A Texas startup wants to make magnet-grade iron from the fine ore that blast furnaces reject
Hertha Metals has started building a US plant that turns low-grade iron ore into 99.95%-pure iron in a single furnace. The case shows how a different process route changes what counts as saleable ore.
- Hertha Metals has started construction of Chalyx in Conroe, Texas, a plant of around 10,000 tonnes a year of high-purity iron and steel, with an investment of US$100 million.
- The process puts iron ore and natural gas into a single furnace and delivers either liquid steel or high-purity iron, depending on the chemistry controlled — including from low-grade fines that blast furnaces and direct reduction cannot take.
- Iron accounts for about 70% of the weight of a neodymium magnet, and the US imports more than 90% of the high-purity iron it uses, mostly from China.
- For ore producers, the message is that the processing route defines a product's value as much as its grade.
Hertha Metals, a company based in Conroe, Texas, announced in October 2026 the start of construction of Chalyx, its first commercial plant. The unit sits next to the Pi100 pilot plant, which runs at around one tonne per day, and is designed to produce about 10,000 tonnes a year of high-purity iron and steel. The company says it has reached 99.95% purity at the pilot. It expects to start operating in late 2027 and to begin magnet-grade production in early 2028. The ore will be domestic, mainly from Minnesota's Iron Range, and the natural gas will come from Texas. The groundbreaking follows a US$133.65 million Series A round co-led by Khosla Ventures and Doerr Capital. Of that round, US$65 million came from the US government through the Industrial Base Analysis and Sustainment program, which is linked to defense. The stated long-term goal is to reach 500,000 tonnes a year.
The classic steel route has several stages, and each one places demands on the ore. The blast furnace needs particle size and mechanical strength, which forces fines to be agglomerated into sinter or pellets. Direct reduction requires high-grade pellets, usually above 67% iron, with little silica and alumina. Whatever falls outside these windows becomes a discounted product, a stockpile or tailings. Hertha's proposal is a single-step pyrometallurgical process. Ore and natural gas go into the same furnace, and control of the bath chemistry decides whether liquid steel or nearly pure iron comes out. According to the company, the furnace removes oxygen, carbon, nitrogen, phosphorus, sulfur and metallic impurities in a single continuous step. It also accepts ores with less than 60% iron, fines and even mill scale. The company also claims three advantages: cost 25% lower than the blast furnace; 30% more energy efficiency than conventional steelmaking; and at least a 50% cut in emissions using natural gas. These are numbers from a company that has not yet operated at commercial scale. The Chalyx plant itself exists to prove them, and until then they should be read as targets. The point that matters to people who work with ore is a different one. The minimum grade that determines whether a material is ore or waste is not a property of the rock. It is a consequence of the available processing technology and of the price of the final product. When a route appears that tolerates more impurities, part of what used to be discarded finds a market. The second lesson is about high-value niches. Iron is a low-priced commodity, but 99.95%-pure iron is a critical input for permanent magnets, which are about 70% iron by weight. The US imports more than 90% of this material, mainly from China. The Department of Defense has banned magnets and materials of Chinese origin from January 1, 2027. In other words, demand for the product exists before the plant does. The financing reflects this: half of the round's capital came from a defense industrial base program, not from a steel development bank.
Brazil has billions of tonnes of low-grade itabirites and fines stored in dams and stockpiles, especially in the Iron Quadrangle. Today, using them requires costly concentration and pelletizing. Process routes that accept leaner ore could change the economics of these stockpiles. That is why they deserve technical follow-up by resource geologists, process engineers and mine planners, even before they are proven at scale. For those working on the magnet and critical minerals supply chain, the case is a reminder that the lowest-profile link — iron — can also be a bottleneck.
What did we learn?
- An ore's cut-off grade depends on the available processing technology: a route that tolerates more impurities turns waste into feedstock.
- 99.95%-pure iron is a different product from ordinary iron and accounts for about 70% of the weight of a neodymium magnet.
- Cost and emissions figures from pre-commercial technologies should be treated as targets until the first commercial plant is running.
Skills Radar
- Ironmaking and iron ore reduction★★★★★
- Techno-economic assessment of emerging technologies★★★★★
- Use of low-grade ore and fines★★★★★
- Permanent magnet supply chain★★★★★
Skills Developed
- Pyrometallurgy
- Process route assessment
- Iron ore value chain
Upward trend
the US defense ban on Chinese materials from 2027 and interest in lower-emission steel routes should attract more capital to processes that accept low-grade iron ore.
Who is this content useful for?
- Engineers
- Geologists
- Researchers
- Managers
- Students
- Companies
To go deeper on this topic
Worth pursuing training in:
- Metallurgical engineering
- Mineral processing
- Materials engineering
- Economic evaluation of mineral projects


