OH/OD product state distributions arising from the reaction of gas-phase O(3P) atoms at the surface of the liquid hydrocarbon squalane C30H62/C30D62 have been measured. The O(3P) atoms were generated by 355 nm laser photolysis of NO2 at a low pressure above the continually refreshed liquid. It has been shown unambiguously that the hydroxyl radicals detected by laser-induced fluorescence originate from the squalane surface. The gas-phase OH/OD rotational populations are found to be partially sensitive to the liquid temperature, but do not adapt to it completely. In addition, rotational temperatures for OH/OD(v'=1) are consistently colder (by 34+/-5 K) than those for OH/OD(v'=0). This is reminiscent of, but less pronounced than, a similar effect in the well-studied homogeneous gas-phase reaction of O(3P) with smaller hydrocarbons. We conclude that the rotational distributions are composed of two different components. One originates from a direct abstraction mechanism with product characteristics similar to those in the gas phase. The other is a trapping-desorption process yielding a thermal, Boltzmann-like distribution close to the surface temperature. This conclusion is consistent with that reached previously from independent measurements of OH product velocity distributions in complementary molecular-beam scattering experiments. It is further supported by the temporal profiles of OH/OD laser-induced fluorescence signals as a function of distance from the surface observed in the current experiments. The vibrational branching ratios for (v'=1)/(v'=0) for OH and OD have been found to be (0.07+/-0.02) and (0.30+/-0.10), respectively. The detection of vibrationally excited hydroxyl radicals suggests that secondary and/or tertiary hydrogen atoms may be accessible to the attacking oxygen atoms.

We report the first measurements of internal energy distributions of the OH produced via a direct mechanism, isolated from other components on the basis of time-of-flight, in the interfacial reaction between gas-phase O((3)P) atoms and the liquid hydrocarbon squalane, C(30)H(62). O((3)P) atoms were generated by laser photolysis of NO(2) above the liquid. Resulting hydroxyl radicals that escape from the surface were detected by laser-induced fluorescence. Time-of-flight profiles demonstrate that the kinetic energy of the fastest OH (nu' = 1) is lower than that of (nu' = 0). Rotational distributions were measured at the rising edge of their appearance for both OH (nu' = 0) and (nu' = 1). They were found to differ substantially more than at the peak of their profiles. They were also less dependent on the bulk liquid temperature. We conclude that the new data confirm strongly that at least two mechanisms contribute to the production of OH. The higher-velocity component has translational and rotational energy distributions, observed cleanly for the first time, consistent with a direct mechanism. The close correspondence of these rotational distributions to those from the corresponding homogeneous gas-phase reaction of O((3)P) with smaller hydrocarbons suggests a very similar, near collinear direct abstraction. This is accompanied by a slower component with kinetic energy and rotational (but not vibrational) distributions reflecting the temperature of the liquid, consistent with a distinct trapping-desorption mechanism.

The relative reactivity of the liquid surface of a long-chain, partially branched hydrocarbon (squalane, C 30 H 62 ) with gas-phase O( 3 P) atoms has been measured as a function of liquid temperature. The O( 3 P) atoms were generated with a superthermal velocity distribution by 355 nm photolysis of NO 2 . Laser-induced fluorescence was used to detect the relative branching into specific OH product vibrational states. The yield of OH(V′)0) proves significantly less dependent on liquid surface temperature than the yield of OH(V′)1). Time-of-flight measurements of the escaping OH provide partially resolved product translational energy distributions. These profiles also differ between OH vibrational states. OH(V′)1) shows overall longer arrival times, but with a clear trend toward earlier times as the surface temperature is increased. OH(V′)0) shows little detectable variation of the distribution of arrival times over the range of temperatures investigated (263-333 K). We discuss the interpretation of these findings, taking account of earlier experimental work, which has indicated significant contributions from distinct "direct" and "trapping-desorption" reaction mechanisms, and new molecular dynamics simulations of the surface structure. There are a number of factors that may contribute, including both energetic and structural effects. It is not possible on the basis of the current evidence to discriminate conclusively between them. Nevertheless, we conclude, on balance, that structural effects may well be the more important. In particular, higher temperatures are predicted to promote more open structures. We speculate that this may enable more OH(V′)1) to escape before it is either vibrationally relaxed or, less probably, undergoes vibrationally enhanced reaction to produce H 2 O.

We have investigated the interfacial reactivity of gas-phase O( 3 P) atoms with a representative range of longchain liquid hydrocarbons. These consisted of two branched molecules, squalane (C 30 H 62 , 2,6,10,15,19,23hexamethyltetracosane) and pristane (C 19 H 40 , 2,6,10,14-tetramethylpentadecane), and three linear ones, n-docosane (C 22 H 46 ), n-tetracosane (C 24 H 50 ) and n-octacosane (C 28 H 58 ). This represents the first systematic investigation of reactions of this type for molecules other than squalane. The O( 3 P) atoms were generated by 355-nm laser photolysis of a low pressure of NO 2 above the liquid surface. The nascent gas-phase OH radical products were detected by laser-induced fluorescence (LIF). Measurements for the linear hydrocarbons were constrained by their vapor pressures to single temperatures slightly (∼1 K) above their respective melting points. Pristane was studied at the lowest temperature practically achievable. Squalane was compared as a reference at the full set of temperatures. Appearance profiles for all of the liquids showed similar characteristic differences between OH V′)0 and 1. LIF excitation spectra were obtained for each of the vibrational levels at both the rising edge and peak of the appearance profiles. We conclude that the observed variations in rotational temperatures are consistent with dual contributions to the reaction mechanism for all the liquids, involving both direct escape and trapping-desorption components of the observed OH, as has previously been proposed for squalane. The relative yields of OH showed some surprising dependences on the liquid, including an unexpectedly strong variation with linear hydrocarbon chain length. These cannot all be explained by the relative reactivity of primary, secondary, and tertiary H-C units. We discuss the possibility that the known "surface freezing" phenomenon for linear hydrocarbons may play a role.

Recent progress that has been made towards understanding the dynamics of collisions at the gas-liquid interface is summarised briefly. We describe in this context a promising new approach to the experimental study of gas-liquid interfacial reactions that we have introduced. This is based on laser-photolytic production of reactive gasphase atoms above the liquid surface and laser-spectroscopic probing of the resulting nascent products. This technique is illustrated for reaction of O( 3 P) atoms at the surface of the long-chain liquid hydrocarbon squalane (2,6,10,15,19,23hexamethyltetracosane). Laser-induced fluorescence detection of the nascent OH has revealed mechanistically diagnostic correlations between its internal and translational energy distributions. Vibrationally excited OH molecules are able to escape the surface. At least two contributions to the product rotational distributions are identified, confirming and extending previous hypotheses of the participation of both direct and trapping-desorption mechanisms. We speculate briefly on future experimental and theoretical developments that might be necessary to address the many currently unanswered mechanistic questions for this, and other, classes of gasliquid interfacial reaction.