Donut Lab's 409 Wh/kg result is real. It still does not prove the chemistry

A scale has finally supplied the number Donut Lab has spent months being asked to produce. In a customer report dated September 1, Finland's VTT measured 409.3 Wh/kg and 804.7 Wh/L from a Donut Battery V1.5 pouch cell. Those are exceptional cell-level figures.

They are also much narrower evidence than the words “solid-state breakthrough” suggest. VTT measured energy, mass and volume. It did not identify the electrolyte, analyse the chemistry, establish cycle life or show that this cell can be manufactured at vehicle scale. The useful reading of the result sits between dismissal and confetti: the number is validated, while the battery story around it remains unfinished.

The seven-page VTT report is admirably plain about its scope. Donut Lab supplied one cell, designated DL6, described as an energy-optimised V1.5 variant. Testing followed the customer's test plan. At 25°C, VTT charged the cell to 4.25 V, discharged it at a gentle 0.1C rate to 2.3 V, and repeated the capacity cycle twice.

The laboratory weighed the pouch and measured its dimensions with a digital scale and calliper. Its summary records the result directly: “The measured gravimetric energy density was 409.3 Wh/kg, while the measured volumetric energy density was 804.7 Wh/L.” The calculation used the second discharge.

That resolves an important question. The tested object stored a great deal of energy for its measured weight and envelope. Mike the Car Geek, reviewing the report on YouTube, acknowledged the figure before concentrating on what the short document leaves out. His concerns include the absence of the underlying mass, dimensions, capacity and charge-discharge curves from the public report. The Electric Viking made the same central distinction more bluntly: the test establishes the energy-density result, not what the cell is made from.

The voltage window is the part the public report will not let you check. The new cell's customer-supplied recommended voltage window is 2.3 to 4.25 V. Earlier V1 testing used a narrower range, according to Mike's comparison of the reports. A wider usable window can increase the measured watt-hours, but the new VTT report does not provide enough raw data to quantify how much of the 409.3 Wh/kg result comes from that change.

Donut Lab says V1.5 uses the same chemistry as its earlier V1 cells, optimised for higher energy density. In the company's announcement of the result, CEO Marko Lehtimäki says that after launch the firm focused on manufacturing the next generation of cells, and that it plans to scale that generation to an industrial level. That is a material clarification because the headline test is not a production-readiness exercise.

It also complicates attempts to combine every published test into a single validation story. VTT's energy-density work concerns DL6, a V1.5 cell. An August Intertek report concerns two different devices supplied for destructive examination. Intertek found their layer arrangement consistent with a bipolar stack design, but its report explicitly states: “Intertek have not verified that the DUT is a functioning battery cell which can be electrically discharged/charged.”

So the public evidence does not yet follow one specimen from electrical performance through chemistry and disassembly. One VTT cell demonstrates high measured energy density. The Intertek samples demonstrate construction consistent with a bipolar stack. Previous tests address other performance claims on other cells. Treating those separate samples as if one battery passed the entire sequence would go beyond the reports.

Cell energy density is not pack energy density. A road-going battery needs cooling, containment, electrical connections, monitoring, crash protection and usable buffers, all of which add mass and volume. The VTT result therefore cannot be dropped straight into a vehicle-range calculator. Nor does a two-cycle, low-rate capacity test say how the cell will age under fast charging, winter use or thousands of mixed driving cycles.

The product claim has moved with it. Donut Lab described its technology at launch as available for production vehicles, yet the same announcement that published the 409 Wh/kg figure now points to the next generation for industrial scaling. A credible route from a hand-supplied V1.5 pouch to repeatable Gen 2 output needs yield data, production samples and vehicle deliveries, not another adjective.

It is no longer reasonable to say Donut Lab has failed to produce an independently measured 400 Wh/kg-class cell. VTT has measured one. It is equally premature to say the report proves a production-ready all-solid-state battery. Chemistry, repeatability, durability and pack-level performance remain separate questions.

The next convincing test would be less theatrical and more demanding: select a production-intent cell from a batch, record its full curves and dimensions, cycle it deeply, verify its materials, then repeat the exercise on multiple samples. At 409.3 Wh/kg, Donut Lab has earned attention. The laboratory paperwork has not yet earned the whole claim.