Solar is one of the cleanest power sources we’ve got. But it could be even greener.

Some manufacturers are going the extra mile to clean up emissions from a critical material in solar panels’ supply chain.

There are only four companies that manufacture polysilicon, a critical material for solar panels and semiconductors, in the United States. This spring, one of them got a big influx of cash. In April, a Korean company called Hanwha Solutions announced it had become the largest shareholder of REC Silicon, which can produce 16,000 metric tons of polysilicon annually from a refinery in Washington State—enough to meet more than a quarter of the U.S. solar industry’s demand. Hanwha, which already operates the largest U.S. solar panel factory in Georgia, described the acquisition as part of a plan to “revitalize the U.S. solar market” by creating a made-in-America supply chain from raw materials to finished products.

If that plan is successful, it would not only demonstrate the U.S. is, in fact, able to make solar panels with domestically sourced materials — a key policy goal of the Biden administration. It would also show that polysilicon refining, the most energy-intensive step in solar manufacturing, could be made considerably greener in the process.

In the pantheon of climate solutions, low-carbon polysilicon may not sound particularly sexy. But it has become a hot topic in the world of solar as corporations and governments start thinking seriously about how to drive their emissions all the way to zero, including in the so-called upstream supply chains that provide materials and components for renewable energy. Already, solar photovoltaic, or PV, panels generate among the lowest carbon emissions of any energy source out there over their entire life cycle, including manufacturing. But as the industry grows, even the relatively small emissions associated with manufacturing PV panels could become significant in aggregate, potentially peaking at levels comparable to the current yearly emissions of large industrialized nations like France or Germany. 

recent study found that in a scenario where the world deploys solar rapidly, PV production could lead to 25 to 30 billion tons of cumulative carbon dioxide emissions by the middle of the century, eating up roughly 10 percent of the remaining carbon budget for limiting global warming to 1.5 degrees Celsius (2.7 degrees Fahrenheit). If we don’t rapidly embrace renewables like solar, we have little chance of meeting that climate target. Still, it is possible to increase our odds by cleaning up polysilicon production, which accounts for roughly half of the climate impact of solar PV….

The largest U.S.-based polysilicon manufacturer, Hemlock Semiconductor, produces polysilicon for both the solar and the semiconductor industry in Hemlock, Michigan. There, the electric grid already includes a significant amount of hydroelectric storage capacity, and it’s getting steadily cleaner as the local utility phases out coal and brings more solar energy online. Hemlock solar commercial manager Phil Rausch told Grist that per kilowatt-hour of energy used, the emissions associated with the electricity the company purchases are about half of its competitors’ electricity emissions. 

Beyond improvements in manufacturing efficiency and getting power from greener grids, a handful of polysilicon makers have turned to alternative processes that are less energy intensive than the Siemens process. Chief among those is the so-called “fluidized bed reactor,” or FBR, process REC Silicon uses at its Moses Lake facility. Hot, silicon-rich gas is fed into a chamber containing pellets of silicon, which grow in size as more silicon crystallizes on them. Because heat is introduced from outside the reactor, no cooling is required, allowing considerable energy and cost savings, according to May.

Industry experts say it is more difficult to produce ultra-high-purity polysilicon with FBR technology compared with the Siemens process, limiting its adoption in the industry. Rausch of Hemlock said that his company investigated the FBR process “extensively” over the last decade but ultimately determined that the polysilicon it produced “did not meet the needs of the industry.” May, however, is confident that REC Silicon can use the method to make polysilicon that meets the increasingly stringent purity requirements of solar manufacturers thanks to a series of recent upgrades to its Moses Lake facility. 

Parr of the Ultra-Low Carbon Solar Alliance is cautiously optimistic. FBR “is a more difficult technology to get better purity from,” he said, but REC Silicon has spent years improving its process. “I think like any technology, it takes a while to perfect it, but it is inherently lower energy, lower carbon technology, so that’s promising.”

Any polysilicon maker that can deliver a greener product — whether that’s due to more efficient production methods or cleaner power sources — is likely to have a growing advantage in the solar market as companies or governments start paying more attention to supply chain emissions. 

“We’re already seeing the downstream buyer becoming much more sensitive to the embodied carbon in the supply chain,” Rausch said. As an example, he notes that several years back, when France began taking the carbon footprint of solar panels into account in its public procurement process for clean energy, companies buying polysilicon from Hemlock started doing better in that market. New labeling schemes, like an eco-label for solar that the Global Electronics Council is developing in partnership with the Ultra-Low Carbon Solar Alliance, are likely to drive further interest in clean polysilicon.