A Theoretical Study of the Oxidation of Hg0 to HgBr2 in the

Environ. Sci. Technol. 2004, 38, 1772-1776

A Theoretical Study of the Oxidation of Hg0 to HgBr2 in the Troposphere M . E . G O O D S I T E , * ,† J. M. C. PLANE,‡ AND H. SKOV† Department of Atmospheric Environment, National Environmental Research Institute, Roskilde, Denmark, and School of Environmental Sciences, University of East Anglia, Norwich, United Kingdom

The oxidation of elemental mercury (Hg0) to the divalent gaseous mercury dibromide (HgBr2) has been proposed to account for the removal of Hg0 during depletion events in the springtime Arctic. The mechanism of this process is explored in this paper by theoretical calculations of the relevant rate coefficients. Rice-Ramsberger-KasselMarcus (RRKM) theory, together with ab initio quantum calculations where required, are used to estimate the following: recombination rate coefficients of Hg with Br, I, and O; the thermal dissociation rate coefficient of HgBr; and the recombination rate coefficients of HgBr with Br, I, OH, and O2. A mechanism based on the initial recombination of Hg with Br, followed by the addition of a second radical (Br, I, or OH) in competition with thermal dissociation of HgBr, is able to account for the observed rate of Hg0 removal, both in Arctic depletion events and at lower latitudes.

allowing it to be globally transported (5-7). Hg0 can thus be transported to the polar regions, where it is removed from the boundary layer during an AMDE with an e-folding lifetime of less than 10 h, being converted to an inorganic oxidized gaseous mercury compound HgXY (2, 3). This compound, which is commonly referred to as reactive gaseous mercury, RGM, is operationally determined in the Arctic. The common method for measuring RGM is by collection onto a KCl-coated annular denuder, followed by the pyrolytic reduction of the captured RGM to Hg0 (8). This technique results in information about the composition of the HgXY family being lost. A number of environmental conditions favorable for AMDEs at high latitudes have been identified. These include the following: a marine/maritime location; calm weather, low wind speeds, and nonturbulent airflow; the existence of a temperature inversion; sunlight; and sub-zero temperatures (9). These conditions are also favorable to the photochemically initiated heterogeneous production of halogen atoms (Br and Cl) and halogen oxide radicals (BrO and ClO), which are assumed to be involved in the mercury oxidative mechanism (9-11). BrO is produced in large quantities (>20 pptv) after polar sunrise in the marine boundary layer, through the so-called “bromine explosion” (4, 12, 13). Gasphase HOBr reacts with a Br- ion in the saline sea ice surface to yield Br2, which is then photolyzed. The resulting Br atoms react with O3 to form BrO, which in turn reacts with HO2 radicals to regenerate HOBr, and the cycle repeats. Several mechanisms have been proposed to explain the oxidation of gaseous elemental mercury to RGM (10, 11). These include the reactions between Hg0 and halogen oxides or halogen atoms to produce HgO, HgBr2, and HgCl2:

Introduction The perennial oxidation of mercury in the Arctic, which occurs simultaneously with the post solar sunrise destruction of ozone (1), potentially doubles the loading of mercury to the Arctic (2). These atmospheric mercury depletion episodes, AMDEs, were discovered in 1995 at Alert in the Canadian Arctic (1) and have since been observed at circum-Arctic locations, the Antarctic, and subpolar locations near sea ice (3 and citations therein). Tarasick and Bottenheim (4) have noted that the frequency of occurrence of boundary-layer ozone depletion episodes has increased since the 1960s, particularly at Resolute in the Canadian Arctic (the only site with a sufficiently long record for proper trend analysis). Those authors postulate that this increase could have resulted from an increase in open leads in the Arctic ice cover, possibly because of climate change induced by increasing levels of greenhouse gases. Furthermore, the increase in frequency of ozone depletion events may explain the increase in mercury levels observed in Arctic biota over the last few decades (4). The current knowledge of AMDEs is summarized in (3). It is clearly important to understand the detailed mechanism of mercury oxidation so that transport and deposition models can be properly parametrized. Mercury exists in the atmosphere primarily in gaseous elemental form, Hg0, which has an atmospheric residence time of approximately 1 year, * Corresponding author present address: University of Southern Denmark, Department of Chemistry, Campusvej 55, DK-5230 Odense M, Denmark. Phone: (45) 6550 2557; fax: (45) 6615 8780; e-mail: [email protected]. † National Environmental Research Institute. ‡ University of East Anglia. 1772

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 6, 2004

Br (Cl) + O3 f BrO (ClO) + O2

(1)

BrO (ClO) + Hg f HgO + Br (Cl)

(2)

Hg + Br (Cl) f HgBr (HgCl) ff HgBr2 (HgCl2) (3) and the possible role of oxidants such as OH, HO2, O(1D and and NO3 that are associated with high levels of NO resulting from photodenitrification processes in the snowpack (14). A recent review of the atmospheric chemistry of mercury can be found in (15). In this paper we will consider the following mechanism for producing HgBrY:

3P),

Hg + Br (+ M) f HgBr

(M ) third body)

HgBr (+ M) f Hg + Br HgBr + Y (+ M) f HgBrY

(4) (-4)

(Y ) Br, I, OH, O2 etc.) (5)

Ariya et al. (11) have shown recently that reaction 4, the recombination reaction between Hg and Br, is surprisingly fast, and so this is our prime candidate to initiate the oxidation of Hg0. However, the analogous reaction of atomic I may also be important, following observations of active iodine oxide chemistry in the mid- and low-latitude marine boundary layer (16, 17). Bauer et al. (18) have recently reported that the reaction Hg + OH is very slow, although two previous studies obtained rate coefficients for this reaction that varied by 2 orders of magnitude. We will therefore investigate the reactions of Hg with both I and OH in this paper. In contrast, the concentration of atomic Cl is extremely low ( i, Pij was calculated by a detailed balance. To simulate irreversible stabilization of HgBr via reaction 4.3, an absorbing boundary was set 24 kJ mol-1 below the energy of the reactants so that collisional energization from the boundary to the threshold was highly improbable. The rate of population of grain i, Ri, VOL. 38, NO. 6, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1773

is given by detailed balance between reactions 4.1 and 4.2:

Ri ) krec,∞[Hg][Br]ηi

(II)

where krec,∞ is the limiting high-pressure association rate coefficient (reaction 4.1) and

ηi )

k-4,ifi

∑

(III)

k-4,ifi

i

where fi is the equilibrium Boltzmann distribution of HgBr(Ei). The microcanonical rate coefficients for dissociation of HgBr were determined using inverse Laplace transformation (25), which links k-1(Ei) directly to krec,∞. In the present case, krec,∞ was expressed in the Arrhenius form A∞ exp(-E∞/RT). Assuming that collisions between Hg and Br are governed by the long-range attractive dispersion force, then A∞ ) 1.67 × 10-10 cm3 molecule-1 s-1 and E∞ ) -423 J mol-1. The microcanonical rate coefficient for dissociation is then given by

k-4,i )

A∞(2πµ)3/2 N(Ei)Γ(1.5)h3

∫

Ei - E∞ - ∆H0o

0

∞

Np(x)[(Ei - E ∆H0o) - x]0.5 dx (IV)

where the density of states of HgBr at energy Ei, N(Ei), was calculated using a combination of the Beyer-Swinehart algorithm for the vibrational modes (including a correction for anharmonicity) and a classical densities of states treatment for the rotational modes; Np(Ei) is the convoluted density of states of Hg and Br; ∆H0o is the Hg-Br bond energy; and µ is the reduced mass of Hg and Br. The ME was expressed in matrix form and then solved to yield k4, the bimolecular recombination rate constant at a specified pressure and temperature. The dissociation rate coefficient, k-4, was calculated by detailed balance with k4. Note that, for these calculations of k4 and k-4, the experimental parameters in Table 1 were employed. Figure 1 illustrates the calculated rate coefficients k4, k-4, and k5 at a pressure of 1 atm N2 over the temperature range 180-400 K. This shows that k4 and k5 have small negative temperature dependencies, as expected for recombination reactions. In contrast, k-4 has a large positive activation energy, approximately equal to the Hg-Br bond energy. The rate coefficients at 1 atm pressure are

k4(Hg + Br f HgBr, 180-400 K) )

1.1 × 10-12 (T/298 K)-2.37 cm3 molecule-1 s-1

k-4(HgBr f Hg + Br, 180-400 K) ) 1.2 × 1010 exp(-8357/T) s-1 k5(HgBr + Br f HgBr2, 180-400 K) )

2.5 × 10-10 (T/298 K)-0.57 cm3 molecule-1 s-1

Reaction 5 is close to the high-pressure limit at 1 atm. Inspection of Table 1 shows that atomic I and OH bond only slightly less strongly to HgBr, so the rate coefficients for these reactions are very similar to k5, essentially at their highpressure limits. Note that the products HgBr2, HgBrI, and HgBrOH (Table 1) are extremely stable against thermal dissociation at temperatures below 400 K. For the recombination reactions of Hg with I and OH, and the dissociation of HgI and HgOH, application of 1774

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 6, 2004

FIGURE 1. Rate coefficients calculated using RRKM theory, plotted as a function of temperature in Kelvin (T/K) for the recombination of Hg with Br, I, and OH and of HgBr with Br (solid lines, left-hand ordinate); and for the thermal dissociation of HgBr, HgI, and HgOH (broken lines, right-hand ordinate). RRKM theory using the data in Table 1 yields (pressure ) 1 atm N2)

k(Hg + I f HgI, 180-400 K) ) 4.0 × 10-13 (T/298 K)-2.38 cm3 molecule-1 s-1 k(Hg + OH f HgOH, 180-400 K) ) 3.2 × 10-13 (T/298 K)-3.06 cm3 molecule-1 s-1 k(HgI f Hg + I, 180-400 K) ) 3.0 × 109 exp(-3742/T) s-1 k(HgOH f Hg + OH, 180-400 K) ) 2.7 × 109 exp(-4061/T) s-1 The temperature dependencies of these four reactions are also illustrated in Figure 1.

Discussion Figure 1 demonstrates several important points with respect to the oxidation of Hg. First, the recombination of Hg with Br or I is surprisingly fast for an atom-atom recombination. The reason is the high density of rovibrational states arising from the low vibrational frequency and small rotational constant of HgBr or HgI (Table 1). Interestingly, for HgBr the theoretical estimate of k4 is about a factor of 3 lower than the recent experimental measurement (11). In fact, we can only match the experimental value if the bond energy of HgBr is increased to over 100 kJ mol-1, about 30 kJ mol-1 higher than the current experimental measurement of 74.9 ( 4 kJ mol-1 (22, 29). However, the recent experimental estimate of k4 was a relative rate measurement that required several significant correction factors (11). Therefore, for the present calculations, we prefer the theoretical estimate of k4 since the vibrational frequency and rotational constant of HgBr used in the present application of RRKM theory are known precisely from laser-induced fluorescence spectroscopy in a supersonic jet (22). In any case, we will show below that the crucial factor that determines the lifetime to convert Hg0 to HgII is the rate at which HgBr decomposes (reaction -4). The rate constant, k-4, is obtained by detailed balance with k4 so that the lifetime of Hg0 is not very dependent on the choice of k4. In the case of reaction 5, the addition of the second bromine to HgBr is predicted to be a very fast reaction, proceeding close to the high-pressure limit (essentially the collision number) at atmospheric pressure. The second point that emerges from Figure 1 is that the recombination reactions of Hg with I and OH are a factor of

FIGURE 2. Contour plot of the lifetime in hours for Hg0 oxidation to HgBr2, plotted as a function of [Br] in parts per trillion and temperature in Kelvin. 3-4 times slower than reaction 4. The principal reason is the smaller binding energies of HgI and HgOH. It should be noted that there is a very large discrepancy in the literature regarding k(Hg + OH), with estimates ranging from 8.7 × 10-14 to 1.6 × 10-11 cm3 molecule-1 s-1 at close to 300 K (18). The current theoretical calculations are in good agreement with the most recent upper limit of 1.2 × 10-13 cm3 molecule-1 s-1 (18). Note, however, that the lifetime of HgOH is only 280 µs at 298 K so that a true kinetic measurement of the recombination reaction would be difficult to achieve in practice. The third point demonstrated in Figure 1 is that the thermal dissociation of HgBr is more than 106 times slower than the thermal dissociation of HgI or HgOH, at temperatures below 300 K. This enormous difference arises from the stronger Hg-Br bond. The dissociation lifetimes of HgI and HgOH are less than 1 s at temperatures above 200 K. Hence, these species will not play a significant role in Hg0 removal. We therefore conclude that Hg0 is oxidized to HgII by recombination with Br. There is then a competition between further addition of Br to form HgBr2, or thermal decomposition of HgBr. The addition of I to HgBr may also be significant in some marine locations; however, the OH concentration in the clean marine boundary layer (typically less than 106 cm-3) is probably too low for OH addition to HgBr to be significant. As shown in Table 1, the addition of O2 to form HgBrO2 will not be an important process because this peroxy radical is so weakly bound that it will dissociate rapidly even at Arctic temperatures. The lifetime of Hg0, against conversion to HgBr2, is then given by

τ ) (k-4 + k5[Br])/k4k5[Br]2

(V)

Note that τ is strictly speaking an upper limit since other processes, yet to be identified, may also remove Hg0. Figure 2 illustrates τ as a function of [Br] and temperature. During springtime in the Arctic, the temperature ranges from about 230 to 260 K. During Hg depletion events [Br] is estimated to vary from 0.2 ppt, when τ will range from 35 to 60 h, to 6 ppt, when τ will be only 0.7-1.5 h (30). A typically observed 10-h lifetime of Hg (2, 10) would correspond to [Br] ) 0.7 ppt at an “average” temperature of 245 K.

At temperatures above 280 K, k2 becomes very fast and so τ increases significantly. In the mid-latitude marine boundary layer, where the concentration of BrO has recently been measured to be around 2 ppt during daytime [A. SaizLopez and J. M. C. Plane, University of East Anglia, personal communication], the atomic Br concentration under photochemical steady state will be e0.1 ppt. This is similar to the atomic I (16) and OH concentrations so that these radicals may also play a role at mid-latitudes, in contrast to the Arctic. Nevertheless, Figure 2 shows that under these conditions τ increases to >4000 h given eq V and its assumptions. This is in sensible accord with the observed global lifetime of more than 1 year (5-7), bearing in mind that HgBr has a rich UV/visible spectroscopy (31) and may therefore have a significant photodissociation rate in the troposphere, which would extend τ even more. Finally, we reiterate that if the measured value of k4 (11) is used in place of the present theoretical estimate (and k-4 is estimated by detailed balance to maintain the equilibrium between HgBr and Hg + Br), then the lifetimes calculated by eq V are little changed. In conclusion, we have shown that a mechanism based on the initial recombination of Hg with Br, followed by addition of a second radical in competition with thermal dissociation, is able to account for the observed rate of Hg0 removal, both in Arctic depletion events and on a global scale.

Acknowledgments The authors wish to thank the Danish Cooperation for Environment in the Arctic (DANCEA), the Danish Research Agency (SNF), and the U.K. National Environmental Research Council (NERC) for support. Michael Goodsite was supported by a COGCI graduate research studentship from the Danish Research Agency and the Department of Atmospheric Environment, NERI-DK.

Literature Cited (1) Schroeder, W. H.; Anlauf, K. G.; Barrie, L. A.; Lu, J. Y.; Steffen, A.; Schneeberger, D. R.; Berg, T. Nature 1998, 394, 331-332. (2) Skov, H.; Christensen, J.; Goodsite, M. E.; Heidam, N. Z.; Jensen, B.; Wåhlin, P.; Geernaert, G. Environ Sci. Technol., submitted for publication. (3) Schroeder, W. H.; Steffen, A.; Scott, K.; Bender, T.; Prestbo, E.; Ebinghaus, R.; Lu, J. Y.; Lindberg, S. E. Atmos. Environ. 2003, 37, 2551-2555. (4) Tarasick, D. W.; Bottenheim, J. W. Atmos. Chem. Phys. 2002, 2, 197-205. (5) Slemr, F.; Schuster, G.; Seiler, W. J. Atmos. Chem. 1985, 3, 407434. (6) Schroeder, W. H.; Jackson, R. A. Chemosphere 1987, 16, 183199. (7) Lamborg, C. H.; Fitzgerald, W.; O’Donnell, J.; Torgersen, T. Geochim. Cosmochim. Acta 2002, 66 (7), 1105-1118. (8) Landis, M. S.; Stevens, R. K.; Schaedlich, F.; Prestbo, E. M. Environ. Sci. Technol. 2002, 36, 3000-3009. (9) Lu, J. Y.; Schroeder, W. H.; Barrie, L. A.; Steffen, A.; Welch, H. E.; Martin, K.; Lockhart, W. L.; Hunt, R. V.; Boila, G.; Richter, A. Geophys. Res. Lett. 2001, 28, 3219-3222. (10) Lindberg, S. E.; Brooks, S.; Lin, C.-J.; Scott, K. J.; Landis, M. S.; Stevens, R. K.; Goodsite, M.; Richter A. Environ. Sci. Technol. 2002, 36, 1245-1256. (11) Ariya, P. A.; Khalizov, A.; Gidas, A. J. Phys. Chem. A 2002, 106, 7310-7320. (12) Barrie, L. A.; Platt, U. Tellus, Ser. B 1997, 49, 450-454. (13) Foster, K. L.; Plastridge, R. A.; Bottenheim, J. W.; Shepson, P. B.; Finlayson-Pitts, B. J.; Spicer, C. W. Science 2001, 291, 471-474. (14) Temme, C.; Einax, J. W.; Ebinghaus, R. Environ. Sci. Technol, 2003, 37 (1), 22-31. (15) Lin, C.-J.; Pehkonen, S. O. Atmos. Environ. 1999, 33, 20672079. (16) McFiggans, G.; Plane, J. M. C.; Allan, B. J.; Carpenter, L. J.; Coe, H.; O’Dowd, C. J. Geophys. Res. 2000, 105, 14371-14385. (17) Allan, B. J.; Plane, J. M. C.; McFiggans, G. Geophys. Res. Lett. 2001, 28, 1945-1948. VOL. 38, NO. 6, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1775

(18) Bauer, D.; D’Ottone, L.; Campuzaon-Jost, P.; Hynes, A. J. J. Photochem. Photobiol., A 2003, 157, 247-256. (19) McFiggans, G.; Cox, R. A.; Mossinger, J. C.; Allan, B. J.; Plane J. M. C. J. Geophys. Res., [Atmos.] 2002, 107, 4271-4280. (20) Steinfeld, J. I.; Francisco, J. S.; Hase, W. L. Chemical Kinetics and Dynamics; Prentice Hall: Englewood Cliffs, NJ, 1989. (21) Lipson, R. H.; Jordan, K. J.; Bascal, H. A. J. Chem. Phys. 1992, 98, 959-967. (22) Jordan, K. J.; Bascal, H. A.; Lipson, R. H.; Melchior, M. J. Mol. Spectrosc. 1993, 159, 144-155. (23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98, Revision A.7; Gaussian, Inc.: Pittsburgh, PA, 1998.

1776

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 6, 2004

(24) Cundari, T. R.; Stevens, W. J. J. Chem. Phys. 1993, 98, 55555565. (25) De Avillez Pereira, R.; Baulch, D. L.; Pilling, M. J.; Robertson, S. H.; Zeng, G. J. Phys. Chem. 1997, 101, 9681-9690. (26) Rollason, R. J.; Plane, J. M. C. Phys. Chem. Chem. Phys. 2000, 2, 2335-2343. (27) Self-D. E.; Plane, J. M. C. Phys. Chem. Chem. Phys. 2003, 5, 1407-1418. (28) Gilbert, R. G.; Smith, S. C. Theory of Unimolecular and Recombination Reactions; Blackwell: Oxford, 1990. (29) Handbook of Physics and Chemistry, 78th ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 1997. (30) Tuckermann, M.; Ackermann, R.; Go¨lz, C.; Lorenzen-Schmidt, H.; Senne, T.; Stutz, J.; Trost, B.; Unold, W.; Platt, U. Tellus, Ser. B 1997, 49, 533-555. (31) NIST Chemistry Web book, http://webbook.nist.gov/. (32) Tossell, J. A. J. Phys. Chem. A 2003, 107 (39), 7804-7808.

Received for review June 30, 2003. Revised manuscript received January 6, 2004. Accepted January 7, 2004. ES034680S