Organic acids in Southern California air: ambient concentrations

Organic acids in Southern California air: ambient concentrations, mobile source emissions, in situ formation and removal processes. Daniel Grosjean. E...
0 downloads 0 Views 2MB Size
Environ. Sci. Technol. 1989, 23, 1506-1514

Organic Acids in Southern California Air: Ambient Concentrations, Mobile Source Emissions, in Situ Formation and Removal Processes Daniel Grosjean DGA, Inc., Suite 205, 4526 Telephone Road, Ventura, California 93003

Ambient levels of organic acids were determined at a California South Coast Air Basin (SCAB) smog receptor site. The most abundant acids, by a substantial margin, were formic acid and acetic acid, with 4- and 8-h-averaged concentrations of 1-13 and 2-16 ppb, respectively. Both acids were present in the atmosphere mostly in the gaseous form. On the average formic acid levels exceeded those of acetic acid, though by a small margin (average ratio of 1.1). Both acids exhibited nighttime maxima. Levels of formic acid and acetic acid each exceeded those of ozone during the 12-h period, 8 p.m. to 8 a.m. Ambient levels of organic acids reflect both direct emissions and in situ formation. Estimates are proposed for emissions from SCAB mobile sources: 6500 kg/day for formic acid (range 1200-13 000 kg/day), 9000 kg/day for acetic acid, 2900 kg/day for other monocarboxylic acids, 400 kg/day for aliphatic dicarboxylic acids, and lo00 kg/day for aromatic acids, for a total of -20000 kg/day. No emission rates could be derived for SCAB stationary sources. Considerations of mechanisms and kinetic data for several reactions, and of the chemical structure and emission rates of the possible acid precursors, point to the ozone-olefin reaction as a major source of organic acids in the atmosphere. In situ reactions produce more formic acid than acetic acid, along with lesser amounts of other carboxylic acids. Net in situ production rates in the SCAB are comparable to emission rates from motor vehicles, i.e., 15000-20000 kg/day. Removal of organic acids from the atmosphere by chemical reactions is negligible. The two important removal processes are dry deposition (gas-phase acids) and removal by rain (particulate-phase acids). Overall, dry deposition is more important than removal by rain and accounts for 92% of the total organic acid deposition budget. SCAB deposition fluxes, in nmol m-2 year-', are 67 for formic acid (of which 95% is by dry deposition), 70 for acetic acid (91% dry deposition), 6.7 for propionic acid (95% dry deposition), and 1.7 for oxalic acid (11% dry deposition). Comparison is made of direct emissions, in situ for formation, and removal processes. Removal estimates exceed production estimates by a factor of about 2-4. This discrepancy may reflect an overestimate of dry deposition velocities, an underestimate of production pathways (direct emissions and/or in situ formation), or both.

-

Introduction Much of the research and regulatory efforts concerning acid deposition have focused in the past on the strong inorganic acids, sulfuric and nitric, and on their respective precursors, sulfur dioxide and oxides of nitrogen. Organic acids have received much less attention. However, recent studies of precipitation chemistry have shown that organic acids may account for a large fraction, up to 64%, of the total acidity in nonurban environments (1). Indeed, organic acids including formic acid and acetic acid now appear ubiquitous in the atmosphere (2-7). In urban southern California, several studies have included reports of organic acids in engine exhaust, ambient air (gas phase and particle bound), and precipitation samples (8-15). Pierson and Brachaczek (16) measured 1506 Environ. Sci. Technol., Vol. 23, No. 12, 1989

organic acids in dew collected at Glendora, CA. Formate was the most abundant of all anions present, organic or inorganic. The main objective of the present study was to contribute additional measurements that would augment the data base from which the role of organic acids in urban pollution and acid deposition could be better understood. Specific issues to be addressed include the nature and ambient levels of organic acids in the South Coast Air Basin, their diurnal variations and phase distribution, and the relationships between organic acids and other indicators of air quality. These results are used to estimate the relative importance of direct emissions, in situ formation in the atmosphere, and removal processes for organic acids in the California South Coast Air Basin (SCAB).

Experimental Methods A detailed description of the sampling and analytical methods employed in this study is already available (14, 17). Only a brief summary is given below. Air samples were collected on August 11-21, 1986, on the Citrus College campus in Glendora, CA, (-40 km east-northeast of downtown Los Angeles) as part of the Carbon Species Measurements Comparison Study (CSMCS),a large multilaboratory air quality study carried out at the Glendora site. Our sampling schedule was that adopted for CSMCS, i.e., 0-8, 8-12, 12-16, 16-20, and 20-24 PDT. The samples were collected -2.5 m above ground by use of open-face 47-mm-diameter dual filter holders connected to a calibrated flowmeter and an air sampling pump. The filter packs employed included a Sartorius 47-mm, 1,2-pm pore size Teflon filter (upstream) and a Gelman AE 47-mm glass fiber filter impregnated with potassium hydroxide (downstream). The collection efficiency of alkaline-impregnated filters for organic acids has been previously established (14). Immediately after sampling, the filters were placed individually in dark (amber) glass vials capped with Teflon-lined screwcaps and containing 10 mL of deionized water and 40 pL of HPLC-grade chloroform added as a biocide. The addition of a biocide is critical to the stability of the samples. Upon return to the laboratory, the samples were sonicated and analyzed directly by liquid chromatography with ultraviolet detection (17). Calibration involved the use of external standards, i.e., dilute solutions of formate, acetate, etc. Replicate samples, filter blanks, and field controls were handled and analyzed in the same way. Analytical detection limits were 12 and 34 ng (100 pL injected) for formate and acetate, respectively. The corresponding ambient air detection limits (4-h samples) were 0.29 and 0.45 ppb, respectively. The relative standard deviation (RSD) for replicate analyses of calibration standards was 2.0% for formate and 3.0% for acetate (n = 15). The RSD for replicate analyses of field samples was 7.3% for formate ( n = 10) and 8.5% for acetate (n = 9). Other analytes, unavoidably present in air samples collected on alkaline traps did not interfere under the conditions employed. These included chloride from HC1, nitrate from HN03, carbonate and bicarbonate from COz,

0013-936X/89/0923-1506$01.50/0

0 1989 American Chemical Society

Table I. Ambient Levels (Gas Phase) of Formic Acid and Acetic Acid

date (1986)

F

A

F

A

sampling perioda 16-20 F A

8/12-13 8/13-14 8/14-15 8/15-16 8/16-17 8/17-18 8/18-19 8/19-20 8120-21

4.5 1.6

7.0 3.5 4.7 4.4 5.0 4.1 2.2 4.1 5.9

2.3 2.7 1.3 6.8 3.0 4.5 3.1 3.2 4.0

3.0 2.9 3.4 4.1 2.2 2.9 2.5 2.1 2.5

1.8 1.3 3.8 5.7 5.0 4.1 2.0 3.0 3.6

8-12

5.5 5.5 4.7 5.3 5.6 8.0

formic acid acetic acid formic + acetic acid ozone nitric acidb hydrogen chlorideb

12-16

20-24

3.5 3.4 5.1 4.5 2.8 3.2 1.9 2.4 2.4

0-8

F

A

F

A

3.5 2.2 4.1 3.6 2.7 4.6 3.5 5.5 2.2

3.0 4.1 4.4 3.3 3.6 3.8 2.1 4.2 3.2

3.7 2.7 7.5 3.3 5.3 4.4 7.2 13.0 10.0

3.4 2.6 2.8 3.3 5.0 4.1 4.1 8.7 16

8-12

12-16

av concn, ppb 16-20

20-24

0-8

5.10 4.54 9.64 48 1.72 1.55

3.40 2.84 6.24 171 6.48 1.39

3.36 3.24 6.60 89 3.84 1.04

3.53 3.52 7.05 3.5 0.82 1.03

6.34 5.55 11.89 5.1 0.42 0.50

O F = formic acid, ppb (1ppb = 1.88 pg m-3). A = acetic acid, ppb (1 ppb = 2.45 pg m-3), not corrected for possible positive bias due to alkaline decomposition of PAN; see text. bFrom Grosjean et al. (17).

and nitrite from partial retention of NO2and from alkaline decomposition of peroxyacetyl nitrate, PAN. Retention of formaldehyde and acetaldehyde as formate and acetate, respectively, was negligible. Earlier studies, discussed in detail by Grosjean (14), showed no evidence for such an artifact. In this study, levels of formic acid and acetic acid measured downstream of DNPH denuders, which remove ambient aldehydes, were identical with those measured without denuders (17). Alkaline filters are rapidly neutralized by atmospheric C02,as evidenced by both pH and carbonate content measurements of filter extracts after sampling (17). This does not affect organic acid collection efficiency, as demonstrated by side-by-side measurements with alkaline and carbonate-impregnated filters (I7), but obviously makes aldehyde interference (if any) negligible, since aldehydes undergo disporportionation to organic acids only in strong alkaline media (14). Levels of formic acid measured in this study with alkaline traps agreed well with those measured at the same time and location by Fourier transform infrared spectroscopy, a method not subject to sampling interferences (Grosjean, Tuazon, and Fujita, unpublished results). Alkaline decomposition of PAN yields acetate. PAN was measured independently by electron capture gas chromatography (17). These measurements, together with liquid chromatography measurements of nitrite and acetate, were used to estimate the maximum acetate bias due to PAN. The maximum bias thus estimated ranged from 19.8 to 34.8% with an average of 22.5% (17). These values are in fact upper limits, since rapid neutralization of the alkaline trap by atmospheric COz results in negligible decomposition of PAN to acetate. For example, PAN does not decompose in annular denuders coated with carbonate or bicarbonate (Grosjean et al., unpublished results). In addition, no close association was found between acetic acidlformic acid concentration ratios and ambient PAN as would be the case if PAN, which decomposes levels to acetate but not formate in alkaline media, had accounted for a significant fraction of the measured acetate.

(In,

Results and Discussion Ambient Concentrations (Gas Phase). Formic and acetic acids were, by a substantial margin, the most

Table 11. Concentration Frequency Distributions of Gas-Phase Formic Acid and Acetic Acid concn, range, ppb 0-1

1.1-2 2.1-3 3.1-4 4.1-5

no. of obsvns formic acetic acid acid 0 5 8 10

8

0 1

15 11 13

concn, range, ppb 5.1-6 6.1-7 7.1-8 8.1-9 >9

no. of obsvns formic acetic acid acid 7 1

3 0 2

2 1 0 1 1

abundant organic acids. Ambient levels of other organic acids, including aliphatic monocarboxylic, aliphatic dicarboxylic, and aromatic acids were below 1ppb. Levels of formic acid ranged from 1.3 to 13 ppb, those of acetic acid ranged from 1.9 to 16 ppb (Table I). Sampling period averaged values (4- or 8-h averages) ranged from 3.4 to 6.3 ppb for formic acid, from 2.8 to 5.5 ppb for acetic acid, and from 6.2 to 11.9 ppb for the sum of formic and acetic acids (13.1-25.0 pg m-3). Also listed in Table I are the period-averaged coneentrations of ozone and of the inorganic acids, nitric acid and hydrogen chloride (17). Formic acid and acetic acid each exceeded the sum of the inorganic acids during the 16-h period 8 p.m. to 12 noon, and together accounted for 74% (ppb basis) of the total gas-phase acids during the study period. It is also interesting to note that, during the entire study, levels of formic acid and acetic acid each exceeded those of ozone during the 12-h period 8 p.m. to 8 a.m. Concentration-frequency distributions are given in Table I1 and exhibit frequency maxima at 3-5 ppb. On the average, levels of formic acid exceeded those of acetic acid, though by a small margin. Concentration ratios are given in Table I11 for each sampling period. Averaged formic acid to acetic acid concentration ratios ranged from 1.00 to 1.14. Diurnal variations are plotted in Figure 1 and are seen to exhibit late evening and nighttime maxima. This diurnal behavior has been reported before for both formic acid (11, 14) and acetic acid (14) at southern California smog receptor sites, on the basis of measurements made over time intervals shorter than those employed in this Environ. Sci. Technol., Vol. 23, No. 12, 1989

1507

Table 111. Formic Acid to Acetic Acid Gas-Phase Concentration Ratios day

8-12

8/12-13 8/13-14 8/14-15 8/15-16 8/16-17 8/17-18 8/18-19 8j19-20 8/20-21 sampling period av

0.64 0.46 1.25 1.10 1.15 2.41 1.33 1.35 1.12

concentrationratios, ppb/ppb 12-16 1620 20-24 (t8 0.77 0.93 0.38 1.66 1.36 1.55 1.24 1.52 1.60 1.2 ~~~

0.51 0.38 0.74 1.27 1.78 1.28

1.17 0.54 0.93 1.09 0.75 1.21

1.09 1.04 2.68 0.98 1.08 1.07

1.05 ~.

1-67

1.77.

1.25 1.50 1.04

1.31 0.69 1.00

1.49 0.62 1.14

~

F)pun 1. D i m 1 variafkns of ambient lav& ofgasghase and acetk acM. Glendora. CA. August 12-21, 1986.

add

study. Diurnal variations of ozone, which exhibit daytime (midafternoon) maxima, are also included in Figure 1for comparison. Comparison with Literature Data. Only a few studies are available for urban levels of carboxylic acids outside of the Los Angela, CA, area. Hoshika (18)measured the Cz-C5 aliphatic monoacids in Nagoya, Japan. Typical concentrations were 2.5 ppb for acetic acid, 0.3 ppb ppb each for the C4 for propionic acid, and (4-50) X and C5isomers (butyric, isobutyric, valeric, and isovaleric). Dawson et al. (2)measured up to 3 ppb formic acid and up to 6 ppb of acetic acid in Tucson, AZ. Early studies carried out in the SCAB involved FT-IR measurements (8-11). Only formic acid can be detected by FT-IR, with a detection limit of 4 ppb for a -1-km optical path length. Hanst et al. (8)first reported up to 72 ppb of formic acid in Pasadena, CA, during the severe July 25, 1973, smog episode. Their results were subsequently corrected downward by a factor of 4 following redetermination of the absorption coefficient (9, 11). Formic acid levels measured by FT-IRare in the range