Skip to content Skip to navigation

Imagine a “cool” data-storage technology that’s just a few atoms thick

Imagine a “cool” data-storage technology that’s just a few atoms thick

An experimental semiconductor material could store data in a new way that minimizes the generation of heat.
May 4, 2016
computer code

The electronics industry would love to create wireless, energy-efficient electronic devices so thin and flexible that they might be woven into the fabric of “smart” clothing. To this end, many researchers are looking at new semiconductor materials that can turn zeroes and ones into knowledge and entertainment far more efficiently than bulky silicon.

Now Stanford researchers offer fresh evidence that a semiconductor material known as molybdenum ditelluride could become the basis for ultrafast, energy-efficient data storage chips that would be just a few atoms thick.

What gives molybdenum ditelluride this potential is its structure: one layer of molybdenum atoms sandwiched between two sheets of tellurium atoms. This 3-atom thick material forms into a regular structure, or crystal, that can exist in two different states – one state that conducts electrons, and the other that doesn’t.

That makes molybdenum ditelluride a semiconductor that can store information – it can conduct electricity or not, to produce digital ones or zeroes – and it does this using far less material than today’s technologies.

But to make such a material useful, engineers must find a way to switch it back and forth between one and zero.

In new computer simulations published in Nature Communications, Yao Li, a graduate student in applied physics, and Evan Reed, an associate professor of materials science, describe how a tiny electric charge, delivered to a carefully designed molybdenum ditelluride chip, could toggle this ultrathin crystal between its zero and one states. This would initiate what researchers call a phase-change – but without relying on heat to initiate the change.

Other approaches to phase-change memory have relied on rapidly heating and cooling a semiconductor material in order to shift the crystal from a conductive to a non-conductive state. But dealing with heat changes can be a problem with high-density circuits.

The electrical charge mechanism proposed by Li could be a way to implement phase-change memory without the heat issues involved with other approaches.

The current favorite material in phase-change technology experiments is a semiconductor called GST. Li said GST memory storage materials are typically about 10 to 100 times thicker than molybdenum ditelluride. The thinness of the latter material suggests that an electric charge rather than heat could cause the structure of molybdenum ditelluride to flip between one and zero.

Whether researchers work with GST or molybdenum ditelluride, heat or an electric charge, making phase-change memory practical is worth the effort.

Li explains that phase-change memory technology could potentially store data about 100 times faster than the silicon-based flash memory currently used in smartphones, laptops, digital cameras and so on. “The next step is to build an experimental data storage device just a few atoms thick,” Li said.

The research team on this project includes Karel-Alexander Duerloo, who earned his PhD from Stanford working on this project, and Kerry Wauson, an undergraduate student at New Mexico State University.