Kyle Proffitt
December 1, 2025 | Researchers from Stanford—along with several additional international facilities and institutions—reported (doi: 10.1038/s41563-025-02356-x) in Nature Materials in October the discovery of an entirely new class of high-valency iron-based lithium intercalation cathode that could translate to lower-cost, high-energy-density batteries. “We’re pushing iron to a new oxidation state and taking it back in a way that has not been demonstrated before,” said co-first author Edward Mu, a PhD candidate in the lab of William Chueh. Battery Power Online spoke with Mu to learn how this new regime of iron redox works inside an intercalation cathode battery.
Iron Redox
Iron participates in redox chemistry readily, both in battery materials like LFP and in biological processes, such as mitochondrial electron transport that couples the electron transfer to oxygen with ATP creation. Those processes involve interconversion between Fe2+ and Fe3+ forms, one electron at a time. LFP is poised to achieve international dominance as a cathode material, despite its lower energy density relative to NMC, and that’s driven by the wide availability of iron and associated low cost. Iron is the second most abundant metal in the earth’s crust, the fourth most abundant element overall, and has global distribution. Making iron-based cathode materials that compete on energy density grounds with NMC would be a major achievement, and this is where battery voltage, a function of iron’s oxidation state, comes into play. If higher iron oxidation states can be reliably harnessed, the voltage increases, with corresponding energy density improvement—the best of both worlds.
Old Material, Prepared a New Way
In pursuit of this idea, Mu and colleagues started with Li4FeSbO6 (LFSO), a known material. It was previously tested as a cathode, and Fe4+ creation was identified during charging, but the evolution of oxygen species also occurred, limiting material stability and utility.
Mu likened their revisit of LFSO to the story of early-stage LFP. LFP started as a very low energy density material, but “the initial discoveries that led to its commercialization and success had to do with making the particle size smaller and coating the material with carbon.” In the prior report testing LFSO as a cathode, ball milling was used to create finer LFSO particles—a step that is necessary to confer conductivity to the as-synthesized LFSO—and Mu and colleagues were stimulated to pursue new synthetic methods to nanosize LFSO that might allow higher oxidation states while maintaining stability. They identified a sol-gel synthesis route that “produces these nano-sized particles basically out of the furnace.” As it turns out, this choice of synthetic route makes a big difference in material properties and associated battery functionality.
Sol-Gel LFSO in Batteries
Coin cell batteries were prepared using sol-gel LFSO, and charge/discharge experiments were performed. The cells charged and discharged while showing a capacity of ~170 mAh/g, but notably, voltage hysteresis was pretty small. Voltage hysteresis is seen as separation between the charging voltage curve and the discharging voltage curve (a battery with no hysteresis would show overlapping curves). “What’s common for these iron materials is more than one volt of hysteresis… we have roughly 200 millivolts,” Mu said. That’s important because the hysteresis represents thermodynamic energy losses that limit usable power for your devices, and low hysteresis indicates material stability.
Under the Hood
What’s really interesting about the new work is that charging the battery pushes iron oxidation to a level, FeV, never before seen in a cathode. A thorough set of spectroscopic experiments was performed to validate this finding, including X-ray diffraction, 57Fe Mössbauer, neutron diffraction, and more. Those experiments provide detailed information about the electronic structure of iron and reliably confirm its formal charge of 3 in the initial, lithiated LFSO, and 5 after charging. The experiments also show that the environment is stable; LFSO experiences an “ordered crystallographic phase transition” during charging, but the original structure is regained after discharge. The iron goes directly from FeIII to FeV and back, skipping over the Fe4+ reported before with milled LFSO.
Correspondingly, the higher oxidation state means higher voltage. “This new redox couple between the FeIII and FeV oxidation states is a very useful redox couple because it’s at high voltage,” Mu explained. LFP caps out at about 3.5 V, whereas the batteries made with this carefully synthesized LFSO operate up to about 4.2 V. “It’s a significant increase in voltage, and that leads to a significant increase in energy density,” he added.
Why the Synthetic Route Matters
The spectroscopy experiments also illuminated how the synthesis method influences high-valency Fe support. Both solid-state synthesized and sol-gel synthesized LFSO adopt a sort of honeycomb structure, with a regular alternating pattern of Fe and Sb, in coordination with oxygen and lithium. Again though, the solid-state synthesized LFSO will not work in a battery without first milling the material, and the spectroscopic experiments in this new report reveal that milling disrupts the honeycomb. Atoms swap places, the electronic environment of iron changes, and oxygen atoms are able to interact and form dimers. “We need the iron atoms to be separate from each other in order for us to access this unique oxidation state; if we break that order, then we actually reduce the amount of the III/V capacity,” Mu explained.
Theoretical Meets Experimental
It may sound straightforward to test the electronic environment of the atoms in LFSO, but Mu credited the many institutions and specialized equipment involved, including the SLAC National Accelerator Laboratory, the Stanford Synchrotron Radiation Lightsource (SSRL), the Advanced Light Source at Lawrence Berkeley National Lab, Argonne, and Oak Ridge National Lab.
He also highlighted theoretical work as a major driver of the project’s success. A collaboration was established with Tom Devereaux, also at Stanford, and his graduate student, Eder Lomeli, who shares first authorship (with one additional co-first author, Hari Ramachandran, now a postdoc in Chueh’s lab at Stanford). “Before coming to them, we could only do comparisons with reference compounds, but we couldn’t really understand the nature of the FeV oxidation state without their key input,” Mu said. Close collaboration between experiment and theory was necessary to understand some of the very complicated spectra. “The study would not be complete without the theoretical validation and understanding this really unusual oxidation state,” Mu added.
Stabilized Holes
There are some more pretty interesting things happening at the atomic level. Fe3+, the state of iron in LFSO after synthesis, has its five outermost electrons in the 3d orbitals. When a battery is charged using LFSO as the cathode, two lithium ions migrate through the separator to the anode, and two electrons migrate via the external circuit. The loss of those electrons, if centered on the oxidation of Fe, would create an Fe5+ state, whereupon one might expect a 3d3 configuration. However, the spectroscopic experiments Mu and colleagues performed reveal that the Fe maintains a more 3d5-like arrangement.
Mu explains better. “When you pull those electrons away … you’re not taking the electrons solely from iron. You’re oxidizing—you’re taking electrons away from—the iron-oxygen bonds”. The explanation goes just a little bit deeper into the realm of physics. “When we charge the battery up, we pull out electrons and we create holes,” Mu explained. Electron holes, or just holes, can be thought of as the place where an electron ought to be within an atom or atomic lattice. “Where these holes like to live, that was kind of the big question here … it turns out they’re primarily centered on oxygen,” Mu said. Because oxygen is a ligand, L, in this cathode, their official notation for iron’s outer shell arrangement in delithiated LFSO is 3d5L2, reflecting two holes centered on oxygen that enable iron to maintain a more stable 3d5 arrangement. This understanding is also why the researchers favor using the formal charge, FeV, over the ionic charge, Fe5+, since the iron retains more of this electron density.
Antimony
Antimony has an important role in the structure. Provided LFSO is nanosized gently, the regular honeycomb structure is retained, and that structure supports hole stability. The report states that, “The added hole density on oxygen enables the high-voltage redox couple, and the key challenge … [is] providing a stable and reversible pathway for the addition and removal of these ligand holes.” The lattice oxygen stabilizes holes through hybridization with iron, and that depends on oxygen’s local coordination environment. The highly charged antimony, Sb5+, forms highly ionic Sb-O bonds, but it does not participate in the redox chemistry. Instead, it acts kind of like a spectator, and its key role is electrostatic; “Sb5+ templates the iron atoms into the ‘honeycomb’ ordering,” and “this prevents the iron atoms from being next to each other, which we show is bad for performance and causes the O-O dimerization behavior,” Mu said.
Antimony unfortunately does not share the same abundance and low cost of iron; it’s more comparable to the contentious cobalt present in NMC cathodes. According to Mu, antimony is less abundant than cobalt, but it is actually cheaper. “It’s not a terrible battery material, but because of its low abundance, we’d like to be able to do it with elements that are more available,” he added.
Next Moves
“The issue that I’m working on right now is understanding and fixing the cycling issues,” Mu said. Despite structural stability in a single charge/discharge cycle, “the capacity fades quite quickly as we continue to cycle the material.” The other big question they’re trying to address is how to get away from antimony. “We’re still thinking about ways to replace antimony, that’s an active research effort.” In one more avenue of pursuit, Mu said that “there’s basically a sodium-ion variant of this material, so we’re thinking about characterizing that too.”
