Characteristics and major influencing factors of sulfate production via heterogeneous transition-metal-catalyzed oxidation during haze evolution in China

Highlights

• TMI-catalyzed pathway varied considerably for different haze events.

• TMI-catalyzed pathway did not always increase in haze aggravation stages with higher aerosol acidity.

• Fe(III), Mn(II) and ionic strength play important role on the production rate of the TMI-catalyzed pathway.

• Crustal particulates from north or northwest may have a non-negligible effect on sulfate formation during haze in Beijing.

Abstract

The heterogeneous transition-metal-catalyzed oxidation (hereinafter referred to as TMI-catalyzed pathway) is an important pathway for atmospheric sulfate formation. While most research has focused on in-cloud reactions of sulfate, studies on its heterogeneous reactions on aerosol surfaces remain scarce. To better understand this pathway, we investigated the temporal changes of sulfate heterogeneous production rates of this pathway during haze evolution, and did not find any pattern in different haze stages. Although previous studies have shown a high sensitivity of the TMI-catalyzed pathway to aerosol pH, sulfate production rates of this pathway did not always increase in haze aggravation stages with higher aerosol acidity. We then explored the major influencing factors on this pathway in various haze evolution stages. The results suggest that the highly variable dissolved Fe(III) and Mn(II) concentrations and ionic strength during haze evolution processes would considerably influence the production rate of the TMI-catalyzed pathway. This further implies that the north or northwest air masses, which often carry crustal-source particulates, may have a non-negligible effect on sulfate formation during the generation and/or dissipation stages of haze in Beijing by influencing aerosol pH and transition-metal contents.

Introduction

Sulfate (SO₄²⁻) is one of the main constituents of atmospheric aerosols and has a significant impact on environmental chemistry and the global climate. The assessment from Haywood and Boucher (2000) showed that the direct radiative forcing of sulfate aerosols ranged from −0.26 to −0.82 W m⁻², indicating a net cooling effect on the Earth's climate. Sulfate aerosols are mainly produced through the oxidation of SO₂ in the atmosphere. Sources of SO₂ include direct emissions of fossil fuel combustion, industrial production, volcanic eruption, or oxidation of reduced sulfur such as dimethyl sulfur (DMS) emitted from marine phytoplankton. As a major component of secondary inorganic aerosols (SIA), the SO₄²⁻ content often increases rapidly during the evolution of haze in China. Taking the haze in January 2013 for example, mass percentages of SO₄²⁻ in PM_2.5 around Beijing increased from 10.3% to 13.4% on non-haze days to 25.1% on haze days (Quan et al., 2014; Zhang et al., 2014). Many other studies also showed that SO₄²⁻ mass fractions of PM_2.5 clearly increased during haze pollution periods (Cheng et al., 2016; Li et al., 2017; Yue et al., 2019). These observations indicate that the formation of SO₄²⁻ is an important driver of the rapid growth of PM_2.5.

The transformation processes from SO₂ to SO₄²⁻ include three major pathways: gas-phase reactions, liquid-phase reactions in clouds, and heterogeneous reactions on the surface and/or within the bulk pre-existing aerosols (Sofen et al., 2014). The gas phase reactions of SO₂ are mainly carried out through OH⋅ radical oxidation (Stockwell and Calvert, 1983), and about 25% of the secondary SO₄²⁻ is produced by this pathway globally (Faloona, 2009). During haze periods in China, the sharp decline of solar radiation can weaken photochemical activities, which may greatly reduce the importance of the SO₂ + OH⋅ gas phase reactions (Zheng et al., 2015). The liquid phase reactions in clouds are considered the primary pathway of SO₄²⁻ production at the global scale, accounting for about 60% of SO₄²⁻ formation (Alexander et al., 2009, 2012bib_Alexander_et_al_2012bib_Alexander_et_al_2009). The reactions include the oxidation of S(IV) (total soluble phase SO₂·H₂O, HSO₃,⁻ and SO₃²⁻ mass) through H₂O₂, O₃, and O₂ catalyzed through transition-metal ions (TMIs) in cloud droplets (Harris et al., 2013; Hoffmann and Calvert, 1985; McArdle and Hoffmann, 1983). Vogt et al. (1996) proposed hypochlorous acid (HOX = HOCl + HOBr) as a potentially important oxidant of S(IV) in the marine boundary layer. Chen et al. (2016) added this reaction to the global model GEOS-Chem and used SO₄²⁻ oxygen isotopes observed in the marine boundary layer as constraints, and found that 33%–50% of the SO₄²⁻ in the marine boundary layer could be formed through this pathway. However, estimates of the contribution from this pathway are limited by the uncertainty of the HOX concentration, its corresponding reaction rate, and the Henry's Law constant. Different from liquid phase reactions in clouds, the contribution of heterogeneous SO₄²⁻ production is usually thought to be minor because of the low aerosol liquid water content (ALWC) compared with clouds (Jacob, 2000). However, recent studies suggested that the traditional gas- and aqueous-phase pathways in cloud droplets were insufficient to explain the rapid SO₄²⁻ production observed during haze periods (Cheng et al., 2016; Huang et al., 2014; Shao et al., 2019; Wang et al., 2014). This underestimation may be due to the fact that some SO₄²⁻ formation mechanisms such as heterogeneous oxidation of SO₂ by NO₂ (Cheng et al., 2016; Gao et al., 2016; Zhang et al., 2016a) and TMI-catalyzed oxidation by O₂ (Li et al., 2017; Shao et al., 2019) are overlooked.

In previous studies, the TMI-catalyzed pathway was mainly considered within the cloud environment because of its high liquid water content. Previous observations of oxygen and sulfur isotopes indicated that such liquid phase reactions in clouds are more important in continental regions with strong anthropogenic and/or natural transition-metal emission sources (Alexander et al., 2009; Harris et al., 2013). Observations and corresponding simulations of SO₄²⁻ oxygen isotopes suggested that in high-latitude regions of the northern hemisphere, where OH⋅ radicals and H₂O₂ concentrations in winter are sufficiently low, the TMI-catalyzed pathway in clouds is the main SO₄²⁻ formation pathway (Alexander et al., 2009). Observations on Mountain Tai showed that the TMI catalyzation was the fastest SO₄²⁻ generation pathway during 12% of the sampling period (Shen et al., 2012). However, this pathway has gained little attention in studies of heterogeneous reactions that occur on aerosol surfaces (e.g., in aerosol liquid water). By incorporating different heterogeneous SO₄²⁻ formation mechanisms into the GEOS-Chem model and combining with measurements of oxygen isotopes (Δ¹⁷O (SO₄²⁻)) (He et al., 2018), Shao et al. (2019) found that the TMI-catalyzed pathway was dominant among heterogeneous SO₄²⁻ production in both heavily-polluted periods (69%) and clean periods (67%), suggesting a considerable contribution of this pathway to atmospheric SO₄²⁻ formation.

A model sensitivity analysis by Cheng et al. (2016) showed that SO₄²⁻ production rates of the TMI-catalyzed pathway are highly sensitive to aerosol pH in heterogeneous reactions, with substantially higher production rates under lower pH conditions. However, their analysis ignored the effect of high ionic strength (I_S) in aerosol water, which would vary considerably during haze evolution and cause non-trivial inhibition to this pathway (Zhang et al., 2015). In addition, soluble Fe(III) and Mn(II) concentrations may vary significantly in different haze stages because of the change in air masses, which may carry sufficient crustal particles from northwest regions to Beijing (Yue et al., 2019). On account of limited research on the SO₄²⁻ heterogeneous TMI-catalyzed pathway and its importance on SO₄²⁻ production, we explored the characteristics of this pathway during haze evolution and considered the main influencing factors of this pathway in different haze stages. The results of this study can be used to better understand SO₄²⁻ formation during haze periods.