The White House prepares a hybrid price floor and tariff regime on the raw material behind every chip and solar panel, aiming at China’s 93.5 percent grip on global supply while U.S. solar developers brace for higher costs
WASHINGTON, August 6, 2026
The Trump administration is preparing to impose a combination of minimum import prices and tariffs on polysilicon and its downstream products, according to reporting by Reuters this week, placing the ultra-pure material that sits at the base of both the semiconductor and solar supply chains at the center of Washington’s economic contest with Beijing.
The action, expected in a presidential proclamation as soon as this month, would conclude a year-long national security investigation into polysilicon imports conducted by the Commerce Department under Section 232 of the Trade Expansion Act of 1962. Four people familiar with the plan told Reuters that the administration intends to pursue a hybrid system pairing a price floor with tariffs on polysilicon and derivative products such as wafers, cells, and modules, an intervention that would reach deeper into commodity markets than nearly any prior action in the current tariff program.
The stated objective is to protect the last two American polysilicon producers, Hemlock Semiconductor of Michigan and Wacker Chemie’s Tennessee operation, from a Chinese industry that produced 93.5 percent of the world’s supply in 2024 and is now pushing into the higher-value semiconductor-grade segment. The cost of that protection, at least initially, will be borne by the U.S. solar developers and chip manufacturers that currently buy inexpensive imported material.
Markets rendered a swift verdict on who wins first. Shares of Corning, which co-owns Hemlock with Japan’s Shin-Etsu Handotai, closed up 9.4 percent on the day of the Reuters report. First Solar rose 4.7 percent, T1 Energy climbed as much as 9.9 percent, Canadian Solar gained 5.6 percent, and Japan’s Toyo jumped as much as 5 percent before settling up 1.7 percent.
The Material at the Bottom of Two Supply Chains
Polysilicon is an ultra-refined form of silicon from which manufacturers slice the wafers that become both solar cells and semiconductor chips. It occupies a chokepoint position with few parallels in industrial policy: a single commodity that feeds the two technologies, energy and computing, that Washington has declared central to national power.
The purity requirements diverge sharply by application. Solar-grade polysilicon requires roughly six to nine nines of purity, while semiconductor-grade material must reach nine to eleven nines, a differential that commands a price premium of three to four times and requires additional purification cycles. Hemlock and Wacker both produce semiconductor-grade material, and preserving that domestic capability is, from a chip-supply perspective, the more strategically sensitive of the two goals the action serves. The Semiconductor Industry Association notes that chipmaking accounts for only about 2.4 percent of global polysilicon demand, which means the solar industry’s much larger appetite effectively underwrites the production base the chip industry depends on.
“They’re finding that they actually do need solar,” Rhone Resch, chief strategy officer at Toyo, which operates a panel factory near Houston and plans a 357 million dollar solar cell facility nearby, told Reuters.
China’s dominance of the material was built deliberately. Beijing provided extensive state support to establish its producers as the world’s dominant suppliers, and cheap hydropower in provinces such as Sichuan gives Chinese plants electricity at roughly two cents per kilowatt-hour in the rainy season, a fraction of typical U.S. industrial rates. That matters enormously because the standard Siemens production process consumes 50 to 100 kilowatt-hours of electricity per kilogram of output. The result was a price collapse from about 39 dollars per kilogram in 2022 to below 4.50 dollars by the end of 2024, which erased margins across the non-Chinese industry.
American producers bear scars from an earlier round of this fight. After China imposed retaliatory anti-dumping duties of up to 57 percent on U.S.-made polysilicon in the 2010s, combined U.S. revenue from the material fell from roughly 1 billion dollars to 107 million dollars between 2011 and 2018, and REC Silicon eventually ceased production at its Moses Lake, Washington facility. Beijing extended those duties for another five years beginning January 14, 2026, keeping the wall around its home market firmly in place even as Washington prepares countermeasures.
What the Proclamation Is Expected to Do
The precise structure of the action has not been announced, and a White House official declined to comment ahead of the president’s decision. The Commerce Department did not respond to requests for comment. But the contours described by people familiar with the deliberations combine three elements.
First, a minimum import price would set a floor below which imported polysilicon cannot be sold in the U.S. market. A price floor differs from a tariff in that it directly targets dumping: rather than adding a percentage at the border, it makes it illegal to sell below the floor, which neutralizes the strategy of pricing below cost to drive competitors out of business. Craig Singleton, senior fellow at the Foundation for Defense of Democracies, told Reuters that “China built its polysilicon dominance through subsidies and chronic overcapacity that pushed prices below sustainable levels and weakened competitors.” He added that “a price floor paired with tariffs could give U.S. producers room to survive, but the policy needs careful calibration so trusted inputs remain available while domestic wafer and cell capacity catches up.”
Second, tariffs would apply to polysilicon and to derivative products, reaching down the value chain into wafers and cells. Industry reporting in Korea and the trade press suggests derivative products could face duties in the neighborhood of 15 percent, though the final rates rest with the president.
Third, and potentially most consequential for investment decisions, two of Reuters’ sources said the plan will allow importers that are investing in U.S. wafer and cell production to offset costs associated with the trade protections. That offset mechanism echoes the design of the January semiconductor action, which exempted chips imported to support the buildout of U.S. manufacturing capacity, and it would convert the tariff from a pure penalty into a lever steering capital toward domestic facilities.
China’s government has already objected. “China urges the U.S. to stop the Section 232 tariff measures as soon as possible, and properly resolve the concerns of all parties through equal dialogue,” a spokesperson for China’s embassy in Washington said.
A Crowded Section 232 Docket
The polysilicon action is one piece of a rapidly expanding Section 232 architecture. The administration has already completed national security investigations and imposed tariffs on autos, steel, aluminum, copper, and lumber, and in January it issued a narrowly targeted 25 percent tariff on certain advanced computing chips alongside the trade and investment agreement with Taiwan. The president is expected soon to release findings from investigations into imported drones and industrial robots, and probes into semiconductors more broadly, pharmaceuticals, critical minerals, and wind turbines remain in progress.
The polysilicon investigation itself opened on July 14, 2025, covering the material and its derivatives, including wafers, cells, and modules. Nearly 50 organizations filed comments. The investigation’s findings, first reported by Reuters this week, conclude that import dependence on a strategic adversary for the foundational material of chips and solar power constitutes a threat to national security within the meaning of the statute.
Notably, the timing intersects awkwardly with developments inside China. The same price collapse that hollowed out Western producers has driven China’s own polysilicon sector into its deepest downturn on record, with operating rates below 40 percent, inventories above 300,000 metric tons, and top producer Tongwei reporting losses of roughly 1 billion dollars across 2024 and 2025. Six of China’s largest producers have discussed a consolidation fund of at least 50 billion yuan, about 7.4 billion dollars, to buy out and permanently retire roughly a third of Chinese capacity. If Beijing succeeds in converting its industry from growth-through-dumping to disciplined pricing power, a U.S. price floor could end up locking in higher revenues for the Chinese producers that survive consolidation, an irony that trade economists have been quick to point out.
The Bill for Solar Developers
For the U.S. solar industry, the arithmetic is uncomfortable. Analysis by Roth Capital Partners managing director Phil Shen, published in June before the final scope was known, modeled a base case in which the Section 232 action adds about 10 cents per watt to the cost of imported solar cells, which translates to an increase of 4.00 to 5.50 dollars per megawatt-hour in power purchase agreement prices. For a typical residential system of six to eight kilowatts, the increase works out to roughly 600 to 800 dollars. In Roth’s worst case, combining a minimum import price with ad valorem tariffs, implied module costs could rise from below 30 cents per watt to as much as 49 cents, an increase of more than 60 percent on imported modules.
Developers are already writing the uncertainty into contracts. Change-in-law clauses have proliferated across active power purchase agreement negotiations, shifting tariff risk between buyers and sellers in anticipation of the proclamation.
The supply math explains why the industry cannot simply substitute away from imports. Outside China, global operational polysilicon capacity is roughly 92,000 metric tons. The United States installs approximately 50 gigawatts of solar per year, and each gigawatt requires about 2,500 metric tons of polysilicon, implying U.S. solar demand alone of at least 125,000 metric tons annually. Non-Chinese supply cannot cover that requirement, which means some degree of import dependence will persist regardless of the tariff structure, and the binding question becomes which imports are permitted at what price.
That is why the trade group Solar Energy Manufacturers for America and a bipartisan group of lawmakers urged Commerce to preserve room for allied suppliers such as South Korea’s OCI Holdings and Wacker’s plants in Germany and Malaysia while domestic capacity scales. The Coalition for a Prosperous America proposed a tariff-rate quota: no duty on the first 40,000 metric tons of polysilicon from allied non-Chinese supply chains, then 10 dollars per kilogram beyond the quota, a 30 gigawatt quota for wafers and cells before a 10 cent per watt duty, and a flat 20 cent per watt tariff on all silicon panels. Whether the administration adopts that structure will determine whether the pain is confined to above-quota volumes or spread across the whole market.
Some manufacturers fear demand destruction more than import competition. “The market might start having projects fall down” if costs climb too high, Martin Pochtaruk, chief executive of Minnesota panel maker Heliene, told Reuters, describing the solar industry’s position in a chip-focused trade action as “collateral damage.”
Implications for Chipmakers and Importers
For semiconductor manufacturers, the direct cost impact is more modest. Semiconductor-grade polysilicon already trades at a multiple of solar-grade prices, and chip fabs pay premiums for purity and supply security as a matter of course. With TSMC’s Arizona fabs, Intel’s expansion, and other CHIPS Act projects creating new domestic demand for semiconductor-grade material, the administration is betting that guaranteed offtake plus trade protection will justify expanded U.S. production. Hemlock received a 325 million dollar CHIPS incentives award in January 2025 to expand semiconductor-grade output, and the Section 45X advanced manufacturing credit adds 3 dollars per kilogram for domestically produced solar-grade material, support that analysts at Intertek CEA estimate has narrowed the import cost gap to under 10 percent.
For importers and trade compliance teams, several practical points stand out. Companies sourcing wafers, cells, or modules should model both a straight tariff and a minimum import price scenario, because the compliance mechanics differ: a price floor puts a premium on accurate invoicing and country-of-origin documentation, and it raises the stakes on transshipment enforcement, an area where Customs and Border Protection has already conducted factory raids in Southeast Asia in connection with solar circumvention. Buyers with long-dated supply agreements should review price adjustment and change-in-law provisions now. And firms with plans to invest in U.S. wafer and cell capacity should watch the offset mechanism closely, because it may convert planned capital expenditure into tariff relief.
Wacker, whose Tennessee plant gives it a foot on both sides of the tariff line, underscored what is at stake in a statement to Reuters: “Without polysilicon, the next steps of the value chain (wafer and chips or wafer and solar cells) are not possible.”
A Long Line of Solar Trade Cases, and Why This One Is Different
The coming proclamation will not be the first American attempt to police solar imports. It will be roughly the seventh major instrument layered onto the sector in fifteen years, and understanding the earlier layers explains both the skepticism and the hope surrounding this one.
Antidumping and countervailing duties on Chinese cells and modules date to 2012, and were promptly blunted when Chinese producers shifted cell production to Taiwan and then to Southeast Asia. A safeguard tariff under Section 201 followed in 2018, imposing global duties on cells and modules above a quota. Circumvention findings in 2023 extended the antidumping orders to Cambodia, Malaysia, Thailand, and Vietnam after Commerce concluded that Chinese producers were finishing goods there to wash origin. The Uyghur Forced Labor Prevention Act added an import ban dimension, with Customs detaining solar shipments connected to Xinjiang polysilicon. Each measure changed trade patterns; none restored American production of the upstream inputs. Panel assembly grew impressively after the Inflation Reduction Act of 2022 created manufacturing credits, but wafers and cells, the capital-intensive middle of the chain, remained overwhelmingly imported, and polysilicon production shrank to two firms.
The lesson the administration appears to have drawn is that taxing finished goods while leaving the feedstock unprotected builds an industry with shallow roots. The polysilicon action inverts the traditional sequence by protecting the base of the chain first and using investment offsets to pull the middle segments onshore behind it. It is, in effect, an industrial strategy executed through a national security statute, with the price floor serving the function that a subsidy would serve in a different political economy: guaranteeing producers a revenue level at which expansion is bankable.
Skeptics note the strategy’s circularity. The tariff raises input costs for the wafer and cell plants the policy hopes to attract, which is why the reported offset mechanism for importers investing in domestic capacity may be the single most important design detail in the proclamation. Executed well, it converts the penalty into a subsidy for exactly the investment the country lacks. Executed poorly, it simply taxes the downstream industry that generates most polysilicon demand, shrinking the market the protected producers depend on.
The Demand Backdrop: Data Centers and a Power-Hungry Grid
The action also lands in the middle of the sharpest electricity demand growth the United States has experienced in a generation. Data center construction tied to artificial intelligence has utilities racing to add generation, and solar paired with storage remains among the fastest resources to deploy at scale. Roth Capital’s estimate that the tariffs could add 4.00 to 5.50 dollars per megawatt-hour to power purchase agreements arrives at a moment when hyperscale buyers are signing contracts as fast as developers can offer them, and when any increase in the cost of new generation feeds directly into the electricity prices paid by households and industry.
That collision between industrial policy and energy policy explains some of the internal tension in the administration’s position. The White House has rolled back federal support for renewable energy while simultaneously courting data center investment that needs power from somewhere. Solar developers argue that taxing their supply chain while demand surges is self-defeating; supporters of the action respond that dependence on Chinese material for a strategic energy source is the greater risk, and that a domestic supply chain, once built, insulates the grid buildout from exactly the kind of foreign leverage the tariffs are meant to counter.
The chip side of the ledger faces its own demand crunch. Memory prices are rising, advanced packaging capacity is scarce, and the CHIPS Act fabs coming online in Arizona, Ohio, and Texas will need reliable feedstock. The Semiconductor Industry Association’s 2.4 percent figure understates the stakes: chips may be a small share of polysilicon demand, but polysilicon is one hundred percent of the silicon wafer, and there is no wafer without it.
The Balancing Act
The administration must now thread a needle of its own making. Set the floor too low and the two domestic producers remain exposed to a Chinese industry with a 35-to-1 capacity advantage. Set it too high and the data-center-driven boom in electricity demand collides with costlier solar power, while chip and electronics prices absorb new input costs. Industry groups representing solar developers and semiconductor buyers have warned Commerce that overreach could raise the cost of power plants and consumer products alike.
What is not in doubt is the direction of policy. With the polysilicon proclamation, the United States will have extended national security tariffs from finished goods down into the raw feedstock of the digital and energy economy, and it will have done so with a novel instrument, the minimum import price, that could become a template for critical minerals and other commodities where Chinese overcapacity has crushed Western producers. For a two-company domestic industry in Michigan and Tennessee, the protection may have arrived just in time. Whether it arrives at an acceptable cost to everyone downstream is the question the market will spend the rest of the year answering.