1

Improved mechanistic model of the atmospheric 1

redox chemistry of mercury 2 3 Viral Shah1*, Daniel J. Jacob1,2, Colin P. Thackray1, Xuan Wang3, Elsie M. Sunderland1,4, 4

Theodore S. Dibble5, Alfonso Saiz-Lopez6, Ivan Černušák7, Vladimir Kellö7, Pedro J. 5

Castro5, Rongrong Wu8, Chuji Wang8 6

7 1 Harvard John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA, USA 8 2 Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA, USA 9 3 School of Energy and Environment, City University of Hong Kong, Hong Kong SAR, China 10 4 Department of Environmental Health, Harvard T.H. Chan School of Public Health, Harvard University, Boston, MA, 11

USA 12 5 Department of Chemistry, State University of New York, College of Environmental Science and Forestry, Syracuse, 13

NY, USA 14 6 Department of Atmospheric Chemistry and Climate, Institute of Physical Chemistry Rocasolano, CSIC, Madrid, 15

Spain 16 7 Department of Physical and Theoretical Chemistry, Faculty of Natural Sciences, Comenius University in Bratislava, 17

Ilkovičova 6, 84215 Bratislava, Slovakia 18 8 Department of Physics and Astronomy, Mississippi State University, Starkville, MS, USA 19

20

Abstract 21

We present a new chemical mechanism for Hg0/HgI/HgII atmospheric cycling, including recent 22

laboratory and computational data, and implement it in the GEOS-Chem global atmospheric 23

chemistry model for comparison to observations. Our mechanism includes the oxidation of Hg0 24

by Br atoms and OH radicals, with subsequent oxidation of HgI by ozone and radicals, re-25

speciation of gaseous HgII in aerosols and cloud droplets, and speciated HgII photolysis in the gas 26

2

and aqueous phases. The tropospheric Hg lifetime against deposition in the model is 5.5 months, 27

consistent with observational constraints. The model reproduces the observed global surface Hg0 28

concentrations and HgII wet deposition fluxes. Hg0 is oxidized almost equally through Br and 29

OH. Ozone is the principal HgI oxidant, enabling the efficient oxidation of Hg0 to HgII by OH. 30

BrHgIIOH and HgII(OH)2 are the initial products, which re-speciate in aerosols/cloud and 31

volatilize to photostable forms. Reduction of HgII to Hg0 takes place largely through photolysis 32

of aqueous HgII-organic complexes. 71% of model HgII deposition is to the oceans. Major 33

uncertainties for atmospheric Hg chemistry modeling include the concentrations of Br atoms, the 34

stability and reactions of HgI, and the speciation of HgII in aerosols/cloud with implications for 35

photoreduction. 36

37

Keywords: mercury modeling, chemical mechanism, mercury oxidation, mercury 38

photoreduction, atmospheric lifetime, mercury deposition 39

40

Synopsis: Mercury spreads globally through the atmosphere and its atmospheric chemistry 41

determines where it deposits to the surface ecosystems. We present an improved and up-to-date 42

model of atmospheric mercury chemistry. 43

3

Introduction 44

Mercury (Hg) is an ecosystem pollutant transported globally through the atmosphere. It is 45

emitted in gaseous elemental state (Hg0) by natural and anthropogenic sources, and cycles in the 46

atmosphere with divalent (HgII) compounds that are highly water-soluble and deposit to 47

ecosystems. Recent theoretical calculations show fast gas-phase reduction of the major HgII 48

species thought to be produced in the atmosphere (Saiz-Lopez et al., 2018; Lam et al., 2019b; 49

Francés‐Monerris et al., 2020; Khiri et al., 2020), posing a challenge for atmospheric models to 50

reproduce the atmospheric Hg mass and lifetime inferred from observations (Saiz-Lopez et al., 51

2020). At the same time, new oxidation pathways to form HgII in the atmosphere have been 52

proposed (Dibble et al., 2020; Saiz-Lopez et al., 2020). Here we integrate these recent 53

developments into a new chemical mechanism for the GEOS-Chem global model (Selin et al., 54

2007; Horowitz et al., 2017) to shed new light on the redox cycling of atmospheric Hg. 55

56

Hg0 is emitted to the atmosphere by mining, fuel combustion, and volcanism, and by 57

volatilization of previously deposited Hg (Pirrone et al., 2010; Streets et al., 2019a). The Hg0 58

oxidation pathways and the speciation of HgII remain highly uncertain (Jaffe et al., 2014; Gustin 59

et al., 2015, 2021). The Br atom is considered to be a major Hg0 oxidant (Lindberg et al., 2002; 60

Holmes et al., 2006; Obrist et al., 2011; Gratz et al., 2015). The oxidation of Hg0 to HgII by Br 61

takes place in two steps, beginning with the formation of a BrHgI intermediate, which reacts 62

further to form HgII (Tossell, 2003; Goodsite et al., 2004, 2012). BrHgI has been thought to react 63

primarily with NO2 and HO2 (Dibble et al., 2012), but preliminary theoretical calculations by 64

Saiz-Lopez et al. (2020) show that BrHgI could also react rapidly with ozone and produce a 65

4

BrHgIIO radical, which can be stabilized to non-radical HgII forms by subsequent reactions (Lam 66

et al., 2019a, b; Khiri et al., 2020; Francés‐Monerris et al., 2020). 67

68

The oxidation of Hg0 by OH has been included in many models (Selin et al., 2007; Travnikov 69

and Ilyin, 2009; De Simone et al., 2014; Dastoor et al., 2015), but its atmospheric relevance has 70

been questioned because of the low stability of HOHgI (Goodsite et al., 2004; Calvert and 71

Lindberg, 2005). Dibble et al. (2020) recalculated the stability of HOHgI and found the OH-72

initiated oxidation pathway to be important only in polluted conditions with high NO2 73

concentrations. Oxidation reactions of Hg0 with ozone (Pal and Ariya, 2004b; Sumner et al., 74

2005) and by BrO (Raofie and Ariya, 2004) have been observed in the laboratory, but are not 75

expected to be atmospherically relevant because the putative product (HgIIO) is weakly bound in 76

the gas phase (Shepler and Peterson, 2003; Filatov and Cremer, 2004; Peterson et al., 2007). 77

78

Partitioning of gas-phase HgII species into aerosols and cloud droplets adds further complexity to 79

the problem. Atmospheric observations indicate that this partitioning is governed by 80

thermodynamic equilibrium (Amos et al., 2012). Once in the condensed phase, HgII may re-81

speciate as different inorganic and organic complexes that then partition back to the gas phase 82

(Lin and Pehkonen, 1999). Aqueous HgII-organic complexes photoreduce to Hg0 in the 83

atmosphere (Yang et al., 2019). 84

85

Although uncertainties in the Hg0/HgI/HgII atmospheric redox cycling remain large, we show 86

here that the most recent laboratory and computational data can be accommodated in a GEOS-87

Chem chemical mechanism that reproduces the main features of atmospheric observations and 88

5

thus provides a basis for predicting Hg deposition to ecosystems. Our work represents a major 89

revision to the previous GEOS-Chem mechanism described by Horowitz et al. (2017). 90

91

Materials and methods 92

Chemical mechanism 93

Table 1 lists the reactions in our chemical mechanism and Figure 1 shows the main reaction 94

pathways. Hg0 oxidation is initiated in the gas phase by the radicals Y ≡ Br, Cl, and OH, forming 95

weakly bound intermediates, YHgI, that further add another radical, Z, to form YHgIIZ: 96

97

Hg! + 𝑌 +M ↔ 𝑌Hg" +M R1 98

𝑌Hg" + 𝑍 + M → 𝑌Hg""𝑍 +M R2 99

100

The reaction of Hg0 with Br is exothermic and barrierless (Tossell, 2003; Goodsite et al., 2004; 101

Balabanov et al., 2005) and its kinetics have been experimentally measured (Ariya et al., 2002; 102

Donohoue et al., 2006). BrHgI has a low bond energy and dissociates thermally within minutes in 103

much of the atmosphere (Dibble et al., 2012; Goodsite et al., 2012), but its association reactions 104

with Z ≡ OH, Br, NO2, HO2, BrO, ClO are also barrierless and fast (Goodsite et al., 2004; Dibble 105

et al., 2012; Jiao and Dibble, 2017a). BrHgIIONO and BrHgIIOOH are thought to be the major 106

products due to the abundance of NO2 and HO2 (Dibble et al., 2012; Jiao and Dibble, 2017a, b). 107

BrHgI does not abstract hydrogen atoms and is inefficient in adding to C=C double bonds 108

(Dibble and Schwid, 2016). It undergoes displacement reactions with certain radicals (Z1≡ NO2 109

and Br) to return Hg0 (Balabanov et al., 2005; Jiao and Dibble, 2017a): 110

111

6

𝑌Hg" + 𝑍# → Hg! + 𝑌𝑍# R3 112

113

This chemistry has been included previously in the GEOS-Chem mechanism (Shah et al., 2016; 114

Horowitz et al., 2017) and other models (Wang et al., 2014; Travnikov et al., 2017; Ye et al., 115

2018). Here we update the rate coefficient for Reaction (R2) based on recent laboratory 116

measurement of the BrHgI + NO2 reaction (Wu et al., 2020). 117

118

The OH-initiated oxidation of Hg0 to HgII also proceeds by the (R1)-(R2) two-step mechanism, 119

and HOHgI is analogous to BrHgI in forming thermally stable HOHgIIZ (Z≡NO2, HO2, etc.) 120

species (Goodsite et al., 2004; Calvert and Lindberg, 2005; Dibble et al., 2020). The Hg0 + OH + 121

M ® HOHgI + M reaction is exothermic and fast (Sommar et al., 2001; Pal and Ariya, 2004a; 122

Hynes et al., 2009), but theoretical calculations by Goodsite et al. (2004) found HOHgI to be so 123

weakly bound that it would thermally decompose rather than further react to form HgII. As a 124

result, this pathway was not considered in past GEOS-Chem mechanisms (Holmes et al., 2010; 125

Horowitz et al., 2017). However, Dibble et al. (2020) found a much higher bond energy for 126

HOHgI and so we reconsider this pathway here. 127

128

Oxidation of Hg0 by Cl atoms is fast (Ariya et al., 2002; Donohoue et al., 2005) and ClHgI is 129

thermally stable, but tropospheric Cl concentrations are low. We include it in our mechanism 130

using GEOS-Chem Cl concentrations from Wang et al. (2021) but find that it accounts for less 131

than 1% of global tropospheric Hg0 conversion to HgII. Horowitz et al. (2017) included the 132

aqueous-phase oxidation of Hg0 by HOCl, OH, and ozone in cloud droplets but found them to be 133

negligible due to the low solubility of Hg0 and we do not include them in our mechanism. 134

7

135

Standard chemical mechanisms for atmospheric Hg, including Horowitz et al. (2017), do not 136

include gas-phase photoreduction of HgII. However, BrHgIIZ (Z≡ NO2, HO2, OH, BrO, ClO) 137

species have been found to rapidly photolyze at tropospheric wavelengths (Saiz-Lopez et al., 138

2018; Francés‐Monerris et al., 2020). The major HgII species, BrHgIIONO and BrHgIIOOH, 139

photolyze on a time scale of minutes (Saiz-Lopez et al., 2018; Lam et al., 2019b). YHgI (Y≡ Br, 140

Cl, OH) species also photo-dissociate rapidly to Hg0. 141

142

Saiz-Lopez et al. (2020) found that including HgI and HgII photolysis in their global model 143

greatly lowered the conversion rate of Hg0 to HgII and led to large overestimate of atmospheric 144

Hg0 concentrations. Their results implied a missing Hg oxidation pathway in current 145

mechanisms, and they suggested the oxidation of BrHgI by ozone: 146

147

BrHg" + O$ → BrHg""O + O% R4 148

149

Reaction (R4) is strongly exothermic (Shepler et al., 2007). Saiz-Lopez et al. (2020) found that it 150

is likely barrierless and produces the BrHgIIO radical. Using methods similar to theirs (density 151

functional theory), we also observe no barrier (Figure S1). However, reactions of ozone can pose 152

significant challenges for theory, so we used more advanced methods, including CASPT2 153

calculations (Figure S1), to confirm the absence of a barrier to reaction. Preliminary 154

experimental data indicate a high rate coefficient and are consistent with the absence of barrier 155

(Figure S2). We find that the analogous reaction of HOHgI with ozone also lacks a barrier and 156

has similar exothermicity to reaction (R4) (Figure S3), reflecting the similarity between BrHgI 157

8

and HOHgI (Dibble et al., 2020). Saiz-Lopez et al. (2020) estimated an upper limit for the rate 158

coefficient of 1×10-10 cm3 molec-1 s-1 for reaction (R4), assuming no steric effects. Here we 159

estimate a rate coefficient of 3×10-11 cm3 molec-1 s-1 for the reaction of YHgI with ozone (Y≡ Br, 160

OH, and Cl). 161

162

The BrHgIIO radical is also formed from the photolysis of certain BrHgIIZ species (Table 1b). Its 163

reactivity mimics that of OH, and it forms stable HgII species by abstracting H atoms from 164

methane and other volatile organic compounds, or by associating with NO and NO2 (Lam et al., 165

2019a, b). Formation of BrHgIIOH by oxidation of methane dominates globally in the kinetics of 166

Lam et al. (2019a, b). Khiri et al. (2020) found that BrHgIIO can also be reduced to BrHgI by CO 167

and the computed reaction rate is fast enough to compete with methane. BrHgIIO photolysis in 168

the troposphere is relatively slow (Francés‐Monerris et al., 2020). Thus, we include the following 169

reactions in our mechanism: 170

171

𝑌Hg" + O$ → 𝑌Hg""O + O% R5 172

𝑌Hg""O + CH& → 𝑌Hg""OH + CH$ R6 173

𝑌Hg""O + CO → 𝑌Hg" + CO% R7 174

175

HgII species are absorbed by aqueous aerosol particles and cloud droplets and dissociate to Hg2+ 176

ions, which re-partition to form inorganic and organic complexes (Lin and Pehkonen, 1999). We 177

refer to total particulate mercury as HgIIP. HgIICl2, HgIICl3-, and HgIICl42- are expected to be the 178

dominant inorganic HgII species in the troposphere because of the abundance of Cl- (Lin and 179

Pehkonen, 1998; Hedgecock and Pirrone, 2001). HgII-organic complexes may also form, 180

9

involving in particular the carboxyl and thiol functional groups (Haitzer et al., 2002; 181

Ravichandran, 2004). In stratospheric sulfuric acid aerosols, HgII likely remains in free ionic 182

form because of the low stability of the HgII-sulfate complex. While the thermodynamics of the 183

HgII-chloride complexes are known (Lin and Pehkonen, 1999), there is little information on the 184

HgII-organic complexes in atmospheric waters. We choose to represent the inorganic and organic 185

complexes by two species – HgIIP(inorg) and HgIIP(org) – and partition the dissolved HgII into 186

these two complexes based on the local relative mass fractions of inorganic and organic aerosol 187

material (Table 1c). Volatilization of HgII from aerosols is as a parameterized species, HgIIX, 188

assumed to be stable against photolysis. HgIIX is modeled after HgIICl2, which does not 189

photolyze at tropospheric wavelengths (Saiz-Lopez et al., 2020). 190

191

Photoreduction of HgII to Hg0 has long been known to occur in atmospheric waters (Munthe and 192

McElroy, 1992). It was initially thought to involve sulfite ions (Munthe et al., 1991) or HO2 193

(Pehkonen and Lin, 1998) as reductants, but it most likely takes place through the direct light 194

absorption by HgII-organic complexes followed by transfer of two electrons from the ligand to 195

Hg (Gårdfeldt and Jonsson, 2003; Si and Ariya, 2008). HgII photoreduction is known to involve 196

HgII-organic complexes in aquatic systems (Amyot et al., 1994; O’Driscoll et al., 2004; Whalin 197

et al., 2007). Aqueous HgII photoreduction frequencies of 0.02–0.2 h-1 have been measured in 198

summertime rainwater samples (Saiz-Lopez et al., 2018), consistent with photoreduction 199

frequencies of HgII-fulvic acid complexes (Yang et al., 2019), but lower than 0.2–3 h-1 observed 200

in fresh and marine waters (Qureshi et al., 2011). In our mechanism, we assume that the 201

photoreduction frequency of HgIIP(org) scales as the local NO2 photolysis frequency (JNO2) and 202

adjust the scaling factor (β in Table 1b) to fit observed atmospheric Hg0 concentrations, 203

10

following Holmes et al. (2010). We obtain a scaling factor β = 4×10-3, corresponding to a 204

tropospheric mean HgIIP(org) photoreduction frequency of 0.13 h-1 in clear sky at noon in 205

summer at 45ºN. 206

207

GEOS-Chem model 208

We implement the chemical mechanism of Table 1 in the global 3-D GEOS-Chem model 209

(www.geos-chem.org; version 12.9.0). The current standard version of the model for Hg is 210

described by Horowitz et al. (2017) and includes dynamic coupling between the atmosphere and 211

surface reservoirs. Here we are focus on the atmospheric reservoir and therefore use gridded land 212

and ocean surface Hg concentrations from Horowitz et al. (2017) as boundary conditions. Other 213

Hg emissions (Figure 1) are also from Horowitz et al. (2017) except that anthropogenic Hg 214

emissions are from Streets et al. (2019b). Total Hg emission in the model is 8.7 Gg a-1, of which 215

0.8 Gg a-1 is HgII and emitted as HgIIX (Figure 1). 216

217

We drive our simulation with assimilated meteorological fields from the NASA Modern-Era 218

Retrospective analysis for Research and Applications, version 2 (MERRA-2) system (Gelaro et 219

al., 2017). We conduct a three-year global simulation (2013–2015) at 4° latitude by 5° longitude 220

resolution following a spin-up period of 15 years to equilibrate the stratosphere. The chemical 221

mechanism is implemented using the Kinetic PreProcessor (KPP; Damian et al., 2002) 222

customized for GEOS-Chem, and the chemical evolution is computed every hour on the model 223

grid. 224

225

11

GEOS-Chem in its ‘full-chemistry’ implementation includes a comprehensive representation of 226

oxidant-aerosol chemistry in the troposphere and stratosphere (Eastham et al., 2014; Wang et al., 227

2019; Hammer et al., 2020; Pai et al., 2020). Aerosol components include sulfate, nitrate, organic 228

aerosol, black carbon, sea salt (two size classes), and dust (four size classes). For computational 229

efficiency, the Hg simulation in GEOS-Chem uses monthly oxidant and aerosol concentrations 230

archived from that full-chemistry simulation. Horowitz et al. (2017) used Br concentration fields 231

from Schmidt et al. (2016) but these are now thought to be too high (Chen et al., 2017) and do 232

not include the known source of bromine radicals from debromination of sea salt aerosols (SSA) 233

(Zhu et al., 2019). Here we use updated oxidant and aerosol fields from GEOS-Chem version 234

12.9 as documented by Wang et al. (2021). This features major updates to the bromine 235

simulation of Schmidt et al. (2016), including mechanistic SSA debromination and less efficient 236

heterogeneous recycling of bromine radicals. The tropospheric mean Br and BrO concentrations 237

are 0.03 and 0.19 pptv, respectively, compared to 0.08 and 0.48 pptv in Schmidt et al. (2016). 238

The distributions of Br and BrO are shifted more to the marine boundary layer (MBL) because of 239

SSA debromination (Figure S4). As pointed out by Wang et al. (2021), available atmospheric 240

observations for BrO offer limited guidance for evaluating the model. Therefore, we also 241

conduct a sensitivity simulation replacing the Br and BrO fields with those from Schmidt et al. 242

(2016). We apply a diurnal scaling to the monthly mean oxidant concentrations using the Y–243

YO–O3–NO (Y≡Br, Cl) photochemical equilibrium for the daytime concentrations of Br, BrO, 244

Cl, ClO following Holmes et al. (2010); a cosine function of the solar zenith angle for daytime 245

OH and HO2; and NO–NO2–O3 photochemical equilibrium for NO2. Br and BrO concentrations 246

in the polar springtime boundary layer are calculated following Fisher et al. (2012). 247

248

12

We treat the transfer of HgII between the gas phase and the aerosol/cloud phase as a kinetic 249

process. Individual gas-phase species Hg'"" are taken up by aerosols and cloud droplets where 250

they are re-speciated to the HgIIP(org) and HgIIP(inorg) forms, and then volatilized (for aerosols) 251

as the HgIIX form. The rate of uptake and volatilization of HgII gaseous species is calculated as 252

(Schwartz, 1986; Jacob, 2000): 253

254

− ()*+!""(+).(/

= 𝑘0/2Hg'""(g)5 (1) 255

()*+""1(+).(/

= 𝑘0/[Hg""(g)]23 , (2) 256

257

where 2Hg'""(g)5 is the number density of Hg'""(g), 𝑘0/ is the mass transfer rate coefficient (s-1), 258

and [Hg""(g)]23 is calculated on the basis of equilibrium between total HgII in the gas and aerosol 259

phases using the empirical equilibrium constant of Amos et al. (2012) as a function of local 260

temperature and mass concentration of fine particulate matter. For cloud droplets, we assume no 261

mass transfer back to the gas phase because of the high solubility of HgII. Uptake on coarse-262

mode SSA follows Holmes et al. (2010). 𝑘0/ for aerosols is calculated as: 263

264

𝑘0/ = ∑ 𝑆4 :5#6$+ &

78;9#

4 , (3) 265

266

where 𝑟4 and 𝑆4 are the effective mean area-weighted radius and surface area per unit volume of 267

air of each aerosol component (j), 𝐷: is the gas-phase molecular diffusion coefficient of HgII gas, 268

𝜈 is the mean molecular speed of HgII gas, and 𝛼 is the mass accommodation coefficient. We 269

take 𝛼 = 0.1 for all HgII gas species in the model since 𝛼 for other highly soluble species 270

13

generally has values of 0.1–0.3 (Davidovits et al., 2006). 𝑘0/ for cloud droplets is calculated 271

similarly but also accounts for entrainment limitation in partly cloudy grid cells (Holmes et al., 272

2019). 273

274

Results and discussion 275

Global atmospheric Hg budget 276

Figure 1 shows the model Hg budget in the lower atmosphere (troposphere) and the major 277

pathways for Hg0/ HgI/HgII redox cycling. The tropospheric mass of Hg is 4 Gg (3.9 Gg as Hg0 278

and 0.1 Gg as HgII). The stratosphere contains an additional 0.8 Gg (not shown in Fig. 1). The 279

tropospheric lifetime of total Hg (Hg0+HgI+HgII) against deposition is 5.5 months. The simulated 280

Hg mass and lifetime are within observationally constrained values of ~4 Gg for the tropospheric 281

Hg mass and 4–7 months for the Hg lifetime (Holmes et al., 2010; De Simone et al., 2014; 282

Horowitz et al., 2017; Saiz-Lopez et al., 2020). The previous GEOS-Chem simulation (Horowitz 283

et al., 2017) had a tropospheric mass of Hg of 3.9 Gg and a lifetime against deposition of 5.2 284

months, similar to ours, but four times as much HgII (0.4 Gg) because of production at higher 285

altitudes (longer lifetime against deposition) and slower photoreduction. 286

287

We find that oxidation of Hg0 to HgI takes place by Br and OH at similar rates. Ozone is the 288

primary oxidant of HgI to HgII as it is far more abundant than NO2 and HO2, which were the 289

main HgI oxidants in previous mechanisms (Horowitz et al., 2017; Saiz-Lopez et al., 2018, 290

2020). Photolysis and thermal decomposition of BrHgI are much slower than its reaction with 291

ozone, so the main fate of BrHgI is oxidation to BrHgIIOH, via BrHgIIO. Although HOHgI is less 292

stable than BrHgI and a smaller fraction of it is converted to HgII, the OH-initiated pathway still 293

14

accounts for one-third of the global HgII production. The chemical lifetime of Hg0 against 294

oxidation to HgII in our model is 4.5 months, compared to 2.7 months in Horowitz et al. (2017) 295

and about 13 months in Saiz-Lopez et al. (2018). Using higher free tropospheric Br 296

concentrations from Schmidt et al. (2016) lowers the tropospheric Hg mass by about 10% due to 297

increased partitioning to HgII and hence faster deposition. Br then contributes 75% of Hg0 298

oxidation (Figure S4). 299

300

Figure 2 shows the zonal distribution of the Hg0 oxidation rate in our standard simulation. Gross 301

oxidation of Hg0 to HgII is fastest in the MBL and in the upper troposphere and largely reflects 302

the Br distribution. Br concentrations are highest near the tropical tropopause due to fast 303

photolysis of BrO and low ozone and temperature (Fernandez et al., 2014; Saiz‐Lopez and 304

Fernandez, 2016). The OH-initiated oxidation pathway contributes most to Hg0 oxidation in the 305

tropical free troposphere, as dissociation of HOHgI is fast at lower altitudes (Dibble et al., 2020). 306

It also dominates in the continental boundary layer, consistent with Gabay et al. (2020), because 307

Br concentrations are low there. 308

309

BrHgIIOH and HgII(OH)2 are the main HgII species initially formed from Hg0 oxidation, but the 310

HgII speciation evolves as these species are processed by aerosol and cloud droplets to form 311

HgIIP particles and HgIIX gas. We find that HgIIX is the most abundant form of HgII in the 312

troposphere, comprising 49% of HgII mass, while HgIIP comprises 22%. The remaining HgII 313

mass is mostly composed of HgII(OH)2, which is more abundant than BrHgIIOH because it does 314

not photolyze. Most of the reduction of HgII to Hg0 is through the aqueous-phase photolysis of 315

HgIIP(org). The photoreduction rate increases with altitude because of stronger UV radiation and 316

15

the higher HgII particle fraction at lower temperatures, and it is faster in the northern hemisphere 317

because of the higher fraction of OA. HgIIP is stable against photoreduction in the stratosphere as 318

it is assumed to be present as free Hg2+. 319

320

The net rate of oxidation of Hg0 to HgII, accounting for HgII reduction, is 43% of the gross Hg0 321

oxidation rate. Net Hg0 oxidation is fastest in the MBL where HgII photoreduction is slower than 322

deposition. Horowitz et al. (2017) found little net oxidation in the lower troposphere because 323

their simulation had little Br in the MBL and did not include the HgI + O3 reaction. They had 324

maximum production in the tropical upper troposphere, but here this is largely canceled by 325

photoreduction and we find areas of net reduction as the HgII-rich tropical upper tropospheric air 326

is transported poleward by the Hadley circulation. Globally, we find that about half of the net 327

oxidation of Hg0 to HgII takes place through the OH-initiated pathway, compared to one-third for 328

gross oxidation, because of the stability of HgII(OH)2 against photolysis. 329

330

Our results differ substantially from the global model simulation of Saiz-Lopez et al. (2020). 331

They found that including the photolysis of HgI and HgII species increased the Hg lifetime to 20 332

months and the tropospheric Hg mass to 7.9 Gg, twice higher than inferred from atmospheric 333

observations. Including the BrHgI+O3 reaction lowered the tropospheric Hg lifetime to 15 334

months, which is still too high. The Hg lifetime in their model would have been even longer had 335

they included the recent findings on the reduction of BrHgIIO by CO (Khiri et al., 2020) and the 336

slower BrHgI+NO2 rate coefficient (Wu et al., 2020). The main reasons why we achieve a shorter 337

Hg lifetime are because we include the HOHgI+O3 reaction, which accounts for half of the net 338

16

chemical loss of Hg0 in our model, and the re-speciation of photolabile HgII species in aerosols 339

and cloud droplets to form the photostable HgIIX gaseous species. 340

341

Spatial distribution of Hg concentrations and deposition 342

Figure 3 shows the modeled zonal distributions of Hg0 and HgII and compares modeled and 343

observed Hg0 concentrations at the surface. There is little variation in Hg0 concentrations with 344

altitude in the troposphere, both in the model and in aircraft measurements (Talbot et al., 2008; 345

Mao et al., 2010; Slemr et al., 2018), consistent with the long lifetime of Hg0. Modeled Hg0 346

concentrations decrease by ~50 ppq within a height of 3 km above the tropopause, which is 347

somewhat lower than the decrease (~70 ppq) observed from aircraft (Slemr et al., 2018), This 348

could reflect excessive mixing across the tropopause in the 4º×5º version of the model 349

(Stanevich et al., 2020). The model captures the observed spatial patterns in surface Hg0 350

concentrations (r = 0.86), which are driven by anthropogenic emissions and the interhemispheric 351

gradient, but it underestimates the observed variability. The model overestimates the observed 352

Hg0 concentrations in the southern hemisphere by about 20 ppq but this could reflect uncertainty 353

in ocean Hg0 emissions (Horowitz et al., 2017). 354

355

HgII concentrations increase with altitude in the troposphere – from 1 ppq near the surface to 15 356

ppq at the tropopause – reflecting the sink from deposition. Concentrations are highest in the 357

subtropics due to subsidence of HgII produced in the tropical upper troposphere (Selin and Jacob, 358

2008; Shah and Jaeglé, 2017). This pattern is consistent with aircraft HgII observations over the 359

US (Gratz et al., 2015; Shah et al., 2016), but the model does not reproduce the observed 360

concentrations (~25 ppq at 6 km altitude), even with the higher Br concentrations from Schmidt 361

17

et al. (2016). We find that the free tropospheric HgII concentrations in the model depend strongly 362

on the representation of aqueous photoreduction. We conducted a simulation in which aqueous 363

HgII photoreduction was limited to liquid cloud droplets and HgIIP(org) formation on aerosol 364

particles was excluded, similar to Saiz-Lopez et al. (2018, 2020), and found a doubling of HgII 365

concentrations in the free troposphere, but the cloud HgIIP(org) photoreduction frequency 366

required to maintain the Hg lifetime was much higher and inconsistent with the rainwater 367

observations of Saiz-Lopez et al. (2018). 368

369

Figure 4 shows the observed and modeled HgII wet deposition flux and the modeled total (wet + 370

dry) HgII deposition flux. The mean Hg wet deposition flux for the global ensemble of sites is 371

25% lower in our model than in the observations. The model shows maximum wet deposition 372

flux over eastern China because of high anthropogenic HgII emissions, but this is not seen in 373

observations, suggesting that China’s HgII emissions may be overestimated due to insufficient 374

accounting of recent emission controls (Liu et al., 2019). The wet deposition flux over China in 375

Horowitz et al. (2017) was much smaller because of fast OA-dependent aqueous-phase HgII 376

photoreduction in their simulation. Our simulation captures the observed spatial variation in HgII 377

wet deposition flux over the continental US and Canada, showing a regional maximum over the 378

Southeast US, although it underestimates the magnitude, reflecting the underestimate of HgII in 379

the free troposphere. This Southeast US maximum is known to result from deep convective 380

scavenging of HgII-rich air (Guentzel et al., 2001; Zhang et al., 2012; Holmes et al., 2016). 381

382

The global HgII wet deposition flux in our standard simulation is 2.6 Gg a-1 and the total (wet + 383

dry) HgII deposition flux is 5.5 Gg a-1. Another 1.2 Gg a-1 is dry deposited to land as Hg0. We 384

18

find that 71% of HgII deposition takes place over the oceans, where it is the main source of Hg 385

for the marine biosphere (Sunderland and Mason, 2007; Mason et al., 2012), and is 15% higher 386

in the northern than in the southern hemisphere. Horowitz et al. (2017) found a higher fraction 387

(82%) of HgII deposition over the oceans because of fast HgII reduction over land due to high 388

OA. Holmes et al. (2010) found that 72% of the HgII deposition takes place over the oceans, but 389

with a higher flux in the southern hemisphere than in the northern hemisphere, reflecting the Br 390

distribution in their simulation. HgII deposition over land in our simulation is concentrated 391

largely in HgII emission hotspots over China, India, and South Africa. Outside of these hotspots, 392

Hg0 dry deposition is the major route for Hg deposition over land. 393

394

Uncertainties in atmospheric Hg redox chemistry 395

We have aimed to provide a mechanistic representation of Hg redox cycling in the atmosphere 396

that reflects current chemical knowledge and matches the major features of the observed 397

atmospheric Hg concentrations and fluxes. This involved a number of assumptions and here we 398

examine the most consequential. 399

400

An important uncertainty is the oxidation rate of Hg0 by Br, reflecting both the reaction rate 401

coefficient and the Br concentrations. Laboratory determinations of the Hg0+Br rate coefficient 402

vary from 3.6×10-13 to 3.2×10-12 cm3 molec-1 s-1 (at 298K, 1 atm) (Ariya et al., 2002; Donohoue 403

et al., 2006; Sun et al., 2016). We use the rate coefficient from Donohoue et al. (2006), which is 404

at the low end, because their measurements are least affected by wall reactions and were made 405

over a range of pressures and temperatures. Using a higher value would require slower 406

conversion of BrHgI to HgII, faster HgII reduction, and/or lower Br concentrations to maintain the 407

19

Hg lifetime in the model. The BrHgI®HgII rate can be slowed by lowering the BrHgI+O3 and 408

increasing the BrHgIO+CO rate coefficients, and while the changes needed are substantial (factor 409

of 10) since the competing BrHgI®Hg0 reactions are currently negligible, they would be within 410

the uncertainties of the theoretically derived rate coefficients (Ariya and Peterson, 2005). The 411

HgII reduction rates could be increased be a faster aqueous HgII photoreduction rate, which 412

currently has limited experimental constraints (Saiz-Lopez et al., 2018; Yang et al., 2019). Faster 413

Hg0+Br kinetics could be offset by lower Br concentrations, but the concentrations used here are 414

at the low end of current models as discussed by Wang et al. (2019, 2021). 415

416

The atmospheric OH concentrations are well known (Holmes et al., 2013) and the Hg0+OH rate 417

coefficient agrees between two independent laboratory studies (Sommar et al., 2001; Pal and 418

Ariya, 2004a), although the pressure and temperature dependences of the rate coefficient need to 419

be further investigated. There are larger uncertainties in the HOHgI+M, HOHgI+O3, 420

HOHgIIO+CO, and HOHgIIO+CH4 reactions that control the branching between HOHgI®Hg0 421

and HOHgI®HgII(OH)2. The HOHgI+M rate coefficient strongly depends on the HO–HgI bond 422

strength, which has not been determined experimentally (Dibble et al., 2020). A moderate 423

change in this rate coefficient could be balanced by proportional changes in the rate coefficients 424

of the other three reactions. Slower dissociation of HOHgI (stronger bond) would increase net 425

Hg0 oxidation in the subtropical free troposphere and help improve the simulation of the 426

observed Hg wet deposition flux maximum over the Southeast US. 427

428

An important part of our mechanism is the photoreduction of HgII to Hg0 in aerosols and cloud 429

droplets, but there is little data to constrain the photoreduction rates. This needs better 430

20

characterization over a range of atmospheric conditions and including identification of the photo-431

reducing species. Dissolved organic carbon is known to be critical for HgII photoreduction in 432

aquatic systems, but the effect of organic ligands on aqueous HgII in the atmosphere has 433

generally been ignored. A better understanding of particulate and cloud HgII speciation, and the 434

implications for photoreduction, would greatly advance our modeling capability. 435

21

Table 1a. GEOS-Chem Hg chemical mechanism: bimolecular and three-body reactions 436

Reaction Rate coefficients a References b

Hg0 + Br + M ® BrHgI + M k0 = 1.46 × 10-32 (T/298)-1.86 (1)

BrHgI + M ® Hg0 + Br + M k0/Keq; Keq = 9.14 × 10-24 exp(7801/T) (2)

Hg0 + OH + M ® HOHgI + M k0 = 3.34 × 10-33 exp(43/T) (3) c

HOHgI + M ® Hg0 + OH + M k0/Keq; Keq = 2.74 × 10-24 exp(5770/T) (4)

Hg0 + Cl + M ® ClHgI + M k0 = 2.25 × 10-33 exp(680/T) (5)

YHgI + O3 ® YHgIIO + O2

(Y ≡Br, OH, Cl) 3.0 × 10-11 (6) d, e

YHgIIO + CH4 ® YHgIIOH + CH3 (Y≡Br, OH, Cl)

4.1 × 10-12 exp(-856/T) (7) e

YHgIIO + CO ® YHgI + CO2 (Y≡Br, OH, Cl) 6.0 × 10-11 exp(-550/T) (8) e, f

YHgI + NO2 + M ® YHgIIONO + M (Y≡Br, OH, Cl)

k0 = 4.3 × 10-30 (T/298)-5.9 k∞ = 1.2 × 10-10 (T/298)-1.9 (9, 10) e

YHgI + Z + M ® YHgIIZ + M (Y ≡Br, OH, Cl; Z ≡HO2, BrO, ClO)

k0 = 4.3 × 10-30 (T/298)-5.9 k∞ = 6.9 × 10-11 (T/298)-2.4 (9, 10) e, g

YHgI + Z (+ M)® YHgIIZ (+ M) (Y≡Br, OH, Cl; Z≡Br, Cl, OH) 3.0 × 10-11 (11) e, h

YHgI + NO2 ® Hg0 + YNO2

(Y≡Br, Cl) 3.0 × 10-12 (9) e

BrHgI + Br ® Hg0 + Br2 3.9 × 10-11 (11)

ClHgI + Cl ® Hg0 + Cl2 1.2 × 10-11 exp(-5942/T) (12)

437

a. The rate coefficients have units of cm3 molec-1 s-1 for bimolecular reactions and cm6 molec-2 s-1 for k0 of 438 three-body reactions. The second-order rate coefficient for three-body reactions is calculated as: 𝑘([𝑀]) =439 ) %!['])*%!['] %"⁄ * 0.6,, where [M] is the number density of air molecules and 𝑝 = (1 +440 (log)-(𝑘-[𝑀] 𝑘.⁄ ))/)0). Only k0 is given when the low-pressure limit dominates in the atmosphere and 441 the second-order rate coefficient is then calculated as 𝑘-[𝑀]. For thermal dissociation reactions, the rate 442 coefficient is calculated as k = k0/Keq where k0 is the rate coefficient of the forward (association) reaction 443 given in the preceding entry and Keq is the equilibrium constant in units of cm3 molec-1. T is absolute 444 temperature in K. 445 446

b. (1) Donohoue et al. (2006); (2) Dibble et al. (2012); (3) Pal and Ariya (2004a); (4) Dibble et al. (2020); (5) 447 Donohoue et al. (2005); (6) Saiz-Lopez et al. (2020); (7) Lam et al. (2019b); (8) Khiri et al. (2020); (9) Wu 448 et al. (2020); (10) Jiao & Dibble (2017a); (11) Balabanov et al. (2005); (12) Wilcox (2009). 449 450

c. The rate coefficient was calculated by Dibble et al. (2020) from the experimental results of Pal and Ariya 451 (2004a). 452 453

22

d. Saiz-Lopez et al. (2020) estimated an upper limit for the rate coefficient of 1.0 × 10-10 cm3 molec-1 s-1, 454 assuming no steric effects. 455

456 e. We assume that the BrHgI + Z rate coefficients hold for HOHgI + Z and ClHgI + Z because of the similar 457

bond energies and reactions pathways for the three species (Dibble et al., 2012, 2020) and that the 458 BrHgIIO+Z rate coefficients hold for HOHgIIO+Z and ClHgIIO+Z. 459

460 f. Khiri et al. (2020) calculated the range for the rate coefficient of the BrHgIIO+CO®BrHgI+CO2 reaction at 461

two temperatures: (9.4–52)×10-12 at 298K and (3.8–29)×10-12 at 220K. We use the mean values at each 462 temperature to determine the temperature-dependent rate coefficient. 463

464 g. We assume that the experimentally determined value of k0 for the BrHgI + NO2 reaction (Wu et al., 2020) 465

holds for this set of reactions too. 466 467

h. These reactions take place at the high-pressure limit in the atmosphere. 468

23

Table 1b. GEOS-Chem Hg chemical mechanism: photolysis reactions a 469

Reaction f J (s-1) b References c BrHgI + hn ® Hg0 + Br 1.0 4.3×10-2 (1)

HOHgI + hn ® Hg0 + OH 1.0 1.6×10-2 (1)

YHgIIOH + hn ® Hg0 + Y + OH ® HOHgI + Y ® YHgI + OH ® YHgIIO + H (Y≡Br, Cl)

0.49 0.35 0.15 0.01

1.3×10-5

(2, 3, 4) d

YHgIIONO + hn ® YHgIIO + NO ® YHgI + NO2 (Y ≡Br, Cl, OH)

0.90 0.10

1.1×10-3 (2, 3, 4, 5) d,e

YHgIIOOH + hn ® Hg0 + Y + HO2 ® YHgIIO + OH ® YHgI + HO2 (Y ≡ Br, Cl, OH)

0.66 0.31 0.03

1.5×10-2 (2, 3, 4) d,e

YHgIIOBr + hn ® YHgI + BrO (Y≡ Br, OH, Cl)

1.0 f 2.4×10-2 (2) e

YHgIIOCl + hn ® YHgI + ClO (Y ≡Br, OH, Cl)

1.0 f 1.4×10-2 (2) e

HgIIBr2 + hn ® BrHgI + Br ® Hg0 + 2Br

0.60 0.40

1.2×10-6 (2, 4) d

HgIIP(org) + hn ® Hg0 1.0 1.9×10-5 this work g

470

a. Photolysis frequencies are calculated using Fast-JX v7.0a (Wild and Prather, 2000) implemented in GEOS-471 Chem by Eastham et al. (2014). f represents the branching fractions for the dissociation channels. 472 HgII(OH)2 and HgIICl2 are not shown in the table because they do not photolyze at tropospheric 473 wavelengths (Strömberg et al., 1989; Saiz-Lopez et al., 2018). 474 475

b. Global annual mean tropospheric photolysis frequencies in GEOS-Chem. 476 477

c. (1) Saiz-Lopez et al. (2019); (2) Saiz-Lopez et al. (2018); (3) Francés‐Monerris et al. (2020); (4) Saiz-478 Lopez et al. (2020); (5) Lam et al. (2019b) 479

480 d. Photolysis cross-sections are from Saiz-Lopez et al. (2020) and branching fractions from Francés‐Monerris 481

et al. (2020) and Saiz-Lopez et al. (2020). 482 483

e. Photolysis cross-sections of the HOHgIIZ and ClHgIIZ (Z=NO2, HO2, BrO, and ClO) species are assumed 484 to be equal to those of BrHgIZ. 485

486 f. Sole photolysis pathway considered in Saiz-Lopez et al. (2018). 487

488 g. The photolysis frequency of this reaction is parameterized as JHgIIP(org) =b JNO2, where JNO2 is the local 489

photolysis frequency of NO2 and the scaling factor b is adjusted to match the global mean Hg0 surface 490

24

observations. For our standard simulation we use β = 4×10-3, and for the simulation with the Schmidt et al. 491 (2016) Br fields we use β = 4×10-2. 492

25

Table 1c. GEOS-Chem Hg chemical mechanism: multiphase processes 493

Reactiona Notes

HgII (g) 23456576,975:;65⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯7 HgIIP b

HgIIP 234565765⎯⎯⎯⎯⎯⎯⎯7 HgIIX (g) c

HgIIP≡ 8HgIIP(org)+HgIIP(inorg)(troposphere)

Hg2+(stratosphere) d

494 a. HgII (g) and HgIIP represent all gas- and particle-phase HgII species; HgIIX (g) represents the unspeciated 495

HgII gas volatilizing from HgIIP. 496 497

b. Uptake rate is given by Eq (1). For clouds the HgII (g) uptake rate accounts for entrainment limitation in 498 partly cloudy grid cells (Holmes et al., 2019). 499

500 c. For tropospheric aerosols only. We assume irreversible uptake in cloud droplets and stratospheric aerosols. 501

The volatilization rate is given by Eq (2) with equilibrium constant between HgII(g) and HgIIP from Amos 502 et al. (2012). 503

504 d. HgIIP in the tropospheric aerosol speciates as HgIIP(org) and HgIIP(inorg) representing HgII-organic and 505

HgII-inorganic complexes. Their concentrations are calculated as >HgIIP(org)? = fOA>HgIIP? and 506

>HgIIP(inorg)? = (1 − 𝑓?@)>HgIIP?, where 𝑓AB is the local mass fraction of organic aerosols computed 507 as𝑓?@ =

C#$C#$*C%&$*C%%$

, with 𝑚?@, 𝑚DE@, and 𝑚DD@ as the mass concentrations of organic aerosols, 508 sulfate-nitrate-ammonium aerosols, and accumulation mode sea-salt aerosols. 509

510 511

26

512 Figure 1. Global tropospheric Hg budget and main Hg redox pathways in our simulation 513 for 2013–2015. The Hg masses and rates are global annual means given in units of Gg and 514 Gg a-1 respectively. The main HgII species in the model and their percent contributions to 515 the total tropospheric HgII mass are listed. HgIIP denotes particulate HgII, which includes 516 HgII-organic complexes (HgIIP(org)), and HgII-inorganic complexes (HgIIP(inorg)). HgIIX 517 denotes the gas-phase HgII species that volatilize from HgIIP and is modeled as HgCl2. 518 Oxidation of Hg0 by Cl atoms is not shown because it accounts only for <1% of the Hg0 519 chemical sink in the troposphere. 520

27

521 Figure 2. Annual (2013–2015) zonal mean gross and net Hg0 oxidation rates in GEOS-522 Chem. The contour lines show the percent contribution of the OH-initiated Hg0 oxidation 523 pathway. The dashed line denotes the annual-mean tropopause. 524

28

525 Figure 3. Annual mean (2013–2015) concentrations of Hg0 and HgII in GEOS-Chem. The 526 top panels are zonal mean concentrations as a function of pressure and latitude. The 527 dashed lines indicate the annual mean tropopause. The bottom panel compares the 528 modeled surface Hg0 concentrations with observations (filled circles and triangles) from the 529 compilations of Travnikov et al. (2017) (courtesy of Hélène Angot) and AMAP/UNEP 530 (2019). Filled triangles represent high altitude sites. We only include observations made 531 between 2010 and 2015. The mean ± standard deviation of the observed concentrations is 532 inset in the bottom panel along with the corresponding model values sampled at the site 533 locations. The color scales are different for each panel. 534

29

535 Figure 4. Annual mean (2013–2015) HgII deposition fluxes in GEOS-Chem. The top panels 536 show HgII wet deposition fluxes overlaid by observations (filled circles and triangles) 537 compiled by Travnikov et al. (2017) (courtesy of Hélène Angot), Sprovieri et al. (2017), 538 AMAP/UNEP (2019) and Fu et al. (2016). Filled triangles represent high altitude sites. We 539 only include observations collected between 2010 and 2015. Values inset are the means ± 540 standard deviations for the global ensemble of sites (left panel) and for the subset of sites 541 over the contiguous US and Canada (right panel). The bottom panel shows total (wet + dry) 542 HgII deposition fluxes. 543

30

ASSOCIATED CONTENT 544 Supporting Information 545 Supplementary Information (PDF) 546

AUTHOR INFORMATION 547 Corresponding Author 548 *Viral Shah (vshah@seas.harvard.edu) 549 Author Contributions 550 V.S. and D.J.J. developed the chemical mechanism with input from C.P.T., E.M.S., T.S.D., and A.S.-L.; 551 V.S. and C.P.T. implemented the mechanism in GEOS-Chem; X.W. developed the current GEOS-Chem 552 halogen simulation; T.S.D, I.Č, V.K, and P.J.C. performed the quantum chemistry calculations for the 553 reactions of BrHg and HOHg with ozone; R.W. and C.W. conducted laboratory experiments on the BrHg 554 and ozone reaction; V.S. and D.J.J. wrote the manuscript with contributions from all coauthors. 555

ACKNOWLEDGMENTS 556 This work was funded by the US EPA Science to Achieve Results (STAR) Program. This work was also 557 supported by the Slovak Grant Agency VEGA (grant 1/0777/19), the high-performance computing facility 558 of the Centre for Information Technology (https://uniba.sk/en/HPC-Clara) at Comenius University, and 559 the US National Science Foundation under awards 1609848 and 2004100. We thank Hélène Angot (CU 560 Boulder) for the surface measurement data. 561

31

REFERENCES 562 AMAP/UNEP: Technical Background Report for the Global Mercury Assessment 2018, Arctic 563 Monitoring and Assessment Programme, Oslo, Norway/UN Environment Programme, Chemicals and 564 Health Branch, Geneva, Switzerland., 2019. 565

Amos, H. M., Jacob, D. J., Holmes, C. D., Fisher, J. A., Wang, Q., Yantosca, R. M., Corbitt, E. S., 566 Galarneau, E., Rutter, A. P., Gustin, M. S., Steffen, A., Schauer, J. J., Graydon, J. A., Louis, V. L. St., 567 Talbot, R. W., Edgerton, E. S., Zhang, Y., and Sunderland, E. M.: Gas-Particle Partitioning of 568 Atmospheric Hg(II) and its Effect on Global Mercury Deposition, Atmos. Chem. Phys., 12, 591–603, 569 https://doi.org/10.5194/acp-12-591-2012, 2012. 570

Amyot, Marc., McQueen, D. J., Mierle, Greg., and Lean, D. R. S.: Sunlight-Induced Formation of 571 Dissolved Gaseous Mercury in Lake Waters, Environ. Sci. Technol., 28, 2366–2371, 572 https://doi.org/10.1021/es00062a022, 1994. 573

Ariya, P. A. and Peterson, K. A.: Chemical Transformation of Gaseous Elemental Hg in the Atmosphere, 574 in: Dynamics of Mercury Pollution on Regional and Global Scales, edited by: Pirrone, N. and Mahaffey, 575 K. R., Springer-Verlag, New York, 261–294, https://doi.org/10.1007/0-387-24494-8_12, 2005. 576

Ariya, P. A., Khalizov, A., and Gidas, A.: Reactions of Gaseous Mercury with Atomic and Molecular 577 Halogens: Kinetics, Product Studies, and Atmospheric Implications, J. Phys. Chem. A, 106, 7310–7320, 578 https://doi.org/10.1021/jp020719o, 2002. 579

Balabanov, N. B., Shepler, B. C., and Peterson, K. A.: Accurate Global Potential Energy Surface and 580 Reaction Dynamics for the Ground State of HgBr2, J. Phys. Chem. A, 109, 8765–8773, 581 https://doi.org/10.1021/jp053415l, 2005. 582

Calvert, J. and Lindberg, S.: Mechanisms of Mercury Removal by O3 and OH in the Atmosphere, Atmos. 583 Environ., 39, 3355–3367, https://doi.org/10.1016/j.atmosenv.2005.01.055, 2005. 584

Chen, Q., Schmidt, J. A., Shah, V., Jaeglé, L., Sherwen, T., and Alexander, B.: Sulfate Production by 585 Reactive Bromine: Implications for the Global Sulfur and Reactive Bromine Budgets, Geophysical 586 Research Letters, 44, 7069–7078, https://doi.org/10.1002/2017GL073812, 2017. 587

Damian, V., Sandu, A., Damian, M., Potra, F., and Carmichael, G. R.: The Kinetic Preprocessor KPP –- 588 A Software Environment for Solving Chemical Kinetics, Comput. Chem. Eng., 26, 1567–1579, 589 https://doi.org/10.1016/S0098-1354(02)00128-X, 2002. 590

Dastoor, A., Ryzhkov, A., Durnford, D., Lehnherr, I., Steffen, A., and Morrison, H.: Atmospheric 591 Mercury in the Canadian Arctic. Part II: Insight from Modeling, Sci. Total Environ., 509–510, 16–27, 592 https://doi.org/10.1016/j.scitotenv.2014.10.112, 2015. 593

Davidovits, P., Kolb, C. E., Williams, L. R., Jayne, J. T., and Worsnop, D. R.: Mass Accommodation and 594 Chemical Reactions at Gas−Liquid Interfaces, Chem. Rev., 106, 1323–1354, 595 https://doi.org/10.1021/cr040366k, 2006. 596

De Simone, F., Gencarelli, C. N., Hedgecock, I. M., and Pirrone, N.: Global Atmospheric Cycle of 597 Mercury: A Model Study on the Impact of Oxidation Mechanisms, Environ Sci Pollut Res, 21, 4110–598 4123, https://doi.org/10.1007/s11356-013-2451-x, 2014. 599

32

Dibble, T. S. and Schwid, A. C.: Thermodynamics Limits the Reactivity of BrHg Radical with Volatile 600 Organic Compounds, Chemical Physics Letters, 659, 289–294, 601 https://doi.org/10.1016/j.cplett.2016.07.065, 2016. 602

Dibble, T. S., Zelie, M. J., and Mao, H.: Thermodynamics of Reactions of ClHg and BrHg Radicals with 603 Atmospherically Abundant Free Radicals, Atmos. Chem. Phys., 12, 10271–10279, 604 https://doi.org/10.5194/acp-12-10271-2012, 2012. 605

Dibble, T. S., Tetu, H. L., Jiao, Y., Thackray, C. P., and Jacob, D. J.: Modeling the OH-Initiated 606 Oxidation of Mercury in the Global Atmosphere without Violating Physical Laws, J. Phys. Chem. A, 124, 607 444–453, https://doi.org/10.1021/acs.jpca.9b10121, 2020. 608

Donohoue, D. L., Bauer, D., and Hynes, A. J.: Temperature and Pressure Dependent Rate Coefficients for 609 the Reaction of Hg with Cl and the Reaction of Cl with Cl: A Pulsed Laser Photolysis−Pulsed Laser 610 Induced Fluorescence Study, J. Phys. Chem. A, 109, 7732–7741, https://doi.org/10.1021/jp051354l, 2005. 611

Donohoue, D. L., Bauer, D., Cossairt, B., and Hynes, A. J.: Temperature and Pressure Dependent Rate 612 Coefficients for the Reaction of Hg with Br and the Reaction of Br with Br: A Pulsed Laser Photolysis-613 Pulsed Laser Induced Fluorescence Study, J. Phys. Chem. A, 110, 6623–6632, 614 https://doi.org/10.1021/jp054688j, 2006. 615

Eastham, S. D., Weisenstein, D. K., and Barrett, S. R. H.: Development and Evaluation of the Unified 616 Tropospheric–Stratospheric Chemistry Extension (UCX) for the Global Chemistry-Transport Model 617 GEOS-Chem, Atmos. Environ., 89, 52–63, https://doi.org/10.1016/j.atmosenv.2014.02.001, 2014. 618

Fernandez, R. P., Salawitch, R. J., Kinnison, D. E., Lamarque, J.-F., and Saiz-Lopez, A.: Bromine 619 Partitioning in the Tropical Tropopause Layer: Implications for Stratospheric Injection, Atmos. Chem. 620 Phys., 14, 13391–13410, https://doi.org/10.5194/acp-14-13391-2014, 2014. 621

Filatov, M. and Cremer, D.: Revision of the Dissociation Energies of Mercury Chalcogenides–Unusual 622 Types of Mercury Bonding, ChemPhysChem, 5, 1547–1557, https://doi.org/10.1002/cphc.200301207, 623 2004. 624

Fisher, J. A., Jacob, D. J., Soerensen, A. L., Amos, H. M., Steffen, A., and Sunderland, E. M.: Riverine 625 Source of Arctic Ocean Mercury Inferred from Atmospheric Observations, Nature Geosci, 5, 499–504, 626 https://doi.org/10.1038/ngeo1478, 2012. 627

Francés‐Monerris, A., Carmona‐García, J., Acuña, A. U., Dávalos, J. Z., Cuevas, C. A., Kinnison, D. E., 628 Francisco, J. S., Saiz‐Lopez, A., and Roca‐Sanjuán, D.: Photodissociation Mechanisms of Major 629 Mercury(II) species in the Atmospheric Chemical Cycle of Mercury, Angew. Chem. Int. Ed., 59, 7605–630 7610, https://doi.org/10.1002/anie.201915656, 2020. 631

Fu, X., Yang, X., Lang, X., Zhou, J., Zhang, H., Yu, B., Yan, H., Lin, C.-J., and Feng, X.: Atmospheric 632 Wet and Litterfall Mercury Deposition at Urban and Rural Sites in China, Atmos. Chem. Phys., 16, 633 11547–11562, https://doi.org/10.5194/acp-16-11547-2016, 2016. 634

Gabay, M., Raveh-Rubin, S., Peleg, M., Fredj, E., and Tas, E.: Is Oxidation of Atmospheric Mercury 635 Controlled by Different Mechanisms in the Polluted Continental Boundary Layer vs. Remote Marine 636 Boundary Layer?, Environ. Res. Lett., 15, 064026, https://doi.org/10.1088/1748-9326/ab7b26, 2020. 637

33

Gårdfeldt, K. and Jonsson, M.: Is Bimolecular Reduction of Hg(II) Complexes Possible in Aqueous 638 Systems of Environmental Importance, J. Phys. Chem. A, 107, 4478–4482, 639 https://doi.org/10.1021/jp0275342, 2003. 640

Gelaro, R., McCarty, W., Suárez, M. J., Todling, R., Molod, A., Takacs, L., Randles, C. A., Darmenov, 641 A., Bosilovich, M. G., Reichle, R., Wargan, K., Coy, L., Cullather, R., Draper, C., Akella, S., Buchard, 642 V., Conaty, A., da Silva, A. M., Gu, W., Kim, G.-K., Koster, R., Lucchesi, R., Merkova, D., Nielsen, J. 643 E., Partyka, G., Pawson, S., Putman, W., Rienecker, M., Schubert, S. D., Sienkiewicz, M., and Zhao, B.: 644 The Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2), J. 645 Climate, 30, 5419–5454, https://doi.org/10.1175/JCLI-D-16-0758.1, 2017. 646

Goodsite, M. E., Plane, J. M. C., and Skov, H.: A Theoretical Study of the Oxidation of Hg0 to HgBr2 in 647 the Troposphere, Environ. Sci. Technol., 38, 1772–1776, https://doi.org/10.1021/es034680s, 2004. 648

Goodsite, M. E., Plane, J. M. C., and Skov, H.: Correction to A Theoretical Study of the Oxidation of 649 Hg0 to HgBr2 in the Troposphere, Environ. Sci. Technol., 46, 5262–5262, 650 https://doi.org/10.1021/es301201c, 2012. 651

Gratz, L. E., Ambrose, J. L., Jaffe, D. A., Shah, V., Jaeglé, L., Stutz, J., Festa, J., Spolaor, M., Tsai, C., 652 Selin, N. E., Song, S., Zhou, X., Weinheimer, A. J., Knapp, D. J., Montzka, D. D., Flocke, F. M., 653 Campos, T. L., Apel, E., Hornbrook, R., Blake, N. J., Hall, S., Tyndall, G. S., Reeves, M., Stechman, D., 654 and Stell, M.: Oxidation of Mercury by Bromine in the Subtropical Pacific Free Troposphere, Geophys. 655 Res. Lett., 42, 10,494-10,502, https://doi.org/10.1002/2015GL066645, 2015. 656

Guentzel, J. L., Landing, W. M., Gill, G. A., and Pollman, C. D.: Processes Influencing Rainfall 657 Deposition of Mercury in Florida, Environ. Sci. Technol., 35, 863–873, 658 https://doi.org/10.1021/es001523+, 2001. 659

Gustin, M. S., Amos, H. M., Huang, J., Miller, M. B., and Heidecorn, K.: Measuring and Modeling 660 Mercury in the Atmosphere: A Critical Review, Atmos. Chem. Phys., 15, 5697–5713, 661 https://doi.org/10.5194/acp-15-5697-2015, 2015. 662

Gustin, M. S., Dunham-Cheatham, S. M., Huang, J., Lindberg, S., and Lyman, S. N.: Development of an 663 Understanding of Reactive Mercury in Ambient Air: A Review, Atmosphere, 12, 73, 664 https://doi.org/10.3390/atmos12010073, 2021. 665

Haitzer, M., Aiken, G. R., and Ryan, J. N.: Binding of Mercury(II) to Dissolved Organic Matter: The 666 Role of the Mercury-to-DOM Concentration Ratio, Environ. Sci. Technol., 36, 3564–3570, 667 https://doi.org/10.1021/es025699i, 2002. 668

Hammer, M. S., van Donkelaar, A., Li, C., Lyapustin, A., Sayer, A. M., Hsu, N. C., Levy, R. C., Garay, 669 M. J., Kalashnikova, O. V., Kahn, R. A., Brauer, M., Apte, J. S., Henze, D. K., Zhang, L., Zhang, Q., 670 Ford, B., Pierce, J. R., and Martin, R. V.: Global Estimates and Long-Term Trends of Fine Particulate 671 Matter Concentrations (1998–2018), Environ. Sci. Technol., 54, 7879–7890, 672 https://doi.org/10.1021/acs.est.0c01764, 2020. 673

Hedgecock, I. M. and Pirrone, N.: Mercury and Photochemistry in the Marine Boundary Layer-Modelling 674 Studies Suggest the In Situ Production of Reactive Gas Phase Mercury, Atmos. Environ., 35, 3055–3062, 675 https://doi.org/10.1016/S1352-2310(01)00109-1, 2001. 676

34

Holmes, C. D., Jacob, D. J., and Yang, X.: Global Lifetime of Elemental Mercury against Oxidation by 677 Atomic Bromine in the Free Troposphere, Geophys. Res. Lett., 33, L20808, 678 https://doi.org/10.1029/2006GL027176, 2006. 679

Holmes, C. D., Jacob, D. J., Corbitt, E. S., Mao, J., Yang, X., Talbot, R., and Slemr, F.: Global 680 Atmospheric Model for Mercury Including Oxidation by Bromine Atoms, Atmos. Chem. Phys., 10, 681 12037–12057, https://doi.org/10.5194/acp-10-12037-2010, 2010. 682

Holmes, C. D., Prather, M. J., Søvde, O. A., and Myhre, G.: Future Methane, Hydroxyl, and their 683 Uncertainties: Key Climate and Emission Parameters for Future Predictions, Atmos. Chem. Phys., 13, 684 285–302, https://doi.org/10.5194/acp-13-285-2013, 2013. 685

Holmes, C. D., Krishnamurthy, N. P., Caffrey, J. M., Landing, W. M., Edgerton, E. S., Knapp, K. R., and 686 Nair, U. S.: Thunderstorms Increase Mercury Wet Deposition, Environ. Sci. Technol., 50, 9343–9350, 687 https://doi.org/10.1021/acs.est.6b02586, 2016. 688

Holmes, C. D., Bertram, T. H., Confer, K. L., Graham, K. A., Ronan, A. C., Wirks, C. K., and Shah, V.: 689 The Role of Clouds in the Tropospheric NOx Cycle: A New Modeling Approach for Cloud Chemistry and 690 Its Global Implications, 46, 4980–4990, https://doi.org/10.1029/2019GL081990, 2019. 691

Horowitz, H. M., Jacob, D. J., Zhang, Y., Dibble, T. S., Slemr, F., Amos, H. M., Schmidt, J. A., Corbitt, 692 E. S., Marais, E. A., and Sunderland, E. M.: A New Mechanism for Atmospheric Mercury Redox 693 Chemistry: Implications for the Global Mercury Budget, Atmos. Chem. Phys., 17, 6353–6371, 694 https://doi.org/10.5194/acp-17-6353-2017, 2017. 695

Hynes, A. J., Donohoue, D. L., Goodsite, M. E., and Hedgecock, I. M.: Our Current Understanding of 696 Major Chemical and Physical Processes Affecting Mercury Dynamics in the Atmosphere and at the Air-697 Water/Terrestrial Interfaces, in: Mercury Fate and Transport in the Global Atmosphere, edited by: Mason, 698 R. and Pirrone, N., Springer US, Boston, MA, 427–457, https://doi.org/10.1007/978-0-387-93958-2_14, 699 2009. 700

Jacob, D.: Heterogeneous Chemistry and Tropospheric Ozone, 34, 2131–2159, 701 https://doi.org/10.1016/S1352-2310(99)00462-8, 2000. 702

Jaffe, D. A., Lyman, S., Amos, H. M., Gustin, M. S., Huang, J., Selin, N. E., Levin, L., ter Schure, A., 703 Mason, R. P., Talbot, R., Rutter, A., Finley, B., Jaeglé, L., Shah, V., McClure, C., Ambrose, J., Gratz, L., 704 Lindberg, S., Weiss-Penzias, P., Sheu, G.-R., Feddersen, D., Horvat, M., Dastoor, A., Hynes, A. J., Mao, 705 H., Sonke, J. E., Slemr, F., Fisher, J. A., Ebinghaus, R., Zhang, Y., and Edwards, G.: Progress on 706 Understanding Atmospheric Mercury Hampered by Uncertain Measurements, Environ. Sci. Technol., 48, 707 7204–7206, https://doi.org/10.1021/es5026432, 2014. 708

Jiao, Y. and Dibble, T. S.: First Kinetic Study of the Atmospherically Important Reactions BrHg• + NO2 709 and BrHg• + HOO, Phys. Chem. Chem. Phys., 19, 1826–1838, https://doi.org/10.1039/C6CP06276H, 710 2017a. 711

Jiao, Y. and Dibble, T. S.: Structures, Vibrational Frequencies, and Bond Energies of the BrHgOX and 712 BrHgXO Species Formed in Atmospheric Mercury Depletion Events, J. Phys. Chem. A, 121, 7976–7985, 713 https://doi.org/10.1021/acs.jpca.7b06829, 2017b. 714

35

Khiri, D., Louis, F., Černušák, I., and Dibble, T. S.: BrHgO• + CO: Analogue of OH + CO and Reduction 715 Path for Hg(II) in the Atmosphere, ACS Earth Space Chem., 4, 1777–1784, 716 https://doi.org/10.1021/acsearthspacechem.0c00171, 2020. 717

Lam, K. T., Wilhelmsen, C. J., and Dibble, T. S.: BrHgO• + C2H4 and BrHgO• + HCHO in Atmospheric 718 Oxidation of Mercury: Determining Rate Constants of Reactions with Prereactive Complexes and 719 Bifurcation, J. Phys. Chem. A, 123, 6045–6055, https://doi.org/10.1021/acs.jpca.9b05120, 2019a. 720

Lam, K. T., Wilhelmsen, C. J., Schwid, A. C., Jiao, Y., and Dibble, T. S.: Computational Study on the 721 Photolysis of BrHgONO and the Reactions of BrHgO• with CH4, C2H6, NO, and NO2: Implications for 722 Formation of Hg(II) Compounds in the Atmosphere, J. Phys. Chem. A, 123, 1637–1647, 723 https://doi.org/10.1021/acs.jpca.8b11216, 2019b. 724

Lin, C.-J. and Pehkonen, S. O.: Two-phase Model of Mercury Chemistry in the Atmosphere, Atmos. 725 Environ., 32, 2543–2558, https://doi.org/10.1016/S1352-2310(98)00002-8, 1998. 726

Lin, C.-J. and Pehkonen, S. O.: The Chemistry of Atmospheric Mercury: A Review, Atmos. Environ., 33, 727 2067–2079, https://doi.org/10.1016/S1352-2310(98)00387-2, 1999. 728

Lindberg, S. E., Brooks, S., Lin, C.-J., Scott, K. J., Landis, M. S., Stevens, R. K., Goodsite, M., and 729 Richter, A.: Dynamic Oxidation of Gaseous Mercury in the Arctic Troposphere at Polar Sunrise, Environ. 730 Sci. Technol., 36, 1245–1256, https://doi.org/10.1021/es0111941, 2002. 731

Liu, K., Wu, Q., Wang, L., Wang, S., Liu, T., Ding, D., Tang, Y., Li, G., Tian, H., Duan, L., Wang, X., 732 Fu, X., Feng, X., and Hao, J.: Measure-Specific Effectiveness of Air Pollution Control on China’s 733 Atmospheric Mercury Concentration and Deposition during 2013–2017, Environ. Sci. Technol., 53, 734 8938–8946, https://doi.org/10.1021/acs.est.9b02428, 2019. 735

Mao, H., Talbot, R. W., Sive, B. C., Youn Kim, S., Blake, D. R., and Weinheimer, A. J.: Arctic Mercury 736 Depletion and its Quantitative Link with Halogens, J Atmos Chem, 65, 145–170, 737 https://doi.org/10.1007/s10874-011-9186-1, 2010. 738

Mason, R. P., Choi, A. L., Fitzgerald, W. F., Hammerschmidt, C. R., Lamborg, C. H., Soerensen, A. L., 739 and Sunderland, E. M.: Mercury Biogeochemical Cycling in the Ocean and Policy Implications, 740 Environmental Research, 119, 101–117, https://doi.org/10.1016/j.envres.2012.03.013, 2012. 741

Munthe, J. and McElroy, W. J.: Some Aqueous Reactions of Potential Importance in the Atmospheric 742 Chemistry of Mercury, Atmos. Environ., 26, 553–557, https://doi.org/10.1016/0960-1686(92)90168-K, 743 1992. 744

Munthe, J., Xiao, Z. F., and Lindqvist, O.: The Aqueous Reduction of Divalent Mercury by Sulfite, Water 745 Air Soil Pollut, 56, 621–630, https://doi.org/10.1007/BF00342304, 1991. 746

Obrist, D., Tas, E., Peleg, M., Matveev, V., Faïn, X., Asaf, D., and Luria, M.: Bromine-induced Oxidation 747 of Mercury in the Mid-latitude Atmosphere, Nature Geosci, 4, 22–26, https://doi.org/10.1038/ngeo1018, 748 2011. 749

O’Driscoll, N. J., Lean, D. R. S., Loseto, L. L., Carignan, R., and Siciliano, S. D.: Effect of Dissolved 750 Organic Carbon on the Photoproduction of Dissolved Gaseous Mercury in Lakes: Potential Impacts of 751 Forestry, Environ. Sci. Technol., 38, 2664–2672, https://doi.org/10.1021/es034702a, 2004. 752

36

Pai, S. J., Heald, C. L., Pierce, J. R., Farina, S. C., Marais, E. A., Jimenez, J. L., Campuzano-Jost, P., 753 Nault, B. A., Middlebrook, A. M., Coe, H., Shilling, J. E., Bahreini, R., Dingle, J. H., and Vu, K.: An 754 Evaluation of Global Organic Aerosol Schemes using Airborne Observations, Atmos. Chem. Phys., 20, 755 2637–2665, https://doi.org/10.5194/acp-20-2637-2020, 2020. 756

Pal, B. and Ariya, P. A.: Gas-Phase HO•-Initiated Reactions of Elemental Mercury: Kinetics, Product 757 Studies, and Atmospheric Implications, Environ. Sci. Technol., 38, 5555–5566, 758 https://doi.org/10.1021/es0494353, 2004a. 759

Pal, B. and Ariya, P. A.: Studies of Ozone Initiated Reactions of Gaseous Mercury: Kinetics, Product 760 studies, and Atmospheric Implications, Phys. Chem. Chem. Phys., 6, 572, 761 https://doi.org/10.1039/b311150d, 2004b. 762

Pehkonen, S. O. and Lin, C.-J.: Aqueous Photochemistry of Mercury with Organic Acids, J. Air Waste 763 Manag. Assoc., 48, 144–150, https://doi.org/10.1080/10473289.1998.10463661, 1998. 764

Peterson, K. A., Shepler, B. C., and Singleton, J. M.: The Group 12 Metal Chalcogenides: An Accurate 765 Multireference Configuration Interaction and Coupled Cluster Study, Molecular Physics, 105, 1139–766 1155, https://doi.org/10.1080/00268970701241664, 2007. 767

Pirrone, N., Cinnirella, S., Feng, X., Finkelman, R. B., Friedli, H. R., Leaner, J., Mason, R., Mukherjee, 768 A. B., Stracher, G. B., Streets, D. G., and Telmer, K.: Global Mercury Emissions to the Atmosphere from 769 Anthropogenic and Natural Sources, Atmos. Chem. Phys., 10, 5951–5964, https://doi.org/10.5194/acp-770 10-5951-2010, 2010. 771

Qureshi, A., MacLeod, M., Sunderland, E., and Hungerbühler, K.: Exchange of Elemental Mercury 772 between the Oceans and the Atmosphere, in: Environmental Chemistry and Toxicology of Mercury, 773 edited by: Liu, G., Cai, Y., and O’Driscoll, N., John Wiley & Sons, Inc., Hoboken, NJ, USA, 389–421, 774 https://doi.org/10.1002/9781118146644.ch12, 2011. 775

Raofie, F. and Ariya, P. A.: Product Study of the Gas-Phase BrO-Initiated Oxidation of Hg0: Evidence for 776 Stable Hg1+ Compounds, Environ. Sci. Technol., 38, 4319–4326, https://doi.org/10.1021/es035339a, 777 2004. 778

Ravichandran, M.: Interactions Between Mercury and Dissolved Organic Matter––A Review, 779 Chemosphere, 55, 319–331, https://doi.org/10.1016/j.chemosphere.2003.11.011, 2004. 780

Saiz‐Lopez, A. and Fernandez, R. P.: On the Formation of Tropical Rings of Atomic Halogens: Causes 781 and Implications, Geophys. Res. Lett., 43, 2928–2935, https://doi.org/10.1002/2015GL067608, 2016. 782

Saiz-Lopez, A., Sitkiewicz, S. P., Roca-Sanjuán, D., Oliva-Enrich, J. M., Dávalos, J. Z., Notario, R., 783 Jiskra, M., Xu, Y., Wang, F., Thackray, C. P., Sunderland, E. M., Jacob, D. J., Travnikov, O., Cuevas, C. 784 A., Acuña, A. U., Rivero, D., Plane, J. M. C., Kinnison, D. E., and Sonke, J. E.: Photoreduction of 785 Gaseous Oxidized Mercury Changes Global Atmospheric Mercury Speciation, Transport and Deposition, 786 Nat. Commun., 9, 4796, https://doi.org/10.1038/s41467-018-07075-3, 2018. 787

Saiz-Lopez, A., Acuña, A. U., Trabelsi, T., Carmona-García, J., Dávalos, J. Z., Rivero, D., Cuevas, C. A., 788 Kinnison, D. E., Sitkiewicz, S. P., Roca-Sanjuán, D., and Francisco, J. S.: Gas-Phase Photolysis of Hg(I) 789 Radical Species: A New Atmospheric Mercury Reduction Process, J. Am. Chem. Soc., 141, 8698–8702, 790 https://doi.org/10.1021/jacs.9b02890, 2019. 791

37

Saiz-Lopez, A., Travnikov, O., Sonke, J. E., Thackray, C. P., Jacob, D. J., Carmona-García, J., Francés-792 Monerris, A., Roca-Sanjuán, D., Acuña, A. U., Dávalos, J. Z., Cuevas, C. A., Jiskra, M., Wang, F., 793 Bieser, J., Plane, J. M. C., and Francisco, J. S.: Photochemistry of Oxidized Hg(I) and Hg(II) Species 794 Suggests Missing Mercury Oxidation in the Troposphere, Proc Natl Acad Sci USA, 117, 30949–30956, 795 https://doi.org/10.1073/pnas.1922486117, 2020. 796

Schmidt, J. A., Jacob, D. J., Horowitz, H. M., Hu, L., Sherwen, T., Evans, M. J., Liang, Q., Suleiman, R. 797 M., Oram, D. E., Le Breton, M., Percival, C. J., Wang, S., Dix, B., and Volkamer, R.: Modeling the 798 Observed Tropospheric BrO Background: Importance of Multiphase Chemistry and Implications for 799 Ozone, OH, and Mercury, J. Geophys. Res. Atmos., 121, 11,819-11,835, 800 https://doi.org/10.1002/2015JD024229, 2016. 801

Schwartz, S. E.: Mass-Transport Considerations Pertinent to Aqueous Phase Reactions of Gases in 802 Liquid-Water Clouds, in: Chemistry of Multiphase Atmospheric Systems, edited by: Jaeschke, W., 803 Springer Berlin Heidelberg, Berlin, Heidelberg, 415–471, https://doi.org/10.1007/978-3-642-70627-1_16, 804 1986. 805

Selin, N. E. and Jacob, D. J.: Seasonal and Spatial Patterns of Mercury Wet Deposition in the United 806 States: Constraints on the Contribution from North American Anthropogenic Sources, Atmos. Environ., 807 42, 5193–5204, https://doi.org/10.1016/j.atmosenv.2008.02.069, 2008. 808

Selin, N. E., Jacob, D. J., Park, R. J., Yantosca, R. M., Strode, S., Jaeglé, L., and Jaffe, D.: Chemical 809 Cycling and Deposition of Atmospheric Mercury: Global Constraints from Observations, J. Geophys. 810 Res. Atmos., 112, https://doi.org/10.1029/2006JD007450, 2007. 811

Shah, V. and Jaeglé, L.: Subtropical Subsidence and Surface Deposition of Oxidized Mercury Produced 812 in the Free Troposphere, Atmos. Chem. Phys., 17, 8999–9017, https://doi.org/10.5194/acp-17-8999-2017, 813 2017. 814

Shah, V., Jaeglé, L., Gratz, L. E., Ambrose, J. L., Jaffe, D. A., Selin, N. E., Song, S., Campos, T. L., 815 Flocke, F. M., Reeves, M., Stechman, D., Stell, M., Festa, J., Stutz, J., Weinheimer, A. J., Knapp, D. J., 816 Montzka, D. D., Tyndall, G. S., Apel, E. C., Hornbrook, R. S., Hills, A. J., Riemer, D. D., Blake, N. J., 817 Cantrell, C. A., and Mauldin III, R. L.: Origin of Oxidized Mercury in the Summertime Free Troposphere 818 over the Southeastern US, Atmos. Chem. Phys., 16, 1511–1530, https://doi.org/10.5194/acp-16-1511-819 2016, 2016. 820

Shepler, B. C. and Peterson, K. A.: Mercury Monoxide:  A Systematic Investigation of Its Ground 821 Electronic State, J. Phys. Chem. A, 107, 1783–1787, https://doi.org/10.1021/jp027512f, 2003. 822

Shepler, B. C., Balabanov, N. B., and Peterson, K. A.: Hg+Br→HgBr Recombination and Collision-823 Induced Dissociation Dynamics, J. Chem. Phys., 127, 164304, https://doi.org/10.1063/1.2777142, 2007. 824

Si, L. and Ariya, P. A.: Reduction of Oxidized Mercury Species by Dicarboxylic Acids (C2−C4): Kinetic 825 and Product Studies, Environ. Sci. Technol., 42, 5150–5155, https://doi.org/10.1021/es800552z, 2008. 826

Slemr, F., Weigelt, A., Ebinghaus, R., Bieser, J., Brenninkmeijer, C. A. M., Rauthe-Schöch, A., Hermann, 827 M., Martinsson, B. G., van Velthoven, P., Bönisch, H., Neumaier, M., Zahn, A., and Ziereis, H.: Mercury 828 Distribution in the Upper Troposphere and Lowermost Stratosphere According to Measurements by the 829 IAGOS-CARIBIC Observatory: 2014–2016, Atmos. Chem. Phys., 18, 12329–12343, 830 https://doi.org/10.5194/acp-18-12329-2018, 2018. 831

38

Sommar, J., Gårdfeldt, K., Strömberg, D., and Feng, X.: A Kinetic Study of the Gas-phase Reaction 832 Between the Hydroxyl Radical and Atomic Mercury, Atmos. Environ., 35, 3049–3054, 833 https://doi.org/10.1016/S1352-2310(01)00108-X, 2001. 834

Sprovieri, F., Pirrone, N., Bencardino, M., D&amp;apos;Amore, F., Angot, H., Barbante, C., Brunke, E.-835 G., Arcega-Cabrera, F., Cairns, W., Comero, S., Diéguez, M. del C., Dommergue, A., Ebinghaus, R., 836 Feng, X. B., Fu, X., Garcia, P. E., Gawlik, B. M., Hageström, U., Hansson, K., Horvat, M., Kotnik, J., 837 Labuschagne, C., Magand, O., Martin, L., Mashyanov, N., Mkololo, T., Munthe, J., Obolkin, V., Ramirez 838 Islas, M., Sena, F., Somerset, V., Spandow, P., Vardè, M., Walters, C., Wängberg, I., Weigelt, A., Yang, 839 X., and Zhang, H.: Five-year Records of Mercury Wet Deposition Flux at GMOS Sites in the Northern 840 and Southern Hemispheres, Atmos. Chem. Phys., 17, 2689–2708, https://doi.org/10.5194/acp-17-2689-841 2017, 2017. 842

Stanevich, I., Jones, D. B. A., Strong, K., Parker, R. J., Boesch, H., Wunch, D., Notholt, J., Petri, C., 843 Warneke, T., Sussmann, R., Schneider, M., Hase, F., Kivi, R., Deutscher, N. M., Velazco, V. A., Walker, 844 K. A., and Deng, F.: Characterizing Model Errors in Chemical Transport Modeling of Methane: Impact of 845 Model Resolution in Versions v9-02 of GEOS-Chem and v35j of its Adjoint model, Geosci. Model Dev., 846 13, 3839–3862, https://doi.org/10.5194/gmd-13-3839-2020, 2020. 847

Streets, D. G., Horowitz, H. M., Lu, Z., Levin, L., Thackray, C. P., and Sunderland, E. M.: Five Hundred 848 Years of Anthropogenic Mercury: Spatial and Temporal Release Profiles, Environ. Res. Lett., 14, 084004, 849 https://doi.org/10.1088/1748-9326/ab281f, 2019a. 850

Streets, D. G., Horowitz, H. M., Lu, Z., Levin, L., Thackray, C. P., and Sunderland, E. M.: Global and 851 Regional Trends in Mercury Emissions and Concentrations, 2010–2015, Atmos. Environ., 201, 417–427, 852 https://doi.org/10.1016/j.atmosenv.2018.12.031, 2019b. 853

Strömberg, D., Gropen, O., and Wahlgren, U.: Non-relativistic and Relativistic Calculations on Some Zn, 854 Cd and Hg Complexes, Chem. Phys., 133, 207–219, https://doi.org/10.1016/0301-0104(89)80202-2, 855 1989. 856

Sumner, A. L., Spicer, C. W., Satola, J., Mangaraj, R., Cowen, K. A., Landis, M. S., Stevens, R. K., and 857 Atkeson, T. D.: Environmental Chamber Studies of Mercury Reactions in the Atmosphere, in: Dynamics 858 of Mercury Pollution on Regional and Global Scales, edited by: Pirrone, N. and Mahaffey, K. R., 859 Springer-Verlag, New York, 193–212, https://doi.org/10.1007/0-387-24494-8_9, 2005. 860

Sun, G., Sommar, J., Feng, X., Lin, C.-J., Ge, M., Wang, W., Yin, R., Fu, X., and Shang, L.: Mass-861 Dependent and -Independent Fractionation of Mercury Isotope during Gas-Phase Oxidation of Elemental 862 Mercury Vapor by Atomic Cl and Br, Environ. Sci. Technol., 50, 9232–9241, 863 https://doi.org/10.1021/acs.est.6b01668, 2016. 864

Sunderland, E. M. and Mason, R. P.: Human Impacts on Open Ocean Mercury Concentrations, Global 865 Biogeochem. Cycles, 21, GB4022, https://doi.org/10.1029/2006GB002876, 2007. 866

Talbot, R., Mao, H., Scheuer, E., Dibb, J., Avery, M., Browell, E., Sachse, G., Vay, S., Blake, D., Huey, 867 G., and Fuelberg, H.: Factors Influencing the Large-Scale Distribution of Hg° in the Mexico City Area 868 and over the North Pacific, Atmos. Chem. Phys., 8, 2103–2114, https://doi.org/10.5194/acp-8-2103-2008, 869 2008. 870

Tossell, J. A.: Calculation of the Energetics for Oxidation of Gas-Phase Elemental Hg by Br and BrO, J. 871 Phys. Chem. A, 107, 7804–7808, https://doi.org/10.1021/jp030390m, 2003. 872

39

Travnikov, O. and Ilyin, I.: The EMEP/MSC-E Mercury Modeling System, in: Mercury Fate and 873 Transport in the Global Atmosphere, edited by: Mason, R. and Pirrone, N., Springer US, Boston, MA, 874 571–587, https://doi.org/10.1007/978-0-387-93958-2_20, 2009. 875

Travnikov, O., Angot, H., Artaxo, P., Bencardino, M., Bieser, J., D&amp;apos;Amore, F., Dastoor, A., 876 De Simone, F., Diéguez, M. del C., Dommergue, A., Ebinghaus, R., Feng, X. B., Gencarelli, C. N., 877 Hedgecock, I. M., Magand, O., Martin, L., Matthias, V., Mashyanov, N., Pirrone, N., Ramachandran, R., 878 Read, K. A., Ryjkov, A., Selin, N. E., Sena, F., Song, S., Sprovieri, F., Wip, D., Wängberg, I., and Yang, 879 X.: Multi-model Study of Mercury Dispersion in the Atmosphere: Atmospheric Processes and Model 880 Evaluation, Atmos. Chem. Phys., 17, 5271–5295, https://doi.org/10.5194/acp-17-5271-2017, 2017. 881

Wang, F., Saiz-Lopez, A., Mahajan, A. S., Gómez Martín, J. C., Armstrong, D., Lemes, M., Hay, T., and 882 Prados-Roman, C.: Enhanced Production of Oxidised Mercury Over the Tropical Pacific Ocean: A Key 883 Missing Oxidation Pathway, Atmos. Chem. Phys., 14, 1323–1335, https://doi.org/10.5194/acp-14-1323-884 2014, 2014. 885

Wang, X., Jacob, D. J., Eastham, S. D., Sulprizio, M. P., Zhu, L., Chen, Q., Alexander, B., Sherwen, T., 886 Evans, M. J., Lee, B. H., Haskins, J. D., Lopez-Hilfiker, F. D., Thornton, J. A., Huey, G. L., and Liao, H.: 887 The Role of Chlorine in Global Tropospheric Chemistry, 19, 3981–4003, https://doi.org/10.5194/acp-19-888 3981-2019, 2019. 889

Wang, X., Jacob, D. J., Downs, W., Zhai, S., Zhu, L., Shah, V., Holmes, C. D., Sherwen, T., Alexander, 890 B., Evans, M. J., Neuman, J. A., Veres, P., Volkamer, R., Huey, L. G., Le Breton, M., Bannan, T. J., 891 Percival, C. J., Lee, B. H., and Thornton, J. A.: Global Tropospheric Halogen (Cl, Br, I) Chemistry and its 892 Impact on Oxidants (submitted), Atmos. Chem. Phys., 2021. 893

Whalin, L., Kim, E.-H., and Mason, R.: Factors Influencing the Oxidation, Reduction, Methylation and 894 Demethylation of Mercury Species in Coastal Waters, Marine Chemistry, 107, 278–294, 895 https://doi.org/10.1016/j.marchem.2007.04.002, 2007. 896

Wilcox, J.: A Kinetic Investigation of High-Temperature Mercury Oxidation by Chlorine, J. Phys. Chem. 897 A, 113, 6633–6639, https://doi.org/10.1021/jp901050d, 2009. 898

Wild, O. and Prather, M. J.: Excitation of the Primary Tropospheric Chemical Mode in a Global Three-899 Dimensional Model, J. Geophys. Res., 105, 24647–24660, https://doi.org/10.1029/2000JD900399, 2000. 900

Wu, R., Wang, C., and Dibble, T. S.: First Experimental Kinetic Study of the Atmospherically Important 901 Reaction of BrHg + NO2, Chem. Phys. Lett., 759, 137928, https://doi.org/10.1016/j.cplett.2020.137928, 902 2020. 903

Yang, X., Jiskra, M., and Sonke, J. E.: Experimental Rainwater Divalent Mercury Speciation and 904 Photoreduction Rates in the Presence of Halides and Organic Carbon, Sci. Total Environ., 697, 133821, 905 https://doi.org/10.1016/j.scitotenv.2019.133821, 2019. 906

Ye, Z., Mao, H., Driscoll, C. T., Wang, Y., Zhang, Y., and Jaeglé, L.: Evaluation of CMAQ Coupled 907 With a State‐of‐the‐Art Mercury Chemical Mechanism (CMAQ‐newHg‐Br), J. Adv. Model. Earth Syst., 908 10, 668–690, https://doi.org/10.1002/2017MS001161, 2018. 909

Zhang, Y., Jaeglé, L., van Donkelaar, A., Martin, R. V., Holmes, C. D., Amos, H. M., Wang, Q., Talbot, 910 R., Artz, R., Brooks, S., Luke, W., Holsen, T. M., Felton, D., Miller, E. K., Perry, K. D., Schmeltz, D., 911

40

Steffen, A., Tordon, R., Weiss-Penzias, P., and Zsolway, R.: Nested-grid Simulation of Mercury over 912 North America, Atmos. Chem. Phys., 12, 6095–6111, https://doi.org/10.5194/acp-12-6095-2012, 2012. 913

Zhu, L., Jacob, D. J., Eastham, S. D., Sulprizio, M. P., Wang, X., Sherwen, T., Evans, M. J., Chen, Q., 914 Alexander, B., Koenig, T. K., Volkamer, R., Huey, L. G., Le Breton, M., Bannan, T. J., and Percival, C. 915 J.: Effect of Sea Salt Aerosol on Tropospheric Bromine Chemistry, Atmos. Chem. Phys., 19, 6497–6507, 916 https://doi.org/10.5194/acp-19-6497-2019, 2019. 917

918

41

TOC art: 919 920

921 922 For Table of Contents only 923