FIGURE 11.30 Range of surface concentrations of NO, NO^, and NO,, at a variety of Northern Hemisphere locations. The asterisk is the mean and the horizontal line the median. The bar represents the central 90% of measured values. SNP, Shenandoah National Park, Virginia; SLD, Schauinsland, Germany; EGB, Egbert, Ontario, Canada; and HAR, Harvard Forest, Massachusetts. (Adapted from Emmons et al., 1997.)

shown by the horizontal lines and asterisks, respectively. The identity of some of the specific sites at which measurements were made are indicated. As expected, remote regions such as Barrow, Alaska, have the smallest concentrations of NO, typically less than 10 ppt. The most polluted (but still rural) areas have concentrations that in the winter are in the ppb range.

Similar conclusions hold for NOx and NOr NOx concentrations range from median values of 25 ppt in Alaska (remote) to 2.3 ppb at Egbert, a rural area in Ontario, Canada, in the summer. NOy ranges from 69 ppt at Barrow, Alaska, to 5.0 ppb at Shenandoah National Park in Virginia (Emmons et al., 1997). Data from a number of studies are archived electronically and can be accessed as described by Carroll and Emmons (1996).

In urban areas, the concentrations can of course be much greater. For example, in Paris in late September 1997, N02 concentrations exceeded 2f0 ppb (Environmental Science & Technology, 1997). In metropolitan Toronto, Canada, peak 30-min average concentrations of ~40 ppb have been reported (Schiff et al., 1986). In the Los Angeles area, maximum 1-h concentrations of about 200 ppb are encountered (Air Quality Management District summary data; see Appendix IV).

Measurements of NO, NOx, NOr PAN, HN03, and 03 in the free troposphere obtained from 1985 to 1995 are summarized by Thakur et al. (1999) and can be obtained electronically as described in that paper.

(5) HN03 Analysis of HN03 at the low levels typically found in the atmosphere is difficult, in large part due to its tendency to adsorb very readily to surfaces. As a result, sampling HN03 in an artifact-free manner is often the limiting aspect in making accurate measurements.

Several different methods exist for measuring HN03, most commonly FTIR and TDLS, which were described earlier. Other techniques commonly used include filters, denuders, transition flow reactors, and scrubbers, followed by analysis of the collected material for nitrate, e.g., by ion chromatography. A modification of the luminol method has also been used. Finally, mass spectrometric methods look very promising as a sensitive and specific method of detection and measurement. A brief description of each of these methods that have not yet been treated follows.

Filters. HN03 is efficiently trapped out on nylon filters. Typically, two or more filters are connected in series. A schematic of such a filter pack was shown in Fig. 11.22 (Anlauf et al., 1988). A Teflon filter first removes particles from the airstream and a nylon filter then removes gaseous HN03. In this particular system, a third filter (Whatman 41 impregnated with an aqueous solution of glycerol and citric acid) was used to trap NH3. After sample collection, each of the filters is extracted separately and nitrate, ammonium, and additional particle components collected on the Teflon filter are measured by ion chromatography. The sensitivity of this method for nitric acid and the other species is determined in part by filter blank values (i.e., nitrate on unexposed filters) and by the total amount collected and hence the sampling time used. Times of

4-6 h are often used, but can be as short as 0.5-2 h (Fehsenfeld et al., 1998).

Other filters have also been used to collect nitric acid, such as Whatman 41 filters impregnated with NaCl (e.g., Anlauf et al., 1986).

Potential interferences in the measurement of nitric acid using this method include removal of gaseous HNO-, on the Teflon particle filter and/or volatilization of particle nitrate collected on this Teflon filter. As discussed in Chapter 7, NH4N03 is a common particle component, but exists in equilibrium with gas-phase NH3 and HN03:

Shifts to the right, e.g., due to a temperature increase, release HN03 and NH3, which are then collected on the nylon filter and Whatman impregnated filters, respectively, and measured as gas-phase nitric acid and ammonia. This was hypothesized to be responsible for higher filter pack values compared to those measured by mass spectrometry under some conditions, particularly at colder air temperatures (<15°C), where the equilibrium (10, — f0) favors relatively larger amounts of ammonium nitrate in air (Fehsenfeld et al., 1998). Talbot et al. (1990) observed higher HN03 concentrations by the nylon filter technique compared to a mist chamber (vide infra) and hypothesized that unknown (perhaps organic) nitrogenous compounds were also being collected on the nylon, forming nitrate. They also showed that 03 at typical concentrations found in the troposphere could react with some unknown substance(s) on the nylon filter to generate a positive artifact. This artifact was significantly reduced by pre-washing the nylon to remove water-soluble adsorbed species.

Denuders. A variety of denuder wall coatings have been used to collect HN03 (Perrino et al., f 990). These include nylon fiber mats (e.g., Durham et al., f987), MgO (e.g., Solomon et al., 1988, 1992), Na2C03/ glycerol (e.g., Ferm, 1986; Koutrakis et al., 1988), and tungsten oxide (W03) (e.g., see Fox et al., f988). A variant of this is the tungstic acid technique. Air containing the HN03 is passed through tubes coated with tungstic acid. When the tube is subsequently heated, the HN03 decomposes and desorbs as NO or N02.

Transition flow reactors. TFRs have also been used to measure gaseous HN03 (e.g., Hering et al., f988). When operated in the configuration used by Durham et al. (1986), where the TFR is upstream of the Teflon particle filter, the problems of adsorbing gaseous HN03 on the particle filter or of evaporation of HN03 from the collected particles onto the gaseous HN03 sampler are avoided.

Scrubbers. Mist chamber scrubbers have also been used for HN03. The airstream passes through a Teflon filter to remove particles and then encounters a mist of water that scrubs the HN03 out of the air. The nitrate concentration is measured in the aqueous scrubbing solution using ion chromatography (Talbot et al., 1990).

Luminol method. As described earlier, N02 undergoes a chemiluminescent reaction with luminol, and this has been used to measure N02. This has also been used to measure HN03 by difference. One airstream passes through a Teflon filter to remove particles, while another passes through a Teflon-nylon filter combination to remove both particles and gaseous HN03. The air is then passed over hot glass beads to convert NO and HN03 to N02, which is measured using the luminol method. The difference in signal between the two filtering methods then gives gaseous HN03 (Hering et al., 1988).

All of these methods have potential interferences, some of which have been described in conjunction with the individual methods. For example, whenever a Teflon prefilter for particles is used, gaseous HN03 can be taken up on this filter, particularly when it has a high particle loading, leading to a negative artifact. Volatilization of ammonium nitrate particles collected on the Teflon filter can also occur, giving a positive artifact. Uptake and conversion of other nitrogen-containing compounds to nitrate gives a positive artifact (e.g., see Koutrakis et al., 1988). In addition, as mentioned at the beginning, HN03 is notoriously "sticky," and the use of short (or ideally, no) sampling lines made of materials such as Teflon that minimize its adsorption are necessary to avoid negative artifacts when it is adsorbed and positive artifacts when it later desorbs. Finally, they are not "real-time" methods in that sampling for periods of time of the order of hours is typically required.

Given these considerations, it is perhaps not surprising that intercomparison studies have shown significant disagreements between the various techniques. Figure 11.31, for example, shows some data from one such intercomparison (the so-called "nitric acid shootout"), in which filter pack (FP), difference denuder (DD), annular denuder (AD), TDLS, and FTtR measurements were all taken simultaneously, ft is seen that the filter pack measurements were consistently higher, and the annular denuder and TDLS consistently lower, than the FTfR measurements, which were used as the standard for comparison. While most measurements in this period were within the wide error bars for the FTIR measurements, note that the range of concentrations from the different methods spans more than a factor of two! in addition, there were significant differences in some of the methods from day to night.

Was this article helpful?

0 0

Post a comment