Low-lying electronic states of FeNC and FeCN: A theoretical journey into isomerization and quartet/sextet competition

Abstract

With several levels of multireference and restricted open-shell single-reference electronic structure theory, optimum structures, relative energetics, and spectroscopic properties of the low-lying (6)Delta, (6)Pi, (4)Delta, (4)Pi, and (4)Sigma(-) states of linear FeNC and FeCN have been investigated using five contracted Gaussian basis sets ranging from Fe[10s8p3d], C/N[4s2p1d] to Fe[6s8p6d3f2g1h], C/N[6s5p4d3f2g]. Based on multireference configuration interaction (MRCISD+Q) results with a correlation-consistent polarized valence quadruple-zeta (cc-pVQZ) basis set, appended with core correlation and relativistic corrections, we propose the relative energies: T(e)(FeNC), (6)Delta(0)<(6)Pi (2300 cm(-1))<(4)Delta (2700 cm(-1))<(4)Pi (4200 cm(-1))<(4)Sigma(-); and T(e)(FeCN), (6)Delta(0)<(6)Pi (1800 cm(-1))<(4)Delta (2500 cm(-1))<(4)Pi (2900 cm(-1))<(4)Sigma(-). The (4)Delta and (4)Pi states have massive multireference character, arising mostly from 11sigma-->12sigma promotions, whereas the sextet states are dominated by single electronic configurations. The single-reference CCSDT-3 (coupled cluster singles and doubles with iterative partial triples) method appears to significantly overshoot the stabilization of the quartet states provided by both static and dynamical correlation. The (4,6)Delta and (4,6)Pi states of both isomers are rather ionic, and all have dipole moments near 5 D. On the ground (6)Delta surface, FeNC is predicted to lie 0.6 kcal mol(-1) below FeCN, and the classical barrier for isocyanide/cyanide isomerization is about 6.5 kcal mol(-1). Our data support the recent spectroscopic characterization by Lei and Dagdigian [J. Chem. Phys. 114, 2137 (2000)] of linear (6)Delta FeNC as the first experimentally observed transition-metal monoisocyanide. Their assignments for the ground term symbol, isotopomeric rotational constants, and the Fe-N omega(3) stretching frequency are confirmed; however, we find rather different structural parameters for (6)Delta FeNC:r(e)(Fe-N)=1.940 A and r(N-C)=1.182 A at the cc-pVQZ MRCISD+Q level. Our results also reveal that the observed band of FeNC originating at 27 236 cm(-1) should have an analog in FeCN near 23 800 cm(-1) of almost equal intensity. Therefore, both thermodynamic stability and absorption intensity factors favor the eventual observation of FeCN via a (6)Pi<--(6)Delta transition in the near-UV.

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

“…We already mentioned in Section 1 (see Table 1) that for FeNC, the r 0 bond length values derived from experiment by Lie and Dagdigian [1] are very different from the r e values calculated ab initio in the present work and in [7]. We will discuss the reasons for this in Section 6 below.…”

“…Table 1 quantifies the 'too-short CN bond length issue' for FeNC by summarizing some salient results of the ab initio calculations to be described below: it lists the electronic energy at equilibrium E e , the equilibrium bond lengths r e (Fe-N) and r e (N-C), the equilibrium bond angle \ e (Fe-N-C), and the equilibrium rotational constant B e . These quantities are compared with the results of the MR-SDCI + Q/[cc-pVQZ (Fe), aug-cc-pVQZ (C, N)] ab initio calculation of [7], and it is seen that the agreement is very good for the structural parameters; our calculation gives a somewhat lower value for E e . However, the table also gives the r 0 structural parameter values derived from experiment by Lie and Dagdigian [1], and it is seen that for the Fe-N internuclear distance, the experimentally derived r 0 -value is 0.07 Å , or 3.9%, longer than the ab initio r e -value from the present work.…”

“…Irrespective of the relatively high weight of the main electronic configuration given here and noted by DeYonker et al [7], we still need multi-reference calculations to treat the 6 D states in A 1 and A 2 symetries in equal weights under the C 2v symmetry constraint.…”

“…In 2004 (that is, subsequent to the experimental observation of 6 D FeNC [1]), DeYonker et al [7] carried out MR-SDCI + Q/[cc-pVQZ (Fe), aug-cc-pVQZ (C, N)] ab initio calculations to predict r e (Fe-N) = 1.940 Å and r e (N-C) = 1.182 Å forX 6 D FeNC. They speculated that the small value of r 0 (N-C) obtained experimentally in [1] could be caused by considerable uncertainties in the underlying, experimentally derived values for the rotational constants B 0 of FeN 12 C and FeN 13 C. However, similar 'too-short' experimentally derived values of r 0 (C-N) have not only been reported for FeNC, but also for a large number of other molecules.…”

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

“…For 6 D FeCN, our preliminary study mentioned above [22], carried out at the MR-SDCI + Q + E rel /[Wachters + f (Fe), cc-pVTZ (C, N)] level of theory, predicted equilibrium bond length values of r e (Fe-C) = 2.066 Å and r e (C-N) = 1.169 Å . DeYonker et al [40] carried out an extensive study of the low-lying electronic states of FeNC and FeCN. They predicted for 6 D FeCN, for example, that r e (Fe-C) = 2.0506 Å and r e (C-N) = 1.1682 Å at the MR-SDCI + Q/[cc-pVQZ (Fe), aug-cc-pVQZ(C, N)] level of theory and concluded, relying on single-reference CCSDT-3/[cc-pVQZ (Fe), aug-cc-pVQZ(C, N)] calculations, that 6 D FeNC should be more stable than 6 D FeCN by about 225 cm À1 .…”