Evidence of Organonitrate Formation at a High Altitude Site, Mahabaleshwar, during the Pre-monsoon Season

The presence of organonitrate and organosulfate was found at Mahabaleshwar, a high altitude site, during the pre-monsoon season of 2016. A Time of Flight Aerosol Chemical Speciation Monitor (ToF-ACSM) was used to measure the organic and inorganic components of non-refractory particulate matter (NR-PM1) aerosol. Positive Matrix Factorization (PMF) was performed on the (i) organics mass spectra of the aerosol (PMFOA), (ii) organics mass spectra merged with inorganics (PMFOA+IOA) and (iii) integrated mass spectra of organics with NO and NO2 ions (PMFOA+NOx) to derive the chemical information on organonitrate and organosulfate. The results of PMFOA were used as a reference for validating the factors obtained through the PMFOA+IOA and PMFOA+NOx results. The analysis of PMFOA resolved four PMF factors: hydrocarbonlike OA (HOA); biomass burning OA (BBOA); oxygenated OA-1 (OOA-1) and OOA-2. The analysis of PMFOA+IOA identified two additional inorganic factors: sulfate organic aerosol (OA) and nitrate OA. Sulfate OA and nitrate OA contributed 36% and 6%, respectively, to the total aerosol mass. Although both originated as secondary organic aerosol, they displayed different diurnal profiles. The results of PMFOA+NOx were used for the quantification and apportionment of nitrate aerosol in two forms, organic nitrate and inorganic nitrate, which contributed 38% and 62%, respectively, to the total nitrate aerosol mass. The diurnal variation in organic nitrate highlights photochemical oxidation and nocturnal oxidation by the nitrate radical as the two major sources. This source apportionment study using a combined (organic and inorganic) dataset provides new source factors and improves our understanding about the sources and chemical nature of submicron aerosols in the atmosphere. However, uncertainties in the quantification of organosulfate remain a limitation.


INTRODUCTION
Organic aerosol (OA) makes up a substantial fraction of tropospheric fine (submicron) particulate matter, accounting for 20-90% of the total aerosol mass.This important component of submicron aerosol is originated from a wide variety of natural (wildfires) and anthropogenic sources (fossil fuel combustion and biomass burning).A significant fraction of OA is also formed as a secondary organic aerosol (SOA) either by the oxidation of volatile organic compounds or aqueous-phase oxidation (Kroll and Seinfeld, 2008;Lukács et al., 2009;Hallquist et al., 2009;Lim et al., 2010).Of the known SOA precursors, isoprene and monoterpenes contributes significantly to the global SOA budget, owing to their large global emission rates (Guenther et al., 1995) and high reactivity with atmospheric oxidantshydroxyl radical (OH), ozone (O 3 ), and nitrate radical (NO 3 ).Several studies have tracked down the nature of origin and processes of OA through fast acquisition of OA mass spectrum using Aerosol Mass Spectrometers and/or Aerosol Chemical Speciation Monitors and analyzing them with positive matrix factorization (PMF) (Paatero and Tapper, 1994;Zhang et al., 2005;Mohr et al., 2012;Carbone et al., 2013;Aurela et al., 2015;Xu et al., 2015a;Iwamoto et al., 2016;Sun et al., 2016).However, most of the previous studies have performed PMF on the OA spectra only, leaving behind the inorganic species (sulfate, nitrate, chloride and ammonium).Recently, a study by Sun et al. (2012) performed PMF analysis on the merged spectra of organic (OA) and inorganic aerosol (IOA), i.e., PMF OA+IOA and identified two additional factors-sulfate-OA (organosulfate) and nitrate-OA (organonitrate).Since the organic and inorganic species undergo different formation mechanisms in the atmosphere, it is believed that the OA factors associated with inorganics may provide insights into the intrinsic relationship between organic and inorganic species.
Both laboratory-generated and air sampling studies have provided the evidence of presence of organosulfate (Iinuma et al., 2007;Liggio et al., 2007;Surratt et al., 2007;Surratt et al., 2008) and organonitrate (Nielsen et al., 1998;Singla et al., Aerosol and Air Quality Research, x: 1-11, xxxx 2 Reemtsma et al., 2006;Ng et al., 2007;Fry et al., 2009;Stone et al., 2009;Bruns et al., 2010;Day et al., 2010;Lim et al., 2016) as an important component of SOA.Due to the lack of measurement approaches, the formation mechanisms and aging of organonitrate (ON) and organosulfate (OS) remains unexplored and is largely ignored in models.Organonitrates have been identified as significant component of NO y in the gas phase (Nielsen et al., 1998;Day et al., 2003;Perring et al., 2009) and is likely to affect the nitrogen cycle and ozone production (Rosen et al., 2004;Mao et al., 2012;Perring et al., 2013).ON are formed in the atmosphere either through photochemical (OH-initiated) or nocturnal (NO 3 -initiated) oxidation reactions of biogenic and anthropogenic VOCs in the presence of NO x (Hallquist et al., 1999;Pitts and Pitts, 1999).NO 3 oxidation of isoprene and monoterpene have been identified as a significant source for ON (Ng et al., 2008;Fry et al., 2009) whereas photochemical oxidation of VOC produces minor ON (Atkinson and Arey, 2003).The low molecular weight ON product tends to remain in the gas phase while larger, multifunctional ON having lower vapor pressure undergoes partitioning and forms SOA.Despite lower ON yields during the daytime, model studies including ON production have reported that photochemically produced ON accounts for the significant fraction of SOA functionalization (Camredon et al., 2007).
Several studies (Romero and Oehme, 2005;Liggio et al., 2006;Reemtsma et al., 2006;Surratt et al., 2008) have reported the enhanced formation of organosulfate through oxidation of several biogenic VOCs including isoprene, limonene, α-pinene, β-pinene, in the presence of sulfate seed aerosol.Thus, the formation of organosulfate is influenced by both biogenic (BVOCs) and anthropogenic (which provides an acidic medium) emissions and it is identified as a secondary (or aged) product.Studies have also indicated the significant contribution of OS to SOA (Gao et al., 2006;Gilardoni et al., 2007;Altieri et al., 2009).Considering the aged nature, organosulfate may play an important role in aerosol microphysics.
So far, the studies on PMF analysis of OA mass spectrum have been performed either at an urban or a rural site, dominated either by anthropogenic or biogenic emissions (Ng et al., 2010;Carbone et al., 2013;Aurela et al., 2015).Aurela et al. (2015) performed PMF on the organic mass spectra for residential area, highway, curbside and described OOA as the major fraction of submicron aerosol.A study by Hao et al. (2014) performed PMF analysis in the mixed environment, i.e., at a semi-urban site, which is influenced by both biogenic and anthropogenic sources.Weber et al. (2007) have also observed the significant impact of anthropogenic activities on biogenic emissions.In the present study, an attempt has been made to comprehend the likely interactions of organic species with inorganic species of the aerosol.The study was carried out in the premonsoon season of 2016 at High Altitude Cloud Physics Laboratory (HACPL), Mahabaleshwar.Mahabaleshwar is a rural tourist town in the Western Ghats of India, where ~80% of the land is covered with forest.Being a tourist place and forest covered site, HACPL observes both fresh anthropogenic emissions (vehicular, fossil fuel combustion, etc.) and biogenic emissions.This study mainly focuses on the identification of organic sulfate and organic nitrate in the ambient air at HACPL during the pre-monsoon season.The quantification of organic and inorganic nitrate has been performed using state of the art PMF methodology.This study is expected to not only strengthen our understanding about the sources, processes and associated chemical characteristics of OA in the ambient air but also improve the chemical composition derived hygroscopicity value (к) of aerosols using the knowledge of two forms of sulfate and nitrate.The hygroscopicity parameter is generally calculated by considering the measured sulfate and nitrate as purely inorganic in origin (hence overestimation) and thus assigns a high hygroscopicity value (~0.6-0.7).This uncertainty in chemically derived hygroscopicity value can be reduced by considering the actual concentration of inorganic species.

Measurement Site
The observatory, HACPL at Mahabaleshwar (17.92°N, 73.65°E), is situated at a height of 1378 m above mean sea level (a.m.s.l.).The map in Fig. 1 gives a picturesque description of the sampling site.Mahabaleshwar serves as a tourist destination during the pre-monsoon season.The site is surrounded with dense vegetation, residential houses, hotels and a rural market.During the study period (March-May 2016), the site experiences high solar radiation (mean: ~471.4 ± 297.9 W m -2 ), high surface temperature (mean: ~26.3 ± 5.5°C), moderate relative humidity (mean: ~46.4 ± 8.2%) and moderate wind speed (mean: ~2.5 ± 1.6 m s -1 ).These meteorological parameters were measured using an Automatic Weather Station (AWS) at an interval of 1 minute.

Time of Flight -Aerosol Chemical Speciation Monitor
Time of Flight -Aerosol Chemical Speciation Monitor (ToF-ACSM) was used for measuring the submicron NR-PM 1 data at an interval of 10 minutes.This high resolution was averaged to 1 hour for the analysis.The chemical information of the mass spectra was extracted by following the protocol of Allan et al. (2004).The calibration and operation of ToF-ACSM is discussed elsewhere in detail (Ng et al., 2011;Fröhlich et al., 2013).In brief, the particle beam impacts on a resistively heated porous tungsten surface (T ≈ 600°C), flash vaporized into the non-refractory constituents that are ionized by electron impact (E kin = 70 eV) and detection by ToF analyzer on the basis of respective mass-to-charge ratios.The mass spectral matrices of organic and inorganic species were extracted for subsequent PMF analysis.The details on QA/QC of ToF-ACSM are given in Mukherjee et al. (2018).

Positive Matrix Factorization
PMF was used to deconvolve the mass spectral matrix into source related factors (Paatero and Tapper, 1994).In this study, Unit Mass Resolution (UMR) matrix (m/z = 12-220) and error matrix was generated for both organic and organic + inorganic species, by using ToFware (Igor Pro; Wavemetrics, Inc.; Oregon, USA), following the protocol of Ulbrich et al. (2009).The ions (m/z) were chosen up to 110 only for PMF analysis due to lower signal to noise ratio (S/N) at larger m/z (> 110).Ions with S/N ratio less than 0.2 were removed while the ions with 0.2 < S/N < 2 were weighted by a factor of 2. Furthermore, the particulate ions, H 2 O + , O + , HO + and CO + were also weighted to avoid the additional weighting by CO 2 + .The rotational ambiguity in both the runs was tested by varying the forcing parameter, f peak , between -1 and 1 with an increment of 0. + (m/z = 16) and NH 3 + (m/z = 17) for ammonium.The sum of chosen ions represents a major fraction of nitrate (95.3%), sulfate (62.5%), chloride (75.6%) and ammonium (99.9%) respectively.These fragment ions correlated strongly (r 2 ≈ 0.99; p-value = 0.005) with the mass concentration of respective inorganic species (Fig. S1).After PMF run, the 4-factor and 6-factor solution was selected as an optimum solution for OA mass spectra and merged mass spectra respectively.The optimal solution for PMF OA and PMF OA+IOA was determined by examining the scaled residuals of PMF fits, the change in Q/Q exp with increasing number of factors, factor correlations with external tracer, EC, and comparing the component mass spectra with reference mass spectra available in AMS spectral database (Ulbrich et al., 2009;Huang et al., 2010;He et al., 2011).The change in Q/Q exp , time series of change in Q/Q exp , scaled residual, factor correlations with external tracers and correlation of the resolved mass spectral profiles with reference mass spectra are provided in the supplementary material (Figs.S2-S4 and Table S1).An optimum f peak value of 0 was used for PMF analysis in both the cases.The PMF analysis of merged mass spectra yielded two additional factors-SO 4 -OA (organosulfate) and NO 3 -OA (organonitrate).

Quantification of Organic Nitrate
In general, the nitrate measured by ToF-ACSM accounts for total nitrate functionality (-ONO 2 ), which could arise from both organic and inorganic nitrates.So far, the direct measurements of organic nitrates are not available.However, a few studies (Bruns et al., 2010;Farmer et al., 2010;Sato et al., 2010) have suggested some indirect methods for quantifying organic nitrate in the atmosphere.Of the reported indirect methods, PMF method was used in this study to apportion the measured total nitrates.This method could unequivocally determine the relative contributions of organic and inorganic nitrates (Sun et al., 2012;Hao et al., 2014;Xu et al., 2015a, b).The PMF analysis of merged (NO 3 + org) mass spectra deconvolved a nitrate inorganic aerosol (NIA) factor along with other organic aerosol factors.The optimal solution for PMF analysis on the merged mass spectra was selected in a similar way as done with the PMF analysis of organic mass spectra.The selected factors were further validated by comparing the mass spectrum and mass concentration (Fig. 5) with the PMF org results (Fig. 3).The total nitrate functionality was apportioned into organic and inorganic signals with the following set of equations: where [OA factor] i is the mass concentration of i th OA factor and f NOi and f NO2i are the mass fraction of NO + and NO 2 + respectively in the i th OA factor.Since the mass spectrum of nitrate (measured by ACSM) was dominated with signals of NO x ions and the sum of which accounts for 95% to total nitrate mass, a factor of 0.95 was applied in Eq. (3).Finally, the nitrate functionality from inorganic nitrate was calculated by subtracting NO 3 org from the total nitrate, measured by ACSM.

Mass Concentration and Diurnal Variation of NR-PM 1 Species
With the mean mass concentration of 6.3 ± 5.1 µg m -3 , organics contributed ~61% to the total NR-PM 1 mass (10.3 ± 8.2 µg m -3 ).This was followed by sulfate, an inorganic constituent of NR-PM 1 with ~23% contribution and an average concentration of 2.4 ± 1.7 µg m -3 .Subsequently, ammonium, nitrate and chloride followed with percentage contribution of ~10%, ~5% and ~1% respectively.The mean mass concentration was measured as 1.0 ± 0.7 µg m -3 for ammonium, 0.6 ± 0.6 µg m -3 for nitrate and 0.1 ± 0.1 µg m -3 for chloride.Each aerosol species exhibited different mean diurnal pattern.Organics, sulfate and nitrate displayed a daytime build-up and evening high, whereas chloride showed a morning and evening high (Fig. 2).Since the diurnal pattern of organics is governed both by primary emissions (anthropogenic activities) and secondary formation (gas to particle partitioning) (Almeida et al., 2014;Xu et al., 2015a;Sun et al., 2016), the sources of organics were identified by performing PMF.The daytime build-up of sulfate and nitrate was attributed to their photochemical formation from precursor gases-SO x and NO x respectively.Sulfate also displayed high background concentration throughout the day.This concentration was likely attributed to the regional transport.Nitrate high during the evening hours was attributed to gas-particle partitioning under high relative humidity and low surface temperature conditions (Fig. 2(f)).The high concentration of chloride during morning and evening hours was ascribed to local anthropogenic activities (biomass and fossil fuel burning).It is important to note that all NR-PM 1 species reflected a significant peak during evening hours.This peak was attributed to the combined effect of source (or source processes) and low PBL height (Fig. S7).The diurnal pattern of ammonium represents the combined result of ammonium chloride, ammonium sulfate and ammonium nitrate.The amount of ammonium (NH 4 + ) required to achieve a charge balance with measured inorganic anions, SO 4 2-, NO 3 -and Cl -, was calculated and observed to be higher than measured NH 4 + (Fig. 2(e)).Few outliers deviating from one-to-one line of calculated vs measured NH 4 + were observed with the slope value of 1.12.The slope value greater than 1 is indicative of possible presence of organic associated nitrate, sulfate and chloride.This overprediction of NH 4 + was correlated with each inorganic anion.The satisfactory correlation of over-predicted NH 4 + with nitrate (r 2 = 0.51; p-value = 0.001), sulfate (r 2 = 0.43; p-value = 0.001) and chloride (r 2 = 0.40; p-value = 0.001) suggested relative presence of organonitrate, organosulfate and organochloride in the ambient air.
To study the possible influence of regional transport on the measured mass concentrations, the Potential Source Contribution Function (PSCF) analysis was performed (Fig. S5).PSCF plot shows that the site is influenced by the possible contribution from Mumbai during the study period.This regional influence was also reflected by the diurnal variation in mass concentration of organics and sulfate.Further, the hourly time series of NR-PM 1 species was studied (Fig. S6).The time series indicate different mass loading of NR-PM 1 species during the study period.
To foresee the effect of air-mass origin on observed mass concentrations, 3 time windows (Windows 1, 2 and 3) were selected randomly.Windows 1 and 3 exhibited comparatively low concentrations while Window 2 showed significantly high concentrations.The hourly backward trajectory analysis during the selected time windows revealed that the air-mass originated from the Arabian Sea (clean air mass) during Windows 1 and 3. On the other hand, air-mass originated mostly from the continent and carried pollutants along the trajectory pathway during Window 2.

Positive Matrix Factorization (PMF)
PMF: Organic Mass Spectra (PMF OA ) PMF analysis of the organic mass spectra yielded four source related factors (Fig. 3).The mass spectrum of Factor 1 was dominated with marker ions at m/z = 41 (C 3 H 5 + ), 55 (C 4 H 7 + ) and 57 (C 3 H 5 O + ).Based on the close resemblance of mass spectrum of Factor 1 with previously reported HOA (hydrocarbon-like OA) spectra by Zhang et al. (2005), Factor 1 was assigned as HOA.The morning and evening peak of HOA were related to vehicular emissions.The afternoon peak of HOA was related to cooking activities based on the fraction of mass sticks at m/z = 55 and 57.Since the mass spectrum of cooking OA (COA) is identical to HOA (Mohr et al., 2012;Sun et al., 2012) and they cannot be separated with UMR (unit mass resolution) data, some contribution from cooking emissions may be anticipated.The second factor showed morning and evening high with dominant marker ions at m/z = 60 (C 2 H 4 O 2 + ) and 73 ).This factor was identified as BBOA-biomassburning OA.The morning and evening peaks of BBOA were attributed to fresh anthropogenic activities (wood burning).Furthermore, the evening peaks of HOA and BBOA were also attributed to low PBL height.The two factors, HOA and BBOA, were validated with EC (an external tracer for anthropogenic activities; r 2 = 0.61, p-value = 0.003 with HOA and r 2 = 0.71, p-value = 0.002 with BBOA).
The Factors 3 and 4 were identified as OOA-1 (oxygenated OA-1) and OOA-2 (oxygenated OA-2), based on the fractional contribution of mass stick at m/z = 44 and O:C ratio.The mass spectrum of OOA-1 was identified with higher contribution from m/z = 44 (~0.31) and higher O:C ratio (~1.27) while OOA-2 was identified with lower contribution at m/z = 44 (~0.15) and low O:C ratio (~0.66).The O:C ratios were calculated by following the methodology of Bougiatioti et al. (2014).Based on the respective O:C ratios, OOA-1 was defined as an aged and processed aerosol and OOA-2 as freshly formed aerosol.The diurnal variation of OOA-1 showed daytime (12:00-16:00) maxima and evening high (18:00-21:00).The daytime high was likely due to secondary formation of aerosols under high solar radiation and temperature.The evening high was attributed to rapid oxidation of freshly emitted gas phase species to SOA under high relative humidity (~65-70%) and low temperature and low PBL conditions.Moreover, high background (1.3-1.8 µg m -3 ) of OOA-1 was also observed throughout the day.This background was probably due to the regional transport of processed/aged aerosols.OOA-1 showed good agreement with sulfate (r 2 = 0.92; p-value = 0.003) and therefore, OOA-1 was assumed to be mixed aerosol (both secondarily formed and regionally transported).Fig. 3. Mass spectral profile and diurnal variation by PMF analysis of OA mass spectrum (PMF OA ) at four-factor solution with f peak = 0. HOA, hydrocarbon-like organic aerosol; BBOA, biomass burning organic aerosol; OOA-1 and OOA-2, oxygenated organic aerosol 1 and 2.
OOA-2 was identified as IEPOX-OA based on fractional contribution of mass stick at m/z = 82 (C 5 H 6 O + ).This mass stick is related with isoprene degradation to SOA.The time series of m/z = 82 ion (r 2 = 0.72; p-value = 0.05) and OOA-2 (r 2 = 0.78; p-value = 0.05) correlated strongly with literature reported IEPOX-OA (isoprene epoxydiol-derived OA) factor (available in the AMS Spectral Database at http://cires.colorado.edu/jimenez-group/AMSsd/).On a diurnal scale, IEPOX-OA displayed daytime maxima, consistent with the isoprene chemistry and solar radiation cycle.IEPOX-OA also exhibited evening high due to low boundary layer height and high relative humidity conditions.The evening high may be attributed to gas particle partitioning under favorable meteorological conditions.In terms of relative contribution, OOA-1 contributed ~33% to the total OA mass followed by HOA (~32%), BBOA (~18%) and OOA-2 (~17%).

PMF: Organic Mass Spectra Integrated with Inorganics (PMF OA+IOA )
PMF analysis on the integrated mass spectra (organic and inorganic) yields 6 PMF factors.Fig. 4 gives the mass spectral profile and corresponding diurnal variation of these factors.Based on the mass sticks identified at m/z = 15 (NH + ), 16 (NH 2 + ), 17 (NH 3 + ), 30 (NO + ), 46 (NO 2 + ), 48 (SO + ), 64 (SO 2 + ), 80 (SO 3 + ), 81 (HSO 3 + ) and 98 (H 2 SO 4 + ), two factors were identified as sulfate OA (SO 4 -OA) and nitrate OA (NO 3 -OA).However, a factor related to ammonium chloride was not identified.This may be due to the lower concentration of inorganic chloride in the atmosphere and thus was difficult to be retrieved by PMF.However, some fraction of chloride (m/z = 35 and 36) was apportioned with BBOA, suggestive of common emission source (biomass burning).The results of PMF OA were compared with PMF OA+IOA .The time series and spectral profile of HOA and BBOA exhibited similar pattern (Figs.3-4), thereby indicative of minor influence of merged PMF analysis over these factors.However, slight differences were observed in the time series of OOA-1 and OOA-2.This may be probably due to the apportionment of some fraction of OOA-1 and OOA-2 to SO 4 -OA and NO 3 -OA.Since the chosen fragment ions of sulfate constitutes 63% of the total sulfate mass (measured by ACSM), we are reporting only the presence of ammonium sulfate and quantification of organic and inorganic constituents of sulfate remains a limitation in this study.Besides this, 95% contribution of NO x ions to the total nitrate mass enabled quantification of two forms (organic and inorganic) of nitrate.This was achieved by following the protocol outlined in Xu et al. (2015b) (see Section Quantification of Organic Nitrate) and will be discussed in Section Quantification of Organic and Inorganic Nitrate.
The fraction of mass sticks related to NH 4 + and SO 4 + accounted for 70% of the total mass of SO 4 --OA factor.This indicated that SO 4 --OA factor is primarily composed of ammonium sulfate.With high degree of oxidation (O:C ratio ≈ 0.7), SO 4 --OA factor displayed a relatively flat pattern along the day.This observation indicated the aged and processed nature of SO 4 --OA, likely contributed by the regional transport.Based on the (i) close resemblance with sulfate diurnal, (ii) high correlation with sulfate mass (r 2 = 0.98; p-value = 0.004) and CO 2 + ion (r 2 = 0.82; p-value = 0.004), this factor may be characterized as the feature of transported aerosol.Nevertheless, remaining 30% of sulfate Fig. 4. Mass spectral profile and diurnal variation by PMF analysis of OA mass spectrum merged with inorganics (PMF OA+IOA ) at six-factor solution with f peak = 0. HOA, hydrocarbon-like organic aerosol; BBOA, biomass burning organic aerosol; OOA-1 and OOA-2, oxygenated organic aerosol 1 and 2; nitrate-OA, nitrate-organic aerosol; sulfate-OA, sulfateorganic aerosol.related ions were found to form association with organics, as evident from the mass spectral profile of OOA-1 and OOA-2.It was inferred that the significant noontime peak of OOA factors may account for the photochemical formation of organosulfate, in agreement with the higher rate of sulfuric acid production.The apportionment of SO 4 + related ions in different factors indicated the sources of sulfate aerosol as (i) long-range transport and (ii) photochemical formation.
Another factor, nitrate OA, was dominantly composed of NH 4 + and NO 3 + related ions.With percentage contribution of 69% to the total nitrate mass, this factor was primarily identified as ammonium nitrate.The diurnal pattern of nitrate OA displayed increase in concentration during the evening hours (19:00-21:00) under enhanced relative humidity and low temperature conditions (Fig. 2(f)).This diurnal profile of nitrate OA synchronized well with the diurnal profile of equilibrium constant between particle phase nitrate and gas phase nitric acid, indicative of the formation of particulate nitrate (Dall'Osto et al., 2009;Petetin et al., 2016).However, lower concentrations were observed during other hours of the day.An appreciable fraction of NO + related ions were found to be associated with OOA factors.These OOA factors displayed a significant noon peak which suggests photochemically produced OOA factors as probable source of organic nitrate (Mather et al., 2004).With the percentage fraction of 36%, SO 4 --OA contributed majorly to the total aerosol mass.This was followed by OOA-1 (21%), HOA (17%), BBOA (10%), OOA-2 (10%) and nitrate OA (6%).

Quantification of Organic and Inorganic Nitrate
PMF analysis of OA mass spectra integrated with NO + and NO 2 + ions (PMF OA+NOx ) was performed to quantify the organic and inorganic components of nitrate.An additional factor, nitrate inorganic aerosol (NIA), was resolved while retaining the previously resolved PMF factors-HOA, BBOA, OOA-1 and OOA-2 (Fig. 5).The mass spectra and diurnal variation of PMF factors are given in Fig. 5 and Table S2 gives the correlation between mass spectra and time series of PMF factors resolved through PMF OA and PMF OA+NOx .A good correlation (r = 0.96-0.98)among different factors validated the individual PMF methodology.In terms of percentage contribution, OOA-1 (32%) contributed maximum to the total aerosol mass followed by HOA (29%), BBOA (17%), OOA-2 (16%) and NIA (6%).The NIA factor was characterized by the signals of NO + and NO 2 + at m/z = 30 and 46 respectively.On mass basis, these two ions formed 65% of this factor, with remaining species from organic fragments.The ratio of NO + /NO 2 + for NIA factor was observed as 1.28, which is very much close to 1.11-a value for NH 4 NO 3 determined during calibration of ACSM.Moreover, the mass concentration of NIA factor correlated strongly (r 2 = 0.96; p-value = 0.004) with the mass concentration of nitrate (as measured with ACSM), thus referring this factor as NH 4 NO 3 factor.Based on different physicochemical properties of organic and inorganic components of nitrate, NO x ions tend to form association with other organic factors as well.Consequently, 62% of NO x -ions were apportioned to NIA factor, 19% to OOA-2, 13% to OOA-1, with minor contributions to BBOA (4%) and HOA (2%) (Fig. 6).Based on this distribution of NO x ions with organic and inorganic factors, an estimation of nitrate in organic and inorganic form was made, referred to as organic nitrate and inorganic nitrate respectively.Considering only the OA factors, approximately 86% of Fig. 5. Mass spectral profile and diurnal variation by PMF analysis of OA mass spectrum merged with NO + and NO 2 + ions (PMF OA+NOx ) at five-factor solution with f peak = 0. HOA, hydrocarbon-like organic aerosol; BBOA, biomass burning organic aerosol; OOA-1 and OOA-2, oxygenated organic aerosol 1 and 2; NIA, nitrate inorganic aerosol.the organic nitrate was found to be associated with OOA factors.
Fig. 7 gives the diurnal variation of organic nitrate (ON) and inorganic nitrate (ION).The mean mass concentration of ON and ION was found to be 0.17 ± 0.1 µg m -3 and 0.29 ± 0.2 µg m -3 respectively.The two forms contributed 38% (organic) and 62% (inorganic) to the total nitrate mass.The diurnal variation of ION closely resembled the variation of nitrate aerosol with higher mass concentration than ON throughout the day.Lower ION concentration observed during the daytime with significant enhancement in the evening (19:00-21:00).This evening high may be due to gas-particle partitioning of HNO 3 under high relative humidity and low temperature conditions (Fig. 2(f)).This feature was evident from Fig. S8 as well where production of ION follows synchronous variation with Kp, an equilibrium constant of partitioning of ammonium nitrate.Kp was calculated using equation: where, T is the ambient temperature in Kelvin.k(298) = 3.35× 10 16 (atm -2 ), a = 75.11and b = -13.5.However, Kp and ION displayed different variation during the morning hours.This variation may be due to (a) less or no availability of precursor gases (sources were mostly inactive) and (b) comparatively high wind speed.Further ON aerosol observed a relatively flat pattern with higher concentration during daytime (14:00-16:00) and evening hours (20:00-22:00).Since formation of ON did not follow the Kp path, it is believed that the daytime high was due to photochemical production of ON from the oxidation of biogenic VOCs in the presence of NO x , while nocturnal oxidation of VOCs by nitrate radical led to nighttime concentration of ON.The sampling site, being a high-altitude forested site observes high solar intensity and moderate surface temperature, thereby making conditions favorable for the enhanced biogenic VOC emissions and hence a significant source for ON aerosol.Beside this, the NO + /NO 2 + ratio for nitrate associated with OOA-2 was observed as ~6.1, much higher than ratio of NH 4 NO 3 .The higher contribution of organic nitrate (38%) to total nitrate explains the statistically different values for NO x ratio.The value of ~6.1 lies close to value of organic nitrate produced from isoprene oxidation (~5; Fry et al., 2009;Bruns et al., 2010) and thus indicative of photochemical oxidation of isoprene as a probable source of organic nitrate at this site.On an average, organic nitrate makes up to 2% of the total NR-PM 1 aerosol with dominance of inorganic nitrate over organic nitrate.To validate the source apportionment result, we have calculated the organic nitrate concentration with conventional NO x ratio method assuming the ratio of NO + /NO 2 + (R ON ) first as 5 (formed from isoprene) and second as 10 (formed from monoterpene).The mean concentration of organonitrate (0.17 ± 0.1 µg m -3 ) obtained from PMF apportionment methodology lies between the mean organonitrate concentration obtained from NO x ratio method (0.19 ± 0.1 at R ON = 5 and 0.15 ± 0.1 at R ON = 10).Fig. S9 shows the correlation of calculated organonitrate by the two methods-PMF method and NO x ratio method.
The results are indicative of the presence of organonitrate and organosulfate in the ambient air.However, while dealing with UMR data of ACSM, some interference (~4%) by organic components were observed at m/z = 30 in this study.Therefore, this may lead to uncertainty in the quantification of organic nitrate and thus the mass concentrations of organic nitrate may be read as an upper limit.Since these aerosol components are the product of atmospheric processing (gas to particle conversion through subsequent adsorption), they are believed to be highly oxidized and aged in nature.These species are hence expected to act as cloud condensation nuclei after attaining optimum size under favorable meteorological conditions.However, the quantification of organosulfate warrants further study.

CONCLUSION
A ToF-ACSM was used to measure the NR-PM 1 aerosol at Mahabaleshwar during the pre-monsoon season of 2016.To obtain intrinsic information on the interactions between organic and inorganic components of the aerosol, PMF analysis was applied to both the organic mass spectra merged with inorganics (PMF OA+IOA ) and the organic mass spectra integrated with NO + and NO 2 + ions (PMF OA+NOx ).The results of PMF OA+IOA provided evidence of the presence of organosulfate and organonitrate based on two additional inorganic factors-sulfate OA and nitrate OA.The major sources of the sulfate OA were regional transport and photochemical formation, as indicated by diurnal variation in this aerosol, whereas the sources of the nitrate-OA were probably gas-particle partitioning of nitric acid and photochemically formed OOA.On average, sulfate OA and nitrate OA contributed 36% and 6%, respectively, to the total aerosol mass.The PMF OA+NOx analysis enabled us to quantify the organic nitrate aerosol, and we determined that the total nitrate functionality received a contribution of ~38% from organic nitrate and a contribution of ~62% from inorganic nitrate.The photochemical oxidation of VOCs by hydroxyl and the nocturnal oxidation by the nitrate radical were recognized as the major sources of organic nitrate during the daytime and late evening, respectively.This research ascertained the relative importance of both biogenic and anthropogenic emissions as sources of organosulfate and organonitrate at this site during the study period.

Fig. 1 .
Fig. 1.Map showing the Google Earth view of HACPL at Mahabaleshwar.

Fig. 2 .
Fig. 2. Mean diurnal variation of (a) organics, (b) sulfate, (c) nitrate, (d) chloride, and (f) temperature and relative humidity.The vertical bar represents standard deviation.(e) shows the correlation between predicted NH 4 + and measured NH 4 + .The dotted line represents one to one line.

Fig. 6 .
Fig. 6.Mass fraction of NO x ions: apportionment NO x ions among PMF factors.