the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Apr 2025
Atmospheric and watershed modelling of HFO-1234ze(E) emissions from prospective pressurized metered-dose inhalers usage
Abstract. HFO-1234ze(E) is a next-generation medical propellant in development for use in pressurized metered-dose inhalers (pMDIs). The chemical structure of HFO-1234ze(E) has the ‘–CF3’ moiety, which makes formation of trifluoroacetic acid (TFA) possible in the atmosphere. To quantify the contribution of these novel pMDIs in forming environmental TFA, we performed an extensive study using a global atmospheric model coupled with detailed watershed modelling. Herein, we included the master-chemical mechanism of HFO-1234ze(E), accounting for all known pathways that may form atmospheric TFA and assumed pMDI usage as the only source of HFO-1234ze(E) emissions. Based on annual pMDI sales data and HFO-1234ze(E) as their sole medical propellant, we estimate annual global propellant emissions of 4.736 Gg/year. Even though pMDI sales are the highest in regions within the northern-temperate zone, model-predicted TFA deposition rates are higher in regions within the tropical zone, suggesting that photolysis reaction of trifluoroacetic aldehyde (TFAA; which does not yield TFA) is dominant in the northern-temperate zone. We used model-predicted TFA deposition rates around the Hudson River, Cauvery River, and Rhine River as an input to our fate-and-transport model of TFA, yielding pMDI usage-based TFA concentrations in surface water, soil and sediments in each of the three modelled watersheds. Our watershed models predict that TFA concentrations in river surface water would vary between 0.8–19.3 ng/L, indicating greater than 500-fold margin-of-exposure for drinking-water TFA. Our results demonstrate that environmental TFA formation due to pMDI usage-based HFO-1234ze(E) emissions do not pose a human health concern.
Competing interests: SGT, PG, HK, MG, and SP are employees of AstraZeneca and hold shares/share options in AstraZeneca. KV, KZ, LMD, KT, FK, YZ, and BY are employees of Ramboll. DKP is an employee of Honeywell.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1485 KB)
-
Supplement
(3128 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1485 KB) - Metadata XML
-
Supplement
(3128 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1031', Anonymous Referee #1, 14 May 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1031/egusphere-2025-1031-RC1-supplement.pdfCitation: https://doi.org/
10.5194/egusphere-2025-1031-RC1 -
AC1: 'Reply on RC1', Shivendra Tewari, 23 Jun 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1031/egusphere-2025-1031-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Shivendra Tewari, 23 Jun 2025
-
RC2: 'Comment on egusphere-2025-1031', Anonymous Referee #2, 23 Jun 2025
The manuscript by Shivendra G. Tewari presents a study of the atmospheric degradation of HFO-1234ze(E) and hydrological fate of one of its degradation products, the environmentally persistent trifluoroacetic acid (TFA). The study is specific for the emissions of HFO-1234ze(E), used as a propellant in metered-dose inhalers, and does not discuss emissions of other use cases. Although, the applied atmospheric and hydrological modelling appear appropriate in the first place, there is very little information on validation of the model results in terms of discussion of previous simulations and observations. In addition, the focus on three selected watersheds for the hydrological part of the paper seems very arbitrary and not suited to provide global worst-case scenarios. Hence, the main conclusion of the study, that the use of HFO-1234ze(E) for inhalers will not pose a risk for TFA in drinking water, may not be valid everywhere on the globe and should be revisited. Although, the manuscript is well-structured and written and only a few additional clarifications are required in this respect, addressing the main concerns may require major revisions of the manuscript before publication.
Â
Major comments
Generalisation of results: Although, it is explicitly said that results are derived for three selected watersheds, the final sentence of the abstract and the final paragraph of the conclusions seems to imply that the use of HFO-1234ze(E) in inhalers does to pose a threat globally. The selection of these watersheds is very arbitrary (L227/228). "prominent" and "large population" seem to be the only criteria. However, for all three continents one can very easily come up with watersheds that have a larger population. Figure 4 shows that there are also word regions where we expect larger TFA deposition. However, in the end it is not the deposition flux but the rainwater concentration that will determine surface water concentrations. Hence, I would have expected that a selection of watersheds would focus on those for which largest TFA concentrations in precipitation are predicted. Previous studies have shown that TFA concentrations will be enhanced in more arid regions and, hence, a look at river systems in such areas would be more helpful than considering the 3 current watersheds, from which we cannot conclude that TFA levels will stay within the safety margins globally.
Â
Isolated view on the TFA budget: There is one more danger in the presentation of isolated results of TFA from a single precursor and use case. Obviously, it is the sum of contributions from all precursors that determines environmental TFA levels. If ten studies like the current for specific use cases all come to the conclusion that individually there is no problem, the sum may still present an environmental problem. It is mentioned that compared to current levels of TFA in precipitation in Germany, TFA from degradation of HFO-1234ze(E) in pMDIs would add less than 1 %. Again, this may not be true globally. I would suggest to put the current use case more into the perspective of the global TFA budget (as much as this is known, for example see Madronich et al. 2023).
Â
Emission scenario: It is assumed that future emissions from pMDIs will follow the same global usage as taken from current pMDI sales (L271). Figure 3 reveals that some world regions are underrepresented with this assumption. Especially emissions in densely populated China seem to be unrealistically low. Is this because sales data from China is potentially incomplete? Even if it is complete, would one not expect that access to pMDIs will increase in China in the future? In general, it would be better to work with projected consumption numbers then with present day values. Since Southeast Asia also seems to be an area of intense TFA deposition (Figure 4b), realistic Chinese emissions seem to be critical for a fair assessment of future TFA levels in this region. Furthermore, instead of using NO as a proxy for the spatial distribution of HFO emissions, it seems more appropriate to use population density directly. NO distributions may be skewed by individual point sources like power plants.
Â
Minor comment
Section 2.1-2.3: There are several questions concerning the setup of GEOS-Chem that need clarification. What is the name/version of the utilised chemistry scheme? How many compounds are treated? How are the NMVOC emissions mapped onto model species? Are there any validation results of GEOS-Chem for the classical air pollutants (O3, NOx, …)? Done as part of this study or published elsewhere for the same setup of the model. These would be helpful to understand if simulated OH levels and, hence, HFO reaction rates are realistic.
L153f: How valid is the use of absorption cross sections of CH3C(O)CHO for CF3C(O)C(O)F? Can this be corroborated from similarities from any know cross sections for other similar molecules?
Figure1 and section 2.3: The figure seems to indicate an alternative path for trifluoroacetic aldehyde (TFAA) to TFA through hydration. This path is not discussed in the text. Is this based on the diol mechanism suggested by Franco et al. (2021) for other aldehydes? Please comment if it was included in the chemistry scheme and if it showed any relevance.
Section 2.4: Does the model take evapotranspiration in the watershed into account? Should to be considered in warmer climates.
Figure 4 and Table 3: Earlier model studies on HFO degradation and TFA deposition all came to the conclusion that wet deposition dominates over dry deposition. This was the case for HFO-1234yf, where the formation of TFA should be fast (e.g., Luecken et al., 2010; Henne et al., 2012, Wang et al., 2018), but also for HFO-1233zd(E) (Sulbaek Andersen et al., 2018) for which TFA formation also proceeds through TFAA. Please comment, why and how this could be different in the present case. This also questions the statement made on line 298 concerning 'akin' deposition fluxes as in (Sulbaek Andersen et al., 2018), which then seems oversimplified.
Figure 4: In order to assess the global impact of TFA deposition, it would be beneficial to add another figure that shows average rainwater concentrations of TFA. Since these are usually strongly enhanced during the summer months, I would suggest to show these with three panels: overall average, summer average, winter average. Derived concentrations could be used in the discussion against currently observed TFA in precipitation (see below).
Section 3.1: What was the total global TFA deposition flux? How does it compare to the HFO emissions and what can be concluded in terms of TFA yields for this compound?
Figure 5 and L325f: To me it remains unclear how the conclusions can be reached from what is presented in the figure. Is it only the spatial correlation between the two quantities? But then TFA concentrations will strongly depend on the removal not just the production pathway. Furthermore, how does the correlation look for other intermediates? This requires additional explanation.
Figure 6: The plot reveals another potential shortcoming in the assessment of maximal TFA concentrations. The global chemistry simulations were performed at relatively coarse resolution. However, precipitation and hence TFA deposition often varies at much smaller scales as can be covered by the global chemistry model. As a consequence actual TFA inputs into individual watersheds may largely differ from the grid cell average of GEOS-Chem. Please add a note of caution and discuss the possible implications.
L340: There seems to be another important simplification for the hydrological modelling which needs to be addressed. It is well know that TFA inputs from the atmosphere have a strong seasonal cycle in midlatitudes (both observed and simulated rainwater concentrations show this). In order to assess maximum concentrations in the watersheds it therefore seems very important to consider the seasonality in the inputs and see how this variability propagates through different strata.
L353f and L362: Both statements seem to suggest that a large fraction of TFA will accumulate in the soil. To me it is not clear on which time scales you are discussing this. In steady state (which apparently is reached quickly), input from the atmosphere should be equal to outflow to the ocean. Or is what you call deep soil a open boundary for the model as well? Furthermore, this strong flux to soil seems to contradict the statement in the footnote of Table 1: " Adsortion/desorption tests results show that TFA is poorly adsorbed to the soil and is considered as a mobile organic compound in the majority of soils investigated." Please clarify.
L378: For the Rhine catchment a more direct comparison between simulated atmospheric inputs and measured TFA in precipitation could be done (see Freeling et al., 2020).
L407f: Why not discuss with the often quoted NOEC for the most sensitive freshwater algae, which is 120'000 ng/L? In addition, as TFA cannot be removed from drinking water at large scale, the discussion of an additional threshold much higher than the one suggested for drinking water seems a bit artificial.
Â
Technical comments
Citations in text: Luecken et al.(Luecken et al., 2010) should be Luecken et al. (2010). Applies to all references that should not include the author.
Andersen et al., 2018 and 2022: Should be Sulbaek Andersen et al., 2018. Sulbaek being part of the surname not the given name.
Figure 6: The labels for the color scale of TFA deposition rates are too small and even with zooming in the pdf cannot be deciphered. Similarly for all other labels.
L375: "Alpine" instead of "Alphine".
Â
Additional references
Franco, B., Blumenstock, T., Cho, C., Clarisse, L., Clerbaux, C., Coheur, P. F., De Mazière, M., De Smedt, I., Dorn, H. P., Emmerichs, T., Fuchs, H., Gkatzelis, G., Griffith, D. W. T., Gromov, S., Hannigan, J. W., Hase, F., Hohaus, T., Jones, N., Kerkweg, A., Kiendler-Scharr, A., Lutsch, E., Mahieu, E., Novelli, A., Ortega, I., Paton-Walsh, C., Pommier, M., Pozzer, A., Reimer, D., Rosanka, S., Sander, R., Schneider, M., Strong, K., Tillmann, R., Van Roozendael, M., Vereecken, L., Vigouroux, C., Wahner, A., and Taraborrelli, D.: Ubiquitous atmospheric production of organic acids mediated by cloud droplets, Nature, 593, 233-237, 10.1038/s41586-021-03462-x, 2021.
Freeling, F., Behringer, D., Heydel, F., Scheurer, M., Ternes, T. A., and Nödler, K.: Trifluoroacetate in Precipitation: Deriving a Benchmark Data Set, Environ. Sci. Technol., 54, 11210-11219, 10.1021/acs.est.0c02910, 2020.
Madronich, S., Sulzberger, B., Longstreth, J. D., Schikowski, T., Andersen, M. P. S., Solomon, K. R., and Wilson, S. R.: Changes in tropospheric air quality related to the protection of stratospheric ozone in a changing climate, Photochemical & Photobiological Sciences, 22, 1129-1176, 10.1007/s43630-023-00369-6, 2023.
Wang, Z., Wang, Y., Li, J., Henne, S., Zhang, B., Hu, J., and Zhang, J.: Impacts of the Degradation of 2,3,3,3-Tetrafluoropropene into Trifluoroacetic Acid from Its Application in Automobile Air Conditioners in China, the United States, and Europe, Environ. Sci. Technol., 52, 2819-2826, 10.1021/acs.est.7b05960, 2018.
Â
Citation: https://doi.org/10.5194/egusphere-2025-1031-RC2
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1031', Anonymous Referee #1, 14 May 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1031/egusphere-2025-1031-RC1-supplement.pdf
-
AC1: 'Reply on RC1', Shivendra Tewari, 23 Jun 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1031/egusphere-2025-1031-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Shivendra Tewari, 23 Jun 2025
-
RC2: 'Comment on egusphere-2025-1031', Anonymous Referee #2, 23 Jun 2025
The manuscript by Shivendra G. Tewari presents a study of the atmospheric degradation of HFO-1234ze(E) and hydrological fate of one of its degradation products, the environmentally persistent trifluoroacetic acid (TFA). The study is specific for the emissions of HFO-1234ze(E), used as a propellant in metered-dose inhalers, and does not discuss emissions of other use cases. Although, the applied atmospheric and hydrological modelling appear appropriate in the first place, there is very little information on validation of the model results in terms of discussion of previous simulations and observations. In addition, the focus on three selected watersheds for the hydrological part of the paper seems very arbitrary and not suited to provide global worst-case scenarios. Hence, the main conclusion of the study, that the use of HFO-1234ze(E) for inhalers will not pose a risk for TFA in drinking water, may not be valid everywhere on the globe and should be revisited. Although, the manuscript is well-structured and written and only a few additional clarifications are required in this respect, addressing the main concerns may require major revisions of the manuscript before publication.
Â
Major comments
Generalisation of results: Although, it is explicitly said that results are derived for three selected watersheds, the final sentence of the abstract and the final paragraph of the conclusions seems to imply that the use of HFO-1234ze(E) in inhalers does to pose a threat globally. The selection of these watersheds is very arbitrary (L227/228). "prominent" and "large population" seem to be the only criteria. However, for all three continents one can very easily come up with watersheds that have a larger population. Figure 4 shows that there are also word regions where we expect larger TFA deposition. However, in the end it is not the deposition flux but the rainwater concentration that will determine surface water concentrations. Hence, I would have expected that a selection of watersheds would focus on those for which largest TFA concentrations in precipitation are predicted. Previous studies have shown that TFA concentrations will be enhanced in more arid regions and, hence, a look at river systems in such areas would be more helpful than considering the 3 current watersheds, from which we cannot conclude that TFA levels will stay within the safety margins globally.
Â
Isolated view on the TFA budget: There is one more danger in the presentation of isolated results of TFA from a single precursor and use case. Obviously, it is the sum of contributions from all precursors that determines environmental TFA levels. If ten studies like the current for specific use cases all come to the conclusion that individually there is no problem, the sum may still present an environmental problem. It is mentioned that compared to current levels of TFA in precipitation in Germany, TFA from degradation of HFO-1234ze(E) in pMDIs would add less than 1 %. Again, this may not be true globally. I would suggest to put the current use case more into the perspective of the global TFA budget (as much as this is known, for example see Madronich et al. 2023).
Â
Emission scenario: It is assumed that future emissions from pMDIs will follow the same global usage as taken from current pMDI sales (L271). Figure 3 reveals that some world regions are underrepresented with this assumption. Especially emissions in densely populated China seem to be unrealistically low. Is this because sales data from China is potentially incomplete? Even if it is complete, would one not expect that access to pMDIs will increase in China in the future? In general, it would be better to work with projected consumption numbers then with present day values. Since Southeast Asia also seems to be an area of intense TFA deposition (Figure 4b), realistic Chinese emissions seem to be critical for a fair assessment of future TFA levels in this region. Furthermore, instead of using NO as a proxy for the spatial distribution of HFO emissions, it seems more appropriate to use population density directly. NO distributions may be skewed by individual point sources like power plants.
Â
Minor comment
Section 2.1-2.3: There are several questions concerning the setup of GEOS-Chem that need clarification. What is the name/version of the utilised chemistry scheme? How many compounds are treated? How are the NMVOC emissions mapped onto model species? Are there any validation results of GEOS-Chem for the classical air pollutants (O3, NOx, …)? Done as part of this study or published elsewhere for the same setup of the model. These would be helpful to understand if simulated OH levels and, hence, HFO reaction rates are realistic.
L153f: How valid is the use of absorption cross sections of CH3C(O)CHO for CF3C(O)C(O)F? Can this be corroborated from similarities from any know cross sections for other similar molecules?
Figure1 and section 2.3: The figure seems to indicate an alternative path for trifluoroacetic aldehyde (TFAA) to TFA through hydration. This path is not discussed in the text. Is this based on the diol mechanism suggested by Franco et al. (2021) for other aldehydes? Please comment if it was included in the chemistry scheme and if it showed any relevance.
Section 2.4: Does the model take evapotranspiration in the watershed into account? Should to be considered in warmer climates.
Figure 4 and Table 3: Earlier model studies on HFO degradation and TFA deposition all came to the conclusion that wet deposition dominates over dry deposition. This was the case for HFO-1234yf, where the formation of TFA should be fast (e.g., Luecken et al., 2010; Henne et al., 2012, Wang et al., 2018), but also for HFO-1233zd(E) (Sulbaek Andersen et al., 2018) for which TFA formation also proceeds through TFAA. Please comment, why and how this could be different in the present case. This also questions the statement made on line 298 concerning 'akin' deposition fluxes as in (Sulbaek Andersen et al., 2018), which then seems oversimplified.
Figure 4: In order to assess the global impact of TFA deposition, it would be beneficial to add another figure that shows average rainwater concentrations of TFA. Since these are usually strongly enhanced during the summer months, I would suggest to show these with three panels: overall average, summer average, winter average. Derived concentrations could be used in the discussion against currently observed TFA in precipitation (see below).
Section 3.1: What was the total global TFA deposition flux? How does it compare to the HFO emissions and what can be concluded in terms of TFA yields for this compound?
Figure 5 and L325f: To me it remains unclear how the conclusions can be reached from what is presented in the figure. Is it only the spatial correlation between the two quantities? But then TFA concentrations will strongly depend on the removal not just the production pathway. Furthermore, how does the correlation look for other intermediates? This requires additional explanation.
Figure 6: The plot reveals another potential shortcoming in the assessment of maximal TFA concentrations. The global chemistry simulations were performed at relatively coarse resolution. However, precipitation and hence TFA deposition often varies at much smaller scales as can be covered by the global chemistry model. As a consequence actual TFA inputs into individual watersheds may largely differ from the grid cell average of GEOS-Chem. Please add a note of caution and discuss the possible implications.
L340: There seems to be another important simplification for the hydrological modelling which needs to be addressed. It is well know that TFA inputs from the atmosphere have a strong seasonal cycle in midlatitudes (both observed and simulated rainwater concentrations show this). In order to assess maximum concentrations in the watersheds it therefore seems very important to consider the seasonality in the inputs and see how this variability propagates through different strata.
L353f and L362: Both statements seem to suggest that a large fraction of TFA will accumulate in the soil. To me it is not clear on which time scales you are discussing this. In steady state (which apparently is reached quickly), input from the atmosphere should be equal to outflow to the ocean. Or is what you call deep soil a open boundary for the model as well? Furthermore, this strong flux to soil seems to contradict the statement in the footnote of Table 1: " Adsortion/desorption tests results show that TFA is poorly adsorbed to the soil and is considered as a mobile organic compound in the majority of soils investigated." Please clarify.
L378: For the Rhine catchment a more direct comparison between simulated atmospheric inputs and measured TFA in precipitation could be done (see Freeling et al., 2020).
L407f: Why not discuss with the often quoted NOEC for the most sensitive freshwater algae, which is 120'000 ng/L? In addition, as TFA cannot be removed from drinking water at large scale, the discussion of an additional threshold much higher than the one suggested for drinking water seems a bit artificial.
Â
Technical comments
Citations in text: Luecken et al.(Luecken et al., 2010) should be Luecken et al. (2010). Applies to all references that should not include the author.
Andersen et al., 2018 and 2022: Should be Sulbaek Andersen et al., 2018. Sulbaek being part of the surname not the given name.
Figure 6: The labels for the color scale of TFA deposition rates are too small and even with zooming in the pdf cannot be deciphered. Similarly for all other labels.
L375: "Alpine" instead of "Alphine".
Â
Additional references
Franco, B., Blumenstock, T., Cho, C., Clarisse, L., Clerbaux, C., Coheur, P. F., De Mazière, M., De Smedt, I., Dorn, H. P., Emmerichs, T., Fuchs, H., Gkatzelis, G., Griffith, D. W. T., Gromov, S., Hannigan, J. W., Hase, F., Hohaus, T., Jones, N., Kerkweg, A., Kiendler-Scharr, A., Lutsch, E., Mahieu, E., Novelli, A., Ortega, I., Paton-Walsh, C., Pommier, M., Pozzer, A., Reimer, D., Rosanka, S., Sander, R., Schneider, M., Strong, K., Tillmann, R., Van Roozendael, M., Vereecken, L., Vigouroux, C., Wahner, A., and Taraborrelli, D.: Ubiquitous atmospheric production of organic acids mediated by cloud droplets, Nature, 593, 233-237, 10.1038/s41586-021-03462-x, 2021.
Freeling, F., Behringer, D., Heydel, F., Scheurer, M., Ternes, T. A., and Nödler, K.: Trifluoroacetate in Precipitation: Deriving a Benchmark Data Set, Environ. Sci. Technol., 54, 11210-11219, 10.1021/acs.est.0c02910, 2020.
Madronich, S., Sulzberger, B., Longstreth, J. D., Schikowski, T., Andersen, M. P. S., Solomon, K. R., and Wilson, S. R.: Changes in tropospheric air quality related to the protection of stratospheric ozone in a changing climate, Photochemical & Photobiological Sciences, 22, 1129-1176, 10.1007/s43630-023-00369-6, 2023.
Wang, Z., Wang, Y., Li, J., Henne, S., Zhang, B., Hu, J., and Zhang, J.: Impacts of the Degradation of 2,3,3,3-Tetrafluoropropene into Trifluoroacetic Acid from Its Application in Automobile Air Conditioners in China, the United States, and Europe, Environ. Sci. Technol., 52, 2819-2826, 10.1021/acs.est.7b05960, 2018.
Â
Citation: https://doi.org/10.5194/egusphere-2025-1031-RC2
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,033 | 218 | 20 | 1,271 | 45 | 17 | 39 |
- HTML: 1,033
- PDF: 218
- XML: 20
- Total: 1,271
- Supplement: 45
- BibTeX: 17
- EndNote: 39
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Shivendra G. Tewari
Krish Vijayaraghavan
Liji M. David
Katie Tuite
Felix Kristanovich
Yuan Zhuang
Benjamin Yang
Cecilia Hurtado
Dimitrios K. Papanastasiou
Paul Giffen
Holly Kimko
Megan Gibbs
Stefan Platz
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1485 KB) - Metadata XML
-
Supplement
(3128 KB) - BibTeX
- EndNote
- Final revised paper