A few months ago, we brought news of a bendable CPU called the Plastic ARM, which was made of amorphous silicon on a flexible substrate. Some such use cases are extremely low-power devices that can be embedded in clothing or slapped onto the surface of irregular objects, allowing them to allow small amounts of autonomous computing. But to meet the low power requirements, a minimal processor is not enough – all the components also have to sip power. And it makes it a poor fit for traditional RAM technology, which requires power to maintain the state of the memory.
But a group at Stanford has now covered it. Researchers have created a form of flexible phase-change memory, which is closer in speed to normal RAM than flash memory, but requires no power to maintain its state. And, while his work initially focused on obtaining something that is flexible to work with, the principles he uncovered during his work should apply to phase-change memory in general.
change of phase
People have made flexible forms of memory before, including flash and ferroelectric RAM, and resistive RAM can be made from materials that are also bendable. But phase-change memory has myriad advantages. It works by connecting the two electrodes through a material that can form a crystalline and amorphous state, depending on how quickly it is cooled after heating. These two states differ in how well they conduct electricity, from which they can be distinguished.
The electrodes provide convenient modes of reading and writing. By applying a high level of current, you can heat the material; Suddenly turning off the current will cool it to the amorphous state, while slowly turning the current down can lead to a crystalline state. Once this is done, the state can be read by passing a very small current and reading the resistance; It is also possible to store more than one bit per device by adjusting the heating to create several different resistance levels. Critically, no current is required to maintain the bit(s) stored in one of these devices, as the crystalline/amorphous difference is constant.
The problem is that resetting the device requires enough current to partially melt the material. So, while the average power usage is low, it is quite high at critical points. This poses a challenge for devices that can be powered by a trickle of charge harvested from environmental sources. Therefore, making a phase-change memory out of a flexible material is not sufficient. You also have to match its performance to specific use cases for flexible devices.
Conveniently, part of the process of making it flexible also provided solutions to improve its performance.
make it flexible
A lot of flexible electronics are built on polymer substrates rather than rigid materials like silicon. In addition to being flexible, most polymers are insulators—they don’t conduct electricity or heat very well. And this phase-change proved to be key to increasing the efficiency of memory.
The gist of the new device is that the phase-change hardware is surrounded by materials that do not conduct heat well. This helps trap the heat needed to partially melt the appliance where it’s needed, meaning you don’t need to generate as much heat in the first place. And this in turn means that you need to apply less power to reset the device.
The device was made by drilling a hole in aluminum oxide. The hole was then filled with alternating layers of tin telluride and tin/gallium telluride, which acted as a phase-change material. The electrodes moved across aluminum oxide to wire the two ends of the device, and it was built on top of a flexible polymer material.
Modeling showed that the combination of aluminum oxide and polymer trapped heat in the hole where the phase-change material was located. This was confirmed by showing that the power requirements to reset the device decreased as the researchers increased the amount of aluminum oxide surrounding the device. At its best, the device’s power requirements were 100 times lower than those of existing devices fabricated on silicon substrates.
All this will be useless if the device does not work properly. But the researchers showed that it could be wrapped around a metal rod that is eight millimeters across and still functioned normally. The performance was similar after 200 bending and straightening cycles, and its storage stability was confirmed to be good for over 1,000 reads. Finally, multibit storage was demonstrated using different resistance levels. So, all in all, it looks like what you want from phase-change memory.
However, the researchers note that the basic principle here—reducing energy use by thermally insulating the material that stores the data—could also be used in more traditional rigid phase-change memory. And it may have some useful applications beyond memory, as other teams have shown that it is possible to train neural networks in phase-change memory instead of relying on repeated rounds of computation. The process is already more energy efficient than using conventional computers for the same task, so increasing the energy efficiency of a phase-change material could make it an even better option.
Science, 2021. DOI: 10.1126/science.abj1261 (About DOI).