Remote an integration time of 5 min have been reported (Williams et al., 1992).
Denuder methods. As described earlier for HN03, diffusion denuders are also used for NH3. For example, tungstic acid coated denuders take up both HN03 and NH3. The ammonia is then thermally desorbed (as NH3), oxidized to NO, and measured using chemilumi-nescence (Braman et al., 1982; Hering et al., 1988). In practice, since HN03 desorbs in the form of NO and N02, the NH3 from the thermal desorption is usually first readsorbed on a second W03-coated tube to separate it from the HN03 signal, and a second thermal desorption followed by conversion to NO used for the measurement of NH3.
Other denuder coatings for NH3 include molybdenum oxide, from which ammonia is thermally desorbed and converted to NO for measurement as in the tungstic acid method; with Mo03 as the coating, some of the NH3 thermally desorbs directly in the form of NO (Langford et al., 1989). In addition, oxalic acid has been used as a coating and the NH3 collected as ammonium ion is obtained by extracting the tube and measuring NH J by ion chromatography (Ferm, 1979).
Filter packs. As shown in Fig. 11.22, NH3 can be collected on impregnated filters in filter packs designed to collect particles and gas-phase nitric acid. Oxalic acid or citric acid on Whatman filters is often used to absorb the gaseous ammonia, which is then measured by extraction into aqueous solution and ion chromatography or by a colorimetric method (e.g., see Anlauf et al., 1988; and Williams et al., 1992).
Williams et al. (1992) have carried out an intercom-parison of PD-LIF with three denuder methods and one filter pack technique using both laboratory-prepared samples and ambient air. All methods agreed to within 10% when measuring an ammonia standard. When a measured amount of NH3 was introduced as a spike into filtered air, the agreement was not as good. The PF-LIF and citric acid denuders gave 87 and 93% response, respectively, but the recoveries using the tungstic acid denuder, the molybdenum oxide annular denuder, and the filter pack were only 68, 37, and 22%, respectively. In the ambient air measurements, the PF-LIF and citric acid denuders were in good agreement. The tungstic acid technique gave values that were well correlated with the PF-LIF values but had an intercept that varied with the concentration range. The filter pack gave concentrations having a response about two-thirds of those methods, possibly due to uptake of NH3 on the Teflon particle prefilter at the lower temperatures of these studies. The molybdenum oxide annular denuder measurements were about 64% of the PF-LIF values above 1 ppb but tended to be higher below this concentration.
Although other fluorescence (e.g., Rapsomanikis et al., 1988; Genfa et al., 1989), chemiluminescence (Maeda and Takenaka, 1992), and photoacoustic or photothermal methods (e.g., De Vries et al., 1995) have been proposed, they have not found widespread use.
Typical tropospheric concentrations of NH3. In remote areas, NH3 concentrations can be quite low, <50 ppt (e.g., Lewin et al., 1986; Alkezweeny et al., 1986), whereas close to sources such as agricultural areas and cattle feedlots, they can be about three orders of magnitude larger. For example, measurements of ammonia made at Boulder, Colorado, and at Niwot Ridge, west of Boulder, are typically ~5.5 ppb and ~200 ppt, respectively (Langford and Fehsenfeld, 1992). The higher value in Boulder is due to the proximity of agricultural areas and cattle feedlots. Similarly, Biermann et al. (1988) observed concentrations of NH3 at Claremont, California, over a range from 57 ppb when the winds were blowing from the direction of an agricultural area with a high density of cattle feedlots and poultry and dairy farms to undetectable levels (<1-2 ppb) when the wind was from other directions.
Was this article helpful?