It has been known for some time from infrared chemiluminescence experiments that a nonthermal rotational distribution of hydrogen halide peaked initially at high rotational quantum number, J, relaxes to a thermal distribution without generating a peak at intermediate J [Discussions Faraday Soc. 44, 183 (1967)]. It is shown in the present study that this characteristic pattern of relaxation is well described by a model according to which ΔJ is unrestricted, except for the relation PJ−ΔJJ=N exp (— CΔE), where PJ−ΔJJ is the probability of a collision−induced transfer from J to J — ΔJ, ΔE is the energy difference between these two rotational states, and N and C are constants. This expression for PJ−ΔJJ ascribes a very much lower probability of rotational deactivation to the higher J levels. Three other, contrasting, models were tested; they were rejected since they failed to describe the observed pattern of relaxation adequately. Upper limits were ascribed to PJ−ΔJJ for ΔJ=1–5 in HCl–H2 collisions. This study provides a further method for correcting infrared chemiluminescence data for modest rotational relaxation and at the same time shows that the simple truncation correction procedure used until now is remarkably good.

The infrared chemiluminescence (``arrested relaxation'') approach has been applied to the measurement of initial vibrational, rotational, and translational energies (V′, R′, and T′) in the products of the exothermic reactions F + H2 → HF + H and F + D2 → DF + D. Detailed rate constants k(V′, R′, T′) are reported as contour plots. The total detailed rate constants into specified vibrational quantum states (summed over the rotational levels of each v′ level) are: (i) for F + H2, k(v′ = 1) = 0.31, [k(v′ = 2) = 1.00], k(v′ = 3) = 0.47; (ii) for F+D2, k(v′ = 1) = 0.28, k(v′ = 2) = 0.65, [k(v′ = 3) = 1.00], k(v′ = 4) = 0.71. Both reactions convert the total available energy quite efficiently into internal excitation of the new molecule. The mean fractions entering vibration plus rotation are: (i) for F+H2, f̄V′ + f̄R′ (= 0.66 ± 0.08) = 0.74; (ii) for F + D2, f̄V′ + f̄R′ ( = 0.66 + 0.08) = 0.74. The fractional conversion of available energy into vibration is comparable to that for the Cl+HI reaction (Part IV of this series) and markedly greater than that for H+Cl2 (Part V). It seems probable that the energy release is predominantly ``repulsive'' in all these reactions, but is more efficiently channeled into product vibration if the attacking atom is heavy (Cl+HI, F+H2). As in the case of other isotopic pairs of reactions (Parts IV and V of this series) there is a parallelism in k(V′) though not in k(v′) between the members of the pair. The present reactions exhibit only a small increase in product rotational excitation with decreasing vibrational excitation (— Δ R̂′ / Δ V′ more closely resembles that for the H+Cl2 reaction than that for the reaction Cl+HI). It follows that the translational energy of the products is markedly greater for successively lower v′ states.

The distribution of vibrational, rotational, and translational energies (symbolized by V′, R′, and T′) have been obtained for the products of the isotopic pair of reactions Cl + HI → HCl + I and Cl + DI → DCl + I. The experimental method was the ``arrested relaxation'' variant of the infrared chemiluminescence technique. Detailed rate constants k(V′, R′, T′) are reported in the form of contour plots for these reactions. The total detailed rate constants into specified vibrational quantum states for Cl+HI (summed over the rotational levels of each v′ level) are k(v′ = 1) = 0.22, k(v′ = 2) = 0.35, [k(v′ = 3) = 1.00], k(v′ = 4) = 0.74; and for Cl + DI, k(v′ = 1) ≈ 0.08, k(v′ = 2) = 0.14, k(v′ = 3) = 0.35, k(v′ = 4) = 0.73, [k(v′ = 5) = 1.00], k(v′ = 6) = 0.05 [relative to the highest rate constant k(v̂′) = 1.00, in each case]. Preliminary energy distributions are also reported for the products of two other reactions, Cl + HBr → HCl + Br and Br + HI → HBr + I, in the X + HY → HX + Y family (X and Y are halogen atoms). The members of this family of reactions channel a substantial fraction of the energy available to the products into vibrational and rotational excitation, and only a small fraction into relative translation. For the Cl+HI and also the Cl+DI reaction the fractions are f̄V′ + f̄R′ (= 0.71 + 0.13) = 0.84; in contrast to f̄T′ = 0.16. As a corollary there is a marked inverse correlation between vibrational and rotational excitation in the reaction products. Despite the fact that the detailed rate constants into specified product quantum states [k(v′, J′)] are markedly different for the isotopic pair of reactions, the fractional conversion of the available energy into vibration, rotation, and translation are in close agreement (to ∼ ± 1%). This close parallelism in f̄V′, f̄R′, f̄T′ is in accord with predictions from classical trajectory studies. It indicates the usefulness of such calculations even for the case of reactions which yield products with widely spaced (vibrational and rotational) quantum levels.

The ``arrested relaxation'' infrared chemiluminescence technique has been used to obtain the distribution of vibrational, rotational, and translational energies (V′, R′, and T′) in the products of the exothermic reactions (i) H + Cl2 → HCl + Cl, (ii) its isotopic analog D + Cl2 → DCl + Cl, and (iii) H + Br2 → HBr + Br. Detailed rate constants k(V′, R′, T′) are reported in the form of contour plots for these three reactions. The total detailed rate constants into specified vibrational quantum states (summed over the rotational levels of each v′ level) are: (i) for H + Cl2, k(v′ = 1) = 0.28, [k(v′ = 2) = 1.00], k(v′ = 3) = 0.92, k(v′ = 4) = 0.1, k(v′ = 5) = 0.05, k(v′ = 6) = 0.005 [the values for v′ = 5 and 6 are from P. D. Pacey and J. C. Polanyi, J. Appl. Opt. 10, 1725 (1971)]; (ii) for D + Cl2, [k(v′ = 1) ∼ 0.1], k(v′ = 2) ≈ 0.3, [k(v′ = 3) = 1.00], k(v′ = 4) = 0.9, k(v′ = 5) = 0.3, k(v′ = 6) = 0.06; (iii) for H + Br2, [k(v′ = 1) = 0.03], k(v′ = 2) = 0.18, [k(v′ = 3) = 1.00], k(v′ = 4) = 0.99, k(v′ = 5) = 0.2, k(v′ = 6) ≤ 0.002. All of these reactions exhibit a comparatively low fractional conversion of the total available energy into internal excitation of the new molecule. The mean fractions entering vibration plus rotation are (i) for H + Cl2, f̄V′ + f̄R′ ( = 0.39 + 0.07) = 0.46; (ii) for D + Cl2, f̄V′ + f̄R′ ( = 0.39 + 0.10) = 0.49; (iii) for H + Br2, f̄V′ + f̄R′ ( = 0.55 + 0.04) = 0.59. The relatively inefficient conversion of reaction energy into vibration is thought to arise from the combined effect of ``repulsive'' energy release and a light attacking atom. As in the case of other isotopic pairs of reactions there is a close parallelism in k(V′) though not in k(v′) between the members of the pair (H+Cl2 and D+Cl2 in the present case). The increased fractional conversion of the available energy into vibration for H+Br2 as compared with H+Cl2 is indicative of a less-repulsive potential-energy surface. This is in accord with expectation, based on the change in barrier height and consequent change in barrier location on the energy surface. These reactions exhibit only a very small increase in product rotational excitation with decreasing vibrational excitation. It follows that the translational energy of the products is markedly greater for successively lower v′ states.

Angular distribution measurements of reactive scattering of a supersonic potassium atom beam by mercuric halide molecules HgX~ are reported with initial kinetic energies E ~ 6 kcal mole -1. Ail~ the reactions exhibit stripping dynamics with strong forward peaking in the centre of mass differential cross sections and large total reaction cross sections, Qr'~ 150 A*. However, there is substantial backward peaking ( ~ forward peak) for HgBr2, HgIz. A major fraction of the reaction exoergicity is disposed into product vibrational excitation. Just as in the alkali atom-halogen molecule reactions, the reaction dynamics are explained by an electron jump in the entrance valley of the potential surface. However, the dissociation of the HgX2-ion involves rather different forces from the X~-ion without having a dramatic effect on the differential reaction cross section.