Graphene has often been considered a kind of “miracle material.”

The nanomaterial’s honeycombed structure consists of a single layer of carbon atoms, so tightly bonded that the graphene has a greater-than-steel strength.

Additionally, graphene is an exceptional conductor, due to its ability to scatter electrons over great distances.

The electron-carrying properties make the nanomaterial an interesting option for nanoscale electronics – except the nanomaterial has always been difficult to pattern. With dimensions below 100 nm, attempts to etch the graphene often ruin its properties.

A team of researchers from the Technical University of Denmark (DTU), however, has found a way to solve the problem, and potentially advance new kinds of graphene-based nanoelectronics.

The results are published in Nature Nanotechnology.

Bjarke Jessen and Lene Gammelgaard, two postdoc students from DTU Physics, first encapsulated graphene inside hexagonal boron nitride — a non-conductive 2D material that is often employed to protect graphene’s properties.

Using electron beam lithography, the team carefully patterned the protective layer of boron nitride (and graphene underneath) with a dense array of ultra-small holes – each with a diameter of approximately 20 nanometers, each spaced just 12 nanometers apart. The arrangement allows 1000 times more electrical current to flow than had been reported in such small graphene structures.

Additionally, the team was able to induce a band gap — a crucial step when making transistors and optoelectronic devices.

“We have shown that we can control graphene’s band structure and design how it should behave,” said Peter Bøggild, Professor at DTU Physics. “When we control the band structure, we have access to all of graphene’s properties — and we found to our surprise that some of the most subtle quantum electronic effects survive the dense patterning.”

Prof. Bøggild spoke with Tech Briefs about his team’s achievements – and other graphene surprises that could lead to new kinds of electronics.

Tech Briefs: In your mind, what makes this work a ‘breakthrough?

Prof. Peter Bøggild, Technical University of Denmark: One of graphene's most appealing properties is how much you can modify, change, tune and improve the electronic and optical properties by doing "something" to it. Every atom is exposed to the environment, and while this can be problematic, it is also a huge opportunity.

One of the most obvious things to do is to pattern it. We are extremely good at patterning, in part thanks to the silicon chip industry and nanotechnology research, so it should be an easy job. But as it turns out, it isn't. When you carve graphene out in narrow lanes, what should happen is that the electrons start to behave in different, very specific ways — depending on the pattern. The so-called "band structure" can be engineered through patterning: nanolithography. But when we and others have tried to do this, we end up with just bad graphene devices.

What we have done, building on other milestone works from German and U.S. research groups over the past five years, is to show that the spell can be lifted: that it is possible to carve out graphene down to the 10-nanometer scale, and get exactly the behavior that theorists have told us it should behave like.

So, the breakthrough, as we see it, lies in putting back nanolithography — the most powerful and versatile method of creating nanostructured materials — in the toolbox. I almost gave up believing this was possible 4 years ago, but some really good postdocs persuaded me that it could be done. And they were right.

An illustration of graphene inside hexagonal boron nitride, made by the Technical University of Denmark (DTU)
An illustration: Graphene first buried in hexagonal borin nitride and then patterned through the top layer. Careful etching of the holes keeps the edges of the holes smooth, allowing the current to flow freely between the holes (Image Credit: DTU)

Tech Briefs: Why has patterning been so challenging on a material like graphene?

Prof. Bøggild: It turns out that the transport of electrons through narrow passages is extremely sensitive to roughness at the edges, as well as contamination — both things that are hard to avoid with the ways that lithography is normally done. When creating a 100-nanometer-wide channel with 2-3 nm roughness, the electrons do not flow easily through — and this also blurs all the sophisticated effects that more elaborate patterns can create.

Groups at Columbia University and at University of Manchester in UK started to use hexagonal boron nitride for encapsulating graphene, and that has changed everything. When encapsulated in this amazing material, graphene suddenly starts to behave like the theorists want it to. A group at Regensburg University in Germany showed that patterns that are drilled into the graphene, but through the hexagonal boron nitride, are more likely to leave the electrons undisturbed, and it is that concept that we have pushed to the extreme.

Tech Briefs: What kinds of future forms of electronics are possible now, because of this research?

Prof. Bøggild: The technique is not perfected — and I cannot say for sure that it will ever be — but our measurements indicate that we have created a bandgap that behaves just like it should. A bandgap is needed to make transistors. The easy answer would be "Now we can make transistors, because we can give graphene a bandgap." But that would be misleading.

I am not sure that nanopatterned graphene transistors will ever be a technology. Silicon and III-V HEMT work great, and I don’t see "holey graphene" able to catch up. But for other types of electronics, such as spintronics and valleytronics that can deliver faster, ultralow power computation perhaps, exact control of the shape will greatly enhance our options for both miniaturizing and customizing the functionality of next-generation electronics.

I am more interested in making a game-changing technology in 10-20 years than trying to beat an existing technology by 20% now. The European project Graphene Flagship just demonstrated room-temperature spintronics in another 2D material, MoS2 (molybdenum disulfide), and it will be extremely exciting to see if we can pull of the same trick with this also very interesting material.

Tech Briefs: Why is nanolithography in graphene such an important task to try to achieve?

Prof. Bøggild: Nanolithography is the heart of modern-day electronics, and I don’t think graphene has a chance to fulfill its potential as a future electronics platform if we can’t even pattern it. So, clearing this roadblock, in our view, is a great relief, and we hope that the community will take up the ball.

Tech Briefs: What is special about graphene and its properties?

Prof. Bøggild: Graphene has the highest mobility of any material, but we have seen recently that this is actually just the tip of the iceberg. A 2D material is a very strange and unusual thing. Squeezing two crystals of graphene — which are essentially pencil dust — and making just the right twist of 1 degree, between the two layers, makes the two layers become a superconductor. That rocked our community at the foundation, because such behavior is entirely unheard of in solid-state physics. And that, I believe, is just one example of the strange property of graphene, that it is bulk material and surface material at the same time.

Graphene can be stacked with atomic precision, together with thousands of other 2D materials, giving us a degree of control of matter that is crazy. Not only can we make up materials with properties that can never, ever be realized by the usual alloys; we can also fabricate them in practice. At the moment, it is still laboratory scale, but we are working hard in the Graphene Flagship and our center of excellence (Center for Nanostructured Graphene) to create these stacked materials on a large scale. Once we learn to master this technique, we have control of the material in the vertical direction (by layering specific sequences of 2D materials into new artificial materials), and in the lateral dimension, through techniques such as the one we have demonstrated in our paper.

What do you think? Write your comments and questions below.