Donut Battery: Finnish start-up promises miracles

Contents

In the automotive world, a donut is actually a circle of rubber abrasion: youngsters try out drifting in supermarket parking lots or on country roads, drawing a black donut with their rear wheels. More recently, the term has come to represent the supposed innovation of a Finnish start-up. Its donut battery, so the promise goes, surpasses the performance data of conventional cells many times over. According to the developers themselves, they have succeeded in bringing a cell with a solid instead of liquid electrolyte to series production readiness, which is supposed to charge from 0 to 100 percent in five minutes and last for 100,000 cycles. That sounds fantastic.

Donut Lab, of course, knows that this performance claim is not very credible, especially regarding durability, and plays with it: the website I donut believe, a play on words combining the company name and the translation "I don't believe it," regularly publishes results of tests conducted at VTT. This independent institute is a kind of Finnish TÜV with a good reputation. The tests, often conducted in such facilities according to the client's specifications, are intended to prove that Donut Lab has indeed achieved a revolution.

"The test procedure is designed to deceive the reader"

But that's exactly what they don't do, says Dr. Joachim Sann from the Center for Materials Research at Justus Liebig University Gießen. The semiconductor physicist specializes in solid-state batteries and had already expressed his skepticism in the podcast Geladen. In an interview with heise/Autos, he elaborates on his criticism: "The test procedure is designed to deceive the reader," Sann states. Important parameters such as the size and weight of the cell are omitted. Other measurement results, on the other hand, contradict Donut Lab's claims.

For example, the Coulomb efficiency of the tested cell is low. In translation: according to the values published by Donut Lab, the cell would not have a lifespan of 100,000 cycles, but only 30 to 50 cycles. For comparison: the battery systems installed in today's electric cars typically last between 1500 and 5000 cycles, depending on the design and cell chemistry. "Donut Lab's data indicate about one percent degradation per charging cycle," explains Dr. Joachim Sann. Thus, after 30 to 50 cycles, the final wear limit of 70 percent is reached.

Commercially available cell in an unusual test protocol

The energy density of 400 watt-hours per kilogram (Wh/kg) also announced by Donut Lab would not be proven by the VTT data. The protocol simply provides no information on this. Estimates based on plausible comparison standards, such as the adhesive tape in the photos of the tests, suggest rather 280 to 300 Wh/kg gravimetric and 600 to 800 watt-hours per liter (Wh/l) volumetric energy density: "This is the data of good, but conventional NMC cells," says Dr. Joachim Sann. He is now convinced that a commercially available cell was tested under an exceptionally demanding test procedure to suggest a breakthrough.

The researcher from Gießen is not alone in the industry; on the contrary. No one attributes any credibility to Donut Lab's statements that they have developed a solid-state cell with 100,000 potential charging cycles and an energy density of 400 Wh/kg. The fact that the company is consistently moving forward is a known strategy: start-ups need to show successes to raise money.

The goal is the purely metallic lithium anode

Nevertheless, it is interesting to consider why research and development are being conducted on solid-state batteries at all. After all, industry giants like Toyota or Mercedes regularly present drivable prototypes. Battery cells with a solid instead of liquid electrolyte have different properties. They can formally have high-temperature resistance and be less flammable. The central argument for an advantage is different: a solid electrolyte is the prerequisite for installing a purely metallic lithium anode.

This metallic lithium anode represents the electrochemical optimum. Almost all lithium cells in use – whether installed in an electric car or a smartphone – have a graphite anode. This pure carbon is reliable, but it slows down charging performance and significantly increases weight. Recently, electric cars have been mixing silicon with graphite to mitigate these weaknesses. Examples include the Porsche Taycan and the Mercedes CLA EQ (Test).

Difficult to handle because it is highly reactive

If it were possible to produce a battery cell with a solid or semi-solid electrolyte and a metallic lithium anode, the gravimetric energy density would increase significantly. An electric car could therefore be much lighter with the same energy content. However, purely metallic lithium is difficult to handle in production because it is highly reactive. Battery companies like CATL, which dominate the market, may be working with hundreds of developers on such problems. The reality of life looks different.

Low Tech in high volumes

One of the most important factors for sales success is not outstanding performance data, but sufficient performance at a low price. In this respect, a triumph of simple LFP cells (for lithium iron phosphate) is currently emerging: in China, the market share in 2025 was 81.2 percent.

This cell chemistry is comparatively inexpensive and robust: in practical tests in winter, the Xpeng G6 and G9 showed that the standardized charging cycle from ten to 80 percent can be completed in twelve minutes. Such LFP cells are neither light nor do they offer high-energy density. New cell chemistries are always in competition with cells produced in large-scale series. The fact that there are constant small improvements is visible. Miracles, however, are unlikely.

(dahe)