Secondary Organic Aerosol Formation from Nitrophenols Photolysis under Atmospheric Conditions

Nitrophenols are important products of the aromatic compounds photooxidation and play a considerable role in urban chemistry. Nitrophenols are important components of agricultural biomass burning that could influence the climate. The formation of secondary organic aerosol from the direct photolysis of nitrophenols was investigated for the first time in a quartz glass simulation chamber under simulated solar radiation. The results from these experiments indicate rapid SOA formation. The proposed mechanism for the gas-phase degradation of nitrophenols through photolysis shows the formation of biradicals that could react further in the presence of oxygen to form low volatile highly oxygenated compounds responsible for secondary organic aerosol formation. The inhibiting effect of NOx and the presence of an OH radical scavenger on the aerosol formation were also studied. For 2-nitrophenol, significant aerosol formation yields were observed in the absence of an OH radical scavenger and NOx, varying in the range of 18%–24%. A gas-phase/aerosol partitioning model was applied assuming the presence of only one compound in both phases. A degradation mechanism is proposed to explain the aerosol formation observed in the photolysis of nitrophenols. The atmospheric impact of nitrophenol photolysis is discussed and the importance for atmospheric chemical models is assessed.


Introduction
A complex mixture of volatile organic compounds (VOC) of anthropogenic and biogenic origin is present in the atmosphere at different mixing ratios [1]. Atmospheric oxidants, temperature, and solar radiation drive the physical and chemical processes taking place in the atmosphere, which can be considered as a photochemical reactor. The reactions and processes occurring within the troposphere have an impact on climate, ecosystems, and human health, affecting mankind's habitability. A strong anthropogenic influence has occurred over the last decades on tropospheric gas and particle composition. Molecules present in the atmosphere may have a considerable effect on the environment even at very low concentrations. Once present in the atmosphere, VOCs can undergo chemical transformations into new, mostly higher oxidized compounds of lower volatility and bigger molecular mass [2,3].
Aromatic hydrocarbons are an important class of VOCs mainly present in the urban atmosphere as a result of anthropogenic activities [4]. They are responsible for secondary organic aerosol (SOA) the main degradation pathway for nitrophenols in the atmosphere [12]. Bardini et al. (2004) studied SOA formation in single photolysis experiments for 2-nitrophenol and a few of methylated isomers in the large EUPHORE chamber as part of a PhD work [39]. The study from Bardini was performed in the presence of a small initial amount of NOx resulting from chamber background and in the absence of an OH radical scavenger [40]. Martin-Reviejo and Wirtz (2005) have suggested nitrophenols as precursors for aerosol formation in the OH radical initiated photooxidation of benzene, toluene, and xylenes [17]. However, their key role in aerosol formation is still unclear.
To improve our understanding of the nitrophenol photolysis as a source of atmospheric HONO and further OH radicals, the influence of an OH scavenger and the irradiation intensity on SOA formation were investigated. Qualitative information regarding the aerosol yield was obtained using a gas/aerosol absorption model. Possible atmospheric implications of the observations are discussed. The present work will complete the investigations on the nitrophenol atmospheric behavior, adding important input on the contribution of this important class of semivolatile organic compounds to SOA formation in the urban area, highlighting the importance of studying new possible mechanisms leading to particle formation in the atmosphere.

Chamber Description and Equipments
Investigations on the SOA formation were performed in the QUAREC 1080 l quartz glass reactor described previously [41]. The reactor is built up by a connection of two quartz glass tubes with a wall thickness of 5 mm and an inner diameter of 47 cm. The reactor has a total length of 6.2 m. The tubes are closed at both ends with two enameled aluminum flanges. The pumping system used for evacuation to an end vacuum of 10 −3 mbar is ensured by a turbo molecular pump system (Leybold-Heraeus PT 450 C) connected to the reactor via a central flange ring. The reactor temperature is monitored by a PT-100 temperature sensor. Teflon blades mixing fans are used for mixing inside the reactor and sampling of gases and particles are performed with Teflon and steel sampling lines. The collection sampling tubes are connected to secondary organic aerosol measurement tools, gas monitors, and gas-chromatography analytical instruments. The inlets and outlets from both closing flanges are linked with specific analytical instrumentation for gas and particle monitoring. The compensation pressure inside the reactor is performed by a calibrated mass flow controller when sampling collection procedures require high volume samples.
The solar light is simulated inside the chamber by using 32 lamps (Philips TL05-40W super actinic) that emit in the range of 300-480 nm with maxima at 360 nm. The lamps are mounted around the reaction vessel and could be switched alternatively by packs of eight to ensure different levels of irradiation. All the experiments performed in the QUAREC chamber are temperature controlled at any level in the range of 283-313 K.
Gas-phase reactants and products were measured inside the reaction chamber employing an FT-IR (Fourier Transform Infrared) spectrometer placed at one end of the chamber and coupled inside the reactor through KBr windows via a gold-coated mirror system. The White-type, multiple-reflection mirror system (base path length 5.91 ± 0.01 m) drives the infrared beam into the reactor using 82 traverses, which is equivalent to a total optical path length of 484.7 ± 0.8 m. A Globar IR light source was used for the FTIR spectrometer. The spectral resolution of 1 cm −1 was used for the recorded spectra of the compounds in this study. The FT-IR spectrometer (NICOLET NEXUS) used in this work was equipped with liquid nitrogen cooled mercury-cadmium-tellurium (MCT) detector.
The formation of SOA from the photolysis of nitrophenols in the QUAREC chamber was investigated with a Scanning Mobility Particle Sizer (SMPS) system. The SMPS consists of an electrostatic classifier TSI 3071 A and a particle counter TSI 3022 A. Aerosol size distributions were measured between 10 and 1000 nm as a function of gas flow and impactor used. Small particles were counted Atmosphere 2020, 11, 1346 4 of 15 by using an ultrafine condensation particle counter (UCPC) TSI 3025 A. The UCPC counter was very useful to observe the rapid formation of small particles following nucleation in the SOA formation.

Experimental Methods and Procedures
All photolysis experiments were performed in the QUAREC chamber using synthetic air and pure nitrogen. Prior to the experiments, an FTIR spectrum and UCPC scan were recorded to test possible background levels of aerosols and the cleanliness of air added to the reactor. All tests showed virtually no particle formation after the photolysis of background air for at least 30 min, the time necessary for one experiment. The reactor was then partially evacuated and, with the reactor under reduced pressure (~400 mbar), the nitroaromatic compounds were introduced into the chamber. The nitrocompound addition to the chamber was performed by using a special steel-glass heated inlet system in a flow of nitrogen/air. The nitroaromatic concentrations were (2-10) × 10 13 molecule cm −3 , which correspond to mixing ratios of 0.8-4 ppmv (1 ppmv = 2.46 × 10 13 molecule cm −3 at 1 atm and 298 K). In order to determine the wall loss rate, before starting the photolysis of the nitrophenols, FTIR spectra were recorded over a period of 15 min in the dark. FT-IR spectra were recorded at a spectral resolution of 1 cm −1 continuously during the experiment by co-adding 128 scans per spectrum over a time period of 2 min. Prior to the start of the experiment, all computers connected to the instruments, used for recording data were timed in order to ensure a good data correlation.
Since it is known that the photolysis of nitrophenol is an important source of nitrous acid (HONO) in the gas phase [11,12] and HONO can photolyze in the atmosphere leading to further OH radical formation, isoprene was injected as OH radical scavenger in some experiments in order to suppress interferences by the reaction of OH radicals with nitrophenols. The scavenger concentration was varied between (10-20) × 10 13 molecule cm −3 . Up to 4 ppmv of NOx (NO with impurities of 2% as NO 2 ) was added in three additional experiments to check for a potential effect of NOx on the aerosol formation yields.

Results
The SOA formation from the photolysis of nitrophenols is investigated for the first time in the present study. The experiments focused on the formation of SOA from the photolysis of 2-nitrophenol; however, some experiments were performed also for 4-methyl-2-nitrophenol.
Results presented in Figure 1 show that no particles are formed in the dark as can be seen in the first five SMPS scans. Additional experiments photolyzing only synthetic air in the chamber for approximately 30 min exhibited only very minor particle formation with less than 200 particle cm −3 , which is negligible. Accordingly, particle formation shown in Figure 1 was solely resulting from the photolysis of nitrophenols.   Figure 1. Rapid aerosol formation was observed in the simulation chamber. Figure 2 also indicates that only a small amount of 2-nitrophenol needs to be photolyzed to initiate aerosol formation. From this figure, it has been observed that photolytic conversion of only 10-15 ppbV of nitrophenol was enough to form aerosols that were detected by the UCPC. Quick particles formation was observed from photolysis of nitrophenol concentrations varying between 0.4-1.2 ppmv. Less than 1 min of photolysis time was necessary to measure a considerable amount of particles, leading to 1.0 × 10 5 particle cm −3 , which is the maximum amount detected by the UCPC. The concentration of nitrophenols used for the aerosol mass was corrected for wall losses. In Figure 1, there is an evident decay of the aerosol concentration when the photolysis lamps were A fast formation of aerosols was detected just seconds after the photolysis lamps were switched on. This rapid formation was observed for both2-nitrophenolic compounds studied in this work. The first SMPS scans ranging from 10 to 750 nm particle size exhibited the formation of very small particles with a mean diameter of only 20 to 40 nm. Figure 1 shows as an example for 2-nitrophenol photolysis, the formation of such small particles, and their growth. Significantly larger particles were observed later in the photolysis experiment, pointing to further coagulation of particles over time. Figure 2 shows the concentration of 2-nitrophenol decreasing and the increase of the SOA mass concentration over time, according to the experiment presented in Figure 1. Rapid aerosol formation was observed in the simulation chamber. Figure 2 also indicates that only a small amount of 2-nitrophenol needs to be photolyzed to initiate aerosol formation. From this figure, it has been observed that photolytic conversion of only 10-15 ppbV of nitrophenol was enough to form aerosols that were detected by the UCPC. Quick particles formation was observed from photolysis of nitrophenol concentrations varying between 0.4-1.2 ppmv. Less than 1 min of photolysis time was necessary to measure a considerable amount of particles, leading to 1.0 × 10 5 particle cm −3 , which is the maximum amount detected by the UCPC.
The concentration of nitrophenols used for the aerosol mass was corrected for wall losses. In Figure 1, there is an evident decay of the aerosol concentration when the photolysis lamps were switched off. The decay of the aerosol concentration could be also influenced by electrostatic deposition of the particles on the reactor walls. The decay of aerosol concentration measured at the end of the experiment was used to correct the aerosol mass when the aerosol yield was calculated. Usually, a correction of 10-15% for the aerosol mass was applied in the present study.
Using the aerosol volume distribution and assuming unity for the aerosol density, the aerosol mass formed in the photolysis experiments was calculated and plotted against the amount of reacted nitrophenols, where the slope represents the SOA yield ( Figure 3). The aerosol mass linearly increased with increasing concentration of both 2-nitrophenol and 2-nitro-p-cresol (4-methyl-2-nitrophenol), and no big difference in the SOA yield was observed for the two nitrophenols in the examined concentration range.
2-nitrophenol needs to be photolyzed to initiate aerosol formation. From this figure, it has been observed that photolytic conversion of only 10-15 ppbV of nitrophenol was enough to form aerosols that were detected by the UCPC. Quick particles formation was observed from photolysis of nitrophenol concentrations varying between 0.4-1.2 ppmv. Less than 1 min of photolysis time was necessary to measure a considerable amount of particles, leading to 1.0 × 10 5 particle cm −3 , which is the maximum amount detected by the UCPC. The concentration of nitrophenols used for the aerosol mass was corrected for wall losses. In Figure 1, there is an evident decay of the aerosol concentration when the photolysis lamps were Atmosphere 2020, 11, x FOR PEER REVIEW 6 of 14 switched off. The decay of the aerosol concentration could be also influenced by electrostatic deposition of the particles on the reactor walls. The decay of aerosol concentration measured at the end of the experiment was used to correct the aerosol mass when the aerosol yield was calculated. Usually, a correction of 10-15% for the aerosol mass was applied in the present study.
Using the aerosol volume distribution and assuming unity for the aerosol density, the aerosol mass formed in the photolysis experiments was calculated and plotted against the amount of reacted nitrophenols, where the slope represents the SOA yield ( Figure 3). The aerosol mass linearly increased with increasing concentration of both 2-nitrophenol and 2-nitro-p-cresol (4-methyl-2-nitrophenol), and no big difference in the SOA yield was observed for the two nitrophenols in the examined concentration range.
The aerosol mass yields from 2-nitrophenol photolysis are shown in Table 1. Three experiments at various concentrations of 2-nitrophenol were performed in this study for each investigated aspect. The SOA yields in the absence of the OH radical scavenger and additional NOx in the reaction mixture varied between 18% and 24%. The yields are higher than those from 2-nitrophenol photolysis in the presence of the OH radical scavenger. Obviously, no SOA yield dependency was observed when the initial mixing ratio of 2-nitrophenol was varied from 0.49 ppmv to 1.23 ppmv.   The aerosol mass yields from 2-nitrophenol photolysis are shown in Table 1. Three experiments at various concentrations of 2-nitrophenol were performed in this study for each investigated aspect. The SOA yields in the absence of the OH radical scavenger and additional NOx in the reaction mixture varied between 18% and 24%. The yields are higher than those from 2-nitrophenol photolysis in the presence of the OH radical scavenger. Obviously, no SOA yield dependency was observed when the initial mixing ratio of 2-nitrophenol was varied from 0.49 ppmv to 1.23 ppmv. Table 1 shows the aerosol formation yields for three series of experiments as follows: (i) photolysis of 2-nitrophenol without NOx and OH scavenger, (ii) photolysis of 2-nitrophenol in the presence of OH radical scavenger, and (iii) photolysis of 2-nitrophenol in the presence of NOx. The highest aerosol yields were observed in the absence of the OH radical scavenger and NOx. This indicates Atmosphere 2020, 11, 1346 7 of 15 either an additional aerosol formation from the OH radical reaction with 2-nitrophenol or an inhibitory effect of the OH radical scavenger and NOx on the SOA formation. Interferences on the aerosol formation from 2-nitrophenol photolysis could result from the SOA formed in the OH radical initiated oxidation of 2-nitrophenol. The OH radical is formed in the reaction system by the photolysis of HONO, which is known to be a product of 2-nitrophenol photolysis [11]. Bejan et al. (2007) have reported a very slow reaction between 2-nitrophenol and OH radicals [12], whereas the reaction of OH radicals with isoprene is very fast (1 × 10 −10 cm 3 molecule −1 s −1 at 298 K) [42]. In order to suppress the reaction of nitrophenols with OH radicals and to investigate the OH radical effect on the SOA yields from 2-nitrophenol photolysis, isoprene (0.5-4 ppmv) was added as OH scavenger in a series of experiments. The isoprene added in these experiments was enough to scavenge >97% of the OH radicals produced during the photolysis. Figure 4 shows aerosol formation and growth during 2-nitrophenol photolysis and in the presence of isoprene as OH radical scavenger. The so-called banana type plot indicates the initial nucleation of the particles and further growing process of SOA. The particle size diameters increase during photolysis by uptake of freshly formed aerosols and condensable gas-phase products and/or reactants on the pre-existing SOA. At the end of the photolysis experiments, the size distribution and number concentration of particles were further measured to allow the calculation of the aerosol wall deposition rate in the dark. The decrease in particle concentration after photolysis is a result of combined processes, i.e., further aerosol coagulation and wall deposition.
SOA formation from 2-nitrophenol photolysis was almost fully suppressed in the presence of NOx. However, the UCPC instrument still showed the formation of a few hundred small particles, but it was not possible to calculate an aerosol yield.
Atmosphere 2020, 11, 1346 8 of 15 nucleation of the particles and further growing process of SOA. The particle size diameters increase during photolysis by uptake of freshly formed aerosols and condensable gas-phase products and/or reactants on the pre-existing SOA. At the end of the photolysis experiments, the size distribution and number concentration of particles were further measured to allow the calculation of the aerosol wall deposition rate in the dark. The decrease in particle concentration after photolysis is a result of combined processes, i.e., further aerosol coagulation and wall deposition. SOA formation from 2-nitrophenol photolysis was almost fully suppressed in the presence of NOx. However, the UCPC instrument still showed the formation of a few hundred small particles, but it was not possible to calculate an aerosol yield.

Expression for SOA Yield
SOA yields have been calculated as the fraction of reactive organic gas compound (ROG) converted to aerosol mass (M) as defined by Odum et al. (1996) [43]. The fractional aerosol yield is given by the expression: where ∆M 0 is the aerosol mass concentration produced by the photolysis of a known amount of nitrophenol, ∆ROG. It has been proposed that SOA could be formed by partition of compounds between gas and particle phase and may be represented by the expression [43]: where Y is the total aerosol mass yield obtained from the photolysis of nitrophenols in terms of the individual single product stoichiometric coefficient α i , the gas/particle partition coefficient K i , and the total aerosol mass M 0 . According to model suggested by Odum, Figure 5 shows the aerosol mass yield as a function of the SOA mass concentration for the photolysis of 2-nitrophenol and 4-methyl-2-nitrophenol, respectively. The lines in Figure 5 are the result of using the gas-phase partitioning model considering the presence of only one photolysis product in the gas and particle phase. The asymptotic aerosol yields approach values representing the aerosol masses rising infinitely. The stoichiometric α coefficient in the Odum's model representation strives to the experimentally determined aerosol formation yield. This indicates that the reaction product(s) formed in the photolysis process (has/have) extremely low volatility and will partition almost entirely to the particle phase. considering the presence of only one photolysis product in the gas and particle phase. The asymptotic aerosol yields approach values representing the aerosol masses rising infinitely. The stoichiometric α coefficient in the Odum's model representation strives to the experimentally determined aerosol formation yield. This indicates that the reaction product(s) formed in the photolysis process (has/have) extremely low volatility and will partition almost entirely to the particle phase.

Proposed Pathways to SOA Formation
In the present study it was not possible to identify the products from the photolysis of 2-nitrophenol because of the low conversion of 2-nitrophenol to photolysis products making their

Proposed Pathways to SOA Formation
In the present study it was not possible to identify the products from the photolysis of 2-nitrophenol because of the low conversion of 2-nitrophenol to photolysis products making their detection by using FT-IR spectrometry impossible. Additionally, ELVOC expected products from the photolysis of 2-nitrophenol are obviously not favored being present in the gas phase. However, a possible mechanism for SOA formation is suggested in Figure 6.
Atmosphere 2020, 11, x FOR PEER REVIEW 9 of 14 detection by using FT-IR spectrometry impossible. Additionally, ELVOC expected products from the photolysis of 2-nitrophenol are obviously not favored being present in the gas phase. However, a possible mechanism for SOA formation is suggested in Figure 6. The proposed mechanism takes into account the instantaneous formation of HONO and the biradical formation route as well as the possible ozonide formation route [11]. The proposed biradical formation, most probably very reactive species, will undergo further chemical reactions with O2 and isomerization processes, which could lead to the formation of acids as has been proposed previously by Alif et al. (1991) [37] for aqueous solutions (Figure 7). Another possible reaction pathway for biradicals is their reaction with the nitrophenol reactant, leading to compounds with higher molecular mass and extremely low vapor pressure. In the presence of oxygen, the possible higher oxygenated compounds formed as gas phase products with low volatility will favor particle formation. The proposed mechanism takes into account the instantaneous formation of HONO and the biradical formation route as well as the possible ozonide formation route [11]. The proposed biradical formation, most probably very reactive species, will undergo further chemical reactions with O 2 and isomerization processes, which could lead to the formation of acids as has been proposed previously by Alif et al. (1991) [37] for aqueous solutions (Figure 7). Another possible reaction pathway for biradicals is their reaction with the nitrophenol reactant, leading to compounds with higher molecular mass and extremely low vapor pressure. In the presence of oxygen, the possible higher oxygenated compounds formed as gas phase products with low volatility will favor particle formation. with O2 and isomerization processes, which could lead to the formation of acids as has been proposed previously by Alif et al. (1991) [37] for aqueous solutions (Figure 7). Another possible reaction pathway for biradicals is their reaction with the nitrophenol reactant, leading to compounds with higher molecular mass and extremely low vapor pressure. In the presence of oxygen, the possible higher oxygenated compounds formed as gas phase products with low volatility will favor particle formation.  [37] in the aqueous phase to explain SOA precursor formation from nitrophenols photolysis.
Nitronic acid formation from the photolysis of 2-nitrophenol by t hydrogen transfer from the hydroxyl functional group to the nitro group has been proposed by Bejan et al. (2006) [11] and Nagaya et al. (2006) [44]. Further, the biradical form resulted as a coproduct of HONO may be responsible for aerosol formation. In a study of the direct photolysis of 1-nitronaphtalene, Healy et al. (2012) suggested the formation of naphthoxy radicals and 2-nitronaphthol as gas-phase products. The evidence of a further self-reaction mechanism of naphthoxy radicals was obtained by using an aerosol time of flight mass spectrometer (ATOFMS) to study SOA composition. However, this study proposed also 2-nitronaphthol formation from nitronaphthalene direct photolysis. The 2-nitronaphthol photolysis could follow a photolysis mechanism as proposed in the present study in  [37] in the aqueous phase to explain SOA precursor formation from nitrophenols photolysis.
Nitronic acid formation from the photolysis of 2-nitrophenol by t hydrogen transfer from the hydroxyl functional group to the nitro group has been proposed by Bejan et al. (2006) [11] and Nagaya et al. (2006) [44]. Further, the biradical form resulted as a coproduct of HONO may be responsible for aerosol formation. In a study of the direct photolysis of 1-nitronaphtalene, Healy et al. (2012) suggested the formation of naphthoxy radicals and 2-nitronaphthol as gas-phase products. The evidence of a further self-reaction mechanism of naphthoxy radicals was obtained by using an aerosol time of flight mass spectrometer (ATOFMS) to study SOA composition. However, this study proposed also 2-nitronaphthol formation from nitronaphthalene direct photolysis. The 2-nitronaphthol photolysis could follow a photolysis mechanism as proposed in the present study in Figure 6 where the hydrogen transfer from the OH group to the nitro group is followed by HONO release and a biradical formation that easily could undergo dimmer formation by self-reaction [45].
The SOA yields for the compounds investigated are almost independent on concentration within the range of experimental errors, as shown in Figure 3. However, a small increase in the SOA formation yield can be seen in Figure 3 for the methylated compound and may be due to the increased reactivity of the methylated isomer.

OH Radical Influence on Aerosol Formation
Methylated 2-nitrophenols react with OH radicals with rate coefficients in the range (2.7-6.7) × 10 −12 cm 3 molecule −1 s −1 [12]. A series of experiments were performed in the presence of isoprene acting as OH radical scavenger to evaluate the effect of OH radicals on the SOA formation. Although particles formation has been reported for the photooxidation of isoprene [46][47][48], the tests performed in the QUAREC chamber have shown almost no aerosol formation at the isoprene concentrations used in the present study [49]. Up to 10 3 particle cm −3 were obtained by OH radical initiated photooxidation of isoprene over a period of 30 min reaction time. These particles measured with the UCPC counter were produced slowly and are much less than > 10 5 particle cm −3 extremely quickly formed from nitrophenol photolysis ( Figure 2).
The data from Table 1 show a considerable decrease in the SOA yields when the OH radical scavenger is present in the reaction mixture. There was an unexpected high concentration of OH radicals; however, the presence of the OH radical scavenger produced a decrease in the SOA formation yield either due to the suppression of the reaction of nitrophenols with the OH radicals or due to the interaction of the scavenger with the intermediates formed from the photolysis of nitrophenols. In the literature, an effect of an OH radical scavenger on SOA formation from alkenes ozonolysis has been reported. The SOA yields are influenced by organic peroxy radicals resulting from the reactions of the OH radicals (formed by alkenes ozonolysis) with the scavenger [50,51]. Keywood et al. (2004) emphasized the effect of acylperoxy radicals resulting from the ozonolysis of alkenes, to contribute significantly to the SOA formation yield [51].

NOx Effect on Aerosol Formation
The present study on the direct photolysis of nitrophenols showed a strong inhibiting effect on the aerosol formation under initial high NOx conditions. The chemical degradation mechanism of the photooxidation of aromatic compounds is known to be influenced by the initial NOx concentration in the reaction system. Both theoretical and experimental studies established the effect of NOx on the photochemistry of aromatic compounds [35,52] and, as a consequence, further the SOA formation [17,[53][54][55]. In the absence of NOx, gas phase chemistry will lead to the highly oxygenated products, which are known precursors for SOA. On the other hand, NOx reactions will lead to organic nitrates and peroxynitrates, which are more volatile and are not favorable to new particle formation [55]. In the present study, it was not possible to verify the effect of NOx on SOA formation at levels of NOx prevailing in the troposphere. The analytical techniques employed in the present study did not allow gas phase product identification and performing experiments under NOx mixing ratios of 10-50 ppbv, which are typical for the urban atmosphere. However, Bardini (2006), in their study, performed in a EUPHORE chamber, where backgrounds of low levels of NOx are usually present, found a substantial amount of aerosols under the low to moderate concentration of NOx.
In a photooxidation study of 3-methyl-2-nitrophenol under atmospheric conditions, Derpmann et al. (2013) has found, employing newly developed capillary atmospheric pressure electron capture ionization mass spectrometry technique, products that have analyte ions with m/z = 137 and m/z = 167, respectively [56]. Their research investigations have identified a product with m/z = 167 as 3-methyl-2-nitro-p-benzoquinone, while for m/z = 137, a molecular formula C 7 H 7 NO 2 is proposed. Although there is an evident inhibiting effect of SOA formation, it could only be speculated that the biradical compound formed through the photolysis of nitrophenols in the reaction with NOx could undergo formation of products with relatively high vapor pressure, which does not partition to particle phase. Also, the products that result under high NOx conditions could be photochemically inactive and to be unavailable in the SOA formation process. Further experimental work will be necessary to assess the effect of different NOx concentrations on the particle formation by direct photolysis of nitrophenols.

Conclusions
The formation of secondary organic aerosol (SOA) from direct photolysis of 2-nitrophenol and 4-methyl-2-nitrophenol was investigated in the present study for the first time. Rapid SOA formation has been observed for both nitrophenols investigated. No effect on the aerosol mass yield of the structural difference between methylated and non-methylated 2-nitrophenol has been observed. The direct photolysis of 2-nitrophenol in the absence of NOx and without OH radical scavenger in the reaction mixture indicated an aerosol formation yield in the range 18%-24%.
The proposed photolytic degradation mechanism leading to SOA involves biradical formation during photolysis, which could react further with oxygen forming peroxy radicals, which will produce further highly oxygenated compounds. These compounds have low volatility and could form SOA.
The presence of the OH radical scavenger and the effect on the SOA yield has also been investigated. The effect of the initial NOx concentration on the SOA yield was tested for 2-nitrophenol photolysis. The corrections of the wall deposition of nitrophenols were performed and photolysis rates were measured.
A slightly higher SOA yield were calculated for higher 2-nitrophenol initial concentrations. The SOA yields measured in the absence of the OH radical scavenger were almost double those measured when isoprene acting as OH radical scavenger was present. The scavenger was proposed to act for suppressing the reactivity of the organic radical and, hence, suggest an inhibitory effect on the aerosol formation. In the direct photolysis of nitrophenols at elevated high NOx conditions, it would become obvious that the SOA formation is strongly inhibited possibly due to the formation of organic nitrates that possess higher vapor pressure and thus do not favor the SOA formation. Methylated nitrophenol photolysis does not produce a considerable difference on the SOA yield related to those SOA yields obtained from 2-nitrophenol photolysis. A partition absorption model was used in the present study to provide additional qualitative information for the aerosol mass yield. The application of Odum's gas-phase partitioning model obtained the best fit of aerosol mass yield against SOA mass, assuming the presence of one aerosol precursor compound in both phases. The study proposed a photolysis degradation mechanism of nitrophenols to explain the experimental observations.
The present study has proven that the direct photolysis of nitrophenols is a potential source for aerosol formation in the atmosphere. Nitrophenols, present in the polluted atmosphere at mixing ratio almost of hundred pptv, are released during agricultural biomass burning. Their study becomes a priority to atmospheric research since their effect on climate and data provided by this study could be useful for atmospheric chemistry models. Further detailed investigations are needed to understand the inside mechanism of the nitrophenol photolysis. The study exhibited a very low reactant conversion through photolysis and a more sensitive technique will be necessary to help to identify the products formed during the photolysis of nitrophenol. The identification of intermediates responsible for aerosol formation could be possible by using different mass spectrometry techniques, helping to validate the proposed mechanism for the direct photolysis of these compounds.