United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
EPA-600/2-80-027
May 1980
Research and Development
Attenuation of
Water-Soluble
Polychlorinated
Biphenyls by Earth
Materials

-------
                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-600/2-80-027
                                        May 1980
ATTENUATION OF WATER-SOLUBLE POLYCHLORINATED
        BIPHENYLS BY EARTH MATERIALS
                     by
                R. A. Griffin
      Illinois State Geological  Survey
           University of  Illinois
           Urbana, Illinois  61801

                     and

               E. S. K. Chian
         School of Civil  Engineering
       Georgia Institute  of  Technology
           Atlanta, Georgia  30332
            Grant  No.  R-804684-01
                Project  Officer

               Richard A.  Carnes
 Solid  and  Hazardous Naste 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
                              U S. Environmental Protection Agenc*
                              Region 5, Library (PL42J)
                              77 West Jackson Boulevard, 121& WW
                              Chicago, II 60604-3590

-------
                                DISCLAIMER
     THIS report has been reviewed by the Municipal Environmental Research
     atory, U.S. Environmental Protection Agency, and approved for publica-
tion.  Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
                                    11

-------
                                 FOREWORD


     The Environmental  Protection Agency was created because of increasing
concern by the public and government 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 paving a
lem— involves defining the problem, measuring its  impact, and starphtng for
solutions.  The Municipal  Environmental  Research Laboratory develops new
and improved technology and systems to preserve and treat public drinHing
water supplies, to minimize the adverse  economic,  social, health, and
aesthetic effects of polution and to prevent, treat, and manap waste water
and solid and hazardous waste pollutant  discharges from municipal and) com^
munity sources.  This publication is one of the products of tfiat research,
it is a vital communications link between the researcher and th§ community.

     Results are reported  from laboratory investigation of the capacity of
earth materials to attenuate polychlorinated biphenyls (RGBs) from waters.
These results are applicable to the design of treatment systems, earth
material liners for sanitary landfills,  and to the land disposal of
pal and hazardous wastes.

                                     Francis T. Mayo, Director
                                     Municipal Environmental Research
                                     Laboratory
                                   m

-------
                                  ABSTRACT

     The aqueous solubility, adsorption,  mobility,  microbial  degradation,  and
volatility of polychlorinated biphenyls (PCBs)  were studied under laboratory
conditions.  The dissolution of Aroclor 1242 in water required five months to
reach equilibrium.  Generally, the water-soluble fractions of the PCB fluids
were richer in the lower chlorinated isomers than in the original mixture  of
isomers in the fluid.  The solubilities of Aroclor  1016, 1221, 1242, and 1254
were 906 ppb, 3516 ppb, 703 ppb, and ^70  ppb, respectively.
     A simple linear relation described the adsorption of water-soluble PCBs
by five earth materials and their low-temperature ashes.  An adsorption con-
stant (K) unique to each adsorbent was obtained.  The adsorption was strongly
correlated to the total organic carbon (TOC) content and surface area of the
earth materials.  TOC was the dominant of these two earth material properties
by a ratio greater than three to one.

     The mobilities of several PCBs in silica-gel,  silica sand, seven soils,
and a coal char were measured with several leaching solutions using the soil
thin-layer chromatography technique.  The PCBs  remained immobile when leached
with water or landfill leachate, but were intensely mobile when leached with
organic solvents.  The mobilities were strongly related to the solubilities of
the PCBs in the leaching solvent.  The mobilities of PCBs in soils leached with
carbon tetrachloride were highly correlated to  the  soil TOC content.
     The degradation of water-soluble Aroclor 1242  by mixed cultures of soil
microorganisms occurred in a short period of time.   The lower chlorinated  iso-
mers were degraded more easily than the higher chlorinated isomers.  The rates
of degradation ranged from the monochloro isomers,  which degraded 100 percent
within 6 hours, to the tetrachloro isomers, which averaged 42 percent degrada-
tion after 15 days.  The predominant organisms  found in the mixed cultures were
Alkaligenes odoranSj Alkaligenes deni-trificans3 and an unidentified bacterium.
     The volatility of PCBs from water was studied  in the presence of soluble
humic acid and suspended soil.  Adsorption of PCBs  by the soil and humic acid
reduced the total amounts of PCBs volatilized.   The volatilization of PCBs
from pure water agreed with theoretical predictions that the half-life of
Aroclor 1242 stripped from water would be about 6 hours.  Higher chlorinated
PCB isomers were found to be less soluble in water, preferentially adsorbed
by soil materials, less mobile in soil, less degradable by microorganisms, and
less volatile from water than lower chlorinated isomers.  Thus, higher chlori-
nated isomers would be less mobile and more persistent in the environment than
lower chlorinated isomers.
     This report was submitted in fulfillment of Grant No. R-804684-01 by the
Illinois State Geological Survey and the Civil  Engineering Department, Univer-
sity of Illinois, under partial sponsorship of the U.S. Environmental Protec-
tion Agency.  Work was completed in October, 1978.


                                      iv

-------
                                   CONTENTS
 Disclaimer	   .11
 Foreword	ni
 Abstract	    iv
 Figures	   V1
 Tables	   vii
 Acknowledgments 	 viii

      1.  Introduction	    1
               Chemical structure 	     1
               Manufacture and industrial uses	    1
               Environmental effects 		    3
               Disposal and soil  attenuation 	    4
               Microbial  degradation 	    5
               Analytical methodology 	     8
               Concluding remarks	    10

      2.  Conclusions  	    H

      3.  Recommendations .	    14

      4.  Materials and methods	    T6
               PCB materials and extraction procedures 	     16
               Earth materials	     19
               Adsorption studies 	    21
               PCB mobility studies 	    22
               Microbial  degradation 	     24
               PCB volatility measurements 	     26
               Analytical methodology 	    28

      5.  Solubility of PCBs in water	    36
               Results and discussion 	    36
      6.  Adsorption of water soluble PCBs by soil materials
            and coal chars	     48
               Results and discussion	    48

      7.  Mobility of  PCBs and Dicamba in soil materials:
            Determination by soil thin-layer chromatography 	    58
               Results and discussion 	    58

      8.  Degradation  of PCBs by mixed microbial  cultures 	    63
               Results and discussion	    63
      9.  Effect of humic acid and soil on volatilization
            of PCBs from water	     82
               Results and discussion 	    82
References.	    87
Appendix	   92

-------
                                            FIGURES


Number                                                                                   Page

  1.   Biphenyl  structure 	  1
  2.   Modified  Erlenmeyer flask  for  extraction of  PCBs from aqueous sample 	 16
  3.   GC chromatogram of internal  standard  (tribromobenzene) in hexane  	 17
  4.   Chromatogram of internal standard and  four PCB  isomers	18
  5.   Flow diagram of experimental protocols used  to  study PCB volatilization from
        aqueous solutions 	26
  6.   Purging system using 1000-mL Erlenmeyer flask	28
  7.   Water-soluble Aroclor 1242 using temperature-programmed column 	 29
  8.   Water-soluble Aroclor 1242 on  second capillary  column 	31
  9.   Water-soluble Aroclor 1242 on  third capillary column	32
 10.   Reconstructed chromatogram of  Aroclor  1016 obtained from mass-spectrometer 	 33
 11.   Teflon-plugged reaction  vessel  used in perchlorination reaction	34
 12.   Schematic block diagram  of the perchlorination  procedure 	 f 	 35
 13.   Comparison of hexane-soluble Aroclor 1242 and used capacitor fluid 	 37
 14.   Solubility of Aroclor 1242 and used capacitor fluid 1242 in water as a
        function of time	40
 15.   GC chromatograms of water-soluble Aroclors and  used capacitor fluid	43
 16.   PCB adsorption by earth  materials	49
 17.   PCB adsorption by low-temperature ashes of earth materials  	 50
 18.   Capillary column analysis  of Aroclor 1242 remaining in solution	51
 19.   Percentage of isomers in water-soluble Aroclor  1242 remaining in solution.
        after adsorption by char	57
 20.   PCB adsorption constants as  function of total organic carbon content 	 57
 21.   PCB adsorption constant  as function of surface  area 	57
 22.   Three-variable regression  analysis of  PCB adsorption constant as function
        of total organic carbon  content and  surface area . 	 57
 23.   Retention of PCB in soils  leached with carbon tetrachloride 	60
 24.   Water-soluble Aroclor 1242 after 15 days of  degradation 	v	65
 25.   Original  water-soluble Aroclor 1242 solution prior to degradation 	66
 26.   Comparison of 5-day blank  and  5-day degradation of Aroclor  1242 using S culture .  .  .69
 27.   Results of 5-day blank of  Aroclor 1242 using the N culture	70
 28.   Results of 5-day degradation of Aroclor 1242 using the N culture  	 70
 29.   Degradation at timed intervals of selected PCB  isomers 	 73
 30.   Degradation at timed intervals of selected PCB  isomers 	 74
 31.   Relative  distribution of PCB isomers in soil  and sludge extracts with
        respect to retention time	79
 32.   Comparison of attenuation  of volatilized PCBs by dissolved  humic acid addition ... 83
 33.   Comparison of attenuation  of volatilized PCBs by soil  addition	85
                                              VI

-------
                                           TABLES


Tables                                                                                    Page

  1.   Characteristics of earth  materials  used  in  adsorption  studies  	 20
  2.   Chemical  characteristics  of earth materials used  in adsorption studies  	  20
  3.   Characteristics of soil materials used in PCB mobility studies	21
  4.   Gas chromatographic conditions  used for  volatilization study  	  27
  5.   Relative  response  of a SSfH  electron capture detector to tribromobenzene
        and PCB isomers	36
  6.   Quantitative composition  and  percentage  of  contribution of  individual isomers
        in Aroclors and  capacitor fluid	'.	38
  7.   Percentage of PCB  isomers in  water-soluble  Aroclors 1242  	  **!
  8.   Comparison of aqueous  solubility of PCBs in Aroclor 1242 to pure isomers 	  ^*
  9.   Quantitative comparison of water-soluble isomers  in Aroclor 1242 with respect
        to time	^2
 10.   Summary of isomer  composition of water-soluble Aroclors and capacitor fluid	**5
 11.   Summary of solubility  of  PCBs and capacitor fluid  in water  .  . .	  46
 12.   Quantitative distribution of  PCBs after  adsorption by  several adsorbents 	  53
 13.   Mobility-of Aroclor 1242  and  1254 and Dicamba in  soil materials leached with
        various solvents 	  "
 14.   Mobility  of Aroclor 1242, Aroclor 1254,  and Dicamba on silica-gel TLC plates
        using various leaching  solvents 	 61
 15.   Results of linear  regression  analysis of retention of Aroclor 1242 and Aroclor
        1254 by soils	61
 16.   Results after 15 and 45 days  of degradation using  culture grown in acetic acid ...  63
 17.   Comparison of mixed cultures  grown  on biphenyl only and those grown on
        acetic  acid	64
 18.   Comparison of sonified extracts and pH 2 extracts	68
 19.   Comparison of degradation of  water-soluble  Aroclor 1242 by three mixed
        cultures of soil  microorganisms 	 71
 20.   Results of kinetic study  of degradation  of  water-soluble Aroclor 1242 by the
        N culture	72
 21.   Rate constants and reaction orders  for selected PCB isomers 	 76
 22.   Cometabolization of PCBs  with sodium acetate by mixed culture microorganisms ....  78
 23.   Relative  amounts of PCB volatilized from water containing humic acid or soil
        compared to control	82
 24.   Relative  attenuation of volatilization of PCBs by  soil as compared with
        attenuation due  to adsorption from solution by  same soil	84
                                             vii

-------
                             ACKNOWLEDGMENTS
     In addition to the authors listed on the title page, the following were
involved in writing parts of this report:  A. K. Au, R. R. Clark, H. Meng,
M. C. Lee, and M. L. Miller.  Their names are also listed in a footnote in
the portion of the Results and Discussion Section to which they contributed1.

     The authors gratefully acknowledge the U.S. Environmental Protection
Agency, Municipal Environmental Research Lab., Solid and Hazardous Waste
Research Div., Cincinnati, Ohio, for partial support of this study under
Grant R-804684-01, Mr. Richard A. Carnes, Project Officer; Monsanto Chemical
Co., St. Louis, Missouri, for supplying the Aroclors; Illinois Power Co.,
Decatur, Illinois, for supplying the used capacitor; American Colloid Co.,
Skokie, Illinois, for supplying the montmorillonite; the Ottawa Silica Co.,
Ottawa, Illinois, for supplying the quartz sand; T. M. Hinsley, for supplying
the sludge-soil samples; and C. N. Goddard, for supplying the Hudson River
sediment samples used in this study.

     The authors also wish to thank Dr. J. H. Kim, Dr. N. F. Shimp,
Dr. R.  R. Ruch, Dr. J. Thomas, Jr., Dr. R. R. Frost, Dr. L. R. Follmer,
Dr. J.  J. Hassett, Dr. W. A. White, F. L. Fiene, J. E. Johnson, A. D. Debus,
Dr. K.  E. Smith, J. M. Masters, L. Kohlenberger, J. D. Steele,
L. R. Henderson, J. F. Ashby, Dr. F. B. DeWalle, R. A. Cahill,
Dr. J.  K. Frost, L. R. Camp, R. A. Keogh, and J. K. Kuhn for assistance
with parts of this research.
                                    viii

-------
                                  SECTION 1
                                INTRODUCTION
     Polychlorinated  biphenyls (PCBs) is
mixtures of  chlorinated organic compounds
much discussion  recently because of their
persist and  accumulate  in the environment
air, water,  soil,  sediments, fish, birds,
all over the world.   PCBs are not acutely
have been observed.   Therefore, PCBs have
hazard to human  health  as well as to the
EHRC, 1976).

CHEMICAL STRUCTURE
a generic term applied  to  certain
   PCBs have become the  subject of
 effect  on the environment.   They
 and have been found  in  samples of
 and mammals (including  humans)
 toxic, but several chronic  effects
 been considered a significant
environment (Gustafson,  1970,  1972;
     PCBs belong  to  a  class  of aromatic chlorinated organic  compounds  that
comprise complex  heterogeneous mixtures of very closely  related  isomers.
The basic biphenyl structure is shown in Figure 1.  PCBs are made  by sub-
stituting chlorine atoms  for one or more of the hydrogen atoms at  the
numbered positions of  the biphenyl  structure.  Possible  chlorine substitu-
ted biphenyl isomers number  209, but commercial preparations contain less
than the total possible;  about 103  have been identified  in various samples
(Widmark, 1967; 1968).  The  different
commercial preparations may  vary from
batch to batch with  regard to the
specific composition of the  mixture
of chlorinated biphenyl isomers (EHRC,
1976; Lloyd et al.,  1976).

MANUFACTURE AND INDUSTRIAL USES

     PCBs have been manufactured in
the United States since 1929 by the
Monsanto Chemical Company; an esti-
mated 800 million pounds  have been
produced since that time.  In 1970,
the year of peak  production, more
than 85 million pounds of PCBs were
produced in the United States alone,
57 percent of which was in the form    Fi9urei.  BIPHENYL STRUCTURE: Positions 2 toe and 2-
n-F flvr^lnv" 19/19 (UE\I   ^Q79^    A               to 6'indicate ten possible positions for chlorine
OT HrOCIOr \mt. ^tltW,  iy/(i;.  An              substitution. Different amounts of chlorine substi-
estimated one-half million pounds             tion form the various PCBS.

-------
per year of PCBs were imported into the United States from foreign manufac-
turers (EHRC, 1976).

     The sole United States manufacturer of PCBs was the Monsanto Chemical
Company located near East St. Louis, Illinois.  Monsanto has marketed its
PCB products in the United States under the trade name AROCLOR.  A four-
digit numbering system was used to identify various Aroclors.  The first
two numbers, 12, specified polychlorinated biphenyls, and the last two
numbers referred to the approximate percentage of chlorine in the mixture.
For example, Aroclor 1242 means a PCB mixture with 42 percent chlorine by
weight (EHRC, 1976; Nisbet and Sarofim, 1972).  A product called Aroclor
1016 came into use after 1971 which did not follow the above numbering
sequence.  This product contained approximately 41 percent chlorine and was
an attempt by Monsanto to produce an Aroclor that could be used in the same
applications as Aroclor 1242; it contained fewer of the higher chlorinated
isomers and was thus more degradable and less persistent in the environment
(Tucker, Litschgi, and Mees, 1975).  Aroclors can also be classified by the
average number of chlorine atoms per molecule.  Thus, Aroclor 1232, 1242,
1248, 1254, 1260, and 1262 correspond to di-, tri-, tetra-, penta-, hexa-,
and hepta-chlorobiphenyls, respectively.  Other trade names for PCB products
produced by foreign manufacturers include:  KANECHLOR and SANTOTHERM (Japan),
PHENOCLOR and PYRALENE (France), FENCLOR (Italy), CLOPHEN (Germany), and
SOVOL (Russia) (EHRC, 1976; Lloyd et al., 1976).

     PCBs come in both solid and liquid forms and have properties appropriate
to a wide range of industrial applications.  In resin form they were used as
protective coatings, plasticizers and extenders, sealers in waterproofing
compounds and putty, asphaltic materials, printing inks, and synthetic adhe-
sives.  In liquid form they were used as dielectrics, hydraulic fluids, ther-
mostats, cutting oils, extreme pressure lubricants, grinding fluids, and heat
transfer media.  As solids they were used to impregnate carbon resistors, as
sealers, and as impregnating agents for electrical apparatus.  Mith few
exceptions, other than environmental concerns, the unique properties of PCBs
qualify them as the best chemical known to make capacitors, transformers,
hydraulics, gas turbines, and vacuum pumps work efficiently and safely
(EHRC, 1976).

     In 1971, because of environmental  concerns, Monsanto voluntarily stopped
production of Aroclor 1260 and restricted the sale  of other PCBs to only
"closed" systems.  Closed systems include PCB-containing insulating fluids
used in electrical transformers and capacitors;  these two applications account
for essentially all the current use of PCBs in the United States (ANSC,  1976).
On October 5, 1976, Monsanto announced that it would cease to manufacture
and distribute PCBs by October 31, 1977.  A timetable set by the United
States Environmental  Protection Agency (EPA) called for gradual phasing out
of PCB manufacturing by January 1, 1979, and a ban on all PCB processing or
distribution in commerce by July 1, 1979 (ES&T,  1977).   These steps should
reduce the introduction of PCBs into the environment in the future, but
millions of pounds of PCBs still  exist in transformers  and capacitors still
in service.  The environmentally safe disposal of these fluids will continue
to be of concern for more than a decade.

-------
ENVIRONMENTAL EFFECTS

     PCBs were largely ignored as environmental  contaminants until 1966 when
Jensen (1966) reported PCB contamination of fish, eagles, and humans.  PCBs
still did not attract much concern as hazardous  chemicals until  the incidence
of contaminated cooking oil in Japan (Risebrough et al., 1968) and of con-
taminated chicken feed in the United States (Duke, Lowe, and Wilson, 1970).
Laboratory studies with animals have shown that  PCBs can cause enlargement
of the liver, induction of hepatic microsomal  enzymes, reproductive failures,
gastric disorders, skin lesions, and tumors in birds and mammals  (Nelson
et al., 1972).

     The best documented case of toxic effects of PCBs in humans is the 1968
"Yusho" incident in Japan.  About 2,000 Japanese people experienced lesions
of the skin, facial swelling, and neurological disorders.  The individuals
with the highest dosages had nausea, lassitude,  anorexia, impotence, and
hematuria.  The severity of the symptoms were directly related to the
quantities of PCBs ingested.  Placental transfer of PCBs had occurred in
pregnant women and had adversly affected the fetuses.  The toxic effects
noted were similar to those reported in animal studies (Nelson  et al. , 1972;
EHRC, 1976).

     Several studies with animals have shown that fatty tissue accumulates
about 10 times more PCBs than liver and more than 100 times more than blood,
heart, kidney, and brain (Burse et al.,1974; Grant, Phillips, and Villeneuve,
1971; and Curley et al., 1971).  A study with refuse workers found that 81
percent had detectable levels of PCBs in their blood  (Hammer et al., 1972).

     PCBs have found their way into the marine and terrestrial environment.
Risebrough et al.  (1968) noted that PCBs, along  with the chlorinated hydro-
carbon pesticides, have become the most widespread and ubiquous synthetic
pollutants in the global ecosystem.  They also noted that PCBs induced hepa-
tic enzymes that degrade estradial, which leads  to aberrations in calcium
metabolism, such as weak, thin egg shells.

     Because PCBs are fat soluble and poorly metabolized, they have entered
the food chain and accumulated in the adipose tissues of fish, birds, and
mammals, including humans  (Jensen et al., 1969;  Hammond, 1972; Solly and
Shanks, 1974; Harvey and Steinhauer, 1975).  The biological magnification of
PCBs that have entered the food chain seems to occur by a factor of 10 to
100 at each step.  Fish and aquatic organisms accumulate PCBs by a factor
of TO4 over concentrations in the ambient waters.  Predators at the top of
the food chain can accumulate PCBs by a factor of 107 over the concentra-
tions found in the ambient waters (EHRC, 1976).   The major problem with
regard to PCB intake by man is that he resides at the top of most food
chains.  Man can therefore amass substantial amounts of PCBs although only
trace levels are present in the waters of lakes  and streams.

     A two-year study of Lake Michigan showed that fish with the highest fat
content contained the highest levels of PCBs (Schacht, 1974).  Other studies
have shown that salmon taken from Lake Michigan  in 1974 contained PCBs in

-------
 excess  of  the  5  ppm  FDA  tolerance  level.   Lake  trout were  reported  to  contain
 PCBs  as  high as  43.8 ppm,  and  carp contained  levels as  high as 51.6  ppm
 (EHRC,  1976).

      The highest concentrations of PCBs are  found in fish that come from the
 Hudson  River in  New  York.   Fish taken from below the General Electric  capac-
 itor  manufacturing plant contained PCBs in the  range of 17 to 78 ppm;  one
 sample  of  rock bass  contained  350 ppm (Ahmed, 1976).  The  bed and bank
 sediments  in a 40-mile reach of the river were  estimated to contain more
 than  400,000 pounds  of PCBs, and concentrations as high as 5,600 ppm were
 found (Maslansky et  al., 1978).

      Old electrical  equipment  has been discarded by many companies and the
 PCBs  have  discharged  into  the  atmosphere and waterways.  PCBs are somewhat
 volatile,  and the migration of PCBs through air is considered to be one of
 the basic  mechanisms  by  which  the ubiquitous presence of PCBs in nature
 occurs  (Persson, 1971).  An average fallout rate of 40  to  80 yg/m2/year of
 PCBs over  continental North America was reported by the Panel on Hazardous
 Trace Substances (Nelson et al., 1972).

 DISPOSAL AND SOIL ATTENUATION

     One problem of disposal involves the high costs and  fees for transport-
 ing PCB wastes to regional incinerators or approved landfills compared to
 simply discarding the wastes.  The costs at disposal  sites that specially
 treat or incinerate wastes may be as high as a dollar per gallon or more.
 The costs  for land burial are generally between 5 to 10 cents per gallon
 (Troise and Kahn, 1978).

     Incineration is  considered the safest method for disposal  of PCB wastes,
 but this method  is extremely costly and has some operating difficulties.
 PCBs do not burn readily; the operating conditions of the incinerator must
 be carefully controlled to prevent the reentry of PCBs  into the environment
 in the stack gases (Duvall and Rubey, 1977; Moon, Leighton, and Huebner,
 1976).

     Nisbet and Sarofim  (.1972) estimated the rate of PCB disposal  in dumps
and landfills in North America in 1970 to be 18,000 tons per year.   In view
of this, land disposal of PCBs and related materials  has concerned many
people (Jordan, 1977; Henderson, 1978) because surprisingly little is known
about the mobility of PCBs in soil  systems.

     The hydrophobic characteristic of PCBs make them easily adsorbed from
aqueous  solution onto available surfaces;  the amount  of PCBs being adsorbed
depends  on the nature of  the surface (Haque,  Schmedding, and Freed,  1974;
Lawrence and Tosine,  1976).  Haque and Schmedding (1976) studied the adsorp-
tion characteristics  of three selected PCB isomers by several  adsorbents
and provided evidence that adsorption increases  as the  number of chlorine
atoms on the isomer increases.

     Iwata, Westlake, and Gunther (1973) studied the  persistence of PCBs in
 six California soils.  Preferential disappearance of some gas chromatograph

-------
(GC) peaks was evident and recovery of Aroclor 1254 was related to the
organic-matter content of the soil.  In a separate study of PCB uptake by
carrots, 97 percent of the residue was found in the peel and very little
translocation into the plant tissue.  Again, the lesser chlorinated biphenyls
were slowly dissipated and the more highly chlorinated biphenyls were not
appreciably affected (Iwata, Gunther, and Westlake, 1974).

     Tucker, Litschgi, and Mees (1975) percolated water through soil  columns
containing PCBs.  In the worst case, less than 0.05 percent of the added
Aroclor 1016 was  leached;  they concluded that PCBs are not readily leached
from soils.  Similarly, Scharpenseel, Theng, and Stephan (1978) found that
very little leaching of PCBs occurred and that the PCBs recovered from the
soil were associated with the soil organic matter.

     Briggs (1973) reported that adsorption of unionized organic compounds
by soils was related to the organic-matter content of the soil and to the
octanol/water partition coefficients of the compound.  On the basis of the
octanol/water partition coefficients for PCBs, he predicted that they would
be immobile in soils.

     Lidgett and Vodden (1970) analyzed waters around a sanitary landfill
for PCBs and found the contamination levels to be below their detection
limits of 4 ppb.  Similarly, Robertson and Li (1976) used GC/Mass Spectrom-
et'ry techniques and failed to detect PCBs in ground water.  Hesse (1971)
reported that runoff from landfills was only a minor source of PCB contami-
nation to the environment.  More recently, Moon, Leighton, and Huebner (1976)
reported that levels of PCBs in ground water in the vicinity of 11 sanitary
landfills were below detection (<1 ppb), but that low levels of PCBs were
found in waters from monitoring wells at several industrial PCB disposal
sites and lagoons.  They concluded from analyses of water and splitspoon
soil samples that PCBs were present in most leachates from land disposal
sites and that PCBs have a strong affinity for soil.  Gresshoff, Mahanty, and
Gortner (1977) have also reported that PCBs have a high affinity for soils.
Leis et al. (1978) studied PCB migration in ground water from 12 dredge
disposal sites in the upper Hudson River Valley of New York.  They concluded
that the velocity of the front of PCB advance was about 2 orders of magnitude
slower than the velocity of the ground water flow.  They further concluded
that inputs of PCB that were due to migration through ground water were
negligible in comparison to other inputs of PCBs.

     The information presently available indicates that PCBs have a strong
affinity for soil and that the nature of the surface, the organic matter
content, and the chlorine content and/or hydrophobicity of the individual
PCB isomers are factors affecting adsorption.  Quantitative data on the
adsorption capacities and factors affecting PCB adsorption by earth materials
are needed to assess the impact of soil attenuation mechanisms for restrict-
ing PCB mobility in the environment.

MICROBIAL DEGRADATION

     The above studies suggest that the interaction of PCBs with soil part-
icles is an important attenuation mechanism.  Another potentially important

-------
 mechanism is  the degradation  of PCBs  by microorganisms.

 Biphenyl  Degradation

      Various  species  of microorganisms  that can  degrade  biphenyl  have been
 reported.   Lunt and Evans  (1970)  were the first  to  report the  microbial
 degradation of biphenyl.   The microorganisms used were gram-negative bac-
 teria found in soil.   They reported that biphenyl was  degraded to 2,3-
 dihydroxybiphenyl.  This compound was then degraded to a a-hydroxy-3-phenyl-
 muconic  semi aldehyde  and then to  phenylpyruvate.

      Catelani  et  al.  (1971, 1973)  reported  that Pseudomonas put-ida could
 degrade biphenyl.  The  biphenyl was degraded  to 2,3-dihydroxybiphenyl  and
 further degraded  to benzoic acid  and  4-hydroxy-2-oxovaleric acid.

      Gibson et  al. (1973)  reported a Beijerinckia species that could  degrade
 biphenyl to 2,3-dihydroxybiphenyl.  They also reported that the intermediate
 in this step,  cis-2,3-dihydroxy-l-phenylcyclohexa-4,6-diene was formed by
 bacteria, whereas the  trans isomers were  formed in  mammals.

 PCB Degradation by Pure Cultures

      Wallnofer  et al.   (1973)  isolated a soil  fungus, Rhizopus japonicas,
which converts  4-chlorobiphenyl to 4-chloro-4'hydroxybiphenyl.

      Ahmed and  Focht  (1973a,  1973b) reported  the degradation of several PCB
 isomers by two  species of AchromobaoteT .  They isolated the two species from
sewage, the first by biphenyl  enrichment and  the second by 4-chlorobiphenyl
enrichment.  They reported that the products  produced by these two species
were  different; this suggests different metabolic pathways for degradation.
They  showed that degradation of the unsubstituted aromatic ring was preferred.
This, and the observation  that no dechlorination occurred, lead to a  build-
up in chlorobenzoic acids.   Recently,  Omori and Alexander (1978) reported
that  certain alkanes can be dehalogenated.  If dehalogenation, along with
degradation, can occur, then PCBs can  be broken down into small nonchlori-
nated compounds.  Kaiser and Wong  (1974) reported the isolation of some
microorganisms that degraded PCBs into metabolites that contained neither
chlorine nor oxygen.   This  is  quite unusual because the PCBs are degraded
aerobically and therefore should be oxidized and should contain oxygen.

      Baxter et al. (1975) measured PCB degradation by a species of Noaard-ia
and a species  of Pseudomonas.   With the  Nooardia species  they reported 88
percent and 95 percent degradation for 52 and 100 days, respectively.  With
the Pseudomonas spectes they reported  76 and 85 percent for the same periods.
Both studies were'done with Aroclor 1242.  They also did  work with single
PCB isomers and in these studies found that PCB isomers with up to 6 chlo-
rines could be degraded to  some degree and that the pattern of chlorine
substitution determines the ease of degradation.   They also found that the
two species would degrade some isomers differently;  for example, the Nocardia
could not degrade 4,4'-dichlorobiphenyl  in 121 days, whereas the Pseudomonas
species degraded it about 50 percent in  15 days.   They also reported a form
of cometabolization,  in that some isomers degraded more slowly when present

-------
as the only carbon substrate, but degraded more rapidly when in a mixture
of isomers or when biphenyl  was present.   An example of this was 2,5,4'-
trichlorobiphenyl.  When alone this isomer degraded only 15 percent in 73
days; however, with biphenyl the degradation was 60 percent in 73 days.

     Furukawa and Matsumura (1976) isolated an Alkali-genes species from a
lake sediment by enrichment with biphenyl.  The PCBs were degraded in two
steps.  The first step was the formation  of a yellow product, identified
as a chlorinated derivative of 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoic acid.
This compound was then broken down into chlorinated benzoic acids.  In add-
ition, the isomers with 4'-chlorine produced longer lasting yellow inter-
mediates.  PCB isomers with chlorines on only one of the rings were degraded
more easily than those with chlorines on both rings.  They also found that
lower chlorinated isomers were more easily degraded and that isomers with
up to five chlorines could be degraded.

     A later study by Furukawa, Tonomura, and Kamibayashi  (1978) with Alka-
ligenes and Aoine-bobaoter used 31 different PCB isomers and showed similar
results.  They found that isomers with 3,6 or 2,2' chlorines were poorly
degraded.  They also reported less degradation as the number of chlorines
per molecule increased.  The two species showed about the  same amount of
degradation of each isomer except for 2,4,6-trichlorobiphenyl.  Several
trichlorobiphenyl isomers could be degraded almost completely in less than
three hours by these microorganisms.

     Sayler,  Shou,  and Colwell  (1977) isolated a species  of Pseudomonas
from sea water that degraded PCBs.  They obtained between  9 and 39 percent
degradation of Aroclor 1254  in 22 days, the percentage depending on the
starting concentration.  In  60 days they found between 63  and 84 percent
degradation, again the percentage depending on the starting concentration.
At 22 days, the lowest degradation was at the lowest concentration, 10 yg/L:
at 60 days the lowest degradation was at the highest concentration, 1000yg/L
Ballschmitter, Unglert, and  Neul  (1977) proposed several schemes for the
microbial breakdown of PCBs  into various metabolites, including chlorinated
benzoic acids.

      It has been  known for  some  time that PCBs could be degraded by mammals
and  birds.  Sundstrom, Hutzinger, and Safe  (1976) presented a review of the
metabolism of PCBs in animals.   Hutzinger, Safe, and Zitko  (1974) also
presented a review of PCBs  in both animals and microorganisms.  Sayler,
Shou, and Colwell  (1977) noted that the degradation of PCBs in  animals could
actually be caused by the action of the intestinal microbiota.

PCB  Degradation in Sewage

     Many PCBs end up in sewers, either through spills or  dumpings, and then
go into activated sludge systems.  Tucker, Litschgi, and Mees  (1975) worked
with  a continuously fed activated  sludge  unit.  At a feed  level of  1 mg/48
hrs., they reported 100 percent  degradation of  biphenyl, 81 percent degrada-
tion  of Aroclor 1221, 33 percent degradation of Aroclor  1016, 26  percent
degradation of Aroclor  1242, and  15 percent degradation  of Aroclor  1254.
With  Aroclor  1221, only the  3-,  4-, and 5-chlorine  isomers were not degraded

-------
 Mihashi  et al.  (1975)  reported  50  percent PCB  degradation  in  activated  sludge
 and  that the degree  of degradation decreased as  chlorine substitution
 increased.

      Herbst et  al.  (1977)  studied  the  fate of  two  radiolabeled  isomers  of
 PCB,  2,5,4'-trichlorobiphenyl and  2,4,6,2',4'-pentachlorobiphenyl.   They
 determined that the  two isomers were poorly degraded  by the activated sludge.
 In addition, most  of the PCBs ended up in the  activated sludge  and  not  in the
 supernatant. Jordan (1977)  reported on the case of Williams, Indiana;
 activated sludge containing  up  to  300  ppm PCBs was used as fertilizer and
 harmed both  the  people and the  wildlife of  the area.

 Inhibition by PCBs

      PCBs  also  inhibit the growth  of microorganisms that have not been
 acclimated to PCBs.  Bourquin and  Cassidy (1975) reported that  a sizeable
 proportion of estuarine  bacteria are sensitive to PCBs.  They pointed out
 that  if  PCBs were present in large  amounts, the  normal microbial heter-
 otrophic  activity could  be disrupted.

      Murado, Tejedor,  and Baluja (1976)  reported that Aroclor 1254 would
 inhibit  the  formation  of mycelium  of AspergiZlus flavus .  They  also found
 that  no  degradation  of PCBs occurs  in  these cultures.  Blakemore and Carey
 (1978) reported  the  inhibiting  of  growth  of two marine bacteria by low
 concentrations of Aroclor 1254.   They  found that the inhibition was dose
 dependent  and that respiration  was  not  inhibited.  Blakemore (1978) also
 reported  the inhibition  of nucleic  acid  synthesis in a marine bacteria and
 subsequent  inhibition  of growth in  these  bacteria.

      It  is clear from  the above studies  that PCBs are somewhat  degradable and
 may come under attack  by soil microorganisms.   If a mixed culture of soil
 microorganisms could be  enriched in organisms that could degrade PCBs at a
 significant  rate, it would then be  feasible to use biological  processes to
 treat PCBs; or PCB wastes could be  inoculated with the culture  before land
 disposal,  assuming that  the organisms would be competitive in the particular
 environment.

 ANALYTICAL METHODOLOGY

 Quantitative Analysis

     The analysis of PCBs has been a matter of estimation;  attempts at
 quantisation of PCBs have most often been done by measuring the  heights of
major gas-liquid chromatographic .peaks,   Koeman et al. (1969), semiquanti-
 tatfvely measured only one of the major peaks,  Skrentny, Hemken, and Dorough
 1971) used two, and Zitko (1971) used  three quantification peaks.   Reynolds
 (1971) applied Koemanls method and took an average of two  or  more major
 peaks in the mixture as the standard.   Hansen  et al.  (1971) averaged the
 heights of five  major peaks.   Likewise, PCBs also have been quantified by
 summing the areas of all the individual peaks  obtained by gas-liquid chroma-
 tographic analyses.   Jensen et al.   (1969) quantitatively estimated  the con-
centration of PCBs  by the total  peak areas.  Armour and Burke  (1970) measured

-------
PCBs in terms of Aroclor 1254 by using the total  area of the peaks.
Risebrough, Reiche, and Alcott (1969)  multiplied  a factor derived from meas-
urements of standard PCB solutions by a microcoulometric detector to the
areas measured by electron capture detector (ECD).  Beezhold and Stout (1973)
used a mixed standard to quantify PCBs.

     All the  gas-liquid chromatography used for  the above studies employed
an electron capture detector on the chromatograph.  Unfortunately, the
electron capture detector does not respond quantitatively the same for all
PCB isomers, as noted by Rote and Murphy (1971),  Cook (1972), Wiencke and
Roach  (1972), and Beezhold and Stout  (1973).  A procedure designed to over-
come this  problem was to perchlorinate the PCB sample (Berg, Diosady, and
Rees,  1972; Armour, 1973; Berg, Rees,  and Ali, 1978; Burkard and Armstrong
(1978).  In the perchlorination procedure, the PCBs were reacted with antimony
pentachloride  (SbCl5) to form decachlorobiphenyl   (DCB).  The advantage of the
perchlorination procedure is that all  the chlorinated biphenyl  is detected
as  one peak by the  GC,  thus  giving improved quantitation and speed of analy-
sis.   In addition,  the  ECD is more sensitive to the  higher  chlorinated com-
pound.  Studies by  Trotter and Young  (1975), however, showed problems in  the
perchlorination reaction with SbCl5 that were due  to trace  amounts of bromine
in  SbCl5 in  the form of SbBrCV  Because the bromine has a  higher affinity
for biphenyl  than  the  chlorine, bromonanochlorinated biphenyl  (BNCB)  is
formed during  perchlorination.  The BNCB  peak would  show up in  a  GC  trace
after  the  DCB  peak.  Further studies  are  needed to improve  the  quantitation
of  PCBs by this procedure.

      Improvements  in PCB analysis can  be made by  the use of  open  tubular
glass  capillary columns which have liquid phases  custom-coated  on the walls
of  the column.  The apparatus and dynamic-coating  procedures of Grob  (1975)
and Grob,  Grob, and Grob  (1977)   can  also be used.

     Generally, liquid  phases for capillary gas chromatography  can  be divided
into apolar, medium polar, and polar  phases.  Polar  phase columns were  notor-
iously short-lived.  Apolar  liquid phases  (OV-17,  OV-101, SE-30,  etc.) were
widely accepted as  the  most  efficient coatings because  of their longer  life;
these  generally are the ones available commercially.  Because  of  the  aroma-
tic characteristics of PCBs, a polar  phase  is needed to achieve maximum  sep-
aration efficiency.  The merit of the glass-surface  pretreatment  methods
given  by Grob  is  that  polar  phases such as  Emulphor  ON-870  can  be coated  on
the glass  and  used for PCB analysis with  increased column life.  In  addition,
the column can easily  be  regenerated.   When a column has deteriorated  to  an
unusable quality,  a cleaning solution can be passed  through the column  to
wash out the  residual  liquid,  and the column can  then be recoated.   In  the
case of silicone  liquid phases,  such  as OV-17, OV-101,  and  SE-30, the  regen-
eration step is  not possible.

Mobi1i ty Measurements

      The technique of  determining pesticide mobility in soils  by soil  thin-
 layer chromatography,  or  soil  TLC, was introduced by Helling and Turner
 (1968). Since the introduction  of the technique, the mobility of a large
 number of  pesticides and  radionuclides in a variety of  soils has  been tested

-------
 (Helling  1971a, 1971b, 1971c; Reeves, Francis, and Duguid, 1977).   Soil
 TLC is a laboratory technique that uses soil  as the adsorbent phase and
 developing solvent such as water, leachate, and organic solvent in  a TLC
 Syu-?™;. The system 1s restively simple and  yields quantitative data on  the
 mobilities of chemical constituents in soils  that appear to correlate well
 with trends noted in the literature (Helling  and Turner, 1968;  Briggs, 1973-
 Reeves et al., 1977).

 CONCLUDING REMARKS

      An exact value for the solubility of PCBs  in water is  difficult to
 determine because PCBs are mixtures of substituted isomeric biphenyls  that
 have different numbers of  chlorine  atoms.   The  structure of the  PCB molecule
 was  shown in figure 1  to illustrate the various  possible combinations  of
 chlorine substitution.   Of the 209  possible isomers, about  103  usually occur
 in  commercial  PCB mixtures (Widmark,  1978).   PCBs  are  hydrophobic and  are
 only slightly soluble  in water.   Isomers  with lower chlorine  substitution
 have been shown  to be preferentially  dissolved  in water (Zitko, 1971;  Hoover,
 (1971).   Another  important factor in  the  quantitative  analysis of PCB  solu-
 bility is that PCBs dissolve  in water  very  slowly.  Therefore, there must
 be  a time-dependent study  before  the absolute value of the  water solubility
 of  PCBs  can  be obtained  (Haque and  Schmedding,  1976).   PCBs are also subject
 to  volatilization (Zitko,  1970) and photodecomposition (Ruzo, Zabik and
 Shiretz,  1972;  Safe and  Hutzinger,  1971;  and Hutzinger,  Safe, and Zitko,
 (1972).   In  view  of the  above  difficulties, care must  be  taken when designing
 experimental  protocols and  interpreting data.

      The  information presently available  regarding  soil  attenuation mecha-
 nisms  indicates that PCBs  have a  strong affinity for soil and are somewhat
 degradable by  microorganisms.  There is no evidence that ground water  around
 sites  containing  relatively low levels of PCBs, such as  sanitary landfills,
 has  become contaminated  with PCBs;  however, surface and  ground waters  around
 some  industrial disposal sites and  around  lagoons:  containing relatively
 large  quantities  of PCBs have  become contaminated by leaching PCBs.   The
 mechanisms of  transport  of PCBs in  the biosphere and the mechanisms  of
 attenuation  in soil are  essentially unknown.  Data on the factors affecting
 PCB  attenuation by  earth materials and  the microbial  degradation of PCBs
 would  provide  a useful basis for determining waste treatment methods, for
 predicting PCB migration under landfills, and for selecting and designing
 future disposal sites.

     The  purposes of the present project were (a) to conduct a literature
 review of pertinent information on the attenuation of PCBs in soil materials
 and their behavior  in the environment; (b) to measure the adsorption capa-
 city of selected earth materials for pure PCBs and PCB wastes; (c)  to
 evaluate quantitatively the effects of biological degradation, volatiliza-
 tion, time, organic carbon  content,  and adsorbent structure on attenuation
 of PCBs;  (d) to use this data to develop a mathematical model  that will
 allow prediction of PCB adsorption and mobility; (e) to measure quantita-
 tively the aqueous solubility of PCBs; and (f) to further develop analytical
 procedures that will allow  quantitative measurement of PCBs contained in
aqueous solutions.


                                     10

-------
                                 SECTION 2
                                CONCLUSIONS

     Studies of the aqueous solubility, adsorption, mobility,  microbial
degradation, and volatility of PCBs have led to the following  conclusions:

1.   The rate of dissolution of Aroclor 1242 and a used capacitor fluid
     (Aroclor 1242 impregnated) in water were essentially identical  and
     required 5 months under the experimental conditions employed in this
     study (the rate of dissolution of PCBs probably depends on the  stirring
     rate).

2.   The solubility of individual PCB isomers was greatly reduced in the
     Aroclor mixture where other PCB isomers were present than when  compared
     to their individual solubilities.

3.   The water-soluble fraction of PCBs was richer in the low-chlorinated
     isomers than that in the original PCB fluids.

4.   Used capacitor fluids can be assumed to be the same as Aroclor  1242 in
     terms of the pattern of isomer distribution in the original fluid and
     the water-soluble fractions, in solubility, and presumably in its
     behavior in the environment.  Age and the "burn out" process do not
     appear to alter appreciably the Aroclor 1242.

5.   Some particular isomers in Aroclor 1242 were relatively soluble com-
     pared to their counterparts with similar degrees of chlorination.^
     They were identified as 2-mono-; 2,4'-di-~; 2,5,2'-tri-; 2,3,2'-tri-;
     2,5,4'-tr>; 2,5,2' ,5'-tetra~; 2,4,2' ,5'-tetra-; 2,4,2',4'-tetra-;
     2,3,2'5'-tetras and 2,4,3',4'-tetra-chlorobiphenyl.

6.   The solubilities of Aroclor 1016, 1221, 1242, 1254 and capacitor fluid
     at room temperature were 906 ppb, 3516  ppb, 703 ppb, ^70 ppb, and 698
     ppb, respectively.

7.   A simple linear relation described the  total adsorption of PCBs from
     aqueous solution by earth materials and yielded an adsorption constant
     (K) unique to each adsorbent.

8.   There were no significant differences between adsorption of water-
     soluble Aroclor 1242 and a water-soluble used capacitor fluid
     (Aroclor 1242 impregnated) by any of the five earth materials.
                                     11

-------
 9.    Adsorption  of  PCBs  by  the  five  earth materials; medium-temperature  coal
      char  (me),  high-temperature  coal  char  (ETC),  Catlin  soil  (cs), mont-
      morillonite  clay  (MO),  and Ottawa  silica  sand  (os)  followed  the series

                         MTC > ETC  >  CS  > MC  >  OS.

 10.   Higher  chlorinated  isomers were preferentially adsorbed over  lower
      chlorinated  isomers.

 11.   Very  highly  significant (.001 level) linear correlations were  found
      between K and  total organic carbon (TOC)\ K and C02-surface  area  (SA)\
      and KS  TOO,  and C02-A4.  Poor correlations were found between K and
      \\z-SA;  and K and ethylene  glycol-5.4.  It  was concluded that  PCB mole-
      cules were unable to penetrate  the interlayer  region of montmorillonite1
      to be adsorbed.

 12.   TOO and C02-S4 were the most  important  properties of earth materials
      controlling  PCB adsorption; TOO was the dominant property by a factor
      greater than three.

 13.   Aroclor 1242, Aroclor 1254, and capacitor fluid remained immobile in
      soils when leached  with water or leachate from the Du Page landfill
      on soil thin-layer  chromatography  (TLC) plates, but were highly mobile
      when leached with carbon tetrachloride.  Dicamba showed the reverse
      trend.

 14.   The mobilities of the PCBs and Dicamba  in thin-layers of soil mat-
      erials and silica-gel were highly related to their solubilities in
      the solvent with which the TLC plates were being leached.

 15.  The higher chlorinated Aroclor 1254 was less mobile in soils leached
     with carbon tetrachloride than the lower chlorinated Aroclor 1242.

 16.  There was a highly significant  (.001  level)  correlation between the
     mobility of PCBs in the soil and the soil  TOO content.

 17.  The degradation of water-soluble Aroclor 1242 by mixed cultures of
     soil  microorganisms occurred to  a large extent in a short  period of
     time.   Degradation rates varied  from monochloro isomers that were
     degraded 100 percent within 6  hours to tetrachloro isomers  that
     averaged 42 percent degradation  after 15 days.

 18.   In general,  lower chlorinated  isomers were more easily degraded than
     higher chlorinated isomers..  Position of chlorine substitution
     affected the rate of PCB degradation.

 19.  The predominant organisms found  in  the  mixed cultures  were  Alkaligenes
     odorans} Alkaligenes denitrifieans, and  an unidentified bacterium.

20.  Volatilization  of PCBs  from aqueous solution was  reduced by the
     presence of soil  and the reduction  was  due to preferential  sorption
     onto the soil particulates.


                                     12


-------
21.   Substantial  binding of PCBs by soluble humic substances occurred,.
     especially for the higher chlorinated isomers,  but such binding was
     less effective than that by soils in reducing volatilization from
     aqueous solution.

22.   Volatilization of PCBs from a landfill leachate can be expected to
     be reduced substantially because of the high content of particulates
     and high concentration of humic substances.   Volatilization favored
     the lower chlorinated isomers.

23.   Higher chlorinated PCB isomers were less soluble in water,  preferen-
     tially adsorbed by soil  materials, less mobile  in soil, less degrad-
     able by microorganisms,  and less volatile from  water than lower
     chlorinated isomers.
                                     13

-------
                                  SECTION 3

                              RECOMMENDATIONS


     The results and conclusions formulated from this study generally dealt
specifically with solubility and attenuation of PCBs from pure aqueous
solutions.  The presence of soluble salts and organic materials in water, as
frequently occurs in landfills and the environment, could drastically alter
some predictions of PCB attenuation.

     Soluble salts reduce the aqueous solubility of hydrophobic organic com-
pounds such as PCBs.  The magnitude of the "salting out" effect and its
influence on attenuation of PCBs by earth materials should be determined.
Likewise, the effect of organic solvents and waste streams of various cate-
gories needs to be examined.

     The effects of organic solvents on increasing the mobility of PCBs in
soils were dramatically illustrated during this study; however, the effects
of organic solvents on PCB adsorption remain unknown.  These effects may be
of particular significance where montmorillonite clays are involved.  The
organic solvent may expand the clay layers and allow access to the interlayer
surfaces; this should substantially increase the amounts of PCBs that are
adsorbed as compared to their adsorption from aqueous solution, where PCBs
were apparently restricted to adsorption on the external surfaces only.  The
interrelationship of surface area and organic carbon content of soils needs
further elucidation to improve predictions of PCB migration through soils.

     With regard to  microbial degradation, it is recommended that studies
be carried out on the effects of soil on the degradation process.  Substan-
tial and rapid degradation was observed in aqueous solutions, however,
selective adsorption of PCBs by the soil and competition from native soil
organisms may drastically alter the degradation.  It is anticipated that
PCB degradation rates in soil will be substantially lower than those observed
in solution cultures.

     The studies of volatility of PCBs were limited in scope.  More work is
needed on the effects of soil properties and conditions on volatilization
before accurate predictions could be made of the flux of PCBs in soils due
to volatilization.

     Thus, these multiple interactions of the properties of soil  and the
chemistry of wastes must be considered when making predictions of the envi-
ronmental impact of PCBs.  A major task still before us is the utilization
of quantitative data on the chemistry of soil to make predictions of PCB
                                     14

-------
migration through soil materials.  It is therefore recommended that further
studies involve a cooperative effort between soil chemists, ground-water
hydrologists, and modelers to implement by computer the prediction process
and to identify possible gaps in knowledge that may still bar the successful
prediction of long-term PCB migration through soil materials.
                                     15

-------
       SECTION 4

MATERIALS AND METHODS
                                   Teflon stopcock
                                    (large)
PCB MATERIALS AND  EXTRACTION  PROCEDURES

     The PCBs Aroclor  1016, 1221,  1242, and 1254 were obtained from the Mon-
santo Chemical Company and were  used  without further purification.  The
capacitor fluid was drained directly
from a burned-out  50-KVA  capacitor
manufactured in 1966 and  originally
impregnated with Aroclor  1242.   This
capacitor was supplied by Illinois
Power Company and  was  scheduled  for
disposal in a landfill.   We believe
that this fluid is representative  of
the type of PCB wastes that are  nor-
mally disposed of  in landfills.

     A 20-mL volume of the individual
PCB fluids was placed  in  a 5-gallon
glass carboy, which had been  cleaned
and baked at 450°  C  overnight.  The
carboy was filled  with water  that
had been passed through a mixed  bed
exchange resin  and an activated car-
bon bed and was finally distilled  in
glass.  The carboy was then sealed
with an aluminum-wrapped  rubber  stop-
per and was wrapped with  aluminum
foil and masking tape  to  protect it
from exposure to light.   The  solution
in the carboy was  agitated gently  with
a magnetic stirring bar coated with
Teflon.  Gentle agitation was used to
avoid breaking the PCB fluids into a
fine suspension.   Water samples  were
collected from the carboys at regular
intervals, and 100 mL  aliquots of  the
PCB-water solution were filtered
through glass wool plugs  tightly
packed into the stems  of  glass fun-
nels.  The first IpO-mL aliquot  was     Figure2  Modifjed Er|enmeyerflaskforextractjonof pcBs
used to saturate the PCB  adsorption            from aqueous sample
          16

-------
capacity of the glass wool and was discarded.  Subsequent  aliquots  were
then used in the studies.  The filtrates were then  extracted  with  three
5-mL portions of either carbon tetrachloride or  hexane  (distilled  in glass
grade, Burdick and Jackson Laboratories, Inc., Muskegon, MI)  in  a  specially
designed extractor, which is shown in figure 2.   Samples to be analyzed
by the perchlorination'procedure were extracted  with  carbon tetrachloride.
The samples to be analyzed by capillary column techniques  were extracted with
hexane.  The extracts were passed through a florisil  column containing
approximately 75 percent florisil (Floridin Co.,  Tallahassee, FL)  and 25
percent Na2S(V  This was done to remove any water  or polar materials that
may have been co-extracted from the samples (Reinert, 1970; Curley et al.,
1975). (The application of florisil, silica gel,  and  alumina  to  clean up
other contaminants from PCBs has been used successfully in many  studies
[Reynolds, 1969; Snyder and Reinert, 1971; Berg,  Diosady,  and Reese, 1972;
and Ahnoff and Josefsson, 1975].)

     The column was 5 mm I.D. x 50 mm long and was  first flushed with sol-
vent.  The sample was then passed through the column  and collected  in a 10
ml volumetric flask.  The column was further eluted with at least  one bed
volume of solvent which was subsequently added to the flask until  exactly
       2.49
         I
              IS
              Figure 3.  A GC chromatogram of internal standard (tribromobenzene) in hexane
                                      17

-------
                    IS
                          3     8
                                                 18
                                                           24
 Figure 4.  A chromatogram of internal standard (IS) and four different PCB isomers;4-mono- (peak 3); 2,4'-di-(peak 8); 2,5,4'-
        tri- (peak 18); and 2,4,2',4'-tetra-chlorobiphenyl (peak 24)
10 mL had  been  collected.  The hexane filtrates from the  florisil columns
were then  spiked with tribromobenzene as an internal standard for quanti-
tative analysis.  The possible interferences from the  solvents and internal
standard were checked.  Figure 3  is  a chromatogram of  a hexane solution con-
taining a  high  concentration of internal standard.  The peaks in the chro-
matogram illustrate some of the impurities present in  the reagents that do
not normally show up on chromatograms of samples or a  standard (Fig. 4).
Figure 4 is a chromatogram of the  internal standard and four PCB isomers.
The major  impurity, shown in Figure  3 as a peak at retention time 2.49 min.,
elutes before any PCB peaks and therefore does not interfere with the
analyses.

     Aqueous samples from adsorption studies with earth materials were
extracted  with  20 ml of either carbon tetrachloride or hexane in three
                                       18

-------
aliquots of 10, 5, and 5 mL each.  The first extract was then placed in a
10 ml Kuderna-Danish (K-D) evaporator (Kontes Glass Co., Vineyard, NO) and
was concentrated to approximately 1 mL in a 100°C water bath by using NaCl
crystals as boiling chips.  The second extract was added to the K-D and the
concentration process was repeated.  The third extract was combined with the
condensed sample to give a final volume of about 6 ml.  The 6 ml of condensed
sample was then passed through a florisil column as described above and was
analyzed for PCBs.

PCB Materials Used in Mobility Study

     The PCBs and Aroclors 1242 and 1254 were obtained from the Monsanto
Chemical Company (St. Louis, MO) and were used without further purification.
The 14C labeled compounds were obtained from New England Nuclear Corporation
(Boston, MA).  Gas chromatographic (GC) traces of the llfC labeled compounds
were identical to those of the Aroclor 1242 and 1254, therefore it was
assumed that there were no significant differences in the respective com-
pounds and that the 11+C labeled PCBs and Aroclors would behave similarly in
studies of mobility.  The specific activities of Aroclor 1242 and 1254
were 0.119 m Ci/mg and 0.096 m Ci/mg, respectively.

     The used capacitor fluid described previously was also used in this
study.

EARTH MATERIALS

Adsorbents

     The adsorbents used  in the PCB adsorption study were  Ottawa  sand (OS),
montmorillonite clay (MC), Catlin silt loam soil (CS), Illinois No. 6 coal
charred at 1800°F (HTC),  and the same coal charred at 1200°F (MTC).  The
two chars were soaked and rinsed with distilled-deionized water and then
oven dried at 110°C overnight.  The coal chars are impure forms of activated
charcoal and were selected for study because of their potentially high
adsorption capacity for PCBs and their possible use as liners for disposal
sites accepting organic wastes.  The sand, clay, and soil represented the
major soil materials and  offered a range of chemical and physical character-
istics.  These are presented in Tables 1 and 2.  The chemical and physical
characterization was performed as described by Griffin and Shimp (1978).

     The organic matter content of the adsorbents was varied by using the
low-temperature ashing (LTA) technique for bituminous coals described by
Gluskoter (1965) and Kuhn, Fiene, and Harvey 0978),  An aliquot of each
sample was ashed except for Catlin soil for which three aliquots were ashed,
each for different time intervals.  The changes in organic matter content
and surface area that are due to ashing are shown in Table 1.

     Surface areas were measured by three methods, each selected to provide
different information about the surface characteristics of the adsorbents.
The BET gas adsorption method was used as described by Thomas and Frost
(1971).  Both N2 and C02 were used as adsorbates.  The third method was
ethylene glycol adsorption as described by Bower and Gschwend (1952).

                                     19

-------
TABLE 1.  Characteristics of earth materials used in adsorption studies.
Surface area

Adsorbent

PH
Ottawa silica sand (OS) 5.2
Montmorillonite clay (MC) 7.0
Montmorillonite clay (LTA)
Catlin silt loam AP
Catlin, 6 hr. LTA
Catlin, 1 2 hr. LTA
(CS) 7.1


Catlin, 336 hr. LTA
Coal char (1200°F)
Coal char (1200°F)
Coal char (1800°F)
Coal char (1800°F)


Constituent
AI203
CaO
Cl
Fe203
P20s
K2O
Si02
TiO2
MgO
Na20
S

Cl
V205
MnO
Na2O
Sn
Cu
Co
Ni
Be
Cr
Mo
Zr
Pb
Zn
F
(MTC) 7.3
(LTA)
(HTC) 7.0
(LTA)
TABLE 2.
High temp.
char
3.77
0.59
0.01
3.04
< 0.01
0.41
12.93
0.25
< 0.01
0.15
2.89
(ppm)

30
86
-
<50
117
< 2.4
19
1.6
26
6.8
60
< 2
36
125

CEC N,
(meq/100g) (m /g)
0.0 0.5
85.0 23.2
23.9
18.1 10.1
12.6
11.4
16.3
9.6 1.6
2.1
1 .7 3.4
10.0
Ethylene
CO2 glycol
(m2/g) (m /g)
0.4 <1
20.1 885
20.2
26.5 68
25.4
24.5
23.8
253 55
214
44 68
120

Sand Silt
100 0
0 0

1 1 .6 60.9







Organic
Clay carbon
0 ^.01
100 0.93
0.13
27.2 4.73
4.37
3.64
1.84
74.04
64.00
76.62
32.14
Chemical characteristics of earth materials used in adsorption studies.
Med. temp.
char
2.93
0.68
0.04
2.52
0.01
0.29
9.36
0.19
< 0.01
0.12
3.28
(ppm)
_
31
54
-
<50
45
< 2.4
8.6
1.3
24
6.3
22
< 2
23
95
Catlin
Ap soil
9.63
0.98
-
4.07
0.09
1.90
69.50
0.74
0.01
0.68
0.06
(ppm)
72
64
543
-
<50
45
9.6
22
3.1
62
4.0
237
16
72
160
Montmorillonite
clay
13.32
5.05
.
2.73
0.05
0.34
59.98
0.23
3.48

0.02
(ppm)

19
30
695
< 50
43
< 2.4
6.8
1.4
27
< 3.4
98
8.2
19
435
Ottawa
silica sand
0.19
0.07
.
0.09
< 0.01
0.01
97.38
0.01
< 0.01
_
0.01
(ppm)
38
< 9
< 2
27
< 50
4.3
< 2.4
< 1
< 1
< 1
< 3.4
80
< 2
< 1.6
< 10
                               20

-------
Soil Materials

     Soil materials representing  a  wide range in characteristics were
selected for use in the mobility  study.  These included a pure silica sand,
seven soils, and medium-temperature (1200°F)  coal  char.  The materials and
some of their chemical and  physical  characteristics are listed in Table 3.
All the soil materials were ground  in  a mortar with pestle and screened
(28 mesh) before use.

ADSORPTION STUDIES

     All glassware was cleaned  with tap water, rinsed with distilled-
deionized water, and  baked  in an  oven  at 450°C for at least 8 hours.  Weighed
amounts of adsorbent  were placed  into  125 mL  brown serium bottles.  An
aliquot of PCB-saturated water  was  taken from the 5-gallon carboy described
earlier and filtered  through glass-wool plugs that were packed tightly into
the stems of glass funnels.  A  60-mL volume of filtrate was then added to
the reaction bottle.  The bottle  was covered  with a sheet of baked aluminum
foil and sealed by a  teflon-coated  septum and a crimp-cap (Wheaton Labora-
tory Products, Millville, NJ, and Pierce Chemical  Co., Rockford, IL).  The
                  TABLE 3. Characteristics of soil materials used in PCB mobility studies.
                                      Surface area

Material
Ava
Silty clay loam
Bloomfield
Loamy sand
Catlin
Silt loam
Catlin
Loam
Cisne
Silt loam
Coal char
(1200°F)
Drummer
Silty clay loam
Flanagan
Silty clay loam
Ottawa
Silica sand

CEC
pH (meq/100g)

4.5 13.1

5.7 0.8

7.1 18.1

8.7 4.7

5.9 8.0
7.3 9.6

6.7 29.4

6.6 23.3

5.2 0.0

N2 C02
(m2/g) (m2/g)

28.3 28.9

1.7 2.2

10.4 26.5

1 1 .5 1 1 .2

6.1 13.0
1 .60 253

22.1 29.1

12.6 32.8

0.5 0.4
Ethylene Organic
glycol Sand Silt Clay carbon
(m2/g) (%) (%) (%) (%)

55 2 69.6 28.4 1.18

2.0 82 10 8 0.21

68 11.6 60.9 27.2 4.73

17 35.9 43.5 18.5 0.57

23 13.8 70.8 14.8 1.30
55 ... 74.04

103 17.9 49.5 31.6 2.17

93 5.4 65.2 29.4 2.62

<1 100 0 0 <0.01
21

-------
 brown bottles  protected the  PCB solutions  from photodecomposition during
 the adsorption process  and the crimp seals prevented volatilization losses
 during the reaction  period.   The bottles were shaken overnight in a water
 bath at 25 ± 0.5°C.

      The bottles  were removed from  the  shaker and  placed  directly into a
 Model  JS-7.5 rotor and  centrifuged  at a constant temperature  of 25 ± 1°C
 in  a Beckman Model J-21B centrifuge for 5  minutes  at 6000 rptn.   They were
 then removed from the centrifuge head,  the seals were broken,  and 40 ml of
 the clear supernatants  were  pipetted for PCB  analysis.

      The effect of reaction-time on PCB adsorption was  studied  by using the
 methods described above.  Two series of samples were prepared,  each contain-
 ing water saturated  with Aroclor 1242 and  a fixed  amount  of montmorillonite
 clay.   The two dosages  of clay were 167 mg/L  and 1667 mg/L.   The  bottles
 were removed from the shaker after  various time intervals and  analyzed for
 their PCB content.   The results  showed  that a shaking time of  8 hours  was
 sufficient to  reach  equilibrium  conditions.   Therefore, all samples were
 shaken  overnight  to  ensure equilibration and  for convenience  of analysis.
 Blanks  were  carried  through  all  experiments to determine  the  background
 levels  of contamination of the  adsorbents  with PCBs  and to evaluate the
 adsorption of  PCBs onto the  surfaces  of the glassware.

      The  amount of adsorption was determined  from the difference  between the
 initial  concentration and the  equilibrium  concentration multiplied  by  the
 volume  of solution.  A  blank was  subtracted and the  amount adsorbed by each
 sample  was  computed  on  a  unit  basis  by  dividing by the dry weight of the
 adsorbent.

 PCB MOBILITY STUDIES

     The  mobilities  of  Aroclor  1242,  Aroclor  1254, and Dicamba  were studied
 using the  soil  TLC technique described  by  Helling  (1971a).  Dicamba, a  pesti-
 cide of known  high mobility, was used as an internal  standard.  The  adsorb-
 ent was  slurried  with water  until moderately  fluid,  and was applied  with  a
 spreader  to  a  clean  glass plate  (20 cm  x 20 cm) that  had  been washed with
 ethanol and  acetone.   The adsorbent was spread to a  thickness of  0.5 mm
 and  then  air dried.  A  horizontal line was  scribed with a  stainless  steel
 spatula at 12  cm  above  the base to  stop solvent movement;  vertical  lines  were
 scribed 2  cm apart to separate the  various  treatments.  The radioactive
 compounds were  spotted  on 2  cm from the base  and leached  10 cm with  a devel-
 oping solvent.  The activity of the  lkC labeled PCB  that was spotted was
 22,000 d.p.m.  (84 ng for  1242, 104 ng for  1254).  The plate was immersed  in
 0.5 cm of  solvent in a  closed glass chamber and was  removed when  the wetting
 front reached  the horizontal  line.  Leaching was thus ascending chromato-
 graphy.  The developed  plate was then removed and air dried.   A piece of
 8 x 10  inch  medical x-ray film was placed  in direct contact with  the devel-
 oped plate for a  period of 2 weeks.   The resulting autoradiograph indicated
 the relative movement of  the compound, which was measured as the  frontal
Rf  of the spot or streak.  The Rf value is defined as the ratio of the dis-
 tance the compound moved  relative to the distance the solvent moved.  The
Rf  value  is  a  quantitative indication of the front of PCB movement and  a

                                     22

-------
reproducible index of mobility.
     The soil  TLC plates were developed with three leaching solvents;
distilled-deionized water, Du Page leachate, and carbon tetrachloride.
Distilled water was further purified by passage through a Milli-Q (Millipore
Corp.) reagent grade (> 10 megohm resistance) water system.  The organic
solvents were glass-distilled from Burdick and Jackson (Muskegon, MI) and
the leachate was collected from the Du Page County, Illinois sanitary land-
fill (well MM-63).  The site description and well location were described
by Hughes, Landon, and Farvolden (1971) and the chemical  characterization
of the Du Page leachate by Griffin and Shimp (1978).  The three leaching
solvents were each extracted with hexane and analyzed for trace contamina-
tion of PCBs by using standard GC techniques and were found to be free
from contamination.

     The mobilities of the two Aroclors and Dicamba were also measured on
silica-gel  (GF-254, type 60, Brinkmann Instr.,  NY)TLC plates leached with
a variety of solvents.  The solvents used were deionized H20; Du Page leach-
ate; carbon tetrachloride; benzene; acetone; 80:20 water:acetone; methanol;
15:85 water:methanol; and 9:91 water:methanol.

     To confirm the data obtained from autoradiographs, zonal extractions
were done on several lanes spotted with PCB.  A given lane was divided into
12 equal segments starting from 1.5 cm below the origin to 10.5 cm above
the origin.  The soil in each segment was carefully transferred to a clean
graduated centrifuge tube to which 0.5 ml of 1:1 C2H5OH:H20 was added; 4 ml
of nanograde hexane was used as the extracting solvent.  The ethanol mixed
with water  at the soil surface and provided a more potar phase than the
hexane whereby the PCB in the soil could be more effectively extracted into
the hexane  layer.  The soil suspensions were sealed with aluminum foil,
mixed with  a vortex mixer for a few seconds, and left to stand for an hour.
This procedure was repeated three times to allow complete extraction of
PCB from the soil, and a recovery of 99.7 ± 1.5 percent of PCB was obtained.
The soil suspension was then centrifuged at 2500 rpm for 5 minutes.  The
final volume of hexane was recorded.  One ml of the radioactive PCB-hexane
solution was pipetted into a scintillation vial containing 10 ml Dioxane
scintillation fluid  (7 g PPO, 0.05 g POPOP; 120 g Naphthalene in 1 L
Dioxane), and the activity was counted for 10 minutes in a Packard Tri-Carb
Liquid Scintillation Spectrometer and the results corrected for quench.   In
the case of nonradioactive PCBs  (capicitor fluid), the hexane extract was
analyzed for PCB  by standard GC techniques.  The conditions for the gas
chromatographic analysis were as follows:

     Instrument:   Varian 2100 series  Sc3H electron capture detector
     Column:  6 ft glass column   2.5% OV-210 and 1.5% OV-17 on 60/80
              Supelcoport G.C.  Bondx

     Injection port temperature:   220°C

     Column temperature:           200°C
     Detector temperature         250°C

     Carrier gas:                  N2, flow 15-20 mL/min
     Electrometer:                range 10~lc amp/mv, attenuation 16

                                     23

-------
 MICROBIAL  DEGRADATION

 Isolation  of Mixed Cultures

     Three mixed cultures of microorganisms were  isolated by biphenyl
 enrichment from three different soil and  sediments.  To isolate the mixed
 cultures,  5 g of soil or sediment and 0.1 g of biphenyl were placed in 100
 mL of the  following mineral media (Gray and Thornton, 1928):
      KN03  1.0 g/L

      MgS04 0.2 g/L

      CaCl2 0.1 g/L
FeCl3 0.02 g/L

NaCl  0.1 g/L

K2HP04 1.0 g/L
The distilled water used to make up this media was saturated with Aroclor
1242.

     The mixture was placed in a glass-stoppered 250 mL Erlenmeyer flask
that was wrapped completely with aluminum foil to keep out light and shaken
at room temperature (26°C).  After 1 week, 5 mL of solution was placed in a
new flask containing 0.1 g of biphenyl and 100 mL of fresh media.  This pro-
cedure was continued until the solution turned yellow.  The yellow product,
the biphenyl metabolite described by Furukawa and Matsumura (1976), indi-
cated that biphenyl was being degraded.  The solution was then transferred
every 3 days to fresh media.

     Three different mixed cultures were isolated in this manner.  The first,
the C-line, was isolated from dried Catlin soil.  The second, the S-line,
was isolated from soil taken from an experimental farm plot at El wood,
Illinois.  Activated sludge containing PCBs from a municipal waste treatment
plant at Stickney, Illinois, was applied to this plot for 7 years.  The
third culture, the N-line, was isolated from Hudson River sediment from
the Fort Miller disposal site in New York (Leis et al., 1978).
ment contained about 2 ppm PCBs.

PCB Degradation Studies
                     The sedi-
     Originally the PCB degradation study was run on the C-line for 15, 30,
and 45 days.  Later studies were run on all three cultures for 5, 10, and
15 days.  For these experiments, 70 mL of a 3-day old culture were added
to 1400 mL of mineral media made with Aroclor 1242 saturated water.  The
solution was stirred rapidly.  Immediately, six 100-mL aliquots were removed
and placed in 250-mL Erlenmeyer flasks, with ground glass stoppers, to
which 10 drops of concentrated HC1  had been added.  The pH of each of these
control samples was about 2.  Six more 100-mL aliquots were placed in
Erlenmeyer flasks as the test samples.  All flasks were completely covered
with aluminum foil and shaken at 27°C ± 2°C.  Two 80-mL aliquots of the
original solution were taken, acidified, and extracted as 0-day, or 0-hour,
blanks.
                                     24

-------
     After a set period of time (depending upon the experiment) 2 control and
2 test flasks were removed from the shaker.  Next, 10 drops of concentrated
HC1 were added to the test flask to stop degradation and to maintain the
same pH for extraction for both the control and test samples.  Eighty ml
of solution from each flask were extracted with 10 ml of hexane.  The solu-
tions were extracted in a modified Erlenmeyer flask as shown in Figure 2.
The solution with hexane was stirred for 10 minutes; then the flask was
inverted and the layers were allowed to separate.  The water layer was
drained off into a beaker, and the hexane was drained into a 25-mL brown
Wheaton bottle (Pierce, Rockford, IL) with a crimp top.  The water was put
back into the extractor and 5 ml of hexane were used to rinse out the beaker.
The hexane was then poured into the extractor and the mixture was_stirred
for 5 minutes.  The hexane was removed as before, and 5 ml of additional
hexane were used to extract the solution.  The Wheaton bottle  containing
the 20 ml of hexane was capped with an aluminum foil covering over a teflon
cap and was stored until analysis.  This extraction procedure was carried
out for all samples.

Soil and Sludge Extractions

     The soil and sludge from the experimental farm plot at  Elwood,  Illinois,
were extracted to determine their PCB content.  The sludge was a composite
of sludge that had been applied to the soil about 3 months earlier.  The
activated sludge had been heated and anaerobically digested  at a municipal
waste treatment plant  at Stickney, Illinois.  The soil was a composite of
soil from 0 to 6 inches deep and was from  a plot that had received a total
of 225 tons of dry solids per acre during  the previous 7 years.  This amounts
to about eight 1-inch  liquid applications  per year.  Applications of sludge
were from June through August.  The soil was tilled from 5 to  6  inches deep.

     The sample soil was dried at 103°C and stored at room temperature prior
to extraction.  The  sludge was kept refrigerated.

     To extract the  soil,  50 g of finely  ground  soil were  placed in  a Soxhlet
 extractor.   About 150  ml of hexane distilled  in  glass were added to  a 250-mL
round bottom  flask,  and the soil was extracted for 48 hours.   The hexane-
cycle time was about 5 minutes.  After extraction was complete,  the  extract
was  treated  to remove  interfering organics  by  adding baked florisil  (450°C
overnight)  in increments to the  round bottom  flask until the solution turned
colorless.  The hexane was then  evaporated to  approximately  0.5  ml  in a
micro K-D evaporator and injected on the  capillary column  for  PCB analysis.

     About  150 ml of sludge were added to  a modified Erlenmeyer  extractor
 (Figure 2)  along with  50 ml of hexane.  This  solution was  stirred for about_
 5 minutes.   The resulting  solution was centrifuged  (Sorvall  table top centri-
 fuge,  Ivan  Sorvall,  Inc. Norwalk, CT) at  1,000 rpm  for  5 minutes to  break  up
 the  sludge-hexane emulsion.  The sludge was  then  extracted again with 50 ml
 of hexane.   Again the  solution was centrifuged to  break  the  emulsion.   The
 resulting extract,  colored deep  yellow, was  cleaned  by  the addition  of  flor-
 isil, as with the soil.  The hexane  then  was  evaporated  to approximately 0.5
ml in a micro K-D evaporator and  injected  on  the  capillary column.
                                     25

-------
 Som'fication Study

      To determine if  the  extraction procedure used to extract  the PCBs from
 the  test solutions could  extract all the  PCBs that were in the microorganisms,
 a  2-day degradation sample was sonified to  disintegrate the microorganisms
 and  then was extracted.   This sonified sample extract was compared to an
 extract done as described previously.

      The sonifier used was a Bronson Sonic  Power (Danbury, CT) sonifier,
 model  number S75 with an  exponential horn tip.   The instrument was operated
 at the  7 level  at 6 to 7  amps for 30 minutes.

      Prior to sonification,  the tip and the  500-mL stainless steel  beaker
 were  rinsed with methanol  and hexane.  After they dried, 240 ml  of a 2-day
 old S-line culture were placed in the beaker.   The beaker was placed in an
 ice bath and allowed to cool  for 15 minutes.   Then the sample was sonified
 for 30  minutes.  Finally,  two 100-rnL aliquots  were acidified with 10 drops of
 concentrated HC1 each and  were extracted by  the  usual  procedure.   In addition
 to the  sonified sample, two  100-mL aliquots  of 2-day S-line  culture were
 acidified and extracted in the normal procedure  and a  2-day blank was run
 and extracted.

 PCB VOLATILITY  MEASUREMENTS

      Experiments were set  up  to compare the  changes in the volatility of
 PCBs  between aqueous solution of PCBs and PCB  solutions containing  different
 amounts  of soluble humic  substances.  The effect of different amounts of
 soil   in  suspension with PCB-saturated water  on the volatilization  of PCBs
was also studied.

     Humic acid standards  were achieved by dissolving  100 mg and  1,000 mg of
commercial  humic acid (Aldrich Chemical  Co., Milwaukee, WI) in 1  liter of
distilled  water and  by adjusting the pH to 12 with NaOH pellets.   The soil
used  in  the  study was the  Ap  horizon of an agricultural  soil, Cat!in  silt
loam.  The properties of the  soil  are listed in  Tables 1  and 2.

                               50 mL PCB—saturated water
 Mixed with 50 mL
 pure water
Mixed with 50 mL water
containing 100 ppm or
1000 ppm humic acid,
equilibrated overnight
                         Purged with IM2 at room temperature, flow
                         rate 200 mL/min for 20 minutes, trap with
                         15 mL hexane in 3 stages
Mixed with 50 mL pure
water, add 0.1 g, 0.5 g
or 0.64 g of Catlin
soil, equilibrated
overnight
                         Hexane evaporated and analyzed by GC-FID


      Figure 5.  Flow diagram of experimental protocols used to study PCB volatilization from aqueous solutions
                                      26

-------
        TABLE 4.  Gas chromatographic conditions used during analyses of RGBs for volatilization study.



   Column: 183cm by 2 mm in diameter glass column packed with 3 percent OV-17 on 100/120-mesh Gas Chrom Q


   Carrier: He,at 40 mL/min; H2:3Q mL/min;and air:250 mL/min


   Injection port: 300°C


   Detector: 300°C


   Temperature programmed from 170°Cto 290 Cat 8 /min


   Attenuator: 16 x 10  amps/mv	—
     The effects of  humic  acid and soil on volatilization of the PCBs were
investigated using the  experimental protocols shown  in  Figure 5.  Triplicates
of each of the samples  described in Figure 5 were  analyzed to obtain repre-
sentative results.

     PCBs dissolved  in  water were purged at room temperature by a stream of
clean nitrogen, and  the purged sample was trapped  in a  three-stage liquid
trap filled with 5 ml of hexane in each trap  (Fig.  6). The nitrogen flow
rate was 200 mL/min.   After 20 minutes, the hexane was  pooled and evaporated
in a Kuderna-Danish  (KD) evaporator to 0.4 ml,  then  was further evaporated
to dryness in a 1-mL vial  by a stream of nitrogen.  Five yL of hexane was
then pipetted into the  vial, and 1 yL of this concentrate was injected onto
a gas-chromatographic column (183 cm by 2 mm  I.D.  glass column packed with
3 percent OV-17 on Gas  Chrom Q) 100/120 mesh  and was detected by a flame
ionization detector  (FID)  on a Hewlett Packard  5830A gas chromatograph
equipped with an automatic integrating recorder.   The gas chromatographic
conditions are given in Table 4.

     The use of an electron-capture detector  (ECD) was  investigated earlier.
The gas-purging process described was found to  carry water vapor into the
hexane; the concentrate when injected resulted  in  an erratic response from
the ECD.  An effort  to dry the hexane layer by  using salts was found to
interfere with quantisation of the PCBs.  Although ECD is highly sensitive
to halogenated compounds and thus would normally be  used for PCB analyses,
FID was used in these experiments.  The sensitivity  of the electrometer of
FID was 16 x 10~12 amps/mv, which was approaching  its sensitivity limit
of 1 x 10~12 amps/mv.

     The GC calibration was accomplished by repeating injections of 1 yL of
500 ppm Aroclor 1242 in hexane onto the GC column  and evaluating the result-
ing chromatograms.   The area integration of the chromatograms then provided
a reference for quantitation of the amounts of  PCBs  contained in the samples
from the hexane traps.

     The trapping  efficiency of the volatilized PCBs in the  hexane traps
(Fig. 6) was evaluated by mixing  15 mL of  hexane with 1 yL of 500 ppm Aro-
clor 1242.   It was  then equally distributed in  the three liquid traps and


                                      27

-------
 stripped by using the identical  purging
 gas-flow rate and time.   The trapping
 efficiency was found to  be 71  percent.

      The purging efficiency was  evalu-
 ated  by  adding 1  ml  of acetone solu-
 tion  containing 0.5  yg of Aroclor 1242
 to  a  purging flask containing  100 ml of
 water (Fig.  6).   This solution was
 purged,  and the contents of the  hexane
 traps  were analyzed  for  PCBs.  The
 resulting  chromatogram was then  used
 to  evaluate the purging  efficiency.

 ANALYTICAL METHODOLOGY

 Gas-Liquid Chromatography
 Instrumental  Parameters

    Apparatus  included a Hewlett
 Packard  (HP) model 5830A (Avondale,
 PA) gas  chromatograph equipped with
 a linearized  63Ni  electron capture
 detector and an  automatic integrator.
 A 183  cm x  2 mm  I.D.  glass column
 packed with  3  percent OV-17  on 100/120
 mesh Gas Chrom. Q  (Analabs Industries,
 North  Haven, CT) was  used for  the
 analysis of  samples  using the  per-
 chlorination procedure.   Helium served
 as a carrier gas at  a  flow rate of
 60 mL/min  under operating temperatures
 for injector,  column,  and  detector at
 300°C, 230°C,  and  300°C  respectively.
                       100 ml saturated
                       PCB solution
Figure 6.  Purging system using 1000-mL Erienmeyer flask
     Samples from hexane extraction were analyzed  in a 40-m wall-coated  open
tubular (WCOT) glass capillary column.  The liquid phase,  Emulphor  ON-870
(Applied Science Lab., State College, PA), was custom-coated  on  the glass
capillary columns using the techniques of Grob (1975) and  Grob,  Grob,  and
Grob (1977).  The Emulphor ON-870 gave better resolution of the  various
isomers than other commercially available capillary-column coating  materials.
Helium and 5-percent methane in argon were the carrier and make-up  gases,
respectively, the linear velocity of the carrier gas being 45 cm/sec.   The
samples were split at a ratio varying from 1:90 to 1:150.  The operating
temperature for injector, column,- and detector were 190°C, 175°C, and  320°C.
All of the results were integrated and recorded on an automatic  integrator.

Capillary Column Analysis

     The capillary columns would sometimes become unstable with  use and
therefore had to be recoated several times.  Varying degrees  of  success
                                     28

-------
ro
to
                                                 Figure 7.  Water- soluble Aroclor 1 242 using a temperature-programmed column

-------
were obtained for each recoating.  As a consequence, the resolution,  reten-
tion times, and column temperatures varied slightly with different columns.

     For the early work on the microbial degradation using the C-line,  the
carrier gas was N2 with a linear velocity of 25 cm/sec.  The column was
programmed from 120°C to 220°C at 20°/min.  Figure 7 shows a chromatogram
of water soluble Aroclor 1242 using the above column.  The number of  theo-

retical plates (w   7) was calculated as follows  (Kaiser, 1976): N    -,=
       I r> j-m      jP&u. Is                                             V&Q t,
5.54 (-g—=£--) where tm = unretained retention time (solvent), and 2>10 and bQ

are the 1/2 height widths at K = 10 on a plot of  1/2 height widths vs. K.

(K.=^-^	where tr = retention time of a certain peak).  For a series  of
similar compounds, such as PCB isomers, N   - = 35,000 plates.

     For the 5-, 10-, and 15-day microbial degradation studies using  the C-,
N-, and S-lines the conditions were as follows:  the carrier gas was He with
a linear velocity of 45 cm/sec.  The column temperature was 192°C and the
split  ratio was 1:150. ®Teal  = 53,000 plates.  Figure 8 shows a chromato-
gram of water-soluble Aroclor 1242 using this column.

     For the microbial degradation kinetic study  using the  N-line, the con-
ditions were as follows:  the carrier gas was N2 with a linear velocity of
45 cm/sec.  The column temperature was 173°C and  the split ratio was  1:90.
Nreal  = 33,000 plates.  Figure 9 shows a chromatogram of water-soluble
Aroclor 1242 using this column, which was also used in the water solubility
and adsorption studies.

     Prior to PCB analysis  using the column described above, a mass spectrum
of Aroclor 1016 was run using an Emulphor-coated WCOT capillary column.
Figure  10 shows the reconstructed chromatogram.   The bottom line is for mass
number  188 or monochlorobiphenyl; the next .line, for mass number 222 or di-
chlorobiphenyl; the third line,  for trichlorobiphenyl;  and the second to the
top line,  for tetrachlorobipheny1.   The  top line on  the figure registers the
total  reconstructed chromatogram.

Quantification and Identification

     Samples analyzed using the capillary column were spiked with tribromo-
benzene as an internal standard.  The quantitation measurements  were carried
out by obtaining the electron capture detector response factor for both the
internal standard and four representative PCB isomers.  The four PCB  isomers
used were 4-mono;. 2, 4'-di-; 2, 5, 4'-tri-; and 2, 4, 2', 4'-tetra-
chl orobi phenyl .  The concentration of stock solution of PCB isomers were
1,502  ppm, 99.98 ppm, 60.04 ppm, and 50.17 ppm, respectively.  The concen-
tration of the stock solution of internal standard, tribromobenzene,  was
24,900 ppm.  A combination standard solution was prepared from these  stock
                                     30

-------
                      3 3 3+23 333   444344
Figure 8.  Water-soluble Aroclor 1242 on second capillary column. Number under peak indicates the number of chlorines on the
       isomer in each peak. Top number is the peak number. Peak number 1 is the internal standard.
solutions by adding  TOO yL of each to  4 ml of hexane,  which were subsequently
condensed to approximately 0.5 ml in a micro Kuderna-Danish (K-D) concen-
trator.   Figure 4 shows a representative chromatogram  of internal standard
and  the  four PCB isomers.

     To  identify each  individual isomer, two approaches  were adapted.
(1)  Seven isomers (which included the  four isomers used  for quantitation and
the  three selected isomers of known structure  for additional  retention-time
data) were injected  into the GC.  By comparing the retention time of each
isomer with the results from the mass  spectrometer, the  seven  particular
                                       31

-------
          1   12 2  22  3 3332 133  3  333444  4 34444
                          2*3
4 44  4 4   4344  34     44
                   Figure 9.  Water-soluble Aroclor 1242 on third capillary column
isomers could be identified.   (2)  By matching the relative position of  the
other isomers to the  isomer distribution pattern reported by Sissons and
Welti (1971) and Webb and  McCall  (1972), the major water-soluble PCB isomers
were identified.

Perchlorination Procedure  for  Analysis of PCB

     The perch!orination of PCBs  to  decachlorobiphenyl (Armour, 1973) was
carried out in 10 mm x  150 mm  hydrolysis tubes (Pierce Chemical Co., Rock-
ford, IL), which had been  modified by adding a 15 mm x 150 mm test tube to
increase the volume (Fig.  11).  Also, the cap of the reaction tube was  modi-
fied so that the teflon insert could be easily removed for cleaning.
Thorough cleaning of the insert between uses was necessary in order to  pre-
vent contamination of succeeding  runs.

     The teflon inserts were cleaned by boiling in a chromic acid solution
for two hours (the chromic acid solution was prepared by mixing 10 to 15 mg
of K2Cr207 in 15 mL of water and  adding sulfuric acid until the resulting
                                      32

-------
co
GO
                                                                                                         18
                                                                                                                           24
                                                                                                                                                    32
                                                                                                                                              30  31  33   34
                                            Figure 10.   A reconstructed chromatogram of Aroclor 1016obtained from mass-spectrometer

-------
red mass dissolved).  The  inserts were  removed  from  the  chromic  acid solution
and rinsed successively with tap water, distilled  water,  methanol,  and carbon
tetrachloride.

     An aliquot of a known volume of a  solution of carbon tetrachloride con-
taining PCBs was placed in a reaction vessel  as described above.  Antimony
pentachloride  (0.5 to 1.0 ml) was added to  the  solution.   Sodium chloride
crystals were  added as a boiling aid.  The  reaction  vessels  were capped se-
curely to prevent losses of PCBs by volatilization.  The  reaction was  carried
out at 200°C ± 5°C for 16  hours.  Preliminary studies  had shown  that 16 hours
at 200°C were  sufficient for complete perchlorination  of  the sample.

     The reaction vessel was allowed to cool  and the caps were loosened.
Then 4 to 5 drops of 6N HC1 were added  slowly (due to  the violent reaction).
To quench the  reaction and to prevent precipitation  of antimony  oxychloride
(Berg, Diosady, and Rees,  1972), 1 ml of HC1  was added.   To  ensure  that the
HC1 was well mixed, the caps were closed tightly and the  vessels  were  shaken
for about 10 seconds.  The caps were then opened and 1.0  ml  of Mi rex (per-
chloropentacyclodecane, Alfred Bader Library  of Rare Chemicals,  a Division of
Aldrich, Inc., Milwaukee, WI) was added (at a concentration  of 500  ppb in a
solution with  benzene), followed by an additional  3 mL of benzene.   The caps
were again closed tightly and the vessels shaken vigorously  for  1 minute.  To
remove any impurities and/or moisture (Armour,  1973),  the upper  benzene layer
was removed with a long-nosed transfer pipette  to  test tubes containing
                                       Na2S04 and  NaHC03.  The benzene solu-
                                       tion was transfered from  the test
                                       tube to  a K-D evaporator,  and NaCl
                                       crystals were added as a  boiling aid.
                                       The  sample  was  then condensed to 0.4
                                       ml in  a  100°C ± 0.5°C water  bath.
                                       About  4  ml  of benzene were added and
                                       condensation  to 0.4 mL was repeated
                                       a second time.  The addition of 4  mL
                                       of benzene  and  condensation  to  0.4 mL
                                       were repeated for  a third  and final
                                       time.  Then 4 mL of hexane were added
                                       to aid in the GC analysis.   A block
                                       diagram  of  the  procedure  is  shown  in
                                       Figure 12.  Analysis  of standard Aro-
                                       clor 1242 solutions yielded  a factor
                                       of 0.737 for conversion of decachloro-
                                       biphenyl  (DCB)  concentrations  to Aro-
                                       clor 1242 concentrations.  Therefore,
                                       all  DCB  concentrations obtained from
                                       the  perchlorinated samples were mul-
                                       tiplied  by  0.737 for  conversion to
                                       Aroclor  1242 concentration.   The GC
                                       instrumental parameters given pre-
                                       viously  were used  for the  analysis.
  Figure 11.
Teflon- plugged reaction vessel used in per-
chlorination reaction
                                      34

-------
      Sample Extraction
           Aqueous

     Quantitatively extract
          with CCI-4.
       Pass CCI4 through
        florisil column.
        Elute with 2-4
       bed volumes CCU.
   Evaporate CCU
 in K-D to ~-0.4 mL.
    Fill to known
volume quantitatively.
     Add aliquot of sample
    quantitatively to reaction
            vessel.
    Add 0.5 to 1.0mLSbCI5.
        Heat sample for
     16to20hrsat200°C.
       Cool. Add ~4mL
         6N HCI. Mix.
   Add quantitatively 1.00 mL
   500 ppb Mirex in benzene.
             1
      Add ~4 mL benzene
          and shake.
                                                        Transfer benzene to
                                                      tube containing NaHCO
                                                         and Na2SO4. Mix.
                                                            KD benzene to 0.5 mL
                                                             in boiling-water bath.
                                                                    ±
                                                          Repeat condensation twice
                                                         withbenzene(4 mL-»0.5 mL)
                                                                     i
                                                         Adjust volume in
                                                         hexane to fit G.C.
                                                            conditions.
Figure 12.   Schematic block diagram of the perchlorination procedure for the analysis of PCB from aqueous and solid samples.
                                                      35

-------
                                  SECTION  5

                         SOLUBILITY OF  PCBs IN WATER1
RESULTS  AND DISCUSSION

Quantitatlon and Identification

     A PCB  mixture contains a variety  of  isomers.   Sissons and Welti  (1971)
have reported that for Aroclor 1242, 1254,  and 1260, the possible number  of
isomers  were 45, 69, and 78.  Because  it  was  not possible to obtain all the
individual  isomers for quantisation and identification of the PCBs analyzed
in this  study, only four isomers were  used  for quantisation.  An assumption
was made that all  of the mono-chlorobiphenyl  isomers gave the same electron
capture  detector response and likewise, the di-, tri-, and tetra-
chlorobiphenyls.  Table 5 lists the relative  response of an electron  capture
detector (63Ni) to the internal standard  and  several PCB isomers.  Although
the responses of the electron capture  detector to  each isomer are not exactly
the same, they are in the same range.  The  isomers with five or more  chlorine
substitutions were quantified by the response of the electron capture detec-
tor to tetrachlorobiphenyl  (2, 4, 2',  4'-).
     To  confirm the above
assumption  for  quantita-
tive analysis,  two stand-
ard solutions were checked.
A standard  containing 40.6
mg of Aroclor 1242 in 10
mL of hexane solution was
prepared.   A 10-yL volume
of solution and 100 yL of
internal standard  (tribro-
mobenzene)  were injected
into 4 mL of hexane,
and then evaporated in a
micro K-D condenser to
approximately 0.5  mL.
Three yL of the concen-
trate were  injected into
  Authors: H.  C.  Lee,
  E.S.K. Chian,  and
  R.A. Griffin
     TABLES. The relative response of a ^Ni electron
            capture detector to tribromobenzene and
            several PCB isomers.
   Chemical
                         Relative response (/Jg/area)
Internal standard
(tribromobenzene)
5.3234 x 10 '
                             3.3772 x 10 :
                             2.2853x10
                             1.9866x 10
                             1.0879x10
                                                                9.6283 x 10'
                                                                9.0012 x 10'
                                      36

-------
      IS
                    10       18
                                              Aroclor   1242
                                                            35
                                            Capacitor  Fluid
                                                       B
Figure 13.  A comparison of hexane-soluble Aroclor 1242 and used capacitor fluid: (A) Aroclor 1242, (B) used capacitor fluid
                                           37

-------
                      TABLE 6. The quantitative composition and percentage of contribution of the
                                 individual water-soluble isomers in Aroclors and capacitor fluid.
Peak
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Total
No. of
Cl
1
1
1
2
2
2
2
2
3
3
3
3
3
2
2&3
3
3
3
3
4
4
4
3&4
4
4
4
4
4
4
4
3
4
4
3
4
5
5
5
5
5


5
5
5



6





Identi-
fication
2
3
4

2,2'
2.4
2.3'
2,4'
2,6,2'
2,5,2'


2,4,2'
3,4'
4,4';2.3,2'
2,4,3'
2,4,4'
2,5,4'
3,4,2'
2,5,2',5'


2,4,2',5'
2,4,2',4'


2,3,2',5'


2,5,3',4'




2,4,3',4'



















AR 1016a
(W L"1) (%)
84.78
13.31
12.74
19.71
18.80
25.94
38.62
171.78
10.29
87.03
2.72

5.78
4.77
49.38
33.52
16.62
110.76
5.43
35.87
5.48
6.95
31.83
27.34
9.18
2.39
19.77
8.16
5.26
14.91
5.27,
3.03
2.12
1.85
14.16


















905.55
9.36
1.47
1.47
2.18
2.08
2.86
4.26
18.97
1.14
9.61
0.30

0.64
0.53
5.45
3.70
1.84
12.23
0.60
3.96
0.61
0.77
3.51
3.02
1.01
0.26
2.18
0.90
0.58
1.65
0.58
0.33
0.23
0.20
1.56



















AR 1221a
(Ha LM) (%)
2572.28
45.75
622.64
13.48
23.50
25.57
33.70
109.59

5.26
1.27
1.87
0.71
4.13
9.07
1.94
1.71
4.73
0.74
1.66
0.64
0.78
1.32
1.08


3.61
0.37

1.48




0.45


















3516.33
73.15
1.30
17.71
0.38
0.67
1.50
0.96
3.12

0.15
0.04
0.05
0.02
0.12
0.26
0.06
0.05
0.13
0.02
0.05
0.02
0.02
0.04
0.03


0.10
0.01

0.04




0.01



















AR 1242b
(/Ug L"1) (%)
121.59

15.07
9.06
21.18
23.76
30.60
138.93
8.33
61.35
2.15
0.48
4.50
2.98
35.30
27.47
9.87
65.07
2.59
22.25
3.59
4.09
19.51
15.88
3.37
1.09
12.83
3.13
2.50
7.21
4.70
2.70
2.63
0.82
9.79

0.99

2.47
2.84













702.67
17.30

2.14
1.29
3.01
3.38
4.35
19.77
1.19
8.73
0.31
0.07
0.64
0.42
5.02
3.91
1.40
9.26
0.37
3.17
0.51
0.58
2.78
2.26
0.48
0.16
1.83
0.45
0.36
1.03
0.67
0.39
0.38
0.12
1.39

0.14

0.35
0.40














AR 1254b
(AO L ') (%)









0.37







1.09
0.28




4.57


2.87

0.23
1.09

2.85

2.33
14.25
0.55

6.46
3.07
2.67
0.29
1.78
2.24
1.72
6.20
0.47

1.48
4.46
0.97
0.80
5.19
1.56
69.84









0.53







1.56
0.40




6.54


4.11

0.33
1.56

4.08

3.34
20.40
0.79

9.25
4.40
3.82
0.42
2.55
3.21
2.46
8.88
0.67

2.12
6.39
1.39
1.15
7.43
2.23

Capacitor fluid3

-------
the gas-liquid chromatograph.  A second standard solution containing capaci-
tor fluid was prepared by adding 28 mg of capacitor fluid to 10 ml of hexane.
The same procedure as with the Aroclor 1242 standard was repeated.  The
results of the analysis showed that 42.77 mg of Aroclor 1242 were quanti-
fied as compared to 40.60 mg of Aroclor 1242 actually present in the sample;
the percentage of error was 5.34 percent.  Versus 28 mg of capacitor fluid
added to the sample, 30.57 mg of capacitor fluid were determined; the per-
centage of error was 9.18 percent.  In both cases, the error was less than
10 percent.  The results show that, considering the errors normally asso-
ciated with PCB analyses, the quantitative method developed in this study
gave satisfactory results.  By using an internal standard and four PCB
isomers, the determinations of concentrations of individual PCB isomers in
a mixture of isomers were relatively accurate.

     GC chromatograms of both hexane-soluble Aroclor 1242 and capacitor
fluid are shown in Figure 13; both have the same isomer distribution pattern.
This indicates that the Aroclor 1242 that was originally impregnated into
the capacitor had not undergone any significant changes during 10 years of
use or during the "burn out" of the capacitor.

     The identification of the water-soluble fractions of several Aroclors
and capacitor fluid are shown in Table 6.  Webb and McCall  (1972) have also
reported data on isomer distributions; peak 3 (4-mono-chlorobiphenyl) was
not reported by them for Aroclor 1242; and peak 10 (2,5,2'-tri-chlorobiphenyl),
peak 18  (2,5,4'-tri-chlorobiphenyl), and peak 19  (3,4,2'-tri-chlorobiphenyl),
were not reported for Aroclor 1254.  In addition, peak 22 and peak 23  (a com-
bination of a tri-chlorobiphenyl and 2,4,2', 5'-tetrarchlorobiphenyl) were
reported by Webb and McCall  (1972) but were not found in Aroclor 1254 in
this study.  The explanation for this apparent discrepancy  seems to be in
the higher resolution of the Emulphor-coated capillary column used in this
study, differences in quantisation techniques, and/or possibly differences
in batches of Aroclors  (EHRC, 1976; Lloyd et al., 1976).

The Time Dependence  of  PCB Dissolution in Water

     One of the major difficulties  in determining the aqueous solubility of
Aroclors is the length  of time necessary to reach a state of equilibrium
between the Aroclors and water.  Figure  14 shows  that about 5 months were
required to reach a  saturation point for both Aroclor 1242  and capacitor
fluid.  Haque, Schmedding, and Freed  (1974) reported that the equilibration
of Aroclor 1254 in water occurred  in about 2 months.  During the  initial
solubilization of about one  week  (Table  7), the  percentages of tri- and
tetra-chlorobiphenyl were high  (about 61 percent  and 32  percent,  respect-
ively);  however, after  the solution approached  equilibrium, the percentages
changed.  The values for mono-, di-, tri-, and  tetra-chlorobiphenyl became
20 percent, 35 percent, 31 percent, and  15 percent, respectively.  That  is,
the composition of water-soluble Aroclor 1242 changed during the  solubili-
zation process and was  also  no  longer the same  as that of the original
Aroclor  1242  fluid.  Higher  portions of  low-chlorinated  isomers predominated
in the water-soluble Aroclor 1242.
                                     39

-------
     A comparison  of  the  solubility of PCB isomers in Aroclor 1242 to the
individual  isomers  (Table 8)  clearly indicates that the solubility of indi-
vidual isomers was  lower  in  the  Aroclor mixture than when measured alone.
It also appeared that the presence  of other chlorinated isomers impaired
the dissolution of  the  lower-chlorinated isomers in the mixture, but, on
the contrary, the dissolution of the higher chlorinated isomers did not seem
to be affected  as   greatly by the  presence of lower chlorinated isomers.

     Table  9 shows  quantitative  values of the water solubility of the indi-
vidual isomers with respect  to time.   The isomers represented by peaks 8, 10,
17, 18, 20, and 24  dissolved  rapidly in the water.   The results shown for
Aroclor 1242 also apply to capacitor fluid.

Solubility  of Aroclors  and Capacitor Fluid

     Representative GC  chromatograms  of the water-soluble PCBs, that is,
Aroclor 1016, 1221, 1242,  and 1254,  and capacitor fluid used in this study,
are shown in Figure 15.   The  quantitative composition and percentages of
contribution of the individual isomers (Table 6) show that in Aroclor 1016
peak numbers 1, 8,  18,  and 20 were  the major peaks  of mono-,  di-,  tri-,
and tetra-chlorobiphenyl.   They  represented 76,  56,  35, and 21 percent of
the individual peaks, respectively.   In Aroclor  1221  the four major peaks
  1,000
                            -i	1	1	1	1	1	1	1	r
       0    20   40    60   80    100   120   140   160   180  200  220   240
                                  Number of days


    Figure 14.  The solubility of Aroclor (AR) 1242 and used capacitor fluid (CF) 1242 in water as a function of time
                                     40

-------
                      TABLE 7. The percentage of PCB isomers in water-soluble Aroclor 1242.
                                   Number of days
Number of Cl
1
2
3
4
7
(%)
a
6
61
32
20
a
29
58
13
30
a
44
47
9
50
a
45
45
11
149
20
34
31
17
251
20
35
31
15
Hexane-soluble
a
12
46
42
 aNo data available because of poor resolution between IS and monochlorinated biphenyl.
of mono-,  di-, tri-, and  tetra-chlorobiphenyl  were represented by peak
numbers  1,  8, 10, and 27,  and the corresponding percentages were 79, 45,
22, and  34 percent.  In Aroclor 1242, peak  numbers 1, 8,  18, and 20 were
the major  peaks of mono-,  di-, tri-, and  tetra-chlorobiphenyl.  The per-
centages of these peaks were 89, 57, 30,  and 22 percent,  respectively.   In
Aroclor  1254 the major peaks of tri-, tetra-,  and penta-chlorobiphenyls were
represented by peaks numbered 34, 35, and 45;  the respective percentages
were  57, 55, and 27 percent.  In capacitor  fluid, the four major peaks
were  represented by peak  numbers 1, 8,  18,  and 20 for mono-, di-, tri-,  and
tetra-chlorobiphenyls, respectively; the  respective percentages were 88,  54,
31, and  22 percent.

      The overall compositions of mono-,  di-, tri-, tetra-, penta-, and
hepta-chlorobiphenyl isomers in the water-soluble fraction of these Aroclors
and capacitor fluid are  listed
in Table  10.  In comparing the
composition of the water-
soluble  Aroclors to the  orig-
inal  Aroclors, such as
reported by Thurston (1971)
and measured in this study,
the water-soluble PCBs were
found to be richer in  the
lower-chlorinated isomers
than  were  the original
PCBs.  Aroclor 1254 is
reported to contain isom-
ers from tetra- to hepta-
chlorobiphenyl in the
original fluid.  The GC
traces of  hexane-soluble
Aroclor  1254  (for example,
Stalling and Huckins,  1971,
and Sissons and Welti,  1971)
 TABLE 8. A comparison of the aqueous solubility of some PCB
        isomers contained in Aroclor 1242 to the pure
        individual isomers.
                          Solubility
Isomer
2,2'
2,5,2'


2,5,2',5'
In mixture
(AOL'1)
                   15.07
                   21.18
                  138.93
                   61.35
  22.25
Individual
(PS L"1)

 4003

 900a

 637b

 248b

  26.5b
aReported by Hoover (1971).
 Reported by Haque and Schmedding (1975).
                                        41

-------
show some  small  peaks in the position of trichlorobiphenyl,  as has also  been
confirmed  in  this study.  The trichlorobiphenyls are  peaks  10, 18, and 19,
shown in Figure  15e.  Because trichlorobiphenyls are  more water soluble
than the more highly chlorinated  biphenyls, detection of trace amounts in
water is reasonable.  The hexa- and  hepta-chlorobiphenyls are enriched in
the Aroclor  1254 fluid and do not dissolve readily  in water.   The low solu-
bility of  these  highly chlorinated isomers in water is apparently the reason
for this situation.
        TABLE 9.
                 A quantitative comparison of water-soluble isomers in Aroclor 1242 with respect to time
Concentration
Peak
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

26

27
28
29
30
31
32

33
34

35
36
37

38
39
40
Total
No.
of Cl
1
1
1
2
2
2
2
2
3
3
3
3
3
2
2&3
3
3
3
3
4
4
4
3&4
4
4

4

4
4
4
4
3
4

4
3

4
5
5

5
5
5

Identi- 7 days 20 days
fication (;Ug L"1) (jUg L."1)
2
3
4

2,2'

2,3'
2,4' 2.90 27.30
2,6,2'
2,5,2' 5.52 22.97

2,4,2'
-
3,4'
4,4';2,3,2' - 8.83
2,4,3'
2,4,4' 2.55
2,5,4' 20.22 35.15
3,4,2'
2,5,2',5' 1.58 6.67

2,4,2',5'
2,4,2',4' 11.12 7.39




2,3,2',5'


2,5,3',4'






2,4,3',4' 1 .96




0.71

46.56 108.31
30 days



1.00
3.18

9.10
71.18
1.40
41.05


-

16.25

4.08
43.28

8.94

0.55
8.93





















208.94
50 days



1.69
1.78
2.93
14.07
78.45
1.72
41.20


1.71

17.19

4.63
48.20

8.78
0.64
0.49
2.11
9.12




0.69
0.74
1.42

0.59





2.18






240.33
149 days
124.60

7.50
5.83
20.80
23.24
19.63
141.85
8.13
62.48


10.00

26.15
29.83
9.83
63.76
2.59
24.05
2.66
3.78
17.95
19.85

4.15

1.11
12.75
4.10

9.15
4.53

4.10
5.43

0.73
13.83


2.53

4.79

671.86
251 days
121.59

15.07
9.06
21.18
23.76
30.60
138.93
8.33
61.35
2.15
0.48
4.50
2.98
35.30
27.47
9.87
65.07
2.59
22.25
3.59
4.09
19.51
15.88

3.37

1.09
12.83
3.13
2.50
7.21
4.70

2.70
2.63

0.82
11.62


0.99

2.47
2.84
702.67
 (-) Indicates peaks too small to quantify.
                                      42

-------
                                             (a) Aroclor 1016
    IS       8   10  IS   1820
    IS I  35678
J
         II Ifl
                10
                                                (b) Aroclor 1221
35
        Figure 15.   GC chromatograms of water-soluble Aroclors and
                   used capacitor fluid.
                            43

-------
                      (c) Aroclor 1242
                      (d) Capacitor Fluid
                     (e) Aroclor 1254
8   , 34I37.38 _40 42  43
                                              51       52   53
              Figure 15.   Continued
                       44

-------
     The comparison of Aroclor 1242 and capacitor fluid as shown  in  Figure
15c and d indicated that the distribution of isomers was  identical,  but
actually the composition was slightly different  (Tables 6 and  10).   The
concentrations of mono-, and di-chlorobiphenyl were slightly higher  in
capacitor fluid; however, each batch of Aroclor may differ slightly  in isomer
composition.  On this basis, the Aroclor 1242 and capacitor fluid were con-
sidered to be essentially identical.  Aroclor 1016 has a  composition similar
to Aroclor 1242 except for isomers with 5 or more chlorines.   The distribu-
tion of isomers in water-soluble Aroclor 1016 and Aroclor 1242 indicated  that
peaks 37 to 40 (pentachlorobiphenyls) were absent in the  chromatogram of
Aroclor 1016 (Figure 15a and c).  This observation was consistent with the
original compositions of Aroclor 1016 and 1242.  Aroclor  1221  (21 percent
chlorine content) and Aroclor 1254  (54 percent chlorine content)  were dif-
ferent in their isomer distribution patterns in water, as shown  in Figure
15b and e.  The lower-chlorinated isomers were predominant in  Aroclor 1221,
whereas higher-chlorinated isomers were predominant in Aroclor 1254.

     The solubility of the PCBs and capacitor fluid in water is  summarized
in Table 11.  It should be noted that the data of MacKay  and Wolkoff (1973)
were calculated values and that the solubility of Aroclor 1016 reported by
Tucker, Litschgi, and Mees (1975) was estimated from the  solubility  of Aro-
clor 1242 on the assumption that Aroclor 1016 contained more of the  lower-
chlorinated isomers than did Aroclor 1242.  The aqueous solubility of Aro-
clors 1016, 1242, and 1254 measured in this study was higher than those
reported in the literature.  The discrepancy in the solubility data  for
Aroclor 1254 may be attributed to the quantisation method employed in this
study.  The response of the electron capture detector to  tetrachlorobiphenyl
was used to estimate the values of the water-soluble fractions of Aroclor
1254, which consisted of a high proportion of tetra- and penta-chlorobiphenyls.
The response factor used for the penta-isomers may have yielded  a high value
for the overall solubility of Aroclor 1254.
            TABLE 10. Summary of isomer composition of water-soluble Aroclors and capacitor fluid.
Number of Cl
1
2
3
4
5
6
7
AR 1221
(%)
92
7
1
a
0
0
0
AR 1016
12
34
35
19
0
0
0
AR 1 242
19
35
31
14
1
0
0
Capacitor fluid
18
32
36
13
1
0
0
AR 1254
0
0
20
56
24
-
-
(-) indicates trace amount.
                                     45

-------
TABLE 11. A summary of the solubility of PCBs and capacitor fluid in water.
AR1016 AR1221 AR 1242 AR 1248 AR 1254 AR 1260
(jUgL"1) (mgL"1) (& L1) (fjg L'1) (VsL1) fa L1}
906 3.5 703 ~70
420 ± 80 340 ± 60
45 ±10
220-250 200
~56
240 54 12 2.7
200 100 50 ~25
43
41
5.0 mg L"1 0.3-3 mg L'1
in fresh in fresh
water water
3.8 mgL"1 0.3-1. 5 mg L"1
in salt in salt
water water
Capacitor
fluid Temperature
(/JQ L"1) (°C)
698 room
<23±2°C)
room
(25±1°C)
room
room
room
25°C
20°C
26°C
4°C
room
Authors
This study
(1978)
Paris et al.
(1978)
Lawrence &
Tosine (1976)
Tucker et al.
(1975)
Haque et al.
(1974)
Mackay &
Wolkoff (1973)
Nisbet &
Sarofin (1972)
Nelson (1972)
Zitko (1970,1971)

-------
     The solubility of Aroclor 1016 and 1242 reported in this study also
appears to be higher than that reported by other investigators.  This appears
to be due to differences in experimental and quantisation methods and pos-
sibly is due to differences in "batches" of Aroclors.  A possible explana-
tion for the lower values reported by earlier workers (Nisbet and Sarofin,
1972; MacKay and Wolkoff, 1973;  and Paris,  Steen,  and Baughman, 1978)  may be
a result of the lack of information on the periods of time needed to reach
equilibrium in water.  This study shows that 5 months were necessary for
Aroclor 1242 to reach equilibrium in aqueous solution, and Haque, Schmedding,
and Freed (1974) have shown that 2 months were necessary for Aroclor 1254.
Whereas, Paris, Steen, and Baughman (1978) reported that saturation of Aro-
clor 1016 and 1242 in water occurred in less than 1 week.  The solubility
of Aroclor 1221 obtained in this study was close to that reported by Zitko
(1970).  Because the solubilities of capacitor fluid have not been reported
previously, no comparison can be made.  Therefore, caution should be used
when interpreting solubility data reported for Aroclors; the effect of the
time-dependence of the solubilization should be considered.
                                     47

-------
                                 SECTION 6

                 ADSORPTION OF WATER SOLUBLE PCBs BY SOIL
                         MATERIALS AND COAL CHARS1


RESULTS AND DISCUSSION

     The data for PCB adsorption by the various earth materials used in this
study are shown in Figure 16 as the amount of PCB adsorbed in relation to the
equilibrium PCB concentration.  Similar results for the low-temperature ashed
samples are shown in Figure 17.

     Several results are evident from the data presented in Figures 16 and
17.  The first result was that, over the concentration range studied, PCB
adsorption by all the earth materials could be described by the linear
relationship:

                                 x/m = KC,                               (1)

where x/m = the amount  of  PCB adsorbed per unit weight of adsorbent (yg/g),
     K = an adsorption constant OnL/g),
and c = the equilibrium PCB concentration (yg/mL).

     The PCB adsorption constant (K) for each adsorbent was obtained from the
slope of the line and is shown in Figure 16 and 17.   The Freundlich adsorp-
tion constant (KF) can be obtained from the foil owing relation:

                                x/m =

Where.£^(yg/g) and n (mL/g)  are constants and the x/m and C are as defined
in equation (1).   It is evident that the PCB adsorption data reported
here are a special case of the Freundlich equation where n = 1.  For the
case n = 1, KF and K are identical, thus Kp values reported in the litera-
ture for other compounds in  units of yg/g can be compared directly with the
K values reported here.

     Another view of the PCB adsorption constant (K)  is as a partition coef-
ficient. The  parameter K is the ratio of the solid phase PCB (x/m} to the
solution phase PCB (c}.  That is, the constant K is  the partition coefficient
between adsorbed and solution phases of PCB in the experimental system used
in this study.  In either view, a simple linear relation (equation [1])
described the adsorption of  PCB from aqueous solution and yielded an  adsorp-

1  Authors:  M.  C. Lee, R. A. Griffin, M. L. Miller, and E.S.K. Chian


                                    48

-------
             400 -
             300-
       -a
       <
              200-
              100-
                                0.1
0.2            0.3            0.4


     PCB Concentration (ppm)
                                                                                          0.5
                                                                                                         0.6

                                                                                                 ISGS 1979
Figure 16.   PCB adsorption by earth materials.  Solid symbols represent data for capacitor fluid; open symbols represent data
            for Aroclor 1242;Xs in open symbols indicate analysis using capillary column.
                                                      49

-------
                                                                Amount PCB Adsorbed (jug/g)
en
O

-------
                     78      10
                               13
                                                                                          B
                                                                                                           ISGS 1979
Figure 18.   Capillary column analysis of Aroclor 1242 remaining in solution: (A) after adsorption on reaction bottle (blank);
            and (B) after adsorption by medium-temperature char
                                                         51

-------
 tion  constant (K)  unique to  each  adsorbent.   Scharpenseel,  Theng,  and  Stephen
 (1978)  also found  a  linear relation  for the  adsorption  of 2,  3,  4-
 trichlorobiphenyl  from n-hexane solution by  montmorillonite and  humic  acid.

      A  second result shown in  Figure 16 is that  there were  no significant
 differences between  adsorption of Aroclor 1242 and  a used capacitor  fluid
 by  any  of  the five adsorbents.  This result  is in agreement with results
 reported previously  for the  solubilities of  fresh Aroclor 1242 and used
 Aroclor 1242 capacitor fluid (Lee et al., 1978).  No significant differences
 in  the  aqueous  solubility or the  distribution of isomers were found; there-
 fore, no differences in adsorption would be  expected.   This result is  signi-
 ficant  because  it  shows that laboratory studies  carried out with fresh Aro-
 clor  1242  can be extrapolated  to  the field,  where aged  fluids that have
 gone  through a  "burn out" process are  found, with a greater degree of  con-
 fidence than previously was  possible.

     A  third result  illustrated in Figures 16 and 17 is the wide differences
 in  adsorption of PCBs  by the various earth materials.   Adsorption  for  the
 five earth materials followed  the series

                          MTC > HTC > CS  > MC > OS.

 Low-temperature ashing  reduced the amounts of PCBs adsorbed by all samples.
 For the series of  Catlin soil  (CS) samples,  the  PCB adsorption decreased as
 the ashing  time increased; this suggests  a relationship between organic car-
 bon content  and adsorption of  PCBs.

 Capillary  Column Analysis of Adsorption

     A  comparison  of the adsorption  of  isomers in the water-soluble Aroclor
 1242 solution by several different adsorbents was investigated by  applica-
 tion of capillary  column GC analysis.  Figure 18 gives  representative  gas
 chromatograms of water-soluble Aroclor 1242 remaining in solution  after
 adsorption  by the  reaction bottle (blank) and by medium-temperature char.
 The quantitative composition and percentage of individual  PCB isomers
 remaining  in  solution after adsorption by several adsorbents  is shown  in
 Table 12.   The total  adsorption found by  the summation of the individual
 isomers is  plotted in Figure 16.   The figure illustrates the close accord
 between the  amounts of  PCBs adsorbed in the two experiments.  Other results
 in  Table 12  indicate that 31  percent of the water-soluble  Aroclor  1242 was
 adsorbed on  the glass surface of the Wheaton reaction bottles.  This  result
 is  in agreement with that of Gresshoff, Mahanty,  and Gortner  (1977) who
 also reported PCB adsorption on glassware.  Zitko (1971) reported that water-
 soluble PCBs were strongly adsorbed by plastic surfaces  such as polethylene
 sheets and tubes.   Eichelberger (1971) also observed that  60 percent  of
Aroclors,  ranging from  1232 to 1268,  in water samples were adsorbed on the
walls of the container or the silt contained in the  river  water.   The results
of this study also confirm the strong affinity of water-soluble PCBs  for
the surfaces of reaction and storage vessels.  The  glass surface preferen-
tially adsorbed peaks 1, 4,  31, and 33 as shown in  Table 12.  This  preferen-
tial adsorption of individual isomers also depends  on the  characteristics
of the adsorbent.   With high  temperature char, peaks 3,  14,  31, and 34


                                     52

-------
      TABLE 1 2.  Quantitative distribution of PCB isomers after adsorption by several adsorbents as analyzed by
               capillary column GC analysis.
Peak
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Total
No. of
Cl
1
1
1
2
2
2
2
2
3
3
3
3
3
2
2&3

3
3
3
3
4
4
4
3&4
4
4
4
4
4
4
4
3
4
4
3
4
5
5
5
5
5

Identi-
fication
2-
3-
4-

2,2'-
2,4-
2,3'-
2,4'-
2,6,2'-
2,5,2'-


2,4,2'-
3,4'-
4,4'-&
2,3,2'-
2,4,3'-
2,4,4'-
2,5,4'-
3,4,2'-
2,5,2',5'-


2,4,2',5'-
2,4,2',4'-


2,3,2',5'-


2,5,3',4'-




2,4,3',4'-






AR1242
(ppb)
121.59

15.07
9.06
21.18
23.76
30.60
138.93
8.33
61.35
2.15
0.48
4.50
2.98
35.30

27.47
9.87
65.07
2.59
22.25
3.59
4.09
19.51
15.88
3.37
1.09
12.83
3.13
2.50
7.21
4.70
2.70
2.63
0.82
9.79

0.99

2.47
2.84
702.67
Blank
(ppb) % remaining
47.67

9.03
3.76
18.04
20.62
24.85
105.58
6.70
49.58
1.38

4.00
2.37
27.32

17.94
8.47
50.19
2.28
18.32
2.76
3.72
14.41
9.26
1.99
0.99
7.21
2.28
2.06
5.45
2.12
1.97
0.73
0.64
7.09

0.59

1.76
1.92
485.05
39.21

59.92
41.50
85.17
86.78
81.21
76.00
80.43
80.81
64.19

88.89
79.53
77.39

65.31
85.82
77.13
88.03
82.34
76.88
90.95
73.86
58.31
53.05
90.83
56.20
72.84
82.40
75.59
45.11
72.96
27.76
78.05
72.42

59.60

72.26
67.61
69.03
HTC (0.1987g)
(ppb) % remaining
21.47

1.14
1.49
8.30
5.25
7.25
31.63
3.16
15.67
0.21

1.03
0.16
5.78

3.47
1.05
6.82
0.55
2.32
0.25
0.58
1.47
0.21
0.21
0.14
0.89
0.25
0.27
0.51
0
0.08
0

0.19





121.80
45.04

12.62
39.63
46.01
25.46
29.18
29.96
47.16
31.61
15.22

25.75
6.75
21.16

19.34
12.40
13.59
24.12
12.66
9.06
15.59
10.20
2.27
10.55
14.14
12.34
10.96
13.11
9.36
0
4.06
0

2.68





25.11
MTC (0.2081 g)
Q
(ppb) % remaining
5.82

0.13
0.42
4.94
2.15
3.56
13.78
2.39
8.86
0.05

0.50
0.14
3.38

2.16
0.55
3.19
0.28
1.22
0.14
0.40
0.64
0.90
0.15
0.08
0.67
0.21
0.15
0.32
0
0.06
0

0.22



0.06
0.04
61.05
12.21

1.44
11.17
27.38
10.43
14.33
13.05
35.67
17.87
3.62

12.50
5.91
12.37

12.04
6.49
6.36
12.28
6.66
5.07
10.75
4.44
9.72
7.54
8.08
9.29
9.21
7.28
5.87
0
3.05
0

3.10



3.41
2.08
12.59
 % remaining compared with AR1242.
 % remaining compared with blank.

were the predominant mono-,  di-,  tri-, and  tetra-chlorobiphenyl  peaks being
adsorbed.   The percentages remaining, compared  with the blank  (glass surface),
were 13, 7, 0, and 0, respectively.  With medium temperature char, peaks 3,
14, 31, and 34 were the predominant mono-,  di-, tri-, and tetra-chlorobiphenyl
peaks being adsorbed.  They  represent 1, 6,  0,  and 0 percent remaining, re-
spectively.  On the other hand,  no preferential adsorption  of  particular
isomers was apparent by either Catlin soil  or sand, but both preferentially
adsorbed higher chlorinated  isomers as a class  over lower chlorinated isomers.

                                       53

-------
                                TABLE 12. Continued.
Peak
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Total
No. of
Cl
1
1
1
2
2
2
2
2
3
3
3
3
3
2
2&3

3
3
3
3
4
4
4
3&4
4
4
4
4
4
4
4
3
4
4
3
4
5
5
5
5
5

Identi-
fication
2-
3-
4-

2,2'-
2,4'
2,3'-
2,4'-
2,6,2'-
2,5,2'-


2,4,2'-
3,4'-
4,4'-&
2,3,2'-
2,4,3'-
2,4,4'-
2,5,4'-
3,4,2'-
2,5,2',5'-


2,4,2',5'-
2,4,2',4'-


2,3.2',5'-


2,5,3',4'-




2,4,3',4'-






HTC (0.0492g)
(ppb) %
15.83

4.36
1.13
9.45
11.36
15.55
66.77
4.43
32.09
0.41

2.37
0.92
15.13

5.24
4.34
25.99
1.34
8.61
0.95
1.57
5.63
4.98
1.12
0.37
2.96
0.44
0.94
2.16
0.28
0.81

0.27
2.53

0.13

0.61
0.58
254.8
remaining
33.21

48.28
30.05
52.38
55.09
62.58
63.24
66.12
64.72
29.71

59.25
38.82
55.38

29.21
51.24
51.78
58.77
47.00
34.42
42.20
39.07
53.74
56.28
37.37
41.05
41.23
45.63
39.63
13.21
41.12

42.19
35.68

22.03

34.66
30.21
52.53
MTC (0.051 1g)
(ppb) %
45.17

1.87
2.94
10.77
6.95
8.45
34.99
3.16
15.34
0.38

1.01
0.24
5.94

6.40
1.36
7.88
0.48
3.11
0.44
0.75
2.23
1.59
0.20
0.11
1.21
0.34
0.29
0.71
0.07
0.13
0.22
0.05
0.61




0.05
165.44
remaining
94.76

20.71
78.19
59.70
33.71
34.00
33.14
47.16
30.94
27.54

25.25
10.13
21.74

35.67
16.06
15.70
21.05
16.98
15.94
20.16
15.48
17.17
10.05
11.11
16.78
14.91
14.08
13.03
3.30
6.60
30.14
7.81
8.60




2.60
34.11
CS (0.0546g)
(ppb) %
45.88

6.88
3.38
13.90
14.72
17.03
73.24
4.71
31.85
0.91

2.46
1.36
16.70

9.63
4.84
28.86
1.27
10.11
1.42
1.84
7.67
5.14
1.24
0.46
3.08
0.96
1.01
2.63
0.83
1.11

0.31
3.29

0.39

0.69
0.74
320.54
remaining
46.25

76.19
89.89
77.05
71.39
68.53
69.37
70.30
64.24
65.94

61.50
57.38
61.13

53.68
57.14
57.50
55.70
55.19
51.45
49.46
53.23
55.51
62.31
46.46
42.72
42.11
49.03
48.26
39.15
56.35

48.44
46.40

66.10

39.20
38.54
66.08
Sand (0.0521 a)
(ppb) %
85.31

12.06
6.09
18.85
19.70
21.96
34.44
6.14
41.79
1.59

3.40
2.17
23.17

17.66
7.15
41.69
1.86
15.40
2.49
3.07
12.67
7.70
1.68
0.76
5.88
1.86
1.66
4.48
2.53
1.63
1.45
0.55
5.93

0.57

1.45
1.66
478.45
remaining
100

100
100
100
95.54
88.37
89.45
91.64
84.29
100

85.0
91.56
84.81

58.44
84.42
83.06
81.58
84.06
90.22
82.53
87.93
83.15
84.42
76.77
81.55
81.58
80.58
82.20
100
82.74
100
85.94
83.64

96.61

82.39
86.46
98.64
|*% remaining compared with AR1242.
 % remaining compared with blank.

     In general,  the  higher chlorinated isomers were more  adsorbable than
the lower chlorinated isomers.   Haque and Schmedding  (1976)  indicated the
extent of adsorption  for the surfaces they studied  followed  the sequence
hexachloro- >  tetrachloro-  > dichloro-biphenyl for  the  isomers chosen.  This
preferential adsorption also depends on the position of the  chlorine atom
on the biphenyl ring; some  particular isomers appeared  to  be more adsorbable
than others in this  study.   They were peaks 1, 3, 4, 14, 31, and 33(Table 12),
                                      54

-------
Peaks 1, 3, and 14 were specifically identified as 2- and 4- monochloro-
biphenyl, and peak 14 as 3, 4'-dichlorobiphenyl.   Peaks 4, 31, and 33 were
identified only as a di-, tri-, and tetra-chlorobiphenyl, respectively.

     The adsorption of each class of isomers is illustrated by the adsorption
data for medium-temperature char in Figure 19.  In this figure, the average
adsorption for all isomers such as monochloro-, and dichloro- have been  con-
sidered together.   It is clear that the higher chlorinated isomers were
preferentially adsorbed.

Earth Material Properties Affecting Adsorption

     The relation between total organic carbon (TOC) content and PCB adsorp-
tion was investigated.  The PCB adsorption constant (K) plotted as a function
of TOC is shown in Figure 20,  A very highly significant (.001 level) corre-
lation was found with a linear regression-relation of

                            K = 255 + 18.5  TOC

                                 r2= 0,87.

Thus, the PCB adsorption constant (K) can be estimated from a knowledge  of
the TOC content of the earth material.

     The relation between surface area (SA) of the adsorbents and PCB adsorp-
tion was also investigated.  Surface area was measured using three adsorbates,
nitrogen, carbon dioxide, and  ethylene glycol.  As is evident from the data
in Table 1, each method can yield large differences in surface area depend-
ing on the surface characteristics of the particular adsorbent.  Nitrogen
adsorption gives a measure of the external surface area and does not pene-
trate pores smaller than about 5 angstroms in diameter.  Carbon dioxide
also measured external surface area, but  is able to penetrate pores smaller
than about 5 angstroms in diameter.  The difference between these two measure-
ments is particularly evident for the medium-temperature char sample, which
has a nitrogen area of 1.6 m2/g and a carbon dioxide area of 253 m2/g.
This large difference is interpreted as a measure of the large proportion of
micropores (< 5 A) contained in the medium-temperature char.  This can be
contrasted to the soil materials, which exhibited relatively small differences
in surface area between the nitrogen and carbon dioxide measurements.

     The third method of measurement was ethylene glycol adsorption.  Ethyl-
ene glycol penetrates the interlayer region of swelling clays such as mont-
morillonite and gives a measure of the total surface area in these types of
samples.  The  difference between the external area measured by nitrogen and
the area measured by ethylene glycol gives an estimate of the "internal" or
"interlayer" surface area.  It is evident that the interlayer surface area
is very high for the montmorillonite clay (Table 1).

     The PCB adsorption constant (K) plotted as a function of CQ2-surface
(SA) is shown in Figure 21.  A highly significant (.001 level) correlation
was found with a linear regression relation of


                                     55

-------
                            K = 230 + 6.64   SA

                                 r2 = 0.82.

Very poor correlations were obtained using nitrogen and ethylene glycol
surface areas.  This is interpreted to  indicate that PCB molecules were
not able to penetrate into the interlayer surface regions  of montmorillonite
clay to be adsorbed.  Thus, the PCB adsorption constant (K) could be esti-
mated from CC-2-SA values.  This result  may be fortuitous since the C02-SA
values parallel the TOC values.  More studies are needed to confirm this
result.

     A three-variable regression analysis of the PCB adsorption constant U),
TOC, and C02--SA were investigated and the results are shown in Figure 22.  A
highly significant (.001 level) correlation was obtained with a linear regres
sion relation of

                       K = 188 + 3.36 SA + 11.4  TOC

                               T2  =   0.95.

     The magnitude of the coefficients  for SA and TOC indicate that TOC is
the dominant property by a factor greater than three.  The best estimates
of K were obtained by a knowledge of both SA and TOC; however, if only one
soil property must be chosen to estimate K, the TOC should be the property
of choice.   Care should be taken when applying this equation since K may
be over-estimated for soils with low TOC content.
                                     56

-------
                                      monochlorobiphenyls

                                  •/\ dichlorobiphenyls

                                      trichlorobiphenyls

                                      tetrachlorobiphenyls

                                      pentachlorobiphenyls
                                                                    2,000
                                0.10
                         Amount of Char (g)
                                                         0.20
                                                   ISGS1979
 Figure 19.   The percentage of isomers in water-soluble Aroclor
             1242 remaining in solution after adsorption by
             medium-temperature char
                                                                 §  1,000 -
                                                                 Q.
                                                                 O
                                                                 CO
                                                                 a
                    20    30    40    50    60

                       Total Organic Carbon (%)
                                                                                                                    70
                                                ISGS1979
 Figure 20.    PCB adsorption constant (K) as a function of
              total organic carbon content (TOO
o
O
    2,000
    1,000
                                    K = 230 + 6.64S-4
                                        r2 = 0.82
          0     40     80    120    160    200    240   280
                       CO2—Surface Area (m /g).
                                                  ISGS1979
  Figure 21.    PCB adsorption constant (K) as a function of
               surface area (SA)
                                                                   2,000
                                                                 I  1,000
                                                                             K= 188 + 3.36S/4 + 11.4rOC
                                                                                     r2 = 0.95
                                                                                               1,000
                                                                                            IS
                                                                                             experimental
                                                     2,000
                                                                                                                  ISGS1979
Figure 22.   Three-variable regression analysis of the PCB
            adsorption constant (K) as a function of total
            organic carbon content (TOO and surface area
            (SA);/fcomputed  as a function of
            ^experimental
                                                           57

-------
                                 SECTION 7

              MOBILITY OF PCBs AND DICAMBA IN SOIL MATERIALS:
              DETERMINATION BY SOIL THIN-LAYER CHROMATOGRAPHY1


RESULTS AND DISCUSSION

     Rf values obtained from autoradiography were compared with those
obtained by zonal extraction.  The agreements were excellent.  Autoradiog-
raphy was  especially  satisfactory for soil TLC studies, as it provides a
qualitative picture of movement (for example, diffusion, tailing) while
allowing measurement  of frontal Rf.  Zonal extraction gave a more quantita-
tive picture of PCB movement.  Also detected by the zonal extraction pro-
cedure are tailing, the origin spot, and the frontal movement of the com-
pound; however, definition of the concentration profile is limited by the
length of the soil segment chosen and by the extraction and analytical
efficiencies.  In summary, the two methods gave essentially identical pic-
tures of PCB movement on TLC plates.

     The mobilities of Aroclor 1242, Aroclor 1254, and Dicamba in several
earth materials expressed as frontal Rf values are summarized in Table 13.
The data for capacitor fluid were identical to Aroclor 1242 and are not
shown in Table 13.  The data show that under the conditions tested, Aroclor
1242 and Aroclor 1254 (and the capacitor fluid) stayed immobile in these
soil materials when leached with water or Du Page leachate, but were highly
mobile when leached with carbon tetrachloride.  Dicamba showed the reverse
trend, being highly mobile in water and in Du Page leachate and quite
stationary when leached with carbon tetrachloride.

     A closer look at the structure of Dicamba and PCB will help explain
the mobilities observed.   Dicamba is 3,6-dichloro-o-anisic acid.  Because
Dicamba has two polar groups (-COOH and -OCH3), hydrogen bonding can occur
between the water molecules and the carboxyl  and methoxy groups of Dicamba;
this increases the solubility of Dicamba in polar solvents like water and
leachate.   Solubility of Dicamba in water is 4,500 ppm (Herbicide Handbook,
1974).

     PCBs are nonpolar in nature and are only very slightly soluble in polar
solvents like water.   Solubility of Aroclor 1242 in water has been deter-
mined to be 200 ppb (Tucker, Litschgi, and Mees,  1975), and Aroclor 1254,
1  Authors:  R. A. Griffin, A.  K.  Au, and E.S.K.  Chian


                                     58

-------
TABLE 13. Mobility of Aroclor 1242, Aroclor 1254, and Dicamba in several soil materials leached with various solvents as measured

        by soil TLC.
                                           Rf values

Ava
Silty clay loam
Bloomfield
Loamy sand
Catlin
Silt loam
Catlin
Loam
Cisne
Silt loam
Coal char
(1200°F)
Drummer
Silt loam
Flanagan
Silt loam
Ottawa
Silica sand

Aroclor
1242
.02

.03
.02
.02
.03
.03
.03
.02
.03
H2O DuPage leachate
Aroclor Aroclor Aroclor Aroclor
1254 Dicamba 1242 1254 Dicamba 1242
.02 1.00 .02 .02 1.00 1.00

.03 1.00 -
.02 .85 .04 .04 .90 1.00
.02 1.00 .03 .03 1.00 1.00
.02 1.00 .03 .02 .99 1.00
.03 .79 .04 .04 .80 1.00
.03 0.99 - - - 1.00
.02 1.00 .06 .05 1.00 1.00
.03 1.00 .03 .03 1.00 1.00
CCI4
Aroclor
1254
.96

'
1.00
1.00
1.00
1.00
1.00
1.00
1.00

Dicamba
.02

"
.02
.03
.05
.03
.03
.02
.02
 56 ppb  (Haque,  Schmedding, and Freed, 1974); however,  PCBs are quite soluble
 in organic  solvents  like acetone, methanol, benzene, or carbon tetrachloride.
 Mobilities  of Aroclor 1242, Aroclor 1254 and Dicamba were tested in silica
 gel  leached with acetone, methanol, benzene, carbon tetrachloride, and
 mixtures  of water-acetone and water-methanol.  The results are given in
 Table 14.   Consistent with the soil TLC data obtained  by leaching with
 carbon  tetrachloride, Rf values of 1.00 were obtained  using the organic
 solvents.   The  data  also indicated that relatively small amounts of water
 (9 percent) in  the methanol-water mixture  significantly reduced the mobil-
 ity  of  the  PCBs compared to pure methanol.  Dicamba mobility followed the
 inverse trend.   It was quite clear that mobilities of  PCBs and Dicamba in
 soil materials  and silica-gel were highly  related to the solubility of PCBs
 and  Dicamba in  the solvent with which the  TLC  plates were being leached.

      The  above  finding has great significance  in the disposal of PCB wastes.
 To prevent  potential migration of PCBs in  a landfill,  PCB wastes and organic
                                       59

-------
 solvents should not be disposed of in
 the same landfill  location, and the
 PCBs should not be allowed to come in
 contact with leaching organic solvents.

      The mobility  of PCBs in carbon
 tetrachloride can  be correlated with
 characteristics of the earth materials
 such as clay content, surface area, PH,
 cation  exchange capacity, and organic
 matter.   The zonal  extraction technique
 yields  a quantitative measure of the
 quantity of PCB that moves to the top
 of  each soil  TLC plate with the solvent.
 Although PCBs have R*  values of 1.00
 for all  the soil materials leached with
 carbon  tetrachloride (Table 13), the
 percentage  of PCBs  that were retained
 against leaching on the lower part of
 the TLC plates  differs  for each of the
 soil  materials.  The mobility measured
 in  this  manner  was  correlated with soil
 material  properties (Table 15).
          roc i
                       ISGS 1979
                                          Figure 23.
Retention of PCB in soils leached with carbon
tetrachloride as a function of total organic
carbon (TOO content of the soils
     The relation  between  the  total
organic carbon content  of  the  soils
and retention of PCB  by soils  leached  with carbon tetrachloride was investi-
gated, and the results  are plotted  in  Figure  23.   The plots indicate there
is a tendency toward  higher percentages  of retention of Aroclor 1254 com-
pared to Aroclor 1242.   This hypothesis  was tested using the paired t-test
statistic (Texas Instruments,  1975)  for  pairs of  data from the same soil.
The result was a highly significant  (.01  level) difference between the
mobility of Aroclors  1242  and  1254  in  carbon  tetrachloride solution, Aroclor
1254 being less mobile.  This  result is  in agreement with Tucker,  Litschgi,
and Mees (1975), Haque  and Schmedding  (1976), and Lee et al.  (1979) who
found the higher-chlorinated PCB  isomers  had  a higher affinity for soil
materials than the lower-chlorinated isomers.

     The results of linear regression  analysis of the data shown in Figure
23 is tabulated in Table 15.   The data yielded a  highly significant (.001
level) correlation between  retention of  PCBs  by soil  and the  TOC content.
The coefficients of determination (r2) were 0.87  and 0.84 for Aroclors 1242
and 1254, respectively.  These results are also congruent with those of
several other workers who  have noted a relation between PCB  retention and
the organic matter in soils  (Briggs, 1973;  Iwata,  Westlake,  and Gunther,
1973; Tucker, Litschgi, and Mees, 1975;  Scharpenseel,  Theng,  and Stephan,
1978; Lee et al., 1979).

     Also included in Table 15 are the results of  linear regression analysis
of the effect of surface area on mobility  of  PCBs.   The results indicate  no
significant correlation with nitrogen  surface  area values,  but significant
correlations were obtained using carbon dioxide and  ethylene  glycol  surface

                                     60


-------
    TABLE 14.  Mobility of Aroclor 1242,Aroclor 1254, and Dicamba on silica-gel TUG plates using various leaching solvents.


                                                                               R- values
Leaching solvent                                  Aroclor 1242                  Aroclor 1254                   Dicamba

Deionized H2O                                       0.15                          0.15                         0.95
Du Page leachate                                      0.15                          0.15                         0.80
80% H20-20% acetone                               0.09                          0.06                         0.30
Acetone                                             1.00                          1.00                         0.88
15% H20-85% methanol                              0.79                          0.79                         0.71
9% H20-91%methanol                               0.80                          0.83                         0.70
Methanol                                            1.00                          1.00
Benzene                                             0.99                          0.95                         0.09
Carbon tetrachloride                                  1.00                          1.00                         0.03
 TABLE 15.  Results of linear regression analysis of retention of Aroclor 1242 and Aroclor 1254 by soils against leaching with
             carbon tetrachloride  (R) versus the soil properties of total organic carbon (TOG) content and surface area measured
             by adsorption of nitrogen (N2-SA), carbon dioxide (C02-SA),  and ethylene glycol (EG-SA) expressed as values of
             the coefficient of determination (r  ).
                                                                           r values
                                                Aroclor 1242                                     Aroclor 1254
R vs. TOG                                           0.873                                            0.84a

R vs. N2-SA                                         0.03NS                                          0.09NS

R vs. CO2-SA                                        0.54b                                            0.54b

R vs. EG-SA                                         0.74a                                            0.43C

R vs. TOG, CO2-SA                                   0.90a                                            0.89a

R vs. TOG, EG-SA	0.93a	0.87a

   Not significant
Significant at 0.001 level
 Significant at 0.05 level
Significant at 0.1 level
                                                        61

-------
area values.  Highly significant correlations were obtained by including both
surface area and TOC content in regressions with the mobility of the two
Aroclors (Table 15).

     The lack of strong correlations and the variability in the results
using the surface area data precludes a conclusion as to the ability of PCB
molecules in carbon tetrachloride solution to penetrate into the interlayer
region of expandable clays or as to the significance of the surface area of
soils in retarding PCB migration.  There is, however, a clear relation
between the TOC content of the soils and their ability to retard PCB migra-
tion.
                                    62

-------
                                    SECTION 8

              DEGRADATION  OF PCBs BY MIXED MICROBIAL CULTURES1


RESULTS  AND DISCUSSION

Preliminary Work

     The first mixed culture to be grown, the C-line,  yielded the yellow-
colored  intermediate described by Furukawa and Matsumara (1976) in  about 5
weeks; the S-line took  about 2 weeks  and the N-line,  less than 1 week to
turn the solutions yellow.   Because the  S-line had some previous exposure

            TABLE 16. Results after 15 and 45 days of degradation using culture grown in acetic acid.
                                       Percentage of removal of peak
                                        when compared to blanks
Peak number
2
4
5
6,7
8
9
11
13
15
16
17

18
22
24
25
26
27
28
32
34
37
41
44,45
Number of Cl
1
2
2
2
2
2
2
3
3
3
2 and 3
mixture
3
3
3
4
4
4
3
4
4
4
4
4
1 5 days
100
29
14
100
100
62
37
0
0
0
0

22
29
10
0
-
0
100
0
0
31
0
0
45 days
100
28
23
100
100
62
37
0
0
0
0

23
56
13
0
-
0
85
0
0
34
0
0
Retention time
in Fig. 25
(min)
10.43
15.30
15.97
16.96
17.19
18.96
19.74
22.04
23.55
24.56
25.83

26.12
27.40
28.29
29.45
29.90
30.65
30.96
31.77
34.34
35.37
38.35
40.96
Total degradation
43
                     44
1  Authors:   R.  R. Clark,  E.S.K.  Chian,and  R.  A. Griffin
                                       63

-------
 to PCBs before isolation  and  the N-line had much more  exposure,  it appears
 that the presence of PCBs in  a particular environment  enhances  the selection
 of microorganisms capable of  degrading PCBs.  In view  of  this,  the mixed
 cultures were grown on  biphenyl  for more than 7 months to ensure that the
 cultures were highly selective toward biphenyl-degrading  organisms before
 the PCB degradation experiments  were conducted.

      The first PCB degradation studies were run for  15, 30,  and  45 days.
 Figure 24 shows a chromatogram of the 15-day extract.   Figure 25 shows the
 chromatogram of the 15-day blank.  Table 16 lists the  peaks  found in the
 PCB samples.  Occasionally peaks cannot be resolved, therefore  these are
 combined in the tables.   As can  be seen, all of the  monochloro  and most of
 the dichlorobiphenyls are easily degraded.  It also  appears  that most of
 the degradation occurred  prior to 15 days.

      The results, however,  were  suspect because the  mixed culture was grown
 in acetic acid for 3 days prior  to being added to the  PCB solution under
 study to eliminate the  biphenyl  in the solution.  This procedure could have
 made the mixed culture  less able to degrade the PCBs because the selective
 pressure toward a biphenyl-degrading culture was eased.

      Table 17 lists the results  of a 7-day study of  PCB degradation using
both a mixed culture grown only on biphenyl and one grown  on  acetic acid for
           TABLE 1 7.  Comparison of mixed cultures grown on biphenyl only and those grown on acetic acid.
 Peak number
                 Number of Cl
                                          Percentage of removal after 7 days when compared to blanks


                                          Biphenyl only                Acetic acid grown
2
3
4
5
6,7
8
11
13
15
16
17
18
22
24
25
27
28
32
33
34
1
1
2
2
2
2
2
3
3
3
2 + 3
3
3
3
4
4
3
4
4
4
100
100
100
-81
100
100
100
0
19
0
66
100
100
93
33
0
100
0
0
0
100
100
7
13
100
100
20
0
0
0
0
0
15
0
0
0
68
0
0
0
 Total degradation
                                             82
                                                                        30
                                       64

-------
en
C71
                                                       Figure 24.   Water-soluble Aroclor 1242 after 15 days of degradation

-------
cn
                                                      Figure 25.   Original water-soluble Aroclor 1242 solution prior to degradation

-------
3 days prior to the experiment.  The culture grown only on biphenyl  did much
better than the culture grown on acetic acid for 3 days.  These results indi-
cate that the data in Table 16 may not be quantitatively correct, but are
probably qualitatively correct in that most of the degradation occurred in
less than 15 days.

      The data in Table 17 show that all the mono- and dichloro-biphenyl
isomers (except peak number 15) were easily degraded in 7 days.  Also, many
of the trichlorobiphenyl  isomers were degraded, but some were resistant to
degradation.  Almost all of the tetrachlorobiphenyl isomers were not degraded
in 7 days.

Sonification Study

To determine if the extraction procedure would extract all the PCBs in
solution, a 2-day degradation sample of the S-line culture was run and soni-
fied before extraction.  In this way, the PCBs that are inside the micro-
organisms would be extracted.  If the pH 2 extraction did not extract all
of the PCBs inside the microorganisms, then the sonified extract would show
higher amounts of PCB isomers and/or more isomer peaks because of isomers
that could not be extracted with the pH 2 extraction.  Table 18 shows the
results of the sonification study.  The sonified extract showed a greater
amount of PCB removal than did the pH 2 extract.  This difference may not
be significant since the error in the analysis may be more than the dif-
ference between the two samples.  Volatilization of PCBs due to localized
heating of the solution near the sonifier tip is another possible explana-
tion.  In addition, all peaks that were missing in the pH 2 extract were
missing in the sonified extract.  Because of this, it was felt that the pH
2 extraction procedure adequately represented the degradation of PCBs that
was occurring in the solutions.

Comparison of the Three Mixed Cultures

      Degradation samples were run for 5, 10, and 15 days using Aroclor 1242
saturated solution inoculated with 3-day old N-, C-, or S-line cultures.
Figure 26 shows a comparison of a 5-day blank and 5-day degradation using
the S culture.  Figure 27 shows a 5-day blank, and Figure 28, a 5-day degrad-
ation, both using the N culture.  Note that the analyses were performed using
a different capillary column.  The C culture was also analyzed on the same
capillary column as the N culture, therefore the chromatograms resembled
the N-culture chromatograms.  As can be seen from these figures, the S-
culture degradation may not be directly comparable to the C- and N-culture
degradation because  of differences in the capillary column.  Table 19 com-
pares the amount of degradation for each culture at 5 and 15 days.  Some
peaks were not reproducibly present, so no percentage of degradation is
reported for them.  Because of analytical difficulties, for those isomers
yielding very small peaks, degradation of less than 20 percent was not
considered significant.  For most of the peaks the degradation was the
same for 5 as for 15'days; this observation indicated that most of the
degradation occurred in less than 5 days.
                                     67

-------
                     TABLE 18.  Comparison of sonified extracts and pH 2 extracts.
 Peak number
                    Number of Cl
          Percentage of degradation


pH 2 extract                  Sonified extract
2
3
4
5
6,7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28,29,30
31
32
33
34
35
36,37
38
39
40
41
42
43
44
45
1
1
2
2
2
2
2
3
3
3
3
2
2 + 3
3
3
3
3
3
3
4
4
4
4
3
4
4
4
4
4
4
4
4
4
4
3
4
4
4
4
4
100
100
54
45
100
95
100
0
8
29
28
100
27
56
13
74
66
100
0
16
.
100
0
100
0
0
17
0
0
0
0
0
2
38
5
5
.
.
.
-
100
100
75
68
100
99
100
34
39
45
51
100
48
70
0
82
75
100
17
35

100
19
100
0
7
80
10
10
11
13
21
4
49
65
2



-
      Most of the  isomers  were degraded to the same degree with  all  three
mixed cultures;  however, some were degraded to different extents.   Differ-
ences in degradation  can be due to different mechanisms or pathways  of
degradation  (Ahmed and  Focht, 197.3a; Baxter et al., 1975; Furukawa,
Tonomura, and Kamibayashi, 1978).   The usual patterns reported  in  other
literature—monochlorobiphenyIs are easily degraded, dichlorobiphenyls
are slightly less  degradational,  and tetrachlorobiphenyls are not  readily
degraded—are seen in the  data reported here.  In  this study, however, the
degradation  occurred  more  rapidly than previously  reported (Ahmed  and Focht,
1973a; Baxter et al., 1975; Furukawa and Matsumura, 1976; Furukawa,
Tonomura, and Kamibayashi, 1978;  Sundstrom, Hutzinger, and Safe, 1976;
Tucker, Saeger,  and Hicks, 1975;  Mihashi et al., 1975).
                                      68

-------
                   5-day blank
                                                           5-day degradation
         Figure 26.  Comparison of 5-day blank and 5-day degradation of Aroclor 1242 using the S culture
Kinetic Study

      A degradation  study was run using the N culture for  3,  6,  12, 24, and
48 hours.  Two samples  and two blanks were taken at each time increment.
Table 20 shows the results of this study.  As with previous  studies, the
lower chlorinated isomers were degraded most rapidly.  Many  of the isomers
showed increasing amounts of degradation with each time increment.  Six of
the peaks are plotted  in  Figures 29 and 30.  The rates of  degradation are
somewhat different for  different isomers; even isomers with  the  same number
of chlorines per molecule have different rates of degradation.  This dif-
ference has also been  reported by Baxter et alI., (1975).   This finding
would indicate that  the position of substitution has a bearing on the
degradational properties  of a particular isomer.  Even with  the  use of a
capillary column, each  particular peak may not be a pure isomer; a peak
                                      69

-------
  -9
     10
               16
              '
           13






             4
                      19
                   18
u
   Figure 27.   Results of 5-day blank of Aroclor 1242 using the N culture
      10
                      IS
               15
           13
                 16
                 I7I8
                         22
Figure 28.    Results of 5-day degradation of Aroclor 1242 using the N culture




                                    70
                                                                                      4.445

-------
 TABLE 19. Comparison of degradation of water-soluble Aroclor 1242 by the three mixed cultures of soil microorganisms.
Peak number
2
3
4
5
6,7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Number of Cl
1
1
2
2
2
2
2
3
3
3
3
2
2 + 3
3
3
3
3
3
3
4
4
4
4
3
4
4
4
4
4
4
4
4
4
4
3
4
4
3
4
4
4
4
4
C
(%)
100
100
64.8
57.3
98.7
-
97.2
12.6
19.8
38.5
48.4
100
41.9
73,2
7.4
77.8
51.3
100
2.9
21.9
-
100
7.2
100
99
8.5
25.2
-
9.7
8.1
0.7
0
0
19.4
3.7
0
20.3
25
0
0
0
0
17
5-day degradation
S
(%)
100
100
100
93
100
100
99
88
67
-
90
-
71
99
100
98
96
-
-
72
-
99
72
100
90
42
-
•,
38
34


39
-
38
44
81
15
30
-
-
-
•
N
(%)
100
100
30.9
98.6
100
99.3
99.7
0
98.7
45.7
15.2
75.2
39.3
86.9
22.0
100
93.9
100
0
99.2
-
60.9
0
100
100
20.1
-
0
27.3
55.6
55.6
0
37.0
-
17.5
37.9
37.9
0
47.9
-
-
21.3
21.3
C
(%)
100
100
71.5
61.7
100
99.5
97.9
3.8
17.3
21.4
34.9
100
42.8
77.0
0
80.6
53.9
100
0
18.7
0
100
0
100
0
0
0
0
31.8
0
0
0
0
22.5
0
0
49.4
-
0
-
-
-
76.1
1 5-day degradation
S
(%)
100
100
100
99
100
100
99
89
71
-
89
-
78
99
100
100
98
-
-
76
-
99
73
100
90
56
-
-
43
49


67
-
51
45
85
0
43
-
-
-
-
N
(%)
100
100
25.3
100
100
100
99.9
0
100
54.8
0
78.2
45.8
96.8
0
100
98.4
100
0
100
0
70.3
0
100
100
13.0
21.6
0
0
80.2
62.8
0
38.2
0
0
100
0
0
62
-
-
-
-
could be  a mixture of isomers.   If it were a mixture, the  isomers may  have
different degradation rates.

      In  order to calculate the  amount of PCBs  left at any particular  time,
the reaction order, n, and the rate constant, K,  must be determined.   The
basic rate equation  is
                                        71

-------
                                   _ _
                                  dt ~

where c is the concentration of a compound, || is the change in concentra-

tion with respect to time, t and K and n are as defined above.   The equation

can be rearranged to
f---
TABLE 20. Results of the kinetic study of the degradation of water-soluble Aroclor 1242 by the N culture.
Percentage of degradation
Peak number
2


6-1
,7

10
11
12
13
14
15
1 C
16
17,18
19,20
21,22
23
24
25
26
27
28,29,30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Number of Cl
1
1
2
2
2
2
2
3
3
3
3
2
2 + 3
3
3
3
3
4
4
4
3
4
4
4
4
4
4
4
4
3
4
4
3
4
4
4
4
4
3hrs.
(%)
92.1
100
0
12.5
72.1
35.5
38.2
0
8.4
41.6
0
100
20.6
17.3
27.1
19.5
0
34.2
18.4
32.2
0
0
0
0
0
0
0
0
24.7
0
0
,
14.3
„

.
0
6hrs.
(%)
100
100
9.8
10.9
93.3
52.9
64.1
3.6
0
18.5
0
100
12.9
9.8
31.9
25.3
0
40.1
0
56.1
0
0
0
0
0
0
0
0
0
0
0
_
0



.
12hrs.
(%)
100
100
21.8
20.8
100
76.7
80
0
5.6
13.7
0
100
15.0
20.1
53.4
40.4
6.0
56.4
0
84.2
. 0
0
62.4
0
0
0
0
0
0
0
0

0




24 hrs.
(%)
100
100
19.7
35.9
100
87.9
88.5
0
20.9
14.3
0
100
15.8
26.9
64.2
49.8
14.4

0
92.9
15.0
0

0
0
0
0
0
0
0
0

0




48 hrs.
(%)
100
100
36.6
69.6
100
90.3
95.5
2.6
55.9
23.1
5.8
100
22.9
45.5
80.0
59.8
37.2
71.5
o
95.9
51.8
0
17.1
0
o
0
0
0
6.2
0
0

0



n
                                     72

-------
Percentage
 Remaining
                            12
                                           24
                                                                          48
                                                      Time in hours

                         Figure 29.   Degradation at timed intervals of selected PCB isomers

                                                  73
                                                                                             ISGS 1978

-------
Percentage
remaining
                3   6
                                                Time in hours

                      Figure 30.    Degradation at tinned intervals of selected PCB isomers
                                                                                                  I SOS 1979
                                                    74

-------
By integrating equation 3, we get equation 4

                         	     	                                 (4)
                         l-n       l-n

Rearranging this
                         _ l-n     r l-n
                         °t     ~  °o

or

                         C l~n  =  (n-l)  Kt + C^~n ;                   (6)

the final equation is then
                         Ct  =  [(n-l)  Kt + CQ ^  ^ -              (7)

     To determine the reaction order for the reduction of sample peaks
 5,  8,  9,  (17-18),  (19-20), and 26  the  following methods were  used:
 Since  the  basic rate equation is

                         _d£ =  -Kd1                                    (2)
                          at
 then
                              dt
 By plotting  inFgf)  vs.  in  C,  the  slope  (n)  and  the  intercept  (in K) can  be
 determined  (|| was determined  from the percentage  remaining  from Table  20;
 C was  taken  as the concentration midway  between  the  two  points  used to
 determine a particular slope
      Table 21  lists  the K and  n values  for  peaks  number  5,  8,  9,  (17-18),
 (19-20),  and 26.   The  rate constant, K,  was calculated using the  percent-
 age of PCB remaining at each time; therefore the  units of K are  (percentage
 remaining) ^~n'per  hour.   C  equals 100 percent  at time zero.  The
                                      75

-------
 constants  in  Table  21,  when  used  in  equation 7,  yield answers that are within
 5  percent  of  the  experimental  data.

     Several  things  should be  pointed  out with regard to these numbers.

 First, the differential method used  here to determine n and K is subject

 to errors  because of the  difficulties  in determining the accurate values

 for -%j--   Tnis  is especially applicable  here because only 5 data points
 were used  to  calculate  the line for  determining  n and K.   However, the

 correlation coefficient (r)  was very high (> 0.95)  for all  but two of the

 peaks.  These two peaks,  (17-18) and 5,  had r values of 0.867 and 0.64

 respectively.   Also  the closeness of the calculated numbers to the experi-
 mental data gives support to the calculated n and K values.


     Second,  the  n value  for peak (19-20)  is very high.   One would expect
n values of 2 or  less.  The  high n value may be  due to the  fact that this
 peak and (17-18)  are a  combination of  two  or more different PCB isomers.
 If the two or more isomers degrade at  rapidly different rates,  then the n
 value for  the combined  peak may be  somewhat unusual.   The  n values, although
 they may not be a true  reaction order  for  a particular isomer,  do allow an
 empirical  prediction using equation  7.

Cometabolization

     PCBs are rarely found in  isolation  in  the environment.   There are
usually more easily oxidizable chemicals such  as  acetic acid  and  amino
acids available for the microorganisms to  utilize.   When microorganisms
are utilizing one compound as an energy  source,  cometabolization  of another
compound may occur (Baxter et al., 1975; Horvath, 1972).  Horvath  (1972)
described cometabolization as "the process  in which  a  microorganism oxidized
a substance without being able to utilize  the energy derived  from  this oxida-
tion to support growth."

                TABLE 21. Rate constants and reaction orders for selected PCB isomers.
Peak number
5
8
9
17,18
19,20
26
Number of Cl
2
2
2
3
3
3
n
0.347
2.211
1.747
2.04
3.72
1.651
K (percentage remaining) /hr
3.44 x 10"1
9.283 x 10"4
7.24 x 10~3
6.1 9 x 10~4
2.92 x 10~7
1.04 x 10"2
                                     76

-------
     Cometabolization of PCBs in the presence of sodium acetate was studied.
Table 22 lists the results after 5 and 15 days of cometabolization and PCB
degradation.   Both cultures degraded PCBs to a large degree, but the cometab-
olization samples produced greatly enhanced degradation of the higher chlor-
inated PCB isomers and other hard-to-be-degraded isomers.

     One possible reason for enhanced degradation of PCBs  in the cometabo-
lization samples could be due to the increased mass of microorganisms present.
When the microorganisms are grown on PCBs alone, very little additional
biomass is produced because of the limited amount of carbon source present
in the solution.  When acetic acid, or any easily degradable substrate, is
present the microorganisms can grow on the acetic acid.  While they are
using the acetic acid they may oxidize the PCBs slightly.   Because the
total mass of microorganisms present has increased, this slight oxidization
by each microorganism adds up to a more complete degradation for the entire
system as a whole.  This type of cometabolization will only work for cul-
tures of microorganisms that are acclimated or selected to PCB degradation.

     Another possible explanation is that the PCBs adsorbed onto the addi-_
tional biomass in the cometabolization samples and were not extracted.  This
is a possibility because PCBs adsorb easily onto nonpolar surfaces such as
bacterial cell walls.  The sonification study showed that the PCBs could
be extracted from the small amounts of biomass  present in the samples con-
taining only PCBs, but did not rule out the possibility of poor extraction
in the presence of large amounts of biomass.

Soil and Sludge Extraction

     Chromatograms of the soil and sludge extracts were compared to chromat-
ograms of Aroclor 1242 and 1254.  By this comparison the peaks in the
extracts that were caused by PCB isomers were determined.  Figure 15 shows
a plot of relative area of the identified PCB peaks versus chromatograph
retention time.  If the PCBs were being degraded in the soil after applica-
tion in the sludge, then there should be a buildup of  the higher chlorinated
isomers  (those with longer retention times).  In addition, there should be
fewer of the lower chlorinated isomers in the soil.  Each point on the graph
represents a peak determined to be a PCB isomer.  The  number over the  point
is the percentage of degradation of that peak in the cometabolization  study.
The degradation from the cometabolization study was used because it would
more closely represent the environmental situation where other organics are
present.

     As  illustrated in Figure 31, the sludge  has more  low chlorine isomers
than the soil, and the soil has relatively more high chlorine  isomers  than
the sludge.  This seems to indicate that there  is degradation  of some  low
chlorine isomers and an accumulation of the nondegradational higher chlorine
isomers.  The cometabolization degradation results agree in part with  this
conclusion.  The two peaks at 25 and 28 minutes are not degraded and  they
build up in the soil, whereas the peak at 26  minutes  is degraded and  does
not build up to the same extent.  This pattern, however, is not obtained for
the peaks at 15.5 and 23 minutes.  This result may not be significant
                                      77

-------
TABLE 22.  Co-metabolization of PCBs with sodium acetate by mixed culture microorganisms.
Peak number
2
3
4
5
6,7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Number of Cl
1
1
2
2
2
2
2
3
3
3
3
2
2 + 3
3
3
3
3
3
3
4
4
4
4
3
4
4
4
4
4
4
4
4
4
4
3
4
4
3
4
4
4
4
4
5 day
PCB
(%)
100
100
54.5
99
100
100
100
0
96.8
96.8
8.2
100
32.9
96.8
0
100
96.7
100
9.1
95
95
24.6
0
100
100
8.0
8.0
0
4.6
50.8
50.8
0
48.1
0
0
100
1.1
0
52.2
0
0
0
0
Percentage of
Cometab.
(%)
100
100
86.0
99
100
100
100
5.0
100
100
38.9
100
90.4
100
0
100
100
100
0
97.4
97.4
100
0
100
100
62.2
100
11.0
19.1
100
100
0
100
15.4
15.4
100
61.7
0
100
0
0
0
50
degradation
PCB
(%)
100
100
45.5
100
100
100
100
6.4
100
98.6
10.4
100
58.9
100
0
100
100
100
0
90
100
57.4
0
100
100
35.2
35.2
0
0
78.1
78.1
0
64.3
13.6
13.6
100
38.7
20.6
75.2
0
0
0
0
15 day
Cometab.
(%)
100
100
78.2
100
100
100
100
9.9
100
100
68.9
100
95.4
100
0
100
100
100
0
90
100
100
0
100
100
91.2
100
0
77.6
100
100
22.1
100
35.9
36.9
100
100
0
100
0
0
0
50
                                           78

-------
10
                    5.0-
                    4.5-
                    4.0-
                    3.5-
        Percentage
        relative
        area
                    3.0-
                    2.5-
                    2.0-
                     1.5-
                     1.0-
                     0.5-
                                                                  15
20            25            30

     Retention time (min)
                                                                                                                         35
                                                                                                                                       40
                                                                                                                                                     45
                                                                                                                                                                   50
           Figure 31.    Relative distribution of PCB isomers in the soil and sludge extracts with respect to retention time.  The number over the points are the percentage degrada-
                        tion of that peak in the cometabolization study (table 22).

-------
because different cultures have been shown in this study to degrade the vari-
ous PCB isomers differently.

     Several things could lead to error in this interpretation.  First, the
soil was dried in an oven, which may have volatilized some of the lower
chlorine isomers.  Second, the data are not quantitative in the sense that
one cannot relate the total amounts of a PCB isomer in the sludge and how
much sludge was applied to the amount of that PCB isomer found in the soil.

Other Observations

     It has been noted that PCBs adsorb to glass surfaces (Gresshoff,
Mahanty, and Gortner, 1977).  In the blanks that were run in this study,
however, no adsorption of PCBs was noted.  One possible explanation of this
is that when microorganisms are present in the water, the PCBs adsorb pre-
ferentially onto them rather than on the glass surface.

     Several other points must be clarified when interpreting the results
of this study.  First, a mixed culture of microorganisms was used in con-
trast to the pure cultures of PCB-degrading microbes used by most other
researchers (Lunt and Evans, 1970; Catelani et al., 1971, 1973; Gibson et
al., 1973; Wallnofer et al., 1973; Ahmed and Focht, 1973a, 1973b; Baxter
et al., 1975; and Furukawa, Tonomura, and Kamibayashi, 1978).  Therefore,
the more rapid degradation of PCBs found in this study compared.to that
previously reported may be partly attributed to the use of the mixed culture.
In addition to the PCB-degrading microbes in such a culture, many other
species of microorganisms may be present that could speed the degradation
of the PCBs by removing potentially inhibitory intermediates.  This mixed-
culture approach is closer to the conditions that exist in the environment
than a pure culture.

     Second, only water-soluble PCBs were utilized.   Water-soluble and total
PCBs have different relative proportions of PCB isomers (Haque, Schmedding,
and Freed, 1974).  Therefore, the degradation of water-soluble PCBs is less
than would be achieved for total PCBs.   Because most potential migration
of PCBs would be in an aqueous environment, the degradation of the water-
soluble isomers are the most environmentally significant.

     The PCB degradation studies all showed the same general trends of other
research (Ahmed and Focht, 1973a, 1973b; Baxter et  al., 1975; Furukawa,
Tonomura, and Kamibayashi, 1978; Tucker, Saeger, and Hicks, 1975; Mihashi
et al., 1975); namely, the lower chlorinated isomers, in general, were more
easily degraded than higher chlorinated ones, and different isomers with
the same number of chlorines degraded at different  rates.  But perhaps most
significant in this study is that PCBs  can be degraded to a large extent in
a short period of time by mixed cultures of microorganisms.   In addition,
if a PCB-degrading culture is used, the microbes will cometabolize PCBs
while using another compound as a source of energy.

     Because two of the mixed cultures  were enriched from soil or sediment
that contained PCBs, and because it appears that there were already PCB-
degrading microbes present, PCBs are expected to be degraded to some extent


                                    80

-------
in the environment.   Koeman et al.  (1969) pointed out that most PCB resid-
uals in animals yield GC chromatographs similar to Aroclor 1260, a rather
highly chlorinated series of PCB isomers.  This has also been noted by other
researchers (Smith,  K. E., Illinois Natural  History Survey, personal com-
munication) since then.  Because Aroclor 1242, a relatively lower chlorinated
series of isomers, accounted for the majority of PCBs produced and marketed
(HEW, 1972), it would appear that the lower chlorinated isomers are degraded
in the environment to some extent.

     The isolation from soil or sediment of microorganisms that can degrade
PCBs readily suggests the possibility of their future inoculation into
wastes containing PCBs before disposal, or biological digestion prior to
disposal.
                                    81

-------
                                  SECTION  9

                      EFFECT OF HUMIC ACID AND SOIL ON
                      VOLATILIZATION OF  PCBs  FROM WATER1

 RESULTS AND  DISCUSSION

      PCBs  are  known to have low solubility  in water (MacKay and Wolkoff,
 1973) and  to have a pronounced tendency  to  adsorb on particulate matter
 (Griffin et  al.,  1978).  The aromatic  character of PCBs is believed  to cause
 them to be adsorbed onto humic substances by hydrophobic interactions.
 Though the structure of humic acid is  still  under extensive investigations,
 a major portion  of humic acid is speculated to be constructed by cross!inked
 phenolics  and  thus to have a distinct  aromatic character.

      The PCB volatility measurements given  in table 23 show that water-
 soluble humic  acid attenuated the volatility of PCBs from aqueous  solution.
 The attenuation  does not seem to be very effective, however, because as
 much as 500  mg/L of humic acid were needed  to reduce the volatility  to 73
 percent of that  of the control, that is, from pure water as shown  by the
 chromatograms  in Figure 32.

      Recent  evidence (Reuter and Havlicek,  1978) has indicated that  humic
 acids are  composed predominantly of an aliphatic substructure (about 70
 percent),  whereas the aromatic moiety  was actually present in the  structure
 as only about  20 to 30 percent of the  total  composition.  This new struc-
 tural information about humic acid explained the experimental observation
 that relatively  few PCBs were adsorbed by dissolved humic acids.   Because
 the binding  of PCBs to humic acid in aqueous solution is relatively  weak,
 the PCBs in  solution and those bound to  humic acid actually formed a dynamic
     TABLE 23.  Relative amount of PCB volatilized from water containing humic acid or soil compared to control.


Control              50 mg/L humic acid                            500 mg/L humic acid
 100%                   87%                                      73%

                  1000 mg/L (47.30)a                5000 mg/L                  6400 mg/L
Control                 Catlin soil                     (236.5r                    (302.7)
 100%                   86%                        50%                       21%

 The numbers in parentheses indicate mg/L of total organic C content.


 1 Authors:   Hsi Meng, E.S.K. Chian, and  R.  A.  Griffin


                                       82


-------
00
CO
                                                         Q.
                                                         O
                                                         h-
                                                                                        B
                                                                                                                                                                ISGS 1978
              Figure 32.    Comparison of attenuation of volatilized PCBs by dissolved humic acid addition. (A) Control. (B) 50 mg/L dissolved humic acid added. (C) 500 mg/L

                           dissolved humic acid added.

-------
equilibrium.   Therefore,  the binding of PCBs by soluble humic  substances
does not  substantially reduce the concentration of free PCBs.   The  volatili-
zation  of PCBs is  thus not effectively influenced by  the  presence  of soluble
humic acids at a pH  of 12; however, this may not be the case when  the humic
acid is present as colloids at a lower pH.  Adsorption of dissolved  PCBs
on the  particulate and nonionized molecules, such as  humic substances, is
expected  to be far more effective than that on the ionized molecules in
true solution;  this  observation is based on the results of this  study of
PCB adsorption onto  Cat!in soil and soluble humic acid.

     Attenuation of  the volatility of water-soluble PCBs by soil adsorption
resulted  in a  greater reduction of the volatile fractions of PCBs  from the
aqueous solution (Table 23) than was the case for the soluble  humic  acid.
This analysis  was  based on the total organic content  in the solution (see
the numbers in  parentheses in Table 23).  Therefore,  the binding of  PCBs
to solids  does  seem  to effectively reduce the amount  of PCB volatilized as
shown by  the results of the Catlin soil addition (Table 23).   Thus the
adsorption of  PCBs on the particulate matter was more effective  than that
on the  soluble  humic matter as shown by the chromatograms in Figure  33.
Table 24  compares  our earlier soil adsorption study (Griffin et  al., 1978)
to the  soil volatility attenuation study presented here.

     As indicated  in Table 24, the results of volatilization and adsorption
are parallel in that for  each the attenuation was more efficient by  adding
a larger  amount of soil.   This correlation further implies that attenuation
of PCB volatilization is  directly related to binding of the PCBs to  the
soil  solids by  an  adsorption mechanism.

     The  conditions  of this experimental study were somewhat different than
would be  expected  under field conditions; however, a water-logged soil  could
result  in  a condition that resembles those conditions.  Also a landfill
site producing  leachate may closely resemble the conditions of our experi-
ments with humic acid,  because an appreciable amount of soluble humic  sub-
stances has been found  in landfill leachates (Chian, 1977).

Change in  Composition of  the Volatile PCBs

     The  percentage  of  lower-chlorinated PCBs in the samples was evaluated.
The early-eluting  larger  peaks in the chromatograms (Figs. 32  and 33)
represented the more volatile and lower-chlorinated PCBs.   The area  under
 TABLE 24.  The relative attenuation of volatilization of PCBs by soil as compared with the attenuation due to adsorption
          from solution by the same soil.


                        1 000 mg/L Catlin soil              5000 mg/L              6400 mg/L
Adsorption attenuation               85                      55                   42

Volatilization attenuation              86                      50                   21
                                     84

-------
00
(Jl
                                                                                                                                                                 ISGS1978
            Figure 33.   Comparison of attenuation of volatilized PCBs by soil addition (see figure 32 for the control).  (A) 1000 mg/L Catlin soil added. (B) 5000 mg/L Catlin soil

                        added. (C) 6400 mg/L Catlin soil added.

-------
 these GC peaks  was  summed  up and  compared  with  the  total  area  of all  the  GC
 peaks in the chromatogram.   The results  of direct injection  of PCB standards
 dissolved in hexane containing 500  ppm Aroclor  1242 showed 52  percent lower-
 chlorinated  PCBs, whereas  purging of  water saturated with PCBs yielded 80
 percent  lower-chlorinated  PCBs in the traps.  By  adding  50 mg/L of humic  acid,
 82  percent of the lower-chlorinated PCBs were found in the traps.   Addition
 of  500 mg/L  of  humic acid  yielded 87  percent of the lower-chlorinated PCB
 isomers  in the  trap.   By the same token, the percentages  of  lower-chlorinated
 PCBs  found in the traps after soil  additions were 80, 79, 74 and 73 percent,
 respectively, with  0,  1000,  5000, and 6400 mg/L soil additions.

       The higher-chlorinated PCBs were adsorbed more effectively by the
 soluble  humic acids;  this  can be  explained by the strong  hydrophobic  bonding
 of  the PCBs  to  the  humic substances.   The  affinity  of higher-chlorinated
 PCBs  to  the  humic substances is reflected  by an increased volatile portion
 of  the lower-chlorinated isomers; however,  the  soil  addition experiments
 showed a  small  decrease and  a leveling off of the amount  of  PCBs  in the traps
 in  the lower-chlorinated form.  This  is  speculated  to be  due to limited
 penetration  of  the  smaller lower-chlorinated isomers  into pores  of the inter-
 layer and  edge  regions of clay minerals  contained in the  Catlin  soil,  and
 thus  a slightly more efficient retention against  volatilization.

       The  most  visible change in  isomer  distribution comes from  comparing
 the hexane-soluble  PCBs and  water saturated with  PCBs.  This can be explained
 by the higher volatility of  the lower-chlorinated isomers and their higher
 solubility in water.

 Comparison with Theoretical  Volatilization Process

      MacKay and Wolkoff (1973)  predicated the  half-life of Aroclor 1242
 volatilized from water to be  6 hours.   Therefore,  the quantity volatilized
 in 20 minutes would correspond to 3 percent of the Aroclor added to the
 purging flask used  in this  study.   They pointed out that the high activity
 coefficient of a sparingly soluble compound in water favors  a high  rate of
 evaporation.   The experimental conditions of this  study agreed with the basic
 assumptions outlined by MacKay and Wolkoff (1973).  When  0.5  yg of PCBs
were introduced  into water followed by purging,  the amount of PCBs volatilized
was  found to be  0.5 percent.   The  trapping efficiency was found to be 71 per-
 cent as evaluated by concentrating a  known amount of PCB in hexane  and
 following  the same  analytical protocol.  Thus,  the actual evaporation
 efficiency was 0.7  percent.  Since 1 vl of the  5 yL  solution was intro-
 duced onto the GC column, 0.7 percent  times 5 gives the total evaporation
 efficiency of 3.5 percent.   This  agrees well with  the value of 3 percent
 volatilized from pure water based on the theoretical calculation of MacKay
 and  Wolkoff's (1973) model.  Similar calculations  for the other experiments
 allow for comparisons.  For example, volatilization of PCBs  from solutions
 containing 500 mg/L humic acid was reduced to 2.6  percent, and volatiliza-
 tion was  reduced to 0.74 percent  from suspensions  containing  6400 mg/L soil.
                                     86

-------
                                        REFERENCES

Ahmed, A. K.,  1976, PCBs in the environment:   Environment, v.  18, p.  6.
Ahmed, M., and D.  D.  Focht, 1973a, Degradation of polychlorinated biphenyls by two species
     of Achromobacter:   Canadian Journal  of Microbiology, v.  19, p.  47-52.
Ahmed, M., and D.  D.  Focht, 1973b, Oxidation of polychlorinated biphenyls by Achromobacter PCB:
     Bulletin of Environmental Contamination and Toxicology, v. 10, p. 70-72.
Ahnoff,  M., and B. Josefsson,  1975, Clean up procedures for PCB analysis on river water
     extracts:  Bulletin of Environmental Contamination and Toxicology, v. 13, no. 2, p.  Ib9-lbb.
ANSC, 1976, American  National  Standards Committee, Askarel Committee responds to PCB peril:
     American National  Standards Institute Reporter, v. 10, p. 3.
Armour, J. A., 1973,  Quantitative pe^chlorinated of polychlorinated biphenyls as a method
     for confirmatory residue measurement and identification:  Journal of the Association of
     Official Analytical Chemists, v. 56, no. 4, p. 987-993.
Armour, J  A., and J. A. Burke, 1970, Method for separating polychlorinated biphenyls from
     DDT and its analogs:  Journal of the Association of Official Analytical Chemists, v. 53,
     no. 4, p. 761-768.
Ballschmiter, K.,  C.  H. Unglert, and H. J. Neul, 1977, Abbau von chlorierten aromaten:
     Mikrobiologischer abbau der polychlorierten biphenyle (PCB): Chemosphere, v. 6, p.  51.
Baxter,  R. A., P.  E.  Gilbert,  R. A. Lidgett, J. H. Mainprize, and H. A.  Vodden, 1975, The
     degradation of polychlorinated biphenyls by microorganisms:  Science of the Total
     Environment,  v.  4, p. 53.                                                             ,
Beezhold, F. L., and V. F. Stout, 1973, The use and effect of mixed standards on the quantitation
     of  polychlorinated biphenyls:  Bulletin of Environmefltal Contamination and Toxicology,
     v.  10, no. 1, p. 10-15.
Berg, 0. W.,'P. L. Diosady, and A. V. Rees, 1972, Column chromatographic separation of poly-
     chlorinated biphenyls from chlorinated hydrocarbon pesticides and their subsequent gas
     chromatographic quantition in terms of derivatives:  Bulletin of Environmental Contamina-
     tion and Toxicology, v. 7, no. 6, p. 338-347.
Berg, 0. W., A. V. Rees, and M. S. Ali, 1978, Simplified techniques for the perchlorination
     of  polychlorinated biphenyls in environmental samples:  Abstracts from the 29th Annual
     Pittsburgh Conference, March 2, 1978, Cleveland, Ohio, p.  579.
Blakemore,  R. P., and A. E. Carey, 1978, Effects of polychlorinated biphenyls on growth and
     respiration of heterotrophic marine bacteria:  Applied Environmental Microbiology, v. 35,
     p.  323.
Blakemore,  R. P.,  1978, Effects of polychlorinated biphenyls on macro-molecular synthesis  by
     a  heterotrophic marine bacterium:  Applied Environmental Microbiology, v. 35, p. 329.
Bourquin, A. W., and S. Cassidy,  1975, Effect of polychlorinated biphenyl formulations on  the
     growth of estuarine bacteria:  Applied Microbiology, v. 29, p. 125.
Bower,  C. A., and  F. B. Gschwend, 1952, Ethylene glycol retention by  soils as a measure of
     surface area and interlaying swelling:  Soil Science Society of  America, Proceedings,
     v.  16, p. 342-345.
Briggs,  G.  G., 1973, A  simple  relationship between soil adsorption of organic chemicals and
     their  octenol/water partition coefficients:  Proceedings  7th British  Insecticide and
      Fungicide Conference, British Crop Protection Council, London WIN6DT, p. 83-86.
Burkard,  L. P., and D.  E. Armstrong, 1978, A modified  perchlorination procedure for the
     determination of polychlorinated biphenyls:  Technical Completion Report, Project
     No.  OWRT A-067-Wis. Water Resources Center, University of  Wisconsin, Madison, Wisconsin.
Burse,  V. W., R. D. Kimbrough,  E. C. Villanueva, R. W.  Jennings, R. E. Linder, and G. W.  Sovocoal,
      1974,  Polychlorinated biphenyls storage, distribution, and recovery:  Liver Morphology
     After  Prolonged Dietary  Ingestion:  Archives of  Environmental Health, v. 29,  p.  301.
Catelani, D., A. Colombi, C.  Sorlini, and V. Treccani,  1973,  2-hydroxy-6-oxo-6-phenylhexa-2,
     4-dienoate:   The meta-clevage produce from 2, 3-dihydroxybiphenyl by  Pseudomonas put-ida:
     Journal of Biochemistry,  v.  134, p. 1063-1066.
Catelani, D., C." Sorlini, and'V7"freccan'i'," 1971, The metabolism of biphenyl by Pseudomonas
     put-ida:  Journal of Biochemistry, v. 134,  p. 1063-1066.
Catelani, D., C. Sorlini, and  V. Treccani, 1971, The metabolism of biphenyl by Pseudomonas
     put-ida:  Experientia, v.  27, p. 1173-1174.
Chian,  E.S.K., 1977, Stability of organic matter in landfill leachates:  Water Research,
     v.  11, p. 225-232.
Cook, J.  W., 1972, Some chemical aspects of polychlorinated biphenyls (PCBs):  Environmental
     Health Perspective, v. 1:3.
                                               87

-------
 Curley, A., V. W. Burse, M.  E. Grim, R.  W.  Jennings, and R.  E.  Linder, 1971, Polychlorinated
      biphenyls:  Distribution and storage in body fluids and tissues of Sherman rats-
      Environmental Research, v. 4, p.  481.
 Curley,_A., V. W. Burse, R.  W. Jennings, E.  C.  Villanueva,  and  R.  D. Kimbrough, 1975,
      Evidence of tetrachlorodibenzofuran (TCDF)  in Aroclor  1254 and the urine of rats  following
      dietary exposure to Aroclor 1254:   Bulletin of Environmental  Contamination and Toxicology,
      V. IT") P « I JO lOO.
 Duke,J  W., J. J. Lowe, and A. J. Wilson,  Jr.,  1970,  A polychlorinated biphenyl  (Aroclor 1254)
      in the water, sediment, and biota of Escambi  Bay, Florida:   Bulletin  of Environmental
      Contamination and Toxicology, v.  5, no.  2,  p.  171.
 Duvall   D.  S., and W.  A. Rubey, 1977,  Laboratory evaluation  of  high-temperature destruction
      of polychlonnated  biphenyls and  related compounds:  U.S.  Environmental  Protection  Agency
      Report 600/2-77/228,  Cincinnati,  OH  45268.
 EHRC, 1976,  Polychlorinated  biphenyls:   Health effects and  recommendations:   Environmental
      Health Resource  Center, Illinois  Institute  for Environmental  Duality,  IIEQ Document No
      76/05,  Chicago,  Illinois.
 Eichelberger,  J., 1971,  Persistence of Aroclors  in  surface waters:   PCB Newsletter,  no.  3,

 ES&T, 1977,  Outlook,  legislation:   The toxic  substances  control  act is  effective  this  month-
      Environmental  Science and  Technology, v.  11,  p. 28.
 Furukawa, K.,  and F. Matsumura,  1976, Microbial  Metabolism of polychlorinated  biphenyls-
      Studies  on the relative degradability of polychlorinated biphenyl  components by
      AlkaUgenes  Sp.:  Journal  of Agricultural and  Food  Chemistry,  v.  24, p.  251.
 Furukawa, K.,  K.  Tonomura, and  A.  Kamibayashi, 1978, Effect  of chlorine substitution on  the
      biodegradability  of polychlorinated biphenyls:   Applied  Environmental Microbiology,  v. 35,

 Gibson, D. J.,  R.  L. Roberts, M.  C.  Wells, and V. M. Kobal,  1973, Oxidation of  biphenyl  by
 n    *  B*tJ«™°*t« species:   Biochemical and Biophysical Research  Communications,  v.  50,' p. 211.
 uuskoter  H   J.,  1965,  Electronic  low temperature  ashing of bituminous  coal:   Fuel, v.  44,

 Grant   D  L.   W E.J. Phillips,  and  D. C.  Villeneuve, 1971, Metabolism of a polychlorinated
      biphenyl  (Aroclor 1254) mixture in  the rat:  Bulletin of Environmental Contamination and
      Toxicology,  v. 6, no. 2, p.  102.
 Gray  P.H.H    and  H. G.  Thornton,  1928,  Soil  bacteria  that decompose certain aromatic  compounds:
      Zentralblatt  Bacteriologie Parasitenkunde, Abteilung II, v  73  p  74
 Gresshoff  P. M., H. K.  Mahanty, and E. Gortner, 1977, Fate of polychlorinated biphenyl
      lArocIor  1242) in an experimental  study and its significance to the natural environment:
      Bulletin of  Environmental Contamination and Toxicology, v.  17, no. 6,  p. 686
briffin  R.  A.,_R. Clark, M.  Lee,  and E.S.K.  Chian,  1978, Disposal  and  removal  of polychlorinated
     biphenyls in  soil, -in  David Schultz  [ed.],  Land disposal  of  hazardous  waste-  U S
     Environmental Protection Agency, Cincinnati,  OH 45268,  EPA-600/9-78-016,  p  169-181
 briffin, R.  A., and N.  F. Shimp, 1978,  Attenuation of pollutants in municipal landfill  leachate
     by clay minerals:    U.S.  Environmental Protection Agency Publication EPA-600/2-78-157
     Cincinnati, OH  45258, 147 p.
 Grob, K.,  1975, Glass  capillary columns in gas chromatography, a tool and a technique-
     Chromatographia,  v.  8, p. 423.
 Grob, K.,  G. Grob, and K. Grob, Jr., 1977, New improved carbonate deposition technique  for
     coating wall  coated open tubular columns:  Chromatographia,  v.  10, p.  181.
Gustafson,  C.  G.,  1970, PCBs—prevalent and  persistent:  Environmental  Science and Technology
     v.  4,  no.  10, p.  814-819.
 Gustafson, C. G.,  [ed.], 1972, Proceedings of Symposium on PCBs—still  prevalent—Still
     persistent:  164th American Chemical Society Meeting, New York, NY, September,  1972.
Hammer, D.  I., J.  F.  Finklea, L. E. Priester, J.  E.  Keil, S.  H.  Sandifer and K.  Bridbord, 1972,
     Polychlorinated biphenyl in the plasma  and  hair of refuse workers:  Environmental  Health
     Perspective,  v.  1, p.  83.
Hammond, A.  L., 1972,  Chemical pollutioli:  Polychlorinated biphenyls1:  Science, v. 175, p. 155-156.
Hansen, D.  J., P.  R.  Parrish, J. I. Lowe, A.  J.  Wilson, Jr.,  and  P.  D.  Wilson, 1971, Chronic
     toxicity, uptake,  and  retention of Aroclor 1254 in two  estuarine fishes:   Bulletin of
     Environmental Contamination and Toxicology,  v.  6,  no.  2, p.  113-]19.
Haque, R.,  and D.  Schmedding, 1975, A method  of measuring the water solubility of hydrophobic
     chemicals:  Solubility of five polychlorinated biphenyls:   Bulletin of Environmental
     Contamination and  Toxicology, v. 14, no.  1,  p.  13-18.
Haque, R.,  D.  W. Schmedding,  and V. H.  Freed,  1974,  Aqueous  solubility, adsorption,  and vapor
     behavior of polychlorinated biphenyl Aroclor 1254:  Environmental  Science and Technology,
     v.  8,  p.  139-142.


                                              88

-------
Haque, R.,  and 0.  Schmedding,  1976,  Studies  on  the adsorption  of  selected  polychlorinated
     biphenyl  isomers on several  surfaces:   Journal  of Environmental  Science  and  Health,

Harvey!1 G.Y, and w! G. Steinhauer, 1975,  Biochemistry of PCB and  DDT in  the North  Atlantic:
     Proceedings of the 2nd International  Conference on Environmental  Biogeochemistry,  Hamilton,
     Canada,  p. 203-225.                                                   ,      ..  ...   ,
Helling  C  S , and B. C.  Turner, 1968, Pesticide mobility:   Determination by  soil  thin-layer

HelliS:Ta19:a?^ia!CpesCticide2moKility1n3s0ils I.   Parameters of thin-layer chromatography:
     Soil  Science Society of America Proceedings, v. 35, p.  732-737.
Helling,,  C. S., 1971b, Pesticide mobility in soils II.  Applications of soil thin-layer
     chromatography:  Soil Science Society of America Proceedings,  v. Jb,p.  /J/-/4J.
Helling, C. S  , 1971c,  Pesticide mobility in soils  III.  Influence of soil properties:   Soil
      Science Society  of America Proceedings, v. 35, p. 743-747.
Henderson, H., 1978, The dilemma of Wilsonville:  Illinois Times, v.  4, p. 4-7.
Herbicide Handbook, 1974, Weed Society of America, 3rd Ed., 425 Illinois Building, 113  N.  Neil
     St., Champaign,  IL  61820.                                        ,,             ^
Herbst, E., I. Scheunert, W. Klein, and F. Korte, 1977, Fate of PCBs—14C in sewage treatment-
     Laboratory experiments with activated sludge:  Chemosphere,  v. 6, p.  725.
Hesse  J  L , 1971, Summary of the Michigan Water Resources Commission PCB monitoring efforts:
     Water Quality Control Division, Water Resources Commission,  Bureau of Water Management,
      Lansing, Michigan.                                                   .          .. . .   ,
HEW,  Department of Health, Education, and Welfare,  1972, PCBs and the Environment:  National
      Technical  Information Service,  COM-72-10419.
Hoover, T. B., July,  1971, Water solubility of PCB  isomers:  PCB Newsletter, no. 3,  p.  4-b.
Horvath, R. S., 1972, Microbiol co-metabolism and the degradation of organic  compounds in
      nature:  Bacteriological  Reviews, v. 36, p.  146.
Hughes, G. M.,  R. A.  London, and R. N. Farvolden, 1971, Hydrogeology of solid waste disposal
      sites  in northeastern  Illinois:  U.S. Environmental Protection Agency,  Solid Waste Manage-
      ment Series, Report  SW-12d., 154 p.                                                 ,Drae\
Hutzinger,  0.,  S. Safe, and V. Zitko, 1972, Photochemical degradation of  chlorobiphenyls (PCBs).
      Environmental  Health  Perspectives, April 15-20.
Hutzinger,  D.,  S. Safe, and V. Zitko,  1974, The  Chemistry of  PCBs:   CRC Press, Cleveland, Ohio.
Iwata, Y.,  F.  A.  Gunther,  and  W. E. Westlake, 1974, Uptake of a  PCB  (Aroclor 1254) from soil
      by carrots under field conditions:   Bulletin of  Environmental  Contamination  and Toxicology,
      v. 11,  p.  523-528.                                          .         ^    ,  un   .  ^  ,
 Iwata Y  ,  W   E.  Westlake,  and F. A. Gunther, 1973, Varying persistence of polychlonnated
      biphenyls  in six California soils under laboratory conditions:   Bulletin of  Environmental
      Contamination  and  Toxicology,  v.  9,  no. 4,  p.  204-211.
Jensen, S.,  1966,  Report  of a  new chemical  hazard:  New Scientist, v.  32, p.  bl^.
Jensen, S.  A.,  A.  G.  Johnels,  M. Olsson,  and G.  Otter!ind, 1969, DDT  and  PCB in marine animals
      from  Swedish waters:   Nature,  v.  224,  p. 247-250.
Jordan, D.,  1977, The town  dilemma:  Environment, v.  19,  p. 6-15.
Kaiser, K.L.E., and  T.  S.  Wong,  1974,  Bacterial  degradation of polychlorinated biphenyls.
      I.   Identification of some metabolic products  from AroclorK 1242:  Bulletin  of Environ-
      mental  Contamination  and  Toxicology, v. 11,  p. 291.
Kaiser, R.,  1976,  Zur richtigen  Messung und Bewertung  von Gutekennzahlen  in  der Chromatographie:
      Die  reale  Trennstutenzahl,  die Tr  inzahl,  die  Dosiergutekennzahl.  Teil  1:   Gas-
      Chromatographie:   Chromatographia,  v.  9,  p.  337.
 Koeman, J.  H.,  M.  C.  Ten  Noever  De  Braun, and R.  H. De  Vos, 1969,  Chlorinated  Biphenyls in
      fish,  mussels  and  birds  from the  River Rhine and  the Netherlands  coastal  area:  Nature,
      v. 221,  p.  1126-1128.
 Kuhn, J.  K.,  F.  Fiene,  and R.  Harvey,  1978, Geochemical evaluation and  characterization of  a
      Pittsburgh No.  8 and a  Rosebud Seam  Coal:   Report for DOE Research Contract  EY-77-X-21-2155,
      Morgantown Energy Technology  Center, NTIS  No.  METC/CR-78/8.
 Lawrence,  J.,  and H.  M. Tosine,  1976,  Adsorption of polychlorinated  biphenyls  from  aqueous
      solutions and  sewage:   Environmental Science and Technology,  v.  10,  no. 4,  p.  381-383.
 Leis, W.  M.,  W.  F.  Beers, J.  M.  Davidson, and  G.  D. Knowles,  1978,  Migration of  PCBs by ground-
      water transport—A case  study  of  twelve  landfills and dredge  disposal  sites  in  the upper
      Hudson Valley, New  York:  Conference  of  Applied Research and Practice  on Municipal  and
      Industrial  Waste,  Box 5571,  Madison, WI   53705,  Sept.  10-13,  1978, p.  539-548.
 Lee, M.  C., R.  A. Griffin, M.  L.  Miller,  and  E.S.K. Chian,  1979, Adsorption  of water soluble
      Aroclor 1242 and used capacitor  fluid  by  soil  materials  and coal  chars:  Journal  of
      Environmental  Science and Health,  (In  press).



                                                89

-------
Lidgett, R. A., and H. A. Vodden, 1970, PCB—The environmental problem:  PCB Conference,
     Wenner-Gren Center, National Environmental Protection Board Stockholm, September 29, 1970,
     p. 88-96.
Lloyd, J. W., R. M. Moore, B. S. Woolf, and H. P. Stein, 1976, Polychlorinated biphenyls:
     Journal of Microbiology, v. 18, p. 109.
Lunt, 0., and W. C. Evans, 1970, The microbial metabolism of biphenyl:  Journal of Biochemistry,
     v. 118, 54 p.
MacKay, D., and A. W. Wolkoff, 1973, Rate of evaporation of low solubility contaminants from
     water bodies to atmosphere:  Environmental Science and Technology, v. 7, p. 611-614.
Maslansky, S. P., R. F. Thomas, R. F. Bonner, Jr., and P. L. Busch, 1978, Selection and design
     criteria of secure landfill facilities for the containment of PCB contaminated river
     debris:  Conference on Applied Research and Practice on Municipal and Industrial Waste,
     Box 5571, Madison, WI  53705, September 10-13, 1978, p. 635-644.
Mihashi, K., T.  Hashinaya, M. Dazai, J. I to, and T. Misono, 1975, Biodegradability of poly-
     chlorobiphenyl:  Kogyo Gijustsuiu Biseibntsu Kogyo Gijustsu Kenkyusho Kenkyn Hokoku,
     v. 46, p. 65.
Moon, D. K., I.  W. Leighton, and D.  A. Huebner, 1976, New England PCB waste management study,
     New Bedford:  Solid Waste Program, Air and Hazardous Materials Division, Region I,
     U.S. - EPA, J.F.K. Federal Building, Boston, MA  02203.
Murado, M.  A., C. Tejedor, and G. Baluja, 1976, Interactions between polychlorinated biphenyls
     (PCBs) and soil microfungi.  Effects of Aroclor 1254 and other PCBs on AspergUlus flavus
     cultures:  Bulletin of Environmental Contamination and Toxicology, v. 15, p. 768.
Nelson, N., P. B. Hammond, I.C.T. Nisbet, A. F. Sarofim, and W. H.  Drury, 1972, Polychlorinated
     biphenyls:   Environmental impact:  Environmental Research, v.  5, p. 249-362.
Nisbet, I.C.T.,' and A.  F.  Sarofim, 1972, Rates and routes of transport of PCBs in the environment:
     Environmental Health Perspectives; no.  1, 21-38.
Omori,  T.,  and M. Alexander, 1978, Bacterial dehalogenation of halogenated alkanes and fatty
     acids:  Applied Environmental Microbiology, v. 35, p.  867.
Paris,  D. F., W. C. Steen, and G. L.  Baughman, 1978, Role of physico-chemical properties of
     Aroclors 1016 and 1242 in determining their fate and transport in aquatic environments:
     Chemosphere, v.  4, p. 319-325.
Peakall, D. B.,  and J.  L.  Lincer, 1970, Polychlorinated biphenyls:   Another long-life wide-
     spread chemical  in the environment:  Bioscience, v.  20, p. 958-964.
Persson, B., 1971, Uptake of chlorinated hydrocarbons by Whitethroats Sylvia eommunis.  aathl.
     in areas sprayed with DDT:  Ornis Scandinavica, v.  2,  p.  127-135.
Reeves, M., C. W. Francis, J. 0.  Duguid, 1977, Quantitative analysis of soil  chromatography.
     I. Water and radionuclide transport:  Environmental  Sciences Division Publication no.  1105,
     Oak Ridge National Laboratory,  Oak Ridge, TN  37830, 170 p.
Reinert, R. E.,  1970, Pesticide concentrations in Great Lake fish:   Pesticides Monitoring
     Journal, v. 3, p.  233.
Reuter, J.  H., and S.  Havlicek, 1978, Private communications:   October, Georgia Institute
     of Technology.
Reynolds, L.  M., 1969,  Polychlorobiphenyls (PCBs)  and their interference with pesticide
     residue analysis:   Bulletin of  Environmental  Contamination and Toxicology, v.  4, no.  3,
     p.  128-143.
Reynolds, L.  M., 1971,  Pesticide residue analysis  in the presence of polychlorobiphenyls
     (PCBs):   Residue Reviews, v.  34, p.  27-57.
Risebrough, R.  W., P.  Reiche, and H.  S.  Olcott,  1969, Current progress in the determination
     of the polychlorinated  biphenyls:  Bulletin of Environmental  Contamination and Toxicology,
     v.  4,  no.  4, p.  192-210.
Risebrough, R.  W., P.  Reiche, D.  B.  Peakall, S.  G.  Herman,  and  M.  N.  Kirven,  1968, Polychlor-
     inated biphenyls  in the global  ecosystem:   .Nature,  v.  220, p.  1098.
Robertson,  J.  M., and E.C.C.  Li,  1976,  Organic leachate threatens ground water quality:
     Water  and Sewage Works, February 1976,  p.  58-59.
Rote, J.  W.,  and P.  G.  Murphy, 1971,  A method for  quantisation  of polychlorinated biphenyl
     (PCB)  isomers:   Bulletin of Environmental  Contamination and  Toxicology,  v.  6, no.  4,
     p.  377-384.
Ruzo, L.  0.,  M.  J.  Zabik,  and R.  D.  Schiretz,  1972,  Polychlorinated biphenyls:   Photolysis
     of 3,  4,  3',  4'-tetrachlorobiphenyl  and 4,  4'-dichlorobiphenyl  in solution:   Bulletin
     of Environmental  Contamination  Toxicology,  v.  8, no. 4,  p.  217-218.
Safe, S., and 0. Hutzinger,  1971, Polychlorinated  biphenyls:   Photolysis  of 2,  4, 6,  2', 4', 6'-
     hexachlorobiphenyl:   Nature,  v.  232,  p.  641-642.
Sayler, G.  S.,  M.  Shou,  and  R.  R.  Colwell,.1977,  Growth  of  an  estuarine Pseudomonas sp.  on
     polychlorinated  biphenyls:   Microbial Ecology,  v.  3, p.  241-245.
                                              90

-------
Schact, R.  A., 1974, Pesticides in the Illinois  water of Lake  Michigan:   Ecological  Research
     Series, U.  S.  Environmental  Protection Agency Publication EPA-660/3-74-002,  Washington,  D.C.
Scharpenseel, H.  W., B.K.G.  Theng, and S.  Stephan, 1978, Polychlorinated  biphenyls  (^C)
     in soils:  Adsorption,  infiltration,  translocation, and decomposition,  in Environmental
     biochemistry and geomicrobiology, The terrestrial  environment:   Ann  Arbor
     Science Publishers, Inc., Ann Arbor,  MI  48106,  v.  2,  p.  619-637.
Sissons, D., and D. Welti, 1971,  Structural identification  of  polychlorinated  biphenyls  in
     commercial  mixtures by  gas-liquid chromatography,  nuclear magnetic  resonance,  and mass
     spectrometry:   Journal  of Chromatography, v.  60, p. 16-32.
Skrentny, R. F., R. W. Hemken, and H.  W.  Dorongh,  1971,  Silo sealants as  a source of poly-
     chlorobiphenyl (PBC) contamination of animal  feed:   Bulletin of Environmental  Contamination
     and Toxicology, v.  6, no. 5, p.  409-416.
Snyder, D., and R.  Reinert,  1971, Rapid separation of polychlorinated biphenyls from DDT and
     its analogues on silica gel:  Bulletin of Environmental Contamination and Toxicology,  v.  6,
     no. 5, p. 358-390.
Solly, S.R.B., and V.'Shanks, 1974, Polychlorinated biphenyls  and organochlorine pesticides
     in human fat in New Zealand:  New Auckland Journal  Science, v.  7, p. 535-544.
Stalling, D. L., and J.  N. Huckins, 1971,  Gas-liquid  chromatography-mass  spectrometry character-
     ization of polychlorinated biphenyls   (Aroclors)  and 36Cl-labeling of Aroclors  1248  and
     1254:   Journal of the Association of  Official Analytical  Chemists,  v. 54, no.  4, p.  801-807.
Sundstrom,  G., D. Hutzinger, and S. Safe,  1976, The metabolism of chlorobiphenyls:   Chemosphere,
     v. 5,  p. 267.
Texas Instruments, 1975, Program manual STI statistics library, Model SR-52 calculator:   Texas
     Instruments, Inc.,  P.O. B.ox 22283, Dallas, TX  75222,  107 p.
Thomas, J., Jr., and R.  R. Frost, 1971, Versatile apparatus for studying reactions  involving
     gas adsorption or evolution:  Illinois Academy of Science Transactions, v. 64, p. 248-253.
Thurston, A., 1971, Quantitative analysis  of PCBs:  PCB Newsletter, July, no.  3, p. 4.
Troise, F., and V. M. Kahn,  1978, High cost of legal  dumping leads some to organized crime:
     The Ground Water Newsletter, Jan. 31, 1978, Water'Information Center, Inc., Huntington,
     NY  11743, v. 7, no. 2.
Trotter, W. J., and S.J.V. Young, 1975, Limitation of the use  of antimony pentachloride for
     perchlorination of polychlorinated biphenyls:  Journal of the Association of Official
     Analytical Chemists, v. 58, p. 466-468.
Tucker, E.  S., W. J.  Litschgi, and W. M. Mees, 1975,  Migration of polychlorinated biphenyls
      in soil  induced  by percolating water:  Bulletin of Environmental Contamination and
     Toxicology, v.  13, p. 86-93.
Tucker,  E.  S., V.  M.  Saeger,  and  0. Hicks,  1975,  Activated sludge primary biodegradation of
      polychlorinated  biphenyls:   Bulletin  of  Environmental Contamination and Toxicology,
      v.  14,  p. 705-713.
Wallnofer,  P. R., G.  Engelhardt, S. Safe,  and 0.  Hutzinger, 1973, Microbial hydroxylation of
     4-chlorobiphenyl and 4-4'-dichlorobiphenyl:  Chemosphere, v. 2, p.  69-76.
Webb, R. G., and A. C. McCall, 1972,  Identities of polychlorinated biphenyl isomers in Aroclors:
     Journal of the Association of Official Analytical  Chemists, July, v. 55, no. 4, p.  746-752.
Widmark, G.,  1967, Possible interference by chlorinated biphenyls:  Journal of Association of
     Official Analytical Chemists, v. 50,  p. 1069.
Widmark, G.,  1968, The OECD study of  analysis of  PCB:   Report from the Institute of Analytical
     Chemistry, Stockholm, Sweden.
Wiencke, W., and J.  Roach, 1972,  Response  to GLC  detectors to the chlorine content in Aroclors,
     PCB Newsletter,  March, no. 4, p. 9-10.
Zitko, V.,  1970, Polychlorinated  biphenyls  (PCB)  solubilized in water by monionic surfactants
     for study of toxicity to aquatic animals:  Bulletin of Environmental Contamination and
     Toxicology, v.  5, no. 3, p.  279-285.
Zitko, V.,  1971, Polychlorinated biphenyls and organochlorine pesticides in some freshwater
     and marine fishes:  Bulletin of  Environmental Contamination and Toxicology, v. 6, no.  5,
     p. 464-470.
Zitko, V.,  0. Hutzinger, and  S. Safe, 1971, Retention times and electron-capture detector
      responses of  some  individual chlorobiphenyls:   Bulletin of Environmental Contamination
     and Toxicology,  v. 6, no. 2, p.  160-163.
                                               91

-------
                                APPENDIX


                          LIST OF PUBLICATIONS


Clark, R. R., E.S.K. Chian, and R. A. Griffin, 1979, Degradation of poly-
     chlorinated biphenyls by mixed microbial cultures:  Applied and Environ-
     mental Microbiology, v. 37, p. 680-685.

Griffin, R. A., A. K. Au, and E.S.K. Chian, 1979, Mobility of polychlorinated
     biphenyls and dicamba in soil materials:  Determination by soil thin-
     layer chromatography:  Proceedings 8th National Conference—Municipal
     and Industrial Sludge, March 19-21, 1979, Miami, FL.  Copyright by
     Information Transfer, Inc., Rockville, MD.

Griffin, R. A., and E.S.K. Chian, in press, Attenuation of water-soluble
     polychlorinated biphenyls by earth materials:  Illinois State Geological
     Survey Environmental Geology Note 86, 98 p.

Griffin, R. A., Robert Clark, Michael Lee, and Edward Chian, 1978, Disposal
     and removal of polychlorinated biphenyls in  soil, in David Schultz [ed.],
     Land disposal of hazardous waste:  Publication EPA-600/9-78-016,
     Cincinnati, OH  45268, p. 169-181.

Griffin, R. A., F. B. DeWalle, E.S.K. Chian, J. H. Kim, and A.  K. Au,  1977,
     Attenuation of PCBs by soil  materials and char wastes, in
     S. K.  Banerji [ed.], Management of gas and leachate in landfills:
     Publication EPA-600/9-77-026, U.S. Environmental  Protection Agency,
     Cincinnati, OH  45268, p. 208-217.

Lee, M. C., E.S.K. Chian, and R.  A.  Griffin, in press, Solubility of poly-
     chlorinated biphenyls and capacitor fluid in water:   Water Research.

Lee, M. C., R. A.  Griffin, M. L.  Miller, and E.S.K.  Chian, in press, Adsorp-
     tion of water soluble polychlorinated biphenyl  Aroclor 1242-and used
     capacitor fluid by soil  materials and coal chars:  Journal  of Environ-
     mental  Science and Health.

Meng, Hsi,  E.S.K.  Chian, and R.  A.  Griffin, (submitted),  Effect of humic acid
     and soil on volatilization of polychlorinated biphenyls  from water:
     Water  Research.
                                     92

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
 EPA-600/2-80-027
                                                           3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
 Attenuation of Water-Soluble Polychlorinated Biphenyls
 by Earth Materials
              5. REPORT DATE             .
               May 1980 (Issuing  Date)
              6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 R. A. Griffin  and  E.  S.  K. Chian
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Illinois  State  Geological Survey
 Urbana, Illinois  61801
              10. PROGRAM ELEMENT NO.
               C73D1C, SOS  #1,  Task 17
              11. CONTRACT/GRANT NO.

                R-804684-01-0
12, SPONSORIWG.AGENCY NAMEANDADDRESS .  .  ,            r.    Au
 Municipal Environmental  Research Laboratory --Cm.,OH
 Office of Research  and Development
 U.S. Environmental  Protection Agency
 Cincinnati, Ohio 45268
              13. TYPE OF REPORT AND PERIOD COVERED
                 Final
              14. SPONSORING AGENCY CODE
                 EPA/600/14
15. SUPPLEMENTARY NOTES

 Richard A. Carnes, Project Officer  (513/684-7871)
16. ABSTRACT

    The aqueous  solubility, adsorption,  mobility, microbial  degradation, and
 volatility of polychlorinated biphenyls (PCBs) were studied under laboratory con-
 ditions.  The dissolution of Aroclor  1242 in water required five months to reach
 equilibrium.  Generally, the water-soluble fractions of the PCB fluids were richer
 in the lower chlorinated isomers than in the original mixture  of isomers in the fluid.
 The solubilities  of Aroclor 1016, 1221, 1242, and 1254 were 906 ppb, 703 ppb, and
 =70 ppb, respectively.
    A simple linear relation described the adsorption of water-soluble PCBs by five
 earth materials and their low-temperature ashes.  An adsorption constant (K) unique
 to each adsorbent was obtained.  The  adsorption was strongly correlated to the total
 organic carbon  (TOC)  content and surface area of the earth  materials.  TOC was the
 dominant of these two earth material  properties by a ratio  greater than three to
 one.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                            c. COS AT I Field/Group
 Biodeterioration
 Mobility
 Solubility
   Hazardous Wastes
68C
13. DISTRIBUTION STATEMENT
 Release to Public
 19. SECURITY CLASS (ThisReport)
   Unclassified
                                                                         21. NO. OF PAGES
   101
                                              20. SECURITY CLASS (Thispage)

                                                 Unclassified
                                                                         22. PRICE
EPA Form 2220-1 (9-73)
93
                                                                 A U.S. GOVERNMENT PRINTING OFFICE: 1980-657-146/5700

-------