The 99.95%-purity iron announced by Hertha Metals is once again putting a spotlight on a raw material that has often gone unnoticed in the neodymium-iron-boron magnet (NdFeB magnet) supply chain. At its demonstration facility in Conroe, Texas, the company says it reached 3N5 purity using only domestic US feedstock and commercial-grade equipment. While rare earths tend to draw the most attention as the star players of the magnet supply chain, iron actually accounts for roughly 70% of a magnet's mass. What's needed isn't ordinary steel in bulk, but iron with carefully controlled carbon and impurity levels.
This milestone doesn't demonstrate the magnetic strength or durability of a finished magnet, nor does it certify the material for defense applications. Still, it's notable that the company—which had previously only laid out concrete future targets—is now reporting that it has hit customer specifications using existing equipment. Hertha claims this is a first for the US in this particular combination, but that's not a third-party-verified industry ranking. Whether the commercial facility the company plans to break ground on this year, called "Chalyx," can actually move toward supply isn't something the purity figure alone can determine.
What 3N5 Means: Iron Makes Up About 69% of a Magnet
99.95% is also expressed as 3N5. By simple subtraction, that leaves a maximum of 0.05%, or 500 ppm, for everything other than iron. However, this isn't a standard that specifies individually how much oxygen, carbon, and nitrogen are allowed. The same goes for phosphorus, sulfur, and residual metals. Hertha explains that it controls all of these simultaneously through a single continuous pyrometallurgical process, but the announcement doesn't disclose element-by-element analytical values or measurement methods.
According to a 2022 US Department of Energy (DOE) supply chain assessment, a typical NdFeB magnet consists of about 69% iron, 1% boron, and 30% rare earths. The core rare earths are neodymium, praseodymium, and dysprosium, with small amounts of cobalt and aluminum as well as terbium and holmium sometimes added. Even if rare earth procurement alone is secured, the manufacturing process won't come together unless magnet makers can also continuously secure the low-carbon, low-impurity iron they require.
In July 2025, Hertha set a target of 99.97% or higher purity for a future facility. Using the same subtraction logic, 99.97% corresponds to a maximum of 0.03%, or 300 ppm, of non-iron content—meaning today's 99.95% leaves 200 ppm more headroom in total impurity content. That said, the earlier figure was a target for a future facility, and the company states that today's 3N5 result meets the specifications of major rare earth magnet manufacturers. This isn't a comparison that can simply be dismissed as falling short. On the other hand, the company hasn't disclosed customer names, the target magnet grades, lot sizes, or qualification results.
What Blocked Domestic Sourcing Wasn't the Existence of Pure Iron, But Cost and Demand
According to the DOE, the iron feedstock traditionally used by US magnet manufacturers was AISI 1001 ultra-low-carbon steel. At the time of that assessment, domestic production was very limited, and imports relied heavily on Germany and, in some cases, Brazil. US companies did have manufacturing capability, but demand volume was insufficient. US-made electrolytic iron was also available, but it required a remelting and solidification process, which was said to drive up costs.
Seen against this backdrop, Hertha's claim isn't really the discovery that "the US can now make high-purity iron." Rather, it's a proposal about whether a domestic route can be built to economically obtain magnet-usable iron from lower-grade feedstock. The company states that its process can cut production costs by more than 20% and emissions by 50% compared to conventional methods. However, the announcement doesn't specify the comparison baseline, yield, energy source, or the scope of the emissions calculation. At this stage, this should be treated as the company's own outlook rather than an independently verified performance comparison.
The state of the facility itself is also worth distinguishing. Hertha is operating a 1-ton-per-day demonstration facility in Conroe. Its 2025 announcement stated that this pilot facility had been running since late 2024. The company's technology page describes the process as "semi-continuous," while the announcement materials use the phrase "continuous" pyrometallurgical process. Either way, what's been confirmed here are material results from a demonstration facility—not data backing long-term commercial operation or the magnetic properties of a finished magnet.
2027: DoD Procurement Will Even Scrutinize Where the Iron Was Mined
The US Department of Defense (DoD) procurement rule DFARS 252.225-7052 restricts the origin of materials contained in certain magnets being procured. Through December 31, 2026, the primary focus is on covered materials melted or produced in covered countries, along with the subsequent magnet manufacturing process. The covered countries are North Korea, China, Russia, and Iran—four countries in total.
Starting January 1, 2027, the conditions expand for NdFeB magnets. The entire supply chain—from the mining of neodymium, iron, and boron through to the finished magnet—becomes subject to traceability. The origin of iron will matter not only for magnet makers' purchasing requirements but also for whether products can be delivered under covered DoD contracts. This is why Hertha's emphasis on "domestic US feedstock" can't be fully explained by purity alone.
However, it would be inaccurate to read this rule as a blanket US regulation banning all Chinese-made magnets. This is a restriction that applies to DoD contracts and procurement, and there are exceptions for certain commercial off-the-shelf (COTS) end items, electronic devices not separately designated, and NdFeB magnets made from recycled material that was crushed and sintered domestically in the US. Non-availability determinations can also be granted in some cases, and contracts below the simplified acquisition threshold are exempt from this clause's restrictions. Domesticating raw material sourcing doesn't automatically guarantee adoption in every single case.
Groundbreaking Pushed to Later This Year—Commercialization Is Just One Link in the Supply Chain
Chalyx is planned as Hertha's first commercial facility. The July 2025 announcement indicated groundbreaking in January 2026, annual capacity exceeding 9,000 tons (30 tons per day), and a purity target of 99.97% or higher. However, the August 18, 2026 announcement now states that groundbreaking is planned for "sometime within 2026." It's not yet accurate to say the plan has progressed from demonstration results to commercial supply—and it has fallen behind the original schedule.
The new announcement doesn't disclose Chalyx's start-of-operations date, initial rated capacity, construction costs, or offtake agreements with customers. Scaling from a 1-ton-per-day demonstration facility to more than 9,000 tons annually involves a significant gap between achieving purity once and consistently producing the same quality with good yield over time. Qualification by magnet manufacturers, and confirmation of strength, coercivity, and durability in finished magnets, are separate processes altogether.
According to the International Energy Agency (IEA), China accounted for 60% of global mining of magnet-grade rare earths, 91% of refining, and 94% of sintered permanent magnet production in 2024. These figures represent the share of magnet-grade rare earths and sintered magnets—not the supply volume of high-purity iron. The IEA estimates that even adding up all publicly announced non-China projects as of early 2026, capacity for metals, alloys, and finished magnets by 2035 would total only about 18,000 tons (18kt) in rare earth content. Magnet manufacturing, it says, is the stage most prone to bottlenecks in diversification efforts.
That roughly 18kt of planned non-China capacity is also counted in rare earth content, not iron tonnage. This figure can't be directly compared against Hertha's production plans to argue about the relative scale of supply. Still, if a bottleneck occurs anywhere along the long chain—from rare earth separation through metal/alloy formation, magnet shaping and sintering, to customer qualification—then domesticating iron supply alone won't translate into finished magnets. Adding one more raw material source and achieving stable magnet supply are two different jobs.
What Hertha is trying to fill in is the iron-related link in that long chain. What should be verified after Chalyx breaks ground isn't the purity figure of 99.97%-plus in isolation—it's whether the company can reproduce it. Actual operational start dates and capacity need to be disclosed, along with yield data and customer qualification results. Only when the origin can also be traced through other raw materials and the sintering process will this 3N5 demonstration become supply capability that genuinely connects to DoD procurement.