Li7P3S11 glass ceramics have high conductivities competitive with liquid electrolytes, making them good candidates as solid-state electrolytes for all-solid-state lithium-ion batteries. However, the metastable nature and performance of Li7P3S11 glass ceramics remain mysterious. Herein, modified Li7P3S11 glass ceramics with compositions of 70Li2S-30P2S5 were prepared via two-step mechanical milling and thermal annealing. Li7P3S11 glass ceramics synthesized using the conventional method (mechanical milling and thermal annealing) were again ball-milled to obtain amorphous 70Li2S-30P2S5 with a peculiar glass structure. Further thermal annealing was carried out to crystallize the glass. The obtained crystalline phase was analogous to the original Li7P3S11 phase, but the conductivity was enhanced by a factor of 1.7. Based on 31P solid-state nuclear magnetic resonance (NMR) spectroscopy, the Li7P3S11 phase contained an additional PS43- unit. A rational deconvolution procedure for the 31P solid-state NMR spectra based on crystalline Li7P3S11 was developed and applied to the samples. The analysis can resolve the additional crystalline PS43- unit in the Li7P3S11 structure. Based on two-dimensional double-quantum 31P NMR spectroscopy, the additional PS43- unit is located adjacent to the P2S74- unit, suggesting that P2S74- is divided into two PS43- units in the Li7P3S11 phase. The flip motion of Li+ was also investigated based on the 7Li spin-lattice relaxation time. The independent activation energy of spin-lattice relaxation with respect to temperature in the Li7P3S11 phase was attributed to a conduction path between the two PS43- units. The findings provide a synthetic route that can be used to develop metastable solid-state electrolytes.
Keywords: conduction path; ion conduction; lithium thiophosphate; mechanical milling; metastable crystal; solid-state NMR; solid-state electrolyte.