Iodine Revisited

Abstract

(1) Previous estimates of the vertical energies of the valence-shell states of the iodine molecule are revised in the light of newer information. Correspondingly, a complete set of estimated potential curves is drawn for those valence-shell states which dissociate into ground-configuration atoms. In the light of this set of curves, conclusions in several recent papers are reviewed and in some cases revised. LeRoy's 0g+ excited state is reconsidered. Predissociations of the 2431, Π30+u(B) state are discussed. It is definitely concluded that the state mainly responsible for magnetic predissociation is the corresponding Π30−u state and not a Σ3u+(0−) state. It is, however, possible that a ∑3u+(0−) state, probably that of the 1441 configuration, makes some contribution to the magnetic predissociation near υ = 10. Spontaneous predissociation and collision-induced predissociation are also discussed. (2) Previous calculations for the estimation of the wavefunctions, relative energies, and transition probabilities from the normal state 2440, 1Σg+(X), of the several states of the 2431 configuration, are revised in the light of newer evidence. It is concluded that of the fairly strong visible absorption, perhaps 20% is due to the transition 1Πu ← X, the rest being 3Π0+u ← X and a little 3Π1u ← X. It is concluded that the 3Π0+u ← X transition moment results largely from case c mixing of 2431, 3Π0+g into X and relatively little directly from the mixing of 1441, 1Σu+ into 3Π0+u, but that a cross term arising from the joint effect of these two admixtures is also important. (3) Contrary to some writers, the conclusion that the intense uv room temperature absorption bands peaking at about 1825 Å (part of the Cordes bands) belong to the transition 1441, 1Σu+ ← X(D ← X) of the VN type, is strongly reaffirmed. The corresponding 1441, 3Σu+(1u) transition is identified with the weak I2 continuous absorption peaking near 2700 Å. Possible minor overlapping contributions from other weakly allowed transitions are discussed. The extension of the D ← X bands to longer wavelengths in high-temperature absorption is reviewed. An equation given by Wieland, aside from some uncertainty in the vibrational numbering, accurately describes the vibrational pattern of the D levels. The high-temperature absorption bands with low-frequency edges at λ 3427 and λ 3263 (Skorko bands) are, respectively, attributed to the transitions 1432, 3Π2g ← 2431, 3Π2u and 1432, 3Π1g ← 2431, 3Π1u which theoretically should be intrinsically intense like the D ← X bands. (4) Starting from semiclassical considerations in an accompanying paper on the role of kinetic energy in the Franck–Condon theory, and taking into account various experimental observations, it is concluded that the uv McLennan bands which culminate in a relatively intense series of fluctuations at about λ3250 are a part of the fluorescence spectrum resulting from excitation into vibrational levels of the D state; the resonance series at higher frequencies belongs to the same fluorescence spectrum (“primary fluorescence spectrum”). Other McLennan bands at longer wavelengths are attributed to transitions down to repulsive curves from other initial electronic states produced by collisions (“secondary fluorescence spectrum”). It is shown that the “fluctuation interval” in the λ3250 group has a magnitude, and varies with exciting frequency, in reasonable accord with the theory. From its magnitude approximate conclusions are drawn about the minimum energy and equilibrium internuclear distance for the D state potential curve. Further experimental work should be of interest. (5) The emission band systems obtained in the presence of considerable pressures of foreign gases (N2 or Ar) are discussed, and reasonable assignments for some of their upper, ion-pair type, initial states are given. In the case of the very strong region with maximum near λ 3425, proposals of Verma, Wieland, and Tellinghuisen are considered, but further work will be necessary before definite conclusions can be reached. Very likely more than one electronic transition may be involved.