Properties and Atmospheric Oxidation of Norpinic Acid Aerosol

Abstract

Norpinic acid is a major semi-volatile oxidation product of α-pinene and β-pinene, two of the most important biogenic atmospheric volatile organic compounds. In this study we characterized the physicochemical properties of norpinic acid aerosol using a variety of techniques, and we investigated its reaction with OH radicals. The Aerosol Mass Spectrometer (AMS) spectrum of norpinic acid was characterized by a pronounced peak at m/z 82 (C₅H₆O⁺), which can be used as its chemical signature. The measured density of norpinic acid particles was 1.3 g cm⁻³. Its saturation concentration at 298 K was estimated to be equal to 8.9 μg m⁻³ using thermodenuder measurements and 12.8 μg m⁻³ using isothermal dilution. Its vaporization enthalpy was equal to 71 kJ mol⁻¹. After reaction with OH radicals for an equivalent atmospheric period of 0.6–5 days under UV radiation and low RH, there were no noticeable changes in the AMS spectrum of the particles, while the wall-loss corrected mass concentration slightly decreased. This suggests that the atmospheric aging products of norpinic acid particles are quite similar to the parent molecule when measured by the AMS, and the aging reactions lead to a small change in particle mass concentration.

1. Introduction

Secondary organic aerosol (SOA) is one of the most important components of ambient organic aerosol (OA) [1]. On a global scale, the biogenic SOA contribution is estimated to be as much as 90% of the total SOA mass concentration [2]. Monoterpenes and, especially, α- and β-pinene are important biogenic SOA precursors [1,3,4]. Numerous studies, e.g., [5,6,7,8,9], have shown that the monoterpene SOA yield is high under typical atmospheric conditions, reaching values up to 30%. Previous studies have focused on the first-generation products of α- and β-pinene reactions, identifying pinic acid, pinonaldehyde, pinonic acid, nopinone, norpinic acid, etc. as some of the most important products in this system [10,11,12,13,14]. These first-generation products are semi-volatile with a wide range of volatilities. For example, the saturation concentrations of pinic and norpinic acid range between 0.9–23 μg m⁻³ and 6.2–21 μg m⁻³, respectively, for temperatures in the 290–310 K range [15], while nopinone and pinonaldehyde are more volatile with saturation concentrations in the range of 175–535 μg m⁻³ (for 290–310 K) [16] and 200 μg m⁻³ at 303 K [17], respectively. These first-generation products can further react to form second- and third-generation products and potentially increase the SOA yield of the parent VOC during a process called chemical aging. Aging of pinonic acid leads to 3-methyl-1,2,3-butanetricarboxylic acid (MBTCA) formation [14,18], while aging of nopinone leads to terpenylic acid production [14], which further reacts with OH radicals, forming terebic acid [19]. Pinonaldehyde oxidation by OH can produce a considerable amount of SOA in certain reaction pathways, but there also several reaction pathways that lead to fragmentation of the precursor carbon backbone and therefore an increase of volatility and evaporation of the corresponding organic aerosol [20].

Norpinic acid (C₈H₁₂O₄) has been found in α-pinene SOA [12,21,22,23,24,25,26] and is also one of the β-pinene SOA components [12,13,14,21]. Norpinic acid has also been measured in the gas phase in laboratory α- and β-pinene oxidation experiments by Yu et al. [21] and it is clearly a semi-volatile SOA component. The norpinic acid yields decrease as NO₂ increases [24].

In addition, norpinic acid has been measured in forests (boreal, spruce, tropical, etc.) in the high Arctic and in urban areas, e.g., [27,28,29,30,31,32,33,34]. In those studies, particulate concentrations ranged between 11 pg m⁻³ and 86 ng m⁻³. The lower concentrations of particulate norpinic acid were observed in the High Arctic (Canada), (11–153 pg m⁻³) and interestingly its mass concentration was lower in the summer compared to winter [30]. Most studies did not observe a significant diurnal variation, except Fu et al. [32] who reported an increase of norpinic acid as light intensity and temperature increased. In forested areas (e.g., Hyytiälä, Indian tropical forests, etc.) norpinic acid particulate mass concentration ranged between zero and 1.6 ng m⁻³ [27,28,31,32]. Studies in the boreal forest at Hyytiälä during the summer have reported a wide range of average concentrations varying from 4.3 to 56 ng m⁻³ [29]. Li et al. [34] measured norpinic acid concentrations in the summer in a forest in west China up to 86 ng m⁻³, while at the same period the norpinic acid concentration in a nearby city was lower and less than 30 ng m⁻³.

Despite the fact that norpinic acid has been detected both in laboratory scale studies and in field campaigns, there are limited studies about its physicochemical properties. Bilde et al. [15] measured, for the first time, the vapor pressure of trans-norpinic acid and reported a value of 11 ± 5.5 μg m⁻³ at 300 K. The melting point of trans-norpinic acid in the same study was 410 K. Kołodziejczyk et al. [35] measured the physicochemical properties of cis-norpinic acid such as its melting point (440.7 K), enthalpy of fusion (14.8 kJ mol⁻¹), heat capacity at the fusion temperature (33.5 J mol⁻¹ K⁻¹), and intrinsic solubility (0.043 mmol dm⁻³). The water solubility of norpinic acid allows it to contribute to the ability of biogenic SOA particulate matter to act as cloud condensation nuclei. Kołodziejczyk et al. [35] also reported that cis-norpinic acid has a solid–solid phase transition at 338 K with an enthalpy of transition equal to 15.6 kJ mol⁻¹.

Previous studies proposed that norpinic and norpinonic acid derive from further oxidation of the first generation (aldehydic) products of the pinenes [10,11]. Yasmeen et al. [36] proposed norpinic acid as a potential tracer for aged α-pinene SOA through degradation of pinic acid. However, Ma et al. [24] suggested that norpinic acid could be also formed as a direct product of α-pinene ozonolysis. After its formation, norpinic acid may react heterogeneously or via gas-phase reactions, forming dimers with extremely low volatility [37]. According to the Master Chemical Mechanism predictions [38,39] norpinic acid reacts with OH radicals with an estimated reaction constant of 6.6 × 10⁻¹² cm³ molecule⁻¹ s⁻¹ (based on a structural-activity relationship, SAR) losing atoms of oxygen, while further reactions lead to open ring products. Yasmeen et al. [19] proposed as products the open ring compounds C₃H₄O₂ (MW = 72) and C₅H₈O₂ (MW = 100). The corresponding lifetime of norpinic acid in the gas phase due to the reaction with OH (assuming OH levels of 1 × 10⁶ molecule cm⁻³) is approximately 1.8 days.

In this work, we used two different atmospheric simulation chambers together with online instrumentation for the characterization of the physicochemical properties of pure norpinic acid aerosol. We measured its aerosol mass spectrometer (AMS) spectrum, and we estimated its density, saturation concentration, and vaporization enthalpy. In addition, we explored the chemical aging of norpinic acid, through its reaction with OH radicals, and the mass spectrum of its products.

2. Experimental Approach and Methods

Pure norpinic acid was synthesized in the Laboratory of Environmental Chemistry in the Institute of Physical Chemistry (Warsaw, Poland), as described in the Supplementary Materials. Briefly, the synthesis took place in two stages, using verbenone as a starting point. The first step was the oxidative cleavage of the double C = C bond using RuCl₃ × H₂O, whereas the second one entailed the haloform reaction using bromine and sodium hydroxide (Figure S1). The nuclear magnetic resonance (NMR) spectra of the synthesized norpinic acid is depicted in Figure S2.

2.1. Experimental Set Up

The experiments were conducted in a 10 m³ Teflon reactor in Patras (FORTH/ICE-HT environmental chamber) and in a similar reactor in Carnegie Mellon University (Figure S3). The use of the two different chambers helps assure that any artifacts related to the walls of the chambers are negligible for the purposes of this work. The RH in the chambers was always kept below 20%. An atomizer (TSI, model 3076) was used to generate aerosol through the corresponding aqueous solutions. Norpinic acid and ammonium sulfate aerosols were introduced into the chamber either from the same solution, or as separate solutions (3–4 g L⁻¹ for each compound). Before entering the chamber, the droplets were dried by passing through a diffusion dryer.

OH radicals were produced by HONO photolysis under UV illumination (J_NO2 = 0.59 min⁻¹ and 0.20 min⁻¹ for FORTH/ICE-HT and CMU chambers respectively), following the same procedure as described in Kostenidou et al. [40]. A HONO aqueous solution was prepared by mixing a fresh sulfuric acid solution (4.9 g L⁻¹) with a fresh sodium nitrate solution (6.9 g L⁻¹) and HONO was introduced into the chamber through a bubbler. Before the photolysis 80–180 ppb of butanol-d9 (98%, Cambridge Isotope Laboratories, Inc.) was introduced into the chamber as an OH tracer [41]. The experimental parameters are shown in

Table 1. Experimental parameters. Experiments 1–4 were conducted in the ICE-HT chamber, while experiments 5 and 6 took place in the CMU chamber.

Facility	ICE-FORTH				CMU
	Exp. 1	Exp. 2	Exp. 3	Exp. 4	Exp. 5	Exp. 6
Initial norpinic acid (μg m−3)	73.3	84.3	11.6	33.1	47	41
Initial (NH4)2SO4 (μg m−3)	31.6 a	82 b	NA	NA	44 b	34 b
AMS Collection Efficiency	0.38 ± 0.05	0.4 ± 0.03	0.4 ± 0.03	0.24 ± 0.03	0.6 ± 0.07	0.75 ± 0.12
Density (g cm−3)	1.29 ± 0.02	1.32 ± 0.03	1.3 ± 0.02	1.29 ± 0.02	ΝA	ΝA
Butanol-d9 (ppb)	0	82	182		17	NA
Equivalent OH exposure (days) c	0	5	2	0	0.6	NA
T (°C)	20.7 ± 0.5	29.1 ± 0.7	22.8 ± 3.7	23.6 ± 1.3	27.3 ± 1.8	28.1 ± 1.4
RH (%)	11 ± 0.3	9.6 ± 0.7	12.8 ± 3.5	16.5 ± 5.3	5 ± 0.8	6 ± 0.9

^a Ammonium sulfate and norpinic acid were introduced by different solutions. ^b Ammonium sulfate and norpinic acid were introduced by the same solution. ^c Assuming a daily average OH concentration of 1 × 10⁶ molecule cm⁻³ [1].

.

The experiments covered a wide range of initial concentrations of norpinic acid aerosol and exposures to OH. In some experiments we used pure norpinic acid to avoid any artifacts from the introduction of seeds on its characterization, in some others we introduced the seeds and the norpinic acid aerosol (externally mixed particles) separately, and we also performed experiments with mixed ammonium sulfate/norpinic acid aerosol (internally mixed). The use of both inert seeds and norpinic acid particles allows a more accurate characterization of the particle losses to the walls of the chamber during the experiment. The use of mixed particles allowed us to estimate the volatility of the norpinic acid at room temperature based on the change of the particle composition.

The mass concentration and the chemical composition were measured by an Aerodyne high-resolution time-of-flight aerosol mass spectrometer (HR-ToF-AMS) sampling at approximately 0.1 L min⁻¹ [42]. The size number and volume distributions were provided by a scanning mobility particle sizer (SMPS, classifier model 3080, DMA model 3081, CPC model 3775, TSI) sampling at 0.5 L min⁻¹.

A thermodenuder [43] was employed for the mass fraction remaining (MFR) measurement of norpinic acid. The thermodenuder used in this study has been described elsewhere [40,44]. The aerosol was passing alternatively either through the thermodenuder, or through a bypass line every 3 min. Each line was analyzed by the HR-AMS and the SMPS. The thermodenuder was operated at temperatures between 25 and 160 °C. The centerline residence time was 14 s.

A Proton Transfer Reaction Mass Spectrometer (PTR-MS, Ionikon Analytik) monitored the gas phase for 47 different m/z’s with a resolution of 10 s. The inlet tube and the reaction chamber were set at 60 °C. The drift tube pressure was 2.2 mbar and the voltage applied was 600 V.

2.2. Data Analysis

To analyze the raw HR-AMS data, we used the standard HR-AMS software SQUIRREL v1.60A and PIKA v1.20A with Igor Pro 6.37 (Wavemetrics). OH levels were calculated based on the butanol-d9 decay (following the signal at m/z 66 in PTR-MS) and using a reaction rate constant with OH of 3.4 × 10⁻¹² cm³ molecule⁻¹ s⁻¹ [41]. For the particle losses calculation inside the thermodenuder, we used the procedure described in Louvaris et al. [44].

The OA volatility was estimated using the dynamic mass transfer model of Riipinen et al. [45]. The model simulates the particle vaporization inside the thermodenuder and predicts the MFR. We estimated both the OA saturation concentration (C*) at 298 K and its effective vaporization enthalpy (ΔH_vap) by minimizing the discrepancy between the measurements and the model predictions. For our simulations, we assumed a mass accommodation coefficient equal to unity.

For the particle wall losses in the experimental chamber, we applied a size-dependent particle correction following the approach suggested by Wang et al. [46]. For the characterization of the wall-loss rate constant, k, dry polydisperse ammonium sulfate particles were used either before or after each experiment. The wall-loss rate constant was quite stable and similar for all the experiments (Figure S4).

For the estimation of the elemental ratios, we applied the method of Canagaratna et al. [47] together with the fragmentation table of Aiken et al. [48], as described in Kostenidou et al. [40]. The approach takes advantage of the high-resolution spectra measured by the AMS, finds the carbon and oxygen content of each m/z signal, and then combines them to calculate the overall O:C for the aerosol. The density of the norpinic acid particles and the collection efficiency (CE) of the AMS were estimated by matching the AMS mass distributions with the SMPS volume distributions using the algorithm of Kostenidou et al. [49]. The conversion of the AMS total mass size distribution to an aerosol volume distribution requires the CE of the AMS and also the unknown density of the OA. The algorithm finds the values of these two parameters that lead to an AMS aerosol volume distribution that is consistent with that measured by the SMPS.

PMF analysis [50,51] was applied to HR-ToF-AMS organic mass spectra in an effort to separate them into fresh and aged norpinic acid aerosol. We followed the procedure of Ulbrich et al. [52].

3. Results and Discussion

3.1. HR-ToF-AMS Mass Spectrum and Elemental Ratios

The average high-resolution AMS spectrum of norpinic acid is depicted in Figure 1. The average unit resolution mass spectrum is shown in Figure S5; the differences between the mass spectra of each individual experiment and the average mass spectrum were small, as their R² ranged between 0.972 and 0.994 (or they had an angle theta between 4 and 12 degrees); the angle theta is described in Kostenidou et al. [53]. The most important characteristic m/z ratios were 59, 65, 67, 68, 73, 77, 79, 82, 83, 85, 95, 100, 108, 111 and 126. Most of them were dominated by oxygenated fragment ions with one or two oxygens. The signal at m/z 82 was quite distinct and mainly due to C₅H₆O⁺. The m/z 82 fragment has also been reported in ambient Positive Matrix Factorization (PMF) biogenic-OA related sources, for example in Borneo, Malaysia [54], at a rural area in Ontario, Canada [55], in tropical rainforests in the central Amazon Basin [56] and at a suburban site in Patras, Greece [57]. In other studies, PMF factors with a characteristic m/z 82 have been attributed to isoprene epoxydiols (IEPOX) OA [58,59,60] due to the isoprene presence and because IEPOX reactive uptake products have a characteristic peak at m/z 82 [61]. However, since m/z 82 is also present in the norpinic acid mass spectrum, the ambient PMF sources containing the m/z 82 should not be attributed only to IEPOX OA but also to monoterpene products.

Table 2. Comparison between the pure norpinic acid mass spectrum and reference mass spectra of ambient biogenic sources as well as mass spectra of biogenic SOA formed in environmental chambers.

	Angle Theta in Degrees (R2)	Reference
Factor 82	48.6 (0.35)	Robinson et al. [54]
IEPOX OA	73.5 (0.01)	Budisulistiorini et al. [58]
Isoprene-OA	31.9 (0.69)	Xu et al. [59]
Biogenic OA	26.3 (0.78)	Kostenidou et al. [57]
MBTCA	24.3 (0.82)	Kostenidou et al. [40]
α-pinene + O3 SOA	27.7 (0.76)	Wang et al. [62]
β-carryophylene + O3 SOA	33.8 (0.65)	Tasoglou et al. [63]

shows the correlation (expressed in both R² and angle theta) between the norpinic acid mass spectrum and the mass spectra of biogenic ambient factors, terpenes SOA and MBTCA. The best correlation was observed for the MBTCA spectrum [40], the biogenic OA factor found in the area of Patras [57] and the α-pinene ozonolysis SOA [46].

The estimated O:C and H:C ratios were 0.49 ± 0.01 and 1.56 ± 0.01 practically the same with the expected values for C₈H₁₂O₄.

The AMS spectra of norpinic acid particles passing through the TD give us an opportunity to investigate any potential changes in the particles as they are exposed to higher temperatures, at least from the AMS limited point of view. After heating the particles up to 60 °C, the AMS spectra were similar to those measured at room temperature with angles theta lower than 3 degrees (Figure S6). At 65 °C, 80% of the norpinic acid had evaporated and the AMS spectrum of the remaining material had an angle of 8° compared to the one measured at room temperature. At 70 °C, with 14% of the material remaining, the corresponding angle θ increased to 20° (Figure S6). This unexpected behavior could be due to different reasons. First, it could be due to impurities in the particles introduced by the aerosol production system. Comparing the mass spectrum of the aerosol produced after the atomization of the water used for our experiments and the TD mass spectra at 70 °C and 90 °C, we found similarities as the angles were relatively low (19 and 15 degrees correspondingly) (Figure S7). This implies that at TD temperatures higher than 65 °C most of the norpinic acid had evaporated and the remaining particles were mainly composed of water impurities. A second explanation could be possible impurities during the synthesis of the norpinic acid. Based on NMR spectrum analysis (Figure S2), these impurities were less than 1–3%, therefore the purity of norpinic acid was 97–99%, which makes this hypothesis unlikely. Another explanation could be reactions taking place in the TD at these higher temperatures and thus changing the norpinic acid. The estimated O:C ratio of the aerosol at 70 °C was 0.62 and its H:C was 1.60. At 70 °C the m/z’s 39, 55, 67, 82, 85, 95 and 100 were lower, while the m/z 44 was higher compared to the AMS spectrum at ambient temperature. This implies that the contribution of some reactions during heating of the aerosol is also possible.

3.2. Density and AMS Collection Efficiency

For the experiments conducted in the ICE-HT smog chamber, the norpinic acid particle density at room temperature was equal to 1.3 ± 0.01 g cm⁻³, while the CE was 0.37 ± 0.05. Using the equation of Kuwata et al. [64], which is based on the elemental ratios, the estimated norpinic acid density was 1.3 g cm⁻³, in excellent agreement with the value estimated by the algorithm of Kostenidou et al. [49]. This supports the applicability of this equation for similar compounds. The CE of the aerosol going through the TD was the same as that of the aerosol going through the bypass line, so there was no need for corrections of the AMS mass concentrations related to different CE values between the heated and non-heated particles.

3.3. Norpinic acid Volatility

3.3.1. Using the Thermodenuder Measurements

The norpinic acid mass fraction remaining (MFR) corrected for particle losses in the TD and calculated based on the AMS measurements is depicted in Figure 2. The corresponding MFR calculated from the SMPS measurements using the temperature-dependent density (Figure S8) was within 10% of that calculated from the AMS measurements.

Following the modelling approach described in Riipinen et al. [45], we used as inputs the corrected AMS MFR data and we estimated a saturation concentration of 8.9 μg m⁻³ and a vaporization enthalpy of 71 kJ mol⁻¹. Our estimated saturation concentration was only slightly higher, but quite similar to the previous estimate of C* = 9 μg m⁻³ reported by Bilde and Pandis [15] for norpinic acid at 298 K. In addition, our ΔH_vap was at the upper part of the range of Bilde and Pandis [15], who suggested a ΔH_vap = 42 ± 51 kJ mol⁻¹.

3.3.2. Isothermal Evaporation of Norpinic Acid in the Chamber

The norpinic acid aerosol introduced in the simulation chamber filled with clean air evaporates partially to establish equilibrium with the gas phase. This evaporation can be used to estimate its saturation concentration at room temperature. The simplest approach is to estimate this saturation concentration, C_norp, as the difference of the initial concentration of norpinic acid aerosol, OA(t₀), and its concentration at equilibrium (corrected for wall losses), OA(t_eq):

$$C_{norp}=OA\left(t_0\right)-OA\left(t_{eq}\right)$$

(1)

The difference in these concentrations is the amount that has been evaporated from the particles and moved to the gas phase. The challenge in this simple approach is that the determination of the two concentrations is not straightforward. First, the aerosol is not well mixed for several minutes after its introduction in the chamber, and therefore the initial concentration of the norpinic acid aerosol, OA(t₀), is quite uncertain. One could, for example, use the peak value, which for Exp. 2 was 86.1 μg m⁻³ (Figure 3a). However, this is probably an upper estimation because it corresponds to a period in which the introduced norpinic acid has not mixed well throughout the chamber. This is supported by the fact that the sulfate concentration (corrected for wall losses) is decreasing, suggesting that the system is still getting mixed. So, the difference of 86.1 − 46.9 = 39.2 μg m⁻³ is probably significantly overestimating the saturation concentration of norpinic acid.

In experiments in which the norpinic acid is produced together with ammonium sulfate from the same solution, the particles contain both the nonvolatile (at room temperature) sulfate and the evaporating norpinic acid. If all particles had the same composition and there was no norpinic acid evaporation, then the OA to sulfate ratio would remain constant with time and would be equal to the mass ratio of the two compounds in the solution that was atomized to produce the particles; however, this is not the case, as shown in Figure 3b. Using the ratio of the OA to sulfate, the saturation concentration of norpinic acid C_norp can be calculated from the equation below:

$$C_{norp}=Sulf\left(t_{eq}\right){\left[\frac{OA}{Sulf}\right]}_0\exp (k_w{t}_{eq})-OA\left(t_{eq}\right)-k_w{{\displaystyle \int }}_0^{t_{eq}}OA\left(t\right)dt$$

(2)

where [OA/Sulf]₀ is the OA to sulfate concentration ratio at time zero (the time at which the injection of the aerosol in the chamber has been completed) and OA(t_eq) and Sulf(t_eq) are the corresponding concentrations at a time t_eq at which the system has equilibrated. The average wall-loss rate constant is k_w and the measured concentration of OA as a function of time is OA(t). In Equation (2) the first term is equal to the concentration of the OA at time zero (before evaporation started), the second is the concentration that remained after evaporation and losses on the walls took place, and the last term is the amount that deposited on the walls. We applied this approach to experiments 2, 5 and 6 and we estimated a saturation concentration of 12.8 ± 4.8 μg m⁻³, which is in good agreement with the value estimated from the thermodenuder measurements. This supports the use of this simple isothermal dilution approach for the estimation of the saturation concentration of the compound of interest.

3.4. Chemical Aging

OH radicals were introduced into the chamber with norpinic acid particles during four experiments (exp. 2, 3, 5 and 6) under UV irradiation. The equivalent OH exposure varied from 0.6 days in experiment 5 to five days in experiment 2, assuming a daily average OH concentration equal to 10⁶ molecule cm⁻³ [65].

The resulting changes in the norpinic acid aerosol concentration were small and could be equal to zero, taking into account the uncertainty due to the wall-loss corrections. For example, during experiment 2 we estimated a decrease of 6% of the organic aerosol concentration based on the AMS data after several hours of exposure to OH (Figure 3a). Based on the SMPS data, the total aerosol volume (which includes the ammonium sulfate) decreased by 3% (Figure 3c), a result consistent with that of the AMS. Experiment 3 showed the same behavior (Figure S9a,b), with the wall-loss corrected organic mass concentration decreasing by 10%, while the wall-loss corrected volume concentration decreased by 5%. The chemical aging due to exposure to OH radicals in experiments 5 and 6 also resulted in a small decrease in the wall-loss corrected total organic mass concentration (10% and 2%, respectively). These results suggest that there was no rapid or large net production of secondary organic aerosol as the norpinic acid particles and vapor reacted with the OH radicals in our experiments. If anything, there could be some small reduction of the particle mass concentration, but this change was within the limits of experimental uncertainty.

In addition, the O:C of the organic aerosol remained practically constant during these aging experiments, changing less than 0.8% in all experiments (Figure 3d and Figure S8c). The small observed change was always an increase.

The theta angle between the mass spectra of the fresh norpinic acid aerosol and the mass spectra after OH exposure and UV irradiation was less than 4.5 degrees for all four aging experiments, even after an OH exposure of approximately five days (Figure 4). These small changes were due to increases in the fragments with m/z 43.018 (C₂H₃O⁺), 57.034 (C₃H₅O⁺), 69.034 (C₄H₅O⁺), 71.049 (C₄H₇O⁺), etc. when OH was introduced into the chamber. Figure S10 illustrates some of the fragments that increased with aging during experiment 2. During experiment 5 we did not observe any significant increase for any m/z during the exposure to OH radicals. This experiment had the shortest exposure to OH (0.6 days of equivalent exposure) and a net decrease of the OA concentration (approximately 10% after correcting for wall losses). This may suggest that the reactions in this experiment resulted in fragmentation and evaporation of a small fraction of the norpinic acid aerosol without any significant addition of new secondary material from the gas phase. In experiment 6, most m/z values remained quite stable after the OH addition (Figure S11). Taking into account that there is evidence of some small decreases of the wall-loss corrected total organic mass concentration by 3%, we conclude that slow fragmentation was likely the only significant process during this experiment as well. Unfortunately, during experiment 6 the OH concentration could not be estimated due to a malfunction of the PTR-MS.

Performing PMF analysis for experiment 2, two factors were extracted. The first factor corresponded to the fresh norpinic acid with an O:C equal to 0.49 (Figure 5a). When the chemical aging started, the mass concentration of the fresh norpinic acid factor started decreasing (Figure 6), reaching 60% of the total organic mass after 17 h. The second factor was probably related to the aged norpinic acid aerosol. It had almost zero mass concentration when the UV lights were off, and it started increasing as soon as the aerosol was exposed to OH (Figure 6). The aged norpinic acid aerosol AMS spectrum had an O:C ratio equal to 0.53 (Figure 5b), very close to that of the fresh. This similarity of the O:C values could explain the fact that the total O:C did not practically change over chemical aging. The compounds produced by the norpinic acid aging had a similar AMS mass spectrum with the fresh norpinic acid (angle theta = 12.5°, R² = 0.948), with only a slightly higher O:C ratio.

The results of the PMF analysis of the other three aging experiments provided little useful information, as the AMS spectra of the fresh and the aged aerosol were quite similar to each other. This is probably due to the smaller OH exposure in these experiments (compared to experiment 2) that did not allow the PMF algorithm to separate the fresh and aged components.

4. Conclusions

In this work we investigated the physicochemical properties of norpinic acid aerosol and its aging potential in two environmental chambers. The saturation concentration (at 298 K) was estimated to be equal to 8.9 μg m⁻³ using thermodenuder measurements and 12.8 μg m⁻³ using isothermal dilution. This suggests that this biogenic secondary organic aerosol component is semi-volatile and will exist in both gas and particulate phases under typical atmospheric conditions. Its vaporization enthalpy was 71 kJ mol⁻¹, while its density was 1.3 g cm⁻³.

A characteristic peak at m/z 82.042 (C₅H₆O⁺), which was observed in the AMS spectrum, can serve as a fingerprint for norpinic acid. Upon chemical aging with OH radicals, which corresponded to an exposure of 0.6–5 days at low RH levels, the wall-loss corrected total organic mass concentration had only a slight net decrease (2–10%). However, these changes were similar in magnitude to the uncertainty of the wall-loss corrections of our measurements. The O:C ratio of the organic aerosol also remained approximately constant (changing by less than 1%) during the exposure to OH. Applying PMF analysis to the AMS data, two factors were determined: one linked to fresh and a second one related to aged norpinic acid. The aged norpinic acid had only a slightly higher O:C ratio (0.53) compared to the fresh, with a rather similar mass spectrum. This may imply that the norpinic acid products have a similar chemical signature on the AMS, and thus it will be quite difficult to distinguish them from the fresh norpinic acid in the ambient atmosphere using this technique. Our results also suggest that the chemical aging of norpinic acid with OH does not result in drastic change of either the secondary organic aerosol concentration or its AMS signature during the timescales and conditions investigated in this study.

The use of particles consisting of both ammonium sulfate and norpinic acid allowed us to measure the saturation concentration of norpinic acid at room temperature using the change of the aerosol composition measured by the AMS. This easy-to-implement technique can be used as a complimentary approach to thermodenuding or other saturation concentration measurement approaches for organic aerosol.

The results of this study can be used to constrain the biogenic SOA contribution to ambient organic aerosol levels using norpinic acid concentration measurements, and they can also provide valuable insights about the simulation of the later-generation chemistry and corresponding SOA production for monoterpenes.