Proton memory is ultracheap but slowBy Kimberly Patch, Technology Research News
A University of New Mexico researcher has found both good news and bad in refining an unusual and potentially cheap memory device made of common Field Effect Transistors.
The bad news is it's relatively slow, and it doesn't remember forever; the good news is it's fast and stable enough for some applications and its still very, very cheap.
A Field Effect Transistor (FET) is a simple device that contains a channel through which electrons can flow. The rate of flow through the channel depends on the amount of voltage going across a dielectric, or naturally nonconducting, silicon oxide gate. "It's called a Field Effect device because by putting voltage across it you generate an electric field and it's this electric field which modulates the connectivity through this channel," said Roderick Devine, a research professor at University of New Mexico.
Devine discovered four years ago that the electric field also drives protons to one side of the device or the other, and when the voltage stops, they stay put.
In addition, depending on which side the protons stop, they essentially prop the gate open or shut. The protons "actually take the place of the original voltage," said Devine. "Depending on where you leave these mobile charges, you can either have this channel conducting or not conducting," he said.
These two states -- open or closed -- can be used to store the ones and zeros of computer memory, said Devine. "If you find it's conducting you can say that it's an on state, if you find it's not conducting, you can say it's an off state."
In putting the memory through itís paces, however, Devine has determined that it is relatively slow. "The fastest devices I've made have actually switched at about 80 milliseconds. We could probably get them to switch at about eight milliseconds. But these are very, very long times compared to normal memory, [which switches in] nanoseconds or picoseconds -- six orders of magnitude [slower]," he said.
The memory also didn't live up to its original potential as nonvolatile memory. Nonvolatile memory, like the flash memory used by digital cameras, does not need to draw electric current in order to retain information. In contrast, a computer has to boot up after being turned off because its Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM) memory is volatile, and empties when the power goes off.
Although the FET memory will remain stable for several hours in the worst conditions and at least a year in the best conditions, it's stability doesn't reach the strict industry standard, which is to hold information for 10 years in the worst conditions, said Devine.
However, there are still significant low-key uses for the memory just because it is so very cheap, he said.
For example, the memory cells integrated into the pixels on flatscreen monitors must be refreshed many times a second because they can't hold information very long and so have to be rewritten. The FET memory devices "which would certainly hold for hours could avoid having to do memory cell refresh -- you could significantly reduce power consumption if you didn't have to refresh your flat-panel display every 12 or 20 milliseconds," said Devine.
"I think it could be potentially interesting," said Jan Van der Spiegel, professor of electrical engineering and chairman of electrical engineering at the University of Pennsylvania. "It's basically just a regular transistor. You don't need to change much on it, so it could be cheap" to produce, he said. "It may lead to inexpensive nonvolatile memories where the speed of writing is not very important."
If it could eventually be made smaller and thus denser it could be appropriate for low-cost, high storage density applications, Van der Spiegel added. "Of course we need to look at what are the competitors. Eventually it all comes down to economics," he said.
If industry finds a need for this type of memory, it could be used fairly quickly, said Devine. "Within six months you could probably produce prototype circuits in an industrial environment," Devine said. "We've already made proper transistors in a clean room using the sort of materials and technological approaches which are perfectly standard," he said.
Meanwhile, Devine is working on refining the technology to increase its retention and speed it up to its full, although still relatively slow, potential.
He's also looking into measuring similar effects in new materials. There may be "different gate dialectics which would allow the thing to switch faster," he said. "Coming online are other materials which will eventually replace silicon dioxide. We don't know what the situation on proton transfer in those materials will be," Devine said.
Devine's research was funded by Sandia National Labs. He published a technical paper on the research in the September 18, 2000 issue of Applied Physics Letters. His previous research on the proton memory effect was funded by Sandia, France Telecom, the Air Force and the University of New Mexico.
Timeline: 6 months; indefinite
Funding: Government, University
TRN Categories: Semiconductors and Materials
Story Type: News
Related Elements: Technical paper, "Electric-Field Dependence of Mobile Proton-Induced Switching in Protonated Gate Oxide Field Effect Transistors," Applied Physics Letters, Sept. 18, 2000
November 29, 2000
Molecular motor shifts speeds
Virtual mic carries concert hall sound over 'net
Researchers ready radio-receiver-on-a-chip
Proton memory is ultracheap but slow
Crystal forms gas-triggered switch
Research News Roundup
Research Watch blog
View from the High Ground Q&A
How It Works
News | Blog | Books
Buy an ad link
Ad links: Clear History
Buy an ad link
© Copyright Technology Research News, LLC 2000-2006. All rights reserved.