Info

0 40 70 120 160 200 NO by Chemiluminescence (ppt)

FIGURE 11.24 Measured NO concentrations using TP-LIF and chemiluminescence during one series of aircraft flights (adapted from Crosley, 1996).

0 40 70 120 160 200 NO by Chemiluminescence (ppt)

FIGURE 11.24 Measured NO concentrations using TP-LIF and chemiluminescence during one series of aircraft flights (adapted from Crosley, 1996).

laser beam to generate NO, causing interference in the NO measurement. However, since the ionization of NO is a two-photon process, the signal is expected to vary with the square of the laser power (P). On the other hand, since the production of NO from other compounds such as N02 requires three photons to generate and then photoionize NO, the dependence of the signal on the laser power is steeper. For example, Lee et al. (1997) report that the signal varies with P115 for NO but P2A for N02.

Figure 11.25 is a schematic diagram of one such REMPI system currently under development for ambient air analysis (Schmidt et al., 1999). The ions are generated in a two-photon process and then separated using time-of-flight mass spectrometry (TOF-MS), which provides an additional means of selectivity. For NO in laboratory air, the current detection limit using this system is 1 ppt. It has also been applied to the measurement of CO and CH3CHO in laboratory systems using a (2 + 1') two-color ionization process, with detection limits in synthetic air in laboratory studies of 10 and 1 ppt, respectively.

(2) NOy NO>( is measured by passing the airstream containing NO and the other oxides of nitrogen over a catalyst to convert all of the other oxides of nitrogen into NO, which is then measured by one of the techniques just discussed. The resulting measurement is taken as the total oxides of nitrogen present.

The most common catalysts used are MoO at 375-400°C or Au at 300°C with added CO or H2. The mechanism of reduction at the surfaces is not clear.

Reaction of the various oxides of nitrogen on the metal surfaces may leave a surface oxide, which is then removed by reaction with the CO, forming C02, or with the H2, forming H20 (e.g., Kliner et al., 1997, and references therein).

This method of measurement of total oxides of nitrogen means that NO>( is defined operationally in terms of compounds that can be reduced to NO over these catalysts. It had been generally accepted that under typical operating conditions, species such as HCN, CH3CN, N20, NH3, and amines are not significantly reduced and hence did not contribute to NO>( (e.g., see Crosley, 1996). However, Kliner et al. (1997) showed that HCN, CH3CN, and NH3 can be converted to NO with high efficiencies under some conditions. For example, 85% of the HCN was converted using H2 and 100% using CO with an Au catalyst at 300°C. Weinheimer et al. (1998) measured conversion efficiencies for HCN using three "outwardly identical" gold converters at 300°C with added CO. The conversion efficiency was 5-7% for ambient air sampled during aircraft flights with or without added water for two of the converters, with the efficiency doubled when synthetic air was sampled on the ground. The third converter had efficiencies for HCN of ~ 30% under all conditions. Bradshaw et al. (1998) reported conversion efficiencies ranging from 6 to 100% for HCN in gold converters. High conversion efficiencies were also found for organic nitrates, with the efficiencies being larger for the smaller nitrates such as nitroethane; differences were also noted between pure gold and gold-plated

308 nm

308 nm

Differential pumping stage and ion source

FIGURE 11.25 Schematic of REMPI-TOF. The conventional electron impact ionization source is just used for ion beam focusing (adapted from Schmidt et al., 1999).

Differential pumping stage and ion source

FIGURE 11.25 Schematic of REMPI-TOF. The conventional electron impact ionization source is just used for ion beam focusing (adapted from Schmidt et al., 1999).

converters, and in the latter case, depended on previous cleaning of the converter.

In short, it is clear that the conversion efficiencies have to be tested for each converter under conditions in which the field measurements are made.

With HCN concentrations of ~170 ppt in the stratosphere and upper troposphere (Coffey et al., f 981; Cicerone and Zellner, 1983; Zander et al., 1988; Schneider et al., 1997), and up to ~900 ppt at times (Rinsland et al., 1998), HCN could contribute significantly to NOy, depending on the conversion efficiency. The same is true of acetonitrile, CH3CN, whose concentrations are less well known; it has been measured over Europe at concentrations in the range of f50-200 ppt (e.g., Hamm et al., 1989) and in the lower stratosphere at concentrations of 110—160 ppt (Schneider et al., 1997). However, much smaller concentrations, of the order of a few tens of ppt, have also been reported in the atmosphere (Knop and Arnold, 1987a, 1987b). High concentrations of NH3 are quite common in the troposphere, particularly near sources such as cattle feedlots (vide infra).

In addition to the potentially varying contributions of compounds such as HCN and NH3 to NO , there are a number of other variables that can impact the measured values. One of the most important is the effect of the sampling lines, which can adsorb and desorb various gases. Nitric acid in particular is well known to be "sticky," readily adsorbing to various surfaces in a manner that is not reproducible and depends on such factors as the amount of water present on the surface. It is therefore not surprising that the agreement between various methods of measuring NO^ is not as good as for NO. Figure 11.26, for example, shows the NO measurements made during the flights for which the NO data are shown in Fig. 11.24 (Crosley, 1996). The slope of the regression line is 1.18 + 0.04, with an r2 value of only 0.37 for the scattered data.

In principle, if each of the compounds contributing to NO is individually measured, their concentrations should sum up to the measured NOr While this is sometimes the case, in many field studies the sum has been found to be less than the measured NO} (e.g., Fahey et al., 1986; Ridley et al., 1990a; Atlas et al., 1992; Parrish et al., 1993; Sandholm et al., 1994; Nielsen et al., 1995; Williams et al., 1997). Cases where the sum of the individual components is significantly less than the NOv measured simultaneously are often referred to as "missing NOr"

Table 11.7, for example, summarizes measurements of the components of NOy made at Niwot Ridge, Colorado, in mid-1987 (Ridley et al., 1990a). The sum of NOx (NO + N02), HN03, particulate nitrate, PAN,

Was this article helpful?

0 0

Post a comment