Transformation of dissolved organic matter (DOM) and 14C-labelled organic contaminants during composting of municipal biowaste

The Science of the Total Environment 278 Ž2001. 1᎐10

Transformation of dissolved organic matter ž DOM/ and 14 C-labelled organic contaminants during composting of municipal biowaste Nicola Hartlieb a,U , Bernd Marschner b, Werner Klein a a

Fraunhofer-Institut fur ¨ Umweltchemie und Oekotoxikologie, Auf dem Aberg 1, D-57392 Schmallenberg, Germany Ruhr-Uni¨ ersitaet Bochum, Geographisches Institut Bodenkunder Bodenoekologie, D-44780 Bochum, Germany

b

Received 17 April 2001; accepted 22 May 2001

Abstract Composting of municipal biowaste in the presence of 14 C-labelled organic contaminants was studied in an attempt to characterize the mobilization potential of dissolved organic matter ŽDOM. for hydrophobic contaminants. The properties and transformation of DOM extracted from municipal biowaste compost with 10 mM KCl at six stages during 370 days of composting were investigated. DOM was fractionated into molecular weight fractions by ultrafiltration, and DOM structure was studied using CPMAS 13 C-NMR- and UV-spectroscopy. The distribution of 14 C-labelled model substances ŽDEHP, pyrene, simazine. upon molecular weight fractions was investigated by ultrafiltration, and association to DOM was studied performing flocculation experiments. The binding capacity of DOM for the model substances was of secondary influence for the mobilization because the intense biochemical reactions during composting pre-dominated the fate of the substances. Composting favoured the degradation of model substances to polar metabolites and supported their binding to the DOM matrix. DEHP and simazine were mainly found in the low- to medium-molecular DOM fraction and showed a small amount of DOM-associated radioactivity Žapprox. 10%.. Pyrene and its metabolites had high affinities to high-molecular DOM. However, a direct relationship between DOM-quality and enhancement of pyrene solubility was not visible. After 120 days of composting DOM showed the highest binding capacity for hydrophobic contaminants. 䊚 2001 Elsevier Science B.V. All rights reserved. Keywords: Compost; Dissolved organic matter; Spectroscopy; Radio-labelled contaminants; Mobilization; Ultrafiltration; Flocculation; Degradation; Transformation

U

Corresponding author. Gluckstr. 7, D-76185 Karlsruhe, Germany. E-mail address: [email protected] ŽN. Hartlieb..

0048-9697r01r$ - see front matter 䊚 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 8 - 9 6 9 7 Ž 0 1 . 0 0 9 0 2 - 0

2

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

1. Introduction The importance of dissolved organic matter ŽDOM. in determining the fate of hydrophobic organic contaminants in the environment has already been shown ŽChiou et al., 1986; McCarthy and Zachara, 1989; Kukkonen et al., 1990; Magee et al., 1991.. DOM, defined as the fraction of organic matter that can pass through a 0.45-␮m filter membrane, is a heterogeneous mixture of compounds with different molecular weight, from polysaccharide fragments to high molecular weight colloids derived from humic substances ŽBuffle and Leppard, 1995.. The solubility of hydrophobic contaminants is enhanced by sorption to DOM, which in turn leads to an increased mobilization due to co-transport with the carrier agent DOM. Compost with its high organic matter content releases compost leachate with high concentrations of DOM. Furthermore, compost leachate shows a higher elution capacity for hydrophobic contaminants such as PAHs compared to other elution agents Žcompost) sewage sludge ) waste disposal leachatersewerage ) rainwater. with more than 70% of PAHs being associated with DOM ŽBusche and Hirner, 1997; Raber and Kogel-Knabner, 1997.. Therefore, compost appli¨ cation implies the potential of mobilising hydrophobic organic contaminants in compost leachate with DOM as a carrier agent. Composting is a well-known system for rapid organic matter stabilization and humification ŽAdani et al., 1995; Chen et al., 1996.. During composting, readily degradable organic matter is used by microorganisms as C and N sources. The end product Žcompost. consists of transformed and slowly degradable compounds, intermediate breakdown products and dead and living microorganisms. Compost has the potential of improving soil fertility and structure ŽChen et al., 1989; Delschen, 1996.. In order to use the advantages of compost application, the hazard potential of contaminant mobilization has to be investigated. The DOM enhanced solubility of hydrophobic contaminants is attributed to shifting the partitioning equilibrium towards the dissolved phase due to the presence of soluble sorbent agents. However, the affinity of hydrophobic contami-

nants for DOM and, therefore, the mobilization potential is dependant on DOM quality. With increasing molecular weight of DOM and increasing amount of hydrophobic binding regions the binding capacity for hydrophobic contaminants increases ŽChiou et al., 1986; McCarthy et al., 1989; Kile and Chiou, 1989; Kukkonen et al., 1990; Raber and Kogel-Knabner, 1997; ¨ Marschner, 1998.. Since degradation and transformation reactions in the course of composting affect the organic matter quality, the DOM binding capacity is subject to changes. Also, transformation reactions affect organic contaminants present in biowaste, therefore, enhanced solubility of hydrophobic contaminants might not only be attributed to the association to DOM but also to the transformation into more polar metabolites during composting. Thus, the aim of the project was to investigate the mobilization of hydrophobic contaminants during composting and to assess which parameters Žtransformation of contaminants or DOM quality. affect their solubility. At present, it is unknown how composting of biowaste will change compost DOM concentration and quality and, in turn, affect the distribution and mobility of hydrophobic organic contaminants. To study the mobilization potential of DOM during different stages of composting and to examine the transformation of organic contaminants, composting was investigated over a period of 1 year using 14 Clabelled model substances ŽDEHP, pyrene, simazine. as tracers for hydrophobic contaminants.

2. Experimental section 2.1. Materials Municipal biowaste was thoroughly mixed with shredded shrubbery as a bulking agent Ž30% vol.. and supplied by the Stratmann composting plant ŽBrilon, Germany.. Properties: organic C content, 28.4% Žmass.; total N content, 1.16%; CrN ratio, 25; pH 6.8; water content, 55%; and loss of ignition, 58.8%. The following model substances have been used: Di-Ž2-ethylhexyl.Žcarboxyl- 14 C.phthalate

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

ŽDEHP., Amersham Life Science, ) 98% purity, 1.96 GBq mmoly1 ; pyrene-Ž4,5,9,10- 14 C. Amersham Life Science, ) 98% purity, 2.04 GBq mmoly1 ; and simazine-R-UL- 14 C Sigma Aldrich, ) 95% purity, 1.65 GBq mmoly1 . 2.2. Spiking procedure The application of each of the model substances Ž 14 C-DEHP, 14 C-pyrene, 14 C-simazine. to the biowaste was carried out individually prior to composting. Initially applied concentrations A 0 : 2.35 mg DEHP Žkg dry mass.y1 s 11.75 MBq DEHP Žkg dry mass.y1 ; 0.56 mg pyrene Žkg dry mass.y1 s 5.6 MBq pyrene Žkg dry mass.y1 ; and 1.17 mg simazine Žkg dry mass.y1 s 9.5 MBq simazine Žkg dry mass.y1 . The substances were spiked in two steps. Initially 200 ml of acetone stock solution of each model substance were added to 520 g quartz sand Ž0.2᎐0.8-mm mesh size, Merck. under continuous mixing. The organic solvent was allowed to slowly evaporate from the sand for 1 h under continuous mixing in an evaporator. For each substance 10 aliquots of the spiked sand were analysed to check the uniformity of the tracer distribution. The variation coefficient of the total 14 C-activity was found to be - 7%. In a second step the quartz sand was successively mixed with 30-kg dry mass of biowaste in small portions and, subsequently, transferred to the bioreactor and to the composting vessels, respectively.

3

temperatures approximately 70⬚C; day 29᎐78 cooling phase at temperatures of approximately 30⬚C; day 78᎐370 maturation phase at 30⬚C. The compost in each bioreactor segment was completely mixed every week during the cooling phase and every second week during the maturation phase. The bioreactor was kept shut during the thermophilic phase. During the maturation phase the compost moisture content was readjusted to 40% by adding water every 2 weeks. Compost material was sampled after sieving Ž- 1 cm. at day 0, 29, 78, 120, 200 and 370. Samples were air-dried until constant weight was obtained. The residual water content was determined by drying at 105⬚C for 24 h. 2.4. DOM elution DOM was eluted from compost by rotationally shaking 100 g of air-dried compost with 1 l of 10 mM KCl solution for 16 h at 8 rpm. After centrifugation at 13 000 rpm for 30 min ŽSorvall Superspeed Angle Rotors GSA. the supernatant was filtered through a - 0.45-␮m glass fibre filter ŽMN GF-5, Macherey-Nagel.. In the filtrates the DOC concentration ŽTotal Carbon Analyzer, TOC-5000r5050, Shimadzu., the pH and the conductivity was determined. The 14 C-activity was measured by liquid ␤-scintillation counting ŽLSC.. LSC was performed in duplicates, and the variation coefficient was less than 5%. Pico Fluor 40 ŽCanberra-Packard Company, Frankfurt, Germany. was used as the scintillation cocktail.

2.3. Composting procedure 2.5. Ultrafiltration Composting was performed in a stainless steel bioreactor Žvolumes 1.8 m3, diameter s 1.5 m., which was aerated from the bottom to the top and had an internal as well as external temperature control. The online detected mass of the bioreactor was recorded daily and the exhaust air composition ŽCO 2 , O 2 , N2 , CH 4 . was analyzed by gas chromatography every 6 h to ensure aerobic conditions. The segmented interior enabled separate composting for each model substance and for two control segments without substance addition. The following three composting stages were simulated: day 0᎐29 thermophilic phase reaching

Separation of DOM in different molecular weight fractions was carried out in pressure filtration cells ŽAmicon 8200 stirred cell. by means of flat membrane filters type NOVA Žmodified PES, PALL Filtron.. The following filtration steps were implemented in succession ŽMWCO s molecular weight cut-off value; kDas kilo dalton.: - 100 kDa MWCO, - 10 kDa MWCO, and - 1 kDa MWCO. The ultrafiltration was carried out in duplicate. The pH and conductivity of DOM solutions varied to a small extent only throughout the composting time ŽpH 7᎐8, con-

4

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

ductivity s 3.7᎐4.6 mS cmy1 ., and were not adjusted prior to ultrafiltration since the binding capacity of DOM is only slightly affected by these parameters ŽRaber et al., 1998.. For each ultrafiltration step, approximately 90% of the solutions were allowed to pass the membrane; the remaining retentate was discarded in order to prevent clogging of the membrane pores. In each ultrafiltrate the concentration of DOC and 14 C-activity was determined Žsee above., and the DOM was characterised by spectroscopic analyses Žsee below.. The amount of DOC and 14 C-activity in each ultrafiltration ŽUF. fraction was calculated by subtracting amounts in low molecular weight fractions from amounts in higher ones: y UF MWCO 100 kDa s fraction DOM solution UF MWCO 100 kDay UF MWCO 10 kDa s fraction UF MWCO 10 kDay UF MWCO 1 kDa s fraction UF MWCO 1 kDa s fraction

) 100 kDa 10᎐100 kDa 1᎐10 kDa - 1 kDa

Preliminary tests have shown that filters, even if they are hydrophilic, sorb hydrophobic contaminants to a large extent. This leads to a falsification of results regarding the distribution between molecular weight fractions. Hydrophobic substances can only pass a filter membrane if they are entrapped in the DOM matrix preventing exchange reactions with the filter or if they are transformed into polar metabolites. In order to elucidate sorption of 14 C-activity to membrane filters a DOM test solution containing a defined concentration of 14 C-pyrene and 14 C-DEHP, and no metabolites was ultrafiltrated in comparison to the DOM solution from the composting experiment containing an unknown amount of metabolites. The test solution was prepared by adding the two model substances separately to DOM eluted from 370-days composted material. Test solutions were ultrafiltrated as triplicates and 14 C-activity was measured in the filtrates as well as the retentates. Retentates were gained by thoroughly rinsing the filter membranes with distilled water. Volumes of filtrates and retentates were determined. The sorbed fraction was calculated by subtracting 14 C-activity in the filtrate and retentate from the 14 C-activity in the test solution.

2.6. Spectroscopic analyses Spectrophotometric analyses of fractionated DOM solutions were performed on a UV-Vis spectrometer Cary-Bio1 ŽVarian.. Samples were placed in a 1-cm quartz window cuvette and scanned for absorbance values at 220, 254, 270 and 330 nm. The ratio of the absorbance at a particular wavelength to the DOC concentration is an estimate of the aromatic nature of the DOM termed SA ␭ , the specific absorbance at wavelength ␭ ŽStevenson, 1994; Chin et al., 1994; Martius et al., 1996.. SA is expressed as per meter Žm. absorbance divided by the DOC in g Crl Žl gy1 my1 .. Solid-state 13 C-NMR spectra were recorded on a Bruker MSL-100 spectrometer at a 13 C-resonance frequency of 25 MHz. The acquisition parameters of the standard CPMAS experiment were the following: 1.2-ms contact time; 1-s recycle delay time; 125-kHz sweep width and 20-Hz line broadening. Freeze-dried samples were placed into 7-mm zirconium dioxide rotors and spun at a frequency of 4.5 kHz. Various structural groups were quantified by integrating the signal peaks in the given chemical shift regions: alkyl-C, 0᎐50 ppm; O-alkyl-C, 50᎐110 ppm; aromatic-C, 110᎐160 ppm; carboxyl-C, 160᎐190 ppm; and carbonyl-C, 190᎐220 ppm. 2.7. Flocculation experiments Flocculation experiments were conducted according to Laor and Rebhun Ž1997. in order to distinguish between freely dissolved and DOMassociated 14 C-activity in fractionated DOM solutions. To optimise flocculation conditions the pH in each solution was adjusted to pH 5 by adding 0.1 M HCl. Flocculation was conducted in duplicates using 50 Ž- 10 and - 100 kDa solutions. and 25 ml Ž- 1 kDa solution. of volume, respectively. Ten percent aluminium sulfate solution wAl 2 ŽSO4 . 3 x was prepared as coagulant agent. The dose required to obtain maximal DOC removal Žup to 60%. was 1 g aluminiumrg DOC. Aluminium sulfate solution was added to the ultrafiltrates mixing vigorously for 3 min. Mixing intensities were reduced, thereafter, for 25 min to

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

5

allow DOM flocs to grow. Flocs were allowed to settle during a period of 30 min without mixing. Concentrations of Ža. DOC and Žb. 14 C-activity were determined in the supernatant before and after flocculation in order to: Ža. control the efficiency of the flocculation; and Žb. to differentiate between freely dissolved and DOM-associated 14 C-activity.

3. Results and discussion 3.1. Properties of dissol¨ ed organic matter

Fig. 1. Composition of DOM made up of molecular weight fractions during composting. The relative amount of each fraction regarding the total DOC in the DOM solution is displayed Žmean value and S.D. from triplicates .. Ža. Grams of DOC eluted from 1 kg compost.

In the course of degradation of easily degradable substances during the thermophilic phase Žday 0᎐29., the amount of extractable DOC decreased by ) 60% and reached constant values after 120 days of composting ŽFig. 1.. After the thermophilic phase Žday 29., DOM had a maximum amount of high molecular DOM Ž) 100 kDa. with decreasing amounts, thereafter ŽFig. 1.. 13 C-NMR spectra revealed that fresh DOM showed a low aromaticity ŽTable 1.; polysaccharides Žalkyl-C and O-alkyl-C. presented the main fractions during the total experimental time. DOM exhibited the highest aromaticity between days 78 and 200 of composting, showing a maximum after 120 days ŽFig. 2, Table 1., thus indicating the accumulation of recalcitrant constituents of aromatic structure. As shown by the absorbance at 254 nm ŽFig. 2., this maximum in aromaticity occurred in all DOM molecular weight fractions. The phase of maximal aromaticity also featured a high amount of high molecular DOM ŽFig. 1., a fraction that is predestined for contaminant binding ŽRaber and Kogel-Knabner, ¨ 1997.. After 370 days of composting, the amount Table 1 Percentage of chemical shift bands of the total CPMAS

13

of high molecular DOM was lower than before ŽFig. 1., and the aromaticity was lower than after 120 days ŽFig. 2.. The evaluation of structural changes of DOM during composting has intensively been studied before by 13 C-NMR spectroscopy, finding a decrease of O-alkyl-C and a relative increase in aromatic-C with time ŽChefetz et al., 1996; Chen et al., 1997.. However, a maximum of aromaticity after 120 days and a decrease with preceding maturity has not been reported. From these results it can be expected that after 120 days of composting the soluble constituents of compost show the highest binding capacity Žmolecular weight and aromaticity. for hydrophobic contaminants. 3.2. Properties of

14

C-contaminants

While flocculation is successful for humic and fulvic acids ŽLaor and Rebhun, 1997. polysaccha-

C-NMR spectra from DOM at different stages of composting

Stage of composting Ždays.

Alkyl-C 0᎐50 ppm

O-Alkyl-C 50᎐110 ppm

Aromatic-C 110᎐160 ppm

Carboxyl-C 160᎐190 ppm

Carbonyl-C 190᎐220 ppm

0 29 120 200

17.7 24.2 19.1 21.2

57.6 40.0 34.8 35.6

12.9 19.1 26.1 24.6

11.1 15.2 18.0 16.5

0.6 1.4 2.0 2.1

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

6

Fig. 2. Specific absorption coefficient Ž ␭ s 254 nm. of molecular weight fractions from DOM at different stages of composting.

rides can hardly be flocculated with coagulant agents. However, this constituent presented the main structural fraction of DOM ŽTable 1.. Therefore, a maximum flocculation efficiency of 60% could be established in this study. Since the binding capacity of polysaccharides for hydrophobic contaminants is much lower compared to fulvic or humic acids, the flocculation efficiency of 60% seems sufficient. The investigations of DOM solutions regarding the distribution of 14 C-activity provided the following results: 3.2.1. DEHP During the course of the incubation, the aqueous extractability of DEHP-borne 14 C-activity varied between 1.0 and 4.9% of the initial activity A 0 , without showing any temporal trend ŽFig. 3.. The concentration of 14 C-activity in DOM solutions lay below water solubility Ž0.029 mg DEHP ly1 ; Rippen, 1989. at any time ŽFig. 3.. The dis-

Fig. 3. Ultrafiltration of DOM solutions containing 14 CDEHP. Percentage distribution of 14 C-activity upon molecular weight fractions at different stages of composting Žmean values and standard deviations of triplicates .. Ža. Micrograms of DEHP equivalents per litre of DOM solution; and Žb. % 14 C-activity in DOM solution of initially applied 14 C-activity A0 .

tribution of 14 C-activity among different DOM weight fractions changed markedly during composting ŽFig. 3.. Initially, equivalent amounts of 14 C-activity were present in the largest and smallest DOM fraction, and less than 10% in the intermediate fractions ŽFig. 3.. After 120 days, over 80 % of the 14 C-activity was found in the two low molecular weight fractions, whereas after 370 days, 14 C-activity was dominant in the highmolecular DOM fraction ŽFig. 3.. Flocculation experiments indicated that the portion of DOM-associated 14 C-activity amounted to less than 15%, while 88᎐94% of 14 C-activity was freely dissolved in solution ŽFigs. 4᎐6.. The extent of DOM-association did not correlate with the size distribution of DOM ŽFig. 6.. Furthermore, sorption experiments with test solutions showed that the parent compound DEHP was

Table 2 Percentage distribution of the model substances DEHP and pyrene upon the fractions ‘filtrate’, ‘retentate’ and ‘filter membrane’ in sorption experiments a Model substance

- 100 kDa filter membrane

- 10 kDa filter membrane

- 1 kDa filter membrane

Filtrate

Retentate

Filter

Filtrate

Retentate

Filter

Filtrate

Retentate

Filter

DEHP Pyrene

1.8 7.6

57.1 36.5

41.1 55.9

0.9 1.7

0.2 1.5

0.7 4.4

0.2 0.4

0.3 0.3

0.4 1.0

a Filtrates of former filtration steps were used for further filtration. Therefore, the summed up percentages of the fractions ‘filtrate’, ‘retentate’ and ‘filter membrane’ provide the percentage of 14 C-activity in the filtrate of the former filtration step.

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

7

Fig. 4. Ultrafiltration of DOM solutions containing 14 Cpyrene. Percentage distribution of 14 C-activity upon molecular weight fractions at different stages of composting Žmean values and standard deviations of triplicates .. Ža. Micrograms of pyrene equivalents per litre of DOM solution; and Žb. % 14 C-activity in DOM solution of initially applied 14 C-activity A0 .

Fig. 5. Ultrafiltration of DOM solutions containing 14 C-simazine. Percentage distribution of 14 C-activity upon molecular weight fractions at different stages of composting Žmean values and standard deviations of triplicates .. Ža. Micrograms of simazine equivalents per litre of DOM solution; and Žb. % 14 C-activity in DOM solution of initially applied 14 C-activity A0 .

strongly retained by the filter membrane by sorption as well as in the retentate ŽTable 2.. Since 14 C-activity in compost leachate passed the filter membranes to a higher extent compared to 14 Cactivity in the test solutions ŽFig. 3, Table 2., the 14 C-activity in the DOM solutions must therefore, have mainly consisted of polar metabolites and to a lower extent of the parent compound DEHP. This corresponds with the results of Weber Ž1989. who found that by composting DEHP contaminated waste, polar metabolites of DEHP reached a portion of 93᎐97% in the aqueous extracts. Weber Ž1989. investigated the displacement of DEHP within the compost heap and found DEHP rather immobile, whereas polar metabolites were

preferentially displaced. In relation to transformation processes during composting, DOM seems to have little effect on the mobilization of DEHP from compost. 3.2.2. Pyrene During the first 200 days of the incubation, the amount of pyrene-borne 14 C-activity in the DOM solution decreased from 2.6 to 1.6% of the initial activity A 0 ŽFig. 4.. After 370 days, the extractable 14 C-activity had increased to 2.4% of A 0 . The concentration of 14 C-activity in DOM solutions lay below water solubility Ž0.14 mg pyrene ly1 ; Rippen, 1989. at any time ŽFig. 4.. Throughout the composting experiment, approxi-

Fig. 6. Differentiation of 14 C-activity in freely dissolved and DOM associated portions by flocculation. Displayed is the amount of DOM associated 14 C-activity removed by flocculation Žmean values of min- and max-values from duplicates..

8

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

mately 50% of 14 C-activity in the DOM solution was retained during the first ultrafiltration step and thus, assumed to be associated with the high molecular DOM fraction ) 100 kDa ŽFig. 4.. This was supported by the flocculation experiments where approximately 20᎐40% of 14 C-activity in the next smaller fraction - 100 kDa was indeed associated with DOM ŽFig. 6.. With decreasing molecular weight the amount of freely dissolved 14 C-activity rose ŽFig. 6.. The dominating association of strongly hydrophobic compounds such as pyrene with the high molecular DOM fraction, corresponds with results of previous sorption studies ŽChin et al., 1997; Raber and Kogel¨ Knabner, 1997.. In contrast, DOM-aromaticity was of minor influence for the association with 14 C-activity, since the amount of 14 C-activity in the fraction ) 100 kDa was relatively low after 120 days of composting, when the DOM-aromaticity was at a maximum ŽFig. 3, Table 1.. Also, during later composting stages DOM aromaticity decreased while 14 C-activity in the high molecular DOM fraction increased. This could be due to different reasons. Ž1. The amount of ‘free’ pyrene increases during composting, so that an increasing amount is sorbed by the ultrafiltrate membranes of the first filtration step and, is thus, falsely regarded as associated with the largest DOM-fraction. To our knowledge, similar processes have not been observed before and are quite unlikely due to the high hydrophobicity of pyrene and its strong affinity to the solid organic matrix. Ž2. During composting, polar metabolites of pyrene are formed that can bind covalently to DOM. This would explain the poor relationship between DOM aromaticity and associated 14 C-activity. Non-bound polar metabolites would also explain the increasing amount of ‘free’ 14 C-activity in the smallest DOM-fraction determined in the flocculation experiments. The formation of such metabolites during composting is very likely, since their presence is the prerequisite for the observed mineralization of PAHs to CO 2 ŽRehmann et al., 1998; Richnow et al., 1998; Hartlieb and Klein, 2001.. While sorption can be reversible with contaminants being in exchange with the compost solu-

tion, the irreversible covalent bonds behave unaffected from partitioning processes and aromaticity implicating that there is no equilibrium with the surrounding solution and no exchange with filter membranes. 3.2.3. Simazine The 14 C-activity in DOM solutions containing simazine was mainly present in the low molecular DOM fraction during the first 200 days of the composting experiment ŽFig. 5.. Only at the end of the experiment, the extractable 14 C-activity was mainly associated with the high-molecular DOM. The concentration of 14 C-activity in DOM solutions lay below water solubility Ž5 mg simazine ly1 ; Rippen, 1989. at any time ŽFig. 5.. As flocculation experiments showed, the 14 C-activity was dominantly freely dissolved in solution ŽFig. 6.. The association to DOM was independent from molecular weight of DOM, i.e. DOM had only little influence on the mobilization of simazine and its metabolites from compost. Sorption to solid organic matter seems to have a stronger influence on triazines than sorption to DOM since the wash-out of atrazine was found to be smaller regarding amount and depth of leaching in humus-rich materials compared to agricultural soils with a small humus content ŽScheunert and Korte, 1985; Weber, 1989.. Also, a reduced leaching of simazine was found after organic matter application on agricultural soils due to organic matter enhanced sorption and degradation processes ŽCox et al., 1999..

4. Conclusions 4.1. Hazard potential of composting regarding mobilization of contaminants The results of this study indicate that degradation and transformation processes in the course of composting had a strong influence on the solubility of the 14 C-labelled model substances. Mineralization of the model substances during 370 days of composting amounted to 86 ŽDEHP., 60 Žpyrene. and 16% Žsimazine. of the initially applied 14 C-activity, respectively ŽHartlieb and Klein,

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

2001.. After 370 days, 1.7 ŽDEHP., 24 Žpyrene. and 65% Žsimazine. of the initially applied 14 Cactivity was non-extractably bound to the compost matrix ŽHartlieb and Klein, 2001.. While in soil the mobilization of hydrophobic contaminants is strongly dependent on DOM content and DOM properties Žmolecular weight, aromaticity. ŽRaber and Kogel-Knabner, 1997; ¨ Marschner 1998; Doring and Marschner, 1998., ¨ these factors played a secondary role during composting. Aromaticity of DOM did not correlate with higher concentrations of 14 C-activity in DOM solutions or in specific molecular weight fractions. The behaviour of the three model substances during composting differed from one another and was dependent on substance-specific properties. However, it can be concluded that the intensive conditions during composting with high temperatures and a high density of microorganisms favoured the transformation of contaminants into metabolites and their fixation to the solid and dissolved organic matter matrix, probably with covalent bonds, thus leading to a reduced mobilization of the model substances. Although the compost material contained high concentrations of the model substances, maximum concentrations of 14 C-activity in DOM solutions were lower compared to the water solubility and did not reflect the potential binding capacity of compost leachate. For example, 70 times higher pyrene concentrations have been documented in compost leachate than were found in this study ŽBusche and Hirner, 1997.. Compost material, in principle, has a high potential of mobilising PAHs, which leads to problems regarding the recultivation of remediated contaminated sites with compost ŽBusche and Hirner, 1997; Raber and Kogel-Knabner, 1997; ¨ .. Doring and Marschner, 1998; Marschner, 1998 ¨ According to the present study, after 120 days of composting the soluble organic compounds exhibited the highest binding capacity for hydrophobic contaminants. Therefore, when applying compost on agricultural land, the use of compost of this stage implicates the highest potential hazard for mobilising contaminants in the soil compared to compost from other stages of composting. It has to be mentioned that the present results are

9

restricted to the conditions of composting, the influence of changing conditions regarding pH or salinity in soil has not been investigated. By affecting the DOM quality these parameters might indirectly have a major effect on contaminant mobilization ŽDoring and Marschner, 1998; ¨ Marschner, 1998.. However, when investigating the hazard potential of composting contaminated biowaste with respect to the mobilization of contaminants by compost leachate, one has to keep in mind that contaminated organic material will always undergo similar degradation processes independently from being in a composting plant or being left in the environment, e.g. on agricultural soil. Therefore, by not composting contaminated biowaste, leaching will not be prevented; it will only be shifted to a different surrounding.

Acknowledgements We would like to thank the Deutsche Bundesstiftung Umwelt for financial support. References Adani F, Genevini PL, Tambone F. A new index of organic matter stability. Compost Sci Util 1995;3:25᎐37. Buffle J, Leppard GG. Characterization of aquatic colloids and macromolecules. 2. Key role of physical structures and analytical results. Environ Sci Technol 1995;29:2176᎐2184. Busche U, Hirner AV. Mobilization potential of hydrophobic organic compounds ŽHOCs. in contaminated soils and waste materials: Part II: Mobilization potential of PAHs, PCBs, and phenols by natural waters. Acta hydrochim hydrobiol 1997;25Ž5.:248᎐252. Chefetz B, Hatcher PG, Hadar Y, Chen Y. Chemical and biological characterization of organic matter during composting of municipal solid waste. J Environ Qual 1996;25:776᎐785. Chen Y, Inbar Y, Hadar Y, Malcolm RL. Chemical properties and solid-state CPMAS 13 C-NMR of composted organic matter. Sci Total Environ 1989;81r82:201᎐208. Chen Y, Chefetz B, Hadar Y. Formation and properties of humic substances originating from composts. In: De Bertoldi M, Sequi P, Lemmes B, Papi T, editors. The science of composting. Bologna: Blackie Academic and Professional, 1996:382᎐393. Chen Y, Inbar Y, Chefetz B, Hadar Y. Composting and recycling of organic wastes. In: Rosen D, Tel-Or E, Hadar Y, Chen Y, editors. Modern agriculture and the environment. Bologna: Kluwer Academic and Professional, 1997.

10

N. Hartlieb et al. r The Science of the Total En¨ ironment 278 (2001) 1᎐10

Chin YP, Aiken G, O⬘Loughlin E. Molecular weight, polydispersity, and spectroscopic properties of aquatic humic substances. Environ Sci Technol 1994;28:1853᎐1858. Chin YP, Aiken G, Danielsen K. Binding of pyrene to aquatic and commercial humic substances: the role of molecular weight and aromaticity. Environ Sci Technol 1997;31: 1630᎐1635. Chiou CT, Malcolm RL, Brinton TI, Kile DE. Water solubility enhancement of some organic pollutants and pesticides by dissolved humic and fulvic acids. Environ Sci Technol 1986;20:502᎐508. Cox L, Hermosin MC, Cornejo J. Effects of two organic amendments Žliquid and solid. on simazine leaching in soil. In: DelRe AAM, Brown C, Capri E, Errera G, Evans SP, Trevisan M, editors. Human and environmental exposure to xenobiotics, 1999:263᎐270. Delschen T. Impacts of long-term application of organic fertilizers on soil quality parameters in reclaimed loess soils of the Rhineland lignite mining area. Plant Soil 1996;213 Ž1᎐2.:43᎐54. Doring U, Marschner B. Water solubility enhancement of ¨ benzoŽa.pyrene and 2,2⬘,5,5⬘-tetrachlorobiphenyl by dissolved organic matter. Phys Chem Earth 1998;23:193᎐197. Hartlieb N, Klein W. Fate and behaviour of organic contaminants during composting of municipal biowaste. In: Rees RM, Ball BC, Campbell CD, Watson CA, editors. Sustainable management of soil organic matter. Wallingford: CABI Publishing, 2001:150᎐157. Kile DE, Chiou CT. Water solubility enhancement of non-ionic organic contaminants. In: Suffet IH, MacCarthy P, editors. Aquatic humic substances: influence on fate and treatment of pollutants, ACS symposium series 219. Washington DC: American Chemical Society, 1989:129᎐157. Kukkonen J, McCarthy JF, Oikari A. Effects of XAD-8 fractions of dissolved organic carbon on the sorption and bioavailability of organic micropollutants. Arch Environ Contam Toxicol 1990;19:551᎐557. Laor Y, Rebhun M. Complexation ᎏ flocculation: a new method to determine binding coefficients of organic contaminants to dissolved humic substances. Environ Sci Technol 1997;31:3558᎐3564. Magee BR, Lion LW, Lemley AT. Transport of dissolved

organic macromolecules and their effect on the transport of Phenanthrene in porous media. Environ Sci Technol 1991;25:323᎐331. Marschner B. DOM-enhanced mobilization of benzŽa.pyrene in contaminated soils under different chemical conditions. Phys Chem Earth 1998;23:199᎐203. Martius GGS, Wiebner A, Stottmeister U. Degradation efficiency and molecular size alteration during the aerobic microbial treatment of lignite pyrolysis deposite water. Appl Microbiol Biotechnol 1996;45:692᎐699. McCarthy JF, Robertson LE, Burrus LW. Association of benzoŽ a.pyrene with dissolved organic matter: prediction of K DO M from structural and chemical properties or the organic matter. Chemosphere 1989;19:1911᎐1920. McCarthy JF, Zachara JM. Subsurface transport of contaminants. Environ Sci Technol 1989;23:496᎐502. Raber B, Kogel-Knabner I. Influence of origin and properties ¨ of dissolved organic matter on the partition of polycyclic aromatic hydrocarbons. Euro J Soil Sci 1997;48:443᎐455. Raber B, Kogel-Knabner I, Stein C, Klem D. Partitioning of ¨ polycyclic aromatic hydrocarbons to dissolved organic matter from different soils. Chemosphere 1998;36Ž1.:79᎐97. Rehmann K, Noll H, Steinberg C, Kettrup A. Pyrene degradation by Mycobacterium sp. strain KR2. Chemosphere 1998;36:2977᎐2992. Richnow HH, Eschenbach A, Mahro B, Seifert R, Wehrung P, Albrecht P, Michaelis W. The use of 13 C-labelled polycyclic aromatic hydrocarbons for the analysis of their transformation in soil. Chemosphere 1998;36:2211᎐2224. Rippen G. Handbuch der Umweltchemikalien. Stoffdaten, Prufverfahren, Vorschriften. II-2.5. Umweltkonzentratio¨ nen organischer chemikalien. Erganzbares Loseblattwerk. ¨ Landsberg: Ecomed Verlagsgesellschaft, 1989. Scheunert I, Korte F. Interactions in the fate of chemicals in terrestrial systems. Ecotox Environ Safety 1985;9:385᎐399. Stevenson FJ. Humus chemistry ᎏ genesis, composition, reactions, 2nd ed. New York and Chichester: John Wiley & Sons Inc, 1994. Weber B. Verhalten von 14 C-markierten Umweltchemikalien bei der Herstellung von Gartenkompost. Dissertationsdruck Frank, Munchen, 1989. ¨