EPA-600/2-78-059
March 1978
Environmental Protection Technology Series
                 EVALUATION OF CHLORINATED
                      HYDROCARBON CATALYTIC
                       REDACTION TECHNOLOGY
                            industrial Environmental Research Laboratory
                                 Office of Research and Development
                                U.S. Environmental Protection Agency
                            Research Triangle Park, North Carolina 27711

-------
                 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.
                        EPA REVIEW NOTICE
 This report has been reviewed by the U.S. Environmental Protection Agency, and
 approved for publication. Approval does not signify that the contents necessarily
 reflect the views and policy of the Agency, nor does mention of trade names or
 commercial products constitute endorsement or recommendation for use.

 This document is available to the public through the National Technical Informa-
 tion Service. Springfield, Virginia 22161.

-------
                                     EPA-600/2-78-059
                                          March 1978
EVALUATION  OF CHLORINATED
   HYDROCARBON CATALYTIC
    REDUCTION TECHNOLOGY
                     by

            Mitchell D. Erickson and Eva D. Estes

               Research Triangle Institute
                 P. 0. Box 12194
            Research Triangle Park, N. C. 27709
               Contract No. 68-02-2612
                 Tasks 17 and 28
              Program Element No. 1BB610
            EPA Project Officer: David K. Oestreich

          Industrial Environmental Research Laboratory
            Office of Energy, Minerals, and Industry
          Research Triangle Park, North Carolina 27711
                  Prepared for

          U.S. ENVIRONMENTAL PROTECTION AGENCY
            Office of Research and Development
               Washington, D.C. 20460

-------
                    EVALUATION OF CHLORINATED HYDROCARBON
                       CATALYTIC REDUCTION TECHNOLOGY
                                     by
                            Mitchell D. Erickson
                                Eva D. Estes
                               John M, Harden
                         Research Triangle Institute
                Research Triangle Park, North Carolina  27709
                         EPA Contract No. 68^02-2612
                                RTI/1430 - 01 .F
            D. K. Oestreich

 U.S. Environmental Protection Agency    U.S. Environmental Protection Agency
Industrial Environmental Research Lab   Industrial Environmental Research Lab
  Research Triangle Park, NC  27711       Research Triangle Park, NC  27711
                         RESEARCH TRIANGLE INSTITUTE
                               P. 0. BOX 12194
                RESEARCH TRIANGLE PARK, NORTH CAROLINA  27709

-------
                                 DISCLAIMER
     This report has been reviewed by the Industrial Environmental Research
Laboratory, U.S. Environmental Protection Agency,  and approved for publication.
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.

-------
                                  ABSTRACT
     A control technique developed for the Environmental Protection Agency for
the catalytic reduction of chlorinated hydrocarbons (specifically polychlori-
nated biphenyls [PCBs] and chlorinated pesticides such as heptachlor and
endrin) was evaluated under laboratory conditions.  The technique involves
elution of polluted water at ambient temperature and at neutral pH through a
column containing a mixture of sand and copper/iron catalyst.  The evalua-
tion found that PCBs are not detectably reduced, but are chromatographi-
cally eluted from the column in order of increasing chlorination.  Thus,
early column eluate fractions were found to contain only the lower chlori-
nated PCBs, giving the illusion that partial reduction had occurred.
     The catalyst was found to partially reduce heptachlor and endrin.  The
carbon skeleton remained intact and chlorines were successively replaced by
hydrogens.  In a stirred flask of the catalytic mixture and pesticide-
spiked water at 60°, the reaction was found to be only partially successful
with most of the products accounted for as the parent compounds or the analogs
representing loss of one chlorine.
     The possibility of chromatographic elution of endrin and/or heptachlor
was briefly investigated using miniature columns and found to be insignificant.
A degradation product of heptachlor upon standing in water was found to be
not heptachlor epoxide, but rather C10HgClc02 (tentative)—apparently a dihy-
droxy derivative of heptachlor.
     An additional on-site demonstration of the catalytic reduction technology
was observed by an RTI and an EPA representative.  This program included obser-
vation of the test procedure, gas chromatography with electron capture detection
and subsequent GC/MS analysis of transported samples at the RTI laboratory.
     An evaluation of the reports generated during the development of the
catalytic reduction technique is presented.  It discusses errors and
omissions in the developmental experimental protocol that led to the
erroneous conclusions that the technique was applicable to PCBs.
                                     ii

-------
                                  CONTENTS


Disclaimer	     ±
Abstract	,	    ±±
Figures	    iv
Tables 	     v
Abbreviations  	    vi
Acknowledgments	   vii

I.    Introduction 	     1

II.   Summary and Conclusions   	     2

III.  Recommendations  	     3

IV.   Materials and Methods  	  .....     4
         Column Preparation  	     4
         Extraction	     4
         Analysis  	     4

V.    Evaluation of PCB Reduction Technology 	     6

VI.   Evaluation of Endrin/Heptachlor Reduction   	    11
         Stirred Flask Experiments  	    11
         Miniature Column Experiments   	    18
         Summary Evaluation of  Endrin and Heptachlor Reduction ....    18

VII.  Critique of Monthly Reports of Contract No. 68-03-2364  	    20

References	•	    22

Appendix	    23
                                     iii

-------
                                   FIGURES

Number                                                                 Page

  1       Chromatograms  (GC/ECD) of extracts  of  copper/iron/sand
          column eluate  fractions:   (A) Fraction 2  (1.35&),  (B)
          Fraction 5  (2.9£),  (C) Fraction 20  (14.3£),  (D)  Fraction
          37  (17.9&),  (E) Fraction 39  (21.9£),  (F)  Fraction  41
          (28£).  A-1016 standard  (B)  is shown as reference  	    3

  2       Chromatograms  (GC/ECD) of extract of sand column eluate
          (A) Fraction 3 (1.4£), (B) Fraction 15 (10.9£) and (C)
          Fraction 19  (23.9£).  A-1016 standard  (D)  is  shown as
          reference 	    9
          Chromatograms  (GC/ECD) of endrin/heptachlor  reduction.
          Experiment No. 1 in a stirred  flask:   (A)  copper/iron/sand
          plus endrin and heptachlor aqueous  solution  (B)  sand  plus
          endrin and heptachlor aqueous  solution (C) endrin  and
          heptachlor aqueous solution  only
          Chromatograms  (GC/ECD) of  endrin/heptachlor  reduction.
          Experiment No.  2  in a stirred  flask:   (A) Mixture
          reacted at 60°  in presence of  copper/iron/sand  (B)
          Mixture reacted at 25° in  presence  of  copper/iron/
          sand   .................  .  .  ........   14

          Total  ion current chromatogram of chemical ionization
          GC/MS  analysis  of products from catalytic reduction  of
          endrin (Experiment 4) ...................   16

          Gas chromatograms of endrin/heptachlor reduction
          experiments using miniature  copper/iron/sand column.
          (A) Column effluent after  3300 m£ — typical of all
          fractions.  (B) Column effluent after  3500 m& — after
          standing in column for 65  hours.  Compound identification
          a) heptachlor,  b) heptachlor epoxide isomer  ........   19

-------
                                   TABLES

Number                                                                Page

  1       Elution of PCBs on Copper/Iron/Sand Column  	    6

  2       Compounds Identified in Electron Impace GC/MS Analyses
          of Endrin/Heptachlor Reduction  (Experiment  3)   	   15

  3       Compounds Identified in Chemical lonization GC/MS
          Analysis of Endrin Reduction  (Experiment 4)	   17
                                      v

-------
                                ABBREVIATIONS




A-1016  —  Aroclor^-1016




ESC     —  Envirogenics Systems Company, El Monte, California




GC/ECD  —  Gas Chromatography/Electron Capture Detection




GC/MS   —  Gas Chromatography/Mass Spectrometry




PCB     —  Polychlorinated Biphenyl




TCB     —  2,3',4',5-tetrachlorobiphenyl




CIS     ••—  Copper/Iron/Sand
                                     VI

-------
                               ACKNOWLEDGMENTS
     We wish to thank Drs. Edo Pellizzari and Denny Wagoner of Research
Triangle Institute (RTI) for their assistance in the planning and
execution of this project.  Major credit for the successful completion
of this project is due to Messrs. Russel P. Cepko, Bobby J. Parker,
Peter Grohse (RTI), and Dr. K. Tomer who assisted with the laboratory
experimentation.  Finally, we wish to thank Mr. David Oestreich of IERL,
EPA, RTP, NC for his helpful discussions and advice.
                                      vii

-------
                                 SECTION I
                              INTRODUCTION
     In an effort to assist industry in compliance with legislation concerning
the permissible levels of toxic substances  in industrial water effluents,
the EPA has undertaken to develop various control technologies which may be
applied by industry. An example of  this research is the catalytic reduction
of chlorinated hydrocarbons (specifically PCBs and chlorinated pesticides
such as heptachlor and endrin). Under a contract to Envirogenics Systems
Company (ESC), El Monte, CA (EPA Contract No. 68-03-2364), a technique for
the chemical removal of PCBs  from water was investigated whereby the aqueous
PCB solution is passed through a column containing sand and a copper/iron
catalyst at ambient temperature and pressure.
     Experiments by Envirogenics indicated  successful removal of PCBs from
water; however, a test at the Gulfbreeze, FL, EPA laboratories discovered the
column effluent was toxic to  fish.  Analysis of effluents of the process by
Gulfbreeze-EPA and Bionomics  detected little or no changes in the concentra-
tion of PCBs as compared to the feed solutions.
     The purpose of this project was to chemically evaluate the technique.
Aqueous solutions of PCBs and chlorinated pesticides were analyzed before and
after treatment with the copper/iron catalyst and examined for changes in
their composition. The main analytical technique used in this study was electron
capture gas chromatography  (GC/EC), which showed whether or not changes had
occurred in the chlorinated compounds. When a species change was noted, the
more sophisticated technique  of gas chromatography-mass spectroscopy (GC/MS)
was used to identify the species. Samples were analyzed for iron and copper
in solution, and as a check for evidence of catalytic dechlorination, samples
were analyzed for ionic chloride.

-------
                                SECTION  II
                         SUMMARY AND  CONCLUSIONS
     The catalytic reduction technology  was evaluated  (according to the
procedures provided by the developing laboratory)  for  PCBs and found totally
ineffective. In essence, the column  (160 cm x 37 mm i.d.) behaves as a
large chromatographic column, requiring  on the order of 40 bed volumes
(30£) to elute all components of the  Aroclor™ -1016 feed solution. No
evidence of catalytic reduction of PCBs  was observed.
     The reduction technique was also briefly evaluated for endrin and
heptachlor. To facilitate the evaluation, the reductions were carried out
in stirred round bottom flasks at elevated temperature (~60°) for extended
times (4-6 hours). Partial reduction  was observed  at these elevated temp-
eratures. GC/MS analysis of highly concentrated batches indicated that the
endrin and heptachlor were being reduced by successive replacement of
chlorine by hydrogen. No evidence for changes in the carbon skeleton were
observed.  The extent of dechlorination, even under these highly favorably
conditions, was slight.  Although all of the C 2H  Cl 0 analogs were found,
including the tentative identification of some C^H^O and related hydr°-
carbons, C10H0C1,0 (endrin) and various  isomers of C -H-drO accounted for
          J.Z O  O                                   -L^ "  J
greater than 80% of the total ion current response in  the GC/MS.  The evalua-
tion of the reduction of heptachlor and  endrin thus showed that the technique
worked, although apparently not well  enough to efficiently reduce the
pesticide to hydrocarbons and mono- and  dichlorinated  species which may be
biodegradable.
     The progress reports generated by the development laboratory were
evaluated to determine the errors and omissions in the experimental protocol
which led the investigators to erroneous conclusions about their results.
A list of the points of criticism is  presented.

-------
                                 SECTION III
                               RECOMMENDATIONS
     Based on the findings of the research presented herein, it is
recommended that no further effort be made to develop and/or implement
on a pilot scale the technology  for  catalytically reducing PCBs. The
brief evaluation of the reduction of heptachlor and endrin by this
technique also indicated  that the technique is insufficient for genera-
ting low- or non-chlorinated organics which are bio-degradable. Thus,
it must be conclusively shown that reduction is complete before any
attempts at commercial application are  undertaken.
     Any further research and development of the chemical reduction of
chlorinated hydrocarbons  in water utilizing copper/iron catalyst must
first address the chemistry of the reactions. Basic research should be
carried out to understand the mechanism(s) and kinetics involved before
any work at optimizing the conditions can be undertaken. Once the chemistry
is well defined, the types of compounds for which the catalyst would be
effective, and thus reaction  conditions, may be predicted.

-------
                                  SECTION IV
                             MATERIALS AND METHODS
COLUMN PREPARATION
      The  columns for the PCS work were constructed of Fisher and Porter
Lab-Crest 37  mm i.d.  Pyrex tubing with a stopcock at the bottom and 0.5 cm
bed of glass  wool. The column packing was prepared by blending the copper
catalyzed iron powder and sand supplied by Envirogenics Systems Company in
a  ratio of 1:3.6.  The packing was poured into the column to a depth of
160 cm and the intersticies filled with water.  The void volume (or bed
volume) of the column was determined to be 750  m .  This column will here-
after be  referred to as the copper/iron/sand (CIS) column.   A duplicate
column was prepared in a similar way using sand without the copper/iron
powder.
EXTRACTION
      Previous experience in these laboratories  (1) has  shown that  PCBs
are much  more soluble in toluene than in hexane and thus are more  readily
extracted from a variety of media,  including water,  using toluene. Accord-
ingly, aliquots of the column effluent and influent were extracted using
toluene as  follows:  a 100 raH aqueous aliquot was extracted  three times with
20 mH portions of  toluene in a separatory funnel.  The combined  toluene
extracts  were dried over sodium sulfate and concentrated on a hot plate in
a flat bottom boiling flask with a  Snyder column,  followed  by blow-down
under a nitrogen stream at 25°.  This recovery procedure has been demonstrated
in our laboratory  to  be  quantitative.
ANALYSIS
      All  gas  chromatography-electron capture detection  (GC/ECD) analyses
were  performed on  a Fisher Victoreen 4400 gas chromatograph with a    Ni
electron  capture detector.  GC/MS analyses were  obtained.using a Finnigan

-------
3300 quadrupole GC/MS with a PDP/12 computer.  The determination of Cu and
Fe was carried out on a Perkin Elmer 603 Atomic Absorption Spectrophoto-
meter equipped with deuterium background correction.  Chloride ion determi-
nations were accomplished using a chloride ion^-specific electrode on an
Orion 801A Digital lonalyzer.

-------
                                 SECTION V
                  EVALUATION OF PCB REDUCTION TECHNOLOGY
     Solutions of PCB mixtures  (ArocloiR-1016 and 2,3',4',5-tetrachloro-
biphenyl  [TCB]) were passed  through  two  columns in parallel experiments.
One column held the CIS mixture, while the  second held only sand.  A
condensed sampling schedule  is  shown in  Table 1.
                                  TABLE  1
Fraction
0
2
5
20
31
37
39
40
41
# Feedstock
A-1016
A-1016
A-1016
A-1016
TCB
TCB
TCB
Distilled water
Distilled water
Volume (£)
0.75
0.1
0.1
0.1
0.3
0.4
2.0
2-0
0.1
Cumulative
0.75
1.35
2.9
14.3
16.8
17.9
21.9
23.9
24.0
Volume Description
Void volume



Change of feedstock


Attempt to flush
column
     After  14.6£  (20  bed  volumes)  of  Aroclor  solution had been passed through
the column,  subsequent  chromatograms  of  the eluate fractions were essentially
identical to one  another.   At that point,  the feedstock was changed to TCB
because  it  is a single  polychlorinated species and thus easier to trace than
the complex Aroclor mixture.
     After  5.4£ (7 bed  volumes)  of TCB solution had been passed  through the
column and  no TCB could be  found in the  eluate, the column was flushed with
2£ (2.7  bed volumes)  of distilled water.   Next, a 0.1  portion of distilled
water was passed  through  and  analyzed.

-------
      The results  of  this experiment,  summarized  in Figure 1, indicate clearly
 that  chromatographic elution  is  taking  place  on  the CIS column.   The lower
 chlorinated isomers  (which have  shorter GC  retention times)  eluted more
 quickly than the  higher isomers.   The lower chlorinated isomers  are more
 water soluble than the  higher chlorinated molecules,  and this  would account
 for their passing through the column  more quickly.   After passing 28$,
 (37 bed volumes)  through the  column,  all isomers present in  the  A-1016
 were  observed,  although the last major  peak (RT «10.2 min) had just begun
 to elute and was  present only as a shoulder on the adjacent  peak (RT «9.4 min).
 The isomers which eluted between 4^8  minutes  were  present in higher concentra-
 tion  since they were not as easily stripped from the column  during feed of
 water and TCB solutions.   The identification  of the major peaks  in Fraction 41
 (the  distilled  water fraction) was confirmed  by GC/MS using  multiple ion
 detection.   The TCB, which was first  applied  to the column about 15£ before
 the experiment  was terminated, was never observed  in the column  eluate.
 This  is to be expected  in light of the  chromatographic  behavior  of the higher
 isomers (also tetrachlorobiphenyls) in  Arochlor    •<• 1016.
                                         (S)
      In the parallel experiment, Aroclor   -  1016  and TCB solutions were
 eluted through  a  column containing sand only.  The results of  this experiment
 were  similar to those observed for the  CIS  column,  although  the  elution
 appeared to be  slightly faster.  Typical gas  chromatograms illustrating the
 chromatographic elution of  successive components of A-1016 are shown in
 Figure 2.
      To determine  the significance of long  contact time within the column,
 two successive  fractions  were collected from  the CIS column  65 hours apart.
 No  significant  reduction  had  taken place after this lengthy  interruption
 of  flow.
      The  above  experiments  clearly illustrate that both the  CIS  column and
 the sand  column act  as  crude  liquid chromatographic columns.   Because of their
greater aqueous solubility, the lower chlorinated  isomers in the PCB mixture
pass  through first.   The  disproportionate amount of lower chlorinated species
in the  initial eluates  from the CIS column  could lead to the erroneous con-
clusion that the higher chlorinated species are• being dechlorinated.

-------
cc
a
g
                TIME (min)
                                 (A) fraction  2  (1.35A)




                                 (B) fraction  5  (2.9£)
                                 (C)  fraction 20  (14.
                                 (D)  fraction 37  (17.9£)
(E)  fraction 39  (21.9£)
                                      (F)  fraction 41  (23£)
                                      (G)  A-1016 standard -shoim  for  reference
                            10   12
           Figure 1.  GC/ECD  chroiuatogratas  of  CIS column eluates.



                                        8

-------
in

1
                           (B) fraction  15 (10.95,)
                                         (A)  fraction 3  (l.U)
                                   (C)  fraction 19 (23.95,)
                                   (D) Ar-1016 standards-shown  for
                                      reference
           2468
                TIME (min)
    Figure  2.   GC/ECD chromatograms of  sand column eluates.

-------
     Only the 2,3',4',5-tetrachlorobiphenyl (TCB)  fed onto  the  column remained
unaccounted for.   It was  suspected that this material and A-1016 were adsorbed
on the column packing.  At  the conclusion of the elution experiments, 50 g of
column packing was collected from the top of each column, vacuum filtered to
remove the aqueous portion,  and extracted with toluene.  The results of this
experiment show  clearly in  both cases, large amounts of  TCB in  the  gas chroma-
togram, although large volumes of non-TCB-containing water  had  passed through
the column  (8.15, or 11 bed  volumes in the case of the CIS  columns and 17.l£
in the case of the sand column) and several days had elapsed.  This presents
further evidence that reductive degradation is not occurring to any appre-
ciable extent.
     The  concentrations of  chloride, as measured using a chloride ion-selective
electrode, were  below the detection limit C~5 ppm) in all  samples analyzed.
Since this measurement yielded no useful information, it was discontinued.
In all fractions analyzed by atomic absorption, the iron concentrations were
found to  be  less than 0.3 ppm and the copper concentrations less  than 0.01
ppm.  The consistently low concentrations of iron and copper in the effluent
do not support a reductive mechanism in which iron is an indicator  of
reduction.
                                      10

-------
                                 SECTION VI
                   EVALUATION OF ENDRIN/HEPTACHLOR REDUCTION
STIRRED FLASK EXPERIMENTS

     The catalytic reduction performance of the copper/iron/sand mixture
was also evaluated for  endrin and heptachlor.  At  the  request of the project
officer, the reduction  effects were first  evaluated by  the  stirred flask
technique.
     An experiment was  performed with three parallel  solutions of endrin and
heptachlor in:  (a) a  flask containing the  copper/iron/sand  mixture (50g);
(b) a flask containing  sand;  and (c)  a flask which contained the endrin/
heptachlor solution only.  All three flasks were heated  to 60° and stirred
for not less than 4 hours.  GC/ECD analysis of  the reaction  mixtures after
several hours indicated that some degradation  of  endrin and heptachlor may
have taken place  (Figure 3).  Since the loss of heptachlor and concurrent
appearance of another peak at RT ~  7  min*  was  observed  in both the flask
containing sand and the flask containing solution only, this reaction cannot
be attributed to catalytic reduction by the copper/iron/sand mixture.
     This peak was, in  fact,  observed in extracts of  the stock solution after
storage at 5° for two days.  Analysis  of this extract  by GC/MS indicated
heptachlor (RT ~ 7 min*) and endrin (RT ~ 11 min*).  The  mass spectrum of the
unknown peak at RT =  8  min*  led to a  tentative structural assignment of
C  H_C1 0_.  The lack of a parent ion precludes a definitive assignment. This
compound would be a product  of addition of two water  molecules and loss of
two chlorines, an unusual  reaction, especially under  these  mild conditions.
     The presence of  this  compound has been reported  numerous times in the
literature (3-5) as a degradation product  of heptachlor in  water. The
compound has been identified in these reports  as  1-hydroxychlordene
(C  C1,H,0, m/e = 352).  Further analytical characterization, beyond the scope
  10  66                                '                               .
of his project, is needed  to resolve  the apparent discrepancies and assign
a structure to this compound.
 *
Retention times were  different in the GC/ECD and GC/MS analyses because  of
differences  in the column temperatures.
                                     11

-------
                                                                            A
ro
CD
M
o
a.
EC
O
o
o
CC
      0
      Figure  3.
Chromatograms (GC/ECD) of endrin/heptachlor reduction.
Experiment No. 1 in a stirred flask.
Top:  A—copper/iron/sand + endrin and heptachlor  aqueous
      solution
Center:  B—sand  + endrin and heptachlor aqueous  solution
Bottom:  C—endrin and heptachlor aqueous solution only
1:  heptachlor
2:  C-LQHgCLjC^ (tentative—see text)
3:  endrin
                                     12

-------
     In a second experiment the effects of temperature were investigated and
it^was found  (Figure  4)  that the degradation was substantially more marked at
60° than at 25°. Thus,  degradation of heptachlor and endrin was established
as being related to the presence of the copper/iron mixture,  although the
degradation products  were unknown. A third flask,  containing distilled water
and the copper/iron/sand mixture, was used as a blank in  this experiment. The
GC/ECD analysis of an extract showed no contamination of  the blank, as expected.
     To establish  the identity of the degradation products,  a reaction was
carried out with much larger amounts of endrin and heptachlor to yield enough
of the reaction products for GC/MS analysis.  The solubility limit of endrin and
heptachlor was exceeded during this experiment.  The results of this experi-
ment  (Table 2) demonstrated that several dechlorinated heptachlor and endrin
species as well as some oxygenated analogs were present.  There was no evidence
of complete  (or even  nearly complete) dechlorination or of  a reverse Diels-Alder
type reaction to form chloro-cyclopentadienes.
     Experiment 3  indicated the complexity of the reaction  products, so a further
experiment with endrin alone was conducted to elucidate the reduction products
of this single starting material. A large batch (19 mg) of  endrin was reacted
at 60° in a stirred  flask and analyzed by chemical ionization mass spectrometry.
The results,  shown in Figure 5, and Table 3, indicate that  endrin is catalytically
degraded  (i.e., replacement of chlorine by hydrogen) to some extent, although
the vast majority  of  the reaction product is found as C ^HgClgO  (endrin) and

     Dechlorinated endrin (G12H14°) was tentatively identified although
the amount found is minor compared to the higher chlorinated analogs.
Thus, even under much more severe reaction conditions than  recommended, the
catalytic reduction is far from complete.
     The results of the stirred flask experiments show that endrin and hepta-
chlor were catalytically reduced by the successive replacements of chlorine
by hydrogen in the presence of the copper/iron/sand catalyst. However, the
evidence indicated that, even at 60° for 4 hours, this process was far from
                                      13

-------
ra
eu
u>

O
Q.
vs
O>
tr

x.
ID
•o

O
o
0)
CC.
                                Time (min.)


      Figure 4.  Chromatograins  (GC/ECD)  of endrin heptachlor reduction.

                 Experiment  No.  2  in a stirred flask:

                 A.  Mixture reacted at 60° in presence of copper/iron/sand

                 B.  Mixture reacted at 25° in presence of copper/iron/sand,

                 See figure  3 for  peak assignments.
                                      14

-------
                                    TABLE 2
CO»CUm,S  IBENHFIKD a ELECTRON IMPACT GC/MS ANALYSIS OF ENDRIN/HEPTACHLO*
                          REDUCTION (Experiment 3)
Compound
— — • 	 	 	 _ 	
C10H7C15
C10H6C16
C10H7C150 (tent)3
and
C10H5C17 (heptachlor)
C10H6C16°
C10HgCl502 (tent)°
C12H10C14°
C12H9C150
C.,H0C1,0 (endrin)
J.Z O D
C12H8C16°C
Retention
Time (min)
	 — 	 .
4.0
4.9

6.4

7.3
7.7
8.8
9.4
9.8
10.8
- 4.4
- 5.6

- 7.3

- 7.7
- 8.6
- 9.4
- 9.8
-10.8
-13
Integrated ,
Response (%)
—————— — — —_ . _^_ — _^__ _ —__»«_»„_ .^^__
0.2
1.2

4.6
•
1.8
25.9
0.3
1.9
17.2
47.0
 ilass spectral quality was  too poor  for  definitive  assignments.
b
 Response values are listed as a percent  of  the  summed integrated areas
 for each peak in the chromatogram..  These values mav correspond approxi-
 mately to relative concentration  although no molar response factors are
 available.

Differences in fragmentation and  chromatographic retention times indi-
 cate at least two isomers with this molecular formula are present in
 addition to endrin.
                                       15

-------
   100.0
   80.0
4-)
•H
w

01
4-1
fi
l-l

4J
C
(1)
u
GO.O
40.0
   20.0
    .0
                                                       10
                                      Time  (min.)
                                                                           15
             Figure 5.
                     Total  ion current chromatogram  of chemical ionization GC/MS
                     analysis  of products from catalytic reduction of  endrin
                     (Experiment 4).   See Table 3  for identification.

-------
                                TABLE 3

  COMPOUNDS IDENTIFIED  IN  CHEMICAL  IONIZATION  GC/MS ANALYSTS OF ENDRIN
                        REDUCTION (Experiment  4)
                                                     Integrated
                                                    Response (%)b
C12H160 (tent.)
C12H13C10
C12H12C12°
C12H11C13°
C12H1QC140
C12H9C150
C10HQC1C0 (endrin)
1 £. O D
C12HgCl50
d 1.3
3.0
4.4
6.2
7.7
9.6
10.8
12.4
- 2.8
- 3.9
- 5.9
- 7.0
- 9.5
- 10.7
- 11.7
- 13.3
4.4
1.1
2.0
0.8
6.1
32.7
45.6
7.4
Retention times represent  boundaries  of  the  integration.  Some compounds
represent two or more  isomers,  as  evidenced  by multiple peaks in the
chroma togram..
Response values are listed as a percent  of the summed  integrated areas for
each peak in the chroma togram.   These values may correspond approximately
to relative concentration  although no molar  response factors are available.
                                    17

-------
complete. These results indicate that a column reduction at room  temperature
with a much shorter contact time would have produced even less dechlorination
of endrin and heptachlor. This hypothesis was tested as described below.
MINIATURE COLUMN EXPERIMENTS
     The reduction of heptachlor and endrin by the copper/iron/sand mixture
was tested using 18 cm x 1.0 cm i.d. (14.1 ml) columns. Two columns were prepared
in parallel: one of the copper/iron/sand mixture and one of the sand only.
The void volume of the columns was determined to be 7 m &  .A mixture of
heptachlor and endrin (200yg/£  of each) was prepared in pH=7 water and
eluted through the columns.
     The sand column was found to be inert. The column effluent contained
endrin, heptachlor, and the compound tentatively identified as C-^B.gCl.0
(vide supra).
     The copper/iron/sand column exhibited some reduction as shown in
Figure 6. The extent of reduction, however, is minor in comparison with the
stirred flask reactions. Even upon sitting in the column for about 65 hours,
endrin and heptachlor are clearly visible and the predominant peak in the
chromatogram corresponds to the compound previously tentatively identified
as CinHQCl,-0,-. Thus, even with extremely long contact time, the reduction is
incomplete and merely represents partial substitution of the chlorines.
SUMMARY EVALUATION OF ENDRIN AND HEPTACHLOR REDUCTION
     Both the stirred flask and miniature column experiments demonstrated
some catalytic reduction of heptachlor and endrin. The extent of  reduction,
however, is slight even when experimental conditions are much more severe
(higher temperature and/or longer times) than those used in the developing
lab.
                                      18

-------
       »
       '5
       =J
       o
       Q.
       cn
       1U
       cs
       UJ
       a
       cc
       o
       o
       UJ
       cc
Figure 6-
                         TIME (min)


Gas chromatograms of endrin/heptachlor reduction experiments

using miniature copper/iron/sand column.

A.  Column  effluent after 3300 m£—typical of all fractions

B.  Column  effluent after 3500 m£—after standing in column

    for 65  hrs.
Compound  identification:   a) heptachlor, b) C-^HgCl^,

                           c) endrin.
                                  19

-------
                                  SECTION VII

           CRITIQUE OF MONTHLY REPORTS  OF CONTRACT NO. 68-03-2364

     The reports generated  during the development of  the catalytic reduction
technology have been  evaluated.   A list of errors and omissions which led to

the erroneous conclusion  that  the technique was  applicable to PCBs follows:

     1)  No preliminary feasibility studies (kinetics, thermodynamics,
         mechanisms,  etc.)  were  carried out to determine whether the
         catalytic reduction could take place at ambient temperature and
         pressure.

     2)  It is disturbing that this technology was found to be unsuccessful
         in batch  (stirred  flask)  reactions (see January 1976 report).  One
         would expect the same type of  reaction  in a flask unless the process
         is a function of chromatography on the  column.  No attempt was
         made by the  development laboratory to explain this observation.

     3)  The suggested increased efficiency of the column as a function
         of the mesh  size (and therefore surface area) of the sand supports
         the theory that  adsorption is  the process rather than reduction.
         However, there was no extraction of the solid bed material to
         investigate  this possibility.

     4)  Observations of  the chromatograins of the inlet (feed) and outlet
         sample from  the March, Gulf Breeze (Bionomics) study yielded
         the conclusion that both chromatograms were essentially identi-
         cal.  No significant  reduction was achieved.  This was proved
         subsequently by  toxicity data  and the high concentration of PCBs
         in Bionomics1 Control samples.

     5)  There was no identification of reaction products (although a
         GC/MS study  was  mentioned in the reports) , and no attempt was made
         to do a complete mass balance  across the column.  Mass balances for
         chloride were conducted in several cases by turbidity.  Chloride
         ion balances at  these concentration levels (at or near the limit of
         detectability) by  a turbidometric procedure are difficult at best.

     6)  A carefully  designed  experiment should have included blanks, controls
         (for example, a  column  packed  with sand only) and internal standards.
         These were lacking in the work done by  the development laboratory.

     7)  Based on the size  of  the columns and the possibility of a chromato-
         graphic mechanism, the  number  of void volumes (water) passed through
         the column were  insufficient.   Extended runs should have been made
         in order to mimic  plant conditions.
                                      20

-------
     Evaluation and justification of catalytic technology in the developmental
portion of the program was primarily based on gas chromatography/electron
capture detection.   Interpretation of gas chromatograms on inlet and outlet
samples (Gulfbreeze, Florida)  demonstrated the difficulties and problems
associated with the lack of GC/MS confirmation.   A stricter control of the
development phase with supporting documentation of the degree of dechlorination
by GC/MS would have controlled boundary conditions of  the experimental program.
                                     21

-------
                                  REFERENCES

1.   M.D. Erickson, R. A.  Zweidinger,  L.  C.  Michael  and E. D. Pellizzari,
     "Environmental Monitoring Near Industrial Sites:  Polychloronaphthalenes,"
     EPA-560/6-77-019, July 1977.

2.   Personal Communication, K... H.  Sweeney,  Envirogenics  Systems Company,
     El Monte,  California, April, 1977.

3.   A. S. Y. Chan,  J. D.  Rosen, and W.  P.  Cochrane, Bull. Environ.
     Contamin.  Toxicol. , Jo, 225-230 (1971).

4.   A. Demayo, Bull. Environ. Contamin.  Toxicol.,  8_, 234-237  (1972).

5.   J. J. Eichelberger and J. J. Lichtenberg, Environ..  Sci. Technol,, 5_,
     541  (1971).
                                        22

-------
                                 APPENDIX
On-site Demonstration  by  Envirogenics  of a  Catalytic Reduction Process for
Endrin/Heptachlor.
                                 INTRODUCTION
     This appendix  describes  the on-site test  procedures, and the analytical
results of a catalytic reduction process of endrin/heptachlor.  AnRTI and EPA
representative  observed the demonstration by Envirogenics of the technology,
on-site sampling and analysis by GC/EC, and returned concentrated aliquots
of the feed materials, blanks and  eluates to RTI  laboratories for GC/MS
analysis.
     The  "Test  Procedure" section  of this appendix was  copied verbatum from
an Envirogenics Systems Company experimental procedures manual.
                                       23

-------
                                TEST PROCEDURE
                        ENDRIN AND HEPTACHLOR REDUCTION
                       ENVIROGENICS DEMONSTRATION TESTS
COLUMN REDUCTION
     Two aqueous .solutions will  be  used,  one  containing approximately 100 PPb
of endrin, and the second approximately 30  ppb  of heptachlor.  Each solution
will be passed through  separate  copper-catalyzed iron  columns at approximately
35 m£/min using freshly-prepared columns  (37  mm by  1600 mm bed), each containing
600 g of copper-catalyzed iron powder  and 2150  g of sand.  Samples of the
catalyzed iron and sand will be provided EPA or  RTI  if  desired.  The runs will
be for a sufficient period to provide  material  for  GC/MS analysis.  Both the
endrin and heptachlor solutions  will also be  passed through sand columns and
eluate samples will be  collected and analyzed.
ANALYSIS
     Envirogenics intends to analyze the samples for endrin and heptachlor by
electron capture and flame ionization  detection gas chromatography, for
chloride by turbidity,  and iron  and copper  by atomic absorption spectrometry.
For purposes of the chloride analysis, it will  be necessary to wash all
columns with water before use.   It  may be advisable to establish a chloride
blank from the eluates  of the  endrin and heptachlor solutions passed through
the sand columns.  In order to avoid losses of  organic material from the
dilute aqueous phases by sorption on the container  walls, the samples will be
extracted with hexane immediately.   The extractions will be made with Burdick
and Jackson glass-distilled hexane  of  a single  lot. The extracts will be
analyzed by EC-GC, composited  (unless  the result  is an obvious sport, in
which case an additional sample  would  be drawn),  and concentrated for GC/MS
and FID/GC in Kuderna-Danish evaporators.  The  extracted aqueous  samples would
be retained for concentration  and chloride  analysis.
                                      24

-------
     Small samples of the extracted effluent will be taken also for iron  and
copper analyses.  These samples would acidified and analyzed for iron and
copper by atomic absorption spectroscopy,
DETAILED PROCEDURE
Columns
     The 37 mm dia x 1800 mm long glass tubes will be thoroughly cleaned  and
the glass wool mat put in place.  For the reduction tests, 2150 g of type
30 sand and 600 g Fe/Cu will be blended and placed in the glass tube.  The
bed will then be washed with 1 £ acetone, drained, and filled with water  by
bottom feed.  Deionized water (2 £) will then be passed through the column.
The sand column to be operated in parallel with the reductant column will
be prepared by placing 2150 g sand in the same sized, similarly prepared
column.  The beds will be examined and suitably treated to remove air
bubbles before starting flow.
Column Feeding
     The endrin feed would be sampled at the beginning of the test, after
6-8 H flow, and after 12 £ flow.  Each sample would be 500 m$L, and would be
extracted with 3-50 m£ portions of hexane.   In making the extractions, it
is important that no aqueous phase be carried over into the organic phase,
so that if phase separation is difficult, 5 m£ of 2% aqueous Na^SO, may be
added to promote separation of the layers.   One m£ portions of the extract
will be reserved for EC/GC analysis, and the remainder composited and con-
centrated to 1.5 m£ in the Kuderna^Dani-sh evaporator.  The 1,5 mil concentrate
will be divided with EPA/RTI taking 1.0 m& and Envirogenics 0.5 m£.  In
addition, 1.3 & portions would be sampled for chloride (and the 3^1/3 S,
samples combined), and 25 mi samples would be taken for AA,  The heptachlor
feed would be sampled similarly.
                                     25

-------
COLUMN RUNS
General

     The columns will  be started and adjusted to a flow of  30-40 mi/min.  Both
the reductive  column and sand column are to be running simultaneously.  The
flow valves will be  adjusted to give a static head of -15 cm.   It  is important
that the column be kept  covered with water at all times.
     The first 2-1/2 I of flow will be required to replace  the  water in the
column at the  start, and will not be extracted.   However, a sample will be
taken for chloride analysis.
Endrin and Heptachlor
     A total of 15 £ of  endrin solution will be run through the reduction column
and the sand column.  Each 3 £ of effluent will be collected, extracted with
three 200 m£ portions  of hexane, and 1 ml of the extract taken  for EC/GC.  If
the result is  not a  sport, the extracts will be combined and concentrated
in the Kuderna-Danish  to 1.5 m£ (from the total 3 £ of hexane extracts).  If
any sample is  obviously  out of line (gross contamination, etc.), then an
additional 100 m£ will be drawn, extracted with three 10 m£ portions of hexane,
and reanalyzed as a  check.  The sample would then either be composited or
rejected by mutual agreement.  If rejected, the test would  be continued for
an additional  3 £.   The  3 £ sample after extraction will have a 1  £ sample
retained for chloride  analysis.  In addition, a 25 m£ sample will  be taken
for AA.  The AA samples  will be acidified with 1 m£ cone HC1.
     At the conclusion of the test (15 £), a 2-1/2 £ portion of deionized
water  (post wash) is passed through the column, and the column  allowed to
drain.  A 1 £  portion  of glass-distilled acetone is then passed through the
column, drained, and analyzed by EC/GC.  The 1 £ acetone washes are continued
until the effluent analysis is clear.
     The heptachlor  test will be carried out similarly, except  that a total
of 50 £ will be passed through the reduction column and the sand column.
                                       26

-------
                          ON-SITE GAS CHROMATOGRAPHY
     All feed solutions  for  endrin and heptachlor were  analyzed three times
during the experimental  phase to document the integrity of the stock (Figures
C-1A and C-1B) .  The  on-site demonstration of the technology for endrin and
heptachlor used one common feed solution to feed both the sand and Fe/Cu columns.
In addition, the concentration of the feed solution was determined by comparison
to standard endrin solution  to within 4% of the  theoretical feed solution.
Aliquots were analyzed from  the effluent Fe/Cu and sand columns periodically
during the tests to follow the extent of reaction.  Six samples were concen-
trated following GC/EC analysis for later GC/MS  analysis.  These included
endrin feed, endrin Fe/Cu effluent,  endrin sand  effluent, heptachlor feed,
heptachlor Fe/Cu effluent, and heptachlor sand effluent.  All gas chromatographic
analyses were performed  on a Perkin Elmer gas chromatograph with an electron
capture detector.  Data  reduction was via a Hewlett Packard data system.  Two
gas chromatographic columns  (OV-17 and DC-200) were utilized for analysis.
The results of the on-site gas chromatography analyses  are as follows:
     1)  Both endrin  and heptachlor feeds compare with  stock standards in
         purity.  The gas chromatographic retention times of the stock endrin
         (6.50 minutes)  and heptachlor (2.15 minutes)  are the same for the
         feed solutions  (Figures C-1A and C-1B).
     2)  No conversion or reaction was observed  in either sand column (without
         Fe/Cu catalyst)  for endrin (Figure C-2A)  or heptachlor (Figure C-3A),
         i.e., the major peak observed was parent endrin (Figure C-1A)  or
         heptachlor (Figure  C-1B).
     3)  Immediate changes in the eluate GC/EC chromatograms for both endrin
         (Figure C-2B) and heptachlor (Figure C-3B) were observed for the
         Fe/Cu catalyst  containing columns.
     4)  Effluent from the Fe/Cu endrin column experiment was shown not to
         contain parent  endrin.   This was demonstrated  by injections of endrin
         (sand column effluent,  Figure C-4A) and effluent from the Fe/Cu column
         (Figure C-4B) on an additional gas chromatographic column (DC ZOO).
         An equal volume injection of the sand and Fe/Cu effluent (Figure C-5)
         on DC-200 is included to show the difference in retention time.
         Presence of  parent  is not eliminated based on  the retention times of
         the margin peaks on the OV-17 gas chromatographic column (Figures C-2A
         and C~2B).   The absence of parent compounds was confirmed later by
         GC/MS analysis.
                                      27

-------
5)  Following the addition of endrin and heptachlor feed to two separate
    Fe/Cu columns, the columns were washed with water and stripped with
    acetone.   GC/EC chromatograms of the acetone wash indicated the
    presence  of additional electron capturing species.
                                23

-------
                                           :   iadrtn :t«d 11
                                          !0:   112-50-f2   3«C«:  7/6/77  .*a*i:
                                           '57- i 7   ?•£?:  ^30   3a»:  i3 ai/ain
                                                                            SO si Hex.
•Uur« C-U.  S*»
              
-------
                                                   SaapU;   !i*?eaciUo« f*«d >2
                                                   i*r.  MQ:   L2Z-51-??   3ac*:  9/3/77   An*l:   12  s:
                                                   Col;   3V-i 7   T*=p:  LM   Ca»:  20 ai/ciin
                                                   laj s   2iO"C   Man;  J30"C   ?ro*:  —*C,'«ia
                                                   D«i   SC    Q4C r«ap:  ;iO"C   -jd«j  1
                                                          Six*:   5 ^i   Coaca:  500 - 3 * 50 ai  Am*.
                                                          flwea^taaa if giuxt* «arftinr.,*»cacai
                                                          t*.  '-?«pciihlor Jttd soluti-so iZ.
S«r. Ho:  122-51-4   3ac»;  9/9/77   inai;  t
Col:  0V 17   T«BD:  ISO   CUi:  20 aJ./=in
IRJ:  220°C   ?ts«:  —*C/aia
3«t:  SC   D«E twap;  liO'C   .-loda:  3
       SU»i  S ai  Cdoea:  3 'Ji»3 x 100 nl
    OV-17
   C-13) -
                      e. So:  122-51-7   3*c«:   J/3/77    Acul:   Z «
                      i:  97-17   7«ap:  130    Jai:   23 ai/sla
                      j:  120*C   San:  220"C    ??B<:   —*C/«ia
                      e:  EC   3«c T*ap;  :40*C    :'^d«:   3
                      «pL. SU«:  S .1  Conca:   IL - 3  x 100 ai H«x.
C-5V  ^** anraoftCacris
       (CC/EC, JV-17 sa
       tec•:  ChcmMea^
                                                                  30

-------
S**eL«:   Stxirin ISt,  ft  F*/C«
S«r.  :*:  121-30*3    D«e«:   3/T/77   Anal;   12  ec
Cat:   3C 200   "tap:   ;iQ    CM:   SO ni/aia
EflJ:   :*0"C   X»n:   220'C    Prog;  —"C/aia
S«e:   EC   0« T«ap:   2iQ*C   «od«:  3
       SU«:  5 tl     Caaea:  a. - 3 * :30 al -T«x.
                                                                                                          J*aoi«:   iaeirlfl iff.  IS "*/Ca
                                                                                                          S«e.  So:   U2-SO-S    3*e«:   5/T-7   ^«t;  :
                                                                                                          Cal:   X  ICO   ttoo:   iiO    lii:  aO r£/*u
                                                                                                          la):   ::0*C    ^ju:  210""    ?rati  — *=' »ia
                                                                                                          3»«:   1C   DtC t*a»:   -^'C   "jM*:  3
                                                                                                                Sin:   5 .i   Caaea:  3L - 3 t 200 ai
crittwweo^raaa of  *odrla r*d«eS
F./Cu :ol^=n (CC/ZC,  X-1^0).
 soiv^rn,  "g£* :   2!*iu*oc !ro»
                                                                 lj*t« J5,
                                                                 an  id* a deal
                                                                         for •edria
                                                                «Lu*C* oa X ICO.
                                                                                          31

-------
                               GC/MS ANALYSIS

INSTRUMENTATION AND ANALYSIS TECHNIQUE

     GC/MS analyses were obtained using a 180 cm x 0.2 cm i.d.  glass column

packed with 2% OV-101 on Chromosorb W-HP maintained at 120°C for  3 minutes,

then programmed at  12°C/min to 230° and held there for the duration of the 'run.

The flow rate was 20 m£/min helium.  GC/MS chromatograms  were obtained using

a Finnigan 3300 quadrupole GC/MS with a PDP/12 computer.

SUMMARY OF RESULTS

     The results of the  GC/MS  investigation of six samples obtained during

the on-site demonstration are  as follows:

     1)  Heptachlor feed - The only species and only major substance found
         was heptachlor  (see Figure C-r-6) .

     2)  Heptachlor sand column - The only  chlorinated species found was
         heptachlor (see Figure C-7).   However, the sample was heavily
         contaminated with C „  and greater  hydrocarbons.

     3)  Heptachlor Fe/Cu eluants - The chlorinated species observed are
         listed below.

           Figure C-9.   Heptachlor eluate concentrated to 1.0 ml.

 Retention Time (min)        Tentative  I.D.           Characteristic Ions

         7.91                   C10H7C15             236,238,240,302,304,306

         8.33                   C1()H7C15             236,238,240,302,304,306

         9.66                   CHCl              283,285,287
         9.91                  CiOHxC14+n           236,238,240
                                    32

-------
         Figure C-8.  Heptachlor eluate concentrated to 100 ul.

Retention Time (min)        Tentative I.D.           Characteristic  Ions

        4.73                  C1()H9C13             168,170,172,234,236

        6.41                  CHCl              168,170,172
        6.66                  C  H Cl              236,238,240,201,203,205,
                               1U '  *             301 to 310

        7.00                  G-mH7C1<             236,238,240,201,203,205,
                               1U '  3             201 to 310
7.41
8.16
8.33
8.41
8.60
? 218,220,222
? 283,285,287,215,217,219,
221,183,185
C1()H7C15 236,238,240
Mixture 283,285,287,183,185
Mixture 236,238,240,318 to 322,
283 to 287
    5)  The predominant species are the higher chlorinated ones with the Cl,
        species being observed only in the very concentrated species (see
        Figure C-8).   The major component has a minimum of 4 Cl's (see
        Figure C-9).   No parent heptachlor was detected.   This confirms
        earlier onr-site GC/EC findings.

    6)  Endrin feed - The chlorinated (and significant) species observed
        were endrin and two isomers (see Figure C^IO).   The other isomers
        were not dieldrin.

    7)  Endrin sand column eluents - The only chlorinated species observed
        were endrin and two isomers (see Figure C-ll).

    8)  Endrin Fe/Cu column eluent - The chlorinated species observed as
        follows:
                                   33

-------
   60.9-
u
V*
o
a-
                                            Tino  (nin)
         Figure C-6.  local ion current chronogram of electron impact GC/MS analysis of
                      heptachlor feed solution extract (sample no. 3, 1.0 ml).
                                           Time  (nin)
        Figure  C-7.   Total ion current chromatogram of electron impact GC/MS analysis of
                     eluate from sand column with heptachlor feed (sample no. 1,  1.0 ml).
                                           Tioe (min)
    Figure  C-8.   Total ion current chromatogram of electron impact GC/MS analysis of
                 concentrated eluate from copper/iron column with heptachlor feed
                 (sample no. 2, 0.1 ml).
                                            Time  (min)
     Figure C-9-  Total  ion current chromatograo of electron impact GC/MS analysis of
                  eluate from copper/iron column with heptachlor feed (sample no.  2,
                  1.0 ml).
                                                           3 4

-------
                                           Time (rain)
      Figure C-10.  Total ion current chromatogram of electron impact GC/MS analysis of
                    endrin feed solution extract (sample no. 5, 1.0 ml).
                                          Time  (min)
     Figure  OH.   Total  ion current  chronatogram  of  electron  impact  GC/MS  analysis  of
                   eluate from  sand column with  endrin  feed  (sample no.  4,  1.0 ml).
                                           Time  (min)


       Figure C-12.   Total ion current chrooatogran of electron impact GC/MS analysis of
                     eluate from copper/iron column with endrin feed (sample no. 6, 1.0 ml).
2
S M.-J
                                         Time (min)
     Figure C-13.   Total  ion current chromatogram of electron impact GC/MS analysis of
                   concentrated  eluate from copper/iron column with endrin feed (sample
                   no.  6,  0.1 ml).
                                                          35

-------
             Figure C-12.  Endrin eluate concentrated to 1.0  ml.
 Retention Time (min)
     =      =

     10.49


     11.49


     11.83


     13.33
Tentative I.D.
  C12H10C14°
   Characteristic Ions
  ^

310,312,314,275,277,247,
249,211,213

344,346,348,309,311,313,
281,283,285,245,247,209,211

344,346,348,309,311,313,
281,283,285,245,247,209,211

245,247,281,283,285,309,
311,313
             Figure C-13.   Endrin  eluate  concentrated to 100 ul.
Retention Time  (min)
      9.66

      9.74
     10.66
     11.08

     11.74
     12.00
     13.66
 Tentative I.D.
   C12H11C13°

   C12H11C13°
   C12H10C14°
   C12H10C14°

   C12H9C15°
       Characteristic Ions
   207,209

   241,243

   276,278,241,243
   275,277,279,240,242,310,
   312,314
   275,277,279,310,312,314

   309,311,313
   245,247,281,283,309,311,313
 The major chlorinated  species  again are the higher chlorinated ones  (see Figure
 C-12).   The major peaks  in  the. dilute  sample are  the  ones  at 11.49 and 11.83 where
 only one chlorine has  been  lost.   No endrin was observed again confirming   _
 earlier GC/EC analysis.   The identifications for  the  lower chlorinated species
 in the concentrated  samples are extremely tentative due to high background
 (see Figure C-13).
                                       36

-------
                                  CONCLUSIONS

•    No conversion  or  reactions were observed on either sand  column  (no catalyst
     present) for endrin or heptachlor,  i.e., the major component remaining was
     parent endrin  or  heptachlor.

•    An immediate change in the parent endrin and heptachlor  was observed in the
     Fe/Cu columns.

•    Effluent from  the Fe/Cu endrin experiment was shown by on-site  gas chroma-
     tography not to contain parent endrin.   This absence of  parent  endrin was
     later confirmed by GC/MS analysis at RTI.

•    The evaluation of the reduction of heptachlor and endrin showed the
     test procedure worked, although apparently not well enough to efficiently
     reduce the  pesticides to hydrocarbons and mono-and dichlorinated species.

•    The biodegradability and toxicity of chlorinated compounds is a function of
     the degree  of  chlorination.  The extent of biodegradability is  predicated
     on reduction to mono- or di-chlorinated compounds.  Based on the above
     requirements and  the results of various controlled experiments, the effluents
     generated by this technology would not be environmentally acceptable.
                                       37

-------
  REPORT NO.
  EPA-600/2-78-059
4. TITLE AND SUBTITLE                                —
Evaluation of Chlorinated Hydrocarbon Catalytic
    Reduction Technology
                          ...     TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
                                                      3. RECIPIENT'S ACCESSION NO.


                                                      5. REPORT DATE
                                                       March 1978
                                                      6. PERFORMING ORGANIZATION CODE
Mitchell D. Erickson and Eva D. Estes
                                                       8. PERFORMING ORGANIZATION REPORT NO
                                                       RTI-1430/F1
                             ADDRESS
 Research Triangle Institute
 P.O.  Box 12194
 Research Triangle Park, North Carolina 27709
                                                       10. PROGRAM ELEMENT NO.
                                                      1BB610
                                                       . CONTRACT/GRANT NO.
                                                      68-02-2612, Tasks 17 and 28
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC  27711
                                                      14. SPONSORING AGENCY CODE
                                                       EPA/600/13
 15. SUPPLEMENTARY NOTES IERL_RTP project officer is David K. Oestreich,  Man Drop 62,
 919/541-2547.
 16. ABSTRACT
          The report gives results of a laboratory evaluation of a control technique
 developed for the EPA for the catalytic reduction of chlorinated hydrocarbons
 (specifically PCBs and chlorinated pesticides such as heptachlor  and endrin). The
 technique involves  elution of polluted water at ambient temperature and at neutral
 pH through a column containing a mixture of sand and copper iron catalyst.  The
 evaluation found that PCBs are not detectably reduced, but are chromatographically
 eluted from the column in order of increasing chlorination.  The  catalyst was found
 to partially reduce heptachlor and endrin.  The possibility of chromatographic
 elution by endrin and/or heptachlor was briefly investigated using miniature columns
 and found to be insignificant.  The report also describes the  observation of an
 additional on-site demonstration of the catalytic reduction technology,  including
 observation of the test procedure, gas chromatography with electron capture
 detection,  and subsequent laboratory GC/MS analysis of transported samples. An
 evaluation of the reports generated during the  development of the catalytic reduction
 technique is presented.  It discusses errors and omissions in the  developmental
 experimental protocol which led to the erroneous conclusions that the technique was
 applicable to PCBs.	         	
                              KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                          b.lDENTIFIERS/OPEN ENDED TERMS
                                                                       n Field/Group
 Pollution
 Chlorohydrocarbons
 Pesticides
 Reduction
 Catalysis
 Gas Chromatography
 3. DISTRIBUTION STATEMENT

 Unlimited
EPA Form 2220-1 (9-73)
                         Biphenyl
                         Heptachlor
                         Endrin
                         Water Pollution
Pollution Control
Stationary Sources
Polychlorinated Biphe-
 nyls
                                          19. SECURITY CLASS (This Report}
                                          Unclassified
13B
07C

07B
07D
                        21. NO. OF PAGES
                        45
                                          20. SECURITY CLASS (This page>
                                          Unclassified
                                                                 22. PRICE
                                         38

-------