Increasing Isoprene Epoxydiol-to-Inorganic Sulfate Aerosol (IEPOX:Sulf_inorg) Ratio Results in Extensive Conversion of Inorganic Sulfate to Organosulfur Forms: Implications for Aerosol Physicochemical Properties

Abstract

Acid-driven multiphase chemistry of isoprene epoxydiols (IEPOX), key isoprene oxidation products, with inorganic sulfate aerosol yields substantial amounts of secondary organic aerosol (SOA) through the formation of organosulfur compounds. The extent and implications of inorganic-to-organic sulfate conversion, however, are unknown. In this report, we demonstrate that extensive consumption of inorganic sulfate occurs, which increases with the IEPOX-to-inorganic sulfate concentration ratio (IEPOX:Sulf_inorg), as determined by laboratory measurements. Characterization of total sulfur aerosol observed at Look Rock, Tennessee, from 2007–2016 shows that organosulfur mass fractions will likely continue to increase with ongoing declines in anthropogenic Sulf_inorg, consistent with our laboratory findings. We further demonstrate that organosulfur compounds greatly modifies critical aerosol properties, such as acidity, morphology, viscosity, and phase state. These new mechanistic insights demonstrate that changes in SO₂ emissions, especially in isoprene-dominated environments, will significantly alter biogenic SOA physicochemical properties. Consequently, IEPOX:Sulf_inorg will play an important role in understanding historical climate and determining future impacts of biogenic SOA on global climate and air quality.

Graphical Abstract

Introduction

Secondary organic aerosol (SOA) formed through the oxidation of volatile organic compounds is a major and globally ubiquitous component of atmospheric fine particulate matter (PM_2.5; aerosol particles ≤ 2.5 μm in aerodynamic diameter).¹ Chemical composition of fine particles determines aerosol physicochemical properties, such as viscosity and phase state, and plays a central role in the effects of SOA on air quality and climate.^2,3 Understanding how SOA forms and interacts with other gas- and particle-phase species is crucial to accurately evaluating its importance in the Earth’s climate system and adverse effects on public health.

Inorganic sulfate species (e.g., SO₄²⁻, HSO₄⁻) are also a significant PM_2.5 component with the capacity to impact atmospheric composition and climate, in part, because of its predicted impact on aerosol acidity, hygroscopicity, visibility and cloud nucleation.^1,4 The oxidation of sulfur dioxide (SO₂) to sulfuric acid (H₂SO₄) increases aerosol acidity, which enhances SOA formation.^4–8 Sulfur in the form S(VI) in aerosols was generally assumed to be primarily present as inorganic sulfate (SO₄²⁻ and HSO₄⁻) ions until more recent studies revealed the presence of organosulfur components in PM_2.5.^9–13 Despite observations that organosulfur compounds are important contributors to SOA mass in a range of environments globally,^14–20 estimations of aerosol acidity and liquid water content typically assume that only inorganic sulfate plays a role.²¹ Correctly identifying the chemical form of sulfur (i.e., inorganic vs. organic), and representing it accurately in atmospheric models is essential as the different forms lead to different aerosol physicochemical properties that will have different predicted impacts on air quality and climate.³

Laboratory studies have demonstrated that acid-driven multiphase chemistry (reactive uptake) of isoprene epoxydiols (IEPOX) is key to explaining the chemical form and extent of SOA formation from photochemical oxidation of isoprene ^22–25 and field measurements have confirmed this is the predominant pathway.^26–31 While chamber studies have shown that organosulfur compounds, specifically organosulfates (OS) ³² formed by the reactive uptake of IEPOX with particulate inorganic sulfate (Sulf_inorg), contribute significantly to IEPOX-SOA,^33,34 the extent and implications of sulfate conversion to organic forms have remained unknown.

Combined laboratory, field, and modeling studies described in this study reveal the hitherto unrecognized effect that acid-driven multiphase chemistry of IEPOX may result in substantial conversion of Sulf_inorg to organosulfur compounds. Laboratory experiments reveal that the extent of Sulf_inorg conversion is greater for high IEPOX:Sulf_inorg ratios (such as observed in the Amazon) than for lower ratios (such as observed in the Southeastern United States (SE U.S.)). Assuming IEPOX concentration remains constant, continuing Sulf_inorg reductions in the Northern Hemisphere as a consequence of SO₂ emission controls are thus expected to greatly increase the fraction of Sulf_inorg converted in the future. High IEPOX:Sulf_inorg scenario also likely characterized pre-industrial atmospheric conditions. Changes in the ratio over time and the consequent changes in SOA chemical composition and physicochemical properties will increase understanding of the evolution of climate change and inform projections for the future.

Experimental Methods

Smog Chamber Experiments.

Experiments were performed in the indoor environmental chamber facility at the University of North Carolina. The experimental setup and analytical techniques used in this work have been described in detail elsewhere ^33,35. Briefly, experiments were carried out under dark and wet conditions (50 ± 4 %, relative humidity (RH)) at 296 ± 1 K in a 10-m³ Teflon chamber. A summary of the experimental conditions is provided in Table S1. Prior to each experiment, the chamber was flushed continuously with clean air for ~ 24 hours corresponding to a minimum of seven chamber volumes until the particle mass concentration was < 0.01 μg m⁻³. Chamber flushing also reduced VOC concentrations below the detection limit (i.e., ~ 75 ppt for IEPOX). Temperature and RH in the chamber were continuously monitored using a dew point meter (Omega Engineering Inc.). Acidified ammonium sulfate seed aerosols were injected into the pre-humidified chamber using a custom-built atomizer with an aqueous solution of 0.06 M (NH₄)₂SO₄ (aq) and 0.06 M H₂SO₄ (aq) until the desired total aerosol mass concentration was achieved. After seed injection, the chamber was left static for at least 30 min to ensure that the seed aerosol stability and uniform mixing. The IEPOX vapor was generated by using a high-purity N₂ (g) flow of 2 L min⁻¹ for 10 min and then 4 L min⁻¹ for 50 min through a heated manifold (60°C) containing a known amount of pure trans-β-IEPOX (synthesized in house³⁶) dissolved in ethyl acetate. This approach to introduce IEPOX has been used in previous studies.^{33, 37, 38}

Aerosol size distributions were continuously measured using a differential mobility analyzer (DMA, BMI model 2002) coupled to a mixing condensation particle counter (MCPC, BMI model 1710) in order to monitor aerosol number, surface area, and volume concentration within the chamber. SOA generated from the reactive uptake of IEPOX was collected using a particle-into-liquid sampler (PILS, BMI model 4001) throughout each experiment. PILS samples were collected at 5 min intervals for offline chemical analysis by ion chromatography (IC) and ultra-performance liquid chromatography interfaced to electrospray ionization high-resolution quadrupole time-of-flight mass spectrometry (UPLC/ESI-HR-Q-TOFMS). The PILS vials were stored in the dark at 2°C immediately after collection and analyzed within 24 h without further pretreatment. PILS operating conditions and dilution correction are described in the SI.

Flow Tube Experiments.

The reactive uptake coefficient of IEPOX (γ_IEPOX) is an important model-relevant parameter that characterizes the heterogenous chemistry of IEPOX.^{24, 36} Direct measurement of γ_IEPOX onto aqueous ammonium bisulfate (ABS) particles was conducted at the University of Washington in an aerosol flow tube coupled to an iodide-adduct chemical ionization mass spectrometry (CIMS); operating conditions have been described elsewhere and in the SI.^37,38

Collection of PM_2.5 samples.

Ambient SOA (PM_2.5) samples from two isoprene-dominated environments (SE-U.S. and Amazon forest) were collected onto quartz filters during three field campaigns: (i) during the 2013 Southern Oxidant and Aerosol Study (SOAS) campaign from 1 June to 15 July 2013 at the Centerville, AL (CTR, AL) ground site; (ii) from 18 July through 1 August 2016 from downtown Manaus, Brazil and (iii) during the Green Ocean Amazon (GoAmazon2014/5) field campaign³⁹ during intensive operating period 2 (IOP2). Additional information regarding filter collection is provided in SI.

Aerosol Chemical Characterization.

Chemical characterization of PM_2.5 and PILS samples was performed by UPLC/ESI-HR-Q-TOFMS (6520 Series, Agilent) operated in the negative ion mode.^40,41 Total water-soluble organosulfur compound mass was determined from the difference between total water-soluble sulfur measured by isotope ratio inductively coupled plasma mass spectrometry (IR-ICP-MS) and sulfate-sulfur (sulfur in the form of inorganic sulfate/Sulf_inorg) measured by IC on the same sample aliquot. Sample extractions, operating conditions, standard preparation, and uncertainty estimates are described in detail in SI. The National Park Service Improve database⁴² was used to evaluate the inorganic sulfate and organosulfur mass concentrations in the SE-U.S from 2007 to 2016.

Microscopy Imaging.

Morphology of aerosol particles collected before, during, and after IEPOX reactive uptake from smog chamber experiments was characterized by atomic force microscopy (AFM) and scanning electron microscopy (SEM). Additional information regarding operating conditions is provided in SI and prior publications.^43–45

Biphasic Microfluidics.

Surface and interfacial tensions inform solute surface-bulk partitioning and, therefore, the availability of compounds for interactions with the ambient gaseous phase, such as heterogenous chemistry and water uptake. Lower interfacial tensions due to organic partitioning affect both particle growth and structure. We measured liquid-liquid interfacial tensions of IEPOX SOA tracer 2-methyltetrols (2-MT) and its OS derivative (3-methyltetrol sulfate, IEPOX-OS) using a biphasic microfluidic platform. Droplet microfluidics as a method to measure interfacial tension for aerosol science applications was first introduced for studying the aging of SOA aerosols,⁴⁶ using a microtensiometry technique described in detail in several studies and in the SI.^47–51 The microfluidic chip is fabricated using standard soft-lithography techniques ^52–54 and mounted on an inverted microscope. Two immiscible liquid phases separately enter the device by pressure-driven flow, ⁵¹ and droplet breakup is induced by a flow-focusing geometry ^55–57 Aqueous solutions containing the SOA are the dispersed (droplet) phase, and silicone oil is the carrier (continuous) phase. The droplets of aqueous SOA then enter a contraction geometry in the microfluidic channel, which induces deformation of the liquid-liquid interface. This deformation, is imaged at high speed, and related to material and flow-field properties ^{48,49,58–61} to calculate the interfacial tension between silicone oil and aqueous SOA solutions. The equation governing droplet deformation is:

$$\alpha {\eta }_c\left(\frac{5}{2\widehat{\eta }+3}\dot{\epsilon }\left(x\right)-u\left(x\right)\frac{\partial D\left(x\right)}{\partial x}\right)\hspace{0.17em}=\sigma \frac{D\left(x\right)}{a_0}.$$ (1)

In the above equation, α is a function of the viscosity ratio $\widehat{\eta }$, which is a ratio of the dispersed phase viscosity (here, the aqueous phase viscosity is close to that of pure water – 1 mPa s – near 100% RH for aerosol droplets), to the continuous phase viscosity η_c (here, the viscosity of silicone oil – 50 mPa s). For solutions containing 1.55 M ammonium sulfate (AS), the viscosity is 1.215 mPa s.⁶² $\dot{\epsilon }\left(x\right)$ is the extensional strain rate in the contraction, u(x) is the velocity of the droplet centroid, D(x) is the deformation, a₀ is the un-deformed diameter of the spherical droplet, and σ is the interfacial tension between the aqueous and oil phase. D(x) is defined by Taylor ^58–60 as $D\left(x\right)=\frac{d_{major}-d_{minor}}{d_{major}+d_{minor}}$, where d_major and d_minor are the major and minor diameters of the deformed droplet respectively, measured using image analysis. A plot of the left side of the equation of motion versus $\frac{D\left(x\right)}{a_0}$ is linear in the small-deformation regime ^58,60. A straight line is fit to the linear portion of the curve, known as a “Taylor plot”, and the slope of the line is equal to the interfacial tension.

Each microfluidic measurement consumes 250–300 μL of aqueous solution, yielding 20–30 droplets. The IEPOX-OS droplet interfacial tensions reported in this work were statistically averaged. All interfacial tension values reported in this work are then normalized by the interfacial tension of pure water with silicone oil measured with microfluidics (30 mN/m). Figures S1 and S2 show Taylor plots and interfacial tensions for 2-MT and IEPOX-OS samples. The surface partitioning of SOA is analogous to that expected to occur in aerosol droplets in the atmosphere, as discussed in Metcalf et al.⁵¹ Techniques for measurement of interfacial tension using biphasic microfluidics are outlined in prior work.^51,63

Viscosity Model.

The viscosity of IEPOX-derived OS is calculated based on a modified version of the Vogel-Tammann-Fulcher (VTF) equation (Eqs. 2 & 3) by Angell et al.^64,65

$$\eta \left(RH\right)={\eta }_{\infty }{e}^{\frac{T_{0}D}{T-T_0}}$$ (2)

where η_∞ is viscosity at infinite temperature and assumed to be 10⁻⁵ Pa s; T₀ is the Vogel temperature; T is the ambient temperature; and D is the fagility parameter, which controls how closely a material follows the Arrhenius law.⁶⁴ When T reaches glass transition temperature T_g, η reaches 10¹² Pa s. Then Eq. 2 becomes

$$\raisebox{1ex}{$T_g$}\!\left/ \!\raisebox{-1ex}{$T_0$}\right.=1+0.0255D$$ (3)

Both D and T_g are required to calculate η for SOA-water mixtures at a given temperature. The T_g of each SOA compound comprising the SOA under dry conditions was estimated based on the improved paramterizaiton described in DeRieux et al.⁶⁶ Then Gordan-Taylor mix rule (Eq. 8 in the SI) was applied to calculate the T_g of SOA mixtures. A detailed description of the workflow and choices of parameters used in calculating the viscosity of IEPOX-derived OS is given in the SI.

Thermodynamic Model of Aerosol pH.

The pH of aerosol particles during the heterogenous reaction processes of IEPOX is calculated with smog chamber data using a thermodynamic model employing two methods. Briefly, the first method uses measured Sulf_inorg concentration as well as the NH₄⁺ concentration at the beginning of the experiment as the input variables. The concentrations of IEPOX-OS and oligomeric OS are not considered. While retaining the inorganic cations and anions considered in the first scenario, the second method takes the anions of IEPOX-OS monomer as well as its dimers and trimers into consideration. The proton balance equation together with the mass balance equations for NH₄⁺, SO₄²⁻, HSO₄⁻, and IEPOX-OS monomers, dimers, and trimers were built to solve the H⁺ ion concentration. The gas-particle equilibrium of NH₃ and NH₄⁺ was considered using a Henry’s law constant of 0.0161 atm M⁻¹. A growth factor of 1.3 was used to calculate the liquid water content of the particles. Similar to the ISORROPIA-II model,⁶⁷ the activity coefficients for all ions were assumed to be 1 in this case. The uncertainties in the growth factor of the particle, the activity coefficient of each ion, and the pKa values of organosulfur species contribute to the overall uncertainty of this model. These factors may affect the exact values of the pH calculated; however, the drastic difference of the pH values between considering and not considering the organosulfur species remains roughly the same. A detailed description of the thermodynamic model is given in the SI.

Global Modeling.

The IEPOX:Sulf_inorg mass ratio has been computed based on numerical simulations performed with the LMDz-OR-INCA global climate-chemistry model. The description of the model representation for sulfate as well as the general features of the LMDz-OR-INCA model can be found in Szopa et al. ⁶⁸ The reaction mechanism of IEPOX is based on St Clair et al.⁶⁹ and considers three different isomers of IEPOX produced by three different geometrical configurations of the isoprene hydroxyperoxide (ISOPOOH) precursor produced through the isoprene oxidation.

The LMDz-OR-INCA model is used with a 3.8° lat × 1.87° lon horizontal resolution and 39 vertical levels. The wind fields are nudged on the ECMWF (European Centre for Medium-Range Weather Forecasts) reanalysis for the year of 2010 (considering first a one-year spin-up). The anthropogenic emissions are those from the Representative Concentration Pathways (RCP) considering the year 2010 in the 8.5 trajectory (compatible with the evolution of radiative forcing equivalent in 2100 to 8.5 W.m⁻²)⁷⁰. The biomass burning emissions correspond to the GFED-v4 inventory for the year of 2010. The anthropogenic, ship and fire emissions of SO₂ are respectively 40.5, 5.9 and 1.1 TgS yr⁻¹. The biogenic emissions are computed by the ORCHIDEE vegetation model as described elsewhere ⁷¹. They lead to a global annual isoprene emission of 466 TgC yr⁻¹ inducing a production of IEPOX of 115 TgC yr⁻¹, which is consistent with the one found by St Clair et al. ⁶⁹

Results and Discussion

IEPOX conversion of Sulf_inorg to organosulfur.

Despite the wealth of studies on the reactive uptake of IEPOX, its reactivity remains poorly constrained.^{24,25,28,33,40,72} We performed controlled chamber experiments in the presence of ABS seed particles (pH = 1.5) at ~50 % RH using atmospherically-relevant ratios of IEPOX:Sulf_inorg (SI Table S1). Figure 1 shows that immediately following IEPOX addition, rapid conversion of Sulf_inorg is observed under all conditions measured by the PILS coupled to an IC system operated with 5-minute resolution. The rate of Sulf_inorg depletion correlates with the formation of IEPOX-OS and oligomeric-OS (quantified by UPLC/ESI-HR-Q-TOFMS for the same PILS samples), which is further supported by computational chemistry modeling by Piletic et al. that suggests IEPOX would readily produce OSs in acidic aerosols.⁷³ IEPOX-OS accounts for most (90–100%) of the Sulf_inorg converted within the first 40–60 min under conditions that mimic IEPOX:Sulf_inorg ratios relevant to both the SE-U.S. (Figure 1A) and the Amazon (Figure 1B). As shown in Figures 1A and B, ~40% of Sulf_inorg injected into the chamber is converted to organosulfur under laboratory conditions that mimic the SE-U.S. conditions, while up to 90% is converted to organosulfur under laboratory conditions that mimic the Amazon. During the next hour following the termination of IEPOX injection, Sulf_inorg as well as total organic sulfate has stabilized, indicating inhibition of IEPOX uptake (discussed below), then a decrease of concentrations of IEPOX-OS commences. A net reduction (up to 30% in one hour) of the three quantified OS species indicates that IEPOX-OS are not stable and react to yield as yet uncharacterized organosulfur compounds. One potential class of species, sulfur-containing oligomers, were observed below quantifiable levels in the positive ion mode.

Figure 1.

Laboratory conversion (RH=50%) of Sulf_inorg to organic sulfate during the reactive uptake of IEPOX at IEPOX (μg m⁻³):Sulf_inorg (μg m⁻³) ratios atmospherically relevant to (A) a laboratory mimic of the SE-U.S. (ratio=1.5) and (B) a laboratory mimic of the Amazon (ratio=17.1). IEPOX is injected into the chamber at t = 0 and stopped at t = 60. IEPOX-OS dimer and trimer concentrations are shown on and expanded scale for clarity. (C) Filled triangular markers indicate the conversion of Sulf_inorg to organic sulfate as a function of IEPOX:Sulf_inorg in chamber experiments at RH~50%. Colors correspond to those in Figure 1D. Red open circles correspond to projected IEPOX:Sulf_inorg over the next 50 years, based on Sulf_inorg reduction rates from Attwood et al.⁷⁵ and Hand et al.⁷⁶, and assuming constant IEPOX. (D) Yearly average IEPOX:Sulf_inorg across the world for present-day conditions using the LMDz-OR-INCA global climate-chemistry model. White color indicates no sulfate (non-sea salt) or IEPOX concentration.

Our results demonstrate the conversion of inorganic-to-organic sulfate is driven primarily by the IEPOX:Sulf_inorg ratio as illustrated in Figure 1C. From our laboratory experiments, the IEPOX:Sulf_inorg ratio appears to be a critical, previously unrecognized factor in the conversion of inorganic-to-organic sulfate. By extrapolating our laboratory findings to ambient conditions, high IEPOX:Sulf_inorg ratios (>2) are common globally, as shown in Figure 1D, especially in equatorial regions and the Southern Hemisphere, and might be important in the formation of organic sulfate 9in such areas. Based on our experimental results, we infer that a higher S(VI) fraction is likely in organic forms in those regions, which are under-sampled relative to the Northern Hemisphere.

Concentration of ambient organosulfur compounds.

Over the past 30 years, air pollution regulations in the Northern Hemisphere have led to decreases in SO₂ concentration of 5.7 % yr⁻¹ in the U.S. and 5.1 % yr⁻¹ in Europe.⁷⁴ Recent studies project a reduction of SO₄²⁻ of ~4.5% yr⁻¹ based on current efforts.^75,76 Figure 2A illustrates this trend and presents the SO₄²⁻ concentration over the last 10 years measured at the Great Smoky Mountain Site (Look Rock, TN) through the IMPROVE monitoring program,⁴² which is strongly influenced by isoprene chemistry.²⁷ A decrease of Sulf_inorg by a factor of 6 is correlated with a subsequent rise in organosulfur as a fraction of total sulfur by a factor of 5. In addition, events with high organosulfur fractions tend to correlate with low Sulf_inorg concentrations (Figure 2B), consistent with a previous analysis¹² and corroborating our laboratory findings. Assuming a constant IEPOX concentration and a reduction of SO₄²⁻ of ~4.5% yr⁻¹, the increase in the IEPOX:Sulf_inorg ratios within the next decades is expected to significantly shift the distribution of S(VI) towards organosulfur (Figure 1C). Since industrialization, Sulf_inorg has significantly increased^77,78 while the concentration of isoprene has remained constant.⁷⁹ This trend indicates that IEPOX:Sulf_inorg was likely much higher during the pre-industrial period than currently, and possibly had a major role in SOA physicochemical properties in isoprene-dominated areas. While the estimated conversion extrapolated from our chamber experiments is likely an upper limit as the reactive uptake of IEPOX is sensitive to aerosol acidity and other factors in the ambient environments (e.g. NO_x, organic coating, RH, etc.), the trend identified in this study is central to understanding future air quality and radiative forcing.

Figure 2.

(A) Correlation between total sulfate and organosulfur fraction and (B) Evolution of organosulfur fraction as a function of Sulf_inorg at the Great Smoky Mountain Site (Look Rock, TN) during summer (May – September) from 2007–2016, using the National Park Service IMPROVE PM_2.5 database. If particulate organosulfur compounds are not present, the sulfate (SO₄²⁻) mass measured by IC should equal three times the sulfur mass concentration measured by X-ray fluorescence (XRF), since the molar mass of sulfur and sulfate are 32 and 96 g/mol, respectively. A value of 3×S/ SO₄²⁻ lower than 1 is caused by the limitation of the analytical techniques to differentiate 3×sulfur from SO₄²⁻.⁴² (C) Average mass concentration of the identified organosulfates + methane sulfonic acid (OS + MSA) and total organosulfur compounds (Org-S) in the PM_2.5 samples collected during the 2013 SOAS campaign as well as the average mass concentrations of the sum of OS + MSA quantified in downwind Manaus and Manaus. AMS data from SE-U.S. and downwind Manaus are also presented.

Molecular characterization of SOA samples collected in isoprene-rich areas show that OS alone contribute a substantial mass fraction of S(VI) in PM: from 8 to 25% in the SE-U.S. and from 20 to 45% in downwind Manaus (Figure 2C and SI Figure S1). From direct comparison between ICP-MS and AMS sulfate data (Figure S3), we applied a correction factor (i.e., 1.28) in order to account for the subset of organosulfur compounds present within the size range of 1 to 2.5 microns. OS mass ranges from 400–1500 ng m⁻³, at the high end of previous studies (SI Tables S2–S4).^{26,27,32,80,81} While the sum of OS in the SE-U.S. is significantly different from that of downwind Manaus, isoprene-OS nevertheless represents the predominant OS in both areas. During the 2013 SOAS campaign, total mass concentrations of organosulfur compounds were also determined by IR-ICP-MS (SI Figure S4). The average total sulfur mass multiplied by three (3 × S) and the total Sulf_inorg measured by IR-ICP-MS and IC, respectively, differs by 300 ± 200 ngSO₄ m⁻³. The difference represents an estimate of total organosulfur compounds and provides an organosulfur/total sulfur ratio similar to that previously observed.^12,13,82 Although total OS + MSA exhibits a strong correlation (r² = 0.78) with the total mass concentration of organosulfur compounds (SI Figure S5), the identified products explain only 50–60% of the total mass of organosulfur quantified in PM_2.5 during the 2013 SOAS field campaign. Assuming similar C:S ratios for identified and unidentified compounds, 16% of the Sulf_inorg is converted into organosulfur compounds (SI Figure S5). In fact, using the fit obtained from all chamber experiments (Figure 1C) and an IEPOX:Sulf_inorg ratio of 0.24 estimated based on collocated measurements during the 2013 SOAS campaign,⁸³ Sulf_inorg conversion to organosulfur is estimated to be ~15%. This is in excellent agreement with the characterizations of field SOA samples presented above.

Atmospheric impact of the acid-driven reactive uptake of IEPOX.

In the following section, we show significant alterations of aerosol physicochemical properties due to high levels of inorganic-to-organic sulfate conversion in the chamber experiments. Explored properties include aerosol acidity, morphology, surface tension, viscosity, and reactivity.

Because acidity is one of the governing factors of atmospheric multiphase chemistry,⁶ the response of condensed-phase acidity to inorganic-to-organic sulfate conversion was further investigated. The widely used thermodynamic models treat sulfate as inorganic S(VI), leading to inconsistencies in charge balance in the condensed phase.^84,85 While organic compounds can reduce the rate of ammonia gas-to-particle partitioning, exclusion of organosulfur compounds such as OSs from thermodynamic models may outweigh this effect and contribute to an even larger discrepancy in acidity. Hence a thermodynamic model was constructed to estimate the acidity of the aerosols from these experiments with and without organosulfur compounds. The model assumes all the components reach thermodynamic equilibrium with the hydrogen ion, while maintaining charge balance and mass balance, similar to previous publications.^67,86,87 The detailed parameterization of the thermodynamic model is described in the SI. As an example, Figure S6 shows the pH of aerosols when not considering the contributions of IEPOX-OS to acidity from the chamber experiments, compared to a thermodynamic box model constructed to take IEPOX-OS into consideration. Aerosol compositions measured at the end of the chamber experiments were used as input to calculate aerosol pH as shown in Figure S6. If the contributions of IEPOX-OS to acidity were not considered, the aerosol pH would have been overestimated by 3.5 and 9.5 units in the laboratory mimic of SE U.S. and Amazon, respectively. The pK_a values of IEPOX-OS remain uncertain. By varying the pK_a values of IEPOX-OS in the range of 0–4, the aerosol acidity changes up to 6 times (0.8 pH units), suggesting the importance of further research on determining the pK_a values of IEPOX-OS and their roles in governing the aerosol acidity.

Acid-driven multiphase chemistry of IEPOX also leads to a modification of the aerosol morphology from a well-mixed sphere to a core-shell structure, shown by height images obtained by AFM at ambient temperature and pressure (Figures 3A–C), as well as AFM amplitude images, AFM phase images and SEM images (Figure S7) of SOA collected from the chamber experiments. It is noted that the AFM images of the laboratory-derived SOA were collected at ambient pressure and RHs close to chamber conditions (30–40%). SEM data were collected after equilibration under vacuum (10⁻³ to 10⁻⁵ Pa). RH was not cycled for these specific particles to check morphology, but has been either cycled or tested at multiple RH values in a recent study.⁸⁸ This study demonstrated that the core-shell structure was not substantially altered by the RH cycling beyond efflorescence and deliquescence of the ammonium sulfate core. Additionally, SEM and AFM images were collected under different conditions, but showed similar morphology, supporting the hypothesis that for the more viscous particles in this study, temperature and RH differences during imaging had minimal effect on the particle morphology. These results are consistent with previous theoretical and semi-empirical studies predicting frequent phase separation for particles in the SE-U.S.^89,90 Larger IEPOX:Sulf_inorg ratios lead to thicker organic shells (e.g., laboratory mimic of Amazon) and can be correlated with the relatively larger amounts of particulate IEPOX-OS and corresponding oligomers (sum of OS corresponds to 27% of Sulf_inorg in Figure 1A and 69% in Figure 1B). Hence, OS and oligomeric-OS may lead to a net modification of the morphology of the OA formed in the atmosphere.

Figure 3.

Left Column (A), (B), and (C). Atomic force microscopy (AFM) height images of ammonium bisulfate seed (ABS), IEPOX-SOA generated from the reactive uptake of IEPOX in the presence of wet acidic aerosol at RH~50% and IEPOX:Sulf_inorg ratios atmospherically relevant to laboratory mimic of SE-U.S. and Amazon, respectively. Middle column (A), (B), and (C). Height maps of impacted particles. (D) Comparison of the measured and predicted viscosity of α-pinene SOA (purple), predicted IEPOX-OS (red) and IEPOX-OS mixtures at 298 K as a function of RH. The data points represent measured viscosity values of α-pinene SOA. The solid lines represent estimated viscosity of α-pinene SOA,⁶⁶ IEPOX-OS, and IEPOX-OS mixtures (IEPOX-OS + Oligomeric-OS). The shaded areas represent the upper and lower bounds of the viscosity estimation for each type of SOA. Model parameters are: glass transition temperature of dry SOA (T_g,org); hygroscopicity (κ), fragility (D), and Gordon–Taylor constant (k_GT). Solid line for - pinene SOA at 278.5K: T_g,org = 0.1, κ = 10, k_GT = 2.5. Solid line for 2-methyltetrol sulfate at 298K; T_g,org = 0.12, κ = 13, k_GT = 2.5.; Solid line for IEPOX-derived OS mixture at 313 K, T_g,org = 0.12, κ = 13, k_GT = 2.5. For upper (lower) bounds (shaded regions): α-pinene SOA; T_g,org = 300 K (268.5 K), κ = 0.1 (0.1), D = 20 (10), k_GT = 2.5 (3.0). 2-Methyltetrol sulfate; T_g,org = 320 K (288 K), κ = 0.10 (0.15), D = 20 (10), k_GT = 2.5 (3.0). OS mixture; T_g,org = 330 K (303 K), κ = 0.10 (0.15), D = 20 (10), k_GT = 2.5 (3.0).

As Sulf_inorg is consumed and OSs are formed, the aerosol phase becomes more viscous and aerosol particle heights increase (Figure 3D). Height images and profiles from AFM were used as a proxy for liquid, semi-solid, or solid phase of the SOA, which is composed largely of an inorganic sulfate seed and organosulfur coating. The observation of more viscous aerosol is supported by the simulated viscosities of molecular tracers based on recent studies,^3,66 which suggest that IEPOX-derived OS has viscosity values 1–4 orders of magnitude higher than α-pinene SOA when the RH levels are below 70% (Figure 3D). Hence, IEPOX-derived OS can significantly increase the viscosity of aerosol particles, compared to isoprene-SOA compounds generated through self-nucleation in atmospheric simulation chambers under high concentrations without the addition of acidic sulfate particles.⁹¹

IEPOX-SOA volatility is significantly lower than structure-based vapor pressures of polyols predict,⁹² which is consistent with the large inorganic-to-organic sulfate conversion reported here. Given the high modelled viscosity of IEPOX-OS and likely higher viscosity values of oligomeric-OS, the OS coating induced inhibition of multiphase chemistry would be larger than coatings of α-pinene SOA.⁹³ In the Amazon rainforest where average RH was higher than 80%, SOA particles were liquid, suggesting limited diffusion will not inhibit multiphase chemistry if RH>80%.⁹⁴ Even though the average RH is close to 80% for both the Amazon and SE-U.S., SI Figure S8 and previous measurements show that the median daytime RH in those regions is consistently below 70% (40–50% during the dry season in the Amazon) during the period of maximum diurnal IEPOX-SOA production.^93,95 Hence, diffusion of gaseous molecules limited by the formation of an OS coating will be more likely to affect multiphase chemistry during daytime in isoprene-rich environments, reducing heterogeneous SOA formation from compounds that are generated by photooxidation reactions.

Change in morphology is further supported by measuring the interfacial tension (IFT) depression of the major IEPOX-SOA products in microfluidic platform experiments. Surface tension and IFT are proxy measures of surface concentrations, as tensions will decrease with increased bulk-to-surface partitioning of surface-active components in the aqueous droplets. Each microfluidic measurement consumes 250–300 μL of aqueous solution, yielding 20–30 droplets. The IEPOX-OS droplet interfacial tensions reported in this work were statistically averaged and reported as the circle markers in Figure 4A. Additionally, Figures S2 and S9 show 2-methyltetrols and IEPOX-OS in methanol aqueous solution and 2-methyltetrols in pure water, respectively. IFT depression was observed in all cases (Figure 4A, Figure S2 and Figure S9), and IFT was lower when AS was present. In fact, lowering of the IFT in salty solutions indicates potential salting out ^96,97 of the organics due to an enhancement of organic activity, driving more organic molecules to the surface. The salting out effect subsequently alters the SOA physicochemical properties by changing morphology to a core-shell structure rather than a homogeneously mixed sphere.⁹³ Salting out effects are quantified by combining the Setschenow equation and a two parameter surface and IFT model.⁹⁸ The Setschenow constant (K_s) indicates salting out when positive for organic-inorganic aqueous systems. The curves in Figure 4A shows model treatment for IEPOX-OS in pure water and in salty water and predicts that IEPOX-OS has a propensity for salting out in the presence of AS, which helps to explain the dependence of shell thickness on the ratio of IEPOX:Sulf_inorg. IEPOX-SOA may alter the climate properties of aerosols, for example, enhancing CCN activity by suppressing the surface tension to enhance cloud droplet formation from aerosols in organosulfur-rich particles, causing larger droplets to form before and during cloud activation.⁹⁹

Figure 4.

(A) Measured and modeled interfacial tension (IFT) of IEPOX-OS. Microfluidic measurements of IFT (circles); two parameter model treatment of SOA in pure water (solid line), using a binary model ^63,100 ; and model treatment of SOA in ammonium sulfate (AS) solution (dashed line), using an adapted form of the binary model adapted for salt-containing organic aqueous solutions that includes known organic model parameters from the solid line and the Setschenow constant as the single adjustable model parameter (dashed line). (B) IEPOX reactive uptake coefficient (γ_IEPOX), obtained from flow tube experiments performed at the University of Washington, on aqueous ammonium bisulfate particles as a function of atmospheric equivalent exposure time defined as the length of the time that an aerosol is exposed to IEPOX gases, assuming gas-phase IEPOX concentration of 1 ppb and an aerosol surface area density of 300 μm² cm⁻³. The atmospheric equivalent exposure time is obtained from the experimental reaction time by multiplying the ratio of experimental-to-ambient concentration of IEPOX (=75) represents an upper limit.

Figure 4B illustrates the limiting effects of IEPOX-SOA products on reactive uptake by aerosol which can most likely be ascribed to a slight decrease in acidity and transformation to a core-shell morphology as a result of the salting out effect induced by organosulfur compounds. The dramatic decrease of γ_IEPOX with increasing atmospheric-equivalent exposure time, highlights the profound modification of IEPOX-SOA on aerosol reactivity, explains the presence of residual IEPOX in the gas phase measured by CIMS (SI Figure S10), and the stabilization of Sulf_inorg (Figures 1A and B) in our laboratory experiments. While previous studies have shown that other organic coatings tend to reduce multiphase chemical processes,^25,34,93 the results presented in Figure 4B provide direct evidence that uptake of IEPOX by sulfate aerosols has a self-limiting effect and indicate that the formation of a viscous IEPOX-SOA coating likely prevents further SOA formation in the ambient environment, primarily during daytime when RH is lower than 80%.

In sum, this study demonstrates that acid-driven multiphase chemistry of IEPOX converts a significant fraction of Sulf_inorg to organosulfur within a range of IEPOX:Sulf_inorg ratios relevant to most isoprene-dominated environments. We further demonstrate through laboratory and field measurements, the substantial conversion of sulfate, underlining the major role of IEPOX in controlling the chemical form of S(VI). Retrospective examination of field data in the SE-U.S. consistently shows that the contribution of organosulfur to S(VI) has been increasing with declining Sulf_inorg. The measured filter-based organosulfur fraction for Amazon, however, is not as much as that determined by the laboratory experiments conducted at the relevant ratios. There are two possible reasons. First, we conducted experiments mimicking both regions under the same initial condition of aerosol acidity and therefore the observed extent of conversion (up to ~90%) for Amazon relevant ratios likely represents the upper bound in the atmosphere. Second, collocated measurements of gas-phase IEPOX have rarely been reported for this area. The actual IEPOX:Sulf_inorg could have been lower than the ratios tested in the laboratory mimic of Amazon. This also calls for the need to enrich the inventory of gas-phase IEPOX measurements to better constrain IEPOX:Sulf_inorg in regions such as Amazon. From chamber experiments, the organosulfur compounds are shown to impact aerosol acidity (up to 9 pH unit increase compared to the conventional case, where organosulfur compounds are not considered) by our thermodynamic model under assumptions similar to those widely used in models. A similar approach could be adopted in updating current thermodynamic models to assess this impact on predicting pH of ambient SOA. IEPOX reactive uptake results in a core-shell morphology supported by microscopic imaging and biphasic microfluidics. Through non-oxidative chemical processes, IEPOX-OS undergoes transformation to as yet uncharacterized organosulfur compounds, together with highly viscous IEPOX-OS, impeding further reactive uptake of IEPOX. Not examined in this work, though, is how the core-shell morphology may affect the prediction of aerosol acidity. It is obvious that the core and shell need to be treated separately in pH calculations. This warrants a systematic investigation in a future study. Worth noting is that the findings regarding IEPOX-OS and its implications for aerosol physicochemical properties have not been included in the presented global model for calculating the global IEPOX:Sulf_inorg. Overall the shift in sulfate distribution within aerosols, changing physicochemical properties, and decreasing multiphase reactivity caused by IEPOX multiphase chemistry have yet to be updated in the regional- and global-scale models to properly predict IEPOX-SOA formation and its impact on air quality and climate. In isoprene-dominated areas, IEPOX-OS, and potentially other biogenic/anthropogenic OS, likely govern the physicochemical properties of aerosol as well as the distribution of inorganic species such as sulfate or ammonium. Consequently, aerosol growth, multiphase reactions, including aging and reactive uptake of other species, and CCN activity change as surface tension, acidity, hygroscopicity and viscosity are modified. These changes could greatly impact atmospheric composition of biogenic SOA formed over isoprene-dominated areas. Hence, changes in SO₂ emissions at different locations around the world over time may have implications for the physicochemical properties of biogenic-derived SOA.