Biodegradation tests of mercaptocarboxylic acids, their esters, related divalent sulfur compounds and mercaptans

Biodegradation tests of mercaptocarboxylic acids, their esters, related divalent sulfur compounds... Mercaptocarboxylic acids and their esters, a class of difunctional compounds bearing both a mercapto and a carboxylic acid or ester functional group, are industrial chemicals of potential environmental concern. Biodegradation of such compounds was systematically investigated here, both by literature search and by experiments (Closed Bottle Test OECD 301D and Manometric Respirometry Test OECD 301F). These compounds were found either readily biodegradable or at least biodegradable to a significant extent. Some related compounds of divalent sulfur were tested for comparison (mercaptans, sulfides, disulfides). For the two relevant monofunctional compound classes, carboxylic acids/esters and mercaptans, literature data were compiled, and by comparison with structurally similar compounds without these functional groups, the influence of COOH/COOR’ and SH groups on biodegradability was evaluated. Thereby, an existing rule of thumb for biodegradation of carboxylic acids/esters was supported by experimental data, and a rule of thumb could be formulated for mercaptans. Concurrent to biodegradation, abiotic processes were observed in the experiments, rapid oxidative formation of disulfides (dimerisation of monomercaptans and cyclisation of dimercaptans) and hydrolysis of esters. Some problems that compromise the reproducibility of biodegradation test results were discussed. . . . . Keywords Ready biodegradability tests Closed Bottle Test Manometric Respirometry Test Difunctional compounds . . . Mercaptocarboxylic acids Mercaptocarboxylic acid esters Mercaptans Biodegradation rules of thumb Introduction minimise the release of persistent chemicals into the envi- ronment. While a persistent chemical is a potential threat to Biodegradability in an aerobic aquatic phase is a key ele- the environment, a chemical that is readily biodegraded to ment in the environmental assessment of chemicals and is products such as CO and H O (mineralisation) will not 2 2 therefore generally considered in national and international cause any harm to the environment. chemicals regulations, e.g. the European Union REACH Mercaptocarboxylic acids and their esters are a chemical Regulation (EU 2006). One aim of such regulations is to class of potential concern, a particular class of difunctional compounds bearing both a mercapto (-SH) and a carboxylic acid or ester moiety (-COOH or -COOR’). Several members Responsible editor: Gerald Thouand of this class are industrial chemicals, some are high production Electronic supplementary material The online version of this article volume compounds of a global production capacity of several (https://doi.org/10.1007/s11356-018-1812-x) contains supplementary thousand metric tons per year. The major part of these prod- material, which is available to authorized users. ucts is used in an industrial setting, so that release to the en- vironment can occur during production and industrial process- * Christoph Rücker ing. However, there is also some use in consumer products, ruecker@leuphana.de e.g. several thousand tons per year of thioglycolic acid and its 1 salts go into cosmetic and cleaning formulations and are there- Institute of Sustainable and Environmental Chemistry, Leuphana fore released to the environment directly or indirectly via sew- University Lüneburg, Universitätsallee 1, 21335 Lüneburg, Germany 2 age treatment plants. It is therefore of interest to better under- Analytical Chemistry Department, Faculty of Pharmacy, Suez Canal stand the environmental impact of this class of chemicals, both University, Ismailia 41522, Egypt of the parent compounds and of any transformation products. Bruno Bock Thiochemicals, 21436 Marschacht, Germany 18394 Environ Sci Pollut Res (2018) 25:18393–18411 The present study should help to obtain a general view of biodegradation of monofunctional compound classes, the these compounds’ biodegradation behaviour, by comparing problem of biodegradation of difunctional or even more com- literature data with our own experimental data to be generated. plex compounds was not adequately envisaged previously. As For measuring biodegradability, a tiered approach is usually a consequence, some computer models for predicting the bio- adopted in regulations, as described in guidance documents degradability of organic compounds are available, but they (OECD 1995, 2006;ECHA 2017). The first tier tests are are built on information obtained largely from simple screening tests described in the OECD technical guidance se- monofunctional compounds, and their predictive ability for ries 301A-F for ready biodegradability (OECD 1992). These multifunctional compounds is therefore questionable (Rücker tests are very stringent, offering unadapted microorganisms and Kümmerer 2012). For example, models such as the present in low concentration and diversity only limited oppor- Biowin models (US EPA) generally assume additive effects tunities for adaptation to and biodegradation of the chemical of various structural fragments, though non-additive interac- tested (test duration 28 days). They do not simulate real envi- tions of functional groups are to be expected. A goal to be ronmental conditions but give an indication of biodegradability aimed at in the future is prediction of any compound’sbiodeg- of test compounds under various conditions (UN 2015). radation behaviour from its molecular structure, including A positive result in such a test (e.g. oxygen consumption complex structures. This will also allow design of environ- 60% or more of theoretical oxygen demand within 28 days in mentally biodegradable chemicals. All this requires a large tests 301D and 301F) can be considered as indication of rapid body of data for building and validating models, data that and more or less complete degradation under most environ- are largely not available hitherto. In view of all combinations mental conditions including biological sewage treatment of functional groups that may be present in an environmental plants (STPs). Such a test result demonstrates conversion to contaminant, there is a huge amount of work to be done, and 2− + stable inorganic products such as CO ,H O, SO and NH our study on mercaptocarboxylic acids is thus a first step in a 2 2 4 4 or NO (OECD 2006; Guhl and Steber 2006). Chemicals direction that needs to be pursued for several other classes of passing an OECD 301 test are classified as readily biodegrad- difunctional compounds. able, and further testing is not normally required (OECD Understanding the behaviour of difunctional compounds 2006;ECHA 2017). However, continuous release of such a obviously requires as a prerequisite thorough knowledge of substance may cause continuous exposure if degradation is how the single functional groups involved influence the re- slow compared to release, and further testing may therefore spective outcome. It is now well known that some molecular be envisaged (OECD 2006). substructures enhance or hinder biodegradation. These find- In contrast, a negative result in an OECD 301 test does not ings were formulated as “rules of thumb” that are based on necessarily mean that the chemical will not be degraded under more or less evidence. Relevant in our context, many carbox- ylic acids and esters are rather well biodegradable. So the rule relevant environmental conditions, but it can be considered as indication of a potentially persistent chemical and may trigger “carboxylic acid or ester groups enhance biodegradation” be- second tier tests (inherent biodegradability, tests OECD 302A- came popular (Howard 2000; Boethling et al. 2007; Cheng C). If necessary, the highest (third) tier is applied, a simulation et al. 2012). However, it turned out that data supporting this of biodegradation either in the aerobic treatment stage of STPs rule were never explicitly compiled, nor was it specified or in environmental compartments such as fresh or marine against which compound a carboxylic acid or ester was to be surface water (ECHA 2017; Kowalczyk et al. 2015). These compared with respect to biodegradability, the compound higher tier tests require considerably more time, technical bearing a methyl group or the one bearing a H atom in place equipment, manpower and funds. Therefore, data from simu- of the COOH/COOR’ group. lation tests are typically not available for most marketed For the other monofunctional compound class of interest chemicals, and biodegradability or persistence of a chemical here, mercaptans, some biodegradability data are available but in the environment is usually judged based on first or second a rule of thumb was not formulated hitherto. tier tests only. Thus, questions to be addressed in the present study were Acompound’s biodegradability depends on its molecular the following. structure and in particular on the functional groups present. Many compounds of environmental concern are complex, 1. Is the rule of thumb for biodegradability of carboxylic containing more than a single functional group, for example acids/esters well based on experimental data? pesticides, herbicides, dyes and pharmaceuticals. Given the 2. Can a rule of thumb be formulated for biodegradability of increasing number and complexity of chemicals and pharma- mercaptans, based on experimental data? ceuticals ending up in the aquatic environment, their biode- 3. What is the biodegradation behaviour of compounds con- gradability (or not so), decisive for their environmental im- taining both a carboxylic acid/ester and a mercaptan func- pact, will increasingly influence environmental quality in the tional group? Can a general rule be given on the biode- future. While some knowledge has accumulated on the gradability of this class of difunctional compounds? Environ Sci Pollut Res (2018) 25:18393–18411 18395 We therefore first searched the literature for biodegrada- 25359-71-1, DiPETMP) and ethoxylated trimethylolpropane tion of carboxylic acids and esters and corresponding com- tris(3-mercaptopropionate) (CAS-RN 345352-19-4, 674786- pounds without these functional groups, in order to con- 83-5, ETTMP 700) are mixtures of esters obtained from an firm or refute the mentioned rule of thumb. In a second excess of 3-mercaptopropionic acid and the respective multi- step, the literature was searched for biodegradation of mer- functional alcohol (trimethylolpropane, pentaerythritol, captans and corresponding hydrocarbons and analogous dipentaerythritol, ethoxylated trimethylolpropane). Along alcohols, in order to obtain a rule of thumb for biodegra- with the major constituent, they contain partially esterified dation of mercaptans. As to the third question, available products as well as thiolesters. experimental biodegradation data for mercaptocarboxylic Myristyl thioglycolate (CAS-RN 84238-40-4, C TG) is a acids and esters turned out to be partially contradictive, mixture of mostly non-branched C alkyl esters of 10-16 which may be explained by their origin from tests per- thioglycolic acid. formed according to various protocols in various laborato- Glyceryl monothioglycolate (CAS-RN 30618-84-9, GMT) is ries. We therefore decided to measure biodegradation of a reaction mixture of glycerol and thioglycolic acid. Main con- several mercaptocarboxylic acids and esters and, for com- stituents are the isomeric monoesters, minor constituents are di- parison, of several miscellaneous divalent sulfur com- and triesters, as well as free glycerol and thioglycolic acid. pounds (mercaptans, sulfides, disulfides) in two standard The mineral medium used in the experiments was prepared ready biodegradability tests, the Closed Bottle Test (CBT, according to the OECD 301 guidelines from deionised water 301D) and the Manometric Respirometry Test (MRT, (Miele Aqua Purification model G 7795, conductivity ≤ 5 μS/ 301F), using a single source of inoculum. Thereby, we cm) and analytical grade reagent salts in the concentrations expected to obtain more consistent results. In addition to specified there (OECD 1992). Effluent of the municipal the OECD 301D and F methodology, LC-UV/MS/MS was sewage treatment plant (STP) of Lüneburg, Germany employed to identify any transformation products. (144,000 inhabitant equivalents) was collected on the day of test start. The effluent sample was filtered and then used as inoculum directly. The Lüneburg STP treats typical municipal Experimental sewage, there is no industry connected to the STP that deals with the compounds studied here. As to the presence of other Acetonitrile and methanol (HiPerSolv CHROMANORM, organic or toxic substances, controls were run according to the LC-MS grade, BDH Prolabo) and formic acid (analytical guidelines: the quality control results and the inoculum blank grade) were purchased from VWR International GmbH results were as required (validity criteria number 2 and 4 (Darmstadt, Germany). Dithiodiglycolic acid (CAS-RN 505- below). 73-7, DTDGA, TGA disulfide) and 3,3′-dithiodipropionic ac- The Closed Bottle Test (CBT, OECD 301D) is consid- id (CAS-RN 1119-62-6, DTDPA, 3-MPA disulfide) were ob- ered the most stringent among the OECD 301 series ready tained from Sigma-Aldrich (Steinheim, Germany), dimethyl biodegradability (RB) tests (OECD 1992). It works at low 3,3′-dithiodipropionate (CAS-RN 15441-06-2, MMP disul- test compound concentration (theoretical oxygen demand 4 6 fide) from abcr GmbH, Karlsruhe, Germany. 1,2,5- (ThOD) ~ 5 mg/L) and low bacterial density (10 –10 colony Trithiepane (CAS-RN 6576-93-8) was purchased from forming units (CFU)/mL). In our CBT modification, we used as Envilytics, Wiesbaden, Germany. All other tested inoculum two drops of STP effluent per litre of mineral solu- organosulfur compounds were provided by Bruno Bock tion. This inoculum amount was enough for degrading sodium Thiochemicals (Marschacht, Germany), were of technical acetate sufficiently (validity criterion 2, see below), at the same grade and were used without purification. Some of these are time being safe with respect to criterion 4. multiconstituent substances (MCSs): isotridecyl thioglycolate The test comprised four completely filled (no headspace) (CAS-RN 57417-85-3, iC TG), isotridecyl 3- bottles (inoculum blank, quality control, test proper, toxicity mercaptopropionate (CAS-RN 1040871-35-9, iC MP), control, each in duplicate) and was run for 28 days at 20 ± 1 °C isooctyl thioglycolate (CAS-RN 25103-09-7, iOTG) and in the dark (OECD 1992). For details, see Table S1 in the isooctyl 3-mercaptopropionate (CAS-RN 30374-01-7, Electronic supplementary material. iOMP) are mixtures of esters obtained from the respective Oxygenconcentrationinthebottleswas monitoreddai- carboxylic acid (thioglycolic acid or 3-mercaptopropionic ac- ly from outside using the Fibox3 system (fiber-optic id) and mixtures of branched mostly C or C primary alco- oxygen meter with temperature sensor, Precision Sensing 13 8 hols (CAS-RN 68526-86-3 and 68526-83-0, respectively). GmbH, Regensburg, Germany) based on a sensor spot in Trimethylolpropane tris(3-mercaptopropionate) (CAS-RN each bottle (Friedrich et al. 2013). The temperature was 33007-83-9, TMPMP), pentaerythrityl tetrakis(3- monitored daily; the pH was measured at days 0 and 28, mercaptopropionate) (CAS-RN 7575-23-7, PETMP), was adjusted to 6.5–8 if necessary at day 0 and was in this dipentaerythrityl hexakis(3-mercaptopropionate) (CAS-RN range at day 28. 18396 Environ Sci Pollut Res (2018) 25:18393–18411 For a CBT result to be valid, five criteria must be met achieve the pass level within 10 days after the first 10% degra- simultaneously: dation occurred (“10 day window”), except for the 301C test. Sufficiently soluble compounds were dissolved in mineral 1. The difference in degradation between replicate test bot- solution and then dispensed into the bottles, whereas com- tles should be less than 20%. pounds of insufficient solubility were, separately for each bot- 2. The biodegradation of reference compound (sodium ace- tle, weighed using a miniature polystyrene weighing boat, tate) in the quality control has to be at least 60% by day 14. thrown into the bottle and then suspended in mineral solution 3. The biodegradation in the toxicity control should be at (“direct weighing”). In both tests at test begin (day 0) and test least 25% of the total ThOD by day 14. end (day 28), samples were taken (except in cases of direct 4. Oxygen consumption in the inoculum blank should not weighing) and were stored at − 80 °C until analysis. Samples exceed 1.5 mg/L by day 28. were analysed for the test compound (primary elimination) 5. Oxygen concentration in the test bottles should not fall and for transformation products (TPs) by HPLC-UV and below 0.5 mg/L at any time. HPLC-UV-MS/MS (ion trap) whenever possible. In most cases, MRT samples rather than CBT samples were analysed The Manometric Respirometry Test (MRT, OECD 301F) since the higher concentration facilitated identification of TPs. uses a higher test compound concentration (50–100 mg For HPLC-UV, a Prominence HPLC instrument ThOD/L according to the guideline, ~ 30 mg ThOD/L in our (Shimadzu, Duisburg, Germany) was used. Separation was 7 8 ® modification) and higher bacterial density (10 –10 CFU/mL performed on a RP-18 column (NUCLEODUR 100-3 C18 according to the guideline, 80 mL STP effluent/L final solu- ec, 2 mm ID, 125 mm, Macherey & Nagel, Düren, Germany) tion in our work) and thus higher bacterial diversity (OECD protected by an EC guard column (NUCLEODUR 100-3 1992). Another difference between CBT and MRT is the im- C18 ec, 4 × 2 mm). Elution was isocratic or gradient using plementation of a further control bottle (sterile control) and the mixtures of 0.1% formic acid in water (solution A) and use of only one bottle for toxicity control. For details, see 100% acetonitrile (solution B). The flow rate was 0.25 mL/ Table S1 in the Electronic supplementary material. The test min, the oven temperature 40 °C and the detector wavelength was run for 28 days at 20 ± 1 °C in the dark with gentle stir- 210 nm. Twenty to fifty microlitres of each sample were ring. Oxygen consumption was recorded daily using an injected without any workup. OxiTop control OC110 system (WTW, Weilheim, HPLC-UV-MS was performed on an Agilent Technologies Germany), measuring the pressure decrease in the headspace series 1100 HPLC instrument (Agilent Technologies, (about one third of the bottle volume), while CO produced Böblingen, Germany). Column, eluent solutions and run pa- was removed by adsorption/absorption to NaOH pellets/conc. rameters were as above. To the chromatograph, a Bruker NaOH solution. The temperature was monitored daily, the pH Daltonic Esquire 6000 plus ion-trap mass spectrometer was was measured at days 0 and 28, was adjusted to 6.5–8if coupled, equipped with an atmospheric pressure electrospray necessary at day 0 and was in this range at day 28. ionisation (ESI) interface and a Bruker data analysis system For a MRT result to be valid, five criteria must be met (Bruker Daltonic GmbH, Bremen, Germany). The mass spec- simultaneously: trometer was operated in the positive mode. For analytical details, see the “Electronic supplementary material”. 1.–3. The first three criteria are the same as in CBT. 4. The oxygen consumption in the inoculum blank should be at most 60 mg/L by day 28. Results and discussion 5. If the oxygen consumption by the test substance at test end is < 60%, then the pH must be within the range 6–8.5. Biodegradation data from the literature The pass level in both CBT and MRT is oxygen consump- Tables 1 and 2 show biodegradation test results published in tion at day 28 of 60% of the theoretical oxygen demand. ThOD the scientific and regulatory literature. The purpose of Tables 1 was calculated under the assumption that sulfur is oxidised to and 2 is to provide a semi-quantitative overview of the com- 2− SO , according to the guideline (OECD 1992). To be consid- pounds’ biodegradability as a function of specific functional ered readily biodegradable, a substance is further required to groups present in the molecules. These data come from differ- ent tests and therefore are not strictly comparable, they should not be used indiscriminately together, e.g. for training of a The 20% difference is usually interpreted as a 20 %points difference, since the unit of biodegradation is % (of a theoretical value). For example, if two biodegradation QSAR model. replicate bottles show 41 and 59% biodegradation, the difference of 18 Unless stated otherwise, results in Tables 1 and 2 were %points is considered still acceptable, though 59 and 41 (in any unit) differ obtained in the 28-day (or 14-day) OECD 301C (MITI-I) test, by ~ 44 or ~ 30.5%, based on the lower or the higher value, respectively, which would be unacceptable when taking the criterion literally. as contained in the J-CHECK database. Results of the 14-day Environ Sci Pollut Res (2018) 25:18393–18411 18397 Table 1 Biodegradation test results (oxygen consumption as % of the theoretical oxygen demand) of carboxylic acids, esters and corresponding compounds without these functional groups, taken from the literature. Data are primarily from the 28-day (or 14-day) OECD 301C (MITI-I) test Biodegradation [%] R-COOH name R R-COOH R-COOCH R-COOC H R-H R-CH 3 2 5 3 Acetic acid ≥74 ≥92 ≥94 n.a. n.a. b a c Chloroacetic acid ≥65 45 75 1 1 d,e 2-Chloropropionic acid 77 RB n.a. 1 0 Dichloroacetic acid ≥97 n.a. n.a. 13 NRB Trichloroacetic acid 7 n.a 67 ≥0 ≥0 Acrylic acid ≥67.8 ≥37 ≥51.5 n.a. 1 Methacrylic acid ≥91 ≥94.3 79.1 1 n.a. Cyanoacetic acid ≥90 99 68 65 n.a. 4-Isopropylbenzoic acid ≥89 n.a. n.a. ≥33 ≥88 2-Chlorobenzoic acid ≥5.6 n.a. n.a. 0 0 2-Methylaminobenzoic acid ≥85 n.a n.a. ≥1.4 1 n.a. not available, The corresponding carboxylic acid was formed to some extent during the test b c d e Test duration 21 days, OECD 301F, ECHA database, Test unspecified, Cheng et al. (2012), Test unspecified, Pizzo et al. (2013) f g h i OECD 301D, ECHA database, OECD 301A, ECHA database, OECD 301E, ECHA database, OECD 301B, ECHA database 18398 Environ Sci Pollut Res (2018) 25:18393–18411 Table 2 Biodegradation results (oxygen consumption as % of the theoretical oxygen demand) of mercaptans and corresponding hydrocarbons and alcohols, taken from the literature. Data are primarily from the 28-day (or 14-day) OECD 301C (MITI-I) test Biodegradation [%] R-SH name R R-SH R-H R-OH Ethyl mercaptan 0 n.a. ≥89 39(301D) 77(301F) 79(301D) n-Dodecyl mercaptan ECHA ECHA ECHA 43-96 (301B) ~74 (306) ECHA; n-Octadecyl mercaptan 0 ECHA 38-69 (301D) ECHA 2-Mercaptoethanol 19 ≥89 ≥90 Thiophenol 0 ≥40 ≥85 b c NRB 2≥0; 96 ; 1-Naphthalenethiol d,e b RB ; fast 2-Naphthalenethiol 0 2≥68 n.a. not available The corresponding disulfide was formed to some extent during the test Test unspecified, Biowin1/2 training set Test unspecified, Pizzo et al. (2013) Test unspecified, Cheng et al. (2012) Biowin5/6 training set (MITI-I model) test are given as “≥xx”, where xx is the numerical result mea- corresponding methyl compound and the compound lacking sured after 14 days. In cases of no result found in the J- that methyl group. Each row of Table 1 contains data for (at CHECK database,the REAXYS and ECHA databases were least) a pair of compounds, so that comparison within a row consulted, then training and test datasets of the Biowin1/2 and shows the impact of a COOH/COOCH /COOC H group on 3 2 5 Biowin5/6 models (US EPA EPI Suite), and as a last resort biodegradability in context R. Comparison between rows then experimental data cited in recent papers on biodegradation shows whether the effect is same or different in different struc- modelling (Cheng et al. 2012; Pizzo et al. 2013). The latter tural contexts as described by the variation of R. sources give results as readily biodegradable/not readily bio- As an exception, we included acetic acid despite the lack of degradable (RB/NRB) only. data for methane and ethane, for the relevance of sodium acetate as the reference compound in biodegradability tests. Carboxylic acids and their esters Examination of Table 1 shows that a carboxylic acid or its methyl or ethyl ester is more easily biodegradable than In Table 1, biodegradation test results are compiled for car- corresponding compounds bearing a CH or H at the same boxylic acids and their methyl or ethyl esters and for the position, and this seems to be generally true. For the pair 4- Environ Sci Pollut Res (2018) 25:18393–18411 18399 methylaminobenzoic acid and N-methylaniline, seemingly a Experimental biodegradation data from the present counterexample, the difference between 0 and 1.4% is within study experimental uncertainty. Thus, the rule of thumb for carbox- ylic acids/esters seems to be well based on facts. We tested biodegradation of 24 substances containing divalent No general trend in biodegradability is seen in Table 1 on sulfur, i.e. mercaptocarboxylic acids, their esters, disulfides, comparing the esters to the carboxylic acids. sulfides and mercaptans. As a first experimental result of the Partial hydrolysis of esters to alcohols and acids is to be present study, for all tested compounds in all biodegradation expected in aqueous solution even at neutral pH and 20 °C experiments no indication of toxicity to the bacteria was found during a reaction time extended as long as 28 days and was in at the tested substance concentrations. fact observed for several esters in the OECD 301C test. Table 3 presents the biodegradation test results obtained in Examples are ethyl acetate and methyl chloroacetate in this study. In case of an invalid result, the test was repeated Table 1, and propyl acetate, sec-butyl acetate, n-butyl acrylate, until a valid result was obtained. For a few compounds, a test methyl 4,4-dimethyl-3-oxopentanoate, acetylsalicylic acid (J- was repeated despite of a valid first result already obtained, in CHECK database). Hydrolysis is a step neutral with respect to order to examine intra-laboratory reproducibility. For all re- oxidation, it does not show up in the oxygen balance but may sults ≥ 60%, the 10-day window was passed. influence the biodegradation rate by producing more water- Tested substances are subdivided in Table 3 into three soluble and bioavailable species. groups. Group 1 contains thioglycolic acid (TGA) and its derivatives (ammonium salt, disulfide and esters). Group 2 Mercaptans comprises 3-mercaptopropionic acid (3-MPA) and its de- rivatives (disulfides and esters). Group 3 contains miscel- Table 2 presents biodegradation test results of mercaptans and laneous compounds of divalent sulfur, e.g. thiolactic acid the corresponding hydrocarbons and alcohols for comparison. (TLA), thioethers (MBT, Di-2-EHTDG) and mercaptans Each row of Table 2 contains data for (at least) a pair of (DMDS, DMPT). compounds, so that comparison within a row shows the im- pact of a SH (or OH) group on biodegradability in structural Discussion of reproducibility context R. Comparison between rows then shows whether the effect is same or different in different structural contexts as Generally, reproducibility is an issue in all biodegradability described by the variation of R. test results. A closer look on our data reveals that poor repro- The limited data in Table 2 show that a mercapto group is ducibility appears on various levels, as detailed in the follow- detrimental to biodegradation compared with the correspond- ing subsections. ing hydrocarbon, in contrast to a hydroxyl group that en- hances biodegradation. This trend is seen in all rows of Intra-experiment reproducibility Table 2 and so may be considered a rule of thumb. A mercapto group, however, does not absolutely pro- Table 4 gathers discrepancies between duplicate bottles within hibit biodegradation, as demonstrated by n-dodecyl mer- an experiment from Table 3 for CBT and MRT, i.e. all those captan and 2-mercaptoethanol. This latter result may be results that are invalid for violation of criterion 1. interpreted as being caused by the presence of structural Discrepancies between duplicate bottles in the same exper- elements both favouring (hydrocarbon chain of intermedi- iment were occasionally observed earlier (Painter and King ate length, OH group) and hindering biodegradation (SH). 1985), the guidelines explain this phenomenon by differences The data for 1-naphthol demonstrate that published exper- in the composition of the inoculum (ECHA 2017,page 214; imental biodegradation test results are generally to be consid- OECD 1992). Nevertheless, some authors consider such re- ered with caution: the 14-day MITI-I test resulted in 0% bio- sults not only “invalid”, but also useless, meaning that “some- degradation, whereas the training set of models Biowin5/6 thing went wrong with the experiment”. The experiment is claims a MITI-I test result of > 60% (“readily biodegradable”), then repeated until a valid result is obtained, and invalid as do Cheng et al. (2012) for an unspecified test. Pizzo results are discarded rather than reported. This practice et al. (2013) even report 96% biodegradation in an unspec- leads to the wrong impression that biodegradation experi- ified test, and the Biowin1/2 training set claims “fast bio- ments are as reproducible as physicochemical measure- degradation” for this compound. The Biowin1/2 data of 1- ments. To correct this view, we here report our invalid naphthol is from the Environmental Fate Database of experiments also and offer alternative explanations. Syracuse Research Corporation, which unfortunately is Moreover, such invalid experimental results are by no no longer publicly available. Further, the four “high” means useless, in particular if both parallel bottles exceed claims may result from tapping thesame sourceratherthan the pass level, as is the case for TGA, DTDGA, 3-MPA coming from four independent experiments. disulfide and MMP disulfide in MRT (Table 4). 18400 Environ Sci Pollut Res (2018) 25:18393–18411 Table 3 Biodegradation test results obtained in the present study. For each experiment, the % O consumption in the two parallel bottles is given, and for valid results the average ±SD, otherwise the criterion rendering a result invalid. For structures, see Table 7 Compound name (abbreviation) Experiment number OECD 301D (CBT) OECD 301F (MRT) [CAS-RN] O consumption/ Average in a valid O consumption/ Average in a valid 2 2 ThOD [%] experiment ±SD ThOD [%] experiment ±SD TGA and derivatives Thioglycolic acid (TGA) 1 31.8, 45.3 38.5 ± 9.5 114.4, 92.1 Not valid, [68-11-1] criterion 1 2 49.2, 51.3 50.2 ± 1.5 78.8, 94.4 86.6 ± 11.1 Diammonium dithiodiglycolate (DADTDG, 1 123.8, 106.4 115.1 ± 12.3 130.0, 132.0 131.0 ± 1.4 TGA disulfide ammonium salt) calculated without [68223-93-8] nitrification 1 76.6, 65.9 71.3 ± 7.6 80.5, 81.7 81.1 ± 0.9 re-calculated with nitrification 2 14.9, 29.2 22.0 ± 10.1 79.4, 90.8 85.1 ± 8.0 calc. with nitrification 3 90.9, 86.0 88.5 ± 3.5 calc. with nitrification Dithiodiglycolic acid (DTDGA, 1 65.9, 62.2 64.1 ± 2.6 87.9, 111.2 Not valid, TGA disulfide) criterion 1 [505-73-7] 2 60.3, 55.7 58.0 ± 3.3 2-Ethylhexyl thioglycolate (2-EHTG) 1* 58.3, 61.2 59.7 ± 2.0 28.2, 54.7 Not valid, [7659-86-1] criterion 1 2* 18.4, 13.9 16.2 ± 3.2 Isotridecyl thioglycolate (iC TG) 1** 1.4, 10.0 Not valid, 40.9, 22.5 31.7 ± 13.0 [57417-85-3] criterion 4 2** 14.6, 9.5 12.0 ± 3.6 Glycol di(mercaptoacetate) (GDMA) 1 62.2 Not valid, 62.1, 69.8 65.9 ± 5.4 [123-81-9] criterion 1 2 82.5, 72.4 Not valid, criterion 4 3 70.0 Not valid, criterion 1 4 79.4, 77.1 78.3 ± 1.6 Glyceryl monothioglycolate (GMT) 1 56.0, 50.7 53.4 ± 3.7 79.8, 80.8 80.3 ± 0.7 [30618-84-9] 3-MPA and derivatives 3-Mercaptopropionic acid (3-MPA) 1 0.4, 32.4 Not valid, 78.7, 87.0 82.8 ± 5.9 [107-96-0] criterion 1 2 27.4, − 7.3 Not valid, 101.4, 90.1 95.8 ± 8.0 criteria 1 and 4 3 0.6, 3.8 2.2 ± 2.3 Dithiodipropionic acid 1 3.9, 1.5 2.7 ± 1.7 121.6, 87.9 Not valid, (DTDPA, 3-MPA disulfide) criterion 1 [1119-62-6] 2 93.7, 100.3 97.0 ± 4.7 Methyl 3-mercaptopropionate (MMP) 1 65.0, 61.7 63.4 ± 2.3 23.3, 12.0 17.7 ± 8.0 [2935-90-2] 2 16.0, 21.0 18.5 ± 3.6 3 74.1, 14.1 Not valid, criterion 1 4 5.4, 12.3 8.8 ± 4.9 Propanoic acid, 3,3′-dithiobis-, 1 6.6, 4.1 5.4 ± 1.8 61.6, 84.2 Not valid, dimethyl ester (MMP disulfide) criterion 1 [15441-06-2] 2 10.6, 13.0 11.8 ± 1.7 14.0, 21.7 17.8 ± 5.4 3* 21.6, 1.3 Not valid, criterion 1 Environ Sci Pollut Res (2018) 25:18393–18411 18401 Table 3 (continued) Compound name (abbreviation) Experiment number OECD 301D (CBT) OECD 301F (MRT) [CAS-RN] O consumption/ Average in a valid O consumption/ Average in a valid 2 2 ThOD [%] experiment ±SD ThOD [%] experiment ±SD Butyl 3-mercaptopropionate (BuMP) 1 55.6, 29.0 Not valid, 15.0, 20.7 17.8 ± 4.0 [16215-21-7] criterion 1 2 57.4, 60.3 Not valid, criterion 4 3 61.9, 62.5 62.2 ± 0.5 4 40.8, 41.1 41.0 ± 0.2 2-Ethylhexyl 3-mercaptopropionate 1* 54.3, 49.8 52.1 ± 3.2 56.4, 27.0 Not valid, (2-EHMP) criterion 1 [50448-95-8] 2* 42.4, 51.6 47.0 ± 6.5 Isooctyl 3-mercaptopropionate (iOMP) 1** 31.1, 32.1 31.6 ± 0.8 − 8.2, 4.6 − 1.8 ± 9.0 [30374-01-7] Isotridecyl 3-mercaptopropionate (iC MP) 1** 10.5, 17.6 14.1 ± 5.0 29.5, − 11.1 Not valid, [1040871-35-9] criterion 1 2** 34.2, 35.8 35.0 ± 1.2 Glycol di(3-mercaptopropionate) (GDMP)12.1, − 0.8 0.7 ± 2.0 73.4, 39.8 Not valid, [22504-50-3] criterion 1 2 10.4, 13.6 12.0 ± 2.3 32.8, 50.4 41.6 ± 12.5 3* 33.4, 41.0 37.2 ± 5.3 Trimethylolpropane tris- 1** 8.2, 9.9 9.1 ± 1.2 32.9, 55.4 Not valid, (3-mercaptopropionate) (TMPMP) criterion 1 [33007-83-9] 2** 4.7, 1.7 3.2 ± 2.1 Dipentaerythrityl hexakis- 1** 1.4, 3.3 2.4 ± 1.4 14.8, 42.4 Not valid, (3-mercaptopropionate) (DiPETMP) criterion 1 [25359-71-1] 2** 12.0, − 5.5 3.2 ± 12.3 Ethoxylated trimethylolpropane tris- 1** 3.1, − 3.2 Not valid, 35.1, 40.0 37.6 ± 3.5 (3-mercaptopropionate) (ETTMP 700) criterion 4 [674786-83-5 or 345352-19-4] 2** 9.2, 8.6 8.9 ± 0.5 Miscellaneous compounds containing divalent sulfur Thiolactic acid (TLA) 1 11.9, 12.8 12.4 ± 0.6 66.6, 75.9 71.2 ± 6.6 [79-42-5] Di(2-ethylhexyl) thiodiglycolate 1** 19.3, 14.0 Not valid, 41.5, 34.3 37.9 ± 5.1 (Di-2-EHTDG) criterion 4 [24293-43-4] 2** 22.7, 33.8 28.3 ± 7.9 Methylene bis(butyl thioglycolate) (MBT) 1* 49.4, 48.5 48.9 ± 0.7 56.6, 63.1 59.8 ± 4.6 [14338-82-0] 2* 52.2, 63.2 57.7 ± 7.8 Bis(2-mercaptoethyl) sulfide (DMDS) 1 − 0.2, − 2.0 − 1.1 ± 1.3 35.8, − 12.9 Not valid, [3570-55-6] criterion 1 2* 1.8, 33.4 Not valid, criterion 1 4-Mercaptomethyl-3,6-dithia-1,8- 1 − 5.0, − 2.7 − 3.9 ± 1.6 octanedithiol (DMPT) [131538-00-6] Multiconstituent substance (MCS), for details, see “Experimental” Result from one test bottle only, the other one became leaky during the experiment *Due to poor solubility, this substance was directly weighed in MRT **Due to very poor solubility, this substance was directly weighed both in CBT and in MRT 18402 Environ Sci Pollut Res (2018) 25:18393–18411 Table 4 Non-reproducibility within the same experiment may contain different numbers of cells of different diversity just by chance. CBT biodegradation [%] MRT biodegradation [%] 3-MPA 0, 32; − 7, 27 TGA 92, 114 MMP 14, 74 DTDGA 88, 111 BuMP 29, 56 2-EHTG 28, 55 Intra-laboratory reproducibility 3-MPA disulfide 88, 122 Few scientists perform a second or third biodegradation ex- MMP disulfide 1, 22; 62, 84 periment if a valid result is already at hand. If such double 2-EHMP 27, 56 measurements are nevertheless done and produce conflicting iC MP − 11, 30 results, each being valid, then such data likely will not be GDMP 40, 73 published. Therefore, this kind of problem also is often not TMPMP 33, 55 apparent, it is not explicitly mentioned in the guidelines, and DiPETMP 15, 42 its prevalence is unknown. DMDS − 13, 36; 2, 33 Table 5 gathers discrepancies from Table 3 between differ- ent experiments performed for the same compound under the same test protocol in the present study, from CBT and MRT. In our experiments, MRT results seem to be more repro- At least some of the discrepancies observed in the ducible than CBT results (Table 5). The CBT results for three present experiments may be explained by experimental out of six compounds led to conflicting assessments (above problems: and below the pass level), while in MRT this happened for one out of four compounds only. This difference can be explained – Poorly soluble substances (2-EHMP, 2-EHTG, iC MP, by the higher bacterial density/diversity in MRT compared to TMPMP, DiPETMP, MMP disulfide in its experiment 3) CBT. were directly weighed into the test bottles, suffering from Discrepancies between two seemingly identical biodeg- difficulties in weighing mg amounts of viscous material radation experiments performed at different times may be into several parallel bottles. due to different lag times. Lag time is the time required for – GDMP (in MRT experiment 1) and MMP disulfide (in adaptation by the microbial population to metabolise the MRT experiment 1) were tested applying the usual substrate efficiently. Different lag times can be caused by substance preparation as a solution despite borderline a variation in the diversity of the bacterial population in two solubility. Incomplete solution then may have resulted experiments, that is, by presence or absence of competent in different amounts of test substance in the two par- bacteria. In a former study, bacterial strains were isolated allel bottles. from activated sludge that not only tolerate organosulfur – In two CBT experiments with GDMA, one of the two carboxylic acid esters, but even were able to use them as parallel bottles became leaky during the test, as mani- sole carbon and energy source (Toups et al. 2010). Such fested by the O concentration in one bottle unexpect- specialist strains are not necessarily those that biodegrade edly increasing after some days of decrease. These these compounds in aerobic real environments or activated bottles were therefore excluded from further consider- sludge, since they may be outcompeted by generalists. ation. However, a miniature leak will have a less ob- However, if sufficient amounts of these compounds are vious effect, the oxygen concentration decreasing available as sole carbon source as in an OECD 301 merely more slowly than in a tight bottle, so that the apparent oxygen consumption in that bottle will lag behind. This is what we observed repeatedly in this Table 5 Reproducibility of test results, valid results only, i.e. each study. Thereby a final difference in apparent biodegra- number is the mean from two bottles differing by no more than 20% dation of > 20% may or may not build up between CBT biodegradation [%] MRT biodegradation [%] parallel bottles. Thus, a possible tiny leak is a suffi- cient explanation for any discrepancies between paral- DADTDG 22, 71, 89 DADTDG 81, 85 lel bottles at day 28. GDMP 1, 12 GDMP 37, 42 – Finally, bacterial density and diversity may differ in TGA 39, 50 3-MPA 83, 96 parallel bottles following from uneven dispersion of MMP 9, 18, 63 MBT 58, 60 bacterial species in the mineral solution. If, e.g. com- MMP disulfide 5, 12 petent bacteria form aggregates, then two samples BuMP 41, 62 (same volume) taken from a highly dilute suspension Environ Sci Pollut Res (2018) 25:18393–18411 18403 Fig. 1 Biodegradation curves of DADTDG in two different CBT experiments. a Experiment 3, lag phase 14 days. b Experiment 2, lag phase 27 days 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 -20 Days Test substance Quality control Toxicity control (measured value) Toxicity control (calculated value) 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 -20 Days Test substance Quality control Toxicity control (measured value) Toxicity control (calculated value) experiment, such specialist strains tend to increase due to treatment plant (STP) was found to show considerable tem- induction of appropriate catabolic enzymes. poral (seasonal) variation (Kim et al. 2013;Juet al. 2014). If the sigmoid biodegradation curve is shifted more or In our case, the inoculum for each experiment was freshly less along the time axis, it hits the fixed 28-day terminus taken from the same source, effluent of the Lüneburg mu- at different points on the biodegradation axis. A shift of nicipal STP, but the delay between subsequent experiments a few days can make the difference between very low with the same substance was between 2 weeks and and very high biodegradation. As an example, consider 9 months, enough time for seasonal variation of the bacte- the biodegradation curves of DADTDG, the ammonium ria population in the STP. salt of TGA disulfide (Fig. 1). DADTDG was readily biodegraded (two valid CBT results, 71 and 89%, and Variation between test results for the same compound two valid MRT results, 81 and 85%), but one valid CBT under different OECD 301 test protocols result of 22% biodegradation was also obtained (exper- iment 2). Unusually long and variable lag phases were For many substances, Table 3 shows strong differences observed in CBT (19 days in experiment 1, resulting in between CBT (OECD 301D) and MRT (OECD 301F) re- 71% biodegradation; 27 days in experiment 2, 22%; sults. Though CBT is meant to be the most stringent test, in 14 days in experiment 3, 89%). On the other hand, in some instances, degradation in CBT was higher than in MRT lag times were in good concordance (9 and MRT. Such a phenomenon has been known for many 10 days), as were the degrees of biodegradation. years, and recent comparisons between 301C and 301X Bacterial populations taken from the same source at dif- test results based on a large experimental basis revealed ferent times are not alike, but are subjected, e.g., to season- that for some compounds, 301C was more efficient, for al variations (Brillet et al. 2016). In particular, the amount others, 301X for each X from {A, B, D, E, F} of rare microbial specialists in activated sludge of a sewage (Kayashima et al. 2014). Biodegradation % Biodegradation % 18404 Environ Sci Pollut Res (2018) 25:18393–18411 Table 6 Results for structurally Chemical class Compound CBT biodegradation [%] MRT biodegradation [%] similar compounds in the same test (valid results only) Mercaptocarboxylic acids 3-MPA 2 83, 96 TGA 39, 50 87 TLA (2-MPA) 12 71 Dithiodicarboxylic acids DTDPA 3 97 DTDGA 64 58 Ethylhexyl esters 2-EHMP 52 47 2-EHTG 60 16 Glycol esters GDMP 1, 12 37, 42 GDMA 78 66 Isotridecyl esters iC MP 14 35 iC TG 12 32 Mercaptocarboxylic acids/esters 3-MPA 2 83, 96 and corresponding disulfides 3-MPA disulfide 3 97 MMP 9, 18, 63 18 MMP disulfide 5, 12 18 TGA 39, 50 87 DTDGA 64 58 Differences between structurally similar compounds Synopsis of biodegradation data in the same test Table 7 is a synopsis of biodegradation test results for Ideally, it is expected that structurally similar compounds mercaptocarboxylic acids and esters and miscellaneous diva- should behave similarly under the same 301 test protocol. lent sulfur compounds. Along with the 24 substances from Table 6 compiles the experimental data (from Table 3)of Table 3, Table 7 contains seven further compounds with avail- structurally similar compounds in the same test. able biodegradation results that were not included in the pres- As Table 6 shows, some pairs of similar compounds more or ent experimental study. less meet the expectation (3-MPA, TGA, TLA; iC MP, Column 3 shows biodegradation data found in the sci- iC TG; 3-MPA, 3-MPA disulfide; MMP, MMP disulfide), entific or regulatory literature (e.g. ECHA database or un- while others do not. In case of a mercaptan and its disulfide, published original study reports). In order to use only reli- concordance is explained by a test of the former being essen- able literature data, the quality of study results was tially a test of the latter, see “Primary elimination and transfor- assessed on a case-by-case basis using the Klimisch scor- mation products”.Table 6 suggests not to overestimate both ing system (Klimisch et al. 1997). Studies with a Klimisch concordant and divergent behaviour of similar compounds in score 1 (reliable without restriction) or 2 (reliable with single biodegradation experiments, given all the experimental restriction) only were considered for column 3. For exam- imponderability discussed above. It is probably naive to expect ple, among many results available in the literature for TGA similar compounds to behave similarly in a single biodegrada- (ranging from 0 to 100%), the reliable value of 67% tion experiment when even repeated experiments on the same (OECD 301D) was chosen from a Klimisch 2 study (van compound can lead to contradictions. We have to realise that a Ginkel and Stroo 1992). biodegradation test result for a compound is only one of several Column 4 shows the highest experimental result for each possibilities, in other words that we should consider biodegra- substance obtained in the present OECD 301D or 301F stud- dation experiments as statistical processes in the sense that they ies (Klimisch 2), taken from Table 3. are governed by some factors difficult to control. Similarity in The last column of Table 7 reports a consolidated assess- the behaviour of similar compounds can then be expected only ment of all available ready biodegradability test results for on the basis of several or even many experiments. Even then, each substance. This overall result is given as readily biode- the often high sensitivity of enzymes to small structural differ- gradable (yes) or not readily biodegradable (no). The proce- ences of potential substrates may lead to unexpected effects dure leading to column 7 entries is based on the relevant within a series of related structures. guidelines saying Environ Sci Pollut Res (2018) 25:18393–18411 18405 Table 7 Synopsis of biodegradation results from the literature (ECHA or BB, i.e. ECHA database or original study reports provided by Bruno Bock) and from the present study Result Result Substance name from liter- Readily from pre- (abbreviation) ature biode- Structure sent study [CAS-RN unless (OECD gradable, (OECD given in Table 3] 301A-F), yes or no 301D or F) source TGA and derivatives Thioglycolic acid 67% 87% (301F) yes (TGA) (301D) Dithiodiglycolic no data 64% (301D) yes acid (DTDGA) found Diammonium 80% (301B) dithiodiglycolate 89% (301D) yes ECHA (DADTDG) 2-Ethylhexyl 82% (301F) thioglycolate 60% (301D) yes ECHA (2-EHTG) Isooctyl thioglyco- 76% (301F) not included late (iOTG) yes ECHA in this work [25103-09-7] Isotridecyl no data thioglycolate 32% (301F) no found (iC TG) Myristyl thioglyco- 82% (301B) not included late (C TG) yes BB in this work [84238-40-4] 18406 Environ Sci Pollut Res (2018) 25:18393–18411 Glycol 70% (301B) di(mercaptoacetate) 78% (301D) yes BB (GDMA) Glyceryl 46% 80% monothioglycolate (301D) yes (301F) (GMT) ECHA 3-MPA and derivatives 96% 3-Mercaptopropi- (301A) 96% (301F) yes onic acid (3-MPA) ECHA Dithiodipropionic no data 97% (301F) yes acid (DTDPA) found Thiodipropionic 93% not included acid (TDPA) (301C) yes in this work [111-17-1] ECHA Methyl 3-mercapto- 46% (301B) 63% (301D) yes propionate (MMP) ECHA Propanoic acid, 3,3'-dithiobis-, no data 18% (301F) no dimethyl ester found (MMP disulfide) Butyl 3-mercapto- no data 62% (301D) yes propionate (BuMP) found 2-Ethylhexyl 3- 70% (301B) mercaptopropionate 52% (301D) yes BB (2-EHMP) Isooctyl 3-mercap- 55% (301B) topropionate 32% (301D) no ECHA (iOMP) Environ Sci Pollut Res (2018) 25:18393–18411 18407 Isotridecyl 3- 17% (301B) mercaptopropionate 35% (301F) no ECHA (iC MP) Glycol di(3-mer- 72% (301B) captopropionate) 42% (301F) yes BB (GDMP) Trimethylolpropane tris(3-mercapto- no data 9% (301D) no propionate) found (TMPMP) Pentaerythrityl tetrakis(3-mercap- 26% (301B) not included topropionate) no ECHA in this work (PETMP) [7575-23-7] Dipentaerythrityl hexakis(3-mercap- 22% (301F) 3% (301F) no topropionate) BB (DiPETMP) Ethoxylated trime- thylolpropane no data tris(3-mercapto- 38% (301F) no found propionate) (ETTMP700) Miscellaneous compounds containing divalent sulfur 38% Thiolactic acid (301A) 71% (301F) yes (TLA) BB 18408 Environ Sci Pollut Res (2018) 25:18393–18411 Di(2-ethylhexyl) dithiodiglycolate 54% (301B) not included no (Di-2-EHDTDG) BB in this work [62268-47-7] Dilauryl thiodi- propionate 82% (301C) not included yes (DLTDP) ECHA in this work [123-28-4] Distearyl 71% thiodipropionate not included (301D) yes (DSTDP) in this work ECHA [693-36-7] Di(2-ethylhexyl) 77% (301B) thiodiglycolate 38% (301F) yes ECHA (Di-2-EHTDG) Methylenebis(butyl 25% 60% thioglycolate) (301D) yes (301F) (MBT) ECHA Bis(2-mercapto- no data ethyl)sulfide 0% (301D) no found (DMDS) 4-Mercaptomethyl- 3,6-dithia-1,8- 3% (301C) 0% (301D) no octanedithiol ECHA (DMPT) Multiconstituent substance (MCS), for details see “Experimental” However, in one MRT experiment of MMP disulfide, both bottles reached the pass level of 60% (62 and 84%), thus failing with respect to criterion 1 Mean value of 56.6 and 63.1% is 59.85%, just below 60% unless rounded “Realising that ready biodegradability tests may some- Results in Table 7 can besummarisedasfollows: time fail because of the stringent test conditions, positive test results should generally supersede negative test re- 1. Simple mercaptocarboxylic acids such as TGA, 3- sults” (ECHA 2017,page 208;OECD 2006, page 3), and MPA and TLA and their simple esters are readily “When contradictory results in ready biodegradabili- biodegradable. ty tests are obtained the positive results could be con- 2. Those esters of the same acids that are not readily biode- sidered valid irrespective of negative results, when gradable are at the same time structurally complex (esters the scientific quality of the former is good and the of branched higher or multifunctional alcohols) and positive test results are well documented, …” (ECHA multiconstituent substances. They nevertheless undergo 2017, page 230). considerable biodegradation in OECD 301 tests, so there is no reason to consider them persistent in the environment. We are aware that these formulations seem to be open to 3. A disulfide motif does not prevent good biodegradability misuse by multiple testing and selective reporting, and this (DTDGA, DADTDG, DTDPA, MMP disulfide, Di-2- is another reason for showing in Table 3 all our experimen- EHDTDG), nor does a sulfide (thioether) (TDPA, tal results. DLTDP, DSTDP, Di-2-EHTDG, MBT). Environ Sci Pollut Res (2018) 25:18393–18411 18409 4. Di- or polymercaptans (without a favourable substructure) 2006)or “… can be used as evidence for inherent bio- are not biodegradable (DMDS, DMPT) in OECD 301D degradability” (ECHA 2017). or 301F tests. By these lines of argumentation, the mercaptoesters tested These observations are in good agreement with the results here, even those not formally readily biodegradable, can pre- shown in Table 2 for mercaptans. That is, the detrimental liminarily be considered inherently biodegradable. influence of a mercapto group can be counterbalanced by a Interestingly, 3-MPA was enzymatically oxidised to the favourable substructure such as an acid or ester group. corresponding sulfinic acid by a special bacterial mutant For mercaptan and sulfide functional groups, this is just the (Bruland et al. 2009) and seems to be an intermediate in the opposite of what is known for their oxygen analogues, hy- metabolism of organic sulfur compounds (methionine, homo- droxyl groups enhance biodegradation while ether groups cysteine) in natural environments (Salgado et al. 2015; are detrimental (Boethling et al. 2007). references cited therein). The higher mercaptoesters that proved not readily biode- Recently, a multilinear model for description and predic- gradable have in common a rather high molecular weight as- tion of biodegradation for general chemicals was proposed sociated with low solubility in water, and some bear branched that uses additive functional group increments and is based higher alkyl groups. Both these characteristics are known to on biodegradation data from several tests. Among all func- be detrimental to biodegradation. Others present at the mole- tional groups considered there, the mercaptan group turned cule’s surface the mercapto groups only, i.e. they can be sub- out the strongest hindering biodegradation, while carboxylic sumed under item 4, not providing a convenient point of at- acid and ester groups were found the most favourable tack for bacteria. Further, all these substances have in common (Vorberg and Tetko 2014). Our results are completely in line their MCS property. For MCSs such as petroleum products or with this picture. mixtures of homologous compounds, e.g. technical surfac- tants, the guidelines acknowledge the onset of biodegradation Primary elimination and transformation products to be delayed and/or the biodegradation curve to be less steep in comparison with the single constituents. Accordingly, the Along with biodegradation by microorganisms, abiotic pro- 10-day window is not applied for such substances (ECHA cesses may occur in the environment or in biodegradation 2017,page 210; OECD 2006, paragraphs 43 and 44). By this tests, e.g. photodegradation, hydrolysis, abiotic oxidation. logic, we expect homologous mixtures also to achieve less Thus, aerobic disulfide formation from mercaptans is well biodegradation than the single constituents after 28 days. known both during OECD 301 tests (footnote a in Table 2) The bacterial adaptation to many isomers, homologues or and by oxidation in air-saturated tap water (TGA, OECD byproducts in low concentration may be slower than adapta- 2009;3-MPA, unpublished, BB). Likewise, hydrolysis of es- tion to a single chemical in higher concentration, since differ- ters during biodegradation tests is known (“Carboxylic acids ent bacterial strains or different degradative enzymes may be and their esters”). With compound 2-EHTG, both these abiotic required. This may also apply to our MCSs that, though not reaction types were observed in the 301C test. homologous series, are complex mixtures of several related In our experiments, based on HPLC-UV-MS/MS monitor- individual compounds. ing, rapid disulfide formation on test day 0 was observed for With respect to inherent biodegradability tests (OECD 3-MPA (→ DTDPA), MMP (→ MMP disulfide) and BuMP 302A-C), the guidelines say (→ BuMP disulfide), accompanied in the ester cases (MMP, BuMP) by formation of the corresponding monoester of “Biodegradation above 20% of theoretical […]may be DTDPA and of DTDPA itself during the test, for details, see regarded as evidence of inherent, primary biodegrad- the “Electronic supplementary material”. ability” (OECD 2006,paragraph 36;ECHA 2017, The dimercapto thioether DMDS was quickly and page 216). completely transformed in CBT and MRT at day 0 to the cyclic disulfide 1,2,5-trithiepane. This disulfide, in contrast As far as substances measured here achieved above 20% bio- to the disulfides mentioned, was not further biodegraded up degradation even in an OECD 301 test, a fortiori we consider to day 28 in CBT or MRT, as it does not contain an ester or this as evidence of their inherent primary biodegradability. acid substructure. A corresponding oxidative cyclisation was Similarly, observed for GDMP, for details, see the “Electronic supple- mentary material”. “When results of ready biodegradability tests indicate Oxidative cyclisation of linear dimercaptans to cyclic that the pass level criterion is almost fulfilled (i.e. disulfides was seen earlier under similar conditions (Houk ThOD […] slightly below 60% […]) such results can and Whitesides 1987;Adamczyk et al. 2015). Formation of be used to indicate inherent biodegradability” (OECD disulfides via oxidation of SH groups can be achieved using a 18410 Environ Sci Pollut Res (2018) 25:18393–18411 wide variety of oxidants, among them molecular oxygen require some time (lag phase) that depends on the particular (Ozen and Aydin 2006; Carriletal. 2007; García Ruano course of expression and biosynthesis of enzymes, growth of et al. 2008; Abaee et al. 2011; Dewan et al. 2012;Shard bacteria and microevolution in the test flask. A conventional et al. 2014), preferably under slightly alkaline conditions, in threshold for ready biodegradability such as 60% O con- what is presumably a dimerisation of thiyl radicals formed by sumption after 28 days may be passed in one experiment but oxidation of thiolate anions (RSH→ RS → RS·,2 RS·→ not in another under seemingly identical conditions. For the RSSR). The transformation mercaptan→ disulfide consumes mercaptans and mercaptoacids/esters considered here, concur- an amount of oxygen (4 RSH + O → 2RSSR+ 2H O) that is rent abiotic reactions such as disulfide formation and ester 2 2 small in most cases, e.g. complete conversion 3-MPA→ 3- hydrolysis complicate the picture. For these reasons, test re- MPA disulfide requires 5% of the ThOD of 3-MPA. In our sults are often not reproducible. Therefore, conclusions on a examples, this step was often so rapid as to occur during day 0 particular compound’s biodegradability cannot be based on a already to a large extent, in such a case the corresponding single experiment. However, a consolidated assessment of all oxygen consumption will not even be measured in CBT or available biodegradation information for a compound class MRT since it happens before the bottles are closed. may lead to a coherent picture. The disulfide functionality, C-S-S-C, is essential in bio- Funding information This study was funded by the European Regional chemistry, e.g. contributing to the conformational integrity Development Fund, grant agreement no. CCI No 2007DE161PR001 in of proteins including highly unusual natural products the Lüneburg Innovation Incubator. (Rücker and Meringer 2002) and in detoxification of xenobi- otics via glutathione. Compliance with ethical standards Conflict of interest D. S. is an employee of Bruno Bock Thiochemicals. Conclusion Open Access This article is distributed under the terms of the Creative New findings from this study are the following: Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, 1. The rule of thumb for biodegradation of carboxylic acids/ distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link esters is well based on experimental evidence: such a to the Creative Commons license, and indicate if changes were made. functional group enhances biodegradation compared to a methyl group or to an H atom. 2. A rule of thumb for biodegradation of mercaptans can tentativelybeestablished basedonexperimentalevi- References dence: a mercapto group hinders biodegradation com- pared to an H atom and even more so compared to a Abaee MS, Mojtahedi MM, Navidipoor S (2011) Diethylamine-catalyzed hydroxy group. dimerization of thiols: an inexpensive and green method for the synthesis of homodisulfides under aqueous conditions. Synth 3. If both kinds of functional groups are present, then an Commun 41:170–176 intermediate behaviour results that strongly depends on Adamczyk J, Bal W, Krezel A (2015) Coordination properties of the compound’s other structural characteristics. Simple dithiobutylamine (DTBA), a newly introduced protein disulfide re- mercaptocarboxylic acids and their esters are either read- ducing agent. Inorg Chem 54:596–606 ily biodegradable or at least biodegradable to some extent, BB (n.d.) unpublished study report or other material from Bruno Bock Thiochemicals i.e. they are not persistent compounds. Boethling RS, Sommer E, DiFiore D (2007) Designing small molecules for biodegradability. Chem Rev 107:2207–2227 Biodegradation even of rather simple organic compounds Brillet F, Maul A, Durand MJ, Thouand G (2016) From laboratory to to CO ,H O, etc. is a multistep process, each step depending 2 2 environmental conditions: a new approach for chemical’s biode- on catabolic capabilities (enzymes) of various microorgan- gradability assessment. Environ Sci Pollut Res 23:18684–18693 Bruland N, Wübbeler JH, Steinbüchel A (2009) 3-Mercaptopropionate isms present. In standardised laboratory biodegradation tests dioxygenase, a cysteine dioxygenase homologue, catalyzes the ini- as defined in the OECD 301 series, some influencing factors tial step of 3-mercaptopropionate catabolism in the 3,3- are not and cannot be standardised, most importantly the in- thiodipropionic acid-degrading bacterium Variovorax paradoxus.J oculum. Thus, in biodegradation experiments bacterial popu- Biol Chem 284:660–672 Carril M, SanMartin R, Domínguez E, Tellitu I (2007) A highly advan- lations are used that differ with respect to the species and tageous metal-free approach to diaryl disulfides in water. Green number of bacteria capable of degrading a particular com- Chem 9:315–317 pound. Necessary enzymes of more or less substrate selectiv- Cheng F, Ikenaga Y, Zhou Y, Yu Y, Li W, Shen J, Du Z, Chen L, Xu C, ity may or may not be present or inducible. Appearance of Liu G, Lee PW, Tang Y (2012) In silico assessment of chemical biodegradability. J Chem Inf Model 52:655–669 bacterial strains actually degrading the test compound may Environ Sci Pollut Res (2018) 25:18393–18411 18411 Dewan A, Bora U, Kakati DK (2012) A mild and efficient protocol for OECD (1995) Organisation for Economic Co-operation and Development, Detailed review paper on biodegradability testing, oxidation of thiols to disulfides in water. Heteroat Chem 23:231–234 Environment Monograph No. 98 ECHA (2017) Guidance on information requirements and chemical safe- OECD (2006) Organisation for Economic Co-operation and ty assessments, Chapter R.7b: Endpoint specific guidance, Development, OECD Guidelines for the testing of chemicals, June 2017, http://www.echa.europa.eu/documents/10162/13632/ Revised introduction to the OECD Guidelines for testing of information_requirements_r7b_en.pdf. Accessed 6 Sept 2017 chemicals, Section 3, adopted 23 March 2006, http://www. ECHA (n.d.) Database: European Chemicals Agency (ECHA), http:// oecdilibrary.org/docserver/download/9730001e.pdf?expires= echa.europa.eu. Accessed 6 Sept 2017 1489593611&id=id&accname=guest&checksum= EU (2006) European Union, Regulation (EC) No 1907/2006 of the 9B023D8E206ABB2DB4C2D77DC8867032.Accessed 6 European Parliament and of the Council of 18 December 2006 Sept 2017 concerning the Registration, Evaluation, Authorisation and OECD (2009) SIDS Initial Assessment Report for SIAM 28: TGA and its Restriction of Chemicals (REACH), establishing a European ammonium salt, webnet.oecd.org/HPV/UI/handler.axd?id= Chemicals Agency, amending Directive 1999/45/EC and repealing 2712f350-08af-4de9-9bc7-0bcfd5f22fce. Accessed 6 Sept 2017 Council Regulation (EEC) No 793/93 and Commission Regulation Ozen R, Aydin F (2006) Oxidation of thiols to disulfides with molecular (EC) No 1488/94 as well as Council Directive 76/769/EEC and oxygen in subcritical water. Monatsh Chem 137:307–310 Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and Painter HA, King EF (1985) Ring test programme 1983–84, assessment 2000/21/EC, Official J. Eur. Union L 2007,L 136, 3–280 (29 of biodegradability of chemicals in water by manometric respirom- May 2007), http://eurlex.europa.eu/LexUriServ/LexUriServ.do? etry. Commission of the European Union, Luxembourg, Report uri=OJ:L:2007:136:0003:0280:EN:PDF. Accessed 6 Sept 2017 EUR 9962 EN Friedrich J, Längin A, Kümmerer K (2013) Comparison of an electro- Pizzo F, Lombardo A, Manganaro A, Benfenati E (2013) In silico models chemical and luminescence-based oxygen measuring system for use for predicting ready biodegradability under REACH: a comparative in the biodegradability testing according to the Closed Bottle Test study. Sci Total Environ 463-464:161–168 (OECD 301D). Clean – Soil Air Water 41:251–257 REAXYS (n.d.) Database https://www.reaxys.com. Accessed 6 Sept 2017 García Ruano JL, Parra A, Alemán J (2008) Efficient synthesis of Rücker C, Kümmerer K (2012) Modeling and predicting aquatic aerobic disulfides by air oxidation of thiols under sonication. Green Chem biodegradation — a review from a user’s perspective. Green Chem 10:706–711 14:875–887 Guhl W, Steber J (2006) The value of biodegradation screening test re- Rücker C, Meringer M (2002) How many organic compounds are graph- sults for predicting the elimination of chemicals’ organic carbon in theoretically nonplanar? MATCH Commun Math Comput Chem waste water treatment plants. Chemosphere 63:9–16 45:153–172, http://match.pmf.kg.ac.rs/electronic_versions/ Houk J, Whitesides GM (1987) Structure-reactivity relations for thiol- Match45/match45_153-172.pdf. Accessed 6 Sept 2017 disulfide interchange. J Am Chem Soc 109:6825–6836 Salgado P, Visnevschi-Necrasov T, Kiene RP, Azevedo I, Rocha ACS, Howard PH (2000) Biodegradation. In: Boethling RS, Mackay D (eds) Almeida CMR, Magalhaes C (2015) Determination of 3- Handbook of property estimation methods for chemicals, mercaptopropionic acid by HPLC: a sensitive method for environ- Environmental and health sciences. CRC Press, Boca Raton mental applications. J Chromatogr B 992:103–108 J-CHECK (n.d.) Database: http://www.safe.nite.go.jp/jcheck/search. Shard A, Kumar R, Saima, Sharma N, Sinha AK (2014) Amino acid and action?request_locale=en. Accessed 6 Sept 2017 water-driven tunable green protocol to access S-S/C-S bonds via Ju F, Guo F, Ye L, Xia Y, Zhang T (2014) Metagenomic analysis on seasonal aerobic oxidative coupling and hydrothiolation. RSC Adv 4: microbial variations of activated sludge from a full-scale wastewater 33399–33407 treatment plant over 4 years. Environ Microbiol Rep 6:80–89 Toups M, Wübbeler JH, Steinbüchel A (2010) Microbial utilization of the industrial wastewater pollutants 2-ethylhexylthioglycolic acid and Kayashima T, Taruki M, Katagiri K, Nabeoka R, Yoshida T, Tsuji T iso-octylthioglycolic acid by aerobic Gram-negative bacteria. (2014) Comparison of biodegradation performance of the OECD Biodegradation 21:309–319 test guideline 301C with that of other ready biodegradability tests. UN (2015) United Nations, Globally Harmonized System of classifica- Environ Toxicol Chem 33:328–333 th tion and labelling of chemicals (GHS), 6 edition, New York and Kim TS, Jeong JY, Wells GF, Park HD (2013) General and rare bacteria Geneva, 2015, Appendix 9, https://www.unece.org/fileadmin/ taxa demonstrating different temporal dynamic patterns in an acti- DAM/trans/danger/publi/ghs/ghs_rev06/English/ST-SG-AC10-30- vated sludge bioreactor. Appl Microbiol Biotechnol 97:1755–1765 Rev6e.pdf. Accessed 6 Sept 2017 Klimisch HJ, Andreae M, Tillmann U (1997) A systematic approach for US EPA (n.d.) Estimation Programs Interface Suite™ for Microsoft® evaluating the quality of experimental toxicological and ecotoxico- Windows, v 4.11, United States Environmental Protection Agency, logical data. Regul Toxicol Pharmacol 25:1–5 Washington, DC, USA, https://www.epa.gov/tsca-screening-tools/ Kowalczyk A, Martin TJ, Price OR, Snape JR, van Egmond RA, Finnegan download-epi-suitetm-estimation-program-interface-v411. CJ, Schäfer H, Davenport RJ, Bending GD (2015) Refinement of Accessed 6 Sept 2017 biodegradation tests methodologies and the proposed utility of new van Ginkel CG, Stroo CA (1992) Simple method to prolong the Closed microbial ecology techniques. Ecotoxicol Environ Saf 111:9–22 Bottle Test for the determination of the inherent biodegradability. OECD (1992) Organisation for Economic Co-operation and Ecotoxicol Environ Saf 24:319–327 Development, OECD Guideline for testing of chemicals – Ready Vorberg S, Tetko IV (2014) Modeling the biodegradability of chemical biodegradability, Paris 1992, http://www.oecd.org/chemicalsafety/ compounds using the Online CHEmical Modeling environment risk-assessment/1948209.pdf. Accessed 6 Sept 2017 (OCHEM). Mol Inf 33:73–85 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Environmental Science and Pollution Research Springer Journals

Biodegradation tests of mercaptocarboxylic acids, their esters, related divalent sulfur compounds and mercaptans

Free
19 pages

Loading next page...
 
/lp/springer_journal/biodegradation-tests-of-mercaptocarboxylic-acids-their-esters-related-YqL2HCX0uH
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2018 by The Author(s)
Subject
Environment; Environment, general; Environmental Chemistry; Ecotoxicology; Environmental Health; Atmospheric Protection/Air Quality Control/Air Pollution; Waste Water Technology / Water Pollution Control / Water Management / Aquatic Pollution
ISSN
0944-1344
eISSN
1614-7499
D.O.I.
10.1007/s11356-018-1812-x
Publisher site
See Article on Publisher Site

Abstract

Mercaptocarboxylic acids and their esters, a class of difunctional compounds bearing both a mercapto and a carboxylic acid or ester functional group, are industrial chemicals of potential environmental concern. Biodegradation of such compounds was systematically investigated here, both by literature search and by experiments (Closed Bottle Test OECD 301D and Manometric Respirometry Test OECD 301F). These compounds were found either readily biodegradable or at least biodegradable to a significant extent. Some related compounds of divalent sulfur were tested for comparison (mercaptans, sulfides, disulfides). For the two relevant monofunctional compound classes, carboxylic acids/esters and mercaptans, literature data were compiled, and by comparison with structurally similar compounds without these functional groups, the influence of COOH/COOR’ and SH groups on biodegradability was evaluated. Thereby, an existing rule of thumb for biodegradation of carboxylic acids/esters was supported by experimental data, and a rule of thumb could be formulated for mercaptans. Concurrent to biodegradation, abiotic processes were observed in the experiments, rapid oxidative formation of disulfides (dimerisation of monomercaptans and cyclisation of dimercaptans) and hydrolysis of esters. Some problems that compromise the reproducibility of biodegradation test results were discussed. . . . . Keywords Ready biodegradability tests Closed Bottle Test Manometric Respirometry Test Difunctional compounds . . . Mercaptocarboxylic acids Mercaptocarboxylic acid esters Mercaptans Biodegradation rules of thumb Introduction minimise the release of persistent chemicals into the envi- ronment. While a persistent chemical is a potential threat to Biodegradability in an aerobic aquatic phase is a key ele- the environment, a chemical that is readily biodegraded to ment in the environmental assessment of chemicals and is products such as CO and H O (mineralisation) will not 2 2 therefore generally considered in national and international cause any harm to the environment. chemicals regulations, e.g. the European Union REACH Mercaptocarboxylic acids and their esters are a chemical Regulation (EU 2006). One aim of such regulations is to class of potential concern, a particular class of difunctional compounds bearing both a mercapto (-SH) and a carboxylic acid or ester moiety (-COOH or -COOR’). Several members Responsible editor: Gerald Thouand of this class are industrial chemicals, some are high production Electronic supplementary material The online version of this article volume compounds of a global production capacity of several (https://doi.org/10.1007/s11356-018-1812-x) contains supplementary thousand metric tons per year. The major part of these prod- material, which is available to authorized users. ucts is used in an industrial setting, so that release to the en- vironment can occur during production and industrial process- * Christoph Rücker ing. However, there is also some use in consumer products, ruecker@leuphana.de e.g. several thousand tons per year of thioglycolic acid and its 1 salts go into cosmetic and cleaning formulations and are there- Institute of Sustainable and Environmental Chemistry, Leuphana fore released to the environment directly or indirectly via sew- University Lüneburg, Universitätsallee 1, 21335 Lüneburg, Germany 2 age treatment plants. It is therefore of interest to better under- Analytical Chemistry Department, Faculty of Pharmacy, Suez Canal stand the environmental impact of this class of chemicals, both University, Ismailia 41522, Egypt of the parent compounds and of any transformation products. Bruno Bock Thiochemicals, 21436 Marschacht, Germany 18394 Environ Sci Pollut Res (2018) 25:18393–18411 The present study should help to obtain a general view of biodegradation of monofunctional compound classes, the these compounds’ biodegradation behaviour, by comparing problem of biodegradation of difunctional or even more com- literature data with our own experimental data to be generated. plex compounds was not adequately envisaged previously. As For measuring biodegradability, a tiered approach is usually a consequence, some computer models for predicting the bio- adopted in regulations, as described in guidance documents degradability of organic compounds are available, but they (OECD 1995, 2006;ECHA 2017). The first tier tests are are built on information obtained largely from simple screening tests described in the OECD technical guidance se- monofunctional compounds, and their predictive ability for ries 301A-F for ready biodegradability (OECD 1992). These multifunctional compounds is therefore questionable (Rücker tests are very stringent, offering unadapted microorganisms and Kümmerer 2012). For example, models such as the present in low concentration and diversity only limited oppor- Biowin models (US EPA) generally assume additive effects tunities for adaptation to and biodegradation of the chemical of various structural fragments, though non-additive interac- tested (test duration 28 days). They do not simulate real envi- tions of functional groups are to be expected. A goal to be ronmental conditions but give an indication of biodegradability aimed at in the future is prediction of any compound’sbiodeg- of test compounds under various conditions (UN 2015). radation behaviour from its molecular structure, including A positive result in such a test (e.g. oxygen consumption complex structures. This will also allow design of environ- 60% or more of theoretical oxygen demand within 28 days in mentally biodegradable chemicals. All this requires a large tests 301D and 301F) can be considered as indication of rapid body of data for building and validating models, data that and more or less complete degradation under most environ- are largely not available hitherto. In view of all combinations mental conditions including biological sewage treatment of functional groups that may be present in an environmental plants (STPs). Such a test result demonstrates conversion to contaminant, there is a huge amount of work to be done, and 2− + stable inorganic products such as CO ,H O, SO and NH our study on mercaptocarboxylic acids is thus a first step in a 2 2 4 4 or NO (OECD 2006; Guhl and Steber 2006). Chemicals direction that needs to be pursued for several other classes of passing an OECD 301 test are classified as readily biodegrad- difunctional compounds. able, and further testing is not normally required (OECD Understanding the behaviour of difunctional compounds 2006;ECHA 2017). However, continuous release of such a obviously requires as a prerequisite thorough knowledge of substance may cause continuous exposure if degradation is how the single functional groups involved influence the re- slow compared to release, and further testing may therefore spective outcome. It is now well known that some molecular be envisaged (OECD 2006). substructures enhance or hinder biodegradation. These find- In contrast, a negative result in an OECD 301 test does not ings were formulated as “rules of thumb” that are based on necessarily mean that the chemical will not be degraded under more or less evidence. Relevant in our context, many carbox- ylic acids and esters are rather well biodegradable. So the rule relevant environmental conditions, but it can be considered as indication of a potentially persistent chemical and may trigger “carboxylic acid or ester groups enhance biodegradation” be- second tier tests (inherent biodegradability, tests OECD 302A- came popular (Howard 2000; Boethling et al. 2007; Cheng C). If necessary, the highest (third) tier is applied, a simulation et al. 2012). However, it turned out that data supporting this of biodegradation either in the aerobic treatment stage of STPs rule were never explicitly compiled, nor was it specified or in environmental compartments such as fresh or marine against which compound a carboxylic acid or ester was to be surface water (ECHA 2017; Kowalczyk et al. 2015). These compared with respect to biodegradability, the compound higher tier tests require considerably more time, technical bearing a methyl group or the one bearing a H atom in place equipment, manpower and funds. Therefore, data from simu- of the COOH/COOR’ group. lation tests are typically not available for most marketed For the other monofunctional compound class of interest chemicals, and biodegradability or persistence of a chemical here, mercaptans, some biodegradability data are available but in the environment is usually judged based on first or second a rule of thumb was not formulated hitherto. tier tests only. Thus, questions to be addressed in the present study were Acompound’s biodegradability depends on its molecular the following. structure and in particular on the functional groups present. Many compounds of environmental concern are complex, 1. Is the rule of thumb for biodegradability of carboxylic containing more than a single functional group, for example acids/esters well based on experimental data? pesticides, herbicides, dyes and pharmaceuticals. Given the 2. Can a rule of thumb be formulated for biodegradability of increasing number and complexity of chemicals and pharma- mercaptans, based on experimental data? ceuticals ending up in the aquatic environment, their biode- 3. What is the biodegradation behaviour of compounds con- gradability (or not so), decisive for their environmental im- taining both a carboxylic acid/ester and a mercaptan func- pact, will increasingly influence environmental quality in the tional group? Can a general rule be given on the biode- future. While some knowledge has accumulated on the gradability of this class of difunctional compounds? Environ Sci Pollut Res (2018) 25:18393–18411 18395 We therefore first searched the literature for biodegrada- 25359-71-1, DiPETMP) and ethoxylated trimethylolpropane tion of carboxylic acids and esters and corresponding com- tris(3-mercaptopropionate) (CAS-RN 345352-19-4, 674786- pounds without these functional groups, in order to con- 83-5, ETTMP 700) are mixtures of esters obtained from an firm or refute the mentioned rule of thumb. In a second excess of 3-mercaptopropionic acid and the respective multi- step, the literature was searched for biodegradation of mer- functional alcohol (trimethylolpropane, pentaerythritol, captans and corresponding hydrocarbons and analogous dipentaerythritol, ethoxylated trimethylolpropane). Along alcohols, in order to obtain a rule of thumb for biodegra- with the major constituent, they contain partially esterified dation of mercaptans. As to the third question, available products as well as thiolesters. experimental biodegradation data for mercaptocarboxylic Myristyl thioglycolate (CAS-RN 84238-40-4, C TG) is a acids and esters turned out to be partially contradictive, mixture of mostly non-branched C alkyl esters of 10-16 which may be explained by their origin from tests per- thioglycolic acid. formed according to various protocols in various laborato- Glyceryl monothioglycolate (CAS-RN 30618-84-9, GMT) is ries. We therefore decided to measure biodegradation of a reaction mixture of glycerol and thioglycolic acid. Main con- several mercaptocarboxylic acids and esters and, for com- stituents are the isomeric monoesters, minor constituents are di- parison, of several miscellaneous divalent sulfur com- and triesters, as well as free glycerol and thioglycolic acid. pounds (mercaptans, sulfides, disulfides) in two standard The mineral medium used in the experiments was prepared ready biodegradability tests, the Closed Bottle Test (CBT, according to the OECD 301 guidelines from deionised water 301D) and the Manometric Respirometry Test (MRT, (Miele Aqua Purification model G 7795, conductivity ≤ 5 μS/ 301F), using a single source of inoculum. Thereby, we cm) and analytical grade reagent salts in the concentrations expected to obtain more consistent results. In addition to specified there (OECD 1992). Effluent of the municipal the OECD 301D and F methodology, LC-UV/MS/MS was sewage treatment plant (STP) of Lüneburg, Germany employed to identify any transformation products. (144,000 inhabitant equivalents) was collected on the day of test start. The effluent sample was filtered and then used as inoculum directly. The Lüneburg STP treats typical municipal Experimental sewage, there is no industry connected to the STP that deals with the compounds studied here. As to the presence of other Acetonitrile and methanol (HiPerSolv CHROMANORM, organic or toxic substances, controls were run according to the LC-MS grade, BDH Prolabo) and formic acid (analytical guidelines: the quality control results and the inoculum blank grade) were purchased from VWR International GmbH results were as required (validity criteria number 2 and 4 (Darmstadt, Germany). Dithiodiglycolic acid (CAS-RN 505- below). 73-7, DTDGA, TGA disulfide) and 3,3′-dithiodipropionic ac- The Closed Bottle Test (CBT, OECD 301D) is consid- id (CAS-RN 1119-62-6, DTDPA, 3-MPA disulfide) were ob- ered the most stringent among the OECD 301 series ready tained from Sigma-Aldrich (Steinheim, Germany), dimethyl biodegradability (RB) tests (OECD 1992). It works at low 3,3′-dithiodipropionate (CAS-RN 15441-06-2, MMP disul- test compound concentration (theoretical oxygen demand 4 6 fide) from abcr GmbH, Karlsruhe, Germany. 1,2,5- (ThOD) ~ 5 mg/L) and low bacterial density (10 –10 colony Trithiepane (CAS-RN 6576-93-8) was purchased from forming units (CFU)/mL). In our CBT modification, we used as Envilytics, Wiesbaden, Germany. All other tested inoculum two drops of STP effluent per litre of mineral solu- organosulfur compounds were provided by Bruno Bock tion. This inoculum amount was enough for degrading sodium Thiochemicals (Marschacht, Germany), were of technical acetate sufficiently (validity criterion 2, see below), at the same grade and were used without purification. Some of these are time being safe with respect to criterion 4. multiconstituent substances (MCSs): isotridecyl thioglycolate The test comprised four completely filled (no headspace) (CAS-RN 57417-85-3, iC TG), isotridecyl 3- bottles (inoculum blank, quality control, test proper, toxicity mercaptopropionate (CAS-RN 1040871-35-9, iC MP), control, each in duplicate) and was run for 28 days at 20 ± 1 °C isooctyl thioglycolate (CAS-RN 25103-09-7, iOTG) and in the dark (OECD 1992). For details, see Table S1 in the isooctyl 3-mercaptopropionate (CAS-RN 30374-01-7, Electronic supplementary material. iOMP) are mixtures of esters obtained from the respective Oxygenconcentrationinthebottleswas monitoreddai- carboxylic acid (thioglycolic acid or 3-mercaptopropionic ac- ly from outside using the Fibox3 system (fiber-optic id) and mixtures of branched mostly C or C primary alco- oxygen meter with temperature sensor, Precision Sensing 13 8 hols (CAS-RN 68526-86-3 and 68526-83-0, respectively). GmbH, Regensburg, Germany) based on a sensor spot in Trimethylolpropane tris(3-mercaptopropionate) (CAS-RN each bottle (Friedrich et al. 2013). The temperature was 33007-83-9, TMPMP), pentaerythrityl tetrakis(3- monitored daily; the pH was measured at days 0 and 28, mercaptopropionate) (CAS-RN 7575-23-7, PETMP), was adjusted to 6.5–8 if necessary at day 0 and was in this dipentaerythrityl hexakis(3-mercaptopropionate) (CAS-RN range at day 28. 18396 Environ Sci Pollut Res (2018) 25:18393–18411 For a CBT result to be valid, five criteria must be met achieve the pass level within 10 days after the first 10% degra- simultaneously: dation occurred (“10 day window”), except for the 301C test. Sufficiently soluble compounds were dissolved in mineral 1. The difference in degradation between replicate test bot- solution and then dispensed into the bottles, whereas com- tles should be less than 20%. pounds of insufficient solubility were, separately for each bot- 2. The biodegradation of reference compound (sodium ace- tle, weighed using a miniature polystyrene weighing boat, tate) in the quality control has to be at least 60% by day 14. thrown into the bottle and then suspended in mineral solution 3. The biodegradation in the toxicity control should be at (“direct weighing”). In both tests at test begin (day 0) and test least 25% of the total ThOD by day 14. end (day 28), samples were taken (except in cases of direct 4. Oxygen consumption in the inoculum blank should not weighing) and were stored at − 80 °C until analysis. Samples exceed 1.5 mg/L by day 28. were analysed for the test compound (primary elimination) 5. Oxygen concentration in the test bottles should not fall and for transformation products (TPs) by HPLC-UV and below 0.5 mg/L at any time. HPLC-UV-MS/MS (ion trap) whenever possible. In most cases, MRT samples rather than CBT samples were analysed The Manometric Respirometry Test (MRT, OECD 301F) since the higher concentration facilitated identification of TPs. uses a higher test compound concentration (50–100 mg For HPLC-UV, a Prominence HPLC instrument ThOD/L according to the guideline, ~ 30 mg ThOD/L in our (Shimadzu, Duisburg, Germany) was used. Separation was 7 8 ® modification) and higher bacterial density (10 –10 CFU/mL performed on a RP-18 column (NUCLEODUR 100-3 C18 according to the guideline, 80 mL STP effluent/L final solu- ec, 2 mm ID, 125 mm, Macherey & Nagel, Düren, Germany) tion in our work) and thus higher bacterial diversity (OECD protected by an EC guard column (NUCLEODUR 100-3 1992). Another difference between CBT and MRT is the im- C18 ec, 4 × 2 mm). Elution was isocratic or gradient using plementation of a further control bottle (sterile control) and the mixtures of 0.1% formic acid in water (solution A) and use of only one bottle for toxicity control. For details, see 100% acetonitrile (solution B). The flow rate was 0.25 mL/ Table S1 in the Electronic supplementary material. The test min, the oven temperature 40 °C and the detector wavelength was run for 28 days at 20 ± 1 °C in the dark with gentle stir- 210 nm. Twenty to fifty microlitres of each sample were ring. Oxygen consumption was recorded daily using an injected without any workup. OxiTop control OC110 system (WTW, Weilheim, HPLC-UV-MS was performed on an Agilent Technologies Germany), measuring the pressure decrease in the headspace series 1100 HPLC instrument (Agilent Technologies, (about one third of the bottle volume), while CO produced Böblingen, Germany). Column, eluent solutions and run pa- was removed by adsorption/absorption to NaOH pellets/conc. rameters were as above. To the chromatograph, a Bruker NaOH solution. The temperature was monitored daily, the pH Daltonic Esquire 6000 plus ion-trap mass spectrometer was was measured at days 0 and 28, was adjusted to 6.5–8if coupled, equipped with an atmospheric pressure electrospray necessary at day 0 and was in this range at day 28. ionisation (ESI) interface and a Bruker data analysis system For a MRT result to be valid, five criteria must be met (Bruker Daltonic GmbH, Bremen, Germany). The mass spec- simultaneously: trometer was operated in the positive mode. For analytical details, see the “Electronic supplementary material”. 1.–3. The first three criteria are the same as in CBT. 4. The oxygen consumption in the inoculum blank should be at most 60 mg/L by day 28. Results and discussion 5. If the oxygen consumption by the test substance at test end is < 60%, then the pH must be within the range 6–8.5. Biodegradation data from the literature The pass level in both CBT and MRT is oxygen consump- Tables 1 and 2 show biodegradation test results published in tion at day 28 of 60% of the theoretical oxygen demand. ThOD the scientific and regulatory literature. The purpose of Tables 1 was calculated under the assumption that sulfur is oxidised to and 2 is to provide a semi-quantitative overview of the com- 2− SO , according to the guideline (OECD 1992). To be consid- pounds’ biodegradability as a function of specific functional ered readily biodegradable, a substance is further required to groups present in the molecules. These data come from differ- ent tests and therefore are not strictly comparable, they should not be used indiscriminately together, e.g. for training of a The 20% difference is usually interpreted as a 20 %points difference, since the unit of biodegradation is % (of a theoretical value). For example, if two biodegradation QSAR model. replicate bottles show 41 and 59% biodegradation, the difference of 18 Unless stated otherwise, results in Tables 1 and 2 were %points is considered still acceptable, though 59 and 41 (in any unit) differ obtained in the 28-day (or 14-day) OECD 301C (MITI-I) test, by ~ 44 or ~ 30.5%, based on the lower or the higher value, respectively, which would be unacceptable when taking the criterion literally. as contained in the J-CHECK database. Results of the 14-day Environ Sci Pollut Res (2018) 25:18393–18411 18397 Table 1 Biodegradation test results (oxygen consumption as % of the theoretical oxygen demand) of carboxylic acids, esters and corresponding compounds without these functional groups, taken from the literature. Data are primarily from the 28-day (or 14-day) OECD 301C (MITI-I) test Biodegradation [%] R-COOH name R R-COOH R-COOCH R-COOC H R-H R-CH 3 2 5 3 Acetic acid ≥74 ≥92 ≥94 n.a. n.a. b a c Chloroacetic acid ≥65 45 75 1 1 d,e 2-Chloropropionic acid 77 RB n.a. 1 0 Dichloroacetic acid ≥97 n.a. n.a. 13 NRB Trichloroacetic acid 7 n.a 67 ≥0 ≥0 Acrylic acid ≥67.8 ≥37 ≥51.5 n.a. 1 Methacrylic acid ≥91 ≥94.3 79.1 1 n.a. Cyanoacetic acid ≥90 99 68 65 n.a. 4-Isopropylbenzoic acid ≥89 n.a. n.a. ≥33 ≥88 2-Chlorobenzoic acid ≥5.6 n.a. n.a. 0 0 2-Methylaminobenzoic acid ≥85 n.a n.a. ≥1.4 1 n.a. not available, The corresponding carboxylic acid was formed to some extent during the test b c d e Test duration 21 days, OECD 301F, ECHA database, Test unspecified, Cheng et al. (2012), Test unspecified, Pizzo et al. (2013) f g h i OECD 301D, ECHA database, OECD 301A, ECHA database, OECD 301E, ECHA database, OECD 301B, ECHA database 18398 Environ Sci Pollut Res (2018) 25:18393–18411 Table 2 Biodegradation results (oxygen consumption as % of the theoretical oxygen demand) of mercaptans and corresponding hydrocarbons and alcohols, taken from the literature. Data are primarily from the 28-day (or 14-day) OECD 301C (MITI-I) test Biodegradation [%] R-SH name R R-SH R-H R-OH Ethyl mercaptan 0 n.a. ≥89 39(301D) 77(301F) 79(301D) n-Dodecyl mercaptan ECHA ECHA ECHA 43-96 (301B) ~74 (306) ECHA; n-Octadecyl mercaptan 0 ECHA 38-69 (301D) ECHA 2-Mercaptoethanol 19 ≥89 ≥90 Thiophenol 0 ≥40 ≥85 b c NRB 2≥0; 96 ; 1-Naphthalenethiol d,e b RB ; fast 2-Naphthalenethiol 0 2≥68 n.a. not available The corresponding disulfide was formed to some extent during the test Test unspecified, Biowin1/2 training set Test unspecified, Pizzo et al. (2013) Test unspecified, Cheng et al. (2012) Biowin5/6 training set (MITI-I model) test are given as “≥xx”, where xx is the numerical result mea- corresponding methyl compound and the compound lacking sured after 14 days. In cases of no result found in the J- that methyl group. Each row of Table 1 contains data for (at CHECK database,the REAXYS and ECHA databases were least) a pair of compounds, so that comparison within a row consulted, then training and test datasets of the Biowin1/2 and shows the impact of a COOH/COOCH /COOC H group on 3 2 5 Biowin5/6 models (US EPA EPI Suite), and as a last resort biodegradability in context R. Comparison between rows then experimental data cited in recent papers on biodegradation shows whether the effect is same or different in different struc- modelling (Cheng et al. 2012; Pizzo et al. 2013). The latter tural contexts as described by the variation of R. sources give results as readily biodegradable/not readily bio- As an exception, we included acetic acid despite the lack of degradable (RB/NRB) only. data for methane and ethane, for the relevance of sodium acetate as the reference compound in biodegradability tests. Carboxylic acids and their esters Examination of Table 1 shows that a carboxylic acid or its methyl or ethyl ester is more easily biodegradable than In Table 1, biodegradation test results are compiled for car- corresponding compounds bearing a CH or H at the same boxylic acids and their methyl or ethyl esters and for the position, and this seems to be generally true. For the pair 4- Environ Sci Pollut Res (2018) 25:18393–18411 18399 methylaminobenzoic acid and N-methylaniline, seemingly a Experimental biodegradation data from the present counterexample, the difference between 0 and 1.4% is within study experimental uncertainty. Thus, the rule of thumb for carbox- ylic acids/esters seems to be well based on facts. We tested biodegradation of 24 substances containing divalent No general trend in biodegradability is seen in Table 1 on sulfur, i.e. mercaptocarboxylic acids, their esters, disulfides, comparing the esters to the carboxylic acids. sulfides and mercaptans. As a first experimental result of the Partial hydrolysis of esters to alcohols and acids is to be present study, for all tested compounds in all biodegradation expected in aqueous solution even at neutral pH and 20 °C experiments no indication of toxicity to the bacteria was found during a reaction time extended as long as 28 days and was in at the tested substance concentrations. fact observed for several esters in the OECD 301C test. Table 3 presents the biodegradation test results obtained in Examples are ethyl acetate and methyl chloroacetate in this study. In case of an invalid result, the test was repeated Table 1, and propyl acetate, sec-butyl acetate, n-butyl acrylate, until a valid result was obtained. For a few compounds, a test methyl 4,4-dimethyl-3-oxopentanoate, acetylsalicylic acid (J- was repeated despite of a valid first result already obtained, in CHECK database). Hydrolysis is a step neutral with respect to order to examine intra-laboratory reproducibility. For all re- oxidation, it does not show up in the oxygen balance but may sults ≥ 60%, the 10-day window was passed. influence the biodegradation rate by producing more water- Tested substances are subdivided in Table 3 into three soluble and bioavailable species. groups. Group 1 contains thioglycolic acid (TGA) and its derivatives (ammonium salt, disulfide and esters). Group 2 Mercaptans comprises 3-mercaptopropionic acid (3-MPA) and its de- rivatives (disulfides and esters). Group 3 contains miscel- Table 2 presents biodegradation test results of mercaptans and laneous compounds of divalent sulfur, e.g. thiolactic acid the corresponding hydrocarbons and alcohols for comparison. (TLA), thioethers (MBT, Di-2-EHTDG) and mercaptans Each row of Table 2 contains data for (at least) a pair of (DMDS, DMPT). compounds, so that comparison within a row shows the im- pact of a SH (or OH) group on biodegradability in structural Discussion of reproducibility context R. Comparison between rows then shows whether the effect is same or different in different structural contexts as Generally, reproducibility is an issue in all biodegradability described by the variation of R. test results. A closer look on our data reveals that poor repro- The limited data in Table 2 show that a mercapto group is ducibility appears on various levels, as detailed in the follow- detrimental to biodegradation compared with the correspond- ing subsections. ing hydrocarbon, in contrast to a hydroxyl group that en- hances biodegradation. This trend is seen in all rows of Intra-experiment reproducibility Table 2 and so may be considered a rule of thumb. A mercapto group, however, does not absolutely pro- Table 4 gathers discrepancies between duplicate bottles within hibit biodegradation, as demonstrated by n-dodecyl mer- an experiment from Table 3 for CBT and MRT, i.e. all those captan and 2-mercaptoethanol. This latter result may be results that are invalid for violation of criterion 1. interpreted as being caused by the presence of structural Discrepancies between duplicate bottles in the same exper- elements both favouring (hydrocarbon chain of intermedi- iment were occasionally observed earlier (Painter and King ate length, OH group) and hindering biodegradation (SH). 1985), the guidelines explain this phenomenon by differences The data for 1-naphthol demonstrate that published exper- in the composition of the inoculum (ECHA 2017,page 214; imental biodegradation test results are generally to be consid- OECD 1992). Nevertheless, some authors consider such re- ered with caution: the 14-day MITI-I test resulted in 0% bio- sults not only “invalid”, but also useless, meaning that “some- degradation, whereas the training set of models Biowin5/6 thing went wrong with the experiment”. The experiment is claims a MITI-I test result of > 60% (“readily biodegradable”), then repeated until a valid result is obtained, and invalid as do Cheng et al. (2012) for an unspecified test. Pizzo results are discarded rather than reported. This practice et al. (2013) even report 96% biodegradation in an unspec- leads to the wrong impression that biodegradation experi- ified test, and the Biowin1/2 training set claims “fast bio- ments are as reproducible as physicochemical measure- degradation” for this compound. The Biowin1/2 data of 1- ments. To correct this view, we here report our invalid naphthol is from the Environmental Fate Database of experiments also and offer alternative explanations. Syracuse Research Corporation, which unfortunately is Moreover, such invalid experimental results are by no no longer publicly available. Further, the four “high” means useless, in particular if both parallel bottles exceed claims may result from tapping thesame sourceratherthan the pass level, as is the case for TGA, DTDGA, 3-MPA coming from four independent experiments. disulfide and MMP disulfide in MRT (Table 4). 18400 Environ Sci Pollut Res (2018) 25:18393–18411 Table 3 Biodegradation test results obtained in the present study. For each experiment, the % O consumption in the two parallel bottles is given, and for valid results the average ±SD, otherwise the criterion rendering a result invalid. For structures, see Table 7 Compound name (abbreviation) Experiment number OECD 301D (CBT) OECD 301F (MRT) [CAS-RN] O consumption/ Average in a valid O consumption/ Average in a valid 2 2 ThOD [%] experiment ±SD ThOD [%] experiment ±SD TGA and derivatives Thioglycolic acid (TGA) 1 31.8, 45.3 38.5 ± 9.5 114.4, 92.1 Not valid, [68-11-1] criterion 1 2 49.2, 51.3 50.2 ± 1.5 78.8, 94.4 86.6 ± 11.1 Diammonium dithiodiglycolate (DADTDG, 1 123.8, 106.4 115.1 ± 12.3 130.0, 132.0 131.0 ± 1.4 TGA disulfide ammonium salt) calculated without [68223-93-8] nitrification 1 76.6, 65.9 71.3 ± 7.6 80.5, 81.7 81.1 ± 0.9 re-calculated with nitrification 2 14.9, 29.2 22.0 ± 10.1 79.4, 90.8 85.1 ± 8.0 calc. with nitrification 3 90.9, 86.0 88.5 ± 3.5 calc. with nitrification Dithiodiglycolic acid (DTDGA, 1 65.9, 62.2 64.1 ± 2.6 87.9, 111.2 Not valid, TGA disulfide) criterion 1 [505-73-7] 2 60.3, 55.7 58.0 ± 3.3 2-Ethylhexyl thioglycolate (2-EHTG) 1* 58.3, 61.2 59.7 ± 2.0 28.2, 54.7 Not valid, [7659-86-1] criterion 1 2* 18.4, 13.9 16.2 ± 3.2 Isotridecyl thioglycolate (iC TG) 1** 1.4, 10.0 Not valid, 40.9, 22.5 31.7 ± 13.0 [57417-85-3] criterion 4 2** 14.6, 9.5 12.0 ± 3.6 Glycol di(mercaptoacetate) (GDMA) 1 62.2 Not valid, 62.1, 69.8 65.9 ± 5.4 [123-81-9] criterion 1 2 82.5, 72.4 Not valid, criterion 4 3 70.0 Not valid, criterion 1 4 79.4, 77.1 78.3 ± 1.6 Glyceryl monothioglycolate (GMT) 1 56.0, 50.7 53.4 ± 3.7 79.8, 80.8 80.3 ± 0.7 [30618-84-9] 3-MPA and derivatives 3-Mercaptopropionic acid (3-MPA) 1 0.4, 32.4 Not valid, 78.7, 87.0 82.8 ± 5.9 [107-96-0] criterion 1 2 27.4, − 7.3 Not valid, 101.4, 90.1 95.8 ± 8.0 criteria 1 and 4 3 0.6, 3.8 2.2 ± 2.3 Dithiodipropionic acid 1 3.9, 1.5 2.7 ± 1.7 121.6, 87.9 Not valid, (DTDPA, 3-MPA disulfide) criterion 1 [1119-62-6] 2 93.7, 100.3 97.0 ± 4.7 Methyl 3-mercaptopropionate (MMP) 1 65.0, 61.7 63.4 ± 2.3 23.3, 12.0 17.7 ± 8.0 [2935-90-2] 2 16.0, 21.0 18.5 ± 3.6 3 74.1, 14.1 Not valid, criterion 1 4 5.4, 12.3 8.8 ± 4.9 Propanoic acid, 3,3′-dithiobis-, 1 6.6, 4.1 5.4 ± 1.8 61.6, 84.2 Not valid, dimethyl ester (MMP disulfide) criterion 1 [15441-06-2] 2 10.6, 13.0 11.8 ± 1.7 14.0, 21.7 17.8 ± 5.4 3* 21.6, 1.3 Not valid, criterion 1 Environ Sci Pollut Res (2018) 25:18393–18411 18401 Table 3 (continued) Compound name (abbreviation) Experiment number OECD 301D (CBT) OECD 301F (MRT) [CAS-RN] O consumption/ Average in a valid O consumption/ Average in a valid 2 2 ThOD [%] experiment ±SD ThOD [%] experiment ±SD Butyl 3-mercaptopropionate (BuMP) 1 55.6, 29.0 Not valid, 15.0, 20.7 17.8 ± 4.0 [16215-21-7] criterion 1 2 57.4, 60.3 Not valid, criterion 4 3 61.9, 62.5 62.2 ± 0.5 4 40.8, 41.1 41.0 ± 0.2 2-Ethylhexyl 3-mercaptopropionate 1* 54.3, 49.8 52.1 ± 3.2 56.4, 27.0 Not valid, (2-EHMP) criterion 1 [50448-95-8] 2* 42.4, 51.6 47.0 ± 6.5 Isooctyl 3-mercaptopropionate (iOMP) 1** 31.1, 32.1 31.6 ± 0.8 − 8.2, 4.6 − 1.8 ± 9.0 [30374-01-7] Isotridecyl 3-mercaptopropionate (iC MP) 1** 10.5, 17.6 14.1 ± 5.0 29.5, − 11.1 Not valid, [1040871-35-9] criterion 1 2** 34.2, 35.8 35.0 ± 1.2 Glycol di(3-mercaptopropionate) (GDMP)12.1, − 0.8 0.7 ± 2.0 73.4, 39.8 Not valid, [22504-50-3] criterion 1 2 10.4, 13.6 12.0 ± 2.3 32.8, 50.4 41.6 ± 12.5 3* 33.4, 41.0 37.2 ± 5.3 Trimethylolpropane tris- 1** 8.2, 9.9 9.1 ± 1.2 32.9, 55.4 Not valid, (3-mercaptopropionate) (TMPMP) criterion 1 [33007-83-9] 2** 4.7, 1.7 3.2 ± 2.1 Dipentaerythrityl hexakis- 1** 1.4, 3.3 2.4 ± 1.4 14.8, 42.4 Not valid, (3-mercaptopropionate) (DiPETMP) criterion 1 [25359-71-1] 2** 12.0, − 5.5 3.2 ± 12.3 Ethoxylated trimethylolpropane tris- 1** 3.1, − 3.2 Not valid, 35.1, 40.0 37.6 ± 3.5 (3-mercaptopropionate) (ETTMP 700) criterion 4 [674786-83-5 or 345352-19-4] 2** 9.2, 8.6 8.9 ± 0.5 Miscellaneous compounds containing divalent sulfur Thiolactic acid (TLA) 1 11.9, 12.8 12.4 ± 0.6 66.6, 75.9 71.2 ± 6.6 [79-42-5] Di(2-ethylhexyl) thiodiglycolate 1** 19.3, 14.0 Not valid, 41.5, 34.3 37.9 ± 5.1 (Di-2-EHTDG) criterion 4 [24293-43-4] 2** 22.7, 33.8 28.3 ± 7.9 Methylene bis(butyl thioglycolate) (MBT) 1* 49.4, 48.5 48.9 ± 0.7 56.6, 63.1 59.8 ± 4.6 [14338-82-0] 2* 52.2, 63.2 57.7 ± 7.8 Bis(2-mercaptoethyl) sulfide (DMDS) 1 − 0.2, − 2.0 − 1.1 ± 1.3 35.8, − 12.9 Not valid, [3570-55-6] criterion 1 2* 1.8, 33.4 Not valid, criterion 1 4-Mercaptomethyl-3,6-dithia-1,8- 1 − 5.0, − 2.7 − 3.9 ± 1.6 octanedithiol (DMPT) [131538-00-6] Multiconstituent substance (MCS), for details, see “Experimental” Result from one test bottle only, the other one became leaky during the experiment *Due to poor solubility, this substance was directly weighed in MRT **Due to very poor solubility, this substance was directly weighed both in CBT and in MRT 18402 Environ Sci Pollut Res (2018) 25:18393–18411 Table 4 Non-reproducibility within the same experiment may contain different numbers of cells of different diversity just by chance. CBT biodegradation [%] MRT biodegradation [%] 3-MPA 0, 32; − 7, 27 TGA 92, 114 MMP 14, 74 DTDGA 88, 111 BuMP 29, 56 2-EHTG 28, 55 Intra-laboratory reproducibility 3-MPA disulfide 88, 122 Few scientists perform a second or third biodegradation ex- MMP disulfide 1, 22; 62, 84 periment if a valid result is already at hand. If such double 2-EHMP 27, 56 measurements are nevertheless done and produce conflicting iC MP − 11, 30 results, each being valid, then such data likely will not be GDMP 40, 73 published. Therefore, this kind of problem also is often not TMPMP 33, 55 apparent, it is not explicitly mentioned in the guidelines, and DiPETMP 15, 42 its prevalence is unknown. DMDS − 13, 36; 2, 33 Table 5 gathers discrepancies from Table 3 between differ- ent experiments performed for the same compound under the same test protocol in the present study, from CBT and MRT. In our experiments, MRT results seem to be more repro- At least some of the discrepancies observed in the ducible than CBT results (Table 5). The CBT results for three present experiments may be explained by experimental out of six compounds led to conflicting assessments (above problems: and below the pass level), while in MRT this happened for one out of four compounds only. This difference can be explained – Poorly soluble substances (2-EHMP, 2-EHTG, iC MP, by the higher bacterial density/diversity in MRT compared to TMPMP, DiPETMP, MMP disulfide in its experiment 3) CBT. were directly weighed into the test bottles, suffering from Discrepancies between two seemingly identical biodeg- difficulties in weighing mg amounts of viscous material radation experiments performed at different times may be into several parallel bottles. due to different lag times. Lag time is the time required for – GDMP (in MRT experiment 1) and MMP disulfide (in adaptation by the microbial population to metabolise the MRT experiment 1) were tested applying the usual substrate efficiently. Different lag times can be caused by substance preparation as a solution despite borderline a variation in the diversity of the bacterial population in two solubility. Incomplete solution then may have resulted experiments, that is, by presence or absence of competent in different amounts of test substance in the two par- bacteria. In a former study, bacterial strains were isolated allel bottles. from activated sludge that not only tolerate organosulfur – In two CBT experiments with GDMA, one of the two carboxylic acid esters, but even were able to use them as parallel bottles became leaky during the test, as mani- sole carbon and energy source (Toups et al. 2010). Such fested by the O concentration in one bottle unexpect- specialist strains are not necessarily those that biodegrade edly increasing after some days of decrease. These these compounds in aerobic real environments or activated bottles were therefore excluded from further consider- sludge, since they may be outcompeted by generalists. ation. However, a miniature leak will have a less ob- However, if sufficient amounts of these compounds are vious effect, the oxygen concentration decreasing available as sole carbon source as in an OECD 301 merely more slowly than in a tight bottle, so that the apparent oxygen consumption in that bottle will lag behind. This is what we observed repeatedly in this Table 5 Reproducibility of test results, valid results only, i.e. each study. Thereby a final difference in apparent biodegra- number is the mean from two bottles differing by no more than 20% dation of > 20% may or may not build up between CBT biodegradation [%] MRT biodegradation [%] parallel bottles. Thus, a possible tiny leak is a suffi- cient explanation for any discrepancies between paral- DADTDG 22, 71, 89 DADTDG 81, 85 lel bottles at day 28. GDMP 1, 12 GDMP 37, 42 – Finally, bacterial density and diversity may differ in TGA 39, 50 3-MPA 83, 96 parallel bottles following from uneven dispersion of MMP 9, 18, 63 MBT 58, 60 bacterial species in the mineral solution. If, e.g. com- MMP disulfide 5, 12 petent bacteria form aggregates, then two samples BuMP 41, 62 (same volume) taken from a highly dilute suspension Environ Sci Pollut Res (2018) 25:18393–18411 18403 Fig. 1 Biodegradation curves of DADTDG in two different CBT experiments. a Experiment 3, lag phase 14 days. b Experiment 2, lag phase 27 days 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 -20 Days Test substance Quality control Toxicity control (measured value) Toxicity control (calculated value) 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 -20 Days Test substance Quality control Toxicity control (measured value) Toxicity control (calculated value) experiment, such specialist strains tend to increase due to treatment plant (STP) was found to show considerable tem- induction of appropriate catabolic enzymes. poral (seasonal) variation (Kim et al. 2013;Juet al. 2014). If the sigmoid biodegradation curve is shifted more or In our case, the inoculum for each experiment was freshly less along the time axis, it hits the fixed 28-day terminus taken from the same source, effluent of the Lüneburg mu- at different points on the biodegradation axis. A shift of nicipal STP, but the delay between subsequent experiments a few days can make the difference between very low with the same substance was between 2 weeks and and very high biodegradation. As an example, consider 9 months, enough time for seasonal variation of the bacte- the biodegradation curves of DADTDG, the ammonium ria population in the STP. salt of TGA disulfide (Fig. 1). DADTDG was readily biodegraded (two valid CBT results, 71 and 89%, and Variation between test results for the same compound two valid MRT results, 81 and 85%), but one valid CBT under different OECD 301 test protocols result of 22% biodegradation was also obtained (exper- iment 2). Unusually long and variable lag phases were For many substances, Table 3 shows strong differences observed in CBT (19 days in experiment 1, resulting in between CBT (OECD 301D) and MRT (OECD 301F) re- 71% biodegradation; 27 days in experiment 2, 22%; sults. Though CBT is meant to be the most stringent test, in 14 days in experiment 3, 89%). On the other hand, in some instances, degradation in CBT was higher than in MRT lag times were in good concordance (9 and MRT. Such a phenomenon has been known for many 10 days), as were the degrees of biodegradation. years, and recent comparisons between 301C and 301X Bacterial populations taken from the same source at dif- test results based on a large experimental basis revealed ferent times are not alike, but are subjected, e.g., to season- that for some compounds, 301C was more efficient, for al variations (Brillet et al. 2016). In particular, the amount others, 301X for each X from {A, B, D, E, F} of rare microbial specialists in activated sludge of a sewage (Kayashima et al. 2014). Biodegradation % Biodegradation % 18404 Environ Sci Pollut Res (2018) 25:18393–18411 Table 6 Results for structurally Chemical class Compound CBT biodegradation [%] MRT biodegradation [%] similar compounds in the same test (valid results only) Mercaptocarboxylic acids 3-MPA 2 83, 96 TGA 39, 50 87 TLA (2-MPA) 12 71 Dithiodicarboxylic acids DTDPA 3 97 DTDGA 64 58 Ethylhexyl esters 2-EHMP 52 47 2-EHTG 60 16 Glycol esters GDMP 1, 12 37, 42 GDMA 78 66 Isotridecyl esters iC MP 14 35 iC TG 12 32 Mercaptocarboxylic acids/esters 3-MPA 2 83, 96 and corresponding disulfides 3-MPA disulfide 3 97 MMP 9, 18, 63 18 MMP disulfide 5, 12 18 TGA 39, 50 87 DTDGA 64 58 Differences between structurally similar compounds Synopsis of biodegradation data in the same test Table 7 is a synopsis of biodegradation test results for Ideally, it is expected that structurally similar compounds mercaptocarboxylic acids and esters and miscellaneous diva- should behave similarly under the same 301 test protocol. lent sulfur compounds. Along with the 24 substances from Table 6 compiles the experimental data (from Table 3)of Table 3, Table 7 contains seven further compounds with avail- structurally similar compounds in the same test. able biodegradation results that were not included in the pres- As Table 6 shows, some pairs of similar compounds more or ent experimental study. less meet the expectation (3-MPA, TGA, TLA; iC MP, Column 3 shows biodegradation data found in the sci- iC TG; 3-MPA, 3-MPA disulfide; MMP, MMP disulfide), entific or regulatory literature (e.g. ECHA database or un- while others do not. In case of a mercaptan and its disulfide, published original study reports). In order to use only reli- concordance is explained by a test of the former being essen- able literature data, the quality of study results was tially a test of the latter, see “Primary elimination and transfor- assessed on a case-by-case basis using the Klimisch scor- mation products”.Table 6 suggests not to overestimate both ing system (Klimisch et al. 1997). Studies with a Klimisch concordant and divergent behaviour of similar compounds in score 1 (reliable without restriction) or 2 (reliable with single biodegradation experiments, given all the experimental restriction) only were considered for column 3. For exam- imponderability discussed above. It is probably naive to expect ple, among many results available in the literature for TGA similar compounds to behave similarly in a single biodegrada- (ranging from 0 to 100%), the reliable value of 67% tion experiment when even repeated experiments on the same (OECD 301D) was chosen from a Klimisch 2 study (van compound can lead to contradictions. We have to realise that a Ginkel and Stroo 1992). biodegradation test result for a compound is only one of several Column 4 shows the highest experimental result for each possibilities, in other words that we should consider biodegra- substance obtained in the present OECD 301D or 301F stud- dation experiments as statistical processes in the sense that they ies (Klimisch 2), taken from Table 3. are governed by some factors difficult to control. Similarity in The last column of Table 7 reports a consolidated assess- the behaviour of similar compounds can then be expected only ment of all available ready biodegradability test results for on the basis of several or even many experiments. Even then, each substance. This overall result is given as readily biode- the often high sensitivity of enzymes to small structural differ- gradable (yes) or not readily biodegradable (no). The proce- ences of potential substrates may lead to unexpected effects dure leading to column 7 entries is based on the relevant within a series of related structures. guidelines saying Environ Sci Pollut Res (2018) 25:18393–18411 18405 Table 7 Synopsis of biodegradation results from the literature (ECHA or BB, i.e. ECHA database or original study reports provided by Bruno Bock) and from the present study Result Result Substance name from liter- Readily from pre- (abbreviation) ature biode- Structure sent study [CAS-RN unless (OECD gradable, (OECD given in Table 3] 301A-F), yes or no 301D or F) source TGA and derivatives Thioglycolic acid 67% 87% (301F) yes (TGA) (301D) Dithiodiglycolic no data 64% (301D) yes acid (DTDGA) found Diammonium 80% (301B) dithiodiglycolate 89% (301D) yes ECHA (DADTDG) 2-Ethylhexyl 82% (301F) thioglycolate 60% (301D) yes ECHA (2-EHTG) Isooctyl thioglyco- 76% (301F) not included late (iOTG) yes ECHA in this work [25103-09-7] Isotridecyl no data thioglycolate 32% (301F) no found (iC TG) Myristyl thioglyco- 82% (301B) not included late (C TG) yes BB in this work [84238-40-4] 18406 Environ Sci Pollut Res (2018) 25:18393–18411 Glycol 70% (301B) di(mercaptoacetate) 78% (301D) yes BB (GDMA) Glyceryl 46% 80% monothioglycolate (301D) yes (301F) (GMT) ECHA 3-MPA and derivatives 96% 3-Mercaptopropi- (301A) 96% (301F) yes onic acid (3-MPA) ECHA Dithiodipropionic no data 97% (301F) yes acid (DTDPA) found Thiodipropionic 93% not included acid (TDPA) (301C) yes in this work [111-17-1] ECHA Methyl 3-mercapto- 46% (301B) 63% (301D) yes propionate (MMP) ECHA Propanoic acid, 3,3'-dithiobis-, no data 18% (301F) no dimethyl ester found (MMP disulfide) Butyl 3-mercapto- no data 62% (301D) yes propionate (BuMP) found 2-Ethylhexyl 3- 70% (301B) mercaptopropionate 52% (301D) yes BB (2-EHMP) Isooctyl 3-mercap- 55% (301B) topropionate 32% (301D) no ECHA (iOMP) Environ Sci Pollut Res (2018) 25:18393–18411 18407 Isotridecyl 3- 17% (301B) mercaptopropionate 35% (301F) no ECHA (iC MP) Glycol di(3-mer- 72% (301B) captopropionate) 42% (301F) yes BB (GDMP) Trimethylolpropane tris(3-mercapto- no data 9% (301D) no propionate) found (TMPMP) Pentaerythrityl tetrakis(3-mercap- 26% (301B) not included topropionate) no ECHA in this work (PETMP) [7575-23-7] Dipentaerythrityl hexakis(3-mercap- 22% (301F) 3% (301F) no topropionate) BB (DiPETMP) Ethoxylated trime- thylolpropane no data tris(3-mercapto- 38% (301F) no found propionate) (ETTMP700) Miscellaneous compounds containing divalent sulfur 38% Thiolactic acid (301A) 71% (301F) yes (TLA) BB 18408 Environ Sci Pollut Res (2018) 25:18393–18411 Di(2-ethylhexyl) dithiodiglycolate 54% (301B) not included no (Di-2-EHDTDG) BB in this work [62268-47-7] Dilauryl thiodi- propionate 82% (301C) not included yes (DLTDP) ECHA in this work [123-28-4] Distearyl 71% thiodipropionate not included (301D) yes (DSTDP) in this work ECHA [693-36-7] Di(2-ethylhexyl) 77% (301B) thiodiglycolate 38% (301F) yes ECHA (Di-2-EHTDG) Methylenebis(butyl 25% 60% thioglycolate) (301D) yes (301F) (MBT) ECHA Bis(2-mercapto- no data ethyl)sulfide 0% (301D) no found (DMDS) 4-Mercaptomethyl- 3,6-dithia-1,8- 3% (301C) 0% (301D) no octanedithiol ECHA (DMPT) Multiconstituent substance (MCS), for details see “Experimental” However, in one MRT experiment of MMP disulfide, both bottles reached the pass level of 60% (62 and 84%), thus failing with respect to criterion 1 Mean value of 56.6 and 63.1% is 59.85%, just below 60% unless rounded “Realising that ready biodegradability tests may some- Results in Table 7 can besummarisedasfollows: time fail because of the stringent test conditions, positive test results should generally supersede negative test re- 1. Simple mercaptocarboxylic acids such as TGA, 3- sults” (ECHA 2017,page 208;OECD 2006, page 3), and MPA and TLA and their simple esters are readily “When contradictory results in ready biodegradabili- biodegradable. ty tests are obtained the positive results could be con- 2. Those esters of the same acids that are not readily biode- sidered valid irrespective of negative results, when gradable are at the same time structurally complex (esters the scientific quality of the former is good and the of branched higher or multifunctional alcohols) and positive test results are well documented, …” (ECHA multiconstituent substances. They nevertheless undergo 2017, page 230). considerable biodegradation in OECD 301 tests, so there is no reason to consider them persistent in the environment. We are aware that these formulations seem to be open to 3. A disulfide motif does not prevent good biodegradability misuse by multiple testing and selective reporting, and this (DTDGA, DADTDG, DTDPA, MMP disulfide, Di-2- is another reason for showing in Table 3 all our experimen- EHDTDG), nor does a sulfide (thioether) (TDPA, tal results. DLTDP, DSTDP, Di-2-EHTDG, MBT). Environ Sci Pollut Res (2018) 25:18393–18411 18409 4. Di- or polymercaptans (without a favourable substructure) 2006)or “… can be used as evidence for inherent bio- are not biodegradable (DMDS, DMPT) in OECD 301D degradability” (ECHA 2017). or 301F tests. By these lines of argumentation, the mercaptoesters tested These observations are in good agreement with the results here, even those not formally readily biodegradable, can pre- shown in Table 2 for mercaptans. That is, the detrimental liminarily be considered inherently biodegradable. influence of a mercapto group can be counterbalanced by a Interestingly, 3-MPA was enzymatically oxidised to the favourable substructure such as an acid or ester group. corresponding sulfinic acid by a special bacterial mutant For mercaptan and sulfide functional groups, this is just the (Bruland et al. 2009) and seems to be an intermediate in the opposite of what is known for their oxygen analogues, hy- metabolism of organic sulfur compounds (methionine, homo- droxyl groups enhance biodegradation while ether groups cysteine) in natural environments (Salgado et al. 2015; are detrimental (Boethling et al. 2007). references cited therein). The higher mercaptoesters that proved not readily biode- Recently, a multilinear model for description and predic- gradable have in common a rather high molecular weight as- tion of biodegradation for general chemicals was proposed sociated with low solubility in water, and some bear branched that uses additive functional group increments and is based higher alkyl groups. Both these characteristics are known to on biodegradation data from several tests. Among all func- be detrimental to biodegradation. Others present at the mole- tional groups considered there, the mercaptan group turned cule’s surface the mercapto groups only, i.e. they can be sub- out the strongest hindering biodegradation, while carboxylic sumed under item 4, not providing a convenient point of at- acid and ester groups were found the most favourable tack for bacteria. Further, all these substances have in common (Vorberg and Tetko 2014). Our results are completely in line their MCS property. For MCSs such as petroleum products or with this picture. mixtures of homologous compounds, e.g. technical surfac- tants, the guidelines acknowledge the onset of biodegradation Primary elimination and transformation products to be delayed and/or the biodegradation curve to be less steep in comparison with the single constituents. Accordingly, the Along with biodegradation by microorganisms, abiotic pro- 10-day window is not applied for such substances (ECHA cesses may occur in the environment or in biodegradation 2017,page 210; OECD 2006, paragraphs 43 and 44). By this tests, e.g. photodegradation, hydrolysis, abiotic oxidation. logic, we expect homologous mixtures also to achieve less Thus, aerobic disulfide formation from mercaptans is well biodegradation than the single constituents after 28 days. known both during OECD 301 tests (footnote a in Table 2) The bacterial adaptation to many isomers, homologues or and by oxidation in air-saturated tap water (TGA, OECD byproducts in low concentration may be slower than adapta- 2009;3-MPA, unpublished, BB). Likewise, hydrolysis of es- tion to a single chemical in higher concentration, since differ- ters during biodegradation tests is known (“Carboxylic acids ent bacterial strains or different degradative enzymes may be and their esters”). With compound 2-EHTG, both these abiotic required. This may also apply to our MCSs that, though not reaction types were observed in the 301C test. homologous series, are complex mixtures of several related In our experiments, based on HPLC-UV-MS/MS monitor- individual compounds. ing, rapid disulfide formation on test day 0 was observed for With respect to inherent biodegradability tests (OECD 3-MPA (→ DTDPA), MMP (→ MMP disulfide) and BuMP 302A-C), the guidelines say (→ BuMP disulfide), accompanied in the ester cases (MMP, BuMP) by formation of the corresponding monoester of “Biodegradation above 20% of theoretical […]may be DTDPA and of DTDPA itself during the test, for details, see regarded as evidence of inherent, primary biodegrad- the “Electronic supplementary material”. ability” (OECD 2006,paragraph 36;ECHA 2017, The dimercapto thioether DMDS was quickly and page 216). completely transformed in CBT and MRT at day 0 to the cyclic disulfide 1,2,5-trithiepane. This disulfide, in contrast As far as substances measured here achieved above 20% bio- to the disulfides mentioned, was not further biodegraded up degradation even in an OECD 301 test, a fortiori we consider to day 28 in CBT or MRT, as it does not contain an ester or this as evidence of their inherent primary biodegradability. acid substructure. A corresponding oxidative cyclisation was Similarly, observed for GDMP, for details, see the “Electronic supple- mentary material”. “When results of ready biodegradability tests indicate Oxidative cyclisation of linear dimercaptans to cyclic that the pass level criterion is almost fulfilled (i.e. disulfides was seen earlier under similar conditions (Houk ThOD […] slightly below 60% […]) such results can and Whitesides 1987;Adamczyk et al. 2015). Formation of be used to indicate inherent biodegradability” (OECD disulfides via oxidation of SH groups can be achieved using a 18410 Environ Sci Pollut Res (2018) 25:18393–18411 wide variety of oxidants, among them molecular oxygen require some time (lag phase) that depends on the particular (Ozen and Aydin 2006; Carriletal. 2007; García Ruano course of expression and biosynthesis of enzymes, growth of et al. 2008; Abaee et al. 2011; Dewan et al. 2012;Shard bacteria and microevolution in the test flask. A conventional et al. 2014), preferably under slightly alkaline conditions, in threshold for ready biodegradability such as 60% O con- what is presumably a dimerisation of thiyl radicals formed by sumption after 28 days may be passed in one experiment but oxidation of thiolate anions (RSH→ RS → RS·,2 RS·→ not in another under seemingly identical conditions. For the RSSR). The transformation mercaptan→ disulfide consumes mercaptans and mercaptoacids/esters considered here, concur- an amount of oxygen (4 RSH + O → 2RSSR+ 2H O) that is rent abiotic reactions such as disulfide formation and ester 2 2 small in most cases, e.g. complete conversion 3-MPA→ 3- hydrolysis complicate the picture. For these reasons, test re- MPA disulfide requires 5% of the ThOD of 3-MPA. In our sults are often not reproducible. Therefore, conclusions on a examples, this step was often so rapid as to occur during day 0 particular compound’s biodegradability cannot be based on a already to a large extent, in such a case the corresponding single experiment. However, a consolidated assessment of all oxygen consumption will not even be measured in CBT or available biodegradation information for a compound class MRT since it happens before the bottles are closed. may lead to a coherent picture. The disulfide functionality, C-S-S-C, is essential in bio- Funding information This study was funded by the European Regional chemistry, e.g. contributing to the conformational integrity Development Fund, grant agreement no. CCI No 2007DE161PR001 in of proteins including highly unusual natural products the Lüneburg Innovation Incubator. (Rücker and Meringer 2002) and in detoxification of xenobi- otics via glutathione. Compliance with ethical standards Conflict of interest D. S. is an employee of Bruno Bock Thiochemicals. Conclusion Open Access This article is distributed under the terms of the Creative New findings from this study are the following: Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, 1. The rule of thumb for biodegradation of carboxylic acids/ distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link esters is well based on experimental evidence: such a to the Creative Commons license, and indicate if changes were made. functional group enhances biodegradation compared to a methyl group or to an H atom. 2. A rule of thumb for biodegradation of mercaptans can tentativelybeestablished basedonexperimentalevi- References dence: a mercapto group hinders biodegradation com- pared to an H atom and even more so compared to a Abaee MS, Mojtahedi MM, Navidipoor S (2011) Diethylamine-catalyzed hydroxy group. dimerization of thiols: an inexpensive and green method for the synthesis of homodisulfides under aqueous conditions. Synth 3. If both kinds of functional groups are present, then an Commun 41:170–176 intermediate behaviour results that strongly depends on Adamczyk J, Bal W, Krezel A (2015) Coordination properties of the compound’s other structural characteristics. Simple dithiobutylamine (DTBA), a newly introduced protein disulfide re- mercaptocarboxylic acids and their esters are either read- ducing agent. Inorg Chem 54:596–606 ily biodegradable or at least biodegradable to some extent, BB (n.d.) unpublished study report or other material from Bruno Bock Thiochemicals i.e. they are not persistent compounds. Boethling RS, Sommer E, DiFiore D (2007) Designing small molecules for biodegradability. Chem Rev 107:2207–2227 Biodegradation even of rather simple organic compounds Brillet F, Maul A, Durand MJ, Thouand G (2016) From laboratory to to CO ,H O, etc. is a multistep process, each step depending 2 2 environmental conditions: a new approach for chemical’s biode- on catabolic capabilities (enzymes) of various microorgan- gradability assessment. Environ Sci Pollut Res 23:18684–18693 Bruland N, Wübbeler JH, Steinbüchel A (2009) 3-Mercaptopropionate isms present. In standardised laboratory biodegradation tests dioxygenase, a cysteine dioxygenase homologue, catalyzes the ini- as defined in the OECD 301 series, some influencing factors tial step of 3-mercaptopropionate catabolism in the 3,3- are not and cannot be standardised, most importantly the in- thiodipropionic acid-degrading bacterium Variovorax paradoxus.J oculum. Thus, in biodegradation experiments bacterial popu- Biol Chem 284:660–672 Carril M, SanMartin R, Domínguez E, Tellitu I (2007) A highly advan- lations are used that differ with respect to the species and tageous metal-free approach to diaryl disulfides in water. Green number of bacteria capable of degrading a particular com- Chem 9:315–317 pound. Necessary enzymes of more or less substrate selectiv- Cheng F, Ikenaga Y, Zhou Y, Yu Y, Li W, Shen J, Du Z, Chen L, Xu C, ity may or may not be present or inducible. Appearance of Liu G, Lee PW, Tang Y (2012) In silico assessment of chemical biodegradability. J Chem Inf Model 52:655–669 bacterial strains actually degrading the test compound may Environ Sci Pollut Res (2018) 25:18393–18411 18411 Dewan A, Bora U, Kakati DK (2012) A mild and efficient protocol for OECD (1995) Organisation for Economic Co-operation and Development, Detailed review paper on biodegradability testing, oxidation of thiols to disulfides in water. Heteroat Chem 23:231–234 Environment Monograph No. 98 ECHA (2017) Guidance on information requirements and chemical safe- OECD (2006) Organisation for Economic Co-operation and ty assessments, Chapter R.7b: Endpoint specific guidance, Development, OECD Guidelines for the testing of chemicals, June 2017, http://www.echa.europa.eu/documents/10162/13632/ Revised introduction to the OECD Guidelines for testing of information_requirements_r7b_en.pdf. Accessed 6 Sept 2017 chemicals, Section 3, adopted 23 March 2006, http://www. ECHA (n.d.) Database: European Chemicals Agency (ECHA), http:// oecdilibrary.org/docserver/download/9730001e.pdf?expires= echa.europa.eu. Accessed 6 Sept 2017 1489593611&id=id&accname=guest&checksum= EU (2006) European Union, Regulation (EC) No 1907/2006 of the 9B023D8E206ABB2DB4C2D77DC8867032.Accessed 6 European Parliament and of the Council of 18 December 2006 Sept 2017 concerning the Registration, Evaluation, Authorisation and OECD (2009) SIDS Initial Assessment Report for SIAM 28: TGA and its Restriction of Chemicals (REACH), establishing a European ammonium salt, webnet.oecd.org/HPV/UI/handler.axd?id= Chemicals Agency, amending Directive 1999/45/EC and repealing 2712f350-08af-4de9-9bc7-0bcfd5f22fce. Accessed 6 Sept 2017 Council Regulation (EEC) No 793/93 and Commission Regulation Ozen R, Aydin F (2006) Oxidation of thiols to disulfides with molecular (EC) No 1488/94 as well as Council Directive 76/769/EEC and oxygen in subcritical water. Monatsh Chem 137:307–310 Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and Painter HA, King EF (1985) Ring test programme 1983–84, assessment 2000/21/EC, Official J. Eur. Union L 2007,L 136, 3–280 (29 of biodegradability of chemicals in water by manometric respirom- May 2007), http://eurlex.europa.eu/LexUriServ/LexUriServ.do? etry. Commission of the European Union, Luxembourg, Report uri=OJ:L:2007:136:0003:0280:EN:PDF. Accessed 6 Sept 2017 EUR 9962 EN Friedrich J, Längin A, Kümmerer K (2013) Comparison of an electro- Pizzo F, Lombardo A, Manganaro A, Benfenati E (2013) In silico models chemical and luminescence-based oxygen measuring system for use for predicting ready biodegradability under REACH: a comparative in the biodegradability testing according to the Closed Bottle Test study. Sci Total Environ 463-464:161–168 (OECD 301D). Clean – Soil Air Water 41:251–257 REAXYS (n.d.) Database https://www.reaxys.com. Accessed 6 Sept 2017 García Ruano JL, Parra A, Alemán J (2008) Efficient synthesis of Rücker C, Kümmerer K (2012) Modeling and predicting aquatic aerobic disulfides by air oxidation of thiols under sonication. Green Chem biodegradation — a review from a user’s perspective. Green Chem 10:706–711 14:875–887 Guhl W, Steber J (2006) The value of biodegradation screening test re- Rücker C, Meringer M (2002) How many organic compounds are graph- sults for predicting the elimination of chemicals’ organic carbon in theoretically nonplanar? MATCH Commun Math Comput Chem waste water treatment plants. Chemosphere 63:9–16 45:153–172, http://match.pmf.kg.ac.rs/electronic_versions/ Houk J, Whitesides GM (1987) Structure-reactivity relations for thiol- Match45/match45_153-172.pdf. Accessed 6 Sept 2017 disulfide interchange. J Am Chem Soc 109:6825–6836 Salgado P, Visnevschi-Necrasov T, Kiene RP, Azevedo I, Rocha ACS, Howard PH (2000) Biodegradation. In: Boethling RS, Mackay D (eds) Almeida CMR, Magalhaes C (2015) Determination of 3- Handbook of property estimation methods for chemicals, mercaptopropionic acid by HPLC: a sensitive method for environ- Environmental and health sciences. CRC Press, Boca Raton mental applications. J Chromatogr B 992:103–108 J-CHECK (n.d.) Database: http://www.safe.nite.go.jp/jcheck/search. Shard A, Kumar R, Saima, Sharma N, Sinha AK (2014) Amino acid and action?request_locale=en. Accessed 6 Sept 2017 water-driven tunable green protocol to access S-S/C-S bonds via Ju F, Guo F, Ye L, Xia Y, Zhang T (2014) Metagenomic analysis on seasonal aerobic oxidative coupling and hydrothiolation. RSC Adv 4: microbial variations of activated sludge from a full-scale wastewater 33399–33407 treatment plant over 4 years. Environ Microbiol Rep 6:80–89 Toups M, Wübbeler JH, Steinbüchel A (2010) Microbial utilization of the industrial wastewater pollutants 2-ethylhexylthioglycolic acid and Kayashima T, Taruki M, Katagiri K, Nabeoka R, Yoshida T, Tsuji T iso-octylthioglycolic acid by aerobic Gram-negative bacteria. (2014) Comparison of biodegradation performance of the OECD Biodegradation 21:309–319 test guideline 301C with that of other ready biodegradability tests. UN (2015) United Nations, Globally Harmonized System of classifica- Environ Toxicol Chem 33:328–333 th tion and labelling of chemicals (GHS), 6 edition, New York and Kim TS, Jeong JY, Wells GF, Park HD (2013) General and rare bacteria Geneva, 2015, Appendix 9, https://www.unece.org/fileadmin/ taxa demonstrating different temporal dynamic patterns in an acti- DAM/trans/danger/publi/ghs/ghs_rev06/English/ST-SG-AC10-30- vated sludge bioreactor. Appl Microbiol Biotechnol 97:1755–1765 Rev6e.pdf. Accessed 6 Sept 2017 Klimisch HJ, Andreae M, Tillmann U (1997) A systematic approach for US EPA (n.d.) Estimation Programs Interface Suite™ for Microsoft® evaluating the quality of experimental toxicological and ecotoxico- Windows, v 4.11, United States Environmental Protection Agency, logical data. Regul Toxicol Pharmacol 25:1–5 Washington, DC, USA, https://www.epa.gov/tsca-screening-tools/ Kowalczyk A, Martin TJ, Price OR, Snape JR, van Egmond RA, Finnegan download-epi-suitetm-estimation-program-interface-v411. CJ, Schäfer H, Davenport RJ, Bending GD (2015) Refinement of Accessed 6 Sept 2017 biodegradation tests methodologies and the proposed utility of new van Ginkel CG, Stroo CA (1992) Simple method to prolong the Closed microbial ecology techniques. Ecotoxicol Environ Saf 111:9–22 Bottle Test for the determination of the inherent biodegradability. OECD (1992) Organisation for Economic Co-operation and Ecotoxicol Environ Saf 24:319–327 Development, OECD Guideline for testing of chemicals – Ready Vorberg S, Tetko IV (2014) Modeling the biodegradability of chemical biodegradability, Paris 1992, http://www.oecd.org/chemicalsafety/ compounds using the Online CHEmical Modeling environment risk-assessment/1948209.pdf. Accessed 6 Sept 2017 (OCHEM). Mol Inf 33:73–85

Journal

Environmental Science and Pollution ResearchSpringer Journals

Published: Apr 17, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off