Phosphate Cosolvent Strategy Tames Water for Long-Lasting Aluminum Batteries

Aqueous aluminum metal batteries have long been one of the most tantalizing promises in post-lithium energy storage. Aluminum metal offers a theoretical capacity of 2980 mAh per gram and an extraordinary 8046 mAh per cubic centimeter, and the element makes up 8.2 percent of the Earth’s crust, making it abundant, cheap, and geographically secure. Yet the same water that makes these batteries safe and inexpensive has also been their undoing. A new study published in Advanced Science now reports a carefully balanced electrolyte design that confronts the problem on two fronts at once, and the results suggest that practical aqueous aluminum batteries may be closer than many assumed.

The core difficulty is chemical. Aluminum ions in water are strongly hydrated, and the water molecules locked into the Al3+ solvation shell are actually easier to electrochemically reduce than free water. During charging, this triggers vigorous hydrogen evolution, which consumes active aluminum, destabilizes the electrode interface, and wrecks cycling stability. Meanwhile, the bulk water in the electrolyte maintains its high reactivity through an extended, interconnected hydrogen bond network, which narrows the electrochemical stability window of the entire system. Corrosion and surface passivation compound the damage, rapidly thinning the aluminum anode until the cell fails.

Previous attempts to fix this usually tweaked one lever at a time, often through trial-and-error screening of additives. The research team behind the new work, led by scientists supported by Singapore’s A*STAR program, argues that this approach conflates two physically distinct phenomena: the chemistry of the bulk solution and the dynamics of the Al3+ cation at the electrode surface. Suppressing water activity in the bulk helps, but excessive disruption of the solvent matrix degrades ionic conductivity and viscosity. Similarly, modifying the cation’s solvation shell can curb interfacial reactions, but too much modification imposes a heavy desolvation penalty that slows the kinetics of aluminum plating and stripping. The team set out to evaluate these two effects independently and then find the composition where both are simultaneously satisfied.

As model cosolvents, the researchers examined four trialkyl phosphates: trimethyl phosphate (TMP), triethyl phosphate (TEP), tripropyl phosphate (TPP), and tributyl phosphate (TBP). These organophosphorus molecules are inherently non-flammable and feature a strongly electron-rich phosphoryl group that acts as a powerful hydrogen bond acceptor. Density functional theory calculations showed that Al3+ binds more strongly to all four phosphates than to water, with binding energy increasing sublinearly as the alkoxy chains lengthen. Electrolytes were prepared by mixing each cosolvent at volume fractions from 20 to 80 percent into a baseline 1 M aluminum triflate solution, and twelve viable formulations were tested in symmetric aluminum-aluminum cells.

The screening revealed a clear trade-off. Ionic conductivity fell steadily as cosolvent content rose, because every phosphate has a lower dielectric constant and higher viscosity than water; at equal volume fractions, conductivity tracked the dielectric constants, ranking TMP above TEP, TPP, and TBP. At the other extreme, cells with 80 percent cosolvent ran for more than 140 hours but suffered overpotentials above 1.1 volts, a sign of sluggish interfacial kinetics. The sweet spot was 40TMP: a moderate overpotential below 0.3 volts combined with cycling stability beyond 200 hours, and an ionic conductivity of 24.6 mS per centimeter, roughly half the baseline value but still far above other reported organic-cosolvent aluminum electrolytes.

Spectroscopy explained why. Fourier-transform infrared spectra showed the water O-H stretching and bending modes shifting to higher wavenumbers as TMP content increased, evidence that TMP’s phosphoryl oxygen outcompetes water for hydrogen bonding and breaks up the cooperative water-water network. Proton nuclear magnetic resonance displayed a two-stage evolution: an initial downfield shift from strong P=O···H-O interactions up to 40TMP, followed by upfield broadening at higher concentrations as the native hydrogen bond network collapses. Differential scanning calorimetry confirmed the disruption, with freezing points depressed by 5.7 degrees Celsius at 60TMP and no detectable freezing peak at 80TMP. Corrosion current density on aluminum electrodes dropped from 11.5 microamperes per square centimeter in the baseline to 3.1 at 80TMP, and the hydrogen evolution onset shifted negatively by 0.51 volts at the highest TMP loading.

Molecular dynamics simulations and 27Al NMR then dissected the solvation shell itself. In the baseline electrolyte, Al3+ carries an average of 5.42 water molecules in its primary shell. In 40TMP, that number falls to 4.82, with roughly half a TMP molecule entering the inner shell, while anion coordination stays essentially constant. Because coordinated water is more easily reduced than free water, removing even one molecule from the shell meaningfully suppresses interfacial hydrogen evolution. But the simulations also exposed the danger of overcorrection: at 80TMP, 2.37 TMP molecules coordinate to each Al3+, and the strong Al3+-TMP binding raises the desolvation barrier, choking deposition kinetics. The 40TMP composition threads the needle between these competing constraints.

One of the most striking findings is that the optimized electrolyte enables the growth of a genuine solid-electrolyte interphase even in a dilute aqueous system. Cross-sectional electron microscopy and X-ray photoelectron spectroscopy of cycled aluminum foils revealed a chemically graded layer: an organic-rich outer region derived from triflate decomposition products, and an inorganic-rich inner region composed mainly of aluminum fluoride and aluminum sulfide. Time-of-flight secondary ion mass spectrometry mapped these components in three dimensions, confirming the hybrid architecture. Notably, no phosphorus was detected in the interphase, meaning TMP does not directly contribute to SEI chemistry. Instead, by suppressing hydrogen evolution, it prevents gas bubbles from physically disrupting the nascent interphase, allowing a robust, continuous film to form where the baseline electrolyte produces only faint, fragmentary signals.

The electrochemical payoff was substantial. Differential electrochemical mass spectrometry showed dramatically less hydrogen generation during plating and stripping, and symmetric cells ran stably for over 500 hours at 0.1 mAh per square centimeter, compared with failure within 120 hours for the baseline. Full cells pairing the electrolyte with a polyaniline cathode, chosen because it stores charge through nitrogen-centered redox rather than proton-coupled intercalation, delivered an average Coulombic efficiency of 99.0 percent and retained 59 percent of capacity after 250 cycles, whereas baseline cells collapsed within 33 cycles. Adding a tin interfacial layer on the aluminum anode, formed by a simple displacement reaction and acting as underpotential-deposition nucleation sites, pushed performance further: the Sn@Al|40TMP|polyaniline cell retained 70 percent of its capacity after 400 cycles and kept more than 70 percent of its capacity even at 1 A per gram. Combustion tests also showed the 40TMP-wetted separator surviving 30 seconds of direct flame exposure, thanks to TMP’s radical-tracking flame-retardant behavior, at a cost premium of only about 7 percent over the baseline formulation.

Beyond the immediate performance numbers, the study’s real contribution may be conceptual. By explicitly separating bulk solution effects from cation solvation dynamics and then optimizing both together, the researchers provide a transferable design framework rather than a one-off recipe. The work demonstrates that the path to stable aqueous aluminum batteries does not run through maximum suppression of water reactivity, but through a precise balance point where water is tamed enough to stop parasitic reactions, yet the aluminum ion can still shed its solvation shell and move quickly across the interface. If that principle holds across other multivalent chemistries, the humble phosphate molecule may have just moved the entire field a significant step closer to batteries that are safe, cheap, and built to last.

Subject of Research: Electrolyte design for stable aqueous aluminum metal batteries

Article Title: Balancing Bulk Solution Chemistry and Cation Solvation for Stable Aqueous Aluminum Batteries

Article References: Jia, B.-E., Wu, G., Gu, H., Liew, J. J., Chen, D., Song, J., Wang, D.-Y., Hu, E., Choo, H. H., Liu, Y., Hu, Y., Song, J., Tang, A., Xing, Z., Zhu, Q., Yan, C., Lv, C., Ng, M.-F., & Yan, Q. (2026). Balancing Bulk Solution Chemistry and Cation Solvation for Stable Aqueous Aluminum Batteries. Advanced Science, 13(55), Article e76540. https://doi.org/10.1002/advs.76540

Image Credits: AI Generated

DOI: 10.1002/advs.76540

Keywords: aqueous aluminum batteries, electrolyte design, trimethyl phosphate, solvation structure, hydrogen evolution, solid-electrolyte interphase, aluminum metal anode, post-lithium energy storage, hydrogen bond network, polyaniline cathode, tin interfacial layer, corrosion suppression