In an extension of our earlier studies at lower temperatures [4,51 the title reaction was measured directly in a flow reactor at temperatures of 600 and 700 K. The pressure of 0.65 mb was chosen that low in order to reduce the contribution of the stabilization channel. OH was used in an excess over CH3. Both reactants along with the reaction products were monitored by mass spectrometry. CH3 profiles served as the major observable quantity for the extraction of rate data. This had to be done by using computer simulation since it was impossible to work under pseudo-first-order conditions.
The obtained total rate coefficients were divided into channel rate coefficients by means of branching ratios as determined by the mass spectrometric measurement of the reaction products. For CH3 + OH, this led to a rate coefficient, k l , into the stabilization channel, and another one, kl,+f referring to the sum of two Hz-eliminating channels yielding the biradical HCOH and to a minor extent H2CO. These latter channels have not been measured before.
In order to distinguish between them we switched over from OH to OD to get ( l'e) CH3 + OD -HCOD + H2
(1'0 -HzCO + HD so that the biradical and/or aldehyde channels could be determined by their by-products Hz and HD, respectively. The use of OD makes it also possible to measure the channel (I'd)
CH3 + OD -lCH2 + HDO through its by-product, HDO.
within our error limits no significant isotope effect takes place.
units of cm, molec, and s:
A comparison of the rate coeficients of both systems, i.e., CH3 + OH and CH3 + OD, indicates that
For the rate coefficient into the HCOH channel, we arrive at a preliminary Arrhenius expression in k1, = 9.1 x 10-l~ exp(-1500/~).
The H2CO channel could not be detected at our lower temperature rendering us with a rate coefficient at 700 K klf(700 K) = 1.7 X 10-l'
Since simulation is needed for the deduction of the total rate coefficients as well as of the branching ratios, an uncertainty factor of 1.5 has to be attributed to these numbers.