An evaluation of the mechanism of nitrous acid formation in the urban atmosphere

Nitrous acid (HONO) has been observed to build in the atmosphere of cities during the nighttime hours and it is suspected that photolysis of HONO may be a significant source of HO radicals early in the day. The sources of HONO are poorly understood, making it difficult to account for nighttime HONO formation in photochemical modeling studies of urban atmospheres, such as modeling of urban O₃ formation. This paper reviews the available information on measurements of HONO in the atmosphere and suggest mechanisms of HONO formation. The most extensive atmospheric measurement databases are used to investigate the relations between HONO and potential precursors. Based on these analyses, the nighttime HONO concentrations are found to correlate best with the product of NO, NO₂ and H₂O concentrations, or possibly the NO, NO₂, H₂O, and aerosol concentrations. A new mechanism for nighttime HONO formation is proposed that is consistent with this precursor relationship, namely, reaction of N₂O₃ with moist aerosols (or other surfaces) to form two HONO molecules. Theoretical considerations of the equilibrium constant for N₂O₃ formation and the theory of gas-particle reactions show that the proposed reaction is a plausible candidate for HONO formation in urban atmospheres. For photochemical modeling purposes, a relation is derived in terms of gas phase species only (i.e., excluding the aerosol concentration): NO + NO₂ + H₂O → 2 HONO with a rate constant of 1.68 x 10^-17 e^6348/T (ppm^-2 min^-1). This rate constant is based on an analysis of ambient measurements of HONO, NO, NO₂ and H₂O, with a temperature dependence from the equilibrium constant for formation of N₂O₃. Photochemical grid modeling is used to investigate the effects of this relation on simulated HONO and O₃ concentrations in Los Angeles, and the results are compared to two alternative sources of nighttime HONO that have been used by modelers. Modeling results show that the proposed relation results in HONO concentrations consistent with ambient measurements. Furthermore, the relation represents a conservative modeling approach because HONO production is effectively confined to the model surface layers in the nighttime hours, the time and place for which ambient data exist to show that HONO formation occurs. The empirical relation derived here should provide a useful tool for modelers until such time as knowledge of the HONO forming mechanisms has improved and more quantitative relations can be derived.