Two previous theoretical investigations of the three body association reaction He++He+He→He+2+He, both of them concerned with molecule formation by a mechanism largely classical in nature, are extended to take into account two hitherto neglected quantum effects, each involving resonant quasibound states (RQS) of the He+2 molecular ion and each involving tunneling through the rotational barrier. One of the quantum effects is the effect that the discrete nature of RQS has on outward tunneling after quasibound states are formed by the classical mechanism; and the other, which was proposed originally by Dickinson, Roberts, and Bernstein, is collisional deexcitation of RQS formed by inward tunneling. While it is found that taking into account the discrete nature of RQS in determining the effect of outward tunneling re− sults in only minimal changes in the rate coefficient, it is also found that the formation of such states by inward tunneling can account for the 10% to 20% discrepancy between our previous calculations and the experimental results of Johnsen, Chen, and Biondi. As in previous work, computations were performed using four assumed forms of the He+3 interaction; and while statistical accuracy now permits some of these interactions to be distinguished from one another, each gives a rate coefficient with magnitude and temperature dependence in agreement with experiment. A considerable amount of discussion is devoted to some of the approximations that have been made, principally our use of the JWKB approximation in identifying RQS and our neglect of collisional breakup of newly formed He+2 ions. It is argued that the inaccuracies introduced by these two approximations are likely to be of opposite sign and to have magnitudes not much greater than the experimental error. The uncertainties in our calculation notwithstanding, it is concluded that the inward tunneling mechanism surely plays a significant, though not dominant, role in He+ recombination. Our calculations of the rate coefficient due to inward tunneling are several times smaller than the very rough estimates accompanying the original suggestion of this mechanism, and it is noted that this difference may be due largely to our estimates of cross sections for collisional deexcitation of RQS, which are obtained using the Monte Carlo trajectory method and which have the appearance of being several times smaller than the simple gas kinetic estimates employed by Dickinson et al. Some other possible explanations of this difference are noted.
Read full abstract