èƵ

Liquid metal particles can self-assemble into electronics

A cheap method for forming the tiny components of chips, such as transistors and diodes, harnesses simple fluid physics to make the electronics self-assembling
A crosshatch pattern of wires created by self-assembling liquid metal particles
Julia Chang / North Carolina State University

Self-assembling electronics made from liquid metal particles could provide a cheaper way of manufacturing computer chips, simply by harnessing the basic physics of how fluids flow through tiny structures.

“The cost of entry in manufacturing electronics and building new chip fabrication plants in the US right now, we’re talking billions of dollars,” says at North Carolina State University. “It’s not cheap.”

Thuo and his colleagues first created a mixture of carbon, oxygen and metallic atoms by using an organic solution to extract charged ions from an alloy of indium, bismuth and tin. This fluid then flowed into moulds made up of tiny channels arranged in patterns. The molecules took on the shape and size of each mould, forming complex 3D structures.

Once the structures had fully assembled, the researchers removed the moulds and applied heat. This set off chemical reactions that combined the molecules into different materials. The final structures consisted of semiconducting metal oxides wrapped in sheets of carbon-based graphene, perfect building blocks for electronics.

The researchers showed that they could use this technique to form a variety of shapes, including rows of wires overlaid one on top of the other to form a crosshatch pattern. They then demonstrated how such shapes could function as transistors and diodes, the tiny electrical switch components within chips. The moulds could also form more complex shapes at a micrometre or even nanometre scale. Thuo says the team should be able to scale up this approach to build even larger and more complex components within chips.

Current semiconductor manufacturing uses high-powered lasers to etch interconnected patterns in silicon chip layers, a process that typically requires billion-dollar factories, called fabs, packed with expensive equipment. By comparison, the self-assembling electronics approach is only constrained by the price of the moulds and the cost of labour.

“If we automate it, which is something that we are exploring, where we use robotic arms to do it, then that cost drops to almost nothing,” says Thuo.

Some chip designs that require multilayer construction will still need to be made using a different process, says at the University of Michigan, who praised the simplicity of this self-assembling process and described it as inherently scalable and elegant.

“We need a wide range of chips,” says Kotov. “This method opens the door for high-quality semiconductors for potentially mass applications as biosensors.”

The researchers are seeking two separate patents based on this process and are pursuing partnerships with semiconductor manufacturing companies.

Thuo envisions a future where people can set up chip manufacturing lines in a more distributed fashion, without a huge, centralised factory. “Instead of being concentrated in one location, like Silicon Valley or here in [North Carolina’s] Research Triangle Park, you can have people all over the country helping you put that stuff together,” he says.

Journal reference

Materials Horizons

Topics: Computing / Electronics / Materials science