1000
The Journal of Physical Chemistry, Vol. 83, No. 8, 1979
Infrared Chemiluminescence Studies of the Branching Ratios
K. Tamagake and
D. W. Setser
H -k CIF Reaction. Energy Disposal and
K. Tamagaket and D. W. Setser" Department of Chemistry, Kansas State University, Manhattan, Kansas 66506 (Received September 13, 1978) Publication costs assisted by the National Science Foundation
Infrared chemiluminescence under arrested relaxation conditions has been used to study the HF and HCl formation channels of the H + C1F reaction at 300 K. From the relative infrared emission intensities the ratio of channels was determined as 4.4 in favor of HC1, the less exoergic channel. The energy disposal is ( / " ) H a = 0.42, ( ~ R ) H C I= 0.14, ( ~ T ) H c=~ 0.44, and (f\-)HF = 0.57, ( ~ R ) H F= 0.10, (fT)HF = 0.33. The rotational distribution for the HF channel is much broader than for the HC1 channel, which is a consequence of two pathways for HF formation. The low J component is attributed t o direct reaction and the high J component to migratory collisions in which the H that intially attacked the C1 end of the molecule migrates and forms HF. Since -57% of the HF distribution can be assigned to a high J component, only about one-half of the HF is formed directly. The migratory aspect of the H + ClF reaction is less important than for the H + IC1 or BrCl reactions recently studied by Polanyi and co-workers. The energy disposal and inferred reaction dynamics for H + ClF are compared to that for H + Clz and H + Fz reactions. Using a numerical procedure, improved HC1 and HF transition probabilities were calculated and these values were used to convert the HC1 and HF rotational line intensities to relative populations. These new HCl and HF Einstein coefficients are presented in the Appendix. Introduction The H + X2 and H + XX' series of reactions (X = halogens) have been extensively investigated by Professor Polanyi and c o - ~ o r k e r s ~using - ~ the arrested-relaxation infrared chemiluminescence technique. The energy disposal from the H + Xz series has been used to beautifully display the dynamical features related to the attack of a light atom upon the heavy diatomic halogen molecule. Their most recent work5 with IC1 and BrCl showed that the HC1 rotational distributions differed dramatically from those of the H X2 reactions. The unusual breadth of the HC1 rotational distribution was attributed to microscopic branching in the HC1 formation channel; the low J component of the distribution was assigned to collisions in which the H interacted directly with C1 in the IC1 or BrCl molecule and the high J component of the distribution was assigned to collisions in which the H migrated to C1 after initially encountering I or Br. In our laboratory we have been studying the reactions of the lowest excited state (metastable) of the rare gas atoms with halogens, interhalogens, and halogen-containing molecules by the flowing afterglow techniquea6 By simulation of the bound-free rare gas halide ~ p e c t r a the ,~ vibrational energy released to the rare gas halide product can be estimated. For purposes of comparison, we have measured the energy disposal for some H atom reactions with the same reagents using the arrested relaxation infrared chemiluminescence technique. We also intend to use the H + C1F reaction as a reference for comparing absolute values of rate constants for a H + RC1 and H + RF (R = polyatomic) series of reactions using the technique that we recently applied to the F + HR reaction^.*^^ After initiating the H + C1F study, we learned that arrested relaxation work was being done by Brandt and Polanyi.lo Since the two sets of data were in good agreement, further work in our laboratory seems unnecessary and we report our experimental work here. The chemiluminescence results can be compared to the chemical laser study of the H + C1F reactionell The rate constants for reaction of H with Clz, F2, ClF, and IC1 are reported in the following paper.12
+
Faculty of Pharmaceutical Sciences, Okayma, Japan.
The H + C1F reaction is particularly amenable to study by the infrared chemiluminescence technique because both channels can be directly observed: H + C1F HClt + F Woo= -42.8 kcal mol-l (la) Woo= -75.6 kcal mol-l HFt C1 (Ib) Assuming that the activation energies for reactions l a and l b are the same as for ClZl3and FZl4(1.2 and 2.5 kcal mol-'), the available energy is 45.5 and 79.5 kcal mol-l for reactions l a and l b a t 300 K. Since some of the vibrational-rotational emission lines for HC1 and HF occur a t nearly the same wavelength, good resolution is essential and interferometric recording of the spectra is very helpful. From the total relative H F and HC1 emission intensities, the ratio of the HF and HC1 formation channels (macroscopic branching) can be determined. Furthermore, if the two channels yielding HFt have rotational distributions that are substantidy different, the ratio of the microscopic channels for HF' formation can be determined. We are especially interested in obtaining a reliable ratio of ( l a ) and ( l b ) from the HC1' and HFt emission intensities; therefore, improved H F and HC1 transition probabilities were calculated and these new values are reported in the Appendix. We also studied the H + Clz reaction in order to demonstrate the reliability of the arrested relaxation technique as currently employed in our laboratory:
-
H
+ C12
-
+
HCl? + C1
AHo, = -45.2 kcal mol-'
(2)
For an activation energyI3 of 1.2 kcal mol-l, the mean available energy for (2) is 47.9 kcal mol-l a t 300 K, which is similar to that for (la). The vibrational and rotational distributions from (la) and (2) can be directly compared since the data were obtained in the same apparatus. Experimental Section The experiments were done by observing the H + C1F infrared chemiluminescence under arrested relaxation conditions with a Digilab FTS-20 interferometer. The reaction vessel was a large aluminum box fitted with an
0022-3654/79/2083-1000$01.00/00 1979 American Chemical Society
Chernilurninescence Studies of the H t CIF Reaction
The Journal of Physical Chemistry, Vol. 83, No.
8, 1979
TABLE I: The H t ClF (and H t Cl,) Product Vibrational Distributions and Branching Ratio . reaction N, : N , : N , : N , : N ,: N , : N , H t ClF --f HClT t F first experiment second experiment H + C ~ F--f ' HFT t c i first experiment second experiment
measured distribution measured distribution
0.17:0.35:0.38:0.10 0.18:0.34:0.38:0.10
measured distribution measured distribution correcteda distribution direct channelb migratoryb channel
0.086: 0.121 :0.134: 0.170: 0.252: 0.206: 0.031 0.151:0.138:0.123:0.158:0.227:0.175:0.027 0.042: 0.097 :0.144: 0,191: 0.277 :0.216: 0.033 0.042:0.036:0.056:0.117:0.12:0.06 0,000: 0.061 :0.088: 0.074: 0.1 57 : 0.1 56: 0.033
HCl/HF
1001
-
3.98 3.90 4.4
H t C1, --f HClT t C1 first experiment 0.13:0.44:0.40:0.03 0.1 1:0.42:0.43 :0.04 second experiment a See text for method of correction of the low u levels for secondary reactions. The N o population is taken t o be zero for See text for method used t o deconvolute the HF distribution into the two microscopic channels. both HClT and HFf'.
inner double-walled aluminum box that could be cooled with liquid N,; the vessel was pumped with a 10-in. NRC diffusion pump. The inner box is of a dewar configuration and was constructed from high conductivity aluminum to ensure that all surfaces were at 77 K. Windows, fitted with NaCl disks, were placed in the ends of outer reaction vessel. The interferometer viewed the reaction zone through one window and a spherical backing mirror was placed behind the second window. The reagents were led to the center of the inner box and allowed to mix via a concentric nozzle arrangement; the diameters of the inner and outer tubes of the nozzle were 1.0 and 1.9 cm, respectively. A, mixture of Hz and H atoms flowed through the inner tube (quartz) and C1F was added through the outer tube. The hydrogen atoms were generated by a microwave discharge in Hz that had been bubbled through liquid HzO to enhance the dissociation in the discharge. The quartz inner tube was treated with phosphoric acid to inhibit loss of H atoms by surface reaction. For ease of maintaining the discharge, a teflon plug with a 4 mm hole was placed a t the end of the H / H z nozzle. The H2 and C1F or Clz flows were -5 and -100 wmol/s, respectively. The pressure measured in the reaction zone for such flow rates was