&EPA
             United States
             Environmental Protection
             Agency
             Municipal Environmental Research  EPA 600/2 78 157
             Laboratory          August 1978
             Cincinnati OH 45268
             Research and Development
Attenuation
of Pollutants
in Municipal Landfill
Leachate
by Clay Minerals

-------
                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-600/2-78-157
                                       August 1978
   ATTENUATION OF POLLUTANTS IN MUNICIPAL
     LANDFILL LEACHATE BY CLAY MINERALS
                     by
       R. A. Griffin and N. F. Shimp
      Illinois State Geological Survey
           Natural Resources Bldg.
           Urbana, Illinois  61801
          Contract No. 68-03-0211
              Project Officer

              Mike H. Roulier
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.
                                      ii

-------
                                 FOREWORD
     The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.

     Research and development, that necessary first step in solving a prob-
lem, involves defining the problem, measuring its impact, and searching for
solutions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal and community
sources, to preserve and treat public drinking water supplies, and to min-
imize the adverse economic, social, health, and aesthetic effects of pol-
lution. This publication is one of the products of that research; it is a
most vital communications link between the researcher and the community.

     This report presents results from laboratory investigation of the
capacity of clay minerals to remove pollutants from municipal landfill
leachates. These results are applicable to the design of clay mineral liners
for sanitary landfills and to the land disposal of municipal and hazardous
wastes.
                                          Francis T. Mayo, Director
                                          Municipal Environmental Research
                                          Laboratory
                                     iii

-------
                                  ABSTRACT

     The first part of this project was a laboratory column study to evaluate
the potential of mixtures of sand and calcium-saturated clay minerals for
attenuating and preventing pollution of water resources by pollutants in
municipal solid waste landfill leachate.  Chloride, Na, and water-soluble
organic compounds (COD) were relatively unattenuated by passage through the
clay-sand columns; K, NH^, Mg, Si, and Fe were moderately attenuated; and
heavy metals — such as Pb» Cd, Hg, and Zn — were strongly attenuated by even
small amounts of clay.  Concentrations of Ca, B, and Mn in the column effluents
increased markedly over the original leachate concentrations.

     Montmorillonite was found to have the highest attenuation capability,
followed by illite and then kaolinite.  Precipitation, with resultant accumula-
tion in the surface layers of the columns, was found to be the principal
attenuation mechanism for the heavy metals Pb, Cd, Hg, and Zn.  The cation
exchange capacity of the clay minerals was concluded to be the dominant
attenuation mechanism responsible for removal of other substances from the
leachate.

     The second part of the project involved batch studies of adsorption of Cr,
Cu, Pb, Cd, Hg, and Zn by montmorillonite and kaolinite from water solutions
and from landfill leachate.  The adsorption in leachate proved to be 50 to 90%
lower in most cases than the clays' adsorption capacity for the metal ions in
pure aqueous solutions.  The adsorption of the cations Cr(III), Cu, Pb, Cd^
Hg, and Zn increased with increasing pH while the anions Cr(VI), As, and Se
decreased with increasing pH.  At pH values greater than about 5.3, precipi-
tation of the heavy metal cations was found to be an important attenuation
mechanism while adsorption was the principal mechanism for the anions over
the entire pH range studied.

     Pollutant adsorption by clay minerals (and hence the mobility of pollut-
ants in clays and clay soils) was significantly affected by other, non-hazard-
ous solutes in the leachates.  This effect was so pronounced that information
on movement of pollutants in one landfill leachate cannot be directly applied
to predicting the movement of the same pollutants present at the same con-
centrations in a different landfill leachate.

     To evaluate the relative pollution hazard for municipal leachates, a
ranking system was developed.  Results of the study are applicable to the use
of clay minerals as liners for sanitary landfills and to the land disposal of
municipal and hazardous wastes.

     This report was submitted in fulfillment of Contract Number 68-03-0211,
by the Illinois State Geological Survey, under the partial sponsorship of the
Environmental Protection Agency.  Work was completed as of August 1975.

                                      iv

-------
                                  CONTENTS

Abstract	iv
List of Figures	vl
List of Tables	lx
Acknowledgments	   x

    1.  Introduction 	   1
    2.  Conclusions	   4
    3.  Recommendations	   7
    4.  Experimental 	   9
    5.  Column Leaching Study	16
    6.  Confirmation of Laboratory  Column Studies by  Comparison
          with Field Data	34
    7.  Effect of pH on Exchange-Adsorprtion or  Precipitation of
          Lead from Municipal  Leachates  by  Clay  Minerals  	  44
    8.  Effect of pH on Copper,  Zinc, and Cadmium Removal from
          Deionized Water and  Landfill Leachate  by  Clay Minerals  ....  61
    9.  Effect of pH on Adsorption  of Chromium from Landfill-
          Leachate by  Clay  Minerals	83
    10.  Effect of pH on Adsorption  of Arsenic and Selenium from
          Landfill-Leachate by Clay Minerals 	  102
    11.  Mercury  Removal from Municipal Landfill-Leachate  by  Clay
          Minerals	113
    12.  Summary  of Adsorption  Studies	  120
    13.  Application of the  Results  to the Problem of  Landfill Design  .  .  123

References	.  .  128
Appendices

    A.  List of  Publications . .	134
    B.  Procedures Used in  Chemical and  Physical Characterization of
          Clay Minerals and Liquid  Samples  	  136

            Clay Mineral Characterization	136
            Atomic Absorption  Methods	136
            Determination of Hg  by  Flameless A.A	  139
            Neutron Activation Analysis  Methods  for Hg and As	141
            X-Ray Fluorescence Methods 	  143
            Optical Emission Spectrometry Methods	143
            Determination of Carbon	143
            Methods for Ammonium, TDS, COD, and  Chloride  	  144
            Methods for Sulfate,  Phosphate, and  Boron	145
            Determination of Base Exchange  Capacity  	  146
            Determination of Surface Area	146

-------
                               LIST OF FIGURES

No.                                                                      Page

1.   Diagram of column apparatus used in leachate study 	    13

2.   Relative column effluent concentrations for several elements as
       a function of pore fraction of leachate passed through columns
       containing (a) 2% mohtmorillonite, (b) 8% montmorillonite, and
       (c) 16% montmorillonite clay	*	    18

3.   Relative column effluent concentrations for several elements as
       a function of pore fraction of leachate passed through columns
       containing (a) 2% montmorillonite, (b) 8% montmorillonite, and
       (c) 16% montmorillonite clay	    19

4.   The attenuation number related to cation exchange capacity (a)
       K, (b) NHi,, (c) Na, and (d) Mg	    24

5.   (a) Ca attenuation number related to cation exchange capacity.
     (b) Mh attenuation number related to clay percentage 	    25

6.   Manganese elution related to percentage of kaolinite leached
       with natural and sterile leachate	    30

7.   Hardness of water expressed as CaCOa, in Silurian Dolomite
       aquifer in DuPage County, northeastern Illinois	    36

8.   "Hardness halo" effect shown as a function of distance (m)
       from the Winnetka and DuPage landfills in northeastern
       Illinois	    37

9.   Hydraulic conductivity of kaolinite-sand columns as a
       function of leaching time	    39

10.  Hydraulic conductivity of montmorillonite-sand columns as
       a function of leaching time	    40

11.  Hydraulic conductivity of illite-sand columns as a function
       of leaching time	    41

12.  The amount of Pb removed from DuPage leachate by kaolinite
       at 25° C plotted as a function of pH	    47

13.  The amount of Pb removed from DuPage leachate by montmoril-
       lonite at 25° C plotted as a function of pH	    48

                                     vi

-------
No.                                                                      Page

14.  The amount of Pb removed from Blackwell leachate by
       kaolinite at 25° C plotted as a function of pH	49

15.  The amount of Pb sorbed per gram of kaolinite at pH 5.0
       and 25° C plotted as a function of the equilibrium Pb
       concentration	52

16.  Pb sorption data for kaolinite and montmorillonite at pH
       5.0 and 25° C plotted according to the Langmuir isotherm
       equation	53

17.  Distribution of Pb (II) species in 4 x lO"1* M Pb(N03)2 and
       uptake by 0.1 g kaolinite from 60 ml of solution	56

18.  The amount of Cu, Zn, or Cd removed from solution per gram
       of kaolinite at pH 5.0 and 25° C, plotted as a function of
       the equilibrium concentration	67

19.  The amount of Cu, Zn, or Cd removed from solution per gram
       of montmorillonite at pH 5.0 and 25° C, plotted as a
       function of the equilibrium concentration	68

20.  Cu, Zn, and Cd removal data for kaolinite in deionized
       water solutions at pH 5.0 and 25° C, plotted according
       to the Langmuir equation (Eq. 8)	71

21.  Cu, Zn, and Cd removal data for montmorillonite in
       deionized water solutions at pH 5.0 and 25° C, plotted
       according to the Langmuir equation. Numerals indicate
       corresponding isotherms in Fig. 19	72

22.  Cu, Zn, and Cd removal data for kaolinite in deionized
       water solutions at pH 5.0 and 25° C, plotted according
       to the competitive Langmuir equation (Eq. 12)	  74

23.  Cu, Zn, and Cd removal data for montmorillonite in
       deionized water solutions at pH 5.0 and 25° C, plotted
       according to the competitive Langmuir equation (Eq. 12)	  75

24.  The amount of Cu, Zn, or.Cd removed from DuPage leachate
       solutions by kaolinite at 25° C, plotted as a function
       of pH	78

25.  The amount of Cu, Zn, or Cd removed from DuPage leachate
       solutions by montmorillonite at 25° C, plotted as a
       function of pH	79

26.  Chromium (VI) adsorption-pH curves for montmorillonite
       at 25° C	87
                                    vii

-------
No.                                                                       Page

27.  Chromium  (VI) adsorption-pH curves for kaolinite at
       25° C ..............................    88

28.  Distribution of Cr(VI) species for various Cr(VI)
       concentrations ..........................    89

29.  Adsorption isotherms for Cr(VI) at pH 4.0 and 25° C  ........    91

30.  Langmuir plots of Cr(VI) adsorption data at pH 4.0 and 25° C.  .  .  •    92

31.  Removal of Cr(III) from solution by kaolinite ...........    94

32.  Chromium  (III) adsorption-pH curves at 25° C ............    96

33.  Chromium  (III) adsorption isotherms at pH 4^.0 and 25° C  ......    97

34.  Langmuir plots of Cr(III) adsorption data at pH 4.0  and  25° C  .  .  .    98

35.  The  amount of As(V) removed from DuPage leachate solutions
       by kaolinite and montmorillonite at 25° C plotted  as a func-
       tion of pH ............................   105

36.  Species distribution diagram for As(V) and Se(IV) .........   106

37.  The  amount of As (III) removed from DuPage leachate solutions
       by kaolinite and montmorillonite at 25° C plotted  as a
     -  function of pH ..........................   107

38.  The  amount of As(V) or As (III) removed from DuPage leachate
       solutions at pH 5.0 and 25° C per gram of clay plotted as
       a  function of the equilibrium arsenic concentration  .......   108

39.  The  amount of Se(IV) removed from DuPage leachate solutions
       by kaolinite and montmorillonite at 25° C plotted  as a
       function of pH ..........................   110

40.  The  amount of Se(IV) removed from DuPage leachate solutions
       at 25°  C and several pH values per gram of clay plotted
       as a function of the equilibrium Se concentration  ........   Ill
 41.   Removal of  Hg  from DuPage  landf ill-leachate  and pure
        solutions plotted as  a function  of pH  at 25° C ..........    116

 42.   Removal of  various forms of Hg from DuPage landf ill-leachate
        solutions by kaolinite plotted as a  function of  pH  at  25° C  .  .  .    117

 43.   Effect  of clay content  on  hydraulic conductivity and  attenuation
        of  Pb, NH4,  and  Cl for a 40-cm thick liner ............    124

 44.   X-ray diffraction patterns of the  clay minerals  used  in
        leachate  pollutant attenuation studies ..............   137


                                    viii

-------
                              LIST OF TABLES
No.                                                                      Page

1.  Chemical Characterization of the Clay Minerals Used in
      Attenuation Studies of Leachate Pollutants ............    10
2.  Summary of Chemical Characteristics of Landfill Leachates
3.  Design of Experiment and Some Physical and Chemical Properties
      of the Column Contents
4.  Rank of Some Chemical Constituents Found in Municipal Leachate
      According to Their Relative Mobility Through Clay Mineral
      Columns .............................    21

5.  Mean Attenuation Number  (ATN) of Some Chemical Constituents
      Found in Municipal Leachates for Three Clay Minerals .......    23

6.  Chemical Constituents in DuPage Leachate, Ranked According to
      Pollution Hazard ....... • ..................    33

7.  Pb Removal Parameters Used to Compute Sorption Isotherms from
      52 ml Reaction Volumes ......................    50

8.  Maximum Removal of Pb from pH 5.0 and 25° C Solutions Computed
      Using the Langmuir Equation ........ . ..........    54

9.  Thickness (cm) of a Square Meter of a 30% Clay Liner Needed to
      Remove Pb from 762 Liters  (201 gal) of Solution Per Year .....    59

10. Total Content of Cu, Zn, or  Cd in 50 ml of Solution ........    69

11. Comparison of Langmuir Adsorption Maximums in Deionized Water
      with CEC Values .........................    76

12. Adsorption Maxima for Cr(VI) by Montmorillonite and Kaolinite
      at 25° C for Various pH Values ..................    93

13. Adsorption Maxima for Cr(III) by Montmorillonite and Kaolinite
      at 25° C for Various pH Values ..................    99

14. Removal of Heavy Metals  from Solutions by Kaolinite at pH 5.0 .  .  .   121

15. Estimated Landfill Liner Thickness Necessary for Attenuation
      of Some Leachate Constituents Per Cubic Meter of Refuse
      During a 20 Yr. Fill Life  ....................   126

                                      ix

-------
                              ACKNOWLEDGMENTS
     In addition to the authors listed, the following were involved in writ-
ing parts of this report:  A. K. Au, Keros Cartwright, R. R. Frost, R. H.
Gilkeson, G. M. Hughes, G. D. Robinson, R. R. Ruch, J. D. Steele, and W. A.
White.  Their names are also listed in a footnote at the beginning of each
section to which they contributed.

     The authors gratefully acknowledge the U.S. Environmental Protection
Agency for partial support of the work under Contract No. 68-03-0211, Cin- •
cinnati, Ohio; the American Colloid Company, Skokie, Illinois, for supplying
the montmorillonite clay; the Minerva Company, Elizabethtown, Illinois, for
supplying the illite clay; and the Ottawa Silica Company, Ottawa, Illinois,
for supplying the quartz sand used in this study.

     The authors also wish to give thanks to those Geological Survey staff
members who gave freely of their time and talents.  Among those are J. M.
Mellske, D. R. Dickerson, R. H. Shiley, W. J. Armon, J. K. Kuhn, J. A.
Schleicher, L. R. Camp, G. B. Dreher, J. K. Frost, L. R. Henderson, M. C.
Charles, S. K. Friesen, R. M. Trandel, and D. B. Heck.

     Special thanks go to R. J. Helfinstine and W. E. Cooper for special
equipment and to P. L. Johnson for typing assistance.
                                      x

-------
                                  SECTION 1

                                INTRODUCTION
     During the past 30 years, the landfill method for disposal of municipal
and industrial waste has been widely used in the United States. More than 90%
of our nation's wastes are now placed on the land for ultimate disposal at
the approximately 14,000 landfills throughout the nation (Garland and Mosher,
1975). These 14,000 landfills accept more than 360 million tons of household,
commercial, industrial, and municipal solid waste per day, at a cost of more
than 4.5 billion dollars annually (Black et al., 1968), As industry in the
United States complies with the Clean Air Act and the Federal Water Pollution
Control Act, the volume of industrial solid wastes, sludges, and liquid con-
centrates of pollutants is expected to double in the next 10 years. The dis-
posal of such huge volumes of solid waste by landfilling is not without its
environmental impact. When refuse buried in a landfill comes in contact with
water, then leachate, a mineralized liquid high in organic substances, is
produced and may move out of the fill and pollute the ground water.

     Garland and Mosher (1975) have cited several examples of pollution by
leachates migrating from landfills. An example of severe economic damage in-
curred by pollution of a drinking water aquifer by leachate from a landfill
occurred in New Castle County, Delaware (Apgar and Satterthwaite, 1975).
Leachate from the landfill migrated more than 800 feet and contaminated the
Potomac drinking water aquifer 4 years after the landfill site had been
closed. The drinking water was contaminated with such high levels of organic
compounds and metal ions that it was no longer potable. To date, this problem
has cost the county $800,000 for interim solutions and, if the dump must be
moved to completely remedy the situation, the cost may reach as high as 20
million dollars. In addition to the monetary costs, the county estimates that
it will take 10 years to restore full use of the aquifer.

     In another case, reported by Garland and Mosher (1975), contamination of
ground water by selenium was found to extend more than 2 miles from a dump
site in Long Island. In this case as elsewhere, heavy-metal contamination may
impart no odor or color to indicate that the water is contaminated.

     The solid waste problem is most acute in the metropolitan areas, where
competition for the available land is intense. A city of 2 million inhabi-
tants generates 5000 tons of solid waste per day (Wirenius and Sloan, 1973),
which rapidly fills the conveniently situated landfill locations. The problem
of finding environmentally acceptable sites close to metropolitan areas is
compounded by urban sprawl and persistent opposition from citizen and en-
vironmental groups. The rapid increase in problems and the costs associated

-------
with transporting refuse for long distances now make it prudent to consider
sites that were previously unacceptable because they failed to meet certain
geologic and hydrologic critera. As of January 1975, Illinois had 2,040 waste
disposal sites recorded with the Illinois State Environmental Protection
Agency. Since 1965, the management of refuse disposal in Illinois has greatly
improved, owing to passage of the Illinois Refuse Disposal Act, which as-
signed the regulation of solid waste disposal to the Department of Public
Health. More comprehensive regulation was provided in 1970 with the creation
of the Environmental Protection Agency by passage of the Environmental Pro-
tection Act (HB3788). Regulations were passed that were designed to insure
that solid waste disposal facilities are located at such sites and operated
in such a manner as will protect the physical environment and public health.
These regulations have been enforced for the past 10 years. During this time,
the Illinois State Geological Survey has assisted the regulatory agencies by
evaluating the hydrogeologic conditions at proposed or operating waste dis-
posal sites. During the past 8 years, the Survey has appraised at about 100
sites annually the conditions relative to pollution hazard. Some sites were
not approved for geologic reasons, including locations in floodplains or
gravel pits, on fractured rock over aquifers, on steep grades, or in areas
of special environmental significance. Other sites were approved but were
never put into operation because of persistent opposition from citizen and
environmental groups or for other reasons.

     The future of landfill disposal is clear. Acceptable disposal sites will
be difficult to find, their location will be approved only after certain
geologic and hydrologic criteria are met, and greater care will be required
in their operation. The relative unavailability of geologically acceptable
sites in close proximity to metropolitan areas and the rapidly escalating
costs associated with transportation of waste materials across long distances
now make it economically feasible to consider physically modifying geo-
logically unacceptable sites that may be ideal in other respects. The
Illinois State Geological Survey has conducted extensive studies of the move-
ments of pollutants through various geologic strata and in a variety of
hydrologic settings at several landfill sites in northeastern Illinois
(Hughes et al., 1971). Cartwright and McComas (1968) also conducted geo-
physical investigations at the same sites. The above studies, and others by
Apgar and Langmuir (1971) and Emrich (1972), have indicated that pollutants
in leachate can be detected at variable distances from a landfill, depending
on the clay content of the soil or the hydraulic conductivity of the soil
strata. It has therefore been suggested that a clay liner in the bottom of
previously unacceptable sites, such as gravel pits or old quarries, could
make them acceptable for disposal of municipal and/or industrial wastes. How-
ever, no sound evidence existed to indicate how thick such a layer must be or
what types of clay minerals were best suited for removal of toxic metals in
the presence of municipal leachate.

     This paper reports the results of a study (1) to investigate and evalu-
ate the attenuating properties of several clay minerals for the pollutants
contained in leachates from municipal solid waste, and (2) to determine the,
capacity of the two major clay mineral types for removing the heavy metals —
Cr, Cu, Pb, Cd, Hg, As, Se, and Zn — from solution and the effect municipal
leachates have on this capacity at various pH values. The study was also de-

-------
signed to provide insight into the mechanisms responsible for attenuation of
heavy metals as well as to evaluate the potential use of clay minerals as
liners for waste disposal sites. Such liners would be used to prevent or
mitigate pollution of ground and surface waters by liquid effluents.

     It is clear that, based on their intrinsic properties alone, municipal
leachates are noxious waste streams that pose a potential threat to public
health (Hanks, 1967; Peterson, 1974). To assess the magnitude of the pollu-
tion hazard of such a stream is a difficult problem and was the subject of
the U. S. Environmental Protection Agency (1973) decision model for screen-
ing, selecting, and ranking hazardous wastes streams. Use of the priority
ranking formulation requires evaluation of the "critical product" (pollution
hazard). At present, no actual waste-stream data for municipal leachates is
available. Therefore, one goal of this study was to provide waste-stream
data that could be used to determine the mobility index of several of the
major pollutants found in municipal leachates £hat had passed through simu-
lated clay mineral landfill liners. The mobility index thus derived could be
used to compute the pollution hazard for most of the chemical constituents
found in municipal leachate.

     This report is presented in separate chapters that describe different
aspects of the work that was performed. For the convenience of the reader,
each chapter includes its own abstract. The column leaching studies (Sections
5 and 6) used "natural" landfill leachate as described in the Experimental
section  (Section 4). Sections 7 through 11 report equilibrium adsorption
"batch" studies using small amounts of clays mixed with aqueous solutions of
metal salts and landfill leachates that had been "spiked" with metal salts.
The results of these investigations will also find application in the land
disposal of industrial and energy production wastes.

-------
                                 SECTION 2

                                CONCLUSIONS
     The results from the column leaching and adsorption studies have yielded
complementary data that have allowed us to make predictions as to the ex-
pected magnitude of the reduction in concentration or attenuation in soil for
most of the common chemical constituents found in landfill leachates. The
data also indicate that several mechanisms may be responsible for attenuation
of the various pollutants under differing environmental conditions.

     Strong evidence is presented supporting the conclusion that the cation
exchange capacity is the principal chemical property of a clay mineral to
effect attenuation. Of the three clays used in this study, montmorillonite
was found to have the highest attenuation capability, followed by illite and
then kaolinite. This order was concluded to be a result of their respective
cation exchange capacities.

     The pH of the leachate was found to significantly affect the amount of
attenuation. It was concluded that the heavy metal cations — Pb, Cd, Cu,
Cr(III), Hg, and Zn — were attenuated primarily by an exchange-adsorption
mechanism which was affected by pH and competition from other cations. How-
ever, at pH values between 5 and 6, a large increase in removal can be ex-
pected due to increased adsorption of metal complex ions and to formation of
insoluble heavy metal hydroxide and carbonate compounds. It was therefore
concluded that at high pH the primary mechanism of attenuation for these ions
was precipitation. The effect of pH on the attenuation of the heavy metal
anions Cr(VI), As, and Se was found to be the opposite of the cations and it
was concluded that precipitation was not an important attenuation mechanism.
Rather, the adsorption of the anions was found to correlate well with the
distribution of certain ionic species in solution. It was concluded that
HCrOIT was the species adsorbing in this study, since Cr(VI) adsorption be-
came zero as the pH was raised to 8.4, corresponding to the disappearance of
HCrO^ from solution in favor of the CrO^ ion. The adsorption of Cr(VI) was
also found to start to decrease as the pH was lowered past 2, corresponding
to the decrease in HCrO^ ion in favor of HaCrO^. Likewise, As and Se adsorp-
tion were also found to correspond to the distribution of HaAsO^ and HSeOJ
species in solution. These results led to the conclusion that the principal
attenuation mechanism for the heavy metal anions was adsorption of the mono-
valent species from solution. It was also concluded that at higher pH values
the heavy metal anions would be significantly more mobile than the cations.

     The relative mobilities of the heavy metals, as determined from equi-
librium adsorption data from pure solutions of the metals at pH 5, were:

-------
       Cr(III) < Cu < Pb < Zn < Cd < As(V) < As(III) < Se < Cr(VI)

The cationic heavy metals are generally adsorbed to a greater degree than are
the anionic heavy metals. However, this ranking is dependent on pH and ionic
competition and therefore changes somewhat in different leachates.

     A significant point shown in the ranking is the importance of the
valence state of an element to the amount of that element removed from solu-
tion by clay minerals. Cr(III) species were removed to a much greater extent
than Cr(VI), and more As(V) was removed than As(III). These data suggested
that safer disposal of certain elements may be achieved if the element werejf
converted to the form most strongly attenuated prior to disposal.

     The formation of Pb and Cd organic complexes in leachates was measured,
and their effect was determined to be of secondary importance to adsorption
and precipitation. This seemed to be due to competition from high concentra-
tions of other cations present in leachates. Due to the complex interactions
between inorganic, organic, and volatile forms of Hg, the mobility and rel-
ative importance of organic Hg complexes could not be accurately assessed.

     The results of this study have led to the conclusion that passage of
leachate through a Ca-saturated clay material will result in high attenuation
of the heavy metals; in moderate attenuation of K, NHi*, Mg, and Si; and in
relatively low attenuation of Cl, Na, and water-soluble organic compounds
(COD). It was further concluded that the oxidation-reduction potential of the
leachate controlled the attenuation of Fe and Mn. Under strongly anaerobic
conditions, Fe and Ma will probably not be attenuated and may even elute in
substantial concentrations due to the dissolution of oxide coatings on the
clay surfaces. However, under mildly anaerobic conditions, substantial at-
tenuation can occur.

     Substantial concentrations of Ca were eluted from the columns. Since a
very highly significant linear regression (r = 0.97) was obtained for the
amount of Ca eluted versus the cation exchange capacity of the clay, and
since the sum of the amount of K, NHi,, Na, and Mg removed from the leachate
agreed within about 3% with the amount of Ca eluted, it was concluded that
Ca elution was due to exchange with the other cations present in the leach-
ate. It was also concluded that this Ca elution observed in the laboratory
experiment corresponded to the "hardness halo" effect observed in field
monitoring wells around sanitary landfill sites.

     Significant reductions in hydraulic conductivity were observed when
landfill leachate was passed through the columns. It was concluded that
microbial action is primarily responsible for the observed reductions and
that hydraulic activity of clay-sand liners placed in the bottom of a land-
fill will decrease after a period of leaching with municipal effluent. It was
further concluded that montmorillonite clays will decrease in permeability to
a greater extent than the other clays. This was assumed to be due to its
tendency to swell to a much greater extent than other types of clay.

     A pollution hazard ranking system was developed. Using this system, it
was concluded that NHit was a 30 times greater pollution hazard than any other

-------
constituent found in the DuPage leachate. It was also concluded that in a
fresh, young leachate, COD or Fe most likely would present the highest pollu-
tion hazard and that the characteristics of each leachate and earth material
must be considered when evaluating their potential for pollution. The well
known variability in composition of municipal landfill leachates (Garland
and Mosher, 1975) and the difficulty in predicting leachate composition be-
fore landfilling is begun will make it difficult to apply this hazard-ranking
system. Nevertheless, the ranking system has the advantage of quantifying the
expected pollution hazard of a given leachate and allows comparisons of the
pollution hazards of one leachate with another. The ranking system also aids
regulators, landfill designers, and researchers by focusing attention on
those chemical constituents in the leachate with the highest potential for
serious pollution.

     The mobility of a given leachate constituent, and hence the thickness of
a clay liner necessary to attenuate the constituent, depends to a large de-
gree on the element and the form of the element, the adsorption capacity of
the earth material, the cations present initially on the exchange complex,
the chemical composition of the leachate, and the pH of the leachate. The
adsorption capacity of the clay minerals and the reversible nature of ex-
change-adsorption reactions have important environmental consequences. Indus-
trial wastes containing heavy metals placed in sanitary landfills could alter
the ionic composition and/or pH of the leachate. A change in pH may release
large amounts of potentially toxic heavy metals into the aqueous phase,
especially in places where precipitates may have accumulated. Other ions in
the waste compete with the heavy metals and may exchange with them, thus al-
lowing metal ions to come into solution. These multiple interactions must be
considered when a disposal site is designed and when the environmental impact
of adding heavy-metal wastes to municipal landfills is assessed.

-------
                                 SECTION 3

                              RECOMMENDATIONS
     This study dealt specifically with attenuation of the inorganic con-
stituents of leachate, but, because leachates contain high concentrations of
organic compounds, also needed is a laboratory investigation dealing with the
attenuation of the specific organic compounds in leachate. The results of our
study, using COD as an indicator of organic-compound movement, showed poor
attenuation of organics. At the same time, the pollution hazard index in-
dicates that the organic fraction of leachate poses a serious pollution
hazard, especially for young leachates. It is therefore recommended that an
attenuation study be initiated, involving a comprehensive organic analysis to
determine removal of specific organics by passage through clay liners and/or
soils.

     The results of this study also illustrated the important role that the
pH of the leachate or waste stream plays in pollutant removal by clay
minerals. It is therefore recommended that, in conjunction with the organic
removal studies, further studies be carried out to determine techniques which
will allow the pH to be manipulated to enhance and economically optimize
pollutant (organic and inorganic) removal by earth materials. Also, landfill
disposal of anionic forms of heavy metals such as Cr(VI), As, and Se should
be closely scrutinized because of their relatively high mobility and the fact
that manipulation of pH conditions to enhance removal of cationic heavy
metals such as Pb, Cd, Zn, Cu, and Cr(III) may actually increase the mobility
of the anionic metal ions.

     Due to the relatively, high toxicity of Hg and the complex interactions
between inorganic, organic, and volatile forms of the element, much more
research is necessary to determine the environmental impact of Hg in landfill
environments. In conjunction with the study of Hg, Pt has also been reported
to be methylated in aquatic systems, thus establishing a previously unknown
mechanism of transfer. The major source of potential redistribution of com-
plexed forms of Pt is expected to come from the disposal into landfills of
spent catalysts from catalytic-converter-equipped automobiles. Thus, it is
recommended that Hg and Pt be studied to quantitatively determine the adsorp-
tion capacity of clays and soils for the inorganic and organic forms of Hg
and Pt. It is also important to determine the magnitude and role of micro-
organisms in the methylation of Hg and Pt to volatile forms under the con-
ditions found in municipal landfills.

     A major task still before us is the utilization of quantitative soil
chemistry data to make predictions of pollutant migration beneath landfills.

-------
It is therefore recommended that further studies be carried out that involve
•a cooperative effort between soil chemists, groundwater hydrologists, and
modelers for the computer implementation of the prediction process and to
identify possible gaps in knowledge that may still bar the successful pre-
diction of long-term pollutant migration from disposal sites.

-------
                                  SECTION 4

                                EXPERIMENTAL
     The clays used in this study were the purified clay minerals kaolinite
(1:1 lattice type), montmorillonite  (2:1 expanding lattice), and illite (2:1
nonexpanding lattice, mica type). These clay minerals were chosen for study
because they are the most common clay minerals in earth materials that would
be used individually or in combinations for landfill sites. Earth materials
containing one or more of these clay minerals can generally be obtained
locally for landfill liners.

     The clays were brought to the laboratory, where they were crushed,
ground, and purified by sedimentation techniques to obtain the <2 ym particle
fraction that contained essentially pure clay minerals. Chemical analyses of
the three clays are given in Table 1. Details of the methods used and results
of X-ray diffraction analyses of the clays are given in the appendix. The
predominantly Ca-saturated, <2 ym fractions of the clays were then used in
the column leaching and heavy metal adsorption studies.

     The municipal landfill-leachates used in this study were collected from
the DuPage County landfill  (well MM 63) and from the monitoring well located
on the Blackwell Forest Preserve landfill. The details of the site descrip-
tions and well locations are given by Hughes et al. (1971). The leachate was
collected by using a tubing pump and large plastic containers that were
equipped with valves to allow continuous purging of high-purity C02 or argon
gas to maintain anaerobic conditions. The DuPage leachate collected initially
for use in the column leaching study was stored under argon. Both the DuPage
and Blackwell leachates collected at a later date for use in the heavy metal
adsorption studies were stored under C02. Purging with C02, a naturally oc-
curring landfill gas, was found to be a more satisfactory method than argon
purging since it permitted the leachates to be stored in the laboratory for
longer periods of time without significant changes in pH. The leachate used
in the column study was collected with a tubing pump and was split into two
53-gallon plastic closed-head drums. One drum was taken to the Argonne
National Laboratories and sterilized by gamma ray irradiation using a cobalt
source which gave a dose of 3.36 x 106 rad at the center of the drum. Both
the sterile and natural drums were stored under refrigeration with the
temperature maintained at 3 to 5° C. The drum containing the microbially
active leachate was stored with argon purged over the top of the leachate,
while the sterilized drum was stored over a 12% ethylene oxide/88% freon gas
mixture which maintained both sterility and an anaerobic environment. Mer-
curic nitrate salt was later added to the drum (Table 2) to maintain steril-
ity. Plate counts were performed on both leachate drums using potato dextrose

-------
           TABLE 1.   CHEMICAL CHARACTERIZATION OF THE CLAY MINERALS  USED
                     IN ATTENUATION STUDIES OF LEACHATE  POLLUTANTS





Element
Kaolin it e (Pike
County, Illinois)


(ppm)
Exch.*



Total
Monttnorillonite
(American Colloid
Co . southern
bentonite)
(ppm)
Exch.*
Total
Illite
(Minerva
Co. Mine)

(ppm)
Exch . *
Total
Ca
Mg
Na
K
NHn
Fe
Mn
Pb
Cd
Zn
B
Al
Si
Ti
Carbon (%)
Total
Organic
Inorganic
CEC
(meq/lOOg)
Surface
area (m2/g]
2,592. 3,700 13,120.
76.8 1.8QO 680.
43.2 929 24.0
87.2 8,200 240.
13.0 40 43.
<2.0 6,600 <2.0
0.06 29 0.02
<2.0 46 <2.0
<0.2 <3 <0.2
0.80 20 1.00
46 -
221,800
217,700
14,700

0.54
0.51
0.03
15.1
34.2
22,300
25,500
178
1,100
38
25,500
25
<15
<3
40
3
95,600
284,800
1,300

0.93
0.92
0.01
79.5
86.0
5,248 23,350.
800. 10,430.
115.2 1,050.
800. 56,270.
50. 62.5
<2.0 28,730.
0.37 <390.
<2.0 93.8
<0,3 18.8
2.5 37.5
43.8
130,100.
226,500.
4,010.

2.19
1.81
0.38
20.5
64.6
* Exch an ge ab1e
                                     10

-------
          TABLE 2.  SUMMARY OF CHEMICAL CHARACTERISTICS OF LANDFILL LEACHATES
/
Component
COD
BOD
TOC
Organic acids
Carbonyls as
acetophenone
Carbohydrates as
dextrose
pH
Eh (m.v.)
TS
TDS
TSS
E.G. (mmhos/cm)
Alkalinity
(CaC03)
Hardness
(CaC03)
Total P
Ortho P
NHit-N '
N03+N02-N
Al
As
B
Ca
Cl
Na
K
Sulfate
Mn
Mg
Fe
Cr
Hg
Ni
Si
Zn
Cu
Cd
Pb
Range of all
values given
by Garland and
Mosher (1975)
(mg/D
40 -
9 -
256 -
-
-

-

4 -
-
0 -
0 -
6 -
^ —
0 -

0 -

0 -
6 -
0 -
0 -
-
-
-
5 -
34 -
0 -
3 -
1 -
0 -
16 -
0 -
-
-
-
-
0 -
0 -
0 -
0 -
89,520
54,610
28,000





9

59,200
42,276
2,685
17
20,850

22,800

154
85
1,106
1,300



4,080
2,800
7,700
3,770
1,826
1,400
15,600
5,500




1,000
10
17
5
DuPage leachate
DuPage used in column
Blackwell Forest leachate used study
Preserve leachate in sorption (mg/1)
(Hughes, 1971b) study
(mg/1) (mg/1) Natural Sterile
39,680.
54,610.
-
-
-

-

7.10
-180.
-
19,144.
-
10.90
3,255.

7,830.

6.
-
-
1.70
2.20
4.31
-
-
1,697.
900.
-
680.
1.66
-
5,500.
. 0.20
-
-
-
-
0.05
<0.05
"
1,362.
-
-
333.
57.6

12.

6.79
-155.
-
5,910.
-
7.20
4,220.

1,100.

<0.1
-
809.
-
<0.1
0.11
33.
49.
1,070.
822.
516.
<0.01
<0.1
204.
4.40
<0.1
0.0008
0.3
15.1
0.03
<0.1
<0.01
<0.1
1,340.
-
-
333.
57.6

12.

6.9
+7.
-
-
-
10.20
-

-

<0.1
-
862.
-
<0.1
0.11
29.9
46.8
3,484.
748.
501.
<0.01
<0.1
233.
4.2
<0.1
0.0008
0.3
14.9
18.8
<0.1
1.95
4.46
10,603.*
-
-
290.
90.1

11.

7.2
+75.
-
-
-
10.42
-

-

<0.1
-
773.
-
<0.1
0.14
28.5
43.2
3,311.
744.
491.
<0.01
<0.1
230.
3.0
<0.1
0.87*
0.3
15.0
16.3
<0.1
1.88
4.26
*Added as- a result of sterilization maintenance
                                           11

-------
agar media. No growth occurred on plates inoculated with the sterilized
leachate while active growth of microbial colonies was observed on plates
inoculated with the natural leachate, thus indicating that sterility had been
achieved.

     The results of chemical analysis of the leachates are presented in Table
2. Chemical procedures used are given in the appendix. For comparison, Table
2 also contains a summary of the range of leachate characteristics found for
more than 20 other leachates as given by Garland and Mosher (1975). It is
useful to note that the two leachates used in this study have widely differ-
ent chemical compositions. The DuPage leachate is approximately 15 years old
and has a lower total salt, phosphate, and sulfate content than the Black-
well. In addition, the organic matter of the DuPage leachate consists mainly
of microbially resistant compounds, which have been found to be more mobile
in soils than are biodegradable compounds (Hughes et al., 1971, Gowler,
1970). The Blackwell leachate, on the other hand, is younger and ranks among
the most concentrated ever reported, especially with regard to BOD and Fe.

     The laboratory apparatus used in the leaching study consisted of labora-
tory columns containing mixtures of clay minerals and washed quartz sand
through which leachate was passed. A diagram of the column apparatus design
appears  in Figure 1. The columns and apparatus were constructed to simulate
the slow (<2 pore volumes per month), saturated, anaerobic flow of leachate
as it is thought to occur at the bottom of a landfill.
                                          ~\
     Pore Volume =   Tl - (  Bulk Density  J    Volume of Column           {l}
                     |_     Particle DensityJ

The entire  column leaching system, was maintained under an argon atmosphere to
maintain anaerobic conditions. A tubing pump lifted the leachate to a 5-gal-
lon plastic carboy which acted as both a constant head device  (Harriot
bottle) and a temperature equilibration reservoir. The leachate was then
passed through the columns and the effluents were collected in graduated
cylinders,  which also allowed measurement of the flow rates. The outflow tube
was maintained above the top of the columns to insure saturated flow. The
level of the outflow tube was moved either up or down to maintain relatively
constant flow rates  throughout the experiment.

     The columns were constructed of 2-inch acrylic tubing, to which man-
ometer outlets were  fitted vertically on the column at five locations. To
simulate field conditions, the leachate containers and columns were either
painted black or masked with black tape to prevent growth of organisms, such
as algae or photosynthetic bacteria, which are not indigenous  to deep refuse
leachate.

     The sand grains were coated with the clays, according to  the methods
given by Grim and Cuthbert (1945) . The clay minerals and sand were then uni-
formly packed in the columns to a depth of 40 cm, except for the 32% and 64%
montmorillonite columns, which were 30 cm thick. The columns were packed to
bulk densities approximating those of naturally occurring glacial tills
(~1.8 g/cc;  Manger,  1963). Table 3 gives some chemical and physical proper-
ties of the column contents. The hydraulic conductivities for each particular

                                     12

-------
                                                                         CVI
55 Gal.         Water  Refrigeration
Leachate Drum   Bath   Unit
 Figure 1.  Diagram of  column apparatus used in leachate study.
                                    13

-------
TABLE 3.  DESIGN OF EXPERIMENT AND SOME PHYSICAL AND CHEMICAL PROPERTIES
          OF THE COLTJMN CONTENTS
Natural Leachate
Treatment
Cation
Exchange
Capacity
(meq./lOOg)
Set A*Set Bt
 Bulk density
	(g/cc)
Initial hydraulic
conductivity
     (cm/sec)
                                   Set A   Set B   Set A
                               Set B
100 Sand
2 Mtt
4 M
8 M
16 M
32 M
64 M
2 Ktt
4 K
8 K
16 K
32 K
64 K
4 Itt
16 I
8K + 81
8M + 8K + 81
0.0
1.4
3.2
7.3
11.9
26.8
56.2
0.7
1.1
1.5
1.8
3.8
9.6
0.8
3.5
2.4
8.8
.1
2.3
4.3
7.2
12.1
24.0
55.5
0.4
0.8
1.4
2.5
3.4
8.5
0.9
3.2
2.3
8.5
1.71
1.71
1.77
1.79
1.87
1.55
1.23
1.68
1.76
1.80
1.87
1.66
1.22
1.80
1.83
1.90
1.64
1.71
1.72
1.74
1.78
1.86
1.52
1.11
1.70
1.74
1.77
1.90
1.55
1.32
1.81
1.91
1.98
1.69
1.27.10-3
9. 45. 10-"
4.34.10-"
4.70.10-"
1.22.10-5
1.27.10-6
3.05.10-7
7.44.10-"
4.78.10"5
9. 90.10-"
2.86.10~5
2.40.10-6
5.45.10-7
8. 17. 10-"
2.68.10"5
1.48.10-6
8.08.10-6
1.80.10-3
7. 93. 10-"
3. 47. 10-"
2. 61. 10-"
1.44.10-5
2.17.10-6
6.83.10-7
4. 53. 10-"
2.76.10'5
8. 25. 10-"
1.92.10-6
4.81.1Q-6
4.57.10'7
7. 16. 10-"
2.19-10-5
1.68.10'6
9.43.10-6
  *Set A - Natural leachate

  tSet B - Sterile leachate

  ttM = Montmorillonite, I = Illite, K = Kaolinite
                                    14

-------
clay content and bulk density  agree with  those  given by Todd  (1959)  for
natural materials. The experimental design  used in  the study  is  also shown  in
Table 3, which gives the percentages  of clay  mineral(s) in  each  column (to
which pure quartz sand was  added  to total 100%) .  The experimental design  in-
cludes a complete geometric progression of  clay percentages from 2%  to 64%
kaolinite and montmorillonite  and two mixtures  of clays as  given in  Table 3.
A 100% sand column was also included. Only  4  and 16 percentages  of illite
were included because kaolinite and illite  have very similar  cation  exchange
and lattice expansion properties, and thus  a  complete geometric  array for
illite was not considered necessary.

     After the leachate and the column contents were characterized,  the
leachate was passed through the columns for periods of between 6 and 10
months, depending on the hydraulic conductivity of  the individual column. The
hydraulic conductivity  (K)  was computed using the relationship:
                                 K
                                     AdH

where  Q  =  flow rate  in cm3 /sec
       A  =  cross sectional area of column in cm2
       dL = length of the column in cm
       dH = head of water in cm.

     During this time effluents from each column were  collected periodically
and  measurements were made  for Na, K, Ca, Mg,  Al, Zn,  Pb,  Cd, Hg, Fe, Mh,
NHi» , B,  Si, Cl, chemical oxygen demand (COD),  Eh, pH,  and  hydraulic  conduc-
tivity.  Finally, after approximately 15 pore volumes were  leached, the clay
mineral  columns were sectioned and the contents  analyzed to  determine the
vertical distribution of chemical constituents in each column.

     Duplicate sets  of columns were used in the  experiment;  one set  of
columns  was leached  with natural effluent, another with sterilized effluent.
The  sterilized treatment was used to determine how gross biological  activity
might  affect hydraulic conductivity of leachate  through clay minerals used as
liners.  The results  of the  experiment were statistically analyzed using the
paired t statistic to determine if there were  significant  differences in the
attenuation of each  chemical constituent between sterile and natural leachate
and  between clay mineral types.  Linear regression and  moving average analysis
were also  performed  on the  column effluent data  to determine relationships
between  hydraulic conductivity,  attenuation, and clay  mineral properties.

     In  addition to  the column leaching study > a series of separate  equi-
librium  studies on the capacity of clays to adsorb eight potentially hazard-
ous  elements — Pb,  Cd, Zn, Cu,  Cr, As, Se, and  Hg —  were performed. From
these  studies, adsorption isotherms for kaolinite and  montmorillonite were
constructed to determine the adsorption capacities as  a function of  concen-
tration, pH, and ionic competition. The details  of the experimental  pro-
cedures  used for each element are included in  the section  of the report deal-
ing  with that particular element.
                                      15

-------
                                  SECTION 5

                           COLUMN LEACHING STUDY1
ABSTRACT
     To evaluate the potential of clay minerals for attenuating the various
chemical constituents of landfill leachate, leachate was collected by an-
aerobic techniques from the 15-year old DuPage County sanitary landfill near
Chicago, Illinois, and passed through laboratory columns that contained vari-
ous mixtures of calcium-saturated clays and washed quartz sand. The columns
were constructed to simulate slow, saturated, anaerobic flow of leachate
through earth materials outside the landfill. Manometers were placed in each
column to measure any changes in permeability. Leachates were run through the
columns for periods of up to 10 months, during which time effluents were
periodically collected and analyzed for 16 chemical constituents. The column
contents were then cut into sections and analyzed to determine the vertical
distribution of chemical constituents in each column.

     Chloride, Na, and water-soluble organic compounds (COD) were relatively
unattenuated by passage through the clay columns; K, NHi* , Mg, Si, and Fe were
moderately attenuated; and heavy metals, such as Pb, Cd, Hg, and Zn, were
strongly attenuated by even small amounts of clay. Concentrations of Ca, B,
and Mn in the column effluents increased markedly over the original leachate
concentrations. The increase in Ca was due to cation exchange with ions in
the leachate. The amount of Ca eluted from the columns was found, by mass
balances, to agree within 3% with the sum of Na, K, NIU , and Mg removed from
solution. The Mn increase probably resulted from a reduction of the oxidized
Mn on clay surfaces by the anaerobic leachate to more soluble ionic species.

     Of the three clays used in the study, montmorillonite had the highest
attenuation capability, followed by illite and then kaolinite. This order
correlates well with the cation exchange capacities of the three clay min-
erals, which appear to be the dominant attenuation mechanism in these clays.
The principal attenuation mechanism for the heavy metals Pb, Cd, Hg, and Zn
was found to be precipitation, with resultant accumulation of the metals in
the surface layers of the columns.

     A ranking system was developed for evaluating the relative pollution
hazard for municipal leachates. The new ranking method overcomes the problems
and objections found in evaluation of the critical product parameter for


JAuthors: R. A. Griffin, N. F. Shimp, J. D. Steele, R. R. Ruch, W. A. White,
          and G. M. Hughes.

                                     16

-------
municipal leachates by a method previously  proposed  to  the U.  S. EPA  for
hazardous wastes.

     Results of the study  are  applicable  to the  use  of  clay minerals  as
liners for sanitary landfills  and  to  the  disposal  of industrial and power
plant wastes in landfills  and  mines.

INTRODUCTION

     The following results and discussion consider data from the columns
leached with normal leachate and do not include  results from the sterilized
leachate. Details of  the effect of sterile  leachate  on  hydraulic conductivity
and chemical constituent attenuation  will be the subject  of Section 6. How-
ever, for those chemical constituents for which  no significant difference in
attenuation between the normal and sterile  leachate  treatment was  found, the
results were pooled to give added  statistical significance to  analysis of
clay-type effects. Those constituents for which  no significant difference was
observed were  Ca, Mg, Na,  K, NHi,,  Pb, Hg, Zn, and  Cd.

COLUMN LEACHING STUDIES

     Hydraulic conductivity and bulk  density measurements of the column con-
tents are presented in Table 3. A  wide range of  hydraulic conductivities,
with values in agreement with  those expected under field  conditions from
similar materials  (Todd, 1959), was observed. The  relatively high  bulk densi-
ties and slow  flow rates used  in this study closely  simulate the conditions
observed in the field, which lends credence to the extrapolation of the re-
sults and conclusions presented here  to field applications using clay liners
of similar composition.

     Results of some  column .effluent  analyses are  shown in Figures 2  and 3
plotted as relative concentration  versus  pore fraction. Relative concentra-
tion is the ratio of  the column effluent  concentration  divided by  the in-
fluent concentration. Thus the "breakthrough" point  for a given element is
where the column effluent  concentration equals the influent concentration and
has a value of one. A pore volume  of  effluent is defined  as the volume neces-
sary to displace the  volume of interstitial liquid in the pore spaces in the
column. The pore fraction  is then  given as  the cumulative volume of column
effluent divided by the pore volume of the  individual column.

     Figures 2 and 3  illustrate the wide  range of  attenuation  observed for
several elements contained in  the  leachate  as it passed through columns con-
taining 2, 8,  and 16% montmorillonite clay.  The  amount  of reduction in con-
centration of  a given element  as it passes  through the  columns is  reflected
by the shift of the curves toward  higher  pore volumes.  The results shown in
Figure 2 for Cl, Na,  NH4,  and  K are qualitatively  in excellent agreement with
the results reported  by Farquhar and  Rovers (1975),  who used soils in their
tests. The attenuation order also  follows the general order of cation re-
placeability given by Grim (1968).

     Figure 3  illustrates  the  negative attenuation or elution  of Ca from the
columns. The relative concentrations  greater than  1  indicate that  Ca, and to

                                      17

-------
             1.20 i
            0.00
                0.0
4.0      6.0       8.0
   Pore  fraction
                                                         10.0
        12.0
             1.20-1
                                         6.0      8.0
                                    Pore  fraction
             1.20 i
             0.00
                        2.0
4.0      6.0       8.0
    Pore  fraction
10.0
12.0
Figure  2.   Relative  column effluent  concentrations for several elements  as
            a function  of pore fraction of leachate passed  through columns
            containing  (a) 2% montmorillonite,  (b)  8% montmorillonite,  and
            (c) 16% montmorillonite clay.
                                       18

-------
            30,0-1
           •£20.0-
           «
           o

           o
           o
           > 10.0-1
           a)
           cr
             0.0
                        2% Montmorillonite   A
                0.0
                fr=T=-t — t—*—V
                                                                 - Mrt
                                 4.0      6.0      8.0

                                     Pore  fraction
                          10.0
                                           12.0
                                               8% moritmorillonite   B
                                                  8.0      10.0     12.0
            30.0 -i
             0.0
               0.0
                       16% Montmorillonite   C
2-.0
4.0       6.0      8.0

    Pore fraction
                                                          10.0
                                          12.0
Figure  3.   Relative column effluent concentrations for several elements  as

            a function of pore  fraction of  leachate passed  through columns

            containing (a) 2% montmorillonite,  (b) 8% montmorillonite,  and

            (c) 16%  montmorillonite clay.
                                       19

-------
a lesser extent Fe and Mn, are eluting from the columns at much higher con-
centrations than the influent leachate at various pore fractions. The area
under the Ca curves in Figure 3 increases in proportion to the percentage of
clay in the column.

     To quantify the observed attenuation, the area under each curve was in-
tegrated between pore fractions 1 and 11. The area between 0 and 1 pore frac-
tion was not included because it was merely the displacement of the deionized
water initially present in the column. The total area was that bounded by 10
pore volumes and relative concentrations between 0 and 1. The relative at-
tenuation number (ATN) was then obtained by subtracting the area under the
curve in each case from the total area and was expressed as a percentage. The
ATN numbers are unique for each element and each clay and express the rel-
ative mobilities of each element through each particular clay or clay mixture
column.

     The attenuation number was computed for all columns and each chemical
constituent studied. The mean attenuation number for each chemical constit-
uent was used to rank the constituents according to their relative mobility
through the clay columns and is reported in Table 4. The constituents Al, Cu,
Ni, Cr, As, S, and POi* were found in such low concentrations in the DuPage
leachate that no attenuation order could be determined. The results showed
that greater amounts of the heavy metals Pb, Zn, Cd, and Hg than any other
element were removed from leachate, yielding an average of about 97% attenu-
ation for  10 pore volumes leached. The data for 2% montmorillonite (Fig. 2a)
show that  Cd and Hg were the most mobile of the heavy metals studied.
Measurable amounts of Cd and Hg appeared in the effluents of the 2% clay
column after about 6 pore volumes were leached. Only traces of Cd and Hg ap-
peared in  the effluents of the 8% and 16% clay treatments (Fig. 2b, 2c). Re-
movals of Pb and Zn were very high in all columns.

     Results obtained from chemical analysis of sectioned columns revealed
large accumulations of all four heavy metals in the surface layers of each
column, including the 100% sand column. These removals could be attributed to
cation exchange replaceability, but a precipitation and/or filtration mechan-
ism appears a more plausible cause. Precipitation could involve formation of
heavy-metal hydroxides or carbonates, brought about by the relatively high pH
found in the column effluents. The average pH of the influent leachate was
6.9, while the average pH of the column effluents rose to values of 7.3-7.9.
The increase in pH could result in precipitation of the heavy metals from
solution in the columns. The precipitation mechanism is given further cre-
dence by the measurements of the effluent concentrations from the sand
columns, which have no cation exchange capacity and an average effluent pH of
7.8. No Pb and markedly reduced concentrations of Zn, Cd, and Hg eluted from
the sand columns. These data, along with the high accumulations of the metals
in the surface layers of the sand columns, indicate that precipitation of
heavy-metal hydroxides and carbonates can be an important attenuation mechan-
ism. This  conclusion has been further verified by our equilibrium studies of
the effect of pH on heavy-metal removal by clay reported in Sections 7, 8, 9,
10, and 11 of this report.

     Filtration of particulate material in the leachate by the earth


                                     20

-------
TABLE 4.  RANK OF CHEMICAL CONSTITUENTS IN MUNICIPAL LEACHATE
          ACCORDING TO THEIR RELATIVE MOBILITY THROUGH CLAY
          MINERAL COLUMNS
Chemical
constituent
Pb
Zn
Cd
Hg
Fe
Si
K
NH4
Mg
COD
Na
Cl
B
Mn
Ca
Mean
attenuation
number
99.8
97.2
97.0
96.8
58.4
54.7
38.2
37.1
29.3
21.3
15.4
10.7
-11.8
-95.4
-656.7
Principal
attenuation
mechanism
Precipitation/Exchange
Precipitation/Exchange
Precipitation/Exchange
Precipitation/Exchange
Oxidation-Reduction
—
Cation Exchange
Cation Exchange
Cation Exchange
Microbial Degradation
Cation Exchange
Dispersion
Artifact?
Elution from Clay
Exchanged from Clay
Relative
mobility
Low


Moderate
High
More Eluted
Than Applied
                             21

-------
materials is also a possible attenuation mechanism. Experiments in which the
DuPage leachate was filtered through 0.45 ym pore size membranes indicated
that only relatively small amounts of metals were retained on the membrane.
It was therefore concluded that filtration was not an important attenuation
mechanism for the heavy metals in this study but may be an important mechan-
ism for many other leachates.

     Moderate attenuation was observed for the leachate constituents Fe, Si,
K, NHit, and Mg, which had values ranging between 58.4% and 29.3% attenuation.
Little attenuation was found for COD, Na, and Cl, values for which ranged be-
tween 21.3% and 10.7% attenuation. The elements Ca, Mn, and B were not at-
tenuated by the clays, but, rather, were found in substantially higher con-
centrations in the column effluents than in the influent leachate.

     To determine what mechanisms were responsible for the observed differ-
ences in attenuation for each leachate constituent, the effect of clay type
was investigated. The mean attenuation obtained for each clay and chemical
constituent is tabulated in Table 5. The results show that no significant
difference in attenuation for the three clay minerals was observed for the
heavy metals (Pb, Cd, Hg, Fe, and Zn) or for B or Cl. Illite and kaolinite
attenuated Si significantly better than montmorillonite. Kaolinite was found
to elute Via in significantly higher amounts than either montmorillonite or
illite. No significant difference in COD attenuation between montmorillonite
and illite was found, and both attenuated COD significantly better than
kaolinite. Montmorillonite was found to attenuate the cations Na, K, NHi* , and
Mg and to elute Ca to a significantly greater degree than illite and kaolin-
ite.

     A precipitation mechanism for the four heavy metals (Pb , Cd, Hg, and
Zn) , as discussed above, is consistent with the fact that no differences in
attenuation were found among the three clays. Because these four metals exist
in solution as cations, a significant clay-type effect would be expected if
cation exchange were the attenuation mechanism. No clay-type effect was ob-
served, which is taken as additional evidence that the primary attenuation
mechanism for these four heavy metals is precipitation.
     To determine whether the higher attenuation of the cations Na, K,
and Mg and elution of Ca by montmorillonite was due to its higher cation ex-
change capacity, the attenuation numbers were plotted as a function of CEC
for the three clay minerals. The results are shown in Figures 4 and 5. The
very highly significant linear regression of attenuation numbers as a func-
tion of CEC led to the conclusion that the principal attenuation mechanism
for Na, K, NHij , and Mg was the cation exchange of these constituents for Ca.

     Ca was the predominant exchangeable cation present on these clays at the
beginning of leaching (Table 1). To further confirm that this mechanism was
responsible, a mass balance for these five cations was computed for data
presented in Figure 2. The sums of the amounts of Na, K, NHi» , and Mg removed
from the leachate agreed within 3% with the amount of Ca eluted. Increases in
the concentration of alkaline earth metals in groundwater preceding a leach-
ate plume have been observed in the field during monitoring of the ground-
water chemistry around landfill sites. Such increases, termed the "hardness

                                     22

-------
TABLE 5.  MEAN ATTENUATION NUMBER  (ATN) OF SOME CHEMICAL CONSTITUENTS
          FOUND IN MUNICIPAL LEACHATES FOR THREE CLAI MINERALS
Mean ATN
Chemical
constituent
Pb
Zn
Cd
Hg
Fe
Si
K
NHt,
Mg
COD
Na
Cl
B
Mn
Ca
Montmorillonite
99.6*
97.7
96.7
98.4
34.8
39.2
58.9
54.8
48.2
24.6
20.6
9.3
-16.lt
-73.2
-885.5
Illite
100.0
98.6
100.0
98.1
82.8
81.6
31.0
31.0
19.7
23.2
16.4
13.5
-12.8
-6.4
-233.3

Kaolinite
99.9
98.1
97.5
95.2
67.6
71.2
23.2
25.1
18.1
16.2
9.7
14.3
-11.5
-266.2
-190.2

All
columns
99.8
97.2
97.0
96.8
58.4
54.7
38.2
37.1
29.3
21.3
15.4
10.7
-11.8
-95.4
-656.7
 *Underlined  means  are not  significantly  different  (0.05).
 tMinus numbers  indicate  elution.
                                  23

-------
      K-ATN = I5.1 + 5.5 CEC
   70-
   10-
           r=
           ' •
   90
                         CEC
             r = .
                 CEC
                                   70-
                                   20-
              NH4+-ATN = I6.
                       r = .96
                                  15   0

                                   100
               Mg-ATN=l0.4+5.l CEC
                      r = .95
                                   80-
                                   70-
                                   60-
                                   50-
                                    0
                                  15  0
                                   40-
                                   30-
                                   20-
    •  Sand
    A  Illite
o Kaolinite
® Montmorillonite
CEC

  a Clay mixtures
Figure 4.   The attenuation number related to cation exchange capacity  (a)
           K,  (b) NHlt, (c) Na, and (d) Mg.
                                  24

-------
          -2400
                    Ca-ATN = 48.8-i49.8 CEC

                          r=.97
            200
           -600
           -500-
           -400
         o
         o>
         Hi
         c
                                       CEC
                  Mn-ATN = -50.6-l6.6

                          (% Kaolinite)

                        r = .95
                                        Mn-ATN=-O.I-5.93
                                                (% Montmorillonite)

                                               r = ,95
                • Sand

                A Illite
 20       30       40

    Clay  content (%)


O Kaolinite

® Montmorillonite
                                                         50
                                                                  60
H Clay mixtures
Figure 5.   (a) Ca  attenuation number related to  cation exchange capacity.

             (b) Mn  attenuation number related to  clay percentage.
                                        25

-------
halo," are likely due to the same mechanism responsible for the Ca elution
from the columns in this study. This is discussed further in Section 6.

      No significant linear regression of attenuation as a function of either
CEC or clay percentage was obtained for Fe, Si, COD, Cl, B, or Mn, except for
a very highly significant linear regression obtained for Mn elution and clay
percentage. The results of the regression of Mn elution as the percentage of
kaolinite arid montmorillonite increased are presented in Figure 5b. No sig-
nificant linear regression could be obtained with illitic clay because only
two illite columns were used. The data illustrated in Figure 5b show that
approximately three times as much Mn eluted from the kaolinite-containing
columns as from the montmorillonite columns. The data presented In Table 1
indicate that kaolinite contains only slightly more total Mn than mdntmoril-
lonite. However, surface Mn is three times more abundant on kaolinite than on
montmorillonite and corresponds to the increased elution from the kaolinite
columns. No correlation of Mn elution was observed with CEC^ and the amounts
of surface-extractable Mn correlate with the amounts eluted from the various
columns. These facts, along with the anaerobic conditions in the columns,
have led to the conclusion that Mn elution is due to the reduction of surface
coatings of Mn compounds on the clays to more soluble reduced ionic species.
The increase in Mn elution from kaolinite columns is proportional to, and
apparently due to, the larger amount of Mn on the surface, where it is readi-
ly available for reduction and solubilization by the anaerobic leachate.

     The behavior of Fe is similar to that of Mn; however, it is more sensi-
tive than Mn to the oxidation potential in the leachate. During the early
stages of leaching, Fe was also solubilized and eluted from the columns, as
is shown in Figure 3. However, unlike Mn, Fe showed a net attenuation of
58.4% when the entire 10 pore volumes leached were considered. The results of
this study, therefore, indicate that Fe may be either eluted or not attenu-
ated by clay liners if the leachate is strongly anaerobic, or it may be
strongly attenuated under weakly anaerobic conditions.

     Other data from this study indicate that attenuation of COD was rel-
atively low after passage of DuPage leachate through clay columns. This re-
sult is in agreement with those of Urioste (1971) , who reported poor removal
of COD when leachate was ponded on soils. The lack of a strong clay effect
arid of a significant correlation of attenuation of COD with either clay per-
centage or CEC indicate that the observed attenuation was probably the result
of microbial degradation of the organic compounds. The fact that the leachate
is relatively old probably accounts for the relatively low reduction in COD.
It has only a small percentage of readily degradable organics and a low nu-
trient status, owing to the lack of PO^ (Table 2).

     Chloride attenuation was also relatively low, only 10.7%. This low at-
tenuation is not surprising, because Cl is considered as a mobile non-inter-
acting anion in soil systems. The low Cl attenuation was not a function of
the type or amount of clay mineral present and is attributed to physical
dispersion in the porous column media, with perhaps a small amount of inter-
action at anion exchange sites on the clay, or to other chemical reactions.

     Since Cl is a negative ion, it would be attracted only by the positive


                                     26

-------
charge at the edge of the clay minerals. Oxygen  and OH  are the negative ions
of the lattice, and any other negative  ion would have to be nearly the same
size as the oxygen ion in order to  coordinate with and  continue or substitute
for the oxygen. Because the chloride  ion is  about two and one half times the
volume of the oxygen ion, it is too large to replace or coordinate with the
oxygen and hydroxyl ions, although  fluorine  can  do so because it is about the
same size as the oxygen and hydroxyl  ions.

     No attenuation of B was observed in this study; rather, a small elution
of boron from the columns took place  throughout  the leaching experiment. This
may be interpreted as a slight solubilization of B from the clay minerals or
sand in the columns. However, no  effect of type  or percentage of clay and no
changes in elution with leaching  time were found during the study. These
facts have led  to the conclusion  that the B  results may be an artifact of the
experiment. It  has been suggested that  B may be  dissolving from the borosili-
cate glass tubing used throughout the apparatus. More importantly, B may be
solubilized from the spun glass wool  used as a filter to keep sand and clay
from migrating  into each of the five  manometers  and the outflow tubing (Fig.
1). The high surface area of the  spun glass  and  the neutral pH of the leach-
ate make it plausible that the boron  elution may be due to contamination. In
any case, no attenuation of boron was observed in this  study.
     The results presented in Figure  4  and Table 5 permit the three clay
minerals to be  ranked, according  to their overall attenuating ability for the
chemical constituents found in municipal leachate, as follows:

                montmorillonite > illite > kaolinite.

     The montmorillonite used in  this study  has  properties similar to those
of smectites produced by weathering of  micas, chlorites, and other crystal-
line minerals.  The cation exchange  capacity  is lower than that for montmoril-
lonite produced from weathering of  volcanic  ash  and basaltic rocks.

     The montmorillonite used produces  clay  material (e.g., Porter's Creek
Clay) that have much higher permeabilities than  those from Wyoming-type mont-
morillonite. Such high permeability should make  this montmorillonite more
useful than some other types of montmorillonite  in making landfill liners in
humid climates. With its high cation  exchange capacity, it would adsorb more
of the cations  in the leachate than either illite or kaolinite. With its
greater permeability, it allows more  water to pass through the liner than do
many other types of montmorillonite.  The low permeabilities of other mont-
morillonite types would probably  increase the hazard of lateral seepage from
the sides of the landfill in humid  climates.

     Some sodium montmorillonites tend  to' shrink when sodium is exchanged for
divalent and trivalent cations or when  salt  concentrations are high. This
shrinkage sets  up tension, which  in turn produces cracks called syneresis
cracks. Syneresis cracks were common  in irrigation ditches in Colorado that
were lined with sodium montmorillonite  (personal communication, R. D.
Dirmeyer, Jr.,  1961). Shrinkage could be reduced considerably by using cal-
cium montmorillonites and mixing  them with other earth  materials  (e.g.,  16 to
32% montmorillonite and 68 to 84% sand).
                                      27

-------
     The kaolinite used in this study is a fine-grained material in which the
crystallinity of the kaolinite crystals is poor. It has a high cation ex-
change capacity compared with other kaolinites. Well crystallized kaolinites
with large crystals have a cation exchange capacity of 1 to 5 meq/100 g. The
kaolinite used in this study would be better suited than most kaolinites for
landfill liners. The permeability of this kaolinite is lower than that of
well crystallized kaolinite with large crystals. The kaolinite in most sedi-
ments has a cation exchange capacity and permeability between those of the
kaolinite used and the well crystallized kaolinites.

     The illite used is similar in cation exchange capacity and permeability
to the illite found in most sediments. The sediment from which the experi-
mental illite was taken contained only one clay mineral — the illite —
whereas most sediments that contain illite also contain other clay minerals.

     It was concluded that the attenuation order was due principally to the
cation exchange capacity of each of the three clays.

Attenuation of Leachate

     During the period of time from collection of the leachate through the
establishment of hydraulic equilibrium, the leachate was stored in a re-
frigerated (3 to 5° C) condition with either argon or sterilant gas being
purged slowly over the top of each respective drum. Chemical analyses were
performed weekly on the leachate to monitor possible changes in composition.
As a result of this monitoring, it was determined that the COD of the sterile
drum was steadily rising. It was determined that the active component of the
sterilant gas, ethylene oxide, was able to react with the chloride ion pre-
sent in the leachate which acted as a nucleophile to produce ethylene chloro-
hydrin (Rosenkranz and Wibdkowski, 1974). When it was discovered that the COD
was rapidly rising, the use of the sterilant gas was discontinued. Argon was
then used as the purge gas for both drums and mercuric nitrate salt was
added to the sterile drum to maintain sterility. Further monitoring of both
drums was continued at approximately 1 week intervals during the 10 month
period during which leaching of the columns occurred. The COD values were
found to remain constant at the value obtained when purging of sterilant gas
was discontinued. The value reported in Table 2 for COD, and for all the
other constituents, is the average of the 37 separate analyses performed
during the 10 month period of leaching.

     Determination of the attenuation of chloride and the other major com-
ponents of DuPage leachate has been described in Section 5. The results of
this study indicated that there was an average 6% greater attenuation of
chloride in the columns leached with sterile leachate than those leached with
the natural leachate. This greater attenuation is attributed to the reaction
of chloride with the ethylene oxide to form ethylene chlorohydrin. Other than
the slight increase in chloride attenuation, no other significant difference
was apparently due to the increase in COD in the sterile leachate as compared
to the natural leachate.

     There were, however, other significant differences between the sterile
and natural leachate treatments which were not attributed to the higher COD

                                     28

-------
of the sterile leachate. Figure  6  illustrates  the  difference observed in Mn
elution from the colums. It  should be noted  that a negative attenuation num-
ber indicates that more Mn eluted  from  the column  than was present  in the
influent leachate. It  can be seen  that  much  higher levels of Mn were found in
effluents from the columns leached with natural leachate. It was concluded in
Section 5 that this elution  was  due  to  reduction of  surface coatings of Mn
oxides on the clays by the anaerobic leachates. This  conclusion is  further
verified by the difference in Mn elution between the  natural and sterile
leachate treatment. This difference  is  attributed  to  the stronger anaerobic
environment provided by the  active microorganisms  present in natural leach-
ate. Inspection of the data  in Table 2  shows that  the average Eh (oxidation
potential) reading of  the natural  leachate was 10  times lower than  the
sterile leachate, even though both were well in the  anaerobic range (Eh read-
ings less than 197 m.v. are  considered  to reflect  anaerobic conditions). A
similar result was obtained  for  Fe in that significantly (.05 level) greater
mobility of Fe was found in  columns  leached  with natural leachate as compared
to those leached with  sterile leachate. A mechanism  similar to that for Mn is
postulated as the reason for the observed differences between the natural and
sterile leachate.

     Those chemical constituents for which no  significant difference in at-
tenuation between the  normal and sterile leachate  was found were Ca, Mg, Na,
K, NHit, Pb, Hg, Zn, and Cd.
POLLUTION HAZARD  OF LEACHATE

      In  addition  to determining the relative  mobilities  of the various chemi-
cal  constituents  of leachate through clays  (liner materials) , their relative
pollution hazard  should be evaluated.  Ranking wastes  in  terms of their exist-
ing  or potential  threats to public health and/or the  environment has been the
subject  of  the  "Priority Ranking System"  suggested  for development by the
U.S. -EPA (1973).

      The priority ranking formula is:

                                  R  =  Q/CP                              {3}

where R  = ranking factor ,
      Q  = annual  production quantity for  the  waste  being ranked, and
      CP =  critical product for the waste being ranked.

      A critical product is the lowest concentration at which any of the
hazards  of  concern become manifest in a given environment, multiplied by an
index representative of the waste's mobility  into that environment. Thus, for
a municipal leachate that would be discharged through a  clay liner or soil
into an  aquifer used for drinking water,  the  toxicity factor could be the
public water supply limits for the given  element, and the mobility index
could be the attenuation numbers derived  above.

      Evaluation of the critical product for municipal leachates moving
through  soils or  clay liners by the U.S. -EPA  (1973) formula  proved to be awk-
ward and unsatisfactory for several reasons.  The  first problem encountered


                                      29

-------
   -600
       0
         O   2  4
8
 16
Koolinite  (%)
Figure 6.  Manganese elution related  to  percentage  of kaolinite leached
           with natural and sterile leachate.
                                  30

-------
was that negative attenuation numbers,  such as  those  obtained  for  calcium,
were not accounted  for by  the formula.  A transformation of  the data would 'be
necessary to express the negative  numbers in a  manner that  could be used in
the CP formulation.

     The second problem encountered was evaluation  of the CP for the four
heavy metals. An upper boundary on the  attenuation  number,  such as the  100%
removal, used to express the heavy metal removed from leachate is not allowed
conceptually by the formula. Instead, what is required is the  actual attenu-
ation as determined by leaching until "breakthrough"  of the particular  ele-
ment is achieved. For Pb ,  an estimated  300 pore volumes would  have to be
leached through an  average column  to achieve breakthrough.  It  was deemed im-
practical to actually measure the  breakthrough  of the heavy metals, and esti-
mating their breakthrough  required additional analytical data  that could
cause errors in estimating the  pollution hazard.

     A third problem with  evaluating the pollution  hazard by using the  CP was
that it was not specific for the waste  being evaluated in that the concen-
tration of the element of  interest in the waste did not enter  into the  evalu-
ation of the hazard. That  this  is  a serious fault can be illustrated by
simple examples: an element with a relatively high  toxicity and mobility in-
dex  could get a high hazard rating, even though only  a trace was present in
the  waste; conversely, an  element  with  a relatively low toxicity and mobility
could receive a low hazard rating  even  though it was  present in very high
concentrations .

     A fourth criticism  of the  CP  rating was that it  was conceptually il-
logical because large CP values indicated a low pollution hazard and, con-
versely, a very small number represented a very high  hazard. It was con-
sidered more logical to  express high pollution  hazards as large numbers when
a relative scale was used  or when  pollution potentials were evaluated.

     To overcome these objections  to the CP formulation of  a pollution hazard
index for municipal leachates,  the ranking equation was changed as follows:

                                R  =   (Q) (HI)                            {4}

where R and Q are as previously defined and
      HI = the pollution hazard index for the waste.

     The pollution  hazard  index (HI) is a toxicity  index for the element
within a given leachate, multiplied by  a mobility index for the element in a
particular leachate-clay system. The pollution  hazard for the  whole leachate
is that for the constituent with the highest hazard within  the particular
leachate .

                           HI  =  ()  (100 - ATN)                        {5}
where  C  = the effective concentration of the chemical constituent,
       DWS =  the drinking water standard (U.S. -EPA,  1972),  and
       ATN =  the attenuation number for the given element.

                                      31

-------
     The effective concentration is defined as the concentration of the
chemical constituent in the leachate plus the concentration of the constitu-
ent that may be leached from the soil or clay. When attenuation is occurring,
the effective concentration is merely the concentration of the constituent in
the influent leachate. When elution from the columns is occurring, as it did
for the three elements B, Ca, and Mn, the effective concentration is the
leachate concentration plus the concentration eluted from the column. Table 6
presents the 15 chemical constituents for which ATN values are available,
ranked according to their pollution hazard, as determined by equation 5.

     The results in Table 6 give a reasonable ranking of the chemical con-
stituents in terms of what would be expected from a gross overview of the
data. The ranking system has the advantage of quantifying the expected pollu-
tion hazard of a given leachate and allows comparisons of the pollution
hazards of one leachate with another. The ranking system also focuses atten-
tion on the chemical constituent with the highest pollution potential. In the
case of DuPage leachate, NffiJ" was found to have the highest pollution hazard.
In a fresh leachate, COD might be expected to have the highest pollution
potential. However, in the DuPage leachate the hazard index clearly indicates
that NHtj is a pollution hazard about 30 times greater than any other con-
stituent found in this particular leachate.

     We feel that the proposed pollution hazard ranking system for municipal
leachates (equation 5) overcomes the objections posed previously for the CP
component of the Priority Ranking System. The toxicity index can in most
cases be computed readily from a chemical analysis of the leachate.

     The evaluation of the toxicity index is flexible in that drinking water
standards need not be the criteria. LDso values, or some other toxicity eval-
uation, can be used in place of drinking water standards. What we think is
important is the computation of the ratio of the actual waste concentration
relative to whichever toxicity evaluator is used. The mobility index, how-
ever, must be determined experimentally or estimated from the data presented
in this paper.  The results of this study indicate that the mobility index
will be a function of the CEC of the earth material, the cations present
initially on the exchange complex, the chemical composition of the leachate,
and the pH of the leachate.
                                     32

-------
TABLE 6.  CHEMICAL CONSTITUENTS IN DUPAGE LEACHATE, RANKED ACCORDING TO
          POLLUTION HAZARD
Chemical
constituent
NHi,
B
COD
Hg
Cl
Ca
Cd
Fe
Na
Mn
K
Mg
Pb
Zn
Si
Effective concentration
D. W. standard
862/0.5
(29-9 + 3.5)/1.0
1340/50
0.87/0.002
3484/250
(46.8 + 307.3)/250*
1.95/0.01
4.2/0.3
748/270
(0.02 + 0.02)/0.05
501/250*
233/250*
4.46/0.05
18.8/5.0
14.9/250*
Toxicity
index
1724.
33.4
26.8
435.
13.9
1.42
195.
14.0
2.77
0.78
2.00
0.93
89.2
3.76
0.06
Mobility
index
62.9
111.8
78.7
3.2
89.3
756.7
3.0
41.6
84.6
195.4
61.8
70.7
0.2
2.8
45.3
Hazard
index
108,440.
3,734.
2,109.
1,392.
1,241.
1,072.
585.
582.
234.
153.
123.
65.7
17.8
10.5
2.7
*Actual value not established by EPA; therefore it was assumed to be -the
 same as chloride.
                                   33

-------
                                  SECTION 6

                  CONFIRMATION OF LABORATORY COLUMN STUDIES
                       BY COMPARISON WITH FIELD DATA1
ABSTRACT
     The results of the laboratory column leaching experiments were checked
at the DuPage County sanitary landfill and at other existing landfills where
detailed field data are available. These field data clearly show a "hardness
halo" corresponding to the Ca release observed in the column experiments. The
relative attenuation rates of some of the ions are also confirmed by the
field data.

     Laboratory results show that the leachate reduced the hydraulic conduc-
tivity of the columns during the experiment. Although similar change in field
hydraulic conductivity was not clearly demonstrated, the field data suggest
that it took place.

     These results suggest that overall pollution from landfill leachate
would be reduced by designing earth material landfill liners for higher per-
meability. Properly designed liners would selectively attenuate the toxic
pollutants from the leachate and allow the ground water to dilute the non-
toxic components which can be tolerated at much higher concentrations without
harmful effects. The study raises some basic questions on monitoring systems
and landfill design which should be addressed by regulatory agencies re-
sponsible for environmental quality.

INTRODUCTION

     This paper principally relates two phenomena noted in the laboratory to
field observations around sanitary landfills: the elution of large amounts of
the calcium ion from the study columns (Fig. 3) and the reductions in hy-
draulic conductivities which resulted from the introduction of leachate to
the clay-sand mixtures. It is felt that the testing of laboratory results in
field situations is necessary before the data may be used in sanitary land-
fill design.

RESULTS AND DISCUSSION

Hardness Halo

     Figure 3 (Section 5) illustrates the negative attenuation or elution of

Authors: K. Cartwright, R. A. Griffin, and R. H. Gilkeson.


                                     34

-------
Ca, Fe, and Mn from the columns containing  2,8,  and  16% montmorillonite clay
in sand. The relative concentrations  greater  than 1 indicate that Ca and to a
lesser extent Fe and Mn are eluting from the  column at much greater concen-
tration than the influent leachate at various pore fractions. The area under
the Ca curves can be seen to  increase in proportion to the percentage of clay
in the column, was quantified in  Section 5  by integrating between pore frac-
tion 1 and 11, and was assigned a relative  attenuation number (ATN) as shown
in Table 4 (Section 5).

     The elution of Ca from the columns  was attributed to an ion-exchange
mechanism; the replacing of the Ca bonded to  the  clays at their cation ex-
change positions by other ions in the leachate.

     Soils in much of Illinois are carbonate  rich, with the clays generally
having Ca in the cation exchange  position,  and have free carbonates in all
except the leached zone of the soils. The presence of excessive hardness in
the vicinity of sources of pollution  has appeared in  a number of articles,
but is rarely discussed as to its origin. An  example  is found in a DuPage
County study  (Fig. 7, from Zeizel et  al.,, 1962).  There are two areas of the
county where the hardness, as CaCOs,  in  the shallow carbonate aquifer exceeds
1000 parts per million  (ppm). The eastern area is a fairly heavily developed
residential area and the glacial  drift,  which protects the aquifer from pol-
lution, is relatively thin. No specific  source of the high hardness can be
established; however, it is most  likely  due to a  high concentration of home
septic systems. The western area  of high hardness near West Chicago is
thought to have resulted from the discharge of large  volumes of waste chemi-
cal salts to surface ponds.

     Other examples can be found  in Anderson  and  Dornbush's (1967) study of a
sanitary landfill in South Dakota, and Walker's  (1969) discussion of ground-
water pollution in Illinois.  Most recently, Henning et al. (1975) showed high
calcium in monitoring wells very  close to a landfill  trench at Mentor, Ohio;
the Ca concentrations both decreased  with distance from the fill and were
lower in the refuse than in the closest  wells.

     Hughes et al.  (1971) published the  results of studies of five landfills
in northeastern Illinois, including the  Old DuPage County landfill from which
the leachate was taken for this study. Monitoring continued for three years
following the completion of that  report. Figure 8 was drawn, using data from
the Winnetka and Old DuPage landfills.

     The Winnetka data  (top,  Fig. 8)  shows  considerable scatter. This may be
partly due to a mixture of points, some  in  the fine-grained alluvium and some
in the glacial till which have somewhat  different properties. However, these
data suggest that the hardness approaches background  within 9 to 15 meters of
the refuse which is somewhat  less than the  distance that the chloride ion
travelled (Hughes et al., 1971).  Note that  the four data points from piez-
ometer nest LW3 follow this pattern.

     The till under the Old DuPage landfill clearly illustrates the increase,
then decrease in hardness. The till is separated  from the refuse by 1 to  1.5
meters of sand. The hardness  returns  to  background within about 1.5 meters of

                                      35

-------
            R 9 E
                                    R 10 E
                                                           R II E
                                   BLOOMINGDALE
                                                       WOOD DALEJ

                                                            r-BENSENVlLL
                                                          LU
                                                          ...
                                                          .1
                                                               ^ELMHURST

                                                            VILLA

                                                            PARK
            WARRENVILLE
                                                   DOWNERS
                                                 *t"1    j V •_. •_. -
                        NAPERVILLE

   HARDNESS  (as CoC03)


      I More than 1000
        500-1000



        300-500



        Less than 300
0
                    5    6  Miles
Figure 7.   Hardness of water,  expressed as CaC03 ,  in  the Silurian Dolomite

            aquifer in DuPage County,  northeastern  Illinois (from Zeizel  et


            al.,  1962).
                                       •

-------
1500-
1000-
i
500-
0-
c
|sooo-
Q.
1*1
O
o 4000-
0 i
w
O
(/)
(rt r\
4, 0-
£ <
k_
o
I
1600-
1200-
800-
i
400-
n
C
Winnetka Landfill-Alluvium and till
• A A Alluvium
V~"\ T Till
/"" \ S Silurian bedrock
*/ \ ^Leaky seal ? * Vertical profile,
/*j A ^ ®^T ' hole LW3
Average value \
of leachote V
\
\ »T
• T N
.T V"-- S T «S S
*T «T «g 	 •-• 	 •-—•--^Background
i i i i 	 	 r 	 1 	 1 —
) 6 12 18 24 30 36 42
Distance (meters)
5 Old Du Page Landfill- Till
Y\
/ \
^Average^V^* « « « ~
volueofV? ! 3 5 5
leachate*v« 	 • 	 • — • 	 •-». Background
3 1.2 2.4 3.6 4.8 6.0 7.2
Distance (meters)
? Old Du Page Landfill - Sand, south side only
•I (oldest part of fill)
/I
/ \
!-\
f X
/ :x^
X^ "~"— 	 	 c
value of leachate
6 12 18 24 30 36 42
Distance (meters)
Figure 8.  "Hardness halo" effect shown as a function of distance (m) from
           the Winnetka and DuPage landfills in northeastern Illinois.
                                    37

-------
travel, about half the distance of the estimated travel of the chloride ion
(Hughes et al., 1971). Note, in particular, the values shown for piezometer
nests LW5 and LW6. These data points are all for the younger, northern part
of the fill; data from the older parts of the fill do not fit the same curve
(all hardness concentrations fall too low).

     The surficial sand transmits leachate-contaminated water south from the
older parts of the DuPage landfill and shows a similar hardness distribution.
The hardness levels are much lower in the old refuse in this area, and all
the values reflect that lower concentration. The hardness returns to back-
ground levels within 9 to 15 meters; however, chloride ion has moved 240 to
300 meters in the permeable sand layer.

     All these data clearly show the presence of the "hardness halo" result-
ing from the movement of leachate into the surrounding till and sands, and
that the rate of the hardness front is less than that of the chloride ion.
The chloride ion is probably the best tracer of this type of pollution in
this environment. The distance of travel varied from slightly less to ap-
proximately 10% of that of the chloride ion; this probably is controlled by
the nature of the materials, cation exchange reactions, concentrations in the
leachate, and ground-water flow rates. Nevertheless, all the data show an in-
crease in the hardness of the water in the sediments over that in the leach-
ate.

Hydraulic Conductivity

     The results of initial hydraulic conductivity and bulk density measure-
ments of the column contents in the laboratory study are presented in Table
3. These data indicate that a wide range of hydraulic conductivities, with
values in agreement with those expected under field conditions from similar
materials (Todd, 1959) were observed. The relatively high bulk densities and
slow flow rates used in this study closely simulate the conditions observed
in the field. This lends credence to the extrapolation of the results and
conclusions of the laboratory studies to those obtained in the field.

     During the initial stage of the experiment, the columns were leached
with deionized water until steady state conditions were achieved. The columns
containing low percentages of clay reached hydraulic equilibrium relatively
rapidly while the high percentage clay columns required leaching for well
over a month to achieve steady manometer readings. When hydraulic equilibrium
was achieved, as indicated by steady flow rates and relatively constant man-
ometer readings from the five manometers located over the entire length of
each column, the leachate was added to the columns.

     Significant reductions in hydraulic conductivity and significant dif-
ferences in the reductions between natural and sterile leachate were ob-
served. The results of hydraulic conductivity changes observed in columns
containing montmorillonite, kaolinite and illite clays are presented in
Figures 9, 10, and 11. The data presented in these figures were statistically
smoothed using the five-member moving-average method to show more clearly the
trends in the data. The raw data were statistically analyzed to determine
whether significant differences in hydraulic conductivity occurred between

                                     38

-------
3
2
1
0
-1
-2
• o
1-3
0
° -4
1-6
L.
c
V
O> |
c
5 -1
0
-1

0
-1
(
: /

''\
T3

-------
 o
 d).
 o
 o
  D

  O
 CD
 o>
 c
 o
6
5
4
3

2
1
0
-1


-3
-4
-5




-9
0
-1
-2
-3
1
0
-1
-2
-3
0
-1
_2
0
_|
I
0
_ 1
I



-
-

/
/
/
I T


•u
s««— -•"—• *N
N Nq 4%

^•^x.....
*"~'v. \ . . .
\. , •
N- 	 	 	 8%
	 -• M
~:^L-^
:rt=«>~.._.
"- 	 	 16%
^-^•=^^_^ 32%
***=*=><&=: -. 	 _ 	 640
	 • 	 -~— --, 	 Of /o

1 i I 1 i
23456

Natural
Sterile











Montmor



















I lonite





Montmoril lonite


	 S





Montmori 1 lonite

. o
Montmorillonite
' • N
Montmorillonite
	 . N
	 . 	 , 	 . g
Montmorillonite 5
==!-_ "
1
7
— -==t N
1 1
8 9
                            Leoching time (months)
Figure 10.  Hydraulic  conductivity of montmorillonite-sand  columns as a

            function of  leaching time.
                                    40

-------
~ 0
0
0) 1
in ~"
E-2
o .,
>._4
1 ~5

D -6
^J
c -7
o
o
o
D O
o

T>
.C
-- 0
O) I
o> '
c
o
£:
l \
°
\


T3
_ *
— O

i


*

_

•^=





|
D 1
~~X. 	 • 	 N Natural
^
\ \ 	 S Sterile
"A "---
\ v^
* "*^ "**.
\ ~-— -s
\
•
\ 4% lllite
•»* -- 	 	 g
* ^*^-^^ .
* i\l

^.. — 	 . 	 	
^•^-... S 16% lllite
"*^^^»
X • 	 -- fvl

c!"~'*-*

s. — .*_•--!-.• -N IOO % Sana
. 5


1 1 1 1 11 1 1
23456789
                      Leaching time  (months)
Figure 11.   Hydraulic conductivity of illite-sand  columns as a function
            of  leaching time.

-------
sterile and natural leachate. The data from the columns containing 4% kaolin
ite were rejected from the analysis when they were found to deviate by more
than 3 standard deviations from the overall mean change in hydraulic conduc-
tivity observed for all other columns. The manometer readings indicated that
the out-flow tubes were plugged. The reason why only 4% kaolinite treatments
had this problem is not clear.

     The results of the statistical analysis indicated that columns leached
with natural leachate had significantly (.05 level) greater reductions in
hydraulic conductivity than those leached with sterile leachate. This result
is illustrated clearly in Figures 9, 10, and 11. Furthermore, the statistical
analysis showed that columns containing montmorillonite had significantly
greater average reductions in permeability than kaolinite or illite and that
there was no significant difference between kaolinite and illite. This result
is consistent with the swelling nature of montmorillonite clay and is not
surprising.

     Field data to support this conclusion are not as clear as for the "hard-
ness halo." Fewer field tests have been made for hydraulic conductivities
than chemical tests for water quality. In addition, field tests may be ac-
curate only to approximately a half-order of magnitude.

     At the Winnetka landfill  (Hughes et al., 1971), 23 hydraulic conduc-
tivity tests were conducted, 7 on refuse, 4 on alluvium, and 7 on till. These
data are too scattered to show any significant differences with distance from
the refuse.

     At the Old DuPage County landfill, 34 field hydraulic conductivity tests
were made, 14 on refuse, 14 on sand, and 6 on till. The data on the hydraulic
conductivity of the sand (all south of the fill) suggest some reduction in
hydraulic conductivity. The 10 tests made on monitoring wells less than 6
meters from the fill have a mean hydraulic conductivity of 4.32 x ID"1* cm/sec
(range 1.9 x 10~3 to 1.9 x 10~7), and those monitoring wells greater than 12
meters from the refuse (4 tests) have a mean conductivity of 2.59 x 10~3
cm/sec (range 7.6 x 10~3 to 9.5 x 10"1*). The data are not statistically sig-
nificant, but they do suggest a hydraulic conductivity reduction similar to
that noted in the laboratory.

     The reductions in hydraulic conductivity observed in the laboratory
study with the DuPage leachate seem particularly significant. This is due to
the fact that the DuPage leachate is approximately 15 years old  (Hughes et
al., 1971) and contains a relatively low percentage of organic compounds
which are readily degradable by microorganisms  (Table 2). In addition, this
leachate has a low nutrient status with both phosphate and sulfate being ab-
sent in detectable quantities. Much higher amounts of microbial growth and
plugging might be expected from a younger leachate.

     The results have led to the conclusion that if clay liners, natural or
man-made, of similar composition to those used in this study are used in
municipal landfills, significant reductions in hydraulic conductivity can be
expected due to microbial growth. Further, slightly higher reductions in hy-
draulic conductivity can be expected from montmorillonite clays, apparently

                                     42

-------
due to their swelling tendency. It was also concluded that Mn and/or Fe may
be leached from clay surfaces in substantial amounts under highly anaerobic
conditions but may not be leached under mildly anaerobic conditions.
                                       43

-------
                                  SECTION 7

                   EFFECT OF pH ON EXCHANGE-ADSORPTION OR
                    PRECIPITATION OF LEAD FROM MUNICIPAL
                         LEACHATES BY CLAY MINERALS1

ABSTRACT
      The  capacity  of kaolinite and montmorillonite clay minerals to remove Pb
from municipal landfill leachates and the mechanisms by which removal is
achieved  were studied to evaluate the potential usefulness of clay minerals
as  liners for waste disposal sites under conditions of varying pH and ionic
competition.
             »
      Montmorillonite was found to remove as much as 5 times more Pb from
various solutions  than did kaolinite. Results indicated that Pb removal was
reduced as much  as 85% by leachate when compared to the amounts removed from
pure Pb(N03)2 solutions. A precipitate was found to form in leachates at pH
values above 5 and was identified as PbC03. The complexing capacity of DuPage
leachate  for Pb  was measured and the extent of complexation was found to be
11%.  The  higher-ionic-strength Blackwell leachate had no measurable com-
plexing capacity for Pb. Increased adsorption of Pb(NOa)2 was found to cor-
respond to the appearance of Pb-hydroxyl species in solution.
      It was concluded that Pb removal from solution is primarily a cation ex-
change-adsorption  reaction that is affected by pH and ionic competition. It
was also  concluded that formation of Pb-organic complexes are of secondary
importance in landfill leachates due to competition from high concentrations
of  other  cations.  At pH values above 6, a large increase in Pb removal from
solution  by clay can be expected, due to either  increased adsorption of Pb-
hydroxyl  complexes or formation of PbCOa in landfill leachates.

      The  thickness of clay liners necessary to remove Pb from solutions of
PbCNOs),  0.1 M NaCl, and landfill leachates at concentrations ranging between
10  and 1000 ppm  Pb and at pH values from 3 to 8 were computed. Some undesir-
able  environmental consequences of the reversible Pb exchange-adsorption re-
action with clay may ensue where pH and ionic, competition are unfavorable.

INTRODUCTION

      This section  reports the results of an investigation, the purpose of
which was to determine the capacity of the two major clay mineral types for
removing Pb from solution and the effect municipal leachates have on this
capacity at various pH values. Another purpose of the investigation was to
JAuthors: R. A. Griffin and N. F. Shimp.

                                     44

-------
gain insight into the mechanisms responsible  for  attenuation of Pb as well as
to evaluate the potential use of clay minerals  as liners  for waste disposal
sites.

     Lead was chosen for study because  documented evidence  shows that low Pb
levels in drinking water can cause  death  to humans.  In  one  case in Australia,
94 adults died of chronic lead poisoning  because  throughout childhood they
drank water collected from roofs of houses painted with lead-pigmented paint
(Henderson, 1955). There is also evidence (Broadbent and  Ott,  1957) that or-
ganic chelates form with hydrolyzable metals  and  that they  may make Pb more
mobile in soils or clay liners when municipal leachates are present than in
effluents that do not contain high  concentrations of organic compounds.

EXPERIMENTAL
     -j
     The Pb removal studies were conducted by placing a known  weight of clay,
between 0.100 and 1.000 g, into a 125 ml  Erlenmeyer  flask.  The weight of clay
used was chosen to give an estimated 20 to 50%  change in  the Pb concentration
of the solution at equilibrium. A 50 ml aliquot of either deionized water,
0.1 M NaCl solution, or leachate and then a 2 ml  aliquot  of a  Pb(NOa)2 solu-
tion were pipetted into the flask.  The  pH of  the  solutions  were adjusted with
either HN03 or NaOH over the pH range of  interest, and  the  volumes of acid or
base added were recorded. The volumes added were  usually  less  than 1 ml. The
flasks were tightly stoppered, and  as a result  the COa  liberated from the
leachate solutions caused a slight  positive pressure in the flask, which
aided in maintaining anaerobic conditions during  equilibration. (For a period
of time after addition of acid, the stoppers  were removed to relieve exces-
sive pressure, and then the flasks  were restoppered.) The results of rate
studies indicated that 4 hours were necessary for Pb in leachate to equili-
brate with kaolinite. This result is in agreement with  Beevers (1966), who
found that Pb(NOa)2 solutions equilibrated in from 1 to 12  hours, depending
on the clay mineral. The samples in this  study  were  shaken  for at least 24
hours in a constant temperature bath at 25 ±  0.5° C  to  insure  equilibration.
The equilibrium pH was recorded, the' samples  were centrifuged, and the solu-
tions were analyzed for their Pb concentration  by atomic  absorption. The dif-
ference between the initial concentration and the equilibrium  concentration
was used to compute the amount of Pb removed  from the solution at the par-
ticular pH by a given clay mineral. This  procedure was  Carried out for a
range of initial Pb concentrations  that varied  between  10 and  1,000 ppm.

     The resulting data were plotted as amount  of Pb removed from solution
per gram of clay versus pH. The resulting family  of  curves  allowed construc-
tion of adsorption isotherms for any individual pH value  of interest through-
out the pH range 3 to 6. Representative adsorption isotherms and plots of the
data according to the Langmuir  (1918) adsorption  equation were constructed at
various pH values to determine the  maximum amount of Pb that could be ad-
sorbed from the various solutions by the  two  clay minerals  and to try to gain
insight into the mechanisms of adsorption. The  Langmuir (1918) equation in
its linear form is:

                              -C_   =  -L  +   ±                            {6}
                              x/m     kb      b

                                      45

-------
where C is the equilibrium concentration of Pb, x/m is the amount of Pb ad-
sorbed per unit mass of clay, 1/b is the slope where b is the adsorption
maximum for Pb, and k equals the slope/intercept where k is a term relating
to the energy of adsorption.

     The ability of the two municipal leachates to complex metal ions, in
particular Pb, was studied. The leachates were centrifuged at 1000 rpm for 10
minutes and then filtered through a 0.45 \m pore-size Millipore membrane held
in an anaerobic bacteria filter holder under argon pressure. The leachate ob-
tained after centrifugation and filtration was considered to contain only
soluble organics and was used in the complexation studies. Successive ali-
quots of Pb were added to the leachate and equilibrated for several hours.
The concentration of free and complexed lead ions in solutions were deter-
mined from pulse polarographic wave heights using the methods and equipment
described by Gadde and Laitinen (1973a).

     Lead was removed from leachate solutions as a white precipitate at pH
values greater than 6. The precipitate was separated from solution on a 0.45
lam Millipore membrane, washed with deionized water and dried at room tempera-
ture. The chemical compound was then identified from its X-ray diffraction
pattern.

RESULTS

     The results of Pb removal from 25° C solutions of DuPage leachate by
kaolinite and montmorillonite clay minerals were plotted as a function of pH
in Figures 12 and 13, respectively. Similar results obtained for Pb removal
from Blackwell leachate by kaolinite are shown in Figure 14.

     The data presented indicate that Pb removal from landfill leachate in-
creases with increasing pH values and with increasing concentration of Pb in
solution. Increasing Pb concentration is indicated on the figures by increas-
ing alphabetical order, and the initial Pb concentration, weight of clay
used, micrograms of Pb added, and volume of solution that correspond to the
alphabetical designations are given in Table 7. A blank (no clay) solution of
leachate with Pb added was carried along through the experiment and the re-
sults of these also appear on the figures.

     Data for Pb removal from leachate plotted as a family of curves of in-
creasing concentration have the advantage that sorption isotherms may be con-
structed from the plot, by using the information given in Table 7, for any
desired pH value from pH 3 to pH 6.

     Pb sorption isotherms can be constructed from these plots by first
placing a vertical line across the family of curves at the pH of interest.
The amount of Pb removed from solution is found on the graph at the points
where the vertical pH line intersects each curve. The equilibrium Pb concen-
trations that correspond to the chosen pH value are then computed from the
amounts of Pb removed at each concentration, as determined from the graph and
the information for each Pb curve given in Table 7 by using the following re-
lation:
                                      46

-------
   7.500
    7.000
Figure 12.  The amount of Pb removed  from DuPage leachate by kaolinite at
            25° C plotted as a  function of pH.
                                    47

-------
   6.000
Figure 13.  The amount of Pb removed from DuPage leachate by montmoril^
            lonite at 25° C plotted as a function of pH.
                                  48

-------
 5.000-
                                                     5.5    6.0    6.5    7.0
2.5
Figure 14.  The amount  of  Pb  removed from Blackwell leachate by kaolinite
            at 25° C plotted  as  a function of pH.
                                    49

-------
TABLE 7.  Pb REMOVAL PARAMETERS USED TO COMPUTE SORPTION ISOTHERMS FROM
          52 ml REACTION VOLUMES
Curve
DuPage leachate -
kaolinite (Fig. 2)








DuPage leachate -
montmorillonite








Blackwell leachate -
kaolinite (Fig. 4)




A
B
C
D
E
F
G
H
I
J
A
B
C
D
E
F
G
H
I
J
A
B
C
D
E
F
Initial Pb
concentration Pb added
(ppm) (micrograms)
9.62
19.23
38.46
57.69
76.92
96.15
115.38
153.85
192.31
384 . 62
9.62
19.23
38.46
57.69
76.92
96.15
115.38
153.85
192.31
384.62
19.23
38.46
76.92
96.15
192.31
288.46
500
1,000
2;000
3,000
4,000
5,000
6,000
8,000
10,000
20,000
500
1,000
2,000
3,000
4,000
5,000
6,000
8,000
10,000
20,000
1,000
2,000
4,000
5,000
10,000
15,000
Clay weight
(grains)
0.500
0.500
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
0.500
0.500
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
i.ooo
                                  50

-------
   Equil.  Pb  =  Pb added  (Ug)  -   (Pb removed  (u/g)  x  wt. clav
                                 sample volume

     Pb sorption curves were constructed by  this method  for several pH values
from the plots given in Figures 12,  13, and  14. Representative  curves are
presented in Figure 15 along with  sorption isotherms  obtained at pH 5.0 and
25° C for Pb sorption from Pb(N03)2  solutions and  0.1 M  NaCl solutions, fol-
lowed by the DuPage and Blackwell  leachates, respectively.

     The sorption curves for the two leachates  show a sharp upswing occurring
at equilibrium concentrations of approximately  200 ppm Pb. Qualitatively
identical curves, also with a sharp  upswing  at  about  200 ppm Pb, were ob-
tained by using montmorillonite clay. The data  presented in Figure 13 may be
used to verify this finding. Sorption isotherms computed from Figures 12, 13,
and 14 at pH 3.0 or 4.0 did not exhibit the  sharp  upswing. A sharp upswing in
a sorption isotherm at higher concentrations is generally viewed as initia-
tion of precipitation of an insoluble compound. The fact that the sharp rise
in Pb removal occurred at  about 200  ppm Pb under the  COa partial pressures in
the flasks, while at pH 4.0 no sharp increase was  observed, is  consistent
with solubility computation which  assume a mechanism  of  PbCOs formation.

     To predict the maximum amounts  of Pb that  could  be  sorbed  by the two
clays from the various solutions,  the kaolinite sorption data plotted in
Figure 15 and the sorption data obtained for montmorillonite were plotted ac-
cording to the Langmuir adsorption isotherm  equation. The results are illus-
trated in Figure 16.

     The Langmuir equation was found to describe the  data obtained for
Pb(1103)2 sorption by both  kaolinite  and montmorillonite  and for Pb sorption
from 0.1 M NaCl solutions  by kaolinite over  the entire concentration range
studied. The results obtained for  the leachates, however, were  somewhat dif-
ferent. The Langmuir equation was  found to describe the  sorption up to con-
centrations between 30 and 40 ppm, at which  point  a sharp change in slope
occurred, giving two distinct linear regions to the isotherms.  The adsorption
maximums computed from the slopes  of the lines  shown  in  Figure  16 are given
in Table 8. From the values in Table 8 a quantitative estimate  of the sorp-
tion differences noted in  Figure  15  and a comparison  of  the sorption capaci-
ties of the two clays can  be made. In Pb(NOs)2  solution, montmorillonite
sorbed approximately five  times more Pb than kaolinite,  while in DuPage
leachate it sorbed less than twice as much as kaolinite. This result indi-
cates that the competitive ions contained in landfill leachate  affect the
relative sorption affinity of Pb  for the montmorillonite; i.e., leachate re-
duced Pb sorption by montmorillonite proportionately  more than  it reduced Pb
sorption by kaolinite.

DISCUSSION

     The data presented above suggest that several mechanisms are responsible
for removal of Pb from solutions  of  varying  ionic  composition and pH.

     Precipitation was found to be an important mechanism in landfill leach-
ates, as is shown by the removal  of  Pb from  the blank solutions, which  con-

                                      51

-------
 16.0
    0  25 50     100
     200           300
Equilibrium  Pb concentration (ppm)
400
500
Figure 15.  The  amount  of Pb sorbed per gram  of  kaolinite at pH 5.0  and
            25°  C  plotted as a function of the equilibrium Pb concentra-
            tion.
                                    52

-------
   0.12
   0.11-
   0.10-
  0.09-
   0.08
   0.07
                                  Koolinite  in Blackwell leachate
                             Kaolinite in
                             Du Page leachate
Kaolinite  in
O.I M NaCI
                              Montmorillonite in
                              Du  Page leachate
      0  25  50
    200          300          400

Equilibrium  Pb  concentration (ppm)
              500
Figure  16-  Pb  sorption data  for kaolinite and montmorillonite at pH  5.0
             and 25° C  plotted  according  to the Langmuir isotherm equation.
                                        53

-------
TABLE 8.  MAXIMUM REMOVAL OF Pb FROM pH 5.0 and 25  C SOLUTIONS
          COMPUTED USING THE LANGMUIR EQUATION

KAOLINITE
Pb(N03)2
0.1 M NaCl
DuPage leachate
Blackwell leachate
MONTMORILLONITE
Pb(N03)a
DuPage leachate
Micrograms/g
Region 1 Region 2

15,914
10,240
1,680 8,530
986 2,401

82,428
1,811 11,133
Meq Pb+^/lOO g clay
Region 1 Region 2

15.36
9.88
1.62 8.23
0.95 2.32

79.56
1.75 10.75
                              54

-------
tained no clay  (Figures  12,  13,  and  14). Losses  of  Pb  from the DuPage leach-
ate were observed at pH  values greater than about 6 and  in Blackwell leachate
at pH values above 5. A  white precipitate  was  observed forming in the leach-
ate solutions at pH values greater than 6.  It  was filtered out, and the
chemical compound was identified by  its X-ray  diffraction pattern as a highly
crystalline PbC03. The peaks were sharply  defined,  and no peaks other than
those attributed to PbC03 were observed. This  is offered as evidence that
PbC03 formation was the  compound responsible for Pb removal from leachate
solutions at the higher  pH values, and it  is presumed  to be the cause of the
apparent formation of a  precipitate  at concentrations  of Pb greater than 200
ppm observed in the sorption isotherms (Figure 15).

     Stumm and  Morgan  (1962) showed  that the occurrence  of metal hydroxyl
species can affect the sorption  of hydrolizable  metal  ions. They found that
the pH at which metal hydroxyl species formed  corresponded to the pH at which
metal ion sorption became significant. To  check  the role of hydrolysis of the
Pb ion on its sorption by clay at various  pH values and  in solutions where
precipitate formation did not effect Pb removal, the distribution of various
hydroxyl species in a 4  x 10"^ M Pb(N03)a  solution  in  the pH range 3 to 8
was obtained from Gadde  and  Laitinen (1973b),  who computed the species dis-
tribution by using the constants given by  Olin (1960). The distribution of
Pb hydroxyl species, along with  data obtained  for sorption by kaolinite from
a solution with an initial  concentration of 4  x 10""* M Pb as Pb(N03)2, is
illustrated in  Figure  17.

     It is evident from  the  plots in Figure 17 that species other than Pb4"4"
are relatively  insignificant (<1%)  at pH values  less than 6. Sorption at pH
values below 6  are not related to the hydroxyl species of Pb, but rather to
Pb4"*" ion. The decrease in Pb sorption at low pH  values is apparently due to
an increase in  competition  for sorption sites, with H4" and its related com-
petitive effects on Pb sorption  caused by  the  dissolution of AT*"1 ' ions from
the clay crystal lattice (Grim,  1968). At  pH values above 6, a sharp rise in
Pb sorption occurred coincident  with the formation  of  hydroxyl Pb species.
It therefore seems likely that at least a  portion of the observed increase
in Pb sorption  with decreasing pH and the  rapid  increase in the amount of Pb
sorbed coincident to the formation  of monovalent Pb-hydroxyl species are
consistent with a cation exchange mechanism for  Pb  removal from solution by
clay minerals.  Table 8 exhibits  further evidence that  cation exchange  is the
principal mechanism for  Pb  removal  by clay minerals.

     The sorption maximums  for Pb(N03)2 solutions,  computed  from  the  slope
of the Langmuir plots, for  kaolinite and montmorillonite are  15.36  and 79.56
meq Pb++/100 g  clay, respectively.  These values  can be compared to  the cation
exchange capacity  (CEC)  values of 15.1 for kaolinite and 79.5 meq/100  g for
montmorillonite that were determined by the ammonium acetate method and re-
ported in Table 1. The CEC  values are within 2%  of  the Pb  sorption  maximums
computed from the Langmuir  equation; i.e., Pb++ sorption is  merely  another
method of measuring the  cation exchange capacity of a clay.

     Further evidence  of a  cation exchange mechanism is  the reduction in Pb
sorption in solutions  containing 0.1 M NaCl and also as  the total salt con-
tent of the two leachates studied increases. The decrease  in Pb sorption is

                                      55

-------
Figure 17.  Distribution of Pb (II)  species in 4 x 10 ^ M Pb(N03)2 and
            uptake by 0.1 g kaolinite from 60 ml of solution.
                                56

-------
attributed to increasing competition for cation exchange  sites by Na+ in the
0.1 M NaCl solutions and to  an  increase  in the divalent cation competition in
the two leachates. For example,  the  Blackwell leachate contains much more Fe
in solution than the DuPage  leachate (Table 2) . Such high levels of com-
peting ions could account  for the  large  reductions  in Pb  sorption observed in
the presence of landfill leachate  compared to the sorption in pure Pb(N03)2
solutions.

     The results for Pb sorption from leachate, plotted according to the
Langmuir adsorption equation, show a distinct two-slope character. This shape
of curve has been attributed to adsorption at sites of distinctly different
energy (Griffin and Burau,  1974; Griffin and Jurinak, 1973). However, in the
present study competition  from  other cations in solution  seems to be re-
sponsible for the change in  slope, because it was observed only in the multi-
component cation systems.  Solutions  of Pb(N03)2 or  NaCl did hot exhibit the
sharp change in slope  for  Pb sorption that was observed for the leachates.

     One of the assumptions  of  the Langmuir equation is that the adsorbent
surface is homogeneous with  respect  to the energy of the  adsorption sites.
However, in a multicomponent cation  system the sites are  occupied by cations
with various retention energies relative to Pb; i.e., Pb  can displace certain
cations, such as Na+, much more easily than it can  replace cations such as
Ca"1"1". This reaction is postulated to affect the shape of  the adsorption iso-
therm by filling the lower energy sites  preferentially; i.e., Pb first ex-
changes with a cation, or  a  group of cations, of similar  exchange energy.
This phase of the sorption is attributed to the initial slope of the Lang-
muir plot. As the concentration of Pb in solution is increased, the chemical
potential gradient  is  increased until it is sufficient to initiate exchange
of the cation, or group of cations,  with the next highest energy of reten-
tion relative to Pb. This  second energy  level of exchange is postulated to
produce the sharp change in slope of the Langmuir plots in the leachate solu-
tions.

     An explanation, other than competition for cation exchange sites, for
the observed reductions in over-all  Pb sorption is  the tendency of Pb to
form metal-organic  complexes with the organic compounds present in landfill
leachate. These organic-metal complexes  can lower the activity of the Pb in
solution, thus reducing the chemical potential gradient for sorption. Gadde
and Laitinen  (1973a) showed that Pb  forms stable complexes with organic com-
pounds found in soils  and  that  these compounds were able  to solubilize Pb
present in different forms in the solid  phase.

     To determine the  role of Pb-organic complexes  in the observed Pb sorp-
tion reductions in  leachates, the complexing capacity of  Pb in the two
leachates was measured.

     In the DuPage  leachate, the extent  of complexation  (ppm complexed) was.
found to be 22 ppm  at  a Pb concentration of 200 ppm. It is clear from the
above data that only about 11%  Pb is complexed, while it  would take more than
50% complexation  to explain the reduction in Pb sorption  by an organic com-
plexing mechanism.
                                     57

-------
     The fact that formation of a Pb-organic complex cannot be used to ex-
plain more than a small fraction of the observed reduction in Pb sorption is
emphasized by the results obtained from using Blackwell leachate. In this
study it was found that addition of successive aliquots of Pb to the leachate
gave approximately the same incremental response (yA-current) in the wave for
free lead ion. Apparently up to 80 ppm total Pb, no complexation of Pb was
observed. It was noted that Fe+  or its weak complex with leachate is pre-
sumed to have produced the large increase in polarographic current observed at
a potential ~1.4 V. From the results of this study, it appears that Pb"*"1" is
either unable to compete with Fe"1"4" or other cations present in large amounts
in Blackwell leachate, or the leachate has little or no complexing capacity.
The former explanation appears to be the more plausible.

     The results of the above studies have led to the conclusion that Pb re-
moval from solution is primarily an exchange-adsorption reaction that is af-
fected by pH and ionic competition. The formation of Pb-organic complexes was
concluded to be of secondary importance in landfill leachates due to competi-
tion from high concentrations of other cations. At pH values above 6, removal
of Pb from solution by clay can be expected to increase substantially, owing
either to increased adsorption of Pb-hydroxyl complexes or to formation of
PbC03 in landfill leachates.

Disposal Site Design Application

     An example of how the data from the study can be used is its application
to the question posed at the beginning of the paper — how thick a proposed
clay liner must be to remove all Pb from landfill leachates, from industrial
waste streams of similar ionic strength to the leachates (0.1 M NaCI), or
Pb(NOs)2 solutions at various pH values and Pb concentrations. The results of
the computations are presented in Table 9.

     The table gives the thickness of a square meter of a 30% clay liner,
packed to a bulk density of 1.60 glee., that contains enough clay to remove
all the Pb from 762 liters  (201 gal) of solution. This particular volume is
the amount generated from a typical sanitary landfill containing municipal
solid waste placed 3 meters (10 feet) deep and having an annual net infil-
tration of 254 mm  (10 inches) (U.S.-EPA, 1974). The thicknesses of the clay
liner given in the table, therefore, effect total removal of Pb for a year
by a square meter of liner at the given concentrations of Pb and pH values.
They are, of course, the minimum thickness possible since they represent an
idealized situation. The actual thickness necessary in a field application
will be somewhat greater to allow for non-equilibrium conditions, physical
dispersion, diffusion, and the normal engineering safety factors.

     The information compiled in Table 9 indicates that only relatively thin
layers of clay, especially montmorillonite, are necessary for removal of Pb
unless the pH values are very acid and the Pb concentrations are high. The
high sorption capacity of clay minerals and the reversible nature of exchange
adsorption reactions have important environmental consequences. Soils and
surface waters may change in ionic composition or pH as environmental con-
ditions change. A sudden decrease in pH may release large amounts of poten-
tially toxic Pb into the aqueous phase, especially in places where PbC03 has

                                      58

-------
TABLE 9.  THICKNESS (cm) OF A SQUARE METER OF A 30% CLAY LINER
          NEEDED TO REMOVE Pb FROM 762 LITERS (201 gallons) OF
          SOLUTION PER YEAR
Pb Concentration



KAOLINITE
Pb(N03)2
0 . 1 M NaCl
DuPage
Blackwell
MONTMORILLONITE
Pb(N03)2
DuPage
10 ppm 100 ppm
at pH at pH
3 58358

<1 — 5.3 1.8 <1
<1 — — 2.3
15.9 2.1 * 28.9 6.4 *
19.8 4.0 * 49.6 11.3 *

<1 - — <1
9.9 1.8 * 13.2 3.7 *
1000 ppm
at pH
358

10.0
15.5
79.4 * *
264.6 * *

1.93 —
18.0 * *
  *Precipitation as PbC03
                                   59

-------
accumulated. Cations, especially di- and tri-valent, compete with Pb and may
exchange with it, thus allowing Pb to come into solution. These multiple
interactions must be considered when a disposal site is designed and the
environmental impact of Pb assessed.
                                     60

-------
                                  SECTION 8

                    EXCHANGE-ADSORPTION OF Cu, Zn, AM) Cd
                      FROM DEIONIZED WATER AND LEACHATE
                         SOLUTIONS BY CLAY MINERALS1
ABSTRACT
     The effect of pH on the  removal  (exchange-adsorption by kaolinite and
montmorillonite clay minerals plus precipitation) of  copper, zinc, and
cadmium in deionized water  (pH  range  4  to  6)  and municipal leachate (pH range
2 to 8) solutions was studied.  Solutions contained up to 1000 ppm Cu, Zn, or
Cd. Families of removal versus  pH curves were obtained that can be used to
construct removal isotherms for specific pH values. With certain exceptions,
very significant increases  in the amounts  removed from both solutions were
observed as the pH rose in  the  pH ranges studied. Precipitation contributes
significantly to removal of Cu  from leachate  above pH 5, Zn above 7, and Cd
above 6.

     Three different pH 5.0 removal isotherms are presented for deionized
water solutions. These isotherms were constructed from removal curves ob-
tained by different experimental methods which were shown to influence the
interpretation of the results.  The differences in the three isotherms are re-
solved by use of a Langmuir-type isotherm  equation that was derived to ex-
press the simultaneous competitive adsorption of two  cations. The equation
reveals that, under certain conditions, the amount of exchange adsorption of
a cation should be independent  of its solution concentration. This proved to
be true for the exchange adsorption of  Cu, Zn, and Cd from deionized water
solutions by the clay minerals  at pH  5.0.

     Isotherms of leachate  removal at pH 5.0  were constructed and compared
with the deionized water isotherms for  the same pH. The amount of removal at
pH 5.0 from leachate is significantly lower than it is from deionized water
solutions because of competition from the  other cations present in the leach-
ate. Competition prevents the amount  of exchange-adsorption at low concentra-
tions from becoming independent of concentration. The leachate isotherms are
specific for the leachate used  and may  not approximate the exchange-adsorp-
tion from another leachate  of different ionic strength and composition.

     The mobility of Cu, Zn,  and Cd in  soils  or clay  minerals is dependent
upon the pH and ionic strength  of the solution as well as on the CEC of the
soils or clay minerals. The CEC value is of little importance at sufficiently


1Authors: R. R. Frost and R.  A. Griffin


                                     61

-------
high  (above 7) pH values because precipitation is then more important  than
cation exchange in the removal of Cu, Zn, or Cd from solution.

INTRODUCTION

      The results of chemical analyses of 20 municipal leachates have been  re-
ported by the U. S. Environmental Protection Agency (EPA, 1974). These analy-
ses showed that leachate was similar to sewage sludge effluent with respect
to its high content of organic matter, nitrogen, phosphorus, and potassium
and also indicated levels as high as 10 ppm for Cu, 1000 ppm for Zn, and 17
ppm for Cd. The accumulation of these potentially toxic heavy metals in soils
makes long-term application of municipal leachates to the land hazardous be-
cause Cu, Zn, and Cd enter the human food chain by accumulating in plants.
However, a lack of basic data on the reactions of these metal ions with soil
colloids in the presence of a complex solution matrix, such as leachate,
hampers efforts to determine what levels of application can be tolerated
without degradation or loss of the soil resource to food crop production.

      The original purpose of this study was to measure the adsorption  of Cu,
Zn, and Cd by kaolinite and montmorillonite clay minerals from a landfill
leachate at pH 5.0. However, the study was later expanded to investigate the
adsorption of low concentrations of heavy metal ions by earth materials from
deionized water solutions and their removal (adsorption plus precipitation)
from  a leachate in the pH range 2.0 to 8.0. The results of this study  give
insights into the mobility of Cu, Zn, and Cd in soils irrigated with leach-
ates  and can be applied to the design of clay liners for municipal and in-
dustrial waste disposal sites.

THEORETICAL

      In any adsorption study, the amount adsorbed is usually measured  as a
function of adsorbate concentration in the medium surrounding the adsorbent.
It is generally desirable to be able to fit the adsorption data to an  adsorp-
tion  isotherm equation so that "parameters" associated with the adsorption
isotherm can be calculated for comparisons and correlation with other  data.

      The Langmuir (1918) equation is given as:

                                 /       KbC                                <•_•!
                               x/m  =  TTKC                              {7}

where x/m = amount of adsorbate adsorbed per gram of adsorbent, C = the equi-
librium ion concentration in solution, b = the adsorption maximum, and K = a
constant related to the bonding-energy of the adsorbate to the adsorbent.
This  equation has been used extensively in studies of adsorption of ions from
solution by soils and clay minerals. Eq. {7} can be rearranged into a  linear
form,  where:

                         C/(x/m)  =  I/(Kb)  +  C/b.                       {8}

     The application of Eq.  {8} to experimental data for Zn adsorption by
soils  (Shuman, 1975)  has produced two linear portions of the plot. Following

                                     62

-------
the lead of Syers et al.  (1973), Shuman  (1975) attributed this result to two
types of adsorption sites. However, the  adsorption model upon which Eq. {7}
is based assumes that the surface of the adsorbent contains only one type of
adsorption site. Therefore, there is no  justification for arguing the exis-
tence of two sites on an  adsorbent surface when a one-site model equation has
been applied to the experimental data. Also, inspection of Eq. {7} shows
that, at low concentrations, it reduces  to:

                                x/m  =   KbC.                              {9}

     Therefore, Eq. {8} has no significance at low concentrations, even
though C/(x/m) values can be calculated  and plotted. If a two-site adsorption
model is to be discussed, then the following equation should be applied to
the experimental data:
where « - fraction of  sites with bonding-energy coefficient KI . In practice,
Eq. {10} would be almost  impossible to apply to experimental data.

     The linear  form of the Langmuir equation  (Eq.  {8}) holds best for the
plateau region of the  adsorption isotherm  (high concentrations) . This is true
regardless of the equilibrium concentrations because the deviations from the
linear plot occur as the  equilibrium concentration  approaches zero. A close
look at Eq. {8}  confirms  this point. At, or near, the plateau region of the
adsorption isotherm,  (x/m) is about constant and, hence, the real plot is
(C/constant) versus C, which  must be a straight line. But at low concentra-
tions, where  (x/m) is  increasing as C is increasing, values of  C/(x/m) can
change at a faster rate than  they do at higher concentrations at which C is
the only variable changing to an appreciable extent. Therefore, all Langmuir
plots  (Eq. {8})  will probably show two or  more straightline segments if data
points are obtained at sufficiently low and high concentrations.

     The most important point ignored by the application of Eq. {7} to ad-
sorption of cations from  solution by soils and clay minerals is that the ad-
sorption process is primarily one of cation exchange, and for every cation
adsorbed one or  more cations  must be desorbed. The  latter' can then compete
for adsorption sites.  Boyd, Shubert , and Adamson  (1947) developed an adsorp-
tion equation for the  simultaneous competitive adsorption of two equally
charged cations, A and B. By  formal analogy with the Langmuir  (1918) equation
for adsorption from a  binary  gaseous mixture the following equation is ob-
tained:

                                    b K /K^   (C./C  )
                         f i \     , _ A   D    A  r>
                         (x/m)A =  ! +


     The linear  form of Eq. {ll} is
                                      63

-------
                           CA/CB  =  _S_     1\                      {12}
                          (x/m)A     KA b     b CB


Where the concentration of A is small compared to the concentration of B, so
that adsorption of A does not produce a detectable change in the concentra-
tion of B (e.g., adsorption on a Ca-saturated clay from solutions with high
Ca content), a plot of C /(x/m)  versus C  should be linear. The plot should
also be linear when (x/mJA approaches the cation exchange capacity (CEC), be-
cause Cg would then become a constant value and would not change as CA
changes. From Eq. {ll} it can be seen that the amount adsorbed, (X/III)A» must
depend upon the ratio of the equilibrium concentrations of the exchanging
cations and not upon the actual concentrations in solution. Where CB»CA,
(x/m)A will become a linear function of CA, and when CA»CB , (x/m)A will be-
come a constant and independent of the concentration of A.

     Equation {11} demonstrates that, under appropriate experimental condi-
tions, the amount of exchange-adsorption should be independent of the solu-
tion concentration of the cation adsorbed. If it is desirable to study the
migration of a cation through soils or clays by measuring its adsorption from
a pure solution, then a fixed weight of soil sample and a fixed solution
volume should be maintained throughout the concentration range being studied.
However, it is not sufficient to measure the exchange-adsorption from pure
cation solutions. For example, to determine how far the Cu, Zn, or Cd in 250
ml of 200 ppm deionized water solutions will migrate down a clay column the
following procedure should be followed. One gram of clay is placed in the
250 ml of solution (initial concentration, C); adsorption will occur to give
an equilibrium concentration, Cj. In the process, (200-Ci) x 0.25 mg of
cation is adsorbed and an equivalent number of moles of exchangeable cations
on the clay will be desorbed. Hence, the solution phase will now contain a
mixture of cations from which the amount of exchange-adsorption measured from
a pure solution with a concentration of Ci. This process should be carried
out stepwise, C to Ci, Ci to C2, .. . , Cn-l to Cn, until Cn -*- 0. Thus, the
amount of adsorption from solution concentration GI will be dependent upon
the initial concentration of the solution from which Ci was derived.

EXPERIMENTAL

     After pH adjustment, all clay-leachate or clay-deionized water suspen-
sions in this study were shaken in a constant temperature bath at 25 ± 0.5°
for at least 24 hours to insure complete equilibration. The equilibrium pH
values of the clay suspensions were measured, the clay suspensions centri-
fuged, and the supernatant solutions were analyzed by atomic absorption
spectroscopy for their Cu, Zn, or Cd concentrations. Blanks (i.e., no clay)
of spiked leachate or deionized water solutions that had been prepared along
with the clay suspensions also were analyzed to determine the initial Cu, Zn,
or Cd concentrations present in the solutions. The amount of Cu, Zn, or Cd
removed from solution by a given clay at a particular pH was calculated as
the (initial equilibrium concentration) x (solution volume/sample weight) /
1000.  The amount of Cu, Zn, or Cd removed from solution was plotted as a
function of pH.
                                     64

-------
     Three types of experiments were  conducted  on  the removal of Cu, Zn  and
Cd by the clay minerals from deionized water  solutions of Cu, Zn  or Cd'ni-
trates. In the first,  1.000 g  of  kaolinite  or montmorillonite and 50-ml ali-
quot s of the appropriate nitrate  solution were  placed into  125-ml Erlenmeyer
flasks. The approximate concentrations chosen were:  25, 50,  100, and 200 ppm
Cu and 10, 40, 200, and 400 ppm Zn  for both kaolinite and montmorillonite;
100, 200, and 500 ppm  Cd for kaolinite;  and 100, 500, and 1000 ppm Cd for
montmorillonite.

     Three or four replicate clay suspensions were prepared  for each concen-
tration of Cu, Zn, or  Cd.  The  pH  values  of  the  replicate clay suspensions
were adjusted to various values in  the pH range 4  to 6 with  dilute HN03 or
dilute NaOH solutions. The use of NaOH was  avoided if possible because Na
ions can compete for adsorption sites.

     In the second type of experiment, the  weight  of the clay sample and/or
the volume of solution were chosen  so that  the  total amount  of Cu, Zn, or Cd
in solution per gram of clay would  be a  constant.  The following quantities
were used: about 12.5  mg of Cu/g  of kaolinite (i.e., 250 ml  at 25 ppm Cu/
0.5g)  at 10, 25, 125,  200, and 500  ppm Cu;  about 62.5 mg of  Cu/g of mont-
morillonite (i.e., 250 ml  at 25 ppm Cu/O.lOg) at 25, 125, and 500 ppm Cu;
about  20.0 mg of Zn/g  of kaolinite  (i.e., 50  ml at 40 ppm Zn/0.10g) at 10,
40, 200, and 400 ppm Zn; about 100  mg of Zn/g of montmorillonite (i.e., 250
ml at  40 ppm Zn/O.lOg) at  10,  40, 200, and  400  ppm Zn; about 25.0 mg Cd/g of
kaolinite  (i.e., 50 ml at  50 ppm  Cd/O.lOg)  at 20,  50, 200, and 500 ppm Cd;
about  125.0 mg Cd/g of montmorillonite  (i.e., 250  ml at 50 ppm Cd/O.lOg) at
50, 100, 200, and  1000 ppm Cd. Three  or  four  replicate suspensions were pre-
pared  for each concentration of Cu, Zn,  or  Cd.  The pH values of the replicate
clay suspensions were  adjusted to various values in  the pH range 4 to 6.

     In the third  type of  experiment, stepwise  removal of Cu, Zn, or Cd from
solution was studied.  Some of  the second type of experiment  that used 250-ml
solutions were taken as the first step in the third  type of  experiment. For
the second step, the supernatant  solutions  from replicate clay suspensions in
the first step were sampled for analysis and  then  mixed together. The com-
bined  solution was used with fresh  clay  samples to prepare three replicate
clay suspensions which were adjusted  to  various pH values in the pH range 4
to 6.  The supernatant  solutions from  the second step were used in a third
step,  etc.

     Experiments on the removal of  Cu, Zn,  or Cd from leachate were carried
out by pipeting 50-ml  aliquots of leachate  into 125-ml Erlenmeyer flasks.
Either the leachate had been spiked with Cu,  Zn, or  Cd nitrates to give the
desired concentration  prior to the  previous step,  or 2.0 ml  of an appropriate
stock  solution was pipeted into the flasks  containing the 50-ml aliquots of
non-spiked leachate. Several replicates  were  prepared for each concentration
of Cu, Zn, or Cd used. The pH  values  of  the replicate spiked leachate solu-
tions  were adjusted to various values throughout the pH range 2 to 8. Appro-
priately sized samples of  either  kaolinite  or montmorillonite,were then
weighed out and transferred to the  flasks.  The  weight of clay used was chosen
so that the amount of  Cu,  Zn,  or  Cd removed from the leachate solutions could
be determined with some precision from the  difference between the initial and

                                      65

-------
final solution concentrations.

     It was discovered during preliminary experiments that, when the pH of  a
clay-leachate suspension was adjusted to a particular value, the pH of the
suspension would rise on shaking and that the removal of Cu, Zn, or Cd in-
creased with increasing pH. Thus, addition of the clay to the spiked leachate
solutions after pH adjustments have been made avoids such potential problems
as the possible irreversible removal of Cu, Zn, or Cd. The irreversible re-
moval could occur if the pH of the prepared clay-spiked leachate suspension
was high and the pH of the suspension was then adjusted to a much lower
value. The experimental procedure used should produce true equilibrium re-
moval data.

     Several individual experiments also were carried out in which clay
samples were placed in a mixture of 50-ml aliquots of pH 5 leachate or de-
ionized water and 2-ml aliquots of different stock solutions. The pH of the
resulting suspensions was repeatedly adjusted to 5.0 until equilibrium was
obtained.

RESULTS AND DISCUSSION

     The amount of Cu, Zn, or Cd removed from deionized water solutions was
plotted versus pH. Except in certain cases, very significant increases in the
amounts removed were observed as the pH rose in the range 4 to 6. The initial
concentration of Cu, Zn, or Cd in solution, the weight of clay used, the
final solution volume after pH adjustments, and the removal versus pH curves
themselves can be used to construct "adsorption" or "removal" (adsorption
plus precipitation) isotherms at different pH values by use of the following
equation:

     •Equilibrium concentration (ppm)  =  initial concentration (ppm)
                Amount removed (mg/g) x wt. clay (g) x 1000.             r.  •,
                           Final solution volume (ml)

Isotherms were calculated from the data recorded for the deionized water
solutions at pH 5.0, and the results are shown as isotherm types I, II, and
III in Figures 18 and 19.

     The type I isotherms (Figs. 18 and 19) are those obtained when 1.00 g  of
kaolinite or montmorillonite was placed in about 50 ml of Cu, Zn, or Cd solu-
tions. The total amounts of Cu, Zn, or Cd present in 50 ml of solution at
different concentrations are given in Table 10. For kaolinite (CEC 15.1 meq/
100 g), 75.5 ymoles of a divalent cation/g would be required for complete ex-
change; for montmorillonite (CEC 79-5 meq/100 g), 397.5 ymoles/g would be re-
quired for complete exchange.  Thus, as indicated in Table 10, at low concen-
trations of Cu, Zn, or Cd, insufficient cations are present in 50 ml of solu-
tion for complete exchange of 1.00 g of clay (especially montmorillonite).
The amount of Cu, Zn, or Cd that can be adsorbed is therefore necessarily
limited by the number of Cu, Zn, or Cd ions initially present in solution.
The number of Cu, Zn, or Cd ions actually adsorbed will depend on how well
the desorbing Ca ions compete with the Cu, Zn, or Cd ions remaining in solu-
tion.  The type I isotherms in Figures 18 and 19 specifically represent the

                                     66

-------
                               KAOLINITE
                    200    300    400    500    600     700
                         Equilibrium concentration  (ppm)
 r
800
900
Figure 18.  The amount of Cu, Zn,  or  Cd  removed  from  solution per gram of
            kaolinite at pH 5.0  and 25°  C, plotted as a function of the
            equilibrium concentration.   Curve  I  - fixed weight clay/fixed
            solution volume; Curve II -  total  amount  of clay exchangeable
            ions/total amount of Cu,  Zn, or Cd ions in solution held about
            constant; Curve III  -  stepwise isotherms; Curve IV - DuPage
            leachate isotherms.  Open symbol data points were obtained
            from clay suspensions  adjusted several times to pH 5.0 instead
            of being calculated  from  removal curves at pH 5.0.
                                    67

-------
                               MONTMORILLONITE
     32-



     28-



     24-



     20-



   ~ 16-








    = 8-
    o
    in
      4-
    E  -
    o 18-
   -o 16-
    0)
      8H
   < 0-
                                                         Zn
              100
200    300    400    500    600    700


     Equilibrium concentration  (ppm)
                                                               800
Figure 19.  The amount  of  Cu,  Zn,  or Cd removed from solution  per gram of

            montmorillonite  at pH  5.0 and 25° C, plotted as  a  function of

            the equilibrium  concentration.  The labels I - IV  and the open

            symbols have same  meaning as in Fig. 18.
                                    68

-------
  TABLE 10. TOTAL CONTENT OF Cu, Zn, OR Cd IN 50 ml OF SOLUTION
Initial concentration
  in solution                   Cu              Zn              Cd
ppm
                                 ymoles in 50 ml solution volumes
4
10
25
40
50
100
200
400
500
1000
3.15
7.87
19.7
31.5
39.4
78.7
157.4
314.8
393.4
786.9
3.06
7.15
17.9
30.6
38.2
76.5
153.0
306.0
382.5
765.0
1.78
4.45
11.1
17.8
22.2
44.5
89-0
178.0
222.5
445.0
                                  69

-------
amount of Cu, Zn, or Cd adsorption by 1.00 g of clay from about 50 ml  of
solution.

     The type II isotherms in Figures 18 and 19 were obtained when the weight
of clay samples and/or solution volumes were adjusted so that the ratios
(total number of Cu, Zn, or Cd ions)/(number of exchangeable cations on the
clay samples) were held about constant (at least 2/1) and independent  of  the
concentrations in the initial solution. Although, at equilibrium, the  equi-
librium concentration of Cu, Zn, or Cd (CA) could not be considered to be
much greater than the concentration of desorbing Ca ions (Cg) , the ratio  of
CA/CB should be a constant and, hence, Eq. {11} says that (x/m)A should be
constant and independent of CA- This is shown by the type II isotherms in
Figures 18 and 19. The "scatter" in the experimental data points for the  type
II isotherms is attributed to experimental errors due primarily to problems
in measuring accurately the small changes in concentration that are due to
adsorption. These errors must then be multiplied by large factors to compute
the amount of metal removed per gram of clay.

     The type I and II isotherms plotted according to the linear form
(Eq. {8}) of the Langmuir equation (Eq. {?}) are shown in Figures 20 and  21.
The Langmuir plots for the type I isotherms (1.00 g clay/50-ml solution)  show
a definite two-slope character (except the plot for Zn-kaolinite) ; in.fact,
three linear segments of the Langmuir plot appear to exist for the Cu-mont-
morillonite plot. The Langmuir plots (Fig. 21) of the type II isotherms for
montmorillonite (Fig. 19; the total amount of Cu, Zn, or Cd initially  present
in solution /g clay was held about constant) show only a one-slope character
throughout the concentration range investigated. The C/(x/m) values defi-
nitely approach zero as the equilibrium concentration approaches zero. This
is precisely what should occur if the competitive Langmuir equation
(Eq. {ll}) is valid, because Eq. {11} says that, if the equilibrium ratio
CA/CB is constant, (X/IH)A is independent of the actual concentration,  and,
therefore, as CA approaches zero, CA/(X/III)A must approach zero.

     The Langmuir plots (Fig. 20) for the Cu and Cd type II isotherms  with
kaolinite (Fig. 18) do show a two-slope character. It was observed during the
Cu adsorption experiments that the pH of the Cu-kaolinite suspensions  in  de-
ionized water decreased when the flask was shaken, thereby indicating  that
hydrolysis of the Cu"1"2 ion was occurring in solution (i.e. , Cu+2 + H20 =
CuOH+ + H+, etc.). Data on hydrolysis of cations that were assembled by Mes-
mer and Baes (1974) show that no appreciable concentrations of CuOH4" or
Cua(OH)2+2 will exist in solutions of low Cu concentration below pH 7, but,
in the pH range 5 to 6, precipitation can occur from solutions of about 400
ppm Cu. Thus, precipitation is a reasonable explanation for the occurrence of
the second slope in the Langmuir plot for Cu-kaolinite. Although similar  ar-
guments could be advanced for the Cd-kaolinite Langmuir plot, it appears
from Mesmer and Baes (1974) that, in the range of Cd-concentrations used,
hydrolysis and precipitation cannot be considered to be contributing factors
to the "adsorption" of Cd by kaolinite around pH 5.0.

     Shuman (1975) obtained two-slope Langmuir plots when he plotted Zn ad-
sorption by Georgia soils data according to Eq. {8}. John (1972) found no
significant correlation between the CEC values and Cd Langmuir adsorption

                                     70

-------
                         KAOLINITE
      0   100200300   0  100200300  0   100200300400
                Equilibrium   concentration  (ppm)
Figure 20.  Cu, Zn,  and Cd removal data for kaolinite in deionized water
           solutions at pH 5-0 and 25° C, plotted according to  the
           Langmuir equation (Eq. 8).  The roman numerals beside the
           plots  indicate the corresponding isotherm in Fig.  18.
                                71

-------
                            Montmorillonite
        100 200 300  0  100 200 300  0  100 200 300 400  500 600 700
                   Equilibrium concentration (ppm)
Figure  21.  Cu, Zn,  and Cd removal data  for montmorillonite in deionized
           water solutions at pH 5.0 and 25° C, plotted  according to the
           Langmuir equation. Numerals  indicate corresponding isotherms
           in Fig.  19.
                                 72

-------
maximums for 30 different  soil samples.  Both Shuman (1974)  and John  (1972)
used a fixed weight of  soil and a fixed  volume of solution  in their  adsorp-
tion measurements. The  results of this study indicate that  both Shuman  (1974)
and John (1972) would have obtained somewhat different results and conclu-
sions if they had  equilibrated their soil samples with sufficient solution
volumes of low-Zn  or Cd concentration so that the total Zn  or Cd content of
the solutions would exceed the CEC values of their respective soil samples.
The appropriate solution volumes will necessarily depend on the CEC  values of
the soil samples  (the larger the CEC, the larger the solution volume re-
quired) and what kind of exchangeable cations are present on the soil sam-
ples.

     The type III  isotherms in Figures 18 and 19 are "stepwise" isotherms.
These isotherms show the competitive effect of the desorbing exchangeable
cations initially  present  on the clay minerals on the removal of other heavy
metal cations from solution.  Thus, the type I and II isotherms in Figures 18
and 19 do not provide sufficient information to predict the migration of Cu,
Zn, or Cd in pure  solutions through soils or clays.

     The adsorption data plotted by using the linear form (Eq. {12}) of the
competitive Langmuir equation (Eq. {11}) are shown in Figures 22 and 23. It
was assumed that the number of ions (presumed to be all Ca, which is more
reasonable for kaolinite than for montmorillonite) desorbed from the clays
into solution equals the amount of Cu, Zn, or Cd adsorbed from solution. It
can be shown that  CA  (ppm)/ACA (ppm) = CA(moles/1) /Cca(moles/1) , where CA =
equilibrium concentration  of Cu, Zn, or  Cd and ACA = ^initial - CA,  so that
(CA/ACA)/(x/m)A versus  (CA/ACA) was plotted in Figures 22 and 23. For step-
wise isotherms, the ACA used included the ACA values from previous steps. Al-
though there is appreciable scatter to the data, reasonably straight lines
(except for Cd-kaolinite) , can be drawn  through most of the data points,
which include all  data  plotted for deionized-water solution isotherms I, II,
and III in Figures 18 and  19.  Some data  points lie above the line drawn
through the data points. At these data points, C/AC is generally small and
the number of ions adsorbed (x/m) is limited by the total number of  ions
initially present  in solution. Negative  deviations from the main linear
region can occur when Eq.  {8} is used at low concentrations. Positive devi-
ations from the main linear region can occur when Eq.  {12}  is used for small
values of C/AC.

     The adsorption maximums calculated  from the type II Langmuir plots in
Figures 20 and 21  and the  competitive-Langmuir plots in Figures 22 and 23 are
given in Table 11.

     The question  arises as to why there is such poor correlation between the
adsorption maximums calculated from the  CEC values and the  experimental ad-
sorption maximums  if cation exchange is  the primary adsorption mechanism.
Bittel and Miller  (1974) have determined the selectivity coefficients for
Cd/Ca on kaolinite and  montmorillonite to be about one (i.e., Ca and Cd ions
compete on about an equal  basis for adsorption sites) , but  they do not
specify a pH for their  experimental measurements. The data  obtained  in this
study indicate that the selectivity coefficients for Cu, Zn, and Cd  on Ca-
saturated kaolinite and montmorillonite  are less than one at pH 5 and ap-

                                      73

-------
    1.0-

    0.5-

    0.0
•g  2.0 H
 x
 ~   1.5-

 3   1-0-
 G 0.5-
 ^
 o o.o
     1.5-
     1.0-
    0.5-
    0.0
                 Cd
Zn
                 Cu
                                KAOLINITE
                     o"
                         -*r~**
       0.0
1.0
2.0      3.0       4.0
           C/AC
5.0
6.0
  Figure 22.   Cu, Zn, and Cd removal data for kaolinite in deionized water
              solutions at pH 5.0  and 25° C, plotted according to the com-
              petitive Langmuir equation (Eq. 12).  The data point symbols
              correspond to those  used in Fig. 18.
                                   74

-------
                     MONTMORILLONITE
       o.o
Figure 23.  Cu, Zn, and Cd removal data for montmorillonite in deionized
           water solutions at pH 5.0 and 25° C, plotted according to the
           competitive Langmuir equation (Eq. 12).  The data point symbols
           correspond to those used in Fig. 19-
                                 75

-------
  TABLE 11.   COMPARISON OF LANGMUIR ADSORPTION MAXIMUMS IN DEIONIZED
             WATER WITH CEC VALUES
                                  Cu             Zn             Cd
Source                         K*    Mt       KM       KM

                                               mg/g
II - Langmuir plots           3.08  18.5     2.86   19.2    5.0    31.2
       (Figs. 21 and 22)

     Langmuir plots           3.33  21.0     3.00   23.6     -     50.0
       (Figs. 23 and 24)

     Calc.  from CEC           4.80  25.3     4.94   26.0    8.45   44.8
       values - Table 1
  *Kaolinite
  tMontmorillonite
                                  76

-------
proach one at hzgher pH. Therefore,  the experimental adsorption maximums for
Cu, Zn, or Cd would not approach the CEC values until the equilibrium ratio
C/CCa was larger than was  the  case in any of our experiments.  Because the ad-
sorptxon of Cu, Zn, or Cd  is pH-dependent , the adsorption maximums will  of
course, show definite pH-dependence.  Prior to the onset  of precipitation
which is dependent on both concentration and PH, decreasing competition from
IT" and increasing concentration of hydrolyzed ions (e.g., Cu2(OHO+2  ZnOH+
and CdOlT) are possible reasons for the "adsorption" to  increase as pH in- '
creases.

     Curves for pH versus  the  removal of Cu, Zn, or Cd from leachate by kaol-
inite and montmorillonite  are  shown in Figures 24 and 25, respectively. The
data listed beside each curve  and in the figure captions were  used in Eq. {13}
to calculate the type IV isotherms at pH 5.0, which are  shown  in Figures 18
and 19. This was done so that  direct comparisons could be made with the de-
ionized-water isotherms. Figures 18 and 19 show that the removal from leach-
ate is appreciably lower than  the removal from deionized-water solutions.
Boyd, Shubert, and Adamson (1947) gave a general expression for the exchange
adsorption of a cation, A, from a mixture of equally charged cations in solu-
tion, which is as follows:

                                         b K. C.
                             \               A  A

     Equation  {14}  says  that  the exchange adsorption of  any one cationic
species at constant concentration will decrease as  the concentrations of
other cationic species increase.  Thus, from solutions of constant ionic
strength, in which  the total  ionic content is  large compared to the total
number of exchangeable cation on any soil or clay mineral sample placed in
that solution,  the  amount  of  Cu,  Zn, or Cd removal  cannot approach the CEC of
the soil or clay mineral at low concentrations, as  is true in the case of ex-
change-adsorption from deionized water (i.e.,  increasing solution volume will
usually not affect  the equilibrium concentration ratio of the cation of
interest to the other cations in solution to any appreciable extent). Only
when the concentrations  of Cu,  Zn, or Cd in solution exceed the combined con-
centrations of all  other cationic species in solution will the amount of ex-
change-adsorption of Cu, Zn,  or Cd approach the CEC of the soil or clay min-
eral sample being tested.  In  practice, except  at very low pH, precipitation
of Cu, Zn, Cd  and other  heavy metal ions as hydroxides and/or carbonates will
occur at moderate concentrations so that precipitation rather than cation ex-
change adsorption can become  the principal mechanism for the removal of heavy
metal ions from solution.

     The sharply rising  portions of the removal curves shown in Figures 24
and 25 can easily be interpreted, with the aid of information assembled by
Mesmer and Baes (1974),  as being caused primarily by precipitation of Cu, Zn,
or Cd carbonates, hydroxides, or hydroxide-carbonates. Removal of Zn and Cd
from the leachate by both  kaolinite and montmorillonite  is .greatly reduced
compared to their removal  from deionized water solutions at pH 5.0. The
amount of Zn and Cd removed is  reduced proportionally about the same for both

                                      77

-------
                            KAOLINITE
 25.0-

 21.0-

 17.0-

 13.0-
t
n
» 9.0-
»

; 5.0-
•
•
i  i.o^
•

' 0.9-
  2  0.8 H
  o>  0.7 H
  o
  E  0.6 H
  0)
  v_
  *-  0.5 H
  c

     0.4 H
0.3-

0.2-

 0.1-

0.0
           Cu
              1000
      192
     52.5ml
                                   887
                               /
                                     52.5ml
                            I   '  I  '  I  '   I
                            2468
                                  PH
                                          Cd
                                                       8
Figure 24.  The amount of Cu,  Zn,  or Cd removed from DuPage leachate
            solutions by kaolinite at 25° C,  plotted as  a function of
            pH.  The plots are labeled with the initial  solution con-
            centration (ppm) of Cu,  Zn,  or Cd from which each plot was
            obtained.  Unless  otherwise indicated, 2.0 g of clay and a
            total leachate solution volume of 50.5 ml were used to
            obtain each data point.
                                  78

-------
                          MONTMORILLONITE
Figure 25.   The  amount of Cu, Zn, or Cd removed from DuPage leachate
            solutions by montmorillonite at 25° C, plotted as a function
            of pH.  The plots are labeled with the initial solution con-
            centration (ppm) of Cu,  Zn,  or Cd from which each plot was
            obtained.  Unless otherwise indicated, 0.5 g of clay and a
            total leachate solution  volume of 50.5 ml were used to obtain
            each data point.
                                  79

-------
kaolinite and montmorillonite, but the reduction in the amount of Cu removed
from the leachate is appreciably greater for montmorillonite than for  kaolin-
ite. This can readily be explained on the basis that the Cu-leachate iso-
therms represents a combination of exchange-adsorption and precipitation.  The
amount of Cu removed by precipitation, for example 1 mg/g, will be about the
same for leachate solutions at pH 5, whether kaolinite or montmorillonite  is
present in solution. However, the 1 mg/g represents about 20% of the CEC for
kaolinite, while it represents only about 4% of the CEC for montmorillonite.
The decrease in removal from leachate, therefore, appears much greater for
montmorillonite than for kaolinite. The actual decrease in Cu exchange-ad-
sorption is probably proportionally the same for both clays. The removal of
Cu from leachate reaches a maximum about pH 7 (Figs. 24 and 25) and then de-
creases for pH values above 7. This behavior is due to the amphoteric  charac-
ter of Cu(OH)2 precipitates, which redissolve in basic solutions by forming
Cu(OH)^, etc. ions.

     The amount of Cu, Zn, or Cd removed from leachate has no significant  ap-
parent dependence on leachate volume, as can be seen from the leachate  iso-
therms (Figs. 18 and 19), on which data points obtained for different  clay
sample weights but constant leachate volumes have been plotted on one  iso-
therm curve. However, the amount removed could become dependent on the  leach-
ate volume if the clay sample is very large.

     Equation {14} implies that at some given ionic strength, where
                                 all.  j Cj

 is  about  constant and large compared to one, (x/m)^ = Constant x C^. The iso-
 therms  for Cd  (Figs. 18 and 19) are linear to quite high concentrations.

     The  isotherms for Cu, Zn, or Cd removal from leachate were not plotted
 according to the linear form of a Langmuir adsorption isotherm equation, al-
 though, most assuredly, straight line plots would have been obtained and
 "adsorption" maximums could be calculated. The reasons for not plotting them
 were, first, none of the leachate isotherm plots have really reached the
 plateau region in the concentration range studied (up to 1000 ppm) . Second,
 any adsorption maximums calculated from the Langmuir plots would be somewhat
 meaningless because the amount of exchange-adsorption from leachate is
 limited by competition, owing to the high ionic strength of the leachate and
 not because all the adsorption sites have been occupied by Cu, Zn, or Cd ions,
 as is implied with a Langmuir adsorption maximum. Third, if an adsorption
 maximum is calculated from the rising part of an isotherm, it represents the
 amount adsorbed at some higher hypothetical concentration that may or may not
 lie on the real adsorption isotherm. Finally, our purpose was to determine
 the maximum amount of Cu, Zn, or Cd that can be removed by kaolinite or mont-
morillonite from leachate at any particular concentration up to approximately
 1000 ppm.  As we have seen for removal from deionized water, with appropriate
experimental conditions, the amount removed is independent of concentration
and is proportional to the CEC values of the clays. It would presumably ap-
proach the CEC at higher (volume)/(weight of clay) ratios. But, with leach-


                                     80

-------
ate, the ionic strength  is not  a variable to be adjusted and  the  conditions
cannot be created where  at low  concentrations of Cu,  Zn, or Cd  the ratio of
the (equilibrium concentration  of Cu,  Zn, or Cd)/(concentration of other
cations) is large. The maximum  amount  removed must  then be a  function of the
ionic strength of the leachate,  the CEC of, the clay sample, and pH of the
leachate. Therefore, the maximum amount removed at  any concentration and pH
from leachate  (or for that matter any  solution) is  simply the value read from
the removal isotherm itself  at  the concentration of interest.

SUMMARY AND CONCLUSIONS

     Under appropriate experimental conditions, the amount of Cu, Zn, or Cd
exchange-adsorbed from deionized water by purified  kaolinite  and montmoril-
lonite clay minerals is  independent of the equilibrium concentration of Cu,
Zn, or Cd. However, the  maximum amount of Cu, Zn, or Cd adsorbed  in our ex-
periments was  related to, but not equal to, the CEC values of the clay min-
erals, probably because  the  desorbing  Ca ions effectively competed with the
Cu, Zn, or Cd  ions present  in solution. If cation-exchange adsorption experi-
ments are carried out at constant (solution volume)/(sample weight) ratios
for Cu, Zn, or Cd in deionized  water,  the amount of adsorption  is necessarily
limited by the total amount  of  Cu, Zn, or Cd that was initially present in
solution at low concentrations.  Therefore, the isotherm obtained  is really a
plot of the amount of Cu, Zn, or Cd removed by a fixed weight of  sample from
a  fixed volume of solution versus concentration. At low concentrations a dif-
ferent isotherm can be obtained simply by using a different constant (solu-
tion volume)/(sample weight) ratio in  the experiments.  To properly simulate
field conditions, soil samples  must be equilibrated with sufficient solution
volumes so that the total metal ion content of the  solutions  exceeds the ad-
sorption capacity  (CEC values)  of the  respective soil samples.  The appropri-
ate solution volumes will necessarily  depend on the CEC of the  soils and the
kind of exchangeable cations present on the soil samples.

     If it is  desirable  to  study the migration of Cu, Zn, or  Cd from de-
ionized water  through soils  or  clays by .means of batch experiments, the ex-
periments can  be carried out stepwise  (i.e., repeated treatments  of a solu-
tion with new  soil or clay  samples) for each initial solution concentration
tested. That is necessary because of the increasing concentration of ex-
changeable cations in solution  as exchange-adsorption of an individual cation
from solution  occurs stepwise towards  zero. This is not a particular problem
in high ionic  strength solutions such  as landfill leachate as discussed be-
low.

     The three different types  of isotherms obtained for the  same range of
Cu, Zn, or Cd  concentration  are easily interpreted  in terms of  a  cation ex-
change-adsorption mechanism. Also, the correct "Langmuir" isotherm equation
to apply to exchange adsorption data is Eq. {11}, which covers  the simultane-
ous competitive adsorption  of two cations, and not Eq. {7>, which  is for a
single cation. Although  strict  application of Eq. {11} requires the exchange-
able cations on the adsorbent to be homoionic, the  principal  conclusion that
"under appropriate conditions"  the amount of exchange-adsorption  is indepen-
dent of concentration will  apply, regardless of the number of exchangeable
cations on the adsorbent.

                                     81

-------
     The adsorption of Cu, Zn, or Cd from leachate by kaolinite and mont-
morillonite clay minerals is presumed to occur via a cation exchange mechan-
ism, but, because the high ionic strength of the leachate is relatively un-
variable in the adsorption experiments, the amount adsorbed does not become
independent of the concentration of Cu, Zn, or Cd in the leachate. Thus,
there is no way to predict the maximum amount of Cu, Zn, or Cd that will be
adsorbed from a given leachate by a given soil or clay mineral, at a given pH
and at some concentration without actually measuring it. At pH 5.0, precipi-
tation of Cu as a hydroxide-carbonate makes a very significant contribution
to the total amount of Cu removed by kaolinite but, because mohtmorillonite
has a higher amount of exchange-adsorption, the contribution of precipitation
to the total amount of Cu removed by montmorillonite is less significant. If
the pH of the leachate is lower than about 6.5, precipitation does not make a
significant contribution to the total amount of Zn and Cd removed by kaolin-
ite and montmorillonite.

     The mobility of Cu, Zn, and Cd in soils or clay minerals is similar to
Pb and is dependent upon solution pH and ionic strength as well as on the CEC
of the soils or clay minerals. This is of little importance at sufficiently
high (above 7) pH values, because precipitation is then more important than
cation exchange in the removal of Cu, Zn, or Cd from solution. However, the
mobility of Cu would apparently reach a minimum at pH 7 and would subse-
quently increase at pH values above 7. At pH 5, Cu, Zn, and Cd in leachate
will be quite mobile in soils or clay minerals with low CEC values, especial-
ly when the ionic strength of the leachate is high. Thus, if adsorption data
at pH 5.0 is obtained either from a leachate very high in ionic strength or
from a low CEC soil or clay mineral, and the mobility of Cu, Zn, or Cd is
calculated, it can be stated that for higher pH values, for higher CEC val-
ues, and/or for leachates low in ionic strength, the mobility would be less
than it is at pH 5.0. These facts add a built-in safety factor to estimations
of adsorption and/or mobility.

     Treatment of a soil with a waste stream or leachate will alter the ex-
changeable cation distribution of the soil; for example, a high Ca soil will
become a high Na soil if treated with a high Na-content waste stream or
leachate. Thus, other potential problems may be created that must be evalu-
ated in addition to the problem of heavy metal toxicity to plants or heavy
metal accumulation in the food chain when waste streams or leachates are dis-
posed of on agricultural land.
                                     82

-------
                                   SECTION 9

                  EFFECT OF  pH ON CHROMIUM ADSORPTION  FROM
                     LANDFILL  LEACHATE BY CLAY MINERALS1
ABSTRACT
     The adsorption of  Cr(VI)  and  Cr(HI)  species  by kaolinite  and montmoril-
lonite clay minerals was  found to  be highly dependent upon  the  pH of the clay
suspensions and the physical-chemical properties of the  clay minerals. Solu-
tion ionic strength was found  to be  of secondary importance to  estimations of
Cr(VI) and Cr(III) adsorption.

     No precipitation of  Cr(VI)  was  detected in the pH range 1.0 to 9.0. Pre-
cipitation of Cr(III) as  an  amorphous hydrated hydroxide starts to occur
above pH 4.5.

     The adsorption of  Cr(VI)  from a given solution decreased as pH in-
creased. The Cr(VI) species  distribution indicated that  the HCrOiT ion was the
Cr(VI) species predominantly adsorbed.  Montmorillonite adsorbed about four
times more Cr(VI) than  kaolinite under similar conditions of pH and ionic
competition* Contrary to  expectations,  less Cr(VI)  was adsorbed from pure
       solutions than from leachate  solutions.
     The adsorption of  Cr(III)  increased  as  the  pH of the suspensions in-
creased. At pH 2.5, the amounts of  Cr(III) adsorbed were consistent with a
cation exchange mechanism involving Cjr3+  ions. As  the pH is raised to 4.0,
the amounts adsorbed correspond to  cation exchange adsorption of the hydro-
lized Cr(III) species,  Cr(OH)t  and  Cr6(OH)it. The  adsorption of Cr(III) is 3%
to 14% lower in leachate than in pure  Cr(N03>3 solutions.

     For a given type of clay,  about 30 to 300 times more Cr(III) than
Cr(VI) is adsorbed depending upon pH and  the ionic competition in solution.
The results of the study suggest that  landfill disposal of Cr(VI) wastes re-
presents a potential pollution  hazard  due to its high mobility in earth ma-
terials and that safe disposal  may  require conversion of Cr(VI) wastes to
Cr(III) before disposal in landfills.

INTRODUCTION   •>

     Efforts to effectively dispose of most  industrial heavy metal wastes
without polluting the environment have thus  far  proved unfruitful. Tradition-
ally, rivers or lakes have been used for  the disposal of these potentially

Authors: R. A. Griffin,  Anna K.  Au, and  R.  R. Frost

                                     83

-------
hazardous discharges. Industrial plants have also disposed of their wastes
through recharge basins or diffusion wells (Welsch, 1955), and into sewer
systems (Nassau County, N. Y., Department of Public Works Sewer Regulations,
1955). All of these disposal methods can contribute to contamination of sur-
face and ground water (Davids and Lieber, 1951). To minimize the problems
caused by heavy metal wastes in sewage treatment, laws have been enacted in
some northern Illinois counties forbidding the disposal of such wastes into
sanitary sewers. That prohibition has increased the pressure for permission
to dispose of these wastes in the available sanitary landfill sites. However,
unless specially designed, sanitary landfills also are a potential source of
surface and ground-water pollution (Walker, 1969). For that reason a demand
has arisen for information about the capacity of earth materials to adsorb
heavy metals from landfill leachates (Fuller, 1975).

     Chromium compounds are widely used in the leather, textile, chemical
manufacturing, metal finishing, and other industries.  Approximately 30,000
tons of chromium-bearing wastes are discharged annually from the metal
finishing industries alone (U.S.-EPA, 1973), and problems of environmental
pollution have arisen. Chromium (VI) contamination of the village wells in
Douglas, Michigan, was reported in 1947 (Davids and Lieber, 1951), and as
early as 1952 chromium was found in high concentration in the ground water of
Nassau County, New York (Welsch, 1955).

     In trace amounts, chromium is an essential element in the diet of some
animals and, presumably, human beings.  However, at sufficiently high concen-
trations, all compounds of chromium are toxic (Smith,  1972). The valence
state of chromium has a considerable influence on its toxicity. It is well
established that Cr(VI) compounds are the most toxic and are usually irri-
tating and poisonous to all tissues (Baetjer, 1956).

     Thus far, the distribution and impact of chromium on aquatic ecological
systems have not received extensive study, so that relatively little is known
about the transfer of the metal from waste streams to earth materials and
then to living systems. Because knowledge of the chromium-leachate system is
scant, the present study was conducted to investigate the effect pH has on
Cr(VI) and Cr(III) adsorption by clay minerals plus precipitation in deion-
ized water and municipal leachate solutions. It was also desired to gain in-
sight into the factors that affect the mobility of chromium as it passes
through soils or clay-mineral layers. These soil or clay layers may be po-
tentially useful as liners for waste disposal sites.

EXPERIMENTAL

     Various concentrations of Cr(VI) in deionized water and DuPage leachate
were prepared using potassium chromate (KaCrOO , and 50-ml aliquots were
pipetted into Erlenmeyer flasks containing either 3 g of montmorillonite or
5 g of kaolinite. The weight of clay used was chosen so that the amount of Cr
removed from the solutions could be determined with some precision from the
difference between the initial and final solution concentrations. Several
replicate suspensions for each concentration were prepared, and their pH
values were adjusted with either HN03 or NaOH to various values in the pH
range 1.0 to 9.0.

                                     84

-------
     In preliminary  experiments, the amount of Cr(VI) adsorption was  found  to
decrease as pH increased,  and pH values of the Cr(VI)-clay suspensions  rose
when the flasks were shaken.  Also, apparent irreversible Cr(VI)  adsorption
occurred if the Cr(VI)-clay suspensions had been equilibrated at one  pH value
and again equilibrated  at  a higher pH value. Therefore, the flasks were
shaken for about  2 weeks  in a constant temperature bath at 25 ±  0.5°  C; the
pH values of the  clay suspensions were measured each day, and acid or base
was added when necessary  to maintain the initial pH value. At least one day
after the final pH adjustment, the pH values were recorded and the suspen-
sions were centrifuged. The supernatant solutions were then decanted  into
plastic bottles and  their pH adjusted to 2.0 to prevent any Cr adsorption by
the container. The equilibrium Cr concentrations C£q (ppm) in the supernatant
solutions were determined by atomic absorption spectroscopy. The initial Cr
concentration  (Cj) in ppm was determined by analyses of blank sample, i.e.,
sample without clay, prepared at the same time as the clay suspensions. In
our experiments,  Cj  concentrations ranging from 5 to 300 ppm Cr  as Cr(VI)
were used.

     The amount of Cr(VI)  adsorbed (x/m) in ing/g clay at a given pH was cal-
culated as  (Cj -  CEq) x VE/1000 where VE = final solution volume (ml)/weight
of clay sample  (g).  The amount of Cr(VI) adsorbed by a fixed amount of  clay
from a given Cj solution  at various pH values was plotted against the pH
values to obtain  an  adsorption-pH curve.

     The experimental procedures for the Cr(III) adsorption studies were
similar to those  used in  the Cr(VI) experiments. It was determined that 0.100
g of the clay minerals  would give the desired precision in determining  the
change in Cr(III) concentrations at equilibrium. Chromium (III)  nitrate
 (Cr(N03)3 • 9H20) was used as a source of Cr(III).  The study of  Cr(III) ad-
sorption was generally  limited to the pH range of 1.5 to 4.5 because  of
Cr(III) precipitation around pH 5. Because of the precipitation, the  pH of
the leachate was  adjusted to about 4 prior to "spiking" with Cr(III). An
initial Cr(III) concentration range of 30 to 800 ppm was chosen  for the ex-
periments.

RESULTS AND DISCUSSION

     Adsorption isotherms can be calculated from a family of adsorption-pH
curves at different  pH  values. The amount of Cr adsorbed (x/m) in mg/g  clay
at a particular pH value  is read from an adsorption-pH curve for a particu-
lar Cx and  the equilibrium concentration, CEq, in ppm is calculated from the
following equation:

                                  r,  e   ^     (x/m) • 1000
                   CEq  (ppm)  =  CX (ppm)  -  	^


where all the parameters  are as previously defined.

     Interpretation  of  the adsorption data was aided by application of  the
Langmuir equation (1918).  In the derivation of the Langmuir equation  it is
assumed that  (a)  the surface is energetically homogeneous   (b) the  adsorbate
adsorbate interaction on the surface is negligible, and (c) the  adsorbed

                                      85

-------
molecules do not  influence neighboring sites.  The Langmuir adsorption equa-
tion  in  its linear  form is:

                                        1   •  C                           {16}
                                   _      i   ,^,
                              x/m     kb     b

where C is the equilibrium concentration of the ion, x is the amount ad-
sorbed, m is the mass of adsorbent, b is the adsorption maximum, and k  is a
constant that relates to the energy of adsorption.

Chromium (VI) Adsorption

     No precipitation of Cr(VI) was observed in the pH range 1.0 to 9.0.
Families of Cr(VI) adsorption-pH curves are presented in Figures 26 and 27.
Several characteristics of Cr(VI) adsorption by the clays appear. First, the
adsorption of Cr(VI) decreases as pH increases. Second, Cr(VI) is not ad-
sorbed by the clays near pH 8.5 and above. Third, the amounts of Cr(VI) ad-
sorbed are small compared to the amounts of exchangeable cations on the clay
samples.

     Diagrams showing the distribution of Cr(VI) species covering the experi-
mental concentration range are presented in Figure 28 and were calculated by
using the constants given in Butler (1964).  Noteworthy is the rapid decrease
in the fractions of HCrOiT and CraO?" species above pH 5 and the corresponding
 increase  in the fraction  of  CrOff"  species, which become the  principal  species
 present  in solution at  about pH  8.5  (Fig. 28). Below pH 2  the  fraction of
       ions decrease rapidly  as the fraction  of HaCrOi*  species  increases.
      The behavior of  the  Cr(VI)  adsorption-pH  curves  in  Figures  26  and 27
 implies that  the HCrO^  ion  is  the principal  ion being adsorbed by the clay
 minerals.  Conversely, the lack of adsorption at pH values  above  8.5 indicates
 that  the CrO*"  ion is not adsorbed  at all by either of the two clays.  The
 mechanism of  Cr(Vl) adsorption by these clay minerals apparently cannot
 neutralize the  two negative charges present  on the Cro£~ ion. On the other
 hand,  Cr207~  ions may be  adsorbed at low pH  values because of its more open
 structure, CrO^ - 0 - CrOl,  which places the two negative  charges an ap-
 preciable distance apart, as opposed to charges on adjacent oxygens that
 occur  in the  CrO£~ structure.  The charge separation on the CrzQ? ion may  al-
 low it to act essentially as two monovalent  ions, with each negative charge
•fulfilling an adsorption  site  or one negative  charge  fulfilling  an  adsorption
 site  and the  other negative charge  being neutralized  by  a  cation in solution.
 If  the distribution of  Cr(VI)  species in solution is  the only factor govern-
 ing adsorption  of Cr(VI)  over  the pH range 2 to 5, a  plateau should be ob-
 served in the Cr(VI)  adsorption-pH  curves; however, the  adsorption  of Cr(VI)
 continues  to  rise with  decreasing pH to a pH value around  2. Therefore, in
 the low pH range the  pH probably modifies the  structures of the  clay minerals
 to  permit  increased Cr(VI)  adsorption to occur. Dissolution of clay minerals
 is  known to occur at  low  pH levels  (Hofmann  et al. , 1956). This  can alter  the
 surface structure and surface  area  of the clays, resulting in changes in
 their  adsorption characteristics as the pH is  lowered.

     Adsorption isotherms were constructed from the Cr(VI) adsorption-pH

                                     86

-------
   0.70-
   0.60-   93)
ol
e 0.50 H
-2 0.40-
E
2 0.30 H

-------
              I    I   I
                           J	u
-3,0.200-
\
o>
 § 0.150-
 o


 E 0.100-
 o
•o

-e 0.050-
o
o 0.000-
           -•185
A.K2Cr04

    SOLUTION
B. DU PAGE

    LEACHATE
             2.0   4.0    6.0   8.0    2.0    4.0    6.0    8.0


                                   PH
  Figure 27.   Chromium (VI) adsorption-pH curves for kaolinite at 25° C.

             Initial Cr(VI) concentrations  (ppm) are indicated beside

             each curve and the equivalence volume for each curve is

             10.Q ml/g kaolinite.
                                 88

-------
    1.0-
 
^—"^1
N
0 0.6-
«4—
 O
 C
 O

I 0.4-
   0.2-
   0:0-
                     a.  I.OxlO"4M
                     b.  I.OxiO"3M
                     c. 6.0 x IO"3M
            1.0
                 I.  HCr04~
                 2.  Cr04=
                 3.  H2Cr04
                 4.  Cr207=
3.0
5.0
PH
7.0
9.0
Figure 28.  Distribution of Cr(VI)  species for various Cr(VI) concentra-
           tions.
                               89

-------
curves in Figures 26 and 27 using Equation {16} at pH values 3.0, 4.0, 5>Q,
and 7.0. Sample adsorption isotherms, constructed at pH 4.0, are shown in
Figure 29. For both clay minerals, more Cr(VI) was adsorbed from DuPage
leachate solutions than from pure KaCrOi* solutions throughout the pH range
3.0 to 7.0. This result is contrary to that which was expected. Evidently,
anions (e.g., Cl~ and HGO^) in the leachate do not compete favorably with
HCrOij ions, or adsorption would have decreased.

     The effect of the Cl~ ion on the adsorption of Cr(VI) by clay was deter-
mined by adsorption experiments carried out with 20 ppm Cr(VI) in deionized
water with and without 1000 ppm of Cl~ added as NaCl. No appreciable change
in the adsorption of Cr(VI) was caused by the Cl~ ion. That more Cr(VI) was
adsorbed from DuPage leachate than from pure KaCrOi* solutions may be the re-
sult of formation of polynuclear complexes in the leachate solutions, organic
or inorganic in nature, which can be adsorbed by the clay. The high ionic
strength of the leachate may also contribute to higher adsorption of Cr(VI)
species by a depression of the diffuse double-layer surrounding the clay
particles  (van Olphen, 1963), which allows more ions to approach the clay
surface and be adsorbed.

     The adsorption isotherms for Cr(VI) at pH values of 3.0, 4.0, 5.0, and
7.0 were plotted according to the linear form of the Langmuir equation (Equa-
tion 16). All the Langmuir plots gave linear regression r2 values 0.99. The
Langmuir plots at pH 4.0 are shown in Figure 30. From the slopes of the Lang-
muir plots, adsorption maxima were calculated and are presented in Table 12.
The difference in the calculated adsorption maxima of montmorillonite and
kaolinite reflects the difference in the probable number of available adsorp-
tion sites, based on comparison of the structural differences and the surface
areas of the two clay minerals (Table 1). The precise mechanism for anion ad-
sorption by clay minerals is uncertain, but we assume, as have others, that
anion exchange plays an important role in the adsorption process. However,
the adsorption maxima presented in Table 12 represent the maximum amount of
Cr(VI) ions adsorbed at some sufficiently high concentration of Cr(Vl) ions
in solution, whereas the adsorption isotherms in Figure 29 represent the
maximum amount of Cr(VI) ions that can be adsorbed at any given concentration
of Cr(VI) ions in solution.

Chromium  (III) Adsorption

     During preliminary experiments on Cr(IIl) adsorption by kaolinite, the
removal curves shown in Figure 31 were obtained. The curve labeled "blank"
in Figure 31, representing removal of Cr(III) from a solution containing no
clay, shows that precipitation becomes a very important mechanism of Cr(IIl)
removal near pH 5.0. The precipitate formed was blue-gray, and X-ray diffrac-
tion patterns of the precipitate showed no definite crystalline structure.
Because only HN03 and NaOH were used to adjust the pH of the Cr(1*03)3 solu-
tion, it is reasonable to believe that the precipitate formed is a chromic
hydroxide. Murray (1956) stated that chromic hydroxide is a hydrous oxide
(Cr20s • nHaO) of indefinite composition that is blue-gray when its water
content is high. It was therefore concluded that chromic hydroxide was the
precipitate of our experiments.
                                     90

-------
   0.60-
   0.50-
 o. 0.40H
 o>
 E
^ 0.30-
o
I 0.20H
o
   0.10-
     0-
                     Montmorillonite in
                     Du Page leachate
                     -X	-X
                         Montmorillonite in K2Cr04 solution
          Kaolinite in Du Page leachate
                          /—Kaolinite in K
                         *o	-o
                                solution
             0
           100           200          300
Equilibrium Cr concentration  (ppm)
    Figure 29.  Adsorption isotherms for Cr(VI) at pH 4.0 and 25° C.
                              91

-------
  400-
   300H
O
 x
o
    100-
     0-
               Kaolinite in
                         solution
                           Koolinite in DuPage leachate
                                         Montmorillonite in
          0
                        Montmorillonite  in
                         Du Page leachate
          100             200
Equilibrium Cr concentration (ppm)
Figure 30.  Langmuir plots of Cr(VI)  adsorption data at pH 4.0 and 25° C.
                             92

-------
TABLE 12.  ADSORPTION MAXIMA FOR Cr(VI)  BY MONTMORILLONITE AND
                          -O
           KAOLINITE AT  25   C  FOR VARIOUS  pH VALUES
                                    mg/g
                                               DuPage
    pH              Pure Solution         Leachate Solution

                               Montmorillonite

    3.0                 0.400                   0.667
    4.0                 0.256                   0.526
    5.0                 0.147                   0.417
    7.0                 0.052                   0.169

                                  Kao Unite

    3.0                 0.093                   0.189
    4.0                 0.044                   0.130
    5.0                 0.032                   0.115
    7.0                 0.015                   0.051
                              93

-------
   100-
   90-
   80-
   70-
0>
I  60H
o
>»
•°  50H
O)

I  40H

-------
     Families of Cr(III)  adsorption-pH curves for the PH range 1.5  to  4.5 are
presented in Figure  32. Cation exchange is generally accepted as  the princi-
pal mechanism for  cation  adsorption by soils and clay minerals. Chromium
(III) is known  (Rollinson,  1956)  to be extensively hydrolyzed in  acid  solu-
ni0^l+SP^ieS SUCh  SS  Cr(°H)  '  Cr(°H)i  °r Cr2(OH)*+ or Cr2(OH12)6+, and
Cr6(OH)15. Therefore,  the increasing adsorption of Cr(III)  as pH  increases
can, in part, be attributed to exchange-adsorption of hydrolyzed  Cr(III)
species other than Cr   ions.

     Adsorption isotherms were constructed from the adsorption-pH curves
(Fig. 32) by using Equation {15}  at pH values 2.5, 3.0,  and 4.0.  The iso-
therms constructed for pH 4.0  are shown in Figure 33. As was expected, the
adsorption of Cr(III)  by  the clay minerals is lower in DuPage leachate than
in pure Cr(N03)3 solutions. However, significantly less  reduction in adsorp-
tion of Cr(lII) from DuPage leachate (3% to 14%) took place than  observed in
tests of Pb, Cu, Zn, and  Cd adsorption from DuPage leachate by the  same clay
minerals. The Cr(III)  species  existing in solution are so strongly  adsorbed
that the cations present  in the DuPage leachate do not effectively  compete
with the Cr(lII) species  for exchange-adsorption sites.

     The adsorption  isotherm data for Cr(III) at pH values  2.5, 3.0, and 4.0
were plotted according to the  Langmuir equation (Equation {16}) and  the Lang-
muir plots for  the pH  4.0 adsorption isotherms are shown in Figure  34. Ad-
sorption maxima calculated from the slopes of the Langmuir  plots  in Figure 34
are presented in Table 13.  The amounts of the various hydrolyzed  Cr(III)
species that could be  adsorbed via a cation exchange mechanism by the mont-
morillonite and kaolinite clay minerals that had CEC values of 79.5 meq/100 g
and 15.1 meq/100 g,  respectively, are presented in Table 13,  because, if
cation exchange is the principal  adsorption mechanism, the  ratio  of the ad-
sorption maxima should be about equal to the ratio of the CEC values for the
two clay minerals.

     At pH 2.5  the adsorption maximum ratio is close to  the CEC ratio  (Table
13) , but the adsorption maxima themselves are higher than those based on ex-
change adsorption  of Cr3+ ions. This implies that some hydrolyzed Cr(III)
species are being  adsorbed even at pH 2.5. The question  arises, however, as
to the validity of using  CEC values obtained from Ntft ion exchange  at pH 7.0
to calculate the expected adsorption maxima at pH 2.5. However, as  there is 'a
general relative correlation between the adsorption maxima  and CEC  values, it
is reasonable to use adsorption maxima calculated from CEC  values for com-
parison purposes.  At pH 4.0, the  ratio of the adsorption maxima increases to
about 13, and the  amount  of Cr(III) adsorbed by montmorillonite is  even
larger than would  be expected from the exchange adsorption  of the Cr6(OH)i5
ion.

     The high adsorption  maximum  ratio at pH 4.0 is the  result of the large
adsorption maximum obtained for montmorillonite, which appears to be in-
directly caused by our experimental procedure. The rates of some  of the
Cr(III) hydrolysis reactions are  very slow, and, in one  reported  case  (Las-
wick and Plane, 1959), about 107  days were required for  Cr(IlI) solutions,
even at elevated temperatures, to reach equilibrium. In  our Cr(III) experi-
ments, the clay mineral suspensions were adjusted to a particular PH value,

                                     95

-------
   150.0-
   100.0-
 o>
 c
    50.0H
 o
 o
     0.0
 .O
 O
   100.0-
 o
    50.0-
     0.0-
A.Cr(N03)3
   SOLUTION
   KAOLINITE
 C. Cr(N03)3
    SOLUTION
MONTMORILLONITE^
                 765
                 ;388
                                   i    .   i    i   I	1	L

                                   B.DUPAGE  LEACH ATE
                                      KAOLINITE
     D. DU PAGE
        LEACHATE
   MONTMORILLONITE
          1.5
      2.5
              3.5
4.5 1.5
  pH
2.5    3.5
4.5
Figure 32.  Chromium (III) adsorption-pH curves at 25° C.  Initial Cr(III)
          concentrations (ppm) are indicated beside each removal curve,
          and the equivalence volume for each curve is 500.0 ml/g kaoli-
          nite.
                              96

-------
   100-
o>
E
•o
0)
T3
o
i_
o
     0-
         Montmorillonite
       in Du Page leachate
             Montmorillonite in
            /Cr(N03)3  solution
                                                   -X
                      Kaolinite in Cr(NO,), solution
Kaolinite in Du Page leachate
                100   200   300  400   500   600   700
                Equilibrium Cr concentration (ppm)
Figure 33.  Chromium (III) adsorption isotherms at pH 4.0 and 25° C.
           The plots shown in dotted lines were obtained from the
           "corrected" adsorption-pH curves.
                              97

-------
    30
  E  20
  V.
  X

 
-------
TABLE 13.  ADSORPTION MAXIMA FOR Cr(III) BY MONTMORILLONITE AND KAOLINITE AT 25° C FOR VARIOUS pH VALUES
(mg/g)
pH 2.5
Pa Lb

17.9 --

3.3 —

5.4 —
pH 3.0 pH 4.0 Values
PL PL Cr+3
Montmorillonite
33.7 32.8 139.6 136.1 13.8
Kaolinite
5.0 5.0 10.7 14.7 2.6
Ratio of Adsorption Maxima
6.5 6.6 13.0 9.3 5.3
Based on CEC for Different Species
Cr (OH) 2+ Cr (OH)t Cr 6 (OH) ft

20.7 41.4 82.7

3.9 7.8 15.6

5.3 5.3 5.3
   a               b
    Pure  solutions,  DuPage  leachate solutions

-------
such as 4.5. After a few hours, hydrolysis caused the pH of the suspensions
to drop. The pH of the suspensions were readjusted to 4.5, but the pH again
dropped because of hydrolysis. The pH of the suspensions were all readjusted
to the desired pH value several times during a period of two weeks. They were
then shaken for 2 days before a final pH value was measured, and the suspen-
sions were centrifuged. What appears to occur in the pH range 3.5 to 4.5 is
as follows. Adsorption plus possible precipitation of Cr(III) takes place at
the higher initial pH values to which the suspensions had been adjusted. But,
when the pH of the suspensions drops due to hydrolysis, desorption of Cr(III)
species from the clay mineral or dissolution of any precipitate formed ap-
parently does not occur at rates fast enough to achieve true equilibrium in
the clay suspension. The pH values of the montmorillonite suspensions showed
larger decreases in pH than the kaolinite suspensions; this apparently pro-
duces much larger differences between the calculated and true equilibrium ad-
sorption maxima for montmorillonite than for kaolinite.

     The adsorption-pH curves shown in Figure 32 were replotted as the amount
of Cr(III) adsorbed versus the highest pH values to which the clay suspen-
sions were adjusted. The adsorption isotherms were calculated from the "cor-
rected" adsorption-pH curves at pH 4.0 and are shown as the dotted isotherms
for montmorillonite in Figure 33. The "corrected" adsorption isotherms for
kaolinite are almost superimposable on the isotherms shown in Figure 33, and,
therefore, were omitted. Adsorption maxima from pure Cr(NOa)3 solutions, cal-
culated by Langmuir plots of the "corrected" isotherms at pH 4.0, are 72.2
mg/g for kaolinite. Thus, the "corrected" adsorption maxima agree with an ex-
change-adsorption mechanism involving hydrolyzed Cr(III) species. We have
learned, as Laswick and Plane (1959) pointed out, that the changes that
Cr(III) species undergo are generally quite slow, which makes the interpre-
tation  of the experiments difficult.

CONCLUSIONS

     The results of this study indicated that Cr(III) yielded the strongest
attenuation of all the heavy metals studied. This suggests that landfill dis-
posal of Cr(III) wastes would exhibit low mobility and probably would initi-
ate fewer pollution problems than would the other heavy metals. Above pH 6,
Cr(III) should be immobile because of precipitation. Below pH 4, Cr(III)
species were strongly adsorbed by both kaolinite and montmorillonite and
would have a relatively low mobility through soils or clay minerals used as
landfill liners. Between pH 4 and 6, the combination of adsorption and pre-
cipitation should render Cr(III) quite immobile.

     In the pH range 1.5 to 4.0, 30 to 300 times more Cr(III) than Cr(VI) was
adsorbed by the clay minerals; and at higher pH values the ratio (Cr(III) re-
moved) /(Cr(IV) removed) became even larger because of increased Cr(III) re-
moval and decreasing adsorption of Cr(VI). The adsorption of Cr(VI) was low
relative to Cr(III), even at very low pH values where Cr(VI) adsorption was
strongest. Cr(VI) adsorption was the weakest of the heavy metals studied and
was markedly reduced as the pH was raised into the alkaline range. Thus, it
would become very mobile at high pH. Since Cr(VI) is the most toxic and mo-
bile form of Cr, landfill disposal of Cr(VI) wastes can potentially cause
serious pollution problems even if the landfill has a thick clay liner. The

                                     100

-------
results of this study suggest that  a  conversion  of Cr(VI) wastes to Cr(III)
by a process such as that devised by  Shiga  (1975) would greatly reduce the
hazard to water resources from  Cr(VI) wastes  disposed of  in landfills.
                                       101

-------
                                 SECTION 10

                   EFFECT OF pH ON ADSORPTION OF As AND Se
                   FROM LANDFILL-LEACHATE BY CLAY MINERALS1
ABSTRACT

     The effect of pH and ionic competition on arsenate (As(V)), arsenite
(As(III)), and selenite (Se(IV)) adsorption by kaolinite and montmorillonite
clay minerals from municipal landfill-leachate solutions were determined.

     The results showed that pH had a strong influence on the amounts ad-
sorbed of all three of the elemental forms studied. Montmorillonite clay was
found to adsorb about twice as much As or Se as kaolinite. Leachate was found
to have little effect on the adsorption of As(V) or Se(IV), while As(III) ad-
sorption was reduced 30 to 50%. It was concluded that the principal adsorp-
tion mechanism was anion exchange of the monovalent species of each elemental
form.

     The results of the study suggest that land disposal of As and Se wastes
under alkaline conditions represents a potentially high pollution hazard.

INTRODUCTION

     As and Se are quite toxic  (U.S.-EPA, 1972) and, therefore, they have a
high potential to produce pollution problems. The U. S. Environmental Protec-
tion Agency (EPA) cites several examples of instances where land disposal of
arsenic wastes have resulted in the poisoning of drinking water wells (U.S.-
EPA, 1973) and selenium has been reported to have polluted ground water as
far as two miles from a dump on Long Island (Garland and Mosher, 1975). As
and Se waste streams may be liquids, suspensions, or sludges. Sample waste
streams from copper, lead, and zinc smelting, from duplicating and photo
equipment manufacturing, and from pharmaceutical industries have been
measured and reported to contain from 1,000 to 30,000 ppm As and from 3,000
to 50,000 ppm Se (Lehman, 1973). These waste streams may be disposed of on
land in lagoons, in landfills, or by spreading. Land spreading of municipal
wastewater effluents (U.S.-EPA, 1975) and sewage sludges which may contain
low concentrations of As and Se is now being considered as a viable alterna-
tive to treatment.

     The purpose of this study was to provide some basic data regarding the
effect of pH on the removal of As and Se from landfill-leachate by kaolinite
and montmorillonite, which are common clay minerals found in soils. The re-

1Authors: R. R. Frost and R. A. Griffin

                                     102

-------
suits of the present study  give insights into whether or not  As  and  Se waste
streams can be safely disposed of in properly designed landfills for munici-
pal solid waste. The results  of the study also provide basic  information on
the mobilities of As and  Se through soils.  This can be of aid in the design
of land disposal systems  for  As and Se waste streams in general.

EXPERIMENTAL

     Appropriate stock  solutions of As(V) and As(III) were prepared  by dis-
solving reagent grade Na2HAs0lf or NaAs02 in deionized water.  Se(IV)  stock
solutions were prepared by  dissolving pure selenium metal in  a minimum amount
of 1:1 HNOs. Some heating of  the solution was necessary to speed up  the dis-
solution. All stock solutions were adjusted to about pH 5 before use.

     Fifty ml aliquots  of leachate were pipeted into 125-ml Erlenmeyer flasks
containing from 1.00 to 5.00  g of either kaolinite or montmorillonite. The
weight of clay used was chosen so that the amount of arsenate, arsenite, or
selenite removed from the leachate solutions could be determined with some
precision from the difference between the initial and final solution concen-
trations. Several replicates  of each clay suspension were prepared.  The pH
values of the replicate clay  suspensions were adjusted with HNOa or  NaOH to
different values over the pH  range 1 to 9.  The clay suspensions  were shaken
overnight and then 2.0  ml of  an appropriate pH 5 stock solution  of either As
or Se was pipeted into  the  flasks. Of the several experimental procedures
possible, the procedure used  gave the most satisfactory results. The As or Se
clay suspensions were then  shaken in a constant temperature bath at  25. ± 0.5°
C for at least 24 hours to  insure complete equilibration. The equilibrium pH
values of the clay suspensions were measured, the clay suspensions centri-
fuged, and the supernatant  solutions analyzed for their As or Se concentra-
tion by atomic adsorption spectroscopy. A NO-CaHz flame was used for As
analyses, and air-CaHa  flame  was used for Se analyses. Blanks (i.e., no clay)
were prepared along with  the  clay suspensions and were also analyzed to de-
termine the  initial As  or Se  concentration. The amount of As  or  Se removed
from solution by a given  clay at a particular pH was determined  as:

                              .   total sol'n vol. after pH adjustments
  (initial -  equilibrium cone.) x 	sample weight.

The amount of As or Se  removed from solution was then plotted as a function
of pH. From  the initial concentration of As or Se in solution, the weight of
clay used, the final solution volume after pH adjustments,  and the removal
versus pH curves themselves,  "adsorption" or "removal" isotherms can be con-
structed at  different pH  values by use of the following equation:

              Equilibrium C (ppm)  =  Initial C (ppm)  -
                      Am't. removed (ug/g)  x Wt. Clay (g).
                          Final solution volume (ml)

RESULTS AND  DISCUSSION

Arsenate (As(V)) and Arsenite (As(III)) Adsorption
                                     103

-------
     Curves for the removal of As(V) from leachate by kaolinite  and  mont-
morillonite versus pH are shown in Figure 35. It is interesting  to note that
the amount of As(V) removed from solution goes through  a maximum at  about pH
5. The distribution of arsenate species in solution as  a function of pH is
shown in Figure 36a. Comparison of Figure 35 with Figure 36a  shows that the
As(V) removal curves follow the monovalent HaAsO^ species  curve  reasonably
well. Thus, it was concluded that the HaAsOIT ion is the principal As(V) ion
being adsorbed by the clay minerals. The non-adsorption or depressed adsorp-
tion of the HAsOij: ion is apparently due to the occurrence  of  negative charges
on adjacent oxygen atoms in the tetrahedral HAsO™ ion which results  in repul-
sion of the ion from the clay surface. Some As(V) precipitation  was  observed
to occur above approximately pH 9 and can be seen on the removal curves in
Figure 35 as the curves turn upward around pH 9.

     Curves for the removal of As(III) from leachate by kaolinite and mont-
morillonite versus pH are shown in Figure 37. In general,  an  increase in ad-
sorption of As(III) is observed as pH increases in the range  3 to 9.  The
montmorillonite removal curves in Figure 37 show an interesting  peak about pH
7. There is some question as to what As(III) species are present in  the
leachate solutions and which As(III) species is actually being adsorbed.  From
a study of As(III) adsorption by an anion exchange resin (Everest and Popiel,
1957), the AssOs on As3(OH)T0 species was proposed as the  species being ad-
sorbed in the pH range 5 to 6, but the fraction of the total  As(III)  in solu-
tion existing as the AsaOs (As3(OH)Io) species is small and the  amount  of ex-
change adsorption was low. Other As(III) species apparently existing in solu-
tion are As2C% (As2(OH)f), As(OH)3, As(OH)^, and As020H= — with the latter
two species becoming important above pH 9. Whatever As(III) species  are
present in solution, the fraction of the total As(III) in  solution present as
monovalent As(III) species will increase as pH increases.  Thus,  the  amount of
As (III) adsorption should increase as pH increases as observed in Figure 37.
There is no apparent explanation for the observed peaks about pH 7 on the
montmorillonite removal curves in Figure 37. Although no peaks are observed
o.n the kaolinite removal curves in Figure 37, they might have appeared if
more data points were obtained in the pH range 6.5 to 7.

     Comparison of removal curves (not shown) for As(V) and As(III)  from de-
ionized water solutions obtained at a single concentration for each  clay
mineral with the corresponding leachate removal curves showed that the  anion
competition present in the leachate surpresses the adsorption by the clay
minerals of As(V) slightly and of As(III) by some 30 to 50 percent.

     Isotherms calculated from the removal curves in Figures  35  and  37  at pH
5.0 are shown in Figure 38. The higher adsorption of As(V) and As(III)  by
montmorillonite compared to kaolinite simply reflect the structural  and sur-
face area differences between the two clay minerals. Anion exchange  sites are
thought to exist primarily at the broken edges of clay minerals  (Grim,  1968).

     Montmorillonite clay was found to adsorb approximately twice as much As
as kaolinite. Examination of the surface area data, as measured  by N2 gas ad-
sorption given in Table 1, suggests that the montmorillonite  has roughly
twice as much edge surface area as the kaolinite. The tetrahedral H2AsO£ ion
can align itself with the silica tetrahedral of the clay lattice and can form

                                     104

-------
         700
         600-
400-

        300-
       o
       E
       0}
         zoo-
         100-
                       Kaolinite
             190
               I O.I
                                                   Montmorillonite
                                                 128.
                                59.O
                                         21.3
                                      pH
                                                          -i	1	r
Figure 35.   The amount of As(V) removed  from DuPage leachate solutions
             by kaolinite and montmorillonite at 25° C plotted as a func-
             tion of pH.  Labels are  the  initial solution concentration
             of As(V) in ppm.  Each datum point was obtained by using
             either 4 g of kaolinite  of 1  g  of montmorillonite in a total
             solution volume of 52.2  ml.
                                    105

-------
       A.  As (32)
B. Se(I2)
 I    i	I	I	I	L
Figure 36.  Species distribution diagram for As(V) and Se(IV).

-------
   400-
 O)
 o
 o
   300-
 E
 O 200
 o
 E
 o>

 V)
   100-
                   J	L_
             Koolinite
                                 68.0
     Montmorillonite
024
                                 I
                                 8
 0
pH
                                                                 8      10
Figure 37.  The amount of As(III) removed from DuPage leachate solutions
            by kaolinite and montmorillonite at 25° C plotted as a func-
            tion of pH.  Labels are the initial solution concentration of
            As(III) in ppm.  Each datum point was obtained by using 4 g
            of clay in a total solution volume of 52.5 ml.
                                   107

-------
jcn
                                            As(3Z) by montmorillonite
                                                   As(3Z:)by  kaolinite
                            As(m) by montmorillonite
                                         kaolinite
                         60
                               80
                                     100
120
                                                  140
                                                         160-
                                                               180
                      Equilibrium concentration  (ppm)
  Figure 38.  The amount of As(V)  or As(III) removed from DuPage leachate
             solutions at pH 5.0  and 25° C per gram of clay plotted as a
             function of the equilibrium arsenic concentration.
                                  108

-------
an extension of the  crystal lattice which has a relatively high bond strength.
The arsenate ion can be  compared to the large As3 (OH) To ion which is not
tetrahedral and thus is  unable to align itself as effectively on the clay
edges. It was therefore  concluded that the principal adsorption mechanism was
anion exchange of  the  monovalent species of each elemental form.

Selenite  (Se(IV))  Adsorption

     Curves for the  removal of Se(IV) from leachate by kaolinite and mont-
morillonite versus pH  are shown in Figure 39. The removal of Se(IV)  goes
through a maximum  in the pH range 2 to 3 and then decreases as the pH in-
creases. The distribution of Se(IV) species in solution as a function of  pH
is shown in the Figure 36b. Although it is not as apparent as for As(V),  the
monovalent HSeOJ ion appears to be the species predominantly adsorbed by  the
clay minerals. It  is evident from the data in Figure 39 that the adsorption
of Se from solution  is rapidly converging on zero removal at approximately pH
10, which is the value at which the HSeO'a species disappears from solution.

     This is evidenced by examining the species distribution diagram pre-
sented in Figure 36b.  The HSeOJ ion has a trigonal-pyramidal configuration
which may account  for  its reduced adsorption; by contrast, the HaAsO^ ion has
a tetrahedral configuration. Thus, the configuration of the HSeO^ ion must be
an inhibiting factor in its adsorption by the clay minerals. On the  other
hand, pH must play a significant role in modifying the clay mineral  surface
structure so that  increasing adsorption occurs with decreasing pH until the
point where a significant fraction of the total Se(IV) in solution is present
as the HaSeOs specie.  At this point, adsorption starts to decline sharply as
the pH is lowered  below a value of 2.

     Comparison of removal curves (not shown) for Se(IV) from deionized water
solutions obtained at  a single concentration for each clay mineral with the
corresponding leachate removal curves showed that the leachate supresses  the
adsorption of Se(IV) by the clay minerals slightly. This is presumed to be
due to competition for anion exchange sites by the high concentrations of
anions present  in  leachate (Table 2) with the HSeO?.

     Isotherms  calculated from the removal curves in Figure 39 at pH values
3.0, 5.0, and 7.0  are  shown in Figure 40. The higher adsorption of Se(IV)  by
montmorillonite compared to kaolinite simply reflects the structural and  sur-
face area differences  between the two clay minerals as previously discussed.

CONCLUSIONS

     The  results of this study indicate that pH has a pronounced effect on
the amounts of As(V),  As(III), and Se(IV) adsorbed from leachate by  clay
minerals. It was concluded that the principal adsorption mechanism was anion
exchange  and, from species distribution diagrams, that the adsorption was
due principally to the monovalent species of each element studied, thus lead-
ing to the strong  pH dependency of the adsorption process.

     The  results of the study suggest that optimum As removal by soil ma-
terials from waste streams would be achieved by conversion of any As (.III J to

                                     109

-------
    600-
    500-
  0>
    400-
  o
  o
  £
  o
  ,£ 300-
  -a
  
  o
  E
  o>
  t.
  
-------
  250-
           I	1	1     I     I
                                      Montmorillonite
          10    20   30   40   50   60   70    80   90   100   110   120
                      Equilibrium  concentration (ppm)
Figure 40.   The amount of Se(IV) removed from DuPage  leachate solutions
            at 25° C and several pH values per gram of  clay plotted as
            a function of the equilibrium Se concentration.
                                  Ill

-------
As(V), and disposal of the waste in montmorillonitic soils at pH 5. Optimum
Se(IV) removal would result from disposal in a montmorillonitic soil at a pH
of 2 to 3. A high mobility and hence potential pollution hazard would be ex-
pected from land disposal of As(V)  or Se(IV) wastes under alkaline condi-
tions.
                                    112

-------
                                  SECTION 11

                        MERCURY REMOVAL FROM LEACHATE
                               BY CLAY MINERALS1
ABSTRACT
     Preliminary investigations regarding  the effect of pH on the amounts of
Hg removed from DuPage leachate solutions  by the two clay minerals, kaolinite
and montmorillonite, are reported.

     Large amounts of Hg were removed  from leachate, both in the presence and
absence of clay. This removal of Hg  increased as the pH was raised. It was
estimated that 70 to 80% of  the Hg was removed  from leachate by precipitation
and/or volatilization, thus  indicating that these were the predominant at-
tenuation mechanisms. Removal by clay  accounted for 20 to 30% of the observed
Hg removal. About two thirds of the  Hg removed  by the clay was found to be
organic-Hg, and about one third was  inorganic-Hg.

     The results indicated that maximum Hg removal from leachate would be
achieved by disposal in montmorillonite clays or soils under alkaline pH con-
ditions.

INTRODUCTION

     Because Hg has been the subject of many publications, its occurrence and
toxicity will be discussed only briefly here. There are two basic sources of
Hg in the environment: as it occurs  in nature,  and as it is redistributed in
nature by man's actions. Hg  is widely  distributed in the air, soil, and water
in low concentrations (Klein, 1972). Man in his utilization of Hg-related
technology has at times redistributed  this Hg in nature. It is this source of
Hg in the environment that we must concern ourselves with since it often re-
sults in the release of dangerously  high Hg concentrations (Goldwater, 1971).

     The main source of environmental  Hg contamination has been in the in-
dustrial sector (D'ltri, 1972a) , and much  of this can be attributed to waste-
water discharge into rivers  and streams (Turney, 1972; Derryberry, 1972). A
great deal of the mercurial  waste, both organic and inorganic, disposed of in
this manner can be directly  or indirectly  converted by anaerobic microbes in-
to mono- or di-methyl-mercury (D'ltri,  1972a; Greeson, 1970), both of which
have been found to be extremely toxic.

     As a result of studies  on Hg contamination and poisoning, the disposal

Authors: R. A. Griffin and  G. D. Robinson


                                     113

-------
of industrial mercurial wastes is carefully controlled. However, the po-
tential for contamination still exists where gaps in knowledge preclude en-
lightened disposal practices. This potential also exists at municipal land-
fill sites accepting mercurial wastes. The leachates generated from these
sites are anaerobic and therefore have the ability to convert mercurial
wastes into their toxic forms. Since Hg is incorporated into many industrial
or consumer products such as paints, Pharmaceuticals, paper products, fluor-
escent lamps, mercury batteries, etc., the indiscriminate disposal of these
products in landfills by an uninformed population represents an important
potential environmental Hg contamination route (D'ltri, 1972b).

     There has been very little research done on the chemical behavior of Hg
in municipal leachates. The information that is available to date on Hg in
municipal leachates is insufficient to evaluate the migration of Hg in land-
fills. It is therefore our purpose here to report the results of studies on
the adsorption of Hg from leachate by clay minerals. These studies were
initiated to determine whether or not clay minerals have potential for use as
landfill liners for attenuation of Hg which may be present in leachates.

EXPERIMENTAL

     In order to determine the amount of Hg removed from solution, a series
of samples were prepared using HgCla in DuPage landfill-leachate, and HgClg
in DuPage landfill-leachate with clay (kaolinite or montmorillonite) added.
The latter two treatments were also used to determine the amounts of organic
and inorganic Hg removed from solution.

     The amounts of organically-bound Hg and inorganic Hg in solution were
determined by the flameless Atomic Absorption Spectroscopy (AA) technique
described in the appendix.

     Eight samples were prepared for each solution in each experiment. Before
addition of Hg, 50 ml aliquots of either deionized-H^Oj DuPage leachate, or
DuPage leachate and 1.000 g kaolinite were added to 125 ml Erlenmeyer flasks.
The pHs of the solutions were then adjusted using 0.1N, IN, 3N, and/or cone.
HNOs to obtain pH values between 1 and 9 (each set of samples spanned this
range). The samples were placed in a shaking waterbath at 25 ± 0.5° C, and
allowed to equilibrate overnight. The pH values were checked after 20-24
hours and 2 ml aliquots of 25 ppm Hg stock solution were added to each flask
giving a total volume of 52 ml and a Hg concentration of approximately 0.962
ppm. The samples were again placed in the shaking waterbaths and allowed to
equilibrate overnight. After 20 to 24 hours the final pH values of the
samples were recorded, and each sample was transferred to a 50 ml centrifuge
tube and centrifuged at 20,000 rpm for 5 minutes. The supernatant solutions
were decanted, acidified, and analyzed for their Hg content using flameless
AA.

     The amount of Hg removed from solution by a given clay at a particular
pH was determined as follows:
                                    114

-------
                           r      (CI  "  CEo)   *  v?
                           CR  =  	T	E                      (19}

where CR = amount  of Hg removed in yg/g clay.
      Cj = initial Hg  concentration in ppm,
      CEq = equilibrium Hg concentration in ppm,
      VF = total solution volume after pH adjustments in ml,  and
      W = weight of clay in grams.
The amount of Hg removed from solution was then plotted as a  function  of pH.

RESULTS AND DISCUSSION

     Some examples of  results obtained for Hg removal from various  solutions
plotted as a function  of pH are presented in Figure 41. It can be seen from
this figure that Hg is removed from solution,  even from a presumably sterile
and pure HgCl2  solution (curve A). In the short time available for  this study
we were not able to satisfactorily isolate all the reasons for this removal
of Hg. Three possible  removal mechanisms are adsorption onto  the walls of  the
glassware and plastic  bottles, precipitation,  or volatilization. At the lower
pH values the Hg removal is thought to have been by adsorption onto the
glassware, although standard solutions may be satisfactorily  stored at pH  2.
It is presumed  that the adsorption in the sample solutions occurred at the
higher initial  pH  values and that the Hg did not have sufficient time  to
totally desorb  after the pH was adjusted to lower values. The increase in  re-
moval in the higher pH range (5-8) is apparently the result of both precipi-
tation and adsorption. Volatilization was not  considered as an important
mechanism for the  Hg losses observed from the HgCl2 solutions since they were
presumably sterile and stoppered.
     An increasing amount of Hg removal is displayed in curve B (leachate
blank) which illustrates the loss of Hg from a leachate solution. The  removal
of Hg from solution here is again probably due to the same mechanisms; that
is, the increase in Hg removal (greater amount of Hg removed  by leachate than
by deionized-H20)  is probably the result of either increased  adsorption or
increased precipitation. However, volatilization losses of organic  Hg  com-
pounds due to microbial transformations in the leachate are also possible.

     Curve C in Figure 41 shows the total amount of Hg removed from leachate
with 1 g of kaolinite  present. The amount of Hg removed by 1  g of kaolinite
is taken as the difference between curve C and curve B in Figure 41. This  is
plotted as curve B in  Figure 42.
     Figure 42  shows  a breakdown of various forms of Hg removal from leachate
solutions. The  amount  of organic- and inorganic-Hg removed by clay  was deter-
mined by placing an  aliquot from each sample into each of two BOD bottles.
One set was digested  following the general procedure while the second  set  was
analyzed undigested. A set of leachate blanks was treated in  the same  manner.
There were then four separate sets in all, which included a digested blank
 (HgCl2 in leachate), an undigested blank, a digested sample (HgCl2  in  leach-
ate with 1 g kaolinite), and an undigested sample.

The total organic-Hg in solution was determined by taking the difference
                                      115

-------
     0.7
     0.6-
     0.5-
c
o
1  0.4H
o

o
E
£  0.2H
o>
x
     O.I -
     0.0
                HgCI2  in leochate
               81 Ig kaolinite
                                              HgCI2  solution
                2.0
                      3.0
4.0
5.0
PH
6.0
7.0
8.0
Figure 41.  Removal of Hg from DuPage  landfill-leachate and pure
            solutions plotted as a function  of pH at  25° C.  The initial
            Hg concentration is 0.96 ppm  and the  final volume is 52 ml.
                                   116

-------
     0.7
     0.6-
  3 Q5~

  0)

  o
  .c


  o  0.4-
  Q>
  e
  o
  I
  d>
     0.3H
     0.2-
      0.1-
     0.0
                Total Hg removed


                 from leachate
                           Total Hg removed by clay
                2.0
3.0
4.0
5.0


PH
6.0
 7

7.0
8.0
Figure 42.  Removal of various forms of Hg from DuPage landfill-leachate

            solutions by kaolinite plotted as a function of pH at 25° C.
                                   117

-------
between the digested and undigested blanks. The amount of organic-Hg in solu-
tion with 1 g kaolinite present was always less than the organic-Hg deter-
mined from the blanks. This difference was then taken as the amount of or-
ganic-Hg removed due to the presence of the clay. Figure 42 illustrates the
relative amounts of organic-Hg (curve C) and inorganic-Hg (curve D) removed
by 1 g kaolinite plotted as a function of pH. These results can be compared
to the total Hg removed by 1 g kaolinite in curve B and with the total Hg re-
moved from solution by all mechanisms in curve A.

     The results given in Figure 42 illustrate that the Hg removal is pH de-
pendent, with increasing amounts of removal as the pH is raised within the
range 2 to 8.

     From the data, it can be stated that approximately two thirds of the
total Hg removed by 1 g kaolinite was organic Hg (curve C) , while about one
third of the Hg removed by the clay was inorganic Hg (curve D). The total Hg
removed from solution (curve A) is apparently the result of several mechan-
isms operating simultaneously, i.e. adsorption by clay, volatilization, and
precipitation. The amount of Hg removed as a result of the presence of clay
was evaluated, as previously described, and accounted for approximately 20 to
30% of the total Hg removed from solution. It therefore appears that precipi-
tation and volatilization account for the largest amounts of Hg removed in
this study.

     At this time, no data is available as to the relative amounts of Hg lost
by volatilization from the leachate as opposed to precipitation. In any
event, large amounts of Hg are lost from leachate solutions both in the
presence and absence of clay.

     In an unreplicated experiment, the removal of Hg from solution by mont-
morillonite was found to be approximately 5 times greater than the removal of
Hg from solution by kaolinite. Since this experiment has not been repeated,
the results are tentative. However, this ratio is similar to the ratio of the
Cation Exchange Capacities (CEC) of these two clay minerals. Since this re-
sult is consistent with those for the previous heavy metal cations studied,
it was concluded that the difference in adsorption between montmorillonite
and kaolinite was due to the cation exchange of the various ionic forms of Hg
which may be present in leachate solutions of variable pH.

CONCLUSIONS

     The results of this study indicate that removal of Hg from leachate
solutions is enhanced by clay minerals and is pH dependent. Substantial
amounts of Hg were removed from leachate by the clays, and these amounts were
concluded to be in proportion to the respective CEC values of the clays. Of
the amount of Hg removed by the clays, it was concluded that about two thirds
was organic Hg and one third inorganic Hg.

     Large amounts of Hg were removed from solution, both in the presence and
absence of clay.  These results lead to the conclusion that adsorption by
clays was not the major mechanism responsible for removal of Hg. Rather, pre-
cipitation and/or volatilization accounted for between 70 and 80% of the Hg

                                     118

-------
removed from the leachate solutions and were concluded to be the predominant
attenuation mechanisms in these experiments.

     The results of this study suggest that maximum removal of Hg from leach-
ate would be achieved by disposal  under alkaline  conditions. It is also sug-
gested that montmorillonitic  clays or soils will  remove substantially more Hg
than kaolinitic  clays or soils.
                                       119

-------
                                 SECTION 12

                        SUMMARY OF ADSORPTION STUDIES
     The results of these studies with heavy metals indicate that pH has a
pronounced effect on the amounts adsorbed from landfill leachates by clay
minerals. It was concluded that the principal adsorption mechanism is cation
and anion exchange, a mechanism that led to the strong pH dependency of the
adsorption process. In addition, species distribution diagrams led to the
conclusion that it is the monovalent species of each element that is
principally adsorbed by anion exchange.

     A comparison of the relative amounts of heavy metals removed at pH 5.0
from 100 ppm equilibrium concentration solutions of the metals studied, both
cationic and anionic, is presented in Table 14. The table indicates that the
cationic heavy metals are generally adsorbed to a greater degree than the
anionic forms. However, this ranking is somewhat pH-dependent, because the
greatest anion adsorption occurs in acid solutions and the greatest cation
adsorption in alkaline solutions. The ranking therefore changes somewhat at
different pH values.

     A significant point shown in Table 14 is the importance of the valence
state of an element to the amount of that element removed from solution by
clay minerals. Cr(III) species are removed to a much greater extent than
Cr(VI) species. The clay minerals removed 30 to 300 times more Cr(III) from
solution than Cr(VI). The table also shows more extensive removal of As(V)
than of As(III). These results indicate that safer disposal of certain ele-
ments may be achieved if, prior to deposition at the landfill or disposal
site, the element is converted to the form that would be most strongly at-
tenuated.

     The information derived from the studies of the various elements indi-
cates that the amounts of heavy metals removed from leachate by clay minerals
depends to a large degree on the element and the form of the element in-
volved, the pH of the leachate, the adsorption capacity of the particular
clay mineral in the liner, and the ionic strength of the leachate.

     The adsorption capacity of the clay minerals and the reversible nature
of exchange-adsorption reactions have important environmental consequences.
For example, if industrial wastes containing heavy metals are placed in a
landfill, changes in the ionic composition or pH of the leachate can occur.
A change in pH may release large amounts of potentially toxic heavy metals
into the aqueous phase, especially in places where precipitates may have ac-
cumulated. Other ions in the waste compete with the heavy metals and may


                                     120

-------
TABLE 14.  REMOVAL OF HEAVY METALS  FROM SOLUTIONS  BY KAOLINITE AT pH 5.0
                                        Amount  removed  at
                                   equilibrium  concentration
                                          of  100  ppm
                                          (pmoles/g)
Element
Cr(III)
Cu
Pb
As(V)
Zn
As (III)
Cd
Se
Cr(VI)
Pure
solutions
769*
55-1
42.3
t
33.6
*
26.7
t
0.62
Du Page
leachate
576*
15.7
12.1
5.3
3.8
2.0
1.9
1.9
1.9
             *Precipitation contributes  to  removal  at pH 5.0.

             tRemovals  from 40 ppm solutions were approxi-
             mately the same as  removals from leachate.

             ^Removals  from 40 ppm solutions were 30 percent
             greater than removals from leachate.
                                   121

-------
exchange with them, thus allowing metal ions to come into solution. These
multiple interactions must be considered when a disposal site is designed
and when the environmental impact of adding heavy-metal wastes to municipal
landfills is assessed.
                                    122

-------
                                  SECTION  13

                      APPLICATION OF THE  RESULTS TO THE
                         PROBLEM  OF LANDFILL DESIGN
LANDFILL DESIGN
     Regarding pollutant control,  the design of a  landfill should take into
account three factors: the hydrologic system governing direction of pollutant
travel, the geochemistry of  the water-sediment system, and the release rate
of unattenuated pollutants to  surface or ground waters.  The first item has
been the subject of a number of papers and will not be addressed here.

     Current landfill design and,engineering practice are to construct clay
liners, either natural or artificial, very thick and containing high clay
percentages.  The motive is  to create relatively impermeable liners that will
contain the leachate and therefore protect the groundwater resources.  This
approach can create difficulties in humid climates where infiltration exceeds
the capacity of the liner to dissipate the leachate.  This causes what is re-
ferred to as the "bathtub" effect  wherein the relatively impermeable clay
liner acts as a bathtub which  fills up with leachate and then overflows.  If
the leachate is not collected  and  treated, the overflow manifests itself in
the form of leachate springs on the surface and results in surface water
pollution and surface environmental degradation instead of ground water pollu-
tion.  Neither form of pollution is desirable; both might be prevented by
using proper design features in the construction of the sanitary landfill.
If one does not choose to collect  and treat the leachate and wishes to avoid
the "bathtub," the results of  this research suggest an alternative landfill
design.

     The chemical attenuation  study reported in Section 5 has indicated that
most of the toxic constituents found in municipal leachates are moderately-
to-highly attenuated by passage through laboratory columns containing rela-
tively low percentages of clay minerals.  If the assumption is granted that
the "bathtub" effect is an undesirable feature of clay liners -, then it
follows that it is desirable to determine the point of "optimal" attenuation,
i.e., the percentage of clay in a  liner material which gives the maximum
attenuation balanced with a  maximized hydraulic conductivity.  Figure 43
represents a dual scaled graph with the initial hydraulic conductivity of
the montmorillonite columns  as given in Table 3 plotted as a function of per-
centage of montmorillonite.  The opposite scale is the attenuation number for
the chemical constituent of  interest also plotted as a function of the per-
centage of montmorilldhite.  The attenuation numbers (Table 5), as reported in
Section 5, are the percentage  of removal of the element from the leachate upon
passage through 10 pore volumes of the clay-sand mixture.  The attenuation

                                    123

-------
                           Hydraulic conductivity
024
8  12  16  20  24  28  32  36  40  44  48  52  56  60  64  68  72
                   Montmorillonite (%)
Figure  43.   Effect of clay content on hydraulic conductivity and
             attenuation of Pb, NH^, and Cl for a 40-cm  thick liner.
             The  bulk densities vary; densities for each clay content
             are  shown in Table 3.
                                124

-------
number scale is given  as  0  at the point of minimum hydraulic conductivity and
100 at the point of maximum hydraulic conductivity.

     For the heavy metals (Pb is used here as an example),  even small amounts
of clay gave almost total removal.  The heavy metals, even though toxic,
represent a minimal pollution hazard in municipal leachates because they are
attenuated very strongly  and are usually present at low levels in leachates.
Therefore they can usually be ignored as far as determining the optimal clay
liner for a given leachate.  At the other extreme are the relatively non-
interacting constituents  represented by Cl.  Chloride as shown here is rela-
tively unattenuated by even large amounts of clay.  Figure 43 suggests that in
order to prevent  chloride migration, relatively impermeable clay liners would
be necessary.  Due  to the non-toxic nature of the chloride ion, it also ranks
low along with the heavy  metals in the pollution hazard index (Table 6).   In
view of the problems  associated with the "bathtub" effect,  it seems unwise
to design clay liners to  optimize chloride attenuation.  Rather, it seems
prudent to design clay liners for optimum attenuation of the most hazardous
constituents  found  in a particular leachate.  In the case of the DuPage leach-
ate used in  this  study, the pollution hazard index ranks NHj as 30 times more
of a pollution hazard than any other constituent found in this leachate.   It
therefore  seems  reasonable to design a clay liner for DuPage leachate that
gives optimal attenuation of NH^ and all the other constituents should also
be attenuated to  relatively safe levels for minimal pollution of the ground-
waters  adjacent  to the landfill site.

     For  the case of NH,  (shown in Figure 43) if one extrapolates the curve,
 it  is apparent that 18-20% montmorillonite would give nearly total removal
 from  the  leachate.   If the 10% liner is doubled in thickness from the 40 cm
used  in this study to 80 cm, it will contain enough montmorillonite to give
nearly  total removal of the Nflt in 10 pore volumes of leachate and still
 retain  the  relatively high hydraulic conductivity of 6 x 10~^ cm/sec.  This
 illustrates  that,  for a given desired hydraulic conductivity, the removal
 capacity  of  the  liner may be adjusted by changing its thickness (at the same
 clay  content)  without greatly affecting transmission of water through it.

LINER THICKNESS

     The  thickness of various mixtures of sand and clay, as represented by
 the total  cation exchange capacity (CEC), to achieve total attenuation of
 selected  relatively mobile ions was calculated (Table 15).   The removal ef-
 ficiency will differ in leachates, depending upon relative ion strength.   The
 efficiencies  used in Table 15 are based on the DuPage leachate used in this
 study.  The  concentrations are given in parentheses under the average concen-
 tration value.  Increasing cation exchange capacity generally reflects in-
 creasing  clay contents.  Thus a thicker liner with greater hydraulic con-
ductivity  and lower CEC may be the optimal liner for attenuation.

     Determining  the release rate of nonattenuated or poorly attenuated con-
taminants  from the  clay liner (natural or man-made) to surface waters or
ground-water  aquifers is  necessary for good landfill design.  In designing a
landfill, a  decision must be made as to which ions should be totally attenuated


                                       125

-------
      TABLE 15.  ESTIMATED LANDFILL LINER THICKNESS NECESSARY FOR ATTENUATION OF SOME LEACHATE
                 CONSTITUENTS PER CUBIC METER OF REFUSE DURING A 20 YEAR FILL LIFE*
NJ
ON
Constituent Initial**
Concentrat ion



NHi,

Na

K

Mg


Ave.
(ppm)
379
(830)
755
(740)
763
(530)
1,609
(240)

Max.
(ppm)
1,106

7,700

3,770

15,600


Ave.
(cm)
32

118

51

226

10
Max.
(cm)
92

1,208

252

2,191

CEC

Ave.
(cm)
16

59

26

113


20
Max.
(cm)
46

604

126

1,096


30
Ave.
(cm)
11

39

17

75



Max.
(cm)
31

403

84

730

       *Assumption:  Bulk density = 1.8 g/cc; 100 liters of leachate generated per m3 of refuse/yr.;
        initial concentration decreases linearly to zero at 20 yr.; removal efficiencies for each
        constituent were estimated using the average values given by Griffin, et al. (1975), NHi*  =  37.1%,
        Na  =  15.4%, K  =  38.2%, Mg  =  29.3%.

      **Concentrations taken from the twenty leachate analyses reported by EPA (1974) ; those in
        parenthesis, (830), are the values of Old DuPage landfill leachate used in this study.

-------
and which should be eventually released to the environment.  The  chloride ion,
which moves essentially  unattenuated,  is the most obvious  example of the latter
type.

CONCLUSIONS

     The results of this study suggest an alternative to landfill design.  The
use of hydraulic conductivity information and the pollution  hazard  rating for
a given leachate or waste stream can allow a different approach than the
prevalent "containment"  type liner systems.  These data suggest that overall
pollution would be lessened by designing landfill liners for higher permea-
bility and by selectively attenuating the most toxic pollutants from
leachate and  allowing the groundwaters to dilute the nontoxic  components
which can be  tolerated at much higher concentrations than  the  toxic without
deleterious effects.   Thus landfill stabilization and use  for  other productive
purposes could be  achieved at much faster rates than can presently  be
achieved by containment type liners.
                                       127

-------
                                 SECTION 14

                                 REFERENCES
Anderson, J. R., and J. N. Dornbush, Influence of Sanitary Landfill on Ground
     Water Quality: J. of Am. Water Works Assoc. v. 59, no. 4, p. 457-470
     (April 1967).

Apgar, M. A., and D. Langmuir, Ground Water v. 9, p. 6  (1971).

Apgar, M. A., and W. B. Satterthwaite, Jr., Ground Water Contamination
     Associated with the Llangollen Landfill, New Castle County, Delaware.
     Proceedings of the Research Symposium, "Gas and Leachate from Landfills:
     Formation, Collection, and Treatment." New Brunswick, New Jersey, March
     15,  16, 1975. U.  S. EPA Nat. Environ. Research Center, Cincinnati, Ohio.

1968 Book of ASTM Standard, Procedure D 1756, "Carbon Dioxide in Coal."
     p. 303-306.

Baetjer,  A. M., in Chromium, J. U. Marvin, {Ed.}, Chapter 4, Reinhold, New
     York, N. Y. (1956).

Beevers,  J. R., Australian Bureau of Mineral Resources, Geology, and Geo-
     physics, Report No. 106 (1966).

Bernas, B., A New Method for Decomposition and Comprehensive Analysis of
     Silicates  by Atomic Absorption Spectrometry: Anal. Chim. Acta  v. 40,
     no.  11, p. 1682-1686 (1968).

Bittel, J. E.,  and R.  J. Miller, Lead, Cadmium, and Calcium Selectivity Co-
     efficients on a Montmorillonite, Illite, and Kaolinite: J. Environ.
     Qual. v. 3, p. 250-253 (1974).

Black, R. J., A. J. Muhich, A.  J. Klee, H. L. Hickman, Jr., and R. D.
     Vaughan, The National Solid Wastes Survey; an interim report. U. S.
     Department of Health, Education, and Welfare. Cincinnati, Ohio, 53p
     (1968).

Boyd, G. E., J.  Schubert, and A. W. Adamson, The Exchange Adsorption of Ions
     from Aqueous Solutions by Organic Zeolites. 1. Ion Exchange Equilibria:
     J. Amer. Chem. Soc. v. 69, p. 2818-2829 (1947).

Broadbent, F. E., and J. B. Ott, Soil Science v. 83, p. 419-427  (1957).
                                     128

-------
Brunauer, S. , P. H. Emmett,  and  E.  Teller,  Adsorption of  Gases  in Multi-
     molecular Layers: J. Am.  Chem.  Soc.  v.  60,  p.  309-319  (1938).

Butler, J. N., Ionic Equilibrium -  A Mathematical Approach,  p.  360-363
     Addison-Wesley, Reading,  Mass.  (1964).                            '

Cartwright, Keros, and M. R. McComas, Ground  Water v.  6.  no. 5  t>. 23-30
     (1968).

Davids, H. W., and M. Lieber,  Water and  Sewage  Works  v.  98, no. 12  p. 528-
     534  (1951).

Davies, R.,  Special Report on  Solid Wastes: Leachate Control. Canadian Water
     and Pollution Control,  v. Ill, no.  12, p.  15-31 (1973).

Derryberry,  0. M. , "Investigation of Mercury  Contamination in the Tennessee
     Valley  Region," in  Environmental Mercury Contamination. R. Hartung and
     B. D. Dinman {Eds.}, Ann  Arbor Science Publishers, Inc., Ann Arbor,
     Mich. p. 76-79  (1972).

D'ltri, F. M., The Environmental Mercury Problem.  Chemical Rubber Co. Press,
     Cleveland,  Ohio, p.  17-33,  p.  24, p. 64  (1972a).

D'ltri, F. M. , "Sources  of Mercury  in the Environment," in Environmental
     Mercury Contamination,  R. Hartung and  B. D.  Dinman {Eds.}, Ann Arbor
     Science Publishers,  Inc., Ann  Arbor, Mich. p. 5-25 (1972b).

Early, A. B., Coal Conversion  Technology and  Solid Waste  Disposal: Time to
     Take Stock. Presented at  the U.  S.  EPA symposium on  Fuel Conversion
     Technology.  St. Louis,  Missouri, May 13-16  (1974).

Elzy, Eugene, F. T. Lindstrom, L. Boersma, H. R.  Sweet, and P. H. Wicks, in
     Disposal of Environmentally Hazardous Wastes. Task Force Report,
     Environmental Health Sciences  Center, Oregon State University, Chapter
     4  (1974).

Enrich, G. H. , Compost Science v. 13, p.  3  (1972).

Everst, D. A., and W. J. Popiel, J.  Chem. Soc.  2433  (1957).

Farquhar, G.  J., and F. A. Rovers.   In: Proceedings  of the Research Symposium
     Gas and Leachate from Landfills:  Formation, Collection, and Treatment.
     New Brunswick, NJ, March  1975.   U.S. Environmental Protection Agency,
     National Environmental  Research  Center,  Cincinnati, OH. (EPA-600/9-76-004)

French, W. J. , and S. J. Adams,  Polypropylene Bottles  in  the Decomposition
     of Silicate Rocks:  Anal.  Chim.  Acta v.  66,  p.  324-328  (1973).

Fuller, W. ,  "Attenuation of  Pollutants in Landfill Leachate by Soils," in
     Proceedings of Research Symposium,  Gas and Leachate  from Landfills:
     Formation,  Collection,  and  Treatment. New  Brunswick, N. J. , March 15,
     16   1975. U. S. EPA, National  Environmental  Research Center, Cincinnati,

                                     129

-------
     Ohio. (EPA-600/9-76-004)

Gadde, R. R., and H. A. Laitinen, Environmental Letters v. 5, p. 91-102
     (1973a).

Gadde, R. R. , and H. A. Laitinen, Environmental Letters v- 5, p. 223-235
     (1973b).

Garland, G. A., and D. C. Mosher, Leachate Effects from Improper Land Dis-
     posal: Waste Age v. 6, no. 3, p. 42-48 (1975).

Giordano, P. M., J. J. Mortvedt, and D. A. Mays, Effect of Municipal Wastes
     on Crop Yields and Uptake of Heavy Metals: J. Environ. Qual. v. 4,
     p. 394-399 (1975).

Goldwater, L. J., Mercury in the Environment: Scientific American v. 224,
     no. 5, p. 15-21  (May 1971).

Gordon, M., Environ.  Sci. Technology v. 9, p. 415-416 (1975).

Gowler, A., Paper presented before Internat. Symp. Geochem. Cosmochem.,
     Tokyo. Sept. 7-9, 1970. Vol. II, The Clarke Company, Washington, D. C.
     (1973).

Greeson, P. E., Biological Factors in the Chemistry of Mercury. U. S. Geolog-
     ical Survey Professional Paper 713, Mercury in the Environment p. 32-34
     (1970).

Griffin, R. A., and N. F. Shimp, WateReuse v. 2, p. 801-811 (1975).

Griffin, R. A., and R. G. Burau, Soil Sci. Soc. Amer. Proc. v. 38, p. 892-897
     (1974).

Grim, R. E., and F. L. Cuthbert, Illinois State Geological Survey Report of
     Investigations 102. Urbana, Illinois (1945).

Grim, R. E., Clay Mineralogy. McGraw-Hill Inc., New York, p. 213, 221 (1968).

Hanks, T. G., U. S. Department of Health, Education, and Welfare, Public
     Health Service Publication 999-UIH-6, Cincinnati, Ohio 179p (1967).

Heitfeld, K. H., and U. Schoettler, Versickert Wohin? (Seeping, where?):
     Umwelt v. 3, p. 57-58 (1973).

Henderson, D. A., Australasian Ann. Med. v. 4, p.  164 (1955).

Henning, R. J., B. L. Manner, R. G. Corbett, and J. L. Jackson, Analysis of
     Groundwater Samples from Boreholes Surrounding Mentor Landfill, Mentor,
     Ohio: G.S.A.  Abstracts with Programs, v. 7, no. 7 (1975).

Hofmann, U., A.  Weiss, G. Koch, A. Mehler, and A.  Scholz, in Clays and Clay
                                     130

-------
     Minerals, A. Swineford,  {Ed.},  p.  273-287, National Academy of
     Sciences-National Research  Council, Pub.  456, Washington D. C.  (1956).

Hughes, G. M. , R. A. Landon,  and R.  N.  Farvolden, Hydrogeology of Solid Waste
     Disposal Sites in Northeastern  Illinois.  U. S. Environmental Protection
     Agency, Solid Waste Management  Series, Report SW-12d,  154p (1971).

Jackson, M. L.,  Soil Chemical Analysis. Prentice-Hall, Inc., Englewood
     Cliffs, N.  J., p.  193-197  (1958).

John, M. K. , Cadmium Adsorption  Maxima  of  Soils as Measured by the Langmuir
     Isotherm: Can. J.  Soil  Sci.  v.  52, p. 343-350 (1972).

Klein, D.,  "Some Estimates of Natural Levels of Mercury in the Environment"
     in Environmental Mercury Contamination. R. Hartung and B. D. Dinman,
     {Eds.}, Ann Arbor  Science Publishers, Inc., Ann Arbor, Mich. p. 25-29
     (1972).

Langmuir,  I., J. Amer.  Chem.  Soc.  v.  40, p. 1361-1403  (1918).

Laswick,  J. A.,  and R.  A. Plane, Amer.  Chem. Soc. v. 81, p. 3564-3567 (1959).

Lehman, J.  P., Guidance for  Conducting  Chain of Evidence Research. Hazardous
     Waste Management Division,  OSWMP,  U.  S. Environmental Protection Agency,
     Washington  D. C.  (June  1973).

Leithe, W. , The  Analysis of  Organic  Pollutants in Water and Waste Water, Ann
     Arbor, Mich. p. 58-59  (1973).

Manger, G.  E. , U. S. Geological  Survey  Bulletin 1144-E (1963).

Mesmer, R.  E., and C. F.  Baes, The Hydrolysis  of Cations: A Critical Review
     of Hydrolytic Species and their Stability Constants in Aqueous Solution.
     ORNL-NDF-EATC-3: Part 3, 141p (1974).

Murray, C.  U. , in Chromium,  J. U.  Marvin,  {Ed.}, Chapter 6, Reinhold, New
     York,  N. Y. (1956).

Nuclear Science  Series  NAS-NS 3002,  Nat. Acad. Science-Nat. Res. Council,
     AEC,  "Radiochemistry of Arsenic."  (1965).

Nelsen, F.  M> and F. T.  Eggertsen, Determination of Surface Area: Adsorption
     Measurements by a  Continuous Flow  Method, Anal. Chem. v. 30, p. 1387-
     1390 (1958).

Olin, A.,  Acta Chem. Scand.  v.  14, p. 814  (I960).

Parham, W.  E. , Clay Mineral  Facies of Certain  Pennsylvanian Underclays.
     Ph.D.  Thesis, Univ.  of  Illinois (1962).

Parker, C.  R., Water Analysis by Atomic Absorption, Springvale, Varian Tech-
     tron Pty- Ltd. p.  9 (1972).

                                     131

-------
Peech, M., L. T. Alexander, L. A. Dean, and J. F. Reed, Methods of Soil
     Analysis for Soil Fertility Investigations. USDA Circular No. 757
     (1947).

Peterson, M. L., Jour. Amer. Public Health Assoc. v. 64, p. 912-914  (1974).

Rollinson, C. L., in Chemistry of the Coordination Compounds, J. C. Bailar,
     Jr., and D. H. Busch {Eds.77 Chapter 13, Reinhold, New York, N. Y.
     (1956).

Rosenkranz, H.  S., and T. J. Wibdkowski, Mutagenicity of Ethylenechloro-
     hydrine. A Degradation Product Present in Foodstuffs Exposed to Ethylene
     Oxide: J.  Agri. Food Chm. v. 22, p. 407-409 (1974).

Ruch, R. R. , E. J. Kennedy, ai*d N. F. Shimp, Illinois State Geological Survey
     Environmental Geology Note 37 (1970).

Shiga, Shoji, Chem. Abstr. v. 82, Abstract no. 102850n  (1975).

Shuman, L. M.,  The Effect of Soil Properties on Zinc Adsorption by Soils:
     Soil Sci.  Soc. Am. Proc. v. 39, p. 454-458 (1975).

Smith, R. G. , in Metallic Contaminants and Human Health, D. H. K. Lee {Ed.},
     Chapter 6, Academic Press, New York, N. Y. (1972).

Stumm, W., and  J. J. Morgan, J. Amer. Water Works Assoc. v. 54, p. 971
     (1962).

Syers, J. K., M. G. Browman, G. W. Smillie, and R. B. Corey, Phosphate Sorp-
     tion by Soils Evaluated by the Langmuir Adsorption Equation: Soil Sci.
     Am. Proc.  v. 37, p. 358-363 (1973).

Todd, D. L., Ground-Water Hydrology, John Wiley and Sons, New York, p. 53
     (1959).

Turney, W. G. ,  "The Mercury Pollution Problem in Michigan" in Environmental
     Mercury Contamination, R. Hartung and B. D. Dinman {Eds.}, p. 29-31
     (1972).

Urioste, J. A., M. S. Thesis, Dept. of Civil Eng., Univ. Wisconsin, Madison
     (1971).

U. S. Environmental Protection Agency. Report to Congress on Hazardous Waste
     Disposal. U. S. Environmental Protection Agency publication SW-115,
     Washington, D. C. 20460, June 30, 1973.

U. S. Environmental Protection Agency; Water Quality Criteria 1972; Environ-
     mental Protection Agency R3-73-033, March 1973.

U. S. Environmental Protection Agency. Summary Report: Gas and Leachate from
     Land Disposal of Municipal Solid Waste. U. S. Environmental Protection
     Agency, National Environmental Research Center, Cincinnati, Ohio 45268

                                    132

-------
     (1974).  Draft report.

U. S. Environmental Protection Agency, Methods  for Chemical Analysis of Water
     and Wastes  1971. Water  Quality  Office, Analytical Quality Control
     Laboratory, Cincinnati,  Ohio  (1971).

U. S. Environmental Protection Agency, Technology Transfer, p. 4-9, October
     1975.

van Olphen, H.,  An Introduction  to Clay  Colloid Chemistry, p. 251-279,
     Interscience, New  York,  N.  Y. (1963).

Walker, W. H.,  Illinois Ground Water Pollution: J. Amer. Water Works Assoc.
     v. 61, p.  31-40  (1969).

Weiss, H. V-, M. Koide, and  E.. D.  Goldberg, Science v. 174, p. 692 (1971).

Welsch, W. F.,  Sewage and Industrial Wastes   v. 27, p. 9, 1065-1069 (1955).

White, W. A.,  Illinois  State Geological  Survey  Circular 282, Urbana, Illinois
      (1959).

Wirenius, J.  D. , and  S. L. Sloan,  Determining the Life of a Landfill Site:
     Public Works  v.  104, no. 9, p.  118-119  (1973).

Zeizel, A.  J. ,  W.  C.  Walton, R.  T. Sasman,  and  T. A.  Pickett, Ground-Water
     Resources of  DuPage County, Illinois:  Illinois State Geological Survey
      and  Illinois State Water Survey Cooperative Ground-Water Report 2,
     p.  103 (1962).
                                      133

-------
                                APPENDIX A

                           LIST OF PUBLICATIONS
Griffin, R. A., and N. F- Shimp. 1975. Leachate migration through selected
     clays. In "Gas and Leachate from Landfills: Formation, Collection, and
     Treatment," E. J. Benetelli and John Cirello, Eds., Ecological Research
     Series, EPA-600-9-76-004, U. S. Environmental Protection Agency,
     Cincinnati, Ohio 45268. p. 92-95.

Griffin, R. A., and N. F. Shimp. 1975. Interaction of clay minerals and
     pollutants in municipal leachate: WateReuse v. 2, p. 801-811.

Griffin, R. A., R. R. Frost, and N. F. Shimp. 1976. "Effect of pH on removal
     of heavy metals from leachates by clay minerals," in Residual Management
     Land Disposal, W. H. Fuller {Ed.}, Ecological Research Series,
     EPA-600/9-76-015, U. S. Environmental Protection Agency, Cincinnati,
     Ohio 45268. p. 259-268.

Griffin, R. A. , and N. F. Shimp. 1976. Effect of pH on exchange-adsorption or
     precipitation of lead from landfill leachates by clay minerals: Environ.
     Sci. and Technology v. 10, p. 1256-1261.

Griffin, R. A., Keros Cartwright, N. F. Shimp, J. D. Steele, R. R. Ruch,
     W. A. White, G. M. Hughes, and R. H. Gilkeson. 1976. Attenuation of
     pollutants in municipal landfill leachates by clay minerals, Part 1 —
     Column leaching and field verification: Illinois State Geological Sur-
     vey, Environmental Geology Notes 78, Urbana, Illinois, 34 p.

Griffin, R. A., R. R. Frost, A. K. Au, G. D. Robinson, and N. F. Shimp. 1977.
     Attenuation of pollutants in municipal landfill-leachates by clay
     minerals, Part 2 — Heavy metal adsorption studies: Illinois State
     Geological Survey, Environmental Geology Notes 79, Urbana, Illinois,
     47 p.

Frost, R. R., and R. A. Griffin. 1977. Effect of pH on adsorption of arsenic
     and selenium from landfill leachate by clay minerals: Soil Sci. Soc. Am.
     Jour., v. 41, p. 53-57.

Frost, R. R., and R. A. Griffin. 1977. Effect of pH on removal of copper,
     zinc, and cadmium from landfill leachate by clay minerals: Jour.
     Environ. Sci. and Health, Part A, v. 12, p. 139-156.

Cartwright, Keros, R. A.  Griffin, and R. H. Gilkeson. 1977. Migration of


                                    134

-------
     landfill leachate through unconsolidated porous media: Advances in
     Groundwater Hydrology, v. 12:(in press).

Griffin, R. A., A. K. Au, and R. R. Frost.  1977. Effect of pH on adsorption
     of chromium from landfill-leachate by  clay minerals: Jour. Env. Sci. and
     Health, Part A, v.  12, p. 431-449.

Cartwright, Keros, R. A.  Griffin,  and R. H.  Gilkeson.  1977. Migration of
     landfill leachate through glacial tills: Ground Water, v. 15,
     p. 294-305.

Griffin, R. A., and  A. K. Au.  1977. Lead adsorption by montmorillonite using
     a  competitive-Langmuir equation: Soil  Sci. Soc. Am. Jour., v. 41,
     p. 880-882.

Griffin, R. A., N. F. Shimp, J. D. Steele,  R. R. Ruch, W. A. White, and
     G. M. Hughes. 1976.  Attenuation of pollutants in  municipal landfill-
     leachate by passage through clay: Environ. Sci. and Technology v. 10,
     p. 1262-1268.
                                      135

-------
                                APPENDIX B
          PROCEDURES USED IN CHEMICAL AND PHYSICAL CHARACTERIZATION
                     OF CLAY MINERALS AND LIQUID SAMPLES
Clay Mineral Characterization
                (W. A. White)

     The kaolinite used in this study was collected from materials of Penn-
sylvanian age in Pike County, Illinois. The site description and location are
given by White (1959) , sample 996N. The montmorillonite used in the study was
southern bentonite from the American Colloid Company. The illite was from the
Minerva Company Mine^of the Allied Chemical Company.

     The clays were brought to the laboratory, where they were crushed,
ground, and purified by sedimentation techniques to obtain the <2 ym particle
fraction that contained essentially pure clay minerals. The purified clay was
then floculated with  1 M CaCla, centrifuged, and spray dried. The clay
minerals present in the <2 ym fraction were identified by X-ray diffraction
techniques (Parham, 1962), and the results are given in Figure 44. The
positions of the diffraction peaks identify which clay minerals are present
in the sample. In addition, the areas under the peaks of the diffraction
patterns give a crude estimate of the relative amounts of each clay mineral
in the sample and were used to prepare the following summary:

     (a) The montmorillonite sample was found to be almost completely
         monomineralic, containing approximately 95% montmorillonite
         and 5% mixed-layer materials close to montmorillonite in
         composition.

     (b) The illite sample was found to be monomineralic, with mica
         minerals predominant. The sample contained approximately 70%
         illite and 30% mixed-layer materials close to illite composition.

     (c) The kaolinite sample could be characterized as a moderately
         pure kaolinite, comparable to the Georgia hard kaolins. The
         sample contained approximately 87% kaolinite, 8% illite, 5%
         mixed-structure material, and a trace of quartz.

Atomic Absorption Methods for Cd, Fe, Mn, Pb, Si, Zn, Na, K, Ca, and Mg
                                            (John Steele and David Heck)


                                    136

-------
              Illite,     ;
               ethylene- ;
               glycol
                                              Montmorillonite,
                                               .ethylene-glycol
                                             Kciolinite M

        36
25             15
   Degrees 29
Figure 44.  X-ray diffraction patterns  of the clay minerals used in
            leachate pollutant attenuation studies.
                              137

-------
     Atomic absorption methods were used for the following determinations:
filtered and acidified leachate (dissolved metals) — Cd, Fe, Mn, Pb, Si,
and Zn; membrane filters (suspended metals) — Cd, Fe, Mn, Pb, and Zn; am-
monium acetate extracts (exchangeable cations) — Cd, Fe, Mn, Pb, Zn, Na, K,
Ca, and Mg; and column sections and original clays (total cations) — Cd, Mn,
Na, Pb, and Zn.

     Atomic absorption measurements were made using a Perkin-Elmer Model 306
Atomic Absorption Spectrophotometer. In the case of the dissolved, suspended,
and exchangeable metal determinations, measurements were recorded directly
from the electronic digital readout in the auto-concentration mode. For the
total cation determinations, absorbance measurements were recorded on a
Perkin-Elmer Model 056 Recorder. The following Perkin-Elmer burner heads were
used: a 4-inch-long slot, flat-head burner using an air-acetylene flame for
all of the dissolved and suspended cation determinations except Si; a 2-inch-
long slot nitrous oxide burner head using a nitrous oxide-acetylene flame for
Si determinations; and a three slot high solids burner head with an air-
acetylene flame for all of the exchangeable and total cation determinations.
Standard single element hollow cathode lamps were used. Corrections for back-
ground absorption were made simultaneously for total cation determinations
using a Perkin-Elmer Deuterium Arc Background Corrector.

     All reagents used are ACS certified reagent grade chemicals, and stan-
dard stock solutions were prepared from high purity metals or compounds.
Calibration standards prepared from diluted stock solutions contained the
following matrices: dissolved metal standards — 1% v/v distilled HN03; sus-
pended metal standards — 4% v/v distilled HC1; exchangeable cation stan-
dards — IN ammonium acetate and 7.5% v/v distilled HNOs; and total cation
standards -- 1.2% v/v 48% HF, 1.5% v/v aqua regia (1:3:1 HN03-HC1-H20) , and
0.54% w/v HsBOa. For total Na determinations, the standards contained, in
addition to the above mentioned matrix for total cations, 1000 ppm Cs as
CsCl. For determination of Ca and Mg in the leachate fraction, the standards
were prepared to contain 1% w/v La20a and 2.5% v/v HC1. For determination of
Na and K, the standards were prepared to contain 1000 ppm Cs as CsCl.

Sample preparation

     The filtered-and-acidified leachate and the ammonium acetate extract
samples required no sample preparation and were analyzed directly except that
for the analysis of Ca and Mg, 1% L^Os w/v and 1% v/v HC1 were added. Also
for the determination of K and Na, the final sample contained 1000 ppm Cs as
CsCl.

Column section dissolution

     The methods described below are modifications of those of Bernas (1968)
and French and Adams (1973). Approximately 0.5 g of the clay-sand mixture,
previously dried at 110° C for several hours, is transferred to a 125 ml
linear polyethylene screw cap bottle. The sample is wetted with 1.5 ml aqua
regia, and 1.2 ml of 48% HF is added. The cap is screwed tight, and the
bottle is placed in a steam bath for 2 hours. The sample and acids are mixed
by an occasional swirling of the bottle during this period. The bottle is

                                     138

-------
removed from the steam bath,  and  9  ml  of  a H3B03  solution  (0.06 g/ml) is
added to the sample to complex  the  fluoride.  The  bottle  is  again sealed and
placed on the steam bath  for  0.5  hours, removed,  and  allowed to cool. The
dissolved sample is transferred to  a 100  ml polyethylene volumetric flask,
the bottle is'washed repeatedly with deionized  H20, and  the washings are
added to the flask. The sample  is diluted to volume with deionized H20 and
returned to the polyethylene  bottle for storage.

Membrane filter dissolution

     The membrane filters were  treated according  to methods modified from
Parker (1972). The membrane is  placed  in  a 250  ml Pyrex  beaker, and 3 ml of
distilled HN03 is added.  The  beaker is covered  with a watch glass and heated
on a hot plate where the  acid is  evaporated to  dryness.  The additions of acid
and the heating steps are continued until a light residue is left. The resi-
due is taken up with 2 ml of  distilled 1:1 HC1  and warmed gently. The sample
is then transferred to a  50 ml  Pyrex volumetric flask, and  the beaker is
washed repeatedly with small  portions  of  deionized H20 and  the washings added
to the flask. The sample  is diluted to volume and transferred to a poly-
ethylene bottle for storage.

Atomic Absorption Spectrophotometric Procedures

     The following analytical wavelengths were  used:  228.8  nm (Cd), 248.3 nm
(Fe), 279.5 nm (Mn), 589.0  nm (Na), 283.3 nm (Pb) , 251.6 nm (Si), 213.9 nm
(Zn) , 422 nm (Ca), 285 nm (Mg) , and 766.5 nm (K). Where necessary, samples
are diluted to bring the  metal  concentration within the  linear portion of the
calibration curve. For total  Na determinations, samples  are diluted 1 to 2
with the addition of 1000 ppm Cs  as CsCl.  In the  case of the suspended, dis-
solved, and exchangeable  metal  determinations,  the metal ion concentrations
are determined directly from  the  electronic digital readout in the auto-con-
centration mode, after appropriate  calibration. For the  total cation deter-
minations, the metal ion  concentrations are calculated by solving for con-
centration in a least squares constructed calibration equation of absorbance
vs. concentration. A new  calibration curve equation is calculated for each
set of analyses.

     An estimate of the average relative  standard deviation for the elements
determined by atomic absorption spectrophotometry is  10  percent or better.

Determination of Hg by Flameless  Atomic Absorption
                (R. A. Griffin  and  G.  D.  Robinson)

     The Hg samples obtained  in the investigation of  the removal character-
istics of Hg from leachate  solutions by clay minerals as a  function of pH
were analyzed by flameless  Atomic Absorption Spectroscopy (A.A.) using a
slight modification of the  procedure described  by the U. S. Environmental
Protection Agency (EPA) (U.S.-EPA,  1971).

     Variations of the standard procedure described by U.S.-EPA (1971) were
implemented in order to gain  maximum sensitivity  from the procedure and the
available instrumentation.  The  apparatus  used was a Perkin-Elmer Model 360

                                     139

-------
A. A. equipped with a Perkin-Elmer Flameless Mercury Analysis  System. The
absorption cell supplied by Perkin-Elmer was modified by replacing  the plas-
tic end windows with quartz windows which were glued in place  with  epoxy
cement. This allowed greater light transmission and higher sensitivity and
stability to be achieved for the Hg analyses.

Reagents used:

Sulfuric Acid reagent grade cone.
Nitric Acid reagent grade cone.
Potassium Permanganate 5% w/v 5g KMnCK in 100 ml DI H20
Potassium Persulfate 5% w/v 5g K2S208 in 100 ml DI H20
Hydroxylanine Hydrochloride 5% w/v 5g NH2OH • HCl in 100 ml DI H20
Stannous Chloride - Dissolve 50g SnCla in 100 ml cone. HCl and 400  ml  DI H20
Stock Mercury Solution 1000 ppm - Dissolve l.OSOg HgO in min.  vol.  1:1 HCl -
     dilute to 1 liter with DI H20
Working Mercury Solutions - Make successive dilutions of the stock  Hg solu-
     tion to obtain a standard containing 0.1 ppm. This should be made fresh
     daily and acidified to .15% HNOa. Note: It was found that 10 and 25 ppm
     working Hg standards could be stored in plastic bottles several weeks
     if acidified to .15% HN03 with no apparent loss or adsorption  of Hg.
     This greatly reduced the steps involved in making fresh working stan-
     dards of .1 ppm.
Waste Hg Absorbing Media - . 1 M KMnOi» and 10% H2SOif.

     A calibration curve was obtained following the procedure  outlined in
U. S.-EPA (1971) substituting a hydroxylamine-HCl solution for the  hydroxyla-
mine sulfate solution, and a stannous chloride solution for the stannous
sulfate solution. It was found that this calibration curve had to be repeated
with each set of samples.

     The samples for Hg analysis were placed in 60 ml plastic  bottles and
acidified with cone. HNOs to pH ~2. A 1 ml aliquot was then taken from each
sample and placed in a 300 ml BOD bottle (these were run in triplicates) ;
the volume in the bottle was then brought up to 100 ml with DI H20. The
samples were then digested to oxidize the organo-mercury compounds  which
would not otherwise respond to the flameless A. A. technique (U. S.-EPA,
1971).

     The samples were digested by addition of cone. H2SOij and  2.5 ml of cone.
HN03 to each bottle with mixing after each addition. One ml of 5% w/v KMnOi,
solution was then added to each bottle, swirled, and allowed to stand for at
least 15 minutes. Two ml of K2S2Os were then added and allowed to stand for
at least 30 minutes. Two ml of NH2OH'HC1 were then added to each bottle to
reduce the excess permanganate. Immediately prior to aerating  through the
absorption cell, 2 to 3 ml of SnCl2 solution were added to each bottle. The
samples were read on the absorbance mode and then converted to concentration
mathematically using linear regression analysis (calibration standards were
prepared fresh daily).  The final Hg concentrations were then subtracted from
the initial concentration (0.962 ppm) to obtain the amount of  Hg removed
from solution.
                                     140

-------
Neutron Activation Analysis  Methods for Hg and As
                      (Joyce  Frost and Larry Camp)

     Irradiations are tarried out in the University of Illinois'  Advanced
TRIGA reactor, with  a thermal neutron flux of 1.4 x 1012  neutrons cnT2  sec'1
(500 KW). The samples are contained in a rotary specimen  assembly.  Rotation
at 1 rpm during  irradiation  ensures an equivalent integrated  flux for each
sample.

     Samples and standards are irradiated in heat-sealed  polyethylene snap-
cap vials which  were previously cleaned with deionized water  and  acetone.

     The gamma-ray spectrometer system consists of a 3" x 3"  Nal(Tl) crystal
connected to a 400-channel analyzer.  The analyses are comparative — e.g. the
activity of an element  in the sample is compared to that  of the element in a
known standard to determine  concentration of the element  in the sample.

Determination of Hg  by  Instrumental Neutron Activation Analysis in  "Spiked"
Liquid Samples

     This procedure  was used for the column effluent samples  (filtered  and
acidified leachate samples), solubilized membrane filter  samples,  and am-
monium extract solutions of  the clay column samples that  originated after the
original sterile leachate was spiked with mercury to a resulting  concentra-
tion of 4.0 ppm.

     A two-hour  irradiation  was carried out on 5 ml portions  of the samples
and two acidified mercuric nitrate standards, one containing  5.02ygHg/ml and
the other 0.502ygHg/ml. The  samples were left to decay one week to  decrease
interference from short-lived radioisotopes, especially 15 hour 2 Na and 36
hour 82Br which  gives a high background in the low energy region  of the gamma
ray spectrum. The activity in the samples and standards due to the  0.077 MeV
gamma ray of 65  hours 197Hg  was then counted. At the same time uranium was
determined in those  column effluent samples that unexpectedly had a high con-
centration of uranium,  up to 0.20 ppm; these were the first several sets of
effluents collected  from the columns that contained montmorillonite. The
uranium activity counted was the 106.1 keV gamma ray of 56.4  hour 239Np; the
standard was 5 ml of a  solution of U02(N03)2 containing 1.02  yg U/ml.

Determination of Hg  in  Liquid Samples by Neutron Activation Analysis with
Radiochemical Separation

     This procedure  was used for the original leachate and for the  precise
determination of mercury in  some of the samples having a  low  mercury content.

     The method  follows closely that of Weiss, Koide and  Goldberg (1971).' The
samples  (1 to 3  irradiation  vials, each containing 10 ml  per  sample) and
standard  (10 ml  of 5.02ugHg/ml) were irradiated for 2 hours.  After a day's
decay, each sample was  transferred to a beaker, and 10 mg of  Hg-^ "carrier was
added. The pH of the solution was adjusted to 1 and elemental mercury pre-
cipitated by the addition of stannous chloride solution.  The  mercury was
separated on a Millipore filter and dissolved in aqua regia.  This solution

                                     141

-------
was neutralized with ammonium hydroxide, and thioacetamide solution was
added. The precipitate of mercuric sulphide was separated on a Millipore
filter; dissolved in aqua regia; transferred to a tared, glass, stoppered
vial; and made up with deionized water to 25 ml. The activity of the solution
due to the 0.077 MeV gamma ray of 197Hg was counted. The radiochemical
separation for the standard (250 yl) plus carrier was begun with the mercuric
sulphide precipitation step. Radiochemical yields (85-98%) were determined by
re-irradiation of a 1 ml aliquot of each sample and the standard.

Determination of Hg in Column Sections and Clay

     One gram of sample was accurately weighed and irradiated for two hours
as previously described. The irradiated sample was then mixed with ~1 gram of
Alundum and transferred to a combustion boat which has had Hg carrier added
as Hg(NOs)a. The boat was then placed in a Vicor tube which in turn is placed
in a Lindberg "Heavy Duty" furnace, such that Oa can flow through and the
combustion gasses trapped in a dry ice cold trap. The furnace is operated for
15 minutes at 1000° C and allowed to cool.

     The boat was removed from the cooled tube (at most, 400° C) and the cold
trap section (dry ice removed) placed end down into a polyethene 100 ml cen-
trifuge tube.

     The foresection of the tube was washed with 10 ml of HNOa, and this
solution via transfer pipplets was used to wash down the cold trap section of
the tube. This was repeated twice more, but HaO substituted for the HNOs,
giving 30 ml of total solution. The centrifuge tube was placed in HaO bath
and heated. When hot (90° C or +), the solution was stirred and solid NH^Br
(AR) added, red Bra was evolved, and the solution was stirred. Solid Ag(NOs)a
was added to remove the excess Br (much of the radioactive Br had isotopical-
ly exchanged) and the solution was centrifuged. The liquid was transferred to
a 4 oz. wide mouth jar and labeled.

     Ten ml of warm (90+° C) HN03-H20 (1-3) was added to the ppt in tube and
again centrifuged. This solution was added to the original 30 ml, and the
solution counted for 77.6 KEV y ray of Hg197 (65.5 hour t^) on a Nal detec-
tor. One ml of this solution was removed and put in 1 dram vial, sealed, and
labeled and  irradiated at 15 minutes at 500 KW-  This was compared with a
"carrier vial" and a "yield" factor found for correction.

Determination of Arsenic in Liquid Samples by Neutron Activation Analysis
with Radiochemical Separation

     Ten ml of sample and 10 ml of an arsenic trioxide standard were irradi-
ated for 1 hour, then left to decay for a day. The sample with 30 mg of As"1"1"1"
carrier added, in a distillation flask,^was oxidized by boiling it for 5 min.
with 3 ml of 30% HaOa, to decompose organic arsenic compounds present  (EPA,
1971). Then, after the addition of hydrochloric and hydrobromic acids,
arsenic was distilled from the solution (NAS-NS #3002, 1965). Elemental
arsenic was precipitated from the distillate by the addition of sodium hypo-
phosphite. The precipitate was separated on a weighed filter paper, re-
weighed, and the activity due to the 0.56 MeV gamma ray of 26.5 hour 76As was

                                     142

-------
counted. The standard was  carried through the same radiochemical separation.

Determination of Arsenic in  Clay  Samples by Neutron Activation Analysis with
a Radiochemical Separation

     The procedure followed  was that  used by Ruch, Kennedy and Shimp  (1970)
for the determination of arsenic  in sediments.  The irradiated clay samples,
with added As"1"*"1" carrier,  are  dissolved with hydrofluoric and perchloric
acids and hydrogen peroxide, and  the  silicon tetrafluoride evaporated off.
The residue was taken up in  hydrochloric acid and arsenic was separated by a
hydrogen bromide distillation; the procedure henceforth was the same  as that
used for liquid samples.

X-Ray Fluorescence Method  for  Mg. Al. Si, Ca. K.  Ti. Fe, and P
                                  (John Kuhn and Ray Henderson)

     The column section samples and clays were  air dried then ground  to ~200
mesh and oven dried  at  105°  C  for 6 hours. From the dried sample,  125 mg was
weighed into a graphite crucible  containing 1.000 g of lithium tetraborate. A
depression made in the  tetraborate prior to addition of the sample prevented
its contact with the crucible  wall. Next, 125 mg  of lanthanum oxide was added
as a heavy-element absorber, and  the  contents of  the crucible' were mixed,
with a  glass rod, as thoroughly as possible without scraping the crucible
wall or bottom. The  mixture  was fused in a furnace for 15 minutes  at  1000° C,
removed, covered with a second crucible and allowed to cool to room tempera-
ture. The resulting  fused  pellet  was  weighed alone to determine fusion loss
and placed in the grinding vial of a  No. 6 Wig-L-Bug with 2 percent by weight
of Somar Mix  (a commercial mixture used as a grinding and plasticizing
agent). The sample was  ground  for 3 minutes, transferred to a die, and
pressed at 40,000 psi.  Samples were then exposed  to X-rays and the data com-
pared to values acquired from  standard rock pellets prepared in the same
manner. The resulting concentrations  determined on the samples are quantita-
tive for those major and minor elements listed  (Mg, Al, Si, Ca, K, Ti, Fe,
and P).

Optical Emission Spectrometry  Method for Be. V. Cr, Co. Ni, Cu, and Mo
                                                          (Gary Dreher)

     Direct reading  optical  emission Spectrometry was used to determine trace
element concentrations  in  the  original clay materials. The procedure  used was
to prepare a mixture of 1  part Ba(N03)2 (5 mg) , 8 parts sample (40 mg) and 31
parts SPEX graphite  powder (Spex  Industries, Inc.) (155 mg) in a 1/2  inch di-
ameter, 1 inch long  polystyrene vial containing two 1/8 inch diameter poly-
styrene beads. The sample  mixture was shaken in a Wig-L-Bug for 60 seconds.
Fifteen milligrams of this mixture was loaded into each of 4 electrodes, 1/8
inch in diameter, having thin  wall craters. The sample electrodes were arced
at 15 A. short-circuit  current for 65 seconds with an arc gap of 4 mm, and
surrounded by a  10 SCFH flow of  a gas mixture 80% in argon and 20% in oxygen.


Determination of Total  Carbon  (TC). Inorganic Carbon (1C), and Organic


                                      143

-------
Carbon (PC)
(David Heck and Larry Kohlenberger)

     Total Carbon (TC) and Inorganic Carbon (1C) were determined in the
column sections and clay samples. Organic Carbon (OC) was determined by dif-
ference of TC and 1C. The procedures are as follows:

     Total Carbon

     A zirconium silicate combustion boat containing a layer of dry alumina
     powder (to prevent sample fusion to boat) is accurately weighed. A
     layer of sample  (1.5 to .30 g) is added, and the boat is reweighed in
     order to obtain the sample weight. A Nesbitt absorption bulb containing
     a bottom layer of magnesium perchlorate and topped with C02 absorbent
     (lithasorb or ascarite) is also weighed. The weighed absorption bulb is
     placed on the exit end of a zirconium silicate combustion tube which is
     heated to 1350° C in a high temperature furnace. With dry oxygen flow-
     ing through the tube continuously, the sample boat is slowly pushed
     into the hot zone at a rate which will neither crack the boat nor blow
     out the sample. After the sample has been in the hot zone for 4 minutes,
     the sample boat is removed, the absorption bulb closed off, and the
     sample allowed to cool. It is reweighed, the C02 absorbed determined by
     difference, and % total carbon calculated.

     Inorganic Carbon

     A modified ASTM  (1968) procedure, D 1756, is used whereby C02 is formed
     by decomposing ~0.5 g sample with acid and absorbing the gas on Litho-
     sorb.

Methods for NHt. TDS. COD, and Chloride
                        (Don Dickerson)
                                                                    r
     Ammonium ion (Nttf) was determined in the filtered (but not acidified)
leachate samples as follows. A suitable volume of leachate (1 ml 0.01 N HC1
is equivalent to 180 mg Nflt/liter) is placed in a Kjeldahl flask and diluted
to -20 ml with NHt free water. The flask is connected to the Kjeldahl dis-
tillation apparatus. The receiver contains 30 ml of 4% boric acid solution.
Fifty percent KOH solution is slowly added into the flask port followed by
5 ml of NaOH-Na2S203 solution (25 g NaOH, 5 g Na2S203, and 75 ml H20). The
solution is steam distilled for 4.minutes after NHt starts coming over. Three
drops of methyl red-bromocresol green indicator is added and the solution is
immediately titrated to the orange-pink end point withf>Q.01 N HC1 solution.

Determination of Exchangeable

     Exchangeable NH£ was determined on the column sections according to a
modified procedure by Jackson (1958). One sample (containing approximately
.8 - 1.0 mg Nfft) is analyzed by the normal Kjeldahl procedure. A second
sample is combined with 20 ml of 1.0 N, acidified NaCl and distilled simi-
larily following the Kjeldohl procedure.
                                     144

-------
Determination of Chloride

     Chloride was determined  in the filtered (but not acidified)  leachate
samples as follows. A sample  containing not more than 4 mg of Cl~ is diluted
to 25 ml with Cl-f ree water and then combined with 25 ml of isopropyl al-
cohol. While stirring,  2 drops  of  bromophenol blue is added and the  pH ad-
justed to slightly  acid by the  addition of 1.0 N NaOH solutions or 1.0 N HN03
solutions as needed.  One half ml of a saturated solution of diphenyl carba-
zone in isopropyl alcohol is  added and the solution titrated to the  endpoint
with 0.1 N Hg4*  (as Hg(N03)2  in dilute HN03).

Determination of Chemical Oxygen Demand (COD)

     The following  procedure, adopted from Leithe, (1947)  was used to analyze
filtered (but not acidified)  leachate samples. A 1-10 ml leachate sample di-
luted to 50 ml with distilled water and 25 ml 0.10 N KaCraO? solution was
added to 0.5 g HgSOij  in a 300-ml ground-joint flask with a Friedrich con-
denser. Concentrated  sulfuric acid (75 ml) containing 1 g AgaSOij  was added in
small portions with thorough  swirling as well as two boiling stones  as anti-
bumping aids, and refluxing was carried out for 2 hours. The mixture was
cooled, washed into a 500-ml  Conical flask, diluted to 350 ml with distilled
water, treated with 3 drops ferroin indicator and titrated with 0.1  N Fe
 (NHif) 2(80^)2 solution to a stable  color change from blue-green to reddish-
brown. A blank value  was determined using 50 ml of distilled water under the
same  conditions.

Determination of Total Dissolved Solids (TDS)

      TDS was determined in the  filtered (but not acidified) leachate samples
as follows. Ten ml  was placed in a weighed beaker, covered with a watch-
glass, and placed  in  an oven  preheated to 250° F until evaporation was com-
pleted. The temperature was then raised to 350-375° C for an hour. The resi-
due was finally  cooled in  a vacuum desicator and weighed.

Methods for Sulfate,  Phosphate, and Boron
                              (David Heck)

      Sulfate was determined  in  the original leachate solutions as follows.
 Fifty ml of sample  was neutralized with concentrated HCl and then 0.5 ml
 added in excess  and the sulfate precipitated in almost boiling solution with
 10 ml of  10% BaCl2  added drop by drop with constant stirring. After  digesting
 one hour until the  BaSOi, settled,  another drop of barium chloride was added
 to make sure the precipitation  was complete, and the BaSOi, filtered  off,
 washed with hot water, ignited  in  the muffle furnace one hour at  1000° C and
weighed as BaSOi*.

 Determination of Phosphate

      Phosphate ion  was determined  in the original leachate solutions as fol-
 lows. Fifty ml of  sample was  brought to boil with 15 cc HN03 and  40  cc H20,
 digested for an  hour, cooled, filtered. The filtered material was washed with
 a dilute HN03-NH^N03  solution.  The residue was discarded, the filtrates were

                                      145

-------
combined and 11 ml ammonium citrimolybdate reagent added, brought  cautiously
to a boil, and then let to stand overnight. The resulting precipitate was
filtered through a fritted crucible, washed with the above solution and
water, dried at 100-105° C, cooled, and weighed.

Determination of Boron

     Boron was determined on the column sections and clay samples  according
to the following procedure.

     A 25-50 mg solid sample was weighted into a clean 4 oz. plastic bottle,
and 10 ml of 2.5 NH2SOn and 4 ml 5% HF added. The bottle was fitted with a
tight cap, swirled, and allowed to stand overnight. The sample was diluted to
a total volume of 20 ml, and 10 ml of .001 M methylene Blue (.374  g/1) and
25 ml of purified ethylene dichloride added. The bottle was shaken for at
least 30 minutes, removed from shaker, and layers allowed to separate. A 5 ml
portion of the lower layer was pipetted into a 25 ml volumetric flask and
diluted to volume with ethylene dichloride. Unknown sample was then compared
to known standards (1-6 ppm range) on a Beckman DBG Spectrophotometer at 660
Vim. A reagent blank was required for standards and samples. Boron was deter-
mined in the leachate fractions according to the same procedure, however, two
ml were used as the sample.

Determination of Base Exchange Capacity
                           (Bill Armon)

     Base Exchange Capacity was determined on the column sections according
to a procedure by Peech (1947), with slight modification. The 10 g sample was
leached with ammonium acetate to remove exchangeable cations and to saturate
the exchange complex with ammonia. The excess ammonia was removed by leaching
with ethyl alcohol and the remaining, exchangeable ammonia removed and deter-
mined by distillation in a Kjeldahl apparatus.

Determination of Surface Area
                 (Joe Thomas)

     The surface areas of the clays were determined by the B.E.T. method
(Brunauer, Emmett, and Teller, 1938) using nitrogen as the adsorbate in a
continuous flow system (Nelsen and Eggertsen, 1958),
                                    146

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 . REPORT NO.
 EPA-600/2-78-157
4. TITLE AND SUBTITLE

 ATTENUATION OF POLLUTANTS  IN  MUNICIPAL LANDFILL  LEACHATE
 BY CLAY MINERALS
                                  3. RECIPIENT'S ACCESSION NO.


                                  5. REPORT DATE

                                  August 1978  (Issuing Date)
                                  6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

 R.  A.  Griffin and N. F.  Shimp
                                  8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS

 Illinois State Geological Survey
 Urbana, Illinois   61801
                                  10. PROGRAM ELEMENT NO.
                                     1DC618
                                  11. CONTRACT/8R£NX NO.

                                     68-03-0211
 12. SPONSORING AGENCY NAME AND ADDRESS
 Municipal Environmental  Research Laboratory--Cin.,  OH
 Office of Research and  Development
 U.S. Environmental Protection Agency
 Cincinnati, Ohio   45268
                                  13. TYPE OF REPORT AND PERIOD COVERED
                                  Final Report 12/72  to 8/75
                                  14. SPONSORING AGENCY CODE
                                  EPA/600/14
 15. SUPPLEMENTARY NOTES

 Project Officer: Mike  H.  Roulier (513) 684-7871
 16. ABSTRACT
      The first part of  this  project was a laboratory  column study of attenuation of
 pollutants in municipal  solid waste landfill leachate by mixtures of sand and calcium-
 saturated clays.  Chloride,  Na, and COD were relatively unattenuated by passage through
 the clay columns; K,  NHd,  Mg, Si, and Fe were moderately attenuated; and the heavy
 metals Pb, Cd, Hg, and  zn  were strongly attenuated  even in columns with small amounts
 of clay.  Calcium, B, and  Mn were higher in column  effluents than in the applied
 leachate.  Precipitation was the principal attenuation mechanism for the heavy metals;
 cation exchange was responsible for any attenuation of the other elements.  The clays,
 in order of increasing  attenuation capacity, were Kaolinite, Illite, Montmorillonite.
      The second part  of the  project involved batch  studies of adsorption of Cr, Cu,  Pb,
 Cd, Hg, and Zn by Montmorillonite and Kaolinite  from  water solutions and from landfill
 leachate.  Adsorption of the cations Cr(III), Cu, Pb, Cd, Hg, and Zn increased with  in-
 creasing pH; adsorption of the anions Cr(VI), As, and Se decreased with increasing pH.
 Above pH = 5.3 precipitation of the cations was  an  important mechanism while adsorption
 was the principal mechanism  for the anions over  the pH range studied.  Because adsorp-
 tion/mobility of any  element was affected by other  solutes in leachate, adsorption
 information on one leachate  may not be directly  applied to predicting adsorption of  the
 same element at the same concentration in another leachate.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b. IDENTIFIERS/OPEN ENDED TERMS
                                                  COS AT I Field/Group
  Attenuation
 *Transport Properties
 *Soil Chemistry
  Contaminants
  Arsenic
  Cadmium
  Copper
Industrial Wastes
Chromium
Iron
Lead (Metal)
Mercury (Metal)
Selenium
Zinc
Pollution Hazard  Rating
Chemical Oxygen Demand
Groundwater Pollution
Clay Liners
Municipal Solid Waste
Leachate
Ammonium (ion)
13B
18. DISTRIBUTION STATEMENT

  RELEASE  TO PUBLIC
                     19. SECURITY CLASS (This Report)
                     UNCLASSIFIED
                          21. NO. OF PAGES
                            157
                                              20. SECURITY CLASS (This page)

                                              UNCLASSIFIED
                                                22. PRICE
EPA Form 2220-1 (Rev. 4-77)
                                            147     -fa U. S. GOVERNMENT PRINTING OFFICE: 1978-757-140/1445 Region No. 5-11

-------