MSU LIBRARIES 4313—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wi11 be charged if book is returned after the date stamped below. REDUCTIVE DECHLORINATION 0F CHLORINATED PHENOLS IN ANAEROBIC UPFLOW BIOREACTORS BY Mary Luise Krumme A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Crop and Soil Science 1986 ABSTRACT REDUCTIVE DECHLORINATION OF CHLORINATED PHENOLS IN ANAEROBIC UPFLOW BIOREACTORS By Mary Luise Krumme The reductive dechlorination of chlorinated aromatic compounds was studied in anaerobic upflow bioreactors with chlorinated phenols as the sole carbon and energy source. Dechlorinating activity was maintained in three separate anaerobic bioreactors for 350, 400 and 190 days. Mineralization of approximately 40% of added chlorophenols to CH4 and C02 was demonstrated. Substrate loading rates of up to 20 mg l'1 d'1 at a hydraulic retention time of two or four days were achieved at a substrate conversion efficiency of greater that 90% as determined by measurement of substrate concentration in the effluent. The ability to degrade a mixture of all three monochlorophenols and 3,4,5-trichlorophenol was demonstrated. The majority of the active biomass was located at the bottom of the bioreactor in the form of a sludge blanket. Three microorganisms which dominated the biomass were identified by morphology as a putative anaerobic phenol- oxidizing bacterium, a Methanosarcina sp. and a Methanothrix sp. TABLE OF CONTENTS List of Tables ................................................... iii List of Figures ................................................... iv Introduction ....................................................... 1 Material and Methods .............................................. 18 Bioreactor Methods ........................................... 18 Analytical Methods ........................................... 25 Experimental Procedures ...................................... 27 Results ........................................................... 36 Discussion ........................................................ 99 Appendix Isolation of anaerobic phenol and benzoate- oxidizing bacteria .......................................... 104 List of References ............................................... 123 LIST OF TABLES Table Page 1. Mineralization of m-chlorophenol to gaseous end products in Bioreactor I ............................................. 52 2. Fate of 14C-para-chlorophenol in Bioreactor II .............. 53 3. Chlorophenol removal efficiency at different substrate loading rates and hydraulic retention times in Bioreactor III .............................................. 57 LIST OF FIGURES Figure Page 1. Proposed pathway for the anaerobic metabolism of phenol ..... 12 2. Diagram of hybrid anaerobic upflow bioreactor ............... 2O 3. Experimental design for trapping radiolabeled material ...... 3O 4. Influent and effluent substrate loading rate for 10. 11. 12. 13. 14. 15. 16. Bioreactor I ................................................ 38 Hydraulic retention time for Bioreactor I ................... 4O Influent and effluent substrate loading rate for Bioreactor II ............................................... 43 Hydraulic retention time for Bioreactor II .................. 45 Influent and effluent substrate loading rate for Bioreactor III .............................................. 48 Hydraulic retention time for Bioreactor III ................. 50 Turnover time of a non-mineralized substrate, ortho- chlorophenol, in Bioreactor I ............................... 56 Influent and effluent substrate loading rate for Bioreactor III during days 40 though 100 .................... 59 Influent and effluent substrate loading rate for Bioreactor III during days 100 through 127 .................. 62 Influent and effluent substrate loading rate for Bioreactor III during days 128 through 169 .................. 64 Influent and effluent substrate loading rate for Bioreactor III during days 171 through 187 .................. 66 Influent and effluent substrate loading rate for 2,4,6-trichlorophenol ....................................... 69 Cumulative substrate in the influent and effluent for 2,4,6-trichlorophenol ................................... 71 Figure 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. page Influent and effluent substrate loading rate for 3,4,5-trichlorophenol ....................................... 73 Cumulative substrate in the influent and effluent for 3,4,5-trichlorophenol ................................... 75 Influent and effluent substrate loading rate for pentachlorophenol ........................................... 78 Cumulative substrate in the influent and effluent for pentachlorophenol ....................................... 80 Location of the biomass in Bioreactor I ..................... 82 Location of the biomass in Bioreactor III ................... 85 Methanogenic activity per column section in Bioreactor I ................................................ 87 Methanogenic activity per column section in Bioreactor III .............................................. 89 Degradative activity per column section in Bioreactor III .............................................. 91 Photomicrograph of bioreactor biomass depicting morphology similar to a putative phenol-oxidizing microorganism ............................................... 94 Photomicrograph of bioreactor biomass depicting morphology similar to Methanosarcina sp ..................... 96 Photomicrograph of bioreactor biomass depicting morphology similar to Methanothrix sp ........................ 98 Phase contrast photomicrograph of the methanogenic phenol-degrading enrichment depicting the proposed phenol-oxidizing microorganism (2000K) ...................... 109 Phase contrast photomicrograph of the methanogenic phenol-degrading enrichment (2000K). The microorganisms found in the enrichment include: A) a proposed phenol- oxidizing bacterium, B) a Methanothrix sp., and C) a presumed hydrogen utilizing methanogen ...................... 111 Figure page 31. Thin section electron micrograph of the proposed phenol- oxidizing microorganism taken from the methanogenic phenol-degrading enrichment culture ......................... 113 32. Methane production from various aromatic substrates in the phenol-degrading enrichment culture ..................... 115 33. Phenol degradation by isolate P1 in co-culture with Desulfovibrio Gll ........................................... 117 34. Benzoate degradation by isolate B4 in co-culture with Desulfovibrio Gll ........................................... 120 INTRODUCTION The Environmental Protection Agency estimated that the United States generates more than 70 billion tons of hazardous waste annually (Thomas, 1984). Because of the vast quantity of waste produced, there is an increasing need for establishing large—scale hazardous waste disposal methods. Biological treatment can eliminate some hazardous compounds, such as chlorinated aromatic compounds, by aerobic or anaerobic degradation to yield inocuous compounds. Some halogenated aromatic compounds, such as the chlorophenols, can be degraded and mineralized in both aerobic and anaerobic systems; however, anaerobic biological treatment can offer many advantages such as lower operational costs and the possibility of novel reactions. Nevertheless, anaerobic biodegradation treatment systems for halogenated aromatics need further development in order to more fully understand the microbial processes involved and to make the processes feasible for large-scale treatment of toxic waste. Chlorophenols are widely used industrial chemicals. Pentachlorophenol (PCP) is registered as an insecticide, fungicide, herbicide, algicide, disinfectant, and an ingredient in antifouling paint (Crosby, 1981). In 1981, the U.S. consumption of PCP was 23,000,000 kg/year (Arsenault, 1976). Approximately 80% of the PCP manufactured was used for commercial wood treatment (Crosby, 1981). Chlorophenols are also precursors for other industrially important 2 herbicides such as 2,4-D and 2,4,5-T (Environmental Protection Agency, 1980). In addition to the deliberate manufacturing of chlorophenols, these chemicals are by-products from other processes such as the chlorination of water for disinfection (Ahlborg and Thunberg, 1980; Environmental Protection Agency, 1980) and the bleaching of paper pulp (Lindberg and Lund, 1980; Salkinoja-Salonen et al., 1981). Because of their widespread usage, chlorophenols are found throughout the environment. The chlorophenols, and in particular 2,4-dichlorophenol, 2,4,6-trichlorophenol and PCP have been placed on the Environmental Protection Agency list of 129 Priority Pollutants (Keith and Telliard, 1979). Many different types of environmental samples have been found contaminated with chlorophenols including soil, river, lake and sea water, sediment, and snow (Salkinoja- Salonen et al. 1981; Salkinoja-Salonen et al., 1984; Paasivirta et al., 1985; Valo et al., 1985; Xie et al., 1986). Chlorophenol contamination has been found not only near point pollution sources such as the Finnish saw mills treating wood with PCP (Valo et al., 1984; Valo et al., 1985), but contamination has also been found in otherwise pristine environments (Paasivirta et a1., 1985). Even man is contaminated with chlorophenols; PCP and other chlorophenols have been found in greater than 90% of human urine samples from occupationally unexposed groups (Edgerton et al., 1981). Chlorophenols, especially PCP, are toxic compounds. Their toxicity has made them both useful as lethal substances and harmful as environmental pollutants. PCP is toxic to nearly all forms of life including man, animals and plants (Crosby, 1981); PCP acts as an 3 inhibitor of oxidative phosphorylation (Weinbach et a1., 1965; Ahlborg and Thunberg, 1980). Numerous fish kills have been reported from PCP amended paddy soils which prompted the Japanese to limit usage of PCP as a herbicide (Crosby, 1981). Methanogens, in particular, are sensitive to PCP; thus the presence of PCP in waste effluent can cause process upsets in anaerobic treatment facilities. PCP has been found to completely inhibit methanogenesis from other carbon sources at concentrations as low as 1 mg/l (Guthrie et a1., 1984; Godsy et a1., 1986). The biodegradation of chlorophenols has been demonstrated in laboratory studies, environmental fate studies and industrial waste treatment processes. Most of the information regarding the biodegradation potential of chlorophenols and other chlorinated aromatic compounds comes from aerobic studies in which oxygen functions as the terminal electron acceptor. In general, ring cleavage precedes dehalogenation and susceptibility to biodegradation decreases with increasing halogenation (Tabak et al., 1964; Dorn and Knackmuss, 1978; Dorn and Knackmuss, 1978a; Knackmuss, 1981; Kobayashi and Rittman, 1982; Ghosal et a1., 1985). In contrast to aerobic systems, biodegradation of halogenated aromatic compounds generally procedes by an initial dehalogenating step resulting in partially or fully dehalogenated products (Tiedje et 31., 1986). In both cases, the resulting aromatic nucleus is more susceptible to either aerobic or anaerobic microbial attack. Dehalogenation under anaerobic conditions was first observed for chlorines on aliphatic carbons such as DDT [1,1,1-trichloro-2,2,-bis (p-chlorophenyl) ethane] and lindane (r-hexachlorocyclohexane) (McRae 4 et al., 1967; McRae et al., 1969). Numerous abiotic and biotic systems have been reported which were able to remove chlorines from non-aromatic carbons under anaerobic conditions (Barker et al., 1965; Miskus et al., 1965; Guenzi and Beard, 1967; McRae et al., 1967; Wedemeyer, 1967; McRae et al., 1969; Zoro et al., 1974; Jagnow et al., 1977). Since these early reports, numerous studies have reported on reductive dehalogenation from aliphatic carbons including compounds such as the halogenated hydrocarbons (Bouwer et al., 1981; Bouwer and McCarty, 1983; Kleopfer et al., 1985; Vogel and McCarty, 1985). One of the earliest observations of dehalogenation from an aromatic nucleus comes from field studies with PCP (Ide et al., 1972; Kuwatsuka and Igarashi, 1975; Murthy et al., 1978). Because PCP was used extensively as an herbicide in rice paddy fields, many investigations were conducted on flooded soils where conditions were either oxygen limited or anaerobic. Ide et a1. (1972) found that reductive dechlorination occurred more readily for chlorines on the ortho or para position resulting in the accumulation of meta- substituted chlorophenols. Kuwatsuka and Igarashi (1975), and Murthy et a1.(1978) also examined the fate of PCP in anaerobic soil and found reductive dechlorination at all positions on the aromatic ring. Pentachloroanisole, the methylated derivative of PCP, was also found as a product. There was no evidence for the complete mineralization of PCP in these studies. Since these early reports, reductive dechlorination has also been observed for the pesticides diuron (3,4-dichlorophenyl-l,1- dimethylurea), (Attaway et al., 1982; Attaway et al., 1982a; Stepp et 5 al., 1985), chlornitrofen (4-nitrophenyl-2,4,6-trichlorophenyl ether), (Yamada and Suzuki, 1983), thiobencarb (S-4-chlorobenzyl-N,N- diethyl thiocarbamate), (Moon and Kuwatsuka, 1984), and techlofthalam (N-(2,3-dichlorophenol)-3,4,5,6-tetrachlorophthalamic acid), (Kirkpatrick et al., 1981). In pond sediment previously treated with diuron, dechlorination of the para-chlorine was observed resulting in the formation of monuron (3-(4-chlorophenyl)-l,1- dimethylurea) which was not further metabolized (Attaway et al., 1982; Attaway et al., 1982a). In further experiments, compounds with chlorines in the meta and para position [3,4-dichlorophenylurea, linuron (3,4-dichloro-pheny1-1-methyl-l-methoxy urea), and isopropyl- 3,4-dichlorocarbanilate] were also dehalogenated at the para-chloro position in anaerobic pond sediment (Stepp et al., 1985). A loss of chlorine from the para position has also been demonstrated for the pesticides thiobencarb and chlornitrofen in flooded paddy soils (Yamada and Suzuki, 1983; Moon and Kuwatsuka, 1984). Similarily, reductive dechlorination of an aryl chlorine was demonstrated for the pesticides techlofthalam in flooded paddy soil (Kirkpatrick et al., 1981). Reductive dehalogenation has been more thoroughly investigated for the halogenated benzoates (Suflita et al., 1982; Horowitz et al., 1983). Substrates were examined in sediment and in an enriched bacterial consortium derived from anaerobic sewage sludge grown on meta-chlorobenzoate. A variety of halogens were removed from the benzoate ring including iodo-, bromo-, fluoro- and chloro-. Bromo- and iodo- were dehalogenated with shorter lag times than either the fluoro- or chloro-. The meta-halogens were more susceptible to 6 attack than either the ortho- or para-halogens. No dehalogenation was observed for either ortho- or para-chlorobenzoate even after one year. Evidence that the mechanism involved in dehalogenation is biological includes: 1) dehalogenation has not been observed in sterile sediments, 2) there are long reproducible lag times before the onset of dehalogenation, and once acclimated to the substrate these lag times are diminished, 3) dehalogenating activity is lost above 39°C and, 4) activity is substrate specific as evidenced by specific cross-acclimation patterns (Horowitz et al., 1983). In addition, a unique dechlorinating bacterium has been isolated in pure culture from meta-chlorobenzoate enrichments (Shelton and Tiedje, 1984). The organism is a Gram negative straight rod with an unusual collar surrounding the cell. It uses both thiosulfate and sulfite but not sulfate as electron acceptors (Linkfield, 1986). The only substrates demonstrated to support growth of the organism were pyruvate and rumen fluid (Shelton and Tiedje, 1984; Linkfield, 1986). The organism dechlorinates meta-chlorobenzoate to stoichiometric amounts of benzoate but does not dechlorinate at other positions on the ring or other chloroaromatic compounds. In addition to the chlorobenzoates, reductive dechlorination of chlorophenols has been investigated in anaerobic sewage sludge. Initial investigations demonstrated the removal of chlorines from monochlorophenols in 10% anaerobic sewage sludge (Boyd et al., 1983). Removal of the chlorine preceded cleavage and subsequent mineralization of the aromatic nucleus as demonstrated by the transient accumulation of phenol. The rate of chlorophenol 7 degradation depended on the position of the chlorine on the ring. Ortho-chlorophenol degraded faster than meta-chlorophenol, and para- chlorophenol persisted over the eight week study. In further studies, all three monochlorophenols were dehalogenated in whole, fresh, anaerobic sewage sludge (Boyd and Shelton, 1984). Mineralization of para-chlorophenol, ortho- chlorophenol, and 2,4-dichlorophenol to CH4 and C02 was confirmed by radiotracer studies using 14C-labeled substrate. The fate of dichlorophenols was also examined in whole, fresh sludge. Reductive dechlorination ortho to the hydroxy moeity was observed for each dichlorophenol with an ortho-chlorine including 2,3-dichlorophenol, 2,6-dichlorophenol, 2,4-dichlorophenol, and 2,5-dichlorophenol. The chlorophenols examined without ortho-chlorines, 3,4-dichlorophenol and 3,5-dichlorophenol, persisted during the study. The anaerobic sewage sludge could be slowly acclimated to chlorophenols by weekly feedings of substrate resulting in more rapid degradation rates. Distinct patterns of acclimation were observed for sludge acclimated to either ortho-chlorophenol, meta-chlorophenol or para-chlorophenol. Sludge acclimated to ortho-chlorophenol degraded both ortho-chlorophenol and para-chlorophenol without a lag but did not degrade meta-chlorophenol. The dichlorophenols that were dehalogenated in fresh sludge, (2,3—dichlorophenol, 2,6- dichlorophenol, and 2,5-dichlorophenol), were not degraded in the ortho-chlorophenol acclimated sludge. Sludge acclimated to meta-chlorophenol degraded both meta- chlorophenol and para-chlorophenol without a lag but did not degrade ortho-chlorophenol. The dichlorophenols with meta- and para- 8 chlorines that persisted in fresh sludge, (3,4-dichlorophenol and 3,5-dichlorophenol) were degraded in the meta-chlorophenol acclimated sludge. Sludge acclimated to para-chlorophenol was able to degrade all three chlorophenols at rates para > meta > ortho. The para acclimated sludge was also able to degrade both 3,4-dichlorophenol and 2,4-dichlorophenol suggesting that dechlorination could occur at all positions on the aromatic ring. Additional experiments with anaerobic sewage sludge from three different Michigan communities confirmed the ability of the sludge to degrade all three monochlorophenols at rates of ortho > meta > para (Mikesell and Boyd, 1985). Additionally, dechlorination was observed for 2,4,6-trichlorophenol, 2,4-D, 2,4,5-T, and pentachlorophenol. The ortho-chlorines were rapidly removed from these substrates with subsequent removal of some of the para- and meta-chlorines. Pentachlorophenol was dehalogenated in the ortho and para positions resulting in the accumulation of 3,5-dichlorophenol which was not further dehalogenated in the 70 day experiment. The initial steps in the degradation of 2,4-D were cleavage of the ether linkage and removal of the ortho-chlorines. The resultant product, para-chlorophenol, was slowly degraded. The initial steps in the degradation of 2,4,5-T were also cleavage of the ether linkage and removal of the ortho-chlorine which resulted in 3,4- dichlorophenol. Further removal of the meta-chlorine resulted in the accumulation of para-chlorophenol which was not further degraded in the study. This is in contrast to earlier studies with 2,4,5-T in which only the para-chlorine was removed resulting in the 9 accumulation of 2,5-D (Suflita et al., 1984). This para- dechlorinating activity, however, was selected for by growth on meta- chlorobenzoate. Further experiments were conducted with pentachlorophenol in anaerobic sewage sludge acclimated to all three monochlorophenols (Mikesell and Boyd, 1986). Again, distinct patterns of acclimation were observed. In general, pentachlorophenol was dechlorinated at both ortho positions in sludge acclimated to ortho-chlorophenol. In para-chlorophenol acclimated sludge, pentachlorophenol was dechlorinated at both the para and ortho positions, and meta- chlorophenol acclimated sludge dechlorinated pentachlorophenol at the meta-positions. The products, tri- and tetra-chlorophenols tended to accumulate in the acclimated sludge. Complete reductive dechlorination and partial mineralization of the aromatic nucleus, however, was observed when all three acclimated sludges were mixed together or when sludge was acclimated simultaneously to all three monochlorophenols. The dechlorinated products, meta-chlorophenol, 3,5-dichlorophenol, and 3,4,5- trichlorophenol accumulated at background levels in the sludge upon repeated pentachlorophenol additions but were completely dehalogenated when pentachlorophenol additions ceased. Mineralization of the pentachlorophenol carbon to CH4 and C02 was also observed in the sludge mixture. Approximately 66% of the added carbon could be accounted for as CH4 and C02. Under methanogenic conditions, once the chlorophenols are completely dehalogenated, the resultant aromatic nucleus can be mineralized to CH4 and C02. Several researchers have reported on the 10 mineralization of phenol by microbial consortia from habitats including anaerobic sewage sludge and rumen fluid (Tarvin and Buswell, 1934; Healy and Young, 1978; Holladay et al., 1978; Balba et al., 1979; Boyd and Shelton, 1984; Fedorak and Hrudey, 1984; Barik et al., 1985; Young and Rivera, 1985; Godsy et al., 1986). The mineralization of phenol has been described by the stoichiometry of the following equation (Healy and Young, 1978): C6H50H + 4 H20 ------- > 3.5 CH4 + 2.5 C02 Several investigators have proposed a pathway for anaerobic phenol metabolism based on the detection of intermediates from phenol in mixed culture experiments (Bakker, 1977; Evans, 1977; Balba et al., 1979; Young and Rivera, 1985). The initial steps in the pathway are the reduction of phenol to cyclohexanone followed by ring cleavage to produce caproate or adipate (Figure 1). The fatty acids are then further oxidized to produce methanogenic precursors (Balba and Evans, 1977). The phenol pathway has not been demonstrated in pure culture experiments because until recently pure cultures of anaerobic phenol- oxidizing bacteria were not available. The free energy change (AGO') for phenol fermentation reactions are positive under standard conditions. When hydrogen is maintained at low partial pressures through the presence of a hydrogen-consuming microorganism, the free energy change for phenol oxidation becomes negative. Thus, anaerobic phenol-oxidizing bacterium can only be isolated on phenol in the presence of a hydrogen-consuming microorganism. Recently, two phenol-degrading organisms have been isolated in co-culture with Wollinella succinogenes, a hydrogen-utilizing 11 Figure 1. Proposed pathway for the anaerobic metabolism of phenol. N8 + Eu # x a x 20m _ Pu. 4|. .II. :NO NXHvHv .AII.TmmUnv N: 2330 o.o.>.: F. mL / 1 / 200.53 ococoxmjofixu 3:98:03? _oco£a {a .H muswfih HOZwIm “.0 533092.32 U_m0¢m<2< l3 bacterium (Barik at al., 1985). Both organisms were able to utilize benzoate, phenylacetate, phenol and hydrocinnamate as their sole carbon source. Other non-aromatic substrates were also able to support growth. An additional organism has been isolated in co- culture with Desulfovibrio sp. which is able to grow on benzoate or phenol (Krumme and Boyd, 1986). To exploit the enzymatic activities involved in the anaerobic dechlorination and mineralization of chlorophenols, a continual, large-scale treatment system needs development. Anaerobic biological reactors have been widely employed for the continual treatment of waste. The advantages of anaerobic reactors versus aerobic reactors are less energy consumption, less sludge production and the production of a useful product, methane (Olthof and Oleszkiewicz, 1982; Klass, 1984). Recently, anaerobic attached film expanded bed reactors have been introduced for waste treatment. In this reactor design, the waste stream enters the bottom of the reactor and passes by the attached microbial biomass where it can be converted to CH4 and C02. Methane gas and effluent are collected at the top. The advantages of an anaerobic attached film expanded bed reactor over conventional treament are shorter retention time, increased hydraulic load, the ability to treat low strength wastes, and an increased tolerance to toxic shock (Switzenbaum and Jewell, 1980; Jewell, 1982; Binot et al., 1983; van den Berg, 1984). A variety of wastes, including waste from sugar refineries and agriculture have been degraded in anaerobic upflow bioreactors (Olthof and Oleszkiewicz, 1982; Speece, 1983). 14 The biodegradation of chlorophenols has also been examined in anaerobic upflow bioreactors. The design scale of the bioreactor varies from small laboratory reactors to large industrial scale reactors. In all of the bioreactor studies, chlorophenols were added to the bioreactor along with other readily utilizable carbon sources. In one study, an anaerobic reactor was fed a mixture of glucose, methanol and acetate to stimulate methanogenesis along with varying concentrations and types of chlorophenols (Woods, 1985). The chlorophenols, di-, tri-, tetra-, and penta-chlorophenol were fed to the bioreactor at influent concentrations varying from 0.1 mg/l to 27 mg/l at a hydraulic retention time (HRT) of 0.5 days. Toxicity to the dechlorinating and methanogenic activity was observed at the higher chlorophenol input concentration. As the chlorophenol concentration in the feed increased from 2.4 mg/l to 8.0 mg/l, dechlorinating activity decreased. Rapid dechlorination was observed for chlorines ortho to the hydroxy moeity. Some meta-chlorines were removed only after four months of acclimation to the chlorophenols. Chlorines in the para position were not removed and no ring cleavage was observed. Rates of dechlorination increased with increasing chlorination of the ring. Thus, the fastest rate of dechlorination was observed for pentachlorophenol. Several different types of microorganisms were enumerated from the bioreactor including aerobes, facultative anaerobes, methanogens and yeast (Hakulinen et al., 1985; Woods, 1985). The isolated facultative anaerobes were able to grow on 2,4,6-trichlorophenol under aerobic but not under anaerobic conditions. When two of the 15 isolates were mixed together, identified as Klebsiella oxytoca and Pseudomonas aeruginosa, anaerobic degradation of 2,4,6- trichlorophenol was observed. No degradation or mineralization products from 2,4,6-trichlorophenol were reported. The significance of these microorganisms in chlorophenol biodegradation was not clear. The Klebsiella oxytoca, along with the other enumerated microbes, decreased in number in the bioreactor with increasing concentration of chlorophenol in the feed. The Pseudomonas aeruginosa remained constant. Yeasts, also isolated from the bioreactor, increased in number with increasing chlorophenol concentration in the feed but the increased number of yeasts had no effect on the rate of chlorophenol degradation. The biodegradation of chlorophenols from pulp waste effluent was examined in a two-stage reactor system consisting of an anaerobic upflow reactor and an aerobic trickling filter (Hakulinen and Salkinoja-Salonen, 1982; Salkinoja-Salonen at al., 1983; Salkinoja- Salonen et al., 1984). The extent of chlorophenol biodegradation due to each system was not determined. The bioreactor system was effective at removing much of the biological oxygen demand, chemical oxygen demand and toxicity from the pulp waste. Chlorophenols, typically found in pulp waste effluent, were also removed by the reactor system. The removal efficiency depended on the loading rate into the bioreactor and the HRT. The efficiency of chlorophenol removal was 93-95% with a six hour HRT and a waste stream containing 0.6-2.2 mg/l chlorophenols. The efficiency dropped to approximately 50% when the HRT decreased to three hours and the waste stream contained 1.5-2.0 mg/l chlorophenols. When the influent 16 and effluent waste streams from the reactor system were analyzed for individual chlorophenols, the authors found that both pentachlorophenol and 2,4,6-trichlorophenol were the most resistant to biodegradation and tended to be the first chlorophenols to appear when the chlorophenol removal efficiency dropped. Mineralization of 60 to 80% of pentachlorophenol to C02 was demonstrated using 140- pentachlorophenol. Methane was not reported as a product. In the present study, we have examined the dechlorination of chlorinated aromatic compounds in anaerobic upflow bioreactors. Chlorophenols have been used in these studies as model chlorinated aromatic compounds because they are a serious environmental pollution problem and represent one class of compounds that are reductively dechlorinated under anaerobic conditions. The dechlorination of chlorophenols in anaerobic upflow bioreactors was studied for several reasons. First, anaerobic upflow bioreactors are currently used for large-scale waste treatment processes. The dechlorination of chlorophenols in the same type of bioreactor would be the first step towards utilizing these unique activities for large-scale treatment. In addition to the potential for waste treatment, anaerobic upflow bioreactors can also be used in the laboratory as a tool for the enrichment of unique activities, such as dehalogenation, by enriching for stable bacterial communities. The bioreactor enrichment technique may be the only feasible method for studying these types of activities when conventional techniques, such as the isolation of individual community members, are difficult or impossible. l7 Bioreactors can also provide a means for acclimating microbial communities to degrade new substrates. New substrates can be slowly but continually introduced into the bioreactors to slowly adapt the microbes to new substrates. This may be an especially useful technique when the substrates of interest, such as the chlorophenols, are toxic to the microorganisms. Specific research objectives were set in order to determine the feasibility of degrading chlorophenols in anaerobic upflow bioreactors. These were: 1) determine whether dechlorinating activity can be maintained in anaerobic upflow bioreactors with chlorophenols as the sole carbon and energy source, 2) determine if mineralization of the chlorophenols occurs in the bioreactors, 3) examine the bioreactor for location of biomass and dominant morphology, 4) determine the substrate range, and 5) determine the substrate loading capacity. MATERIAL AND METHODS Bioreactor Methods Bioreactor Design. Hybrid anaerobic upflow bioreactors were used for the continual treatment of a synthetic chlorophenol waste stream. A diagram of the reactor is shown in Figure 2. The cylindrical plexiglass reactor was approximately one liter in volume with a height of 12 inches and an inner diameter of three inches. The ends of the reactor were conical shaped and threaded for easy removal. Cajon fittings were used to connect either 1/4 inch I.D. glass or stainless steel tubing to the cylinder. Sampling ports sealed with butyl rubber stoppers were located throughout the length of the bioreactor. Two areas inside the bioreactor were designed for the containment of the biomass. A fine mesh screen located at the bottom of the reactor was designed for the accumulation of a sludge blanket. The center of the bioreactor, designed for the attachment of the biomass, contained approximately 250 mls of 4 mm etched glass beads held in place by an additional screen. Gaseous end products were separated from liquid effluent in a glass gas release structure attached to the top of the cylindrical bioreactor. Gas was collected in an inverted graduated cylinder 19 Figure 2. Diagram of hybrid anaerobic upflow bioreactor. 20 Figure 2. GLASS GAS RELEASE STRUCTURE CAJ ON FlTTlNG—e— E s \ \ GAS COLLECTION THREADED / PLEXIGLASS\ / TOP I \ I \ rI’ . CONICAL \- SHAPE. 31 3 IN. I.D. PLEXIGLASS—p- CYLINDER WASTE RESERVOIR 8m. 4mmETCHED GLASSBEADS ,. .plma 1' ' MESHSCREEN __/TO SUPPORT M ghhiensign GLASS BEADS SLUDGE BLANKET \ I \\ ’I \ I \‘ ’l \ I FEED *RECYCLE RESERVOIR ”NE PUMP r' 21 placed in an acid-brine solution (2% HCl, 10% NaCl). The top of the graduated cylinder was fitted with a butyl rubber stopper for taking quantitative gaseous measurements. The glass gas release structure also contained a port for the collection of effluent in a one liter Erlenmeyer flask. The feed reservoir consisted of a one liter Erlenmeyer flask fitted with a rubber stopper. The feed flask was completely sealed by the rubber stopper and held in place by plastic tape. Tygon and stainless steel tubing connected the bottom of the feed flask through the stopper to a piston-action pump (Fluid Meter, Inc. RP6 pump equipped with a 1/4” ceramic piston pump head). Additional tygon and stainless steel tubing connected the pump to the bottom of the reactor. A slow continual stream of oxygen-free nitrogen was supplied to the feed flask headspace via the rubber stopper. The oxygen-free nitrogen gas was provided by a tank connected in series to a two- stage regulator (Airco 9600), a low pressure regulator (Matheson Model 70) and an adjustable relief valve (Union Carbide Corp. RV5570). The pressure release valve was set at 1-2 psi to prevent excess pressure build-up in the feed flask. The feed flask, containing anaerobic media, was replaced as needed by continually flushing the flask with oxygen-free nitrogen. Heat distribution inside the bioreactor was provided by eight aluminum strips (7.5" x 3/4" x 1/4") attached to the outside of the column. Column heat was obtained by means of a heat resistor tape wrapped around the outside of the aluminum strips and attached to a variable transformer. The temperature was controlled either by the 22 setting of the transformer or by connecting the transformer to a temperature controller (YSI Model 64RC, Yellow Springs Instrument Co., Inc.) equipped with a YSI Series 600 thermistor attached to the outside of the column. Temperature was monitored by a thermometer placed inside the bioreactor through a top well sealed by a Cajon fitting. The entire bioreactor was placed inside a wooden box packed with styrofoam for insulation. Inoculation Procedure. Strict anaerobic techniques were used throughout the inoculation procedure. The column was initially flushed with oxygen-free nitrogen for one hour. Flushing probes extended approximately half-way down into the column to insure adequate flushing. While flushing, reduced anaerobic mineral media, (RAMM), was added to the column using a specifically designed inoculation flask. Ports were sealed and the column was allowed to stand overnight before slowly draining the column while flushing. The column was inoculated using the inoculation flask and the ports were sealed. The inoculation flask consisted of a two liter Erlenmeyer flask sealed by a rubber stopper. A five-inch, ll-gauge stainless steel needle was attached to tygon tubing which went through the rubber stopper and extended into the bottom on the flask. An additional stainless steel needle pierced the rubber stopper and extended into the headspace of the flask. The flask was filled with either RAMM or inoculum while continually flushing with oxygen-free nitrogen. To inoculate the column, oxygen-free nitrogen was connected to the needle that pierced the stopper. Excess pressure caused the liquid 23 in the flask to escape through the additional needle which was placed inside the column through a top port. 0 ce nocu um. All three bioreactors were inoculated with either acclimated anaerobic sewage sludge or sewage sludge enrichment cultures that were obtained from the acclimated anaerobic sewage sludge. Anaerobic sewage sludge, obtained from a primary anaerobic sewage sludge digestor, (Jackson, MI), was acclimated by weekly feedings (20 ug/ml) of chlorophenols. The anaerobic sludge had been acclimated to the chlorophenols for approximately two years prior to use in the columns. Anaerobic enrichments were obtained by periodic transfers (25%) of acclimated sludge into RAMM. Enrichments were maintained by feeding approximately 10-20 ug/ml chlorophenols as the sole carbon and energy source every two days. Both acclimated sludge and enrichment culture were periodically monitored for chlorophenol content to prevent a toxic build-up of the substrate. Bioreactors I and III were initially inoculated with an enrichment that had been degrading meta-chlorophenol for approximately one year. The enrichment was obtained by two transfers of 50% and 25% from acclimated sludge. The bioreactors were inoculated by completely filling the bioreactor with the enrichment culture. Bioreactor II was initially inoculated with an equal mixture of sludge that had been acclimated to either ortho-, meta-, or para- chlorophenol. The bioreactor was seeded with a 1:1 mixture of the sludge and RAMM. When activity in Bioreactor II diminished at day 200 (see Figure 6), the bioreactor was reseeded with 100 ml (10%) of 24 a 111:1 mixture of sludge acclimated to either ortho-, meta- or para- chlorophenol. Feed Media. Feed media for all bioreactor experiments was reduced anaerobic mineral media (RAMM). The mineral salts media consisted of (per liter): 0.27 g of KHZPOA, 0.35 g of KZHPOA, 0.53 g of NHACl, 0.10 g of MgC12'6H20, 79 mg of CaClZ'2H20, 20 mg of FeClz'4H20, 1.0 ml of a trace metals solution and 1.0 ml of a vitamin solution. The trace metals solution consisted of (per liter): 0.5 g of Mn012'4H20, 0.05 g of H3B03, 0.05 g of ZnClz, 0.03 g of CuC12, 0.01 g of Na2M004'2H20, 0.50 g of CoC12'6H20, 0.05 g of NiClz'6H20, and 0.05 g of Na25e03. The vitamin solution consisted of (per liter): 0.02 g of biotin, 0.02 g of folic acid, 0.06 g of lipoic acid, 0.05 g of thiamin, 0.05 g of riboflavin, 0.05 g nicotinic acid, 0.10 g of pyridoxal-HCl, 0.05 g of pantothenic acid, 0.05 g of cyanocobalamin, and 0.05 g p-aminobenzoic acid. Media, prepared as needed in one liter Erlenmeyer flasks, was boiled for five minutes to remove oxygen and then cooled to approximately 35°C while being sparged with a 20% C02/80% N2 oxygen-free gas mixture. After cooling, 2.4 g NaHC03 and 0.24 g NaZS was added to the media and the flask was tightly stoppered. Chlorophenol substrates were added to the media just prior to use from anaerobic stock solutions of 6 mg chlorophenol/ml H20. Effluent was collected and measured daily to determine the amount of feed media which entered the bioreactor. Daily samples of the effluent were frozen for future high pressure liquid chromatographic (HPLC) determination of the chlorophenol concentration. 25 Analytical Methods Methane. Methane was collected as a gaseous product in the inverted cylinder placed in the acid-brine solution. Total volume of gas collected was measured daily and samples of the gas were taken periodically for methane quantification. Methane was quantified by injecting 0.2 ml of the gas into a Carle gas chromatograph equipped with a poropack Q column and a microthermistor detector. Dissolved methane in the effluent was extracted according to the procedure developed by Robinson et. a1 (1981) for extracting dissolved H2. Samples of the effluent were taken after separation of gaseous products in the glass gas release structure (Fig. 2), and before exposure to the atmosphere, by piercing the tygon tubing with a 22-gauge needle and slowly withdrawing a sample into a glass 50 ml syringe fitted with a one way stopcock. Methane was extracted from the effluent and then analyzed by the Carle gas chromatograph as previously described. The following equation was used to determine the amount of methane dissolved in the effluent: N - X[1+o((A/B)]/A where N- moles of CH4 ml'1 of extracted effluent, A- volume (ml) of aqueous phase in the 50 ml syringe, B- volume (ml) of the gas phase of the syringe, and X- moles of CH4 in the gas phase of the syringe at equilibrium. The value X is calculated by multiplying the number of moles of CH4 m1"1 of gas in the gas bubble, as determined by gas chromatography, by the volume of the gas bubble. A Bunsen absorption coefficient,°‘, of 0.03046 was used. e - a xt ctio . Effluent samples containing 3,4,5- trichlorophenol were concentrated using Sep-Pak C18 cartridges 26 (Millipore) before HPLC analysis. An effluent sample of 50 ml was concentrated in the cartridge and eluted with 2.0 ml of methanol. Recovery efficiency of the trichlorophenol was determined by spiking known quantities of trichlorophenol in spent effluent and concentrating by Sep-Pak extraction. Effluent spiked with a trichlorophenol concentration of 0.1-2.0 ug/ml was concentrated 25 times at a recovery of 80%. Analysis of Chlorophenols. Effluent samples were thawed and filtered through 0.45 uM HA Millipore filters. Monochlorophenols were routinely measured with a Waters HPLC consisting of a model 6000A pump and a model 441 UV absorbence detector set at 280 nm. The sample injection valve (Rheodyne 7125) was fitted with a 20-ul loop. The analytical column was a Waters Radial PAK C18 cartridge held in a ROM-100 Radial Compression Module. Peak areas were measured with a Hewlett Packard 3390A integrator. The trichlorophenols and pentachlorophenol were measured with a Waters HPLC consisting of a model 6000A pump, a model 45 pump, and a model 720 systems controller, coupled with a model 480 Lamba Max variable wavelength UV absorbance detector. The wavelength was set at 300.5 nm for pentachlorophenol and 280.0 nm for trichlorophenols. The analytical column was a uBondapak C18 column (Waters Assoc.). Mobile phase for both HPLC systems consisted of a 5% aqueous acetic acidzacetonitrile mixture with a flow rate of 2.0 ml/min. The mobile phase was adjusted to give a retention time of approximately 3.5 min for the monochlorophenols (1:1 mixture); 5 min for trichlorophenols (4:6 mixture) and 7 min for pentachlorophenol (3:7 27 mixture). Mixtures of the monochlorophenols were separated by using a mobile phase mixture of 8:2 resulting in retention times of 5-7 min. Experimental Procedures General. All experiments were conducted utilizing one of three bioreactors. Preliminary experiments were conducted with Bioreactor I which was initially seeded with an enrichment culture that had been degrading meta-chlorophenol. Mineralization studies were also performed with Bioreactor I. After nearly one year of operation (350 days), the bioreactor was terminated and studies were performed utilizing the bioreactor contents. Bioreactor II, in operation for 400 days, was initialy seeded with acclimated sludge that had been degrading all three monochlorophenols. After an initial start-up period, two mineralization studies were performed utilizing radiolabeled substrate. Bioreactor II was used for three additional experiments involving the fate of higher chlorinated phenols; 2,4,6- trichlorophenol; 3,4,5-trichlorophenol and pentachlorophenol. Bioreactor III was originally seeded from a meta-chlorophenol degrading enrichment. After an initial start-up period, four experiments were run sequentially to determine the effect of loading rate and (HRT) on the substrate conversion efficiency. The experiments were terminated after 190 days and the bioreactor contents were further characterized. Migg;glizg£igg_§£g§i§§. The mineralization of meta- chlorophenol was examined in Bioreactor I by quantitative recovery of 28 the mineralization product, CH4. The study took place during days 159 through 202 of operation with an average substrate loading rate of 10.2 mg l"1 d"1 and an average HRT of 6.7 days. The substrate conversion efficiency was 100%. Both gaseous and dissolved CH4 were recovered and quantified. The total amount of theoretical CH4 obtained from the total amount of added substrate chlorophenol during the 46 day period was determined based on the stoichiometry of the following equation: C6H50H + 4 H20 -------- > 3.5 CH4 + 2.5 C02 (Healy and Young, 1978) The mineralization of chlorophenol in Bioreactor II was examined in two separate experiments by determining the fate of 14C-para- chlorophenol. The first study took place from day 112 to 157 of operation with an average total substrate loading rate of 21.2 mg l'1 d'1 and an average HRT of 8.3 days. The average substrate conversion efficiency was 99%. Uniformly ring-labeled 14C-para-chlorophenol (obtained from Pathfinders with a specific activity of 11.61 mCi/mmol and purity of 98%) was added as a spike to the feed media. The spiked media contained 57.8 uCi of 14C para-chlorophenol and 150 mg/l of unlabeled substrate (a 1:4:1 mixture of ortho-:meta-zpara- chlorophenol) and was pumped into the bioreactor over a 48 hour period. Labeled material was recovered in a series of traps as depicted in Figure 3a. Effluent and any gaseous products were initially collected in a stoppered one liter Erlenmeyer flask. Before daily collection and measurement of the effluent, the contents of the flask were acidified with HCl and flushed with nitrogen gas by piercing the rubber stopper with a needle. Exhausted gas passed through a series 29 Figure 3. Experimental design for trapping rabiolabeled material. 30 Figure 3a. EXIT o T muuzm C 2 “PS COLLECTION Figure 3b. H IMRTED CYLINDER FOR “3310 SAMPLES 002 Traps. I Al] N. [MM EFFLLUENT 002 Traps Non-co2 Volatile COLLECTION Traps 31 of four traps (scintillation vials containing 10 ml of a COZ-trapping cocktail) to trap any C02 before venting to the atmosphere. Traps were changed daily and radioactivity was measured. Duplicate 1.0 ml samples of the effluent were added to universal cocktail (Scinti- verse II, Fisher Scientific) and counted. All samples were counted in a Beckman Model L88100 Scintillation Counter. Samples of effluent were also collected for HPLC analysis of the chlorophenol content. The second study took place from day 157 through 202 of operation with an average substrate loading rate of 15.5 mg 1'1 d"1 and average HRT of 7.2 days. The substrate conversion efficiency was 99%. Para-chlorophenol was added as a spike to 300 ml of feed media containing 53.4 uCi of [U]~14C-para-chlorophenol and 150 mg/l of total chlorophenol (1:4:1 mixture of ortho-:meta-zpara-), and was delivered to the bioreactor over a 72 hour period. Labeled material was recovered in a series of traps as depicted in Figure 3b. Gaseous products were separated from the effluent by means of the glass gas release structure. Gaseous products passed through a series of C02 traps before collection of non-C02 gases in the inverted cylinder. Samples of the gas collected in the inverted cylinder were periodically placed in 3 m1 Venoject vials (Terumo Medical Co.) for later analysis. Radioactivity was measured in the gaseous samples stored in Venoject vials by oxidizing the gas to 002 in a Harvey Biological Material Oxidizer Model 0x200 and collecting the resultant C02 in scintillation fluid. The effluent, containing any dissolved 140, was collected in a sealed Erlenmeyer flask. The flask was connected to a series of C02 32 traps followed by an additional series of traps (scintillation vials containing 20 ml of: 0.375 g 2,5-diphenyloxazole, 0.100 g dimethyl POPOP in 1.0 1 of toluene). Before daily collection of the effluent, tubing to the bioreactor was clamped off, the effluent was acidified with HCl and flushed with nitrogen gas. The effluent was then collected, the volume was measured and duplicate 1.0 ml samples were placed in scintillation fluid and counted. Effluent samples were also taken for high pressure liquid chromatographic determination of the chlorophenol concentration. Turnover lime and Sorption Studies. Bioreactor I was employed to determine the turnover time and recovery of a non-mineralized compound in the bioreactor during days 334 to 341 of operation. The average HRT was 4.1 days. Ortho-chlorophenol was used as the non- mineralized compound due to its similarity to the mineralized substrate, meta-chlorophenol. Ortho-chlorophenol was added to the bioreactor by injecting the substrate (3.0 ml of a 6.0 mg chlorophenol/ml H20 stock for a total of 18.0 mg) through tygon tubing that connected the pump to the bioreactor. Effluent samples were monitored daily for the presence of ortho-chlorophenol by HPLC. Irichlorgphenols and Pentachlorophenol Studies. Three experiments were run with Bioreactor II to determine the fate of higher chlorinated phenols in the bioreactor. The substrates used were 2,4,6-trichlorophenol, 3,4,5-trichlorophenol and pentachlorophenol. During these experiments, influent loading rates of the monochlorophenol substrates were decreased. Between experiments, the loading rates of the monochorophenols were increased to approximately 20 mg total substrate 1'1 d'l. 33 The higher chlorinated phenol substrates were added to the feed media just prior to use from stock solutions. Effluent was monitored for monochlorophenol, the higher chlorinated substrate and any possible chlorinated products. For the experiment with the substrate 3,4,5-trichlorophenol, effluent samples were first concentrated by Sep-Pak extraction (see analytical methods). Substrate Loading Rate and Hydraulic Retention Time Studies. Bioreactor III was used to determine the capacity of the bioreactors to degrade chlorophenols at different loading rates and HRT's. Substrate loading rates at a constant HRT varied by changing the concentration of chlorophenol in the feed media. HRT varied by changing the volume of feed media pumped into the bioreactor. The chlorophenol removal efficiency was calculated from the average influent substrate loading rate and the average effluent substrate concentration as determined by HPLC. Tea 0 rocedures. After 350 days of operation, Bioreactor I was disassembled and the contents were examined. Strict anaerobic techniques were employed thoughout the procedure by continually flushing the bioreactor with oxygen-free nitrogen. The bioreactor was divided into ten non-equal portions as shown in Figure 21. Portions of each section were collected for protein analysis, chlorophenol analysis, methane productivity and microscopic observations. Samples for protein determination (5.0 or 10.0 ml) were centrifuged at 10,000 rpm for 15 min, resuspended in phosphate buffer, and acidified overnight with concentrated HCl. Samples containing glass beads were vortexed and washed with phosphate buffer 34 to remove the biomass from the beads. After acidification, the samples were neutralized with NaOH and the protein was solubilized with 2N NaOH. Samples were centrifuged at 1200 x g for 15 min to remove any remaining particulate matter and then assayed according to the procedure of Lowry et a1. (1951). The bioreactor contents were also analyzed for the presence of chlorophenol. A 10.0 ml aliquot from each section was extracted with 5.0 ml acetonitrile, filtered through 0.45 uM Millipore HVLP filters and analyzed by HPLC as previously described. Methane production from each column section was also determined. A 25 ml portion of each section was placed into a 71 m1 serum bottle while continually flushing the bioreactor and the serum bottle with oxygen-free nitrogen. Substrate chlorophenol was added to each bottle to give a final concentration of 48 ug/ml before sealing the bottle with a butyl rubber stopper and incubating at 37°C in the dark. Methane production was monitored periodically as previously described. After 190 days of operation, Bioreactor III was disassembled and the contents were examined. Strict anaerobic techniques were employed throughout the procedure. The bioreactor was divided into five non-equal portions as shown in Figure 22 and analyzed for protein, dechlorinating activity, methane production and microbial morphology. Protein was determined as previously described. Duplicate samples were employed in the analysis. Dechlorinating activity and methane production were determined simultaneously, in triplicate. Methane production was determined as previously described for Bioreactor I. Periodic samples, (1.0 ml), 35 of the aqueous phase were taken and frozen. The samples were thawed, extracted with 0.5 ml acetonitrile, filtered though HVLP filters and assayed by HPLC as previously described. A Zeiss Photo-1 photomicrosc0pe was utilized for microscopic observations and photomicrographs. The bioreactor material was embedded on a slide in 1.5% agarose for the photomicrographs. RESULTS General. One of the primary objectives of the bioreactor experiments was to determine whether dechlorinating activity could be maintained in the bioreactors for long periods of time with chlorophenols as the sole carbon and energy source. An additional primary objective was to determine feasible substrate loading rates and hydraulic retention times. Three bioreactors were employed for these general studies. Bioreactor I and III were fed only meta- chlorophenol as the sole carbon source while Bioreactor II was fed a mixture of ortho-, meta-, and para-chlorophenol. Bioreactor I was a preliminary, long term experiment designed to answer the two primary objectives. Figure 4 describes the influent and effluent substrate loading rates for the entire operation of Bioreactor I including the start-up period. The hydraulic retention time (HRT) varied throughout the length of operation as depicted in Figure 5. After inoculation of the bioreactor at day 0, start-up conditions were maintained for approximately 100 days (Figure 4). During the initial 40 days of operation influent feed media was applied to the one liter bioreactor at a rate of 50 ml/day. The concentration of chlorophenol in the feed media was 60 mg/l to give a loading rate of approximately 3 mg 1'1 d-l. 37 Figure 4. Influent and effluent substrate loading rate for Bioreactor I. 38 Satan—3m Eon-3m Aux—v.5 2.5... 33.3.5 325:. III. , ‘41 ., I . . i 41 . F 5:55... 0 v» 00v 00...... hNN mNN MNN pNN opN hpN m—N n—N P L p . L p . . . . . L p . O I O— I ON 3 II on I O? I Or. on 0.0.30...“ 03.0.0an .3 83...... ScoraotoEorfilqim (0m) 0‘0“:an [pun 72 Figure 17. Influent and effluent substrate loading rate for 3,4,5- trichlorophenol. 73 on» P 0.0.30...“ .c0:.§u .... Agog 0...... On» - 0*” ON» 0 P 0.00.09“ 2.00:... .. 0.. I 0.. r u r «.0 .. 0.0 a 3 .. 0... n 0.0.. 05000.. _ocm.._n_o._o_ro_¢lm....m . Z 8:0... (App J00 .I0m 40d. 60:) upucqns 74 Figure 18. Cumulative substrate in the influent and effluent for 3,4,5-trichlorophenol. 75 0mm 300:5 .0.0._. + 300...... .0.0._. a A933 05.... 0.7% ON” ON» 0 pm - . 0.0...0nam 02.0.3550 _ocm£n_o..o_£o_..._.lm.fi.m .3 .25 (5”) ”Mains lam 76 indicating that degradation of the trichlorophenol occurred. Again, no dehalogenated products were detected in the effluent. The third experiment was performed utilizing pentachlorophenol as the higher chlorinated substrate. The average influent total substrate loading rate of the monochlorophenols was 7.55 mg 1'1 d'l. The average substrate conversion efficiency of the monochlorophenols was 86%. The average loading rate of the pentachlorophenol was 1.61 mg 1'1 d'1 during days 371 through 389 (Figure 19). The total cumulative amount of pentachlorophenol added to the bioreactor and detected in the effluent is depicted in Figure 20. Approximately 75% of the added pentachlorophenol could be detected in the effluent indicating that degradation was minimal. No lesser chlorinated products were observed in the effluent. Two different HRT's were used to determine the effect on the degradation of pentachlorophenol. During days 371 to 378, a HRT of 2.6 days was employed and during days 379 to 400 a HRT of 5.1 days was employed. The change in HRT had no apparent effect on the degradation of pentachlorophenol. Biomass Location. Bioreactors I and III, with meta-chlorophenol as the sole carbon source, were examined for the location of biomass. These experiments were performed at the termination of Bioreactor I and III after 350 and 190 days of operation, respectively. The location of the biomass, as protein, in Bioreactor I is shown in Figure 21. The bioreactor was divided into ten non-equal portions. Section 1 and 2, at the top of the bioreactor contained very little protein indicating that washout of the cells was minimal. Sections 3-5, which contained the glass beads for the attachment of 77 Figure 19. Influent and effluent substrate loading rate for pentachlorophenol. 78 fan _ 0.0.3.01.” 99.03—th +. L n0>00v 05:. can 00“ NDM Ohm . . . . 0.0...0nam 2.03.:— *5” . 0.0m 050004 _oco£Qo..o_r_oo.cmn_ I 0... I D.— I N I N.N I Q.N I 0.N I O.N I n .ma 0HSMHm nu (A00 40:! um .106 501) ”0.140an 79 Figure 20. Cumulative substrate in the influent and effluent for pentachlorophenol. 80 10M 0 c0: . .0 0 + . . .u . h a0>00v 05... ND” . .52.... .20... 000 000 050 050 n - n n D «- «- 0.0...0A:n 02.0....an —OC0£QOLO—£oo*cml .ON muawfim (5w) «nu-nus m0; 81 Figure 21. Location of the biomass in Bioreactor I. S0cfl0n Bioreacfor l l- I J l 1 MM” 3 .3 w § Km N I" ::::> 2 <3 0000 3 8800 960% 4 800 (X% 0 I 00 5 0000 O 042 l l 0.4 0.5' 0.0 - 1 Pr0f0ln (mg/ ml) 1.2. 83 the biomass, also contained very little protein, but at slightly higher levels than section 1 and 2. Attachment of the biomass to the glass beads was minimal. Sections 6-10 contained increasing amounts of biomass with the most significant amount of biomass located in section 9 and 10 near the screen designed for the accumulation of a sludge blanket. The distribution of protein was similar in Bioreactor III. (Figure 22). The majority of the protein was located at the bottom of the bioreactor near the screen. Methane production and a chlorophenol degradation assay were used to determine the location of active biomass in both bioreactors. Although each section of the bioreactor produced small amounts of methane, the sections near the sludge blanket, sections 9 and 10 for Bioreactor I and sections 4 and 5 for Bioreactor III produced significantly more methane (Figures 23 and 24). A chlorophenol degradation assay was also used to determine the dehalogenation activity in Bioreactor III (Figure 25). Chlorophenol was rapidly degraded in the bottom section of the bioreactor in both sections 4 and 5. No dissappearance of added chlorophenol over a 20 day period was observed for sections of the bioreactor above the sludge blanket, including the sections containing the glass beads. Dominant Morphology. The dominant morphology in the bioreactor was determined by microscopic examination of the bioreactor contents. Although not the sole organisms present, three organisms dominated the bioreactor contents. These were: 1. a putative phenol-degrader 2. a Methanosarcina sp. 3. a Methanothrix sp. 84 Figure 22. Location of the biomass in Bioreactor III. 85 Bioreactor Ill Figure 22. ‘g A Pr0t0ln (I00 9/ ml) 86 Figure 23. Methanogenic activity per column section in Bioreactor I. 87 Agaonv 05.... ON— . IL /O— O? m...— ION Ion IO! Ion Tom ION. Ion Too 100p b.)..0< 0300000502 _ Lo.oomco.m .mm ohswhg (urn) 000m”. 88 Figure 24. Methanogenic activity per column section in Bioreactor III. 89 Agnes 05.... Don . DON 00— P L b 3 \ 11w! IIWI 0.. F. '0‘ t.>..0< 07.000.53.02 .2 to.oomto.m 00 .qm muswam IO. ION ID” ID“ ON. (an) 000010" 90 Figure 25. Degradative activity per column section in Bioreactor III. 91 Du A0>00v 05? a; N— O— F r p . . r O . f 1T TO- m.\: In. ION IOM me. Or. >000< :0..0.00..u0O ___ to.ooo..o.m .nN 005000 an /—. - Oi (nu/6n) |0u0qdoa0|qg—0;0n 92 The morphology of the first organism, a putative phenol-degrader (Figure 26), was similar to that observed in other phenol-degrading enrichment cultures and has been isolated in co-culture from the phenol enrichments (Krumme and Boyd, 1986). The second microorganism was similar in morphology to Methanosarcina sp. (Figure 27), and the third organism was similar in morphology to Methanothrix sp. (Figure 28). Although present in chlorophenol enrichment cultures, these three organisms did not dominate the enrichment cultures to the extent that they dominated in the bioreactor. Numerical quantification was not possible due to the presence of particulate matter in both enrichment culture and bioreactor contents which obscured observations. 93 Figure 26. Photomicrograph of bioreactor biomass depicting morphology similar to a putative phenol-oxidizing microorganism (4000X). 94 95 Figure 27. Photomicrograph of bioreactor biomass depicting morphology similar to Methanosarcina Sp. (4000X). 96 97 Figure 28. Photomicrograph of bioreactor biomass depicting morphology similar to Methanothrix Sp. (4000X). 98 b - O ' 0“. 'u 'Q""-3' 43143" 3‘.“ K ,,-‘\‘", ', 0 [I.‘. DISCUSSION We have demonstrated with three separate bioreactors, that reductive dechlorinating activity can be maintained in anaerobic upflow bioreactors. The anaerobic conditions necessary for reductive dehalogenation were maintained with chlorophenols as the sole carbon and energy source. The bioreactors operated for 350, 400 and 190 days suggesting that reductive dechlorinating activity can be maintained indefinitely given the proper conditions. In addition to reductive dechlorination, mineralization of the chlorophenols to CH4 and 002 was demonstrated in the bioreactors. Approximately 40% of the added carbon was accounted for as gaseous end products. Gaseous products, CH4 and C02, may have accounted for the remaining (60%) of the carbon but was not recovered due to the difficulty in recovering and detecting small quantities of gaseous products from large effluent volumes over long time periods in an open flow-through system. Mineralization of the chlorophenol to CH4 and C02 was additional evidence that strict anaerobic conditions were maintained in the bioreactors. In the first two bioreactor experiments, we attempted to determine a range of feasible substrate loading rates and HRT's while maintaining a substrate conversion efficiency of greater than 90%. We were concerned with both washout of the biomass from short HRT's 100 and toxicity of the chlorophenol substrate from high loading rates. Short-term evidence, with Bioreactor I, suggested that a substrate loading rate of 60 mg 1‘1 d'1 at a HRT of two days was possible. During long-term mineralization studies, we demonstrated that substrate loading rates of 10 mg 1'1 d-1 at a HRT of four days could be maintained at 100% substrate conversion efficiency for nearly 100 days. With Bioreactor II, short-term evidence also suggested that substrate loading rates of 70 mg 1'1 d"1 at a HRT of two days and a substrate conversion efficiency of 100% was feasible. These loading rates were achieved using a mixture of all three monochlorophenols. Again, during the mineralization studies we demonstrated that loading rates of approximately 18 mg 1'1 d"1 at a HRT of approximately eight days and a substrate conversion efficiency of approximately 100% could be maintained for over 100 days. In order to more fully define maximum substrate loading rates at a specific HRT, experiments were performed with Bioreactor III, which utilized only meta-chlorophenol as the sole carbon source. Loading rates of 20 mg 1'1 d"1 at HRT's of two and four days were achieved at substrate conversion efficiencies of greater than 90%. However, when the loading rate was increased to 30 mg 1'1 d"1 at both retention times, the substrate conversion efficiency decreased to approximately 60%. It appeared that the biomass was adapted to degrade only 20 mg l'1 d'1 because the decrease in substrate conversion efficiency was equal to the difference between 20 and 30 mg l'1 d'l. Toxicity of the substrate at the higher loading rate did not appear to effect the biomass since it retained the ability to degrade the lower '1 d'l. substrate loading rate of 20 mg 1 In future experiments it 101 may be possible to achieve higher substrate loading rates in bioreactors by either allowing a sufficient period of time for the biomass to adapt to a higher substrate loading rate or by some other means of increasing the amount of active biomass in the bioreactor. We have demonstrated the ability to dehalogenate and degrade a mixture of all three monochlorophenols in anaerobic upflow bioreactors. Dehalogenating activity was maintained on a 1:4:1 mixture of ortho-:meta-zpara-chlorophenol for 400 days. These results suggest that anaerobic upflow bioreactors may be capable of treating more complex waste streams containing higher chlorinated phenols since dechlorination occurs at all positions on the aromatic ring. Batch culture experiments with acclimated anaerobic sewage sludge have demonstrated the ability to dechlorinate higher chlorinated phenols such as the dichlorophenols and pentachlorophenol. (Boyd and Shelton, 1984; Mikesell and Boyd, 1985). Additional experiments with anaerobic sewage sludge acclimated to all three monochlorophenols have demonstrated the ability to dechlorinate and partially mineralize pentachlorophenol (Mikesell and Boyd, 1986). Thus, the potential for dechlorinating higher chlorinated phenols exists. We have demonstrated the partial dissappearance of one higher chlorinated phenol, 3,4,S-trichlorophenol, in the bioreactor capable of dechlorinating all three monochlorophenols. The lack of dechlorination in the bioreactors of the other higher chlorinated phenols we tried, (i.e., 2,4,6-trichlorophenol and pentachlorophenol), may be the result of improper experimental conditions such as the loading rate or the HRT. 102 Anaerobic upflow bioreactors can be used as a tool in studying unique microorganisms and their activities. Bioreactors are an especially important tool when the microbial activity of interest is the result of interactions between members of microbial communities. When the microbes responsible for a unique activity are difficult to identify in a microbial community and isolate by conventional methods, such as those involved with the dechlorination and degradation of chlorinated phenols, bioreactor studies may be the only feasible method of studying these unique microbial activities. Anaerobic upflow bioreactors can be an alternative to conventional enrichment culture techniques where periodically portions of the enrichments are transferred into fresh media. The undesirable microbes and metabolic end-products are eliminated and the microbes of interest are selected. Conventional enrichment culture techniques may be lengthy due to slow growth of the organisms after transferring portions of the enrichment. Enrichment of the desirable microbial community can be achieved faster in anaerobic upflow bioreactors for two reasons. First, the majority of the microbial biomass remains in the bioreactor, usually in the form of a sludge blanket or attached to a solid support material. Second, enrichment is continual; substrate is continually added to the bioreactor while end-products are continually removed. An upflow bioreactor also provides a feasible method of enrichment when the substrate of interest is toxic to the microbes, as is the case with chlorophenols, because low concentrations of the substrate can be continually applied to the bioreactor. 103 Enrichment of a microbial community occurred with the bioreactors fed only chlorophenols as a carbon source. Although a dechlorinating microorganism could not be identified, three other microorganisms dominated the biomass. These were 1) a putative phenol-degrader, 2) a Methanosarcina sp. and 3) a Methanothrix sp. APPENDIX Isolation of Anaerobic Phenol and Benzoate-Oxidizing Bacteria Introduction Phenol and phenolic compounds are widely used industrial chemicals. They are also toxic compounds and have been placed on the Environmental Protection Agency list of 129 Priority Pollutants (Keith and Telliard, 1979). The anaerobic biodegradation of phenol and related compounds has been extensively studied with mixed cultures of bacteria in waste treatment processes and other environments (Tarvin and Buswell, 1934; Healy and Young, 1978; Holladay et al., 1978; Balba et al., 1979; Boyd and Shelton, 1984; Fedorak and Hrudey, 1984; Young and Rivera, 1985; Godsy et a1, 1986). However, there has been only one previous report of anaerobic phenol- oxidizing bacteria isolated in pure culture (Barik et a1, 1985). The microorganisms, in co-culture with the hydrogen-consumer Wolinella succinogenes, were able to anaerobically oxidize phenol and other aromatic compounds. Anaerobic phenol-oxidizing bacteria are presumed to be obligate syntrophs with hydrogen-consuming microorganisms when grown on phenol, because the free energy change,AG°' , for phenol fermentation is negative unless low partial pressures of hydrogen are maintained. Thus, isolation of phenol-oxidizing bacteria on phenol must be carried out in the presence of hydrogen-consuming microorganisms. 105 In the present study we describe the isolation of two bacteria in co- culture with the hydrogen consumer Desulfovibrio 611 from phenol- degrading enrichment cultures. Material and Methods Methanogenic Phenol Enrichment. A stable methanogenic phenol- degrading enrichment was established from anaerobic sewage sludge with phenol as the sole carbon and energy source. The enrichment was established over a two to three year period by periodically transferring 25-50% of the enrichment into fresh medium (RAMM, as previously described). Approximately 100-200 ug/ml phenol was added to the enrichment every two days. Sulfidogenic Enrichments. Sulfidogenic enrichments were established from the methanogenic phenol enrichment by transferring 25% of the methanogenic enrichment into media (RAMM) containing 2.5 mM 304- and either 2.5 mM phenol or 10 mM benzoate. The enrichment was also inoculated with 25% of a turbid culture of Desulfovibrio 611. The sulfidogenic enrichment was allowed to acclimate for approximately one month prior to use. Isolation. Serial dilutions of the sulfidogenic enrichments were inoculated into anaerobic roll tubes containing 5.0 ml RAMM, 5% clarified rumen fluid, 2% agar, 2.5 mM 804-, and either 2.5 mM phenol or 10 mM benzoate. The sulfidogen, 0.5 ml of a turbid culture, was also added to each tube to insure an adequate lawn of hydrogen consumers. Colonies were picked from the roll tubes after 8-12 weeks by a bent Pasteur pipette and transferred into 10 m1 RAMM containing 2.5 mM 504' and either 2.5 mM phenol or 10 mM benzoate. The cultures 106 were maintained by periodic transfers of 5% of the volume into fresh media. Strict anaerobic techniques were employed throughout. Incubations were at 37°C and in the dark. Characterization. The substrate range for the methanogenic phenol enrichment was determined by transferring 50 ml of active culture into triplicate serum bottles containing 5 mM of either phenol, benzoate, 4-hydroxybenzoate, phenylacetate or p-cresol. Methane production was periodically monitored. Fermentation balance and time course studies for the two isolates were performed in triplicate by transferring 5% of a turbid co-culture into RAMM containing 2.5 mM 804- and either 2.5 mM phenol or 10 mM benzoate. Periodic samples of the aqueous phase were taken and frozen until analysis. C emica nal s s. Phenol and benzoate were quantified by using a Waters high pressure liquid chromatography system consisting of a Model 6000A pump and a model 441 UV absorbance detector set at 280 nm. The sample injection valve (Rheodyne 7125) was fitted with a 20- ul loop. The analytical column was a Waters Radial PAK C-l8 cartridge held in a RCM-100 Radial Compression Module. Peak areas were measured with a Hewlett Packard 3390A integrator. The mobile phase was a 3:7 mixture of acetonitrile: 5% acetic acid to give retention times of approximately three minutes. The mobile phase flow rate was set at 2.0 ml/min. The samples were thawed and filtered through a 0.45 um HA Millipore filter before analysis. Acetate was quantified by using a PE-9000 gas chromatograph equipped with a flame ionization detector and a 2-m glass column (2 mm ID) packed with Carbopack C-0.3% Carbowax 20 M-0.l% H3P04. The 107 oven temperature was 90°C and the nitrogen carrier gas flow was set at 30 to 40 ml/min. Samples were acidified with formic acid (20% by volume) before injection. Methane was quantified by injecting 0.2 m1 of headspace gas into a Carle gas chromatograph equipped with a porapack Q column and a microthermistor detector. Protein was analyzed according to the procedure of Lowry et al., (1951). Results Phenol Enrichment. The phenol enrichment contained three dominant morphological types, tentatively identified by their morphology as: 1) a putative phenol-oxidizing bacterium, 2) a Methanothrix sp. and 3) a hydrogen-utilizing methanogen (Figures 29 and 30). The phenol-oxidizing bacterium was a Gram-negative oval shaped rod, 0.8 um by 1.2-3.3 um in length. A thin section electron micrograph of the phenol-oxidizing bacterium is shown in Figure 31. The methanogenic phenol enrichment was also able to degrade other aromatic substrates including benzoate, 4-hydroxybenzoate, phenylacetate and p-cresol as determined by the production of methane. The substrate p-cresol exhibited a longer lag time before the onset of methane production than the other aromatic substrates tested (Figure 32). Isolate PlI Isolate P1 was obtained on phenol from the sulfidogenic phenol enrichment. It was morphologically similar to the putative phenol-oxidizer found in the enrichments. P1 in co- culture with Desulfovibrio Gll degraded phenol to approximately equal molar amounts of acetate (Figure 33). The generation time was 108 Figure 29. Phase contrast photomicrograph of the methanogenic phenol-degrading enrichment depicting the proposed phenol-oxidizing microorganism (2000X). 109 Figure 30. 110 Phase contrast photomicrograph of the methanogenic phenol-degrading enrichment (2000K). The microorganisms found in the enrichment include: A) a proposed phenol- oxidizing bacterium, B) a Methanothrix sp., and C) a presumed hydrogen-utilizing methanogen. 111 112 Figure 31. Thin section electron micrograph of the proposed phenol- oxidizing microorganism taken from the methanogenic phenol-degrading enrichment culture (95,000X). 113 114 Figure 32. Methane production from various aromatic substrates in the phenol-degrading enrichment culture. 115 $050.... 02: com 00 . on. 00 3 P n p — .000.U-a 4 0.000000 0 0.0.000. >03... X .22... .r 0.003000 10-? 0 I —.O I NO I ”O I NO I 0.0 I GO r . I —.— 0030000.... 0:00:02 N.— HZMZIUEZM JOZwIn. .235. ("I“) ammuw 'IViOi 116 Figure 33. Phenol degradation by isolate P1 in co-culture with Desulfovibrio Gll. 117 uh