OSLO, Norway Advancing its proposition that polymer electronics technology could soon be used to create complete electronic systems in plastic, Opticom ASA is developing a PC-card format polymer memory subsystem for a customer, with plans to complete the subsystem before the end of 1999. The R&D house has also designed a polymer-over-silicon demonstrator chip due to be delivered from a European wafer fab in February.
The company believes that polymer-based memory systems could be used to provide large-capacity memories held in multiple layers, either standalone or laid down over conventionally processed silicon dice that would contain associated logic or microprocessor systems. The plastic systems could be manufactured using inexpensive reel-to-reel continuous production processes, according to Opticom.
Since its founding in 1994, the company has been researching conjugated polymers as conductors and semiconductors. It claims to have achieved a number of milestones in 1997, including a single-layer polymer memory structure, a polymer radio transmitter, an active memory film on an organic substrate, and submicron feature sizes. An overview of its work is available online.
Company chairman Thomas Fussell said Opticom has been sustained in its research partly by shareholders' funds and partly by technology licensing. Computer-games maker Eidos (London) has taken a 15 percent shareholding in Opticom, Fussell said, and the companies have formed "a joint venture to exploit our technology in the field of computer games."
Opticom increased its activity in 1997 and now claims 39 research projects going on around the world, including efforts with Lucent Technologies Inc. and connector maker AMP Inc. Opticom has also set up a unit in Sweden, Thin Film Electronics AB, which conducts research with the University of Linkping.
Johan Carlsson, senior research manager at Thin Film Electronics, said Opticom's concept of "a passively addressed crosspoint matrix in which no components are single-crystal materials" promises several advantages over other memory subsystems. Those include stackable memory architectures, with or without plastic substrates as carriers, and the smallest possible memory cell with inherently high yield. The passive crosspoint memory structure avoids the use of space-consuming transistors, Carlsson said.
Such systems have the potential for very high data-transfer rates because of the large area and parallel readouts that are possible when contacts are arranged in multiple layers, Carlsson said. Finally, such systems should be low-cost, because of the ease of reel-to-reel processing and because they are founded on low-cost materials.
More than read-only
Carlsson said that the crosspoint matrix architecture is not restricted to a read-only memory function, as would be the case with a metal-programmed cross-point silicon memory. "We can do ROM, write-once-read-many or erasable memory, depending on the function of the polymer," he said. Control and test logic would vary according to the function.
Memory cells are created with the technique by applying current at a certain voltage to a top and bottom electrode. The organic material is modified at the point where the electrodes cross each other. By changing the voltage level, information can be written, read or erased.
Opticom is developing an all-plastic memory subsystem that will fit on a credit-card-sized PC card. "The first version of the card will contain 1 Gbyte with a data-transfer rate of 0.5 Gbyte/second. The access time will be around 50 ns," said Carlsson. To achieve the high capacity, Opticom will lay down memory arrays in multiple layers.
For now, Opticom will not divulge details of the conjugated polymer materials it uses in its process. "The materials are the core of our technology and therefore cannot be disclosed," said Carlsson, though he did say Opticom is using "conducting, insulating and semiconducting polymers" in its systems.
"The field of organic electronics has now matured such that organic materials with good insulating, semiconducting and conducting properties can be found," he added. "Among other materials, we are considering polythiophenes and BCB as semiconducting and insulating materials, respectively. In the hybrid version, we are also working with conventional metals and silicon.
"The switching materials that constitute the active part of the memory can switch in less than 10 ns. The inherent read time is less than the switching time. The layers will be deposited through meniscus coating and defined through a novel technique proprietary to Opticom."
The company is also developing an additive deposition/patterning process.
Carlsson offered one hint as to the nature of the proprietary definition process: "We apply multiple layers first and then define the layers in situ, using a post-processing technique," he said.
As for the lifetime and endurance characteristics of Opticom's polymer memory technology, Carlsson said that "qualified lifetime tests have not been performed yet" but that "stability of the memories seems not to be an issue. The memory switches well at elevated temperatures up to 150°C, which was the limit of the test. Read/write-cycle tests of the material up to 1 million cycles have been performed without significant degradation."