A Four Carbon Organonitrate as a Significant Product of Secondary Isoprene Chemistry.

Oxidation of isoprene by nitrate radicals (NO3) or by hydroxyl radicals (OH) under high NOx conditions forms a substantial amount of organonitrates (ONs). ONs impact NOx concentrations and consequently ozone formation while also contributing to secondary organic aerosol. Here we show that the ONs with the chemical formula C4H7NO5 are a significant fraction of isoprene-derived ONs, based on chamber experiments and ambient measurements from different sites around the globe. From chamber experiments we found that C4H7NO5 isomers contribute 5%-17% of all measured ONs formed during nighttime and constitute more than 40% of the measured ONs after further daytime oxidation. In ambient measurements C4H7NO5 isomers usually dominate both nighttime and daytime, implying a long residence time compared to C5 ONs which are removed more rapidly. We propose potential nighttime sources and secondary formation pathways, and test them using a box model with an updated isoprene oxidation scheme.

Introduction

Isoprene dominates biogenic non‐methane hydrocarbon emissions, contributing around 50%, followed by monoterpenes, 15%, and sesquiterpenes, 3% (Guenther et al., 2012). Isoprene reacts mainly with hydroxyl radicals (OH), ozone (O3), or nitrate radicals (NO3) (Wennberg et al., 2018), influencing surface ozone concentrations and secondary organic aerosol (SOA) formation. SOA is a major component of submicron‐sized tropospheric aerosol (Shrivastava et al., 2017) and affects human health and the climate (Glasius & Goldstein, 2016; Hallquist et al., 2009).

The oxidation of isoprene by OH radicals under low and high NOx conditions has been studied extensively (D'Ambro et al., 2017; Kleindienst et al., 2007; Kroll et al., 2005; L. Lee et al., 2014; Novelli et al., 2020; Peeters et al., 2014; Schwantes et al., 2019; Thornton et al., 2020; Wennberg et al., 2018) compared to NO3‐initiated oxidation (Kwan et al., 2012; Ng et al., 2008; Schwantes et al., 2015; Vereecken et al., 2021; Wennberg et al., 2018; Wu et al., 2021; Zhao et al., 2021). NO3 is formed during nighttime from the reaction of nitrogen dioxide (NO2) with ozone. Oxidation initiated by NO3 radicals leads to significant formation of organonitrates (ONs) which add to the ONs produced during daytime oxidation under high NOx conditions (Hamilton et al., 2021; Kiendler‐Scharr et al., 2016). ONs can act both as a reservoir and as a permanent sink of NOx (Kenagy et al., 2020; Kiendler‐Scharr et al., 2016), and can contribute to SOA formation (Bryant et al., 2020; Fry et al., 2018; Kiendler‐Scharr et al., 2016; Lee et al., 2016; Xu et al., 2021; Zaveri et al., 2020). Investigating isoprene‐originated ONs formation is necessary to understand isoprene's effects on atmospheric NOx, HOx and ozone formation (Li et al., 2019; Schwantes et al., 2020; Vasquez et al., 2020).

The dominant gas phase nitrated oxidation products from isoprene + NO3 include compounds like isoprene nitrooxy hydroperoxides (INP) (C5H9NO5), dihydroxy nitrates (IDHN) (C5H9NO5), carbonyl nitrates (ICN) (C5H7NO4), hydroxy nitrates (IHN) (C5H9NO4), and hydroxy hydroperoxy nitrates (IHPN) (C5H9NO6), among others (Schwantes et al., 2015; Vereecken et al., 2021; Wennberg et al., 2018; Wu et al., 2021). These compounds have also been observed in the ambient atmosphere (Lee et al., 2016; Schwantes et al., 2015; Xu et al., 2021; Ye et al., 2021).

To provide new insights into this important nitrate‐isoprene chemistry, an extensive experimental campaign focusing on isoprene oxidation by NO3 was performed in the large atmospheric simulation chamber SAPHIR in 2018. In this study we focus on the formation and fate of the isomers of one of the most ubiquitously detected ONs, C4H7NO5, and how their atmospheric fate changes between night and daytime. The experimental findings are linked to our observations of these compounds around the globe and chemical mechanisms are proposed to support our observations.

Materials and Methods

All experimental studies and field observations in this study utilized a high resolution time‐of‐flight chemical ionization mass spectrometer (Aerodyne Research Inc., hereafter CIMS) to measure nitrated organic products, using iodide as the primary reagent ion (B. H. Lee et al., 2014). The CIMS was deployed for simulation chamber experiments within a comprehensive study on nighttime isoprene chemistry, for dedicated flow reactor studies, and at several field campaign sites providing diurnal concentration profiles of selected organic nitrates.

The comprehensive nighttime isoprene chemistry study was conducted in August 2018 in the atmospheric simulation chamber SAPHIR (Section S1 in Supporting Information S1) (Fuchs et al., 2017; Rohrer et al., 2005) at Forschungszentrum Jülich, Germany (Brownwood et al., 2021; Dewald et al., 2020; Vereecken et al., 2021; Wu et al., 2021). One goal of the campaign was to explore different oxidation regimes by applying conditions enhancing the contribution of different reaction pathways (i.e., RO2 + RO2, RO2 + HO2 or unimolecular RO2). Nevertheless, the dominant loss of RO2 were the reactions with HO2 or NO3, as detailed in a previous manuscript (Brownwood et al., 2021). Here we focus our analysis on four selected experiments: one experiment enhancing HO2 by propene ozonolysis (exp. 1), one favoring the RO2 + RO2 pathway (exp. 2), and two experiments in which a nighttime to daytime transition was achieved by opening the roof of the chamber after the oxidation products of isoprene and NO3 had accumulated (exp. 3 & 4, Table S1 in Supporting Information S1). The daytime chemistry favored either oxidation by OH in exp. 3, or only photolysis by scavenging OH by CO addition in exp. 4. Exp. 3 favored RO2 isomerization and also contained seed aerosol to test effects of heterogeneous chemistry, whereas exp. 4 favored the RO2 + RO2 reactions. This had minor effects on the general evolution of gas‐phase products, that is, only a small fraction of the accretion ONs would partition to the particle phase (Wu et al., 2021). Experiments took place in Gothenburg using the laminar‐flow Go:PAM reactor (Tsiligiannis et al., 2019; Watne et al., 2018) to test for a possible formation of the target compound(s) C4H7NO5 from NO3 oxidation of methyl vinyl ketone (MVK) (Table S8 in Supporting Information S1) and to constrain the instrument's sensitivity to ONs (Lopez‐Hilfiker et al., 2016) (Table S3 in Supporting Information S1). Here NO3 was introduced following the decomposition of N2O5 added via a diffusion source (Sections S2 and S5 in Supporting Information S1).

The ambient concentration profiles of the target isoprene products C4H7NO5 were characterized during six field campaigns. Two campaigns took place in Asia, in Changping (near Beijing) (Le Breton, Wang et al., 2018) and Hong Kong as part of a project on photochemical smog in China (Hallquist et al., 2016). Two additional sites were located in Europe, in Jülich, Germany, during the JULIAC campaign and in Gothenburg, Sweden. Finally, data from Alabama, in the south‐eastern USA during the Southern Oxidant and Aerosol Study (SOAS) (Lee et al., 2016) and from the Amazon rainforest were used to illustrate the omnipresence of the C4H7NO5 isomers (Section S3 in Supporting Information S1). The C4H7NO5 signal has also been observed in the free troposphere as part of flight measurements over the south‐eastern USA during the Southeast Nexus campaign. However, a detailed discussion on those flights will be presented elsewhere.

The University of Gothenburg CIMS (GU‐CIMS) was used for most field and laboratory measurements. The measurements from the SOAS campaign used the University of Washington CIMS (UW‐CIMS), while the Amazon rainforest measurements used the University of Manchester CIMS (UMan‐CIMS). Additional information on operational characteristics of CIMS during each campaign are given in Table S6 in Supporting Information S1. A bulk ON sensitivity factor of 4.8 ncps ppt−1, derived during the SAPHIR experiment was used to convert the measured ONs signal to concentrations by GU‐CIMS, where potential variability between ON was investigated using a voltage scanning method (see Section S2 in Supporting Information S1). UW‐CIMS used a weighted isomer distribution of the isoprene‐derived ON (C5H9NO4, IHN) and UMan‐CIMS used the isoprene‐derived IEPOX as a proxy calibrant. More details on calibrations, sensitivity estimations and assumptions can be found in the SI (Section S2 & S3 in Supporting Information S1).

Results and Discussions

Experiments in the Atmospheric Simulation Chamber SAPHIR

During the experiments in the SAPHIR chamber, 24 mononitrates, 22 dinitrates and 18 accretion products were identified using the CIMS. Mononitrates dominated the spectrum ranging from 80.7% to 96.4% of the measured ONs, followed by dinitrates (3.3%–17.9%) and accretion products (0.2%–1.5%). Formation of ions assigned to the chemical composition C4H7NO5 (Figure S2 in Supporting Information S1) were evident in all experiments. C4H7NO5 signal was identified by the CIMS as an important nitrated product(s) together with the primary products C5H9NO5 (hydroperoxide nitrates, INP), C5H7NO4 (carbonyl nitrates, ICN), and C5H9NO4 (hydroxy nitrates, IHN). Herein INP forms from RO2 + HO2 reaction whilst ICN and IHN largely form from RO2 + RO2 reaction.

Figure 1 depicts the time evolution of C4H7NO5 (black), and the three other major primary oxidation products, C5H9NO5 (blue), C5H7NO4 (purple), and C5H9NO4 (cyan), during the four selected experiments (Table S1 in Supporting Information S1). The relative contribution (red) of C4H7NO5, expressed as the ratio of C4H7NO5 over the total measured ONs signals by CIMS is also shown. The relative contribution is estimated assuming the same sensitivity for all the measured ONs. Generally, the time‐series of the sum of measured mononitrates, dinitrates and accretion products followed the total gas‐phase alkyl nitrates time evolution measured by a Thermal Dissociation Cavity Ring‐down Spectrometer during the campaign (Brownwood et al., 2021), suggesting the CIMS total ONs signal includes the majority of the formed ON. In the dark the primary products increased rapidly after the injection of oxidant precursors and isoprene (Figure 1), especially C5H9NO5 (INP), the dominant primary nighttime ON measured by the CIMS in all experiments. The C4H7NO5 increased slowly and steadily, suggesting that there was no strong sink during the NO3‐dominated nighttime oxidation. This behavior would be typical for a closed shell product. However, the yield of C4H7NO5 strongly depended on the chemical regime. In exp. 1 (Figure 1a) RO2 + HO2 reactions dominated (91%), while the RO2 + RO2 reactions were responsible for only a 4% loss (Brownwood et al., 2021). In exp. 2 (Figure 1b), the RO2 + RO2 reactions contributed up to 13% and the RO2 + HO2 reactions 53% of the loss rate of RO2 (Brownwood et al., 2021). Under dominant HO2 conditions (exp. 1) the C4H7NO5 formation was lower, and its relative contribution was always less than 5%, while the C5H9NO5 (INP) contribution to ONs ranged between 21% and 33% due to the enhanced importance of the HO2 reaction for the primary peroxy radical from isoprene + NO3 reaction. When the RO2 + RO2 reactions were more important (exp. 2), the relative contribution of C4H7NO5 increased, ultimately reaching over 15% and becoming the second major remaining product with a contribution close to that of C5H9NO5.

Figure 1

(a) Exp. 1 favoring RO2 + HO2 reactions. (b) Exp. 2 favoring RO2 + RO2 reactions. (c) Exp. 3 Nighttime‐daytime transition focusing on the effect of the OH oxidation during daytime. (d) Exp. 4 Nighttime‐daytime transition focusing on photolysis, using CO scavenger to suppress OH chemistry.

Toward the end of exp. 1 and 2, when all isoprene was consumed, oxidant precursors were added without additional isoprene to enhance the oxidation of the products. For the experiment favoring RO2 + HO2 (exp. 1), this addition did not have a substantial effect on the products, including C4H7NO5. For the conditions where RO2 + RO2 became more important (exp. 2), the signal of major primary products decreases, thus the relative contribution of C4H7NO5 to the total ONs further increases from around 10% before the enhanced NO3 oxidation to more than 15% at the end of the experiment. Thus, the RO2 + RO2 regime enables a subsequent acceleration of secondary chemistry and multi‐generation products (Wu et al., 2021).

Nighttime to daytime transitions were included in exp. 3 and 4 by exposing the reaction mixture to sunlight by opening the roof of SAPHIR after a period of dark NO3 oxidation (Figures 1c and 1d). The period of NO3 oxidation was similar to exp. 2 without the last oxidant‐only addition. The relative contribution of C4H7NO5 to the total ONs was around 10% at the end of the nighttime period. Under daytime conditions in exp. 3, the ONs concentrations are expected to decrease under low NO conditions, as the carbonyl nitrates from isoprene react with OH radicals or are rapidly photolyzed (Müller et al., 2014; Xiong et al., 2016). This was also observed here for most of the ONs, for example, C5H7NO4 (ICN) and C5H9NO4 (IHN). However, the signal of the major product C5H9NO5 did not decrease, that is, either C5H9NO5 is not affected by daytime chemistry, or there are processes counteracting the loss, for example, formation of a nitrooxy hydroxyepoxide (INHE) by OH oxidation as described by Schwantes et al. (2015). The most pronounced change was the strong increase of the C4H7NO5 signal, which became the dominant nitrated product(s), increasing from 10% at the end of nighttime to over 40% after an hour of sunlight. At the onset of this enhanced increase of C4H7NO5, there was no isoprene left in the chamber, clearly demonstrating the multi‐generational sources of this product(s). In accordance with Schwantes et al. (2015) and Wennberg et al. (2018) we propose that all three major primary C5 ONs (C5H9NO5 (INP), C5H7NO4 (ICN), and C5H9NO4 (IHN)) can react with OH to form products with the chemical formula C4H7NO5. After one and a half hours, C4H7NO5 started decreasing due to chamber dilution, decreased availability of its precursors and slow but persistent removal processes, such as reaction with OH, photolysis.

To separately study the effect of photolysis, CO was added as an OH scavenger during the daytime period in exp. 4 (Figure 1d). In contrast to exp. 3, C5H9NO5 decreased extremely rapidly as soon as the roof was opened. This difference to the daytime period of exp. 3 supports the existence of a formation pathway of C5H9NO5 by OH oxidation that can balance out the significant loss by photolysis, as observed in exp. 4. In contrast, the C4H7NO5 signal still increased in exp. 4 but to a lesser extent than in exp. 3. From the experiments in the simulation chamber, we conclude that formation pathways of C4H7NO5 compound(s) exist in the dark (5%–17% of measured ONs), while during the following daytime, OH oxidation of multi‐generational NO3 products, together with a smaller contribution by photolysis, leads to a significant formation of C4H7NO5, ultimately contributing more than 40% of measured ONs.

Ambient Measurements

C4H7NO5 and the other major ONs observed in the chamber experiments were measured in six different field locations around the world. Overall, C4H7NO5 signal was the dominant isoprene‐derived nitrate measured during both daytime and nighttime at all the sampling sites (Figure S4 in Supporting Information S1). The ratio of C4H7NO5 to the major primary C5 ONs (C5H9NO5, C5H7NO4, and C5H9NO4) often exhibited values above one throughout the day. However, the diurnal profile of C4H7NO5 differed from site to site (Figure 2).

Figure 2

Median diurnal profile of the C4H7NO5 and NOx in six different locations, with the 10th, 25th, 75th and 90th percentile. The gray areas in the plot indicate the nighttime period. Only the nitrogen dioxide profile is depicted in Amazonia. The map with the isoprene emissions is adapted from McFiggans et al. (2019).

In Hong Kong – an isoprene‐rich area with influence from anthropogenic emissions (Peng et al., 2022) – there were two peaks, one during daytime and one during nighttime. The nighttime peak becomes more prevalent if the period with high ONs formation is selected (between 14–25 November, Figure S5 in Supporting Information S1). Then, the median maximum value of the nighttime peak increases from 107 to 135 ppt and the daytime from 139 to 159 ppt. The fraction of C4H7NO5 in the total measured isoprene‐derived ONs ranged from 5% to 40% during the Hong Kong campaign (Figure S5 in Supporting Information S1).

The profile in Changping (near Beijing, China) showed a higher contribution during night than during day in conjunction with higher variability compared to Hong Kong. The isoprene mean diurnal profile had a peak at 14:00 and the isoprene concentrations were almost always above zero, even during nighttime (Le Breton, Hallquist et al., 2018). The local meteorology has a pronounced impact on the type of air masses reaching Beijing, for example, the wind speed has high values during the day and low values during the night (Le Breton, Wang et al., 2018).

The two locations in Europe are characterized by low regional isoprene emissions, especially Gothenburg, Sweden, and as expected the C4H7NO5 signal was much lower compared to the other sites. In Gothenburg, C4H7NO5 had a weak peak around noon, whilst there was a clear wide peak in the early evening in Jülich, Germany.

The Southeastern USA represents an area with high isoprene concentrations and low to modest NOx emissions outside of urban areas. Centreville is a rural site with low average NOx concentrations (Edwards et al., 2017; Lee et al., 2016). There, C4H7NO5 had a diurnal profile with a clear peak during daytime with high variability. The strong increase during morning (8:00–10:00, local time) was likely due to sampling of the residual layer after the morning breakup of the nocturnal boundary layer, enabling both production and downward transport to contribute to increasing C4H7NO5.

Finally, the Amazon rainforest measurements showed a strong daytime peak. The site is remote from human sources, exhibiting very low NOx concentrations (average = 0.62 ppb), and consequently lower C4H7NO5 compared to the other isoprene‐rich areas, Hong Kong with much higher NOx (average = 5 ppb) and southeastern USA with slightly higher NOx levels (average = 0.67 ppb). The corresponding ozone concentrations at the measurement sites are given in SI (Table S5 in Supporting Information S1).

The correlations of C4H7NO5 with the other major primary isoprene‐derived ONs (C5H9NO5, C5H7NO4, and C5H9NO4) varied at the different sites (Table S4 in Supporting Information S1). The isoprene concentration, the origin and chemical age of the air masses, and meteorology all influenced the correlation slope and the correlation coefficient R2. For example, in Gothenburg local isoprene emissions are low and the correlation was likely driven by variability in air mass origin, with primary and multi‐generational isoprene products like C4H7NO5 exhibiting high correlation. In contrast, in areas with high isoprene emissions, such as the Amazon rainforest, the correlation between first and multi‐generational products was significantly weaker, illustrating the influence of air mass aging and the expected sequential production from isoprene.

Based on the chamber experiments, C4H7NO5 was the most abundant isoprene‐derived ON measured by the CIMS during daytime, but not during nighttime. The ambient measurements also showed C4H7NO5 was the dominant isoprene‐derived ON during daytime. However, at the ambient measurements C4H7NO5 was also the highest isoprene‐derived measured ON signal during nighttime. The dominant role for C4H7NO5 observed during nighttime could be due to efficient accumulation. Under daytime ambient conditions C4H7NO5 can be also produced in relatively high amounts via OH oxidation of other isoprene‐nitrated products (C5H9NO5 (INP), C5H7NO4 (ICN), and C5H9NO4 (IHN)). The chamber experiments showed that C4H7NO5 did not have any major losses by NO3 and O3. C4H7NO5 was also not affected drastically by the additions of isoprene or extra non‐OH oxidant during the experiments. The lack of a carbon‐carbon double bond gives C4H7NO5 a low reactivity toward the NO3 and O3 oxidants dominant during the night. This produces longer residence times than other major ONs that enhance accumulation of the produced C4H7NO5.

Formation and Accumulation of C4 Compounds

A key feature from the experiments in the chamber was the pronounced increase in concentration of C4H7NO5 during transitions from night to day conditions. At some of the ambient sites there was also such a tendency, but it was often concealed by the overwhelming contribution of the daytime oxidation of freshly emitted isoprene, especially for the sites with large isoprene emissions (Amazon rainforest, SE‐USA, and Hong Kong). A major result from our work is the demonstration that C4H7NO5 isomers are multi‐generational products of several C5 compounds. In the nighttime‐daytime transition this was evident both with and without OH radical chemistry (Figures 1c and 1d).

During daytime oxidation, C4H7NO5 isomers are expected to form by OH‐initiated oxidation via a number of pathways. For example, two major unsaturated products from OH‐initiated isoprene oxidation are MVK and methacrolein (MACR). OH addition to these compounds, followed by O2 addition to form a peroxy radical and subsequent NO reaction to alkoxy radicals, leads to the formation of nitrooxy ketones and nitrooxy aldehydes, with the chemical formula C4H7NO5 (Jenkin et al., 2015; Praske et al., 2015). OH oxidation of C5 ONs (C5H9NO5 (INP), C5H7NO4 (ICN), and C5H9NO4 (IHN)) can also lead to C4H7NO5 compounds (Schwantes et al., 2015; Wennberg et al., 2018). The initial peroxy radicals from the aforementioned C5 ONs also form alkoxy radicals from reaction with NO, or HO2 where a fraction can decompose to C4H7NO5 isomers (Novelli et al., 2021; Vereecken & Peeters, 2009; Wennberg et al., 2018).

In contrast to the daytime formation of C4H7NO5, where several pathways from different precursors have been suggested, little is known about nighttime formation. The most plausible reactions forming a C4H7NO5 isomer are listed in the supplemental and are summarized here.

Although MVK and MACR oxidation by NO3 radicals is slow (Kwok et al., 1996) these major products could provide a persistent source of C4H7NO5 (Reactions R1 and R2). To verify this, further experiments were performed in an oxidation flow reactor, the Go:PAM (Table S8 in Supporting Information S1) showing a direct source of C4H7NO5 from NO3‐initiated MVK oxidation (Figure S10 in Supporting Information S1). However, the estimated maximum contribution from this pathway is very low and cannot explain the observed formation in the SAPHIR chamber.

Formation of C4H7NO5 during nighttime may also be due to the NO3‐initiated oxidation of first‐generation hydroxy nitrate isomers (IHN, C5H9NO4) (Reactions (R3), (R4) and R5). Here C4H7NO5 isomers are formed via decomposition of produced alkoxy radicals, with the specific pathway depending on the isomer. The structural differences of the IHN‐isomers also affect their rate constants with NO3 radicals, spanning an order of magnitude (Pfrang et al., 2006; Wennberg et al., 2018). These pathways release NO2 back to the system, but such an increase of NO2 was not observed in the SAPHIR chamber experiments. Also, structure‐activity relationship (SAR) estimates that the decomposition channel is not dominant (Novelli et al., 2021). Thus, these reactions are likely insignificant.

Another potential pathway is the further oxidation of hydroxy carbonyls (HC4CCHO and HC4ACHO, C5H8O2) by NO3 (Reactions R6 and R7). The peroxy radicals formed from oxidation of C5H8O2 can undergo isomerization and decomposition leading to the formation of a C4H7NO5 nitrooxy ketone or nitrooxy aldehyde (Figure S6 in Supporting Information S1) (Wu et al., 2021). The peroxy radicals C5H8NO7 can also form a C4H7NO5 nitrooxy carbonyl by decomposing (Reactions R8 and R9) (Wennberg et al., 2018). Those parent peroxy radicals (C5H8NO7) can be formed either by further autoxidation of the initial peroxy radical formed by NO3 oxidation (i.e., isomerization and O2 addition) or by OH‐initiated oxidation of C5H7NO4 isomers. Finally, peroxy acids and epoxides formed by the NO3‐initiated oxidation of isoprene (Vereecken et al., 2021) may decompose forming C4H7NO5 isomers (Reactions R10 and R11). However, this chemistry is not well‐known and needs further attention.

To test potential contributions of the suggested pathways, the FZJ‐NO3‐Isoprene mechanism presented by Vereecken et al. (2021), which in turn is based on the MCMv3.3.1 (Jenkin et al., 1997; Jenkin et al., 2015; Saunders et al., 2003) and on the more explicit descriptions from the CalTech mechanism (Wennberg et al., 2018) was expanded by the additional pathways as described above and summarized in the SI (Section S4 in Supporting Information S1). The model considered Reactions (R6), (R7), (R8), (R9). Figure 3 shows the calculated trends of four selected isomers (i.e., MACRNB, MVKNO3, HMVKANO3 and MACRNO3, with the corresponding structures shown, see also Table S7 in Supporting Information S1) for the nighttime to daytime transition experiment in the SAPHIR chamber (exp. 4, Figure 1d) with OH scavenger, where only photolysis was important. The model predicts that the dominant C4H7NO5 nitrooxy isomer under nighttime conditions should be the aldehyde MACRNB followed by MVKNO3, HMVKANO3 and MACRNO3 in lower concentrations (Figure 3). The overall trend of the model matches the behavior of the measurements. However, there remains a significant discrepancy in the absolute concentrations (see Section S4, Figures S7–S9 in Supporting Information S1). It is not clear why there is a such large discrepancy and as outlined in the SI some measurement concerns are addressed. Still, the general measurement agreements with for example, the total alkyl nitrates (Brownwood et al., 2021) and Br‐CIMS measurements (Section S2 in Supporting Information S1) clearly illustrate that the C4H7NO5 isomers are important products. This discrepancy could be due to unknown formation pathways for example, secondary formation of C4H7NO5 isomers from decomposition of epoxides and peroxides, highlighted as major contributing species in the model. It is not clear if produced epoxides and peroxides are unstable and can decompose into C4 products (C4H7NO5 being one of them) in the gas phase or on available surfaces, thus explaining why only low concentration of these were observed. Finally, the possibility of some production by OH oxidation under dark conditions cannot be ruled out, but certainly the contribution was small as isoprene loss due to NO3 was calculated to be around 90% (Brownwood et al., 2021) and the impact of isoprene + OH on the total yield of C4H7NO5 was below 5% for all modeled experiments (more details in Section S1 in Supporting Information S1). However, further exploration and evaluation of these pathways are beyond the scope of this work.

Figure 3

Comparison of the measured (black) and modeled (green) C4H7NO5 formation during exp. 4 (nighttime to daytime transition using an OH scavenger). The “full FZJ‐NO3‐isoprene” sum of the four main isomers is compared against the I‐CIMS measurements.

In the model, the increase of C4H7NO5 under the OH scavenged/photolytic conditions is due to the formation of the nitrooxy aldehydes MACRNO3 and MACRNB. The photolysis of two hydroperoxy aldehydes (HPALD, with chemical formula C5H8O3), formed via isomerization of isoprene hydroxy peroxy radicals, can lead to C4H7NO5 after a subsequent NO2 addition (Wennberg et al., 2018). One isomer forms the nitrooxy aldehyde MACRNO3 while the other forms the nitrooxy ketone MVKNO3. However, these pathways do not represent the major loss of HPALDs (Wennberg et al., 2018). Xiong et al. (2016) have suggested that C5H7NO4 (ICN) can dissociate via photolysis and then react with O2 and HO2 to form a vinyl hydroperoxide with chemical formula C4H7NO5. Since no mechanistic description was given, in this work the photolysis was implemented as given in Wennberg et al. (2018). Müller et al. (2014) have suggested that photolysis is the dominant sink of the isoprene‐derived carbonyl nitrates such as MACRNO3, MVKNO3 and HMVKANO3 under atmospheric relevant conditions, but in the model, rapid formation from HPALD and other sources counteracts their loss, leading to constant or increasing concentrations.

In addition to photolysis, the residual nighttime primary products can also react with OH radicals during daytime. We attribute the rapid C4H7NO5 formation in exp. 3 (Figure 1c) to the OH‐initiated oxidation of the three other major ONs (C5H9NO5, C5H7NO4, and C5H9NO4), which has also been proposed in previous studies (Wennberg et al., 2018) (Section S4 in Supporting Information S1). The efficiency of these pathways in forming C4H7NO5 must be high in order to fully explain the observations in both the findings in the SAPHIR chamber and the field observations.

Atmospheric Implication and Conclusion

The C4H7NO5 isomers are important products of isoprene oxidation in NOx‐influenced regions. Ambient measurements showed that C4H7NO5 compound(s) typically have higher concentrations than the other three major ONs (C5H9NO5 (INP), C5H7NO4 (ICN), and C5H9NO4 (IHN)) during both night and day. For nighttime conditions this finding appears in contradiction to our chamber measurements, where C4H7NO5 was only dominant during daytime. We found that C4H7NO5 isomers are multi‐generation products, with no remaining C=C bonds, in the isoprene degradation mechanism, formed from both OH and NO3‐radical initiated oxidation where observations in ambient air can be expected from air mass aging processes.

C4H7NO5 nighttime production was investigated here in detail. We suggest that the decomposition of the C5H8NO7 peroxy radicals from NO3‐initiated chemistry, the oxidation of hydroxy carbonyls (HC4CCHO and HC4ACHO) (C5H8O2) by NO3 and the decomposition of nitrated epoxides and peroxides are mainly responsible for nighttime production. The relative contribution of C4H7NO5 to total measured ONs increased in chamber experiments when RO2 loss was enhanced by RO2 + RO2 reactions. Furthermore, the chamber experiments showed that C4H7NO5 formation was lower during nighttime when RO2 + HO2 reactions were dominant. According to model calculations, the isomers MACRNB and MVKNO3 have the highest contribution to C4H7NO5 formation under dark conditions. Although most of the other ONs, generated initially in higher yields, react away after transition into the daytime, C4H7NO5 concentration increased, indicating a slower reactivity together with continuing or enhanced production.

The lack of a carbon‐carbon double bond lowers its reactivity and thus increases its lifetime. Slow oxidation and photolytic reactions of C4H7NO5 lead to longer lifetimes than those of the C5 ONs formed in higher yields from isoprene. This can explain the higher effective concentrations of C4H7NO5 in the residual boundary layer and dominance in various ambient conditions. This suggests that C4H7NO5 isoprene ONs could be important as a long‐term organic reservoir species of NOx, in comparison to the other more reactive isoprene‐derived ONs. The importance of further understanding the properties of the different isomers is highlighted by a recent study (Vasquez et al., 2020) which showed that the isoprene nitrate isomer 1,2‐IHN can efficiently remove NOx from the atmosphere, whereas other isomers cannot. To further understand the distribution of isomers and their specific chemistry, further studies are needed using a broader range of methods. Especially a focus on the predicted dominant product family of the nitrated epoxides (Vereecken et al., 2021), whose secondary chemistry and therefore potential for forming C4H7NO5 is largely unknown, is a necessity. Finally, the isoprene‐derived ONs, of which C4H7NO5 is a dominant species, can thus be important for the formation of ozone, with ONs formed during nighttime affecting the initiation of tropospheric ozone formation during the following day.

Conflict of Interest

The authors declare no conflicts of interest relevant to this study.

Supporting Information S1

Click here for additional data file.