A theoretical study of FeNC in the 6Δ electronic ground state

Abstract

We report an ab initio calculation, at the MR-SDCI + Q + E rel/[Roos ANO (Fe), aug-cc-pVQZ (C, N)] level of theory, of the potential energy surface for 6Δ i FeNC. From the ab initio results, we have computed values for the standard spectroscopic parameters of FeN 12C and FeN 13C. Analytical representations of the potential energy surfaces have been fitted through the ab initio points, and the resulting functions have been used for directly solving the rotation–vibration Schrödinger equation by means of the MORBID program and by means of an adiabatic-separation method. For 6Δ i FeNC, our ab initio calculations show that the equilibrium structure is linear with r e (Fe–N) = 1.9354 Å and r e (N–C) = 1.1823 Å. We find that the bending potential is very shallow, and the MORBID calculations show that the zero-point averaged structure is bent with the expectation values 〈 r (Fe–N)〉 = 1.9672 Å, 〈 r(N–C)〉 = 1.1866 Å, and 〈 ρ ¯ 〉 = 180 ° - 〈 ∠ ( Fe – N – C ) 〉 = 13 ° . The experimentally derived bond length r 0 (N–C) = 1.03(8) Å reported for 6Δ i FeNC by Lie and Dagdigian [J. Chem. Phys. 114 (2001) 2137–2143] is much shorter than the corresponding ab initio r e-value and the averaged value from MORBID. Our calculations suggest that this discrepancy is caused by the inadequate treatment of the large-amplitude bending motion of 6Δ i FeNC. It would appear that for floppy triatomic molecules such as FeNC, r 0-values have little physical meaning, at least when they are determined with the effects of the large-amplitude bending motion being ignored, i.e., under the assumption that the r 0 structure is linear.