the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Bromine and Iodine in Atmospheric Mercury Oxidation
Abstract. We investigate atmospheric oxidation of mercury Hg(0) by halogens, initiated by Br and I to yield Hg(I), and continued by I, Br, BrO, ClO, IO, NO2 and HO2 to yield Hg(II) or Hg(0), using computational methods with focus on creation of data for determining the impact of rising iodine levels. We calculate reaction enthalpies and Gibbs free energies using the Coupled Cluster singlets, doublets, and perturbative triplets method (CCSD(T)) with the ma-def2-TZVP basis set and effective core potential to account for relativistic effects. Additionally, we investigate reaction kinetics using variational transition state theory based on geometric scans of bond dissociations at the CASPT2/ma-def2-TZVP level. We compare the results obtained from both methods to help define the uncertainty. Our results provide insights into the mechanisms of these reactions, and the data produced get us closer to determining iodine’s impact on mercury depletion events and on the atmosphere as a whole. The reaction ·HgBr+Br·->HgBr2 was found to be twice as fast as HgI·+I·->HgI2 with reaction rate coefficients of 8.8x10−13 and 4.2x10−13 cm3 molecule−1 s−1 respectively. The BrHg·+BrO·->BrHgOBr reaction was about 7.2 times faster than the ·HgI+IO·->IHgOI reaction with their rates being 3.3x10−14 and 4.6x10−15 cm3 molecule−1 s−1 respectively. We investigate the Hg•XOY (X and Y being halogen) complexes. From the reactions investigated including iodine, the reaction with the most plausible chance of impacting the mercury lifetime in the atmosphere is HgI·+I·->HgI2.
Status: final response (author comments only)
-
CC1: 'Comment on egusphere-2025-116', Henrik Skov, 31 Jan 2025
There is a wrong citation:
ref. 27. Goodsite, M. E., Skov, H., Christensen, J. H., Heidam, N. Z., Jensen, B., and W˚ahlin, P.: Fate of elemental mercury in the Arctic during atmospheric mercury depletion episodes and the load of atmospheric mercury to the Arctic. Antarctic springtime depletion of atmospheric mercury, Environmental Science & Technology, 38, 2373–2382, 2004.
Should be:
Skov, H. Christensen, J. Goodsite, M.E. Heidam, N.Z. Jensen, B. Wåhlin, P. and Geernaert, G. (2004) “The fate of elemental mercury in Arctic during atmospheric mercury depletion episodes and the load of atmospheric mercury to Arctic” ES & T. vol. 38, 2373-2382.
Citation: https://doi.org/10.5194/egusphere-2025-116-CC1 -
CC2: 'Reply on CC1', Svend L. Bager, 01 Feb 2025
Thanks for letting us know. We will make sure the correction is included.
Citation: https://doi.org/10.5194/egusphere-2025-116-CC2
-
CC2: 'Reply on CC1', Svend L. Bager, 01 Feb 2025
-
RC1: 'Comment on egusphere-2025-116', Anonymous Referee #1, 12 Feb 2025
Bager et al. investigate the reactions between Hg(I) (primarily HgBr, but also HgI) and various atmospheric species (I, Br, BrO, ClO, IO, NO₂, and HO₂) leading to the formation of Hg(II) or Hg(0) using computational methods. They propose that these reactions could play a dominant role in mercury oxidation during Arctic Mercury Depletion Events (AMDEs) in spring.
However, both experimental and theoretical studies (Shat et al., 2021; Gómez Martín et al., 2023) have firmly established that the reaction between HgBr and O₃ is the primary pathway to Hg(II), with a measured rate constant of k(HgBr + O₃) = 7.5 × 10⁻¹¹ cm³ molecule⁻¹ s⁻¹.
During extreme AMDEs, ozone mixing ratios may drop below 1 ppb. Even at such low levels, given that the reaction between HgBr and O₃ is orders of magnitude faster than the reactions considered in this study, these alternative pathways would have negligible impact.
For example, assuming 10 ppt of Br and 1 ppb of O₃:
Species
Mixing Ratio (ppt)
Concentration (molecule cm⁻³)
k (cm³ molecule⁻¹ s⁻¹)
k' (s⁻¹)
Br
10
2.50 × 10⁸
8.00 × 10⁻¹³
2.00 × 10⁻⁴
O₃
1000
2.50 × 10¹⁰
7.50 × 10⁻¹¹
1.88
This indicates that the removal rate of HgBr by O₃ is still four orders of magnitude faster than its removal by Br.
To reach a point where these rates are comparable:
Species
Mixing Ratio (ppt)
Concentration (molecule cm⁻³)
k (cm³ molecule⁻¹ s⁻¹)
k' (s⁻¹)
Br
10
2.50 × 10⁸
8.00 × 10⁻¹³
2.00 × 10⁻⁴
O₃
0.1
2.50 × 10⁶
7.50 × 10⁻¹¹
1.88 × 10⁻⁴
Thus, ozone mixing ratios would need to drop to sub-ppt levels—an extremely unlikely scenario, even in the most extreme AMDEs.
Rather than questioning the computational methods used in this study, my concern is the atmospheric relevance of its conclusions. The authors appear to have overlooked key aspects of mercury atmospheric chemistry and have overstated the implications of their results.
Since ACP is a journal focused on atmospheric chemistry and physics, this paper does not meet the necessary standards for publication in this venu
Citation: https://doi.org/10.5194/egusphere-2025-116-RC1 -
AC1: 'Reply on RC1', Stephan P. A. Sauer, 25 Apr 2025
Author Response:
Certainly, this is correct as regards bromine chemistry. We feel that the same cannot be said for the iodine chemistry, where the concentrations could be higher, and the reaction rates are not as well characterized.
Changes:
We have added the following passage:
Mercury reactions as a whole are rather poorly characterized due to the experimental difficulties and the potential health risks. This is also true to some extent for bromine and iodine reactions. Quantum chemistry is therefore an especially valuable tool for characterizing reaction paths. Special consideration must be taken when working with heavy nuclei. A particular emphasis of this paper, in addition to validating the computational methods that we use, is to better characterize the iodine reactions. While it is known that ozone oxidation of HgBr, and not Br, is very likely to be the main pathway for Hg(II) formation under virtually all atmospheric conditions, the same cannot be said with certainty about the iodine reactions. This is both because the reaction rates themselves are not well characterized, and because the atmospheric chemistry of iodine is different from that of bromine.
Author Response:
This goal of the study is to compare sets of reactions that include iodine and bromine, to determine whether iodine reactions could have an impact on ozone deposition. This is a relevant line of inquiry also due to the rising levels of iodine in the atmosphere driven by climate change.
The VTST method was chosen because it fits well with the reactions considered in the article, being able to describe the rate of the six reactions in the same manner and thereby making a solid basis for comparison. We have obtained rate constants that are slower than for example the one described by Goodsite 2004 (2.5E-10 cm3/(molecule s)) and Balabanov 2005 (3.0E-11 cm3/(molecule s)). The reason is explained in the article.
We can of course see and understand the relevance of the HgBr + O3 reaction. It would be useful to characterize the reaction using the same VTST method in future work.
Changes:
We have added the following passage:
For future work, investigating the reaction rates of ·XHg + O3 → XHgO + O2 , ·XHg + ·NO2 → XHgONO and ·XHg + ·OOH → XHgOOH is likely to be beneficial.Citation: https://doi.org/10.5194/egusphere-2025-116-AC1
-
AC1: 'Reply on RC1', Stephan P. A. Sauer, 25 Apr 2025
-
RC2: 'Comment on egusphere-2025-116', Anonymous Referee #2, 15 Feb 2025
This paper describes high level electronic structure calculations to explore the oxidation pathways for elemental Hg(0) by reactions with Br and I atoms. The objective is to assess the importance of I relative to Br in the atmospheric oxidation of Hg to Hg(II). The ab initio calculations are carried out at a high level of theory, and rate coefficients are then estimated using variational transition rate theory. Overall, the study makes some potentially useful contributions to understanding Hg oxidation.
However, there is at least one major problem to address. Table VIII compares results from the present study with previous work. Although the temperature and pressure is not stated (!), one assumes it is 298 K and 1 bar. The reaction HgBr + Br --> HgBr2 is a recombination reaction. According to the authors, this has a second-order rate coefficient of 8.8e-13 cm3 molecule-1 s-1, and hence a third-order rate coefficient of ~ 4e-32 cm6 molecule-2 s-1. This seems extremely slow for a reaction that is exothermic by 71 kcal mol-1 (Table I) and involves exclusively very heavy atoms, so must have a high density of rovibrational states in the HgBr2 adduct. Strangely there is no discussion as to why this reaction is so unexpectedly slow (by around 2 orders of magnitude). Inspection of Table VIII shows an earlier estimate from Goodsite et al. that is close to the HgBr + Br capture rate, which is closer to what would be expected.
Can the authors demonstrate that their TST approach yields rate coefficients for any of these Hg reactions that agree well with experiment? There do not appear to be any examples in Table VIII.
Other points include:
The references are in the format of a physical chemistry journal, not ACP!
All calculated energies are quoted to 0.01 kcal mol-1 which implies a level of accuracy that is certainly not warranted.
Page 3 discusses observations of IO in the Arctic, but fails to cite the first observations made (at Hudson Bay): Mahajan, A. S., M. Shaw, H. Oetjen, K. E. Hornsby, L. J. Carpenter, L. Kaleschke, X. Tian-Kunze, J. D. Lee, S. J. Moller, P. Edwards, R. Commane, T. Ingham, D. E. Heard, and J. M. C. Plane (2010), Evidence of reactive iodine chemistry in the Arctic boundary layer, J. Geophys. Res. 115, art. no.: D20303.
Page 4: the original mechanism for Br oxidation was formulated by Goodsite et al (not Goodsite).
Page 5: in the computational details, there is no mention of how spin-orbit coupling is included (particularly important for I)
Page 6: no mention of the temperature at which G is calculated (presumably 298 K)
Citation: https://doi.org/10.5194/egusphere-2025-116-RC2 -
AC2: 'Reply on RC2', Stephan P. A. Sauer, 25 Apr 2025
1) Reviewer Comment:
However, there is at least one major problem to address. Table VIII compares results from the present study with previous work. Although the temperature and pressure is not stated (!), one assumes it is 298 K and 1 bar.Author Response:
Thank you, yes this is correct and we have now noted it in the paper.2) Reviewer Comment:
The reaction HgBr + Br --> HgBr2 is a recombination reaction. According to the authors, this has a second-order rate coefficient of 8.8e-13 cm3 molecule-1 s-1, and hence a third-order rate coefficient of ~ 4e-32 cm6 molecule-2 s-1. This seems extremely slow for a reaction that is exothermic by 71 kcal mol-1 (Table I) and involves exclusively very heavy atoms, so must have a high density of rovibrational states in the HgBr2 adduct. Strangely there is no discussion as to why this reaction is so unexpectedly slow (by around 2 orders of magnitude). Inspection of Table VIII shows an earlier estimate from Goodsite et al. that is close to the HgBr + Br capture rate, which is closer to what would be expected.
Author Response:
It is true that the reaction constant is a high pressure reaction constant, but due the lack of a low pressure constant we have not estimated the pressure dependent constant of the reaction. We note that both Goodsite 2004 and Balabanov 2005 use units of cm3/(molecule*s)) for the rate constants.3) Reviewer Comment:
Can the authors demonstrate that their TST approach yields rate coefficients for any of these Hg reactions that agree well with experiment? There do not appear to be any examples in Table VIII.Author Response:
There is one example of an experimental value in Table 8 of Greig et al. (2002) who report 7x10-17 for the HgBr+Br → HgBr2 reaction. We have not been able to find an experimental determination for the rate of the IHg + I → HgI2 reaction.4) Reviewer Comment:
Other points include: The references are in the format of a physical chemistry journal, not ACP!
Author Response:
Corrected in the LaTeX template5) Reviewer Comment:
All calculated energies are quoted to 0.01 kcal mol-1 which implies a level of accuracy that is certainly not warranted.Author Response:
Thank you, we have revised all of these numbers to 0.1 kcal/mol.6) Reviewer Comment:
Page 3 discusses observations of IO in the Arctic, but fails to cite the first observations made (at Hudson Bay): Mahajan, A. S., M. Shaw, H. Oetjen, K. E. Hornsby, L. J. Carpenter, L. Kaleschke, X. Tian-Kunze, J. D. Lee, S. J. Moller, P. Edwards, R. Commane, T. Ingham, D. E. Heard, and J. M. C. Plane (2010), Evidence of reactive iodine chemistry in the Arctic boundary layer, J. Geophys. Res. 115, art. no.: D20303.
Author Response:
Thanks for letting us know about his reference, it has now been included.7) Reviewer Comment:
Page 4: the original mechanism for Br oxidation was formulated by Goodsite et al (not Goodsite).
Author Response:
Thanks you we have corrected the citation.8) Reviewer Comment:
Page 5: in the computational details, there is no mention of how spin-orbit coupling is included (particularly important for I)Author Response:
We have added the following description to the Computational Methods section:
In the application of ECPs, the core electrons are replaced by effective potentials, corresponding to 60 electrons for Hg and 28 electrons for I.9) Reviewer Comment:
Page 6: no mention of the temperature at which G is calculated (presumably 298 K)
Author Response:
The temperature at which G is calculated is now noted in tables 1, 2, 3 and 4.Citation: https://doi.org/10.5194/egusphere-2025-116-AC2
-
AC2: 'Reply on RC2', Stephan P. A. Sauer, 25 Apr 2025
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 137 | 0 | 0 | 137 | 0 | 0 |
- HTML: 137
- PDF: 0
- XML: 0
- Total: 137
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1