Tiny technology leads to big data breakthrough for Edmonton scientists

Researchers at the University of Alberta have mastered the art of writing computer memory at the atomic level, a new technology which exceeds the capabilities of current hard drives by 1,000 times.

'Five years ago, this wasn't even something we thought possible'

Roshan Achal, left, and his PhD supervisor Robert Wolkow, a University of Alberta professor and leader in the field of nanotechnology research. (John Ulan for the University of Alberta)

Researchers at the University of Alberta have mastered the art of writing computer memory at the atomic level, a new technology which exceeds the capabilities of current hard drives by 1,000 times.

Using nanotechnology, the team of Edmonton scientists have designed the most dense, solid-state computer memory ever created.

"Essentially, you can take all 45 million songs on iTunes and store them on the surface of one quarter," said Roshan Achal, a PhD student in Department of Physics at the University of Alberta and lead author on a new research study published in Nature Communications. 

"Five years ago, this wasn't even something we thought possible."

The technology works at the atomic level and allows for the writing and rewriting of computer data using hydrogen atoms.

The design is perfect to deal with the deluge of information in a data-driven society, Achal said. 

Other similar technologies have been developed but only work in very cold, cryogenic conditions. These memory chips, however, are more stable and can withstand temperatures that are room temperature or higher. 

Atoms in binary code

"We really have automated this process to make it faster and more reliable to start building these perfect structures," said Achal whose work is supervised by U of A professor Robert Wolkow, a pioneer in nanotechnology. 

To create the memory chips, researchers insert a tiny wafer of silicon into what is called a scanning tunnelling microscope, and coat the wafer with hydrogen atoms.

You are able to rip that atom right out of there.-   Roshan   Achal

By plucking atom after atom off the surface with a tool similar to a pair of tweezers, they encode the silicon with data binary code, a language of ones and zeros used by computers.

They're building atomic structures with 100 per cent accuracy and memories with a density of 138 terabytes per square inch, Achal said.

"You apply a little bit of voltage, similar to a battery voltage, and you are able to rip that atom right out of there," Achal said. 

"This allows us to write by removing the atoms, or rewrite by putting atoms back and removing different atoms." 

Rewriting memory

The "ah-ha" moment for Achal and his team came when a fellow graduate student "accidentally" realized that atoms could be replaced and manipulated with the microscope.

That led to the "fun idea" of creating memory, Achal said — but it would be about a year before they perfected the technique. 

He anticipates it will be another decade before the process is ready for the commercial market.

It's still a bit too slow and too cumbersome for individual users, Achal said.   

"To move atoms, it requires a very precise motor but the trade-off you make for that precision is speed. You can only move so fast with a motor that precise.

"But there are shortcuts people can make and I know there is active research into these types of motors, making them faster and more applicable." 

When the time comes, their memory work could help archive massive amounts of internet data,  Achal said.  

A single silicon wafer could store all published information on sites like Wikipedia, Achal said. 

"We're basically laying the foundation for future memory applications." 

The paper, titled "Lithography for robust and editable atomic-scale silicon devices and memories", appears in the current issue of Nature Communications.