Material Information

Coupled-processes interactions of contaminants, bacteria, and surfaces
Physical Description:
xiii, 170 leaves : ill. ; 29 cm.
Bellin, Cheryl A., 1963-
Publication Date:


Subjects / Keywords:
Soil microbiology   ( lcsh )
Microorganisms   ( lcsh )
Bioremediation   ( lcsh )
Soil ecology   ( lcsh )
Soil and Water Science thesis Ph. D
Dissertations, Academic -- Soil and Water Science -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1993.
Includes bibliographical references (leaves 157-169).
General Note:
General Note:
Statement of Responsibility:
by Cheryl A. Bellin.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030382223
oclc - 31022288
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
    List of Tables
        Page vii
    List of Figures
        Page viii
        Page ix
        Page x
        Page xi
        Page xii
        Page xiii
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
    Chapter 2. Chemodynamics of N-heterocyclic compounds in abiotic systems: Batch and flow-through techniques
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
    Chapter 3. Alteration of surfaces by bacterial biomass
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
    Chapter 4. Quinoline biodegradation in flow-through systems
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
    Chapter 5. Summary and conclusions
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
    Biographical sketch
        Page 170
        Page 171
        Page 172
Full Text







Copyright 1993


Cheryl A. Bellin


This dissertation was the collaborative effort of many people that either

directly or indirectly facilitated the completion of this document. Working on a

project entitled coupled processes is not easily accomplished in one laboratory

needless to say by one individual. During the last four years I received

assistance from many people, which I greatly appreciate. First, I would like to

thank Dr. Suresh Rao, for his insightful method of teaching and guidance. His

questions and discussions generated enthusiasm and challenged me to a

higher level of thinking. I would like to thank my committee members, Drs.

Bitton, Nkedi-Kizza, Hatfield, and Rhue for their helpful comments and

suggestions during my doctoral program. I especially thank Dr. Bitton for

facilitating the surface characterization of the bacterial isolates used in these

studies and Dr. Nkedi-Kizza for his critical review of the abiotic behavior of


I would like to thank my friends and colleagues in the lab, Drs. Linda

Lee, Denie Augustijn, Itaru Okuda, and Ms. Dongping Dai. Most of the time ...

the lab was a great place to interact and work. I would like to especially thank

Linda for her personal and professional insight when it was asked for and even

when it was not. I thank Ron Jessup and Gerco Hoogeweg for their modeling

efforts. I also thank Christianne Smethurst and Ann Benner for their assistance

in the laboratory. I give special thanks to Candace Biggerstaff who added

levity, friendship, and help in finishing this dissertation.

Several people in the Soil and Water Science Department have

contributed to the work presented in this dissertation. In particular, I would like

to thank Dr. Sylvia Coleman for her guidance and use of her laboratory for

microbial preparations, Chris Pedersen and Dr. Amiel Jarstfer for the use of

their laboratory and their insight on microbiology ecology, Dr. Willie Harris for an

introduction to clay mineralogy and x-ray diffraction, and Kevin Cubinski and Dr.

Ann Wilke for their help in designing the continuously stirred flow-through


I would like to acknowledge the financial support provided by the State of

Florida via a Soil and Water Science research assistantship and additional

funding provided through Battelle Pacific Northwest Laboratories (PNL) from the

Department of Energy.

I thank Dr. John Zachara and Dr. Cal Ainsworth for their insights on

quinoline sorption, Dr. Jim Fredrickson and Dr. Fred Brockman for the bacterial

isolates used in this study and the other Battelle PNL staff for my enhancing my

summer fellowship experience.

I thank my mother and father for their continued support and pride in the

work I was doing. I also want to thank Dave Cantlin for his friendship, support,

and patience that were essential throughout my stay at UF and in Florida.


ACKNOW LEDGEMENTS ..................................... iii

LIST O F TABLES .......................................... vii

LIST O F FIG URES ......................................... viii

A BSTRA CT .............................................. xi


1 INTRODUCTION .................................. 1

Overview of the Problem ............................. 1
S option . . . . . . . . . . . . . . . . . . . . 7
Biodegradation .................................... 9
Transport ...................................... 18
Research Objectives ............................... 21

TECHNIQ UES .................................. 25

Introduction .................................... 25
Quinoline Sorption Dynamics ....................... 28
Research Question and Tasks ...................... 35
Materials and Methods ............................ 35
Results and Discussion ........................... 39
Sum m ary ..................................... 73


Introduction .................................... .. 75
Research Question and Tasks ...................... 78
Materials and Methods ............................ 79
R results ....................................... 83

Discussion ..................................... 98
Sum m ary .................................... 101

SYSTEM S .................................... 103

Introduction ................................... 103
Quinoline Biodegradation Dynamics .................. 121
Research Question and Tasks ..................... 126
Material and Methods ............................ 127
Results and Discussion .......................... 132
Summary ..................................... 145

5 SUMMARY AND CONCLUSIONS ................... 148

Sum m ary .................................... 148
C conclusions .................................. 154

REFERENCES .......................................... 157

BIOGRAPHICAL SKETCH .................................. 170


Table pag

2-1. Soil properties before and after steam autoclaving .............. 36

2-2. Column parameters for sterile soil columns................... 48

2-3. Summary of estimated transport parameters for quinoline. ........ 57

3-1. Column parameters and Kf values for quinoline, naphthalene, and
45Ca in sterile and inoculated Norborne soil columns ............ 87

4-1. Nutrient concentration (mg/L) extracted from the Norborne soil
colum n .......................................... 133


Figure age

2-1. Calcium (1) and quinoline (0) BTCs: a) pH 6, v = 0.162 cm/s and
b) pH = 6.9, v = 0.063 cm/s. Lines correspond to equilibrium
(solid) and first-order models (dash). (from Szecsody and Streile,
1992) .. . . . . . . . . . . . . . . . . . . . . . . 27

2-2. Quinoline speciation diagram and the protonated and neutral
species structures ..................................... 29

2-3. Quinoline sorption isotherms for three soils normalized to their
cation exchange capacity and to the fraction of protonated
species. .......................................... 41

2-4. Stirred batch reactor (a) and quinoline sorption onto the Norborne
soil fraction < 50 jim (b) (where C = quinoline filtrate
concentration and Co = the initial quinoline concentration)........ 43

2-5. Sorption of quinoline on the Norborne soil in the presence of 2-
hydroxyquinoline .................................... 44

2-6. Examples of breakthrough curves for PFBA and 3H20 in Norborne
soil columns ......................................... 46

2-7. Quinoline and 45Ca breakthrough curves with flow interruptions in
0.005 M (closed symbols) and 0.05 M (open symbols) CaCI2
Norborne soil columns ................................ 49

2-8. Quinoline breakthrough curves in 0.005 M (closed symbols) and
0.05 M CaCI2 (open symbols) in pH adjusted Norborne soil
colum ns. ...... .. ... .. ... .. .. .. ... ....... .. .... 51

2-9. Repeated flow interruptions for quinoline in a 0.05 M CaCI2 (pH
6.2) Norborne soil column and bicontinuum model fit ............ 55

2-10. Conceptual diagram of quinoline sorption onto smectite clay
m inerals. ........................................... 60

2-11. Isotopic exchange of 12C-quinoline and 14C-quinoline in 0.05 M
CaCl2 (pH 6.2) in the Norborne soil ....................... 63

2-12. Breakthrough curves of quinoline in Eustis soil with 0.005 M CaCl2
and 30% methanol ................................... 67

2-13. Structural representation of organic matter (adapted from Bbhar
and Vandenbroucke, 1987) .............................. 69

2-14. Scanning electron micrograph of an organic soil at 6000 x and
1000 x ............................................ 72

3-1. Measured BTCs for PFBA (N) in a sterile column and for Quinoline
in a sterile (@), 3N3A inoculated (*), and B53 inoculated (0J) soil
column. Column designations are given in parenthesis
corresponding to Table 3-1 ............................. 84

3-2. Measured BTCs for 45Ca in sterile (@) and B53 inoculated (0 and
*) soil columns. Column designations are given in parenthesis
corresponding to Table 3-1............................... 85

3-3. Measured BTCs for Naphthalene in a sterile (*) and a B53
inoculated (0) soil column. Column designations are given in
parenthesis corresponding to Table 3-1 ..................... 86

3-4. Measured BTCs for PFBA (*), 45Ca (0), Quinoline (J), and
Naphthalene (0) in a B53 inoculated soil column............... 92

4-1. Schematic of sorption and biodegradation in soil aggregates (C
and C = the solute concentration in the pore water inside the
aggregate and the bulk solution, respectively) (adapted from
Mihelcic and Luthy, 1988c).............................. 106

4-2. The impact of varying the sorption partition coefficient on
biodegradation (L/kg) in the presence of aggregates with radii of
0.05 cm. (From Scow and Hutson, 1992)................... 108

4-3. Data (symbols) for aggregates with different radii and DSB model
simulations (solid lines) of mineralization of 50 ng 14C-labeled
glutamate/mL in the presence of gel exclusion beads. (From Scow
and Alexander, 1992).................................. 109

4-4. Measured and simulated BTCs for 2,4,5-T developed with the two
region model for the two cases of no degradation (C =0) and
degradation (a >0). (From Gamerdinger et al., 1990) .......... 111

4-5. Simulation of naphthalene degradation in soil suspensions. The
lines were generated using the bicontinuum model with first order
biodegradation kinetics. (model input parameters from Guerin and
Boyd, 1992).. ...................................... 113

4-6. Simulation using the bicontinuum model with first-order
biodegradation kinetics assuming irreversible sorption .......... 115

4-7. Naphthalene mineralization for strain NP-Alk in a soil free (o),
Colwood (a) and Oshtemo (b) soil slurries with 66.7 (0), 133 (),
or 200 (0) mg/mL (From Guerin and Boyd, 1992). ............ 116

4-8. Naphthalene mineralization time courses for strain 17484 in a soil-
free control and Capac (a) and Colwood soil suspensions (From
Guerin and Boyd, 1992) ............................... 117

4-9. Conceptualization of quinoline biodegradation in the presence of
sm ectite clay m inerals................................. 122

4-10. Schematic of CSFTR system used to monitor quinoline
biodegradation ..................................... 129

4-11. Quinoline biodegradation in a Norborne soil column with limiting
nutrients ........................................... 137

4-12. Biodegradation of quinoline and production of 2-HQ by the 3N3A
isolate in the CSFTR .................................. 140

4-13. Alteration of bacterial activity upon introduction of Norborne clay
and silt as measured by the change in biodegradation of quinoline. 142

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Cheryl A. Bellin

August 1993

Chairperson: P.S.C. Rao
Major Department: Soil and Water Science

Bioavailability and biodegradation of organic solutes in soils are

thought to be controlled by coupled sorption and transformation processes.

The principal hypothesis is that sorbed substrates are unavailable to

microorganisms. The fact that microorganisms may actively change the local

environment further complicates the issue by altering the magnitude and

kinetics of sorption and degradation. The importance of coupled sorption-

biodegradation processes is recognized in regard to the impact on

environmental contamination and bioremediation.

Bioremediation technologies have generally had limited success in

achieving adequate levels of cleanup, primarily because of constraints on

bioavailability of sorbed contaminants. Thus, understanding the interactions

among sorption, biodegradation, and transport processes is needed to

elucidate rate-limiting mechanisms of contaminant biodegradation.

Quinoline, an ionizable organic base, is a contaminant of interest found in

energy-derived waste materials and products. Batch reactors were used to

measure quinoline equilibrium sorption coefficients in the absence of physical

constraints. Miscible displacement studies were conducted to simultaneously

measure quinoline sorption and biodegradation. The quinolinium cation was the

predominant species sorbed via cation exchange. However, the bicontinuum

sorption nonequilibrium model was inadequate in describing the measured

breakthrough curves for quinoline displacement through "sterilized" soil

columns. Quinoline-surface complexes limit the desorption and redistribution

within the sorbent matrix and thus, are likely to be unavailable for degradation.

Addition of bacteria (quinoline-nondegrader) reduced quinoline sorption

and retardation in soil columns, which were attributed to biomass-induced

changes in quinoline speciation and blockage of surface sites. In columns

inoculated with a quinoline-degrader, quinoline was rapidly degraded and

biodegradation kinetics could not be measured. The continuously stirred flow-

through reactor was used as an alternate technique to monitor rapid

biodegradation kinetics (kb < 0.5 seconds'1) and to measure the response to

imposed perturbations. Introduction of sorbent particles at steady state (i.e.,

biodegradation of quinoline to 2-hydroxyquinoline and other metabolites)

resulted in two responses: 1) addition of soil particles required readaptation of

the bacterial isolate and caused reduced degradation rates; and 2) soil particles

reduced 2-hydroxyquinoline uptake and degradation, while quinoline

biodegradation was not altered. In this case, bacterial activity may have been

reduced upon bacteria-sorbent association.


The motivation for this dissertation arose from my desire to work with

microorganisms and to determine their potential for bioremediation of

contaminated soils, aquifers, and sediments. There is an illusion that bacteria

are fragile, delicate creatures. The reality of the situation, after working with

them for the last few years, is that they at times seem to have a mind of their

own. They have the capability to alter their environment in order to enhance

their very existence. I believe that their potential in bioremediation practices is

unlimited if we can only come to understand how they interact with their

environment. As Marshall (1976) stated:

It is my belief that many microbiologists fail to appreciate the effects of
interfaces on microbial populations, despite the widespread occurrence
of solid-liquid, gas-liquid, and liquid-liquid interfaces in natural microbial
habitats. ... Importance must be given to the nature, distribution, and
unique physicochemical properties of interfaces, the interaction between
microorganisms and interfaces, and the modifying effects of interfaces on
the ecology of microorganisms. (v)

Overview of the Problem

The improper use and accidental release of toxic organic compounds

into the environment have led to widespread contamination of soils and

aquifers. Treatment of contaminated materials has included excavation,

incineration, vapor extraction, and soil washing technologies. These treatments

are often costly and only result in a transfer of the contaminant from one phase

to another. However, implementation of above ground and in situ

bioremediation practices may lead to degradation of organic contaminants.

Bioremediation practices using laboratory tested microbial populations

have failed to achieve adequate levels of cleanup for reasons which will be

discussed. Failures are not surprising because frequently laboratory studies

investigate processes in isolation and attempt to extrapolate to field sites where

temperature, pH, soil water content, and microbial populations vary daily and

seasonally. As Rao et al. (1993a) so accurately described:

Most laboratory-scale experiments, and some field-scale studies, are
designed for investigating environmental processes in isolation; at least
attempts are made to do so by controlling most variables except the one
whose impact upon the system is being investigated. In real-world
scenarios, even in the simplest of laboratory experiments, however, the
rates and magnitudes of a reaction or a process are often controlled by
one or more other processes, each of which may have its own set of
unique control variables at different spatial and temporal scales. This is
indeed the case for laboratory experiments and field studies on fate and
transport of organic chemicals in soils and aquifers. An explicit
understanding of the coupling and feedback among simultaneous
processes is essential in explaining experimental observations and for
developing predictive models. (1)

To illustrate the importance of process coupling, Rao et al. (1993a)

presented two different scenarios. In one case, sorption renders the

contaminant unavailable for biodegradation and in the second case

biodegradation is unaffected by sorption. The first example suggests that

differences in biodegradation may be due to the soil sorption capacity and/or

variations in microbial activity. The second example suggests that differences in

biodegradation are due solely to variations in microbial activity and bioavailability

sorptionn of contaminants) is not a factor.

The use of bioremediation technology is hinged upon improving existing

knowledge of the controlling processes and their appropriate coupling such that

the probability and predictability of remediating a contaminated site are

increased. To fulfill this task it is necessary to 1) determine the reasons for

bioremediation failures; 2) develop predictive coupled-process models for

describing contaminant fate in the environment; and 3) determine the

ramifications of introducing bacteria or stimulating bacterial growth in soil and

aquifer materials to promote biodegradation of contaminants.

The success of bioremediation of contaminated soils and groundwater is

limited due to (1) the ability to degrade chemicals to an acceptable level and (2)

the ineffectiveness of laboratory-tested microorganisms to biodegrade

chemicals under field conditions. Understanding the physical and chemical

constraints of biodegradation in soils and aquifers may improve the designs of

bioremediation programs and provide an understanding of the reasons for

chemical persistence. Therefore, information is needed regarding microbial

transformations of organic chemicals in soil-water systems, as affected by the

interaction of chemical, physical, and biological processes.

A lack of consideration of physico-chemical and biological processes can

result in discrepancies between model predictions and experimental

observations. Investigation of organic chemical behavior in natural systems

and development of solute transport models that account for biodegradation

and sorption are necessary to adequately predict the environmental behavior of

such chemicals. These models can then be used to gain insight into the

processes that affect the fate of chemicals in the environment, for prescribing

management strategies that prevent or minimize groundwater contamination,

and for designing effective remediation procedures for contaminated sites.

Coupled-process models attempt to describe contaminant sorption,

degradation, and water flow by incorporating pertinent processes controlling the

fate of contaminants. Mathematical descriptions of existing coupled-process

models were reviewed by Brusseau et al. (1992). Development of an unbiased

coupled-process model requires a multidisciplinary approach. However, models

often contain a particular emphasis on a single process depending on the

researcher's background. The conceptual basis for the coupling of sorption

and biodegradation during transport was presented by Rao et al. (1993b).

Emphasis was given to the importance of adequately describing contaminant

sorption and the impact of the biomass on contaminant behavior.

Various levels of complexity arise when describing the processes that

control contaminant behavior. Frequently models are limited by the ability to

accurately measure the parameter of interest. When dealing with aquifer

materials, steady water flow is assumed. However, the unsaturated zone adds

seasonal variations in soil water content and temperature which directly or

indirectly impact the primary processes controlling the fate of contaminants. A

description of the sorption dynamics is primarily concerned with equilibrium or

rate-limited reactions, whereas microbial processes require descriptions of

microbial kinetics (e.g., growth and biodegradation) and biomass distribution.

Extensive data have been gathered describing individual processes that

determine the behavior of hydrophobic organic compounds (HOCs).

Equilibrium sorption coefficients (Kp) for HOCs can be estimated from aqueous

solubility and octanol-water partition coefficients among others (cf., Green and

Karickhoff, 1990; Gerstl, 1990). The sorption mass-transfer coefficients (k2) can

be estimated for a variety of soils and HOCs from the inverse, log-log

relationship noted between k2 and Kp (Brusseau and Rao, 1989a) or Koc

(Augustijn, 1993). Specific interactions between ionizable organic acids and soil

caused deviations from the behavior of HOCs (Brusseau and Rao, 1989a).

Complex sorption interactions of organic bases such as the nitrogen

heterocyclic compounds (NHCs) in soil have not been adequately investigated

to assess if this relationship is valid for NHCs.

The estimation of the model parameters related to biomass growth

dynamics of specific degraders and substrate degradation kinetics in soil and

aquifer materials is somewhat uncertain. Monod-type equations are used to

describe the behavior of pure culture systems. However, these models did not

adequately describe degradation in mixed culture laboratory systems (Scow et

al., 1986; Simkins et al., 1986). Therefore, these models are not likely to predict

field-scale observations. Blackburn (1989) claims that laboratory-scale

predictions of field-scale observations are destined to fail because of the

complexity of the spatial scales of interest (for further discussion see Rao et al.,

1993a). Blackburn (1989) suggested that the Heisenberg Uncertainty Principle

applies to microbial dynamics which states that by simply making an

experimental observation (since most experimental techniques are invasive,

though in some cases noninvasive techniques may be used), the system is

perturbed and is no longer an adequate representation of the original system.

Despite these arguments, complex degradation models have been developed

that incorporate availability of electron acceptors and electron donors, nutrients,

and the oxygen status in aquifers (Widdowson et al., 1987; 1988; MacQuarrie

and Sudicky, 1990). Because of the inability to describe the parameters at the

field scale, many of these models are not validated.

Existing coupled process models are highly limited by a lack of

experimental observations (laboratory and field scales) that quantitatively

demonstrate the effects of process coupling, specifically the manifestation of

such coupling on contaminant migration/degradation rates and profiles.

Laboratory studies coupling sorption, degradation, and transport are limited to

HOCs; most are conducted in batch reactors. The simultaneous sorption,

transformation, and transport of NHCs in dynamic soil systems has not been

studied. NHCs can exist in their protonated or neutral form depending on the

pH in the system. Therefore, to estimate the fate of these compounds, an

adequate representation of the appropriate linkages between the controlling

processes is essential. For these compounds, variations in pH will have

ramifications on the microbial community and their activity as well as on the

sorption dynamics. The following section is a review of the key processes that

control the fate of organic compounds and discuss the factors important in

developing a coupled- process model.


The distribution of HOCs between the solid and solution phases is

characterized by an equilibrium sorption partition coefficient (Karickhoff et al.,

1979; Chiou et al., 1983). Most often the Freundlich isotherm is used:

S_ Kf C1/n (1-1)

where S is the sorbed concentration (gg/g), Kf = Freundlich sorption coefficient

[mL(1/n) 4g[1-(1/n)]/g], C = equilibrium solution concentration (Ag/mL), and 1/n

= Freundlich isotherm constant. Equilibrium sorption models are often used in

solute transport models. However, equilibrium assumptions are generally

inadequate in describing local-scale and field-scale sorption because

nonequilibrium conditions predominate.

Sorption nonequilibrium for HOCs can be described using the

bicontinuum model (Brusseau and Rao, 1989b). Conceptually, the model

describes partitioning of compounds into the soil organic phase or adsorption

of compounds onto surfaces. Nonequilibrium sorption is represented by a two-

step process in which sorption in the first domain is instantaneous, while mass

transfer constraints limit sorption in the second domain. Thin organic coatings

distributed throughout the soil may result in minimal constraints for sorption

mass transfer, whereas sorption into large organic particles may increase solute

diffusion due to limited accessibility of sorptive regions. Factors that limit the

rate of HOC sorption that have been proposed include intraparticle diffusion

(IPD) (Wu and Gschwend, 1986; Ball and Roberts, 1991) and intraorganic

matter diffusion (IOMD) (Brusseau et al., 1991). Regardless of the actual

mechanism responsible for rate-limited sorption, contaminants are likely to

reside within the interior regions of the sorbent matrix. The consequences of

this occurrence on biodegradation will be discussed in the next section.

Sorption of NHCs has been described by the Freundlich isotherm

(Zachara et al., 1986, Ainsworth et al., 1987). Linearity of the sorption

isotherms varied, approaching a linear isotherm at low concentrations and

surface coverages (Ainsworth et al., 1987). The protonated species is the

predominant form of NHCs sorbed and is expected to sorb primarily onto cation

exchange sites. These sites may be associated with phyllosilicate minerals or

organic matter. In either case, sorption is likely to be rate limited due to

migration into clay interlayers and aggregates or organic matter matrices.

Given the complexity of exchange reactions involving organic cations, the

bicontinuum model may not adequately describe the behavior of NHCs in soil

materials. This aspect will be explored further in a later section (see Chapter 2).



Biodegradation is a dominant mechanism affecting organic chemical

transformations in soils and aquifers. Microbial degradation of most small

organic compounds (molecular mass < 600) occurs intra-cellularly (Bitton et al.,

1988). Thus, the rate of biodegradation is limited by the dynamics of 1)

physical-chemical processes (e.g., solubility, sorption, hydrodynamic dispersion)

that leads to a lowering of solute concentration in the solution phase; 2) soil or

environmental factors that limit physiological activity of the appropriate microbial

consortia; 3) microbial factors that limit substrate uptake by the microorganisms

(e.g., cell permeability and hydrophobicity); and 4) intra-cellular genetic or

biochemical factors (e.g., presence of appropriate enzyme systems, presence

and expression of genes) that limits utilization of the compound. The

recalcitrance of different organic chemicals in a specific soil, or the variations in

degradation rates of a specific compound in several soils, may be explained to

a large extent by understanding these key factors.

Inoculation of soils and aquifers with microorganisms capable of readily

degrading chemicals may result in a partial or complete lack of contaminant

removal due to various environmental stresses not present under laboratory

conditions. Contaminant persistence may result from the following factors

(Madsen, 1985; Goldstein et al., 1985): 1) low substrate concentrations not

supporting microbial growth; 2) microorganisms encountering toxins or

predators; 3) microorganisms using more readily available carbon sources; and

4) introduced microorganisms not reaching the contaminated site.

Enhanced on-site or in-situ biodegradation provides a method for

removing organic contaminants in soils and aquifers. Utilizing indigenous

microorganisms is preferable to "inoculation" or injection because they are

already adapted to the local environment. However, in the subsurface

environment, the complex interaction between microorganisms, substrates, and

surfaces may alter this process. Biodegradation rates may be limited by

chemical properties of substrates, interactions of the substrate with surfaces, or

simply by the lack of necessary enzymes (Madsen, 1985). Availability of slightly

soluble substrates may be controlled by the rate of dissolution (Stucki and

Alexander, 1987; Miller and Bartha, 1989; Huang and Chou, 1990), or by low

aqueous concentrations which may not induce the necessary enzymes for

biodegradation (Madsen, 1985). Similarly, sorption of the substrate by soil may

reduce substrate concentrations in solution below levels necessary for enzyme


Sorption of substrates might also enhance biodegradation rates by

decreasing the substrate concentration to levels that are not toxic to

microorganisms responsible for degradation (van Loosdrecht et al., 1990).

Sorption more likely reduces or inhibits biodegradation rates in soils (Stotzky

and Rem, 1966; Madsen, 1985; van Loosdrecht et al., 1990). For example,

sorption was found to decrease the amount of substrate available to

microorganisms capable of degrading several compounds, including diquat

(Weber and Coble, 1968), benzylamine (Subba-Rao and Alexander, 1982; Miller

and Alexander, 1991), alkylamines (Wszolek and Alexander, 1979), glucose

(Gordon and Millero, 1985), 2,4-Dichlorophenoxyacetic acid (Ogram et al.,

1985), amino acids (Dashman and Stotzky, 1986), toluene (Robinson et al.,

1990), benzidine (Weber, 1991), quinoline (Smith et al., 1992), and flumetsulam

(Lehman et al., 1992). Degradation was adequately described by a second-

order rate equation with the assumption that only solution-phase chlorproham

and dibutyl phthalate are biodegraded in the presence of sediments (Steen et

al., 1980).

Biodegradation of contaminants may be limited when contaminants are

sequestered within the organic or inorganic components of the sorbent matrix

that are not directly accessible to microorganisms. Biodegradation may also

be limited by mass transfer (IPD and IOMD) from the interior of the sorbent to

the exterior solution. Bioavailability is limited in these examples because intra-

aggregate pores are too small to be accessible to bacteria (Steinberg et al.,

1987, Scow and Alexander, 1992). The substrate sorbed within organic matter

is accessible only after desorption or diffusion out of the sorbent matrix. Mass

transfer constraints have been shown for sorption/desorption of hydrophobic

organic compounds (HOCs) in soils and sediments (Wu and Gschwend, 1986;

Brusseau and Rao, 1989b; Brusseau et al., 1991), for biodegradation of HOCs

(Rijnaarts et al., 1990; Robinson et al., 1990), and for denitrification (Myrold and

Tiedje, 1985). For naphthalene, which exhibits reversible sorption/desorption

(Mihelcic and Luthy, 1988a,b), biodegradation was not dependent upon

desorption kinetics from fine-sized material (Mihelcic, 1988). For larger

particles, biodegradation of naphthalene was dependent upon intra-particle

diffusion from the solid-phase to the solution-phase, which suggests mass

transfer constraints or reduced bioavailability of the sorbed naphthalene

(Mihelcic and Luthy, 1988c).

For quinoline, highly selective cation exchange reactions may control

mass transfer from the soil to solution, thereby limiting biodegradation. Smith et

al. (1992) suggested that biodegradation of quinoline in dispersed clay

suspensions is limited by desorption of the highly stable quinolinium ion surface

complex. However, it is not known if these same rate-limiting steps control

biodegradation rates in soils and sediments or if diffusion-limited mass transfer

constraints (IOMD, IPD) are operative. For this reason, mechanistic models

coupling the sorption, degradation, and transport in soil and aquifer systems

are needed to understand the rate-limiting steps of organic chemical


Effects of Surfaces on Biodearadation

At the cellular-scale, the influences of surfaces on bacterial activity have

been monitored indirectly in a variety of disciplines. Reported observations

suggesting the influence of surfaces on bacterial activity have been dismissed

because of possible secondary responses occurring at the surfaces (van

Loosdrecht et al., 1990). Ogram et al. (1985) demonstrated that sorbed 2,4-

dichlorophenoxy acetic acid (2,4-D) was protected from biodegradation and that

only the solution-phase 2,4-D was degraded by free and attached bacteria. The

degradative activity of free and attached bacteria, however, could not be

differentiated. In a similar study, 2,4-D was suggested to be degraded by

bacteria in the sorbed and solution phase (Zou et al., 1992); however,

degradation rates were thought to be faster by "free" bacteria rather than

sorbed-phase bacteria. Aamand et al. (1989) also suggested that only bacteria

in the solution phase were degrading the aquifer contaminants.

More recently, Guerin and Boyd (1992) argued that a bacterial isolate (P.

putida 17484) was capable of utilizing sorbed naphthalene from the surface,

contrary to the paradigm that degradation occurs intracellularly. Another

bacterial isolate (NP-Alk) was thought to be unable to degrade naphthalene in

the sorbed- phase. Therefore, organism-specific properties must be considered

in determining the potential for degradation. These observations will be

discussed further in Chapter 4. Determining the influence of surfaces on

biodegradation and whether or not bacteria have the ability to degrade

contaminants in the sorbed or solution phase is still unresolved. Further, a

predictive model requires knowledge of the distribution of the active microbial

biomass and microbial growth dynamics (e.g., contingent upon substrate,

nutrient, and electron acceptor concentration and bacterial population) in

combination with factors discussed above.

Biomass Distribution

Microbial biomass is subject to sorption and transport processes.

Therefore, bacteria may exist in the soil either sorbed (attached) or in solution

(free). Physical, chemical, and microbial factors controlling the distribution of

bacteria in porous media have recently been summarized by Harvey (1991),

Lindqvist and Enfield (1992b), and Tan et al. (1992). Bacteria grow after they

attach to surfaces if essential carbon and energy sources are available. Growth

and development of bacterial colonies generally is followed by the production of

extracellular polysaccharides and promote the formation of bacterial biofilms

(van Loosdrecht et al., 1990; Fletcher, 1991). Under nutrient- and substrate-rich

conditions, as may be the case near waste disposal sites, biofilms may be


Mathematical models for biodegradation are developed assuming that

the microbial biomass may be distributed in biofilms, microcolonies, or uniformly

throughout the porous medium (Baveye and Valocchi, 1989). The assumption

of microbial biofilms suggests that surfaces are uniformly coated by biofilms in

which the degradation of the contaminant and the utilization of the electron

acceptor takes place (Rittman and McCarty, 1980). The microcolony approach

suggests that bacteria exist in discrete microcolonies and that growth and

substrate utilization rates correspond to the microbial population (Molz et. al.,

1986; Marshall, 1992). Recent microscopic evidence suggests that bacteria

exist in microcolonies with bacterial cells extending out into the soil pore spaces

(Vandevivere and Baveye, 1992). The difficulty in mathematically describing the

dimensions of the biofilms and microcolonies limits the utilization of these

models in soils and aquifers. The uniform microbial description, commonly

used in solute transport models, makes no assumptions about the distribution

of bacteria (e.g., discrete colonies or biofilms) in solution or on the surfaces

(Corapicoglu and Haridas, 1985; Kindred and Celia, 1989). This concept

suggests that overall growth and metabolism are not influenced by the microbial


Biomass Impacts on Contaminant Sorption and Transport

Growth or addition of bacteria may drastically alter the chemical, physical

and microbiological environment of soil surfaces (Fletcher, 1991). Chemical

properties of soil surfaces may be altered by bacterial biomass thereby

influencing contaminant transport (Stucki et al. 1992; van Loosdrecht et al.,

1990; Stotzky, 1966). Physical alterations including blockage of pores by

bacterial biomass and blockage of sorptive regions in the soil may occur

altering water flow and sorption contaminants (Tan et al., 1992; Vandevivere

and Baveye, 1992). Bacterial transport (e.g., solution phase bacteria) and their

facilitation of contaminant migration was recently demonstrated (Lindqvist and

Enfield, 1992a). The impact of bacterial biomass is becoming recognized as an

important process influencing contaminant sorption and transport (Rao et al.,

1993b). Therefore, the impact of bacterial biomass near hazardous waste sites

is of interest.

Environmental Factors Influencing Biodegradation

Environmental variables may be significant in surface soils where

microbial communities are in direct contact with the soil atmosphere. Seasonal

cycles in temperature and soil-water content distinguish this zone from aquifer

systems that may exhibit more constant conditions. Groundwater temperatures

are relatively constant; however, temperatures may be as low as 10 to 15C

which may reduce microbial activity. Surface fluctuations in temperate regions

may reduce bacterial activity throughout the winter months. In contrast,

bacterial activity will likely be high in warmer, tropical environments. Variations

in temperature over the usual range of interest (5-40C) are not likely to

influence the degradation pathway, only the rate of microbial degradation and

the microbial density. Changes in soil-water content, on the other hand, may

influence microbial communities and their activity.

Quantitative and qualitative differences result when observing aerobic and

anaerobic degradation. Deep, saturated aquifers may be depleted in oxygen

and bacterial populations may be limited by the availability of alternate electron

acceptors (NO3, SO4, CO3). In oxygen depleted zones, fermentation results in

incomplete degradation of contaminants. Flow heterogeneities may create

zones of mixing thus supplying adequate nutrients and cofactors to stimulate a

diverse and numerous group of microorganisms. On the other hand, a

contaminated area can turn an oxygenated aquifer into an anoxic region, if the

heterotrophic respiration exceeds oxygen input or recharge. In well-drained

soils and shallow aquifers, microbial populations are predominantly aerobic,

utilizing gaseous or dissolved oxygen as an electron acceptor which would

degrade organic contaminants to metabolites and ultimately mineralized to C02,

H20, and other elements. Even in a well-drained soil, however, anaerobic

regions (e.g., microsites) may develop as oxygen is depleted potentially altering

the end products of metabolism.

Biodegradation Models

Specific growth rates of microbial populations have been represented by

a variety of mathematical models (Pirt, 1975; Alexander and Scow, 1989; Bazin

and Menell, 1990). The empirical power rate model:

ac kbCn (1-2)

where kb is the biodegradation rate constant (1/T), simplifies to a first-order

kinetics model when n = 1 (Hamaker, 1972). Concern over the use of this

model is expressed as it is often presented with no theoretical justification for its

use (Bazin et al., 1976).

The description of the microbial growth rate when it is restricted by the

concentration of a growth-limiting substrate is given by the Monod equation

which was developed from enzyme kinetics:

A max --S (1-3)

where g is the specific growth rate of the biomass (1/T), Amax is the maximum

specific growth rate (1 /T), S is the substrate concentration (M/L3) and Ks is the

substrate half saturation constant (M/L3). This equation is commonly used to

describe the bacterial growth upon contaminant degradation.

Often, organic contaminant degradation is limited by availability of an

electron acceptor or an additional carbon substrate. The modified Monod

equation couples the dependence of bacterial growth on another carbon

substrate or electron acceptor:

A- [ S A mO] (1-4)
K T+S 4I<_+O

where 0 is the oxygen concentration (M/L3) and K0 is the oxygen half

saturation constant (M/L3). Equations may also incorporate an inhibition

coefficient to account for growth rate limitations due to a toxic feedback

mechanism (Harvey and Widdowson, 1992). In order to adequately describe

contaminant behavior, all parameters necessary for these models must be

measured at the particular scale of interest.


The governing differential equation that serves as the basis for most

coupled-process models used in soils and aquifers is

a(ec) [v D vq [v qC] [ ]a(PS) ,i (1-5)
at at

where C = solution-phase concentration (M/L3); S = sorbed-phase

concentration (M/M); t = time (T); p = soil bulk density (M/L3); 0 = fractional

volumetric water content dimensionlesss); D = hydrodynamic dispersion

coefficient (L2/T); x = distance (L); q = Darcy flux for water flow (L/T); and Oi

= rates (M/L3T) of loss or gain via various sinks and sources. In Eq (1-5),

multi-dimensional, advective-dispersive solute transport in a heterogeneous

porous medium under transient water flow conditions (first two terms on the

r.h.s.) is coupled to sorption dynamics (third term on r.h.s.) and biodegradation

kinetics (last term on r.h.s.). Differences in published models arise from the

specific manner in which sorption and degradation kinetics are modeled,

whether transient or steady flow is considered, and if one- or multi-dimensional

transport is of interest.

For one-dimensional steady, saturated water flow conditions in a

homogeneous medium, eq (1-5) can be restated as

8C DC vC C EaS + 1 (1-6)
at aW2 ax 9at -0

where v= (q/0) is the average pore-water velocity (L/T).

Assuming that sorption can be represented by the bicontinuum sorption

model with a Freundlich isotherm and that first-order biodegradation kinetics

apply to biodegradation (o = kbOC), eq (1-6) is restated as follows:

[1+C-k I ac -2 8C ac P S2_
[I+---P F] acn D2" D -v --.C kbC (1-7)
0 at ax2 ax 0 at

where kb represents the pseudo first-order rate constant (1/T) for

biodegradation (assumed to occur only in the solution phase), Kf is the

Freundlich sorption coefficient (mL0/n) jg[1-(1/n)1/g), and 1/n is the Freundlich

sorption isotherm coefficient. Note that the Freundlich model (eq 1-1) is used

to represent equilibrium sorption isotherms. Thus, isotherm nonlinearity may be

accounted for with this model which results in nonlinear mass transfer and

mixed order (1/n) equations. The model may be written in nondimensional

form (Nkedi- Kizza et al., 1989):

ac* +(,e R-1)(1/n)C*(1/n)-18 C*
ap ap

1 a2C* 1ac* ) yC (1-8a)
P ax2 aX ap

(1 R-)LaS: (C*(l/n)- S*) (1-8b)
by defining the following dimensionless parameters: C* = C/Co, p = vt/L, X =

x/L, y = kv/L, S* = [S2/1-F)KC(1/n)-1], R = [1 + ((p/0) KfCo(1/n)l)] is the

retardation factor, which represents equilibrium sorption; P = vL/D is the Peclet

number, which represents the hydrodynamic dispersion in the column;

p3 = {[1 + (Fp/6)KfCo'/n)-']/R} represents the fraction of instantaneous

retardation; o = {[k2(1-p)RL]/v} is the Damkohler number, which is

proportional to the ratio of hydrodynamic residence time (L/V) to the reaction

time (1/k2); L is the length of the column (L); F is the fraction of sorption in the

instantaneous regions; k2 is the first-order rate coefficient (1/T).

At the field scale, heterogeneous flow fields are often assumed to be

represented as being macroscopically homogeneous (MacQuarrie and Sudicky,

1990). The effects of local-scale pore-velocity variations are represented by a
"macrodispersion" term for the whole flow field. MacQuarrie and Sudicky (1990)

showed that such an approach can lead to a serious overestimation of

substrate degradation rate as a result of far greater mixing of the substrate and

dissolved oxygen plumes predicted to occur at the local scales when a macro-

dispersion concept is employed. This is a clear demonstration of the

importance of appropriately understanding local-scale physical heterogeneity in

explaining and predicting macro-scale observations of biodegradation.

Similarly, heterogeneities in the flow fields may create mixing zones where high

concentrations of electron donors (i.e., organic acids produced by fermentation

processes active in oxygen limited regions) and acceptors (i.e., oxygen) create

high microbial populations and degradation capacities.

Research Objectives

Over the past two decades, studying each factor that influences the

environmental behavior of organic chemicals in isolation has resulted in the

accumulation of an extensive database on several key processes. As Rao et al.

(1993a) pointed out:

Having made impressive advances in our understanding of the key
processes (transport, transformations, and sorption), it is now important
to examine the linkages between these processes. Coupled-processes
models provide the stimulus for a paradigm shift--from the reductionist
approaches to the relational approaches--where an investigation of the
inter-relations among the processes is considered even more important
than the examination of individual processes themselves. (8)

The primary objective of this dissertation research is to investigate soil-

solute-microorganism interactions and their importance in contaminant

persistence and transport in soil and aquifer materials. Reactions that are

important in coupling sorption, biodegradation, and transport of quinoline will be

investigated. From these results, the bioavailability of NHCs and thus, the

success of bioremediation practices will be assessed.

The following questions are proposed to address the following solute-

sorbent-microorganism interactions:

(1) Solute-sorbent What sorption processes limit bioavailability of

NHCs in remediation practices? Is the nonequilibrium sorption of

NHCs accurately described by the bicontinuum model?

(2) Microorganism-sorbent-solute: Do bioremediation practices

influence NHC sorption and transport?

(3) Solute-microorganism: What essential nutrient and oxygen

contents are required for biodegradation?

(4) Microorganism-sorbent Is bacterial activity (i.e., biodegradation)

altered in the presence of surfaces?

The following chapters address the questions stated above by studying

quinoline sorption and degradation. Quinoline, a NHC, is a contaminant found

in energy-derived waste materials and products and has the potential to be

transported to the subsurface soil and groundwater (Zachara et al., 1986).

Quinoline sorption has been characterized in batch systems using clay minerals

and soils. Desorption was recently shown to limit biodegradation of quinoline to

its primary metabolite (2-hydroxyquinoline) in batch systems (Smith et al., 1992).

In Chapter 2, process-level sorption kinetics of quinoline are examined and the

utility of the bicontinuum model is evaluated. Understanding the behavior of

quinoline sorption in flow through systems is necessary to determine the

processes controlling bioavailability. Equilibrium and mass transfer coefficients

for sorption and desorption were measured using batch and miscible

displacement techniques. The bicontinuum sorption model coupled with the

advective-dispersive solute transport model (during one-dimensional steady,

water flow) was used to assess the behavior of NHCs. This information was

then used to determine the rate-limiting processes controlling bioremediation

practices of NHCs.

The impact of biomass on the sorption and transport of three solutes

(naphthalene, 45Ca, quinoline) in a subsurface soil are investigated in Chapter

3. These compounds were selected because of their known interactions in soil

(i.e., cation exchange or hydrophobic partitioning). Miscible displacement

techniques were used to measure sorption and transport of the above

compounds during steady, saturated water flow conditions through

homogeneously packed, sterile or bacterial-inoculated soil columns. Pre-

inoculation of the Norborne soil with bacteria (108 cfu/g) simulates

contaminated subsurface soils and aquifers where bacterial populations may be

high. In this chapter I investigated the consequences of biostimulation practices

that attempt to remediate contaminated sites.

In Chapter 4, I explored the process coupling of sorption and

biodegradation of quinoline in flow through systems. First, growth limiting

factors were determined using miscible displacement techniques. This was

accomplished by monitoring breakthrough of quinoline and its primary

metabolite, 2-hydroxyquinoline, in bacterial-inoculated columns. Second, a

continually stirred, flow-through reactor was designed to monitor rapid

biodegradation kinetics and to assess the impact of surfaces on

biodegradation. These studies were used to determine processes important in

coupling sorption and biodegradation by investigating the impact of surfaces on

bacterial activity, and the conditions necessary for successfully remediating

contaminated sites.

Insights gained during my investigation of coupled processes and the

arduous task of dealing with living organisms are summarized in Chapter 5.

The significance, failures, and future opportunities of this research are also

presented in this chapter.



Sorption of NHCs may occur via cation exchange of the protonated

species on clay minerals or in organic matter and/or via partitioning of the

neutral species into organic matter. In contrast, sorption of HOCs occurs

primarily via "partitioning" into the organic phase. The dynamics of HOC

sorption have been conceptualized and described by the bicontinuum sorption

model (Karickhoff, 1980; Brusseau et al. 1991, Ball and Roberts, 1991).

However, the adequacy of this model to describe the behavior of NHCs is


Nonequilibrium sorption has been separated into transport- and sorption-

related processes. Transport-related nonequilibrium affects both sorptive and

nonsorptive compounds and results from heterogeneities in the flow paths.

When using non-aggregated media in packed-column laboratory studies,

transport-related nonequilibrium is generally determined to be negligible.

Sorption-related nonequilibrium results from specific solute-sorbent interactions

or diffusive mass transfer constraints. Organic matter is considered to be a

flexible polymer-like substance (Behar and Vandenbroucke, 1987) in which

diffusional constraints within the matrix (IOMD) cause sorption nonequilibrium of

HOCs. Nonequilibrium may also result from IPD (intraparticle diffusion) inside

microporous particles which contain organic coatings. HOCs are not likely to

exhibit chemical nonequilibrium because sorption occurs via partitioning

(Karickhoff et al., 1979; Chiou et al., 1983). Sorption of inorganic cations has

been shown to be rapid onto cation exchange sites and limited only by diffusion

to/from the exchanger surface (Nkedi-Kizza et al., 1989). Brusseau et al.

(1991) suggested that compensation of charge (i.e, cation sorption) likely

occurs near surfaces of organic matter; therefore, diffusional constraints of

HOCs and cations differ because of the path length and sorbent matrix.

Specific interactions of NHCs with the sorbent as well as and mass transfer

constraints within organic matter or phyllosilicate minerals are likely to limit

sorption of NHCs. Sorption of the quinolinium ion (i.e., cationic form of NHC)

onto predominantly organic matter associated CEC sites was suggested to be

faster than sorption of the neutral species (i.e., similar to HOCs) into the organic

matrix (Brusseau et al., 1991).

Quinoline, is a contaminant found in energy-derived waste materials and

products. Therefore, it was selected as a probe to evaluate the bicontinuum

sorption model and to further characterize the sorption dynamics of NHCs. A

first-order model did not adequately describe the complex interaction of

quinoline sorption onto clay modified alumina where 90% of the sites were

suggested to be readily available (Figure 2-1; Szecsody and Streile, 1992).













Figure 2-1. 0

Pore Volumes

0 50 100 150 200 250 300 350
Pore Volumes
alcium (0) and quinoline (o) BTCs: a) pH 6, v = 0.162 cm/s and b)
-H = 6.9, v = 0.063 cm/s. Lines correspond to equilibrium (solid)
id first-order models (dash). (from Szecsody and Streile, 1992).

Therefore, the mechanisms influencing quinoline sorption must be accurately

determined to assess the conceptual validity and adequacy of the bicontinuum


Quinoline Sorption Dynamics

Figure 2-2 describes the ionization of quinoline between the protonated

(QH+) and neutral species (Q) as a function of pH. Mathematically, the

ionization of quinoline is represented by

QH+ Q0 + H+ (2-1)

Ka [Q][H+] (2-2)

where Ka is the ionization constant.

Sorption of quinoline has been characterized in batch systems using soil

and clay materials (Ainsworth et al., 1987, Zachara et al., 1988; 1990), and in

column studies using modified and pure clays (McBride et al., 1992; Szecsody

and Streile, 1992). Quinoline sorption was adequately described by the

Freundlich isotherm (see Chapter 1). These studies suggest that the

quinolinium ion (QH+) is the predominant species sorbed via cation exchange

at low concentrations. As surface coverage increases, quinoline likely

occupies lower energy sites and multiple layers of quinoline at the sorbent

surface may form. More importantly, sorption varies with pH reflecting quinoline

ionization (Fig. 2-2, eq 2-1) and preferential retention of the organic cation.


0.75 -D$

. 0.50 pKa= 4.92

0 2 4 6 8 10

Figure 2-2. Quinoline speciation diagram and the protonated and
neutral species structures.

Sorption of the quinolinium ion has been shown even at pH values as much as

2 units greater than its ionization constant (pKa = 4.92) (Zachara et al., 1986;

Smith et al., 1992). Therefore, in a Ca+2 saturated homoionic soil, the following

cation exchange reaction can be used to describe quinoline exchange with


CaR2 + 2QH -- 2QHR + Ca+2 (2-3)

where QH + is the aqueous concentration of the quinolinium ion, Ca +2 is the

aqueous concentration of Ca +2, CaR2 is the Ca on the exchanger complex, and

QHR is the quinoline on the exchanger complex. The equilibrium constant

describing this reaction is given as follows:

Ke = [(QHR)2 (Ca+2)] (2-4)
[(CaR2)(QH+) 2]

where () refers to the activity of QH + and Ca +2 in the solution and exchange

phase. The conditional equilibrium constant (Kex) or Vanselow selectivity

coefficient (K ) for eq 2-3 is depicted as

Kv [XQHR (Ca2)](2-5)
[XcaR2(QH )2I

where X is the mole fraction, (QH +) is the activity of QH + in solution, and (Ca +2)

is the activity of Ca +2 in solution. In eq 2-5, the activities in the exchanger

phase are represented by X. The selectivity coefficient (K) is related to the

equilibrium constant (Ke), if the reaction is reversible, by the relationship:

Kv Kex fca+ 2 (2-6)

where the activity coefficients in the solid phase (f) of the exchanging ions

convert activity to mole fraction.

Quinoline and other NHCs form complexes with negatively charged solid

surfaces such as clay layer silicates (Zachara et al., 1986; 1987; 1988).

Selectivity coefficients (Kv) were developed for comparing the affinity of one

cation versus another to occupy a cation exchange site. The exchange of

quinoline and Ca +2 does not solely consider cation exchange because of the

strong quinoline-surface complexes. In this example, Kv includes the exchange

of quinoline and Ca 2 and the stability of the quinoline complexes on the

exchange phase. In eq 2-5, Kv > 1 indicates selectivity for QH + in the solid

phase whereas KV < 1 indicates Ca +2 is preferred. The high quinolinium

exchange selectivity coefficient on Na-montmorillonite (Kv = 200 to 1300) and

clay isolated from the Norborne soil (K, = 104 to 108) suggests that strong

quinoline-surface complexes are formed (Ainsworth et al., 1987; Zachara et al.,

1990). In these soils and pure clay minerals, Kv varied with pH and with

surface coverage which was suggested to be due to sorption of the neutral

species, occupation of high energy sites at low surface coverages, and surface

condensation. Reconfiguration of the quinoline molecule to a planar position

within interlayers of clay minerals may contribute to the hysteretic behavior

(Zachara et al., 1986; 1990) implying constraints to quinoline desorption. The

implication of this on quinoline transport in soils and aquifers will be examined in

a later section (Chapter 5).

A high Kv for quinoline suggests that quinoline may be favored over

inorganic cations on the exchange complex. Other NHCs (e.g., acridine,

pyridine), were shown to reduce quinoline sorption in low pH soils (4.7) where

compounds are protonated and sorption occurs via cation exchange (Zachara

et al., 1987). However, competition in soils where the neutral species

predominates (pH 7) was not apparent.

Predictive models have not been developed which adequately describe

the sorption and transport of NHCs (Szecsody and Streile, 1992). Sorption of

NHCs has been shown to be dependent upon the pH and cation exchange

capacity of the sorbent matrix. Therefore, accounting for these factors with an

individual parameter would enable the use of a predictive model for soils that

vary in their cation exchange capacity and pH. If the predominant sorption

mechanism is cation exchange, normalization of quinoline sorption to QH + and

the CEC of the soil of may be described by

S, KC1/n (2-7)

where S is the sorbed concentration [mol QH +/molc(-)], Ki = Freundlich-type

sorption coefficient [(L(1/rnmol QH +[-(n)/mol(-)], C, = equilibrium solution

concentration [mol QH +/L], and 1/n = isotherm constant. This relationship

resembles a Freundlich-type isotherm where the Kif describes the sorption of

NHCs accounting for variations in the cation exchange capacity and pH of the


The ion exchange of quinoline and Ca+2 in a system initially saturated

with Ca +2 was described in eq 2-5 and represented by KV. Freundlich

isotherms are not considered to be ion exchange isotherms. However,

assuming sorption of the protonated species onto cation exchange sites and

the fraction of the CEC occupied by quinoline is small, the K, may be related to

the Kv by the following relationship:

KV fc2 OH N Kv (2-8)

where N is the normality of the background electrolyte solution. However,

Zachara et al. (1988) predicted, based on eq 2-1, that the total sorbed quinoline

exceeded the fraction of quinoline existing as the quinolinum ion. Additional

sorption of quinoline could have been due to sorption of the neutral species,

clustering of the sorbate, surface condensation, or protonation of quinoline at

the exchanger surface (Ainsworth et al., 1987; Zachara et al., 1988). However,

measurement of enhanced acidity, thus, protonation of quinolinium at soil

surfaces, is not a trivial task. Sorption of the neutral species and cooperative

adsorption have been reported (Ainsworth et al., 1987) to occur at high surface

coverages via entropic or van der Waals forces.

Sorption of quinoline onto soils (pH 4 to 7) was thought to occur via

cation exchange in the presence of cosolvent mixtures [volume fraction of

cosolvent (f) < 0.4] (Zachara et al., 1988). Fu and Luthy (1986a) suggested

that cosolvents decreased quinoline sorption in response to an increase in

quinoline solubility. Quinoline isotherms at high concentrations (25 to 1000

mg/L) were suggested to be linear in water-methanol systems up to fo = 0.5

(Fu and Luthy, 1986b). Sorption at low concentrations (" 0.15 gg/mL) was

suggested to be nonlinear in aqueous systems (1/n = 0.75) and in

methanol/water solutions (20 vol % methanol; 1/n = 0.67) (Zachara et al.,

1988). Isotherm linearity has been shown to increase upon addition of

cosolvents for partitioning of solutes into an organic matrix; however, if ion

exchange predominates specific interactions with cation exchange sites may be

altered. For organic bases and acids, addition of solvents increases the fraction

of neutral species (Perrin et al., 1981; Lee, 1993). In the presence of

cosolvents, changes in the pKa values for organic bases are minimal (Perrin et

al., 1981). However, substantial increases in pKa values for organic acids have

been shown due to solute-solvent interactions resulting in decreased sorption of

phenolic compounds and increased sorption of carboxylic acids (fc > 0.2) (Lee,


Considering that the quinolinium ion sorption occurs predominately onto

cation exchange sites at low surface coverages, one could envision rate-limited

desorption of quinoline out of interlamellar regions of clay minerals and

aggregates or intra-organic matter regions. Such mass transfer constraints

delay the release of contaminants leading to persistence, inadequate

remediation, and limited bioavailability.

Research Question and Tasks

The primary objective of these studies are to investigate the process-level

kinetics of quinoline sorption by soils addressing the question: what are the

rate-limiting processes controlling bioremediation practices of NHCs?

Equilibrium and mass transfer coefficients for sorption and desorption were

measured as a function of pH, molarity (M) and sorbent. The bicontinuum

nonlinear sorption model coupled with the advective-dispersive solute transport

model was used to assess quinoline sorption and transport during one-

dimensional, steady water flow.

Materials and Methods


The soils used in this study and their properties are presented in Table 2-

1. Soils were sterilized for 30 min by steam autoclaving 50 g samples that were

brought to 15% water content and incubated for 24 hours. The process was

repeated two additional times and the soil was used in all subsequent

experiments unless otherwise noted. The soils used in the batch and column

experiments were initially saturated with Ca +2. Cation exchange measurements

were measured at the pH of the soil (See Table 2-1).


Pentafluorobenzoic acid (PFBA; 150 mg/L) and 1-2O (6000 cpm/mL)

were used as conservative, nonsorbing tracers to assess the hydrodynamic

Table 2-1. Soil properties before and after steam autoclaving.

pH in 0.005 CEC location of
Soil M CaCl2 f0C cmol(-)/kg CEC

Eustis 5.3 0.0039 3.20 organic matter and kaolinitic
Sterile Eustis 5.4 0.0032 4.44 clay minerals
Norborne 6.4 0.0015 11.91 smectite clay minerals and
Sterile Norborne 6.4 0.0015 11.76 organic matter
Webster 6.9 0.037 47.9 organic matter and
smectite clay minerals

dispersion and extent of physical nonequilibrium conditions prevailing during

transport through the soil columns (Brusseau and Rao, 1989a). Quinoline

concentrations in the influent solutions for the column studies ranged from 4 to

10 mg/L. 'C-quinoline (Sigma) and spiked to obtain solutions at 10,000

cpm/mL. Batch studies were conducted for 2-Hydroxyquinoline (2-HQ) and

quinoline over the concentration range of interest at either 1 to 10 and 1 to 5

mass to volume ratios. Isotopic exchange of 40Ca and 45ca (6,000 cpm/mL)

was also investigated. Aqueous solutions of the chemicals were prepared in

filter-sterilized (0.2 pm) 0.005 or 0.05 M CaCl. Background matrix solutions

(0.005, 0.05 M CaCI2) were filter sterilized (0.2 gm) to minimize biodegradation

of organic solutes.

Experimental Setup

Batch techniques (Nkedi-Kizza et al., 1985) were used to assess

sorption/desorption kinetics and equilibrium constants for quinoline in sterile

systems. A stirred batch reactor was used to measure quinoline sorption

kinetics. The soil fraction < 50 gm was used in the stirred batch reactor to

minimize separation of the soil suspension. The soil fraction (2 g) was added to

150 mL 0.005 M CaCIl At various time intervals, the suspension was sampled

and immediately separated through a 0.45 jim teflon filter. The filtrate (C) was

analyzed to determine the quinoline concentration at various time intervals for 4

days. Flow-through column techniques (Brusseau et al., 1990) were utilized to

determine sorption rate coefficients for quinoline using sterile background matrix

solutions. The sterile soil was packed into a Kontes glass column (5 cm long,

2.5 cm i.d.). Bed supports on both ends of the column consisted of a teflon

diffusion mesh with a glass membrane porous filter (1 jim). The pumps and

tubing were disinfected by rinsing with methanol. The glass columns and

solution vessels were sterilized by autoclaving. After packing, approximately

150 pore volumes of 0.005 or 0.05 M CaCl2 solution were pumped through the

column to achieve saturated, steady water flow conditions. Experiments were

conducted under saturated, steady water flow conditions at pore water

velocities of 15 to 90 cm/hr. In displacement studies, the molarity (0.005 M,

0.05 M) and pH of the displacing solution were varied.

Solute concentrations were monitored continuously or by collecting

column effluent fractions. Flow through UV detection (Gilson Holochrome or

Milton Roy LDC) was monitored continuously at 230 nm for quinoline and 2-HQ

and 254 nm for PFBA. Detector response was recorded using a strip chart

recorder (Fisher Series 5000). Effluent samples were collected intermittently

and analyzed by HPLC-UV techniques (Gilson 115 UV detector, Gilson Model

302 pump, Waters WISP 710B autosampler, HP333492A Integrator) to verify

sample purity and to compare the initial solute concentration to the maximum

effluent concentration. Quinoline and 2-HQ were eluted from a reversed-phase

column (Supelco LCPAH column) at a flow rate of 1 mL/min with a mobile

phase of 10/10/80 (v/v/v) methanol, acetonitrile and water adjusted to pH 2

with HCI. Soil column effluent pH was monitored on-line using an Ingold

microelectrode (Lee et al., 1991). Effluent fractions of the radiolabeled

compounds were collected with an automatic sample collector (ISCO Model

273). The activity of each radiolabeled compound was assayed using a liquid

scintillation counter (Searle Delta 300).

Data Analysis

Retardation factors (R) were calculated from area above the BTC for

quinoline and naphthalene (Nkedi-Kizza et al., 1987); a linear extrapolation

technique was used to extend the BTCs to C/Co= 1 in order to estimate the

area above the BTC. For 45Ca pulses, the R was calculated by moment

analysis techniques (Brusseau et al., 1990). The curve fitting program CFITIM

(van Genuchten, 1981), which is based on nonlinear least-squares optimization

techniques, was used to estimate the Peclet number (P) from the BTC for

PFBA. For nonsorbed solutes (R= 1), two model parameters can be optimized:

P and the solute pulse size (J). Since the pulse size was determined

experimentally, only the value for P was estimated by fitting to the measured

BTC for PFBA or 3120p. For sorbed solutes (R> 1), five model parameters can

be optimized: P,R, /, o, and J. For 4"5Ca and naphthalene BTCs, R was fixed

(estimated as described above), J was experimentally determined, P was fixed

as the value estimated from PFBA BTCs, and the values of nonequilibrium

sorption parameters (8 and () were estimated from parameter optimization

using the CFITIM program. For quinoline BTCs, the curve fitting program

FLOINT (Brusseau et al., 1989) with nonlinear sorption isotherms was used to

estimate the parameters when flow interruption techniques were used to

enhance the investigation of sorption nonequilibrium processes.

Results and Discussion

Sterilization Techniques

Initial batch studies were conducted to characterize the sorption of

quinoline and to assess techniques used for soil sterilization. Batch sorption

experiments were conducted using three nonsterilized air-dry soils and two soils

sterilized by steam autoclaving techniques. Autoclaving had minimal effect

(<2%) on the properties of the Norborne soil (Table 2-1). CEC measured by

45Ca isotopic exchange (Babcock and Schulz, 1970) and the MgNO3 extract

procedure (Rhue and Reve, 1990) resulted in similar values for nonsterilized and

autoclaved soils (See Table 2-1). Measurement of 45Ca isotopic exchange over

time suggested that cation exchange on Norborne soil was completed within

the first 5 minutes. Isotopic exchange, thus, migration of 4"9a into the interlayer

exchange sites, was virtually instantaneous. The CEC of the Eustis soil

increased about 28% after autoclaving. The standard deviation of the CEC

estimates for this sample, however, was high. Nonuniformity in soil sampling

may have caused some of this error. On the other hand, the increase may

have been caused by release of organic acids, alteration of the organic matter

structure, or a change in the interfacial pH though the bulk pH is the same.

The soils (Table 2-1) varied in pH, cation exchange capacity, and location

of charge. The quinoline sorption isotherm, plotted on a log-log scale, was

normalized to the protonated species (QH +) in the sorbed and solution phases

and the CEC of the soil (mmol(-)/g). The sorption data for all soils can be

represented by a single scaled isotherm (Figure 2-3), suggesting that quinoline

sorption occurs primarily via cation exchange. Sorption isotherms were

nonlinear (1/n = 0.68 to 0.8) over the concentration range investigated. At

higher concentrations (Figure 2-3), sorption of quinoline increases in the

Norborne and Webster soil. The S-type sorptive behavior for these soils occurs

at high concentrations (100 mg/L), where > 95% of quinoline is present as the

neutral species.

Cooperative interactions between the sorbed species and multilayer

sorption has been suggested to enhance quinoline sorption clay minerals at

high concentrations (Ainsworth et al., 1987). However, at this concentration

less than 1% of the cation exchange sites are occupied by quinoline. This

behavior may result from aggregation of sorption sites where quinoline sorption

occurs in collocation with clay mineral aggregates or organic matter. The


- D

Sterile Eustis
Sterile Norborne




Log C, [mol QH+/L]

Figure 2-3.

Quinoline sorption isotherms for three soils normalized
to their cation exchange capacity and to the fraction of
protonated species (See eq 2-7).





Eustis soil (pH 5.3) has a higher fraction of QH + present for the same initial

quinoline loading than the other soils (pH 6.4 and 6.9) (see Figure 2-2). The

isotherm nonlinearity for the Eustis soil remains constant at high concentrations

of QH +. This suggests that sorption of the neutral species may be occurring in

the higher pH soils. Another possible explanation is that high energy cation

exchange sites are the first sites occupied by quinoline, followed by sorption

onto lower energy sites such has been shown for sorption of inorganic

compounds (O'Connor et al., 1983).

Investigation of quinoline sorption kinetics suggested that sorption

occurred via a three step process (Figure 2-4). About 20% of quinoline sorption

occurred onto readily available or instantaneously accessible sorption sites.

These sites have typically been thought to exist on external regions of the

sorbent matrix (Brusseau and Rao, 1990). However, these sites may include

external sites or readily accessible internal sites depending upon the

architecture of the sorbent (Okuda, 1993). Sorption of quinoline occurs

predominantly on cation exchange sites located within organic matter and

smectite minerals. The slower rates of quinoline sorption likely correspond to

sorption and redistribution in the internal less-accessible regions of the sorbent.

In a binary solute batch system, quinoline sorption at low concentrations

was unaffected by the presence of its primary degradative metabolite, 2-

hydroxyquinoline (2-HQ), at pH 6.8 (Figure 2-5). The data points at the highest

quinoline concentration had the greatest amount of scatter in the data which



-0.2 -





Figure 2-4.


S 10 20 30 40 50 60
Time (hr)
Stirred batch reactor (a) and quinoline sorption onto
the Norborne soil fraction < 50 prm (b) (where C =
quinoline filtrate concentration and Co = the initial
quinoline concentration).


2HQ (mg/L) **
0 0
0 1



34 8 12
C (mg/L)

0 4 8 12 1
C (mag/L)

Figure 2-5.

Sorption of quinoline on the Norborne soil in the
presence of 2-hydroxyquinoline.





E 30




caused variation in the 1/n values. McBride et al. (1992) suggested that by

adding 2-HQ (5 and 20 mg/L) quinoline sorption in soil columns was reduced

as much as 23%. Competitive adsorption has been shown for NHCs such as

pyridine, quinoline, and acridine (Zachara et al., 1987) where the compounds

adsorb onto the same limited number of cation exchange sites. For HOCs,

competitive sorption is not likely because sorption occurs via partitioning (Chiou

et al., 1983). 2-HQ exists in its neutral form (pKa = 1.7) in the Norborne soil.

The predominant mechanism of 2-HQ sorption is hydrophobic partitioning, while

quinoline sorption occurs predominantly onto cation exchange sites. Therefore,

competitive sorption was not expected. If however, organic matter is located in

conjunction with the phyllosilicate minerals (Stevenson, 1985) quinoline sorption

may have been reduced due to the interference of 2-HQ and quinoline sorbing

in the same location of the organic matter-mineral complex. This behavior may

become more apparent in column studies (McBride et al., 1992) where

diffusional mass transfer constraints further limit sorption. These studies

suggest that 2-HQ production upon quinoline biodegradation is not likely to

reduce quinoline sorption by competing for available sorption sites.

Sorption Dynamics

Physical characterization. The 3H20 and PFBA breakthrough curves

(BTCs) for all soil columns were symmetrical and sigmoidal in shape (e.g.,

Figure 2-6) suggesting the absence of transport-related nonequilibrium. Peclet

numbers (P) were all greater than 80 indicating minimal hydrodynamic

0 9*

0 0







0 *

1 2 3 4 5

Pore Volumes (p)

Figure 2-6.

Examples of breakthrough curves for PFBA and 3H20
in Norborne soil columns.






dispersion (Table 2-2). Slight retardation (R z 1.15) of BTCs for 3H20 on the

Norborne soil suggests that this tracer was sorbed. Sorption of 3H20 onto a

soil high in iron oxide content that contains predominately kaolinitic clay

minerals has been previously reported (Nkedi-Kizza et al., 1982). The Norborne

soil also contains iron oxides with 2:1 type clay minerals (Zachara et al., 1990);

thus, 3H20 sorption is likely. Sorption of 3H20 may indicate that water is

exchanged with hydrated sorbed ions on the clay surface (Szecsody and

Streile, 1992). Batch studies were conducted to measure 3H20 sorption onto

sterile Norborne soil. The sorption coefficient (Kd) was 0.03 (+ 0.001) mL/g.

These Kd values are consistent with retardation factor (R) values ranging from

1.09 to 1.12 observed in different columns. The pore volumes determined by

3H20 after correcting for sorption resulted in similar pore volumes as

determined using gravimetric methods, and the BTC data for displacement.

3H20 was not sorbed onto the Eustis soil (R ; 1.0).

Chemical characterization. Monitoring 45Ca and quinoline sorption and

transport under specific chemical and physical conditions (e.g., molarity of

solution, pH, and pore-water velocity) will help understand mechanisms

influencing quinoline behavior. The data for 45Ca and quinoline were utilized to

explore the accessibility of cation exchange sites by an inorganic cation and an

organic cation. Nonequilibrium sorption was explored by observing isotopic

exchange of both 45Ca/40Ca and 14C-quinoline/12C-quinoline, as well as the

exchange of quinoline for calcium. The behavior of these two solutes were

Table 2-2. Column parameters for sterile soil columns.

Column ID



Norborne soil columns:
BQ5 0.005
A 0.005
B 0.05
BQ3 0.005
BQ8 0.005
BQ10 0.005
Floint 0.05
pH5.1 0.005
pH4.7 0.05

Eustis soil columns:

BQ2 0.005
DCMA 0.005

p e

pH g/cm3 mL/cm3


5.3 1.79
5.3 1.75



*nd = not determined

compared in a soil where sorption occurred primarily in organic matter (70%)

and kaolinitic minerals, and in a soil where sorption occurred primarily on

smectite type minerals and organic matter.

Figure 2-7 shows the BTC for 45Ca in 0.005 and 0.05 M CaCl2. The

retardation factor for 45Ca in the 0.005 M CaCl2 soil column is 37.6, whereas

the R in 0.05 M CaCl2 is 5.0. The sorption coefficient (Kd) of 45Ca is related

directly to the CEC of the soil, and inversely to the normality (N) of the

background electrolyte solution (Kd CEC/N) (Wilklander, 1964). Therefore, a

factor-of-ten increase in N should result in a 10-fold decrease in Kd. This was

indeed the case for sorption coefficients for 45Ca in the sterile 0.005 M CaC12



E cP. oo...
* 0 0 0

ED a




> 19.2 h

17.8 h

Flow Interruption

- 0

- DO






Pore Volumes (p)

Figure 2-7.

Quinoline and 45Ca breakthrough curves with flow
interruptions in 0.005 M (closed symbols) and 0.05 M
(open symbols) CaCL2 Norborne soil columns.





o Quinoline pH 6.2

* Quinoline pH 7

S45 Ca


. . i c i iI

column (1.0 mL/g) and 0.05 M CaCl2 column (11.0 mL/g). In contrast, ionic

concentration molarityy) of background matrix had minimal impact on quinoline

sorption at pH > 6.2 (Figure 2-7). The pH of the 0.005 M CaCl2 column is 7

and the pH of the 0.05 M CaCl2 column is 6.2. The fraction of protonated

species is greater at pH 6.2 (5%) versus pH 7 (1%). The decrease in pH in the

lower background matrix concentration (0.005 M) column may compensate for

the decrease in sorption due to higher ionic concentration. Batch studies at pH

6.2 for 0.05 M CaCl2 and in pH 6.8 for 0.005 M CaCl2 suggest that sorption

(Kd) is greater (z11%) as the molarity of the background matrix solution

decreases. Charge compensation in the diffuse double layer at higher

electrolyte concentrations may reduce the sorption of quinoline. In a subsoil

with a pH 7, the effects of ionic strength on quinoline sorption were negligible

(Zachara et al., 1986).

The influence of pH is evident upon comparing the BTCs in Figure 2-7

and 2-8 at the same background electrolyte concentrations. A decrease in pH

results in a increase in quinoline sorption. Increased sorption at lower pH

values is expected based on the increase in the fraction of QH The influence

of background electrolyte concentration was not clearly determined. Previous

investigation suggested that sorption decreased 60% at pH values near its pKa

when the ionic strength increased from 0.001 to 0.1 M CaCI2 (Helmy et al.,

1983; Zachara et al., 1986).





0 50 100 150 200
Pore Volumes (p)

Figure 2-8.

Quinoline breakthrough curves in 0.005 M (closed
symbols) and 0.05 M CaCI2 (open symbols) in pH
adjusted Norborne soil columns.


Measuring the influence of background electrolyte concentration on

quinoline was confounded by a simultaneous change in electrolyte

concentration and pH (Figure 2-7). Poising the soil pH at some value other

than the natural pH is often difficult. Repeated flushing of the soil column with

0.005 M CaCI2 resulted in a pH 6.9. The final pH after flushing the soil

column with 0.05 M CaCl2 ranged from 6.2 to 6.4, decreasing the pH about 0.6

pH units. As the pH of the soil approaches the pKa of the compound of

interest, sorption is increasingly sensitive to slight pH changes (Figure 2-2).

Therefore, sorption measurements of ionizable compounds must be conducted

at a constant pH.

The use of nutrient solutions was shown to alter the sorption of quinoline

(McBride et al., 1992). As a result, use of buffers was avoided. To alter the soil

pH, HCI may be added to the system. The addition of other ions may change

the overall ionic strength and the cation exchange complex, thereby influencing

quinoline sorption and possibly the phyllosilicate mineral structure. A titration

device was used to maintain a constant pH of soil-suspensions while quinoline

sorption was measured (Zachara et al., 1990). However, this procedure does

not lend itself to use in flow-through column techniques. In these column

experiments at the lower pH values, the background electrolyte solution was

adjusted with HCI and flushed until the pH was essentially constant ( 0.3 pH

units). Soil columns were flushed at 0.5 mL/min for about 2 weeks. Additional

acid was not added to adjust the pH of the quinoline solution due to changes in

electrolyte concentrations and ionic composition. Therefore, the pH was not

adequately controlled. Experimental techniques must be carried out with the

utmost detail when investigating the behavior of ionizable compounds. A

controlled experiment with the system poised at a particular pH value has not

been conducted to accurately measure the influence of ionic strength on

quinoline sorption in soil columns.

Flow interruption. The accessibility of the cation exchange sites (i.e, clay

interlayer positions and organic matter) was evaluated by examining the

dynamics of 40Ca/45Ca isotopic exchange and exchange of quinoline for 4Ca.

The 45Ca BTC was symmetrical and showed about a 5% drop in concentration

after an 18-hour flow interruption in the 0.005 and 0.05 M CaCl2 columns

(Figure 2-7). This suggests that cation exchange and diffusion into clay

interlayer sites and organic matter regions was rapid, and that near equilibrium

conditions were attained under flow conditions for the column. Flow

interruptions suggested that migration of 45Ca into interlayer sites and organic

matter matrices was not limiting mass transfer or isotopic exchange kinetics.

Szecsody and Streile (1992) also found isotopic exchange of 40Ca/45Ca to be

rapid in columns packed with clay-modified alumina. Exchange of 40Ca/45Ca in

organic matter was rapid and not limited by mass transfer into the organic

matrix (Nkedi Kizza et al., 1989). They speculated that the compensation of

charge may occur at the exterior of the organic matter matrix and Ca does not

necessarily need to migrate within the sorbent.

Considerable asymmetry of the quinoline BTC at pH 7 in the sterile

Norborne soil (0.05 and 0.005 M CaCl2) is indicative of nonequilibrium behavior

during displacement of quinoline for 4Ca (Figure 2-7). A large drop in effluent

concentration (;35 to 50%) during the flow interruptions greater than 17 hours

indicates strong nonequilibrium behavior (Brusseau et al., 1989). Figure 2-9

shows the nonequilibrium behavior upon repeated flow interruptions in the 0.05

M CaCl2 Norborne soil column at pH 6.2. The first flow interruption (at 20

hours) results in a 50% drop in concentration. Subsequent flow interruptions

(24 hours) suggested that quinoline sorption is rate-limited into interlayer

positions of phyllosilicate minerals and possibly into interior regions of organic

matter matrices.

Symmetrical BTCs for PFBA and 3H20 preclude physical nonequilibrium

constraints (e.g., mobile-immobile water) as a possible reason, and 45Ca cation

exchange was rapid. Therefore, quinoline sorption nonequilibrium must be due

to other constraints.

O'Loughlin et al. (1991) reported that sorption of a N-heterocyclic

compound (2-methyl pyridine) into 2:1 clay interlayers was rate-limited, whereas

sorption onto edge-sites of kaolinite was rapid which suggests that steric

hindrances are limiting sorption. However, similar molecular dimensions of

quinoline (1.02 nm X 0.76 nm x 0.36 nm; Weast, 1984) and hydrated Ca (0.6

nm; Bohn et al., 1979) suggest that size considerations alone are not likely to

account for the observed sorption nonequilibrium of quinoline. Szecsody and

Pore Volumes (p)

Figure 2-9.

Figure 2-9.

Repeated flow interruptions for quinoline in a 0.05 M
CaCl2 (pH 6.2) Norborne soil column and bicontinuum
model fit.






Streile (1992) attributed sorption nonequilibrium to kinetic constraints from site-

specific chemical processes between the quinoline and montmorillonite. Over

the concentration range used in this study, the protonated form is likely the

predominate species sorbed via cation exchange. In the bulk solution of the

soil columns, quinoline exists essentially in the neutral form. Zachara et al.

(1990) demonstrated that even when pH values are pH (pKa +2) and most of

quinoline exists in its neutral form, the quinolinium ion is still the predominant

form sorbed. In addition, surfaces can be up to 2 units lower in pH than the

bulk solution pH (Bates, 1973) and protonation reactions are rapid. Therefore,

availability of quinolinium ions in solutions is not likely to limit sorption.

The bicontinuum model provided an inadequate description of quinoline

behavior in Norborne soil columns (Figure 2-8, 2-9). The frontal portion of the

curve adequately describes the rapid access to the easily accessible external

sites. Nonlinearity of the quinoline sorption also caused self sharpening of the

front of the BTC. The model fits were optimized for nonequilibrium parameters

B and o (Table 2-3), and are shown in Figure 2-9. The quinoline displacement

in the column adjusted to pH 4.7 was conducted at 0.5 mL/min, whereas

displacement studies in the other three columns listed in the Table 2-3 were

conducted at 2 mL/min. The Norborne soil has 0.16% organic matter in

addition to smectite clay minerals. The large fraction (0.5) of sites

instantaneously accessed by quinoline was attributed to sorption on edge sites

(as much as 20%) and easily accessible interlamellar sites of smectite minerals

Table 2-3. Summary of estimated transport parameters for quinoline.

ID pH R Kf 0 B F k2

BQ5 7.0 11.0 3.18 0.762 (0.52-1.0) 0.536 (0.47-0.59) 0.493 1.336
Floint 6.2 12.6 3.87 0.814 (0.41-1.22) 0.508 (0.43-0.59) 0.466 1.337
pH4 4.7 28.6 8.58 0.178 (0.17-0.18) 0.821 (0.79-0.85) 0.814 0.074
BQ10 3.0 140 43.1 0.261 (0.11-0.41) 0.503(0.26-0.75) 0.499 0.041
BQ2 4.75 0.69 1.727 (0.98-2.47) 0.507 (0.41-0.60) 0.375 8.286
DCMA 5.3 11.0 1.82 0.984 (0.39-1.58)) 0.535 (0.42-0.65) 0.488 2.615

values in parenthesis are 95% confidence intervals.

and exterior regions of organic matter. Simultaneously describing the large

fraction of instantaneously accessible sites and the slow redistribution of

quinoline within the clay interlayers is not possible using the bicontinuum model.

Therefore, rapid sorption of the quinolinium ion followed by the rate-limited

diffusion of quinoline into the phyllosilicate minerals is not an accurate

conceptualization for quinoline sorption.

Specific chemical interactions (e.g., hysteresis, reconfiguration of the

molecular arrangement) likely limit desorption, and steric hindrances may limit

redistribution within phyllosilicate minerals. The molecular configuration was

suggested to change from an upright position to a planar position within clay

minerals (Zachara et al., 1988). As a result, desorption is strongly inhibited due

to delocalization of charge over the entire molecular surface. Subsequent

migration within interlamellar regions may be restricted due to desorption and

redistribution of the quinoline molecule and limited accessibility due steric to


The significance of the interlayer spacing in this smectite clay mineral

during quinoline sorption is apparent, given that the majority (up to 80%) of the

charge associated with the clay mineral originates in the interlayer spacing from

isomorphic substitution. The predominant form of clay in the Norborne soil is

biedellite which is characterized by substitution of Al+3 for Si+4 in the

tetrahedral layer. The clay fraction was isolated from the Norborne soil and

prepared for X-ray diffraction to measure changes in d-spacing upon

replacement of quinoline for 40Ca. Mounts were prepared by placing a known

amount of clay suspension onto a clay tile and saturating with 1 M CaCI2. The

sample was rinsed with deionized water to remove excess Ca. The tile was

equilibrated for about 48 hours at both 56 and 87 % relative humidity and the d-

spacing was measured. Sufficient quinoline was then added to the clay tile to

occupy 1% of the total sites. Measurements of the d-spacing were repeated at

56 and 87 % relative humidity. A decrease in the d-spacing upon addition of

quinoline would indicate the collapse of the clay interlayers and a potential

source of nonequilibrium sorption.

No obvious changes in d-spacing were indicated in the 1% quinoline

saturated samples compared to the Ca saturated samples at either relative

humidity. A decrease in the d-spacing was detected upon decreasing the

relative humidity. The d spacing was 1.6 nm at 87% relative humidity of which

0.92 nm is occupied by an octahedral and tetrahedral layer. Therefore, the

interlamellar region is approximately 0.68 nm. This procedure was limited by

the fact that only 1 % of the total CEC sites were occupied by quinoline; 99% of

the exchange sites were occupied by Ca. Therefore, no changes were

detected. To enable the detection of d spacing changes a larger fraction of

sites would need to be saturated with quinoline. However, saturating the

exchange complex with quinoline would likely alter the sorption mechanism and

would not be comparable to low quinoline concentrations (see Figure 2-3).

Figure 2-10 conceptualizes the process hypothesized for quinoline

sorption onto smectite clay minerals. The size of the interlayer spacing of the

smectite clay, the Ca, and quinoline are approximately drawn to scale.

Quinoline replaces Ca on edge and readily accessible interlayer CEC sites

(Figure 2-10a), representing the fraction of instantaneous sites (F) associated

with the clay minerals. After this initial step, quinoline must desorb and migrate

further within the interlamellar region of the clay mineral. Displacement of Ca

by quinoline in interlayer regions may be physically constrained (Figure 2-10a),

which may contribute to sorption nonequilibrium. Ca is hydrated and initially

occupies CEC sites in the interlayer positions. Smith et al. (1992) suggested

that quinoline displaced interstitial water upon reorientation to a planar position

on the surface. Therefore, the hydration energies associated with quinoline and

40Ca may be important in understanding rate-limitations of quinoline sorption.

Reorientation of quinoline to planar position

( )
^^\ $I0

Redistribution of quinoline within clay interlayers

Collapse of clay interlayers
Figure 2-10. Conceptual diagram of quinoline sorption onto smectite clay minerals.

0 0

Quinoline may also be drawn into a planar orientation (Figure 2-10Ob)

delocalizing the charge over the whole quinoline molecule. At this stage,

quinoline molecules in the solution phase may pass further into the interlamellar

regions of the clay mineral due to compensation of the electrostatic charge by

the previously sorbed quinoline molecule (Figure 2-10c). Some of these sites

may essentially be inaccessible once quinoline has occupied the interior of clay

minerals and formed a stable surface complex. After breakthrough and

washout of quinoline from the Norborne soil columns, mass balance suggested

that 5 to 10 % of the quinoline introduced into the column remained on the soil.

Repeated washing with 80% methanol was insufficient to completely wash out

residual quinoline within the interior clay aggregates. Introduction of a cation

more selective for the exchange complex than quinoline would be a more

efficient method for removing quinoline from the exchange complex.

The inability to successfully remove residual quinoline from interlayer

positions further supports that strong quinoline surface complexes are formed

or that the interlayers have collapsed (Figure 2-10Od). This depicts the

tetrahedral layer charge being drawn to the quinoline molecule and collapse the

interlayer spacing restricting further migration into this region. Electrical

neutrality must be maintained at all times suggesting that two quinoline

molecules must replace one calcium. Therefore, the total collapse of the

interlayer regions is not likely. However, formation of strong surface complexes

and several molecules sorbed in the interlayers may create a buildup of

molecules redistributed throughout the interlamellar region.

Isotopic exchange of 12C-quinoline/14C-quinoline was measured to

determine the exchange of quinoline molecules during displacement with 0.05 M

CaCl2 in a Norborne soil column (Figure 2-11). The breakthrough of quinoline

was first monitored in a 0.05 M CaC12 background matrix solution. After 7 flow

interruptions (3 shown in Figure 2-9), the equilibrium solution concentration was

98% of the influent concentration. At this point, a solution of 14C-quinoline and

12C-quinoline (same total concentration) in 0.05 M CaCI2 was introduced into

the column. The breakthrough of 14C-quinoline was delayed following

preconditioning with 12C-quinoline and the drop in relative concentration (25%

versus 50%) during flow interruption decreased. Apparent increased retention

(delayed breakthrough) in the 14C-quinoline column may have been caused by

decreased pH (6.1). However, the change in pH causes about a 1% increase

in the fraction of QH+ and it is not likely to cause this shift in breakthrough.

Two cases will be presented as alternatives for the isotopic exchange data.

First, the decreased drop in relative concentration of the 14C-quinoline versus
12C-quinoline BTC suggests that equilibrium is more readily approached by 14C-

quinoline. A decrease in rate-limited sorption sites would result in a reduced

drop during the flow interruption. This may occur if quinoline surface

complexes are formed in interlamellar regions. However, this would

simultaneously decrease cation exchange capacity, resulting in early quinoline

breakthrough (lower R). In fact breakthrough was delayed, therefore, this was

reasoned not to be a viable option.

0 0
0 *
0 0
0 0
0 *




0 *)
*oe* 0
*0 0


Pore Volumes (p)

Figure 2-11.

Isotopic exchange of 12C-quinoline and 14C-quinoline
in 0.05 M CaCl2 (pH 6.2) in the Norborne soil.






0 12 C-Quinoline

* 14 C-Quinoline



Another possibility is that the 140C-quinoline approaches equilibrium more

rapidly than the 12C-quinoline. The sigmoidal shape of the BTC for 14C-

quinoline is indicative of equilibrium sorption and a linear isotherm is expected

from exchange of 12C- and 14C-quinoline. Quinoline sorption isotherms were

nonlinear 1/np0.7 in batch systems upon exchange of quinoline for 4Ca. The

self-sharpening front for the 12C-quinoline BTC is indicative of nonlinear sorption

(Brusseau and Rao, 1989b). The sharp front may also indicate nonequilibrium

conditions suggesting access into the interlayer positions and replacement of

calcium is difficult. Initial access of quinoline into interlayer positions may

enhance subsequent access of interlayer regions due to charge compensation

and reorientation (Figure 2-10c). To test this hypothesis, a BTC of 14C-

quinoline on the backside tail (5-10% residual quinoline) could be conducted. If

the sharp front occurred on the BTC then it would suggest that as the percent

of quinoline on the exchange complex increased access to other interlayer sites

would increase.

From these two cases, nonequilibrium conditions prevail upon

introduction of quinoline, suggesting that some sites are extremely constrained

by diffusional and chemical factors. A fraction of the sites are considered to be

unavailable and thus, bioavailability is likely to be limited.

In the pH 3 and pH 4.6 columns, the minimal drop in effluent

concentration during flow interruption (Fig. 2-6) suggested that quinoline

sorption is near equilibrium. However, the relative concentration only reaches

95-98% after the flow interruption. In the pH 6.2 column, C/Co approached 1

rapidly after the flow interruption. In the pH 6.2, 4.7, and 3 columns, the flow

interruption resulted in a 50, 18, and 2% drop in relative concentration,

respectively. Diffusion into clay interlayer positions is pH dependent. At first

glance it appears that nonequilibrium is greater at higher pH values. However,

as the pH decreases the fraction of protonated species increases and R

increases which may alter access into these regions. A larger R, a highly

selective exchange coefficient, and steric hindrances may further limit quinoline

entry into the clay interlayer positions. Rate-limited sorption of quinoline may be

related to both the magnitude of selectivity coefficients (Ainsworth et al., 1987)

and the ability of quinoline to delocalize its charge over the entire surface of the

compound (Zachara et al., 1990).

Replacement of hydrogen ions for 40Ca on exchange sites may alter the

clay interlayer environment, thus, quinoline migration into interlayers. As a

result, the ability to access interlayer positions as the pH decreases may be

further constrained. It may be possible that mass transfer is restricted beyond

the time allowed for flow interruption (8.8 h, pH 3 column), modeling the data

assuming flow interruption occurred for a longer period of time (16.6 d) would

result in a large drop during flow interruption. The model fit (granted the error

associated with the use of this model) suggests that mass transfer is more

constrained than indicated for the 8.8 hour flow interruption. However, a flow

interruption for as much as 10 days in a pH 5 column resulted in only an 8%

drop in the relative concentration and approached a relative concentration of

98%. The k2 values determined from model fits for the lower pH columns are

less than the higher pH columns (Table 2-3). Sorption is about 25 times faster

at higher pH values than in lower pH soils. The trend indicates reduced access

to interlayer positions as pH decreases.

Consideration must be given to differences in the nature of organic

matter versus clay minerals in describing the quinoline sorption. Diffusion of

quinoline into interlamellar regions was suggested to be rate-limited. However,

sorption of quinoline into organic matter matrices was not clearly defined in the

Norborne soil due to the presence of smectite clay minerals. Therefore, column

studies were conducted on a Eustis soil where a majority of the CEC is located

in the organic fraction of the soil and the remainder are associated with the

kaolinitic clay minerals.

A large drop in effluent concentration (;35%) during the 17.8 hr flow

interruption in the Eustis soil column is indicative of nonequilibrium sorption into

organic matter matrices (Figure 2-12). It was suggested that sorption of the

neutral species behaved similarly to HOCs (IOMD), while sorption of the

quinolinium ion onto exterior regions of organic matter was rapid (cation

exchange) (Brusseau et al., 1991). However, flow interruption techniques

enhanced detection of sorption nonequilibrium and suggested that

nonequilibrium conditions predominated in organic matter matrices (Eustis soil)

and phyllosilicate minerals (Norborne soil). Access to rate-limited sites was


0 0

00 0

0 15.5 h
* 15.5 h


07 Oete

23 h

Pore Volumes (p)

Figure 2-12.

Breakthrough curves of quinoline in Eustis soil with
0.005 M CaCI2 and 30% methanol.









* 0.01 NCaCI2

o 30 % Methanol

suggested to be faster into the organic matter matrix than into the clay minerals

(Table 2-3). Organic matter is thought to be a flexible deformable organic

polymer; therefore, migration into this type of matrix may be less restricted than

into interlamellar regions of clay minerals.

Nonequilibrium due to IOMD arises due to restricted diffusion within the

polymer-like matrix of organic matter (Brusseau et al., 1991). Specific

interactions of quinoline with functional groups of organic matter are likely to

change the nature of this flexible organic polymer. Redistribution of charge

upon migration of quinoline within the organic matter may cause the matrix to

collapse around the quinoline molecule and restrict diffusion. Brusseau et al.

(1991) suggested that Ca sorption occurred at the exterior or the organic matter

because of the long range interaction of electrostatic charge. Alternatively, Ca

migration into interior regions of organic matter may be rapid. The

nonequilibrium behavior of quinoline suggests that quinoline migration into the

interior regions of the organic matter is rate limited. Hydrophobic and cation

exchange interactions may occur with the quinoline molecule due to charge

separation on the molecule.

Figure 2-13 is a proposed schematic diagram of organic matter (B6har

and Vandenbroucke, 1987). The structure is composed of regions of randomly

distributed hydrophobic and hydrophilic regions comprised of aromatic and

aliphatic structures, respectively. Envision quinoline migration into this organic

matrix: specific interactions between quinoline and hydroxyl groups may occur


\;,^^0 0

00/ 0 ( 0

NH ...

Figure 2-13. Structural representation of organic matter (adapted
from Behar and Vandenbroucke, 1987).

followed by redistribution of charge and reconfiguration of the matrix around the

quinoline molecule. The hydrophobic portion of the molecule may associate

and partition into the aromatic region.

Addition of cosolvents increases solubility of organic compounds and

decreases sorption. In addition, the organic matter matrix may swell, increasing

accessibility to the interior of the organic matter matrix thereby reducing

sorption nonequilibrium (Nkedi-Kizza et al., 1989; Lee et al., 1991). However,

the fraction of instantaneous sites (F) decreased as the matrix swelled because

the surface area to volume ratio decreases (Lee et al., 1991).

Other specific solute-solvent and solvent-sorbent interactions increase the

complexity of describing sorption of ionizable organic compounds in mixed

solvents systems (Lee et al., 1992). The pKa of acidic functional groups

associated with the sorbent may increase upon addition of solvents (Lee et al.,

1992). Thus, in the presence of solvents at a given pH, the functional groups

become more neutral and reduce electrostatic interactions. In addition, the pKa

of the quinoline decreases upon addition of cosolvents. Therefore, at a given

pH the amount of neutral species present increases. These solvent-sorbent

and solute-solvent interactions may enhance the migration of molecules within

the matrix by increasing the permeability and reducing specific quinoline-sorbent

interactions, thereby reducing sorption nonequilibrium. However, the fraction of

instantaneous sites may decrease.

Addition of methanol (30%) reduced quinoline sorption (Figure 2-12).

Quinoline solubility increases with increasing volume fraction methanol

corresponding to a decrease in sorption upon solvent addition (Fu and Luthy,

1986a). Transport parameters for two Eustis soil columns are presented in

Table 2-3. Cosolvent effects on solubility and sorption of quinoline is

confounded by specific solvent-sorbent and solvent-solute interactions. The self

sharpening front is indicative of isotherm nonlinearity. Sorption of quinoline in

up to 40% methanol was nonlinear (Zachara et al., 1988). However, sorption

isotherms of pesticides have shown increased linearity upon addition of

cosolvents (Nkedi-Kizza et al., 1985).

Direct observation of the organic matter surfaces was attempted by

taking a scanning electron micrograph (SEM) of an organic soil (Figure 2-14).

The soil was dried at 60C and gold coated to prepare the sample. The soil

was not fixed with glutaraldehyde or dehydrated with solvents to minimize

structural changes due to fixative agents. The SEM shows the heterogeneity

association with the surface of organic matter (Figure 2-14a). However, the

interior of the organic matrix which is the major sink for HOCs is not visualized

using this technique. We do, however, come to appreciate the complexity of

the organic surface and the relative scale at which we need to view the organic

matrix to observe the location of the contaminants within this region. For

example, the magnification of this SEM is 6000 X. The scale provides reference

to the size of quinoline. The quinoline molecule conceptualized in Figure 2-10 is

1 nm in length, suggesting that about 3000 quinoline molecules

would fit along the scale key. One begins to envision molecules diffusing

through this heterogeneous media and the concept of rate-limited sorption.

..-. ,

,,. ,. b.. j '

Figure 2-14.

6000x (a)

electron micrograph of an organic soil at
and lOOOx (b).



The SEM photograph also shows fungal spores that have been preserved

within this organic soil. Further investigation of the organic matter surface

revealed fungal mats forming on the organic matter surface (Figure 2-14b). The

prolific growth of fungal spores and hyphal mass depicts the colonization of the

soil surface by microbial biomass and the possibility of altering contaminant

sorption and transport.


Quinoline is sorbed predominately on cation exchange sites on clay and

organic matter. Sorption is therefore dependent on quinoline speciation as

influenced by pH. Kinetics of ion exchange are rapid; therefore, quinoline

sorption is controlled by accessibility of sites, most likely through surface

complexation or inaccessibility due to steric hindrances. Quinoline sorption

potentially occurs via a three-step process -- an initial rapid phase sorbing onto

instantaneously accessible sites, followed by a reorientation of the molecule on

the surface, and subsequent redistribution within the organic matrix and

interlamellar regions of phyllosilicate minerals. Therefore, conceptually the

bicontinuum model is not adequate to describe quinoline sorption.

Quinoline sorption within phyllosilicate minerals and organic matter is

rate-limited. Sorption of quinoline on the outer edges of smectite clay minerals

may impede access of other quinoline molecules. A buildup of molecules at

clay interlayers may occur as desorption and migration into interlamellar regions

is limited. Therefore, access to the interlayer position may be blocked if

quinoline migration and redistribution is rate-limited. In addition, specific

quinoline-sorbent interactions (reorientation and charge delocalization) limit

desorption from the surface. On the other hand, if sorption occurs onto a

preconditioned quinoline soil containing phyllosilicate minerals access to

interlayer regions may be enhanced due to compensation of charge by the

preexisting quinoline. Sorption within organic matter is likely limited by specific

electrostatic interactions which cause reconfiguration of the organic-type

polymers. Both sorbents restrict migration into interior regions causing rate-

limited sorption.

The bioavailability of quinoline sorbed within either sorbent matrix is likely

to be limited. As indicated by repeated washing of the Norborne soil, 5 to 10 %,

of the solute remains sorbed. This fraction is therefore, rendered unavailable to

the microorganisms based on the location of the solute and the microorganism

(See Chapter 4 for further discussion). The distribution is microbial biomass in

the organic soil (Figure 2-14) suggested that microbial biomass may proliferate

and cover the soil surface.

The addition of microbial biomass to soils and aquifers may substantially

alter the nature of the sorbent surface (Figure 2-14). In the absence of

biodegradation, the impact of biomass on contaminant sorption and transport is

of great interest.



Bioremediation practices attempt to increase microbial activity or populations

in order to degrade organic contaminants present in soils or aquifers. Indigenous

microbial activity and/or populations may be increased by providing nutrients

essential for bacterial growth, or axenic bacterial cultures known to degrade

specific compounds may be injected directly into contaminated sites. Growth or

addition of bacteria may drastically alter the chemical and physical characteristics

of solid surfaces (Fletcher, 1991). Therefore, the impact of bacterial biomass on

contaminant behavior in porous media near hazardous waste sites is of interest.

In addition to contaminant biodegradation, addition of bacteria to porous

media may result in: 1) bacterial growth or transport through the porous media

leading to pore clogging as a result of physical straining; 2) biosorption and

bacterial migration facilitating contaminant transport; and 3) bacterial sorption onto

soil surfaces altering the sorption capacity. Although bacterial migration through

sandy soils and aquifers is well documented, bioremediation attempts have failed,

among other reasons, due to the inability of injected bacteria to reach

contaminated sites (Gibson and Sayler, 1992). Physical, chemical, and microbial

factors controlling bacterial transport in porous media have recently been

summarized (Harvey, 1991, Lindqvist and Enfield 1992b, Tan et al., 1992).

Bacterial transport may be limited by physical constraints imposed by the porous

media, such as soil structure and pore size distribution (Lindqvist and Enfield,

1992b). Straining or filtration occurs in soils and aquifers when bacteria are too

large to pass through soil pores; this results in pore clogging, which restricts

further penetration of bacteria (Herzig et al., 1970; Harvey, 1991). Once bacteria

become clogged in the soil pores, water flow is also restricted, and the path of

water flow can be altered (Vandevivere and Baveye, 1992).

Chemical constraints, such as adsorption of bacteria, may also limit bacterial

migration through soils and aquifers (Harvey et al., 1989; Harvey, 1991, Bales et

al., 1991; Tan et al., 1992). Bacteria that are hydrophobic and are minimally

charged have the greatest potential to sorb onto surfaces; however, many other

factors may influence bacterial attachment (van Loosdrecht et al., 1987). Because

of bacterial adsorption by soils (Daniels, 1972) and clay minerals (Stotzky and

Rem, 1966), the contaminant sorption capacity of the soil may be altered. Bacteria

grow after they attach to surfaces if essential carbon and energy sources are

available. Growth and development of bacterial colonies generally coincide with

the production of extracellular polysaccharides and promote the formation of

bacterial biofilms (van Loosdrecht et al., 1990; Fletcher, 1991). Bacterial biomass,

therefore, contains live and dead cells and cell exudates extracellularr polymers).

Under nutrient- and substrate-rich conditions, as may be the case near wastes

sites, biofilm formation may create diffusional barriers leading to nonequilibrium

sorption of contaminants. This is generally the case for wastewater treatment by

filtration through activated carbon beds (Speitel et al., 1989; Rittman and McCarty,

1978). Bacterial biomass may physically alter the accessibility of sorption sites,

thereby reducing contaminant sorption. To further complicate the problem,

bacterial biomass may act as an additional sorbent, thereby increasing

contaminant sorption.

Sorption by various microorganisms in aquatic systems has been shown for

hydrophobic organic chemicals (HOCs) (Baughman and Paris, 1981; Tsezos and

Bell, 1989), metals (Scott and Palmer, 1990), and organic amines (Crist et al.,

1992). A consensus on biosorption mechanisms has not been reached, and

usually no distinction is made between sorption onto extracellular regions and

absorption into the cells. Properties such as aqueous solubility and log Kow (Kow

= octanol water partition coefficient) for the contaminant (Selvakumur and Hsieh,

1988) and bacterial lipid content (Bitton et al., 1988) have been correlated to

biosorption of HOCs. Biosorption of trace metals has been shown to occur via

adsorption onto extracellular bacterial capsules with minimal intracellular uptake

(Scott and Palmer, 1992). Sorption of organic amines by algae has also been

described by mechanisms including ion exchange and hydrophobic bonding (Crist

et al., 1992). Occurrence of biosorption and bacterial migration, regardless of the

underlying mechanisms, suggests the potential for biofacilitated transport of

contaminants. Lindqvist and Enfield (1992a) demonstrated bacterial-facilitated

transport of two HOCs (dichloro-diphenyl-trichloroethane and hexachlorobenzene)

in sand columns. Biosorption technology has been commercialized to mobilize

metals in the mining industry (Ehrlich and Brierley, 1990). However, biofacilitated

transport of NHCs bases has yet to be demonstrated.

Research Question and Tasks

At the field scale, the question of interest is: what are the consequences

of bioenhancement or bioaugmentation practices in attempts to remediate

contaminated sites? Specifically, do bacteria alter the sorption and transport of

NHCs? In this chapter I examine the impact of bacterial biomass on the

sorption and transport of three solutes (naphthalene, 45Ca, and quinoline) in a

subsurface soil. These compounds were selected because of their known

specific interactions in soil: 1) naphthalene was selected to probe hydrophobic

interactions with the nonpolar organic phase; 2) 45Ca was selected to probe

electrostatic interactions with the cation exchange sites; and 3) quinoline, a N-

heterocyclic organic base, was selected because it can exist as a neutral

organic compound interacting with the organic phase or as a quinolinium ion

interacting with cation exchange sites. Miscible displacement techniques were

used to measure sorption and transport of the above compounds during

steady, saturated water flow conditions through homogeneously-packed, sterile

or bacterial-inoculated, soil columns. A fine-textured silt loam soil (Norborne;

fine-loamy, mixed, mesic Typic Argiudoll) was chosen for these experiments

because of the extensive characterization of quinoline sorption by this soil

(Zachara et al., 1988; 1990). Sorption of naphthalene by the organic fraction of

soil is well documented (Chiou et al., 1983; Karickhoff et al., 1979). Pre-

inoculation of the Norborne soil with bacteria (108 cfu/g) simulates

contaminated subsurface soils and aquifers where bacterial populations may be


Materials and Methods


The Norborne soil was used for these studies (Table 2-1). Glassbeads

(average diameter 150 /m; Alltech Associates) and inert quartz sand (< 2 mm)

were used as inert solid support material. All sorbents were sterilized using

steam autoclaving as referenced in Chapter 2.


Pentafluorobenzoic acid (PFBA; 150 Ag/mL) was used as conservative,

nonsorbing tracer to assess the hydrodynamic dispersion and extent of physical

nonequilibrium conditions prevailing during transport through the soil columns

(Brusseau et al., 1989). Quinoline and naphthalene concentrations in the

influent solutions for the column studies ranged from 4 to 10 pg/mL. Isotopic

exchange of 40Ca and 45Ca (6,000 dpm/mL) was also investigated. Aqueous

solutions of the chemicals were prepared in filter-sterilized (0.2 jim) 0.005 or

0.05 M CaCI2. Sorbates were monitored by HPLC-UV for quinoline and

naphthalene, and by radio-assay techniques for 45Ca (See Chapter 2).

Bacterial Strains and Culture Conditions

A strain of Pseudomonas sp. 3N3A capable of degrading quinoline and a

mutant strain (B53) derived from the 3N3A strain [obtained from Brockman et

al. (1989)]. Incorporation of two proteins for bacterial enumeration rendered the

organism incapable of degrading quinoline (McBride et al., 1992). The B53

isolate was used to determine the impact of biomass on sorption and transport

of quinoline where degradation was not a factor.

The B53 and 3N3A strains were grown for 17.5 hours on tryptic soy

broth (3 g/L) at 28 C on a rotary shaker (100 rpm). Bacterial cells were

harvested by centrifugation, washed two times and diluted to the desired

bacterial density with the appropriate background matrix solution (0.005 or 0.05

M CaCl2). Bacteria were allowed to equilibrate overnight in the desired matrix

prior to each experiment. Plate counts were done using tryptic soy agar (TSA)

and 4 day incubation periods at 28C. Plate counts were verified by visual

inspection of bacterial suspensions using a hemacytometer. A phase-contrast

microscope (Wild Neenbrugg) was used for counting the bacteria in the


Bacterial Inoculation

A 0.5 mL-aliquot of the appropriate bacterial suspension was placed in

an aspirator. The sterile soil (50 g) was thinly spread on aluminum foil and the

bacterial suspension was sprayed on the soil in a fine mist to uniformly

distribute the bacteria. The soil sample was mixed thoroughly to ensure

homogeneous distribution of the bacteria. The aspirator was rinsed with a 0.5-

mL aliquot of filtered (0.2 j/um) CaCI2, and the rinsate was sprayed on the soil.

The initial inoculation rate was 106 cfu/g soil unless otherwise indicated. The

soil was mixed again, and a subsample was taken for water content

determination. The soil-water content following bacterial addition ranged from 5

to 10%.

Column Studies

Miscible displacement techniques were used to characterize the transport

of PFBA, 45Ca, quinoline, and naphthalene. The sterile or bacterial inoculated

soil was packed into a Kontes glass column (5 cm long, 2.5 cm i.d.) as

described in Chapter 2. After packing, approximately 150 pore volumes of

0.005 or 0.05 M CaCI2 solution were pumped through the column to achieve

saturated, steady water flow conditions and uniform bacterial populations (108

cfu/g). Soil columns varied in bacterial density and type (sterile, or inoculated

with either B53 or 3N3A isolate) and in ionic strength (0.005 or 0.05 M) of the

displacing solution. Solute concentrations were monitored continuously or by

collecting column effluent fractions. Dissolved oxygen (DO) in the soil column

effluent was measured at different pore-velocities from 0.6 to 90 cm/hr. A

vessel was purged with N2, effluent from the column introduced, and DO

measured with a dissolved oxygen electrode (Yellow Springs Instruments 5750).

Sorption of quinoline by live cells of the B53 and 3N3A isolates was

measured at a bacterial density of 108 cfu/mL. The initial quinoline

concentrations were 1, 4, and 8 /g/mL. Bacterial suspensions were

equilibrated with quinoline at 5 C for 1 hr to minimize intracellular uptake and

possible biodegradation by the 3N3A isolate. Biosorption of 45Ca was

measured at room temperature (22-25C). Samples were centrifuged for 20

min at 1,250 g at 5C to separate the cells from the aqueous phase. Quinoline

solution concentrations were measured by HPLC to monitor for possible

biodegradation products. Biosorption was calculated as the difference in the

initial and final solution concentrations. Miscible displacement techniques

described earlier were employed to measure biosorption by bacteria "attached"

to glass microbeads. Glass microbeads (average diameter 150 ,m; Alltech

Associates) were inoculated with 107 cfu/g, packed into a column, and

saturated with 0.05 M CaCI2 for 48 hours at a pore-water velocity of 13.5 cm/hr.

BTCs for quinoline, 45Ca, and naphthalene were measured simultaneously by

injecting a mixture of these three solutes on the column; this was done so that

BTCs for all three solutes were obtained under identical hydrodynamic and

microbial conditions. Effluent fractions were collected and monitored by the

techniques stated above.

Surface Accessibility

A comparison of the estimated values of the bicontinuum sorption model

parameters (Chapter 1 and 2) for the sterile and inoculated soil columns were

used for a quantitative assessment of: 1) the hydrodynamic impacts, based on

P; 2) the changes in equilibrium sorption capacity, based on K ; and 3) the

accessibility of sorption regions, based on F and k2.


The behavior of PFBA in sterile and bacterial-inoculated columns is

represented by the PFBA breakthrough curve (BTC) in Figure 3-1. BTCs for

quinoline (0.005 M CaCI2) in a sterile and inoculated (B53 and 3N3A isolates)

columns are also shown in Figure 3-1. BTCs for 45Ca and naphthalene (0.05 M

CaCl2) in sterile and inoculated B53 columns are shown in Figure 3-2 and 3-3,

respectively. The PFBA BTCs for all soil columns were symmetrical and sigmoidal

in shape, which suggests the absence of physical nonequilibrium (Brusseau and

Rao, 1989b), and P > 98 is indicative of minimal hydrodynamic dispersion.

Quinoline and naphthalene sorption was reduced in inoculated soil columns

(Figures 3-1 and 3-3). 45Ca sorption (Figure 3-2) was not reduced in the B53

inoculated soil columns. The shift in the 45Ca BTC in the two bacterial-inoculated

soil columns (BQ11 and BQ112) and the sterile column (B) resulted from

differences in the bulk densities (p) and volumetric water contents (0) of the

various columns (Table 3-1). Therefore, direct comparison of R for different

columns is misleading. The impact of bacteria on sorption and transport of

quinoline, 45Ca, and naphthalene was assessed by comparing the Kf values in

sterile and inoculated columns. The Kf values verified that sorption of quinoline

and naphthalene was reduced in inoculated columns, whereas 45Ca sorption was

not significantly different.

The following results are from a series of experiments that were conducted

to deduce the causes of reduced quinoline and naphthalene sorption. Experiments






1 0.2


* 0

- m




* Sterile (B)

* 3N3A (BQ6)

E B53 (BQ9)




Pore Volumes (p)

Figure 3-1.

Measured BTCs for PFBA (0) in a sterile column and
for Quinoline in a sterile (0), 3N3A inoculated (*), and
B53 inoculated (D) soil column. Column designations
are given in parenthesis corresponding to Table 3-1.


E] El

c. O

0 *
*0 0
0 +
OJ 4!
$ O 0
. .. ,* ^ oo .

5 10 1
Pore Volumes (p)

Figure 3-2.

Measured BTCs for 45Ca in sterile (*) and B53
inoculated (0 and *) soil columns. Column
designations are given in parenthesis corresponding to
Table 3-1.





* Sterile (B)
o B53(BQ11)
* B53(BQ112)



0 0 0
0 0

0.6 o
S0.4 o .

0.2 Sterile (B)
"0 0 B53(BQ11)
06 d j I I -----I

0 5 10 15
Pore Volumes (p)

Figure 3-3. Measured BTCs for Naphthalene in a sterile (0) and a
B53 inoculated (o) soil column. Column designations
are given in parenthesis corresponding to Table 3-1.

Table 3-1. Column parameters and Kf values for quinoline, naphthalene, and
45Ca in sterile and inoculated Norborne soil columns.

CaCI2 p 6 Kf

Column ID mol/L pH g/cm3 cm3/cm3 Quinoline Naphthalene 45Ca

Sterile, BQ5 0.005 7.0 1.48 0.44 3.11 --- 10.0
Sterile, B 0.05 7.0 1.54 0.42 3.11 0.946 1.17
B53, BQ9 0.005 6.8 1.46 0.45 1.39 ..
B53, BQ11 0.05 6.6 1.44 0.41 1.05 0.555 1.05
B53, BQ112 0.05 6.7 1.44 0.46 -- 1.06
3N3A, BQ6 0.005 6.9 1.39 0.46 2.42 ..

focused on distinguishing between the processes that may influence

contaminant sorption and transport including altered water flow resulting from

pore blockage, biofacilitated contaminant transport, and/or altered sorption

capacity of soil.

Pore Blockage

Pore blockage or straining of bacteria was investigated by measuring

BTCs for a nonadsorbed tracer (PFBA) once a day for 7 days following

bacterial inoculation. Variations in pore volume determinations or asymmetrical

BTCs would indicate changes in physical characteristics of the column. In all

cases, the BTCs measured for PFBA were symmetrical (indicative of no

changes in hydrodynamic characteristics) and the pore volume determined by