On the Cl
*(
2P
1/2
) Reactivity and the Effect of Bend Excitation in the Cl + CH
4/CD
4Reactions
†Jingang Zhou,‡Jim J. Lin,‡,§ Bailin Zhang,‡and Kopin Liu*,‡,#
Institute of Atomic and Molecular Sciences (IAMS), Academia Sinica, P.O. Box 23-166, Taipei, Taiwan 106, Department of Applied Chemistry, National Chiao Tung UniVersity, Hsinchu, Taiwan 300, and
Department of Chemistry, National Taiwan UniVersity, Taipei, Taiwan 106 ReceiVed: February 12, 2004
The reaction of atomic chlorine with methane was studied in crossed-beam experiments at a collisional energy of Ec ≈ 4.7 kcal/mol. Using a time-sliced product imaging technique, several outstanding issues about this
reaction were addressed in this work. It was found that the reactivity of the spin-orbit excited Cl*(2P
1/2) atom
toward methane is negligibly small, in agreement with the assertion of recent studies. The excitation of methane bending/torsional modes exhibits only modest enhancement in reactivity, in contrast to the previous experiments. The product angular distributions are mainly backscattered for Cl + CD4, and sideways scattered for Cl +
CH4. The shapes of the distributions from the ground state and the bend-excited methane are remarkably
similar.
I. Introduction
The reaction of Cl + CH4f HCl + CH3has an important
role in the ozone production/depletion cycle in the stratosphere.1
As a result, it has been under extensive investigations, both experimentally2-5and theoretically,6-14over the past decades.
The reaction is endothermic by 1.21 kcal/mol and has a rate of 1.0× 10-13cm3molecule-1s-1at 298 K. Kinetics experiments
show non-Arrhenius behavior with a temperature-dependent activation energy of 2.4-3.2 kcal/mol and an approximate pre-exponential factor of A≈ 1 × 10-11cm3molecule-1s-1, which is∼30 times smaller than the hard-sphere collision rate at room temperature.2-5Such a small value of A is often indicative of
tight steric hindrance for surmounting the barrier to reaction. Non-Arrhenius behavior generally is not unusual and can be attributed to many factors. For the activated Cl + CH4reaction,
H-atom transfer is involved in chemical rearrangement. At low temperatures, tunneling must occur and its contribution to the deviation from linear Arrhenius plot cannot be neglected. In addition, Ravishakara and Wine5 ascribed the non-Arrhenius
behavior to differential reactivity of the spin-orbit excited Cl*(2P
1/2) reactants that are accessible under thermal conditions.
This hypothesis is counter-intuitive. Hence, since then, numerous kinetics studies have been devoted to solving this intriguing proposition.15-19However, because the measured quenching rate
constants are typically∼300 times faster than the rate of the ground-state Cl(2P
3/2) reaction, the collision of Cl*(2P1/2) + CH4
is simply overwhelmed by the physical quenching process, yielding Cl(2P
3/2) + CH4. Therefore, these kinetics studies cannot
decipher the relative reactivities of Cl*(2P
1/2) and Cl(2P3/2).
By analyzing the speed-dependent spatial anisotropy of the CH3product in a PHOTOLOC experiment, Kandel and Zare20
concluded that the Cl*+ CH
4reaction is not significant in the
collision energy range of 3-6.7 kcal/mol. They instead sug-gested the dominant role of the Cl(2P
3/2) + CH4(V2or V4) 1)
reaction in contributing to the non-Arrhenius behavior. A reinvestigation by Kim et al.,21using BrCl near 420 nm as the
photolysis Cl-atom source (which gives a Cl*yield of∼50%)
confirmed the two conclusions mentioned previously. The estimated vibrational enhancement factor in the rate from these studies, which is ∼200 (or 80) times faster for one quantum excitation of V2(torsion) or V4(bend) mode of CH4 (or CD4),
compared to the vibrational ground-state reaction,20,21 is,
however, substantially larger than the enhancement factor of 30 ( 15 that has been observed for the asymmetric stretching mode excited (V3) 1) CH4reaction.22,23This is in sharp contrast
to the earlier, rather indirect experiments in which no enhance-ment was found for either bending or torsional excited CH4
reaction.24,25In a very recent study of this reaction using the
PHOTOLOC approach coupled with velocity-map ion imaging, Bass et al. found that the estimated rate enhancements for bend-/ torsion-excited methanes are heavily dependent on how one treats the experimental resolution in fitting the data.26 With
proper accounting for the speed distribution of the photolyzed Cl reactant, a satisfactory simulation of the product image can be obtained without invoking the contribution from the vibra-tional excited methanes; otherwise, a significant enhancement factor will be concluded.
Theoretical investigations predicted, at most, only modest enhancement factors for the V2- and V4-mode excited
react-ants.12-14Very recently, Michelsen and Simpson27,28modeled
the non-Arrhenius behavior of this reaction from the viewpoints of both the kinetics and dynamics. They concluded that the curvature in the Arrhenius plot at temperatures above room temperature can be explained by the rate enhancement from the symmetric (V1) or asymmetric stretch (V3) excited CH4.
Non-Arrhenius behavior at lower temperatures is taken into consid-eration by both tunneling and a modest rate enhancement from bending or torsional excited CH4.
The purpose of the present study is to clarify the aforemen-tioned uncertainty or discrepancy on the vibrational enhancement †Part of the special issue “Richard Bersohn Memorial Issue”.
* Author to whom correspondence should be addressed. E-mail: kpliu@ gate.sinica.edu.tw.
‡Academica Sinica.
§National Chiao Tung University. #National Taiwan University.
7832 J. Phys. Chem. A 2004, 108, 7832-7836
10.1021/jp040128t CCC: $27.50 © 2004 American Chemical Society Published on Web 04/16/2004
factor for bending and/or torsional excited CH4. For
experi-mental reasons, most of experiments were performed for the Cl + CD4reaction.
II. Experiment
The experiments were conducted with the crossed-beam apparatus that has been described in detail previously.29In brief,
a dc high-voltage discharge source was used to generate the Cl-atom beam (3.5% Cl2in helium, at 6 atm.).30,31To prolong
the stability of the discharge beam, a scheme of pulsed high voltage with a pre-ionization pulse was adapted.29,32-34 The
amounts of Cl(2P
3/2) and Cl*(2P1/2) were not determined in this
study. The previous investigation, using a slightly different discharge scheme, indicated that the relative yield of Cl*(2P
1/2)
can be quite significant.30 A neat CH
4 or CD4 beam was
delivered from a heatable, pulsed valve (an Evan-Lavie valve) at 6 atm. Two source temperaturess340 and 463 K, from the thermocouple measurementswere used to vary the initial populations of V2- and V4-mode excited CD4. Their speedss1.1
and 1.29 km/s, respectivelyswere measured using two fast ionization gauges. To isolate the effects of the initial vibration excitation, the intersection angles of the two molecular beams were adjusted accordingly, so that the experiments under different source temperatures yield approximately the same collision (translation) energy of Ec ) 4.78 ( 0.03 kcal/mol.
The reaction product CD3(or CH3) was interrogated by (2 +
1) resonance-enhanced multiphoton ionization (REMPI), using the time-sliced ion velocity imaging technique.29,32-34The laser
frequency of the probe was fixed at the peak of the X2A 2ff
3pz2A2 00
0 Q-head for better image resolution. Consequently,
only the low N-states of the vibrational ground state of methyl radicals were sampled.32REMPI spectra, nonetheless, indicate
that they represent a very significant fraction of total reactivity.
III. Results and Discussion
A. Identification of Product Images. Figures 1a and 1b present the two raw images of CD3products at 340 and 463 K,
respectively. After the density-to-flux correction,29the
corre-sponding product velocity-flux contour maps, d2σ/dµd(cos θ),
are displayed in Figures 1c and 1d. Both images feature three backscattered, ring-like structures. The reaction of Cl + CD4 f CD3 + DCl is endothermic by 2.34 kcal/mol. Using the
conservations of energy and momentum, the most intense ring can readily be assigned to the ground-state reaction, Cl(2P
3/2) + CD4(V ) 0) f CD3(V ) 0) + DCl(V′) 0). The two outer
rings, labeled 0*and 0**, with faster recoil speeds, must then
orginate from internally excited reactants. The possible candidate from the Cl beam is the spin-orbit excited Cl*(2P
1/2) state that
lies 882 cm-1(2.52 kcal/mol) above the ground Cl(2P 3/2) state.
The dramatic increase of the 0*and 0**signals that is observed
when the secondary beam source is heated (see Figure 1d versus Figure 1c), however, implicates the vibrationally excited CD4
reactants as the major contributors. Among them, the excitations of the two low-frequency modes V2(1092 cm-1or 3.12 kcal/
mol, double degeneracy) and V4(998 cm-1 or 2.85 kcal/mol,
triple degeneracy) of CD4 are the prime candidates. The
outermost ring has a maximal kinetic energy of∼7.6 kcal/mol, as seen below. Hence, it is most likely from the Cl(2P
3/2) + Figure 1. Two raw images of the CD3(V ) 0) products from the Cl + CD4reaction at the same collision energy (Ec) 4.78 kcal/mol) but under different source temperatures of the CD4beam ((a) 340 and (b) 463 K). Overlaid on the images are the Newton diagrams. The dashed circles, labeled as 0, 0*, and 0**, represent the anticipated CD3(V′) 0) product speeds from the V ) 0, V
4) 1, and V4) 2 states of the CD4reactants, respectively. In panels c and d, the corresponding product velocity-flux contour maps are displayed. The obvious background in the forward direction, which was caused by the collision of CD4with the metastable He*in the discharged Cl beam, was discarded.
CD4(V4) 2) f CD3(V ) 0) + DCl(V′) 0) reaction, although
some minor contributions from the other excited states of CD4
such as V2) 2, V1and/or V3) 1 cannot be ruled out. The origin
of the middle ring, peaking at an average kinetic energy release of 4.8 kcal/mol, is somewhat ambiguous, because of the proximity of the energetics between Cl*(2P
1/2) and CD4(V4or V2) 1), which is beyond our energy resolution.
To differentiate the aforementioned two possibilities, the image of the Cl + CH4reaction was acquired and shown in
Figure 2a. To enhance the signals from the hot-band reaction,
the source temperature of the CH4beam was raised to 465 K,
for which a beam speed of 1.42 km/s was measured. Despite the background interferences, two clear rings in the sideways/ backward direction are observed. After subtracting the back-ground and performing the density-to-flux correction, the product velocity-flux contour map is shown in Figure 2b. Energetically, the inner feature corresponds to the Cl(2P
3/2) +
CH4(V ) 0) f CH3(V ) 0) + HCl(V′) 0) ground-state reaction.
The outer ring is separated from the inner one by ∼3.5 kcal/ mol, which is consistent with the energetics of 3.75 kcal/mol
Figure 2. Similar to Figures 1b and 1d, except the reaction is Cl + CH4at Ec) 4.60 kcal/mol. Obviously, the background problem is more severe when m/e ) 15 was detected. In addition to the two localized beam-related backgrounds, there is a widespread background centered around the laboratory origin, which was caused by the interactions of the UV laser with the pump-oil background. These backgrounds were subtracted, and the resultant contour map is shown in panel b.
Figure 3. CD3product speed distributions for the two source temperatures (a) 340 and (b) 463 K. The labels are the same as those in Figure 1.
The small shift of the peaks from the labeled marks reflect the rotational excitations of the two product rotors. The corresponding angular distributions are shown in panels c and d. The slight difference in shapes between panels c and d for the same labeled reaction could be the experimental uncertainty and/or could result from the slight variations of the CD4beam characteristics. (Note that the peaks in panel b are slightly narrower than
those shown in panel a.)
from the Cl(2P
3/2) + CH4(V4) 1) f CH3(V ) 0) + HCl(V′)
0) reaction. Nothing obvious, as will be shown below, is present at the energy anticipated for Cl*(2P
1/2) + CH4(V ) 0), clearly
demonstrating the insignificance of its contribution or reactivity, which confirms the previous assertion by Zare’s group.20,21
Analogously, we conclude thqat the assignment of the middle ring for the CD3 product predominantly originates from the
Cl(2P
3/2) + CD4(V4) 1) reaction.
B. Reactivity Enhancement of Bend-Excited Methane. Analysis of the product images provides more-quantitative determination of the relative reactivity. Integrating the contour map over the scattering angles, weighted by sinθ to account for the solid angle factor, yields the product speed distribution. The results for the CD3 products under the two source
temperatures are presented in Figure 3a. Three peaks are clearly resolved and can be assigned as noted previously. The integrated area of each peak then yields the respective contribution, S ) nσ, from the process of interest. The results from the speed distributions are summarized in Table 1. Also indicated in the table are the anticipated Boltzmann populations of the excited
V2and V4modes (the ground-state population is set as one),
assuming that the vibrational populations in our beam are similar to source thermal values. As is observed, both the ratios of S*/
S0and S**/S0of the two temperatures are, within 10%, in accord
with the thermal population enhancement factors of the V4mode,
e.g., (S*/S
0)463K/(S*/S0)340K≈ (nV4) 1/n0)463K/(nV4) 1/n0)340K, and
the agreement is slightly less so for the V2 mode. This
comparison suggests little vibrational cooling in our CD4(CH4)
beam, despite rather strong supersonic expansion that yields cold rotational and translational distributions.
Based on this information, we can now estimate the vibra-tional enhancements in reactivity (σ*/σ
0andσ**/σ0, as shown
in Table 1). The range of values is dependent on if we consider the V4mode solely or both the V4and V2 modes (with equal
reactivity). Roughly speaking, each quantum excitation results
in an∼3-fold enhancement in cross section. Similar analysis can be performed for the Cl + CH4reaction, as shown in Figure
4. Assuming that the vibrational population of the CH4beam is
equilibrated at the source temperature of 465 Ksin analogy to the CD4beamsand with S*/S0≈ 0.13, we deduce a reactivity
enhancement factor of 2-4 (depending on if V2) 1 is active
or not) for bend-excited CH4 at Ec ) 4.60 kcal/mol. These
factors are significantly smaller than the factor of∼80 for Cl
+ CD4and∼200 for Cl + CH4that were obtained by Zare’s
group20,21 but are more consistent with the suggestion by
Michelsen and Simpson.27,28
It is instructive to compare the vibrational enhancement factor to the reagent translational factor. The excitation function for the ground-state Cl + CD4reaction has recently been measured
in this laboratory. The unpublished results indicate that, with equivalent amounts of additional energy in the translational degrees of freedom, the enhancement factors in the reaction cross sections become∼3.5 and 4.3 times larger in the order of increasing Ec. Hence, the vibrational enhancement from the V4(and/or V2)-mode excitation is not much different from the
reagent translational energy, and, thus, is not mode-specific. This comparison should have a strong bearing in future modeling, such as the Michelsen’s approach,27,28of the kinetic behaviors
of this reaction.
A recent theoretical investigation indicated that the V4mode
of CH4 is intimately coupled to the reaction coordinate and
adiabatically correlated to the V2(umbrella) mode of the CH3
product.14This analysis corroborates the classical intuition of
the geometric change from the pyramidal structure of the CH3
moiety in the transition state to the planar methyl radical product. Yet, the observed enhancement factor of the vibrationally nonadiabatic product channel of the ground-state CH3does not
seem to be mode-specific. Further work on the formation of the V2-mode excited products, which will be the adiabatically
correlated channel, is planned.
Figure 4. Similar to Figure 3, except the reaction is Cl + CH4. The residual background was removed when partitioning the peaks in the speed
distribution. Note the absence of any significant signals at the location anticipated for the Cl*(2P
1/2) + CH4(V ) 0) reaction. The hot band feature
is more consistent with the assignment of V4) 1 than V2) 1. TABLE 1: Vibrational Enhancement Factors in the Cl(2P
3/2) + CD4(W2/W4) Reactiona S*/S0 S**/S0 nV4) 1/n0 nV4) 2/n0 nV2) 1/n0 nV2) 2/n0 340 K 0.161 0.0061 0.044 7.1× 10-4 0.02 2.1× 10-4 463 K 0.474 0.053 0.137 6.6× 10-3 0.068 2.3× 10-3 ratio(463 K/340 K) 2.94 8.7 3.1 9.3 3.4 11 σ*/σ0orσ**/σ0 3 ( 1 7.5 ( 2
aNote: The relative cross sections are deduced from S ) nσ.
We note that the effect of bend excitation of CH4on the
product HCl rotational distribution has also been investigated theoretically.35A cold unimodal rotational distribution of HCl
was found with ground-state reaction of CH4,35aand a unimodal,
although slightly hotter, distribution was found for the bend fundamental.35bThese results are in accord with the present
experiment and the previous observation.22The calculations also
predicted a bimodal HCl rotational distribution for the second bend-excited CH4. If this prediction also holds for CD4, this
would not be supported by the experiments on CD4(ν4) 2),
judging from the clear separation and the shape of our dσ/dµ distributions (Figures 3a, 3b, and 4a). Further theoretical studies are warranted.
C. Product Angular Distributions. Figures 3c and 3d show that the product angular distribution from the ground-state reaction of Cl + CD4is backward scattered and has a sharp
cutoff toward the forward hemisphere. This is in qualitative agreement, although quantitative discrepancies are discernible, with the previous measurements20,21 and with the theoretical
expectation that the transition state is highly collinear along the Cl-D-C axis. The observed angular distribution is very similar to that found for the Cl(2P
3/2) + H2reaction36but significantly
different from the F + CD4reaction and its isotopic variants.32-34
Although the product angular distribution for the Cl + CH4
ground-state reaction also indicates an abrupt cutoff, it exhibits a distinct sideways-scattered peak (see Figure 4b). The origin of the striking difference between the two isotopic reactions, probably reflecting the difference in the cones of acceptance of the two reactions, is currently under investigation. The present study also shows that the angular distribution for the bend-excited methane reaction is predominantly backward or sideways scattered with a shape similar to the respective ground-state reaction. This is to be contrasted with the previous finding when one quantum of asymmetric CH stretch (V3) 1) of CH4was
excited, for which a significant change in product angular distributions was noted.22,23
IV. Conclusions
To summarize, many important issues about the Cl + CH4
reaction have been addressed in this work, from which several conclusions can be drawn.
(1) The reactivity of Cl*(2P
1/2) toward CH4/CD4 seems
negligibly small, which is in sharp contrast to the previous, still controversial, finding for the analogous Cl/Cl* + H
2/D2/HD
reactions.30,31,36-39
(2) The supersonic expansion is not efficient in cooling the vibrational degrees of freedom of the CD4/CH4reactants.
(3) At variance with previous experiments,20,21the vibrational
enhancement factor with V4- and/or V2-mode excitation of CH4/
CD4in forming the ground-state CH3/CD3is quite modest (∼3
times per quantum excitation), which is not much different from the translational enhancement factor with equivalent amounts of energy.
(4) The DCl/HCl product rotational distributions are cold and unimodal, which is in accordance (or at variance) with a recent theoretical prediction for the ground-state (or the bend-excited) CH4reaction.
(5) Both the ground and bending/torsional excited methane reactants yield predominantly backscattered (for reaction with
CD4) or sideways scattered (for CH4) products, and their angular
distributions are remarkably similar.
Acknowledgment. This work was supported by the National Science Council of Taiwan (K.L. was supported by NSC No. 2113-M-001-040, and J.J.L. was supported by NSC No. 92-2113-M-001-044) and Academia Sinica. We thank M. Brouard and K. Han for sending us a copy of the manuscripts of refs 26 and 39, respectively, prior to their publication.
References and Notes
(1) Solomon, S. ReV. Geophys. 1999, 37, 275 and references therein. (2) Zahniser, M. S.; Berquist, B. M.; Kaufman, F. Int. J. Chem. Kinet.
1978, 10, 15.
(3) Keyser, L. F. J. Chem. Phys. 1978, 69, 214.
(4) Pilgrim, J. S.; Mcllroy, A.; Taatjes, C. A. J. Phys. Chem. A 1997, 101, 1873.
(5) Ravishankara, A. R.; Wine, P. H. J. Chem. Phys. 1980, 72, 25. (6) Truong, T. N.; Truhlar, D. G.; Baldridge, K. K.; Gordon, M. S.; Steckler, R. J. Chem. Phys. 1989, 90, 7137.
(7) Troya, D.; Millan, J.; Banos, I.; Gonzalez, M. J. Chem. Phys. 2002, 117, 5730.
(8) Wang, X.; Ben-Nun, M.; Levine, R. D. Chem. Phys. 1995, 197, 1. (9) Nyman, G.; Yu, H.-G.; Walker, R. B. J. Chem. Phys. 1998, 109, 5896.
(10) Yu, H.-G.; Nyman, G. J. Chem. Phys. 1999, 110, 7233. (11) Dobbs, K. D.; Dixon, D. A. J. Phys. Chem. 1994, 98, 12584. (12) Duncan, W. T.; Truong, T. N. J. Chem. Phys. 1995, 103, 9642. (13) Espinosa-Garcia, J.; Corchado, J. C. J. Chem. Phys. 1996, 105, 3517.
(14) Corchado, J. C.; Truhlar, D. G.; Espinosa-Garcia, J. J. Chem. Phys.
2000, 112, 9375.
(15) Wang, J. J.; Keyser, L. F. J. Phys. Chem. A 1999, 103, 7460. (16) Matsumi, Y.; Izumi, K.; Skorokhodov, V.; Kawasaki, M.; Tanaka, N. J. Phys. Chem. A 1997, 101, 1216.
(17) Tyndall, G. S.; Orlando, J. J.; Kegley-Owen, C. S. J. Chem. Soc., Faraday Trans. 1995, 91, 3055.
(18) Bryukov, M. G.; Slagle, I. R.; Knyazev, V. D. J. Phys. Chem. A
2002, 106, 10532.
(19) Chichinin, A. I. J. Chem. Phys. 2000, 112, 3772.
(20) Kandel, S. A.; Zare, R. N. J. Chem. Phys. 1998, 109, 9719. (21) Kim, Z. H.; Alexander, A. J.; Bechtel, H. A.; Zare, R. N. J. Chem. Phys. 2001, 115, 179.
(22) Simpson, W. R.; Rakitzis, T. P.; Kandel, S. A.; Lev-On, T.; Zare, R. N. J. Phys. Chem. 1996, 100, 7938.
(23) Simpson, W. R.; Rakitzis, T. P.; Kandel, S. A.; Orr-Ewing, A. J.; Zare, R. N. J. Chem. Phys. 1995, 103, 7313.
(24) Vijin, V. V.; Mikheev, A. N.; Petrov, A. K.; Molin, Y. N. React. Kinet. Catal. Lett. 1975, 3, 79.
(25) Chesnokov, E. N.; Strunin, V. P.; Serdyuk, N. K.; Panfilov, V. N. React. Kinet. Catal. Lett. 1975, 3, 131.
(26) Bass, M. J.; Brouard, M.; Cireasa, R.; Vallance, C. submitted to J. Chem. Phys.
(27) Michelsen, H. A.; Simpson, W. R. J. Phys. Chem. A 2001, 105, 1476.
(28) Michelsen, H. A. Acc. Chem. Res. 2001, 34, 331.
(29) Lin, J. J.; Zhou, J.; Shiu, W.; Liu, K. ReV. Sci. Instrum. 2003, 74, 2495.
(30) Dong, F.; Lee, S.-H.; Liu, K. J. Chem. Phys. 2001, 115, 1197. (31) Lee, S.-H.; Lai, L.-H.; Liu, K.; Chang, H. J. Chem. Phys. 1999, 110, 8229.
(32) Lin, J. J.; Zhou, J.; Shiu, W.; Liu, K. Science 2003, 300, 966. (33) Zhou, J.; Lin, J. J.; Shiu, W.; Liu, K. J. Chem. Phys. 2003, 119, 4997.
(34) Zhou, J.; Lin, J. J.; Liu, K. J. Chem. Phys. 2003, 119, 8289. (35) (a) Skokov, S.; Bowman, J. M. J. Chem. Phys. 2000, 113, 4495. (b) Bowman, J. M., private communication.
(36) Lee, S.-H.; Liu, K. J. Chem. Phys. 1999, 111, 6253.
(37) Alexander, M. H.; Capecchi, G.; Werner, H.-J. Science 2002, 296, 715.
(38) Manolopoulos, D. E. Science 2002, 296, 664. (39) Xie, T.-X.; Zhang, Y.; Han, K.-L., unpublished work.
