EPA-600/2-77-186a
September 1977
Environmental Protection Technology Series
EVALUATION OF LEACHATE TREATMENT
Volume I:
Characterization of Leachate
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA--600/2-77-186a
September 1977
EVALUATION OF LEACHATE TREATMENT
Volume I
Characterization of Leachate
by
Edward S. K. Chi an
Foppe B. DeWalle
Environmental Engineering
Department of Civil Engineering
University of Illinois at Urbana-Champaign
Urbana, Illinois 61801
Contract No. 68-03-0162
Project Officers
Dirk Brunner
James A. Heidman
Richard A. Carnes
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. Approval does not signify that the contents necessarily reflect
the views and policies of the U.'S. Environmental Protection Agency, nor
does mention of trade names or commercial products constitute endorsement
or recommendation for use.
11
-------
FOREWORD
The Environmental Protection Agency was created because of increasing public
and government concern about the dangers of pollution to the health and wel-
fare of the American people. Noxious air, foul water , and spoiled land are
tragic testimony to the deterioration of our natural environment. The com-
plexity of that environment and the interplay between its components require
a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solution
and it involves defining the problem, measuring its impact, and searching
for solutions. The Municipal Environmental Research Laboratory develops new
and improved technology and systems for the prevention, treatment, and man-
agement of wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and treatment of
public drinking water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution. This publication is one of the
products of that research; a most vital communications link between the re-
searcher and the user community.
This study involved extensive analysis of different organics and inorganics
present in leachate samples from landfills located in different regions of
the United States. These analyses were then used to predict the effective-
ness of different biological and physical-chemical treatment methods for
contaminant removal. Laboratory bench-scale studies were then performed
to determine the treatability of leachates of different compositions.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
-------
ABSTRACT
The present study involved an extensive analysis of both organics and inor-
ganics present in leachate samples collected from landfills located in dif-
ferent regions of the United States. Bench-scale performance data for
selected leachate treatment methods were obtained. The character of the
analyzed leachate samples was described using ratios of different parameters
such as COD/TOC, BOD5/COD, VS/FS and (total organic carbons present in the
free volatile fatty acids)/TOC. Leachate from recently constructed land-
fills, characterized by high values for the listed ratios, was best treated
by aerobic or anaerobic, biological treatment processes. Physical-chemical
treatment processes, especially activated carbon and reverse osmosis, were
found most effective in treating leachate from stabilized old landfills and
were also useful for removing organics in effluent of biological units.
The leachate samples were subjected to membrane ultrafiltration, gel permea-
tion chromatography and specific organic analysis to separate different
molecular weight fractions and to determine the main classes of organics and
associated functional groups. The majority of the organics were able to
permeate a 500 MW UF membrane, indicative of their low molecular weight.
Most of the organics in the 500 MW UF permeate were free volatile fatty
acids. Fulvic-like material with a relatively high carboxyl and aromatic
hydroxyl group density was found to be the next largest group of organics.
Membrane fractionation and organic analysis of leachate samples collected
from different landfills commonly showed a decrease in the free volatile
fatty acid fraction with increasing age of the fill.
Biological degradation studies showed the sequential removal of different
classes of organics present in leachate from a recently constructed solid
waste lysimeter. Four sequential phases were recognized: 1) removal of
iv
-------
high molecular weight humic carbohydrate-like organics, 2) removal of free
volatile fatty acids, 3) removal of bacterially excreted carbonyl compounds
and amino acids, and 4) removal of high molecular weight carbohydrates pro-
duced during the third phase. A decreased ability of bacteria to degrade
organics was noted for free volatile fatty acids, amino acids, carbohydrates,
humic acids and fulvic acids, respectively.
Analyses of the different heavy metals in the leachate samples showed that
high concentrations are generally present; iron and zinc were found in es-
pecially high concentrations. Determination of metals in leachate by flame
atomic absorption spectroscopy was subject to large matrix interferences that
were circumbented by use of the standard addition technique. Membrane frac-
tionation analysis showed that the majority of the metals permeated the 500
MW UF membrane, indicating that chelation of most metals by refractory
organics in leachate plays a minor role in metal attenuation processes; an
exception was iron, most of which was associated with the 100,000 MW UF re-
tentate. Biological treatability studies showed that the metal concentra-
tions in leachate generally do not cause any toxic effects due to the forma-
tion of hydroxides and carbonates in aerobic systems and sulfides and
carbonates in anaerobic systems.
This report was submitted in fulfillment of Contract No. 68-03-0162 by the
University of Illinois under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period June 30, 1972 to
November 30, 1974, and work was completed as of September 1976.
-------
TABLE OF CONTENTS
Page
Foreword i i i
Abstract iv
List of Figures 1>x
List of Tables x111
Acknowledgements xv
I. Recommendations for Further Research XV1
II. Characterization of Sanitary Landfill Leachates and Its
Significance to Leachate Treatment
Conclusions 1
Introduction 2
Materials and Methods 4
Results and Discussion 7
References 32
III. The Use of Membrane Ultrafiltration and Specific Organic
Analysis for Characterization of Organic Matter in Leachate
Conclusions 36
Introduction 38
Materials and Methods 48
Results 50
Discussion 72
References 79
IV. The Sequential Removal of Organic Matter in Leachate During
Aerobic biological Degradation
Conclusions 85
Introduction 87
Materials and Methods 92
Results and Discussion 94
References 110
V. Metals and Their Interaction with Organic Matter in Leachate
Conclusions 115
Introduction 118
Materials and Methods 126
Results and Discussion 130
References 147
vi i
-------
VI. Evaluation of Colorimetric Tests Used for Characterization
of Organic Matter in Leachate
Conclusions 150
The Determination of Carboyxl Groups with the
Hydroxylamine Test 152
The Determination of Free Volatile Fatty Acids by Gas
Liquid Chromatography 168
The Determination of Carbonyl Groups with the
Dinitrophenylhydrazine Method 174
The Determination of Aromatic Hydroxyl Groups with the
Folin-Denis Method 186
The Determination of Carbohydrates with the Anthrone Test 191
The Determination of Proteins with the Lowry, Ninhydrin
and Kjeldahl Method 196
The Determination of Extractable Organics Using Hexane
and Butanol 204
vm
-------
LIST OF FIGURES
Number Page
1. Changes in Chemical Oxygen Demand, Turbidity, Color and
Conductivity of Leachate During Storage at 5°C in a Capped 10
Bottle
2. Changes in Oxidation Reduction Potential (ORP), pH and
Suspended Solids (SS) of Leachate During Storage at 5°C
in a Capped Bottle 11
3. Changes of Ratios of COD/TOC and Percent of Free Volatile
Fatty Acid Carbon to TOC of Leachate Samples Versus the
Age of Landfills 15
4. Changes of Ratios of VS/FS and BOD/COD of Leachate Samples
Versus the Age of Landfills 17
5. Changes of S04/C1 Ratio, Oxidation Reduction Potential
(ORP) and pH of Leachate Samples Versus the Age of
Landfills 19
6. Eluate of the 500 MW UF Retentate of the UI Leachate
Sample on a Sephadex G-75 Column as Characterized by
Total Organic Carbon, Specific Organics and Functional
Groups 51
7. Effect of Parallel and Sequential Membrane Separation on
the Retention of Organic Carbon in Centrifuges Leachate
by Reverse Osmosis and Ultrafiltration Membranes having
Different Molecular Weight Cutoffs. 54
8. Characterization of Different Membrane Ultrafiltration
Fractions for Specific Organics and Functional Groups 57
9. IR Spectra of the Residue of the Hexane and Butanol
Extracts: (a) Hexane Extract of 10,000 MW UF Retentate
(b) Butanol Extract of 10,000 MW UF Retentate (c) Butanol
Extract of Acidified 10,000 MW UF Retentate (d) Butanol
Extract of 500 MW UF Retentate (e) Hexane Extract of 500
MW UF Permeate 61
10. Characterization of Different Membrane UF Fractions in
a Leachate Sample Collected from a Control (a) and
Recirculation (b) Landfill in Sonoma County 66
11. Eluate of the 500 MW UF Retentate of the Dupage County
Leachate Sample on a Sephadex G-75 Column as Characterized
by Total Organic Carbon, Specific Organics and Functional
Groups 68
IX
-------
LIST OF FIGURES (continued)
Number Page
12. Percentage of TOC Retained with Either 10,000 MW
(a) or 500 MW and 1000 MW (b) UF Membranes in Leachate
Samples Collected from Landfills of Different Ages 70
13. Carbohydrates (a), Proteins (b) and Aromatic Hydroxyls
(c) in Leachate Samples Collected from Landfills of
Different Ages. 71
14. Molecular Weight Distribution of Organics as Reported in
Different Studies and Determined by Dialysis and Membrane
Ultrafiltration 77
15. Elution Pattern of the Ultrafiltration Retentate on
Sephadex G-75 before Aeration of the Leachate with
Activated Sludge 95
16. Change of pH, DO, ORP, TOC and Conductivity During
Aeration of the Wastewater 97
17. Change of Suspended Solids, Turbidity and Filtered
Turbidity During Aeration of the Wastewater 98
18. Change of Proteins, Carbohydrates, Phenolic Hydroxyl-,
Carbonyl- and Carboxyl Substances During Aeration of
the Wastewater 99
19. Elution Pattern of the Reverse Osmosis and Ultrafiltration
Retentate on Sephadex G-75 During Aeration of the Leachate
(a) Corresponding with Maximum Substrate Uptake (b) After
Uptake of the Readily Available Carbon Source 101
20. Change of Filtered Turbidity, DO, pH, TOC and Conductivity
During a Longer Period of Aeration with a Higher Strength
Leachate 107
21. Change of Carbohydrates, Proteins (Amino Acids) and Phenolic
Hydroxyls, Carbonyl and Carboxyl Containing Substances
During a Longer Period of Aeration with a Higher Strength
Leachate 108
22. Application of the Standard Addition Method to Determine
The Actual Heavy Metal Concentration in the Cincinnati
Field Site Leachate Sample 131
-------
LIST OF FIGURES (continued)
Number Page
23. Percentage Depression in the Heavy Metal Determinations
for Leachate Samples, and Synthetic Salt Solutions 132
24. Decrease of COD, Ca, Fe and Zn Relative to the Salt
Content of Leachate Samples with Increasing Age of
Landfills. 138
25. Relative Distribution of Metals in Each of the
Different Molecular Weight Fractions as Obtained by
Membrane Ultrafiltration 140
26. Relation Between the Magnitude of the High Molecular
Weight Organic Fraction and the Amount of Ca and Fe
Retained with the 10,000 MW UF Membrane 143
27. Relation Between the Magnitude of the High Molecular
Weight Organic Fraction and the Amount of Cu and Zn
Retained with the 10,000 MW UF Membrane 145
28. Molar Absorbances of Free Volatile Fatty Acids and
Different Aromatic Compounds 156
29. Absorption Spectra of Free Volatile Fatty Acids (a)
and Branched Free Volatile Fatty Acids at a Concentra-
tion of Approximately 0.25 mmole/1 157
30. Absorption Spectra of Acetic Acid and Different Aromatic
Acids and Hydroxyl Compounds 159
31. Effect of Varying Iron Concentration (a) and Varying
Times Allowed for the Iron Complex Formation (b) 162
32. Effect of Varying Hydroxyl Amine Concentrations (a)
and Varying Times Allowed for the Hydroxamate Formation (b) 164
33. Calibration Curves for C2-Cg Volatile Fatty Acids 171
34. Gas Chromatogram of Free Volatile Fatty Acids Detected
in Leachate of the Boon County-EPA Landfill 173
35. Effect of (a) Incubation Time on the Color Formation of
Pentanone with 2,4 Dinitrophenylhydrazine (b) Effect of
Length of Waiting Time after the Color Formation on
Fading of Color of Acetone 179
-------
LIST OF FIGURES (continued)
Number Page
36. Molar Absorbances of Different Carbonyl Compounds 180
37. Absorption Spectra of Different Aromatic Carbonyl
Compounds 182
38. Absorption Spectra of Different Aliphatic Carbonyl
Compounds and Leachate Fractions 183
39. Spectrum of the Fulvic Acid Fraction of the Dupage
Landfill Leachate and the Standard Tannic Acid 188
40. Absorbence of Different Aromatic Hydroxyl Compounds 189
41. Effect of Hydrolysis Time on Absorbence of Different
Molecular Weight Fractions in Leachate 193
42. Interference in Carbohydrate Analysis by Aromatic
Hydroxyl Compounds 194
43. Effect of Standard Additions of Tannic Acid in Leachate
on the Outcome of the Lowry Test (a) and Relation Between
Results of the Aromatic Hydroxyl Test and Lowry Test as
Determined on Different Leachate Samples (b) 198
44. Effect of Digestion Time on Outcome of Ninhydrin
Protein Test (a) and Relation Between Protein Concen-
tration as Measured with Kjeldahl Method and Ninhydrin
Method Using Digestion 201
45. Butanol Extraction of Tannic Acid, Starch and Bovine
Albumin 205
46. Hexane Extraction of Free Volatile Fatty Acid Mixture 207
47. Butanol Extraction of Acidified Free Volatile Fatty
Acid Mixture 208
xi i
-------
LIST OF TABLES
Number Page
1. Description of Sites from Which Leachate Samples were
Collected 5
2. Composition of Leachate from Different Sources as
Measured by Different Researchers 8
3. Composition of Leachate from Different Sources as
Measured at the University of Illinois 13
4. Results of Treatment Efficiencies Obtained in Different
Biological Treatment Studies 22
5. Results of Treatment Efficiencies Obtained in Different
Physical Chemical Treatment Studies 25
6. Proposed Relationship between COD/TOC, BOD/COD, Absolute
COD and Age of Fill to Expected Efficiences of Organic
Removal from Leachate 30
7. Dialysis Studies Characterizing the Molecular Weight
Distribution of Aquatic Organics 42
8. Membrane Ultrafiltration Studies Characterizing the
Molecular Weight Distribution of Aquatic Organics 44
9. Organic Matter Concentrations in Landfill Leachate
and their Relative Composition 47
10. Chemical Analyses of the Sephadex G-75 Eluate of the
500 MW UF Retentate 52
11. Effectiveness of the Hexane and Butanol Extraction to
Remove Organics from the Different Molecular Weight
Fractions 60
12. Infrared Absorption Bands and their Tentative Assignments
for the Residue of the Hexane and Butanol Extracts 62
13. Organic Composition of Humic Substances Fractionated
According to Molecular Weight 74
14. Chemical Interferences Observed in the Metal Determinations
by Flame Atomic Absorption Spectroscopy 122
xi 11
-------
LIST OF TABLES (continued)
Number Page
15. Synthetic Leachate Samples Used to Evaluate Matrix
Interferences in the Heavy Metal Determination
Containing 1 mg/1 of Cr, Ni, Cu, Pb and Cd 128
16. Chemical Characteristics of Leachate Samples in which
the Heavy Metals were Tested by the Standard Addition
Method 134
17. Metal Concentrations in Leachate Samples Collected
from Landfills Having Different Ages 137
18. TOC in Redissolved Dried Residue of the Hexane and
Butanol Used to Extract a Free Volatile Fatty Acid
Solution; the Residue was Redissolved in 100 ml
Distilled Water 209
xiv
-------
ACKNOWLEDGEMENTS
The support of the US-EPA Solid and Hazardous Waste Research Division
of the Municipal Environmental Research Laboratory in Cincinnati is
acknowledged with sincere thanks. The chemical analysis and treatment
studies were performed by Richard L. Davison, Laboratory Chemist, and
the Research Assistants J. T. Y. Chu, Y. Chang, G. Velioglu, and F. M.
Saunders. Invaluable help was also provided by Laboratory Assistants
B. MacPherson, J. Hansen, C. Stroupe, T. Brozozowski, M. Sweeny, P. J.
Strange, B. Clark, G. S. M. Yu, J. Young, and J. Hunsicker. The follow-
ing researchers assisted to a great extent the facilitation of leachate
sample collection: Reinhardt and Ham, County of Sonoma andEmconAssociates,
Solid and Hazardous Waste Research Laboratory, Mao and Pohland, Apgar and
Langmuir, Fungaroli, and Hughes, et al.
xv
-------
I
RECOMMENDATIONS FOR FURTHER RESEARCH
Considerable spread was observed in some of the results of grab sample
analyses from different landfills when these analyses were correlated
with the age of the landfill. The cause of these variations are suspected
to be the individual characteristics of the different landfills and the
local climatic conditions. Those factors affecting the trend should be
more accurately defined to allow prediction of future impacts on the
environment and the long-term efficacy of leachate treatment methods.
Additional, more specific chemical analysis of the organic fractions
identified is recommended to ascertain the presence of environmentally
significant organics and the presence of organics significant in leachate
treatment.
Additional basic research is necessary to more clearly define the limita-
tions of the different separation techniques such as membrane ultrafiltra-
tion and gel permeation chromatography using naturally occurring organics
of different molecular weight and chemical composition. This should be
followed by extensive degradation and identification studies to more
clearly define the chemical nature of each of the different molecular
weight fractions.
xvi
-------
II
CHARACTERIZATION OF SANITARY LANDFILL LEACHATES
AND ITS SIGNIFICANCE TO LEACHATE TREATMENT
CONCLUSIONS
The present study conducted an extensive analysis of both organics and
inorganics present in leachate samples collected from landfills located
in different parts of the United States. Several ratios were determined
for leachate such as COD/TOC, BODg/COD, VS/FS, and total carbon present
in the free volatile fatty acids to TOC. These ratios reflected results
of the extensive analysis of the organic matter in leachate and can be
used to predict the effectiveness of either biological or physical -
chemical treatment methods with a given leachate. Some of these ratios,
such as COD/TOC, are also used as an internal check on the reliability
of the results of chemical analysis of leachate samples.
When the organic analysis was related to the leachate treatability, it
was noted that leachate collected from recently leaching landfills is
best treated by aerobic or anaerobic biological treatment processes as
they are most effective in removing the free volatile fatty acids that
are present in large quantities. Physical-chemical treatment processes
are most effective in treating leachate from stabilized landfills or to
further remove organic matter in effluent of biological units treating
leachate. Of all the physical-chemical processes evaluated, activated
carbon and reverse osmosis gave the best removal of organic matter.
-------
INTRODUCTION
The sanitary landfill method for the ultimate disposal of solid
waste continues to be widely accepted and used. To minimize environ-
mental insult, it is necessary that the selection of the site be given
special attention and that the site be properly designed, constructed,
and operated. After the solid waste is placed in the fill, problems
may still develop. One such problem is caused by infiltrating rainwater
and the subsequent movement of liquid (leachate) out of the fill into
the surrounding soil. This leachate often contains a high concentration
of organic matter and inorganic ions, including heavy metals. A number
of incidences have been reported where leachate has contaminated the
surrounding soil and polluted an underlaying groundwater aquifer or
nearby surface water. In one of the recent cases of groundwater contam-
ination by leachate, private wells located 300 m downstream from the
Llangollon landfill (completed and closed in 1968) in New Castle County,
Delaware, were heavily polluted and subsequently abandoned (R. F. Weston
Inc., 1974). The municipal water supply from Genesco, Illinois, was
polluted by leachate from a garbage dump located about 500 m north of
the well field, while private wells located about 200 m from a landfill
in Kane County, Illinois, showed bacterial and chemical pollution
(Walker, 1969). One way to avoid or correct such situations is to
collect and treat leachate. Treatability of leachate, however, is
related to its chemical composition, especially the nature of the organic
matter.
The composition of leachate has been investigated in several studies.
One of the first studies, reported by Merz (1952), was concerned with
inorganic ions leached out of incinerator ash dumps. Other studies,
published later, were conducted with leachate generated from sanitary
landfills handling municipal solid waste (1954). Lysimeter studies,
conducted by Qasim and Burchinal (1970), Fungaroli and Steiner (1971),
-------
Pohland (1972), and Pohland and Maye (1970), determined organic matter
in addition to the inorganic constituents by analyzing the leachate for
COD, BOD, free volatile fatty acids, organic-nitrogen, and in one study
lignins and tannins.
An extensive chemical analysis of leachate generated from landfills
located in different parts of the United States was conducted at the
University of Illinois. Particular emphasis was given to detailed
chemical analysis of both the organic constituents and heavy metals.
Since the methods developed for organic analysis are elaborated and time-
consuming (DeWalle and Chian, 1974), a number of parameters were identi-
fied to represent the more complex actual composition of organic matter
in leachate. Laboratory-scale biological and physical-chemical treatment
processes were evaluated for their ability to remove the individual con-
taminants in leachate. Leachate from the pilot-scale lysimeter installed
at the University of Illinois was used throughout the treatability study.
-------
MATERIALS AND METHODS
Representative leachate samples from a number of completed lysimeters
and pilot- and field-scale sanitary landfills were carefully collected
and stored at refrigeration temperature (5°C) for analysis. These sites
are located in the United States at representative locations and cover a
range of temperature and precipitation.regimens (Table 1).
The gross properties of leachate, such as chemical oxygen demand (COD),
total organic carbon (TOC), pH, conductivity, oxidation reduction poten-
tial (ORP), turbidity, suspended solids (SS), volatile solids (VS) and
nitrate (N03-N) etc., were immediately analyzed upon arrival in the
laboratory according to procedures outlined in Standard Methods (American
Public Health Association, 1971). These were followed by extensive
analysis of the organic matter and heavy metals in leachate.
A unique and comprehensive fractionation scheme was developed for the
concentration and separation of organic compounds of various molecular
weights using ultrafiltration and reverse osmosis membranes and gel per-
meation chromatography. A Hewlett Packard 5750 B gas chromatograph
equipped with a column containing a NPGS liquid phase was used for the
analysis of free volatile fatty acids. The heavy metals were determined
with atomic adsorption using a Beckman Unit, Model 485 (Fullerton, Ca.).
It was found that any analytical methods relying on colorimetric tech-
niques were undesirable with the undiluted leachate due to the interfering
color contributed by leachate, unless extensive dilution of samples was
practical (Chian and DeWalle, 1975). Standard addition methods were
employed to determine the magnitude of the interfering effect.
The leachate used to evaluate its treatability by biological and physical-
chemical processes was obtained from a laboratory lysimeter (1.52 m
diameter by 3.04 m high) containing 1620 kg of shredded (3.8 cm grate
-------
TABLE 1.—Description of Sites fromWhich Lcachate Samples were Collected
Location
(1)
University of Illinois
Urbana, 111.
Oscar Mayer Treatment
Plant Madison, Wise.
County of Sonoma,
Santa Rosa. Calif.
Cincinnati Boon County
EPA landfill, Ohio
Georgia Institute of Tech-
nology, Atlanta, Ga.
Pennsylvania State Uni-
versity, College Park,
Drexel University.
Kennett Square, Pa.
Winnetka Landfill
Illinois
Dupage County Landfill
Illinois
Code
(2)
UI
UMC
MUNC
SONCON
SONREC
CINC
ATLCON
ATLREC
PACP-67
PHKS-12
LW-17
LW-5B
LW-5B
•Source: U.S. Department of Commerce
Solid
waste fill
characteristics
(3)
milled; indoor
lysimeter
unmilled. covered:
lysimeter
milled, uncovered:
lysimeter
control ; pilot fill
recirculation pilot fill
unmilled; pilot fill
control; lysimeter
recircuiation
lysimeter
unmilled; pilot fill
unmilled; pilot fill
field scale fill
field scale fill
field scale fill
Date
solid Date
waste samp.e
placed collected
(4) (5)
December 1972 February 1, 1973
August 1972 November 20, 1973
August 1972
October 1971 July 9, 1973
June 1971 October 2, 1973
September 1971 December 20, 1973
September 1971
November 1967 February 17, 1973
May 1966 February 15, 1973
December 1967 October 30, 1972
September 1956
Februarv 1959
Weather Characteristics"
Average
temperature,
in
degrees
Celcius
<6)
10.9
7.2
7.8
12.7
16.0
8.1
12.3
9.4
Rainfall,
in
millimeters
(7)
1.012
768
153
1,017
1,228
9SO
1,118
S06
References
(8)
This study
Reinhardt and Ham
(1973)
County of Sonoma
Emcon AssccM973)
Solid and Hazardous
Waste Research
Laboratory (1973)
Mao and Pohland
11973)
Apgar and Langmuir
(1971)
Fungaroli (197 1)
Hughes, et al.(1971)
National Oceanic and Atmospheric Administration. Environmental Data Service. National Climatic Center, Ashville. N.C.
-------
opening) refuse obtained from the Hartwell section of Cincinnati, Ohio.
It was compacted to 0.33 ton/m3 (556.9 pound/yd3). Water was applied
daily to bring the solid waste to field capacity within 30 days, and
thereafter an equivalent of 0.89 mm precipitation per week was added to
generate sufficient leachate for evaluation of leachate treatment methods,
-------
RESULTS AND DISCUSSION
The composition of leachate samples from different sanitary landfills,
as reported in the literature, show a large variation. Table 2 summar-
izes these ranges of leachate composition; these data show that the age
of the landfill, and thus the degree of solid waste stabilization, has a
significant effect on the composition of leachate. Other factors that
contribute to the variation of data are solid waste characteristics (e.g.,
the composition and size of the waste and degree of compaction), the
moisture content and degree of rainwater infiltration, temperature,
sampling technique, and analytical methods. Other factors such as land-
fill geometry and interaction of leachate with its environment prior to
sample collection also contribute to the spread of data. Most of these
factors are rarely defined in the literature, making interpretation and
comparison with other studies difficult, if not rather arbitrary.
The time lapse between sample collection and analysis is one such factor
that is seldom defined. Figures 1 and 2 depict changes in chemical
oxygen demand (COD), turbidity, color (absorbance at 400 nm of the 1:10
diluted sample), pH, suspended solids (SS), oxidation reduction potential
(ORP), and specific conductivity of leachate collected from the lysimeter
at the University of Illinois. Analyses were performed immediately after
sampling and frequently during storage at 5°C in a capped bottle for a
twelve day period (17,280 minutes) during which time small aliquots were
removed for analysis. It is seen from these figures that the ORP showed
the greatest change and became more positive immediately after sampling.
However, it stabilized after approximately one day (1440 minutes), indi-
cating that this parameter should be determined immediately at the time
of sampling. The other parameters such as SS, color, and turbidity
started increasing in magnitude after approximately one day and, unlike
ORP, they continued to increase with time. The sharp rise of these
parameters shown in Figure 1 and 2 is amplified by the use of the loga-
rithmic time scale. As expected, these changes were observed to occur
more rapidly when the sample was exposed to aerobic conditions and stored
at room temperature (Chian and DeWalle, 1975).
-------
Table 2. Composition of Leachate from Different Sources
as Measured by Different Researchers
oo
Source
Parameter
Tine Span
(year)
COD
BOD 5
TOC
pH
TS
TDS
TSS
Specific
Conductance
Alkalinity
(CaCOa)
Hardness
(CaC03)
Total P
Ortho-P
NH4-N
N03+N02-N
Ca
Cl
Na
Sulfate
Kn
Mg
Fe
Zn
Cu
Pb
SHWRL (1973)
Cincinnati
1
16,000-22,000
7,500-10,000
5.2-6.4
10,000-14,000
10,000-14,000
100-700
6,000-9,000
800-4,000
3,500-5,000
25-35
23-33
0.2-0.8
900-1,700
600-800
450-500
295-310
400-650
75-125
160-250
210-325
10-30
NOTE: All figures in rng/t
* Based on 20 Day 30D
Fungaroli
(1973)
Drexel U.
2
1,000-51,000
3.7-8.5
0-42,000
10-26,500
0-9,700
0-5,500
0-130
0-482
4.7-2,340
0-7,700
25-450
0-1,716
0-167
0.9-9
Reinhart and
Ham (1973)
3.5
2,700-10,650
1,550- 3,450
5.9-3.1
4,028- 7,790
1 ,800
1,400
12
80
400
20
330
5.5
Hughes
et al .
(1971)
Dupaqe
LH-5B
5.5
8,000
14,000*
6.3
6,794
5,810
2,200h
1.2
0.7A
308
1,330
810
610
2
0.06
450
6.3
0.4
v-0.5
<0.05
0.5
Huahes
et'al.
(1971)
Dupaqe
M:-i-63
8.5
2,940
4,550*
5,910
4,720
2,250+
0.17
0.5A
447
946
615
220
1
0.09
725
12
0.05
<0.5
<0.05
1.0
Huahes
et'al .
(1971)
Dunace
MM-61
14.5
360
125*
1,104
1,630
690+
0.3
0.14A
156
205
63
85
i
0.24
110
106
0.10
<0.5
<0.05
1.0
Hughes
et al.
(1971)
Dunaqe
LW-6i:
16.0
40
225*
7.0
1,198
2,250
540+
8.9
1.6
102
135
74
100
2
0.06
90
0.6
4.5
<0.5
-0.05
1.0
except Specific Conductance which is measured as pmhos/cm and
Hughes
et"al .
(1971)
Winnntka
LH-17
2.1
4,260
6,500*
2,306
3,370
2.'s
0 5A
"100
429
298
168
13.6
1.14
198
96
1.5
<0.5
<0.05
1.0
pH as pH
Hughes
et el.
(1S71)
Winnetka
LK-5B
12.5
415
250*
994
3,030
970+
2. 75
0.43A
109
701
348
220
1
0.1
200
13.6
0.6
<0.5
<0.05
0.5
units
Hughes
et al.
(1971)
'•.'inr-etka
LW-13
1C. 1
153
105*
584
1 ,450
71 0+
1 .3
0.2A
'l09
70
34
39
i;
•J
0.2
75
11
0.1
<0.5
<0.05
1 0
1 * U
Merz
(1974)
Riverside
B-l
1 . O
81-33,100
R fi - 7 f,
•J . U / .O
730-9.500
'650-8,120
I; . i. ~ L. y
On o (-..-v
.2-3^0
115-2,570
96-2,350
85-1 ,805
23-1,860
64-410
6.5-305
* Blackwell value calculated from Hg and Ca Concentration
A Using
only NO analyses, NO not
determined
-------
Table 2 continued
V£>
Meichtry (1971) Sonotiia (1973)
Mission Canyon
County of County of
Parameter
Time Span
COD
B005
TOC
1 WU
pH
TS
TDS
TSS
Specific
Conductance
Alka 1 inity
(,CaC03)
Hardness
Total P
Orthc P
N03-I-N02-N
Ca
Ci
Na
K
S04
Mn
Mg
Fe
Zn
Cu
Cd
Pb
3-18-68
>3
76,800
10,900
5.75
44,900
172
9,860
22,800
0.24
o
7,200
660
767
68
1,190
15,600
2,820
3-24-71
>6
3,042
903
7.4
13,409
220
8,677
8,930
0.65
270
216
2,355
1,160
440
19
8,714
4.75
Control
2
3,260-22,400
2,250-19,200
4.7-5.4
724-14,080
16-200
240- 3,920
0-2.8
0-81
0-0.90
20-1,082
55- 490
80- 338
4.8- 1 48
20- 250
2.6- 608
22-1,050
0.23-9.0
0.2-0.44
0
Recirculate
2
22,700-89,520
15,500-33,600
4.6-6.5
14,600-21,010
22-600
1,680- 7,900
1.4-79.2
880-194
0-4. It,
561-1 ,800
920-1 ,565
408-1 ..010
260-910
257-1 ,0<0
316-672
165-300
85-95
<0.2-0.4
<0. 05-0. 1 6
<0. 0-2.0
Qasiir. and
Burchinal
(1970)
Cylinder C
0.33
24,200-33,360
5.25-5.88
36,250-59,200
16,200-20,850
8,580-10,950
87-128
828-1.106
2,405-2,790
1,465-2,310
871-1,439
2,970-3,770
320-768
325-420
546-860
Pohland
(1972) Ga.
Inst. Tech
Control
3
4,320-12,000
2,500-11 .COO
1,230- 5,000
5.2-5.6
2,442-12,500
34-610
558-2,280
450-1,940
2.8-26
56-137
125-750
98-385
64-143
81-156
3-1U
26-75
9-95
Pohland
(1973) Ga.
Inst. Tech
Recircul ate
3
4,280-9,288
2,750-6,900
256-2,793
4.8-5.4
1,627-6,918
12-385
302-1 ,370
370-1 ,040
0.65-22
68-114
60-435
Sl-2^3
62-109
1 2-1 33
4-65
17-63
4-110
Range of
A"! ! Val ves
40-89,520
81-33,360
256-28,000
3.7-8.5
0-59,200
584-44,900
10-700
2,810-16,300
0-20,350
o -?_,-? oo
0-22,800
0-130
6. 5-85
0-1,106
0.2-10.29
60-7,200
4.7-2,467
0-7,700
2C-3 .770
1- 1 ,Cbci
0.03-125 -
17-15.600
0-2,820
On "i n
- J / U
0-9. 9
0.03-17
<0. 10-2.0
-------
10-
100 1000
Time, min.
IOOOO
Figure 1. Changes in Chemical Oxygen Demand, Turbidity, Color, and Conductivity
of Leachate During Storage at 5°C in a Capped Bottle
-------
1000-
500-
10
--IOO
5.9
5.8
5.7
x
a.
5.6
5.5
100 1000
Time, min,
10,000
Figure 2.
Changes in Oxidation Reduction Potential (ORP) pH and Suspended Solids
(SS) of Leachate During Storage at 5°C in a Capped Bottle
-------
The changes that took place with the color parameter corresponded to a
visual change of the sample color from light yellow to dark brown, as
reported by many other investigators, and are caused by the oxidation
of the ferrous ion to the ferric form. The latter then forms ferric
hydroxide colloids which contribute to the brownish color of leachate
and the increase in suspended solids containing large quantities of iron.
As a result of this, turbidity also increases with time of storage.
Since a daily sample was taken from the container, exposure of the
leachate remaining in the container to atmospheric oxygen is the major
cause contributing to the decrease in COD and the increase in suspended
solids. Storage under anaerobic conditions and at low temperature is
therefore necessary. Thus, the necessity of establishing strict sam-
pling procedures and testing schedules for leachate analysis is evident.
Since the samples had to be collected from different parts of the United
States, it was not practical to conduct on-site measurements; the most
sensitive parameters however, were measured in the laboratory within 48
hours after collection. The remaining gross properties of leachate
samples were analyzed within two weeks after sample collection while
being stored at 5°C.
Results of chemical analysis of leachate samples collected from the
different sites are presented in Table 3. It 1s seen from Table 3 that
there is a wide variation in concentrations of organic matter in leachate
as measured by the gross properties of COD, TOC, and BODg. Values of COD
vary from 71,670 mg/1 with leachate sample "MUNC" obtained from an 8-
month old uncovered landfill containing mulled refuse to 412 mg/1 with
sample "LW-17" obtained from an approximately 17-year old landfill. The
concentration of inorganics as measured by fixed solids (FS) also
12
-------
TABLE 3.—Composition of Leachate from Different Sources as Measured at the University of Illinois (Ul)
Sourc*
parameter.
in
milligram!
par liter
(1)
COD
TOC
BOD,
Bicarbonate alkalinity
(CaCO,)
pH (pH unit)
Conductance, in micro-
mhos per centimeter
ORP, in millivolts
Turbidity, in Jackson
lurbidit) units
SS
FSS
TS
FS
Org-N
NH4-N
NO -N
NOj-N
Total-P
Ortho-P
Cl
Ca
Mg
Fe
Na
K
Ul
(2)
49,300
17,060
24,700
668
563
13.700
-60
75
139
92.5
33.989
15,5*6
544.7
392.6
0.5
BDL"
21.5
6.5
I.IIO
1.480
3,750
650
2,200
1.360
1,140
MUNC
(3)
71.680
27,700
57.000
459
5.97
16.800
-132
84)
202
167
55.348
22.895
945
1,028
10.25
O.O4
98
29
1,558
2,467
3.900
1.140
1,046
1.580
2.300
UMC
(4)
16.580
5,906
9.960
284
5.59
5,420
-220
270
192
110
7.930
3,475
78.5
347.4
4.25
0.04
85
85
77
474
572
220
91
330
900
SONCON
(S)
12.800
4.476
8.100
1.576
5.09
10.000
+68
110
55
26.5
12,579
5,823
1074
2
7.12
BDL
1.20
0.80
181
462
1.030
515
301
240
74
SONREC
(6)
12.050
3.921
9,300
1,265
5.99
12.500
-95
96
204
91
14.643
6,647
257.4
254.5
1.80
BDL
28.8
22
176
1.010
1.220
750
540
675
300
CINC
(7)
45,750
13.840
22,000
175
525
9.450
NM*
71
8.9
75
32.145
13.603
31
247.7
9.8
0.19
31.6
28
909
2,0%
2.325
530
735
750
860
ATLCON
(8)
7,021
3,273
4.410
142
5.72
2,810
+72
23
472
277
4.120
2,273
27.8
49.2
6.5
BDL
1.85
1.7
103
lift
520
52.5
810
75
620
ATLREC
(9)
134
93.2
6.4
1,025
6.62
978
+ 130
61
68
36
911
RI2
1.4
0.4
BUL
0.5
0.25
30.3
60.6
133
35
U.5
147
•Not Measured.
••Below Delectable Limits.
PACP-
67
(10)
2.062
'MO
I. ISO
1 49
6.46
l.5*>
-7'
43
163
1375
2,091
1.294
4.4
21.3
0.90
BDl.
0.70
0.55
19.5
50.2
215
75
560
17.5
35
PHKS-
12
(ID
3.X06
1,303
nnC
I5K
6.37
3.115
-75
58
923
901
4.225
2.706
2(1.1
62.2
1 23
BDl.
4.35
34
29
265
330
S8
330
113
70
I.W-5B
(12)
412.1
210
55.2
3,520
7.15
4,515
-H63
625
386
246
2.704
2.072
8.4
299.5
0.55
0.05
138
3.0K
7.4
715.5
76
186
40
370
450
LW-17
<13i
31.1
70
3.9
1.301)
^.25
1.4(10
•i97
I3li
226
1:8
1 .458
1.144
3.2
3.S
0.45
(l.0f>
4.%
2 73
3'J.5
IV7.4
2M
HI
1 5
91
39
-------
fluctuates from a high value of 22,895 mg/1 to a low value of 812 mg/1.
Other organic and inorganic contaminants also follow the trend of a
decrease in strength with the increase in age of the landfills.
Since the concentration of the different pollutants will show a consid-
erable variation on a day-to-day basis, it is more beneficial to use the
ratios of different parameters instead of their absolute values. Further
analysis of data in Table 3 reveals that many ratios of chemical prop-
erties, such as COD/TOC, BOD/TOC, VS/FS, S04/C1, etc., reflect the com-
position of the organic matter in leachate and are in turn related to the
age of the landfill. The parameter found to be useful in relating to the
composition of organic matter was the ratio of chemical oxygen demand to
total organic carbon (COD/TOC) (DeWalle and Chian, 1974).
The change in the COD/TOC ratio with the age of the landfills or lysi-
meters is shown in Figure 3a. These ratios were calculated and plotted
from data shown in Table 3. It can be seen from Figure 3a that the COD/
TOG ratio tends to decrease as the age of the landfill increases and this
ratio varied from 3.3 for a relatively young fill to 1.16 for an old fill.
The maximum possible COD/TOC ratio for several organic compounds is 4.0,
and it can be as low as 1.3 for organics containing carboxyl groups
(Rickert and Hunter, 1971). A decrease in this ratio in the leachate
samples represents a more oxidized state of the organic carbon which
becomes less readily available as an energy source for microbial growth.
These organics are generally degradation products of microbial activity
and increase as the age of the landfill increases. The resulting leach-
ate thus becomes less amenable to biological treatment.
The decrease in COD/TOC ratio is reflected by a gradual change of the com-
position of the organic matter as measured with theBOD^or free volatile
fatty acids test. Since the biochemical oxygen demand (BOD) test 1s
predominantly a biological test, it generally reflects the b1odegradab1lity
of the organic matter in leachate and is, therefore, in itself a direct
14
-------
Time, years
Figure 3. Changes of Ratios of COD/TOC and Free Volatile Fatty
Acids as Percentage of TOC of the Sample Versus Age of the Landfill
15
-------
measure of the treatability of leachate by biological processes. Similar
to COD/TOC ratio, the BOD/COD ratio shows a decrease as the age of the
landfills increase (Figure 3b). This relationship corroborates results
from a recent study of Miller ejt aj_., (1974). The calculated ratio of
BOD/COD based on his data showed a decrease from 0.47 to 0.07 within a
period of twenty-three years. They are compared to a decrease from 0.8
to 0.048 as found in this study within a time span of seventeen years
(Figure 3b). The BOD test, however, is subject to more variation than
the TOC analysis, since the value obtained depends on the dilution of the
leachate and the amount of seed added. Readily oxidizable inorganics
such as Fe will contribute to both COD and BOD values; however, their
effect was found to be relatively small. It is even further reduced by
taking the ratio of the two parameters.
Specific organic analysis showed that the free volatile fatty acids
constituted the main fraction present in leachate. Since free volatile
fatty acids are readily biodegradable (Chian and DeWalle, 1975), a
decrease in the ratio of carbon present in free volatile fatty acids to
TOC corroborates the decrease in BOD/COD ratio. Figure 4b depicts the
decrease.of this ratio from 0.49 to 0.05 with increasing age of the
landfill. Measurement of the individual volatile acids using GC, showed
a relative increase in odd numbered fatty acids. The ratio of carbon
present in the odd- and even-number volatile free fatty acids found in the
present study increased from 0.27 to 0.55 during the first two-year
period when appreciable amounts of fatty acids are present in leachate.
Acetic acid is the main catabolic product produced by microorganisms
during anaerobic fermentation. This is followed by condensation reac-
tions to form other acids of higher molecular weight such as butyrate.
The formation of odd numbered fatty acids occurs when repeated catabolic
and anabolic metabolisms proceed in the presence of microorganisms
responsible for producing them. The increase in the odd-number fatty
acids, such as propionic and valeric has been reported by Burchinal
(1970), Songonuga (1969) and Thompson (1969) to form at the later stage
of leachate production.
16
-------
0.5
4b
10
Time, years
15
Figure 4. Changes of Ratios of Volatile Solids/Fixed Solids (VS/FS) and
BOD/COD of Leachate Samples Versus the Age of Landfills
17
-------
Due to its initially biodegradable nature, the organic compounds decrease
more rapidly than the inorganics with increasing age of landfill. The
decrease of organics is due to anaerobic methane fermentation and washout.
The inorganics only decrease due to washout by infiltrating rainwater.
The ratio of total volatile solids to total fixed solids (VS/FS) there-
fore decreases with the age of the fill (Figure 4a). It is seen that
this ratio decreases gradually from a value of 2.0 for a relatively
young fill to 0.2-0.25 for old landfills. Such a decrease was in corrob-
oration to results of Qasim and Burchinal (1970). It was found that the
calculated ratio of this parameter decreased from 2.06 to 1.64 and from
1.88 to 1.45 during a period of 121 days and 163 days with leachate
obtained from simulated landfills having a depth of 3.5 m and 2.3 m,
respectively.
Other parameters, such as sulfate to chloride ratio, oxidation reduction
potential (ORP) and pH, also reflect the degree of stabilization of the
landfill and thus the leachate generated from the fill. Figure 5a
depicts the decrease of the ratio of sulfates to chlorides in leachate
with respect to the age of the landfill obtained from this study. This
rapid decrease is attributed to the decrease in concentration of sulfates
as a result of anaerobic conditions prevailing in the landfill in which
sulfate is reduced to sulfide. The latter is precipitated out with iron
and other heavy metals. A similar decrease in this ratio was calculated
from data reported by Pohland (1972) with leachate generated from both the
control and the recirculating landfills. Since chlorides represent a
nonbiodegradable parameter, it can be used to assess the extent of
leachate contamination (Anderson and Dornbush, 1967).
The trend in the S04/C1 ratio is inversely related to the ORP parameter.
The largest decrease in SCL/C1 ratio corresponds with the lowest ORP
values confirming the high degree of anaerobiosis in the landfill. The
increase in ORP readings with time, as shown in Figure 5b, also reflects
the degree of stabilization of leachate. Figure 5c shows the increase
18
-------
1,0
0,8
0,6
0,4
0,2
5a
a.
-------
in pH of leachate with time, which reflects the decrease of the concen-
tration of the partially ionized free volatile fatty acids.
The potential use of these parameters is twofold. It can be used to
establish a generalized trend between the leachate composition and the
age of the landfill, and can be employed as an internal check on the con-
sistency of the analytical results. The COD/TOC ratio, for example, can
be used to determine the reliability of either the COD or TOC measure-
ments. If the results of the analysis of a given leachate sample yield
a ratio of COD/TOC greater than the theoretical maximum value of 4.0 for
soluble organics, either the COD or TOC value is in question.
The ultimate goal of relating these parameters to the composition of
organic matter and the age of the landfill, is, however, to establish
useful criteria for the choice of specific treatment processes best suited
for the removal of organic contaminants from leachate. Several investi-
gators have studied the treatment of sanitary landfill leachate and prom-
ising results have been obtained with a number of treatment processes.
Review of results of their studies indicates that the specific process
suitable for the treatment of a given leachate is related to its chemical
composition which is in turn related to the degree of stabilization of
the refuse or the age of the landfill. The second phase of this study
was therefore conducted to relate the performance of each treatment method
to the composition of the organic matter in leachate. Since this approach
is elaborate and time-consuming, several leachate parameters were selected
to predict the effectiveness of each treatment process. Determining the
value of these parameters and knowing the age of the landfill will enable
the design engineer to select suitable treatment processes for the
removal of organic matter in leachate.
Understanding the removal efficiencies of each treatment process requires
a basic knowledge of the mechanisms resulting in the removal of specific
organic fractions in leachate. Chian and DeWalle (1975b) studied the
20
-------
aerobic biodegradation of leachate collected from a lysimeter having an
age of five months and showed four distinct phases of substrates utili-
zation by microorganisms. Carbohydrates were found to be utilized in
the first phase followed by uptake of fatty acids in the second phase.
Catabolites, such as amino acids, accumulated during the active phase
of acid metabolism, were found to be assimilated in the third phase. The
humic carbohydrate-like materials, having a molecular weight greater than
50,000, which were accumulated in the third phase, were removed in the
fourth phase. The resulting residual refractory organic matter remaining
beyond the fourth phase consisted mainly of fulvic acid-like materials
with a molecular weight predominantly between 500 and 10,000. DeWalle
and Chian (1974) showed that the high-molecular-weight humic-carbohydrate
fraction affected bacterial flocculation. This high-molecular-weight
fraction was poorly removed by activated carbon treatment of the activated
sludge effluent, while the low molecular weight fulvic acid-like
materials were removed effectively (DeWalle and Chian, 1974b). Both
the humic and fulvic substances are relatively inert to biological
degradation. Based on these findings, it is anticipated that leachate
generated from young fills in which the organic matter mainly consists
of free volatile fatty acids, can be readily degraded by biological means.
Leachate from old fills containing organics corresponding to the refrac-
tory material excreted by microorganisms in the later two phases of
substrate utilization, is more amenable to physical-chemical treatment.
Aerobic biological treatment of leachate has been studied by many inves-
tigators as summarized in Table 4. Between 58 and 99% removal of COD
was accomplished by either aerated lagoon or activated sludge systems.
The range of BOD/COD of leachate fed to these systems varied between
0.45-0.81. All leachate samples employed in these studies were obtained
from recently installed landfills and lysimeters. Leachate treatment
using anaerobic digesters or anaerobic filters resulted in COD removals
of 87 to 99% (Table 4). The BOD/COD ratio's of the leachate samples
used for these studies ranged from 0.45 to 0.81. Results of the present
21
-------
TABLE 4.—Result of Treatment Efficiences Obtained in Different Biological Treatment
Studies
Bio-
logical
process
(1)
Aerobic
Anaerobic
Aerobic/
Anaerobic
Author
(2)
Boyle and
Ham
Roy Weston
Inc.HW
This study
Boyle and
Ham
Rogers (1*73)
This study
Boyle and
Ham »»74>
Foree and
Reed .(!»»>
This study
Ini-
tial
COD
(3)
8.8(10
15.800
3.550
500
139
30,000
10.600
12,900
16,500
5,500
1,300
30.000
—
510
1,000
BOO/
COD
W>
0.80
COD /
TOC
(5)
—
0.45
0.64
0.52
0.03
0.65
0.79
0.45
0.62
0.78
0.81
0.65
0.18
—
—
3.45
3.20
1.56
2.1
2.9
_
281
2.92
2.82
—
2.90
—
2.53
2.35
Treatment
system
(6)
Aerated lagoon
Aerated lagoon
Aerated lagoon
Aerated iagoon
Aerated lagoon
Aerated lagoon
Anaerobic
digester
Anaerobic
digester
Anaerobic
digester
Anaerobic
digester
Anaerobic
filter using
lime treated
leachate
Anaerobic
filter
Aerated lagoon
treatment of
anaerobic di-
gester efflu-
ent
Aerated lagoon
treatment of
anaerobic fil-
ter effluent
Aerated lagoon
treatment of
anaerobi< fil-
ter effluent
Percent-
age
1 COD
re-
moval
(7)
74
98
77
58
0
99
93
92
99
93
87
97
40
22
17
Detan-
tion
time
'81
5d
10 d
0.6 d
0.3 d
7.7 d
Id
10 d
10 d
lid
10 d
1.2 d
27 d
*d
1 d
Id
22
-------
study on treatment of leachate with activated sludge, aerated lagoon
and anaerobic filter agree well with the above studies in that leachate
collected from recently installed fills is amenable to both aerobic and
anaerobic biological treatment.
Only one study tested the effectiveness of activated sludge treatment of
leachate obtained from a 6-year old completed landfill (R. F. Weston Inc,
1974). Results of that study showed no decrease in COD after an aera-
tion period of 184 hours due to the refractory nature of the organics
as reflected by the ratios for COD/TOC and BOD/COD of leachate samples
which were as low as 2.1 and 0.03, respectively. Thus, biological
processes are not effective to remove organics in leachate from stabil-
ized landfills that have been generating leachate for a substantial
time period. A similar observation was made in the present study. It
was found that the percentage COD removal in the anaerobic filter
decreased with decreasing BOD/COD ratio's of the leachate from the
laboratory-scale lysimeter.
These results are further substantiated by studies on aerobic treatment
of effluents from anaerobic filters and anaerobic digesters treating
leachate (Table 4). Boyle and Ham (1974) reported that only 40%
removal of the COD material in effluent from the anaerobic digester
was accomplished by polishing it with an aerobic process. The BOD/COD
ratio of the effluent from the anaerobic digester was 0.18 as compared to
0.80 for thfc untreated leachate. Poor COD removal in effluents from
anaerobic filters when treated with aerobic biological processes was
also experienced by Foree and Reid (1973) and DeWalle and Chian (1976)
(Table 4). Comparison of the COD/TOC and BOD/COD ratios of effluent
from anaerobic units treating leachate with those determined in leachate
from landfills having different ages (Figure 3a and 3b) shows that the
effluent of anaerobic processes is comparable to leachate from landfills
of intermediate age. Since a substantial part of the biodegradable
organic matter is already removed in the landfill, biological treatment
23
-------
methods are only moderately effective in removing the remaining organic
matter in leachate.
No systematic studies have been conducted using anaerobic lagoons.
Nordstedt (1970) observed a 70% BOD removal in the holding pond having
an influent BOD of 778 mg/1. Davies (1973), however, only observed
an 11% BOD reduction at a detention time of 12.5 days. These results
therefore tend to indicate that biological processes are only effective
when the leachate is in constant contact with the bacterial solids.
None of the studies indicated the need to remove heavy metals prior to
biological treatment. Formation of metal hydroxides in aerobic systems
and formation of metal sulfides under anaerobic conditions will
effectively eliminate the majority of the heavy metals present in
leachate.
Several studies have tested the effectiveness of physical-chemical treat-
ment of both raw leachate and biologically treated leachate samples.
Chemical precipitation of leachate from a recently leaching landfill was
reported by several investigators (Table 5) and a COD reduction of 0
to 40% in leachate having BOD/COD ratio's of 0.04 to 0.75. In spite of
relatively low COD removals, chemical precipitation was found by most
investigators to be very effective in removing color and iron from
leachate. However, relatively large dosages of lime were required
(Table 5).
The above results can be explained by a related study reported by Chian
and DeWalle (1976) which showed that lime treatment predominantly
removed the fraction of organic matter having a molecular weight larger
than 50,000. This fraction is initially present in relatively low
concentrations in young leachate and almost absent in old leachate.
Since the 50,000 molecular weight fraction was found to increase and
then decrease during the last phase of substrate utilization, lime
treatment will be most effective with leachate from medium aged
24
-------
TABLE 5.—Results of Treatment Efficiencies Obtained in Different Physical-Chemical
Treatment Studies
Treatment
process
(1)
Chemical
Precipita-
tion
Activated
Carbon
and Ion-
Exchange
Adsorp-
tion
Author
(2)
Cook and Force
<»74)
Ho. et al (|»74)
Karr (1912)
Roy Western Inc.
Rogers (l«73>
Simensen and
nitial
COD
131
14.900
9.100
M.IOO
10.800
558
366
4.800
139
3.400
1.240
Odegaa,d(l»7l>
Thornton and
Blanc (l»73)
Van Fleet, et al.
(1174)
This study
Cook and Foree
(W14I
Ho. et al (1974)
K«rrKH71) '
Pohlandand
K»ng(|»i5)
Roy Weston Inc.
Van Fleel. et al.
(1*74)
1.234
1.234
5.0J3
12.923
2.0UO
2.820
330
3.290
4.920
7.213
5.500
184
120
127
2.000
BOD/
COD
(4)
0.45
0.75
0.75
0.74
0.27
U.ll
0.66
0.04
0.81
0.66
0.68
0.68
0.60
0.57
0.36
0.65
0.07
0.45
0.75
0.75
0.66
0.18
0.18
0.04
0.36
COD/
TOC
(5)
.'.45
—
—
—
—
—
2.73
2.1
—
2.78
2.88
2.88
—
—
—
2.89
2.57
3.45
—
2.73
1.5
1.5
2.1
—
Treatment
system
IS)
Line
Ferric chloride
Aium
Lime
Lime treatment of
anaerobic Ji^estor
effluent
Lime treatment of
anaerobic digestor
effluent polished
by aerated lagoon
Alum and lime
FerrosuJfate
Lime
Lime
Lime and Aeration
Iron and aeration
Alum and aeration
Lime
Lime
Alum
Lime
Activated carton
batch treatment of
aerated lagoon
effluent
Activated carbon
column treatment
of litrte pretreated
teas hate
Activated carbon.
batch
Activated Carbon.
column
Activated carbon
batch
Carbon batch treat-
ment of activated
sludge effluent
Ion exchange treat-
ment of activated
sludge effluent
Activaled carbon.
batch
Activated carbon
column treatment
of leachate
Per-
centage
COD
removal
17)
13
16.3
5.3
?.S
t ^
:*
40
13
0
0
Dosages
(8)
2.760 mg/KalCHl,
I.IXIO mg/l
l.'.XX)'ng/l
1,840 mg/l
2.700 mg, i
, .400 mg/l
2 .250 mg/l
AI,(SO4),and
800 mg/l CaO
2.500 mj/l
FeS04 7H,0
I.OnOmg/l
I.OOOmg/!
8 2 10 ml saturated
0
11
24
26
31
26
'0
81
34
59
60
91
58
85
71
lime/! leachate
noOmg/l Fed,
i«0mg/l Alj(SO4),
l,350mg/I
1.200 mg/l
1.700 mg/l
450 mg/l
—
15 min HRT. after
initial volume
turnovers
16.000 mg/l
45 min HRT after 3
volume turnovers
160.000 mg/l
10.000 mg/l
5.000 mg/l cation
and Anionic resin
mixture
10.000 mg/l
25
-------
TABLE 5.—Continued
(1)
Chemical
Oxidation
Reverse
Osmosis
(2)
This study
Cook and Force
(1*74)
Ho, et al(i«u>
Kan-OMZI
Roy Western Inc.
This study
Roy Western Inc.
This study
(31
632
546
527
932
522
330
1.500
7,162
4.800
139
139
1,230
627
265
53,330
33,300
900
536
(4)
0.65
<0.l
<0.l
—
<0.l
0.07
0.75
0.75
0.66
0.04
0.04
—
—
—
0.65
0.65
—
—
(SI
2.89
2.55
2.44
2.9
2.7
2.57
—
—
2.73
2.1
2.1
2.9
2.5
2.1
2.89
2.89
2.9
2.5
(6)
Activated curbon
column treatment
of alum pretreatcti
leachate
Activated carbon
column treatment
of leachate
Activated carbon
column treatment
of effluent of
aerated lugoon
Ion exchange col-
umn treatment of
effluent of aerate*!
lagoon
Activated carbon
column treatment
of effluent of
anaerobic filter
Activated carbon
column treatment
of aerated efflu-
ent of
anaerobic filter
Chlorination
Chloritution with
calcium hypo-
chlorite
Ozonatton
Chlorination
Chlorination with
calcium hypo-
chlorite
Ozonation
Ozonation of an-
aerobic filter ef-
fluent
Ozonation of aerat-
ed lagoon effluent
Reverse osmosis
Reverse osmosis of
leachate at pH
5.5, cellulose ace-
tate membrane
Reverse osmosis
of leachate at
pH 8.0. cellulose
acetate memb.
Reverse osmosis of
anaerobic filter ef-
fluent DuPont 8-9
Reverse osmosis of
aerated lagoon ef-
fluent, cellulose
acetate membrane
(7)
94
70 decreased to
13 after 140 B\
70
50
50
70
33
i
37
22
0
22
37
48
80
56
89
98
95
(8)
65 ml bleach /I
sample
8.000 mj/l Ca(CIO);
after 1 hr
4 hr, 7,700 mg
Oj/1 • hr.
2.000 mg/l Cl,
1.000 mg/ 1 Ca(CIO),
4hr34mgO,/l • hr
3hr. 600mgO,/i • hr
3hr, 400 mg
O,/l ;lir.
80% Permeate yield
50% Permeate yield
50% permeate yield
77% permeate yield
30% permeate yield
26
-------
fills. Low COD removals are therefore observed at both high and low
BOD/COD ratio's while it is largest for intermediate BOD/COD ratio's.
This mechanism is further supported by results of Ho et al_. (1974)
who showed that the COD removal improved from 3 percent to 29 percent
when the leachate was pretreated by aerobic processes.
Activated carbon treatment of raw leachate generally gives better removal
of organic matter than is observed with chemical precipitation (Table 5).
Using relatively large dosages of activated carbon, between 34 percent to
as high as 85 percent COD removals were obtained (Table 5). Column
studies with activated carbon resulted in COD removals of 59 to 94
percent. However, the column effluent was only evaluated for the first
few bed volumes. A carbon column study of larger duration conducted in
this laboratory with a leachate having a COD/TOC ratio of 2.89, showed
an initial TOC removal of 70 percent which decreased to 13 percent
after 140 bed volumes indicating that treatment of leachate by activated
carbon is not feasible due to the large quantities of activated carbon
required.
The above studies were conducted using leachate obtained mostly from
young fills. Since free volatile fatty acids are the main components
in such leachate, the overall removal of organics by activated carbon
is not expected to be as high as compared with the more efficient
biological methods discussed previously. Kipling (1965) reported
that activated carbon removal of acetic acid was never greater than
20 percent. A recent study reported by Giusti et a_l_. (1974) showed
that the percentage reduction by activated carbon of acetic, propionic
and butyric acids was 24, 33 and 60 percent, respectively*at a dosage
of 5,000 mg/1. Although carbon adsorption increased as the homologous
series of acid in aqueous solution is ascended (Traubes1 rule), valeric
and caproic acids are not present in large quantities in leachate. The
variations in reported data on COD or TOC removal by activated carbon
in young leachate are therefore attributable to both the varying
27
-------
proportion of low and high molecular weight free volatile fatty acids
and the varying magnitude of the volatile fatty acid fraction as
observed by Burchinal (1970) Songanuga (1969) and in the present
study.
Since the effluent of biological units treating leachate is comparable
in organic matter composition to leachate from an intermediate aged
landfill, the results obtained in studies treating such effluent with
activated carbon can be used to predict the effectiveness of activated
carbon treatment of leachate from intermediate aged fills. Cook and
Foree (1974) and Pohland and Kang (1975) found that activated carbon
resulted in higher COD removals from biologically pretreated leachate
than that from the untreated leachate (Table 5). The present study
showed that organic matter removal by activated carbon increased with
increasing biological stabilization or decreasing COD/TOC ratio's.
Increased adsorptive capacities were found to correspond with leachate
having relatively low free volatile fatty acid fractions.
Only one study reported COD removal by activated carbon with old stabil-
ized leachate having a BOD/COD ratio of 0.04 (R. F. Weston Inc., 1974).
It was found that activated carbon was able to reduce 85 percent of the
COD. The reported surface concentration was 0.17 mg COD/mg activated
carbon which is high as compared to values obtained for free volatile
fatty acids. DeWalle and Chian (1974b) have shown that the major organic
fraction removed by activated carbon was a fulvic acid-like material
having a molecular weight (MW) between 500 and 10,000. Since, this MW
fraction increases as percentage of the organic matter as the age of
the landfill increases, activated carbon is expected to be more
effective in treating biologically stabilized leachate.
Extensive studies have been reported on the use of oxidants as a means of
removing organic matter in leachate. COD removals varying between 0 and
48 percent (Table 5) were obtained. Results by Ho et a!., (1974)
28
-------
indicated that calcium hypochlorite produced the best results among all
the oxidants evaluated, such as chlorine, potassium permanganate and
ozone. Ozonation of leachate from an old fill was reported by R. F.
Weston Inc, (1974) which showed a COD removal of 22 percent after a
four-hour test. All tests, however, showed that the use of oxidants
was less effective than activated carbon in removing the organic matter
from such stabilized leachate. Ozonation of leachate from a recently
generating landfill is also not promising because of strong resistance
of fatty acids, especially acetic acid, to ozone.
Of all the physical-chemical methods evaluated in the present study,
reverse osmosis membrane treatment was found to be most effective in
removal of COD. Between 56 and 70 percent removal of TOC was obtained
with the conventional cellulose acetate membrane while the removal
increased to 85 and 88 percent with the polyethylenimine, NS-100,
membrane at 85 percent product water recovery and 40.8 atm. Since low
rejection of undissociated fatty acids by membranes was responsible for
the leakage of TOC into the permeate (Chian and Fang, 1973), the
performance of the membrane process was improved to 85-88 and 93-94
percent respectively when the pH of the leachate was increased from
5.5 to 8.0. Good TOC removal with cellulose acetate membranes was
also reported by R. F. Weston Inc, (1974) with an old leachate having
an average BOD/COD of 0.05 and COD/TOC less than 2. In addition to
efficient TOC removal, removal of total dissolved solids was as high
as 99 percent. However, severe membrane fouling was experienced with
leachate and biological pretreatment of leachate prior to membrane
processes is necessary.
The studies summarized above clearly show that the use of physical-
chemical treatment processes treating young leachate does not produce
the degree of organic removal that can be accomplished with biological
processes. However, good results are observed with stabilized leachate
collected from old fills. Similar good results are obtainable with
29
-------
leachate which has been stabilized biologically with both anaerobic and
aerobic processes. Table 6 summarizes the proposed treatment processes
for leachate as characterized by four parameters that have to be known,
i.e., COD/TOC and BOD/COD ratios, absolute COD concentration and age of
the fill. Using these values, a first approximation can be made to select
proper treatment processes for the removal of organic matter present in
leachate. However, if there are less than four such parameters agreeable
to each other, more uncertainty is expected in the selection of the proper
treatment processes.
Since only a limited number of leachate samples were analyzed in the pre-
sent study, it was difficult to give the equivalent age of the so-called
young, medium, and old fills to indicate at what age of the landfill the
physical-chemical methods become less effective than the biological
methods. A larger number of leachate samples from landfills having dif-
ferent ages should therefore be analyzed for these parameters in relation
to their treatability. However, the results in Figures 3 and 4 tend to
indicate that the leachate, which is amendable to biological treatment,
is generated from the first five years during the life of the landfill,
while the intermediate phase corresponds to an age of five to ten years.
Leachate generated after approximately ten years is best treated by phy-
sical-chemical methods. These ranges of landfill ages are very tentative,
and will depend on such factors as the duration of landfill construction,
time for the fill to reach field capacity, density of the refuse, etc.
The present conclusions are based on leachate samples collected from
widely different fill conditions, and therefore should be representative
of most situations.
30
-------
TABLE 6.—Proposed Ralationship between COD/TOC. BOD/COD, Absolute COD,
and Age of Fill To Expected Efficiencies of Organic Removal from Leachate
Character of Leachate
COO/
TQC
(1)
>2.S
2.0-2.8
<2.0
BOO/
COD
(21
>o.s
0.1-0.3
<0.1
Ag«
of
fill
13)
Young
<<5 yr)
Medium
(5yr-IOyr)
Old
OlOyr)
COO. in
milligrams
per liMr
(4)
> 10.000
500-10,000
<500
Effectiveness of Treatment Processes
Biolog-
ical
treat-
ment
(51
Good
Fair
Poor
Chem-
ical
precipi-
tation
(mast
lima
dose)
(6)
Poor
Fair
Poor
Chem-
ical
oxida-
tion.
Ca
(CIO),
(7)
Poor
Ftir
T»it
o,
(8)
Poor
Fair
Fair
Re-
verse
os-
mosis
(9)
Fair
Good
Good
Acti-
vated
car-
bon
(10)
Poor
Fair
Good
Ion
ex-
change
resins
I")
Poor
Fair
Fair
31
-------
REFERENCES
Anaerobic Biological Treatment Processes, Pohland, F. G., Editor, Advances
1'n Chemistry Sev
pp. 1-10 (1971),
in Chemistry Series 105, American Chemical Society, Washington, D.C.,
TWTT:
Andersen, J. R. and Dornbush, J. N. "Influence of Sanitary Landfill on
Ground Water Quality," Journal of American Hater Works Association,
Vol. 59, No. 4, pp. 457-470 (1967).
Apgar, M. A. and Langmuir, D. "Groundwater Pollution Potential of a Landfill
Above the Water Table," Ground Water. Vol. 9, No. 6, pp. 76-96 (1971).
Bender, M. E. et aj_., "On the Significance of Metal Complexing Agents in
Secondary Sewage Effluents," Environmental Science & Technology, Vol.
4, No. 6, pp. 520-522 (1970).
Boyle, W. C. and Ham, R. K. "Treatability of Leachate from Sanitary Landfills,"
Journal Water Pollution Control Federation. Vol. 46, No. 5, pp. 860-875
(1974).
Burchinal, J. C. "Microbiology and Acid Production in Sanitary Landfills,"
A summary report, U.S. Department of Health, Solid Waste Program Grant
No. UI 00529, (1970).
Chian, S. K. and Mateles, R. I. "Growth of Mixed Cultures on Mixed Substrates -
1. Continuous Culture," Applied Microbiology. Vol. 16, No. 9, pp.
1337-1344 (1968).
Chian, E. S. K., Cupta, A., and Fileccia, R. J., "Evaluation of Methods for
the Recycle of Wastewater from Washracks Located at Army Installations,"
Final Report, Contract No. DACA 88-73-C-1004, Letter Report E-40, CERL,
Department of Army, Champaign, Illinois (1974).
Chian, E. S. K., and Fang, H. H. P. "Evaluation of New Reverse Osmosis
Membranes for the Separation of Toxic Compounds from Water," AIChE
Symposium Series, Water-1973, pp. 497-507.
Chian, E. S. K., and DeWalle, F. B., "Sequential Substrate Removal in
Activated Sludge Systems Fed with Naturally Ocurring Wastewater,"
Progress in Water Technology, Vol. 7, No. 2, pp. 235-241 (1975).
Chian, E. S. K., and DeWalle, F. B., "Compilation of Methodology Used for
Measuring Pollution Parameters of Sanitary Landfill Leachate," Final
Report, Contract 68-03-2052, Solid and Hazardous Waste Research
Laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio
(1975).
32
-------
Chian, E. S. K., and DeWalle, F. B., "Removal of Organic Matter by Chemical
Precipitation," Submitted to Journal of Water Pollution Control Federa-
tion for publication (1976).
Chian, E. S. K., and DeWalle, F. B., "The Composition of Organic Matter in
Sanitary Landfill Leachate," Paper accepted for publication in
Environmental Science and Technology (1975).
Cook, E. N. and Foree, E. G., "Aerobic Biostabilization of Sanitary Landfill
Leachate," Journal Water Pollution Control Federation, Vol. 46, No. 2,
pp. 380-392 (1974).
Davies R., "The Hydrogeologist in Action," Mater and Pollution Control. Ill,
12, 21-25 (1973).
DeWalle, F. B. and Chian, E. S. K., "The Kinetics of Formation of Humic
Substances in Activated Sludge Systems and Their Effect on Flocculation,"
Biotechnology and Bioengineering. Vol. 14, No. 7, pp. 739-755 (1974).
DeWalle, F. B. and Chian, E. S. K., "Removal of Organic Matter by Activated
Carbon Columns," Journal of Environmental Engineering Division, ASCE,
Vol. 100, No. EE5, pp. 1089-1104 (1974).
DeWalle, F. B., and Chian, E. S. K., "Treatment of Leachate with Anaerobic
Filter," Paper presented at the First Chem. Congress North America,
Mexico City, Mexico (1976).
Foree, E. G. and Reid, V. M., "Anaerobic Biological Stabilization of
Sanitary Landfill Leachate," Technical Report TR 65-73-CE17, Department
of Civil Engineering, University of Kentucky, Lexington, Kentucky (1973).
Fungaroli, A. A., "Pollution of Subsurface Water by Sanitary Landfills,"
Report SW-12RG Environmental Protection Agency, Washington, D.C. (1971).
Fungaroli, A. A., and Steiner, R. L., "Laboratory Study of the Behavior of
a Sanitary Landfill," Journal of Water Pollution Control Federation,
Vol. 43, No. 2, pp. 252 (1971).
Giusti, D. M., Conway, R. A. and Lawson, C. T., "Activated Carbon Adsorption
of Petrochemicals," Journal Water Pollution Control Federation, Vol.
46, No. 5, pp. 947-965 (1974).
Ho, S., Boyle, W. C., and Ham, R. K., "Chemical Treatment of Leachates
from Sanitary Landfills," Journal Water Pollution Control Federation,
Vol. 46, No. 7, pp. 1776-1791 (1974).
Hughes, G. M. e_t a_l_. "Hydrogeology of Solid Waste Disposal Sites in
Northeastern Illinois," Report SW-12d, U.S. Environmental Protection
Agency, Washington, D.C. (1971).
33
-------
Karr, P. R. Ill, "Treatment of Leachate from Sanitary Landfills," Special
Research Problem, Georgia Institute of Technology, School of Civil
Engineering , (1972).
Kipling, J. J. Adsorption from Solution of Non Electrolytes, Academic
Press, New York, N.Y. (1965).
Knoch, J. and Stegman, R. "Experiments of the Treatment of Landfill
Leachate," Refuse and Waste (Germ.) Vol. 6, pp. 160-166 (1971).
Mao, M. C. M., and Pohland, F. G. "Continuing Investigations on Landfill
Stabilization with Leachate Recirculation, Neutralization and Sludge
Seeding," Special Research Report, Georgia Institute of Technology,
School of Civil Engineering, Atlanta, (1973).
Meichtry, T. M. "Leachate Control Systems," Paper presented at the Los
Angelos Regional Forum on Solid Waste Management, (1971).
Merz, R. C. "Investigation of Leaching of Ash Dumps," Publication No. 2,
State Water Pollution Control Board Sacramento, California (1952).
Merz, R. C. "Final Report on the Investigation of Leaching of a Sanitary
Landfill," Publication No. 10, State Water Pollution Control Board,
Sacramento, California (1954).
Miller, D. W., DeLuca, F. A. and Tessier, T. L. "Ground Water Contamination
in the Northeast States," Report on Contract 68-01-0777, Office of
Research and Development, USEPA, Washington , D.C. (1974).
"Municipal Solid Waste Generated Gas and Leachate," Summary Report by
Disposal Branch, Solid and Hazardous Waste Research Laboratory,
National Environmental Research Center, USEPA, Cincinnati, Ohio
(1974).
Nordstedt, R. A., "Land Disposal of Effluent from a Sanitary Landfill,"
Journal Water Pollution Control Federation. Vol. 47, 7, pp. 1961-1970
(1975).
Pohland, F. G., "Landfill Stabilization with Leachate Recycle," Interim
Progress Report, Grant EP 00658-01, Solid Waste Research Division,
USEPA (1972).
Pohland, F. G., and Kang, S. J., "Sanitary Landfill Stabilization with
Leachate Recycle and Residual Treatment," AIChE Sumposium Series,
Water-!974. II Municipal Wastewater Treatment, pp. 308-318 (1975).
Pohland, F. G., and Maye, P. E., "Landfill Stabilization with Leachate
Recycle," Paper presented at the 3rd Annual Environ. Engr. and Sci.
Conference, Louisville, Kentucky, (1970).
34
-------
Qasim, S. R. and Burchinal, J. C., "Leaching of Pollutants from Refuse
Beds," Journal Sanitary Engineering Division ASCE, Vol. 96, No. SA1,
pp. 49-56 (1970).
Qasim, S. R. and Burchinal, J. C., "Leaching from Simulated Landfills,"
Journal Water Pollution Control Federation, Vol. 42, No, 3,
pp. 371-379 (1970).
Reinhardt, J. J., and Ham, R. K., "Final Report on a Demonstration Project
at Madison, Wisconsin, to Investigate Milling of Solid Wastes Between
1966 and 1972, Vol. 1," U.S. Environmental Protection Agency, Office
of Solid Waste Management Program, Washington, D.C. (1973).
Rickert, D. A. and Hunter, J. V., "General Nature of Soluble and Particulate
Organics in Sewage and Secondary Effluent," Water Research, Vol. 5,
pp. 421-436 (1971).
Rogers, W. P., "Treatment of Leachate from a Sanitary Landfill by Lime
Precipitation Followed by an Anaerobic Filter," M.Sc. Thesis Dept.
of Civil Engineering, Clarkson College of Technology, 100 pp. (1973).
Simensen, T. and Odegaard, H., "Pilot Studies for the Chemical Coagulation
of Leachate," Norwegian Institute of Water Pollution Research, Blindern,
Oslo, (1971).
Songanuga, 0. 0., "Acid, Gas and Microbial Dynamics in Sanitary Landfills,"
Ph.D. Dissertation, West Virginia University, Morgantown, West Virginia,
(1969).
Solid and Hazardous Waste Research Laboratory "Interim Report 1, Test Cell
1 Boon County Field Site," National Environmental Research Center,
Cincinnati, Ohio (1973).
Standard Methods for the Examination of Water and Wastewater, 13th ed.,
APHA, AWWA, WPCF (1971).
Thompson, F. C., "Microbiology and Fatty Acid Production in Sanitary
Landfills," M.S. Thesis, West Virginia University, Morgantown, West
Virginia (1969).
Thornton, R. J. and Blanc, F. C., "Leachate Treatment by Coagulation and
Precipitation," Journal Environmental Engineering Division ASCE,
Vol. 99, No. E4, pp. 535-539 (1973).*
Van Fleet, S. R. et al_., "Discussion, Aerobic Biostabilization of Sanitary
Landfill LeacFate," Journal Water Pollution Control Federation, Vol.
46, pp. 2611-2612 (1974]L
R. F. Weston Inc., West Chester, Pennsylvania (1974).
Walker, W. H., "Illinois Groundwater Pollution," Journal American Water Works
Association, Vol. 61, pp. 31-40 (1969).
35
-------
Ill
THE USE OF MEMBRANE ULTRAFILTRATION AND SPECIFIC ORGANIC ANALYSIS
FOR CHARACTERIZATION OF ORGANIC MATTER IN LEACHATE
CONCLUSIONS
Membrane ultra-filtration, gel permeation chromatography, and specific organic
analysis were used to separate and determine the main classes of organics
present in variable strength leachate having COD values ranging from 81 to
71,680 mg/1. Ultrafiltration (UF) of a leachate sample obtained from an
experimental lysimeter showed that 72.8% of the organics permeated the 500
MW UF membrane, and further analysis showed that most of these were present
as low molecular weight free volatile fatty acids. Comparison of the gel-
permeation TOC distribution with the TOC in the retentate of the UF membranes
Indicated that the nominal molecular weight estimated from the membrane ultra-
filtration is approximately four times higher than the absolute molecular
weight obtained with the former method. Evaluation of the serial UF proce-
dure, using either membranes of stepwise increasing pore sizes or stepwise
decreasing pore sizes, showed that the presence of high molecular weight
organics and ferric hydroxides will decrease the molecular weight cutoff
of a membrane. The nominal molecular weight cutoff of a membrane can
change by a factor of ten as a result of the sequence during the ultrafil-
tration. It was therefore concluded that the parallel UF procedure resulted
in a more representative separation than the serial procedure.
Further UF separation and gel permeation followed by specific organic
analysis and IR identification of the membrane-separated organics showed
that the next largest organic fraction was a fulvie-like material with a
relatively high carboxyl and aromatic hydroxyl group density. A small
percentage of the organics (i.e., 6%) consisted of a high molecular weight
humic-carbohydrate-like complex also characterized by a significant amount
of hydrolyzable amlno adds.
36
-------
A carbon balance of the organics present in each fraction showed that 86%
of the carbon in the 500 MW UF permeate and 50% of the carbon in the high
molecular weight fraction was identified. Only 20% of the fulvic-acid-like
fraction was identified, indicating that it consisted of complex molecular
structures. Solvent extraction of the different membrane fractions showed
the presence of esterified fatty acids associated with the high molecular
weight humic fraction.
Membrane fractionation of leachate samples collected from different land-
fills showed a decrease of free volatile fatty acids in the 500 MW UF
permeate with increasing age of the landfill. Further membrane fractiona-
tion showed that the high molecular weight humic-carbohydrate-like organics
experience a more rapid decrease than the intermediate molecular weight
fulvie-like organics.
Comparison of the distribution of specific organics in the different
molecular weight fractions with the results obtained by other researchers
with organics obtained from widely different environments shows a large
similarity, indicating that universal bacterial processes are governing
the character of these organics. The present study measured a relatively
small amount of high molecular weight organics as compared to other studies.
This is likely a result of the anaerobic environment during the leaching
process which enhances acid fermentation of complex high molecular weight
organics. The present study characterized stabilized landfill leachate
by relative organic distributions of the following three groups: fatty
acids < carbohydrates, proteins, and humic-like substances < aromatic
hydroxyl, carboxyl and fulvic-like substances. This pattern reflects
increased resistance of the organics to bacterial degradation.
37
-------
INTRODUCTION
Major complications generally arise in the identification of general
classes of organic matter present in wastewater, surface water and
groundwater, due to the complex nature of these compounds. Many investi-
gators have studied the techniques for concentration, separation and
identification of different classes of organics. For example, Hunter
and Heukelekian (1965) used primarily solvent extraction techniques to
obtain fractions such as lipids, fatty acids, detergents, phenols,
carbohydrates and lignins in domestic wastewaters. The magnitude of
each fraction was then determined using gravimetric, colorimetric and
chromatographic methods. When the organic matter concentrations in the
water are low, a preconcentration step is generally employed. Myrick
and Ryckman (1963), for example passed 20,000 a of river water through
an activated carbon column, whereupon, the dried carbon was extracted
with chloroform, ethanol, acetone and benzene.
Extraction or adsorption techniques are generally not effective in con-
centrating polar organics; these can, however, be concentrated by freeze
drying, evaporation or distillation. The separation of such polar
organic compounds is generally more cumbersome. Polar volatile organics
can be detected by gas chromatographic techniques, while polar non-
volatile organics, often having high molecular weights, can be separated
by fractional precipitation, gelpermeation chromatography or membrane
ultrafiltration. Rebhun and Manka (1971), for example, have determined
humic acid gravimetrically after precipitation at pH 2 and fulvic acid
by extraction in butanol. Gjessing and Lee (1967) used evaporation
under reduced pressure followed by gelpermeation chromatography to
obtain different organic fractions separated according to molecular
size. Physical or chemical analysis was conducted only after the
separation step was completed.
38
-------
A disadvantage of the above methods is that only relatively small
amounts of organic matter are obtained. For that reason DeWalle and
Chian (1974) employed bench-scale membrane reverse osmosis and ultra-
filtration units to obtain relatively large amounts of organics of
different molecular sizes. After the membrane fractionation step,
further separation was realized using gel permeation chromatography
followed by functional group analyses.
SEPARATING PROPERTIES OF MEMBRANES
The molecular properties of membranes have been recognized for more than
100 years in dialyses, ultrafiltration and reverse osmosis processes
(Ferry, 1936) but have not found wide application until recently. Black
and Christman (1963), for example, used dialysis membranes with pore
o o o
sizes of 35 A, 48 A and 100 A to obtain different fractions of colored
organics present in surface water. Using ultrafiltration membranes
with different molecular size cutoffs,Gjessing (1973) and Schindler and
Alberts (1974) studied the molecular weight distribution of organics in
surface waters. Fang and Chian (1975) employed reverse osmosis to concen-
trate and separate organics having relatively low molecular weights.
Dialysis is based on diffusion in which relatively small solutes permeate
the membrane, while larger solutes are retained due to the seiving effect
of the membrane (Kesting, 1971). Ultrafiltration separates from a solu-
tion those organics whose molecular dimensions are ten or more times
larger than ^he solvent, by pressurizing the solvent to flow through a
membrane (Michaels, 1968). The solute transport during ultrafiltration
occurs by viscous flow mechanisms rather than diffusive mechanisms such
as is the case in reverse osmosis. Because of these different transport
mechanisms the rejection efficiency of ultrafiltration membranes is
generally independent of the applied pressure, while the rejection for
reverse osmosis membranes increases to an asymptotic maximum with
increasing pressure (Michaels, 1968). Some deviations of this general
behavior have been noted. For example, the rejection in membrane ultra-
filtration may decrease beyond a certain pressure due to elastic distor-
39
-------
tion of the membrane while it may increase with increasing solute con-
centration due to higher solute drag on the pore walls. As a result
of such viscous drag, the flux will also decrease. The occurrence of
concentration polarization (i.e., the formation of a boundary layer of
a substantially higher concentration than the bulk retentate concentra-
tion) will decrease the apparent rejection of the membrane. Such con-
centration polarization will generally not decrease the flux through
the membrane due to the low osmotic pressure of the relatively high
molecular weight organics (Blatt et_jil_., 1970). Formation of a gel
upstream of the membrane with a lower porosity than the membrane will
increase the rejection (Michaels, 1968).
Craig ie_t a]_. (1957) have concluded that molecular size of the solute is the
main factor determining its permeation and that charge effects play a
less significant role in membrane ultrafiltration. Mathematical
expressions accounting for the steric hindrance to entry into a pore,
the viscous drag at the wall of the assumed capillaries in the membrane
and the friction between the solute and the walls of the capillary,
indicate that the probability of permeation rapidly decreases when the
solute diameter becomes more than half of the pore diameter. The
greatest selectivity occurs when the solute diameter is 20 to 30% of
the pore diameter (Renkin, 1954).
DIALYSIS MEMBRANES
The earliest membranes consisted of animal derived material, but were
later replaced by synthetic materials such as cellulose (cellophane,
cuprophane) or nitrocellulose (collodion, gradocol) membranes. The
pore size of the first manufactured cellophane membranes (Visking
O
Division, Union Carbide, Chicago, IL) was generally 40-60 A
O
(McBain and Kistler, 1931) but was later reduced to 20 A (Morton,
o
1935). Thereafter, it was increased to 30 A (Seymour, 1940; Madras,
o o
1949) and presently varies between 38 A (Renkin, 1954) and 46 A
(Durbin, 1960). The pore size of uncoated cellophane (Dupont, Stillwater,
DE) generally varies between 62 A (Renkin, 1954) and 82 A (Durbin, 1960).
40
-------
The cellulose membranes from Visking Company (Chicago, IL) have a
o
stated pore size of 48 A and are reported to retain organics with a
molecular weight above 12,000. The molecular weight cutoffs, however,
have been found to vary by a factor of ten among the different inflated
diameters of the seamless tubing (Craig e_t ail_., 1957). Cellulose
membranes with molecular-weight cutoffs of 12,000, 6,000 and 3,500
are also commercially available (Spectrum Medical Industries, Los
Angeles, CA). Several investigators have used cellulose membranes to
concentrate and separate organics in wastewater and surface waters
o o
(Table 7), using pore sizes of 48 A (Black and Christman, 1963), 40 A
(Gjessing, 1971), 25 A (Werschaw and Burcar,1967) and 24 A (Packham, 1974;
Dor, 1975). Several studies in Europe using 1KB ultrafilters (1KB Comp.,
Upsola, Sweden) and were able to obtain organics with a molecular weight
larger than 30,000 (Povoledo and Gerletti, 1968).
A wider range of pore sizes can be obtained using collodion dialyses
membranes. Elford (1931) made ether alcohol collodion membranes with
o o
reproducible and well-defined pore sizes ranging from 100 A to 30,000 A,
with the former pore size able to retain bovine serum albumin (MW
65,000). Elford and Ferry (1936) made a collodion membrane with a pore
o
size of 80 A able to retain ovalbumin (MW 45,000). A membrane with a
o
pore size of 20 A was able to retain sucrose (MW 342) as reported by
Ferry (1936). The pore size measured by these researchers using
Poiseuille's law, however, are about twice as large as the actual size
(Renkin, 1954) since they did not take into account the viscous drag
and the presence of blind pores. Using special curing steps collodion
membranes can be made that are able to retain glucose (MW 180) but
allow passage of urea (MW 60) (Carr, 1959). Most collodian membranes
o
have pore sizes of 100 to 150 A (Meschia and Setnikar, 1958), 60 to
o o
80 A (Jacobs, 1972) and approximately 30 A (Robbins and Mauro, 1960).
Membranes with molecular weight cutoffs of 20,000, 60,000, 70,000,
80,000 and 160,000 are commercially available (Schleicher and Schuell,
Keene, NH). Black and Christman (1963) used a collodion membrane with
o
a pore size of 35 A to concentrate and separate organics in surface
water.
41
-------
Table 7. Dialysis Studies Characterizing the Molecular
Weight Distribution of Aquatic Organics
ro
Percentage
Retained
Organic Matter
Author
Earth and
Acheson (1962)
Black and
Christman (1963a)
Bunch et al .
(1961)
Gjessing
(1971)
Martin et al.
(1972)
Packham
(1964)
Shapiro
(1964)
Source
tap water
surface water
secondary effluent
surface water
fungal humic acid
fungal fulvic acid
organics from surface fhumic acid
water concentrated by Jhumic acid
ion exchange [fulvic .acid
butanol extractable
organics
• \\.
Membrane
cellulose dialysis
membrane retains
MW > 10,000 0
colloidon membrane D 35A0
cellophane membrane D048A
membrane filter D 100A
cellulose dialysis
membrane
cellgphane membranes
D 40A
dialysis membrane
o
dialysis membrane D 24A
membrane filter D 100A0
dialysis membrane D 24A
cellulose dialysis
membrane
. wv* i 1 1\«\«
(*)
91
87.5
13
27.4
66
97%
29%
82
0
18
•^30-35
^
Parameter
color
color
color
COD
TOC
weight
weight
weight
weight
weight
weight
UMWCII ui a i 1 1
(mg/1)
0.08
142
142
142
48.9
3600
1400
DM
Parameter
Poly-
saccharide
COD
COD
COD
COD
weight
weight
-------
ULTRAFILTRATION MEMBRANES
New types of highly permeable ultra filtration (UF) membranes consisting
of anisotropic polyelectrolyte complexes (Amicon, Lexington, Mass) and
cellulose acetates (Abcor, Cambridge, Mass) have become available since
1960. Isotropic microporous cellulose ester membranes, such as
Pellicon molecular filters (Millipore, Bedford, Mass.) and Nuclepore
membranes (Nucleopore Pleasanton, CA), manufactured by etching of radia-
tion damaged sites, have become available only recently (Good, 1974).
These membranes have a sharper molecular cutoff than most ultrafilters
owing to their isoporous structure. Cellulose acetate ultrafilter are
also commercially available (Biorad, Richmond, CA).
Several studies have used membrane ultrafiltration to characterize
molecular sizes of aquatic organics and their experimental procedures
are listed in Table 8. When more than one ultrafiltration membrane is
used, a parallel procedure can be selected in which the sample is
divided in different aliquots that are each passed through the indi-
vidual membrane (Gjessing, 1975; Gjessing, 1970). In the sequential
procedure, the sample is passed either through the membrane with the
largest pore size whereupon the permeate is passed through the
membrane with the next smaller pore size (Calder and Kearsley, 1974)
or it is first passed through the membrane with the smallest pore size
whereupon the retentate is passed through the membrane with the next
largest pore size (Doig and Martin, 1974; Gjessing, 1970). None of
the studies, however, evaluated the difference among the three pro-
cedures. After comparing a dialysis membrane with the ultrafiltration
membranes UM-10 and XM-50, Gjessing (1971) noted that 66% of the organic
carbon was retained by a cellophane membrane with a stated pore size
O
of 40 A and a reported molecular-weight of 10,000; the ultrafiltration
membranes with molecular-weight cutoffs of 10,000 and 50,000 retained
94% and 50%,respectively, on a comparable sample. Interpolation of the
rejection values would indicate that 66% of the organic carbon would
be retained with a UF membrane with a molecular-weight cutoff of 30,000
which is about three times larger than the reported molecular-weight
43
-------
Table 8. Membrane Ultrafiltration Studies Characterizing the
Molecular Weight Distribution of Aquatic Organic
Ultrafiltration
Organic
carbon
content Filtration
Washing of Concentration
flutnor
Barber (1968)
Calder and
Kearsley (1974)
Cameron et al.
(1972)
Doig and
4* Martin (1974)
-P»
Ghassemi
(1367)
Gjessing
(1971)
Gjessing
(1973)
Gjessing
(1970)
Membranes
UM3
UM2, PM10, XM50,
XM100A
XM50
UM05, UM2, UM10
UM1
UM10, XM50
UM2, UM10, XM100
UM2, UM10, UM20E
Source
Atlantic Ocean
marsh water
salty marsh water
soil humic acid
3 water extract
of 200g soil
vacuum concentrated
surface water
4x concentrated
surface water
surface water
30x concentrated
surface water
Pretreatment
no
glass fiber
filter
class fiber
filter
pyrophosphate
extract
0.22
paper filter
glass fiber
filter
glass fiber
filter
qlass fiber
filter
(mg/1)
0.6-1.9
6.2-9.5
3.8-5.4
61
15.2
12.5
19-28
..v.ww.v.. KM •*• 1 1 I 1 ^ VI WW IV»^.II^I U Ir I UII
Steps retentate Factor
no two 7x
sequential yes
high MW->-low MW
sequential yes
high MW+low MW
sequential yes
low MW-»high MW
no lOx
parallel
parallel yes 40x
sequential yes till
Tow MW^-high MW permeate
<5 mg/1 TOC
Schindler and UM2, UM10, XM50,
Alberts (1974) XM100
Wilander
(1972)
Zajic and
Knettig (1971)
UM2, PM10
UM10, XM50
lake water
surface water
hydrocarbon
solution
glass fiber 1.2-7.8 parallel
12-180 parallel
0.45p, dialysis
with 48A cellu-
lose membrane
sequential
low MW-»high MW
no
yes
lOx
-------
cutoff of the dialysis membrane. Similarly, Ferry (1936) noted that
while iso-electric serum albumin would not penetrate a collodion ultra-
filtration membrane, it was still able to dialyze through a collodion
membrane with a pore diameter one third of the UF membrane. El ford
(1933) also observed that substances that would pass collodion membranes
during dialysis will be retained during ultrafiltration using the same
membrane as a result of partial blocking of the pores.
COMPARISON OF MEMBRANE SEPARATION METHODS WITH OTHER SEPARATION TECHNIQUES
Only a few studies confirmed the obtained molecular-weight distribution
of aquatic organics by independent methods. Sieburth and Jensen (1969)
noted that organics included in Sephadex 6-15, thereby having a molecular
smaller than 1500, all permeated through a dialysis membrane, while DeHaan
(1974) noted that organics included in G-25, thereby having a molecular
weight smaller than 5000, passed through a dialysis membrane. Gjessing
(1973) further showed that membrane fractionation using ultrafiltration
membranes gave better separation than gel permeation chromatography.
The organics excluded from Sephadex G-75, thereby having a molecular
weight larger than 50,000, for example, still permeated partially (19%)
through a 10,000 MW UF membrane, while the organics included in the G-10,
thereby having a molecular weight smaller than 500, were partially
retained (29%) by a 100,000 MW UF membrane. The combined use of membrane
ultrafiltration followed by gel permeation of the UF retentate, however,
gave the best separation. A quantitative comparison of the two separa-
tion processes was not made,si nee the UF was monitored by TOC and the
Sephadex by absorbance (Gjessing, 1973).
It was therefore the purpose of the present study to quantitatively
compare membrane ultrafiltration and gelpermeation chromatography. The
present study also evaluated the parallel versus the sequential UF
separation procedures. Although, most of the previous studies have
tested water with a low organic matter content (Table 8), the present
study evaluated the applicability of UF separation using high strength
leachate. The leachate samples having chemical oxygen demand (COD)
45
-------
values ranging from 81 to 71,680 mg/1 were obtained from solid waste
fills of different ages and subject to different climatic conditions
(Chian and DeWalle , 1976). Such leachate is generated by infiltrating
rainwater which leaches pollutants from the solid waste. An in-depth
knowledge of the organic composition is necessary to assess the potential
for groundwater pollution and to predict the effectiveness of treatment
processes to remove pollutants from the leachate. A summary of
previous studies (Table 9) showed similar ranges of organic matter
concentration. It was also noted that the major fraction consisted
of free volatile fatty acids which was also reflected in the high
biochemical oxygen demand (BOD) relative to the chemical oxygen demand.
The present study also included the constituents listed in Table 9
but used membrane ultrafiltration, gel permeation chromatography and
specific organic analyses to account for the unidentified major organic
fractions.
46
-------
Table 9. Organic Matter Concentrations in Landfill Leachate
and their Relative Composition
Author
Organic Matter
concentration
Relative organic matter
composition
Burrows and
Row (1975)
County of
Sonoma (1973)
Hughes et al.
(1971)
Lin (1966)
Mao and
Pohland (1973)
Qasim and
Burchinal (1970)
170,000 mg/1 COD
3260-22,400 mg/1 COD
1000-50,000 mg/1 COD
20,000 mg/1 COD
4320-12,000 mg/1 COD
33,360 mg/1 BOD
79% of COD as fatty acids
4% of COD as alcohols
40% of COD as fatty acids
80% of COD as BOD
25% of COD as fatty acids
75% of anions from ionized
fatty acids
80% of COD as fatty acids
74% of COD as BOD
35% of organic matter as
aromatic hydroxyls
32% of organic matter as
nitrogeneous organics
47
-------
MATERIALS AND METHODS
The ultrafiltration membranes used in this study were UM05, UM2, DM10,
and XM100A (Amicon, Lexington, MA) with nominal molecular-weight cutoffs
of 500, 1000, 10,000 and 100,000, respectively. After ultracentrifuga-
tion of one liter of leachate at 30,000 rpm for 30 minutes, the super-
natant was concentrated 5 fold with a 500 MW UF membrane. The retentate
was then desalted by diluting to twice its volume with distilled water
followed by a reduction to the original retentate volume; this step
was repeated a second time. The percentage retained was calculated from
a material balance. The 500 MW UF membrane was selected since it has
the lowest molecular-weight cutoff that is able to separate the organic
matter from the inorganic salts that are present in high concentration
in the leachate.
Following the ultrafiltration step, separate aliquots of the UF retentate
were further separated into different molecular-weight components using
Sephadex 6-75 (1000-50,000 MW) and G-200 (1000-200,000 MW) columns
(Pharmacia, Piscataway, NJ). After the molecular weight distribution
of the 500 MW UF retentate was established by the use of the Sephadex
columns, other ultrafiltration membranes with higher molecular weight
cutoffs were employed to obtain sufficient quanties of each of the
different molecular weight fractions for chemical analyses.
The ultrafiltration membrane fractions and Sephadex fractions were
characterized by TOC (TOC Analyzer 915A, Beckman, Fullerton, CA),
COD, carbohydrates as dextrose (anthrone test), amino acids as lysine
(ninhydrin test after 24 h. acid digestion with 6N HC1), carboxyl groups
as acetic acid (hydroxyl amine test), carbonyl groups as acetophenone (2,
4-dinitrophenylhydrazine test), and aromatic hydroxyl groups as tannic
acid (Denis test). Dissolved free amino acids were measured with an
automated Technicon TSI amino acid analyzer(Tarrytown, NY).
48
-------
Each membrane fraction was also separated according to its polarity by
two solvent extractions: hexane, which separates the less polar lipids,
hydrocarbons and fatty acids from the water phase, and butanol, after
pH adjustment of the aqueous layer to 2, which separates humic sub-
stances from the water phase (Rebhum and Manka, 1971; Packham, 1964;
Christman and Ghassemi, 1966). The hexane extraction was performed
twice on a 20-ml sample of the concentrated leachate with 2 portions of
20-ml hexane heated to 45°C, which after vigorous shaking formed an
emulsion. This was eliminated by addition of 5 ml ethanol and 0.5 ml
of IN ^SO,, after which the hexane layer was removed from the separatory
funnel. After acidification of the aqueous layer to pH 2, it was
further extracted twice with two 20-ml aliquots of butanol. An addi-
tional 3.5 ml of ethanol was added to break the emulsion. After separa-
tion of the solvent layers from the water phase, the solvents were
removed.
Infrared (IR) spectra were run on the residues of the hexane and butanol
extracts using a Beckman Model 20A (Fullerton, CA). Gas chromatographic
(GC) analyses of the free volatile fatty acids were conducted using a
Hewlett Packard 5750 B Chromatograph (Palo Alto, CA) equipped with a
flame ionization detector and a column consisting of 20% neopentyl glycol
succinate and 2% phosphoric acid on Chromosorb PAW 60-80 mesh (Chian
and Mateles, 1969).
The most extensive organic analysis was conducted on a leachate sample
obtained from a large experimental lysimeter installed at the University
of Illinois. The lysimeter - 1.51 m in diameter and 3.02 m in height -
was filled with 1525 kg of milled solid waste (obtained from the Hartwell
section of Cincinnati, Ohio) which consisted of 44% paper and 12% food
waste. Based on the analysis of this leachate sample the most important
parameters were selected and subsequently measured on leachate samples
collected from selected landfills in different locations of the United
States.
49
-------
RESULTS
It was observed that the color of the leachate sample collected from
the underdrain of the lysimeter changed rapidly from light yellow to
dark green, and that after prolonged storage a dark brown iron hydroxide
precipitate accumulated on the bottom of the sample container. The
sample, with a TOC of 17,060 mg/1 and a COD of 49,300 mg/1, has a
corresponding iron concentration of 2200 mg/1. Since the colloidal
iron hydroxide would interfere with the membrane separation processes,
the leachate sample was collected under anaerobic conditions. To
prevent further oxidation of the iron, the sample was processed as
rapidly as possible through the ultracentrifugation and membrane
ultrafiltration step under nitrogen atmosphere.
SEPARATION BY MEMBRANE ULTRAFILTRATION AND GEL PERMEATION
Only 104 mg/1 or 0.6% of the volatile residue of 18,403 mg/1 was removed
with the ultracentrifuge indicating that most organics are present in
the soluble fraction. Subsequent membrane ultrafiltration showed that
27.2% of the initial TOC of the sample was retained by the 500 MW UF
membrane: thus, because the majority passed the membrane, they may be
identified as low-molecular-weight compounds. When the 500 MW UF
retentate was applied on a Sephadex G-75 column, the TOC distribution
in the eluate showed that 22% of the 500 MW UF retentate or 6% of the
original TOC was excluded from the column as it eluted between 20 and
36 ml (Figure 6). This exclusion indicates a molecular weight larger
than approximately 30,000-50,000 MW. The application of the 500 MW UF
retentate to a Sephadex G-200 column resulted in essentially the same
elution pattern. This fraction, therefore, had a molecular weight
considerably larger than 50,000 (Fisher, 1965). The results in
Figure 6 further showed that the majority of the 500 MW UF retentate
consisted of organics with a molecular weight smaller than 1000 to
3000 (Table 10).
50
-------
320r 48r 800
^280- 42- 700-
80 -|320
100
0 -1
I
160 <»
120
80
40
0
Elution Volume, m/
Figure 6. Eluate of the 500 MW UF Retentate of the UI Leachate Sample on a Sephadex
G-75 Column as Characterized by Total Organic Carbon, Specific Organics and Functional Groups.
-------
Table 10. Chemical Analyses of the Sephadex G-75
Eluate of the 500 MW UF Retentate
High MW Intermediate MW Low MW
Fraction (la) Fraction (2a) Fraction (3a)
Estimated MW
E luation volume
UOC of 500 MW
UF retentate
>(30,000-50,000)
20-36 ml
22%
^3000-30,000
36-66 ml
10%
<( 1000-3000)
66-100 ml
68%
%TOC of original 6%
sample
% of organic carbon 26%
consisting of
carbohydrates
% of organic carbon 4%
consisting of aromatic
hydroxyl compounds
% of organic carbon 5%
consisting of carbonyl
compounds
% of organic carbon 5%
consisting of carboxyl
compounds
2.5%
42%
2%
2%
5%
18.7%
7%
5%
5%
6%
52
-------
Substantial quantities of the different molecular weight fractions were
obtained using UF membranes having larger molecular weight cutoffs than
the previously used 500 MW UF membrane. Based on the gel permeation
chromatographic data, both a 10,000 MW and a 100,000 MW UF membrane
were selected. When another aliquot of the centrifuged leachate
was passed through a 10,000 MW UF membrane, 9.9% of the TOC of the
sample was recovered in the retentate (Fraction 3b, Figure 7). The
results indicate that the 10,000 MW UF membrane not only retained the
Sephadex fractions la and 2a in (Table 9) but also part of the low-
molecular-weight Sephadex fraction 3a. As 9.9% of the initial TOC would
represent organics eluted from the 6-75 up to 69 ml (Figure 6) corres-
ponding to an apparent molecular weight of 2500, it would indicate that
the nominal molecular weight estimated from the membrane ultrafiltration
is approximately four times higher than the absolute molecular weight
obtained from the gel permeation. When the 10,000 MW UF permeate was
subsequently passed through a 500 MW membrane, the sum of the TOC of
the 10,000 and 500 MW retentate (Fractions 3b and 4b, Figure 7) was only
19.6% of the original TOC of the sample, as compared to 27.2% when an
aliquot was passed through the 500 MW UF directly (Fraction 6b, Figure
7). Since the 10,000 MW UF permeate when applied to the 500 MW UF
was devoid of large quantities of high molecular weight compounds and
ferric hydroxide colloids, the absence of these compounds appeared to
decrease the rejection with the 500 MW UF.
To further investigate this apparent decrease in rejection, an aliquot of
the original 500 MW UF retentate was diluted to its original strength and
pH and passed through the 10,000 MW UF membrane. The retentate (Fraction
8b, Figure 7) was found to represent 21.8% of the original TOC as com-
pared to 9.9% when the centrifuged leachate directly permeated the 10,000
MW UF membrane. This indicates that a relatively large concentration of
high-molecular-weight organics together with ferric hydroxides will
decrease the nominal molecular-weight cutoff of the membrane. This was
further exemplified by the application of fraction 8b to a 100,000 MW UF
53
-------
Centrifuged Leachote
in
-p.
1C
10 ooo Fraction 3b
In^T ^J t f\f\ of
5<
MY*
n. Fraction Ib
MIA/ DO * Retentate
MW|TO 67%
1
«*M^ »w
f) Fraction 4b
i,c— Retentate -1
UF 9.7%
Retained
19.6%
5C
MM
lorir
IVA/JL
MVv
innnn Fraction 8b
uS^-^ Retentate
WI™
Y\ Fraction 6b
f,ip-*-Retentate
UF 27.2%
^ 21.8 %
,
faction 2b Fraction 5b Fraction 7b Fraction 9b Froctk
ermeate Permeate Permeate Permeate Perme
> Fraction lOb
'— Retentate
11.1%
Mb
Figure 7. Effect of Parallel and Sequential Membrane Separation on the
Retention of Organic Carbon in Centrifuged Leachate by
Reverse Osmosis and Ultrafnitration Membranes having
Different Molecular-Weight Cutoffs.
-------
membrane after dilution to its original strength. The 100,000 MW UF membrane
was found to retain 11.1% of the original TOC (Fraction lOb, Figure 7)
which is larger than the 9.9% (Fraction 3b, Figure 7) obtained when the
sample directly permeated the 10,000 MW UF membrane. Thus, the sequence
of membrane ultrafiltration steps may change the nominal molecular-
weight cutoff of a membrane by a factor of more than 10. When the 500
MW UF retentate permeated the membranes with the next highest molecular-
weight cutoff, most of the salts were already removed and present in
the 500 MW UF permeate. The reduced salt content may result in an
increase of the diffuse double layer and an increase of the size of
the molecule (Gjessing, 1971), which in turn will increase its
rejection. The above results indicate that parallel ultrafiltration
will result in a more representative separation than serial ultrafil-
tration. Because the 500 and 10,000 MW UF membranes are best suited
to separate the organics in leachate, they were selected and used in
the parallel procedure to separate the organics in other leachate
samples.
SPECIFIC ORGANIC ANALYSIS OF DIFFERENT MOLECULAR-WEIGHT FRACTIONS
Extensive organic analyses were conducted on the organic fractions
obtained from the gel permeation and membrane ultrafiltration steps.
Analyses of the specific organics present in the Sephadex fractions
showed considerable differences (Figure 7; Table 10) with relatively
high concentrations of carbohydrates observed in the high-molecular-
weight fraction and substantial quantities of aromatic hydroxyl and
carboxylic compounds present in the low-molecular-weight fraction.
Although the TOC data apparently indicate that fraction 3a is rather
homogeneous (Figure 6), the colorimetric tests show that this is not
the case since the maxima of the different organic compounds do not
elute at the same volume, and the specific compounds associated with
organics of decreasing molecular weight are carbohydrates (maximum
at 77 ml), carbonyl - (81 ml), carboxyl - (83 ml) and aromatic
hydroxyl compounds (85 ml).
55
-------
Relative constribution to the carbon in each molecular weight fraction
was then calculated (Table 10). It was thereby assumed that the carbon
in the model compounds used in the colorimetric tests reflected the
actual structures and that a certain organic structure did not respond
to more than one colorimetric test. A TOC/weight ratio of 0.4 for
carbohydrates and carboxyl compounds and a ratio of 0.5 for aromatic
hydroxyl and carbonyl compounds were used to calculate the values in
Table 10. Although the model compound in the carbonyl test had a TOC/
weight ratio larger than 0.5, the value of 0.5 was selected since the
carbon generally does not exceed more than half of the weight of humic
substances (Konanova, 1966).
Organic analyses of the membrane UF fractions 3b and lOb (Figure 7)
showed that 23% and 24% of the organic carbon consisted of carbohydrates
which is comparable to the value of 26% obtained in the larger than
50,000 MW Sephadex fraction. Analyses of the other molecular weight
fractions (Figure 8) showed that the relative carbohydrate content
decreased with decreasing molecular weight, thereby, confirming the
results of the Sephadex analyses. These data also show that the
carbohydrates and aromatic hydroxyl groups can be selected as indicators
for the presence of high-and low-molecular-weight organics (Fraction
lOb and 9b, Figure 7) respectively, since the carbohydrates are mainly
present in the high-molecular-weight fraction while the aromatic
hydroxyls are present in highest concentration in the low-molecular-
weight organics.
The protein test was run on the different membrane fractions and its
values followed a trend similar to the carbohydrates in the fractions
larger than 10,000 MW (Figure 8). A surprisingly high content of amino
acids was found in the 500 MW permeate indicating the presence of free
amino acids that permeate the 500 MW UF membrane. Qualitative analysis
of the unhydrolyzed 500 MW UF permeate with a Technicon automated
amino acid analyser showed in decreasing order thepresence of ornithine,
lysine and valine. Dissolved free amino acids are generally detected
56
-------
30
§20
3$
en
10
I
<500MW
Fraction 7b
A Carbohydrates As Dextrose
D Proteins As Leucine
• Carboxyl As Acetic Acid
• Car bony I As Acetophenone
A Aromatic Hydroxyl As Tannic Acid
/^Percentage Of TOC Identified In
V_yEiach Molecular Weight Fraction
5OO-
10,000 MW
FractionSb
Molecular Weight Fraction
10,000- >IOOPOOMW
100,000 MW Fraction KDb
Fraction Mb
Figure 8. Characterization of Different Membrane Ultrafiltration
Fractions for Specific Organics and Functional Groups
-------
when nitrogen is not a limiting nutrient source; and the presence of
248 mg/1 NH^-N in the leachate sample indicated that this was probably
the case. Comparison of the amino acids measured with ninhydrine method
after 6N HC1 hydrolysis to the organic nitrogen content measured with
Kjeldahl method and multiplied by 6.25 (Bunch eit al_., 1966) showed that
90% of the nitrogeneous compounds were present as free or bound amino
acids.
Analyses of the carboxyl compounds showed that the organics associated
with this group increased in the membrane fractions having decreasing
molecular weight (Figure 8) confirming the results in Figure 6. A
titration with NaOH was made using Fraction 8b and 9b (Figure 7) which
after acidification with HC1 showed a weak inflection point at a pH
of 4.30 for fraction 8b and at 4.37 for fraction 9b. A sharper inflec-
tion point was noted at 7.2 for both fractions. Assuming that the
amount of base necessary to raise the pH of the solution from the first
inflection point to the second is due solely to the dissociation of the
carboxyl group, it was calculated that the density of carboxyl groups
was 10.1 meq/gC for fraction 9b. The corresponding density as measured
with the hydroxyl amine test indicated that 8.8% of the carbon of that
membrane fraction was associated with the carboxyl group. The titra-
tion resulted in a density of 6.6 meq/g C for fraction 8b. The corres-
ponding value measured with the hydroxylamine test was 5.2%, showing
that both the titration and hydroxyl amine test resulted in values of
the same relative magnitude.
Owing to the presence of the free volatile fatty acids that pass the
500 MW UF membrane, the highest percentage of carboxyl groups was present
in the 500 MW permeate. Further analysis of the 500 MW UF permeate by
GC showed the presence of acetic acid (1748 mg/1), propionoic acid
(509 mg/1), isobutyric acid (312 mg/1), butyric acid (3074 mg/1), isoval-
eric acid (717 mg/1), valeric acid (564 mg/1), and hexanoic acid (1488
mg/1). The carbon from these free volatile fatty acids comprised 78%
of the TOC of the 500 MW UF permeate or 49% of the initial TOC of the
leachate sample.
58
-------
The results of the analyses showed that as much as 86% of the carbon
present in the 500 MW UF permeate was identified, followed by 50%
identified in the 100,000 MW UF retentate. The intermediate fractions
were least characterized indicating that they consisted of complex
molecular structures. Of the total organics present in the sample
7U was identified.
SOLVENT EXTRACTION OF DIFFERENT MOLECULAR WEIGHT FRACTIONS
The last step in the characterization scheme consisted of solvent
extraction to separate the different membrane UF fractions according
to their polarity. The hexane extraction of the 10,000 MW UF reten-
tate (Fraction 3b, Figure 7) showed that this step extracted only 13.8
mg/1. If most of the hexane extract consisted of palmitic acid
(Shorland, 1963), to give an organic carbon/weight ratio of 0.75, this
fraction would represent only 0.5% of the organic carbon in the 10,000
MW UF retentate (Table 11). The residue of the hexane extract had a
relatively white to light brown color and a very sharp and pungent
odor. Since the hexane residue contained some dark brown iron hydroxide
colloids originating from the water layer, the hexane layer was filtered
prior to drying.
The IR spectrum of the residue (Figure 9a) showed strong absorption
bands near 2900 cm"1 and 1470 and 1390 cm, indicating the dominant
influence of aliphatic structures (Table 12). The band at 1720 cm"
represents the C=0 stretch of the carbonyl in the carboxyl group
whereas the shoulder at 1750 cm reflects the fraction of carbonyls
associated with ester linkages which also shows at 1270 cm (Table 12).
The IR spectrum therefore, would indicate that the majority of this
fraction consists indeed of fatty acids some of which are still present
in ester linkages. Colorimetric analysis of the redissolved hexane
residue showed presence of carbohydrates. These results therefore
indicate that the residue of the hexane extract will slightly overestimate
the concentration of fatty acids because other organic fragments are
also extracted.
59
-------
Table 11. Effectiveness of the Hexane and Butanol Extraction to Remove
Organics from the Different Molecular Weight Fractions
Membrane UF
Fraction
10,000 MW UF
retentate
(fraction 3b)
Hexane extract
as percentage
of the carbon
0.5
Extraction of
specific organic
0.8% of carbohydrates in
UF retentate extracted
with hexane
Butanol extract
as percentage
of the carbon
2.5
Extraction of
specific organics
4% of carbohydrates
and 100% of aromatic
hydroxyls present in
UF retentate extracted
with butanol
500 MW UF retentate
combined
10,000 MM and
100,000 MW UF
permeate (fraction
9b and lib)
100% of aromatic
hydroxyls present
in UF retentate
extracted with
butanol
500 MW UF permeate
(fraction 7b)
1.5
0.5% of carbohydrates
10% of aromatic hydroxyls
present in UF permeate
extracted with hexane
65% of aromatic
hydroxyls present
in UF permeate
extracted with butanol
-------
Wavelength, microns
35 4 4.5 5 55 6 66 7 8 9 K) 12 14 16 20 30
4000
3000
2000
1600
1200
800
400
Wavenumber, cm
-i
Figure 9. IR Spectra of the Residue of the Hexane and Butanol Extracts:
(a) Hexane Extract of the 10,000 MW UF Retentate (b) Butanol Extract of
the 10,000 MW UF Retentate (c) Butanol Extract of the Hydrolysed 10,000
MW UF Retentate (d) Butanol Extract of the 500 MW UF Retentate (e) Hexane
Extract of the 500 MW UF Permeate
61
-------
Table 12. Infrared Absorption Dands and their Tentative Assignments
for the Residue of the Hexane and Butanol Extracts
Hexane
extract of
Butanol
extract of
Acidified
butanol
extract of
Butanol
extract of
500 MW UF
Hexane
extract of
500 MW UF
10,000 MW UF 10,000 MW UF 10,000 MW UF retentate
retentate retentate RETENTATE (Fraction 9b permeate
(Fraction 3b)(Fraction 3b)(Fraction 3b) and lib) (Fraction 7b)
3400
2960
2860
2730
2380
1750
1720
3420
2990
2960
2880
3400
2920
2840
1720
1630
1720
1640
3460
2980
2950
2880
2350
1740
1650
1620
1530
1470
1390
1270
1100
1030
810
740
730
1440
1390
1460
1390
1470
1390
1080
1160
1100
1050
1180
760
3440 H bonded 0-H stretch,
N-H stretch
2960 methyl C-H stretch
2930 methylene C-H stretch
2880 aldehyde C-H stretch
carboxyl 0-H stretch
phosphorus P-OH stretch
carbonyl C=0 stretch
in esters
1720 carbonyl C=0 stretch
in carboxyl groups
N-H bending, C=C stretch, C=0
stretch of ketones, amides
and carboxylate ions
N-H bending
1550 N-H bending, C-N stretch
of peptides
1440 CH2 scissor bending; benzene
skeletal vibration
1390 Symmetric C-CH? bending; 0-H
bending of carboxylic acid.
Symmetric 0-C-O Stretch of
carboxylate ion
Symmetric C-O-C stretch of
esters C-OH stretch of
carboxylic acid
aromatic structures
1100 C-0 stretch of prim alcohol
aromatic structures
C-N stretch of amino groups
C-0 stretch of prim alcohol
aromatic structures
out of plane bending of NH-
aromatic out of plane bending
of C-H
CH2 rocking vibrations
62
-------
The residue of the butanol extraction of the 10,000 MW UF retentate
was of a greater magnitude than that of the hexane extract. If 50%
of the weight of the butanol residue consisted of carbon, 2.5% of the
organic carbon in the 10,000 MW UF retentate is comprised of this
group (Table 11). The residue of the butanol extract had a dark brown
oily appearance and an acidic and "moldy" odor. The IR spectrum of
the butanol residue of the 10,000 MW UF retentate of the lysimeter
sample (Figure 9b) showed the dominant influence of hydroxyl groups
such as those present in carbohydrates (1080 cm" ) and carbonyl groups
(1640 cnf ). The strong peak at 1390 cm is probably due to the
scissor bending of a CH2 group adjacent to a carbonyl group. The IR
spectrum is almost identical to a humic - carbohydrate-like high molecular
weight fraction isolated from effluent from an activated sludge unit
(DeWalle and Chian, 1974). Further acidification of the dry residue with
HC1 and subsequent drying reduced the peak at 1390 cm" . This reduc-
tion indicates that part of the peak was contributed by the symmetric
0-C-O stretch of carboxylate ions. Since the carbonyl stretch at 1630
cm did not shift to 1720 cm" , however, the majority of the carbonyl
groups in this fraction are present not as carboxyl carbonyls, but as
aldehydes and ketones. Moreover, the reduction of the peak at 1080
cm may indicate that some carbohydrates were hydrolyzed (Tan, 1970).
Colorimetric tests confirmed the presence of carbohydrates and aromatic
compounds. While only 4% of the carbohydrates present in the 10,000
MW UF retentate were removed, all of the aromatic hydroxyl groups were
extracted with the butanol (Table 11). If the model compound reflects
the actual extracted organics, it can be calculated that about 30% of
the weight of the butanol extract of the 10,000 MW UF retentate
was associated with aromatic hydroxyl compounds. Since the aromatic
compounds only represent about 3% of the carbon in this fraction,
it was concluded that the butanol preferentially extracted this group.
The above results of the extraction were confirmed by using model compounds
that resemble the different classes of organics present in the 10.000MW UF
63
-------
retentate. The removal of 93% by butanol extraction of a 1000 mg/1
tannic acid solution demonstrated the affinity of butanol towards
aromatic hydroxyl compounds. Only 14% of a bovine albumin solution
was extracted while no removal was observed with a starch solution.
Extraction of the 500 MW UF retentate, i.e., fractions 9b and lib
combined (Figure 7), with hexane did not result in a residue. This
indicates that the fatty acids or other non-polar organics are only
present in the high molecular weight fraction, a conclusion similar
to the one reached by Ogner and Schnitzer (1970). The butanol extract,
however, was substantially larger and represented 9% of the TOC
(Table 11). The butanol layer was not as dark brown as that of the
10,000 MW retentate (Fraction 3b, Figure 7). Before complete evapora-
tion of the butanol a white semi-crystalline substance was visible which
darkened after the last butanol had evaporated. The appearance in the
IR spectrum (Figure 9) of strong influences of the carbonyl in the
carboxyl group at 1740 cm (Table 12) indicated that the carboxyl
group was more dominant in this fraction than in the butanol extract
of the 10,000 MW UF retentate. This trend parallels the results of
the hydroxylamine test for carboxyls (Figure 8). Comparison with other
IR spectra (Black and Christman, 1963; Stevenson and Goh, 1971) showed
that this fraction resembled a fulvic acid. The butanol was able to
extract almost 100% of the aromatic hydroxyls present in this fraction
(Table 11).
The hexane extract of the 500 MW permeate represented 1.5X of the organic
carbon of that fraction. The IR spectrum (Figure 9) showed two dominant
peaks at 1550 and 1440 cm" representing the asymmetric and symmetric
0-C-O stretch of the ionized carboxyl group. These peaks indicate that
the hexane extracted some of the free volatile fatty acids present in
the 500 MW UF permeate. The small peak at 1110 cm and the broad
peak at 3440 cm"1 may indicate that some carbohydrates and hydroxyl
compounds were also extracted with the hexane, a possibility supported
by colorimetric tests (Table 11).
64
-------
The butanol extract of the 500 MW UF permeate represented 1% of the
carbon of that fraction. The IR spectrum was similar to that of the
hexane extract of the 500 MW UF retentate with the exception that the
peak near 2930 cm" was smaller, probably because of the extraction
of lower molecular weight free volatile fatty acids, whle the peak
near 3440 cm was broader.
The above observations indicate a short-coming of the extraction scheme
using hexane and butanol in that the hexane extraction may overestimate
the amount of lipids since some organic fragments are also extracted.
In addition the hexane will extract some low molecular weight free
volatile fatty acids when present in the sample. The butanol used to
extract the fulvic acid fraction may also extract some of the free
volatile fatty acids. Lamar and Goerlitz (1966) found that although
humic substances were the main fraction obtained with butanol extrac-
tion at acidic pH, up to 10% of the residue consisted of fatty acids.
On the other hand, the butanol does not extract all of the fulvic-like
material in the 500 MW UF retentate. DeHaan (1972) similarly noted
that only 50% of the color was extracted with butanol at pH 2. Butanol,
however, does have a preference for aromatic hydroxyl compounds present
in this fraction. DeHaan (1972) noted that butanol preferentially
extracts organics with a high carboxyl and carbonyl content.
FRACTIONATION AND ANALYSES OF OTHER SAMPLES
The effectiveness of the membrane fractionation was also tested on two
other leachate samples obtained from a control and recirculation pilot
landfill of intermediate age, operated since October 1971 by the City
of Santa Rosa (CA) (County of Sonoma, 1973). The membrane separations
were conducted with two aliquots passed separately through either the
500 MW or 10,000 MW UF membrane. The 10,000 MW UF membrane was found
to retain 13% and 12% respectively, whereas the 500 MW UF retained 30%
and 22% respectively of the initial TOC of the sample. The results of
the specific organics present in each fraction as shown in Figure 10
show a similar trend as observed in Figure 8. Comparison shows that
65
-------
UJ
c
41*
2
o
*
Carbohydrates As Dextrose
Proteins As Leucine
Carboxyl As Acetic Acid
Car bony I As Acetophenone
Aromatic Hydroxyl As Tannic Acid
/'Percentage Of TOC Identified In
Molecular Weight Fraction
<500MW 500-IO,OOOMW >IO,OOOMW
Molecular Weight Fraction
Figure 10. Characterization of Different Membrane UF Fractions in a
Leachate Sample Collected from a Control (a) and a Recirculation
(b) Landfill in Sonoma County
66
-------
the identified organics represent a lower percentage of the organic
carbon. Since the landfill is located in a warmer climate, a relatively
larger percentage of the carbohydrates and proteins may have been
degraded at the higher temperature. The carboxyl group, on the other
hand, is present in relatively large amounts. As it represents one
of the most oxidized forms of organic matter, a higher carboxyl group
density would indicate a higher degree of oxidation. The greater
stabilization of the Sonoma samples is further substantiated by the
lower BOD/COD ratio of 0.62 and 0.65 as compared to 0.80 for the
University of Illinois sample.
The membrane fractionation was finally tested with a sample obtained
from a relatively old landfill in Dupage County in Northern Illinois
(Hughes et §1., 1971). Concentration through a 10,000 MW UF membrane
did not result in retention of any material, while concentration of
a separate aliquot through a 500 MW UF membrane only retained 5% of the
original TOC of the sample. As the fatty acids were present in this
sample in a concentration below that of the detection limit of the GC,
it was concluded that most of the organics were present as low molecular
weight refractory molecules. The 500 MW UF retentate was further
separated using a Sephadex G-75 column (Figure 11). The TOC data show
that only 7% of the 500 MW UF retentate corresponding to 0.5% of the
original TOC is present as the high molecular weight humic carbohydrates,
The percentage of carbohydrates in this MW fraction is approx-
imately equal to the percentage in the high molecular weight fraction
of the leachate obtained from younger fills. The concentration of
carbohydrates therefore, tends to reflect the magnitude of the high
molecular weight fraction independent of the age of the sample. Since
the phenolic hydroxyl group obtained relatively high concentrations
in the low molecular weight fraction of the 500 MW UF retentate, it
may be used as an indicator for the fulvic like material.
67
-------
<»
S «8
c
O tt>
?!=: «6
01
O I—
c+ n> m
o> o» — '
o-s o c
-J _i. 3- CU
us M a> c-»-
O» CD rt- (D
3 o. n>
a o — i(o
Q.H- n>
O> (71
•n — • o o
C _3 O
3 O
O -* Qi 3
rl-ca S
-J. Oi OO
o 3 a> c=
3 -•••a ~n
o» o 3-
— ' 01 73
C~> Q. (D
cr> a> a> en a>
TJ
n> o o
o o -t»
-h C rt-
o § m
0«
vt
114
o 12
10
^
fa
o 6
.O
8 4
2
- 9
- 8
E
-5
h-Q-2
- I
u 0
18 n 18
- 16 -
o»
(Q
(D
/\ NCorboxyl 1
50
Elution Volume, ml
16
14
12 p
3
10 £
o>
o
81
6 §
a
O
2
0
-------
The last part of the study consisted of analyzing seven leachate samples
in addition to the four that were already studied in greater detail.
The samples were collected from different landfills in the United States
representing a range of climatic conditions. When the free volatile
fatty acids, calculated as percentage of the TOC of the sample, were
related to the age of the landfill, the results showed a gradual decrease
of the magnitude of this fraction with increasing age of the fill.
Since these organics are relatively biodegradable, a similar decrease
was observed in the BOD/COD ratios. Separate aliquots were passed
through the 10,000 MW UF and the 500 MW UF during membrane separation
of the seven samples. However, since some of these samples resulted
in a substantial flux decrease, a 1000 MW instead of the 500 MW UF
membrane was used. The results of the separation, shown in Figure 12a
and 12b, illustrate the gradual decrease of both the high molecular
weight humic-carbohydrate-like complex and the lower molecular weight
fulvie-like material with increasing age. It was noted that the high
molecular weight fraction as obtained in the 10,000 MW UF retentate
showed a more rapid decrease then the low molecular weight fraction
as obtained in the 500 MW UF retentate. This pattern may indicate that
the former fraction is more liable during progressing biodegradation.
Results of the membrane fractionation reflect the changes observed for
carbohydrates, proteins and aromatic compounds as measured in the
unfractioned leachate samples (Figure 13). The carbohydrates and
proteins present in relatively large quantities in the high molecular
weight fraction show a relatively more rapid decrease with increasing
age of the fill than the aromatic hydroxyl compounds mainly present
in the low molecular fulvic-like fraction. These data may therefore
indicate that increased stability with respect to biodegradation is
observed for free volatile fatty acids < carbohydrates, proteins, humic-
like substances < aromatic hydroxyl, carboxyl compounds, and fulvic-
like substances.
69
-------
~g 40
o> o
SE
£2 30
S§ 2°
I0
8|
£i
•
A
(a)
*°
eu
•g
•g
'5
ff " ' I 1
(b)
/\
/ .^7»
« I/ ^rA
*2 t ^V
8| I0
•-E
02
||
^ Qj
^ \
' \
\.
\
10
Time, years
15
Figure 12. Percentage of TOC Retained with Either 10,000 MW (a) or
500 W and 1000 MW (b) UF Membranes in Leachate
Samples Collected from Landfills of Different Ages
70
-------
c D
^1 4
$^
"to 2
w
oo o
"c 0>
Sf
>. 0>
ll
K °<
1 — 1 1 1
X
CT *
^ x
^ ^
\^
0
£ \^
' V —Hi
A •!
•
T
^^-* •*"" ~ ~ """""" "^- — -
3 -^ • "^' 1—
Sf '
> ' 5 10 IS
Time, years
Figure 13. Carbonhvdrates (a), Proteins (b) and Aromatic
Hydroxyls (c) in Leachate Samples Collected
from Landfills of Different Ages.
71
-------
DISCUSSION
MEMBRANE SEPARATION SEQUENCE
The results of the separation of organic matter by membrane ultrafil-
tration indicate that a larger percentage of the TOC was retained when
the sample was directly passed through the 500 MW UF membrane as com-
pared to the sequential membrane separation through the 10,000 MW
followed by the 500 MW UF membrane. Blatt et^ a]_. (1970) similarly
indicated that formation of a polarized layer by the highest molecular
species resulted in an increased retention of the intermediate molecular
weight fraction present in milk whey; the lowest molecular weight
fraction, however, passed through the membranes unimpeded. They also
noted that although the rejection of human serum albumin (65,000 MW)
with a 100,000 MW UF membrane was zero, its rejection increased rapidly
when y-globulin (160,000 MW) was added. The removal of high molecular
weight organics with the 10,000 MW UF membrane, therefore, reduced
the degree of polarization at the 500 MW UF membrane during permeation
of the 10,000 MW UF permeate. In order to give similar degrees of
polarization to each of the ultrafiltration membranes, a parallel
ultrafiltration procedure would give more representative results. The
least reliable results are obtained when the retentate of the lowest
MW UF membrane is passed through UF membranes with successively larger
cut offs due to the accumulation of high molecular weight organics
that experience such a large degree of polarization, and the absence of
salts that would otherwise reduce the double layer of the high
molecular weight organics.
COMPARISON OF SEPARATION TECHNIQUES
Comparison of the membrane ultrafiltration and the gelpermeation chroma-
tography showed that the molecular weight estimated by the former
method was approximately four times higher than by the latter method.
However, this difference may be due to formation of a polarized layer
72
-------
which increased the rejection of the membrane. Formation of such a
polarized layer may explain why Gjessing (1971) and Ferry (1936) noted
that similar rejections were obtained when the UF membrane had a pore
size about three times larger than a dialysis membrane. The opposite
effect, was also observed with a sample obtained from a stabilized
landfill as no organics were present in the 10,000 MW UF retentate
while the excluded Sephadex fraction of the 500 MW UF retentate
indicated that 0.5% of the initial TOC should have been retained with
the 10,000 NW UF. Discrepancies between the separation techniques may
also be due to the nature of each process. Baker and Stratman (1970)
showed that membrane rejection of globular organics is higher than
that of linear molecules of the same molecular weight due to the
deformation and stretching of the latter. A linear molecule for
example, was found to have the same retention as a protein one tenth
its molecular weight. This behavior is exactly opposite to that
noted for both gel permeation chromatography and dialysis, as they are
based on diffusion through microporous capillaries. For example,
linear molecules are eluted in the same fraction as protein twice
their molecular weight (Fisher, 1965). The discrepancies between the
results of the gelpermeation and membrane ultrafiltration could
indicate that in the sample obtained from a recent fill globular high
molecular weight organics are more dominant, while organics from the
stabilized fill have more linear organic molecules. However, additional
research is necessary before final conclusions can be made.
ORGANICS IN DIFFERENT MOLECULAR WEIGHT FRACTION
The results of the analysis of the organics separated by membrane
fractionation and gelpermeation chromatography tend to confirm earlier
studies that separated soil or aquatic humic substances. Table 13
shows that the carbohydrates and hydrolyzable amino acids are present
in high concentration in the high molecular weight fraction while
carboxyl groups, aromatic hydroxyl groups, color and fluorescence
are contributed by fractions of lower molecular weight. Most of the
above mentioned studies used methods for functional group analysis
73
-------
Author
Origin
Table 13. Organic Composition of Fractionated Humic Substances
Organic Fraction Chemical Composition
Dubach et al.
(1964)
Schnitzer and
Skinner
(1968)
soil fulvic acid high MW excluded from G-75
low MW included in G-75
soil fulvic acid high MW excluded from G-50
low MW excluded from G-10
Swift and soil humic acid
Posner (1972)
Leenheer and
Malcolm (1973) soil fulvic acid
different Sephadex grades
Povoledo and
Gerletti
(1968)
Rashid and
King (1971)
neutral charged high MW fraction
negative chared low MW fraction
riverwater organics neutral charged fraction
unfractiona'ted
lake sediment
fulvic acid
marine sediment
humic acid
high MW excluded from G-25
low MW included in G-75
high MW excluded from G-200
low KW included in G-50
Reuter and
Perdue (1971)
river fulvic acid 'high MW excluded from G-50
low MW included in G-50
21% nitrogenous organics, 25% as carbohydrates
9% nitrogenous organics, 5% as carbohydrates
Specific absorb, increased with decreasing MW
4.8 meq/g (9.6 meq/gC) carboxyl groups
3.4 meq/g aromatic hydroxyl groups
7% nitrogenous organics
8.7 meq/g (16.9 meq/gC) carboxyl groups
6.2 meq/g aromatic hydroxyl groups
2% nitrogenous organics
C=0 and C-OH decreased with decreasing MW
28% nitrogenous organics in high MW fractions
13% nitrogenous organics in low MM fractions
43% carbohydrates
20% nitrooenous organics
10% carbohydrates; 14% nitrogenous organics
contained most of color and carboxyl groups
5% carbohydrates
8% carbohydrates
1% hydro1yz?b!e amino acids
contains half of carbohydrates and protein's
contains most of the color, fluorescence and
aromatic hydroxyls
2.4 meq/g (4.0 meq/gC) carboxyl groups
0.3 meq/g aromatic hydroxyl groups
28% nitrogenous organics
3.6 meq/g (7.8 meq/gC) carboxyl groups
1.0 meq/g aromatic hydroxyl groups
26% nitrogenous organics
4.2 meq/g carboxyl groups
9.6 meq/g carboxyl groups
-------
developed in the soil science (Schnitzer and Kahn, 1972). Many tests,
however, give arbitrary results and the outcome of the carboxyl and
aromatic hydroxyl test, for example, is dependent upon the generally
unknown spectrum of dissociation constants (Van Dyk, 1966). The
colorimetric tests selected in the present study for functional group
analyses may therefore be less arbitrary. All of the studies using
samples from widely different environments, indicate a similar distri-
bution of specific organics in the different molecular weight fractions.
This may indicate that similar bacterial or chemical processes are
governing the composition of the organics in natural environments.
Several studies do indicate that bacterial processes are the most
important. Identification of individual sugars in soil polysaccharides,
for example, showed that xylose and glucose were present in lower con-
centrations than in the original plant material, while galactose,
mannose, arabinose and uronic acid were present in higher concentra-
tions. This relative distribution indicates that most of the sugars
are derived from microorganisms (Forsyth, 1950). The presence of
relatively large amounts of amino acids in the high molecular weight
fraction is probably the result of excreted bacterial amino acids
that preferentially attach to this fraction. Broadbent (1968) noted
that labelled inorganic nitrogen and organic carbon often appear in
the high molecular weight humic fraction as acid hydrolysable amino
acids. The aromatic compounds present in the 500 MW UF retentate
may have been excreted by fungi and bacteria as orsellinic acid,
salicylic acid, hydroxybenzoic acid and phenols and condensed to
larger molecules as a result of the pH increase of the solution
(Hoak, 1963).
MOLECULAR WEIGHT DISTRIBUTION
Related studies that used dialysis or ultrafiltration membranes as
listed in Table 7 and 8 were plotted as percentage of the organic
matter retained versus nominal molecular weight cutoff (Figure 14).
The results in Figure 14 indicate that the majority of the aquatic
organics have molecular weights ranging from 10,000 to 100,000.
75
-------
No systematic differences were noted between the molecular weight
distributions determined by dialysis or membrane ultrafiltration.
As compared with the other studies, the samples analyzed in the present
study have the largest percentage of low molecular weight organics.
This difference is mainly due to the large fraction of free volatile
fatty acids generated during anaerobic acid fermentation. Since
most of the investigations in Figure 14 analyzed aerobic stabilized
organics, the absence of low molecular weight organics is to be
expected.
STABILITY OF ORGANIGS IN DIFFERENT MOLECULAR WEIGHT FRACTIONS
Bacterially derived organics are subject to degradation. Degens et al.
(1964), for example, noted a rapid decrease of carbohydrates with
increasing time of burial or increasing depth in sediments while a
less rapid decrease was observed for ami no acids. No reduction in
amount of humic substances was noted, while aromatic compounds some-
times showed an increase with depth of burial. Ishiwatari (1971)
noted that carbohydrates showed the largest relative decrease with
increasing depth of burial while hydrolyzable amino acids experienced
a less rapid reduction. Humic acids did not show any decrease and
in fact showed some increase. Swain (1970) also noted a gradual
decrease of the carbohydrate content with increasing depths in lake
sediments and an occasional increase of the organic nitrogen followed
by a subsequent decrease. Aromatic compounds are generally less subject
to decomposition than aliphatic compounds such as peptides and carbo-
hydrates, and incorporation of proteins and carbohydrates in aromatic
hydroxyl compounds renders the complex less susceptible to bacterial
degradation (Benoit and Starkey, 1968). The results of these studies
therefore agree with the present findings in that increased stability
was noted for: free volatile fatty acids < carbohydrates < proteins
< aromatic hydroxyls.
The more rapid decrease of the high molecular weight humic-carbohydrate-
like material as compared to fulvic acid-like material as found in the
76
-------
Mtmbront Diolysis
Membrane Ultrafiltration
-s
(D
Martin etol (1972) Humic Acid
Martin etol (1972) Fulvic Acid
Packham (1964) Humic Acid
Packham (1964) Fulvic Acid
Shapiro (1964)
Black And Christmon (1963)
G jessing (1971)
Bunch etal
JDO
n> —•
3 01
01 «-!--$
O. CO S
<-«• ffi
3 c -••
n>
• Barber (1968)
O Ghossimi (1967)
• Colder And Keorsley (1974) Fresh Water
• Colder And Keorsley (1974) Salt Water
A Ooig And Martin 41974)
O This Study
Dialysis Membrane (Diameter)
24& 35! 40& 48A
T
T T
* Cameron etal (1972)
0 Wilander (1972)
* 6 jessing (I97O)
V 6jessing (1971)
A Gjessing (1973)
X Schindler And Alberts (1974)
500&
1*2(264)
Membram Ultrafiltration (Diameter)
UMOOOX) PM30M6A) XM»OA(IO4A)
PMIO(36A) UM2pEO6A) XM5O(64&L
IQPOO
Molecular Weight
oopoo
-------
present study may be due to aerobic biological stabilization as the
ORP in the leachate increased from -200 mv to about +200 mv with
increasing age of the landfill. As the high molecular weight fraction
only decreased in magnitude but not in relative composition (percentage
carbohydrates and amino acids), one may conclude that entire parts of
the high molecular weight molecule were hydrolysed which then become
more susceptible to further enzymatic and bacterial degradation. The
more rapid decrease of the high molecular weight fraction as compared
to the fulvic like materials is probably not due to its preferential
coagulation, since no correlation could be established between the
magnitude of this fraction and the concentration of Ca and Fe which
could have caused such coagulations. Preferential leaching to the
subsoil, is also not likely to cause a more rapid decrease in the high
molecular weight fraction since such leaching would rather cause a
decrease of the low molecular weight fulvic-like material.
78
-------
REFERENCES
Benoit, R. E. and Starkey, E. L., "Inhibition of Decomposition of
Cellulose and Some other Carbohydrates by Tannin," Soil Sci, 105,
291 (1968).
Baker, R. W. and Strathman, H., "Ultrafiltration of Micromolecular
Solution with High Flux Membranes," Jour. Appl. Polymer Science
14, 1197 (1970).
Barber, R. T., "Dissolved Organic Carbon from Deep Waters Resists
Microbial Oxidation," Nature 220. 274 (1968).
Barth, E. F. and Acheson, N. H., "High Molecular Weight Materials in
Tap Water," Jour. Amer. Water Works Assn. 54, 959 (1962).
Black, A. P. and Christman, R. F., "Characterization of Colored Surface
Waters," Jour. Amer. Water Works Assn. 55_, 753 (1963).
Black, A. P. and Christman, R. F., "Chemical Characteristics of Fulvic
Acids," Jour. Amer. Water Works Assn. 55_, 897 (1963).
Blatt, W. F., "Solute Polarization and Cake Formation in Membrane
Ultrafiltration: Causes, Consequences and Control Techniques,"
in Membrane Science and Technology, Ed. Flinn J. E., p. 47 (1970).
Broadbent, F. G., "Nitrogen Immobilization and Relation to N Containing
Fractions of Organic Matter," in "Isotopes and Radiation in Soil
Organic Matter Studies," Internet. Atom. Energy Agency, Vienna,
p. 131 (1968).
Bunch, R. L. je_t al_., "Organic Materials in Secondary Effluents,"
Jour. Water Pollut. Control Feder. 33_, 122 (1961).
Burrows, W. D. and Rowe, R. S., "Ether Soluble Constitutents of Landfill
Leachate," Jour. Water Pollut. Control Feder. 4£, 921 (1975).
Cameron, R. S. £t al_., "Molecular Weight and Shape of Humic Acid from
Sedimentation and Diffusion Measurements and Fractionated Extracts,"
J. Soil Sci. 23_, 394 (1972).
13
Calder, J. A. and Kearsley, F., "Molecular Weight Distribution and C
Content of the Dissolved Organic Carbon in a Salt Marsh," Paper
presented at the 37th Ann. Meeting, Americal Soc. Limnol.
Oceanography, University of Washington, Seattle (1974).
Carr, C. W., "Graded Collodion Membranes for Separation of Small
Molecules," Science 125. 1245 (1957).
79
-------
Chian, E. S. K. and Mateles, R. S., "Growth of Mixed Cultures on Mixed
Substrates I. Continuous Culture," Applied Microbiology 16, 1337
(1968). ~~
Chian, E. S. K. and DeWalle, F. B., "Sanitary Landfill Leachates and
their Treatment," J. Env. Engr. Div. ASCE. 102. EE2, 411 (1976).
Christman, R. F. and Ghassemi, M., "Chemical Nature of Organic Color
in Water," Jour. Amer. Water Works Assn. 58_, 723 (1966).
County of Sonoma, " Sonoma County Refuse Stabilization Study," Second
Annual Report, Dept. of Public Works, Santa Rosa, CA (1973).
Craig, L. C. et aj_., "Dialysis Studies II Some Experiments Dealing with
the Problem of Selectivity, J. Amer. Chem. Soc. 79. 3729 (1957).
Degens, E. T. e£al_., "Biochemical Compounds in Offshore California
Sediments and Seawater," Geochim. Cosmochim. Acta 28. 45 (1964).
DeHaan, H., "Molecular Size Distribution of Soluble Humic Compounds
from Different Natural Waters," Freshwater Biolog. 2, 235 (1972).
DeWalle, F. B. and Chian, E. S. K., "Removal of Organic Matter by
Activated Carbon Columns," Jour. Env. Engr. Div. ASCE 100, 1089
(1974). a
DeWalle, F. B. and Chian, E. S. K., "Kinetics of Formation of Humic
Substances in Activated Sludge Systems and their Effect on Floc-
culation," Biotechnol. Bioengrng. 16. 739 (1974).
Doig, M. T. and Martin, D. F., "The Effect of Naturally Occuring Organic
Substances on the Growth of a Red Tide Organisms," Water Research
8, 601 (1974).
Dor, I., "High Density Dialysis Culture of Algae and Sewage," Water
Research 9. 251 (1975).
Dubach, P. ejt a]_., "Chemical Investigation of Soil Humic Substances,"
Geochim. Cosmochim. Acta 28, 1567 (1964).
Durbin, R. F., "Osmotic Flow of Water Across Permeable Cellulose
Membranes," J. Gen. Physiol. 44, 315 (1960).
Elford, W. J., "A New Series of Graded Collodion Membranes Suitable
for General Bacteriological Use, Especially in Filterable Virus
Studies," Jour. Pathol. Bacteriol. ^4_, 505 (1931).
Elford, W. J., "Principles Governing the Preparation of Membranes Having
Graded Porosities," Faraday Soc. Trans. 33_, 1094 (1937).
Elford, W. J. and Ferry, J. D., "The Ultrafiltration of Proteins through
Graded Collodion Membranes." Biochem. Jour. 3_2, 84 (1936).
80
-------
Fang, H. H. P. and Chian, E. S. K., "Removal of Alcohols, Amines and
Aliphatic Acids in Aqueous Solution by NS-100 Membrane," Jour.
Appli. Polymer Sci. 19, 1347 (1975).
Ferry, J. D., "Ultrafliter Membranes and Ultrafiltration," Chem. Rev.
18., 373 (1936).
Fischer, C., "An Introduction to Gel chromatography," Amer. Elsevier
Publ. Comp. Inc., New York (1965).
Forsyth, W. G. C, "Studies on the More Soluble Complexes of Soil Organic
Matter, 2," Biochem. J.. 46. 141 (1950).
Ghassemi, M., "Chemical Nature and Physical Properties of Color Causing
Organic Substances in Natural Waters," PhD dissertation, Dept. of
Civil Engineering, University of Washington, Seattle, 142 p. (1967).
Gjessing, E. T. and Lee, G. F., "Fractionation of Organic Matter in
Natural Waters on Sephadex Columns," Env. Sci. Techn. ]_, 631 (1967).
Gjessing, E. T., "Ultrafiltration of Aquatic Humus," Env. Sci. Techno!.
4, 437 (1970).
Gjessing ,E. T. , "Effects of pH on the Filtration of Aquatic Humus Using
Gels and Membranes," Schweiz. Zeitschr. Hydro!. 3JU 592 (1971).
Gjessing, E. T., "Gel and Ultra Membrane Filtration of Aquatic Humus:
A Comparison of the Two Methods," Schweiz. Zeitschr. Hydro!. 35,
286 (1973)
Jacobs, S., "The Distribution of Pore Diameters in Graded Ultrafilter
Membranes," Filtration Separation, 9, 525 (1972).
Good, C. M. et aj_., "Molecular Filtration, Application in Virus
Purification and Binding Studies," Amer. Laboratory 6, 71 (1974).
Hoak, R. D., "Recovery and Identification of Organics in Water," in
Advances in Water Pollution Research, Ed. Southgate B.A., Pergamon
Press, Vol. 1, p. 163 (1963).
Hughes, G. M. ejb al., "Hydrogeology of Solid Waste Disposal Sites in
N.E. Illinois?1" Report SW12d US-EPA, Washington, D.C. (1971).
Hunter, J. V. and Heukelekian, H. , "The Composition of Domestic Sewage
Fractions," Jour. Water Pollut. Control Feder. 37. 8, 1142 (1965).
Ishiwatari, R., "Molecular Weight Distribution of Humic Acids from Lake
and Marine Sediments," Geochemical Journal 5. 121 (1971).
Kesting, R. E., "Synthetic Polymer Membranes," McGraw Hill, New York,
307 p. (1971)
81
-------
Konanova, M. M., "Soil Organic Matter," Second Ed., Pergamon Press,
New York (1966).
Lamar, W. L. and Goerlitz, D. F., "Organic Acids in Naturally Colored
Surface Waters," Geol Survey Water Supply Paper 1817A, Washington,
D.C. (1966).
Leenheer, J. A. and Malcolm, R. L., "Fractionation and Characterization
of Natural Organic Matter from Certain Rivers and Soils by Free
Flow Electrophoresis," U.S. Geological Survey Water Supply Paper
1817 E, 14 p. Washington, D.C. (1973).
Lin, Y. H., "Acid and Gas Production from Sanitary Landfills," PhD
dissertation, Department of Civil Engineering, University of
Virginia, Morgantown, WV (1966).
Madras, S. et al., "A Preliminary Study of the Permeability of Cellophane
to LiquWTrr Can. J. Res. 27. B, 764 (1949).
Mao, M. C. M. and Pohland, F. G., "Continuing Investigations and Landfill
Stabilization with Leachate Recirculation," Special Research Report,
Department of Civil Engineering, Georgia Inst. Techn., Atlanta, GA
(1973).
Martin, J. P. et aj_., "Synthesis of Phenols and Phenolic Polymers by
HendersanuTa" Formuladea in Relation to Humic Acid Formation," Soil
Sci. Soc. Amer Proceed 36. 311 (1972).
McBain, J. W. and Kistler, J., "Ultrafiltration as a Test for Colloidal
Constituents in Aqueous and Non Aqueous Systems," J. Phys. Chem.35. 130 (1931)
Meschia, G and Stniker, I.,"Experimental Study of Osmosis Through a
Celloidon Membrane," J. Gen. Physio. 42, 429 (1958).
Michaels, A. S., "Ultrafiltration," in Progress in Separation and Purifi-
cation, Vol 1, E.S. Perry Ed., Interscience, New York, p. 297 (1968).
Morton, T. H., "The Dyeing of Cellulose with Direct Dyestuffs; the
Importance of the Colloidal Constitution of the Dye Solution and
the Fine Structure ov the Fiber," Trans. Faraday Soc. 31, 262 (1935).
Myrick, H. N. and Rykman, D. W., "Considerations in the Isolation and
Measurements of the Organic Refractories in Water," Jour. Amer.
Water Works Assn. 55, 6, 783 (1963).
Packham, R. F., "Studies of Organic Color in Natural Water," Proceed
Soc. Water Treatment Exam 13, 316 (1964).
Porter, M. C. and Michaels, A. S., "Membrane Ultrafiltration,'1 Chem.
Techno!. 1, 56 (1971).
82
-------
Povoledo, D. and Gerletti, M.s "Recent Studies on the Sedimentary Organic
Matter from Lake Maggiore, Northern Italy," Mitt. Intern. Verein
Limnol 14, 145 (1968).
Qasim, S. R. and Burchinal, J. C., "Leaching From Simulated Landfills,"
Jour. Water Pollut. Control Feder. 42, 371 (1970).
Rashid, M. A. and King, L. H., "Chemical Characteristics of Fractionated
Humic Acids Associated with Marine Sediments," Chem. Geol 7, 37
(1971).
Rebhun, M. and Manka, J., "Classification of Organics in Secondary
Effluents," Env. Sci. Techn. £, 7, 606 (1971)
Renkin, E. M., "Filtration, Diffusion and Molecular Sieving through
Porous Cellulose Membranes," J. Gen. Physio!. 38., 2, 225 (1954).
Reuter, J. H. and Perdue, G. M., "Chemical Characterization of Dissolved
Organic Matter and its Influence on the Chemistry of River Water,"
OWRC Report A-026-GA, Environ. Research Center, Georgia Inst. Technol.
p. 33 (1971).
Robbins, E. and Manro, A., "Experimental Study of the Independence of
Diffusion and Hydrodynamic Permeability Coefficients in Collodion
Membranes," J. Gen. Physiol. 43, 523 (1960).
Schnidler, J. E. and Alberts, J. J., "Analysis of Organic Inorganic
Associations of Four Georgia Reservoirs," Arch. Hydrobiol. 74
429 (1974). ~~
Seymour, W. B., "The Preparation of Cellophane Membranes of Graded
Permeability," J. Biol. Chem. 134. 701 (1940).
Schnitzer, M. and Skinner, S. I. M., "Gelfiltration of Fulvic Acid,
A Soil Humic Compound," in Isotopes and Radiation in Soil Organic
Matter Studies, IAEA, Vienna, p. 41 (1968).
Schnitzer, M. and Kahn, S. U., "Humic Substances in the Environment,"
Marcel 1 Dekker, Inc., New York (1972).
Shapiro, J., "Effect of Yellow Organic Acids on Iron and other Metals
in Water," Jour. Amer. Water Works Assn. 55_, 1062 (1964).
Shorland, F. B., "The Distribution of Fatty Acids in Plant Lipids,"
in Chemical Plant Taxonory, Ed. Swain T., Academic Press, New York
(1963).
Sieburth, J. M. and Jensen, A., "Studies on Algel Substances in the
Sea II. The Formation of Gelbstoff, Humic Substances, by Exudates
of Phaeophuta." J. Exp. Mar. Biol. Ecol. 3, 265 (1969).
83
-------
Stevenson, F. J. and Goh, K. M., "Infrared Spectra of Humic Acids and
Related Substances," Geochimica et Gsmochim. Acta 35, 471 (1971).
Swain, F. M., "Non Marine Organic Geochemistry," Cambridge University
Press, Cambridge, (1970).
Swift, R. S. and Posner, A. M., "Nitrogen, Phosphorus and Sulfur
Contents of Humic Acids Fractionated with Respect to Molecular
Weight," J. Soil Science 23, 50 (1972).
Tan, K. H. and McCreery, R. A., "The Infrared Identification of a
Humopoly Saccharide Ester in Soil Humic Acid," Soil Sci. Plant
Anal. 1, 75 (1970).
VanDyk, H., "Some Physico Chemical Aspects of the Investigation of
Humus," in "The Use of Isotopes in Soil Organic Matter Studies,"
Proceed. Symp FAO, IAEA, Brunswick, Volkenrode, F.R.G., Sept. 1963,
Pergamon Press, Oxford, p. 129 (1966).
Werschaw, R. L. and Burcar, P. J., "Physical Chemical Properties of
Naturally Occuring Polyelectrolytes - I Sodium Humate," Proceed.
3rd Ann. Amer. Water Resources Conference, Amer Water Resources
Assn., Urbana, p. 351 (1967).
Wilander, A., "A Study on the Fractionation of Organic Matter in Natural
Waters by Ultrafiltration Techniques," Schweiz. Zeitschr. Hydro!.
34, 2, 190 (1972).
Zajic, J. E. and Knettig, E., "Flocculants from Paraffinic Hydrocarbons,"
Developm. Industr. Microb. 12, 87 (1971).
84
-------
IV
THE SEQUENTIAL REMOVAL OF ORGANIC MATTER IN LEACHATE
DURING AEROBIC BIOLOGICAL DEGRADATION
CONCLUSIONS
The present study showed the sequential uptake of different classes of
organics during aerobic biological degradation of leachate obtained from
a recently installed fill. Using RO and UF membrane fractionation
followed by specific organic analysis it was found that the leachate
degradation occurs in four phases. The first phase was characterized
by a removal of high molecular weight humic carbohydrate-like organics
as measured by gelpermeation chromatography and specific organic analysis.
Its removal may be due to adsorption onto bacterial floes and by
bridging single particles thus causing a decrease in the filtered
turbidity. The second phase is characterized by the decrease of the
free volatile fatty acids. This is also reflected in a decrease of
the oxidation reduction potential, conductivity and dissolved oxygen.
During the second phase of the fatty acid utilization, an accumulation
of intermediates containing the carbonyl group and amino acids was
detected and are most likely derived from the intracellular amino acid
pool of the bacteria. After removal of these intermediates, representing
the third phase of substrate removal,high-molecular-weight carbohydrates
are excreted by the bacteria resulting in the further flocculation of
the bacteria. The fourth phase consists of the removal of these high-
molecular-weight humic-carbohydrate-like organics. The rate of organic
matter removal and the amount removed gradually decreases when the
removal progresses to phase four. The control mechanism that governs
the sequential utilization appears to be catabolite inhibition, i.e.,
inhibition by the catabolite on the functioning rather than on the
synthesis of the induciable enzymes. The present study also noted
increasing stability for fatty acids, amino acids, carbohydrates, humic
85
-------
and fulvic acids. Since a similar sequence is noted in the relative
composition of leachate samples from landfills of increasing age, only
leachate from recently installed fills, containing relatively large
amounts of fatty acids, will be amenable to biological treatment,
while such methods will be less effective when the leachate contains
large amounts of humic and fulvic acids.
86
-------
INTRODUCTION
Organic matter analysis of leachate samples collected from solid waste
landfills of different ages showed that the free volatile fatty acid
fraction gradually decreased while organics such a humic-carbohydrate-
like compounds and fulvie-like materials became more predominent with
increasing age of the fill (Chian and DeWalle, 1976). The organic matter
composition in leachate is expected to have a large impact on the treat-
ment efficiency of biological units treating such waste since some
classes of organics are more biodegradable than others. Several pilot-
and full-scale aerobic and anaerobic units treating leachate have been in
operation but consistently high efficiency was generally not obtained
due to changes in strength and variations in composition of the leachate
(Ham et_ aj_., 1976). It was therefore the purpose of this study to follow
the removal of specific classes of organics in leachate during aerobic
biological degradation. Such tests simulate the organic matter removal
in a plug-flow activated sludge unit treating leachate.
It is well known that microorganisms in a completely-mixed reactor are
more efficient in the removal of single substrates than a plug-flow
reactor (Herbert £t al_., 1956). The completely-mixed reactor is pre-
ferred because of the autocatalytic nature of the reaction and the
apparent zero order kinetics for single substrate utilization. However,
with a complex multi-substrate medium such as wastewater, the plug-
flow reactor, particularly with sludge recycle, is preferred (Eckenfelder
and O'Connor, 1961). Since uptake rates of individual substrates may
be markedly different in the presence of other substrates, the overall
reaction kinetics may no longer be zero order (Mateles and Chian, 1969).
Well-known examples of such substrate interaction is the occurrence of
diauxic growth in batch pure culture (Monod, 1949), sequential substrate
utilization in batch mixed cultures (Gaudy, 1962; Gaudy et al_., 1963a;
Gaudy et al_., 1963b; Gaudy et.al, 196*; Stumm-Zollinger, 1968) and
sequential substrate removal in continuous pure or mixed cultures
87
-------
(Komolrit and Gaudy, 1966; Ghosh and Pohland, 1972). All of the above
studies, however, were conducted using mixtures of pure carbon substrates,
In no instance was the substrate comprised of more than two or three
different carbon sources. There have been no reported studies on the
sequential uptake of substrates by microorganisms in naturally occurring
wastewaters known to contain multiple substrates (Painter, 1971). The
study reported herein demonstrates the sequential uptake by micro-
organisms of the different chemical groups present in leachate using
batch-mixed cultures. The leachate was generated from a pilot-scale
sanitary landfill. Results of this study show that the same control
mechanisms that govern the sequential uptake of substrates by bacterial
cells in well-defined synthetic media are also applicable to complex
wastewaters.
The efficiency of activated sludge systems depends on the ability of
microorganisms to degrade the different food sources present in the
influent. The carbon leaving the effluent of an activated sludge unit
is mostly refractory or inert material that the bacteria are not able
to degrade. It may consist of polymerized waste products of the
bacteria which are excreted into the solution. Other refractory
materials may consist of cell material or inert material from lyzed
cells. Some of these slowly degradable products have been referred to
as humus-like substances and their presence may represent 20 to 50
percent of the chemical oxygen demand (COD) of the effluent. These
humic substances have a molecular weight ranging from several hundreds
to tens of thousands (Schnitzer and Kahn, 1972). They are able to
chelate heavy metals and prevent their removal in treatment plants,
protect pathogenic organisms by coating and impair the color of the
effluent.
Only a few studies have analyzed the organic composition of the effluent
of biological treatment plants. Painter je_t a_l_. (1961) were able to
classify 25 percent of the organic matter while Bunch et. jal_. (1961)
identified 35 percent. The remaining part was thought to consist of
-------
humic material. Dialysis tests with secondary effluent showed that
21 to 49 percent of the soluble COD was retained by a cellulose dialysis
membrane indicating presence of high molecular compounds (Bunch et a!.,
1961).
Rebhun and Manka (1971) calculated that 39.3 to 45.2 percent of the COD
in activated sludge effluent consisted of humic substances; however,
their extraction technique may have overestimated the fulvic acid
fraction. Free volatile fatty acid and other compounds could be extracted
by the butanol used for extracting the fulvic fraction which, in turn,
may increase the weight of this layer.
A preliminary study by Bender et a^L (1970), using gel-permeation
chromatography, showed the presence of two molecular weight fractions
in activated sludge effluent. Both fractions had a complexing capacity
for heavy metals. A similar technique employing Sephadex columns was
used by Obiaga and Ganczarczyk (1972) and Minear and Christman (1966)
who also found a low molecular weight fraction in the effluent.
A recent development in the identification of organic matter in waste-
water is the combined use of membranes and Sephadex columns to analyze
the higher molecular weight substances. Goldsmith et^ aK (1971) first
characterized the molecular weight fractions of proteins in cheese
whey using Sephadex columns while employing membrane ultrafiltration
to concentrate the high-molecular-weight protein fraction. A similar
technique was used by A. D. Little, Inc. (1971). No use has yet been
made of reverse osmosis and ultrafiltration membranes to concentrate
wastewater or effluent for analyzing organic matter. In previous
studies in this laboratory,initial use was made of extraction techniques
to obtain the high molecular weight humic and fulvic acids. Because
of the unsatisfactory nature of these procedures, new methods were
adopted and extensive use was made of ultrafiltration (UF) and reverse
osmosis (RO) membranes in combination with Sephadex separation
(DeWalle and Chian, 1974).
89
-------
A relatively large number of studies concerning the nature of refractory
organics have been conducted in the area of soil biochemistry and recent
reports show humic acid to have a molecular weight ranging from 5,000 to
100,000 while the molecular weight for fulvic acid varied between 2,000
and 10,000 (Hurst and Surges, 1967). They also showed that about 50
percent of the total carbon of humic acid might be present in aromatic
structures, 20 percent in functional groups and 30 percent in aliphatic
structures. A large fraction of the oxygen is contained in the func-
tional groups especially the carboxyl group which can be as high as
8.9 meq/g for fulvic acid and 3.7 meq/g for humic acid.
The formation of humic and fulvic acids is not yet well understood.
Several investigators have proposed that fungi and bacteria are able to
convert notyet stabilized organic material into final humic and fulvic
matter. Waksman (1936) formulated that humic substances are derived
primarily from the interaction of plant lignin with proteins from the
soil microorganisms which generates a humic nucleus to which undecomposed
residues are attached. An opposing view has been presented by Kononova
(1966) who indicated that some soil fungi would convert a food source
to simple phenols, amino acids and carbohydrates which then condensed
extracellularly into a brown organic residue which was found to closely
resemble natural humic acid. The author presented evidence for
quinoid-amino acid melanoidin type humic compounds.
Similar studies were conducted by Martin and Haider (1969) and Hurst
and Wagner (1972) who grew fungi on a glucose-asparagine-salt solution
for several weeks and detected numerous phenolic compounds in the medium.
After the cessation of the excretion of phenolic compounds, a humic-like
polymer appeared in the medium accompanied by a decrease in the free
phenolic compounds. The humic acid was similar to soil humic acids in
its slow rate of microbial decomposition, number of carboxyl and phenolic
hydroxyl groups and molecular weight as determined by gel filtration.
The humic substances may have been formed intracellularly and released
into the solution after cell lysis.
90
-------
Besides the relatively stable humic and fulvic acids, it has been found
that certain high-molecular-weight carbohydrates, alone or in combina-
tion with humic material, are also resistant to microbial attack. The
most stable compounds were isolated from exocellular polysaccharides
(Finch
-------
MATERIALS AND METHODS
The batch aeration tests were conducted In an 80 £ vessel filled with
60 a of diluted leachate. Leachate,representative of a young landfill,
was obtained from a pilot-scale lysimeter filled with ground solid
waste and was extensively analyzed for its organic and inorganic con-
stituents. The bacteria culture was obtained from an activated sludge
unit receiving both municipal sewage and leachate at a concentration
of a half volume percent of the waste stream. Air was supplied to
the mixture at a rate of 2.8 £/min. The first aeration test (A) was
conducted with a leachate diluted to a concentration of 91 mg/1 TOC
and an added bacterial density of 20 mg/1 MLSS. In the second run
(B) both constituents were increased approximately five fold and in
the third test (C) the leachate concentration was double the amount
used in the second test. The leachate and the bacteria were diluted
with distilled deionized water supplied with N and P in similar
quantities as used in BOD dilution water (APHA, 1971).
At the end of each test run the solids were removed by centrifugation
and the centrifugate was concentrated with an Abcor (Cambridge, Mass.)
BEU-301 reverse osmosis unit using an AS-197 membrane.
The higher molecular weight organic matter was further concentrated
with an Amicon (Lexington, Mass.) ultrafilter cell using an UM-05
untrafiltration membrane with a nominal 500 MW cut-off. The retentate
was separated into its components using Sephadex G-75 (Pharmacia,
Piscataway, N.J.) columns which separate carbohydrate-like material
between a molecular weight of 1,000 and 50,000. The high-molecular-
weight fraction of the G-75 was further separated with a 6-150 column
and the low molecular weight fraction with a G-25 column (Gjessing and
Lee, 1967). The total organic carbon (TOC) analysis established the
distribution of the organic matter in the different molecular weight
92
-------
fractions. Proteins in the different molecular weight fractions were
measured with the ninhydrin method after extensive digestion (Stevenson
and Cheng, 1970). The more sentivive Lowry test could not be used,
since many constituents with a phenolic hydroxyl group, such as humic
and fulvic acids, interfere (1955). The carbohydrates were determined
with the anthrone test (Dreywood, 1946).
Functional groups in the different molecular weight fractions were also
measured with colorimetric tests. The carbonyl groups were determined
with 2,4-dinitrophenylhydrazine, which detects ketones but does not
measure carbonyl in carboxyl groups or aldehydes in sugars, as described
by Lappin and Clark (1951) and modified by Sanders and Schubert (1971).
The carboxyl groups were determined with the hydroxyl amine test as
reported by Montgomery et al_. (1962) and Lehmann and Wilhelm (1968).
The phenolic hydroxyl group was determined with the tannin test as
reported in Standard Methods (APHA, 1971). These colorimetric tests
were also conducted on filtered samples using 0.45 y Millipore (Bedford,
Mass.) membrane filter. Samples were removed at regular intervals during
the aeration test to determine changes in the monitored parameters over
time. Other parameters monitored during the aeration test were turbidity,
after filtration through a Whatmann 4 filter paper (W and R Ldt,
Madestone, G. B.), suspended solids, pH, dissolved oxygen (DO), oxida-
tion-reduction potential (ROP), conductivity, and total organic carbon
(TOC)•
93
-------
RESULTS AND DISCUSSION
An extensive analysis of the leachate used in this study showed that
51 percent of the TOC consisted of free volatile fatty acids with
butyric and acetic acid present in the highest concentrations followed
in concentration by caproic acid, iso-valeric acid, propionic acid,
valeric acid and isobutyric acid, respectively.
The high-molecular-weight fraction, as obtained in the ultrafilter
retentate, represented 27.2 percent of the TOC of the leachate. When
the retentate was further separated on a Sephadex G-75 column, two
large peaks, as measured with the TOC, became apparent (Figure 15).
The last eluted peak representing the low-molecular-weight material
(69 percent of the TOC of the UF retentate) evolved before the salt
peak. It was, therefore, considered to have a molecular weight larger
than 1000, possibly between 1000 and 5000. The high-molecular-weight
material representing 21.6 percent of the TOC of the ultrafilter
retentate evolved at an elution volume similar to that of the Sephadex
10 MW grade dextran, and is, therefore, larger than 50,000. The
application to G-25 and G-150 did not result in further significant
separation. A similar result was observed by Gjessing and Lee (1967).
Analysis showed that the carbohydrates, phenolic, hydroxyl and carbonyl
groups were equally distributed between the two fractions; however, the
carboxyls were only observed in the low molecular fraction. Infra-red
spectra of the two molecular-weight fractions showed close resemblance
to humic and fulvic acid spectra as reported by Goh and Stevenson
(1971) and Stevenson and Goh (1971). The high molecular weight frac-
tion resembled a humic acid with a strong peak near 1620 cm , which
corresponds to the carbonyl adsorption bond in quinones and diketones.
Also, it had some characteristics of a polysaccharide. The low molecular
weight fraction resembled a fulvic acid with a strong peak near 1720
cm" , which corroborated the result of the carboxyl test. Although
94
-------
10
01
SO -1320
280
240 E
U
200 <
o
ISO
K
120 9
so
4O
o*— o1—
so
100
Elution Volume, ml
Figure 15. Elution Pattern of the Ultrafiltration Retentate on Sephadex
G-75 before Aeration of the Leachate with Activated Sludge
-------
further characterization studies with mass spectrometry are still under-
way, the high-molecular-weight substance will be referred to as humic
acid (HA) and the low-molecular-weight fraction as fulvic acid (FA).
The unidentified fraction of the organic matter (21.8 percent of the
TOC) may have consisted of fulvic acid fragments with a molecular weight
less than 500 as detected in the UF permeate. This fraction gave a
positive phenolic hydroxyl test. Most of the inorganic anions were
present as chlorides and sulfates. Calcium, iron and sodium were the
major cations.
The result of the aeration tests showed that organics in the leachate,
as measured with the TOC, were removed in four steps; the first mainly
representing the removal of the humic, carbohydrate-like substances;
the second representing the uptake of the free volatile fatty acids; and
the third corresponding to the removal of intermediates released during the
second stage, while refractory organics are removed in the fourth stage.
In order to gain insight into the mechanism of the formation of refrac-
tory material, test A was terminated at the beginning of the second
phase during maximum fatty acid uptake. Test B was terminated at the
end of the third phase, when the excreted intermediates were exhausted
(53 hrs, see Figure 18). The third test (C) was aerated for 450 hours.
The changes of the different parameters during aeration of test B are
presented in Figure 16, 17 and 18. Figure 16 shows that the organic
matter was removed in sequence,with the first phase occurring between
zero and 10 hours; the second phase between 10 and 23 hours; and the
last phase between 23 and 48 hours. In the first phase, it shows a
decrease in TOC from 580 mg/1 to 480 mg/1. In this period the suspended
solids show their first increase (Figure 17). An increase was also
observed in the ORP, pH and conductivity, but no large oxygen demand
was exerted. The carboxyl test measuring the free volatile fatty acids
96
-------
)r— 8.0r— 8,IOi
r- BOO
zro — 7,0
260
250
— 6.0
— 7,90
— T.70
— 5,0 — 7,50
0.
or
o
£40
q
d
230
— 3.0
— E.O
210
2OO
— 1.0
; 7.30
— 7.10
— 6,90
— 6.TO
U. 0
6.50
,4 i IO
Time, hours
Figure 16, Change of pH, DO, ORP, TOC and Conductivity
During Aeration of the Wastewater
-------
30r— 300.0
00
20
—>J 20QP
I
«
O.
10
01— 0«— 0
— 150
10
20 30
Time, hours
40
Figure 17. Change of Suspended Solids, Turbidity and Filtered
Turbidity During Aeration of the Wastewater
50
-------
UD
US
4O
», 3°
>x
r
•T
Q
i
i M
u
»
S
^
o
— ° 200
100
•^ 0
— Z4
— 21
-i.18
•>v
?
-•15
O
tt
«
-°I2
•>
«*
o
-0 9
>»
|
W
~ D
-
f\
^ O
— 8.O
- ™i
-•,6-0
r
- -o 5.0
< 1
U
'E.
c
-^4.0
^T
~«
0
•
^
-°- 2.0
— 1.0
^M rt
fc-« g
1 i i I i
w
1 /-Proteins / ~
x /\ /
^^-^O-^v . i Corbohydrates •*./
v IJ^^X* /
>>-^v 7 ?~t^^\ /
— ^^ S> 1 X\ '
X / / ^ V
"•^fi" * /\l / VVTI ^ "
' ""••— «-5^. / \ * \ ^ ^x»"
/ *^/ y\' \ Ovx*^'^ ~
it / v » ^^c1*^
A\ / / *^^\ Nv
- ^_v— • ^ **»• 1 \ 'v*^ /- Phenolic OH
^" \' \ \ s ,
% 1 • \ v>— — O
1 \ X
l« >• \
l\ \ \
' » X \/~ Corbonyl
/ » » D
— • » % v
_ / v corto \.\
j,.— o — o--0"" "° \/" " °Xy "^^"^
— °* % \
II 1 1 *
) 10 20 30 40 50
Time, hours
300
*<
^v.
200 g
_
_C
^J
3
•>
_l
CO
<
«0
100 -5
k_
Q.
0
Figure 18. Change of Proteins, Carbonhydrates, Phenolic Hydroxyl-, Carbonyl-and
Carboxyl Substances During Aeration of the Wastewater
-------
present in the diluted leachate also remained constant (Figure 18),
while the increase in conductivity indicated dissociation of these
acids as the pH increases.
The analysis of the humic material, obtained after termination of test
A with subsequent RO and UF concentration of the effluent, showed that
the amount of substances greater than 500 MW had dropped from 27.2
percent of the initial TOC to 7.10 percent of the TOC of the solution
at end of test A. Of the different components detected in the 500 MW
UF retentate, a large reduction was observed in the carbohydrates. The
ratio of the carbohydrates to TOC (as mg dextrose/mg TOC) in the humic
acid and fulvic acid peak decreased from 0.47 to 0.35 and from 0.23 to
0.16 respectively (Figure 15 and 19). These ratios were obtained using
the carbohydrate value divided by the maximum TOC value of the fraction.
A similar decrease in ratios was observed for the phenolic hydroxyl and
carbonyl in the humic acid fraction. The increase in carboxyl/TOC
ratio may indicate the higher resistance of this group to bacterial
attack or a conversion from hydroxyl and carbonyl groups to the carboxyl
group. During this first stage the fulvic acid is removed to a larger
extent than the humic acid fraction with the result that the ratio of
humic to fulvic increased from 0.42 to 0.77. This preferential removal
may be due to the higher initial concentrations of the fulvic compared
to the humic acid. The rapid increase in suspended solids and a
decrease of the filtered turbidity during the first 5 hour period
(Figure 17) may indicate that the humic-carbohydrate-!ike materials
have flocculating properties and are able to bridge single
particles. These data may therefore indicate that the humic carbohydrate-
like materials are removed by adsorption onto the bacteria floes. The
presence of a high iron concentration (75 mg/1) and its oxidation during
aeration may also have enhanced the floe formation.
It should be noted that the filtered turbidity (Figure 17) and the
carbohydrates (Figure 18) show a small increase at the end of the first,
and the beginning of the second phase. It could well be that at this
100
-------
-*»*
2 I
Elutlon Volumt, ml
Elution Volum*, mi
Figure 19. Elution Pattern of the Reverse Osmosis and Ultra-filtration
Retentate on Sephadex G-75 During Aeration of the Leachate
(a) Corresponding with Maximum Substrate Uptake
(b) After Uptake of the Readily Available Carbon Source
101
-------
point the degradable part of the adsorbed humic carbohydrate-like
materials is consumed by the cells and that some of these degraded
carbohydrate segments are subsequently released into the solution
(Figure 18). Redispersion of the flocculated particles due to decrease
of such bridging materials is evidenced by the increase in the filtered
turbidity (Figure 17). Utilization of carbohydrates in the first
phase is also evidenced by the inhibition of the fatty acid uptake
(Figure 17). This was shown by Chian and Mateles (1968) to occur with
butyric acid in the presence of glucose. The uptake rate of mono-
saccharides by the cells occurs more rapidly than that of fatty acids.
The lower rate of substrate utilization observed in the first phase as
compared to the second phase fatty acid utilization is, therefore, most
satisfactorily explained by the rate limiting step of hydrolysis. It
is necessary to hydrolyze the high-molecular-weight carbohydrates in
the media to make them available to the cells (Banerji crt ah, 1968).
The second phase of the TOC removal is characterized by the decrease of
the free volatile fatty acids as measured with the carboxyl test
(Figure 18). Shortly before the beginning of the acid uptake the
oxidation reduction potential (Figure 16) shows a decrease, indicating
the decrease of the oxidation state of the oxidative co-enzymes. The
decrease in conductivity parallels the actual removal of the fatty
acids from the solution. Only after the acid is taken up in the
Tricarboxylic Acid Cycle (TCA cycle) is an oxygen demand exerted which
explains the decrease of the DO in the solution lagging behind the
actual uptake.
During the second phase of the fatty acid utilization, an accumulation
of an intermediate containing the carbonyl group was first detected in
the medium after about 19 hours of aeration (Figure 17). A similar
accumulation was observed for proteins or amino acids at a much higher
concentration. The excretion of intermediates during the active meta-
bolism of monosaccharides was realized by Gaudy e£ afh(1964) Mateles
and Chian, 0969). The intermediates were later found to contain acetic
102
-------
acid. These authors observed that as much as 30-40% of the carbon
substrates, particularly when fed in the form of mixtures, was excreted
as intermediates. However, only 10% of the carbon substrates lost as
intermediates was identified as acetic acid when fed with glucose
alone, and 15% when fed with glucose and lactose. A complete material
balance of the excreted intermediaries has never been attained due to
the lack of information on the nature of other excreted compounds
(Mateles and Chian, 1969).
While intracellular amino acid pool accumulation has been reported in
batch cultures of Saccharomyces cerevisiae (Brown, 1969; Moat et al.,
1969) Bacillus megaterium (Nelson and Kornberg, 1970), and Escherichia
coli (Brada, 1970), only one study reported the accumulation of extra-
cellular amino acids. Clark e_taj_. (1972) observed such an increase
in cultures of Bacillus licheniformis when the cells changed from
vegetative to sporulation metabolism. Most of the authors observed that
the total intracellular amino acid pool increased during exponential
growth and decreased rapidly thereafter. This agrees well with this
observation which shows that the extracellular proteins (or amino acids)
decrease rapidly toward the end of the active metabolism of fatty acids.
It was calculated that the maximum amount of proteins excreted accounted
for 35-50% of the total weight of the fatty acids initially present 1n
the medium, which is appreciable (Figures 18).
The authors cited above also reported that only a few amino acids made
up the majority of the pool. Between 70-80% of the amino acids in the
pool consisted of glutamate followed by 10-20% alanine and in some cases
small amounts of aspartate. A review of amino acid biosynthesis shows
that each of these amino acids is derived from an a-keto acid (a-keto-
glutaric acid, pyruvic acid and oxaloacetic acid, respectively). Only
glutamic acid is formed from direct amidation of a-ketogluraric acid
with ammonia. Aspartic acid and alanine are formed by transamination of
oxaloacetic acid and pyruvate with glutamic acid. The abundance of
103
-------
glutamic acid found in the anrino acid pool as reported by the above
authors is therefore well explicable.
Based on these observations, it was expected that the excreted amino
acids and keto containing compounds found in this study were predominatly
glutamate and glutamine, respectively. Excretion of substantial amounts
of alanine during active metabolism of fatty acids is unlikely, since
pyruvate is not an intermediate in the catabolic pathways of fatty acids.
Excretion of the other a-keto acids, such as o-keto glutaric acid and
oxaloacetic acid, is also unlikely, since the intermediaries in the
catabolic pathway are not expected to accumulate in substantial amounts
in the cells. It is worthwhile to note that both acetic acid and gluta-
mate represent intermediates being excreted at the branching point of
the metabolic pathways. The former is excreted prior to acetyl-CoA
entering into the citric acid (Krebs) cycle for energy yielding steps
of oxidative phosphorylation for ATP production. Both of these excre-
tions appear to be regulated by the delicate enzyme control systems within
cells to prevent from unnecessary waste of energy.
As shown in Figure 16, the rate of TOC removal started decreasing at
about the twenty-ninth hour of aeration. At this time, the dissolved
oxygen (DO) level increased rapidly as a result of exhaustion of fatty
acids. This is indicated by the diminishing of the carboxyl group in
the medium (Figure 18). This transition into a slower rate of substrate
uptake in terms of TOC removal represents the third phase of substrate
utilization. In this phase, continuing removal of excreted proteins
(amino acids), carbonyl (keto)-containing and phenolic-OH containing
substances were observed (Figure 18).
During the removal of these intermediates in the third phase, the carbo-
hydrate concentration starts to increase again. This also corresponds
with a decrease in the filtered turbidity. The increase in the carbo-
hydrate concentration reflects an increase in the amount of refractory
material. After termination of test B, the solids were removed by
104
-------
centrifugation. The supernatant was concentrated with the RO and UF
module. Results of TOC analysis showed that the amount of organic matter
retained by the 500 MW ultrafiltration membrane had increased from 7.1
percent to 21.6 percent. This represents an actual increase in concen-
tration of the refractory material in the solution from an estimated
20.7 mg/1 during the middle of the second phase to 39.9 mg/1 at the end
of the third phase. The most notable increase was shown by the carbo-
hydrate in the larger than 50,000 MW fraction (Figure 19). As a result
of this increase, the carbohydrate to TOC ratio of the humic acid
fraction increased from 0.35 to 0.67.
The fulvic acid fraction was characterized by a strong increase in
carbonyl-bearing structures as shown at an elution volume of 90 ml
(Figure 19). This appears to be at an expense of a reduction of the
carboxyl groups, since the ratio of carboxyl to TOC (as mg acetic acid/
rug TOC) decreased from 1.32 at the end of test A to 0.158 at the termi-
nation of test B. It is also seen from Figure 19 a and b that the humic
to fulvic acid ratio decreased from 0.77 (Figure 19a) to 0.55 (Figure 19b).
This may have been the result of the preferential adsorption and bridging
of the humic carbohydrate-like polymers onto the bacteria during the
flocculation stage. The fulvic fraction showed a decrease in the
carbohydrate/TOC ratio from 0.16 in the middle of the second phase
(test A) to 0.06 at the end of the third phase (test B). This lower
carbohydrate content may have prevented its adsorption onto the sludge
particles by not providing sufficient polymers for bridging to the
bacterial mass.
This observation may also explain the settling behavior observed in a
continuous control and test unit employed during further investigations.
The test unit receiving various volume percentages of leachate and
municipal sewage generally experienced poorer settling characteristics
than the control unit treating only municipal sewage. The carbohydrate
to TOC ratio of the humic acid fraction in the effluent of the control
105
-------
unit was 0.68 while it was only 0.37 in the effluent of test unit. This
apparent shortage of humic-carbohydrate-like polymers and the resulting
lack of bridging may have caused the observed impairment of the floc-
culation and the higher TOC and color in the effluent of the latter
unit. A large part of this TOC, which mostly consists of humic sub-
stances, may not be able to adsorb onto the sludge cells and be removed
along with the sludge. The higher humic acid to fulvic acid ratio
in the effluent of the test unit as compared to the control unit would
substantiate this conclusion. This observation has an important impact
on activated sludge sewage treatment plants. These units are operated
with respect to optimum carbon removal, which may differ from the condi-
tions required for optimum floe formation and removal of humic sub-
stances. If more stringent requirements will apply to the presence
of residual organics in the effluent, the treatment plant or its opera-
tion may have to be modified in order to ensure optimum removal of these
substances.
A subsequent experiment (test C) was conducted with a feed containing
approximately twice the amount of the leachate as used in the previous
experiment and tested through a much longer period of aeration, 450
hours (Figure 20, 21). A decrease of the humic-carbohydrate-like
materials produced by the cells along with the others can be seen
clearly at a time starting from 160 hours of aeration (Figure 21).
This represents the fourth phase of substrate utilization which is
carried out at even a slower rate as is shown by the lower rate of
TOC removal (Figure 20).
These four distinct phases of substrate utilization not only demon-
strate clearly the occurrence of sequential substrate removal in a
natural wastewater, but also explain well the shape of the inverse
sigmoidal curve commonly observed for the BOD removal in the wastewater.
Figure 20 depicts a typical inverse sigmoidal curve for the TOC removal
from the wastewater under this study.
106
-------
15
9.0
8.0
7.0
6.0
.t 10
O 4.0
Q
3.0
2.0
1.0
1— 0
8.0
7.0
6.0
5.0 «—
900
100
zoo
3OO
4OO
Time, hours
— 2.55 «10*
— 2.50x10*
E
o
•v>
O
"I
4.
1
— 2.45 x I04 O
500
2.40x10*
Figure 20. Change of Filtered Turbidity, DO, pH, TOC and Conductivity During
a Longer Period of Aeration with a Higher Strength Leachate
-------
o
00
60
50
9
40
o.
o
O
U
« 2O
o
o
U
10
I5OO
- x
•§500
O
U
80
70
8 60
0
X
— 5°
O 40
O
w
£ 30
O
O 20
10
O1
900
Time, hours
Figure 21. Change of Carbohydrate, Proteins (Amino acids) and Phenolic Hydroxyl-,
Carbonyl- and Carboxyl Containing Substances During a Longer
Period of Aeration with a Higher Strength Leachate
-------
The overall control mechanism that governs the sequential utilization
of substrates appears to follow the mechanisms described by Gaudy et al.
(1964), i.e. catabolite inhibition, wherein the effect of the catabolites
is on the functioning, rather than on the synthesis of the inducible
enzymes. This is evidenced by the smooth and the continuing curve
observed for TOC removal and the absence of the lag period required
for the utilization of the less preferred substrate after the exhaustion
of the more preferred one.
The present study shows that the relative composition of the substrate
i.e., the leachate, has a large impact on the rate of organic matter
removal and on the total amount removed. Increased stability is noted
for fatty acids,ami no acids, carbohydrates, humic and fulvic acids.
As pointed out by Chian and DeWalle (1976) a similar sequence is noted
in the relative composition of leachate samples collected from landfills
of increasing age. This in turn indicates that leachate of young fills,
containing substantial amounts of fatty acids, is amenable to biological
degradation while such treatment methods will be less effective when
the leachate contains large amounts of fulvic and humic acids. Such
leachate is therefore preferably treated by physical-chemical methods.
109
-------
REFERENCES
American Public Health Association, Standard Methods for the Examination
of Water and Wastewater, Editors M. J. Taras et al., American Public
Health Association, Washington, D.C. (1971).
Banerji, S. K., Ewing, B. B., Engelbrecht, R. S. and Speece, R. E.,
"Mechanism of Starch Removal in the Activated Sludge Process,"
J. Water Pollut. Control Fed. 40, 16 (1968).
Barker, S. A. and Somers, P. J., "Bacterial and Fungal Polysaccharides,"
in The Carbohydrates, Vol. 11B, Editors W. Pigman, ejb a]_., Academic
Press Inc., New York (1970).
Bender, M. E., et al_. , "On the Significance of Metal Complexing Agents
in Secondary Sewage Effluents," Environmental Science and Technology,
4, 520 (1970). -
Broda, P., "Der Aminosauerpool von Escherichia coil bei Aminosaure-
aushungerung," Monatsch. Chem. 102, 811 (1970).
Brown, C. M., "The Ami no Acid Pools of Yeast," J. Gen. Microbiol. 58,
...
Bunch, R. L. £t al,., "Organic Materials in Secondary Effluents,"
Journal Mater Pollution Control Federation, 33., 122 (1961).
Busch, P. L. and Stumm, W., "Chemical Interaction in the Aggregation of
Bacteria," Environmental Science and Technology. ,2, 49 (1968).
Chian, S. K. and Mateles, R. I., "Growth of Mixed Cultures on Mixed
Substrates - I. Continuous Culture," Appl . Microbiol. 16, 1337
(1968). --
Chian, E. S. K. and DeWalle, F. B., "Sanitary Landfill Leachates and
their Characterizations," Jour. Environ. Engr. Div. ASCE, 102,
EE2, 411 (1976). - --
Clark, V. L. et al. , "Changes in Free Amino Acid Production and Intra-
cellular A"mino Acid Pools of Bacillus lichem'-formis as a Function
of Culture Age and Growth Medium." J. Bacteriol. 112. 715 (1972).
DeWalle, F. B. and Chian, E. S. K., "Removal of Organic Matter by
Activated Carbon Columns," Environ. Engr. Div. ASCE 100, 1089 (1974).
Dreywood, R., "Qualitative Tests for Carbohydrate Materials," Ind. Eng.
Chem. Anal. Ed. 18. 499 (1946).
110
-------
Dunican, L. K. and Seeley, H. W., "Extracellular Polysaccharide
Synthesis by Members of the Genus Lactobacillus, Conditions for
Formation and Accumulation," Journal General Microbiology. 40,
297 (1965). ~~
Eckenfelder, W. W. and O'Connor, D. J., "Biological Waste Treatment,"
Perganon Press, New York (1961).
Finch, P. et_ al_., "The Biochemistry of Soil Polysaccharides," in Soil
Biochemistry. Vol. 2, Editors A. D. McLaren and J. Skujins,
Marcel Dekker Inc., New York (1972).
Friedman, B. A. e_t a]_., "Structure of Exocellular Polymers and Their
Relationship to Bacterial Flocculation," Journal Bacteriology, 98,
1328 (1969). ~~
Gaudy, A. F., Jr., "Studies on Induction and Repression in Activated
Sludge Systems," Appl. Microbiol. ]_0, 264 (1962).
Gaudy, A. P., Jr., Gaudy, E. T. and Komolrit, "Multicomponent Substrate
Utilization by Natural Populations and a Pure Culture of Escherichia
Coli," Appl. Microbiol. 1]_, 157 (1963).
Gaudy, A. F., Jr., Komolrit, K. and Gaudy, E. T., "Sequential Substrate
Removal in Response to Qualitative Shock Loading of Activated
Sludge Systems," Appl. Microbiol. 12, 280 (1964).
Ghosh, S. and Pohland, F. G., "Kinetics of Assimilation of Multiple
Substrate in Dispersed Growth Systems," Hater Res., Peragamon Press,
6., 99 (1972).
Gjessing, E. and Lee, G. F., "Fractionation of Organic Matter in Natural
Waters on Sephadex Columns," Environmental Science and Technology,
1, 631 (1967).
Goh, K. M. and Stevenson, F. J., "Comparison of Infrared Spectra of
Synthetic and Natural Humic and Fulvic Acids," Soil Science. 112,
392 (1971).
Goldsmith, R. L. etal_., "Industrial Ultrafiltration," in "Membrane
Processes in Industry and Biomedicine," Proc. 160th National Meeting
American Chemical Society, September 1970, Editor M. Bier, Plenum
Press, New York (1971).
Ham, R. K. e_t al_., "Pilot Plant Treatment of Sanitary Landfill Leachate,"
Final Report Grant R-801814, Department of Civil Engineering,
University of Wisconsin, Madison, p. 156 (1976).
Herbert, D. e_t aJL, "The Continuous Culture of Bacteria, a Theoretical
and Experimental Study," J. Gen. Microb. ]±, 601 (1956).
Ill
-------
Hurst, H. M. and Burges, N. A., "Lignin and Humic Acids," in Soil
Biochemistry. Vol. 1, Editors A. D. McLaren and G. H. Peterson,
Marcel Dekker Inc., New York (1967).
Hurst, H. M. and Wagner, G. H., "Decomposition of 4C Labeled Cell Wall
and Cytoplasmatic Fraction from Hyaline and Melanic Fungi," Soil
Science Society of American Proceedings, 36, 597 (1972).
Komolrit, K. and Gaudy, A. F., Jr., "Biochemical Response of Continuous
Flow Activated Sludge Processes to Qualitative Shock Loadings,"
J. Water Pollut. Control Fed. 38, 85 (1966).
Konanova, M. M., Soil Organic Matter, 2nd ed., Pergamon Press Inc.,
New York (1966]"!
Lappin, G. R. and Clark, L. C., "Colorimetric Method for Determination
of Traces of Carbonyl Compounds," Analytical Chemistry, 23, 541
(1951).
Lehmann, G. and Wilhelm, G., "Colorimetric Method for the Determination
of Esters of Phthalic Acid in the Presence of Esters of Other
Carboxyl Acids," Zeitschrift fur Analytische Chemie, 238. 415 (1968).
Little, A. D., Inc., "Characterization and Separation of Secondary
Effluent Compounds by Molecular Weight," Water Pollution Control
Series, Project No. 16020, FEN 03-71, Environmental Protection
Agency, Washington, D.C. (1971).
Martin, J. P. and Haider, K., "Phenolic Polumers of Stachybotrys atra,
Stachybotrys chartarum and Epiroccum nigrum in Relation to Humic
Acid Formation," Soil Science, 107, 260 (1969).
Mateles, R. I., and Chian, S. K., "Kinetics of Substrate Uptake in Pure
and Mixed Culture," Environ. Sci. and Techn., 3,, 569 (1969).
McKinney, R. E., "Staining Bacterial Polysaccharides," Journal Bacteriology,
66_, 453 (1953).
Minear, R. A. and Christman, R. F., "A Method for Evaluating the
Biodegration of Lignin Sulfonates," Paper presented at the 33rd
Annual Meeting of the Pacific Northwest Water Pollution Control
Association, Portland, Oregon (1966).
Moat, A. G., "Alternation in the Amino Acid Content of Yeast During
Growth Under Various Nutritional Conditions," J. Bacteriol. 98,
573 (1969).
Montgomery, H. A. C., "The Rapid Colorimetric Determination of Organic
Acids and Their Salts in Sewage Sludge Liquor," Analyst, 87., 949
(1962).
112
-------
Monod, J., "The Growth of Bacterial Cultures," Ann. Rev. Microbiol.
3., 371 (1949).
Nelson, D. L. and Kornberg, A., "Biochemical Studies of Bacterial
Sporulation and Germination. XVIII. Free Amino Acids in Spores,"
J. Biol. Chem. 245. 1128 (1970).
Obiaga, T. I. and Ganczarczyk, J., "Biological Removal of Lignin from
Kraft Mill Effluent. II. Changes in Molecular Size Distribution,"
University of Toronto, Department of Civil Engineering, Publication
72-08 (1972).
Painter, H. A. ejt al_., "Composition of Sewage and Sewage Effluents,"
Institute of Sewage Purification. 1961. 302.
Painter, H. A., "Chemical, Physical and Biological Characteristics of
Water and Waste Effluents," Water and Water Pollution Handbook,
Vol. 1, Ciaccio, L. L. (Ed.), Marcel Dekker Inc., New York, N.Y.
(1971).
Pavoni, J. L. e_t al_., "Bacterial Exocullular Polymers and Biological
Flocculation," Journal Water Pollution Control Federation, 44,
415 (1972).
Peter, G. and Wuhrmann, D., "Contribution to the Problem of Biofloccula-
tion in the Activated Sludge Process," Proceedings 5th International
Water Pollution Research Conference, July 1970, Pergamon Press
Inc., New York (1971)
Rebhun, M. and Manka, J., "Classification of Organics in Secondary
Effluents," Environmental Science and Technology, _5, 606 (1971).
Sanders, E. B. and Schubert, J., "Spectrophotometric Analysis of Carbonyl
Compounds in the Presence of Carbohydrates without Prior Separation,"
Analytical Chemistry, 43, 59 (1971).
Schnitzer, M. and Khan, S. U., "Humic Substances in the Environment,"
Marcel Dekker, Inc., New York (1972).
Smit, C. J. B. ejt al_., "Determination of Tannins and Related Polyphenols
in Foods; Comparison of the Loewenthal and the Folin-Denis Methods,"
Analytical Chemistry 27. 1159 (1955).
Stevenson, F. J. and Cheng, C. N., "Amino Acids in Sediments: Recovery
by Acid Hydrolysis and Quantitative Estimation by a Colorimetric
Procedure," Geochemica et Cosmochimica Acta. 34_, 77 (1970).
Stevenson, F. J. and Goh, K. M., "Infrared Spectra of Humic Acids and
Related Substances," Geochimica et Cosmochimica Acta, 35, 471
(1971).
113
-------
Stumm-Zollinger, E., "Substrate Utilization in Hetergeneous Bacterial
Communities," J. Water Pollut. Control Fed. 4£, R213 (1968).
Waksman, S. A., Humus, Wiliams and Wilkins, Baltimore, MD (1936).
114
-------
METALS AND THEIR INTERACTION WITH
ORGANIC MATTER IN LEACHATE
CONCLUSIONS
It was noted in the present study that heavy metals are present in
relatively high concentrations in landfill leachate. The determina-
tion of heavy metals by flame atomic absorption spectroscopy is sub-
ject to relatively large matrix interferences in heavily polluted
leachate samples; this effect is less noticeable in more dilute
samples. The matrix interference causes a reduction in the absorbance
in the flame at the analyte wavelength resulting in a lower apparent
metal concentration. It was observed that Cu was subjected to the
largest matrix interferences followed by Cr, Ni, Pb, Cd and Zn respec-
tively. The interfering effect can be eliminated by using the standard
addition technique in which a sample is spiked with a known amount of
the metal. The obtained recovery factor is than used to correct the
measurement of the unspiked sample. It was noted that the interference
in the Cr determination was both related to the absolute salt content
and to the relative salt content (with respect to the heavy metal
concentration), indicating that simple dilution of the polluted leachate
sample will not totally eliminate the matrix interference. In addition
it was noted that also Cl resulted in lower Cr recoveries as measured
with the standard addition technique. The percentage depression in the
Ni determination was best correlated with the absolute salt content and
the concentration of Ca and Fe. The depression in the Cu determination
may have been the result of the presence of phosphate compounds, while
the Pb analyses showed that both a high relative salt content and the
presence of sulfates decreased the recovery of the metal. It was further
noted that La had to be added to the leachate samples as a releasing agent
during the Ca determination.
115
-------
Analysis of the heavy metal content in 12 leachate samples collected
from landfills of different ages and subjected to varying stabilization
and leaching periods, showed that the heavy metal content decreased
more rapidly with age than other inorganic components. This is most
1ikely the result of adsorption and precipitation reactions which in turn
are enhanced by an oxidation reduction potential gradually increasing
with increasing age of the landfill.
Extensive fractionation of the different molecular weight fractions
present in leachate showed that the majority of the metals permeated
the 500 MW UF membrane indicating that chelation of the metals by
refractory organics is not a major mechanism in metal attenuation
processes. A deviation of this general pattern was noted for iron
which was associated for a considerable part with the larger than
100,000 MW UF retentate. Measurement of the ion exchange capacity and
carboxyl group density indicated that only a small amount of the iron
in this fraction can be chelated and that most of it is apparently
present as ferric hydroxides. Membrane ultrafiltration of additional
leachate samples showed that the percentage Ca and Fe retained with
the 10,000 MW UF membrane increased with increasing magnitude of the
high molecular weight organic fraction. The increase of the calcium
content in the UF retentate may be due to chelation by the functional
groups of the organic matter, while the increase in iron content is
probably due to interaction of hydroxide complexes and the organics by
adsorptive processes. The copper retention showed considerable
competitive effects indicating that other metals such as Fe, Ca and
Zn are better chelated. Treatability studies of leachate with aerobic
and anaerobic biological systems showed that substantial amounts of the
the metals are removed by these processes, again indicating that
chelation is not a major factor. Removal in aerobic systems is likely
due to formation of metal hydroxides and carbonates, while removal in
anaerobic systems is the result of formation of insoluble metal sulfide
and carbonate. Thus despite the relative high concentrations of heavy
116
-------
metals detected in leachate, no physical chemical pretreatment step
is necessary to remove these metals prior to biological treatment.
117
-------
INTRODUCTION
Concentration of Heavy Metals in Leachate
Leachates generated by infiltrating rainwater in landfills constructed
from recently deposited solid waste will contain high concentrations of
organic and inorganic pollutants. Merz (1954) measured maximum concen-
trations of 2570 mg/1 calcium (Ca), 410 mg/1 magnesium (Mg), and 305
mg/1 iron (Fe) in leachate from a lysimeter filled with solid waste. The
maximum sodium (Na) and potassium (K) concentrations of 1805 and 1860
mg/1, respectively, were approximately equal. The major anions were
chloride (Cl) and sulfate ($04) with a maximum concentration of 2350
mg/1 and 710 mg/1.
Qasim and Burchinal (1970) also found that Na, K, Ca, Mg, and Fe were
the major cations. They showed that increasing the height of the fill
reduced the relative amount of organic and inorganic pollutants leached
from the solid waste. The longer distance that the organic pollutants
had to travel through the refuse enabled the bacteria present in the fill
to partially remove the degradable organics, and the percentage of
organics with a biochemical oxygen demand (BOD) extracted per dry weight
of solid waste decreased from 5.2% to 2.6% when the height of the fill
was increased from 0.76 m to 3.05 m. A similar pattern was noted for
most inorganic pollutants. Iron, however, showed a deviation from this,
since the amount extracted per dry weight of solid waste increased by
30% when the height of the fill increased from 0.76 m to 3.05 m. As this
pattern is the inverse of that observed for the degradable organics the
relative increase in iron content may indicate that a substantial portion
is chelated by refractory organics, that are formed by bacteria after the
removal of more biodegradable organics.
118
-------
More recent leachate studies have included several other metals in their
measurements. Fungaroli and Steiner (1971) showed that zinc (Zn) is
generally present at a concentration one-tenth of that of iron in leachate
obtained from a lysimeter. They also showed that the iron concentration
was proportional to the volume of the generated leachate indicating that
the infiltration flow has to be sufficiently high to leach the iron.
Since the higher leachate production corresponded to a lower pH, caused
by a higher free volatile fatty acid content, the variation in the iron
content may also be the result of its pH dependent solubility. The Zn
increase lagged behind that of the iron content, indicating that the
former element was only leached after the solid waste degradation had
proceeded substantially. The nickel (Ni) concentration generally ranged
from 0.2 to 0.3 mg/1 and its variation with time was generally the
opposite of the iron content.
Lower heavy metal concentrations were measured in leachate from actual
landfills as compared to leachate generated from lysimeters filled with
solid waste. A survey by Hughes &t at. (1971) of several full-scale
landfills in Illinois showed that the iron and zinc content were generally
less than 100 mg/1 and 5 mg/1, respectively. Several heavy metals such
as copper (Cu) and cadmium (Cd) were usually below their detectable limits.
The County of Sonoma study (1973) measured the heavy metals content in
leachate from a control pilot landfill and a pilot landfill in which the
collected lea-hate was reapplied onto the fill, resulting in a higher
moisture content of the solid waste. A calculation based on the reported
data showed that the higher moisture content did not have a major effect
on the composition of the inorganic salts. Using the chloride concentra-
tion as the base for comparison it was computed that the ratio of Ca/Cl
was 1.17 for the control fill and 1.10 for the recirculation fill. The
relative zinc content was slightly lower for the control fill since the
Zn/Cl ratios were 0.014 and 0.022, respectively. The relative iron con-
tent, •however, was approximately 10 times higher for the control fill
based on the Fe/Cl ratios of 1.72 and 0.18, respectively. The relatively
119
-------
higher iron content may be the result of the higher chelation of the
organics present in the control fill since the relative organic matter
content based on the chemical oxygen demand (COD)/chloride (Cl) ratio was
86 for the control fill and 29 for the recirculation fill.
Interferences in the Heavy Metal Determination
Many techniques are presently available to determine heavy metals in
aquatic systems and three basic methods can be distinguished: electronic
transition methods (molecular absorption spectrometry, molecular fluor-
escence spectrometry, atomic absorption spectroscopy, atomic fluorescence
spectrometry, atomic emission spectroscopy, X-ray emission spectroscopy),
neutron activation methods,spark source mass spectrometry, gas chromato-
graphy and electrochemical techniques (Minear and Murray, 1973). Many
studies in the area of water pollution use the flame atomic absorption
spectroscopy and this method was also employed in the present study. The
method involves the atomization of the sample by flame and the absorption
of radiation at the specific wavelength of the metal. The radiation source
used is a hollow cathode lamp consisting of a cathode made of the same
element as the metal under analysis (Willard at <*£., 1965).
Leachate samples will present numerous difficulties in flame atomic absorp-
tion analysis for heavy metals such as Cr, Ni, Fe, Pb, Cu, and Zn, due to
the presence of complex matrices. Similar interferences have been encount-
ered in the analysis of heavy metals in sea water which has an average
salt concentration of 35,000 mg/1 (Home, 1969). In comparison, leachate
samples can have a fixed solids content of up to 23,000 mg/1 (Chian and
DeWalle, 1976).
Interference effects in flame atomic absorption can be classified into
three main groups: 1) spectral, 2) non-atomic absorption, and 3) chemical
(Dean, 1960).
120
-------
Spectral or line interference can often be avoided by judicious choice of
an analyte resonance line and by using a narrow band-width as close as
possible to the width of the selected line emitted by the hollow cathode
tube. Kahn and Manning (1972) showed that absorption band-widths in a
flame can be as small as 0.03 A. The best resolution obtained in commer-
o
cially available units, however, is only 0.3 A, while most units have
monochromatic spectral band-widths of 2 A or more.
Non-atomic absorption can result from the flame itself, molecular absorp-
tion from the elements in the samples or light scattering caused by
unburned particles in the optical path. These effects are generally most
pronounced at shorter wave lengths in the far UV portion of the spectrum.
This background absorption effect can be eliminated by using a two-line
method, i.e., subtraction of the absorption at a non absorbing line close
to the analyte line, and by using a deuterium background correction or
employing a hydrogen hollow cathode lamp.
The last type of interference effects, which are element specific, are
expected to cause complications in leachate analyses. Specific chemical
interferences have long been known in atomic absorption spectroscopy
(Table 14). Alkali and alkaline earth elements such as Na and K are
subject to a wide range of interferences, primarily because they ionize
in the flame and are thus unavailable for atomic absorption of their
resonance line energy emitted by the hollow cathode lamp (Johnson and
Schrank, 1964). This situation can be corrected, however, by the addi-
tion of a more easily ionized element such as Cs at a concentration level
considerably higher than that of the analyte atom. Ca and Mg present a
more serious problem since they are interfered with by a number of common
cations and anions, such as SO^ , PO^, CO^, HCO^, Na+, K , and Al
(Ramakrishna, 1968). An inorganic releasing agent, lanthanum, for
instance, has been found effective in suppressing interferences and
prevent formation of oxides in the flame (Adams, 1966).
121
-------
Table 14
Chemical Interferences Observed in the Metal Determination
by Flame Atomic Absorption Spectroscopy (Hermann, 1963)
Interferent
Al
As
B
Ba
Be
Br
co?
Ca
Cd
Cl
C104
Co
Cr
Cs
Cu
Fe
104
K
HC03
Li
Mg
Mn
Mo
NH4
NO 3
Na
NaCl
Ni
P
P04
Pb
Rb
S
Si
Sn
Sr
S04
Te
Ti
TiF6
W
Zn
Zr
Organics
Citrate
* d =
** e =
**4
Ca Cd
d*
d
e**
d
e
dde*
eed
d
d
d
e
d
ed
d
d
d
dde
d
d
e
d
d
d
d
d
eed
depression
enhancement
Analyte
r* **i
Cr Cu Fe K
d e
d
d
e dde
d ed
e
d
e
d e
d e
e
edd
d
d
d de
d
d
e
e e
d
e e
d
dee
d
e
eed eed
d
(r*
Mg
d
de
d
e
e
e
e
e
d
e
dde
d
eed
**** *****
Na Ni Pb Zn
d
d
eed
d
d
eed
d
d
d
d e
dde e
d
d
ed e
ed e
eed
dde
d
dde e
d
d
e
d
d
d
e
e
d
d
d
eed eed
*** dde = depression at low and intermediate concentration and
enhancement at high concentrations
****no information yet available
122
-------
Detailed specific interferences for heavy metals such as Fe, Cu, Cr, Ni,
Sr, and Al are given by Herrmann (1963) and Yanagisawa
-------
these authors concluded that the micro nutrients were largely associated
with low molecular weight dialyzable organic constituents. Schnitzer and
Kahn (1972) reported that the major chelation of the heavy metals involved
the combined presence of phenolic and carboxylic groups and that minor
chelation took place through the carboxyl group alone. Broadbent and
Bradford (1952) also showed that carboxyl and phenolic groups, attached to
heterocyclic compounds, are important functional groups for the Cu binding
capacity. Broadbent (1951) further concluded that while the carboxyl
groups complexed both Cu and Ca, other functional groups reacted selectively
with Cu to the exclusion of Ca.
Shapiro (1964) showed that fulvic acids present in fresh water are associated
with iron. Christman (1970) presented evidence that the majority of the
iron was complexed and associated with all size fractions of the organic
matter. A strong chelation agent was able to remove some of the iron but
it was least efficient in removing the iron from the fraction with a
molecular weight larger than 50,000. Considerable amounts of heavy metals
can also be expected to adsorb to the ferric hydroxide colloids (Barsdate
and Matson, 1966).
In the case of leachate, the organic complexing is most likely governed by
the Fe, Ca, and Mg concentrations which are competing with the heavy metals
for the ligand sites. If complexing occurs, it can be an important
mechanism in the transport of heavy metals in groundwater. Its attenuation
would then be govered less by the solubility of their hydroxides but by
the ability of the soil materials to remove the cheland.
The above survey showed that only a limited number of studies have investi-
gated the presence of heavy metals in landfill leachate. None of the
studies investigated whether the atomic absorption spectroscopy used to
measure the heavy metal content was interfered by matrix effects of the
high salt concentration in the leachate. Based on the results listed in
Table 14 such is indeed expected. Some chelation of the heavy metals is
124
-------
also expected, the magnitude of which would depend upon the concentration
of Ca and Mg competing for the absorption sites. It was therefore the
purpose of this study to determine the actual concentration of heavy
metals in leachate and to measure the magnitude of the matrix effect of
the salts. The second part of the study measured the interaction of the
heavy metals with the organic matter and determined the molecular weight
fractions that showed the strongest chelation with the heavy metals.
125
-------
MATERIALS AND METHODS
The leachate samples were collected under strictly anaerobic conditions
followed by ultracentrifugation and 0.45 y membrane filtration to remove
any solids present in the sample. Prior to flame atomic absorption
measurements, the organic matter in the sample was removed using a rigorous
digestion procedure, which also served to eliminate any metal hydroxydes.
The procedures used for pretreating the leachate sample are those
recommended by the US-EPA (1971) and 50 ma of sample was gently digested
with two portions of 3 nu concentrated HNOo whereafter 3 m£ of 1:1 HC1
was added. Silicates that had formed on the bottom of the beaker were
removed by filtration. The atomic absorption unit used for the present
study was a Beckman Model 485 (Fillerton, Cal.).
It was found that the high salt content in leachate greatly limited direct
aspiration of the sample. A high solids burner was tested, but it did not
have the desired sensivitity. Chelation with ammonium pyrrolidine dithio-
carbamate (APDC) and extraction into methylisobutyl ketone (MIBK) was
evaluated next. However, the high salt content caused the formation of
an intractable emulsion which likewise would not aspirate.
The final method used in this study was to dilute the leachate and to use
the standard addition technique (Christian, 1969), in which the sample is
spiked with increasing amounts of the analyte. This method can only be
used when the absorbance is linear over the concentration range involved,
while the relative magnitude of the chemical interference should also be
constant within that range. The non atomic absorption will have to be
measured at each standard addition since otherwise the calculated concen-
tration would be greater than it actually is.
The determination of the alkali and alkaline earth metals did not give any
complications. It was found satisfactory to add 10,000 mg/£ La to the
126
-------
leachate samples to suppress the ionization of Ca and Mg. To each sample
analyzed for Na and K, 1000 mg/£ of Cs was added to suppress the ionization
of the analyte atom. The elements Cd, Cr, Cu, Ni, Pb and Zn were determined
using the standard addition method.
Leachate samples were collected from 12 landfills of various ages located
in different regions of the United States and representing different
climatic environments (Chian and DeWalle, 1976). The strength of the
organic pollutants varied from 81 mg/i to 71,080 mg/£ COD, while the
inorganics varied between 812 mg/£ to 22,895 mg/£ fixed solids. As a
result of these various strengths the magnitude of the depression also
varied considerably. In order to further evaluate some of the interfer-
ences and the applicability of the standard addition method to leachate
analysis, four synthetic samples were devised containing 1 mg/£ of each
of the heavy metals. The composition of the major ions in these synthetic
samples as listed in Table 15 were chosen to resemble actual concentrations
that were found in analyzing the various leachates (Chian and DeWalle, 1976)
The first two samples contained a low total dissolved solids (TDS) content
while the last two samples had a salt content approximately five times
higher. The first and the third sample had a relatively high sulfate
content while the second and the fourth has a high chloride concentration.
The next phase of the study consisted of fractionating the organic matter
in a leachate sample obtained from a lysimeter installed at the University
of Illinois. The organics were separated according to their molecular
weight using membrane ultrafiltration techniques and gel permeation
chromatography (DeWalle and Chian, 1974). The ultrafiltration membranes
used were UM05, UM10 and XM100A having nominal molecular weight cutoffs
of 500, 10,000 and 100,000 (Amicon, Lexington, Mass.), respectively. The
different ultrafiltration membrane fractions were characterized by total
organic carbon (TOC), chemical oxygen demand (COD), carbohydrates
(anthrone test), proteins (ninhydrin test after 24 hr acid digestion),
carboxyl groups (hydroxyl amine test), carbonyl groups (2,4 dinitrophenyl
127
-------
Table 15
Synthetic Leachate Samples Used to Evaluate Matrix Interferences
in the Heavy Metal Determination Containing 1 mg/1 of Cr, Ni,
Cu, Pb and Cd.(a)
Elements
Cl
Na
so4
K
Ca (as CaCO-)
Mg (as Mg[C2H302]-4H20)
Fe3+
IDS
1
(mg/1)
50
32
1000
670
100
30
50
2246
Sample
2
(mg/1)
1000
640
50
34
100
30
50
2218
Number
3
(mg/1 )
50
32
1000
670
2000
500
500
10482
4
(mg/1 )
1000
640
50
34
2000
500
500
10454
(a) All samples were diluted 1:2 when the standard additions were added,
resulting in concentrations half of those listed above
128
-------
hydrazine test), and aromatic hydroxyl groups (Folin-Denis test). After
the heavy metal distribution in each of the molecular weight fractions
was established, additional leachate samples previously analyzed for
their total heavy metal content, were fractionated with a 10,000 MW
ultrafiltration membrane whereafter the heavy metals in each molecular
weight fraction were determined.
129
-------
RESULTS AND DISCUSSION
The results of the standard addition technique in a 1:8 diluted leachate
sample are shown in Figure 1. The leachate sample was obtained from the
experimental US-EPA pilot landfill at the Boone County Field Site, Ky and
referred to as the Cincinnati sample (SHWRL, 1973). The sample had an
initial COD of 45,750 mg/a and a fixed solids content (FS) of 13,603 mg/£.
The depression caused by the constituents in the leachate ranged from 40
percent for Cr to 5 percent for Cd (Figure 22). The depression was calcu-
lated as the tangent of the function between the amount recovered and
amount added, divided by the tangent of the 100% recovery function. When,
therefore, a heavy metal determination is made without application of the
standard addition the measurement can be as much as 40 percent lower than
the actual value. When the determination is made on a leachate sample
without prior dilution, the matrix of the sample may even result in a
higher depression and further underestimate the actual value.
In addition to the Cincinnati sample (CINC) two samples collected from a
milled uncovered (MUNC) and an unmilled covered (UMC) solid waste lysimeter
in Madison, Wisconsin (Reinhardt and Ham, 1973) also experienced consider-
able depression. Two samples obtained from a pilot solid waste lysimeter
(UI 2/1 and UI 9/21) at the University of Illinois (Chian and DeWalle,
1976) were also evaluated. The percent depression of these five samples
was conveniently plotted against the wavelength at which each element was
measured (Figure 2a). The results of the five samples are qualitatively
similar as the magnitude of the depression decreased from Cu to Cr, Ni,
Pb, Cd and Zn respectively. Several factors may have caused the depres-
sion in the recovery of metal analysis.
Comparison of the percent depression with the salt concentration in the
diluted leachate samples indicated that some relation existed. For
example, the highest depressions for Cr were observed with the samples
130
-------
0.3
Q2-
I 1
100% Recovery
Line
0.162 Amount Added, mg/JJ
Figure 22. Application of the Standard Addition Method to Determine the Actual
Heavy Metal Concentration in the Cincinnati Field Site Leachate Sample
131
-------
100
& 50
c
I
— CINC 1,700mg/Jf FS(l:8dil.)
-MUNC 4,579 mg/f FS(l:5dil.)
• UMC 695 mg/{ FS (i:5dil.)
JI2/1 780 mg/J FS (l:20dil.)
ZnCd
19/21 10,015 mg/ff FS(II2)
J
Pb3000 Cu NiCr Fe 4000
Wavelength (A)
100
0
a>
o.
a>
O
50
0
-Solution 3, 5241 mg/i FS
-Solution 4, 5227mg/J FS
ZnCd
Solution 2,1109mg/x FS
Solution 1,1123 mg/* FS
Pb3000 Cu NiCr Fe 4000
Wavelength (A)
Figure 23. Percentage Depression in the Heavy Metal Determinations
for Leachate Samples (a) and Synthetic Salt Solutions (b).
132
-------
having salt contents of 1700 mg/fc, 4570 mg/x. and 10,015 mg/n while the
lowest depression occurred in the samples with a salt content of 695
and 780 mg/z. The relationship between depression and salt content,
however, is certainly not linear for Cr. The Cincinnati sample, for
example, with a salt content of 1700 mg/a resulted in a depression of
40 percent, while the University of Illinois sample (III 9/21) having a
salt content of 10,015 mg/a in the 1:2 dilution, gave a depression of
only 23 percent. Comparison of the depression of the heavy metal analysis
of leachate samples with those of the synthetic salt solutions (Figure 2b)
confirmed the effect of the concentration of inorganic salts on the Cr
determination. The highest depression was indeed found for solutions 3
and 4 with a TDS of 5,900 mg/£ and 5,821 mg/£, respectively. In addition
to the absolute salt content, the relative salt concentration also seems
to be an important parameter. For example, the ratio of fixed solids/Cr
in the CINC, MUNC, UMC and UI (2/1) samples decreased from 68,015 to
45,790, 19,306 and 866, respectively, corresponding with a percentage
depression which decreased from 40% to 25%, 14% and 11%, respectively.
The results with the synthetic leachate samples also show that the
chloride concentrations have a pronounced effect on the depression of
the absorbance since the 1000 mg/s, Cl" in solution 2 resulted in a
depression of 22 percent, while the 50 mg/a in solution 1 caused a depres-
sion of only 8 percent; both solutions have a almost identical salt
concentration. Similar conclusions with regard to the chloride effect
can be drawn from the chemical analysis of the tested leachate samples
as listed in Table 16. This, however, is contrary to the results pre-
sented in Table 14, which showed that chloride actually enhances the Cr
absorbance.
Although most of the chemical interferences in the Ni determination tend to
enhance the absorbance (Table 14), the constituents present in leachate
cause an apparent depression (Figure 23). However, the depression for
Ni in leachate samples and the synthetic salt samples is generally lower
than that observed for Cr. When the percentage depression was correlated
with different factors that could cause such depression, the best correlation
133
-------
Table 16
Chemical Characteristics of Leachate Samples in Which
the Heavy Metals were Tested by the Standard Addition Method
Source
Parameter
(mg/1 )
COD
TOC
BOC
Acetic acid
Propionic acid
Isobutyric acid
Butyric acid
Isovaleric acid
Valeric acid
Caproic acid
Bicarbonate alkalinity
PH (-)
Cond (ymho/cm)
ORP (mv)
Turb (2Tu)
Suspended Solids (SS)
Fixed Suspended Solids (FSS)
Total Solids (TS)
Fixed Solids (FS
Org-N
NH4-N
N03-N
N02-N
Tot-P
Ortho-P
$04
Cl
University
of Illinois
(UI 2/1)
49,300
17,060
24,700
4,370
1,050
570
5,620
1,220
960
2,400
668
5.63
13,700
-60
75
139
92.5
33,989
15,586
544.7
392.6
0.5
BDL**
21.5
6.5
1,110
1,480
Madison
(MUNC)
71,680
27,700
57,000
6,690
3,180
360
9,270
770
1,260
1,920
459
5,97
16,800
-132
80
202
167
55,348
22,348
945
1,028
10.25
0.04
98
29
1,558
2,467
Madison Cincinnati
(UMC) (CIN)
16,580
5,906
9,960
510
255
31
480
40
265
565
284
5.
5,420
-220
270
192
110
7,930
3,475
78.
347.
4.
0.
85
85
77
474
45,750
13,840
22,000
1,340
660
340
1,700
520
460
1,090
175
59 5.25
9,450
*
71
8.9
7.5
32,145
13,603
5 31
4 247.7
25 9.8
04 0.19
31.6
28
909
2,096
As a result of the standard addition the samples were diluted 1:20, 1:5,
1:5, and 1:8, respectively.
134
-------
was obtained with the fixed solids content in the diluted leachate sample.
The result of the synthetic sample analysis indicated that neither sulfates
nor chlorides had any effect. High correlations were noted for Ca and Fe
possibly indicating that multivalent cations are the main cause of the
observed depression.
The results for Cu show no interference in the synthetic salt solution
while the depression was relatively large in all leachate samples except
for the two UI samples. A main difference between the leachate samples
and the synthetic salt solutions is that only the former contain nitrogenous
substances, such as ammonia, nitrate and nitrite, and phosphates. Of these,the
relative phosphate concentrations showed the best correlation with the
percent depression. The parameter that showed the best correlation
with the percent depression, however, was the relative zinc concentration,
indicating that the heavy metals can have an effect on each other.
Evaluation of the interference for Pb indicates that a high salt content
increases the depression. At a high relative salt content, however, the
depression tended to decrease again. The results with the simulated
leachate samples indicates that a high sulfate content also increases the
depression. The determination of Zn and Cd is generally not strongly
interfered with by matrices present in leachate samples and the measure-
ment can be made on a sample without using the standard addition technique.
The metals present in relatively high concentrations in leachate such as
Ca, Mg, and Fe were next evaluated. When Fe was added at a concentration
of 50 mq/a and 500 mg/£ to the four synthetic salt solutions which was
subsequently diluted 1:10 and 1:100 to result in a final concentration of
5 mg/£ no depression was observed; instead a slight enhancement in all
four samples was noted. The synthetic salt solutions were also used to
determine the recovery of the Ca ion. Solutions 1 and 2 were diluted
1:20 and solutions 3 and 4 diluted 1:400 to result in a final Ca concen-
tration of 5 mg/Ji. The measurement of the Ca showed that less than 100
percent was recovered. The depressions in the four salt solutions were
135
-------
12, 11, 12 and 15 percent, respectively, The addition of 10,000 mg/£ La,
however, was found to eliminate the interfering effect. The effect of
omitting the La addition on the Mg determination was less pronounced and
the Mg values in the absence of the La were even slightly higher than in
the samples to which La was added. The Mg determinations also showed that
the salt solution with the highest salt content showed the highest enhance-
ment effect.
Analysis of metals in less polluted leachate samples collected from addi-
tional landfills described by Chian and DeWalle (1976) was not interfered
with by the matrix of the sample since the relative or absolute salt
content was sufficiently low at the dilutions used. The results of the
flame atomic absorption measurements for each of the 12 leachate samples
are listed in Table 17. The values for the different landfills show a
gradual decrease corresponding with an increasing age of the landfill.
Such a decrease is expected since continuous leaching will remove the
metals from the refuse, and since the leachate experiences a gradual
pH increase with the increasing age of the fill that decreases the solu-
bility of each of the metals. Although the trend of decreasing concen-
trations with increasing age is quite noticeable, considerable variation
does occur at a specific landfill age. To eliminate these variations the
relative concentration of the metal was calculated with respect to the
other inorganic salts as measured as fixed solids (FS). These ratios
were then plotted against the age of the landfill (Figure 24). The
relative Ca, Fe and Zn concentrations show a gradual decrease. As the
Ca is the most soluble of the three metals shown, its decrease with age
is the least rapid. Its decrease is probably the result of its percipi-
tation as carbonate salt. The decrease of the relative Fe and Zn concen-
trations follows a pattern inverse to that of the oxidation reduction
potential. Their relative concentrations do not seem affected by the
organic matter decrease. Their more pronounced decrease after a five
year period is more abrupt than noted for the organic matter decrease,
which may indicate that most of the metals are not effected and thus not
chelated by the organics in leachate.
136
-------
Table 17
Metal Concentrations in Leachate Samples Collected
from Landfills Having Different Ages
Source
Parameter
U. of U of
University Wisconsin Wisconsin
of Illinois Madison Madison
(UI 2/1) (MUNC) (UMC)
Sonoma, Ca
Sonoma, Ca Recircu- SHWRL
Control lation Cincinnati
(SONCON) (SONREC) OH (CIN)
Ca 3750 3900 572 1030
Mg 650 1140 220 515
Fe 2200 1046 91 301
Na 1360 1580 330 240
K 1140 2300 900 74
Zn 104 370 13 22
Cd 0.1 0.375 0.05 0.
Pb 1.25 1.65 0.80 1,
Cu 0.48 0.75 0.65 1.
Ni 13 2.6 0.23 0,
Cr 18 0.5 0.18 0,
Age of fill (yr) 0.25 0.33 0.33 1.64
02
1
0
5
11
1220
750
540
675
300
15
0.008
0.48
0.1
0.6
0.15
1.64
2325
530
735
750
860
52
0.36
3.12
0.71
1.3
0.2
2.3
Ca
Mg
Fe
Na
K
Zn
Cd
Pb
Cu
Ni
Cr
Age of fill (yr)
GIT GIT Atlanta
Atlanta Recircula- College Kennet Winnetka, Winnetka,
Control tion Park, PA Squire, PA Illinois Illinois
(ATLCON) (ATLREC) (PACP-67) (PHKS-12) (LW-17) (LW-5B)
520
52.5
810
75
620
74
17
1.16
0.48
0.4
0.03
2.25
133
35
0.5
44
147
0.05
0.03
0.10
0.03
0.05
0.03
2.25
215
51.5
560
37.5
35
6. 5
0.01
0.21
0.15
0.09
0.015
5.3
330
85
330
113
70
2.5
0.005
0.2
0.21
0.32
0.095
6.75
254
81
1.5
91
39
0.16
0.003
0.48
0.10
0.08
0.04
4.83
76
186
4
370
450
1.02
0.003
0.54
0.32
0.12
0.04
16
137
-------
COD
FS
Co
FS
Fe
FS
Zn
FS
(a)
5 \0
Age 01 Landfill, years
15
Figure 24. Decrease of COD, Ca, Fe and Zn Relative to the Salt Content
of Leachate Samples with Increasing Age of Landfills
138
-------
To study the interaction of heavy metals with the organic matter in the
leachate sample an extensive fractionation was made of the different
molecular weight fractions followed by their organic analysis. The first
fractionation conducted on the leachate sample collected from the University
of Illinois lysimeter using a 500 MW UF membrane showed that the majority
of the organics had relative low molecular weights since 72.8 percent of
the TOC was recovered in the permeate. The majority of the metals were
also recovered in the permeate and only 2.4 percent of the Mg and 2.5
percent of the Ca were measured in the 500 MW UF retentate, while a
slightly larger percentage (5.5 percent) of Zn was retained. The majority
of the iron on the other hand was recovered in the UF retentate and only
41 percent was detected in the permeate.
The 500 MW UF retentate was then diluted to its original concentration and
subsequently passed through a 10,000 MW UF membrane which now retained
21.8 percent of the original TOC. When the diluted 10,000 MW UF retentate
was subsequently passed through a 100,000 MW UF membrane the percentage
TOC decreased further to 11.1 percent of the original TOC. The results
of the organic analysis for different classes of compounds and functional
groups of the 4 membrane fractions showed a decreasing carbohydrate content
with decreasing molecular weight. The proteins followed a trend similar
to that of the carbohydrates in the fractions larger than 500 MW. A
surprisingly high content was also found in the 500 MW UF permeate
indicating the presence of amino acids. The carboxyl group analysis
showed that the carboxyl group density increases with decreasing molecular
weight. The highest percentage was present in the 500 MW UF permeate
due to the presence of the free volatile fatty acids. Gas chromatographic
analysis showed free volatile fatty acids, predominantly acetic and
butyric acid. The carbon of these acids comprised 78 percent of the TOC
of the 500 MW UF permeate or 49 percent of the initial TOC of the leachate
sample.
Each of the fractions that were analyzed for functional groups were also
subjected to heavy metal analysis. The results in Figure 25 show that
139
-------
Q>
o
'S
fc.
"c
o>
o
o
o
0.15
O
o Magnesium (Mg)
Calcium (Ca)
D Iron (Fe)
x Zinc (Zn)
C C
a> o
$_•!-
0> -tJ
«t- (O
M- S-
•r- -l->
Oi—
•r-
O) M-
-C (O
+-> i-
u c
10 tO
111 i_
•— >»
CO O3
+J
^"S
tr o *o
c -a
o o
3 (O
03
•r- (/)
S. C
•M O
O U
tO
IOOJOOOMW
100,000 MW
en o
-------
the relative concentration of metals (expressed as mMole/g organic carbon
generally decreases with increasing molecular weight, a trend parallel to
that observed for the carboxyl group. Calcium, magnesium and iron were
the metals retained in highest quantity followed by zinc. These results
tend to confirm the model studies of Sturm and Bilinsky (1973) who
calculated that high concentrations of Ca and Mg will prevent the chela-
tion of substantial amounts of heavy metals. The results obtained for
iron in the present study forms an exception in this respect, since it
has a low relative concentration in the 500 MW UF permeate and high
relative concentrations in the 100,000 MW retentate. Similarly Shapiro
(1964) and Christman (1970) reported relatively high amounts of iron
associated with the high molecular weight organics as obtained by gel
permeation chromatography.
The sum of the different metals in the 500-10,000 MW fraction represents
1.30 mMole/g C and assuming that the iron is also in the divalent form,
this equals 2.6 meq/g organic carbon. The carboxyl group density as
determined by titration and confirmed with the hydroxyl amine test is
equal to 10.1 meq/g C, indicating that approximately 26 percent of the
carboxyl groups have divalent metals as counter ions. As a result of this
divalency, these metals may well be strongly retained or chelated by the
organic matter. The other carboxyl groups may not have been dissociated
at the pH of 6.5, i.e. the pH at which the fractionation was performed.
An undetermined amount of K and Na ions will also be present as counter
ions in each of the molecular weight fractions. However these monovalent
cations are less preferred than divalent cations and are therefore not
expected to be present in large quantities. The metals present in the
10,000-100,000 MW UF fraction are equal to 1.75 meq/g organic carbon which
also represents approximately one fourth of the carboxyl content of 6.6
meq/g C. Due to the high iron content the metals would represent 377
percent of the carboxyl group content in the 100,000 MW UF retentate, thus
indicating that the majority of the iron is present as colloids, possibly
ferric hydroxides. The formation of such colloids was shown to occur
141
-------
directly after collection of the leachate sample as a result of a notic-
able increase of the oxidation reduction potential despite the strict
anaerobic collection procedures (Chian and DeWalle, 1976). The Zn shows
an interesting pattern in that its relative concentration in the larger
than 500 MW fractions tends to be the opposite of that of the iron
indicating that some competition for chelation sites may occur.
The last phase of the study consisted of fractionating 8 of the 12 col-
lected leachate samples according to their molecular weight using a
10,000 MW UF membrane to determine the amount of heavy metals present in
the UF retentate. The 10,000 MW UF membrane was selected since it was
best suited to obtain the high molecular weight fraction. The percentage
of the organic matter retained was then related to the percentage of each
of metals retained. The results in Figure 26 show that both the percen-
tage Ca and Fe retained with the 10,000 MW UF membrane increases with
increasing magnitude of the high molecular weight organic fraction. The
calcium content showed a larger than proportional increase with increasing
magnitude of the high molecular weight fractions, while the retained iron
increased proportional to the high molecular weight organic fraction. The
observed increase in the amount of calcium retained with increasing size
of the high molecular weight fraction may be due to chelation of the
calcium by the carboxyl - and aromatic hydroxyl groups of the high
molecular weight organic fraction. Its more than proportional increase
may indicate that the amount of functional groups associated with this
organic fraction increases with increasing concentration of this fraction.
As most of the iron is present as ferric hydroxide colloids, the increase in
amount retained with increasing size of the high molecular weight fraction
indicates that both parameters show a significant interaction. The high
molecular weight organics could be adsorbed onto the colloidal particles,
resulting in a larger complex which is therefore better retained with the
10,000 MW UF membrane. When both the Ca and Fe were plotted in the form
of a Freundlich isotherm with the metal content in the 10,000 MW retentate
expressed as meq/g organic carbon and the concentration in the permeate as
142
-------
u_
5
o
8
20
7
*/
sS
5g 40
£>
E
fe
SS2°
c
0)
o
/
10
20
30
Percentage Of TOC Retained With The IO.OOOMWUF
Membrane
Figure 26. Relation Between Magnitude of the High Molecular Weight Organic
Fraction and the Amount of Ca and Fe Retained with
the 10,000 MW UF Membrane
143
-------
equilibrium concentration, the curve for the Ca ion showed a slight
decrease while the Fe showed an increase in amount retained with increasing
equilibrium concentration. Since the magnitude of the total iron concen-
tration parallels that of the calcium concentration, this may indicate
that the iron is able to partially displace the Ca in the high molecular
weight fraction at increasing iron concentration.
The results for Cu (Figure 27) show that the percentage retained gradually
decreases with increasing magnitude of the high molecular weight fraction.
Only one sample having a substantial concentration of high molecular
organics deviated from this trend as it also showed a large amount of
copper retained. The Zn showed a similar result as the Cu for small sizes
of the high molecular weight fraction. At high percentages of the TOC
retained with the 10,000 MW UF membrane the percentage Zn retained showed
an increase.When the data were plotted as a Freundlich isotherm, the
amount of Cu adsorbed per gram organic carbon showed a significant decrease
at increasing equilibrium concentrations. The amount of Zn adsorbed per
gram organic carbon, however, experienced a gradual increase at higher
equilibrium concentrations. The increase was lower than that observed
for iron possibly indicating iron is complexed preferentially to the zinc,
which in turn is preferred to Ca and Cu at the concentrations studied.
Since ion exchange mechanisms may also exert a considerable effect, a
change in absolute concentration of each of the metals may change the
preference of chelation as noted above. The increase of amount adsorbed
at higher equilibrium concentrations for Zn and especially Fe may explain
their generally high concentrations in leachate as compared to heavy
metals of similar solubility such as Cd, Cu, Pb, Cr, and Ni.
The above results clearly indicate that the majority of the metals are not
associated with organics having a molecular weight larger than 500.
Interaction of the metals with the low molecular weight organics may well
occur and has to be further evaluated. Such interaction, however, may
be of an indirect nature since a high concentration of free volatile fatty
144
-------
*
o
H
s S
•I I
20
I0
8.
o
0
£
£
%
10
20
30
Percentage Of TOC Retained With The 10,000
MWUF Membrane
Figure 27. Relation Between the Magnitude of the High Molecular Weight
Organic Fraction and the Amount of Ca and Fe Retained
with the 10,000 MW UF Membrane
145
-------
acids as found in high concentrations in the 500 MW UF permeate, will
result in a lower pH of the solution and therefore a higher solubility
of the metals.
Treatability studies with leachate in the second phase of the research
project showed that aerobic biological systems generally resulted in a
higher removal of metals as compared with anaerobic systems. While both
anaerobic and aerobic treatment resulted in large reductions of the iron
content, the reduction of Cr and Mg was larger in the aerobic treatment
unit. No heavy metal toxicity has been observed in the aerobic units
treating leachate but one such instance was observed in the anaerobic
filter. The toxicity caused a gradual deterioration of the effluent COD
of the unit and a gradual increase in Fe, Cu and Zn in the effluent.
When the Fe and Cu concentration exceeded 3 mg/£ and 1 mg/ji, respectively,
the methane bacteria were inhibited thus causing a sharp increase in free
volatile fatty acids in the effluent and a decrease in gas production of
the unit. Addition of sulfide to precipitate the heavy metals eliminated
the toxicity. Sulfate in the leachate is reduced in the anaerobic filter
to sulfide which then precipitates the heavy metals. The presence of
0.45 filterable ferric hydroxide colloids as found in the present study
is apparently not presenting the removal of iron in the anaerobic filter;
in effect it may enhance it. The possible chelation of zinc is apparently
not effecting its removal as a zinc sulfide precipitate.
146
-------
REFERENCES
Adams, P. G., e£ oJL., "Critical Factors in the Determination of the
Alkaline Earth Elements in Glass by Atomic Absorption Spectrometry,"
Analytical Chemistry, 38., No. 4, 38 (1966).
Barsdate, R. J. and Matson, W. R., "Trace Metals in Arctic and Subarctic
Lakes with Reference to the Organic Complexes of metals Proceedings
International Symposium Radioecological Concentration Processes,
Stockholm, Pergamon Press, Oxford, (1966).
Broadbent, F. E. and Bradford, G. R., "Cation Exchange Groupings in the
Soil Organic Fraction," Soil Science, 74_, 447 (1952).
Broadbent, F. E., "Soil Organic Matter-Metal Complexes: II Cation
Exchange Chromatography of Copper and Calcium Complexes," Soil Science,
84, 127 (1951).
Chian, E. S. K. and DeWalle, F., "Sanitary Landfill Leachates and their
Treatment," Jour. Environ. Eng. Div., ASCE, 102. 411 (1976).
Chian, E. S. K. and DeWalle, F. B., "Compilation of Methodology for
Measuring Pollution Parameters of Landfill Leachate," Ecological
Research Series, EPA-600/3-75-011, U.S. Environmental Protection
Agency, Cincinnati, p. 164 (1976).
Christian, G. D., "Medicine, Trace Elements and Atomic Absorption
Spectroscopy," Analytical Chemistry, Vol. 41_, 25, (1969).
Christman, R. F., "Chemical Structures of Color Producing Organic Sub-
Stances in Water," Symposium on Organic Matter in Natural Waters,
University of Alaska, D. W. Hood, ed., Institute of Marine Science
Occasional Publication No. 1 (1970).
County of Sonoma, "Sonoma County Stabilization Study: Second Annual
Report," Department of Public Works, Santa Rosa, California (1973).
DeWalle, F. B. and Chian, E. S. K., "Kinetics of Formation of Humic Sub-
Stances in Activated Sludge Systems and their Effect on Flocculation,"
Biotechnology and Bioengineering, 16, 739 (1974).
Environmental Protection Agency, "Methods for Chemical Analysis of Water
and Wastes," National Environmental Research Center, Analytical
Quality Control Laboratory, Cincinnati, Ohio (1971).
147
-------
Fungaroli, A. A. and Steiner, R. L., "Laboratory Study of the Behavior
of a Sanitary Landfill," Journal Water Pollution Control Federation,
43, 252 (1971).
Geering, H. R. and Hodgson, J. F., "Micronutrient Cation Complexes in
Soil Solutions: III. Characterization of Soil Solution Ligands and
their Complexes with Zn2+ and Cn2+," Proceedings Soil Science Society
America. Vol. 33., 54 (1969).
Gilbert, P. T., "Absorption Flame Photometry," Analytical Chemistry, 34,
210R (1962).
Herrmann, L., Chemical Analysis by Flame Photometry, Wiley-Interscience,
New York (1963).
Hodgsen, J. F., zt at., "Micronutrient Cation Complexing in Soil Solution:
II. Complexing of Zn and Cn in Displaced Solution from Calcareous
Soils," Proceedings Soil Science Society America, 3jD, 723 (1966).
Home, R. A., "Marine Chemistry," Wiley Interscience, New York (1969).
Hughes, G. M., ztoJL.* "Hydrogeology of Solid Waste Disposal Sites in
N.E. Illinois," Report SW-12d, U.S. Environmental Protection Agency,
Washington, D.C. (1971).
Jenne, E. A., "Controls on Mn, Fe, Co, Ni, Cu and Zn Concentrations in
Soils and Water: The Significant Role of Hydrous Mn and Fe Oxides,"
Trace Inorganics in Water. Adv. Chemistry Series, 73^337 (1968)
Johnson, R. W. and Schrank, W. G., "Mechanisms of Interactions of Alkali
and Alkaline Earth Elements by Flame Photometry," Applied Spectroscopy
J8., 144 (1964).
Kahn, H. L. and Manning, D. C., "Background Correction in Atomic Absorption
Spectroscopy, Amer. Laboratory, 4, 51 (1972).
Merz, R. C., "Final Report on the Investigation of Leaching of a Sanitary
Landfill," Publication No. 10, State Water Pollution Control Board,
Sacramento, California (1954).
Minear, R. A. and Murray, B. B., "Methods of Trace Metal Analysis in
Aquatic Systems," Trace Metals and Metal Organic Interactions in
Natural Waters, P. Singer, ed., Ann Arbor Science Pub., Inc., Ann
Arbor, Mich. (1973).
Qasim, S. R. and Burchinal, J. C., "Leaching from Mimulated Landfills,"
Journal Water Pollution Control Federation, 42, 371 (1970).
148
-------
Ramakrishna, T. V., vt at., "The Determination of Calcium and Magnesium
in Acetylene Flames," Analytica Chimica Acta, 40, 347 (1968).
Roos, J. T. H. and Price, W. J., "Mechanism of Interference and Releasing
Action in Atomic Absorption Spectrometry - I," Spectrochimica Acta,
26B, 279 (1971).
Schnitzer, M. and Kahn, S. U., Humic Substances in the Environment,
Marcel Dekker, Inc., New York (1972).
Shapiro, J., "Effect of Yellow Organic Acids on Iron and Other Metals
in Water," Journal American Water Works Association, 56, 1062 (1964).
Siegel, A., "Metal Organic Interactions in the Marine Environment,"
Organic Compounds in Aquatic Environments, S. D. Faust and J. V.
Hunter, Eds., Marcel Dekker, Inc., New York (1971).
Sipalo-Zuljeric, J. and Wolf, R. H. H., "Sorption of Lanthanum, Cobalt
and Iodide Ions at Trace Concentrations on Ferric Hydroxide,"
Mikrochimica Acta. 1973. 315 (1973).
Slowey, J. F., Jeffrey, L. M. and Hood, D. W., "Evidence for Organic
Complexed Copper in Sea Water," Nature, 214, 377 (1967).
Stumm, W. and Bilinsky, H., "The Apparent Supersaturation," Proceedings
6th International Conference on Water Pollution Research, Jerusalem,
1972, Advances in Water Pollution Research, S. H. Jenkins, ed.,
Pergamon Press, Oxford (1973).
Willard, H. H., at aJi., Instrumental Methods of Analysis, Van Nostrand
C., Inc., Princeton, N.J., (1965).
Yanagisawa, it at., "Cationic Interferences in the Atomic Absorption
Spectrophotometry of Chromium," Analytica Chimica Acta, Vol. 52^
386 (1970).
149
-------
VI
EVALUATION OF COLORIMETRIC TESTS USED FOR
CHARACTERIZATION OF ORGANIC MATTER IN LEACHATE
CONCLUSIONS
The present study evaluated several colorimetric tests used to charac-
terize functional groups and general classes of organics present in the
different molecular weight fractions in leachate. Although most studies
in the area of soil science use titrimetric methods for functional group
analysis these methods were not applicable in the present study due to
their low sensitivity. Colorimetric methods are preferred only when
small amounts of organic matter are present. The hydroxyl amine test
for measurement of carboxyl groups and free volatile fatty acids
was evaluated extensively. The latter group of organics can also
be measured using gas chromatography, employing m-cresol as an internal
standard. The carbonyl groups were measured with the dinitrophenyl-
hydrazinemethod while the aromatic hydroxyl groups were determined with
the Folin-Denis method. The anthrone test was used to determine carbo-
hydrates, proteins were measured with the ninhydrin method and the
Kjeldahl method, while fats and oils were measured as hexane extractables.
Evaluation of the hydroxyl amine test showed that the spectra obtained
with the different molecular weight fractions in leachate corresponded
to that of carboxyl groups attached to an aliphatic carbon chain , and
was dissimilar from a carboxyl group attached to dihydroxyl benzene rings
or hydroxylnapthalenes. It was further noted that the molar absorbance
of most free volatile fatty acids is similar? however, this does not
hold after introduction of a methyl or hydroxyl group in the aliphate
carbon chain. Evaluation of the 2,4-dinitrophenyl hydrazine method showed
that most carbonyl groups were present as aliphatic aldehydes and not
as aliphatic ketones or aromatic aldehydes, a conclusion based on the
150
-------
specificity of the spectra of each group of compounds. The Folin-Denis
test used to determine aromatic hydroxyl groups was found to be one of
the most sensitive ones; however, the model compound has to represent
the actual compound as closely as possible. The anthrone test for carbo-
hydrate analysis was sometimes interfered with by the formation of brown
colored products possibly as a result of the formation of furfural rings.
Selection of a short hydrolysis time was found to eliminate this;
aromatic hydroxyl compounds did not interfere with the test. The
Lowry method, commonly used to measure proteins, was found to be
greatly interfered by aromatic hydroxyl compounds; spiking of the sample
with 1 mg/1 tanm'c acid resulted in an apparent protein concentration of
3.8 mg/1 and the less sensitive but more accurate ninhydrin test was
therefore selected. The peptide bonds in this test are hydrolyzed by
refluxing with 6 N HC1 for 20 hours followed by addition of the ninhydrin
reagent to the solution. Comparison of the Kjeldahl method and the
ninhydrine method showed the results obtained with each test agree well
with each other. Evaluation of extraction techniques using hexane and
butanol followed by gravimetric determination of the residue showed that
significant amounts of free volatile fatty acids are coextracted to result
in erroneous high readings of the dried residue. The butanol, often used
to extract fulvic acid, was found to preferentially extract aromatic
hydroxyl compounds.
151
-------
THE DETERMINATION OF CARBOXYL GROUPS WITH
THE HYDROXYL AMINE TEST
THE HYDROXYLAMINE TEST
The hydroxylamine method used to determine carboxylic acids or carboxyl
groups was first reported by Lessen in 1869 (Goddu e_t aj_., 1955). However,
it was not until 1934 that Feigl ejt al_. (1972) used the hydroxamic acid
together with complexed ferric iron as a spot test for organic acids.
Keenan (1945) applied the method for determining acetic acid esters while
Lipmann and Tuttle (1945) used it to determine the concentration of
organo-phosphate esters at a wavelength of 540 nm. Sulfate and phosphate
were found to interfere with the test by forming iron complexes competing
with the hydroxamine acid complex for iron. Hill (1946) used the test
to determine fats in ether solution using wavelength of 520 nm. Hestrin
(1949) confirmed the findings of Lipmann and Tuttle (1945) that phosphate
could reduce the color at 520 nm which interference could only be elim-
inated by using excess ferric chloride. Hestrin (1946) also noted that
the pH of the final complex did not have any effect on the color between
pH of 1.0 and 1.4 and that equal molar quantities of acetic, propionic
and butyric acid resulted in the same amount of color. Branched acids
however, gave a lower color intensity. Peptides did not interfere with
the test, but ketonic substances did so slightly as a result of oxime
formation, a result confirmed by Buckles and Thelen (1950). Goddu et al.
(1955) conducted the test in anhydrous ethanol and found that at least a
fivefold excess of iron is necessary for satisfactory final color develop-
ment. Aromatic esters were found to have maximums at 530 nm while for
aromatic esters it shifted to 550 nm.
Aksnes (1957) noted that excess hydroxylamine could cause a fading of the
color by reducing the ferric to ferrous iron. Absorption readings therefore
have to be made directly after the ferric chloride addition. The maximum
color was formed at a pH of 1.4 and increasing the pH caused the absorp-
tion maximum to shift to a wavelength of 440 nm corresponding to a brown
152
-------
color instead of the purple color. Bayer and Reuther (1956) reported that
acetic, propionic, butyric, valeric and caproic acid gave the same molar
absorption. Lower molar readings however, were obtained with isobutyric
and isovaleric acid and also with formic acid, confirming the results of
Hestin (1949). Morgan and Kingsbury (1959) reported that all fatty acid
esters gave a maximum absorption at 515 nm and that the chainlength has
neither an effect on the wavelength at which the maximum absorption occurs
nor does it affect themolarabsorption thereby confirming Bayer's results.
Goldberg and Spoerri (1958) noted that the presence of an electron releas-
ing group in the molecule increased the rate of hydroxylamine formation
which in turn corresponded to a lower color intensity. The closer the
electron donor group is located near the carboxyl group, the more pro-
nounced its effect, since methyl ester of toluate with the methyl group
of the toluate in the ortha, meta and para position respectively, gave
readings of 1.05, 1.29, and 1.42 times higher than found for methyl acetate,
A similar result was obtained with aliphatic esters, thus explaining the
lower reading obtained with branched acids. Aromatic compounds were
found to give a hyperchromic effect as compared to aliphatic compounds.
An electron pulling group, such as the hydroxyl group in methyl-p-hydrox-
ybenzoate, resulted in an absorbance 1.79 times that of methyl acetate,
while a hydroxyl group in the para position gave an absorbance 1.45 times
higher. An ortho methoxy group was less effective as an electron pulling
group than a hydroxyl group and resulted in a reading 1.60 times higher
than methyl acetate. The authors did not investigate benzene rings with
more than one hydroxyl group, nor did they investigate the different
spectra since all compounds were measured at 540 nm.
Lehmann and Wilhelm (1967) found, contrary to Hestin (1949), that excess
ferric iron concentrations did not influence the absorption. They did
note a salt effect which caused a depression of the absorption. The
effect was more pronounced with increasing salt content and increasing
valence of the ions. Using ferric sulfate instead of ferric chloride
also lowered the absorbance readings noticeably. The ferric chloride
also resulted in the slowest fading of the formed complex as compared
153
-------
to ferric sulfate. The maximum color development with the ferric hydrox-
ylamine acid complex occurred at a pH of 1.8.
Montgomery e£ al_. (1962) proposed the hydroxy 1 amine methods for measuring
free volatile fatty acids in digester liquor and they included an esterifi'
cation step with ethylene glycol in their method. More than 13 mg/1 of
acetic acid could be detected with this method; however, when colored
samples were processed, a sample blank had to be included in order to
obtain such sensitivity. The maximum color was formed at pH 1.6, and
the absorbance was measured at 500 nm. They also noted, contrary to
earlier findings, that the equimolar absorbances of propionic and butyric
acid were only 88 and 87 percent of those obtained with acetic acid; the
addition of the esterification step may well have caused this discrepancy
with earlier investigations.
The above review showed that only one study used the hydroxylamine method
for measuring fatty acids, while all other studies used the method to
detect esters. Only one study investigated to some extent the response
of aromatic esters; however, including the esterification step may well
lead to different results. Other aromatic compounds that could be
present in the fulvic and humic acids (Schnitzer and Kahn, 1972) were
not tested.
PROCEDURES
The carboxyl test as used by Montgomery et.al_. (1962) to determine free
volatile fatty acids was investigated and its applicability to aromatic
acids was also evaluated. The reagents and procedures were as follows:
Reagents:
1. 50% sulfuric acid
2. ethylene glycol
3. 4.5 N NaOH
4. 10% hydroxylamine hydrochloride (10 gms in 100 ml HC1)
5. 20 gms ferric chloride, 20 ml concentrated sulfuric made up
to 1 liter with distilled water
Sequence:
1. place 1.0 ml sample in tube
2. add 1.5 ml ethylene glycol
154
-------
3. add 0.20 ml 50% H2$04
4. stir mixture well for 2.0 min, at 25°C
5. cool to room temperature in water bath and add 0.50 ml
hydroxylamine solution
6. add 2.0 ml NaOH solution,mix well and wait 1-2 minutes
7. add 10.0 ml ferric chloride solution
8. add 5.0 ml distilled water
9. read absorbance at 500 nm after 30 minutes
The reaction that takes place during the test is as follows:
Ester H+ S
Formation: R-COOH + OHCH2CH2OH 9- R-C-OCH2CH2OH + H20
Hydroxamate „ nH- „ ,
Formation: R-C-OCH2CH2OH + NH2OH U3 R-C-N-H + OHCH2CH2OH
Color S °-H ... „+
Formation: R-C-N-H + Fe " R-C-N-H
it i
0 0-H
Colored 1/3 Fe+++
Complex
COMPOUNDS GIVING A POSITIVE TEST
The results of the tests with the different free volatile fatty acids are
shown in Figure 28. Acetic acid gave the highest molar absorbance while
the other fatty acids such as propionic, valeric and caproic acid gave
identical absorbances 79 percent lower than observed for acetic acid,
which result agrees with Montgomery e_t aJL (1962). Formic acid gave only
33 percent of the absorbance of acetic acid. As all absorbances were
measured at 500 nm, the spectra of the different compounds were evaluated
to determine whether 500 nm was indeed the optimum wavelength. The
spectra of the different acids at a concentration of approximately 2.5
mM/1 are shown in Figure 29a. The maximum absorbance is indeed observed
between 500 and 510 nm which is indeed best suited for measuring the
purple color of the solution. The introduction of a methyl or hydroxyl
155
-------
en
(Q
n>
ro
oo
a>
a-s
-her
a> w
-$ o
n> -$
J»o
-s n>
o -n
o n>
o a>
•a *z
o o
=i a>
a-rt
a>
P>
cfr
o
VI
o>
o.
Q50
Caproic Acid
Valeric Acid
/ ''I*
/Gallic /
AcidV /
} r X
Dihydroxy Ben zoic Acid ...rj
9.-"""
.
10 15
Molar Concentration, mM
-------
Acetic Acid
Propionic Acid
Butyric Acid
Valeric Acid
Caproic Acid
« 0,08
i
0,06
0,04
0,02
400
-Valeric Acid
3 Hydroxy Butyric Acid
500
600
700
Wavelength, nm
Figure 29. Absorption Spectra of Free Volatile Fatty Acids
(A) and Branched Free Volatile Fatty Acids at a
Concentration of Approximately 0.25 mmole/l
157
-------
group into the aliphatic acid does not change the maximum absorption
(Figure 29b) but only reduces its magnitude. Although the introduction
of an electron donating group such as a hydroxyl group was expected to
increase the absorption, such effect was not observed.
Several aromatic compounds which could be present in the fulvic and
humic acid fractions were tested for their ability to give a color reac-
tion:
OH
catechol
resorcinol
COOH
HO | OH
OH
pyrogallol gallic acid salicylic acid
COOH
COOH
hydroquinone benzoic acid m-hydroxyl
benzoic acid
Only the first four gave a reaction (Figure 28) while no color development
was observed with the latter five compounds. The color was different from
that observed with the aliphatic acids since catechol gave a brown color,
gallic acid and pyrogallol a green color, tannic acid a grey color, while
the color from salicylic acid was violet and identical to those of the
aliphatic acids. The spectra of the different aromatic compounds are shown
in Figure 30. The spectrum of salicylic acid (800 mg/1) is almost
identical to that of caproic acid (1000 mg/1) with the exception that
the absorbance decreases less rapidly at higher wavelengths. The ratio
of the wavelength at 560 to that of 500 nm is 88 for salicylic acid and
only 74 for caproic acid. Such internal ratios can thus be used to
158
-------
en
10
O
§ 3=»
T3 O
O ->•
c. a.
3
Q. O>
V) 3
CL
-$
3
rl-
0.8r—i—i—i—i—i i "t—i—i
Blank
Acetic Acid
Catechol
Caproic Acid
•• Gallic Acid
Tannic Acid
0.61— PyroGallol
3-Hydroxy-2-Napthoic Acid
Phenyl Acetic Acid
0.51 2.4 Dihydroxy Benzoic Acid
„ Salicylic Acid
o
o
JS
<
600
550
500
Wavelength, nm
450
400
-------
distinguish between different compounds. It should be realized however
that the ratio can only be used when the absorption maxima are of equal
magnitude. As catechol and pyrogallol are unable to form hydroxamates
it was concluded that they formed a complex such as:
It is not yet known whether the color in salicylic acid is due to the
reaction of the hydroxyl in the carboxyl group, which gives the normal
3+
hydroxamate colored complex, or a complex in which the Fe is coordinated
by the carbonyl of the carboxyl group and the oxygen of the ortho hydroxyl
group.
OH
C/
\0
o
To further evaluate this hypothesis a para nitro salicylic acid, or other
ortho-hydroxyl benzoic acids in which the nitro group deactivates the COOH,
will have to be tested. The inability of benzoic acid or m- or o-hydroxy-
benzoic acid to form a colored complex also has to be further investigated
and it has to be determined whether this is due to the inability of the
carboxyl group to esterify or due to the difficulty to form hydroxamic
acids.
160
-------
EFFECT OF VARYING FERRIC IRON CONCENTRATIONS
The effect of varying the concentration of the ferric iron solution was
*t"++
next evaluated. The Fe concentrations chosen were 1.0, 2.0, 4.0 and
8.0 mM, while in the standard procedure the Fe concentration is 3.7 mM
It was found that at the organic acid concentration studied, with a
hydroxamate concentration of 0.50 mM in the final solution, Fe ceased
to be limiting beyond 8.0 mM Fe (Figure 31). It was concluded that in
3+
the standard procedure with an Fe concentration of3.7mM, small changes
in Fe concentration would have no noticeable effect on the color
intensity.
FADING OF FORMED COLOR
Readings were taken at 0 and 30 minutes, 2 hours, and 16 hours with
salicylic acid, formic acid and valeric acid in order to determine the
stability of the different complexes. It was found that the color of
salicylic acid reached its maximum at 30 min, then decreased about 25
percent after 16 hours (Figure 31). Both formic acid and valeric acid
showed an increased color after 2 hours as compared to the reading after
30 minutes. After 16 hours the color of the valeric acid solution had
not yet faded to the original 30 minute reading, while the color in the
formic acid solution decreased to about 66 percent of its original value
after 16 hours. It was therefore concluded that if readings are taken
at 30 minutes +. 1 minutes, none of the colors have yet begun to fade
and the least error is incurred with the different compounds. The result
further show that not adhering strictly and systematically to this time
limit will cause erroneous results.
EFFECT OF VARYING HYDROXYLAMINE CONCENTRATION
Several tests were made in order to determine the effect of changing the
hydroxylamine concentration. Hydroxylamine concentrations of 5.0, 10.0,
20.0 and 40.0 mM were tested with a concentration of 400 mg/1 of acetic
acid. The normal procedure requires 35 mM NhLOH, with concentrations
referred to final volumes. It was found that at 35 mM the color formation
161
-------
4mM Catechol
10 mM Acetic Acid
10 mM Caproic
I
Test Procedure
i I
10 20 30
Ferric Iron Concentration, mM
40
o
V)
JQ
0.20
e QIO
Test Procedure
I
1000 mg/J
Valeric Acid
1000 mg/J?
Formic Acid
10
Time, hours
15
20
Figure 31. Effect of Varying Iron Concentration (A) and Varying
Times Allowed for the Iron Complex Formation (B)
162
-------
was limited by some factor other than hydroxylamine (Figure 32a),
possibly the acetic acid concentration. It was therefore concluded
that the selected hydroxylamine concentration would not cause any limitation
in the hydroxamate formation.
EFFECT OF VARYING HYDROXAMATE FORMATION TIME
The length of the reaction period that the hydroxylamine is exposed to
the carboxylester was also evaluated since increasing time periods would
increase the hydroxamate formation. Tests were made at the standard
temperature of 25°C for various time periods. The procedure was also
tested at 100°C to determine if an increased temperature would affect
the hydroxamate formation. It was found that at 25°C the reaction is
essentially complete and stable after a few minutes and only a slight
increase was observed over a following period of 15 minutes (Figure 32b).
At 100°C, however, a decrease in the violet color was observed from 1.0
to 2.0 minute incubation, followed by a subsequent rise at 4.0 minutes
due to the formation of a brown colored complex. It therefore appears
that high temperatures promote the decomposition of the hydroxamate and
hasten the formation of these brown colored products. Since it was
shown that the reaction is nearly complete upon mixing at 25°C, the
ferric chloride solution can be added after 1 to 2 minutes, thereby
confirming the validity of the test procedures.
ANALYSIS OF LEACHATE SAMPLES
The leachate samples were tested and the amount of carboxyl groups were
calculated as acetic acid equivalent. As the COOH measured in the membrane
permeates is derived from fatty acids, a violet color was developed.
The membrane retentates also gave a violet color and a spectrum similar to
that of acetic acid, reflecting carboxyl groups which are probably associated
with either aromatic carboxyl groups in combination with an ortho-phenol
group, or from carboxyl groups attached to an aliphatic carbon chain which
in turn may be attached to an aromatic ring. It is clearly not associated
with dihydroxybenzoic acids or hydrbxy napthoic acid.
163
-------
0)
o
c
o
jQ
O
U>
.Q
0,2
o 0,1
o
400 mg/f Acetic Acid
Test Procedure
t I
10 20 30 40 50
Hydroxylamine Concentration, mmole/Jf
1000 mg/t
1000 mg/I Valeric Acid (25°C)
Valeric Acid (IOO°C)
Yellow Brown
Violet
Test Procedure
I
I
60
100 200
Hydroxamate Formation Time, min.
Figure 32. Effect of Varying Hydroxyl Amine Concentrations (A) and
Varying Times Allowed for the Hydroxamate Formation (B)
300
164
-------
CONCLUSIONS
The hydroxylamina colorimetric test for organic acids is reliable and
reproducible for the compounds tested. It detects all aliphatic acids,
with identical molar absorption coefficients except for acetic acid and
formic acid. However.the molar absorbance decreases when methyl or
hydroxyl groups are introduced in the aliphate carbon chain. This implies
that the acid selected as standard has to resemble the most common acid
present in the solution as closely as possible. The test also detects
aromatic acids, especially aromatic carboxyl groups associated with an
ortho phenol group, as commonly occurs in humic and fulvic acids.
165
-------
REFERENCES
Aksnes, G., "The Complex Between Hydroxartric Acids and Ferric Ions and
the Use of the Complex for Quantitative Determination of Hydroxamic
Acid, Acyl Derivatives and Ferric Salts," Acta Chemica Scandinavica,
n_, 710 (1952).
Bayer, E. and Reuther, K. H., "Photometric Determination of Acyl Groups;
The Analytical Application of Ferric Hydroxamic Acid Complex,"
Chemische Berichte, 89_, 2541 (1956).
Buckles, R. E. and Thelen, C. J., "Qualitative Determination of Carboxylic
Esters," Analytical Chemistry, 2£, 646 (1950).
Feigl, F., et^ al., Spot Tests in Inorganic Analysis, Elsevier Publishing
Company, Amsterdam (1972).
Goddu, R. F. e_t ajL, "Spectrophotometric Determination of Esters and
Anhydrides by Hydroxamic Acid," Analytical Chemistry, 27., 1251 (1955)
Goldenberg, V. and Spoerri, P. E., "Colorimetric Determination of Carboxylic
Acid Derivatives as Hydroxamic Acids," Analytical Chemistry, 30,
132 (1958).
Hestrin, S., "The Reaction of Acetylcholine and Other Carboxylic Acid
Derivatives with Hydroxyl Amine and its Analytical Application,"
Journal of Biological Chemistry, 8£ 249 (1949).
Hill, U. T., "Colorimetric Determination of Fatty Acids and Esters,"
Analytical Chemistry. J8., 317 (1946).
Keenan, A. G., "Colorimetric Determination of Ethyl Acetate," Canadian
Chemical Processing. 29_, 857 (1945).
Lehmann, G. and Wilhelm, G., "Colorimetric Determination of Esters of
Phtalic Acid with the Hydroxyl Amine Method in the Presence of Esters
of Other Carboxyl Acids," Z. Anal. Chem., 238. 415 (1967).
Lipmann, F. and Tuttle, L. C., "A Specific Micromethod for the Determina-
tion of Acylphosphates," Journal of Biological Chemistry, 159, 21
(1945).
Montgomery, H. A. C. et aJL, "The Rapid Colorimetric Determination of
Organic Acids ancTtheir Salts in Sewage Sludge Liquor," The Analyst.
87., 949 (1962)
Morgan, D. M. and Kingsbury, K. J., "A Modified Hydroxamic Acid Method for
Determining Total Esterified Fatty Acids in Plasma," The Analyst. 84_
409 (1959).
166
-------
Schnitzer, M. and Kahn, S. U., Humic Substances in the Environment,
MarceT Dekker, Inc., New York (1972).
167
-------
THE DETERMINATION OF FREE VOLATILE FATTY ACIDS BY
GAS-LIQUID CHROMATOGRAPHY
THE GC ANALYSES
The use of gas-liquid chromatography has been growing rapidly because of
sensitivity, resolution and speed of the analysis. It has thereby super-
seded the tedious and less accurate methods of column, paper and thin
layer chromatography (Mueller ert a_]_., 1956). The first use of gas-liquid
chromatography in the water pollution area was reported by Lamar and
Goerlitz (1963), which method was followed by numerous other inves-
tigators. Due to advancing technology new liquid phases were developed
that gave better separation of the mixture studied. Andrews e_t aJL (1964)
used a polyethylene glycol PEG 20,000 + H^PO* on firebrick to determine
acids in digesters while Hindin e£ a_l_. (1964) used the lower polymer
PEG 4000 + H-PO^ on Chromasorb W to determine the fatty acids in sewage.
Murtaugh and Bunch (1965) measured the acids in effluent of several
treatment plants by using a 6 mm by 1830 mm column packed with 20 percent
(by weight) diethylene glycol adipate polyester and 2 percent phosphoric
acid (85 percent concentrated HgPO^) on 60-80 mesh Chromasorb W with a
helium gas flow rateof 50 ml/minute. The column temperature was maintained
at 125°C for formic (C.,) to isobutyric (C4) and 150°C for butyric (Cj) to
caproic acid (Cg).
A similar glycol ester was used by Mateles and Chian (1969) to detect fatty
acids in activated sludge by employing a 20 percent neopentylglycol
succinate (NPGS) + 2 percent H-PO, on 60-80 mesh firebrick adapted after
Baumgardt (1964). A 3.17 mm x 1830 mm coiled stainless steel column
(type 316) was installed. Compared with Baumgardt (1964) the oven temp-
erature was 175°C instead of 165°C, injector temperature 265°C instead of
195°C, nitrogen gas carier flow of 39.8 ml/nrin instead of 26 ml/min and
hydrogen flow rate to detector 18 ml/min instead of 20 ml/min. The air
flow rate to the detector was 290 ml/min. In all cases sample sizes
injected into the GC were between 0.5 and 1.0 microliter.
168
-------
PROCEDURES
The columns used in the present study are similar to those of Mateles and
Chian (1969). As solid support, acid washed chromosorb P (AW) was employed.
The hydrogen and air flow rates to the flame detector however were increased
to 50 ml/min and 525 ml/min, respectively. Also, a molecular sieve trap
was installed in the outlet of the nitrogen gas tank. This trap was regen-
3
erated for every other tank of N2, i.e., 18.7 m of N2, in order to ensure
a steady baseline in the recorder. The detector of the Hewlett Packard
5750B gas chromatograph was set at Range 1 and an attenuation of 64 and
128. The Sargent-Welch SRG recorder was set at 1 mv. Phosphoric acid,
that was added to the glycol stationary phase to improve the polarity,
proved to be unstable at higher temperatures. Much better separation will
therefore be obtained with a recently developed stationary phase FFAP
(free fatty acid phase), a reaction produced between polyethelenglycol
20,000 and a nitroterephthalic acid. Improved separation was also obtained
by using polyaromatic polymers as packing material. Van Huysste&n (1970)
used 3 percent FFAP on a styrene-divinyl benzene copolymer and was able
to analyze fatty acid concentrations in digesters as low as 5-10 mg/1.
The order of magnitude of response per unit weight of acid was Cg, i-C^,
C2, C4, i-C5> C5, and Cg. Because the concentration of fatty acids in
leachate is usually quite high the NPGS phase gave results accurate enough
to not necessitate the more sensitive FFAP columns.
METHOD USED FOR CALCULATING FATTY ACID CONCENTRATIONS
Ideally, the area under a peak is proportional to the concentration of the
compound. Because the area of a triangle is equal to the height times half
the basewidth.the concentration is proportional to the height if the base
width does not change appreciably between separate injection. This occurs
if operating conditions are fixed, small samples are used and the gas flow
rate is constant. The concentration of these acids in a sample can be
calculated by comparing the peak height with peak heights of acids in a
standard solution of a known concentration. Also, by adding increasing
but known amounts of acids to the sample and by extrapolation to zero
169
-------
concentration one can calculate the concentration of the acids present.
The most accurate method, which was also employed by Mateles and Chian
(1969) is the use of an internal standard which has a larger retention
timethan any of the known acids. After testing many substances, m-cresol
was selected as the best internal standard at a concentration of 200 mg/1
The calibration curves (Figure 33) relate the ratio of fatty acid/
m-cresol peak height to the acid concentration in the standard. The
calibration curves show that the i-C, acid can be detected most accurately
to as low as 20 ppm. Hexanoic acid has the lowest sensitivity. Linearity
of peak height with increasing acid concentration was observed in all
cases. A similar amount of 250 mg/1 m-cresol is added to the diluted
sample to be analyzed and from the calculated ratio of acid to m-cresol
peak height the actual fatty acid concentration is determined from the
calibration curve.
OPERATIONAL COMPLICATIONS ENCOUNTERED
Most problems encountered with the gas chromatograph are related to
"ghosting" which is defined as the appearance of peaks with the same
retention time as the previous injected acids when only water is injected.
Baumgardt (1964) suggested that injector assembly and glass wool column
plugs should be cleaned frequently. Distilled water samples should be
injected to clean the column. Mateles and Chian (1969) minimized
ghosting by coating the stainless steel wool in the injection port with
15 percent phosphoric acid. Ghosting is more pronounced in sensitive
columns. Van Huyssteen (1970) experimented with this by injecting 2 yl
distilled water after 2 yl of 25 mg/1 C2-Cg solution or 1 yl of dis-
tilled water after 1 yl of 750 mg/1 C2-Cg solution. Although Mateles and
Chian (1969) usually worked in the range of 100-1000 ppm volatile fatty
acids and a m-cresol concentration of 1000 ppm, our experiences with these
high concentrations were less satisfactory. For example, at a concentration
of 1000 ppm Cp-Cg the ratio of acetic acid m-cresol in sequential injec-
tions could increase from 1.22 to 1.30 to 1.38, respectively. Other non-
linearities and ghosting phenomena were also experienced which led us to
use lower concentrations of acids, in the range of 50 to 300 ppm and
using the internal standard of 250 ppm m-cresol.
170
-------
4,0
3,0
JC
O
o
V)
o
I 2.0
o
1,0
50
100
200
Butyric
Isobutyric
Propionic
Isovoleric
Valeric
Acetic
Caproic
Formic
300
Figure 33.
Fatty Acid Concentration (mg/O
Calibration Curves for C2-Cg Volatile Fatty Acids
171
-------
An example of the analysis of a leachate sample from the Boone County
Field Site of SHWRL, EPA collected on October 2, 1972 is shown in Figure
34. The dilution employed is used to analyze for propanoic-, isobutyric-,
isovaleric-, valeric-, and hexanoic acid. A dilution 4 times higher
enables the determination of acetic- and butyric acid.
REFERENCES
Andrews, J. F., Thomas, J. F. and Pearson, E. A., "Determination of
Volatile Acids by Gas Chromatography," Hater and Sewage Works, 4
206 (1964). ~~
Baumgardt, B. R., "Practical Observations on the Quantitative Analysis
of Free Volatile Fatty Acids in Aqueous Solutions by Gas Liquid
Chromatography," Dept. Bulletin No. 1, Dept. Dary Science, University
of Wisconsin, Madison (1964).
Hindin, E., e_t a]_., "Analysis of Volatile Fatty Acids in Sewage by Gas
Chromatography," Water and Sewage Works. 4_, 92 (1964).
Lamar, W. L. and Goerlitz, D. F., "Characterization of Carboxylic Acids
in Unpolluted Streams by Gas Chromatography," Journal American Water
Works Association. 55_, 797 (1963).
Mueller, H. F., Buswell, A. M. and Larson, T. E., "Chromatographic
Determination of Volatile Acids," Sewage and Industrial Wastes, 28,
255 (1956) ~~
Mateles, R. I. and Chian, S. K., "Kinetics of Substrate Uptake in Pure
and Mixed Cultures," Environmental Science and Technology, 3, 569
(1969).
Murtaugh, J. J. and Bunch, R. L., "Acidic Components of Sewage Effluents
and River Water," Journal Water Pollution Control Federation, 37_
410 (1965).
Van Huyssteen, J. J., "The Determination of Short Chain Fatty Acids in
Aqueous Solution by Gas Liquid Chromatography," Water Research, 4_
645 (1970).
172
-------
L-*—jj-Acet
I ^-But
8 Ik-Prop
r
Range R = I
Attenuation A= 64
mv = I
Gas Flowrate - 40ml/min
Inj, Port Temp = 265°C
Detect or Temp = 265° C
Column Temp - 175°C
Material - 20% Neopentyl
Glycol Succinate-
27o H3P04 On
Chromosorb PAW
60-80 Mesh
:m-Cresol
/-Isobut
Isoval
-Valeric
Hexanoic
6 8
Time, min
10
12
14
Figure 34. Gas Chromatogram of Free Volatile Fatty Acids Detected in
Leachate of the Boon County-EPA Landfill
173
-------
THE DETERMINATION OF CARBONYL GROUPS WITH THE
DINITROPHENYL HYDRAZINE METHOD
THE HYDRAZONE METHOD
The most widely used methods for determining carbonyl groups are the
oxime- and phenylhydrazone methods. The phenylhydrazone method which
uses 2,4 dinitrophenylhydrazine as the active reagent to attach to the
carbonyl group is interfered by excess reagent. This interference
can be minimized by either selective extraction of the intense yellow
derivative using a hydrocarbon solvent (Lohman, 1958) and measuring
the absorbance at 340 nm or by addition of alkali to the solution of
the derivative and measuring the wine red complex at 430-480 nm (Hanna,
1966). The molar absorptivity and the absorption maximum is dependent
on the nature of the compound and is mainly influenced by the saturated
or unsaturated character of the compound. The 2,4 dinitrophenylhydrazine
method can also be used to precipitate and separate the different
aldehydes that react with it using thin layer chromatography (Meyboom,
1968) or gas chromatography (Soukup, 1964; Fiacchia e_t aj_., 1967).
The determination of small amounts of carbonyl compounds was first
reported by Mathewson (1920) who measured acetone in water samples
after extraction of the complex with carbon tetrachloride and measuring
the absorbance at 435 nm. Barrenscheen and Dreguss (1931) and Barta
(1934) used the purple color developed in alkaline solutions to determine
concentrations of pyruvaldehyde and furfual respectively. The 2,4
dinitrophenylhydrazine has a proton donating character as a result of
the two nitrogroups and will release a proton after addition of base,
which causes the formation of wine red quinoidal compounds:
OH" ? /=V '?' /=\
_ fi = R — 8-C_>= fiH - N = R^N—4T/" N - N = F
-Q| 101 ^N - 01
II - - I
I fM I U|
I0|
174
-------
Iddles and Jackson (1934) used the precipitate of 2,4 dinitrophenyl-
hydrazine and different ketones and aldehydes as a quantitative method
to determine the concentration of the compounds. A saturated 2,4
dinitrophenylhydrazine solution in 2 NHC1 was added to the aqueous sample
after which a precipitate formed within 1 hour. An increased reaction
time or increased temperature decreased the yield of the precipitate due
to hydrolysis. When small amounts of aldehydes are present in aqueous
solution, addition of 2,4 dinitrophenylhydrazine will only cause a
turbidity and no precipitate in the sample, which turbidity can be
measured quantitatively with a nephelometer.
Lappin and Clark (1951) determined the 2,4 dinitrophenylhydrazine complex
of carbonyl compounds at a wavelength of 480 nm in an alkaline methanol
solution. Excess hydrazine reagent gave a light yellow color in alkaline
solution, the absorbance of which was corrected for by means of a blank.
They added 1 ml of saturated 2,4 dinitrophenylhydrazine to 1 ml of sample
dissolved in methanol. The sample with the methanol was first acidified
by adding 1 drop of concentrated HC1. The mixture was heated for 5 minutes
at 100°C or 30 minutes at 500°C and after cooling 5 ml of a methanol KOH
(10 g KOH, 20 ml HgO diluted to 100 ml with methanol) was added to give a
wine red color. They reported that the molar absorption and the maximum
wavelength were independent of the structure of the aldehyde compound.
Mendelowitz and Riley (1953) found that most aliphatic ketones gave a
maximum absorption at 430 nm instead of 480 nm as used by Lappin and
Clark (1951). They also noted that a fading of the color at 430 nm would
occur after formation and that the absorption maximum is influenced by
the structure of the compound. After addition of KOH to raise the pH of
the solution a potassium chloride precipitate was observed. Jones et al.
(1956) also found that the absorption maximum is influenced by the
structure of the compound. They found that in neutral solutions
aliphatic ketone hydrazones had absorption maxima at 364-367 nm which
shifted to 430-440 nm in basic solutions. A second peak appeared at
175
-------
526-535 nm which did not decrease rapidly with time. Aliphatic aldehyde
hydrazones had maxima at 344-358 nm in neutral solutions which shifted to
426-435 nm in basic solutions. A second peak appeared at 515-523 nm
which faded more rapidly with time than observed for the ketonic compounds.
Formaldehyde hydrazone exhibited the most rapid decrease with 50 percent
loss in 15 minutes. Phenyl acetaldehyde hydrazone was the most stable
but still lost 30 percent of its color after 60 minutes. Aromatic ketone
hydrazones had a maximum at 386-399 nm in neutral solutions which
shifted to 458-490 nm in basic solution; no peak was observed at 520 nm.
Aromatic aldehyde hydrazones absorbed at 381-390 nm which maximum shifted
to 460-486 nm in a basic solution. Similarly to the aromatic ketones,
no second maximum was observed at 520 nm.
Addition of water to the hydrocarbon-alcohol solvent reduced the problem
of the KC1 precipitates after addition of the basic methanol (Jordan and
Veatch, 1964). They noted similarly to Mendelowitz and Riley (1953) that
after the addition of base the absorbance must be determined within the
interval of 8-15 minutes, since the color will fade beyond that time.
They dissolved the sample in 5 ml 3:7 hexane-alcohol solution after which
2 ml of a saturated 2,4 dinitrophenylhydrazine alcohol solution and 0.1 ml
of concentrated HC1 was added. After heating at 55°C for 30 minutes it
was diluted to 25 ml with a basic alcohol solution (59 g KOH, 180 ml H20,
diluted to 1 1 with alcohol). They observed a lower maximum wavelength
for aromatic ketones than mentioned by Jones et^ al_. (1956) since aceto-
phenone and benzophenone gave maximum absorption at 446 nm and 430 nm
respectively. They confirmed the earlier findings that aldehyde hydrazones
fade more rapidly than ketonic hydrazones.
Sanders and Schubert (1971) modified the Lappin and Clark (1951) method
by carefully acidifying the mixture of the aqueous carbonyl compound and
the 2,4 DNPH methanol with 1.0 ml of 0.01 N HC1 to result in a final pH
of 2.5-2.7. After heating at 80°C for 10 minutes the sample was cooled
for 10 minutes after which 5 ml of methanol KOH was added. After 10
minutes the absorbance was measured at 515 nm. Sugars did not interfere
176
-------
significantly in the pH range specified but did interfer at low pH
because the conversion of the cyclic hemiacetal to acyclic aldehyde is
accelerated. Aldehydes gave a relative constant absorption when the
hydrazone was formed between pH 2.5 and 2.6, but the ketonic hydrazones
gave better absorbances at pH 2.5 instead of 2.7 due to the slower rate
of formation at the higher pH. The optimum heating temperature was 80°C
since at 100°C, as recommended by Lappin and Clark (1951), sugars caused
a substantial interference. A 10 minute boiling time at 80°C is
sufficient to form hydrozones with ketones and aldehydes while only a
fraction of the sugars have yet reacted.
PROCEDURE
The test is based on the formation of a red colored substance on the
addition of base to the 2,4 dim'trophenylhydrazone in an alcohol solution
(Lappin and Clark, 1951), and can be used as a colorimetric method for
detecting carbonyl compounds in a water solution.
A saturated solution of 2,4-DNP was made up in 1.0 N HC1.
It was found that the 2,4 dim'trophenyl hydrazone precipitated from the
water layer during the test and that no quantitative colorimetric results
were obtained indicating that the presence of methanol is required to
increase its solubility. It was also found that the methanol contained
too many impurities to give reliable results. One liter of methanol was
therefore distilled from a 2,4-DNP-HCl mixture, in which the 2,4-DNP was
recrystallized twice. The procedure is now as follows:
a) Add 1 ml sample
b) Add 1 ml saturated 2,4-DNP in methanol
c) Add 1 drop concentrated HC1 (pH 2.5 - 2.7)
d) Heat at 50°C for 30 minutes
e) Cool, add 5.0 ml methanolic KOH
f) Read at 480 nm after 15 minutes
177
-------
INCUBATION TIME
The effect of the elevated temperature during hydrazone formation was
first studied. A precipitate was indeed formed, however too much methanol
was found to evaporate. A lower incubation temperature of 50°C instead
of the 80°C as recommended by Sanders and Schubert (1971) was therefore
selected. The evaluation of the length of the incubation period with
15 mg/1 pentanone showed that the reaction is essentially complete after
a few minutes (Figure 35a), and that the recommended 30 minute period
can be reduced if necessary. These results with regard to the length of
the incubation time confirm the findings of Sanders and Schubert (1971).
FADING OF COLOR
The fading of the color beyond the recommended waiting time of 15 minutes
after the color formation was found to be linear with time and acetone
showed a decrease of 16 percent of the original color after 120 minutes
(Figure 35b). All the leachate samples showed fading of the color.
Using different test compounds it was noted that the fading was more
rapid for aldehydes than for ketones, which emphasizes the importance
of obtaining spectrophotometric readings at a standardized time, i.e.
15 minutes.
ABSORBANCE OF DIFFERENT COMPOUNDS
The absorbances of the different compounds are shown in Figure 36. The
results contradict the results of Lappin and Clark (1951) but confirm the
findings of Jones et, al_. (1956) in that aldehydes generally give higher
absorbances than ketones and that aromatic structures also enhance the
absorbance. Although insertion of functional groups in the benzene ring
was reported to increase the absorbances, such was not found for piperonal
which has two oxygen groups in the meta- and paraposition connected with
a methyl group. When a methyl group is attached to an aromatic aldehyde
group such as in acetophenone, the absorbance decreases due to the presence
of the electron donating character of the methyl group. Acids having a
carbonyl oxygen in the carboxyl group such as valeric and salicylic acid
178
-------
p-pentonone
(15 mg/*.)
10 20 30 40 50 60
Incubation Time (min) At 50° C
70
E
c
O
00
1,0
a>
u
O
.a
§ 0,5
X)
15 Minutes
30 Minutes
60 Minutes
120 Minutes
0
10
20
30
40
50
Acetone Concentration (mg/X)
Figure 35. Effect of (a) Incubation Time on the Color Formation
of Pentanone with 2,4 dinitrophenylhydrazine (b) effect of length
of waiting time after the color formation on fading of color
of acetone
179
-------
1,5
1,0
g
0)
u
o
X}
§
0,5
Benzaldehyde
Piperonal
Acetophenone
0,1 0,2
Molar Concentration, mmole/I
Figure 36. Molar Absorbances of Different Carboxyl Compounds
180
-------
did not give any response in the test. Formic acid was the only acid
that interferes but formic acid is present in the aldehyde form in only
a small percentage. Dextrose at a concentration of 1000 mg/1 did not
give any response with the test. The spectraof the different compounds
in alkaline media show that the wavelength of 480 nm as selected by
Lappin and Clark (1951) is only valid for aromatic compounds (Figure 37)
and p-dimethylaminobenzaldehyde did indeed have a maximum at 480 nm. In
the absence of functional groups attached to the benzene ring the
absorbance maximum even shifts to lower wavelengths, and maxima for ben-
zaldehyde and acetophone were 464 and 458 nm respectively. In accordance
with Jones et al_. (1956) only a small shoulder was noted near 550 nm
unless the aldehyde group attains a more aliphatic character such as with
acetophone which has a noticable shoulder at 560 nm. Much higher wave-
lengths were observed for quinones which absorb at 540 nm and compounds
with two aldehyde groups such as glyoxal which forms a cyclic hydrazone
The aliphatic aldehyde compounds all have absorption
maxima near 430 nm (Figure 38a) which maximum tends to shift to higher
wavelength with increasing aliphatic chain length. In accordance with
Jones (1956) it was noted that aliphatic ketones have a more pronounced
shoulder at 540 nm than aliphatic aldehydes.
The spectra of the different molecular weight fractions in leachate are
shown in Figure 38b and one can conclude that they resemble aliphatic
aldehydes. The shoulder at 540 nm is less pronounced than found for
propanal but is similar to that of valeraldehyde. The conclusion,that
the carbonyl group is present as an aldehyde instead of a ketone.is
important since oxidation converts the carbonyl into a carboxyl group
which oxidation cannot take place with ketones.
181
-------
n>
00
ro
cr
tn
o
-i
•a
o
3
CO
o
sr _
o T
-h —*
» $
- 3
1.8
1.6
1.4
1.2
o
01
o-
o
r>
o
-a
O
Q8
Q6
0.4
Q2
0
Benzaldehyde
\ /-p-dimethylamino
""•^ ' Benzaldehyde
Glyoxal
400
500 600
Wavelength , nm
700
-------
00
GO
co
oo
cr
to
o O
o -I
3-0
-O r*
O -"•
C O
3 3
O.
CO
3 O
O. H-
r-3
(0
cv o
O -h
o> o
rt- -<•
n> -h
-h
-n n>
o> n>
o 3
o 3>
3 —«
CO _i.
T3
O>
O
O
CU
-s
cr
o
Q8
Q6
I
Q2
0
1.0
1
o
08
"
Q4
Q2
-Formic Acid
-Valeraldehyde
-Acetone
-Propanol
500 mw UF Retenlate
lOjOOOmw UFRetentate
500 mw UF Retenlate
10,000 mw UF Permeate
500 600
Wavelength, nm
700
-------
REFERENCES
Ahmad, M. et^ al_., "Radiolysis of Ethylene Glycol Aqueous Solutions,"
Journal Chemical Society, ]_3_, 945 (1968).
Barrenscheen, H. K. and Dreguss, M., "A Colorimetric Micromethod to
Determine Methyl Glyoxals," Biochemische Zeitschrift, 233, 305 (1931).
Barta, L., "Method to Determine Small Quantities of Furfural, Methyl
Furfural and Oxymethyl Furfural," Biochemische Zeitschrift, 274,
212 (1934).
Bel isle, J., "Non Aqueous Spectrophotometric Determination of Carbonyl
Functions," Analytica Chimica Acta, 43, 515 (1968).
Fiacchia, M. F. et^al., "A Method for Sampling and Determination of
Organic CarbonyT Compounds in Automobile Exhaust," Environmental
Science and Technology, ]_, 915 (1967).
Hanna, J. G., "Chemical and Physical Methods of Analysis," in "The
Chemistry of the Carbonyl Group," Ed. Patai S., Interscience Publ.
London (1966).
Hanna, J. G., and Siggia, S., "Carbonyl Functional Groups and Derived
Functions," in "Treatise on Analytical Chemistry," Ed. Kolthoff,
•I. M. and Elving, P. J., II, Vol. 13, Interscience, New York (1966).
Iddles, H. A. and Jackson, C. E., "Determination of Carbonyl Compounds by
Means of 2,4-dinitrophenylhydrazine," Industrial and Engineering
Chemistry. Anal. Ed.. 6_, 454 (1934).
Jones, L. A., "Spectrophotometric Studies of Some 2,4 Dinitrophenylhydrazines,"
Analytical Chemistry. 28., 191 (1956).
Jordan, D. E., and Veatch, F. C., "Spectrophotometric Determination of
Trace Concentrations of Carbonyl Compounds," Analytical Chemistry,
3i, 120 (1954).
Lappin, G. R., and Clark, L. C., "Colorimetric Method for Determination
of Traces of Carbonyl Compounds," Analytical Chemistry, 23_, 41 (1951).
Lehman, F. H., "Spectrophotometric Determination of Carbonyl Oxygen,"
Analytical Chemistry. 30, 972 (1958).
Mathewson, W. E., "The Combination of Fractionation with Spectrophotometry
in Proximate Organic Analysis," Journal American Chemical Society,
42, 1279 (1920).
184
-------
Mendelowitz, A., and Riley, 0. P., "The Spectrophotometric Determination
of Long Chain Fatty Acids Containing Ketonic Groups," Analyst, 78,
704 (1953).
Meyboom, P. W., "Separation and Identification of the 2,4 Dinitro
Derivative of Unsaturated Aldehydes and Methylketones by Using Thin
Layer Chromatography," Fats, Soaps and Paints, 70, 477 (1968).
Roberts, J. D., and Screen, C., "Separation of 2,4 Dinitrophenylhydrazones
by Chromatographic Adsorption," Industrial and Engineering Chemistry,
Anal. Ed., J8., 335 (1946).
Sanders, E. B., and Schubert, J., "Spectrophotometric Analysis of Carbonyl
Compounds in the Presence of Carbohydrates without Prior Separation,"
Analytical Chemistry, 43, 59 (1971).
Schepartz, A. I., and Danberg, B. F., "Flavor Reversion in Soybean Oil. V.
Isolation and Identification of Reversion Compounds in Hydrogenated
Soybean Oil," Journal American Oil Chemistry Society. Z7_, 367 (1950).
Soukup, R. J., et^ al_., "Gas Chromatographic Separation of 2,4 Dinitrophenyl'
hydrozone Derivatives of Carbonyl Compounds," Analytical Chemistry.
36, 2255 (1964).
Weeks, B. M., et_ ail_., "Reactions of Alanine with the Reducing Species
formed in Water Radiolysis," Journal Physical Chemistry. 69, 4131
(1965).
185
-------
THE DETERMINATION OF AROMATIC HYDROXYL GROUPS WITH
THE FOLIN-DENIS METHOD
THE FOLIN-DENIS TEST
The phenolic hydroxyl group concentration has been determined in humic
acids by subtracting the value found for the carboxyl group (calcium
acetate method) from the value found for the total acidity (bariumhydroxide
method) which procedure is commonly used by soil scientists (Schnitzer
and Khan, 1972).
When relatively small amounts are available, only colorimetric methods
are applicable. The method used in this study was the Folin-Denis (1912)
test which test is also listed in Standard Methods (1971) for the deter-
mination of lignin-tannin like material. The Standard Methods (1971)
adapted modifications introduced by Berk and Schroeder (1942). The blue
color is produced after addition of alkali to a mixture of the Folin-Denis
reagent and the sample,which color is measured at 700 nm. The reaction
mechanism is based on the oxidation of the phenolic hydroxyl group and
the reduction of the phospho molybdotungstic acid. The method was tested
by Smit jit a_l_. (1955) for different aromatic compounds having an hydroxyl
group. He noted that more than one hydroxyl group in the benzene ring
did not give a doubling of the color absorbance. He also concluded that
the test is very accurate but that the standard should reflect the tested
compound as closely as possible, since otherwise erroneous conclusions
are drawn with regard to measured concentrations.
FOLIN CIOCALTEAU TEST
The Folin-Denis test was improved by Folin-Ciocalteau (1927) to eliminate
a precipitate that sometimes formed during the test. The precipitate con-
sisted of the sodium salt of the tungsten and molybdenum complex and it
was found that the formation could be prevented by addition of the more
186
-------
soluble Li salts. The modified test also forms the basis for the Lowry
protein test. The Folin-Ciocalteau reagent was used by Singleton and Rossi
(1965) to measure phenolic compounds and by Kloster (1974) to measure
tannins in water. The latter showed that reducing substances that are
also present in the sample will increase the measured value. The largest
interference is caused by ferrous iron and 2 mg/1 will give a color
equivalent of 1 mg/1 tannic acid. The interferences are similar as those
encountered in the phosphate test in which phosphate forms a complex
with the molybdic acid which is then reduced with SnCl~ or ascorbic acid
2+
to form a blue color. Interferences for the phosphate test include Fe ,
sulfide and nitrite.
PROCEDURES
Since no precipitation was observed with the Folin-Denis reagent it was
used instead of the more complicated Folin-Ciocalteau test. The procedure
is identical to the tannin-lignin test as listed in Standard Methods.
Since most of the organic fractions measured showed a slight leveling off
at 620 nm (Figure 39), this wavelength was selected for all measurements.
RESULTS
To obtain a clearer understanding of the reaction, different aromatic
compounds having more than one hydroxyl group in the benzene ring, were
tested. As shown in Figure 40, the intensity of the colored product
depends on whether the location of the OH groups is in the ortho, meta,
or para position. When the ring has three OH groups, such as in gallic
acid one of the groups may not participate in the reaction. Further tests
are necessary to gain a better understanding of the mechanism of this
test. It can be concluded that this is a sensitive test for phenolic
hydroxyl compounds, however, the standard compound has to reflect the
composition of the humic and fulvic acids as closely as possible.
187
-------
0,15
0,10
o
JQ
i-
O
V)
JO
0,05
Dupage
Fulvic Acid
Fraction
Tannic Acid (5mg/J?)
500 600
Wavelength, nm
700
Figure 39. Spectrum of the Fulvic Acid Fraction of the Dupage Landfill
Leachate and the Standard Tannic Acid
188
-------
1,0
246
Concentration , mg/J?
Figure 40. Absorbence of Different Aromatic Hydroxyl Compounds
189
-------
REFERENCES
Berk, A. A. and Schroeder, D., "Determination of Tannin Substances in
Boiler Water," Ind. Engin. Chem. Anal. Ed., ]£, 456 (1942).
Folin, 0. and Ciocalteau, V., "On Tyrosin and Tryptophane Determinations
in Proteins," Journal of Biological Chemistry. 7£, 627 (1927).
Folin, 0. and Denis, W., "On Phosphotungstic-Phosphomolybdic Compounds as
Color Reagents," Journal of Biological Chemistry. ]2, 239 (1912).
Kloster, M. B., "The Determination of Tannin and Lignin," Journal American
Water Works Association. 66_, 44 (1974).
Schnitzer, M. and Kahn, S. U., Humic Substances in the Environment, Marcel
Dekker, Inc., New York (1972).
Singleton, V. L. and Rossi, 0. A., "Colorimetry of Total Phenolic with
Phosphomolybdic-Phosphotungstic Acid Reagent," American Journal of
Enology and Viticulture, ]j>, 144 (1965).
Smit, C. J. B., et al., "Determination of Tannins and Related Polyphenols
in Foods," Analytical Chemistry. 27., 1159 (1955).
190
-------
THE DETERMINATION OF CARBOHYDRATES WITH THE ANTHRONE TEST
THE ANTHRONE TEST
The most common colorimetric methods of determining carbohydrates are the
phenol sulfonic acid method and the anthrone method. Because the former
method procuces a yellow color the reaction might be interferred with by
the yellow brown color of the leachate; the anthrone test on the other
hand gives a blue-green color reaction. The color in this test has been
attributed to the reaction product of 5-hydroxymethyl furfural or furfural
and anthrone (9,10-dihydroxyl-9-oxoanthracene) a reduction product of
anthra quinone in an acid solution (Dreywood, 1946). Glucose can further
decompose to laevulinic acid, black humic products, formic acid and
formaldehyde (Amihoff, 1970). The different sugars display a wide range
of color intensities, and in this respect the test is inferior to the
phenol-sulfuric acid test in which the different sugars give a more
identical response. It is therefore of great importance to select a
standard that closely resembles the building units of the polysaccharide
to be determined. The test has been applied successfully in the determin-
ation of hexosen, pentosen, deoxy sugars, hexuloses, heptuloses, glycogen,
and cellulose(Seifer et al_, 1950; Black, 1951). Some compounds will
interfere slightly; for example, fatty acids, some proteins (Peterson
and Rose, 1951) and a high salt content (Lanner, 1960). No interferences
were given by 1 M Na, 1 M K, 1 M P04, 1 M NH3, 0.1 M Cl, 0.05 M Mn,
0.01 M Ca and 0.01 M Ba (Koehler, 1952). Precautions have to be taken
when particulate matter is analyzed (Biggs and Wetzel, 1968). When
cellulose or glycogen like material has to be determined, the solution
should be boiled in acid for a relatively long period as compared to
single sugars. Boiling during an extensive period, however, polymerizes
the furfural units into brown humic complexes which will interfere with
the analysis.
191
-------
PROCEDURES
The tests reported here have been conducted according to the recommendations
of the International Biological Program, (Golterman and Clyroo, 1969), in
which 5 ml of sample is boiled in 25 ml 58 percent sulfuric acid containing
140 mg/1 anthrone reagent for 15 minutes. The anthrone reagent is obtained
by dissolving 0.1 g of anthrone in 500 ml concentrated H^SO. to which 200
ml 1^0 and 1 g of thiourea is added as an oxidant. The boiling time of
15 minutes was selected because higher readings are obtained when heated
longer than that period due to a brown color formation which absorbs at
540 nm instead of 620 nm.
RESULTS
When the test was run on the whole leachate and the different molecular
weight fractions, all resulted in a blue color and the absorption spectra
were similar to that of glucose, indicating that glucose is a good standard
to use. The color was less well developed in the 500 MW permeate as com-
pared to other membrane fractions as it had traces of a brown color
corresponding to a shift in the absorption maximum from 620 nm to 540 nm,
which change became stronger after 15 minutes boiling. All fractions
showed a continuous increase in absorbance with increasing boiling time
(Figure 41), probably due to the formation of brown humin products. The
most rapid increase was observed for glucose when the boiling time was
increased from 15 to 30 minutes. This resulted in lower apparent concen-
trations for each of the fractions when expressed as glucose equivalent at
the 30 minute as compared to the 15 minute digestion time. The values
after 4 hours boiling were higher than the 15 minute values when expressed
as glucose equivalent since the absorbence of each of the fractions would
increase while the values for glucose would remain constant. Since a
15 minute boiling time would result in a minimum change toward a brown
color, this period was therefore selected as the optimum boiling period.
Aromatic hydroxyl compounds that resemble the composition of the low
molecular weight fulvic acid fraction did not cause significant interference
in the test (Figure 42). In combination with glucose however, they may
192
-------
03
~ 02
^—
Q>
O
C
O
JO
k.
O
JQ
0.1
/
p
1 f J
i
T
/
r
A'
/
b
i i ^r*^ ' '
/ 10,000 MW
^/ Froction9b
^
^^—^^^^ ^^^™ ^^^_ ^^^B ^^^^^B ^^^ »••
' /^ 100 Glucose
/
>IOODOOMW
Fraction lOb
^
^^^^ ^^ ^^^
^^^^
s"' 50,OOOMW|^phadex
^-"ITI- D-
^•^— """ 3000 MW Sephadex _
KVXX)-^
lOoToOOMW
Fraction lib
*- -4-
i i i 1
100 200 300 400
Hydrolysis Time , minutes
Figure 41. Effect of Hydrolysis Time on Absorbence of Different
Molecular Weight Fractions in Leachate
193
-------
LJ
600 1000
Concentration, mg/l
1500
Figure 42. Interference in Carbohydrate Analysis by Aromatic
Hydroxyl Compounds
194
-------
cause formation of brown products (Feeny, 1969) and may have partially
caused the brownish color development in the 500 MW UF permeate.
REFERENCES
Amihoff, D. et^ al_., "Analytical Methods for Carbohydrates," The Carbohydrates,
W. Pigma, et al_., ed., Vol. I IB, Academic Press, New York, 1970.
Black, H. C., "Determination of Sodium Carboxyl Methyl Cellulose in
Detergent Mixtures," Analytical Chemistry. Vol. 23, 1792 (1951).
Biggs, R. B. and Wetzel, C. D., "Concentration of Particulate Carbohydrate
at the Halocline in Chesapeake Bay," Limnology and Oceanography, Vol.
13, 169 (1968).
Dreywood, R., "Qualitative Test for Carbohydrate Material." Indr. Engin.
Chem. Anal. Ed.. J8., 499 (1946).
Feeny, P., "Spectrophotometric Method for Rapid Determiantion of Oak Leaf
Condensed Tannin," Analytical Chemistry, 41_, 1347 (1969).
Golterman, H. L. and Clymo, R. S., "Methods for Chemical Analysis of Fresh
Waters," International Biological Programme Handbook No. 8, Blackwell
Scientific Publications, Oxford (1969).
Koehler, L. H., "Differentiation of Carbohvdrates by Anthrone Reaction Rate
and Color Intensity," Analytical Chemistry. 24, 1976 (1952).
Launer, H. F., "Determination of Carbohydrates," Methods in Carbohydrate
Chemistry. R. L. Whistler et al. eds., Vol. I, Academic Press, New
York, 1960.
Peterson, R. and Rose, D., "Estimation of Glycogen by the use of Anthrone,"
Canadian Journal of Technology, 29_, 317 (1951).
Seifert, S., e_t al_., "The Estimation of Glycogen with the Anthrone Reagent,"
Arch, of Biochem.. 25, 191 (1950).
195
-------
THE DETERMINATION OF PROTEINS WITH THE LOWRY, NINHYDRIN
AND KJELDAHL METHOD
THE LOWRY METHOD
The Lowry method is the most commonly employed protein test in laboratory
determinations. The test was developed by Folin and Ciocalteau (1927) and
significantly modified by Lowry ert al_. (1951) who added the copper reagent
used in the biuret reaction. The test is based on the reduction of phospho-
tungstic and phosphomolybdic acid and the concurrent oxidation of the
phenolic hydroxyl group of aromatic amino acids that usually make up 10
percent of any protein, resulting in a reaction product that gives a blue
color. Miller (1959) modified the test by using an elevated temperature
(50°C), a smaller volume of a more concentrated alkaline copper reagent,
and a larger quantity of a more dilute phenol reagent. Schacterle and
Pollack (1973) simplified the procedure even further by using only two
reagents.
The amino acids mainly responsible for the reaction are tyrosine, tryptophan
and the purines, guanine, xanthine and uric acid. A recent evaluation of
the Folin test by Shao-Chia and Goldstein (1960) reported that although
a large part of the color is caused by the reaction of the phenolic hydroxyl
in the aromatic amino acids, any peptide bond will yield some color
depending on the sequence of certain amino acids. They found that some
amino acid sequences have a larger impact on,the color than others.
Hydrolysis reduced the color of albumin by more than two-thirds as it
eliminated the protein bonds. Cleavage of the disulfide bonds in insulin
by oxidative reactions also gave a loss of about one-third of the total
color. Ji (1973) reviewed the interferences of the Lowry test and stated
that some carbohydrates, lipids, di- and tri-carboxylic acids and single
amino acids like glycine and glycylglycine interfere with the test. Other
limitations of the test are that the amount of color varies with different
196
-------
proteins while other aromatic compounds present in the sample contribute
to the color. The main advantages of the test are [1] that it is as
sensitive as the Nessler reagent but does not require prior digestion
of the sample, [2] it is several fold more sensitive than the ninhydrin
method, [3] it is 10 to 20 times more sensitive than measurement of the
UV adsorption and a hundred times more sensitive than the biuret method
(Lowry e_tal_., 1951; Young, 1963).
In the water pollution area the Folin test has been used extensively by
Orme-Johnson and Woods (1964), Walters e_t a_l_. (1968) and Rebhun and Manka
(1971). The latter used it for determination of proteins in treatment
plant effluent. Phenolic hydroxyl groups in humic acid in the effluent,
however, will greatly interfere with the test. Because high protein
values were initially obtained in this study, as compared to the protein
concentrations calculated from the organic nitrogen values multiplied by
6.25, the Lowry test was evaluated with respect to interferences from
aromatic hydroxyl compounds.
When increasing amounts of tannic acid, containing numerous aromatic
hydroxyl groups, were added to a 1:100 and 1:200 diluted leachate sample,
increasing apparent amounts of proteins were measured with the Lowry
test (Figure 43a). It was found that 1 mg/1 tannic acid gave a response
identical to 3.8 mg/1 proteins, clearly illustrating that this protein
test is interfered with by the presence of aromatic hydroxyl groups.
When the values obtained with the Lowry test using bovine albumin as a
standard and the aromatic hydroxyl test using tannic acid as a standard
(Figure 43b) a similar result was obtained as 1 mg/1 tannic acid corres-
ponded to 3.8 mg/1 proteins, indicating that a relatively low protein
content reactive to the Lowry test is present in leachate.
THE NINHYDRIN METHOD
A test which is very accurate but several times less sensitive than the
Lowry test uses the reaction between ninhydrin (triketohydrindenehydrate)
and the individual amino acids. Moore and Stein (1948) tested the
197
-------
-5
0 5
Amount Of Tannic Acid Added, mg/Jt
10
v>w
_o
f*
o
H 500
?V»00
&E
ol300
o
•§, 200
X
u
'•§ 100
E
2
i i y \
/
o /
~ f~ *™
/
x/c
/ D
y
/
~ /
_ ^° (b)
/ O
II
ijllr. «rr 1 1 1
1000
2000
3000
Protein As Measured With The Lowry Test, mg/l
Figure 43. Effect of Standard Additions of Tannfc Acid in Leachate on the
Outcome of the Lowry Test (a) and Relation Between Results of the
Aromatic Hydroxyl Test and Lowry Test as Determined on Different
Leachate Samples
198
-------
reaction extensively to obtain a reliable method for the automatic analysis
of amino acids. Yemm and Cocking (1955) described the first step in the
reaction as the oxidative deamination of amino acid with the formation of
ammonia and the reduction of ninhydrin to hydrantin. The ammonia then
condenses with the hydrindantin to form a colored complex.
The test has been used in Painter and Viney (1959) to measure free amino
acids in sewage. Stevenson and Cheng (1970) used the ninhydrin method after
extensive acid hydrolyses to determine amino acids in sediment samples. A
sample is hydrolyzed with 6 N HC1. After removal of HC1 the solution is
made alkaline and ammonia and amine are removed. Sufficient 6 N HC1 is
then added to dissolve metallic hydroxides after which the amino acids are
analyzed according to the method of Yemm and Cocking (1955), with the
exception that excess sodium citrate is applied to chelate metals. The
standard amino acid, usually leucine, is subjected to an identical diges-
tion step which generally reduces its concentration by 5-10 percent. It
was shown that some amino acids, especially tryptophane, are less resis-
tant to digestions than other amino acids.
PROCEDURES OF THE NINHYDRIN METHOD
After several trials the following procedure was selected as the most
satisfactory one:
1. Add 5 ml sample to 25 ml of 6 N HC1 and reflux for 20 hours.
2. Filter and wash residue.
3. Adr4 2-3 drops phenolphthalein, titrate to pink color with
6 N NaOH.
4. Boil to remove NH^.
5. Add 6 N HC1 to dissolve metal hydroxide and dilute to 100 ml
with boiled water.
6. Take a 1-ml aliquot and add 1 ml sodium citrate solution.
7. Add 4 ml of freshly prepared ninhydrin reagent mix as described
by Yemm and Cocking (1955).
8. Place in boiling water bath for 3 minutes.
199
-------
9. Remove from bath and immediately add 10 ml of 47.5 percent
ethanol water mixture.
10. Cool to room temperature and read absorbence at 570 m.
RESULTS OF THE NINHYDRIN METHOD
The results of the test with different leachate samples were found to be
accurate and reproducible; a duplicate test with the U of I leachate, for
example, showed values of 2420 and 2440 mg/1.
The evaluation of the digestion time showed that high readings are even
obtained without hydrolysis (Figure 44a) indicating that most proteins are
present in leachate as single amino acids. Analysis with an Automated
Technicon Amino acid analyzer showed that most amino acids were present as
ornithine,lysine and valine confirming the high organic nitrogen values
measured in the 500 MW UF permeate. Since these amino acids do not have
aromatic hydroxyl groups, they will not register with the Lowry test. A
digestion time of 20 hours was found to be suitable (Figure 44).
Comparison of the protein concentration determined with both the Kjeldahl
method and the ninhydrin test showed that the results obtained with each
test agree well with each other (Figure 44b). The protein value obtained
with the Kjeldahl method was calculated by multiplying the value for N as
obtained after digestion and distillation by a factor of 6.25 (Bunch et a 1.,
1961).
200
-------
2000
2
JZ
1000
0>
e
0_
I
1
12 24
Boiling Time, hr
6 bUUU
&
O
it •<
"SI
o ^4000
3 3
o .9
0 >
II
o i_
|i2000
u
c
o
o
c
~£
O f\f
/•"
~" /
/
0 /
X
/ ^-100 %, Recovery
- o X
f/
_
n/"
Q'
X 1
2000
4000
6000
Protein Concentration Determined With Ninhydrin
Method, mg/J?
Figure 44. Effect of Digestion Time on Outcome of Ninhydrin Protein Test
(a) and Relation Between Protein Concentration as Measured
with Kjeldahl Method and Ninhydrin Method Using Digestion
201
-------
REFERENCES
Bunch, R. L., e_t aJL, "Organic Materials in Secondary Effluents," Journal
Water Pollution Control Federation, 33, 122 (1961).
Folin, 0. and Ciocalteau, V., "On Tyrosin and Tryptophane Determinations
in Proteins," Journal of Biological Chemistry, _73_, 627 (1927).
Ji, T. H., "Interference by Detergents, Chelating A-ents and Buffers
with the Lowry Protein Determination," Analytical Biochemistry, 52^
517 (1973).
Johnson Orme, W. H. and Woods, C. E., "Colorimetric Determination of
Proteins and Free Amino Acids," Mater and Sewage Works, 111. 339
(1964).
Lowry, 0. H., e_t al_., "Protein Measurement with the Folin Phenol Reagent,"
Journal of Biological Chemistry, 193, 265 (1951).
Miller, G. L., "Protein Determination for Large Number of Samples,"
Analytical Chemistry. 3j[» 964 (1959).
Moore, S. and Stein, W. H., "Photometric Ninhydrin Method for Use in the
Chromatography of Amino Acids," Journal of Biological Chemistry, 176,
367 (1948).
Painter, H. A. and Viney, M., "Composition of a Domestic Sewage," Journal
of Biochemical and Microbiological Technology and Engineering, 1,
143 (1959).
Rebhun, M. and Manka, J., "Classification of Organics in Secondary
Effluents," Environmental Science and Technology, 5_, 606 (1971).
Schacterle, G. R. and Pollack, R. L., "A Simplified Method for the
Quantitative Assay of Small Amounts of Protein in Biological Material,"
Analytical Biochemistry. 51, 654 (1973).
Chia, S. and Goldstein, A., "Chromogenic Groupings in the Lowry Protein
Determination," Biochemical Journal, 7J5, 109 (1960).
Stevenson, F. J. and Cheng, C. N., "Amino Acids in Sediments: Recovery by
Acid Hydrolysis and Quantitative Estimation by a Colorimetric
Procedure," Geochtmica et Cosmochimica Acta, 34_, 77 (1970).
Walters, C. F., e_t al_., "Microbial Substrate Storage in Activated Sludge,"
Journal Sanitary Engineering Division, ASCE, 94, 257 (1968).
Yemm, E. W. and Cocking, E. C., "The Determination of Amino Acids with
Ninhydrin," Analyst, 80, 209 (1955).
202
-------
Young, E. G., "Occurrence, Classification, Preparation and Analysis of
Proteins," Comprehensive Biochemistry, M. Florkin and E. H. Stotz,
eds., Vol. 7, Elsevier Publishing Corp., Amsterdam (1963).
203
-------
THE DETERMINATION OF EXTRACTABLE ORGANICS USING
HEXANE AND BUTANOL
THE SOLVENT EXTRACTION METHOD
The solvent extraction technique has been used to fractionate organics
according to their polarity. Ethanol is for example used to obtain the
hymatomelanic acid from the pH 2 precipitated humic acid. Several inves-
tigations in the soil science area have used extraction techniques using
benzene-methanol, chloroform-methane!, acetyl acetone, hexa methylene,
tetramine, dodecylsulfate, urea, formic acid, phenol and acetone to obtain
humic materials (Schnitzer and Kahn, 1972). The solvent extraction scheme
of Bunch et al_. (1961) as modified by Rubhun and Manka (1971) was initially
used in the present study and methanol was used to extract the fulvic
acids from the solution. Unrealistic high values however were obtained
for fulvic acid concentrations. A more in-depth evaluation showed that
when the permeate of the 10,000 MW ultrafilter was extracted, both fulvic
acids and short-chain free volatile fatty acids were removed from the
solution. The latter compounds were detected when the residue of the
dried butanol layer was resuspended and GC analysis showed substantial
amounts of free volatile fatty acids.
EXTRACTION OF MODEL COMPOUNDS
To study more precisely how many other compounds besides fulvic acid are
extracted, a butanol extraction was performed on model compounds that
resemble the organics that are expected to be present in leachate such as
fulvic acid (tannic acid), carbohydrates (starch), and proteins (bovine
albumin). The results in Figure 45 show that the butanol extraction of
tannic acid removed most of the compound from the water layer. A compari-
son of the gravimetric and the colorimetric determination showed that the
latter method resulted in lower values as compared with the former method.
The extraction of the bovine albumin is less effective than that of
tannic acid, however since it is extracted to a considerable extent it
204
-------
1500
ro
CD
01
o»
£ 1000
u
s
"x
u
500
Tannic Acid Removed (Gravimetric)
^/.
Tannic Acid Removed (Colorimetric)
Bovine Albumin (Color!metric)
Starch (Gravimetric)
I
Figure 45.
1000 2000
Initial Concentration (mg/A)
Butanol Extraction of Tannic Acid, Starch and Bovine Albumin
-------
will interfere in the gravimetric determination of the fulvic acid as
obtained in the dried butanol residue. The extraction of starch is
almost negligible and will not result in any interference.
A second extraction test was conducted with a mixture of free volatile
fatty acids as encountered in fresh leachate (formic acid 2000 mg/1,
acetic acid 5000 mg/1, propionic acid 1000 mg/1, butyric acid 5000 mg/1,
isobutyric acid 1000 mg/1, valeric acid 1000 mg/1, isovaleric acid 1000 mg/1
and caproic acid 6000 mg/1). The extraction was performed similarly to
that employed in the leachate samples; a 100 ml aliquot is extracted with
two 100 ml of hexane. Following the hexane extraction the sample is
acidified to pH 2 to obtain a possible precipitate of humic acid, after
which the water layer is extracted with butanol. The results of both
tests show that large amount of free volatile fatty acid are extracted
into the butanol and hexane layer (Figure 46 and 47). After the drying
of the hexane and butanol layer at 60°C for 2 days the residues were
redissolved in 100 ml water and the TOC was measured. Because of the
extensive evaporation of the fatty acids only a fraction was left in the
residue as shown in Table 19. The TOC of the redissolved hexane
residue was found to be proportional to the initial acid concentration
in the water layer while the subsequent butanol extraction removed more
fatty acids at the lowest fatty acid concentration. The results also
show that the accuracy of the residue determination can vary by more
than 50%. For these reasons the extraction method was only applied
using the hexane to determine oils or lipids and butanol to detect fulvic
acids in the high molecular weight retentate. Application of the extraction in
the 500 MW UF permeate will lead to substantial interferences due to
the presence of free volatile fatty acids in leachate.
206
-------
ro
o
3000 -
o»
E
«
c
o
X
Q>
X
H 2000 —
•o
x
LU
Z 1000 —
o
i
Isovaleric Acid
aleric Acid
1000 2000 3000
Initial Acid Concentration (mg/«.)
4000
Figure 46. Hexane Extraction of Free Volatile Fatty Acid Mixture
-------
1000
2000
Acid Present Before Extraction (mg/JO
Figure 47. Butanol Extraction of Acidified Free Volatile Fatty Acid Mixture
-------
Table 18
TOC In Redissolved Dried Residue of the Hexane and Butanol
Used to Extract a Free Volatile Fatty Acid
Solutions the Residue Was Redissolved in 100 ml Distilled Water
TOC of Redissolved TOC of Redissolved
TOC of Water Layer Hexane Layer Butanol Layer
(mg/1) (mg/1) (mg/1)
Test 1 Test 2 Test 1 Test 2
9318 (initial stock)
465 (1:2)
2329 (1:4)
167
61
1
145
98
4
120
120
190
120
180
160
REFERENCES
Bunch, R. L. et al_., "Organic Materials in Secondary Effluents," Journal
Water PoTTution Control Federation. 33, 122 (1961).
Rebhun, M. and Manka, G., "Classification of Organics in Secondary
Effluents," Environmental Science and Technology, 5_, 606 (1971).
Schnitzer, M. and Kahn, S. N. "Humic Substances in the Environment,"
Marcek Dekker, Inc., New York, 1972.
209
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
\. REPORT NO.
EPA-600/2-77-186a
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Evaluation of Leachate Treatment
Volume I
Characterization of Leachate
5. REPORT DATE
September 1977 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Edward S.K. Chi an
Foppe B. DeWalle
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Civil Engineering
University of Illinois at Urbana-Champaign
3217 Civil Engineering Bldg.
Urbana, Illinois 61801
10. PROGRAM ELEMENT NO.
PE 1DC618
11. CONTRACT/GRANT NO.
68-03-0162
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory-_cin.
Office of Research and Development
U.S, Environmental Protection Agencv
Cincinnati, Ohio 45268
OH
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Volume I of a two-volume final report.
Project Officer: Dirk Brunner (513-684-7871)
IB.ABSTRACT An extensive analysis of organics and inorganics present in leachate from
landfills located in different regions of the United States was performed to evaluate
methods for treating leachates. Bench-scale performance data and evaluation of leach-
ate treatment methods are reported in Volume II. Gross properties of the leachate
quality varied, greatly between different landfills. Membrane ultrafiltration, gel per-
meation chromatography, and specific organic analysis were used to separate different
molecular weight fractions and to determine the main classes of organics and associ-
ated functional groups. The majority of the organics were able to permeate a 500 MW UF
membrane indicative of their low molecular weight. Membrane fractionation and organic
analysis of leachate samples collected from different landfills generally showed a de-
crease of the free volatile fatty acid fraction with increasing age of the fill. Bio-
logical degradation studies showed the sequential removal of different classes of
organics. Four sequential phases were recognized: 1) removal of high molecular
weight humic carbohydrate-like organics; 2) removal of free volatile fatty acids- 3)
remova of factenally excreted carbonyl compounds and amino acids; 4) removal of high
molecular weight carbohydrates produced during the third phase. Membrane fractionation
analysis showed that the majority of the metals permeated the 500 MW UF membrane, in-
dicating that chelation of most metals by refractory organics in leachate plays a
minor role in metal attenuation processes; an exception was iron, most of which was
associated with the 100,000 MW UF retentate.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/GtOUp
hemical analysis
tefuse disposal
.caching
Colorimetric analysis
Waste treatment
Solid Waste
Leachate Analysis
Organic Analysis
Sanitary Landfill
13B
8. DISTRIBUTION STATEMENT
RELEASE UNLIMITED
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. Or PAGES
226
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
210
AHA gnomon nwTMomcfcii77-797-056/6»6
------- |