Additive And Bio Remediation Of Soil

Colloids and Surfaces A: Physicochemical and Engineering Aspects 162 (2000) 1 – 14 www.elsevier.nl/locate/colsurfa

Effect of additives on biodegradation of PAH in soils T. Sobisch c,*, H. Heß a, H. Niebelschu¨tz b, U. Schmidt a a

AERES Angewandte Umweltforschung GmbH, Rudower Chaussee 29 (OWZ), 12484 Berlin, Germany b ARGUS Umweltbiotechnologie GmbH, Reuchlinstraße 11 – 13, 10553 Berlin, Germany c L.U.M. Gesellschaft fu¨r Labor-, Umweltdiagnostik & Medizintechnik mbH, Rudower Chaussee 5, 12489 Berlin, Germany Received 22 September 1997; accepted 27 April 1999

Abstract In a series of experiments with soil samples of contaminated sites we investigated the effect of special surfactant combinations and other additives on the kinetics and extent of solubilization and degradation of PAH. Biodegradation tests showed a positive effect of adding surfactant combinations T10 and T15. The effect was more pronounced when samples exhibited low degradation effectivity without surfactants. Addition of surfactants counterbalanced the dependence of bioavailability of the individual PAH compounds on their solubility in pure water. Changing the composition of surfactant combinations different tendencies were observed for the extent of solubilization and degradation, i.e. very high concentrations of PAH in the liquid phase may have toxic effects on the soil microflora. However, addition of a sorbent had the most remarkable effect on residual pyrene concentration in a sandy soil, whereby the contamination was transferred to the sorbent phase. Sorption to the sorptive phase and solubilization are substantially faster processes than degradation. © 2000 Elsevier Science B.V. All rights reserved. Keywords: PAH; Pyrene; Biodegradation; Solubilization; Soil clean-up; Surfactant combinations

1. Introduction Bioremediation, often a very economical option for soil decontamination, has restricted applicability for soils contaminated with polycyclic aromatic hydrocarbons (PAH’s). Biodegradation of PAH’s in soil – water systems of contaminated sites has been shown to be limited by certain factors. Addition of surfactants to * Corresponding author. Fax: +49-306-7198189. E-mail address: [email protected] (T. Sobisch)  This article was originally submitted to the IAP ’97 Special Issue.

improve bioavailability seems a viable option but has been conversely discussed in literature [1,2]. In previous investigations we identified special surfactant combinations which enhanced the biological activity (respiration rates) in tar–oil–water systems without preferential degradation of surfactants added [3]. To evaluate the applicability of these effects for remediation of contaminated soils and to support commercialization we investigated the effect of different surfactant combinations of this type on the kinetics and extent of solubilization and degradation in a series of experiments with soil samples of contaminated sites.

0927-7757/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 7 - 7 7 5 7 ( 9 9 ) 0 0 2 1 7 - 4

2

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

2. Experimental

2.1. Soil samples All soil samples, except sample A (fraction of soil fines from a soil washing plant), were homogenized to avoid ill reproducibility of the initial composition by passing through a 4 mm screen and by mixing. Sample A: fraction of soil fines (B 63 mm) containing cationic flocculants from a soil washing plant with an initial PAH concentration of 88 mg/kg (16 individual PAH compounds according to the EPA 610 priority pollutant list). Sample B: sandy soil of a former gaswork site with an initial PAH concentration of 646 mg/kg. Sample C: sandy soil of a former gaswork site with an initial PAH concentration of 1086 mg/kg, having a lower content of 5 and 6 ring compounds than sample B. Soil graines had a film of mineral oils (total petroleum hydrocarbons according to DEV H-18: 1000 mg/kg) on their surface. Sample D: subsample of C with an initial pyrene concentration of 417 mg/kg. After saponification (see below) an additional concentration of 3 mg/kg pyrene was determined, where from it is supposed that the fraction of PAH fixed in the organic soil matrix is low. Sample D was further characterized by wetsieving into size fractions 4 – 2 mm, 2 mm–63 mm, 63–20 mm after mechanical disintegration in a laboratory attrition cell (30 g soil, 30 g water, 10 min at 80 rpm). The dispersion of soil particlesB 20 mm was centrifuged at 3600 rpm for 5 min. The size fractions had pyrene concentrations and distribution summarized in Table 1. Table 1 Distribution of size fractions and pyrene contamination in sample D Size fraction

Wt%

Pyrene (mg/kg)

4–2 mm 2–0.063 mm 63–20 mm B20 mm

1.7 93.8 3.4 1.1

360 230 2020 1440

Calculating the overall pyrene concentration from the four fractions results in 306 mg/kg pyrene. The difference between this and the initial value seems to be due to the removal of PAH with the wash water, mainly along with dispersed oil (resulting from the initial oil film on the particles).

2.2. Soil microcosms Results of soil microcosm studies are averaged from two or three sets run in parallel. Deviations were normally in the range of 5%. Procedure A: The extent of the removal of PAH was investigated in soil suspensions (20 g soil–180 ml mineral salt solution). The mineral salt solution (pH: 6.8) was prepared by adding 1 g (NH4)2SO4, 0.3 g NaH2PO4, 0.5 g KH2PO4 and 0.05 g MgSO4 to 1 l of tapwater. The nonsterilized soil samples were amended with different surfactant combination. After shaking for 10 days at a frequency of 125/min the soil was removed by filtration and air dried. PAH’s were analyzed by HPLC according to EPA 610. Procedure B: The extent and kinetics of solubilization and removal of pyrene from sample D was investigated in soil slurries (10 or 20 ml solution/5 g soil). Nonsterilized soil samples were amended with nutrients (per g of soil: 2.38 mg (NH4)2SO4; 1.67 mg NaH2PO4; 0.425 mg KH2PO4; 1.09 mg K2HPO4) and different additives. Tap water was used. After shaking for a distinct period of time at a frequency of 150/min the soil was removed by centrifugation for 15 min at 4500 rpm, air dried, extracted with methanol in an ultrasonic bath at 40°C for 1 h. Extracts were centrifuged at 3500 rpm for 5 min. Optionally the extracted soil was rinsed with methanol, again separated by centrifugation and treated with 80 ml boiling aqueous methanolic KOH (2N) to have a measure for the fraction of pyrene not available by simple extraction. After slurry tests pyrene concentration in the liquid and solid phase was determined by UV-derivative spectrometry [4]. The extent of PAH sorption to the container walls was measured by equilibration with methanol after removal of soil and aqueous phase.

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

2.3. Additi6es Surfactant combinations: T10, T15, T16, T35 and T36 are each mixtures of a nonionic hydrophilic and a nonionic hydrophobic component. Combinations had low water solubility but could be easily dispersed in water. The components are biodegradable and are available at the market at reasonable prices. Different ratios between hydrophilic and hydrophobic compounds have been used, marked with 1,/2,/3 in order of increasing content of the hydrophobic compound, i.e. T10/1, T10/2 and T10/3. Other additives: polymeric compounds and sorbents (activated carbon, wood, S1) were also tested. All additives (incl. surfactant combinations) mentioned in this paper are available from L.U.M. GmbH, Berlin. 3. Results and discussion

3.1. Relation between reduction of PAH concentrations measured and biodegradation In this paper microcosm studies are only investigated by the variation of PAH concentrations in soil and solution. Besides biodegradation of PAH there are several biotic and abiotic factors which may contribute to the variations measured. In principle the following processes may play a role in the reduction of the initial PAH content, i.e. sorption to container walls, volatilization, photochemical degradation, oxidation and physical, chemical or biological bound residue formation. However, PAH initially not available for extraction may be released by physical, chemical and biological processes too. Sorption of PAH to container walls during the microcosm studies varied between 0.5 and 3% of the initial PAH content and was taken into account by a correction. Volatilization should only have a measurable effect for two and three ring PAH, but will be relatively small during 1 or 2 weeks. Variations caused by photochemical and oxidative degradation are expected to be of minor importance in case of the present investigations. However, ‘bound residue formation’ or even ‘bound residue destruction’ could interfere seri-

3

ously with experimental results. Only a fraction of the overall PAH content in contaminated soils is available by solvent extraction. This fraction can be increased by solvent extraction with additional saponification, however, the fraction of non-extractable PAH may amount up to more than 60% [5]. On the other hand, bound residues or non-extractable PAH are connected with soil organic matter and clays, but the content of organic matter and clays is very low for soil samples B, C and D. The low percentage of pyrene additionally extractable after saponification may indicate that the overall content of non-extractables is also low. In the case of soil fines (sample A) no distinction between ‘real’ biodegradation and bound residue formation can be made.

3.2. Effect of special surfactant combinations on slurry biodegradation tests with different contaminated soils Fig. 1 illustrates some typical results of biodegradation tests in suspension (procedure A). The variation of the content of selected PAH (phenanthrene as three ring, pyrene as four ring, benzo(a)pyrene as five ring, benzo(g,h,i)perylene as six ring compound and the total of the 16 EPA compounds) before and after biological action (10 days) demonstrates a positive effect of adding T10/1 (1%) to different soils (sample A: fraction of soil fines from a soil washing plant with relatively low contamination, sample B and C: both sandy soil of a former gaswork site, sample B with a higher content of five and six ring compounds). The soil microflora, when amended only with the mineral salt medium, exhibited no PAH-degrading activity in the case of soil fines (sample A), had only a moderate effect on the PAH content of sample B but reduced the amount of lower molecular weight PAH compounds substantially in the case of sample C. The addition of T10/1 to sample A, having an overall low bioavailability of PAH, caused a reduction of the 16 EPA compounds by 36%. The PAH content of sample B was reduced by 44% and of sample C by 90%. Addition of surfactants may be especially useful if the entire degradation effectivity of the soil

4

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Fig. 1. Effect of surfactant addition on slurry degradation tests (10 days).

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Fig. 2. Removal of individual PAH compounds as function of their original aqueous solubilities.

5

6

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Fig. 3. Effect of composition of surfactant combinations on solubilization and removal of pyrene (sample D, 5 g/20 ml, 2% surfactant per soil).

microflora is low, as in the case of sample A. As could be expected with sample C, in the presence of mineral oils with appropriate bioavailability, surface active bioproducts may be released (see below), which would have a positive effect on the physico-chemical and biological removal of PAH without addition of technical surfactants. Bioavailability of individual PAH compounds is related to their solubility in the aqueous phase. This trend is also evident for the degradation tests with no surfactant added (Fig. 2). The percentage of the individual compounds remaining in soil after treatment is depicted against their aqueous solubility (S). The literature data for the solubilities are compiled from refs. [6,7]. Unreasonably high values were excluded, which was also the case for the only available data for indeno(1,2,3-ed)pyrene (0.062 mg/l [7]). The addition of 1% T10/1 or T15/1 to the soil caused the fraction of the individual compounds remaining in soil to be nearly independent on their original aqueous solubilities for sample B. In the case of sample C with higher biological activity the dependence on the solubility values is more pronounced. The addition of T10/1 and T15/1 weakens this tendency.

The solubilizing action of the surfactant combinations used is not solubilization in terms of ‘preparation of thermodynamically stable isotropic solutions of substances otherwise only slightly soluble’ [8], i.e. micellar solubilization. Because of these surfactant combinations are only dispersed in the water phase, the PAH will partition between the soil and the dispersed surfactant phase. Therefore, the critical volume necessary for solubilization of PAH of high molecular weight [9] may not be a limiting factor, as in the case of micellar solubilization. It is obvious from Fig. 2 that addition of 1% T15/1 caused a reduced removal of PAH, whereas addition of 0.1% was almost as effective as the amendment with 1% T10/1.

3.3. Effect of surfactant composition on the extent of solubilization and degradation of pyrene Fig. 3 shows the influence of the composition of surfactant combinations on solubilization and degradation for sample D, a subsample of sample C with an initial pyrene concentration of 417 mg/kg. The residual pyrene concentration in soil and in the water phase was analyzed after one week shaking of the soil suspension (5 g/20 ml) amended

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

7

Fig. 4. Effect of surfactant concentration on solubilization and removal of pyrene (sample D, 7 days).

with nutrients and 2% surfactant per soil. Changing the quantity of the hydrophobic compound between 37 and 70% only a slight variation in residual and solubilized amounts of pyrene was observable. With T15, were only the hydrophobic component in T10 was changed (same hydrophilic compound), the solubilizing power decreased by more than 50%. Replacing the hydrophilic component in T10 and T15 by a slightly more hydrophilic surfactant, which differed only in the degree of ethoxylation (HLB: 0.5), solubilization was enhanced (T10“ T36) or slightly reduced (T15“ T35). On the other hand, if the hydrophilic compound in T10 was substituted by a surfactant of the same HLB value ( “ T16), the degree of solubilization was reduced by an order of magnitude. Therefrom it is evident that the structure of the hydrophilic compound has a marked influence on the solubilization effect of the formulation. Interestingly, the extent of biodegradation, deduced from the total of pyrene determined in soil and solution, is the highest for the surfactant combinations with low solubilizing power and vice versa. This may be caused by the very high concentration of PAH in the liquid phase, which may have had a toxic or inhibiting effect on the soil microflora.

During the slurry tests the fraction of pyrene extractable after additional saponification remained on the low initial level. Hence biological fixation of PAH might not have been occurred.

3.4. Influence of surfactant concentration on solubilization and residual concentration of pyrene The degree of solubilization and concentration of pyrene in soil (sample D) after one week shaking was measured as function of concentration of T15/2 (Fig. 4). As could be expected, the degree of solubilization is increased with increasing surfactant concentration. On the other hand the total residual pyrene content in the soil slurry (soil+ liquid phase) is lowest at a surfactant concentration of 0.5%. Lower surfactant concentrations (0.1–0.3%) exhibited a negative effect on residual concentration. Increasing the soil–water ratio to 5 g/10 ml the degree of degradation as well as the fraction of pyrene solubilized from soil is enhanced. Enhanced solubilization corresponds to higher ratios of PAH in the liquid phase per surfactant added at concentrations of 0.5% T15/2. This might be due to the formation of active bioproducts (biosurfactants or biopolymers) released by the soil microflora, maybe preferentially during degradation

8

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Fig. 5. Effect of different additives on solubilization and removal of pyrene (sample D, 5 g/10 ml, 7 days).

of the mineral oil contaminants. Release of such bioproducts is reflected by increased foaming, which had been observed after a distinct period of shaking. It seems that dilution of the soil microflora has an adverse effect on its overall activity, i.e. biodegradation and production of active metabolites. Another indication of the production of active metabolites is the enhanced solubility of pyrene in the soil slurries amended only with nutrients. In the case of 5 g soil/20 ml the pyrene concentration in the liquid phase reaches a value of 0.18 mg/l after one week, which is only slightly above the value for its solubility in pure water (0.13 mg/l), but a value of 3.1 mg/l for 5 g soil/10 ml. After 2 weeks the concentration of pyrene increases to a value of 0.9 mg/l (5 g soil/20 ml). Enhanced mutual attrition of soil grains may play also an important role in increasing the extent of solubilization if (technical or biological) surface active substances are present to stabilize mineral oils in suspension which will carry PAH into the liquid phase.

3.5. Screening of effecti6ity of different additi6es The performance of different surfactant combinations and other additives (polymeric compounds and sorbents) was further tested under standard conditions, i.e. shaking over a period of 1 week, 5 g soil (sample D)/10 ml. Selected results are compiled in Fig. 5. Compared with the sample not amended with additives (only with nutrients) addition of various surfactant combinations caused only a further reduction of the residual pyrene concentration in soil of about 10% (5 g/10 ml values). In the case of the formulation with high solubilization power T36/2 this could be reduced further by 10%, but degradation was completely inhibited. Addition of 0.1% of the polymeric compound Z18 reduced the amount of pyrene remaining in soil to 58% of the initial value. When 0.1% Z18 and 0.5% T15/2 were added simultaneously a reduction to 52% was reached. In these cases the relatively high concentrations of pyrene in the liquid phase did not inhibited degradation.

Fig. 6. Effect of different additives on the kinetics of pyrene removal (sample D, 5 g/10 ml).

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14 9

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Fig. 7A. Kinetics of pyrene removal from sample D (without additives).

10

Fig. 7B. Kinetics of pyrene removal from sample D (0.5% T15/2).

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

11

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Fig. 7C. Kinetics of pyrene removal from sample D (0.1% Z18).

12

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Fig. 7D. Kinetics of pyrene removal from sample D (5% SI + 0.1% Z18).

13

14

T. Sobisch et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 162 (2000) 1–14

Addition of 5% of the sorbent S1 reduced contamination of soil substantially. S1 was far more effective than other sorbents tested, i.e. activated carbon or wood. Without further additives present pyrene content of soil was lowered to 44%, when 41% of the initial amount of pyrene was concentrated in the sorbent phase. With additional 0.5% T15/2 or 0.1% Z18 these levels reached 36/44% (T15/2) and 34/43% (Z18).

3.6. Kinetics of pyrene remo6al and solubilization Fig. 6 shows the kinetics of pyrene removal from sample D. Figs. 7A, 7B, 7C and 7D summarize the variation of the pyrene fraction in soil and the liquid phase or bound to the sorbent. In the presence of 5% of sorbent S1 + 0.1% Z18 a fast removal of pyrene was observed. After a few days the residual pyrene concentration approached 34% of the initial value. Correspondingly, 43% of pyrene was concentrated in S1 (Fig. 7D). In all other cases the pyrene concentration reduced only slowly with only a moderate positive effect of adding 0.5% T15/2 or 0.1% Z18. The fraction of pyrene solubilized approaches a maximum during the first week of shaking, remaining nearly constant after a rapid decline. Sorption to the sorbent and solubilization are substantially faster processes than degradation. Degradation, measured as difference between initial values of pyrene and the total amount determined in the soil slurries, is small in all cases after 3–5 weeks.

4. Conclusions Biodegradation tests showed a positive effect of adding surfactant combinations T10 and T15. The effect was more pronounced when samples exhibited low degradation effectivity without surfactants. Addition of surfactants counterbalanced the dependence of bioavailability of the individual PAH compounds on their solubility in pure water. By variation of composition and concentration of surfactant combinations different tendencies were observed for the extent of solubilization and degradation, i.e. very high concentrations of PAH

in the liquid phase may have toxic effects on the soil microflora. Addition of the sorbent S1 had the most remarkable effect on residual pyrene concentration in a sandy soil, whereby the contamination was transferred to the sorbent phase. Sorption to the sorbent and solubilization are substantially faster processes than degradation. However, under no circumstances removal rates and residual concentrations were in the order of values which would be necessary for a competitive decontamination process. Therefore a new two stage bioreactor process was developed (patent pending, described elsewhere [10,11]), which enhances removal rates drastically. Acknowledgements We gratefully acknowledge the support of the Bundesministerium fu¨r Bildung, Forschung und Technologie under grant number FKV 0211001J6 and FKV 0150402J6. References [1] S. Laha, R.G. Luthy, Environ. Sci. Technol. 25 (1991) 1920. [2] B.N. Aronstein, M. Alexander, Appl. Microbiol. Biotechnol. 39 (1993) 386. [3] K. Ahrens, E. Winsel, T. Sobisch, Poster presented at Kontaktforum Biotechnologie, DECHEMA Jahrestagungen ‘95, Wiesbaden 1 June (1995). [4] T. Sobisch, Deutsche Bundesstiftung Umwelt, Sanierung von Bo¨den - Beispiele aus der Praxis, pp. 213 – 233. [5] A. Eschenbach, R. Wienberg, B. Mahro, Schriftenreihe Biologische Abwasserreinigung 7, Technische Universita¨t Berlin, 1996, p. 63. [6] US Environmental Protection Agency, Water related environmental fate of 129 priority pollutants, EPA 440/4-700296, (1979). [7] U. Starke, M. Herbert, G. Einsele, in: D. Rosenkranz, G. Einsele and H.-M. Harreß (Eds.), Handbuch BodenschutzBoS 9. Lfg. X/91, Erich Schmidt Verlag, 1991, p. 1680. [8] P.H. Elworthy, A.T. Florence, C.B. Macfarlane, Solubilization by surface-active agents, Chapman and Hall, London, 1968. [9] H.B. Klevens, J. Phys. Chem. 54 (1950) 283. [10] T. Sobisch, H. Niebelschu¨tz, TerraTech 5 (1998) 58. [11] T. Sobisch, H. Niebelschu¨tz, Contaminated Soil ‘98, Proceedings of the Sixth International FZK/TNO Conference on Contaminated Soil, 17 – 21 May, Edinburgh, Thomas Telford Publishing, Ltd., London, 1998, pp. 1189 –1190.