Sequential Anaerobic-AerobicTreatment of Soil Contaminatedwith Weathered Aroclor 1260E M M A R . M A S T E R , † V I V I A N W . - M . L A I , ‡

B I A N C A K U I P E R S , ‡ W I L L I A M R . C U L L E N , ‡

A N D W I L L I A M W . M O H N * , †

Department of Microbiology and Immunology,University of British Columbia, 300-6174 UniversityBoulevard, Vancouver, British Columbia, V6T 1Z3 Canada,and Environmental Chemistry Group, Department ofChemistry, University of British Columbia, 2036 Main Mall,Vancouver, British Columbia, V6T 1Z1 Canada

Soil contaminated with weathered Aroclor 1260 wasbioremediated by sequential anaerobic and aerobiclaboratory-scale treatment. The initial concentration was59 µg of PCBs/g of soil. Following 4 months of anaerobictreatment with an enrichment culture, all of the majorcomponents in Aroclor 1260 were completely or partiallytransformed to less chlorinated PCB congeners. The majorproducts of reductive dechlorination were 24-24-tetrachlorobiphenyl and 24-26-tetrachlorobiphenyl, andthe average chlorine substituents per PCB moleculedecreased from 6.4 to 5.2. The molar concentration ofPCBs did not decrease during the anaerobic treatment.All of the major products formed during the anaerobictreatment were degraded in the subsequent aerobic treatmentusing Burkholderia sp. strain LB400. After 28 days of theaerobic treatment, the concentration of PCBs was reducedto 20 µg/g of soil. PCBs were not significantly removedin aerobic treatments unless they were bioaugmented withLB400. Also, PCB degradation was not detected in soilbioaugmented with LB400 without prior anaerobic treatment.These results confirm the potential for extensive biologicaldestruction of highly chlorinated, weathered PCBcongeners in soil.

IntroductionDespite the extreme persistence of highly chlorinated PCBsin the environment, the potential exists for their extensivedegradation by naturally occurring microorganisms. Severalaerobic bacteria that degrade PCBs have been isolated(1-12). PCB degradation by these organisms proceedsthrough a series of enzymatic steps that result in the degra-dation of PCBs to chlorobenzoic acids and chloroaliphaticacids (13-17). Generally, these products are excreted by cellsand further transformed by other organisms. The initialenzymes used in aerobic PCB transformation are those ofthe biphenyl degradative pathway (8, 16), and PCB biodeg-radation is thought to occur fortuitously during biphenylmetabolism (1, 14, 17). Most of the PCB-degrading bacteriathat have been isolated transform PCBs that have up to four

chlorine substituents (18). However, a few bacterial isolates,including Burkholderia sp. strain LB400, can transformcertain PCBs with up to six chlorine substituents (5, 18).

PCB mixtures used for industrial purposes in NorthAmerica were mainly sold under the trade name Aroclor.Aroclor mixtures are designated by four digits; the first twoindicate that the mixture is composed of chlorinated bi-phenyls and the last two digits indicate the percent chlorineby weight. Industrial use of Aroclor 1260 has resulted incontamination of many soil and water sites with PCBscontaining more than seven chlorine substituents. Such PCBcongeners are very recalcitrant to aerobic microbial me-tabolism. Several researchers have demonstrated anaerobicmicrobial reductive dehalogenation of PCBs (19-24). Al-though a bacterium capable of dehalogenating hydroxylatedPCBs was recently isolated (25), isolation of bacteria capableof PCB dehalogenation has been unsuccessful to date.Consequently, various anaerobic sediments containing mixedpopulations of anaerobic microorganisms have been usedin experiments investigating PCB dehalogenation. It iscurrently believed that certain PCB congeners serve as cata-bolic electron acceptors under anaerobic conditions (26).Interestingly, different dechlorinating microbial consortiaappear to have different specificities for PCBs (19). Initialanaerobic treatment of soil contaminated with highlychlorinated PCBs then can potentially transform PCBsresistant to aerobic microbial attack to less chlorinatedcongeners that are susceptible to aerobic microbial miner-alization (1).

Sequential anaerobic-aerobic degradation of PCBs hasbeen previously demonstrated. Anid et al. (27) demonstratedsequential anaerobic-aerobic treatment of dehalogenatedPCB-contaminated soil spiked with Aroclor 1242. In theirexperiment, the soil was incubated anaerobically for 76 weeksand then aerobically for 96 days. Moreover, this report showedthat anaerobic soil can be aerated with H2O2 and that theaerobic step is enhanced by bioaugmentation. Shannon etal. (28) also reported sequential treatment of soil contami-nated with Aroclor 1242 and show profiles of PCBs remainingin a soil matrix after anaerobic and aerobic treatments.However, the details of their experimental method were notrevealed. Finally, Evans et al. (29) demonstrated treatmentof soil containing weathered PCBs, probably from Aroclor1248. In their experiment, the anaerobic treatment continuedfor 19 weeks, followed by a 19-week aerobic treatment thatincluded bioaugmentation with strain LB400. However, thisstudy showed that, when using LB400 as an aerobic inoculum,the initial anaerobic treatment did not contribute to totalPCB removal.

Here we show for the first time, a detailed quantitativeanalysis of sequential anaerobic-aerobic treatment of soilcontaminated with weathered Aroclor 1260. The soil usedwas obtained from Saglek, Laborador, Canada. We verifiedthe reproducibility of our method using several replicateexperiments. We report on the specificity of both theanaerobic and aerobic treatments and the kinetics of theaerobic step.

MethodsAnaerobic Enrichment Culture Preparation. A mixedanaerobic culture (BK81) was enriched from marine sedi-ments having a history of PCB contamination and collectedfrom Esquimalt Harbor, BC, Canada. Sources of contaminatedsoil and sediment and the procedure of enrichment werepreviously published (20). The BK81 enrichment culture usedin this study was transferred as described (20) four times.

* Corresponding author phone: (604)822-4285; fax: (604)822-6041;e-mail: wmohn@interchange.ubc.ca.

† Department of Microbiology and Immunology.‡ Environmental Chemistry Group, Department of Chemistry.

Environ. Sci. Technol. 2002, 36, 100-103

100 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 1, 2002 10.1021/es001930l CCC: $22.00 2002 American Chemical SocietyPublished on Web 11/28/2001

Preparation and Sampling of Anaerobic Treatments.Four slurries were prepared in 125-mL serum bottles bymixing 15 g of weathered PCB-contaminated soil from Saglekand 7.5 g of anaerobic pond sediment with 90 mL of mineralmedium as previously described (20) with the followingmodifications. The medium used in this experiment did notcontain resazurin or reductant. The medium was not spargedwith nitrogen or autoclaved. The medium was added tomixtures of soil and sediment in serum bottles under aheadspace of air. Serum bottles were crimp-sealed with butylrubber stoppers. The bottles were shaken for 2.5 h and thenincubated stationary at 21 °C in the dark. After 8 weeks ofincubation, oxygen was biologically removed from the bottlesand methanogenesis was detected. Methane was measuredby using a gas chromatograph equipped with a Haysep DB,100-120 mesh, 10 in. × 1/8 in. packed column and a thermalconductivity detector. The slurries were then autoclaved andinoculated (10%) with the enriched culture, BK81. 4-Bro-mobenzoate (200 µg/g of soil plus sediment) was added tothe sediment slurries, as this has been shown to stimulatedevelopment of PCB dechlorination activity (30). Serumbottles were crimp-sealed with Teflon-faced rubber septaand further stationary incubation was at 21 °C in the dark.

Samples of approximately 8 mL were removed from theslurries after inoculation and after 3 and 4 months ofincubation. Sampling was done inside an anaerobic chamber(N2-CO2-H2, 85:5:10), and samples were frozen in extractiontubes with Teflon-lined caps until extraction. Prior toextraction, the samples were thawed and centrifuged (2 minat 1500g), and the supernatant (medium) was removed anddiscarded. Preliminary tests showed that the supernatantdid not contain detectable PCBs.

Aerobic Culture Preparation. Burkholderia sp. strainLB400 (5) was grown at 21 °C to midexponential phase inmineral medium (18) supplemented with 200 mg/L biphenyl.At midexponential phase, cells were washed and concen-trated in mineral medium to 5 × 107 cells/mL.

Preparation and Sampling of Aerobic Treatments. Inpreliminary experiments, anaerobically treated soil waswashed to remove potentially inhibitory soluble compounds(e.g., sulfide), aerated, or washed and aerated before bio-augmentation with LB400 grown on biphenyl. The effect ofadding biphenyl to the aerobic treatments was also tested.These preliminary tests showed that washing the anaero-bically treated soil or adding biphenyl did not improve theaerobic treatment. However, the amount of biphenyl addedmight not have been significant compared to the amount ofbiphenyl that was carried over in the inoculum. As a result,for this experiment the anaerobically treated soil andsediment slurries were pooled in a 500 mL Erlenmeyer flaskand then aerated on an orbital shaker. Slurry aliquots of 15g were transferred into six 125-mL serum bottles, and then10 mL of mineral medium was added to each bottle. Sixadditional slurries were prepared in 125-mL serum bottlesby mixing 15 g of the above soil and 10 mL of mineral medium.Biphenyl (100 µg/g) was added to each treatment. LB400was added to three of the six anaerobically treated and threeof the six nonanaerobically treated slurries, to 2 × 106 cells/g. Serum bottles were crimp-sealed with Teflon-faced rubbersepta and incubated at 21 °C on an orbital shaker at 200 rpm.Following 0, 3, 7, 14, and 28 days of incubation, 2-g samplesfrom each aerobic treatment were transferred to Teflon-linedscrew capped glass tubes. Samples were stored at -20 °Cbefore extraction. Each sampling permitted equilibration ofculture headspace with air.

Extraction of PCBs From Soil. The slurries were extractedby shaking and vortexing soil twice with an equivalent volumeof acetone and then twice with an equivalent volume ofhexane. Tubes were centrifuged (2 min at 1500g) betweenextractions. The extracts were pooled and evaporated to

approximately 0.5 mL using nitrogen gas and then passedthrough a Pasteur pipet packed with hexane washed Florisiltopped with sodium sulfate. PCB congeners were eluted fromthe Florisil column using hexane, and eluates were collectedin glass chromatographic vials.

Analysis of PCBs. Samples were analyzed with a gaschromatograph fitted with a DB5-ms column (30 m × 0.25mm × 0.25 µm) and coupled to a mass spectrometer (VarianSaturn model 4D ion trap). The sample volume injected was2.5 µL, the temperature of the splitless injector was held at260 °C, and the temperature of the transfer line was 280 °C.The column temperature program used to analyze extractsfrom the anaerobic treatments was as follows: 104 °C for 3min, increased at 20 °C/min to 180 °C, increased at 2.5 °C/min to 272 °C, increased at 30 °C/min to 290 °C, and heldat 290 °C for 2 min. The column temperature program usedto analyze extracts from the aerobic treatments was asfollows: 104 °C for 3 min, increased at 20 °C/min to 160 °C,increased at 2.5 °C/min to 260 °C, increased at 50 °C/min to290 °C, and held at 290 °C for 3 min. The mass spectrum ofeach GC peak was used to determine if the peak correspondedto a PCB and, if so, to determine the number of chlorinesubstituents. Congener assignments were made by com-parison of the retention times with those in Aroclor 1221,1242, 1254, and 1260 (Accu Standard) and comparison of therelative retention times with published results (31, 32).Individual congener standards were used to quantify majorproduct congeners that are present in traces or not presentin the Aroclor standards. Linear three-point calibration curveswere generated for all congeners using either pure congenersor using the weight percent contributions of the componentspresent in Aroclor standards (31, 32). Amounts of PCBs (µg)per g of dry soil plus sediment were calculated, using as aninternal standard, 23456-2345-nonachlorobiphenyl, whichis present in the contaminated soil, to correct for variabilityin sample size and PCB recovery. Importantly, 23456-2345-nonachlorobiphenyl is not significantly degraded after 16weeks of anaerobic incubation (20), and aerobic biodegra-dation of this PCB congener has not been demonstrated.

Results and DiscussionAnaerobic Treatment of PCB-Contaminated Soil. All majorcomponents of weathered Aroclor 1260 initially present inSaglek soil were completely or partially removed after 4months of anaerobic incubation (Figure 1). However, ex-tensive PCB dechlorination was apparent after 3 months.Weathered Aroclor 1260 can have a higher proportion ofrecalcitrant, highly chlorinated PCB congeners than wheninitially synthesized (unpublished). Consequently, this dem-onstration of dechlorination of weathered Aroclor 1260 isencouraging for the application of bioremediation of soilwith a history of contamination with PCBs.

Dechlorination of PCBs during the anaerobic process wasindicated by significant decreases in the amounts of highlychlorinated PCBs and corresponding increases or appearanceof less chlorinated PCBs (Figure 1). PCB homologue classeswith 5-8 chlorine substituents were removed. The majorproducts of dechlorination were 24-24-tetrachlorobiphenyland 24-26-tetrachlorobiphenyl. Other less abundant prod-ucts were 23-2-trichlorobiphenyl or 26-4-trichlorobiphenyl(coeluting) and 235-2-tetrachlorobiphenyl or 24-25-tetra-chlorobiphenyl (coeluting). The pattern of dechlorinationobserved resembled Pattern N as described by Bedard andQuensen (19).

As expected, the number of moles of PCBs before andafter the anaerobic treatment did not change significantly(Table 1). This result confirms that the anaerobic processresults in PCB dechlorination but in little or no degradationof the biphenyl molecule (Table 1). Based on the reductionof chlorine substituents per biphenyl molecule, the weight

VOL. 36, NO. 1, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 101

of total PCBs should have been reduced by 11%. However,we did not detect significant differences in values for µg ofPCBs/g soil before and after the anaerobic step (Table 1).The inhomogeneity of the PCBs appears to have caused largevariability in their measurement, which is typical of envi-ronmental samples. Further, the anaerobic treatment mayhave increased the extraction efficiency of PCBs by alteringPCB-soil interactions and by converting the PCBs to less-chlorinated, less hydrophobic congeners. Despite this un-certainty about the total PCB concentration, the evidencefor their extensive dechlorination is very clear.

Aerobic Treatment of PCB-Contaminated Soil. In PCB-contaminated Saglek soil that was not treated anaerobically,90.2 ( 19.8% of Aroclor 1260 remained after the aerobictreatment. However, the amount of PCBs removed duringthis treatment was not significant (Students t-test, P < 0.1).This result was not surprising since the majority of PCBcongeners present in Saglek soil are substituted with 6-8chlorine atoms. Consequently, most of these congeners arerecalcitrant to biodegradation by LB400 (5, 18).

In aerobic treatments bioaugmented with LB400 andpreviously treated anaerobically, all major PCB congenersgenerated during the anaerobic stage were completely orpartially removed (Figure 1). In treatments bioaugmentedwith LB400, 67% of total PCBs was removed. Bioaugmentationwith LB400 of anaerobically treated soil was necessary toachieve significant aerobic PCB-degradation (Figure 2).Significantly, this result indicates that the products ofanaerobic PCB dechlorination are available for aerobic

microbial degradation and that the anaerobic treatment ofthe soil did not yield inhibitors of aerobic PCB removal. PCBcatabolism is also indicated by a decrease in the total molesof PCBs during the aerobic treatment (Table 1). The selectivityof the aerobic process for catabolism of less chlorinated PCBcongeners is indicated by an increase in the average numberof chlorine atoms per biphenyl molecule of PCBs (Table 1).

Notably, not all congeners known to be degraded by LB400were removed (5, 18, 33). Also, removal of certain PCBcongeners with seven chlorine substituents was detected.The range of PCBs degraded by pure cultures of LB400 mightdiffer from that of LB400 in soil. Also, detectable loss ofparticular heptachlorobiphenyls may be due to indigenousmicrobial activity or sampling variation. Certain PCB con-geners that were present in the original soil sample wereaerobically degraded only after the anaerobic treatment. Thisresult suggests that highly chlorinated PCBs may inhibit PCBbiodegradation. Alternatively, the anaerobic treatment mightincrease the bioavailability of weathered PCBs. Interestingly,the extent of PCB degradation in soil bioaugmented withstrain LB400 was similar at 7 days and 28 days (Figure 2), andnone of the congeners were completely removed by LB400

FIGURE 1. Sequential treatment of PCB contaminated soil. Histograms show the major congeners of weathered Aroclor 1260 (>0.5 µg/gof soil) identified by positions of chlorine substitution. Bars indicate standard error; n ) 3. Percent removal in each treatment phase isindicated for PCBs that were significantly degraded (Student’s t-test, P < 0.1).

TABLE 1. Summary of Treatment of PCB Contaminated Soila

treatmentnmol of

PCBs/g of soilµg of

PCB/g soilav Cl/PCBmolecule

initial soilb 157.4 (15.7) 58.7 (5.7) 6.4 (0.031)after anaerobic stepc 162.1 (45.4) 58.3 (14.4) 5.2 (0.068)after aerobic stepc 55.5 (10.5) 19.6 (3.6) 5.5 (0.12)

a Values are means (standard errors). b n ) 4. c n ) 3.

FIGURE 2. Total PCBs during aerobic treatment of anaerobicallytreated soil (n ) 3; bars indicate standard error): 9, aerobic treatmentwithout inoculation and [, aerobic treatment with inoculation withBurkholderia sp. strain LB400.

102 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 1, 2002

(Figure 1). The remaining PCBs may be strongly sorbed tothe soil matrix, and poorly available for microbial degradation.Also, metabolites resulting from aerobic PCB biodegradationmay have accumulated during the aerobic treatment, whichcan inhibit further PCB degradation by LB400 (14). Thus, itwould be interesting to determine if coinoculation of LB400and certain chlorobenzoic acid degraders would increasetotal aerobic PCB removal (34). Finally, we noted that thebiphenyl added to each treatment was completely consumedafter 7 days. LB400 removes highly chlorinated PCBs lessefficiently in the absence of biphenyl, than in the presenceof biphenyl (35). It is possible then that adding more biphenylto the soil treatments during incubation would increase theextent of PCB removal (36).

In conclusion, we show that sequential anaerobic-aerobictreatment of soil contaminated with highly chlorinated andweathered PCB congeners results in initial decrease in majorcomponents of Aroclor 1260, followed by aerobic degradationof the resulting less chlorinated PCBs. Overall, the concen-tration of PCBs was decreased by 67% (from 59 µg/g of soilto 20 µg/g of soil), the number of moles PCBs was decreasedby 65%, and the average number of chlorine atoms per PCBmolecule decreased from 6.4 to 5.5. This successful labora-tory-scale demonstration of treatment of weathered PCBsindicates great potential for PCB bioremediation. Ourmethods are currently being applied to design simple soilslurry bioreactors that can be used on-site. Future experi-ments will attempt to establish if total aerobic PCB removalis enhanced by coinoculation of additional PCB-degradingbacteria and chlorobenzoic acid degrading bacteria orrepeated addition of biphenyl during the aerobic treatment.

AcknowledgmentsThis research was supported by funding provided by theCanadian Department of National Defence, a CanadianNatural Science and Engineering Research Council (NSERC)Strategic Project Grant, and an NSERC graduate scholarshipto E.R.M.

Literature Cited(1) Abramowicz, D. A. Crit. Rev. Biotechnol. 1990, 10, 241-251.(2) Ahmed, M.; Focht, D. D. Can. J. Microbiol. 1973, 19, 47-52.(3) Ahmad, D.; Masse, R.; Sylvestre, M. Gene 1990, 86, 53-61.(4) Bedard, D. L.; Wagner, R. E.; Brennan, M. J.; Haberl, M. L.; Brown,

J. F., Jr. Appl. Environ. Microbiol. 1987, 53, 1094-1102.(5) Bopp, L. H. J. Ind. Microbiol. 1986, 1, 23-29.(6) Boyle, A. W.; Silvin, C. J. Hassett, J. P.; Nakas, J. P.; Tanenbaum,

S. W. Biodegradation 1992, 3, 285-298.(7) Clark, R. R.; Chian, E. S. K.; Griffin, R. A. Appl. Environ. Microbiol.

1979, 37, 680-685.

(8) Furukawa, K.; Miyazaki, T. J. Bacteriol. 1986, 166, 392-398.(9) Gibson, D. T.; Roberts, R. L.; Wells, M. C.; Kobal V. M. Biochem.

Biophys. Res. Commun. 1973, 50, 211-219.(10) Master, E. R.; Mohn W. W. Appl. Environ. Microbiol. 1998, 64,

4823-4829.(11) Peloquin, L.; Greer, C. W. Gene 1993, 125, 35-40(12) Seto, M.; Kimbara, K.; Shimura, M.; Hatta, T.; Fukuda, M.; Yano,

K. Appl. Environ. Microbiol. 1995, 61, 3353-3358(13) Bergeron, J.; Ahmad, D.; Barriault, D.; Larose, A.; Sylvestre, M.

Can. J. Microbiol. 1994, 40, 743-753.(14) Brenner, B.; Arensdorf, J. J.; Focht, D. D. Biodegradation 1994,

5, 359-377.(15) Furukawa, K.; Hayase, N.; Taira, K.; Tomizuka, N. J. Bacteriol.

1989, 171, 5467-5472.(16) Furukawa, K. Biodegradation 1994, 5, 289-300.(17) Hernandez, B. S., Arensdorf, J. J.; Focht, D. D. Biodegradation

1995, 6, 75-82.(18) Bedard, D. L.; Unterman, R.; Bopp, L. H.; Brennan, M. J.; Haberl,

M. L.; Johnson, C. Appl. Environ. Microbiol. 1986, 51, 761-768.(19) Bedard, D. L.; Quensen III, J. F. Microbial Trans. Degradation

Toxic Org. Chem. 1995, 127-216.(20) Kuipers, B.; Cullen, W. R.; Mohn W. W. Environ. Sci. Technol.

1999, 33, 3579-3585.(21) Quensen III, J. F.; Tiedje, J. M.; Boyd, S. A. Science 1988, 242,

752-754.(22) Williams, W. A. Chemosphere 1997, 34, 655-669.(23) Wu, Q.; Sowers, K. R.; May, H. D. Appl. Environ. Microbiol. 1998,

64, 1052-1058.(24) Wu, Q.; Sowers, K. R.; May, H. D. Appl. Environ. Microbiol. 2000,

66, 49-53.(25) Wiegel, J.; Zhang, X.; Wu, Q. Appl. Environ. Microbiol. 1999, 65,

2217-2221.(26) Mohn, W. W.; Tiedje, J. M. Microbiol. Rev. 1992, 56, 482-507.(27) Anid, P. J.; Ravest-Webster, B. P.; Vogel, T. M. Biodegradation

1993, 4, 241-248.(28) Shannon, M. J. R.; Rothmel, R.; Chunn, C. D.; Unterman, R. In

Bioremediation of chlorinated and polycyclic aromatic hydro-carbons; Hinchee, R. E., Leeson, A., Semprini, L., Ong, S. K.,Eds.; Butterworth-Heinemann: Boston, MA, 1994; pp 354-358.

(29) Evans, B. S.; Dudley, C. A.; Klasson, K. T. Appl. Biochem.Biotechnol. 1996, 57/58, 885-894.

(30) Bedard, D. L.; van Dort, H.; Deweerd, K. A. Appl. Environ.Microbiol. 1998, 54, 1786-1795.

(31) Frame, G. M. Fresenius J. Anal. Chem. 1997, 357, 701-713.(32) Williams, W. Personal communication to W. W. Mohn.(33) Haddock, J. D.; Horton, J. R.; Gibson, D. T. J. Bacteriol. 1995,

177, 20-26.(34) Focht, D. D. Curr. Opin. Biotechnol. 1995, 6, 341-346.(35) Mondello, F. J. J. Bacteriol. 1989, 171, 1725-1732.(36) Barriault, D.; Sylvestre, M. Can. J. Microbiol. 1993, 39, 594-602.

Received for review December 4, 2000. Revised manuscriptreceived October 3, 2001. Accepted October 17, 2001.

ES001930L

VOL. 36, NO. 1, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 103