The Trump administration is preparing to set a price floor and impose tariffs on polysilicon and related products, according to four people familiar with the plan — an intervention that would reach deeper into commodity markets than almost any prior trade action and strikes simultaneously at the supply chains for both semiconductor chips and solar panels. The move follows a year-long national security investigation and is expected to culminate in a presidential proclamation within days.
China’s manufacturers produce 93.5% of the world’s polysilicon, leaving the United States with essentially two domestic producers — Hemlock Semiconductor in Hemlock, Michigan, and Wacker Chemie in Charleston, Tennessee — to supply the feedstock that underpins every silicon chip and every solar panel manufactured in the country. Protecting those two producers from being further undercut by Chinese dumping is the policy’s stated objective. The bill for that protection, however, will initially be paid by the US solar developers and chip manufacturers that currently source cheap imported polysilicon.
Polysilicon is the ultra-pure feedstock from which silicon wafers are sliced for use in both chip fabrication and solar cells. It is, as the industry sometimes describes it, the material at the very base of two supply chains simultaneously — a chokepoint position with no modern equivalent in semiconductor manufacturing.
The Chinese polysilicon industry reached 93.5% of global output in 2024, with nine of the world’s ten largest producers based there. The top four — Tongwei, GCL Technology, Daqo New Energy, and Xinte Energy — together held 65% of global output. This dominance was not accidental: Beijing provided significant state support to its domestic industry to establish Chinese firms as the dominant global suppliers and is now seeking to push its companies further into the higher-value semiconductor-grade segment.
The price consequences of this concentration have been severe. Prices fell from around $39 per kilogram ($17.69 per pound) in 2022 to below $4.50 per kilogram ($2.04 per pound) by the end of 2024, wiping out margins for Western producers and triggering shutdowns across the non-Chinese industry. As of July 7, 2026, the China benchmark price for mono-grade polysilicon stood at approximately CNY 32,286 per metric ton — about $4.75 per kilogram ($2.15 per pound) — according to OPIS market data, as a seasonal rebound in hydropower-fueled Chinese output pushed supply back toward 100,000 metric tons (approximately 110,231 short tons) per month.
Read more: Forced Labor Import Tariffs on 60 Countries: Companies Face June 22 Deadline to Testify
To understand why China dominates polysilicon — and why a price floor can shield US producers from import competition without fully solving the cost gap — requires understanding how polysilicon is made.
The dominant production method is the Siemens process, a form of chemical vapor deposition invented in 1954. Metallurgical-grade silicon (about 98% pure, produced from quartzite) is reacted with hydrochloric acid to produce trichlorosilane gas, which is then purified through distillation to strip out metallic impurities. The purified trichlorosilane is introduced into a sealed reactor where thin silicon rods are resistively heated to approximately 1150°C (2102°F). Ultra-pure silicon deposits onto the hot rod surface — a process that runs for roughly 60 hours per batch, producing rods of 120mm (4.7 inches) diameter before they are harvested, broken into chunks, and packed in inert gas for shipment.
The energy consumed in heating those rods is enormous — roughly 50 to 100 kilowatt-hours per kilogram of polysilicon produced. That energy intensity makes electricity price the single most important variable in production economics. Chinese facilities in Sichuan province, located near major hydropower installations, access electricity at roughly $0.02 per kilowatt-hour during the rainy season — approximately one-quarter to one-fifth the industrial electricity rate paid by Hemlock and Wacker at typical US grid prices. That structural advantage does not disappear when the US imposes a price floor. It simply means that protected US producers can survive in the US market; they still cannot export competitively or match Chinese costs on a global basis.
For semiconductor fabrication — the application that feeds TSMC Arizona, Intel, and other advanced US chip fabs — polysilicon must achieve nine-to-eleven-nines purity — 99.999999999% pure silicon — far exceeding the six-to-nine-nines standard for solar-grade material. This purity differential commands a three-to-four-times price premium and requires additional purification cycles beyond what solar producers need. Hemlock and Wacker both produce semiconductor-grade material; preserving domestic production of this higher-specification product is, from the chip-supply perspective, the more critical of the two objectives this tariff serves.
On July 14, 2025, the US Department of Commerce Bureau of Industry and Security announced a formal investigation into the effects on US national security of imports of polysilicon and its derivatives, including silicon wafers, solar cells, and solar modules. The investigation operates under Section 232 of the Trade Expansion Act of 1962, a national security import statute that authorizes the president to restrict imports of products that threaten to impair national security.
Section 232 gives the president broad authority to impose tariffs, quotas, or minimum prices on covered imports. Its remedies can include price floors — formally called Minimum Import Prices under Section 232 (MIPs) — which set a floor below which imported goods cannot legally be sold in the US market. A Minimum Import Price is structurally different from a tariff: rather than adding a percentage or dollar amount to the import value, it directly establishes the floor sales price, targeting dumping below production cost.
By January 2026, Trump had signed proclamations on semiconductors and critical minerals that directed Commerce and the US Trade Representative to negotiate agreements on critical minerals, with a related report due July 13, 2026. The polysilicon action now expected is the culmination of that process. Among the Section 232 investigations still pending are those covering pharmaceuticals, commercial aircraft, unmanned aircraft systems, wind turbines, and industrial machinery.
Read more: South Carolina Cell Factory Hits 3 GW as TOPCon Patent Battle Stalls Rivals
The domestic polysilicon industry arrived at this moment in significantly diminished condition. Three major US producers once competed: Hemlock Semiconductor (Hemlock, Michigan), Wacker Chemie (Charleston, Tennessee), and REC Silicon (Moses Lake, Washington). Between 2011 and 2018, after China imposed retaliatory anti-dumping duties of as much as 57% on US-made polysilicon, combined US revenue from the material fell from approximately $1 billion to $107 million. REC Silicon subsequently ceased polysilicon production at its Moses Lake facility after failed attempts to achieve semiconductor-grade purity.
Hemlock and Wacker remain in production, both for solar and semiconductor markets. In January 2025, the Biden administration awarded Hemlock a $325 million CHIPS Act grant to expand semiconductor-grade polysilicon production at its Michigan campus. The IRA’s Section 45X Advanced Manufacturing Production Credit provides an additional $3 per kilogram ($1.36 per pound) incentive for domestically produced solar-grade polysilicon, which analysts at Intertek CEA estimate has narrowed the cost gap with imports to less than 10%.
Outside China, the entire world has only approximately 92,000 metric tons (about 101,413 short tons) of operational polysilicon capacity. China’s operational capacity reached 3,250,000 metric tons (about 3.58 million short tons) in 2024 — a 35-to-1 ratio that no tariff can close in the near term.
Paradoxically, the same price collapse that devastated US producers has pushed China’s own polysilicon sector into its deepest recorded downturn. China’s producers have been cutting output to historically low levels, with average operating rates dropping below 40%, as inventories ballooned to more than 300,000 metric tons (331,000 short tons) and losses mounted across the industry — Tongwei reported losses of approximately CNY 7 billion ($1 billion) over 2024 and into 2025, while GCL Technology, Daqo New Energy, and Xinte Energy reported similar losses.
In July 2025, six of China’s largest polysilicon producers announced discussions around creating a fund of at least 50 billion yuan capacity acquisition plan (approximately $7.4 billion USD) to acquire and permanently shut down roughly one-third of China’s existing polysilicon capacity — an estimated 1 million metric tons (1.1 million short tons) of lower-quality production lines. OPIS analyst Summer Zhang told pv magazine that the plan’s effectiveness would “hinge largely on the timing and scale of capital injections” and that key governance and funding details remained undefined.
The consolidation effort reflects a structural shift in China’s competitive strategy — from growth-through-dumping to growth-through-pricing-power — that makes the timing of the US intervention uniquely complicated. A price floor imposed while China is actively trying to reduce its own supply would, if successful, leave Chinese producers with higher revenues from a smaller US market share, while the consolidation fund removes the most inefficient capacity that was driving prices below cash costs.
Meanwhile, China has reinforced its own market barriers. Beijing extended anti-dumping duties on solar-grade polysilicon imported from the US and South Korea for another five years starting January 14, 2026, with US producers facing duties of 53.3% to 57% — keeping the retaliatory framework that gutted US producers in the 2010s firmly in place even as Washington prepares its own countermeasures.
The cost implications of the expected proclamation differ sharply between the two industries this material serves.
For US solar, the analysis from Roth Capital Partners managing director Phil Shen — reported in June 2026 before the proclamation’s final scope was known — modeled a base case in which Section 232 tariffs would add approximately $0.10 per watt to the cost of imported solar cells. That amount, Roth’s research suggests, correlates to a $4.00 to $5.50 increase per megawatt-hour in power purchase agreement prices — a material increase for utility-scale solar developers working with thin margins. In a worst-case scenario combining a Minimum Import Price with ad valorem tariffs, Roth’s modeling found that implied module costs could rise from the current sub-$0.30 per watt to as high as $0.49 per watt — roughly a 63% increase on imported modules. Solar developers have already begun implementing “change-in-law” clauses across active PPA negotiations in anticipation of the volatility.
The Coalition for a Prosperous America, one of nearly 50 organizations that submitted public comments to Commerce during the Section 232 investigation, proposed a tariff-rate quota structure: no tariff on the first 40,000 metric tons (44,092 short tons) of polysilicon imports from allied countries with non-Chinese supply chains, then $10 per kilogram ($4.54 per pound) on out-of-quota imports; a 30-gigawatt quota on imported wafers and cells before a 10 cents per watt out-of-quota tariff; and a flat 20-cents-per-watt tariff on all silicon solar panels. Whether the administration adopts this structure, or a different one, will determine whether the cost impact is confined to above-quota volumes or applied broadly.
For the US chip industry, the situation is different. Semiconductor-grade polysilicon already commands a 3× to 4× price premium over solar-grade material, and chip manufacturers have long operated under a different cost model — they pay premium prices for guaranteed purity and supply security. The CHIPS Act investments in domestic chip fabrication (TSMC Arizona, Intel Ohio) have created new domestic demand for semiconductor-grade polysilicon at precisely the moment the Section 232 investigation was concluding. Tariff-induced cost increases may be significant in dollar terms but are less likely to threaten the economics of chip fab projects that are themselves receiving billions in federal support.
The non-Chinese supply chain as a whole — 92,000 metric tons (101,413 short tons) of operational capacity worldwide — cannot meet total US solar demand alone. With the US averaging approximately 50 gigawatts of annual solar installations, and each gigawatt requiring roughly 2,500 metric tons (2,756 short tons) of polysilicon, the US needs at least 125,000 metric tons (137,789 short tons) of polysilicon per year for solar alone. The gap between available non-Chinese supply and US solar demand means some degree of continued import dependency is unavoidable in the near term, no matter the tariff structure.
A Minimum Import Price (or price floor) sets a minimum legal sales price below which imported polysilicon cannot be sold in the US market. A tariff, by contrast, adds a percentage or dollar amount to the imported product’s value at the border — it raises the price but doesn’t directly constrain what a seller can charge. A price floor is specifically designed to combat below-cost dumping: if Chinese producers are selling polysilicon at $4.75 per kilogram ($2.15 per pound) while their full production costs are higher, a price floor set above that level would prevent the dumped material from reaching US buyers at the dumped price. The enforcement challenge is that a price floor can be circumvented through transshipment (routing material through a third country) or mis-invoicing — the same vulnerabilities documented in CBP’s enforcement actions on solar panels.
China’s dominance traces to the Siemens process’s extreme energy intensity — producing one kilogram (2.2 pounds) of polysilicon consumes roughly 50 to 100 kilowatt-hours of electricity. Chinese facilities in Sichuan, located near major hydropower dams, access power at approximately $0.02 per kilowatt-hour during the rainy season — far below the $0.07 to $0.09 per kilowatt-hour typical of US industrial electricity. That structural gap persists even when US producers receive the IRA’s $3-per-kilogram ($1.36 per pound) Section 45X production credit and CHIPS Act capital grants. A price floor can prevent Chinese producers from undercutting US competitors in the US market; it cannot give Hemlock or Wacker access to Chinese hydropower.
The short answer is yes, with a range. Roth Capital Partners modeled a base case of approximately $0.10 per watt added to the cost of imported solar cells — which translates to $600 to $800 in additional cost for a typical 6-to-8 kilowatt residential system. In a worst-case scenario involving both a Minimum Import Price and ad valorem tariffs, the increase could be substantially larger. The actual impact will depend on the specific structure of the presidential proclamation when it is issued, how much of the supply chain is covered, and how quickly domestic production can scale to offer competitive alternatives. Solar developers have begun adding “change-in-law” provisions to contracts precisely because the tariff scope and timing remain uncertain.
Nothing does so automatically. Transshipment — routing Chinese-made goods through a third country to shed their country-of-origin designation — is the most documented circumvention technique in solar supply chain enforcement. US Customs and Border Protection has conducted raid operations on Vietnam factories doing exactly this with solar panels, and the “substantial transformation” test used to determine country of origin has been contested in courts. A polysilicon price floor would face the same vulnerability unless accompanied by rigorous origin documentation requirements and enforcement capacity — both of which the administration’s broader solar tariff actions have struggled to provide comprehensively.
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.

source