Donut Labs Responds to Critics

Support CleanTechnica’s work through a Substack subscription or on Stripe.

There has been a frenzy of response to Donut Labs’ announcement of the world’s first solid-state battery. Donut Labs produced a response video to critics. The video says everyone jumped the shark, and independent testing is coming. It drops hints all over the place. It tells their strategy and why they elected to reveal information the way they did. In my opinion, it is completely consistent with a small company attempting to survive in a world of larger players. I don’t believe they are lying. I am sorry. This is all I have to offer those only interested in an executive summary. Others, read on. There might be something useful or entertaining beyond the headline. Some may even enjoy the writing. I do have a private reputation as a muckraker and a writer. I think that may no longer be a secret. Yes, I do more than technical writing. And, yes, I have a ton of skeptics throwing spaghetti at me. No matter. I’m used to it. It comes with the territory.

The Story

When Donut Labs announced the world’s first production solid-state battery, I knew what would happen. The response came slowly at first, then grew like a breaking wave. 

One could hardly expect the world’s largest companies researching solid-state batteries for years without result to stay silent while an upstart stole their thunder and limelight. You could see the panicked response as they struggled to explain why, out of the blue, an unknown threatened to grab the crown and brass ring before they did. Experts now had to explain to the click-hungry social media crowd and to their own corporate bosses and investors how they lost the race to an upstart, and why it took them so long and yet they still have not accomplished what this company claims. It causes severe changes in plans, and apologies to Tom, whose expertise and expressions I admire, but it does not mean companies will change plans — because they have NIH, the “not invented here” principle. That is, companies will not necessarily all flock to get Donut Labs tech. Some of them will stubbornly refuse any outside tech in favor of their own despite pressure. Others will attempt to steal it, and more on that later. 

I spent my entire career in the tech industry observing the interplay between technology companies, news organizations, and startups in Silicon Valley. None of this kind of claim, counterclaim, and inter-company rivalry is new to me. On the contrary, it is to be expected in industries in which the stakes are in billions and the advancements come rapidly, year by year, and month by month. I wrote recently about the social media storm regarding recent sodium-ion advancements, and the stark difference between projections of sodium-ion production timelines and CATL’s announcement of the Naxtra battery. One must tread carefully when making statements about what cannot be done. Exaggerations come in both directions. 

When I heard reports that major companies expressed skepticism, I was not surprised. Years ago, I wrote a story about the Tesla Semi, titled “Does Tesla Semi Break the Laws of Physics?” I proved that it was technically possible based on known empirical data and readily available information to conclude that it was possible to make an electric truck with the specifications and performance claimed. The story title was based on a response from a large, well established company that expressed skepticism. I could not verify cost claims, only performance. That kind of math and research is normal for senior design engineers. The makers of the Semi might not choose to make the vehicle that way, but they could, and I could only assert that it was possible, not predict what tradeoffs they might make. It is possible to predict what is technically possible, not what human events and behavior will follow. When the vehicle was built, and the test results agreed with my calculations, it was no surprise to me. Technology is not magic to me. I know how the sausage is made. 

As an experienced design engineer, my career has been spent attempting to justify my designs in design reviews, and have them sent into production in a venture of great cost and investment risk. No one took my sole word for it. I had to prove it with months of simulations, data, and screening. It took a huge amount of effort by me and others to give birth to those products in the real world. It is with this background that I viewed the endless torrent of social media opinion and speculation following Donut Labs’ announcement. 

As a design engineer, I know there are two general responses to claims of novel product performance. One is opinion, and the other is based on ideas and facts. The latter is the first order of business for me. It has been my professional practice to sort that out. I am familiar with doing necessary diligence to prove a design production worthy, but as a design engineer, I know the only way to silence critics is to produce a successful product. I have agonized over the prospect of failure versus success many, many times. After years of experience, I learned how to deal with it and became comfortable with it, comfortable enough to trust myself and realize I could make mistakes, but I would be a fool if I did not trust myself instead of a horde of critics. Select your critics wisely.

What Donut Labs Device Isn’t

Right off, some of the responses to Donut Labs were speculative opinion and false. Donut Labs storage technology is not a supercapacitor. This one is easy. Capacitors work on physics that depends on a simple equation.

E = 1/2 C ( V ) 2

where E is energy, C is capacitance, and V is voltage. 

Donut Labs declared an operating range of 2.5 to 4 volts. Capacitive storage relies on voltage squared for greatest energy. Capacitor voltage storage does not use small voltage swings. It uses the most voltage it can endure to optimize energy. Normally, that voltage is in hundreds of volts, limited by the dielectric strength of the insulator material between the plates of the capacitor. This idea is a non starter if Donut Labs is making a drop-in replacement for the lithium batteries it previously had in the Verge motorcycle. That is too much evidence pointing toward the idea of a capacitor being a nonstarter. You can forget that. 

The fact that this idea has circulated and has legs in social media circles is testament to the fact that news organizations of the past with large professional screening staffs are now officially a thing of the past, replaced by a social media fray hungry for clicks and not overly concerned with journalistic reputation. No doubt, the idea will persist long after this article debunks it and the product becomes an every day reality, but you cannot completely stop rumors with fact. My advice to journalists without technical backgrounds in this matter is, don’t take sides. If you are not an expert in something, do something else. Do journalism. Report what you find, hopefully in a balanced way. Taking sides and changing your mind back and forth is bad form. Don’t lay shade on CleanTechnica either. I might not like it. It might be a game changer. Game changer and holy grail are two phrases that may soon be banned by English teachers. Don’t mess with them.

We Couldn’t Do It, So Neither Could You

It should come as no surprise that self-interest dictates that companies that could not succeed at solid-state technology might respond by saying it is impossible. Sure enough, they did. I don’t want to further embarrass them, but a response sans proof is just another unsubstantiated claim. It is not just bad form to do it that way, it is a bad idea. Making uninformed claims without full knowledge is apt to end with egg on the face. The fact that one party could not do it is not proof that another party could not. It is only proof that the measures they took failed. Without knowledge of exactly what Donut Labs did to achieve true solid-state batteries and other performance claims, it is not possible to come to any conclusion based on particular measures. When and if full disclosure and reveal of Donut Labs is available, then we can fully test and determine. Without that, the only path forward is using generally based known physical laws and information. Since all parties are attempting to create designs that must be based on physical principles, and we are entering the territory of new knowledge and advancement, this is a matter of learning, not just measurement or speculation.

What Do We Know?

We know few things we can firmly base conclusions on. Unfortunately, there are very few things we do have. We’ve only had the announcement and statements of the company, pictures, and video as clues to the story. Now we have the response video from Donut Labs. The statements of the company combined with those leave a trail of bread crumbs that lead to the most likely results. We can test the self-consistency of the claims and likelihood of possibilities, and guesstimates.

The Claims

The videos reveal much information. It is not a lithium battery. It does cost about the same as lithium batteries or lower. It works on a voltage range compatible with lithium, from 2.5V to about 4V. It is planned as a drop-in replacement for a lithium battery, although exact details are not available. The stated voltage range is greater than lithium (about 3V to 4V). It works up to 100°C and down to -40°C. In videos, they have said, that is as far as they have extended testing. That means it could be more. It has double the energy density (400 Wh/kg) of the best LFP and about 50% more than average NMC. If it fits into a space replacing the prior lithium packs and the range is doubled, it has something like double the volumetric energy density as well. How could this be?

For starters, there are batteries that have high energy density at that level and beyond today. CATL’s condensed batteries are capable of 500 Wh/kg. The recently announced semi-solid state batteries in FAW EVs are claimed capable of more than that as well. These both have some measure of planned production. There is more than just a possibility batteries can achieve those specifications. 

There is a small video showing high-speed charging in five minutes. Some of the Donut Labs people are using Verge motorcycles with the advanced battery. Can a battery do 12C charging? Oh, yes, certainly. A123 Labs was doing 25C years ago. Bill Dube and Eva Hakkonen used them years ago on the Killacyle to set electric drag race records. I met Bill and Eva. Nice people. Battery people wear non-conductive, non-metallic rings. LFP batteries are capable of excessive peak currents, like 60C for short periods of time. In the videos, Donut Labs says they do the 5 minute, 12C charging, but that is only if they are thermally managed with liquid cooling. For the Verge motorcycle which is air cooled, they do 5C. This is consistent information. 

The battery is claimed to have a cycle life of 100,000. I have to say, I don’t care. If it were a measly 25,000, it would be good enough. Seriously. How does anyone know a lifetime of 100,000 cycles? Present day battery testing does elevated temperature and limited laboratory cycles to project lifetime. Higher temperature increases aging, allowing accelerated testing. Accurate coulombic testing has been pioneered for some time by Jeff Dahn and others. It works. It means the total charge is measured accurately to determine how much charge is lost over cycling. Let’s not waste time over whether it is 25,000 or 100,000 cycles. Some people even thought 25,000 is the highest previously. It isn’t. Recently defunct Natron Energy claimed 50,000 cycles. Why is that? In concession to the TLDR crowd, it comes down to how stable the materials are. This education comes courtesy of Shirley Meng, foremost battery educator and researcher. If so inclined, you owe it to yourself to absorb her teachings. 

The miracle of modern intercalating batteries is that they shuttle ions, ionized atoms, or molecules (big bowling balls), not electrons (little peas), back and forth from cathode to anode. The anode and cathode must retain intact structures for the life of the battery. The structure of the cathode makes tunnels like bowling ball receivers at the bowling alley that allow ions to park themselves in a regular, crystal-like structure, in rows. As the ions pass through the anode and cathode, they may expand the structure or undergo other transformations that modify them and reduce their capability. If those structures are stable, the battery life can be quite long. This is the story with lithium titanate, which has been known for many, many years. The idea that batteries can have very high cycle life is not new, only the particular claim. For my part, this is nearly irrelevant, because if they only had cycle life of 50,000, it would hardly be a disaster. How do they know? They probably tested it in the ways everyone tests batteries. Did they test correctly? I don’t know. We will find out when independent testing is done. Is it likely to be high cycle life? Yes. Will it matter. Not much.

What about high temperatures? 100C is pretty high. That is the temperature of boiling water. It doesn’t just survive that temperature, it works at that temperature. It better not have a paper label on it. We do know that sodium-ion batteries can endure several hundred degrees C without becoming a conflagration. Other sodium-ion batteries work up to 70C. 

In reality, the claims are not that far beyond the realm of possibility and beyond what people familiar with batteries are aware of. Their performance is a shock to lay people unfamiliar with tech, and researchers and companies trying to achieve those levels who are unable to succeed. The battery performance is not entirely beyond the realm of possibility compared to known batteries, but we do not have the data, and we do not know how rigorously it was tested and measured. It is very common for product specifications to be tightened for yield. It would not strike me as unusual if the final production cells had tighter specs. The Wh/kg will still be near that level, but the temperature range and cycle life may be less for production. 400 Wh/kg is for pouches, 350 Wh/kg is for prismatic. I don’t expect those numbers to change that much.

We know a few other details that are interesting. There is some production. It is assembled by machines, not by hand. One of their products is a powered trailer used to lower truck fuel consumption. They created a battery storage trailer to charge the other trailers, because no adequate charging infrastructure was available to charge their trailer product demos. The storage needed to do that is sizable, in the 10 MWh realm. This is consistent with at least some factory production. The claim of one GWh production is about the size of a good pilot line. Finland has a reasonable level of battery infrastructure.

What About Solid State?

This is the biggest mystery in all of it. CATL, BYD, Toyota, and others have struggled without success for many years. In the wake of the Donut Labs announcement, predictably, the others have responded with plans in the near future or limited production, anything that feeds the limitless hunger of the social media piranha frenzy. It is important to understand why solid state is so difficult. I mentioned the regular, cage-like crystal structure of the cathode and how expansion and contraction could damage it. Solid state has one more problem. It is solids, not liquids. That cathode and the anode both have a bumpy surface. Any solid electrolyte is composed of solids. A liquid can fill the pores, gaps, and bumps. Up until now, efforts have been made to force the solids into contact with high pressure. This has not been a production-worthy, low-cost approach. There are low quantities available from Quantumscape for example. Whatever anyone does, it has to solve the problem mating an uneven solid surface to a solid. Some of the attempts have failed because solid structures that expand and contract can crack. The result has been short cycle life.

What Could It Be?

There is a video that guesses what it might be and does find some useful information about the company. I give Miss GoElectric credit for digging up facts. The speculative part is interesting for detectives and might be entertaining, but probably has some mistakes in it. There are some patents and there was a deal with another company, Nordic Nano. What could fit from the clues? Let me be direct. The voltage range from 2.5V to 4V is a dead ringer for a sodium-ion battery. We know from Naxtra that a sodium-ion battery can have a self-formed, or anodeless, construction. That improves performance. We know if we pay attention to Shirley Meng, a researcher’s goal is to do both a self-forming anode and solid state. We also know from her teachings that if this is done, 400 Wh/kg values are expected. We also know that sodium-ion tech has extended temperature range operation, in liquid electrolyte, and that solid state will improve that. We know that sodium-ion tech has impressive safety, and high temperatures will not cause it to explode or burn rapidly. So, is there anything really that unbelievable about it being a solid-state, sodium-ion battery? No, not really. It is not the sodium part of it, or the performance that is so unusual. The only part we have not seen any particular knowledge of is solid state. So that is what we need to talk about. Any tech can be solid state. It can be sodium, lithium, or others. Solid state is only about the electrolyte. If it has a self-forming anode, the anode is not the issue. The cathode is. Then we need to know what type of cathode. We already know TAQ is a kind of cathode that is more than capable of 400 Wh/kg, for one. That does not mean this is what they used, but it does lend credibility to the possibility of such performance. How they get the solids in contact is the key mystery. There is no way they are going to reveal it. Every major company in the world wants to know, and some of them probably have detectives in the field trying to discover it.

That is all we really know. The speculation about Donut Labs’ solid-state battery being a scam is overdone. Let’s wait until more is revealed. Don’t call it a scam until the independent testing is done. There is too much evidence that they are not lying. If it was a scam artist, promising product and giving timelines would be the worst way to do it. Giving too many details would be a poor choice. Leave scams for the experts like Elizabeth Holmes and Trevor Milton. The last thing they would want to do is give a date for independent testing.