Powers to be

No matter how long-lasting a battery company says its product is, it's unlikely any will ever claim their devices could last 200 years or more — unless nuclear microbatteries one day become commercially available.

Unlike traditional batteries that rely on chemical reactions to produce energy, nuclear microbatteries transform the heat created by the decay of radioactive isotopes into electricity.

They are unlikely to be used to power your TV remote control in the near future, but the rice grain-sized devices might be used to run a variety of micro electromechanical systems (MEMS) soon.

Nuclear microbatteries' size and endurance makes them ideal for use in situations where MEMS are embedded in the concrete foundations of roads and bridges to monitor condition, or stationed high in the air to check the atmosphere for pollutants or hazardous biological terror agents.

Funded by a three-year, $300,000

grant,

researchers are attempting to perfect a prototype so that it is able to generate all of its potential electricity.

Lead researcher James Blanchard and his assistant Rui Yao have already developed a method of converting the heat created by radioactive decay into electricity — by using thermocouples — and they have recently devised a microscale insulation to cover and heat up the battery, thereby increasing the amount of electricity it produces.

Prototypes of the insulator, which consists of very thin layers of silicon separated by poor conducting semicircular silicon oxide pillars, were constructed in the clean room facilities at Wisconsin's Centre for Applied Microelectronics. The insulator's heat radiation and conduction was studied with elaborate computer models developed by Blanchard and Yao. So far, they say, the models suggest that at smaller scales and at low vacuum conditions, the prototype is just what they need to create an efficient thermoelectric battery. However, the insulation they are using is a little thicker than what they ultimately want to achieve.

'The thinnest we've made is about 45 microns,' said Blanchard. 'We'd like to get to half of that. There's nothing fundamental preventing us from getting to 20. It just requires the use of different processing equipment.'

With an earlier battery prototype the team developed, power output was entirely determined by the energy voltage of the particles emitted by the decaying radioactive material. The microbattery consisted of a tiny semiconductor diode made up of two layers of silicon.

The first layer, called

-type material, was doped with an element that gives it an excess of conducting electrons, while the second, called

-type material, was deficient of electrons. The device generated electric current when the diode was met with radioactive material — which, in the case of the prototype, was the beta-particle emitting isotope nickel 63 (Ni63).

When beta particles, which are high-energy electrons, were shot into

-type silicon, it displaced the spare electrons into the

-type, which caused current to flow across the semiconductor.

This prototype, however, had certain limitations on the amount of electricity it could produce. This is because silicon begins to break down when the energy of the electrons hitting the lattice rises above 250 kiloelectronvolts (keV). While Ni63 produces electrons with a maximum energy of 66.9 keV, there are few other isotopes that won't damage the silicon.

With this knowledge, the researchers decided to try testing a different, more conventional, method that transforms the heat created by radioactive decay into electricity. To do this, they used thermocouples — devices with two junctions, each made from a pair of different conductors that are held at different temperatures. The output is determined solely by the temperature difference between the junctions, so thermocouples can work with any radioisotope — regardless of the voltage of the particles it emits.

'You can't use high-energy beta particles with the silicon diode, but with thermocouples, all we need is low thermal conductivity,' said Blanchard. 'A little bit of damage [to silicon insulation] would probably help energy conversion, so you really don't care.'

When the thermocouple device prototype was first tested, it produced a few tens of nanowatts. With the new insulation concept, Blanchard said it is possible to create a thermoelectric device that produces about a milliwatt, which is enough to make it a viable source of power for a MEMS application.

While the team's first microbattery prototype used beta particle-emitting Ni63, Blanchard said with a thermoelectric device it would be preferable to work with alpha-emitting isotopes.

'Generally, when people use a thermoelectric converter they prefer alpha particles because their energy density is much higher,' he said. He added that the conversional efficiency of thermoelectrics are an average of five per cent, with beta cells being slightly worse and alpha cells being slightly better.

The technology involved with nuclear microbatteries has been used before on grander scales. NASA's Cassini probe, which was launched in 1997, used them to power its mission to study Saturn and its moons.

The probe, however, used a battery the size of a dishwasher. The challenge for the Wisconsin team has been to miniaturise that technology.

Despite the potential for nuclear microbatteries, there are still those who will be sceptical about their safety. While eyes often widen with alarm at the mention of the word 'nuclear,' Blanchard and Yao say there is little radioactive material actually involved.

The microbatteries are said to be no more dangerous than a cracked smoke detector, which contains small amounts of radioactive materials.

Moreover, many of the proposed applications for these microbatteries are situations in which a battery-operated system is placed in a remote location and left alone for years.

For example, the half-life of Ni63 isotope is 102 years. This means that a battery made with it would not need to be changed for more than 200 years.

Still, potential human exposure to radioactive material wouldn't be an issue with one proposed application — using nuclear microbatteries to power swarms of nanospacecraft in the atmosphere of Mars or Venus or around Earth's orbit.

Blanchard admits nuclear microbatteries will not be available out of the laboratory for another few years. The researchers are still deciding on their power and size, as well as working out ways to keep costs down — expenses increase with the amount of radioisotope used. 'There is not a lot of development work left,' he said. 'It's just a matter of getting more funding and making these design decisions and putting it all together.'