Curio's recycling chemistry moves from simulants to actual spent fuel

Curio has taken $2.5 million from the Department of Energy's ARPA-E to test two steps of its NuCycle fuel-recycling process against actual used nuclear fuel and transuranic elements, at Pacific Northwest National Laboratory and Savannah River National Laboratory ^[1][4][5]. The move off simulated materials matters because that is where separations chemistry usually discovers what it got wrong ^[11].

The award is a cost share: about $1 million comes from Curio, putting the total project value near $3.5 million, with the company carrying roughly 29 percent ^[2][1][2]. It follows a $5 million ARPA-E project that began in 2023 and ran side-by-side laboratory demonstrations of the full NuCycle process at Idaho National Laboratory, Oak Ridge National Laboratory and PNNL ^[9]. Curio says that work hit its go/no-go milestones in September 2025 and produced data on the chemistry and reaction rates ^[10]. So the new money is half the size of the previous grant ^[3], but it is buying a harder test rather than more of the same one.

The two operations under examination are selective fluoride volatility and electrowinning, both described as central to NuCycle ^[3]. At PNNL, the work is fluorination to separate uranium from other radioactive materials ^[6]. At Savannah River, it is developing electrowinning equipment that uses high temperatures and electricity to pull transuranic elements out of molten salt ^[7]. Between the two projects, four national laboratories have now had hands on the process ^[4]. Splitting the flowsheet into unit operations and testing each against real feed is the conventional way to find out whether performance holds up under realistic conditions ^[11], and it is also the way to accumulate the kind of data a regulator will eventually ask for.

Which is the other half of the announcement. Curio will produce preliminary engineering designs for the fluorination and electrowinning equipment, plus a regulatory engagement plan, white papers on regulatory questions and an analysis of potential exemptions ^[12]. It says it is targeting deployment by the end of 2027 and pursuing a Part 70 operating license application ^[13]. The ambition behind that is a laboratory and demonstration facility able to process up to 4,000 metric tons of used fuel ^[8].

Note the distance between those two facts. Lab-scale testing of two unit operations on real material is early-stage work; a 4,000-tonne facility is an industrial plant ^[5][8]. The regulatory workstream here is still at the stage of white papers and exemption analysis, not a submitted application ^[12][13]. Curio has signed agreements or memoranda of understanding with NuScale, Framatome and Sargent & Lundy ^[14], which is customer and engineering interest, not committed offtake as described.

Chief executive Ed McGinnis framed the award as evidence that DOE "intends to embrace new technologies" that reduce stored waste and capitalise on untapped domestic reserves, and credited the Trump administration's actions on used-fuel recycling ^[15][16]. ARPA-E's remit is high-risk energy technology with potential strategic and commercial impact ^[17], which is a fair description of where this sits.

What to watch: whether the fluorination and electrowinning results on real fuel are published in enough detail to compare against the simulant runs, whether a Part 70 application is actually filed, and whether any of the three named industry partners converts an MOU into a funded commitment ^[13][14].