Shaker Sadasivam, MEMC Vice President of R&D

Shaker Sadasivam (Source: MEMC)

Shaker Sadasivam is senior vice president of R&D at MEMC Electronic Materials (St. Peters, Mo). He has been with the company since 1993, and served as president of MEMC Japan, director of U.S. technology, corporate director for operations technology, and director of R&D for MEMC Korea. Sadasivam has both a Ph.D. and M.S. in chemical engineering, in addition to an M.S. in business administration. MEMC Electronic Materials is the world's largest public company solely devoted to the supply of wafers to semiconductor device manufacturers.

SI: From your perspective, has the CTO's role changed over the last five years or so?

Sadasivam: Viewing it from the lifetime of plain planar CMOS and what is emerging now in the way of alternative transistor design and device requirements, as well as new materials, yes, it has changed significantly. Previously, one primarily dealt with larger wafer diameters, mainly mechanical product attributes, and crystal defectivity — everything related to bulk silicon. Most of the action took place in the fabs, with nearly nothing happening in the materials sector other than to continue improving parametric performance by 30% each time a node shrank. The explosion in diversification of both product and applications has changed all that. We're now doing this at a rate that would have seemed impossible five years ago.

SI: So the requirements have increased?

Sadasivam: Most definitely! Now you must contend with many requirements, plus determine which are going to be important going into the future. What was primarily development and process technology is now process technology and product technology. We must continue working on process technology while making improvements in bulk silicon requirements, which are also important for new materials such as SOI, strained silicon, or any of the other new ones that are coming along because many of them will be built on silicon.

Then you have the traditional dimension of product technology, which is mainly aimed at mobility enhancement and power reduction. Even in that space, the complexity of what customers require has also increased. Before, when they specified flatness or particles, for example, they referred to only one number and the requirement was met by the measurement tool that you had in place. Now, you have various and specific definitions of what planar is, and you must track what exactly happens at the edge and how do you characterize this. And it is different for everyone: Some want a wafer that is flat all the way to the edge, others one that isn't. Even within the ITRS, what used to be a simple number has grown in complexity, and on top of everything you have new materials coming in whose parameters we are yet to define. All this growth in complexity has changed the role of the CTO. Also, now one must understand the product technology that is out there, since necessarily not all the technologies are being developed in-house by silicon wafer suppliers. Licensing of those technologies becomes important, as does choosing from among all these alternative technologies.

SI: Once you establish your technology roadmap, do you just fine-tune it or do you sometimes have to make major changes to it?

Sadavisam: Well (laughing), the famous — or infamous — red brick wall keeps being pushed out further, which is good news. However, the industry's innovativeness, and how it has managed to come up with solutions, can never be underestimated. We've found ways to do what was once thought could only be done in SOI; people have been working on process strain and other tricks within the fab. Here is where the fine-tuning takes place: When exactly does a given process solution intersect a technology node? We always search for ways to enable the fast adoption of new materials solutions.

SI: You mentioned the notorious red bricks. What's your opinion of the ITRS?

Sadasivam: As a pointer to general directions or general trends in technology, it's useful. But you must keep in mind that it is an average. We have customers who are well ahead of it in terms of being on the leading edge, and there are others who trail it.

SI: What is your R&D philosophy?

Sadavisam: The R&D effort must be set up so as to anticipate and meet customers' problems. There are three things that, as a wafer supplier, we can do for our customers. One is to provide them with products that can reduce their costs. The second is to provide products that enhance performance. The third is asset life extension. Our R&D activities are aimed at understanding the commonality of problems the industry faces across its various segments. This enables our solutions to fit one or more categories within the customers' timelines.

SI: Nanotechnology is on the horizon. How do you view this as a supplier?

Sadasivam: From a process perspective, we're looking at increased complexity. The same thing holds for measurement. All aspects of production will require a steep redefinition in order for us — for anyone — to provide the needed solutions and then be able to measure and quantify them.

SI: Is there any challenge connected with this that you consider as especially major?

Sadavisam: Measurement, without a doubt. New definitions of what measurement will be needed before we can tackle developing processes that will meet these new, more exacting requirements. Next comes developing a process technology that will meet those requirements.

SI: Is coming up with these solutions becoming more difficult?

Sadavisam: There can be no doubt of that — at all levels. In the case of planarity, for example, we used to talk about 3 mm edge exclusion for flatness. Now people want to go to 2 and soon 1 mm edge exclusion. You now also have added components besides flatness; nanotopography is certainly one of these. When you consider particles at the small levels being looked at, how you monitor the potential sources of these particles and control them really becomes a challenge.

SI: From a silicon wafer supplier's viewpoint, what do you consider as the device manufacturer's greatest challenge over the next five years?

Sadasivam: The limitations of silicon. As we continue scaling, silicon is hitting its limits in terms of mobility. Therefore, managing performance and power is, to me, the biggest challenge that the design engineer faces.

SI: Partnering seems to have become a key concept in the industry. Do you see this continuing?

Sadavisam: Most certainly! In our case, mainly because of the integration challenges associated with all the new materials that are coming in and being considered. Silicon wafer requirements — flatness, for example — are primarily driven by lithography's needs. At arm's length, you could probably provide what the customer requires, and he would most likely be satisfied with what he got. This is no longer possible, because now there are subtle interactions with more process steps than with just lithography. Many of these interactions weren't important before, but now must be fully understood. Subtle variations in material properties and their interactions with wafer mechanic parametrics are playing a much more significant role than they ever did before.

SI: Are there any industry trends that we ought to be paying more attention to?

Sadasivam: From our perspective, consolidation is the biggest, most important trend taking place in the industry. Before 1990, there were some 20 silicon wafer suppliers for what was then the primary 150 mm market. By the mid-1990s, there were probably half that many supplying 200 mm. Now, with 300 mm, there are four. Besides wafer diameter, which was the main consideration, there has been a diversification of key parameters — so not only is the process technology required to meet advanced requirements increasing in terms of complexity, technology and resources, but you have the added component of new products and new materials. All this will probably accelerate consolidation even more. Another major trend, which has been going on now for some time, is that customers look for full-service suppliers. They want to work with suppliers who will provide everything, including a portfolio of solutions.