Atmospheric hydroperoxides in West Antarctica links to stratospheric ozone and atmospheric oxidation capacity

Frey, et al Atmospheric ROOH in West Antarctica Atmospheric hydroperoxides in West Antarctica: links to stratospheric ozone and atmospheric oxidation capacity Markus M Freya,d, Richard W Stewartb, Joseph R McConnellc, Roger C Balesd a Department of Hydrology and Water Resources, University of Arizona, 1133 E North Campus Drive, Tucson AZ 85721, USA b Atmospheric Chemistry and Dynamics Branch, National Aeronautics and Space Administration (NASA) Goddard Space Flight Center (GSFC), Greenbelt, Maryland, USA c Desert Research Institute, Divison of Hydrologic Sciences, 2215 Raggio Parkway, Reno, NV 89512, USA d University of California, Merced, 4225 N Hospital Road, Bldg 1200, Atwater, CA 95301, USA Corresponding author: Markus M Frey University of California, Merced 4225 N Hospital Rd., Bldg 1200 Atwater, CA-95301 mfrey@ucmerced.edu, phone: 209-205 8565, fax: 209-724 4459 Frey, et al Atmospheric ROOH in West Antarctica Abstract The troposphere above the West Antarctic Ice Sheet (WAIS) was sampled for hydroperoxides in 2000 to 2002 at 21 locations on 2-month-long summer traverses, as part of US ITASE (International Transantarctic Scientific Expedition) First time quantitative measurements using an HPLC method showed that methylhydroperoxide (MHP) is the only important organic hydroperoxide occurring in the Antarctic troposphere, and is found at levels ten times that predicted by photochemical models During three field seasons, means and standard deviations for hydrogen peroxide (H2O2) were 321±158 pptv, 650±176 pptv and 330±147 pptv MHP was detected, but not quantified in December 2000; levels in summer 2001 and 2002 were 317±128 pptv and 304±172.2 pptv Results from firn air experiments and diurnal variability of the two species showed that atmospheric H2O2 is significantly impacted by a physical snow pack source between 76 and 90 S, whereas MHP is not We show strong evidence of a positive feedback between stratospheric ozone and H2O2 at the surface Between November-27 and December-12 in 2001, when ozone column densities dropped below 220 DU (means in 2000 and 2001 were 318 DU and 334 DU), H2O2 was 1.7 times that observed in the same period in 2000 and 2002, while MHP was only 80% of the levels encountered in 2002 Photochemical box model runs suggest that NO and OH levels on WAIS are closer to coastal values, while Antarctic Plateau levels are higher, confirming that region to be a highly oxidizing environment The modeled sensitivity of H2O2 and particularly MHP to NO offers the potential to use atmospheric hydroperoxides to constrain the NO background and thus estimate the past oxidation capacity of the remote atmosphere Index Terms: 0365 Atmospheric Composition and Structure: Troposphere: composition and chemistry; 0322 Atmospheric Composition and Structure: Constituent sources and sinks; 1610 Global Change: Atmosphere (0315, 0325); 0736 Cryosphere: Snow (1827, 1863); 0724 Cryosphere: Ice Cores (4932) Keywords: hydrogen peroxide, methylhydroperoxide, Antarctica, air-snow exchange, stratospheric ozone, atmospheric oxidation capacity Frey, et al Atmospheric ROOH in West Antarctica Introduction Atmospheric photooxidants are responsible for the removal of atmospheric pollutants including carbon monoxide (CO), methane (CH 4), nitric and nitrogen oxides (NO, NO 2), dimethyl sulfide (DMS), sulfur dioxide (SO2) and halogenated compounds and thus control particle formation, the buildup of greenhouse gases and ultimately climate change Atmospheric hydroperoxides (ROOH) contribute to the oxidizing power of the troposphere [Lee, et al., 2000], defined as the total burden of ozone (O 3), HOx radicals and hydrogen peroxide (H 2O2), and also constitute an important radical reservoir There is increasing evidence that snow packs in polar regions influence the composition of the overlying atmosphere by uptake and release of photochemically active species such as NOx, organic acids, formaldehyde (HCHO) [Dominé and Shepson, 2002], including also H2O2 [Hutterli, et al., 2004; Hutterli, et al., 2001] Direct measurements of elevated HOx levels in the air above South Pole [Mauldin, et al., 2001; Mauldin, et al., 2004] and observed changes in ground levels of O in relation to the stratospheric O3 depletion in spring time [Jones and Wolff, 2003] show that snow-atmosphere interactions can alter the budget of atmospheric oxidants in the tropospheric boundary layer In addition, ice core records of the highly water soluble H2O2 [Anklin and Bales, 1997; Sigg and Neftel, 1991] offer the potential to reconstruct past changes in the oxidation capacity of the atmosphere if the processes controlling deposition and long-term preservation are quantitatively understood [McConnell, et al., 1997a] In the presence of sunlight, water vapor and O3, the short-lived OH radical is produced and converted into peroxyradicals through the oxidation of CO, CH or other non-methane hydrocarbons (R1 and R2): CO + OH + O2  CO2 + HO2 (R1) CH4 + OH + O2  H2O + CH3O2 (R2) The main photolytic source of ROOH is the combination of peroxyradicals (R3 and R4): HO2 + HO2  H2O2 + O2 (R3) RO2 + HO2  ROOH + O2 (R4) Frey, et al Atmospheric ROOH in West Antarctica However, competing reactions R5 and R6 with NO, which represent also the core of photochemical O3 production in the troposphere, can suppress the formation of hydroperoxides depending on the level of NO present [Kleinmann, 1991; Stewart, 1995]: NO + HO2  NO2 + OH (R5) NO + CH3OO (+ O2)  NO2 + HCHO + HO2 (R6) Observations from South Pole during the polar day show indeed that H 2O2 increases, at first linearly with increasing NO, and then starts to decrease once NO levels exceed 100 pptv [Hutterli, et al., 2004] With methane as the only significant organic peroxy radical precursor at remote sites, methylhydroperoxide (MHP, CH3OOH) is expected to be the dominant organic peroxide occurring in the gas phase above Antarctica Sink reactions for hydroperoxides include photolysis (R7 and R8) and attack by the OH radical (R9 and R10), with the former regenerating radicals from the temporary reservoir H2O2 + hν  OH (R7), (λ < 355 nm) CH3OOH + hν (+ O2)  HCHO + OH + HO2 (R8), (λ < 360 nm) H2O2 + OH  H2O + HO2 (R9) CH3OOH + OH  CH3OO + H2O (R10) Wet deposition affects the highly water soluble H 2O2 and less so higher organic peroxides This follows from the fact that the Henry’s Law constant for MHP is only 0.1% of that for H 2O2 [Lind and Kok, 1994] Dry deposition of ROOH above snow and ice surfaces can also impact atmospheric levels, as model studies at South Pole showed [Hutterli, et al., 2004] It has to be noted though that deposition is reversible for these gases and changes in temperature of the surface snowpack drive a cycle of emission and uptake until the top layer is buried by additional accumulation and disconnected from further exchange [Hutterli, et al., 2001; Wolff and Bales, 1996] Deposition of H2O2 to ice sheets thus provides the potential to reconstruct or at least constrain past levels of atmospheric radicals through ice core analysis Current photochemistry of hydroperoxides in the remote atmosphere has been little studied so far above Antarctica The only data available are summer time observations of H 2O2 at South Pole [Hutterli, et al., 2004; McConnell, et al., 1997b], spot measurements from a traverse Frey, et al Atmospheric ROOH in West Antarctica in Dronning Maud Land between 73 and 76S [Fuhrer, et al., 1996] and observations over various seasons at the coastal Neumayer Station, including a year-round data set of H 2O2 and methylhydroperoxide [Jacob and Klockow, 1993; Riedel, et al., 2000] (see Table for overview) The main goal of our study was to understand factors controlling the photochemistry of hydroperoxides in the background atmosphere above Antarctica, away from anthropogenic and biogenic emission sources This is motivated by two particular reasons: one, hydroperoxides are cited in the current literature to be potential ‘diagnostic tools’ of atmospheric oxidation capacity (e.g Riedel, et al., 2000 and, Lee, et al., 2000), however the relationship between hydroperoxides and ambient HOx radical levels has not yet been quantified in the polar regions And second, only a quantitative understanding of current relationships between ROOH, oxidant levels, solar radiation and climate, will allow interpreting past variability in ROOH reconstructed from ice cores in terms of atmospheric change To this we need to understand the influence of snowpack sources and enhanced UV from stratospheric ozone depletion on atmospheric hydroperoxide concentrations Developing this understanding across a wide gradient of temperature, accumulation rate, latitude and elevation enables obtaining a consistent and more sound explanation for the variability of hydroperoxides above an ice sheet and its relation to atmospheric radicals Methods We report atmospheric concentrations of atmospheric hydroperoxides in ambient and firn interstitial air measured above the West Antarctic Ice Sheet (WAIS) in the vicinity of past and future deep ice core drill sites The measurements from a ground traverse are unique in the sense that for the first time a data set of atmospheric observations across latitudes, similar to measurement campaigns on oceanic vessels, was obtained, covering much of WAIS Sinks and sources of ROOH were investigated, including radiative conditions, atmospheric properties and snow-atmosphere exchange The suitability of hydroperoxides as a constraint for the current atmospheric HOx-NOx budget was evaluated using a photochemical box model 2.1 Measurement Sites Measurements were carried out in summer beginning in 2000 to 2002 at 21 locations on 2-month-long traverses (Figure 1; Table 1), as part of the US component of the International Transantarctic Scientific Expedition (ITASE) The traverse routes connected WAIS regions low in Frey, et al Atmospheric ROOH in West Antarctica elevation and closer to the coast with the cold and dry East Antarctic Plateau (76 - 90 S latitude, 1232 - 2810 m elevation and –11.8 to –28.1 ºC mean on-site air temperature) Each site was occupied typically for 1-4 days and atmospheric experiments were carried out 500 m upwind from the main camp in a Scott Polar tent (summer 2000) and using a heated Weatherhaven shelter mounted onto a sledge as a mobile laboratory (summer 2001-2002) 2.2 Gas phase measurements Mixing ratios of atmospheric hydroperoxides were monitored continuously using a 2channel method, where the total concentration of peroxides was determined in channel [e.g Jacobi, et al., 2002], while in channel 2, the amount and speciation of individual peroxides were measured (Figure 2) For each channel ambient air was drawn through a separate PFA (¼” I.D.) intake line mounted at ~1 m above the snow surface Sample intake lines were insulated and heated above ambient temperature to prevent line loss due to condensation Flow rates of typically 1.4 STP-L min-1 were regulated by Aalborg mass flow controllers within 0.05 STP-L min-1 and recorded by a data logger Peroxides were scrubbed from the air stream into aqueous solution in temperature controlled glass coils of 55 cm length, mm I.D and 17 mm O.D Coils were mounted with silicone heat sink paste onto an aluminum cylinder, insulated and kept at 5±0.1 C using Peltier elements and temperature controllers (CNI3243, Omega International Inc.) Note that due to the gradual improvement of the detector set up in between seasons, no temperature control was available in 2000, only for channel in 2001 and for both channels in 2002 In channel total peroxides were then analyzed by reaction with 4-ethylphenol as a hydrogen donor and subsequent detection of the fluorescence signal at pH >11 Every 30 minutes peroxide free air was pumped for 10 minutes through the intake lines to monitor the baseline Baseline air was generated by passing ambient air through a column filled with manganese dioxide-copper oxide mixture (Hopcalite, Callery Chemical Company, USA) and then pumped through a separate line from inside the instrument out to a tee-junction at the beginning of the sample intake Discrete samples were injected into a High Pressure Liquid Chromatography (HPLC) system (channel 2) with a 6-port HPLC injection valve (model C2 with PAEK body, Valco Instruments Co Inc.) using a 2-position standard electric actuator (Valco Instruments Co Inc.) (Figure 2) Automatic injection of a 912 l sample every 10 minutes was triggered by a programmed CR10X data logger (Campbell Scientific Inc.) Separation of H 2O2 and organic Frey, et al Atmospheric ROOH in West Antarctica hydroperoxides was achieved on a INERTSIL ODS-2 column (metal free PEEK lining, Alltech Associates, Inc.) using 10-3 M H2SO4 solution with EDTA added as the mobile phase as described previously [Kok, et al., 1995; Lee, et al., 1995] Post-column chemistry and detection was the same as in channel Both channels had separate detectors but a common excitation source to assure covariation of the signal The electrical output from both channels was monitored at a rate of Hz with 10 s averages recorded on a laptop computer Temporal resolution of discrete sampling in channel was constrained by flow rates and sample loop volume and was typically 4-6 chromatograms per hour, each one of them representing a ~5 average of sampled air On the other hand resolution of continuous sampling in channel was limited by dispersion and estimated to be better than minutes The instrument response was calibrated in the field 1-2 times per day injecting liquid standards of H2O2 and MHP While H2O2 is commercially available, CH3OOH standards needed to be synthesized in the lab prior to each field season based on published procedures [Lee, et al., 2000; Rieche and Hitz, 1929] Methyl hydroperoxide was synthesized from the reaction between dimethyl sulfate and hydrogen peroxide under strongly alkaline conditions 38 ml of 40% KOH was added slowly over 40 minutes to a mixture of 45 ml of 30% hydrogen peroxide, 75 ml H 2O and 30 g dimethyl sulfate Since the reaction takes place under release of heat, the mixture was stirred and cooled constantly in order to prevent disintegration of the reaction product The liquid was acidified and about half of it distilled at 98ºC yielding diluted solutions of methyl hydroperoxide with concentrations of 0.9-1.5 M (4-7%) Permanganate titration to check for the presence of hydrogen peroxide was negative suggesting that H 2O2 contamination of the synthesis product was less than 10-4 M MHP standards were diluted to 10 -1 M, filled into brown glass bottles with EDTA added and stored at 5ºC Pre and post field season calibration of H 2O2 and MHP standard stock solutions using permanganate and iodometric titration with FeCl as a catalyst respectively showed no significant concentration change The decrease in MHP concentration of the stock solution after >1 yr of storage was less than 4% Firn interstitial air was sampled at sites (01-5, 02-1, 02-2, 02-3, 02-4, 02-5) evenly distributed between 76ºS and 90ºS The sampling procedure involved periodically alternating the intake line between ambient and firn air every ~30 minutes over a total time of 4-6 hours In order to draw air from the firn interstitial pore space, a hole was pre-cored to about ~10 cm depth, the intake line inserted down to this depth and fixed in this position and finally snow was packed around the line to reduce mixing with ambient air The use of only one sample intake line had the advantage of not introducing a systematic error due to different sample intakes Frey, et al Atmospheric ROOH in West Antarctica 2.3 Data processing Channel output, which contained 4-6 data points per hour, was linearly interpolated to original sample times of channel (1 value per 10s) and was used to correct channel for contributions of organic peroxides to the total peroxide signal After linear baseline correction the signal was converted into liquid concentration based on the current standard response Chromatogram peaks were evaluated using height rather than area as the parameter related to concentration Gas phase mixing ratios and measured aqueous phase concentrations of hydroperoxides are related by the ratio between liquid (Φl) and gas flow (Φg) rates and the collection efficiency CE of the glass coil scrubber: c g cl V H 2O  l MW  g CE (1) where MW is the molecular weight of the respective chemical species, ρH2O density of water, V the molar gas volume CE is defined as: CE   cg ' cg (2) with gas phase concentrations before (cg) and after the coil (cg’) Collection efficiencies for MHP determined in the lab at 5.8 ºC and 13.0 ºC were 0.86 and 0.75, while they were unity for H 2O2 at both temperatures Since the experiment compares well to calculations assuming Henry’s equilibrium, CE in equation (1) can be written in terms of measured quantities T0  l p0  g , STP  T l  KH pR p0  g , STP KH pR CE(T , p ) (3) and becomes a function of coil pressure p and coil temperature through the temperature dependence of the respective Henry’s Law constant KH [Lind and Kok, 1994] Coil temperatures were controlled within 0.1 K, however observed ambient pressure during the measurements Frey, et al Atmospheric ROOH in West Antarctica varied between 690 and 840 mbar throughout all ITASE seasons, which translates into CE variations of ~4%s Coil pressure, being a function of gas flow rates and intake line dimensions, was not monitored Instead a constant downward correction of ~140 mbar measured for the current experimental set up (M Hutterli, pers communication) was applied to recorded ambient pressures Liquid flow rates were corrected for evaporation occurring in the coil scrubbers with corrections being generally on the order of 5% The limit of detection (LOD), defined as times the baseline standard deviation, for H 2O2 from channel was 50 pptv during the first two seasons and 30 pptv during the last season The precision was usually better than 20 pptv MHP measurements from channel had a LOD of