Is this the smallest battery in widespread use?

-June 17, 2014

No matter how low power your data acquisition system is, it still needs a source of energy. You may be able to implement some sort of energy-harvesting scheme (although often not practical for many reasons), and so a discrete battery or ultracapacitor may be the only viable option.

That battery may be bigger than your entire circuit and become the dominant physical feature the design as well as constraint on the package size. Batteries' construction and performance doesn't scale up or down easily either physically or electrically, although their energy capacity (watt hours) is roughly proportional to their volume for a given chemistry.

So what do you do if you need a really, really tiny battery, and there's nothing commercially available? You do what any good engineering team would do: you design and fabricate your own. That's what the researchers at Pacific Northwest National Laboratory did for salmon tags in a fisheries project being run by the U.S Army Corps of Engineers, to track salmon migratory paths. The electronics of the tags were down shrunk to the point where there was the option to inject the tracker rather than insert it surgically; the latter option being more time consuming, costly, and stressful for the fish.

The microbattery they devised (see "A battery small enough to be injected, energetic enough to track salmon") uses lithium as the anode and carbon fluoride for the cathode, and is a cylinder about the size of a grain of rice, 6 mm long and 3 mm wide, weighing 70 milligrams (see figure below). To keep the internal impedance low - critical to the battery's performance as a micropower source - it's made in layers to increase the internal surface area. As an added challenge, it has to be efficient at the lower temperatures where the fish often love.

Not much energy capacity but it's still enough for the mission: hundreds of these 6×3-mm batteries are in use and powering injectable fish tags (Courtesy of Pacific Northwest National Laboratory).

Although the press release doesn't specify the energy capacity of this tiny power source, it does say that the density of this chemistry and technology is about 240 W-Hr/kg, much greater than more common silver oxide, which stores around 100 W-Hr/kg. Doing a little math, I calculate its energy capacity at about 17 mW-Hr - not much, for sure, but apparently enough to allow the transmitter to send a 744-microsecond signal burst every three seconds, for three weeks. (That's also an impressive demonstration of ultra-low-power circuit design; let's not ignore that aspect of the project.)

When I first saw the headline about this tiny battery, I assumed it was an experimental development and feasibility demonstration at a lab or university, perhaps using two ultratiny, dissimilar metal probes in an electrolyte or some other proprietary technique - the kind of attention-getting project where one or just a few of the devices had actually been made and evaluated (and there's nothing wrong with that, of course). But the press release said that about 1,000 of these units had been made, and 700 were already in use in injected tags — that's a genuine reality check. Although the batteries are handmade using a precise layering technique described briefly, it seems to me that the fabrication process could be automated if the need was there (of course, that's easy for me to say!).

Have you seen any smaller batteries in actual use, in comparable numbers? What were the circumstances and what was the underlying technology?

And one question for the folks at PNNL: what happens if you accidentally catch one of these salmon and eat the battery? Besides any medical effects, do you become part of the data set?


Loading comments...

Write a Comment

To comment please Log In