Document 82wLx8KB0j7VzaB9rpQ4vk0OK

Human and Ecological Risk Assessment: An International Journal ISSN: 1080-7039 (Print) 1549-7860 (Online) Journal homepage: https://www.tandfonline.com/loi/bher20 Environmental Risk Assessment of Trifluoroacetic Acid Jean Charles Boutonnet , Pauline Bingham , Davide Calamari , Christ de Rooij , James Franklin , Toshihiko Kawano , Jean-Marie Libre , Archie McCulloch , Giuseppe Malinverno , J. Martin Odom , George M. Rusch , Katie Smythe , Igor Sobolev , Roy Thompson & James M. Tiedje To cite this article: Jean Charles Boutonnet , Pauline Bingham , Davide Calamari , Christ de Rooij , James Franklin , Toshihiko Kawano , Jean-Marie Libre , Archie McCul-loch , Giuseppe Malinverno , J. Martin Odom , George M. Rusch , Katie Smythe , Igor Sobolev , Roy Thompson & James M. Tiedje (1999) Environmental Risk Assessment of Trifluoroacetic Acid, Human and Ecological Risk Assessment: An International Journal, 5:1, 59-124, DOI: 10.1080/10807039991289644 To link to this article: https://doi.org/10.1080/10807039991289644 Published online: 14 Jan 2011. Submit your article to this journal Article views: 824 View related articles Citing articles: 27 View citing articles Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=bher20 H uman and Ecological Risk Assessment: Vol. 5, No. 1, pp. 59-124 (1999) Environmental Risk Assessment of Trifluoroacetic Acid Jean C harles Boutonnet (Ed.),1 Pauline Bingham,2 Davide C alamari,3 C hrist de Rooij,4 James Franklin,5 T oshihiko Kawano,6 Jean-Marie L ibre,7 Archie McC ulloch,8 G iuseppe Malinverno,9 J. Martin O dom,10 G eorge M. Rusch,11 Katie Smythe,12,* Igor Sobolev,13 Roy T hompson,14 and James M. T iedje15 1Elf Atochem S.A., Centre d'Application de L evallois, 95 rue Danton, BP 108, 92303 L evallois-Perret Cedex, France; 2Rhodia, L td., St. Andrews Road, Avonmouth, Bristol BS11 9YF U .K.; 3Department of Structural and Functional Biology, U niversity of M ilan, Via Ravasi 2, 21100 Varese VA I taly; 4Solvay S.A., rue de Ransbeek 310, B-1120 Brussels, Belgium; 5Solvay S.A., rue de Ransbeek 310, B-1120 Brussels, Belgium; 6Daikin I ndustries L td., 1-1 N ishi-H itotsuya, Settsushi, O saka 566 Japan; 7Elf Atochem S.A., 4 Cours M ichelet, L a Dfense 10, Cdex 42 92091 Paris la Dfense, France; 8I CI Chemicals & Polymers L td., Environment Department, P.O . Box 13, T he H eath, Runcorn, Cheshire WA7 4Q F U .K.; 9Ausimont S.p.A, via L ombardia 20, 20021 Bollate (M ilan), I taly; 10E.I . DuPont de Nemours & Company, Experimental Station (E328-B47), Wilmington, DE 19898; 11Allied Signal I nc., P.O . Box 1139, M orristown, N J 07962- 1139; 12AFEAS Program O ffice, 1333 H Street N W, Washington, DC 20005; 13Chemical & Polymer T echnology I nc., 5 Rita Way, O rinda, CA 94563; 14Zeneca L td., Brixham Environmental L aboratory, Freshwater Q uarry, Brixham, Devon T Q 5 8BA U .K.; 15Center for M icrobial Ecology, M ichigan State U niversity, East L ansing, M I 48824-1326 ABST RAC T T he M ontreal Protocol was developed in 1987 in response to concerns that the chlorofluorocarbons (CFCs) were releasing chlorine into the stratosphere and that this chlorine was causing a depletion of stratospheric ozone over Antarctica. T his international agreement called for a phase out of these CFCs. Industry initiated a major effort to find replacements that are safe when properly * Primary contact. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. Boutonnet used and safe to the environment. T he toxicology and environmental fate of these first generation replacements has been studied extensively. It was determined that the new substances break down in the environment to give predominantly carbon dioxide, water and inorganic salts of chlorine and fluorine. T he only exception is that some substances also break down to yield trifluoroacetic acid (H T FA), a substance resistant to further degradation. Recognizing this, industry embarked on a research and assessment program to study the potential effects of trifluoroacetate (T FA) on the environment and to investigate possible degradation pathways. T he resultsof these recentlycompleted studies are summarized below and described in further detail in this paper. T rifluoroacetic acid is a strong organic acid with a pKa of 0.23. It is miscible with water and its low octanol/ water partition coefficient (log Pow =-2.1) indicates no potential to bioaccumulate. I ndustrial use is limited and environmental releases are very low. Some additional T FA will be formed from the breakdown of a few halogenated hydrocarbons, most notably H FC-134a (CF3CH 2F), H CFC-124 (CF3CH FCl), and H CFC-123 (CF3CH Cl2). As these substances have only been produced in limited commercial quantities, their contribution to environmental levels has been minimal. Surprisingly, environmental measurements in many of diverse locations show existing levels of 100 to 300 ngl-1 in water with one site (Dead Sea) having a level of 6400 ngl-1. T hese levels cannot be accounted for based on current atmospheric sources and imply a long-term, possibly pre-industrial source. Generally, soil retention of T FA is poor although soils with high levels of organic matter have been shown to have a greater affinity for T FA when contrasted to soils with low levels of organic matter. T his appears to be an adsorption phenomenon, not irreversible binding. T herefore, T FA will not be retained in soil, but will ultimately enter the aqueous compartment. M odeling of emission rates and subsequent conversion rates for precursors has led to estimates of maximum levels of T FA in rain water in the region of 0.1 gl-1 in the year 2020. T FA is resistant to both oxidative and reductive degradation. While there had been speculation regarding the possibility of T FA being degraded into monofluoroacetic acid (M FA), the rate of breakdown of M FA is so much higher than for T FA that any M FA formed would rapidly degrade. T herefore, there would be no buildup of M FA regardless of the levels of T FA present in the environment. Although highly resistant to microbial degradation, there have been two reports of T FA degradation under anaerobic conditions. In the first study, natural sediments reduced T FA. H owever, even though this work was done in replicate, the investigators and others were unable to reproduce it in subsequent studies. In the second study, radiolabeled T FA was removed from a mixed anaerobic in vitro microcosm. L imited evidence of decarboxylation has also been reported for two strains of bacteria grown under highly specific conditions. T FA was not biodegraded in a semi-continuous activated sludge test even with prolonged incubation (up to 84 days). Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 60 Environmental Risk Assessment of T rifluoroacetic Acid T FA does not accumulate significantly in lower aquatic life forms such as bacteria, small invertebrates, oligochaete worms and some aquatic plants including L emna gibba (duckweed). Some bioaccumulation was observed in terrestrial higher plants, such as sunflower and wheat. T his result appeared to be related to uptake with water and then concentration due to transpiration water loss. When transferred to clean hydroponic media, some elimination of T FA was seen. Also, more than 80% of the T FA in leaves was found to be water extractable, suggesting that no significant metabolism of T FA had occurred. At an exposure level of 1200 mgl-1 of sodium trifluoroacetate (N aT FA) -- corresponding to 1000 mgl-1 H T FA -- no effects were seen on either Brachydanio rerio (a fish) or Daphnia magna (a water flea). With duckweed, mild effects were seen on frond increase and weight increase at the same exposure level. At a concentration of 300 mgl-1 no effects were observed. T oxicity tests were conducted with 11 species of algae. For ten of these species the EC50 was greater than 100 mgl-1. I n Selenastrum capricornutum the no-effect level was 0.12 mgl-1. At higher levels the effect was reversible. T he reason for the unique sensitivity of this strain is unknown, but a recovery of the growth rate was seen when citric acid was added. T his could imply a competitive inhibition of the citric acid cycle. T he effect of T FA on seed germination and plant growth hasbeen evaluated with a wide variety of plants. Application of N aT FA at 1000 mgl-1 to seeds of sunflower, cabbage, lettuce, tomato, mung bean, soy bean, wheat, corn, oats and rice did not affect germination. Foliar application of a solution of 100 mgl-1 of N aT FA to field grown plants did not affect growth of sunflower, soya, wheat, maize, oilseed rape, rice and plantain. When plantain, wheat (varieties Katepwa and H anno) and soya were grown in hydroponic systems containing N aT FA, no effects were seen on plantain at 32 mgl-1, on wheat (Katepwa) and soya at 1 mgl-1, or on wheat (H anno) at 10 mgl-1; some effects on growth were seen at, respectively, 100 mgl-1, 5 mgl-1, 5 mgl-1, and 10 mgl-1 and above. T FA is not metabolized in mammalian systems to any great extent. It is the major final metabolite of halothane, H CFC-123 and H CFC-124. T he half-life of T FA in humans is 16 hours. As expected, the acute oral toxicity of the free acid is higher than the one of the sodium salt. T he inhalation L C50 (2 hour exposure) for mice was 13.5 mgl-1 (2900 ppm) and for rats it was 10 mgl-1 (2140 ppm). T hus, T FA is considered to have low inhalation toxicity. T he irritation threshold for humans was 54 ppm. As one would expect of a strong acid, it is a severe irritant to the skin and eye. When conjugated with protein, it has been shown to elicit an immunological reaction; however, it is unlikely that T FA itself would elicit a sensitization response. Repeat administration of aqueous solutions have shown that T FA can cause increased liver weight and induction of peroxisomes. Relative to the doses (0.5% in diet or 150 mgkg-1day-1 by gavage) the effects are mild. In a series of Ames assays, T FA was reported to be non-mutagenic. Its carcinogenic potential has not been evaluated. Although T FA was shown to accumulate in amniotic fluid following exposure of pregnant animals to high levels of halothane (1200 ppm), no fetal effects were seen. L ikewise, a reproduction Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 61 Boutonnet study that involved exposure of animals to halothane at levels up to 4000 ppm for 4 hours per day, 7 days per week, resulted in no adverse effects. Given the high levels of halothane exposure, it is unlikely that environmental T FA is a reproductive or developmental hazard. O verall the toxicity of T FA has been evaluated in stream mesocosms, algae, higher plants, fish, animals and humans. It has been found to be of very low toxicity in all of these systems. T he lowest threshold for any effects was the reversible effect on growth of one strain of algae, Selenastrum capricornutum, which was seen at 0.12 mgl-1. T here is a 1000-fold difference between the noeffect concentration and the projected environmental levels of T FA from H FCs and H CFCs (0.0001 mgl-1). Based on available data, one can conclude that environmental levels of T FA resulting from the breakdown of alternative fluorocarbons do not pose a threat to the environment. Key Words: hydrochlorofluorocarbons(H CFCs), hydrofluorocarbons(H FCs), CFC alternatives, trifluoroacetate (T FA), environmental fate, ecotoxicity. IN T RO DU C T IO N AN D PERSPEC T IVE T he M ontreal Protocol was developed in 1987 in response to concerns that the chlorofluorocarbons (CFCs) were releasing chlorine into the stratosphere and that this chlorine was causing a depletion of stratospheric ozone over Antarctica. T his international agreement called for a phase out of these CFCs. Industry initiated a major effort to find safe replacements. T he Alternative Fluorocarbons Environmental Acceptability Study (AFEAS) was formed in 1989 by several chemical companies from Europe, Japan, and the U nited States with the aim of assessing the environmental acceptability of a range of hydrochlorofluorocarbons (H CFCs) and hydrofluorocarbons (H FCs). Studies byindependent researchersover the next six yearswere coordinated with work supported by U .S. and European government agencies. T he results showed that, unlike CFCs, alternative fluorocarbons will break down readily in the lower atmosphere to form simple inorganic species alreadypresent in the environment (NASA, NO AA and AFEAS, 1995 and references cited therein). H owever, a few of the H CFCs and H FCs can be expected to form trifluoroacetyl halides that will dissolve in environmental water to give trifluoroacetate (T FA) salts. In 1991, in response to concerns that this would introduce a new and potentially hazardous material into the environment, AFEAS initiated a research program to determine the environmental fate of T FA and to provide a timely and accurate forecast of potential ecological impact. T he initial concern had been that H FC-134a (1,1,1,2-tetrafluoroethane, an alternative to CFC-12) would form T FA, which has relatively low mammalian toxicity, and that the T FA would subsequently react in the environment to give monofluoroacetic acid (M FA), which is acutely poisonous at low doses. T heoretical and experimental data (Emptage, 1994) quickly removed these concerns about M FA ac- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 62 Environmental Risk Assessment of T rifluoroacetic Acid cumulation in the environment, but there was still a clear need to examine the ecological impact of T FA. A panel of advisorswasconvened in 1992 to outline research that would provide a valid and systematic approach to assessing the impact of T FA. T he experts suggested a strategy to identify key "pressure" points in the major biogeochemical cycles -- carbon fixation, mineralization, nitrogen fixation and bioaccumulation in plants -- where T FA could conceivably exert a major influence. Furthermore, plans were devised for assessing whether T FA can be biodegraded in the environment. T his approach was then used to guide an evolutionary research program that was executed over the last 4 years through collaborative efforts between academic and industrial laboratories. T he minimum levels of quality control were compliance with "Good L aboratory Practice" guidelines or publication of the results in peer-reviewed journals. A number of physico-chemical and biodegradation studies have been completed, in addition to biological investigations. During the course of the work it became clear that T FA is not a new environmental contaminant. It is present in contemporary air, precipitation and surface waters from around the world, and before significant amounts of H FC or H CFC precursors have been produced and released. T he program described here represents one of the most comprehensive efforts to assess the environmental impact of a chemical ever undertaken by industry. T he risk assessment is a synthesis of the research results and an assessment based on best available data. I t consists of a comparison of the anticipated future levels of T FA, arising from decomposition of H FCs and H CFCs, with concentrations that would exert environmental effects. M ore specifically, the risk assessment covers: general substance information, sources, distribution and releases into the environment, exposure measurements, degradation, accumulation, effects assessment, and risk characterization. In this article, the following abbreviations will be used: T FA: trifluoroacetate H T FA: trifluoroacetic acid NaT FA: sodium trifluoroacetate G EN ERAL SU BST AN C E IN FO RMAT IO N Identification of the Substance N ame: CAS number: Synonym: Formula: M olecular weight: T rifluoroacetic acid 76-05-1 H TFA CF3CO 2H 114 Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 63 Boutonnet N ame: Sodium trifluoroacetate CAS number: Synonym: Formula: M olecular weight: 2923-18-4 NaT FA CF3CO 2N a 136 Physico-C hemical Properties H T FA Physical state: Specific gravity: M elting point: Boiling point: Vapor pressure: Water solubility: Partition coefficient n-O ctanol/ water (log): Colorless fuming (hygroscopic) liquid 1484 kgm-3 at 25C -15.3C 72C 105.7 hPa at 20C M iscible in all proportions (Feenstra-Bieders and O lthof, 1992) T hus, 1997) pK a: pH (1% solution): pH (100 mgl-1 distilled water): pH (10 mgl-1 distilled water): pH (100 mgl-1 buffered ISO water): pH (10 mgl-1 buffered ISO water): H enry's L aw Constant KH=1.1 U V/ vis absorption: Conversion Factors (air, 1 atm, 25C): NaT FA Physical state: M elting point: Water solubility: pH (2% solution): Partition coefficient n-O ctanol/ water (log): -2.1 (Feenstra-Bieders and O lthof, 1992; -0.2 (calculated according to Rekker) 0.325 (ClogP for Windows V 1.0.0) 0.5 (SRC's KO WWIN v1.52) 0.23 1 3.1 7.4 - 7.75 4 7.4 (T hus and van Dijk, 1996) 10-2 Pam-3mol-1 at 25C (Bowden et al., 1996) No absorption at >250 nm (van Dijk, 1992a) 1 mgl-1 =214 ppm; 1 ppm =4.66 mgm-3 White (hygroscopic) powder 207C (decomposition) 625 gl-1 ( 25C) 7 -3.31 (SRC's KO WWIN v1.52) Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 64 Environmental Risk Assessment of T rifluoroacetic Acid SO U RC ES Production and Known Sources According to Elliott (1994), trifluoroacetic acid is manufactured by only three companies: H alocarbon Products Corporation in the U nited States, and Rhne-Poulenc and Solvay in Europe. T he quantities involved are relatively small -- of the order of 1000 metric tons per year ( J. Franklin, personal communication), an estimate that is consistent with the quoted revenue of the major U .S. producer (M cConville, 1996). T he processes are enclosed, with effluent being subject to local requirements for disposal of chemical waste. Similar chemical waste streams can also arise from processes in which trifluoroacetate is produced as a byproduct. From one such plant (DuPont - Fayetteville, North Carolina), the site effluent contained about 3 ppm of trifluoroacetate (Chumley, 1992). Environmental trifluoroacetate can also be produced during the oxidation of several organofluorine compounds released to the atmosphere by human activities. H alothane and isoflurane anaesthetics, which have been in use since the 1970s, yield trifluoroacetate ( R. Atkinson, personal communication) . Also, some of the fluorocarbon alternatives to CFCs decompose in the atmosphere to form trifluoroacetate (NASA, NO AA, and AFEAS, 1995) and, while these sources are currently small, they could become significant in the future and merit the closer scrutiny given below. U ses T rifluoroacetic acid is widely used in the fine chemicals industry and as a laboratory reagent: for derivatizing carbohydrates, amino acids and peptides; as a catalyst in esterification reactions and the Beckmann rearrangement of oximes to amides, and for protein synthesis (Elliott, 1994). In all cases, the material is either consumed or becomes part of a chemical waste stream. T he quantities released into the atmosphere are very small indeed. Emission Pattern T he known sources can be expected to yield fluxes of trifluoroacetate into the environment that differ significantly in both geographical distribution and the first compartment into which they are released. For example, only if there are fugitive emissions of vapor will material that is used as a chemical intermediate give rise to a point source of atmospheric contamination, otherwise releases can be expected to be to the aqueous compartment. T he H enry'sL aw constant (KH) for trifluoroacetic acid estimated from measurements by Bowden et al. (1996) is 8.95 0.1 10-3 molkg-1atm-1, which translates into a value for KH of 1.1 10-2 Pam-3mol-1. With such a value, and because trifluoroacetic acid is totally miscible with water and has a log KOW value of -0.2, the preferred environmental compartment will be water, rather than air, ground, or biota. T he rules set out by Ballschmiter (1992) state that, if K H is less than 2 Pam-3mol-1 and log KO W is less than 4, the material should preferentially distribute into the aqueous compartment of the environment. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 65 Boutonnet T hus, trifluoroacetic acid that is emitted from processes as vapor, or trifluoroacetate that is formed from other materials by reaction in the atmosphere will partition rapidly into cloud, rain, or surface water. Substances that decompose in the atmosphere by oxidation or photolysis will give rise to diffuse fluxes of decomposition products into the atmosphere as the first compartment. T hese substances have atmospheric lifetimes ranging from about 1 to 40 years (WM O , 1994) so that most decomposition occurs at or near the background concentration of the substance after it has become well mixed in the atmosphere. T he decomposition mechanisms depend on physical and chemical properties and processes that are relatively well understood (NASA, NO AA, and AFEAS, 1995) so that it is possible to calculate accurately the future atmospheric concentrations of precursors to trifluoroacetate, given a particular set of scenarios for emissions. Furthermore, the flux of trifluoroacetic acid, into the atmosphere and thence into rain and surface water, can be estimated from the precursor concentrations with confidence and provides the possibility of verifying the magnitude of known fluxes against environmental observations. Concentrations of trifluoroacetate observed in the air, in precipitation and in surface waters in Europe are several orders of magnitude larger than the known source fluxes would allow (Frank et al., 1996). Similar concentrations have been observed in contemporarywater and air samplesfrom Nevada, U SA; Canada; Australia and South Africa (Zehavi and Seiber, 1996; Frank and Klein, 1997; Grimvall et al., 1997), suggesting that there is one or more large unknown source of environmental trifluoroacetate. T he sizes of the fluxes of the known source gases are not in doubt; their concentrations calculated from estimated emissions are consistent with observations, and additional contemporary sources, substantially larger than those known, must be invoked to explain the observed environmental concentrations of trifluoroacetate. T he current environmental burden from the decomposition of man-made fluorocarbons, and the anticipated future burden resulting from future emissions of such precursors, should be evaluated in the context of the background concentrations that have been observed. DIST RIBU T IO N AN D REL EASES IN T O EN VIRO N MEN T From T rifluoroacetic Acid Production and U se With a pKa of 0.23, the strength of trifluoroacetic acid approaches that of mineral acids. U nlike the mineral acids, it is miscible not just with water but with fluorocarbons and most common organic solvents including methanol, benzene, carbon tetrachloride, acetone, ether, and hexane. It is a good solvent for proteins, leading to its use in protein synthesis. I t is useful for making derivatives of carbohydrates, amino acids and peptides from which the trifluoroacetyl protective group can be removed relatively easily (Elliott, 1994). T rifluoroacetic acid is a useful catalyst for esterifications of alcohols, acylations of aromatics (H alocarbon, 1967) and in the Beckmann rearrangement of Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 66 Environmental Risk Assessment of T rifluoroacetic Acid oximes to amides (H uber, 1955). T hese uses are typical of the fine chemicals industry and of laboratories, and the total global production of trifluoroacetic acid is expected to reflect the small scale of the uses. Although there are no audited data for production and use, informal estimates suggest that the quantities involved are of the order of 1000 metric tons per year. T rifluoroacetic acid is stable to oxidation and so can be prepared by a number of routes that involve the oxidation of compounds that have -trifluoromethyl groups. T hus, alkaline permanganate oxidation of 2,3-dichlorohexafluorobut-2-ene gives an 87% yield of the acid (H enne and T rott, 1947) and, in a manner similar to atmospheric degradation processes, photochemical oxidation of 1-chloro-2,2,2-trifluoroethane or 1,1-dichloro-2,2,2-trifluoroethane with oxygen gives high yields of trifluoroacetyl chloride that can be hydrolyzed to trifluoroacetic acid (H aszeldine and Nyman, 1959; Dittmann, 1975) . L ike the use processes, the production processes involve reactions in contained and/ or aqueous systems from which fugitive losses of trifluoroacetic acid to the atmosphere are small. In many cases the use is as a raw material so that most of it is wholly converted and cannot be released as trifluoroacetic acid or its salts. From both production and use, any persistent loss is likely to be in the form of solution in chemical waste streams, the disposal and treatment of which will be subject to local controls. Deliberate production and use of trifluoroacetic acid does not appear to be a significant contributor to current global environmental levels. By-Product from C hemical Syntheses Production of hexafluoropropylene oxide can result in an aqueous waste stream containing trifluoroacetate. T he concentration of trifluoroacetate measured in one plant outfall was 3 ppm or less (Chumley, 1992) and, while this may or may not be typical of such processes, the compartment and the concentration are consistent with the properties discussed in a previous section. Atmospheric O xidation of Fluorinated H ydrocarbons In recent years the atmospheric decomposition of halocarbons, particularly those fluorocarbons that could replace CFCs, has been the subject of intensive study. It is now agreed that the mechanism involves an initial abstraction of the hydrogen atom in the molecule byatmospheric hydroxyl radicalsto yield water and a haloalkyl radical. T he latter reacts rapidlywith oxygen to give a haloalkylperoxy radical that undergoes further reactions, generally with nitric oxide or hydroperoxyradicals. T he product in both cases is a haloalkoxy radical, which can react further in one of three ways, depending on its molecular composition: C-Cl bond cleavage, C-C bond cleavage or hydrogen abstraction (Cox et al., 1995; M cCulloch and Sidebottom, 1993; M idgley, 1995). T hus, the general reactions of those haloalkoxy radicals that could form precursors to trifluoroacetate are: Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 67 Boutonnet CF3CYlO CF3CYO +Cl ( 1) CF3CYZO CF3 +CYZO ( 2) CF3CYH O +O 2 CF3CYO +H O 3 ( 3) where Y and Z can be Br, Cl, F, or H . T he ultimate products of many of these reactions are fluoride and chloride ions and carbon dioxide. T hese are neither new to the environment nor expected to be introduced at rates that would significantly enhance the background levels (WM O , 1989). H owever, if Y is either Br, Cl, CF3, or F then the products of reactions (1) and (3) are trifluoroacetyl halides that can, on hydrolysis with atmospheric or surface water, form trifluoroacetic acid. T hus, trifluoroacetic acid is by no means a universal product of the degradation of fluorinated hydrocarbons; the parent compound requires a CF3 group and a halogen atom on the carbon adjacent to it. Substances that fulfill these criteria are: CF3CH ClBr 1-bromo-1-chloro-2,2,2-trifluoroethane `H alothane' CF3CH ClO CH F2 1-chloro-2,2,2-trifluoroethyl difluoromethyl ether `I soflurane' CF3CH Cl2 1,1-dichloro-2,2,2-trifluoroethane H CFC-123 CF3CH FCl 1-chloro-1,2,2,2-tetrafluoroethane H CFC-124 CF3CH 2F 1,1,1,2-tetrafluoroethane H FC-134a CF3CH FCF3 1,1,1,2,3,3,3-heptafluoropropane H FC-227ea H FC-125 (1,1,1,2,2-pentafluoroethane) matches the criteria but undergoes solely C-C bond cleavage according to reaction (2); reaction (1) would require the cleavage of a C-F bond which isnot possible under atmospheric conditions. H alothane and isoflurane anaesthetics are expected to form totally trifluoroacetyl chloride or bromide (probably the former) (R. Atkinson, personal communication). H CFC-123 is converted 98% to trifluoroacetyl chloride, H CFC124 100% to trifluoroacetyl fluoride and H FC-227ea yields equal amounts of carbonyl difluoride and trifluoroacetyl fluoride (Zellner et al., 1994). H FC134a, however, can undergo both reactions (2) and (3) so that it forms a mixture of trifluoroacetyl fluoride, formyl fluoride and carbonyl difluoride (T uazon and Atkinson, 1995; Cox et al., 1995). T he extent to which each reaction will proceed in the atmosphere depends on the local temperature where decomposition is occurring and the local pressure, particularly partial pressure of oxygen for reaction (3). At sea-level, only some 18% of the H FC-134a de- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 68 Environmental Risk Assessment of T rifluoroacetic Acid composing is converted to trifluoroacetyl fluoride but, with increasing altitude (and consequently reducing temperature and pressure), the yield of trifluoroacetyl fluoride increases and it becomes the major product at the tropopause. An average yield for the troposphere is calculated to be in the region of 40% (Cox et al., 1995; Franklin, 1993; M idgley, 1995). Recently, it has been suggested that these conversions to trifluoroacetyl fluoride might be overestimates and that the true value could be as little as 12% or one third of the value now accepted. T his view is based on work done in a laboratory system that contained N O ( Wallington et al., 1996) , which is thought to enhance the scission reaction via a relatively stable CF3CFH O O N O intermediate. I f this is truly representative of the atmospheric behavior of H FC-134a, it would reduce the expected flux of T FA from this source by a factor of about three. A yield of 40% is, however, used throughout the calculations of source strengths in this assessment on the grounds that this is the value used in most of the atmospheric modeling studies and that it represents a maximum from the H FC-134a source. T rifluoracetyl fluoride is stable towards photolysis in the troposphere but trifluoroacetyl chloride (and bromide) will photolyze there, reducing the apparent yields of these acid halides from the parent compounds to about 60% of stoichiometry in the case of the acid chloride (Cox et al., 1995). T he acid halides will be taken up by atmospheric or surface water relatively rapidly; the rates are governed by the rate constants of the hydrolysis reactions forming trifluoroacetic acid and their H enry's constants for aqueous solution. I n all cases, the maximum lifetime for the acid halide would be of the order of 30 daysand the minimum lifetime, which isgoverned more bycloud formation probabilities, is about 10 days (Kolb et al., 1995a,b). T hese lifetimes are long compared with regional mixing times so that, having been formed in the atmosphere, the trifluoroacetyl halide would be dispersed on a continental scale before being deposited as trifluoroacetic acid. Nevertheless, the lifetimes are short compared to those of the parent compounds which are governed by the rates of reaction with hydroxyl radicals. T he atmospheric lifetimes of the parent compounds range from about 1 year, in the case of the anaesthetics, halothane and isoflurane (Brown et al., 1990), through 1.4 years for H CFC-123 (WM O , 1994), 6.1 years for H CFC-124, and 14.6 years for H FC-134a (IPCC, 1996a). In all cases, such lifetimes mean that the compounds will be hemispherically, if not globally, dispersed and so they will react at concentrations that are relatively constant geographically. L ocal variations in hydroxyl radical concentrations are liable to have more profound effects on the rates and extent of reactions (M adronich and Dentener, 1995; Kanakidou et al., 1995). C urrent Releases T he atmospheric concentrations of the H CFCs and H FC-134a are determined. O n the contrary, there are no concentration measurements and only informal estimates for releases of the anaesthetics (halothane and isoflurane) Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 69 Boutonnet to the atmosphere. It has been assumed that halothane remains the most used of the fluorocarbon anaesthetics at a maximum of 1500 metric tons per year globally, and that all the material used is emitted to atmosphere (M cCulloch, 1995b). T he figure is consistent with the productive capacity and market share of the major manufacturer (C&P News, 1995 and 1996). T he same references suggest that production has been at or near this level for several years so that the fluxes of these anaesthetics and their atmospheric decomposition can be assumed to be in equilibrium. If the use of isoflurane were half as much as that of halothane, the total quantity of trifluoroacetyl chloride formed by the anaesthetics would be in the region of 1540 metric tons per year; given that this material photolyzes, the effective flux would be 930 metric tons per year, equivalent to a global deposition rate of 800 metric tons per year of trifluoroacetic acid. T he potential H CFC precursors to trifluoroacetic acid are at the point of their limits of detection in the atmosphere (P.G. Simmonds, personal communication), inferring that their concentrations are, at most, 0.1 parts per trillion by volume (pptv). T he upper limits for the atmospheric burdens of these compounds are 2500 metric tons of H CFC-123 and 2300 metric tons of H CFC-124, consistent with the values given in M cCulloch (1995b) for lower limits of detection. T he value for H CFC-123, coupled with its atmospheric lifetime of 1.5 years, equates to a decomposition rate of 1690 metric tons per year which, after allowing for the photolysis of the trifluoroacetyl chloride product, gives a deposition rate for trifluoroacetic acid of 760 metric tons per year. Similarly, the deposition rate for trifluoroacetic acid from H CFC-124 is calculated to be 320 metric tons per year. Contemporary concentrations of H FC-134a in the atmosphere are 2.5 pptv in the Northern H emisphere and 1.2 pptv in the Southern H emisphere (M ontzka et al., 1996; O ram et al., 1996). T his corresponds to an atmospheric burden of approximately 31,000 metric tons. U sing the atmospheric lifetime of 14.6 years, the decomposition rate is calculated to be 2140 metric tons per year. At a 40% yield, this would amount to 960 metric tons per year of trifluoroacetic acid deposited globally from this source. H FC-227ea has yet to be detected in the atmosphere and so the potential contribution from it has been assumed to be zero now. T herefore, the maximum estimate for the total contemporary deposition rate of trifluoroacetate from fluorinated hydrocarbons is 800 +760 +320 +960, or a total of 2800 metric tons per year. Scenarios for Future Releases of Precursors A lternative Scenarios With the exception of the anaesthetics, demand for which is static or even falling, the fluorocarbon precursors to trifluoroacetic acid are expected to be released in progressively larger amounts throughout the coming decades. Production and use of H CFCs is controlled under the M ontreal Protocol, so that the emissions of H CFC-123 and H CFC-124 will become limited within the next Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 70 Environmental Risk Assessment of T rifluoroacetic Acid 25 years and will eventually fall to zero. T here is no specific regulatory limit on H FCs and so, theoretically, growth of H FC-134a emissions could continue unabated. In order to judge the potential future effects of these materials on the environment, various scenarios have been proposed, based on several methodologies. 1. For the Enquete Commission of the German Parliament (DB, 1994) on life cycle analysis applied to H FC-134a and other substitute refrigerants, emissions of H FC-134a were calculated to reach 160 kT yr-1 by 2020, with production in that year running at 310 kT yr-1. T he calculations were based on projection of underlying demand, using projected regional GDPs, and views on the extent of substitution and integrity of containment. 2. T he demand for alternative fluorocarbonsover the period to 2020 wasestimated by M cCulloch (1994, 1995a) from a detailed analysis of the historical market for CFCs and estimates for their replacement by H CFCs and H FCs. T he values for H FC-134a production and emissions in 2020 were almost identical to those in DB (1994). No data for H CFCs 123 and 124 were provided by this work. 3. In order to gauge the potential climate change effect from future emissions of fluorocarbons, the Intergovernmental Panel on Climate Change (IPCC) has provided scenarios for H CFCs and H FCs. T hese were developed by projecting CFC demand and proposing almost total substitution of that demand using H CFCs and H FCs. T he scenarios, which cover all greenhouse gases, were originally proposed in 1992 to replace the 1990 scenarios that were considered unacceptable by the parties to the Rio Convention. T hey have been updated in the light of changing regulations and the latest version is contained in the "Second Assessment Report" (IPCC, 1996b). O nly for H CFC-123 is the emission specified in mass units; it is proposed that it reaches a maximum of 130 kT yr-1 at the end of this century, falling to 13 kT yr-1 by 2020. Emissions of H CFC-124 are not mentioned; presumably they are expected to be very low. For H FC-134a, the scenario output is in terms of atmospheric concentrations and values of 200-230 pptv are proposed for the year 2020, the variability arising from different assumptions about overall economic growth. T he equivalent emissions can be back-calculated from these concentrations and, for H FC-134a, amount to 400-470 kT yr-1. H FC-227ea is not included specifically in this scenario. T he much larger values proposed in the IPCC scenarios reflect the underlying assumption that the materialsare used to substitute in all historic markets for CFCs, including immediately dispersive uses such as aerosol propellants. T hey may be unrealistically high. For example, the IPCC scenarios anticipated that the current atmospheric concentration of H FC-134a would be 13 to 14 pptv. T he actual measurements, as described above, show a global average Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 71 Boutonnet of about 2 pptv (M ontzka et al., 1996; O ram et al., 1996). For reasons such as this the IPCC scenarios are not used in this assessment. Development of a Standard C oncentration of T FA T he possible future deposition rate of trifluoroacetic acid has been calculated using mathematical models of the lower atmosphere. In addition to the chemical reactions themselves, these models take account of the geographical distributions of: T he concentration of hydroxyl radicals with which the parent compounds react first of all T he emissions of the parent compounds T he rainfall that scavenges trifluoroacetic acid from the atmosphere (Kanakidou et al., 1995; Rodriguez et al., 1994). Because most of the rainfall and the highest regional concentrations of hydroxyl occur in the tropics, the mass deposition rate of trifluoroacetic acid is highest in this region. Starting from the scenario described in M cCulloch (1994), Kanakidou et al. (1995) calculate that the tropical concentration would be up to 2 molm-2yr-1 and that this would be equivalent to 1 nmoll-1 of rainwater. T he latter value would correspond to 100 ngl-1. O ver Europe the deposition rate in 2020 would be about 0.5 molm-2yr-1 and, over the U nited States, levels would be up to double this. T hese values are of the same order of magnitude as those calculated by Rodriguez et al. (1994), on the basis of a scenario from the U .S. Environmental Protection Agency; the latitudinal variation (up to a factor of 40) and longitudinal variation (a factor of 3) are also similar. In the year 2020, the modeled decomposition flux of H FC-134a was calculated to be 115 kT yr-1, with a further 20 kT yr-1 each of H CFCs 123 and 124. T he implied global deposition of trifluoroacetic acid then would be 155 kT yr-1. With such close agreement on potential future deposition rates, this quantity (rounded to 160,000 metric tons per year), and a corresponding rainwater concentration of 0.1 gl-1 (0.0001 mgl-1) are adopted as standards for this risk characterization. Accumulation of T FA in Aquatic Ecosystems T he objective of this section is to review whether conditions could occur in the environment which could lead to a significant accumulation of T FA in natural water bodies. It appears to be possible to accumulate a riverborne flux of T FA from "Normal" concentrations of a few hundreds of nanograms per liter to several thousands of nanograms per liter in receiving waters from which there is no outflow, only evaporation. In the case of the River Jordan/ Dead Sea system (T able 3), consideration of the volumes and flows indicates accumulation over many hundreds of years. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 72 Environmental Risk Assessment of T rifluoroacetic Acid Recently, the assumption has been made by T romp et al. (1995) that T FA can accumulate to very high concentrations in seasonal wetlands, which are special aquatic ecosystems that dry out periodically and are replenished only by rainfall. It is necessary to examine if the hypothesis developed in that work could occur in real conditions. T romp et al. (1995) conducted a sensitivitystudyof the potential for enhancement of T FA concentrations in such ecosystems. In the study, several assumptions were used concerning both the conditions that might lead to higher concentrations of T FA in rainwater (in comparison with expected average concentrations calculated by mathematical models; see previous sections) and the characteristics of the seasonal wetlands that would favor accumulation. A calculation presented in the paper showsthat a seasonal wetland, receiving rainfall for a total period of 30 years at a local (enhanced) T FA concentration of 1 mgl-1, with an evaporative loss rate of 5 yr-1, and a loss of T FA out of the wetland by seepage (or other physical or biological process) at a rate of 0.1 yr-1, would reach a concentration of 100 mgl-1 at the end of that period. H owever, this is simply a mathematical calculation that has been conducted by adding together the consequences of extreme conditions without regard for the probability that they could occur simultaneously. In the real environment it is highlyimprobable that all of the necessary conditions for accumulation will happen together and be maintained for decades. T he step-by-step review of those conditions carried out below indicates that the probability that such accumulation will take place over decades is close to zero. First, the seasonal wetland would need to be located near a large urban area, where the assumption is made that local conditions could lead to enhanced T FA concentration in rainwater. T hese local conditions would have to meet several criteria. 1. T he concentrations of precursors to T FA should be higher than average, due to local emissions. A factor of 10 to 20 is suggested in T romp et al. (1995), but observations in the L os Angeles area (notorious for local pollution) show an historic factor of about 3 for CFC-11 (Bastable et al., 1990). CFC-11 was used in the blowing of polyurethane foams and in whole building air conditioning. T he most significant T FA precursor is expected to be H FC-134a, which is used in refrigeration and air conditioning applications where smaller emissions factors are expected and further improvements in containment are likely in the coming decades. T he appropriate enhancement factor therefore is less than 3. 2. T he oxidizing capacity of the local atmosphere (including the O H radical concentrations) would be higher due to local pollution, thus increasing the rate of conversion of the precursors. A possible factor of 10 is quoted in T romp et al. (1995), but calculations based on actual observations of tropospheric ozone, NO x, water vapor, carbon monoxide, methane and non-methane hydrocarbons at Riverside, California in June Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 73 Boutonnet suggest that the O H field enhancement in that portion of the L os Angeles Basin does not exceed a factor of 2 (Ko et al., 1995). 3. It would be necessary to have an appropriate combination of pollution event and rainfall, the most effective rainout of T FA occurring when rain immediately followed a pollution event. T he implicit assumption in T romp et al. (1995) is that only the most effective rainout occurs and that it is repeated systematically throughout the year for decades. T he real situation, for example, in the L os Angeles Basin, is that rainfall happens preferentially in the winter months, when photochemical activity in the atmosphere is lowest (Ko et al., 1995). T he same authors pointed out that no rainfall event was observed during June 1990 and, in that case, T FA would be dispersed (on a continental scale) before being rained out. It is expected that pollution events would last more than one day, allowing precursor concentrations to increase; this would be possible under an inversion layer that trapped the gases within the "boundary" layer. O bservations show that such highly polluted conditions are associated with low wind speed. O n the other hand, rainout of the T FA formed within the pollution event needs a change in meteorological conditions bringing in a new air mass, with clouds and rain. For effective rainout of the T FA, this change in meteorology must not affect the stability of the boundary layer and must not disperse the T FA before rainout. It is not clear that such a mechanism could exist in the real atmosphere and, even if it did, it would be an exceptional event. An attempt to calculate the resulting increase of the average T FA concentration from uncorrelated rain and pollution events gives a value of the order 15% increase at most (Ko et al., 1995). T he probability of maintaining conditions that give total local rainout over a period of decades is therefore not far from zero. Secondly, the long-term accumulation in seasonal wetlands requires: 4. A closed system, maintained for decades, where seepage or loss by other physical processessuch asaeolian transport would be small. Vernal pools, examples of seasonal wetlands suggested to be most at risk, have been shown to have significant seepage (Sefchick et al., 1994). 5. An evaporation rate high enough to have a severalfold increase in T FA concentration each year. T romp et al. (1995) chose a value of 5 yr-1 for this parameter, but the largest evapoconcentration factor for vernal pools is 2.2 as reported by Sefchick et al. (1994) and inferred from chloride ion concentration that is indicated as the best conservative solute with sodium ion. It is also important to point out that such evapoconcentration factors reflect concentrations variations but not accumulation from year to year. Evapoconcentration factors vary with the type and location of the wetland and with the ion concentrated and thisobservation yields the most appropriate value. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 74 Environmental Risk Assessment of T rifluoroacetic Acid Each condition listed above, even taken alone, has little probability of occurring. For accumulation to take place, not only do the conditions have to occur together but this combination of conditions has to be maintained for decades. In conclusion, although some accumulation of T FA may take place in aquatic ecosystems like vernal pools, accumulation to high concentrations as described in T romp et al. (1995) appears to be highly improbable. Soil Adsorption T he degree to which T FA adsorbs to soils may affect its bioavailability to plants and bacteria, the general mobility of this compound through soils and aquifers, and, of course, the concentration to which T FA could accumulate in the soil. T he soil-water milieu may vary greatly in terms of mineral and organic compositions, as will the flux of cations and anions through these matrices. T herefore, if there is an interaction of T FA with specific components of the soil, this interaction may be site-specific and non-uniform in occurrence. Clearly, the problem of assessing soil adsorption in a controlled manner and in a way that is useful for developing a global perspective is very difficult. T wo laboratory studies were conducted to directly assess the partitioning of T FA between water and a variety of soil types under controlled laboratory conditions. T he first study by van Dijk (1992b) showed that T FA did not adsorb to soil. T he second study by Richey et al. (1997) showed that T FA generally interacted only weakly with most soils but was strongly adsorbed by some soils which contained high levels of organic matter. T he data are not necessarily contradictory but may reflect the heterogeneity of soils tested. T he studies are discussed in more detail below. T he first study examined the interaction of sodium T FA with three different soil types, obtained from locations in Northern Europe. T he salient properties of the soils from van Dijk (1992b) follow. San d Sandy L oam L oam % Clay 2.7 8.3 16.5 % Silt 6.4 27.8 46.8 % Sand 90.9 64.1 36.7 pH 5.5 6.5 7.5 % O rganic Carbon 4.3 1.3 2.6 T hese soils represent a relatively narrow range in terms of organic carbon and clay content; pH varied from mildly acidic to slightly alkaline. T he researchers tested a soil/ water mixture comprising 2 g soil and 10 ml of 10 M calcium acetate in mixture with 40 M sodium T FA. T he supernatants were assayed for T FA after 16 hours of incubation. T he results of the analyses indicated that <3% of the added T FA had been retained by the soil. T his was equivalent to partition coefficients (Kd, soil concn./ aqueous concn.) of <0.2 lkg-1. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 75 Boutonnet A second study was performed on 54 different soils from disperse environments located predominantly in North America and Europe (Richey et al., 1997). Rather than list the 54 soils individually, the ranges of important parameters are summarized here. pH % Clay % O rganic M atter 2.76 - 5.85 1% - 27% 1.4% - 93.3% In these experiments adsorption isotherms were determined for 1:5 or 1:20 soil:solution ratios (for high organic or high mineral soils, respectively) over a range of T FA concentrations: 0, 2, 4, 7, 10, 20, 30, 40 M NaT FA. T FA adsorption wasdetermined bythe difference in supernatant T FA concentrationsafter an equilibration period of 24 hours. O f the 54 soils tested, 20 soils showed no significant retention of T FA. O f the remainder, 23 gave Kd values <2 (high to intermediate mobility) and 8 soils exhibited low mobility of T FA (Kd =2 to 10). O nly three soils gave Kd values >10; the maximum Kd reported was 20 (60% retention) for a peat core containing 93% organic matter. At all sites where T FA retention wasdetermined for both organic and mineral soils, greater retention was observed for the organic soil. T FA adsorption was compared to that of other ions: bromide, chloride, fluoride, nitrate, and sulfate. T he results indicated that T FA was not the most strongly retained of the common ions with the relative order for a high (82.5%) organic soil as follows: fluoride (most strongly retained) >>sulfate > chloride >T FA >bromide >nitrate. For intermediate (24.8%) organic matter soil: fluoride =sulfate >chloride =T FA >bromide >nitrate. Anions were also shown to competitively inhibit T FA adsorption. For example, 150 M sulfate -- not an excessive environmental sulfate concentration -- caused 50% reduction in T FA adsorption. In comparing the two studies it is clear that a much larger representation of soil types was investigated in the work by Richey et al. (1997) and that the organic content of many of the soils in that work was strikingly higher. O rganic content was determined by Richey et al. to be in positive correlation with T FA adsorption. T he relatively low organic content of the soils in the van Dijk study may have led to the accurate conclusion for that study that T FA did not adsorb. O ne disturbing aspect of the van Dijk study was the use of 10 mM calcium acetate in the T FA equilibration mixture, use of such high concentration of a T FA analogue might have suppressed T FA binding. It should be noted that the T FA concentration used in both studies was 3 to 4 orders of magnitude higher than that anticipated in the environment in the year 2020. Field manipulation experiments were conducted at two sites at the H ubbard Brook Experimental Forest (New H ampshire, U SA). Field results generally agreed with laboratory studies, indicating that there is considerable variability in soil retention of T FA. T FA adsorption occurred mainly in the surface soil, where organic content was highest, and diminished with soil depth. Additions of T FA to an upland forest soil were largely transported with drain- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 76 Environmental Risk Assessment of T rifluoroacetic Acid age water (>70%). In contrast, additions of T FA were strongly retained in a wetland with <5% of the added T FA exiting the system in water. T he remainder of the T FA in the wetland was found in soil and plant tissues (L ikenset al., 1997). As in the laboratory studies of Richeyet al. (1997), T FA mobility in field plots (Berger et al., 1997) was found to be comparable with that of bromide, which is widely used as a conservative hydrologic tracer. It can be concluded that typical soils do not exhibit significant retention of T FA, but that some partitioning to organic-rich soils is observed. Although none of the investigators examined the reversibility of T FA binding, based on the maximum observed partition coefficient (Kd) of 20 (Richey et al., 1997), the maximum concentration bound to soil -- in equilibrium with the projected rainfall concentration from CFC alternatives of 0.0001 mgl-1 -- would be 0.002 mgkg-1. EXPO SU RE MEASU REMEN T S Aquatic C ompartment A number of groups have determined T FA in the aquatic environment at concentrations ranging upwards of a few nanograms per liter. Air, rain, and surface waters in Europe were extensively sampled and analyzed for trifluoroacetate during 1995 by Frank et al. (1996) and Frank and Klein (1997). T his built on previous work by the same authors. T he analytical methods were standardized and measurements made in air and rainwater samples during 1993 and 1994 (Frank et al., 1995). T he results of their determinations of the T FA concentrations in the Roter M ain river in 1995 are given in T able 1; Figure 1 shows the analyses for both 1995 and 1996. T here is no discernible seasonal trend. T able 2 shows the results of T FA determinations in river and lake waters in Europe and the approximate locations of the sample points. T he locations of European seawater, air, and precipitation sample points are shown in Figure 2. T able 1. T FA content of the Roter Main River at Bayreuth, G ermany (Frank et al., 1996). Date of Sample T FA (ngl-1) M arch 1995 April 1995 M ay 1995 June 1995 O ctober 1995 November 1995 December 1995 Average 160 27 80 140 110 280 110 60 140 Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 77 Boutonnet Figure 1. T FA in the river Roter M ain at Bayreuth. . Surface water in and around Reno, Nevada, showed far higher concentrations of trifluoroacetate using a comparable, but different, analytical method (Zehavi and Seiber, 1996). T he highest concentration that was recorded exceeded that in the Dead Sea (Frank et al., 1996) by a factor of 6. In both cases, if the samples of the receiving waters and the rivers feeding them are representative, accumulation over many hundreds of years would be necessary to account for the observations. T able 3 shows the concentrations reported for surface waters in Israel, Russia, Brazil, South Africa, and the U nited States. T able 4 shows the seawater concentrations recorded by Frank and co-workers for samples from the Baltic and North Seas and the Atlantic and Pacific O ceans. Figure 3 shows the sampling locations for all of the non-European surface water determinations, together with the locations of precipitation samples. T able 5 lists the results of determinations of trifluoroacetate in contemporary and ancient spring, mineral, and tap waters. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 78 Environmental Risk Assessment of T rifluoroacetic Acid T able 2. Present T FA concentration in lakes and rivers in Europe (Frank et al., 1996; Frank and Klein. 1997) L ocation Sample date R. Elbe (Wittenberg, Germany) R. Danube (Regensburg, Germany) R. Pegnitz (Nrnberg, Germany) R. M istelbach (Bayreuth, Germany) R. M ain (M ainleus, Germany) R. Neckar (T bingen, Germany) R. Elbe (H amburg-Altona, Germany) R. Weiser (Bremerhaven, Germany) R. Rhine (Duisburg-Ruhort, Germany) R. Warnow ( Rostock, Germany) R. Nabb (Schwandorf, Germany) R. Seine (Paris, France) R. L oire (Nantes, France) R. Rhine (Bregenz, Austria) R. Kemiojki (T rytie, Finland) L ake Fichtel (Fichtelberg, Germany) L ake Weienstadt (Weienstadt, Germany) L ake Constance (Constance, Germany) L ough Skannive (Ireland) L ough Ahalia (Conemara, I reland) 24 Feb 95 25 M ar 95 26 M ar 95 24 M ar 95 3 Apr 95 23 Apr 95 7 Jul 95 6 Jul 95 6 Jul 95 9 Jul 95 9 Jul 95 Jul 95 Jul 95 14 Sep 95 15 Jul 96 17 M ay 95 17 M ay 95 14 Sep 95 1 Nov 95 1 Nov 95 a S.D. is the Standard Deviation of n analyses. T FA determination ng.1-1 n S.D .a 200 3 10 200 3 2 150 3 3 215 3 9 40 3 16 260 3 5 100 5 14 100 4 7 630 4 15 60 4 25 130 3 14 40 330 55 3 9 210 5 20 70 3 25 115 2 100 60 3 8 20 3 21 <10 3 -- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 79 Boutonnet Figure 2. Approximate locations and concentrations of environment T FA measured in Europe Air and precipitation samples: B, Bayreuth; G , Gdansk; MG , M rmo Glacier; and MH , M ace H ead. Seawater determination are shown in italic; rivers and lakes are in normal print. While the whole of the world has not been covered by the sampling program, it is apparent that trifluoroacetic acid is a ubiquitous contaminant of surface waters over a large part of the earth's surface. It is also present in air and rain samples (see subsequent section) and in samples of old ice. T he determinations in enclosed lakes in Nevada and Israel that could indicate long-term accumulation, over several hundred years, would point to a source that is not just a present day industrial activity. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 80 Environmental Risk Assessment of T rifluoroacetic Acid Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 81 Boutonnet Figure 3. Approximate locations and concentrations of environment T FA other than in Europe. Air and precipitation samples: D, Davis, California; MC , M ount Cook, New Zealand; Q ML , Q ueen M aud L and, Antarctica; R, Resolute, Canada; and Reno, Nevada. Seawater measurements are shown in italic; rivers and lakes are in normal print. T errestrial C ompartment T here are no data on background concentrations of T FA in soil or on soil surfaces. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 82 Environmental Risk Assessment of T rifluoroacetic Acid Atmosphere O ver the same period of time that they were determining trifluoroacetate concentrations in surface waters, Frank and co-workers conducted an extended campaign to sample and analyze air and precipitation in Bayreuth, Germany. T he results are shown in T able 6 for 1995. Figures 4 and 5 show the air and precipitation analyses for both 1995 and 1996. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 83 Boutonnet Values determined during the earlier campaign, in 1993 and 1994 when the method was being developed, were of a similar order, ranging from 18 to 620 pgm-3 for air ( with one outlier at 3230 pgm-3) and <3 to 120 pgg-1 ( ngl-1) for rainwater (Frank et al., 1995). T he determinations are consistent with the levels observed in surface waters (see previous section). T he observation that concentrations in air under the tree canopy in a forest are consistently greater than concentrations in a clearing ( reproduced in Figure 6) was interpreted by Frank and Klein (1997) to indicate that the forest canopy was involved in the deposition mechanism. Analyses have also been performed on air and precipitation samples from the western U SA (Zehavi and Seiber, 1996) and on samples of precipitation from a wide range of remote locations in the Northern and Southern H emispheres (Grimvall et al., 1997). T he data are recorded in T able 7. Similarly to their surface water determinations, Zehavi and Seiber (1996) reported levels of T FA in rain and fog that were high but consistent with the measurements of Frank et al. (1996). Grimvall and co-workers have shown not only that T FA is widely distributed -- from Arctic to Antarctic -- at low concentrations, but that it was present in precipitation in the Swedish Arctic some 400 years ago. T he levels in old ice from the M rmo Glacier are similar to present day concentrations. Given the lower limit of detection of Grimvall's method, the results in T able 7 are consistent with those in T able 5. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 84 Environmental Risk Assessment of T rifluoroacetic Acid Figure 4. T FA in air near Bayreuth from M arch 1995 to O ctober 1996. T he width of each column represents the duration of the sampling; from November 1996 the length of a typical sampling period was 5 to 7 days. T he significance of the discovery of low concentrations of T FA in precipitation in remote areas is that it indicates a geographically dispersed source. With an atmospheric lifetime (towards rainout) of a few weeks, it would not be possible for T FA that is generated in the northern hemisphere to be present in southern hemispherical precipitation. T here are too few results to map the deposition of T FA definitively, so that a global mass balance is not possible. It is, however, instructive to examine the actual deposition in a well characterized location, such as Bayreuth, and compare it with the deposition from known sources predicted by atmospheric models. G lobal Measurements and T heir Significance for Future Emissions As described in a previous section, the current potential sources of trifluoroacetic acid would provide, at most, 2800 metric tons per year distributed throughout the world. It is also apparent, from the atmospheric modelling that has been carried out to assess the possible future deposition of trifluoroacetic acid, that this quantity will not be deposited evenly over the globe but will fall preferentially on tropical regions. M odels of future deposition can be used to geographically distribute the current contribution. Kanakidou et al. (1995) calculated that, from a global flux of 160,000 metric tons per year of tri- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 85 Boutonnet Figure 5. T FA in precipitation near Bayreuth from M arch 1995 to O ctober 1996. Expanded scale from M ay to September 1996. fluoroacetic acid, the deposition rate in Europe would be 0.5 molm-2yr-1; scaling the current contribution to this would imply a deposition rate of 0.5 2800/ 160,000 or 8.8 nmolm-2yr-1, equivalent to 1 gm-2yr-1. T he annual rainfall at N urnberg, near Bayreuth, is 620 mmyr-1 (van der L eeden, 1975) so that the expected concentration in rainwater there from current sources is 1/ 0.62 gm-3 or 1.6 gm-3, which equates to 1.6 ngl-1. T he measured concentration at Bayreuth averages 100 ngl-1 so the expected concentration of trifluoroacetic acid in rain from known sources is less than1/ 60th of that observed. DEG RADAT IO N O verview T he ionic character of T FA precludes simple partitioning of the molecule into fatty tissues of animals. As mentioned previously, it is expected that the compound will be found in aqueous compartments such as lakes, rivers, oceans, estuaries, and associated sediments. T he steadystate levelsof T FA within these compartments will depend on the ambient input concentrations of Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 86 Environmental Risk Assessment of T rifluoroacetic Acid Figure 6. T FA in precipitation at a forest site and nearby clearing (1-km south of the U niversityof Bayreuth). T he higher column of each pair alwaysrepresents the T FA contents in the precipitation collected under the spruce tree canopy, including tree canopy runoff. T FA in precipitation and T FA removal processes. T he ultimate fate of T FA will largely depend on four processes: A. Biodegradation B. Abiotic mineralization C. Accumulation or uptake of T FA by organisms (see Accumulation Section) D. Adsorption (see earlier section) Processes A and B are truly degradative and contribute to a net loss of T FA from the system; they will be the focus of discussion in this section. Process C may result in transformation of T FA into a biological constituent molecule, a fluorinated metabolite of T FA, or simply temporary sequestration. Process D is not true removal of the T FA from the environment but rather sequestration of the compound onto, for example, soil particles, resulting in less mobility or bioavailability. Biodegradation Reductive dehalogenation is well known to be an effective natural mechanism for destruction of chlorinated hydrocarbons. H owever, defluorination by this mechanism is thought to be much more difficult due to the lower reduc- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 87 Boutonnet Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 88 Environmental Risk Assessment of T rifluoroacetic Acid tion potential of the C-F bond (-2 volts). M onofluoroacetate (M FA) is known to be degraded hydrolytically by enzymes generically known as fluoroacetate dehalogenases. In fact, microorganisms containing these enzymes have been isolated and shown to hydrolyze M FA with release of fluoride. H owever, these organisms do not degrade difluoroacetate (DFA) or trifluoroacetate. T his is consistent with the relative stability of these compounds (T FA>>DFA>M FA) (Emptage, 1994). T FA may be considered to be an analogue of a central biological metabolite -- acetate. T ransformation of T FA byenzymatic systemsspecialized for acetate metabolisms might be another route for T FA transformation resulting either in mineralization or incorporation into cellular material. H owever, rates for T FA transformation by this route have been estimated using purified enzyme systems and found to be less than 0.5% of the rate of acetate (Peijnenburg et al., 1994), which suggests that T FA will not compete effectively with acetate in biological systems. Demonstration of T FA Biodegradation T he studies summarized below represent a variety of empirical approaches to demonstrate T FA biodegradation in field- or laboratory-derived microbial cultures. M easurements Determination of the biodegradative potential of microorganisms for a specific compound is typically carried out using one transformation, or lack of transformation (of T FA), which can be inferred from one or more of the following analyses: T FA-dependent microbial growth, T FA-dependent oxygen consumption, appearance of defluorination or decarboxylation products such as methane or carbon dioxide using [1-14C]-labeled T FA, fluoride anion, and direct demonstration of T FA loss. M icrobiological Approaches 1. Enrichment for degradative bacteria using the compound in question (T FA) as the sole carbon and energy source in the presence of added nutrients such as ammonium or phosphate. In this case bacteria, which can utilize the compound, will grow preferentially. 2. M icrobial growth on standard carbon and energy sources is independent of the test compound; however, biodegradation of the test compound is dependent on growth and metabolism of the bacteria on natural or standard carbon sources. Standard biodegradation tests employing organicrich inocula such as digestor sludge is one example of this approach. T his is a co-metabolic mechanism where transformation of the test compound does not result in any specific advantage for the organism or does it result in any enrichment of bacteria that specifically degrade the test compound. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 89 Boutonnet 3. Intrinsic metabolic capability of a naturally occurring sample. T ransfor- mation of the test compound is assessed in a naturally occurring sample without any additional nutrient or carbon source. A naerobic Degradation of T FA : R eductive Defluorination Reductive processes are expected to be operative under anaerobic or highly reducing conditions such as those referred to as "methanogenic" or "sulfatereducing." Anticipated products from the reductive defluorination of T FA are ultimately methane, fluoride, and carbon dioxide. Early studies using unamended marine sediment (San Francisco Bay) or freshwater sediments incubated with 14C-labeled T FA clearly demonstrated a reductive defluorination process resulting in sequential formation of DFA, M FA, acetate and methane (Visscher et al., 1994). T his sequence was observed under methanogenic and to a lesser degree sulfate-reducing conditions. O ther respiratory acceptors such as nitrate, ferric iron or oxygen failed to produce defluorination although some fluoroform was reported under aerobic conditions. T hisphenomenon hasonlybeen observed in certain field samplesbythe one laboratory. T he results also showed that T FA inhibited methanogenesis at low concentrations (10 M ) and that at higher concentrations of T FA, the compound could inhibit its own biodegradation. Reinvestigation of samples from the same field sites ( by the same laboratory) failed to confirm the initial results (M atheson et al., 1996). Another laboratory reported that marine sediments from the sites that had previously been shown to be active by Visscher et al. (San Francisco Bay) for methanogenic T FA biodegradation were unable to degrade T FA (Emptage et al., 1997). T he initial results of Visscher et al. have not been replicated despite numerous attempts by the original investigators as well as others. Chauhan et al. (1995) carried out anaerobic closed microcosm bottle incubations inoculated with soils obtained from globally diverse sites as well as anaerobic digestor sludge. T hese were assessed for transformation of [2-14C]labeled T FA to 14CO 2. T hese tests showed no evidence for biotransformation with either nutritionally amended or unamended bottles. An assessment of anaerobic biodegradation by using the "sequential column microcosm" was also conducted by the same group. T his approach involves a succession of compartmentalized but interdependent microbial communities that are distinguished on the basis of mode of respiration. T he sequence starts with the most reducing conditions (i.e., methanogenic) and progresses toward more oxidizing physiologies by separate additions of sulfate, iron, nitrate, and ultimately oxygen into the sequence of chambers downstream of the initial methanogenic chamber. In this way a broad cross-section of microbial metabolism may be individually assessed in one experiment. [1-] or [2-14C]-labeled T FA input and 14C-labeled T FA output as well as total radioactive counts were monitored for each chamber and inferred from the loss in steady-state levels of T FA. I n addition, fluoroform and 14CO 2 were assayed in the methanogenic chamber. Results from these experiments showed sustained Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 90 Environmental Risk Assessment of T rifluoroacetic Acid and long-term loss of 25 to 30% of the incoming T FA in the methanogenic chamber. O nly very minor amounts of 14CO 2 were detected in the headspace of the culture vessel. H owever, total countsof radioactivityexiting the chamber were only slightly diminished. T he results suggest that either the T FA was complexed such that it did not migrate on the H PL C column and was thus invisible by this method, or that it was transformed to another soluble aqueous phase metabolite that was not recovered in the H PL C analysis. No losses were noted for subsequent chambers involving sulfate reduction, iron reduction, nitrate respiration, photosynthetic sulfide respiration, or aerobic dark respirations. No definite product was ever identified from the putative methanogenic transformation; however, a conversion of T FA to a nonvolatile product was hypothesized to account for the losses. T he inference that some biotransformation occurred based on loss alone must be viewed cautiously until direct evidence for the "nonvolatile product" can be obtained. A erobic Degradation of T FA : Decarboxylation Compelling evidence for decarboxylation of T FA was obtained using pure cultures of A zoarcus tolulyticus tol-4, or P seudomonas putida strain KZ6R. T he researchers demonstrated release of labeled carbon dioxide from T FA labeled in the 1-position (Chauhan et al., 1995). Decarboxylation activity was strictly dependent on the substrates on which the cells were grown. A . tolulyticus had to be pre-grown on toluene as the carbon source with nitrate as the electron acceptor, whereas P . putida had to be aerobically pre-grown on 4-chlorobenzoate. Neither organism exhibited the activity during its normal growth on these substrates but rather only after harvesting the cells, concentrating them and resuspending them in fresh medium. T he T FA decarboxylation activity of the cells was dependent on the age of the culture and the metabolic state of the cells. T he resting cells rapidly lost the ability to decarboxylate T FA. H owever, the cellswere capable of transforming up to 16% of millimolar levelsof T FA during the optimal period of activity. T he exacting nutritional requirements, and the instability of the activity argue against extrapolating this observation to the natural environment outside the laboratory. It is not known at this time how effective the transformation is at very low levels of T FA anticipated in the environment. It is also unclear as to what the biochemical mechanism is other than the correlation with aromatic catabolism. A erobic C losed Bottle T ests using Diverse I nocula T he transformation of [1-14C] T FA was assessed in aerobic microcosms inoculated with composite samples consisting of a blend of soils from geographically diverse regions of the globe, digestor sludges, or organic rich soil fractions from sites in M ichigan (Chauhan et al., 1995). No evidence for aerobic mineralization was obtained from analysis for 14CO 2 in these incubations. A semicontinuous activated sludge test was conducted by van Ginkel and Kroon (1992a,b) according to EEC, O ECD, and ISO test guidelines. T he pro- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 91 Boutonnet cedures assessed fluoride release from aerobic digestor sludge incubations that included co-metabolic substrates yeast extract, acetate or peptone and vitamin B12. T he semi-continuous sludge test demonstrated that T FA had no effect on the efficacy of the sludge as a biodegradation agent. No fluoride was detected in the effluent from the SCAS units even though there was an approximately 20% reduction of T FA in the effluent (van Ginkel and Kroon, 1992a). Closed bottle tests were also performed. T hese incubations typically lasted for periods of up to 84 days after which fluoride was measured. No detectable fluoride was observed in the bottle tests. T he tests were considered negative for T FA biodegradation (van Ginkel and Kroon, 1992b). A erobic Biodegradation of T FA by M onooxygenase-C ontaining Bacteria Certain bacteria contain enzymes (monooxygenases) that are capable of inserting oxygen into aliphatic and aromatic hydrocarbons. By virtue of this, these bacteria are capable of aerobic degradation of a wide variety of chlorinated hydrocarbons such as trichloroethylene (T CE), dichloroethylene and vinyl chloride. T o test for the ability of monooxygenases to degrade T FA, a set of methylotrophic and propanotrophic bacteria were grown and tested against compounds known to be degradable (i.e., T CE) as well as T FA. DeFlaun (1996) tested nine strains comprising P seudomonas mendocino KR1, P seudomonas putida F1, M ethylosinus trichosporium, M ycobacterium vaccae, as well as five isolates (designated ENV2C, ENV2D, ENV2R, ENV2W, and ENVO B) that were known to degrade H FCs or H CFCs. T hese strains were tested against [1-14C] T FA with T CE as a positive control (Deflaun, 1996). Assays were performed for 14CO 2, fluoride, M FA, and T FA. T he results showed that even with cultures which vigorously degraded T CE within 24 hours there was no detectable decarboxylation or dehalogenation of T FA after 13 days. No M FA or fluoride was detected in any incubation. Abiotic Mineralization I rradiation of natural waters by sunlight can produce reactive species such as H O , RO 2, and O 2H and solvated electrons. T he only ones capable of oxidizing or reducing T FA are H O and solvated electrons (eAq-). H owever, the concentrations of these species in natural waters are likely to be very low and their rates of reaction with T FA, together, do not support a role for these agents in the destruction of T FA (M aruthamuthu et al., 1995; M aruthamuthu and H uie, 1995). O ne possible photocatalytic mechanism involving ferric- or manganeseassisted photodecarboxylation ( photo-Kolbe) reactions was shown to be effective for degradation of chloroacetates and mono- or difluoroacetates but not for trifluoroacetate ( M aruthamuthu and H uie, 1994) . T rifluoroacetic acid has been shown to undergo photooxidation on a variety of iron oxyhydroxide surfaces (Pehkonen et al., 1995). Similarly with the Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 92 Environmental Risk Assessment of T rifluoroacetic Acid other fully halogenated acetic acids, this reaction proceeds by a photo-Kolbe mechanism to yield H F and CO2, but the rate is so low that it is unlikely to have much environmental relevance. T FA was shown to be inert to photocatalytically generated oxidizing species on semiconductor surfaces. T FA was also shown to be inert to redox- and radical-induced degradation except at pH s above 10.5 (Asmus et al., 1994). I n a study of the rates of reductive dehalogenation versus structure, t1/ 2 times for reductive transformation of a series of chlorocarbons as well as T FA (based on structure-activity relationships) were obtained under anaerobic conditions in the presence of sediment. T he reductive dehalogenation reactions were presumably microbial in origin, but this was not directly demonstrated in the study. I n contrast to the fully chlorinated ethanes and ethenes, which are readily dehalogenated, the t1/ 2 for T FA -- obtained by extrapolation -- was found to be over 200 years in these biologically active sediments (Peijnenburg et al., 1994). T hese results are consistent with those obtained in other laboratories investigating the reductive dehalogenation of T FA but contrast with the report by Visscher et al. (1994). Due to the recycling of T FA between water and air and the rapid rainout of T FA from clouds (10 days), it is unlikely that there will be any aqueous phase degradative process that could compete with rainout kinetically. O xidation of T FA by N O 3, SO 4-, Cl2-, O H followed by decarboxylation may occur with a t1/ 2 on the order of 80 years. Ferric-assisted photodecarboxylation was shown to be ineffective for T FA (Wine, 1994). C onclusion T FA is a highly stable molecule to known abiotic or biological degradation mechanisms. T he microbiological studies summarized here involved both empirical screens of naturally occurring soils as well as focused tests of known degradative mechanisms. T FA biodegradation was investigated in aerobic and anaerobic bacterial cultures with pure isolates as well as crude environmental samples. T hese studies have not identified evidence for a widespread, environmentally significant, biological mechanism for defluorination of T FA. T herefore, the assessment is that T FA will be very long-lived in the environment. T his prediction suggests that renewed emphasis be placed on identifying the background levels of T FA in the environment as well as how this T FA may compartmentalize. Despite published data claiming rapid reductive defluorination of T FA (Visscher et al., 1994), the evidence for this mechanism of T FA removal must be considered hypothetical until the original result can be further confirmed or supported. T he weight of theoretical considerations as well as the combined experience of the five laboratories involved in these projects suggests that confirmation is unlikely. T his does not rule out the possibility that the result was correct but due to a rare concatenation of physical and biological factors. T he result reported by Chauhan et al. (1995) that clearly demonstrated decarboxylation, but not defluorination, of T FA maybe of limited environmental Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 93 Boutonnet significance as a mechanism for destruction of T FA in the environment. Experience with a wide variety of real environmental samples by that laboratory, as well as other laboratories, suggest that decarboxylation of T FA is not widespread in nature or is it easily induced where it is present. Indeed, even the laboratory strains only produce the activity under very controlled conditions. T he fate of the CF3- moiety in the decarboxylation reaction is unknown. T herefore, for the purpose of this risk assessment, we will assume, as a reasonable worst case, that no degradation of T FA is likely to occur in the environment. ACCU MU LAT IO N T he physico-chemical properties of T FA (high solubility and low Kow) indicate that the molecule would not partition significantly to the lipid phase of cells and therefore significant bioconcentration by this mechanism would not be expected. H owever, other mechanisms that might lead to a bioaccumulation of T FA have been investigated. Due to the structural similarity of T FA to acetate, a major biosynthetic precursor molecule, studies were performed to assess the level of cellular incorporation of T FA into biomass. Bott and Standley (1994) investigated the incorporation of [2-14C] T FA into microbial cell mass, small invertebrates, and plants. T he cellular components were assayed and comprised the lipid, carbohydrate, and proteinaceous fractions. Aquatic communities, including bacteria, aquatic invertebrates, and plants, were set up in flowing water mesocosms to provide a source of biological samples for experimentation. O ne of these communities was exposed to NaT FA (30 gl-l) for nearly 2.5 years; the other was an unexposed control run for a comparable period of time. Results showed that sediment microbial communities exposed to T FA in mesocosms for approximately 2.5 years incorporated more radiolabeled T FA than communities exposed for <1 year, with rates increasing from -1 10-l3 to 22 10-13 g T FAcell-lday-l ( Bott and Standley, in review). T he cause of the change, whether due to a microbial adaptation to the T FA exposure or a change in some other environmental variable(s) is unclear. Incorporation was correlated positively with the amount of T FA added to samples and exposure time. M icrobial incorporation of radiolabeled T FA by cells cultured from those sediments ( at a test concentration of 10 gl-l) demonstrated incorporation into biomolecule fractions such as lipids, proteins and residual (e.g., cell wall) materials. I ncorporation into these fractions in bacteria increased from 0.005% of total counts to 0.024% over this 2.5 year period ( Bott and Standley, 1994; Standley and Bott, in review) . O ligochaete worms exposed to 40 g N aT FAl-l also showed low uptake of T FA into biomolecules. T he oligochaetes contained radiolabel that was not extractable as T FA; it was primarily associated with the protein fraction. T he presumptive incorporation product from T FA has not been identified. Jewelweed also contained radiolabel that was not extractable as T FA, but the levels of label incorporated ranged from Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 94 Environmental Risk Assessment of T rifluoroacetic Acid 0.0069% of the total exposure for roots to 0.012% of the total for leaves after a 32-day exposure. Apparent incorporation factors were low for other macroinvertebrates and plants. T hese studies suggest that microbial biomass is not a major sink for T FA, but that there is an association of T FA with biomass that could be due to biochemical incorporation. As part of a study to determine the possible defluorination of T FA by the freshwater green alga, Selenastrum capricornutum, van Dijk (1996) reported a bioaccumulation factor (BCF) of approximately 10, based on radiolabeled residues in cells filtered from the medium. H owever, the study did not correct for any radioactivity from the medium that may have been retained on the filter due to evaporation or adsorption and therefore may have overestimated the radioactivity associated with the cells. In a toxicity study with L emna gibba (duckweed, a floating aquatic higher plant) described in another section, 14C-labeled residues of T FA were determined after 7 days exposure to a range of concentrations in the culture medium (Smyth et al., 1993). T he bioconcentration factors were low, ranging from 1.0 to 1.6, reflecting the hydrophilic properties of the substance and the lack, in this species, of a differentiated water transport (transpiration) system. In terrestrial higher plants, the uptake and transpiration of water, to replace that lost by the leaves as a consequence of photosynthetic gas exchange, provides a mechanism for the bioaccumulation of T FA. Prior to the AFEAS program, root uptake and transport of T FA in the xylem tissues had been demonstrated in tomato seedlings (Rollinset al., 1989), using nuclear magnetic resonance imaging. H owever, the exposure concentration was very high (approximately 51,000 mg H T FAl-1 as a buffered solution) and the exposure duration was too short ( 17 hours) to evaluate the bioaccumulation potential. Subsequently, sunflower (H elianthus annuus) seedlings were exposed to a single concentration (2 g H T FAl-1) of 14C-radiolabeled trifluoroacetic acid in the aqueous (hydroponic) medium surrounding the roots (T hompson et al., 1994) . Plants were removed at intervals and the concentration of 14C residues determined in root, stem, and leaf tissue. After 12 days the remaining plants were transferred to clean medium and sampled for a further 4 days. 14C residues in the leaf tissues increased continuously over the period of exposure, with a BCF of approximately 22 after 12 days. T he stem tissue behaved similarly but with a lower rate of accumulation ( 12-day BCF of approximately 5) . Root tissue reached apparent equilibrium after 5 days exposure with a BCF of approximately3. T he accumulation rate wassomewhat lessthan would be expected from passive influx in the transpiration stream without efflux. T he concentrations measured in the root medium also suggested either a concurrent efflux or some partial barrier to influx. All tissues showed a decline in 14C-residue concentrations on transfer to clean medium. Although this can be attributed largely to dilution by growth, a significant quantity of radiolabel was lost to the medium surrounding the roots and it was concluded that the plants showed some excretion (depuration) of accumulated radiolabel to the root medium. M ore than 80% of the 14C residues in the leaves were found to be extractable in water after tissue maceration Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 95 Boutonnet (T hompson and Gillings, 1996). Fractionation of the extract using ion chromatography showed the residues co-eluted with a trifluoroacetate standard spiked into leaf extract, suggesting that no significant metabolism of T FA had occurred. As part of a 5-week hydroponic toxicity study with wheat (T hompson et al., 1995) (described in a subsequent section), the accumulation of 14C residues of trifluoroacetate was also monitored. At 1 mg N aT FAl-1, the bioconcentration factor in the total aerial tissue (shoots), based on fresh weight, increased continuously over the exposure period, to a final value of 27. T he concentrations in the tips of the shoots were approximately four times greater than in the remaining shoot tissues. T his factor was used to estimate a tissue concentration of approximately 1000 mgkg-1 (fresh weight) corresponding with tissue necrosis observed in the shoot tips at 10 mg N aT FAl-1, after 28 days. After 35 days, a BCF of 43 was determined for the whole shoots. Davison and Pearson ( 1997) analyzed tissues of soya and wheat exposed to NaT FA by determination of inorganic fluoride, after fusion with sodium carbonate to cleave any C-F bonds. Concentrations were expressed as NaT FA in dry tissue, after background fluoride subtraction. For soya exposed to 5 mgl-1 under hydroponic conditionsfor 44 days, the concentration in the tissue of the oldest (first trifoliate) leaves reached a plateau after approximately 20 days. Each successive set of leaves attained higher tissue concentrations, except for the fifth trifoliates that were still expanding at final harvest but already contained a higher concentration than had been reached by the older, second trifoliates. As symptoms first appeared on the third trifoliate leaves, the tissue concentration was approximately 150 mg N aT FAkg-1 dry weight. I t was concluded that the stage of development of the leaf is important and that T FA is most toxic to young, expanding leaves and has little effect on mature leaves with a similar tissue concentration. I n a further study with soya at 1 mg N aT FAl-1, Davison and Pearson (1997) found that the slight symptoms observed were associated with tissue concentrations of 160 to 190 mg N aT FAkg-1 dry weight. T issue concentration in the whole shoots of wheat exposed to the same concentration (which caused no effects) reached a plateau at approximately 55 mg N aT FAkg-1 dry weight after 25 days and had decreased slightly after 43 days. At 5 mg N aT FAl-1, the tissue concentration was approximately 190 mg T FAkg-1 dry weight when growth was beginning to be inhibited. T he final leaf/ shoot tissue concentrations(on a dryweight basis) for the different hydroponic studies are summarized in T able 8. For comparison, these are also expressed as bioconcentration factors based on tissue fresh weights, although it should be noted that fresh weights were not determined (and are therefore estimated) for some of the studies. Different analytical methods were also employed (see above). At concentrations at or below the no effect level of 1 mgl-1, bioconcentration factors ranged from 5.4 to 27. Frank (1994) reported concentrations of T FA in spruce needles sampled from T bingen, Germany, of 98 to 195 ng.g-1 (presumed to be based on fresh weight). Recent measurements(Frank et al., 1996) suggest an average concentration of approx- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 96 Environmental Risk Assessment of T rifluoroacetic Acid imately 100 ngl-1 in rainwater, suggesting a bioconcentration factor of 1 to 2. Although conifer needles are relatively long lived, this may indicate that the relatively low water usage of such leaves may enable excretory translocation to become more significant. T here is good evidence to suggest that the toxicityof T FA to plants is related to the tissue concentration of the accumulated material that is dependent on the transpiration flow from the roots. H owever, because the mode of action appears to involve specific processes in the leaf, possibly related to leaf expansion, it is not possible to specify precisely the critical tissue concentrations in terms of the whole leaf. H owever, the lowest tissue concentration associated with adverse effectswas150 mgkg-1 dryweight, in expanding soya leaves(Davison and Pearson, 1997). It would appear that when exposure is at or below the Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 97 Boutonnet threshold for effects, the tissue concentrations tend to plateau and accumulated T FA isreduced bygrowth and byexcretion via the roots. L eaf-fall represents an ultimate removal mechanism for any residual T FA in mature leaves. EFFEC T S ASSESSMEN T M icroorganisms Effect of T FA on M ethanogenic C ommunities A vital part of the carbon cycle involves the formation and degradation of acetate during anaerobic methanogenesis. T his process occurs globally but can be differentiated in terms of specific environments. Commercially or environmentally important methanogenic populations are found in anaerobic digestors, sediments, and rumen of ruminant animals. T he principal microorganisms are generically referred to as methanogens, many species of which specialize in the transformation of acetate to methane and carbon dioxide. M ethanogenesis was assessed for any toxic effects of T FA by measuring the rates of methane formation from anaerobic digestor samples, sediment samples, and rumen samples in the presence of increasing concentrations of T FA up to 10 mM (1114 ppm) according to Emptage et al. (1997). T he results indicated that there was no significant effect of T FA, even at the 10 mM concentration, on any incubation. M onofluoroacetate did however exert toxicity, as expected, at levels as low as 10 mM . T hese results contrast with those reported by Visscher et al. (1994), which suggested inhibition of methanogenesis at low concentrations (0.1 mM T FA). T he discrepancy in the results cannot be accounted for at this time except to note that the toxicity was observed in the same sedimentswhich exhibited rapid T FA biodegradation (see a previoussection). T FA biodegradation was not observed in the study by Emptage et al. (1997). O verall, the results suggest that T FA is inert in these systems and that endogenous methanogenesis was neither stimulated nor inhibited. Effect of T FA on A ctivated Sludge A semicontinuous activated sludge test was conducted by van Ginkel and Kroon (1992a) according to EEC, O ECD, and ISO test guidelines. T his test also indicated that T FA had no discernible effect on the performance of the sludge for catalyzing biodegradation of organic carbon. Effect of T FA on A cetate M etabolism by M icrobial C ommunities Acetate is a key intermediate in most living organisms. Ecologically, the mineralization of acetate to carbon dioxide is a key link in the biogeochemical carbon cycle. T herefore, it is essential to know whether T FA, which is structurally close to acetate, could interfere with acetate metabolism. Experiments were conducted to assess the interaction of T FA with the metabolism of [1-14C] acetate by freshwater microbial communities which had not been previously exposed to T FA (Bott and Standley, in review). T he approach comprised additions of non-labeled T FA or non-labeled acetate to parallel in- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 98 Environmental Risk Assessment of T rifluoroacetic Acid cubations containing [1-14C] acetate. Evidence for interaction with metabolism of acetate would then be inferred from the effects observed on incorporation of radiolabel into biomass or evolution of 14C carbon dioxide. According to Bott and Standley ( 1994) , none of the differences in 14C acetate metabolism between the unamended control and any of the T FA or acetate treatments were statistically significant (Dunnetts's test, p >0.05). Between 85 and 90% of the [1-14C] acetate radioactivitywas found in the biomass and there was no significant difference between controls and T FA-treated incubations. T hese results suggest that T FA at concentrations several orders of magnitude higher than those anticipated in the environment did not impact acetate mineralization to carbon dioxide or did it affect incorporation of acetate into cellular material. Similar experiments were performed using samples derived from T FAtreated ( 30 gl-1 3-months exposure) or control mesocosms. Experimental T FA additions were 200, 20, and 2.4 gl-1. Cell-specific rates of acetate mineralization rates were similar for samples from the T FA-treated and control mesocosms. With samples from each mesocosm there were never statistically significant differences in [ 1-14C] acetate metabolism over the range of T FA concentrations tested. T he percentage of radioactivity in the biomass ranged between 80 and 90% and there were no significant differences between unamended controls and T FA-treated samples. At extraordinarily high nonradioactive acetate additions of 4 to 450 mgl-1, there was suppression of 14C acetate metabolism, as would be anticipated from expected isotopic dilution of 14C acetate. At 440 mgl-1, T FA also suppressed 14C acetate metabolism, although to a lesser extent than did cold acetate suggesting some weak interaction of T FA with acetate metabolism at these concentrations. T hese competition experiments using benthic microbial communities also clearly demonstrated that T FA and acetate were not strongly competitive compounds in terms of mineralization or incorporation into biomass. T he effect of T FA has been investigated in three species of free-living nitrogen-fixing bacteria (Nagel and O dom, 1997). T he species were selected on the basis of phylogenetic diversity and because they have been used in many laboratory studies on the biochemistry of nitrogen fixation. A common aerobic soil microorganism (A zotobacter vinelandii), a freshwater photosynthetic bacterium (R hodobacter capsulatus), and a common anaerobe (C lostridium pasteurianum) were the test species. T he effect of T FA on these species was determined during nitrogen-dependent growth, growth on fixed nitrogen (ammonium ion), and on nitrogenase activity itself. T he experiments were designed to determine whether T FA wasspecificallytoxic to nitrogen fixation versusa more general physiological aspect. N o effect of T FA on growth either by N2 fixation or with ammonium ion as nitrogen source was noted even at concentrations as high as 1 mM (100 ppm) T FA with either the R . capsulatus or A . vinelandii strain. C . pasteurianum was only tested for T FA toxicity during growth on fixed nitrogen due to difficulties in obtaining significant rates of N2 fixation with this strain. Increases in production of molecular hydrogen were noted at very high Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 99 Boutonnet T FA concentration (>50 ppm) in R hodobacter capsulatus. T his effect could be due to either stimulation of nitrogenase activity or inhibition of hydrogen reutilization at these very high T FA concentrations. Aquatic O rganisms O nly acute toxicity studies are available; unless otherwise stated these were carried out with NaT FA. A number of studies employed radiochemical analysis of the test solutions, which confirmed the nominal concentrations and demonstrated no significant decline during the test period. In view of this and the known stability of the compound and low potential for adsorption, the remaining studies were carried out without analysis of water concentrations. All endpoints reported here are based on nominal concentrations. Acute toxicity in fish, determined in Brachydanio rerio, did not reveal any effect at 1200 mgl-1 during 96 hours of exposure. For water flea Daphnia magna, the same result was obtained after 48 hours of exposure. T he acute NO EC for these two species, representing primary and secondary consumers, is 1200 mgl-1 N aT FA (corresponds to 1000 mgl-1 of T FA). A study using H T FA with Daphnia magna showed a 24-hour EC50 of 55 mgl-1 (Rhne-Poulenc, 1995); this was attributed to a pH effect. A toxicity study on duckweed (L emna gibba) showed EC50 values for frond increase and weight increase of respectively: 1100 and 1200 mgl-1. For both endpoints, a N O EC of 300 mgl-1 has been recorded (Smyth et al., 1993). T oxicity tests with sodium trifluoroacetate in algae were conducted on 11 different species: Selenastrum capricornutum, C hlorella vulgaris, Scenedesmus subspicatus, C hlamydomonas reinhardtii, Dunaliella tertiolecta, Euglena gracilis, P haeodactylum tricornutum, N avicula pelliculosa (Smyth et al., 1994a), Skeletonema costatum (Smyth et al., 1994b), A nabaena flos-aquae(Smyth et al., 1994c) and M icrocystis aeruginosa. T hese 11 species belong to 4 different classes: Chlorophyceae (4 freshwater and 1 marine species), Euglenophyceae (1 freshwater species), Cyanophyceae (2 freshwater species) and Bacillariophyceae (1 freshwater and 2 marine species). Selenastrum capricornutum (alternative names: R aphidocelis subcapitata or P seudokirchneriella subcapitata) was the most sensitive species for sodium trifluoroacetate. T hisunicellular alga isthe most frequentlyused for ecotoxicitydetermination under laboratory conditions. Based on the results of five toxicity tests with Selenastrum capricornutum, a concentration of 0.12 mgl-1 (120 gl-1) can be considered a toxicitythreshold concentration. Adverse effects on the growth of this species were not found at this concentration. For the remaining 10 algal species the EC50 values were all higher than 100 mgl-1 (see T able 9). Algal studies were also conducted with potential metabolites of trifluoroacetate, like difluoroacetate and monofluoroacetate. M onofluoroacetate is a toxic substance which mode of action is the inhibition of the citric acid cycle. Studies with Selenastrum capricornutum and Scenedesmus subspicatus revealed ef- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 100 Environmental Risk Assessment of T rifluoroacetic Acid fects on growth at sodium monofluoroacetate concentrations which were several orders of magnitude lower than effect concentrations of sodium trifluoroacetate (see T able 10). Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 101 Boutonnet T here is some evidence that exposure to trifluoroacetate results also in the inhibition of the citric acid cycle. Selenastrum capricornutum exposed to sodium trifluoroacetate showed a recovery of the growth when citric acid was added. Furthermore, there seemsto be a correlation between the sensitivityof an algal species for trifluoroacetate and monofluoroacetate. Both compounds are not toxic for C hlorella vulgaris (see T able 10). O ne semi-field study with mesocosm streams has been conducted with sodium trifluoroacetate (Bott and Standley, 1995) to study the potential effects of trifluoroacetate on freshwater algal communities and primary productivity. T he long-term exposure to a mean sodium trifluoroacetate concentration of 31 to 32 gl-1 had no effect on the primary productivity of the diatom dominated algal flora. Effects on organic carbon excretion which were related to high levels of T FA were noted in some experiments. T FA did not alter the algal species composition in the stream mesocosm. In conclusion, based on the results of laboratory toxicity tests in fish, Daphn ia, duckweed, and in a large number of algal and based on the results of the semi-field study with stream mesocosms, an exposure of an aquatic ecosystem to a sodium trifluoroacetate concentration of 0.12 mgl-1 did not show adverse effect. T errestrial O rganisms T errestrial P lants Inputs of T FA to the terrestrial environment will be exclusively via wet and dry deposition from the atmosphere. Due to the high water solubility of T FA and its salts, the assessment of terrestrial effects should be based primarily on precipitation and soil-water concentrations. Prior to 1993, there was little information available on the toxicity of T FA to higher plants. Poignant (1957) had observed phytotoxicity to T riticum vulgare, Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 102 Environmental Risk Assessment of T rifluoroacetic Acid by pre-emergence application, at and above 1500 mg N aT FAl-1. H e also reported the same level of activity to maize (Zea mays), meadow grass (P oa annua), and sweetcorn (A lopecurus agrestis). Ingle (1968) reported significant inhibition of shoot growth of germinating wheat seeds and root growth of germinating tomato seeds at 114 and 285 mg H T FAl-1 (buffered solutions), respectively; T FA showed the weakest phytotoxicity of the seven halogenated aliphatic acids tested. Atmospheric exposure of a conifer, T axus baccata var. Repandens, to H T FA had been investigated by gradually increasing the vapor concentration from 100 to 2100 ppm over a period of 21 days (Freist, 1986). No effects were observed after 19 days (up to 1900 ppm). Slight necrosis of the tips of the young needles was observed at 2100 ppm, which was attributed to the acidity of condensation water which formed on the needles. Because of these indications of phytotoxic potential, the effect of T FA has been investigated for a large number of terrestrial higher plant species. U nless otherwise stated, the experiments employed sodium trifluoroacetate and the concentrations are expressed as the sodium salt. In preliminary studies with sunflower, wheat, and mung bean, using O ECD protocols, the exposure concentrations (1, 10, 100, and 1000 mgkg-1) were expressed by dry weight of soil (Windeatt and T hompson, 1993a,b). As explained above, the results from such tests are of limited value since the soil moisture levels, and thus the soil water concentrations of T FA, fluctuate widely between plant watering occasions. H owever, theyprovided evidence that further studies on higher plants were necessary. Significant effects on germination and emergence (14 days from seed sowing) were observed at 1000 mgkg-1 dry soil for wheat and mung bean and at 10 mgkg-1 dry soil for sunflower. Subsequent vegetative growth was more sensitive; after a further 14 days, a significant reduction in the weight of aerial tissue was observed at 10 mgkg-1 dry soil for wheat and mung bean and at 1 mgkg-1 dry soil for sunflower. T he effects of aqueous exposure to T FA on seed germination was investigated (T hompson and Windeatt, 1994) for 10 species of terrestrial plant, including dicotyledons (sunflower, cabbage, lettuce, tomato), leguminous dicotyledons ( mung bean, soybean) , and monocotyledons ( wheat, corn, oats, rice). T he seeds were added to filter papers soaked with solutions of T FA; except for small seed species, the seeds were also presoaked in the same solutions. No effects were observed for any species at the maximum concentration tested that was 1000 mgl-1. Seedlings of plantain (P lantago major) and wheat (T riticum aestivum variety Katepwa) were exposed to a range of T FA concentrations in the aqueous (hydroponic) medium surrounding the roots. After 14 days, growth of plantain was affected at 100 mgl-1 of sodium trifluoroacetate, but there was no effect at 32 or 10 mgl-1 (T hompson, 1995). Growth of wheat was affected at the lowest concentration tested ( 32 mgl-1) in the preliminary 14-day study and this species was tested again in a longer study at lower concentrations (T hompson et al., 1995). After 5 weeks, wheat growth was inhibited at 10 mgl-1 with tissue necrosis evident first in the shoot and leaf tips; there was no effect at 1 mgl-1 of T FA. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 103 Boutonnet In a similar hydroponic study (Davison and Pearson, 1997), growth of the same variety of wheat was inhibited at a concentration of 5 mgl-1; the effects became apparent after 30 to 35 days, with more severe effects at 10 and 100 mgl-1. As in the earlier study, there was no significant effect at 1 mgl-1 at the end of the exposure (43 days). U sing the same system, another variety of wheat (H anno) showed lower sensitivity; no effects were found at 5 mgl-1 and only slight growth inhibition at 10 mgl-1 after 52 days exposure. U nder these hydroponic conditions, soya bean plants, which had proven more sensitive than several other species to foliar application of T FA (see below), gave results essentially similar to those of the sensitive wheat variety. In a preliminary experiment, symptoms of toxicity were observed at 5 and 10 mgl-1 but were more severe at the higher concentration, becoming apparent after 20 days. Subsequently, exposure concentrations of 5 mgl-1 and 1 mgl-1 were tested, in separate experiments, for 43 and 33 days, respectively. At 5 mgl-1, effects on growth ( dry weight) were observed after approximately 30 days, with symptoms appearing more severe in the younger, expanding leaves. At 1 mgl-1, there was no significant effect on final weight (33 days). H owever, in some plants approximately 25% of the 4th and 5th trifoliate leaves (the youngest) appeared to show slight symptoms (slight rounding of the tip and barely discernible necrosis of the margin), but this was variable between plants. O ther studies (Davison and Pearson, 1997) have investigated the effects of foliar application of solutions of NaT FA. Seedlings (7 days old) of seven species of terrestrial plant (sunflower, soya, wheat, maize, oilseed rape, rice, and plantain) were field-grown, to ensure that the leaves were fully hardened, sprayed (60 microns droplet size) with solutions of NaT FA for 7.5 hours and kept wet with the mist for a further 16.5 hours. T he soil and plant roots were protected from the application. After 3 weeks, there was no effect of NaT FA on the height, leaf number, stomatal conductance, final harvest weight or chlorophyll and carotenoid concentrationsof anyof the speciesat the maximum concentration tested which was 100 mgl-1. T here was a significant decrease in specific leaf area only for wheat at 100 mgl-1. A subsequent study, at only 100 mgl-1, used a similar system but with laboratory grown plants (wheat, maize, sunflower and soya), both with and without protection of the soil and roots from the spray (Davison and Pearson, 1997). O nly soya showed any symptoms of toxicity. T hese were apparent by both routes of exposure, suggesting that laboratory grown plants were more sensitive to foliar application than those field-hardened; however, the symptoms were much more severe when the NaT FA solution was able to reach the roots. In a further experiment, soya seedlings were exposed to a range of concentrations of NaT FA applied to the soil surface in a volume equivalent to 10 mm of rainfall every 3 days for 44 days. T here was no effect at 1, 5, and 10 mgl-1 on dry weight, leaf size, or stomatal conductance (transpiration) and no visible injury. At 100 mgl-1, symptoms of toxicity were apparent and leaf size ( but not dry weight) was reduced, suggesting T FA affected leaf expansion. Although this regime simulated relatively high rainfall ( 100 mm per month) , the effec- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 104 Environmental Risk Assessment of T rifluoroacetic Acid tive concentration was more than one order of magnitude higher than by hydroponic exposure. T his suggests that the hydroponic exposures represent the worst case situation, possibly because transpiration rates are maximized. Studies by Emerich (1997) have shown that T FA at a concentration of 1 mgkg-1 of soil had no effect on the germination or growth of soybean seedlings. Also, 1 mg T FAkg-1 had no discernible effect on the nitrogen-fixing capacity of these soybean plants as determined by plant nodule development or by direct assay of nitrogenase activity of these soybean nodules. T oxic effects were, however, observed in the development of the plants and in the nodulation patterns at 10 and 100 mg T FAkg-1. Comparable results were obtained when the studies were carried out under hydroponic conditions, at concentrations of 1, 10, and 100 mg T FAl-1. T here were no significant effects on any parameter at 1 mgl-1, with pronounced effects on plant development at 10 and 100 mgl-1. It can be concluded that the most significant route of exposure for terrestrial plants is uptake from soil-water via the roots. T FA enters the transpiration stream and is taken to the shoots and leaves where there is a tendency for T FA to bioaccumulate as the water evaporates. For the most sensitive species, with long-term, continuous exposure under conditions allowing high transpiration rates, no effect on the growth of plants occurs at 1 mgl-1 (see T able 11). T his concentration is four orders of magnitude higher than the projected rainwater concentration of T FA derived from CFC alternatives. T errestrial I nvertebrates Because of the low toxicity of T FA to aquatic invertebrates (see a previous section) and other higher organisms, it was not considered necessary to test terrestrial, soil-dwelling organisms. M ammals Distribution, T oxico-K inetics, and M etabolism T he distribution half-life of T FA in rabbits was about 20 minutes (Kinoshita, 1989). After intravenous administration in rats, approximately 58% of the T FA was distributed in the total body water, whereas plasma proteins were binding 10% (H oladay and Cunnah, 1976). In human blood, T FA binds mainly to albumin (44 to 54%) and only little (4 to 14%) to erythrocytes (Dallmeier and H enschler, 1981; H oladay and Cunnah, 1976). After NaT FA administration via drinking water to male Wistar rats the amount of organically bound fluorine in the plasma and the liver reached a steady level within 3 days of treatment with a 1:1 ratio of liver to plasma concentrations (Stier et al., 1972). In pregnant mice, the T FA concentration ratio between amniotic fluid/ maternal plasma became >1 at 24 hours after an intravenous infusion. Approximately 20 to 30% of the T FA was bound to amniotic fluid and plasma macromolecules (Ghantous et al., 1986). Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 105 Boutonnet ( Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 106 Environmental Risk Assessment of T rifluoroacetic Acid T FA has been shown to be the ultimate biotransformation product of H CFC123 (2,2-chloro-1,1,1-trifluoroethane) (H arriset al., 1991; O lson et al., 1991). In a study with H CFC-123 in lactating Sprague-Dawley rats, T FA was determined in milk. T he measured concentration in milk increased 5 to 23% after saponi- Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 107 Boutonnet fication, indicating that a part of the T FA was covalently bound to peptides or proteins (Buschmann, 1996). T FA was also identified as a major metabolite of halothane. A study (Kinoshita, 1989) using rabbits exposed to halothane demonstrated that the biliary route is the most important excretory pathway for biotransformed T FA. A large quantity of the acid is excreted into the bile, reabsorbed in the gastrointestinal tract, and appears in the circulating blood. T he elimination halflife after intravenous injection or oral administration of T FA was 34 and 48 hours, respectively. Enterohepatic circulation is one of the probable mechanisms of the long lasting excretion of T FA. Kawahara et al. (1988) have reported salivary excretion of T FA in surgical patients and M irkov et al. (1988) have reported excretion of T FA in the digestive juice after halothane inhalation in guinea pigs. T he majorityof metabolized H CFC-123 isexcreted as T FA in the urine (Vinegar et al., 1994). Following a 6-hour exposure of lactating Sprague-Dawley rats to 1000 ppm H CFC-123 and after 1 hour of nursing (15 ml milk per litter containing 50 gml-1 T FA), each litter excreted over 16 hours an average quantity of 79 g T FA in 9.1 ml urine. T his study illustrates the relative high concentration in the mothers milk and the prolonged retention of T FA by the rat pup (Buschmann, 1996). In adult rats the plasma half-life after intravenous injection was 30 hours (H oladay and Cunnah, 1976). Intravenous-injected T FA in two human volunteers was completely (95 to 100%) excreted via the urine over 63 to 72 hours (plasma half-life 25 to 32 hours) (H oladay and Cunnah, 1976). In a 2-week (5-day week) exposure study with 22 ppm halothane the T FA concentration in the blood in the second week was only slightly higher compared to that of the first week. T FA level in human blood and urine are correlated linearly with levels in blood ( 0.2 to 9 mgl-1) being approximately three times higher (Dallmeier and H enschler, 1981). T FA has not been found to be metabolized to any appreciable extent in rats (Fraser and Kaminsky, 1988). A cute T oxicity Acutely, trifluoroacetic acid would be categorized as being moderately toxic by the oral route of administration having a L D50 of a 2 to 5% solution in the range of 200 to 400 mgkg-1 body weight in rats and mice (Patty, 1963). O ral L D100, L D10, and L D0 values were reported by another laboratory to be 1000, 500, and 250 mgkg-1 body weight, respectively (Kheilo and Kremneva, 1966). T he sodium salt of trifluoroacetic acid is much less toxic than the free acid. No deaths were observed when mice were administered intraperitoneally up to 5000 mgkg-1 of N aT FA; whereas, a dose of 150 mgkg-1 H T FA caused death in 2 out of 5 mice, a result which is comparable to that obtained with an equimolar dose of hydrochloric acid ( Blake et al., 1969) . O ther reported L D50 values obtained with the sodium salt using the intraperitoneal route of administration include: >4000 mgkg-1 (Rosenberg, 1971) Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 108 Environmental Risk Assessment of T rifluoroacetic Acid and >2000 mgkg-1 (Airaksinen and T ammisto, 1968; Airaksinen et al., 1970). An intravenous L D50 for N aT FA was reported to be 1200 mgkg-1 for mice (Airaksinen and T ammisto, 1968). I nformation on the acute inhalation toxicity of H T FA is limited to a single Russian report ( Kheilo and Kremneva, 1966) . Values were reported for L C50 for both mice and rats. T he L C50 for a 2-hour exposure in mice was reported to be 13.5 mgl-1 ( 2900 ppm) ; additional values for L C100 and L C10 ( 2 hour) were 20.4 mgl-1 ( 4366 ppm) and 9.2 mgl-1 ( 1968 ppm) , respectively. T he 2-hour L C50 for the rat was reported to be 10 mgl-1 ( 2140 ppm) with additional values for L C100 and L C10 ( 2 hour) of 11.5 mgl-1 ( 2461 ppm) and 8.3 mgl-1 ( 1776 ppm) . According to these investigators, for rats the threshold concentrations are 4 mgl-1 ( 856 ppm) based on "the temperature reaction of the body" and 1.5 mgl-1 ( 321 ppm) based on "neuromuscular excitability." T he threshold concentration for irritation in humans exposed for one minute is 0.25 mgl-1 ( 54 ppm) . Following treatment of Swiss male mice with either a single intraperitoneal dose of 1000 mgkg-1 or 2000 mgkg-1 N aT FA, hepatocytes were observed to have a cloudy swelling with slight accumulation of fat accompanied by an increase in liver glycogen at the lower dose. At the higher dose, vacuolization was also noted (Rosenberg and Wahlstrom, 1971). At 2000 mgkg-1 N aT FA given to mice intraperitoneally, the glucose-6-phosphate dehydrogenase activity in livers and erythrocytes was slightly increased and a transient decrease in glutathione and N ADPH content was observed in the liver at 12 hours after administration. T his decrease was also observed in erythrocytes at 24 hours (Rosenberg, 1971). T he relative low toxicity of T FA is demonstrated by comparison with two closely related chemicals: trifluoroethanol (T FE) and trifluoroacetaldehyde (T FAld). T he intraperitoneal L D50sof T FE, T FAld and N aT FA were 158 to 195, 650 and >2000 mgkg-1, respectively, in male Swiss mice (Airaksinen and T ammisto, 1968; Airaksinen et al., 1990). When these three chemicals were each administered to 10-week old, male AlpK/ AP strain (Wistar-derived) rats with single intraperitoneal (L loyd et al., 1986) or single oral (L loyd et al., 1988) doses of 10 and 25 mgkg-1 body weight, T FE and T FAld caused a dose-related reduction in testis weight within 3 days, which was accompanied by morphological changes. In marked contrast, T FA did not cause any observable testicular effects on weight or morphology. U sing an in vitro Sertoli/ germ cell co-culture system obtained from the rat, at concentrations of 0.1 to 1 mM , T FAld produced dose-related effects including increased germ cell loss of pachytene and dividing spermatocytes. T he germ cell losses were accompanied by leakage of the pachytene spermatocyte marker enzyme lactic acid dehydrogenase-X (L DH X ) and decreased lactate and pyruvate production. With T FA increased cell loss and increased L DH X leakage became only apparent at 10 mM . No response was observed with 10 mM T FE (L loyd et al., 1986). In a similar study (Williams, 1997) in isolated L eydig cell cultures, Sertoli cell only cultures and Sertoli/ germ cell co-cultures obtained from the Sprague-Dawley rat T FAld had also the most wide ranging Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 109 Boutonnet and severe effects on the function of isolated testicular cells, affecting all three cell systems. T FA and T FE show only a small effect on L eydig cell function and essentiallyno effect on the Sertoli/ germ cell cultures. T he apparent difference of activity of T FE between in vitro and in vivo is explained by the metabolism of T FE in vivo to T FAld. T FE, especially when muscular fibrillations were induced by atropineneostigmine, caused a drop in the lactate and pyruvate pool in muscle and liver of mice and guinea pigs, probably by inhibition of glycolysis. I n guinea pig muscle a decrease of AT P was also seen. N aT FA had little, if any, effect on AT P, lactate and pyruvate. Both N aT FA and T FE had no effect on citrate values, showing that they do not block the Krebs' cycle ( Airaksinen and T ammisto, 1968) . I n a perfusion experiment with rat liver, Stier et al. ( 1972) showed that N aT FA enhanced lactate and pyruvate turnover and uptake without apparent stimulation of the tricarboxylic acid cycle or glucose synthesis. Single intraperitoneal doses of 2.1 mmolkg-1 of T FE and T FAld produced significant bone marrow and intestinal toxicity, which is characterized by leucopenia and loss of intestinal dry weight. T his eventually leads to a lethal septicemia in male Wistar rats. H T FA, when administered at 240 mgkg-1 (a dose equimolar to T FE and T FAld) did not produce any similar toxic effects (Fraser and Kaminsky, 1988). R epeated Dose T oxicity When H T FA or N aT FA were added to drinking water to a concentration of 114 gl-1 ( 1 N l-1) , male Sprague-Dawley rats rejected the solutions which resulted in dehydration and body weight loss ( Blake et al., 1970) . As a consequence, an increase ( <30%) in the liver-to-body weight ratio was observed within 10 days. I n contrast, daily intragastric administration of 1 mlkg-1day- 1 of 1N N aT FA ( 114 mgkg-1day-1) to rats, for 8 days, did not cause dehydration or body weight loss and did not significantly affect the liver-to-body weight ratio. Furthermore, 1 N N aT FA did not affect the duration of hexabarbital-induced hypnosis suggesting no change in the metabolic activity of the liver. In one other report (Stier et al., 1972), N aT FA was administered via drinking water to male Wistar rats such that the animals were provided with approximately 130 mol NaT FA per 100 g body weight per 24 hours (equivalent to 150 mgkg-1day-1, based on an assumed normal consumption of water) for 5 to 6 days. After 5 days of treatment, the relative liver weight was increased by about 43%. Furthermore, the glycogen content of the liver was decreased. M ale Wistar rats were fed trifluoroacetate, perfluorobutyrate, and perfluorooctanoate in their diets for 5 to 14 days at concentrations of 5000 ppm (500 mgkg-1day-1) for the T FA, and 2500 ppm for the latter two compounds (Just et al., 1989). All three compounds induced hepatomegaly, anorexia, and peroxisome proliferation. T he relative liver weight increased by 20 to 30%. With T FA, the body weight gain was similar to controls. Perfluorobutyrate and Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 110 Environmental Risk Assessment of T rifluoroacetic Acid perfluorooctanoate were found to be more active than T FA. T he rise in activity of peroxisomal fatty acid -oxidation was only slight for T FA, amounting to about a 2-fold increase, whereas the increase in activity was 9- to 10-fold after perfluorobutyrate and perfluorooctanoate treatment. M ale rats and male guinea pigs were fed diets containing 7500 ppm (rat: 750 mgkg-1day-1) for 25 days. A decrease of body weight without concomitant changes in food consumption was observed for rats. Decreases in serum cholesterol, triglycerides, glucose, and insuline levels were reported in rats and to a lesser extent in guinea pigs. Increased liver weights and diffuse liver hypertrophy were observed in treated rats, but in guinea pig the liver weights were reduced. T he rate of hepatic peroxisomal -oxidation was increased in rats but not in guinea pigs supporting the conclusion that peroxisome induction is a rodent-specific phenomena (Warheit, 1993). Studies on repeated inhalation exposures of animals to H T FA are limited to one report (Kheilo and Kremneva, 1966). Rats were exposed for 4-hours daily 6 times a week at concentrations of 0.4 to 0.7 mgl-1 (86 to 150 ppm) H T FA for 5 months. Effects observed included: irritation of mucous membranes, increased proteinuria, and "altered neuromuscular excitability." T hese investigators noted that chronic exposure to 0.025 to 0.05 mgl-1 (5 to 11 ppm) caused only very slight symptoms and suggested that they can be considered to be close to the threshold concentrations for chronic exposures. In conclusion, the major target organ in T FA-exposed rats is the liver, showing mild effects (increased weight, hypertrophy and induction of peroxisomes) at 150 mgkg-1day-1 by gavage for 5 days. I rritation Application of solutions of H T FA to the skin of rats or guinea pigs, at concentrations of 20% or greater, caused marked coagulation, necrosis; 10% concentration was moderately severely irritating and concentrations of from 2 to 5% were moderately irritating ( Patty, 1963) . Kheilo and Kremneva ( 1966) reported H T FA to be corrosive to rabbit skin. C orrosivity As with any strong acid, severe eye irritation would be expected from contact with H T FA (Patty, 1963). O cular irritation was observed during the Russian acute inhalation studies in rats and mice (Kheilo and Kremneva, 1966). Sen si ti za ti on T FA, when conjugated to human and guinea pig albumin, was shown to induce a specific delayed-type hypersensitivity in guinea pigs (M athieu et al., 1974) . Erythema and induration were observed 24 to 48 hours after intradermal challenge with antigen, as evidence for cutaneous hypersensitivity. Perivascular mononuclear infiltration was observed, histologically. T his prompted the investigators to conclude that T FA may serve as a hapten able to elicit specific cellular immune reponses. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 111 Boutonnet Several reports relating to cellular-mediated immune response have appeared in the literature. T hese have included demonstrations of trifluoroacetylated neoantigens such as: T FA-albumin (Reves and M cCracken, 1976a,b), T FA-ovalalbumin (Ford et al., 1984), T FA-neoantigen (present in microsomal carboxylesterase) (Satoh et al., 1984), and trifluoroacetylated hepatocytes (Satoh et al., 1985) that have resulted from exposure to halothane or T FA. H owever, the role of T FA in cell-mediated immunity is questionable because T FA, as well as halothane, did not stimulate DNA formation in cultured lymphocytes obtained from the blood of human subjects previously exposed to halothane (Waldron and Ratra, 1972), and the delayed hypersensitivity response to T FA had no effect on halothane-induced liver damage in mice and rats (Ford et al., 1984). Furthermore, skin test hypersensitivity reactions with T FA, prepared as a complex with autologous serum protein, did not correlate with hepatic necrosis induced by halothane exposure in guinea pigs (Reves and M cCracken, 1976b). M utagenicity NaT FA was tested for mutagenicity in the standard Ames assay (Blake et al., 1981). NaT FA was found to be nonmutagenic to Salmonella typhimurium strains T A98, T A100, and T A1535, both in the presence and absence of Aroclor 1254induced rat liver or rat testes S-9 activation systems. H T FA was tested in another laboratory (Waskell, 1978) using Salmonella typhimurium strains T A98 and T A100, both with and without rat liver S-9 activation, and was also found to be nonmutagenic. Additionally, H T FA was tested using the repair-deficient strains T S24, T A2322, and T A1950 of Salmonella typhimurium. H T FA did not inhibit the growth of any of the above DN A repair-deficient strains relative to a normal repair-proficient hisG , thus indicating a lack of genetic activity. I n yet another laboratory (Baden et al., 1976), H T FA or urine from patients anesthetized with halothane were all found to be nonmutagenic to Salmonella typhimurium strains T A98 and T A100, either in the presence or absence of Aroclor 1254-induced rat liver S-9 activating medium. C arcinogenicity No information was found with regard to the carcinogenic potential of T FA. T oxicity for R eproduction T FA has not been tested for teratogenicity or fetotoxicity. RISK C H ARAC T ERIZAT IO N Present and Future Environmental C oncentrations As described in detail in previous sections, the present contributions to environmental levels of trifluoroacetic acid can be summarized as follows. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 112 Environmental Risk Assessment of T rifluoroacetic Acid P roduction and U se of T rifluoroacetic A cid T rifluoroacetic acid is manufactured at about 1000 metric tons per year and is widely used in the fine chemicals industry and as a laboratory reagent. I t is either consumed or becomes part of a chemical waste stream. T he quantities released into the atmosphere from this source are very small indeed and do not appear to be a significant contribution to global environmental levels. U se of Fluorocarbon A naesthetics H alothane and isoflurane anaesthetics, which have been in use since the 1970s, both yield trifluoroacetate when degraded. H alothane is the most used of the fluorocarbon anaesthetics at a maximum of 1500 metric tons per year globally. All the material used is assumed to be emitted to the atmosphere. T he global deposition rate of trifluoroacetic acid from this source is estimated to be 800 metric tons per year. U se of C ertain C FC Substitutes CFCs substitutes -- H CFCs and H FCs -- break down in the environment to give predominantly carbon dioxide, water, and inorganic salts of chlorine and fluorine. H owever, some substances, most notably H FC-134a, H CFC-124, and H CFC-123, also break down to yield trifluoroacetic acid (H T FA), which is resistant to further atmospheric degradation. Because the H CFCs and H FCs have until now been produced only in limited commercial quantities, their contribution to present environmental levels of T FA has been estimated to be small -- around 2000 metric tons per year. In total, the present known emission sources of all fluorocarbon precursors (anaesthetics +CFC substitutes) will yield trifluoroacetic acid at a rate of 2800 metric tons per year. T o derive future environmental concentrations, predicted production and releases of the relevant H CFCs and H FCs, along with their rates of transformation into T FA, have been modeled. Substances that decompose in the atmosphere by oxidation or photolysis will give rise to diffuse fluxes of decomposition products in the atmosphere as the first compartment. H CFCs and H FCs that break down to trifluoroacetic acid have atmospheric lifetimes ranging from about 1.4 to 14.6 years, so that most decomposition occurs at or near the background concentration of the substance after it has become well mixed in the atmosphere. T he decomposition mechanisms depend on physical and chemical properties and processes that are relatively well understood so that it is possible to calculate accurately the future atmospheric concentrations of precursors to trifluoroacetate, given a particular set of scenarios for emissions. I t has been calculated ( see earlier section for details) that in the year 2020 the global deposition of trifluoroacetic acid as a breakdown product of the H CFCs and H FCs will be about 160,000 metric tons per year. Based on the physico-chemical properties of T FA, the preferred environmental compartment will be water, rather than air, ground, or biota. T hus, trifluoroacetic acid that is emitted from processes as a vapor, or trifluoroacetate Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 113 Boutonnet that is formed from other materials by reaction in the atmosphere will partition rapidly into cloud, rain, or surface water. I n rain water a range of concentrations has been calculated, taking into account geographical variations (O H radical concentrations, amount of rain, regional releases of parent compound). Because most of the rainfall and the highest regional concentrations of hydroxyl occur in the tropics, the mass deposition rate of trifluoroacetic acid is highest in this region. It was calculated that in 2020 the tropical concentration would be 2 molm-2yr-1 or 100 ngl-1 in rainwater. O ver Europe, the deposition rate would be about 0.5 molm-2yr-1 and over the U nited States levels would be up to double this. For this risk assessment, a maximum level of T FA in rain water in the region of 0.1 gl-1 (0.0001 mgl-1) in the year 2020 was adopted as the standard for the predicted environmental concentration. In surface water, it appears to be possible to accumulate a riverborne flux of T FA in lakes or seas from which there is no outflow, only evaporation. H owever, consideration of the volumes and flows indicates significant local accumulation over global levels takes many hundreds of years. A possible buildup in special ecosystems, such as seasonal wetlands (special aquatic ecosystems that dry out periodically and are replenished only by rainfall), has been postulated but significant buildup over background levels can only be anticipated if a number of low probability events occur at the same time and endure for a long period of time. T herefore, although some accumulation of T FA may take place in special ecosystems like vernal pools, accumulation over several orders of magnitude appears to be highly improbable. Generally, soil retention of T FA is poor although soils with high levels of organic matter have been shown to have a greater affinity for T FA when contrasted with soils with low levels of organic matter. T his appears to be an adsorption phenomenon, not irreversible binding. Based on the maximum observed partition coefficient (Kd) of 20, the maximum concentration bound to soil -- in equilibrium with the projected rainfall concentration (from CFC alternatives) in 2020 of 0.0001 mgl-1 -- would be 0.002 mgkg-1. Degradation mechanisms of T FA that could mitigate future concentrations of this compound have been studied quite extensively. Given the high thermodynamical stability of the CF3 group, no significant abiotic mineralization process has been observed or is forecast in environmentally relevant conditions. T FA is highly resistant to both microbial oxidative and reductive degradation. Although biodegradation has been observed under specific conditions, the relevance of these results to the real world are considered to be doubtful. For the purpose of this risk assessment, we will assume, as a reasonable worst case, that no degradation of T FA is likely to occur in the environment. T FA's low octanol/ water partition coefficient (log Pow =-2.1) indicates no potential to bioaccumulate. T FA does not accumulate significantly in lower aquatic life forms such as bacteria, small invertebrates, oligochaete worms and some aquatic plants including L emna gibba (duckweed). In terrestrial higher plants such as sunflower and wheat, some bioaccumulation was seen ( bioconcentration factors up to 43 based on fresh weight). T his appeared to be related to uptake with water and then concentration due to transpiration water loss. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 114 Environmental Risk Assessment of T rifluoroacetic Acid In the event that T FA is degraded, it may be transformed into monofluoroacetic acid (M FA). H owever, as the rate of breakdown of M FA (hydrolytically by fluoroacetate dehalogenases) is so much higher than for T FA, any M FA formed would rapidly degrade. T herefore, there would be no buildup of M FA regardless of the levels of T FA present in the environment. Effect Assessment Given the fact that deposited T FA would remain in water, a number of studies have been conducted aiming to derive a no-effect concentration in the aquatic compartment. Standard acute tests on fish and D aphn ia carried out with N aT FA show that they are insensitive to large concentrations ( up to 1 gl-1 T FA) . T his low toxicity of T FA is reinforced by toxicological data on mammals developed previously, namely, in the context of T FA being a metabolite of several anaesthetics ( see earlier section ) . O n the other hand, the no-effect concentration for the standard algae species Selen astru m capricorn u tu m is around 0.10 mgl-1 ( as T FA) . T o see if this sensitivity was general among algae, 10 other species belonging to 4 different classes have been tested; these other species were not sensitive to N aT FA ( no-effect concentration >100 mgl-1) . Even among the chlorophyceae, Selen astru m capricorn u tu m seems to be unique; therefore, even if this species was affected above 0.1 mgl-1, the ecology of the system (i.e., the functionality) where it lives would not be affected. Consequently, it is considered that for the protection of the aquatic environment the no-effect concentration of 0.1 mgl-1 of this alga can be used. Extensive research has been devoted to effects on higher plants, as they could be exposed to T FA in rainwater through leaves and stems and T FA in pore water through roots. A number of species were tested, with particular emphasis on those having an important role in feeding people and cattle. Exposure of the leaves was shown to be less important than exposure of the roots. By application to the soil, the most sensitive species was sunflower, with an effects threshold of 1 mg N aT FAkg-1 ( dry soil) . T he same soil concentration had no effect on the germination, growth and nitrogen-fixing capacity of soya seedlings. H owever, because of the limited binding of T FA to soil, effects are better defined as concentrations in the soil water. By exposure of the roots to aqueous solutions, the no-effect concentration was 1 mg N aT FAl-1 for the long-term growth of wheat and soya. T herefore, as a conservative figure, the no-effect concentration in soil-water is considered to be 0.1 mgl-1. Risk Assessment In comparing the present and future environmental concentrations of T FA ( maximum of 0.0001 mgl-1 in rainwater in 2020) with the no-effect concentrations in both surface water and soil (0.1 mgl-1), it is concluded that no noticeable risk for the environment can be anticipated as there is a 1000-fold difference. T he local enrichments in certain aquatic systems and organic rich soils do not influence this conclusion. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 115 Boutonnet For the time being, an important question remains concerning the origin of the large present levels of T FA that have been measured in the environment (fresh and marine surface waters, rain, and air) and cannot be explained by the known industrial sources. T hese levels, in the range of 0.1 gl-1 for water and 50 ngl-1 for air, are roughly60 timeswhat isestimated to arise from known sources today and more or less equal to what is anticipated in 25 years from now. AC KN O WL EDG MEN T S T he authors would like to thank the principal investigators who contributed to the AFEAS T FA research program: A.G. Berends, A.H .C. Groeneveld and N.R.M . van Dijk (Solvay Duphar), T homas Bott and L aurel Standley (Stroud Water Research Center), Craig Criddle (M ichigan State U niversity), Alan Davison (U niversity of Newcastle), Charles Driscoll (Syracuse U niversity), David Emerich (U niversity of M issouri), H artmut Frank (U niversity of Bayreuth), Anders Grimvall (L inkping U niversity), M alcolm Ko (Atmospheric and Environmental Research, Inc.), Gene L ikens (Institute of Ecosystem Studies), Ron O remland (U .S. Geological Survey), Steve Schwarzbach (U .S. Fish and Wildlife Service), James Seiber (U niversity of Nevada - Reno), and C.G. van Ginkel ( formerly Akzo) . I n addition, we wish to acknowledge Karlheinz Ballschmiter (U niversity of U lm), Grard Blake (U niversity of Savoie) and Paul Falkowski (Brookhaven National L aboratory), who served as advisors or reviewers. T he T FA research program was sponsored by the Alternative Fluorocarbons Environmental Acceptability Study (AFEAS), a consortium of international chemical manufacturers. AFEAS members include: AlliedSignal, Inc. (U SA), Asahi Glass Co., L td. (Japan), Ausimont S.p.A. (Italy), Daikin Industries, L td. (Japan), E.I. DuPont de Nemours & Co., Inc. (U SA), Elf Atochem S.A. (France), H oechst AG (Germany), ICI Chemicals & Polymers L td. (U K), L aRoche Industries Inc. (U SA), Rhodia L td. (U K), and Solvay S.A. (Belgium). T he program is managed by RAND Environmental Science & Policy Center (Washington, DC). R E FE R E N C E S Airaksinen, M .M . and T ammisto,. T . 1968. T oxic actions of the metabolites of halothane: L D50 and some metabolic effects of trifluoroethanol and trifluoroacetic acid in mice and guinea pigs. A nn. M ed. Exp. Biol. Fenn. 46(2), 242-248. [C.A. 69:65967q] Airaksinen, M .M ., Rosenberg, P.H ., and T ammisto, T . 1970. A possible mechanism of toxicity of trifluoroethanol and other halothane metabolites. A cta P harmacol. T oxicol. 28(4), 299-304. [C.A. 73: 75271e] Asmus, K-D., Fliount, R., H ungerbuhler, H ., and M akogon, O . 1994. Degradation of trifluoroacetic acid (T FA) heterogenous processes. Proceedings of a Workshop on Decomposition of T FA in the Environment, February 8 to 9, 1994. Published by Alternative Fluorocarbons Environmental Acceptability Study (AFEAS), T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005 U SA. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 116 Environmental Risk Assessment of T rifluoroacetic Acid Atallah, M .M . and Geddes, I.C. 1977. Effect of halothane metabolites on repeated halothane anesthesia in rats. I sot. R adiat. R es. 9(2), 65-69. [C.A. 92:174526r] Baden, J., Wharton, R., H itt, B., Brinkerhoff, M ., Simmon, V., and M azze, R. 1976. M utagenicity of volatile anesthetics. Fed. P roc. 35, 410 [ Abstract N o. 1132] . Ballschmiter, K. 1992. T ransport and fate of organic compounds in the global environment. A ngew. C hem. I nt. Ed. Engl. 31, 487-515. Bastable, H .G., Rogers, D.P., and Schoran, D.E. 1990. T racers of O pportunity and Pollutant T ransport in Southern California. A tmos. Environ. 24B, 137-151. Berends, A.G. 1996. T he toxicity of sodium trifluoroacetate to the marine alga Dunaliella tertiolecta. T echnical Report 56834/ 20/ 96, Solvay Duphar, Weesp, T he Netherlands. Berends, A.G. 1996. T he toxicity of sodium trifluoroacetate to the freshwater alga Euglena gracilis. T echnical Report 56834/ 41/ 96, Solvay Duphar, Weesp, T he Netherlands. Berends, A.G. 1996. T he toxicity of sodium trifluoroacetate to the marine alga P haeodactylum tricornutum. T echnical Report 56834/ 39/ 96, Solvay Duphar, Weesp, T he N etherlands. Berends, A.G. 1993. T he toxicityof sodium trifluoroacetate to the alga C hlorella vulgaris. T echnical Report 56835/ 29/ 94, Solvay Duphar, Weesp, T he Netherlands. Berends, A.G. and Groeneveld, A.H .C. 1996. T he toxicity of sodium trifluoroacetate to the freshwater alga C hlamydomonas reinhardtii. T echnical Report 56834/ 19/ 96, Solvay Duphar, Weesp, T he Netherlands. Berends, A.G. and M olenaar, J.A. 1993. T oxicity of sodium trifluoroacetate to the alga Selenastrum capricornutum at low concentrations. T echnical Report 56635/ 61/ 92, Solvay Duphar, Weesp, T he Netherlands. Berends, A.G., Keetelaar-Jansen, W.A.J., and van Dijk, N.R.M . 1994. A comparison of the toxicity of sodium trifluoroacetate, sodium difluoroacetate, sodium monofluoroacetate and sodium fluoride to the alga Scenedesmus subspicatus. T echnical Report 56835/ 61/ 94, Solvay Duphar, Weesp, T he Netherlands. Berger, T .W., T artowski, S.L ., and L ikens, G.E. 1997. T rifluoroacetate (T FA) retention in a northern hardwood forest soil. Environ. Sci. T echnol. 31, 1916-1921. Blake, D.A., Cascorbi, H .F., Rozman, R.S., and M eyer, F.J. 1969. Animal toxicityof 2,2,2trifluoroethanol. T oxicol. A ppl. P harmacol. 15, 83-91. Blake, D.A., Barry, J.G., and Cascorbi, H .F. 1970. Effect of trifluoroacetate on the growth of rat liver. Naunyn-Schmiedesbergs. A rch. P harmakol. 265(5), 474-475. [C.A. 72:108792k] Blake, D.A., DiBlasi, M .C., and Gordon, G.B. 1981. Absence of mutagenic activity of trifluoroethanol and its metabolites in Salmonella typhimurium. Fund. A ppl. T oxicol. 1, 415-418. Bott, T .L . 1994. T he effect of T FA on biological processes in streams. Workshop on the Environmental Fate of T rifluoroacetic Acid, M arch 3-4, 1994, M iami Beach, FL . Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. Bott, T .L . and Standley, L .J. 1995. T he interaction of trifluoroacetate with microbial communities. Abstract for platform presentation for the annual meeting for the American Society for M icrobiology, Washington, DC, U SA. Bott, T .L . and Standley, L .J. 1994. Potential effects of T FA on freshwater algal communities and productivity. Final report to AFEAS by Stroud Water Research Center (Project SP91-18.11/ BP-93.20). Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 117 Boutonnet Bott, T .L . and Standley, L .J. 1994. T he effect of T FA on biological processes in streams. Proceedings of a Workshop on the Environmental Fate of T rifluoroacetic Acid, M arch 3-4, 1994, M iami Beach, FL . Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005, U SA. Bott, T .L . and Standley, L .J. T rifluoroacetate, an atmospheric breakdown product of hydrofluorocarbon refrigerants, and freshwater benthic microbial communities: incorporation and effects on acetate metabolism. Submitted to Environ. Sci. T echnol. Bowden, D.J., Clegg, S.L ., and Brimblecombe, P. 1996. T he H enry's L aw constant of trifluoroacetic acid and its partitioning into liquid water in the atmosphere. C hemosphere 32(2), 405-420. Brown, A.C., Canosa-M as, C.E., Parr, A.D., and Wayne, R.P. 1990. L aboratory studies of some halogenated ethanes and ethers: M easurements of rates of reaction with O H and of infrared absorption cross-sections. A tmos. Environ. 24A(9), 2499-2511. Buschmann, J. 1996. Crossover study with H CFC 123 in lactating Sprague-Dawley rats including additional studies on milk production and metabolites in offspring urine. Fraunhofer Institute of T oxicology and Aerosol Research. Final report 95/ 9. Buschmann, J., Bartsch, W., Koch, W., and Preiss, A. Inhalation Study with H CFC 123 in L actating Sprague-Dawley Rats. Submitted to: Fund. A ppl. T oxicol. C&P News. 1995. Plant overhaul record tops off tough year. Published by ICI Chemicals & Polymers L td., Runcorn, U K, December 1995. C&P N ews. 1996. Product partnership reaches sales peak. Published by I CI Chemicals & Polymers L td., Runcorn, U K, June 28, 1996. Chauhan, S., Criddle, C., and T iedje, J. 1995. Potential for microbial transformation of trifluoroacetic acid. Final Report to AFEAS by M ichigan State U niversity (Project SP91-18.11/ BP93-15). Chumley, F.G. 1992. Ecological testing of an industrial wastestream that contains trifluoroacetate (T FA): L ack of evidence for biodegradation of T FA. Interim report, DuPont Central Research & Development, Wilmington, DE 19880. Cox, R.A., Rattigan, O ., H ayman, G.D., H oward, C.J., Nielsen, O .-J., Sidebottom, H .W., and Zellner, R. 1995. Atmospheric Degradation M echanisms -- Summary of Discussions. Ch. 2 of NASA, NO AA and AFEAS (1995). Dallmaier, E. and H enschler, D. 1981. H alothane hazards for the team in the operation theatre. Deutsch. M ed. W ochenschr. 106(11), 324-338. D.B. (Deutscher Bundestag). 1994. H FC 134a and other CFC substitutes. Section 4.3.3 of Responsibility for the Future, O ptions for Sustainable M anagement of Substance Chains and M aterial Flows, Economica Verlag, Berlin. Davison, A.W. and Pearson, S. 1997. T oxicity of T FA to plants. Final Report to AFEAS by U niversity of Newcastle (Project SP91-18.23/ BP96-31). DeFlaun, M .F. 1996. Aerobic biodegradation study of trifluoroacetic acid (T FA). Final Report to DuPont from Envirogen Inc., Princeton Research Center, 4100 Q uaker Bridge Rd. L awrenceville, NJ 08648 (Report #57202). Dittmann, A.L . 1975. U .S. Patent 3,883,407 to H alocarbon Products Corp., M ay 13, 1975. Elliott, A.J. 1994. Fluorine compounds organic (acetic acids). Vol. 11, pp. 546-550 of Kirk-O thmer Encyclopoedia of Chemical T echnology, Wiley, New York, NY. Emerich, D. 1997. Effects of T FA on symbiotic nitrogen fixation by soybean. Final Report to AFEAS by U niversity of M issouri (Project SP91-18.24/ BP95-32). Emptage, M ., T abinowski, J.A., and O dom, J.M . 1997. T he effect of fluoroacetates on methanogenesis in samples from selected methanogenic environments. Environ. Sci. T echnol. 31, 732-734. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 118 Environmental Risk Assessment of T rifluoroacetic Acid Feenstra-Bieders, G.C. and O lthof, J.A. 1992. Determination of the solubility of trifluoroacetic acid and sodium trifluoroacetate in various aqueous media. Solvay Duphar Report 56630/ 182/ 91. Ford, D.J., Coyle, D.E., and H arrington, J.F. 1984. Effects of hypersensitivity to a halothane metabolite on halothane-induced liver damage. A nesthesiol. 60(2), 141-143. [C.A. 100:132470w] Frank, H . 1994. Determination of phytotoxic haloacetic acids in the environment. AFEAS Workshop on the Environmental Fate of T rifluoroacetic Acid, M arch 3-4, 1994, M iami Beach, FL . AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. Frank, H . 1997. Determination of T rifluoroacetic Acid (T FA): M ethods, Validations and Assessment of Environmental O ccurrence. Final Report to AFEAS by U niv. Bayreuth (Project SP91-18.19/ BP94-26). Frank, H ., Klein, A., and Renschen, D. 1996. Environmental trifluoroacetate. N ature 382, 34. Frank, H ., Renschen, D., Klein, A., and Scholl, H . 1995. T race analysis of airborne haloacetates. J. H igh R esol. C hromatogr. 18, 83-88. Frank, H . and Klein, A. 1997. T rifluoroacetic Acid: Assessment of Environmental Relevance. Final Report to AFEAS by U niv. Bayreuth (Project SP91-18.19/ BP94-26). Franklin, J. 1993. T he atmospheric degradation and impact of 1,1,1,2-tetrafluoroethane (H ydrofluorocarbon 134a). C hemosphere 27(8), 1565-1601. Fraser, J.M . and Kaminsky, L .S. 1988. 2,2,2-T rifluoroethanol intestinal and bone marrow toxicity: T he role of its metabolism to 2,2,2-trifluoroacetaldehyde and trifluoroacetic acid. T oxicol. A ppl. P harmacol. 94, 84-92. Freist, F. 1986. Weiterfuhrende U ntersuchungen bezuglich der Pflanzenschadigungen in der U mgebung des Werkes Wimpfen. Kali Chemie H annover Versuchsbericht. Ghantous, H ., Parnerud, I., Danielsson, B.R.G., and Dencker, L . 1986. Distribution of halothane and the metabolites trifluoroacetic acid and bromide in the conceptus after halothane inhalation by pregnant mice. A cta P harmacol. T oxicol. 59, 370-376. Grimvall A., Born, H ., von Sydow, L ., and L aniewski, K. 1997. Analysis of T FA in samples of rain, snow and ice collected at remote sites. Presentation to AFEAS, Washington, DC. Groeneveld, A.H .C., de Kok, H .A.M ., and van den Berg, G. 1992. T he toxicityof sodium trifluoroacetate to the alga Selenastrum capricornutum. T echnical Report 56635/ 52/ 92, Solvay Duphar, Weesp, T he Netherlands. H alocarbon. 1967. T rifluoroacetic acid brochure. H alocarbon Products Corporation, H ackensack NJ. H arris, J.W., Pohl, L .R., M artin, J.L ., and Anders, M .W. 1991. T issue acylation by the chlorofluorocarbon substitute 2,2-dichloro-1,1,1-trifluoroethane. P roc. N atl. A cad. Sci. 88, 1407-1410. H aszeldine, R.N. and N yman, F. 1959. O xidation of polyhalogen compounds. I I. Photolysis and photochemical oxidation of some chlorofluoroethanes. J. A m. C hem. Soc. 387-396. H enne, A.L . and T rott, P. 1947. Improved preparation of trifluoroacetic acid. J. A m. C hem. Soc. 69, 1820. H oladay, D.A. and Cunnah, R. 1976. Distribution and elimination of trifluoroacetic acid in man. Paper presented at the Annual M eeting American Society of Anesthesiologists. [As cited in: H oladay, D.A. (1977); Absorption, biotransformation, and storage of halothane. Environ. H ealth P erspectives 21, 165-169.] Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 119 Boutonnet H uber, M .L . 1955. U .S. Patent 2,721,199 to E.I. DuPont de Nemours & Co., Inc., O ct. 18, 1955. Ingle, L .M . 1968. Response of wheat and tomato seedlings to seven halogenated aliphatic acids. Proc. West Virginia. Acad. Sci., Biological Section 40, 1-11. IPCC (Intergovernmental Panel on Climate Change). 1996a. Climate Change 1995 : T he science of climate change. Contribution of Working Group 1 to the Second Assessment Report of the IPCC, H oughton, J.T ., M eira Filho, L .G., Callander, B.A., H arris, N., Kattenberg, A., and M askell, K. Eds., T able 2.2 pp. 92-93, U niversity Press, Cambridge, U K, 1996. IPCC (1996b) ibid., T ables 2.5d and 2.5e, pp. 100-102. Just, W.W., Gorgas, K., H artl, F.U ., H einemann, P., Salzer, M ., and Schimassek, H . 1989. Biochemical effects of zonal heterogeneity of peroxisome proliferation induced by perfluorocarboxylic acids in rat liver. H epatology (Baltimore) 9(4), 570-581. Kanakidou, M ., Dentener, F.J., and Crutzen, P.J. 1995. A global three-dimensional study of the fate of H CFCs and H FC-134a in the troposphere. J. G eophys. R es. 100(D9), 18781-18801. Kawahara, M ., Akita, S., T akeshita, T ., Fujii, K., and M orio, M . 1988. Salivary excretion of trifluoroacetic acid (T FAA) after halothane anesthesia. J. A nesthesia 2, 161-164. Kheilo, L .S. and Kremneva, S.N. 1966. Comparative assessment of the toxicity of trifluoroacetic acid and pentafluoroproprionic acids. G igiena T ruda i P rof. Zabol. 10(3), 13-17. Kinoshita, H . 1989. Enterohepatic circulation of trifluoroacetic acid. A pharmacokinetic study. M asui to Sosei ( H iroshima J. A n aestesia) 25(2), 155-164. Ko, M .K.W., Kothamarthi, V.R., Sze, N.D., Rodriguez, J., and T romp, T .K. 1996. Potential Deposition of T rifluoroacetic Acid Near U rban Areas. Final Report to AFEAS by Atmospheric & Environmental Research, Inc. (Project SP91-18.22/ BP94-25). Kolb, C.E., Worsnop, D.R., and Rattigan, O . 1995a. Removal Processes -- Summary of Discussions, Chap. 4 of NASA, NO AA and AFEAS (1995). Kolb, C.E., Worsnop, D.R., Zahniser, M .S., De Bruyn, W.J., Shorter, A., and Davidovits, P. 1995b. H eterogeneous atmospheric chemistry of alternative halocarbon oxidation intermediates, Chap. 6 of H alon Replacements, American Chemical Society, Washington, DC, 1995. L ikens, G.E., T artowski, S.L ., Berger, T .W., Richey, D.G., Driscoll, C.T ., Frank, H .G. and Klein, A. 1997. T ransport and fate of trifluoroacetic acid in upland forest and wetland ecosystems. Accepted by P roc. N atl. A cad. Sci. U SA 94, 4499-4503. L loyd, S.C., Blackburn, D.M ., and Foster, P.M .D. 1986. T rifluoroethanol and its oxidative metabolites: Comparison of in vivo and in vitro effects in the rat testis. Fd. C hem. T oxicol. 24(6/ 7), 653-654. L loyd, S.C., Blackburn, D.M ., and Foster. P.M .D. 1988. T rifluoroethanol and its oxidative metabolites: Comparison of in vivo and in vitro effects in the rat testis. T oxicol. A ppl. P harmacol. 92, 390-401. M adronich, S. and Dentener. F. 1995. Dosage Delivery of U ltimate Products -- Summary of Discussions. Chap. 5 of NASA, NO AA and AFEAS (1995). M aruthamuthu, P. and H uie, R.E. 1994. Free radical and photo-Kolbe reactions. Proceedings of a Workshop on Decomposition of T FA in the Environment, February 8- 9, 1994, Washington DC. Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. M aruthamuthu, P. and H uie, R.E. 1995. Ferric ion assisted photoxidation of haloacetates. C hemosphere 30(11), 2199-2207. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 120 Environmental Risk Assessment of T rifluoroacetic Acid M aruthamuthu, P., Padmaja, S., and H uie. R. 1995. Rate constants for some reactions of free radicals with haloacetates in aqueous solution. I nt. J. C hem. K inetics 27, 605- 612. M atheson, L .J., Guidetti, J.R., Visscher, P.T ., Schaefer, J.K., and O remland, R.S. 1996. Summary of research results on bacterial degradation of trifluoroacetate (T FA). U SGS O pen File Report 96-219. Prepared in cooperation with the Alternative Fluorocarbons Environmental Assessment Study (AFEAS). Published by U .S. Geological Survey, Information Services, Box 25286, M ail Stop 417, Denver Federal Center, Denver, CO 80225-0046. M athieu, A., DiPadua, D., M ills, J., and Kahan, B. 1974. Experimental immunityto a metabolite of halothane and fluroxene. A naesthesiol. 40(4), 385-390. [C.A. 81:114554m] M cConville, D.J. 1996. H alocarbon grows to stay in race. C hemical W eek 54. M cCulloch, A. 1994. Sources of hydrochlorofluorocarbons, hydrofluorocarbons and fluorocarbons and their potential emissions during the next twenty-five years. E n viron. M on. A ssess. 31, 167-174. M cCulloch, A. 1995a. Future consumption and emissions of hydrofluorocarbon (H FC) alternatives to CFCs: Comparison of estimates using top-down and bottom-up approaches. Environment I nternational 21(4), 353-362. M cCulloch, A. 1995b. M echanisms of atmospheric degradation of H CFCs and H FCs and formation of trifluoroacetic acid. Workshop on the Environmental Fate of T rifluoroacetic Acid, M arch 3-4, 1994, M iami Beach, FL . Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. M cCulloch, A. and Sidebottom, H .W., Eds. 1993. Kinetics and M echanisms for the Reactions of H alogenated O rganic Compounds in the T roposphere, Commission of the European Communities ST EP-H alocside/ AFEAS Workshop, M arch 23-25, 1993, Dublin. Available from AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. M idgley, P.M . 1995. Alternativesto CFCsand their behaviour in the atmosphere, in Volatile O rganic Compoundsin the Atmosphere, H ester, R.E. and H arrison, R.M ., Eds., I ssues in Environmental Science and T echnology, No. 4, T he Royal Society of Chemistry, Cambridge, U K. M irkov, M .I., M orio, M . Kawahara, M ., Yuge, O ., Kinoshita, H ., and Fujii, K., 1988. Excretion of trifluoroacetic acid as a metabolite of halothane in digestive juice. J. A nesthesia 2, 133-138. M ontzka, S.A., M yers, R.C., Butler, J.H ., Elkins, J.W., L ock, L .T ., Clarke, A.D., and Goldstein, A.H . 1996. O bservations of H FC-134a in the remote troposphere. G eophys. R es. L ettrs. 23(2), 169-172. Nagel, E. and O dom, J.M . 1997. T he effect of T FA on growth and nitrogen fixation by nitrogen fixing bacteria. DuPont Report submitted to AFEAS. NASA, NO AA and AFEAS. 1995. Proceedings of the workshop on the atmospheric degradation of H CFCs and H FCs, November 17-19, 1993, Boulder, CO . Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005 U SA. O lson, M .J., Johnson, J.T ., O 'Gara, J.F., and Surbrook, S.E. 1991. M etabolism in vivo and in vitro of the refrigerant substitute 1,1,1,2-tetrafluoro-2-chloroethane. Drug M etab. Dispos. 19(5), 1004-1011. O ram, D.E., Reeves, C.E., Sturges, W.T ., Penkett, S.A., Fraser, P.J., and L angenfelds, R.L . 1996. Recent tropospheric growth rate and distribution of H FC-134a (CF3CH 2F). G eophys. R es. L et. 23(15), 1949-1952. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 121 Boutonnet Patty, F.A. 1963. I ndustrial H ygiene and T oxicology, Vol. II, 2nd revised edit, pp. 1801- 1803. Pehkonen, S.O ., Siefert, R.L ., and H offmann, M .R. 1995. Photoreduction of iron oxyhydroxides and the photooxidation of halogenated acetic acids. Environ. Sci. T echnol. 29, 1215-1222. Peijnenburg, W., Rorije, E., and Wolfe, N.L . 1994. A Q SAR appproach to reductive transformation of T FA. Proceedings of a Workshop on Decomposition of T FA in the Environment, February 8-9, 1994, Washington, DC. Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. Poignant, P. 1957. Recherches sur l'activit phytotoxique de quelques drivs trihalogns de l'acide actique. C .R . A cadmie des Sciences, P hysiologie vgtale 245, 1022- 1024. Reves, J.G. and M cCracken, L .E. 1976a. Failure to induce hepatic pathology in animals sensitized to a halothane metabolite and subsequently challenged with halothane. A nesth. A nalg. 55(2), 235-242. [C.A. 85:386k] Reves, J.G. and M cCracken, L .E. 1976b. H epatic pathology and skin test reactions to trifluoroacetylated autologous protein after repeated halothane anesthesia in the guinea pig. Br. J. A naesth. 48(5), 419-424. [C.A. 85:14068v] Rhne-Poulenc. 1995. M aterial Safety Data Sheet, v. 6 Dec. 1995. Richey, D.G., Driscoll, C.T ., and L ikens, G.E. 1997. Soil retention of trifluoroacetate. Environ. Sci. T echnol. 31, 1723-1727. Rodriguez, J.M ., Ko, M .K.W., H eisey, C.W., T romp, T .K., and Sze, N.D. 1994. M odeling dosage projections for trifluoroacetic acid, Workshop on Decomposition of T FA in the environment, February 8-9, 1994, Washington, DC. Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. Rollins, A., Barber, J., Elliott, R. and Wood, B. 1989. X enobiotic monitoring in Plants by19F and 1H nuclear magnetic resonance imaging and spectroscopy: U ptake of trifluoroacetic acid in L ycopersicum esculentum. P lant P hysiol. 91, 1243-1246. Rosenberg, P.H . 1971. Decrease in reduced glutathione and NADPH and inhibition of glucose-6-phosphate dehydrogenase activity caused by metabolites of fluoroxene and halothane. A nn. M ed. Exp. Biol. Fenn. 49(2), 84-88. [C.A. 75:86904f] Rosenberg, P.H . and Wahlstrom, T . 1971. H epatotoxicity of halothane metabolites in vivo and inhibition of fibroblast growth in vitro. A cta P harmacol. T oxicol. 29(1), 9-19. [C.A. 74:98048x] Satoh, H ., M artin, B.M ., Schulick, A.H ., Christ, D.D., Kenna, J.G., and Pohl, L .R., 1989. H uman anti-endoplasmic reticulum antibodies in sera of patients with halothane-induced hepatitis are directed against a trifluoroacetylated carboxylesterase. P roc. N atl. A cad. Sci. U SA 86, 322-326. Satoh, H ., Fukuda, Y., Anderson, D.K., Ferrans, V.J., Gillette, J.R., and Pohl, L .R. 1985. Immunological studies on the mechanism of halothane-induced hepatotoxicity: immunohistochemical evidence of trifluoroacetylated hepatocytes. J. P harmacol. Exp. T her. 233(3), 857-862. Sefchick, J., Skorupa, J., and Schwarzbach, S. 1994. Review of evaporative concentration potential of conservative solutes in selected wetland systems. Final Report to AFEAS by U .S. Fish and Wildlife Service (SP91-18.6/ BP93-22). Smyth, D.V., T hompson, R.S., and Gillings, E. 1993. Sodium trifluoroacetate: T oxicity to the duckweed (L emna gibba). Brixham Environmental L aboratory Report BL 4872/ B. Smyth, D.V., T hompson, R.S., and Gillings, E. 1994a. Sodium trifluoroacetate: T oxicity to the freshwater diatom N avicula pelliculosa. Brixham Environmental L aboratory Report BL 4968/ B. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 122 Environmental Risk Assessment of T rifluoroacetic Acid Smyth, D.V., T hompson, R.S., and Gillings, E. 1994b. Sodium trifluoroacetate: T oxicity to the marine alga Skeletonema costatum. Brixham Environmental L aboratory Report BL 4980/ B. Smyth, D.V., T hompson, R.S., and Gillings, E. 1994c. Sodium trifluoroacetate: T oxicity to the blue-green alga A nabaena flos-aquae. Brixham Environmental L aboratory Report BL 4967/ B. Standley, L .J. and Bott, T .L . 1994. Distribution and Radiolabel in Biomolecules of aquatic organisms exposed to 14C T FA, an atmospheric breakdown product of hydrofluorocarbon refrigerants. Final Report to AFEAS by Stroud Water Research Center (Project SP91-18.13/ BP93-21). Standley, L .J. and Bott, T .L . 1995. Distribution and radiolabel in biomolecules of aquatic organisms exposed to 14C T FA, an atmospheric breakdown product of hydrofluorocarbon refrigerants. Abstract for a poster presentation for the annual meeting of Society of Environmental T oxicology and Chemistry, Denver, CO . Standley, L .J. and T .L . Bott. Biomolecular incorporation of trifluoroacetate, an atmospheric breakdown product of hydrofluorocarbon refrigerants, by aquatic organisms. Submitted to Environ. Sci. T echnol. Stier, A., Kunz, H .W., Walli, A.K., and Schimassek, H . 1972. Effects on growth and metabolism of rat liver by halothane and its metabolite trifluoroacetate. Biochem. P harmacol. 21, 2181-2192. T hompson, R.S. 1995. Sodium trifluoroacetate: T oxicity to P lantago major by aqueous exposure of the roots. Brixham Environmental L aboratory Report BL 5474/ B. T hompson, R.S. and Gillings, E. 1996. Characterisation of accumulated residues in leaf tissue of Sunflower (H elianthus annuus). Brixham Environmental L aboratoryReport BL 5648/ B. T hompson, R.S. and Windeatt, A.J. 1994. Sodium trifluoroacetate: Effects on seed germination (ten species, by aqueous exposure). Brixham Environmental L aboratory Report BL 5197/ B. T hompson, R.S., Stewart, K.M . and Gillings, E. 1994. T rifluoroacetic acid: Accumulation from aqueous solution by the roots of Sunflower (H elianthus annuus). Brixham Environmental L aboratory Report BL 5042/ B. T hompson R.S., Stewart, K.M ., and Gillings, E. 1995. Sodium trifluoroacetate: T oxicity to wheat (T riticum aestivum) in relation to bioaccumulation (by aqueous exposure of the roots). Brixham Environmental L aboratory Report BL 5473/ B. T hus, J.L .G. 1997. Addendum report: determination of the partition coefficient of trifluoroacetic acid. Solvay Duphar Report 56834/ 45/ 97. T hus, J. and van Dijk, C. 1996. Solvay Duphar Environmental Research Report JL GT / ik/ ERD/ fax/ 96.135. T romp, T .K., Ko, M .K.W., Rodriguez, J.M ., and Sze, N.D. 1995. Potential accumulation of a CFC-replacement degradation product in seasonal wetlands. N ature 376, 327- 330. T uazon, E.C. and Atkinson, R. 1995. T ropospheric degradation products of hydrochlorofluorocarbons and hydrofluorocarbons, Chap. 5 of H alon Replacements, American Chemical Society, Washington, DC. van der L eeden, F. 1975. Water Resources of the World, Water Information Center Inc., Port Washington, NY. van Dijk, N.R.M . 1992a. U V-Vis absorption of trifluoroacetic acid. Solvay Duphar Report 56635/ 85/ 91. van Dijk, N.R.M . 1992b. Adsorption of sodium trifluoroacetate (NaT FA) to three different soils. Solvay Duphar Report 56635/ 91/ 91. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 123 Boutonnet van Dijk, N.R.M . 1996. Possible biotransformation of sodium trifluoroacetate by the alga Selenastrum capricornutum. Solvay Duphar Report 56834/ 13/ 96. van Ginkel, C.G. and A.G.M . Kroon. 1992a. Biodegradability of trifluoroacetic acid sodium salt in a modified SCAS test. Report to Akzo Chemicals International, Akzo Research L aboratory, Valperweg 76, 6824 BM Arnhem, T he Netherlands. van Ginkel, C.G. and A.G.M . Kroon. 1992b. Cometabolic transformation of trifluoroacetic acid, sodium salt, in batch cultures. Report to Akzo Chemicals International, Akzo Research L aboratory, Valperweg 76, 6824 BM Arnhem, T he Netherlands. Vinegar, A., Williams, R.J., Fischer, J.W., and M cDougal, J.N. 1994. Dose-dependent metabolism of 2,2-dichloro-1,1,1-trifluoroethane: A physiologically based pharmacokinetic model in the male Fischer 344 rat. T oxicol. A ppl. P harmacol. 129, 103-113. Visscher, P.T ., Culbertson, C.W., and O remland, R.S. 1994. Degradation of trifluoroacetate in oxic and anoxic sediments. N ature 369, 729-731. Waldron, B.A. and Ratra, CK. 1972. Response of lymphocytes cultured after halothane anesthesia to halothane and trifluoroacetic acid. Brit. J. A naesth. 44(11), 1150-1154. [C.A. 78:67170j] Wallington, T .J., H urley, M .D., Fracheboud, J.M ., O rlando, J.J., T yndall, G.S., Sehested, J., M gelberg, T .E., and N ielsen, O .J. 1996. Role of excited CF3CFH O Radicalsin the Atmospheric Chemistry of H FC-134a. J. P hys. C hem. 100(46), 18116-18122. Warheit, D.B. 1993. M echanistic studies with H CFC-123. Dupont Company, H askell L aboratory Report H L R 828-92. Waskell, L . 1978. A study of the mutagenicity of anesthetics and their metabolites. M ut. R es. 57, 141-153. Williams, J. 1997. I n vitro studies on the effects of fluorocarbons H CFC 123, H FC 134a, H CFC 141b and certain potential metabolites of H FC 134a in testicular cell cultures. Zeneca Central T oxicology L aboratory Report CT L / E/ 187. Windeatt, A.J. and T hompson, R.S. 1993a. Sodium trifluoroacetate: Determination of its effect in soil on seed germination and earlyplant growth of Sunflower (H elianthus annuus) and M ung Bean (P haseolus aureus). Brixham Environmental L aboratory Report BL 4920/ B. Windeatt, A.J. and T hompson, R.S. 1993b. Sodium trifluoroacetate: Determination of its effect in soil on seed germination and early plant growth of wheat (T riticum aestivum). Brixham Environmental L aboratory Report BL 4761/ B. Wine, P. 1994. Cloud chemistry of T FA. Proceedings of a Workshop on the Decomposition of T FA in the Environment, February 8-9, 1994, Washington, DC. Published by AFEAS, T he West T ower -- Suite 400, 1333 H Street NW, Washington, DC 20005. WM O (World M eteorological O rganisation). 1989. Scientific assessment of stratospheric ozone: 1989. WM O Global O zone Research & M onitoring Project Report No 20, Volume II, WM O , PO Box 2300, Geneva, CH . WM O . 1994. Scientific assessment of ozone depletion: 1994. WM O Global O zone Research & M onitoring Project Report No 37, WM O , PO Box 2300, Geneva, CH . Zehavi, D., Seiber, J.N., and Wujcik, C. 1995. Development and application of a method for analysing trifluoroacetic acid in environmental samples. A mer. C hem. Soc. Environmental C hem. Bull. 35(1), 190-191. Zehavi, D. and Seiber, J.N. 1996. An analytical method for trifluoroacetic acid in water and air samples using headspace gas chromatographic determination of the methyl ester. A nal. C hem., 68, 3450-3459. Zellner R., Bednarek, G., H offmann, A., Kohlmann, J.P., M rs, V., and Saathoff, H . 1994. Rate and mechanism of the atmospheric degradation of 2H -heptafluoropropane (H FC-227). Ber. Bunsensges. P hys. C hem. 98(2), 141-146. Copyright 1999, CRC Press L L C -- Files may be downloaded for personal use only. Reproduction of this material without the consent of the publisher is prohibited. H um. Ecol. Risk Assess. Vol. 5, No. 1, 1999 124