Hacker News new | past | comments | ask | show | jobs | submit login
NanoTritium 20-Year Betavoltaic Battery (citylabs.net)
107 points by peter_d_sherman 20 days ago | hide | past | web | favorite | 87 comments

I'm not an electronics person, but as I understand it this amount of energy is typically harvestable from ambient rf in populated areas.

See the first few paragraphs of this paper for a bunch of references: https://www.ntu.edu.sg/home/nprivault/papers/ambient_harvest...

I think these are awesome ... I've got an application for a very low duty-cycle device (operates for seconds each hour) and, paired with a suitable capacitor (or super-capacitor), I can completely pot the entire system in epoxy, mount that in foam in a tennis ball and still have about a decade long service life!

EDIT: No, it's not a dog toy

If it's inside a tennis ball, it only stays not a dog toy as long as you keep it away from the dog.

And given my dog, it'd stay a dog toy for a very short half life after me got to it.

Nothing less than Kong toys in his world, everything else is a snack.

Caps leak a lot. Take it into account. I would probably use something like a very simple buck boost dc to dc converter to constantly charge a battery to provide buffer that will be enough to accommodate spikes.

I just want a sealed, finned RTG I can drop in the deep end of my pool and forget about.

You don't even need the TG part if you just want to heat the water. Drop an alpha source with a heat sink in there and it'll heat the water directly.

Of course, all the isotopes suitable for use in an RTG are a) extremely dangerous and b) extremely expensive. Pu-238 makes gold look like dirt price wise, it's around 5-10 million dollars per kilogram and is chemically and radiologically extremely dangerous, but minor details.

>> Pu-238 makes gold look like dirt price wise, it's around 5-10 million dollars per kilogram and is chemically and radiologically extremely dangerous, but minor details.

The price is actually infinite, since the global supply is somewhere between 20-30kg total. No one will sell you any no matter the price.

You can actually buy microgram quantities if you know where to look, it's occasionally used in laboratories as a test source. You could buy larger quantities if you had a really good justification and were willing to pay for the cost of producing it, that's effectively what NASA is doing.

Doable with Americium. Start hoarding old fire detectors.

Maybe you heard about the kid who did exactly that? https://www.snopes.com/fact-check/eagle-scout-nuclear-reacto...

I'm curious, what do you need to power in the deep end of your pool? lights?

I think it's about heat power not electric power

This is cool stuff but I'm not sure what application this is targeted at. Primary lithium cells have a 15 to 20 years operating life, and contrary to those batteries you can draw more power on demand when you need it and stay asleep at low current the rest of the time. Energy density seems higher but not that much.

15 to 20 years operating life on a single charge? Or 15 to 20 years operating life if you recharge them occasionally?

"primary lithium cell" means a non-rechargeable lithium battery.

Primary means not rechargeable. Secondary is rechargeable.

This is not a battery per se, it's really a generator. Over 20 years this thing is worth something like 20 watt-hours, but is far smaller than the equivalent primary cell (because it doesn't store energy, at least not chemically) and doesn't have issues with current leakage. It's meant for powering super long-lived low-power devices, one example that comes to mind that could benefit would be the container monitors used by the IAEA for monitoring nuclear material.

The energy density of this betavoltaic battery is only about half that of an LiMnO2 primary-cell battery, about 370 mWh/cc versus 650 mWh/cc. Examples of LiMnO2 batteries are the extremely common CR2032 or CR2450 coin cells.

Self-discharge isn't much of an issue with LiMn02 batteries, either, at only 1% per year.

And, of course, LiMnO2 cells have the benefits other commentators have mentioned, such as the ability to supply more than just 75 uA short-circuit.

Hmm, I was looking at the wrong figure on the page, the higher power model is much bigger than I was estimating from the pictures. Still, I imagine these at least have more stable performance at extreme temperatures. Would be nice if they had a real datasheet but I don't see one.

What is the point of a 40-pin connector when the spec sheet only indicates that two of them are used (exactly as one would expect for a battery)?

Two reasons: it's a standard form-factor, which makes it easier to work with and the additional pins hold it securely to the board.

Mounting? You solder all 40 pins to hold the device onto the board even if you only need 2 pins. Presumably the pins are repeated across all sides so you can pick any pin that makes routing more convenient.

That would make sense, but according to the spec sheets the pins are not repeated--only two of them are functional, the rest are dead.

Seems like an oversight that can be easily fixed.

Saves them money on packaging costs. Wire bonding isn't free.

Presumably the battery+shielding sets minimum volumetric requirements, and those form factors meet them while providing easy and robust circuit integration using existing standards.

Lithium batteries already beat this on every metric and aren't exotic. http://www.tadiranbat.com/long-life-xol.html

I'm pretty sure Lithium batteries don't produce continuous power for 20 years...

Non rechargeable cells can have that lifetime.

I doubt at -40C to 80C.

Then you doubt for no reason these ones form Energizer are -40 to +60c..


And quick googling shows batteries with a wider positive range.

If you scroll through that PDF you'll notice the temperature capacity chart. At -40°C you're going to have to limit your power drain a lot to get rated capacity. At low temperature the chemical reactions get too slow to take advantage of the full capacity of the battery, that's simply a reality you can't get around.

A "betavoltaic" would deliver full rated capacity at any temperature, regardless of changes or duration.

Are these actually being sold now? I first saw these advertised by citylabs years ago with vague wording on availability and maybe-pricing in the $3k range.

My understanding is there have been betavoltaic cells for a long time, and with improvements in semiconductor manufacturing technology, they've been able to move towards less radioactive materials such as the tritium used by citylabs. These things exist, but of course, tritium is not cheap.

Strontium 90 seems like a good candidate. It and what it decomposes into are both pure beta emitters. Beryllium 10 also looks good, though very low power.

Someone recently used nickel 63 too.

> Someone recently used nickel 63 too.

Yeah the Russians, right? (https://newatlas.com/nickel-nuclear-battery-design/54884)

If these are going on IoT/Sensor network devices, doesn't this add to the e-waste issue? A lot of these sensor devices are never recovered. Now we have radioactive e-waste.

It adds to the e-waste issue if the device is ground up or burnt.

Some, but not all, of the properties of tritium are somewhat friendly. The half-life is comparatively short at 12.5 years, so it isn't left for future generations. It is a beta-emitter, so shielding is straightforward. The daughter, helium-3, is quite benign. It can be readily encapsulated (see tritiated key fobs).

If it does get loose from encapsulation, it goes everywhere that hydrogen goes, which is largely everywhere. On the plus-side, gaseous tritium mixes and dilutes quickly in the atmosphere. The key is not to ingest concentrated quantities of it. The decay energy is very low, which makes assaying for tritium contamination quite challenging.

If you're looking for e-waste of radiological concern, americium smoke detectors might matter more. I'm not sure.

Maybe someone throws a handful of these into a grinder to harvest prescious metals or something. Grinder cracks the encapsulation of like 10-20 of these in sequence with an operator standing nearby. Now is it an issue?

Hydrogen diffuses very quickly. I would imagine no issue. There could be an issue if it were to burn, then the inhaled water could linger in your system for a while (although water is constantly excreted too). But then the dosage is so small, and it's alpha radiation (readily absorbed by nearby water molecules). Sure you would get a better estimate by running some calculations, but I suspect the dosage even of a handful of those totally inhaled as water vapor would be sub-background.

Compare it to mercury (which some years back was still in batteries), which evaporates easily when heated, stays in your body for very long times (and is cumulative), and has clear neurological impact.

I wouldn't worry about even a large number of those breaking.

Considering what other kind of particulate matter would be coming out of a machine grinding electronics up, tritium (which would simply float away) should be among the least of this individual's worries.

As long as the chip is in one piece, it shouldn’t be an issue. And if the chip gets smashed open... tritium is hydrogen. It’ll just float away and dissipate.

You should do more research on Tritium, it isn't the same as waste from a nuclear fission reactor

Doesn’t mean it’s good. It’s an inhalation hazard.

Not likely to be a practical problem given the tiny amount that's going to be in one of these chips. Unless you get a whole bunch of them together and open them all at once.

To me the bigger environment question is how the tritium is produced and what impacts that might have.

I mean I wouldn't eat bananas and would never fly commercial airlines if you're concerned about that.

You can’t just respond that way to every radiation worry. You need numbers.

For context: 50 microwatts * 20 years = 9 watt-hours, comparable to an 18650 cell.

So 1000 of these and I can power a 1W LED. But after 20 years it will only be 0.3W. This is comparable to RF power harvesting. What are the use cases for such minuscule amounts of power? Remote sensors?

Pacemakers and remote senors are some of the big ones I know of. Though I don't know if modern pacemakers use them still.

“In the past, small "plutonium cells" (very small 238Pu-powered RTGs) were used in implanted heart pacemakers to ensure a very long "battery life". As of 2004, about ninety were still in use. By the end of 2007, the number was reported to be down to just nine. The Mound Laboratory Cardiac Pacemaker program began on 1 June 1966, in conjunction with NUMEC. When it was recognized that the heat source would not remain intact during cremation, the program was cancelled in 1972 because there was no way to completely ensure that the units would not be cremated with their users' bodies.“

There are quite a few videos on youtube (nerdrage ? forgot) where people use light radioactive material to energize photovoltaic cells.

Poor man's RTG. Love it.

Tritium vials with a solar panel

This seems infinitely more plausible that this, which was doing the rounds last week:


“Releasing 3600 _whole_ electrons! Or even 14400!!!”

(That’s about a billionth of a 1 pico farad capacitor discharging. I suspect you need some reasonably exotic laboratory equipment to even detect 14400 electrons...)

God that website is terrible. It is so incredibly redundant about the value of infinite power, as if anyone would need an explanation as to why infinite energy would be useful, but gives so little detail or explanation about what the device actually is.

Absolutely screams investor scam to me.

What kind of battery/capacitor could you make with lithography/chip making techniques? Genuine question. I know there was an announcement about this recently. I know it would be far more expensive than the roll of tape that current lithium batteries use, but I have to imagine that Apple is working on this for the iPhone/etc

Is this the recent announcement you were thinking of? https://xnrgi.com

Maybe. I thought I saw something from IEEE; it may have been this company.

edit: Looks like yes: https://spectrum.ieee.org/energy/the-smarter-grid/the-return...

edit again: also this from 2017 https://spectrum.ieee.org/semiconductors/design/how-to-build...

It's the only one I know of applying standard semiconductor wafer techniques.

If it works out, I think this one (https://spectrum.ieee.org/energywise/energy/environment/a-gl...) is a more promising battery technology in the long run due to lifespan. It appears to be far more preliminary at this point however.

Tritium's beta decay process emits electrons with 5.7 keV energy. I wonder why these devices output such low voltages as 2.4 V. A beta decay electron should be able to jump to a cathode at nearly -5700 volts.

Very likely they have some integrated voltage downstepping.

Doing 5.7kV circuits is not your typical power engineering 101

So who’s gonna be the first to order, like, 400 of the 100 micro watt ones, to power a RaspberryPi web server for 30 years? (Which will, of course, have trashed it’s flash memory by Xmas...)

if by 400 you mean 40000

Ahh yeah. _Almost_ within an order of magnitude... (Good enough for government work, right?)

I mean this question at face value because I don't want to be that guy -- what are the applications?

This seems like a very small amount of power with very specific and unusual benefits over other power sources.

One use is in military/aviation radios to keep encryption keys in sram. The power can be disconnected and the key wiped in the event of a crash or impending enemy capture. If the key was in flash it might be recovered.

How do you know that?

Lots of IOT stuff has low power requirements and VERY low duty cycles. You may not be able to do a lot with a continuous 50μA•, but if you have a sensor that sleeps for an hour, takes a reading, reports it (over BTLE or wifi or even an epaper screen), and then goes back to sleep, this chip would allegedly be enough to keep your device running continuously for decades.

• I think some of the lower power BTLE chips can actually advertise continuously for less than 50μA on average



I keep hearing conflicting information on the legality of importing tritium into the US. Is it controlled? Is it legal to import?

I have tritium keychains from china. Probably not legal but aliexpress or wish. For the gun sights trijicon is the only game in town to get tritium installed.

IANAL, but I think tritium distributers must be specially licensed by the NRC in the US.

Cool, but if you use Carbon-14 you get more microwatts and thousands of years.

So what are some actual uses for these at such low power?

Besides what’s in the article, some Bluetooth low energy SoCs can operate on a power budget of less than 50 µA. Also keep in mind that with components like super capacitors, power can be “charged up” (very roughly like filling a large bucket with a small water hose so you can get one big toss of water) in order to get a larger pulse of power.

... medical implants...

Powered by radioactive battery?

Sure. Tritium releases very low energy beta particles; even a thin casing will render this safe.

I could swear that, as a kid near Chicago in the early 1980s, I had a tritium-backlit Timex or Casio digital watch.

That said, I can't find it online...and gave it away decades ago.

I could be wrong but tritium on watches usually wasn’t typically used on the face itself, but in gas tubes or vials on the hands and hours markers. Ball, Marathon and Luminox are three popular watch brands that use them currently.

Evidently radium stopped being used for lume in 1970, and while tritium faces may have been a replacement around that time, tritium as used only has a lifespan between 12-24 years.

Tritium was commonly used in luminescent paint on mainstream watches (Rolex, Omega, et al) until the late 90's/early 2000's.

Tried as early as in the 60s with pacemakers:


Applications are open for YC Summer 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact