http://www.aip.org/pnu/2005/729.html
A room-temperature, palm-sized nuclear fusion device has been reported by a UCLA collaboration, potentially leading to new kinds of fusion devices and other novel applications such as microthrusters for MEMS spaceships.
The key component of the UCLA device is a pyroelectric crystal, a class of materials that includes lithium niobate, an inexpensive solid that is used to filter signals in cell phones. When heated, a pyroelectric crystal polarizes charge, segregating a significant amount of electric charge near a surface, leading to a very large electric field there. In turn, this effect can accelerate electrons to relatively high (keV) energies (see Update 564).
The UCLA researchers (Brian Naranjo, Jim Gimzewski, Seth Putterman) take this idea and add a few other elements to it. In a vacuum chamber containing deuterium gas, they place a lithium tantalate (LiTaO3) pyroelectric crystal so that one of its faces touches a copper disc which itself is surmounted by a tungsten probe. They cool and then heat the crystal, which creates an electric potential energy of about 120 kilovolts at its surface.
The electric field at the end of the tungsten probe tip is so high (25 V/nm) that it strips electrons from nearby deuterium atoms. Repelled by the positively charged tip, and crystal field, the resulting deuterium ions then accelerate towards a solid target of erbium deuteride (ErD2), slamming into it so hard that some of the deuterium ions fuse with deuterium in the target.
Each deuterium-deuterium fusion reaction creates a helium-3 nucleus and a 2.45 MeV neutron, the latter being collected as evidence for nuclear fusion. In a typical heating cycle, the researchers measure a peak of about 900 neutrons per second, about 400 times the "background" of naturally occurring neutrons.
During a heating cycle, which could last from 5 minutes to 8 hours depending on how fast they heat the crystal, the researchers estimate that they create approximately 10-8 joules of fusion energy. [To provide some perspective, it takes about 1,000 joules to heat an 8-oz (237 ml) cup of coffee one degree Celsius.]
By using a larger tungsten tip, cooling the crystal to cryogenic temperatures, and constructing a target containing tritium, the researchers believe they can scale up the observed neutron production 1000 times, to more than 106 neutrons per second. (Naranjo, Gimzewski, Putterman, Nature, 28 April 2005).
The experimental setup is strikingly simple: "We can build a tiny self-contained handheld object which when plunged into ice water creates fusion," Putterman says. (More information at http://rodan.physics.ucla.edu/pyrofusion .)