Grantee Research Project Results

2011 Progress Report: Chemistry of Secondary Organic Aerosol Formation from the Oxidation of Aromatic Hydrocarbons

EPA Grant Number: R833752
Title: Chemistry of Secondary Organic Aerosol Formation from the Oxidation of Aromatic Hydrocarbons
Investigators: Ziemann, Paul J. , Arey, Janet , Atkinson, Roger
Institution: University of California - Riverside
EPA Project Officer: Chung, Serena
Project Period: October 1, 2007 through September 30, 2010 (Extended to March 31, 2012)
Project Period Covered by this Report: October 1, 2010 through September 30,2011
Project Amount: $514,464
RFA: Sources and Atmospheric Formation of Organic Particulate Matter (2007) RFA Text | Recipients Lists
Research Category: Particulate Matter , Air Quality and Air Toxics , Air

Objective:

In this project we are developing a quantitative understanding of the kinetics, products, and mechanisms of secondary organic aerosol (SOA) formation from the photooxidation of aromatic hydrocarbons, and will provide this information to the scientific community in a form that can be readily incorporated into SOA modules used in air quality models for predicting atmospheric organic PM_2.5 concentrations. These types of models are used widely to evaluate the potential effects of aerosols on global climate, air pollution and visibility, and human health, all of which are important problems confronting society.

Progress Summary:

In year 1 of this program, environmental chamber experiments were carried out to identify and quantify dicarbonyl products formed from reactions of OH radicals with toluene, o-, m- and p-xylene and 1,2,3-, 1,2,4- and 1,3,5-trimethylbenzene. Gas-phase products were collected using denuders coated with XAD resin and O-(2,3,4,5,6-pentafluorobenzyl)hydroxylamine (PFBHA) to derivatize carbonyl-containing products for GC/MS analysis. The 1,2-dicarbonyls glyoxal, methylglyoxal and biacetyl were observed, as were 8 of 9 possible unsaturated 1,4-dicarbonyl co-products. Compared to their potential co-product 1,2-dicarbonyls, unsaturated 1,4-diketones had similar formation yields, whereas all but one unsaturated 1,4-dialdehyde and keto-aldehyde had lower yields. These results provide new product yields from aromatic reactions that can be used as inputs to atmospheric models. In addition, the photolysis rate of 2-formylcinnamaldehyde was measured by monitoring its time dependent signal during the naphthalene-OH reaction using atmospheric pressure ionization mass spectrometry (API-MS). 2-Formylcinnamaldehyde is a major product of the OH radical-initiated reaction of naphthalene, the atmospherically most abundant polycyclic aromatic hydrocarbon, whose oxidation has been suggested as a possible source of SOA in urban atmospheres. Results were used with those from our earlier, 1997, study to determine a 2-formylcinnamaldehyde formation yield of 56%. Combined with other previously observed and quantified products, we now can account for ~92% of naphthalene reaction products under conditions where the NO₂ concentration is greater than ~60 ppbv.

In year 2 of this project, the formation yields of glyoxal were measured from the OH radical-initiated reactions of naphthalene, 1-methylnaphthalene, 1,4-dimethylnaphthalene, acenaphthene and acenaphthylene, using solid phase microextraction (SPME) fibers pre-coated with PFBHA for collection of glyoxal and GC-FID for analysis. In the presence of NO_x, glyoxal was observed as a first-generation product from these PAHs, with yields of 5%, 3%, 2%, 10-15% and <2%, respectively, and with a yield from naphthalene in the absence of NO_x of 3%. Second-generation formation was obvious from the 1-methylnaphthalene, 1,4-dimethylnaphthalene and acenaphthene reactions. Simultaneous measurements of phthaldialdehyde from naphthalene, of 2-acetylbenzaldehyde from 1-methylnaphthalene and of 1,2-diacetylbenzene from 1,4-dimethylnaphthalene suggest that these aromatic dicarbonyls are co-products to glyoxal. The formation yields of glyoxal and methylglyoxal were measured from the gas-phase OH radical-initiated reactions of toluene, o-, m- and p-xylene, and 1,2,3-, 1,2,4- and 1,3,5-trimethylbenzene as a function of the NO₂ concentration [(0.1-4) ppm]. Glyoxal and methylglyoxal were collected onto SPME fibers pre-coated with PFBHA and analyzed as their oximes by GC-FID. The glyoxal and methylglyoxal yields generally decrease with increasing NO₂ concentration. However, for formation of glyoxal from 1,2,3-trimethylbenzene and of glyoxal and methylglyoxal from 1,2,4-trimethylbenzene, the yields were independent of the NO₂ concentration within the experimental errors. These data allow, by a very short extrapolation, glyoxal and methylglyoxal yields appropriate for atmospheric conditions.

In years 2 and 3 of this project, 2-formylcinnamaldehyde formation from OH + naphthalene was investigated in the absence of NO_x (using O₃ + alkene to generate OH radicals) and in the presence of NO_x at 0.1 and 1 ppm NO_x. SPME fibers were used for sample collection, and our data show that 2-formylcinnamaldehyde is formed in the absence (as well as in the presence) of NO_x, indicating that the OH-naphthalene adduct + O₂ reaction forms 2-formylcinnamaldehyde. Based on our previous 2-formylcinnamaldehyde yield at ppm levels of NO₂, the 2-formylcinnamaldehyde yields at 0.1 ppm NO_x and in the absence of NO_x are 43% and ~20%, respectively. In the absence of NO, RO₂ + RO₂ and RO₂ + HO₂ radical reactions will dominate and a lower yield of 2-formylcinnamaldehyde is expected if 2-formylcinnamaldehyde is formed from an alkoxy radical. Our data then suggest that the 2-formylcinnamaldehyde formation yield is not too dissimilar from the reactions of the OH-naphthalene adducts with O₂ and NO₂. We investigated the formation of unsaturated 1,4-dicarbonyls (and other products) from the OH radical-initiated reactions of furans (furan, 2- and 3-methylfuran and 2,3- and 2,5-dimethylfuran), using PFBHA-coated denuders, gas chromatography and API-MS. These studies are valuable for understanding the chemistry of aromatic reactions because the unsaturated 1,4-dicarbonyls formed from these furans are the same as those formed from aromatics, but are formed in higher yields and with fewer co-products so they can be used more easily to investigate the subsequent kinetics and products of the 1,4-unsaturated dicarbonyl reactions. We find that furan forms HC(O)CH=CHCHO, 2-methylfuran forms CH₃C(O)CH=CHCHO, 3-methylfuran forms HC(O)C(CH₃)=CHCHO, 2,3-dimethylfuran forms CH₃C(O)C(CH₃)=CHCHO, and 2,5-dimethylfuran forms CH₃C(O)CH=CHC(O)CH₃, with formation yields in the presence of NO of 82 ± 9%, 31 ± 5%, 38 ± 2%, 8 ± 2%, and 24 ± 3%, respectively. The formation yield of CH₃C(O)CH=CHC(O)CH₃ (mainly the cis-isomer) from OH + 2,5-dimethylfuran in the absence of NO also was measured, and determined to be 34 ± 4%. API-MS analyses showed the formation of additional products from all of the furans studied, which are attributed to (taking furan as an example) HC(O)CH=CHCOOH and/or the isomeric hydroxylactone. In addition, for OH + 2,5-dimethylfuran API-MS analyses showed the formation of a product attributed to the keto-ester CH₃C(O)OCH=CHC(O)CH₃. Using API-MS to monitor the 1,4-unsaturated dicarbonyls, the concentration-time dependence of the 1,4-unsaturated dicarbonyls has been studied from the five furans available; the results indicate that in our chambers with blacklamp irradiation, removal of the unsaturated 1,4-dicarbonyls is dominated by reaction with OH radicals, with OH radical reaction rate constants of (6 ± 2) ´ 10^-11 cm³molecule^-1s^-1, and showing no evidence of rapid photolysis or wall loss rate of the unsaturated 1,4-dicarbonyls.

We have investigated formation of cresols from the OH + m-xylene and OH + p-cymene reactions to assess the importance of dealkylation (e.g., OH + m-xylene -> cresol + CH₃). We see no evidence for cresol formation from either reaction (<1% of any cresol isomer from m-xylene and <2% of any cresol isomer from p-cymene). Formation of 4-methylacetophenone, a product expected after H-atom abstraction from the CH(CH₃)₂ group, was observed from p-cymene, and a yield of 14.8 ± 3.2% measured. Inclusion of other products arising after H-atom abstraction from the methyl and isopropyl substituent groups indicates that H-atom abstraction from these substituent groups accounts for 20 ± 4% of the overall OH radical reaction for p-cymene. The formation yields of dimethylnitronaphthalenes have been measured from the reactions of OH radicals with 1,7- and 2,7-dimethylnaphthalene (these are the dominant dimethylnitronaphthalenes during daytime in the atmosphere) as a function of NO₂ concentration. The yield of 1,7-dimethylnitronaphthalenes from OH + 1,7-dimethylnaphthalene is ~0.25%, and that for 2,7-dimethylnitronaphthalenes from OH + 2,7-dimethylnaphthalene is ~0.07%, both with a dependence on the NO₂ concentration similar to that for naphthalene.

During year 4 of this program, we have investigated the behavior of PAHs and nitro-PAHs in ambient particles, collected on filters, exposed to O₃, OH radicals in the presence of NO_x, and NO₃ radicals in the presence of N₂O₅ and NO₂. This work has been conducted at UCR in collaboration with Prof. Staci Simonich of Oregon State University, and particulate matter from Riverside and other locations in the Los Angeles basin and from Bejing, China, has been exposed. Chemical and data analyses are still underway. However, for exposure to NO₃ radicals in the presence of N₂O₅ and NO₂, it appears to a first approximation that formation of nitro-PAHs occurs for freshly emitted particulate matter (e.g., collected in downtown Los Angeles during nighttime) while little or no change in PAHs or nitro-PAHs occurs for particulate matter that has been photochemically aged (e.g., particulate matter collected in Riverside). A more comprehensive picture will emerge after the chemical and data analyses are complete.

Additionally, in year 4 we have investigated the products formed from OH + cycloalkanes as a means to better understand the chemistry of cyclic radical intermediates. In particular, we investigated whether or not the cycloalkoxy radicals undergo isomerization (mass spectral peaks in the SOA from OH + cyclododecane and cyclopentadecane suggest the presence of products formed subsequent to cycloalkoxy radical isomerization). We investigated the products from cyclohexane, cycloheptane, cyclooctane and cyclodecane using API-MS and gas chromatography. The formation of cyclic hydroxyketones (expected products of the cycloalkoxy radical isomerization) were observed for the cycloheptane, cyclooctane and cyclodecane reactions, but not from cyclohexane. Formation yields of the cycloketones (formed from cycloalkoxy + O₂) and cyclic hydroxyketones were measured, and it appears that isomerization of cycloalkoxy radicals is important for ³C₇ cycloalkanes.

The approach to SOA studies has built on our growing understanding of gas-phase chemistry and has sought to determine the specific gas-phase reaction products and conditions that lead to SOA formation. The chemistry is complicated because products of OH radical-initiated reactions of aromatics are influenced by concentrations of NO, NO₂, RO₂• and HO₂ radicals. NO₂ can compete with O₂ in reactions of the aromatic-OH adduct, and NO, NO₂, RO₂•, and HO₂ radicals can all compete with each other in reactions with RO₂• radicals. Furthermore, it appears that oligomers are a major component of SOA and the reactions that lead to their formation are complex and influenced by the specific mix of gas-phase products, RH, and acidity. In years 2 and 3 of this project, a large number of environmental chamber reactions were carried out to investigate the yields of SOA formed from OH radical-initiated reactions of m-xylene, 3-methylfuran, and 2-methylfuran under different NO, NO₂, and RH regimes. It is currently known that a variety of products are formed from aromatic reactions. Two of these are aromatic products that include aldehydes and phenolic compounds. Non-aromatic products include bicyclic peroxides, diunsaturated dicarbonyls and unsaturated epoxydicarbonyls formed by ring opening, and 1,2-dicarbonyls and their unsaturated 1,4-dicarbonyl co-products formed by ring opening and further decomposition. The major products identified to date for the m-xylene reaction are (yields in parentheses) aromatic aldehydes (<10%), methylglyoxal (40%) and its unsaturated 1,4-dicarbonyl co-products 2-methyl-2-butene-dial (5-14%) and 4-oxo-pentenal (~12-34%), glyoxal (~6%) and its co-product (1-2%), and phenolic compounds (~10%). It has been thought that highly multifunctional, second-generation products formed by the reactions of the unsaturated 1,4-dicarbonyls with OH radicals might be significant contributors to SOA formation. To investigate this possibility, the yields of SOA formed from the reactions of m-xylene were compared with those formed from similar reactions of 3-methylfuran and 2-methylfuran, which form 2-methyl-2-butene-dial and 4-oxo-pentenal in higher yields than from the m-xylene reaction. SOA yields/reacted unsaturated 1,4-dicarbonyl were estimated from the measurements and it was observed that under all conditions the SOA yields from the m-xylene reactions were much larger (often by more than an order of magnitude) than could be explained by the products of unsaturated 1,4-dicarbonyl reactions. This indicates that the products formed from reactions of unsaturated 1,4-dicarbonyls were not by themselves responsible for SOA formation from m-xylene. This conclusion was supported by real-time and temperature-programmed thermal desorption particle mass spectra, which showed that the SOA formed from the m-xylene and methylfuran reactions had very different compositions. In year 4, measurements of UV-Vis spectra, elemental composition, functional groups, and mass spectra, followed by data interpretation that built on the results of the gas phase products analysis described above, showed that the SOA formed in the furan and methylfuran reactions with OH radicals is composed primarily of oligomeric esters formed by the particle phase reactions of the first generation unsaturated oxoacid and hydroxylactone products. This is one of the few studies to definitively identify oligomeric products in SOA (most are highly complex mixtures) and also demonstrates the facile nature of ester forming reactions.

It has often been reported that SOA formation from the reactions of aromatics is much higher in the absence of NO, with the usual explanation being that under these conditions organic peroxides are formed that contribute significantly to SOA formation. This hypothesis has never been tested, so in year 3 we measured the mass of organic peroxides in SOA formed from OH-initiated reactions of toluene, m-xylene, p-xylene, 1,3,5-trimethylbenzene, and benzaldehyde in the presence and absence of NO using a method we developed previously. Results were quite consistent, with SOA mass fractions of organic peroxides of 5.7–6.8% for reactions in the presence of NO and 16.4–21.3% in the absence of NO. These results provide the first direct evidence that organic peroxides contribute significantly to SOA formation from aromatic oxidation, especially in the absence of NO, but that other components are also important. One potential explanation for the other components is that methylglyoxal and/or phenols, which are not formed in the methylfuran reactions, or some unknown products, are important for SOA formation, either alone or through oligomer-forming reactions with the products of the unsaturated 1, 4-dicarbonyl reactions. This possibility was explored by adding methyglyoxal and phenols separately to a chamber containing the products of methyfuran reactions, but no additional SOA was formed. Having ruled out a role for unsaturated 1,4-dicarbonyls and 1,2-carbonyls in SOA formation, in year 4 additional experiments and data analysis were carried out to attempt to identify the SOA products of aromatic oxidation. Interesting similarities and differences in the mass spectra and thermal desorption profiles of SOA formed from the reactions of toluene, m-xylene, p-xylene, and 1,3,5-trimethylbenzene with OH radicals in the presence and absence of NO_x provide strong evidence that the SOA consists of a mix of first and second generation aromatic carbonyls and phenolic compounds, bicyclic compounds, and ring-opened intact diunsaturated dicarbonyls and unsaturated epoxydicarbonyls and their hydroxycarbonyl OH reaction products. Whereas the hydroxycarbonyls may exist in the particles as monomers, it is more likely that all the compounds are present in the form and hemiacetals and peroxyhemiacetals, with the ring-opened dicarbonyls playing a key role and again demonstrating the important role of oligomers in SOA formation from aromatic oxidation reactions.

A variety of studies also was carried out in years 3 and 4 on the effects of RH, acidity, and NH₃ on SOA formation, with data analysis ongoing. SOA formation from the reactions of benzaldehyde (a toluene-OH radical reaction product) with OH radicals in the absence of NO_x also has been explored, with HPLC analyses carried out to identify SOA products. Although the products are clearly aromatic, they do not correspond to any that would be expected from known chemistry: benzoic acid, benzoyl peroxide, benzyl anhydride, or peroxybenzoic acid (the latter compound was synthesized while the other standards were commercially available). This is a fascinating result, and it is hoped that further study will identify these products and the apparently unusual reaction pathways by which they are formed.

All but a few of the originally proposed studies have been completed (see below for remaining studies), and considerable progress has been made toward meeting goals. This is especially the case in the areas of gas phase product identification, quantification, and kinetics. SOA studies have focused primarily on effects of conditions on yields from OH radical reactions with aromatics and related methylfurans, and a variety of methods for product identification have been employed, including mass spectrometry, spectrophotometry, and functional group and elemental analysis. We currently believe we know which first and second generation products participate in SOA formation and the mechanisms by which this occurs. The experimental data obtained in this project thus far include reaction rate constants, product branching ratios, and yields of a number of products from the OH radical-initiated reactions of aromatic hydrocarbons and related methylfurans. In upcoming experiments it is expected that the identity of particle phase products will be confirmed and that data on particle phase reactions needed to model their formation via oligomerization will be obtained. These data can be used by atmospheric modelers as inputs into detailed chemical mechanisms, which in turn can be used directly or after condensation in urban and regional airshed computer models.

Future Activities:

We will re-determine the formation yields of unsaturated 1,4-dicarbonyls from the OH radical-initiated reactions of selected aromatic hydrocarbons using the PFBHA-coated denuder technique, by adding 2,5-hexanedione as an internal standard into the chamber after the reaction (the 2,5-hexanedione concentrations can be measured accurately by GC-FID without derivatization), as we have done in our recent OH + furans study. We also will conduct experiments to further investigate SOA formation from OH radical-initiated reactions of toluene, m-xylene, p-xylene, 1,3,5-trimethylbenzene, and nonylbenzene in the presence and absence of NO_x. A major focus will be the role of oligomer formation, which will be probed via environmental chamber studies of these reactions in the presence of alkenes (which we know form hemiacetals in SOA) and solution studies of hemiacetal and peroxyhemiacetal formation from reactions of unsaturated carbonyls with alcohols, phenols, and hydroperoxides. Joint studies of SOA formation using real-time particle and gas mass spectrometric analysis also will be carried out to further investigate oligomer formation from unsaturated dicarbonyls. The experiments with nonylbenzene, which will form much less volatile products than the other aromatics, will allow us to investigate whether the reason that products of OH reactions with unsaturated 1,4-dicarbonyls do not form SOA is solely due to their volatility. If so, then they may form SOA if compounds are present with which they can form oligomers. Work also will continue on the preparation of a significant number of additional new manuscripts for publication.

Journal Articles on this Report : 8 Displayed | Download in RIS Format

Publications Views
Other project views:	All 34 publications	16 publications in selected types	All 16 journal articles

Publications
Type	Citation	Project	Document Sources
Journal Article	Arey J, Obermeyer G, Aschmann SM, Chattopadhyay S, Cusick RD, Atkinson R. Dicarbonyl products of the OH radical-initiated reaction of a series of aromatic hydrocarbons. Environmental Science & Technology 2009;43(3):683-689.	R833752 (2008) R833752 (2009) R833752 (2010) R833752 (2011) R833752 (Final)	Abstract from PubMed Full-text: ES&T-Full Text HTML Exit Abstract: ES&T-Abstract Exit Other: ES&T-Full Text PDF Exit
Journal Article	Aschmann SM, Arey J, Atkinson R. Extent of H-atom abstraction from OH + p-cymene and upper limits to the formation of cresols from OH + m-xylene and OH + p-cymene. Atmospheric Environment 2010;44(32):3970-3975.	R833752 (2010) R833752 (2011) R833752 (Final)	Full-text: ScienceDirect-Full Text HTML Exit Abstract: ScienceDirect-Abstract Exit Other: ScienceDirect-Full Text PDF Exit
Journal Article	Aschmann SM, Arey J, Atkinson R. Reactions of OH radicals with C₆-C₁₀ cycloalkanes in the presence of NO:isomerization of C₇-C₁₀ cycloalkoxy radicals. Journal of Physical Chemistry A 2011;115(50):14452-14461.	R833752 (2011) R833752 (Final)	Abstract from PubMed Full-text: ACS-Full Text HTML Exit Abstract: ACS-Abstract Exit Other: ACS-Full Text PDF Exit
Journal Article	Aschmann SM, Nishino N, Arey J, Atkinson R. Kinetics of the reactions of OH radicals with 2-and 3-methylfuran, 2,3-and 2,5-dimethylfuran, and E-and Z-3-hexene-2,5-dione, and products of OH + 2,5-dimethylfuran. Environmental Science & Technology 2011;45(5):1859-1865.	R833752 (2010) R833752 (2011) R833752 (Final)	Abstract from PubMed Full-text: ES&T-Full Text HTML Exit Abstract: ES&T-Abstract Exit Other: ES&T-Full Text PDF Exit
Journal Article	Nishino N, Arey J, Atkinson R. Formation and reactions of 2-formylcinnamaldehyde in the OH radical-initiated reaction of naphthalene. Environmental Science & Technology 2009;43(5):1349-1353.	R833752 (2008) R833752 (2009) R833752 (2010) R833752 (2011) R833752 (Final)	Abstract from PubMed Full-text: ES&T-Full Text HTML Exit Abstract: ES&T-Abstract Exit Other: ES&T-Full Text PDF Exit
Journal Article	Nishino N, Arey J, Atkinson R. Yields of glyoxal and ring-cleavage co-products from the OH radical-initiated reactions of naphthalene and selected alkylnaphthalenes. Environmental Science & Technology 2009;43(22):8554-8560.	R833752 (2009) R833752 (2010) R833752 (2011) R833752 (Final)	Abstract from PubMed Full-text: ES&T-Full Text HTML Exit Abstract: ES&T-Abstract Exit Other: ES&T-Full Text PDF Exit
Journal Article	Nishino N, Arey J, Atkinson R. Formation of nitro-products from the gas-phase OH radical-initiated reactions of toluene, naphthalene and biphenyl: effect of NO₂ concentration (erratum). Environmental Science & Technology 2010;44(9):3644-3645.	R833752 (2011) R833752 (Final)	Full-text: ES&T-Full Text HTML Exit Abstract: ES&T-First page Exit Other: ES&T-Full Text PDF Exit
Journal Article	Nishino N, Arey J, Atkinson R. Formation yields of glyoxal and methyglyoxal from the gas-phase OH radical-initiated reactions of toluene, xylenes, and trimethylbenzenes as a function of NO₂ concentration. The Journal of Physical Chemistry A 2010;114(37):10140-10147.	R833752 (2010) R833752 (2011) R833752 (Final)	Abstract from PubMed Full-text: ACS-Full Text HTML Exit Abstract: ACS-Abstract Exit Other: ACS-Full Text PDF Exit

Supplemental Keywords:

absorption, chemicals, environmental chemistry, global climate, oxidants, particulates, PAHs, regional and climate models, toxics, tropospheric, VOC

Relevant Websites:

http://www.envisci.ucr.edu/faculty/arey.html Exit
http://www.envisci.ucr.edu/faculty/atkinson.html Exit
http://www.envisci.ucr.edu/faculty/ziemann.html Exit

Progress and Final Reports:

Original Abstract

The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.