A generalization of the Dubinin-Radushkevich equation for the filling
of heterogeneous micropore systems. J. Colloid Interface Sci. 59: 184-185.

Stoeckli, 11F. 1989. On the description of micropore distributions by various
mathematical models. Carbon 27:962.

Stoeckli, H.F., L. Ballerini, and S. DeBernardini. 1989. On the evolution of micropore
widths and areas in the course of activation. Carbon 27:501-502.

Stoeckli, H.F., P. Rebstein, and L. Ballerini. 1990. On the assessment of nricroporosity
in active carbons, a comparison of theoretical and experimental data. Carbon 28:907-909.

Xu, S., G. Sheng, and S.A. Boyd. 199_. Use of organoclays in pollution abatement. Adv.
Agron. (in press).

Chapter 4 Function of Trimethylphenylammonium Cations in
Smectite for Adsorption of Aqueous Organic Compounds

Abstract

Several organoclays were formed by exchanging Ca2+ in a Ca2*-saturated smectite
partially or fully with trimethylphenylammonium (TMPA) cations. The functions of TMPA
cations in these organoclays as adsorbents for organic compounds in aqueous solutions were
then examined. TMPA cations were found to take a partially random distribution in the
interlayers of mixed Ca/TMPA-smectites. The presence of TMPA and its partially random
distribution result in far less water associated with the clay surfaces and interlayers. In these
clays, water molecules are held by Ca2+ ions only via coordination and electrostatic attraction,
composing about two hydration shells around each Ca". The interspersing of TMPA and Ca2+
ions prohrbits the formation of a network of larger hydration shells around Ca2+. Calculations
showed that water molecules held in any other fashion, e.g. via weak hydrogen bonding to
the siloxane surface, are completely removed during the adsorption of organic compounds.
These results demonstrate that the amount of water associated with the clay surfaces and
interlayers depends on the nature of the exchange cation, and not on the amount of available
siloxane surface area per se. In TMPA-smectites, the TMPA cations firnction as non-hydrated
pillars, and sorption of organic solutes occurs predominantly on the adjacent siloxane
surfaces, which are hydrophobic in nature. The hydration shell around Ca2+ in mixed

Ca/TMPA-smectites obscures some of the siloxane surfaces. This diminishes sorption

72

73

capacity because organic solutes cannot displace the water of hydration.

Key Words - Adsorption, Organoclay, Smectite, TMPA.

Introduction

Due to the isomorphous substitution in the aluminosilicate layers, natural clay minerals
usually have a net negative charge which is balanced by alkali-metal and alkaline-earth-metal
cations such as Na+ and Ca”. The strong hydration of these inorganic cations makes the
mineral surfaces hydrophilic. Early studies (Stark, 1948; Hanson and Nex, 1953; Call, 195 7;
Spencer et al., [969; Spencer and Cliath, 19 70) on vapor phase sorption by soils and clays
demonstrated that pesticides and other relatively non-polar organic compounds are not
strongly adsorbed on such mineral surfaces.

The surface properties of clay minerals can be dramatically changed by simple ion-
exchange of inorganic cations with a variety of organic cations such as quaternary ammonium
cations of the form [(CH,),NR]" or [(CH3)2NRR']". Previous studies have shown (Mortland
et al., 1986; Boyd et al., 1988a, b, c; 1991) that the organoclays can be used to effectively
sorb organic compounds from water. i

The sorptive properties of the organoclays formed with quaternary ammonium cations
depend largely on the size of R- group (Boyd et al., 1991). Mortland et al. (1986) and Boyd
et al. (1988c) found that when R is a large alkyl group such as hexadecyl, the organoclays
formed fiom smectite were efl‘ective sorbents for phenol and chlorophenols compared to
unmodified smectite. A recent study (Sheng et al., 1996) showed that clays modified with

hexadecyltrimethylarmnonium cationic surfactant (HDTMA) effectively sorb neutral organic

74

contamirmnts via multiple mechanisms. For aliphatic compounds such as trichloroethylene and
carbon tetrachloride, partitioning into the organic phase derived from the conglomeration of
HDTMA is the singular sorption mechanism by HDTMA-clays, consistent with previous
observations and with the concept of solute partitioning into soil organic matter (Chiou et al.,
19 79, 1983). The sorption isotherms generally display type III (convex to the abscissa
commencing at the origin) compared to the linear isotherms for soil organic matter. This is
due to the much higher degree of sorption into the HDTMA phase, as compared to soil
organic matter, which substantially changes the composition of HDTMA phase and causes
the solvency of HDTMA phase for solutes to increase. Multiple mechanisms are involved in
the sorption of aromatic compounds by HDTMA-clays. Aromatic molecules solvate the
cationic ammonium centers and alkyl chains of HDTMA, and are concomitantly adsorbed on
the vacated mineral surfaces. Solvation leads to a cooperative adsorption and yields a sigmoid
isotherm. The combination of solvation (a sigmoid isotherm) and partitioning (a type III
isotherm) produces a double-sigmoid isotherm for aromatic solute sorption by HDTMA-
clays.

In contrast to the solvation and partitioning behavior of organoclays exchanged with
large organic cations such as HDTMA, the organoclays formed with small organic cations
display adsorptive properties. Due to their small size, organic cations such as
tetramethylammonium (TMA) or trimethylphenylammonium (TMPA) do not form a
continuous organic phase on the clay surfaces and in the interlayers. Rather, they are
physically isolated, leaving unobscured (fi’ee) siloxane mineral surfaces between the

ammonium pillars. Lee et al. (1989, 1990) studied the adsorption of benzene and substituted

75
benzenes by TMA-smectite. Sorption of organic compounds as vapors by the dry clay and as
solutes from water was examined. Sorption by the dry clay manifested type II isotherms and
was not strongly dependent on sorbate size. However, shape-selective adsorption of aromatic
molecules from water by TMA-smectite was evidenced by high uptake of benzene and -
progressively lower uptake of larger aromatic molecules. This phenomenon was ascribed to
the shrinkage of interlarnellar cavities by water.

Jaynes and Boyd (1990) reported that trimethylphenylammonium (TMPA)-smectite
was also an efl‘ective adsorbent of water soluble aromatic hydrocarbons including benzene,
toluene, ethylbenzene, p-xylene, butylbenzene, and naphthalene, and did not show strong
shape-selective adsorption characteristics of TMA-smectite. It was suggested that the water
hydration of TMPA as compared to TMA accounted for the different adsorption
characteristics of the corresponding organoclays. In a subsequent study (1991), the charge
density of smectite was reduced first, resulting in different TMPA contents in the organoclays.
Adsorption was found to increase as layer charge and TMPA content decreased. They
concluded that the siloxane surface of smectite is hydrophobic, and aromatic molecules were
mainly adsorbed on the hydrophobic siloxane surfaces. TMPA cations had little direct efi‘ect
on adsorption, and functioned as non-hydrated ”pillars” to keep the smectite interlayers open
for adsorption.

Considering the important efi‘ects of charge reduction on the adsorption of aromatic
hydrocarbons by TMPA-smectites, it was of interest to evaluate the adsorptive properties of
mixed Cal’I'MPA-srnectites. Difi‘erences would be expected between charge-reduced TMPA

smectite and mixed Ca/TMPA-smectite because charge reduction leaves the hydrophobic

76
siloxane surfaces available for the adsorption of organic compounds while Ca2+ in mixed
Ca/I'MPA-smectite will occupy the surfaces and may be hydrated. Furthermore, differences
in the adsorptive properties would be expected between TMA/Ca-smectite, which does not
contain an aromatic moiety, and CafI'MPA-smectite which does. In the latter case, aromatic
rings of TMPA may interact directly with the 1: electron system of aromatic adsorbate
molecules. The overall objective of the current work is to evaluate the firnction of TMPA and

Ca” cations in the adsorption of organic compounds on mixed Ca/‘I'MPA-smectites.

Experimental Section

Smectite from a Wyoming bentonite (designated SAC), which has Na as the primary
exchangeable cation and a cation-exchange capacity (CEC) of ~90 cmoleg was obtained
fiom the American Colloid Company (Chicago, IL). The <2 um clay fractions were obtained
by the wet sedimentation method, and subsequently Ca-saturated. Trimethylphenylammonium
(TMPA) bromide (Aldrich Chemical Co., Milwaukee, WI) was used to prepare organoclays.
To obtain mixed Ca/TMPA clay, aqueous TMPA solutions were added to 2 liters of smectite
suspensions containing 25 grams of Ca-saturated smectite in amounts less than the CEC of
the clay. To fully saturate the smectite with TMPA, the total amount of TMPA added was
equal to 3 times the CEC of the smectite. The mixtures of the smectite suspensions and
TMPA solution were stirred overnight at room temperature (~23°C). The clay suspensions
were then washed with distilled water repeatedly until free of bromide ions (as indicated by
AgNO3 test), and subsequently quick-frozen, freeze-dried. Organic carbon (0C) contents

were determined using Dohrrnann DC-190 high temperature TOC analyzer (Rosemount

77

Analytical Inc., Santa Clara, CA), and used to calculate the percent of CEC occupied by
TMPA cations (Table 4.1).

The batch equilibration technique was used to quantitate TMPA adsorption and Ca”
release. Ca2+ saturated smectite (0.1g) was weighed into a series of 25-ml Corex glass
centrifirge tubes containing difi‘erent volumes of aqueous TMPA bromide solutions (3.6
cmol/L), corresponding to 0.16, 0.32, 0.4, 0.6, 0.8, 0.96, 1.2, 1.6, 2.4, 4.0, 6.0, 8.0 times
CEC of the smectite, followed by adding distilled water to bring the total volume to 20 ml.
The tubes were shaken on a reciprocating shaker at room temperature for 4 days. After
centrifirgation at 6000 rpm (RCF=4302 g) for 10 min, a portion of each supernatant was
taken and diluted to the proper concentration. TMPA concentrations were measured using
an ultraviolet absorption spectrophotometer (Hewlett-Packard 8452A, Diode Array,
Germany) at 254 nm. Ca2+ concentrations were measured using an atomic absorption
spectrophotometer (Perkin-Elmer 1 1003, Norwalk, CT). The amounts of TMPA adsorbed
were calculated by the differences between the initial and equilibrium concentrations.
Adsorption of TMPA cations and release of Ca2+ were plotted against the equilibrium TMPA
concentration.

Basal spacings were determined by X-ray difi‘raction analysis. TMPA-smectites with
difi‘erent exchange degrees were analyzed by dropping aqueous suspensions on glass slides
which were subsequently covered by teflon tape to prevent them from drying. X-ray
diffraction patterns were recorded using CuKa radiator and a Philips APD3 720 automated
X-ray difiactometer using an APD3 521 goniometer fit with a 0-compensating slit, a 0.2-mm

receiving slit, and a diffracted-beam graphite monochromator, from 2 to 10°20, in steps of

78

0.02°20, 1 sec/step.

Table 4.1 Properties of TMPA-smectites

 

 

sample - % of CEC doc,
clay name OC% occupied by TMPA (A)
1 SAC-TMPA. 17 1.66 17 14.42, 18.89
2 SAC-TWA38 3.57 38 14.12, 18.97
3 SAC-TMPA.65 5.88 65 14. 16
4 SAC-TMPA” 6.75 75 14.95
5 SAC-TMPALO 8.84 100 14.38

 

Adsorption isotherms of neutral organic compounds on TMPA-smectites were
obtained using batch equilibration technique. Benzene, o-xylene, ethylbenzene, n-
butylbenzene, and trichloroethylene (HPLC-grade) were obtained commercially (Aldrich
Chemical Co., Milwaukee, WI), and used without firrther purification. TMPA-smectites
(0.2g) were weighed into 25-ml Corex glass centrifirge tubes, followed by adding 25 ml of
distilled water. Different amounts of each organic compound were delivered directly into the
tubes as the pure liquid using a nricroliter syringe. The maximum amount of each compound
added corresponded to a final equilibrium concentration of ~50 % of its water solubility. The
tubes were capped immediately with teflon-backed septa and shaken for 24 hours on a
reciprocating shaker at room temperature. Preliminary experiments showed that the

adsorption process reached equilibrium within 18 hours. After equilibration, the tubes were

79

centrifirged at 8000 rpm (RCF=4302 g) for 10 minutes. 5 ml of supernatant was extracted
with 10 ml of carbon disulfide in a glass vial.

Analysis of the organic compounds in the CS2 extracts was by gas chromatography.
A Hewlett-Packard 5890A gas chromatograph was used with flame ionization detector and
a packed column with N2 as the carrier gas. Peak areas were recorded by a Hewlett-Packard
3392A integrator. Concentrations were calculated from the measured peak areas using a
series of external standards. Recoveries were ~90%. Amounts adsorbed were calculated from
differences between the amount of organic compounds added and that remaining in the final
equilibrium solutions. lsothenns were made by plotting the amounts adsorbed against the

equilibrium concentrations in solution.

Results

The adsorption of TMPA cations by, and the release of Ca2+ fi'om, a Cay-saturated
smectite is shown in Figure 4.1. The curves relating TMPA uptake and Ca2+ release to the
equilibrium concentration of TMPA are initially superimposable forming a unique S-shaped
isotherm. At a TMPA content of about 80 cmol/kg, corresponding to about 90% of CEC, the
curves begin to diverge. After this, the amount of TMPA cations bound to smectite exceeds
the amount of Ca2+ released, which plateaus at about 90 cmol/kg. The amount of TMPA
cations bound to smectite increases steadily beyond the CEC of smectite; at an aqueous
TMPA concentration of 2.3 cmol/L, the amount of TMPA cations bound to smectite was
twice the CEC. At this point the amount of TMPA added was 8 times the CEC of the clay.

The lowest equihbrium concentration of TMPA cations in solution appears in the intermediate

80

 

 

   

 

 

 

a 200 I-‘IIIITTTTTI'TTIIII. 200 >
£0 —0— TMPA Adsorbed :: 3 <3:
E 160 —0— Ca Released ‘3' _: 160 ,5,
v I; I
8 120 -"— -‘ 120 E
a, :' I m
5 :E : 8
'8 80 j: - 80 CD
'35 .. I Q
.. :: : ’3‘
§ 40 '3? """"" '5 4o 9.
.. ---------- 1 K
5 o - " 0 g
0.01 0.1 1 10 20 18 16 14
Ce(cmol/L) d-Spacing (A)

Figure 4.1 Adsorption of TMPA cations by and release of Ca2+ from Ca”-
smectite (left), and corresponding basal spacings (right)

exchange degree of Ca2+ with TMPA cations.

Changes in basal spacings with the degree of TMPA exchange was observed (Figure
4.1). At TMPA levels up to about 16 cmol/kg, a single spacing of ~19A was observed.
Between 16 and 38 cmong of TMPA level, two spacings of about 19A and 15A were
present. The intensity of 15A peak increased and the intensity of 19A peak decreased as the
TMPA level increased over this range. At TMPA levels above 38 cmol/kg, a single spacing
of ~15A was present.

Figure 4.2 shows the adsorption isotherms for benzene, o-xylene, ethylbenzene, n-

81

       

 

 

 

   

  
 

so -: o SAC-TMPA.17
50 _: D SAC-TMPA.38
4o _: - SAC-TMPA.65
v SAC-TMPA.75
2° benzene ‘: o SAC-TMPA1.0
o ‘.
0 300 600 900 1200
§ 100 :‘ 50 :‘
E3“ so :5. 4o -:
E 60 -j 30 -:
. - ‘ 3 ,/ .E
g 40 f E 20 .' ' 3
f. 20 2' o-xylene -: 10 :' ethylbenzene 4
5 1 a a
O 0 ‘ 0 ‘
g 0 50 100 150 200 0 40 80 120 160
40 . .
E 60 1
30 1 1
20 _: 40 '5
1o 5 2° - ‘1
n-butylbenzene ; trichloroethylene.
o 1 o ‘

 

 

0481216 0200400600800
Equilibrium Concentration, Ce(mg/L)

Figure 4.2 Adsorption of organic compounds by TMPA-smectites from aqueous
solutions

82

butylbenzene and trichloroethylene. All the organoclays (mixed Ca/TMPA-smectites and
TMPA-smectite) were effective adsorbents manifesting type I isotherms for all solutes except
o-xylene. The occurrence of type I isotherms indicates that adsorption occurred in the
interlayer pores and on the surfaces of the organoclays. The extent of the adsorbate-adsorbent
interaction was difi‘erent for the different adsorbate molecules, as evidenced by the curvature
of the isotherms. The isotherms of trichloroethylene were less curvilinear than those of the
alkylbenzenes, indicating a comparatively weaker interaction between trichloroethylene and
the organoclays than between the alkylbenzenes and the organoclays. The isotherms of o-
xylene were type V, indicating a weaker interaction between o-xylene and the organoclays
than between o-xylene molecules themselves.

The degree of exchange of Ca2+ with TMPA cations also afi‘ected solute uptake. The
adsorptive ability of the TMPA-smectites increased continually as the TMPA content
increased fiom 0.17CEC (17% of the CEC) to 0.65CEC. TMPA contents of higher than
0.75CEC, however, caused a decrease in the adsorptive ability of the clay. As indicated in
Figure 4.2, the adsorptive ability of SAC-TMPAl.0 is lower than that of SAC-

TMPA.65/SAC-TMPA.75.

Discussion

The distribution and interaction of organic cations in the interlayers of smectite have
a pronounced efi‘ect on the adsorption of organic compounds. The exchange reaction of
TMPA cations on smectite and the release of Ca2+ from smectite in this study provided useful

information about the distribution and interaction of TMPA cations in the interlayers of

83

smectite. In this study, the unique S-shaped adsorption isotherm of TMPA, and the extra
adsorption of TMPA over the CEC indicate strong lateral interaction between the adsorbed
TMPA cations. Xu and Boyd (1994, 1995) observed a similar S-shaped isotherm for the
exchange of Na“ but not Ca2+ by HDTMA in smectite. In the case of Ca-smectite, the face to
face aggregation of clay sheets limited access of HDTMA to the interlayers. This promoted
lateral interactions among adsorbed HDTMA even at very low loadings. For Na-smectite, the
clay sheets are fillly expanded. This manifested a more random distribution of HDTMA at low
loadings, and minimized lateral interactions. As the loading increased, lateral interactions
commenced, causing a decrease in the equilibrium concentration of HDTMA, and hence the
S-shaped isotherm. In the case of TMPA exchange onto Ca-smectite, lateral interactions
manifest a collapse of the clay layers. As a result, TMPA cations on opposing clay surfaces
are brought in close proximity to each other, perhaps producing strong it -n interaction
between the phenyl groups. This result agrees with the findings of Jaynes and Boyd (1991)
who proposed that TMPA cations on opposing clay surfaces interact through it -n
interactions, producing a more stable arrangement for the organic cations, and also creating
larger areas of the siloxane surface between ammonium ions for the adsorption of organic
compounds. Washing with distilled water removes the excess of TMPA cations from smectite
over the CBC, as indicated by OC content measurements in Table 4.1.

The S-shaped isotherm of TMPA adsorption also strongly suggests that Ca2+ and
TMPA cations are not segregated in the different interiayers of smectite, although segregation
of organic cations in smectite is commonly observed (McBride and Mortland 1973). As

alluded to above, segregation of HDTMA cations exchanged onto Ca2*- smectite and

84

vermiculite (Xu and Boyd 1994, 1995) occurred due to the limited expansion of the clay
layers and large size of HDTMA cations. The random distribution of TMPA and Ca” cations
in Ca2*-smectite may occur due to the smaller size of TMPA (compared to HDTMA) and
hence greater access to the interlayer exchange sites. The random distribution is evidenced
by the two distinct basal spacings at low and high TMPA levels. This random distribution was
also observed for mixed Ca/TMA-smectite based on the surface area measurements (Lee et
al., 1989).

The smectite surface itself is inhomogeneous, containing a spectrum of charge
densities. This is indicated by changes in basal spacings as a function of TMPA level. TMPA
cations, owing to their small size, are preferentially adsorbed on the surfaces of high charge
density (Faver and Lagaly, 1991), and successively on surfaces of lower charge density.
TMPA-smectite collapses over a range, albeit a narrow one, of TMPA levels. This indicates
that although Ca2+ and TMPA cations are distributed randomly on surfaces with similar
charge densities, they may segregate to some extent in smectite as a whole. Overall, TMPA
cations are neither completely segregated nor randomly distributed. Rather, the distribution
of TMPA cations in the interlayers of smectite could be described as partially random
distribution.

Our results show that partial replacement of Ca2+ in smectite with TMPA cations
significantly increases the adsorptive ability of smectite even though part of siloxane surface
of smectite is still covered by Ca2+ ions. To illustrate this, our results are compared to those
for reduced charge TMPA-smectites having no exchangeable Ca2+ reported by Jaynes and

Boyd (1991). Although adsorption isotherms on TMPA-smectites are better described by the

85

Dubinin-Radushkevich (DR) equation (Sheng et al., 1996), the data were still fit to the
Langmuir equation to allow comparison to the data reported by Jaynes and Boyd (1991). The
monolayer adsorption capacities Q, of benzene, ethylbenzene, n-butylbenzene and
trichloroethylene on the mixed CafI'MPA-smectites and TMPA-smectite described here and
on the reduced charge TMPA-smectites of Jaynes and Boyd (1991) are summarized in Table
4.2. The monolayer adsorption capacities of these organic compounds versus organic carbon
(from TMPA) content are presented in Figure 4.3. The steep initial rise of the curves is
additional evidence that Ca2+ and TMPA cations are not completely segregated, but rather
that some TMPA cations occur in patches. Otherwise, the monolayer adsorption capacity
would be linearly correlated to TMPA content (OC content) in smectite (dashed line in Figure
4.3).

In order to determine the function of TMPA cations in the adsorption of organic
compounds, SAC-TMPA 75 in our study and SAz-0.6Li-250-TMPA (indicates a high-charge
smectite from Arizona, SAz, with lithium added in an amount of 60% of the CEC to reduce
charge at 250°C) fiom Jaynes and Boyd (1991) compared because these two smectites have
similar TMPA contents (6.75% and 6.70% 0C contents, respectively) and the interlayers in
SAz-0.6Li-250-TMPA are not collapsed. The monolayer adsorption capacities of SAC-
TMPA.75 are lower than those of SAz-0.6Li-250-TMPA by ~20% (17.6% for benzene,
17.7% for ethylbenzene, and 23.2% for n-butylbenzene). This is near to but slightly less than
the 25% of the CEC occupied by Ca”. This implies that the siloxane surface of TMPA-
smectites is only partially covered by hydrated Ca” ions, the extent of which roughly

corresponds to the percent of the CEC occupied by Ca”. Hydration of TMPA or the siloxane

86

surface around TMPA cations is not indicated because this would significantly depress their

adsorptive ability of the mixed Ca/TMPA-smectites for neutral organic compounds because

the hydrophobic surface would be no longer available for adsorption.

Table 4.2 Monolayer adsorption capacities of organic compounds on TMPA-smectites

 

monolayer adsorption capacity (mg/g)

 

 

 

organo-clay O.C.%
type ethyl- n-butyl- trichloro-
benzene benzene ethylene
Ca/TMPA-smectite
SAC-TMPA.” 1.66 61.38 38.26 22.40 100.80
SAC-TMPA.38 3.57 74.38 42.66 30.74 126.47
SAC-TMPA.65 5.88 77.29 50.45 36.83 128.49
SAC-TMPA.75 6.75 78.17 48.83 35.54 135.91
SAC-TMPA1.0 8.84 76.95 42.09 28.59 112.70
Reduced charge TMPA-smectite“
SAz-l .0Li-250-TMPA 1.33 64.60 43.51
SAz-0.8Li-250-TMPA 3.42 87.66 40.93 41.67
SAz-0.6Li-250-TMPA 6.70 94.79 59.3 1 46.28
SAz-0.3Li-250-TMPA 9.33 71.92 28.58 25.35

 

* Data fiom Jaynes and Boyd (1991).

To further illustrate this, SAC-TMPA75 is again chosen to calculate the area covered

by dehydrated Ca2+ cations. A crystal radius of 0.99A for Ca” is used in the calculation. In

87

 

 

 

    

 

 

 

 

 

 

 

 

100 80
75 - 60
50 - 40
’6
5’ 25 _ d 20 g
g . , . ' ethylbenzene - g
0 r l l l . l .
2 l l l l g
E o
3 40 - 120 ’3‘
E . g
30 - 90
20 .. 60
10 - so
n-butylbenzene . trichloroethylene q
o L . .1 1 r 1 4 L L 1 r n a I r 1 r 1 L 1 L 0
0 2 4 6 8 0 2 4 6 8 10
Organic Carbon Content (%)

Figure 4.3 Dependence of monolayer adsorption capacity on organic carbon
content on CafTMPA-smectites (square) and reduced charge TMPA-smectites
from Jaynes and Boyd (1991) (circle) for benzene, ethylbenzene, n-
butylbenzene and trichloroethylene

88

SAC-TMPA.75 in which 25% of CEC is occupied by Ca2+ ions, the area covered by
dehydrated Ca2+ would be 2.09 mzlg. Jaynes and Boyd (1991) reported a surface area of
239.6 mzlg for SAz-0.6Li-250-TMPA. If it is supposed that the surface areas of SAC-
TMPA 7 5 without Ca2+ ions and of SAz-0.6Li-250-TMPA are the same or very close, and
the 20% difference of adsorption capacities between SAC-TMPA.75 and SAz-0.6Li-250-
TMPA is produced via covering the siloxane surface by Ca” and water molecules, then the
area covered by water molecules is approximately 45.83 m’Ig (239.6><20%-2.09). Then, the
total number of monolayer water molecules on the surface is ca. 7.66><102° lg (using the
radius of water molecule equal to 1.38 A (Hunt, 1965)). Thus, the ratio of the number of
water molecules to that of Ca” ions in SAC-TMPA.75 is 11.3. If the surface in SAC-
TMPA.75 was completely covered by Ca” and water (2.09 mz/g by Ca”, and 237.5 mzlg by
water), then the ratio of HZO/Ca2+ would be 58.8. Similarly, if the water vapor isotherm on
the same Cay-smectite reported by Boyd et al. (1988b) is used to calculate the ratio of
HZO/Ca2+ in the smectite, the limiting adsorption amount of water vapor on smectite obtained
is ~224 mg/g by extrapolating the linear portion of the isotherm to p/po=1. This adsorption
amount gives the ratio of HZO/Ca2+ 54.5 at p/po=l. This number is very close to that in SAC-
TMPA 75 with the monolayer coverage by water. Thus, the amount of water on the surface
of SAC-TMPA 75 is much less than the monolayer amount, demonstrating that a large part
of the siloxane surface is not covered by water and therefore for the adsorption of organic
compounds. Furthermore, these calculations indicate that the presence of TMPA cations in
the mixed Ca/TMPA-smectites subsequently reduces the amount of water associated with

each Ca2+ ion, and with the surfaces as a whole. Thus, the nature of the exchange cation, not

89

the amount of siloxane surface area per se, dictates the degree to which water is associated
with the clay surfaces.

It appears that upon replacement of Ca2+ by TMPA cations, the overall interaction
between water molecules and Ca2+ ions weakens. The partially random distribution of TMPA
and Ca2+ cations increases the percentage of the surfaces which are covered by the weakly
held water film. As the neutral organic molecules are introduced into the interlayers of
smectite, they are able to displace weakly bound water molecules from the siloxane surfaces
of smectite. However, those water molecules close to Ca2+ ions (in inner region) are still
strongly held by coordination bonds and electrostatic forces. The ratio of HZO/Caz“ of 11.3
suggests that most of the water molecules present in SAC-TMPA.75 are associated with two
hydration shells around Ca2+ since the coordination number for Ca2+ with H20 is equal to 6.

In general, the replacement of Ca2+ by TMPA enhances the adsorptive ability of the
resultant clay for neutral organic compounds. However, TMPA cations may also have a
negative effect on the adsorption of neutral organic compounds at high TMPA levels. At low
TMPA levels, TMPA cations are separated too far to interact with each other. They firnction
only to make the hydrophobic siloxane surface available for adsorption. The higher the
content of TMPA is, the more surface is available for the adsorption of organic compounds.
However, when the amount of TMPA cations in smectite is high enough, i.e. over 75% of
CEC, the size of the adsorptive regions between neighboring ammonium ions diminishes
sufficiently so that some of this surface cannot accommodate the adsorbate molecule. As a
result, the adsorption amount is lower at high levels of TMPA cations than at intermediate

levels.

90

Lee et al. (1989) showed that water molecules shrink the interlarnellar cavities in
TMA-smectite, resulting in a strong selective adsorption of aromatic compounds. This
presumably results fi'om the hydration of TMA cations which have an appreciable hydration
energy of ~32 kcal/mol. TMPA-smectites do not show this water induced shape selectivity,
indicating that TMPA is less strongly hydrated. This maintains the size of the adsorptive
regions between TMPA ions, even in the presence of water. However, adsorbates of sufiicient
size may still be sterically excluded. In TMPA-smectites, all alkylbenzenes with the same
width except o-xylene showed high afinity for the surface. However, the presence of adjacent
substituents in o-xylene molecules increases the size of the molecule in two dimensions and

this creates steric hindrance which leads to the weak adsorbent-adsorbate interaction.

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.
Organic Substances and Sediments in Waters, Vol 1. Baker, R. S. (ed.) p181-200.

Boyd, S.A., Lee, J.-F., and Mortland, M.M. 1988a. Attenuating organic contaminant
mobility by soil modification. Nature 333:345-347.

Boyd, S.A., Mortland, M.M., and Chiou, C.T. 1988b. Sorption characteristics of
organic compounds on hexadecyltrimethylammonium-smectite. Soil Sci. Soc. Am. J.
52:652-657.

Boyd, S.A., Sun, S., Lee, J.-F., and M.M. Mortland. 1988c. Pentachlorophenol
sorption by organo-clays. Clays and Clay Minerals 36:125-130.

Call, F. 1957. The mechanism of sorption of ethylene dibromide on moist soils. .1. Sci.
Food Agric. 82630-639.

Chiou, C.T., Peters, L.J., and Freed, V.H. 1979. A physical concept of soil-water
equilibria for nonionic organic compounds. Science 206:831-832.

91

Chiou, C.T., Porter, P.E., and Schmedding, D.W. 1983. Partition equilibria of nonionic
organic compounds between soil organic matter and water. Environ. Sci. T echnol.
17:227-231.

Faver, 11., and Lagaly, G. 1991. Organo-bentonites with quaternary alkylammonium
ions. Clay Minerals 26:19-32.

Hunt, J.P. 1965. Metal Ions in Aqueous Solution. WA Benjamin, Inc. New York.
Hanson, W.J., and Nex, R.W. 1953. Diffusion of ethylene dibromide in soil. Soil Sci.
76:209-214.

J aynes, W.F., and Boyd, S.A. 1990. Trimethylphenylammonium-smectite as an effective
adsorbent of water soluble aromatic hydrocarbons. .1. Air Waste Manage. Assoc. 40: 1649-
1653.

Jaynes, W.F., and Boyd, S.A. 1991. Hydrophobicity of siloxane surfaces in smectite as
revealed by aromatic hydrocarbon adsorption from water. Clays and Clay Minerals
392428-436.

Lee, J.-F., Mortland, M.M., Boyd, S.A., and CT. Chiou. 1989. Shape-selective
adsorption of aromatic molecules from water by tetramethylammonium-smectite. J.
Chem. Soc., Faraday Trans. 1 85:2953-2962.

Lee, Jo'Fo, Mortland, M.M., Chiou, C.T., Kile, D.E., and Boyd, S.A. 1990.
Adsorption of benzene, toluene, and xylene by two tetramethylammonium-smectites
having difi‘erent charge densities. Clays and Clay Miner. 38:113-120.

McBride, M.B., and Mortland, M.M. 1973. Segregation and exchange properties of
alkylammonium ions in a smectite and vermiculite. Clays and Clay Minerals 21 :323-329.

McBride, M.B., Pinnavaia, T.J., and Mortland, M.M. 1977 . in Fate of Pollutants in
the Air and Water Environments. John Wiley and Sons, New York. Part I, p145.

Mortland, M.M., Sun, S., and Boyd, S.A. 1986. Clay-organic complexes as adsorbents
for phenol and chlorophenols. Clays and Clay Miner. 342581-585. ,

Sheng, G., Xu, S., and Boyd, S.A. 1996. Surface heterogeneity of
trimethylphenylammonium-smectite as revealed by adsorption of aromatic hydrocarbons
fiom water. (Submitted to Clays Clay Miner.)

Spencer, W.F., Cliath, M.M., and Farmer, W.J. 1969. Vapor density of soil-applied
dieldrin as related to soil-water content, temperature, and dieldrin concentration. Soil Sci.
Soc. Am. Proc. 33:509-511.

92

Spencer, W.F., and Cliath, M.M. 1970. Desorption of lindane fiom soil as related to
vapor density. Soil Sci. Soc. Am. Proc. 34:574-578.

Stark, F.L., Jr. 1948. Investigations of chloropicrin as a soil fumigant. New York
(comell) Agr. Exp. Sta. Mem. 178:1-61.

Xu, S., and Boyd, S.A. 1994. Cation exchange chemistry of hexadecyltrimethylarnmo-
nium in a subsoil containing vermiculite. Soil Sci. Soc. Am. J. 58: 1382-1391.

Xu, S., and Boyd, S.A. 1995. Cationic surfactant sorption to a vernriculitic subsoil via
hydrophobic bonding. Environ. Sci. T echnol. 29:312-320.

Chapter 5 Conclusions and Significance

Sorption of neutral organic contaminants by organoclays

Naturally occurring aluminosilicates (e.g. smectites and verrniculites) do not sorb
significant amounts of hydrophobic organic contaminants (HOCs) in presence of water due
to the preferential adsorption of water around the metallic exchangeable cations (Prost,
1975). The sorptive characteristics of soils and clays for neutral organic compounds (N OCs)
are dramatically improved when the inorganic exchangeable cations are replaced by organic
cations such as quaternary ammonium cations (QACs) (Lee et al., 1989a; 1990; Boyd et al.,
1988 a, b; 1991; Jaynes andBoyd 1991a). For example, as shown in Figure 5.1, sorption
coefficients for several common organic groundwater contaminants (e. g. benzene, toluene,
xylene) are increased by over 2 orders of magnitude by hexadecyltrimethylammonium
(HDTMA) modification of a subsoil containing vermiculite and illite as major clay nrinerals
(Boyd et al., 1988a; Lee et al., 1989a). The large increase in sorptive capability of modified
soil is partially due to increase in organic carbon content resulting fi'om HDTMA adsorption,
and partially due to the fact that HDTMA-derived organic matter is 10-30 times more
effective than natural organic matter for sorbing NOCs (Lee et al., 1989a).

The sorption properties of a given modified clay or soil depend in part on the nature
of the organic modifiers. Replacing the inorganic exchangeable cations with small organic
cations such as tetramethylammonium (TMA), trimethylphenylammonium (TMPA) and

tetrarnethylphosphonium (TMP) increases the sorption of NOCs (Lee et al., 1990; Jaynes and

93

Soil Concentration, 0 (nln/ku)

 

 

 

     
            
    
    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9000 Ie.v,.,-. 5400 eti.trr 21001111.,.,.,r
r- a .. l- c
O
7500 - r 4500 r- 4 1750 " T
. . 1 . _ sr.crair81 ,
°/. Clay
0 ... _. .. _. ._
60 0 3600 K=34.5 1400
b -I b J "‘
4500 - ~ 2700 - ~ 1050 —-
I- . p -
3000- - 1800- - 700 —
1500 ._ . benzene _ 900 toluene 350 ethylbenzene
- 'l
o A 1.1 L l! ‘1 I." L o a
0 100 200 300 400 .500 600 0 70 140 210 280 350 420 0 20 40 60 80 100 120
aooo'riltlrflifi ‘55":I'irl'l'1r 600a: ‘l‘l'l'
.d . e I . l
2500 ~ - 1375 — 500 - _
- benzene - ~ toluene - - ethylbenzene -
2000 - - 1100 — - 400 - _
. . .
1500- - 825- - 300— -
F . " d
1000 -- - 550 - - 200 - Oshtemo Bt -
i- +HDTMA-Treated . _ . - 6.3% Clay .
soo— ' — 275- - loo -—
O
- {Untreated 4 - . .
0 Li 14L 1 L r L g 1“ o h 14 1 14 1 1- [A1 1-1 u 1 L 1 1 a
0 120 240 360 480 600 720 0 35 70 105 140175 210 0 25 50 75 100 125 150

Aqueous Phase Concentration. C [mg/l)

Figure 5.1 Enhanced sorption of common groundwater contaminants following
treatment of soil with hexadecyltrimethylammonium. Sorption isotherms of
benzene, toluene, and ethylbenzene from water by untreated and treated St. Clair
(a-c) and Oshtemo (d-f) Bt horizon soils are shown. Reproduced from Lee et al.
(19893)

95

Boyd, 1990, 1991b; Kukkadapu and Boyd 1995). The sorption is characterized by
Langmuir-type isotherms, indicating surface adsorption as the primary retention mechanism
(Jaynes and Boyd 1991b). These clays are commonly referred to as "adsorptive organoclays"
(Boyd et al., 1991). The fact that the amount of NOC adsorbed by TMPA-smectite was
directly proportional to the surface area of the organoclay but inversely proportional to the
layer charge of the clay or its organic carbon content (Lee et al., 1990; Jaynes and Boyd
1991b) demonstrates that either direct interaction of sorbed NOCs with the exchangeable
TMPA cations does not occur, or, additional mechanism(s) occur simultaneously (Jaynes and
Boyd 1 991 b). Jaynes and Boyd (1991b) proposed that in the presence of bulk water the
primary adsorptive sites for aromatic hydrocarbons are the uncharged siloxane oxygens, and
that small organic cations such as TMPA function primarily as pillars to prop open interlayers
and satisfy the cation exchange capacity. The above interpretation is consistent with the
arguments presented by others that uncharged siloxane surfaces in montmorillonites are
hydrophobic, and that the hydrophilicity of untreated montmorillonites is mainly due to the
strong hydration of exchangeable metal cations (Prost, 1 9 75 ,° Sposito and Prost, 1982). The
neutral surfaces of natural clays are apparently obscured by water which hydrates the metal
exchangeable cations. Replacing these highly hydrated metal cations with much less hydrated
organic cations exposes the neutral surface between the organic exchange cations, leading to
the adsorption of NOCs (Jaynes and Boyd 1991b). In this model, the adsorptive organoclays
are viewed as a porous adsorbent. The dimension of those micropores is fixed by the size of
the QACs, the charge density of the clays and by the distribution of the QACs in the interlayer

regions. This is important because the size of the micropores may impose certain steric

96
constraints on organic contaminant adsorption (Lee et al., 1989b, 1990).

For short chain QACs, cation exchange is the singular adsorption mechanism. The
distribution of the short chain QACs in the interlayers should correspond closely to the
distribution of negatively-charged sites. Due to the heterogeneity of layer charge distribution,
the distribution of QACs such as TMPA in interlayers of smectite will be heterogeneous. This
manifests difi‘ererrt pore sizes within a given adsorptive organosmectite (Sheng et al., 199_).
The heterogeneity of pore sizes influences the sorption of organic contaminants of difl‘erent
molecular dimensions. For example, among benzene and several alkylbenzenes, benzene can
be adsorbed inside both small and large micropores of TMPA-smectite, manifesting the
largest sorption capacity among the solutes and the most heterogeneous distribution of
adsorption energies. In contrast, larger molecules like propylbenzene can be adsorbed only
in the larger pores. As a result, the sorption capacity of organosmectite for propylbenzene
is only 23% of that for benzene (Sheng et al., 199_). In addition, the sterically restricted
adsorption of propylbenzene resulted in a more narrow distribution of adsorption energies.
This explains the observation that the adsorption isotherm of benzene does not exactly fit the
Langmuir equation, which assumes a uniform adsorption energy, while that of propylbenzene
does.

The above discussion has focused on organoclays in which the clay is completely
saturated with short chain QACs. There may be instances where the clay is only partially
saturated with QACs because the last 10-20% of the exchange sites on layer silicates such as
montmorillonite have comparatively low aflinities for QACs (Xu and Boyd 1995a). It has

been demonstrated that adsorption capacity of TMPA-smectite for benzene, alkylbenzenes

97

and TCE increases dramatically with the fraction of exchange sites occupied by TMPA until
loading reaches 65%-75% of the CEC (Sheng, 1996). Further increases of TMPA loading
beyond this level actually decreases the adsorption capacity of the organoclays.

Replacing the inorganic exchangeable cations of clays with long chain organic cations
such as HDTMA results in an organoclay with fundamentally difi‘erent sorptive properties
than adsorptive organoclays (Boyd et al., 1988 a, b, 1991; Jaynes and Boyd 1991a; Lee et
al., 1989a). Sorption of NOCs from water by these organoclays manifests type IH (concave)
or double-sigmoid isotherms (Figure 5.2), suggesting different mechanisms may be
responsible for sorption of NOCs. These organoclays are commonly referred to as
organophilic clays. In organophilic clays, the large organic moieties (e. g. C-16 alkyl moieties
of HDTMA) fixed on the mineral surface interact to form an organic phase. For HDTMA
modified clay, the micropolarity sensed by a fluorescent probe, pyrene, decreased as HDTMA
loading increased, suggesting that modification resulted in the formation of an organic phase
much less polar than the untreated clay surface (Brahimi et al., 1992; Kalyanasundarwn and
Ihomas, I 9 7 7).

The hydrophobic phase of such clays functions as a powerful partition medium for
NOC sorption. Compared to linear isotherms for natural soil organic matter (Figure 5.2),
Type III isotherms result fi'om the high sorption ability of the hydrophobic phase for NOCs.
In these organoclays, the weight of the sorbed compound, e.g. chlorobenzene, may exceed
that of the HDTMA phase itself (Figure 5.2). The presence of such high amounts of a sorbed
NOC increases the solvency of the hydrophobic phase for the NOC, which in turn manifests

a Type III isotherm.

98

 

 

 

 

 

50 I. I I I I I I I. I I I I I I I l I I I I l I T I I 0.5
:m“°" 5°" 5 8.13%? has“ 3
. I trichloroeth Iene .
A 40 _ SACH D chlorobenz ne _ 0 4
0’ - A trichloroeth lene " 1 '
>8 - SAZH A ch orobenz ne , , .
D ' ‘ d
~5- 3o — 2‘ - 0.3
8 : . . ‘ :
O - . '4 ..
<13 20 - - 0.2
C P O A ‘ '
é ” .' I
< 10 - 1 0.1
I ‘ c g 1
o - ' ._._ -._. .—l r r I r r 0.0
0.0 0.2 0.4 0.6 0.8 1.0

Solute Relative Concentration

Figure 5.2 Type III isotherms for trichloroethylene and double-sigmoid
isotherms for chlorobenzene sorption by low-charge
hexadecyltrimethylammonium (HDTMA)-smectite (SACH) and high-charge
HDTMA-smectite (SAzH), compared to the linear isotherms for trichloroethylene
and chlorobenzene sorption by a muck soil. Data from Sheng et al. (1996a)

The function of the hydrophobic phase also depends on the arrangement of organic
cations in the interlayers of clay, which in turn, depends on the charge density of the clay. For

vermiculite, long chain QACs such as HDTMA adopt a vertical arrangement as illustrated in

99

Figure 5.3 (paraflin type). The tails of adsorbed HDTMA are highly flexible and the
hydrophobic phase functions as a solubilizing medium for NOCs analogous to a bulk organic
solvent. Solute partitioning has been suggested as a major HOC sorption mechanism for this
type of organophilic clays, based on the observation of linear sorption isotherms. Our careful
study using organoclay synthesized fiom both non-swelling and swelling 2:1 clay minerals
with different layer charges demonstrates that double sigmoid adsorption isotherms are
commonly observed for organophilic clays (Figure 5.2). The double sigmoid isotherms can
be explained by combination of two sorption mechanisms, which are related to the
arrangement of QACs in the interlayer region (Sheng et al., 1996a). In low charge smectites,
HDTMA forms a flat-lying bilayer in the interlayers. The tails of the adsorbed HDTMA
interact strongly with the mineral surfaces and with each other. The hydrophobic phases are
relatively rigid and differ in this regard from a fiee solvent phase. As a result, NOC sorption
at low relative solute concentrations (aqueous concentration/water solubility) is mainly due
to the solvation of the adsorbed HDTMA and the mineral surfaces. Inserting organic
contaminants between the hydrophobic "tails” of adsorbed HDTMA and the mineral surfaces
is enhanced as solution concentrations increase. Ultimately, this results in detachment of the
tails from surfaces and increases the d-spacing of the organoclay. As tails become more
flexible, a more continuous solvent-like hydrophobic phase forms in the interlayer, and
partitioning becomes a predominant sorption mechanism. The solvation and swelling is more
easily achieved in high-charge smectite (e. g. SAz-l), in which the HDTMA tails adopt a
pseudotrimolecular arrangement, manifesting weaker surface-tail interactions. Thus, the NOC

sorption isotherm for HDTMA-SAz-l is more linear at low relative concentration than that

100

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. ,meeeem b waeanm\
W33"?
8 ii '85 wee-ease“
Mr W...

 

 

c \\\\§&s>ie~laia\\\\m (if? ¥W\\\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\\\\s\®aa\ake<\\\\‘
Pseudotrimolectular Layer Paraffin Complex
A~12.1A A>12.4A

Figure 5.3 Possible interlayer arrangements of hexadecyltrimethyl-
ammonium and corresponding interlayer spacings (A). Reproduced from
Jaynes and Boyd (1991a).

101
for HDTMA-SWy-l.

The sorptive characteristics of organoclays in single organic solute systems may be
difi‘erent fi'om those in multi-solute systems, the latter being more representative of actual
contaminated sites. These difi'erences may arise, for example, if sorption of one solute causes
interlayer expansion, or a change in the solvency of organic phase in the interlayers. We
found the sorption of one contaminant may have synergistic or antagonistic effect on the
sorption of another, depending on the nature of the co-solute (Sheng et al., 1996b). For
example, the sorption of trichloroethylene (TCE) by HDTMA-modified smectites is much
lower in a single solute system than that in the presence of other solutes such as nitrobenzene
and carbon tetrachloride. Similarly, the presence of TCE increased the sorption of
nitrobenzeneand carbon tetrachloride. In contrast, the presence of ethyl ether decreased the
sorption of TCE by HDTMA-smectites. The synergistic efi‘ects occur because the cosolute
caused interlayer expansion and/or increased solvency of the sorptive phase for the primary
sohrte. Depending on the nature of the cosolute, the solvency of the sorptive phase may also
decrease, as in the case of ethyl ether, causing an antagonistic effect. However, sorption of
most common NOCs should cause an increase in the solvency of the sorptive phase of
organophilic clays (e.g. HDTMA-smectite) and hence produce a synergistic, rather than

antagonistic, efl‘ect on contaminant removal from water.

Application of organoclays to pollution abatement
Based on the thorough understanding of the sorptive properties, organoclays may be

efi‘ectively used in pollution abatement.

102

Formation of organoclays involves the use of surfactants. Cationic surfactants such
as QACs are combined with layer aluminosilicate clay minerals to form organoclay. The
resultant organoclays have greatly enhanced sorptive capabilities for a variety of organic
cOntaminants and are able to immobilize hydrophobic organic contaminants dissolved in
water. In a similar application, anion surfactants such as sodium dodecyl sulfate (SDS) are
used to replace the native exchangeable anions on iron and aluminum oxides (Park and Jafi'é,
1993, 1995). SDS adsorbed on oxide surfaces forms hemimicelles (F uerstenau, 19 70) that
substantially enhance the capability of oxides to sorb hydrophobic organic compounds fi'om
water via partitioning into the hernimicelles. The primary application of such modified oxides
is water treatment. An interesting point is that some iron oxides (e. g. magnetite and
maghemite) are magnetic and thus SDS-treated iron oxides can be separated from water or
soil suspension with magnets (Park and Jafi'é 1995).

Difi‘erent organic cations such as quaternary ammonium- (Boyd et al. 1988b, 1991),
phosphonium- (Kukkadapu and Boyd 1995), and alkylpyridinium- (Wagner et al., 1994)
compounds may be used to produce organoclays. The resultant organoclays can have vastly
improved and unique sorptive properties towards organic contaminants, depending on the
nature of both organic cations and the types of layer silicates used. Importantly, organoclays
have potential of being used both ex-situ and in-situ. For example, organoclays can be used
in place of or in conjunction with activated carbon for water purification (Beall, 1985).

Contaminant transport in soils is usually dependent on the extent of contaminant
sorption and the hydraulic conductivity of the soil. The hydraulic conductivity may, however,

be sufficiently low in landfill barriers such as clay liners and bentonite slurry walls, that

103

molecular difiirsion of contaminants becomes a dominant transport mechanism. Addition of

organoclays to landfill baniers will retard the mobility of contaminants through landfill

barriers by greatly enhancing the extent of contaminant sorption. The inclusion of Na-

bentonite in such barriers is needed because they form highly dispensed particles in water.

These small particles fill voids in the liner material (e. g. soil) and effectively reduce hydraulic

conductivity.

Gullick et al. (1995) used the following one-dimensional contaminant transport model

under steady-state, uniform flow with equilibrium sorption in saturated, homogeneous,

isotropic landfill barriers (Weber et al., 1991) to predict the effect of TMPA-bentonite

addition to landfill barriers on contaminant transport:

where: C

X

and t

E
at

2
=Dh.a_c.-vx£-p. (1-8) a_q a_C..
ax’ 6x 9 6C 01

 

contaminant concentration;

coeflicient of hydrodynamic dispersion;
average linear groundwater velocity;
density of landfill barriers;

sorbed concentration;

porosity of landfill barriers;

distance in the x direction;

time.

[5.1]

The third term on the right side of the above equation describes the effect of sorption on

contaminant transport in landfill baniers. Figure 5.4 shows the model predictions, under given

104

 

  
  

 

 

 

 

1.0 , —
I
‘ I
I
0'8 No 0.1% I
~—- Sorption TM PA-bent. I’
o 0.6 — I
O l
a I
0.4 — ’
I
—- l
I
0.2 - I
I
_ I
0.0 J I 11111 1 llllluJ/ 11111111

1 1 O 100 1 000
Time (years)

Figure 5.4 Simulated transport of 1,2,4-trichlorobenzene through
bentonite liners with and without 0.1% trimethylphenyl (TMPA)-bentonite
added. Conditions: Kc = 10‘7 cm/sec, dhldl = 1, D, = 2.26 x 10‘ cmzlsec, e
= 0.40, pb = 1.60 g/cm3, x = 1m, Co = 5000 ug/l, Freundlich adsorption
parameters K, = 22.67 and n = 0.808. Reproduced from Gullick et al.
(1995). -

105

conditions, of transport of 1,2,4-trichlorobenzene through landfill barriers with and without
addition of TMPA-bentonite. The addition of 0.1% TMPA-bentonite to landfill barriers
reduces 1,2,4-trichlorobenzene transport by tens of times. Smith and Jaffé(1994) evaluated
benzene transport through landfill liners containing organo-bentonites. In this study, Ottawa
sand (88%) was amended with untreated Wyoming bentonite (8%) (primarily Na-bentonite)
and either 4% untreated bentonite or 4% organobentonite. The mixtures were compacted to
simulate sand-bentonite landfill liners. Hydraulic conductivities of the organobentonite
composite liners were low, all on the order of 10" cm/sec. A one-dimensional solute transport
model was used to simulate benzene transport in conventional sand-bentonite liners and in
liners containing two sorptive organo-bentonites, namely HDTMA-bentonite and
benzyltrimethylammonium (BTMA)-bentonite. The addition of the organobentonites
substantially delayed maximum benzene breakthrough fi'om about four years in the
conventional liner to about 275 years in the organo-bentonite amended liners (Figure 5.5).
Perhaps the most unique feature and important advantage of this chemistry is that it
can be applied in-situ. Subsoils and aquifer materials contain negatively charged clay minerals
and hence possess a cation exchange capacity (CEC). In earlier research, Boyd et al.
demonstrated that sorption of common organic groundwater contaminants by aquifer
materials or soils can be increased by at least two orders of magnitude by using cationic
surfactants to convert soil clays into sorptive organoclays (Boyd et al., 1988a, 1991; Lee et
al., 1989). These results suggested the possibility that aquifer materials or subsoils could be
modified in-situ via injections of cationic surfactant solutions, and the modified soil materials

could function as "sorptive zones. " Such sorptive zones, if pr0perly placed, could intercept

106

 

2000

 

 

 

 

 

 

 

 

 

   
    
 
 
 

 

 

 

 

C A
1500 T v = 9.67x 70‘5 m/day
: Fi_—.1.44
A1000:
3.1.3 :
:1 500—
\ ..
m .
g Obtrrl111111111’1’11'1'4“
a o 5 10 15 2o 25
811200
t. . B
.9 - '
g 900— v: 1.94 X 10'4 m/a'ay
: . Fi=6.38
£2 600—
E
6 300
O h
c:
O OlqlrIOLJ—l1lll'!l""Llfi
O o 20 40 so 80 100
93 so 0
63 :
E1 40:
CD I v = 2.29 x 10‘4 m/day
CD 30-
20}-
10:
0C , ,,,!,,.,ltnn.
O 500 1000 1500 2000

Time (Years)

Figure 5.5 Benzene transport through landfill liners containing organophilic
bentonite: '(A) 88% sand and 12% Na-bentonite; (B) 88% sand, 8% Na-bentonite
and 4% hexadecyltrimethyl-ammonium-bentonite; and (C) 88% sand, 8% Na-
bentonite and 4% benzyltrimethylammonium-bentonite. R is the retardation factor,
v is the linear velocity. Reproduced from Smith and Jaffé (1994).

107

and immobilize contaminant plumes containing dissolved organic chemicals as shown
schematically in Figure 5.6. The immobilized contaminants could then be detoxified by
various chemical or biological means, as for example through bioremediation using native
microbial populations or introduced organisms. Containing contaminant migration in this
fashion has the important advantages of preventing firrther downgradient aquifer
contamination, and of concentrating contaminants in a defined zone that can be managed to
enhance remediation. For example, in the case of bioremediation, nutrients and/or oxygen

could be added to stimulate microbial activity.

108

 

      
 

 

 

 

 

 

 

 

 

 

 

 

 

l e\e) aamuemaaaacfl
geafiaaegafiwnm
«a? u 5:181)“. “madame.
”than? u linflM/l/
act—Ewan O W"”.”.WWM//>o DE;

 

 

 

 

I .rorIVIOII/
I uwnflfltoflfllflc confide;

 

 

 

 

 

              

 

   

fer material to create contaminant

sorption zone, and coupled sorption and biodegradation of organic contaminants

for groundwater cleanup.

’9 .
. 9’ " ’/
l.|I l Nuisances» l m $980,448.”?

 

     

  

‘IIIaII
IIIIII.
I'll-I
II‘I
all-III.-
Illllr’a
I‘ll-II;
-IIII:
ll'llr
Illlll.
I'll;

Ills"; lull
I'llllo Ill
Illlrlv ll"
11": ll-

lull"
‘lllll
II I
II.

(III

managed...
e.«snow-«oueuoueuoueuouev ll
o99909040909oeeoeooeeeoeeoom
00909999009900,
voodoo.90909999999999.0900.on
.(909909009009900900099999...
. . I.ooeoeeeooeooooeeoeooosox
Al‘s—PDT: peleeueuewenfise .
ensues. =8 .2 283 ..

=o>> 20:00.5. UESm

 

 

 

111”

 

 

Figure 5.6 Proposed in-situ modification of aqu

109

References

Beall, G.W. 1985. Process for treating organics contaminated water. US. Patent
P4, 5 17,094.

Boyd, S. A., Lee, J.-F., and Mortland, M.M. 1988a. Attenuating organic contaminant
mobility by soil modification. Nature 33:345-347.

Boyd, S. A., Mortland, M.M., and Chiou, C.T. 1988b. Sorption characteristics of
organic compounds on hexadecyltrimethylammonium-smectite. Soil Sci. Soc. Amer. J.
52:652-657.

Boyd, S. A., Jaynes, W.F., and B. S. Ross. 1991. Immobilization of organic
contarrrinants by organo-clays: application to soil restoration and hazardous waste
contaminants, pp181-200. In: R A Baker (ed) Organic substances and sediments in
water, vol. 1, Lewis Publishers, Chelsea, MI.

Brahimi, B., Labbe, P., and Reverdy, G. 1992. Study of the adsorption of cationic
surfactants on aqueous laponite clay suspensions and laponite clay modified electrodes.
Langmuir 8: 1908-1918.

Fuerstenau, D. W. 1970.1nterfacial processes in mineral/water systems. Pure Appl.
Chem. 24:135-164.

Gullick, R.W., Weber, W.J., Jr., and Gray, D.H. 1995. Organic contaminant transport
through clay liners and slurry walls. In: Reactions of organic pollutants with clays. The
Clay Minerals Society Pre-meeting Workshop. Baltimore, MD.

Jaynes, W. F., and Boyd, S.A. 1990. Trimethylphenylammonium-smectite as an efi‘ective
adsorbent of water soluble aromatic hydrocarbons. .1. Air Waste Mgmt. Assoc. 40: 1649-
1653.

Jaynes, W. F., and Boyd, S.A. 1991a. Clay mineral type and organic compound sorption
by hexadecyltrimethylammonium-exchanged clays. Soil Sci. Soc. Amer. J. 55 :43 -48.

Jaynes, W. F., and Boyd, S.A. 1991b. Hydrophobicity of siloxane surfaces in smectites
as revealed by aromatic hydrocarbon adsorption fiom water. Clays Clay Miner. 39:428-
43 6.

Kalyanasundaram, K., and Thomas, J.K. 1977 . Environmental effects on vibronic band
intensities in pyrene monomer fluorescence and their application in studies of micellar

110

systems. J. Am. Chem. Soc. 99:2039-2044.

Kukkadapu, RK, and Boyd, S.A. 1995. Tetramethylphosphonium- and
tetramethylammonium-smectites as adsorbents of aromatic and chlorinated hydrocarbons:
effect of water on adsorption efficiency. Clays Clay Miner. 43:318-323.

Lee, J.-F., Crum, J.R., and Boyd, S.A. 1989a. Enhanced retention of prganic
contaminants by soils exchanged with organic cations. Environ Sci. T chnol. 2321365-
1372.

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

Lee, J.-F., Mortland, M.M., Chiou, C.T., Kile, D.E., and Boyd, S.A. 1990.
Adsorption of benzene, toluene and xylene by two tetramethylammonium-smectites having
difl‘erent dharge densities. Clays Clay Miner. 38:113-120.

Park, J.-W., and Jafi'é, J.R. 1993. Partitioning of three nonionic organic compounds
between adsorbed surfactants, nricelles, and water. Environ. Sci. Tech. 27:2559-2565.

Park, J.-W., and Jaffé, J.R. 1995. Phenanthrene removal from soil slurries with
surfactant-treated oxides. J. Environ. Engr. 121 :430-43 7.

Prost, R. 1975. Interactions between adsobed water molecules and the structure of clay
minerals. In: S. W. Bailey (ed.) Proc. Int. Clay Conf. 1975. Applied Publishing Ltd.,
Wilwette, Illinois.

Sheng, G. 1996. Function of trimethylphenylammonium cations on smectite for
adsorption of aqueous organic contaminants. Chapter 4. PhD Thesis. Michigan State
University. East Lansing, Michigan.

Sheng, G., Xu, S., and Boyd, S.A. 1996a. Mechanism(s) controlling sorption of neutral
organic contaminants by surfactant derived and natural organic matter. Environ. Sci.
Technol. 30:1553-1557.

Sheng, G., Xu, S., and Boyd, S.A. 1996b. Cosorption of organic contaminants fiom
water by hexadecyltrimethylammonium-exchanged clays. Water Res. 30: 1483-1489.

Sheng, G., Xu, S., and Boyd, S.A. 199_. Surface heterogeneity of
trimethylphenylammonium-smectite for adsorption of aqueous aromatic hydrocarbons.
Clays Clay Miner. (submitted)

111

Smith, J.A., and Jaffé, RR. 1994. Benzene transport through landfill liners containing
organophilic bentonite. J. Environ. Engr. 120:1559-1577.

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

Wagner, J., Chen, B., Brownawell, B.J., and Westall, J.C. 1994. Use of cationic
surfactants to modify soil surfaces to promote sorption and migration of hydrophobic
organic compounds. Environ. Sci. Tech. 28:231-237.

Weber, W.J., Jr., McGinley, P.R., and Katz, LE. 1991. Sorption processes and their
efi‘ects on contaminant fate and transport in subsurface systems. Water Res. 25:499-528.

Xu, S., and Boyd, S.A. 1995. Cationic surfactant sorption to a vermiculite subsoil via
hydrophobic bonding. Environ. Sci. Tech. 29:312-320.

ICH RN STRT N V.

to E u r Lraannres
llll ill WI Hll llllllllillilllWll
74009

312930156