Solid-State Batteries in Cold Weather: What Low-Temperature Tests Really Show

The cold-performance comparison in QuantumScape's Q1 2024 shareholder letter has an important footnote: all plotted data points followed charging at approximately 25°C. It therefore addresses discharge at the plotted temperatures—not charging at those same cold temperatures.

The chart also reports equivalent-area electrode discharge capacity, not vehicle range. Its comparisons have different origins: QuantumScape tested its prototype and a commercial energy cell in its laboratory, while the commercial power-cell data came from a manufacturer specification sheet. Read it as the developer's disclosed comparison, not a uniform independent pack test.

For a buyer, the distinction changes the follow-up. If the application needs cold charging, that question still needs its own evidence. A useful discharge result is neither a failed charging test nor proof of cold-charging capability. A later recovery measurement after cold exposure would answer another question again.

Lowest temperature alone also hides the electrical condition. A 2025 Nature Communications paper by Hong and colleagues reports these results at −60°C for its Li-In/LGPS–L1.25NTCl/LiCoO2 research configuration:

The corresponding Figure 4f/g tests use a 3.0–4.5 V potential window referenced to Li+/Li. These are the authors' reported experimental values, not AntBattery measurements or pack ratings.

The meaningful reading is not “a battery works at −60°C.” Keep the current and measured output attached to that statement. A result at one current does not establish the same output at another, and the paper's specific-capacity unit should not become a vehicle-range or pack-energy claim.

This is a defined laboratory configuration with its own methods and fixtures. The example is useful because it exposes what a temperature-only headline leaves out; it does not prescribe a commercial cell's charging conditions or reproduce the experiment as a buyer's test procedure.

A third distinction concerns where the temperature applies. The July 2026 photothermal-study abstract indexed by PubMed describes a PEO-based composite electrolyte in a transparent research battery. Light-induced heating raises the internal cell temperature during operation in subzero surroundings.

That thermal input is part of the reported approach, not a detail to remove when shortening the headline. “Operation in cold ambient conditions with intentional heating” is a different claim from an unheated cell operating internally at the ambient temperature.

Only the abstract supports the account here. It does not establish a commercial pack implementation, complete temperature history or net energy benefit. Before using such a result for an application decision, ask whether the intended system includes the relevant thermal input. Do not assume a conventional enclosure reproduces a transparent research device's conditions.

These examples are not a ranking of developers or chemistries. They show three ways a shortened claim can lose its meaning: omitting the charge sequence, omitting the current, or omitting intentional heat input. Keep the missing distinction in the question you send to a supplier.

For example: “The document shows discharge after warmer charging, but our application requires a separately supported cold-charge condition.” Or: “We need output at the intended load, not only evidence that the cell operated at the stated temperature.” If the reference used active heating, ask whether the proposed application's evidence includes it.

AntBattery's product range includes semi-solid offerings; the independently described research configurations above are not their specifications. Use the existing sample acceptance plan for the broader evaluation process. Here, the task is narrower: establish which cold-use claim the available test actually answers and which still needs product-specific evidence.

Do not try an unsupported charging condition to fill the gap. The exact manufacturer's instructions and responsible evaluation team govern testing. An unanswered question is neither proof of failure nor permission to proceed.