\ I /
                   United States
                   Environmental Protection
                   Agency
Hazardous Waste Engineering
Research Laboratory
Cincinnati OH 45268
                   Research and Development
EPA/600/S2-86/113 March 1987
&EPA         Project  Summary
                   Catalytic
                   Dehydrohalogenation:  A
                   Chemical  Destruction  Method
                   for  Halogenated  Organics
                    Dehydrohalogenation shows poten-
                   tial as a means for converting certain
                   halogenated organics in wastes to inor-
                   ganic salts and gaseous aliphatic com-
                   pounds.
                    Dehydrohalogenation is a dehalo-
                   genation/elimination reaction that is
                   initiated by a strong base. The resulting
                   products are the halide salt, water, and
                   an elimination compound.
                    A novel reagent, sodium or potas-
                   sium hydroxide mixed with  a
                   polyethylene glycol, is a very effective
                   dehydrohalogenation  agent. This
                   reagent is shown to dehalogenate six
                   organic compounds that are represen-
                   tative of low molecular weight com-
                   pounds encountered in hazardous
                   wastes: CCI4, CHCI3, CH2CI2, C2H4Br2,
                   and CCI3NO2. Kinetics data for the reac-
                   tions of this reagent with the six com-
                   pounds is given to allow reactor design
                   and calculation of destruction effi-
                   ciency.
                    This Project Summary was devel-
                   oped by EPA's Hazardous Waste Engi-
                   neering Research Laboratory, Cincin-
                   nati, OH, to announce key findings of
                   the research project that is full docu-
                   mented in a separate report of the same
                   title (see Project Report ordering infor-
                   mation at back).

                   Introduction

                   Background
                    Halogenated organic compounds ac-
                   count for a major portion of toxic and
                   persistent hazardous wastes. Various
                   techniques have been proposed for de-
                   stroying these constituents. Incinera-
tion is a widely used destruction tech-
niques, but has limitations.
  There have been other methods pro-
posed and used for destroying halo-
genated species, such as ultraviolet
(UV) and UV/ozone degradation,
biotreatment with specially adapted mi-
croorganisms, and sodium metal reduc-
tion. Each of these techniques has
shown promise for destruction (or
transformation) of the halogenated spe-
cies, and each of these technologies has
certain applications wherein it is the
method of choice.
  Preliminary work has been initiated
toward development of an alternate
method for the  destruction of halo-
genated organics that is based on a
classical organic chemistry technique
for dehalogenation. It involves reaction
of the halogenated species with caustic
to produce an elimination reaction. The
products of the reaction are the halide
salts, water, and a multiple bond on the
organic molecule at the site of the de-
halogenation.

        H H
        I   I
      R-C-C-X + NaOH->
        i   I
        H H

        H     H
           \
                NaX + H20
             H
This reaction mechanism is referred to
as dehydrohalogenation.

-------
  Typical dehydrohalogenation reac-
tions are carried out in an initially anhy-
drous system in the presence of solid
caustic, or a small amount of water may
be used as a catalyst. These conditions
lead to very vigorous, and even uncon-
trollable, reactions. In order to make this
technology useful for the treatment of
hazardous waste, the following criteria
must be met:
  1. The reaction system must accom-
    modate wastes containing water.
  2. The reaction must be smooth and
    controlled.
  3. The methodology must provide
    treatment efficiency equivalent to
    or exceeding existing treatment
    techniques.
  The full report documents an experi-
mental program to validate the treat-
ment of halogenated waste compounds
by means of a novel dehydrohalogena-
tion reagent that meets the above listed
criteria. This reagent, composed of
caustic mixed  with a relatively small
quantity of  polyethylene glycol, has
been shown to be effective in rapid yet
controlled dehydrohalogenation in or-
ganic syntheses. The polyethylene gly-
col acts  as a catalyst in the reaction. In
this program, the reagent mixture cho-
sen for experimentation was potassium
hydroxide (KOH) and tetraethylene gly-
col (TEG). This reagent will be referred
to subsequently as KTEG.

Objectives
  There were four objectives in this pro-
gram:
  1. Validation of the efficacy of KTEG
    to destroy ethylene dibromide,
  2.  General, qualitative observation of
    the reactivity of KTEG with a series
    of  halogenated and nonhalo-
    genated compounds to determine
    those showing promise for treat-
     ment with KTEG,
  3.  Determination of reaction kinetics
     data to allow design of treatment
     systems and calculation of treat-
     ment efficiencies, and
  4.  Conceptual design  of  a possible
    treatment system.

Experimental Program
  The experiments for both the qualita-
tive observations and the kinetic studies
were carried out in a bench scale reactor
system  (Figure 1). The reactor vessel
was a 1  L three-neck flask. Attached to
the flask were a thermometer, a chilled
reflux condenser, a sample withdrawal
system, and a mechanical stirrer.  The
mechanical stirrer was used instead of a
                     Thermometer
                                         Gas Sampling Port
          Syringe
 Sample Vial
                                                                  Inverted
                                                                  Graduatet
                                                                  Cylinder
        Thermometer
                                       Temperature Control Bath
                           High Torque Stirrer
Figure 1.    Experimental set-up.
magnetic stirrer because the density of
the reaction mixture increased with
time; if a magnetic bar were used it
would become stalled in the slurry.
  A typical run is outlined below: 280 g
of potassium hydroxide (KOH) was dis-
solved in 180 ml water in the 500 ml
3-neck flask. A 5 ml portion of n-nonane
was added to the flask as an internal
standard. A 5 mL portion of ethylene
dibromide (EDB) was added, the system
was closed, and stirring was initiated.
The  system was  observed closely for
several minutes to assure that no reac-
tion was taking place. An initial sample
was taken from the reaction vessel, and
then 10 ml of tetraethylene glycol (TEG)
was added to the vessel.
  Reaction began immediately, as evi-
denced by vigorous bubbling, and gas
displacement versus time was plotted.
Gas samples and liquid samples were
taken at several intermediate reaction
times  and at the end of the reaction.
After the gas evolution had stopped, an
additional 5 mL  portion of EDB  was
added to the reaction mixture, and the
reaction resumed.  The sampling and
gas displacement readings were taken
as before. The entire process was  re-
peated until the  mixture  became  too
thick to  stir  and  the reaction  showed
signs of slowing.
  The reflux condenser was chilled to
0°C  (±5°C)  to  prevent escape of the
moderate boiling  point halogenated
species.  Gases generated in the reac-
tion passed through the reflux con-
denser and into the inverted graduatec
cylinder by means of a \" ID flexible
tube. Gas samples were taken at the toj
of the reflux column through a septum
closed sample port by means of a gas
tight syringe.

  Liquid samples were removed fron
the reactor through the suction mecha
nism  shown in  Figure 2. The empt^
sample vial was retained in place soleh
to prevent system pressure loss. Thre<
mL of water was placed in a 5 mL sam
pie vial and frozen. To remove a sampU
from the reacting mixture, the empt>
vial was removed and the syringe wai
filled with  approximately 5 mL of air
Static pressure within the sample tube
would  usually prevent any leakage  o
reactor contents during the sample via
transfer. The vial containing the deion
ized ice was transferred to the syringe
assembly, the  j mL of air pushec
through to clear the sample line, and the
syringe withdrawn 1 mL to create a suc-
tion and draw sample from the reaction
flask into the ice vial. The sudden cole
and dilution from the melting ice haltec
the reaction. This facilitated a reason
able delay in actually getting the sample
to analysis. The samples were then re
frozen for  preservation. Such sample;
taken at  intervals during the  reactior
provided process "snapshots" and en
abled detailed determination of the re
action kinetics. Analyses of the gas anc
liquid samples were performed by G
FID.
\

-------
                                                I
    Empty Vial in Place to
   Prevent Loss of Pressure
       in the Reactor
       Sample
                                              3 mL Ice
                                    To Reactor
                     Collecting Sample
                     from the Reactor
  Figure 2.    Liquid sample collection.
  Results and Conclusions

  General and Qualitative
  Observations
    A total of seven compounds were
  ested in the dehydrohalogenation sys-
  tem. These compounds are listed in
  Table 1 along with qualitative descrip-
  tions of the results of the dehydrohalo-
  genation. It should be noted that several
  chemicals appeared to react with the te-
  traethylene glycol (TEG) catalyst under
  the reaction conditions. The mecha-
  nisms for these non-catalytic TEG reac-
  tions are not  known at this time,  but
  such reactions always resulted in an
  overall slowdown of the desired dehy-
  drohalogenation. Carbon disulfide, for
  example, a nonhalogenated test com-
  pound, stoichiometrically reacted with
  the TEG to form a sticky brown sludge;
  this reaction prevented all further reac-
  tion  of the reagent with any halo-
  genated compounds. Carbon tetrachlo-
  ride  and chloropicrin both showed
  evidence  of slight consumption of  the
  catalyst,  but this did not  seriously  im-
  pair their dehalogenation. The effects of
  such compounds in  real world situa-
  tions can be overcome by  addition of
  excess reagent.
    Chloroform showed an  unusual reac-
  tion with  the TEG. The chloroform re-
  acted vigorously when added to the sys-
  tem,  and  it evolved a large volume of
^os very  quickly. If a large volume of
^•oroform was added to  the reactor, a
gray-colored foam formed, and this
foam solidified into a stable open-celled
spongy material. A portion of this foam
was cleaned with a series of solvents
and analyzed by infrared spectroscopy
(IR). From  this cursory analysis, the
foam appeared to be a substituted form
of the TEG.
  Chloroform could be  added in small
portions to the reactor without foam for-
mation. However,  under  these condi-
tions, the chloroform reacted smoothly
and consumed a small  amount of the
TEG in a similar manner to CCI4 and
chloropicrin.
  The reactions with ethylene dibro-
mide and ethylene dichloride were fast,
controlled,  and showed no consump-
tion of the TEG. The reaction products
consisted of vinyl bromide or vinyl chlo-
ride, acetylene, and the  halogen salt as
a reactor residue.

Determination  of Reaction
Kinetics
  The reaction rate constants for dehy-
dorhalogenations by the KTEG reagent
vary by temperature. Table 2 shows rate
constants for six compounds over the
temperature range of 21°-50°C.
  These rate constants may be used to
design reactors and to determine the re-
action efficiencies  for KTEG systems.
Since this particular dehydrohalogena-
tion reaction is a first order  (or pseudo
first order), nonreversible reaction, the
level of destruction  is dependent only
                                       upon time of contact of the halogenated
                                       compound and the KTEG reagent. This
                                       means that any desired level of destruc-
                                       tion may be accomplished by designing
                                       the  reaction system  for proper resi-
                                       dence times.
                                        The values of ka in Table  2 may be
                                       easily converted to other  units as
                                       needed. For example, the rate constant
                                       for ED6 at 21°C on a  molar basis is:
                                                                               90.2
                                              mLEDB      2.179gEDBx
                                           mole TEG • min    ml EDB
                                           1 mole EDB
                                          187.87 g EDB
               = 1.05
   mole EDB
mole TEG • min
Reactor Design
  Conceptual designs of reactors for
treatment of halogenated wastes were
developed by utilizing the kinetics data
described above  and the parameters
listed below:
WASTE
COMPOSITION  - 95% CCI4, 5% EDB
                (V/V)
REACTOR TYPE - (A) Plug flow, tubular
                (B) Batch
OPERATION
TEMPERATURE - 50°C
FEED RATE     - (A) 1 gal/min
                (B) 60 gal/hr
DESTRUCTION
EFFICIENCY    - 99.999% of the EDB
  Table 3 shows the necessary lengths
for a tubular,  plug-flow reactor accord-
ing to the tubular  inner diameter.
  Alternately, a batch reactor sized to
handle a feed rate of 60 gallons per hour
would have a volume of 550 to 600 gal-
lons. The batch reactor could operate at
atmospheric pressure, and it could be
fitted  with manual or automatic valves
for filling and discharging the reactor
contents.
  A practical  design for any dehydro-
halogenation  waste treatment reactor
must  be based on  realistic consider-
ations of material costs and reactor size
as well as its capability for achieving the
desired treatment efficiency. The ulti-
mate cost of treatment of a unit volume
of waste is determined by adding all as-
sociated costs (system construction, op-
eration, waste disposal, etc.), and divid-
ing that sum  by the volume of waste
treated. Thus, even though operating
costs  of a continuous feed process are
generally  lower than those of  a batch
feed process, the high initial cost of the
tublar reactor might never be offset,  if
there were relatively small volumes of
waste to be treated.

-------
Table 1. Compounds Investigated
Chemical Results
Carbon disulfide (CS2)

Methylene chloride (CH2CI2)
Chloroform (CHCI3)




Carbon tetrachloride (CCIJ

Quantitatively reacts with the TEC to form a brown
sludge.
Reacts slowly. Consumes some TEG.
Rapid, uncontrollable reaction if added too quickly.
Forms a stable, open-celled foam. Slow addition pro-
duces a controlled reaction without foam formation.
Consumes some TEG.

Moderate reaction rate. Consumes some TEG.

Table 3. Representative
Lengths
Diameter (in.)

2
4
6

8

W

12
Tubular Reactc
Length (ft

1,056
264
117

86

42

29
                   Cl  Cl
                I   \   I
Ethylene dichloride \ —C—C-
                V   i   i
                   Br Br
Ethylene dibromide\ — C— C-
                V   i   i
               Cl

Chloropicrin \ Cl—C—NO2

               Cl
Fast but controlled reaction. No evidence of reaction
with TEG.
Slightly faster rate than with EDC, but still controlled.
No TEG consumption.
Moderate reaction rate. Consumes some TEG.
Table 2.    Reaction Rate Constants by Temperature

                                      ka (mL Constituent/mole TEG • min)

      Compound
21"C
30°C
40°C
50°C
Ethylene dibromide (EDB)          90.2          153           267           449

Ethylene dichloride (EDC)         102           163           265           419

Carbon tetrachloride              16.0          22.3          31.6           43.7

Chloroform                       9.8          21.3          48.0          103

Methylene chloride                0.10          0.36          1.36           4.78

Chloropicrin                      5.72         N/D          N/D           N/D

N/D = no data.
  Modification  of key design  parame
ters, such as temperature of operatic
or feed rate, could change the size r<
quirements  of the tubular reactor suff
ciently to make it more cost effective a
a treatment  system design. The desig
parameters  of  a dehydrohalogenatio
reactor system  must be derived on
case-by-case  basis for the  specifi
wastes to be treated. This may be don
by using techniques similar to those de
scribed in the full report.

Recommendations
  Dehydrohalogenation technology i
to be developed and evaluated as a pos
sible alternative treatment of halocai
bon wastes. In  determining the relativ
advantages  and disadvantages of thi
technique, it is  necessary to compare
to existing treatment methods.
  The potential advantages of dehydrc
halogenation include cost, energy sa\
ings, materials recovery, and the  nor
production  of  harmful by-products
Dehydrohalogenation by-products ten
to be hydrolysis products of the parer
molecule; it is highly unlikely that corr
plex aromatic species will be produce'
in these reactions. The potential disac
vantages of dehydrohalogenation in
elude the production of waste gases, 01
ganics, and brine that may  requir
disposal.
  Future research with  this techniqu
should explore its practicality for treat
ment  of contaminated soils  and fo
scrubbing halocarbons from  gaseou
streams. The technique should b
tested  on actual wastes,  such as distilla
tion bottoms from  chlorinated solver
recycle, to determine the effects of com
plex waste  matrices on the treatmen
efficiency of the technique.

-------