phosphate sorption in soils and reference oxides under
TRANSCRIPT
University of Massachusetts Amherst University of Massachusetts Amherst
ScholarWorks@UMass Amherst ScholarWorks@UMass Amherst
Doctoral Dissertations 1896 - February 2014
1-1-1989
Phosphate sorption in soils and reference oxides under varying Phosphate sorption in soils and reference oxides under varying
pH and redox conditions. pH and redox conditions.
Nadim Khouri University of Massachusetts Amherst
Follow this and additional works at: https://scholarworks.umass.edu/dissertations_1
Recommended Citation Recommended Citation Khouri, Nadim, "Phosphate sorption in soils and reference oxides under varying pH and redox conditions." (1989). Doctoral Dissertations 1896 - February 2014. 6076. https://scholarworks.umass.edu/dissertations_1/6076
This Open Access Dissertation is brought to you for free and open access by ScholarWorks@UMass Amherst. It has been accepted for inclusion in Doctoral Dissertations 1896 - February 2014 by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact [email protected].
PHOSPHATE SORPTION IN SOILS AND REFERENCE
OXIDES UNDER VARYING pH AND REDOX CONDITIONS
A Dissertation Presented
by
NADIM KHOURI
Submitted to the Graduate School of the University of Massachusetts in partial fulfillment of the
requirements for the degree of DOCTOR OF PHILOSOPHY
May 1989
Department of Plant and Soil Sciences
c Copyright by Nadim Khouri 1989
All Rights Reserved
PHOSPHATE SORPTION IN SOILS AND REFERENCE
OXIDES UNDER VARYING pH AND REDOX CONDITIONS
A Dissertation Presented
by
NADIM KHOURI
Approved as to style and content by:
j
Peter L. M. Veneman, Chairperson of Committee
tc/ ■fytc--
Michael S. Switzenbaum, Member
Baker, Department Head
lent of Plant and Soil Sciences
Dedication
This dissertation is dedicated
to the memory of my parents,
Mary Debbas Khouri
and Joseph Khouri.
ACKNOWLEDGMENTS
I would like to thank Dr. Peter Veneman for his trust, support,
and friendship throughout the course of my studies at the University
of Massachusetts and for suggesting the research area for this
dissertation.
I am thankful to Drs. John Baker, Allen Barker, Daniel Hillel,
and Michael Switzenbaum for their guidance as members of the
dissertation committee. Collectively, they were extremely patient in
attempting to keep my work focused and my objectives clear.
Individually, I feel very privileged to have shared their friendship
and science.
I feel privileged to have been associated with the Soil
Morphology group. I thank Steve Bodine for the daily discussions we
had and for guiding me through the soils characteristics data used in
Chapters 4 and 5; he has had a major role in generating this basic
data. Judy Bartos artfully performed metal analyses reported in
Chapter 3 and Mohammad Mahinakbarzadeh ran quite a few of the P
sorption isotherms of Chapter 4. Bill Donnelly's help in typing parts
of this dissertation came at a crucial time. Also, to Ron Lavigne, Deb
Kozlowski, Mike Reed and David Lindbo I express my gratitude for the
moments we had in Rm 18.
I am grateful to the following persons who advised and assisted
me during the construction of the experimental setup to control redox
potentials of oxide suspensions: Sami Haddad, Keith Shimeld, and Drs.
David Curran, Les Whitney, and Haim Gunner.
v
I would also like to thank Mrs. Lucy Yin and Mr. Lou Raboin for
conducting the electron microscope-related work of my research. Ed
Ostrowski and Ed Weist ran the "BET" analysis, and the Geology
Department allowed me to use their XRD equipment for the
characterization of metal oxides.
I thank Carol Cumps, Fred Schulten, and Martha Robinson for
repeatedly devoting time, energy and quite a bit of imagination in
addressing my concerns and those of my family. We always felt welcome
at the Office of Foreign Students and Scholars.
In Amherst, I feel privileged to have met some wonderful people.
In particular, I would like to thank Dr. and Mrs. Mack Drake, Val
Veneman, Rev. and Mrs. Philip Ward, and the staff of the Hampshire-
Franklin Day Care Center their help and encouragement made this
dissertation possible.
At the UNDP/World Bank program in Washington I am grateful for
the understanding of my supervisors and colleagues who have allowed me
to pursue a double-life of sorts for the last year. I am particularly
grateful to Carl Bartone for his confidence in me and to Saul
Arlosoroff and Fred Wright for their support.
My daughters, Lina and Maya, will soon realize how essential
their role was in this dissertation. Finally, I would like to express
my gratitude to my wife, Hoda, for her love. I feel that it is from
our companionship that I derived the will to produce this
dissertation, and the strength to accomplish whatever good comes out
of my life.
vi
ABSTRACT
PHOSPHATE SORPTION IN SOILS AND REFERENCE
OXIDES UNDER VARYING pH AND REDOX CONDITIONS
MAY 1989
NADIM KHOURI, B.S., AMERICAN UNIVERSITY OF BEIRUT
M.S., AMERICAN UNIVERSITY OF BEIRUT
Ph.D., UNIVERSITY OF MASSACHUSETTS
Directed by: Professor Peter L. M. Veneman
Despite numerous studies of phosphate (P) retention by soils
there is no unified methodology for the determination of P sorption
capacity (PSC). This dissertation establishes a framework for the use
of chemical models to evaluate PSC obtained through different methods
under varying environmental conditions. Chemical modeling too often is
an exercise in numerical fitting of thermodynamic data to reactions
that do not necessarily apply to field conditions. This study assessed
the PSC of two hydroxides and binary mixtures of these hydroxides
under different redox conditions. Subsequently, the PSC in selected
Massachusetts soils was studied to evaluate the significance of Fe and
A1 in true soils. Finally, the significance of P sorption in the
disposal of wastewater on land was quantified by determination of the
number of acres necessary to treat a certain amount of wastewater and
to ascertain the relative role of PSC determination techniques in this
land limiting constituent analysis procedure.
Two oxyhydroxides, goethite (a-FeOOH) and gibbsite (/-A^O^) were
synthesized and physically and chemically characterized. Phosphorus
vii
sorption capacities for the goethite and gibbsite were approximately 100
and 70 mmol kg~^, respectively, with maximum PSC values at pH 3 to 4.
The sorption additivity hypothesis, stating that each particle in
a mixed system reacts as in a monocomponent system, was validated for
the binary mixture of goethite and gibbsite at an initial P in
solution of 200 mmol m and at different pH values.
The applicability of mathematical and chemical models to land
application system is dependent on the determination of the maximum
capacity of the system under varying natural conditions. Iron is
affected by the oxidation-reduction potential; reducing conditions
solubilize Fe and possibly release sorbed P into solution. However, In
some instances, researchers have reported increases in P sorption
under reducing conditions. The speciation of Al theoretically should
not be affected by changes in soil redox.
To establish the relative contribution of Al and Fe to PSC, the
latter was evaluated under controlled redox conditions in the
laboratory. Special reactor vessels were constructed, filled with
simulated wastewater, and the redox potential microbially maintained
at ranges of +500 to -50 mV, representing oxidizing to reducing
conditions.
A substantial increase in P sorption was observed in reducing
environments containing 1:1 mixtures of goethite and gibbsite. This
increase was due entirely to Fe11 */Fe^ transformations. Compared to
the initial monocomponent systems, reoxidation of previously reduced
minerals increased the PSC but not to the extent of the continuously
reduced suspensions. The principal mechanism of P sorption in the
viii
reduced reactors could not be determined but was attributed to surface
reactions in the solid phase. Gibbsite surfaces may act as catalysts
to facilitate the formation of reduced iron minerals, possibly
vivianite Fe^CPO^).8^0. The research results indicated the
significance of PSC analysis for long term disposal and its
implications for the possible movememt of reduced colloids to the
groundwater.
In the present study, the elements that could be included in
such models are identified by correlating the PSC of representative
Massachusetts soils to selected soil characteristics. Five PSC indices
(Langmuir, Freundlich, Bache-Williams, Interpolation, and Equilibrium)
were derived. PSC was significantly correlated to Fe, Al, organic C,
and pH. Pedotransfer regression equations to predict P sorption in
soils were developed with Al and Fe as regressors.
ix
TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS. v
ABSTRACT. vii
LIST OF TABLES. xii
LIST OF FIGURES.xiii
Chapter
1. INTRODUCTION. 1
2. BACKGROUND. 7
Introduction. 7
The Phosphate Controversy. 7
The Langmuir Equation: An Empirical Model . 8
Empirical vs Mechanistic Models. 12
Phosphorus Retaining Factors in Soils . 13
The Use of Reference Oxide Minerals . 13
Research Needs. 14
3. PHOSPHORUS SORPTION ON REFERENCE OXIDES:
pH AND REDOX EFFECTS. 16
Introduction. 16
Phosphorus Retention Mechanism. 17
Phosphate Sorption by Goethite. 18
Phosphate Sorption by Gibbsite. 19
Phosphate Sorption by Mixtures of Oxides. 20
Redox Effects on P sorption. 22
Materials and Methods . 24
Reference Oxides . 24
Sorption Isotherms. 25
pH Effects. 26
Chemical Characterization. 26
Crystallographic Characterization. 26
Sorption by Mixture of Oxides. 27
Redox Effects. 28
Experimental Apparatus. 28
Experimental Procedures. 32
Phosphate Isotherms. 33
Microscopic Analyses. 33
x
Results and Discussion. 34
Optical and Chemical Characterization
of the Oxides. 34
Phosphate Sorption by Monocomponent Systems. 40
Sorption Isotherms of Binary Mixtures. 43
pH Effects. 43
Redox Effects. 46
Phosphorus Sorption in Oxidized Systems. 48
Phosphorus Sorption in Reduced Systems. 50
Conclusions. 55
4. PHOSPHORUS SORPTION CAPACITY IN ACID SOILS. 58
Introduction. 58
Materials and Methods. 60
Soil Sampling and Characterization. 60
Phosphorus Sorption Capacity. 61
Statistical Analysis. 62
Results and Discussion. 63
Conclusions. 71
5. ROLE OF PHOSPHORUS SORPTION CAPACITY IN
SUITABILITY ASSESSMENT FOR WASTE DISPOSAL. 72
Introduction. 72
Materials and Methods. 74
Wastewater Characteristics. 74
Soils. 77
Phosphorus Sorption Capacity. 77
Assessment of Soil Assimilative Capacity. 81
Results and Discussion. 82
Conclus ions. 89
6. DISCUSSION AND CONCLUSIONS. 90
APPENDIX. 93
REDOX POTENTIAL MEASUREMENT
A TOOL FOR SOIL INVESTIGATIONS. 93
REFERENCES. 108
xi
LIST OF TABLES
Table 3.1. Demonstration of the independence of specific edge
area from the dimension of crystal thickness (D) in
gibbsite, assuming a perfectly hexagonal crystal. 39
Table 3.2. Maximum phosphate sorption capacity of the two
oxides derived by five methods: indirect surface
measurement (BET), geometric calculations (TEM),
batch equilibration (Pexd), Langmuir (PL), and
interpolation (Pj). 42
Table 4.1. Range in properties of 57 surface and subsurface
horizons in 23 representative Massachusetts soils.... 65
Table 4.2. Correlation matrix of selected soil characteristics
and PSC indices from 23 Massachusetts soils. 66
Table 4.3. Selected results of stepwise regression for
different PSC indices and various soil properties.
In each model, all variables are significant
at the 0.15 level. 68
Table 4.4. Summary of statistical parameters for regression
models with Fe^ and Al^, or Fep and
Alp as independent variables. 70
Table 5.1. Representative wastewater characteristics for raw
municipal sewage and secondary effluent. 76
Table 5.2. Range in properties of selected surface and
subsurface horizons of two Massachusetts soil series. 78
Table 5.3. Phosphate sorption capacity of two soils using
different estimation methods. PE y and Pg g refer
to PSC values based on final soil’solution
concentrations of 7 and 8 mg L~ , respectively. 80
Table 5.4. Land limiting constituent analysis for application
of 1,000 nr d_i of raw municipal wastewater
on two representative New England soil series. 83
3 -1 Table 5.5. Acreage required for the disposal of 1,000 m d
of secondary effluent using land limiting constituent
analysis in two representative New England soil
series. 84
Table 5.6. Summary of potential land limiting constituents
as affected by PSC method, soil series, and type
of effluent.
xii
LIST OF FIGURES
Figure 1.1 The chemical modeling process. 3
Figure 3.1 Cross-section of the reactor used for controlling
redox potential in soil suspensions. 29
Figure 3.2 Cross-section of the salt bridge (a), and platinum
electrode (b). 30
Figure 3.3 Diagram of electronic switching mechanism used to
control the Eh in soil suspensions. 31
Figure 3.4 Titrimetric surface acidity characterization of
goethite. 35
Figure 3.5 Titrimetric surface acidity characterization of
gibbsite. 36
Figure 3.6 Summary of major characteristics of the two oxides.. 38
Figure 3.7 Phosphate sorption in goethite, gibbsite, and a 1:1
mixture (1.9 g L” ) of both oxides. Electrolyte:
0.1 M NaCl, pH: 5.5-6.5,
concentrations:
initial solution P
0-500 mmol m” .. 41
Figure 3.8 1:1 Phosphate sorption of goethite, gibbsite and
mixture (1.9 g L” ) at different pH levels;
electrolyte* 0.1 M NaCl, initial P concentration:
100 mmol m~ (53 mmol P Kg” solids). 44
Figure 3.9 Phosphate sorption of goethite, gibbsite and a 1:1
mixture (1.9 g L” ) at different pH levels.
Electrolyte* 0.1 M NaCl, initial P concentration:
200 mmol m” (106 mmol P Kg” solids).. 45
Figure 3.10 Phosphate sorption of goethite, gibbsite and a
1:1 mixture (1.9 g L"^) at different pH levels.
Electrolyte* 0.1 M NaCl, initial P concentration:
300 mmol m” (160 mmol P Kg”^ solids).
Figure 3.11 Phosphate sorption on 1:1 goethite and gibbsite
suspension in simulated wastewater. pH=6.5.
Figure 3.12 Phosphate sorption in goethite and gibbsite (1:1)
suspension in simulated wastewater. Initial
P concentration=300 mmol m .
Figure 3.13 Phosphate sorption in 1.9 g L ^ gibbsite
suspension in simulated wastewater. pH=6.5.
47
49
51
52
xiii
Figure 3.14 Phosphate sorption expressed on volume basis.
pH=6.5. 54
Figure 5.1 Land limiting constituent (LLC) concept in land
treatment design. 75
xiv
CHAPTER 1
INTRODUCTION
Sustainability is increasingly being considered as a prerequisite
for any decision-making in policies affecting all areas of human
activities. In environmental management, the hazards of improper waste
disposal have been recognized by decision-makers and the general
public alike (Time, 1989). According to Turner (1988), the sustainable
use of environmental resources entails what he called "Conservation
Rule One" that states:
"Maintenance of the regenerative capacity of renewable resources,
and avoidance of excessive pollution which could threaten
biospherical waste assimilation capacities and life support
systems."
The evaluation of waste assimilation capacity of a given system,
and at a given time, has to be compared to a reference or "threshold
value". Liverman et al. (1988) emphasized the importance of such
values for the evaluation of sustainability of all natural systems.
In the case of waste disposal on land, the threshold value would be
the point at which the land application system can no more "absorb"
one or more of the constituents of the waste material applied.
Models that are commonly used to describe and predict the
physical, chemical, and biological interactions occurring in the soil-
waste system (Iskandar, 1981) are needed to evaluate the assimilative
capacity for land application systems. This modeling should be aimed
at (i) identifying the major parameters and processes controlling the
soil-waste system, (ii) selecting suitable sites for waste disposal,
(iii) designing waste disposal systems that will promote maximum waste
1
attenuation, and (iv) designing appropriate monitoring systems to
ensure that the land application schemes are performing as planned.
Models that concern the transformation of pollutants in soils are of a
chemical nature and generally involve two levels of abstraction (Fig.
1.1). Firstly, the real-world components relevant to the system are
identified (soil particles, solutes, etc.). Secondly, the chemical
components are abstracted into mathematical systems of equations
generally of a thermodynamic nature. The outcome of the mathematical
manipulations is generally a prediction of the chemical speciation
state, i.e., which valences of components will be dominant. The
results of the chemical models then can be interpreted and validated
against real-world experiments.
One comprehensive experiment rarely can be designed to include
all the steps in Fig.1.1. Nevertheless, the process can be used to
combine the results of individual chemical modeling studies and
communicate with other models such as water transport models.
Mathematical systems that handle chemical speciation issues require
very sophisticated levels of numerical capabilities (e.g. Westall,
1979; 1982). The models generally assume that a system is at
equilibrium and that reactions are governed by thermodynamics. Strict
application of thermodynamics to the soil solution usually is not
justified (Bohn and Bohn, 1986) and a more thorough knowledge of the
chemical reactants, processes, and products under natural conditions
could help refine, adapt, or correct the results of numerical
speciation modeling (Sposito, 1985).
2
Figure 1.1. The chemical modeling process.
3
Processes occurring at the solid/liquid interface tend to defy
direct thermodynamic interpretation (Stumm, 1986; Morel and Gschwend,
1987). Phosphorus is one of the more important pollutants whose
movement is strongly dependent on interactions with solids. It is
considered rather immobile once it is applied to land because of its
strong tendency to adsorb to soil particles. There is evidence,
however, that P is reaching the groundwater (Gschwend and Reynolds,
1987). Field evaluation of P movement in soils is costly and its
applicability often limited. Chemical laboratory systems simulating
the soil-P system can provide an insight into the type of possible
forms of P in natural systems. The overall objective of this research
was to study the fate of P in the soil system, in particular with
respect to the disposal of municipal wastewater on land.
This dissertation is centered on the abstraction phase depicted
in Fig.1.1, as it relates to the retention or sorption of P by soils.
The different chapters and appendix in this dissertation follow a
"scientific article" format and are, therefore, somewhat independent.
In each chapter the specific objectives of the experiments are
identified, accompanied by a review of the appropriate literature and
a description of the methodology addressing each objective.
In Chapter 2, an overview of the major issues pertaining to the
importance of non point sources of P pollution and soil assimilative
capacity is presented as a review of literature.
In Chapter 3, a system of oxide suspensions was used in the
laboratory under various environmental conditions to evaluate
phosphorus sorption capacity (PSC). Iron oxide (goethite; a-FeOOH) and
4
aluminum oxide (gibbsite; -A^O^) minerals were generated in the
laboratory, characterized, and tested for their capacity to retain P
in monodisperse systems under aerated conditions and a variable pH.
This procedure defined the behavior of each parameter in an inorganic,
aerated environment. In a second stage, the oxides were combined in
experiments to determine the extent of particle-particle interactions
and to assess the effect such interactions have on P sorption. The
principal objective was to verify whether or not the oxides behaved
the same in a combined system, as they did separately.
Further simulation of real-world conditions was achieved through
the study of binary mixtures of Fe and Al oxides under various redox
conditions. In enclosed reactors, low redox potentials were obtained
by microbial activity, whereas positive redox potentials were attained
by manipulating the amount of air available in the system. The
objective of this experiment was to verify that Al oxides were
practically unaffected by reducing conditions, and that reduced forms
of Fe caused significant changes in the P sorption capacity of the
system. A review of applicable oxidation-reduction reactions in soils
and the problems inherent to the thermodynamic consideration of redox
reactions in natural systems is presented in the Appendix.
In Chapter 4, the major soil components that are responsible for
P sorption were identified through a correlation study aimed at
verifying the importance of Fe and Al for P sorption in Massachusetts
soils. The capacity of a soil to retain P was evaluated by a number of
methods that can be used for site evaluation purposes and land
application system design.
5
In Chapter 5, the role of P-soil assimilative capacity was
assessed in the disposal of various types of municipal wastewater. The
particular role of PSC was evaluated and its effect on land
application system design was studied. A comprehensive discussion of
the research results and a summary of its conslusions is provided in
Chapter 6.
The research presented in the following chapters is an attempt to
quantify P sorption capacity of reference oxides and representative
Massachusetts soils by using empirical as well as mechanistic
approaches. The elements required to improve the long-term prediction
of soil assimilative capacity were identified and evaluated in this
study. The experiments presented in the following chapters had the
following overall objectives:
a. to evaluate the effect of varying pH and redox conditions on P
sorption by reference oxides;
b. to develop statistical models of P retention in soil based on
empirically derived PSC parameters;
c. to determine the effect of soil PSC evaluation on the design
of land application systems.
6
CHAPTER 2
BACKGROUND
Introduction
Current efforts in pollution control legislation are aimed at
increasing the role of the land as recipient of waste materials. By
aiming at the "elimination of discharge of pollutants" in water
systems, the Clean Water Act (PL 92-500) has affected land application
by embracing the ecological approach to waste disposal (Noss, 1984).
Land is not only the final depository of waste treatment residues, but
also contributes to the actual treatment of wastewaters (Westman,
1972).
The waste assimilative capacity of a soil is a measure of the
ability of the soil to retain and possibly transform wastewater
constituents (Settergren and Tennyson, 1979). The capacity of soils to
act as "living filters" depends on soil environmental conditions that
govern the physical, microbial, and chemical processes by which
contaminants are transported, sorbed, or degraded in the soil
(Norstadt et al., 1977). Phosphate retention is of particular interest
because wastewaters, in general, are rich in phosphates that can
promote eutrophication in receiving waters (Vollenweider, 1968).
The Phosphate Controversy
Point sources as well as non-point sources contribute P loading
to receiving water bodies. Reduction of this effect can be achieved by
improved point-source control. In a study of the lower Great Lakes,
Switzenbaum and co-workers (1980) estimated that between 11 and 14 %
P loading reduction can be achieved by treating municipal plant
effluents. They also found that 24 to 47 % of the total P loading was
from non-point sources other than direct runoff or atmospheric
precipitation. While it is clear that P seeping through the soil
contributes to this significant non-point fraction, it is difficult to
quantify this contribution.
A survey of current literature shows that there is no consensus
regarding the threat of P percolation through soils. The spectrum of
opinions ranges from a virtually unlimited capacity to retain P
(Canter and Knox, 1985, p.75) to the belief that groundwater seepage
is a significant source of P to surface water bodies (Belanger et al.,
1984). Hook (1983) reviewed both sides of the P controversy and
concluded that field investigations of P movement are difficult to
interpret because of "inherent problems" in sampling and analysis of
soils and soil water, and because of the complexity of P chemistry in
the soils.
While efforts are being made continuously to elucidate P
behavior in soil, several methods for the determination of PSC are
being used currently - although major aspects of P chemistry are still
unknown. Phosphate retention capacity and P release in the soil
solution presently are estimated by one of the models described below.
The Langmuir Equation: An Empirical Model
Any equation that is used for the evaluation P retention capacity
8
of soils should answer the following "real world" questions:
a) If an effluent with a certain concentration of pollutant is
applied to a soil, how much of the contaminant will be
retained by the soil particles?
b) What is the maximum amount of pollutant that can be retained
by the soil particles?
Langmuir originally developed his adsorption model to represent the
retention of gases on solid surfaces. It was only later that this
equation was applied to the adsorption of solutes to soil particles
(Olsen and Watanabe, 1957). The following equation (Langmuir, 1918)
has been used more than any other model to describe and predict the
two parameters identified above:
q = abc/(l+ac) [1]
where, q — amount of sorbed P per unit of soil
a — constant
b — maximum amount of P that can be sorbed per unit of soil
c — concentration of solute, at equilibrium with q.
Experimentally, parameters c and q can be evaluated by
equilibrating soil samples in solution containing different amounts of
P and measuring the amount of P remaining in solution after
equilibrium is attained (Nair et al., 1984). The experimental
procedure has to be performed at constant temperature, which is
probably the reason why the plot of q vs. c is called an "isotherm".
Theoretically, and following equation 1, the isotherm is a smooth
concave curve with a horizontal asymptote equal to the maximum
adsorption value b. This can be demonstrated by considering a modified
form of equation 1:
q - b/[l+(l/ac)] [2]
It is clear that (1/ac) approaches zero, and q approaches b, as c (the
concentration of P in solution) increases. The limit of q when c tends
9
towards infinity is thus b. In other words, b is the maximum sorption
capacity. Veith and Sposito (1977) reported another relationship that
can be derived from equation 1 between q and q/c:
q - -(1/a)(q/c) + b [3]
The term q/c is known as the distribution coefficient. A more
widely used linear transformation of the Langmuir equation is:
c/q — (l/b)c + 1/ab [4]
Equation 4 has been traditionally used to determine maximum sorption
capacity in soils (Olsen and Watanabe, 1957).
The Langmuir equation and its derived forms are based on the
following assumptions (Mansell and Selim, 1981):
1. There is a constant energy of sorption for each contact
surface (no solute-solute reaction, and same intensity of surface
bond).
2. Sorption occurs at localized sites.
3. Maximum sorption corresponds to a complete layer that is one
molecule thick.
These assumptions are similar to the ones advanced by Langmuir
(1918) for the representation of gas-surface interactions. Harter and
Baker (1977) reviewed the major difficulties in transposing these
assumptions to the solute-surface system. A major difference is the
nature of the phenomenon described. The solute system involves the
exchange of a sorbed entity with one or more molecules of solute,
whereas no such exchange is required in the gaseous system. In
addition, in case of high solute concentration, physical bonds could
link solutes to one another in multilayer adsorption (Larsen, 1967)
which contradicts one of the basic assumptions cited. Another
assumption that would not be justified, in general, is the uniqueness
10
of type of bond involved. Soils have more than one type of site for
ion adsorption, leading to a range (not only one) of bonding energies
(Holford et al.f 1974). Finally, it is virtually impossible to
differentiate between adsorption of solutes and surface precipitation,
especially as the concentration of solute increases (Veith and
Sposito, 1977).
Upon testing the Langmuir model with soil samples data many
researchers reported that they did not obtain the linear relationship
of equation 4. Those who did (e.g. Olsen and Watanabe, 1957)
attributed the close fit of their results to the Langmuir equation to
the fact that they had used low concentrations of P. In general, the
results obtained by sorption studies followed a curvilinear
relationship when equation 4 was used. Nevertheless, the Langmuir
equation is generally considered suitable for the empirical
description of P retention in soils and for the interpretation of
analytical data (Harter, 1984).
Numerous variations of the simple Langmuir equation have been
suggested. They can be grouped under the form:
n
q - 2 (at b£ c±)/(l+&± cj_) [5]
i Equation 5 represents the juxtaposition of n simple Langmuir terms.
Values for n of up to n—6 have been reportred to model P sorption by
soils (Goldberg and Sposito, 1985). Holford et al. (1974) obtained a
satisfactory fit for a model where n=2 with soil experimental data.
They interpreted this as indicating the presence of two levels of
bonding energies. However, Sposito (1982) considered the terms a^ and
b^ mere curve-fitting parameters without mechanistic significance.
11
Furthermore, he demonstrated the "Interpolation Theorem":
"Any sorption reaction for which the distribution coefficient is
a finite, decreasing function of the amount sorbed, q, and
extrapolates to zero at a finite value of q, can be represented
mathematically by a two-surface Langmuir equation."
Most soils would have a typical two-layer Langmuir shape with q
linearly approaching zero at increasing values of q/c and q linearly
approaching b (the maximum sorption) at lower values of q/c. The
latter method of maximum sorption capacity determination is most
convenient because it removes the forced linearization of the simple-
Langmuir model and uses the general applicabiity of the two-layer
model.
Empirical vs. Mechanistic Models
In the previous section, the Langmuir equation was presented as
one of the most widely used empirical models of P sorption. Other
empirical models have been reviewed and compared elsewhere (e.g.
Travis and Etnier, 1981; Ratkowsky, 1986). One way of using the
results obtained on a given set of soils to another set is through the
development of statistical relationships between soil components and
the parameters of the empirical models. Another approach is to
develop mechanistic models that have, as parameters, the particular
soil components and characteristics that are effective in P sorption
related within specific chemical reactions representing the sorption
processes that are thought to occur in the soil.
An effective model incorporates the factors that are believed
pertinent to the problem at hand (Hillel, 1977). In the case of P
retention, it is mainly chemical modeling that considers P retaining
factors while mathematical models only describe the retention effect
12
without explaining it (Sposito, 1985). The identification and
quantification of soil properties responsible for P retention will
increase the applicability of empirical models as well as define the
parameters of mechanistic levels.
Phosphorus Retaining Factors in Soils
In any soil system the factors that affect P retention can be
grouped into: (i) the amount of P-retaining components, and (ii)
environmental factors that control the behavior of the P-retaining
components. In non-calcareous soils, Fe and Al are the most effective
such components as was demonstrated in numerous studies (e.g. Syers
and Iskandar, 1981; Singh and Tabatabai, 1977; Laverdiere and Karam,
1984). This has led to the development of land evaluation criteria for
waste-P disposal that are exclusively based on the levels of Al and Fe
in the soil (Breeuwsma et al., 1986).
An appropriate model of P retention in acid soils should thus
include Fe and Al oxides and the environmental conditions that could
affect their behavior. Soil pH and redox conditions have been shown
to significantly affect P retention (e.g. Holford and Patrick, 1979).
Quantifying this effect has been achieved more successfuly for
pH (e.g. Barrow, 1984) than for oxidation-reduction conditions.
The Use of Reference Oxide Minerals
Recent progress in the understanding of P retention can be
attributed largely to the study of aqueous synthetic mineral systems.
Oxides of Fe and Al generally are present in the soil in very low
13
concentrations and therefore can be much more effectively studied in
synthetic suspensions (Dixon et al., 1983). The relevance of such
laboratory studies to actual field conditions naturally is a crucial
consideration.
Goldberg and Sposito (1984a; b) determined that there was
significant correlation between the results obtained by different
researchers working with synthetic oxides to retention data obtained
on a variety of non-calcareous soils. Most of the chemical speciation
models used in the field of environmental planning are based on
thermodynamic data obtained from reactions of pure chemical products
in controlled experiments (Sposito, 1985).
The gap between laboratory studies and in-situ research can be
reduced through the inclusion of a greater number of relevant factors
in laboratory experiments or through the improvement of field
techniques in order to identify the various contributing processes.
The research presented herein is an attempt at the former approach.
Research Needs
In this Chapter, it is argued that the evaluation of PSC models
is hindered by the fact that the chemistry of P sorption is not well
understood. Another obstacle is the difficulty to include appropriate
soil environmental factors.
Oxide-P laboratory studies have improved our knowledge of P
behavior in soil. Such studies, however, should be extended to include
mixtures of well defined adsorbents (Hingston, 1981). In practice,
such studies should result in improved quantification of the retention
14
process in order to be used in transport models (Sposito, 1985) and
site selection for land application of wastes (Veneman, 1982; Kleiss
and Hoover, 1986).
15
CHAPTER 3
PHOSPHORUS SORPTION ON REFERENCE OXIDES:
pH AND REDOX EFFECTS
Introduction
Many areas in environmental research involve the determination of
macroscopic processes using parameters microscopic in nature. For the
evaluation of pollutant retention by soils, two macroscopic approaches
are used commonly. The first is the partition coefficient between
solution and solid components (Lyman, 1982). The second approach is
the use of sorption isotherm data (Tofflemire and Chen, 1976). While
both evaluations are based on specific assumptions concerning single-
particle behavior (e.g. energy of adsorption, maximum monolayer ionic
coverage, particle charge), these microscopic parameters are
quantified by the use of macroscopic, indirect techniques of an
empirical nature (Harter and Baker, 1977).
In practice, sorption coefficients are often used as mere curve -
fitting parameters (in pollution transport models, for example),
precluding any measurement -direct or indirect- of the sorption
phenomenon itself (Jury, 1983). An assessment of sorption processes
based on independently measurable parameters that relate to accepted
mechanistic sorption models (Westall and Hohl, 1980) will improve
significantly the validity of environmental models (Sposito, 1985).
Modeling of P retention in acid soils has been studied
extensively and it was shown that Fe and A1 hydroxides can be used as
16
reference models that mimic the sorption properties of soil (Goldberg
and Sposito, 1984a; b).
Phosphorus Retention Mechanism
Phosphate ions adsorb to and desorb from the oxide surfaces
through ligand exchange (Stumm et al., 1980). This adsorption
mechanism involves the release of ligand hydroxyl ions (0H~) from
metal ions (S) in an oxide mineral, and its replacement with another
ligand, in this case phosphate ions. The reactions are as follows:
S0H(s) + H+ - S0H+2(s) [1]
S0H(s) - S0'(s) + H+(aq) [2]
S0H(s) + H3P04(aq) = SH2P04(s) + »2° [3]
S0H(s) + H3P04(s) - SHp0-4(s) + H2° + H+(aq) [4]
S0“ (s) H- H3P04(aq) = spo"4(s) + H20 + 2H+(aq) [5]
In general (Goldberg and Sposito, 1985):
aS0H(s)+HbP04(aq)b'3+cH+(aq) “ SaHcP04(s)+bH20(s)+<a-b>0H‘(aq) t6)
where b-3 is the degree of protonation of phosphate ions. Each one of
equations [1] to [5] has an equilibrium constant Ka(^s of the form (Morel
and Gschwend, 1987):
^ads ^chem * ^coul
where Kchem — chemical (or intrinsic KSa(intr)) energy term
Kcoul “ coulombic (electrostatic) energy term
[7]
17
In equation [4] for example, the intrinsic constant KSa(intr) can be
written (Goldberg and Sposito, 1984a):
[SHP0-4][H+]
KSa(intr) “ --- exP<-F A1) («] [S0H][H3P04]
where F — Faraday constant (Cmol )
= surface potential (V)
R = molar gas constant (Jmol-^ K"^)
T = absolute temperature (K)
Reactions [1] and [2] are independent of the ion being adsorbed (i.e.
phosphates) and can thus characterize the particular surfaces being
investigated (Hohl et al., 1980).
Phosphate Sorption by Goethite
Goethite was used by Sigg and Stumm (1980) as a model metal
hydrous oxide to illustrate ligand exchange reactions involving
phosphate ions. They concluded that P ions formed "inner sphere
complexes" by binding directly to surface Fe atoms. Atkinson et al.
(1974) studied dry samples of goethite treated with P ions and
estimated that most of P is sorbed as binuclear complexes, to the
surface of goethite. In this type of complex, two atoms of surface Fe
are connected to two oxygen atoms and finally to one P ion. Based on
the total surface area, such an arrangement would lead to an
average coverage of 2.8 +0.7 pmol P m according to Borggaard (1983).
Assuming a monodentate complex between surface 0-H and P ions,
Anderson et al. (1985) estimated the average capacity of goethite for
P coverage at 4.5 /imol m ^. Goldberg (1983) reported values between
18
_ 2 2.5 and 7 /imol P m of goethite. The area occupied by one phosphate
2 2 molecule is 0.235 nm which leads to 7.1 /xmol Pm (Van Riemsdijk and
Lyklema, 1980a) . The capacity of 7 /imol P m~ would thus provide the
complete coverage of goethite with phosphate molecules. Such coverage
would only occur, at adsorption capacity, at the edge surface of
goethite, because it has the 0-H groups that contribute to P-
adsorption (Atkinson et al., 1974).
Phosphate Sorption by Gibbsite
Gibbsite crystals, like goethite, have their reactive 0-H groups
at the edge surface. This surface was found by Van Riemsdijk and
Lyklema (1980a) to represent about 14% of the total BET area and to
-2 -2 adsorb 7.1 /xmol Pm of edge area or 1.0 /imol Pm of BET area.
Based on total BET area, P adsorption capacity of gibbsite was found
to vary between 1.6 and 2.8 /imol P m’^ (Goldberg, 1983).
Van Riemsdijk and Lyklema (1980a; b) studied sorption reactions of
P on gibbsite surfaces at increasing P concentrations and found that
adsorption is the rate controlling P retention mechanism at low P
_ 3 concentrations. At P concentrations greater than 5000 mmol m , P
precipitated on the gibbsite surface (Van Riemsdijk and Lyklema,
1980a; b).
The various sorption models that combine chemical and coulombian
effects have been reviewed extensively (Hingston, 1981; Goldberg and
Sposito, 1985; Sposito, 1985; Morel et al., 1981; Vestall and Hohl,
1980; Dzombak, 1986; Westall, 1987). These models rely on a thorough
characterization of the solid surfaces, including size, reactivity,
and electrostatic parameters (James and Parks, 1982).
19
Empirical and mechanistic methods of sorption evaluation assume a
relationship between experimental, crystallographic, and chemical
representations of the phenomenon. This relationship should, however,
be tested for any set of adsorbent/ion system. Such a verification
would evaluate the sorption capacity of the surfaces and also
characterize these surfaces so that the effects of system
perturbations can be compared to reference systems.
Phosphate Sorption by Mixtures of Oxides
Phosphate-retaining constituents of the soil generally are
assumed to have a simple cumulative effect on the macroscopic sorption
capacity of the soil. This assumption underlies site suitability
assessment for waste disposal and the computation of equilibrium
constants for chemical speciation programs (Goldberg and Sposito,
1984a). No explicit validation of this assumption has been proposed
and recent reviews have emphasized the need to study ion retention by
mixtures of adsorbents (Hingston, 1981). Multicomponent studies can be
of particular interest in systems where changing environmental
conditions affect each one of these components differently.
Common models describing anion sorption, assume a constant
sorption energy for each contact surface (Langmuir equations), or
decreasing bonding energies with increasing site coverage by sorbed
ions (Freundlich equation; Mansell and Selim, 1981). Surface
heterogeneities are not considered by these models and are averaged
out for simplicity (Van Riemsdijk et al., 1986). Extension of such
models to include heterogeneous surfaces is possible by adapting
20
heterogeneous gas-solid models (Sircar, 1984) to the electrolyte-solid
system (Van Riemsdijk et al., 1986). While these models are empirical
in nature, they may lead to an improved description of multicomponent
systems. Parker and Jardine (1986) evaluated the effects of surface
heterogeneities on ion transport models by considering the solid phase
equivalent fractions of exchanging ions and assigning particular
thermodynamic parameters to each of the fractions. Their model
predicted the behavior of surfaces that have heterogeneous ionic
affinities on the same particle. Other models concentrate on
heterogeneity caused by the presence of mixtures of oxides (Honeyman,
1984). Assuming that all particles of each oxide in the system behave
in an average manner, the representation of the whole system of oxides
could be obtained by summing the separate thermodynamic parameters in
the system. The weighting factor of each oxide would be dependent on
its mass fraction in the system. Schwarz et al. (1984) applied this
concept to the evaluation of electrical properties of mixed oxide
suspensions. Using silica-alumina mixtures, they verified that the
zero point of charge of various mixtures was linearly affected by the
weight fraction of the pure components. Kuo and Yen (1988) suggested
replacing the weight fraction with a surface area fraction that would
represent the surface charge density of each oxide.
Whether weight fraction or surface area fraction is used, there
is an underlying assumption of non-interference between oxide
particles. As suggested by Honeyman (1984), such an assumption is not
necessarily justified in all cases. He enumerated the following
interactions that alter the non-interference assumption and,
consequently, the partial weight or surface area concept, (i)
21
coagulation or heterocoagulation that may block surface sites, (ii)
changes in the surface charge, and (iii) oxide dissolution. Honeyman
(1984) used the term "adsorptive additivity" to describe the non¬
interference concept. In the study reported in the following section,
the concept will be referenced as the "sorptive additivity" because of
the difficulty in identifying the phenomenon (adsorption vs.
precipitation) responsible for P retention by oxides.
The sorptive additivity concept should be tested on systems of
mixtures of oxides to aid in the extrapolation from monocomponent
systems to more complicated systems that are closer to the soil
conditions.
Redox Effects on P Sorption
The succession of wet-dry cycles in the soil enhances generally
its capacity to retain P. For example, Hill and Sawhney (1981)
reported a regeneration of soil retention sites below septic system
leaching fields receiving intermittent doses of wastewater. Bradley et
al. (1984) found that extractable P decreased in soils that are
waterlogged seasonally. In non-calcareous soils, this increase in P
retention was attributed to the formation of Al-P and Fe-P minerals
(Sah and Mikkelsen, 1986). It is expected that only in cases where
flooding causes an increase in soluble Fe, subsequent drying increases
the P retention capacity of soils (Holford and Patrick, 1981). While
details concerning the exact mechanisms involved in the regeneration
of sorption sites are still unknown, there is little disagreement in
the literature on the effect of wetting and drying cycles. The
effect of flooding on the availability of P, however, is not known.
22
Reducing soil conditions generally are associated with the
release of P into the soil solution (Ponnamperuma, 1972; Savant and
Ellis, 1964). This action is attributed to the reduction of ferric to
ferrous ion and the release of P held by ferric ions (Bradley et al.,
1984). Such conditions are believed to contribute to P contamination
of water sources by land disposal systems in areas with high water
tables (Gilliom and Patmont, 1983). In some cases, reducing conditions
have also been shown to enhance P retention by soils (Holford and
Patrick, 1979). This rather surprising effect is believed to be caused
by the formation, under suitable pH conditions, of a highly reactive
ferrous-ferric iron complex (Khalid et al., 1977).
Previous studies of redox-phosphate interaction have failed to
positively identify P-retaining compounds under reduced conditions
(Willett, 1983), and have considered generally only precipitation-
dissolution equilibrium reactions (Willett, 1985), and not sorption
reactions. Willett and Cunningham (1983) published a report on P-
reference oxide interactions under controlled pH and redox conditions.
They used H2 to decrease redox levels, a method which leads to rapid
redox changes not believed to represent real soil conditions. A
microbially mediated system could improve the simulation of such
conditions and their effect on soil chemical processes.
The objective of this research was to determine the effect of pH
and reduction-oxidation conditions on oxide-P systems in the presence of
chemical constituents of wastewater. This was attained by pursuing the
following specific objectives:
23
(i) characterize the monocomponent goethite and gibbsite
particles, and define their PSC using various methods;
(ii) test the sorption additivity concept in the goethite-
gibbsite system; and
(iii) evaluate the effect of varying pH and redox conditions
on mixtures of goethite and gibbsite oxides and assess
their PSC values.
Materials and Methods
Reference Oxides
Goethite (a-FeOOH) crystals were prepared by reacting a 0.4M
ferric nitrate (Fe(NO^) ^ • 9^0) solution with a NaOH solution at an
OH/Fe ratio of 1.5 for 50 hours (Atkinson et al., 1968). Sodium
hydroxide was then added to adjust the pH to 11.8. The precipitate was
maintained at 60°C for 4 days and dialyzed until no nitrate was
detected in the water bath.
Gibbsite (AI2O3.3H2O) crystals were prepared by adding 2.5M
NaOH solution to a 1.4M aluminum chloride (AICI3.6H2O) solution until
pH 4.6 was reached (Gastuche and Herbillon, 1962). The resulting
suspension was heated to 40° C for 2 hours then dialyzed against
distilled water for 23 days at 60° C with a daily change of distilled
water.
Goethite and gibbsite were identified by X-ray diffraction
(Schwertmann and Taylor, 1977; Hsu, 1977) and characterized by
electron microscopy and BET analysis to facilitate specific surface
24
area computations (McLaughlin et al., 1981). The oxides were stored as
aqueous suspensions throughout the experiment.
Sorption Isotherms
Five-milliliter aliquots of oxide suspensions containing about 75
mg of adsorbent were added to 45-ml polycarbonate centrifuge tubes
containing from 0 to 20 /imol Na2HP0^ before bringing the total volume
to 40 ml with 0.1M NaCL solution. Most tubes were run in duplicate.
Initial pH was adjusted to pH 5. The range of P concentrations
selected (0 to 300 mmol m after equilibration), was relatively low
and associated with adsorption with little possibility for
precipitation to occur especially at the relatively short
equilibration times used (Goldberg, 1983).
The centrifuge tubes were gently shaken for 24 hours on a
reciprocating shaker before centrifugation at 12,500 rpm for 30
minutes. Supernatant solutions were passed through a 0.20 /xm Metricel
filter, and the pH and P content (Murphy and Riley, 1962) were
measured in the filtrate. The amount of P sorbed was calculated by the
difference between the initial and final P in the centrifuge tube
filtrate.
Two parameters of maximum PSC (Pjj= Langmuir maximum, and P^=
Interpolation) were derived from the P sorption isotherm as follows:
(i) Pj^ was calculated from a linear form of the following
equation:
q = a PLc/(l + ac) [1]
where, q = moles of P sorbed per unit weight of soil, a = a constant,
c — moles of P in solution at equilibrium.
25
(ii) Pj was derived graphically by plotting the partition
coefficient (=q/c) against q and determining the intercept with the
q-axis (Sposito, 1982).
pH Effects
The same batch equilibration procedure as above was used to study
pH effects on P sorption. The initial P concentration was identical in
all centrifuge tubes (200 mmol m ) and pH levels were varied between
3 and 11 by the addition of IN HC1 or IN NaOH. The maximum sorption
value obtained in this experiment was considered the maximum
experimental PSC (PeXp)•
Chemical Characterization
The electrochemical properties of the oxides were determined by
acid-base titration and the procedure of Hohl et al. (1980) was
used to determine the intrinsic constants and the point of zero charge
(ZPC) of the oxides.
Crystallographic Characterization
The oxide crystals were observed by transmission electron
microscopy (TEM) to determine (i) the degree of homogeneity of the
crystals and (ii) the average dimensions of goethite and gibbsite
particles. Phosphate sorption capacities were estimated from the
distribution of reactive sites of the goethite (Hingston et al., 1972)
and gibbsite crystals (Van Riemsdijk and Lyklema, 1980a). The
thickness of the oxides was determined by gold shadowing in which gold
26
was sprayed at a 60° angle on oxides mounted on glass slides and the
resulting shadow was measured on electron micrographs. This thickness,
for both oxides, indicates the "edge area" (Van Riemsdijk and Lyklema,
1980a) where most of the 0H(H) groups available for P adsorption are
o located. By assuming an average area of 0.235 nm per mole of P
_ 2 adsorbed, equivalent to 7.1 /imol m (Van Riemsdijk and Lyklema,
1980a), an estimate of the maximum adsorption capacity of the edge
area can be calculated. In general, P adsorption capacity can thus be
expressed as:
q = 7.1 [ A r / (V X)] [9]
where q =
A =
r =
V =
X =
P adsorption capacity (mmol kg~ or /imol g" )
total surface area of oxide (mz)
fraction of total oxide surface area in edge area
total volume of oxides (nr)
density of the oxide (g m-'5)
Surface area was also determined from BET analysis (Everett and
Otterwill, 1969). The BET isotherm for N2 adsorption of each of these
oxides was obtained after drying of duplicate samples , and the BET
value was used to evaluate maximum PSC using equation 9 with the BET
surface value as parameter A (Everett and Otterwill, 1969).
Sorption by Mixture of Oxides
Phosphate sorption in the mixture of oxides was evaluated by
transferring 2.5-ml aliquots from each of the 15-g L stock
suspensions to 40-ml centrifuge tubes and running the same
equilibration, filtration, and P measurement procedures previously
presented.
27
The effect of pH on P sorption by the mixture of oxides was
evaluated at initial P solution concentrations of 100, 200, and 300
-3 mmol m . Adjustment of pH levels was done through the addition of a
few drops of HC1 or NaOH at concentrations of 0.2 or 1.0 M. Adsorbed P
was then determined as described previously.
Redox Effects
Experimental Apparatus. A system to control redox potential
conditions in soil suspensions was adapted from Patrick et al.
(1973). Experimental containers were constructed from clear acrylic
plastic (plexiglas) as shown in Fig.3.1.
Redox potential was measured continuously with a platinum
electrode (Fig.3.2) constructed according to Dirasian (1968) and
tested in a poised ferrous/ferric solution (Light, 1972). The Pt
electrodes were tested periodically during the course of the
experiments. Each electrode was connected to a voltmeter (Accumet
Model 805 MP, Fischer Scientific). The voltmeter had
microprocessor-based features that allowed the use of minimum and
maximum limit controls. If the redox potential in the reaction
container fell below a certain preset value, an electric signal
would activate a solenoid valve (Dynamo single solenoid 4-way valve.
Model 02110k00), which, through an electronic switching mechanism
admitted air into the reactor (Fig.3.3).
28
5
1. Pressure regulator 2. Gas inlet
3. Glass electrode 4. Salt bridge
5. Reference electrode
6. Platinum electrode
7. Brass o-seal connectors 8. Septum stopper
9. Stirring bar
10. Air trap
Figure 3.1. Cross-section of the reactor used for controlling redox
potential in soil suspensions.
29
KC1 agar
Semi-permeable membrane
___ Epoxy cement
t—-Acrylic tubing
_Copper wire
_Silver solder
_Epoxy cement
_Platinum wire
(a) (b)
Figure 3.2. Cross-section of salt bridge (a) and platinum
electrode (b).
30
To pH
meter
pin 14
Figure 3.3. Diagram of electronic switching mechanism used to control
the Eh in soil suspensions.
31
Experimental Procedures. Six reaction vessels were used to
control redox conditions. Each reactor contained the following
mixture:
2.0 L of simulated wastewater
2.0 g of reference oxides
50 mg of dextrose
1.0 g of straw
1.0 g soil.
The composition of simulated wastewater was adapted from Van
Riemsdijk et al. (1977) to represent approximately one tenth of the
strength of municipal wastewater. Per liter of suspension the
following amounts of reagent grade chemicals were added and allowed to
dissolve:
73 mg NaCl
30 mg CaS04.2H20
25 mg CaC^^I^O
20 mg NH4C1
14 mg MgCl2
8 mg NaF
5 mg KC1
4 mg KNO^
5 mg KH2P04
The ingredients in the reactor mixture, minus the oxides, were
allowed to equilibrate. Purpose of this preliminary step was to
promote microbial action and remove the largest amount of P possible
from solution. When P was no longer detected in solution the oxide
suspensions were introduced. Five reactors received 1:1 mixtures of
goethite and gibbsite, and one reactor received gibbsite only. The
equilibrium redox potential in the six containers was adjusted to
reach and maintain one of the following Eh values: +500, +300, 0, or -
50 mV. These values cover the range of redox potentials common in
well-drained to waterlogged soils (Patrick and Mahapatra, 1968). The
reactor vessels were placed on magnetic stirrers allowing continuous
mixing of the suspensions.
32
During the incubation period, the pH was monitored and manually
maintained at pH 6.5 by addition of IN NaOH or IN HC1. Monitoring of
the pH was required several times each day at the beginning of
incubation, but little adjustment was needed by the time redox
potentials stabilized. Holford and Patrick (1981) found the pH
stabilized after 8 days of incubation of suspensions where Eh was
controlled and the pH was not controlled.
Phosphate Isotherms. At the end of each incubation period
(approximately one week after the system stabilized at any given Eh
and pH levels), samples from the stirred suspensions were removed and
transferred to centrifuge bottles for P sorption evaluation. Samples
from the reactors in the reducing treatments were removed
anaerobically with a syringe and added to air-tight centrifuge
bottles containing solutions of P and filled with N gas (Holford and
Patrick, 1981).
Microscopic Analyses
Microprobe analysis was attempted on samples extracted from the
reactors (Lee and Kittrick, 1984; Sawhney, 1973) to determine the
location of P-sorption sites. The degree of resolution of the
instrument was, however, insufficient for the size of the oxides
involved.
To visually inspect the shape of the goethite and gibbsite
minerals, samples of the suspensions were mounted on glass slides and
observed under the transmission electron microscope. Samples from
33
reactors with a reducing environment were dried in a N2-atmosphere.
This was done as a precaution to prevent reoxidation although many
researchers who studied the micromorphology of anaerobic environments
did not provide such anaerobic conditions during sample preparation
(Lee and Kittrick, 1984; Jeanson, 1972).
Results and Discussion
Optical and Chemical Characterization of the Oxides
The goethite crystals exhibited a somewhat poorly defined, lath¬
like structure and tended to aggregate strongly. They appeared similar to
synthetic goethite prepared by Dixon et al. (1983). Gibbsite appeared
as pseudohexagonal, plate-like crystals, the typical form described
by Hsu (1977).
The electrochemical properties of the oxides were determined
based on a theoretical representation of P sorption onto the crystal
surface sites. It is generally accepted that P adsorption occurs
through ligand exchange in accordance to reactions [1] to [5].
According to that theory, the first two protonation-dissociation (or
acid-base) constants are independent of P, depend only on surface
characteristics of the oxides and can be determined by acid-base
titration.
For goethite, the intrinsic acid and base constants were 8.4 and
10.6 respectively, with the ZPC equal to the average of these two
values or 9.5 (Fig.3.4). The constants for gibbsite were equal to 3.2
for acidity, 6.1 for basicity, and 4.6 for ZPC (Fig.3.5).
34
PH
Figure 3.4. Titrimetric surface acidity characterization of goethite.
35
PH
Figure 3.5. Titrimetric surface acidity characterization of gibbsite.
36
These values are not in agreement with others reported in the
literature. Goldberg and Sposito (1984a), for example, reviewed Fe and
A1 oxide data and found that the ZPC of Fe and Al oxides averaged 8.0
and 8.2 respectively. Theoretically, such discrepancies may be
attributed to interactions of the background electrolyte (here,
NaNO^), with the oxide surface instead of just serving as a swamping
electrolyte (Sposito, 1981a). In practice, such interactions have not
been shown to occur between NaNO-j and the oxides of Fe and Al
(Goldberg, 1983).
The major properties of both oxides are summarized in Fig.3.6.
The principal characteristic for P sorption is the specific area, in
particular the specific edge area. The two methods used for the
evaluation of total specific area yielded significantly different
dimensions for goethite, with a 145 % difference between the BET and
the TEM measurements. The tendency of goethite to strongly aggregate
probably caused a decrease in BET surface compared to the specific
area of dispersed crystals. Also, the poor crystallinity of goethite
may have caused overestimation of the specific surface area.
In contrast to goethite, BET and TEM measurements in gibbsite
resulted in practically identical values. The specific area measured
43 m^ gwhich is in the range of values reported in the
literature (Goldberg, 1983).
The edge area was evaluated by gold shadowing and by TEM
calculations. It should be noted that the estimation of specific
surface area for gibbsite is independent of crystal thickness. Both
methods can be used to verify one another. This is illustrated in
Table 3.1 which shows that variable crystal thicknesses (D)
37
(a) Goethite.
72 nm_
5 U^y -3 V Density: 4.3 kg dm
Specific Surface Area: BET - 53 + 1
TEM - 130 + 20 m" g"1
ZPC - 9.5
(b) Gibbsite.
L - 225 + 17 nm; D - 23 + 3 nm
Density: 2.4 kg dm"^
o .1 Specific Surface Area: BET - 43 + 3 m g
TEM - 44 + 4 m2 g'1
Edge Area — Approximately 15 % of total area
ZPC - 4.6
Figure 3.6. Summary of major characteristics of the two oxides.
38
Table 3.1. Demonstration of the independence of specific edge
area from the dimension of crystal thickness (D) in
gibbsite, assuming a perfectly hexagonal crystal.
D L
Total
Specific
Area
Specific
Edge
Area
Edge Area
Percent of
Total Area
2 -1 2 -1 -nm- m g m g %
17 208 57.0 8.0 14
17 225 56.4 7.4 13
17 242 55.9 6.9 12
20 208 49.7 8.0 16
20 225 49.1 7.4 15
20 242 48.5 6.9 14
23 208 44.2 8.0 18
23 225 43.6 7.4 17
23 242 43.1 6.9 16
26 208 40.1 8.0 20
26 225 39.5 7.4 19
26 242 38.9 6.9 18
39
approximately yield the same specific edge area for a given size of
crystal represented by the dimension L (L refers to the diagonal
dimension of a hypothetical hexagon as shown in Fig.3.6b).
Phosphate Sorption by Monocomponent Systems
Of the two oxides, goethite had the highest P sorption capacity
(Fig.3.7). No maximum P sorption plateau was reached, due probably to
P precipitation on the surface of the oxides at the highest P
concentrations used. In order to avoid precipitation, the P adsorption
envelopes (sorption variation with pH) were run at initial
concentrations of 200 mmol m , and the maximum experimental sorption
capacity was determined.
Table 3.2 summarizes the results of the different techniques used
to evaluate P sorption capacity of the two oxides. In goethite the
BET, P_v_, and PT led to comparable results of around 110 mmol P kg"^.
The Pj and especially TEM resulted in higher estimates of PSC. It was
to be expected that would be greater than P^, because the isotherm
did not reach a plateau (see discussion in Chapter 2). As discussed
previously, the poor crystallinity of goethite may cause a much greater
effective surface area and higher PCS values.
Phosphorus sorption capacity values for gibbsite varied from
about 52 to 130 mmol P kg Results from the BET and TEM estimates
where very close because of the similar evaluation of the size of
gibbsite crystals. The difference between Pj^ and P^ is greater in
gibbsite than in goethite probably because of the relatively steeper
slope of the gibbsite isotherm at higher P concentrations (Fig.3.7).
40
PH
OS
PH
AT
E
SO
RB
ED
(MM
OL
/KG
)
Figure 3.7. Phosphate sorption in goethite, gibbsite, and a 1:1
mixture (1.9 g L"^) of both oxides. Electrolyte: 0.LM
NaCl, pH: 5.5- 6.5, initial solution P concentrations:
0-500 mmol m""^.
41
Table 3.2. Maximal phosphate sorption capacity of the two
oxides derived by five methods: indirect surface
measurement (BET), geometric calculations (TEM),
batch equilibration (P^p) , Langmuir (P^), and
interpolation (Pj).
_Maximum_P_Sorption_Capacity_
Oxide BET TEM Pexp PL Px
mmol kg~^
Goethite 113 + 2 277 +43 110 110 165
Gibbsite 52 + 6 53 + 4 75 77 130
42
Sorption Isotherms of Binary Mixtures
The isotherms obtained from the binary mixtures and each of the
monocomponent suspensions are shown on Fig.3.7. The isotherm of the
binary system is located clearly between the goethite and gibbsite
isotherms, with goethite having the highest P sorption capacity.
pH Effects
At the lowest initial P concentration of 100 mmol m , the curves
for goethite and the mixture of oxides were essentially identical.
(Fig.3.8). They showed a decrease in P sorption with increasing pH.
At pH<4.5 gibbsite dissolved but the sites available on the goethite
particles must have been sufficient for the sorption of P in the
binary suspensions. The apparent P-goethite affinity was also the
case at higher pH levels although goethite and gibbsite were present
in equal concentrations. Whether the shape of the curves in Fig.3.8
is an indication of the high affinity of P on goethite could not be
verified. From the fact that at pH>5 the three systems showed
parallel curves, one could conclude that gibbsite contributed also to
the sorption process.
At initial P concentrations of 200 mmol m , the mixture of
oxides had a retention capacity that closely followed the average
between the goethite and the gibbsite monocomponent sorption
capacities (Fig.3.9). The mixture envelope was generally closer to
the goethite evelope than to the gibbsite. However, the concentration
of P in solution was clearly beyond the retention capacity of goethite
particles at all pH levels.
43
SO
HB
EO
PH
OS
PH
AT
E
(MM
OL
/KG
l
130. 0-
120. O-
1 10. O-
100. 0-
90. 0-
80. O-
70. O-1 I
ODSORStiNTs
—■— GOETWITE mix TUBE
~ GI9BSITE
Figure 3.8. Phosphate sorption of goethite, gibbsite and a 1:1 mixture
(1.9 g L ) at different pH levels. Electrolyte: O.LH
NaCl initial P concentration: 100 mmol m"^ (53 mmol P kg" solids).
44
PH
OS
PH
AT
E
SO
RB
ED
(MM
OL
/KG
)
130. 0-
120. O-
1 10. 0-
100. 0-
90. 0-
80. O-
70. O-
60. 0-
SO. 0-
«0. 0-
30. 0-
20. 0-
10. 0-
. 0-
1. o
PCSOR0ENTS —— GOETWITE --«■* MIXTURE
GI93SITE
I II.'. 3. O S. 0 7. 0 9. 0 11.0 13. 0
PH
Figure 3.9. Phosphate sorption of goethite, gibbsite and a 1:1 mixture
(1.9 g L’^) at different pH levels. Electrolyte: 0.1M
NaCl initial P concentration: 200 mmol m (106 mmol • 1
P kg" solids).
45
_ 3 At the higher concentration of 300 mmol P m , the mixture
envelope did not follow the general trend of the monocomponent systems
(Fig.3.10). The monocomponent suspensions had a generally smooth
decrease of P retention with increasing pH levels whereas the mixed
suspension had two features that could not be explained adequately,
(i) around pH 4, the mixture of oxides retained more P than would be
expected by the additivity concept, and (ii) the mixture curve
fluctuated at pH>5 instead of smoothly and gradually decreasing (Fig.
3.10).
There is no specific Fe-Al-P complex reported in the literature
(Lindsay and Moreno, 1960; Stumm and Morgan, 1981) to explain the
apparent interaction between goethite and gibbsite. Samples of the P-
treated mixtures were observed by electron microscopy and did not show
any new deposits on the minerals. There was no difference in the
shape of the oxides, even at low pH levels. We can thus assume that
the only effect of very low pH (<4) is to dissolve gibbsite and
probably increase the potential of the remaining sorption sites to
retain P.
Redox Effects
Three distinct colors developed in response to different redox
conditions in the reactors containing goethite-gibbsite mixtures. The
reactors with oxidizing environments maintained at Eh values of +500
and +300 mV, had a brown color described by 10YR 6/6 (Munsell Color,
1975). The reactors with reducing environments, after about two weeks
46
SO
RB
ED
PH
OS
PH
AT
E
(MM
OL
/KG
l
Figure 3.10. Phosphate sorption of goethite, gibbsite and a 1:1 mixture
(1.9 g L'^) at different pH levels. Electrolyte: 0.1M
NaCl, initial P concentration: 300 mmol m(160 mmol P
kg' solids).
47
at 0 and -50 mV, developed a greenish color of 5Y 5/6, characteristic
of reduced iron in the soil. The reducing environments subsequently
were reoxidized by discontinuing the flow of N2 and allowing air to
enter. The color quickly turned from green to brick red (5YR 5/8). The
reactor containing a monocomponent suspension of gibbsite did not show
any color changes due to alteration in redox conditions.
Observed under the electron microscope, samples from the reducing
reactors showed very few goethite particles remaining,whereas the
shape and abundance of gibbsite particles were unchanged. Gibbsite
remained unchanged in samples from the oxidized reactors, whereas
goethite minerals were rare. Instead, smaller particles that may have
been Fe precipitate formed aggregates somewhat similar to goethite.
Examination of reoxidized samples by X-ray diffraction did not reveal
any Fe in crystalline form. It was surmised that after dissolution and
reoxidation, Fe precipitated as amorphous ferric oxides. To confirm
this interpretation, extracts from three of the reactors were analyzed
for dissolved Fe. The +500 mV treatment had 0.14 mg Fe L”^, upon
reduction to -50 mV, dissolved Fe reached 281 mg L ^. Finally, iron in
the reoxidized reactor dropped to 3 mg L ^. Aluminum varied much less
than Fe between treatments. The levels reached 0.47, 0.53, and 2.28 mg
Al L ^ for the +500 mV, the -50 mV, and the reoxidized reactor
respectively. The increase of Al in the reoxidized reactor can not be
explained.
Posphorus Sorption in Oxidized Systems
Phosphorus sorption (Fig.3.11) reached a maximum of 40 mmol kg
in both oxidized treatments of +300 and +500 mV. This level of PSC
48
140 Sorbed P (mmol/kg)
120 -
100 m
80 e
-50 mV
Figure 3.11. Phosphate sorption on 1:1 goethite and gibbsite
suspension in simulated wastewater. pH—6.5.
49
represents a reduction of more than 50% from the sorption capacity if
the oxides were in a "pure" environment (without biological activity
or wastewater constituents) as reported in Chapter 2. There, the Na
and Cl ions did not compete with phosphate for sorption sites. Of the
major compounds that compete with P for sorption sites, organic acids
produced by the microbial activity may be predominant (Goldberg, 1985;
Evans 1985; Sibanda and Young, 1986).
The effect of pH on P sorption (Fig.3.12) showed a trend of
decreasing sorption at higher pH values although sorption peaks occur
at much lower levels than in the case of suspensions in purely
inorganic systems (Fig.3.9.)
Gibbsite, as a monocomponent suspension, in an oxidizing
environment shows P sorption isotherm similar to the mixture of
goethite and gibbsite (Fig.3.13). This behavior is in agreement with
the findings by Goldberg and Sposito (1984a) that Fe and Al oxides have
a similar P-sorption capacity.
Phosphorus Sorption in Reduced Systems
Binary suspensions of goethite and gibbsite showed a significant
increase in P sorption capacity under reducing conditions (Fig.3.11).
This result was consistent with results reported by Holford and
Patrick (1979) who postulated that at pH 6.5, increased sorption was
due to the precipitaton of fresh Fe(0H)2* Another possibility was the
unintentional precipitation of amorphous Fe(0H)2 because of accidental
oxidation during the experimental procedures (Holford and Patrick,
1979).
50
Sorbed P (mmol/kg)
Figure 3.12. Phosphate sorption in goethite and gibbsite (1:1)
suspension in simulated wastewater.
Initial P concentration—300 anol ■
Sorbed P (mmol/kg)
Figure 3.13. Phosphate sorption in
simulated wastewater.
1.9 g I/1
pH—6.5.
gibbsite suspension in
52
In the system containing gibbsite as a monocomponent sorbent,
there was no apparent effect of redox on P sorption, especially at P
equilibrium concentrations below 200 mmol m (Fig.3.13). This
confirms that, generally, A1 is not significantly affected by
differences in redox potential (Lindsay, 1979). In binary systems the
increase in P sorption with reducing conditions, thus, has to be
explained by Fe and P interactions.
A series of experiments were conducted to study which mechanism
was responsible for the increased sorption. To establish whether the
increased P retention was due only to the oxides or to compounds in
solution, one of the reducing reactors (Eh=-50 mV) was sampled.
Minimizing contamination with O2, the extracted suspension was rapidly
filtered and the filtrate collected in a series of centrifuge tubes
containing increasing levels of P in solution. Because no solids were
_ 3 in the filtrate, the results of P retention were expressed in mmol m
of filtrate. The goethite/gibbsite suspension isotherm (Fig.3.13) was
converted to mmol Pm of suspension allowing comparison
with the filtrate results (Fig.3.14). Although some P appeared
retained by the filtrate, the amount was practically insignificant
compared to the retention capacity of the whole suspension. In
general, the share of sorbed P contributed by the filtrate was around
l/10th of the P sorbed by the whole suspension.
53
Sorbed P (mmol/cu.m.)
Figure 3.14. Phosphate sorption expressed on volume basis. pH
54
Accidental reoxidation of solids in the centrifuge tubes appeared
insignificant. Reoxidation would yield an amorphous Fe oxide of the
type obtained in the two reduced reactors that subsequently were
reoxidized. The isotherm of such products was determined (Fig.3.11).
Although the sorption capacity of the reoxidized mixed suspension is
clearly larger than the capacity of the crystalline oxides (Fig.3.11)
the isotherm of the reoxidized material reached a plateau at around
220 mmol P kg~^ whereas the reduced mixture had a sorption capacity
exceeding 700 mmol P kg~^.
One possible mechanism for the increase in P sorption under
reduced conditions could be the formation of colloidal precipitates of
Fe-P on the surface of gibbsite particles. Morel and Gshwend (1987)
suggested the occurrence of such colloids as an explanation to some of
what they call "strange" behavior of sorbates in natural waters. They
postulated that these colloids have solute-like properties but can
aggregate to form larger particles. The existence of such colloids
would have important applications in the area of non-point source
pollution and the movement of P in soils and ground water (Gschwend
and Reynolds, 1987).
Conclusions
From the study of the goethite and gibbsite monocomponent
systems the following conclusions were drawn:
i. The evaluation of sorption capacity based on particle size gave
comparable results when BET and TEM procedures were used on the
highly crystalline gibbsite minerals, whereas TEM values were
twice as large as BET in the poorly crystalline goethite.
55
ii. Of the empirical, isotherm-derived methods, the P exp and PL
parameters were in agreement for both oxides. These two
parameters, however, are dependent on the particular P
concentrations used during the equilibration phase. On the other
hand, the third empirical parameter, Pj, is found by
extrapolation and thus represents a "potential" that is
independent of the particular concentrations at which it was
evaluated.
iii. The two systems of adsorbents studied here were characterized by
measurable properties that, although not always consistent, can
be used to monitor the changes of surface-ion interactions under
variable environmental conditions.
The study of binary mixtures of the references oxides led to the
following conclusions:
i. For the adsorbents studied, the sorptive additivity concept was
validated for pH values of 6.0 and at levels of solution P that
exceed the sorption capacity of each of the oxides, and
ii. Non-adherence to the additivity concept occurred at initial P
concentrations of 300 mmol m , at pH values of 4.0 to 4.5 and
_ 3
6.5 to 7.0, and in initial P concentrations of 100 mmol m at
pH<4.5. These effects were possibly caused by particle-particle
interation P-precipitation at high solution P levels, and to
excess available sorption sites on goethite at low P levels.
Finally, from a study of mixtures of goethite-gibbsite
suspensions under varying redox potentials it was concluded that:
56
i. An apparatus to study soil chemical reactions under controlled pH
and Redox conditions was successfully adapted to the study of
reference oxides. By controlling redox through microbial activity
the behavior of particular soil components can be studied in
"living" environments that are often "non-thermodynamic" and
cannot be simulated in inorganic media,
ii. At redox potentials exceeding 300 mV, P sorption in 1:1 goethite
and gibbsite suspensions (1.9 g L”^) reached a maximum between 35
to 40 mmol kgA 1.9 g Lsuspension of gibbsite alone showed
a similar maximum P sorption value,
iii. In a reducing environment (Eh = -50 mV), the goethite and
gibbsite mixture exhibited a very sharp increase in P sorption
capacity due to dissolution of goethite and subsequent
precipitation, probably as an amorphous ferrous oxide,
iv. P sorption onto gibbsite appeared independent of the redox
potential.
v. Repeated wetting and drying cycles in soils may result in
degradation of goethite into less crystalline Fe solids that will
increase the P sorption capacity of the soil unless Fe is leached
out.
57
CHAPTER 4
PHOSPHORUS SORPTION CAPACITY IN ACID SOILS
Introduction
Land-use decisions may be facilitated by the application of soil
survey information. Current soil survey interpretations can be
improved by better correlations between available data and selected
soil properties and integrating these "pedotransfer" functions in
appropriate computer modules (Bouma, 1988). Phosphate retention data
are used increasingly to estimate the total phosphorus sorption
capacity (PSC) of a soil in relation to soil fertility (Laverdiere and
Karam, 1984; Ping and Michaelson, 1986) or environmental pollution
(Breeuwsma et al., 1986).
Studies undertaken with acid soils have shown that factors such
as clay content, organic matter, and extractable Fe, Al, and Ca
influenced PSC (White, 1981). No unifying formulations have been
derived, however, of the relationship between PSC and selected soil
properties if research results from different regions are combined
(Berkheiser et al., 1980). Consequently, PSC correlation studies have
been limited to the regional level (Tofflemire and Chen, 1976;
Polyzopoulos et al., 1983; Laverdiere and Karam, 1984). A regional
approach ensures uniformity of analytical methods and soil description
criteria and provides essential information to users of this data.
While correlation studies can lead to predictive regression equations,
58
no cause-effect relationships can be inferred from these type of
analyses.
The role of Fe and Al oxides in enhancing P retention in acid
soils has been well established (Goldberg and Sposito, 1984b). Simple
transformation equations, including extractable Fe and Al, have been
used to obtain PSC as a component in the evaluation of land for the
application of agricultural wastes (Breeuwsma et al., 1986). This
simplified procedure cannot be applied to all acid soils, since the P
retention capacity of some of these soils can be affected by other
parameters such as extractable Ca (Smillie et al., 1987). The
hypothesis that, in acid soils, the main agents of P sorption are Fe
and Al, is worth testing because of the wealth of information
available on these reference compounds in soil survey data bases and
the potential of quantitatively adapting results of Al and Fe oxide
experiments to soil simulation models.
One of the major difficulties in incorporating PSC values in
land suitability assessment procedures is the selection of a PSC index
that is simple, accurate, and field-applicable (Stuanes, 1982).
Sorption indices can be relative or absolute indicators of PSC.
Relative indices compare and rank various soils as to their relative
sorptive capacity, whereas an absolute index, in addition to being
suitable for making comparisons, can be used to calculate actual P
loading rates for wastewater or fertilizer applications.
The equilibrium sorption (Pgl Singh and Tabatabai, 1977;
Loganathan et al., 1987), Langmuir maximum retention (P^; Veith and
Sposito, 1977) and the interpolation maximum (Pjj Sposito, 1982) are
absolute indicators of P sorption in soil. The Freundlich index (PpJ
59
Bache and Morel, 1983) and the Bache-Williams phosphorus index (Pgyl
Williams, 1971) are examples of relative P sorption indicators. While
four of the PSC indices have been used extensively, the more recent Pj
method has received little attention.
The present study is aimed at evaluating the relationship between
PSC and selected soil properties in various acid Massachusetts soils
and to assess the effect of different PSC determination techniques on
the design of waste disposal systems. Specific objectives of this
study are to (i) compare PSC values of selected horizons in
representative Massachusetts soils by different measuring techniques,
(ii) correlate these PSC values with selected soil properties, and
(iii) develop regression equations to predict appropriate PSC values
from these soil parameters, including an evaluation of statistical
models that use Fe and A1 as unique regressors.
Materials and Methods
Soil Sampling and Characterization
Fifty-seven samples were obtained from surface and subsurface
horizons from 23 Massachusetts soil profiles located throughout the
state. Chemical and physical properties were determined on the
fraction of air-dried soil passing a 2-mm sieve. Particle-size
analysis was performed by wet-sieving to determine the sand fraction
and by measurement of the silt and clay fractions either by the
pipette method or the hydrometer procedure (Gee and Bauder, 1986).
When necessary, organic matter was removed by oxidation with H2O2. The
pH was measured in 1:1 (w:v) suspensions of soil in 0.01M CaC^.
% 60
Cations extracted by a 1M ammonium acetate solution at pH 7.0, were
determined by atomic absorption spectrometry (Soil Survey Staff,
1972). Cation exchange capacity (CEC) was obtained by distillation of
ammonium retained on the soil from pH 7.0 buffered ammonium acetate
(CECpHy) or by the sum of extractable cations (CECp^g ^ Survey
Staff, 1972). Extractable acidity was determined by a barium chloride-
triethanolamine method (Peech et al., 1962). Iron and A1 concen¬
trations were measured by atomic absorption analyses of extracts of
dithionite-citrate at pH 7 (Fe^, Al^) and pyrophosphate at pH 10
(Fep, Alpj Soil Survey Staff, 1972). Base saturation was calculated
from the measured values for extractable bases and CEC, and organic C
was determined by acid-dichromate digestion (Nelson and Sommers,
1986).
Phosphorus Sorption Capacity
The P-sorption procedure was adapted from Nair et al. (1984).
Duplicate samples were shaken for 24 hours in 0.01 M CaC^ solutions
containing K^PO^ at initial concentrations of 0 to 516 mmol m (0-16
mg L”^). Phosphorus in solution was determined after equilibration at
25 ± 1°C (Murphy and Riley, 1962). Final P at equilibrium subtracted
from the initial P value yielded the amount sorbed. The sorption
isotherms relating P sorbed to P in solution were used to calculate
the following PSC parameters:
(i) P£ was obtained from the sorption isotherm by determining the
level of P sorbed corresponding to the estimated P levels of the soil
solution. An equilibrium level of 216 mmol Pm (7 mg L ) was
61
assumed, which is the approximate P content of raw wastewater (U.S.
EPA, 1981).
(ii) P^ was calculated from a linear form of the following
equation:
q = a PLc/(l + ac) (1)
where, q = moles of P sorbed per unit weight of soil, a = a constant,
c = moles of P in solution at equilibrium.
(iii) Pj was derived graphically by plotting the partition
coefficient (=q/c) against q and determining the intercept with the
q-axis. (Sposito, 1982).
(iv) The Freundlich index (Pp) was determined by regression of a
linear, logarithmic transformation of equation 1:
log q = log (Pp) + r log (c) (2)
where Pp and r are emperical constants with 0<r<l.
(v) The value for the Bache-Williams index (Pgy) was derived from
the following equation:
PBW = / loS c' (3)
where qr = sorbed P (in mg/lOOg soil) after addition of 50 mmol P kg 3
soil, and c' = soluble P (in mmol m ) at equilibrium after addition of
50 mmol P kg^ soil.
Statistical Analysis
The data were evaluated statistically by deriving a correlation
matrix between principal soil characteristics and various PSC values
(SAS Institute, Inc., 1985). From this matrix, variables with the
highest correlation were used in a stepwise, multiple regression
62
by stepwise elimination or by inclusion of variables (Draper and
Smith, 1981).
Results and Discussion
Ranges in values for the soil properties and the various PSC
procedures are summarized in Table 4.1. The PSC methods were
correlated significantly (PC0.001) with one another (Table 4.2). In
this respect, PE exhibited the best overall correlation (r>0.75) with
the other PSC parameters. Holford (1979) discussed that a high degree
of correlation between P indices did not imply that they were
equivalent. In fact, the effectiveness of any PSC index depended on
its particular application such in as plant uptake or in P transport
(Holford, 1982).
A significant (P<0.001), positive correlation was found between
each of the PSC indices and Fe^, Al^, Fep, Alp, CEC, extractable
acidity, and in most cases organic C. Phosphate retention was
correlated negatively with pH and base saturation (P<0.01). No
significant correlation was found between various PSC values and
different particle size fractions. Extractable Ca was not correlated
significantly with Pj and P^, and only correlated weakly (P<0.10) with
Pp, Pgy. and Fg- Except for the PL value, all PSC indices had higher
correlation coefficients with extractable Al than with extractable Fe.
This Al-PSC correlation was reported also in previous studies
(Vijayachandran and Harter, 1975; Diggle and Bell, 1984; Laverdiere
and Karam, 1984; Ping and Michaelson, 1986).
For each PSC index, the correlation coefficients obtained with
the different Fe and Al extractants were comparable (Table 4.2).
Sodium dithionite removes amorphous and crystalline Fe compounds,
63
whereas pyrophosphate extracts the amorphous and organic forms of Fe
(Bascomb, 1968). These extractants are less discriminating in the case
64
Table 4.1. Range in properties of 57 surface and subsurface horizons
in 23 representative Massachusetts soils.
Soil
Property
Unit Range Mean Standard
Deviation
Sand % 6-99 53 23
Silt % 1-75 38 17
Clay % 0-50 10 10
Fen % 0.01-2.49 0.37 0.64
Fed % 0.13-4.73 1.18 0.97
A1P % 0.01-1.01 0.21 0.20
A1d % 0.02-1.15 0.26 0.22
Organic C % 0.1-11.4 1.4 2.2
Base sat. % 1-99 15 26.2
pH 1 3.0-6.6 4.6 0.8
extract. Ca cmol(+) kg 0.01-12.5 1.2 2.5
exchang. acid. cmol(+) kg' 1.1-50.3 13.7 14.4
CECpH 8.2
CECpH 7.0
cmol(+) kg” 1.1-50.4 15.0 14.2
cmol(+) kg”F 0.9-34.7 10.0 8.5
PSC parameters:
fe mmol kg”F 1.9-66.5 16.4 15.2
PL mmol kg” 4.5-69.6 16.7 13.6
Pt mmol kg 8.1-145.3 32.4 33.4
pp 8-1552 295 292
PBW 7-99 25 20
65
Table 4.2 Correlation matrix of selected soil characteristics
and PSC indices from 23 Massachusetts soils.
pH FeP Alp F°d Ald 0RG. C
0.62424
(0.0001)
-0.09768
(0.5232)
-0.19977
(0.1936)
0.14576
(0.2793)
-0.09868
(0.4693)
-0.03353
(0.8044)
-0.38851
(0.0092)
-0.37032
(0.0145)
-0.13488
(0.3262)
-0.27843
(0.0415)
-0.49601
(0.0001)
0.84900
(0.0001)
0.82908
(0.0001)
0.77666
(0.0001)
0.50059
(0.0005)
0.67503
(0.0001)
0.94367
(0.0001)
0.45965
(0.0017)
0.75145
(0.0001)
0.35099
(0.0074)
0.41155
(0.0016)
Pf Pi P| Pf PBW
-0.21017
(0.1348)
-0.12873
(0.3489)
-0.14539
(0.3772)
-0.25935
(0.0514)
-0.21389
(0.1101)
EXTC. Ca
-0.38997
(0.0051)
-0.44615
(0.0008)
-0.40379
(0.0132)
-0.35475
(0.0079)
-0.35439
(0.0079)
pH
0.77587
(0.0001)
0.64302
(0.0001)
0.47457
(0.0070
0.71277
(0.0001)
0.72209
(0.0001)
FeP
0.86167
(0.0001)
0.63825
(0.0001)
0.63424
(0.0002) 0.90377
(0.0001)
0.89926
(0.0001)
Alp
0.72994
(0.0001)
0.67847
(0.0001)
0.45995
(0.0032)
0.67255
(0.0001)
0.67733
(0.0001)
0.88158
(0.0001)
0.69705
(0.0001)
0.64502
(0.0001)
0.33750
(0.0001)
0.87752
(0.0001)
Ald
0.51805
(0.0001)
0.69779
(0.0001)
0.65562
(0.0001)
0.35700
(0.0064)
0.41361
(0.0014)
ORGANIC C
0.78741
(0.0001)
0.75523
(0.0001)
0.91824
(0.0001)
0.89493
(0.0001) P£
0.89778
(0.0001)
0.66499
(0.0001)
0.68638
(0.0001) pl
0.57564
(0.0001)
0.57959
(0.0001) P|
0.90504
(0.0001) pf
of A1 (McKeague et al., 1971). The amorphous forms of Fe and Al oxides
are associated clearly with P sorption (Johnson et al. , 1986). Stuanes
(1984) cautioned against using extractable Fe and Al to indirectly
derive PSC, chiefly because of the great variation in published
results. For land application of wastes, the usefulness of soil Fe and
Al correlations with PSC is complicated further by the possibility
that the Fe and Al that retain P are present already in the waste
applied (Chang et al., 1983).
Stepwise regression, including exchangeable Ca, organic C, pH,
Al^, Fe^, Al^, and Fep as independent variables, resulted in
prediction equations (Table 4.3). Validation of these models required
additional analytical and field work, which was beyond the scope of
the present study. All models were significant at the 0.1% level. The
equation for Pj had the least predictive ability. In addition to
possible interactions between soil parameters not considered in these
regression equations, the low correlation may result from inherent
inaccuracies in determining Pj. The estimation of this parameter was
facilitated by the use of concentrated P solutions to obtain the P-
isotherms, but at high concentrations other phenomena such as
precipitation may affect the results. Nair et al. (1984), therefore,
recommended that P-equilibrium studies be performed only at low P
levels. On the other hand, regression equations are empirical, and the
gain in additional data achieved by the use of higher P levels in the
equilibration solutions may outweigh the need to know which P-
retention mechanism predominates.
67
Table 4.3. Selected results of stepwise regression for different PSC
indices and various soil properties. In each model, all
variables are significant at the 0.15 level.
Model R2
PE ~~ 18.3 + 60.1Ald - 3.6pH 0.800***
PL = -18.4 + 8.9Fed + 5.3C + 4.8pH - 2.4Ca 0.847***
Pi = -8.6 + 147.4Ald - 27.5Fe 0.620***
Pv = 16.4 + 1183A1H - 22Ca P 0.848***
PBW = 3.27 + 76.02Ald + 2.54Fed - 1.27Ca 0.902***
***, significant at the 0.001 probability level.
Fed and Ald, dithionite-extractable Fe and Al.
Fep and Alp, pyrophosphate-extractable Fe and Al.
68
Dithionite-extractable A1 was a component in four of the five
regression equations (Table 4.3). Autocorrelated variables that
individually correlated highly with the PSC indices, often were not
retained in the stepwise regression procedure. Conversely, calcium,
which correlated very poorly with any of the indices, was retained in
three of the equations, evidence that Ca contributed to that part of
the variation not covered by Al-related parameters. In our study, the
contribution of Ca was, however, always correlated negatively with PSC
indicating a pH-type of effect rather than the positive correlation
reported by Smillie et al. (1987).
If extractable Fe and Al were used as independent variables to
o predict the PSC indices, a distinct difference between the R values
of the models of the various PSC indices was observed (Table 4.4).
Most of the two-variable models, including either Fe^ and Al^ or Fep
and Alp, were highly significant, but Fe and Al together covered only
38 to 50% of the variability in the P^ and Pj values. On the other
hand, Fe and Al covered 74 to 88% of the total variation in each of
the other three PSC methods. Relative PSC indices, Pp and Pgy, can be
predicted with a fairly high degree of confidence from dithionite
extractable Fe and Al. This result confirms earlier work by Burnham
and Lopez-Hernandez (1982) who found a good relationship between Pgy
and soil classification at the series level.
The correlation between extractable sesquioxides (Fe and Al) and
predictors of absolute PSC (Pg, Pg» an<^ Pj) was not as well defined as
that of the relative PSC predictors. The regression results (Tables
4.3 and 4.4) indicated that inclusion of a pH-related factor
significantly improved the predictive potential of the models.
69
Table 4.4. Summary of statistical parameters for regression models
with Fe^ and Al^, or Fep and Alp as independent
variables.
- Model considered -
PSC
Parameter
No. of
samples Fed j A1d
RZ
Fe„ + Al Pr2 P
39 0.766*** 0.743***
PL 41 0.501*** 0.490***
Pi 29 0.378** 0.424***
PF 43 0.874*** 0.818***
PBW 43 0.880*** 0.809***
-t— aa A A. AT » > ° ignificant at the 0.01 and 0.001 probability levels.
respectively.
and Al^, dithionite-extractable Fe and Al.
p and Alp, pyrophosphate-extractable Fe and Al.
70
Computer modeling to assess differences in assimilative capacity
between soils can be facilitated by inclusion of transfer functions
based on relative PSC values. Modeling of actual P sorption in soils
is complicated and should consider additional variables in the
transfer functions. These parameters need to be based on more
mechanistic models of P sorption.
Conclusions
Phosphorus sorption capacity was correlated highly with
dithionite and pyrophosphate extractable Fe and Al, pH, exchangeable
Ca, and organic C. Models including Fe and Al, exclusively, were
effective in explaining P sorption as measured by PE, Pp, and PBW
methods, and were less significant in the case of PL and Pj. The
latter was attributed, in part, to inaccuracies inherent to the
determination of these PSC parameters. All PSC parameters were
correlated significantly with one another. The results emphasize the
need to refine analytical techniques determining P^ and Pj indices,
because these parameters lead to estimates for absolute retention
capacities, compared to Pp and PBW> which give only a relative index
of soil P retention. More accurate determination of Pj is of special
interest because it leads to higher estimates of PSC than the other
absolute PSC indices, which underestimate the absolute PSC value.
71
CHAPTER 5
ROLE OF PHOSPHORUS SORPTION CAPACITY IN
SUITABILITY ASSESSMENT FOR WASTE DISPOSAL
Introduction
Land application of waste materials is a practice encouraged by
legislation to control environmental pollution (Massey, 1983).
Common waste application methods include overland flow, sludge
application, injection wells, evaporation ponds, sanitary landfills,
waste compost application, and on-site sewage disposal.
Except for contained systems, such as lined landfills, waste
application techniques rely on physical, chemical, and microbiological
processes in the soil to treat waste. Pollutant treatment processes
include volatilization, ion exchange, adsorption, precipitation,
oxidation, reduction, biodegradation, or uptake by living organisms.
Ideally, all these components should be considered in the assessment
of the assimilative capacity of a particular soil for waste disposal
(King, 1985).
The role of soil scientists in land application of wastes is to
assess the suitability of a particular site, especially the
determination of the soil assimilative capacity for various waste
constituents (Loehr, 1985). In practice, assimilative capacity is
evaluated by considering the factors that affect treatment efficiency.
Loehr and Overcash (1985) considered the following site specific
factors of importance: climate, topography, hydrogeology, vegetation,
and edaphic conditions. The latter include percolation rate, pH,
72
organic matter content, cation exchange capacity (CEC), and phosphorus
sorption capacity (PSC). These characteristics may be used to evaluate
the potential effect of general site conditions on the efficacy of
land treatment (Veneman, 1982; Steele et al., 1986). With respect to
soil chemical properties that contribute to pollutant attenuation,
organic C is used routinely to estimate soil assimilative capacity for
organic compounds (Lyman, 1982), whereas cation exchange capacity
often is indicative of heavy metal assimilative capacity (Houck et
al., 1987). To determine potential phosphorus sorption, however, a
myriad of methods are used to estimate the total assimilative capacity
(Sibbesen, 1981; Travis and Etnier, 1981).
Determination of the PSC generally has been concerned with
defining either a "maximum PSC" (Olsen and Watanabe, 1957; Stuanes,
1984; Van Riemsdijk and van der Linden, 1984) or an index parameter
that can be used to evaluate the relative PSC of soils (Bache and
Williams, 1971). Comparison of the relative PSC of two sites is
technically less challenging than the determination of an absolute PSC
value for each soil. Ultimately, however, absolute values of PSC are
needed in the integrated design of waste disposal systems.
Land limiting constituent (LLC) analysis is a convenient way of
integrating the two major technical considerations in land disposal,
waste characteristics and site assimilative capacity (Loehr and
Overcash, 1985). The concept of LLC analysis is outlined in Fig. 5.1.
The loading rate of each waste component is estimated and divided by
the projected waste assimilative capacity of the soil. The resulting
area-value is a measure of the total soil mass required to retain,
73
detoxify, or otherwise absorb waste components in an environmentally
sound fashion in accordance with the nature of each waste component.
The LLC procedure determines which waste consitituent has the highest
land area requirements for disposal. After an initial LLC analysis is
completed, two actions are possible depending on the nature of the
LLC: (i) further treatment of the effluent at the source in order to
reduce the value of LLC, or (ii) acquisition of additional land based
on LLC land area requirements.
In this study, a LLC analysis was used to evaluate whether or not
PSC was the limiting factor in land disposal of municipal wastewater
on acid soils. Specific objectives of the study were to: (i) compare
the results of different methods to determine PSC in two
representative Massachusetts soils, (ii) determine the effect of the
PSC determination method on potential waste loading rate, and (iii)
determine the sensitivity of system design for land disposal of
municipal wastewater to soil phosphorus loading capacity.
Materials and Methods
Wastewater Characteristics
Characteristics of raw municipal sewage and effluent from a
secondary treatment plant were used in this study. A daily flow of
1000 m^ was assumed, corresponding to the expected sewage flow for a
population of 3,000 people (U.S. EPA, 1981). Average wastewater
characteristics were obtained from the literature (Table 5.1). The
concentration of inorganic P reported (8mg L ^) is in the higher limit
of average P levels in raw wastewater. Annual loads of relevant
wastewater components were calculated by multiplying the elemental
74
WASTE CHARACTERISTICS
volume of flow
chemical composition
PH oxygen demand
solids content
I
I
I
SITE ASSIMILATIVE CAPACITY
kg ha"^ yr~^
I I
-| II V
AREA REQUIREMENT
ha
(LLC is waste consitutent
with largest area requirement)
I
I
I
WASTE LOADING
kg yr
I
I
SITE CHARACTERISTICS
climatic conditions
topography
geology, groundwater
soil properties
vegetation
LLC
REDUCTION
AT SOURCE
II II II V
ENGINEERING
DESIGN
LAND
ACQUISITON
Figure 5.1. Land limiting constituent (LLC) concept in land treatment
design (after Loehr and Overcash, 1985).
75
Table 5.1. Representative wastewater characteristics for raw municipal
sewage and secondary effluent (Source: Thomas and Law, 1977;
U.S. EPA, 1981).
PARAMETER RAW SEWAGE SECONDARY EFFLUENT E m-3 _ gm
Solids 450 320
COD 500 100
BOD 200 40
TOTAL N 40 20
INORG. P 8 7
B 1 0.8
Cd 0.02 0.004
Cr 0.20 0.12
Cu 0.10 0.06
Fe 1.0 0.5
Hg 0.01 0.005
Mn 0.14 0.13
Ni 0.08 0.05
Pb 0.50 0.25
Zn 0.40 0.12
76
composition by the total volume of the wastewater. Waste treatment was
assumed to occur through overland flow.
Soils
Two soils were selected for this study. The first was a Quonset
sandy loam (Typic Udipsamment, mixed, mesic) and the second a
Lanesboro silt loam (Typic Dystrochrept, coarse-loamy, mixed, mesic).
A summary of the major characteristics of these soils is in Table 5.2.
Phosphorus Sorption Capacity
The PSC of the major horizons of each soil was determined by
performing batch equilibrium sorption studies at 22° C to obtain P-
isotherms (Nair et al., 1984). Duplicate soil samples passing a 2-mm
sieve were equilibrated for 24 hours in Kl^PO^ solutions with initial
-1 - 3 concentrations ranging from 0 to 16 mg P L (0 - 516 mmol m ) .
The amount of P remaining in solution after equilibration was
determined (Murphy and Riley, 1962) and the amount of P sorbed was
calculated by subtracting the amount of final P in solution from the
initial P. The resulting P-sorption isotherms were used to evaluate
the following three parameters of "maximum" PSC:
(i) Equilibrium Sorption (Pg):
The P£ was obtained from the sorption isotherm by matching the
level of sorbed P with an assumed P-level expected in the soil
solution after application of the waste (Loehr et al., 1979). In this
simulation, we considered final concentrations of 7 and 8 mg P L
77
Table 5.2. Range in properties of selected surface and subsurface
horizons of two Massachusetts soil series.
Quonset Lanesboro
Soils:
Sand, % 51-86 37-69
Silt, % 12-39 25-53
Clay, % 1.5-8 5-12
PH , 4.3-4.9 3.1-4.7
CEC, cmol(+).kg” 3.4-18.0 3.5-48.8
Fed, % 0.52-0.83 0.67-4.73
A^, % 0.11-0.47 0.11-0.88
Organic C, % 0.2-3.1 0.3-11.4
PSC (mmol kg~l):
7.7-26.6 8.5-54.2
PL 6.7-39.0 10.1-69.6
pi 15.3-145.3 27.1-122.7
78
(ii) Langmuir-Maximum Retention (P^):
This index is the theoretical maximum adsorption capacity as
described in the equation:
q — aPLc/(l+ac) [1]
where, q = moles of P sorbed per unit weight of soil, a = constant,
c = moles of P in solution, at equilibrium. Values for P^ were
calculated from a linear transformation of Equation 1 (Olsen and
Watanabe, 1957).
(iii) Interpolation Maximum (Pj):
This method refers to the interpolation theorem proposed by
Sposito (1982). The maximum PSC was found by defining the partition
coefficient:
Kd = q / c [2]
and plotting Kd against q. Pj was then determined by extrapolating
to intercept the q-axis.
A weighted average of PSC (Tofflemire and Chen, 1976) for each
soil was computed by considering a one-meter reactive depth of soil
and applying the following equation:
n
t = (Sdj^ x qi) /100 [3]
i
where, t — average maximum PSC (g kg ^), i = horizon down to 1 m
depth, n = total number of horizons down to 1 m, d = thickness of
horizon i (cm), q = maximum PSC of horizon i (g kg ^). To transform
the results from g P kg ^ soil to kg P ha ^ an average soil bulk
density of 1.3 Mg m'3 was assumed. These calculations were repeated
for each of the three methods of estimating PSC (Table 5.3).
79
Table 5.3. Phosphate sorption capacity of two soils using different
estimation methods. P£ y and P£ g refer to PSC values
based on final soil solution concentrations of 7 and 8 mg
L~ , respectively.
PHOPHATE SORPTION PARAMETERS
SOIL HORIZON PE y PE g PL P-j-
- mg p kg'1 -
Quonset A 825 860 1207 4500
Bwl 700 760 963 3300
C 237 250 209 475
profile average 379 403 444 1386
Lanesboro A 1275 1375 2156 3750
Bwl 1680 1760 2087 3800
Cr2 263 294 313 840
profile average 615 660 717 1608
80
Assessment of Soil Assimilative Capacity
The following wastewater constituents were considered potentially
limiting and were included in the LLC analysis: total N, Cd, Cr, Cu,
Mn, Ni, Pb, and Zn, and inorganic-P (Table 5.1). The soil assimilative
capacity of the waste constituents included in the LLC analysis, was
evaluated using methods discussed hereafter.
To simulate the N-assimilative capacity, the maximum wastewater
loading rate (W) was calculated (Loehr et al., 1979):
W= [C+A(p-ET)-c'p]/(Y-A) [4]
where, C = plant N-uptake (about 150 kg yr"'*'), A = allowable N
concentration in percolating water (10 g m ), p = precipitation (1 m
yr-^), ET = evapotranspiration (0.6 m yr~^), Y = concentration of N in
-3 applied wastewater effluent (g m ), and c' = N concentration in
_ 3
precipitation p (0.5 g m ). Values for p and ET that are
representative of central Masschusetts were obtained from Yuretich et
al., (1986). The resulting W (in m yr-^) was used to calculate the
maximum N loading rate (kg ha ^ yr ^):
M = W x Y [5]
where, M = maximum allowable N loading. It was assumed that 35% of
applied N would be lost through denitrification (U.S. EPA, 1981). The
area of land required was determined by dividing the value for M by
the waste-N generation rate G (kg yr ^).
The soil assimilative capacity for heavy metals, in general, can
be obtained from laboratory column studies (Fuller and Warrick, 1985),
or by single (Tirsch and Baker, 1976) or repetitive (Amacher et al.,
1986) batch equilibrium experiments from which sorption isotherms can
be derived. In this simulation, we used maximum application levels of
81
heavy metals that in the field produced no toxicity in plants. These
levels have been termed the "USDA Guidelines" (Loehr et al., 1979).
The magnitude of the values generally are affected by the CEC or the
texture of the soil. Recommended maximum allowable levels (Overcash
and Pal, 1979) for soils with CEC values < 5 cmol (+) kg"1 (Quonset)
and 5 to 15 (Lanesboro), were used in the simulation (Tables 5.4 and
5.5).
Results and Discussion
Not included in the LLC analysis, but nevertheless of practical
importance were site-specific factors such as topography and land use,
soil hydraulic characteristics and soil pH (Fig. 5.1). Sites with a
high chemical retention capacity may have a low permeability that
could limit the volume of liquid wastes applied. A neutral, well-
buffered soil is often considered ideal for land application (Overcash
and Pal, 1979). The relatively low pH of most southern New England
soils could be a limiting factor for the disposal of wastes except
that the pH can be raised by co-disposal with alkaline waste
materials or by liming. In this study, it was assumed that the site-
specific factors were equally non-limiting in both sites or could be
corrected by practical management methods.
The LLC analysis indicated the importance of N in sewage as a
potential pollutant, because N was limiting in both soils (Table 5.4).
Secondary treatment significantly reduced the N content (Table 5.5),
thereby, lessening the area-limiting role of N. Overall, this
reduction decreased land requirements by more than 20% for the Quonset
soil and 55% for the Lanesboro soil. The assumption that N-
82
Table 5.4. Land limiting constituent analysis for application of
1,000 nr d~ of raw municipal wastewater on two
representative New England soil series.
Constituent
Loading
rate
kg yr_1
- Quonset
Site
assimilative
capacity
kg ha~^ yr"^
Area
required
ha
- Lanesboro -
Site Area
assimilative required
capacity
kg ha'l yr~^ ha
Total N 15,000 385 39 385 39
Inorganic P
7
29,000
105 28 172 17
PL 115 25 206 14
PI 360 8 418 7
Cd 7 0.2 35 0.4 18
Cr 70 20 3.5 40 2
Cu 40 5 8.0 10 4
Mn 50 15 3.3 20 2.5
Ni 30 2 15 4 7.5
Pb 180 20 9 40 4.5
Zn 150 10 15 20 7.5
83
Q 1 Table 5.5. Acreage required for the disposal of 1,000 m d of
secondary effluent using land limiting constituent
analysis in two representative New England soil series.
Quonset - Lanesboro
Constituent
Loading
rate
kg yr"1
Site
assimilative
capacity
kg ha”^ yr"^
Area
required
ha
Site
assimilative
capacity
kg ha-^ yr"^
Area
require*
ha
Total N 7,500 577 13 577 13
Inorganic P
7
2,600
99 26 160 16
PL 115 23 202 13
PI 360 7 418 6
Cd 5 0.2 25 0.4 12.5
Cr 40 20 2 40 1
Cu 20 5 4 10 2
Mn 50 15 3 20 2.5
Ni 20 2 10 4 5
Pb 90 20 4.5 40 2.5
Zn 35 10 3.5 20 2
84
assimilative capacity was independent of soil type, resulted in
identical land area requirements for both soils. As a result, in the
case of land disposal of raw sewage, where N was the LLC (Table 5.4),
no difference was found between soils in the amount of land required
(39 ha).
When components other than N were considered (Table 5.5), the land
requirement for the Lanesboro soil was 38% less than for the Quonset
series. The latter, because of lower CEC and PSC values, exhibited
less assimilative capacity than the Lanesboro. The relative role of
Cd in LLC analyses appeared dependent on the PSC evaluation technique
(Table 5.6). Cation exchange capacity based guidelines for heavy metal
application vary greatly among regions and states. More restrictive
guidelines for sludge application than the ones determined in this
study, result in lower estimated assimilative capacities and in larger
land areas.
Different methods of PSC evaluation led to varying area
requirements for P sorption. Theoretically, the equilibration PSC
method should be the only index to be used because, until now, it is
the only parameter that derives from the examination of the mechanism
of sorption. If it is accepted that a state of equilibrium is
established between sorbed and solution ions at any given
concentration of ions, then it should be impossible for the solid
particles to reach complete coverage with sorbed ions. Although this
might happen at very high concentrations of P. The use of P^ and Pj
as PSC indices assumes that even at relatively low P levels (7 and 8 g
m ^) total coverage of the particle may occur. Although this is known
85
Table 5.6. Summary of potential land limiting constituents as
affected by PSC method, soil series, and type of
effluent.
PSC
parameter
- Quonset - - Lanesboro -
Raw municipal Secondary Raw municipal Secondary
sewage effluent sewage effluent
PE
PL
PI
N>Cd P>Cd N P>Cd
N>Cd Cd>P N P>Cd>N
N>Cd Cd N Cd>N
86
to be theoretically false, there is an interest in using indices that
reflect field conditions, even if these indices are only empirical.
In the case of secondary effluent application to the Quonset
soil, the estimated maximum PSC values (Table 5.3) resulted in area
requirements of 26, 23 and 7 ha, representing a decrease of about 70%
in area requirement because of differences in PSC determination.
Table 5.6 shows that P was not limiting for raw waste applications.
For secondary effluent, however, P exhibited a land limiting role.
The Pg method resulted in P becoming the LLC, indicating the need for
land acquisition to provide adequate land area for treatment. If the
P^ maximum was considered, P remained the principal LLC in the
Lanesboro soil, whereas Cd was of primary concern in the Quonset soil,
because of its lower heavy metal retention. Phosphorus was not of any
LLC importance if the Pj method was used.
The lack of standard methodology for the evaluation of PSC in
soils has resulted in wastewater treatment designs where P is not
considered in the LLC analysis (e.g., Houck et al., 1987). Another
common approach is to consider that P content for the percolating soil
water is negligible and that only runoff P is of environmental
importance. Modeling of P loading through the use of the universal
soil loss equation has been successful in the evaluation of land
spreading of animal manure (Moore and Madison, 1985). The significance
of P loading through percolation may become more evident if large
amounts of waste are applied on land as is quite common in western
Europe. The extent of this seepage-P loading under various soil and
climatic conditions seems to be the major data lacking for the use of
PSC values in the design of land application systems.
87
The present study indicates that the PSC evaluation methods used
can have significant effect on the design of land application systems.
As to the applicability of these methods, Schweich and Sardin (1981)
found a correlation between batch equilibration experiments and column
studies. Interpretation of field experiments has generally been
difficult because of technical problems in the sampling of seepage
soil water (Uebler, 1984) and the uncertainties in extrapolating
laboratory data to the field.
Phosphorus generally is considered an immobile constituent (Loehr
and Overcash, 1985) that will accumulate in soil until the maximum PSC
is reached. Persistently high waste P removal rates have been detected
in the field as well in the laboratory. Nagpal (1985) in a column P-
loading study of almost 3 years, obtained 60 to 90% P removal at the
end of the experiment. Almost from the beginning, however, the P-
concentration in the aqueous extracts from the columns was 1 to 2 mg
L~^. Sorption processes in the soil environment apparently remained
significant but did not result in total P removal.
The contribution of seepage P on surface water quality is
unknown. Recent studies show that river water quality generally is
improving in the USA, whereas lakes and reservoirs are deteriorating
due to non-point source pollution (Humenik et al., 1987). Changes in
lake quality often are related to minute increases in P levels because
this element frequently is the growth limiting nutrient. In evaluating
the potential polluting effect of P, this element should not be
considered an immobile species, but rather a "somewhat mobile"
species. Perhaps a certain allowable seepage P load should be
88
considered as in the case of nitrate. This consideration will become
possible only when more information on the relative importance of
runoff and seepage P is gathered. If only runoff is significant, P
does not affect the design of land application systems through LLC
analysis. If dissolved P is significant, models of P movement through
soils could be adapted to field conditions in order to define a more
meaningful maximum PSC than the one generated by Langmuir-type
isotherms (Stuanes and Enfield, 1984).
Conclusions
i. In addition to facilitating engineering design of land
application systems, LLC analysis offers a suitable framework for
comparing methods to evaluate assimilative capacity of particular
soils.
ii. Phosphorus overloading can be significant when wastewaters are
disposed of through overland flow. The chemical characteristics
ot two representative southern New England soils indicate a
potential for P pollution due to seepage, in addition to runoff,
especially for secondary wastewater application,
iii. No routine PSC method is accepted commonly. Current methods
result in an underestimation of the soil retentive capacity.
Engineering design of waste application systems is affected
strongly by the particular PSC method used.
89
CHAPTER 6
DISCUSSION AND CONCLUSIONS
The detrimental effect of P in the environment is manifested
through its effect on surface water eutrophication mainly. While all
the mechanisms controlling eutrophication are not elucidated yet, it
is known that P is in almost all cases the limiting nutrient for
eutrophication to take place.
There are two accepted mechanisms of P entry in surface water
bodies: through runoff of P enriched soil particles, and through
internal enrichment by the release of sediment-P. A third route,
seepage of P through soil, is generally less accepted. In all three
routes the reaction of P at solid surfaces controls the release of P
in quantities that can be used by aquatic organisms that initiate
eutrophication.
The present research focused on elements that allow the
quantification and prediction of P interactions with solid surfaces
under varying pH and redox conditions. In the first experiment,
goethite and gibbsite were used as reference oxides under increasingly
complex conditions to simulate wastewater-P-solids effects. In
reducing environments a sharp increase in P retention capacity of
the mixture of oxides was observed. This increase was not observed in
monocomponent gibbsite systems. It is believed that this increase was
due to the dissolution of Fe (III) and subsequent precipitation of a
90
ferrous-P compound (possibly vivianite). Gibbsite might have played a
catalytic role to precipitate the Fe(II)-P compound on its surface.
In the second experiment, empirical parameters of PSC of selected
non-calcareous Massachusetts soils were determined and pedotransfer
functions were statistically derived. Extractable Fe and A1 were
essential regressors in the predictive equations for the five PSC
indexes that were studied. Based on the conclusions of the first
experiment, it would seem advisable to alter the relative importance
of Fe and A1 parameters in the empirical models for the long term
prediction of soil PSC.
Under conditions of systematic wastewater application Fe could be
mobilized and transfered to lower levels of the profile, or into the
groundwater. The PSC of the soil would thus be expected to decrease
with time. This decrease would not only be caused by saturation of
retention sites of the soil particles, but also due to Fe leaching
from the upper soil horizons. Furthermore, the Fe that is leached out
of the soil might have a PSC much higher than that of the soil. Such a
high PSC would mean that reduced Fe-P complexes could present a means
of P loading in surface waters through groundwater. Mobile forms of P
would influence the design of land application systems.
Traditionally, PSC of a soil has rarely been considered a
limiting factor in the design of municipal wastewater land application
systems. Where P was considered, it was generally by assuming that P
was immobile in a soil horizon until the PSC of this horizon became
saturated with P. This approach was used to compare PSC with other
factors that should enter in the design of the size of land
application systems. Depending on the particular PSC evaluation
91
method, P could become a limiting factor for the application of
effluents of secondary treatment plants. Other limiting factors were N
and Cd.
A conceptual change might be required to further reduce the
potential hazards from P in the aquatic environment. This change would
require labelling P as "somewhat mobile", assigning a limit
concentration of P in seepage water and then calculating the
permissible rate of application of wastewater after all the other
potential pollutants have been considered as well. Nitrate application
rates are currently being evaluated using this concept.
In conclusion, the work presented here established the potential
impact of waste-P application on soils and the effect of such an
impact on the design of a land application system. Such an approach
should increasingly be introduced in the design of management systems
to limit the impact of wastewater application on the environment.
92
APPENDIX
REDOX POTENTIAL MEASUREMENT
A TOOL FOR SOIL INVESTIGATIONS
Introduction
Oxidation-reduction reactions are characterized by electron
transfer. Such reactions involve inorganic, organic, and biochemical
reactants and take place in all types of natural systems (Bohn, 1971).
In the soil system, oxidation-reduction (redox) reactions affect soil
formation and classification (Collins and Buol, 1970), soil fertility
(Schwab and Lindsay, 1983a,b), and chemical transport (Holford and
Patrick, 1979). This appendix is a review of the principal methods of
redox potential measurement, their application in the various fields
of soil science, and their use in laboratory experiments involving the
control of redox and pH levels.
Rationale for Redox Potential Measurement
Redox Reactions
At equilibrium, any reaction involving the transfer of electrons
(e) can be considered the summation of two half reactions in the
following manner (Sposito, 1981b):
mAQX + nH+ + e = pA^ed + 9H2°(1) (reduction) [1]
sBred - tBQX + e (oxidation) [2]
mAox + sBred + nH+ - pA^ + tBQX + qH20 (redox) [3]
93
Where AQX and Are(^ are the oxidized and reduced states of a chemical
species A (same for B) and e is the electron in solution that is
transfered from B QX (electron donor) to AQX (electron acceptor).
Two parameters -electron activity, and electrode potential- have
been suggested for the measurement of the relative tendency of an
environment (generally a solution) to attract or release electrons.
These two parameters are briefly discussed in the following two
sections.
Electron Activity
Considering that the electron is a reagent in oxidation and
reduction equations, a parameter pe can be defined as follows:
pe = -log{e) [4]
where (e) is the electron "activity" at equilibrium, in either
equation [1] or [2]. The term activity is not used here in its strict
chemical sense because the electrons are not free in solution.
Nevertheless, such a representation is useful because it expresses the
"electron pressure" in terms that are conceptually similar to the
treatment of pH in acid-base equilibria (Stumm and Morgan, 1981).
Consequently, a high pe would indicate a low electron activity and
high oxidizing conditions (i.e. regularly higher tendency to take
electrons). Low pe values indicate that reducing conditions prevail.
Electrode Potential
Similarly to pH, pe is actually determined by the electrochemical
measurement of an Electromotive Force (Emf) in volts. Eh is the
94
electrode redox potential and represents the voltage of the redox
reaction when measured against the standard hydrogen electrode.
The electrode potential Eh represents in fact the free energy
change ( G), (products - reactants) of each half reaction in calories,
G of equation [1], for example, would be:
G = G° + RT In[{A}/{B}] [5]
where G° is the free energy change at unit activity and {A} and {B}
the activities of the oxidizing chemical and its reduced form
respectively. By converting equation [5] to electrical energy the
Nernst equation is obtained:
Eh = Eq + RT/nF ln[{A}/{B}] [6]
where EQ = potential when the redox couple is at unit activity, and F is
the constant of Faraday in heating units.
Standard Electrode Potential (EQ) values for naturally occurring
redox reactions can be found in the literature (Stumm and Morgan,
1981).
As mentioned in the beginning of this section, pe values can be
derived from Eh measurements:
pe - (F/2.3 RT) Eh [7]
at 25°C, equation [7] can be written:
pe = Eh/0.0591 [8]
Lindsay (1979) found that, in a given soil suspension, pe
generally decreases when pH is raised; the value of pe + pH remaining
the same. He suggested that pe + pH be used as a parameter to describe
the redox status of suspensions.
95
Equilibrium vs. Non-Equilibrium Systems
In the previous sections the redox potential of systems that are
in equilibrium has been discussed. According to Bohn (1968),
equilibrium is reached when "the net donation and acceptance of
electrons is zero". But soils and other surface natural systems are
rarely in equilibrium for the following reasons: (i) the heterogeneity
of soil components, (ii) the exchanges with the surroundings, (iii)
the temperature and pressure variations and (iv) the slow reaction
rates. These factors would preclude the application of equilibrium
thermodynamics to redox reactions in the soil.
In ground water chemistry, such a conclusion was reached by
Lindberg and Runnells (1984) who found no internal equilibrium between
the various redox couples in 600 ground water samples. The difference
between measured and calculated Eh reached 1000 mV in many cases.
Measured pH was determined through the use of a platinum electrode and
calculated Eh was derived from the use of sample ionic concentrations
in published equilibrium equations.
Equations combining all significant redox couples in a non-
equilibrated system have been presented by Bohn (1971). It has been
demonstrated, however, that in natural systems the "final" redox
potential is not determined by all the redox couples present, but only
by the ones of significant concentration. Some redox couples do not
affect solution Eh, even at high concentrations. Runnells et al.
(1987), for example, showed that the Se(VI)/Se(IV) dynamics in
solution were not affected by Eh.
The redox couples that are considered to affect the soil redox
potential mostly are chemical entities that derive from the following
96
elements: iron, manganese, sulfur, nitrogen, oxygen and hydrogen
(Bohn, 1971). Sposito (1981a) added carbon to this list and, in the
case of contaminated soils, mentioned arsenic, chromium, copper and
lead. These "redox-significant" elements can even be used as indirect
indicators of approximate Eh values (Stumm and Morgan, 1981).
Even if the quantification of redox potential is a difficult and
rather "inaccurate" exercise, thermodynamics can define the trends and
possibilities of redox couples in the soil system. Furthermore, Eh
measurement is the only suitable means to quantify redox conditions in
reducing environments (Bohn, 1971). This is discussed in the following
section.
Waterlogged vs. Well-Drained Soils
Of the main contributors to soil Eh mentioned in the previous
section, oxygen is the most significant element in systems that are
exposed to the atmosphere. Oxygen is one of the most powerful
oxidizing agents in natural systems. When a soil is waterlogged, the
submerged portion -except for a thin layer in contact with air- is
generally depleted of oxygen and goes from oxidizing conditions to
reducing conditions (Patrick and Mahapatra, 1968). Waterlogging would
thus greatly widen the range of expected Eh values. While aerated
soils usually have Eh values of 400 to 700 millivolts, waterlogged
soils may have an Eh of as low as -300 volts. The wide range of Eh
values, the high concentration of reduced components and the
difficulties in the measurement of oxygen content all contribute in
making Eh the preferred tool for measuring the oxidation reduction
97
status of waterlogged soils (Patrick and Mahapatra, 1968). Reddy and
Patrick (1983) noted that it is not enough to study the effect of
water logging on the reactivity and mobility of soil constituents but
that it was necessary to relate such results to the O2 content of
water.
In general. Eh measurements in well-drained soils have been
unsatisfactory for a number of experimental reasons as well as the
narrowness of redox potential range (Ponnamperuma, 1972). A more
fundamental reason to favor the credibility of Eh measurements in
flooded soils is related to the relative stability of microbial
dynamics in such closed systems. For well oxidizing environments,
other methods such as O2 uptake from solution can be used (McBride,
1987).
The Biological Factor
This general discussion on redox potential would be incomplete
without considering the role of microbial activity in redox reactions.
Most redox reactions that occur in natural systems need microbial
organisms as mediators (Stumm and Morgan, 1981). These organisms do
not actually oxidize or reduce a substance, they make redox reactions
-that are thermodynamically possible- actually occur.
A thermodynamically possible reaction is one where conditions
favor the formation of products that are more stable than reactants.
Microbes cannot affect the thermodynamics of redox reactions, only
their kinetics. This catalytic role involves the acceleration of the
redox reactions and the enhancement of coupling between reduction and
oxidation half reactions (Lovley, 1987). In this respect, a
98
thermodynamic description of redox reactions should be more applicable
in a closed biological system (such as a flooded soil) than in an
open, aerated system (Sposito, 1981b).
In general, any effect on redox due to waterlogging or other
changes in the soil environment is in fact due to a shift in microbial
population and activity (Patrick and Mahapatra, 1968).
Redox potential studies and applications
In this section, the principal methods that have been used for Eh
measurement will be reviewed. First, the in-situ redox studies and
their limited applications will be discussed; then, the applications of
Eh values in laboratory measurements will be covered in greater
detail, and the different methodologies employed for Eh and pH control
will be presented and discussed.
In-Situ Eh Measurements
In a review article, Pormamperuma (1972) acknowledged the
usefulness of Eh measurements only as "semi-quantitative" indicators
of oxidation-reduction status, with limited diagnostic value for
agronomic interpretation. He mentioned that, in the past, in-situ
redox potential measurements have mainly dealt with the following
areas of investigation: (i) study of drainage problems, (ii) Mn, P and
Si release during waterlogging, (iii) solubilization and deposition of
minerals.
More recent applications of in-situ Eh measurements include: (i)
diagnosis of an aquic moisture regime in profiles without low chroma
99
color (Vepraskas and Wilding, 1983), (ii) correlation of Eh values
with natural vegetation productivity along a catena (De Laune et.
al., 1983), and nutrient distribution in submerged soils (Pedrazzini,
1983).
In-situ redox potential is usually measured with a voltmeter
using a palatinum electrode and a reference electrode (Stumm and
Morgan, 1981). The installation and retrieval of the Pt electrode is
often problematic in flooded soils and permanently placed electrodes
have been found to decrease in accuracy with time (Bailey and
Beauchamp, 1971; Fox and Rolfe, 1982). Use of a KCl-salt bridge
facilitates collection of year-round Eh measurements in seasonally dry
soils. (Pickering and Veneman, 1984; Veneman and Pickering 1983).
In-situ quantification of redox potential is used cautiously by
soil investigators, mainly because of the difficulties in measurements
and the limited use of Eh values obtained. Fox and Rolfe (1982)
devised a method to transport soil samples under N2 atmospheric
conditions, to the laboratory where Eh measurements can be performed
more accurately.
Laboratory Redox Potential Studies
Laboratory redox investigations usually are aimed at confirming
or refuting the theoretical behavior of chemical entities in the soil
as depicted by thermodynamically-derived stability diagrams (Lindsay
et al., 1981). The first attempts to test "calculated" stability
diagrams were done in aqueous solutions with the exclusion of soil
particles. Under such conditions Collins and Buol (19/0) found that Mn
and Fe followed the predicted results satisfactorily when each
100
chemical was used alone, but not when a mixture of Fe and Mn was used,
indicating the complexity of chemical interactions that can be
expected in mixed systems.
Bohn (1968) reported that many of the earlier soil redox studies
were unsatisfactory because of measurement defects caused by the use
of inadequate electrodes. He tested the response of platinum, gold,
and graphite electrodes to different oxidation states and found that
platinum was the most suitable. Platinum electrodes are still the most
widely used electrodes for Eh measurement (Mueller et al., 1985).
In the following sections some of the most recent methods used
for laboratory Eh-pH investigations will be discussed, and their
principal findings will be presented.
Soil Suspension Studies
In these studies the soil particles are kept in suspension to get
uniform conditions in all the systems studied and to expose a maximum
particle surface area (Letey et al., 1981).
Patrick et al. (1973) devised an apparatus that allowed the
control of pH and Eh conditions in a soil suspension. The main
features of the apparatus included: platinum electrodes for Eh
monitoring, a pH electrode, and two meter relays for Eh and pH
control. Redox potential was controlled by air (to increase Eh) and
nitrogen (to obtain a low potential). The pH was adjusted by the
controlled addition of either 0.5 N HCL or 0.5 N NaOH. Solenoid valves
permitted the automatic control of Eh to pre-set values. Attainable Eh
values with this system were reported to vary between +600 and -250 mV
(Patrick et al., 1973).
101
The set-up described above has been used to study particle
interactions with mineral ions, organic compounds, and gaseous
compounds. Of the main redox-relevant chemicals cited previously, soil
Fe and Mn have been studied most. Gotoh and Patrick (1974) studied the
effect of Eh and pH on Fe species in soil suspensions. Redox potential
was increased by the supply of one of several mixtures of O2 and N2,
and decreased by the supply of glucose to increase microbial activity
(leading to more reducing conditions). Iron fractionation and analysis
methods under reducing conditions were presented and the results
suggested that Fe activity in a flooded soil was mainly controlled
by the solubility of ferric oxyhydroxide.
Patrick and Henderson (1980) compared Fe and Mn behavior in
suspension and soil free systems. They described a method for sample
collection and preservation in a reducing N2 atmosphere and found that
Mn solubility was dependent almost completely on pH with little Eh
effect, while Fe transformations were more Eh-dependent. They also
found that Fe reduction in soil suspension occured at higher Eh and pH
values than in soil-free systems, probably due to microbial activity
(Patrick and Henderson, 1980).
An apparatus with the same features as described previously was
used by Schwab and Lindsay (1983a). They added CO2 to the oxidizing
mixture in order to keep a constant CO2 content and control the pH of
the calcareous soil suspension used. They found that in calcareous
soils Fe4* * activity was controlled by FeCO^ (siderite) when pe + pH
is less than 8 and by Fe3(0H)g (ferric hydroxide) when pe + pH is
greater than 8 (Schwab and Lindsay, 1983a). As for Mn, they found that
102
Mn solubility in a calcareous soil was controlled by MnCO-^
(rhodocrosite) when pe + pH 16 (Schwab and Lindsay, 1983b). They
also used a system developed by Reddy et al. (1976) to study the
effect of Eh and pH on uptake of Fe and Mn by plants. Manganese uptake
was found to decrease at pH values below 7, probably because of
competition from Fe++ (Schwab and Lindsay, 1983a; b).
While research is still in progress to elucidate many of the
remaining unanswered questions concerning Fe (Savant and McClellan,
1987) and Mn behavior in soil, a number of soil chemical substances
have been studied under varying Eh-pH conditions.
Sims and Patrick (1978) studied the effect of redox on the
distribution of certain micronutrients. They found that Fe, Mn, Zn,
and Cu that get solubilized under reduced conditions, became complexed
by organic matter occurring in the controlled reduced environment. In
sulfur rich soils, heavy metals solubility was found to be enhanced
because of increased acidity resulting from reducing conditions.
Sorption and mobility of phosphate in an acid soil suspension
were investigated by Holford and Patrick (1979). At pH values between
5 and 8, phosphate release was found to increase with more reducing
conditions. Also, at any given redox potential value (from -150 to
+500) increasing pH caused decreasing P levels in the soil solution.
It was concluded that Eh and pH were affecting P availability by
influencing Fe phosphates solubility (Holford and Patrick Jr., 1979).
The exact effect of Eh-pH on soil P, however, is still not elucidated
and needs more investigation, especially in view of the existing need
to assign soil P sorption indexes for the rational use of soils
(Burnham and Lopez-Hernandez, 1982).
103
As to organic compounds, Gambrell et al. (1984) demonstrated that
reduced conditions in soil suspensions can enhance the degradation of
DDT and Permethrin.
The study of equilibrium between the different species of a soil
component often involves a gaseous phase. Hunt et al. (1980) modified
the system of Patrick et al. (1973) to collect and analyze gaseous
emanations from incubated soil suspensions. The system consisted of
two independent units, one aerobic and the other anaerobic. This study
demonstrated the interaction between aerobic and anaerobic processes
for the production of ethylene, an important step in the C-cycle (Hunt
et al., 1982).
Gaseous products involved in the soil N-cycle were studied by
Letey et al. (1981), who modified Eh and pH levels in incubated jars of
soil slurries. They found that ^0 was reduced more rapidly than N0^.
In addition. Smith et al. (1983) found that, under certain redox
conditions, the soil could act as a sink for ^0.
In general, redox studies of stirred suspensions are useful and
have incorporated the soil factor in the study of chemical equilibrium
in soil solutions. In many cases, however, there is a need to study
redox potential of soil samples that are more closely related to field
conditions where soil particles are not in suspension (Patrick and
Henderson,1981).
Soil Column Studies
Some of the recent studies of soil Eh effects have dealt with
monitoring the changes in redox status in soil columns subjected to
varying physical and biochemical conditions.
104
In a column experiment, Baligar et al. (1980) found that redox
potential increased with increasing moisture stress and decreasing
bulk densities. Their results were related to soil aeration status.
However, no definite relationship was found between soil texture and
Eh. Moisture-Eh relationships were studied by Callebaut et al.
(1982). Their column studies showed that oxygen diffusion rate (ODR)
above a water table increased when the water table was lowered.
Oxygen-rich soil layers usually gave inaccurate Eh readings and no
definite relationship between water table depth and Eh was
established.
The gradual establishment of reducing conditions after saturation
of soil columns was studied by Farooqi and De Mooy (1983). They
observed that the peak in microbial activity coincided with the lowest
pe + pH values and occured after 48 hours of saturation. Microbial
activity and Eh were found related to the availability of organic
matter, especially C in the soil (Farooqi and De Mooy, 1983). The role
of organic matter was also investigated by Ghosh et al. (1982)
using sterile mixtures of soil, metals and humus. Potentiometric
titrations showed that humic substances can change the valence status
of Fe, Cu and Mn thus affecting the general redox conditions in the
soil even without the action of microorganisms (Ghosh et al., 1982).
The soil column studies cited above, monitored Eh as affected
by changing soil environments. To study the effect of redox
potentials on soil constituents it is necessary to be able to control
Eh levels. Patrick and Henderson (1981) devised a system for the
control of redox potential in compacted soil cores by the use of
105
saturation water and based on the ORD-Eh relationship mentioned
previously (Callebaut et al., 1982). The apparatus was used to study
the behavior of Fe and Mn at variable reducing conditions. The results
of the core study were in agreement with the soil suspension studies;
Mn and Fe were reduced at Eh levels of +400 and +200 mV respectively.
Soil texture was found to be a major concern for the success of the
procedure suggested; samples with a fine texture had to be mixed with
sand to permit successful Eh adjustment (Patrick and Henderson, 1981).
This of course introduces a new factor that may significantly alter
the natural soil conditions.
Conclusions
Reactions involving electron exchange occur in all natural
environnmental systems. This appendix covered the measurement and
control of redox potential and reviewed some of the applications to
soil investigations. The main findings and conclusions can be
summarized as follows:
i. The theory behind Eh quantification is sound (and as valid as pH
measurements) but field measured Eh values are still considered
semi-quantitative, principally because the theoretical state of
thermodynamic equilibrium is practically never present in the
natural environment.
ii. Eh measurement is most adapted to estimate "electron pressure" in
reducing environments and under controlled laboratory conditions.
iii. In-situ studies have yielded significant results, mainly
concerning the behavior of Fe and Mn in soils. Field
measurements, however, are limited because of practical
106
difficulties and do not allow for study where the redox potential
is controlled.
iv. Laboratory studies of the effect of Eh-pH levels on various soil
components concluded that some of the major effects are
modification of surface-water interfaces, and the dissolution of
solids (mainly oxides).
v. The results of soil column studies are questionable because they
require changing the nature of the soil under study.
107
REFERENCES
Amacher, M.C. , J. Kotuby-Amacher, H.M. Selim, and I.K. Iskandar. 1986.
Retention and release of metals by soils: evaluation of several
models. Geoderma 38:131-154.
Anderson, M.A., H.I. Tejedor-Tejedor, and R. R. Stanforth. 1985.
Influence of aggregation on the uptake of phosphate by goethite.
Environ. Sci. Technol. 19:632-637.
Atkinson, R.J., R.L. Parfitt, and R.C. Smart. 1974. Infra-red study of
phosphate adsorption on goethite. J. Chem. Soc. Faraday Trans. I.
1472-1479.
Atkinson, R.J., A.M. Posner, and J.P. Quirk. 1968. Crystal nucleation
in Fe(III) solutions and hydroxide gels. J. Inorg. Nucl. Chem.
30:2371-2376.
Bache, B.W. , and E.G. Williams. 1971. A phosphate sorption index for
soils. J. Soil Sci. 22:289-301.
Bailey, L.D., and E.G. Beauchamp. 1971. Nitrate reduction and redox
potentials measured with permanently and temporarily placed
platinum electrodes in saturated soils. Can. J. Soil Sci. 51: 51-
58.
Baligar, V.C., F.D. Whisler, and V.E. Nash. 1980. Soybean seedling
root growth as influenced by soil texture, matric suction and
bulk density. Commun. Soil Sci. Plant Anal. 11:903-915.
Barrow, N.J. 1984. Modeling the effects of pH on phosphate sorption by
soils. J. Soil Sci. 35:283-297.
Bascomb, C.I. 1968. Distribution of pyrophosphate-extractable iron and
organic carbon in soils of various groups. J. Soil Sci. 19:251-
268.
Belanger, T.V., D.F. Mikutel, and P.A. Churchill. 1984. Groundwater
seepage nutrient loading in a Florida lake. Water Res. 19:773-
781.
Berkheiser, V.E., J.J. Street, P.S.C. Rao, and T.L. Yuan. 1980.
Partitioning of inorganic orthophosphate in soil water systems.
CRC Critical Rev. Environ. Control 10:179-224.
Bohn, H.L. 1968. Electromotive force of inert electrodes in soil
suspensions. Soil Sci. Soc. Am. Proc. 32:211-215.
Bohn, H.L. 1971. Redox potentials. Soil Sci. 122:39-45.
108
Bohn, H.L., and R.K. Bohn 1986. Solid activity coefficients of soil
components. Geoderma 38:3-18.
Borggaard, O.K. 1983. Effect of surface area and mineralogy of iron
oxides on their surface charge and anion-adsorption properties.
Clays Clay Minerals 31:230-232.
Bouma, J. 1988. When the mapping is over, then what? p. 3. Soil
Resources: Their inventory, analysis and interpretation for use
in the 1990's. Intern. Interactive Workshop, 22-24 March, 1988.
Univ. Minnesota, Minneapolis, MN.
Bradley, J., I. Vimpany, and P.J. Nicholls. 1984. Effects of
waterlogging and subsequent drainage of a pasture soil on
phosphate and oxalate-extractable iron. Aust. J. Soil Res.
22:455-461.
Breeuwsma, A., J.H.M. Wosten, J.J. Vleeshouver, A.M. Van Slobbe, and
J. Bouma. 1986. Derivation of land qualities to assess
environmental problems from soil surveys. Soil Sci. Soc. Am. J.
50:186-190.
Burnham, C.P., and D. Lopez-Hernandez. 1982. Phosphate retention in
different soil taxonomic classes. Soil Sci. 134:376-380.
Callebaut, F., D. Gabriels, W. Minjauw, and M. De Boodt. 1982. Redox
potential, O2 diffusion rate and soil gas composition in
relation to water table in two soils. Soil Sci. 134:149-156.
Canter, L.W., and R.C. Knox. 1985. Septic tank system effects and
groundwater quality. Lewis Publishers, Inc. Chelsea, MI, 336 p.
Chang, A.C., A.L. Page, R.H. Sutherland, and E. Grgurevic. 1983.
Fractionation of phosphorus in sludge-affected soils. J. Environ
Qual. 12:286-290.
Collins, J.F., and S.W. Buol. 1970. Effects of fluctuations in the
Eh-pH environment and/or manganese equilibria. Soil Sci. 110:111
118.
De Laune, R.D., C.J. Smith, and W.H. Patrick, Jr. 1983. Relationship
of marsh elevation, redox potential, and sulfide to Spartina
alterniflora productivity. Soil Sci. Soc. Am. J. 47: 930-935.
Diggle, A.J., and L.C. Bell. 1984. Movement of applied phosphorus
following the mining and revegetation of mineral sands on
Australia's east coast. Aust. J. Soil Res. 22:135-148.
Dirasian, H. 1968. Electrode potentials - Significance in biological
systems. Part 1: Fundamentals of measurement. Water Sewage Works
40:420-425.
109
Dixon, J.B., D.C. Golden, F.G. Calhoun, and P.R. Buseck. 1983.
Synthetic aluminous goethite investigated by high resolution
transmission electron microscopy. Proc. Meetings Electron Microscopy 41:192-193.
Draper, N.R., and H. Smith. 1981. Applied regression analysis. 2nd
ed., John Wiley, New York. 709 p.
Dzombak, D. A. 1986. Toward a uniform model for the sorption of
inorganic ions on hydrous oxides. Ph.D. diss. Massachusetts
Inst. Technology, Cambridge, MA.
Evans, A. Jr. 1985. The adsorption of inorganic phosphate by a sandy
soil as influenced by dissolved organic compounds. Soil Sci.
140:251-255.
Everett, D.H., and R.H. Otterwill (eds.). 1969. Surface area
determination. Butterworths, London. 405 p.
Farooqi, M.A.R., and C.J. De Mooy. 1983. Effect of soil saturation
on redox potential and microbial activity. Commun. Soil Sci.
Plant Anal. 14:185-197.
Fox, R., and G. Rolfe. 1982. A method for collecting soil solutions
from flooded soils. Soil Sci. Soc. Am. J. 46:1332-1333.
Fuller, W.H., and A.W. Warrick. 1985. Soils in waste treatment and
utilization. Vol. I and II. CRC Press, Inc., Boca Raton, FL.
Gambrell, R.P., C.N. Reddy, V. Collard, G. Green, and W.H. Patrick,
Jr. 1984. The recovery of DDT, kepone, and permethrin added to
soil and sediment suspensions incubated under controlled redox
potential and pH conditions. J. Water Pollut. Control Fed.
56:174-182.
Gastuche, M.C., and A. Herbillon. 1962. Etude des gels d'alumine:
crystallisation en milieu desionise. Bull. Soc. Chim. Fr. 1404-
1412.
Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis. In: A. Klute
(ed.) Methods of soil analysis. Part 1. 2nd ed. Agronomy
9:383-412. Am. Soc. Agron., Madison, WI.
Ghosh, K. , A. Chattopadhyay, and C. Varadachari. 1982. Electron
exchange behavior of humic substances with Fe, Cu, and Mn. Soil
Sci. 135:193-196.
Gilliom, R.J., and C.R. Patmont. 1983. Lake phosphorus loading from
septic systems by seasonally perched groundwater. J. Water
Pollut. Control Fed. 55:1297-1305.
110
Goldberg, S. 1983. A chemical model of phosphate adsorption on
oxide minerals and soils. Ph.D. diss. University of California, Riverside, CA.
Goldberg, S. 1985. Chemical modeling of anion competition on goethite
using the constant capacitance model. Soil Sci. Soc. Am. J. 49:851-856.
Goldberg, S., and G. Sposito. 1984a. A chemical model of phosphate
adsorption by soils: I. Reference oxide minerals. Soil Sci. Soc.
Am. J. 48:772-778.
Goldberg, S., and G. Sposito. 1984b. A chemical model of phosphate
adsorption by soils: II. Non calcareous soils. Soil Sci. Soc. Am. J. 48:779-783.
Goldberg, S., and G. Sposito. 1985. On the mechanism of specific
phosphate adsorption by hydroxylated mineral surfaces: A review.
Commun. Soil Sci. Plant Anal. 16:801-821.
Gotoh, S., and W.H. Patrick, Jr. 1974. Transformation of iron in a wa¬
terlogged soil as influenced by redox potential and pH. Soil
Sci. Soc. Am. Proc. 38:66-71.
Gschwend, P.M., and M.D. Reynolds. 1987. Monodisperse ferrous
phosphate colloids in an anoxic groundwater plume. J. Contam.
Hydro1. 1:309-327.
Harter, R.D. 1984. Curve-fit errors in Langmuir adsorption maxima.
Soil Sci. Soc.Am. J. 48:749-752.
Harter, R.D., and D.E. Baker. 1977. Applications and misapplications
of the Langmuir equation. Soil Sci. Soc. Am. J. 48:779-783.
Hill, D.E., and B.L. Sawhney. 1981. Removal of phosphorus from
wastewater by soil under aerobic and anaerobic conditions. J.
Environ. Qual. 10:401-405.
Hillel, D. 1977. Computer simulation of soil water dynamics: a
compendium of recent work. Int. Dev. Res. Center, Ottawa, 214 p.
Hingston, F.J. 1981. A review of anion adsorption, p. 51-90. In: M.A.
Anderson, and A.J. Rubin(eds.). Adsorption of inorganics at solid-
liquid interfaces. Ann Arbor Science, Ann Arbor, MI.
Hingston, F.J., A.M. Posner, and J.P.Quirk. 1972. Anion adsorption by
goethite and gibbsite: I. The role of the proton in determining
adsorption envelopes. J. Soil Sci. 23:177-192.
Hohl, H., L. Sigg, and W. Stumm. 1980. Characterization of surface
chemical properties of oxides in natural waters, p. 1-31. In:
M.C. Kavanaugh, and J.O. Leckie (eds.). Particulates in water.
Adv. in Chem. Series, No. 189. Washington, DC.
Ill
Holford, I.C.R. 1979. Evaluation of soil phosphate buffering indices.
Aust. J. Soil Res. 17:495-504.
Holford, I.C.R. 1982. The comparative significance and utility of the
Freundlich and Langmuir parameters for characterizing sorption
and plant availability of phosphate in soils. Aust. J. Soil Res. 20:233-242.
Holford, I.C.R., and W.H. Patrick, Jr. 1979. Effects of reduction and
pH changes on phosphate sorption and mobility in an acid soil.
Soil Sci. Soc. Am. J. 43:292-297.
Holford, I.C.R., and W.H. Patrick Jr. 1981. Effects of duration of
anaerobiosis and reoxidation on phosphate sorption
characteristics of an acid soil. Aust. J. Soil Res. 19:69-78.
Holford, I.C.R., R.W.M. Wedderbum, and C.E.G. Mattingly. 1974. A
Langmuir two-surface equation as a model for phosphate adsorption
by soils. J. Soil Sci. 25:243-255.
Honeyman, B.D. 1984. Cation and anion adsorption at the oxide/solution
interface in systems containing binary mixtures of adsorbents: an
investigation of the concept of adsorptive additivity. Ph. D.
diss. Stanford University, Stanford, CA. UMI No. 8420552.
Hook, J.E. 1983. Movement of phosphorus and nitrogen in soil following
application of municipal wastewater, p 241-255. In: Chemical
mobility and reactivity in soil systems. Soil Sci. Soc. Am.
Special Publication number 11. Madison, WI.
Houck, C.P., S.P. Putnam, and R.C. Loehr. 1987. Initial operations
report on sludge farm. J. Environ. Engin. 113:970-978.
Humenik, F.J., M.D. Smolen, and S.A. Dressing. 1987. Pollution from
nonpoint sources. Where we are and where we should go. Environ.
Sci. Technol. 21:737-742.
Hunt, P.G., T.A. Matheny, R.B. Campbell, and J.E. Parsons. 1982.
Ethylene accumulation in southeastern coastal plain soils: Soil
characteristics and oxidative-reductive involvement. Commun. Soil
Sci. and Plant Anal. 13:267-278.
Hunt, P.G., T.A. Matheny, J.E. Parsons, and R.B. Campbell. 1980. An
aerobic-anaerobic interactive system for soil suspensions. Soil
Sci. Soc. Am. J. 44:1328-1329.
Hsu, P.H. 1977. Aluminum hydroxides and oxyhydroxides p. 99-138. In:
J.B. Dixon and S. Weed (eds.) Minerals in soil environments. Soil
Sci. Soc. Am,. Madison, WI.
Iskandar, I.K. 1981. Modeling wastewater renovation. John Wiley, New
York. 410 p.
112
James, R.O., and G.A. Parks. 1982. Characterization of aqueous
colloids by their electrical double-layer and intrinsic surface
chemical properties, p. 119-216. In: E. Matijevic Surface Colloid
Sci., Vol. 12, Plenum Press, New York.
Jeanson, C. 1972. Etude microscopique de depots de fer, de manganese
et de calcium dans un sol experimental, leur association avec des
microorganismes. Rev. Ecol. Biol. Sol. 9:479-489.
Johnson, D.W. , D.W. Cole, H. Van Miegroet, and F.W. Homg. 1986.
Factors affecting anion movement and retention in four forest
soils. Soil Sci. Soc. Am. J. 50:776-783.
Jury, W.A. 1983. Chemical transport modeling: current approaches and
unresolved problems. In: D.W. Nelson (ed.) Chemical mobility and
reactivity in soil systems. Soil Sci. Soc. Am. Special Publ. No.
11., Madison, WI.
Khalid, R.A., W.H. Patrick Jr., and R.D. Delaune. 1977. Phosphorus
sorption characteristics of flooded soils. Soil Sci. Soc. Am. J.
41:305-310.
King,L.D. 1985. Design and management considerations in applying
minimally treated wastewater to land. p. 15-26. In: Utilization,
treatment, and disposal of waste on land. Proc. Workshop.
Chicago, IL. 6-7 December, 1985. Soil Sci. Soc. Am., Madison, WI.
Kleiss, H.J., and M.T. Hoover. 1986. Soil and site criteria for on¬
site systems, p. 111-128. In: K.W. Brown (ed.) Utilization,
treatment, and disposal of waste on land. Soil Sci. Soc. Am.
Madison, WI.
Kuo, J-H., and T.F. Yen. 1988. Some aspects in predicting the point of
zero charge of a composite oxide system. J. Colloid Interface
Sci. 121:220-225.
Langmuir, I. 1918. The adsorption of gases on plane surfaces of glass,
mica, and platinum. J. Am. Chem. Soc. 40:1361-1403.
Larsen, S. 1967. Soil Phosphorus. Adv. Agron. 19:151-206.
Laverdiere, M.R., and A. Karam. 1984. Sorption of phosphorus by some
surface soils from Quebec in relation to their properties.
Commun. Soil Sci. Plant Anal. 15:1215-1230.
Lee, F.Y., and J.A. Kittrick. 1984. Electron microprobe analysis of
elements associated with zinc and copper in an oxidizing and an
anaerobic soil environment. Soil Sci. Soc. Am. J. 48:548-554.
Letey, J., N. Valoras, D.D. Focht, and J.C. Ryden. 1981. Nitrous oxide
production and reduction during denitrification as affected by
redox potential. Soil Sci. Soc. Am. J. 45:727-730.
113
Light, T.S. 1972. Standard solution for redox potential measurements. Anal. Chem. 44:1038-1039.
Lindberg, R.D., and D.D. Runnells. 1984. Ground water redox
reactions: An analysis of equilibrium state applied to Eh
measurements and geochemical modeling. Science 255:925-927.
Lindsay, W.L. 1979. Chemical equilibria in soils. J. Wiley Inc., New
York. 449p.
Lindsay, W.L., and E.C. Moreno. 1960. Phosphate phase equilibria in
soils. Soil Sci. Soc. Am. Proc. 25:177-182.
Lindsay, W.L., M. Sadiq, and L.K. Porter. 1981. Thermodynamics of
inorganic N transformation. Soil Sci. Soc. Am. J.45:61-66.
Liverman, D.M., M.E. Hanson, B.J. Brown, R.W. Meredith. 1988. Global
sustainability: toward measurement. Environ. Management 12:133-
143.
Loehr, R.C. 1985. The role of soil science in the utilization,
treatment, and disposal of waste, p. 305-318. In: Utilization,
treatment, and disposal of waste on land. Proc. Workshop.
Chicago, IL. 6-7 December, 1985. Soil Sci. Soc. Am., Madison, WI.
Loehr, R.C., W.J. Jewell, J.D. Novak, W.W. Clarkson, and G.S.
Friedman. 1979. Land application of waste. Vol. I and II. Van
Nostrand Reinhold Co., New York.
Loehr, R.C., and M.R. Overcash. 1985. Land treatment of wastes:
concepts and general design. J. Environ. Engin. 111:141-160.
Loganathan, P., N.O. Isirimah, and D.A. Nwachuku. 1987. Phosphorus
sorption by Ultisols and Inceptisols of the Niger delta in
southern Nigeria. Soil Sci. 139:330-338.
Lovley, D.R. 1987. Organic matter mineralization with the reduction
of ferric iron: A review. Geomicrobiol. J. 5:375-399.
Lyman, W.J. 1982. Adsorption coefficient for soils and sediments, p.
4.1-4.33. In: W.J. Lyman, W.F. Reehl, and D.H. Rosenblatt, (eds.)
Handbook of chemical property estimation methods. Mc.Graw Hill,
New York.
Mansell, R.S., and H.S. Selim. 1981. Mathematical models for
predicting reactions and transport of phosphorus applied to
soils. In: I.K. Iskandar (ed.) Modeling waste water renovation.
J. Wiley, New York.
Massey, D.T. 1983. Features of federal law that encourage adoption of
land application of wastewater. J. Water Res. Planning Management
109:121-133.
114
McBride, M.B. 1987. Adsorption and oxidation of phenolic compounds
by iron and manganese oxides. Soil Sci. Soc. Am. J. 51:1466- 1472.
McKeague, J.A., J.E. Brydon, and N.M. Miles. 1971. Differentiation
of forms of extractable iron and aluminum in soils. Soil Sci. Soc. Am. Proc. 35:33-38.
McLaughlin, J.R., J.C. Ryden, and J.K. Syers. 1981. Sorption of
inorganic phosphate by iron and aluminum containing components. J. Soil Sci. 32:31-36.
Moore, I.C., and F.W. Madison. 1985. Description and application of an
animal waste phosphorus loading model. J. Environ. Qual. 14:364-
369.
Morel, F.M.M. 1983. Principles of aquatic chemistry. John Wiley, New
York. 319 p.
Morel, F.M.M., and P.M. Gschwend. 1987. The role of colloids in the
partitioning of solutes in natural waters, pp 405-422. In: W.
Stumm (ed.) Aquatic Surface Chemistry. J. Wiley, New York.
Morel, F.M.M., G.J. Yeasted, and J.C. Westall. 1981. Adsorption
models: a mathematical analysis in the framework of general
equilibrium calculations, p. 263-294. In: M.A. Anderson, and A.J.
Rubin (eds.). Adsorption of inorganics at solid-liquid interfaces.
Ann Arbor Science, Ann Arbor, MI.
Mueller, S.C., L.H. Stolzy, and G.W. Fick. 1985. Constructing and
screening platinum microelectrodes for measuring soil redox
potential. Soil Sci. 139:558-560.
Munsell Color. 1975. Munsell soil color charts. Munsell Color Comp.
Baltimore, MD.
Murphy, J., and J. P. Riley. 1962. A modified single solution method
for the determination of phosphate in natural waters. Anal. Chim.
Acta 27:31-36.
Nagpal, N.K. 1985. Long-term phosphorus sorption in a Brunisol in
response to dosed-effluent loading. J. Environ. Qual. 14:280-285.
Nair, P.S., T.J. Logan, A.N. Sharpley, L. E. Sommers, M.A. Tabatabai,
and T.L. Yuan. 1984. Interlaboratory comparison of a standardized
phosphorus adsorption procedure. J. Environ. Qual. 13:591-595.
Nelson, D.W., and L.E. Sommers. 1986. Total carbon, organic carbon,
and organic matter. In: A.L. Page et al. (ed.) Methods of soil
analysis. Part 2. 2nd ed. Agronomy 9:539-579. Am. Soc. Agron.,
Madison, WI.
115
Norstadt, F.A., N.P. Swanson, and B.R. Sabey. 1977. Site design and
management for utilization and disposal of organic wastes, p.
348-376. In: L.F. Elliot, and F.J. Stevenson (eds.) Soil for
management of organic wastes and wastewaters. Soil Sci. Soc. Am., Madison, WI.
Noss, R. 1984. The US approach to water pollution control. Environ. Int. 10:217-223.
Olsen, S.R., and F.S. Watanabe. 1957. A method to determine a
phosphorus adsorption maximum of soils as measured by the
Langmuir isotherm. Soil Sci. Soc. Am. Proc. 21:144-149.
Overcash, M.R., and D. Pal. 1979. Design of land treatment systems for
industrial wastes; theory and practice. Ann Arbor Sci. Publishers
Inc., Ann Arbor, MI.
Parfitt, R.L., A.R. Fraser, J.D. Russell, and V.C. Farmer. 1977.
Adsorption on hydrous oxides. II. Oxalate, benzoate and phosphate
on gibbsite. J. Soil Sci. 28:40-47.
Parker, J.C., and P.M. Jardine. 1986. Effects of heterogeneous
adsorption behavior on ion transport. Water Res. Research
22:1334-1340.
Patrick, W.H. Jr., and R.E. Henderson. 1980. Reduction and reoxidation
cycles of manganese and iron in flooded soil and in water
solution. Soil Sci. Soc. Am. J. 45:855-859.
Patrick, W.H. Jr., and R.E. Henderson. 1981. A method for controlling
redox potential in packed soil cores. Soil Sci. Soc. Am. J.
45:35-38.
Patrick, W.H. Jr., and I.C. Mahapatra. 1968. Transformation and
availability to rice of nitrogen and phosphorus in water-logged
soils. Adv. Agron. 20:323-359. Am. Soc. Agron., Madison, WI.
Patrick, W.H. Jr., B.G. Willingood, and J.T. Moraghan. 1973. A
simple system for controlling redox potential and pH in soil
suspensions. Soil Sci. Soc. Am. J. 37:331-332.
Pedrazzini, F.R. 1983. Mineral nitrogen distribution, redox potential,
and pH of rice paddy soils profiles in northern Italy. Commun.
Soil Sci. Plant Anal. 14:309-320.
Peech, M., R.L. Cowan, and J.H. Baker. 1962. A critical study of the
BaC^-triethanolamine and ammonium acetate methods for
determining the exchangeable hydrogen content of soils. Soil Sci.
Soc. Am. Proc. 26:37-40.
Pickering, E.W., and P.L.M. Veneman. 1984. Moisture regimes and
morphological characteristics in a hydrosequence in central
Massachusetts. Soil Sci. Soc. Am. J. 48:113-118.
116
Pi-ng» C.L., and G.J. Michaelson. 1986. Phosphorus sorption by major
agricultural soils of Alaska. Commun. Soil Sci. Plant Anal.
17:299-320.
Ponnamperuma, F.N. 1972. The chemistry of submerged soils. Adv.
Agron. 24:29-88. Am. Soc. Agron., Madison, WI.
Polyzopoulos, N.A., V.Z. Keramidas, and H. Kiosse. 1983. Phosphate
sorption by some Alfisols of Greece as described by commonly used
isotherms. Soil Sci. Soc. Am. J. 49:81-84.
Ratkowsky, D.A. 1986. A statistical study of seven curves for
describing the sorption of phosphate by soil. J. Soil Sci.
37:183-189.
Runnells, D.D., R.D. Linberg, and J.H. Kempton. 1987. Irreversibility
of Se(VI)/Se(IV) redox couple in synthetic basaltic ground water
at 25°C and 75°C. Mat. Res. Soc. Symp. Proc. 84:723-733.
Reddy, C.N., A. Jugsujinda, and W.H. Patrick, Jr. 1976. System for
growing plants under controlled redox potential-pH conditions.
Agron. J. 68:987-988.
Reddy, K.R., and W.H. Patrick, Jr. 1983. Effects of aeration on
reactivity and mobility of soil constituents. In: Chemical
mobility and reactivity in soil systems. SSSA Special
Publication No. 11. Soil Sci. Soc. Am., Madison, WI.
Sah, R.N., and D.S. Mikkelsen. 1986. Transformations of inorganic
phosphorus during the flooding and draining cycles of soil. Soil
Sci. Soc. Am. J. 50:62-67.
SAS Institute, Inc. 1985. SAS/STAT guide for personal computers,
version 6. SAS Institute, Inc, Cary, NC. 378 p.
Savant, N.K., and R. Ellis Jr. 1964. Changes in redox potential and
phosphorus availability in submerged soil. Soil Sci. 98:388-394.
Savant, N.K., and G.H. McClellan. 1987. Do iron oxides influence soil
properties and nitrogen transformations in soils under wetland
rice-based cropping systems? Commun. Soil Sci. Plant Anal.
18:83-113.
Sawhney, B.L. 1973. Electron microprobe analysis of phosphate in soils
and sediments. Soil Sci. Soc. Am. Proc. 37:658-660.
Schwab, A.P., and W.L. Lindsay. 1983a. Effect of redox on the
solubility and availability of iron. Soil Sci. Soc. Am. J.
47:201-205.
Schwab, A.P., and W.L. Lindsay. 1983b. The effect of redox on the
solubility of manganese in a calcareous soil. Soil Sci. Soc.
Am. J. 47:217-220.
117
Schwarz, J.A. , C.T. Driscoll, and A.K. Bhanot. 1984. The zero point of
charge of silica-alumina oxide suspensions. J. Colloid Interface
Sci. 97:55-61.
Schweich, D., and M. Sardin. 1981. Adsorption, partition, ion exchange
and chemical reaction in batch reactors or in columns- A review.
J. Hydro1. 50:1-33.
Schwertmann, D., and R.M. Taylor. 1977. Iron oxides, p.145-180. In:
J.B. Dixon, and S.B. Weed (eds.) Minerals in soil environments.
Soil Sci. Soc. Am., Madison, WI.
Settergren, D., and L.C. Tennyson. 1979. The importance of soil in
evaluating potential wastewater disposal sites in the Missouri
Ozarks. Water Res. Bulletin. 15:1353-1364.
Sibanda, H.M., and S.D. Young. 1986. Competitive adsorption of humus
acids and phosphate on goethite, gibbsite and two tropical soils.
J. Soil Sci. 37:197-204.
Sibbesen, E. 1981. Some new equations to describe phosphate sorption
by soils. J. Soil Sci. 32:67-74.
Sigg, L. and W. Stumm. 1980. The interaction of anions and weak acids
with the hydrous goethite surface. Colloids Surfaces 2:101-
117.
Sims, J.L. , and W.H. Patrick, Jr. 1978. The distribution of
micronutrient cations in soil under conditions of varying redox
potential and pH. Soil Sci. Soc. Am. J. 42:258-262.
Singh, B.B., and M.A. Tabatabai. 1977. Effects of soil properties on
phosphate sorption. Commun. Soil Sci. Plant Anal. 8:97-107.
Sircar, S. 1984. Adsorption of gases on heterogeneous adsorbents. J.
Chem. Soc. Faraday Trans. 1. 80:1101-1111.
Smillie, G.W. , D. Curtin, and J.K. Syers. 1987. Influence of
exchangeable calcium on phosphate retention by weakly acid soils.
Soil Sci. Soc. Am. J. 51:1169-1172.
Smith, C.J., M.F. Wright, and W.H. Patrick, Jr. 1983. The effect of
soil redox potential and pH on the reduction of nitrous oxide. J.
Environ. Qual. 12:186-188.
Soil Survey Staff. 1972. Soil survey laboratory methods and procedures
for collecting soil samples. Soil Survey Investigations Report
No.1. U.S.D.A. Soil Cons. Serv., Govern. Printing Office,
Washington, DC.
Sposito, G. 1981a. The operational definition of the zero point of
charge in soils. Soil Sci. Soc. Am. J. 45:292-297.
118
Sposito, G. 1981b. The thermodynamics of soil solutions. Oxford
Clarendon Press. Oxford, England.
Sposito, G. 1982. On the use of the Langmuir equation in the
interpretation of "adsorption" phenomena: II. The "two-surface"
Langmuir equation. Soil Sci. Soc. Am. J. 46:1147-1152.
Sposito, G. 1985. Chemical models of inorganic pollutants in soils.
CRC Crit. Rev. Environ. Control 15:1-24.
Steele, D.P., G.W. Petersen, and D.D. Fritton. 1986. Soil potential
ratings for on-site wastewater disposal. J. Environ. Qual. 15:37-
44.
Stuanes, A.O. 1982. Phosphorus sorption by soil; a review, p. 145-152.
In: A.S. Eikum and R.W. Seabloom (eds.) Alternative wastewater
treatment. D. Reidel Publishing Co., Dordrecht, The Netherlands.
Stuanes, A.O. 1984. Phosphorus sorption of soils to be used in
wastewater renovation. J. Environ. Qual. 13:220-224.
Stuanes, A.O., and C.G. Enfield. 1984. Prediction of phosphate
movement through some selected soils. J. Environ. Qual. 13:317-
320.
Stumm, W. 1986. Coordinative interactions between soil solids and
water - An aquatic chemist's point of view. Geoderma 38:19-30.
Stumm, W., R. Kummert, and L. Sigg. 1980. A ligand exchange model for
the adsorption of inorganic and organic ligands at hydrous oxides
interfaces. Croatica Chemica Acta. 53:291-312.
Stumm, W.J., and J. Morgan. 1981. Aquatic chemistry. 2nd ed. Wiley
Interscience Publication, John Wiley, New York.
Switzenbaum, M.S., J.V. De Pinto, T.C. Young, and J.K. Edzwald. 1980.
A survey of phosphorus removal in lower Great Lakes municipal
treatment plants. J. Water Pollut. Control. Fed. 25:2628-2633.
Syers, J.K., and I.K. Iskandar. 1981. Soil-phosphorus chemistry, p.
571-599. In: I.K. Iskandar (ed.) Modeling wastewater renovation,
land treatment. John Wiley, New York.
Time. 1989. Endangered Earth. Planet of the Year. Time Magazine
133:25-72.
Thomas, R., and J.P. Law. 1977. Properties of waste waters, p. 46-72.
In: L.F. Elliott and F.J. Stevenson (eds.) Soils for management
of organic waste and waste waters. Am. Soc. Agron., Madison, WI.
119
Tirsch, F.S., and J.H. Baker. 1976. Heavy metal reactions in soils, p.
II.1~II.26. In: D.D. Adrian (ed.) Suitability of New England soil
in accepting waste effluents, phase I. Water Res. Research
Center. Univ. Massachusetts, Amherst, MA.
Tofflemire, T.J., and M. Chen. 1976. Evaluation of phosphate
adsorption capacity and cation exchange capacity and related data
for 35 common soil series in New York state. Technical Report No.
45, NY State Dept, of Environ. Conservation, Albany, NY.
Travis, C.C., and E.L. Etnier. 1981. A survey of sorption
relationships for reactive solutes in soil. J. Environ. Qual.
10:8-17.
Turner, K. (ed.) 1988. Sustainable Environmental Management,
Principles and Practices. Belhaven Press, London/Westview Press,
Boulder, CO. 292 pages.
Uebler, R.L. 1984. Effect of loading rate and soil amendments on
organic nitrogen and phosphorus leached from wastewater soil
absorption system. J. Environ. Qual. 13:475-479.
U.S. Environmental Protection Agency. 1981. Land treatment of
municipal wastewater. EPA 625/1-81-013. U.S.E.P.A., Center for
Environmental Research Information, Cincinnati, OH.
Van Riemsdijk, H., G.H. Bolt, L.K. Koopal, and J. Blaakmeer. 1986.
Electrolyte adsorption on heterogeneous surfaces: adsorption
models. J. Colloid Interface Sci. 109:219-228.
Van Riemsdijk, W.H., and J. Lyklema. 1980a. Reaction of phosphate with
gibbsite (A1(0H)^) beyond the adsorption maximum. J. Colloid
Interface Sci. 76:55-66.
Van Riemsdijk, W. H., and J. Lyklema. 1980b. The reaction of phosphate
with aluminum hydroxide in relation with phosphate bonding in
soils. Colloids and Surfaces 1:33-44.
Van Riemsdijk, W.H., and A.M.A. van der Linden. 1984. Phosphate
sorption by soils: II. Sorption measurement technique. Soil Sci.
Soc. Am. J. 48:541-544.
Van Riemsdijk, W.H., F.A. Westrate, and J.Beek. 1977. Phosphates in
soils treated with sewage water: III. Kinetic studies on the
reaction of phosphates with Al compounds. J. Environ. Qual. 6:26-
29.
Veith, J.A. , and G. Sposito. 1977. On the use of the Langmuir equation
in the interpretation of "adsorption" phenomena. Soil. Sci. Soc.
Am. J. 41:697-702.
120
Veneman, P.L.M. 1982. The septic tank as a wastewater treatment and
disposal facility. Water Resour. Res. Center Publ. No. 128.
University of Massachusetts, Amherst, MA. 45 p.
Veneman, P.L.M., and E.W. Pickering. 1983. Salt bridge for redox
potential measurements. Commun. Soil Sci. Plant Anal.
14:669-677.
Vepraskas, M.J., and L.P. Wilding. 1983. Aquic moisture regimes in
soils with and without low chroma colors Soil Sci. Soc. Am. J.
47:280-285.
Vijayachandran, P.K., and R.D. Harter. 1975. Evaluation of phosphorus
adsorption by a cross section of soil types. Soil Sci. 119:119-
126.
Vollenweider, R.A. 1968. Scientific fundamentals of the eutrophication
of lakes and flowing waters. Organization for Economic
Cooperation and Development. Technical Report DA5/SCI/68.27,
Paris, France.
Westall, J. 1979. MICROQL: I. A chemical equilibrium program in basic.
Swiss Federal Inst, of Technology, Duebendorf, Switzerland.
Westall, J.C. 1987. Adsorption mechanisms in aquatic surface
chemistry. 3-32. In: W. Stumm (ed.) Aquatic Surface Chemistry.
Wiley, NY.
Westall, J.C. 1982. FITEQL: A computer program for determination of
chemical equilibrium constants from experimental data. Report 82-
01, Dept, of Chemistry, Oregon State Univ., Corvallis, OR.
Westall, J., and H. Hohl. 1980. A comparison of electrostatic models
for the oxide/solution interface. Adv. in Colloid and Interface
Sci. 12:265-294.
Westman, W.E. 1972. Some basic issues in water pollution control
legislation. Am. Scientist 60:767-773.
White, R.E. 1981. Retention and release of phosphate by soil and soil
constituents. CRC Critical Rev. Applied Chem. 2:71-114.
Willett, I.R. 1983. Oxidation-reduction reactions pp.417-416 In:
Soils, An Australian Viewpoint. Division of Soils, CSIRO, CSIRO:
Melbourne/Academic Press: London.
Willett, I.R. 1985. The reductive dissolution of phosphated
ferrihydrite and strengite. Aust. J. Soil Res. 23:237-244.
Willett, I.R., and R.B. Cunningham. 1983. Influence of sorbed
phosphate on the stability of ferric hydrous oxide under
controlled pH and Eh conditions. Aust. J. Soil Res. 21:301-308.
121
Yuretich, R.F., P.J. Stekl, and S.J. Pohanka. 1986. Hydrogeology and
mineral weathering reactions in watersheds of central
Massachusetts. Publication No. 152. Water Res. Research Center.
Univ. Massachusetts, Amherst, MA.
122
