# A Quest for Effective 19F NMR Spectra Modeling: What Brings a Good Balance Between Accuracy and Computational Cost in Fluorine Chemical Shift Calculations?

## TL;DR

This paper investigates the best computational methods for modeling 19F NMR spectra, focusing on balancing accuracy and cost.

## Contribution

The study introduces and evaluates locally dense basis sets for fluorine chemical shift calculations.

## Key findings

- The pcS-3/pcS-2 LDBS scheme is the most balanced for fluorine chemical shift calculations.
- Different basis sets and functionals were compared against CCSD as a reference method.
- Basis set family, size, and valence-splitting levels significantly affect calculation accuracy.

## Abstract

This work proposes a systematic study of different computational schemes for fluorine Nuclear Magnetic Resonance (19F NMR) chemical shifts, with special emphasis placed on the basis set issue. This study encompasses two stages of calculation, namely, the development of the computational schemes for the geometry optimization of fluorine compounds and the NMR chemical shift calculations. In both stages, the performance of different density functional theory functionals is considered against the method of coupled-cluster singles and doubles (CCSD), with the latter representing a theoretical reference in this work. This exchange-correlation functional study is accompanied with a basis set study in both stages of calculation. Basis sets of different families, sizes, and valence-splitting levels are considered. Various locally dense basis sets (LDBSs) are proposed for the calculation of 19F NMR chemical shifts, and their performance is assessed by comparison of the calculated chemical shifts with both theoretical and experimental reference data. Overall, the pcS-3/pcS-2 LDBS scheme is recommended as the most balanced locally dense basis set scheme for fluorine chemical shift calculations.