KAIST Advances Giant Batteries Toward Commercialization for AI Data Centers

The rapid expansion of artificial intelligence data centers is creating a new demand for energy-storage systems capable of operating at enormous scale. These facilities consume electricity continuously, placing pressure on power grids and increasing the need for systems that can store renewable energy and deliver it reliably when sunlight and wind power fluctuate. Researchers at the Korea Advanced Institute of Science and Technology (KAIST) have now reported a manufacturing advance that could bring one of the leading candidates for this role—vanadium redox flow batteries—closer to commercial deployment.

A KAIST team led by Professor Hee-Tak Kim has developed a faster and more stable method for producing the vanadium electrolyte used in these large batteries. The redesigned process reduces production time by approximately 67 percent, cutting it to about one-third of the duration required by the conventional method. The researchers say the approach also reduces impurities, lowers energy and equipment requirements, and allows the key catalyst to be reused more than 2,500 times without a significant loss of performance.

Vanadium redox flow batteries, or VRFBs, store energy in liquid electrolytes held in external tanks. During charging and discharging, the electrolytes flow through an electrochemical cell, where vanadium ions change their oxidation states and either absorb or release electrical energy. Unlike lithium-ion batteries, whose energy capacity is largely tied to the size and number of their cells, flow batteries can be expanded by increasing the volume of electrolyte in the tanks. This makes them particularly attractive for grid-scale storage, renewable-energy facilities, and data centers that require large reserves of electricity.

The technology also offers a safety advantage. VRFB electrolytes are water-based and nonflammable, substantially reducing the fire risks associated with many conventional battery systems. However, the chemical composition of the electrolyte must be carefully controlled for the battery to operate efficiently. The standard starting material has an average vanadium oxidation state of +3.5, commonly written as V3.5+. Producing this composition at industrial scale has traditionally been slow, expensive, and technically demanding.

The conventional manufacturing route uses two reduction stages. First, a chemical reducing agent—typically oxalic acid—causes vanadium ions to gain electrons, lowering their average oxidation state. The electrolyte is then subjected to electrochemical reduction, in which an electric current adjusts the remaining vanadium ions to the desired V3.5+ composition. That second stage requires a costly flow-battery stack and substantial electrical power, adding both capital expenses and operational complexity to the production process.

The KAIST researchers discovered that the problem was not limited to the final electrochemical step. Their analysis showed that the chemical reduction itself slows dramatically when the average oxidation state reaches approximately +4.1. This intermediate condition acts as a kinetic bottleneck: the reaction proceeds relatively efficiently before this point, but then decelerates sharply, much like traffic accumulating where a highway narrows. The slowdown extends the manufacturing process and limits the practicality of producing large quantities of electrolyte.

To bypass this bottleneck, the team combined chemical and catalytic reduction in a redesigned sequence. Chemical reduction is used during the earlier, faster stage, while a platinum-on-carbon catalyst, known as Pt/C, takes over when the vanadium reaches an average oxidation state of about +4.1. The catalytic route accelerates electron transfer through the slowest portion of the process, allowing the production system to avoid the rate-limiting region rather than forcing the chemical reaction to continue through it.

The new method also addresses a chemical-quality problem. Conventional processing can leave residual oxalic acid in the electrolyte, where it may act as an impurity and contribute to performance degradation inside the battery. By switching to catalytic reduction at the critical stage, the KAIST process eliminates the remaining oxalic acid while producing the targeted V3.5+ composition. According to the researchers, the Pt/C catalyst maintained its effectiveness through more than 2,500 reuse cycles, an important result for a process intended for industrial operation rather than laboratory-scale demonstrations.

“This study combined reaction engineering principles with thermodynamic predictions to identify the rate-determining step in the chemical reduction and redesigned the electrolyte production process to overcome this major bottleneck to the commercialization of large-scale batteries,” Kim said. The study, led by doctoral researcher Kyunghwa Seok, was published in Advanced Energy Materials under the title “Streamlined V3.5+ Electrolyte Production by Leveraging Chemical and Catalytic Reductions.” The authors say the advance could help reduce manufacturing costs and accelerate the use of vanadium flow batteries in AI data centers, renewable-energy installations, and other applications requiring dependable, long-duration energy storage.

Web References: https://doi.org/10.1002/aenm.71029

References: Kyunghwa Seok, Minseong Kang, and Hee-Tak Kim, Advanced Energy Materials

Keywords

Vanadium redox flow batteries, VRFBs, energy storage, AI data centers, renewable energy, vanadium electrolyte, catalytic reduction, platinum-on-carbon catalyst, grid-scale batteries, long-duration energy storage