The material at the heart of the work, Li7PS6, is a lithium argyrodite that exists in two crystalline forms: a high-temperature phase and a low-temperature phase. The high-temperature version is traditionally produced by firing lithium sulfide and phosphorus pentasulfide together at 600 to 700 degrees Celsius, and it conducts reasonably well once hot. The low-temperature phase, by contrast, has historically been a poor performer, with reported ionic conductivity of roughly 1.6 × 10⁻⁶ S·cm⁻¹ at 40 degrees Celsius when made by conventional solid-state reaction. That figure places it far below the threshold needed for practical battery operation, which is why most researchers have either worked at high temperatures or turned to other argyrodite chemistries such as Li6PS5Cl.

The Vietnamese team, led by corresponding author Nguyen Huu Huy Phuc, approached the problem from two directions at once. First, they abandoned the furnace entirely in favor of liquid-phase synthesis. A suspension of Li3PS4 was prepared by stirring lithium sulfide and phosphorus pentasulfide in tetrahydrofuran at 50 degrees Celsius for 24 hours. Adding more lithium sulfide dissolved in ethanol then triggered an immediate reaction that produced a yellowish solution, which was evaporated at about 90 degrees Celsius under reduced pressure and heat-treated at only 200 degrees Celsius for one hour. The result was phase-pure low-temperature Li7PS6, and remarkably, its ionic conductivity at 25 degrees Celsius measured 2.3 × 10⁻⁵ S·cm⁻¹, roughly ten times higher than the same phase prepared by solid-state reaction.

The second lever was chemical substitution. The researchers first ball-milled lithium sulfide and aluminum sulfide into a Li3AlS3 glass ceramic, then introduced controlled amounts of this dopant precursor into the ethanolic reaction mixture. The target composition, Li7−4xAlxP1−xS6−3x, embodies a simple charge-balance logic: each trivalent Al³⁺ ion that replaces a pentavalent P⁵⁺ ion in the crystal lattice must be accompanied by the removal of four lithium ions and three sulfide ions to keep the structure electrically neutral. Four compositions were prepared, spanning x values of 0, 0.1, 0.2 and 0.3, and all were handled in an argon-filled glove box with oxygen and moisture levels held below 1 ppm and 0.1 ppm respectively, reflecting the acute sensitivity of sulfide electrolytes to air.

Proving that the aluminum actually entered the crystal structure, rather than sitting in an inert second phase, required a battery of characterization techniques. X-ray diffraction patterns for the samples with x equal to 0.1 and 0.2 showed the signature of low-temperature Li7PS6 with no trace of the Li5AlS4 crystal that would betray undissolved dopant. Rietveld refinement against the structural model of the related argyrodite Li7Zn0.5SiS6 confirmed a tetragonal lattice in the I4/m space group, and revealed that the a and b lattice axes and the unit-cell volume grew steadily with aluminum content while the c axis stayed nearly constant. Scanning electron microscopy with energy-dispersive spectroscopy showed the electrolyte forming nanoparticles tens of nanometres across that agglomerate into secondary particles tens of micrometres wide, with aluminum uniformly dispersed throughout. Raman spectroscopy added further confirmation: the PS4³⁻ vibrational peaks shifted slightly to lower wavenumbers in the doped samples, a shift previously observed when P⁵⁺ is replaced by lower-valent cations in related argyrodites.

The doping story had a clear limit, however. At x equal to 0.3, the diffraction patterns revealed a mixture of both the low-temperature and high-temperature phases, and the Raman spectrum showed the isolated Li–S and Al–S peaks of unreacted Li3AlS3. Together these observations mark the solubility ceiling of aluminum in the low-temperature argyrodite lattice, and they explain why the highest doping level did not produce the best conductor. Crystalline content, calculated from the ratio of refined peak area to total diffraction profile area, ranged from 57 to 71 percent across the series, but the team found that conductivity did not track crystallinity, indicating that the crystal structure itself, rather than the amorphous fraction, governs lithium transport.

The electrical measurements delivered the headline result. Alternating-current impedance spectroscopy, performed on pellets pressed at 570 MPa with carbon paper blocking electrodes, gave room-temperature ionic conductivities of 2.3 × 10⁻⁵, 1.2 × 10⁻⁴, 7.4 × 10⁻⁵ and 3.2 × 10⁻⁵ S·cm⁻¹ for x equal to 0, 0.1, 0.2 and 0.3 respectively. The sweet spot at x equal to 0.1 represents a fivefold improvement over the undoped liquid-phase material and roughly fifty times the conductivity of solid-state-prepared low-temperature Li7PS6. The activation energy for conduction fell from 0.526 eV in the undoped sample to 0.368 eV at x equal to 0.1, the lowest of the series, before rising again at higher doping levels where parasitic phases appeared. Direct-current polarization confirmed that the material is a genuine ionic conductor: its electronic conductivity, measured at 1.8 × 10⁻⁸ S·cm⁻¹, is nearly four orders of magnitude below its ionic conductivity, a prerequisite for suppressing self-discharge in a real cell.

Perhaps the most scientifically valuable part of the study is the dissection of why the doped material conducts better. Using the Jonscher universal dielectric response to extract the lithium hopping frequency from the impedance data, the team separated total conductivity into two factors: how easily lithium ions jump between sites, and how many mobile charge carriers exist in the first place. The activation energy for hopping migration plummeted from 0.479 eV in undoped Li7PS6 to just 0.122 eV at x equal to 0.1, while the activation energy for mobile-ion formation rose from 0.223 to 0.321 eV. In other words, aluminum substitution makes each individual lithium jump dramatically easier, at the modest cost of slightly reducing the population of mobile carriers, and the mobility gain wins decisively. Dielectric loss analysis told the same story from a different angle: the migration energy extracted from the temperature dependence of the loss peaks was smallest for x equal to 0.1, and the characteristic migration time of the doped samples lengthened by about an order of magnitude relative to the undoped material, signaling a shift from short-range to long-range lithium diffusion.

Practical stability was not neglected. A symmetric lithium cell built with the x equal to 0.1 composition, cycled at a current density of 0.1 mA·cm⁻² with one-hour charge and discharge steps, ran for more than 100 cycles without any severe increase in overvoltage, indicating that the aluminum-doped electrolyte is reasonably stable in direct contact with lithium metal. Taken together, the results sketch a compelling recipe for next-generation solid electrolytes: a low-temperature, solution-based synthesis that avoids energy-hungry firing and the phase segregation it causes, combined with a precisely tuned aliovalent substitution that flattens the energy landscape for migrating lithium ions. If the same strategy transfers to other argyrodite systems, the manufacturing bottleneck that has kept solid-state batteries confined to laboratories may become considerably easier to break.

Subject of Research: Aluminum-doped argyrodite-type Li7PS6 solid electrolytes synthesized by wet-chemical routes and their lithium-ion conduction dynamics

Article Title: Li-ion dynamics and relaxation in wet-chemical synthesized argyrodite-type Li7−4xAlxP1−xS6−3x solid electrolyte

Article References: Quang, V. A., Toan, T. V., Tu, T. A., Ty, N. M., Anh, L. T., Anh, L. T. Q., & Phuc, N. H. H. (2026). Li-ion dynamics and relaxation in wet-chemical synthesized argyrodite-type Li7−4xAlxP1−xS6−3x solid electrolyte. Discover Electrochemistry, 3(1), Article 32. https://doi.org/10.1007/s44373-026-00120-7

Image Credits: AI Generated

DOI: 10.1007/s44373-026-00120-7

Keywords: solid-state batteries, argyrodite, Li7PS6, solid electrolyte, ionic conductivity, aluminum doping, liquid-phase synthesis, lithium-ion hopping, impedance spectroscopy, activation energy, sulfide electrolyte, Rietveld refinement