• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 Abstract
 Introduction
 Literature review
 Areas of investigation, sampling...
 Evaluation and comparison of methods...
 Copper binding by sewage, swamp...
 Fate of heavy metals in a wetland...
 Conclusions
 Bibliography
 Biographical sketch






Title: Heavy metal complexation with naturally occurring organic ligands in wetland ecosystems /
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099511/00001
 Material Information
Title: Heavy metal complexation with naturally occurring organic ligands in wetland ecosystems /
Physical Description: vi, 212 leaves : ill., map ; 28 cm.
Language: English
Creator: Tuschall, John Richard, 1950-
Publisher: University of Florida
Publication Date: 1981
Copyright Date: 1981
 Subjects
Subject: Wetland ecology -- Florida   ( lcsh )
Heavy metals   ( lcsh )
Sewage disposal in rivers, lakes, etc   ( lcsh )
Environmental Engineering Sciences thesis Ph. D
Dissertations, Academic -- Environmental Engineering Sciences -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida 1981.
Bibliography: Bibliography: leaves 199-211.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by John Richard Tuschall, Jr.
 Record Information
Bibliographic ID: UF00099511
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000297260
oclc - 08365838
notis - ABS3633

Downloads

This item has the following downloads:

PDF ( 7 MBs ) ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
        Page iv
    Abstract
        Page v
        Page vi
    Introduction
        Page 1
        Page 2
    Literature review
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
    Areas of investigation, sampling procedures, and analytical methods
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
    Evaluation and comparison of methods for determining CL and B’
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
    Copper binding by sewage, swamp water, and peat-extractable organic matter
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
    Fate of heavy metals in a wetland ecosystem
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
    Conclusions
        Page 196
        Page 197
        Page 198
    Bibliography
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
        Page 211
    Biographical sketch
        Page 212
        Page 213
        Page 214
        Page 215
Full Text










HEAVY METAL COMPLEXATION WITH NATURALLY OCCURRING
ORGANIC LIGANDS IN WETLAND ECOSYSTEMS




by


JOHN RICHARD TUSCHALL, JR.


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY
OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1981














ACKNOWLEDGMENTS


This dissertation would not have been possible without the

assistance of many. I would like to express my sincerest appreciation

to the chairman of my supervisory committee, Dr. P. L. Brezonik, for

his assistance and guidance throughout this research project. Addi-

tionally, I would like to thank the other members of my committee-

Drs. W. B. Arbuckle, G. R. Best, J. E. Singley, and J. J. Street--for

their continued interest and guidance during this project.

The assistance of Jim Butner, Bill DeBusk, Pete Straub, and Bob

Tighe in field efforts and sample analyses was invaluable. Special

thanks go to Carl Miles, who always had time to discuss my half-baked

ideas and who adroitly aided in many other phases of my research.

Additional thanks go to all the other students and friends who provided

moral and technical support.

Appreciation is extended to Jenny Carter for a highly professional

job of typing this manuscript and to Randy Koper, Rick Cabrera, and

Ruben Belen, Jr., for carefully drafting the figures.

I especially wish to thank my wife for her continued patience and

support during my graduate studies and to my parents, who have always

encouraged me to pursue further education.














TABLE OF CONTENTS



ACKNOWLEDGMENTS.................................................. ii

ABSTRACT............................................................. v

CHAPTER I- INTRODUCTION..............................................1

CHAPTER II-LITERATURE REVIEW....................................... 3

A. Heavy Metal Content of Sewage............................... 3
B. Fate of Metals in the Wetlands Environment...................6
C. Organic Composition of Sewage and Swamp Water...............11
1. Sewage Organics.............................. ....... 12
2. Swamp Organic Matter...................................13
D. Methods for Quantifying Metal-Organic Matter Interactions...16
1. Physical Separation Techniques.........................19
2. Potentiometric Titrations..............................23
3. Ion Exchange......................................... 27
4. Spectrophotometric Methods.............................30
5. Miscellaneous....................................... 33
E. Methods of Fractionating Organic Matter.....................34

CHAPTER III-AREAS OF INVESTIGATION, SAMPLING PROCEDURES,
AND ANALYTICAL METHODS...........................................38

A. Jasper Study............................................ 38
1. Site Description .................................... 38
2. Sampling and Analytical Methods........................40
B. Waldo Study...... ......................................... 42
1. Site Description................................... 42
2. Sampling and Analytical Methods........................44
C. Microcosm Study............................................. 45
D. Analytical Methods for Determining CL and '................ 49
1. Anodic Stripping Voltammetry...........................50
a. Theoretical considerations for determining
CL and 6' by ASV .................................. 52
2. Ion-Selective Electrode................................55
3. Fluorescence Quenching.................................56
4. Continuous Ultrafiltration.............................57
5. Competing Ligand/Differential Spectroscopy.............62
6. Theoretical Considerations for Determining o'
by Scatchard Analysis ..................................66
E. Chemical Characterization and Fractionation of
Sewage and Swamp Water......................................67
F. Copper Binding by Peat Extract..............................71









1. Extraction ............................................. 71
2. Molecular Weight Fractionation.........................71
3. Binding Capacity and Stability Constant Determination..72

CHAPTER IV-EVALUATION AND COMPARISON OF METHODS FOR
DETERMINING CL AND B' ........................................74

A. Evaluation of the ASV-Complexometric Titration..............75
B. Comparison of Five Methods to Determine CL and p'...........90
1. ASV Titration ......................................... 90
2. Ion-Selective Electrode Titration.....................105
3. Fluorescence Quenching................................113
4. Continuous Ultrafiltration............................121
5. Competing Ligand/Differential Spectroscopy............130
6. Comparison of Results.................................138
C ............................................... 138
S............................ ...... ...............140

CHAPTER V-COPPER BINDING BY SEWAGE, SWAMP WATER, AND
PEAT-EXTRACTABLE ORGANIC MATTER................................ 144

A. Sewage Organics......................................... 144
B. Swamp-Water Organics.......................................149
C. Soil Organic Matter: Molecular Size Fractionation
and Metal Binding Ability ..................................154
1. Molecular Size Fractionation.........................154
2. Metal Binding of Peat Extract.........................158

CHAPTER VI-FATE OF HEAVY METALS IN A WETLAND ECOSYSTEM............167

A. Heavy Metals in Basin Swamp................................167
B. Fate of Metals Added to Waldo Swamp........................173
C. Microcosm Study............................................177
1. Cadmium ............................................ ... 184
2. Copper ................................................ 187
3. Manganese ............................................. 190
4. Zinc .................................................. 191
5. Summary ............................................... 192

CHAPTER VII- CONCLUSIONS...........................................196

BIBLIOGRAPHY .......................................................199

BIOGRAPHICAL SKETCH.............................................212















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



HEAVY METAL COMPLEXATION WITH NATURALLY OCCURRING
ORGANIC LIGANDS IN WETLAND ECOSYSTEMS

by

JOHN RICHARD TUSCHALL, JR.

December 1981


Chairman: P. L. Brezonik
Major Department: Environmental Engineering Sciences


The mode, rate, and extent of heavy metal uptake by freshwater

wetlands were investigated as well as heavy metal complexation by a

variety of soluble organic. In one study, the levels of Cd, Cu, Mn,

Ni, Pb, and Zn entering Basin Swamp from the sewage treatment plant in

Jasper, Florida, were low (0.1-30 jg/L). Consequently, no conclusions

could be made regarding heavy metal cycling in Basin Swamp. Another

study involved adding Cd, Cu, Mn, and Zn to septage from Waldo, Flor-

ida, and pumping the wastewater through a 10 x 40-m corridor in a

cypress swamp. Soluble concentrations of all four metals decreased

within the corridor to levels close to those considered acceptable for

discharge to surface water. Rates of immobilization were calculated to

be a minimum of 7.2, 36, and 72, and 72 g metal/ha*day-1 for Cd,

Cu, Mn, and Zn, respectively. In addition, microcosms (model ecosys-

tems that simulated a swamp) were constructed for a laboratory








investigation of factors that control metal immobilization rates. Of

the factors-dissolved organic carbon, iron, calcium, sulfide, and

pH-the latter two most greatly affected metal immobilization rates.

Heavy metal interactions with naturally occurring organic and

model compounds were examined by complexometric titrations using five

separate techniques. Three of the techniques used--anodic stripping

voltammetry (ASV), ion-selective electrode, and fluorescence quench-

ing-were previously published methods, and the other two methods-

continuous flow ultrafiltration and competing ligand/differential

spectroscopy-were developed for this study. Each procedure was used

to determine the available metal binding ability of the organic solu-

tions and the conditional stability constants of the metal-organic com-

plexes. For each solution, agreement among the procedures was good

except for the ASV method, which produced results lower than the

others. Additionally, the ASV method yielded a conditional stability

constant for copper with ethylenedinitrilotetraacetic acid that was

five orders of magnitude lower than other published values. The low

ASV results for copper with swamp-water organic was found to be caused

by the reducible nature of organically completed copper.















CHAPTER I
INTRODUCTION



Increasing rates of pollution of natural waters by sewage and

other anthropogenic sources has resulted in an increased environmental

consciousness during the past quarter century and specifically has

resulted in increasing concern about heavy metal cycling in the envi-

ronment. The costs of achieving "zero discharge" (as required by 1985

by the Federal Water Pollution Control Act of 1972) by conventional

waste-treatment methods have been found to be enormous. Furthermore,

the stresses our burgeoning population has placed on our current fresh-

water resources have created considerable interest in utilizing land

disposal of sewage effluents as a low-cost means of achieving tertiary

treatment of wastewater and simultaneous recharge of freshwater aqui-

fers.

In warm climates, such as the southeastern United States, flowing

wetland ecosystems are a feasible alternative to in-plant tertiary

wastewater treatment. However, the potential use of wetlands for sew-

age disposal is hampered by our lack of knowledge of heavy metal cyc-

ling in a wetland ecosystem. Specifically, the toxicity and mobility

of heavy metals in wetlands remain in question. Although mechanisms of

metal toxicity to the biota is currently an active area of research, a

consensus among researchers is that toxicity is dependent upon the

chemical form of heavy metals. Additionally, heavy metal transport in









an ecosystem varies among the species of metal. Thus, the necessity to

speciate heavy metals is paramount in understanding heavy metal cyc-

ling.

One of the major obstacles to understanding heavy metal cycling in

the environment is the lack of analytical methods that are capable of

accurately speciating heavy metals. Consequently, the nature and

extent of heavy metal-organic interactions in aquatic systems are not

well understood.

The intent of the research proposed here was to investigate the

mechanisms that govern the fate of heavy metals in a wetland ecosystem.

Specifically, the objectives of this research were

1. to evaluate the precision and accuracy of analytical methods

that are used to speciate heavy metals;

2. to quantify heavy metal interactions with organic matter from

sewage, swamp water, and peat sediment; and

3. to determine the fate of heavy metals in cypress wetlands.














CHAPTER II
LITERATURE REVIEW



A. Heavy Metal Content of Sewage


The concentration of heavy metals in treated sewage effluent is

highly variable and extremely difficult to typify, primarily because of

the variable nature of industrial loading to municipal sewage treatment

plants. Oliver and Cosgrove (1974) monitored influent and effluent of

a conventional secondary sewage treatment plant hourly during a 3-day

period and reported metal concentrations varied as much as 100-fold.

Presumably, metal disposal from industrial batch-type processes caused

the observed fluctuations. Although the variable levels of heavy

metals in sewage suggest that generalizations are at best tenuous, the

concentrations of heavy metals in the secondary effluent of the Los

Angeles sewage treatment plant are in the same general range as those

reported by Klein et al. (1974) for wholly domestic sewage in New York

and by Oliver and Cosgrove (1974) for a Canadian sewage treatment plant

receiving domestic and industrial waste (Table II-i). To that extent,

the values listed in Table II-1 can be considered "typical" of munici-

pal sewage effluents.

The reported levels of metals in treated sewage are considerably

higher than typical background levels of metals in freshwater

(Table II-I). The largest differences between concentrations of metal

in wastewater and freshwater appear for nickel and zinc. These metals





Table II-1. Concentrations
in freshwater.


(ug/L) of heavy metals in secondary sewage from domestic sources and


Secondary Effluent Secondary Effluent
Ontarioa Los Angelesb Secondary Effluent
New Yorkc
Dissolved Dissolved Typical
Total (0.45 1) Total (0.2 v) Total Freshwaterd

Cd Ie <1e 8.59 7.49 13i 0.1
(<1-5)f (<1-1) (5)h
Cr 60 <10 55 38 30 1
(<20-680) (<10-20) (85)
Cu 80 70 27 19 140 3
(10-220) (<10-180) (55)
Pb 15 6 48 45 --- 3
(1-50) (<1-14) (85)
Mn 40 30 25 24 --- 8
(20-70) (10-40) (26)
Ni 270 220 116 109 40 0.5
(<30-720) (<30-580) (165)
Zn 560 400 103 90 160 15
(160-2410) (60-1750) (170)

aoliver and Cosgrove (1974).
bChen et al. (1974).
cKlein et al. (1974).
dBowen (1979).
eMean of 84 8-h composites.
fRange of 84 8-h composites.
gMean of two monthly composites.
hMedian of 12 monthly composites.
I'ean of 270 daily samples.








(especially zinc) are more ubiquitous and less toxic than most of the

other metals listed. Conversely, cadmium, which is the most toxic of

the metals reported, was found at the lowest concentration. Further-

more, the metals in treated sewage are present primarily in the dis-

solved state (Table II-i), because heavy metals associated with partic-

ulates are effectively removed by conventional biological waste treat-

ment (Chen et al. 1974).

Levels of heavy metals in domestic sewage effluent from three

communities in Florida--Wildwood, Jasper, and a trailer park in Gaines-

ville--were found to be 10-100 times lower than the values reported in

Table II-1 for treated domestic sewage in New York and Los Angeles

(Klein 1976; Boyt et al. 1977; Carriker 1977; Brezonik et al. 1981).

In these cases, reasons for the very low (i.e., near background)

levels of metals are not readily apparent. Domestic sources of copper

and zinc, for instance, include water pipes, foods, and other consum-

ables (such as some shampoos, vitamins, and drugs), and, hence, one

would expect to find these metals in domestic sewage.

Perhaps the lower levels of metals in sewage from the non-indus-

trialized areas in Florida reflect the atmospheric concentrations of

metals. Williams et al. (1974) reported atmospheric levels of heavy

metals to be up to 100 times higher in industrial areas than in remote

ones; hence, near New York and Los Angeles, higher levels of metals are

expected in surface runoff, which often enters the sewage treatment

facility.

Nonetheless, the presence of heavy metals in sewage effluent is

potentially hazardous, and even in low concentrations, long-term dis-

charge can be detrimental to a receiving ecosystem.









B. Fate of Metals in the Wetlands Environment


Heavy metals applied to wetland ecosystems may experience three

pathways of transport and transformation: (1) uptake by plants;

(2) movement in water to groundwater or surface water; and (3) immobil-

ization in the soil matrix. The fate of heavy metals in freshwater

cypress domes (ponds) was investigated by Klein (1976) and Carriker

(1977), but the concentrations of heavy metals in the waste effluent

(which was wholly from domestic sources) were too low to determine the

ultimate fate of the metals. Similarly, Boyt et al. (1977) reported

low concentrations of zinc, copper, and lead in the effluent from the

Wildwood, Florida, sewage treatment plant and in the swamp receiving

the waste. However, it should be noted that heavy-metals issues were a

minor aspect of the study, and samples were collected on only one

occasion at a few stations. In another study, Mudroch and Capobianco

(1979) investigated a marsh that has received treated domestic sewage

since 1919. Concentrations of metals in the surface water and sediment

cores in the marsh were low and variable, and no trends were observed.

Natural (background) levels and cycling of metals in undisturbed

swamp and marsh areas within Okefenokee Swamp were investigated by

Casagrande and Erchull (1977). Concentrations of metals were reported

for surface water, peat sediment, and 11 species of vegetation, and

these results are summarized in Table 11-2. Many of the reported

values are less than the limit of detection since analysis employed the

flame mode of atomic absorption spectroscopy, and hence the utility of

much of the data is limited. In another study (Casagrande and Erchull

1976), metals in peat sediments (from the same locations described










Table 11-2. Distribution of metals in Minnie's Lake Swamp and Chesser Prairie Marsha.



Cr Cu Hg Mn Ni Pb Zn


Minnie's Lake Swamp

Surface water (mg/L)

Peat (0-6 cm)
(mg/kg dry wt.)

Vegetation (mg/kg dry wt.)

Mean of 14 samples
Minimum
Maximum

Chesser Prairie Marsh

Surface water (mg/L)

Peat (0-6 cm)
(mg/kg dry wt.)

Vegetation (mg/kg dry wt.)

Mean of 14 samples
Minimum
Maximum


<0.04 <0.02 <0.0001 0.01 <0.06 <0.06 0.17


<16 12.5 0.73


6 18 14


0.26 196
0.05 7
0.98 701


<0.04 <0.02 0.0001 0.01 <0.06 <0.06 0.10


12 18 0.34


3 22 32.5


0.93 46
0.20 13
2.04 131


aCasagrande and Erchull (1977).








above) were examined in detail. The authors concluded that the metals

were associated with the organic fraction of the peat sediments. How-

ever, metal concentrations did not correlate with depth of sediment,

and hence, if the influx of metals to the swamp (and therefore the

sediments) has increased due to cultural changes, the metals in the

sediment were redistributed due to (1) the low pH of the sediment

(pH 4), and (2) periodic oxidation of peat during droughts and fires

(Casagrande and Erchull 1976).

Heavy metal cycling in a swamp containing naturally high levels of

copper was reported by Dykeman and de Sousa (1966). Copper in the

Tantramar Swamp, located in New Brunswick, Canada, orginates from a

mineral deposit in the bedrock beneath a 60-m layer of glacial drift

and is circulated by percolating groundwater. The organic peat

sediments contain up to 7% copper, yet the swamp supports a luxuriant

growth of larch, black spruce, and ground cover species typical of

other wetland areas in the region. The authors concluded that the

reason for the apparent lack of copper toxicity to the swamp flora is

due to the strong complexation of copper with the organic peat soil.

Although conclusive research about the fate of heavy metals in

wetlands is sparse, several researchers have investigated heavy metal

cycling in systems similar to freshwater wetlands. Banus et al. (1975)

examined the fate of lead, zinc, and cadmium added to a salt marsh

ecosystem and reported that most of the lead and cadmium added to the

marsh was retained by the surficial sediments, while a small portion

was taken up by the marsh grass. The salt marsh acted as a partial

sink for zinc, with substantial export of zinc to deeper water.

Jennett and Linnemann (1977) reported excellent retention of zinc and








lead applied to Missouri soil under laboratory conditions. The absorp-

tion capacity of the soil approached 100% of the cation exchange capac-

ity (CEC), illustrating the utility of the soil for land disposal of

waste from milling and mining operations.

Waterhyacinths, which are frequently found near the nutrient-rich

effluents of sewage treatment plants, have been shown to accumulate

heavy metals (Dinges 1978). The leaves and stems of the hyacinth cul-

ture that received treated sewage contained high levels of Cr, Cu, Fe,

Hg, Mn, Ni, and Zn, whereas Ag, Cd, and Pb contents were below the

reported detection limits.

Empirical evidence suggests that sediment-bound heavy metals are

associated with several geochemical phases, the principal ones being

hydrous metal oxides, clays, organic, carbonates, and sulfides (Jenne

1968; Khalid et al. 1977). Since the bioavailability and recycling of

heavy metals depend on the chemical phase with which a metal is asso-

ciated in sediments (Jenne and Luoma 1975), the quantitative distribu-

tion of metals among the various phases is of considerable importance.

To determine the distributions, several researchers have proposed

schemes by which metals are selectively extracted from natural sedi-

ments to determine metal partitioning (Presley et al. 1972; Gupta and

Chen 1975; Engler et al. 1977; Tessier et al. 1979). However, these

extraction schemes are laden with difficulties, such as low recovery of

metal and non-specific extraction. Moreover, there is a considerable

lack of information on desorption/adsorption kinetics (Luoma and Jenne

1976; Guy et al. 1978). At present, sediment extraction schemes cannot

adequately assess heavy metal partitioning in the solid phase.








Even though the chemical form of metals in sediments is not amen-

able to accurate analysis, several reports have indicated that heavy

metals can be immobilized by insoluble organic matter similar to the

organic peat found in wetlands. Chaney and Hundemann (1979) added

industrial effluent containing 100 ppm cadmium to 60-cm long columns

containing peat moss, and they found 2 ppb cadmium in the effluent.

Analysis of the column following the passage of 14 L of effluent showed

that most of the cadmium was present in the upper few centimeters of

the column, indicating that organic peat moss was efficient in removing

cadmium. In a similar study by Coupal and Lalancette (1976), peat moss

was used in pilot plant scale (20,000 gpd) to remove Hg, Pb, Sb, Cu,

Ni, Zn, and Cr from a variety of industrial wastes. These workers

found the organic peat to be effective in reducing the levels of all

metals to concentrations ranging from 0.02 to 0.2 mg/L.

In light of the previous discussion, a prior one would expect the

insoluble peat, which usually composes a major portion of sediments in

wetlands, to be at least partially responsible for immobilizing soluble

heavy metals. A variety of factors, such as the concentration of

soluble organic matter, the pH of the water, and sulfide levels, would

be expected to affect the rate and degree of metal immobilization by

peat sediments. However, the influence of these and other environmen-

tal parameters on the fate of heavy metals has not been addressed


satisfactorily to date.








C. Organic Composition of Sewage and Swamp Water


Both sewage effluent and humic-colored surface water (such as

those in cypress swamps) are rich in organic matter that may alter the

rate at which metals are removed from solution. For instance, Davis

and Leckie (1978) proposed three mechanisms by which soluble organic

matter can affect adsorption of metals to solids (either suspended

particulates or sediments):

(1) organic ligands not adsorbed to solid surfaces can decrease

adsorption of metals due to complexation of the metals;

(2) organic ligands adsorbed to surfaces, but not capable of

ligand-bridging with metals, can decrease adsorption by

decreasing the availability of surface sites; and

(3) complex-forming organic ligands strongly adsorbed to surfaces

can function as effective ligand bridges and increase adsorp-

tion.

In addition, soluble organic matter has been reported to enhance uptake

of heavy metals by duckweed (Carriker 1977), stimulate algal growth in

seawater (Barber 1973; Sunda and Guillard 1976; Anderson and Morel

1978), and reduce toxicity of heavy metals to fish (Stiff 1971b; Zitko

et al. 1973). At this time, no consensus exists regarding the mechan-

ism by which organic matter ameliorates the toxic effects of heavy

metals to the biota.

The literature concerning the importance of complexation of heavy

metals by organic matter is vast and reviewed thoroughly elsewhere

(Singer 1973; Leckie and James 1974; Jenne and Luoma 1975; Reuter and

Perdue 1977). However, because the extent of heavy metal complexation








is related to the chemical nature of organic matter (i.e., function-

ality, molecular size and shape), the reported organic composition of

sewage and cypress swamps will be discussed in the following sec-

tion.


1. Sewage Organics

In the case of a cypress swamp receiving heavy metal-containing

waste, the metals likely would be completed initially with the organic

matter present in the wastewater. The most detailed characterization

of organic matter in secondary sewage was reported by Manka et al.

(1974). They investigated three Israeli treatment facilities, which

included trickling filter, stabilization pond, and extended aeration-

activated sludge treatments. The organic composition was similar among

the different facilities, with average values (as percent of COD) as

follows: humics, 43%; proteins, 22%; anionic detergents, 15%; ether

extractables, 14%; carbohydrates, 6%; and tannins, 1.5%.

The composition of total nitrogen in the influents of two munici-

pal sewage treatment plants in Wisconsin was studied by Hanson and Lee

(1971). Average total hydrolyzable amino acid levels were similar for

the two influents (2.1 and 2.7 mg N/L). Ammonia and urea represented

most of the remainder of total nitrogen, although the results were var-

iable and a significant portion of organic nitrogen remained unidenti-

fied. Keller et al. (1978) examined the behavior of soluble organic

nitrogen from a secondary effluent on a column of Sephadex G-15 and

found that the organic nitrogen was distributed over the entire molecu-

lar weight range of the column (i.e., <165->1800 daltons). Their








results imply that the organic nitrogen in secondary sewage effluents

is heterogeneous, and not of a single molecular size.


2. Swamp Organic Matter

The colored organic matter typically found in surface water of

swamps and marshes has been classified collectively as humic substan-

ces. Additionally, the terms organic color, yellow organic acids,

gelbstoff, humates, fulvic acids, and humic acids have been used in the

literature to identify this class of compounds.

Hunic substances are subdivided into three fractions: (1) humic

acid, which is soluble in alkaline solution (0.1-0.5 M NaOH) but is

precipitated by acidification (pH 1); (2) fulvic acid, which is the

humic fraction that remains soluble in both alkaline and acidic solu-

tions; and (3) humin, the fraction that cannot be extracted by acid or

base (Kononova 1966). Although this fractionation scheme is arbitrary

and the fractions are molecularly heterogeneous, Schnitzer and Khan

(1972) have reported some properties that distinguish fulvic and humic

acids. Fulvic acid has lower color, carbon content, and molecular

weight and higher CEC, oxygen content, and density of functional groups

than does humic acid. Therefore, due to the more soluble nature of

fulvic acid, the humic substances in natural water are predominantly

the fulvic acid fraction.

The structure and function of humic substances have been the

topics of much research, and because several excellent reviews are

available (Faust and Hunter 1971; Schnitzer and Khan 1972; Povoledo and

Golterman 1973; Gjessing 1976; Carriker 1977; Miles 1979; Hartenstein








1981), only a brief discussion of the chemical nature and dynamic role

of humic substances in natural waters is presented here.

No one has satisfactorily characterized humic substances, and no

consensus exists regarding the structure of these compounds; however,

several hypothetical structures have been proposed that account for

their elemental composition, degradation products, acidic properties,

and complexation capacities (Christman and Ghassemi 1966; Schnitzer

1971; Stevenson and Ardakani 1972; Gamble and Schnitzer 1973). The

structures consist mainly of complex hydrophilic polyelectrolytic poly-

mers of benzene rings that are variously substituted with phenolic,

carboxylic, or methoxy groups, together with large amounts of aliphatic

carboxylic acids. Two hypothetical structures of humic materials--one

that is loosely held together by hydrogen bonds and another that is

more rigidly bonded-are illustrated in Figure II-1. The elemental

composition varies among sources and analytical techniques, but typical

ranges are: C, 45-63%; 0, 30-35%; H, 3-6%; N, 0.5-5% (Gjessing 1976;

Schnitzer and Khan 1972). Molecular weights of these polymers have

been reported to range from a few hundred to a few hundred thousand,

although recently a consensus has been growing that aquatic humic sub-

stances are approximately 1000 daltons (Wilson and Weber 1977; Reuter

and Perdue 1981; Schnitzer 1981). These workers based their conclu-

sions on colligative property techniques such as vapor phase osmometry

and cryoscopy. Interestingly, Shapiro (1964) concluded similarly that

aquatic humic material was of low molecular weight (using cryoscopy),

but because several others reported much higher molecular weights for

humics (Black and Christman 1963; Gjessing and Lee 1967; Ghassemi and

Christman 1968; Gjessing 1971), Shapiro's results were disbelieved











0 ,OH
O OH OH 0 OH C' 0
II I 11 I II
HO-C q OH------------O=C C-OH----O=C C-OH

HO C-OH ---------- O=C OH-------- O=C C-OH
I/ I I II
OH O OH C OH OH
O'OH


0 OcH 6OH OH 0OH

HO-C OH H
0 *O00 0 _Q
HO-C C=O--------HO-C C=OH-------HO'OH OH
0, C. OH TO OH OH
HO O=C C= O
OH C OH
HO 0




COOH
QOH

0 N
COOH =0 Q H
(CS t (C) ( C) n(C)

OCH3 CH300 OCH3 -CH3 H
OH OH OH 0

OCH3
o H O0
f-o 0 H :I- 0-(C)R- (C)--
-0 O O H H O O H
OH 0 0 0
0 -N

Figure II-1. Proposed structure of humic substances from
Schnitzer (1971) (top) and Christman and Ghassemi
(1966) (bottom).








until recently. The high molecular weights were determined using gel

chromatography, ultrafiltration, and dialysis techniques, which can

cause aberrant results due to adsorption and charge repulsion of the

organic solute with the membranes and gel beads (Tuschall and Brezonik

1980; Truitt and Weber 1981b).

Heavy metal complexation with humic materials has been reported by

many researchers, and several reviews are available (Singer 1973; Rubin

1974; Reuter and Perdue 1977). Of the metals commonly found or

expected in a wetland environment, copper and iron typically form the

strongest complexes with humics; manganese and zinc form the weakest

complexes; and lead, nickel, and cadmium form complexes of intermediate

stability. The conditional stability constants, 8' (defined in

Section II-D), vary greatly due to both the variability of the methods

used to determine 8' and the variability of the conditions of analysis,

especially pH. Nevertheless, the overall implications of past research

are that humic materials can complex heavy metals to an appreciable

extent, thus altering the toxicity, solubility, and ultimately the fate

of heavy metals in the environment.



D. Methods for Quantifying Metal-Organic Matter Interactions


The most commonly employed parameters for quantifying the extent

and stability of complexation between heavy metals and naturally occur-

ring organic matter are available metal binding ability of the organic

matter, also called completing capacity (CL), and the conditional

stability constant (9'). Available metal binding ability is defined as

the quantity of a metal ion titrant that is completed by the ligands in







a water sample and is usually expressed in units of moles/L (Allen et

al. 1970; Chau 1973; Hanck and Dillard 1977b). Typically, CL is

determined by titrating a sample with the metal of interest and,

following each addition of titrant, measuring the concentration of com-

plexed metal (e.g., ion-exchange technique [Crosser and Allen 1976]),

the concentration of uncomplexed or weakly completed metal (e.g.,

ion-selective electrode [Buffle et al. 1977], anodic stripping voltam-

metry [ASV] [Shuman and Woodward 1973, 1977], equilibrium dialysis

[Zunino and Martin 1977; Truitt and Weber 1981b]), or the concentration

of uncomplexed ligand (e.g., fluorescence quenching [Saar and Weber

1980]). The point at which the ligand can no longer complex additional

metal ion is the available metal binding ability (or complexation capa-

city) of the solution, which can be determined graphically.

In describing the quantitative stability of a complex for the

equilibrium condition

aM + bL = MaLb (2.1)
the thermodynamic stability constant, e, is defined:

(MaLb)
S -- (2.2)
(M)a(L)b
where (M) is the activity of free metal ion, (L) is the activity of the

reactive species of ligand, and (MaLb) is the activity of the com-

plex. The value of B is a true constant, although the value alone says

nothing about the distribution of M and L under existing solution con-

ditions such as pH or concentrations of other competing species, nor

does reveal anything about the distribution of M and L in reference








to their total concentrations, since the metal and ligand can exist in
forms other than M+b, L-a, and MaLb (other forms include

MOH, MCO3, HL, H2L, etc.). Furthermore, B is extremely difficult

(or impossible) to determine accurately for environmental samples,

which contain a heterogeneous mixture of ligands, since the various

side reactions and their attendant constants are seldom known and are

exceedingly difficult to obtain. For these reasons, another parameter,

the conditional stability constant, B', was proposed by Ringbom (1963)

for the equilibrium condition aM + bL = MaLb. The conditional

stability constant, B', is defined by Ringbom (1963) as follows:

(NLab)
B' (2.3)
(M')a(L')b

where (M') is the concentration of metal not completed with L, and (L')

is the concentration of L not completed with M. By definition, B' is

not a constant, but instead is a function of the solution conditions

and varies with concentrations of competing species such as hydrogen

ion, calcium, and carbonate, and also with metal and ligand concentra-

tions (Ringbom 1963; Mantoura and Riley 1975; Bresnahan et al. 1978;

Buffle 1980; Gamble et al. 1980). Nevertheless, B' is a useful param-

eter for predicting the relative distribution of metal for specific

environmental conditions.

Numerous methods exist for determining the available metal binding

capacity (CL) of naturally occurring organic matter and conditional

stability constants (B') for the resulting complexes. In the following

section, the most commonly used procedures are described briefly and

the utility of each method for naturally occurring organic matter is








discussed. These methods can be grouped into major classes as

follows.


1. Physical Separation Techniques

The gel permeation chromatography method described by Mantoura and

Riley (1975) involves a dynamic equilibrium between organic matter that

is injected onto a chromatographic column of porous gel beads and metal

ions in the elution buffer. The complexes formed between the macromol-

ecular organic matter and the metal ions are excluded from the gel

beads and travel down the column more rapidly than the free metal ions,

which permeate the gel. A plot of metal concentration versus volume of

eluant (Figure 11-2) shows a peak due to completed metal (peak A) at

the same volume in which the organic ligand eluted, and a negative peak

(metal deficiency) at the volume corresponding to elution of low

molecular weight solutes (peak B). The areas of the two peaks deter-

mine the amount of metal completed, and knowledge of the initial ligand

concentration and the background metal concentration permits calcula-

tion of conditional stability constants. For a 1:1 complex (i.e., ML),

8' is calculated as follows:

(ML) (Mbound)
'= ----- = (2.4)
(M')(L') (Mf)(Lt-Mbound)
where ML = Mbound = area of peak A (Figure 11-2), Mf is the un-

complexed metal concentration (i.e., metal concentration in eluant),

and Lt is the total ligand concentration. Multiple chromatographic

runs at varying metal concentrations can yield information about dif-

ferent metal binding sites on one ligand or sequential complexation by

a mixture of ligands.























o COMPLETED
METALS


z
O
C.)
-J


\ METAL

UJ ORGANIC DEFICIENCY
U
SMATTER B

O

<



0 Vo Vt
ELUANT VOLUME


Figure 11-2. Theoretical gel filtration chromatogram for the interaction
of organic matter with metal ions.








Unfortunately, this method is limited to only those organic

ligands that are excluded by the gel beads. The minimum exclusion

limit for Sephadex brand gels (G-10) is 700 daltons, and for Biogel

brand gels (P-2) is 1800 daltons. Solute-gel charge interactions are

significant, and adsorption of organic matter to the gel (especially

humic substances at low ionic strength and pH) often necessitates con-

ditions of analysis that are not typical of the natural system (i.e.,

1 M ionic strength and pH 8.0) (Klein 1976; Miles 1979). However, this

method is applicable to a variety of metals (with some exceptions such

as Fe). Mantoura and Riley (1975) and Mantoura et al. (1978) reported

conditional stability constants with Cd, Co, Cu, Hg, Mn, Ni, and Zn

using the gel filtration chromatographic technique.

Membrane separation is another method available to measure CL

and s' of metal-organic complexes. The two types of membranes most

commonly used for studying metal-organic matter interactions are dialy-

sis tubing and ultrafiltration membranes. For the dialysis method, a

sample is placed inside semipermeable cellulose-based tubing, which

subsequently is suspended in a large reservoir of buffered water.

Small ions and compounds diffuse through the membrane until equilibrium

is reached, whereas large organic molecules (and heavy metals completed

to them) are prevented from passing through the tubing. Analysis of

the two solutions for total metal concentrations at equilibrium indi-

cates the quantity of metal bound to the retained organic matter. The

initial concentrations of metal and organic matter can be varied, and a

complete binding analysis can be performed. This method has been used

to study protein-metal interactions (Klotz et al. 1946; Gurd and Good-

man 1952; Malmstrom 1953; Osterberger 1973) and metal binding by humic








substances (Benes et al. 1976; Guy and Chakrabarti 1976; Zunino and

Martin 1977; Truitt and Weber 1981b). Unfortunately, the dialysis pro-

cedure suffers from several important problems. Metals and organic

matter may adsorb to the membrane; the membranes are permeable to some

environmentally important organic compounds; and the necessity to

repeat the experiment at several different metal or ligand concentra-

tions makes the procedure tedious and time-consuming (Ramamoorthy and

Kushner 1975; Zunino and Martin 1977; Truitt and Weber 1981b). On the

other hand, the dialysis procedure works well at pH and ligand concen-

trations typically found in natural waters, and most metals can be

speciated by dialysis.

The other common membrane-separation technique is ultrafiltration.

In this procedure, a membrane composed of a non-cellulosic polymer lies

on the bottom of a pressurized, stirred cell that contains a solution

of the metal and ligand of interest. Pressure forces water containing

small ions and compounds to pass through the membrane, whereas large

compounds are retained. Analysis of the retained and filtered frac-

tions indicates the extent of metal binding with high molecular-weight

organic matter. Ultrafiltration is subject to the same advantages and

disadvantages as the dialysis method, although the former method has

the added advantage of controlled and variable rate of separation

because the system is pressurized.

Most of the ultrafiltration studies reported to date were designed

to speciate ambient levels of metals, and therefore filtrations were

performed at a single metal and ligand concentration (Gjessing 1970;

Schindler et al. 1972; Benes and Steinnes 1974; Guy and Chakrabarti

1976; Giesy and Briese 1979; Hoffman et al. 1981). Titrations of








organic matter with heavy metals have been avoided, apparently due to

the tedium involved in performing individual experiments for each addi-

tion of metal. Ultrafiltration also has been used simply to fraction-

ate organic matter into different size ranges, and other metal-specia-

tion techniques were used to investigate interactions of heavy metals

with the fractionated organic matter (Allen 1976; Smith 1976; Hoffman

et al. 1981).


2. Potentiometric Titrations

Anodic stripping voltammetry (ASV) has been used frequently in

studying metal-organic matter interactions in natural waters, and the

method of complexometric titration has been developed to determine the

available metal binding capacity (CL) and conditional stability con-

stant (') (Matson 1968; Shuman and Woodward 1973, 1977; Chau and Lum-

Shue-Chan 1974). The method involves titrating water samples that con-

tain excess ligand with a heavy metal and analyzing the "uncomplexed"

metal by ASV. The endpoint of such a titration is interpreted as a

measure of the completing capacity (CL) of a water sample. The

original procedure was adapted by Shuman and Woodward (1973, 1977) to

include the direct determination of a conditional stability constant

(8') for metal-ligand complexes that are not reduced under the condi-

tions of metal ion analysis.

The ASV analysis consists of two steps: an initial plating step

and a subsequent stripping step. In the former step, a potential more

negative than the reduction potential of the analyte metal ion is

applied to the solution, thus reducing the metal to an uncharged state.








The elemental metal plates onto the electrode, which is typically mer-

cury, and thus the analyte is concentrated during the plating step:
+N
Maq + 2e" + Hg -- > MHg. (2.5)

Subsequently, the electrode potential is increased linearly with time

in the stripping step, causing the metals that have been concentrated

on the electrode to be oxidized and to flow back into solution. Oxida-

tion occurs at a characteristic potential (Ep) for each metal; con-

current with oxidation is a release of electrons. The electron flow is

measured as current (i) during the scanning step. A typical ASV scan

of a 10 pM copper solution is illustrated in Figure 11-3. If all other

variables such as stirring rate, deposition time and voltage, scan

rate, and electrode surface area remain constant, the measured peak

current (ip) is directly proportional to the concentration of reduc-

ible metal originally in solution. Hence, a plot of peak current

(ip) versus metal added produces a titration curve from which the

values of CL and ' can be obtained. The theoretical considerations

of calculating CL and p' are discussed further in Section III-D.

The ASV titration method is a simple technique that can be used at

metal, ligand, and pH levels found in natural waters. However, the

authors who developed the method stated the following underlying

assumptions: (1) ip is proportional to free and labile metal;

(2) the ML complex is non-reducible; (3) the ML complex does not

kinetically dissociate appreciably during the plating step (Shuman and

Woodward 1977). Although Shuman and Woodward (1977) stated that these

assumptions appear reasonable for natural water samples, the validity

of the assumptions has not been demonstrated clearly. There is reason

to believe that the dissociation or direct reduction of metal-ligand





































I _I I I I I I


0.2


25





20


-1
m

15
o


z


-0.4 -0.3 -0.2 -0.1 0.0 0.1
POTENTIAL (VOLTS)

Figure 11-3. Typical ASV scan of a sample containing 10 pM copper.









complexes can cause low results. For example, Hanck and Dillard

(1977b) reported up to 34% relative error in determining the complexa-

tion capacities of ethylenedinitrilotetraacetic acid (EDTA) solutions

using indium as the titrant, and they assumed that complex dissociation

caused the low results. Nonetheless, copper is commonly used to

titrate natural water samples (Chau and Lum-Shue-Chan 1974; O'Shea and

Mancy 1976; Shuman and Woodward 1977; Sugai and Healy 1978; Tuschall

and Brezonik 1980) because it forms more stable complexes with most

organic ligands than do most other environmentally significant metals

such as zinc and cadmium (Irving and Williams 1948). However, the

extent of complex dissociation or reduction remains a potential draw-

back of using copper as titrant.

Ion-selective electrodes have been used widely in characterizing

interactions of heavy metals with naturally occurring organic matter

(Stiff 1971a, 1971b; Gardiner 1974; Buffle et al. 1977; Bresnahan et

al. 1978; Swallow et al. 1978; Giesy and Briese 1979; McKnight and

Morel 1979; Saar and Weber 1980). Presently, the only heavy metals for

which ion-selective electrodes are commercially available are Cu+2,
Cd+2, Pb+2, Hg+2, and Ag+1. These electrodes respond to the aquated

ion only and not to any complex forms of metal.

The electrode response is described by the Nernst equation:

E = E + 2.303 R log aRT (2.6)
nF log aMn (2.6)


where E is the standard electrode potential for the metal being mea-

sured, a N1+n is the activity of metal ion, and 2.303 RT/nF =

29.6 mV at 25C for a divalent ion (n = 2) and 59.2 mV for a monovalent

ion (n = 1) (Ross 1969). Hence, for each 10-fold change in activity,









the electrode response (E) should change by 29.6 mV for Cu+2,

Hg+2, Cd+2, Pb+2 and 59.2 mV for Ag+1.

Ion-selective electrodes allow direct measurement of heavy metal

ion activity at pH and ionic strength levels found in most natural

waters. However, a minimum total heavy-metal concentration is required

to obtain accurate potentials for heavy-metal ion activity; thus

typical environmental levels of metals cannot be detected using

ion-selective electrodes. For instance, the limit of detection of the

Cu+2-ISE is about 10-7 M Cu+2 (6.5 ug Cu+2/L) for a solution

containing aquated copper (Cu+2) as the only form of copper. How-

ever, for a solution with completed copper in addition to aquated cop-

per (Cu+2), the ISE will measure levels of Cu as low as 10-10 M

if the total concentration of copper is 10-7 M or more (McKnight

and Morel 1979). For this reason, ion-selective electrodes are best

suited for laboratory studies involving sample titrations with heavy

metals. Disadvantages of using ion-selective electrodes include

lengthy calibration procedures that need to be performed frequently due

to drift, and broad titration curves that make CL difficult to deter-

mine accurately.


3. Ion Exchange

Shubert (1948) developed the ion-exchange technique to determine

CL and 8' for a metal-organic matter system, and, subsequently, the

method of calculating s' has been revised and adapted by Zunino et al.

(1972), Crosser and Allen (1976), and Stevenson and Ardakani (1972).

Basically, the ion-exchange procedure is performed by adding a strong

cation exchange resin (e.g., Dowex 50) to a series of vessels








containing the ligand of interest at several concentrations of metal.

Flasks are mixed for a specified time (e.g., 24 h), and the super-

natants are analyzed for total soluble metal by atomic absorption spec-

troscopy. The amount of metal taken up by the resin is the difference

between the amount added initially and the amount remaining in solution

at equilibrium. The soluble metal concentration is related to the

extent of metal complexation: that is, completed metal that does not

adsorb to the resin will increase the concentration of soluble metal

compared to that in the absence of ligand. Theoretical plots of metal

bound by the resin versus total soluble metal for three possible cases

are illustrated in Figure 11-4. The curves vary according to the sta-

bilities of the metal-ligand complexes formed, the ligand concentra-

tion, and the number of ligands present.

The principal advantage of the procedure is that any metal that

can be analyzed by atomic absorption spectroscopy can be used, unlike

methods using ASV or ion-selective electrodes, in which only a few

metals can be analyzed. However, the ion-exchange procedure does not

work well with weak ligands because partitioning of metal between solu-

tion and resin strongly favors the resin, and hence only a small pro-

portion of metal remains in solution. Thus for ligands that form weak

complexes, a weak cation exchange resin such as manganese dioxide is

preferable, although preparation of a reproducible suspension of MnO2

is difficult and highly dependent on technique (van den Berg and Kramer

1979). Another serious problem with the ion-exchange technique is

adsorption of ligand to the surface of the resin, thus reducing both

the extent of metal exchange between solution and resin and the soluble

ligand concentration.























Z /
z

A A C B










0
00









TOTAL SOLUBLE METAL
COMPLETEDD & FREE)

Figure II-4. Theoretical curves for ion-exchange equil-
ibrium method (A: no ligand present;
B: strong ligand; C: mixture of two
ligands, both weaker than ligand in B).
m
-J//
I-y7















ligands, both weaker than Uigand in B).








4. Spectrophotometric Methods

The fluorescent properties of naturally occurring organic matter

have been used to quantify metal-organic interactions (Levesque 1972;

Saar and Weber 1980). Fluorescence of organic matter is diminished

(quenched) by binding with paramagnetic metal ions such as Cu+2,
Co+2, and Ni+2 (i.e., ions with one or more unpaired elec-

trons), and thus the intensity of fluorescence can be used as a measure

of ligand not bound to paramagnetic metal ions. Metal binding ability

is determined simply by titrating organic matter with metal and record-

ing fluorescence after each addition. Although the technique is simple

to perform and sensitive to low levels of ligand, only fluorescent

ligands and paramagnetic metal ions can be analyzed by this method.

Job (1928) introduced the method of continuous variation for

determining CL and 8'. In this method, differential spectroscopy is

used to measure the absorbance of solutions containing various levels

of metal and ligand. One condition necessary to perform this analysis

is that the ML complex absorbs light at a characteristic wavelength,

and, ideally, the other metal and ligand species do not absorb appreci-

ably at the analytical wavelength. Background absorbance is usually

subtracted optically by placing the sample without added metal in the

reference path, and the sample with added metal in the sample path

(i.e., differential spectroscopy).

By mixing V volumes of ligand (concentration Lo) with 1-V

volumes of metal (concentration Mo = Lo/r), where V varies from 0

to 1, the concentration of the MLx complex (i.e., A absorbance) will

be maximal when

B' Mo rx-1 [(x+r)Vmax-x]x+1 = (r-l)x [x-(x+)Vmaxl]. (2.7)








Thus a plot of A absorbance versus the fractional volumes of metal and

ligand (Figure 11-5) is used to determine Vmax, and ultimately 8'

using equation 7. However, the method of continuous variation is

usually not sensitive to low levels of metal and ligand, and it suffers

from the underlying assumption that only one complex is formed through-

out the titration.

The concept of differential spectroscopy has been applied further

to the determination of s' of pure compounds by monitoring the concen-

tration of metal completed with a competing ligand (Rossotti and

Rossotti 1961). In this procedure, a competing ligand (A) that has a

known stability constant with the titrant metal is added to a solution

of the metal (M) and ligand (L) of interest, and the following

equilibria are established:

M + L = ML (2.8)

M + A = MA. (2.9)

By measuring the concentration of MA (using differential spectroscopy)

and knowing the analytical concentrations of M, L, and A, the

conditional stability constant of the ML complex can be calculated.

The competing-ligand method using differential spectroscopy has

been applied to environmental samples by Anderegg et al. (1963) and

Harris et al. (1979), who used EDTA as the competing ligand to deter-

mine the conditional stability constants for iron with several

microbially produced siderophores. Campbell et al. (1977) reported a

competing-ligand procedure to study binding of zinc by natural ligands

in river water. The competing ligand was a zinc-specific ligand called

zincon (2-carboxy-2'-hydroxy-5'-sulfoformazylbenzene), which forms a

red complex with zinc. The concentration of the zinc-zincon complex,




















LU







< I I



0 2 0.4 0.6 0.8 1.0
V
[ORGANIC LIGAND] -
[METAL]

Figure 11-5. Theoretical curve for method of continuous
variations.








monitored in the presence of naturally occurring organic matter using

differential spectroscopy, was used to determine the complexation

capacity (CL) of the river water samples.

Unfortunately, many of the strong absorption bands for metal-

ligand complexes are in the 200-300-nm range (Martin 1974), which is

the same region that natural organic matter absorbs strongly (Schnitzer

and Khan 1972). Although the background absorbance can be subtracted

optically, the accuracy of measuring a small MA peak in the presence of

a strong absorption band is diminished. Nevertheless, if an appropri-

ate competing ligand is used, the method has the advantages of being

simple to perform, sensitive to low levels of metal and ligands, and

often applicable to more than one metal (Rossotti and Rossotti

1961).


5. Miscellaneous

In addition to the methods discussed in the preceding section,

several other methods for quantifying metal-organic interactions have

been proposed, although their use for natural water studies has been

slight. These include a copper solubilization method (Kunkel and

Manahan 1973), a potentiometric titration using a pH electrode (Steven-

son 1977), and a Co(III) complexation technique (Hanck and Dillard

1977a). Apparently procedural limitations (such as low sensitivity,

restricted sample pH, and excessive sample manipulation) have prevented

these methods from receiving widespread use.

As the above review indicates, none of the currently available

methods for quantifying metal-organic interactions is problem-free, and

many suffer from serious limitations. The need to critically evaluate








the accuracy of existing methods and to develop new and better methods

is apparent if metal-organic interactions in natural waters are to be

characterized accurately.



E. Methods of Fractionating Organic Matter


Interactions between heavy metals and naturally occurring

organic matter have been studied by numerous researchers, and typically

these studies include determination of conditional stability constants

for the metal-organic complex (Mantoura and Riley 1975; Guy and

Chakrabarti 1976; Buffle et al. 1977; Bresnahan et al. 1978; Davis and

Leckie 1978). Although several different methods have been used to

determine stability constants for copper-organic complexes, one

interesting trend emerged when Scatchard plots were used to calculate

the constant: namely, that two sites of metal bonding were observed,

one strong and one weak (Mantoura and Riley 1975; Guy and Chakrabarti

1976; Bresnahan et al. 1978; Giesy 1978).

A priori, one would expect the strong site to be very important in

heavy metal partitioning. Fractionation of the organic matter prior to

measuring metal binding ability may determine if the strong and weak

binding sites exist as separate entities.

Rebhun and Manka (1971) developed a fractionation procedure that

involved a series of chemical precipitations and solvent extractions to

fractionate the organic in secondary sewage. Their scheme was used to

isolate anionic surfactants, carbohydrates, tannins, proteins, fulvic

acids, humic acids, hymatomelanic acids, and ether extractables.

However, chemical precipitation and solvent extraction techniques








suffer from the problems of contamination and poor recoveries, and

extreme care must be exercised to minimize these problems.

Sirotkina et al. (1974) fractionated dissolved organic matter from

the Moscow River by using columns of ion-exchange celluloses followed

by further separation on gel permeation bead columns. Their scheme,

illustrated in Figure 11-6, involved application of a sample to an

anionic exchange column (DEAE cellulose), onto which the anionic organ-

ics (acidic fraction) were adsorbed and retained. The cationic and

neutral organic matter in the elute was subsequently applied to a

cation exchange column (CM cellulose), which retained the cationic

organic matter. The adsorbed compounds were released from the columns

by rinsing with 0.1 N NaOH (DEAE column) or 0.1 N HCl (CM column). The

desorbed organic matter and the neutral fraction were separated further

using a column of Sephadex gel. The authors claimed better than 90%

recoveries of organic matter, which is superior to most reported recov-

eries for similar techniques using other ion exchange resins (i.e.,

divinyl benzene or styrene polymer supports). These resins typically

adsorb organic matter irreversibly (Sirotkina et al. 1974; Leenheer

1981).

An approach similar to that described by Sirotkina et al. (1974)

was used by Tuschall and Brezonik (1980) to isolate the proteinaceous

fraction of organic matter from water. They applied an acidified

sample to a strong cation exchange column (P cellulose), desorbed the

retained organic matter with NaOH, and desalted the isolated fraction

with a Sephadex column. Recoveries of organic matter from the column

were greater than 90%.








Natural water
concentrate


DEAE cellulose
column Filtrate


0.1 N
NaOH
desorption


Acidic group


Sephadex
column


CM cellulose o-- Neutral
column Filtrate group


0.1 N
HC1
desorption


Basic group


Sephadex
column


Sephadex
column


High
Molecular
Weight
Fraction
(fulvic acids;
polyphenols)


Low
Molecular
Weight
Fraction
phenolss;
carboxylic acids)


High
Molecular
Weight
Fraction
(proteins)


Low
Molecular
Weight
Fraction
(amino acids)


High
Molecular
Weight
Fraction
(polysaccharides)


Figure 11-6. Scheme for fractionation of dissolved organic matter (Sirotkina et al. 1974).


Low
Molecular
Weight
Fraction
(sugars)





37

Leenheer (1981) used nonionic and ion-exchange resins to fraction-

ate organic matter into hydrophobic (base, acid, and neutral fractions)

and hydrophilic (base, acid, and neutral fractions) classes. The

author claims to have overcome the problems of poor recoveries reported

by others when using ion-exchange resins. An extensive cleanup of

resins and a detailed desorption procedure were used to recover at

least 80% of the adsorbed material.














CHAPTER III
AREAS OF INVESTIGATION, SAMPLING PROCEDURES,
AND ANALYTICAL METHODS



A. Jasper Study


1. Site Description

Basin Swamp, located near Jasper, Florida, was selected as an area

to study the efficiency and effects of using cypress strands for ter-

tiary treatment of wastewater during the 2-year period January 1979

through December 1980. This swamp has been receiving wastewater efflu-

ent from the City of Jasper since 1914. Raw sewage was released into

the swamp from 1914 to 1951; primary treated effluent was released from

1952 to 1972; and secondary effluent has been released since then.

Typically, the plant discharges 1150 m3/day (0.3 MGD) into the swamp

except during periods of heavy rainfall, when the rate increases to

2200 m3/day (0.6 MGD).

The wastewater that leaves the Jasper sewage treatment plant flows

first into an area that is now a marsh (Upper Basin Swarp; Figure

III-I). A railroad embankment divides this area into two sections

(East Swamp and West Swamp), connected by two culverts. The West

Swamp, which receives the effluent directly, is dominated at the north-

ern end by herbaceous plants such as elephant ear (Colocasia

escylentum) and pennywort (Hydrocotyle umbellata). Bushes and small

trees (e.g., red maple [Acer rubrum]) are more common in the middle and












SCALE Drainage
SCALE Ditch
1500 750 0 1500 3000 FEET Sewage
STreatment
2 (Effluent) Plant
iger Creek (













1- 4- Ba sBasin n railroad
SRoad Seaboard


Figure III-1. Surface water sampling stations at study site near Jasper, Florida.









southern end of the marsh, and cattails (Typha latifolia) line both the

east and west boundaries. Unconsolidated sediments are more than a

meter deep in several places.

Surface water from Upper Basin Swamp channelizes and converges

with Bell Creek, which flows into Lower Basin Swamp (Figure III-I).

Vegetation in Lower Basin Swamp is similar to that described for Upper

Basin Swamp, although more cypress trees (Taxodium distichum) are pres-

ent in the downstream swamp than in the upper area. Further down-

stream, channelized flow continues (Tiger Creek) and surface water flow

terminates at a sinkhole (Tiger Sink).


2. Sampling and Analytical Methods

Samples for heavy metals were collected at several points through-

out the swamp and along the channelized flow downstream, including the

sinkhole into which Tiger Creek flows (Figure III-1). The heavy metal

content in surface runoff was measured at the drainage ditch from the

City of Jasper (station 1), Bell Creek (station 8), and a tributary of

Tiger Creek (station 15). In addition, samples were collected at both

the influent and effluent of the sewage treatment plant. Groundwater

quality was monitored at the wells shown in Figure 111-2.

Surface water samples were collected directly (by hand) in

acid-washed plastic sample bottles, and groundwater samples were

collected with a hand vacuum pump by drawing water through Tygon tubing

into a sample bottle. Prior to sampling the wells, the water in the

well pipe was pumped out and discarded so that fresh percolate could be

collected. All samples for heavy-metals analysis were filtered through

a 1-um polycarbonate filter and preserved with high purity nitric acid
























"W 0 CYPRESS TREES

Figure 111-2. Locations of groundwater wells at Basin Swamp study site near Jasper, Florida.








(Ultrex) at the rate of 5 mL/L of sample. Cadmium, copper, lead, and

nickel were analyzed using a flameless graphite-furnace technique with

an atomic absorption spectrophotometer (AAS), whereas manganese and

zinc were analyzed by direct aspiration in the flame mode of the AAS.

Instrument settings were those recommended by the manufacturer (Varian

1972).



B. Waldo Study

1. Site Description

The City of Waldo, Florida (population 1OOQ), has been discharg-

ing wastewater into a series of cypress wetlands for over 45 years.

During that time, the sewage from Waldo has been treated only by a con-

crete septic tank, and, at most, primary treatment was effected. The

septic tank effluent (septage) flowed through a ditch and into a 2.6-ha

cypress wetland on the eastern side of a railroad embankment (Figure

111-3). After traveling approximately 1.6 km, the surface water flowed

through a culvert under the railroad tracks and into a large (2500 ha)

cypress swamp, which eventually drains into Lake Altho and ultimately

into the Santa Fe River.

The swamp forest is dominated by an even-age stand of pondcypress

(Taxodium distichum var. nutans), with a subcanopy of swamp black gum

(Iyssa sylvatica var. biflora) and red maple (Acer rubrum) (Nessel

1978). The understory includes a variety of shrubs, vines, and herbs.

For average water conditions, the surface water was approximately

0.5-1.0 m in depth. A dense mat of duckweed (Lemna sp.) covered most













40 m


CYPRESS STRAND


Figure 111-3. Map of experimental site at Waldo, Florida.








of the surface water during the warm months, and waterhyacinths (Eich-

hornia crassipes) grew luxuriantly in some areas.

In order to confine a portion of swamp so that heavy metals could

be added and monitored quantitatively, two corridors were constructed

in the downstream area of the wetland. This area was thought to be

impacted less by antecedant discharge of wastewater than the upstream

area, which was closer to the source (Figure III-3). Septage was

pumped through plastic pipes to the experimental plots where the waste-

water flow was bifurcated, with one plot receiving septage spiked with

heavy metals (plot M) and the other plot receiving septage only (plot

C). In addition, a reference station was established outside the

experimental plots (station R).

Wastewater was pumped at a flow rate between 10 and 40 L/min in

each plot, which produced a hydraulic retention time in the plots of

approximately 3-5 days. Metals were added using a constant-pressure

Mariotte bottle such that the final metal concentration in the waste-

water approximated the levels reported for municipal wastewater in New

York and Los Angeles (Chen et al. 1974; Klein et al. 1974; see Table

II-1), namely: Cd, 10 ig/L; Cu, 50 ug/L; Mn, 100 wg/L; and Zn, 100

ug/L. The chloride salts of each metal were used to make the stock

solution.


2. Sampling and Analytical Methods

Surface water samples were collected for metal analyses from

March 3, 1981, to April 1, 1981. Subsequent sampling was prevented by

severe drought because the amounts of both swamp water and septage were

insufficient to continue the experiment. Sampling stations were








located at the sewage discharge pipe and at 10, 20, 30, and 40 m from

the discharge point in both the corridor receiving sewage spiked with

heavy metals and at the corridor receiving sewage only. Triplicate

samples were taken at each station, and in most cases the samples were

composite. Samples were preserved with nitric acid (U.S. EPA 1979)

and autoclaved (1210C, 2 atm) for 1 h to release colloidal and particu-

late-bound metals. Copper, zinc, and manganese were analyzed using a

Perkin-Elmer model 5000 AAS in the flame mode, and cadmium and some

low-level samples for copper were analyzed by the graphite-furnace

technique with an AAS. Instrument settings were those recommended by

the manufacturer (Perkin-Elmer 1980).



C. Microcosm Study


Twenty microcosms were constructed by adding 800 g of wet peat,

a layer of partially decomposed leaves (litter), and 3 L of filtered

water to a 4-L jar in a manner simulating the structure of the swamp

ecosystem at Waldo (Figure 111-4). The peat, litter, and surface water

were collected from an undisturbed area in Waldo Swamp and kept at 4C

until needed. After the microcosms were constructed they were purged

with a slow stream of nitrogen for 1 month to allow the sediments and

water to reach steady state. Several distinct horizontal layers in the

sediment were visible through the plastic containers, and apparently

the vertical structure of the sediments had been re-established.

Initially, the intent was to vary redox potential (Eh), dis-

solved organic carbon (DOC), pH, calcium, and iron levels in the micro-

cosms and to observe the correlation of these factors with metal loss.




















Sealed Top





4L Plastic Jor










-Leaves


Figure 111-4. Illustration of microcosm.








However, Eh was not easily regulated. Control of Eh was attempted

by allowing conditions in the microcosms to become reduced and subse-

quently purging the system with a variable amount of oxygen. This

method produced inconsistent levels of Eh, and therefore another

approach was taken. The Eh was controlled externally by purging the

microcosms with various levels of hydrogen sulfide (H2S). The three

levels of H2S used in the study were <0.1, 0.5, and 5.0 mg H2S/L

(Table III-1). The lowest value was achieved by purging the systems

with nitrogen and enough oxygen to maintain oxidizing conditions ('t1-2

mg 02/L), which were similar to field conditions in Waldo Swamp dur-

ing the winter months. The other two levels were maintained by

bubbling a mixture of nitrogen and H2S into the water of the micro-

cosms. Levels of Eh were measured during the preliminary experimen-

tal work, and at H2S concentrations of <0.1, 0.5, and 5.0, Eh was

+480, -100, and -100 mV, respectively. Although the levels of H2S

and Eh are interdependent, the level of H2S was varied in the

experiment, and hence the factor H2S is used in subsequent discus-

sions.

The five factors examined in this study--DOC, iron, calcium, sul-

fide, and pH--were tested at two levels such that all possible combina-

tions of the two levels were made (Table III-1). This factorial design

experiment had 32 (i.e., 25) different treatments. The intermediate

levels were performed in replicate so the variability of the experiment

could be determined. The experiment was conducted in two stages using

20 microcosms for each stage. After the microcosms were used once, the

water was acidified to pH 2 to desorb metals from surface sites, and

the acidified water was replaced with fresh surface water. In addi-








Table III-1. Factors and levels for factorial microcosm experiment.



Levels

High Low Intermediate
Factors + 0


DOC (mg/L) 50 17 33

pH 7.0 4.0 5.5

Calcium (mg/L) 40 4.0 22

Iron (mg/L) 5.0 0.3 2.5

Sulfide (mg/L) 5.0 <0.1 0.5








tion, several systems containing only surface water were examined con-

currently with the microcosms (which contained both water and peat

sediments).

Calcium and iron levels in the microcosms were increased by the

addition of the chloride salts. DOC was reduced from the original

level of 50 mg/L by passing the water through an exchange column of

DEAE-cellulose and mixing with untreated water to obtain the desired

level of DOC. The pH was adjusted with HC1 or NaOH, and sulfide was

increased by continuously bubbling H2S into the containers. Sulfide

concentrations were measured using the fluorescein procedure described

by Natusch et al. (1972). The water in the microcosm was gently circu-

lated by a constant stream of nitrogen (2 L/h).

After the conditions were set, the microcosms were spiked with

cadmium, copper, zinc, and manganese at levels of 10, 50, 100, and 100

pg/L, respectively. The overlying water was sampled at 24-h intervals,

and the levels of metal remaining in solution were determined by atomic

absorption spectroscopy. The treatment parameters DOC, iron, and cal-

cium were adjusted at the beginning of each experiment only, whereas

the sulfide and pH levels were maintained at constant values throughout

the experiment.



D. Analytical Methods for Determining CL and Q'


This section describes the instrumentation and methodology used

to evaluate and compare five methods for determining CL and e'.

Three of the methods, ASV, ISE, and fluorescence, are published, and

two have received widespread use. The fluorescence method was pub-








lished only recently. Titrations using these three methods were per-

formed according to each published procedure. The other two proced-

ures, ultrafiltration and the competing ligand method, were adapted

from studies not directly related to metal speciation in natural

waters. For the latter two methods, the development of each procedure

is described in this section in addition to the final experimental pro-

cedure, because these details are not documented elsewhere.

Evaluation of the ASV complexometric technique used solutions con-

taining a single organic compound such as EDTA, histidine, and Des-

feral, for which a' for copper has been published. For the comparative

study, two colored-water samples from Waldo and Basin swamps and three

homopolyamino acids were titrated. The conditions of pH (6.25) and

ionic strength (0.1 M KNO3) were identical for each of the five pro-

cedures used in the comparison study.


1. Anodic Stripping Voltammetry

Titrations were performed according to the procedures described by

Shuman and Woodward (1973, 1977), except that the stripping step of ASV

was performed in the differential pulse (DP) mode rather than the lin-

ear (DC) mode. Instrumentation consisted of a PAR model 174 polaro-

graphic analyzer using a Kemula-type hanging mercury drop electrode

(HMDE). A saturated calomel reference electrode (SCE) was connected to

the solution by a 0.01 M KNO3 bridge. Instrument settings were 50 mV

modulation amplitude and 1 s-1 pulse rate. Pseudopolarograms were

constructed by the method of Figura and McDuffie (1979) to determine

appropriate plating potentials. Cadmium was plated at -0.8 V versus

SCE and scanned at 5 mV/s. Copper was plated at -0.3 V versus SCE and








scanned at 2 mV/s. Zinc was plated at -1.2 V versus SCE and scanned at

5 mV/s. The initial potential was applied for 60-300 s with stirring,

followed by a 15-s quiescent period. Analyses were performed with

50-mL samples containing 30-50 mM sodium acetate buffer, 10-100 mM

KN03, or 1 mM KNO3 (Desferal). All electrolytes were purified by

electrolysis.

Cyclic voltammetry was performed using an HDME in the DC mode.

The scan rate was 50 mV/s; solution conditions are described in the

results section.

Nitrogen gas, used to deoxygenate and stir each test solution, was

regulated by an in-line ball flowmeter. The pH was adjusted with NaOH

after deoxygenation and solution pH was checked at the end of each

titration to assure constant pH. A gel-filled pH electrode was used so

that chloride contamination of samples was avoided.

Solutions were prepared from distilled and deionized (Milli-Q)

water, and all chemicals were reagent grade except Desferal (Ciba-

Geigy) and the polyamino acids (Miles). Metal solutions were added in

PL quantities to avoid dilution, and prior to analysis, samples were

equilibrated 15 min after each addition of metal. Rapid equilibration

of each ligand with metal was verified by analyzing solutions that had

equilibrated in quartz test tubes up to 3 h. In all cases, equilibra-

tion was achieved within 15 min. Analyses were duplicated after each

metal addition, and each organic compound was titrated at least twice.

Linear portions of titration curves were modeled by least squares lin-

ear repression.








a. Theoretical considerations for determining CL and a' by ASV.

With each addition of metal M, the reaction with the ligand L proceeds

according to:


aM + bL = MaLb. (3.1)

The concentration of metal not bound to L, which is in equilibrium

with MaLb, is determined by ASV after each addition of metal. A

plot of the resulting peak current, ip, versus the concentration of

metal added (Figure 111-5), can be used to determine the available

ligand concentration, CL, and, ultimately, the conditional stability

constant, g', of the metal-ligand complex.

Shuman and Woodward (1977) derived the following equation to

determine the conditional stability constant of the MaLb complex:

i Su1/a (CM/a)1/a (3.2)
ip 71/a CL(CM _CM)b/a ] (3.2)


The symbol CM is the concentration of metal added, CL is the avail-

able ligand concentration (determined graphically), Su is the slope

of the upper region of the titration curve, and ' is the conditional

stability constant. In the case of large organic molecules, CL is

interpreted as the molar concentration of the individual functional

groups that act independently as ligands.

Given eqn. 3.2, a plot of ip versus


[( /a]/[(CL jM) ]


in the region of the titration curve well before the equivalence point

produces a slope of (Su,/B')/a. The conditions stated by Shuman




























A ip-
ACM s


CL
ADDED METAL, CM

Figure 111-5. Theoretical ASV titration curve of ligand with metal.








and Woodward (1977) when they derived eqn. 3.2, require CM < CL, so

only the values obtained from the first few additions of titrant are

plotted for this curve. By inserting reasonable integers for a and b,

and plotting ip versus


[(C)1/a]/[(CL b/a
a J(CL- 'CM)J


several curves are obtained for each titration. The curve most closely

approaching a straight line will indicate the stoichiometry of reaction

3.1. The slope of the most nearly linear line, SL, is equal to

(Su/')1/a and the ratio of Su/SL can then be used to deter-
mine p', the conditional stability constant of the MaLb complex.

Several assumptions are required for this procedure to work:

1. the concentration of available ligand was assumed to be in

excess prior to the addition of the metal titrant;

2. metals bound to ligand of interest more weakly than the

titrant metal are displaced by titrant metal;

3. peak current, ip, is proportional to the concentration

of metal not bound to the ligand of interest; and

4. the metal-ligand complex does not kinetically dissociate

during the plating step.

Assumptions (3) and (4) are addressed by this research and discussed in

Chapter IV.

In addition to the method of Shuman and Woodward (1977) to deter-

mine conditional stability constants, the ASV data were handled by the

Scatchard procedure, which is discussed at the end of this chapter

(section III-D-6).








2. Ion-Selective Electrode

Complexometric titrations of swamp samples and model compounds

were performed using an ion-selective electrode (ISE) in a manner simi-

lar to that described for the ASV procedure, except that a copper-ISE

was used to detect uncomplexed copper, rather than a mercury electrode.

Samples (50 mL) were added to a polyethylene beaker and were maintained

at 25C in a constant temperature bath. Potentials were measured with

an Orion Cu-ISE (model 94-29) and a double junction reference electrode

containing KNO3 in the outer reservoir. The pH was maintained at

6.25 0.05 by purging with a mixture of CO2 and N2 gases, and

solutions were continually stirred with a silanized glass paddle con-

nected to a stirring motor. Ionic strength was adjusted to 0.1 M with

KNO3, and copper was added in UL quantities to avoid dilution. The

electrode potential was monitored continually with a strip-chart

recorder to ensure that equilibrium had been obtained at each titration

point. Between samples, the Cu-ISE was soaked in a solution of

10-3 M EDTA for 10 min to remove excess copper from the electrode

surface. The system was calibrated frequently by titrating deionized

water adjusted to the same ionic strength and pH as for the samples.

The activities of ionic copper were calculated by:


pCu+2 = pCut + (Estd Esample)/s (3.3)

where pCu+2 and pCut are the negative logs of copper ion activ-

ity and total copper concentration, respectively, Estd and

Esample are the potentials (mV) of the standards and samples at

identical pCut, and s is the slope of the standard calibration curve

in mV.








3. Fluorescence Quenching

The method described by Saar and Weber (1980) was used to deter-

mine CL and B' for samples using copper as titrant. Samples were

placed in test tubes, ionic strength was adjusted to 0.1 M using

KNO3, and copper was added to each solution in increasing amounts

using micropipets to avoid appreciable dilution. Subsequently, pH was

adjusted to 6.25 using HC1 or NaOH. Relative fluorescence of each

sample was measured using an Aminco Model SPF125 at an excitation wave-

length of 350 nm and an emission wavelength of 450 nm. Solutions con-

taining increasing levels of copper were analyzed until the copper con-

centration was such that flocculation occurred.

The concentration of copper bound (Cub) at each titration point

was calculated using the equation:


Percent Quenched/57 = V = (Cub)/(LT) (3.4)

where (LT) is the total ligand concentration and 57 is an empiric-

ally derived coefficient (Saar and Weber 1980). Therefore, at any

point in the titration, the concentration of uncomplexed copper (Cuf)

was calculated by:


(Cuf) = (Cut) (Cub) (3.5)

where (Cut) is the total (analytical) concentration of copper in

solution. These data were treated further by the method of Scatchard

(1949) to produce conditional stability constants for the copper

complexes (see section III-D-6).








4. Continuous Ultrafiltration

Several studies have used the process of ultrafiltration for quan-

tifying metals associated with high-molecular-weight organic in nat-

ural waters (Gjessing 1971; Guy and Chakrabarti 1976; Schindler et al.

1972; Hoffman et al. 1981). The membrane filters allow low-molecular-

weight compounds and ions to pass through the filter, whereas high-

molecular-weight organic and the metals associated with them are

impermeable to the membrane and are retained. The most common applica-

tion of ultrafiltration to natural water studies has been the determin-

ation of metal distribution by filtering a sample at a single metal and

ligand concentration. For this approach, filtration must be repeated

if different concentrations of metal and ligand are to be evaluated.

One application of ultrafiltration that obviates the necessity of

repeating the experiment for each concentration of binding component

was reported for binding of microsolutes (e.g., steroids, oligopep-

tides) to large proteins (Ryan and Hanna 1971). The method is a con-

tinuous ultrafiltration technique in which a pressurized reservoir con-

taining a solution of the binding component of interest is continuously

fed into a stirred ultrafiltration cell, which contains the high-molec-

ular-weight organic matter initially with no binding component (Figure

III-6). By collecting discrete fractions of effluent, the free or

unbound concentration of microsolute can be evaluated for various con-

centrations of total microsolute. The level of bound microsolute can

be calculated for each fraction collected by performing a simple mass

balance, since the total quanity of microsolute added at any point

during the titration is known.





















N2
HMW .
ORGANICS
*. : ... "

S'."PRESSURIZED

FILTRATE ,FILTER







METAL
FRACTION SOLUTION
COLLECTOR









Figure III-6. Diagram of apparatus for method of continuous ultrafil-
trat ion.








The continuous ultrafiltration procedure reported by Ryan and

Hanna (1971) was adapted in this study so that the extent of metal

binding with organic matter could be evaluated. However, several dif-

ficulties appeared during the development of the method. The type of

filtration membrane was found to be important. The small pore-size

filters have more ionic character than the large porosity filters, and

solute interaction with ionic membranes produced erratic results. On

the other hand, with increasing porosity, retention of organic matter

diminished. The Amicon UM-10 (10,000 dalton) membrane was found to be

the best compromise. The molecular weight cutoff values assigned to

the filters are approximate and apply to globular proteins and sugars.

Thus, membrane rejection is not related solely to molecular weight,

since retention is a function of both size and charge of solute.

Another parameter found to control the success of this procedure

was the flow rate through the membrane. At fast flow rates (0.5-1.0

mL/min) and with a purely inorganic matrix, the concentration of metal

in the retentate eventually exceeded the concentration of metal in the

feed solution, apparently due to partial rejection of ionic metal by

the membrane. A flow rate of 0.2 mL/min alleviated the problem of

enhanced concentrations of metal in the retentate, but even at that

flow rate, the metal concentration in the filtrate was lower than that

in the retentate for purely ionic solutions. Ionic metal was binding

to the membrane and causing low results. It was discovered that the

ionic strength of the solution influenced the extent of metal adsorp-

tion to the filters, and subsequently, all solutions were adjusted with

KN03 to 0.1 M, the lowest level found to prevent metal loss to the

membrane.








After the above problems were rectified, the samples from Waldo

Swamp, Basin Swamp, and several model compounds were analyzed using the

following procedure. Samples were filtered through 1-um pore size

Nucleopore membrane filters, and 70 mL of the filtrate was placed above

an Amicon UM-10 ultrafiltration membrane in a 70-mL-capacity filtration

cell (Amicon Model 52). Each sample was rinsed with a solution of 0.1

M KNO3 at pH 6.25 to remove the nonretained compounds. During the

rinsing process, great care was taken to ensure that the sample was not

contaminated with metal from a previous titration. Several membranes

were dedicated to rinsing samples and were replaced during metal titra-

tions. The cell and tubing were rinsed with dilute acid prior to add-

ing samples or solutions. After rinsing, the sample was transferred

from the cell to an acid-washed bottle. The membrane was changed and

equilibrated with a dilute solution of the metal of interest for 1 h.

Subsequently, the membrane was rinsed with 15 mL of metal-free buffer

and the sample was added to the cell. The reservoir, which contained

the solution of metal (and 0.1 M KNO3 at pH 6.25), as connected to

the ultrafiltration cell, and nitrogen gas was applied at a pressure of

2 atm. The sample was stirred continuously, and effluent was collected

in 3-8-mL fractions with a Gilson GME automatic collector. After col-

lection, samples were acidified with nitric acid and analyzed by atomic

absorption spectroscopy (Perkin-Elmer Model 5000).

The data generated with the continuous ultrafiltration technique

were used to calculate the quantity of metal bound, Mb, and metal not

bound, Mf, to the retained ligand for each fraction collected by

using a simple balance of materials,








Mb = Madded Mf Mout (3.6)

where Madded and Mout are the total amounts of metal added to

the cell and metal in the cell effluent, respectively. Equation 3.6

can be expanded to:


Mb = Vout(Mo) Vcell(Me) Mout (3.7)


where Vout is the volume of effluent collected, (Mo) is the

initial metal concentration (in the feed solution), and (Me) is the

measured metal concentration in the effluent (which is equal to the

free metal concentration in the cell at the time the fraction was col-

lected). Therefore, the amount of metal bound to the organic ligand at

any point throughout the titration can be determined by eqn. 3.7. A

programmable calculator was used to reduce the experimental data such

that for each fraction collected, both free and bound metal were com-

puted. These data were treated further by the method of Scatchard to

determine conditional stability constants for the metal-organic com-

plexes (see section III-D-6).

To perform the continuous ultrafiltration titration accurately and

to compute the levels of metal bound and not bound to retained organic

matter by the methods described, the following conditions are

necessary:

1. the ligand of interest is retained by the membrane;

2. the cell contents are completely mixed;

3. the effluent metal concentration is equal to the unbound

metal concentration in the cell;

4. the cell volume remains constant; and

5. the effluent is collected in accurately measured volumes.








5. Competing Ligand/Differential Spectroscopy

This procedure involves adding a competing ligand (A) that has a

known stability constant with the titrant metal to a solution of the

metal (M) and ligand (L) of interest, and thus establishing the follow-

ing equilibria:


M + A = MA (3.8)


M + L = ML. (3.9)


By quantifying either A or MA, and knowing the analytical concentra-

tions of M, L, and A, the value of the conditional stability constant

for the ML complex can be determined. Details of such a calculation

are presented later in this section.

Several organic compounds were examined for use as competing

ligands in this study. Of the ligands examined, most absorbed light in

the regions below 300 nm, which is an area that naturally occurring

organic absorb strongly. Scans of aqueous solutions of nitrilotriace-

tate, citric acid, phenylalanine, and histidine revealed only neglig-

ible absorption between 300 and 600 nm. The addition of copper to

these ligands did not affect any of their absorption spectra at wave-

lengths above 300 nm. However, salicylate (anionic form of o-hydroxy

benzoic acid) was found to absorb light strongly at 291 nm, and the

addition of copper caused a bathochromatic shift (i.e., to longer wave-

lengths) in the absorption band of salicylate (SAL) to a wavelength of

320 nm at maximal intensity. The concentration of the Cu-SAL complex

was found to be proportional to the absorbance at 320 nm. Because








salicylate, at the concentrations used, absorbed appreciably at 320 nrm,

this background absorbance was subtracted optically by performing dif-

ferential spectroscopy; namely, the sample without added metal was

placed in the reference path of the spectrophotometer, and the sample

with added metal was placed in the sample path. Thus the instrument

measures the difference between the two samples, which in this case is

the absorbance due to CuSAL.

Sample preparation consisted of pipeting 50.0 mL of sample into

each of several 125-mL flasks and adding 0.50 mL of 10-1 M salicy-

late. The ionic strength was increased to 0.1 M with KNO3 and copper

was added in various quantities to each flask. The pH was adjusted to

6.25 0.05 using small aliquots of NaOH or HCl to minimize dilution.

Sample containing no additional copper was placed in the reference cell

(either 1-cm or 4-cm pathlength) of a Perkin-Elmer Model 552 spectro-

photometer and samples with copper added were placed in the sample

cell. Absorbance was measured at 320 nm.

Absorbance values were standardized by titrating a solution of

10-3 M salicylate with copper. The concentration of CuSAL at each

level of total copper concentration was calculated by computer program

(MINEQL) and a plot of absorbance versus CuSAL concentration is pre-

sented in Figure 111-7. The lower limit of this procedure using a 4-cm

cell is 2.0"10-7 M CuSAL. Measurement of lower concentrations

was attempted with a 10-cm cell, but total absorbance was beyond the

maximum absorbance measurable with the instrument.

The absorbance values obtained for samples were converted to con-

centrations of CuSAL using the data in Figure 111-7. Hence, for the

competing equilibria,




























4cm cell


Icm cell


2 3 4 5

CuSAL,IO5M


Figure III-7. Calibration curve for competing ligand/differential
spectroscopy method.








CuL = Cuf + L (3.10)

Cuf + SAL = CuSAL (3.11)

where L is the ligand of interest, Cuf is uncomplexed copper, and

SAL is salicylate, equilibrium concentrations of Cuf, CuL, and L were

calculated by using the mass balance of each component, the value of

B'CuSAL, and the empirically derived concentration of CuSAL. The
mass balance equations for this experiment are:


CuT = (Cuf) + (CuL) + (CuSAL) (3.12)

SALT = (SAL) + (CuSAL) (3.13)

LT = (CuL) + (L) (3.14)

where the parentheses indicate aqueous concentrations and the subscript

T indicates the total concentration of each species. The conditional

stability constant for the Cu-SAL complex was calculated from the ther-

modynamic stability constant (Sillen and Martell 1971) by the method of

Ringbom (1963) (q.v., section IV-A). The value of B'CuSAL at pH

6.25 was calculated to be 103.5 and thus

B' = io3.5 = (CuSAL)
103.5 Cuf)(SAL) (3.15)


The empirical concentration of CuSAL for each titration point was

inserted into eqns. 3.13 and 3.15, and (Cuf) was calculated by simul-

taneously solving the two equations. Subsequently, (CuL) was deter-

mined using eqn. 3.12. These data were treated further by Scatchard

analysis to determine '.








6. Theoretical Considerations for Determining
B' by Scatchard Analyis

Conditional stability constants, 8', can be estimated by using the

method of Scatchard (Scatchard 1949; Guy and Chakrabarti 1975; Mantoura

and Riley 1975), in which it was assumed that

B' = (Mb)/(Mf)(nLT Mb) (3.16)

where (Mb) and (Mf) are concentrations of metal bound and not

bound to ligand, LT is the total ligand concentration, and n is the

number of binding sites per ligand molecule. Equation 3.16 can be

rearranged to


(Mb)/(LT)(tMf) = B'[n (Mb)/(LT)]. (3.17)

By substituting V for (Mb)/(LT), the final form of the equation

becomes


V/(Mf) = B'(n V). (3.18)

Thus a plot of V/(Mf) versus V will produce a curve with slope -a',

and n can be determined from the intercept value. This data analysis

has been attributed to Scatchard, and a plot of V/(Mf) versus V is

termed a Scatchard plot.

Originally, the method was intended for use with pure compounds

containing one or more distinct binding sites. Thus for a single

ligand with one class of binding site, a titration with a metal will

produce a straight line with slope -B' and an intercept of n on the

V axis. However, the typical curve obtained by titrating environmental








samples with metal is asymptotic due to the variable strength of metal

binding. Many researchers who use Scatchard plots for determining '

for metals with naturally occurring organic classify the sites into

two categories with one "strong" site and one "weak" site (Figure

111-8). However this approach is too simplistic to characterize the

variety of complexes expected in environmental samples. For the data

obtained here, subtraction of the "strong" site from the "weak" site

produced plots that remained curved. The slope of the "strong" site is

too steep to have much effect on the remainder of the titration curve.

Therefore, the conditional stability constant data generated here were

treated as a continuum, and the method commonly used to correct the

plots for contributions from the "strong" and "weak" sites was not per-

formed. Instead, the curves were divided into sections of V,

and the data were fitted to a straight line. The ranges of V were

selected by visually sectioning the Scatchard plot from the method that

produced the widest range of values (ISE method). Three segments were

used to fit the experimental data. Nonetheless, the data are curvilin-

ear, and the only intent of fitting segments of the data to straight

lines was so that a comparison could be made among the five analytical

procedures used to determine the levels of completed and uncomplexed

metal.



E. Chemical Characterization and Fractionation
of Sewage and Swamp Water


Swamp water samples were collected in 5-L quantities from undis-

turbed areas in Waldo Swamp and Basin Swamp. Secondary sewage effluent




























o SLOPE = -,
("STRONG" SITE)




no,
"-- .. .


\


\" o SLOPE = -b
o-- ("WEAK" SITE)
S 0 +nb


V= (metal bound)/(total ligand)

Figure 111-8. Theoretical Scatchard plot for titration of
organic matter with metal. Dashed lines are
resolved contributions (assuming that only two
classes of binding occurred).


I


\ A'n + 9'nb








from the Jasper sewage treatment plant was obtained from the outfall of

the oxidation pond that flows directly into Basin Swamp, and septage

was collected from the effluent of the septic tank that received sewage

from Waldo, Florida. All samples were returned to the laboratory and

filtered through a 0.5-pm precombusted (4500C) Reeve Angel 984-H glass

fiber filter. Samples were preserved at 40C until needed.

The two sewage samples were treated further to reduce the carbon-

ate and ammonium concentrations. Carbonate and copper form an ion pair

complex that is not measured by ISE. Thus copper-organic complexation

is difficult to assess for samples containing appreciable levels of

carbonate. Ammonium was removed from samples because the fluorescence

procedure for quantifying amino acids is reactive to ammonium, although

to a lesser degree than for amino acids. Therefore, the pH was lowered

to 4, and the sewage samples were purged with dry nitrogen for 3 h at

35C to remove carbonates. Subsequently, the pH was increased to 9.5

and purged with dry nitrogen for 3 h at 350C to remove ammonium.

The inorganic composition of each sample was quantified by deter-

mining concentrations of dissolved organic carbon (DOC), dissolved

organic nitrogen (DON), and the dissolved hydrolyzable amino acid

nitrogen (DHAAN). DOC was analyzed by direct injection using a Beckman

Model 915 Total Organic Carbon Analyzer equipped with a Model 865

Infrared Analyzer (APHA 1981). Organic nitrogen was determined after

Kjeldahl digestion by the automated phenate method (U.S. EPA 1979).

The amino acid nitrogen was determined by the fluorescamine technique

(Udenfriend 1972) before and after hydrolysis. Samples to be hydro-

lyzed were placed in precombusted (4500C) ampules to which an equal

volume of 12 N HCl was added. The ampules were then purged with dry








nitrogen, sealed, and autoclaved (1200C) for 1 h. After cooling, the

ampules were opened and evaporated (45C) to dryness in a vacuum desic-

cator. The hydrolysate was quantified by the fluorescamine method

using an Aminco Model SPF125 spectrofluorometer.

Sewage samples were fractionated by ultrafiltration using Amicon

Diaflo ultrafiltration membranes (UM-02 and UM-10) and a magnetically

stirred Amicon Ultrafiltration Cell (Model 52), under nitrogen pressure

of 2 atm. Before adding samples, each filter was flushed with 200 mL

deionized water to remove contaminants.

The swamp water samples were fractionated using diethylaminoethyl

(DEAE) cellulose, which is an anion exchange material. DEAE cellulose

(Bio-Rad) was prepared by the following procedure: the exchanger was

soaked in 1 M HC1, filtered, rinsed with deionized water, and then

soaked in 1 M NaOH, filtered, and rinsed with deionized water. This

process was repeated five times. The purified exchanger (15 g) was

poured into a 2.5 cm by 25 cm glass column and allowed to settle. The

column was rinsed with 100 mL of 0.01 M NaCl at pH 7, and 1-L samples

were applied individually to the column. The retained organic were

desorbed by gradient elution using 0.1 M Prideaux buffer (0.1 M in

phosphoric, acetic, and boric acids) and 1.0 M NaCl. Subsequently, an

increasing pH gradient was used (0.1 M Prideaux buffer and 0.2 M NaOH).

The Prideaux buffer with NaOH provides a nearly linear pH gradient up

to pH 12 (Curtis et al. 1981). Following desorption, the recovered

organic were desalted using ultrafiltration with a continuous flow of

deionized water.

Copper binding experiments using complete and fractionated samples

were performed using the methods of continuous ultrafiltration and ISE.








Experimental details of these procedures are described in section

III-D.



F. Copper Binding by Peat Extract


1. Extraction

Waldo peat was extracted using the method of Randhawa and Broad-

bent (1965), as follows: 200 g of wet A, horizon was mixed with 400

mL of 0.1 N HCI and shaken for 1 h. Supernatant was decanted and

discarded. Eight hundred milliliters of 0.5 N NaOH solution was

added, and the contents was shaken for 4 h. One hundred milliliters

of extract was diluted to 2 L and filtered through Whatman 41 paper.

The pH was adjusted to 5 by additions of acetic acid, resulting in an

additional 20% dilution. Aliquots of this solution were filtered

through 0.45 pm Millipore filter paper prior to use.


2. Molecular Weight Fractionation

Gel filtration chromatography (GFC) was performed with Sepha-

cryl, a Sephadex gel with a wide fractionating range (250,000 to

5,000 M.W. as globular proteins). The preswollen gel was added to a

2 x 100-cm glass column and rinsed at 40 mL/min for 4 h with 1 M Tris

buffer (pH = 8.6). Operating flow rate was 17 mL/min, and 5-mL

fractions were collected automatically by a Gilson fraction collector.

The column was rinsed thoroughly (3 times) after each injection. Blue

dextran 2000 (M.W. = 2 x 106) and KCI were used to determine void

volume (Vo) and total volume (Vt), respectively. Chromatograms of

extract and reinjected fractions were produced by monitoring UV

absorbance at 280 nm.








Ultrafiltrations were performed using Amicon Diaflo membranes

and a magnetically stirred Amicon Ultrafiltration cell (Model 52)

under nitrogen pressures of 1.1-2 atm. A 50-mL aliquot of extracted

organic matter was added to the cell containing an XM-50 (50,000 dal-

ton cutoff) or XM-300 (300,000 dalton cutoff) membrane. After 25 mL

of sample was filtered, the pressure was reduced and the retained

portion of the sample was collected for analysis. The organic con-

tent of the filtered and retained fractions was quantified spectro-

photometrically at 460 and 280 nm.


3. Binding Capacity and Stability Constant Determinations

The ion-exchange method reported by Crosser and Allen (1976) was

used to measure the copper binding ability of the peat extract.

Dowex AG 50W-X8 strong cation exchange resin (200-400 mesh) was pre-

pared by adding 10 M NaOH to the resin until pH remained neutral and

was followed by batchwise equilibration with 0.02 M acetate buffer.

Resin was suction dried and 1.0-g amounts weighed into each of 20

250-mL flasks. Then 50 mL of 0.02 M acetate buffer (pH = 5.9) was

added to each flask, and 1 mL of peat extract was added to each of

the 10 designated flasks. All flasks were shaken to equilibrate for

4 h at 25C. Copper (as sulfate) was added with micropipettes to

yield Cu concentrations ranging from 0 to 6.3 x 10-4 M in both

control and ligand-containing flasks. Flasks were shaken an addi-

tional 17 h. Total soluble copper was determined by atomic absorp-

tion spectrophotometry (Varian model 1200).

Additionally, the copper completing capacity of the extracted

organic matter and the conditional stability constant of the copper-






73

organic complex were investigated by ASV using the methods of Chau et

al. (1974) and Shuman and Woodward (1977). Instrumentation and set-

tings were those described previously (section III-D-1) except that a

scan rate of 5 mV/s was used. Copper was plated at -0.4 V versus SCE

for 75 s (60 s stirring, 15 s quiescent) prior to stripping. Analy-

ses were performed on 50-mL aliquots of sample extract (which was

diluted 1:10) to which 50 pL of 6-M acetate was added to provide a

6-1M supporting electrolyte at a pH of 5.9. The solution was deaer-

ated 15 min with dry nitrogen to remove oxygen prior to analysis.

Copper was added sequentially and analyzed until a complete titration

curve was obtained.













CHAPTER IV
EVALUATION AND COMPARISON OF METHODS
FOR DETERMINING CL and 8'


The first section of this chapter is the result of an evaluation

of the ASV titrimetric method for determining metal binding capacity,

CL, and conditional stability constants, S', using model compounds

such as EDTA, Desferal, and histidine. The subsequent sections des-

cribe a comparison of five methods to determine CL and 5' using

colored water samples from Waldo and Basin swamps, and three homopoly-

amino acids-polyaspartic acid, polyarginine, and polyalanine. Titra-

tions of surface water samples were performed with aliquots from com-

posite samples, which contained 50 and 25 mg C/L for Waldo and Basin

swamps' samples, respectively. For comparative purposes, all Scatchard

calculations were performed by using the same molar concentration of

ligand-10-4 M (Waldo Swamp) and 5-10-5 M (Basin Swamp). The

Scatchard analysis weights B' according to the molarity of binding

sites (not molarity of total ligand), and thus the final value of B' is

independent of the total molar value attributed to a given ligand con-

centration. The polyamino acids were homopolymers with molecular

weights of 13,900 (polyarginine), 3,900 (polyalanine), and 5,400 (poly-

aspartic acid).

Because a comparison among the five procedures was intended,

identical conditions were used for each method; namely, pH 6.25 and

ionic strength 0.1 M KN03. The pH was selected as the best compro-








mise between the increasing concentrations of inorganic complexes of

metal at higher pH and decreasing organic complexation of metals at

lower pH. The ionic strength of 0.1 M KNO3 was used because it was

the minimum level found to prevent adsorption to ultrafiltration mem-

branes.



A. Evaluation of the ASV-Complexometric Titration


The ASV technique has been used widely to determine CL and R'

for natural waters (Chau and Lum-Shue-Chan 1974; O'Shea and Mancy 1976;

Smith 1976; Shuman and Woodward 1977; Sugai and Healy 1978; Shuman and

Cromer 1979; Srna et al. 1980; Tuschall and Brezonik 1980; Hoffman et

al. 1981), but surprisingly no evaluation of the method has been

reported using ligands with known stability constants with copper. To

date, the only published validation or calibration of the procedure was

reported by Shuman and Woodward (1973) who titrated ethylenedinitrilo-

tetraacetic acid (EDTA) with cadmium and concluded that the procedure

produced accurate results. However, copper is commonly used to titrate

natural water samples because it forms much more stable complexes with

most organic ligands than does cadmium (Irving and Williams 1948).

Because of the discrepancies observed between the ion-exchange tech-

nique and the ASV titration in the determination of CL and B' (q.v.,

Chapter V), a systematic evaluation of the ASV procedure appeared

necessary to determine its validity.

One of the requisite criteria for performing the titrimetric ASV

procedure is that the metal-organic complex be nonreducible at the mer-

cury electrode at a potential significantly separated from the








reduction of ionic metal (Shuman and Woodward 1973). Consequently,

ligands that form strong complexes with copper were sought so that B'

could be determined experimentally and compared to existing literature

values. Figure IV-1 illustrates the structures of the organic

compounds selected for titration with copper and analysis by ASV.

The first compound titrated was EDTA, and it became apparent why

S'CuEDTA had not been determined previously by this method. For

solutions of EDTA (no copper added), a broad peak was observed at 0.0 V

versus SCE, which is the same potential at which copper is oxidized.

Originally, the peak was believed to be caused by contamination of the

EDTA or buffer solutions with copper, and all solutions were purified

by electrolysis. However, the peak remained. Further investigation

showed that this peak decreased in size as EDTA was titrated with

copper up to a Cu:EDTA ratio of 0.4 (Figure IV-2), and subsequently a

peak at 0.0 V versus SCE increased in size with further addition of

copper. Apparently the initial peak was due to the cathodic shift in

the oxidation of mercury by EDTA, and the subsequent peak (at Cu:EDTA

ratios above 0.4) was due to reducible copper. In addition to copper,

several other metals were found to suppress the interfering mercury

peak; namely, lead, nickel, cadmium, zinc, and calcium. Therefore a

metal that would suppress or mask the interfering peak was added to the

EDTA solution so that the initial copper peaks could be quantified

accurately. Calcium was selected since it forms a weaker EDTA complex

than do most of the heavy metals.

Because the ASV procedure produces a conditional stability con-

stant (B') and most literature values are thermodynamic stability con-

stants (B), one must be converted to the other so that a direct













HOOCCH2 CH2COOH

N CH2CH2 -N
/HOOCCH \
HOOCCH2 CH2COOH


H2N CONH CONH

(CH2)5 (CH2)2 (CH2)5 (CH2)2 (CH2)5 CH3
\ / \ / \
N- C N-C N-C
I II I II I II
HO 0 HO 0 HO 0


EDTA


DESFERAL


CH2 -COOH

N -CH2-COOH

CH2-COOH


HO HC= CH -CO\
OH


p HYDROXYCINNAMIC ACID


NH2
/ CH2-CH-COOH
N NH




HISTIDINE

Figure IV-1. Organic compounds used in ASV study.


0
HO-C

OH

OH OH

GALLIC ACID





OH

OH OH

PYROGALLOL




















































1.0i A


-0.4 -0.3 -0.2 -0.1


Figure IV-2.


Anodic
(5 pM)
tions.


0.0 +0.1 V


stripping peaks for a solution of EDTA
containing copper at varying concentra-








comparison can be made. For the model compounds, the thermodynamic

stability constants from Sillen and Martell (1971) and Ringbom (1963)

were converted to conditional values by the method of Ringbom (1963) as

follows:

[ =ML] MmLn (4.1)
S'MmLn [M'][L'] a n'


where M' is the metal not completed by L, L' is the ligand not

completed by M, aM = [M']/[M]free, and aL = [L']/[L]free-

For instance, in the case of the titration of EDTA with copper, with

hydrogen and calcium as competing cations:

[Y'] 1 [H] [H]2 [H]3 + [H]4 + [Ca]BaY-
EDTA = y]free K4~ K4K3 K4K3K2 K4K3K2K1I

(4.2)


After substituting the appropriate constants (pK4 = 10.34; pK3 =

6.24; pK2 = 2.75; pK1 = 2.07; and PkCaY = 10.7) (Ringbom 1963)

and concentrations of analysis (pH = 7.0; pCa+2 < 5.3), the aL

value was calculated to be 105.4. Similarly, the aCu was cal-

culated to be 100.6 for pH = 7.0 and acetate concentration of 0.03
M. Therefore, the thermodynamic stability constant for Cu-EDTA (0 =
1018.8) was converted to the conditional value (s') as follows:

u-EDTA 1018.8 = 1012.8
Cu-EDTA aCu aEDTA 100.6 105.4 (4.3)


It should be noted that as Cu+2 replaced the calcium completed by

EDTA during the titration, the free calcium concentration, [Ca+2],








increased from 10-6.3 M to '10-5.3 M. Thus, the conditional

stability constant (a' = 1012.8) represents a lower limit based on

the extreme case (i.e., [Ca+2] = [Ca]T = 10-5.3 M).

To confirm the assumption that the Cu-EDTA complex was

sufficiently strong and not reducible at the plating potential used for

analysis, a pseudopolarogram was constructed using the method of Matson

(1968) (Figure IV-3). Each point on the plot represents the amount of

reducible copper at a specific plating potential for various solutions.

Figure IV-3 shows that a copper solution devoid of chelating agents is

plated maximally at -0.3 V but that Cu-EDTA is not reduced 100% until a

plating potential of -0.8 V versus SCE is applied. Hence a plating

potential of -0.3 V versus SCE should be sufficiently negative to plate

ionic and inorganic species of copper without directly reducing the

copper completed to EDTA.

The results of the titration of EDTA with copper (Figure IV-4)

produced a conditional stability constant, 8', equal to 107.7,

which is over five orders of magnitude lower than other published

values (B' = 1012.8) (Ringbom 1963; Sillen and Martell 1971) for

similar solution conditions. This discrepancy was apparently caused by

dissociation of the complex throughout the initial stages of the

titration ([Cu] << [EDTA]), thus producing a slope (SL) much greater

than expected. If complex dissociation were not occurring, the lower

curve in Figure IV-4c (SL for CuEDTA) would be nearly horizontal with

a slope of 1.2 x 10-6.

Calcium (and other cations) have been reported to cause a shift in

the reduction potential of CuEDTA and thus may possibly enhance strip-

ping current due to enhanced rates of dissociation of CuEDTA (Bril and

























0.0 -0.2 -0.4 -0.6 -0.8


PLATING POTENTIAL (V)


Figure IV-3.


DPASV pseudopolarograms of copper (2 pH) plus NTA
(5 pM) or EDTA (5 pM) in the presence and absence
of calcium (5 pM) at pH = 7.0.



















































CM/(C_- C)


Figure IV-4.


DPASV titration curve of
(A) 5 PM EDTA with copper;
(B) 2 x 10-5 M EDTA
with cadmium; and (C) plots
of metal (Cr) and ligand
(CL) concentrations for
initial additions of copper
and cadmium.








Krumholz 1954; Rajput et al. 1978). Further experiments were performed

to determine whether the low value of B'CuEDTA that was measured

here resulted from the presence of Ca+2. I was able to quantify

several points relatively early in an EDTA-copper titration in the com-

plete absence of calcium, since the interfering peak subsided and

shifted anodically as copper was added. The resulting peak heights at

CM/(CL CM) values of 0.4 and 0.67 (see Figure IV-4c) did not
differ by more than 10% from those in the presence of 5, 10, and 20 pM

of calcium, and thus calcium did not enhance the dissociation or reduc-

tion of CuEDTA under the conditions of analysis. Additionally, pseudo-

polarograms were constructed for CuEDTA solutions in the presence and

absence of calcium (Figure IV-3). The half-wave potential for CuEDTA

was -0.55 V versus SCE in the absence of calcium, and -0.50 V versus

SCE in the presence of 5 pM calcium. This shift of 0.05 V in the

reduction potential of CuEDTA was not large enough to cause the

observed stripping currents when the plating step was performed at -0.3

V. These results clearly demonstrate that the low value of B'CuEDTA

was not caused by the use of Ca+2 in the EDTA solutions.

Shuman and Woodward (1973) reported that complex dissociation

occurred with CdEDTA and proposed a method to correct for it. Their

correction amounted to a 38% increase in B' for CdEDTA at pH = 4.5.

EDTA was titrated with cadmium using the differential pulse (DP) mode

(Figure IV-3) to compare with the titration reported by Shuman and

Woodward (1973), which was done in the DC mode. Experimental condi-

tions were identical to those reported by Shuman and Woodward (1973),

except for mode (DP) and EDTA concentration. The higher EDTA

concentration used by Shuman and Woodward (4.45 to 17.8 x 10-5 M)








produced nonlinear titration curves in the region beyond the equiva-

lence point when analyzed by DPASV; thus lower levels of EDTA (2-20 PM)

were used. The DP mode was selected so that ligand and pH levels

typically found in the environment could be investigated, and because

under those conditions, DC mode often would require unreasonably long

deposition times. After correcting for complex dissociation, I calcu-

lated B' for CdEDTA to be 108.5, which compares well with the

values of Ringbom (1963) (6' = 108.8) and Shuman and Woodward

(1973) (0' = 108.1) for the same solution conditions.

In order to compensate for dissociation of the Cu-EDTA complex

using the method of Shuman and Woodward (1973), the correction factor

would need to be 105-.1 to produce an accurate conditional stability

constant. Instead, the correction procedure produced negative inter-

cept values, which are physically meaningless and could not be used to

correct 6' of CuEDTA.

Possibly, partial dissociation of the Cu-EDTA complex during the

time scale of measurement by ASV is atypical of the behavior of (ther-

modynamically) strong copper-organic complexes. Therefore several

other organic compounds were investigated. Desferal, a trihydroxamate

siderochrome, is a natural microbial growth factor that simulates the

functionality and molecular weight (656 daltons) of aquatic and soil-

derived humic substances (Wilson and Weber 1977; Reuter and Perdue

1981; Schnitzer 1981). Desferal was recently reported to complex

copper with 6' = 108.4 (McKnight and Morel 1979), based on

titrations with a copper ion-selective electrode. Titration of Des-

feral (10-5 M) with copper using DPASV under the same conditions of

pH (6.25) and supporting electrolyte (10-3 M KNO3) reported by








McKnight and Morel (1979) revealed that the Cu-Desferal complex was

reducible at the mercury electrode (Table IV-1). In solutions of Des-

feral, copper produced somewhat lower (by 10-50%) and broader peaks

than did equal concentrations of free copper ion; however, on an areal

basis the peaks were of equal size. The peak potential (Ep) of

Cu-Desferal was 100 mV cathodic of Ep for uncomplexed copper, and

thus we were unable to plate copper ion without reducing the Cu-Des-

feral complex. A similar situation was found with Cu-nitrilotriacetic

acid (NTA) in that discrete plating of copper ion was not possible

(Figure IV-3).

On the other hand, cadmium in the presence of Desferal was only

partially reduced during the plating step of ASV. The addition of

Desferal to a 2 pM cadmium solution decreased ip, but even at Desfer-

al concentrations of 1 mM, ip did not approach 0.0 vA, suggesting

that cadmium forms a weak, nonreducible complex with Desferal (O'Shea

and Mancy 1976).

Several other organic compounds were studied to determine the

extent of copper reduction during analysis by DPASV. Histidine formed

a complex with copper that was completely reducible at the mercury

electrode (Table IV-1) despite calculations predicting that more than

99% of the copper was completed with histidine under the conditions of

analysis (2 x 10-6 M Cu; 2 x 10-4 M histidine; for Cu(HIS)2,

log 82 = 18.1 and log 1'2 = 16). Three other small compounds,

p-hydroxycinnamic acid, gallic acid, and pyrogallol, which are simple

(monomeric) models of humic material found in natural waters, were

analyzed in solutions of 2 1M copper. At ligand concentrations up to

10-4 M, the stripping currents were identical to those for






Table VI-1. Effect of model organic compounds on ASV analysis of Cu and Cd.


Ep (volts vs. SCE)

Metal Reducible
Model Metal in presence (Metal nature of
compound titrant pH of ligand* ion) complex**


EDTA Cu 7.0 0.065 (0.0) Complex predominantly
nonreducible

Cd 4.5 0.60 (- 0.60) Complex nonreducible

Desferal Cu 6.25 0.07 (+ 0.025) Complex reducible

Cd 6.25 0.58 (- 0.58) Complex nonreducible, but weak

Bovine Serum
Albumin Cu 7.0 0.0 (0.0) Initial complex nonreducible
subsequent complexes partially
reducible

Histidine Cu 7.0 0.0 (0.0) Complex reducible

p-hydroxycinnamic
acid Cu 7.0 0.0 (0.0) Complex reducible

Gallic acid Cu 7.0 0.0 (0.0) Complex reducible

Pyrogallol Cu 7.0 0.0 (0.0) Complex reducible


*Most negative value of Ep obtained during titration.
**For plating potentials of -0.3 V (Cu) and -0.8 V (Cd).








uncomplexed copper. No stability constants for these three compounds

are available in the literature to calculate the extent of copper com-

plexation, but some complexation seems likely at these concentrations

of ligand.

A large molecular weight organic compound, bovine serum albumin

(BSA), was used as a model of the macromolecular dissolved organic

nitrogen in natural waters and was titrated with copper. A very

strong, nonreducible complex was observed initially, followed by weaker

binding. No discernable break was found in the titration curve, thus

precluding an accurate determination of '. Bovine serum albumin has

one very strong binding site and at least 11 other weak binding sites,

some of which act more like ion exchange sites than binding sites

(Klotz and Curme 1948). The nature of BSA exemplifies the problems

involved in analyzing a heterogeneous mixture of organic compounds,

such as those found in natural waters. For example, with naturally

occurring organic matter others (Mantoura and Riley 1975; Guy and

Chakrabarti 1976; Bresnahan et al. 1978; Giesy 1978) have reported at

least two distinct classes of binding sites (based on Scatchard plots).

The binding sites typically varied from strong to weak as metal

concentration increased. The ASV titration procedure does not identify

the various classes of available binding sites, and calculation of 8'

from such titrations is based on the first few copper additions (CM

CL [Shuman and Woodward 1977]), which probably represent only the
stronger binding sites, not the average. Thus for a mixture of

compounds with various binding sites, the B' probably would be

overestimated by the ASV procedure, the convoluting problems of complex

dissociation and direct complex reduction aside.








It is concluded that, as currently practiced, the complexometric

titration with copper, using ASV with a stationary mercury electrode,

is not an accurate method either for determining completing capacities

(CL) of natural water samples or for estimating stability constants

(B') of copper with natural organic matter. The titrimetric ASV

procedure has not been validated with any of the pure compounds

investigated using copper as the titrant. Correct CL values could

not be obtained for a wide variety of organic compounds. Although

titrations of EDTA gave the correct value of CL, the calculated a'

was over five orders of magnitude in error.

Inaccuracies in determining completing capacities by this method

arise from at least two sources, namely:

1. Rapidly dissociating or directly-reducible metal-organic

complex under conditions that predict an "inert" complex.

2. Lack of plating potentials for some environmentally

significant metal-organic complexes that are sufficiently

separated from the reduction of copper ion.

For a variety of organic ligands that form thermodynamically strong

complexes with copper, at most slight decreases in ip were found when

the plating step was performed at -0.3 V, which was the least negative

potential that could be used to plate copper ion under my experimental

conditions. The ASV copper-titration method thus apparently does not

measure CL for: (1) most inorganic ligands (Ernst et a]. 1975; Shuman

and Woodward 1977); (2) many small organic ligands that form moderately

strong copper complexes (e.g., NTA, histidine); and (3) some large,

naturally occurring organic species that also have relatively high B'








values with copper (e.g., Desferal). However, when using the method

with natural water samples, one does obtain a titration curve with an

"apparent" equivalence point. The work with model compounds (above)

does not rule out the possibility that naturally occurring organic

(e.g., proteins, humics) form non-reducible complexes. Further evalu-

ation of the ASV method with humics from natural waters is presented in

the following section, and the work described there does provide an

explanation for the "positive" results obtained with natural water

samples.

It has been demonstrated that the above problems occur frequently

with copper as a titrant, but the possibility that other metal ions may

be more suitable titrants for completing capacities is not precluded by

this work. Cadmium apparently forms more ASV-nonlabile complexes with

organic ligands than does copper, probably due to the electron config-

uration of the d-orbitals, which are filled for Cd(II) (d10) but

only partially filled for Cu(II) (d9). However, cadmium forms

complexes with organic ligands that are much less stable (thermodynam-

ically) than does copper, and the completing capacities thus obtained

may not be comparable to the true completing capacities of natural

organic compounds with copper. Hanck and Dillard (1977b) used indium

(III) as a titrant for completing capacities because In(III) typically

forms stronger complexes than copper. However, the same problems of

interpreting such completing capacities relative to binding capacities

for other metal ions still remain, and the environmental significance

of indium itself is rather small.

The same deficiencies stated above for completing capacities apply

to the estimation of conditional stability constants. In addition, the








problem observed with BSA may be common to macromolecular organic

matter in natural waters; i.e., the average conditional stability con-

stant determined by this method is based on the first few additions of

metal, which probably does not accurately represent the average value

for all ligand sites on the macromolecule. Although the latter was

observed only with copper, it should occur with other metal titrants as

well.



B. Comparison of Five Methods to Determine CL and 8'


1. ASV Titration

Although I could not verify the validity of the ASV method using

model compounds, it is possible that the method gives accurate results

for complexation by natural water macromolecules such as humics. Many

researchers have used it for this purpose, assuming it produces valid

results. Because of its widespread use, the method was included in the

comparative study in spite of the reservations derived from work in the

previous section. If the ASV method produced results that were compar-

able to the other four methods, then perhaps the method works for

natural water organic ligands despite the discrepancies obtained with

model compounds. However, results inconsistent with the other four

methods would cast further doubt on the validity of the ASV method.

The titrations of Waldo and Basin swamps' samples with copper at

pH 6.25 and 0.1 M KN03 resulted in the curves illustrated in Figure

IV-5. The data were treated in two ways, first by the method of Shuman

and Woodward (1977) to determine CL and 8' and additionally by the





















WALDO
SWAMP


z

S20 -

CL



10




0
O 5 10 15 20

40-
BASIN
SWAMP









20 -2





10



I 7 CL= 5.BuM

0 5 10 15
COPPER ADDED, jjM


20 25


Figure IV-5. ASV titration curve of Waldo Swamp and
Basin Swamp samples with copper.








method of Scatchard (1949). The former method produced CL values of

8.5 and 5.8 pM (Figure IV-5) and B' values of 2.5.105 and 5.2-105

for copper with Waldo and Basin swamps' waters, respectively (Table

IV-2).

Because the procedure described by Shuman and Woodward (1977)

applies only to the ASV technique, the data were treated by Scatchard

analysis so that a direct comparison with other procedures could be

made. The data in Figure IV-5 were converted to concentrations of

copper bound, (Cub), and not bound, (Cuf), to the soluble organic

matter of interest. The slope of the curve after the equivalence

point, Su, was assumed to represent the behavior of uncomplexed

metal, and therefore ip values prior to the equivalence point were

divided by Su to determine concentrations of uncomplexed copper. The

concentrations of completed copper were calculated from the difference

between the concentrations of total and uncomplexed copper. Subse-

quently, values of V (where V = [Cub]/[Lt]), and V/(Cuf) were calculated

and plotted (Figure IV-6).

The Scatchard plots were segmented into three sections so that a

comparison could be made among the five procedures. The slope of the

most nearly linear line in each section (determined by linear regres-

sion) is the reported value of a'. Scatchard analysis of both Waldo

and Basin swamps' samples produced values in the lower two ranges of

V only. For the Waldo Swamp sample, similar ' values were

obtained for each range (1.8-105 and 2.3-105; Table IV-2).

These values are similar to B' determined using the Shuman and Woodward

(1977) method of calculation (o' = 2.5-105). Scatchard analysis of

the Basin Swamp sample resulted in two different slopes, and B' values














12
WALDO
SWAMP
10 *

I 8


6 .





2-



0 0.01 0.02 0.03
V


5
BASIN
SWAMP

4\





I2 3









0 0.02 0.04 0.06 0.08


Figure IV-6. Scatchard plots for copper titrations of water
samples using ASV technique. Arrows delimit
linear segments (dashed lines).








Table IV-2. Copper binding capacities and conditional stability con-
stants for water samples and model compounds using ASV.



Copper Molarity
Binding of V
Capacity V Range Range 8' log 8'


Waldo Swamp 8.510-6 M 0-0.02 5.2:10-6 1.8:105 5.25
0.021-0.2 4.7.10-6 2.3"105 5.36
0.21-0.8 --- --- ---

Basin Swamp 5.8-10-6 M 0-0.025 2.2-10-6 1.1-106 6.04
0.026-0.125 5.0-10-6 2.0-105 5.30
0.126-0.32 --- --- ---


were 1.1*106 and 2.0*105 for the two ranges (Table IV-2). By

comparison, the 8' value determined by the Shuman and Woodward method

was 5.2-105. It should be mentioned that the Shuman-Woodward

method of calculation gives 8' values that are based on the early part

of the titration, and hence 8' values calculated thusly should be

related to the initial value determined in a Scatchard plot.

The addition of cadmium and zinc to Waldo Swamp and Basin Swamp

waters resulted in completely reducible species of each metal. There-

fore, no values of CL or 8' could be determined for cadmium and zinc

by the ASV-titrimetric procedure.

The discussion in subsequent sections reveals a significant dis-

crepancy in the Scatchard-derived values of 8' and CL for copper

determined with data from the ASV technique and those values determined

using data from the other methods. Consequently, reasons for this

discrepancy were sought. An underlying assumption stated for the ASV

technique is that the metal-ligand complex be nonlabile (i.e., not

reduced at the mercury electrode) (Shuman and Woodward 1973, 1977). As

shown in the previous section, complexes of copper with a variety of




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs