Chemmphere, Vol. 39, No. I, pp.

81-87, 1999
Q 1999 Elsevia Science Ltd. All rights reserved
0045-6535/59/ s - see frunt matter

PII: !ao45-6535(!98)00590-6

IN SITU DECOMPOSITION OF PAHs IN SOIL AND DESORPTION OF

ORGANIC SOLVENTS USING MICROWAVE ENERGY

RUDOLPHA. ABRAMOVlTCH,*HUANG BANGZHOU,DOROTAA. ABRAMOVITCH’and

SONGJIANGAO

*Department of Chemistry, Clemson University, Clemson, SC 2%34-190

‘Department of Science and Mathematics, Anderson College, Anderson, SC 29621
(Received in Gemany 24 August 1998; accepted 15 October 1998)

Abstract. The decomposition of individual PAH’s (ber@aJpyrene, benxo~fluoranthene,

dibenz~hJanthracene. benz@mhracene, I-nitropyrene and benxo~quinoline) in soil in an open vessel using

microwave energy is reported. No PAH’s remain in the soil after remediation. An attempt to use “C-1abelled

benzo~pyrene to determine the fate of that contaminant on mmediation led only to the observation of the

formation of 0.3% of “COz. None of the other decomposition products could be extracted, isolated or quantitied.

Using a similar procedure, a variety of organic solvents (cyclohexane, l,l,l-trichloroethane, tetrachloroethylene,

benzene, toluene and xylene) could be efficiently desorbed and collected from contaminated soil. These

experiments offer a valid model for the _- in situ remediation of these pollutants from contaminated soil.
0 1999 Elsevier Science Ltd. All tights reserved

Introduction
A large number of studies have been carried out on soil mmediation and those for PCBs have been

summarixed in an earlier paper.’ Bioremediation of polychlorinated biphenyls (PCBs) and polycyclic aromatic

hydrocarbons (PAHs) in soil has been discussed.2 An assessment of the various industrially-developed bench-

scale technologies for mmediation of sediments contaminated with PCBs and PAHs has been carried out.’ These

inch&d the B.E.S.T.’ solvent extraction process using triethylamine, the ReTec low temperature thermal

desorption technology, the Holo-Flitem screw processor method, the ZIMPRO wet oxidation process which

oxidizes inorganic and organic contaminants under aqueous conditions at 175-320’ C and pressures of 20 MPa, and

SoilTech’s anaerobic thermal process (ATP) technology. The results were variable. For example, ReTec’s

processor desorbed >96% of the PCBs and %O% of the PAHs, white ZIMPRO’s oxidation process decomposed

only 29% of the PCBs but >98.9% of the PAHs. EtrN solvent extraction removed 9549% of the PCBs and 65-

90% of the PAH’s.

We have recently reported a model bench-scale method for the _-

in situ remediation of soils contaminated

with PCBs and other chlorinated aromatics using microwave energy.’ We now report the extension of these studies

to the remediation of soils contaminated with PAHs, as well as to the desorption of organic solvents tiom soils.

81
82

Experimental Section
Microwrve Decompodtionof PAHr
Benzo@Jpyrene, benzo~fluoranthene, diben&hJanthracene, benz@mhracene, 1nitropyrene and
benxo[flquinobne were purchased from AccuStandard Inc., New Haven, CT. The soil was collected from around
the Clemson University campus and cleaned as described previously.’ The cleaned soil was shown to be free of
any extractable PAHs.
For the remediation of soil contaminated with PAHs the procedure used was the same as that described for
the bench-scale h & remediation of PCBs-contaminated soil.” The clean soil was spiked with the individual the
bench-scale --
in situ remediation of PCBs-contaminated soil.’ The clean soil was spiked with the individual PAH
(the PAH was dissolved in redistilled, dried CH2Cls. the sohttion was added to the soil which was then thoroughly
mixed and air dried, eventually under water pump vacuum at room temperature to constant weight). For example,
soil (8 g) was spiked with benxo[alpytcne (1.008 mg). See Table 1 for other loading rates. In ah cases, the spiked
soil was left to stand for more than a week before extraction or microwave irradiation to ‘age’it. An ahquot of the
spiked soil was extracted repeatedly with pure CH&, the solution concentrated and analyzed quantitatively by
gas chromatography to recover the PAH almost quantitatively (biphenyl was used as the internal standard for the
quantitative analysis). The spiked soil (8.0 g) was placed in the quartx insert of the vessel in Fig 2 of ref. 4 and
then treated with a solution of Na#O4(200 mg) in water (2.2 ml). One pencil lead (from a # 2 pencil) (3.9 cm long,
2.0 mm dia.)’ was inserted in the soil such that it was covered by a thin layer of soil. The vessel was then placed in
a Kenmore domestic microwave oven (max. power 750 w), covered with the funnel and condenser (Fig 2 of ref. 4),
the condenser end was connected to a water pump, and the mixture was irradiated according to the time and power
sequence indicated in Table I. After the mixture had cooled to room temperature, the condenser and
funnel/distillation arm were washed separately with pure CH2C12,and the mostly vitrified soil was crushed and
extracted repeatedly with the same solvent. The washings/extracts were concentrated and each was analyzed
qualitatively (GCIMS) and quantitatively’ (GC). The same procedure was used for all the PAHs tested, and the
results are summarized in Table 1 (only a few representative runs are shown: many more were actually carried out
with similar results). Runs were also carried out using iron wire (one run with copper wire) instead of pencil lead,
as were blank runs (no PAHs) to determine what products might distil or be extracted following remediation. Iron
wins (6.5 x 2.0 mm) were inserted in the soil so that the tops were just below the surface. After much of the
water had evaporated, the soil was irradiated in the Kenmore domestic microwave oven (750 w max. power) for 10
min at 5% power, 10 min at 10% power, 10 min at 30% power and IO min at 100% power. After the mass had
cooled to room temperature the soil was found to have vitrified to a height of 5.5-6.0 cm, while the top l-l.5 an
was not vitrified.
 83

Mkrowrve Decompodtiaa of Beazo~pyrene-7-“C.

Thel~belkdcompoundin solutionwas purchasedfrom Sigma. The solution was diluted to I ml with pure,
dry toluene reauhing in an activity of 50 mCi/ml. 100 ml of the solution was added to a solution of non-
radioactive benzoupyrene (0.64 mg) in CH$& (7 ml), and this was added to 5 g of clean soil. One g of spiked
soil (2 mCi) was extracted with pure CHsClz, treated with scintillation fluid (Ecolite; ICN cat. # 882475) and
counted in a Beckman Liquid Scintillation counter LS-3 133L. Seventy-five percent of the benzoupyrene was
extractable; the rest was probably held tightly by the soiL6 Three gmms of spiked soil were treated with a
solution of Na$Oa (75 mg) in water (0.6 ml), placed in the quartz vessel and pencil lead was inserted (U.S.A.
Super Lo&bond Wallace lNVADER 5” #3). (NOTE: This is much narrower than the #2 pencil lead used above,
resuhing in the soil not being heated to the same high temperatures as in the non-labelled material experiments.
This was done so as to ensure that not all the benzo[_a]pyrenewas decomposed in the soil, to try and determine if
fragments or transformation products could be observed). After allowing the soil to stand for lh, it was irradiated
with microwave energy as before: 5 min (20% power), 5 min (30%). and 3 min (100%). The soil was allowed to
cool and then the condenser and fimnelkiistillation arm were washed repeatedly with pure CH&12, and the soil was

extracted separately continuously with the same solvent. Counting of the extracts showed the presence of 1.4% of
the original radioactivity in the distillates , and 43.4% in the soil extracts. When the procedure was repeated and
the off-gases passed also through a 0.1N NaOH solution in a tube connected to the condenser, so that gases evolved
were swept through by the low vacuum passed through the alkaline solution aRer any benzo~pyrene and/or
decomposition products had been trapped in the condenser, 0.3% of the total radioactivity was trapped as COz.
GClMS analysis of the condensed distillates and of the soil extractives indicated that they were virtually all
unchanged benxo~pyrene. Acidification of basitication of the remediated soil followed by extensive extraction
failed to yield any additional extractable material, not was any more 14C02released. As before: attempts were
made to detect any other material remaining in the soil by dissolving the remediated soil in aqua regia, diluting the
suspension somewhat (the soil precipitated out on greater dilution), adding scintillation fluid and counting but, as
before, no radioactivity was detected, probably owing to the scintillation fluid used not being stable to the strongly
acidic conditions. Thus, approximately 55% of the benz~pyrene decomposed but only a trace amount of that
could be accounted for, in the form of C02.

Deaorptioo of Organic Solvents ftom Soil Using Mkrowave Energy

Individual solvents (I g) were placed in the quartz tube (Fig. 2 of ref. 4) and covered with clean soil (10 8)“’
which reached to a height of 5.5 cm above the bottom of the tube. The apparatus waa moditied by adding three
momcondcnaer
veesels in series to the one in Fig. 2 of ref. 4, and anaching the water pump to the last of the four
condensers. The tint condenser was cooled in ice-water, the second in and ice-salt bath, and the third
TABLE 1. Microwave Decomposition of Individual PAHs in Soil
~~~~~--~~_----____--____-____----_----__--___-_----__---~---------_-________________________
PAH
m
0 (blank) EtEEs GUMS parent ion txaks at Y?
(3.9 cm x 2.0 mm) m/g 71,85,212,2i4,240,256.
284, 324,330, 338. 352.366,
374,408,423,436,45 1

Iron wire M(5). lo(10). lo(M), lo(lOO) y5y3Sayz ion peaks at 0

(3.5 cm x 1.5 mm)
Pencil lead 10(5), lo(10). lo(u)). 5(100) G&MS &nt ion peaks at 240. 0
(3.9 cm x 2.0 mm) 284.338,352,300,394,408,423

~4itlPY~~ Rncil lead 15(10), 5(20), 500). 3(W W&IS parent ion peaks at 0
(1.008) (3.9 cm x 2.0 mm) 57.71. 85.99

Benzo[hlfluoranthenc Ironwire 10(5), lo(lO), lO(30). lo(lOO) benzo[hlfluoranthene (0.0033 mg) 0

(0.992) (3.8 cm x 2.0 mm) ~62~Zj0.33%); trace peaks at
1 .

Coppcr~ 10(5), lo(lO), lo(30). lo(lOO) F2SM=nt km peairsa 0

(4.0 cm x 2.0 mm) .
Pencillead 10(5), lo(lO), lo(30). lo(lOO) Benz a anthracene (0.034 mg) 0
(3.8 cm x 2.0 mm) &2 Y 8 (3.54%); trace peaks at 212,
226.228,240,254,296

1-Nitropyrenc Ironwire 10(5), lo(lO), lwo), 10(100) GCYMS parent ion peak at 256 (trace) 0
(0.80) (3.6 cm x 2.0 mm)

Baiz.o&uinolinc Ironwire lo(5). 1WO). 1qw. lqloo) Benzo[Dquinoline (0.015 mg) 0

(1.432) (3.8 cm x 2.0 mm) (l.O5%);tmcepeakat~256

??
In all cases a solution of N+SO, (200 mg) in water (2.2 ml) was added to the soil, which was then
kt stand in air for 1 hr.
 85

Discussion

The soil used had a particle size of 0.74 mm or less [composition: sand (0.74 - 0.05 mm) 57.4%; silt (0.05 -
0.002 mm) 22.7%; clay (< 0.002 mm) 19.9%) and wss cleaned as described previously.’ The soil used had a
particle size of 0.74 mm or less [composition: sand [(0.74 - 0.05 mm) 57.4%. silt (0.05 -O/O02mm) 22.7%; clay
(<0.002 mm) 19.94/o].’ This included a final extraction with hot, pure methylene chloride to remove all extractable
materials. To get a baseline reading of what extractable materials would be obtained (owing to the expected
decomposition of any unextractable organic materials e.g. humus, bacteria, present in the soil before remediation)
from the bulk soil following its microwave treatment, samples of clean soil were irradiated for comparable periods
of time as used in the remediation processes, and then the soil and condenser+distillation arm were extracted
separately with pure CH& In no case was any material extracted from the soil itself. A largenumberof~
minor peaks were observed in the GCh4!S of the extractives from the condenser snd distillation arm when pencil
lead was used as the conductor of microwave energy (Table I). The odd numbered parent ion peaks are
undoubtedly nitrogen-containing molecules having one, or an odd number, of nitrogen atoms, and probably arise
from proteinaceous materials in the soil. When an iron wire (or nail) was used as the conductor, only three very
small off-gas peaks were observed. The soil was then spiked with individual PAH’s at the loading levels indicated
in Table 1, and an aliquot was extracted with CH&& to establish that the PAH’s could be recovered virtually
quantitatively from unremediated soil. Microwave irradiation was then carried out as per the schedule shown in
Table 1. In no case was any extractable product recovered from the soil after remediation! With benxoupyrene
(MW 252), most of the products recovered in trace amounts from the extraction of the condenser and distillation
arm matched those observed t?om the blank with two exceptions: M+, -_
m/z 380 and -_
m/z 394. These two peaks
were much too small (estimated to be well below the 0.5% level) to permit any characterixation of the materials.
When the irradiation time was reduced, no significant material was desorbed and collected. In the csse of
benzo~fluoranthene, and using iron wire as the conductor, 0.33% of the spiked PAH distilled over and was
collected. Traces of compounds with hIi ions at 276 (CdmO) and 322 (&&so1 ?) were observed, and while
structures can be written for these, they would be hi&ly speculative. Copper wire was used as the conductor in
the remediation of diben&hJanthmcene (MW 278); once again, nothing was extractable from the remediated soil,

but traces of two products @‘xx206 (Ct#tOO ?) and -_

m/z 234 (C17Ht40?)] were found in the extractives from the

condenser and distillation atm. When ber@Janthracene was the cant aminsnt 3.54% distilled unchanged and was
collected, together with trace amounts of compounds that had parent ions at p& 212,226.228, 240,254,2%. Of
these, the 212 and 240 peaks arose from the blank soil. The peak at 254 could correspond to C,&I220 for which
structure (1) is given as a possibility (among numerous others), and that at 2% could correapond to CIsHt604
(structure 2 could be a possible one). There were traces of a peak corresponding to M’ 256 in both of the
86

distillates from I-nitropyrene and benxo@quinoline, resulting most likely from soil background. Alternatively,
one could attribute it to a molecule C&,Hr& from I-nitropyrene, and to CrcHu, from benxo[flquinoline. In
addition, 1.05% of the benzo[flquinoline distilled unchanged.

As in the case with the PCB’s4 we made an attempt at uncovering the fate of the decomposed PAH’s by
using benxoupyrene-7-“C. As be.fore.4we used much narrower pencil lead in the hope that only partial
decomposition would occur and we would then be able to determine the structures of the residual material. Again,
this proved not to be the case. Up to 1.7% of the labelled be&aJpyrene was desorbed, while ca. 43% of it could
be extracted from the remediated soil (>75% of the radioactivity could be extracted from the soil before
remediation). A trace (0.3%) was accounted for in the form of COr, but up to 55% of the bemx$aJpyrene was no
longer extractable. Thus, it would appear that, as in the case of the PCB’s, whatever fragments are fotmed from
the PAH’s will probably be very tightly bound to the partially vitrified soil, which is reasonable since volatile
organic compounds such as cyclohexanol and cyclohexanom form unimolecular layers on Montmorillonite-Ca in
the cold and these are retained indefinitely under vacuum6 At higher temperatures some of the organic compounds
may be entrapped in the interliunellar regions of the clay,’ while at still higher temperatures either mineralixation
(as indicated by the small amount of 14C@evolved), oxidations, reductions, self-condensations and fragmentations
could lead to products which are tightly bound to the clay lamellas and further stabilized by encapsulation
attending partial vitrification, such that they cannot be treed by extended treatment with HCl or NaOH solutions.
Unfortunately, we did not possess the means of quantifying the unextractable residual 14Cleft in the soil.
Finally, we were able to show that a variety of organic solvents (see Table 3) could be desorbed, some
quantitatively, using this technique and an iron conductor, and collected in a condenser outside the microwave oven
Thus, these methods may be valuable leads in the development of jt~ gi& rented&m methods for soils
contaminated with PAH’s and organic solvents in those cases where the remediated soil will not be used to grow
plants, but on which buildings could be built. We are hoping to use the experience w here to develop much
milder microwave b&r remediation procedures that will not harm the soil and permit plant growth, perhaps a&r
see&g the remediated soil with its original bacterial flora, samples of which would have been taken prior to
remediation and grown --
in vitro for that purpose.
 87

Literrture Cited

(I) Abramovitch, R.A.; Huang BangZhou; Davis, M.; Peters, L.C. ChemosDhere in press.
(2) Mueller, J.G.; Lantz, SE.; Devereux. R.; Berg, J.D.; Pritchard, P.H. “Studies in the Microbial Ecology of
PAH Biodegradation,” i Bionmedhtioa of Cbloriarted and PolyeyclieAromatie liydrocarboo
Compouod~(Hinchee, R.E. &&., edts), Lewis Publishers. Boca Raton, FL, 1994, pp. 218-230.
(3) Meaghw-HartzelI, E.; Dial, C. Microfiche, 1994, EP 1.2: AS 717.
(4) Abramovitch, R.A.; Huang BangZhou; Abramovitch, D.A.; Song, Jiangao. Chemosphere in press.
(5) Chang, T.C.; Karr, C., Jr. Anal. Chim. Acta 1%9,2l_ 474.
(6) Glaeser, R Comet. Rend. Acad. Sci.. Ser. C. 1951,232, 1496.
(7) Kubicki, J. D.; Droh, M. J.; Schroeter, L. M.; Apitz, S. E. Environ. Sci. Technol. l!M7,3l_ 1151.