Other hydroperoxides have also been detected at small concentrations in air. For example, Hewitt and Kok (1991) reported the presence of 1-hydroxyethyl hydroperoxide as well as an unidentified compound, perhaps hydroxybutyl hydroperoxide, in air in rural Colorado.
In remote areas, CH3OOH is generally the major, and often the sole detectable, organic hydroperoxide present (e.g., see Staffelbach et al., 1996). This is not surprising, since CH4 is often the major organic in such regions, and hence the CH302 + H02 reaction is important.
A three-channel approach was developed by Lee et al. (1993) to distinguish H202 from hydroxymethyl hydroperoxide and total peroxides. In this approach, one channel is used to scrub the air sample into a POPHA solution to obtain total peroxides. In a second channel, the air sample is scrubbed into Fenton reagent solution at a pH of 3. This converts the H202 into OH radicals:
The OH radicals are trapped by reaction with benzoic acid, forming hydroxybenzoic acid, which is measured by fluorescence. Organic peroxides ROOH form RO + OH~, but the derivatives of benzoic acid formed by the reaction of the RO radicals do not fluoresce under the conditions chosen to measure H202. Thus, in principle, this second channel measures H202. However, in practice, it was found to give about a 30% response to hydroxymethyl hydroperoxide as well, so that the results from this channel must be corrected for this contribution (Lee et al., 1993).
In the third channel of this instrument, air is scrubbed into a solution containing Fenton reagent at a pH of 9. At this high pH, hydroxymethyl hydroperoxide is rapidly hydrolyzed to give H202. Thus the third channel gives the sum of H202 and hydroxymethyl hydroperoxide and the difference between this and the second channel (corrected) gives hydroxymethyl hydroperoxide.
An intercomparison of TDLS, POPHA, and a lumi-nol chemiluminescence method for H202 was carried out using zero air, irradiated VOC-NOx mixtures, and ambient air (Kleindienst et al., 1988a). The TDLS and two POPHA methods using different sampling approaches (continuous scrubbing and a diffusion scrubber, respectively) were in reasonably good agreement. However, the luminol method exhibited positive and negative interferences under different conditions and hence has not since been applied extensively to ambient air measurements.
Another intercomparison was carried out at the Mauna Loa Observatory in 1991 and 1992 (Staffelbach et al., 1996). TDLS was used to measure H202. In addition, the POPHA method with catalase was used to distinguish between H202 and organic peroxides, the POPHA method with aqueous solubility differences was employed to discriminate between these compounds, and HPLC was used to separate and detect different hydroperoxides. For H202, while the measurements using the wet chemical methods and TDLS showed similar trends, there was a significant amount of scatter in individual measurements. For example, the correlation coefficients for plots of the TDLS versus the two POPHA techniques varied from 0.20 to 0.60. HPLC showed that CH3OOH was the only organic hydroperoxide present. However, individual measurements of CH3OOH made using this method compared to those using POPHA with catalase only had a correlation coefficient of 0.14 in wet air and 0.49 in dry air, whereas the corresponding correlation coefficients for the two POPHA measurements of organic hydroperoxides were 0.33 and 0.48, respectively.
In short, while wet chemical techniques are valuable for measurement of H202 and organic hydroperoxides, the absolute accuracy and precision remain a subject of concern and research.
Typical tropospheric concentrations. Measurements of H202 in air up to approximately 1990 are summarized by Gunz and Hoffmann (1990) and Sakugawa et al. (1990). Concentrations of H202 observed in air near the earth's surface are typically about 1-5 ppb (e.g., Daum et al., 1990; Van Valin et al., 1990; Claiborn and Aneja, 1991; Tremmel et al., 1993; Lee et al., 1993; Das and Aneja, 1994; Tanner and Schorran, 1995; Macdonald et al., 1995; Staffelbach et al., 1996; Sanhueza et al., 1996; Ridley et al., 1997; Martin et al., 1997; Balasubra-manian and Husain, 1997; Weinstein-Lloyd et al., 1998). Although it may initially appear surprising, concentrations in remote and rural areas are not tremendously different from those in more polluted urban areas. For example, Heikes et al. (1996) reported levels of 0.3-5 ppb in the marine boundary layer, and Weinstein-Lloyd et al. (1998) measured concentrations of 1-4 ppb in the continental boundary layer midday in a rural area in the southeastern United States. The reason is that although there is a great deal more photochemical activity in the polluted regions, which might be expected to lead to H202, there is also more NO. Since H202 is formed by the H02 + H02 self-reaction and since HOz also reacts rapidly with NO, higher NO levels tend to inhibit the formation of the peroxide.
In addition to H202, methyl hydroperoxide has also been measured. For example, in air over Hawaii, concentrations of ~ 0.1 -0.5 ppb were typical, although concentrations as high as 1.6 ppb have been observed in remote areas (Staffelbach et al., 1996; Heikes et al., 1996; Sanhueza et al., 1996; Ridley et al., 1997). Weinstein-Lloyd et al. (1998) measured concentrations up to ~2.5 ppb in the rural continental boundary layer. There are insufficient studies to firmly establish the relative contribution of CH3OOH and perhaps other organic hydroperoxides to the total atmospheric levels, but the data available indicate that H202 is the major hydroperoxide present in air. For example, Tanner and Schorran (1995) found that H202 typically comprised about 90% of the total peroxides in the Grand Canyon area of the United States and Ayers et al. (1996) found in a limited set of measurements at Cape Grim, Tasmania, that H202 was ~60% of the total. Similarly, the median concentration of CH^OOH measured by Weinstein-Lloyd et al. (1998) in the rural continental boundary layer was 1.7 ppb, representing about a third of the total measured hydroperoxides.
Hydroxymethyl hydroperoxide has also been identified and measured in rural continental areas as well as in the marine boundary layer. For example, Lee et al. (1993) report concentrations as high as 5 ppb in rural Georgia, and Heikes et al. (1996) measured concentrations from 0.6 to f.6 ppb in the marine boundary layer over the south Atlantic Ocean. The median value in the rural continental boundary layer in the southeast United States was reported to be 0.97 ppb, with individual measurements ranging from ~0.2 to 3 ppb (Weinstein-Lloyd et al., 1998).
In summary, H202 is ubiquitous in air throughout the troposphere. CH3OOH and HOCH2OOH have also been observed, generally at smaller concentrations than H202. There is at present little evidence for significant contributions of larger organic peroxides.
i. HOx Free Radicals
As seen throughout this book, OH and HOz are central to the chemistry of both the lower and upper atmosphere. As a result, accurate measurement of their concentrations is critical to a quantitative understanding of atmospheric chemistry.
(1) OH Estimates of globally averaged OH concentrations have been obtained by applying a mass balance type of approach to certain compounds whose major removal from the atmosphere is believed to be reaction with OH. For example, the emissions of methylchloroform, CH3CCI3, are well known and its concentrations have been measured at a number of locations around the world. Using 3-D models, one can calculate the concentrations of OH and their geographical distribution that remove CH3CC13 at appropriate rates to generate the measured concentrations (e.g., see Alt-shuller review, 1989; Spivakovsky et al., 1990; Prinn et al., 1992, 1995; and Krol et al., 1998). A similar approach has been taken using l4CO (Brenninkmeijer et al., 1992).
In particular air masses, estimates of OH concentrations have also been derived from the relative rates of decay of a series of hydrocarbons in the air mass whose rate constants for reactions with OH are well known (e.g., Blake et al., 1993). Alternatively, organics can be added as tracers; criteria for the choice of suitable compounds are discussed by Davenport and Singh (1987). However, such approaches can be complicated by the effects of transport and mixing of the air mass with ones of different composition (e.g., McKeen et al., 1990) and by the possible contribution of oxidants other than OH to the decay of the organic (e.g., Blake et al., 1993). In addition, average OH concentrations rather than point or local measurements are derived from such data, and both this and the mass balance approach are indirect. However, Ehhalt et al. (1998) have proposed an alternate approach to using hydrocarbon concentrations to determine OH, which minimizes the assumptions inherent in this method.
Clearly, direct techniques for measuring OH are needed that provide concentrations either at a point or over relatively restricted spatial scales. Two (absorption and laser-induced fluorescence) are direct, spectroscopic methods and two others (mass spectrometry and a radiocarbon method) rely on conversion of OH to another species that is measured. Each of these approaches and some of the intercomparisons that have been carried out are discussed briefly in the following sections. A good overview of these methods is found in a review by Eisele and Bradshaw (1993) and articles by Crosley (1994, 1995a, 1995b) and papers in a special issue of the Journal of the Atmospheric Sciences [52 (19), October 1, 1995],
Absorption spectroscopy. OH undergoes an allowed transition between its X2I1 ground state and the first electronically excited A2£+ state. Because it is a small species, absorption lines due to the individual vibrational and rotational transitions can be resolved experimentally. As a result, it has a very characteristic banded absorption structure around 308 nm whose features make it an ideal candidate for DOAS measurements.
Atmospheric OH has been measured for a number of years using its absorption. For example, vertical column abundances of OH have been measured in a number of studies using the sun as a light source (e.g., Burnett and Burnett, 1981, 1996; Burnett et al., 1988; Burnett and Minschwaner, 1998). Over the past several decades, beginning with the measurements of Perner et al. (1976), absorption spectroscopy has been used to make measurements of OH over much shorter paths in the troposphere. The fundamental principles behind this technique have been described earlier in the discussion of DOAS spectrometry. Here we briefly discuss some of the aspects of the measurements that are unique to OH, as well as some typical applications.
Figure 11.42a shows an energy level diagram and some of the allowed lines in the v" =0 level of X2I1 to the v' = 0 level of A2 2+ (Mount, 1992). Figure 11.42b shows the absorption spectrum of OH obtained using a butane flame as the source in this case. The emission of a frequency-doubled dye laser whose full width at half-maximum is 0.41 nm is also shown (Dorn et al., 1995a). The laser emission is sufficiently broad n1
OH Reference Spectrum
Was this article helpful?