# Response to Comment on "Atmospheric Hydroxyl Radical Production from Electronically Excited NO2 and H2O"

See allHide authors and affiliations

Science  17 Apr 2009:
Vol. 324, Issue 5925, pp. 336
DOI: 10.1126/science.1166877

## Abstract

Carr et al. failed to detect hydroxyl radical formation from the reaction of excited state nitrogen dioxide with water, contrary to our findings. We present several reasons, based on energetic and spectroscopic considerations, why the OH radicals we observed from this reaction are not likely to be due to multiphoton excitation as they suggest.

The reaction of electronically excited nitrogen dioxide with water, NO2*+ H2O, can be an important source of atmospheric OH radicals if its reaction rate is sufficiently fast (13). Carr et al. (2) report a rate for this reaction that differs from the rates we reported (3) by more than an order of magnitude. In their study, Carr et al. report being unable to detect any OH from the reaction in the vicinity of 560 nm, whereas we reported detecting OH products over the region from ∼560 to 635 nm by electronically exciting NO2 in the presence of water vapor (3). The reasons for the differences in the results of the two studies are not obvious, although it is well known that the reaction under consideration is a difficult one to study because of rapid quenching of the electronically excited NO2 by water and the low product yields (13). The main difference in the approaches between the two studies apparently involves the laser power density used to excite the NO2 reagent and the calibration reaction used to quantify OH yields.

In our experiment, we used a lens to introduce the excitation laser into the reaction cell and electronically excite NO2 (3). This was necessitated by the low electronic absorption cross section of NO2 over the red end of the spectral range covered in our study. By contrast, Carr et al. did not use a lens, and thus their measurements used an unfocused laser beam (2). With regard to the OH calibration reaction, we used OH from the vibrational overtone–induced unimolecular dissociation of CH3OOH (5νOH) to calibrate OH yields, whereas Carr et al. used the photodissociation of acetone followed by the reaction of the resulting acetyl radical with O2 for their calibrations. Carr et al. suggest that the difference in their results and those of our study are likely due to multiphoton effects contributing to the OH signal in our study arising from the greater laser power density created by the presence of the lens. Carr et al. cite the presence of an intercept appearing in the linear power dependence plot in (3) as indicative of multiphoton effects. However, the authors do not provide any particular mechanism to suggest how the OH radicals we observed are produced by the multiphoton excitation process, nor do they provide any data on the effect of increasing laser power density in their measurements on the OH yield. In (3), we suggested that the intercepts in our power dependence plots likely arose from our detection electronics. Indeed, we looked at the power dependence for OH formation from the unimolecular dissociation of CH3OOH (5νOH) and HNO3 (5νOH), which are both linear processes, using the same experimental setup as that used for the NO2*+H2O study and found small intercepts in the power dependence plots for these unimolecular processes as well (intercepts of ∼1 to 2.5 mJ, corresponding to 6 to 8% of the peak power range). Hence, the presence of an intercept in an otherwise linear power dependence plot does not in itself signify the occurrence of multiphoton effects.

Based on the OH spectrum recorded and the action spectrum presented in (3), we have little doubt that we observed OH from the excitation of NO2 in the presence of water. Could the OH we detected arise from multiphoton effects? As reported in (3), we considered the possibility of several multiphoton mechanisms and found no compelling evidence for their occurrence. First, multiphoton dissociation of NO2 can result in the production of electronically excited oxygen atoms such as O(1D). However, it is well known that a common characteristic of O(1D) atoms is that they react rapidly with H2 to produce OH radicals (4). When we introduced H2 into the experimental cell in place of H2O, we did not observe any OH being produced (3). This lack of OH formation, which would be expected if O(1D) were present, nullifies the production of these electronically excited atoms through multiphoton excitation of NO2. A second multiphoton process that was also considered involved the possibility of producing translationally hot ground state O(3P) atoms from the multiphoton dissociation of NO2. These oxygen atoms could in principle then react with water to produce OH radicals. This scenario was ruled out on the basis that the reaction of O(3P)+H2O is endothermic by roughly 17 Kcal/mole (4). Measurements of Brouard and Vallance (5) show that O(3P) atoms generated from the photodissociation of NO2 at 308 nm have an average translational energy of 0.3 eV (∼6.9 kcal/mole) with a full-width at half-max of 0.2 eV. As we also observed OH production for NO2 excitation wavelengths between 610 and 630 nm, which corresponds to a two-photon excitation energy equivalent to exciting between 305 and 315 nm, the results of Brouard and Vallance then suggest that the average translational energy of the oxygen atoms potentially generated by two-photon dissociation of NO2 at these wavelengths will not be sufficiently energetic to initiate the O(3P)+H2O reaction. Hence, this multiphoton mechanism for OH production was also ruled out. Third, the possibility of populating long-lived bound excited quartet electronic excited states of NO2 through multiphoton excitation, which then reacts with H2O, has also been considered. According to the calculations of Bera et al. (6), NO2 has several excited quartet states in the vicinity of 3.61 to 4.37 eV. Accessing these bound excited electronic states of NO2 from the ground state requires the involvement of spin-forbidden transitions. These bound excited states could potentially be accessed through sequential two-photon excitation by the intermediate A 2B2 electronic state of NO2 initially populated through excitation in the visible range. However, spectral features of bound-to-bound transitions such as these should be reflected in their characteristic structured excitation spectrum. The action spectrum generated from the NO2*+ H2O reaction in (3), however, match that expected from the traditional X 2A1 → A 2B2 excitation of NO2 in the visible range. Also, in addition to being spin-forbidden with respect to their photochemical formation, the subsequent bimolecular reaction of these quartet NO2 states with H2O to form OH + HONO is expected to be spin-forbidden as well. Finally, we also considered the possibility that the primary source of OH signal in our experiments were due to the collision of electronically excited nitrogen dioxide, NO2*, with vibrationally excited water through the following reaction sequence. $Math$(1) $Math$(2)

The pseudo-first order rate constant for OH production through this mechanism would be expected to go as the square of the NO2 concentration. Our experiments using different starting NO2 concentrations did not show this squared dependence for the pseudo-first order rates; hence, this mechanism was ruled out as the main source of the observed OH signal, although it could contribute to small background levels as noted in (3). Thus, based on the above reasoning, it is difficult to justify the suggestion that the OH signal observed in our study and the rate we measured arise from multiphoton effects.

View Abstract