Observations of cyanogen bromide (BrCN) in the global troposphere and their relation to polar surface O 3 destruction

Bromine activation (the production of Br in an elevated oxidation state) promotes ozone destruction and mercury removal in the global troposphere and commonly occurs in both springtime polar boundary layers, often accompanied by nearly complete ozone destruction. The chemistry and budget of active bromine compounds (e.g., Br 2 , BrCl , BrO , HOBr ) reflect the cycling of Br and affect its environmental impact. Cyanogen bromide ( BrCN ) has recently been measured by iodide ion high-resolution time-of-flight mass spectrometry ( I - CIMS), and trifluoro methoxide ion time-of-flight mass spectrometry ( CF(3)o---CIMS) during the NASA Atmospheric Tomography Mission second, third, and fourth deployments (NASA ATom), and could be a previously unquantified participant in active Br chemistry. BrCN mixing ratios ranged from below the detection limit (1.5 pptv ) up to as high as 36 pptv (10 s average) and enhancements were almost exclusively confined to the polar boundary layers in the Arctic winter and in both polar regions during spring and fall. The coincidence of BrCN with active Br chemistry (often observable BrO , BrCl and O 3 loss) and high CHBr3 / CH2 Br-2 ratios imply that much of the observed BrCN is from atmospheric Br chemistry rather than a biogenic source. Likely BrCN formation pathways involve the heterogeneous reactions of active Br ( Br 2 , HOBr ) with reduced nitrogen compounds, for example hydrogen cyanide ( HCN /CN- ), on snow, ice, or particle surfaces. Competitive reaction calculations of HOBr reactions with Cl - / Br - and HCN / CN - in solution, as well as box model calculations with bromine chemistry, confirm the viability of this formation channel and show a distinct pH dependence, with BrCN formation favored at higher pH values. Gas-phase loss processes of BrCN due to reaction with radical species are likely quite slow and photolysis is known to be relatively slow ( BrCN lifetime of similar to 4 months in midlatitude summer). These features, and the lack of BrCN enhancements above the polar boundary layer, imply that surface reactions must be the major loss processes. The fate of BrCN determines whether BrCN production fuels or terminates bromine activation. BrCN reactions with other halogens ( Br - , HOCl , HOBr ) may perpetuate the active Br cycle; however, preliminary laboratory experiments showed that BrCN did not react with aqueous bromide ion ( < 0.1 %) to reform Br 2 . Liquid-phase reactions of BrCN are more likely to convert Br to bromide ( Br - ) or form a C - Br bonded organic species, as these are the known condensed-phase reactions of BrCN and would therefore constitute a loss of atmospheric active Br. Thus, further study of the chemistry of BrCN will be important for diagnosing polar Br cycling.