ROTATIONALLY RESOLVED RATE CONSTANT MEASUREMENTS FOR REMOVAL OF CH ( A : A and B : E-) BY KETENE

ion reaction of CH(a4Z-) has an activation barrier. It has been noted that on the basis of spin and orbital symmetry conservation rules, some similarity can be established between the collisional quenching of CH and CH2; CH(aaE-) parallels the kinetic behaviour of CH2(X3B1) in which a barrier for abstraction reactions has been found; CH(X2II) undergoes insertion and addition reactions in resemblance with the alA state of CH2. Removal rate constants of CH(X2II) by several diatomic and polyatomic molecules are faster than for CH(AZA) although some exceptions have been found. For the higher lying CH(B2E-) more limited information is available;8,,6 quenching studies carried in atmospheric and low pressure hydrocarbon flames8, show that the B2E state is quenched more rapidly than the A2A state;1 however, recent measurements for several quenchers indicate that enhancement of quenching rates in going from the A2A to the BE states, does not always take place. In a recent work we have found that multiphoton dissociation of ketene with an ArF laser leads to formation of CH in several electronically excited states and we have observed that ketene is a very efficient quencher of the A2A and B2E states of CH. In order to accumulate data which could provide information about the quenching mechanisms of the CH radical in collisions with fast quenchers, we have studied the removal by ketene of several vibrational and rotational energy levels of the electronic A2A and B2Z states of CH.

Despite the amount of data reported, the mechanisms involved in the quenching processes are not completely understood. Studies of the cross-section dependence on temperature and rotational quantum number indicate that the quenching mechanism differs from that established for other hydrides 3,6,3 where collisional deactivation seems to be controlled, by long-range multipole attractive forces; some exceptions could be the quenching of CH(A2A) by highly efficient collision partners as acetone, 4 ammonia and chloromethanes.
The quenching properties of other electronic states of CH have also been the subject of a number of studies. The a4Z electronic state of CH in collisions with H2 and hydrocarbons shows rate constants 100-1000 times smaller than those obtained for CH(X2II); 5 this is in agreement with calculations predicting that the H-atom 207 208 JORGE LUGUE et al.
abstraction reaction of CH(a4Z-) has an activation barrier. It has been noted that on the basis of spin and orbital symmetry conservation rules, some similarity can be established between the collisional quenching of CH and CH2; CH(aaE-) parallels the kinetic behaviour of CH2(X3B1) in which a barrier for abstraction reactions has been found; CH(X2II) undergoes insertion and addition reactions in resemblance with the alA state of CH2. Removal rate constants of CH(X2II) by several diatomic and polyatomic molecules are faster than for CH(AZA) although some exceptions have been found. For the higher lying CH(B2E-) more limited information is available; 8,,6 quenching studies carried in atmospheric and low pressure hydrocarbon flames 8, show that the B2E state is quenched more rapidly than the A2A state; 1 however, recent measurements for several quenchers indicate that enhancement of quenching rates in going from the A2A to the BE states, does not always take place. 6 In a recent work we have found that multiphoton dissociation of ketene with an ArF laser leads to formation of CH in several electronically excited states and we have observed that ketene is a very efficient quencher of the A2A and B2E states of CH. In order to accumulate data which could provide information about the quenching mechanisms of the CH radical in collisions with fast quenchers, we have studied the removal by ketene of several vibrational and rotational energy levels of the electronic A2A and B2Z states of CH.
EXPERIMENTAL CH in the AZA, and B2] electronic states were produced following multiple photon dissociation of gas samples of ketene at room temperature, with an ArF excimer laser. Fluorescence from excited CH photofragments was observed at right angles with respect to the photodissociation laser. The fluorescence was focused onto the entrance slit of a 0.5 rn Jarrell-Ash monochromator and viewed by a photomultiplier EMI 9816QB. Time resolved fluorescence was recorded by a 40 MHz digital oscilloscope Tektronix 2430A (risetime of 12 ns and vertical resolution of 8 bits). Typically, 32 decay traces were averaged before being transferred to a personal microcomputer to be sto, red and analyzed.
Samples of ketene were synthesized by pyrolysis of acetic anhydric vapour at 500C and then degassed and purified by trap-to-trap distillation. Massspectrometric analysis of samples indicated no more than 1% impurities. Pressures were measured by a capacitive manometer MKS Baratron, type 227A (pressure range up to 1 Torr).

RESULTS
Quenching of CH(AZA) a) Quenching of CH(A2A, v" 0 and v" 1) Rate constants for removal of the v' 0 and v' 1 levels of CH(AZA) were measured. Time resolved emission from the overlapping Q-branches of the CH(AZA, v' 0 ---> ROTATIONALLY RESOLVED RATE 209 X2II, v"= 0) and CH(A2A, v' 1 --X2II, v" 1) bands at h 430.7 _+ 1 nm was recorded. We estimate that contribution to the total emission from v' 1 is smaller than 50% within the above bandwidth.
Decay traces at pressures of ketene in the range from 40 to 200 mTorr were fitted to single exponentials. A typical decay trace and semilogarithmic plot are shown in Figure 1. Linear Stern-Volmer plots obtained representing reciprocal decay lifetimes versus ketene pressure are shown in Figure 2. The rate constant for total quenching of CH(AZA) by ketene obtained from the slope of the Stern-Volmer linear graph is listed in Table 1. Deactivation of CH(A2A, v' 2) was studied by recording the time resolved fluorescence at 432.5 _+ 0.2 nm at different ketene pressures in the range from 30 to 150 mTorr; the isolated spectral region corresponds to the Q-branch of the CH(AZA, v' 2 --X2II, v" 2) transition without rotational resolution.
Fluorescence traces were analyzed finding that they followed monoexponential decays in the whole range of ketene pressures investigated. Stern-Volmer analysis of decay data lead to the linear representation shown in Figure 2; the rate constant for total removal by ketene and collision-free lifetime are listed in Table 1. To our knowledge no measurements of the radiative lifetime for the v' 2 level of CH(AZA) have been reported n the literature; similar values than those measured for v' 0 and for rotational levels with N' < 11 of v' 1 would be expected; 7,8 however the collision-free lifetime measured in the present work is shorter, which would be in agreement with previous indications of weak predissociation of this level. 7 210 JORGE LUGUE et al. good monoexponential decays for traces recorded at pressures of ketene below 80 mTorr. At higher ketene pressures, small deviations from the monoexponential behaviour at times longer than 600 ns could be observed; the latter could indicate that at high pressures of ketene, relaxation of levels with high rotational quantum fiumber can result in some population of the lower rotational levels, similarly to the behaviour observed in a number of systems. 4,[9][10][11][12][13][14][15][16][17][18][19][20][21] In order to separate this small long-living contribution, traces obtained at ketene pressure over 80 mTorr were analyzed from the point of maximum fluorescence intensity up to 15% of the maximum value; the analyses showed that in this region a single-exponential decay dominate the time resolved fluorescence.
Decay Measurements of the total disappearance rate constants of levels with higher rotational quantum numbers were also performed; in this case the experiments were carried at , 398.5 + 1.2 nm; within this wavelength interval, only rotational levels with N' 11-13 contribute to the emission. Pressure-dependent decay lifetimes were obtained fitting single exponentials to the recorded traces. Stern-Volmer analysis of traces obtained for low rotational levels gave a very good fitting to a straight line; the slope was about 25% smaller than that obtained for rotational levels with N' 11-13. Both linear plots are shown in Figure 4. Zero-pressure intercepts for low rotational levels lead to a value slightly slower than for higher N'; this is in accordance with previous measurements indicating that the radiative lifetimes for the B:Estate have a slight dependence on rotational quantum number. 7 Total removal rate constants and collision free lifetimes are listed in Table 1.

DISCUSSION
For a particular rotational level of CH(A2A, v' 0), the rate constants listed in Table  0 50 100 150 200 Pressure / retort Figure 4 Removal rates of several rotational levels of CH(B2Z-, v' 0) obtained from analysis of fluorescence decays" a) from levels with rotational quantum number N' 5-7; b) from levels with N' 11-13. analyzing the relative intensities of resolved rotational lines belonging to the Rbranch of the CH(A2A, v' 0 X2II, v' 0) emission spectra, recorded at different pressures of ketene. The analysis showed that CH(A2A, v' 0) is formed with rotational populations which cannot be described in terms of a single Boltzmann type distribution; 23,24 however we found that collisions with ketene in the pressure range from 40 to 250 mTorr produced only moderately small changes in the rotational distribution; the latter is also in agreement with the time resolved experiments reported above, in which only small deviations from monoexponential decays were obtained. From the spectral observations mentioned above, the contribution of rotational relaxation to the total disappearance of each level can be roughly estimated as less than 15% of the total rate constant. Therefore we conclude that, although some rotational relaxation takes place, the main contribution to the removal of the A2A state by ketene arises from electronic quenching. The rate constants measured in this work for quenching of several N' levels of CH(A2A, v' 0) in collisions with ketene, are of the same magnitude to those reported for colliders such as CS2 and large hydrocarbons and faster than those observed for NH3, 3'6 acetone 4 and the lowest alkane series. 2--4,6 For the fastest quenchers, such as NH3, it has been said that long range attractive forces are probably involved in the collisional relaxation mechanism. The same mechanism seems to take place in the collisional quenching by acetone; for the latter, measurements carried for rotational levels up to N' 7 show that increasing rotational quantum number leads to decreasing rate constants 14.
Removal of CH(A2A, v' 0) by ketene is not enhanced with rotational quantum number up to N' < 16. However it has been pointed that when highly efficient polar quenchers are involved, this behaviour would also be compatible with a relaxation mechanism dominated by the attractive part of the potential. 13 For several fast quenchers of CH(A2A), with high dipole moments, we have calculated thermally averaged cross sections at 300 K assuming a multipole model 22 and using the orientationally averaged expressions for the multipole forces. Calculated and experimental cross sections are given in Table 2. For ketene and ammonia a qualitative correlation is obtained; for the chloromethanes, although they are slower quenchers, the calculated cross sections are somewhat faster than those calculated for ketene and ammonia. Different reaction probabilities for each quencher, following complex formation, could account for the deviations; if this were the case, the lower probabilities observed for the chloromethanes would suggest that addition to the unsaturated bonds of ketene is more efficient than reactivity towards carbon-halogen bonds. 2) shows a significant increase of total deactivation rate by ketene with vibrational energy. Vibrational energy transfer in electronically excited states of other hydrides has been shown to be competitive with that for electronic quenching; 13 however for CH(AA) in atmospheric pressure flames, rate for vibrational energy transfer from v' 1 to v' 0 is only a small fraction of the electronic relaxation For v' 2, measurements of the ratio between vibrational energy transfer and electronic relaxation have not been reported in the literature; the higher efficiency of removal rate constant observed with increasing vibrational energy of CH(A2A) could reflect the contribution of vibrational relaxation; on the other hand, collisional energy transfer to the rotational manifold of v' 0 and v' 1 of CH(B2E-), lying on the same energy region, could significantly enhance the deactivation rate for the v' 2 level of AA.
Regarding the various terms contributing to the total removal rate constants measured for the B2E state, similar considerations to those discussed above could apply. We have found that rotational populations obtained at pressures of ketene between 80 and 500 mTorr can be fitted to hot Boltzmann distributions with temperatures which do not differ in more than 6% over the whole range of ketene pressures investigatedTM. This seems to indicate that the main contribution to removal of CH(B2E-) by ketene arises from electronic quenching. The total disappearance rate constant measured for the B2E state in the low N' levels is of the same magnitude as that obtained for the high rotational levels of A:A; at higher N' the rate is about 25% faster. In atmospheric pressure flames of methane, oxygen and nitrogen, dependence of rate constants on rotational quantum number was also observed although its degree was found to vary with the collisional environment of the flame. Comparison with recent data obtained for a wide number of quenchers, 6 shows that removal rate by ketene is of the same magnitude as that measured for hydrocarbons; for both, ketene and hydrocarbons, quenching of the B2E state is faster than for the A2A state. This increase of rate constant is only moderate in the case of ketene compared to the hydrocarbons where quenching of the B2E state is more than twice as efficient as for the A2A state. , 6 The enhancement of B2E state quenching rates observed with the hydrocarbons has been tentatively explained by the participation of an efficient B2Z ---) A2A energy transfer in which the energy mismatch would be disposed as vibrational excitation of the C-H stretching vibration in the hydrocarbons, rendering the process near resonant, t6 The smaller enhancement of rate constant in going from the AZA to the B2 states observed by ketene where less number of the appropriated vibrational modes are present, could fit within the above picture. In flames, electronic energy transfer from CH(B2Z-, v' 0) to CH(AZA, v' 1) has been observed. Under our experimental conditions, characterized by the broad rotational distribution in which both states are formed and by the low ratio between total populations of B2 to AZA, an increase of the A2A state populations due to energy transfer from B2 would be difficult to observe. Some final conclusions could be tentatively drawn: removal rate constants measured in the present work for the AZA state are close to those obtained if collisional complex formation is assumed; the rates measured with ketene are faster than those reported for large alkanes which is in agreement with previous results indicating that addition to the g-bonded systems of the unsaturated molecule enhances the quenching efficiency; further work would be needed to identify the reactive channels contributing to the total removal of excited CH by ketene. More systematic work with a large number of quenchers wotdd also be required in order to assess the influence of vibrational and electronic energy content on the removal rates; in collisions with ketene, rates tend to increase with increasing energy which would be consistent with the higher number of collisionally coupled relaxation channels available at higher energies.