Anthro Energy has broken ground on the conversion of an existing manufacturing site in Louisville, Kentucky into an advanced electrolyte plant. TechCrunch reported the groundbreaking on August 18, 2026 U.S. time, with the company targeting production start in 2028. Annual output is set at 12,000 tons, equivalent to 25GWh of battery capacity. What the groundbreaking raises is whether a liquid precursor can be impregnated evenly into every corner of mass-produced cells and cured uniformly.
The company's Proteus works by injecting a liquid-phase precursor into cells using existing equipment, wetting the pores of the electrode and separator, and then cross-linking and solidifying it during the formation process (including initial charge-discharge cycles). It attempts to handle both the wettability of a liquid electrolyte and the mechanical properties of a solid/semi-solid electrolyte within existing cell manufacturing processes. When reading the plant plan, before translating the large figure of 25GWh into a number of cells produced, one must look at how far this material capability can be carried over into a reproducible cell process.
25GWh Is a Converted Electrolyte Capacity Figure
25GWh is not a direct capacity to produce cells or EVs; it is a figure obtained by converting the electrolyte manufactured into battery capacity terms. TechCrunch reported it as equivalent to more than 300,000 EVs. A simple calculation gives about 83.3kWh per vehicle, but this is merely a convenience calculation using exactly 300,000 as the denominator for the reported "more than 300,000," and is nothing more than a conversion of material capacity. It does not mean the plant will assemble cells for more than 300,000 vehicles.
The same distinction is needed for weight. Dividing the annual 12,000 tons by 25GWh gives 480 tons per GWh. This ratio shows the relationship between the capacity conversion figure the company presents and its material production volume. Since the energy density of individual cells, the amount of electrolyte input, and customer-specific designs have not been disclosed, one cannot use this to pin down cell specifications or vehicle-model-specific supply volumes.
On the funding side, the company has received a $24.9 million grant from the U.S. Department of Energy (DOE), and an $18.4 million figure is presented as an Inflation Reduction Act (IRA) 48C investment tax credit. The former is a grant, the latter an investment tax credit—they are not funding of the same nature. The disclosed breakdown of project costs alone does not allow one to calculate a subsidy rate.
The plant's product mix is also not fixed to Proteus alone from the outset. According to TechCrunch, while customers are conducting evaluations, the plant will also manufacture other companies' formulations, with the plan being to increase the proportion of Proteus as customers complete validation and align accordingly. The plant's operation marks a milestone of the start of manufacturing, but commercial adoption of Proteus will only be signified once customer cell qualification has progressed further.
Entering as a Liquid, Curing Inside the Cell
Proteus's aim is to build a network inside the cell after reaching the pores as a liquid. Anthro describes it as both solid and semi-solid. It is a design that bridges the manufacturing gap between the conventional electrolyte-filling process and the process of forming interfaces with a solid electrolyte, through precursor impregnation and solidification during the formation process.
The related patent WO2024173770A2 describes a representative procedure in which a battery stack is wetted with a solution containing a polymer precursor, initiator, and plasticizer, forming a covalently bonded gel electrolyte network. Candidate plasticizers listed include ordinary liquid solvents such as EC and DEC, and the disclosed range spans from under 10 wt% to over 50 wt%. Therefore, knowing that gel electrolyte examples exist is a separate matter from knowing the composition ratio of the commercial Proteus.
The same patent includes examples with precursor viscosity of 10–50 cP, impregnation for at least 6 hours, a fixed pressure of 5–1,000 psi, and a curing example of 7 hours raised to a maximum of 80°C. It also mentions degassing as needed, ionic conductivity exceeding 0.1 mS/cm at 25°C, and elastic modulus of 0.1 MPa or higher. However, these are representative examples given in the patent and cannot be definitively taken as the standard specifications of a mass-production line, the process times of commercial cells, or guaranteed values for Proteus.
Compatibility with existing equipment suggests the possibility of trials without wholesale equipment replacement. But this is not equivalent to saying the process need not be re-qualified. It is necessary to confirm how the conditions of impregnation, curing, and formation processes tie into cell design and customer formulations, and to verify in mass production whether uniform solidification occurs. The company itself has not stated that re-qualification is unnecessary.
The Boundary Between Solid and Semi-Solid
If an electrolyte that solidifies is uniformly called an "all-solid-state battery," the significance of the liquid-phase precursor and plasticizer used by Proteus is lost. Anthro itself describes Proteus as both solid and semi-solid, and the patent's representative procedure forms a gel electrolyte network. For this reason, the distinction between wetting pores with a liquid and how much liquid phase remains after curing is a point that must be separated when interpreting performance and manufacturing.
Speaking to TechCrunch, the CEO said that cells using Proteus could be 10 to 15 times stronger than liquid-based ones. This is a claim from the company's own side, not a published comparative test of customer cells. Commercial composition, and cell-level gravimetric energy density (Wh/kg) and volumetric energy density (Wh/L) from third-party testing, have not been disclosed. Cycle life, fast-charging capability, and low-temperature characteristics also cannot be confirmed. Safety evaluation test results, customer names, and contracted volumes have not been disclosed either. The same goes for yield and pricing. One cannot conclude dendrite suppression, non-flammability, or long lifespan from the solidification process alone.
As for the technology's lineage, the company explains that Stanford Chemical Engineering licensed 2019 research to Anthro. That research separated dynamic hydrogen bonding from ion-conducting segments and reported conductivity of about 1.2×10^-4 S/cm at room temperature, and about 2×10^-4 S/cm with plasticizer added. Toughness was reported at 29.3±1.4 MJ/m3, with conduction maintained even at 200% elongation. It can be read as a starting point for a design that combines mechanical strength with ionic conductivity.
Even so, there is no basis for regarding the material in this research as an identical formulation to the current Proteus. Nor can the research's measured values be substituted for the output values of commercial cells. What Anthro needs to demonstrate by 2028 is not simply reproducing the research's figures, but showing that performance and manufacturability can be achieved together even when impregnation and curing are repeated under customer cell conditions.
Commercial Trials Through 2028
By the 2028 production target, the value of the plant will not be determined by annual tonnage alone. Customers will evaluate the electrolyte with their own cell formulations, and the manufacturing side will confirm curing uniformity and yield under those conditions. The process time and cost obtained there will determine whether the proportion can be shifted from other companies' formulations to Proteus. The phased product mix reported by TechCrunch is premised on this sequence.
Customer qualification is broader than evaluation of material samples alone. If the precursor wets the pores of the electrode and separator before cross-linking and solidifying during the formation process, it will be necessary to confirm how variation in impregnation and curing conditions affects reproducibility from cell to cell. The patent's examples of 6 hours, 5–1,000 psi, and up to 80°C indicate the range of conditions to be evaluated, but they do not indicate the conditions the plant will actually adopt.
The timeline of policy also needs to be separated when reading demand for 2028. The U.S. Internal Revenue Service (IRS) states that the clean vehicle tax credit will not be recognized for new, used, or commercial vehicles acquired after September 30, 2025. Because of this change, 2028 demand cannot be tied solely to the old 30D consumer credit. On the other hand, there is also no basis for generalizing that the value of Anthro's stated approach—using sourcing that does not fall under FEOC from day one—disappears simply because this tax credit has ended.
Whether the Louisville plant can narrow the gap in mass production cannot be judged from the converted figure of 25GWh alone. What must first be confirmed is the curing time, yield, and price at the point when the proportion of Proteus is increased in customer cells. Only once third-party test results for cell-level gravimetric energy density (Wh/kg) and volumetric energy density (Wh/L), cycle life, fast-charging capability, and low-temperature characteristics are all disclosed can one gauge whether the material capability has truly moved closer to commercial supply capability.