DFT Approach for Predicting ¹³C NMR Shifts of Atoms Directly Coordinated to Nickel

Abstract

The aim of this work is to develop a simple density functional theory (DFT) method that can be safely applied to calculate 13C NMR shifts in diamagnetic Ni complexes. A comparative analysis of calculated vs experimental 13C NMR shifts for a wide range of nickel complexes (157 in total) was carried out. A number of functional–basis set combinations were tested. On the whole, the NMR shifts of carbon atoms directly bonded to nickel can be calculated within the framework of the Kohn–Sham theory level using a number of hybrid functionals. The linear scaling procedure significantly improves results (root-mean-square error (RMSE) = 4.0–4.6 ppm). In the complexes coordinated with solvent and/or counterions, the predictions are less accurate (RMSE = 6.4 ppm). In the Ni–diimine alkyne complexes, calculations underestimate the NMR shielding (by 7–10 ppm) possibly due to the influence of high-spin states. At the same time, pure generalized gradient approximation (GGA) functionals show limits in their predictive power, particularly for the N-heterocyclic carbenes (NHCs) and π-donating carbons. Care has to be taken when interpreting 13C NMR spectra of Ni complexes using empirical dependencies, which may be misleading.

Which DFT factors influence the accuracy of 1H, 13C and 195Pt NMR chemical shift predictions in organopolymetallic square-planar complexes? New scaling parameters for homo- and hetero-multimetallic compounds and their direct applications

“…For evaluation purposes, the TPSSTPSS functionals, the def2-TZVPP basis set, the IEFPCM solvent model, the ultrafine integration grid and the GIAO method were adopted as ''fixed parameters'' when the other factors were studied. The mean absolute error (MAE) is the chosen parameter in order to evaluate the general fitting goodness between the computed and experimental data sets: The experimental chemical shifts were recovered from the literature, below the references for each compound: [1,2,3,4,5,6,7,16,17,18,19,20,21,22, Pt-Me, Pt-iso, Pt-u], 31 [8,9,10], 85 [11,12] 86 [13,14,15,24,25], 87 [23], 88 [a, b]. 83 For practical reasons, when signals were identified as overcrowded unresolved resonances, the average value was taken into consideration.…”

“…20 Although in silico , theoretical specialists have focused their efforts on NMR simulations of molecules cointaining atoms belonging to the first three rows of the periodic table, 21 theoretical predictions of chemical shifts of NMR active nuclei in organometallic compounds are still in their infancy. 22,23…”

“…20 Although in silico, theoretical specialists have focused their efforts on NMR simulations of molecules cointaining atoms belonging to the first three rows of the periodic table, 21 theoretical predictions of chemical shifts of NMR active nuclei in organometallic compounds are still in their infancy. 22,23 Nowadays, the most reliable quantum mechanical algorithms for chemical shift simulations are based on linear regression analysis. This method involves the use of a series of reference compounds (probe set) to reduce, or in the better case cancel systematic errors and methodological limitations.…”

Nuclear Magnetic Resonance

“…41 Absolute NMR shielding scales in methyl halides were derived from experimental and calculated nuclear spin-rotation constants. 42 Latypov et al 43 reported on accurate prediction of 13 C nuclear shieldings for atoms directly coordinated to diamagnetic nickel ion. Rusakova and Rusakov 44 analyzed 77 Se and 125 Te NMR structure parameters as well as their structural applications by quantum mechanical calculations.…”

“…116 DFT calculations as support of NMR experimental data were also used for structural characterization of phosphate species adsorbed on g-alumina. 117 Among rarely used NMR nuclei is 43 Ca. On the other hand, calcium ion is widely distributed in construction materials, as well as in biomaterials.…”