US researchers have claimed a breakthrough that promises nuclear battery technology with a lifespan measured in decades.

The project, which is in development at the US University of Rochester, has demonstrated an enhanced fabrication method that "in its roughest form" is already 10 times more efficient than current nuclear batteries.

The academics claim that once refined the technology has the potential to be nearly 200 times more efficient.

"Our society is placing ever-higher demands for power from all kinds of devices," said Philippe Fauchet, professor of electrical and computer engineering at the University of Rochester, and co-author of the research.

"For 50 years people have been investigating converting simple nuclear decay into usable energy, but the yields were always too low.

"We have found a way to make the interaction much more efficient, and we hope these findings will lead to a new kind of battery that can pump out energy for years."

The technology is geared toward applications where power is needed in inaccessible places or under extreme conditions.

Since the battery should be able to run reliably for more than 10 years without recharge or replacement, the scientists point out that it would be perfect for medical devices like pacemakers, implanted defibrillators, or other subcutaneous devices that would otherwise require surgery to replace or repair.

Likewise, deep-space probes or deep-sea sensors, which are beyond the reach of repair, also would benefit from the technology.

Betavoltaics, the method that the new battery uses, has been around for half a century, but its usefulness was limited due to its low energy yields.

The latest battery technology makes its successful gains by dramatically increasing the surface area where the current is produced.

Instead of attempting to invent new, more reactive materials, Fauchet's team focused on turning the regular material's flat surface into a three-dimensional one.

The scientists explained that, similar to the way solar panels work by catching photons from the sun and turning them into current, the science of betavoltaics uses silicon to capture electrons emitted from a radioactive gas, such as tritium, to form a current. As the electrons strike a special pair of layers a current results.

What has traditionally held these batteries back is that so little current is generated, but Fauchet decided that to catch more of the radioactive decay, it would be best not to use a flat collecting surface of silicon, but one with deep pits.

The pits, or wells, are only about a micron wide (about four ten-thousandths of an inch), but are more than 40 microns deep.

After the wells are "dug" with an etching technique, their insides are coated with a material just a tenth of a micron thick, which is the best thickness to induce a current.

Digging these wells in a random fashion yields a 10-fold increase in current over the conventional design.

The team is already working on a technique to create and line the wells in a much more uniform, lattice formation that should increase the energy produced by as much as 160-fold over current technology.

"Our ultimate design has roughly 160 times the surface area of the conventional, flat design," said Fauchet.

"We expect to be able to get an efficiency that very nearly matches, and we're doing this using standard semiconductor industry fabrication techniques."