:4
F. a.
733:: \.
.|I‘

.1...

I ‘
Vivi:
O

in...
a

x.
a.

 

 

 

 

 

 

'

u;

‘4. .
. .1...

z. a.
«b (—
..iir .

1r. .343?

mmmwmfi .
A In. .
...L .

i
.115:
.3... 45...}
;..

a. 1&4 . .
imaufipmvfiwn.

.

0 .1}.

.1“. Livia.“ ..

”in .. Pl. «My»?

1.42:5. .xi
1.3,.
x .

4‘
4:3!

“03.03

35......
v if

. uni.
Plkfl

. I~l:‘ .EI.

_ A. $32.

.. a.” . uh.
tub.) {XLKH {zit
fl 1 x a .

..

“mum."
Jr“

Hutu. ..
BUR
.

a 4.—

, mW
., aw

if?

9.". ‘1.

affiuris

an“? .aLL: u. x
r) a. \
v..

\ :r.
:U «I

 

.91
a.

at:
am

'1

 

I, LIBRARIES
;-i..._.,:;--_.;; MICHIGAN STATE UNIVERSITY
2 EAST LANSING, MICH 48824-1048

4* av
;__. I};

/géfil*/33’7

This is to certify that the
dissertation entitled

QUANTIFICATION OF THE AVAILABILITY OF CLAY
SURFACES IN SOILS FOR SORPTION OF
NITROAROMATICS

presented by

Simone Melanie Charles

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Crop and Soil Sciences-
Environmental Toxicology

 

 

/« //2//

ajor Professor’SvSignature

17th December, 2004

 

Date

MSU is an Affinnative Action/Equal Opportunity Institution

-.-.-.-.-.-.-.-,-.-I—._ —~. o---c-o—4--—.--.-.-.-.-.-~-o-p--

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

1

10:1 t1 sfizéhét

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DatoDuo.p65-p.15

 

QUANTIFICATION OF THE AVAILABILITY OF CLAY SURFACES IN SOILS FOR

SORPTION OF NITROAROMATICS

Simone Melanie Charles

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Crop and Soil Sciences

2004

ABSTRACT
QUANTIFICATION OF THE AVAILABILITY OF CLAY SURFACES IN SOILS
FOR SORPTION OF NITROAROMATICS
By

Simone Melanie Charles

Nitroaromatics (NACS) comprise an important class of environmental contaminants that
pose a potential toxic threat to ecological and human health. To improve our
understanding and modeling of the fate and transport of NACS in the environment, the

effect of the interaction of soil clays with SOM must be better understood.

The influence of exchangeable cation on the adsorption of p-nitrocyanobenzene (p-NCB)
and 1,4-dinitrobenzene (1,4-DNB) by Webster A and B horizon soils, which contains
smectite as their most abundant clay mineral, was determined. On removal of SOM from
whole soils, sorption of p-NCB and 1,4-DNB increased indicating an overall negative
contribution of SOM to their sorption by whole soil. The removal of the clay-Sized
fraction from whole soil decreased sorption of these compounds by a factor of at least 10.
This suggests that the interaction between SOM and soil clays suppresses sorption of p-
NCB and 1,4—DNB by soils. We also determined, independently, the effect of
exchangeable cation on sorption of 1,4-DNB by SOM and on sorption of p-NCB by
whole soil, the clay-sized fraction from that soil and SOM. No effect of exchangeable
cation type was found for sorption of p-NCB and 1,4-DNB by SOM. For homoionic clay-

sized soil fractions, the sorption of p-NCB was generally greater by monovalent-cation-

exchanged clays than divalent-cation-exchanged clays, and increased with decreasing
cation hydration energy (Li+~Na+<K+<Cs+ and Ca2+~Mg2+<Ba2+). The same trends were

observed for homoionic soils exchanged with monovalent and divalent cations.

Two methods were used to assess the availability of clay surfaces (fa) in soils for sorption
of NACS. One method offered earlier by Karickhoff (1984) involves the summation of
the sorption of solutes to SOM and swelling clays. We also developed an alternative
approach for determining fa that alleviates methodological limitations of the Karickhoff
approach. Using three NACS (p-nitrocyanobenzene (NCB), 2,4-dinitrotoluene (DNT) and
1,4-dinitrobenzene (DNB)) and four loam soils, sorption of each NAC by whole soil and
soil from which SOM was removed was determined. For each NAC, increased sorption
was observed with the removal of SOM from whole soils indicating diminished
availability of mineral surfaces due to obscuration or blockage by SOM; the fa values
were less than 1. For each NAC, we found a strong correlation between the f, values and
the relative contents of SOM and smectite in the soils (> 0.60). For each soil, the fa values
decreased in the order DNT > DNB > NCB. This suggests that the nature of the NAC
may also influence the mechanism driving sorption site availability. We found a strong
correlation between the fa values and the strength of sorption of each NAC to smectites (_>_
0.90 in three of four cases). Our studies reveal that for organic solutes with polar
functional groups (e.g., N02) and in soils containing smectite clays, solute uptake from
water will be strongly influenced by the type of exchangeable cation(s) present, the
relative contents of SOM and sorptive clays (e.g., smectites) in the whole soil as well as

the nature of the solute.

ACKNOWLEDGEMENTS

“In the confrontation between the stream and the rock, the stream always wins — not

necessarily through strength, but through persistence ” Anonymous

The process of this study entailed character building as well as academic achievement. I
am certain that I will continue to grow professionally and as an individual. My
appreciation extends to all those with whom I have interacted during the course of my

studies; such interactions facilitated my learning and growth.

I specially thank my advisor, Dr. Stephen Boyd, for his guidance, financial support, and
encouragement. I have grown professionally through his leadership and I truly was

honored to work with him.
I extend gratitude to Dr. Brian Teppen for his constant encouragement and support during
my studies. His willingness to repeatedly assist on all aspects of this study was greatly

appreciated.

Dr. Hui Li made a significant contribution to this study and for this, as well as his willing

assistance on all aspects of the study, I am grateful.

iv

To Dr. Carrie Laboski and Dr. Thomas Voice, I extend many thanks for their assistance
with various aspects of the study, and for facilitating discussions about the project and

what questions needed to be addressed.

I thank the College of Agriculture and Natural Resources and the department of Crop and
Soil Science for their financial support during the final months of this study. In addition, I
thank the faculty and staff of the CSS department with specific thanks to Dr. Kravchenko
who worked tediously with me on the statistics of this project. Thanks also to my
colleagues, both past and present, who have assisted me to some degree to reach my

goals.

To my family, Melvin, Sheila, Marlene and Suzanne Charles, I extend great thanks for
their prayers, support, life-long encouragement, and their faith in my ability to complete
this task. To Rev. Mark and Cheryl McKeel, Dr. Cheron Herrera—Lezama and my dear
friends, I also say thank you. Bountiful thanks to God for His Graces given that I might

accomplish this task.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................ viii
LIST OF FIGURES ........................................................................ xii

CHAPTER 1 EFFECTS OF EXCHANGEABLE CATION ON SORPTION OF
NITROCYANOBENZENE TO SURFACE SOILS

Abstract ............................................................... 1
Introduction ............................................................... 2
Experimental Section ...................................................... 5
Results and Discussion ............................................. 11
Conclusion ............................................................... 33
References ............................................................... 34

CHAPTER 2 AVAILABILITY OF CLAY SURFACES IN SOILS FOR
ABSORPTION OF ORGANIC CONTAMINANTS

Abstract ............................................................... 38
Introduction ............................................................... 40
Experimental Section ...................................................... 45
Results and Discussion ............................................. 51
Conclusion ............................................................... 65
References ............................................................... 66

vi

CHAPTER 3 SORPTION OF P-NITROCYANOBENZENE TO SOILS: THE
RELATIONSHIP BETWEEN SOIL PROPERTIES AND
AVAILABILITY OF CLAY SURFACES

Abstract ............................................................... 71
Introduction ............................................................... 72
Experimental Section ...................................................... 74
Results and Discussion ............................................. 80
Conclusion ............................................................... 96
References ............................................................... 97

CHAPTER 4 SORPTION OF NONIONIC ORGANIC COMPOUNDS TO SOILS:
THE COMPOUND SPECIFICITY OF AN AVAILABILITY INDEX

Abstract ............................................................... 1 02
Introduction ............................................................... l 04
Experimental Section ...................................................... 106
Results and Discussion ............................................. 112
References ............................................................... 1 32
CONCLUSION ........................................................................ 136

vii

LIST OF TABLES

Table 1.1 .................................................................................. 14

Measured SOM, clay content and CEC for Webster A and B (whole soils and soils from
which SOM was removed), peat, Houghton muck (HM) and Brookston loam (BL) soils.
Calculated Qom values derived from sorption isotherm data for trichlorobenzene (TCB),
p-nitrocyanobenzene (p-NCB) and 1,-4 dinitrobenzene (1,4-DNB) by K- and Mg2 -
saturated soils at an equilibrium concentration of 5 mgL'l.

Table 1.2 ................................................................................. l7

Freundlich sorption equation Kf and n values used to calculate Qom values derived from
sorption isotherm data for trichlorobenzene (TC132), p-nitrocyanobenzene (p-NCB) and
1,4—dinitrobenzene (1, 4- -DNB) by K- and Mg2 -saturated soils at an equilibrium
concentration of 5 mgL'l.

Table 1.3 ................................................................................. 27

Comparison of amount of solute sorbed per unit mass of sorbent (Q) at an equilibrium
concentration of 5 mgL for p-nitrocyanobenzene sorption by homoionic Webster A
clay- sized fraction and Webster A unfractionated soil (Q values also normalized to lowest
Q value).

Table 1.4 ................................................................................. 28

Freundlich sorption equation Kf and n values used to calculate the amount of solute
sorbed per unit mass of sorbent (Q) at an equilibrium concentration of 5 mgL for p-
nitrocyanobenzene sorption by homoionic Webster A clay-sized fraction and Webster A
unfractionated soil.

Table 2.1 ................................................................................. 60

Calculated f values for p-nitrocyanobenzene (p-NCB) sorption by Webster A soil using
Karickhoffs equation [1] (using p-NCB sorption by K- and Mg2 -saturated peat,
Houghton muck and Brookston loam to determine the Km, term), average f values for
K- and Mg2 -saturated sorbents and our alternative equation [2].

viii

Table 2.2 ................................................................................. 63

Calculated fa values for diuron sorption by Webster A soil using Karickhoffs equation
[1] (using diuron sorption by K+- and Mg2+-saturated peat to determine the K0m term),
average fa values for K+- and Mg2+-saturated sorbents and our alternative equation [2].

Table 3.1 ................................................................................. 81

Physicochemical characteristics of Webster and Clarion soils (whole soil and soils from
which SOM was removed).

Table 3.2 ................................................................................. 83

Freundlich equation Kr and n values for p-nitrocyanobenzene sorption by K+- and Mg2+-
exchanged Webster and Clarion (A and B horizons) whole soils and soils from which
SOM was removed.

Table 3.3 ................................................................................. 87

The f, values for p-nitrocyanobenzene (NCB) sorption by Webster (A and B horizons)
and Clarion (A and B horizons) estimated using Karickhofi’s equation [1] (using solute
sorption by K+-Saturated peat, Houghton muck, or Brookston loam to determine the Km
term).

Table 3.4 ................................................................................. 88

The fa values for p-nitrocyanobenzene (NCB) sorption by Webster (A and B horizons)
and Clarion (A and B horizons) estimated using Karickhoffs equation [1] (using solute
sorption by Mg2+-saturated peat, Houghton muck, or Brookston loam to determine the

Kom term).

Table 3.5 ................................................................................. 90

Correlation (I value) between fa values for sorption of p-nitrocyanobenzene estimated
using Karickhoff equation [1] (using solute sorption by K+-saturated peat, Houghton
muck or Brookston loam to determine the K0m term) and the alternative equation [2] with
percent organic matter (%OM), percent smectite (%smectite) and ratio of organic matter
to smectite (%OM/%smectite).

ix

Table 3.6 ................................................................................. 92
Linear regression describing the relationship between fa values and %OM/%smectite.

Table 4.1 ................................................................................. 113
Physicochemical characteristics of Webster and Clarion soils (whole soil and soils from
which SOM was removed).

Table 4. 2 ................................................................................. 118
Freundlich equation Kr and n values for p-nitrocyanobenzene,1,4-dinitrobenzene and
2 ,4-dinitrobenzene sorption by K- and Mg2 -exchanged Webster and Clarion (A and B
horizons) whole soils and soils from which SOM was removed.

Table 4.3 .................................................................................. 122
The fa values for p-nitrocyanobenzene (NCB), 1,4-dinitrobenzene (DNB), and 2,4-
dinitrotoluene (DNT) sorption by Webster (A and B horizons) and Clarion (A and B
horizons) estimated using Karickhoffs equation [1] (using solute sorption by K+-
saturated peat to determine the K0m term).

Table 4. 4 ................................................................................. 123
The f values for p-nitrocyanobenzene (NCB),1,4-dinitrobenzene (DNB), and 2, 4-
dinitrotoluene (DNT) sorption by Webster (A and B horizons) and Clarion (A and B
horizons) estimated using Karickhoff’s equation [1] (using solute sorption by Mg2 -
saturated peat to determine the K0'“ term).

Table 4.5 .................................................................................. 124
The fa values for p-nitrocyanobenzene (NCB), 1,4-dinitrobenzene (DNB), and 2,4-
dinitrotoluene (DNT) sorption by Webster (A and B horizons) and Clarion (A and B
horizons) estimated using our alternative equation [2].

Table 4.6 .................................................................................. 126
Correlations between fa values for sorption of p-nitrocyanobenzene (NCB), 1,4-
dinitrobenzene (DNB) and 2,4-dinitrotoluene (DNT) estimated using Karickhoff equation
[1] (using solute sorption by K+-saturated peat to determine the K0'“ term) and our
alternative equation [2] and percent organic matter (%OM), percent smectite (%smectite)
and ratio of organic matter to smectite (%OM/%smectite).

Table 4.7 ................................................................................. 127

Amount of solute sorbed to K+-beidellite (Qbeidellite) and peat (pr) soil at an equilibrium
concentration of 10 mg L'1 and the ratio of extents of p-nitrocyanobenzene (NCB), 1,4-
dinitrobenzene (DNB), and 2,4-dinitrotoluene (DNT) sorption by SOM and K+-beidellite.
Regression correlations between fa values estimated using Karickhoff equation [1] (using
solute sorption by K+-satIIrated peat to determine the K0... term) and our alternative
equation [2] and chdcimc and the ratio of extents of solute sorption by SOM and K+-

beidellite (Kom/Ksmectite)-

xi

LIST OF FIGURES

Figure 1.1 ................................................................................. 13
X-Ray spectra of Webster A (1) and B (2) soils, Mg2+-exchanged, air dried.

Figure 1.2 ................................................................................. 15

Adsorption isotherms for sorption of p-nitrocyanobenzene (NCB) and 1,4-dinitrobenzene
(DNB) by Webster A (KL-saturated) soil fractions.

Figure 1.3 ................................................................................. 16
Adsorption isotherms for sorption of p-nitrocyanobenzene (NCB) and 1,4-dinitrobenzene
(DNB) by Webster B (K+-saturated) soil fractions.

Figure 1.4 ................................................................................. l9

Adsorption isotherms for sorption of 1,2,4-trichlorobenzene (TCB) by (a) K+-saturated
peat, Houghton muck, Brookston loam, Webster A, Webster B, Webster A and B from
which bulk SOM was removed and (b) sorption isotherms normalized to amount of SOM
in soils.

Figure 1.5 ................................................................................. 20

Adsorption isotherms for sorption of 1,2,4-trichlorobenzene (TCB) by (a) Mg2+-sat11rated
peat, Houghton muck, Brookston loam, Webster A, Webster B, Webster A and B from
which bulk SOM was removed and (b) sorption isotherms normalized to amount of SOM
in soils.

Figure 1.6 ................................................................................. 21

Adsorption isotherms for sorption of p-nitrocyanobenzene (p-NCB) by (a) Kt and Mg”-
saturated peat, Houghton muck, Brookston loam and (b) sorption isotherms normalized to
amount of SOM in the soils.

Figure 1.7 ................................................................................. 22
Adsorption isotherms for sorption of 1,4-dinitrobenzene (DNB) by (a) K+- and Mg2+-
saturated peat, Houghton muck, Brookston loam and (b) sorption isotherms normalized to
amount of SOM in the soils.

xii

Figure 1.8 ................................................................................. 24
Adsorption isotherms for sorption of p-nitrocyanobenzene by K+-saturated reference

clays (illite IMt-2; beidellite; montmorillonite SWy-2).

Figure 1.9 ................................................................................. 26
Adsorption isotherms for sorption of p-nitrocyanobenzene (p-NCB) by clay-sized

fraction of Webster A soil exchanged with various cations (Cs+, K+, Na+, Li+, Ba2+, Mg”,

and Ca2+).

Figure 1.10 ................................................................................. 32

Adsorption isotherms for sorption of p-nitrocyanobenzene by Webster A soils exchanged
with various cations (CS+, K+, Na+, Li+, Ba2+, Mg”, and Ca2+).

Figure 2.1 ................................................................................. 52
Adsorption isotherms for sorption of p-nitrocyanobenzene (p-NCB) by (a) KT- and Mg”-
saturated peat, Houghton muck, Brookston loam and (b) sorption isotherms normalized to
amount of SOM in the soils.

Figure 2.2 ................................................................................. 53
Adsorption isotherms for sorption of diuron by K+- and Mg2+-saturated peat, Houghton
muck, Brookston loam.

Figure 2.3 ................................................................................. 55
Adsorption isotherms for p-nitrocyanobenzene (p-NCB) and diuron sorption by K+- and
Mg2+-saturated by Webster A soil (whole soil and soil from which SOM was removed).

Figure 3.1 ................................................................................. 82

X-ray diffraction spectra for Mg2+-saturated clay-sized fraction isolated by Webster and
Clarion (A and B horizons) soils.

Figure 3.2 ................................................................................. 84
Adsorption isotherms for sorption of p-nitrocyanobenzene by K+- and Mg2+-saturated
Webster and Clarion (A and B horizon) soils.

xiii

Figure 3.3 ................................................................................. 93
Linear regressions of f8 values obtained from data analysis of p-nitrocyanobenzene (p-
NCB) sorption data using Karickhoff equation [1] and our alternative equation [2] across
4 specified concentrations (5, 10, 15, 20 mg L'l) versus ratio of SOM to smectite content
(%OM/%smectite) across 4 soils (Webster A, Webster B, Clarion A, Clarion B). Sorption
of p-NCB to K+-peat, Houghton muck, or Brookston loam was used to obtain the Korn
term in Karickhoff equation [1].

Figure 4.1 ................................................................................. 114
Isotherms for sorption of p-nitrocyanobenzene (p-NCB), 2,4-dinitrotoluene (DNT), and
1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Webster A (WA) and Webster A
soil from which SOM was removed (WAO).

Figure 4.2 ................................................................................. 115
Isotherms for sorption of p-nitrocyanobenzene (NCB), 2,4-dinitrotoluene (DNT), and
1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Webster B (WB) and Webster B
soil from which SOM was removed (WBO).

Figure 4.3 ................................................................................. 116
Isotherms for sorption of p-nitrocyanobenzene (NCB), 2,4-dinitrotoluene (DNT), and
1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Clarion A (CLA) and Clarion A
soil from which SOM was removed (CAO).

Figure 4.4 ................................................................................. 117
Isotherms for sorption of p-nitrocyanobenzene (p-NCB), 2,4-dinitrotoluene (DNT), and
1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Clarion B (CLB) and Clarion B
soil from which SOM was removed (CBO).

Figure 4.5 ................................................................................. 130
Isotherms for sorption of p-nitrocyanobenzene (NCB), 2,4-dinitrotoluene (DNT), and
1,4-dinitrobenzene (DNB) by K+-beidellite.

Figure 4.6 ................................................................................. 131
Isotherms for sorption of 1,3,5-trinitrobenzene by K+- and Mg2+-saturated Webster B
(WB) and Webster B from which SOM was removed (WBO).

xiv

CHAPTER 1
EFFECTS OF EXCHANGEABLE CATION ON SORPTION OF

NITROCYANOBENZENE TO SURFACE SOILS

ABSTRACT

The influence of exchangeable cation on the adsorption of p-nitrocyanobenzene (p-NCB)
and 1,4-dinitrobenzene (1,4—DNB) by Webster A and B horizon soils, which contain
smectite as their most abundant clay mineral, was determined. On removal of soil organic
matter (SOM), sorption of p-NCB and 1,4-DNB increased indicating an overall negative
contribution of SOM to their sorption by whole soil. This suggests that SOM suppresses
sorption of p-NCB and 1,4-DNB by soil mineral components. Removal of the clay-sized
fraction from whole soil decreased sorption of these compounds by a factor of at least 10.
We determined, independently, the effect of exchangeable cation on sorption of 1,4-DNB
by SOM, and on sorption of p-NCB by whole soil, the clay-Sized fraction of soil and
SOM. No effect of exchangeable cation type was found for sorption of p-NCB and 1,4-
DNB by SOM. For homoionic clay-sized soil fractions, the extent of p-NCB sorption was
generally greater for monovalent-cation-exchanged clays than divalent-cation-exchanged
clays, and increased with decreasing cation hydration energy (CS+>K+>Na+~Li+ and
Ba2+>Mg2+~Ca2+). The same trends were observed for homoionic soils exchanged with

monovalent and divalent cations.

INTRODUCTION

Nitroaromatic compounds (NACS) are commonly used as pesticides, explosives, and
intermediates in the synthesis of dyes, ammunition and solvents. Their uses and
manufacture have resulted in contamination of soils, sediments, quarries and aquifers. In
order to assess the fate of NACS in the environment, their mobility and reactivity in the
soil must be understood. Sorption processes play an important role in determining
mobility, translocation, and (bio)availability of these solutes in soil. Clay minerals and
organic matter are generally considered the two most active soil components in the
sorption of aqueous phase organic contaminants. For many nonpolar organic compounds,
sorption is highly correlated with the organic carbon (OC) content of a soil or sediment
and is relatively independent of other sorbent properties (Chiou et al., 1979, 1983;
Karickhoff et al., 1979). This has been observed as linear relationships between the
organic carbon normalized sorption coefficient (K00) and the octanol-water partition
coefficient (KW) (a direct relationship) and water solubility (8..) (an inverse relationship)
of nonpolar organic compounds (Chiou, 2002). In these instances sorption has been
conceptualized as solute partitioning into SOM (Chiou et al., 1979, 1983; Karickhoff,
1979). It has been further suggested that adsorption of such solutes by mineral phases in
soils is suppressed by the preferential adsorption of water (Chiou, 2002). However, many
pesticides and organic contaminants, e.g. NACS, contain polar functional groups that can
engage in specific interactions with clays even in the presence of bulk water. In fact,
compared on a unit mass basis, adsorption of NACS from water by smectite clays was

greater than that by SOM (Sheng et al., 2002). Thus, clay minerals may play a dominant

role in sorption of certain polar organic molecules, particularly in geosorbents with low

SOM contents (Weissmahr etal., 1999).

Several studies have reported the high affinity of NACS for smectites in aqueous solution
(Haderlein et al., 1996; Weissmahr et al., 1998; Boyd et al., 2001; Sheng et al., 2002;
Johnston et al., 2001; Johnston et al., 2002). This has been attributed to the formation of
electron donor (clay) — acceptor (NAC) complexes facilitated by the electron
withdrawing properties of the nitro (N02) group (Weissmahr et al., 1997). Alternatively,
complexation of the N02 groups of NACS with exchangeable cations on the clay has been
suggested as a primary sorption mechanism (Boyd et al., 2001). Partial solute
dehydration associated with adsorption in smectite interlayers of optimal distance (ca.

12.3 A) also provides favorable energy for adsorption (Boyd et al., 2001; Li et al., 2004).

Several studies document the importance of the nature of the exchangeable cation on
sorption of NACS by clays. Sorption of NACS by reference smectite clays was favored by
weakly hydrated exchangeable cations and a low surface charge density of the clay
(Haderlein et al., 1993; Weissmahr et al., 1997; Boyd et al., 2001; Sheng et al., 2002).
These factors combine to manifest optimal interlayer distances, and promote NAC
interactions with exchangeable cations (Boyd et al., 2001). Significant adsorption of
NACS by smectite clays was observed in the presence of weakly hydrated cations (e.g.
K+, Cs+, NHX), while strongly hydrated cation-systems (e.g. Na+, Mg”, Ca”) exhibited
low adsorption of NACS (Haderlein et al., 1993). These results agree with the lower

sorption observed for 2,4-dinitro-o-cresol (DNOC) sorption by smectite clays saturated

with more strongly hydrated cations (Sheng et al., 2002). Johnston et al. (2002) reported
Shifts in the N02 vibrational modes of 1,3,5-trinitrobenzene (1,3,5-TNB) on its
adsorption by K+-Smectite using Fourier Transformed Infrared Spectroscopy (FTIR).
They attributed these shifts in N02 vibrational modes (vasym and vsym (~NO) bands) to
complexation of K+ by N02 groups. These Shifts were not observed for smectites
saturated with more strongly hydrated cations (i.e. Na+, Mg”, Ca2+ and Ba”). Overall,
cation hydration seems to be a central determinant of NAC sorption by reference smectite
clays. Sorption of NACS occurs primarily in the interlayers of these clays, and cation
hydration determines the interlayer distance, the effective size of adsorption domains
between cations, and the ability of N02 groups to complex directly with exchangeable

cations (Boyd et al., 2001; Johnston et al., 2001, 2002; Sheng et al., 2002; Li et al., 2004).

Although the effect of cation saturation on NAC sorption by homoionic clays is well
documented, little is known about the effect of exchangeable cation type on sorption by
soils with relatively high SOM contents, or on the relative contributions of soil organic
and mineral components on NAC sorption and how this might be affected by the nature
of the exchangeable cation. The objective of this study was to document the effect of
exchangeable cation type on NAC sorption by whole soil, SOM and the soil clay-mineral

fraction.

EXPERIMENTAL SECTION

A clay loam Webster soil (fine loamy, mixed, mesic Typic Haplaquoll) (Soil Survey
Staff, NRCS) (A and B horizons) was fractionated by standard procedures (Kunze &
Dixon, 1986). Carbonates (CO3), SOM, iron oxides (Fer), and clay-sized particles were
sequentially removed from each soil horizon. Carbonates were removed using 0.5N
sodium acetate acidified to pH 5 at 80°C, followed by the removal of SOM using 30%
H202 at 80°C, followed by the removal of free Fer using the citrate-bicarbonate-
dithionite method. The clay-sized fraction (<2.0 um) was obtained by wet sedimentation
of the CO3-OM-Fe-removed soil fraction (Whittig & Allardice, 1987). The fraction
remaining after the removal of CO3, SOM, Fer and clay-sized particles was termed the
residual soil fraction. The masses of CO3 and Fer removed, and the mass of residual
fraction remaining were determined gravimetrically. Additional soils used were a
Pahokee peat, Houghton muck and Brookston loam. The SOM content of whole soils and
the amount of SOM remaining after extraction of bulk SOM using H202 were determined
by the Michigan State University (MSU) Soil Testing Laboratory using dry combustion
and a Leco carbon analyzer (St. Joseph, Michigan) (Nelson and Sommers, 1996). The
content of clay-sized particles in the Webster soils was determined by the MSU Soil

Testing Laboratory using the hydrometer method (Gee and Bauder, 1986).

To prepare homoionic sorbents, soil and soil fractions (ca. 100 g) were washed with 200
ml of a 0.1M solution of the cation chloride salt overnight, thereafter centrifuged and the
supernatant discarded; this procedure was repeated five times. The soils were then

washed with approximately 200 ml of Millipore Milli-Q (Billerica, Massachusetts)

deionized H2O, thereafter centrifuged and the supernatant tested with AgNO3; the
procedure was repeated until a negative Cl' ion test with AgNO3 was obtained. The soils

were then freeze-dried and stored at room temperature (23 :l: 1°C) until used.

Para-nitrocyanobenzene (p-NCB), 1,4-dinitrobenzene (1,4-DNB) and 1,2,4-
trichlorobenzene (1,2,4-TCB) were purchased from Aldrich Chemical Company
(Milwaukee, Wisconsin) with purities of 97%, 98%, and 99% respectively, and used as
received. For p-NCB, 1,4-DNB, and 1,2,4-TCB, the water solubilities (SW) at 25°C are
1650, 69 and 49 mg L", respectively; the log Kow values are 1.19, 1.46, and 4.02,

respectively (Howard and Meylan, 1997).

Sorption of solutes by K+-saturated soil and soil fractions were measured in triplicate
using a batch equilibrium technique. This involved adding aqueous solutions of p-NCB,
1,4-DNB, or 1,2,4-TCB over a range of initial concentrations (Ci, 1-42 mg L’1 for p—NCB
and 1,4-DNB; 1-30 mg L'1 for 1,2,4-TCB) to a known mass of soil. Initial solutions of the
compounds were prepared in 0.05M solution of the cation chloride salt. Total initial
volumes of 5 ml (p-NCB, 1,4-DNB) or 7 ml (TCB) were pipetted into 7.5 ml borosilicate
glass vials containing various amount of sorbent (0.1-0.4g for p-NCB and 1,4-DNB;
0.04-0.35g for 1,2,4-TCB). The vials were rotated continuously ovemight (24 hours) in
the dark at room temperature (23 i 1°C); preliminary studies showed equilibrium was
reached within 24 hours (data not provided). The liquid and solid phases were separated
by centrifugation at 4068g for 40 min, and the solute’s equilibrium aqueous concentration

(Ce) measured using a Perkin-Elmer high performance liquid chromatograph (HPLC)

(Perkin-Elmer, Norwalk, Connecticut). This instrument consisted of a binary liquid
chromatography pump 250 with a series 200 autosampler and a uv-visible detector set at
a suitable adsorption wavelength for the respective compound (1. 254, 262, and 228nm
for p-NCB, 1,4-DNB, and 1,2,4-TCB, respectively). An Alltech platinum EPS C18
column (Deerfield, Illinois) was used for p-NCB and 1,4-DNB. A Varian microsorb-MV
C18 column (Palo Alto, California) was used for 1,2,4-TCB. The mobile phase was an
isocratic mixture of methanol (HPLC grade) and Milli-Q deionized water (methanol:
water ratios of 55:45 for p-NCB, 65:35 for 1,4-DNB, and 92:8 for 1,2,4-TCB) with a
flow rate of 1.0 m1 min'l. The amount of solute sorbed by each sorbent was calculated as
the difference between C; and Cc. Compound recoveries in control vials without sorbent

were above 95%. Because of the high solute recoveries, the data were not adjusted.

The effect of exchangeable cation type on sorption by SOM was evaluated using either
organic soils relatively devoid of mineral matter or a mineral soil devoid of swelling
clays (i.e. smectites) implicated in the sorption of NACS (Boyd et al., 2001; Haderlein et
al., 1993). The sorbents used were the Pahokee peat (euic, hyperthermic Lithic
Medisaprist), Brookston loam (fine-loamy, mixed, superactive, mesic Typic Argiaquoll)
and Houghton muck (euic, mesic Typic Haplosaprist) (Soil Survey Staff, NRCS). The
peat was purchased from the International Humic Substances Society (BS103P) (Denver,
Colorado). The Brookston loam was obtained from Ionia, Michigan, and the Houghton
muck from Bath, Michigan. Sorption isotherms for p-NCB and 1,4-DNB, and 1,2,4-TCB

by these soils were obtained using the batch equilibrium technique described above.

These soils were saturated with K+ or Mg2+ to evaluate the effect of cation type on

sorption of NOCs.

The unfractionated Webster soil (A horizon) and its isolated clay-sized fraction were
saturated with different exchangeable cations (CS+, K+, Na+, Li+, Mg”, Ca2+) to
determine the effect of exchangeable cation on sorption of p-NCB. Batch equilibrium
experiments were conducted using the M“+-saturated soil clay fractions and whole soils
with p-NCB as the sorbate. Sorption of p-NCB by K+-saturated reference clays which are
representative of the identified clay types in the soil clay fraction (see below) were also
measured using batch equilibrium experiments as described above. The clays used were:
Wyoming montmorillonite (SWy-2), illite (IMt-2) and kaolinite (KGa-2) obtained fi'om
the Clay Minerals Society Source Clay Repository (Columbia, MO) and beidellite from

Ward’s Natural Science (Rochester, NY).

X-ray diffraction (XRD) was used to identify the clay mineralogy of the clay-sized
fractions (Whittig and Allardice, 1987). Oriented samples of Mg2+- and K+-saturated soil
clay-Sized fractions were prepared by dropping suspensions of the clay-sized materials on
glass Slides, then air-drying, heating at 500°C, and treatment with ethylene glycol.
Samples were analyzed using CuKa radiation and a Philips APD 3720 X-ray

diffractometer (Philips Electronic Instruments, NJ).

Semi-quantitative estimates for the amount of smectite in the unfractionated Webster

soils was attempted using the measured cation exchange capacity (CEC) of the

unfractionated soils ((Michigan State University (MSU) Soil Testing Laboratory using
the ammonium acetate method (Sumner and Miller, 1996)), and an average CEC of a soil
smectite and SOM of 80 and 160 cmol kg", respectively, using the equation:- CEC =
(160 cmol kg")(fom) + (80 cmol kg")(fcm) (1)

where for“ and fcm are the fractional SOM and smectite clay contents of the soil. The
calculation assumed that the CEC of the soil originated solely from SOM and soil
smectites. By comparing the X-ray diffraction peak areas associated with the non-
smectite minerals to the smectite peak, we estimated the relative fractional amounts of
each non-smectite non-quartz mineral present in the clay-sized fraction as a fraction of
the smectite peak (illite/mica, 0.125; kaolinite, 0.125). Multiplying the percent smectite in
the soil obtained using equation (1) (e.g., 24%) by the relative fractional amount of each
non-smectite mineral determined from the relative peak areas in the X-ray diffraction
graph, we calculated the amount of non-smectite minerals (e.g. illite/mica, 3%; kaolinite,
3%). We subtracted the estimated smectite content (e.g. 24%) from the total clay content
(e.g. 31%) to obtain the amount of non-smectite minerals (e.g. 7%) and compared this

value to the sum of the non-smectite minerals estimated using the peak areas (3% + 3%).

Statistical Analysis

The sorption coefficient (K2) and the constant related to the sorption isotherm curvature
(n) were estimated using the Freundlich equation, Q=K¢C.", where Q is the sorbed solute
concentration per unit mass of the sorbent (mg g'l), Kr is the sorption coefficient, n is a
constant related to the isotherm curvature (concave or convex). Comparisons of sorption

isotherms were made by comparing the K2 and n values for each soil-solute system using

nonlinear regression in SAS (SAS Institute, 1999). A significant difference between
isotherms was assigned when there was a significant difference (or = 0.05) between Kf

and n or Kf values of compared isotherms (Laboski and Lamb, 2004).

10

RESULTS AND DISCUSSION

The Webster soils (fine loamy, mixed, mesic Typic Haplaquoll) contain smectites (Figure
1.1). Nitroaromatics are known to be sorbed significantly by smectite clays (Haderlein et
al., 1996; Weissmahr et al., 1998; Boyd et al., 2001; Sheng et al., 2002; Johnston et al.,
2001; Johnston et al., 2002). The smectite contents of the Webster A and B soils,
calculated using the CECS of the soils, are 24 and 16% smectite, respectively (Table 1.1)

and were therefore chosen as the study soils.

Removal of SOM from the Webster A and B horizons increased sorption of p-NCB and
1,4-DNB, with K+ as the saturating exchangeable cation (Figure 1.2, 1.3). In whole soil it
appeared that the overall effect of SOM was to suppress sorption of NCB and 1,4-DNB,
though SOM itself is undoubtedly a sorptive sink for these compounds. Karickhoff
(1984) proposed that, if SOM blocked clay sorption sites, then SOM removal could result
in increased sorption by clays, although this effect had not been demonstrated
experimentally until now. Removal of the clay-sized fraction decreased sorption by a
factor of at least 10 in all soils for both p-NCB and 1,4-DNB compared to unfractionated
soil, with K+ as the saturating cation (Figure 1.2, 1.3). Given the very low sorption of
NACS by free Fer (data not reported) and the established high affinity of smectites for
NACS (Haderlein and Schwarzenbach, 1993; Haderlein et al., 1996; Weissmahr et al.,
1997; Haderlein et al., 1993 Boyd et al., 2001; Johnston et al., 2002; Johnston et al.,
2001; Sheng et al., 2002; Li et al., 2003), our data supports the premise that SOM and soil
clays are the two components that significantly influence NAC sorption. The contribution

by soil clays to sorption of p-NCB and 1,4-DNB was greater than the contribution by

11

SOM. The contribution of CO3 and the residual fraction to sorption of the solutes was
minimal for all soils (Figure 1.2, 1.3). We found low levels of CO3 present in the
unfractionated soils (data not shown) consistent with the findings of Laird et a1. (1994)

for a Webster soil.

The influence of exchangeable cation on sorption of organic solutes by soil will depend
on its influence on the dominant soil components responsible for solute sorption, i.e.
SOM and clay. Trichlorobenzene (TCB) is a nonpolar organic compound whose sorption
by soil results predominated (if not exclusively) from partitioning into SOM (Chiou,

2002)

The amount of TCB sorbed (Qom) by K+-saturated Houghton muck, Brookston loam and
Webster soils at an equilibrium concentration of 5 mg L'1 (Table 1.1) was calculated
using the Kf and n parameters from the sorption data normalized to the amount of SOM
in the soils fitted to the F reundlich equation (Q = Kr * Ce"; Qom = Q / fom, where Q is the
sorbed solute concentration per unit mass of the sorbent (mg g'l), K2 is the sorption
coefficient, 11 is a constant related to the isotherm curvature (concave or convex), fom is
the fractional SOM content of the soil). The values ranged from 2.15 to 3.60 mg g"; less
than a factor of 2 (Table 1.2). The Qom values for TCB sorption by Mg2+-saturated
Houghton muck, Brookston loam, peat and Webster soils at an equilibrium concentration
of 5 mg L'1 ranged from 1.15 to 3.30 (Table 1.2); less than a factor of 3. The sorption
isotherms for TCB sorption by K+- and Mg2+- saturated organic and mineral soils

normalized to the

12

 

 

 

 

 

q
E
M
c -I
*3
f "r . ”I” 1 1
1 I ' N
2 MWw-J
0 5 10 15 20 25 30

Angle 2 theta

Figure 1.1. X-ray spectra of Webster A (1) and B (2) soils, Mg2+-exchanged, air dried.

(s, smectite; m/i, mica and/or illite; k, kaolinite; q, quartz)

13

20 82: $58 £8088 >20 .3 33888 mm comp—om own—moon 038:8? «o: - 8
3882-20 m newton guano? l om? m newton 83.0.33 I m?
3882-20 < 805.8: 66903 I O<3 < 88.8: 88303 I <3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a: 8 a: a. 8..” 3m - S R to .52 as on?
a. s. s. a: 3m 2.... - em a no .52 as 9;»
a: a: a. 8 m: 2a 8 E R 8 .52 .36 m?
E a: a: s. 43 2a 2 em 3. mm :32 a6 <3
EMO—
Sd e2 98 $6 on... 8a 2 - em ed as has .8
:3 one «2 one new Nam as - - hm... gamma 2:
3o o2 $5 2: 2a 3a - - - was 8&8 .8.—
..»2 . .x I»: .x a»: .2 a.
8.8-.. _ 82... 89 an:
8:853. 32.x. 55.x. 23:; =8
OZOm .WSES a CEO
80°
.8388

mo 8088 20m 8283.“ \ AOV 835m .«o mme :83 “on C098 828 we 85088 n :50 .74 w8 m .«o 82888888 885283 8w

8 £8 vou888-+~w2 98 QM .3 80:88 5754.. 5 0808308868; 98 502.3 unenconoamxooEcd .50: 058380205

com 8% 8350mm noun—cm 80¢ 33.8w $28, :50 vogue—mo .mmom va Emo— 89385 98 AEIV x38 counwnom

.83 55382 33 20m 533 89¢ £8 28 £8 BonBV m 98 < Bane? 8m 85 ES 8088 .3? .20m 38802 .2 2an

l4

 

0.5
WA

~CO3

-CO -OM

residual a

<40.

0.4 ~

 

Concentration of NCB sorbed (mg g")

 

 

 

 

0.0 f I

0 10 20 30
Aqueous NCB concentration (mg L")

 

 

 

 

1.0
- 0 WA
‘7 0 -CO3
"‘ ' -co on a
E 0.8 . v reeldual ‘
.. v b
b
E ,.e ‘
8 0.6 « 2 2
m ’5’
z
a
h
g 0.4 .
.° C "
E «.0
s 2 "
8 0. ‘ I", 7 v WC
0 I” ..
U a . "
0.0 V ' I I I I I
0 5 10 15 20 25 30

Aqueous DNB concentration (mg L")

Figure 1.2. Adsorption isotherms for sorption of p-nitrocyanobenzene (NCB) and 1,4-

dinitrobenzene (DNB) by Webster A (K+-saturated) soil fractions.

(carbonates removed, -C03; C03 and SOM removed, -CO3-OM; residual fraction,
fraction remaining after removal of CO3, OM, free F er and clay-sized fraction, residual)

(isotherms with different letters are statistically different at p < 0.05)

15

 

0.8
WB NCB

-CO3 a
-CO3-OM
residual

QQOO

   

0.6 i

  
  
  
  
  
   

0.4 .

0.2 ‘

 

Concentration of NCB sorbed (mg g")

 

 

0.0 :7
0 10 20 30
Aqueous NCB concentration (mg L")

 

.3
0)

WB

603
£03-OM
reeldual

.1)

9

_l

I.
<I<oe

.5
N
4

A
o
I

.0
m
1

  

.o
A

1

 

Concentration of DNB sorbed (mg
_0 o
N on

 

 

I r

0 Aqsueous 0N1: concenti‘ftlon (mgzlg') 25
Figure 1.3. Adsorption isotherms for sorption of p-nitrocyanobenzene (NCB) and 1,4-
dinitrobenzene (DNB) by Webster B (TC-saturated) soil fractions.
(carbonates removed, -CO3; CO3 and SOM removed, -C03-OM; residual fraction,
fraction remaining after removal of CO3, OM, free Fer and clay-sized fraction, residual)

(isotherms with different letters are statistically different at p < 0.05)

16

20 85 858 £8688 .83 .3 38888 mm 8888 0888 28238 8: - 8
m 888: 8803 I m?
< 888: 83363 I <3

3388-20 m 888 88303 I 0m?
3382-20 < 888: 88303 I 0.35

 

 

17

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 8 8 8 8 8 8 8 N9— wmd 3.— 9.6 On?

8 8 8 8 8 8 8 8 $6 cod mm; wmd 03$

8 8 8 8 8 8 8 8 no; 906 m _ A mmd m3

8 8 8 8 8 8 8 8 36 $6 mad mwd <3
96 56 3.6 mod $6 36 mmd Rd mod 2 A wad 3d a:—
Cmd end mvd wad mod 3 .o wmd Rd mud mad mwd 3.0 2:
mod mmd omd omd omd _m.o nmd wmd Nod m fl ._ mud ac; 80.—

Z J— : cv— Z .v— a .v— Z cv— : J—

8 .6. as: L. a8 .0. =5
88..— moz-.. 8...
8888

Lo 8688 20m _8ouom¢ \ Ag 888m mo 88 :8 Con 338m 828 Mo H8588 n :50 .74 w8 m we 828.8888 8:883 8w

8 28m uafihdmm-Awwz 8w fix .3 8:88 5758.3 888808888.— 88 502-3 80830830888 AmOC 808380203

8m 88 8.8508 828.88 80¢ 33.88 328 :50 885:8 8 B8 828 a 88 CM 8383 8888 now—83$ .NA 2an

fractional SOM content of each soil overlapped (Figures 1.4, 1.5). The similarities of Qom
values for TCB sorption by K+-saturated organic and mineral soils indicates that SOM
was the dominant, if not singular, sorbent phase for TCB, and that SOM functioned very

similarly in this regard among the soils tested.

When p-NCB and 1,4-DNB were used as the solutes, the QOlm values at an equilibrium
concentration of 5 mg L'1 (Table 1.1) calculated using the Freundlich Kf and n values
(Table 1.2) for their sorption by ICC-saturated peat, Houghton muck and Brookston loam
fell within a relatively narrow range of 0.56 to 1.07 and 0.56 to 1.36mg g", respectively;
both being a range of less than 3. For p-NCB and 1,4-DNB sorption by Mg2+-saturated
peat, Houghton muck and Brookston loam the range in Q0m at 5mg L'1 was 0.38 to 0.55
and 0.47 to 0.62mg g'1 respectively (Table 1.1); both being a range of less than 2. The
SOM-normalized sorption isotherms for NAC sorption by K+- and Mg2+- saturated soils
almost overlapped (p-NCB) and overlapped (1,4-DNB) (Figure 1.6, 1.7). The Brookston
loam soil is devoid of smectite clays implicated in the mineral phase sorption of NACs,
so for all three sorbents, SOM is likely the dominant sorption phase. A study of
dichlorobenzene and carbon tetrachloride by a much larger group of soils revealed
variations in Kom (where Qom = Kom * Cc) of a factor of ~3 among soils (Kile et al., 1995),
similar to those we observed among organic and mineral soils for TCB, p-NCB, and 1,4-
DNB sorption. Hence an organic soil (devoid of sorptive clays) can be reasonably used to
evaluate sorption of NOCs by SOM, and importantly to evaluate the independent
contribution of SOM to sorption of polar compounds by mineral soils where sorption by

both SOM and mineral components is expected. In addition, the overlap between the Qom

18

 

Poet

Bmdumnknm
lflmmmmwmmx
“humuA

Webster 8

Webster A OM-removed
Webster 8 CIA-removed

00-0400

   
   

Concentration of TCB sorbed (mg g")

 

 

Aqueous TCB concentration (mg L")

 

20
0 Peat

18 7 O Houghton muck u
v Brookston loam

16 ' v WebsterA D

14 . I Webster 8 D
U Webster A OM-removed

12 - o Webster 8 OM-removed

  

 

Amount of TCB sorbed per gram SOM (mg g")
23

 

 

0 2 4 6 8 10 12 14 16 18
Aqueous TCB concentratlon (mg L")

Figure 1.4. Adsorption isotherms for sorption of 1,2,4-trichlorobenzene (TCB) by (a) K'-
saturated peat, Houghton muck, Brookston loam, Webster A, Webster B, Webster A and
B from which bulk SOM was removed and (b) sorption isotherms normalized to amount

of SOM in soils.

l9

 

1 .2
Peat

Brookston loam
Houghton muck
Webster A

Webster 8

Webster A OM-removed
Webster 8 OM-removed

1.0 a

0.8 ~

90-440.

    
 

0.6 r

0.4 ‘

 

0.2

Concentration of TCB sorbed (mg g")

 

0.0
O 2 4 6 8

Aqueous TCB concentration (mg L")

 

 

 

 

v?" 5

g e Peat

g o Brookston loam

; 4 1 v Houghton muck a

8 v WebsterA

E I Webster B

2 o Webster A OM-removed . D
3’ 3 ‘ e Webster 8 OM-remo - 9

a .

2 '4 V ’ ‘

0 I

I

8 ob . v

I.- I
g / I ' . O

a .

5 o ’ . . .

O 2 4 6

Aqueous TCB concentration (mg L")

Figure 1.5. Adsorption isotherms for sorption of 1,2,4-tn'chlorobenzene (TCB) by (a)
Mg2+-saturated peat, Houghton muck, Brookston loam, Webster A, Webster B, Webster
A and B from which bulk SOM was removed and (b) sorption isotherms normalized to

amount of SOM in soils.

20

 

 

 

 

 

2.0
I Peat (K)
v; 0 Houghton muck (K)
a v Brookston loam (K)
E, 1 5 _ v Peat (Mg)
' I Houghton muck (Mg)
8 D Brookston loam . 9
g ,
m
‘2’ 1.0 - 3
to. C
c ‘c
3 -.
§ 0.5 ~ - Q) .
g j. 0 .” "V‘I‘,
8 "J I __ . _ =.
0.0 ““qu- - mi r r
O 10 20 3O 4O
Aqueous NCB concentration (mg L")
3‘0 e Peat (K)
O Houghton muck (K)
2.5 . v Brookston loam (K)
v Peat (Mg)
I Houghton muck (M9)
2.0 - D Brookston loam (Mg)

 

 

 

Amount of NCB sorbed per gram SOM (mg g")

 

40

Aqueous NCB concentration (mg L")

Figure 1.6. Adsorption isotherms for sorption of p-nitrocyanobenzene (p-NCB) by (a)
K+- and Mg2+-saturated peat, Houghton muck, Brookston loam and (b) sorption isotherms

normalized to amount of SOM in the soils.

21

 

2'0 Peat (K)

Houghton muck (K)
Brookston loam (K)
Peat (Mg)

Houghton muck (Mg)
Brookston loam (Mg)

1.5 4

UIQ‘O.

    
  
 

 

Concentration of one sorbed (mg a")
3 3

 

 

 

.0
o

o 5 10 15 20 25 so
Aqueous DNB concentration (mg L")

 

e Peat (K)

o Houghton muck (K)
2.5 4 v Brookston loam (K)
v Peat (Mg)

I Houghton muck (Mg)
2.0 « D Brookston loam (Mg)

d
o
1

 

 

Amount of DNB sorbed per gram SOM (mg g")
at

 

 

0 5 10 1 5 20 25 30
Aqueous DNB concentration (mg L")

Figure 1.7. Adsorption isotherms for sorption of 1,4-dinitrobenzene (DNB) by (a) K‘-
and Mg2+-saturated peat, Houghton muck, Brookston loam and (b) sorption isotherms

normalized to amount of SOM in the soils.

22

values for sorption of TCB, p-NCB and 1,4-DNB by the K+- and Mg2+-saturated soils as
well as the narrow range in Qom values (i.e. less than a factor of 3) indicates that the IC-
and Mg2+-saturated systems function similarly for the sorption of NOCs. The residual
SOM (i.e., SOM remaining after H202 extraction of bulk SOM) sorbed TCB similarly to
bulk SOM in the mineral and organic soils. This is evidenced by the similarity in the
calculated Q0m values at an equilibrium concentration of 5 mg L'1 (Table 1.1) for TCB
sorption by K+- and Mg2+-saturated peat, Houghton muck, Brookston loam, Webster soils
and SOM-removed soils. Our results therefore provide evidence that, in systems where
sorption is dominated by SOM, sorption is not strongly influenced by the saturating
cation. Among the group of solutes and soils characterized in this study it appears that
differences in sorption induced by different cations saturating the CEC would result from

their effect on the sorptive characteristics of the clay-sized mineral fraction and not the

SOM.

XRD analysis indicated that the minerals present in the clay fractions of the Webster A
and B horizons included kaolinite (ca. 2-3%), quartz (ca. 1-11%), smectite (ca.16-24%)
and mica and/or illite (ca. 2-3%) (Figure 1.1). For the Webster A sorption experiments
for p-NCB sorption by K+-saturated kaolinite, illite, montmorillonite (SWy-2) and
beidellite showed that the smectite clays (i.e., SWy-2 and beidellite) were by far the most
effective sorbents for p-NCB, with the montmorillonite (SWy-2) at least 2x more
effective than beidellite (Figure 1.8). These results agree with previous studies that report
the affinity and the adsorption capacity of clays for NACs increased in the order kaolinite

< illite << montmorillonite (Haderlein et al., 1996). Thus, among the soil minerals

23

 

  
  
  
 
 

 

 

 

2.5
O K-Illite
7" O K-Beidellite
: v K-Kaolinite
E 2.0 a v KSWy—2
'U
0
.D
'6
n 1.5 4
m
U
2
"6
c 1.0 - b
.2
‘6
b
5 o 5 -
o .
C
8
C
d
o 0 V ' ‘F=='=I=Fl=:=='-—
0 5 1O 15 20 25

Aqueous NCB concentration (mg L")

Figure 1.8. Adsorption isotherms for sorption of p-nitrocyanobenzene by K+-saturated

reference clays (illite [Mt-2; beidellite; montmorillonite SWy-2).

(isotherms with different letters are statistically different at p < 0.05)

24

present in the soils studied, smectites appear to be the dominant sorptive phase. Previous
studies have demonstrated that, for pure smectite samples, NAC sorption is affected
strongly by the type of exchangeable cation (Li et al., 2003; Boyd et al., 2001; Sheng et
al., 2001, 2002; Johnston et al., 2001, 2002; Weissmahr et al., 1997; Haderlein et al.,

1996)

Isotherms for NAC sorption by soil clay-sized fraction were nonlinear and generally
exhibited saturation-type curvature approaching a maximum surface concentration
(Figure 1.9). The concentration of p-NCB at an equilibrium concentration of 5 mg L'1
sorbed by homoionic clay—sized fractions of Webster A (Table 1.3) was calculated by
fitting the sorption isotherm data to the Freundlich equation (Kf and n values in Table
1.4). Sorption of p-NCB increased in accordance with decreasing hydration energy, i.e.,
Li+~Na+<<K+<Cs+ and Ca2+~Mg2+<Ba2+ (Table 1.3; Figure 1.9). Overall, Cs+ and K’-
exchanged soil clays were much more effective sorbents than Na+, Li+, Ca2+, Mg2+, or
Bay-exchanged soil clays. These results are consistent with results reported using
reference clays showing higher NAC sorption by smectites saturated with more weakly
hydrated cations. Haderlein et a1. (1993; 1996) found the extent of sorption of NACS was
greater for K+ and NHf-clays compared to Al”, Ca2+, Mg2+, and Na+-clays (Arizona
montmorillonite and kaolinite). Similar results were obtained by Weissmahr et al. (1997,
1999) for NAC sorption to various reference minerals (illite, kaolinite and SAz-l). Sheng
et a1 (2002) measured 4,6-dinitro-o-cresol (DNOC) sorption to SWy-2 smectites and
observed that K+ and Cs+-exchanged SWy-2 clays sorbed DNOC more effectively than

Al”, Ba”, Ca2+, or Na+-exchanged SWy-2 reference clays.

25

 

 

1.6 - ' Ki
0 Na‘“

1.4 - v Mg2+
v Ca2+

12" I L?
C]

1.0 ~ . Cs+

   
 
   

.0
O)
l

 

 

 

.0
.A
4

 

 

Concentration of NCB sorbed, Q (mg g")
0 O
'N on

 

 

p
o

 

VI

0 5 10 15 20 25 30
Aqueous NCB concentration, C(. (mg L")

Figure 1.9. Adsorption isotherms for sorption of p-nitrocyanobenzene (p-NCB) by clay-
sized fraction of Webster A soil exchanged with various cations (Cs+, K+, Na+, Li+, Bay,
Mg“, and Ca2+). Insert is a magnification of adsorption isotherms for sorption of p-NCB

by clay-sized fraction of Webster A soil exchanged with Na+, Li+, Ba”, Mg2+, and Ca2+.

26

73 $2080 2: mo Omo awash“ as ._-w NE own mo 85 08.35 050098 owes; 5 £538 .«o :92 336»: :o Baum ”
.832 .& .Eom use 3:3 E3 332% :2 .32 .3032 .> a

 

 

 

 

 

 

 

 

 

: 3.8 a; 23 em 22- «U
3 >2: 3 2.2 9. $2. a:
E 3.2 ad ems E meT «m
w; mm. K 3” we. _ m mm. :3. 5
:4 502 A: $8 8 mum- «Z
a.» 3.3 H ”.3 Sam? 3 vow- v—
92 2.8m Sq: wadow mm mmm- any
.x.
”.8393.
.5 «32:: 212.. a:
O .333 Avwv— way 0 .833 9.»! 95 3.3 3:23. 3:96.»:
3 uni—«5.8: O O 3 team—«8.8: O O «Econ—m no Rho—.m— ooam 55.5
:8 33559223 .8325 637.235

 

 

.038, 0 $032 8 Bum—ago:

8E 0.8 was?» 0 d8 wowmcomaofimg < 63903 98 5323 BE??? < 8533 £56an 3 £598 oaowaonoagoobfi

d do.“ _.qum we 5:828:00 83.9533 5 3 Ag Sachem («o 3.2: :5. 5Q 398m 838 Mo 280:3 mo manta—=00 .m._ 2an

27

 

 

 

28

 

 

 

 

 

 

38d :85 53 :85 5
285 $86 m and 083 u:
326 flood :23 ”Sod 3
33¢ Rood 523 $56 5
C32 $85 E? 2.86 «z
3%.: 385 Save $85 v.
$23 $85 ammo ”$3 6

z i z _ i

:8 33:33.33: 553.: coma->20 .855

$8 33:03.8? <

.8390? was cough,“ Edwina—o < 6690? 083080: ‘3 H8398 ocoNconowaoobEé com _.A we m we :ouwbcooaoo “nuisance an

3 2% 2888 we mmmE E5 8a won—8m 838 «o 230:8 05 03328 8 com: 83? c 98 .M .8333 5:98 com—950$ 42 2an

Sorption of NACS by smecties occurs largely on the clay interlayers, and cation hydration
determines the interlayer distance, the size of adsorption domains (i.e. regions of siloxane
surface unobscured by water), and the ability of N02 groups to complex directly with
exchangeable cations (Boyd et al., 2001). Considering the hydrated radii of Cs+, K+, Na+,
Li‘“, Ba”, Ca2+, and Mg2+ (3.6, 5.3, 7.9, 10.0, 8.8, 9.6 and 10.8 A respectively)
(Evangelou, 1998), it is apparent that lower cation hydration manifests larger adsorption
domains between cations. Large hydration spheres and strongly held waters of hydration
would act to inhibit the direct interaction of N02 groups with the interlayer cation, as
proposed by Boyd et al. (2001), Johnston et al. (2001, 2002) and Li et al. (2004). The
hydration sphere water molecules are less strongly held (lower enthalpy of hydration) by
ions with larger ionic radii (e.g. K4) compared to ions with smaller ionic radii (e.g., Na+)
and lower valence. These factors would account for the greater p-NCB sorption occurring
by Cs+- and K+-exchanged soil clay fraction (ionic radii 0.169 and 0.133 nm; Athd -258
and -304 kJ mol", respectively) compared to that by soil clay fractions exchanged with
Na‘“, Li", Ca2+, Mg2+ or Be2+ (ionic radii, 0.095, 0.060, 0.099, 0.065 or 0.135 nm; Athd, -

375, -481, -1515, -1258, or -1258 k] mor‘, respectively).

The effect of enthalphy of hydration can be illustrated by comparing K+ and Bay.
Although these ions have similar ionic radii, at an equilibrium concentration of 5mg L",
our K+-exchanged soil clay fraction sorbed p-NCB to a greater extent (69 times greater)
than the Bay-exchanged soil clay fraction (Table 1.3). The greater hydration sphere and

hydration energy (-1290 kJ mol") of the Ba2+ ion compared to the K+ ion (-314 kJ mol'l)

29

suggests that ion-dipole interactions between the exchangeable cation and polar
functional groups on the NAC would require the displacement of more strongly held
water molecules in the case of Bay-exchanged system. The removal of water requires
energy and so, for direct cation-solute interactions, the Baz+-exchanged system would
require a larger energy input than the K+ system. Supporting F TIR studies observed -NO
vibrational band shifts for 1,3,5-trinitrobenzene sorption on K+-smectite but not for Ba”-

smectite @oyd et al., 2001; Johnston et al., 2001, 2002).

Li et a1. (2004) calculated an enthalpy of interaction involved in complex formation
between the solute and the exchangeable cation. Their analyses considered other energy
yielding or consuming processes including water removal from the clay and partial
dehydration of the solute in the interlayer. The interaction energies between K+-SWy-2
and 1,3,5-TNB, 1,3-dinitrobenzene (1,3-DNB) and 1,4-DNB at 80% dehydration of the
solute were -99, -69, and -73 kJ mol", respectively. The calculated enthalpies of
interaction between Ca2+-exchanged SWy-2 for these compounds were lower ranging
from -30 to -38 kJ mol". The calculated 'interaction enthalpies help explain the
observation of greater NAC sorption by the less strongly hydrated system (e.g. K+) than
by the more strongly hydrated system (e.g., Ca2+). Finally, Cs+ and K+-saturated smectites
manifest interlayer distances of ~ 3 A (dom ~ 12.3 A) in the presence of sorbed NACs,
and this distance appears optimal for sorption of planar NACS (Li et al., 2004; Sheng et
al., 2002; Sheng et al., 2001). These spacings closely match the width of p-NCB and
through simultaneous interactions with the opposing siloxane sheets the adsorbed

molecules are largely dehydrated (assuming 80% dehydrated in the above calculation of

30

interaction enthalpy), an energetically favorable process (Li et al., 2004). Smectites
saturated with the other cations (e.g. Na“, Ba”, Ca2+) always swell to > 15 A in the
presence of bulk water. The energy requirement for compressing the interlayers, and
dehydrating the solute, may be larger than the energy gain from solute dehydration

thereby making the overall process energetically less favorable.

The effects of saturating cation on sorption of p-NCB by whole soils were similar to
those by the isolated soil clay fractions (Table 1.3; Figure 1.10). Cs+ and K+-soil were
approximately 14 and 9 times more effective than Na+ or Li+-soil for p-NCB sorption at
an equilibrium concentration of 5 mg L'1 (Table 1.3); sorption by Cs+- and K+-saturated
soils was also much higher than by the divalent-cation-exchanged soils (Table 1.3; Figure
1.10). Among the divalent cations, the extent of p-NCB sorption at an equilibrium
concentration of 5 mg L'1 also increased with decreasing hydration energy

(Mg2+<Ca2+=Ba2+) (Table 1.3).

31

.0
oi

 

 

 

 

 

 

 

 

 

: . CS+
'5: o K“
g v Na+ o
.. 0 4 - v Li“
S- ' Ba? .
D ” ,,
o 0.3 - a
‘9 .r
m 0
U
z
“6 0.2 « "
C ,
o O
'5
E 7,
g 0.1 -
o a
5 -v‘-’7
0 (I: - 57- ""'
0.0 i ‘ ‘ l I r l l
O 10 20 30 4O 50

Aqueous NCB concentration, C. (mg L")

Figure 1.10. Adsorption isotherms for sorption of p—nitrocyanobenzene by Webster A
soils exchanged with various cations (Cs+, K+, Na+, Li+, Ba”, Mg”, and Ca2+).
Insert is a magnification of adsorption isotherms for sorption of p—nitrocyanobenzene by

Webster A soils exchanged with Na+, Li+, Ba”, Mg“, and Ca”.

32

CONCLUSION

Adsorption of NACs by Webster soil was strongly influenced by the nature of the
dominant exchangeable cation due to their effects on the sorptive characteristics of
smectite clays present in this soil. The observed order of sorptivity of p-NCB as related to
the saturating exchangeable cation could be explained by the hydration energy of
exchangeable cations and its effect on the molecular-scale environment of the clay
interlayers. Increasing the hydration energy of the exchanged cation decreases the extent
of sorption of p-NCB by the clay-sized soil fraction and whole soil due to diminished
effective size of adsorption domains between exchangeable cations, less favorable
interactions between the exchange cation and N02 groups of sorbed NACS, and to greater
interlayer separations which disallows solute dehydration. Exchangeable cation type did
not strongly influence sorption of organic solutes by SOM. Thus in soils lacking smectite
clays, sorption of nonionic organic solutes will probably not be strongly influenced by the
type of exchangeable cation. Conversely, for organic solutes with polar functional groups
(e.g., N02), and in soils containing smectite clays, solute uptake from water will be

strongly influenced by the type of exchangeable cation(s) present.

33

REFERENCES

Boyd, S.A., G. Sheng, B.J Teppen, and CT Johnston. 2001. Mechanisms for the
adsorption of substituted nitrobenzenes by smectite clays. Environ. Sci. Technol.
35:4227-4234.

Chiou. C. 2002. Partition and adsorption of organic contaminants in environmental
systems. John Wiley and Sons, New Jersey.

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

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

Evangelou, V. 1998. Environmental soil and water chemistry: principles and applications.
John Wiley & Sons, Inc., New York.

Gee, G. and J. Bauder. 1986. Particle-size analysis. p. 383-411. In A. Klute (ed.) Methods
of Soil Analysis. Part I — Physical and mineralogical methods. 2nd ed. Agronomy. 9.
ASA, Madison, WI.

Ghosh, K and M. Schnitzer. 1979. UV and visible absorption spectroscopic investigations
in relation to macromolecular characteristics of humic substances. J. Soil Sci. 30:735-
745.

Haderlein, S., and R. Schwarzenbach. 1993. Adsorption of substituted nitrobenzenes and
nitrophenols to mineral surfaces. Environ. Sci. Technol. 27:316-326.

Haderlein, S.B., K. Weissmahr, and R. Schwarzenbach. 1996. Specific adsorption of
nitroaromatic: explosives and pesticides to clay minerals. Environ. Sci. Technol. 30:612-
622.

Howard, P., and W. Meylan. 1997. Handbook of physical properties of organic
chemicals. Lewis Publishers, Boca Raton, FL.

34

Johnston, C., M.F. de Oliveira, B.J. Teppen, G. Sheng, and SA. Boyd. 2001.
Spectroscopic study of nitroaromatic-smectite sorption mechanisms. Environ. Sci.
Technol. 35:4767-4772.

Johnston, C., G. Sheng, B.J. Teppen, S.A. Boyd, and M. F. de Oliveira. 2002.
Spectroscopic study of dinitrophenol herbicide sorption on smectite. Environ. Sci.
Technol. 36:5067-5074.

Karickhoff, S. 1984. Organic Pollutant Sorption in aquatic systems. J. Hydraulic Eng.
110:707-735.

Karickhoff, S., D. Brown, and T. Scott. 1979. Sorption of hydrophobic pollutants on
natural sediments. Water Res. 13:241-248.

Kile, D.E., C.T. Chiou, H. Zhou, H. Li, and O. Xu. 1995. Partition of nonpolar organic
pollutants from water to soil and sediment organic matters. Environ. Sci. Technol.
29:1401-1406.

Kunze, G., and J. Dixon. 1986. Pretreatment for mineralogical analysis. p. 92-99. In A.
Klute (ed.) Methods of Soil Analysis. Part I — Physical and mineralogical methods. 2nd
ed. Agronomy. 9. ASA, Madison, WI.

Laboski, C.A.M., and J .A. Lamb. 2004. Impact of manure application on soil phosphorus
sorption characteristics and subsequent water quality implications. Soil Sci. 169: 440-
448.

Laird, D., Y. Yen, W. Koskinen, T. Steinheimer, and R. Dowdy. 1994. Sorption of
Atrazine on soil clay components. Environ. Sci. Technol. 28: 1054-1061.

Li, H., G. Sheng, B. Teppen, C. Johnston, and SA. Boyd. 2003. Sorption and desorption
of pesticides by clay minerals and humic acid-clay complexes. Soil Sci. Soc. Am. J.
67:122-131.

Li, H., B.J. Teppen, D. Laird, C. Johnston, and SA. Boyd. 2004. Thermodynamics of
nitroaromatic compound adsorption from water by smectite clay. Environ. Sci. Technol.
38:5433-5442.

35

Nelson, D. and L. Sommers. 1996. Total carbon, organic carbon, and organic matter. p.
961-1010. In D. Sparks (ed.) Methods of Soil Analysis. Part III - Chemical methods.
Agronomy. 5. ASA, Madison, WI.

SAS Institute. 1999. SAS 8.1. SAS Institute, Inc., Cary, NC.

Sheng, G., C.T. Johnston, B.J. Teppen, and SA. Boyd. 2001. Potential contributions of

smectite clays and organic matter to pesticide retention in soils. J. Agric. Food Chem.
49:2899-2907.

Sheng, G., C.T. Johnston, B.J. Teppen, and SA. Boyd. 2002. Adsorption of dinitrophenol
herbicides from water by montmorillonites. Clays Clay Miner. 50:25-34.

Soil Survey Staff, Natural Resources Conservation Service, United States Department of
Agriculture. Soil Series Classification Database [Online]. Available at
http://soils.usda.gov/soils/technical/classification/scfile/index.htrnl (verified 10 Feb.
2004).

Sumner, M. and W. Miller. 1996. Cation exchange capacity and exchange coefficients. p.
1201-1229. In D. Sparks (ed.) Methods of Soil Analysis. Part III — Chemical methods.
Agronomy. 5. ASA, Madison, WI.

Weissmahr, K., S. Haderlein, and R. Schwarzenbach. 1997. In Situ Spectroscopic
investigations of adsorption mechanisms of nitroaromatic compounds at clay minerals.
Environ. Sci. Technol. 31:240-247.

Weissmahr, K., S. Haderlein, and R. Schwarzenbach. 1998. Complex formation of soil
minerals with nitroaromatic explosives and other pie—acceptors. Soil Sci. Soc. Am. J.
62:369-378.

Weissmahr, K., M. Hildenbrand, R. Schwarzenbach, and S. Haderlein. 1999. Laboratory
and Field Scale Evaluation of Geochemical Controls on Groundwater Transport of
Nitroaromatic Ammunition Residues. Environ. Sci. Technol. 33:2593-2600.

36

Whittig, L and W. Allardice. 1987. X-ray diffraction techniques: Section 12-3.3. p. 338-
342. In A. Klute (ed.) Methods of Soil Analysis. Part I — Physical and mineralogical
methods. 2nd ed. Agronomy. 9. ASA, Madison, WI.

37

CHAPTER 2
AVAILABILITY OF CLAY SURFACES IN SOILS FOR ADSORPTION OF

ORGANIC CONTAMINANTS

ABSTRACT

In soil, clay and soil organic matter (SOM) are usually intimately associated with each
other. Clay surface coverage by SOM may limit the efficacy of the clay mineral fraction
for the sorption of organic contaminants and pesticides in soil. Accordingly, one would
predict that the sorption of such solutes by soil might differ from the summation of their
sorption to the soil’s isolated components. Two methods were used to assess the
availability of clay surfaces in a Webster soil for sorption of p-nitrocyanobenzene (p-
NCB), a solute known to be strongly sorbed by smectite clays. One method offered
earlier by Karickhoff (1984) involves the summation of the sorption of p—NCB to SOM
and swelling clays. In order to determine the fractional availability of the clay surfaces
(i.e., the fraction of the clay mineral surface available for sorption within soil, fa) for p-
NCB sorption utilizing Karickhoff‘s equation, several assumptions must be made and/or
procedural difficulties overcome in the determination of each term of the equation. We
developed an alternative approach for determining f, that alleviates these methodological
limitations. Preliminary results reveal good agreement between fa values obtained from
both methods of analysis for p-NCB but not diuron sorption. Generally, for p-NCB
sorption, f, values varied between 0.55 and 0.71. For diuron sorption, using our

alternative equation for data analysis, fa values varied between 0.28 and 0.47. This

38

suggests that organic matter does suppress sorption of p-NCB and diuron by Webster A

soil by 29 to 45% and 53 to 72%, respectively.

39

INTRODUCTION

Nitroaromatic compounds (NACS) are commonly used as pesticides, explosives,
intermediates in the synthesis of dyes, ammunition, and solvents. Their uses have resulted
in contamination of soils, sediments, surface waters, and aquifers. Understanding NAC
mobility and reactivity in the environment is needed to accurately predict their
environmental fate. Clay minerals and soil organic matter (SOM) are generally
considered the two most active soil components in the sorption of aqueous phase organic
contaminants such as NACs. For compounds with functional groups (e.g. N02) capable
of engaging in specific interaction with clays, adsorption to soil clays may occur to a

greater extent than to SOM (Sheng et al., 2001; Sheng et al., 2002).

Significant sorption of NACs from water by clays, particularly smectites, is well
documented. The high affinity of NACs for smectites has been attributed to electron
withdrawing properties of the N02 group which may promote formation of electron
donor-acceptor complexes (Haderlein and Schwarzenbach, 1993; Haderlein et al., 1996;
Weissmahr et al., 1997), and/or complexation of the NAC with the exchangeable cations
on the clay (Boyd et al., 2001; Johnston et al., 2001, 2002; Sheng et al., 2002; Li et al.,
2003). Sorption of NACs by smectites is strongly influenced by the nature of the
exchangeable cation. Monovalent cations with comparatively low hydration energies
(e.g., K+) manifest much higher NAC sorption than more strongly hydrated mono- or
divalent cations (e. g. Na+, Mg”, Ca“). In addition, the increasing number of N02 groups,

a planar molecular structure, and the absence of “bulky” substituents (Haderlein et al.,

40

1996; Haderlein and Schwarzenbach, 1993; Sheng et al., 2002; Johnston et al., 2001; Li

et al., 2003) promote NAC-clay interactions.

The extent of interaction of NACs with clay minerals may, however, be limited by the
availability of the clay sorption sites within soil. Soil organic matter may obscure clay
surfaces hence reducing the efficacy of the clay mineral fraction for the adsorption of
NACs. Organic matter associated with clays is known to facilitate aggregate formation
through cohesion of clay and/or clay-humus complexes into larger particles through H-
bonding, coordination with polyvalent cations, and/or cementation or encapsulation
(Stevenson, 1994). The extent to which the clay surfaces in any soil are coated by organic
substances depends on the nature and amount of clay and SOM (Stevenson, 1994). Using
scanning electron microscopy Laird et al. (2001a) observed, in the clay fraction of a
Webster Ap soil from which SOM was not removed (i.e., whole clay fraction), both large
aggregated structures (5-20 um) and discrete particles (0.5-2 pm). They described the
surfaces as having a diffuse-porous appearance and suggested that the surfaces were
covered with humic materials. Scanning electron micrographs of the clay fraction treated
with H202 to remove humic materials showed the presence of numerous individual
mineral grains, only a few small aggregates and no large aggregates (Laird et al., 2001b).
This suggested the involvement of humic substances in aggregate formation and the
destruction of aggregates through H202-treatment. Laird et al. (1994; 2001b) found that
the SOM associated with the clay fraction of a Webster Ap soil was dependent on the
particle size of the soil clay fractionate. The fine clay fraction (<0.02 um) was relatively

enriched with aliphatic carbon, bound hydroxyls, and amines and was more humifred

41

based on the C/N ratio of the associated SOM (Laird et al., 2001a), as compared to the

coarse (0.2-2.0 um) and medium (0.02-0.2 um) clay fractions.

The influence of organic components on the sorptive properties of the soil clay mineral
fraction is still not fully understood despite numerous studies on organic compound
sorption by synthetic clay-organic complexes. Li et a1 (2003) did not find any measurable
impact of humic acid (HA) on NAC sorption by K+—saturated HA-smectite complexes.
However for Cay-saturated HA-smectite complexes, HA appeared to obscure some
mineral surfaces resulting in reduced sorption of 4,6—dinitro-o-cresol. Pusino et al. (1992)
compared experimentally measured sorption coefficients of metolachlor by a HA-clay
complex with those estimated assuming an independent adsorptive behavior by the clay
and humic components. They found that the estimated Kd values were higher than the
experimental values indicating that the humics reduced the availability of the clay
surfaces for sorption. Pusino et al. (1994) found reduced sorption of triclopyr by HA-clay
complexes compared to pure clay systems. Celis et al. (1999) found that HA coating on
clay-ferrihydrite binary complexes reduced the sorption of 2,4-dichlorophenoxyacetic
acid. Celis et al. (1997) found reduced sorption of thiazafluron by Ca2+-Wyoming
montmorillonite when coated with HA. Removal of the HA from the clay surface resulted
in increased sorption of thiazafluron. Celis et al. (1999), Celis et al. (1998) observed
decreased specific N2 BET surface areas of synthetic HA-clay complexes compared to
corresponding clays devoid of HAS. They proposed that the HAs were associated with the

external surfaces of the clay.

42

Karickhoff (1984) presented an equation that could be used to experimentally assess the
availability of soil clay surfaces for sorption of nonionic organic contaminants (NOCs).
This equation summed NOC sorption to clay and SOM at a specific relative
concentration, which incorporated the potentially reduced availability of sorption sites on
clays in whole soil:

Qwhole soil = Qmin + Qsom = (fa * Kmin * Ce * fmin) + (Kom "' Ce * fom) [1]
where Qwho;e so" is the total solute mass sorbed per unit mass of whole soil (mg g"), Qmin
and Qsom are the mineral fraction (i.e., soil without SOM)-sorbed and SOM-sorbed solute
per unit mass of sorbent, respectively, Cc is the aqueous solute equilibrium concentration,
and Km and K0m represent the sorption coefficients of the solute by the soil fraction from
which SOM was removed (mineral fraction) and SOM, respectively, fmm and f0.m are the
fractional mineral fraction and SOM contents of soil, and f, represents the fractional
availability of sorption sites on the clay in whole soil (i.e. the fraction of the sorptive
surfaces that are available in unfractionated soil). Fractional availability can range from 0
(unavailable) to 1 (100% available). Karickhoff (1984) equates the mineral fraction with
the clay fraction owing to the high surface area of smectites in the clay fraction. Note that
equation 1 assumes that the clays and SOM are the only sorptive materials in soil. This is
probably a good assumption for NACs, which exhibit insignificant sorption to A1 or Fe

(hydr)oxides, carbonates, or quartz G—laderlein and Schwarzenbach, 1993).

A few studies have shown how the extent of sorption of NOCs to soil or HA-clay

complexes does not equate to the sum of its individual components (Pusino et al., 1992;

Celis et al., 1997; Liu et al., 2002). However, there appear to be no studies that have

43

attempted to experimentally quantify the extent to which clay surfaces are available for
sorption in soil. The objective of this study was to quantify the fractional availability of
mineral surfaces for sorption of NACs by whole soil using the Karickhoff equation and

our newly proposed method of data analysis (described below).

44

EXPERIMENTAL SECTION

The A horizon of a Webster clay loam (fine loamy, mixed, mesic Typic Haplaquoll) was
air-dried and sieved to remove coarse fragments (>2 mm). The soil was fractionated by
standard procedures (Kunze & Dixon, 1986). Carbonates (CO3) and SOM were
sequentially removed from each soil horizon. Carbonates were removed by stirring the
soil with 0.5N sodium acetate acidified to pH 5 while heating at 80°C, followed by the
removal of SOM with 30% H202 at 80°C (denoted OM-removed soil). Homoionic soils
and soil fractions were prepared by washing the soils with 200 ml aliquots of 0.1M
solution of KCl or MgCl2 overnight, thereafier centrifuging and discarding the
supernatant. This procedure was repeated four times. The soils were then washed with
200 ml aliquots of Milli-Q deionized H2O, thereafier centrifuged and the supernatant
tested with AgNO3. The procedure was repeated until a negative Cl' ion test with AgN03
was obtained. The soils were then freeze-dried and stored at room temperature (23 i 1°C)
until used. Para-nitrocyanobenzene (p-NCB) was purchased from Aldrich Chemical
Company (Milwaukee, Wisconsin) with a purity of 97% and used as received. Its water
solubility (SW) at 25°C is 1650 mg L'1 and its log K0w value is 1.19 (Howard and Meylan,
1997). Diuron (1,1-dimethyl-3—(3,4-dichlorophenyl)urea was purchased from Aldrich
Chemical Company (Milwaukee, Wisconsin) with a purity of 98% and used as received.

Its Sw at 25°C is 42 mg L'1 and its log K0W value is 2.75 (Howard and Meylan, 1997).

The clay-sized particles (<2.0 um) were obtained from soil using wet sedimentation after
removal of CO3, SOM and free iron oxides (Fer) (Whittig & Allardice, 1987), then

saturated with K+ and Mg2+ as described above. Dispersed suspensions of Mg2+- and K‘-

45

saturated clay-sized particles were dropped on to glass slides to produce clay films. The
clay films were then treated by air-drying, heating at 500°C, and ethylene glycol. X-ray
diffraction patterns of the treated clay films were used to identify the clay mineralogy of
the clay-sized fraction of the soil (Whittig & Allardice, 1987). The clay samples were
analyzed using CuKa radiation and a Philips APD 3720 X-ray diffractometer (Philips

Electronic Instruments, NJ).

Semi-quantitative estimates for the amount of smectite in the unfractionated Webster A
soil was attempted using the measured cation exchange capacity (CEC) of the
unfractionated soils using the ammonium acetate method (Sumner and Miller, 1996)
(Michigan State University (MSU) Soil Testing Laboratory), and an average CEC of a
soil smectite and SOM of 80 and 160 cmol kg", respectively, using the equation:

CEC = (160 cmol kg")(fom) + (80 cmol kg")(fcm) [2]
where for“ and fcm are the fractional SOM and smectite clay contents of the soil. The
calculation assumed that the CEC of the soil originated solely from SOM and soil
smectites. By comparing the X-ray diffraction peak areas associated with the non-
smectite minerals to the smectite peak, we estimated the relative fractional amounts of
each non-smectite non-quartz mineral present in the clay-sized fraction as a fraction of
the smectite peak (illite/mica, 0.125; kaolinite, 0.125). Multiplying the percent smectite in
the soil obtained using equation (2) (e.g., 24%) by the relative fractional amount of each
non—smectite mineral determined from the relative peak areas in the X-ray diffraction
graph, we calculated the amount of non-smectite minerals (e.g. illite/mica, 3%; kaolinite,

3%). We subtracted the estimated smectite content (e.g. 24%) from the total clay content

46

(e.g. 31%) to obtain the amount of non-smectite minerals (e.g. 7%) and compared this

value to the sum of the non-smectite minerals estimated using the peak areas (3% + 3%).

Removal of SOM with H202 may result in the liberation and/or oxidation of metals
previously bound to SOM which could result in the formation of sesquioxides. The
potential contribution of sesquioxides to p-NCB sorption was assessed. For this purpose
ferric hydroxide was prepared by mixing FeCl; with KOH; KOH was present in excess of
stoichiometric quantities needed to react with all of the ferric ions in the system. The
Fe(OH)3 suspension was centrifuged for 30 min at 4068g, the supernatant discarded and
the pellet washed five times with 200 ml portions of 0.1M KCl. The F e(OH)3 was then
washed with deionized H2O to a negative Cl' ion test with AgNO3, freeze-dried and
stored at room temperature (23 i 1°C). X-ray diffraction analysis was performed on an
air-dried film of Fe(OH)3 prepared by using a disposable pipette to drop a dispersed

suspension of Fe(OH)3 on to a glass slide.

The effect of exchangeable cation type on sorption by SOM was evaluated using either
organic soils relatively devoid of mineral matter or a mineral soil devoid of swelling
clays (i.e. smectites) implicated in the sorption of NACs (Boyd et al., 2001; Haderlein
and Schwarzenbach, 1993). The sorbent used was the Pahokee peat (euic, hyperthermic
Lithic Medisaprist) (Soil Survey Staff, NRCS). The peat was purchased from the
International Humic Substances Society (BS103P) (Denver, Colorado). Sorption

isotherms for diuron by these soils were obtained using the batch equilibrium technique

47

described above. These soils were saturated with K+ or Mg2+ to evaluate the effect of

cation type on sorption of diuron.

Sorption isotherms for p-NCB sorption by K+- and Mg2+-saturated unfractionated soils
and SOM-removed soils as well by Fe(OH)3 were obtained, in triplicate, using a batch
equilibrium technique. This involved adding aqueous solutions of p-NCB (1-42 mg L")
or diuron (1-30 mg L") over a range of initial aqueous concentrations (C,) to a known
mass of soil. The p-NCB and diuron solutions were prepared in 0.05M KCl or MgCl2
corresponding to the saturating cation. Five ml of the solute solutions were pipetted into
7.5 ml borosilicate glass vials containing various amounts of sorbent (between 0.2 and
0.4 g for soil). The vials were rotated continuously for 24 hours in the dark at room
temperature (23 i 1°C); preliminary studies showed equilibrium was reached within 24
hours. The liquid and solid phases were separated by centrifugation at 4068g for 30 min,
and the solute’s equilibrium concentration in water (Ce) was determined using a Perkin-
Elmer high performance liquid chromatograph (HPLC) (Perkin-Elmer, Norwalk, CT)
consisting of a binary LC pump 250 with a series 200 autosampler and a uv-visible
detector set at It 254 nm for p-NCB and A 248 nm for diuron. A platinum EPS C 18
column from Alltech Associates was used (Deerfield, IL). The mobile phase was an
isocratic mixture of methanol (HPLC grade) and Millipore Milli-Q (Billerica,
Massachusetts) deionized water (methanol: water ratio of 55:45 for p-NCB and 75:25 for
diuron) with a flow rate of 1.0 ml min". The amount of solute sorbed was calculated as
the difference between the initial aqueous concentration (Ci) and the final aqueous

concentration (Ce). Compound recoveries in control vials without sorbent were above

48

95% and not used to adjust the data. The extent of solute sorption was quantitated by
fitting the data to the F reundlich equation, Q=K¢Ce", where Q is the sorbed solute
concentration per unit mass of the sorbent (mg g"), Kf is the sorption coefficient, n is a
constant related to the isotherm curvature (concave or convex). The Michigan State
University (MSU) Soil Testing Laboratory determined the SOM content of the
unfractionated (3.87% OM) and OM-removed soil (0.23% OM) using dry combustion
and a Leco carbon analyzer (St. Joseph, Michigan) (Nelson and Sommers, 1996), and the
gravimetric clay content (31%) of the Webster soil by the hydrometer method (Gee and

Bauder, 1986).

Data Analysis

Calculation of f, values

Each sorption isotherm data set was fitted to the Freundlich equation, with an r2 >09 in
all cases. Using the Kf and n values obtained, the Q value was calculated for four
equilibrium aqueous phase concentrations, 5, 10, 15, 20 mg L". Using either the
Karickhoff [1] or our alternative equation (presented below), four corresponding f, values
were obtained for each soil (i.e., one for each concentration). Standard deviations of each
f, value associated with each concentration were calculated using error propagation
procedures. For the Karickhoff equation, only the K+-saturated system was used because
preliminary results indicated that this system provides sufficient difference in sorption
between the unfractionated and OM-removed soils to facilitate accurate determination of
f, values. The independently obtained K0m values utilized when analyzing fa by equation

[1] were previously measured as discussed in Chapter 1.

49

Estimation of degree of OM mineral surface coverage

The maximum percent of the smectite clay surface covered by the SOM was estimated
using Lagaly’s formula area per carbon atom of 5.67 A2 (Lagaly, 1981), a percent carbon
in SOM of 62% (Stevenson and Cole, 1999), an average smectite surface area of 750 m2
g" (Sposito, 1984), the estimated percent smectite in unfractionated soil, and the
measured percent SOM in the unfractionated and OM-removed soil. The operating
assumption underlying this exercise was that all carbon from SOM resides on smectite
surfaces. The calculation therefore provides an estimate for the maximum extent of
coverage of the sorptive mineral surface by the SOM present. So: with 3.87% SOM and
24% smectite in Webster A (WA) soil:

lg of WA has a surface area of oc = ((0.62 x 0.0387)/12)(6.023 x 1023)(5.67)(1O'2°) =
68.27 m2,

1 g of WA has a smectite clay surface area = (750 x 0.24) = 180 m2, therefore

The maximum percent of smectite clay surface covered by CC = (68.27/180) = 38%.

Statistical Analysis

The sorption coefficient (Kf) and the constant related to the sorption isotherm curvature
(n) were estimated using the Freundlich equation, Q=Kch". Comparisons of sorption
isotherms were made by comparing the Kr and n values for each soil-solute system using
nonlinear regression in SAS (SAS Institute, 1999). A significant difference between
isotherms was assigned when there was a significant difference between Kf and n or Kf

values of compared isotherms (Laboski and Lamb, 2004).

50

RESULTS AND DISCUSSION

Sorption of p-NCB and diuron by Fe(OH)3 was negligible (data not shown) and therefore
sesquioxides are not expected to act as sorbents for p-NCB or diuron in whole soils. This
is consistent with studies by Haderlein and Schwarzenbach (1993) who reported that
NACs did not adsorb significantly to aluminum and iron (hydr)oxides, carbonates or
quartz. In addition, Laird et al. (1994) found no indication of surface coatings of particles

in the fine clay fraction of a Webster soil by free Fer.

Considering the two soil components responsible for NAC sorption by soils, SOM and
swelling clays, sorption of NACs by SOM is not influenced by the nature of the
saturating cation. We observed no statistically significant effect of exchangeable cation
(K+ and Mg”) on p-NCB sorption by peat, Houghton muck and Brookston loam soils or
diuron sorption by peat. This was reflected in the observed proximity of the sorption
isotherms for p—NCB sorption by K+- and Mg2+-saturated peat, Brookston loam and
Houghton muck each normalized to the SOM content in each soil (Figure 2.1). These
soils are composed of either predominantly SOM (peat and muck) or are devoid of clays
known to sorb NACs (i.e. swelling clays) and therefore sorption in these systems is
dominated by SOM. We also observed no statistically significant effect of exchangeable
cation (K+ and Mg”) on diuron sorption by K+- and Mg2+-saturated peat (Figure 2.2).
These results provide evidence that sorption of p-NCB and diuron by SOM, in systems
where sorption is dominated by SOM, is not strongly influenced by the saturating cation.
This can be explained by sorption to SOM being a partition process (Chiou, 2002) and

such a dissolution mechanism is not expected to be influenced by the

51

 

 

 

 

 

 

 

 

,_ 1'8 e Put-K a
7, J o Peat-Mg
g 1-4 v HM-K
_, v HM—Ilg :
a 12 . I BL-K .9
- o BL-Mg
3 10 " e.
o .
a e
8 0.8 - ,
z 0
U- 0
° 0.6 - °
5 . v. -
z 04 v, 1
g a ..
2 0.2 9’ - '
8 {‘1 g .. -: "‘ ':-'-.-3:‘ :- =‘
0.0 5“-c. I r I
o 10 20 30 4o

Aqueous NCB concentration, C. (mg L")
"a 3'0 0 Peat (K) b
‘5” O Houghton muck (K)
‘5' 2.5 . v Brookston loam (K)
o v Peat (Mg)
‘2 I Houghton muck (Mg)
2 2.0 . D Brookston loam (Mg)
E . .

1.5 d v
E . ' 'v .-
o ‘15 -
«ii 1.0 - o ;- ' ,/‘..’
g ’3' a”:
'0- e {’45, I
3 0.5 4 _7”‘
5 14-!
e
< 0.0 u I I I
0 1O 20 30 40

Aqueous NCB concentration, C. (mg L")

Figure 2.1. Adsorption isotherms for sorption of p-nitrocyanobenzene (p-NCB) by (a)
K+- and MgZI-saturated peat, Houghton muck, Brookston loam and (b) sorption isotherms

normalized to amount of SOM in the soils.

52

 

3.5

0 peat-K Diuron

 

Concentration of diuron sorbed. 0 (mg 9")

 

 

 

0.0 V I j’ I fl I I I

0 2 4 6 8 1 0 1 2 14 1 6
Aqueous diuron concentration, C. (mg L")

Figure 2.2. Adsorption isotherms for sorption of diuron by K+- and Mg2+-saturated peat,

Houghton muck, Brookston loam.

53

saturating cation. It appears that any significant effects of cation saturation observed for
p-NCB or diuron sorption to soil and soil fractions would result from the effect of cation
saturation on the sorptive characteristics of the clay-sized mineral fraction and not the
SOM. The clay types in the soil clay-sized fraction were identified as mica and/or illite,
smectite, vermiculite, and kaolinite. This was consistent with the clay mineralogy

reported by Laird et al. (1991) for a Webster soil.

Unfractionated and SOM-removed soils saturated with K+ (Kr 0.0528 and 0.0922 L kg";
n 0.5983 and 0.5012, respectively) sorbed p-NCB to a much greater extent compared to
Mg2+-saturated soils (K, = 0.0067 and 0.0007 L kg"; n = 0.6219 and 0.9033,
respectively) (Figure 2.3). These results are consistent with several studies that report
stronger sorption of NACS by reference clays and soils saturated with cations of
relatively low hydration energies (e.g. K+, Cs+) compared to that by sorbents saturated
with cations of relatively higher hydration energies (e.g. Na+, Ca2+, Mg”) (Li et al., 2003;
Sheng et al., 2002; Haderlein et al., 1996; Haderlein and Schwarzenbach, 1993;
Weissmahr et al., 1997, 1999; Johnston et al., 2002). Cations of larger hydrated radii
(e.g., Mg”) occupy a larger surface area of the clay interlayer thereby diminishing the
effective size of adsorption domains between cations and facilitating less NAC sorption
compared to clays saturated with cations of smaller hydrated radii (e.g., KT). In addition,
hydration water associated with cations having smaller enthalpies of hydration (e.g., K+,
Athd -314 kJ mol") are less strongly held compared to water molecules hydrating
cations with larger enthalpies of hydration (e.g., Mg2+, Athd -1258 kJ mol"). Large

hydration spheres and strongly held waters of hydration would act to inhibit the direct

54

 

    
  
   

0.5
p-NCB e WA-K
a o WAC-K
0.4 - v WA-Mg
v WAC-M9

.°
U
r

P
N
i

l

   

P
.s

 

 

 

  

 

P
o

 

r

0 10 20 30
Aqueous NCB concentration, C. (mg L")

I

Concentration of NCB sorbed, Q (mg g")

 

0.20
0.18 '
0.16 '
0.14 -
0.12 i
0.10 -
0.08 -
0.06 ‘
0.04 r
0.02 -

0.00 '
0 5 10 15 20 25 30
Aqueous diuron concentration, C. (mg L")

WA‘K Diuron

<10.

JV C

 

 

 

 

I

Concentration of diuron sorbed, 0 (mg 9")

Figure 2.3. Adsorption isotherms for p-nitrocyanobenzene (p-NCB) and diuron sorption

by K+- and Mg2+-saturated by Webster A soil (whole soil, WA, and soil from which SOM
was removed, WAO).

Isotherms with the different letters are statistically different at p < 0.05.

55

interaction of N02 groups with the interlayer cation (Boyd et al., 2001; Johnston et al.,
2001, 2002; Li et al., 2004). Li et al. (2004) calculated an enthalpy of interaction between
the NAC and the exchangeable cation. This calculation accounted for the associated
processes of water removal required (an energy consuming process) for direct interaction
between the cation and the NAC, and partial (80%) dehydration of the solute in the clay
interlayers (see below). The interaction energies between K+-SWy-2 and 1,3,5-
trinitrobenzene (1,3,5-TNB), 1,3-dinitrobenzene (1,3-DNB) and 1,4-DNB were -99, -69,
and -73 kJ mol", respectively (Li et al., 2004). The calculated enthalpies of interaction
between Cay-exchanged SWy-2 for these compounds ranged from -30 to -38 kJ mol"
(Li et al., 2004). These interaction enthalpies reflect greater NAC sorption by the less
strongly hydrated system (e.g. K+) than by the more strongly hydrated system (e.g., Ca2+).
Previously, KI-saturated smectites were reported to manifest interlayer distances of ~ 3 A
(door ~ 12.3 A) in the presence of sorbed NACs (Li et al., 2004; Sheng et al., 2002; Sheng
et al., 2001); a distance which closely matches the width of p-NCB and appears optimal
for sorption of planar NACs. That is, through simultaneous interactions with the opposing
siloxane sheets the adsorbed molecules are largely dehydrated, an energetically favorable
process (Li et al., 2004). Smectites saturated with cations such as Na+, Ca2+, or Mg2+
always swell to _>_ 15 A in the presence of bulk water. The energy requirement for
compressing the interlayers, and dehydrating the solute, may be larger than the energy
gain from solute dehydration thereby making the overall process energetically less
favorable. Therefore Mg2+-saturated systems exhibit lower NAC sorption than K+-

saturated systems (Figure 2.3).

56

Sorption of p-NCB by the Mg2+—saturated soil was greater than that by the Mg2+-saturated
SOM-removed soil (Figure 2.3). Removal of SOM may liberate some clay surface sites,
but, as discussed above, Mg2+-exchanged clays have low affinities for aqueous phase
NACs. Hence, in the Mg2+-saturated SOM-removed soil, clays are relatively inactive as
sorbents for sorption of p-NCB. Rather, SOM is the principal sorbent phase hence with

its removal p-NCB sorption decreases (Figure 2.3).

The K+-saturated OM-removed soils statistically sorbed more p-NCB than the K‘-
unfractionated soils (Figure 2.3). This indicates that the presence of SOM in whole soil
reduces p-NCB sorption. Although SOM acts as a sink for p-NCB, its removal may
liberate clay sorptive surfaces which could serve as additional sites for p-NCB sorption.
Additional sorption of p-NCB by these unobscured sites apparently exceeds that afforded
by SOM so that the net result of SOM removal is to increase p-NCB sorption compared
to that by whole soil. Suppressed sorption by OM suggests f, values (i.e. the fraction of
the sorptive surfaces that are available in unfractionated soil) less than 1. To
quantitatively assess fa, we used two approaches; the Karickhoff equation [1] and an
alternative equation introduced herein. One potential difficulty with the determination of
f,, using Karickhoffs equation [1] is the variation in the K0m values for the test solute and
its potential influence on the calculated fa value. Kile et al. (1995) reported a variation of
a factor of 2 to 3 for K0c (organic carbon normalized partition coefficient) values for
dichlorobenzene and carbon tetrachloride for a series of soils and sediments. For solutes
that sorb preferentially to SOM the effect of the variation in K0m values may be large

compared to that for solutes that preferentially sorb to smectites (e. g., NACS). In addition,

57

in order to determine the isolated SOM contribution to NAC sorption by a specific soil, a
surrogate for that SOM must be used as a sorbent since the SOM cannot be isolated in
total or without alteration. This difficulty introduces error into the equation since Kom is

not obtained by direct measurement on the system being analyzed.

An alternative equation to the Karickhoff equation [1] is proposed that eliminates reliance
on Kom to determine clay surface fractional availability, fa, in whole soils. It involves
measuring the difference in NAC sorption between sorbent systems whose cation
exchange capacity (CEC) was saturated with different cations. This method is based, in
part, on the observation that NAC sorption by SOM whose CEC is exchanged with
various cations was unaffected by the nature of the saturating cation (Figure 2.1 as
determined in Chapter 1). Using peat, Brookston loam and Houghton muck, each
saturated with K+ or Mg2+ ions, as the sorbents for p-NCB, we determined that p-NCB
sorption by the corresponding K+- and Mg2+-saturated sorbent did not differ (Figure 2.1).
The method is also based on the observation that NAC sorption by clay is strongly
dependent on the cation saturating the clay exchange sites (Sheng et al., 2002; Johnston et
al., 2001, 2002; Li et al., 2003; Haderlien et al., 1996; Haderlein and Schwarzenbach,
1993; Boyd et al., 2001). For example, K+-smectites have high affinities for NACs while
Mg2+-smectites do not (Haderlein et al., 1996; Haderlein and Schwarzenbach, 1993;
Johnston et al., 2001; Boyd et al., 2001; Sheng et al., 2002). We therefore propose if two
experiments are done in which NAC sorption is measured for a soil saturated with K+

versus Mg2+ cations, then from equation 1:

58

Qwhole soil, K sat - Qwhole soil, Mg sat = {fa "‘ {(Kmin, Ksat * Cc * fmin)’ (Kmin, Mg sat * Ce * fmin)}} +
{(Kom,1< sat "‘ Ce “ fem)- (Kora, Mg sat * Cc * fom)} [2]
where Qwhote ,0“, K ,3, and Qwho.e ,0", Mg ,3. are the total contaminant masses sorbed by whole
soil whose CEC is saturated with either K+ and Mg2+, respectively; Km, K 5,, and Km)“, Mg
5,. represent the partition coefficients normalized to soil fraction from which SOM was
removed (mineral) for sorbent systems saturated with either K“ and Mg”, respectively;
Kom, K 5,, and Kom, Mg 5,. represent the partition coefficients normalized to SOM for sorbent
systems saturated with either K+ and Mg”, respectively; fmin is the fration of the whole
soil remaining after removal of bulk SOM; for" is the fractional SOM contents of soil; fa
represents the fractional availability of clay surfaces in whole soil.

Since Kom, K sat = Kom, Mg ,3, then:

Qwhole soil, K saturated - Qwholc soil, Mg saturated =

{fa * {(Kmin.1( sat "' Ce * fmin)’ (Kmin, Mg sat * Cc * fmin)}} [3]
Equation [3] eliminates the Korn and fom terms, which are needed in the Karickhoff [1]
equation. Within both the Karickhoff [1] and alternative [3] equations, we accounted for
the extent of p-NCB sorption by the residual SOM associated with the SOM-removed
fraction. We determined that SOM remaining after removal of bulk SOM sorbs p-NCB
similarly to that of bulk SOM. However, in the alternative equation, the effect of K+-
saturated residual SOM on p-NCB sorption cancels out that of Mg2+-saturated residual
SOM. When K+-saturated peat, Houghton muck or Brookston loam were used as the
surrogate sorbent for determining the organic contribution in equation [1], the f, values

did not differ greatly (Table 2.1). When Mg2+-saturated sorbents were used as the SOM

59

Table 2.1. Calculated f, values for p-nitrocyanobenzene (p-NCB) sorption by Webster A
soil using Karickhoff’s equation [1] (using p-NCB sorption by K+- and Mg2+-saturated
peat, Houghton muck and Brookston loam to determine the K0m term), average fa values

for K+- and Mg2+-saturated sorbents and our alternative equation [2].

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Equation Concentration 1'. values 1‘. values Average f.
(mg L") (lC-saturated) (Mg2+-saturated) values
Karickhoff 5 0.55 (0.02) 0.59 (0.02) 0.57 (0.02)
(peat) 10 0.60 (0.02) 0.63 (0.02) 0.62 (0.02)
15 0.63 (0.03) 0.66 (0.03) 0.65 (0.03)
20 0.65 (0.03) 0.68 (0.03) 0.67 (0.03)
Karickhoff 5 0.60 (0.02) 0.62 (0.02) 0.61 (0.02)
(Houghton 10 0.65 (0.02) 0.66 (0.02) 0.66 (0.02)
muck) 15 0.69 (0.03) 0.69 (0.03) 0.69 (0.03)
20 0.71 (0.03) 0.71 (0.03) 0.71 (0.03)
Karickhoff 5 0.58 (0.02) 0.61 (0.02) 0.60 (0.02)
(Brookston 10 0.62 (0.02) 0.65 (0.02) 0.64 (0.02)
loam) 15 0.64 (0.03) 0.68 (0.03) 0.66 (0.03)
20 0.66 (0.03) 0.70 (0.03) 0.68 (0.03)
Alternative 5 0.61 (0.02) 0.61 (0.02) 0.61 (0.02
10 0.65 (0.02) 0.65 (0.02) 0.65 (0.02
15 0.68 (0.03) 0.68 (0.03) 0.68 (0.03)
20 0.70 (0.03) 0.70 (0.03) 0.70 (0.03)

 

 

 

Across each concentrations, f, values calculated using K+-saturated Karickhoff-peat <
Karickhoff-BL < Karickhoff-HM = Alternative (p 5 0.05). The fa values at each
concentration were statistically different (5ppm < 10ppm< 15ppm < 20ppm) (p < 0.05).
Standard deviations in parentheses.

60

surrogate, f, values did not differ greatly from those determined using the K+-saturated
sorbents (Table 2.1). This is a direct manifestation of our determination that, on a unit
SOM basis, the sorption isotherms for p-NCB sorption by K+- and Mg2+-saturated peat,
Houghton muck and Brookston loam were similar (Figure 2.1). The assumption is that
these soils could be used to determine the isolated SOM contribution of Webster soil for
sorption of p-NCB. Comparing the fa values for p-NCB sorption derived using the
Karickhoff [1] and alternative [3] equation, the values did not differ greatly (Table 2.1).
This is because the SOM term in the Karickhoff equation is very small (only ~20% of
whole soil p-NCB sorption at Cc = 10 mg L") compared to the mineral fraction term. As
noted earlier, NACs sorb to a greater extent to clays than to SOM (Sheng et al., 2001) and
therefore the organic contribution and, by extension the error associated with the term, is

small, when using NACs as the probe solute to estimate fa values.

For diuron, the K+-saturated OM-removed soil (Kf, 0.0137 L kg"; n, 0.6043) sorbed less
than K+-saturated unfractionated soil (Kr, 0.0277 L kg"; n, 0.6274) for Webster A soil
(Figure 2.2). This demonstrates that SOM enhanced sorption of diuron by unfractionated
soil and that removal of SOM diminished sorption in excess of any enhanced sorption
that might result from increasing mineral surface availability. Sheng et al. (2001)
observed greater sorption of diuron to Houghton muck soil (representing SOM)
compared to K+-SWy-2 smectite. Parker and Rate (1998) suggest that surface coatings of
SOM may promote sorption of the solute to soil if the SOM surface coating is a more
effective sorbent than the clay particles they obscure. Murphy et al. (1990) found

increased sorption of hydrophobic compounds (e.g., anthracene (log K0W 4.45)) by

61

hematite and kaolinite coated with natural humic substances compared to the pure
minerals. Sanchez-Martin and Sanchez-Camazano (1991) observed that the removal of
organic matter from soils resulted in less chloridazon (log Kow 1.14) sorption compared to
whole soil. Cox et al. (1998) reported similar findings for imidacloprid (log K0W 0.57)
sorption by soil clay fractions and model sorbents. Li et al. (2003) observed enhanced
dichlobenil (log K0w 2.74) sorption by humic acid (HA)-modified Cay-smectites (SWy-2
and SAz-l) compared to that by unmodified clays. Celis et al. (1998) observed a similar
trend for atrazine (log K0w 2.51) sorption by Ca-SWy-HA complexes. Pusino et al. (1992)
observed decreased sorption of metolachlor (log K0w 3.13) sorption by a sandy loam and
a loamy sand soil after removal of SOM. Hence, the sorption of diuron by the liberated
sorption sites may not exceed that by the removed SOM and thus overall sorption

decreases (Rebhun et al., 1992).

For diuron, using equation [1], negative fa values were calculated using K+- and Mg2+-
saturated peat to determine the SOM contribution to solute sorption (Table 2.2). In
contrast, positive fa values were obtained when calculated using our alternative equation
[2] (Table 2.2). The negative fa values obtained using equation [1] point out an inherent
weakness in the data analysis suggested by Karickhoff (1984) especially for
compounds/soils where SOM is the dominant sorptive component. We suggest that
because of the sensitivity of the Karickhoff equation to the accurate determination of
SOM content and Kom, the alternative equation [2] is more reliable for fa determinations.

This is particularly true for compounds that sorb to SOM to a greater extent than to

62

Table 2.2. Calculated f, values for diuron sorption by Webster A soil using Karickhoffs

equation [1] (using diuron sorption by K+- and Mg2+-saturated peat to determine the Kom

term), average f, values for K3 and Mg2+-saturated sorbents and our alternative equation

[2].

 

 

 

 

 

 

 

 

 

 

Equation Concentration 1'. values 1'. values Average 1'.
(mLL'l) (1C -saturated) (Mi-saturated) values

Karickhoff 5 -0.06 (-0.05) -0.46 (-0.04) -0.26 (-0.05)
(peat) 10 -0.20 (-0.03) 06400.04) -0.42 (-0.04)
15 -0.30 (-0.04) -0.77 (-0.06) -0.53 (-0.05)
20 -0.38 (-0.06) -0.86 (-0.08) -0.62 (-0.07)

Alternative 5 0.47 (0.07) 0.47 (0.07) 0.47 (0.07)
10 0.40 (0.05) 0.40 (0.05) 0.40 (0.05)

15 0.34 (0.04) 0.34 (0.04) 0.34 (0.04)

20 0.28 (0.05) 0.28 (0.05) 0.28 (0.05)

 

 

 

 

 

63

 

smectite clays, such as diuron (log K0w 2.68) (Sheng et al., 2001; Nkedi-Kizza et al.,
1983). Our results indicate that 29% to 45% (i.e., ca. 37%) of clay surfaces are obscured
from p—NCB sorption by SOM in Webster A soil. The estimate for the clay basal surface
coverage by the 3.87% SOM in the Webster A soil, which included the swelling clay
interlayer surface area and hence incorporates the SOM coverage of interlayer basal
surfaces, was 38%. The similarity of the surface coverage estimate (38%) to the
complement of the average fa value suggests that SOM covers the basal surfaces of the
soil clays. We know that p-NCB and other NACs (Sheng et al., 2001, 2002; Li et al.,
2003), as well as organic cations such as hexadecyltrimethylammonium (Jaynes and
Boyd, 1991a), tetramethylammonium (Lee et al., 1989, 1990), and
trimethylphenylammonium (Jaynes and Boyd, 1991b), intercalate between clay layers
when they sorb to smectites so SOM may be occupying those interlayer sites.
Obscuration of clay surfaces may also result from aggregate formation facilitated by
SOM through the bridging of clay particles or reduction of repulsive forces between clay
particles (Stevenson, 1994; Walker and Crawford, 1968). When diuron sorption data
were analyzed using equation [2], the fa values ranged from 0.28-0.47 (i.e. ca. 0.38)
indicating that, for diuron sorption, 53% to 72% of clay surfaces are obscured by SOM in

Webster A soil.

64

CONCLUSION

Our study reveals that the association of SOM with clay suppressed sorption of p-NCB
and diuron by whole soil by at least 29 to 45% and at least 53 to 72%, respectively. In
addition, the method of sorption data analysis proposed by Karickhoff (1984) for
determining clay surface availability in whole soil may not be applicable for analyzing
diuron sorption. F ate and transport models that utilize sorption parameters based on the
additive effect of reference clay and humic material sorption may deviate from reality

and may overestimate the amount of NAC sorbed.

65

REFERENCES

Boyd, S.A., G. Sheng, B.J Teppen, and CT Johnston. 2001. Mechanisms for the
adsorption of substituted nitrobenzenes by smectite clays. Environ. Sci. Technol.
35:4227-4234.

Celis, R., J. Comejo, M. Hennosin, and W. Koskinen. 1998. Sorption of atrazine and
simazine by model associations of soil colloids. Soil Sci Soc. Am. J. 62: 165-171.

Celis, R., L. Cox, M. Hermosin, and J. Comejo. 1997. Sorption of thiazafluron by iron-
and humic acid-coated montmorillonite. J. Environ. Qual. 26:472-479.

Celis, R., C. Hermosin, L. Cox, and J. Comejo. 1999. Sorption of 2,4-
dichlorophenoxyacetic acid by model particles simulating naturally occurring soil
colloids. Environ. Sci. Technol. 33:1200—1206.

Chiou. C. 2002. Partition and adsorption of organic contaminants in environmental
systems. John Wiley and Sons, New Jersey.

Cox, L., W. Koskinen, R. Celis, P. Yen, M. Hennosin, and J. Comejo. 1998. Sorption of
imidacloprid on soil clay mineral and organic components. Soil Sci. Soc. Am. J. 62: 911-
915.

Gee, G. and J. Bauder. 1986. Particle-size analysis. p. 383-411. In A. Klute (ed.) Methods
of Soil Analysis. Part I — Physical and mineralogical methods. 2nd ed. Agronomy. 9.
ASA, Madison, WI.

Haderlein, S., and R. Schwarzenbach. 1993. Adsorption of substituted nitrobenzenes and
nitrophenols to mineral surfaces. Environ. Sci. Technol. 27:316-326.

Haderlein, S.B., K. Weissmahr, and R. Schwarzenbach. 1996. Specific adsorption of
nitroaromatic: explosives and pesticides to clay minerals. Environ. Sci. Technol. 30:612-
622.

Howard, P., and W. Meylan. 1997. Handbook of physical properties of organic
chemicals. Lewis Publishers, Boca Raton, FL.

66

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

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

Johnston, C., M.F. de Oliveira, B.J. Teppen, G. Sheng, and SA. Boyd. 2001.
Spectroscopic study of nitroaromatic-smectite sorption mechanisms. Environ. Sci. and
Technol. 35:4767-4772.

Johnston, C., G. Sheng, B. Teppen, S.A. Boyd, and M.F. de Oliveira. 2002.
Spectroscopic study of dinitrophenol herbicide sorption on smectite. Environ. Sci.
Technol. 36:5067-5074.

Karickhoff, S. 1984. Organic Pollutant Sorption in aquatic systems. J. Hydraulic Eng.
110:707-735.

Kile, D., C. Chiou, H. Zhou, H. Li, and O. Xu. 1995. Partition of nonpolar organic
pollutants from water to soil and sediment organic matters. Environ. Sci. Technol.
29:1401-1406.

Kunze, G., and J. Dixon. 1986. Pretreatment for mineralogical analysis. p. 92-99. In A.
Klute (ed.) Methods of Soil Analysis. Part I — Physical and mineralogical methods. 2“".
Agronomy 9.

Laboski, C.A.M., and J .A. Lamb. 2004. Impact of manure application on soil phosphorus
sorption characteristics and subsequent water quality implications. Soil Sci. 169: 440-
448.

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

Laird, D. 2001a. Nature of clay-humic complexes in an agricultural soil: II. Scanning
electron microscopy analysis. Soil Sci. Soc. Am. J. 65:1419-1425.

67

Laird, D., P. Barak, E. Nater, and R. Dowdy. 1991. Chemistry of smectitic and illitic
phases in interstratified soil smectite. Soil Sci. Soc. Am. J. 55:1499-1504.

Laird, D., D. Martens and W. Kingery. 2001b. Nature of clay-humic complexes in an
agricultural soil: 1. Chemical, biochemical, and spectroscopic analyses. Soil Sci. Soc.
Am. J. 65:1413-1418.

Laird, D., Y. Yen, W. Koskinen, T. Steinheimer, and R. Dowdy. 1994. Sorption of
atrazine on soil clay components. Environ. Sci. Technol. 28:1054-1061.

Lee, J., M. Mortland, C. Chiou, D. Kile, and SA. Boyd. 1990. Adsorption of benzene,
toluene, and xylene by two tetramethylammonium-smectites having different charge
densities. Clays Clay Miner. 38: 113-120.

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

Li, H., G. Sheng, B. Teppen, C. Johnston, and SA. Boyd. 2003. Sorption and desorption
of pesticides by clay minerals and humic acid-clay complexes. Soil Sci. Soc. Am. J.
67:122-131.

Li, H., B.J. Teppen, D. Laird, C. Johnston, and SA. Boyd. 2004. Thermodynamics of
nitroaromatic compound adsorption from water by smectite clay. Environ. Sci. Technol.
38:5433-5442.

Liu, W., J. Gan, and S. Yates. 2002. Influence of herbicide structure, clay acidity, and
humic acid coating on acetanilide herbicide adsorption on homoionic clays. J. Agric.Food
Chem. 50:4003-4008.

Murphy, E., J. Zachara, and S. Smith. 1990. Influence of mineral-bound humic
substances on the sorption of hydrophobic organic compounds. Environ. Sci. Technol.
24:1507-1516.

Nelson, D. and L. Sommers. 1996. Total carbon, organic carbon, and organic matter. p.
961-1010. In D. Sparks (ed.) Methods of Soil Analysis. Part III — Chemical methods.
Agronomy. 5. ASA, Madison, WI.

68

Nkedi-Kizza, P, P. Rao, and J. Johnson. 1983. Adsorption of diuron and 2,4,5-T on soil
particle-size separates. J. Environ. Qual. 12:195-197.

Parker, A., and 1E. Rate. 1998. Environmental interactions of clays — clays and the
environment. Springer-Verlag, Berlin, Germany.

Pusino, A., W. Liu, and C. Gessa. 1992. Influence of organic matter and its clay
complexes on metolachlor adsorption on soil. Pestic. Sci. 36:283-286.

Pusino, A., W. Liu, and C. Gessa. 1994. Adsorption of triclopyr on soil and some of its
components. J. Agric. Food Chem. 42:1026—1029.

Rebhun, M., R. Kalabo, L. Grossman, J. Manka, and Ch. Rav-Acha. 1992. Sorption of
organics on clay and synthetic humic-clay complexes simulating aquifer processes. Wat.
Res. 26:79-84.

Sanchez-Martin, M., and M. Sanchez-Camazano. 1991. Adsorption of chloridazon by
soils and their components. Weed Sci. 39:417-422.

SAS Institute. 1999. SAS 8.1. SAS Institute, Inc., Cary, NC.

Sheng, G., C. Johnston, B. Teppen, and SA. Boyd. 2001. Potential contributions of

smectite clays and organic matter to pesticide retention in soils. J. Agri. Food Chem.
49:2899-2907.

Sheng, G., C. Johnston, B. Teppen, and SA. Boyd. 2002. Adsorption of dinitrophenol
herbicides from water by montmorillonites. Clay Clay Min. 50:25-34.

Soil Survey Staff, Natural Resources Conservation Service, United States Department of
Agriculture. Soil Series Classification Database [Online]. Available at
http://soils.usda.gov/soils/technical/classification/scfile/index.html (verified 10 Feb.
2004)

Sposito, G. 1984. The surface chemistry of soils. Oxford Univ. Press, New York, NY.

69

Stevenson, F. 1994. Humus Chemistry, Genesis, Composition, Reactions. John Wiley
and Sons, Inc., New York, NY.

Stevenson, F. and M. Cole. 1999. Cycles of Soil. Carbon, Nitrogen, Phosphorus, Sulfur,
Micronutrients. John Wiley and Sons, Inc., New York, NY.

Sumner, M. and W. Miller. 1996. Cation exchange capacity and exchange coefficients. p.
1201-1229. In D. Sparks (ed.) Methods of Soil Analysis. Part III — Chemical methods.
Agronomy. 5. ASA, Madison, WI.

Walker, A., and D. Crawford. 1968. The role of OM in adsorption of the triazine
herbicides by soil. P. 91-108. In lntemational Atomic Energy Agency. Isotopes and
radiation in SOM studies. International Atomic Energy Agency, Vienna.

Weissmahr, K., S. Haderlein, and R. Schwarzenbach. 1997. In situ spectroscopic
investigations of adsorption mechanisms of nitroaromatic compounds at clay minerals.
Environ. Sci. Technol. 31:240-247.

Weissmahr, K., M. Hildenbrand, R. Schwarzenbach, and S. Haderlein. 1999. Laboratory
and field scale evaluation of geochemical controls on groundwater transport of
nitroaromatic ammunition residues. Environ. Sci. Technol. 33:2593-2600.

Whittig, L and W. Allardice. 1987. X-ray diffraction techniques: Section 12-3.3. p. 338-
342. In A. Klute (ed.) Methods of Soil Analysis. Part I — Physical and mineralogical
methods. 2'”. Agronomy 9. ASA, Madison, WI.

70

CHAPTER 3
SORPTION OF P-NITROCYANOBENZENE TO SOILS: THE RELATIONSHIP

BETWEEN SOIL PROPERTIES AND AVAILABILITY OF CLAY SURFACES

ABSTRACT

The swelling clay and soil organic matter (SOM) contents of soils tend to correlate, yet
they are typically considered to be independent sorbent phases. The inter-association of
SOM with soil clays (particularly swelling clays) modified the efficacy of the soil clays
for the sorption of p-nitrocyanobenzene (p-NCB). Removal of SOM from the whole soils
increased sorption of p-NCB by the remaining soil fraction. The SOM thereby suppressed
sorption of p-NCB by whole soil. We quantified the suppression of p-NCB by the whole
soil as 17 to 53% (i.e. the fraction of the soil sorptive surfaces unavailable for p-NCB
sorption was between 47 and 83%). The extent of suppression of p-NCB sorption was

strongly related to the relative contents of SOM and smectite in the soils tested.

71

INTRODUCTION

Soil organic matter (SOM) and soil clays are intimately associated with each other in
whole soil, through mechanisms including cation exchange, Van der Waals interactions,
water bridging, cation bridging and ligand exchange (Parker and Rate, 1998). Using a
Webster A soil, Laird (2001) identified two types of clay-associated humic substances,
(1) a diffuse filamentous film covering basal 2:1 phyllosilicate surfaces in the medium
(0.02-0.2 um) and fine clay (< 0.02 pm) fractions and (2) discrete particles of high-
density metal-humic complexes in the coarse clay (0.2-2.0 um) fraction. Nkedi-Kizza et
al. (1983) found that SOM was primarily associated with the < 2 pm clay-sized fraction
of Webster soil. A major role played by SOM might be to neutralize repulsive forces that
might otherwise keep clay particles apart thereby facilitating cohesion of the clay

particles (Walker and Crawford, 1968).

The association of SOM with clays may influence the ability of either soil clays or SOM
to sorb nonionic organic compounds (NOCs) such as nitroaromatics (NACs). For
example, the interaction of pesticides with clay minerals may be limited by the
availability of the clay sorption sites within soil. Our previous study (Chapter 2) revealed
that availability of clay surfaces for NAC sorption by Webster soil was limited by the
association of SOM with soil clays. Despite the long-standing notion that SOM obscures
some of the clay surface area in soils, there is a paucity of evidence that quantitates this

effect. Karickhoff (1984) proposed an equation to account for the individual contributions

72

of SOM and soil clays to NOC sorption, and for reduced availability of clay surfaces for

sorption due to coverage by SOM:

Qwhole soil = (fa "' Kmin * Ce * fmin) + (Kom * Ce * fom) [1]

where Qwho.c so” is the total solute mass sorbed per unit mass of whole soil (mg g"), CC is
the aqueous solute equilibrium concentration, and Km,“ and Kom represent the sorption
coefficients of the solute to soil fraction from which SOM was removed (mineral)
fraction and SOM, respectively, fmm and f0an are the fractional mineral and SOM contents
of soil, and f,l represents the fractional availability of sorption sites on the clay (i.e. the
fraction of sorptive surfaces available for NOC sorption in whole soil). An inherent
assumption in this equation is that the sorptive capacity of the SOM is unaffected by the
presence of mineral matter. An alternative method of determining f, as previously
outlined involved measuring the difference in NAC sorption between sorbent systems
whose cation exchange capacity (CEC) is saturated with different cations. This method is
based on the premise that NAC sorption by clay is strongly dependent on the cation
saturating the clay exchange sites (Li et al., 2003; Boyd et al., 2001; Sheng et al., 2001,
2002; Johnston et al., 2001, 2002; Weissmahr et al., 1997; Haderlein et al., 1996)
whereas NAC sorption by SOM is unaffected by the nature of the saturating cation
(Chapter 1). Both the Karickhoff approach and our alternative method revealed similar
results, i.e., ~ 30-45% of p-NCB sorptive surfaces in a Webster A soil were rendered
unavailable in the presence of SOM (Chapter 2). The objective of this study was to
determine the availability of mineral sorptive surfaces for NAC sorption among a larger

group of soils, and to relate availability to soil properties.

73

EXPERIMENTAL SECTION

Both the A and B horizons of two soils, a Webster clay loam (fine loamy, mixed, mesic
Typic Endoaquoll) and a Clarion sandy clay loam (fine loamy, mixed, mesic Typic
Hapludoll) (Soil Survey Staff, NRCS), were air-dried and sieved to remove coarse
fragments (>2mm). The soils were fractionated by standard procedures (Kunze & Dixon,
1986). Carbonates (CO3) and SOM were sequentially removed from each soil horizon.
Carbonates were removed by stirring the soil with 0.5N sodium acetate acidified to pH 5
while heating at 80°C, followed by the removal of SOM with 30% H202 at 80°C (denoted
OM-removed soil). Homoionic soil fractions and whole (unfractioned) soils were
prepared by washing the soils with 200 ml aliquots of 0.1M solution of KCl or MgCl2
overnight, thereafter centrifuging and discarding the supernatant. This procedure was
repeated four times. The soils were then washed with 200 ml aliquots of Millipore Milli-
Q (Billerica, Massachusetts) deionized H2O, thereafter centrifuged and the supernatant
tested with AgNO3. The procedure was repeated until a negative Cl' ion test with AgN03
was obtained. The soils were then freeze-dried and stored at room temperature (23 i 1°C)
until used. Para-nitrocyanobenzene (p-NCB) was purchased from Aldrich Chemical
Company (Milwaukee, Wisconsin) with a purity of 97% and used as received. Its water
solubility (Sw) at 25°C is 1650 mg L" and its log K0w value is 1.19 (Howard and Meylan,

1997)

Sorption of p-NCB by unfractionated soils and OM-removed soils (K+- and Mg”-
saturated) were measured in triplicate using a batch equilibrium technique over a range of

initial aqueous phase concentrations (Ci) (1-42 mg L") to a known mass of soil. Initial

74

solutions of p-NCB were prepared in 0.05M KCl or MgCl2 electrolyte solutions to
correspond to the saturating cation of the study soils. Total volumes of 5 ml were pipetted
into 7.5 ml borosilicate glass vials containing various amounts of sorbent (between 0.2
and 0.4 g). The vials were rotated continuously overnight (24 hours) in the dark at room
temperature (23 i- 1°C) to ensure complete sorption; preliminary studies showed
equilibrium was reached within 24 hours. The liquid and solid phases were separated by
centrifugation at 4068g for 30 min, and the solute’s equilibrium concentration in water
(Ce) measured. The concentrations of p-NCB in the supematants were determined by a
Perkin-Elmer high performance liquid chromatograph (Perkin-Elmer, Norwalk,
Connecticut) consisting of a binary LC pump 250 with a series 200 autosampler and a uv-
visible detector set at It 254 nm. A platinum EPS C18 column from Alltech Associates
was used (Deerfield, Illinois). The mobile phase was an isocratic mixture of methanol
(HPLC grade) and Milli-Q deionized water (methanol: water ratio of 55:45) with a flow
rate of 1.0 ml min". The amount of p-NCB sorbed was calculated as the difference
between the initial aqueous concentration (C,) and the final aqueous concentration (Ce).
Compound recoveries in control vials without sorbent were above 95% and not used to

adjust the data.

The extent of sorption of p-NCB by each soil was described by fitting the data to the
Freundlich equation, Q = KfCe", where Q is the sorbed concentration of p-NCB per unit
mass of soil (mg g"), Kr is the Freundlich constant, and n is an exponent related to
isotherm shape (Chiou, 2002). The n and Kf sorption parameters were used to determine

the sorption per gram of soil or clay at specific equilibrium concentrations within the data

75

range (i.e., @ 5, 10, 15 and 20 mg L"). The SOM content of the unfractionated and OM-
removed soils was determined by the Michigan State University (MSU) Soil Testing
Laboratory using dry combustion and a Leco carbon analyzer (St. Joseph, Michigan)
(Nelson and Sommers, 1996). Soil cation exchange capacity (CBC) and gravimetric clay
content was determined by the MSU Soil Testing Laboratory using the ammonium
acetate method (Sumner and Miller, 1996) and the hydrometer method (Gee and Bauder,
1986), respectively. Soil pH was determined using a pH meter and 1:1 water to soil ratio

(Sparks et al., 1996).

X-ray diffraction (XRD) was used to identify the clay mineralogy of the clay-sized
fractions (Whittig & Allardice, 1987). Oriented samples of Mg2+- and K+-saturated clay-
sized particles were prepared by dropping suspensions of the clay-sized materials,
isolated from the OM-removed soils on glass slides, air-drying, heating at 500°C, then
treatment with ethylene glycol. Samples were analyzed using CuKa radiation and a
Philips APD 3720 X-ray diffractometer (Philips Electronic Instruments, NJ). Semi-
quantitative estimates for the amount of smectite in each unfractionated soil were
determined assuming an average CEC for a soil smectite and SOM of 80 and 160 cmol

kg", respectively, as described in Chapter 1.

Data Analysis

76

Calculation of f3 values

The chief goal of this work was to estimate the availability (fa) of smectite surfaces in the
study soils for p—NCB sorption. Two independent equations were used for the calculation
of f... First, a modified equation from Karickhoff (1984) [1]:

Qwhole soil = (fa "' Kmin * Ce * fmin) + (Kom * Cc * fem) [1]
where Qwhole 50,-. is the total solute mass sorbed per unit mass of soil (mg g"), Ce is the
solute aqueous equilibrium concentration, and Km,“ and K0m represent the sorption
coefficients of the solute by soil fraction from which SOM was removed (mineral) and
SOM, respectively, fmin and fom are the fractional mineral and SOM contents of soil,
respectively, and f, represents the fractional availability of sorption sites on the clay (i.e.
the fraction of sorptive surfaces available for NOC sorption in whole soil).

Second, an alternative equation [2] developed by us in Chapter 2:

Qwhoie soil, K+ - Qwhole soil, Mg2+ = {fa * {(Kmin. K+ "‘ Ce * fmin)‘ (Kmin, Mg2+ sat * Ce * fmin)}}[2]
where Qwholc so“, If and Qwholc 50,1,Mg2+ is the total solute mass sorbed per unit mass of soil
(mg g") whose CEC is saturated with either K+ and Mg2+, respectively, Cc is the solute
aqueous equilibrium concentration, fa represents the fractional availability of sorption
sites on the clay, fmin is the fractional mineral content of the soil, and Kmin, If and Kmin,
Mg2+ represent the partition coefficients normalized to soil mineral fi'action saturated with

either K+ and Mg”, respectively.
Within both the Karickhoff [1] and alternative [2] equations, the extent of p-NCB

sorption by the residual SOM associated with the OM-removed soil must be accounted

for. However, in equation [2], the sorptive contribution of K+-saturated residual SOM

77

cancels out that of Mg2+-saturated residual SOM, thus obviating the need to estimate K0m
(Chapter 2). Sorption data were fitted to the F reundlich equation with r2 > 0.9 in all cases.
Using the Kf and n values obtained, the Q value (amount p-NCB sorbed per mass of soil)
was calculated for four concentrations that were within the data range (@ 5, 10, 15, and
20 mg L"). Using either equation [1] or [2], four fa values were obtained for each soil
(i.e., one for each concentration). Standard deviations around each f, value were
calculated using error propagation procedures. For the Karickhoff equation, only the K‘-

saturated system was used.

Estimation of degree of OM mineral surface coverage

The hypothetical percentage of the mineral surface covered by a specified amount of
SOM present was estimated using Lagaly’s formula area per C of 5.67 A2 (Lagaly, 1981),
a percent carbon in SOM of 62% (Stevenson and Cole, 1999), an average smectite

2 g" (Sposito, 1984), the estimated percent smectite in

surface area of 750 m
unfractionated soil, and the measured percent SOM in each unfractionated and OM-
removed soil as described in Chapter 2. The operating assumption underlying this

estimate was that all C from SOM resides on smectite surfaces.

Sensitivity Analysis and Validation of fa values

Validation of the fa values was determined by manipulating equation [1] to give the

equation: fa = _ [(Kom/Kmin) * (fom/fmin)] + (K/Kmin*fmin) [3]

78

where plotting fa versus fem/fa“n for each soil should give an intercept of 1 when the SOM
content is zero (i.e., 100% availability of sorptive surfaces). Linear regression equations
describing the relationship between f, values and the ratio of SOM to smectite content
(fom/fsmwitc‘ or %OM/%smectite), to SOM content and to smectite content in
unfractionated soils were obtained. If the calculated f. values are reasonable, the intercept
of the regression equations should tend to unity (i.e., no mineral blockage by SOM) when

SOM content tends to zero.

Statistical Analysis

The sorption coefficient (Kf) and the constant related to the sorption isotherm curvature
(n) were estimated using the Freundlich equation, Q=K¢Cc". Comparisons of sorption
isotherms were made by comparing the Kr and n values for each soil-solute system using
nonlinear regression in SAS (SAS Institute, 1999). A significant difference between
isotherms was assigned when there was a significant difference between K; and n or Kr

values of compared isotherms (or = 0.05) (Laboski and Lamb, 2004).

79

RESULTS AND DISCUSSION

Nitroaromatics are known to be sorbed significantly by smectite clays (Haderlein et al.,
1996; Weissmahr et al., 1997; Boyd et al., 2001; Sheng et al., 2002; Johnston et al., 2001,
2002). The Webster and Clarion soils contain significant amounts of smectite clays (13-
24% smectite) (Table 3.1, Figure 3.1) and were therefore chosen as the study soils. The
extent of NAC sorption by smectites is cation-dependent. Reference clays exchanged
with cations of relatively lower hydration energies (e.g., K+) sorb NACS to a greater
extent than those exchanged with cations of relatively higher hydration energies (e.g.,
Mg2+) (Boyd et al., 2001; Johnston et al., 2001, 2002; Sheng et al., 2002; Haderlein and
Schwarzenbach, 1993; Weissmahr et al., 1997). We found that K+-exchanged soils
sorbed more NACs than Mg2+-exchanged soils (Chapters 1 and 2). In this study we also
found that the K+-saturated unfractionated and OM-removed soils sorbed more p-NCB
compared to Mg2+—saturated soils (Table 3.2, Figure 3.2). This can be attributed to the
lower hydration energy of the K+ ion which facilitates greater interaction between the
saturating cation and the solute and larger sorption domains on clay interlayer surfaces
(Li et al., 2003; Sheng et al., 2002; Haderlein et al., 1996; Haderlein and Schwarzenbach,

1993; Weissmahr et al., 1997, 1999; Boyd et al., 2001; Johnston et al., 2001, 2002).

For each soil, with K+ as the saturating cation, OM-removed soil sorbed more p-NCB
than unfractionated soil (Figure 3.2). Sorption of p-NCB by the Webster B and Clarion A
horizons doubled after SOM was removed, while sorption increased by about 30% when

SOM was removed from the Webster A and Clarion B horizons (Figure 3.2). These data

80

98“ 08v mount—a 8388a Ewan Bantam”
5.820 .0 ”BE—o9. .M 65: SEE“ «£8 .52 ”838:6 .m

 

 

 

 

 

 

 

 

 

81

_ - - whod CmU
:32 N30
2 oidzw : R and 26 >28 50
h - - owmd O<U
Emo— .33
av O.M.§.m m _ mm 80:6. 0 _ .0 35m <40
2 - - am.— On?
~ m 00:93 3 _ m co _ .m cos sac— 520 95
N - - mmmd 93$
<3
wm 09.52% em 3 oh: ”no :32 .920
«owfiogo cocoa...“
«out: Lao—295:. can:
>29 .x. emu—U 8392:er Ana—o .x. 26.x. In 255,—. :8

 

 

 

 

 

 

 

 

 

.3589720 m cows—U .OmU mm cots—U .mAU 60582-20 <

856 .05 m... SEC .16 62682-20 m c2363 .83 mm c2363 .95 66682-20 < .2363 .055 E sees? .35

@9688 we? 20m 5023 88% £8 can :8 2085 £8 ”SEED Ea L830? mo 85:38:30 Homfiosooommzsm ._.m 033.

 

 

L I
_

 

Intensity

l "[1 1"

 

 

z.
“—2
<—.o

 

 

 

 

_ a 1 ll
* 2 5‘ WW
d - wwamwgwthJ‘u-f
0 5 10 15 o 25 30

Angie 2 their:

Figure 3.1. X-ray diffraction spectra for Mg2+-saturated clay-sized fraction isolated by
Webster and Clarion (A and B horizons) soils.
Graphs represented are (a) Clarion B, (b) Clarion A, (c) Webster B, and (d) Webster A.

s-smectite; m/i, mica and/or illite; k, kaolinite; q, quartz.

82

Table 3.2. Freundlich equation Kr and n values for p-nitrocyanobenzene sorption by K“-

and Mg2+-exchanged Webster and Clarion (A and B horizons) whole soils and soils from

which SOM was removed.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soil K; 11
Webster A K+-whole 0.0528 0.5983
KI-OM removed 0.0922 0.5012
Mgziwhole 0.0067 0.6219
Mg2+-OM removed 0.0007 0.9033
Webster B KI-whole 0.0634 0.5587
K‘C-OM removed 0.1452 0.5106
Mg2+-whole 0.0018 0.8151
Mgz‘zorw removed 0.0009 0.9736
Clarion A K+-whole 0.0222 0.6427
K+-OM removed 0.0736 0.4821
Mg2+-whole 0.0031 0.6434
Mg2*-0M removed 0.0010 0.7079
Clarion B KI-whole 0.0912 0.5326
K+-OM removed 0.1280 0.4661
Mg2+-whole 0.0007 0.9304
Mgz‘t-OM removed 0.0009 0.8185

 

83

 

0 0°
33281.3

.0
9

Concentration of NCB sorbed (mg g"

 

 

 

 

 

 

 

    

 

 

 

 

 

 

7" 0135
E OLA-K
‘3’ 430‘ CAO-K
cu-m
0.25-
3 «on;
3 0120- b
2
3 0.15
5 01101
E 0.06
O _ %c’
§0oo +' e "".
0 o 10 20 so
mammoth

Concentration of NCB sorbed (mg 9“

Concentration of NCB sorbed (mg g")

 

p
a

8

P
5

5

3

 

p
.5

 

 

 

 

0.7
0.6 .
0.5 r
0.4 1
0.3 -

0.2 -'

 

0.1

0.0

400.

    

CLB-K
CBC-K a

 

10
Aqueous NCB concentration (mg L")

 

Figure 3.2. Adsorption isotherms for sorption of p-nitrocyanobenzene by K+- and Mg”-

saturated Webster and Clarion (A and B horizon) soils.

(isotherms marked with different letters are significantly different at p < 0.05)

WA/WB — Webster A/ Webster B horizon

CLA/CLB - Clarion A/ Clarion B horizon

WAO/WBO — Webster A/B horizon, OM removed

CLA/CLB — Clarion A/B horizon, OM removed

84

indicate that SOM suppresses sorption of p-NCB by the soil mineral fraction despite
SOM also acting simultaneously as an independent sorptive sink. Literature reports
decreased sorption of organic solutes to reference minerals when associated with SOM.
Liu et al. (2002) observed the sorption of four herbicides (metolachlor, acetochlor,‘
alachlor, and propachlor) by Wyoming montmorillonite (SWy-Z), humic acid (HA)
extracted from a Webster clay loam, and clay-HA complexes containing various ratios of
clay and SOM. They found that the measured sorption of the herbicides by the clay-HA
complexes were less than the sum of sorption by clay and by HA in the same ratios as the
complexes. Celis et al. (1997) found that coating Ca-montomorillonite (Ca-SWy-l) with
commercial HA and soil-extracted HA decreased the sorption of thiazafluron by the clay-
HA complex compared to clay alone. Subsequent removal of the HA coating on the clay-
HA complex increased thiazafluron sorption compared to the clay-HA complex. Pusino
et al. (1992; 1994) compared the experimental sorption coefficients (Kd values) obtained
for the sorption of metolachlor (1992) and triclopyr (1994) by bentonite montmorillonite-
HA complexes with the K, values calculated assuming independent sorption of
metolachlor by HA and clay. The experimental Kd values were lower than the calculated
Kd values. Although these studies indicated that the presence of SOM associated with
clays results in decreased sorption of NOCs by the clays, these studies utilized synthetic
clay-SOM complexes. Our data indicates that in natural soils, removal of SOM from soil
increases sorption of NOCs indicating that in natural soils, as in synthetic systems, SOM

obscures clay surfaces for sorption of NOCs.

85

The phenomenon of reduced sorption in the presence of SOM is exemplified by f. values
less than 1. We used our sorption data for the K+-saturated soils and applied the
Karickhoff equation [1] to calculate fa values (i.e. the fraction of the sorptive surfaces that
is available in whole soil) for the four soil horizons. We used K+- and Mgz+-saturated
peat, Houghton muck and Brookston loam to determine the SOM contribution to overall
solute sorption. This approach requires an estimate of K0'“ for sorption of p-NCB, which
we derived previously using three SOM surrogates (Chapter 2). The similarity in fa
values calculated using K+- and Mg2+-saturated SOM surrogates (Table 3.3, Table 3.4)
indicates that the influence of cation on p—NCB sorption by SOM is minimal. We also
applied our alternative equation [2] to estimate fa; this approach utilizes sorption data
from K+- and Mg2+-saturated soils and eliminates the need for an independently estimated
Kmm values. These fa values (Table 3.3, Table 3.4) obtained for a particular soil using
equation [1] did not differ significantly from those obtained using equation [2]. That is,
two independent methods for estimating fa (equations [1] and [2]) yielded similar results.
Also, the fa values were relatively independent of sorbent material (muck, peat or loam)
used as a surrogate for pure SOM (Table 3.3, Table 3.4) when using equation [1].
However, fa values varied across soils and soil horizons (i.e., f, values for Clarion A <
Webster B < Wester A < Clarion B) (p < 0.05). The fa values, averaged across both
equations and all K+-saturated SOM surrogates, for Clarion B (ca. 0.83 i 0.03) were
greater than for Clarion A (ca. 0.36 i 0.05) and those for Webster A (ca. 0.64 i 0.04)
were higher than for Webster B (ca. 0.47 i 0.01) (p < 0.05). Despite various amounts of

SOM initially present in the soils and different amounts being removed from the soils, f,

86

88382.8 5 8283.8 28:85 .32. v 3 m cots—U v < 88.83 v m .8303 v < 85.0 8.8 828,
a 483980 some 8... .39: v 3 8:888:00 9888 288.886 £85388 8: 083 mos—9 u .m 8683 8 m 836 v 3 ASA—mom v 8.32 vege—
v 8&3 8388880 8888 28898 3.828886 083 $3.? .8 .m 8683 838 28m :8 88m 836 v 3 0288.882 w 21.885282
v amucoiotaw v anon—$05883. 9.6: 888—528 mos—8, .8 88388380 :8 8088 .< cots—U 28 < .8303 8m .God v 8 038:8“?
namfiofiovaz n Emfioiotax u 809.coiom8v— win: 888328 .828.» u 88388880 :8 8888 .m 85.0 28 m .8303 .8...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

88.8 S8 :88 :8 A88 8.8 28.8 888 om

A88 38 98.8 98 A88 88 28.8 888 2

A88 88 28.8 888 A88 88 A88 38 8

$88 888 28.8 38 88.8 3.8 38.8 68 8 388537..
88.8 88 A88 88 88.8 8.8 38.8 888 cm

88.8 £8 :88 £8 :88 :8 98.8 $8 2

A88 88 28.8 .88 £88 88 A88 88 8 8.8. 88.828
88.8 888 28.8 88.8 38.8 88 A88 88 m 82:35.
8.8 58 98.8 88 @888 8.8 98.8 :8 8

A88 38 :88 8.8 38.8 8.8 88.8 $8 2

38.8 $8 :88 R8 888 :8 A88 88 8 22:: 88.88.58
88.8 $8 :88 28 A88 28 A88 888 m 8233.5.
888 88.8 :88 R8 88 :8 98.8 88 cm

48.8 38 28.8 28 A88 :8 ~88 $8 2 A 8
. . . . . . . . .3 8
A8 8 N88 :8 8mg 88 838 :8 8888 8
88.8 $8 :88 R8 388 3.8 38.8 38 m 82:88.
A mud—5 .88: Bafieta.

m .8220 < .8220 m 8.393 < .8883 flirt—89:80 :5 can nouns—.m—

 

 

 

AE88 sex 2: 088.888

8 :82 coamxooa 8 £02: :oEwnoE Java wBfiBmmVM .3 8:98 828 magma H: 8633 mrcofiowavm mam: 38:53

@888: m can <8 8.530 98 @588: m 98 <8 .8683 .3 8:98 AmUZv 058305880588 8.« mos—8, _w 2.: .m.m 238,—.

87

80858.8 5 28.838 8.88% .Amod v 3 m 885 V < .2803 v m .8283 v < 885 8.8 82.?
d .8880 88 8.... .396 V 3 88.38080 $08.... .88.:8 biosmufim 8.. 083 828, d .m 8.803 8 m .Amcd v 3 ASE—ow v Enam— v.82:
v 883 88.8888 808.. 89.85% £88.83... 083 82.3 d .m .3803 8088 £88 :8 .8..— .God v 3 038E028. w Encoiotav.
v amutofiotnv— v “meantofiomug. mama: 88.8.8 .83? d 88388080 :8 8088 :8. 8.8.0 88 < .8803 8m 83.0 v 3 02.8.8.2
"amntofiowav. n 23E0508M u .80....toiot8v. mama: 88—8—8 mos—g d 88.888080 :8 888.. .m 88—0 .28 m .8883 8...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.88 S8 :88 3.8 38.8 8.8 .88 28 cm

.88 £8 :88 8.8 .88 88 .88 .88 2

.88 88 28.8 8.8 .88 8.8 88.8 $8 8.

.88 8.8 98.8 28 88.8 2.8 .88 .88 m 3.3.5.?

.88 S8 :88 $8 .88 88 .88 888 cm

.88 28 28.8 8.8 .88 88 8.8 .88 2 .58.

.88 S8 :88 R8 .88 $8 .88 2.8 8. 58.8....

88.8 888 488 N8 .88 2.8 88.8 .88 m 82.8.8.5.

48.8 $8 :88 8.8 48.8 8.8 .88 :8 cm

.88 28 28.8 8.8 .88 8.8 .88 888 2 22::

.88 88 28.8 888 .88 $8 .88 88 8. 88.8888...

88.8 888 28.8 .88 .88 2.8 88.8 88 m 88.8.2.5.

.88 8.8 28.8 8.8 .88 8.8 .88 .88 cm

Q8 28 28.8 888 .88 5.8 98.8 88 2 . «2..

.88 $8 :88 28 .88 2.8 .88 $8 8. .

88.8 888 4.8.8 .88 .88 2.8 .88 88 w 885.85.
A -1— ”Ev 60m:

m 8.8.0 A. acts—U m .8303 < 8.883 8:"58280 888.83 55
. E... 8:85am

 

 

 

8:8. sov— ofi 0588.8

8 E8 8% 08m 8 n 2.8 :9 mac... .88 88888- $2 .8 8.8.8 8: cm wag: . 8.880 m o 0.8M mafia 888.80
_ x v. A .8 +~ a . _ . . rt .3 . . w .

A2888: m .88 <3 8.8.0 88 A2888: m 88 <3 8.283 3 8.8.8 307C 8880888058-.. .8 82.3 8.. on... .vd 038...

88

relates to the available sites in the context of whole soil. We accounted for the sorption of
p-NCB by residual SOM (i.e., SOM remaining after removal of bulk SOM) in each OM-
removed soil. Thus, by our analysis, an average of 36 to 83% of the sorptive surfaces are
‘available’ for p-NCB sorption in the unfractionated soil horizons, with more clay surface

becoming available for sorption afier SOM is removed.

Data analysis revealed that fa values are strongly negatively correlated to ratios of SOM
and smectite content of soil. From 83-88% of the difference in f, values among soil
horizons could be attributed to differences in ratios of SOM and smectite in the soils
(Table 3.5). This is reasonable, given that the SOM and clay are the active sorbents of p-
NCB and that, of the minerals present in each soil, smectites have the highest affinities
for p-NCB (Chapter 1). The availability of sorptive clay sites does not correlate
particularly well with either SOM content, or smectite content alone (Table 3.5). The fa
values and %OM were moderately, negatively correlated with 40-50% of the difference
in fal values among soil horizons attributed to differences in SOM content. This suggests
that the availability of sorptive surfaces in whole soil decreases with an increase in SOM
content. This may only operate to a certain extent as Murphy et al. (1990) and Rebhun et
al. (1992) suggested that at high concentrations of HA, multiple layering of HA on the
mineral surface occurs. The fa values and %smectite were weakly correlated with only
about 20% of the difference in f8| values attributed to differences in smectite content in the
soil horizons (Table 3.5). The positive correlation suggests that increasing the smectite

content of a soil should increase the overall availability of clay mineral sorptive surfaces

89

Table 3.5. Correlation (r value) between fa values for sorption of p-nitrocyanobenzene
estimated using Karickhoff equation [1] (using solute sorption by K+-saturated peat,
Houghton muck or Brookston loam to determine the Km term) and the alternative
equation [2] with percent organic matter (%OM), percent smectite (%smectite) and ratio
of organic matter to smectite (%OM/%smectite). Correlations were performed across the

Webster and Clarion (A and B horizons) soils (a = 0.05).

 

 

 

 

 

Equation used and %OM %smectite %OM/%smectite
SOM surmate
Karickhoff —J)eat ~0.51 0.17 -0.88
Karickhoff — Houghton -O.42 0.24 -O.82
muck
Karickhoff — Brookston loam -O.48 0.20 -O.87
Alternative -0.45 0.21 -0.83

 

 

 

 

 

 

9O

for a given SOM content. This is because the amount of surface that a given amount of
SOM can obscure theoretically would decrease with increasing smectite content. Plotting
fa values against %OM/%smectite indicates that the fa values calculated are reasonable
since, for each equation and each SOM-surrogate used, the intercepts are approximately
equal and all approach unity at 0% SOM (Table 3.6; Figure 3.3). That is, in the absence
of SOM the smectite surfaces should be fully available for sorption of p-NCB. The near-
unity intercepts also suggest that other types of coatings on smectites that might be

envisioned do not significantly reduce availability.

While this research clearly indicates that natural SOM renders soil clay surfaces
unavailable for sorption of organic solutes, one can only postulate possible mechanisms.
First, SOM could cover soil clay basal surfaces and render them unavailable. We know
that p-NCB and other NACs (Sheng et al., 2001, 2002; Li et al., 2003), as well as organic
cations such as hexadecyltrimethylammonium (Jaynes and Boyd, 1991a),
tetramethylammonium (Lee et al., 1989, 1990), and trimethylphenylammonium (Jaynes
and Boyd, 1991b), intercalate between the clay layers when sorbed by smectites so SOM
may be occupying those interlayer sites. Stevenson ( 1994) points out that although lab
studies have reported organic compound sorption in the interlarnellar surfaces of clays, it
has not been unequivocally established that such complexes occur in natural soils.
However, recovery of amino acids from clay-rich subsoils by acid hydrolysis is
incomplete unless preceded by HF pretreatment (Stevenson, 1994) suggesting that small
molecules from SOM are able to intercalate clays. If all carbon (C) in SOM were

assumed to cover the basal surfaces of soil smectites, then the formula for surface

91

Table 3.6. Linear regression describing the relationship between fa values and ratio of

SOM and smectite soil contents.
Regressions were performed across the Webster and Clarion (A and B horizons) soils.

Intercept is the fa value at which %OM is zero.

 

 

 

 

 

 

 

 

 

 

 

Equation used Slope 12:33:; [£21.12th 1'2
Karickhoff (peat) -2.73 (0.40) 0.92 (0.06) 0.86-0.98 0.77
Karickhoff (Houghton -2.43 (0.45) 0.91 (0.07) 0.84-0.98 0.68
muck)
Karickhoff (Brookston -2.63 (0.41) 0.92 (0.06) 0.86-0.98 0.75
loam)
Alternative -2.36 (0.43) 0.91 (0.06) 0.85-0.97 0.69

 

Standard deviations in parentheses.

92

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

0.8
0.7
0.6 i
d“
0.5
0.4
0.3 .
0.2 V I I I 0.2 I T fl 1
0.05 0.10 0.15 0.20 (105 010 0.15 OZ)
%OM/ %srrectite (yaw/W
0 9 09
3 0 Marnie
0.8 ={fit-+0.91
03 . . Y
3413
0.7
07‘
0.6 . I09
.9 as
0.5 1
0.4 ~ 05‘
0.3 . Q4 .
o
0.2 I I r T 03 I ' I I
0.05 0.10 0.15 0.20 Gm 010 0.15 02)
%OM/ %amcfite mm

Figure 3.3. Linear regressions of f8 values obtained from data analysis of p-
nitrocyanobenzene (p-NCB) sorption data using Karickhoff equation [I] and our
alternative equation [2] across 4 specified concentrations (5, 10, 15, 20mg L") versus
ratio of SOM to smectite content (%OM/%smectite) across 4 soils (Webster A, Webster
B, Clarion A, Clarion B). Sorption of p-NCB to K+-peat, Houghton muck, or Brookston

loam was used to obtain the Kom term in Karickhoff equation [1].

93

coverage used in the alkylammonium method (Lagaly, 1981) predicts that this C would
cover 49, 38, 31 and 10% of the smectite surface area in the Clarion A, Webster A,
Webster B, and Clarion B horizons, respectively. Note that these values are in modest
agreement with the complements of the average fa values (i.e., 1- f,; the fraction of clay
surface obscured by SOM) for each horizon (Table 3.3) which are 0.64, 0.36, 0.53, and
0.17, respectively. This agreement may be stricly fortuitous, but would provide a good
explanation for the strong correlation between f, and the %OM / %smectite ratios (Table
3.5). However, some authors (Ahlrich, 1972; Celis et al., 1997, 1999) argue that
intercalation of natural SOM (at least large SOM molecules) does not occur, in
opposition to the aforementioned suggestions. Celis et al. (1998) measured the specific
surface area (SSA) (N2 adsorption) of a reference smectite before and after the sorption of
humic substances by the smectite. They found that the SSA decreased indicating that the
external surfaces were obscured by the humic acid coating. Their X-ray diffraction dam
spacings did not change with humic acid coating suggesting that intercalation of humic
acids did not occur. Celis et al. (1997), Celis et al. (1999), and Li et al. (2003) found

similar results.

Alternatively, SOM may bind to edge sites of clay minerals and thereby facilitate
aggregate formation, and/or inhibit shrinking and swelling of the smectite interlayers or
may partially block the entrance of sorbates into the interlayer region. This is quite
reasonable, as edge sites would typically carry positive charges and readily form
complexes with carboxylate groups of SOM, thereby creating bridges between clay

particles. We propose that not all of the smectite surfaces are obscured by SOM bridges,

94

coatings or aggregation because (a) cation exchange appears to happen readily, which
implies that interlayers are accessible to soil solution, and (b) strong sorption of p-NCB
occurs even before removal of SOM, and we are confident that most of this sorption
occurs on smectite basal surfaces. Further studies would be needed to elucidate the

mechanisms behind the obscuration of sorptive surfaces in soil smectites.

95

CONCLUSION

Individual effects of SOM and clay are not easily ascertained because in soils, SOM is
intimately associated with clay. Accordingly, clay and SOM in soils undoubtedly
function in concert more than as separate entities. Our study reveals that the association
of SOM with clay obscures sorptive surfaces of soil clays, with availability being
dependent on the relative contents of SOM and sorptive clays (e.g., smectites) in the

whole soil.

96

REFERENCES

Ahlrich, J. 1972. The soil environment. p. 3-43. In C. Goring and J. Hamaker (ed.)
Organic chemicals in the soil environment. Vol. 1. Marcel Dekker, New York.

Boyd, S.A., G. Sheng, B.J Teppen, and CT Johnston. 2001. Mechanisms for the
adsorption of substituted nitrobenzenes by smectite clays. Environ. Sci. Technol.
35:4227-4234.

Celis, R., J. Comejo, M. Hermosin, and W. Koskinen. 1998. Sorption of atrazine and
simazine by model associations of soil colloids. Soil Sci Soc. Am. J. 62:165-171.

Celis, R., L. Cox, M. Hermosin, and J. Comejo. 1997. Sorption of thiazafluron by iron-
and humic acid-coated montmorillonite. J. Environ. Qual. 26:472-479.

Celis, R., C. Hermosin, L. Cox, and J. Comejo. 1999. Sorption of 2,4-
dichlorophenoxyacetic acid by model particles simulating naturally occurring soil
colloids. Environ. Sci. Technol. 33: 1200-1206.

Chiou. C. 2002. Partition and adsorption of organic contaminants in environmental
systems. John Wiley and Sons, New Jersey.

Gee, G. and J. Bauder. 1986. Particle-size analysis. p. 383-411. In A. Klute (ed.) Methods
of Soil Analysis. Part I — Physical and mineralogical methods. 2nd ed. Agronomy. 9.
ASA, Madison, WI.

Haderlein, S., and R. Schwarzenbach. 1993. Adsorption of substituted nitrobenzenes and
nitrophenols to mineral surfaces. Environ. Sci. Technol. 27:316-326.

Haderlein, S.B., K. Weissmahr, and R. Schwarzenbach. 1996. Specific adsorption of
nitroaromatic: explosives and pesticides to clay minerals. Environ. Sci. Technol. 30:612-
622.

Howard, P. and W. Meylan. 1997. Handbook of physical properties of organic chemicals.
Lewis Publishers, New York, NY.

97

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

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

Johnston, C., M.F. de Oliveira, B.J. Teppen, G. Sheng, and SA. Boyd. 2001.
Spectroscopic study of nitroaromatic-smectite sorption mechanisms. Environ. Sci.
Technol. 35:4767-4772.

Johnston, C., G. Sheng, B.J. Teppen, S.A. Boyd, and M. F. de Oliveira. 2002.
Spectroscopic study of dinitrophenol herbicide sorption on smectite. Environ. Sci.
Technol. 36:5067-5074.

Karickhoff, S. 1984. Organic Pollutant Sorption in aquatic systems. J. Hydraulic Eng.
110:707-735.

Kunze, G., and J. Dixon. 1987. Pretreatment for mineralogical analysis. p. 92-99. In A.
Klute (ed.) Methods of Soil Analysis. Part I — Physical and mineralogical methods. 2nd
ed. Agronomy. 9. ASA, Madison, WI.

Laboski, C.A.M., and J .A. Lamb. 2004. Impact of manure application on soil phosphorus
sorption characteristics and subsequent water quality implications. Soil Sci. 169: 440-
448.

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

Laird, D. 2001. Nature of clay-humic complexes in an agricultural soil: II scanning
electron microscopy analysis. Soil Sci. Soc. Am. J. 65:1419-1425.

Lee, J., M. Mortland, C. Chiou, D. Kile, and SA. Boyd. 1990. Adsorption of benzene,
toluene, and xylene by two tetramethylammonium-smectites having different charge
densities. Clays Clay Miner. 38: 113-120.

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

98

Li, H., G. Sheng, B. Teppen, C. Johnston, and SA. Boyd. 2003. Sorption and desorption
of pesticides by clay minerals and humic acid-clay complexes. Soil Sci. Soc. Am. J.
67:122-131.

Liu, W., J. Gan, and S. Yates. 2002. Influence of herbicide structure, clay acidity, and
humic acid coating on acetanilide herbicide adsorption on homoionic clays. J. Agric.Food
Chem. 50:4003-4008.

Murphy, E., J. Zachara, and S. Smith. 1990. Influence of mineral-bound humic
substances on the sorption of hydrophobic organic compounds. Environ. Sci. Technol.
24:1507-1516.

Nelson, D. and L. Sommers. 1996. Total carbon, organic carbon, and organic matter. p.
961-1010. In D. Sparks (ed.) Methods of Soil Analysis. Part III — Chemical methods.
Agronomy. 5. ASA, Madison, WI.

Nkedi-Kizza, P, P. Rao, and J. Johnson. 1983. Adsorption of diuron and 2,4,5-T on soil
particle-size separates. J. Environ. Qual. 12:195-197.

Parker, A., and J .E. Rate. 1998. Environmental interactions of clays — clays and the
environment. Springer-Verlag, Berlin, Germany.

Pusino, A., W. Liu, and C. Gessa. 1992. Influence of organic matter and its clay
complexes on metolachlor adsorption on soil. Pestic. Sci. 36:283-286.

Pusino, A., W. Liu, and C. Gessa. 1994. Adsorption of triclopyr on soil and some of its
components. J. Agric. Food Chem. 42:1026-1029.

Rebhun, M., R. Kalabo, L. Grossman, J. Manka, and C. Rav-Acha. 1992. Sorption of
organics on clay and synthetic humic-clay complexes simulating aquifer processes. Wat.
Res. 26:79-84.

SAS Institute. 1999. SAS 8.1. SAS Institute, Inc., Cary, NC.

99

Sheng, G., C. Johnston, B. Teppen, and SA. Boyd. 2001. Potential contributions of
smectite clays and organic matter to pesticide retention in soils. J. Agric. Food Chem.
49:2899—2907.

Sheng, G., C. Johnston, B. Teppen, and SA. Boyd. 2002. Adsorption of dinitrophenol
herbicides from water by montmorillonites. Clay Clay Min. 50:25-34.

Soil Survey Staff, Natural Resources Conservation Service, United States Department of
Agriculture. Soil Series Classification Database [Online]. Available at
http://soils.usda.gov/soils/technical/classification/scfile/index.htrnl (verified 10 Feb.
2004)

Sparks, D., A. Page, P. Helmke, R. Loeppert, P. Soltanpour, M. Tabatabai, C. Johnston,
and M. Sumner. 1996. Methods of soils analysis: Part 3 — Chemical methods. Soil
Science Society of America, Madison, WI.

Sposito, G. 1984. The surface chemistry of soils. Oxford Univ. Press, New York, NY.

Stevenson, F. 1994. Humus chemistry, genesis, composition, reactions. John Wiley and
Sons, New York, NY.

Stevenson, F. and M. Cole. 1999. Cycles of Soil. Carbon, Nitrogen, Phosphorus, Sulfur,
Micronutrients. John Wiley and Sons, New York.

Sumner, M. and W. Miller. 1996. Cation exchange capacity and exchange coefficients. p.
1201-1229. In D. Sparks (ed.) Methods of Soil Analysis. Part III — Chemical methods.
Agronomy. 5. ASA, Madison, WI.

Walker, A., and D. Crawford. 1968. The role of OM in adsorption of the triazine
herbicides by soil. P. 91-108. In International Atomic Energy Agency. Isotopes and
radiation in SOM studies. International Atomic Energy Agency, Vienna.

Weissmahr, K., S. Haderlein, and R. Schwarzenbach. 1997. In Situ Spectroscopic
investigations of adsorption mechanisms of nitroaromatic compounds at clay minerals.
Environ. Sci. Technol. 31:240-247.

100

Weissmahr, K., M. Hildenbrand, R. Schwarzenbach, and S. Haderlein. 1999. Laboratory
and field scale evaluation of geochemical controls on groundwater transport of
nitroaromatic ammunition residues. Environ. Sci. Technol. 33:2593-2600.

Whittig, L and W. Allardice. 1987. X-ray diffraction techniques: Section 12-3.3. p. 338-
342. In A. Klute (ed.) Methods of Soil Analysis. Part I — Physical and mineralogical
methods. 2nd ed. Agronomy. 9. ASA, Madison, WI.

101

CHAPTER 4
SORPTION OF NONIONIC ORGANIC COMPOUNDS TO SOILS: THE

COMPOUND SPECIFICITY OF AN AVAILABILITY INDEX

ABSTRACT

Efficacy of the clay mineral fraction of soil for the sorption of nonionic organic
compounds (NOCs) may be reduced by its association with soil organic matter (SOM).
An understanding of the influence of the clay and SOM association on the fate and
transport of NOCs would provide important information that would enhance quantitative
predictions of sorption of NOCs by soil compositions. Previously we quantitatively
assessed the fractional availability (fa) of soil clay surfaces for sorption of p-
nitrocyanobenzene (p-NCB). For p-NCB that sorbs preferably to smectite clays the
relative contents of SOM and smectite was the primary determinant of the availability of
smectite surfaces in soil. In this study we used three nitroaromatic (NAC) compounds (p-
NCB, 2,4-dinitrotoluene (DNT) and 1,4-dinitrobenzene (DNT)) and four loam soils to
further evaluate the availability of soil smectite surfaces for sorption. Sorption of each
NAC by whole soil and soil from which SOM was removed was determined. Increased
NAC sorption was observed with the removal of SOM from whole soils indicating
diminished availability of mineral surfaces due to obscuration or blockage by SOM. The
fa values differed with NAC suggesting the nature of the NAC may also influence the
mechanism driving sorption site availability. We found a strong correlation between the

strength of sorption of each NAC to smectites (and the relative sorption capacity of each

102

NAC to SOM and smectites) and the efficacy of SOM to obscure sorption sites in whole

soil.

103

INTRODUCTION

The ever increasing manufacture and use of synthetic organic chemicals has amplified the
need for the development and improvement of models to predict their fate in the
environment. Fate and transport models utilize factors representing processes that affect
the fate of these chemicals in the environment. A significant fate process for nonionic
organic compounds (NOCs) in soils is their sorption by clay minerals and soil organic
matter (SOM), the two most important components in soil for NOC sorption. Fate and
transport models for fate of NOCs in soils often include terms representing sorption of
the NOC to SOM (e.g., Kom, fom) but usually ignore the potential contribution of soil
clays to sorption. Soil clays may play an equal or dominant role, compared to SOM, in
the immobilization of certain important classes of NOCs, for example, nitroaromatic
compounds (NACS) (Sheng et al., 2001). However, the sorptive contributions of SOM
and clay components may be interrelated since these materials are often associated in

soils.

Several studies investigating the sorption of NACs by synthetic humic acid (HA)-clay
complexes have shown that the extent of sorption of NOCs to the complex does not
equate to the sum of sorption to the individual components assuming independent and
additive adsorption behavior (Pusino et al., 1992, 1994; Celis et al., 1997; Liu et al.,
2002). It is suggested that SOM may adhere to clay surfaces and form bridges between
clay packets, resulting in aggregate formation (Celis et al., 1999; Pusino et al., 1992,

1994; Onken and Traina, 1997), inhibit shrinking and swelling of the smectite interlayers,

104

and/or may partially block the entrance of sorbates into the interlayer regions (Walker

and Crawford, 1968).

We determined that the association of SOM with clays in soils modifies the efficacy of
the clay mineral fraction for the sorption of NOCs in soil (Chapter 2 and 3). Our previous
studies revealed that p-nitrocyanobenzene (p-NCB) sorption by Webster and Clarion soils
from which SOM was removed was greater than that by whole soil (Chapter 2 and 3)
suggesting that the association of SOM with clay minerals obscures sorptive surfaces. In
the case of p-NCB, 55 to 71% of the soil clay surfaces were available for sorption by a
Webster A soil whose SOM and smectite contents were ca. 4% and 24%, respectively
(Chapter 2). The efficacy of SOM to obscure sorptive surfaces (i.e., the fractional
availability (f,) of sorptive surfaces) was strongly related to the ratio of SOM to smectite
content in soil (Chapter 3). However, the effect of solute characteristics on availability of
soil clay surfaces has not been studied. Liu et al. (2002) calculated a parameter ‘a’ that
represented the relative change in adsorption capacity of HA-clay complexes, that is, the
fraction of the complex that is potentially hindered from sorption by SOM. The ‘a’ values
obtained for metolachlor, acetochlor, alachlor and propachlor were -0.73, -0.79, -0.62
and -0.53, respectively suggesting that the presence of HA on the complex differentially
decreases sorption depending on the solute used. However, their study used synthetic HA
and clay complexes so that translation to whole soil is uncertain. Our objective was to
quantify the fractional availability of mineral sorption sites within whole soils for three

solutes.

105

EXPERIMENTAL SECTION

Two mineral soils, Webster (fine loamy, mixed, mesic Typic Endoaquoll) and Clarion
(fine loamy, mixed, mesic Typic Hapludoll) (A and B horizons) (Soil Survey Staff,
NRCS), and an organic soil (Pahokee peat; a Euic, hyperthermic Lithic Medisaprist) were
air-dried and sieved to remove coarse fragments (>2mm). The mineral soils were
fractionated by standard procedures (Kunze & Dixon, 1986). Carbonates (C03) and soil
organic matter (SOM) were sequentially removed from each soil horizon. Carbonates
were removed using 0.5M sodium acetate buffer acidified to pH 5 at 80°C, followed by
the removal of SOM using 30% H202 at 80°C (denoted OM-removed soil). Homoionic
soils and soil fractions were prepared by washing the soils with approximately 200 ml
portions of 0.1M solution of KCl or MgClz overnight, thereafter centrifuging at 4068g for
30 min and discarding the supernatant. This procedure was repeated four times. The soils
were then washed with 200 ml portions of Millipore Milli-Q (Billerica, Massachusetts)
deionized H20, thereafter centrifuged and the supernatant tested with AgNO3; the
procedure was repeated until a negative Cl' ion test with AgNO; was obtained. The soils
were then freeze-dried and stored at room temperature (23 i 1°C) until used. A K‘-
saturated reference clay, a montmorillonite, beidellite obtained from Ward’s Natural
Science (Rochester, NY), was prepared as outlined above, freeze-dried and stored at
room temperature until used. Clay minerals present in the soils, both A and B horizons,
were previously identified as mica and/or illite, smectite, and kaolinite (Chapter 1). The
Michigan State University (MSU) Soil Testing Laboratory (East Lansing, Michigan)

determined the soil gravimetric clay content by the hydrometer method (Gee and Bauder,

106

1986), and the SOM content using dry combustion and a Leco carbon analyzer (St.

Joseph, Michigan) (Nelson and Sommers, 1996).

Dinitrobenzene (1,4-DNB) was purchased from Sigma Chemical Company (St. Louis,
Missouri) with a purity of 98%. Nitrocyanobenzene (p-NCB) and 2,4-dinitrotoluene (2,4-
DNT) were purchased from Aldrich Chemical Company (Milwaukee, Wisconsin) with
purities of 97% and 97%, respectively. Trinitrobenzene (1,3,5-TNB) was purchased from
ChemService Inc. (West Chestnut, PA) with a purity of 99%. All NOCs were used as

received.

Sorption of NACs by soils, soil fractions and K+-saturated beidellite was measured, in
triplicate, using a batch equilibrium technique. This involved adding aqueous solutions of
p-NCB (1 to 42 mg L'l), 1,4-DNB (1-42 mg L'l), and 2,4-DNT (1-34 mg L'l) over a
range of initial aqueous concentrations (Ci) to a known mass of soil. Solutions of the
compounds were prepared in 0.05M KCl or MgClz corresponding to the sorbent
saturating cation. Total volumes of 5 ml were pipetted into 7.5 ml borosilicate glass vials
containing between 0.2 and 0.4 g of sorbent. The vials were rotated continuously
overnight (24 hours) in the dark at room temperature (23 i 1°C) to ensure complete
sorption; preliminary studies showed that sorption equilibrium for each solute was
reached within 24 hours (data not reported). The liquid and solid phases were separated
by centrifugation at 4068g for 30 min. The concentration of NAC in the supernatants (Cc)
were determined by a Perkin-Elmer high performance liquid chromatograph (HPLC)

(Perkin-Elmer, Norwalk, Connecticut) consisting of a binary LC pump 250 with a series

107

200 autosampler and a uv-visible detector set at the following wavelengths: p-NCB, 254
nm; 1,4-DNB, and 2,4-DNT, 265 nm. A platinum EPS C18 column (Alltech, Illinois)
was eluted with an isocratic mixture of methanol (HPLC grade) and Milli-Q deionized
water (methanol: water ratio of 55:45 for p-NCB; 65:35 for 1,4-DNB, and 2,4-DNT) with
a flow rate of 1.0 ml min". The amount of solute sorbed was calculated as the difference
between the initial (C5) and final (Cc) concentration in the supernatant. Compound
recoveries in control vials without sorbent were above 95% and not used to adjust the
data. Sorption data were described using the Freundlich equation, Q = Kth", where Q is
the solute sorbed per unit mass of sorbent (mg g'l), Kf is the F reundlich constant, and n is

an exponent describing the shape of the sorption isotherm (Chiou, 2002).

Removal of organic matter with H202 may result in the liberation and oxidation of heavy
metals. The potential contribution of sesquioxide presence and/or formation afier H202
treatment to solute sorption was determined. K+-saturated ferric hydroxide, F e(OH)3, was
prepared as previously outlined (Chapter 1) using FeC13 and KOH. Sorption of p-NCB,
1,4-DNB, and 2,4-DNT by Fe(OH)3 was measured, in duplicate, using a batch

equilibrium technique as described earlier.

Data Analysis

Calculation of fractional availability (fa) of mineral surfaces

The fa values were calculated using the equations presented by Karickhoff (1984) and in

Chapter 2. The two equations are given below:

Qwhole soil = Qmin + Qsom = (fa * Kmin * Ce * fmin) + (Kom * Ce * fom) [1]

108

where Qwhole so“ is the total solute mass sorbed by unit mass of whole soil (mg g'l), Qmin
and Qsom are the solute mass sorbed by unit mass of SOM-removed soil and SOM (mg g'
1), respectively, Cc is the aqueous solute equilibrium concentration, and Km and K0'“
represent the sorption coefficients of the solute by the soil fraction from which SOM was
removed (mineral) and SOM, respectively, fmin and fom are the fractional mineral fraction
and SOM contents of soil, and f,l represents the fractional availability of sorption sites on
the clays in whole soil (i.e. the fraction of the sorptive surfaces that are available in

unfractionated soil).

Using the F reundlich parameters, Kr and n, the Q value (amount NOC sorbed per unit
mass of sorbent) was calculated for four selected concentrations that were within the data
range (i.e., @ 5, 10, 15, 20 mg L"). Standard deviations for each fa value associated with
each concentration were calculated using error propagation procedures (Skoog and West,
1986). For the Karickhoff equation, the K+-saturated system was used and peat was used
as the SOM to estimate K0m values for p-NCB, 1,4-DNB, and 2,4-DNT. Within both the
Karickhoff and alternative equations, we accounted for the extent of solute sorption by
the residual SOM associated with the OM-removed fraction; the residual SOM being the

OM remaining afier extraction of bulk SOM using H202.

Sensitivity Analysis of f, values

Manipulating a modified form of Karickhoffs equation (Chapter 3) we derived a linear

equation relating f8 and the SOM and smectite contents:

fa = (“Kom / Kmin) X (fom/fmin) + (Kwhole / Kmin * fmin) [3]

109

Plotting fa versus fom/fmin for each soil should give an intercept of 1 when the SOM
content is zero (i.e., 100% availability of sorptive surfaces). The sorption coefficients for
sorption of a solute by SOM (i.e., Km) and by the smectites in the OM-removed soil (i.e.
Kmm) would differ for different solutes. To determine the Km (analogous to Ksmwim)
term in equation [3], the extent of sorption of each solute by peat soil and K+-beidellite
were described by fitting the data to the Freundlich equation as previously described. The
n and Kf sorption parameters were used to determine the sorption per gram of sorbent
(peat and K+-beidellite) at an equilibrium concentration of 10 mg L". Correlation
coefficients describing the relationship between f, values and the ratio of SOM to
smectite content (fem/fsmcme or %OM/%smectite), to SOM content, to smectite content in
unfractionated soils, to the sorption distribution coefficient (i.e., Kd) for solute sorption
by K+-beidellite, and to the ratio of sorption of solute by SOM and smectite at an
equilibrium concentration of 10 mg L'l (Kom/Ksmectite) were obtained. If the calculated f,l
values are reasonable, the intercept of a linear regression equation relating f. with
fom/fsmectite should tend to unity (i.e., no mineral blockage by SOM) when %OM tends to

ZCI'O .

Statistical Analysis

The sorption coefficient (Kf) and the constant related to the sorption isotherm curvature
(n) were estimated using the Freundlich equation, Q=Kch". Comparisons of sorption
isotherms were made by comparing the Kf and n values for each soil-solute system using

nonlinear regression in SAS (SAS Institute, 1999). A significant difference between

110

isotherms was assigned when there was a significant difference (a = 0.05) between Kf

and n or Kf values of compared isotherms (Laboski and Lamb, 2004).

111

RESULTS AND DISCUSSION

The Webster and Clarion soils contain significant amounts of smectite clays (13-24%
smectite) (Table 4.1) and were therefore chosen as the study soils. The amount of free
Fer associated with the Webster and Clarion soils is less than 1%. For p-NCB, 1,4-
DNB, and 2,4-DNT sorption by Fer synthesized in the laboratory was negligible (data
not reported). We therefore conclude that Fer did not contribute to sorption of p-
nitrocyanobenzene (p-NCB), 1,4-dinitrobenzene (1,4-DNB), and 2,4-dinitrotoluene (2,4-
DNT) in the soils studied. This is consistent with studies by Haderlein and
Schwarzenbach (1993) who reported that NACS do not adsorb significantly to aluminum

and iron (hydr)oxides, carbonates or quartz.

The sorption isotherms for p-NCB, 1,4-DNB, and 2,4-DNT sorption by each soil were L
shaped (Figure 4.1—4.4). For the concentration range used for each solute, the amount
sorbed did not plateau for any soil hence the suitability of the Freundlich equation to
describe these isotherms (Table 4.2) (Chiou, 2002). Sorption of p-NCB, 1,4-DNB and
2,4-DNT by the Mg2+-saturated soil was equal to or greater than that by the Mg2+-
saturated OM-removed soil (Figure 4.1-4.4). Removal of SOM may liberate clay surface
sites, but the sorption by the liberated Mg2+-exchanged clay sorption sites would be low
due to the large hydrated radius of the Mg2+ cation. Cations with large hydrated radii
obscure a greater portion of the clay interlayer sorption domain compared to cations with
smaller hydrated radii resulting in lower NAC sorption by Mg2+-saturated clays and soils
compared to K+-saturated clays and soils (Li et al., 2003; Sheng et al., 2002; Haderlein et

al., 1996; Weissmahr et al., 1997, 1999; Johnston et al., 2002). In addition, sorption

112

ILJf... é... . . .2

 

C38 oomv mounts.“ ofiooEm Emma 3883mm“
5.56 .0 Max:268. .M 65: 8:65 some .52 638:6 .mb

 

 

 

 

 

 

 

 

 

_ - - Rod 25
:82 >26
2 o.m.§.m E R 82 Re 35m 39
h - - 82 o5
:32 ago
a. o£.§.m 2 mm 83 Se 5% Eu
2 - - 22 on?
: o.m.§.m 2 a 83 as $230 95
m - -. RS 955
<3
mm. orgasw cm a 23 ”3 $230
”$3959 .3 .8399:
3.22::—
ooatsm _. >20 3:00:53 comma 20.x. In 9:55;. _mem
an? .x. 33.9 .x.

 

 

 

 

 

 

 

 

 

3382-20 m 8:5 6% Wm 856 dd €852-20 <

855 .05 8 8E6 .50 6232-20 m bag»? .83 mm 5283 .95 6382.20 < page? .035 2.. 5303 .35

.9308»: mm? 20m £033 Bob £8 was :8 2936 $8 .855 was “8363 no 8338835 _mflaonoofimmsm ._.v 033.

113

 

 

 

     

 

 

 

 

 

 

 

 

$0.5 ‘9 10
g 0 MK 3
g o W s
0*“ v M ”a“
8 0.31 g Q3.
2 o
'5 0.2« g M-
5 s
0.1‘ ‘5 0.2
8
C
o
0.0 4 0 no . .
0 10 3 $ 0 5 10 15 20 25 30
WWW(mL) WMW(NU)

 

   

1.2
O WA-K
O WAG-K
1 .0 ‘ ' WA“
V

9
O

o
h

 

Concentration of DNT sorbed (mg g")
.o 9
N a

 

 

P
c

0 5 10 15 20 25
Aqueous DNT consultation (mg L")

Figure 4.1. Isotherms for sorption of p-nitrocyanobenzene (p-NCB), 2,4-dinitrotoluene
(DNT), and 1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Webster A (WA) and

Webster A soil from which SOM was removed (WAO).

114

 

 

9
a

 

  
  
   

9 d d
o lo t
440.

J

 

 

a:

_I

  
 

Concentration of NCB sorbed (mg g“)

 

 

Concentration of DNB sorbed (mg g")
p g
a

 

 

8

 

 

1.4‘

4‘0.

1.2 .

Id
0
r

  

I
_l_ j

 

P
‘

Concentration of DNT sorbed (mg g' )
a :

 

P
o

0 5 10 15 20 25
Aqua. DNT Won (nu L")

Figure 4.2. Isotherms for sorption of p-nitrocyanobenzene (NCB), 2,4-dinitrotoluene
(DNT), and 1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Webster B (WB) and

Webster B soil from which SOM was removed (WBO).

115

 

 

O
O
V
V

MM

9

J

a
BM

   

4* 1

  
   

p
d
o

 

ntra

 

 

 

Concentration of uca sorbed (mg 9")

§ §

 

 

 

 

3
0‘0.

8

  

L

E

F

 

 

Concentration of DNT sorbed (mg g")

g

 

0 5 10 15 m 3 30
Amour: WT Wm (nu L")

Figure 4.3. Isotherms for sorption of p-nitrocyanobenzene (NCB), 2,4-dinitrotoluene
(DNT), and 1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Clarion A (CLA) and

Clarion A soil from which SOM was removed (CAO).

116

 

 

 

 
   
  

  
    
 

 

 

6‘07 $1.1
g g o QB-K
506- £11. 0 cab-K o
v 01.3w
30.5- 310- v caom
a a
now 030.81
0 z
z a
3&3: “5&6
S 5
3502‘ 30.44
g g
30.1 00.2
0
c c
800 80.0

 

 

 

 

° 1.0+

.09
ac:

 

on

 

 

 

Concentration of ONT 3 rho

Figure 4.4. Isotherms for sorption of p-nitrocyanobenzene (p-NCB), 2,4-dinitrotoluene
(DNT), and 1,4-dinitrobenzene (DNB) by K+- and Mg2+-saturated Clarion B (CLB) and

Clarion B soil from which SOM was removed (CBO).

117

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rmdd mvood nmmnd mmood ooKd flood 3.688 Endgame;
Nvood wmood omood omood 3 mod noood o_o:3-+~w2
ooomd omomd _mwmd oommd mvovd mom—d @9688 EO-+M
ommmd .2 3d wmmvd _vmmd owmmd moood flosawvm m 83.85
_omnd mmood Sid mmood _oond o_ood 3382 20-+Mw2
Good oo_od oommd owood Novnd mmood 03:3-me
Smmd Comd hmmvd owe—d 3.de omnod @2688 29.3
oomvd com _d mde monod vowed ommod o_on>»-+v_ < 83.8.0
woood m Sod _wvwd ooood vmwwd good 3382 20-...ng
mogd omood Sid flood movwd Cood flosawwwoz
vommd vwmvd omvvd _womd omfid 9.3 .o @0388 EO-+¥ m
mwovd m _omd mwmmd moomd owmmd Vmood o_o:3-+v_ 88303
owvwd food envod mmood omowd _oood 3888 EOJNWZ
Rood mm _ od whomd momod o y Nod $ood 038$-me
nwomd vommd voovd mummd mmomd Good 3888 20-3. <
wmmvd monmd ooovd Rood owomd nmmod o_o:>»-+v_ 88303
n- ov— : ov— a ov—
oflo—u—Ouchumflm—T‘N ofioufloa—Ohumflm—YVJ QEONflon—OflahochumZufl m—mcm

 

 

 

 

 

.3588 mm? 20m 533 88m $8 98 £8 20:3 @8088: m 98 <V corn—U was “2303 cowgnoxo-+~m2 Ea

-+v_ .3 ”8398 0808090888-»..N 98 0580908868; 6808308839885 .88 $29» 8 98 CM 5883 com—988m :3. 033.

118

of NACS through direct coordination of the N02 group with the interlayer cation is
hindered by the waters of hydration for more highly hydrated cations (e.g., Mg2+)
resulting in lower energies of interaction between the NAC and the cation and hence
lower NAC sorption compared to less strongly hydrated cation systems (e.g., K+) (Li et
al., 2004). Hence, the clays in the Mg2+-saturated OM-removed soil are relatively inactive
as sorbents for sorption of the NACS. The SOM in the Mg2+- saturated whole soil is the
principal sorbent phase hence with its removal p-NCB, 1,4-DNB and 2,4-DNT sorption

decreases in the Mg2+-saturated OM-removed soil (Figure 4.1-4.4).

The K+-saturated soils (unfractionated and OM-removed) sorbed p-NCB, 1,4-DNB and
2,4-DNT to a greater extent compared to Mg2+-saturated soils (Figure 4.1-4.4). These
results are consistent with several studies that report stronger sorption of NACS by
smectite clays saturated with cations of relatively low hydration energies (e.g. K+, Cs+)
compared to that by systems saturated with cations of relatively higher hydration energies
(e.g. Ca2+, Mg”) (Li et al., 2003; Boyd et al., 2001; Sheng et al., 2001 , 2002; Haderlein et
al., 1996; Weissmahr et al., 1997, 1999; Johnston et al., 2001, 2002) and with the fact that
smectite is the most abundant clay mineral in the Webster and Clarion soils (Chapter 3).
This result can be attributed, in part, to direct interaction between N02 groups of the
NACS and K+; complexation of this nature is inhibited by the stronger hydration of Mg2+
in the Mg2+-saturated soils (Boyd et al., 2001; Johnston et al., 2001, 2002; Sheng et al.,
2002; Li et al., 2003). Lower hydration of K+ (versus Mgr“) on the exchange sites also
manifests larger adsorption domains between exchangeable cations for NOC sorption

which facilitates interaction of the solute with the clay basal surface (Lee et al., 1990;

119

 

Jaynes and Boyd, 1991a, b; Haderlein and Schwarzenbach, 1993; Haderlein et al., 1996;
Weissmahr et al., 1997, 1998). Also, lower hydration of K+ (versus Mg”) gives rise to
smaller interlayer distances of ~3 A which are optimal for the solute to interact with
opposing siloxane sheets and thereby minimize its contact with water. In Chapter 1 we

reported similar trends for p-NCB sorption by surface soils.

The K+-saturated OM-removed soils sorbed more p-NCB, 1,4-DNB and 2,4-DNT than
the K+-unfractionated soils (Figure 4.1-4.4). Removal of SOM, therefore, increased
sorption of the NACS by each soil although SOM undoubtedly acts as an independent
sorptive sink for these solutes (Sheng et al., 2001). This suggests that SOM obscures
sorptive surfaces in soils. This effect is most likely related to obscuration of smectite
surfaces since they are the clays having the greatest affinities for NACS. It is possible to
observe the suppression of sorption in the presence of SOM because of the preferential
sorption of p-NCB, 1,4-DNB and 2,4-DNT to smectite clays compared to SOM as well as
the use of systems saturated with cations (e.g., [(4’) that facilitate relatively significant
sorption by smectites. Literature reports decreased sorption of organic solutes to smectite
clays when associated with SOM (Liu et al., 2002; Celis et al., 1997; Pusino et al., 1992,
1994). Although these studies indicated that the presence of SOM associated with
smectite clays results in decreased sorption of solutes by the clays, these studies utilized
synthetic clay-SOM complexes. Our data indicates that in natural soils, removal of SOM
from soil increases sorption of solutes indicating that in natural soils, as in synthetic

systems, SOM obscures clay surfaces for sorption of NACs.

120

 

The fa values (i.e. the fraction of mineral surfaces that are available in unfractionated soil
for solute sorption) derived using the Karickhoff and alternative equations, are given in
Tables 4.3, 4.4 and 4.5, respectively. The fa values derived using Karickhoft‘s equation
and K+-saturated peat were similar to those derived using Mg2+-saturated peat (Tables
4.3, 4.4). For each equation, the fa values differed across soils for each NOC. For p-NCB,
1,4-DNB and 2,4-DNT, the fa values calculated for each soil using the Karickhoff (1984)
equation [1] (Tables 4.3, 4.4) were similar to those calculated using the equation [2] we
proposed in Chapter 2 (Table 4.5). Considering the NACs (p-NCB, 1,4-DNB and 2,4-
DNT) both equations yielded f, values for Clarion B that were greater than for Clarion A;
fa values for Webster A and B were similar (Tables 4.3, 4.4, 4.5). If all carbon (C) in
SOM were assumed to cover the basal surfaces of soil smectites, the estimated maximal
percent of the smectite basal surface covered by SOM (calculated using the Lagaly
(1981) alkylammonium method as outlined in Chapter 3) in whole soil was greater for
Clarion A (49%) compared to Clarion B (10%) and similar for Webster A (38%) and
Webster B (31%) which follows the trend for the complement of the f, vales (i.e., 1- f,)
for each horizon. One conclusion from this analysis is that the basal surfaces of smectite
clays are partially covered by SOM as suggested by Stevenson (1994). We know that
NACs intercalate between clay layers When they sorb to smectites (Sheng et al., 2001,
2002; Li et al., 2003). Small organic molecules can also occupy clay interlayer sites
(Jaynes and Boyd, 1991a, b; Lee et al., 1989, 1990). Another plausible mechanism is that

SOM blocks access to the interlayer of swelling clays.

121

 

.maoafiaa a 82338 838% .828 u 828
V mUZ 88 839 d 05 88888388 808 8 .m 82.85 Sm .HZQ v mZQ v mUZ 88 8883 d 05 80888888 8080 8 .m 8830?
888 < 8830? .< 82.820 Sm .God v 3 m 88303 v m 88.85 H < 88303 v < 82.85 So 329, no .m80888888oo some $88
.HZQ 8m .God v 8 m 82.85 v < 8830? v m 8330? v < 83.820 So 8883 d .m8oueb8oo8oo some 8808 £75 888 mUZ 8n— .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

838 ad 688 So Q8 8.8 888 So om

88.8 85 8.8 88 98.8 2: A88 83 2

688 So #88 $8 688 2: 88.8 83 8

80.8 83 :88 and 888 35 §8 93 w 98332.88

888 83 888 88 43.8 m3 888 w; on

$88 83 88.8 88 888 E. E8 who 2

A88 88 688 38 A88 88 A88 Ed 8

:88 e; 888 83 888 $8 888 88 m «58.32888

#88 £5 :88 Rd 888 :3 9.88 86 on

@8 33 :88 m2 888 :3 88.8 $8 2

A88 83 :88 NS 8.8 83 88.8 88 S

@838 :88 Rd 888 as 988 m2 A Mm M 058323882
8 .3303 < .3203 -8 a

m— 8385 < 838—0 838— 838— 8.8%...83880 com: U<Z

 

 

122

.888 sov— 05 088.88% 8 83 88833
-+v~ .3 88898 828 w8m3 a: 88883 armada—05M w8m8 888888 $808.88 m 888 <v 828—0 888 $808.88 m 888 .3 88303

B 8898 8.28 usasoefivwm as. .828 05838883; .888 0583053388 8 329 8 25 .2. 2.5

880580.88 2 828328 8.88885 .975 u mZQ
v m52 88 838, _w 05 82888888 8080 8 .m 82.85 .88 .HZQ v mZQ v M52 88 32? d 05 82888888 8080 8 .m 8830?
888 < 8830? .< 82.85 80”— Amod v 8 m 83,303 v m 82.85 H < 8833 v < 82.85 88 mos—S, a £82888oo8oo 8080 $8.88
HZQ Sm Amod v 3 m 82.85 v < 88303 v m 8830? v < 8285 How 82? «.2 “8888888383 8080 $088 .949 888 mUZ 8cm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

85.8 23 898 :8 80.8 8.2 28.8 22. cm
888 38 $88 $8 288 2: 28.8 mg 2
688 88 28.8 88 688 8: 688 88 8
888 23 888 88 A288 38 23.8 88 w 05232858
28.8 83 888 88 23.8 mg 888 83 cm
85.8 :3 83.8 23 80.8 :8 ~88 a; 2
20.8 a; 888 88 28.8 %d 288 Ed 2
:88 Rd :58 m2 888 38 2.8.8 :8 m 8382.98.58
988 $8 :88 38 no.8 83 28.8 o; cm
8.8 23 :88 as 28.8 83 58.8 88 2
28.8 88 :88 one no.8 83 888 88 8
$88 a; :88 Rd 288 23 £88 83 m 958383282
8 82.3.0 < 82.3.0 8 .3303 < .8303 :8 mi
53¢— ch— flfiuflhu—SOEOU 50m: U42

 

 

A888 sex 05 0288808 9 808 808888

-832 as 82888 328 M883 H: 82883 mrmofiotavm m2? 8088280 $802.58 m 888 «G 80885 28 $882.88 m 888 «8 88303

E 8888 8.28 883858848 Ba .8288 2583853 .8028 8503588888 :8 82$ 8 28. .3 288

123

 

. $805558 E 323338 Emu—8% .HZD v mZD
n mUZ 8m 838, d 05 downbcooaoo some 8 am coca—U 8m .HZQ v mZQ v mUZ S.“ 833 d 05 .cowabcoocoo :23 as .m “3303
was < bane? .< coin—O 8m Amod v 3 m Sumac? v m cots—U n < “8303 v < cord—U Sm mos—g d .mcouwbcoocg some $83
.HZQ you Amod v 3 m corn—U v < H8303 v m .8333 v < cord—U 8m 82? am .mcouwbcoocoo some £88 £75 28 mUZ Sm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:38 go A88 o3 ~38 2: A88 .38 cm

A88 23 898 $8 698 2: A88 88 2

698 «do :98 as 698 2: A88 86 2

A88 58 $8 Ed :58 3o ~88 ”no m 95322.35
$98 93 48.8 $5 A38 Ed A38 So cm

9.8 So :38 86 A88 NZ A88 :3 2

A88 83 5.8 and :58 $5 A88 ”3 2

:98 Ed A88 86 A88 85 48.3 m; m 055.3955:
A88 36 98.8 33 A88 83 58 Ed om

A88 23 :c8 ~28 A88 83 A88 3o 2

§8 23 :98 a2 §8 88 A88 35 8

$8 a; 2:8 98 §8 93 A88 So m «5238:3222
m Evin—U < Evin—U m 53393 «33 < 582.03 «.33 acuflwflM—Hflcb tom: DO Z

 

 

A8 cognac @3358.“ So wai— uoumfiumo Amcoutos m 98 <8 cots—U was Amcomto: m 98 <v “8303

.3 8:88 Cze 88E§E3.~ Ea .528 8%:3333; .5qu 853053083 5 82? a 2F .2. ”BS

124

The f, values for p-NCB, 1,4-DNB, and 2,4-DNT were strongly negatively correlated
(61-88%) with the ratio of SOM to smectite contents (%OM/%smectite) in soils; the
percent correlation being in the order p-NCB > 1,4-DNB > 2,4-DNT (Table 4.6). This is
reasonable since smectites have the highest afiinity for NACS among soil clays (Chapter
1). The fa values for p-NCB, 2,4-DNT and 1,4-DNB were moderately negatively
correlated with the %OM contents of the soil horizons with 45-51%, 24-27% and 21-33%
of the difi‘erence in f, values between soil horizons attributed to differences in SOM
contents, respectively (Table 4.6). The fa values for p-NCB, 2,4-DNT and 1,4-DNB were
weakly correlated with the %smectite contents of the soil horizons with 17-21%, 31-33%
and 33-43% of the difference in f, values between soil horizons attributed to differences
in smectite contents, respectively (Table 4.6). Linear regression analysis of the f, values
and the ratios of SOM to smectite contents of the soil horizons indicates that the f, values
calculated are reasonable since, for each regression, the intercepts are approximately
equal and all approach unity at 0% SOM (Table 4.6). The fa values for the NACS, for
each soil horizon, decreased in the order 2,4-DNT > 1,4-DNB > p-NCB, suggesting that
the mechanism driving sorption site availability is influenced, in part, by the nature of the

solute.

Data analysis revealed that the fal values for the Webster A, Webster B and Clarion A
soils across solutes were strongly negatively correlated with Kom/Ksmectitc with 68-89% of
the differences in fa values between solutes attributed to differences in the relative
affinities of the solute for SOM and smectite in these soils (Table 4.7). For each of these

soils, the fa values and Qbeidellite (i.e., sorption distribution coefficient for sorption of each

125

 

Table 4.6. Correlations between f. values for sorption of p—nitrocyanobenzene (NCB),
1,4-dinitrobenzene (DNB) and 2,4-dinitrotoluene (DNT) estimated using Karickhoff
equation [1] (using solute sorption by K+-saturated peat to determine the KOlm term) and
our alternative equation [2] and percent organic matter (%OM), percent smectite
(%smectite) and ratio of organic matter to smectite (%OM/%smectite).

Correlations were performed across the Webster and Clarion (A and B horizons) soils (or

 

 

 

 

 

 

 

= 0.5).

Equation NOC %OM %smectite °/oOM/°/osmectite Intercept

used used

Karickhoff DNT -0.27 0.31 -O.64 1.07 (0.08)
DNB -0.33 0.33 -O.75 0.92 (0.06)
NCB -O.51 0.17 -0.88 0.85 (0.04)

Alternative DNT -O.24 0.33 -0.61 1.08 (0.08)
DNB -0.21 0.43 -0.66 0.91 (0.06)
NCB -O.45 0.21 -0.83 0.85 (0.04)

 

 

 

 

 

 

 

Standard deviations in parentheses

126

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$83 8 2: 8m: 82
£33 82: 38.". ES
8.? 3.8 $8 :8 :3 85 N3 86 ~82. 5.2 8N3 Ea 35.5.?
$36 2 z: 88: 82
833 OB: £52 mza
:8 3.8 $8 NS- 83 :3 3o 85 N83 5.? an: 8.75 usages.
Bu <5 In? <3 50 <5 ImB <3 22.5%.: A.» «.5 A.» ”.5 Ba. 3....
gagged. :wdé 2:330 u _...._o .._.__._...._o a
88.22.:6 ..____._._zo\..ao . . 0oz .822. m

 

 

 

 

 

 

 

 

8:85.858 3502398 23 20m
3 coma—cm 828 Mo 3568 .«o 2:2 05 93 2:320 28 _S 53950 03862? .50 28 9:8 50M 05 058883 8 Son @88me
-+v~ 3 8:98 328 mammsv H: 553:8 boiotaM mama: 33858 82? d 5953 macaw—oboe commmohwom 65323-...»—
wcw 20m 3 E598 Crzav oaoflouobWEVWN can AmZQV ocoNnonobEm?vJ AmUZv ocoucoaofiwxoobmcd mo 35:8 no can“ 05

EB 74 we 3 mo :oumbaoocoo Saw—£253 an S :8 AaaOv “won 98 A§=§£Ov oE—oEonfiM 8 3&8 82% mo “55:2 .58 £an

127

NAC to K+-beidellite clay at an equilibrium concentration of 10 mg L'l) were strongly
correlated with >90% of the difference in f, values for each soil across NACS attributed
to the difference in affinity of each NAC for smectite clay (Figure 4.5, Table 4.6). These
results suggest that the greater the extent of sorption of the NAC to swelling clays, the
less impact SOM may have on availability of sorptive surfaces in soils. For Clarion B, fa
values were negatively correlated with Kem/Ksmwm with 9-11% of difference in fa values
for this soil across solutes attributed to differences in the relative affinities of the solute
for SOM and smectite clay in this soil. The f, values for Clarion B, across solutes, were
correlated to Qbeidellite with 47-49% of the difference attributed to difference in affinity of
each NAC for smectite clay. This may be due to the small initial amounts of SOM in this
soil which we propose covers only 10% of the clay surface in whole soil (Chapter 3).
Thus we suggest the efficacy of SOM to obscure sorptive clay surfaces in this case is

probably minimal.

Our data suggest that fa values for NACS approach 1 in cases where the NAC is strongly
sorbed (presumably by smectite) and the OM/smectite ratio low. For example, f. values
for 2,4-DNT sorbed by the Webster B soil approach 1 (Table 4.3, 4.4). Since in theory fa
can not be larger than 1, it was of interest to test our method for evaluating fa using a
compound that sorbs more strongly to smectite clays than DNT, viz., 1,3,5-
trinitrobenzene (1,3,5-TNB) (Haderlein et al., 1996). We measured sorption of 1,3,5-
TNB to K+- and Mgztexchanged Webster B whole and OM-removed soils using a batch
equilibrium technique (Figure 4.6), and calculated fa values for the selected

concentrations 5, 10, 15, and 20 mg L‘l, as previously described. The determined fa

128

 

values were close to unity ranging from 1.05 i 0.04 to 1.09 i: 0.09 and 1.06 :h 0.04 to
1.10 :t 0.09 using equation [1] and [2], respectively, for data analysis. This observation
helps validate our approach for the experimental evaluation of clay mineral surface
availability. These results suggest that the relative SOM and smectite contents of the soil
horizons as well as the nature of the solute, primarily its adsorption capacity to swelling
clays, influence the efficacy of SOM to obscure sorptive surfaces of soil clays and

therefore influence the availability of these sites in whole soil for the sorption of NACS.

129

 

O NCB
o DNB a
v DNT

0|
1

o"

h
1

O
..‘

N
L
\‘

 

d
I
.\
.

 

 

Concentration of solute sorbed, Q (mg g“)
a
Q

o ‘ I I I I I
0 5 10 15 20 26 30

Aqueous solute concentration, C. (mg L")

Figure 4.5. Isotherms for sorption of p-nitrocyanobenzene (NCB), 2,4-dinitrotoluene
(DNT), and 1,4-dinitrobenzene (DNB) by K+-beidellite.

(isotherms with different letters are statistically different, or = 0.05)

130

 

 

 

a." 3.5
0’ e WB-K . TNB
2’30. 0 WBO-K .
f; v WB-Mg
- 00
g 2.5 _ V WBO Mg ; O
o o
a 20
m - Q 0)
E o ' .’
00
“6 15‘ /
= s
o O
a 10- .5
t-
s /o
3 0.5
5 .
U 0.0 '37- 7 7 7 — ..,_—_____ 7.7__.____77 ‘
0 5 10 15 20 25

Aqueous TNB concentration (mg L")

Figure 4.6. Isotherms for sorption of 1,3,5-trinitrobenzene by IC- and Mg2+-saturated

Webster B (WB) and Webster B from which SOM was removed (WBO).

131

 

REFERENCES

Boyd, S.A., G. Sheng, B.J Teppen, and CT Johnston. 2001. Mechanisms for the
adsorption of substituted nitrobenzenes by smectite clays. Environ. Sci. Technol.
35:4227-4234.

Celis, R., L. Cox, M. Hermosin, and J. Comejo. 1997. Sorption of thiazafluron by iron-
and humic acid-coated montmorillonite. J. Environ. Qual. 26:472-479.

Celis, R., C. Hermosin, L. Cox, and J. Comejo. 1999. Sorption of 2,4-
dichlorophenoxyacetic acid by model particles simulating naturally occurring soil
colloids. Environ. Sci. Technol. 33:1200-1206.

Chiou. C. 2002. Partition and adsorption of organic contaminants in environmental
systems. John Wiley and Sons, New Jersey.

Gee, G. and J. Bauder. 1986. Particle-size analysis. p. 383-411. In A. Klute (ed.) Methods
of Soil Analysis. Part I — Physical and mineralogical methods. 2nd ed. Agronomy. 9.
ASA, Madison, WI.

Haderlein, S., and R. Schwarzenbach. 1993. Adsorption of substituted nitrobenzenes and
nitrophenols to mineral surfaces. Environ. Sci. Technol. 27:316-326.

Haderlein, S.B., K. Weissmahr, and R. Schwarzenbach. 1996. Specific adsorption of
nitroaromatic: explosives and pesticides to clay minerals. Environ. Sci. Technol. 30:612-
622.

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

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

Johnston, C., M.F. de Oliveira, B.J. Teppen, G. Sheng, and SA. Boyd. 2001.
Spectroscopic study of nitroaromatic-smectite sorption mechanisms. Environ. Sci. and
Technol. 35:4767-4772.

132

 

Johnston, C., G. Sheng, B. Teppen, S.A. Boyd, and M. F. de Oliveira. 2002.
Spectroscopic study of dinitrophenol herbicide sorption on smectite. Environ. Sci.
Technol. 36:5067-5074.

Karickhoff, S. 1984. Organic Pollutant Sorption in aquatic systems. J. Hydraulic Eng.
110:707-735.

Kunze, G., and J. Dixon. 1986. Pretreatment for mineralogical analysis. p. 92-99. In A.
Klute (ed.) Methods of Soil Analysis. Part I — Physical and mineralogical methods. 2'”.
Agronomy 9. ASA, Madison, WI.

Laboski, C.A.M., and J .A. Lamb. 2004. Impact of manure application on soil phosphorus
sorption characteristics and subsequent water quality implications. Soil Sci. 169: 440-
448.

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

Lee, J ., M. Mortland, C. Chiou, D. Kile, and SA. Boyd. 1990. Adsorption of benzene,
toluene, and xylene by two tetrarnethylammonium-smectites having different charge
densities. Clays Clay Miner. 38: 113-120.

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

Li, H., G. Sheng, B. Teppen, C. Johnston, and SA. Boyd. 2003. Sorption and desorption
of pesticides by clay minerals and humic acid-clay complexes. Soil Sci. Soc. Am. J.
67:122-131. '

Li, H., B.J. Teppen, D. Laird, C. Johnston, and SA. Boyd. 2004. Thermodynamics of
nitroaromatic compound adsorption from water by smectite clay. Environ. Sci. Technol.
38:5433-5442.

Liu, W., J. Gan, and S. Yates. 2002. Influence of herbicide structure, clay acidity, and
humic acid coating on acetanilide herbicide adsorption on homoionic clays. J. Agric.Food
Chem. 50:4003-4008.

133

Nelson, D. and L. Sommers. 1996. Total carbon, organic carbon, and organic matter. p.
961-1010. In D. Sparks (ed.) Methods of Soil Analysis. Part III — Chemical methods.
Agronomy. 5. ASA, Madison, WI.

Onken, B., and S. Traina. 1997. The sorption of pyrene and anthracene to humic acid-
mineral complexes: Effect of fractional organic carbon content. J. Environ. Qual. 26:126-
132.

Pusino, A., W. Liu, and C. Gessa. 1992. Influence of organic matter and its clay
complexes on metolachlor adsorption on soil. Pestic. Sci. 36:283-286.

Pusino, A., W. Liu, and C. Gessa. 1994. Adsorption of triclopyr on soil and some of its
components. J. Agric. Food Chem. 42: 1026-1029.

SAS Institute. 1999. SAS 8.1. SAS Institute, Inc., Cary, NC.

Sheng, G., C. Johnston, B. Teppen, and SA. Boyd. 2001. Potential contributions of

smectite clays and organic matter to pesticide retention in soils. J. Agri. Food Chem.
49:2899-2907.

Sheng, G., C. Johnston, B. Teppen, and SA. Boyd. 2002. Adsorption of dinitrophenol
herbicides from water by montmorillonites. Clay Clay Min. 50:25-34.

Skoog, D. and D. West. 1986. Analytical chemistry: an introduction. CBS College
Publishing, New York, NY.

Soil Survey Staff, Natural Resources Conservation Service, United States Department of
Agriculture. Soil Series Classification Database [Online WWW]. Available URL:
http://soils.usda.gov/soils/technical/classifimn/scfile/index.html (verified 10 Feb.
2004)

Stevenson, F. 1994. Humus chemistry, genesis, composition, reactions. John Wiley and
Sons, New York, NY.

134

 

Walker, A., and D. Crawford. 1968. The role of OM in adsorption of the triazine
herbicides by soil. p. 91-108. In International Atomic Energy Agency. Isotopes and
radiation in SOM studies. lntemational Atomic Energy Agency, Vienna.

Weissmahr, K., S. Haderlein, and R. Schwarzenbach. 1997. In situ spectroscopic
investigations of adsorption mechanisms of nitroaromatic compounds at clay minerals.
Environ. Sci. Technol. 31 :240-247.

Weissmahr, K., S. Haderlein, and R. Schwarzenbach. 1998. Complex formation of soil
minerals with nitroaromatic explosives and other pie-acceptors. Soil Sci. Soc. Am. J.
62:369-378.

Weissmahr, K., M. Hildenbrand, R. Schwarzenbach, and S. Haderlein. 1999. Laboratory
and field scale evaluation of geochemical controls on groundwater transport of
nitroaromatic ammunition residues. Environ. Sci. Technol. 33:2593-2600.

135

 

CONCLUSION

The molecular and environmental factors that control the sorption of nitroaromatics
(NACS) in the environment must be more accurately understood for the purposes of
controlling their mobility and reactivity in the-environment. Soil clays and soil organic
matter (SOM) undoubtedly function in concert than as separate entities due to their
intimate association in surface soils. Our studies indicate that adsorption of NACS by
surface soils is strongly influenced by the nature of the dominant exchangeable cation. In
addition, the association of SOM with clay reduces the availability of sorptive surfaces in
soils, with availability being dependent on the relative contents of SOM and sorptive
clays (and the nature of the clays) in the soil as well as the nature of the solute in the
whole soil. We determined that the equation proposed by Karickhoff is best suited to
compounds that sorb significantly to clays such as NACs such that the sorption to the
clay fraction is large enough to facilitate accurate calculation of f,. This issue was
encountered when using diuron as the solute. Fate and transport models that utilize
sorption parameters based on the additive effect of reference clay and humic material

sorption may deviate from reality and may overestimate the amount of NAC sorbed.

136

          
 

llittlirtELr‘ul‘lflmililljlMilli"

1 2