On the Efficiency of the Density Functional Theory (DFT)-Based Computational Protocol for 1H and 13C Nuclear Magnetic Resonance (NMR) Chemical Shifts of Natural Products: Studying the Accuracy of the pecS-n (n = 1, 2) Basis Sets

Abstract

The basis set issue has always been one of the most important factors of accuracy in the quantum chemical calculations of NMR chemical shifts. In a previous paper, we developed new pecS-n (n = 1, 2) basis sets purposed for the calculations of the NMR chemical shifts of the nuclei of the most popular NMR-active isotopes of 1–2 row elements and successfully approbated these on the DFT calculations of chemical shifts in a limited series of small molecules. In this paper, we demonstrate the performance of the pecS-n (n = 1, 2) basis sets on the calculations of as much as 713 1H and 767 13C chemical shifts of 23 biologically active natural products with complicated stereochemical structures, carried out using the GIAO-DFT(PBE0) approach. We also proposed new alternative contraction schemes for our basis sets characterized by less contraction depth of the p-shell. New contraction coefficients have been optimized with the property-energy consistent (PEC) method. The accuracies of the pecS-n (n = 1, 2) basis sets of both the original and newly contracted forms were assessed on massive benchmark calculations of proton and carbon chemical shifts of a vast variety of natural products. It was found that less contracted pecS-n (n = 1, 2) basis sets provide no noticeable improvement in accuracy. These calculations represent the most austere test of our basis sets as applied to routine calculations of the NMR chemical shifts of real-life compounds.

Getaway from the Geometry Factor Error in the Molecular Property Calculations: Efficient pecG-n (n = 1, 2) Basis Sets for the Geometry Optimization of Molecules Containing Light p Elements

“…0.46/0.25 ppm for 1 H, 4.47/2.09 ppm for 13 C, 7.0/3.28 ppm for 15 N, 9.85/6.92 ppm for 31 P, 4.31/2.08 ppm for 19 F, and 4.67/3.13 ppm for 29 Si, calculated on the CCSD or (/) DFT(B97-2) equilibrium geometries, respectively. These are very considerable numbers, given that the best contemporary accuracy of the NMR chemical shift calculations can be characterized by absolute deviations from the experiment of about 0.3–0.5 ppm for 1 H, − 1–2 ppm for 13 C, − 5–6 ppm for 15 N, , 4–5 ppm for 31 P, ,, 1–5 ppm for 19 F, , and 5–7 ppm for 29 Si. , …”