o I ’ b4. 1 l 2 v. p h. I. L u-..... u- n . K . u . . u,‘ r . It. I 3’1 u l 44. n h J -1\ 1 F v‘vvw I \ r J.» « ¢ » or t 5. ‘ 1|. 2 a \ . A . o A a ‘ \ . a v . .. x 4 1 1 c . l . a u . L x . , t . v u A? v .A v 3. vl \ d o A I .1 . . I Its ‘ .yL. » u . .l ‘l A. ‘ I l 1" l 6 V ‘t .0 I. l I. )0 I ucfl - Ir . I Nut 4 In». . . o ‘ I . L . _ - t x n z o .t k? t . q > ‘ )1 on o. 2 t v 0 . ' ‘ V \ I ‘ b O". n I. ~ I I V ‘ I - ‘ v I. l A n I 4 .4 1% 1 l I a I .v o c . A l. I _ 12!! do .- \ VI. ! \ D D! u u 0 0 .o 1‘. n8 . firm. IhD‘ I v . . L... 1 ’I. N Jaw 1- | O u I Q I . I | not. v ~ .fil‘ .4 WM" —% to 00' I i ‘ ‘ ‘ Q s - . 'I ‘ ‘5' I I . ,t .‘I‘: t .V ' t. . V ‘1‘ ‘I‘ .1! .. E. 14.111.315 . All!!- o. Avv. I..- l! . gut-$5.31. . a u n to I ilk? .0 .a ~Q. ‘. C '1 O ‘5 ‘1' 'C n I.‘ ‘ ¢ 1 I ‘ o \ I u . .- 'Il‘mfi‘l Q. o v v Q U..dvt\\ cu...“ . oVflA.‘ 1|. |!l.v.|!¢v.|! .. .1... 11211 Li. -1: ofoKIIllllflbhhqltfllflNUnflnvhunn? 9.11% I i- 307'. 3.1.! v‘:|;nnv|iql n5~.uul .1 . .o v.... . n..u.‘g. .l\ ultlevQ¢ :- . .w rill-cu” .s . . x. t... «I. . £59.11...‘ - . -31.. “In”? 1.: air . .7. ..R...Pt.n..l... , . - ”.11. 4‘ | $ A . v ‘11 v.“ c u c . ufi I C I . n . 1 . .-v .I .x. ;... . v ‘. A, 0' ..A. '1!" II‘ 1,.. . v. a. . . $3.1M ' 3 l 1 1l ‘ . ‘stzgm $11 3‘1: M Villt." '\ 3. I‘ ' L . . '., ,|_\ . ‘-{'l‘~ l':1 £ .,5 3| a. I Er." W‘: l. '- ' I“ I . *3.‘ ‘,.“ '3 it .- THESIS This is to certify that the dissertation entitled MICROBIAL TRANSFORMATIONS 0F ACETANILIDE HERBICIDES presented by LINDA LEIGH McGAHEN has been accepted towards fulfillment of the requirements for PH.D. degreein CROP s SOIL SCIENCES JAMES Tl EDJE Major professor 5-21-82 Date MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES “ BETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be Charged if book is returned after the date stamped below. ‘— ‘MICROBIAL TRANSFORMATIONS.OF ACETANILIDE HERBICIDES BY Linda Leigh McGahen A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of ' DOCTOR OF PHILOSOPHY Department of Crop and Soil Sciences 1982 ABSTRACT MICROBIAL TRANSFORMATIONS OF ACETANILIDE HERBICIDES By Linda Leigh McGahen The release of xenobiotics into the environment has elicited a great deal of concern about their biological and chemical transfor- mation due to the toxicity of many of them or of their transformation products. Research on the biodegradation of xenobiotics under aerobic conditions is common (reviewed in Chapter 3). However, exposure to anaerobic environments is possible whenever the rate of diffusion .of oxygen into a site is less then the rate of respiration, which is the case for most sediments in streams, lakes, and estuaries, or a flooded or poorly drained field. I chose two acetanilide herbicides as model compounds to study their biodegradation and disappearance under anaerobic conditions, diethatyl [2-chloro:§f(2',6'-diethylphenyl)fi§fmethyl(ethylcarboxylate) acetamide] and metolachlor [2-chlorojgf(2'-ethyl-6'-methy1pheny1)- ‘Ef(deethoxy-l-methylethy1)acetamide]. Sediment from a eutrophic lake was chosen as a biologically active anaerobic system to study the transformation of the herbicides. A simple, inexpensive two- 14 step procedure was developed to extract and clean-up the ~C-labeled herbicide samples from the lake sediment. After eight weeks incubation in lake sediment under anaerobic 14C-ring label from conditions, approximately equal amounts of the diethatyl (452) and metolachlor (41%) was extracted from nonsterile sediment; the remainder was apparently bound through biological Linda Leigh MbGahen mechanisms or strongly adsorbed in the sediment. Diethatyl was not detectable in the extract whereas 34% of the metolachlor remained; predicted half-lives were 1.2 - 1.5 weeks (diethatyl), and 3.7 - 6.0 weeks Cmetolachlor). However, the time for disappearance of the 140- labeled rings were comparable (6.6 weeks, diethatyl; 5.6 weeks, metola- chlor). Also, degradation rates under aerobic or anaerobic conditions are comparable to rates found under aerobic conditions. Two nonpolar transformation products of both herbicides were identified. Both herbicides were transformed by simple reductive 'dechlorination of the chloroacetyl moiety with the addition of a proton, or reductive dechlorination with addition of a thiomethyl group. This is the first time a thiomethyl product has been reported for this class of herbicides under aerobic or anaerobic conditions. CODE FOR.MICROBIOLOGICAL BEHAVIOR TI-DU SHAL'I‘ NOT DEGRADE THAT FOR WHICH THOU' HAS NO APPETITE. THOU SHALT NOT DEGRADE THAT WHICH THOU CANNOT TAKE INTO THY BELLY. THOU SHALT N01" DEGRADE'. THOSE COMPOUNDS WHICH WILL DO THEE HARM. THOU SHALT NOT DEGRADE THAT WHICH IS INACCESSIBLE. THDU SHALT NOT DEGRADE THAT WHICH IS FOUND IN A HDSTILE ENVIRONMENT. ii ACKNOWLEDGEMENTS I would like to thank Dr. James Tiedje for his enthusiasm, patience, friendship, and his support and knowledgeable guidance during my years at Michigan State University. I also wish to express my sincere thanks to my guidance committee, each one being helpful and supportive. These Professors, besides Jim Tiedje, are Dr. Fumio Matsumura, Dr. Donald Penner, and Dr. Matthew Zabik. I also appreciate the support and encouragement of Dr. Arthur WClcott (Frofessor Emeritus) and Dr. Frank Dazzo. Lastly, without the chemical ionization and high-resolution mass spectromety data pro- vided by Dr. Robert Minard (Penn State) and the electron ionization mass spectromety data provided by the Mass Spectrometry facility at 'Michigan State, this work would not have been possible. My sister, Carol McGahen, deserves much credit and praise for her excellent, patient typing of this dissertation and its revisions while my parents.deserve a commendation for providing me with their endless support. Finally, my studies were supported throughout by a graduate research assistantship from'Michigan State University, for which I am very grateful. iii LIST OF TABLES LIST OF FIGURES TABLE OF CONTENTS INTRODUCTION AND EXPERIMENTAL OBJECTIVES . . . CHAPTER I. CHAPTER II. CHAPTER III. APPENDIX A APPENDIX B LITERATURE CITED . . . . . . . . . METHODS FOR THE EXTRACTION, CLEANeUP, AND ANALYSIS OF HERBICIDES INCUBATED IN LAKE smmm O O O O O I O O O O O O 0 MATERIALS AND METHODS . . . . . . RESULTS . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . LITERATURE CITED . . . . . . . . . ANAEROBIC METABOLISM OF TWO ACETANILIDE HERBI- CIDES IN A LAKE SEDIMENT . . . . . MATERIALS AND METHODS . . . . . . . RESULTS . . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . LITERATURE CITED . . . . . .7. . . MICROBIAL METABOLISM OF ACETANILIDE HERBICIDES UNDER AEROBIC CONDITIONS -- A REVIEW. . . . . GENERAL METHODS . . . . . . . . . . ACETANILIDE METABOLITES . . . . . . LITERATURE CITED . . . . . . . . . EXPERIMENTS OR RESULTS WHICH SUPPORT THE RESEARCH IN CHAPTERS 1 AND 2 . . . ANALYSIS OF POLAR.METABOLITES FORMED FROM DIETHATYL OR.METOLACHLOR BY CHAETOMIUM GLOBOSUM DURING RESTING CELL EXPERIMENTS . . MATERIALS AND METHODS . . . . . . . RESULTS AND DISCUSSION . . . . . iv Page vii ya 10 13 14 24 42 53 57 59 68 94 112 117 118 119 137 138 152 152 155 Table CHAPTER I CHAPTER II LIST OF TABLES page Characterization of sediment, soils, and sludge used, and percent solids used in experiments . . . 15 Properties of herbicides used. . . . . . . . . . . 13 Summary of experimental set-up for all extraction experiments . . . . . . . . . . . . . . 23 Adsorption by and extractability of herbicides8 in sediment or soils . . . . . . . . . . . . . . . 29 Loading capacity of a 20% acetonitrile extractR of herbicide—amended sediment on a Sep-Pak €13 cartridge . . . C O . . C . . . C C . . . . C C O 40 Summary of previous extraction methods and re- coveries of various herbicides and insecticides from soils and sediments . . . . . . . . . . . . . 43 Disappearance of diethatyl and metolachlor in sterile and nonsterile sediment . . . . . . . . . 73 Least squares fit of data of diethatyl and metolachlor disappearance to an exponential curvea: prediction of parent herbicide and ring- label half-life under anaerobic conditions . . . . 81 Mass spectral data (EI) for diethatyl metabolites D1 and D2 and metolachlor metabo- lites m and M2 0 O O O O O O O O O O O O O O O O 84 Isotopic distributions calculated by the Kratos MS 9/50, and from EI and CI MS data . . . . . . . 37 High resolution mass spectral data: calculations of molecular formulas for parent and fragment ions 0 O O O O O O I I O O I O O O O O O O O O 0 O 90 Table CHAPTER III APPENDIX A Page Summary of evidence.for degradation of several acyl-aor acetanilide compounds by microorgan- ism. O O O O O O I C O O O O O O O O O O O O O O 133 R . Efficiency of Sep-Pak C18 separation versus hexane extraction of a 202 acetonitrile extract of diethatyl from sediment and soils. The effect of sample filtration prior to separation of extraction is also reported . . . . . . . . . . . 142 vi LIST OF FIGURES Figure Page CHAPTER 1 1 Separation of nonEolar compounds from solutions using Sep-Pak 018 cartridges-. . . . . . . . . . . 21 2 'Adsorption of diethatyl and metolachlor by sediment . O O O C C C C C C C O C O C C O Q C I O 25 '3 Effectiveness of extraction of 50 ppm diethatyl (H) and.12.5 ppm metolachlor (H) by different concentrations of acetonitrile. . . . . . 27 4 HPLC analysis of diethatyl and metabolites (top), or metolachlor and metabolites (bottom) formed during anaerobic incubation in sediment (Chapter II) . . . ... . . . . . . . . . . . . . . 31 5 _ HPLC chromatograms of diethatyl (top row) and ‘metolachlor (bottom row) extracted from;f (l) ' 0.02 M phosphate buffer (50 ppm diethatyl; 12.5 ppm metolachlor); (2) sediment; (3) sandy loam soil; (4) sludge; and (5) muck soil . . . . . . . . 34 6 HPLC chromatograms of cyanazine (top row) and chloramben (bottom row) extracted from: (1) 0.02 M phosphate buffer (cyanazine only); (2) sediment; (3) muck soil; or (4) sandy loam soil . . 35 7 HPLC chromatograms of chloramben extracted from sludge as described in Materials and Methods. Chromatograms are: (l) methanol phase and (2) water phase of a 202 acetonitrile extract and (3) methanol phase and (4) water phase of a water extracted control . . . . . . . . . . . . . . . . . 36 CHAPTER II 1 Flow diagram of preparation of serum bottles containing sediment for long-term anaerobic incubation experiments . . . . . . . . . . . . . . 64 2 Flow diagram of extraction and analysis procedure . . . . . . . . . . . . . . . . . . . . . 68 vii Figure 3a &'3b 10 CHAPTER III Inhibition of methane production in sediment incubated with various concentrations of (a) diethatyl or (b) metolachlor . . . . . . . . Percent methane (v/v) produced after four weeks incubation versus diethatyl (H) or metola- chlor (H) concentration . . . . . . . . . . Disappearance of 14C-ring labeled diethatyl incubated in sterile and nonsterile sediment under anaerobic conditions . . . .:. . . . . . . Gas-liquid chromatograms of extracts (see Materials and Methods) from sediment amended 'with diethatyl, metolachlor,_or water (control) after eight weeks incubation . . . . . . . . . . Highéperformance liquid chromatographic (HPLC) analysis of extracts (see Materials and Methods) of sediment amended with diethatyl or water (control) after eight weeks incubation . . . . . Disappearance of 14C-ring labeled metolachlor incubated in sterile and non-sterile sediment under anaerobic conditions . . . . . . . . . . .‘ High-performance liquid chromatographic (HPLC) analysis of extracts (see Materials and Methods) of sediment amended with metolachlor after eight weeks incubation . . . . . . . . . . . . . . . . Structures of diethatyl and metolachlor and their metabolites D1 and D2, and M1 and M2 . . . . . Degradation of 140 ring- and carbonyl-labeled diethatyl by g. globosum in resting cell experi- ment 8 O O I O O O O O O O O O O O O I O I O O 0 Degradation of metolachlor by resting cells of £0 810bo 8m 0 O O O O O O O O O O O O O I O O O A.comparison of the similarity of products reported by several authors (8, 13, 5, l) for fungal metabolism for several acylanilides . . viii Page 70 71 74 75 76 78 8O 88 120 121 123 Figure APPENDIX A 1 2a 2b 3c APPENDIX B 1 chomparison of other products from fungal metabolism of the indicated herbicide (see Figure 3 for additional details) . . . . . . . . Several proposed metolachlor degradation products and alternative structures consistent ‘with the mass spectra and reasonable metabolic pathways . . . . . . . . . . . . . . . . . . . . Decrease in pH in normal and autoclaved sediment, 0.02 M phosphate buffer, and distilled, deionized water by increasing flow rate of 002 . . . . . . Gas-liquid chromatographic analysis of diethatyl and metabolites produced under anaerobic conditions in sediment (Chapter 2) . . . . . . . . . . . . Gas-liquid chromatographic analysis of metolachlor and metabolites produced under anaerobic conditions in sediment . . . . . . . . Possible fragmentation scheme for diethatyl metabolite D1 . . . . . . . . . . . . . . . . . Possible fragmentation scheme for diethatyl metabolite D2 . . . . . . . . . . . . . . . . . Possible fragmentation scheme for metolachlor metab01ite M1 0 O O O O O O O O O O O O O O O 0 Possible fragmentation scheme for metolachlor metabOIite M2 0 I O O I O I O O O O O O O O O O Temperature-programmed gas-liquid chromatographic analysis of derivatized (BSTFA) polar metabolites formed from diethatyl by Chaetomium globosum . . Temperature-programmed gas-liquid chromatographic analysis of derivatized (BSTFA) polar metabolites formed from metolachlor by Cheetomium globosum . ix Page 126 128 139 145 146 148 149 150 151 158 159 Figure Page 3 High-performance liquid chromatographic analysis of underivatized polar metabolites formed from.diethatyl by ChaetOmium Elabosm I O O O O O O O O O I I O O O O O O I O O 4 High-performance liquid chromatographic analysis of underivatized.polar metabolites formed from metolachlor by ChaetOmiumiglobosum . . . . . . . . 162 161 INTRODUCTION AND EXPERIMENTAL OBJECTIVES Concern about contamination of the environment by xenobiotics has resulted in a great deal of research towards assessing their fate, and the passage of the Toxic Substances Control Act which mandates rational and systematic evaluation of chemicals introduced into commerce. Environmental hazard assessment involves endeavor in many areas, from biomagnification and residue studies in organisms to research on photodecomposition. Another aspect of major importance is the biodegradability of the substance by microorganisms. So far, most research on the biodegradation of xenobiotics has been performed under aerobic conditions. For example, in the compendium Microbial Transformations of Non-Steroid Compounds, author Kieslich (1976) cites 1932 references, of which only 25 deal with transformations occurring under anaerobic conditions. That relatively so little research focus' on the fate of xenobiotics under anaerobic conditions may be explained by the fact that methods providing anaerobic conditions are rather difficult, expensive and foreign to most investigators [methods were developed in 1950 by Hungate (1950), and have been improved by Bryant and Robinson (1961) and Macy et a1. (1972)]. Secondly, the importance of anaerobic habitats and the potential for biodegradation under anaerobic conditions has not been recognized. However, anaerobic conditions exist wherever the diffusion of 0 is less than the rate 2 at which it is metabolized. For example, flooded or even moist soils and lake sediments are frequently anaerobic or have anaerobic macro and micro sites. Therefore, for example, pesticides which were applied to fields that become flooded or which are adsorbed onto organic matter and clay particulates and carried with run-off into stream and lake sediments would be exposed to anaerobic conditions. Williams (1977) reviewed the transformation of pesticides under anaerobic conditions. Of 161 examples from the 306 references, 29% of the studies used rumen fluid, 19% were done with pure cultures or isolates from anaerobic environments, 16% used "flooded", "submerged", or "rice" soils, 10% used gut microflora or feces, 4% each used sludge or lake sediment, and 2% used mixed cultures. It is interesting that so much work has been done with pure cultures when anaerobic degradation often requires several interacting species which makes metabolism more thermodynamically favorable. For aerobic organisms, such interaction is not necessary. Under anaerobic conditions, metabolism can be optimized by interspecies hydrogen transfer (Zehnder, 1978) which involves a coupled fermentation between a heterotroph, which ferments a substrate, producing C02 + H2, and a methanogen (or sulfate reducer), which utilizes the H2. The methanogenic step "pulls" along the fermentation of the substrate (degradation step). Complete fermentation of various compounds to and H CO by methanogenic consortia has been well documented. 2 2 Methanogenic fermentation of benzoate is included in a review by Evans (1977). In all cases, the aromatic nucleus is reduced (dearomatized) before ring cleavage and metabolism to various possible fatty acids (butyrate, propionate, acetate) and 002. Recently, Healy and Young (1979) have characterized the anaerobic biodegradation to methane of vanillin, vanillic acid, ferulic acid, cinnamic acid, benzoic acid, catechol, protocatechuic acid, phenol, pfhydroxybenzoic acid, syringic acid and syringaldehyde. Decomposition of ferulic acid to methane was found to occur via reductive (dearomatization) pathway quite similar to that reported for benzoate (Healy et al., 1980). Similarly, Evans (1977) notes in his review, Rhodopseudommonas palustris grew photosynthetically under anaerobic conditions on benzoate and a denitrifying Pseudomonas sp. grew on benzoate using nitrate as an electron acceptor. In both cases the benzene nucleus was reduced prior to cleavage. Williams (1977) and Sethunathan (1978) reviewed the degradation of pesticides by anaerobic microorganisms from various systems (rumen, sediment, flooded soils, gut microflora, pure culture, etc.) and in rice soils, respectively. Various characteristic reactions occur under anaerobic conditions such as dehalogenation, hydrolysis of phosphate and aliphatic esters and amides, reduction of nitro and sulfate groups to amines and sulfide, and dealkylation of ye and ‘Qfalkyl compounds. Two other unusual reactions have been noted. Bollag and Russel (1976) found that 2,4-dichloroaniline was acetylated by a denitrifying Paracoccus, and Marthy and Kaufman (1978) found that the nitro group on the fungicide pfchloronitrobenzene was replaced with a thiomethyl group under anaerobic conditions. Polar products have frequently been found but not identified, and, in soil systems, degradation products may become bound. However, neither 14C02 nor 14CH4 were evolved from 14C-ring-labeled herbicides in any anaerobic system. Complete degradation of the substituted ring apparently does not occur. As with many aerobic transformations of pesticides, biodegradative reactions may be due to "cometabolism", which is probably due to the compound serving as a secondary substrate for a nonspecific enzyme. Many reports [reviewed by Williams (1977) and Sethunathan (1978)] indicated that some reactions such as dechlorination may be catalyzed chemically. Glass (1972) proposed that dehalogenation of DDT to TDE could be due to ferrous ion (Fe+2), although Zoro et a1. (1974) disagree that the reaction would be as quick or extensive as proposed. Data from.Zoro and coeworkers (1974) instead support that of Miskus et al. (1965) which suggests that reduced iron porphyrin systems can nonenzymatically reductively dechlorinate compounds such as DDT 3 and toxaphene. This was further substantiated by Khalifa et al. (1976) and Saleh and Casida (1978). Reduced cytochrome P450 has also been found to nonenzymatically dechlorinate DDT, toxaphene. mexacarbate, parathion, dieldrin and lindane (Matsumura and Esaac, 1979 and 1978; Esaac and Matsumura, 1980; Khalifa et al., 1976; Saleh and Casida, 1978). ,Reduction of nitro groups and hydrolysis of some pesticides may also be catalyzed chemically. Recently, wahid et a1. (1980) reported that the nitro moiety of parathion is "instantly" (within 5 3) reduced to an amine when parathion is added to a soil which has been prereduced by flooding for 60 days. Interestingly, no reaction occurred in prereduced soils sterilized by autoclaving, although it still occurred in soils sterilized by irradiation or sodium azide. Since the latter two sterilization methods may not inactivate enzymes as the former method (autoclaving) would, the reaction may be due to soil enzymes. Although degradation rates and pesticide half-lives are not usually calculated by researchers, Sethunathan (1978) extrapolated data from various studies to present the relative stability of various insecticides, herbicides, and fungicides in anaerobic and aerobic soils. Under anaerobic conditions, the half-lives of 22 of the 34 examples were between 4 and 25 days; there was no degradation of three chlorinated insecticides, (chlordane, dieldrin and one endrin sample; endrin.was degraded in the two other studies). Uhder aerobic conditions, the half-lives of 15 of the 34 examples were between 9 and 50 days, while there was no degradation of seven of the chlorinated insecticides (BHC, DDT, chlordane, dieldrin, endrin, or parathion) and two fungicides (PCNB and DCNA). (There were 13 insecticides, 13 herbicides and four fungicides; three pesticides were used in two or three different studies, totaling 34 samples.) The longest average half-life calculated under anaerobic conditions was 105 days (for picloram), although under aerobic conditions, trifluralin and picloram had half-lives of 150 and 135 days, respectively.* Although the compounds used for this comparison of half-lives are somewhat representative of those which have been studied for their biodegradability under anaerobic conditions, they are not necessarily representative of those compounds currently widely used, at least in the United States (DDT, dieldrin, methoxychlor, chlordane, aldrin, and endrin, for example, are not used). It seems that the potential for biodegradation under anaerobic conditions may occur as readily as it does under aerobic conditions and that the rates of degradation are of the same order of magnitude. However, the extent of biodegradation in the environment may be more limited under anaerobic conditions than under aerobic, although binding of products to soil components (organic matter?) appears to occur readily in anaerobic environments (Golab et al., 1979; Probst et al., 1967; Golab et al., 1970; Takase et al., 1978; Roberts and Standen, 1978; Murthy and Kaufman, 1978; Ambrosi et al., 1977; Belling and Kivonak, 1978; Wang and Broadbent, 1973). A greater degree of decomposition in soils is frequently thought to be associated with higher organic matter content or with amendments *However, my estimation of trifluralin half-life from data by Probst et a1. (1967) is 17 days under aerobic field conditions and 95 days under "dry", aerobic (50% field capacity) conditions. of organic matter. Actually, reports vary. No increase in decompo- sition was found in soils incubated anaerobically or flooded and amended with microcrystalline cellulose powder (PCNB; Murthy and Kaufman, 1978) or alfalfa (dieldrin, endrin, DDT, DDD; Guenzi et al., 1971). Amendments of air dried and ground alfalfa meal, rice straw or rice hulls (DDT; Parr et a1. 1970) also did not increase decompo- sition in anaerobic or flooded soils. However, other studies have noted an increase in decomposition in anaerobically incubated or flooded soil amended with rice straw (endrin; Beard, 1968); glucose, alfalfa, or rice straw (PCNB, DCNA; wang and Broadbent, 1973); and glucose only -- not alfalfa or rice straw -- (DDT; Parr et a1. 1970). In some cases, degradation under anaerobic conditions is enhanced in soils which have a higher organic matter content vs. those which are lower in organic matter [parathion, Sethunathan (1973); and oxadiazon, Ambrosi et a1. (1977)]. The effect of the organic matter may be to "increase microbial activity and hasten the Eh drop" (Sethunathan, 1978). It may also increase adsorption and binding of xenobiotics or their degradation products. In a flooded soil system, volatilization of ethalfluralin and trifluralin was decreased compared to that in soil which was air dried or at field capacity, although for fluchloralin volatilization was decreased by soil which was air dried or flooded, compared to high volatilization for soil at field capacity (Savage, 1978). Therefore, in a flooded soil system, volatilization is decreased compared to a moist and/or air dried soil and partitioning into or onto the clay and organic matter becomes a major fate, especially for less polar compounds. A significant portion of the pesticides in anaerobic environments would be expected to be associated with the particulates. Therefore, even if the water solubility of a pesticide were fairly high (500 ppm, e.g.) it may readily partition into the clay/organic matter present depending on the partition coefficient of the compound, its ionization (if any), and the type, state, or charge of the organic matter present. Anaerobic systems tend to have a fairly neutral pH which is frequently (in soils) buffered by ferrous (Fe+2) ion. Since substituents such as nitro and sulfate groups on organic compounds tend to be reduced to amines and sulfides under anaerobic conditions, ionic interactions of compounds with organic matter in anaerobic systems may be different than in more oxidized and perhaps negatively charged aerobic systems. My research centered on the following questions: 1. Can a simple method be developed for the efficient extraction, clean-up and analysis of the acetanilide herbicides from anaerobic lake sediment? What are the usual methods of extraction from sediment or soil and can a procedure be developed which quickly and efficiently extracts the come pounds, minimizes clean-up time, and directly prepares the samples for analysis by high-performance liquid chromatography? 2. What is the potential use of the clean-up method for samples of various sizes? What is the feasibility of using it for other herbicides and in other habitats (sludge and flooded soils)? 3. Uhder anaerobic conditions, are the acetanilide herbicides degraded and to what extent is this biological versus chemical? 4. What are the products formed —- if the herbicides are degraded -- under anaerobic conditions? 5. What is the rate at which herbicide transformations occur? 6. Do the herbicide transformation products become bound in the sediment? Representative compounds from the acetanilide group were chosen because these herbicides are widely used in corn and soybeans (alachlor, metolachlor, diethatyl, propachlor), and in rice soils (butachlor). Basic research on their biodegradation under aerobic or anaerobic conditions is of considerable interest. Due to their widespread use, they could be exposed to aerobic or anaerobic conditions in soils, or be carried into stream or lake sediments. 10. 11. 12. LITERATURE CITED Ambrosi, D., P. C. Kearney, and J. A. Macchia. 1977. Persistence and metabolism of oxadiazon in soils. J. Agric. Food Chem. 25: 868-872. Bollag, J.-M., and S. Russel. 1976. Aerobic versus anaerobic metabolism of halogenated anilines by a Paracoccus sp. Microbial Ecol. 3:65-73. Bryant, M. P., and I. M. Robinson. 1961. An improved nonselective culture medium for ruminal bacteria and its use in determining diurnal variation in numbers of bacteria in the rumen. J. Dairy Sci. 4451446-1456. Esaac, E. G. and F. Matsumura. 1978. A novel reductive system involving flavoprotein in the rat intestine. Bull. Environ. Cont. Toxic. .12‘15'22° Essac, E. G., and Matsumura. 1980. Mechanisms of reductive dechlorination of DDT by rat liver microsomes. Pest. Biochem. Evans, W. C. 1977. Biochemistry of the bacterial catabolism of aromatic compounds in anaerobic environments. Nature 270:17-22. Glass, B. L. 1972. Relation between the degradation of DDT and the iron redox system in soils. J. Agric. Food. Chem. 295324-327. Golab, T., R. J. Herberg, J. V. Gramlich, A. P. Raun, G. W. Probst. 1970. Fate of benefin in soils, plants, artificial rumen fluid, and the ruminant animal. J. Agric. Food Chem. .1§:838-844. Golab, T., w. A. Althaus, and H. L. Wooten. 1979. Fate of [14c], trifluralin in soil. J. Agric. Food Chem. .21:163-179. Gowda, T. K. 8., and N. Sethunathan. 1976. Persistence of endrin in Indian rice soils under flooded conditions. J. Agric. Food Guenzi, W. D., and W. E. Beard. 1968. Anaerobic conversion of DDT to DDD and the aerobic stability of DDT in soil. Soil Sci. Amer. Proc. .32:522-524. Guenzi, W. D., W. E. Beard, and F. G. Viets, Jr. 1971. Influence of soil treatment on persistence of six chlorinated hydrocarbon insecticides in the field. Soil Sci. Soc. Am. Proc. 225910-913. 10 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 11 Healy, J. B., Jr., and L. Y. Young. 1979. Anaerobic biodegradation of eleven aromatic compounds to methane. Appl. Environ. Microbiol. 38:84-89. Healy, J. B., L. Y. Young, and M. Reinhard. 1980. Methanogenic decomposition of ferulic acid, a model lignin derivative. Appl. Environ. Microb. 32;436-444. Belling, C. S., and A. E. Krivonak. 1978. Biological character- . istics of bound dinitroaniline herbicides in soils. J. Agric. Food Chem. '26:1164-1172. Hungate, R. E. 1950. The anaerobic mesophilic cellulolytic bacteria. Bacteriol. Rev. .14:1-49. Khalifa, 8., R. L. Holmstead, and J. E. Casida. 1976. Toxaphene degradation by iron (II) protoporphyrin systems. J. Agric. Food Chem. .25‘277'282- Kieslich, K. 1976. 1976. Microbial transformations of non-steroid cyclic compounds. Georg Thieme Verlag, Germany. 1262 pp. Macy, J. M., T. E. Snellen, and R. E. Hungate. 1972. Use of syringe methods for anaerobiosis. J. Clin. Nutr. ‘2§:1318-1323. Matsumura, F., and E. G. Esaac. 1979. Degradation of pesticides by algae and aquatic microorganisms. Pages 371-387 $£.M' A. Q. Khan, J. J. Lech, and J. J. Menn, eds. Pesticide and enobiotic metabolism in aquatic organisms. ACS Symposium Series, v. 99, washington, D.C. Miskus, R. P., D. P. Blair, and J. E. Casida. 1965. Conversion of DDT to DDD by bovine rumen fluid, lake water, and reduced porphyrins. J. Agric. Food Chem. .1§;481-483. Murthy, N. B. K., and D. D. Kaufman. 1978. Degradation of pentachloronitrobenzene (PCNB) in anaerobic soils. J. Agric. Food Chem. “26:1151-1156. Parr, J. F., G. H. Willis, and S. Smith. 1970. Soil anaerobiosis: II. Effect of selected environments and energy sources on the degradation of DDT. Soil Sci. 110:306-312. Probst, G. W., T. Golab, R. J. Herberg, F. J. Holzer, S. J. Parka, C. vander Schans, and J. B. Tepe. 1967. Fate of trifluralin in soil and plants. J. Agric. Food Chan. ‘1§:592-599. 25. 26. 27. 28. 29. 30. 31. 32. 33. 35. 12 Roberts, R. R., and M. E. Standen. 1978. Degradation of the herbicide flamprop-methyl in soil under anaerobic conditions. Pest Biochem. Physiol. .g:322-333. Saleh, M. A., and J. E. Casida. 1978. Reductive dechlorination of the toxaphene component 2,2,5-endo, 6fgxg, 8,9,10-hepta- chlorobornane in various chemical, photochemical, and metabolic systems. J. Agric. Food Chem. .2§:583-590. Savage, K. E. 1978. Persistence of several dinitroaniline herbicides as affected by soil moisture. weed Sci. 265465-471. Sethunathan, N. 1973. Degradation of parathion in flooded acid soils. J. Agric. Food Chem. .213602-604. Sethunathan, N., and R. Siddaramappa. 1978. Microbial degradation of pesticides in rice soils. Pages 479-497 in Soils and rice. International Rice Research Institute, Los Banos, Philippines. Takase, I., T. Nakahara, and K. Ishizuka. 1978. Degradation of 3-(3-chfioro-4-chlorodifluoro-methylthiophenyl)-1,1-dimethy1urea (Clearcide ) in paddy soils. J. Pesticide Sci. §;9-19. Wahid, P. A., C. Ramakrishna, and N. Sethunathan. 1980. Instan- taneous degradation of parathion in anaerobic soils. J. Environ. Qual. ‘2:127-130. Wang, C. H., and T. E. Broadbent. 1973. Effect of soil treatments on losses of two chloronitrobenzene fungicides. J. Environ. Qual. 5511-515. Williams, P. P. 1977. Metabolism of synthetic organic pesticides by anaerobic microorganisms. Residue Reviews. .99363‘135- Zehnder, A. J. B. 1978. Ecology of methane formation. Pages 349'376.$E.R° Mitchell, ed., water Pollution Microbiology. vol. 2. John Wiley and Sons, Inc. Zoro, J. A., J. M. Hunter, G. Eglinton, G. C. ware. 1974. Degradation of pjpf-DDT in reducing environments. Nature 247: 235-237. CHAPTER I METHODS FOR THE EXTRACTION, CLEAN-UP, AND ANALYSIS OF HERBICIDES INCUBATED IN LAKE SEDIMENT Extraction and clean-up procedures for compounds added to soils or sediments are quite varied. Polar solvents alone or in combi- nation with nonpolar solvents are frequently used for extraction. Clean-up may consist of partitioning a compound extracted with a polar solvent into a nonpolar solvent, and/or clean-up on an alumina or florisil column (see discussion). These extraction, partitioning and clean-up procedures are frequently time-consuming and nonpolar solvents, especially, may foam or emulsify when used as extractants, adding to the time and difficulty of sample preparation. I developed a two-step procedure to (1) simply and efficiently extract the herbicides from a lake sediment, and (2) subsequently clean-up and separate nonpolar and polar compounds in the extract. The compounds were initially extracted from the sediment (952 water, 52 solids) by the addition of acetonitrile to a final concentration of approximately 20% acetonitrile (v/v). The solution was centri- fuged and a portion of the supernatant (20% acetonitrile extract), 18R was added to a Sep-Pak C cartridge (waters Assoc., Milford, MA), which contains an octadecylsilyl (018) microparticulate reverse-phase R packing. Compounds in the extract are separated by the Sep-Pak C18 cartridge on the basis of their affinity for the 018 packing.‘ Essen- tially, nonpolar compounds partition into the cartridge and polar 13 14 compounds pass through, remaining in the 20% acetonitrile ("water") phase. NOnpolar compounds are subsequently eluted with a less polar, organic solvent such as methanol ("methanol phase"). The water and methanol phases are clean enough to directly inject into a high- performance liquid chromatography (HPLC) system. MATERIALS AND METHODS Soils, sediment, and sludgggused and preparation The extractability of diethatyl and metolachlor was compared among two soils, lake sediment, sludge, and phosphate buffer or de- ionized water. The samples were; 1) anaerobic sediment from.the pelagic zone of Wintergreen Lake, a hardwater, eutrophic lake in Kalamazoo County, Michigan; 2) Miami sandy loam soil (Typic haplu- dalf); 3) Carlisle muck soil (Typic metasaprist); and 4) digested, secondary, anaerobic sludge obtained from a sewage treatment plant in Jackson, Michigan. The two soils were further characterized by Smith et a1. (1978). The organic matter content and pH for each sample are presented in Table 1. Experiments were run in either 26-mL or 160-mL serum bottles, using 10 g or 50 g (wet weight; see Table 1) samples, respectively. Sediment was weighed into serum'bottles under aerobic conditions using a cut-off lO-mL plastic syringe to add it to the bottles. To compare the adsorption and extractability characteristics, soil samples or the glass bead control was added to bottles so that the dry weight of the sample was equivalent to the dry weight of sediment. Sludge was an exception; 50 g of sludge was added to each bottle, without regard to percent dry weight 15 .w on no unwfioa Hmuou m camuno ou Houm3 Hmoouwvvm poo: use meadow «o w H.m o>mn ones as H.« a mom. x w e.eaV soon sass so is H.m n mas. x m m.ov EmoH meson wofiououooo moaeamm .w on «0 unmaos Hmuou m o>o£ ou nouns Hmoofiufipvm 0: memo: was Aw H.m n moH. x w omv Hmuuouma vaaom w H.m moamuooo usuaapom w on .ouomouonH .mam>«uooemou .mvuaom Nm.om poo mm.w~ mum wagon zone was amoH modem mafia? :.mpfiaom: N~.oH ma pooaavmm w an .oamamxo Mom ponaamauos whoa moadamm HH< .uomaavom mo m om :« mvHHom mo uoooam onu ou .vom: mos summon muonemosm z No.0 mamas .mvmoa macaw Mom udooxo nouns wonwcoaoe .ooaawumav nous w on so OH mo unwaos Hmuou m on unwaoun mama moadamm+ H.m ~.oa ~.H o.~H o.~ .. Aaouuaoo summsnv meson mmmaw o.om o.euo.n -- u- o.“ o.mm mmuaam e.ga w.om a.m o.Hm m.e o.mm son: m.o m.me e.H o.om o.o ~.~ anon Assam o.om N.OH o.oa o.~a o.o oeuom unassumm mauuon madame mauuon onEMm ou popes Hmowwfiuo a“ On pumps Hmofiwwuo nu ma Auv nouuma +uswaoz «Haamm Aev meaaom +unmam3 seesaw any muaaom uaamwuo «Haamm cameos w on madamm m CH .mucoaauooxm ofi vmm=.mvHHom unwound pom .oom: owvoam mom .mafiom .ucmawvmm mo :ofiumuauouomuoso .H «Heme 16 because it would have had to have been concentrated to obtain a dry weight equivalent to that of sediment. Data for Figures 2 and Table 4 (and Figure 1, Appendix A) are from experiments obtained using 26-mL bottles (10 g samples), while data for Figures 3 through 7 and Tables 4 and 5 are from experiments using l60-mL bottles (50 g samples). To determine the extent of adsorption of diethatyl and metolachlor in sediment over time, 10 g sediment was put in 26-mL vials. The sediment was amended with 10 mL of a 100 ppm herbicide solution (final herbicide concentration, 50 ppm; the final sample was 952 water, 52 solids, v/v). The amended sediment samples were centri- fuged and the supernatant was separated using the Sep-Pak 018R pro- cedure; no extractant, including water, was used (results in Figure 2). Several aspects of the extraction of the four herbicides (diethatyl, metolachlor, cyanazine, and chloramben from several systems and clean-up by Sep-Pak C18R procedure were elucidated in the last experiment (see the last three sections of the results). These aspects are: 1) problems with the extraction and clean-up of polar compounds, using chloramben as the primary example and 2) the feasibility of using a pre-rinse to help clean-up samples and completely remove polar compounds from the Sep-Pak C18R cartridge; 3) the adsorption, extractability, and clean-up of the_four herbi- cides from lake sediment, sludge, two soils, and phosphate buffer; and, lastly, 4) the capacity for loading the Sep-Pak C18R cartridges with the 20% acetonitrile extract from sediment. 17 Chemicals Non-radiolabeled diethatyl [2-chlorofiNf(2',6'-diethylpheny1) fimeethyl(ethylcarboxylate)acetamide] and 14Cfgfring labeled diethatyl (100 mg; specific activity, 0.207 uCi/mg) were gifts from Boots/Hercules, Inc., Wilmington, DE (courtesy of Dr. Donald Black). Non-radiolabeled diethatyl was purified according to the method for alachlor described by Tiedje and Hagedorn (1975). Diethatyl was recovered as white needle-shaped crystals after concentrating the hexane extract in a flash evaporator and freezing it. It was re- crystalized until it was 99.9% pure by GLC. Metolachlor [2-chloro- ‘Nf(2'-ethyl-6'-methy1phenyl)jNe(2-methoxy-l-methylacetamide] and 14C- ‘Ufring-labeled metolachlor (1.1 mg; specific activity, 26.4 uCi/mg) were provided by Ciba-Giegy Corp., Agricultural Division, Greensboro, NC (courtesy of Dr. Homer LeBaron). Nonlabeled metolachlor was 97.5% pure by GLC analysis. Thin-layer chromatography and analysis with a plate scanner for 14C-labeled diethatyl and metolachlor revealed no additional spots other than those for the parent compounds. Cyanazine [2-[[4-chloro-6-(ethylamdno)fsftriazine-Z-yl]amino]-2- methylpropionitrile] and chloramben [3-amino-2,S-dichlorobenzoic acid], were gifts from Shell (courtesy of Mr. Norman Gannon, Worthington, 0H), and Amchem Products, Inc. (courtesy of Dr. Amin H. Furrer, Jr., Columbus, OH), respectively. Percent purity, accord- ing to company analysis, was: cyanazine, 97%; and chloroamben, 97.1%. Solubilities of the compounds in water were (ppm, between 20-250 C): diethatyl, 105; metolachlor, 530; cyanazine, 160; chloramben, 700. Properties of the herbicides are described in Table 2. 18 .33" to... .833583 . neonqsn- .ooon-uoaso onus» is. coouuuu-oa «qua-cu .ooou ans-z . .53 sou-so .36 .0033 .l. 3. .503: 3:939: «33. sis-usua- souuouuouuou 033:2. house-38... .93....- 5n.— n 3030:3033“. so in: 2.1.8. 33:3 32.. .33. 6 o2. 8s a... 8“ .88 98.1336 u n no In no 4 .. 4 __ Ian! _ .138 3 on: a: 9: «on \UII 02-3350 «39 3.3qu 3:193 :3 5!. .39. no .83qu 3.: :3 n so none-on «noun-I own-duo 3 an: .30. 3 9.30:— n N sauna-lunar. oceans—x. 38333 .333.— 6 can» can 9; .63 no .35 ga- 4:29:30: 0 0n- voaaoou u no. u Soho .3 . A v . «Jun «usage-Gannon: .- dd 0.! Id 3’ Guns 0 Gnu “On 0 a 8 .- Idhuzu.dn a 6 Al: no»... «Add... coda-«unaccuod gonna 3 5:333 a: 3:3 nauseous: .303qu .32. 3303») no 0.3.3.3.: .u .3: l9 Deionized water was prepared from distilled water which was char- coal filtered, deionized and redistilled; 0.02 M phosphate buffer, pH 7.0, and the herbicide solutions were prepared using deionized water. Solutions of 50 mL of the appropriate 100 ppm solution of each of herbicide dissolved in deionized water were added to each sample bottle. Total concentration of each herbicide in the sediment, soils, or buffer was 50 ppm, since the total volume of the 50 g sample plus 50 mL herbicide solution was 100 mL except for metolachlor data presented in Figure 2, and Tables 4 and 5. The metolachlor concentra- tion in these experiments was 12.5 ppm; the volume was 100 mL. Glass distilled Burdick and Jackson or Baker resi-analyzed aceto- nitrile and methanol were used for extraction procedures and HPLC analysis. All solvents for HPLC analysis, except for the use of deionized water, mentioned above, were commercially prepared high- performance liquid chromatography grade reagents. Extraction procedures Samples containing 50 g sediment or soil plus deionized water which had been amended with 50 mL herbicide solution had a volume approximately equal to 100 mL. Acetonitrile, methanol, or water were added in volumes of 5.3, 11.1, 17.1, or 25 mL to obtain 5, 10, 15, or 20% (v/v) "extractions." In 26-mL sample bottles containing 10 g sample plus 10 mL herbicide solution, 5 mL acetonitrile was added to obtain a 20% (v/v) acetonitrile extract. All samples were shaken for 45-60 seconds, emptied into glass centrifuge tubes (for small samples) or stainless steel centrifuge bottles (for large samples), 20 and centrifuged at 8000 x g for 15 min. Portions of the supernatant were used for analysis. If the sample contained 14C-labeled herbi- cide, an 0.5 mL sample was taken to count the total radioactivity. A 3 mL sample was added to a Sep-Pak C18R cartridge (Waters Assoc., Milford, MA) containing a bonded reverse-phase packing. The Sep- Pak 018R cartridge was coupled to a leur-lock syringe. Non-polar compounds partitioned into and, depending on their affinity for the reverse-phase packing, were retained by the Sep-Pak C18R cartridge, as the sample passed through the cartridge. The "water" which passes through I have termed the "water phase." The cartridge was then re- moved from the syringe, the syringe plunger pulled out, and the_ syringe and cartridge were recoupled. Nonpolar compounds were removed 18R from the Sep-Pak C cartridge by the addition of 3 mL methanol R (methanol phase). Following elution of the Sep-Pak C18 cartridge, the cartridge was washed with 3-4 mL methanol, and rejuvenated with R 3-4 mL water; it could then be reused. The Sep-Pak C18 procedure is diagrammed in Figure 1. Instrumental analysis High-performance liquid chromatography (HPLC) analysis was done on a Varian 5000 microprocessor-controlled HPLC fitted with a Rheodyne 5020 6-part injection valve which had a 20 uL sample loop. Columns used were either a Partisil ODS (Varian), or a Hibar II Lichrosorb column (Whatman, Inc., Clifton, NJ), having the same dimensions and packing: 30 cm x 3.2 mm i.d., containing 10 um reverse-phase C18 packing. Compounds were detected using a Varian UV detector at 254 nm, 21 .Hooonuos NOOH Sufi? pousao mauomsvmmnom mum mpcSOQEoo umaooooz .nwooucu Lama amazonaoo umHod saws: mpoooaaoo umaoeloos mowmumu nowna wofixomo ommnolomum>ou cocoon m camusoo mmmpwuuuoo may .mmmpauuumo o xemuemm wofim=.moo«uzaom aoum mvcboeaoo umaoeoo: mo sowumumamm .H ouowam and a 8.31% we}... umeu>um-2o was?) 22m: 288% moauameux e «(son to zo_h A 1\ ' k ’D/Z kP} ' P/Z q J 4 4 9 4 '4812'4812'48'48'43 TIME (MIN) METOLACHLOP > /2 1" V2 9; c: k C ,1 M- '48M'48'48'48'53 TIME (MIN) 33 Examples of chromatograms from HPLC analysis of the samples after extraction with 20% acetonitrile and Sep-Pak C18R separation are shown in Figure 5 (methanol phase, diethatyl and metolachlor, all systems), Figure 6 (methanol phase, cyanazine and chloramben, all systems except sludge), and Figure 7 (methanol and water phases of chloramben in sludge). HPLC separation was excellent and the extracts were quite clean. Acetonitrile was a better extractant than 20% methanol for diethatyl, metolachlor, and cyanazine in sediment, muck, and sludge (Table 4).‘ Chloramben seemed to have more affinity for sediment than soil sludge. Recovery of chloramben was about 10% higher in muck, sludge, and sandy loam samples and 20% higher in sediment samples extracted with 20% acetonitrile vs. water controls. In buffer, recovery of chloramben in samples extracted with 20% acetonitrile was 10% lower than in water controls, whereas recovery of the other three herbicides from buffer was approximately equal between the 20% acetonitrile extract and the water controls. Again, a compound such as chloramben which is fairly water—soluble, will 18R cartridge when an have even less affinity for the Sep-Pak C organic solvent is present. Extraction of metolachlor was as efficient as that for the other herbicides, even though metolachlor was amended at only one-forth the concentration (12.5 ppm) of the other herbicides (50 ppm). The extraction efficiency of metolachlor amended at 12.5 ppm is also comparable to that for metolachlor amended at 50 ppm (compare with zero-time data in Table 1, Chapter II). 34 1 2 3 4 S r--*--\ r--45--\ t--‘5--\ r--i¥--\ l--‘--\ Diethatyl ML 33L -ML lb 1 LLJ U) o 4 Z 33 l_l"""IF"I'—'1l—1'"||"_l"""ll'"T'—I 3 036036 036036036 0: Metolachlor a: 1 D g.— I.) LLJ [.— LLJ D J I I») I4 H Ry f r—r—I r—r—I r—I—1r—r—1 r—I—I 0360.36036036036 TIME (MIN) Figure 5. HPLC chromatograms of diethatyl (top row) and metolachlor (bottom row) extracted from: (1) 0.02 M phosphate buffer (50 ppm diethatyl; 12.5 ppm.metolachlor); (2) sediment; (3) sandy loam.soil; (4) sludge; and (5) muck soil. Samples were extracted and prepared for HPLC analysis as described in Materials and Methods. Retention times are 4.6 and 4.4 minutes for diethatyl and metolachlor, respectively. 35 1 2 3 4 Cyanazine I I. I .1 I“ W _§§ ‘4 4 4 .. 4 Q laelaelae'se Chloramben C2 C3 p. U LLJ p. LU C3 i L ; z .‘r I .‘J .‘J TIME (MIN) Figure 6. HPLC chromatograms of cyanazine (top row) and chloramben (bottom row) extracted from: (1) 0.02 M phosphate buffer (cyanazine only); (2) sediment; (3) Muck soil; or (4) sandy loam soil. The first injection of chloramben is not an extract but instead is a chromatogram of the "total", unextracted solutifin because chloramben was not extractable by the Sep-Pak Cl8 cartridges under the conditions used. The concentration of both cyanazine and chloramben in the first chromatogram for each was 50 ppm. Extraction, clean- up and analysis of the sediment and soil; samples was done as noted in Materials and Methods. Arrows indicate injec- tion time. Retention times were 2.6 and 2.5 minutes for cyanazine and chloramben, respectively. DETECTOR RESPONSE Figure 7. *e N? f A. 9 3w. |'_'_|f—F-'| l'""'"l 021. 24 240 TIME (MIN) ~ HPLC chromatograms of chloramben extracted from sludge as described in Materials and Methods. Chromatograms are: (l) methanol phase and (2) water phase of a 20% acetonitrile extract and (3) methanol phase and (4) water phase of a water extracted control. Arrows indicate in- Jection times. Retention time of chloramben was 2.5 minutes. I I. N O O 37 The adsorption of each herbicide, in relation to the others, can be compared among the systems (sediment, sludge, soils), as well as comparing the adsorption of the herbicides among themselves in each system. Only qualitative statements on the degree of adsorp- tion among samples and among the herbicides have been made here. These indications should be confirmed by appropriate adsorption studies to determine the Freundlich adsorption coefficient on a per- cent organic carbon basis (Hassett et. al., 1980, and Reinhold et. al., 1979). Generally adsorption is determined using various con- centrations of each herbicide in each system and equilibrating the samples for 24 hours. This was not done in these experiments. Comparison of the percent extracted in water controls (Table 4) can be subtracted from.100%, which indicates the degree of adsorption. Adsorption of diethatyl, metolachlor, and cyanazine is greater in muck soil than other systems, while adsorption in sediment is almost equal to that in sludge for diethatyl and metolachlor. No signifi- cant adsorption of diethatyl, metolachlor, and cyanazine occurred in sandy loam.soil. The degree of adsorption of the four herbicides among theme selves can also be compared. In muck and sludge, the series from the most adsorbed to the least is: diethatyl 3_metolachlor > cyanazine (muck only) > chloramben; and in sediment it is diethatyl 3 metolachlor > chloramben > cyanazine. However, compound solu- bilities, in ppm are: 105 (diethatyl) < 160 (cyanazine) < 530 (metolachlor) < 700 (chloramben). Hydrophobic sorption increases ‘with decreasing compound polarity, or decreasing solubility in 38 water. Thus, on the basis of compound solubilities it is somewhat surprising that metolachlor is more adsorbed than cyanazine in all systems and that in sediment the adsorption of cyanazine is much less than that of diethatyl and metolachlor and almost 10% less than that of chloramben, the most hydrophilic herbicide present. Perhaps the phenyl rings of diethatyl and metolachlor have more hydrophobic character than the nitro-substituted ring of cyanazine. Diethatyl and metolachlor may also form hydrogen bonds more easily than cyanzine through the carbonyl of ketone and/or ester groups in the alkyl groups on the nitrogen (see compound structures, Table 3). The anaerobic nature of the sediment may create a qualitative difference in the nature of the organic matter, enhancing the sorp— tion through hydrogen bonding which would especially increase the adsorption of chloramben above that in the other systems. If the organic matter in sediment contains more reduced functional groups such as amines, sulfides, or alcohols, an acidic compound such as chloramben would have a greater tendency to adsorb through ionic or hydrogen bonding interactions than it would to more oxidized, negatively charged functional groups present in the organic matter of aerobic soils.“ 18R Loading_capacity of Sep-Pak C cartridges R The Sep-Pak 018 cartridges were also tested for their loading capacity. Various amounts, from 3 to 45 mL, of the 20% acetonitrile extracts from sediment of diethatyl or metolachlor were added to R a Sep—Pak C18 cartridge and eluted with equal volumes of 100% methanol. 39 Results are shown in Table 5, and indicate that the total loading capacity of the Sep-Pak C18R cartridges was not reached. The diethatyl sample contained about 45 mg/mL diethatyl, which is exactly what was recovered even when 2250 mg (total) diethatyl was added to the cartridge. This is impressive since the 20% acetoni- trile extract of sediment contains a great deal more than just the extracted diethatyl. Whitney and Thaler (1980), determined that the total capacity of the Sep—Pak C18R cartridge for the bile salt taurocholate was in the range of 50 mg. They also found they could suction fluid qhantities (urine samples) of at least 750 mL at flow rates up to 20 mL/min without significant loss of the bile salts. Because of the debris in the 20% acetonitrile extract of sediment, I would not expect to be able to load such a large volume on the cartridge, although the capacity may be much greater than the 45 mL sample volume which I tried (Table 5). As in this study, Whitney and Thaler (1980) were able to reuse the Sep-Pak 018R cartridge by washing with methanol and regenerating with water. In my studies, R the Sep-Pak C18 cartridge efficiency appeared to decrease after more than six uses; the sediment was leaving debris on the cartridges which was not removed by the methanol wash or during re- generation with water. Acetonitrile was not used to elute the samples or regenerate the cartridges due to its greater toxicity, but it may be more effective. The cartridges were not reused more than six times for all studies in this thesis. 40 o.mem c.mNH o.m~ m.~m emcee w: swoon I m.NH m.HH m.HH m.HH m.HH wouo>ouou Aa\w: I uoaaomfiouoz omum com com and emcee m; Hmuou . on me me “a me wouo>oomu As\wn u Heumeuman Aaa\mav me as o m unmaHoom ca sowumuucooaoo AAEV woven oomuuxo mo oasao> Hmfiuaao .owvfiuuumo M mac 3mmlmom o co uooaavom voodoamlomaowauon mo uoouuxo oawuuwaouoom Now e no hufiomemo wsavoog .m oHan 41 R The ability of the Sep-Pak C18 cartridge to retain the herbicides during passage of large volumes of fluid is valuable in the study of herbicide residues, for example, in lake water. The presence of trace amount of herbicides could be detected by passing a large quantity of a solution through the Sep-Pak 018R cartridge, and eluting the sorbed herbicides with much smaller quantities of solvent. In a preliminary experiment, 50 ppm and 100 ppm solutions of diethatyl were concentrated by addition of each of the samples to a Sep-Pak C18R cartridge and eluting with only one-half the sample volume. Peak areas, from HPLC analysis, were compared with that of the 100 ppm.sample which had been eluted with an equal volume of the initial .sample size. Ratios calculated by dividing the peak areas of the concentrated samples by the unconcentrated Sample were 0.84 and 1.62 for the 50 ppm and 100 ppm samples. Ideally, they should be 1.00 and 2.00, but efficiency was lost due to the use of 80:20 methanol- water as an eluant rather than 100% methanol. However, these re- sults do indicate a potential for concentrating samples. 42 DISCUSSION Generally, extraction procedures can be classified into three major approaches (deduced from Table 6): 1) use of polar and nonpolar solvent combinations such as benzene-isopropanol, chloroform-diethyl ether, or hexane-acetone (1 through 19); 2) use of polar solvents alone (acetone, methanol or acetonitrile), or in combination with water or each other (20 through 29); 3) use of polar solvents, occas- ionally in combination with nonpolar solvents or base, under reflux (Soxhlet) conditions (30 through 37). Initially extractions may be followed by the humic acid extractants sodium pyrophosphate, Na4P207, or 0.5 - 1.0 N NaOH (6, 16, 22, 23, 25, 26, 27) or by Soxhlet or alkaline steam distillation (24, 30, 31, 33). Frequently extracts are partitioned into nonpolar solvents and analyzed directly (20, 22, 25, 33), or partitioned and cleaned on alumina or florisil columns prior to analysis by gas-liquid chromatography or HPLC (8, 21, 24, 28, 31, 34, 35, 36). Lastly, except for one case, samples are frequent- l4 1y combusted to C02 after extraction is completed to determine the 14co-1abe1 (3, 4, 6, 7, 10, 14 to 17, 21, 24, 28 amount of residual to 31, 33, 35, 37). Extraction (shake) times vary from about one minute to 20h (Table 6). In the experiments in these studies (Chapter I and II), samples were only shaken for 45-60 seconds. Time dependency of shaking on the extraction efficiency is important, but whether longer shake times would have increased the amount extracted here is not known. 43 ensue. A¢.«u0uu .1.u.uu..ueoo “Avocaduuuu.av 3+. n..«a.oo A...u..a Ana: v..o«»«uu.a .1..«A.uu A: .ssuc..~u AA.A. .aouou. u...u..s .un A: ...u..s ..n A. u..u..u.aau o deco. .u¢e~ . Aa+.. ”go-scsuuucu a . ounces; cocoo_. ...o« .u uuuuuuu unnuuuu u...u“u“w“avuoo u..«.«.. .A.... ..o«u.:a.ou Au nu u.n-«~. «as ea .uoacono. «cues-on. ...a-.. a». Aaouoo... acouuoo NAA Aa+lv n.u.u.eo. A9..»~... a as. .v a nu .uoqsuoa .Aos.4u.l A: s 00‘ v.voo~u no o_so.o-e< u.ao~o< .ca- e. .A~.~. one-cos - scene... A. o c. ~sz....eoco~eu.e case... ...oa .x A.~ c couu.o«.. a.) . u.:A.oo Au an. as“ o e .. ~o- q. : ~.o A; a an un— .u..>Aou nu .ea... u0.uuu. .sq. on .oqa. uuSAsu .J.£. .usuucu.>o ac.u. A~.nv «03.30ueoou u .e..e.a A. v on u»u.£u.ua .uo~au.«< .uulas .Auoa .o . .Bd..uu< .uoaau.u:- o Auuav do..aaueo.« I .3....A A. a nu .uonau.50un .ueusu.u< Ago» .n Isa... .usuusu .« a «A .«suua ...u...uoouu«. u..«o«.o .Asl.. .v.».:a.oo A4 nude. as solace ..u«—oa «a AA.A~A~ occu.u. Ae¢.:u.. 4 .c.u¢.a .In A. A « u.u.l as. cogsu.u.~ Adam .9 Aaao .. oaeuoau .uo. nos-:a-au .o 4 cm £253 e.g.) u «9.22.. AA .33.: .3 n .unA . u 1.»u.uuu. nu. . A + . a an .A«.A.Av .90uou. u .u.u.u. «Ann. I .ccncoa A. c an .o..«v.uo u«4¢n.. uuuou .n .uan coaacaune .uu~ coda.-a-a~»g..: can on .A~.~. cogs. exec-«s - nuououosau Au Auosoa.~.ue~ .aoa. m .ucA aeggueuuucem .AA.uouv u0u9.uu.. . an. a. ..c..c.a An seagu.u.m ..o«4u.u.s ..u«Aon.u.. .30.; .:AA .vuuqau.a u..u.¢ nu. ma . oa~.l u «oe.su.l 4 .cou.u. A. 1 ~A «annex ..o«suouuuson vooooAu .Aaon .N nuance .dunsu. use. u.a.noa ace: soaouuauuug .uoaaa-oo Au.a... a:43..~u ..«I n .aql nu ....n.n .IN A. a. .xg... .ao.... - .a.... .. a— nun ...«uuauoa uaou.e Au ...Ou.o. A. I . v.~.a.~uu use. .u.uu*.>cou Aden .A ..o«u..aa.uu «scone. u.Aossoa\u.th .o: ...- Avassae. unco. .o u. «nan vuuoaeo.- o..«u .o«u0.uun. u. .3.3. ..u..>A.. .o«m.a:o.u ...esoo.oo .oa .muamaavom use wagon scum outflowuoomow use mooaownuon maofium> mo mowuo>ooou one moonuoa coauomuuxo moow>oun mo mumaaom .0 dance 44 3 3...... in A. s 2 .320... :8 8. 8a A. 3.3.38... 0. 03 N8 «n ill I) c« a =8 u :8 d- . .6 35 no .3.— A«.n.3 3.3.0.350. «50373.03. «30. A. . o~« nan...« «0o. 3 2 Acetic-.3308: . o c.3303.» 0« on A0 n« on 3.3 . « .0 0 0.53.. .«g 1.0.21.0 A0 .««.«n .0.««.« h 20......— AA v A: .033: Ad". «1.0.0.0430... .3 A. a no 00.3300. 3.. n n6 ll! «.0 0 3|! .« « 0001.0 .« A« u . «.0 3 on .A.... .30. 3|... 1.. 0:16 1.00.000. .« 6.0.213 .« A0 2 Am: 8? =8 .8. 2 48...... S I. .03: :8 8.30 + :8 a... .— ~ .25 Bo...o2?~8.53 :a A. s 8 5.5.8.0-... .33 «n.« 3.33013 In. “~83 .0 .30 0.53.0 .3... 330.2100 A0 .A«u«a«v 3...... 2 4.80.32 38 2.9.... 3 .00 3... $8103.28»... 2 .25 «31.34880. .3 A. a n 820.... :8 .A....0.0. on 17>... 1.32. 0.10 «0 Nan 0:33 1.0.380 A. inch-00.30 A. .33» 230.0. 300.00.. a: 3.5100 A103... 0...... A0 .A«.$ «330.013.000.30 A. .380... A. «. no... 88. 03... 3 A9133 «0.301 31.—005.34....«303... A. II 8.21.03.33.03... .38 .5. o« 1.3.: a. .35...— 150580 0.80.. .8 A. .03.. 3...... so. 3 « 10.0. and. 3 3.... .280... A. I « A3310: 0..» 03-.» ..««o. .3. .33.: .on Adusuoauflu I nun 30% non A. A13 01.0. «3037-0003030 .In A. v 3 .0335 .93.. .3. A033... «3 ..I.«.3 .30.”! D. A. ANIXBV n« no. A3030». + use A. A3030: 30. 30.0. 30.«§.T.80.0. A. a 0 5 0a.”... .38 en 3.. .«0‘. .na a nu A13 0.10. «303750000030 .8 A. 1 an .3... 0.3.: .30. . 33033.8 0.3«3 0.3.8:}.«0. 6- .0.- .122. ~33 u. u. I: .3030..- 0'04 300.00.. .0 3.4. 30.25.. 3:03 .352..."- .0- As.0.oov 0.3.0. .o« .o« .2 .o« .n« .v« .n« .N« .«d .o« .0 1L5 .mn-.. ...... "can"... 30...... 30.1.... 1.183 A .3333... 8333:... .3... .3303..— ..§ 10.0.03»... 2.3!...«0 6:333... .0 3.0... 0...... . u 49...... ..N A. . e 33.32.... .3... n .3 A13... Nu. A. 3.33.... A. .l.«.. 3.3.: A333... .3... «.1302. .32.... .0... 1.333.... 2.3:...«0 .: .2 32...... .82.... a... .3 A. .8. n 6.8.8... .2... .83.... n .3... A. . 8 8...... .868. .38 .2 as. us. A. .n.....o... .u. A.. A. . A....... ........ ... . no u... A. 3.3.3.. .3 A. . 3.3.3.533 .30.. .« o 3...... 3.3.... 3.31.. 1.0.3.3... . on 3.5.3.. 2......“ .3 as. - ... A. . a... a. A. .- .o a no. - .3 A. A85: .3... a 838...... A. . 3 .Egg .288. .3... .8 88...... a... 3.88.... A8 .3... u« .3338 a .8838... .3 A. 8. an .838. .8... - 838...... .8 A. . 8. . .88.... 8... .8... .2 «3...: «.6... .0..— 1.0.33... A.» .31.:«0 8.... .a. .8... 8.1 .88... 3.8.8.. An. .33.... AS A. «.33... u: I n« .. 3.3.3... 83303 «2...... .n« A. 3... a: 0 a. a. + . 3.33.0... .9330... .30.... .IN A. 1 3 c...0....o..0«......u 1.1.0: .«3... ..N on An... .31.... «3. u. 33...... .2. A. u. A. 2...... 3.. A3 68.8.. .81.... A3 .38... as A2 -8. A.» 8.... .A. a. .53... 8.82.8. I... «an A: A. .3.....«.l.. .3. A3 .tx» .33 .3331». 2.3.3. A. .53.. 8 an A. 2...... .3 A. 1 no 5.3.... Lin 2|... .33 .3 .3... «5...... .0... .u... .8 0...... J0..— ... 3.3.. A. 0...... 3!. 1.6.33... A. 3.3.8 .. n. .8.- n a... A. a: A. . n« .30.. .N no .N .«I 3 ..u 3.00.... «a u. .«A. . o.« 3... .. n6 1.... A35: .3... u «0...... .«.A. 1 ns 33...... «.8 .s... 5.3.. .«g. nan .n €03.38. .u .3 9.... .- ........au ..:0 nun .«A. .33....— AEV 1.3.33... 390...... .3. .« A. .n .s. A. a... a . ......... .n. A. . .. .m....... A........ .... ... 3.0... «u... .31.. 3.3.03. .0... 1.833... A. 973.3 A1321]... no... A. 1.0.33... A. .1.1..«u .. n. I: A. 3...... in A. v .3 33533.. 6...... 3... .3 .0358 .2... ... .... A....... 8.... .o a. ...u 15......- ..I.. 33...... a. 3... 3.33.. ...u......« “Ill... .0. A1.0...v o .«..h 46 a: N8 13.3.2 .3033 A: A‘.080v O Odd-P a... A. .8... 3...... A... 13.82.... .u an a: A. «.3... .1 an £13 «2.3:. u: I u a .n A. s 3 33.3.2..- 38 .3 n .leo 33.3: «3...: .33.... 3.7.33 «.3... .1 n .28.... A. a c E v.13: .30. .on n... 3 . a .. N8 _. .u A... 3...». A... «.3 . A3 .3.- . .— Iaao. 33.0: .5330». n u o "a "a nuA .- a... .333... 15.5. .33.... 2:73.30 . on A... 33".“... 2 2 z a o .32.... A. 83.8. i 8 .38... a. a .2 A: 2.3.3... - 4- 4 I .0 .3338. .A'.. 63.31. A a a 3 A: 3.3.53... 33.0. .33.»... .28.... AA. agave—Ks 03!. 0:05.39. ail—3a .3 Bali. .. M 2.... 3.3.0... «3.3.8.... A. 3.74.70 .03 A. I. main-0H0 no 05 33... .33. 3.8.033. I 3...... A. 1 3A 3...... A... .3 2.3... 34.40.... A33 .338... A... .83.... .8... A: 33.... .838... 5:. a: I a: 33.: 3.3.3.... £83.23... .8... 3. 3.28.2 nan I u. 313 .SAJA. .2513. .3 o 5.3.3:... .33.: A3 .SuuswhausguI. a: I awn A153 .0334... A: AA 3333-533“. I. 0 as I a. A. a «A 3.3... .335.- A. a . finiauonelc 30a .3 .33... was I «S :8. 3 3...... 3...... 3...... n3 .3 :1... A. an .33.. u .33... .0 C25..- 3. 3 .53 .3... u coo 33.5. "no A. A. 8 .33... 38 .3 3d: 33.. 3030.! an 903...: €332.23 “Aug duly sun: 1.3.3:...— .u.u..oa.>. .1338. u 4... 97.3.3 .3 an. . 3:3 5. a 39-250.- A.uoh £13.03 .8 fl 1393...... u a...) I 37.3.3... A; a 3 .333.» £33 .3... I .3333... A. 3.5 .83... 0 ~33 3a.... 3 :33... £258. 3 «A u... .3qu o A 5.3.... 39.5.... .3 an. I non a: I a: u A~83v 83.3.... 3.8m nun u. a: I and .u a: I no man I and .AA. :0. v3.3.3 A: IIIIIIIIIIII A. a... a . .. ~ 33...»... 3.... m8... 8.. A: A. u. I a. 8. I u. A. o a 6...... .3 A. an I an an I a... A. a... 8. 3 .35. «3.... in A. 9-..; 3 GI. u hIn . n .33.»... 3.4. 3.3 .3... and»... I an: I n4 A. 1 3A hIn...~ =0. .8 3.03238 23.3.3 9.... .0: Jon A33... «33 «o 5 .5.— CD30... .3. 3.30.38. .0 3.... {eggs guzaonu a; .o- 47 .wouo: unashmnuo mamaas .vmuaammua ma amum coaumsnaou m:u no .mmoum new» Iomuuxm cucunaou mnu .noum nowuumuuxm 50mm ha wouo>oomu Hmnmalo Hugo» mo unseen «nu "huo>oomm «H :.m:nvw:mmau: mm? uomuuxm :uwns ou whammy An no Am mzu "zanlamoau: Amy .Haom mau cu mmavamou canon mo opamwma w .NooqH :fimuno ou coaumanaou am: up no nouavwxo oHaamm m a« vmumanaoo mg and mucm>aom nufia vmuomuuxm camp was snags mavwmou Haom ou mummou mhmaam :vmumanaou: :ANV van AC 3.. mm 3an 3 3.3 .momz 5.3 coauuwuuxm you mac can coaumanaoo new mac ..m.m .mcoauuoa oBu Gus“ vmwu>wv mg manamm a NH . :.Q=I:mmao: :a.vocoauama mu mficu .vmawnaoo mum muumuuxw any ma "coauowuuxa mo umvuo us» a“ Au.n.mv voumwfl mum muam>Hom .mfimmn >\> m fiO fimuflmmwhh GHQ mUGMPHOm A3 A8 A8 wmafinaoo wo.moaumm .vmuammwum mum Amuonusm ha cm>fim may mmawu coauumuuxw and van: muaw>aom Aav .Aav mnucoa Ho .Asv muse: .Avv mhmv .Asv mxmma aw ohm muawu aouumnaoaH .vmuoa mausumcuo mamas: vmamnuauu ohm «vasomaou «a mouocu00h c A6388 0 manna manna 48 Nonpolar solvents frequently foam or emulsify when used as extract- ants. Emulsions can be dispersed by the addition of an antifoam agent, 2% NaCL, or by freezing the extract. But these procedures cost time or add additional difficulties to the clean-up and analysis procedures. An extractant such as acetonitrile, which is miscible in water and also mixes well with particulates, is ideal for an aqueous medium such as sediment or flooded soils, and difficulties with foaming are not encountered. Use of stronger extractants in the sediment created difficulties. Addition of sodium pyrophosphate to Wintergreen Lake sediment resulted in much darker-colored, extracts with more interfering material (as determined by HPLC analysis of a sample after clean-up), than did the 20% acetonitrile extracts. In the extraction experiment for Figure 3, extracts from sediment were more intensely colored (darker yellow to brown) as the concentration of acetonitrile used for extraction was increased. Basic or acidic extractants were not used because I have observed that diethatyl hydrolyzes easily under those conditions. Chou (1977) has also observed the hydrolysis of alachlor and diethatyl under basic conditions. Generally, a higher ratio of acetonitrile-water (80:20 or 70:30, v/v; see 13, 14 in Table 6, and Smith, 1981) is used for extraction than what was used here (202, or 20:80). Smith (1981) found that acetonitrile-water (90:70 v/v) was only about one-half as efficient as acetonitrile-ammonium hydroxide at pH 9.0 (70:30 v/v) for atrazine residues in soils weathered 12 months, although there was no significant 49 difference between acetonitrile-water (70:30 v/v) and acetonitrile- water-acetic acid (70:30:2.5 v/v/v) for the extraction of benzoylprop or flamprop from similar lZ—month old soil samples. Smith also found that the addition of water prior to extraction did not increase the amount of herbicides extracted. In my studies, a higher ratio of acetonitriledwater such as 25:70 or 30:70 (v/v) may have been more efficient, especially for diethatyl, but I was concerned that an acetonitrile concentration which was too high would cause the herbi- cides and their nonpolar metabolites to not be retained on the Sep- R Pak C18 cartridge unless the solution was diluted with water. Metola- R chlor and its metabolites were partly removed from a Sep-Pak C18 cartridge when a 40:60 or 50:50 (v/v) methanol-water pre-rinse was applied to the cartridge (Figure 4). Acetonitrile is a stronger sol- vent and may limit retention at lower concentrations if mixed with water for use as a pre-rinse. Although this initial extraction procedure is not exhaustive, examin- ation of the data from the propylene oxide sterilized samples in Chapter 11 indicated that the compounds, especially the parent herbicides, were extracted as efficiently after 8 weeks incubation as they were at zero time. Since Figure 2 (Chapter I) indicates that most of the diethatyl and metolachlor added were adsorbed within 2 days in sterile and non- sterile sediment, the extraction procedure seems to extract at least 85 to 902 of the diethatyl and over 90% of the metolachlor if its disap- pearance is due to simple adsorption. Chou (1977) has found that a large percent of the products formed from alachlor and/or diethatyl 50 in a resting cell experiment using a pure culture of Chaetomium globosum.were almost immediately inextractable (bound) when added to sterilized soil or humic or fulvic acid extracts of soil. About 52% of the products formed by g; globosum were bound in soil within 4 days, and extracted humic and fulvic acids each bound 42% within 6 days. Katon and Lichtenstein (1977) have had similar results with parathion. Hsu and Bertha (1976) and Bollag et al. (1978) have found that a large pro- portion of various substituted anilines, which are potential inter- mediates in the degradation of several pesticides, including the acetanilides, become bound very quickly, within 24 h. However, these bound residues require extraction by alkaline hydrolysis under reflux conditions. A disadvantage in extracting the entire sample rather than separa- ting the particulates from the aqueous fraction is that it does not distinquish between completely soluble compounds or metabolites which are not adsorbed and those which are. However, from Figure 2 and work by Chou (1977), alachlor, diethatyl, and their metabolites are generally closely associated with sediment and 3011 and I would not expect signifi- cant fractions to be in the water. The second step required for the preparation of samples for analysis, is an effective method of separating polar from nonpolar com- ponents and for sample clean-up prior to HPLC analysis. This is provided R by the Sep—Pak C18 procedure. Growth culture media, soils, or sedi- R ments should always be centrifuged or filtered prior to Sep-Pak C18 separation to remove particulates. A pre-rinse of water should be used 51 with buffers or salty solutions prior to eluting compounds for the Sep- R Pak C18 cartridge with an organic solvent because a precipitate will form when the "salt" is mixed with the organic solvent. Also, when a small sample volume (8 to 10 mL) is used, there may be considerable error if polar metabolites are not rinsed off the Sep-Pak C18R car- tridge (e.g., with a water pre-rinse) prior to elution of a nonpolar sample. Although the loading capacity of the sediment extract on the Sep- R Pak C18 cartridge (Table 5) compares favorably with the results of Whitney and Thaler (1980), parameters needed for the use of Sep-Pak R 018 cartridges for large volumes, higher loading, faster flow rates, and other more polar (such as chloramben) or more nonpolar compounds than the acetanilides should be determined on an individual basis. Retention of an acidic compound such as chloramben could be improved by acidifying the extract to protonate the compound. Compounds which are more nonpolar than the acetanilides may require a stronger eluant -than methanol; e.g., acetonitrile. HPLC chromatograms (Figure 5 R through 7) indicate that the Sep-Pak C18 cartridge is excellent for clean-up of samples extracted from flooded soils, sediment, and even sludge. Furthermore, the samples are in a solvent which can be directly injected into an HPLC for analysis. Florisil and alumina are effective in the clean-up of extracts of environmental soil or water samples (Table 6). Tenax-GC, a polymer of diphenylphenylene oxide, has been shown to be effective in the extraction of compounds from environ- mental soil or water samples (Shiaris et al., 1980). However, compounds .52 are generally eluted by gravity from these columns (thus much slower) using nonpolar solvents (hexane, benzene, diethyl ether, chloroform) which are inappropriate as solvents for injection of samples onto reverse-phase (C18) HPLC columns. Furthermore, these columns need to be packed by hand, except for florisil and silica gel which are now available in Sep-Pak cartridges (waters Assoc. Milford, MA). The Sep-Pak cartridges offer a simple, rapid method of clean-up because they are pre-packed, they can be attached to leur-lok syringes so that the sample and solvent can be quickly "pushed" through them. Also, in the case of the cartridges containing reverse~phase (C18) packing, debris in the sample which would permanently adhere to and ruin the reverse-phase packing in the column on the HPLC is removed by the Sep-Pak 018R cartridge. Furthermore, it was possible to extract and separate 50 to 75 samples within one week, preparing 200 to 300 samples for 14C-analysis and 50 to 75 samples for HPLC analysis, not R including water phases of the separation. Reuse of the Sep-Pak 018 cartridges up to 6 times costs about $0.25 per sample, not including solvent costs. However, use of the cartridges also saves time for the researcher and minimizes the use of solvents used for normal extrac- tion procedures; both factors help minimize costs. 10. ll. LITERATURE CITED Anderson, J. P. B., and K. H. Domsch. 1980. Influence of selected pesticides on the microbial degradation of 14C- triallate and 14C-diallate in soil. Arch. Environ. Contam. Toxicol. ‘9:llS-123. Adhya, T. K., S.-Barik, and N. Sethunathan. 1981. Stability of commercial formulation of fenitrothion, methyl parathion, and parathion in anaerobic soils. J. Agric. Food Chem. 22590- 93. Ambrosi, D., P. C. Kearney, and J. A. Macchia. 1977. persis- tence and metabolism of oxadiazon in soils. J. Agric. Food Chem..2§:868-872. Beestman, G. B., and J. M. Deming. 1974. Dissipation of acetanilide herbicides from soil. Agron. J. §§:308-311. Beland, F. A., S. 0. Farwell, R. D. Greer. 1974. Anaerobic degradation of l,l,l,2-tetrachloro-2,2-bis (pfchlorophenyl)- ethane (DTE). J. Agric. Food Chem. ‘22:1148-1149. Bollag, J.-M., P. Blattmann, and T. Laanio. 1978. Adsorption and transformation of four substituted anilines in soil. J. Camper, N. D., K. Stralka, and H. D. Skipper. 1980. Aerobic and anaerobic degradation of profluralin and trifluralin. J. Environ. Sci. Health. 315:457-473. Chou, S.-F. J. 1977. Fate of acetanilides in soils and poly— brominated biphenyls (PBB's) in soils and plants. Ph.D. Dissertation, Michigan State University, E. Lansing, MI. 90 PP- Clark, J. M., and F. Matsumura. 1979. Metabolism of toxaphene by aquatic sediment and a camphor-degrading pseudomonad. Arch. Environ. Contam. Toxicol. .§:299-308. de Vos, R. H., M. C. ten Noever de Brauw, and P. D. A. 01thof. 1974. Residues of pentachloronitrobenzene and related com- pounds in greenhouse soils. Bull. Environ. Contam. Toxicol. .li‘567'571° Doyle, R. 0., D. D. Kaufman, G. W. Burt, L. Douglass. 1981. Degradation of cis-permethrin in soil amended with sewage sludge or dairy manure. J. Agric. Food Chem. ‘22:412-414. 53 12. 13. 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 54 Ferris, I. G., and E. P. Lichtenstein. 1980. Interactions between agricultural chemicals and soil microflora and their effects on the degradation of 14C-parathion in a cranberry soil. J. Agric. Food Chem. .2§:lOll-1019. Golab, T., W. A. Althaus, and H. L. WOoten. 1979. Fate of 14C-trifluralin in soil. J. Agric. Food Chem. iglzl63-l79. Golab, T., R. J. Herberg, J. V. Gramlich, A. P. Raun, G. W. Probst. 1970. Fate of benefin in soils, plants, artifical rumen fluid, and the ruminant animal. J. Agric. Food Chem. $83838-844. Guenzi, W. D., and W. E. Beard. 1968. Anerobic conversion of DDT and DDD and aerobic stability of DDT in soils. Soil Soc. Am. Proc. 32:522-524. Hargrove, R. S., and M. G. Merkle. 1971. The loss of alachlor from soil. weed Sci. 12:652-654. Hassett, J. J., J. C. Means, W. L. Banwart, and S. G. WOod. 1980. Sorption and properties of sediments and energy-related pollutants. EPAr600/3-80-041. Athens, GA 30605. Herbes, S. B., and L. R. Schwall. 1978. Microbial transfor- mation of polycyclic aromatic hydrocarbons in pristine and petroleumrcontaminated sediments. Appl. Environ. Microbiol. .§§:306-316. Hill, B. D., 1981. Persistence and distribution of fenvalerate residues in soil under field and laboratory conditions. J. Han, T.-S, and R. Bartha. 1976. Hydrolyzable and non- hydrolyzable 3,4-dichloroaniline-humus complexes and their respective rates of biodegradation. J. Agric. Food Chem. ‘gi:ll8-122. Iwata, Y., M. Ittig, and F. A. Gunther. 1977. Degradation of 9,97dimethyl §7Icarboethoxy)benzyl] phosphorodithioate (phenthoate) in soil. Arch. Environ. Contam. Toxicol. 6:1- 23. Katan, J., and E. P. Lichtenstein. 1977. Mechanisms of pro- duction of soil-bound residues of 14C-parathion by micro- organisms. J. Agric. Food Chem. ‘25:1404-1408. Kaufman, D. D., J. R. Plinnner, P. C. Kearney, J. Blake, and F. S. Guardia. 1968. Chemical versus microbial decomposition of amitrole in soil. weed Sci. 16:266-272. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 55 Khan, S. U. and H. A. Hamilton. 1980. Extractable and bound (nonextractable) residues of prometryn and its metabolites in an organic soil. J. Agric. Food Chem. 28:126-132. Martens, R. 1978. Degradation of the herbicide (14C)—dichlofop- methyl in soil under different conditions. Pestic. Sci. .9:127-134. Mathur, S. P. and J. G. Saha. 1977. Degradation of lindane- 14C in a mineral soil and in an organic soil. Bull. Environ. McCall, P. J., S. A. Vrona, and S. S. Kelley. 1981. Fate of uniformly carbon-14C-ring labeled 2,4,5-trichlorophenoxyacetic acid and 2,4,-dichlorophenoxyacetic acid. J. Agric. Food Chem. 32:100-106. Muir, D. C. G. 1980. Determination of terbutryn and its degradation products in water, sediments, aquatic plants, and fish. J. Agric. Food Chem. '2§:7l4-7l9. Murthy, N. B. K., and D. D. Kaufman. 1978. Degradation of pentachloronitrobenzene (PCNB) in anaerobic soils. J. Agric. Food Chem. ‘2651151-1156. Reinbold, K. A., J. J. Hassett, J. C. Means, and W. L. Banwart. 1979. Adosrption of energy-related organic pollutants: a literature review. EPAr600/3-79-086, Athens, GA 30605. Roberts, T. R., and M. E. Standen. 1978. Degradation of the herbicide flamprop-methyl in soil under anaerobic conditions. Pest. Biochem. Physiol. 2:322-333. Rosenberg, A., and M. Alexander. 1980. 2,4,5-Trichlorophenoxy- acetic acid (2,4,S-T) decomposition in tropical soil and its cometabolism by bacteria in vitro. J. Agric. Food Chem. 28:705-709. Shiaris, M. P., T. W. Sherrill, and G. S. Sayler. 1980. Tenax- GC extraction technique for residual polychlorinated biphenyl and polyaromatic hydrocarbon analysis in biodegradation assays. Appl. Environ. Microbiol. .32:l65-l71. Siddarame Gowda, T. K., and N. Sethunathan. 1977. Endrin decom-position in soils as influenced by aerobic and anaerobic conditions. Soil Sci. .124:5-9. Smith, A. E. 1981. Comparison of solvent systems for the extraction of atrazine, benzoylprop, flamprop, and trifluralin from.weathered field soils. J. Agric. Food Chem. 129:111-115. 36. 37. 38. 39. 40. 41. 42. 56 Spillner, C. J., Jr., J. R. DeBraun, and J. J. Menu. 1979. Degradation of fenitrothion in forest soil and effects on forest soil microbes. J. Agric. Food Chem. .21:1054-1060. Takase, I., T. Nakahara, and K. Ishizuka. 1978. Degradation of 3-(3-chloro-4-chlorodifluoromethylthiophenyl)-l,l-dimethylurea (ClearcideR) in paddy soils. J. Pesticide Sci. “3:9-19. Venkateswarlu, K., and N. Sethunathan. 1978. Degradation of carbofuran in rice soils as influenced by repeated applications and exposure to aerobic conditions following anaerobiosis. J. Agric. Food Chem. .26:1l48-1151. Ward, C. T., and F. Matsumura. 1978. Fate of 2,3,7,8-tetra- chlorodibenzofgfdioxin (TCDD) in a model aquatic environment. Arch. Environ. Contam. Toxicol. “1:349-457. Whitney, J. 0., and M. M. Thaler. 1980. A simple liquid chromato- graphic method for quantitative extraction of hydrophobic come pounds from aqueous solutions. J. Liquid Chrom. .§:545-556. wu, T. L. 1980. Dissipation of herbicides atrazine and alachlor in a Maryland corn field. J. Environ. Qual. .l:459-465. Yoshida, T., and T. F. Castro. 1970. Degradation of Gamma- BHC in rice soils. Soil Sci. Soc. Am. Proc. .34:44-442. CHAPTER II. ANAEROBIC METABOLISM OF TWO ACETANILIDE HERBICIDES IN A LAKE SEDIMENT Acetanilide herbicides are widely used for broadleaf weed control in corn and soybeans. Although field conditions are generally con- sidered aerobic, the herbicides may be exposed to anaerobic conditions if they reach aquatic sediments through run-off, if they are exposed to poorly drained conditions in the field, or if they reach a sewage disposal system. Several studies have shown that they are biodegrad- able by pure fungal cultures (Tiedje and Hagedorn, 1975; McGahen and Tiedje, 1978; Smith and Phillips, 1975; Chahal et al., 1976; and Kaufman et al., 1971), soils (Chou, 1977; Hargrove and Merkle, 1971; Beestman and Deming, 1974; and EPA.Metolachlor Pesticide Registration Standard, 1980), and in xitrg by an acylamidase from a soil fungus (Lanzilotta and Pramer, 1980). These studies were all conducted under aerobic conditions, although reactions in the gastrointestinal system of rats may be occurring under anaerobic conditions. Leavitt and Penner (1979) have shown that the acetanilide herbicides alachlor [2- chloro-Z',6'-diethy1fo(methoxymethyl)acetanilide)], and metolachlor [2-chlorofigf(2-ethy1-6-methylphenyl)fo(2-methoxy-l-methy1ethyl)acetamide)] conjugate in vitae (chemically), under anaerobic conditions, with glutathione and other thiols. However, the potential for the anaerobic biodegradation of the acetanilides is not known. Anaerobic biodegradation of pesticides in many types of systems (pure cultures of microorganisms, rumen fluid, sediment, flooded soils, 57 58 sludge, gut microflora, corpses, and by chemical means) was reviewed by Williams (1977). Sethunathan (1978) reviewed anaerobic transfor- mations of pesticides in rice soils. Generally, reactions involved dehalogenation (DDT, PCP, aldrin, linuron, and 2,4,5-T), hydrolysis of phosphate and alkyl esters (parathion, diazinon, malathion, and carbaryl), reduction of nitro and sulfoxide groups (parathion, fensulfothion, amitrole, trifluralin, and propanil), and dealkylation of N7 and gralkyl compounds (atrazine). Some reactions such as nitro group reduction, hydrolysis, and, in particular, dehalogenation, may ' in part be performed by chemical means. Other reactions which occur under anaerobic conditions are acetylation (Bollag and Russel, 1976) and replacement of groups such as a nitro moiety by a thiomethyl moiety (Murthy and Kaufman, 1978). Frequently polar metabolites are formed, and some reports have noted that the products are concurrently bound in soil systems (Williams, 1977); and Sethunathan, 1978). Volatilization generally decreases in flooded soil systems, and 14CO2 production, particularly from ring- labeled compounds, has not been found from pesticides incubated anaerobically in any system (Williams, 1977; and Sethunathan, 1978). Also, degradation rates and pesticide half-lives have generally not been reported. A major frustration with much of the literature is that frequently well-controlled anaerobic conditions were not used. Soils, for example, which are merely flooded and incubated without .shaking may be aerobic, especially on the surface. Therefore, the object of the research in this chapter was to investigate the potential for anaerobic biodegradation of acetanilide 59 herbicides in a well-controlled anaerobic system. Two products were identified from each herbicide, polar products were noted, degradation rates were calculated and the apparent extent of binding of the herbi- cides (probably their biodegradation products) was determined. MATERIALS AND METHODS Instrumental analysis High-performance liquid chromatography (HPLC) analysis was done on a varian 5000 microprocessor-controlled HPLC fitted with a Rheodyne 5020 6-part injection valve which had a 20 uL sample loop. The analyti- cal column used was a Varian Partisil column (30 cm x 3.2 mm i.d.) containing 10 um reverse-phase C18 packing. A precolumn (6 cm x 0.5 mm i.d.) which was dry packed with 30 to 40 um Lichrosorb reverse- phase C18 packing preceeded the column. Components in the eluant were detected by either a Varian UV absorbance detector (254-um wave- length) or a Hitachi multiple wavelength, single beam UV absorbance detector operated at 220-nm wavelength. Ratios of methanol-water for HPLC solvent systems were (1) 80:20 for diethatyl analysis, and (2) 75:25 for metolachlor analysis; the flow rate was 2 mL/min. A Spectra Physics 4100 computing integrator monitored retention times and peak areas during HPLC analysis. The concentration of herbi- cide in each sample was computed using external standards and a stanr dard curve based on peak areas. HPLC separation of diethatyl metabolites was obtained using reverse phase separation and a solvent system of methanol-water (70:30). Separa- tion could be increased by altering the methanol-water solvent system 60 to ratios of 65:25 (diethatyl metabolites) or 63:37 (metolachlor metabo- lites). ‘Metabolites of both herbicides were only visible at 220 nm although the parent compounds are also visible at 254 nm. Gas-liquid chromatographic analysis (GLC) was done on a Perkin- Elmer 900 gas chromatograph with a flame-ionization detector. A Porapak Q column (1 m.x 3.2 mm) was used for methane analysis. Operating conditions were: detector and injector, 1500 C; oven 600 C; carrier gas flow rate, 30 mL/min. Peak areas were measured using a Spectra- Physics Minigrator. For the parent herbicides and their metabolites, a 2 m x 2 mm (i.d.) glass column packed with 32 SP2100 on 80/100 mesh Supelcoport (Supelco, Inc., Bellefonte, PA.) was used. Operating conditions for temperature programmed runs to separate metabolites were: detector and injector, 2500 C; oven 125-2100 C (60/min) with a gas flow rate of 30 mL/min. Chemical and electron ionization mass spectra were obtained with a Hewlett-Packard 5985 gas chromatograph-mass spectrometer (CC/MS). Isobutane was used for chemical ionization mass analysis (CI), while electron impact mass spectra (EI) were done with an ionizing voltage of 70 eV. The column and column conditions used for GC/MS were compar- able to those described above for separation of the parent herbicide and herbicide metabolites. High resolution mass spectra were run on a Kratos MS 9/50 mass spectrometer. The metabolites from each herbicide were added as a mix; they were not separated or purified prior to obtaining the high resolution data. 61 14 Liquid scintillation counting (LSC) for C-label was done on a Beckman 8000 liquid scintillation counter. An internal 137Cs stan- dard produced an "H#", which is a measure of the energy of the Compton's edge and correlates directly to the counting efficiency of the sample. A standard curve produced from prepared quenched standards correlated the H# to the counting efficiency. Beckman Economy Premix was used to count 0.5 mL samples from the total sample and the methanol and water phases after separation. The samples were counted in 7-mL glass vials which were held in polypropylene shell vials while in the counter. Reagents Glass-distilled Burdick and Jackson or Baker Resi-analyzed aceto- nitrile and methanol were used for extraction procedures and HPLC analysis. All solvents for HPLC analysis were liquid chromatography grade reagents. Deionized water was prepared by charcoal filtration, deionization and redistillation of distilled water. Propylene oxide for sterilizing sediment was purchased from Baker Chemical Co. The acetanilide herbicide diethatyl [2-chlorofiN72',6'-diethylphenyl) fomethyl(ethylcarboxylate)acetamide] was obtained from Boots/Hercules, Inc., Wilmington, DE, as a formulation. It was purified according to the method described for alachlor by Tiedje and Hagedorn (1975). Diethatyl was recovered as white needle-shaped crystals after concen- trating the hexane extract in a flash evaporator and freezing it. It was recrystalized until it was 99.92 pure by GLC. Also, 14C-U— ring-labeled diethatyl (specific activity, 0.297 uCi/mg) was provided by Boots/Hercules, Inc. Metolachlor [2-chlorofo(2'-ethyl-6'-methyl- phenyl)f§f(Z-methoxy-l-methylethyl)acetamide], a liquid, was provided 62 by Ciba-Giegy Corp., Agricultural Division, Greensboro, NC. It was 97.52 pure by GLC analysis. 14C-U-ring-labeled metolachlor (1.1 mg; specific activity, 26.4 uCi/mg) was also provided by Ciba-Giegy Corp. Solutions of diethatyl and metolachlor were prepared by dissolving 0.1 g of each herbicide in 1 L deionized water; this yielded 100 ppm or 0.32 mM diethatyl, 0.35 mM metolachlor. The 100 ppm.solutions were used for addition to sediment or for preparing standard curves for HPLC analysis. Radiolabeled diethatyl and metolachlor were dis- solved in ethanol: diethatyl, 4.00 x 107 7 dpm/mL; metolachlor, 4.01 x 10 dpm/mL. Solutions containing radiolabeled herbicide were prepared by adding the appropriate amount of 14C-Uering-labeled herbicide dis- solved in ethanol to a flask, evaporating the ethanol, and adding deionized water and unlabeled herbicide to obtain a final concentra- tion of 100 ppm. Final concentrations of radiolabeled herbicides 3 in standard solutions were: diethatyl, 9.01 x 10 dpm/mL; metola- 3 chlor, 5.27 x 10 dpm/mL. Fortification Sediment from the pelagic zone of a eutrophic hardwater lake in Michigan (Wintergreen Lake, Kalamazoo County) was obtained using an Eckman dredge. The sediment pH was 6.5 and consisted of 652 organic matter, 35% marl, and some clay. Mean temperatures of the sediment range from 150 C (summer) to 40 C (winter). Additional character— istics of the sediment and of the sedimentary seston are described by Molongoski and Klug (1980a and b). Sediment was weighed in 50 g portions (about 50 mL) into 160-mL serum bottles under aerobic con- - ditions using a cut-off lO-mL plastic syringe to avoid excessive mixing 63 and subsequent saturation of the sediment with air. Sterile samples 'were prepared by the addition of 0.1 mL propylene oxide per gram wet weight of sediment (5 mL total) to the stoppered bottle, then held at 4° C for 2 to 3 days. Following the sterilization period the propylene oxide was removed by flushing the bottles through a vent needle in the stopper with a mixture of CO :N2 (flow rates: CO ° 2 2° 20 mL/min; N2: 80 mL/min). Procedures for preparation of the sedi- ment, fortification and flushing are outlined in Figure 1. Diethatyl and metolachlor studies The 50 g samples of sterile and nonsterile sediment in bottles were unstoppered and flushed with a mixture of filtered (to sterilize) 02-free 52 C02: 95% N2 by inserting a canula into each bottle (Figure l). Flushing continued in each bottle for 3 to 5 min. to insure anaerobio- sis prior to the addition of 50 mL of the herbicide solution. The herbicide solution was also flushed with the same anaerobic gases to insure that no 02 was added with the herbicide. Herbicide was added by anaerobic pipet to achieve a final concentration of 50 ppm in the sediment. The concentration of 14C-radiolabel was: diethatyl, 4500 dpm/mL; metolachlor, 2635 dpm/mL. Bottles to serve as controls received deionized water instead of herbicide. After the herbicide solution was added, the bottles were restop- pered by pushing the stopper partly into the bottle, then removing the canula, pushing the stopper firmly into place and sealing the bottles with an A1 crimp seal (Figure 1). This procedure produced an environment which was initially anaerobic and insured that the 64 ofiooummom anoulmooa How uaoaaooo wcfiowmuooo moHuuon Epsom mo cowuouoeoue mo amuwmfio scam .nuoosfiuooxo cowuoosooa .H shaman .pu<¢—uu az< n4<_> uu.m—¢usoce —l n aa< .< "au~—a—¢upu 9 . 59:3... .. 3 . mQOIhwz 65 sediment was never exposed to oxygen during preparation, herbicide addition, or incubation. To determine the amount of radioactivity added per milliliter of fluid in bottles containing sediment, buffer samples containing 6.5 g of glass beads were made to 50 g total weight with 0.02 M phosphate buffer, pH 7.0. The glass beads simulated the solid matter remaining after sediment was dried at 1040 C. (Solid matter was 132 of the initial wet weight of the sediment.) The buffer samples were amended with 50 mL of 100 ppm herbicide solution. The total 140- label (dpm/mL) in the buffer samples was used to calculate the 14C-label present in the total, methanol and water phases of the sediment extracts. Two major experiments were run, the first was incubated for 8 weeks and the second for 4 weeks; all treatments in the eightdweek experiment had five replicates. Results were so uniform between replicates, however, that only duplicate samples were used in the four-week experiment which was sampled at 0, l, and 4 weeks. Sediment for the eight-week experiment was collected in February, while that for the fourdweek experiment was collected in April. Also, since preliminary experiments indicated that over 90% of herbicide added was extracted at zero time, no zero time samples were done for the eight-week experiment. Samples were incubated at 280 C without shaking. Methane pro- duction was measured weekly after the samples had been shaken vigorous- ly to insure equilibrium of methane in the sediment with the headspace. 66 To determine whether the decrease in the rate of methane production noted in the above experiments was dependent on herbicide concentration, 50 mL solutions of unlabeled diethatyl or metolachlor were added to sediments to obtain samples having 0, 5, 10, 25, and 50 ppm herbicide concentrations. Extraction procedures R The Sep-Pak C18 method, described in the previous chapter and outlined in Figure 2, was used to extract and clean-up the herbicides and their metabolites. To extract the sediment, the bottles were opened and 25 mL acetonitrile was added to the sediment/herbicide solution. The final concentration of acetonitrile to sediment solu- tion was about 20%, v/v. The bottles were restoppered, shaken for 30 to 40 seconds, reopened and the contents poured into stainless steel centrifuge bottles. The solution was centrifuged at 8000 x g for 15 min. A small portion (0.5 mL) was removed to count the total 14C-label recovered while a second, 3 mL portion was added to a Sep- R Pak c18 cartridge (Waters Assoc., Milford, MA) containing C18 reverse phase packing. The cartridge was coupled to a luer-lock syringe so that the sample and solvents could be sequentially pushed through the cartridge. Polar ccompounds and polar herbicide products were eluted through the Sep-Pak C18R cartridge (termed water phase) when the initial 3 mL sample was added to the cartridge, while the more nonpolar, hydrophobic compounds such as the parent herbicide and some herbicide products were retained on the packing. Subsequently, the R nonpolar compounds were eluted from the Sep-Pak C18 cartridge with 67 3 mL of methanol (termed methanol phase). A portion (0.5 mL) was removed from both phases to count 14C-label, while HPLC analysis for the parent herbicides was done on the methanol phase only. The remaining supernatant was combined and refrigerated at 4° C for future analysis. A 100 mL sample of combined supernatant frac- tions from the eightdweek samples of each herbicide was extracted by three successive extractions with 33 mL hexane. The hexane frac- tions were combined, dried with anhydrous sodium sulfate, and concen- trated to a small volume in a flash evaporator. The sample was trans- ferred to a 26-mL glass bottle, concentrated further under a stream of N2 (N-evap, at 600 C water bath) and brought to 10 mL with hexane for GLC and GC/MS analysis. A small portion.was removed from the 10 mL hexane sample, brought to dryness using the N-evap, and redissolved in an equal volume of methanol for HPLC analysis. The pellets which remained after the supernatants were removed from the centrifuged samples were re-extracted with 100 mL acetonitrile, and centrifuged at 8000 x g for 15 min. The supernatants (termed 1002 acetonitrile extract) from the replicates were combined and 0.10 and 0.25 mL aliquots were removed from each for LSC. The small samples were used for LSC due to the dark brown color of the 1002 acetonitrile extract which caused extremely high quenching. Extraction and analysis procedures are outlined in Figure 2. 68 .ousvooouo mfimhaman was dowuumuuxo mo amummav scam o m c .N musmfim n u — mzohw uvp hzaou wzopm m—m»4u=o .o .oldu one» as ovuuuouus accuse can «ooonuu0— uo auo>ooou «no vulauo< .0 .ou«l«~ can-unease sown: as: aw ouaooon one» no vouuooou no: conuauuouucoo «sauce «Au coca. axon: usuuo an couuouucuocou «Auonuunv ecu you son: on: :«A: no o:~a>on .cuuv nooanusuuo no: uu500 vuouoaoo use 0:. .3003 «£000 a. none can: nanosuuuo use non .uon no I u Ihou «so no oouuuoau «nuuoocoexo ca uou nude Joualusmuu 0:. nu300 00¢«Al00 0c. .nuswuo .nuao0 uo uqu counse- uu-ua .I 0.0 00. «a. 0.«a o0 0.n 0a. ««. «.«0 A0 0.n 00. 0a. «.00 0 «nuohn00u 0.0 00. «A. «.00 U0 «.0 oo.« «u. 0.no o0 0.n 00. 0«. 0.00 0 docasoo2100u 0.0 «o. «a. 0.«0 00 0.0 co.« «a. 0.00 00 «.n 00. a". 0.00 0 can: 0002004090: «.0 0s. 00. 0.0« 00 0.5 as. oo. «.n0 00 «.0 no. «a. «.00 0 uauoauu0u 0.0 no. «a. c.«« 00 0.0 00. 0g. n.«0 00 o.« «0. 0«. 0.«0 0 «oousuuzno0n «.« no. «0. o.on« o0 «.« co. «0. 0.00m 00 n.« oo.~ «0. 0.0o~ 0 Damn A>H<=hu~a “020030 mu «u Annxmusv I anv o0 nuh<0=uz~ m~mrau:u «ouucocoexo on on oooauaoeeuouv nousuauouul 0:. shunguowv uo auav 0o uuu now-30a amend .« canoh 82 rate constant (week-1) and t is the time in weeks that the samples were analyzed. A half-life prediction was derived from the rate constant, k. An extraction efficiency of 95% was assumed for the zero time samples of the eight-week experiment since no zero time samples were extracted. From other experiments, the average percent 14C-label recovered at zero time in the total and methanol phases and the percent of parent herbicide determined by HPLC analysis was 95.8% i 6.02. These calculations indicate that the half-life of diethatyl was from 1.2 to 1.5 weeks, whereas that for metolachlor was from 3.7 to 6.0 weeks. The half-life for nonpolar extractable compounds includ- ing the parent herbicide was indicated by the 14C-label in the methanol phase and was between 2.9 and 5.5 weeks for diethatyl and between 3.8 and 6.2 weeks for metolachlor. Finally, analysis of the total 14C-labeled extractable 14C-label indicates that the half-life of the ring may be between 4.2 and 8.3 weeks for diethatyl and 3.8 and 6.4 weeks for metolachlor. The coefficient of determination is greater than 0.90 for all least squares fit of the data except for measure- ments of l4C-label in the methanol and total phases of diethatyl eight- week or combined four- and eight-week data (Table 2). It appears that while diethatyl itself may be transformed twice as quickly as metolachlor under anaerobic conditions, the rate of disappearance of the ring label of both herbicides is approximately equal. Lastly, transformation rates may be slower at in §i£u_temperatures (4 - 100 C) than the rates determined at the temperature (280 C) at which these experiments were run. 83 Identification of metabolites In the text below, the metabolites are referred to by the numbers assigned to them on the chromatograms (Figure 6), but are preceded by a "D" if it is a metabolite from diethatyl or an "M" if it is from metolachlor. The GC/MS data are presented as m/z followed in parenthe- sis by the peak intensity as a percent of the base peak (Table 3). Background was subtracted from all GC/MS data by using data from one point on either side of the metabolite peak. Peaks which show isotopic distributions are also presented (Table 3) while calculations of the isotopic distributions for parent peaks and ion fragments are presented in Table 4. The computer on the Kratos MS 9/50 was used to calculate the isotopic distribution from the molecular formulas of parent ions and is compared to the isotopic distributions1 obtained by EI and CI for parent and fragment ions. Several minor peaks in the GC/MS data of each metabolite are presented because they represent the isotopic distribution of the compound and indicate the presence or absence of unusual elements. GC/MS data and fragmentation schemes are located in Appendix A, Figure 3 (a through d). High resolution mass spectral data are presented in Table 5. The Kratos MS 9/50 measures the m/z of the ion fragments and calculates potential molecular formulas for each ion fragment detected in the. scan. Elements included in the calculations are submitted to the 1The intensity of the peaks for ions having one and two mass units higher molecular weight (P+l and P+2 or M+l and M+2, respectively) than.the parent (P) and fragment (M) ions was calculated and is given as a percent of the parent or fragment ion intensity. 01 D2 ml: 279 278 277 262 268 236 232 206 188 162 160 61 325 323 280 279 278 263 262 252 251 250 235 236 217 216 206 190 189 188 162 91 63 61 (21‘) ( 1.5) ( 13.6) ( 68.6) ( 2.2) ( 16.6) ( 30.7) ( 13.2) ( 81.6) ( 8.0) (100.0) ( 25.5) ( 12.5) ( none) 0.2) 3.3) 0.7) ( 1.8) ( 12.0) ( 20.3) (100.0) ( 0.2) ( 0.6) ( 3.3) ( 12.6) ( 19.5) ( 3.0) ( 19.7) ( 38.1) ( 3.3) ( 10.6) ( 79.7) AAA ( 66.6) ( 12.0) ( 2.6) ( 66.8) Table 3. 8i) Maaa apectra1 data (21) for diethatyl metabo1itea D1 and D2 and metolachlor metabolite: M1 and M2. fragment (m.w.)b M - C! M - C3368: . u - c3300 M - €33C320 ml: 232 - CD ml: 268 ml: 236 ml: 206 ml: 188 - Cl - CH CE 0! CO MUN N ml: 278-- 00 ml: 262 - CH CH ml: 262 ml: 278 ml: 262 ml: 250 ml: 235 ml: 250 ml: 216 ml: 236 ml: 190 CH 5C3 - Cl 3C! 3 2 - CIBCHZOH - 3c]3 - 0328633 CH3C3203 (15) (27) (63) (65) (28) (66) (66) (12) (28) (65) (61) (28) (27) (28) (61) (66) (66) (66) (61) (28) (66) (28) RI - relative intensity of the ion fragment fragment 10:: from parent peak or intermediate ion fragment to obtain ion II: 269 206 177 162 91 61 297 296 295 252 251 250 236 226 223 177 162 91 61 (11‘) ( CI ) ( 65.8) ( 8.9) (100.0) ( 5.6) ( none) (< 0.1) ( 0.2) ( 1.0) ( 1.6) ( 6.7) ( 26.7) ( 15.0) ( 0.9) ( 2.6) ( 15.2) ( 7.3) ( 6.3) (100.0) ( 6.6) ( 22.9) fragment (m.w.)b “4' H - C330C32 ml: 206 - 0830! ll: 177 _- C33 M+ H - C330C32 M - CB3SC32 ml: 250 - CHZCH ml: 236 - C320 ml: 223 . cazs ml: 206 - CHZCHCH CE3SC32 (65) (27) (15) (65) (61) (27) (30) (66) (62) 85 instrument by the operator. As the ion fragments for each metabolite had already been determined by CI and EI GC/MS, the appropriate mass measurements (as m/z) for the major ion fragments of each compound were extracted from the high resolution mass spectral data. Structures for all major ion fragments found by El analysis were previously derived and the molecular formulas of these structures were compared to those calculated by the Kratos MS 9/50. The Kratos MS 9/50 calcu- lates the molecular weight to the forth decimal place for each formula it derives for each measured mass. The calculated mass is subtracted from.the measured mass to obtain the millimass unit, MMU. Parts per million (PPM) are obtained by dividing the MMU by the measured mass. By convention, a formula is considered correct if it is within 10 PPM of the measured mass. In this case, if more than one calculated formula was within the 10 PPM range or the molecular formulas derived from the El data were outside the 10 PPM range, the additional formulas derived by the Kratos MS 9/50 could be eliminated because calculations indicated there were too few double bonds (< 6.0). All fragments had to contain the phenyl ring, which means that any possible struc— tures derived from the formulas calculated by Kratos MS 9/50 had to have at least four double bonds. Most formulas could be eliminated by these criteria (superscripts in data in Table 6 indicate where this occurred). Overall, the high resolution mass spectral data strong- ly supports the structures derived for the metabolites. The parent ion of diethatyl metabolite 1 (D1) was m/z 162. The parent ion was confirmed by CI data. There was no chlorine. Mass spectral data for D1 are in Table 3. A tropilium ion, m/z 91 (12.5), 86 was also present and indicates the presence of the phenyl ring. The. isotopic distribution of the parent and selected fragment ions indi- cated that no unusual elements such as sulfur or chlorine are present (Table 4). High resolution MS data confirm this (Table 5). The molecular ion of this metabolite was 34 atomic mass units (amu) lower than that for the parent herbicide, although the fragmentation pattern was similar to that for diethatyl. The loss of 34 amu indicated that the chlorine (35Cl) was removed and replaced with a proton (reductive dechlorination). The compound was positively identified as §f(2',6'- diethylphenyl)f§fmethyl(ethylcarboxylate)acetamide (C N03), 16H23 Figure 10. The parent ion of metolachlor metabolite 1 (M1) was m/z 249 (0.132), and was visible by CI only; mass spectral data are in Table 3. Similarly to metabolite Dl, isotopic distribution of the parent and selected fragment ions showed no unusual elements such as sulfur or halogens (Table 4), and the molecular formula was confirmed by the high resolution MS data (Table 5). As with diethatyl.metabolite Dl, the molecular ion for M1 was also 34 amu lower than that for the parent herbicide, indicating reductive dechlorination with the addition of a proton. The structure was identified as Er(2'~methyl-6'-ethy1phenyl)f§§ (2-methoxy-l-methyl)acetamide (C N), Figure 10. 15H23°2 Metabolites D2 and M2 are interesting because they both contain a thiomethyl group in place of the chlorine. The mass spectral data of both thiomethyl metabolites are discussed together below because certain important components which support the presence of the thiomethyl group are common to the fragmentation patterns of both. 87 . 003nm! mega-now nou ooov no: cowuannnnmqv omouoau uo conuousoamo neuamloo .o .ooquouaofiou nounoaou up some «+0 oz .0 .hdouonaooo ano> venomooa on noooeo axon: .onouonosn «Hm >0 canoe nuance»:« sou ano> .n -- -- 50.5 .. can emu z~oo~=55o .. .. on.n m nNN nna mzon5=~5o .. .. ou.n a «nu can mzoc~=05o 5a.. so.“ .ma.. m nan n65 mzaon~=65o 5: . . . n «N n5 as «n 55 am an me 56 no“ man 562 c a o no5=uunoou= .. .. no.5 .. ocN ecu zoa5=n5o no no.5 a--- .. 55a sen z~6n~=n5o 5: u- -- 6a.5 .. can Nam enoo~=n5o .. u- on.“ m «an own mzoo~=05o .. n- as.“ m emu an“ mzucc~=n5o mm.e 55.6 Ne.n a nun n~n memon~=n5o «a oe.~n no.~n no.a~ 5o n5m 55m 5ozno-nc5o 5nuaeuo5a u- .. c..~ .. can «an zo¢5=n5o no ~6.~ ~5.~ -- gnu new anon~e65o me .0400 Ho «0 « acumenm «+3 no ax! use! maomnoh nauaam no «+0 ex! Iwonu non nonaoouox announce moo0n300num00 admouomu ocunoflso no noonmm .ouov a: no 0cm «0 aonu 0cm .onxa 0x nonenx 0:0 >0 vouoaaoaoo moouuaannumwv unmouomu .0 names .«x won H: 0cm .«0 won .3 mouHHonmuoa nwosn 055m noflnooaonoa 055m 730553.20 mo mononoonnm .OH 955505.55. Saw a)»: «2 (O/ 855.055 1520105505.. 0 05.. :Nm a)»: «0 (DC 55.055100205108555: m m .505 85:... .2 (O/ - 550201055505: 0 05.. .555 85a. .0 /C< _ 5152056005105: 0 o .555 .55. 5.35.0320... 5 5 .0 100.2020 :00 I 0 05.. N 2:.” «\E. ._.>._. xN’Q ° g2 ~CHz 92 CH2 0 is observed from several of the fragment ions from D2 and M2 (see MS fragmentation schemes, Appendix A). Also, the mass spectra of D2 and M2 each contain strong peaks at m/z 61 having intensities of 46.82 for D2 and 22.9% for M2. There are no peaks at m/z 61 in the spectra of the other metabolites or parent compounds. Silverstein and Bassler (1976, p 28) state the "methyl primary sulfides cleave + at the o8 bond to give the m/z 61 ion, CH -S-CH2, which itself is 3 9O Table 5. High resolution mass spectral data: calculations of molecular formulas for parent and fragment ions. Metabo- Measured Inten- c H N 0 s MMU PPM lite Mass* sity D1 277.1687Pg 19.0 16 23 1 3 0 0.9 3.4 248.1301 4.6 14 18 1 3 0 1.5 5.9 232.1349 3.8 14 18 1 2 0 1.1 4.8 02 323.1552P 1.9 17 25 1 3 1 -0.3 -1.0 278.1225M 01 2.0 15 20 1 2 1 1.0 3.6 262.1488 ’ 55.4 15 20 1 3 0 4.5 17.18 250.1266 0.3 14 20 1 1 1 0.1 0.2 235.1564 4.7 14 21 1 2 0 -0 8 3.4:: 13 17 1 3 0 --- --- 234.1510 19.7 14 20 1 2 0 1.6 6.7 217.1058 4.7 13 15 1 2 0 -4.5 -20.53 216.1024 31.9 13 14 1 2 0 -0.0 -0.2 204.1377 34.2 13 18 1 1 0 -1.2 -5.7 190.1201 3.7 12 16 1 1 0 -3.1 -16.2c 189.1120 20.8 12 15 1 1 0 -3.4 -17.8c 188.1073 100.0 12 14 1 1 0 -0.2 -1.0 162.1277M’D2 90.2 11 16 1 0 0 -0.6 -3.8 160.1125 58.8 11 14 1 0 0 -0.1 -0.8 147.1056 31.4 10 13 1 - 0 0 0.8 5.2 146.0977 50.4 10 12 1 0 0 0.7 5.0 M1 249.1722P 1.0 15 23 1 2 0 -0.7 -2.8 204.1386 5.3 13 18 1 1 0 -0.3 -1.3 M2 295.1568P 0.3 16 25 1 2 1 -3.8 -12.9c 250.1249 3.2 14 20 1 1 1 -1.7 -6.88 234.1473 3.4 14 20 1 2 0 -2.1 -8.8 223.1021 2.1 12 17 1 1 1 -1.0 -4.6 177.115 2.4 11 15 1 1 0 -0.1 -0.8 162.1262“’M1’Mz 00.0 11 16 1 0 0 -2.1 -12.7a’f 162.0887 20.1 10 12 1 1 0 -3.2 -19.58 161.1153 5.3 11 15 1 0 0 -5.1 -31.8 146.0940 26.4 10 12 1 0 0 -3.0 -20.33 134.0959 8.3 0 12 1 0 0 -1.1 -8.1 91 Table 5, cont'd tt b 1’ Footnotes MMU - millimass units; e.g., if the mass measured was 323.1152, (D2, parent pead), the actual calculated molecular weight of the formula shown is within -0.3 MMU of the measured mass. To get the calculated molecular weight, subtract the MMU shown from the measured mass 323.1552 - (-0.0003) - 323.1555. PPM - parts per million; convention requires that all formulas chosen fit within 10 PPM of the measured mass intensity and, unless a superscript is present, the formula shown has the closest mass to the measured mass. It is computed as: MMU - PPM measured mass 6 ‘0'0003 = -1.0 x 10‘ . -6 (XIO )’ e°g" 323.1552 PPM was out of range (> 10 PPM); a different formula calculated by the Kratos MS 9/50 came within range but was eliminated because calculations indicated there were too few double bonds (< 4.0). b2. Formula calculated by the computer was within range (bl), but does not match the formula of the expected fragment ion (b2). PPM was out of range but no other formula was calculated by the Kratos MS 9/50. The formula shown fit the expected formula for the ion fragment; calculated double bond character also fit. All metolachlor metabolite high resolution mass spectral data was averaged from five scans, although some peaks were present in only two (m/z 295.1568), three (m/z 249.1722), or four (m/z 250.1249 and m/z 223.1021) of the scans. (Diethatyl data was from only one scan.) Other calculated formulas were within range, but contained chlorine. CI data definitely indicated that no chlorine was present. There is the possibility of a second structure with a different formula within mass range 162. 92 Table 5, cont'd "P" indicates parent ions, "M" indicates base peak ions from EI data; all others are fragment ions. M,Dl, M,D2 or M,Ml, M,M2 indicate which metabolite the base peak (M) belongs to. For D1, data for masses 262.1488, 234.1510, 204.1377, and 189.1120 to 146.0977 were identical to and included within the data for D2; for M1, data for masses 177.1152, 162.0887, 161.1153, 146.0940, and 134.0949 were also present in the data for M2. 93 visible as a moderate to strong peak." Further evidence for the presence of the Sfmethylthiomethyl moiety on the nitrogen is found by examination of the parent peak and certain ion fragments to which this group is still attached. Comparison of the isotopic distribu- tion of parent or ion fragments at m/z 323, 278 or 250 for metabolite D2 and m/z 295, 250, and 223, for metabolite M2 (Table 4), indicates that metabolites'DZ and M2 both contain a sulfur. The abundance of 3“ 32s is about 4.474. Therefore. in the mass spectra S relative to of any sulfur-containing compound, ions which retain the sulfur would have a sister peak two mass units Gn/z) greater than themselves which are at least 4.4% of the height of the "lower-mass" peak. So, for D2, the parent peak at m/z 323 and fragment ions which retain the sulfur Om/z 278 and m/z 250) were found to have sister peaks with isotopic distributions of 5.4 to 5.92 (Table 4). Another fragment ion which does not retain the sulfur, m/z 262, has an isotopic distri- bution of only 1.92. A similar result is seen for M2 (Table 4) which also has a thiomethyl group, D1 and M1, in contrast, do not contain the thiomethyl group and their parent peaks and major ion fragments have isotopic distributions of much less than 4.4%. 94 DISCUSSION .The transformation products of diethatyl and metolachlor are be- lieved to be identified as shown in Figure 10. No known mechanism can clearly be associated with the production of these products by microorganisms under anaerobic conditions. In the review below, dechlorination of compounds anaerobically is shown to be facilitated by cytochromes, especially cytochrome P450, in mammals, algae, and Pseudomonas putida. Secondly, dechlorination can be facilitated by the addition of a simple organic thiol compound such as glutathione, cysteine or methionine with subsequent cleavage by a C-S lyase to a thio- or a thiomethyl group. This elimination-addition reaction has only been shown to occur to plants and animals, and chemically under anaerobic conditions. A potential mechanism which may be imr portant to the addition of the thiol compound is presented, although some combination of the cytochrome-mediated reductive dechlorination or the eliminationeaddition reaction involving thiol compounds is possible. Simple reductive dechlorination has been found to occur anaero- bically on many compounds including DDT, toxaphene, and lindane. Casida and coworkers (Khalifa et al., 1976; Saleh and Casida, 1978; Miskus et al. 1965) have found that under anaerobic conditions, rat liver microsomes supplemented with NADPH can mediate the dechlorina- tion of toxaphene or DDT, suggesting the involvement of reduced cytochrome P They also found that these reactions can be mediated 450' nonenzymatically by reduced iron (II) protoporphyrin systems such as hematin. Esaac and Matsumura (1980, 1978, and Matsumura and Esaac, 95 1979) have found two systems in rat liver which anaerobically mediate reductive dechlorination. One system is heat labile and requires NADPH and the other is heat stable and requires FAD (which probably is reduced to FADH). Inhibition, induction, and spectral change studies have shown that the NADPHrrequiring system involved a reduced cytochrome P -mixed function oxidase system. 450 Matsumura and Esaac (1979) found the dechlorination activity of an FMN-mediated system in rat liver and intestine, two algal species (most active), and in Pseudomonas putida (least active). Dechlorinat- ed and sometimes water soluble products were found from parathion, toxaphene, mexacarbate, DDT, lindane and dieldrin after anaerobic incubation with FMN. This evidence coupled with that of Casida and his coworkers indicated that reductive dechlorination or hydrolysis may readily occur in anaerobic sediment. Dead intact or lysed algal cells which have sedimented or natural anaerobic flora in the sedi- ment could be involved in the reactions. A second mechanism of dehalogenation may be promoted by the nucleophilic attack of simple organic thiol compounds such as cysteine, 2-mercaptoethanol, or methionine. Sander (1979) has reviewed the nonenzymatic dehalogenation of substituted uracils via thiol-promoted reactions. Thiol compounds react with pyrimidine bases such as S-iodo- or S-bromouracil anaerobically under mild conditions of temperature and pH. The halogen is liberated with or without the addition of a thiol compound such as cysteine. Addition of the thiol compound generally occurs across a double bond or a single bond which can be resonance stabilized, and the product may or may not retain the thiol compmmd (Scheme 1) . 96 0 . fl . u . . “ a K “N511: HN\H2‘_.HN«£H xic"H'{ H 0:” H s 0‘kN H ,1” '01)): H o‘H H 0 H c” H S H 50y: H scn “N H *~ H 0 a SCya Scheme 1. E2 Hal (from Sander, 1979). Sander proposed that the "E2 Hal" (elimination) mechanism was the most reasonable process of those proposed. Subsequently, he postulated an enzymatic mechanism which would involve a nucleo- philic attack by an active-site thiol anion on a halogen atom via the E2 Hal process. The halogen would be eliminated and the active site of the enzyme regenerated by a reduced pyridine nucleotide such as NADP (Scheme 2): in IHN’QIrs—o‘I HN {4.31 0*“ H H84 3 04‘" H 1{\;s—A- H 514 9 ) ""2 ENZ HN’I” 0%: H ‘ H 's NAOPH -' ‘ 1' ts H '«E H 3‘— . 's ‘—d '5 N2 8812 ENZ Scheme 2. Enzymatic mechanism (from Sander, 1979). 97 Sander also showed that these mechanisms are not always applied across double bonds, as in the "SNZ mechanism" (not shown here). How- ever, the potential for resonance stabilization through keto-enol tautomerization may increase the possibility of a nucleophilic attack by a thiol compound. Such may be the case with the chloroacetanilide moiety of the acetanilide herbicides. Metolachlor and another acetani- lide, alachlor [2-chloro-2',6'-diethylf§f(methoxymethyl)acetamide], react nonenzymatically under anaerobic conditions with glutathione (Leavitt and Penner, 1979). The glutathione replaces the chlorine on the chloroacetyl moiety, forming a covalent bond between the sulfur on the glutathione and the methylene of the acetyl moiety. Alachlor also conjugates anaerobically with cysteine, dithiothreitol, or ethanethiol. This nonenzymatic dehalogenation may proceed via keto- enol tantomerization with addition of a thiol compound (Scheme 3): NAU’H R u H 1 I 1\N/C\C£Cl ___’ 1\N’c‘c/ _\_/_9 \N’E\C H R 2 H R 2 RS' H RS‘ Cl 2 RS H39 0 R‘ 3 H3COH2CHC'; HJCOCFlzc-g H3C0H2C- (R‘ NA) (RN/K) (RN/K) 18.-AG“ ff <2 Scheme 3. Keto-enol tautomerization of acetanilide moiety and conjugation with a thiol-compound. 98 In this scheme, the acetyl group becomes conjugated to the thiol compound. However, it may be possible for the reaction to proceed without retention of the thiol compound, as in the E2 Hal mechanism, above. These reactions are interesting but conjugates of cysteine, methionine, glutathione or other thiol-compounds would require further alterations to form the methylthiol derivative. Various mechanisms have been proposed by different researchers. Lotlikar et al. (1966) found that esters of thydroxy amines and amides of 2-acetylaminofluorene formed a methylthio derivative in_vi££g under anaerobic conditions through reaction with methionine or methionyl- glycine. The products were relatively unstable and subsequently under— went an "internal nucleophilic attack of the carbonyl group on the y-carbon with the formation of homoserine lactone which hydrolyzes to yield the [3-methylthio] derivative (scheme 4): . Q (g 0209090310 methionyl- / NICCHa CH3 -CH 000' “Wm ‘H I _3_, A—. IS-CH3 ‘\ OCCH H C O u 3 2 Wu 0 H3: ,c-FN 2 HCH2c00H )R H NH2 9 ° ’CCH3 "2%-? CH c00H . H' 6}"\“ ’ Hzflcfi’o . ”ZN 2 SCH3 I ‘N H H2 Scheme 4. Hydrolysis of methionylglycine conjugate of an ester of 2-acety1aminofluorene to form a methylthio conjugate (from Lotlikar et al., 1966). 99 Cu et al. (1977) found that rabbits formed a urinary metabolite from the oral administration of 3-(5-nitro-2-fury1)-2-(2-furyl)acrylamide which was identified as 2-(B-carboxypropionyl)-3-(S-methylthiol-Z- furyl)acrylamide. They proposed that the introduction of the methyl- thio group (SCH3) on the furyl moiety was from methionine, directly, or from cysteine which was cleaved off and later methylated (Scheme 5): .CCH COOH Ucm 2 H 3cs C‘CNHZ 0 Scheme 5. Thiomethyl product, 2-(B-carboxypropionyl)-3- (5-methylthio-2-furyl)acrylamide. Tateishi et al. (1978) discovered that a methylthio conjugate of bromazepam was formed in 22££2.by the action of §fadenosyl methionine (SAM). The mercapturic acid-bromazepam conjugate was hydrolyzed at the sulfur, transferring the sulfur to the bromazepam moiety and sub- sequently adding a methyl group to the sulfur (Scheme 6): :1 . a. 5) I \ ' SCH3 Scheme 6. Methylthio bromazepam (Tateishi, 1978). 100 There are numerous examples of conjugate formation between acetani- lides, especially propachlor, and simple thiol compounds in rats (Bakke, 1980) or glutathione in other mammals, birds, insects, and plants (Hutson, 1976). In comparing the metabolism of propachlor in germ— free, bile-cannulated, and normal, control rats, Bakke and coworkers (1980), concluded that the rats formed water-soluble glutathione conju- gates which were cleaved to form a mercapturic acid-conjugate prior to elimination through the urine. However, the gut flora of the normal rat transformed the gluthathione-, cysteine-, and mercapturic acid- propachlor conjugates to many other various products including a simple methylthio product: 0 (RI-N-C CH R2 2SCI-13). Thus, a conjugate formed from a simple thiol compound in the sediment may be reduced to a methylthiol compound by indigenous microflora. An interesting enzyme, L-methionine y-lyase, has been isolated from Pseudomonas ovalis by Tanaka and his coworkers (1977). This type of enzyme could be responsible for the cleavage of herbicide-thiol compound conjugates to a thiol or methylthiol. L-methionine y-lyase catalyzes and elimination reaction on several amino acids such as L- methionine, L-ethionine, Sfmethyl-L-methionine, DL-methionine sulfone or sulfoxide, L-methionine-DL-sulfoximine, or L—homocysteine. In the case of L-methionine, the enzyme releases o—keto butyrate, ammonia, and methanethiol. The lyase can also catalyze replacement reactions, such as the transfer of the thiomethyl group of L-methionine to lOl ethanethiol to get L-ethionine. The enzyme was also capable of re- placing the methylthio group of methionine with thiol-substituted alkanethiols or arylthio alcohols (See Scheme 7). The lyase replace- ment activity was only tested on thiol-substituted compounds, but it CHZSH C CHCOO . SCH H “3C 2H2“ r toluene- . . 2 g 3 ° (hiol L- methnonme )_ 3 § ‘< 3 a D 5- a 2 :1 E 5‘ n . v CHZSCH2CHZCHCOOH CH3$H , NHZ unanno- tMol Scheme 7.. Replacement of the methylthio group of methionine with an arylthio alcohol (from Tanaka et al., 1977). would be interesting to see if it operated on compounds with other substituents such as a chlorine group. This lyase could also do a B-replacement reaction between Sfmethyl-L-cysteine and ethanethiol to yield Srethyl-L-cysteine and methanethiol. The lyase from Pg. ovalis catalyzed these reactions anaerobically, under nitrogen, although those of other microorganisms tested (one strain each of the following was tested: Bacillus, Agrobacterium, Alkaligenes, Aerobacter, and Brevibacterium) did not operate anaero- bically. The lyase activity of three of the eight Pseudomonas species tested were active anaerobically, although the lyase concentration in Pa. ovalis was the most abundant. 102 This type of enzyme might be able to do a replacement reaction between a chloroacetyl herbicide and L-methionine or L-cysteine or some other suitable thiol-substituted compound. It might also cleave an L-methionine leaving the methylthio group on the herbicide. Enzymes which conjugate glutathione to various pesticides are fairly widespread, occurring in mammals, birds, insects and plants (Hutson, 1976). Furthermore, glutathionejgftransferase from mouse liver (Guddewar and Dauterman, 1979), and corn and peas (Frear and Swanson; 1970, 1973) appear to operate well aerobically when reduced glutathione is added. To my knowledge, no enzyme comparable to glutathionefotransferase, "aralkyl transferase" or "acetyl-CoA acetyl transferase" which conjugates glutathione via the removal of the functional group (methoxy, halogen, etc), the breakage of an ester or ether bond, or transfer across a double or resonance stabilized bond (Hutson, 1976) has been demonstrated in microorganisms. From this study and the fact that these metabolites of acetanilides have not been detected in studies with soils or microorganisms performed under aerobic conditions, I assume that anaerobic conditions are necessary. A.methylthio product of the fungicide pentachloronitro- benzene (PCHB) was found in greenhouse (de Vos et al., 1974) and flooded or moist field soils (Murthy and Kaufman, 1978). The product was identified as pentachlorothionanisole (PCTA). As de Vos et a1. (1974) were only examining a number of greenhouse soils for residues of PCNB, there was no clear indication of incubation conditions. How- ever, soils used by Murthy and Kaufman (1978) were either moistened to 75% of 1/3 bar soil moisture content (moist soil) or flooded and 103 kept under a nitrogen atmosphere. PCTA was found under both conditions, but, upon coupling their evidence with that of others, they proposed that PCTA.was one mode of dissipation of PCNB from anaerobic soils. However, they made no speculations on the mechanism of formation in their report. So far we have shown that dehalogenation can occur under anaerobic conditions via a reduced cytochrome system, such as cytochrome P450, or through conjugation with a simple thiol compound. The former pro- cess would result in metabolites D1 and M1, while the latter, with further manipulation, might produce metabolites D2 and M2 (Figure 10). While no clear mechanism can be identified from these studies, especially for the formation of the thiomethyl product, D2 and M2, it is possible to speculate about what sort of thiol compounds of microbial origin might be involved and what type of enzyme might be responsible for the production of the final thiomethyl products. V01atile sulfur compounds are produced by microorganisms as end- products in the anaerobic metabolism of methionine, allylcysteine, cysteine, §7methylcysteine, dimethyl-B-propiothetin [(CH3)ZSCH2CH COOH], 2 etc. Compounds which are frequently produced are hydrogen sulfide, methylmercaptan (CHasH), and dimethylsulfide (CH3SCH (Kadota and 3) Ishida, 1972). Hydrogen sulfide is also produced by the reduction of sulphate in the sediment. Secondly, transformation could occur through chemical or bio- logical addition of the simple volatile thiol compounds mentioned above or their precursors. Guard et al. (1981) have reported that the methylation of trimethyltin hydroxide [CH3)38n0H] to tetramethyltin [CH3)4Sn] can occur through biotic and abiotic mechanisms in anoxic 104 sediment. While sterile samples were weakly active in methylating the (CH3)3Sn0H, methylation in nonsterile samples was significantly enhanced by the addition of acetate and sulfide (Na28-9H20). Further- more, a Lewis-base redistribution of the (CH3)38nOH to obtain (CH3)4Sn was induced by sodium thioglycolate in seawater. They proposed the following mechanism, where L is thioglycolate: (CH3)3SnOH + xL 2 (CH3) 3SnOHLx _). 2(CH SnOHLx ._ (C33)45n + (C33)ZSnOL + H 0 3)3 2 2 Although they believe this biomethylation is a minor process, the evidence for induction of the methylation by a sulfur-ligand with the production of both the methylated and conjugated products is inter- esting. Also, the addition of thioglycolate to the trimethyltin hydroxide appears to be a nucleophilic substitution which is chemically similar to the addition of glutathione, methionine, or other compounds to the acetanilides, as reviewed above. No clear mechanism.has been deduced for the formation of products D1 and D2 or M1 and M2. However, possibly a simple cytochrome-induced reductive dechlorination as well as a mechanism involving a sulfur ligand with a simple thiol compound could be involved. Mediation of the dechlorination reactions, especially addition of the thiomethyl moiety, appears to be biological. Conjugates of thiol compounds such as cysteine or Sfmethylcysteine to the herbicides could perhaps be cleaved to the thiomethyl by a C—S lyase such asethe one investigated by Tanaka et a1. (1977). However, production of volatile sulfur come pounds such as methylmercaptan is ubiquitous among microorganisms, 105 although I do not know its importance as a product in anoxic sediment; hydrogen sulfide production from fermentative decomposition is probably by far more important. Carbon dioxide production ‘C02 production was not monitored because many studies have shown that essentially no 14CO2 was produced from 14C-ring-labeled pesti- cides incubated in anaerobic soils (including rice soils), or in xitrg rumen liquor or rumen cultures, although some pesticides are metabo- lized [Roberts and Standen, 1978; Martens, 1978; Belasco and Harvey, 1980; Murthy and Kaufman, 1978; Ambrosi et al., 1977; Williams and Feil, 1971; and Sethunathan and Siddaramappa (review of rice soil), 1978]. Considering the high adsorptive capacity of this sediment (up to 60% disappears at zero time, although 95% or more can be extract- 14 ed with acetonitrile, Chapter I) the fact that little CO has been 2 found from the 14C-ring-labeled acetanilide, alachlor, even in aerobically incubated soils (Chou, 1974; Taylor, 1972), and the indica- l4 tions from many other researchers that CO2 is not formed from ring— 14 CO2 or 14GB.4 would be produced from complete degradation of the phenyl ring. labeled herbicides anaerobically, it is doubtful that any Instead, the products are assumed to be readily adsorbed or bound to the sediment. Herbicide disappearance Both herbicides showed extensive anaerobic degradation in non- sterile Wintergreen Lake sediment. The concentration of diethatyl was below detectable limits (1 to 3 ppm) in nonsterile sediment, 106 whereas 62% of the diethatyl remained in sterile sediment after eight weeks incubation. This transformation of diethatyl in the sterile sediment indicated that 38% of the diethatyl may have been degraded by nonbiological means, with the formation of bound and water soluble products. Metolachlor was not as extensively degraded: 34% of the parent herbicide remained in the nonsterile sediment and 92% in the sterile sediment. Disappearance of metolachlor in the sterile sediment indicated that about 82 of the degradation was due to nonbiological mechanisms while 58% was due to biological mechanisms. The predicted half-lives of diethatyl and metolachlor under anaerobic conditions in the sediment were 1.2 to 1.5 weeks and 3.7 to 6.0 weeks, respectively. The higher recalcitrance of metolachlor compared to diethatyl under anaerobic conditions was also typical of aerobic studies with soils and pure culture. Chou (1977) determined that the half-life of diethatyl was 0.4 and 0.6 weeks in soils under laboratory conditions. From data presented in the EPA.Metolachlor Pesticide Hegistration Standard, (1980), I calculated that the half— life of metolachlor may be 4.6 to 6.9 weeks under field conditions. Lastly, when the herbicides were individually incubated with the soil fungus, Chaetomium globosum, 552 of the metolachlor and none of the diethatyl remained after 144 hours (McGahen and Tiedje, 1978). The estimated half-life of diethatyl and metolachlor in pure culture were about 0.6 weeks and 1.3 weeks, respectively, although these values cannot be directly extrapolated to the environment. Structural differences between the two herbicides cause diethatyl to be more susceptible to biochemical as well as chemical alterations under aerobic or anaerobic conditions. I have found that diethatyl 107 is easily altered under acidic (pH 2 to 4) or basic (pH 8.5) con- Iditions and forms polar metabolites (my observations), but metolachlor does not. The chloroacetyl moiety of both may be expected to have the same stability; however, the ethylcarboxylate of diethatyl may hydrolyze easily at the ester bond, while the ether bond in the 2- methoxy-l-methylethyl group of metolachlor seems much more stable. Also, the more linear ethylcarboxylate group of diethatyl may allow enzymatic alterations to occur more readily than the analogous group of metolachlor which has a methyl branch. In the diethatyl experiments, when the amount of 14C-label in. l4C-label extracted and multi- 14 the water phase is normalized to the plied by 100 to obtain a percent, the amount of C-label in the water phase of the sterile sediment is almost exactly equal to that in the nonsterile for all extraction times (Table 1). For example, in the sterile sediment after eight weeks incubation, 33 and 842 of the 140- 1abe1 were in the water phase and the total sample, respectively, so 392 [(33/84 x 100 - 392] of the 14 14 C-label was in the water phase. Similarly, 402 of the C-label was in the water phase of nonsterile sediment at eight weeks. Comparison of these values between the sterile and nonsterile samples for each time point may indicate a steady chemical alteration of diethatyl. However, in absolute terms, a great deal more 14C-label was in the total sample and the water phase of the sterile sediment as compared to the nonsterile. Transformation of diethatyl in the sterile sediment may have been due, in part, to incomplete inactivation of available biochemical transformation mechanisms. Fletcher and Kaufman (1979) found that about 33% more pfchloroaniline was bound in ethylene oxide and/or 108 propylene oxide sterilized soils than in autoclaved soils. They also found that there was only slight inhibition of soil peroxidase activity by propylene or ethylene oxide as compared to 100% inhibition by auto- claving. Although part of the degradation of diethatyl to polar products may have been due to chemical alterations, it seems that the mechanism of alteration of diethatyl to a polar product or products was enhanced in nonsterile sediment, since, it terms of absolute concentration, there was less 14C-label in the total sample and water phase of the nonsterile sediment than in the sterile sediment. In nonsterile sediment, these alterations could be catalyzed by endo- or exoenzymes of live microorganisms. As there was almost no alteration of metola- chlor to polar products, I assume that that was due to the higher stability of the ether bond in the 2—methoxy-l-methylethyl moiety of metolachlor versus the relative instability of the ester bond in the ethylcarboxylate group of diethatyl. Also, since the chloroacetyl group, as well as the phenyl rings, of both herbicides should be chemi- cally nearly identical, it is clear that the formation of a polar product or products from diethatyl must be related to its ethylcaboxy- late moiety since no polar products are formed from.metolachlor. The formation, in nonsterile sediment, of the two extractable products from each herbicide was due to biological means since they were not found in the sterile sediment. It was also clear that the decreasing extractability of the 14C-labeled herbicides was due to biological 14C-label from activity. It is interesting to note that 55% of the diethatyl and 59% of that from.metolachlor was not recovered by the 202.acetonitrile extract. So, approximately equal amounts of both 109 herbicides were strongly adsorbed or bound in the sediment, indicating this occurred at approximately equal rates, although diethatyl was much more subject to biological or chemical alteration. The average half-life of the total extractable 14C-label for diethatyl was 6.6 i 2.1 weeks and for metolachlor was 5.6 i 1.6 weeks (Table 2) indicating that "disappearance" of both was occurring at approximately equal rates. Adsorption and/or binding, therefore, does not appear to depend on the rate of biological or chemical alter- ation of the herbicides, since virtually all of the diethatyl was transformed, but the amount of metolachlor remaining was only slightly less than the total amount of 14C-label extracted. It may depend on another transformation of coupling step which occurs more slowly than the degradation of diethatyl but perhaps at about the same rate as the transformation of metolachlor. It is not strictly dependent on the amount of organic matter because usually 84 to 882 of the 14C- label was recoverable with 20% acetonitrile in the sterile samples. Thus, it required some biological mediation to either create products which were more tightly adsorbed or bound or to mediate the binding. Inhibition of methane production Although the reason the herbicides cause a reduction in the rate of methane production is not known, it may be due to an inhibitory effect on heterotrophs which produce substrates such as short-chain (volatile) fatty acids, carbon dioxide, and hydrogen (H2), which the methanogens further utilize for energy and growth (Zehnder, 1978, and Zeikus, 1977). Syntrophic associations between the methane bacteria and other Hz-producing organisms have frequently been llO demonstrated (McInerney and Bryant, 1981, and Bogels et al., 1980). Therefore, significant alteration or a decrease in the amount of the products produced by heterotrophs which the methanogens utilize could affect methane production. Some researchers have found shifts in heterotrophic end-product composition in methane fermentors by the addition of trifluralin (Williams and Feil, (1971) or monensin (Slyter, 1979). However, in the former there was no evidence that the compounds were not affecting the methanogens directly and thus causing changes in the end-product balance, while in the latter, there was evidence that the monensin altered the hydrogen production and composition of fermentation products. However, Brock (p. 695, 1979) has noted that various compounds such as "acetylene (CH 5 CH), ethylene (CH2 - CH2), chloroform (CHCL3), carbon tetrachloride (CCLA) and DDT [l,l,l-trichloro-2,2-bis (p-chlorophenyl) ethane]" inhibit methanogene- sis "more or less specifically." Methane production is also inhibited by dichloroacetamide (Slyter, 1979). The inclusion of DDT in this list indicates that perhaps large molecular weight compounds can be taken up by the methanogens, although only 602, Hz, acetate, formate, and perhaps a few other small compounds are utilized as substrates. The inhibitors all seem to be CO2 analogs (including DDT). The chloro— acetyl group on the acetanilides may be inhibitory for the same reason, but there is no proof as to their ability to directly or indirectly inhibit methanogenesis. Conclusion The acetanilide herbicides diethatyl and metolachlor were de- graded under anaerobic conditions in Wintergreen Lake sediment. Two products were identified from each herbicide. One product from each 111 ‘was simply dechlorinated and the second product from each was dechlor- inated with the addition of a thiomethyl moiety. No specific mechanism for the formation of these products could be deduced. Addition of the acetanilides to the sediment resulted in inhibi- tion of methane production and was dependent on the concentration of the herbicides. After eight weeks incubation, the diethatyl concen- tration was below detectable limits, whereas 34% of the metolachlor remained. The half-life of diethatyl was about 1.2 to 1.5 weeks where- as that for metolachlor was 3.7 to 6.0 weeks. However, disappearance 14 of C-labeled ring of diethatyl (6.6 t 2.1 weeks) was close to that for metolachlor (5.6 i 1.6 weeks). 10. 11. 12. LITERATURE CITED Ambrosi, D., P. C. Kearney, and J. A. Macchia. 1977. Persistence and metabolism of oxadiazon in soils. J. Agric. Food Chem. 25:868-872. Bakke, J. E., J.-A°. Gustafsson, and B. E. Gustafsson. 1980. Metabolism of propachlor by the germfree rat. Science. 210:433- 435. Beestman, G. B., and J. M. Deming. 1974. Dissipation of acetani- lide herbicides from soil. Agron. J. ‘§§:308-311. Belasco, I. J., and J. Harvey, Jr. 1980. In vitro rumen metabo- lites of 14C—labeled oxamyl and selected metabolites of oxamyl. J. Agric. Food Chem. 28:689-692. Bollag, J. -M., and S. Russel. 1976. Aerobic versus anaerobic metabolism.of halogenated anilines of a Paracoccus sp. Microbial Ecol. 3: 65- 73. Brock, T. D. *1979. The biology of microorganisms, 3rd edition. 802 pp. Prentice-Hall, Inc., Englewood Cliffs, N.J. 07632. Chahal, D. S., I. S. Bans, and S. L. Chopra. 1976. Degradation of alachlor [2-chlor01Nf(methoxymethyl)-2',6'-diethy1acetanilide] by soil fungi. Plant and Soil. 423689-692. Chou, S.—F. 1974. Biodegradation of alachlor [2-chloro-2', 6'-diethy1-N-(methoxymethyl)acetanilide] in soils. ‘M.S. Thesis, Michigan State University. 22 pp. Chou, S.-F. 1977. Fate of acylanilides in soils and polybrominat- ed biphenyls (PBB's) in soils and plants. Ph. D. Dissertation, Michigan State university. 90 pp. de Vos, R. H., M. C. ten Noever de Brauw, and P. D. A. 01thof. 1974. Residues of pentachloronitrobenzene and related compounds in greenhouse soils. .Bull. Environ. Contam. Toxicol. 11:567- 571. EPA Metolachlor Pesticide Registration Standard. 1980. U.S. Environmental Protection Agency, Office of Pesticides and Toxic Substances, Special Pesticide Review Division, Washington, D.C. 20460. Esaac, E. G. and F. Matsumura. 1978. A novel reductive system involving flavoprotein in the rat intestine. Bull. Environ. Contam. Toxicol. 19:15-22. 112 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 113 Esaac, E. G., and Matsumura. 1980.. Mechanisms of reductive dechlor- ination of DDT by rat liver microsomes. Pest. Biochem. Physiol. 13:81-93. Fletcher, C. L. and D. D. Kaufman. 1980. Effect of sterilization methods on 3-ch1oroaniline behaviOr in soil. J. Agric. Food Chem. 28:667-671. Freer, D. S. and H. R. Swanson.- 1970. Biosynthesis of §f(4- ethylamino-6—isopropyl-amino-ngftriazino) glutathione: partial purification and properties of a glutathione §7transferase from corn. Phytochemistry. 9:2123-2132. Frear, D. S. and H. R. Swanson. 1973. Metabolism of substituted diphenylether herbicides in plants. 1. Enzymatic cleavage of flurodifen in peas (Pisum sativum L.)‘ Pest. Biochem. Physiol. 3:473-482. Guard, H. H., A. B. Cobet, and W. M. Coleman III. 1981. Methyla— tion of trimethyltin compounds by estuarine sediments. Science. 213:770-771. Guaddewar, M. B., and W. C. Dauterman. 1979. Studies on gluta- thionefigftransferase preparation from mouse liver which conjugates chlorofigftriazine herbicides. Pest. Biochem. Physiol. 12:1—9. Hargrove, R. S., and M. G. Merkle. 1971. The loss of alachlor from.soi1. Weed Sci. 912:652-654. Hutson, D. H. 1976. Glutathione conjugates. Pages 103-131-gg D. D. Kaufman, G. G. Still, G. D. Paulson, S. K. Bandal, eds. Bound and conjugated pesticide residues. American Chemical Society Symposium Series, vol. 29, washington, D.C. Kadota, H., and Y. Ishida. 1972. Production of volatile sulfur compounds by microorganisms. Annual Review Microbial. 26:127- 138. Kaufman, D. D., J. R. Plimmer, and J. Iwan. 1971. Microbial degradation of propachlor. Abstract, 162nd meeting American Chemi- cal Society. A21. ’ Khalifa, S., R. L. Homstead, and J. E. Casida. 1976. Toxaphene degradation by iron (II) protoporphyrin systems. J. Agric. Food Chem. 24:277-282. Lamoureux, G. L., L. E. Stafford, and F. S. Tanaka. 1971. Metabol— ism of 2-chlorofiN7isopropylacetanilide (propachlor) in the leaves of corn, sorghum, sugarcane, and barley. J. Agric. Food Chem. g: 346—350. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 114 Lanzilotta, R. P., and D. Framer. 1970. Herbicide transformation. II. Studies with an acylamidase of Fusarium solani. Appl. Micro- biol. .19:307-313. Leavitt, J. R. C. and D. Penner. 1979. In vitro conjugation of glutathione and other thiols with acetanilide herbicides and EPTC sulfoxide and the action of the herbicide antidote R-25788. J. Agric. Food Chem. 27:5330536. Lotlikar, P. D., J. D. Scribner, J. A. Miller, E. C. Miller, 1966. Reaction of esters of acromatic Nehydroxy amines and amides with methionine‘in_vitro: model for in vivo binding of amine carcino- gens to protein. Life Sci. 5: 1263— 1269. Martens, R. 1978. Degradation of the herbicide [14C]-dichlofop- methyl in soil under different conditions. Pestic. Sci. 9:127- 134. Matsumura, F., and E. G. Esaac. 1979. Degradation of pesticides by algae and aquatic microorganisms. Pages 3717387.$E.M- A. Q. Khan, J. J. Lech, and J. J. Menn, eds. Pesticide and xenobiotic metabolism in aquatic organisms. American chemical Society Symr posium Series, vol. 99, Washington, D.C. McGahen, L. L., and J. M. Tiedje. 1978. Metabolism of two new acylanilide herbicides, Antor Herbicide (H—22234) and Dual Gmetola- chlor) by the soil fungus Chaetomium globosum. J. Agric. Food Chem. 26:414-419. McInerney, M. J. and M. P. Bryant. 1981. Anaerobic degradation of lactate by syntrophic associations of Methanosarcina barkeri and Desulfavibrio species and effect of H2 on acetate degradation. Appl. Environ. Microbiol. 41:346-354. Miskus, R. P., D. P. Blair, and J. E. Casida. 1965. Conversion of DDT to DDD by bovine rumen fluid, lake water, and reduced porphy- rins. J. Agric. Food Chem. 13:481-483. Molongoski, J. J., and M. J. Klug. 1980. a. Quantification and characterization of sedimenting particulate organic matter in a shallow hypereutrophic lake. Freshwater Biol. 10:497-506. Molongoski, J. J., and M. J. Klug. 1980. b. Anaerobic metabolism of particulate organic matter and sediments of a hypereutrophic lake. Freshwater Biol. 10:507-518. Murthy, N. B. K., and D. D. Kaufman. 1978. Degradation of penta- chloronitrobenzene (PCNB) in anaerobic soils. J. Agric. Food Chem. 26:115161156. 36. 37. 38. 39. 40. ' 41. 42. 43. 44. 45. 46. 47. 48. 115 On, T., K. Tatsumi, H. Yoshimura. 1977. Isolation and identifica- tion of urinary metabolites of AF=2 I3-(5-nitro-2-fury1-2-(2- furyl)acrylamide] in rabbits. Biochem. Biophys. Res. Comm. 75:401-405. Roberts, T. R. and M. E. Standen. 1978. Degradation of the herbi- cide flampropmethyl in soil under anaerobic conditions. Pest. Biochem. Physiol. 9:322-333. Saleh, M. A., and J. E. Casida. l978.~ Reductive dechlorination of the toxaphene component 2,2,5-endo, 6-gxg. 8,9,10-heptachloro- bornane in various chemical, photochemical, and metabolic systems. J. Agric. Food Chem. 26:583-590. Sander, E. G., 1978. Reactions of sulfur nucleophiles with halo- genated pyrimidines. Pages 273-297 in E. E. van Tamelen, ed., Bioorganic Chemistry. Academic Press, New York, N.Y. Sethunathan, N., and R. Siddaramappa. 1978. Microbial degrada- tion of pesticides in rice soils. pp. 479-497, Soils and rice. International Rice Research Institute, Los Banos, Philippines. Slyter, L. L. 1979. Monensin and dichloroacetamide influences on methane and volatile fatty acid production by rumen bacteria .13 vitro. Appl. Environ. Microbiol. 37:283-288. Smith, A. E. and D. V. Phillips. 1975. Degradation of alachlor by Rhizoctonia solani. Agronomy J. 'QZ:347-349. Tanks, H., N. Esaki, and K. Soda. 1977. Properties of L-methionine- lyase from Pseudomonas ovalis. Biochem. 16:100-106. Tateishi, M., S. Suzuki, and H. Shimizu. 1978. The metabolism of bromazepam in the rat - identification of mercapturic acid and its conversion in vitro to methylthio-bromazepam. Biochem. Pharm. 27:809-810. Taylor, R. W. 1979. Degradation of alachlor [2-chloro-2',6'- diethyl-Nf(methoxymethyl)acetamide] in soils and by microorganisms. 8M.S. Thesis, Michigan State university. 25 pp. Tiedje, J. M. and M. L..Hagedorn, 1975. Degradation of alachlor by a soil fungus, Chaetomium globosum. J. Agric. Food Chem. 23:77-81. Vogels, G. D., W. F. Hoppe, and C. K. Stumm. 1980. Association of methanogenic bacteria with rumen ciliates. Appl. Environ. Microbiol. 40:608-612. Williams, P. P. 1977. Metabolism of synthetic organic pesticides by anaerobic microorganisms. Residue Reviews. ‘§§:63-135. 49. 50. 51. 116 Williams. P. P., and V. J. Feil. 1971. Identification of triflura- lin metabolites from rumen microbial cultures. Effect of triflura- lin on bacteria and protozoa. J. Agric. Food Chem. 19:1198-1204. Zehnder, A. J. B. 1978. Eealogy of methane formation. Pages 349-376 ig_R. Mitchell, ed., water pollution microbiology. vol. 2. John Wiley and Sons, Inc. Zeikus, J. G.. 1977.) The biology of methanogenic bacteria. Bacterial. Reviews. 41:514-541. CHAPTER III MICROBIAL METABOLISM OF ACETANILIDE HERBICIDES UNDER AEROBIC CONDITIONS -- A REVIEW The acetanilide class has been one of the most successful of the current generation of herbicidal chemicals. Their control of annual grasses plus several problem.broad-leaf weeds has made this class well suited for use in American agriculture. In this report, we review information.on microbial metabolism of several acetanilides and then summarize the general types of reac- tions and products that seem to be common for this group of chemicals. The chemicals included are alachlor, diethatyl, metolachlor, butachlor, and propachlor for which metabolite information exists plus less de- tailed information on persistence of propanil, dicryl, and solan.1 The work summarized in this report is limited to studies with microbial cultures. All reported examples of metabolism are with fungal cultures; to our knowledge no evidence of bacterial metabolimm of the acylanilides have been observed. Pure culture studies such as these lend themselves to easier and a more complete analysis of 1Diethatyl [2-chloro-N-(2', 6' -diethy1pheny1)-N-methyl(ethylcarboxy- 1ate)acetamdde], Boots/Hercules, Agricultural Division; metolachlor [2-chloro-N-(2'-ethyl-6'-methylpheny1)-N-(2-methoxy-1-methylethyl) acetamide]. Ciba-Giegy Corp., Agricultural Division; alachlor [2- chlaro-2' ,6'-diethyl-N-(methoxymethyl)acetanilide], [2-ch1ora-2',6'- diethyl-N-(butaxy-methyl)acetanilide, and propachlor (2-chloro-N- isoprapylacetanilide), Monsanto Agricultural Products Co.; prapanil (3,4-dichlorophenyl-propiananilide), Rahm and Haas Ca.; dicryl (3',4'- dichloro-Z-methyla-crylanilide), product discontinued; salan (3'- chloro-2-methyl-p-valerotoluidide), product discontinued. 117 118 products and identification of chemical structures. Furthermore, it is easier to elucidate the pathway of degradation under these more controlled and restricted conditions. Pure culture studies, however, should not be used as evidence-for these products or reactions in nature. Information from culture studies is more properly used to aid analogous research on soils and sediments by providing an idea of which.metabolites might be expected and how to analyze for them. We feel that by conducting both pure culture and soil studies in con- cert, one has a better chance at resolving the fate and behavior of a pesticide in nature. Studies with pure cultures are usually conducted either with the microbial culture growing on complex medium to which the pesticide is added or with previously grown microbial cells incubated with pesti- cide but under conditions where growth does not occur, usually because the growth substrate is absent. The organisms in the latter case are termed "resting cells" because they do not proliferate though all of their metabolic capacities are still present and functioning. The resting cell approach is often preferred because the cell density can be high and thus the reaction proceeds at a faster rate and be- cause no medium components are present to interfere with analyses. GENERAL'METHODS Resting cells were used for all of our work summarized here (McGahen and Tiedje, 1978, and Tiedje and Hagedorn, 1975). Our general procedure is to grow a pure culture of fungal or bacterial cells in large quan- tities an a suitable growth medium (in the presence of pesticide if that yields a more active resting cell preparation). Cells are then 119 ‘harvested by filtration or centrifugation, washed two or three times in buffer to remove residual medium contaminants, and resuspended in a buffer containing the compound to be degraded. This procedure is conducted aseptically if the resting cell incubation period is long enough to be influenced by introduced contaminants. We used the soil fungus, Chaetomium globOsum, in our studies of acetanilide metabolism since we found this strain to be the most active of several examined (11). Subsamples of the resting cell-herbicide suspension taken at various times were extracted and analyzed for herbicide disap- 140-1868188 herbi- pearance by gas-liquid chromatography (GLC) and, if cide was used, by liquid scintillation spectrometry of the organic extract and the remaining aqueous phase (4). At the end of the experi- ment, the entire solution was filtered to remove fungal cells and extracted. The extract was concentrated and analyzed for metabolites which were identified by GLC with mass-spectrometry (MS). ACETANILIDE‘METABOLITES The pattern of degradation of the diethatyl is presented in Figure 1. By GLC analysis, most of the parent compound disappeared within 24 hours, and was no longer detectable after 48 hours. Hawever, using either carbonyl or ring 14C-labeled diethatyl, we noticed no loss in total radioactivity, while a reduction in organic extractable radioactivity occurred with a concurrent increase in the per cent 14C-diethatyl remaining in the aqueous fraction. The same fungus, however, had more difficulty metabolizing metolachlor (Figure 2). Degradation was analyzed by GLC only. Even after incubation periods of up to six days, 50 percent of the metolachlor still remained. Figure l. 120 CARBONYL LABEL M _ E \ c 400E009 PHASE -5. O ,A—-""" 8'- fl—‘— u_ ’.A\ 0 [I A; a! ,’ *0 A, "(i-ORGANIC PHASE ANTOR (glc) ‘ ‘ k 0 1 ‘ ‘ ‘ 3 ‘ ‘ O O 40 80 l20 IOO '- (1)9» RING LABEL 80 . \ .3 "C-AQUEOUS PHASE d E 60 _ A\~am now aofiaooonoa Howoom nom AH .0 .ma .00 onoSuoo Hono>oo an oounooon ouoaoono wo.hufinoananm 030 «a :omnnomaoo < eo5xu—(2—0.° . 10.2(000 .m anawwm 124 ring, and subsequent closure. This indoline was first found for alachlor (11). It has since been syntheSized by Monsanto chemists who confirmed the identity (J3 Malik, personal communication). The indoline derived from alachloralsa ca-elutes from GLC and high-perfor- mance liquid chromatography (HPLC) with the diethatyl indoline product. The MS also match. We are now confident of the identity of the indoline products. Chen and Wu, who researched butachlor degradation by the soil fungus, Mucor sufui, also claim to have found this same indoline metabolite (1). Since formation apparently depends on the ethyl groups, then propachlor should not form this indoline, and none has been re- ported. The explanation for lack of isolation of this product from metolachlor is less clear; it may lie in the fact that g. gldboSum apparently could not readily remove the alkyl group from the nitrogen. The product in column III, also an indoline, is formed when the chloroacetyl group, rather than the alkyl substituent is removed from the nitrogen. This was not reported for butachlor or propachlor and is now considered unlikely for alachlor because of the expected insta- bility of the remaining alachlor moiety when the chloroacetyl group is removed and only the methoxymethyl group remains attached to the nitrogen. Tiedje and Hagedorn (13) originally reported a methoxy- methylalachlor metabolite (see Figure 4, column III) but its existence is now questioned because of its chemical instability. Mansanto scientists attempting to synthesize the compound noted that only its decomposition products - the aniline derivative and methanol -- could be detected (J. Malik, personal communication). Furthermore, we were not able to repeatedly detect this product in fungal cultures. 125 The compound in column IV has two alterations labeled "A" and "B" on different positions of the diethatyl structure. The "A" version contains a hydroxyl in place of the chlorine on the chloroacetyl group and the "B" alteration shows a dehydrogenation of the ethyl substituent on the ring. A compound having only the "B" dehydrogenation alteration as well as a compound having the exact structure shown here with both the "A" and "B" alterations on the same molecule were found as transfor- mation products of diethatyl. A product having both the "A" and "B" transformations together on the same molecule was also found from metolachlor, but not from alachlor or butachlor. However, propachlor was dechlorinated with hydroxylation (an "A" alteration), farming the only product identified from propachlor (5). The authors state that other products were formed by Fusarium axysporum from propachlor which were not identified (5). The replacement of the chloro group by a hydroxy found for propachlor is a reaction that we expect to be more common; the analytical procedures used probably did not allow its detection in other cases. In concluding a discussion of Figure 3, it is worth noting that metolachlor was markedly more resistant to 'N-dealkylation than were the other acetanilides. The only N-dealkyla- tion metolachlor product (column I, Figure 3) was a relatively minor product, while the demethoxylated metolachlor product (column III, Figure 4) appears to be a major metabolite (7). Figure 4 contains some miscellaneous biodegradation products, possibly unique to each herbicide. Butachlor has a product, shown in column I, which is dealkylated, and dechlorinated with hydroxylation. The aniline product in column II was found from both alachlor and 126 oononono: oouoowoon one mo Bonaooouoi Howoow aonw mnoooono nonuo mo oooanooaoo < .Aoawonoo Hoooaunvoo now m onowwm ooov I (.2? + 55.3205: .3. o £40.25: 0.2:... 52.05.3052 5:2. « .eaieet 030! u.“ 0 n a m 0 2 WZUZOUZ .029 Z 19 :0: 05: 5: 2...... «.0300; 3.0.: firemoJ... Exam 0.... o: >. :— .- 0~U39000 ZO=(O<0009 00 axv<0000 .6. 058.4615»... \ .05zowzoh. 05: noozuepoa 4.) (.8? zomzx .05zowzu52005zu5zu5205: 00:34: I .m. 83.80:..‘0 .o5zowzo5xoo5x 35:012.: I do 83.50:...9 35.3%. wzofsnz 5: 5:422... .l (O/\ .6. 05543.3 .05: 0wzonzwounzu5: 623050409 10.2(000 .0 onowam 127 butachlor. No attempt was made to look for aniline products in the diethatyl and metolachlor studies. In column III, compounds which may be the major metabolites for both diethatyl and metolachlor degra- dation by E. globosum are shown. .93 globosum formed an acid by hydroly- sis of the ester bond on the N-alkyl group of diethatyl, removing the ethyl group. Similarly, the fungus split the ether bond of the .N-alkyl group on metolachlor, leaving a hydroxyl. A final product identified from‘butachlor may be the dubutoxylated product shown in column III. The reaction is similar to the demethylation or deethylation of metolachlor and diethatyl, respectively, except that the butachlor product was not hydroxylated. The last two products are shown in column IV. The diethatyl imine product was found in fungal resting cell flasks. Scientists of Hercules, Inc., noted that this compound was also formed by photo- degradation (D. Black, personal communication). In our case, it was presumed to be a biological product since the flasks were incubated in the dark and it was not detected in control uninoculated flasks. The last metolachlor product in this figure, column IV, has the chloro- acetyl group removed from the nitrogen. It may be a precursor to the previously identified indoline (Figure 3). g, globosum also produced some products from metolachlor which were particularly difficult to identify (Figure 5). At this point, they are still unconfirmed in that the structures are based solely an interpretation of mass spectra. The structures III, V, and VI are the original tetrahydroquinoline identifications from.our previous report (7); the numbers assigned to the three products which eluted at different times on GLC are conserved here. The two "morpholine" 128 .omoszuoo ofiaooouoa odomoomoon ooo onuoooo owns 050 £053 nooomnmaoo mononoonnw o>nuoononao poo onoooono connooonwoo noaaooaoooa oomoaona Hono>om at .03 5P4: 205.. :50 «5% 5.5.0. .5 8...... 2.. $5 .\... l K 565 0).. Fauna; OQK .505: 555 a}. .n 565 {I .< .0 .e. 32.59200: 05.24235: :>.> :5 502.5oz.:ooaonz<:: .3 555 o\... :5 35 as... I « 0 Z 20 800 8 on: :0 a 5: :3 555 0).. 505 ox... 3:; n no. 2 5:0 :00 05:05.5 a 5.. . .m onownm 129 structures A and B, are alternatiyes for tetrahydroquinOIine struc- tures III and IV. These alternatiye structures were suggested by Ciba-Giegy Corp. scientists (H. LeBaron, personal communication). The compounds in the upper and lower left with mass 265 are exactly the same structure but drawn differently to suggest how tetrahy- droquinoline product III and morpholine product A, both haying mass 263, may be formed from.this compound which was a proposed inter- mediate in our metolachlor degradation scheme (7). Structue "II" at the lower right of Figure 5, with mass 269, is a metolachlor metabolite and, if it were dechlorinated and sub- sequently hydroxylated, it could be an intermediate for morpholine structure B (mass 233). One problem with the formation of lower morpholine structure A is that the closure must be formed by a reaction between the methyl on the alkyl group and a hydroxyl on the dechlorinated[hydroxylated chloroacetyl group. If this reaction is biological, it seems out of character for the fungus to attach this methyl. No other instance of attack on the alkyl group of metolachlor has been demonstrated, other than a demethylation, leaving a hydroxyl. As 9. globosum can dehydrogenate the ethyl group on the ring, it may be possible for it to catalyze a reaction with the methyl and the remaining part of the chloroacetyl group. It is logical for metolachlor product II (Figure 5) to become dechlorinated, hydroxylated. and subsequently form the morpholine structure B with mass 233. Also, Ciba-Giegy Corp. scientists have- apparently isolated this product as a photodecomposition product of 130 metolachlor (H. LeBaron, personal communication). However, the morpho- line structure A with.mass 263'Was not found as a photodecomposition product. One further complication is that of the three metolachlor products found which have questionable identities (pictured as products III, V, and VI) two had a mass of 233. A second structure similar to morpholine B with mass 233 could not be derived although it is possible to derive a second tetrahydroquinoline form with mass 233 (Figure 5, structure VI). These products (tetrahydroquinoline or morpholine structures) could also be formed by condensation reactions of hydroxylated precursors during GLC analysis. Providing that these metabolites were formed by Q, globosum. the eventual verification of these structures will result in an interesting addition to the types of reactions catalyzed by microorganisms. From examination of Figure 3, 4 and 5, we can summarize the types of reactions of acetanilides which microorganisms may catalyze. These are listed below: 1. Dealkylation from the nitrogen or removal of part of the alkyl group on the nitrogen, particularly breaking ether and/or ester bonds. 2. Dehydrogenation of the ethyl substituent on the ring. 3. Dehalogenation of the chlorine from.the chloroacetyl moiety. 4. Hydroxylation which frequently occurs in conjunction with dehalogenation and dealkylation. 5. Indoline ring formation. 6. Tetrahydroquinoline or morpholine formation, which are uncomfirmed metolachlor products. 131 A primary reaction is complete or partial removal of the alkyl group on the nitrogen which frequently occurs as a result of breakage or hydrolysis of ether or ester bonds. This is termed demethoxylation, deethoxylation, etc., depending on the type of group removed. A second reaction is dehalogenation, or the removal of the chlorine from the chloroacetyl moiety. Dehalogenation has also occurred on halogens which are ring substituents on other acylanilides (4). Frequently, one finds that a hydroxylation reaction occurs on a site which has been dealkylated or dehalogenated. Two other reactions that occurred on acetanilides which have ethyl or diethyl substitutents on the ring are dehydrogenation of an ethyl group and, probably subsequently, indoline ring formation. Further, more complex products such as tetrahydroquinoline_or morpholine are possible. The prevalence of ring closure reactions suggests that these types of products bear closer scrutiny in studies on pesticide fate. One would expect their precursors also to be very reactive with soil organic matter, thereby providing a mechanism for covalent incorporation of pesticide products into humus. So far, these reaction types are derived from.experiments using single, isolated organisms, under aerobic conditions in buffer solutions with a relatively high concentration of both biomass and herbicide. Also, the metabolites identified represent only the organic-extractable compounds. Polar products are definitely prevalent as indicated by Figure l where substantial 14C-label from diethatyl entered the polar phase. Formation of polar products has also been shown for alachlor (11). Clearly further work needs to be done to elucidate the polar metabolites. 132 Kaufman and Blake tested the ability of several soil microorganisms to degrade acyl- or acetanilide herbicides (4). Table 1 presents a selection of some of the compounds and organisms tested which showed activity. Degradation was assessed by assaying for the release of aniline or chloride from the parent compounds. The "+" indicates that aniline or chloride was found. The zero indicates none found, but does not necessarily indicate lack of degradation. The indication, then, is that all the organisms can degrade propanil and dicryl to the extent of releasing aniline, although only the fungi Aspergillus ustus, Fusarium oxysporum, and Penicillium chrysogenum also release chloride from propanil. The remaining fungus, Trichaderma veride, and the two bacterial species, Pseudomonas striata and Achromobacter do not release the chloride from propanil. All organisms released chloride from dicryl. None of the organisms released the aniline moiety from propachlor, although A. ustus, F. gxysporum and the bac— terium Achromobacter sp. released chloride. Similarly, F. oxysporum, .2. chrysogenum and T, viride released aniline and chloride from solan, while the two bacteria only released the chloride and A. 2§£g§_did neither. Table 1 illustrates what has become an observation worth noting, namely that it is more common for fungi to be active in pesticide metabolism than bacteria, especially when compounds of moderate to lengthy persistence are considered. For the acetanilides, this is particularly true in our experience and is evident from the work of Kaufman and Blake (4) and Chen and we (1). The former authors found eight active fungal species and two bacterial species in the study 133 .a05umnsuca shop 5 umumm Emfisowuo can hp downspouwoo mo xoma mumoaosfi hawummmooo: uo: mooo use mouomuoo mood omuooaos5 :o: "monomoua mmumowoca :+:n .mcmlhm~5m Annmav .Em3005m .Hoflm afiom .mmem was smamamx 805mm + + + o o . o + + .mm nouomnoaounu< + o + o o o + + .em mommaovaomm + o + o + o + + .mm mauooonuwua + o + + + o + + .3 5553535.. + + + + + o + + .am 5555535 c + + + o o a... + .9. 535595253 smaom Hoanomaoum ahuofia Hammeoum smaom uoaaomdoum ahuoan stmmoum Bogaowuo oouomuoo oouuoanu oouoouoo ma5H5s< .mamfiamwuoouoda up mossoqaou wowawsmuooo uo Iahom Houo>mm mo sowumoouwoo mom moswo5>o mo huoaasm .H manna 134 summarized in Table 1 while the latter authors found ten fungi active on butachlor but only two bacterial species. As stated earlier, the above studies were done in pure culture and do not necessarily reflect degradation in nature. There are several factors which can account for the differences in culture and nature results. First, mixed cultures or organisms such as found in nature, can sometimes degrade compounds that cannot be degraded by each organism in isolation. Though there are no examples of this phenomenon for the acetanilides, but there are examples for other chemicals. One example is for silvex which could be degraded to 14CO2 and chloride in the presence of a mixed culture of aquatic organisms but could not be degraded by any one of the bacteria acting alone (8). The second case is the anaerobic degradation of the aromatic ring by a consortia of bacteria which include methanogens, fatty acid degraders and organisms which reduce the aromatic ring (2, 6, 9). In this case, the association is obligate since the methanogens serve as the required sink for H maintaining this gas concentration very 2 low so that the earlier fermentation steps are thermodynamically feas- ible. A second environmental influence on degradation in nature can -be due to the presence of additional carbon and energy sources. In some cases, additional carbon stimulates degradation generally due to an increase in concentration of the degrading enzymes caused either by growth of~a cometabolizing population or by causing induction of specific enzymes (3). In other cases, reduction in degradation due to addition of carbon is noted. This can be caused by selection of 135 a population less favorable to degradation or to catabolite repression, especially by glucose, which represses more diverse pathways of carbon flow. For the acetanilides we have examined, we have noted no changes in their metabolism due to the presence of additional carbon. A third factor is the availability of the pesticide to the competent microorganism. Since these xenobiotics can be sorbed, leached, volatilized, hydrolyzed or be spacially protected from the microbes, the substrate may not be available for attack by the organisms. In our experience with the acetanilides, the chemicals seem available to eventual biological attack though this could be the rate controlling step. Finally, one is generally not certain of the predominance or the activity of the organisms studied in culture in the habitats of interest. Most organisms in nature are starving and they may not have enough energy to initiate degradation, (e.g., enzyme synthesis, cell uptake) even though they may possess the genetic information which could result in degradation. Also, it is very difficult to determine which organism(s) is responsible for degradation when the compound does not serve as a growth substrate. In one attempt to show a possible role of Q. globosum, we added a heavy suspension of germinating spores to soil to see if we could stimulate alachlor degra- dation. However, no change in rate was observed from the uninoculated control (S. F. Chou and J. M. Tiedje, unpublished data). Thus, we are uncertain of the role of this fungus in alachlor degradation in soil. 136 In an earlier workshop on microbial biotransformations, Tiedje (10) identified seven areas of research needs for future studies on biodegradation, and one area that seemed to be overemphasized. These topics are thought to be important to better predicting and evaluating biodegradation and thus are worth brief mention here. They are (1) better methods to determine "the permanence of binding" of xenobiotics to sediments and organic matter so that the significance of this fate can be confidently evaluated, (2) a better understanding of behavior of xenobiotic degradation at low substrate concentrations, (3) a better characterization of the mechanisms and kinetics of cometabolism, (4) a better understanding of the influence of "starvation physiology" on the metabolism of xenobiotic substrates (amplified above), (5) a better understanding of the anaerobic metabolism of xenobiotics and its significance and the types of prevalent reactions, (6) more use of assay systems which have the salient features of the receiving habitat, and (7) a more general focus on accurately predicting the fate of xenobiotics in nature. The topic which seems to be receiving adequate, if not too much emphasis, is the effect of xenobiotics on pure cultures of microorganisms since this has proven not to be a productive and effective way to evaluate general toxicological concerns. 10. 11. LITERATURE CITED Chen, Y.-L., and T.-C. wu.‘ 1978. Degradation of the herbicide butachlor by soil microbes. Paper presented of the IVth Inter- national Congress on Pesticide Chemistry (IUPAC). Zurich, Switzerland, 1978. Available through Y.-L. Chen and T.-C. Wu, Dept. of Agricultural Chemistry, National Taiwan university, Taipei, Taiwan, Republic of China. Fina, L. R., R. L. Bridges, T. H. Coblentz, and F. F. Roberts. 1978. The anaerobic decomposition of benzoic acid during methane fermentation. III. ‘The fate of carbon four and the identifica- tion of propanoic acid. Arch. Microbiol. 118:169-172. Horvath, R. S. 1972. Microbial co-metabolism.and the degradation of organic compounds in nature. Bact. Rev. 36:146-155. Kaufman, D. D. and J. Blake. 1973. 'Microbial degradation of several acetamide, acylanilide, carbamate, toluidine and urea pesticides. Soil Biol. Biochem. 5:297-308. Kaufman, D. D., J. R. Plimmer and J. Iwan. 1971. Microbial degradation of propachlor. American Chemical Society Abstracts, l62nd Meeting, washington, D.C. PEST-21. Keith, C. L., R. L. Bridges, L. R. Fina, K. L. Iverson, and J. A. Cloran. 1978. The anaerobic decomposition of benzoic acid during methane fermentation. IV. Dearomatization of the ring and volatile fatty acids formed on ring rupture. Arch. Micro- biol. 118:173-176. McGahen, L. L., and J. M. Tiedje. 1978. Metabolism of two new acylanilide herbicides, Antor Herbicide (H-22234) and Dual (metolachlor) by the soil fungus Chaetomium globosum. J. Agric. Food Chem. 26:414-419. Ou, L. T. and H. C. Sikka. 1977. Extensive degradation of silvex by synergistic action of aquatic microorganisms. J. Agric. Food Chem. 25:1336-1339. Shalomi, E. R., A. Lankhorst, and L. R. Prins. 1978. Methanogenic fermentation of benzoate in an enrichment culture. ‘Microb. Ecol. 4:249-261. Tiedje, J. M. 1980. An attempt at identifying research needs for studies on microbial transformations. IE_D. Schlessinger, ed. Microbiology 1980. American Society for Microbiology, washington, D.C. ' Tiedje, J. M., and M. L. Hagedorn...1975. Degradation of alachlor by a soil fungus, ChaetOmium globosum. J. Agric. Food Chem. 137 APPENDIX APPENDIX A EXPERIMENTS OR RESULTS WHICH SUPPORT THE RESEARCH IN CHAPTERS 1 AND 2. The first section of this Appendix presents experiments which (1) test the effect on the pH of flushing the sediment with a C02:N2 mixture, and (2) further investigate and compare the Sep-Pak C18R separation procedure (see Chapter 1) with that of extraction with hexane. The second section presents (l) chromatograms from.the gas— liquid chromatographic (GLC) analysis and (2) proposed fragmentation schemes for diethatyl and metolachlor and their metabolites which were produced in the sediment under anaerobic conditions. The effect of the flushing procedure on sedimentng In anaerobic experiments, sediment was weighed into 26-mL serum bottles under aerobic conditions. After the serum bottles were stop- pered, the headspace was flushed to remove any 02 with a mixture of COZ:N2, having approximate relative flow rates of 20:80 mL/min, C02:N2. u 0 However, the pH of the samples might have been affected by the 002 added by flushing. To test this, 26-mL serum bottles with 10 g sedi- ment or autoclaved sediment, or 10 mL samples of 0.02 M phosphate buffer (pH 7.0) or deionized water were stoppered and flushed for 15 min using various flow rates of COzzNZ. The "0" samples were not flushed at all. Data are presented in Figure l. The flushing procedure caused very little change in pH in normal or autoclaved sediment samples, although the pH of the buffer decreased from pH 7.0 in samples which.were not flushed to pH 5.9 in samples which were'flushed with C02:N2 at approximate flow rates of 40:60 mL/min. The pH of 138 139 o 10 20 30 1.0 so APPROXIMATE FLOW RATE OF co2 (ML/MIN) _ Decrease in pH in normal and autoclaved sediment, 0.02 M phosphate buffer, and distilled, deionized water by increas- ing flow rates of 002. legend: (0—0) 0.02 M phosphate buffer; (H) normal sediment; (L-A) autoclaved sediment; (.—.) distilled, deionized water. Figure l. 140 deionized water decreased from pH 5.6 in unflushed samples to pH 3.7 in samples flushed with CO at flow rates of 20:80 mL/min. Buffer 2‘N2 solutions could be used as controls for chemical decomposition of the herbicides although due to difficulties incurred in later experi- ments where the 0.02 M phosphate buffer, pH 7.0, was diluted to 0.01 M by the addition of an equal volume of 100 ppm herbicide solution, a stronger buffer concentration such as 0.05 M is recommended to permit the solution to be diluted by 502, flushed, and still maintain pH 7.0. R Comparison of the extraction efficiency of the Sep-Pak 018 procedure with hexane extraction; the loss of herbicides on polycarbonate filters The extraction efficiency of the Sep-Pak'C18R procedure was come pared to that of hexane using 10 g of soil or sediment samples or 10 mL 0.02 M phosphate buffer (pH 7.0) amended with 10 mL of diethatyl solution. Final concentratin of the diethatyl was 50 ppm and the 14C-label concentration was 4000 dpm/mL. Samples were initially extracted by the addition of 5 mL acetonitrile (20% acetonitrile); Sme water was added to controls. After shaking and centrifuging the samples, 0.5 mL was analyzed for total 14C-label. A 5 mL portion of each 20% acetonitrile extract was removed and extracted twice with 2.5 mL portions of hexane and the hexane phases were combined. A 0.5 mL portion each of the hexane and water phases was analyzed for 14C-label. A third, 3 mL portion of the extract, was removed and R separated using the Sep-Pak C18 procedure (Chapter 1). Samples were eluted with 100% methanol. A 0.5 mL portion each of the methanol and water phases was analyzed for 14C-label. To determine if 141 filtration of samples was a useful method of sample clean-up prior R to Sep-Pak 018 separation and HPLC analysis, 10 mL samples was filtered through 0.4 uM polycarbonate filters with glass fiber pre- filters (Nucleopore Corp., Pleasanton, CA). The filters were not rinsed with any solvent or water after filtration to remove adsorbed diethatyl, and they were not reused. A 0.5 mL sample was analyzed for total 14Cj—label which passed through the filter, and a 5 mL and R a 3 mL portion were extracted with hexane or by the Sep-Pak C18 method, respectively. Data are presented in Table 1. Total recoveries of diethatyl are consistent with those presented in Table 4, Chapter 1. Inconsis- R tent results between samples extracted by the Sep-Pak'C18 method and those extracted with hexane are noted with an saterisk (Table 1). R Here, Sep-Pak 018 separation resulted in some high values in the water phases of unfiltered samples from sediment, muck, and sandy loam. However, in most cases, the amount of ll‘C-label in the methanol R phases from the Sap-Pak C18 separation was comparable to that in the hexane phase from hexane extracts. As consistent results have 'R been obtained using the Sep-Pak C18 method in all subsequent experi- ments, it is considered reliable and comparable to extraction with R a nonpolar solvent such as hexane. The advantage of the Sap-Pak C18 method is that emulsions are not created by extraction of the herbi- cide from sediment with 20% acetonitrile extraction followed by Sep- R Pak 018 separation. Secondly, the samples are cleaned by the Sep- R Pak 018 cartridge and are in the solvent which is compatible with the solvent system of the HPLC system. 142 R Efficiency of Sep-Pak C18 separation versus hexane extraction of a 20% acetonitrile extract of diethatyl from sediment and soils. The effect of sample filtration prior to separation Table l. or extraction is also reported. Percent Extracted Treatment of Acetonitrile Water extract & sample Phase SepePak Hexane Sep-Pak Hexane Sediment Total 91 60 Organic 81 91 54 62 Water 17 5 3 5 Filtered Total 56 sediment Organic ND 74 ND ND Water ND 7 ND ND Muck soil Total 81 42 Organic 80 86 38 41 water 4 6 3 4 Filtered Total 46 26 Muck soil Organic 40 36 38 26 water 12* 5 3 4 Sandy loam. Total 97 ~ 95 Organic 94 98 82 98 water 22* 7 3 5 Filtered sandy Total 68 loam soil Organic ND 73 ND ND water ND 6 ND ND Buffer & glass Total 102 98 beads Organic 101 116 98 94 Water 3 7 3 10 Filtered Buffer Total 75 & glass beads Organic 78 72 ND ND water 5 9 ND’ ND ND - Not determined * - These values are too high and are considered anomalous in comparison with values obtained in all other experiments (Chapters I and II). 143 Recovery of ll'C—label was 30 to 402 less if 20% acetonitrile extracts were filtered in addition to the usual extraction procedure, R and then extracted with hexane or separated by the Sep-Pak 018 method (Table 6). The concentration of 14C-label in buffer samples which were filtered was also 25% less than in those which were not filtered. If the filter is compatible with an organic solvent it could be rinsed to remove the herbicide. However, that would dilute the sample, and the increase in organic solvent may cause the herbicides to not be R adsorbed on the Sep-Pak C18 cartridge or be efficiently extracted by hexane. A note concernipg sample preparation for HPLC analysis A precolumn should always be used on the HPLC system to ensure longer column life. Caution is recbmmended with samples containing buffer or solutions which are "salty." Concentrations stronger than 0.02 M to 0.03 M precipitate in the water and/or methanol phase after Sep-Pak C18R separation, or after injection into the HPLC system if an organic solvent is used in the mobile phase. The precipitate clogs the pump valves and column frits. The Sep-Pak C18R does not clean up media used for bacterial or fungal growth culture experiments with- out centrifugation. That is, centrifugation of debris and cells from the samples is mandatory prior to Sep-Pak C18R separation. Filtration through a filter compatible with the compounds which are being analyzed or from which the compounds can be extracted may be as appropriate as centrifugation. This was determined from experiments on the potential for degradation of 2,6-diethylaniline by bacterial and fungal cultures while I was in the lab or Dr. J.-M. Bollag, February to June 1981. 144 Gas-liquid chromatographic analysis of diethatyl and metolachlor metabo- lites produced in sediment under anaerobic conditions Figures 2a and b show the gas-liquid chromatographic (GLC) analysis of extracts from sediment amended with 50 ppm diethatyl or metolachlor. Samples were incubated anaerobically for eight weeks, and extracted to obtain the 20% acetonitrile extract, as described in Chapter 2. A portion of the combined 20% acetonitrile extracts from five samples (between 100 and 150 mL) was partitioned three times with hexane, each portion being one-third the volume of the sample. The combined hexane fractions were dried with anhydrous sodium sulfate, and concen- trated to a small volume (3 to 10 mL). GLC analysis was done as described in Chapter 2, except the column temperature was programmed between 1300 C - 2250 C (6O/min) to separate metabolites. A Spectro- Physics 4100 computing integrator monitored retention times and peak areas during GLC analysis. Samples were analyzed within 1.5 months after extraction. GLC analysis of the diethatyl extract of sediment (Figure 2a) shows a third metabolite peak compared to that in Figure 6, Chapter 2. Aparently, the transformation product of diethatyl which.was respon- sible for peak 3 was labile and no CC/MS data was obtained to identify it. ‘ No additional peaks were seen in the chromatogram of the metola- chlor extract (Figure 2b) compared to that shown in Figure 6, Chapter 2. 145 r1 LL! 0') Z O O. U) LU O: E) 3 a DIETHA‘I’YL {B o 2 1 . I T 1 ' l r I . r f I a 12 '15 20 TIME (MIN) Figure Za. Gas-liquid chromatographic analysis of diethatyl and metabolites produced under anaerobic conditions in Sediment (Chapter 2). 146 ‘1 DJ METOLACHLOR CD a ti ., '2 U5 “J m 1‘ m C) .— C) UU .— UJ C3 1 I U I ' ' ‘ l ' I v I ' l . ‘r' I. .8 12 15 20 21. 28 TIME (MIN) Figure 2b. Gas-liquid chromatographic analysis of metolachlor and metabolites produced under anaerobic conditions in sediment. 147 Mass-spectral data Possible transformation schemes for diethatyl metabolites D1 and D2 and metolachlor metabolites M1 and M2 are presented in Figure 3, a through d. Metabolites were produced from the incubation of diethatyl or metolachlor in sediment under anaerobic conditions (Chapter 2). All pertinent facts are presented in Chapter 2. 148 .Ho ouaaonmuoa ahumnuo5v Mom season sofiumucmawmum oaowmmom .mm ouswam 2.... so 5...... .J z 5o 3.. so . 3855. 3. 3.. as... 5: 3.. 855.5. +55 5... 2.0: .0 0" 'Ux.~ 30,-30430 ’0 I Inca: WOUuIUn—L 3 .7 ._, .52 3.. 5... study «as (.3. ozuc «5 + 7 2.2.5... .57 63.5... 555 Q... .. . LO< 5.55.4.2: .“Hw a). {omitsm 5...... . .62.... to \es 2......» . 30. 530+: .tu «)3 o *0 9 5:83. N2 (.3 06 .1 . O w 5.3 2.9... .o . .. f .\.- w .5 .5 QUVU/eoo...: .2: {I :00 ~‘I AV °2213U ntUUZ'Ud: / dOZ..T-O.U 355.5. «5.8.. m 8:: 3.5.2.5555} ’\ AU :50 v 0 a . x... 2.55: uzlof W « 353... 149 .Na muaaonmuoa .Hmumnumao now 9550.... caumusmewmum 35:530.” .58 «.5:me a: 5.: :0 5:05.80 .ol 5203:: aged: '0. (I ‘1‘ ) aflnd. and (to 1p 2...... so . “ onunzlu... 3535. .3 I 23525 2105.. / / 5 (AA: 3 55: 52.55.22... 50 .150 38.53555 0... 5 w. s. 5.5:...