Quantum-chemical study of 7Li NMR shifts in the context of delithiation of paramagnetic lithium vanadium phosphate, Li3V2(PO4)3 (LVP)

Abstract

The lithium NMR shifts of three paramagnetic materials important in the charging/discharging processes of lithium vanadium phosphate cathode materials have been studied by large-scale quantum-chemical methodology. Namely, the 7Li NMR shifts of the fully lithiated Li3V2(PO4)3 (LVP3.0), and of the partly delithiated Li2.5V2(PO4)3 (LVP2.5) and Li2V2(PO4)3 (LVP2.0), have been computed and analyzed using a recently proposed approach (A. Mondal, M. Kaupp J. Phys. Chem. C 123 (2019) 8387-8405) that accounts for the Fermi-contact, pseudo-contact, as well as orbital shifts, combining periodic computations with an incremental cluster model. LVP3.0 and LVP2.0 exhibit three and two unique Li sites, respectively, which could be assigned to their experimental 7Li NMR signals. In case of LVP2.0, the computations clearly assigned the signals at 143 ppm and 77 ppm to the Li(1) and Li(2) sites, respectively, even though the latter is connected to Vb(+III d2 sites and the former to Va(+IV) d1 sites. LVP2.5 is the most complex of these three materials, exhibiting a 50% occupation of Li(3) sites, which generates much more complicated Li NMR spectra with seven peaks that partly are closely spaced. Exploring three different occupation patterns, the computations can clearly assign five of the seven signals to one type of Li site and give most probable assignments for the two remaining signals. Notably, the calculations support seven signals to be assigned to LVP2.5, while previous interpretations took two of the signals as being entirely due to contamination by LVP2.0. The accuracy of the computations could probably be improved further by full DFT optimization of large super-cell structures. This work suggests that first-principles computations of NMR shifts of extended paramagnetic solids provide an important tool for the analysis of even rather complex NMR spectra.