Silanano launches the first new commercial battery chemistry since 1991 | Hacker News Reader

Silanano launches the first new commercial battery chemistry since 1991

silanano.com

82 pointsspullara33 comments

For anyone interested in what the tech is (very hard to find on their website) here is a quote from a techcrunch article https://techcrunch.com/2019/11/04/battery-tech-startup-sila-...

"Sila Nanotechnologies has developed a drop-in silicon-based anode that replaces graphite in lithium-ion batteries without requiring changes to the manufacturing process. The company claims that its materials can improve the energy density of batteries by 20% and has the potential to reach 40% improvement over traditional li-ion."

Is that consistent with their claim that it's a fundamentally new chemistry?

No. If they had a real breakthrough, they wouldn't be selling it in a fitness wristband.

There are many "solid state battery" companies. Solid Power, QuantumScape, Ampicera, ProLogium. Toyota. ProLogium seems to be the only one actually shipping a battery, but the technology seems to be getting close to working. The next generation of batteries should not have the thermal runaway/catches fire if damaged problem.

I don't think this is a valid argument against this being a real breakthrough. The first major production lithium ion batteries were for Sony camcorders.

Not all that different from wearables, and obviously lithium ion batteries were a pretty major breakthrough.

Camcorders sipped amps, wristband seeps microamps up to single digit milliamps.

Anything that has radios in it, such as devices with bluetooth, are going to see peak currents in the amps.

I've designed and worked on a number of devices that are very similar to these wrist bands. The typical modes are uA draw for long sleep periods, 10s of mA while collecting data/doing whatever for milliseconds, and 500mA to 2 or 3A for a few hundred ms every once in a while to transmit data.

So low ESR, high power capable batteries are very desirable for this sort of device. Especially when you consider that 2A draw is likely around 20C out of a wearable battery!

Please point me to one of devices you worked on that is wearable and draws 3A when transmitting. I would also love to see a BLE device with power consumption in the amps, especially considering TX power limits in the standard. Its normal for Class 3 BLE radio to consume same current when transmitting as when receiving (in <20mA range).

Indeed. All breakthrough techs initially suck. They're a shitty, borderline-useless, but FUNCTIONAL breakthrough. And that's really the thing - they might be a joke, but if they do work, and if they're just barely usable for some niche product (specifically in some new space that other solutions were not really viable for), then the giant deathball of applied corporate R&D starts rolling on them.

SSDs are probably the absolute bellweather of this, since their early forms held mere kilobytes. They were uniquely useful because they were non-volatile, and immune(-ish) to the kinds of shaking disruptions that befoul magnetic media (usually via head crashes).

The thing about these things is the corporate R&D deathball. Once it gets rolling, they can breed a mouse to be the size of an elephant. The more profound/insight-driven challenge lies in getting that initial mouse, and getting that initial foothold of a market opportunity.

Batteries of course were another great example. The reason we have viable electric cars, instead of trashy ones with a 20-mile range, is entirely thanks to the consumer electronics industry (broadly including power tools, etc). They needed batteries and paid for the R&D. Fast forward 25 years, and "the best we can do" jumped from a 20-mile range, to a 250-mile range.

>SSDs are probably the absolute bellweather of this, since their early forms held mere kilobytes

You are three orders of magnitude wrong, first commercial SSDs were in the ballpark of HDD capacity and perfectly sufficient for OS install (~1990).

>instead of trashy ones with a 20-mile range

again even EV1 had over 100 miles range, first electric cars in 1910 had 30–40 miles range.

Not sure what that has to do with anything. They built a plant for mostly research but its enough to sell for wearables.

As they increase scaling they will start do phones and then cars.

Its about scaling, but that is totally outside if they have a 'breakthrough' chemistry.

There has been a lot of research in to replacing the anode in lithium ion batteries with materials other than graphite. Some have been successful, but only to limited levels.

A pure silicon anode would be a huge step up in energy storage, but there are a ton of issues that need solving down this path.

Tesla has slowly been working towards integrating more and more silicon in to their anodes. They announced plans to try and do effectively what SilaNano is doing here at their "battery day" last year, so it is a pretty big deal that they are pulling it off in a production battery.

Its not pure silicon. Its a silicon composite particle. The difference between other commercial silicons (like that in every Tesla) is that you not combining silicon (silicon oxide usually) with graphite.

The problem is we don't actually know what is in Sila Nanoparticles, but I am pretty sure its not 100% silicon.

Removing graphite is the claim to fame here.

Correct. Tesla, and many others, are working on similar concepts where it is silicon nano particles encapsulated with some other material. I'm betting that's what SilaNano is doing.

This, in theory, allows the silicon to expand without cracking, which has been the major issue with silicon anodes so far.

Here is an image from Tesla's "battery day" that gives a visual idea about the issue/potential solution: https://cdn.shopify.com/s/files/1/0173/8204/7844/articles/SS...

No, Tesla is actually not using Composite Nano particles.

They are using micro particles that are coated with a polymer and then mixed with graphite and a binder.

This is quite different. In that version you contain the silicon expansion not by having complex composite particle but rather by having the polymer contain the cracking.

See this image: https://youtu.be/YTW_Q63bEEY?t=929

Sila on the other hand is creating a more complex silicon particle that is partly empty and can expand without cracking. They have not released a picture, but Talga resources has a silicon composite product and they have a picture:

See this: https://youtu.be/WeTAC6Xe4Uw?t=1557

In general Limiting Factor on yt has by far the best videos on this:

#5 The Science Behind Tesla Silicon - https://www.youtube.com/watch?v=YTW_Q63bEEY

#7 Tesla Silicon is Disrupting Silicon Disruption - https://www.youtube.com/watch?v=AL6xp4fLdFw

He also has the interview with Talga about their Si product:

Mark Thompson - Talga https://youtu.be/WeTAC6Xe4Uw?t=1506

And also with interview with Sinanode:

Sinanode: Low Cost Silicon Nanowire Coated Graphite for OEMs - https://www.youtube.com/watch?v=ySCFZIWMYQA

There is also :

Professor Shirley Meng: The Future of the Anode (C, Si, Li) - https://www.youtube.com/watch?v=0ktsgwzUh3A

It's hard to say really. Nickel-cadmium refer to the electrodes, but lithium ion refers to the electrolyte, and lead-acid refers to both the lead electrodes and the acid electrolyte. Switching out the electrode may qualify as a new chemistry, although 'fundamentally new' might be a stretch.

No. Enovix beat them to market (seemingly by just a few months) with a 100% silicon anode. These two companies will change the world, the only question will be is it Enovix (which makes whole batteries) or Sila (which only makes anode materials).

> These two companies will change the world [...]

That is a bold affirmation. Do you have data to back up your claim?

It's a prediction of the future. Even with all the data in the world there's no way you could make an accurate call.

A data backed assumption is still better than a wild guess.

The sun will rise tomorrow, this is a prediction of the future, I'd say it will turn out to be pretty accurate.

When it comes to predicting something like the impact of a new technology or the success of a startup, I'd argue it's worse.

At least a wild guess doesn't pretend to be anything more than hope and speculation.

What matters is how good your batteries are compared to price and at what scale. What combination of silicon/graphite or other materials is used doesn't really matter.

I for one am more exited about Tesla mixing metallic silicon with graphite, rather then these complex composite particles. I think that will be much easier to scale.

Silanano published a white paper last year that gives a pretty good overview of lithium-ion batteries and explains why their silicon anode is a big deal.

https://silanano.com/news/futureofenergystorage/

It's very approachable even for someone without a chemistry or EE background.

Anodes are really not a that big deal, even if you manage to replace them with a solid block of lithium.

At max, they make 20%-25% of the modern lithium cell mass. So, even if you manage to just use lithium, you will still need few percents of it that for the lithium itself, and some mechanical spongy structure to hold it.

Moreover, "solid" lithium implies solid electrolyte, and solid separator, which will also add to the mass.

Thanks that was a really really interesting read. Probably deserve its own HN submission.

Some more info on the product page: https://www.whoop.com/thelocker/whoop-4-0-vs-3-0-whats-new/

The new device has 17% improved battery capacity which allows it to maintain the same battery life at 2/3 the size. Not groundbreaking overall, but could make a big impact for wearables. 2mm shaved off an Apple watch, for example, would make it way more comfortable.

"completed and tested 55,000 iterations of our technology"

Over the 10 years they state it took them, that means over 15 versions/iterations per day (including Sundays and holidays). Colour me puzzled. I wonder just what it is they mean by "iteration"...

You can try multiple variations at the same time. A research team could test roughly 200 variants per week to get this number and be able to test each variant for the entire week.

Good point. I was more-or-less assuming you'd want to see the results from one variant before deciding how to proceed, but you're right - parallel testing would be quicker, if "less efficient" (in some sense).

The company is valued at $3.3 billion, but their news article sounds like an over promising kickstarter pitch. The whole "secret ingredient" thing and the feel of the entire website makes me think they just want traction for a soon to come IPO.

Si anode batteries weren't already being made? Are they really the first company to produce it, or is this a press release? Remember studying them years ago in Stanford Ph240 class.

https://www.theverge.com/2021/9/8/22662979/whoop-fitness-tra...

Appears to be first mass commercialization of the tech.

Tesla (and likely others) uses Si in their anodes but its not Si anode. Its a graphite anode with 2-10% silicon oxide.

you can find articles from several years ago that discuss using silicon anodes : https://www.google.com/search?q=site%3Aieee.org+silicon+anod...

Reply on news.ycombinator.com