Formation of Inorganic Sulfate and Volatile Nonsulfated Products from Heterogeneous Hydroxyl Radical Oxidation of 2-Methyltetrol Sulfate Aerosols: Mechanisms and Atmospheric Implications

Chemical transformation of 2-methyltetrol sulfates (2-MTS), key isoprene-derived secondary organic aerosol (SOA) constituents, through heterogeneous hydroxyl radical (•OH) oxidation can result in the formation of previously unidentified atmospheric organosulfates (OSs). However, detected OSs cannot fully account for the sulfur content released from reacted 2-MTS, indicating the existence of sulfur in forms other than OSs such as inorganic sulfates. This work investigated the formation of inorganic sulfates through heterogeneous •OH oxidation of 2-MTS aerosols. Remarkably, high yields of inorganic sulfates, defined as the moles of inorganic sulfates produced per mole of reacted 2-MTS, were observed in the range from 0.48 ± 0.07 to 0.68 ± 0.07. These could be explained by the production of sulfate (SO4•–) and sulfite (SO3•–) radicals through the cleavage of C–O(S) and (C)O–S bonds, followed by aerosol-phase reactions. Additionally, nonsulfated products resulting from bond cleavage were likely volatile and evaporated into the gas phase, as evidenced by the observed aerosol mass loss (≤25%) and concurrent size reduction upon oxidation. This investigation highlights the significant transformation of sulfur from its organic to inorganic forms during the heterogeneous oxidation of 2-MTS aerosols, potentially influencing the physicochemical properties and environmental impacts of isoprene-derived SOAs.


Section S1: HILIC/ESI-HR-QTOFMS analysis for characterization of OSs.
A 50 μL aliquot of each PILS sample was drawn and diluted in 950 μL acetonitrile (ACN, HPLC grade, Fisher Scientific) after collection in order to achieve the solvent composition of the organic mobile phase.The aliquots were then immediately stored in the dark at −20 °C prior to HILIC/ESI−HR-QTOFMS analysis.2-MTS and OSs were characterized using an Agilent 6500 Series UPLC system coupled with an electrospray ionization (ESI) source and a Quadrupole-Time-of-Flight Mass Spectrometer (Agilent 6250).The aliquot was injected into a Waters ACQUITY UPLC ethylene bridged hybrid amide (BEH-Amide) column (2.1 × 100 mm, 1.7-µm particle size, Waters) at 35 °C.The mobile phases consisted of two sets of eluents: eluent (A) containing 0.1% (w/w) of ammonium acetate in H 2 O solution, and eluent (B) containing 0.1% (w/w) of ammonium acetate in an ACN-H 2 O solution (95:5 vol/vol).Both eluents were adjusted to a pH level at 9.0 using NH 4 OH (TraceMetal Grade, Fisher Scientific).The eluent gradient was 0% of eluent A for initial 4 min, then increased to 5.6% A over next 6 min and held constant for another minute, decreased to 0% from 11 min to 11.5 min, and held constant till 15 min.The ESI source within the mass spectrometer was operated in negative ion mode, and the mass spectra were recorded from mass-to-charge ratio (m/z) 60 to 1000 at high-resolution mode (4GHz).Data was analyzed by Mass Hunter Version B.06.00 Build 6.0.633.0 software (Agilent Technologies).
Detected OSs are summarized in Table S1.As a first approximation, an estimation for the abundance of these newly formed OS products relative to reacted 2-MTSs (in the term of OS yield) were also derived based on their signal intensities and corresponding responses factors (RFs) in HILIC/ESI−HR-QTOFMS analysis: where ∆ i[  ]  refers to detected change in the signal intensity (peak area) of a certain OS species (OS i ) at a given OH exposure, j;   is the RF of OS i during HILIC/ESI−HR-QTOFMS analysis.Considering the correlation between RFs of various OS standards and their retention times (RTs) elucidated in previous study, 1 the RF of specific OS species in this study was approximated to that of either methylsulfate or 2-MTS (Table S1), depending on which compound exhibited the closest RT.The quantification of 2-MTS and methylsulfate (a tracer in the standard) was based on the calibration curve of 2-MTS and methylsulfate standards in HILIC/ESI−HR-QTOFMS (methylsulfate sodium salt, Sigma Aldrich, Q100).S1).

Section S2: The generic reaction mechanism of heterogeneous • OH oxidation of organosulfates (OSs) involving the inorganic sulfates formation
Generally, the chemical transformation of organics can proceed both by the addition of a new functional groups into parent molecules (functionalization) and the fragmentation of the carbon skeleton via chemical bond scission (fragmentation).As shown in Scheme S1, heterogeneous • OH oxidation of OSs initiates with hydrogen atom abstraction from a C-H bond, leading to the formation of an alkyl radical (R • ), which can instantly react with an oxygen (O 2 ) molecule to form a peroxyl radical (RO 2 • ).The self-or cross-reactions of RO 2 • can result in two carbonyl products (R1), 2 10 an alcohol product and a carbonyl product (R2), 3 and α-hydroxylperoxy radical can undergo unimolecular HO 2 • elimination (R3). 4Simultaneously, self-or cross-reactions of RO

Section S3: IC analysis for quantification of inorganic sulfates
The amount of inorganic sulfate (SO 4 2− ) formed upon heterogeneous oxidation was quantified using the IC method.A 25 μL of each PILS sample was injected into an IC system (ICS 3000, Thermo Fisher) equipped with an AS11-HC guard column (IonPac, 2×50 mm, Thermo Scientific) and anion-exchange column (2×250 mm, Thermo Scientific).The temperature of the conductivity detector (Dionex) interfaced to the IC was set at 35 °C.The 35-min potassium hydroxide (KOH) eluent program utilized for the IC was as follows: KOH was increased from 1 mM to 30 mM KOH from 0 to 25 min, then ramped to 84 mM KOH from 25 to 30 min, decreased to 1 mM 10 until 30.1 min and held constant until 35 min.The quantification of eluent peaks was based on calibration chromatograms of ammonium sulfate standards (Sigma Aldrich, purity ≥ 99.0%).It is known that bisulfate ion (HSO 4 − ) is being converted into SO 4 2− upon mixing with the alkaline eluent. 10,11 s a result, the concentration of SO 4 2− quantified by the IC method represents a total amount of HSO 4 − and SO 4 2− produced upon heterogeneous oxidation at a given • OH exposure.In addition, peroxydisulfate ion (S 2 O 8 2− ) could be a potential product resulting from sulfate/sulfite radical reaction (RS4, Scheme S1).Our recent study has confirmed the S 2 O 8 2− formation in IC chromatograms as a result of the heterogenous oxidation of another organosulfur species (hydroxymethanesulfonate, (OH)CH 2 SO 3 − ) using 200 mM of NaOH as the eluent in 20 IC. 11 We acknowledge that the detection of peroxydisulfate (S 2 O 8 2− ) could provide a more comprehensive information of the distribution and evolution of aerosol sulfur mass during the oxidation of 2-MTS aerosols (Figure S4).However, the IC analytical method in this study was primarily employed for the quantification of inorganic sulfate and did not have the capability to detect the presence of S 2 O 8 2− .

Section S4: Oxidation kinetics of 2-MTS
Based on the aerosol composition data characterized at different extents of oxidation, 2-MTS was found to be oxidized by • OH at a significant rate (Figure 1a).Approximately 64% of 2-MTS was reacted at the highest • OH exposure.Heterogeneous oxidation kinetics can be quantified by measuring the decay of 2-MTS over the • OH exposure, and expressed by the effective second-order heterogeneous • OH rate constant, k: 10 where I is the signal intensity of 2-MTS at a given • OH exposure, I 0 is its signal intensity before oxidation, and [OH]×t is the • OH exposure.The k was fitted to be 5.36 ± 0.11 × 10 −13 cm 3 molecule −1 s −1 for 2-MTS aerosols with a surface-weighted diameter of 109 ± 2 nm, and was in a good agreement with our previous work (k = 4.9 ± 0.6 × 10 −13 cm 3 molecule −1 s −1 for an aerosol size of 119 ± 4 nm). 12Based on the kinetic data, the e-folding lifetime of 2-MTS against heterogeneous • OH oxidation,  = 1/[], is estimated to be 14.4 ± 0.3 days assuming the ambient • OH concentration of 1.5 × 10 6 molecules cm -3 .Given atmospheric aerosols of a comparable size that exhibit an average lifetime of 5-12 days against wet or dry deposition, 13 heterogeneous • OH oxidation of 2-MTS could be a comparable removal processes and warrants attention 20 during the simulation of chemical transport models to better predict the atmospheric abundance of 2-MTSs.
Section S5: Correction for inorganic sulfate formation from the 2-MTS oxidation

Impurities in synthesised 2-MTS standard
The 2-MTS standard (C 5 H 12 O 7 S − , counter ion NH 4 + ) investigated in this work was synthesized following previously published procedure 14 and its purity was determined to be 60.2% using proton nuclear magnetic resonance ( 1 H NMR) spectroscopy with impurities including 25.9% (w/w) of methylsulfate (CH 3 SO 4 − ), 4.5% of ammonium sulfate ((NH 4 ) 2 SO 4 , AS) and C 2 -C 4 OSs accounting for the residual mass. 12Here, the sulfate detected by IC before oxidation ([SO 4 2− -AS]) was attributed to the presence of (NH 4 ) 2 SO 4 as shown in Figure S3a.It is also acknowledged that the heterogenous • OH oxidation of methylsulfate can also lead to the inorganic sulfate formation. 8,10 owever, its existence in 2-MTS standard cannot fully account for the significant increase in inorganic sulfate signal observed by IC.Assuming a conservative sulfate yield of one, the amount of SO 4 2− formed from the oxidized methylsulfate contributed approximately 22.2% -28.7% of the total SO 4 2− detected by IC after oxidation (Figure S5).Despite the potential uncertainties, these results suggest the presence and further reactions of the trace impurities could not fully explain the observed inorganic sulfate formation (Table S2) and was corrected for the quantification of sulfates from 2-MTS oxidation using Eqn.S3: where the SO 2- 4 -corrected i , SO 2- 4 -IC i and SO 2- 4 -CH 3 SO - 4 i are the corrected inorganic sulfate formation resulting from 2-MTS oxidation, the total inorganic sulfate amount measured by IC and the inorganic sulfate formed from methylsulfate oxidation derived from its decay at a given • OH exposure, respectively.This corrected inorganic sulfate formation was used for the determination of sulfate yields.

Hydrolysis of OS species
It is noted that the acid-catalysed hydrolysis of OSs in aqueous phase can lead to the formation of polyols and H 2 SO 4 . 15,16 ecent research has indicated that the lifetime of 2-MTSs against hydrolysis is approximately 28 days in a neutral solution and 10 days in a 0.1 M sulfuric acid solution with a pH below 1. 17 In our experimental setup, the aerosol residence time in the OFR was maintained at approximately 128 s for all • OH exposure levels.After oxidation, the aerosol particles were collected onto PILS and subsequently analyzed within two weeks or less using IC and HILIC/ESI-HR-QTFOMS.Given that, it is expected that the hydrolysis of 2-MTS and newly formed OSs product would likely have a negligible impact on sulfate formation during such shorter timescales.In addition, the hydrolysis of OS species can lead to the formation of polyols.For instance, xylitol (C 5 H 12 O 5 ) can be formed through the hydrolysis of major OSs products (i.e., C 5 H 11 O 8 S − ).This xylitol has a low volatility with saturation vapor pressure of 1.71 ×10 −6 Pa, 18 comparable to that of organosufates 19 and thus tends to stay in aerosol phase after formation.Consequently, the absence of these C 5 polyol species and other alcohol products in our mass spectra could support the proposition that the hydrolysis of these OSs holds minor significance in the sulfate formation.More importantly, an obvious increase in the amount of inorganic sulfates with increasing • OH exposure levels, as demonstrated in Figure S4 and S5, is observed and likely a result of heterogenous reactions.Taken together, although we could not completely rule out the possibility of the formation of sulfate via the hydrolysis of OS products, the enhanced inorganic sulfates formation is anticipated to be mainly due to the heterogenous • OH reactions of 2-MTS.10 Measurement precisions for the concentration of species i (σ Ci ) are propagated from precisions of volumetric measurements, chemical composition measurements, and blank sample variability and sample repeatability. 20For simplicity, the following equations are used to calculate the uncertainty associated with our PILS sample solutions: = 0.05 ( 6) ) where B i = average amount of species i in blank samples B ij = the amount of species i found in blank sample j C i = the concentration of species i M i = amount of species i in sample solution n = total number of samples in the sum SIG Bi = the root mean square error (RMSE), the square root of the averaged sum of the squared σ Bij STD Bi = standard deviation of the blank samples σ Bi = blank precision for species i σ Bij = precision of the species i found in blank sample j σ Ci = propagated precision for the concentration of species i σ Mi = precision of amount of species i σ V = precision of sample volume V = sample volume (25 μL for IC analysis, 50 μL for LC/ESI-HR-MS/MS analysis) The precisions (σ Mi ) were determined from duplicate analysis of samples.When duplicate sample analysis is made, the range of results, R, is nearly as efficient as the standard deviation since two measures differ by a constant (1.128σBased on the calculation of sulfate yield (Eqn. S4) and the uncertainties propagated from measurement precisions, the uncertainty for sulfate yield can be 10 calculated from Eqn. S5.

Section S7: Formation of formic acid and glycolic acid
IC analysis from our prior study observed increased signal intensities of glycolate ((OH)CHCOO − ) and formate (HCOO − ) ions in the chromatograms after the oxidation of 2-MTS, 12 indicating the formation of formic acid (HCOOH, FA) and glycolic acid (C 2 H 2 O 3 , GA) during the heterogeneous • OH oxidation of 2-MTS as elucidated in Scheme 1.In addition, given the high concentration of • OH in the reactor, non-sulfated organic products (i.e., C 3 H 6 O 2 , C 4 H 8 O 3 and C 4 H 8 O 4 ) can undergo further heterogeneous oxidation and lead to the formation of FA and GA as higher-generation products.Among these organics, C 4 H 8 O 4 exhibits lower volatility (Table S3), making it more 10 susceptible to •

Section S8: The heterogenous reactivity and chemistry of 2-MTS in atmospheric aerosols
This work investigated the kinetic and chemistry for heterogeneous • OH oxidation of aqueous 2-MTS aerosol at a single RH (~60%).However, the difference in our laboratory and ambient reaction conditions may introduce discrepancies in observed reaction kinetic and sulfate yield for heterogenous • OH oxidation of 2-MTS in atmospheric aerosols.In the following discussion, we address the possible differences and their potential impacts on the heterogeneous reaction kinetics and mechanisms of atmospheric 2-MTS aerosol.

Low oxidant ( • OH) level
Firstly, the OFR experiments typically employ a higher concentration of oxidants to simulate long-term atmospheric oxidation processes within a shorter experimental timescale by assuming that reaction time and oxidant concentration are interchangeable. 7,21 n this work, the • OH concentration (5.86 × 10 10 molecules cm −3 to 1.54 × 10 10 molecules cm −3 ) was significantly higher than ambient • OH concentration of 1.5 × 10 6 molecules cm −3 , 22 which may introduce difference in the heterogeneous oxidation chemistry such as sulfate yield between our experiments and the real atmosphere.As shown in Scheme 1, the inorganic sulfate formation requires the formation of alkoxy radical (RO • ) though the self-or cross-reactions of peroxy radical (RO 2 • ) (R4).A recent study has suggested that branching ratios for RO • formation resulted from self-or cross-reactions of RO 2 • could be smaller under lower RO 2 • concentration. 23Given that the formation rate and concentration of RO 2 • would be lower under low • OH level conditions, it is expected that the yields of inorganic sulfate reported in this work would be higher than those observed in the atmosphere.Additionally, in the atmosphere, bimolecular reactions with HO 2 24 or NOx 25 may also compete with the self-or cross-reactions of RO 2 • although RO • is more likely formed from the RO 2 • + NO reactions and subsequently undergoes bond dissociation.Therefore, future laboratory or modeling investigations are highly desired to explore the effect of • OH concentration on the kinetic and inorganic sulfate formation of heterogeneous • OH oxidation of 2-MTS considering the presence or absence of HO 2 • or NOx.

Atmospheric humidity condition
In addition, the atmospheric humidity condition (i.e., relative humidity, RH) has been suggested to impact the heterogenous reaction rates of OSs, 8,10 primarily by influencing the aerosol liquid water content.Laboratory study has reported that aerosols containing the ammonium salt of 2-MTS can absorb or desorb water reversibly when the RH increase or decrease in the range of 10%-90% but did not show a distinct phase transition, 26 suggesting that these aerosols likely exist in an aqueous state when RH exceeds 10%.Regarding the RH effects on the heterogenous reactivity of 2-MTS, we anticipate two main influences.On one hand, under higher RH conditions, enhanced water uptake leads to a decrease in the bulk and surface concentration of 2-MTS and dissolved • OH.Consequently, the collision probability between 2-MTS and • OH in the bulk solution and/or on the aerosol surface decreases, thereby slowing down the overall oxidation rate.On the other hand, as RH decreases to a certain extent, the aerosol particles become highly concentrated.This high concentration may lead to a reduction in the reaction rate, as the viscosity of aerosols generally increases with solute concentration.The increased viscosity hampers the diffusion of 2-MTS within the aerosol and towards the aerosol surface for oxidation, thereby lowering the overall reactivity.As a result, the formation rate of inorganic sulfates and volatile non-sulfated products is expected to decrease or increase with the heterogenous reaction rate while the sulfate yield may be not significantly affected as it describes the formation of inorganic sulfates relative to the reaction of 2-MTS.These hypotheses can find support in our previous study which examined the effect of RH (from 75 to 85%) on the oxidative kinetics and sulfate yield of methylsulfate upon heterogeneous • OH oxidation based on experimental measurement and model simulation. 8,10 e found higher RH could lower the overall reaction rate between methylsulfate aerosol and • OH through dilution effect while the sulfate yield is not very sensitive to the RH change.Further study is desired to fully elucidate how and to what extent RH changes alter the heterogeneous reactivity of 2-MTS and sulfate yield.

Aerosol composition and physicochemical properties
Atmospheric aerosols are complex mixtures, encompassing organic compounds, inorganic salts, and numerous other species, and thus cannot be fully represented by a simplistic, binary experimental system.The presence of inorganic salts in atmospheric aerosol have been found can affect the heterogeneous reaction kinetics of organic species, including OSs. 27,28 Our prior study has found that the presence of ammonium sulfate (AS) in aqueous sodium dodecyl OS aerosols can significantly enhance the heterogenous • OH reaction rate by a factor of 3 as the mass ratio of AS to dodecyl sulfate increased from 0.0 to 0.77.Molecular dynamic (MD) simulations indicated that this enhancement in oxidation rate could be attributed to the preferential accumulation of ammonium ions (NH 4 + ) at the air-water interface, along with the strong attraction between NH 4 + and • OH, which facilitated the closer proximity of diffused • OH to the surface-active dodecyl sulfate ions.Despite this enhancement, reaction products remained largely consistent across the experimental aerosol compositions, suggesting a negligible impact of AS on the reaction mechanisms.Given the surface-active nature of OS species, 29 a similar reactivity enhancement is anticipated for other OSs in the presence of cations exhibiting similar surface affinity and a pronounced attraction to • OH as NH 4 + .
Additionally, the presence and abundance of inorganic salts can impact the ionic strength of atmospheric aerosols, thereby potentially impacting the heterogeneous and multiphase reactions.A few laboratory studies have found that high ionic strength condition can accelerate the rate of photochemical degradation of many organic compounds within aerosol and the reactive uptake of ozone on organic aerosol surface. 30,31 his may be partly explained by that for reaction system containing an ion and a neutral reactant, the reaction rate constant can increase with the increase of ionic strength, known as the primary kinetic salt effect. 32Taken together, the types and concentration of inorganic ions (e.g.ionic strength) within aerosol could potentially alter the overall heterogeneous kinetics of OSs while it remains an open question whether the salts alter heterogeneous reactivity chemically and thus warrant future investigations.
Furthermore, it is noteworthy that the heterogeneous • OH oxidation process of 2-MTS can have an impact on the physicochemical properties of 2-MTS aerosols.This impact arises from both the ongoing conversion of 2-MTS to inorganic sulfate, considering that sulfur in its inorganic and organic form exhibit distinct properties 26, 33- 35 , and the evaporation of non-sulfated organic products during the oxidation process.For instance, compared to inorganic sulfates, 2-MTS exhibits lower hygroscopicity 26 as evidenced by measured volume growth factor (G f ) at 60% RH for aerosols containing ammonium sulfate (G f =~2) or 2-MTS ammonium salt (G f = 1.13).Therefore, the aerosol water uptake behaviour is expected to be enhanced upon the oxidation of 2-MTS aerosol, which may potentially impact the oxidation rate to some extent.Besides, considering the comparable pKa value of 2-MTS in its neutral form (C 5 H 12 O 7 S, pKa = -2.37) 19to sulfuric acid (pKa = -3.0),aerosols composed of 2-MTS anion (C 5 H 11 O 7 S − ) and its counterion NH 4 + prior to oxidation are expected to be slightly acidic.Upon the formation and dissociation of inorganic sulfates as well as the subsequent dilution resulted from enhanced water uptake, a mild decrease in aerosol pH is anticipated.The increased acidity may enhance the evaporation of low volatility organic acids products 36 and ultimately affect the aerosol mass loading.Consequently, the resulting variation in the physicochemical properties of 2-MTS aerosols may subsequently exert an influence on its heterogeneous reactivity.
Overall, we acknowledge that discussion provided above serves the purpose of gaining further qualitative insights into the potential factors driving the differences in reaction kinetics and sulfate yield observed in this experimental study compared to those in atmospheric aerosols.Further laboratory and modelling studies under more atmospherically relevant conditions are necessary to specifically address the effects of reaction conditions and aerosol properties on the heterogeneous reactivity and chemistry (e.g., sulfate formation) of 2-MTS and other isoprene-derived organosulfates.c This indicates whether the OS detected in this study.Some of our previously detected OSs were missing in this study, probably due to the low concentration of these minor OS products in diluted solutions collected by PILS.d The ranges of OS yield upon oxidation were calculated using Eqn.S1.
e The quantification of this OS species for yield determination used the response factor 10 of 2-MTS (RT = 7.17 and 8.86 min) in the HILIC/ESI-HR-QTOFMS run as an approximation.
f The quantification of this OS species used the response factor of methylsulfate (RT = 2.03 min) in the HILIC/ESI-HR-QTOFMS run as an approximation.e These values refer to the mole fraction of specific organic species remaining in the 10 aerosol phase based on equilibrium gas-aerosol partitioning calculations 40 using the measured aerosol mass loading (300-400 µg m -3 ).These values were calculated by assuming a neutral solution condition and thus can serve as upper limits for aerosolphase fraction of organic acids as their evaporation is generally enhanced with increasing acidity 36 .

Figure S2 .
Figure S2.The extracted ion chromatograms (EICs) detected by HILIC/ESI-HR-QTOFMS in negative ion mode, and the formation evolution profile of major C 5 OSs formed upon heterogeneous • OH oxidation of 2-MTS (only major OS products with 10 yield larger than 0.001 are presented in this figure, other OS products can be found in TableS1).
Scheme S2.A schematic diagram of experimental setup and chemical analysis for the heterogeneous • OH oxidation of 2-MTS aerosols.

Figure S4 .
Figure S4.The sulfur mass distribution and evolution of sulfur-containing species including 2-MTS, methylsulfate, OSs products and inorganic sulfates upon the oxidation of 2-MTS aerosols (The blue bars represent the sulfur mass attributed to 2-MTS molecules, and the sulfur mass associated with OS products is determined based on the estimated abundance of OSs, calculated using the methodology outlined in Section S1).

Figure S5 .
Figure S5.The extracted ion chromatograms (EICs) of methylsulfate from HILIC/ESI-HR-QTOFMS analysis in negative ion mode before (a) and after (b) heterogeneous • OH oxidation (RT = 2.03 min); (c) the normalized decay of methylsulfate upon heterogenous • OH oxidation; (d) the total amount of SO 4 2− at different • OH exposures measured by IC (pink bar) and the amount of SO 4 2− formed from heterogeneous • OH oxidation of methylsulfate (purple bar).
OH oxidation.Take the C 4 H 8 O 4 as an example, 3 carbon sites (carbon 1, 3, 4) of the C 4 H 8 O 4 are available for hydrogen abstraction by • OH to initiate the oxidation process.As shown in Scheme S3, both FA and GA can be formed during the further • OH oxidation of C 4 H 8 O 4 .In a similar manner, the • OH oxidation of C 3 H 6 O 2 , C 4 H 8 O 3 can contribute to the production of FA and GA.Therefore, the absence of these non-sulfated organic products in the HILIC/ESI-HR-QTOFMS analysis, along with the enhanced production of FA and GA detected in the IC analysis, could be partly explained by the occurrence of further • OH oxidation of these non-sulfated organic products leading to higher generation products.20 Scheme S3.Proposed mechanisms for formic acid (CH 2 O 2 ) and glycolic acid (C 2 H 4 O 3 ) formation resulted from further heterogenous • OH oxidation of non-sulfated organic products (i.e., C 4 H 8 O 4 ).

Figure S6 .
Figure S6.The normalized decay of aerosol mass loading and surface-weighted diameter measured by SEMS upon heterogeneous • OH oxidation of 2-MTS aerosols, starting from the initial mass loading of 393 ± 6 µg m −3 and diameter of 109 ± 2 nm before oxidation.Here the aerosol mass loading was monitored assuming a unit density of 1 g cm −3 for 2-MTS aerosols upon oxidation as a first approximation.

Figure S7 .
Figure S7.The correlation between aerosol mass loss (measured by SEMS) and sulfate 10 formation (measured by IC) resulted from the heterogeneous • OH oxidation of 2-MTS aerosols.

a
This value was measured byKahlbaum (1894).37b  The saturation vapor pressures are estimated by EVAPORATION model38 at 293 K, which calculates the vapour pressure of an organic compound based on its molecular structure and functional groups.c This value was measured by Dauber (1989). 39d C* represents the effective vapor pressure. 2

Uncertainty for the sulfate yield
Mi = R).Based on the blank samples and duplicate samples, coefficients needed for determining uncertainty are given in following table:

Table S1 .
OSs formed upon the heterogeneous • OH oxidation of 2-MTS aerosols.
a These OSs were observed by HILIC/ESI-HR-QTOFMS after the heterogeneous • OH oxidation of 2-MTS aerosols in our previous filter-based study.12bVariousisomers are formed.Only one isomer was shown here for simplicity.
a The corrected inorganic sulfate from 2-MTS oxidation as a function of OH exposure was calculated using Eqn.S3.bThe sulfate yield was calculated using Eqn.S4. 10

Table S3 .
Volatility of non-sulfated products formed upon the heterogeneous • OH oxidation of 2-MTS aerosols.