Critical Materials Can Make or Break America’s Semiconductor Supply Chains
The United States must increase its domestic capacity for high-purity minerals, gases, and chemicals, or else American semiconductor capacity—and national security—will remain highly dependent on foreign factors outside of its own control.
In order to meet this forecasted increasing demand, Taiwan, Japan, and South Korea are in the process of expanding their industrial gas capacity. Major gas companies are upgrading their capacity throughout East Asia. Yet at the same time, China’s share of the gas markets is bound to grow through 2025 as it pursues semiconductor self-sufficiency, as outlined in 14th Five-Year Plan. This is concerning for Washington, and there are a few bright spots when it comes to U.S. gas capacity. For now, this capacity is bound to go up, since on-site gas supply facilities are generally built alongside major semiconductor fabrication facilities. For example, Linde is investing $600 million in gas plants in Arizona to supply chip manufacturer TSMC’s new facility.
Chemicals
Approximately five hundred chemicals are used in semiconductor fabrication, including photoresists, slurries, and wet chemicals like sulfuric acid. Also necessary are commodity chemicals such as isopropyl alcohol and hydrochloric acid. These chemicals have no alternatives given their unique properties. As one can guess, the need for these high-purity chemicals is driving an increasing demand, particularly for extreme ultraviolet (EUV) photoresists.
Once again, East Asia dominates in market share. One telling example is that Japanese firms such as Tokyo Ohka Kogyo Co., JSR Corp., and Shin-Etsu Chemical Co. control 90 percent of EUV photoresists, a critical chemical for leading node logic chips. China, as expected, is a major source of raw materials in wet chemicals, including hydrofluoric acid. On the other hand, the United States currently does not produce any high-purity chemicals for leading-node logic chips, and is not expanding its capacity for wet chemicals. As it stands, U.S.-based fabrication plants for advanced nodes source high-purity chemicals from overseas.
Chemical supplies are particularly at risk of disruption, due to low ultra-pure chemical production capacity. Experts predict a 37 percent price hike in U.S. demand for sulphuric acid, and 46 percent for ammonium hydroxide. TECHCET, an electronic materials advisory firm, has warned that “there may be shortages experienced by 2023.” The unique properties of these chemicals prevent fabrication plants from substituting alternative chemicals, and the different chemical formulations from each supplier inhibit plants from quickly switching to alternative suppliers. This, in addition to factory accidents and new regulations, has created a recipe for disaster.
Not that the preexisting situation was any better: chemical supplies are often sole-sourced from a single country or firm, inflating any associated geopolitical risks. Taiwan, for example, is a major source of ultra-high purity isopropyl alcohol. A China-Taiwan military conflict would obviously have a serious impact. Notably, Japan produces about 90 percent of fluorinated polyimide and 70 percent of hydrogen fluoride worldwide, and was the source for 90 percent of South Korean photoresist imports. When Japan restricted exports of photoresists, hydrogen fluoride, and fluorinated polyimides to South Korea in July 2019, this cost Korean semiconductor exports around $8.4 billion a month.
With a clear present and future need, South Korea and Taiwan are increasing investment to boost chemical production. Seoul has established a goal for domestic suppliers to provide 70 percent of electronic materials by 2023. Likewise, China is investing heavily in photoresist production, and Chinese photoresist companies are projected to experience massive growth in the next three to five years. Japan, for its part, is expected to dominate the semiconductor chemical market in the near future, as fab-supplier relationships are generally long-term.
As for the United States, Washington only faces challenges in growing chemical capacity: U.S . chemical plants, on average, cost five to six times more than Asian counterparts, while taking two to three years to build.
The Situation is Obvious: Immediate Action is Required
Washington must incentivize quality improvement and capacity growth for semiconductor materials to protect U.S. national security. Yet while bills like the CHIPS Act target research and fabrication, they painfully neglect minerals, gases, and chemicals. Even if current fabrication goals are achieved, the heavy reliance on East Asian markets continues to pose significant risks in the long run. Without sufficient materials, semiconductor fabrication is impossible. A White House report is clear on this matter: “A disruption in the supply of any of these materials could have far-reaching impacts on semiconductor production.”
The United States must increase its domestic capacity for high-purity minerals, gases, and chemicals, or else American semiconductor capacity—and national security—will remain highly dependent on foreign factors outside of its own control.
Shubham Dwivedi is a Fulbright Doctoral Fellow at Georgetown University’s Science, Technology, and International Affairs Program.
Gregory D. Wischer is a Master’s Candidate at Georgetown University’s Security Studies Program.
Image: Reuters.