Nog

M ^

Jf

00:00 00:00 00:00 00:00 00:00 9/24/93 9/25/93 9/26/93 9/27/93 9/28/93

FIGURE 11.28 Ratio of sum of individual compounds (NO + N02 + PAN + PPN + HNO3 + particulate nitrate) to total measured NO,, under two types of overall meteorological conditions: (a) episodes with winds from the south and east with fresh emissions and (b) winds primarily from the west with cleaner but more aged air (adapted from Williams et al, 1997).

00:00 00:00 00:00 00:00 00:00 9/24/93 9/25/93 9/26/93 9/27/93 9/28/93

FIGURE 11.28 Ratio of sum of individual compounds (NO + N02 + PAN + PPN + HNO3 + particulate nitrate) to total measured NO,, under two types of overall meteorological conditions: (a) episodes with winds from the south and east with fresh emissions and (b) winds primarily from the west with cleaner but more aged air (adapted from Williams et al, 1997).

to cr to cr

Easterly upslope flow

Westerly downslope flow

FIGURE 11.29 Ratio of measured individual compounds of NO,, to total NO,, at Idaho Hill, Colorado, with easterly winds (more polluted air) and with westerly winds (cleaner but more aged air), respectively (data from Williams et al, 1997).

Easterly upslope flow

Westerly downslope flow

FIGURE 11.29 Ratio of measured individual compounds of NO,, to total NO,, at Idaho Hill, Colorado, with easterly winds (more polluted air) and with westerly winds (cleaner but more aged air), respectively (data from Williams et al, 1997).

conditions of westerly flow (but where the air was more aged), the total NOy was smaller, 1.4 ± 0.4 ppb, but only 77% was accounted for by the individual compounds. Interestingly, the deficit appeared to correlate with 03, suggesting the compounds responsible are photochemically generated. A similar observation has been made in Denmark by Nielsen et al. (1995, 1998), who also report that the fraction of total NO that is in the form of particulate nitrate is small (0.17-0.28%). The deficit in accounting for NO>( at Idaho Hill also decreased as the air temperature decreased, which may reflect a correlation of temperature with the age of the air mass and/or that the species responsible for the missing NO are thermally unstable (Williams et al., 1997).

As seen from the VOC-NOx chemistry in Chapter 6, organic nitrates are among the expected products of the oxidation of hydrocarbons in air containing NOx. Williams et al. (1997) have considered the possible contribution of simple alkyl nitrates but, based on other measurements of these species, indicate that it is unlikely they are responsible for a significant portion of the "missing NO ."

Multifunctional organics are also possible contributors. Nielsen et al. (1998) have examined the possible contribution of multifunctional compounds to "missing NO " in both the gas and particle phases. As discussed in Chapter 9, compounds with sufficiently high vapor pressures (> 2xl0~5 Torr) exist essentially completely in the gas phase, those with low vapor pressures (< 2 X 10"9 Torr) in the condensed phase (i.e., on or in particles), and those in between the two extremes as both gases and particles. Nielsen and co-workers have developed a relationship between the expected vapor pressure of a multifunctional compound and its structure:

— (2.144 + 0.070)(no. of nitrate groups) -(1.961 + 0.057)(no. of OH groups)

This relationship is based on data for 183 compounds, including C7-C29 hydrocarbons, C,-C18 alcohols, C2-Cl0 diols, C5-C18 carbonyls, C,-C20 alkyl nitrates, and C2-C3 hydroxynitrates and dinitrates. Based on this analysis, Nielsen et al. (1998) suggest that the organic nitrates found in particles are probably bi- and multifunctional compounds and that they may also contribute to gas-phase NO and NOz.

Such multifunctional compounds, however, are very difficult to collect, identify, and quantify and, in fact, need to be specifically targeted if they are of interest for a particular study. As a result, such compounds usually go undetected but may be responsible for some of the "missing NO ."

In addition, given that the efficiency of conversion of compounds such as HCN and NH3 over the catalysts may be higher than thought under some conditions (e.g., Kliner et al., 1997; Weinheimer et al., 1998; Brad-shaw et al., 1998), these compounds may also be responsible for a substantial portion of the "missing NO ." However, Williams et al. (1998) argue that such interferences, if they exist in their measurements, are too small to account for the magnitude of the "missing NO>(" in their studies.

Because of the sensitivity of NO>( measurements to the particular catalyst used, its recent exposure, cleaning, etc., agreement between various measurements of

NO,, and between NO,, and the sum of individual

compounds would not necessarily be expected, especially in aged air masses and/or other types or air masses where compounds other than NO and N02 contribute significantly to NOr Indeed, this is the case (e.g., see discussion by Bradshaw et al. (1998) and Williams et al. (1998)). Agreement is generally reasonably good at higher concentrations and when NOx is a major portion of NO , e.g., in urban and suburban areas (Williams et al., 1998).

(3) NOx and N02 NOx is defined as the sum of (NO + N02). NO can be measured by the techniques described earlier. N02 is one of the compounds contributing to NO and in a relatively "young" air mass is often the primary contributor. However, separating out its contribution from other compounds contributing to NO>( obviously requires a different approach.

One approach that has been used is to photolyze the N02 at wavelengths below 400 nm to form NO and then measure the NO using chemiluminescence or TP-LIF as discussed earlier (Kley and McFarland, 1980; Ridley et al., 1988; Gao et al., 1994). The reactions are as follows:

NO + O, —^ NOf + 02 (in gas phase), (9a) N02 + 02 (on cell wall). (9b)

From the differential equations for the change in N02 and NO with time, i.e., d[N02]/dt and d[NO]/dt, based on reactions (8), (9a), and (9b), it can be shown that the fractional conversion of NOz to NO is given by Eq. (N) (Kley and McFarland, 1980; Gao et al., 1994):

k is the photolysis rate constant for N02, reaction (8), ky = kt).d + ki)h, and r is the residence time of the air in the photolysis cell. Fractional conversions of up to ~0.65 have been observed ((Kley and McFarland, 1980; Ridley et al., 1988; Gao et al., 1994). Photolysis of NOz at 353 nm using a XeF excimer laser has also been used (Sandholm et al., 1990). Measurement precision and detection limits are determined by a number of factors, including an artifact due to desorption of NOx from the walls of the reaction vessel during irradiation. Gao et al. (1994) report the latter is equivalent to ~ 20-40 ppt using synthetic air in the laboratory, but in ambient air, may limit measurements of concentrations below 100 ppt.

As discussed earlier, TDLS can be used to measure N02. The detection limit cited for a path length of 33.5 m in a ground-based study is ~ 150 ppt (Mackay et al., 1988) and 25 ppt in an aircraft study (Schiff et al., 1990). The detection limit for DOAS with a path length of 800 m is ~4 ppb (Biermann et al., 1988).

Finally, matrix isolation combined with electron spin resonance has been used for N02 as well as for other free radicals such as HOz, ROz, and NO, (Mihelcic et al., 1985, 1990, 1993; Zenker et al., 1998). Trace gases in a sample of air (typically about 8 L) are trapped in a D20 matrix at 77 K and the ESR spectrum obtained. Any paramagnetic species present has a characteristic ESR spectrum that can be used to identify it and, using reference spectra, obtain its concentration. Since N02 is the paramagnetic species present in the largest concentration, it is easily detected and measured.

Several intercomparison studies for N02 have been carried out (e.g., Fehsenfeld et al., 1990). At concentra tions of N02 above 400 ppt, measurements using the photolysis of N02 and chemiluminescence for the NO generated by photolysis were in reasonably good agreement with TDLS measurements. At levels above about 300 ppt, the photolysis and luminol method corrected for ozone and PAN agreed reasonably well, with the slope of the corrected luminol versus photolysis data being 1.09 (Fehsenfeld et al., 1990).

An airborne intercomparison study (Gregory et al., 1990a) was also carried out using two photolysis methods (the 353-nm laser photolysis with TP-LIF detection of NO and a Xe arc lamp photolysis with chemiluminescence detection of NO) as well as TDLS. Overall, for N02 up to 200 ppt, the techniques agreed with the average values of all three by 20% or better and with each other to within 30%. However, below 50 ppt, there was very little correlation between the various measurement techniques (Gregory et al., 1990a).

An informal intercomparison study of N02 measurements was carried out in a remote atmosphere at Izana, Tenerife (Zenker et al., f998). Three techniques were used: TDLS, photolysis with a chemiluminescence detector, and matrix isolation-ESR. Agreement between the three methods was good, with plots of data from one technique against the others having slopes within experimental error of unity. For example, TDLS and the photolysis technique plotted against the matrix isolation measurements had slopes of 0.90 ± 0.47 and f.04 + 0.34, respectively, over a range of NOz concentrations from ~ f 00 to 600 ppt.

In summary, there are a variety of methods of measuring N02 that are reasonably accurate for higher concentrations, particularly those found in polluted areas. However, at smaller concentrations found in the remote troposphere, there are significant discrepancies between the various methods.

In addition to these techniques, there are passive samplers for N02 that have been used for unique situations such as indoor measurements. For example, in the Palmes Tube, N02 diffuses through to a surface coated with triethanolamine and is trapped in the form of NO^. The nitrite is subsequently measured colori-metrically (e.g., see Boleij et al., 1986; Miller, 1988; and Krochmal and Gorski, 1991). As with most, if not all, such wet chemical methods, interferences can arise, for example, from PAN (Hisham and Grosjean, 1990) and HONO (Spicer et al., 1993b).

(4) Typical levels of NO, N02, and NOy Figure If.30 shows a summary of measurements of surface concentrations of NO, NOx, and NO>( made at a variety of remote to rural sites in North America and Europe (Emmons et al., 1997). The bars encompass the central 90% of the values and the medians and means are

10000 !L 1000

z 100

Was this article helpful?

0 0

Post a comment