ACCURATE DISSOCIATION ENERGIES FROM ROTATIONAL PREDISSOCIATION AND LONG-RANGE FORCES: B $^{1}\Pi$ LiH

$^{1}$ R.B. Bernstein, Phys. Rev. Letters 16, 385 (1966). $^{2}$ For a list, see W. C. Stwalley, Chem. Phys. Letters 7, 600 (1970). $^{3}$ R. Velasco, Can. J. Phys. 35, 1204 (1957).""Author Institution: Department of Chemistry, University of Iowa$Bernstein^{1}$ pointed out in 1966 that the centrifugal barriers of the effective potentials governing rotational predissociation were simply obtainable if the ordinary potential could be expressed as $$ V = D_{0} - C/r^{n}. \eqno{(1)}$$. He (and $others^{2}$) then used these formulas with predissociation data to obtain the constants $D_{0}$ and C for several diatomic molecules, with known long-range values of n. Unfortunately, it now appears that the centrifugal barriers considered in these cases were not at sufficiently large distances for (1) to hold. A search of the rotational predissociation literature produced a $case^{3}$ clearly in the long-range region: v = 2 of B $^{1}\Pi$ LiH. Since we have long-range predissociation in only a single vibrational level, we have chosen to calculate the long-range forces accurately and then use the predissociation data to obtain a precise dissociation limit (with respect to v = 0, J = 0 in the ground state). The effect of higher inverse power terms and tunnelling has also been considered. Our value for the dissociation limit (34 $492.8 \pm 0.5 cm^{-1}$) may be compared with the value given by $Velasco^{3}$ (34 $495 \pm 2 cm^{-1})$. If the width of either J = 2 or 3 of v = 2 were known even roughly, the dissociation limit uncertainty could be reduced to $ \mathop {<}\kern -8pt {\vphantom {1 \over 2}}^{\sim} 0.2 cm^{-1}$