The shape of the 21-valence-electron systems like the triatomic interhalogens is not well known due to lack of sufficient experimental and theoretical studies. So far only two such interhalogens, namely, CIF2 and C13, have been experimentally characterized [1,2]. Recently, Ungemach and Schaefer [ 3 ] have reported a b initio SCF calculations on CIF2 using basis sets ranging from minimal to extended plus polarized CI. Their best results indicate that the bond angle of this radical is 148.8", which compares favorably with the experimental estimate ( I 36" f 15") of Mamantov et al. [I]. Based on spectroscopic observations, Nelson and Pimentel [2] have concluded that the isovalent species, Cl3, is linear. No theoretical study has, however, been made to substantiate this result. We have, therefore, decided to perform CNDO/2 and INM3 calculations on these radicals to see whether predictions from such calculations tally with the a b initio and experimental findings. The proposed study would also throw some light on the relative performance of the CND0/2 and I N D O methods in the calculation of ground state properties of triatomic interhalogens, in general.
Recently, Deb et al. [4]have reported CNDO/2 calculations on the CIF2 radical using spd basis set for CI and sp for F. Their conclusion that this radical is linear is in conflict both with experiment [ I ] and ab initio SCF study [3]. There is also further evidence [5][6][7][8][9][10][11][12][13][14] that CNDO/2 (spd) calculations yield unreliable and inconsistent results for the interbond angles in molecules containing second-row atoms. We have, therefore, preferred not to included orbitals in the present study.Calculations Angular potential curves of CIF2 and C13 are shown in Figure 1. Calculated equilibrium geometries and other molecular properties of the two radicals are listed in Table I, which includes also the available experimental and ab initio theoretical values. It is observed that the CNDO/2 calculations give the geometry of the CIF2 radical in very good agreement with the ab initio SCF study [ 3 ] . The predicted bent structure for Cl3, however, disagrees with the experimental finding [2]. Although neither experimental nor ab initio theoretical value is available for the bond distance in Cl3, the present calculated value seems to be justifiable when compared with the ab initio calculations of Cl, [9]. The CI-F and CI-CI stretching force constants are overestimated to a great extent, while the bending force constants are in g o d agreement with experiment. Very low values for the barriers to linearity in CIF2 and C13 are in conformity with their large bond angles and small bending force constants. It may be noted that in the case of strongly bent HAB-type molecules [ 14) having bond angles in the range of 100°-l loo, both barrier to linearity and bending force constants are considerably higher than the corresponding values of CIF2 and C13. As regards the charge distribution * Teacher Fellow; on leave from the