Organic Memory Gains Momentum

Hybrid prototypes promise to give conventional memory a run for the money

4 min read

Memory chips may one day rely on organic compounds to store information, just like that gray matter in your brain. In the past few months, several research groups reported promising organic memory prototypes--from devices that function as dynamic random-access memories (DRAMs) to high-capacity, nonrewritable storage media similar to CD-ROMs. Organics won't make your computer any more powerful, but because they don't require clean-room fabrication and endless rounds of photolithography to produce, they might make memory chips and other storage devices much easier to manufacture--and therefore much cheaper.

Organics enthusiasts hope to open a crack in the US $27 billion market for semiconductor memories. But their approach is not to offer all-organic devices. Instead, to ease the technology's adoption, they plan to mesh their materials with existing silicon processing technology.

Among the chief challenges in developing such hybrid devices is finding organic compounds that don't simply burn to cinders when submitted to the extreme manufacturing and operating conditions required for today's silicon electronics. Recently, scientists at the University of California, Riverside, reported finding one such material, an organic molecule known as porphyrin, which works as a charge-storage element and occurs naturally as part of hemoglobin and chlorophyll.

Application of relatively low voltages to a synthetic version of porphyrin make it switch back and forth from an electrically neutral to a positive charge, so that the two electronic states can encode the zeros and ones of digital information, says David Bocian, a chemistry professor at Riverside. A single layer of porphyrin, in effect, could replace the capacitor in standard silicon DRAMs, which is more costly to produce.

Bocian and co-workers have shown that a layer of porphyrin molecules on top of a silicon substrate can work as a memory device that survives temperatures up to 400 °C, a condition typically encountered during the manufacturing of silicon chips. Just as important, the device can undergo trillions of read/write cycles, a requirement of DRAMs in real-world operation. What's more, porphyrin holds its charge for minutes instead of the tens of milliseconds of a traditional DRAM.

That advantage should reduce power consumption substantially, because the chip will not have to refresh its memory cells nearly as often. Start-up ZettaCore Inc., in Denver, Colo., of which Bocian is a founder, is already testing porphyrin-based prototype memory chips and working to implement first-generation products in the next couple of years.

In Another organics effort, electrical engineering professor and IEEE Fellow Stephen Forrest, at Princeton University, in New Jersey, also adopted a hybrid approach. But instead of a rewritable memory, his group teamed up with researchers at Hewlett-Packard Laboratories (H-P Labs), in Palo Alto, Calif., to build a prototype write-once-read-many-times (WORM) device. The permanent media could be used as very low cost--even disposable--high-capacity storage devices for images, for example.

The organic molecule used by the Princeton researchers is a conductive polymer known by the acronym PEDOT, which normally is employed as an antistatic coating on television screens and electronics packaging, to form a fuse. When enough current is put through it, the polymer changes its conductivity state permanently--going from conducting to insulating. A blown fuse, then, can be read as a zero and an intact fuse as a one. Each 1-bit PEDOT memory cell is 100-200 nanometers in diameter, about one-fifth the area required to store a bit on a CD. A million bits could fit into a square millimeter of paper-thin material, and layers could be stacked on top of each other, considerably increasing storage density.

"We have read these things thousands of times, and no changes have been observed."

The device consists of an array of crisscrossed electrodes with a PEDOT fuse and a silicon thin-film diode sandwiched between them at each crossing point [see diagram, " Minuscule Memory"]. Manufacturing such arrays should be cheap, since it uses inexpensive raw material and has none of the pricier chip-making requirements, such as a clean-room environment or multiple rounds of photolithography. It could eventually be done with simple pattern-printing technology.

Also, unlike CD drives, PEDOT devices require no moving read heads for writing and reading bits. A memory element is instead read by applying a test voltage to a row and measuring the current flowing from a column. "So you read the entire array quickly by scanning this test voltage from row to row while observing currents flowing in the column lines," says Forrest. "We have read these things thousands of times, and no changes have been observed." Further tests are necessary, he says, under different environmental conditions, and if everything goes right, the memory could be available within five years or so.

The Semiconductor Industry has shown much interest in organic memories, because it is getting prohibitively expensive to maintain the growth in performance of memory chips with current silicon technology. Big companies such as Intel, AMD, IBM, and Philips have in-house groups working on organic memory devices or have partnered with university groups and start-ups in the field. "There's an extreme need for new technologies," says Bob Merritt, a vice president at semiconductor-industry research company Semico Research Corp. in Phoenix, Ariz.

A possible market entry point for such new memory technologies, Merritt says, is applications that are mobile and for which battery life is crucial. Today's most sophisticated cellphones, for example, package multiple types of memory, each useful for a particular function--one for data applications, another to store the phone's software, a third for high-speed communications processes. "That's a packaging solution to a semiconductor problem," he says. Ultimately, organic compounds could have the right properties to lead to a cheaper and more power-efficient "universal memory" that would replace all others.

This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions