Evaluation of HF product distributions deduced from infrared chemiluminiscence. II. F atom reactions

Abstract

The HF product state distributions from F+CH 4(CD 4), C 2H 6, CH 4-nCl n series, CH 3F, CD 3OH, CH 3OD and H 2S have been studied by the infrared chemiluminiscence technique in two different laboratories with two types of reactors: a fast-flow system with = 1 Torr of Ar buffer gas and a low-pressure, cold-wall system (usually called the arrested-relaxation method). The same HF Einstein coefficients were used in both laboratories to convert intensities to populations and emphasis is placed upon evaluation of the reliability of the resulting HF vibrational-rotational distributions. Arrested-relaxation data from Frankfurt for the F+D 2O reaction, which populates DF(υ⩽2) are also presented. Well-behaved data were obtained for CH 4(CD 4), C 2H 6 and H 2S in both laboratories and in both reactors. For the CH 4-nCl n series the HF(υ, J) distributions depend upon the design of the arrested-relaxation vessel and upon the operating conditions, especially flow rates. The lowest-pressure arrested-relaxation distributions strongly disagree with flow-reactor results. It is argued that the flowing-afterglow data provide initial HF(υ) distributions for 300 K Boltzmann reaction conditions. The AR data are compared to this work. One possibility for explaining the arrested-relaxation results, which fits many of the diagnostic tests, is the existence of two reaction channels: one directly gives HF, the second involves formation of adducts with F atoms, which subsequently react to give HF in low vibrational levels. For CH 3F the discrepancy between flow-reactor and arrested-relaxation results most likely stems from an interfering secondary reaction. Unusually rapid HF(υ)/DF(υ) vibrational relaxation is observed for methanol. However, initial HF(υ)/DF(υ) vibrational distributions in the flow reactor were obtained. Arguments based on surprisal analysis suggest that the F+H 2O and D 2O systems yield inverted HF(υ)/DF(υ) primary distributions.