When looking into innovation in the solar manufacturing space, most of it usually comes from solar cells and modules. The shift from p-type to n-type solar cells is still ongoing, with TOPCon already dominating the market over PERC, but work on the next generational leap is underway.
Further upstream, things seem more straightforward, especially with polysilicon. The production of polysilicon is predominantly achieved via the Siemens process, although the fluidised bed reactor (FBR) process also exists, with what seems like very little room for new ways to produce polysilicon for the solar industry.
Try Premium for just $1
- Full premium access for the first month at only $1
- Converts to an annual rate after 30 days unless cancelled
- Cancel anytime during the trial period
Premium Benefits
- Expert industry analysis and interviews
- Digital access to PV Tech Power journal
- Exclusive event discounts
Or get the full Premium subscription right away
Or continue reading this article for free
US polysilicon company Highland Materials aims to bridge that gap with a new proprietary technology. “We believe there’s a lot of opportunity to improve the way polysilicon is being made, because that’s what we’re doing,” says Richard Rast, CEO at Highland Materials. The company recently secured a long-term lease at a former nuclear site to build what will be a 20,000MT polysilicon manufacturing plant once fully operational.
With only a select few companies with operational polysilicon capacity in the US – primarily Wacker Chemie and Hemlock Semiconductors – the entry of a new player would help increase the domestic capacity of polysilicon in the US, while also being able to compete within the market globally. And for Rast, this will be achieved through innovation. “Innovation gives us the opportunity to compete straight up against China.”
Innovating through recycling
One of the innovations in polysilicon production that Highland Materials aims to bring is through recycling, something that cannot be achieved with the Siemens process, says Rast. He mentions that around 30-40% of polysilicon intended for solar ends up as kerf, essentially “sawdust”, that’s generated at the end of the wafer process. Highland intends to recycle that kerf into new material.
The topic of recycling has gained significant importance in the solar industry over the past couple of years, especially when talking about recycling solar modules that reach their end of life or are damaged.
Obviously, when talking about innovation, there is also the uncertainty of how well that new process works and, more importantly, its compatibility with n-type technology. Rast explains that Highland Materials made independent tests last year with two European laboratories to prove that the company’s new process works for n-type.
“The evidence seemed to be fairly conclusive. It’s our expectation that we’ll be producing samples from a pilot line during the construction phase that we can provide to customers so they can test on their own,” adds Rast.
Reducing costs compared with the Siemens process
Rast explains that the other advantage compared with the Siemens process is cost, both in building the plant and its long-term operational costs. Rast forecasts that the construction of a polysilicon plant similar to Highland Materials’ planned one in Tennessee can be achieved at half the capital cost typically associated with the Siemens process, while also reducing the construction period by 40%. The operating costs could be nearly 40-50% lower too, and use half as much power, Rast claims.
The cost of setting up a new polysilicon plant in the US is significant, Rast explains, highlighting how building a 20,000 metric ton polysilicon facility, using a traditional process, could require an investment of US$1.5-2 billion.
The numbers are similar when talking about building a 5GW annual nameplate capacity for wafers or ingots, perhaps explaining the ongoing disparity between available domestic capacity for modules, and to a lesser extent, solar cells, compared with the shortfall of operational wafer capacity in the US. Only a limited number of companies aim to build capacity in the coming years.
The polysilicon market in the US is quite similar to wafers, in that there are only a select few companies producing capacity in the US. “The market can potentially absorb between 75,000 and 80,000 metric tons without too much problem. And today, the market capacity is only roughly half of that in the United States for domestic polysilicon. It’s our intent to plug ourselves into that shortfall,” says Rast.
Producing 80,000 metric tons of polysilicon would put the US market in the range of producing 40GW of annual domestic capacity, which Rast seems confident can be achieved in spite of current market uncertainties.
“For planning purposes, we believe that it is likely to happen, regardless of what happens from a policy point of view, because the demand is there,” he says. “Solar is still the cheapest power and the quickest to deploy power production capacity that can be added.”
Despite the continuing uncertainty around the solar industry, both at the upstream and downstream levels, these numbers show that there will still be a need for a domestic supply chain in the US, especially factoring in the tariffs and other policy developments, such as the Section 232 investigation on polysilicon.
“We’re feeling pretty bullish about it. Whether it really happens is still to be determined. But we feel a lot less uncertainty now than we did a month ago.”
“It’s time for us to decide if we want to be competitive, and if so, we’ve got to figure out how to be smarter,” concludes Rast.
Richard Rast will be speaking at PV CellTech USA 2025 on 7-8 October 2025, our third PV CellTech conference dedicated to the US manufacturing sector. Full details on the event are available here.
