PCBs AND ASSOCIATED AROMATICS 949
teristic peaks of Aroclor 1260. After 30 minutes reaction time
the peaks attributable to the more highly chlorinated conge-
ners are either missing entirely or are much reduced. Peaks
attributable to the constituents of Aroclor 1242 are missing
from the original chromatogram but some are present in the
sample after 30 minutes reaction time. The fractional reduc-
tion of PCBs in the presence of a relatively large quantity of
HCBz is much smaller than the fractional reduction of HCBz
in the same time.
A similar experiment was performed with a starting
emulsion of Aroclor 1260 in water at 22,500 ppb PCB.
Approximately 320 mg of Aroclor 1260 was dissolved in
100 mL of warm methanol and added to 7.9 L of water. The
electrochemical cell arrangement was set up in the solution
with a supporting electrolyte.
After about five minutes the emulsion had become a
clear solution.
After thirty minutes the PCB concentration decreased
from 22,500 ppb PCB to 27.9 ppb PCB. The last sample was
taken after 90 minutes and the PCB concentration was found
to be 7.1 ppb PCB. The chromatogram of the final solution
was significantly different from that of the starting material
and resembled a mixture of Aroclor 1260 and Aroclor 1242.
The penta-, hexa- and septachlorinated biphenyls appeared
to have completely reacted while trace quantities of di-, tri-
and tetrachlorinated isomers remained.
The reaction was complete at 90 minutes. A sample was
analysed to determine any residual PCB content as well as
the complete range of chlorobenzene isomers and benzene.
Trace quantities of 1,2,3-trichlorobenzene, tetraCBz, penta-
CBz were found at sub ppb levels while hexaCBz was 4.3
ppb. Trichloro- and lower chlorinated benzenes were below
the method detection limit of 0.3 ppb.
Other work by the author has shown that the action of
sodium reagents on PCBs causes the sequential removal
of chlorine atoms from the biphenyl molecule until a
polymerized polyphenyl sludge is obtained. The change
in the relative peak heights shown in the chromatograms
of samples taken at different times was clearly indicated.
The later peaks contain C 12 H 3 Cl 7 isomers which appear to
lose chlorine and add to the earlier peaks with six, five
and fewer chlorine atoms. The PCBs which elute at shorterretention times probably contain five and four chlorine
atoms. Both of these peaks are relatively larger than in
the starting material which indicates that the more highly
chlorinated PCB congeners react faster than the less chlo-
rinated congeners. The kinetics of the process were not
established in detail because the dechlorination process
takes place too rapidly under the conditions used to allow
precise correlations to be developed.
The reduction in concentration by three orders of magni-
tude in 30 minutes reaction time is very similar to the reduction
observed in the HCBz experiment.
Chemical reaction from a relatively concentrated PCB
solution of Aroclor 1260 has been found to be reduced
from an initial value of 254 ppm PCB to 4.6 ppm PCB in
30 minutes.
At this point in time it is uncertain whether the final
product is polyphenyl hydrocarbons or whether the reac-
tion continues to break the phenyl-phenyl bond. There is
no observable precipitation of polyphenyls or other organic
residue. The turbidity of the initial solution caused by the
emulsion is completely eliminated and points to the probable
elimination of organics from the water. This is supported by
the observation that total organic carbon (TOC) is reduced
simultaneously with PCB dechlorination.
A further experiment was conducted with a 3,300 ppm
Aroclor 1260 emulsion in water. The process reduced the
concentration to 0.2 ppm in 30 minutes.
The following table summarises the data (Table 58).
The electrochemical process by its nature is easily con-
trolled and safe. The system is closed and has no emissions,
it operates at ambient temperature.
In any chemical dechlorination system in which there
is a possibility of the formation of hydroxyl radicals or
hydroxylated species there is a possibility that incomplete
reaction might produce hydroxylated species which are
more water soluble than the original PCB. While the tox-
icity of the hydroxylated product might be less than that
of the original PCB it is much more likely to be dispersed
into the environment because of its increased water solubil-
ity. The electrochemical process yields products which do
not contain chlorine as evidenced by the chromatograms of
samples taken from the process at different reaction times.TABLE 57
Contaminant of Concern (ppm) Initial concentration (ppm) Final concentration (ppm)
PCBs (Aroclor 1260) 3,300 0.2
PCBs (Aroclor 1260) 254 4.6
PCBs (Aroclor 1260) 22.5 0.007
Hexachloro-benzene 11.5 0.01
Monochloro-benzene 39.4 0.0005
2,3,7,8-TCDF 0.5 (ppt) 0.03 (ppt)
AOX 43 0.97C016_003_r03.indd 949C016_003_r03.indd 949 11/18/2005 1:12:46 PM11/18/2005 1:12:46 PM