the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Trifluoroacetate (TFA) in Precipitation and Surface Waters in Switzerland: Trends, Source Attribution, and Budget
Abstract. Sources and budgets of the persistent, anthropogenic compound trifluoroacetate (TFA) are poorly quantified across different environmental media. Recently, introduction of hydrofluoroolefins and continued use of other fluorinated compounds has increased environmental levels of TFA. Here, we present concentrations of TFA observed in precipitation and surface waters in Switzerland during three years of continuous monitoring and in archived samples, collected since 1984. Atmospheric simulations attribute TFA to precursor gases. Mean observed TFA concentrations were 0.30 to 0.96 μg L-1 across 14 precipitation sites and 0.33 to 0.88 μg L-1 across 9 river sites in 2021–2023 – four-to-six-fold increase since 1996/1997. Simulated atmospheric degradation of known TFA precursors accounted for 60–70 % of observed deposition (40–54 % hydrofluoroolefins and 12–17 % long-lived fluorinated gases). Atmospheric deposition amounted to 24.5±9.6 Mg yr-1. TFA terrestrial inputs in Switzerland from plant protection products (PPP) and veterinary pharmaceuticals, estimated from literature, ranged from 3.9 to 13.2 Mg yr-1. These inputs were balanced by exports through the major rivers, 31±4 Mg yr-1. In croplands, TFA inputs from the degradation of PPP were 2.5–3 times larger than those from atmospheric deposition. Archived precipitation samples revealed that TFA was formed in the atmosphere before the introduction of known atmospheric precursors, whereas in the 1990ies TFA deposition increased along with precursors. However, simulated atmospheric degradation underestimates summertime deposition five-fold. Continued use of fluorinated compounds is likely to enhance TFA deposition in the future. Additional environmental monitoring and source attribution studies are paramount for refining the risk assessment of TFA.
- Preprint
(11329 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 08 Aug 2025)
-
RC1: 'Comment on egusphere-2025-2861', Anonymous Referee #1, 09 Jul 2025
reply
The authors discuss the trends, budget and deposition of TFA in Switzerland. The study is very comprehensive, including measurements of TFA in rainwater, rivers, lakes, modelling of the deposition of TFA from fluorinated gases, and estimated of TFA from pharmaceuticals and plant protection products. The paper has a good and extensive introduction of TFA. The methods, measurements and model calculations are described in great detail. The results are also described in detail with all their uncertainties and caveats. I want to compliment the authors with this paper, well done. I only have some smaller textual comments.
Some specifics comments on the text:
L9: Specify the region the deposition of 24.5 Mg applies to.
L11-12: “In croplands … deposition”. I suggest removing this sentence. Although interesting, it disrupts the flow of the abstract and is does not add relevant information for the remainder of the abstract.
L12: It is not clear how old the “archived samples” are. Are they from the 1990s?
L17: You write about the risk assessment, but later in the paper you mention that there is no consensus of the risks of TFA (L81). I suggest you rephrase the sentence, e.g. “for refining the assessment of TFA sources for potential health and environmental risks.”.
L53-58: This a very long sentence and therefore hard to read. Please rephrase.
L474: Data from WWTP could indeed be valuable, but are they a new source not accounted for yet? Do the WWTPs not discharge their water on the rivers?
L657: “The gap in explained deposition…”. Where the gap refers to is not clear from the previous sentences. Please add some text here.
P30, table 2: There is a typo in the compound name: It should be HFC-43-10mee, not HFC-41-10mee
Citation: https://doi.org/10.5194/egusphere-2025-2861-RC1 -
RC2: 'Comment on egusphere-2025-2861', Anonymous Referee #2, 14 Jul 2025
reply
The authors present a comprehensive assessment of TFA concentrations recorded in precipitation and surface waters in Switzerland 2021-2023. They compare these measurements with those taken from archived water samples and demonstrate significant increases in TFA concentrations in recent years. They also conduct a number of modelling studies to determine TFA deposition from the atmosphere and determine that this cannot be the sole driver of the observed TFA increases. They propose terrestrial inputs from PPP and veterinary pharmaceuticals as a more significant contributor to TFA contamination in surface waters. They also suggest an atmospheric TFA source that predates the introduction of known precursors (though it seems they are suggesting a further anthropogenic source, in which case this should be made clearer in the abstract).
The report is generally well written and presents an important study into TFA contamination at a national level. However, there a few areas that need further discussion or that have been omitted. Firstly, there should be full justification of the veracity of the archived water samples as these form a major basis for many of the conclusions drawn. Secondly, the relationship between CF3CHO and TFA formation has been omitted from this study. Given the recent literature on the potential impacts of CF3CHO on TFA concentrations and the uncertainty it introduces to TFA yields etc. it should at least be discussed and could go some way to closing the gap seen between the modelled deposition and the measurements.
More generally, tables and figures could be better placed throughout the manuscript (i.e. near the text the references them) and more accessible for those with colour-blindness (though I appreciate this is difficult to achieve with figures such as those presented here).
Specific comments:
Line 7: unnecessary ‘-‘
Line 8: Where does the 60-70% come from? From the other % range seems like it should be 52-71%?
Line 11: Presumably this value refers to total atmospheric deposition (wet + dry?)
Line 14: 1990s instead of 1990ies (also occurs …)
Line 24: There are some reports that suggest TFA may bioaccumulate in plant material
Line 39: ‘As a consequence’ as opposed to ‘In consequence’, ‘potential risk’ as opposed to ‘risk’
Line 56: () within ()
Line 117: Is the conducted test sufficient to prove the veracity of the archived samples?
Line 169: Error of 10% for the emissions estimates – please justify this number?
Line 197: You should justify why the model was run for 60 days, as from a trajectory viewpoint it seems far too long to give a reliable distribution
Line 241: What is the timescale of hydrolysis of TFF to TFA?
Line 247: Why did you use The Henry’s Law constant of HNO3 when there are reported values for TFA? There is uncertainty in the TFA value that may mean it diverges from the HNO3 value. Given the importance of this parameter in wet deposition and the importance of this loss to the atmospheric fate of TFA, the uncertainty of this would be a useful thing to explore. Regardless, the value used should be given in the text with ref
Line 255-275: There should be mention of recent publications relating to CF3CHO that dispute TFA yields, specifically the 4-60% value for HFO-1336mzzZ, and introduce extra uncertainty that hasn’t been accounted for in this study as it stands
Line 274: Typo ‘emissions’
Line 280: Are the samples below the estimated quantification limit considered in the average? How can you quantify a value below the limit of quantification? What is the limit of detection? How is the limit of quantification derived?
Line 281: Is this 0.58 ± 0.58 correct?
Line 283-303: Are you using T-dependent rate constants for HFO+OH? Some of the HFOs you are considering have a negative relationship between k and T
Line 311: Could deposition not also be a driver of load seasonality (in addition to discharge seasonality)?
Line 327-329: ‘underlines’ and ‘underlining’ should be replaced with ‘supports’ and ‘reinforcing’
Line 330-334: Does this evidence not serve as an argument against pre known-precursor TFA?
Line 343: ‘No significant TFA gradients with depths were detected in the other lakes’ so concentrations at depth were the same as the surface? Were the values elevated?
Line 383: Replace ‘the latter’ with ‘this’
Line 391: ‘large HFO emissions in the Po Valley’ ? Are these emissions simulated in the model at an appropriate resolution to be able to make this claim?
Line 400: ‘compared with’ instead of ‘compared to’
Line 409: Remove ‘as well’ and replace with also earlier in the sentence
Line 415: It also suggests that if these unknown sources exist, they are likely to be anthropogenic
Line 426: What is the estimated combined uncertainty of the simulated TFA deposition? It should be reported here along with the specific contributions of the TFA yield vs the precursor emissions
Line 430-431: This sentence ‘faster (slower) … increased (decrease)’ is not worded clearly.
Line 448: ‘which is in the range smaller 0.3%’ does not make sense and there is also an unclosed ‘(‘
Line 451: ‘For non-atmospheric sources of TFA’
Line 484: Need an ‘is’ between ‘discharge’ and ‘below’
Line 489: This should be more specific e.g. ‘PPPs may be a more relevant source of TFA to environmental aqueous phases (i.e rivers and lakes) in Switzerland’
Line 494: Remove () from reported flux to be consistent
Line 503: Missing ‘(‘
Line 511: TFA in rainwater in Bayreuth?
Line 514-518: Clarify that you’re talking about TFA concentration in rainwater and in archived plant material
Line 528: Was 2019 when HFO use started to increase exponentially? If yes, I would explicitly make this link
Line 542: You should be specific as to what other sources you are suggesting (agricultural run-off? Natural sources?)
Line 548: Do these processes release fluorocarbons into the environment via wastewater etc only or is there a potential gaseous emission?
Line 627: Also there’s massive uncertainty in the yields?
Figure 2 & 3: Additional x-ticks representing the months or every 3 months would make these figures easier to interpret where variations are not exactly halfway through the year
Table 2 & 3: Should present the 3 yields used rather than the range, lifetime of ‘1’ should be reported as ‘1.0’ for consistency, last column in each needs () around units
Citation: https://doi.org/10.5194/egusphere-2025-2861-RC2 -
CC1: 'Comment on egusphere-2025-2861', Tim Wallington, 14 Jul 2025
reply
Henne et al. report the results of a comprehensive set of measurements of trifluoroacetate (TFA) in surface waters in Switzerland. There is substantial research interest in the sources, concentrations, and effects of trifluoroacetate in the environment. This paper is an important contribution to better understanding the levels, sources, and trends of trifluoroacetate in the environment. We have the following comments for the authors to consider.
First, the United Nations Environmental Effects Assessment Panel of which some of us are members has assessed that the risks to human and ecosystem health from trifluoroacetate formed as a degradation product of ODS replacements are currently de minimis (Neale et al., 2025). For balance in the introduction where the toxicological effects are discussed the authors may wish to mention the assessment of the UNEP panel and note the need for further work to reconcile the divergent assessments in the literature.
Second, there are new measurements published by the German UBA (Umweltbundesamt) of trifluoroacetate in samples from the Atlantic Ocean collected from surface waters (n = 33) and from seven distinct depth profiles (n = 41) in 2022–23. Using these data, Neale et al. estimated that the mass of TFA measured in the oceans in the late 1990s and early 2000s, assuming even distribution of 200 ng L−1, was about 500–1000 times higher than the estimated total anthropogenic TFA input to the environment (including Montreal Protocol gases, pesticides, pharmaceuticals, and industrial uses) over the period 1930–1999. The evidence for the contribution of one or more natural source(s) of TFA to the marine environment is relevant and should be mentioned.
Third, using the flux of trifluoroacetate (TFA) measured at several locations on the Rhine River in the Netherlands (2017-2023), an average flux into the Rhine basin of ca 0.5 kg TFA km-2 yr-1 can be estimated (Neale et al., 2025). This is of a similar magnitude to that estimated in Switzerland by Henne et al. There was no discernable trend in the flux of TFA in 2017-2023 suggesting either that degradation of HFO-1234yf was not a major contributor to TFA in the Rhine basin for 2017–2023, or that its increasing contribution was masked by compensating decreases in contributions from other sources. A discussion of how their findings compare with those of Neale et al. (2025) would be a useful addition.
Fourth, it should be possible to propagate the error bars for the relevant parameters for both the precursor measurements and the TFA measurements/modelling results. This would allow for better comparison of the contributions to TFA accounted for and the “unaccounted” remainder. On the same note, greater clarity on how the TFA deposition fluxes from the individual precursors were calculated would be beneficial for the reader, for example which molar yield of TFA from HFO-1233ze was used for the calculations. This combined information would be informative and most interesting. Possibly, it could hint at additional atmospheric sources of TFA or lacking/incorrect understanding of the atmospheric oxidation chemistry involved.
Fifth, the statement in the conclusions “Therefore, it is fundamental to continue efforts to abandon the use of fluorinated compounds, wherever possible, to avoid further, continued accumulation of TFA. Both industry and policy makers are called to increase their level of ambition” is very generalized and simplistic. It fails to consider that not all fluorinated compounds degrade to produce TFA and that the stoichiometry is such that yields are not always molar. It also fails to recognize that the CF3- group acts as a pseudo halogen that increases the efficacy of pharmaceuticals and pesticides to the direct and indirect benefits of humans and the environment. As discussed in Neale et al. (2025), TFA in the environment is present as salts that are highly water soluble and easily excreted. TFA-salts do not biomagnify in food webs and there are no known biochemicals or receptors that interact with TFA, although it is a moderately strong acid (pKa = 0.23), it is unreactive. There are wide margins of safety between current and predicted future concentrations in surface- and ground-waters levels of concern for human and environmental health. While continuous monitoring would be useful in quantifying future rates of change in concentrations, this should be focused on key matrices and should include measurements of systemic doses in the general population, such as those conducted in the NHANES program [1].
T. J. Wallington, S. Madronich, O. J. Nielsen, K. R. Solomon, M. P. Sulbaek Andersen
References
Neale et al. Environmental consequences of interacting effects of changes in stratospheric ozone, ultraviolet radiation and climate: UNEP Environmental Effects Assessment Panel, Update 2024, Photochem. Photobiol. Sci., 24, 357 (2025). https://doi.org/10.1007/s43630-025-00687-x
UBA. (2024). Untersuchung von aktuellen Meerwasserproben auf Trifluoressigsäure. TZW: DVGW-Technologiezentrum Wasser, Karlsruhe, Umweltbundesamt, Dessau-Roßlau, Germany. https://www.umweltbundesamt.de/publikationen/untersuchung-von-aktuellen-meerwasserproben-auf
[1] https://www.cdc.gov/exposurereport/data_tables.html
Citation: https://doi.org/10.5194/egusphere-2025-2861-CC1
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
197 | 64 | 11 | 272 | 6 | 11 |
- HTML: 197
- PDF: 64
- XML: 11
- Total: 272
- BibTeX: 6
- EndNote: 11
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1