Researchers Build a Transistor Out Of a Single Electron

Illustration for article titled Researchers Build a Transistor Out Of a Single Electron

University of Pittsburgh researchers have assembled a key piece of tech that will help enable a future generation of extremely powerful quantum computers as well as advanced electronic materials and better computer memories. Their single-electron transistor is the first of its kind made entirely from oxide-based materials, an important aspect that allows it to work as a solid-state memory.


SketchSET—or sketch-based single-electron transistor—contains a tiny 1.5-nanometer-wide island at its core that operates with just one or two electrons at a time. The ability to work at such small scales makes it ideal for several advanced computing applications, like quantum processors that would be orders of magnitude more powerful than existing supercomputers.

The tiny transistors could also enable ultradense memory that packs far more information into far less space than existing memory.


But the subatomic possibilities don't end there. That tiny central island at the heart of the technology could also be used as an artificial atom, the researchers say. These in turn could be used to develop new classes of electronic materials, like superconductors with exotic properties not see in natural materials.

But the real upside here is in the intended application: computing. Because the transistors are oxide-based, they possess a ferroelectric property that allows them to act as solid-state memory. That means that even when they are powered down they can control the number of electrons on the island. The number of electrons dictate the ones and zeros that are the basis of computer memory, meaning that a future computer based on such transistors could hold its data even without external power. [Science Daily]

Share This Story

Get our newsletter


I am not sure how many people understand the implication this has. What this article doesn't describe are the limitations of modern semiconductors as a transistor medium. The current materials being used have their physical limitations which companies like Intel are rapidly approaching. At their current rate we will meet the limitations (channels being 1 atom wide), within the next decade which would stagnate the progression of computing potential to something not much better than where it is today. With a channel being 1 electron the future computer would be exponentially better than where we're at today.

As a side note, with this research being done, I wish I had spent more of my grad work focusing on on nano-tech rather than RF communications. Both fields are incredibly interesting and I encourage people to take a class in electrical engineering (or materials in this case) to learn a little and see if it's something they'd enjoy doing. This research is a key example of why we need more engineers and why it's actually fun to be an engineer.