Theoretical Evidence from UQCISD Calculations against the Linear π-Configuration of α-Silylimidoyl Radicals As Determined by ESR Spectroscopy

Abstract

Geometries of imidoyl radicals Y−C(•)NMe have been determined by means of ab initio UHF/VTZP calculations. The local radical structure is found to be essentially indipendent of the electronic nature of the α-substituent, the ∠YCN bond angle being about 128°. This structural information is in contrast to that obtained from the magnitude of the experimental isotropic 13C hyperfine splitting (hfs) constants indicating that imidoyls bearing electropositive α-substituents such as silyl groups are nearly linear about Cα. UQCISD/VTZP calculations on H3Si−C(•)NH confirm that α-silylimidoyl radicals should be largely bent at Cα. At variance with what was previously found in other localized radicals, UMP2/DZP//UHF/VTZP calculations do not closely reproduce the hfs constants at the radical center but underestimate largely their values owing not only to slow convergence of the doubles amplitudes in Møller−Plesset perturbation theory but mainly to neglect of single excitations. Excellent agreement with experiment has been achieved at the UQCISD/DZP//UHF/VTZP level.

Imidoyl Radicals in Organic Synthesis

“…Since pseudooverlap is maximum for a linear geometry, this specific imidoyl radical was hence suggested to be close to linear at C , similarly to what proposed for the isoelectronic , -bis(trimethylsilyl)--(trimethylsilyl)vinyl radical, which has a comparable a( 13 C ) value of 28.1 G [23]. Nevertheless, some theoretical calculations seem in contrast with the conclusions drawn on the basis of the EPR experimental data, suggesting that the radical structure is essentially independent of the electronic nature of the -substituent, the N=C • -C angle being virtually constant at about 128° [24].…”

Role of Standard Diffuse Functions for Computing Hyperfine Splitting Constants in Radical Anions

“…The effect of electron correlation has been estimated employing the QCISD method . Core electrons have been held frozen since their inclusion is not expected to significantly influence optimum geometries and was found to give a small contribution to hfs constants using an unrestricted Hartree−Fock (UHF) reference determinant since spin-polarization effects are included explicitly. , …”

“…Good hfs constants for moderate-sized radicals were computed with the Møller−Plesset second-order perturbation theory (UMP2) using the small full double-ζ (DZP) and valence triple-ζ 6-311G(d,p) (VTZP) basis sets, which requires much less computational resources. However, this approach provides unreliable hfs constants when the underlying unrestricted Hartree−Fock (UHF) reference wave function does not describe well the doublet state − and even the determination of the basic structure is problematic . Hence, the evaluation of hfs constants with the UMP2 method must be checked case-by-case against the UQCISD method, , for example studying the simplest member of a class of radicals at both levels .…”

Overestimation of the stability of the π-delocalized versus the σ-localized configuration in radicals by current density functionals: the case of vinylacyl radicals

“…However, the hfs constants computed with these methods are largely in error as found previously for the isoelectronic imidoyl radicals [Y±C(·)@N-R] [32]. Comparison of the hfs values computed including only double excitations in perturbation theory (MP3) and in the coupled-cluster approach (UCCD), which takes account of the eect of double excitations through all orders of perturbation, indicates that only a minor part of this discrepancy is due to a slow convergence of double amplitudes.…”

“…The improvement due to contributions from single excitations estimated perturbatively [MP4(SDQ)] is small. Indeed, it was shown that single excitations should be included iteratively in post-HF calculations to obtain a correct spin density distribution when a localized unpaired r electron can interact with a lone pair in the presence of a polarizable p system [32].…”