- 16 June 2005 -

Another power approach - atomic batteries [excert]

The typical future-tech scenario calls for millions of low-powered radio frequency devices scattered throughout our environment -- from factory-floor sensor arrays to medical implants to smart devices for battlefields writes Eric Brown in MIT's Technology Review.

Because of the short and unpredictable lifespans of chemical batteries, however, regular replacements would be required to keep these devices humming. Fuel cells and solar cells require little maintenance, but the former are too expensive for such modest, low-power applications, and the latter need plenty of sun.

A third option, though, may provide a powerful - and safe - alternative is the Direct Energy Conversion (DEC) Cell, a betavoltaics-based 'nuclear' battery that can run for over a decade on the electrons generated by the natural decay of the radioactive isotope tritium.

It's developed by researchers at the University of Rochester and a startup, BetaBatt, in a project described in the May 13 issue of Advanced Materials and funded in part by the National Science Foundation.

Because tritium's half-life is 12.3 years (the time in which half of its radioactive power has been emitted), the DEC Cell could provide a decade's worth of power for many applications.

Clearly, that would be an economic boon - especially for applications in which the replacement of batteries is highly inconvenient, such as in medicine and oil and mining industries, which often place sensors in dangerous or hard-to-reach locations.

Betavoltaic devices use radioisotopes that emit relatively harmless beta particles, rather than more dangerous gamma particles. They've actually been tested in labs for 50 years - but they generate so little power that a larger commercial role for them has yet to be found.

So far, tritium-powered betavoltaics, which require minimal shielding and are unable to penetrate human skin, have been used to light exit signs and glow-in-the-dark watches. A commercial version of the DEC Cell will likely not have enough juice to power a cell phone - but plenty for a sensor or pacemaker.

The key to making the DEC Cell more viable is increasing the efficiency with which it creates power. In the past, betavoltaics researchers have used a design similar to a solar cell: a flat wafer is coated with a diode material that creates electric current when bombarded by emitted electrons.

However, all but the electron particles that shoot down toward the diodes are lost in that design, says University of Rochester professor of electrical and computer engineering Phillipe Fauchett, who developed the more-efficient design based on Gadeken's concept.

The solution was to expose more of the reactive surface to the particles by creating a porous silicon diode wafer made up of one-micron wide, 40 micron-deep pits. When the radioactive gas occupies these pits, it creates the maximum opportunity for harnessing the reaction.

The process is easily reproducible and cheap, says Fauchett, necessary if the DEC Cell is to be commercially viable.

Fabrication techniques may be affordable, but tritium itself -- a byproduct of nuclear power production -- is still more expensive than lithium in cell-phone battery. The cost is less of an issue, however, for devices designed specifically to collect hard-to-get data.

Cost is only one reason why Gadeken says he will not pursue the battery-hungry consumer electronics market. Other issues include the regulatory and marketing obstacles posed by powering mass-market devices with radioactive materials and the large battery size that would be required to generate sufficient power.

Still, he says, the technology might some day be used as a trickle-recharging device for lithium-ion batteries.

And the next step is to adapt the technology for use in very tiny batteries that could power Micro-Electro-Mechanical Systems (MEMS) devices, such as those used in optical switches or the free-floating "smart dust" sensors being developed by the military.

Another betavoltaics device, under development at Cornell University, is also targeting the MEMS power market. The Radioisotope-Powered Piezoelectric Generator, due in prototype form in a few years, will combine a betavoltaics cell with a tritium-powered electromechanical cantilever device first demonstrated in 2002.

Source: http://www.technologyreview.com/articles/05/06/wo/wo_061605brown.asp