Scientists have invented a computer memory system that uses protons instead of electrons to store data. The new device, described in today's issue of Nature, should be easy to manufacture and could eventually provide a cheap alternative to magnetic hard drives and other storage media.
Silicon chips are the engine under a computer's hood, providing the short-term memory for crunching numbers. But while engineers would like to design computer memory out of a single material--silicon--it's expensive to design silicon chips that store numbers for days on end. Instead, computers often rely on magnetic disks or tape for long-term storage.
But defects in silicon's crystal structure could open the way to using the material for long-term data storage, Karel van Heusden of Sandia National Laboratory in Albuquerque, New Mexico, and his colleagues have shown. When chips are prepared from silicon dioxide crystals, they often have imperfections in the crystal lattice that spawn rogue protons. Ordinarily these particles are a menace--they roam the crystal, knocking off electrons and sometimes triggering data-storage errors. But the researchers were able to regiment the protons in a defect-riddled silicon dioxide wafer by putting it in an electric field, which got the stray protons to migrate to one side or the other--storing a 1 or 0.
Van Heusden's group found that they could align thousands of protons on a single chip that stay put after the field is turned off. They also found that it takes less electricity to shuffle protons on a chip than to corral electrons in a floating gate, an insulated island designed on some silicon chips for long-term data storage.
Some experts find the new approach mind-boggling. "The very idea of using protons rather than electrons to move charge encounters intuitive resistance," says John Roberts, an electrical engineer at Cambridge University in the U.K. "People know how to control electrons. With protons, there's a lack of experience." Hard drives won't go away anytime soon, adds George Brown of Texas Instruments' Research and Development division. "Potentially it will be very inexpensive to implement, but a lot of work remains to be done" to move from a proof of concept to manufacturing, he says. However, he says, "it's a very exciting possibility that we can get protons to dance in this way."