A ' f
EPA
Municipal Environmental Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati. Ohio 45268
EPA-600/8-77-013
September 1977
CATALYTIC HYDRODECHLORINATION OF
POLYCHLORINATED PESTICIDES AND
RELATED SUBSTANCES;
AN EXECUTIVE SUMMARY
-------
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 "SPECIAL" REPORTS series. This series is
reserved for reports targeted to meet the technical information needs of specific
user groups The series includes problem-oriented reports, research application
reports, and executive summary documents. Examples include state-of-the-art
analyses, technology assessments, design manuals, user manuals, and reports
on the results of major research and development efforts.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/8-77-013
September 1977
CATALYTIC HYDRDDECHLORINATION OF
POLYCHLORINATED PESTICIDES AND RELATED SUBSTANCES
An Executive Summary
by
Ebon Research Systems
Silver Spring, Maryland 20901
Contract No. 68-03-2460
Project Officer
Robert Landreth
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and 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 is that necessary first step in problem solution
and it involves defining the problem, measuring its impact, and searching for
solutions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems for the prevention, treatment, and management
of wastewater and solid and hazardous waste pollutant discharges from
municipal and community sources, for the preservation and treatment of public
drinking water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution. This publication is one of the
products of that research; a most vital communications link between the
researcher and the user community.
This report is an executive summary of a study undertaken by the
Department of Chemical Engineering, Worcester Polytechnic Institute. In
presenting their data in this form, it is hoped that the information will be
more easily accessible to a larger population of readers.
Francis Mayo, Director
Municipal Environmental Research
Laboratory
111
-------
ABSTRACT
A study was undertaken of the catalytic conversion of chlorinated
pesticides and other environmentally undesirable chlorinated materials to
acceptable compounds. The results of this study show that chlorine can be
catalytically removed and replaced by hydrogen to produce relatively non-
toxic hydrocarbons.
The experimental foundation for a hydrodechlorination process is
established. This batch process involves use of a supported nickel catalyst,
ethanol as a solvent and sodium hydroxide as an acid acceptor of the hydrogen
chloride by-product. Temperatures less than 150°C and hydrogen pressures on
the order of 50 atmospheres are required.
A reactivity sequence is established based on carbon-chlorine bonding
wherein olefinic chlorine is the least reactive. Reaction models are
proposed and the relative rates of steps in the reaction sequence are deter-
mined for Aroclor 1248 and DDE.
Removal of ortho-substituted chlorine is the limiting reaction in
hydrodechlorination of Aroclor. Aldrin and dieldrin are the most difficult
compounds to hydrodechlorinate because of steric hindrance. Removal of
aromatic chlorine is the limiting reaction in the hydrodechlorination of DDT
and DDE.
This report is submitted in fulfillment of Contract No. 68-03-2460 by
Ebon Research Systems under the sponsorship of the U.S. Environmental
Protection Agency. Research data for the report was summarized from research
conducted by the Department of Chemical Engineering, Worcester Polytechnic
Institute for Contract No. R 802-857-01 under the partial sponsorship of the
U.S. Environmental Protection Agency. The work on this report was completed
as of May 1977.
IV
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables vi
Acknowledgments vii
1. Introduction 1
2. Conclusions and Recommendations 3
3. Materials and Methods 5
4. Analytical Techniques 12
5. Results 19
6. Process Design 47
References 49
v
-------
FIGURES
Number Page
1 Pulsed microreactor 6
2 Continuous vapor phase reactor 7
3 Low pressure gas reactor 8
4 High pressure teflon lined autoclave 10
5 High pressure rocking autoclave 11
6 Aroclor 1248 chromatogram 13
7 Representative mass spectra of DDE/ODD 15
8 Analysis of toxaphene hydrodechlorination 18
9 Aroclor 1248 relative product distribution data 1 23
10 Aroclor 1248 relative product distribution data II 24
11 Aroclor 1248 relative product distribution data III 25
12 Experimentally observed Aroclor 1248 kinetics 27
13 Aroclor 1248 hydrodechlorination as characterized by
o- vs. (m + p) substitution 28
14 DDT hydrodechlorination 30
15 DDD hydrodechlorination 31
16 DDD selectivity plot 32
17 DDE Selectivity plot (A) 35
18 DDE Selectivity plot (B) 35
19 DDE relative rate constants 38
20 Selectivity plot DDE 8 43
21 Selectivity plot DDE 2, 4 44
22 Pesticide detoxification process flow sheet 48
TABLES
Number Page
1 Retention Times for Major Aldrin Products 17
2 Retention Times for Major Dieldrin Products 17
3 Pulse Microreactor Aroclor 1248 Hydrodechlorination 20
4 Preliminary Liquid Phase Aroclor Experimental Conditions.... 21
5 Experimental Conditions for Batch Kinetic Studies Conducted
in Teflon Autoclave 40
vi
-------
ACKNOWLEDGMENTS
We are indebted to the authors of CATALYTIC CONVERSION OF HAZARDOUS
AND TOXIC CHEMICALS: CATALYTIC HYDRODECHLORINATION OF POLYCHLORINATED PEST-
ICIDES AND RELATED SUBSTANCES, Rene B. LaPierre, Ehud Biron, David Wu,
Laszlo Guczi, Wilmer L. Kranich and Alvin H. Weiss.
VI1
-------
-------
SECTION 1
INTRODUCTION
Many of the widely used insecticides and other organic chemicals that
have been found to persist in the abiotic and biotic environments are
chlorinated compounds. Stocks of these compounds, whose toxicity has
warranted their restriction at the market, must be destroyed or converted to
harmless biodegradable or potentially useful chemicals. Removal of some or
all of the chlorine atoms generally effects this transformation.
Catalytic replacement of chlorine with hydrogen was the widely applicable
process selected as the model system for this investigation. This process is
adaptable to a trailer-mounted unit, an objective of some EPA officials. The
compounds chosen for study were: (a) Aroclor 1248 (representative of
polychlorinated biphenyls, PCB's); (b) 2,2- bis (p-chlorophenyl)-1,1,1-
trichloroethane (DDT); 2,2-bis(p-chlorophenyl)-l,T-dichloroethylene (DDE),-ODD,
2,2-bis(p-chlorophenyl)-l,l,-dichloroethane; (c) Dieldrin and aldrin; and
(d) toxaphene. Extensive experimental data are reported for the first two
classes, whereas limited data have been obtained for the latter two classes.
LITERATURE
Recent studies of hydrodechlorination according to the general reaction:
(1) R-C1 + H2 > R-H + HC1 (Where R = alkyl, aryl, vinylic or
allylic group)
have established that the ease of reaction is strongly dependent on the
structure of the carbon associated with the chlorine. The above reaction is
usually carried out in the presence of an acid acceptor such as NaOH. Carbon
configurations encountered in the study were generally: mono and
polychlorinated aliphatic, mono and dichlorinated vinylic and allylic, and
aromatic. It has been reported that electron polarization followed by
rupture of the C-C1 bond is the mechanism of dechlorination of mono- and
dichlorinated butene (1-7).
Transition metals (Pt, Pd, Rh, Ni) are known to be effective catalysts,
and the activity sequence Pd > Pt > Rh has been established for removal of
chlorine from chlorinated benzene (8-12). Product inhibition of catalyst has
been proposed, as catalyst activity in batch systems decreases with
conversion (11-15). Presence of an acid acceptor has been shown to lessen
damage to the catalyst (9). Reaction rates are affected by the solvent
chosen for liquid phase reactions (12) and are related to catalyst activity.
-------
The reactivity sequence:
Cl Cl Cl
I I '
-CH_CI <-CHCI,- c - c- < -c -c - ci
2 It I
H H
H
has been established for simple aliphatics. For simple compounds containing
double bonds, the reactivity increases in the sequence; aliphatic chlorine,
vinyl chlorine, olefinic chlorine (17-20).
For more complex molecules, it has been reported that PCB can be
quantitatively hydrodechlorinated at 180°C, and that 1,1-dichloro,
(2-bis-p-chlorophenyl) ethane is easier to hydrodechlorinate than trichloro,
(2-bis-p-chlorophenyl) ethane (21).
This report addresses the question of whether or not chlorine atoms
bonded to different groups in a single molecule retain the relative
reactivity as determined from studies of simple molecules. It is also
concerned with the relative hydrodechlorination activity of polychlorinated
compounds.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
This study established the hydrodechlorination stoichiometry of several
pesticides and related substances. Results of preliminary process variable
(temperature and Hz pressure) studies were also obtained. It was found that
a pulsed microreactor interfaced with a gas chromatography/mass spectroraetry
(GC/MS) facility is a powerful tool in determining catalyst activity and
reaction selectivity, and that a continuous gas phase reactor is useful for
evaluating catalyst stability. Liquid phase stirred reactors were used for
obtaining kinetic data and establishing reaction paths. The liquid phase
reactors would form the basis for the most feasible processing system.
The major conclusions drawn from the data obtained are as follows:
1) Inexpensive supported Ni catalysts, such as the 61% Ni on kieselguhr
used in these studies, are adequate for the hydrodechlorination of pesticides
and related substances. Reactions catalyzed with 10% Pd on charcoal
exhibited similar stoichiometry to those observed with nickel. The apparent
activation energies observed during liquid phase DDE hydrodechlorination were
virtually identical on both catalysts. Liquid phase Ni catalyzed reactions
must of course be carried out in the presence of an acid acceptor such as
NaOH to avoid dissolution of the Ni metal. The Ni catalyst was observed to
deactivate during aromatic hydrodechlorination at temperatures of 130°C and
above, which places an upper limit on reaction temperature. When needed,
higher catalyst loading may be used to increase specific rates in a constant
volume reactor.
2) Polar solvents such as ethanol are recommended over hydrocarbon
solvents such as xylene. Comparative experiments on DDT hydrodechlorination
show that higher reaction rates and less complex product distributions are
obtained when a polar solvent is used. Gas phase DDE kinetic studies also
show that xylene competes for the catalyst surface and therefore acts as an
inhibitor. In addition, polar solvents allow for the use of soluble acid
acceptors such as NaOH, which increase observed reaction rates in the liquid
phase.
3) Ni catalyzed Aroclor (PCB) hydrodechlorination is a consecutive
process characterized by the number of chlorine atoms per molecule and the
degree of ortho versus (meta plus para) chlorine substitution. The relative
rate constants obtained by reaction path analysis indicate that product
inhibition effects are expected in PCB hydrodechlorination. The relative
hydrodechlorination rate of ortho versus (meta plus para) PCB isomers was
determined to be 0.52. Removal of ortho substituted chlorine is the limiting
step in complete hydrodechlorination of PCB.
-------
4) DDT reacts rapidly with hydroxyl ions to produce DDE. DDE
hydrodechlorination itself is characterized by concerted removal of both
olefinic chlorines followed by rapid saturation of the dechlorinated olefin.
Removal of aromatic chlorine from DDE is a stepwise process, and is the rate
limiting step in producing the hydrocarbon product 1,1-diphenyl ethane. DDE
hydrodechlorination is characterized by concerted removal of both aliphatic
chlorines and stepwise removal of aromatic chlorine. Aliphatic hydrodechlor-
ination is the rate limiting step for ODD. Olefinic hydrodechlorination is
more pressure dependent than aromatic hydrodechlorination, and may be used to
effect changes in intermediate selectivities. Liquid phase reactions, both
Pd and Ni catalyzed, are characterized by concerted steps where both olifinic
and aromatic chlorines are removed in one trip to the catalyst surface.
5) Aliphatic chlorines, while least reactive, may also be catalytically
removed. Studies on toxaphene hydrodechlorination show that most of the
chlorine is easily removed. The decreased reactivity of the remaining
chlorine is believed to be due to steric effects.
6) Bridged non-planar compounds are the most difficult to react due to
steric effects. Studies on aldrin and dieldrin hydrodechlorination show
these compounds are the least reactive, even though they contain highly
reactive olefinic chlorine. It appears that the non-chlorinated part of the
aldrin structure is preferentially adsorbed, inhibiting the access of the
chlorinated part of the molecule to the catalyst surface.
7) The liquid phase hydrodechlorination of pesticides and related
substances is the favored process alternative for a large scale unit. The
commercial utility of the partially and totally dechlorinated products has
not been established, neither has a meaningful assessment of toxicity yet
been made.
-------
SECTION 3
MATERIALS AND METHODS
Design of an appropriate catalytic hydrodechlorination process involves
selection from a wide variety of reactor configurations, reactant
compositions, catalyst forms, and operating conditions. In this study the
five reactors described below were used to investigate the possible modes of
operation.
PULSED MICROREACTOR
The pulsed microreactor used (Figure 1) was the tubular glass liner of
the injection part of a Perkin Elmer 900 gas chromatograph. The gas
chromatograph could be connected to a mass spectrometer for product
identification. Carrier gas transported the injected reactants through the
reactor, which was placed in an aluminum heating block capable of varying
temperature from ambient to 400°C. Catalyst beds, typically 2 to 10 mm long
containing 2 to 10 mg, were pre-reduced with carrier gas (H2) at 250°C. The
organochlorine reactant (0.5 to 2jul) was then injected, and the reaction
products monitored chromatographically. The residence time, calculated on
the basis of carrier gas flow rate, was of the order of 10 seconds. The
pulsed microreactor was employed as a screening device, allowing approximate
determination of operating conditions.
CONTINUOUS VAPOR PHASE REACTOR
In the continuous flow process (Figure 2), the chlorinated organic was
vaporized into a stream of hot carrier gas and was passed through a heated
tubular (6mmID) glass reactor, where the gas reaction proceeded. The product
collection device had a cold trap. Catalyst beds typically 1 to 3mm long
containing 20-30 mg catalyst were located at the isothermal center of the tu-
bular electric furnace. The solvent used in the reactor system was p— xylene.
The total gas flow rate was measured with a bubble flow meter. The"
continuous reactor was well suited for catalyst deactivation studies.
LIQUID PHASE REACTORS
Batch processes studied employed supported metal catalysts slurried in a
well stirred reaction vessel. This configuration allowed use of high
pressure coupled with high temperature. Three reactors of this type were
used:
(1) Low Pressure Glass Reactor
This reactor (Figure 3), designed to measure H2 uptake and product
5
-------
COLUMN
CONNECTION
HEATER BLOCK
009-1324
SEPTUM CAP
009-1356
^INJECTOR
SEAL ASSY
009-0284
INJECTOR
009- 0183 ( 1/8-IN )
009- 0164 i 1/4-IN I
CARRIER
'SUPPLIED WITH INJECTOR MODIFICATION KIT 009-0295
- Model 900 Liquid Sample Injection System (Top View)
catalyst bed
glass'
wool
isothermal
Figure 1. Pulsed microreactor
-------
helium
drierite
ascarite
hydrogen
electric furnace
to flow meter
sample collection device
ice bath
Figure 2. Continuous vapor phase reactor
-------
.' • f"
-» ' - ,
• .7.» ,
copper
deoxygenator
5A molecular sieve bed
sampling
device
j acketed
reactor
magnetic
stirrer
differential
manometer
hydrogen
nitrogen
Figure 3. Low pressure glass reactor
-------
distribution simultaneously, was surrounded by a water jacket which was
connected to a thermostat. The system included a mercury-filled burette for
H2 uptake measurement and a sampling port from which a small amount of
product could be withdrawn for chromatographic analysis.
(2) High Pressure Stirred Reactors
(a) Teflon-Lined Autoclave
A 750cc teflon-lined autoclave (Figure 4) was used for liquid phase
studies at elevated hydrogen pressures and temperatures up to 250°C. The
autoclave was equipped with a stirrer, thermocouple, and high pressure
sampling device. In the operating procedure, solvent, catalyst and acid
acceptor were heated under a high pressure N2 atmosphere to reaction
conditions before H2 was added by difference at time zero.
(b) Rocking Autoclave
Preliminary high pressure liquid phase experiments were conducted in a
100 cc glass-lined rocking autoclave (Figure 5), which could be operated at
pressures up to 13,000 psi. The reactor shell was secured horizontally in a
heating jacket and the entire unit was rocked to obtain mixing. This system
was operated at 750 psi with time zero being the time at which H 2 was charged
to the reactor.
-------
hydrogen
copper
deoxygenator
nitrogen
5A molecular sieve bed
induction
stirrer
thermocoupl e-
O I
£
4 blade
turbines
sampling
valve
Figure 4. High pressure teflon-lined autoclave
-------
copper deoxygenator
(£)©
hydrogen
nitrogen
-o
5A molecular
~~^~\ „ sieve bed
eccentric
drive
Figure 5. High pressure rocking autoclave
-------
SECTION 4
ANALYTICAL TECHNIQUES
GC-MS techniques were used for identification and quantitative analysis
of all reactants and products studied during the course of these investiga-
tions on the hydrodechlorination of pesticides and related substances. The
equipment used was a Perkin-Elmer 900 dual flame ionization detector gas
chromatograph interfaced with a DuPont 21-491 double focusing mass spec-
trometer. A CSI model 208 automatic digital integrator was used for quanti-
fying FID output. Mass spectra were recorded using a CSI 5-124A recording
oscillograph. Molecular weight determination and structural information
obtained from the mass spectra of individual chromatographic peaks identified
reaction intermediates in the complex systems studied. The FID response was
found to be insensitive to chlorine content, thus allowing conversion and
rate data to be obtained from comparison of the area under the peaks on the
chromatograms.
AROCLOR ANALYSIS
The total number of polychlorinated biphenyl compounds has been reported
to be 210 (22, 23). Consequently, identification of each individual isomeric
compound in an unknown sample was not practical. In this study, product
compositions were quantified under two classifications. Firstly, unknowns
were grouped under the number of chlorine atoms per molecule. Secondly,
within each of these groups, products were further characterized by total
ortho versus total (meta + para) substitution of chlorine. Gas chromatograph
mass spectrometer analysis provided all the information necessary for the
first determination (25, 26, 27). A chromatogram representative of this type
of determination is given in Figure 6. The peaks, each of which is repre-
sentative of a particular chlorinated biphenyl, are labled with the number of
chlorine atoms contained in that molecule. The Aroclor 1248 used in this
study was found to contain an average of 3.92 chlorine atoms per molecule and
to have a molecular weight of 2892. Webb et al. (24) identified the major
components of Aroclor 1221-1254 using gas chromatograph techniques. The
quantification under chlorine substitution was obtained from the gas chromat-
ograph data of this study by cross referencing with the published work of
Webb.
1, 1-DIPHENYL ETHANES/ETHYLENES ANALYSIS (DDT, ODD, DDE)
Two different columns were used for analysis of chlorinated 1,1-diphenyl
ethane/ethylene reactants and products. The mass spectra of a number of
2-chloro substituted l,l-bis(p-chlorophenyl) ethanes and ethylenes have been
reported (27, 28) and the spectra of corresponding hydrocarbons are avail-
able (29). However, mass spectra of reaction intermediates produced in
12
-------
'14
1
3
A
q
3
r
\
3
V
numbers denote
C1 atoms/molecule
16
'18 '20
Retention time (mins)
Fiqure 6. Aroclor 1248 chronatogram, 9'xl/8" 3% Ov-17 on chrcmosorb W-HP,
122-250°C @ 5°C/inin,He 30cc/hin.
13
-------
this study were not available. The process followed in developing these
spectra and identifying the reaction products is outlined below.
ANALYSIS OF DDT AND ODD
The mass spectrometer fragmentation of every reported 1,1-bis
(p-chlorophenyl) ethane is characterized by the following reaction:
- +
H
- 0C1
0C1 - C -
Based on this information, similar behavior was proposed for reaction
intermediates and this behavior was in fact observed. These reactions are
diagrammed along with the relevant mass spectra in Figure 7. In addition to
identification of the major fragmentation products, discrimination between
products and reactants based on differences in aromatic chlorine content is
thus obtained using the mass spectrometer. Differences in retention time in
the gas chromatograph of reactants and products differing only in aliphatic
(the R group above) chlorine content in conjunction with the observed mass
spectra, allowed complete identification and quantification of the compounds
involved in a given reaction.
ANALYSIS OF DDE
The mass spectra of chlorinated 1,1-diphenyl ethylene reactants and
products reported in the literature (28, 29) and obtained in this study were
analyzed. In all cases, fragmentation of the olefinic bond, which would be
analogous to the initial reaction in the DDT MS analysis did not occur. This
complicates the analysis, as the mass spectra obtained for DDE reaction
intermediates containing the same number of chlorine atoms per molecule were
similar. Compounds containing olefinic chlorine could not be distinguished
from compounds containing aromatic chlorine based on mass spectra alone.
Chromatographic retention time data was used to complete the identification
with the exception that the identity of 1 and 2 chloro compounds were
ambiguous.
ALDRIN AND DIELDRIN ANALYSIS
Partial identification of aldrin and dieldrin and their
hydrodechlorination products was carried out using the GLC/MS system. The
mass spectrometer fragmentations of dieldrin and aldrin were characterized by
retro-Diels-Alder(RDA) reactions (27) accompanied by expulsion of HC1 or Cl
as in the reaction:
14
-------
100
le Intensity
a
01
0
m e
100
>-
o
11
PJ
11
Si
0
100
Ul
u
•5
D
n
167
-1
REACTION INTERMEDIATES
C3-C-0+
m/e 167
H ~] t
-CHC12. 0-C-O 1
1 CriCl
m/e 250
ll
150 200 250
201
i n
iJl,
H •+ '! — I +
ClO-C-0 -ci fl. I l(/-C-i/ '
m e 201 ^ CH-.C1
250
I . l.i. . , 1 i,,
- - - — , -- - - -i , - - -. - - — — — , , , , , , — 1 '• ' i-1 —
150 200 2;0
ClO-r-OCl4' -ur. (.[O-i'-Mf I' '
3 i
m/e 235 •« fll,
i
•i /e 2?Q
2"0
,ll 1
I
. ,,l,| ll
, ,1 ll ll
e 130 ' -00 ' j;,o
Figure 7. Representative mass spectra of DDE/DDD
15
-------
Ci2H8Cl50 m/e = 343
dieldrin
"e
.j
Similar spectra were obtained for dieldrin hydrodechlorination products
according to the reaction:
C1ZK9C150
HC1.
Because these spectra were characterized by RDA reactions, no information
could be obtained on which chlorine was removed during hydrodechlorination.
Mass spectra of aldrin, where one olefin had been saturated during
hydrodechlorination, yielded more information, allowing further but not
complete identification. The extent of analytic capability is indicated in
Table 1 for aldrin and in Table 2 for dieldrin.
TOXAPHENE ANALYSIS
Toxaphene is a complex mixture of polychlorinated camphene isomers.
Toxaphene and its hydrodechlorination products were analyzed using GS-MS
techniques. Mass spectra have been reported (30) for toxaphene and analyzed
(31) in conjunction with the data of this study. The determination
illustrated in the chromatogram of Figure 8 is at best qualitative.
TOXICOLOGY
As a part of this study, EPA arranged that starting materials and
products used in this study would be examined for toxicity by the Toxicology
Division, Biomedical Laboratory, Edgewood Arsenal. The LD50 of the test
solution was--injected intravenously into the tails of white mice. Some of
the test results are reported as preliminary, while others show that the
solvent (hexane) is more toxic than the pesticide. It was not possible to
distinguish between the starting materials and ordinary acceptable chemicals
(i.e. solvents used) using the techniques employed. Therefore, the use of
this toxicity test data for assessment of environmental acceptability of
material used in this study was not recommended by the researchers.
16
-------
TABLE 1. RETENTION TIMES FOR MAJOR ALDRIN PRODUCTS
(91 x 1/8" 3% OV-17 on chromosorb W-HP 102-250°C @ 8°C/min 30 cc/min He)
R.T. (min) MS IDENTIFICATION
10'6 C12H16C12
14.1 C12H15C13
15.1 C H Cl
1212 1+
15.6 C H Cl (ALDRIN)
128 6
16.2 C H Cl
i2 1 1 5
16.4 C H Cl
1210 6
TABLE 2. RETENTION TIMES FOR MAJOR DIELDRIN PRODUCTS
(91 x 1/8" 3% OV-17 on chromosorb W-HP 102-250°C @ 8°C/min 30 cc/min He)
R.T. (min) MS IDENTIFICATION
18.0 • C12H10C140
18.4 C -H.CIO (DIELDRIN) + C. H Cl 0
i/ob 1Z95
17
-------
00
0 Cl
C10H18
C10"14
1 Cl
2 Cl
3 Cl
AC:
ci
C10H9C13
10H12C14
C10»8C14
1Q 12 14
Retention time (min)
16
1 8 20
Figure 8. OV-17 GC/MS analysis of toxaphene hydrodechlorination products,
9'xl/8" 3% OV-17 on chronosorb W-HP, 82-250°C @ 8°C/inin.
-------
SECTION 5
RESULTS
AROCLOR 1248
Aroclor 1248, a clear viscous oil containing 48 wt% chlorine, v^s the
polychlorinated biphenyl (PCB) used as the model reaction mixture. Three
sets of experiments were conducted using Aroclor 1248. Firstly, the pulsed
microreactor was used to verify reactivity of the volatilized PCB in the
presence of the Pd catalyst. Secondly, preliminary experiments in the teflon
lined autoclave established the reactivity in the liquid phase using the Ni
catalyst. Thirdly, continuous sampling experiments in the teflon lined
autoclave provided data for rate and selectivity analysis. Chlorine
conversion, X, is used as the indicator of extent of reaction and is defined
as:
n
X = 1 - i=0
n
EM
i=0
where d^is the percentage of biphenyl containing i chlorine atoms per
molecule, andd^Jis the initial percent of a given molecule containing i
chlorine atoms.
1. Gas Phase Pulsed Microreactor Screening
Four experiments were conducted at the same temperature (220°C), pressure
(2.3 atm.Hz), space time (10" seconds), and feed composition (0.7 wt% Aroclor
in n-heptane). The concentrations of Pd catalyst on 5mg of A1203 support
were varied. These experiments demonstrated the feasibility of catalytic
hydrodechlorination of Aroclor 1248. The results are presented in Table 3. A
semi-leg plot of (1-x) versus g Pd indicates that the chlorine conversion
approximates first order behavior in Pd.
19
-------
TABLE 3. PULSE MICROREACTOR AROCLOR 1248 HYDRODECHLORINATION
220°C 2.3 ATM H2 T - .Ql sec.
Catalyst wt (mg)
wt % Pd
00o
00i
002
003
00t
005
X
5.0
.00
.003
.013
.305
.493
.189
.013
4.6
.05
.134
.058
.129
.337
.236
.091
.304
5.0
.10
.296
.068
.140
.279
.166
.046
.469
5.0
.35
.758
.043
.094
.078
.026
.855
2. Preliminary Liquid Phase Experiments
In these experiments the teflon-lined autoclave was initially charged
with Aroclor in ethanol using 61% Ni on kieselguhr as catalyst, and with the
acid acceptor, NaOH. The reactor was then heated, pressurized with H2 ,
maintained near these conditions for the reaction time, and finally cooled.
The contents were analyzed at this final condition. Temperature, H2
pressure, reaction time, and ratio of weight of reactant to weight of
catalyst were varied. The conditions and results are summarized in Table 4.
The following conclusions were drawn from this set of experiments:
1) Complete hydrodechlorination of PCB can be obtained using Ni catalyst
at elevated H2 pressure (50 atm) and temperature (115°C) in a six
hour reaction period.
2) The reaction has positive dependence on H2 pressure above 25
atmospheres.
3) Catalyst poisoning occurs at 175°C placing an upper limit on reaction
temperature.
3. Continuous Sampling Liquid Phase Experiments
Experiments were conducted at four temperatures (68°, 80°, 100°, 130°C)
in the teflon-lined autoclave. In each experiment, the H2 pressure (50 atm),
catalyst (Ni), acid acceptor (NaOH), and initial reactant composition (2 wt%
Aroclor in ethanol) were reproduced. Samples.were withdrawn and analyzed at
20
-------
TABLE 4 PRELIMINARY LIQUID PHASE AROCLOR EXPERIMENTAL CONDITIONS
Run
Aroclor 1248 (gm)
Aroclor 1248 (gm mole)
Ethanol (gm.)
Wt. % Aroclor 1248
NaOH gms
NaOH (gm-mole)
NaOH
Aroclor 1248
Ni on Kieselguhr
wt. Aroclor
wt. catalyst
Reaction time (hrs)
Pressure (Bar)
Temperature °C
Chlorine Conversion
1
5.27
.0180
98.6
5.07
4.0
0.100
5.5
2.5
2.1
2
50
100
.628
3
4.57
.0157
86.8
5.00
2.6
.065
4.1
0.55
8.3
2
50
100
.293
5
11.0
.0377
110.5
9.05
6.4
.160
4.2
1.1
10.0
4
50
100
.574
15
13.3
.0455
131.8
7.38
7.3
.183
4.0
1.39
9.6
2
50
.100
.559
16
12.19
.0417
110
9.98
6.66
.167
4.0
1.25
9.75
6
50
(50-42)
100-115
1.000
17
11.29
.0387
102
9.97
6.12
.153
3.95
1.12
10.08
2
25
(25-21)
100-110
.390
18
12.01
.0411
108
10.01
6.50
.164
3.99
1.23
9.76
2
50
(50-48)
25
.099
19
12.96
.0444
117
9.97
7.04
.176
3.96
1.30
9.97
4
50
(54-47)
100
.888
20
10.67
.0365
96
10.00
5.80
.145
3.97
1.08
9.88
1.2
O(atm)
25-100
0.00
21
11.45
.0392
103
10.00
6.23
.156
3.97
1.11
10.32
2
50
(50-60)
175
.107
-------
intervals during the reaction. Selectivity data are plotted versus chlorine
conversion in Figures 9, 10 and 11. Selectivity data obtained at these four
temperatures were sufficiently similiar that they were treated as one set of
data. Two models ware developed to interpret the reaction results.
(a) Consecutive Reaction Scheme
The observed selectivity data were filled to a consecutive reaction
'54 •*•!» 3
> 004
k21
scheme described by a set of first order reaction path expressions.
5lt
" ^
323
= k<1,00, - k,
The best fit was obtained with the following set of relative rate constants.
k,
k,
0.40
0.23
0.36
0.40
22
-------
130°C
D 100
• 80
o 60
0.4
0,2
0
Figure 9. Aroclor 1248 relative product distribution data (50 atm.H2,
Ni on kieselguhr).
23
-------
• 130 C
100
80
o 60
.2
.4
.6
0.8
1.0
1.0
Figure 10. Aroclor 1248 relative product distribution data (50 atm.I
Ni on kieselguhr).
24
-------
.6
.2
130°C
100
80
60
.6
1.0
Figure 11. Aroclor 1248 relative product distribution data
(50 atm. E2' N^ on kieselguhr).
25
-------
The curves obtained using this set of constants are superimposed on the
data in Figures 9, 10 and 11. A semi-log plot of (1-X) versus time for data
is given in Figure 12. An analogous plot obtained from the model would be
convex and consequently in disagreement with the data. A Langmuir Hinshelwood
kinetic model which takes into account adsorption effects provided insight
into this problem.
(b) Relative Rates of Isomeric Reactions
In the course of these studies it was observed that, within each group of
PCB isomers containing the same number of chlorine atoms per molecule,
certain isomers were selectively produced and reacted. In all cases, during
a reaction the amount of ortho substituted isomers increased relative to the
amount of meta and para isomers present. First order rate expressions were
proposed.
dO = k °
d9 °
dMp = k MP
~
Dividing these expressions and integrating gives:
log _0
Oo = k
log MP k
MP° np
A log-log plot (Figure 13) fits a straight line as predicted by the model. /
differing slope for the 130°C experiment is indicative of catalyst
deactivation.
DDT, DDE, ODD
Three reactors were used in this portion of the study. The pulsed
microreactor was used to screen for the thermal and catalytic reactivity of
DDT and ODD. The continuous gas phase reactor was used to hydrodechlorinate
ODD and DDE. The teflon-lined autoclave was used in a series of
dechlorinated reactions involving DDT and DDE.
1. Gas Phase Pulsed Microreactor Experiments
The three sets of experiments conducted in this reactor established the
practicality of catalytic hydrodechlorination of DDT, ODD, and DDE.
a) Using no catalyst and no hydrogen and with helium as the carrier gas, it
was found that 0.5 /ul samples of 9.5 wt% DDT in p-xylene underwent thermal
decomposition over a range of temperatures (200-360°C). DDE was the major
reaction product.
26
-------
,05
60°C
• 80
D 100
• 130
\
4
8
16
12
Time (hrs.)
Figure 12. Experimentally observed Aroclor 1248 kinetics.
20
22
27
-------
1.0
0.5
0.2
MfP
0.1
0.05
0.01
0.5
0.2 0 0.1
0°
0.05
Figure 13. Aroclor 1248 hydrodechlorination as characterized by
o~ vs. (mf) substitution.
28
-------
b) A set of six experiments was conducted using 0.5/ul samples of 5.2 wt% DDT
at a pressure of 2.3 atm, a temperature of 220°C, and a space time of .01
seconds. The first two experiments investigated thermal reactions in the
empty reactor with He and H2 as the carrier gases. The remaining four
experiments were conducted with H2 carrier gas, Al2O3support, and varying
concentrations (0.0, 0.05, 0.10, 0.35 wt%) of Pd catalyst. Chromatograms for
the products of these experiments are presented in Figure 14. The results
show that the thermal reaction was minimal but that Al203(no Pd) catalyzed
hydrodechlorination of the aliphatic part of DDT, with DDE as the major
reaction product. Aromatic hydrodechlorination was affected only with Pd
present.
c) This set of six experiments was identical to set (b) with the exception
that O.Sjul samples of 5.1 wt% ODD were injected. The results, presented in
Figure 15, show that the thermal reaction was minimal, that the
hydrodechlorination catalyzed by A1203 occurred to a lesser extent than with
DDT, and that ODD was generally less reactive than DDT with Pd catalyst
present.
(2) Continuous Gas Phase Reactor Experiments
(a) ODD
In a single experiment, the rate of ODD hydrodechlorination at a
temperature of 220°C and hydrogen partial pressure of 65 ± 5 mm Hg was
measured by varying ODD partial pressure from 0.214 mm Hg to 0.529 mm Hg, and
by varying space time from 0.0582 seconds to 0.0674 seconds. It was found
that the initial reaction, which is diagrammed in Figure 16 along with the
data, occurred in a single step without intermediate desorption.
(b) DDE
The following three sets of experiments were conducted in the continuous
gas phase reactor with DDE as reactant:
i) An experiment, conducted at 200°C, hydrogen partial pressure
of 555 mm Hg, DDE partial pressure of 0.688 mm Hg, space time
of 0.0482 seconds, and with 0.35 wt% Pd on A1203 catalyst,
demonstrated that catalyst activity declined rapidly during the
first 24 hours of a run, but stabilized thereafter. All
kinetic data was taken after the catalyst had been on stream
for more than one day.
ii) In an experiment conducted at conditions identical to
experiment (i) above, with the exception that no Pd was
present, it was found that A12 03 did not catalyze
hydrodechlorination of DDE. This result differed from that
obtained in the pulsed microreactor.
iii) In a series of five experiments conducted at three tempera-
tures and three partial pressures of hydrogen, rate and
29
-------
1000
100
8 10
j-i
§ i
1000
a; 100
a 10
s
§
1000
100
10
>n
Empty Reactor
Empty Reactor
50 mq oc1 - AI-0,
4 6 mq 005% PaV-AI2Oj
Aromotic I Aliphatic
Reactions I Reactions
6 10
12
I t
50mq 001% PaV-
I
HO
50mg 035%
V< K
5 10 15 20
TimeCmin.)
5 10 15
Time (min.)
20
Figure 14. DDT hydrodecMorination
(30 scc/^nin.gas,220±10°C,19psig).
30
-------
1000
100
LQ
S'
50 mg oi -AI203
16
H2
46 mg 005%
Aromatic I Aliphatic
Reactions I Reactions
10
I
6 9
16
. ill.1 ll
T
H2
50 mg O 10% Pd
6 9 '2 16
I
10
(5;
50 mg 035% Pdo<-AI203
10 15
Time (min.)
20 0
5 10 15
Time (min.)
20
Figure 15,
EOD hydrodechlorination
(30 scc/min. gas, 225* 5°C, 19 psig)
31
-------
Product distribution
Ul
CTl
(D
I
I
o -
o -
o
•
U)
o.
h •
8?
3 01
I
CO
-------
selectivity data were obtained by varying space time and DDE
partial pressure. Five major DDE (species VI)
hydrodechlorination products were observed. These compounds,
which amounted to 95% of the total products, are labeled with
Roman numerals in the following list:
I,1-biphenylethane
H
1-chlorophenyl, 1-phenylethane Cl(|>-C-$ II
H
1,1-bis-parachlorophenylethane Cl-C-4>Cl III
CH3
1,1-biphenylethylene dichloride (/HC-0 IV
CO..,
1-chlorophenyl, 1-phejiylethylene 0-C-^Cl v
dichloride 1L,
CUL2
Chlorine conversion was used as an extent-of- react ion parameter and is
j j i=4
defined as ™o~™ , where 0= ^~? id . , 0ois 4, based on the chlorine
~~ = 1
content of starting material, i=number of chlorines per molecule, and
Cli= mole fraction of compounds having i chlorines per molecule.
Representative selectivity data along with temperature and pressure
conditions at which they were obtained are presented in Figures 17 and 18.
The solid curves and reaction constants are a computer fit of the data
calculated from the reaction path model represented above the curves. The
complete reaction scheme suggested by the data is as follows:
33
-------
VI
_f-C;-^
CC12
V H(|)-C-
H i
I
^32
H(^-C-i-Cl
V21
III
II
The following three assumptions were employed in the modelling:
1) The reactor is back-mixed. This assumption is supported by the relative-
ly high value of the vessel dispersion number. 2) Langmuir Hinshelwood
kinetics describe the reaction system. 3) First order functionality
represents the rates. The rates were:
rvr = - ^ + k63) VI
= k
65
V
rIV = k54 V - k41 IV
= k63 ^ ' k32
r = k32 III + k52 V - k2l II
II
rl ~ k21
34
-------
ODDE
• V
A III
X IV
II
v I
T
0.2 0.4 0.6
Chlorine conversion
0.8
Figure 17. DDE Selectivity Plot (A)
(T=200°C, P « 140 ± 30rrm. Hg)
2
35
-------
Product distribution
00
to
CD
" 1
10
•1 ro
H
-------
Define relative rates:
ry* = rV , rw* = rIV , rin* = rill , r*= rll , r-,-* = rl
rVI rVI rVI rVI rVI
and V* = V , IV* = IV , III* = III , II*= II , I* = I_
VI VI VI VI VI
Also define relative rate constants:
a = kgs d = k63 b = k52 e = k32
k65+k63 k65+k63 k65+k63 k65+k63
c = k54 g = k2, f = k41
k65+k63 k65+k63
then:
rv* = a - (b + c) V*
rm*= d - e III*
riv* = cV* - IV
rI;[* = elll* + bV* - gll*
r-j-* = IV* + gll* and a 4- d = 1
Rate constants defined above relative to (kg,-+k63) were matched with experi-
mental results. - .
''• " ''97-• v l' ' ". k -•»«•--'•'* '• •
-------
200°C, PH2=60nri Hg 200°C, % = 140nm Hg 200°C, PH =580 ran Hg
^ "^
0.63
0.17
0.37
• \
0.59
' ^j
0.12 0.44
iT
0.08 0.15
y i
V
0.56
r ^
0.55
t- *!
0.32 0.28
'-" -^
0.72
0.2
v
0.09 0.08
f 1
0.30
0.0
• ..._.:
Q.Jb
0.46
0.42
170°C, PH=580 ran Hg
0.72
0.28
0.32
230°C, PH-,=580
2
0.66
0.02 0.34
0.18
0.22
0.23
0.03
0.29
Figure 19. DDE relative rate constants
( 0.35% Pd on «*<- A12O3)
38
-------
Temperature changes did not appreciably affect the relative rates of olefinic
and aromatic hydrodechlorination at constant temperatures. If it is assumed
that olefinic hydrodechlorination is of order n with respect to H2 partial
pressure and aromatic hydrodechlorination is of order m with respect to H2
partial pressure, it can be shown that the relative rate constants of
olefinic reaction decreases as H2 partial pressure decreases. The same trend
is demonstrated by a Langmuir Hinshelwood model based on dissociative
adsorption of hydrogen and bimolecular surface reaction of DDE with hydrogen.
Correlations, based on assumptions regarding the relative sizes of
equilibrium adsorption constants, demonstrated that the rate of
hydrodechlorination of DDE increases linearly with the product of DDE partial
pressure and the square root of the H2 partial pressure. Similar
correlations demonstrated that the rate of removal of aromatic chlorine and
of olefinic chlorine from DDE should be linearly proportionate to the product
of DDE partial pressure, and a function of the hydrogen partial pressure.
For aromatic chlorine, this function was the square root of the partial
pressure of hydrogen; while for olefinic chlorine the function was the
partial pressure of hydrogen.
An additional result of these experiments is the following reactivity
sequence. Initial hydrogenation of DDE did not occur and the fact that only
trace olefinic intermediates were observed may indicate the existence of a
non-equilibrium adsorption-desorption sequence. A comparison of ODD and DDE
reactions conducted at similar conditions used in conjunction with the fact
that olefinic chlorines are more reactive than aromatic chlorines, gives the
following reactivity order for hydrodechlorination in the ODD, DDE sequence.
In the ODD, DDE sequence:
I
-C- 0 - - C -
II I '
Cl, Cl CHC12
More reactive *Less reactive
3. Liquid Phase Reactions in the Teflon Lined Autoclave
A preliminary experiment was conducted with DDT in the teflon-lined
autoclave to determine if a hydrocarbon solvent, xylene, and the acid
acceptor Ca(OH)2 could be used in the liquid phase hydrodechlorination
process with Ni catalyst. Only trace amounts of the desired product were
formed, indicating that the acid acceptor should be soluble in the reaction
mixture and that ethanol as solvent is favored over xylene.
A more complete set of experiments using ethanol solvent and Pd and Ni
catalyst were carried out in the teflon-lined autoclave to investigate:
1) liquid phase stoichiometry; 2) reaction paths and associated rate
constants; 3) the effects of catalyst loading and stirrer speed on reaction
rate; and 4) the effect of hydrogen pressure and temperature on rates and
selectivity in DDE hydrodechlorination. Conditions for these experiments are
summarized in Table 5.
39
-------
TABLE 5 EXPERIMENTAL CONDITIONS FOR BATCH DDE KINETIC STUDIES CONDUCTED IN TEFLON AUTOCLAVE
99 gm ETOH, 1.0 gm DDE., .5526 gm NaOH
Activation Reaction
Exp. No. Catalyst Catalyst Wt. Temp. H~ Pressure Temp.
(rag.) (°C) (bar) (
•C)
DDE 1 Pd Black 30.0 220 50 100
DDE 2 10% Pd on C 30.0
DDE 3
DDE 4
DDE 5
DDE 5R
DDE 5R 1
DDE 6
DDE 7
DDE 8
DDE 9
DDE 10
DDE 11
DDE 11R
10.0
20.0
30.0
\
\
\
\
/
Stirrer
Speed
(RPM)
1500
60
60±5
59.5±1.5
, 24-47
900
1500
5 22-42
7 100±4
20 24-47
| 98±2
50 80.5±2.5
I 81.5±2.5
DDE 12 61% Ni on K 50.0 220 50.0 99±1
DDE 12R
DDE 13
DDE 13R
DDE 14
DDE 14R
DDE 15
DDE 15R
,
370
220
370
220
370
220
370
J,
100
81.5±3.5
80±1
60.5±2.5
60±3
27-42
25-41
f
LS-4* 10% Pd on C 62.1 2/!0 1 "atm 23
* 1.008 gm. DDT, 100 cc EtOH, 0.7421 gm NaOH
-------
(i) Liquid Phase Stoichiometry
Similarly to the gas phase reaction, both Ni and Pd catalyzed concerted
removal of both olefinic chlorines in the liquid phase reaction. ODD was not
observed, indicating that olefinic hydrodechlorination preceded olefinic
hydrogenation. The major difference between gas and liquid phase hydrode-
chlorination of DDE was that concerted or simultaneous removal of aromatic
and olefinic chlorines occurred in the liquid phase reaction. The overall
Stoichiometry for liquid phase hydrodechlorination of DDE is characterized
as:
VI
V H6-C-6-C1
IV
III
II
ii) Reaction Paths
Similar to the treatment described for Aroclor 1248, DDE
hydrodechlorination Stoichiometry can be described by a set of normalized
first order differential equations which simulate batch reaction behavior.
41
-------
d V = k65VI - (k52 + k5J V
d VI
d IV = k51tV -
d VI
-(k +k +k + k ) VI
65 63 62 61
d III = kegVI - k32III
d VI
- (k65 + k63 + kb2 + k61)VI
d VI
d I = k61VI + k21II + kltlIV
d VI _
There is no unique solution to this set of equations. However, simplifying
assumptions can be applied to yield unique solutions. The results obtained
through this approach are drawn over representative Figures 20 and 21. With
the values of the relative rate constants obtained, it is possible to
determine the total initial olefinic and aromatic hydrodechlorination rate
constants. For Pd catalyzed runs, it was found that olefinic
hydrodechlorination rate is more dependent on H2 pressure than the aromatic
chlorine reaction rate, and that the aromatic rate was favored at low
pressures. Selectivities were insensitive to temperature in the range 20 to
100°C for both Ni and Pd catalyzed reactions.
iii)Catalyst Loading and Stirrer Speed
Variation of the stirrer speed did not significantly affect initial rates
or intermediate selectivities, indicating that mass transfer limitation did
not occur over the range of speeds studied.
For the Pd catalyst, it was found that initial DDE reaction rate is not
proportional to catalyst loading as would be expected at low levels of
catalyst. It was proposed (32) that this was due to poison of some of the
catalyst by impurities. Due to this effect, all other Pd and all Ni
catalyzed experiments were conducted at higher levels of catalyst loading.
iv) Effect of Hydrogen Pressure and Temperature
DDE reaction rates in both Ni and Pd catalyzed experiments were not
accurately described by simple power law kinetics. Attempts were made to
describe the effect of changes in hydrogen pressure on the rate of DDE
disappearance for Pd catalyzed reactions. A log-log plot of initial rate
42
-------
Product distribution
00
u>
-------
Chlorine Conversion
.2 ,4 '.6
Chlorine conversion
Figure 21. Selectivity plot DDE 2, 4.
44
DDE
V
III
II
-------
versus H2 pressure had a slope of 0.72.
The kinetic behavior of Ni-catalyzed DDE hydrodechlorination is quite
different. The DDE concentration versus time curves are characterized by:
1) an induction period; 2) zero order behavior at intermediate conversions;
and 3) a shift to higher order functionality at high conversions. The nickel
catalyst requires reduction (33) and it was found that the length of the
induction period could be shortened by increasing prereduction temperature.
The induction period may also be due to product desorption rate limitation
(34). The remaining characteristics are consistent with the adsorption
reaction model of a disappearing component.
C. ALDRIN AND DIELDRIN
Pulsed microreactor and high pressure liquid phase reactions were carried
out to determine hydrodechlorination activity of aldrin and dieldrin.
Gas Phase Pulsed Microreactor
A set of six aldrin hydrodechlorination experiments were conducted at a
single temperature (220°C), pressure (2.3 atm), and space time (.01 seconds).
The first two experiments investigated thermal reactions in the empty reactor
with He and Ha carrier gas. The remaining four experiments were conducted
with H2carrier gas, A120 support and varying concentrations
(0.00, 0.05, 0.10, 0.35 wt%) of Pd catalyst. No thermal reaction occurred
but A1203 with no Pd converted 11% of aldrin to a less volatile compound
(aldrin with one olefin hydrogenated). As the concentration of Pd increased,
hydrogenation was quantitative while hydrodechlorination proceeded in
parallel. Aldrin chlorine was not highly reactive. One experiment using
hexachlorocyclopentadiene showed nearly complete hydrodechlorination
indicating that steric effects are important in aldrin dechlorination.
Liquid Phase Experiments
A set of preliminary dieldrin hydrodechlorination experiments were
conducted in the rocking autoclave. All employed Ni catalyst while
temperature, Ha pressure, time, and catalyst concentration were varied. The
major conclusion was that only at high temperature, high Ha pressure and high
catalyst concentration were significant amounts of completely
hydrodechlorinated product formed. Initial hydrogenation of the olefin was
not observed which was consistent with the DDE results.
Two experiments were conducted in the teflon-lined autoclave to determine
the reactivity of aldrin and dieldrin. The conditions were the same for both
experiments. Detailed analysis was not completed, but results indicate that
the aldrin olefin is hydrogenated and two chlorines removed, before the re-
action slows down. Dieldrin is not hydrogenated but does lose two chlorines
before the reaction stops. More severe conditions are required for further
reaction.
45
-------
D. TOXAPHENE
Toxaphene is a mixture of polychlorinated camphenes. Most of the
chlorine is aliphatic in nature. Thousands of isomers exist (34-36),
precluding quantitative analysis based on number of chlorine atoms per
molecule.
Liquid Phase Experiments
Four experiments were performed to determine overall reactivity. All
were performed at the same temperature, H2 pressure, and ratio of weight of
catalyst to weight of toxaphene. Reaction time varied. Under these
conditions, most chlorine was removed in the first two hours and molecules
containing only one or two chlorine atoms were resistant to further
hydrodechlorination. Toxaphene was more easily hydrodechlorinated than
aldrin or dieldrin, even though most of the chlorine is aliphatic. This
further substantiates the belief that steric factors account for the
non-reactivity of aldrin and dieldrin.
46
-------
SECTION 6
PROCESS DESIGN
Final design of a hydrodechlorination process is beyond the scope of this
project but quantitative features of the process may be proposed with the aid
of the data outlined.
Qualitative Aspects
Due to the variability of the pesticide and solvent compounds to be
processed, a batch process is indicated in spite of the possible advantages
of continuous systems. This same variability in the volatility of the feed
favors liquid phase processing. Ethyl alcohol is recommended as a solvent
over j>-xylene because it is a better solvent for the acid acceptor, and
having lower vapor pressure, it will contribute less to the pressure in the
liquid phase reactor. Although catalyst life under service conditions has
not been determined, the Girdler G49A nickel on kieselguhr used in this study
was satisfactory for both economic and activity reasons. Operation with the
HC1 as a by-product is unacceptable, and NaOH as acid acceptor was found to
be superior to Ca(OH)2 . Details of product recovery will depend on the
nature of the feed and the severity of the processing.
Quantitative Aspects
Operation at high temperature is favored on the basis of rate
considerations, but processing complications indicate that operation at
around 100°C is preferred. The solution step should be carried out at
atmospheric pressure, while the attainment of reasonable rates in the reactor
will require a partial pressure of H2 of 20 to 50 atmospheres.
Process Description
A conceptualized batch process employing ethanol, sodiun hydroxide,
catalyst and hydrogen is given in Figure 22. The process includes extractive
solution of pesticides in ethanol, reaction at high H2 pressure, recovery of
ethanol and extractive removal of soluble inorganics from the waste stream.
47
-------
condenser
ethanol liquid
solvent
holding
tank
containers,
wetting
agents
' rotating
high pressure
hydrogen
pump
/
-------
REFERENCES
1. Nbller, H., W. Low and P. Andreu. Naturwissenschaften, 51:211, 1964.
2. Andreu, P., M. Heunisch, E. Schmitz and H. Nbller. Z Fur Naturforschung,
196:649, 1964.
3. Noller, H., H. Hantsche and P. Andreu. Angewondte Chemie, 76:645-646,
1964.
4. Nbller, H., P. Andreu, E. Schmitz, A. Zahlout and R. Balesteros. A fur
Phipikolische Chemie N. F., 49:299-309, 1966.
5. Noller, H., G. Kabiersch and P. Andreu. A. Naturforschung, 236:894-900,
1968.
6. Nbller, H., H. Hantsche and P. Andreu. J. Catal., 4:354-362, 1965.
7. Nbller, H., P. Andreu and M. Hunger. Angeu. Chem. Internat. Edit., 10:
172-181, 1971.
8. Berg, O.W., P. L. Drosody and G. A. V. Pees. Bui. Environ. Cont. Tox.,
7(6):338-347, 1972.
9. Freifelder, M. Practical Catalytic Hydrogenation; Techniques and
Applications, John Wiley & Sons, Inc., New York, 1971.
10. Kammerer, H., L. Horner and H. Beck. Chem. Ber., 91:1376-1379, 1958.
11. Southwick, A. Unpublished Observations, Englehard, Ind., 1962.
12. Hasbrouck, L. Unpublished Observation, Englehard, Ind., 1966.
13. Meschke, R.W. and W.H. Hartung. J. Org. Chem., 25:137, 1960.
14. Baltzly, R. and A. P. Phillips. J. Am Chem. Soc., 68:261, 1946.
15. Reinecke, M.G. J. Org. Chem., 29:299, 1964.
16. Horner, L., L. Schlafer and H. Kommerer. Chem. Ber., 92:1700, 1959.
17. Weiss, A.H. and K. A. Krieger. J. Catal., 6:167, 1966.
18. Weiss, A.H., B. S. Gambhir and R. B. Leon. J. Catal., 12:245, 1971.
49
-------
19. Weiss, A.H. and B. S. Gambhir. Proc. V. Intl. Congress on Catal., Vol 2:
1319. North Holland Publ., Amsterdam, 1973.
20. Gambhir, B.S. and A. H. Vfeiss. J. Catal., 131:243, 1973.
21. Fleck and Haller. J. Amer, Chem. Soc., 67:1419-1420, 1945.
22. Smith, D. H. Anal. Chem., Vol. 47(7), June 1975.
23. Hutzinger, 0., S. Safe and V."Zitko. The Chemistry of PCB'S CRC Press,
Cleveland, Ohio, 1974.
24. Webb, R. G., and A.C. McCall. J. of the ADAC, 55(4), 1972.
25. Lao, R. C., R.S. Thomas, J.L. Monkman and R.F. Pottie. Symp. on the
Identification and Measurement of Environmental Pollutants, Nat. Res.
Council, Ottawa, Canada 144, 1971.
26. Perry, R. Symp. on the Identification and Measurement of Environmental
Pollutants. Nat. Res. Council, Ottawa, Canada 130, 1971.
27. Safe, S. and 0. Hutzinger. Mass Spectroscopy of Pesticides and
Pollutants. CRC Press, Cleveland, Ohio, 1973.
28. Bonelli, E. J., P.A. Taylor and W.J. Morris. American Laboratory 29-35
July 1975.
29. A. P. I. Research Project 44, Selected Mass Spectral Data Chem. Thermo.
Properties Center. Texas A&M University, College Station, Texas.
30. Holmstead, R. L., S. Khalifa and J.E. Casida. J. Agr. Food Chem.,
22(6), 1974.
31. Weininger, S. J. Interdepartment Communication.
32. Satterfield, C. N. Mass Transfer in Heterogenous Catalysis. M.I.T.
Press, Cambridge, Mass., 1970.
33. Anderson, J. R. Structure of Metallic Catalysts. Academic Press, London
1975.
34. Weiss, A. H., R.B. LaPierre and J. Shapira. J. Catal., 16:332, 1970.
35. Casida, J. E., R.L. Holmstead, S. Khalifa, J.R. Knox, T. Ohsawa, K.J.
K.J. Palmer and R.Y. Wong. Science, 183:520, 1974.
36. Smith, D. H. Anal. Chem., 47, June 1975.
50
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
EPA-600/8-77-013
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
CATALYTIC HYDRODECHLORINATION OF POLYCHLORINATED PEST-
ICIDES AND RELATED SUBSTANCES
An Executive Sunniary
5. REPORT DATE
September 1977 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Ebon Research Systems
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Ebon Research Systems
10108 Quinby Street
Silver Springs, Maryland 20901
10. PROGRAM ELEMENT NO.
1DC618
11. CONTRACT/G-FMrfW-NO.
68-03-2460
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Protection Agency—Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Executive Summary
14. SPONSORING AGENCY CODE
EPA/600/14
is. SUPPLEMENTARY NOTES See EPA-600/3-77-018 (PB-262 804/AS) "Catalytic Conversion of Haz-
ardous and Toxic Chemicals: Catalytic Hydrodechlorination of Polychlorinated Pesticides
and Related Substances" by R.B. LaPierre, E. Biron, D. Wu, L. Guczi, W.L. Kranich, A.H.
/feiss. Jan. 1977. Project Officer: Robert E. Landreth (513) 684-7876.
16. ABSTRACT
A study was undertaken of the catalytic conversion of chlorinated pesticides
and other undesirable chlorinated compounds to acceptable compounds. This study
shows that chlorine can be catalytically removed and replaced by hydrogen to
produce relatively non-toxic hydrocarbons.
The batch process involves use of a supported nickel catalyst, ethanol as
solvent, and sodium hydroxide as an acid-acceptor of the hydrogen chloride by-
product.
A reactivity sequence is established based on carbon-chlorine bonding wherein
olefinic chlorine is the least reactive. Reaction models are determined for Aro-
clor 1248 and DDE.
Removal of ortho-substituted chlorine is the limiting reaction in hydrodechlor-
ination of Aroclor. Aldrin and dieldrin are the most difficult compounds to hydro-
dechlorinate because of steric hindrance. Removal of aromatic chlorine is the
limiting reaction in the hydrodechlorination of DDT and DDE.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Aldrin
DDT
Pesticides
DDE
Deildrin
Hydrodechlorination
Polychlorinated biphen-
yls
Toxaphene
7C
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
59
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
51
GOVERNMENT PRINTING OFFICE 1977-757-056/6539
------- |