If you have read anything about cleaning up steel in the last few years, it was probably about hydrogen. Swap the coal for hydrogen gas, let the hydrogen pull the oxygen off the iron ore, and you get water vapor coming off instead of carbon dioxide. Sweden has an entire plant built around that idea.<\/p>\n
Boston Metal spent a decade arguing you do not need any of it. No hydrogen, no coal, no carbon capture, and no process water either. What you need is a box roughly the size of a school bus, iron ore dissolved in molten electrolyte, and enough current to hold the whole mess at 1,600 degrees Celsius until the metal separates itself out and pools at the bottom.<\/p>\n
The box works. Boston Metal proved that in February 2025. What went wrong afterward had nothing to do with the chemistry.<\/p>\n
MOE stands for molten oxide electrolysis. An inert anode sits in an electrolyte loaded with iron ore and gets electrified, and once the cell reaches 1,600 degrees the electrons split the bonds holding the iron oxide together. Oxygen gas comes off the top. High-purity liquid metal collects at the bottom and gets tapped out the way you would tap any furnace.<\/p>\n
The appeal is in what it deletes. The conventional route needs coke production, then sintering and pelletizing the ore, then the blast furnace, then a basic oxygen furnace to refine whatever comes out of that. Boston Metal’s own MOE Steel documentation<\/a> describes the process as replacing all four of those steps and sending molten metal straight to ladle metallurgy.<\/p>\n The technology came out of MIT. Professor Emeritus Donald Sadoway, Professor Antoine Allanore and James Yurko founded the company in 2013<\/a>, after Sadoway and Allanore worked out that an iron-chromium alloy was cheap enough and tough enough to serve as the anode. Hang onto the chromium. It comes back later.<\/p>\n Scale was always the hard part. The lab cells were the size of a coffee cup and made a few grams at a time. The commercially interesting ones are the school-bus-sized units, and getting there meant cramming multiple anodes into a single cell instead of one.<\/p>\n Boston Metal started up its multi-anode industrial cell in Woburn, Massachusetts in January 2025, let it run for several weeks, and tapped roughly a ton of material on February 17. MIT Technology Review reported the run at the time.<\/p>\n That is a real number from a real machine rather than a rendering, which puts it ahead of most green steel announcements. It is also a rounding error. The world makes about 2 billion metric tons of steel a year and emits over 3 billion metric tons of carbon dioxide doing it. A single American mill like the one in Osceola, Arkansas rolls 142,000 tons of steel coil<\/a> on its own.<\/p>\n Boston Metal says steelmaking accounts for about 10% of global CO2 emissions. MIT Technology Review puts it at roughly 8% of global greenhouse emissions, and MIT’s own materials science department has used a range of 7 to 9 percent. Pick whichever you like, because it is an enormous slice by any of those measures.<\/p>\n A cell of that size makes a ton or two a month. Adam Rauwerdink, the company’s senior vice president of business development, told MIT Technology Review in March 2025 that truly industrial equipment would produce that much in about a day.<\/p>\n Boston Metal’s Brazilian subsidiary runs a plant at Coronel Xavier Chaves, near S\u00e3o Jo\u00e3o del-Rei in Minas Gerais, pulling niobium, tantalum and tin out of mining waste. On January 30, 2026, something in it failed.<\/p>\n In its notice to Massachusetts the company called it an unforeseen industrial accident that caused a critical equipment failure. CEO Tadeu Carneiro later told MIT Technology Review<\/a> the problem was the refractory system, the lining that insulates the reactor and stops it corroding. Electrolyte leaked out. Operators shut the system down and removed the metal, and according to the company nobody was hurt and there was no environmental damage.<\/p>\n No injuries, no spill, and a plant that can be repaired. The financial damage was severe anyway.<\/p>\n Stopping the plant meant Boston Metal missed an operational milestone tied to a financing deal that was still pending. As a result of the accident and the inability to meet that milestone, vice president Natalie Alarcon wrote to the state on February 20, the company lost access to committed capital that was essential to sustaining operations at its Woburn facilities.<\/p>\n The Massachusetts WARN filings lay it out. Boston Electrometallurgical Corporation, doing business as Boston Metal, filed on February 20, 2026 for 71 layoffs in Woburn. An updated filing on March 12 moved the date to April 21. Same 71 people.<\/p>\n The timing is the part that stings. On February 11 the Healey-Driscoll administration had announced $950,308 for Boston Metal<\/a> through the Business Builds Capital Grant Program, earmarked for facility upgrades that would let it recover critical materials at commercial-grade purity. The WARN notice landed nine days later.<\/p>\n Boston Metal declined the award and no money was disbursed, a spokesperson for the Executive Office of Economic Development told the Daily Hampshire Gazette in March.<\/p>\n The rescue arrived in May. Boston Metal raised $75 million, taking its lifetime total past $500 million, with existing backers joined by Tata Steel Limited. Carneiro said the delay had put serious stress on the company’s cash flow, and that the repaired Brazilian plant should be ready to start up in September 2026.<\/p>\n The money is not mainly for steel. It is for critical metals: niobium, tantalum and tin at the Brazilian plant now, with vanadium, nickel and chromium named as the next targets.<\/p>\n Seaver Wang, director of climate and energy at the Breakthrough Institute, gave MIT Technology Review the cleanest explanation anyone has managed on this: “Nobody wants to pay a green premium for steel\u2014hence niobium.”<\/p>\n The arithmetic is on his side. Boston Metal puts annual global niobium production at around 120,000 tons, which is nothing against 2 billion tons of steel. But Rauwerdink told Canary Media back in 2023 that a kilogram of niobium might go for $40 while crude steel sells for about 40 cents, and at a hundred to one a small plant can pay for itself in a way a small steel plant never will.<\/p>\n The scrap route already exploits that gap from the other end. When four German steam generators were cut out of a dead reactor and shipped to a Swedish smelter, about 95% of that 1,200-ton load left as ordinary ingots<\/a>, because remelting existing steel has always been cheaper than making it from rock.<\/p>\n Here is where the chromium comes back. The inert anode at the center of an MOE cell is an iron-chromium alloy, and in 2025 Boston Metal signed a memorandum of understanding with Finnish stainless producer Outokumpu partly to secure chromium feedstock for those anodes.<\/p>\n Chromium sits on the US critical minerals list. According to the USGS Mineral Commodity Summaries 2026<\/a>, the United States mined no chromite whatsoever in 2024 or 2025, and net import reliance ran at an estimated 79% of apparent consumption last year, with recycled stainless scrap covering the remaining 21%. Of the chromite ore that did arrive between 2021 and 2024, 96% came from South Africa.<\/p>\n The Department of Energy picked Boston Metal in November 2023 to negotiate up to $50 million for a plant in Weirton, West Virginia making ultrapure chromium metal. Nearly three years on, that American chromium plant is still a plan.<\/p>\n The MOE Steel page on Boston Metal’s site was updated on August 18. It still says the cells are about the size of a school bus, still says they run at 1,600 degrees, still says a demonstration plant is coming. What it does not say any more is when.<\/p>\n In March 2025 Rauwerdink had that plant online in late 2026 and operating in 2027. The current wording is that it will come online in the coming years. The same page also notes that Boston Metal will not produce steel itself at all; the plan is to license the technology to steelmakers and sell them the anodes.<\/p>\n Electrifying industrial heat is not the barrier it used to be. A stack of brick and toaster wire in Kern County has been making high-pressure steam past 1,832\u00b0F on sunshine alone<\/a> for a year, for an oil producer, because the economics happened to work there.<\/p>\n None of which is a failure of the machine. A cell in a Massachusetts industrial park really did pour liquid steel with no coal anywhere near it, and that is more than most green steel pitches can put on the table. It is just that steel sells for 40 cents a kilo, niobium sells for a hundred times that, and a refractory lining in Minas Gerais turned out to get a vote.<\/p>\n","protected":false},"excerpt":{"rendered":" If you have read anything about cleaning up steel in the last few years, it was probably about hydrogen. Swap … <\/p>\nA ton of steel came out of the bottom in Woburn<\/h2>\n
Then the refractory let go in Minas Gerais<\/h2>\n
71 jobs in Woburn, and a state check nobody cashed<\/h2>\n
Nobody wants to pay extra for green steel<\/h2>\n
The anode is made of the one metal America does not dig up<\/h2>\n
The demonstration plant no longer has a date<\/h2>\n