AbstractAs a means to remediate soil contaminated by polycyclicaromatic hydrocarbons, we investigated a combined pro-cess involving ethanol washing followed by a Fenton oxi-dation reaction. Artificial loamy soil was contaminated withvarious representative polycyclic aromatic hydrocarbons(i.e., fluorene, anthracene, pyrene, benzo(b)fluoranthene, orbenzo(a)pyrene) at concentrations ten times higher thanregulatory soil standards of The Netherlands or Canada,and then washed four times in ethanol, which reduced theconcentration of polycyclic aromatic hydrocarbon contam-ination to below the regulatory standard. Fenton oxidationof ethanol solutions containing anthracene, benzo(a)py-rene, pyrene, acenaphthylene, acenaphthene, benz(a)an-thracene, benzo(j)fluoranthene, or indeno(1,2,3-cd)pyreneshowed a removal efficiency of 73.3%–99.0%; by contrast,solutions containing naphthalene, fluorene, fluoranthene,phenanthrene, or benzo(b)fluoranthene showed a removalefficiency of 9.6%–27.6%. Since each of the nonremediatedpolycyclic aromatic hydrocarbons, excluding benzo(b)fluo-ranthene, are easily biodegradable, these results indicatethat the proposed treatment can be successfully applied topolycyclic aromatic hydrocarbon-contaminated soil thatdoes not contain high concentrations of benzo(b)fluo-ranthene. The main reaction products resulting from Fenton oxidation of ethanol solutions containinganthracene or benz(a)anthracene were anthraquinon orbenz(a)anthracene-7,12-dione, respectively; while 1,8-naphthalic anhydride was produced by solutions of ace-naphthylene and acenaphthene, and 9-fluorenone by a flu-orene solution.

Key words PAH-contaminated soil · Ethanol soil washing ·Fenton oxidation · Reaction product identification

Introduction

Polycyclic aromatic hydrocarbons (PAHs), which containmore than two benzene rings, are refractory organic com-pounds commonly produced by incomplete combustion of fossil fuels.1 Some PAHs with more than four benzene rings have accumulated in soil due to their strong microbialresistance.2

Clean-up processes for PAH-contaminated soil havebeen studied extensively over the last two decades, most ofthem being bioremediation processes based on microbialdegradation.3–9 Unfortunately, however, these processes are not only time-consuming, but are of limited value; theyhave almost no degradational effect on four-ring or greaterPAHs.2

With the aim of developing a means to accelerate PAH decomposition prior to microbial degradation, achemical–biological treatment has been applied, i.e., anadvanced oxidative process with Fenton’s reagent,10,11

although practical application is difficult owing to the highcosts of the chemicals required.

Soil washing has been applied to PAH- and heavy metal-contaminated soil because of its simplicity, low capital andoperating costs, easy maintenance requirements, and rela-tively good removal efficiency. On the other hand, furthertreatment of the washing solution (i.e., water, surfactant,and solvent) is needed; hence, this is only a partial remedy.

According to a U.S. EPA report, highly PAH-contaminated soil (30–1400mgPAH/kg) was washed andbiodegraded by the Biogenesis system.12 Compressed air-hot water (90°C) was used in this system, and a re-moval efficiency of 65%–73% was obtained. After aerobicbiodegradation of the washed water, the total PAH removalefficiency was 85%–88%. In spite of long-term biodegrada-tion (120 days), the treated soil could not meet the soil stan-dards of The Netherlands or Canada. Table 1 shows the soilstandards of The Netherlands and Canada.

In the process of Fenton treatment, a Fenton reactionproduces an OH radical (HO·) in such a way that oxidationof refractory organics and an organic radical occurs.13,14

J Mater Cycles Waste Manag (2000) 2:24–30 © Springer-Verlag 2000

Byung-Dae Lee · Masaaki Hosomi

Ethanol washing of PAH-contaminated soil and Fenton oxidation of washingsolution

Received: June 9, 1998 / Accepted: March 24, 1999

ORIGINAL ARTICLE

B.-D. Lee (*) · M. HosomiDepartment of Chemical Engineering, Tokyo University ofAgriculture and Technology, 2-24-16 Naka, Koganei, Tokyo 184-8588,JapanTel. +81-42-388-7070; Fax +81-42-381-4201e-mail: leebd@cc.tuat.ac.jp

25

With the above treatment history of PAH-contaminatedsoil in mind, we now describe a new, combined ethanol-washing/Fenton-oxidation process which is used for thispurpose. That is, we report on the applicability of ethanolwashing of artificial soil contaminated with represen-tative PAHs, and also on further degradation treatment ofPAH-containing washing solutions by a Fenton oxidationprocess. In addition, for various PAHs we identify the mainoxidation product generated in the Fenton reaction inethanol.

Methods

Materials

The following materials were used for PAHs: special grade naphthalene (NAP) (95%, Waco); acenaphthylene(ACEL) (98%, AccuStandard); acenaphthene (ACE)(97%, Aldrich); fluorene (FLU) (95%, Waco); fluoranthene(FLUT) (95%, Waco); 9-fluorenone (FLUO) (99%, Waco);phenanthrene (PHE) (98%, Aldrich); anthracene (ANT)(99%, Aldrich); anthraquinone (ANTQU) (97%, Aldrich);pyrene (PYN) (98%, Waco); benz(a)anthracene (BAA)(99%, Aldrich); benz(a)anthracene-7,12-dione (BAADI)(97%, Aldrich); benzo(b)fluoranthene (BBFT) (99%,Aldrich); benzo(j)fluoranthene (BJFT) (98%, AccuStan-dard); benzo(a)pyrene (BAP) (97%,Aldrich); indeno(1,2,3-cd)pyrene (INDE) (98%, AccuStandard); 1,8-naphthalicanhydride (NAPAN) (purity unknown, Aldrich). ForFenton oxidation reagents, the following were used: H2SO4

(95%, Waco); NaOH (97%, Waco); FeSO47H2O (99%,Waco); H2O2 (30%, Waco); ethanol (99.5%, Kokusan). Forthe GC–MS and HPLC analyses, dichloromethane (99.5%,Kokusan), acetonitrile (99.8%, Kokusan), and water (HPLCgrade) were used. Table 2 summarizes the chemical prop-erties of PAHs.15

As simulated soil, we used a loamy soil, namely akadama,which is representative of Japanese soil and has the follow-ing properties: mesh size No. 10–20; organic carbon content1.6%; base-exchange capacity 17meq/100g soil. Soil pH was6.5 and was measured by following the standard methodsfor the examination of soil: 25ml water or 1N of KCl solu-tion was added to 10g soil [i.e., 1 :2.5 (soil :water or KClsolution)], and then the pH of the water or KCl solution wasmeasured by a pH meter.16

Artificially contaminated soil

After being prepared in individual Teflon bottles, the loamysoil was artificially contaminated by FLUT (50mg/kg soil), ANT (50mg/kg soil), PYN (100mg/kg soil), BBFT (10mg/kg soil), or BAP (10mg/kg soil). The concentrationsof these contaminants were ten times higher than thoseallowed under the soil standards. Briefly, 6ml of ethylacetate (50mg FLUT, ANT, or PYN/l; or 10mg BBFT orBAP/l) was added to 6g of soil, after which the mixture wasagitated reciprocally without capping at 100r.p.m. for 24hto evaporate the ethyl acetate.

Ethanol washing of PAH-contaminated soil

All washing was done at 30°C. We first added 18ml freshethanol and washed and agitated (200r.p.m.) 6g of eachPAH-contaminated soil for 24h, and then removed 12ml of the supernatant. The remaining soil was washed again inthe same way after adding 12ml of fresh ethanol. This was repeated two more times (for a total of four washings).For similarly prepared controls, we substituted 12ml ofdichloromethane for ethanol. Each supernatant was ana-lyzed by HPLC to determine the PAH concentration.

Fenton oxidation of PAHs in ethanol

We performed Fenton oxidation of ethanol solutions con-taining NAP, ACEL, ACE, FLU, FLUT, PHE, ANT, BAA,PYN, BBFT, BJFT, BAP, or INDE. That is, a solution com-bining ethanol (10ml) with each prepared PAH (150mg/l)was added to teflon bottles and the pH was adjusted to 3.5by adding 1M H2SO4 or NaOH prior to adding 0.5M Fe2+

(2.7ml) and 30% H2O2 (4ml). After a variety of timeperiods, the reaction was terminated by adding four to eightdrops of concentrated H2SO4 to lower the pH to less than1.0. All of the Fenton oxidation experiments were con-ducted in the dark at 30°C.

Analysis

It is noted that the pH values cited in this paper were notexactly the same as those found in pure water because weused a normal pH meter (Yokogawa, model pH82) whileadjusting the pH in ethanol solution containing at least 15%or 30% water (i.e., 2.7ml of 0.5M Fe2+ and/or 4ml of 30%H2O2 solution added to 10ml of ethanol). After Fenton oxi-dation, samples were filtered through a 0.22-mm hydropho-bic membrane filter (Durapore, Millipore) and subjected to HPLC analysis. For the GC–MS analysis, to allow forliquid–liquid extraction (i.e., ethanol/dichloromethane), weadded 10ml of dichloromethane followed by reciprocal agitation (200r.p.m.) at 30°C for 1 day and filtration througha 0.22-mm hydrophilic membrane filter (Durapore,Millipore). Table 3 summarizes the GC–MS analysis conditions.

Table 1. Soil standards of The Netherlands and Canada for residentialareas

PAH Soil standards (mg/kg soil)

Fluoranthene (FLUT) 5a

Anthracene (ANT) 5a

Pyrene (PYN) 10a

Benzo(b)fluoranthene (BBFT) 1b

Benzo(a)pyrene (BAP) 1b

a The Netherlandsb Canada

26

Table 2. Chemical properties of PAHs

PAH Molecular Boiling Log Kow Vapor pressureweight point (°C) (mmHg)

Naphthalene (NAP)

128.1 218 3.36 2.30 ¥ 10-1

Acenaphthylene (ACEL)

154.2 279 3.92 1.55 ¥ 10-3

Acenaphthene (ACE)

152.2 280 4.05 NDa

Fluorene (FLU)

166.2 298 4.12 1.00 ¥ 10-3

Fluoranthene (FLUT)

202.3 384 5.22 1.00 ¥ 10-2

Phenanthrene (PHE)

178.2 340 4.52 6.80 ¥ 10-4

Anthracene (ANT)

178.2 400 4.45 1.95 ¥ 10-4

Pyrene (PYN)

202.3 393 5.18 6.85 ¥ 10-7

Benz(a)anthracene (BAA)

228.3 438 5.61 5.00 ¥ 10-9

Benzo(b)fluoranthene (BBFT)

252.3 480 6.35 ND

Benzo(j)fluoranthene (BJFT)

252.3 480 ND ND

Benzo(a)pyrene (BAP)

252.3 495 5.99 5.49 ¥ 10-9

Indeno(1,2,3-cd)pyrene (INDE)

276.3 ND 6.58 1.00 ¥ 10-10

ND: no data available

27

The concentrations of PAHs and their reaction productsafter Fenton oxidation were measured by isocratic HPLC(CCPE, Tosoh) with reverse phase separation (SupelcosilLC-PAH, 15cm ¥ 4.6mm i.d.) at 254nm and 25°C. Themobile phase was optimized at 80% acetonitrile and 20%HPLC-grade water. Fenton oxidation products were alsoconfirmed by GC–MS with electron impact and chemicalionization capability (5890-2, Hewlett Packard) and a dataacquisition system (Model 138,Wiley) equipped with a silicacapillary column (DB-5HT, 30m ¥ 0.25mm ¥ 0.1mm, J&WScientific). All analyses were performed in quintuplicate (n = 5). The data shown in the results indicate a 95% confi-dence interval. We determined the main reaction productsof PAH in ethanol produced by Fenton oxidation.

Results and discussion

Ethanol washing of PAH-contaminated soil

Figure 1 shows PAH concentrations after each ethanolwashing of FLUT-, ANT-, PYN-, BBFT-, and BAP-contaminated soil. Also shown are the corresponding stan-dards for residential area soil contamination in The Nether-lands (FLUT, 5mg/kg soil;ANT, 5mg/kg soil; PYN, 10mg/kgsoil) and Canada (BBFT, 1mg/kg soil; BAP, 1mg/kg soil).Note that after one washing the ANT concentration isbelow its standard, whereas four washings are required for PYN-, BBFT-, and BAP-contaminated soil. Also, for allcases, 73%–91% of PAHs are removed by the first wash-

ing. As a result, the regression curves between ln C/C0 and soil washing times followed pseudo-first-order. Theseresults suggested that the removal efficiency in eachwashing time is proportional to the remaining PAH con-centration in the soil. After each successive washing the concentrations are remediated with >90% efficiency to below their respective soil standards. In comparison,dichloromethane washing in the control experimentsshowed at least 88% recovery for the same PAH-contaminated soils (data not shown).

Fenton oxidation of PAHs in ethanol

The difference in boiling points between PAHs and ethanolallows concentration of PAHs. The initial concentration ofPAHs at 150mg/l was set relatively higher than that in thewashing experiment for better identification of the reactionproducts.

The Fenton reaction depends greatly on pH, with a pHrange from 3 to 4 providing the optimal condition fordecomposition of most organic compounds. As preliminaryexperiments using a BAP–ethanol solution showed theoptimal removal efficiency at pH = 3.5 (data not shown), weused this pH level. Figure 2 shows decomposition profilesof PAHs in ethanol after adjusting the Fenton oxidationreaction to pH = 3.5, where most PAHs show more than70% removal efficiency, i.e.,ACE,ACEL,ANT, PYN, BAA,BJFT, BAP, and INDE. This degradation, which was identi-fied in all cases, is followed by a pseudo-first-order reaction

Fig. 1. Effect ofethanol washingtimes on PAH-contaminatedsoil

Table 3. GC–MS analytical conditions

GC Injection mode SplitlessInjection volume 1.0mlInjection temperature 250°CColumn temperature 40°C (5min hold) Æ 10°C/min Æ 370°C (5min hold) Æ endCarrier gas/pressure He/8.0psi

MS Detector temperature 280°CIonization method Electron impact

28

and is easily degraded by Fenton oxidation. Of these PAHs,BAP’s degradation rate after 3h gave the largest rate coef-ficient of 0.397min-1 (Table 4).

The degradation of FUL and FLUT, on the other hand,shows a relatively low removal efficiency of 27% after 60min, which is consistent with previous results,7 whileNAP, PHE, and BBFT show almost no degradation eventhough PHE and BBFT have identical molecular weights

and chemical properties very similar to those of ANT andBJFT, respectively. The slight differences in chemical struc-tures are thought to be responsible for variations in degra-dation by Fenton oxidation in ethanol, although the reasonfor these differences among removal efficiencies remainsunclear and is left for further study.

Fortunately, NAP, FLU, FLUT, and PHE biodegrade rel-atively easily in comparison with other PAHs containing

Fig. 2. Effect of Fenton oxidation on the remediation of some PAHs contained in the ethanol washing solution. “Reg.” indicates the calculatedregression line

29

more than four benzene rings, with successful bioremedia-tion of soil contaminated with these PAHs having been well described.2 Consequently, for treating NAP-, FLU-,FLUT-, or PHE-contaminated soils, an ethanol-washing/bioremediation process is more effective than the ethanol-washing/Fenton-oxidation process presented.

Most importantly, those PAHs containing more than fourbenzene rings, excluding BBFT, which is known to be nonbiodegradable, were effectively degraded within a 3-hreaction period. Table 4 summarizes the ratios of recoveredreaction products and decreased PAHs, the final removalefficiency after a 3-h reaction period, and the respectivereaction coefficients and main reactions products produced.Note that 82% of ANT was converted into ANTQU, whichwas the largest conversion percentage of three quantifica-tions carried out. Also, in the reaction products produced,FLUO and ANTQU were detected.These are the metaboliccompounds of biodegradation of FLU and ANT, respec-tively,7,9 which indicates that Fenton oxidation improves thebiodegradability of FLU and ANT.

Regarding the percentage of Fenton-oxidized PAHs con-verted into the main reaction product identified for FLU

and BAA, it was realized that their conversion into FLUOand BAADI was relatively low.The conversion percentagesof other PAHs into the main reaction products identifiedwere not determined because no distinct peaks (exceptethanol reaction products) appeared on GC–MS and HPLC chromatographs. Although NAPAN was identifiedby GC–MS and HPLC analysis to be the main reactionproduct produced by ACE and ACEL, we could not quan-tify the conversion percentage because the purity of a stan-dard NAPAN solution was unknown.

Conclusions

Our main results are listed below.

1. One ethanol washing effectively remediated anthracene-contaminated soil (50mg/kg soil) to below the soil stan-dards of The Netherlands, while four successive washingswere required to do the same for soil contaminated withfluorene (50mg/kg soil) or pyrene (100mg/kg soil) [The

Table 4. Summary of Fenton oxidation results of PAHs in ethanol washing solutiona

PAH Reaction product Removal Recovered reaction Reaction Rate constantefficiency products/decrease- order (day-1)(%) PAHs (%)

Naphthalene (NAP) NIb 9.6 –c – –

Acenaphthylene (ACEL) 1,8-naphthalic anhydride 99.0 – Pseudo first 0.216

(NAPAN)Acenaphthene (ACE) 99.0 – Pseudo first 0.060

Fluorene (FLU) 9-fluorenone (FLUO) 27.6 13.1 – –

Fluoranthene (FLUT) NI 26.7 – – –Phenanthrene (PHE) NI 21.3 – – –Anthracene (ANT) Anthraquinon (ANTQU) 99.0 82.0 Pseudo first 0.267

Pyrene (PYN) NI 99.0 – Pseudo first 0.039Benz(a)anthracene (BAA) Benz(a)anthracene-7,12-dione 74.3 44.3 Pseudo first 0.022

(BAADI)

Benzo(b)fluoranthene (BBFT) NI 14.6 – – –Benzo(j)fluoranthene (BJFT) NI 73.3 – Pseudo first 0.005Benzo(a)pyrene (BAP) NI 99.0 – Pseudo first 0.397Indeno(1,2,3-cd)pyrene (INDE) NI 99.0 – Pseudo first 0.209a Based on the reaction period in Fig. 3b NI: No reaction product was identified by GC–MSc – indicates no data obtained

O

O

O

O

O

O OO

30

Netherlands], or benzo(b)fluoranthene (10mg/kg soil) orbenzo(a)pyrene (10mg/kg soil) [Canada].

2. Fenton oxidation of ethanol solutions containinganthracene, benzo(a)pyrene, pyrene, acenaphthylene,acenaphthene, benz(a)anthracene, benzo(j)fluoran-thene, or indeno(1,2,3-cd)pyrene showed a high removalefficiency over a 3-h reaction period; by contrast, thosecontaining naphthalene, fluorene, fluoranthene, phenan-threne, or benzo(b)fluoranthene showed a low removalefficiency over a long reaction period. Since each of thenonremediated PAHs, excluding benzo(b)fluoranthene,are easily biodegradable, these results indicate that theproposed treatment can be successfully applied to PAH-contaminated soil that does not contain high concentra-tions of benzo(b)fluoranthene. This case requires furtherstudy.

3. The main reaction products produced by Fenton oxidation of ethanol solutions containing anthracene or benz(a)anthracene were anthraquinon orbenz(a)anthracene-7,12-dione, respectively, while 1,8-naphthalic anhydride was the product produced by solutions of acenaphthylene and acenaphthene, and 9-fluorenone by a fluorene solution. It was of particularinterest that the anthracene-contaminated solutiondegraded into 82% anthraquinon.

Our final goal is to develop a clean-up process for PAH-contaminated soil. A future research need identifiedin this study is to degrade the reaction products of Fentonoxidation completely. For this subject, we need to conductfurther studies on ethanol washing with distillation, and theaerobic or anaerobic biodegradability of Fenton oxidationproducts.

References

1. Warman K (1985) PAH emissions from coal-fired plants. In: Hand-book of polycyclic aromatic hydrocarbons. vol 2. Emission sources

and recent processes in analytical chemistry. Marcel Dekker, NewYork, pp 21–59

2. Cookson JT (1995) Bioremediation engineering: design and appli-cation. McGraw-Hill, New York, pp 114–120

3. Barnaley EA (1975) The bacterial degradation of fluoranthene andbenzo(a)pyrene. Can J Microbiol 21:1004–1008

4. Buswell JA, Eriksson KE (1979) Aromatic ring cleavage by thewhite rot fungus Sporotrichum pulverulentum. FEBS Lett 104:258–260

5. Haemmerli SD, Leisola MSA, Sanglard D, Fiechter A (1986) Oxi-dation of benzo(a)pyrene by extracellular ligninases of phane-rochaete chrysosporium. J Biol Chem 261:6900–6903

6. Heitkamp MA, Franklin W, Cerniglia CE (1988) Microbial metab-olism of polycyclic aromatic hydrocarbons: isolation and charac-terization of pyrene-degrading bacterium. Appl Environ Microbiol54:2549–2555

7. Hammel KE, Green B, Gai WZ (1991) Ring fission of anthraceneby a eukaryote. Proc Natl Acad Sci USA 88:10605–10608

8. Field JA, de Jong E, Costa GF, de Bont JAM (1992) Biodegrada-tion of polycyclic aromatic hydrocarbons by new isolates of whiterot fungi. Appl Environ Microbiol 58:2219–2226

9. Selifonov SA, Grifoll M, Gurst JE, Chapman PJ (1993) Isolationand characterization of (+)-1,1a-dihydroxy-1-hydrofluoren-9-oneformed by angular dioxygenation in the bacterial catabolism of fluorene. Biochem Biophys Res Commun 193:67–76

10. Martens DA, Frankenberger JR (1995) Enhanced degradation ofpolycyclic aromatic hydrocarbons in soil treated with an advancedoxidative process–Fenton’s reagent. J Soil Contam 4:175–190

11. Gauger WK, Srivastava VJ, Hayes TD (1991) Enhanced biodegra-dation of polyaromatic hydrocarbons in manufactured gas plantwastes. Gas, Oil, Coal, Environ Biotechnol 3:75–92

12. US EPA (1993) Biogenesis soil washing technology: innovativetechnology evaluation report. EPA/540/R-93/510

13. Walling C (1975) Fenton’s reagent revisited. Acc Chem Res8:125–131

14. Mertz JH, Waters WA (1949) Some oxidation involving the freeradical hydroxyl. J Chem Soc 515–525

15. WebKis-Plus Chemical Database (1998) National Institute forEnvironmental Studies (Japan)

16. Japan Soil Fertilizer Science Association (1992) Standard methodsfor the examination of soil