Hydrochlorofluorocarbons (HCFCs) and
hydrofluorocarbons (HFCs) are used as CFC substitutes in a variety of applications.
A substantial body of experimental and theoretical work has been undertaken
to determine the atmospheric chemistry and environmental impact of these
compounds. While some minor uncertainties exist, our current understanding
of the atmospheric chemistry of the commercially important HFCs and HCFCs
is now well established (Wallington et al., 1994; 1995). A generalized
scheme for the atmospheric oxidation of a haloalkane that could degrade
to give trifluoroacetic acid is given in Figure 6.4.
|Fig. 6.4. Generalized scheme for the atmospheric oxidation of a halogenated organic compound, CF3CXYH (X = Cl or F, and Y = Cl, H, Br, or CF3). Radical intermediates are enclosed in ellipses. Typical lifetime estimates are given in parenthesis.|
Oxidation is initiated by reaction with OH radicals giving a halogenated alkyl radical which adds O2 to give the corresponding peroxy radical (RO2). Peroxy radicals react with three important trace species in the atmosphere: NO, NO2, and HO2 radicals. Reactions with HO2 and NO2 delay, but do not prevent, the conversion of peroxy (CF3CXYO2) into alkoxy (CF3CXYO) radicals. Reactions of haloperoxy radicals with NO are rapid and give the alkoxy radical with essentially 100% yield. The atmospheric fate of the alkoxy radical, CF3CXYO, is either decomposition or reaction with O2. Decomposition occurs by C-C bond fission, or by the elimination of a Br, Cl, or CF3 group. The atmospheric fate of CF3C(O)X (X=F or Cl) is dominated by incorporation into rain-cloud-sea water followed by rapid hydrolysis to trifluoroacetic acid. Photolysis is a competing loss mechanism for CF3C(O)Cl and limits its conversion into CF3C(O)OH to 60% (Cox et al., 1995). There are no competing loss processes for CF3C(O)F; it is converted entirely into CF3C(O)OH. Although CF3C(O)OH is produced in aqueous phase chemistry, is highly soluble and partitions into the water phase (Bowden et al., 1996), the evaporation of cloud droplets can transfer CF3C(O)OH to the gas phase where it can react with OH radicals. However, this reaction is slow (Carr et al., 1994; Møgelberg et al., 1994) and is only a minor (<5%, Kanakidou et al., 1995) loss of CF3C(O)OH. The main atmospheric fate of CF3C(O)OH is rain-out to the surface.
To assess the potential for a halocarbon to produce trifluoroacetic acid, it is necessary to quantify the yield of trifluoroacetyl halide in its gas phase oxidation mechanism. This has been established for all the commercially significant halocarbons. Six compounds, listed in Table 6.2, have been identified that degrade to give trifluoroacetyl halide. With the exception of HFC-134a, the atmospheric oxidation mechanisms of the compounds given in Table 6.2 are relatively simple. Atmospheric oxidation of halothane (Bilde et al., 1998), isoflurane, and HCFC-123 (Edney et al., 1991; Tuazon and Atkinson, 1993a; Hayman et al., 1994) gives CF3C(O)Cl of which approximately 60% undergoes hydrolysis to give CF3C(O)OH. Oxidation of HCFC-124 (Tuazon and Atkinson, 1993a; Edney and Driscoll, 1992) and HFC-227ea (Zellner et al., 1994; Møgelberg et al., 1996) gives CF3C(O)F which is then converted into CF3C(O)OH. Current and projected use of HFC-227ea is minor. The oxidation mechanism of HFC-134a is complicated by two factors. First, vibrationally excited CF3CFHO* radicals are formed in the CF3CFHO2 + NO reaction and two thirds of these vibrationally excited radicals undergo decomposition via C-C bond scission on a time scale short compared to that needed for chemical reactions (Wallington et al., 1996). Second, reaction with O2 competes with decomposition via C-C bond scission for the available thermalized CF3CFHO radicals (Wallington et al., 1992; Tuazon and Atkinson, 1993b; Rattigan et al., 1994). The net effect is that the molar yield of CF3C(O)F, and hence CF3C(O)OH, is approximately 0.13 (Wallington et al., 1996).
An estimate for the concentration of CF3C(O)OH in rain water can be obtained as follows. Halothane and isoflurane have been used as anesthetics for many years. It is estimated that the current emissions of halothane and isoflurane are 1500 and 750 tonnes/yr, respectively (Boutonnet et al., 1998). The atmospheric lifetimes of halothane and isoflurane are short (see Table 6.2) compared to the time scale over which they have been emitted into the atmosphere. Assuming that these compounds are in steady state then their degradation gives a combined yield of 1540 tonnes/year of CF3C(O)Cl. Accounting for photolysis, this results in a global deposition rate of 800 tonnes/year of CF3C(O)OH. Current atmospheric concentrations of HCFC-123 and HCFC-124 are at, or below the detection limit of 0.1 ppt (Boutonnet et al., 1998). Thus, upper limits for atmospheric burdens of these gases are 2500 tonnes for HCFC-123 and 2300 tonnes for HCFC-124. Combining these burdens with the lifetime values given in Table 6.2 and accounting for photolysis of CF3C(O)Cl it can be concluded that upper limits for the flux of CF3C(O)OH from HCFC-123 and HCFC-124 are 760 and 320 tonnes/year, respectively. HFC-134a is present at a concentration of 2.5 pptv in the Northern Hemisphere and 1.2 pptv in the Southern Hemisphere (Montzka et al., 1996; Oram et al., 1996) from which it can be estimated that the present atmospheric burden of HFC-134a is 31,000 tonnes. Combining this burden with the lifetime and yield values in Table 6.2 gives a CF3C(O)OH flux of 300 tonnes/year from the atmospheric oxidation of HFC-134a. Combining the contributions from these known sources gives an upper limit for the estimated contemporary TFA formation rate of 800 + 760 + 320 + 300 = 2180 tonnes/year. The known TFA precursors are relatively long lived and therefore distributed on a global scale. The annual global rainfall is 4.9 x 1017 liters (Erchel, 1975) and the global average TFA concentration in rainwater is expected to be less than 5 ng/l. Detailed computer modeling studies using appropriate global OH fields, halocarbon emissions, and rainfall patterns have confirmed that the simple approach adopted above provides a reasonable estimate of expected TFA concentrations in rainwater (Rodriguez et al., 1993; Kanakidou et al., 1995; Wild et al., 1996).
To assess the future contribution of HFCs and HCFCs to levels of TFA in rainwater we need to consider future emission scenarios for these compounds which are reviewed in detail elsewhere (Boutonnet et al., 1998). Use of halothane and isoflurane anesthetics is anticipated to remain constant, or perhaps decline. Production of HCFC-123 and HCFC-124 is regulated under the Montreal Protocol and will be phased out. Production of HFC-134a is not regulated and anticipated to increase significantly. Using an emission scenario from the US Environmental Protection Agency (EPA) combined with a three-dimensional atmospheric model, Rodriguez et al. (1994) calculated that in the year 2020 the atmospheric decomposition fluxes of HFC-134a, HCFC-123, and HCFC-124 will be 115, 20, and 20 kTonnes/yr. Multiplying by the appropriate CF3C(O)OH molar yields in Table 6.2 and making the necessary adjustments for the molecular weights involved, a flux of 16.7 + 8.9 + 16.7 = 42.3 kTonnes/yr is estimated and translates to a global average concentration in rainwater of 86 ng/l. Kotamarti et al. (1998) estimated the global average TFA concentration in rain water for 2010 to be in the range 100-160 ng/l.
Table 6.2: Compounds known to produce TFA (CF3C(O)OH) in the atmosphere.
|Compound||Molecular weight||Common name||Molar CF3C(O)OH yield||Atmospheric
|CF3CHClBr||197.5||Halothane||0.6||1.2 years |
|CF3CHClOCHF2||184.5||Isoflurane||0.6||5 years |
|CF3CHCl2||153||HCFC-123||0.6||1.5 years |
|CF3CHFCl||136.5||HCFC-124||1.0||6.0 years |
|CF3CH2F||102||HFC-134a||0.13||14.6 years |
|CF3CHFCF3||170||HFC-227ea||1.0||36.5 years |
 Orkin and Khamagonov (1993);  Brown et al. (1989);  WMO (1989);  IPCC (1996).
Water bodies characterized by little or no outflow and high evaporation rates may have the potential to accumulate TFA. Tromp et al. (1995) developed a concentration-dependent model and estimated that TFA concentrations of 100 mg/l could be achieved in this type of water body in as few as 30 years, with rainfall concentration only 1 mg/l. In contrast, a recent analysis by Boutonnet et al. (1998) questioned the validity of the assumptions inherent in the Tromp et al. study, concluding that accumulation of TFA in seasonal wetlands "appears to be highly improbable". The potential for accumulation of TFA remains unclear.
High concentrations of TFA have been
observed in contemporary water and air samples, suggesting the existence
of one or more large unknown sources. Samples of rain and surface waters
(oceans, rivers, lakes, and springs ) have been obtained from many geographical
areas (USA, Canada, Australia, South Africa, Germany, Israel, Ireland,
France, Switzerland, Finland) and show that TFA is a ubiquitous contaminant
of the hydrosphere (Frank et al., 1996; Zehavi and Seiber, 1996; Grimvall
et al., 1997; Wujcik et al., 1998), with values up to 40900 ng/l
(Zehavi and Seiber, 1996). The average TFA concentration in rain water
observed in Bayreuth during 1995 was 100 ng/l (Frank et al., 1996). The
source of the currently observed levels is unknown and puzzling. The observed
TFA concentrations are orders of magnitude larger than those predicted
to result from the atmospheric degradation of the replacement HCFCs and