Previous studies have shown that 1,4-hydroxycarbonyls, which are often major products of the atmospheric oxidation of hydrocarbons, can undergo acid-catalyzed cyclization and dehydration in aerosol particles to form highly reactive unsaturated dihydrofurans. In this study the kinetics of dehydration of cyclic hemiacetals, the rate-limiting step in this process, was investigated in a series of environmental chamber experiments in which secondary organic aerosol (SOA) containing cyclic hemiacetals was formed from the reaction of n-pentadecane with OH radicals in dry air in the presence of HNO3. A particle beam mass spectrometer was used to monitor the formation and dehydration of cyclic hemiacetals in real time, and SOA and HNO3 were quantified in filter samples by gravimetric analysis and ion chromatography. Measured dehydration rate constants increased linearly with increasing concentration of HNO3 in the gas phase and in SOA, corresponding to catalytic rate constants of 0.27 h–1 ppmv–1 and 7.0 h–1 M–1, respectively. The measured Henry’s law constant for partitioning of HNO3 into SOA was 3.7 × 104 M atm–1, ∼25% of the value for dissolution into water, and the acid dissociation constant was estimated to be <8 × 10–4, at least a factor of 104 less than that for HNO3 in water. The results indicate that HNO3 was only weakly dissociated in the SOA and that dehydration of cyclic hemiacetals was catalyzed by molecular HNO3 rather than by H+. The Henry’s law constant and kinetics relationships measured here can be used to improve mechanisms and models of SOA formation from the oxidation of hydrocarbons in dry air in the presence of NOx, which are conditions commonly used in laboratory studies. The fate of cyclic hemiacetals in the atmosphere, where the effects of higher relative humidity, organic/aqueous phase separation, and acid catalysis by molecular H2SO4 and/or H+ are likely to be important, is discussed.

Yields of secondary organic aerosol (SOA) were measured for OH radical-initiated reactions of the 2- through 6-dodecanone positional isomers and also n-dodecane and n-tetradecane in the presence of NOx. Yields decreased in the order n-tetradecane > dodecanone isomer average > n-dodecane, and the dodecanone isomer yields decreased as the keto group moved toward the center of the molecule, with 6-dodecanone being an exception. Trends in the yields can be explained by the effect of carbon number and keto group presence and position on product vapor pressures, and by the isomer-specific effects of the keto group on branching ratios for keto alkoxy radical isomerization, decomposition, and reaction with O2. Most importantly, results indicate that isomerization of keto alkoxy radicals via 1,5- and 1,6-H shifts are significantly hindered by the presence of a keto group whereas decomposition is enhanced. Analysis of particle composition indicates that the SOA products are similar for all isomers, and that compared to those formed from the corresponding reactions of alkanes the presence of a pre-existing keto group opens up additional heterogeneous/multiphase reaction pathways that can lead to the formation of new products. The results demonstrate that the presence of a keto group alters gas and particle phase chemistry and provide new insights into the potential effects of molecular structure on the products of the atmospheric oxidation of volatile organic compounds and subsequent formation of SOA.

The linear C15 alkene, 1-pentadecene, was reacted with NO3 radicals in a Teflon environmental chamber and yields of secondary organic aerosol (SOA) and particulate β-hydroxynitrates, β-carbonylnitrates, and organic peroxides (β-nitrooxyhydroperoxides + dinitrooxyperoxides) were quantified using a variety of methods. Reaction occurs almost solely by addition of NO3 to the C═C double bond and measured yields of β-hydroxynitrate isomers indicate that 92% of addition occurs at the terminal carbon. Molar yields of reaction products determined from measurements, a proposed reaction mechanism, and mass-balance considerations were 0.065 for β-hydroxynitrates (0.060 and 0.005 for 1-nitrooxy-2-hydroxypentadecane and 1-hydroxy-2-nitrooxypentadecane isomers), 0.102 for β-carbonylnitrates, 0.017 for organic peroxides, 0.232 for β-nitrooxyalkoxy radical isomerization products, and 0.584 for tetradecanal and formaldehyde, the volatile C14 and C1 products of β-nitrooxyalkoxy radical decomposition. Branching ratios for decomposition and isomerization of β-nitrooxyalkoxy radicals were 0.716 and 0.284 and should be similar for other linear 1-alkenes ≥ C6 whose alkyl chains are long enough to allow for isomerization to occur. These branching ratios have not been measured previously, and they differ significantly from those estimated using structure–activity relationships, which predict >99% isomerization. It appears that the presence of a −ONO2 group adjacent to an alkoxy radical site greatly enhances the rate of decomposition relative to isomerization, which is otherwise negligible, and that the effect is similar to that of a −OH group. The results provide insight into the effects of molecular structure on mechanisms of oxidation of volatile organic compounds and should be useful for improving structure–activity relationships that are widely used to predict the fate of these compounds in the atmosphere and for modeling SOA formation and aging.

In this study, C8–C14 n-alkanes were reacted with OH radicals in the presence of NOx in a Teflon film environmental chamber and isomer-specific yields of alkyl nitrates were determined using gas chromatography. Because results indicated significant losses of alkyl nitrates to chamber walls, gas–wall partitioning was investigated by monitoring the concentrations of a suite of synthesized alkyl nitrates added to the chamber. Gas-to-wall partitioning increased with increasing carbon number and with proximity of the nitrooxy group to the terminal carbon, with losses as high as 86%. The results were used to develop a structure–activity model to predict the effects of carbon number and isomer structure on gas–wall partitioning, which was used to correct the measured yields of alkyl nitrate isomers formed in chamber reactions. The resulting branching ratios for formation of secondary alkyl nitrates were similar for all isomers of a particular carbon number, and average values, which were almost identical to alkyl nitrate yields, were 0.219, 0.206, 0.254, 0.291, and 0.315 for reactions of n-octane, n-decane, n-dodecane, n-tridecane, and n-tetradecane, respectively. The increase in average branching ratios and alkyl nitrate yields with increasing carbon number to a plateau value of ∼0.30 at about C13–C14 is consistent with predictions of a previously developed model, indicating that the model is valid for alkane carbon numbers ≥C3.

A series of C8–C16 n-alkanes were reacted with OH radicals in the presence of NOx in an environmental chamber and particulate 1,4-hydroxynitrate reaction products were collected by filtration, extracted, and analyzed by high-performance liquid chromatography with UV absorption and electron ionization mass spectrometry (HPLC/UV/MS). Observed mass spectral patterns can be explained by using proposed ion fragmentation mechanisms, permitting the identification of each hydroxynitrate isomer. Reversed-phase retention of these compounds was dictated by the length of the longer of two alkyl chains attached to the 1,4-hydroxynitrate subunit. 1,4-Hydroxynitrates were quantified in particles using an authentic analytical standard for calibration, and the results were combined with gas chromatography measurements of the n-alkanes to determine the molar yields. Yields based on analyses of particles increased with increasing carbon number from 0.00 for C8 to an average plateau value of 0.130 ± 0.008 for C14–C16, due primarily to corresponding increases in gas-to-particle partitioning. The value at the plateau, where essentially all 1,4-hydroxynitrates were in particles, was equal to the average total yield of C14–C16 1,4-hydroxynitrates. The average branching ratio for the formation of C14–C16 1,4-hydroxynitrates from the reaction of NO with the corresponding 1,4-hydroxyperoxy radicals was 0.184 ± 0.011. This value is ∼20% higher than the plateau value of 0.15 for reactions of secondary 1,2-hydroxyperoxy radicals and ∼40% lower than the plateau value of 0.29 for reactions of secondary alkyl peroxy radicals, both of which were reported previously. The branching ratios determined here were used with values reported previously to calculate the yields of C7–C18 alkyl nitrates, 1,4-hydroxynitrates, and 1,4-hydroxycarbonyls, the three products formed from the reactions of these n-alkanes.