us..-

‘ . 5.>-'
.m.» A
.4 2 ~..-,.

_ ,—
52:.
,4 .

.. ..‘
A. .
c.

.

    

4 “-7 - —.—~.
.ANA-"' .;:’3“""‘.;.-
2 4,1"; r .D.‘f.’.‘t:_‘$'v-

‘ t
- rz~v<;...._-

v -'
v" ,‘

$ 1.“;

gf

a

   

W

_ - :‘v. ‘,o ‘

FA. V‘

. fin».-.

 

 

 
 
 
   
 
 
 
  
  
 
    
  
  
   
   
           
   
 
 
 
  

5321.5.

  

  
   

‘ ‘

 

   

 
 
 
 

    

. whirl 1‘
‘1“ J 75' ". '
. fin}?! .. 3:1};4.‘
K ' '7 ;‘;‘.'I‘: \l‘vW
"' M ‘ “n3. ‘3 u
A.» - Hf! n 11'
. “'11 3:3

t’z‘g i
'1

r.
H‘

.-_‘._ -‘Nl. « .

 
 
  
  
   
   
   

 
    

    
 
 

   
 
 

 

    

”4-31"

Ci Wit 4—
; .-' o; ‘3‘ W: ‘ F ’1! ..
-:2£'aaatrsgr§i tgifi' "‘mgifr- :t
.ul‘L'm .;' " . . 3.. I
:fig'flaé .,; i ”
313' ‘v' M
0

01-.“

.__.“'
...__.
.37“
—-- --*-v::"

     
      
      
 

 

 
 
  
    

Iliiiliiiiililifiil‘llii‘li’l‘lifi’lflilililifiliil

a< 3 1293 02058 2072

This is to certify that the
dissertation entitled
Effects of Aging on the Sorption/Desorption
Behavior of Organic Chemicals in Soils and
Relation to Bioavailability

presented by

Michael Steven Sharer

has been accepted towards fulfillment
ofthe requirements for

Ph.D. degree in W1 Soil
Chemistry/Toxicology

 

DateW ?

MSU is an Affirmative Action/Equal Opportunity Inslitun'on 042771

 

 

 

 

 

 

LiBRARY
Michigan State
University

 

 

 

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

 

. 4'"

(Pin

(7,
20

- n n
JHNDE

A—

UJD

 

—..

2
v

 

 

 

 

 

 

 

 

 

 

 

 

11/00 W.MM

 

 

THE EFFECTS OF AGING ON THE SORPTION/DESORPTION BEHAVIOR OF
ORGANIC CHEMICALS IN SOILS AND RELATION TO BIOAVAILABILITY
By

Michael Steven Sharer

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Crop and Soil Sciences

1999

ABSTRACT
EFFECTS OF AGING ON THE SORPTION/DESORPTION BEHAVIOR OF
ORGANIC CHEMICALS IN SOILS AND RELATION TO BIOAVAILABILITY
By

Michael Steven Sharer

Observations on the behavior of pesticides in field-weathered soils have
demonstrated that increased soil-chemical contact time (aging) may lead to increased
persistence of contaminants in the environment. Aged chemicals can act as long-term
sources of groundwater contamination, and aging may reduce the effectiveness of soil
remediation technologies such as soil washing, pump and treat, and in-situ
bioremediation. Environmental fate characterization based on laboratory studies of
freshly added chemicals fail to consider the effects of aging and hence may underestimate
the persistence of organic compounds in the environment. For example, although the soil
fumigant ethylene dibromide (EDB) is readily biodegradable, relatively water soluble,
and volatile, EDB residues have been found in field soils 20 years after its last known
application. However, the causes of increased persistence due to aging are not well
understood. One possible cause of increased persistence is a reduction in the
bioavailability of contaminants to degrading microorganisms. This decrease in
bioavailability is likely due to changes in the physical/chemical disposition of the
contaminants due to aging. For example, an increase in sorption to soils due to aging can
limit the extent of biodegradation of organic contaminants. Also, desorption from soils,
which may decrease with aging, is thought to be the limiting factor controlling the rate of

degradation of organic contaminants by degrading microorganisms.

This study describes three experiments designed to test the role of aging and its
effect on the sorption/desorption dynamics and bioavailability of chemicals in soils. One
experiment examined the sorption/desorption behavior of chlorobenzene on four soils
(soil organic carbon content from 0.4 — 7.8 %) after soil-chemical contact times (aging) of
24 hours and 14 months. Another experiment examined the sorption/desorption behavior
of four compounds (EDB, chlorobenzene, 2,4-D, and atrazine) on one soil type after
aging times of 24 hours, 1 month, and 14 months. Lastly, the sorption/desorption
behavior and bioavailability of biphenyl to Pseudomonas putida P106 was studied afier
aging times of 24 hours and 8 months.

Increases in sorption due to aging were found for some chemicals (2,4-D, EDB)
but not others (chlorobenzene). Large differences in sorption/desorption behavior for
similar compounds (EDB and chlorobenzene) were observed, indicating that
generalizations about compound behavior in soils are inappropriate. Desorption profiles
were consistent in that aged chemicals showed a much larger desorption resistant fraction
after long aging periods and this was true for all soil types. Desorption rates were
unaffected by aging time. Biodegradation of biphenyl decreased with increased aging
time and the desorption resistant fraction (observed in the desorption profiles) appeared to

be inaccessible to microbes.

I would like to dedicate this work to my mother and father.

iv

ACKNOWLEDGEMENTS

I would like to thank the members of my graduate committee, Dr. Stephen Boyd
(Chair), Dr. Thomas Voice, Dr. Matthew Zabik, Dr. Sharon Anderson, and Dr. Boyd Ellis
for their guidance, support, and teaching during my tenure at Michigan State.

I would also like to thank Dr. Stephen Boyd and the USDA for their financial
support for this research project. Just as important were the interactions and discussions
with the members of Dr. Boyd’s lab and people in the other labs on the fifth floor of

PSSB.

TABLE OF CONTENTS

List of Tables ................................................................................. viii
List of Figures ................................................................................ x
Chapter 1 Effects of Aging on the Sorption/Desorption Behavior
of Organic Chemicals in Soils and Relation to Bioavailability
Introduction and Literature Review ....................................... 1
List of References .......................................................... 18
Chapter 2 Effects of Aging on the Sorption/Desorption Behavior

of Chlorobenzene on Soils with a Range of Organic

Carbon Content ............................................................. 23
Abstract ..................................................................... 23
Introduction ................................................................. 24
Experimental Section ...................................................... 27
Results ....................................................................... 32
Discussion ................................................................... 42
List of References .......................................................... 48

Chapter 3 Effects of Aging on the Sorption/Desorption Behavior of EDB,

Chlorobenzene, 2,4-D, and Atrazine .................................... 51
Abstract ..................................................................... 51
Introduction ................................................................. 52
Experimental Section ...................................................... 54
Results ....................................................................... 59

vi

Discussion .................................................................. 66
List of References .......................................................... 79

Chapter 4 Effects of Aging on the Bioavailability and Sorption/Desorption

Behavior of Biphenyl in Soils ........................................... 82
Abstract ..................................................................... 82
Introduction ................................................................ 83
Experimental Section ...................................................... 86
Results ....................................................................... 91
Discussion ................................................................... 97
List of References .......................................................... 104

vii

Table 2.1

Table 2.2

Table 2.3

Table 2.4

Table 2.5

Table 2.6

Table 3.1

Table 3.2

Table 3.3

Table 3.4

LIST OF TABLES

Characterization and properties of the soils used for
sorption/desorption studies with CB .................................... 28
Linear sorption model (x/m = Kp 0 C.) data after 24 hours and 14
months of aging ............................................................ 32
Freundlich isotherm (x/m = K 0 Ce") parameters after 24 hours and

14 months of aging ......................................................... 34
Organic matter normalized sorption values (Km) for

chlorobenzene ............................................................... 35
Chlorobenzene desorption data after 24 hours and 14 months

of aging ...................................................................... 36
Desorption equation parameters (y = aoefk“) + boeH‘Zt) ) for
chlorobenzene ............................................................... 36
Linear sorption model (x/m = Kp - C.) data for EDB, CB,

AT, and 2,4-D .............................................................. 60
Organic matter normalized sorption values (Km) for EDB,

CB, and AT ................................................................. 62
Freundlich isotherm (x/m = K 0 Ce") parameters after 24 hours,

1 month, and 14 months of aging ........................................ 62
Desorption data (fun at 4 hours of desorption) afier 24 hours,

1 month, and 14 months of aging ........................................ 64

viii

Table 3.5

Table 4.1

Table 4.2

Table 4.3

Desorption equation (y = aoeH‘") + boem‘) ) parameters after

24 hours, 1 month, and 14 months of aging ............................ 65
Freundlich isotherm (x/m = K 0 Ce") parameters after 24 hours

and 8 months of aging time ............................................... 93
Desorption equation (y = aoeH‘") + boeH‘Z‘) ) parameters after 24
hours and 8 months of aging .............................................. 94
Comparison of biological and chemical assays of biphenyl after

24 hours and 8 months of aging .......................................... 100

Figure 1.1

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

LIST OF FIGURES

Schematic diagram of the Two-Box Model (from Guerin

and Boyd, 1993) ............................................................ 8

Sorption isotherms of CB on the four soil types after 24

hours and 14 months of aging ............................................ 33

Desorption of CB on Colwood A after 24 hours and 14 months

of aging ...................................................................... 37

Desorption of CB on Capac A after 24 hours and 14 months

of aging ...................................................................... 38

Desorption of CB on Schoolcrafi A after 24 hours and 14

months of aging ............................................................ 39

Desorption of CB on Capac B after 24 hours and 14 months

of aging ...................................................................... 40

Sorption isotherms of CB, EDB, 2,4-D, and AT after 24 hours

and 14 months (or 1 month for AT) of aging ........................... 61

Desorption kinetics of 2,4-D after a 24 hour and 14 month

aging period ................................................................. 68

Desorption kinetics of 2,4-D on Capac A afier a 1 month

aging period ................................................................. 69

Desorption kinetics of 2,4—D on Capac A after a

desorption period of 2800 hours .......................................... 70

Desorption kinetics of AT after aging periods of

Figure 3.6

Figure 3.7

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Figure 4.6

24 hours and 1 month ...................................................... 71
Desorption kinetics of EDB afier aging for 24 hours

and 14 months ............................................................... 72
Desorption kinetics of EDB after a desorption period of up

to 650 hours .................................................................. 73
BP sorption on Capac A after 24 hours and 8 months of aging ...... 92
BP sorption on Capac B after 24 hours and 8 months of aging ...... 92
Desorption kinetics of BP on Capac A after 24 hours and

8 months of aging .......................................................... 95
Desorption kinetics of BP on Capac B after 24 hours and

8 months of aging ........................................................... 96
Mineralization kinetics of BP to 14C02 by P. putida afier

different aging periods on Capac A ...................................... 98
Mineralization kinetics of BP to 14CO; by P. putida after

different aging periods on Capac B. For legend, see figure 4.5 ...... 99

xi

Chapter 1

EFFECTS OF AGING ON THE SORPTION/DESORPTION BEHAVIOR OF

ORGANIC CHEMICALS IN SOILS AND RELATION TO BIOAVAILABILITY

Introduction and Literature Review

An understanding of the sorption/desorption dynamics of organic contaminants in
field-weathered soils is crucial to understanding contaminant fate and behavior in the
environment. Sorption/desorption processes are known to affect biodegradation, which is
one of the main routes for the dissipation of contaminants in the environment (Alexander,
1994). Sorption of contaminants to soils is known to limit the extent of biodegradation in
the environment. Desorption may be the kinetically slow step controlling the rate of
contaminant biodegradation in the environment (Carmichael, et al., 1997). Soil-chemical
contact time (aging) has been shown to play a dominant role in the sorption/desorption
dynamics of contaminants in soils. The effects of aging may reduce the effectiveness of
soil remediation technologies (soil washing, pump and treat, and bioremediation) and
aged chemicals can act as long term sources of groundwater contamination. The main
purpose of this work is to determine the effects of aging on the sorption/desorption and
biodegradation dynamics of contaminants in soils.

According to partition theory, sorption of organic contaminants in soil-water
systems is due primarily to partitioning into bulk phase soil organic matter (SOM)
(Chiou, et al., 1979). Partitioning of organic contaminants into SOM is analogous to the
dissolution of an organic chemical in a bulk phase organic solvent. The extent of uptake

depends on the SOM content and the relative solubility of the contaminant in water and

SOM. This process is distinct from adsorption, which refers to the condensation of a
solute or vapor on the adsorbent surface. Partition theory predicts linear and singular
isotherms whether approached fi'om the sorptive or desorptive direction. Most laboratory
studies dealing with sorption and desorption of organic contaminants and pesticides have
assumed rapid (24 hours or less) and reversible equilibrium. There are numerous
examples that demonstrate these assumptions may not be valid. Deviations from these
assumptions could have major effects on predictions of contaminant transport and
biodegradation. One study indicated that sorption was a slow process and could take
hundreds of days to reach equilibrium and that sorption on aquifer material was an order
of magnitude greater than that predicted by partition theory (Ball and Roberts, 1991).
Another study found that the equilibrium sorption coefficient (K,,) may increase between
30 % and tenfold between short and long equilibration times (Pignatello and Xing, 1996).
These investigators also observed that sorption isotherms might become increasingly
non-linear with time (Xing and Pignatello, 1996). There are several recent examples in
the literature that sorption and desorption isotherms are ofien not reversible. One study
found that desorption was not complete and reversible even when the adsorption time
scale was short (7 days) (Pavlostathis and Jaglal, 1991). F u et a1. (1994) found that
sorption was rapid (equilibrium reached in 1-7 days) in a low organic carbon (0C) (0.157
%) sandy sediment but that a large fraction of naphthalene (over 40 %) was not removed
by successive desorption steps. Desorption hysteresis and incomplete desorption have
also been observed in long contaminated field soils (Steinberg, et al., 198 7; Scribner, et
al., 1992; Connaughton, et al., 1993). '

Despite a great deal of empirical evidence that soil-chemical contact times (aging)

can affect the sorption/desorption process, there are relatively few long-term, laboratory
controlled aging studies which explicitly measure sorption and desorption parameters.
Most long-term aging studies have utilized field-contaminated soils. Steinberg et al.
(1987) found that 1,2-dibromoethane (EDB) persisted in an agricultural field up to 19
years after its last known application. This is despite the fact that EDB is volatile (vapor
pressure = 13.8 mm Hg), moderately water-soluble (water solubility = 4250 mg/L), has
low affinity for soils, and is degraded rapidly by soils organisms under both aerobic
(Pignatello, .I.J., 1986) and anaerobic conditions (Castro and Belser, 1968). In the
majority of these field studies, freshly added chemicals are "spiked" into clean soils and
parameters such as extractability (Hatzinger and Alexander, 1995), biodegradability
(Steinberg, et al., 198 7; Hatzinger and Alexander, et al., 1995), bioavailability (Scribner,
et al., 1992), or desorption rate (Steinberg, et al., 1 98 7; Scribner, et al., 1992;
Pavlostathis and Mathavan, 1992; Pavlostathis and Jaglal, 1991) are measured for both
freshly added and field-aged chemicals. Conclusions in these types of studies have been
consistent, demonstrating that field-aged chemicals all have reduced extractability,
bioavailability, biodegradability, and a reduced extent of desorption. However,
information in field studies can be limited because of uncertainty of the total chemical
mass involved (Harmon and Roberts, 1994) and uncertainty (in some cases) about
soil/chemical contact times.

Attempts to reproduce results from field studies in long term, laboratory
controlled aging studies have been inconclusive. Furthermore, there are very few long
term (> 3 month), laboratory controlled aging studies which explicitly measure sorption

and desorption parameters. In a seven-day, laboratory aging study, Wu and Gschwend

(1986) found that sorption kinetics were controlled by an intraparticle diffusion
mechanism. Their results indicated that sorption and desorption should be completely
reversible given the appropriate time scale in accordance with a reversible diffusive
exchange mechanism. Another study (Connaughton, et al., 1993) using aged (3 months)
and relatively "unaged" (3 day) naphthalene in soil also suggested a completely reversible
diffusion mechanism, although reversibility was not seen on the time scale of their
experiment. The desorption of hexachlorobenzenes aged on sediments from days to
weeks was found to be completely reversible (Karickhofir and Morris, 1 985). There was
no irreversible fraction observed even for samples aged for up to 28 days. Other studies
have demonstrated that sorption is a slow process which may take hundreds of days to
reach equilibrium (Ball and Roberts, 1991; Pignatello and Xing, 1996). The studies
mentioned above would implicate a diffiisive transport mechanism for the
sorption/desorption behavior of organic chemicals in soils and tend to contradict
observations from field studies.

Several other laboratory studies document the irreversibility of contaminant
sorption by soils. F u at al. (1994) found that sorption of PAH's (polynuclear-aromatic
hydrocarbons) in low 0C (0.157 %) surface sediments was rapid (reached in 1-7 days)
and that desorption was not reversible even for samples that were aged for short time
periods (1 or 7 days). They found that the irreversible (non-desorbable) fraction was 30—
50 % of the adsorbed amount. According to calculations based on diffusion coefficients,
sorption which takes place for days may take years for desorption to be completed (Kan,
et al., 1994). Pavlostathis and Jaglal (1991) also observed that even short-term (7 days)

sorption / desorption was not reversible. Numerous other investigators have found that

the irreversible fi'action of a chemical increased with increased aging time (Pavlostathis
and Mathavan, 1992; McCall and Agin, 1985; Pignatello, 1 990(1). Results from these
studies are consistent with observations from field studies, but in contrast to results from
studies demonstrating reversibility of the sorption/desorption process. Furthermore,
another study found that aging had little effect on the desorption curves of
tetrachloroethene and l,2-dibromo-3-chloropropane from aquifer sediments (although
aging was for only 5 and 30 days) (Pignatello, 1991). However, it is not possible to say
whether aging affected the desorption rate or not, since this parameter was not measured.
In contrast, it has also been observed that aging does effect the desorption rate of
picloram in soil where aging was done for 0, 28, and 200 days (McCall and Agin, I 985).
It is clear from the above examples that the observed effects of aging on contaminant
dynamics in laboratory studies have been inconclusive. The studies conducted to this
point in time have found no consistent effects of aging and have not identified a causative
mechanism for the observed effects.

Results of the studies cited above have led to the formulation of conceptual
models to aid in the understanding of the processes involved in the sorption/desorption
behavior of organic chemicals in soils. There are several excellent reviews of these
models available (Pignatello and Xing, 1996; Pignatello, 1989; Hatzinger and
Alexander, 1995), so they will be discussed only briefly here. The first model, the
sorption-retarded pore diffusion model (SRPD) (Pignatello and Xing, I996), assumes
complete sorptive and desorptive reversibility based on an intraparticle diffusion
mechanism. Movement of a chemical into and out of an idealized, spherical sorbent (soil

particle) is highly retarded due to microporosity and tortuosity of the sorbent (Wu and

Gschwend, 1986). Movement of the chemical may be slowed by partitioning into and out
of organic matter on the pore walls. Another model, the organic matter diffusion (OMD)
model assumes that the sorbate interacts mainly with the organic matter in the sorbent
and it diffuses through the organic matter polymer, which is viewed as a lattice-like
structure (Brusseau and Rao, 1989a; Pignatello and Xing, 1996). The organic polymer
may be in a condensed (glassy) or relaxed (rubbery) polymeric state, with sorbate
diffusion being faster in the relaxed state.

It is also possible that sorption and desorption behavior is the result of a surface
adsorption process in which strong chemical or physical bonds may take place between
the sorbate and the soil particle surface (Hatzinger and Alexander, 1995). The soil
surface may be the outer surfaces of clays, interlayer spaces of clays, or the surface of
organic matter. Crucial to this model is a determination of the surface area of organic
matter, but standard methods for this process have been questioned recently. lsaacson
and F rink (1984) proposed that the irreversible fraction of phenols and some substituted
phenols may be the result of hydrogen bonding of these compounds to humic material.
Other researchers have also implicated specific adsorption interactions with soil
materials, although these types of interactions are limited to ionic compounds and those
with polar functional groups (Nearpass, 1976; Hayes, 1970; Li and F elbeck, 1972; Boyd
and Mikesell, 198 7). Recently, some experimental evidence has suggested that surface
adsorption models may be able to explain sorption/desorption behavior in soils (Kan, et
al., 1994). Since it has been demonstrated that surface adsorption can lead to hysteresis
(Adamson, 1990) even on model sorbents, it seems reasonable to consider a surface

adsorption model along with the SRPD and OMD conceptual models.

There is also some evidence that the sorption of organic contaminants may cause
conformational changes in the structure of soil humic and fulvic acids. Evidence
indicates that this psuedomicellar behavior of humic and fulvic acids has the ability to
sequester hydrophobic species in the solution environment (Engebretson, et al., 1996;
Engebretson and von Wandruszka, 1994). Physical entrapment due to structural changes
would explain the existence of a desorption resistant fraction often observed in
sorption/desorption studies. Computational molecular modeling approaches have
demonstrated that organic contaminants such as atrazine can be trapped in structural
voids of humic acids by means of H-bonding (Schulten, 1995). The importance of the
physical structure of SOM and its constituents and its role in organic contaminant
entrapment in the environment has not been studied in-depth and needs further analysis.

There are also several mathematical models, both mechanistically and non-
mechanistically based, which have been used in an attempt to predict sorption and
desorption behavior. The two most widely used mathematical models are the two-box
model and the radial diffusion model. The two-box model (Cameron and Klute, 1977)
assumes that sorption/desorption occurs on "fast" and "slow" sites in the soil (see Figure
1). The model describes two possible compartment configurations, a series version
(Co <=> S1 a S2) and a parallel version (S14: Co <=> S2). Although mathematically
identical, the series version makes more physical sense because the S1 compartment can
be conceptualized as the exterior of the sorbent and the S2 compartment as the interior of
the sorbent. The rate constants describe fast sorption (K) or slow desorption (kd) between
the compartments. The variables S1 and S2 give the fraction of chemical in each

compartment. This model makes no assumptions about the mechanisms of soil/chemical

interactions. The radial diffusion model (Wu and Gschwend, 1986) assumes the sorbent

is a perfect sphere where movement of chemical into the sorbent is governed by retarded

 

 

 

K

r-A

   
 

   
 

Sr 7-1 52

kd

Co

 

 

 

 

 

 

 

Dissolved Labile Non-Labile
Figure 1.1. Schematic diagram of the Two-Box Model (from Guerin and Boyd, 1993).

diffusion. This retarded diffusion is due to movement of the sorbate through soil
micropores and / or microscale partitioning of the sorbate into and out of organic matter
on pore walls. This model uses only on fitting parameter, the effective diffusion
coefficient (Dcff) where:
Dcff= Dm*n / ((1-11) "‘ Ps "' Kp)

Dm = pore fluid diffusivity of the sorbate, n = porosity of the sorbent, p5 = specific gravity
of the sorbent, Kp = equilibrium sorption coefficient of the specific soil/solute pair, and
(l-n) = the mass of the sorbent. This model assumes a mechanism of interaction between
the soil and sorbate, and in order for the model to be correct, sorption and desorption
should be completely reversible. Although other mathematical models have been
employed to predict sorption/desorption behavior, most are variations of the ones
discussed above.

There are very few studies that have explicitly analyzed the role of soil organic

matter in the aging process. Given the dominant role of SOM with respect to organic

contaminant interactions with soil, it seems reasonable to expect that the effects of
contaminant aging may be different in soils with different SOM content, especially if the
DMD model is valid. Although no studies have examined the role of SOM and aging
explicitly, scrutiny of the literature demonstrates that there may be a correlation between
SOM content and the effects of aging. In a review paper on aging, Loehr and Webster
(1996) thought that the effect of aging time on sorption and desorption may be more
pronounced in soils with high organic carbon content.

With respect to the forward process, sorption, some investigators have found that
equilibrium does not appear to be reached quickly (in 24 hours or less) while in other
cases, equilibrium appears to be approximated within this time frame. Ball and Roberts
(1991) investigated the long-term sorption of tetrachloroethene and tetrachlorobenzene
on very low organic carbon (0C) (0.021 %) aquifer materials and found that equilibrium
may take hundreds of days to be established. Another study looked at the sorption of
PAH's on relatively low OC surface sediments (0.157 %) and found that sorption was
complete in one day (F u, et al., 1994). It is not clear whether the differences in OC
content between sorbents in these studies were significant enough to account for the
observed behavior. Pignatello and Xing(1996) reviewed slow sorption and desorption
behavior and found that sorption increased between 1.3 and 10 times for long and short
aging times, although there appeared to be little correlation to SOM content. The same
two investigators found that sorption isotherms may become more non-linear with aging
time (between 1 and 180 days) (Xing and Pignatello, 1996). They also found that for
dichlorophenol (DCP) the largest relative increase in the sorption coefficient (K) between

1 and 180 days occurred in soil (1.74 % 0C) rather than in peat (54.07 % 0C) (Xing and

Pignatello, 1996).

For the reverse process, desorption, some investigators have noticed the
relationship between a larger resistant (non-desorbable) fraction and increased SOM
(Pignatello, 1990b; Carroll, et al., 1994) for chemicals aged in soils. At least one
investigator found that the rate and extent (which determines the resistant fraction) of
desorption of aged trichloroethylene, toluene, and xylene were not correlated to SOM
content, nor to specific surface area or water solubility (Pavlostathis and Mathavan,

I 992). In one of the most comprehensive studies done on aging (McCall and Agin,
1985), there appeared to be a greater effect on Kdapp (the apparent equilibrium
distribution coefficient) of picloram aged in low OC soils (0.45 %) than in high 0C soils
(2.9 %). The Kdapp was measured by a 2 minute desorption protocol after aging
picloram on these soils for 0 and 300 days. In the low 0C soil the Kdapp increased 5.4
times, but only 1.6 times in the high OC soil. Also measured were Kdapp values after
sorption for 100 days and for desorption (after 100 days of sorption), the difference in
these values represents the degree of hysteresis between a single sorption and desorption
equilibrium point after 100 days. The difference in the values for the high OC soil was
only 1.08 times (effectively the same, no hysteresis) whereas in the low OC soil the
difference was 11.25 times. These changes indicate that the effect of aging is greater in
low 0C soils than in high 0C soils. This may be in contrast to what some researchers
have expected, but it seems to agree with the results found by Xing and Pignatello (1996)
and their study with DCP on peat and soil. Furthermore, assuming that the 7-8 fold
difference in OC content was the reason for the discrepancy in the sorption results found

by Ball and Roberts (1991) and Fu et al. (1994), the correlation between lower 0C

10

content and larger aging effects would explain their results as well.

The study discussed in the following chapter was initiated because of the lack of
long term, laboratory controlled aging studies explicitly measuring sorption and
desorption parameters and because of the need to look at the role of SOM content in the
aging process. In this study, the sorption isotherms of chlorobenzene (CB) on four soil
types with a range of SOM content (from 0.69 to 13.42 %) were measured after 24 hours
and 14 months of aging. Then, desorption kinetics were measured for chlorobenzene on
each soil after the sorption times mentioned above, with desorption being followed for up
to four months. This was done in an effort to examine the effect of aging on one
compound in several soils to establish the role of SOM content in the aging process.

There has been very little work examining the role of sorbate properties in the
aging process and how this may affect sorption/desorption behavior. Several researchers
have observed that increasingly hydrophobic compounds (as measured by Kp , the
equilibrium partition coefficient) exhibit slower sorption/desorption behavior in soils
(Karickhofl and Morris, 1 985; Brusseau and Rao, [98%). In contrast, Oliver (1985)
found that desorption behavior was largely independent of adsorbate properties even for a
wide variety of aged non-polar organics in sediments. These compounds exhibited nearly
a four order of magnitude range in hydrophobicity, as measured by K0“, (octanol-water
partition coefficient). If hydrophobicity affects desorption as observed by Karickhoff and
Morris (1985), it seems likely that these differences would have been observed in the
aged sediments by Oliver (1985) given the range in hydrophobicity of the solutes.
Another study (Staufler and Maclntyre, 1986) tried unsuccessfully to correlate sorption

coefficients to sorbate dipole moments. In their discussion, these investigators stated that

11

sorbate polarizability might be a better parameter with which to predict sorption
behavior. They did find a positive correlation between sorbate molar volumes and
sorption coefficients. The aforementioned studies only compared behavior of non-polar
organics, which are largely unreactive with soil components. Their behavior in soil is
expected to be governed by a single mechanism, solute partitioning into SOM (Chiou, et
al., 1979).

Several other studies have examined the aging behavior of both polar (or
compounds with polar functional groups) and non-polar organic compounds. Xing and
Pignatello (1996) examined sorption isotherms of dichlorobenzene (non-polar) and
dichlorophenol (polar) in soil after 0 and 30 days of aging and found that the Kp increased
only 1.3 times for dichlorobenzene (DCB; log KP = 3.38) but 2.7 times for dichlorophenol
(DCP; log Kp = 2.75). This would suggest that the non-polar compound, although
slightly more hydrophobic, had faster sorption kinetics than the polar compound and
therefore approached equilibrium faster. These results are in agreement with the results
of another study (Brusseau and Rao, I989a) in which linear free energy relationships
(LFER's) were observed between sorption rate coefficients (log k) and sorption
equilibrium coefficients (log KP) for both polar and non-polar compounds. For polar
(Type II) and non-polar (Type I) compounds with similar partition coefficients, the polar
compounds exhibited substantially slower sorption/desorption kinetics. They attributed
this to the increased reactivity of polar compounds or compounds with polar functional
groups. The polar compounds must diffuse through organic matrices (a physical non-
equilibrium process or PNE) and are subject to specific chemical interactions due to their

polar functional groups (a chemical non-equilibrium process or CNE), whereas the

12

relatively unreactive non-polar compounds are subject only to the former process. When
the results of Xing and Pignatello (1996) are viewed in the context of these findings, it is
reasonable to expect that DCP would exhibit a larger increase in Kp upon aging than
DCB. Based upon the data of Brusseau and Rao, the relatively similar log Kp values for
DCB and DCP would translate to a log k value for DCP that is two orders of magnitude
slower that that for DCB. So it is apparent, from this data, that DCB would approach
equilibrium much more rapidly than DCP.

Brusseau and Rao (I989a) presented evidence that polar compounds are rate-
lirnited due to diffusion (PNE) and CNE whereas non-polar compounds are rate limited
solely by diffusion. Polar compounds, such as atrazine, picloram, and chlorophenols,
which have reactive fimctional groups, are known to form specific interactions with soil
(Hayes, 1970; Li and Felbeck, 1972; Nearpass, 1976). Given the temporal nature of
bond formation, polar compounds which interact by dual mechanisms (at least) may
demonstrate dramatic sorption and desorption differences from non-polar organics, which
have a single and general sorption mechanism, if enough aging time is allowed for these
specific interactions to form. The literature indicates some evidence of this occurrence.
Winkelrnan and Klaine (1991) found that bound residues of atrazine in soil microcosms
increased with increased aging time. Ma and Selim (1994) also found that sorption and
desorption behavior of atrazine was highly hysteretic and this hysteresis increased with
aging time. They also found that atrazine had not reached sorption equilibrium even after
22 days. The same behavior was observed for alachlor (Xue and Selim, 1995). Alachlor
was characterized in this study as a moderately non-polar compound, but they stated that

sorption of alachlor occurred by at least two types of mechanisms: hydrophobic binding

13

to aliphatic parts of humic acid (Senesi, et al., 1994) and by coordination binding to clay
and organic matter surfaces (Bosetto, et al., 1993). Alachlor has also been shown to
penetrate the interlayer space of montrnorillonite (Bosetto, et al., 1993).

In contrast, Boesten and van der Pas (1988) found no sorption/desorption
hysteresis of cyanizine (a triazine herbicide similar to atrazine) and metribuzin, both
polar compounds, in soil suspension. Interestingly, this study incorporated no long term
aging component, as sorption and desorption were followed for only 24 hours. This may
account for the lack of hysteresis observed in this study, in contrast to the two previously
mentioned studies, in which aging was performed for up to 22 days (Ma and Selim, 1994;
Xue and Selim, 1995). Another study on atrazine (Moreau and Mouvet, 1997) found that
atrazine sorption in soil (0.65 % 0C) was complete after 72 hours and that desorption
was completely reversible. One possible reason for this conclusion was that sorption was
not followed for longer than 96 hours, in which case slow approach to equilibrium would
not have been observed. Furthermore, desorption may have appeared to be reversible if
sufficient time (only 3-4 days in this case) was not allowed for bound residues to form, as
may have occurred in the 22 day aging studies. Conclusions fi'om these studies would
suggest that aging time may have more of an effect on the dynamics of polar compounds,
which interact with soils by two or more mechanisms, than for non-polar compounds.

In an attempt to better understand the effects of aging on the dynamics of polar
and non-polar compounds in soils, the study in chapter 3 was undertaken to explicitly
measure sorption and desorption parameters for chlorobenzene, EDB, atrazine, and 2,4-D
on a Capac A soil. Sorption isotherms and desorption kinetics were measured after 24

hours, 1 month, and 14 months of aging for each compound, with desorption being

14

followed for up to four months. This study was initiated because of the lack of long term,
laboratory controlled aging studies and because of the need to examine the role of sorbate
prOperties in the aging process.

Sorption/desorption behavior of organic chemicals in soils may be the most
important factor governing their bioavailability. Bioavailability of organic contaminants
to bacteria is an important determinant of the contaminants' environmental persistence.
Bioavailability also determines the toxicological and herbicidal activity of chemicals (i.e.,
pesticides) in the environment. Sorption to soils is known to play a role in bioavailability
by limiting biodegradation of a chemical in the environment (Wszolek and Alexander,
1979; Steen et al., 1980). Most researchers have assumed that before a sorbed molecule
can be taken up by a bacterium, it first has to diffuse to the solid-liquid interface, partition
into the water, and then must diffuse to the bacterium through the aqueous phase (Harms
and Zehnder, 1995). This assumption appears to be correct based on studies by Ogram,
et al. (1985) and Steen et al. (1980), although a study by Guerin and Boyd (1992)
concluded that generalizations regarding availability of sorbed contaminants are
inappropriate. Guerin and Boyd (1992) demonstrated that different bacterial species have
different abilities to use substrates and may, in fact, be able to use sorbed substrates. This
may be due to organism specific properties such as attachment, motility, or affinity for
the particular substrate.

If bacteria can only use dissolved substrates, sorption can limit bioavailability in
two ways: 1) decreasing aqueous concentrations to low levels due to sorption or 2)
limiting high theoretical biodegradation rates due to small desorption rates (a desorption

limited process) (Harms and Zehnder, 1995). These factors will limit the initial

15

biodegradation rate but not the extent biodegraded, assuming bacteria are able to enhance
desorption. Several studies have concluded that bacteria are able to enhance desorption,
which means that desorption is slow relative to biodegradation and that biodegradation
increases desorption. Harms and Zehnder (1995) found that 3-chlorodibenzofuran
desorption was driven (enhanced) by cells attached to Teflon sorbents. They concluded
that bioavailability was dependent on the specific affinity of a bacteria for a specific
substrate and the tendency of that organism to adhere to a surface. Rijnaarts, et al.,
(1990) also found that microbes enhanced desorption and that initial rates of
biodegradation were consistently higher than initial desorption rates. The study of
Guerin and Boyd (1992) found that one bacterial species was able to enhance desorption
whereas another species was not.

An area where more research is needed is studying the effect of aging (soil-
chemical contact time) on the bioavailability of contaminants in soils. Since many field
soils have been contaminated for periods of months to years, conclusions from short-term
bioavailability studies may be inappropriate for long-term contaminated sites.
Observations from field studies have demonstrated that aged chemicals are less
bioavailable than fieshly added ones (Steinberg, et al., 198 7; Weissenfels, et al., 1992;
Scribner, et al., 1992). The study by Weissenfels, et al. (1992) demonstrated that the
strength of PAH sorption to soils, which may increase with aging, decreased the amount
of PAH degraded and that freshly added chemicals do not accurately mimic the behavior
of aged chemicals. They also found that the aged PAH's on a high OC soil (13.6 %) were
less bioavailable than aged PAH's on a low OC soil (1.0 %). The extent of freshly added

phenanthrene mineralized was found to be greater than the extent mineralized for aged

16

phenanthrene in a muck soil (11.22 % OC) (Hatzinger and Alexander, 1995). However,
no difference was observed between the extent mineralized for freshly added (0 days) and
aged (84 days) phenanthrene in a loamy soil (2.33 % OC). Rates of degradation were
decreased for aged phenanthrene on both the loam and the muck. In a sandy soil (1.34 %
OC), they found no difference in the rate or extent of degradation for aged versus freshly
added chemical. 4-Nitrophenol was added to the loam and muck soils and mineralization
was assessed at 0, 40, and 103 days. In both soil types, rates and extents of
mineralization decreased with increased aging time. These results show the effect of
aging on biodegradation and would seem to indicate that aging has more of an effect on
bioavailability in high OC soils than in low OC soils, in agreement with the results found
by Weissenfels (1992).

In a study by Guerin and Boyd (1993) on the bioavailability of naphthalene,
increased aging time (up to 300 days) decreased initial mineralization rates. Although a
portion of the sorbed naphthalene pool was initially available for mineralization, access
to this material diminished as aging time increased. Guerin and Boyd (1993) also
observed that sorbed naphthalene appeared to be more available to bacteria in low OC
soils than in high OC soils. Using the two-box model, they attributed this to a larger
fraction of the compound residing in a non-labile state in the high OC soil as compared to
the low OC soil. Only one study has explicitly measured both desorption and
biodegradation rates for freshly added and aged chemicals (Carmichael, et al., 199 7).
Desorption rates of chrysene and phenanthrene (both PAH's) were found to be much
faster than observed mineralization rates for freshly added compounds. They also found

that desorption rates for field-aged chemicals were equal to or slower than mineralization

l7

rates, so it was concluded that desorption of aged PAH's may control their fate and
availability. This may explain why the persistence of PAH's has been observed even in
soils that have active populations of PAH-degrading organisms.

In order to better understand the effects of aging on bioavailability and on the fate
of chemicals in soils, we initiated a long term aging study explicitly examining sorption,
desorption, and biodegradation in soils (see chapter 4). In this study, we measured
sorption isotherms and desorption rates for biphenyl on Capac A and B soils horizons
after 24 hours and 8 months of aging. We also measured biphenyl degradation rates and
extents on these soils using a Psuedomonas species capable of completely mineralizing
biphenyl in order to observe the effects of aging on bioavailability. The parameters for
the biological and chemical assays were then compared in an effort to better understand

the effects of aging on both biological and physico-chemical processes.

List of Referen ces

Adamson, A.W. 1990. Physical Chemistry of Surfaces. John Wiley & Sons, New
York.

Alexander, M. 1994. Biodegradation and Bioremediation. Academic Press, New York.

Ball, W.P. and Roberts, P.V. 1991. Long-term sorption of halogenated organic
chemicals by aquifer material. 1. Equilibrium. Environ. Sci. T echnol. 25:1223-1236.

Boesten, J.J.T.I. and van der Pas, L.J.T. 1988. Modeling adsorption/desorption
kinetics of pesticides in a soil suspension. Soil Sci. 146:221-231.

Bosetto, M., Arfaioli, P., and Fusi, P. 1993. Interactions of alachlor with homoionic
montrnorillonites. Soil Sci. 155:105-113.

Boyd, S.A. and Mikesell, MD. 1987. Reactions of chlorophenols in soils. P. 272.
Agron. Abst. ASA, Madison, WI.

Brusseau, M.L. and Rao, P.S.C. 1989a. The influence of sorbate-organic matter

18

interactions on sorption non-equilibrium. Chemo. 18: 1691-1706.

Brusseau, M.L. and Rao, P.S.C. 1989b. Sorption non-ideality during organic
contaminant transport in porous media. CRC Crit. Rev. Environ. Control. 19:33-99.

Cameron, D.A. and Klute, A. 1977. Convective-dispersive solute transport with a
combined equilibrium and kinetic adsorption model. Water Resour. Res. 13:183-188.

Carmichael, L.M., Christman, RE, and Pfaender, F.K. 1997. Desorption and
mineralization kinetics of phenanthrene and chrysene in contaminated soils. Environ.
Sci. T echnol. 31 : 126-132.

Carroll, K.M., Harkness, M.R., Bracco, A.A., Balcarcel, RR. 1994. Application of a
permeant/polymer diffusional model to the desorption of polychlorinated biphenyls from
Hudson river sediments. Environ. Sci. T echnol. 28:253-258.

Castro, CE. and Belser, ND. 1968. Biodehalogenation. Reductive dehalogenation of
the biocides ethylene dibromide, 1,2-dibromo-3-chloropropane, and 2,3-dibromobutane
in soil. Environ. Sci. Technol. 2:779-783.

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

Connanghton, D.F., Stedinger, J.R., Lion, L.W., and Shuler, M.L. 1993. Description
of time-varying desorption kinetics: Release of naphthalene from contaminated soils.
Environ. Sci. T echnol. 27:2397-2403.

Engebretson, RR. and von Wandruszka, R. 1994. Microorganization in dissolved
humic acids. Environ. Sci. T echnol. 28:1934-1941.

Engebretson, R.R., Amos, T., and von Wandruszka, R. 1996. Quantitative approach
to humic acid associations. Environ. Sci. Technol. 30:990-997.

Fu, 6., Kan, A.T., Tomson, MB. 1994. Adsorption and desorption hysteresis of
PAH’s in surface sediment. Environ. T oxicol. Chem. 13:1559-1567.

Guerin, W.F. and Boyd, S.A. 1992. Differential bioavailability of soil-sorbed
naphthalene to two bacterial species. Appl. Environ. Microbiol. 58:1142-1152.

Guerin, W.F. and Boyd, S.A. 1993. Bioavailability of sorbed naphthalene to bacteria:
Influence of contaminant aging and soil organic carbon content. p. 197-208. In Sorption
and Degradation of Pesticides and Organic Chemicals in Soil. SSSA special publication
no. 32. SSSA, Madison, WI.

Harmon, T.C. and Roberts, P.V. 1994. Comparison of intraparticle sorption and
desorption rates for a halogenated alkene in a sandy aquifer material. Environ. Sci.

19

T echnol. 28: 1650-1660.

Harms, H. and Zehnder, A.J.B. 1995. Bioavailability of sorbed 3-chlorodibenzofiiran.
Appl. Environ. Microbial. 61:27-33.

Hatzinger, RB. and Alexander, M. 1995. Effect of aging of chemicals in soil on their
biodegradability and extractability. Environ. Sci. T echnal. 29:537-545.

Hayes, M.H.B. 1970. Adsorption of triazine herbicides on SOM, including a short
review on SOM chemistry. Res. Review. 32:131-174.

Issacson, P.J. and Frink, CR. 1984. Nonreversible sorption of phenolic compounds by
sediment fractions: The role of sediment organic matter. Environ. Sci. T echnol. 18:43-
48.

Kan, A.T., Fn, G., and Tomson, MB. 1994. Adsorption/desorption hysteresis in
organic pollutant and soil/sediment interaction. Environ. Sci. T echnal. 28:859-867.

Karickhoff, S.W. 1980. Sorption kinetics of hydrophobic pollutants in natural
sediments. P. 193-205. In R.A. Baker (ed) Contaminants and Sediments, Vol. 2. Ann
Arbor Sci. Publ., Ann Arbor, MI.

Karickhoff, S.W. and Morris, KR. 1985. Sorption dynamics of hydrophobic
pollutants in sediment suspensions. Environ. T axical. Chem. 4:469-479.

Li, G. and Felbeck, G.T. 1972. A study of the mechanism of atrazine adsorption by
hurrric acid from muck soil. Soil Sci. 113:140-148.

Loehr, RC. and Webster, M.T. 1996. Behavior of fresh vs. aged chemicals in soil. J.
Soil Contam. 5:361-383.

Ma, L. and Selim, H.M. 1994. Predicting atrazine adsorption-desorption in soils: A
modified second-order kinetic model. Water Resour. Res. 30:447-456.

McCall, P.J. and Agin, G.L. 1985. Desorption kinetics of picloram as affected by
residence time in the soil. Environ. T axical. Chem. 4:37-44.

Morean, C. and Monvet, C. 1997. Sorption and desorption of atrazine, deethylatrazine,
and hydroxyatrazine by soil and aquifer solids. J. Environ. Qua]. 26:416-424.

Nearpass, D.C. 1976. Adsorption of picloram by humic acids and humin. Soil Sci.
121 :272-277.

Ogram, A.V., Jessup, R.E., On, L.T., and Rao, P.S.C. 1985. Effects of sorption on

biological degradation rates of (2,4-dichlorophenoxy)acetic acid in soils. Appl. Environ.
Microbial. 49:582-587.

20

Oliver, B.G. 1985. Desorption of chlorinated hydrocarbons from spiked and
anthropogenically contaminated sediments. Chemo. 14:1087-1106.

Pavlostathis, S.G. and Jaglal, K. 1991. Desorptive behavior of trichloroethylene in
contaminated soil. Environ. Sci. T echnal. 25:274-279.

Pavlostathis, S.G. and Mathavan, G.N. 1992. Desorption kinetics of selected volatile
organic compounds fiom field contaminated soils. Environ. Sci. T echnol. 26:532-538.

Pignatello, J.J. 1986. Ethylene dibromide mineralization in soils under aerobic
conditions. Appl. Environ. Microbial. 51 :588—592.

Pignatello, J.J. 1989. Sorption dynamics of organic compounds in soils and sediments.
P. 45-79. In Sawhney, B.L., Brawn, K. (eds). Reactions and Movement of Organic
Chemicals in Soils. SSSA special publication no. 22. SSSA, Madison, WI.

Pignatello, J.J. 1990a. Slowly reversible sorption of aliphatic halocarbons in soils. 1.
Formation of residual fi'actions. Environ. T axical. Chem. 9:1107-1115.

Pignatello, J.J. 1990b. Slowly reversible sorption of aliphatic halocarbons in soils. II.
Mechanistic aspects. Environ. Toxicol. Chem. 9:1117-1126.

Pignatello, J .J . 1991. Desorption of tetrachloroethene and 1,2-dibromo-3-
chloropropane from aquifer sediments. Environ. Toxicol. Chem. 10:1399-1404.

Pignatello, J.J. and Xing, B. 1996. Mechanisms of slow sorption of organic chemicals
to natural particles. Environ. Sci. T echnal. 30:1-11.

Rijnaarts, H.H.M., Bachmann, A., Jumelet, J.C., Zehnder, A.J.B. 1990. Effect of
desorption and intraparticle mass transfer on the aerobic biomineralization of alpha-

hexachlorocyclohexane in a contaminated calcareous soil. Environ. Sci. T echnal.
24:1349-1354.

Schulten, H.R. 1995. The three-dimensional structure of humic substances and soil
organic matter studied by computational analytical chemistry. Fresenius J. Anal. Chem.
3 5 1 :62-7 3.

Scribner, S.L., Benzing, T.R., Sun, S., and Boyd, S.A. 1992. Desorption and
bioavailability of aged sirnazine residues in soil from a continuous corn field. J. Environ.
Qual. 21:115-120.

Senesi, N., Brnnetti, G., Cava, P.L., Miano, T.M. 1994. Adsorption of alachlor by
humic acids fi‘om sewage sludge and amended and non-amended soils. Soil Sci.
157:176-184.

Stauffer, TB. and Maclntyre, W.G. 1986. Sorption of low-polarity organic

21

compounds on oxide minerals and aquifer material. Environ. T oxical. Chem. 5:949-955.

Steen, W.C., Paris, D.F., and Banghman, G.L. 1980. Effects of sediment sorption on
microbial degradation of toxic substances. p. 477-482. In R.A. Baker (ed. ) Contaminants
and Sediments, Vol. I . Ann Arbor Sci. Publ., Ann Arbor, MI.

Steinberg, S.M., Pignatello, J.J., and Sawhney, B.L. 1987. Persistence of 1,2-
dibromoethane in soils: Entrapment in intraparticle nricropores. Environ. Sci. T echnol.
21:1201-1207.

Weissenfels, W.D., Klewer, H.J., and Langhoff, J. 1992. Adsorption of polycyclic
aromatic hydrocarbons (PAH's) by soil particles: Influence on biodegradability and
biotoxicity. Appl. Microbiol. Biotechnol. 36:689-696.

Winkelman, D.A. and Klaine, SJ. 1991. Degradation and bound residue formation of
atrazine in a westem Tennessee soil. Environ. T axical. Chem. 10:335-345.

Wszolek, RC. and Alexander, M. 1979. Effect of desorption rate on the
biodegradation of n-alkylamines bound to clay. J. Agric. F oad Chem. 27:410-414.

Wu, S. and Gschwend, RM. 1986. Sorption kinetics of hydrophobic organic
compounds to natural sediments and soils. Environ. Sci. T echnol. 20:717-725.

Xing, B. and Pignatello, J.J. 1996. Time-dependent isotherm shape of organic
compounds in soil organic matter: Implications for sorption mechanism. Environ.
T axical. Chem. 15: 1282-1288.

Xue, S.K. and Selim, H.M. 1995. Modeling adsorption-desorption kinetics of alachlor
in a Typic Fragiudalf. J. Environ. Qua]. 24:896-903.

22

Chapter 2

EFFECTS OF AGING ON THE SORPTION/DESORPTION OF CHLOROBENZENE

ON SOILS WITH A RANGE OF ORGANIC CARBON CONTENT

Abstract

Soil-organic chemical contact time (aging) has been found to cause increased
persistence of these compounds in the environment. There are few studies that have
tested the effects of long-term aging and the role of organic matter content on
contaminant sorption/desorption behavior and fate in soils. In this study, the
sorption/desorption behavior of chlorobenzene on four soil types was evaluated after
aging for periods of 24 hours and 14 months. The four soil types ranged in organic
carbon content from 7.8 to 0.4 %. Sorption isotherms were prepared and evaluated after
each aging period to observe changes in the uptake of chlorobenzene. Desorption kinetic
profiles were generated after each aging period to observe changes in chlorobenzene
release from soil, with desorption followed for up to four months. The data presented
here demonstrates no increase in the uptake of chlorobenzene on 3 of 4 soils after 24
hours and 14 months of aging, indicating that equilibrium was reached quickly (24 hours
or less). Initial desorption rates appear to be similar despite the differences in aging time,
and desorption equilibrium was not reached on any soil type, indicating a large non-
desorbable fraction. This non-desorbable fraction increased for all soil types with longer
aging time.

Keywords — Sorption, desorption kinetics, chlorobenzene, aging, soils

23

Introduction

There are many studies which indicate that sorption and desorption in soils are not
rapid, reversible processes, despite the assumption by many researchers that this is the
case. Deviations from this assumption could have major effects on contaminant transport
and fate. A study by Ball and Roberts (1991) indicated that sorption of
tetrachlorobenzene was a slow process and could take hundreds of days to reach
equilibrium, and that sorption coefficients on a low organic carbon (OC) aquifer material
were an order of magnitude greater than those predicted by partition theory (Chiou,
1979). Similarly, Pignatello and Xing (1996) showed that equilibrium sorption
coefficients (KP) may increase between 1.3 and 10 times for long and short soil-chemical
contact times (aging) and that sorption isotherms may become more non-linear with
increased aging (Xing and Pignatello, 1996). Data from field studies have found that
certain compounds (e. g. EDB, sirnazine) demonstrate unexpected persistence in soils
compared to laboratory studies of binding and degradability in freshly spiked samples
(Steinberg, et al., 1987; Scribner, et al., 1992). The formation of such a protected
fraction was attributed to long pesticide-soil contact times. Numerous other investigators
have found that the irreversible (non-desorbable) fraction of a chemical increased with
increased aging (Pavlostathis and Mathavan, 1992; McCall and Agin, 1985; Pignatello,
1990(1). Assuming these non-desorbable chemicals are biologically unavailable, this may
explain their measured persistence. From these examples, it seems apparent that soil-
chemical contact time (aging) does have an effect on sorption/desorption behavior, hence
fate, and that these processes are not rapid and reversible. The specific effects of aging

time on contaminant behavior, however, are not well defined or understood.

24

Despite a great deal of empirical evidence that aging can affect the
sorption/desorption process, there are relatively few long term, laboratory controlled
aging studies which explicitly measure sorption and desorption parameters. Most long-
tenn aging studies have utilized field-contaminated soils. Steinberg et al. (1987) found
that 1,2-dibromoethane (EDB) persisted in an agricultural field up to 19 years after its
last known application. This is despite the fact that EDB is volatile (vapor pressure =
13.8 mm Hg), moderately water-soluble (water solubility = 4250 mg/L), has low affinity
for soils, and is degraded rapidly by soil organisms under both aerobic (Pignatello, 1986)
and anaerobic conditions (Castro and Belser, 1968). In the majority of these field
studies, fi'eshly added chemicals are “spiked” into clean soils and parameters such as
extractability (Hatzinger and Alexander, 1995), biodegradability (Steinberg, et al., 198 7;
Hatzinger and Alexander, et al., 1995), bioavailability (Scribner, et al., 1992), or
desorption (Steinberg, et al., 198 7; Scribner, et al., 1992; Pavlostathis and Mathavan,
I992; Pavlostathis and Jaglal, 1991) are measured for both the freshly added and field-
aged chemicals. Conclusions in these types of studies have been consistent in that field-
aged chemicals all demonstrate reduced extractability, bioavailability, biodegradability,
and a reduced extent of desorption. However, information from field studies can be
limited because of uncertainty of the total chemical mass involved (Harman and Roberts,
1994) and uncertainty (in some cases) about soil-chcnrical contact times. Furthermore,
most aging studies have not explicitly measured sorption and desorption parameters.

There are very few, if any, studies that have explicitly analyzed the role of soil
organic matter (SOM) in the aging process. Given the dominant role of SOM as a

sorptive phase for organic contaminants in soils (Chiou, 1979), it seems reasonable to

25

expect that the effects of contaminant aging may be different in soils with different SOM
contents. Although no studies have examined directly the role of SOM and aging,
scrutiny of the literature demonstrates that there may be some relation between SOM
content and the effects of aging. In a review paper on aging, Loehr and Webster (1996)
concluded that the effect of aging time on sorption/desorption processes may be greater
in soils with high organic carbon contents. It was observed recently that for
dichlorophenol (DCP) the largest relative increase in the sorption coefficient (K) between
1 and 180 days occurred in soil (1.74 % OC) rather than in peat (54.07 % OC) (Xing and
Pignatello, 1996). For the reverse process, desorption, some investigators have noticed
the relationship between a larger resistant fraction and increased SOM (Pignatello,
I990b; Carroll, et al., 1994) for aged chemicals. In one of the most comprehensive
studies on aging (McCall and Agin, 1985), there appeared to be a greater effect on Kdapp
(the apparent equilibrium distribution coefficient) of picloram aged in low OC soils (0.45
%) than in high OC soils (2.9 %). The Kdapp was measured by a 2 rrrinute desorption
protocol after aging picloram on these soils for 0 and 300 days. In the low OC soil, the
Kdapp increased 5.4 times between 0 and 300 days, but only 1.6 times in the high OC
soil. Also measured were Kdapp values after sorption for 100 days and for desorption
(after 100 days of sorption); the difference in these values represents the degree of
hysteresis between a single sorption and desorption equilibrium point after 100 days. The
difference in the values for the high OC soil was only 1.08 times (effectively the same;
no hysteresis) whereas in the low OC soil the difference was 11.25 times. These changes
indicate that the effect of aging is greater in low OC soils than in high OC soils. This

may be in contrast to what some researchers have expected, but it agrees with the results

26

found by Xing and Pignatello (1996) and their study with DCP on peat and soil.

In this study we have conducted long term, laboratory controlled aging studies
with the objective of explicitly measuring changes in sorption and desorption parameters
with aging, and to evaluate the role of SOM content in the aging process. Sorption
isotherms of chlorobenzene (CB) on four soil types with a range of SOM content (from
0.69 to 13.42 %) were measured after 24 hours and 14 months of aging. The degree of
sorption is measured by KP, the equilibrium partition coefficient, which is determined by
the slope of the sorption isotherm. Then, desorption kinetics were measured for
chlorobenzene (CB) on each soil after the sorption times mentioned above, with

desorption being followed for up to four months.

Experimental Section

The soils used in this study were from the Colwood A horizon (fme-loamy,
mixed, mesic Typic Haplaquoll), Schoolcraft A horizon, and the Capac A and B horizons
(fine-loamy, mixed, mesic Aerie Ochraqualf). Soils were collected from areas located in
southern Michigan and air-dried, ground, and sieved (2 mm). Sand, silt, clay, and OC
content were done by the Michigan State University Plant and Soil Nutrient Lab (see
Table 2.1). All soils were then gamma irradiated with 5 MRad at the University of
Michigan Ford Nuclear Reactor Laboratory using a cobalt-60 irradiator (dose rates by
Renter-Stokes ion chamber). Soils were stored in 500-mL Nalgene bottles during
irradiation and remained sealed until use.
W The amounts of a particular soil and solution used to develop

sorption isotherms were based on preliminary sorption experiments. The linear isotherm

27

Table 2.1. Characterization and properties of the soils used in this study.

 

 

Sail ° ° ' ° y Type "/2 OC *CEC
Colwood A 64.2 20.7 15.1 sandy loam 7.8 43.0
Capac A 54.6 24.0 21.4 sandy clay loam 3.3 24.4
Schoolcraft A 58.6 22.0 19.4 sandy loam 1.4 11.2
Capac B 60.0 20.0 20.0 sandy clay loam 0.4 9.6

 

*(CEC= cmol(+)/kg)

equation (x/m = K*Ce) was employed to estimate the amount of soil and solution needed
to sorb 50 % of the added solute while also nrinirrrizing headspace volume. The tubes
used were 20-mL glass ampules that were autoclaved prior to use. All additions of soil
and solution to the ampules were carried out in a sterilized laminar flow hood. A mass of
soil was weighed and transferred to the ampule using a metal spatula along with a plastic
funnel. A 0.005 M CaCl2 solution, which had been autoclaved for 25 minutes and
amended with 200 mg/L NaN3, was added to the ampules using a sterilized repipettor.
These functions were carried out in the laminar flow hood and all items contacting soil or
solution were autoclaved prior to use (spatulas, funnels, etc.).

Additions of the l‘lC-chlorobenzene (CB) (Sigma Chemical Co., radiochemical
purity > 97 %) stock solution (in acetone) and flame sealing of the ampules were carried
out as follows. An aliquot of CB stock solution was added to the ampule using a 10 or 25
111 Hamilton syringe and the ampule was immediately flame sealed using an Ampulmatic
flame sealer unit (Bioscience, Inc.). Aliquots used for the sorption isotherms were 3, 5,

10, and 15 ul of a low concentration CB stock and 3, 10, and 15 ul of a higher

28

concentration CB stock solution. The low concentration stock solution was 83-mg/L 14C-
CB and the high concentration stock solution contained the same concentration of l4C-CB
and an additional 28 mg/mL of unlabelled (“cold”) CB. The stock solutions were stored
in a -4° C freezer in amber reacti-vials with mininert valves when not in use. Each of the
seven sorption isotherm points were done in triplicate. After flame sealing, the ampules
were equilibrated (“aged”) for either 24 hours or 14 months. To insure homogeneity, the
tubes were shaken (reciprocating shaker) for either 24 hours (for 24-hour aging period) or
72 hours (for 14-month aging). After shaking, the tubes were either analyzed
immediately (24 hour aging) or stored in the dark with periodic shaking (14 month
aging). For analysis, each ampule was centrifuged at 635 x g for 7-9 minutes, the ampule
broken at the neck, and a lmL aliquot of the aqueous phase was added to 8 mL of
scintillation fluid (RPI, Inc., Safety-Solve). The samples were shaken and the next day
analyzed by liquid scintillation counting (Beckrnan LS 6500 multi-purpose scintillation
counter). Values for each triplicate set of ampules were averaged and standard deviations
calculated. The data were plotted as the amount sorbed (ug/ g) versus concentration in
solution (ug/L). Linear regression was used to compute the slope of the line, which
corresponds to the equilibrium sorption coefficient (KP). The Kp values were compared
at each of the aging periods for each soil type.

W: Soil and solution were added to the ampules as described above,
with the same amounts of soil and solution used as were used for the sorption
experiments. A fixed volume of CB stock solution (5 uL) was added to all of the
desorption ampules and the ampules were then flame sealed. The ampules were aged for

the appropriate time period (24 hours or 14 months) and shaken and centrifirged as

29

described above. Then, a quadruplicate set of ampules were opened by cutting the stem
(1.6 — 2.0 cm from the top of the ampule) using a lAt-inch glass drill bit (Sommer & Maca,
Co.), then sampling the solution phase (200 uL) using a 200 uL Gilson Pipetrnan. This
sample was analyzed using scintillation counting (afier 24 hours) to arrive at the
equilibrium concentration in solution (Co) before desorption was initiated. Then a larger
aliquot of the solution phase was removed (z 75-80 % of the total volume of solution)
using a 25 mL gas-tight Hamilton syringe. The same volume of flesh solution
(autoclaved 0.005 M CaCl2 with 200 mg/L NaN3) was then added to the ampule, again
using the 25-mL Hamilton syringe. The syringe was rinsed with clean solution between
withdrawal of the old solution and addition of the new. The ampule was then re-flame
sealed and shaken by hand. At the appropriate time interval, the quadruplicate set of
tubes was centrifuged at 635 x g (7-9 rrrinutes) and the remainder of the stems were
broken at the necks and a 1 mL aliquot of the aqueous phase was sampled and analyzed
by scintillation counting (after 24 hours). This gives the concentration of CB that desorbs
into solution after a given desorption time (Cf). The amount desorbed is: Cf— [Co + (the
mass of CB remaining in the residual solution) -:- (volume of residual solution)]; since
only 75-80 % of the total solution volume was removed, an appreciable amount of
solution remained, the residual solution, which mixed with the fresh solution added to
initiate desorption. Desorption time intervals were: 0.083, 0.5, l, 2, 4, 8, 12, 24, 72, 216,
648, and 2880 hours. Since four ampules were used for each desorption time interval, a
total of 48 ampules were prepared in the same manner and incubated for the 24 or 14
month aging period. The concentration desorbed Cf —- (C0 + residual CB) was converted

to the fraction of the total remaining in the soil (at i=0, the fraction in the soil is 1 and the

30

fraction desorbed is 0).

W5: CB concentrations were monitored during the aging regime to
ensure that chemical or biological degradation did not occur. CB was analyzed by HPLC
utilizing a Perkin-Ehner 250 binary liquid chromatography pump connected in series to a
Waters 480 UV spectrophotometer and a INUS Beta Ram l4C detector. Several ampules
were chosen at random at several time points during the aging period (up to 14 months)
and analyzed for possible degradation products. Samples were centrifuged (635 x g for
7-9 minutes) and a 3 to 4 mL aliquot of the solution was passed through a 0.2 um teflon
filter. The filtrate (z 200 uL) was analyzed on the system described above.
Chromatograms were analyzed and were compared to those obtained initially from the
analysis of CB at the beginning of the aging period. Peaks comprising the
Chromatograms had not changed qualitatively (based on retention times and peak shape)
during the 14 month aging. Elution of the compound as a single peak on both the UV
and 14C detector was evidence of no degradation.

Statisticalanalysis; Sorption isotherms were analyzed for linearity using linear and non-
linear regression analysis. Isotherms were fit to the Freundlich model (x/m = KCe") and
the 95 % confidence interval (CI) for n was generated for the model. Isotherms were
also analyzed for differences in slope between 24-hour sorption and 14 month sorption.
This was accomplished by performing linear regression analysis with indicator variables.
The variability in the model coefficients was tested for significance at the .995 level.
Nonlinear regression analysis was performed on the desorption data using Si gmaplot
graphing software. Parameters for the desorption equation were estimated using the

Marquardt-Levenberg algorithm and standard deviations of each parameter are computed.

31

Results

ngtionlsmherms: Comparing Kp values fi'om each aging period, it is evident
that apparent sorption equilibrium is reached quickly on most soils tested (see Figure
2.1). For Schoolcraft A, Capac A, and Colwood A, the Kp after 24 hours is nearly
identical to Kp after 14 months (see Table 2.2). The slopes of these isotherms are not
statistically different (at the 0.005 level) between each aging period (P =0.015, 0.057,
0.295 for Colwood A, Capac A, and Schoolcraft A, respectively). The exception is CB

sorption on Capac B, where a statistical difference (at 0.005) between isotherm slopes

Table 2.2. Linear sorption model (x/m = Kp-Ce) data after 24 hours and 14 months of

 

aging.
W _rf
CB on Col. A 24 hour 14.2 0.9967
14 month 13.1a 0.9976
CB on Cap. A 24 hour 3.96 0.9978
14 month 4.153 0.9988
CB on Schl. A 24 hour 1.19 0.9967
14 month 1.22a 0.9967
CB on Cap. B 24 hour 0.29 0.9990
14 month 0.36” 0.9991

 

' no statistically significant change in sorption at the .005 level
b a statistically significant increase in sorption at the .005 level

32

 

 

 

  

 

    
   

 

 

 

16-1 1
r 18-2
1&2 J
t 1&3
1&3~ 2
L 1e4
g 194--
g 1&5 1&5
l 1&2
E 1&2a
r 1&3
1&3 -
~ 1&4
«>41
Kp=1.19UKg “ "*5
2
1&5 - r =.9999
r I I I . l 166
1 10 100 1000 10000 1 10 100 1000 10000

Wm In Salmon (uglL)

. 24 hour aging 1:1 14 month aging

Figure 2.1. Sorption isotherms of CB on the four soil types after 24 hours and 14 months
of aging.

33

Table 2.3. Freundlich isotherm (x/m = KOCe“) parameters after 24 hours and 14 months

 

of aging.

KIQL‘ZLCJ.) n_(9_5_°Az_CaL.)
CB on Col. A 24 hour -0.004 — 0.017 0.908 -— 1.274
14 month 0.017 - 0.056 0.826 - 0.943

CB on Cap. A 24 hour 0.010 — 0.013 0.865 - 0.893
14 month 0.009 - 0.010 0.905 - 0.908

CB on Schl. A 24 hour 0.001 — 0.007 0.789 — 0.961
14 month 0.005 - 0.006 0.832 - 0.858

CB on Cap. B 24 hour 0.000 — 0.001 0.942 - 0.971
14 month 0.001 - 0.001 0.895 - 0.938

 

was found to exist (P =0.000). For this soil type, which has the lowest OC content (0.4

%), Kp increased by 22 % from 24 hours to 14 months (Table 2.3). Isothenn data was fit

to both the linear sorption model and the Freundlich model (x/m = K*Ce"). The data

demonstrating isotherm linearity, or near linearity, in the range tested, is the 24 hour CB

sorption on Capac B (95 % CI. of Freundlich coefficient, n = 0.942 - 0.971) data and the

24 hour CB sorption on Colwood A (n = 0.908 — 1.27) data. Linearity is assumed when

the 95 °/o CI. of "n" includes 1. The Freundlich equation parameters for all isotherms are

listed in Table 2.3. When the isotherm data were fit using the linear isotherm model, r2

values for all isotherms fell between .9967 and .9991 (linear plot). Parameters for the

linear isotherm model are listed in Table 2.2. Kom values (calculated from K) appear to

agree with literature values for CB on all soil types (Table 2.4).

mm: A double exponential decay equation (y=a*e(-k1*t) + b*e(-

k2*t)) was used for non-linear regression analysis of the desorption data. Use of this

Table 2.4. Organic matter normalized sorption values (Km) for chlorobenzene.

 

 

Soil K,,24 hr. K,,14 mo. 1K0", (exp.) 2K0... (calc.)
(L/Kg) (LZKg) (UKg) (LZKg)
Colwood A 14.2 13.1 105.81 72.56
Capac A 3.96 4.15 69.77 72.56
Schoolcraft A 1.19 1.22 49.42 72.56
Capac B 0.29 0.36 42.15 72.56

 

‘K,,,, (exp.) = 09,24 hr.)/ tom (fraction of organic matter); fom = (% OC/lOO) . 1.72. 2K0m(calc.)
from log Kom = 0.904 "‘ 10g Kow - 0.779 (Chiou, et al., 1983).

equation assumes that there are fast and slow sites (or “boxes”) for sorption and
desorption, similar to the two-box model (Cameron and Klute, I 97 7). The main
parameters used to analyze the desorption data were the fast desorption rate constant (k1),
the slow desorption rate constant (k2), fraction of chemical in the fast compartment (a),
and the fraction of chemical in the slow compartment (b). Also utilized in the analysis
was the fi'action desorbed (Ida) which gives a measure of the extent of total desorption
after a given desorption period. This value was calculated by dividing the total amount
observed to desorb (Tdcs) divided by the amount that should desorb assuming complete
desorption reversibility (Eqdcs).

in. = Tdu / E110:-
On the desorption plots (Figures 2.2 - 2.5) the value for Eqd,s can be seen as the
asymptote that should, theoretically, be approached by the desorption curve. If the
experimental desorption curve reaches the asymptote, then fa... = 1. The experimental I“,

values after 24 hour and 14 month aging periods are summarized in Table 2.5. The

35

Table 2.5. Chlorobenzene desorption data after 24 hours and 14 months of aging.

 

Jam—1M %m
CB/Col. A .93 .75 19 %
CB/Cap. A .81 .66 19 %
CB/Schl. A .80 .60 25 %
CB/Cap. B .76 .70 8 %

 

 

Table 2.6. Desorption equation parameters (y = anew” + boeH‘m) for chlorobenzene.

 

CB/Col. A 24 hour
14 month

CB/Cap. A 24 hour
14 month

CB/Schl. A 24 hour
14 month

CB/Cap. B 24 hour
14 month

 

i b kIflIIiL—_J$L(hr1)
0.25 0.75 19.6 :1: 2.9 1.76-5 i 66-6
0.25 0.75 18.3 :t 4.3 3.46-17 1' 96-6

0.24 0.76 19.1 :1: 2.9 1.36-5 i 66-6
0.18 0.82 26.1 i 6.6 5.86-6 i 56-6
0.25 0.75 14.0 :1: 4.0 1.76-5 i 1e-5
0.15 0.85 23.6 i6.6 2.56-5 i 56-6
0.13 0.87 12.3 :1: 3.4 1.86-5 i 66-6
0.10 0.90 314 2.06-5 i 8e-6

 

36

.mea mo 3:88 3 EB muse: vm Home < “009.200 :0 m0 mo :ocEOmoQ .N.N oSwE

3.505 2.5...

 

 

 

 

8mm 88 8mm 88 83 88 com o 88 8mm 88 83 82 com o

OW.C r L _ _ _ _ _ _ _ _ _ _ _ _ 00.0
8.0 I I and 4.
. m
85 I I 85 a
D
. 1 II . O
8 o 8 o u
93 I I Ed 8
m

2.0 I l I who
a 8.: u a... m.
86 I who u a... I 8.0 m.
u
36 I I mmd 5
one I . I 85 m.
as "a. «8. I". S
8.0 I m£m< ficoE S. m:_m< .52. em I 8.0 nlh.

 

 

 

 

 

 

 

8.? If it 2:

37

.mea do 3:88 3 can 8:0: vm comm < 880 :0 m0 m0 catacomom .md ecu—ma

8505 05:.

 

 

 

 

 

 

 

 

 

 

88 88 88 89 82 o8 d 88 88 88 82 82 com d
00.0 h m _ _ _ _ _ _ _ _ _ _ _ 00.0
Ed I T Ed
8d I I 8d
8d I m m m I 8d
. 80
8d I 8 d u a I 8d
8d I .. 8d
Ed. "at 8.1%
8 o I m:_m< 5:9: 3. m:_m< So: am T 8 o
2: 8 TI 8e

 

 

 

1103 u! Burugeurau uogroerd

38

.wEwm MO 20008 3 one 8:0: 4m coon < taco—oonom 80o m0 o0 00:9085 .010. 2:me

8.505 0.5...

comm ooom comm oooN oom_. ooo_. com o

coo

moo

cno

mud

coo

moo

coo

moo

ooé

_ L _ _

F

_ &

ocom oomN oooN oomr ooc_. com

H _ _ _

o

_

o

_

 

coo

 

 

8d u it

8.11
I 8:2 £85 3

 

   

 

 

 

Ed. "at
o:_o< .50... em

 

mod

I ono

I mno

I coo

I mod

I coo

I moo

 

 

co...

"as u! Bugugewoa uorroerg

39

ono

.88: co £808 3 one 8:0: vm 80o: m 08:0 :0 m0 o0 cacaowofl .m.m 0.8mm:

$505 08:.

ocom comm ocom oomF oco_. com o

_

_ _ _ _ _ _

comm ccom oom_. ooo_.

_

_

_

_

ocm

_

 

mud 1

 

cod 1

mod I

cod 1

mod I.

 

 

 

24. u N\_
984. 5:2: 2

cho

I mno

 

 

 

=8. nu.
o:_o< .50: em

 

I cco

I mod

I oo.o

I moo

 

 

oo...

oo._.

"as u! Bugugewau uorroerd

40

desorption equation parameters for the experimental data are given in Table 2.6.

In no case did all of the CB desorb completely. Even for samples that were aged
for only 24 hours and then desorbed for 4 months, the most that desorbed (I'm) is
approximately 0.93 (for Colwood A) (Figure 2.2 and Table 2.5). On the other soil types
for the 24 hour aging time, the fraction desorbed is approximately 0.7 6080, which
means 20—24 % is bound “irreversibly” (on the time scales used in this study) after only 1
day of sorption (Figures 2.3, 2.4, 2.5 and Table 2.5). For CB on all soil types, the 1d,,
decreased with increased aging time, which means that more was bound “irreversibly”
for the samples that were aged longer (Table 2.5). This finding is in agreement with what
many other investigators have found. The largest change in f“. was on Schoolcraft A;
0.80 desorbed after aging for 24 hours but only 0.60 desorbed after 14 months of aging
(Figure 2.4 and Table 2.5). After aging for 14 months and then initiating desorption for
four months, 40% of CB was found to resist desorption on Schoolcraft A.

There appears to be no effect of aging on the fast desorption rate constant (kI)
(Table 2.6). Initial desorption rates appear to be very rapid regardless of the aging time
and desorption appears to be essentially complete after one to two hours of desorption
time. Based on the modeling results using the double decay equation, the rate constant kI
was the same for each soil type regardless of whether the sample was aged for 24 hours
or 14 months. For example, k1 for CB desorption from Colwood A after 24 hours of
aging was 19.6 i 2.9 hr”. The k1 after 14 months ofaging was 18.3 i 4.3 hr".

There also does not appear to be a consistent effect of aging on the slow
desorption rate constant k2 (Table 2.6). Again, the values for k2 from the double decay

equation are the same for each soil type regardless of the aging time. In a few cases the

41

variability in the k2 estimates are large enough so that comparisons and conclusions are
not possible, but in all other cases they are the same. For example, R2 for CB desorption
from Capac A after 24 hours of aging is 1.35 x 10'5 i 5.7 x 1045 hr'l. The k2 after 14
months of aging is 5.80 x 10*5 i- 4.8 x 10‘ In“.

Although there was no increase in sorption due to aging (except on Capac B),
aging did seem to effect the distribution of chemical within the soil matrix. For all soil
types, except Colwood A, there was a decrease in the fraction of chemical residing in the
fast compartment (a) and an increase in the fraction residing in the slow compartment (b)
(Table 2.6). The fraction in “b” from the modeling data is analogous to the “irreversible”
fraction (1 - f2”). On Colwood A, the amount in each compartment remained constant.

Reasons for this behavior are not clear at this time.

Discussion

There seems to be little or no effect of aging on the apparent sorption equilibrium
of CB in all soil types. Although there was an increase in Kp for the Capac B soil, the
increase was very small. It appears that for CB, equilibrium is approached quickly in
soils of moderate or high SOM content. The results from Table 2.4 demonstrate good
agreement between calculated Kom and experimental K0m , which can be expected to vary
up to 2 to 3 times for different soils (Kile, et al., 1995). This information provides some
evidence that SOM is the dominant sorptive phase for CB in the soils tested.

Results from these experiments did not show any clear increase in non-linearity
with aging time. This is in contrast to the observations of Xing and Pignatello (1996) and

Weber and Huang (1996). Xing and Pignatello (1996) observed increases in non-

42

linearity of isotherms for dichlorobenzene (DCB) and dichlorophenol (DCP) between 1
and 180 days. This was attributed to the fact that SOM has both partition and adsorption
domains (dual-mode sorption) and that the adsorption domains are reached only given
long soil-cherrrical contact times since these domains are internal in the SOM matrix
(Xing and Pignatello, 1996). The adsorption component gives the isotherm increasing
non-linearity, which only becomes apparent with long soil-chemical contact times. The
isotherms observed in this study do not appear to be consistent with a dual-mode sorption
mechanism. According to the results presented here, in one case (Colwood A), isotherm
linearity decreased with aging time, but in another case (Capac A), linearity increased
with aging time (Table 2.2). These results also did not show increased non-linearity in
relatively low OM soils. The most linear isotherms were found on the highest and lowest
OM soils, Colwood A and Capac B, respectively (see Table 2.2).

The use of the linear sorption model versus the Freundlich model is an issue that
is difficult to resolve. In cases where the F reundlich exponent was found to be farthest
from 1 (Schoolcraft A data), apparently indicating non-linearity (when n=1, the isotherm
is linear), r2 values for the linear model are greater than 0.9967 (linear plot) (see Tables
2.2 and 2.3). If the 95 % CI. of “11” does not include 1, then the Freundlich model
provides a better fit to the data than the linear model, and the use of the more
parameterized (Freundlich) model is justified statistically. So, in some cases, the
Freundlich model is more appropriate, but in other cases the linear model is more
appropriate. On a practical level, which ovenides statistical significance, the linear
model provides a very good approximation of the sorption behavior of CB on these soils

and a very accurate fit to the data. For reasons of practicality, consistency, and accuracy

43

the linear model is the model of choice for this study.

The statistically significant increase in Kp on Capac B with aging time is
consistent with evidence from other studies indicating that aging has a greater effect on
sorption in relatively low OM soils (McCall and Agin, I985; Xing and Pignatello, 1996).
However, the magnitude of the increase observed here is relatively small compared to
that observed by Ball and Roberts (1991). They observed an order of magnitude increase
in the long-term sorption of tetrachlorobenzene in an aquifer material (0.021 % O.C.)
over the amount of sorption expected based on organic matter partitioning. Here, the
long-term sorption of CB on Capac B is approximately the same (52.32 L/Kg) as the
amount of sorption expected based on organic matter partitioning (72.56 L/Kg).
Although, it is possible that this discrepancy could be accounted for by the differences in
the chemicals studied and the OC content of the soils. As mentioned earlier, it was
observed that for dichlorophenol (DCP) the largest relative increase in the sorption
coefficient (K) between 1 and 180 days occurred in soil (1.74 % OC) rather than in peat
(54.07 % OC) (Xing and Pignatello, 1996). The magnitude of the increase for DCP due
to aging is a: 175 %; in this study, the increase for CB is only z 22 %. Results from
McCall and Agin (1985) also demonstrate an increased effect of aging in low OM soils.
There appears to be a trend of increased effect of aging in soils of low OC content, but
the magnitude of this effect does not appear to be consistent.

The desorption kinetics of CB in this study shows that there is an increase in the
“irreversible” fraction (or a decrease in 1",.) with increased aging time for all soils. This
is in agreement with observations from many other studies (Pignatello, I990a;

Pavlostathis and Mathavan, 1992). There does not appear to be any correlation to the

increase in irreversible fiaction and SOM content, but this relationship may be obscured
by the large error associated with the Capac B desorption data (Figure 2.5). This soil was
the most difficult to work with because of its low sorption capacity and the resultant need
for a high soil:solution ratio. This made the soil slurry difficult to remix after
centrifugation, which was a problem during desorption kinetics where precise sampling
times were important. This could have helped contribute to the large uncertainty of some
of this data If one neglects the Capac B data, there does appear to be a greater increase
in the irreversible fraction with time in soils with decreased SOM content, but with only
three soils this conclusion is difficult to make with any certainty.

Another surprising observation is that for CB and soils aged for only 24 hours,
there is a significant irreversible fi'action even when desorption was performed for up to
four months. This data set indicates that diffusion is unlikely to be the main process
governing sorption/desorption behavior. Previous work has suggested that sorption and
desorption of organic chemicals in soils is reversible in accordance with a retarded
diffusion mechanism (Wu and Gschwend, 1986). The data presented here, along with
data fi'om other recent studies on the irreversibility of PAH sorption/desorption (Kan, et
al., 1994; Fu, et al., 1994), appear contradictory to the aforementioned hypothesis. The
size of the desorption resistant fraction observed for PAH's of 30-50 % (Fu, et al., 1994)
is very similar in magnitude to the desorption resistant fi'action for CB found in this
study. The rapid formation of an “irreversible” (desorption resistant) fraction suggests a
specific physical or chemical sorption interaction, rather than retarded diffusion, as the
dominant mechanism.

It is very difficult to discern from the data whether the slow “tailing” of chemical

45

in the desorption graphs has actually "plateaued" or if it is approaching the asymptote at a
very slow rate. Even with quadruplicate samples, the error bars are large enough to make
this conclusion unreachable. In some cases there appears to be a “plateau” and in other
cases it appears that there may be an extremely slow net desorption process still taking
place. Regardless, on a practical level it appears that the net desorption process has
stopped after approximately one day and there is little or no change fi'om one day to four
months. This means there is a significant irreversible (non-desorbable) fraction of CB,
which forms rapidly with aging. This finding was unexpected fiom the standpoint that
CB is a non-polar compound thought to be unreactive in soils.

Examination of the kinetic parameters shows that there appears to be no effect of
aging on the slow or fast desorption rate constants (k1 or k2). In all cases, the rate
constants are nearly the same between 24 hours and 14 months. These results are in
contrast to McCall and Agin (1985) who observed no change in the fast desorption rate
constant with aging time, but there was a decrease in the slow desorption rate constant.
Their data was for picloram, which is a larger, more polar molecule than CB, aged in
soils for 1 and 200 days. These results also contrast those of Carmichael, et al. (1997)
which found that the desorption rates (not rate constants) for fresh and added PAH's were
different by several orders of magnitude. If there are actual differences in k1 due to
aging, which are not easily discemable with the method used here, it would appear that
these differences must be extremely small and perhaps negligible on a practical level.
These results are not easily attributable to simple intraparticle diffusion. If diffusion were
operative, one would expect the shape of the desorption curve to be markedly different

for the 14 month sample. The curve should show a long, slow "tailing" towards the

46

asymptote and have reduced kinetic coefficients. Schulten (1995) has suggested that the
interaction of SOM and organic chemicals is one in which the humic substances are
penetrated and disrupted by an organic compound, which increases the overall energy
state of the complex. Given an appropriate period of time (aging) the complex finds a
more relaxed state due to structural or conformational rearrangements characterized by
shrinking of the void space around the pollutant molecule. The complex is now more apt
to sequester the compound permanently, possibly leading to what is observed as the
irreversible fiaction on our desorption profiles. Although the kinetic coefficients and the
amount sorbed do not change with aging, the distribution of compound within the fast
and slow compartments does change with aging. In all cases, except Colwood A, the
fraction of CB in the "slow" compartment (b) increased with aging time. This could
indicate that movement into the "slow" compartment should not be viewed as diffusion
deep into the soil matrix, but as a static process where the compound becomes bound at
one physical location on the soil particle. The slow compartment may not be a physically
distant area of the soil particle, but a kinetically limited site located in the same region as
non-kinetically limited sites.

Field studies have also shown markedly decreased bioavailability of aged
pesticide residues. In the field, the rapidly desorbable fraction (observable as the fraction
in “a” in the desorption plots) would be expected to move into the aqueous phase to
reestablish equilibrium as pesticides are removed by biodegradation, leaching, plant
uptake, or volatilization. This fraction of the soil-applied pesticides would be rapidly
dissipated, leaving only the irreversible fraction. This fraction would likely be highly

immobile and very difficult to access, and hence difficult to degrade. Increased aging

47

times would promote the formation of such a bound fraction. This is exactly what has
been observed for compounds such as EDB and sirnazine that have been applied
continuously over long periods of time (Steinberg, et al., 198 7; Scribner, et al., 1992).
The findings discussed herein show that aging of CB in soils, as performed in a
laboratory setting, has an effect on the distribution of CB within the soil matrix and
decreases the amount desorbed. Aging of CB in soils had little or no effect on other
factors such as the rate of desorption and the extent of sorption. The effect of aging on
apparent sorption equilibrium may be related to SOM content, with soils of low organic
content approaching equilibrium more slowly. The relation between SOM content and
desorption is still unclear as no conclusive trends were established, although it is possible

that desorption is less reversible in soils with very low OM content.

List of References

Adamson, A.W. 1990. Physical Chemistry of Surfaces. John Wiley & Sons, New
York.

Ball, W.P. and Roberts, P.V. 1991. Long-terrn sorption of halogenated organic
chemicals by aquifer material. 1. Equilibrium. Environ. Sci. Technol. 25:1223-1236.

Cameron, D.A. and Klute, A. 1977. Convective-dispersive solute transport with a
combined equilibrium and kinetic adsorption model. Water Resour. Res. 13: 183-1 88.

Carmichael, L.M., Christman, RE, and Pfaender, F.K. 1997. Desorption and
mineralization kinetics of phenanthrene and chrysene in contaminated soils. Environ.
Sci. T echnal. 31 : 126- l 32.

Carroll, K.M., Harkness, M.R., Bracco, A.A., Balcarcel, RR. 1994. Application of a
permeant/polymer diffusional model to the desorption of polychlorinated biphenyls from
Hudson river sediments. Environ. Sci. T echnal. 28:253-258.

Castro, CE. and Belser, ND. 1968. Biodehalogenation. Reductive dehalogenation of

the biocides ethylene dibrorrride, 1,2-dibromo-3-chloropropane, and 2,3-dibromobutane
in soil. Environ. Sci. T echnol. 2:779-783.

48

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

Chiou, C.T., Porter, P.E., Schmedding, D.W. 1983. Partition equilibria of non-ionic
organic compounds between soil organic matter and water. Environ. Sci. T echnal.
1 7 :227-23 1 .

Engebretson, R.R., Amos, T., and von Wandruszka, R. 1996. Quantitative approach
to humic acid associations. Environ. Sci. T echnal. 30:990-997.

Fn, G., Kan, A.T., Tomson, M.B. 1994. Adsorption and desorption hysteresis of
PAH’s in surface sediment. Environ. T axica]. Chem. 13:1559-1567.

Harmon, T.C. and Roberts, P.V. 1994. Comparison of intraparticle sorption and
desorption rates for a halogenated alkene in a sandy aquifer material. Environ. Sci.
T echnal. 28:1650-1660.

Hatzinger, PB. and Alexander, M. 1995. Effect of aging of chemicals in soil on their
biodegradability and extractability. Environ. Sci. T echnol. 29:537-545.

Kan, A.T., Fu, G., and Tomson, M.B. 1994. Adsorption/desorption hysteresis in
organic pollutant and soil/sediment interaction. Environ. Sci. T echno]. 28:859-867.

Kile, D.E., Chiou, C.T., Zhou, H., Li, H., Xu, 0. 1995. Partition of non-polar organic
pollutants from water to soil and sediment organic matter. Environ. Sci. Technol.
29:1401-1406.

Loehr, RC. and Webster, M.T. 1996. Behavior of fresh vs. aged chemicals in soil. J.
Soil Cantam. 5:361-383.

McCall, P.J. and Agin, G.L. 1985. Desorption kinetics of picloram as affected by
residence time in the soil. Environ. T axical. Chem. 4:37-44.

Pavlostathis, S.G. and J aglal, K. 1991. Desorptive behavior of trichloroethylene in
contaminated soil. Environ. Sci. T echnol. 25:274-279.

Pavlostathis, S.G. and Mathavan, G.N. 1992. Desorption kinetics of selected volatile
organic compounds from field contaminated soils. Environ. Sci. T echnal. 26:532-538.

Pignatello, J.J. 1986. Ethylene dibromide mineralization in soils under aerobic
conditions. Appl. Environ. Microbial. 51:588-592.

Pignatello, J.J. 1990a. Slowly reversible sorption of aliphatic halocarbons in soils. 1.
Formation of residual fractions. Environ. Toxicol. Chem. 9:1107-1115.

Pignatello, J.J. 1990b. Slowly reversible sorption of aliphatic halocarbons in soils. II.

49

Mechanistic aspects. Environ. T axical. Chem. 9:1117-1 126.

Pignatello, J.J. and Xing, B. 1996. Mechanisms of slow sorption of organic chemicals
to natural particles. Environ. Sci. T echnal. 30:1-11.

Schulten, H.R. 1995. The three-dimensional structure of humic substances and soil
organic matter studied by computational analytical chemistry. Fresenius J. Anal. Chem.
3 5 1 262-73.

Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M. 1993. Environmental
Organic Chemistry. John Wiley & Sons, Inc., New York, NY.

Scribner, S.L., Benzing, T.R., Sun, S., and Boyd, S.A. 1992. Desorption and
bioavailability of aged simazine residues in soil from a continuous corn field. J. Environ.
Qua]. 21:115-120.

Steinberg, S.M., Pignatello, J.J., and Sawhney, B.L. 1987. Persistence of 1,2-
dibromoethane in soils: Entrapment in intraparticle nricropores. Environ. Sci. T echnol.
21:1201-1207.

Weber, W.J., J r., Huang, W. 1996. A distributed reactivity model for sorption by soils
and sediments. 4. Intraparticle heterogeneity and phase-distribution relationships under
nonequilibrium conditions. Environ. Sci. T echnal. 30:881-888.

Wilke, CR. and Chang, P. 1955. Correlation of diffusion coefficients in dilute
solutions. A.I.Ch.E. Journal. 1:264-270.

Wu, S. and Gschwend, P.M. 1986. Sorption kinetics of hydrophobic organic
compounds to natural sediments and soils. Environ. Sci. T echnal. 20:717-725.

Xing, B. and Pignatello, J.J. 1996. Time-dependent isotherm shape of organic

compounds in soil organic matter: Implications for sorption mechanism. Environ.
T axicol. Chem. 15:1282-1288.

50

Chapter 3

EFFECTS OF AGING ON THE SORPTION/DESORPTION BEHAVIOR OF EDB,

CHLOROBENZENE, 2,4-D, AND ATRAZINE

Abstract
Field studies have demonstrated that prolonged pesticide-soil contact times

(aging) leads to unexpected persistence of these compounds in the environment.
Although this phenomena is well documented in the field, there have been very few
controlled laboratory studies that have tested the effects of long-term aging and the role
of differing sorbates on contaminant sorption/desorption behavior and fate in soils. This
study was initiated to examine the sorption/desorption behavior of chlorobenzene (CB),
1,2-dibromoethane (EDB), atrazine (AT), and 2,4-dichlorophenoxyacetic acid (2,4-D) on
one soil type after 24 hours, 1 month, and 14 months of aging. Sorption isotherms were
prepared and evaluated after each aging period to observe changes in the uptake of each
compound by soil. Desorption kinetic curves were generated after each aging period to
observe changes in release from soil, with desorption being followed for up to four
months. The data indicates increases in sorption for all compounds except CB, indicating
that for 2,4-D, EDB, and AT, equilibrium was not reached in 24 hours or less. Initial
desorption rates appear to be the same despite differences in aging times. Desorption
equilibrium was not reached for any chemical or any aging time, except 2,4-D after aging
for 24 hours. In this case, desorption was found to be reversible and equilibrium was
attained. In all other cases, a large non-desorbable fraction was evident.

Keywords — atrazine, 2,4-D, EDB, aging, sorption, desorption kinetics, soils

51

Introduction

Although laboratory studies of organic contaminant sorption by soils often
employ equilibration times of 24 hours or less, several recent studies have indicated that
sorption is not a rapid process and may take long time periods (hundreds of days) to reach
equilibrium (Ball and Roberts, 1991; Pignatello and Xing, 1996). Furthermore, several
desorption studies have found that the desorption resistant (irreversible) fi'action of a
chemical increases with increased soil-chemical contact time (aging) (Pignatello, 1990;
Pavlostathis and Mathavan, 1992). Data from field studies have indicated that certain
field weathered or aged compounds (e. g. EDB, simazine) demonstrate unexpected
persistence despite being readily degradable by soil microorganisms (Steinberg, et al.,
198 7; Scribner, 1992). These observations suggest that aging does affect
sorption/desorption behavior, bioavailability, and biodegradability of organic
contaminants and pesticides in soils. Despite the dramatic effects attributed to aging on
contaminant binding and persistence, there are very few long-term, laboratory controlled
studies which explicitly measure sorption/desorption parameters after different aging
times. Consequently, little is known about the effects of aging on contaminant fate as
might be influenced by sorbent and sorbate properties.

Several researchers have observed that increasingly hydrophobic compounds
exhibit slower sorption/desorption kinetics in soils (Karickhafl and Morris, 1985;
Brusseau and Rao, 1989b). In contrast, Oliver (1 985) found that desorption behavior of
aged non-polar organics in sediments was largely independent of adsorbate properties.
The compounds studied exhibited nearly a four order of magnitude range in

hydrophobicity, as measured by K0W (octanol-water partition coefficient), but only

52

included relatively non-polar chlorinated hydrocarbons. Several other studies have
examined the effects of aging of both polar (and/or compounds with polar functional
groups) and non-polar organic compounds. Xing and Pignatello (1996) examined
sorption isotherms of dichlorobenzene (non-polar) and dichlorophenol (polar) in soil after
0 and 30 days of aging and found that the Kp increased only 1.3 times for
dichlorobenzene (DCB; log Kp= 3.38) but 2.7 times for dichlorophenol (DCP; log Kp=
2.75). This would suggest that the non-polar compound, although slightly more
hydrophobic, had faster sorption kinetics than the polar compound. These results are in
agreement with the results of another study (Brusseau and Rao, 1989a) in which inverse
linear free energy relationships (LFER) were observed between sorption rate coefficients
(log k) and sorption equilibrium coefficients (log K) for both non-polar and polar
compounds. Among compounds with similar partition coefficients, the polar compounds
exhibited considerably slower sorption/desorption kinetics. They attributed this to the
increased reactivity of polar compounds (and/or compounds with polar functional
groups). Brusseau and Rao ([9890) suggested that polar compounds must diffirse
through soil organic matter (a physical non-equilibrium process or PNE) where they are
also subject to specific chemical interactions due to their polar ftmctional groups (a
chemical non-equilibrium processes or CNE). These authors presented evidence that
polar compounds are rate-lirnited due to diffusion (PNE) and CNE whereas non-polar
compounds are rate limited solely by diffusion. Polar compounds, such as atrazine,
picloram, and chlorophenols, which have reactive functional groups, are known to form
specific interactions with soil components (Hayes, 1970; Li and F elbeck, 1972;

Nearpass, 1976). When the results of Xing and Pignatello (1996) are viewed in this

53

context, DCP would be expected to exhibit a larger increase in Kp upon aging than DCB,
as was observed. Based upon the data of Brusseau and Rao (1989a), the relatively
similar log Kp values for DCB and DCP would translate to a log k value for
dichlorophenol that is two orders of magnitude slower than that for dichlorobenzene.
The objective of this study was to evaluate the effects of aging on sorption
dynamics of polar and non-polar compounds in soils utilizing long-term laboratory
controlled aging regimes. Sorption isotherms and desorption kinetics of chlorobenzene
(CB), EDB, atrazine (AT), and 2,4-D on a Capac A soil were measured after 24 hours, 1

month, and 14 months, with desorption being followed for up to four months.

Experimental Section

A Capac A horizon soil (fine-loamy, mixed, mesic Aeric Ochraqualf) was used in
this study. The soil was collected fi'om southern Michigan, air dried, ground, and sieved
(2 mm). Sand, silt, clay, and organic carbon (OC) content were determined by the
Michigan State University Plant and Soil Nutrient Lab. This soil is a sandy clay loam
with a 3.3 % OC content. It is composed of 54.6, 24.0, and 21.4 % sand, silt, and clay,
respectively. The soil was then gamma irradiated with 5 MRad at the University of
Michigan Ford Nuclear Reactor Laboratory using a cobalt-60 irradiator (dose rates by
Reuter-Stokes ion chamber). The soil was then stored in 500 mL Nalgene bottles during
irradiation and remained sealed until use.
Somfignlsotherms; The quantities of soil and solution used to develop sorption
isotherms corresponded to the amount needed to sorb z 50% of the added solute while

minimizing headspace volume. Either 20 mL glass ampules (for EDB and

54

chlorobenzene) or 20 mL Corex glass tubes (for 2,4-D and AT) were utilized for the
sorption/desorption experiments. The ampules were closed by flame sealing and the
tubes were closed using teflon lined screw caps. The tubes and ampules were autoclaved
prior to use. A mass of soil was weighed and transferred to the ampule/corex tube using
an autoclaved metal spatula along with an autoclaved plastic funnel. A 0.005 M CaCl2
solution, which had been autoclaved for 25 minutes and amended with 200 mg/L NaN3,
was added to the tubes using a sterilized repipettor. All additions of soil and solution to
the ampules and tubes were carried out in a laminar flow hood and all items contacting
soil or solutiOn were autoclaved prior to use (funnels, spatulas, etc.).

Acetone stock solutions of the following compounds were prepared: 14C-
chlorobenzene (CB)(radiochemical purity (RP) > 97 %), 14C-l,2 dibromoethane
(EDB)(RP > 98 %), ”C-2,4-dichlorophenoxyacetic acid (2,4-D) (RP > 97 %), and ”C-
atrazine (AT)(RP > 93 %) (were obtained from Sigma Chemical Co.). Stock solutions
for each compound consised of a low concentration stock, in which the specific
compound is all 14C labeled, and a high concentration stock, which contains l4C
compound and non-labeled (“cold”) compound. The stock solutions were stored in
amber reacti-vials with mininert valves in 3 -4° C fieezer while not in use. An aliquot of
the stock solution of a specific compound was added to the ampule/corex tube using a 10,
25, or 100 uL Hamilton syringe and the ampule was immediately flame sealed using an
Ampuhnatic flame sealing unit (Bioscience, Inc.); Corex glass tubes were immediately
sealed with screw-on caps. Aliquots used for the sorption isotherms were from both the
low and high concentration stock solutions to cover a 3-4 order of magnitude range in

initial solution concentration. Seven initial solute concentrations were utilized (in

55

triplicate) for each compound. After sealing, the ampules/corex tubes were equilibrated
(“aged”) for either 24 hours, 1month, or 14 months. Immediately after closure the
ampules/tubes were shaken on a reciprocating shaker for either 24 (for 24 hour aging
period) or 72 hour (for 1 and 14 month aging) periods. After shaking, the ampules/tubes
were either analyzed immediately (24 hour aging) or stored in the dark with periodic
shaking (l and 14 month aging). Prior to analysis, ampules were centrifuged at 635 x g
for 7 to 9 minutes and a lmL aliquot of the aqueous phase was removed and added to 8
mL of scintillation fluid. Tubes were handled in the same manner except that they were
centrifuged at 6722 x g for 7 to 9 rrrinutes. The samples were then shaken, stored in the
dark overnight, and the next day analyzed by liquid scintillation counting (Beckrnan LS
6500) to determine the aqueous phase concentrations; the amount of solute sorbed was
determined by difference. Linear regression was used to compute the slope (or sorption
coefficient, KP) of the sorption isotherm on a plot of amount sorbed (ug/Kg) versus
concentration in solution (ug/L).

W Soil and solution were added to the ampules/tubes as described
above, with the same amounts of soil and solution used as for the sorption experiments.
A single initial solute concentration was used for each chemical. These were then aged
for the appropriate time period (24 hours, 1 month, or 14 months) and shaken and
centrifirged as described for the sorption isotherms. Then, a quadruplicate set of ampules
were opened by cutting the stem (1.6 — 2.0 cm from the top of the ampule) using a lAI-inch
glass drill bit (Sommer & Maca, Co.), and sampling the solution phase (200 uL) using a
200 uL pipetrnan. Tubes were sampled in triplicate by removing the caps and taking a 1

mL sample of the solution phase. This sample was analyzed using liquid scintillation

56

counting (after 24 hours) (Beckrnan LS 6500) to arrive at the equilibrium concentration
in solution (Co) before desorption was initiated. Then a larger aliquot of the solution
phase was removed (z 75-80 % of the total volume of solution) using a 10 or 25 mL gas-
tight Hamilton syringe. The same volume of fresh solution (autoclaved 0.005 M CaCl2
with 200 mg/L NaN 3) was then added to the ampule/tube, again using the 10 or 25 mL
Hamilton syringe. The syringe was rinsed with clean solution between withdrawal of the
old solution and addition of the new. The ampule was then re-flame sealed and shaken
by hand, and the tubes were re-capped and shaken by hand. At the appropriate time
interval, the quadruplicate set of ampules were centrifuged at 635 x g (7 to 9 minutes),
the remainder of the stems were broken at the neck, and a 1 mL aliquot of the aqueous
phase was sampled and analyzed by liquid scintillation counting (after 24 hours). This
gives the concentration of CB that desorbs in solution after a given desorption time (CI).
The amount desorbed is: Cf — [C0 + (the mass of CB remaining in the residual solution) +
( volume of residual solution)]. The tubes were treated in the same manner as the
ampules, except that triplicate samples were analyzed and these were centrifuged at 6722
x g for 7 to 9 minutes. Desorption time intervals were: 0.083, 0.5, l, 2, 4, 8, 12, 24, 72,
216, 648, and 2880 hours. Since four ampules were used for each desorption time
interval, a total of 48 ampules were prepared in the same manner and incubated for the 24
hour, 1 month, or 14 month aging period. Thirty-six tubes were prepared in the same
manner and incubated for the aging periods stated above. The concentration desorbed CI
— (Co + residual CB) was converted to the fraction of the total remaining in the soil (at

=0, the fi'action in the soil is 1 and the fi'action desorbed is 0).

Anabflicalflocedmes; High-perfonnance liquid chromatography (HPLC) analysis was

57

used to determine the radiochemical purity and stability of the l4C-CB, l4C-EDB, l4C-
AT, and l4C-2,4-D for the 14 month aging period. After receiving the 14C compounds,
acetone stock solutions were prepared and stored in a —4° C freezer until they were ready
to use. Prior to first use, an aliquot of the stock solution was analyzed utilizing a Perkin-
Elmer 250 binary liquid chromatography pump connected in series to a Waters 480 UV
spectrophotometer and a INUS Beta Ram 14C detector. Radiochemical purity was
estimated by injecting z 0.025 uCi from the acetone stock solution. Elution of the
compound as a single peak on both the UV and 14C detector was evidence of
radiochemical purity. The relative peak areas of the parent compound and other non-
sirnultaneously eluting peaks was used to quantitate the radiochemical purity. The
compounds were delivered from Sigma as 93-98 % radiochemically pure and our results
verified this. This procedure was repeated before each aging experiment. After the 14
month aging period, several tubes (for each compound) were chosen at random and
analyzed for possible degradation products. Samples were centrifirged and a 3 to 4 mL
aliquot was passed through a 0.2 um teflon filter. The filtrate (z 200 uL) was analyzed
by HPLC as described above. Chromatograms were analyzed and were compared to
those obtained initially from the analysis of radiochemical purity. Peaks comprising the
chromatograrns had not changed qualitatively (based on retention times and peak shape)
during the 14 month aging period, except for AT. AT showed significant chemical
degradation (> 20 % ) for the 14 month period, but not for the 1 month aging period.
WSorption isotherms were analyzed for linearity using non-linear
regression analysis. Isotherms were fit to the Freundlich model (x/m = K* Ce") and the

95 % confidence interval (C.I.) for n was calculated. Isotherms were also analyzed for

58

differences in slope between 24 hour sorption and 14 month sorption. This was
accomplished by performing linear regression analysis with indicator variables. The
variability in the model coefficients were tested for significance at the 0.99 level.
Nonlinear regression analysis was performed on the desorption data. Parameters for the
desorption equation were estimated using the Marquardt-Levenberg algorithm and

standard deviations of each parameter are computed.

Results
W: Comparing Kp values (Figure 3.1 and Table 3.1) for the aging
periods of 24 hours and 1 month, there is very little difference in apparent sorption for
any compound except 2,4-D. EDB, CB, and AT had no statistically significant (at the
0.01 level) increase in sorption from 24 hours to 1 month. 2,4-D did show a statistically
significant (P = 0.001) increase in the slope of the sorption isotherm between 24 hours
and 1 month (Table 3.1). At 14 months of aging, 2,4-D and EDB both had an increase in
Kp fiom the corresponding values at 24 hours and 1 month (P = 0.000 for both), whereas
CB still showed no increase in Kp for any aging period. Also, AT showed a statistically
significant, albeit small, decrease in K" at 1 month. Sorption of AT at 14 months could
not be evaluated due to chemical conversion of AT to hydroxy-AT. This was confirmed
with HPLC analysis by matching the retention times of the degradation product and an
authentic standard. There appears to be good agreement between experimental K0m
(calculated from K) and calculated K0m values for all compounds (Table 3.2).

The isotherm data were fit to both the F reundlich sorption model (x/m=K*Ce"),

where x/m is the concentration in soil and Ce is the aqueous phase concentration, and the

59

Table 3.1. Linear sorption model (x/m = Kp*Ce) data for EDB, CB, AT, and 2,4-D.

 

Kali-[112) r3
EDB on Cap. A 24 hour 0.73 0.9999
1 month 0.72“ 0.9965
14 month 1.26” 0.9992
CB on Cap. A 24 hour 3.96 0.9978
1 month 3.37 0.9995
14 month 4.15‘ll 0.9988
AT on Cap. A 24 hour 3.24 0.9980
1 month 2.82c 0.9932
14 month ND. ND.
2,4-D on Cap. A 24 hour 0.62 0.9981
1 month 0.66” 0.9992
14 month 0.91” 0.9961

 

‘ no statistically significant increase in sorption from 24 hour sample at the .01 level
b a statistically significant increase in sorption from 24 hour sample at the .01 level
c a statistically significant decrease in sorption from 24 hour sample at the .01 level

linear sorption model. The Freundlich isotherm data showed a slight degree of non-
linearity as indicated by the F reundlich coefficient “n”. The most linear isotherms were
found to be for 24 hour 2,4-D sorption (n = 0.939 - 1.18; 95 % CL), 24 hour EDB
sorption (n = 0.966 - 0.995), and 1 month CB sorption (n = 1.00 -1.03). The Freundlich
isotherm parameters are listed in Table 3.3. When the data was fit using the linear model
(n=1), all isotherms had 12 values of 0.9932 or greater (Table 3.1).

W A double exponential decay equation (f=a*e(-k1*t) + b*e(-k2*t)
was used for non-linear regression analysis of the desorption data. Use of this equation
assumes that there are fast and slow sites (or "boxes") for sorption and desorption, similar

to the two-box model (Cameron and Klute, 1977). The parameters used to analyze the

60

1 10 1(1) 1W 10000 1 10 1(1) 1(DO 10000

 

 

Armuntsorbod(uglKg)

   

 

 

 

 

1&2 -
- 1&2
1&3 1
1&3
1&4 «
1&4
1&5 ~
1&8
1&1 -
1&2 : 19.3
14 Month Aging
I83- VII-sown: /
/
2_ / 1&4
164 J r "-9993 / 24 r i
Kp=0.617 [Mg
2
19.5 s r =.9997 1&5
ie+1 1e+2 1e+3 1e+4 1e+5 1 10 1(1) 1000
Concentration in solution (uglL)
. 24 hour aging a 14 month aging

Figure 3.1. Sorption isotherms of CB, EDB, 2,4-D, and AT after 24 hours and 14
months (or 1 month for AT) of aging.

61

Table 3.2. Organic matter normalized sorption values (Kom) values for EDB, CB, and

 

 

AT.

K,,24 hr. Kp,14(for AT, 1 mo.) K0w 1K0m(exp.) 2K0m(calc.)

) (L/IggL—JLlKg)
EDB 0.73 1.26 57.54 12.77 6.49
CB 3.96 4.15 831.76 69.77 72.56
AT 3.24 2.82 363.08 57.08 34.30

 

lK0... (exp.) = (Kp,24 hr.)/ fom (fraction of organic matter); fom = (% OC/ 100) * 1.72. 2I(,,,,,(calc.)

from log K0,m = 0.904 "' log K0,, — 0.779 (Chiou, et al., 1983).

Table 3.3. Freundlich isotherm (x/m = K*Ce") parameters after 24 hours, 1 month, and

 

14 months of aging.
W aim.)
CB 24 hour 0.010 — 0.013 0.865 — 0.893
1 month 0.002 — 0.003 1.004 — 1.032
14 month 0.009 — 0.010 0.905 - 0.908
EDB 24 hour 0.001 — 0.001 0.966 - 0.995
1 month 0.002 — 0.004 0.827 - 0.884
14 month 0.002 — 0.003 0.915 — 0.935
AT 24 hour 0.005 — 0.007 0.875 — 0.926
1 month 0.005 -— 0.012 0.772 — 0.901
14 month ND. ND.
2,4-D 24 hour 0.000 — 0.001 0.939 — 1.180
1 month 0.001 — 0.002 0.905 — 0.953
14 month 0.001 — 0.006 0.794 — 0.952

 

62

desorption data were the fast desorption rate constant (k.), the slow desorption rate
constant (k2), fraction of chemical in the fast compartment (a), and the fraction of
chemical in the slow compartment (b). Also utilized in the analysis was the fraction
desorbed (fag), which gives a measure of the extent of total desorption. This value was
calculated by dividing the total amount observed to desorb (T46) divided by the amount
that should desorb assuming complete desorption reversibility (Eqdm).
fdee = Tau / Equa

On the desorption plots (Figures 3.2 - 3.7) the value for Eqdas can be seen as the
asymptote that should, theoretically, be approached by the desorption curve. If the
experimental desorption curve reaches the asymptote, then fd,, = 1.

The desorption curves are shown in Figures 3.2 - 3.7. In only one case did all of
the sorbed chemical desorb ahnost completely. This was for the 24 hour sorption of 2,4-
D, where fdc, is approximately 0.90 (Figure 3.2). For the 14 month and 1 month samples,
a considerably smaller fraction of 2,4-D desorbed (0.58 and 0.68, respectively) (Figures
3.2 and 3.3, respectively). When the desorption intervals for 2,4-D were increased
beyond four hours to 4 months, the 13,, increased to .98 and .74 for the 24 hour and 14
month 2,4-D samples, respectively (Figure 3.4). This means that for the 2,4—D sample
aged 24 hours, 98 % desorbed after a four month desorption interval, while 90 % had
desorbed after a four hour desorption interval.

For all compounds tested, the, decreased with increased aging time. This means a
greater portion of chemical was bound “irreversibly” for the samples that were aged
longer. The largest change in 123,, fi'om 24 hours to lmonth was for AT (28.6 %) (Table

3.4 and Figure 3.5). The largest change in fdcs from the 24 hour aging period to the 14

63

month period was for EDB (46.1 %) (Table 3.4 and Figure 3.6). For EDB, fde, = 0.36 for
the 14 month aging period, meaning that nearly 64 % of EDB was bound irreversibly to
soil after 14 months of aging followed by 4 hours of desorption (Figure 3.6). There was
very little, if any, increase in this value even after desorption was extended to 4 months
(Figure 3.7). The fdc, values are summarized in Table 3.4.

For this study, when comparing desorption data among all compounds, a four
hour desorption interval was utilized so that the sorption period substantially exceeded
the desorption period. This was done to minimize significant re-sorption occuning along
with desorption, since these processes occur simultaneously, especially for the 24 hour
aged samples. Essentially, this is the method of initial rates for calculating rates of
chemical reactions (Atkins, 1994).

There does not appear to be any effect of aging on the fast or slow desorption rate
constants (kI or k2, respectively) for any compounds tested (Table 3.5). In all cases, the

rate constant is the same for each aging period, or in the case of EDB, the variability in k

Table 3.4 Desorption data (fdes at 4 hours of desorption) after 24 hours, 1, and 14 months.

 

 

fag, , 24 hr. I'd”, 1mo. % decrease In”, 14 mo. % decrease
(24 hr. to 1 m2.)_____(21hL_Lo_L‘Lms.)
EDB/Cap. A .67 .60 11.4 % .36 46.1 %
CB/Cap. A .83 .66 19.9 % .60 27.0 %
AT/Cap. A .79 .56 28.6 % ND. ND.
24D/Cap. A .90 .68 24.7 % .58 35.3 %

k

Table 3.5. Desorption equation (Fa-em”) + boe("‘2”)) parameters after 24 hours, 1
month and 14 months of aging.

 

EDB 24 hour
1 month
14 month

CB 24 hour
1 month
14 month

AT 24 hour
1 month
14 month

2,4-D 24 hour
1 month
14 month

 

a b krflflJ—kzfllri)
0.16 0.84 31.3 i 13 2.06-15 i 46-3
0.10 0.90 301.2 i' 4e+4 5.8e-3 i 36-4
0.05 0.95 287.5 i‘ 264-5 386-4 i 26-3
0.22 0.78 24.0 i 3 1.16-2 i 3e-3
0.20 0.80 23.0 i 5 3.7e-3 i 58-3
0.17 0.83 30.5 i 6 266-4 :1: 3e-3
0.27 0.73 21.4 i 4 1.26-2 i 76-3
0.15 0.85 24.7 :1: 4 696-3 i 26-3
N.D. N.D. N.D. N.D.
0.27 0.73 15.7 :1: 5 1.46-2 :1: 16-2
0.20 0.79 11.9 i 5 3.46-15 i le-2
0.15 0.85 9.5 i 1 6.16-3 i 26—3

 

65

is too large to draw any conclusions (see Table 3.5).

Aging had a dramatic effect on the distribution of the chemical within the soil
matrix. For all compounds tested, the fraction of compound residing in the fast
compartment decreased progressively with aging time (Table 3.5) and the fraction in “b”
increased. The fraction in “b” is similar to the “irreversible” fraction (1- fdes)- The
largest decrease in the fraction in the fast compartment (a) between 24 hours and 1 month
was for AT (45.4 % decrease ). The largest decrease between 24 hours and 14 months
was for EDB (68.1 % decrease).

Discussion

We observed a compound specific effect of aging on the apparent sorption
equilibrium. The effect was largest for EDB and was not observed for CB. EDB showed
a 74 % increase in Kp between 24 hours and 14 months equilibration, but had no
measurable increase between 24 hours and 1 month. The Kp for AT actually decreased
slightly between 24 hours and 1 month equilibration; chemical degradation prohibited
evaluation of the K9 for AT aged for 14 months. The AT desorption curves generated
utilizing 1 and 4 month desorption intervals show clear evidence of re—sorption,
indicating that sorption equilibrium had not been reached at 1 month. Based on this, it
seems likely that the decrease in Kp for AT at 1 month as compared to 24 hours was a
manifestation of sorption non-equilibrium. In contrast, 2,4-D had a statistically
significant increase in the apparent 24 hour Kp at both 1 month and 14 months of aging
(see Table 3.2). These data demonstrate that for some compounds the 24 hour Kp may
underestimate the degree of sorption in soils.

There is very good agreement between the experimental K,m and K0m values

66

calculated from empirical relationships between K0w and K0m (Table 3.2). Measured Kom
values utilizing different soils can be expected to vary up to 2 to 3 times (Kile, 1995).
There is also good agreement between the values generated here and literature values,
except in the case of EDB. For CB, a K0", of 47.86 L/Kg has been reported by Chiou, et
al. (1983), which is very close to the value of 72.56 L/Kg reported here. Reported K0m
values for AT vary widely from 27 to 230 L/Kg (converted from K0,) (Montgomery,
1993), but the values reported here are within this range. Reported K0m values reported
for EDB do not agree as well as those for the other compounds. Values between 21.1 to
94 L/Kg are reported by Montgomery (1993) and between 59.9 and 78.0 L/Kg are
reported by Steinberg, et al. (1 98 7). These values are much larger than the values
reported here, by an order of magnitude in some cases. The earlier reported K0m values
for EDB seem very high given its relatively low K0W value. An empirical equation
(Chiou, et al., 1983) based on Kow predicts a K0m of 12.77 L/Kg (Table 3.2). The
experimental value of 6.49 L/Kg reported here is much closer to the expected Kom for
EDB than the values reported earlier.

Results from this study demonstrated no clear increase in non-linearity with
aging. The CB sorption isotherm is more linear at 14 months (n=0.905-0.908) and 1
month (n=1.00-1.03) than at 24 hours (n=0.864-0.893), whereas the EDB isotherm is
most linear at 24 hours (n=0.966-0.995) and least linear at 1 month (n=0.827-0.884).
These results are in agreement to a previous study using chlorobenzene on several soil
types (Sharer, 1999) showing no clear increase in non-linearity with increased
aging time. These results are in contrast to those of Weber and Huang (1996) and Xin g

and Pignatello (1996). Xing and Pignatello (1996) observed increases in non-linearity of

67

00:0: 8%: :808 3 0 :80 80: vm 0 80:0 D-WN :0 8:08: 8038803 .m.m 08mm:

.0505 08:.

 

 

 

 

 

 

 

 

 

m a F d a 8

8d _ _ _ _ _ 8d
mod I I 8d
8d I 8. u as: I 8d
mod I 8.1. as: a. I 8d
Ed I I Ed
Ed I I Ed
8.0 I I 8d
8d I I 8d
8d I 08. n N._ 08. u n. I 8d
8.0 I o:...o< 5:08 3 oEo< .50: cm I 8.0
8.0 8.0

”as u! Bugugewaa uogroergl

68

1.00

 

 

 

 

 

 

 

'3 .
In 0-95 7 1 month Aging
.8
d ”-90 1 r’ = .965
.5 -
.8 0.85 a r
I!
E l l
o 0.80 — 9 1
r: l
C
0 0.75 a 1
0:. fdoo 3 .680
0
g 0.70 _
II.

0.65 _

0'60 I I T 1 I

o 1 2 3 4 5

Time (hrs.)

Figure 3.3. Desorption kinetics of 2,4-D on Capac A after a 1 month aging period.

isotherms for dichlorobenzene (DCB) and dichlorophenol (DCP) between 1 and 180
days. This was attributed to the fact that SOM has both partition and adsorption domains
(dual-mode sorption) and that the adsorption domains are reached only given long soil-
chemical contact times since these domains are internal in the SOM matrix (Xing and
Pignatello, 1996). The adsorption component gives the isotherm increasing non-linearity
that only becomes apparent with long soil-chemical contact times. The isotherms
observed in this study do not exhibit consistent increases in non-linearity with increased
contact time, so this data would not appear to support a dual-mode sorption mechanism.
The linear and Freundlich sorption models were used to fit the experimental data.
Visual inspection of the data show the isotherms to be very linear. Although, in some

cases, Freundlich coefficients indicate some non-linearity (i.e. 1 month EDB sorption

69

8:0: comm :0 3808: 005880: 0 80:: < 0093 :0 Sim :0 00:08: 80080009 in 08?:

8.505 08:.

 

 

 

 

 

88 88 88 88 d8: 83 d8 0 88 88 88 d8: 82 08 o
00.0 _ P E F p _ : _ _ r _ _ _ _ 00.0
and I I and
. I 000 J.—
ood I no I .— I 00.0 m
19
8d I .u a... I 8d m.
3. .— W m u
Ed I I E d W.
Ed I I Ed m.
U
8d I I 8d m.
I 6
Ed I I 8d m.
. I .I I . S
80 8.0.": 08 Ia. 80 MI.
80 I 053 5:9: 3 053 52. E I as

 

 

 

 

 

ooé #1 oo...

70

.5008 _ 05 8:0: vm 00 £02.50 mfiwd 00cm I_I< 00 80003 gun—omen .m.m oSwE

3.505 0.5...

 

 

 

 

 

 

 

 

 

 

m w m d v 8

8d _ _ L r _ 8d
mod I 8.. u :3 r 8d
Ed I on. u .8. I 8d
Ed I I 2d
8d I I 8d
8d I I 8d
8d I I 8d
mad I 0:34. >2. On mEm< .50: VN I mad
8.? f 84

nos u! Bugugewaa uonomg

71

88:08 3 000 8:0: vm 00.0 wEwd Bad mam .00 800:3 0039809 .06 0SwE

3.505 08:.

 

 

 

 

 

 

 

 

o v m N r d w m
0.50 _ _ _ H _ _ _ 0N.0
. I . I .
B d 3.. an". m» 0
8.0 I 8:2 5:2: 2 . s. r as
no u 0 fi
00d 0 F I 8d
8. .I.. .20
8d I I 8.0
000.0 nu.
00.0 I w/WI/W/WIW r 00.0
mEm< .50: Va
8.? 4 8.0

 

nos u! Bugugewaa u0!132.l_-]

72

.050: owe 00 00 00 202000: 000000800 00% mam .00 000002 000000009 SM 05me

$505 08:.
2: d8 8... 8.. d8 d8 80 d 2: d8 8.0 8v 8 d8 2: d 87

 

 

 

 

 

 

 

 

 

05.0 _ _ _ _ _ _ _ _ _ _ _ _ 0 _ 05.0
3
Ed I I Ed m
3
9
85a. 5.5:. 3 o
and I 1 000 U
“a
a
8d I 00d w
W M
u
8d I T 8d 6
ml
mad I T 0 8d mu
m:_m< .50: vu H

8.0 t 8.0

73

isotherm, n = 0.827-0.884), the r2 values for the linear model are never below 0.9932. If
the 95 % CI. of "n" for the Freundlich model includes one, the linear model is more
appropriate statistically because the extra parameter becomes unnecessary. Conversely,
if one is outside the confidence interval, then the use of the more parameterized
(F reundlich) model is justified statistically. For this data set, in some cases the linear
model is more appropriate and in some cases the F reundlich model is more appropriate.
We are unable to conclude that aging time has more of an effect on the sorption
behavior of polar (Type II) compounds (Brusseau and Rao, 1989a) than on non-polar
(Type I) compounds, such as CB and EDB. Although there was no effect of aging on the
sorption coefficients of CB, there was a dramatic effect on EDB. For the polar (Type H)
compounds, 2,4-D had an increase in sorption afier 1 month, whereas AT did not. EDB
has a lower Kp value than CB and both are Type I compounds, and therefore, according
to Brusseau and Rao ([9890), EDB should have faster sorption kinetic behavior. They
developed linear free energy relationships (LFER’s) which predict that compounds with
higher sorption coefficients (KP) should have slower sorption/desorption behavior. Their
data predicts that EDB should have a log k (sorption rate coefficient) approximately 5
times larger than CB. Their data also predicts that for Type I and Type H compounds
with similar Kp values, the Type 11 compound should have a log k approximately two
orders of magnitude smaller than the Type I compound. This behavior was not observed
here and hence it is apparent that these generalizations do not always explain the behavior
of specific compounds such as EDB. Likewise, these concepts do not explain why 2,4-D
showed an increase in sorption after 1 month whereas AT did not.

One of the most interesting aspects of this study is the difference in

74

sorption/desorption behavior between CB and EDB upon aging. Although there was no
increase in apparenth for CB with increased aging time, a large increase in Kp was
observed for EDB subject to the same aging regimen. These results are very surprising
given that both compounds are halogenated hydrocarbons, are relatively non-polar, and
have similar molecular volumes (0.144 nm3 and 0.170 nm3 for EDB and CB,
respectively). Furthermore, using the Wilke-Chang relationship, calculated diffusivities
in water are nearly equal for these compounds (9.22 x 10’6 cmZ/sec for CB; 10.15 x 10'6
cmz/sec for EDB) (Wilke and Chang, 1955). Based on these physico-chemical
characteristics, one would predict a priori that these compounds should behave similarly
in soil. If any differences in sorption characteristics existed, one would predict that CB,
rather than EDB, would have an increase in apparenth due to aging. Previous
researchers have concluded that increasingly hydrophobic compounds (water solubilities
of CB and EDB are 500 and 4250 mg/L, respectively) approach sorption equilibrium
more slowly. Our observation that CB equilibrates very rapidly and EDB very slowly,
contradicts this relationship and suggests that compound specific interactions may be an
important determinant of sorption kinetic behavior. Interestingly, EDB has been shown
to have unexpected persistence in field soils (Steinberg, et al., 1987) and results presented
here indicate that EDB exhibits sorption behavior that is unpredictable from its physical
and chemical properties.

Spectral evidence has shown that compounds like EDB and 1,2-DCA exist as
different conformational isomers in soils. This may help in part to explain the
unexpected behavior of EDB observed in field studies. These compounds are known to

associate more strongly with clay surfaces in the gauche conformer (Aochi, et al., I 992)

75

and with humics in the anti conformer (Aochi and Farmer, 199 7). Sorption of these
molecules is thought to occur by a pore filling mechanism in soils as deduced from
spectroscopic evidence (Aochi and Farmer, 1997). This pore filling mechanism has been
used to help explain the resistant fraction of DCA and EDB in soils and clays. The
resistant fi'action of DCA in clay minerals appears to be consistent with the existence of
the chemical in a narrow network of pores where it experiences intense electrical fields
(Aochi and Farmer, 1995). It is possible that the increased rotational freedom about the
C-C single bond of this molecule enables it to reach a lowered (stable) energy state in
soils, which could account for the unexpected sorption/desorption behavior of EDB. A
relatively inflexible, rigid molecule like CB may not be able to reach this more relaxed
energy state in soils.

Desorption data shows that there is an increase in the “irreversible” (non-
desorbable) fraction (or a decrease in I“. ) with increased aging time for all compounds
tested (Table 3.4). This is in agreement with observations from a previous study done on
CB (Sharer, 1999) where the irreversible fraction of CB was found to increase on four
soil types with increased aging time. Other studies have also found that the irreversible
fraction increases with equilibration time (Pignatello, 1990; Pavlostathis and Mathavan,
1992). There does not appear to be any correlation between the magnitude of f4” and
compound type. Although at 1 month AT and 2,4-D show the largest decrease in {4“, at
14 months EDB has the largest decrease and CB the smallest (see Table 3.3).
Furthermore, the actual fraction desorbed afier the one month aging period is very similar
among compounds. After 14 months of aging, EDB has the largest bound fraction.

Based on the work of Brusseau and Rao (1989a) indicating that polar compounds are

76

subject to slower sorption/desorption behavior, our hypothesis was that there would be a
greater “irreversible” fraction due to aging for 2,4-D and AT than for CB and EDB.
However, these data indicate that aging does not have more of an effect on the desorption
behavior of polar compounds such as 2,4-D and AT than on non-polar compounds.

The distribution of chemical within the fast (a) and slow (b) compartments in soil
does change with aging, with an increase in the slow fi'action for longer periods of aging
for all compounds tested. In contrast, the kinetic data (k1 and k2) shows that aging has no
effect on the rate of desorption. These results are in contrast to McCall and Agin (1985)
who observed no change in the fast desorption rate constant with aging time, but there
was a decrease in the slow desorption rate constant. Their data was for picloram, a
relatively polar compound, aged in soils for 1 and 200 days. These results also contrast
those of Carmichael, et al. (1997) which found that the desorption rates (not rate
constants) for fresh and field contaminated PAH's were different by several orders of
magnitude. It should be mentioned that although k1 is very difficult to measure
accurately by the method described here, it seems apparent that if aging does have an
effect on k1, the effect must be very small and perhaps negligible. As was also the case in
a previous study (Sharer, 1999), it was found that in most cases a substantial fraction of
the compound was not desorbed even for samples aged for only 24 hours and desorbed
for up to four months. This observation is consistent with the findings of Kan, et al.
(1994) and Fu, et al. (1994) who showed that the desorption behavior of PAH's in soils
was irreversible. The size of the irreversible fraction of PAH’s was found to be 30-50 %
of the total mass applied after seven days of sorption and up to hundreds of days of

desorption. It is interesting to note that 2,4-D was the only compound that demonstrated

77

sorption reversibility after 24 hour contact times (0 98 % desorbs afier 4 months of
desorption) (Figure 3.2). However, the sorption reversibility disappeared upon aging
with the soil. Afier 1 month, 0 92 % of 2,4-D desorbed and after 14 months of aging,
only 0 73 % desorbed (Figures 3.3 and 3.4). The desorption resistant fraction seen in the
24 hour samples is an indication that diffusion may not be the main mechanism
controlling desorption behavior, as these samples would be expected to show ideal (or
near ideal) sorption/desorption behavior afier sorption for only 24 hours and desorption
for four months.

This study has demonstrated that aging has an effect on sorption for some
compounds but not for others on a common soil type. EDB and 2,4-D show clear
indications of an increase in sorption, whereas CB does not, and AT does not for the 1
month sorption period. Indirect evidence indicates that AT sorption does increase after a
longer period of equilibration. Aging has no effect on the desorption rates of any
compound tested. Aging does affect the amount of a compound desorbed with an
increase in the irreversible (non-desorbable) fi'action with increased aging time. There
does not seem to be a correlation between the extent of desorption and compound type,
indicating that aging does not have more of an effect on the desorption of 2,4-D and AT
than on CB and EDB. Aging also has an effect on the distribution of a chemical between
the fast and slow sites in soil, with a greater portion residing in the slow fraction for
longer aging times. This trend was observed for all chemicals tested. It seems clear that
the behavior of chemicals in soils is compound specific and can not always be predicted
from physical-chemical characteristics. A much larger data set is needed to ascertain

whether the behavior of EDB is anomalous due to some unknown chemistry or if many

78

compounds exhibit behavior that is unpredictable due to aging. Based on our results
generalizations about effects of aging on the sorption behavior of organic chemicals in

soils should be avoided.

List of Referen ces

Adamson, A.W. 1990. Physical Chemistry of Surfaces. John Wiley & Sons, Inc., New
York, NY.

Aochi, Y.O., Farmer, W.J., and Sawhney, B.L. 1992. In situ investigation of 1,2-
dibromoethane sorption/desorption processes on clay mineral surfaces by diffuse
reflectance infiared spectroscopy. Environ. Sci. Technol. 26:329-335.

Aochi, Y.0. and Farmer, W.J. 1995. Spectroscopic evidence for the rate-lirnited
accumulation of a persistent fraction of 1,2-dichloroethane sorbed onto clay minerals.
Environ. Sci. T echnol. 29:1760-1765.

Aochi, Y.O. and Farmer, W.J. 1997. Role of rnicrostructural properties in the time-
dependent sorption/desorption behavior of 1,2-dichloroethane on humic substances.
Environ. Sci. T echnol. 31:2520-2526.

Atkins, P.W. 1994. Physical Chemistry. W.H. Freeman and Co., New York, NY.

Ball, W.P. and Roberts, P.V. 1991. Long-tenn sorption of halogenated organic
chemicals by aquifer material. 1. Equilibrium. Environ. Sci. Technol. 25:1223-1236.

Brusseau, M.L. and Rao, P.S.C. 1989a. The influence of sorbate-organic matter
interactions on sorption non-equilibrium Chemo. 18:1691-1706.

Brusseau, M.L. and Rao, P.S.C. 1989b. Sorption non-ideality during organic
contaminant transport in porous media. CRC Crit. Rev. Environ. Control. 19:33-99.

Brusseau, M.L. and Rao, P.S.C. 1991. Influence of sorbate structure on
nonequilibrium sorption of organic compounds. Environ. Sci. T echnol. 25: 1501-1506.

Brusseau, M.L., Jessup, RE, and Rao, P.S.C. 1991. Nonequilibrium sorption of
organic chemicals: Elucidation of rate-limiting processes. Environ. Sci. Technol.
25:134-142.

Cameron, D.A. and Klute, A. 1977. Convective-dispersive solute transport with a
combined equilibrium and kinetic adsorption model. Water Resour. Res. 13:183-188.

79

Carmichael, L.M., Christman, RE, and Pfaender, F.K. 1997. Desorption and
mineralization kinetics of phenanthrene and chrysene in contaminated soils. Environ.
Sci. T echnol. 31:126-132.

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

Chiou, C.T., Porter, P.E., Schmedding, D.W. 1983. Partition equilibria of non-ionic

organic compounds between soil organic matter and water. Environ. Sci. T echnol.
1 7 :227-23 1 .

Engebretson, R.R., Amos, T., and von Wandruszka, R. 1996. Quantitative approach
to humic acid associations. Environ. Sci. T echnol. 30:990-997.

Fu, G., Kan, A.T., Tomson, M.B. 1994. Adsorption and desorption hysteresis of
PAH’s in surface sediment. Environ. T oxicol. Chem. 13:1559-1567.

Hayes, M.H.B. 1970. Adsorption of triazine herbicides on SOM, including a short
review on SOM chemistry. Res. Reviews. 32: 131-174.

Kan, A.T., Fu, G., and Tomson, M.B. 1994. Adsorption/desorption hysteresis in
organic pollutant and soil/sediment interaction. Environ. Sci. T echnol. 28:859-867.

Karickhoff, S.W. and Morris, KR. 1985. Sorption dynamics of hydrophobic
pollutants in sediment suspensions. Environ. T oxicol. Chem. 4:469-479.

Kile, D.E., Chiou, C.T., Zhou, H., Li, H., Xu, 0. 1995. Partition of non-polar organic
pollutants from water to soil and sediment organic matter. Environ. Sci. T echnol.
29:1401-1406.

Li, G. and Felbeck, G.T. 1972. A study of the mechanism of atrazine adsorption by
humic acid from muck soil. Soil Sci. 113: 140-148.

McCall, P.J. and Agin, G.L. 1985. Desorption kinetics of picloram as affected by
residence time in the soil. Environ. Toxicol. Chem. 4:37-44.

Montgomery, J. 1993. Agrochemicals desk reference: Environmental data. Lewis
publishers, Chelsea, MI.

Nearpass, D.C. 1976. Adsorption of picloram by humic acids and humin. Soil Sci.
121:272-277.

Oliver, B.G. 1985. Desorption of chlorinated hydrocarbons from spiked and
anthropogenically contaminated sediments. Chemo. 14: 1087-1 106.

Pavlostathis, S.G. and Mathavan, G.N. 1992. Desorption kinetics of selected volatile

80

organic compounds from field contaminated soils. Environ. Sci. T echnol. 26:532-538.

Pignatello, J.J. 1990. Slowly reversible sorption of aliphatic halocarbons in soils. 1.
Formation of residual fi'actions. Environ. T oxicol. Chem. 9:1107-1115.

Pignatello, J.J. and Xing, B. 1996. Mechanisms of slow sorption of organic chemicals
to natural particles. Environ. Sci. T echnol. 30:1-11.

Schulten, H.R. 1995. The three-dimensional structure of humic substances and soil
organic matter studied by computational analytical chemistry. Fresenius J. Anal. Chem.
35 1 :62-73.

Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M. 1993. Environmental
Organic Chemistry. John Wiley & Sons, Inc., New York, NY.

Sharer, M. 1999. Effects of aging on the sorption/desorption of chlorobenzene on soils
with a range of organic carbon content. Chapter 2. PhD Thesis. Michigan State
University. East Lansing, MI

Steinberg, S.M., Pignatello, J.J., and Sawhney, B.L. 1987. Persistence of 1,2-
dibromoethane in soils: Entrapment in intraparticle micropores. Environ. Sci. T echnol.
21:1201-1207.

Weber, W.J., J r., Huang, W. 1996. A distributed reactivity model for sorption by soils
and sediments. 4. Intraparticle heterogeneity and phase-distribution relationships under
nonequilibrium conditions. Environ. Sci. Technol. 30:881-888.

Wilke, CR. and Chang, P. 1955. Correlation of diffusion coefficients in dilute
solutions. A]. Ch.E. Journal. 1:264-270.

Wu, S. and Gschwend, P.M. 1986. Sorption kinetics of hydrophobic organic
compounds to natural sediments and soils. Environ. Sci. T echnol. 20:717-725.

Xing, B. and Pignatello, J .J. 1996. Time-dependent isotherm shape of organic

compounds in soil organic matter: Implications for sorption mechanism. Environ.
Toxicol. Chem. 15:1282-1288.

81

Chapter 4

EFFECTS OF AGING ON THE BIOAVAILABILITY AND SORPTION

IDESORPTION BEHAVIOR OF BIPHENYL IN SOILS

Abstract

Aging (soil-chemical contact time) is known to lead to unexpected persistence of
organic contaminants in the environment. A highly mobile and easily degradable soil
fumigant, EDB, has been found in agricultural field soils almost twenty years after its last
known application. The effects of aging on bioavailability, one of the main variables for
determining the persistence of organic chemicals in the environment, is a poorly
understood process. There are very few studies that have tested the effects of long-term
aging on chemical dynamics in soils and have related these effects to bioavailability tests
of these aged compounds. In this study, sorption/desorption behavior of biphenyl (BP)
on two soil types (Capac A and Capac B) was evaluated for aging times of 24 hours and 8
months. Then bioavailability experiments of BP on the same soils were performed after
aging times of 24 hours to 6 months. Sorption isotherms and desorption kinetic profiles
were prepared to analyze changes in uptake and release, respectively, due to aging of BP
from a physico-chemical standpoint. Mineralization kinetics of BP to 14C02 by a strain
of Pseudomonas putida was used to assess changes in bioavailability due to aging from a
biological standpoint. Data indicates that there is an increase in sorption with aging time
for BP on both soil types. The rate of BP desorption does not change with increased
aging time. The extent of BP mineralization was found to decrease with aging time.

Keywords — Bioavailability, biphenyl, sorption, desorption, aging, soils

82

Introduction

Sorption/desorption behavior of organic chemicals in soils is an important
determinant governing the bioavailability of these compounds to bacteria. Limited
bioavailability of these compounds may cause increased persistence in natural
environments such as soils. Bioavailability also determines the toxicological and
herbicidal activity of organic contaminants and pesticides in the environment. The
bioavailability of a compound can be limited by sorption to soils which decreases the
extent of degradation by microbes (Wszolek and Alexander, 1979; Steen, et al., I 980).
Most researchers have assumed that before a sorbed molecule can be degraded by a
bacterium, it first has to diffuse to the solid-liquid interface, then partition into the
aqueous phase, where it is available for microbial degradation (Harms and Zehnder,
1995). The conclusion that sorbed substrates are unavailable to bacteria, and that
desorption is a prerequisite for degradation, have been validated in studies by Ogram, et
al. (1985) and Steen, et al. (I 980). Other researchers (Guerin and Boyd, 1992) have
concluded that generalizations regarding availability of sorbed contaminants are
inappropriate. They provided evidence for the direct access of soil-sorbed naphthalene
by one species of bacteria but not another. This may be because of organism specific
properties such as attachment, motility, or affinity for the particular substrate.

In cases where bacteria can only use dissolved substrates, sorption can limit
bioavailability in two ways: 1) Decreasing aqueous phase concentrations to low levels
due to sorption and thereby limiting rates of degradation, and 2) Limiting the extent of
biodegradation due to small desorption rates (a desorption limited process) (Harms and

Zehnder, 1995). These factors will limit the initial biodegradation rate but limited extents

83

of biodegradation can be overcome if bacteria are able to enhance desorption. Several
studies have concluded that bacteria are able to enhance desorption, indicating that
desorption is slow relative to biodegradation and that biodegradation increases

desorption. Harms and Zehnder (1995) found that 3-chlorodibenzofi1ran desorption was
driven by cells attached to teflon sorbents. They concluded that bioavailability was
dependent on the specific affinity of a bacteria for a specific substrate and the tendency of
that organism to adhere to a surface. Rijnaarts, et a1. (1990) also found that microbes
enhanced desorption and that initial rates of biodegradation were consistently higher than
initial desorption rates. The study of Guerin and Boyd (1 992) found that one bacterial
species was able to enhance desorption whereas another species was not.

Because many field soils have been contaminated for periods of months to years,
conclusions from short-term bioavailability studies may be inappropriate for situations
where the soil-chemical contact time (aging) is long. Observations from field studies
have demonstrated that aged chemicals are less bioavailable than fi'eshly added ones
(Steinberg, et al., I 98 7; Scribner, et al., 1992). A study by Weissenfels, et al. (I 992)
demonstrated that the strength of PAH sorption to soils, which may increase with aging,
decreased the amount degraded. They also concluded that freshly added chemicals do
not accurately mimic the behavior of aged chemicals. Hatzinger and Alexander (1995)
showed that the extent of freshly added phenanthrene mineralized was greater than the
extent mineralized for aged phenanthrene in a muck soil (11.22 % OC). They found no
difference in the extent mineralized for freshly added (0 days) and aged (84 days)
phenanthrene in a loamy soil (2.33 % OC). Rates of degradation were decreased for aged

phenanthrene on both the loam and the muck. In a sandy soil (1.34 % OC), they found

84

no difference in the rate or extent of degradation for aged versus freshly added
phenanthrene. 4-Nitrophenol was added to the loam and muck soils and mineralization
was assessed at 0, 40, and 103 days. In both soil types, rates and extents of
mineralization decreased with increased aging time. These results show the effect of
aging on biodegradation and indicate that in many cases aging has an effect on both the
rate and extent of biodegradation.

One study has explicitly measured both desorption and biodegradation rates for
freshly added and aged chemicals (Carmichael, et al., 1997). Desorption rates of
chrysene and phenanthrene (both PAH’s) were found to be much faster than observed
mineralization rates for fi'eshly added compounds in this study. They also found that
desorption rates for field-aged chemicals were equal to or slower than mineralization
rates, so it was concluded that desorption of aged PAH’s may control their fate and
availability. This may explain why the persistence of PAH’s has been observed even in
soils that have active populations of PAH-degrading organisms. These results are in
agreement with those of Guerin and Boyd (1993) which found that initial mineralization
rates (IMR’s) for aged compounds were slower than for freshly added.

In order to better understand the effects of aging on bioavailability and on the fate
of chemicals in soils, we initiated a long-term aging study explicitly examining sorption,
desorption, and biodegradation in soils. In this study, sorption of biphenyl on Capac A
and B soil horizons after was evaluated 24 hours and 8 months of aging. Also, biphenyl
desorption rates were measured after these periods of aging. Finally, biphenyl
biodegradation in these soils using a Psuedomonas species was evaluated to observe the

effects of aging on bioavailability. Results from the biological and chemical assays were

85

then compared in an effort to better understand the effects of aging on both biological and

physico-chemical processes.

Experimental Section

Soils used in this study were Capac A and B horizon (fme-loamy, mixed, mesic
Aeric Ochraqualf ) soils collected from southern Michigan. These soils were prepared
using the same methods as in chapter 2. Capac A has a sand, silt , and clay content of
54.6, 24, and 21.4 %, respectively, and an OC content of 3.3 %. Capac B has a sand, silt,
and clay content of 60, 20, and 20 %, respectively, and an OC content of 0.4 %. All soils
were then gamma irradiated with 5 MRad at the University of Michigan Ford Nuclear
Reactor Laboratory using a cobalt-60 irradiator (dose rates by Reuter-Stokes ion
chamber). Soils were stored in 500 mL Nalgene bottles during irradiation and remained
sealed until use.
W The amounts of a particular soil and solution used to develop
sorption isotherms were based on preliminary sorption experiments. The linear isotherm
equation (x/m = KCe) was employed to estimate the amount of soil and solution needed
to sorb 50 % of the added BP while also minimizing headspace volume. The tubes used
were 20 mL glass ampules which were autoclaved prior to use. All additions of soil and
solution to the ampules were carried out in a sterilized laminar flow hood. A mass of soil
was weighed and transferred to the ampule using a metal spatula along with a plastic
funnel. A 0.005 M CaClz solution, which had been autoclaved for 25 minutes and
amended with 200 mg/L NaN3, was added to the ampules using a sterilized repipettor.

These functions were carried out in the laminar flow hood and all items contacting soil or

86

solution were autoclaved prior to use (spatulas, fimnels, etc.).

Additions of the 14C-Biphenyl (BP) (radiochemical purity > 98 %; Sigma
Chemical Co.) stock solution (in acetone) and flame sealing of the ampules were carried
out in a chemical fume hood. An aliquot of BP stock solution was added to the ampule
using a 10 or 25 ul Hamilton syringe and the ampule was immediately flame sealed using
an Ampulmatic flame sealer unit (Bioscience, Inc.). Aliquots used for the sorption
isotherms were 2, 5, 10, and 20 ul of a low concentration BP stock and 10, 25, and 70 ul
of a higher concentration BP stock solution. The low concentration stock solution was
initially 2.84 uCi/mL 1‘tC-BP and the high concentration stock solution contained initially
16.6 uCi/mL l4C-BP. The stock solutions were stored in a -4° C freezer in amber reacti-
vials with mininert valves when not in use. Each of the seven sorption isotherm points
were done in triplicate. After flame sealing, the ampules were equilibrated (“aged”) for
either 24 hours or 8 months. To insure homogeneity, the tubes were shaken
(reciprocating shaker) for either 24 hours (for 24 hour aging period) or 72 hours (for 8
month aging). After shaking, the tubes were either analyzed immediately (24 hour aging)
or stored in the dark with periodic shaking (8 month aging). For analysis, each ampule
was centrifirged at 635 x g for 7-9 minutes, the ampule broken at the neck, and a lmL
aliquot of the aqueous phase was added to 8 mL of scintillation fluid (RPI, Inc., Safety-
Solve). The samples were shaken and the next day analyzed by liquid scintillation
counting (Beckrnan LS 65 00 multi-purpose scintillation counter). Values for each
triplicate set of ampules were averaged and standard deviations calculated. The data
were plotted as the amount sorbed (ug/Kg) versus concentration in solution (ug/L).

Linear regression was used to compute the slope of the line, which corresponds to the

87

equilibrium sorption coefficient (KP). The Kp values were compared at each of the aging
periods for each soil type.

Wes; Soil and solution were added to the ampules as described above,
with the same amounts of soil and solution used as were used for the sorption
experiments. A fixed volume of BP stock solution (10 uL) was added to all of the
desorption ampules and the ampules were then flame sealed. The ampules were aged for
the appropriate time period (24 hours or 8 months) and shaken and centrifuged as
described above. Then, a quadruplicate set of ampules were opened by cutting the stem
(1.6 - 2.0 cm from the top of the ampule) using a lAI-inch glass drill bit (Sommer & Maca,
Co.), then sampling the solution phase (200 uL) using a 200 uL Gilson Pipetrnan. This
sample was analyzed using scintillation counting (after 24 hours) to arrive at the
equilibrium concentration in solution (Co) before desorption was initiated. Then a larger
aliquot of the solution phase was removed (0 75-80 % of the total volume of solution)
using a 25 mL gas-tight Hamilton syringe. The same volume of fresh solution
(autoclaved 0.005 M CaClz with 200 mg/L NaN3) was then added to the ampule, again
using the 25 mL Hamilton syringe. The syringe was rinsed with clean solution between
withdrawal of the old solution and addition of the new. The ampule was then re-flame
sealed and shaken by hand. At the appropriate time interval, the quadruplicate set of
tubes was centrifuged at 635 x g (7-9 minutes) and the remainder of the stems were
broken at the neck and a 1 mL aliquot of the aqueous phase was sampled and analyzed by
scintillation counting (afler 24 hours); this gives the concentration of BP that desorbs in

solution after a given desorption time (Cf). The amount desorbed is: Cf - [Co + (the mass

of BP remaining in the residual solution) -:— (volume of residual solution)]; since only 75-

88

80 % of the total solution volume was removed, an appreciable amount of solution
remained, the residual solution, which mixed with the fresh solution added to initiate
desorption. Desorption time intervals were: 0.083, 0.5, 1, 2, 4, 8, 12, 24, 72, 216, 648,
and 2880 hours. Since four ampules were used for each desorption time interval, a total
of 48 ampules were prepared in the same manner and incubated for the 24 or 8 month
aging period. The concentration desorbed Cf - (Co + residual BP) was converted to the
fraction of the total remaining in the soil (at t=0, the fraction in the soil is 1 and the
fraction desorbed is 0).

MW Sorption isotherms were analyzed for linearity using non-linear
regression analysis. Isotherms were fit to the Freundlich model (x/m = KOCe“) and the
95 % confidence interval (CI) for n was generated for the model. Isotherms were also
analyzed for differences in slope between 24 hour sorption and 8 month sorption. This
was accomplished by performing linear regression analysis with indicator variables. The
variability in the model coefficients were tested for significance at the 0.99 level.
Nonlinear regression analysis was performed on the desorption data using Si gmaplot
graphing soflware. Parameters for the desorption equation were estimated using the
Marquardt-Levenberg algorithm and standard deviations of each parameter are computed.
Wigwam: To evaluate the availability of soil-sorbed biphenyl to the
biphenyl degrader (Psuedomonas putida P106), a mineralization assay, adapted from the
method described by Guerin and Boyd (1992), was used. Soil-free controls and soil
slurries were set up in 160-ml serum bottles sealed with Teflon-coated butyl stoppers.
Sterile soil (7.5 g) (gamma irradiated as for the sorption/desorption experiments) was

placed into each bottle, and variable volumes of sterile phosphate buffer (20 mM) and

89

non-radioactive biphenyl stock solution were added to obtain the appropriate

concentrations. Approximately 0.1 uCi of 1‘fC-biphenyl was added to each serum vial
separately from the addition of “cold” biphenyl. Soil slurries were shaken for 24 hours
and then stored in dark with periodic shaking for the prescribed aging period. Soil-free

controls were prepared in sterile soil extracts. Cells harvested at early stationary phase

were used as inocula. Cell density in serum bottles was ~106 CFU/ml for strain P106.

At precise time intervals, 1 ml suspension and 1 m1 headspace were withdrawn from each
bottle and injected into to a sealed tube which served as a 1"'C02 trap. The tube

contained 1ml of 2 N HCl and was sealed with a rubber stopper. Filter paper in the tube
was soaked with 2N NaOH which was used to trap the 14C. After the assay, the filter
paper strip was transferred to a scintillation vial containing 10 ml scintillation fluid, and
the center cup which held the filter paper was washed with 2 ml 50% (v/v) ethanol to
remove any residual 14C. Samples were left overnight in the dark before being analyzed
by liquid scintillation counting.
W5; Percentage of the initial radioactivity converted to 14CO;)_ was plotted
versus time. Mineralization data were fitted to a first-order kinetic equation:

P = Pmax [l - exp(-kot)]
where P is the percentage of initial radioactivity mineralized, Pmax is the maximal
percent mineralized, k is the first-order rate constant, and t is time. The initial
mineralization rate (IMR, ugo L'lomin") is the product of Pmax, k, and initial biphenyl

concentration:

- -1
IMR = k (min 1) o Pmax o [Biphenyl] (pg-L )

90

In soil free systems, at substrate concentrations below the half saturation constant of the
organism, and with non-growing cells, mineralization rates are linearly proportional to
the substrate concentration. Thus, a linear control line was established relating IMR to
aqueous biphenyl concentrations. In soil slurries if sorbed biphenyl is unavailable to
bacteria, only biphenyl in the aqueous phase can be degraded in soil slurries. Under this
assumption, IMR values of soil slurries should be equal to those in a soil free control with
the same aqueous phase biphenyl concentration, i.e. they should fall on the soil free
control line. IMR of soil slurries above the soil-free control line, indicates that bacteria

have access to sorbed biphenyl or that desorption is rapid relative to degradation.

Results

The sorption isotherms of biphenyl (BP) on Capac A and B show that BP had a
statistically significant increase (at the 0.01 level) in sorption from 24 hours to 8 months
(Figures 4.1 and 4.2). The increase in sorption on Capac B was less than the increase on
Capac A. Sorption data were fit to the Freundlich equation, x/m = KCe", where x/m is
the concentration of sorbed BP (rig/Kg), Ce is the aqueous BP concentration (ug/L) , K is
the sorption coefficient, and n is the coefficient of linearity (Table 4.1). When sorption
data are fit to the linear form of the Freundlich equation (n = 1), the isotherms appear
linear in the range tested (12 values range from .9970 - .9997 for the linear plot).
Evaluation of the F reundlich isotherm parameters show some evidence of non-linearity (n
slightly less than 1). For Capac A, these values appeared to increase (become more
linear) from the 24 hour to 8 month aging interval. For Capac B, these values remained

essentially constant.

91

Concentration in Solution (pg/L)

 

 

 

 

0.01 a
’5
¥
G
.3-
'2

0.001 — 3
“
C
3
O
E
<
0.0001

 

Figure 4.1. BP sorption on Capac A after 24 hours and 8 months of aging.

Concentration in Solution (pg/L)

 

  

 

 

 

10—3 - a
8
mm E] f
Kp-BJO LIKg 3
19-4 — ,3
‘0'
3
o Attenuator: E
Kp-6.35 LIKg ‘1
1°‘5 7/ r2-.9994
I l l
1 10 100

Figure 4.2. BP sorption on Capac B after 24 hours and 8 months of aging.

92

Table 4.1. Freundlich isotherm (x/m = KOCe") parameters after 24 hours and 8 months

 

of aging time.
KIWI.) MI.)
BP / Cap. A 24 hour 0.099 — 0.140 0.843 — 0.914
8 month 0.115 — 0.148 0.907 — 0.961
BP / Cap. B 24 hour 0.009 - 0.011 0.905 — 0.930
8 month 0.009 — 0.012 0.888 — 0.940

 

Desorption kinetic data was evaluated using the double decay equation (y = ace“

k") + boe('k2‘)) where y is the fraction of chemical remaining in the soil, k1 and k2 are the
fast and slow desorption rate constants, respectively, and “a” and “b” are the fraction of
chemical residing in the fast (a) or slow (b) sites. Parameters from the double decay
equation are listed in Table 4.2. Also utilized in the analysis of desorption data was the
fraction desorbed (fdu), which gives a measure of the extent of total desorption. This
value was calculated by dividing the total amount observed to desorb (Tdes) divided by
the amount that should desorb assuming complete desorption reversibility (Eqdes):

fad = Tries / Eqdu
On the desorption plots (Figures 4.3-4.4), the value for Eqch can be seen as the
asymptote that should, theoretically, be approached by the desorption curve. If the
experimental desorption curve reaches the asymptote, then {m = 1. Also listed in Figures
4.3 and 4.4 are the 13 values which indicate the goodness of fit of the double decay
equation to the experimental data. The experimental data points (average of
quadruplicate values) are shown along with a plot of the regression line generated by the

double decay equation. For this study, when comparing desorption data, a four hour

93

Table 4.2. Desorption equation (Fa-em”) + boe(’k2°°) parameters after 24 hours and 8

 

 

months of aging.
.0 b Jim—kiwi)
BP / Cap. A 24 hour 0.16 0.84 13.8 i 5.3 1.7e-2 i 8.5e-3
8 month 0.16 0.84 15.2 i 2.0 9.4e-3 j: 2.8e-3
BP / Cap. B 24 hour 0.26 0.74 20.8 i 1.9 2.5e-2 i- 3.2e-3
8 month 0.22 0.78 14.6 i 3.8 2.6e-2 i 8.8e-3

 

desorption interval was utilized so that the sorption interval substantially exceeded the
desorption period. This was done to minimize significant resorption occuning along with
desorption, since these processes occur simultaneously, especially for the 24 hour aged
samples which may not have reached sorption equilibrium. Essentially, this is the
method of initial rates for calculating rates of chemical reactions (Atkins, 1994).

Aging appeared to have little effect on the desorption rate constants of BP. The
fast desorption rate constant (k1) is nearly identical on Capac A after 24 hours and 8
months of aging (13.77 :t 5.3 hr" and 15.23 i 2.0 hr", respectively). This trend also
holds on Capac B after 24 hours and 8 months ofaging (20.8 i 1.9 hr'1 and 14.57 i 3.8
M”, respectively). The slow desorption rate constant (k2) also varies little with aging (see
Table 4.2). The fraction of BP residing in the fast (a) and slow (b) compartments changes
little with aging on Capac A. The fraction in each compartment does change with aging
on Capac B (Table 4.2). Correspondingly, the fraction desorbed (8,...) after four hours
decreases for Capac B (.67 to .58) whereas fa, actually increases slightly on Capac A

with aging (fi'om .70 to .76) (see Figures 4.3 and 4.4).

94

.w0mw0 .00 0:0008 w 000 050: 0N 00000 < 00000 00 mm .00 000003 00080000 .020 05mm:

0.505 95...

 

cod

mod I

end I

E. I J

00.0

I. mod

 

 

mud

owd I

mwd I

omd I

mad I

 

 

n00.000
m:_m< 500:. w

 

E. .. a

 

050.00.
0590 .50: VN

and

I mud

I owd

I mad

I omd

I mad

 

 

oo...

oo._.

nos uI BquIewaa uonoerd

95

.w0ww0 .00 0500.0 w 000 050: Va 00:00 m 00.00 00 mm .00 00:00.: 00000000D .3... 05wE

A0505 0.0:.

 

 

 

 

 

 

 

 

 

m V m N . v m
00.0 _ _ _ _ 8 _ 00.0
Ed I I 9d
Ed I I 9d
8d I I 8d
Ed I an. H 00.00 5. n 00.0.— I 8d
8d I I 8d
8d I I 8d
Ed I I Ed
Ed I I Ed
8d I . I 8d
Ed I cam. II. N. co _. u N. I 8d
8d I 9.50 5:00. 0 0:50 .50: g I 8d
Ed I I 8d
2: oo._.

uos u! BquIewaa uoIroeH

96

Biological data for the mineralization of 1‘lC-BP to 14CO; by Pseudomonas putida
P106 is shown in Figure 4.5 and 4.6. On both Capac A and B, the extent of BP
mineralized to C02 decreases progressively with the length of aging time prior to
inoculation of the soil slurries (1 day to 6 months) with P. putida P106.

Table 4.3 gives a direct comparison of the biological and chemical data. Data
from the respective experiments were compared using the first order decay model
(Ao=AIe('k")), where A0 is the amount of chemical at t=0 hours, AI is the amount at any
point in time, t is time in hours, and k is the first order rate constant. For this study, the
first four hours of desorption were compared directly to the first four hours of
biodegradation. The biological experiments measured biodegradation for 6 hours, so
comparisons between the two systems beyond 6 hours was not possible. Four hours was
the longest common time point measured for these systems. It is clear that the rate
constant for biological degradation is much faster than the rate constant for desorption by
almost two orders of magnitude and this trend is consistent on both soil types. An
important observation is that neither the biological nor the desorption rate constant

decrease with aging, which was highly unexpected.

Discussion

Sorption isotherms of BP indicate that sorption is nearly complete on Capac B
after 24 hours, but not on Capac A. Although isotherms on both soils showed a
statistically significant (at the .01 level) increase in sorption, the increase in sorption on
Capac B was relatively small. The sorption isotherms can be fit adequately to the linear

form of the Freundlich equation (11 = 1); r2 values were > .9970. When fitting the data to

97

Time (hrs.)

 

50
40 - D 0 Q
C B {j 1:]
30, BB 8 0
O o\°
B 0
B O O O .E
20 r 6 ON
5 A 6 month 0
o 4 month 0
1O — O 2 month "
5 v 1 month
CI 2 weeks
L3 1 d
0 5 C} o Cogtyrol

 

 

 

I I I I I I I

0 100 200 300 400 500 600

Figure 4.5. Mineralization kinetics of BP to 1“CO; by P. putida after different aging
periods on Capac A.

98

Time (hrs.)

 

40
35‘— o o .
C. [j
. I: DU [5 O
25“ US$50 o ’ \°
0g .0 o
20- Q.. .5
Or A‘ 0
15- A ‘ O
‘A
A 0
A v
10— A ‘-
5...
0 — {3

 

 

 

I I I I I I I

0 100 200 300 400 500 600

Figure 4.6. Mineralization kinetics of BP to MC02 by P. putida after different aging
periods on Capac B. For legend, follow figure 4.5.

99

Table 4.3. Comparison of biological and chemical assays of biphenyl after 24 hours and

 

6 to 8 months of aging.

CthaLdata Biglegicamata
WA
24 hour 65 4.44e-2 .70 .30 1.31 .38
6/8 month“ 96 3.56e—2 .76 .24 1.45 .25
001209.13
24 hour 6.4 7.45e-2 .67 .33 1.27 .28
6/8 month“ 6.7 6.78e-2 .58 .42 1.46 .14

 

KP = L/Kg ; all data based on the first 4 hours of desorption or mineralization.
* chemical assay done after 8 months of aging, biological done after 6 months of aging
the conventional form of the Freundlich equation, there is some evidence of non-linearity
from the fact that the n values are less than 1. The values of n, however, remain constant
or increase slightly when comparing data at 24 hours of equilibration time to those at 8
months. This is in contrast to the results of Xing and Pignatello (1996) who observed
increasing non-linearity of sorption isotherms (decreases in n) for dichlorobenzene and
dichlorophenol on a sandy loam soil after equilibration times of 1 and 180 days. This
was attributed to the fact that SOM has both partition and adsorption domains (dual-mode
sorption) and that the adsorption domains are reached only given long soil-chemical
contact times since these domains are internal in the SOM matrix (Xing and Pignatello,
1996). The adsorption component gives the isotherm increasing non-linearity that only

becomes apparent with long soil-chemical contact times. The isotherms observed in this

100

study and in previous studies (Sharer, 1999a; Sharer, 1999b) do not appear to be
consistent with a dual-mode sorption mechanism.

The desorption data indicates that the desorption rate of BP is the same whether
the sample has been aged for 24 hours or 8 months on both soil types. This behavior is
consistent on both Capac A and Capac B. This is in agreement with results flom two
studies using a variety of compounds and soils (Sharer, I999a; Sharer, 1999b), although
in contrast to studies by McCall and Agin (1985) and Carmichael, et al. (I 99 7). McCall
and Agin (1985) found that the fast desorption rate constant did not decrease with
increased aging time, but the slow desorption rate constant did, for freshly added and
aged picloram. The study by Carmichael, et al. (1997) found that desorption rates, not
rate constants, decreased dramatically for flesh versus field contaminated (aged) PAH's.
The flaction of compound residing in the fast and slow compartments changed on the
Capac B soil, but remained the same on the Capac A soil. For Capac B, 26 % of sorbed
BP was located in the fast compartment after 24 hours of aging and 22 % after 8 months
of aging. For Capac A, the flaction in the fast compartment remained constant at z 16 %
and the flaction desorbed actually increased slightly with aging (flom 70 to 76 %). Both
these results are unexpected and are in contrast to almost all other data obtained by this
desorption technique. Biphenyl desorption on Capac B was characterized by a decrease
in the flaction desorbed with increased aging time. This result is consistent with the
general observations flom two previous studies (Sharer, 1 999a; Sharer, 1999b) and from
results by Pignatello (I 990) and Pavlostathis and Mathavan (1992).

The biological data showing the mineralization of l4C-BP clearly demonstrates a

progressive decrease in the amount degraded with increased aging time on both soil

101

types. For example, the amount mineralized to CO; after a 24 hour aging period is 32-33
%, and after a 6 month aging period, the amount mineralized is only 14-15 %. It would
be expected that if an increase in sorption occurs for BP due to increased aging, this
should cause a concomitant decrease in the initial rate of biodegradation of BP, due to a
lower solution concentration. This would manifest itself as a reduction in the slope of the
initial part of the mineralization curve Giigures 4.5 and 4.6). From inspection of these
graphs, there does not appear to be such a reduction in slope, except possibly in the case
of the 6 month aging of BP on Capac B. Given the fact that the desorption rates of BP do
not change with aging, a reduction in the initial rate of biodegradation would only be
expected due to a reduction in the initial solution concentration and this was not indicated
flom the sorption isotherms for Capac B. There was a decrease in initial solution
concentration for BP on Capac A due to aging. It seems likely that the decrease in initial
solution concentration due to aging was not large enough to become evident as a
reduction in the initial slope of the mineralization curves. These sorption/desorption
results are consistent with the observation that there is no reduction in the rate of
biodegradation due to aging as reflected by the mineralization curves themselves and by
the first order mineralization rate constants, which also do not change with aging (Table
4.3).

The mineralization curves of BP on Capac B are consistent with the desorption
data. The mineralization curves appear to have no initial decrease in slope with increased
aging, with the possible exception of the 6 month aged sample. This is consistent with
the sorption data which shows that sorption of BP on Capac B increases only very

slightly with aging. Also, the desorption rate of BP does not change with aging, so again

102

desorption would not be expected to influence differentially the initial slope of the
mineralization curves. There is, however, a decrease in the irreversible (non-desorbable)
flaction of BP with aging, along with an increased flaction of BP in the slow
compartment based on the modeling data. This would cause a decrease in the extent of
mineralization, consistent with our observations (Figure 4.6).

There is a decrease in the extent of mineralization of BP in Capac A after 2 and 6
months of aging; the magnitude of this decrease was considerably smaller than in the
Capac B soil. The results are difficult to resolve in terms of the sorption/desorption data
which show an increase in sorption but no increase in the irreversible (non-desorbable)
flaction with aging. The increase in sorption due to aging is greater on Capac A, so it
would be expected that there would be a larger decrease in the extent mineralized on
Capac A. However, it should be noted that the sorption/desorption behavior of BP on the
Capac A soil is unique. All other combinations of compounds and soils (Sharer, I999a;
Sharer, 1999b) show an increase in the non-desorbable flaction with aging.

A comparison of the first order desorption rate constants and biodegradation rate
constants show the latter to be two orders of magnitude greater. This suggests that the
systems described here are desorption limited (desorption slow relative to biodegradation,
Table 4.3) and that the irreversible flaction is not accessed by microbes. However,
distinguishing whether the irreversible (non-desorbable) flaction is truly irreversible or
just extremely rate-limited is difficult due to the time scale of the mineralization assay. It
may be that for the time scale of the mineralization assay this portion of BP did not
desorb, but desorption may occur on a longer time scale typical in field situations.

However, studies of field weathered soils have indicated the existence of a truly

103

biologically unavailable flaction which may correspond to the non-desorbable flactions
observed here. For example, EDB, a compound readily degraded by soil microorganisms
in non-aged laboratory assays, is known to persist in field soils for nearly 20 years after
its last application (Steinberg, et al.,198 7). Thus, our data and the observations flom
field weathered soils indicate that the non-desorbable flaction observed here may not be
accessible to microbes on any pertinent time scale. The 24 hour aging desorption data
supports this assertion. After aging for 24 hours and then desorbing for four months,
there is still an appreciable flaction that is non-desorbable for both BP and EDB (Sharer,

I999b)

List of Referen ces
Atkins, P.W. 1994. Physical Chemistry. W.H. Freeman and Co., New York.

Carmichael, L.M., Christman, RE, and Pfaender, F.K. 1997. Desorption and
mineralization kinetics of phenanthrene and chrysene in contaminated soils. Environ.
Sci. T echnol. 31 :126—132.

Guerin, W.F. and Boyd, S.A. 1992. Differential bioavailability of soil-sorbed
naphthalene to two bacterial species. Appl. Environ. Microbiol. 58:1142-1152.

Guerin, W.F. and Boyd, S.A. 1993. Bioavailability of sorbed naphthalene to bacteria:
Influence of contaminant aging and soil organic carbon content. p. 197-208. In Sorption

and Degradation of Pesticides and Organic Chemicals in Soil. SSSA special publication
no. 32. SSSA, Madison, WI.

Harms, H. and Zehnder, A.J.B. 1995. Bioavailability of sorbed 3-chlorodibenzofuran.
Appl. Environ. Microbiol. 61 :27-33.

Hatzinger, P.B. and Alexander, M. 1995. Effect of aging of chemicals in soil on their
biodegradability and extractability. Environ. Sci. T echnol. 29:537-545.

McCall, P.J. and Agin, G.L. 1985. Desorption kinetics of picloram as affected by
residence time in the soil. Environ. T oxicol. Chem. 4:37-44.

104

Ogram, A.V., Jessup, R.E., Ou, L.T., and Rao, P.S.C. 1985. Effects of sorption on
biological degradation rates of (2,4-dichlorophenoxy)acetic acid in soils. Appl. Environ.
Microbial. 49:582-587.

Pavlostathis, S.G. and Mathavan, G.N. 1992. Desorption kinetics of selected volatile
organic compounds from field contaminated soils. Environ. Sci. T echnal. 26:532-538.

Pignatello, J.J. 1990. Slowly reversible sorption of aliphatic halocarbons in soils. 1.
Formation of residual flactions. Environ. T axical. Chem. 9:1107-1115.

Pignatello, J.J. and Xing, B. 1996. Mechanisms of slow sorption of organic chemicals
to natural particles. Environ. Sci. T echnal. 30:1-11.

Rijnaarts, H.H.M., Bachmann, A., Jumelet, J.C., Zehnder, A.J.B. 1990. Effect of
desorption and intraparticle mass transfer on the aerobic biomineralization of alpha-

hexachlorocyclohexane in a contaminated calcareous soil. Environ. Sci. Technol.
24:1349-1354.

Scribner, S.L., Benzing, T.R., Sun, S., and Boyd, S.A. 1992. Desorption and
bioavailability of aged simazine residues in soil flom a continuous corn field. J. Environ.
Qual. 21:115-120.

Sharer, M. 1999a. Effects of aging on the sorption/desorption of chlorobenzene on
soils with a range of organic carbon content. Chapter 2. PhD Thesis. Michigan State
University. East Lansing, MI

Sharer, M. 1999b. Effects of aging on the sorption/desorption behavior of EDB,
chlorobenzene, 2,4-D, and atrazine. Chapter 3. PhD Thesis. Michigan State University.
East Lansing, MI

Steen, W.C., Paris, D.F., and Baughman, G.L. 1980. Effects of sediment sorption on
microbial degradation of toxic substances. p. 477-482. In R.A. Baker (ed. ) Contaminants
and Sediments, Vol. 1. Ann Arbor Sci. Publ., Ann Arbor, MI.

Steinberg, S.M., Pignatello, J.J., and Sawhney, B.L. 1987. Persistence of 1,2-
dibromoethane in soils: Entrapment in intraparticle micropores. Environ. Sci. T echnal.
21:1201-1207.

Weissenfels, W.D., Klewer, H.J., and Langhofl‘, J. 1992. Adsorption of polycyclic
aromatic hydrocarbons (PAH's) by soil particles: Influence on biodegradability and
biotoxicity. Appl. Microbial. Biatechnal. 36:689-696.

Wszolek, RC. and Alexander, M. 1979. Effect of desorption rate on the
biodegradation of n-alkylamines bound to clay. J. Agric. Food Chem. 27:410-414.

105

"IIIIIIIIIIIIIIIIIIs