Laser-Induced Fluorescence Study on Reactions of C / ( 2 P ) and C ( D , S ) with Alcohols Leading to CH ( X 2 ]-I ) Radicals at Thermal Energy

Thermal energy reactions of C+(2p) and C(1D,1S) with CH3OH, C2HsOH, n-C3H7OH and i-C3H7OH to give CH(X2H) radicals have been investigated by observing laserinduced fluorescence (LIF) spectra of CH(AmX) in a flowing afterglow. In the hydride ion abstraction by C (2p), only CH3OH exhibited the CH(A--X) LIF spectrum with a vibrational distribution of 1.0 0.62 for v"= 0 and 1, respectively. On the other hand, the CH(A--X) LIF spectra exclusively from v" 0 were observed in the reactions of C(1D,1S) with all alcohols studied. By isotopic substitution for CH3OH it was shown that the CH(X) radicals arise from the selective attack on the methyl site in the reactions of C/(2P) and C(1D, S) with methanol.


INTRODUCTION
Although rate constants for reactions of the C+(2P) ground-state ion and the C(1D,1S) metastable atoms with various molecules have been determined at thermal energy, [1][2][3][4][5] little information has been obtained 333 on internal energy distributions of products. We have recently used a flowing afterglow apparatus coupled with LIF detection to determine internal energy distributions ofproducts in ground electronic states.  In this paper, we report a flowing afterglow-LIF study on thermal energy reactions of C+(ZP) and C(1D,1S) with a series of aliphatic alcohols leading to CH(X2II).
Hydride ion abstraction from methanol by C/(2P) has recently attracted considerable interest, because the product CH2OH + ion is an important precursor in the interstellar formation of formaldehyde, ethanol and dimethyl ether. 11 The rate constant for the hydride ion abstraction has been determined to be 1.2 + 0.1 10 -1 cm3s -1 in an ion-cyclotron-resonance study of Anicich et al. and 5.2 10-10 cm3s in a selected-ion-flow-tube study of Adams and Smith.2. Very recently, Curtis and Farrar 12 have studied the hydride ion abstraction dynamics by using a crossed-beam technique in the relative energy range of 1.2-2.8 eV. They found that the hydride ion abstraction occurs predominantly at the methyl position on the basis of an isotopic labelling experiment with CH3OD. A barycentric polar flux contour map for the CH2OH + production demonstrated that the hydride ion abstraction proceeds through a direct stripping reaction in which large impact parameter collisions lead to product formation. To our knowledge, such kinetic and dynamic studies have not been performed on hydride ion abstraction from other alcohols.
The reactions of C(3p,1D,1S) with alcohols have been studied by Skell and Harris 13 in the condensed phase at near 77 K. They found that all aliphatic alcohols react with carbon atoms to undergo deoxygenation and insertion into every OH and CH bonds from the product analysis and isotopic labelling experiments. No work, however, has been done on these reaction systems in the gas phase.

EXPERIMENTAL
The flowing afterglow-LIF apparatus used in the present study was identical to that employed for studying the C(3p) + NO, N20 reactions leading to CN(X2E /)8 and the C / (2p) + hydrocarbon reactions leading to CH(X2II) 1 except for the installation of an ion-collector grid (nickel 20 mesh) before a reaction zone. The ground-state C/(2P) ions were generated through the dissociative charge-transfer reaction: He + + CO C+(P) + O + He. (1) On the other hand, the C(1D,1S) metastable atoms were prepared through a microwave discharge of Ar/CO mixture or through the following dissociative excitation involving radiative cascade from upper electronic states: The presence of C(1D) in the reaction zone was confirmed by observing the CH(AmX) LIF spectrum from the C(1D) + H2 reaction. 7'14'5 Although no direct information could be obtained on the presence of C(aS) in the reaction zone, it must be involved in the flow system on the basis of the fact that a weak C(P---S:248 nm) atomic line resulting from reaction (2) appears around a CO gas inlet. Reactions by C / (2p) and C(1D,1S) were distinguished using an ion-collector grid placed between the first entry port of CO and the second one of sample gas. We found that one grid to which a negative potential of 40-80 eV was applied with respect to the stainless-steel reaction cell sufficiently removed ionic active species without a great reduction of concentrations of active species. The .pressure in the reaction zone was 0. 6-1.4 Torr. for He, 5-70mTorr for CO and 5-10 mTorr for sample gases.
A tunable N2 laser pumped dye laser (Molectron UV22-DL14P) was used to detect LIF spectra of CH(A--X). The laser was tuned to the Av 0 sequence in the 425-435 nm region and the resulting LIF signal was monitored through an interferance filter. A cut-off filter was employed to record the total LIF signal from which the relative vibrational distribution of CH(X) was determined. LIF signals of a Hamamatsu Photonics R955 photomultiplier were fed to a boxcar integrator (NF Circuit Block Co.: BX-531) for signal averaging. The boxcar output was recorded on a chart recorder.

RESULTS AND DISCUSSION
In Table i is represented the energetics for the formation of CH(X) by reactions of C / (2p) and C(3p,D ,1S) with five aliphatic alcohols studied here. Thermochemical and spectroscopic data listed in Refs. [16][17][18][19][20]   were used for the calculation. It is clear from Table 1   The rotational distribution of CH(X) is thermalized at about 300 K, indicating rotational relaxation by collisions with He, CO and CH3OH is complete during the time between formation and detection. The vibrational population ratio between the v"= 0 and v"= 1 levels was determined by the same manner as that described previously, l The N1/N0 ratio in the C+(2P) + CH3OH reaction was estimated by subtracting the band intensity of CH(AX) LIF due to C(1D,1S) + CHaOH (e.g. Figure l(b)) from that due to C+(2P), C(1D,S) / CHaOH (e.g. Figure l(a)). Although the N1/N0 ratio was invariant over the He pressure range of 0.8-1.3 Torr and the CHaOH pressure range of 2-10mTorr, it was sensitive to the CO gas pressure. It was difficult to examine directly the dependence of the N/Na ratio on the CO gas pressure because He active species (He(23S), He+, and He2+) are not completely quenched by CO and directly react with CH3OH to yield CH(X) radicals at low CO pressures below about 10mTorr. Therefore, the correction for vibrational relaxation by CO was made by reference with the variation of the N1/N0 ratio in the C(3p) + HI reaction, 9 C(3p) atoms being produced through the Ar(3p0.2) + CO reaction. The nascent N1/N0 ratio was found to be 0.62 + 0.05. This ratio is smaller than a value of 0.84 predicted from a statistical prior distribution based on a three body model in which CH2OH + ions are treated as an atom 7 and a value of 0.70 based on an extended three body model in which three rotational degrees of freedom are included, e7 If the reaction dynamics of the hydride ion abstraction are governed by kinematics associated with the H + L--H' mass combination (H heavy, L light), a large fraction of the available energy is expected to be released as a vibrational energy of CH(X). e8 In such a case, much higher vibrational excitation of CH(X) is predicted based on a large available energy deduced from Table 1. The low vibrational excitation of CH(X) suggests that the dynamic factor associated with potential energy surfaces is more importantthan the kinematic factor in the present system. It also implies that most of the available energy is deposited into the internal energy of CHOH + and relative kinetic energies of products. The release of a large fractional energy to the internal energy of CHOH + not to CH vibration, may be explained by the fact that the abstraction process is sufficiently short that the residual alkoxyl radical cation does not relax before the products separate and the radical cation stabilization energy is converted into internal motion of CH2OH +.
In similar LIF measurements on larger aliphatic alcohols, no CH(A--X) LIF spectra was observed from the hydride ion abstraction from C2HsOH, C2HsCH2OH, and (CH3)2CHOH by C / (2p), indicating that the Habstraction is an unfavorable product channel for these larger alcohols. Although rate constants and product branching ratios have not been measured for the reactions of C + (2p) with these larger alcohols, by analogy with the known reaction for CH3OH, 1'2 the formation of parent cations by charge-transfer and that of fragment ions by dissociative charge-transfer and/or by insertion and elimination are probably dominant product channels for these alcohols with lower ionization potentials (IP)" IP values of CH3OH, C2H5OH, n-C3HTOH and i-C3HTOH are 10.83, 10.47, 10.22 and 10.12 eV respectively. 16 CH(A--X) LIF spectra exclusively from v" 0 were detected from the reactions of C(1D,1S) with the same substrates as in the case of CH3OH. This suggests that no energy is channelled into the CH(X) vibration. Possible reaction pathways of C(1D,1S) + alcohols leading to CH(X) radicals are direct abstraction of a hydrogen atom and insertion into a CH or OH bond followed by elimination of CH(X)  (5) C gg H---OCH3 CH + CH30. (6) As shown before, processes (4) and (6) are unimportant, though corresponding processes can not be ruled out in the cases of the other larger alcohols. In general, processes such as (3) and (4) can be distinguished from insertion/elimination processes (5) and (6) by observing the product vibrational excitation. 25'29 A large fraction of available energy is channelled into CH(X) vibration in the former case, while a much smaller fraction is deposited into CH(X) vibration in the latter case. On the basis of the present spectral observation, it is reasonable to assume that CH(X) radicals are produced in the reactions of C(1D,1S) with aliphatic alcohols via an insertion on a CH bond to form a long-lived intermediate species followed by elimination of the newly formed CH group. Most of available energy is probably released as an internal energy of alkoxyl radical and relative kinetic energies of products. However, the possibility that lifetimes of chemically activated intermediate complexes are sufficiently long to allow collisional deactivation by the He buffer gas cannot be completely ruled out under the relatively high He pressure used in the discharge flow system. In order to understand specific behaviors of low vibrational excitation and site selectivity in the C(aD,1S) + alcohol reactions to form CH(X), further experimental studies such as measurements of rotational distribution and angular distribution of products are required under single collision conditions.