Polysilicon tariff stacks on top of existing duties, pushing combined levies on Chinese solar derivatives above 65%

Trump's 6 August proclamation under Section 232 of the Trade Expansion Act of 1962 imposes a 15% ad valorem tariff and a minimum import price (MIP) regime on polysilicon and its downstream derivatives, effective 4 December 2026^[1]. The measure lands on top of existing Section 301 duties and anti-dumping and countervailing duty orders, meaning combined levies on Chinese-origin solar derivatives could exceed 65%.

What the proclamation actually does

The tariff structure has two distinct layers. Raw polysilicon is subject only to the MIP-based duty, while downstream derivatives face both the MIP floor and the 15% ad valorem tariff - a design that explicitly penalises importing processed material over the raw feedstock.

The MIP floors are^[1]:

• $21 per kilogram for polysilicon
• $100 per kilogram for polysilicon ingots and wafers
• $0.22 per watt for solar cells
• $0.38 per watt for solar modules

The Section 232 tariff replaces a narrower safeguard tariff on solar cells and modules that expired in February. Country-specific treatment applies: the United Kingdom faces a 10% Section 232 tariff, while the EU, Japan, South Korea, Taiwan, Switzerland, and Liechtenstein are subject to a combined Section 232 and Column 1 duty rate that equals 15%.

CBP may assess additional duties, issue penalties, or ban importers from bringing covered products into the US for noncompliance.

Why polysilicon sits at the centre of both solar and chip supply chains

Polysilicon is a critical input for semiconductor and solar product manufacturing, supporting virtually every sector of the modern economy, including defense, cybersecurity, communications, artificial intelligence, renewable energy, and advanced manufacturing.

The administration's rationale rests on two structural facts. The US share of global polysilicon production capacity fell from 50% in 2005 to less than 2% in 2024, and around 90% of the world's polysilicon production now originates in China. Commerce Secretary Howard Lutnick found that since 2020, global production of polysilicon has grown by more than 270% and inventories reached a record high of 400,000 tons by the end of 2024.

Domestic semiconductor manufacturing depends on solar because the solar industry's larger demand for polysilicon helps support production of the material required for chips. The chip industry accounts for 2.4% of global polysilicon demand, according to the Semiconductor Industry Association.

The data centre procurement angle

The proclamation arrives as hyperscalers are executing some of the largest solar PPAs in history to power AI data centres. The average North American solar PPA price rose 4.7% during the first quarter of 2026, while solar and wind PPA costs have increased 13% and 24%, respectively, since this time last year. The new MIP floors on modules at $0.38/W set a hard price floor for imported panels that will feed directly into project-level capital costs for any developer sourcing modules offshore.

Tariff uncertainties and ongoing Section 232 investigations have already been adding direct development costs to the PPA bottom line. With the proclamation now finalised, developers have a 120-day window - until 4 December - to reprice contracts or accelerate procurement under existing terms.

The United States currently has two major polysilicon production facilities: Hemlock Semiconductor in Michigan and Wacker Chemie's facility in Tennessee. Both companies welcomed the administration's action, with Hemlock parent Corning saying it would encourage continued investment in US capacity, while Wacker pointed to the importance of polysilicon production for semiconductor supply-chain resilience.

Onshoring incentive and what comes next

Companies investing in US polysilicon, ingot, wafer, or cell production may negotiate individually tailored onshoring agreements with Commerce for duty-free equipment and covered product imports during construction, provided construction begins by January 20, 2029.

Combined duties on Chinese-origin solar derivatives could exceed 65% when the Section 232 tariff is stacked on top of existing Section 301 tariffs (currently 50%) and applicable AD/CVD orders. That arithmetic will sharpen the calculus for any data centre developer weighing imported modules against domestically assembled alternatives.

The 120-day runway before the measures take effect is the immediate pressure point. Procurement teams with open module orders or unsigned PPAs referencing imported equipment will need to assess exposure before 4 December. The Commerce Department's onshoring incentive programme - and the terms it offers to companies that break ground on US capacity before January 2029 - is the secondary variable to track.