Environ. Sci. Technol. 2003, 37, 446-453
Air Quality Model Evaluation Data for Organics. 6. C3-C24 Organic Acids M . P . F R A S E R , * ,‡ G . R . C A S S , †,‡ A N D B. R. T. SIMONEIT§ Environmental Engineering Science Program, California Institute of Technology, Pasadena, California 91125, and College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, Oregon 97331
The atmospheric concentrations of 47 carboxylic acids in the semivolatile and particle phases are quantified in the Los Angeles area, as part of a larger study of the vaporphase, semivolatile, and particle-phase organic compounds. Variations in the spatial and temporal distributions of acid concentrations are analyzed to determine whether atmospheric formation or primary emissions are responsible for the observed levels. Relatively low molecular weight aliphatic dicarboxylic acids (e.g., butanedioic acid, hexanedioic acid, and propanedioic acid) and some n-alkanoic acids (e.g., n-octanoic acid and n-nonanoic acid) are found at an offshore sampling location at levels comparable to urban area concentrations indicating that these compounds or their atmospheric precursors may be derived from long-range transport or natural background sources. Some aromatic carboxylic acids (e.g., benzoic acid and 1,2-benzenedicarboxylic acid) have spatial and temporal distributions suggesting that formation from anthropogenic emissions of gaseous precursors dominates their atmospheric concentrations. Additionally, the distributions of aliphatic carboxylic acid concentrations known to be emitted from primary sources (e.g., hexadecanoic acid and octadecanoic acid) are consistent with direct emissions as the dominant source of these compounds.
Introduction Much of the research on monocarboxylic acid concentrations in the atmosphere has focused on the water soluble low molecular weight acids, including formic and acetic acids (1-7). These acids are present in the gas-phase, partition into the aqueous aerosol droplets, and thus are important to the chemistry of fogwater and cloud droplets. The present research focuses instead on the higher molecular weight acids which are not as water soluble. While present at lower ambient concentrations, these compounds partition between the vapor phase and organic particulate matter. Higher molecular weight acids constitute the most abundant class of organic compounds found in fine particulate matter in the urban atmosphere (8). Thus, they are of interest because of the need to reduce fine particulate matter concentrations * Corresponding author phone: (713)348-5883; fax: (713)348-5203; e-mail:
[email protected]. Present address: Department of Civil and Environmental Engineering, Rice University, Houston, TX 77005. † Deceased. ‡ California Institute of Technology. § Oregon State University. 446
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 3, 2003
in order to attain compliance with the National Ambient Air Quality Standard for fine particles (9). Organic acids are relatively stable in the atmosphere, with removal dominated by wet and dry deposition processes (10, 11). They are emitted to the atmosphere directly from natural and anthropogenic primary sources and are also produced by secondary formation from gas-phase hydrocarbons and certain oxygenated species via atmospheric chemical reactions (12, 13). Previous studies of the Southern California atmosphere show that aliphatic dicarboxylic acids display a diurnal variation that indicates secondary formation due to atmospheric chemical reactions (13, 14). Oxidation of cyclic olefins and diolefins is suggested as one possible formation route (15). Analysis of particulate matter collected during 1982 in Southern California shows that both aliphatic dicarboxylic acids and aromatic polycarboxylic acids display similar spatial and seasonal characteristics, suggesting that atmospheric formation may be the main source of aromatic polycarboxylic acids in the atmosphere as well (8). In a study of urban particulate matter in Tokyo, Kawamura and Ikushima (16) reported that C2-C4 aliphatic diacids are present at higher concentrations during the summer months, showing positive correlations between total dicarboxylic acid concentrations (normalized to total aerosol carbon) and oxidant concentrations, again suggesting that diacids are produced by reactions which occur in photochemical smog. However, the highest molecular weight aliphatic diacids (C6-C10) showed high concentrations in Tokyo in both summer and winter samples (16), leading to the question of whether the same routes of emissions and atmospheric chemical reaction govern the ambient concentrations of different homologues within this compound class. The role of primary emissions of both aliphatic and aromatic diacids was investigated in the Los Angeles area (17). Analysis of exhaust samples from gasoline and diesel engines showed ratios between aliphatic and aromatic diacids that are similar to those seen in the urban atmosphere, which along with the overall high emission rates measured in such studies suggest that direct emission from motor vehicles is an important contributor to the atmospheric concentrations of dicarboxylic acids (17, 18). The present study reports experimental results on the ambient concentrations of 47 organic acids in the Los Angeles atmosphere. These data have been collected as part of a larger study of the vapor-phase, semivolatile, and particlephase organic compounds (19, 20). Some of the questions raised above concerning the importance of primary emissions versus secondary formation of higher molecular weight organic acids will be addressed by statistical analyses. The spatial and diurnal variations in the ambient concentrations of individual acids will be compared to the ambient concentrations of other classes of organic compounds whose ambient concentrations are known to be governed by primary emissions or by secondary formation due to atmospheric chemical reactions.
Experimental Methods Collection of Samples. Samples were collected during the summer of 1993 at five sites in and around Southern California as part of an effort to measure the ambient concentrations of as many particle-phase, semivolatile, and vapor-phase organic compounds as possible for use in the evaluation of an airshed model that tracks the ambient concentrations of individual organic compounds (19, 20). Four urban sampling sites included Long Beach, near a high concentration of industrial emission sources in the Los Angeles Harbor, central 10.1021/es0209262 CCC: $25.00
2003 American Chemical Society Published on Web 12/18/2002
Los Angeles, near the center of the freeway system of Southern California, and Azusa and Claremont, two downwind, receptor locations in the inland valleys. Four-hour integrated samples centered within consecutive 6-h intervals were collected over a 2-day photochemical smog episode on Sept 8-9, 1993. Meteorological conditions during the experiment exhibited the low mixing depths and stagnant transport conditions that occur during the summer months in Southern California, leading to peak ozone concentrations approaching 0.3 ppm. A complete description of the experiment, inducing the samplers deployed, sample preparation, extraction techniques, and organic compound quantification methods is presented elsewhere (19, 20); only brief summaries will be given here. Organic acids were collected using two separate samplers: (1) a high-volume dichotomous sampler (21) with quartz fiber (QF) filters followed by polyurethane foam (PUF: ILD ) 30, density ) 0.022 g cm-3) cartridges to trap particle and vapor-phase organic compounds and (2) a lowvolume particulate matter sampler with potassium hydroxide (KOH) impregnated glass fiber filters placed in stacked filter holders downstream of Teflon particle prefilters. The high volume dichotomous sampler QF filters and PUF cartridges were employed to collect particle-phase and high molecular weight vapor-phase organic acids, respectively. While this sampling arrangement is prone to artifacts in the gas-particle partitioning of compounds, the goal of the study is to quantify the total concentration of acids present in the atmosphere, not to measure partitioning. The KOH impregnated filters deployed with the low-volume sampler were used to measure the higher concentrations of low-molecular weight gas-phase acids as reported in Nolte et al. (7). The QF filters (Pallflex QAO 2500) were baked at 550 °C for at least 8 h before use to reduce their contaminant blank values. These filters were stored in prebaked aluminum foil packets for transportation to the field experiment. After sampling, the filters were placed in prebaked glass jars, sealed with Teflon lined lids, and returned to the laboratory where they were stored at -21 °C until analysis. Polyurethane foam plugs (density ) 0.022 g cm-3, ILD ) 30) were cleaned by repeated washing in dichloromethane (DCM). Each PUF cartridge consisted of a bed of 7.6 cm diameter foam segments 12.5 cm deep packed within an aluminum housing. Each PUF cartridge was sealed with Teflon film, deployed to the field for sampling, resealed, returned to the laboratory where the foam was transferred into precleaned glass jars with Teflon lined lids, and frozen until analysis.
Analytical Methods The samples were extracted from the collection media, concentrated, and analyzed by gas chromatography-mass spectrometry (GC-MS) using a Finnigan 4000 system accompanied by a NOVA-4C data station. A complete description of the extraction and analysis techniques has been given previously (20, 22) and will only be summarized here. Before extraction, samples were spiked with the following perdeuterated compounds to monitor extraction efficiency and loss in the concentration steps: (1) for PUF sampless n-decane-d12, n-pentadecane-d32, n-tetracosane-d50, benzoic acid-d5, phenol-d5, hexanoic acid-d11, and decanoic acid-d19 and (2) for filter samplessn-tetracosane-d50. After extraction and concentration to approximately 400 µL, each filter and polyurethane foam extract was divided into two aliquots. One of these was derivatized with diazomethane to convert carboxylic acids into their methyl ester analogues for quantification by GC-MS. All carboxylic acids were quantified as their methyl ester derivatives. Analysis of the underivatized samples did not show any methyl esters of interest, removing
the possibility that the actual airborne compounds measured were originally present as methyl esters. The GC-MS analyses were carried out with a DB-1701 capillary column (30 m × 0.32 mm i.d.; bonded phase 86% dimethyl, 14% cyanopropylphenyl polysiloxane; J&W Scientific, Rancho Cordova, CA). Response of the GC-MS instrument was monitored by 1-phenyldodecane as a coinjection standard with each sample. Compound identification and quantification were accomplished by comparison to authentic standards when available and included the following: benzoic acid, 2-methylbenzoic acid, 3-methylbenzoic acid, 4-methylbenzoic acid, 2,3-dimethylbenzoic acid, 2,4dimethylbenzoic acid, 2,5-dimethylbenzoic acid, 2,6-dimethylbenzoic acid, 3,4-dimethylbenzoic acid, 3,5-dimethylbenzoic acid, dehydroabietic acid, 7-oxodehydroabietic acid, 1,2-benzenedicarboxylic acid, 1,3-benzenedicarboxylic acid, 1,4-benzenedicarboxylic acid, 4-methyl-1,2-benzenedicarboxylic acid, 1,2,3-benzenetricarboxylic acid, 1,2,4-benzenetricarboxylic acid, 1,3,5-benzenetricarboxylic acid, 1,2,4,5benzenetetracarboxylic acid, propanedioic acid, butanedioic acid, butenedioic acid, methylbutanedioic acid, pentanedioic acid, hexanedioic acid, octanedioic acid, nonanedioic acid, n-octanoic acid, n-decanoic acid, n-dodecanoic acid, ntetradecanoic acid, n-hexadecanoic acid, n-hexadecenoic acid, n-octadecanoic acid, n-octadecenoic acid, n-eicosanoic acid, and n-docosanoic acid. The probable identity of unknown compounds was determined by comparison to the NIST/EPA mass spectral libraries and by fundamental interpretation of the mass spectrometric fragmentation patterns. When authentic standards were not available, the response of the instrument to similar compounds was used to estimate the response factor and also to support the probable identification of the compound. Multiple injections of a known quantity of authentic organic standards were used to estimate the precision of the quantification by GCMS, which is approximately (24%. Analysis of field and laboratory blanks taken during this sampling program revealed traces of normal alkanoic acids on both QF filters and PUF cartridges and 1,2-benzenedicarboxylic (phthalic) acid on the QF filters. For alkanoic acids measured on the PUF, blank levels were of the same order of magnitude as the ambient measurements. For this reason, the data for normal alkanoic acids collected on polyurethane foam are suspect and are not reported. For phthalic acid, the mass of contamination was equivalent to an atmospheric concentration of 0.6 ng m-3 (using an average air sampling volume). Since these blank values are minor compared to the ambient concentrations, blank values are subtracted from mass quantified in all samples to report an atmospheric concentration. Corrections for extraction efficiency and blowdown losses were calculated by quantifying a perdeutereated recovery standard in each sample. Benzoic acid-d5 and n-tetracosaned50 were used for the recovery standard for PUF samples and QFF samples, respectively. The ambient concentrations reported here include these corrections for recovery and extraction efficiencies as estimated by the perdeuterated standards. Recovery during extraction and blowdown ranged from 21% to 110% between samples and averaged 61% for all QF filter samples; it ranged from 42% to 98% between samples and averaged 65% for samples collected by PUF plugs. The loss value for each individual sample, and not the average recovery, was used to correct ambient concentrations measured in that sample.
Results and Discussion The ambient concentrations of 24 alkanoic, 3 alkenoic, and 20 aromatic acids, including resin acids and polycarboxylic acids, were measured during this experiment. The averages and ranges of their concentrations are listed in Table 1 for VOL. 37, NO. 3, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
447
TABLE 1. Average (and Range) of Urban Ambient Concentrations of High Molecular Weight Carboxylic Acids Measured at Four Urban Sites in Southern California over the Period September 8-9, 1993 vapor phase semivolatile organic acids by PUF sampling (ng/m3)
compound
fine particulate organic acids (ng/m3)
n-Alkanoic Acids n-octanoic acid n-nonanoic acid n-decanoic acid n-undecanoic acid n-dodecanoic acid (lauric acid) n-tridecanoic acid n-tetradecanoic acid (myristic acid) n-pentadecanoic acid n-hexadecanoic acid (palmitic acid) n-heptadecanoic acid n-octadecanoic acid (stearic acid) n-nonadecanoic acid n-eicosanoic acid n-heneicosanoic acid n-docosanoic acid n-tricosanoic acid n-tetracosanoic acid
n.r.a n.r.a n.r.a n.r.a n.r.a
2.4 (0.9-4.8) 4.9 (2.1-11.7) 3.2 (1.7-5.3) 3.1 (0.0-18.7) 12.1 (3.4-33.8)
n.r.a n.r.a
4.4 (1.2-7.7) 22.1 (7.7-50.8)
n.r.a n.r.a
8.7 (1.5-27.4) 72.6 (10.2-196.7)
n.r.a n.r.a
3.2 (0.5-13.9) 33.8 (4.9-99.3)
n.r.a n.r.a n.r.a n.r.a n.r.a n.r.a
1.7 (0.0-7.0) 3.6 (0.0-12.1) 0.9 (0.0-4.3) 2.6 (0.0-8.7) 1.0 (0.0-6.6) 2.6 (0.0-6.3)
n-Alkenoic Acids n.r.a n.r.a
n-hexadecenoic acid n-octadecenoic acid (oleic acid)
2.9 (0.0-7.7) 7.7 (0.0-34.2)
Alkanedioic Acids propanedioic acid butanedioic acid methylbutanedioic acid pentanedioic acid hexanedioic acid octanedioic acid nonanedioic acid
38.6 (0.3-473.5) 58.5 (0.5-440.1) 13.8 (0.2-115.8) 20.4 (1.9-195.8) 7.5 (0.0-24.1) 1.4 (0.2-2.0) 2.9 (0.3-7.0)
Alkenedioic Acids butenedioic acid
0.3 (0.0-1.1)
Aromatic Carboxylic Acids benzoic acid 2-methylbenzoic acid 3-methylbenzoic acid 4-methylbenzoic acid 2,4-dimethylbenzoic acid 2,5-dimethylbenzoic acid 2,6-dimethylbenzoic acid 3,4-dimethylbenzoic acid 3,5-dimethylbenzoic acid 1,2-benzenedicarboxylic acid (phthalic acid) 1,3-benzenedicarboxylic acid 1,4-benzenedicarboxylic acid 4-methyl-1,2-benzenedicarboxylic acid 1,2,3-benzenetricarboxylic acid 1,2,4-benzenetricarboxylic acid 1,3,5-benzenetricarboxylic acid 1,2,4,5-benzenetetracarboxylic acid 2,6-naphthalenedicarboxylic acid
299.7 (41.5-962.2) 10.2 (0.7-35.7) 75.0 (3.7-322.8) 57.8 (4.3-253.8) 21.6 (2.0-91.8) 25.8 (4.9-84.5) 6.1 (1.4-20.2) 52.2 (5.8-188.4) 52.0 (4.9-277.8) 80.2 (4.1-242.4) 4.9 (0.3-12.9) 5.4 (0.9-17.2) 54.1 (1.2-262.8) 14.8 (0.7-74.5) 18.9 (2.0-52.7) 0.3 (0.0-0.9) 1.0 (0.0-5.1) 2.4 (0.5-6.3)
Resin Acids dehydroabietic acid 7-oxodehydroabietic acid a
0.4 (0.0-3.4)
