The basic architecture of a liquid metal battery is deceptively simple. Two liquid metal electrodes of different densities sit in a container, one above the other, with a molten salt electrolyte sandwiched between them because it is denser than the light negative electrode but less dense than the heavy positive electrode. Because everything is liquid, there are no fragile solid microstructures to crack or degrade, which is why these cells can tolerate tens of thousands of charge and discharge cycles. The concept traces back to fused-salt galvanic cells studied at Argonne National Laboratory in the 1960s, but it was the 2013 landmark review by Donald Sadoway’s group at MIT and the 2014 demonstration of a lithium-antimony-lead cell in Nature that transformed the idea into a serious contender for stationary storage. Sodium, with its abundance, low cost, and favorable electrochemistry, has emerged as the most attractive negative electrode metal for grid applications.

Yet sodium brings a fundamental problem that has dogged the technology for years. Molten sodium dissolves appreciably in conventional single-cation halide molten salts, such as the sodium chloride-based systems long used as electrolytes. Dissolved sodium, along with solvated electrons in the melt, can shuttle across the electrolyte to the positive electrode without passing through the external circuit, causing severe self-discharge and dragging Coulombic efficiency down to levels that would be unacceptable in a commercial product. The review highlights how researchers have attacked this problem by reformulating the electrolyte itself. Instead of relying on a single cation, teams have developed multi-cation molten salts that blend sodium with lithium, calcium, potassium, or other ions, fundamentally changing the chemistry of the melt and suppressing the dissolution of sodium metal.

The mechanism behind this improvement is one of the most technically interesting threads in the review. In a multi-cation electrolyte, the additional cations compete for space in the salt structure and alter the activity of sodium in the melt, lowering its solubility and thereby choking off the parasitic dissolution pathway that causes self-discharge. Multi-cation salts also tend to have lower melting points than their single-cation counterparts, which allows the entire battery to operate at reduced temperatures. That matters enormously for engineering: lower operating temperatures mean less thermal stress on seals and containers, slower corrosion of structural materials, and reduced parasitic energy consumption to keep the cell molten. Studies from groups in China and Germany, including work on multi-cationic electrolytes published in Energy Storage Materials and the Journal of Power Sources, have demonstrated sodium-based cells with markedly improved cycling stability and efficiency when these mixed-salt electrolytes are employed.

The positive electrode presents a different set of challenges. In a sodium-based liquid metal battery, the positive electrode is typically a molten alloy of antimony, bismuth, tin, or combinations of these metals, and during discharge, sodium ions arriving from the electrolyte alloy into this pool. The trouble is that sodium diffuses sluggishly through many of these alloy electrodes, which limits how much of the active material can actually be used and therefore caps the energy density of the cell. The review describes design principles for overcoming this bottleneck, including engineering liquid-phase diffusion pathways within the electrode and tuning the alloy composition to accelerate interfacial reaction kinetics. Ternary systems such as bismuth-antimony-tin have shown synergistic stabilization mechanisms, with bismuth contributing a self-healing action that helps the electrode recover its structure over repeated cycles, an approach the review’s lead author has explored in earlier work on lithium-based systems.

Interfacial behavior between the liquid electrodes, the molten salt, and the solid current collectors emerges as another decisive factor. Wetting, the tendency of a liquid to spread across a solid surface, governs how well the molten positive electrode makes electrical contact with its current collector, and poor wetting can raise internal resistance and destabilize cycling. Researchers have found that improving wettability at the positive electrode enhances the cycling stability of bismuth-based cells, and the review emphasizes that controlling these interfacial phenomena is as important as bulk chemistry. The review also covers the charge and discharge mechanisms themselves, noting that mass transport in the liquid phases creates asymmetries between charging and discharging behavior, since convection patterns driven by thermal gradients and electrical currents redistribute species differently depending on the direction of operation.

That last point opens onto one of the most fascinating aspects of liquid metal battery physics: the multiphysics coupling inside the cell. Because the electrodes and electrolyte are all molten, the battery is a fluid-dynamical system as much as an electrochemical one. Thermal convection of the Rayleigh-Bénard type, driven by temperature differences between the bottom and top of the cell, interacts with electro-vortex flows generated by the current entering and leaving the electrodes, and with solutal convection caused by composition gradients during alloying and dealloying. A comprehensive review of the fluid mechanics of sodium-zinc liquid metal batteries published in Applied Physics Reviews catalogued these coupled phenomena, and the new Ionics review integrates them into a thermal-flow-electrical-corrosion framework. How these flows behave determines whether the salt layer stays cleanly between the electrodes or whether metal droplets could short the cell, so understanding and regulating them is essential for scaling up.

Scaling up is precisely where the engineering challenges become daunting. A grid-storage module must be sealed reliably against oxygen and moisture ingress for decades while operating at several hundred degrees Celsius, and the review identifies sealing, thermal management, and corrosion as the central integration problems. Molten salts and molten metals are aggressive corrosive media, and current collectors in particular must survive prolonged contact with them; research on stainless steel and coated collectors for lithium-antimony-tin cells illustrates the materials science effort underway. Thermal management systems, studied numerically for liquid metal battery modules, must keep the entire cell within its operating window without wasting energy. Cost, the ultimate arbiter for grid storage, depends on cheap raw materials but also on manufacturing reliability and the ability to build large-capacity cells without defects.

Looking forward, the review maps out four directions that its authors believe will determine whether sodium-based liquid metal batteries make the leap from laboratory to substation. Computation-driven materials screening promises to accelerate the discovery of optimal electrolyte and electrode combinations by simulating thermodynamics and kinetics before any crucible is filled. Multiphysics collaborative regulation aims to exploit, rather than merely tolerate, the fluid dynamics of the cell to enhance mass transport and performance. Highly reliable packaging must deliver the hermetic, corrosion-resistant enclosures that decades of service demand. And full-lifecycle recycling, already demonstrated for key materials of low-cost liquid metal batteries in work published in ACS Sustainable Chemistry and Engineering, could give the technology a decisive sustainability advantage, since the constituent metals and salts can in principle be recovered and reutilized with relative ease compared with the complex chemistries of conventional lithium-ion cells.

The stakes for getting this right are considerable. Recent analyses of long-duration storage, including studies of the value of storage in zero-emissions grids and of Europe’s vulnerability to prolonged wind droughts known as Dunkelflaute events, all point to the same conclusion: decarbonized electricity systems will need vast quantities of affordable storage capable of discharging for many hours or days. Sodium-based liquid metal batteries, with their abundant raw materials, tolerance of deep cycling, and potential for decades-long lifetimes, fit that requirement profile unusually well. The review by Cui, Du, and Wang makes clear that the remaining hurdles are real, spanning electrolyte chemistry, electrode kinetics, fluid mechanics, and heavy engineering, but it also shows a field that has moved from proving the concept to systematically dismantling each obstacle. If the multi-cation electrolyte strategy and the accompanying electrode and packaging advances continue on their current trajectory, the strange and elegant idea of storing electricity in pools of molten metal may soon be humming quietly beside the power plants it helps to tame.

Subject of Research: Sodium-based liquid metal batteries for grid-scale energy storage

Article Title: Research progress in Na-based liquid metal batteries: multi-cation systems, interfacial mechanisms, and engineering challenges

Article References: Cui, K., Du, Y., & Wang, L. (2026). Research progress in Na-based liquid metal batteries: multi-cation systems, interfacial mechanisms, and engineering challenges. Ionics. https://doi.org/10.1007/s11581-026-07567-z

Image Credits: AI Generated

DOI: 10.1007/s11581-026-07567-z

Keywords: sodium liquid metal batteries, molten salt electrolytes, multi-cation systems, grid energy storage, self-discharge, alloy electrodes, interfacial wetting, electro-vortex flow, thermal management, corrosion, recycling, long-duration storage

The technology at the heart of the project is thermochemical energy storage, or TCES, a method of preserving energy through reversible chemical reactions rather than through sensible heat in water, molten salt or rock. In a TCES system, heat drives an endothermic reaction in a solid material, storing energy in the chemical products; later, the reverse reaction releases that energy as high-temperature heat. The approach offers several advantages that conventional storage technologies struggle to match. Chemical storage can hold heat for long periods with relatively low standby losses, it can operate continuously, and it can deliver heat at the high temperatures demanded by advanced manufacturing processes. For industries that need a low-cost, large-scale storage solution, those qualities make TCES an increasingly attractive candidate.

The obstacle, Li explains, is that almost nothing is known about the granular flow inside these reactors. Engineers can measure how much material enters and leaves the vessel, but the motion of particles within the bed itself has been extraordinarily difficult to observe. Optical techniques are generally restricted to surfaces or to reactors with transparent walls, which are impractical at extreme temperatures. Interior-probing methods such as X-ray tomography or magnetic resonance imaging can penetrate the bed, but they are complex, expensive and limited in resolution. At higher temperatures the problem compounds further, because particle properties shift under the combined influence of gas flow, heat transfer and chemical reactions. The result is a field that has advanced with very little experimental data to guide it.

To close that gap, Li and his team in UCF’s Thermal Energy Storage Development Lab will build a high-temperature test rig designed to measure particle mechanical and flow properties under reactor-relevant conditions. The rig will allow the researchers to quantify how metal oxide particles behave as they are heated, fluidized and cycled through reaction states at temperatures approaching 1,500 degrees Celsius. Those measurements will then feed into computational models, giving simulations an experimental anchor they have never had. With validated models, the team can examine the gas-particle, interparticle and particle-wall interactions that experiments alone cannot easily resolve, effectively reconstructing the invisible interior of a working reactor on a computer.

The simulation campaign is central to the project’s strategy. Li plans to use the models to evaluate different reactor designs and operating conditions, identify undesirable flow behavior before it causes problems, and develop new strategies for controlling how particles move through the system. The most promising of those particle-flow control strategies will then be tested experimentally on the rig to determine whether they genuinely improve overall reactor performance. This iterative loop between measurement and modeling is what distinguishes the effort from previous work in the field: rather than inferring particle behavior from indirect signals, the team intends to measure it directly and then design around what they learn.

The material focus of the experiments is metal oxide chemistry, particularly magnesium-manganese-oxide. Compounds in this family can repeatedly store and release energy through reversible reactions, functioning much like a battery that charges and discharges, except that the stored quantity is heat rather than electricity. Cycling stability matters enormously for a storage medium that may undergo thousands of charge-discharge cycles over a plant’s lifetime, and magnesium-manganese oxides have emerged as promising candidates for high-temperature thermochemical cycles. Understanding how these specific particles flow, abrade, agglomerate and interact with reactor walls at temperature is essential to translating laboratory chemistry into dependable industrial hardware.

The implications extend well beyond energy storage alone. Li points out that biofuel production, food processing and pharmaceutical manufacturing, despite producing very different products, share a critical common feature: large quantities of particles or granular materials must move through processing equipment while simultaneously exchanging heat and mass. A better understanding and control of granular flow could improve reactor uniformity and efficiency in biofuel processing, achieve more consistent heating and drying in food production, and enhance particle handling and process consistency in pharmaceutical manufacturing. In other words, the fundamental data generated by this project could ripple across multiple sectors of the economy that have long managed granular flows largely by empirical trial and error.

Timing, Li argues, is another reason the work matters now. The rapid growth of artificial intelligence has created an enormous appetite for electricity, and data centers increasingly strain grids that are simultaneously absorbing variable renewable generation. TCES could address part of that challenge by providing long-duration energy storage that helps integrate renewable electricity and improve grid flexibility. In such a scheme, electricity would be converted into high-temperature stored energy, retained for long periods with relatively low standby losses, and then discharged as heat or power when demand peaks. That ability to time-shift large quantities of energy across hours or days is precisely what many grid planners identify as a missing piece of the decarbonized energy system.

The NSF award also carries an educational mission through the UCF Thermal Energy Storage Ambassadors, or TESA, initiative. Through TESA, UCF students and researchers will bring energy storage concepts to K-12 students, community colleges and local communities using accessible demonstrations and hands-on activities. Li’s stated goal is not only to teach students about energy storage but to show them how the fundamental subjects they learn in school, including physics, chemistry, mathematics and engineering, can be applied to solve real-world energy challenges. Building a pipeline of curious students, the project suggests, is as much a part of the energy transition as building better reactors.

Li brings substantial experience to the effort. He serves as a core faculty member of UCF’s Center for Advanced Turbomachinery and Energy Research and is principal investigator on more than $4.6 million in research funding from federal agencies and industry, including the National Science Foundation and the U.S. Department of Energy. He previously received a $3.8 million DOE grant to develop an integrated particle-based thermochemical energy storage system for concentrating solar power, and he has collaborated with the Florida High Tech Corridor and Duke Energy on advancing electrically heated thermochemical storage systems for long-duration energy storage. With the new NSF-supported rig, that accumulated expertise will be directed at a question that has stubbornly resisted measurement: what actually happens to a particle inside a reactor too hot, too opaque and too complex to see into. The answer could help determine whether thermochemical storage fulfills its promise as a backbone technology for industrial heat and a stabilizer for a renewable-powered grid.

Subject of Research: Experimental characterization of particle flow and mechanical properties in high-temperature thermochemical energy storage reactors

Article Title: UCF researcher to characterize particle properties in thermochemical energy storage reactors for first time

Article References: UCF researcher to characterize particle properties in thermochemical energy storage reactors for first time. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: thermochemical energy storage, granular flow, metal oxide particles, magnesium-manganese-oxide, high-temperature reactors, energy storage, renewable energy integration, long-duration storage, computational modeling, National Science Foundation, University of Central Florida, data center energy demand