Environ. Sci. Technol. 2009, 43, 1108–1114
have very different chemical and physical properties. For example, although both NO2- (reaction 3) and HONO (reaction 4) in solution undergo direct photolysis to form NO and •OH (4, 6, 8), it is unclear whether H2ONO+ behaves similarly
Photochemistry of Nitrous Acid (HONO) and Nitrous Acidium Ion (H2ONO+) in Aqueous Solution and Ice CORT ANASTASIO* AND LIANG CHU Atmosphere Science Program, Department of Land, Air, and Water Resources, University of California, One Shields Avenue, Davis, California 95616-8627
Received September 11, 2008. Revised manuscript received November 17, 2008. Accepted December 15, 2008.
We have examined the photochemistries of two N(III) species, nitrous acid (HONO) and nitrous acidium ion (H2ONO+), in solution and ice. Although the light absorption spectra for these two species are very similar, their photochemical efficiencies are quite different: the •OH (and NO) quantum yield for HONO is approximately 8 times greater than that of H2ONO+ at 274 K. The temperature dependent expressions for the •OH (and NO) quantum yields are ln(Φ(HONO f •OH) ) (7.14 ( 0.57) - (2430 ( 160)/T and ln(Φ(H2ONO+ f •OH) ) (3.16 ( 0.67) - (1890 ( 180)/T. The temperature dependence for H2ONO+ includes both solution and ice data (255-283 K), suggesting that its ice photochemistry is occurring in a quasi-liquid environment. The quantum yields for HONO and H2ONO+ are independent of wavelength, in contrast to NO2-. On the basis of the pH dependence of N(III) photolysis, our results are consistent with recently reported pKa values of 1.7 for H2ONO+ and 2.8 for HONO. Using our results in a kinetic model of nitrogen chemistry illustrates that the fluxes of HONO and NOx from sunlit snow can be explained by nitrate photolysis and are pH dependent because of a competition between HONO evaporation and N(III) reactions on ice grains.
Introduction HONO is an important photochemical source of hydroxyl radical (•OH) and nitric oxide (NO) in a wide range of environments, from urban regions to polar snowpacks (1-3). This photolysis of HONO occurs both in the gas phase as well as in aqueous phases such as cloud and fog drops (4-6). The chemistry of HONO is complicated by the fact that in the aqueous phase it can both dissociate to nitrite (reaction 1) and be protonated to form nitrous acidium ion (also called nitroacidium ion, nitrosoacidium ion, or hydrated nitrosonium ion) (reaction 2) HONO a H+ + NO2+
HONO + H a H2ONO
(1) +
(2)
In addition, under very acidic conditions (e.g., at least 45 wt % H2SO4; ref 7), the nitrous acidium ion dehydrates to form free nitrosonium ion (NO+). Understanding the speciation of N(III) (i.e., NO2-, HONO, H2ONO+, and NO+) is important because the different forms * Corresponding author phone: (530) 754-6095; fax: (530) 7521552; e-mail:
[email protected]. 1108
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 4, 2009
NO2- + hν(λ 5) are significantly lower than austral summer values recently measured at Halley, Antarctica (30 - 40 nmol L-1 in the top 20 cm; ref 22). This difference might be because the model is missing snowpack sources of N(III) or, possibly, because the measurements are too high due to a positive interference by peroxynitrite (23). The pH dependence of the N(III) lifetime in the QLL (Figure 5a) reflects the fact that the three forms of N(III) have different rate constants for loss. As shown in Figure 5b, HONO volatilization and photolysis are the dominant N(III) sinks in the QLL across a wide range of pH values (roughly pH 0.5 to 4), but at very low and near neutral pH values the reactions of H2ONO+ (photolysis) and NO2- (photolysis and reaction with O3), respectively, dominate. Figure 5c illustrates the modeled fluxes of NO, NO2, and HONO into the firn air, expressed as the equivalent surface fluxes into the boundary layer if there were no losses or transformations in the firn 1112
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 4, 2009
air. The dominant flux is that of NO2, which is produced directly from nitrate photolysis (Figure 4); the decrease in NO2 release at pH values below 2 has been estimated from experimental data (19) (see Figure S2 in the Supporting Information). The modeled NO emissions rate increases slightly with decreasing pH, and ranges from 20% (at pH 7) to 70% (at pH 0) of the NO2 flux. Although nitrate is the original source of NOx and N(III) in our model, N(III) photolysis is a significant source of NO, as suggested recently by calculations (8) and measurements (22). The flux of HONO (Figure 5c) has the same general pH dependence seen for the HONO mole fraction (Figure 5a), with a peak between pH 1.5-3.5. Figure 5c indicates that snow with weakly acidic or alkaline pH values is not a source of HONO, as observed in the field (24); similarly, in very acidic snow, HONO fluxes are suppressed because H2ONO+ becomes the dominant N(III) species. The right-hand side of Figure 5c shows observed snowto-air fluxes of NOx and HONO at Summit during summer. In general, our modeled values are at the upper end of the observations in both categories, which is consistent with the idea that nitrate photolysis is the major source of snowpack NOx and HONO (3, 19, 25-29). While a number of our model parameters are poorly constrained, the fact that the modeled emissions are at the high end of the observations suggests there might be sinks for NOx and N(III) in the Summit snowpack that are missing from our model, either in the QLL or firn air. As shown in gray in Figure 4, some of the potential QLL sinks include formation of peroxynitrous and peroxynitric acids, and oxidation of halides by these species
FIGURE 5. Effect of quasi-liquid layer pH on modeled nitrogen chemistry in the Summit surface snowpack (0-20 cm) at midday on the summer solstice. The model is the same as described in Figure 4 and Table S2 in the Supporting Information. (a) Calculated QLL concentrations (in terms of bulk (melted) snow volumes) for each N(III) species and for total N(III) (N(III)T). The dotted line represents the overall lifetime of N(III)T in the QLL. (b) Fraction of N(III) in the QLL that reacts via the five N(III) sinks. (c) Modeled fluxes from the QLL into the firn air, expressed as the equivalent surface fluxes that would result if all of the HONO and NOx released in the top 20 cm of snowpack was emitted from the surface. The box plots at the right of panel (c) show the observed snow-to-air fluxes measured by Honrath et al. (30) at Summit during midday (10:30-13:30 local time) in the summer (05 June-03 July, 2000). Key to box plots: square ) mean flux; circle ) median flux; shaded box ) middle 67% of observations; vertical bar ) middle 95% of observations. and H2ONO+. These reactions are possibly important sources of HOONO, HOONO2, and reactive halogens to the firn air and boundary layer. In addition to QLL pH, the mass transport of HONO from the QLL into the firn air affects both HONO release and the concentration of N(III) in the QLL. If HONO release is very fast (see Figure S3 in the Supporting Information) then the
N(III) concentration is very low (
