Chipmakers Look To New Materials

Graphene, the wonder material rediscovered in 2004, and a host of other two-dimensional materials are gaining ground in manufacturing semiconductors as silicon’s usefulness begins to fade. And while there are a number of compounds in use already, such as gallium arsenide, gallium nitride, and silicon carbide, those materials generally are being confined to specific niche applications.

Transition metal dichalcogenides (TMDCs), a class of 2D materials derived from basic elements—principally tellurium, selenium, sulfur, and oxygen—are being widely explored by researchers for their use as semiconducting materials. These include molybdenum disulfide (MOS₂), molybdenum diselenide (MOSe₂), molybdenum ditelluride and molybdenum telluride (MOTe₂), tungsten disulfide (WS₂), and tungsten diselenide (WSe₂), which are among the materials being tested for use in chips.

TMDCs are functioning as semiconductors in conjunction with graphene (a carbon allotrope) as an electrical conductor, and monolayer hexagonal boron nitride (also known as white graphene) as an electrical insulator. These materials can be used in electronic devices, energy and harvesting devices, and for flexible and transparent substrates. TMDCs are also being combined with silicon substrates, to give good old silicon a few more years to shine. And 2D materials can be printed on paper substrates, opening up a whole new field of paper-based devices, such as sensors.

Monolayer graphene is highly conductive – overly so. The 2D carbon material has no bandgap, however, limiting its use in integrated circuits. Bilayer graphene, in contrast, can be tuned to have a bandgap. Bilayer graphene films on silicon carbide can be better controlled, researchers have found. Trilayer graphene can also be tunable to produce a bandgap, needed to develop field-effect transistors in a semiconductor device.

Fig. 1: Graphene. Source: Cambridge Nanosystems.

IDTechEx Research forecasts the graphene market will increase to more than $300 million in 2027, with 3,800-plus tons of graphene shipped in that year.

“The market will be segmented across many applications, reflecting the diverse properties of graphene,” says Khasha Ghaffarzadeh, research director at IDTechEx. “In general, we expect functional inks and coatings to reach the market earlier. This is a trend that we forecasted several years ago and is now observed in prototypes and small-volume applications. Indeed, IDTechEx Research projects that the market for functional inks and coatings will make up 21% of the market by 2018. Ultimately, however, energy storage and composites will grow to be the largest sectors, controlling 25% and 40% of the market in 2027, respectively.” (Graphene, 2D Materials and Carbon Nanotubes: Markets, Technologies and Opportunities, 2017-2027.)

IBM has touted carbon nanotubes as silicon’s successor for several years, and has reported on progress in developing transistors based on CNTs.

Fig. 1: CNT transistor with 40nm footprint. Source: IBM

“Graphene appears to be an excellent material for sensor applications due to properties such as each atom being exposed to the local environment, ultra-high electron mobility, heat conductivity, and the fact that it’s very strong and flexible,” says Timothy Schultz, project lead for microsystems technology at Robert Bosch. “In Europe, we’ve had graphene research going since 2012. We’re part of the European Graphene Flagship. Here in the U.S., among other things we’re doing, we’re actively working with the University of California at Berkeley and Stanford on graphene-related research.”

This isn’t always a straight line forward, though. “As with any new technology, it’s a bumpy road to industrialization,” Schultz says. “We see three key obstacles. Number one, manufacturing graphene at wafer scale is difficult, and this is due to getting good uniformity over a large area, low defect density, and possible contamination risks with bringing it into a standard semiconductor line. And, of course, when you talk about the manufacturing, among all the different processes or different ways there are for manufacturing graphene. Should we do direct versus transfer-free? Also, we require a strong entry sensor and application. A rule of thumb for us is it needs to be an order of magnitude better than whatever particular parameters are important in the state of the art. We need to identify a core application and get a strong return on investment. And then finally, it’s integrating the graphene into the device and maintaining graphene properties during manufacturing. We may end up having special requirements on the substrate, on the subsequent layers, and special requirements on how to contact the graphene.”

Rahul Raut, director of strategy and technology acquisition at Alpha Assembly Solutions, noted his company has been working with MacDermid Performance Solutions and with the National Graphene Institute at the University of Manchester in the United Kingdom. “Currently, one major challenge is the ability to produce high-quality graphene engineered or tailored for specific applications, in a scalable high-volume-manufacturing method,” says Raut. “Alpha’s work with NGI has led to the development of a HVM-capable synthesis method that engineers large-size graphene flakes which, when made from this process, possess a unique combination of properties that make them suitable for use in electronic materials and other related applications.”

Other markets
Graphene’s use goes well beyond just electronics because of its conductivity capabilities. For example, the U.K.’s Haydale Ltd., which focuses on nanomaterials treatment, is working with a Korean company to incorporate graphene into cookware, according to Haydale CEO Ray Gibbs. It also is working with Huntsman Advanced Materials, which is designing materials such as boat hulls to structural adhesives for creating high-strength, low-weight automobile bodies.

“In parts, less is more,” Gibbs says. “That’s the key thing in nanomaterials.”

Felice Torrisi, a lecturer at the University of Cambridge’s Graphene Centre, pointed to graphene and 2D materials being used in inks for smart fabrics and wearable electronics, as well. The university’s researchers were able to print boron nitride, graphene, and silver on textiles to form two transistors – fully flexible, not rigid. “We can also spray-coat these devices on 3D surfaces,” he says.

Thomas Swan & Co., meanwhile, is looking at graphene applications in epoxy resin coatings because of its thermal conductivity, and is prototyping graphene-based inks. “We started working with graphene in 2013,” says Dimitris Presvytis, the company’s advanced materials research leader, and has since focused on industrial applications of graphene and boron nitride. “We have high hopes for this. It could find a lot of potential applications—even automotive or aerospace.”

The U.K.’s Centre for Process Innovation – Graphene Application Centre has worked with industry to integrate graphene into coatings, composites, inks, membranes, and sensors, as well. “Many of the claimed properties of graphene can be realized cost-effectively,” says Tom Taylor, director of new business.

The CPI can help scale up engineering designs in various electronics technologies, according to Taylor. “We have pilot-scale equipment that we make available for companies, investors, to use. The idea, the model is, you’ve got something new you want to do that you’re currently not set up to do. Before making that capital investment, you can come to us. You’re effectively paying to rent equipment by the day. It’s all about de-risking innovation. The economic model is you pay for a day’s work, rather than spend a few million pounds on equipment. The area we’ve homed in on is the graphene space – areas that we think offer economic advantage.”

Setting standards
Standards are beginning to show up in this area, as well. Terrance Barken, executive director of The Graphene Council in the U.S., says his organization is working on standards for graphene.

“We think the future of it is going to be graphene combined with other technologies, other materials, not necessarily on its own,” says Barken. “The challenge is, how do you bring all these disparate parties together to accelerate the development process and to accelerate the cross-fertilization of ideas, so that actual applications get developed faster. Strictly speaking, graphene is a single layer of carbon atoms, if that’s the purest form, but there are many forms of graphene actually. You have bulk graphene. This is multilayer graphene, which might be three, four, five, six, seven layers. There’s even material on the market that might be 50 layers or 100 layers of carbon, that is being labeled as graphene, which it clearly is not.”

There need to be baseline definitions for all of these materials. “For an industrial application, for people to use a material on a regular basis, it needs to be reproduced, and for industrial use, you also need usually more than one source of supply,” Barken says. “If you’re going to have more than one source of supply, those suppliers need to be supplying the same material so you can use it in your industrial processes. Hence, you need to have some kind of standardization process, at least to some extent. What’s important is that there is now an ISO standard organization agreed definition of what graphene is and what graphene is not. And the definition is that material that includes and up to 10 layers of carbon can be considered a graphene material. And if it’s 11 layers of carbon or more, it’s considered micro graphite. That doesn’t mean that material that is more than 10 layers is useless. It is material that’s still useful, but we shouldn’t be calling it graphene because it just confuses the market and it’s not accurate.”

Conclusion
Silicon is still going strong, but it is getting more difficult to extend silicon into some new markets as well as new process nodes. CMOS has been running out of steam for several nodes, and that becomes more obvious at each new node.

As companies such as Lam Research and Applied Materials have been saying for some time, the future will depend on new materials. That opens the door for a variety of new options, including graphene, 2D materials, and carbon nanotubes. Some of these materials will be used on their own, while others will be combined with silicon to extend Moore’s Law and improve performance, power and area far beyond where it is today.

Related Stories
Ruthenium Liners Give Way To Ruthenium Lines
Materials to replace copper at 5nm and beyond are starting to come into focus.
Materials For Future Electronics
Flexible electronics, new memory types, and neuromorphic computing dominate research.