Article pubs.acs.org/est
Atmospheric Peroxides in a Polluted Subtropical Environment: Seasonal Variation, Sources and Sinks, and Importance of Heterogeneous Processes Jia Guo,†,‡ Andreas Tilgner,‡ Chungpong Yeung,† Zhe Wang,† Peter K. K. Louie,§ Connie W. Y. Luk,§ Zheng Xu,∥ Chao Yuan,∥ Yuan Gao,† Steven Poon,† Hartmut Herrmann,‡ Shuncheng Lee,† Ka Se Lam,† and Tao Wang*,† †
Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, China Chemistry Department, Leibniz Institute for Tropospheric Research (TROPOS), 04318 Leipzig, Germany § Environmental Protection Department, the Government of the Hong Kong Special Administrative Region, Hong Kong, China ∥ Environment Research Institute of Shandong University, Shandong University, Jinan, Shandong 250100, China ‡
S Supporting Information *
ABSTRACT: Hydrogen peroxide (H2O2) and organic peroxides play an important role in atmospheric chemistry, but knowledge of their abundances, sources, and sinks from heterogeneous processes remains incomplete. Here we report the measurement results obtained in four seasons during 2011−2012 at a suburban site and a background site in Hong Kong. Organic peroxides were found to be more abundant than H2O2, which is in contrast to most previous observations. Model calculations with a multiphase chemical mechanism suggest important contributions from heterogeneous processes (primarily transition metal ion [TMI]-HOx reactions) to the H2O2 budget, accounting for about one-third and more than half of total production rate and loss rate, respectively. In comparison, they contribute much less to organic peroxides. The fast removal of H2O2 by these heterogeneous reactions explains the observed high organic peroxide fractions. Sensitivity analysis reveals that the role of heterogeneous processes depends on the abundance of soluble metals in aerosol, serving as a net H2O2 source at low metal concentrations, but as a net sink with high metal loading. The findings of this study suggest the need to consider the chemical processes in the aerosol aqueous phase when examining the chemical budget of gas-phase H2O2.
1. INTRODUCTION Hydrogen peroxide (H2O2) and organic peroxides are the products of atmospheric photo- and HOx chemistry and contribute to the oxidation capacity of the troposphere. Peroxides can be a source of HOx and ROx through photolysis and can also act as the terminal sinks for these radicals through dry and wet deposition. In addition to their importance in gasphase photochemical reactions, H2O2 and organic peroxide compounds such as methyl peroxides and peroxyacetic acid (PAA) are also capable of oxidizing SO2 into sulfate in the aqueous phase, thus contributing to the acidification of cloud droplets and aerosol particles.1,2 The photochemical sources of peroxides in the gas phase have been explored by numerous researchers, with the bimolecular combination of HO2 or RO2 being found to be the dominant source of peroxides in this phase.3−5 Peroxides can also be produced by the ozonolysis of alkenes, a production pathway that is independent of radiation and may be prominent in low photochemical conditions. When high levels of NO are present, peroxide formation is suppressed, as the peroxy © 2013 American Chemical Society
radicals react with NO to form NO2. The reaction between O3 and HO2 acts as a similar suppression mechanism. The H2O2 photolysis, reaction with OH radicals, and dry deposition are believed to be the main H2O2 sinks in the gas phase. In aqueous phase, although it was found some 20 years ago that transition metal ion (TMI) reactions were capable of producing and consuming H2O2,6−8 the dependence of the H2O2 budget on aerosol composition is still not well-known because of the complexity of the multiphase system and the wide variation in aerosol composition across sites.9,10 In a recent field study, a higher aerosol aqueous-phase H2O2 concentration than that predicted by gas-particle partitioning (Henry’s Law) was observed in California, suggesting net H2O2 formation in the aqueous phase.11,12 However, current modeling investigations Received: Revised: Accepted: Published: 1443
July 20, 2013 December 19, 2013 December 20, 2013 December 20, 2013 dx.doi.org/10.1021/es403229x | Environ. Sci. Technol. 2014, 48, 1443−1450
Environmental Science & Technology
Article
60%−85% for MHP and 100% for HMHP).17,19,20 Due to lack of information on organic peroxides speciation at our sites, we did not correct their concentration with any assumed efficiency, thus the organic peroxides concentration reported here represents a lower limit of their real atmospheric value. Other trace gases, aerosols compositions, and several meteorological parameters were also measured during the campaigns. They are used to constrain the model simulation, and details of the measuring instruments are described in the Supporting Information. 2.2. Modeling Procedures with RACM and CAPRAM. A zero-dimensional chemical model comprising the gas-phase regional atmospheric chemical mechanism (RACM scheme with 237 reactions)21 coupled to a chemical aqueous-phase radical mechanism (CAPRAM2.4 scheme with 438 reactions)10 was used to investigate the peroxide production and loss at the observation sites. The RACM simulates chemical reactions leading to production and loss of peroxides in gas phase, whereas CAPRAM investigates mass transfer between gas phase and deliquescent aerosols and chemical reactions in aerosol aqueous phase. The combined model is constrained with the measurements of gas-phase species O3, NO, NO2, HONO, SO2, and CO and 52 volatile organic compounds (VOCs, including formaldehyde, other aldehydes, and ketones), aerosol size distribution, aerosol major ions and trace elements, and meteorological parameters (temperature, relative humidity, boundary layer height). The detailed reaction list and recent developments concerning the CAPRAM mechanism can be found in the literature10,22 and at the CAPRAM home page (http://projects. tropos.de/capram). Phase transfer of chemical species between gas and aqueous phases is calculated in the model using the resistance model approach proposed by Schwartz et al.23 The aerosol radius used for evaluating the mass transfer rate was set as the corresponding radius of volume mean mass transfer constant (Kmt) calculated from all aerosol size bins (see more details in the Supporting Information). An Aerosol Inorganic Model (E-AIM model IV)24 was used to calculate the liquid water content (LWC), aqueous-phase ions concentrations (including pH), ionic strength, and the activity coefficients of major ions (i.e., H+, Na+, NH4+, Cl−, NO3−, SO42‑) with the input of field measured species concentrations and the ambient relative humidity. Under conditions of large ion strength such as for deliquescent aerosols, the activity coefficient should be taken into account when modeling aerosol aqueous chemistry.25 The CAPRAM2.4 mechanism has been further extended by considering the activity coefficients in chemical equilibrium and in ionic reactions. The activity coefficients for neutral species and trace metal ions are calculated by Pitzer ion interaction model26 with parameters from Christov27 and Millero et al.28 The Henry’s law constants of gaseous species were corrected by the activity coefficients for the salting out effect.25 Further details about the calculation method for the activity coefficients are provided in the Supporting Information. The concentrations of iron (Fe), manganese (Mn), and copper (Cu) used in the model were constrained with the semiobservational data estimated from the integrated filter concentrations and time profile of continuously measured inorganic ions in aerosol, by assuming a fixed fraction of transition metal concentration in total cationic concentration for episode days and clean days. In our study, the urban metal soluble fractions (Fe: 10%; Mn: 70%; Cu: 20%) derived from the metal solubility range in Deguillaume et al.29 were used for
rarely consider the aqueous interfacial processes in the atmospheric chemistry of H2O2. Hong Kong is a metropolis situated on the South China coast. Its coastal and subtropical environment favors peroxide production because of the abundant water vapor and strong solar radiation. However, the large amounts of NOx emissions from the numerous vehicles including ships in Hong Kong serve to suppress that production. Although many studies have investigated photochemical pollution in the region, to date there has been no investigation of the mixing ratio, sources, and sinks of peroxides and their role in the region’s atmospheric photochemistry. Hence, the present study was carried out to fill this research gap. This research constitutes the first observation of the seasonal behavior of atmospheric peroxides in China. An observation-based model (OBM) with a coupled gas-phase regional atmospheric chemical mechanism (RACM) and chemical aqueous-phase radical mechanism (CAPRAM2.4) was subsequently designed to investigate the gas- and aqueousphase contributions to the H2O2 budget. The overall objective of this work was to investigate the role of aerosol heterogeneous processes in determining the chemical budget of peroxides in humid environments.
2. EXPERIMENTAL AND MULTIPHASE BOX MODEL 2.1. Sites and Instrumentation. Measurements of H2O2, organic peroxides, and related species were taken at two sites in Hong Kong in 2011−2012: a regional background site at Hok Tsui (HT) and a suburban site at Tung Chung (TC). Locations of the two sites are shown in Figure S1 of the Supporting Information. The HT site is the Hong Kong Polytechnic University’s background air monitoring station, which is located at the southeastern tip of Hong Kong Island (22.217° N, 114.25°E, 60 m above sea level) on a cliff with a 270° view of the South China Sea. Most of the time, the station is upwind of the urban areas of Hong Kong and the Pearl River Delta (PRD). It has been involved in several long-term and intensive studies of photochemistry.13 In the current study, measurements were taken at the HT site in May, August, and December of 2011 and in February of 2012 (i.e., one month in each season). The TC site is an air-quality monitoring station operated by the Hong Kong Environmental Protection Department. It is located in the northern part of Lantau Island in western Hong Kong, an area known to suffer from serious photochemical pollution and poor visibility.14,15 The site is in a newly developed residential town adjacent to Hong Kong International Airport and several highways. In addition to the local emission sources, both clean background air and aged urban plumes can be sampled at the TC site, making it a suitable place for investigating the photochemical evolution of plumes from various urban areas in the region. The measurements at TC were taken in August and December of 2011 and February and May of 2012. H2O2 and organic peroxides were measured with an Aerolaser AL-2021 analyzer based on the enzyme-catalyzed fluorescence technique.16 The technique is sensitive to all peroxides in the solution, with two parallel channels used to distinguish between H2O2 and organic peroxides. The detection limit of the analyzer was
Atmospheric Peroxides in a Polluted Subtropical Environment: Seasonal Variation, Sources and Sinks, and Importance of Heterogeneous Processes Jia Guo,†,‡ Andreas Tilgner,‡ Chungpong Yeung,† Zhe Wang,† Peter K. K. Louie,§ Connie W. Y. Luk,§ Zheng Xu,∥ Chao Yuan,∥ Yuan Gao,† Steven Poon,† Hartmut Herrmann,‡ Shuncheng Lee,† Ka Se Lam,† and Tao Wang*,† †
Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, China Chemistry Department, Leibniz Institute for Tropospheric Research (TROPOS), 04318 Leipzig, Germany § Environmental Protection Department, the Government of the Hong Kong Special Administrative Region, Hong Kong, China ∥ Environment Research Institute of Shandong University, Shandong University, Jinan, Shandong 250100, China ‡
S Supporting Information *
ABSTRACT: Hydrogen peroxide (H2O2) and organic peroxides play an important role in atmospheric chemistry, but knowledge of their abundances, sources, and sinks from heterogeneous processes remains incomplete. Here we report the measurement results obtained in four seasons during 2011−2012 at a suburban site and a background site in Hong Kong. Organic peroxides were found to be more abundant than H2O2, which is in contrast to most previous observations. Model calculations with a multiphase chemical mechanism suggest important contributions from heterogeneous processes (primarily transition metal ion [TMI]-HOx reactions) to the H2O2 budget, accounting for about one-third and more than half of total production rate and loss rate, respectively. In comparison, they contribute much less to organic peroxides. The fast removal of H2O2 by these heterogeneous reactions explains the observed high organic peroxide fractions. Sensitivity analysis reveals that the role of heterogeneous processes depends on the abundance of soluble metals in aerosol, serving as a net H2O2 source at low metal concentrations, but as a net sink with high metal loading. The findings of this study suggest the need to consider the chemical processes in the aerosol aqueous phase when examining the chemical budget of gas-phase H2O2.
1. INTRODUCTION Hydrogen peroxide (H2O2) and organic peroxides are the products of atmospheric photo- and HOx chemistry and contribute to the oxidation capacity of the troposphere. Peroxides can be a source of HOx and ROx through photolysis and can also act as the terminal sinks for these radicals through dry and wet deposition. In addition to their importance in gasphase photochemical reactions, H2O2 and organic peroxide compounds such as methyl peroxides and peroxyacetic acid (PAA) are also capable of oxidizing SO2 into sulfate in the aqueous phase, thus contributing to the acidification of cloud droplets and aerosol particles.1,2 The photochemical sources of peroxides in the gas phase have been explored by numerous researchers, with the bimolecular combination of HO2 or RO2 being found to be the dominant source of peroxides in this phase.3−5 Peroxides can also be produced by the ozonolysis of alkenes, a production pathway that is independent of radiation and may be prominent in low photochemical conditions. When high levels of NO are present, peroxide formation is suppressed, as the peroxy © 2013 American Chemical Society
radicals react with NO to form NO2. The reaction between O3 and HO2 acts as a similar suppression mechanism. The H2O2 photolysis, reaction with OH radicals, and dry deposition are believed to be the main H2O2 sinks in the gas phase. In aqueous phase, although it was found some 20 years ago that transition metal ion (TMI) reactions were capable of producing and consuming H2O2,6−8 the dependence of the H2O2 budget on aerosol composition is still not well-known because of the complexity of the multiphase system and the wide variation in aerosol composition across sites.9,10 In a recent field study, a higher aerosol aqueous-phase H2O2 concentration than that predicted by gas-particle partitioning (Henry’s Law) was observed in California, suggesting net H2O2 formation in the aqueous phase.11,12 However, current modeling investigations Received: Revised: Accepted: Published: 1443
July 20, 2013 December 19, 2013 December 20, 2013 December 20, 2013 dx.doi.org/10.1021/es403229x | Environ. Sci. Technol. 2014, 48, 1443−1450
Environmental Science & Technology
Article
60%−85% for MHP and 100% for HMHP).17,19,20 Due to lack of information on organic peroxides speciation at our sites, we did not correct their concentration with any assumed efficiency, thus the organic peroxides concentration reported here represents a lower limit of their real atmospheric value. Other trace gases, aerosols compositions, and several meteorological parameters were also measured during the campaigns. They are used to constrain the model simulation, and details of the measuring instruments are described in the Supporting Information. 2.2. Modeling Procedures with RACM and CAPRAM. A zero-dimensional chemical model comprising the gas-phase regional atmospheric chemical mechanism (RACM scheme with 237 reactions)21 coupled to a chemical aqueous-phase radical mechanism (CAPRAM2.4 scheme with 438 reactions)10 was used to investigate the peroxide production and loss at the observation sites. The RACM simulates chemical reactions leading to production and loss of peroxides in gas phase, whereas CAPRAM investigates mass transfer between gas phase and deliquescent aerosols and chemical reactions in aerosol aqueous phase. The combined model is constrained with the measurements of gas-phase species O3, NO, NO2, HONO, SO2, and CO and 52 volatile organic compounds (VOCs, including formaldehyde, other aldehydes, and ketones), aerosol size distribution, aerosol major ions and trace elements, and meteorological parameters (temperature, relative humidity, boundary layer height). The detailed reaction list and recent developments concerning the CAPRAM mechanism can be found in the literature10,22 and at the CAPRAM home page (http://projects. tropos.de/capram). Phase transfer of chemical species between gas and aqueous phases is calculated in the model using the resistance model approach proposed by Schwartz et al.23 The aerosol radius used for evaluating the mass transfer rate was set as the corresponding radius of volume mean mass transfer constant (Kmt) calculated from all aerosol size bins (see more details in the Supporting Information). An Aerosol Inorganic Model (E-AIM model IV)24 was used to calculate the liquid water content (LWC), aqueous-phase ions concentrations (including pH), ionic strength, and the activity coefficients of major ions (i.e., H+, Na+, NH4+, Cl−, NO3−, SO42‑) with the input of field measured species concentrations and the ambient relative humidity. Under conditions of large ion strength such as for deliquescent aerosols, the activity coefficient should be taken into account when modeling aerosol aqueous chemistry.25 The CAPRAM2.4 mechanism has been further extended by considering the activity coefficients in chemical equilibrium and in ionic reactions. The activity coefficients for neutral species and trace metal ions are calculated by Pitzer ion interaction model26 with parameters from Christov27 and Millero et al.28 The Henry’s law constants of gaseous species were corrected by the activity coefficients for the salting out effect.25 Further details about the calculation method for the activity coefficients are provided in the Supporting Information. The concentrations of iron (Fe), manganese (Mn), and copper (Cu) used in the model were constrained with the semiobservational data estimated from the integrated filter concentrations and time profile of continuously measured inorganic ions in aerosol, by assuming a fixed fraction of transition metal concentration in total cationic concentration for episode days and clean days. In our study, the urban metal soluble fractions (Fe: 10%; Mn: 70%; Cu: 20%) derived from the metal solubility range in Deguillaume et al.29 were used for
rarely consider the aqueous interfacial processes in the atmospheric chemistry of H2O2. Hong Kong is a metropolis situated on the South China coast. Its coastal and subtropical environment favors peroxide production because of the abundant water vapor and strong solar radiation. However, the large amounts of NOx emissions from the numerous vehicles including ships in Hong Kong serve to suppress that production. Although many studies have investigated photochemical pollution in the region, to date there has been no investigation of the mixing ratio, sources, and sinks of peroxides and their role in the region’s atmospheric photochemistry. Hence, the present study was carried out to fill this research gap. This research constitutes the first observation of the seasonal behavior of atmospheric peroxides in China. An observation-based model (OBM) with a coupled gas-phase regional atmospheric chemical mechanism (RACM) and chemical aqueous-phase radical mechanism (CAPRAM2.4) was subsequently designed to investigate the gas- and aqueousphase contributions to the H2O2 budget. The overall objective of this work was to investigate the role of aerosol heterogeneous processes in determining the chemical budget of peroxides in humid environments.
2. EXPERIMENTAL AND MULTIPHASE BOX MODEL 2.1. Sites and Instrumentation. Measurements of H2O2, organic peroxides, and related species were taken at two sites in Hong Kong in 2011−2012: a regional background site at Hok Tsui (HT) and a suburban site at Tung Chung (TC). Locations of the two sites are shown in Figure S1 of the Supporting Information. The HT site is the Hong Kong Polytechnic University’s background air monitoring station, which is located at the southeastern tip of Hong Kong Island (22.217° N, 114.25°E, 60 m above sea level) on a cliff with a 270° view of the South China Sea. Most of the time, the station is upwind of the urban areas of Hong Kong and the Pearl River Delta (PRD). It has been involved in several long-term and intensive studies of photochemistry.13 In the current study, measurements were taken at the HT site in May, August, and December of 2011 and in February of 2012 (i.e., one month in each season). The TC site is an air-quality monitoring station operated by the Hong Kong Environmental Protection Department. It is located in the northern part of Lantau Island in western Hong Kong, an area known to suffer from serious photochemical pollution and poor visibility.14,15 The site is in a newly developed residential town adjacent to Hong Kong International Airport and several highways. In addition to the local emission sources, both clean background air and aged urban plumes can be sampled at the TC site, making it a suitable place for investigating the photochemical evolution of plumes from various urban areas in the region. The measurements at TC were taken in August and December of 2011 and February and May of 2012. H2O2 and organic peroxides were measured with an Aerolaser AL-2021 analyzer based on the enzyme-catalyzed fluorescence technique.16 The technique is sensitive to all peroxides in the solution, with two parallel channels used to distinguish between H2O2 and organic peroxides. The detection limit of the analyzer was
