Wonder Material?

The fact that new materials continue to be discovered and introduced into innovative electronic devices is a major feat. Semico is analyzing the challenges of new product discovery, development and commercialization. Graphene is just one of the products under investigation. This write-up is taken from a full essay in this month’s Semico IPI report.

Since 2004, when two scientists at Manchester University, Andre Geim and Konstantin Novoselov, isolated a single sheet of graphene and then performed electrical measurements and characterizations on the material, graphene has become the superstar of the material world. Today, I Googled graphene, and got 11,500,000 results, including definitions, companies, recent research findings, images and in-depth articles. And the patents are piling up. Samsung was reported as one of the leaders in this area with more than 200 patents.

Graphene isn’t the only 2D material, but it’s definitely the media star. The benefit of all the attention has been an influx of research activity and dollars. But as the application research gets underway, has graphene lost some of its luster? In the August 2014 issue of the Semico IPI Report, we reviewed all the positive characteristics and benefits of graphene. From a public relations perspective, 2014 was a good year for graphene. Now it’s time to face reality. When will all these new graphene products be on the market? When will graphene make an impact on the semiconductor industry? The answer is, probably not real soon. That’s not to say progress isn’t being made. But how long will investors be enamored by the promises of graphene?

In December 2014 at the IEEE IEDM meeting, Imec presented a paper that demonstrated the first integrated optical modulator utilizing graphene. The graphene electro-absorption modulator (EAM) is capable of 10Gb/s modulation speeds. Integrated optical modulators exhibiting high speed, small footprint and broadband athermal operation are in high demand for future chip-level interconnects. Although 10Gb/s is far from the speed capabilities of silicon (50Gb/s) or SiGe (28Gb/s), this breakthrough is significant for a couple of reasons. First, this proves that graphene can result in a working product that exhibits low insertion loss, low drive voltage, high thermal stability, and broadband operation. All this is obtained within a compact footprint. Second, a working product now enables continued modeling and further development to optimize the device and improve manufacturing techniques and efficiencies.

In the early stages of research, each development step forward usually uncovers even more challenges. So if graphene is not ready for commercial chip production, when will it be? Semico has compiled a list of hurdles yet to be cleared.

Usually the first question that arises with the commercialization of a new material is the cost. How does this material compare to other materials being used today? Graphene isn’t even close to being stacked up against existing materials on cost. At this point, cost is not a major concern, because it is expensive but affordable enough for research purposes. If production volumes can be increased, the basic principles of economics will take over, and prices should fall. There are more than 70 producers and suppliers of graphene materials, but there is no single company that can ramp production to levels required for semiconductor production. Volume production is still a challenge. The material itself is still in a developmental phase and can be confusing even to players in the supply chain. Graphene comes in many different forms, can be manufactured in a variety of processes, and currently is being developed for a number of applications in all areas. Very few graphene producers understand the semiconductor manufacturing process or where graphene would fit into the process. Graphene is being tested for paints and lubricants all the way up to 3D printers, display materials, batteries and electronic devices. A significant amount of education and communication needs to occur before the material is segmented into categories that reflect quality, form, delivery and appropriate price for different applications. Standards need to be developed for bringing graphene into the fab environment, contamination being only one of the negative effects to be avoided.

Material properties and interaction are, of course, still in the discovery phase. Imec and others already have found that there is a tradeoff between material quality and its beneficial properties. Higher quality graphene removes metals that can interfere with graphene’s contact resistance. In order to counteract the loss of resistance, more steps must be added to the process. In order to lower cost, graphene is combined with other materials. Other materials can be sandwiched around, in between and on top of graphene. The resulting interactions between these adjacent materials are numerous.

I once heard a new memory technologist describe the interest level in each new memory technology as a rollercoaster ride. The enthusiasm for new technologies like graphene rides a two-year rollercoaster that cycles through peaks of over-exuberance, followed by valleys of disillusionment. Graphene may not yet be in a valley of disillusionment, but I suspect it will go through several cycles of exuberance and disillusionment before we see a real application in the electronics world.

For more on the list of graphene challenges, which includes a discussion on process equipment, in-depth material pricing breakouts, media hype, and next steps, check out Semico’s January IPI Report or contact [email protected].