Battery patents and public information are two halves of one read when assessing a battery technology’s manufacturing readiness. A patent shows what a company claims to have demonstrated and how the material or cell for which it seeks protection was made. Public information – product announcements, pilot-line statements, conference presentations, journal papers, specification sheets – shows what the company says it will ship, and when. Neither makes sense alone.
A patent without the public context is an experiment without a destination. A public claim without the patents behind it is a press release. Read together, the two sources reinforce each other, and the patent half requires a disciplined AI methodology to allow for industry-wide KPI benchmarking. Below are the numbers I look for.
The First Gate: Fundamental Properties, at Any Scale
Ionic conductivity comes first. For an EV cell, the electrolyte has to conduct above 10-3 S/cm (one millisiemens per centimeter) at room temperature. Below that line, power, fast charging and cold-weather operation are lost before any manufacturing question arises. Fortunately, the library of liquid, solid-liquid and all-solid electrolytes that fulfill this benchmark has massively grown in recent years.
Laboratory values are entirely sufficient for this check. That is what makes patents valuable even though, in my reading of the solid-state battery patent literature published since 2022, roughly nine in ten patent families disclose laboratory parameters only, fewer than one in ten disclose pilot parameters, and under two percent disclose anything at industrial scale. The lab-scale majority still allows benchmarking of the properties that must be in place before upscaling, on three axes:
• Performance – ionic conductivity, electrochemical stability window, transference number (the share of the current carried by the working ion), critical current density.
• Inherent safety – non-flammability, thermal onset temperature, gas formation – for sulfide electrolytes specifically hydrogen sulfide.
• Cost – the raw-material floor of the formulation, which no process improvement lowers.
A chemistry that clears this gate is worth reading for manufacturing readiness. One that does not is not, however elegant its process patents. Getting the order right raises the odds of a successful product launch along the whole value chain – materials supplier, cell maker, pack integrator and end application alike.
Seven Numbers of Manufacturing Readiness
Once a chemistry clears the gate, seven further numbers show how far it is from a production cell.
1. Cell format and size. Battery material pellet, coin cell, single-layer pouch, multilayer pouch, ampere-hour class. From material proof-of-concept to industrial-size cell proof-of-concept.
2. Stack pressure. Tens of megapascals is a laboratory fixture. A pack can realistically hold at most one to two megapascals, and ideally can be operated without excess pressure. Whatever pressure a cell needs, the module has to be engineered to deliver – in weight, volume and cost.
3. Areal loading and areal capacity. Thin electrodes inflate rate capability and cycle life. A commercial cell needs roughly five or more milliampere-hours per square centimeter on the cathode side.
4. Test temperature and rate. A cycling result at 60 °C hides room-temperature shortfalls. In the polymer-electrolyte literature, results reported at or above 60 °C outnumber those at or below 0 °C by close to four to one.
5. Voltage cycled versus voltage claimed. In the same body of literature, stability windows measured by voltammetry cluster near 4.9 V. Cells are cycled to a median of about 4.2 V, and only some 40 percent reach 4.3 V. The claimed window is the chemistry's promise, the cycled cut-off is the product.
6. Coulombic efficiency to the third decimal. For an anode-free cell, where the metal anode is plated from the cathode's inventory on every charge, 99.8 percent per cycle leaves 67 percent of that inventory after 200 cycles. Eighty percent retention after 500 cycles requires 99.955 percent.
7. Process parameters and process chain. Tape width of at least five centimeters, line speed in meters per minute, kilogram batch size, continuous web length of a roll-to-roll run, press conditions for cells with an edge length of at least ten centimeters. And the steps themselves: mixing, coating, calendering and stacking run on an existing line, while sintering, hot pressing and isostatic pressing tend to need new equipment. Routes that keep the electrolyte-filling step and cure the electrolyte inside the assembled cell are the clearest drop-in path, but tend to produce less well-defined polymers compared to separate polymer manufacturing steps.
Why Next-Generation Cells, and for Whom
Each application re-weights these numbers, which is why the same patent can be irrelevant for one market and decisive for another.
Stationary storage, including the battery buffers now being installed next to AI data centers, requires cost per kilowatt-hour per cycle, calendar life and inherent safety. A non-flammable cell changes siting, fire-code exposure and cooling load for a facility that cannot tolerate downtime. Weight barely matters.
Humanoid robots need volumetric energy density and fast charging, and they operate next to people, so thermal-runaway risk weighs as heavily as watt-hours per liter.
Drones are governed by gravimetric energy density and discharge power, often at low temperature.
EVs need all of it at once: watt-hours per liter for the pack, watt-hours per kilogram for range, 10 to 80 percent charge in 5–12 minutes, safety, and a cost that beats the incumbent technology.
Why Cost Favors the Cheapest Materials
Process costs usually fall with scale – equipment amortization, yield learning, throughput. Raw-material costs do not fall below their commodity floor. Once every process cost has been competed down, the material floor decides who wins. That is why three approaches deserve particular attention:
Low-cost silicon anodes – abundant and high in capacity, needing the right process technology to shape the silicon into a nano- and microscale architecture that accommodates the swelling.
“Anode-free” designs – the anode material disappears, paid for with the efficiency requirement in number six, which likely requires an all-solid electrolyte at the interface deposited with a manufacturing technology that avoids interface defects at acceptable cost.
Manganese-rich, high-voltage cathodes – manganese is cheap and abundant, but in liquid cells it tends to dissolve into the electrolyte and migrate to the anode, and the voltage required to fully extract its energy oxidizes the liquid. A solid electrolyte that holds 4.5 V and carries no liquid to transport dissolved manganese makes the cheapest cathode usable in terms of raw materials.
The Bottom Line
Clear the ionic conductivity gate first, then read the seven numbers, always against what the company says in public. Scale competes process cost away, not raw-material cost. Cheap silicon anodes, “anode-free” designs and manganese-rich cathodes are structural advantages, not incremental ones – provided that process costs for the novel manufacturing steps fall far enough for adoption beyond niches.
Pirmin Ulmann is co-founder and CEO of the energy storage innovation and patent monitoring subscription service b-science.net, Stäfa / Zurich, Switzerland.
