LEMAF research is the cover of Journal of Chemical Physics issue

New LEMAF publication is the cover of the Journal of Chemical Physics

Study from the LEMAF group got the cover image in the volume 155, issue 7 of August 2021 of the Journal of Chemical Physics.

As physicists, materials scientists, and engineers continue striving to enhance and improve batteries and other energy storage technologies, a key focus is on finding or designing new ways to make electrodes and electrolytes.  One promising avenue of research involves solid-state materials, making possible batteries free of liquid electrolytes, which can pose fire and corrosion hazards.  An international group of researchers joined with scientists at the Argonne National Laboratory to investigate the structure of crystalline and amorphous compounds based on the NASICON system.  The work was published in the Journal of Chemical Physics.

Because of their high ionic conductivity, materials with a NASICON structure are prime candidates for a solid electrolyte in sodium-ion batteries.  They can be prepared in various forms, including glass, which gives them the usefulness of moldable bulk materials.  In this work, the research team specifically studied the NAGP system [Na_1+xAl_xGe_2-x(PO₄)₃] with x = 0, 0.4 and 0.8 in both crystalline and glassy forms.  Working at several different facilities, they used a combination of techniques, including neutron and X-ray diffraction, along with ²⁷Al and ³¹P magic angle spinning and ³¹P/²³Na double-resonance nuclear magnetic resonance spectroscopy.  The glassy form of NAGP materials was examined both in its as-prepared state and after thermal annealing, so that the changes on nucleation could be studied.

Neutron powder diffraction measurements were performed at the BER II reactor source, Helmholtz-Zentrum Berlin (HZB), Germany, using the fine resolution powder diffractometer E9 (FIREPOD), followed by Rietveld analysis.  Further neutron diffraction observations were conducted at the Institut Laue-Langevin using the D4c diffractometer and at the ISIS pulsed neutron source using the GEM diffractometer.  X-ray diffraction studies were performed at beamline 6-ID-D of the Advanced Photon Source. 

The NAGP x = 0 and x = 0.4 compounds are classified as space group R while the x = 0.8 compound is space group Rc.  The x = 0 phase shows tetrahedral PO₄ motifs linked by bridging oxygen (BO) atoms to four octahedral GeO₆ motifs.  This permits Na⁺ ions to reside at the interstices, allowing ionic conductivity.  In the glassy NAGP phases, the formation of sub-octahedral Ge and Al-centered units leads to non-bridging oxygen (NBO) atoms.  Upon annealing, the fraction of NBO units decreases as the Ge and Al coordination numbers increase.  Again, the ionic conductivity increases with the concentration of Na⁺ ions in the glassy NAGP material. 

Based on the Ren and Eckert model for vitreous sodium phosphosilicate, the researchers propose a model for the x = 0 glass in which superstructural units are formed.  In these units, P⁽³⁾ phosphate motifs with 3 bridging oxygen (BO) and one NBO atoms are converted to P⁽⁴⁾ phosphate motifs with 4BO atoms, thereby converting GeO₄ to GeO₆ motifs and increasing the size of the superstructural units.  Annealing the as-prepared glass results in an increase to the Ge coordination number and the fraction of P⁽⁴⁾ motifs, which provide the nucleation sites for crystal growth.

The work reveals some substantial differences between the crystalline and glass phases of NAGP, which affect the ionic conductivity of the various materials.  The investigators note that the fraction of NBO atoms appears to play a significant role, possibly altering the Na⁺ ion mobility, and suggest this as an area of further study.  Additional investigations of the NAGP structure promise valuable insights not only into its own complex characteristics but also those of the NASICON system in general — a necessary step in the realization of the practical application of these materials.

From our NMR viewpoint this study serves as an example of how one can select a suitable deconvolution model for a poorly resolved lineshape (in this case ³¹P MAS-NMR) based on REAPDOR data, to arrive at a structural model consistent with diffraction data. Furthermore, the NMR data serve to monitor the nucleation process, involving the successive conversion of P⁽³⁾ to P⁽⁴⁾ units, resulting in the disappearance of non-bridging oxygen atoms.