Written by two members of the consortium's supplier relations division, the study traces potential problems to "latent supply-chain issues." Neil Gayle and Dave Anderson insist that the semiconductor industry can eliminate or minimize these threats by systematically analyzing the supply chain and taking preemptive actions.

International Sematech's Critical Materials Council oversees work to analyze such risks and draw up contingency plans, the authors note. The council was established in 1994 after an explosion and fire at a Sumitomo chemical plant in Japan greatly reduced the industry's supply of important epoxy cresol novalac resin. Sumitomo was the source of more than half of the world's supply, the study says. The resin is used in device molding.

The authors cite examples of other supply shortages. Competition from the electronics, power plant, and aerospace industries for scarce tantalum supplies caused a scarcity and a price spike in 2000, for instance. Chipmakers had to scramble to find supplies of the metal, which is used in interconnect barrier metallization.

Unstable international politics are another potential pitfall, Gayle and Anderson contend. The Democratic Republic of Congo is a major source of cobalt, which is used as a transistor gate electrode. In another example, the authors say that Russia is the most likely future source of growing demand for the helium used in wafer cooling.

An overabundance of suppliers for certain materials can also threaten availability, the authors emphasize. The competition can lead to drops in profit margins and an eventual shakeout of the market. Silicon wafer suppliers are exhibit number one in this regard. On the other hand, too few suppliers can lead to monopolies that drive up prices and, ultimately, manufacturing costs.

By sponsoring industry workshops, Sematech's Critical Materials Council has brought together suppliers of critical materials and member companies to discuss mutual concerns in a confidential setting, the study points out. The meetings have elucidated issues such as the expectations of the chipmakers and forecasting issues. The workshops enabled member companies to react better to the hydroxylamine explosion at a Nisshin Chemical plant in 2000, the authors write.

The authors enumerate at least four future concerns for materials and the supply chain. One, protecting intellectual property rights will become increasingly important, especially for small firms, as suppliers form alliances for sharing the cost of research. Two, Europe's stringent new environmental regulations may make it harder for materials suppliers to conduct business. Three, China's growth could worsen global chipmaking overcapacity. Four, further consolidation is expected in certain industry segments, with silicon wafer companies again the prime example.

The authors predict that new "specialty products" will appear as processes reach their technical limits. However, they say the long-term effect of this shift is unknown right now. "In particular, the impact on semiconductor productivity and the growth of the industry needs to be understood."