I used to assume a nuclear reactor starts itself. Load the fuel, pull the control rods, let physics take it from there.<\/p>\n
A fresh core is nearly silent, though. Almost nothing in it is splitting yet, so the neutron detectors that are supposed to tell operators what the core is doing have close to nothing to count, and if you are the person bringing that reactor up, the bottom of the instrument range is the last place you want your readings sitting. So crews put a neutron source into the core on purpose and watch the count climb off it. The material inside that source is californium-252. One lab makes it for the entire western world.<\/p>\n
That lab is Oak Ridge National Laboratory, and on August 4 it published a change to the chemistry<\/a> that gets californium-252 ready to leave the building, updating the post a week later. One molecule swapped into one step now strips 95 percent of the unwanted isotopes out of the solution in a single cycle. ORNL says the separation runs about a week faster than before, which cuts it roughly in half.<\/p>\n Californium-252 begins as curium targets sitting inside a reactor, soaking up neutrons. When those targets come out, technicians dissolve them in concentrated nitric acid, and what is left in the vessel is a mess. Fission throws off plenty of things besides californium, and americium, curium and cerium all ride along in the same solution. They have to come out before anybody has californium worth shipping.<\/p>\n For decades ORNL pulled them out with bis(2-ethylhexyl) phosphoric acid, HDEHP for short. HDEHP is an acid itself, and the way it grabs a metal ion is by letting go of a hydrogen ion. It cannot do that in a bath already swimming in them. So before the extraction would work at all, the crew had to bring the acidity of the solution down.<\/p>\n That took about a week.<\/p>\n All of that waiting happens inside a shielded hot cell, which is arguably the most expensive room in the building to occupy. Laetitia Delmau, the ORNL radiochemist behind the change, put the economics in one line: “If you can save time, it saves you a lot of money.” The lab has not published what an hour in there actually costs, so that is the argument rather than a dollar figure.<\/p>\n Delmau walked into the Radiochemical Engineering Development Center during a californium separations campaign in 2015 with a different idea. Use a neutral molecule instead. A neutral molecule holds onto the metals without dumping acid of its own into the solution, so it works in the nitric acid that is already there, and the week of adjustment stops being necessary.<\/p>\n That year she was handed one drop of the running campaign’s waste, diluted far enough that a glovebox would do instead of the hot cell. Two years on, at the next campaign, she got a small amount of the real dissolved-target solution and worked with it inside the cell. Two years later she got more of it and adjusted the concentration of her ligand. She then scaled from 6 percent of the 2023 campaign material to a full-scale test during the 2025 campaign.<\/p>\n One cycle took out 95 percent of the americium, curium and cerium. Same result, half the time.<\/p>\n ORNL did not name the molecule in the August post, which is a slightly frustrating thing to leave out of an announcement about a molecule. The lab’s own citation for Delmau’s Glenn T. Seaborg actinide separations award describes what she built as a diglycolamide-based system replacing the older cation-exchange one, and her published work on californium campaign solutions runs on tetraoctyldiglycolamide, TODGA in the literature. So I am confident about the chemical family and a lot less confident about the exact formulation that ran in 2025.<\/p>\n Federal policy is doing most of the pushing. One executive order has the Department of Energy working to get 10 new large reactors under construction by 2030<\/a> and 5 gigawatts of power uprates at plants already running. Another gives the Army until September 30, 2028 to have a reactor operating on a domestic military base. Several companies are building small fission reactors that need californium-252 to start, and every one of those startups wants a source. All of them trace back to the same Tennessee hot cell.<\/p>\n Then there is the older customer list, which has not gone anywhere: fuel rod scanners, port security, coal analysis, oil well logging and instrument calibration. Oak Ridge is also where the DOE parked a legacy radioisotope generator that a US startup is now recycling into seabed power for underwater drones<\/a>. Isotope inventories in Tennessee have a way of finding new jobs.<\/p>\n So why not just make it more often? ORNL runs a californium campaign every two years, and Samantha Schrell, the lab’s Cf-252 program manager, looked at going annual and decided against it. She cited the money, the hot cell hours it would eat and the way californium-252 decays. A batch loses half of itself to decay in 2.645 years, so producing more of it earlier does little good if it sits in a capsule waiting for a buyer.<\/p>\n So the lab is coming at the problem from two other directions. More reactor cycles with curium targets loaded into them each year, and chemistry that hands back a week on the other end.<\/p>\n The High Flux Isotope Reactor has been running at Oak Ridge since 1965, at 85 megawatts, and ORNL rates it as the country’s strongest reactor-based neutron source. Curium targets go in there to come out as californium. Everything downstream, including Delmau’s new cycle, waits on that schedule.<\/p>\n If you have followed America’s isotope supply at all, you already know this shape. One facility, no domestic backup. The country runs the same single-source math on its medical isotopes<\/a>, and the DOE keeps a handful of other buildings alive for the same reason, including a 1,028-foot separations plant in South Carolina<\/a> where nobody has set foot on the hot side since 1955.<\/p>\n ORNL’s own account of the finishing steps is oddly domestic for a transuranic element. Technicians melt the californium into a pellet and roll it out on a jeweler’s mill until the wire is under a tenth of an inch thick. Segments of that wire get welded into capsules, cleaned, checked for leaks, and handed to an industrial consortium that seals the finished sources and ships them to customers.<\/p>\n The next batch of wire waits on the reactor. HFIR’s status page<\/a> listed the reactor at 0 megawatts on September 1, shut down at the end of Cycle 516B, with startup for Cycle 517 scheduled for September 8.<\/p>\n Image credit:\u00a0Genevieve Martin\/ORNL<\/em><\/strong><\/p>\n","protected":false},"excerpt":{"rendered":" I used to assume a nuclear reactor starts itself. Load the fuel, pull the control rods, let physics take it … <\/p>\nSo what is the acid doing in the middle of all this?<\/h2>\n
Delmau started with one drop in 2015<\/h2>\n
Demand is heading up<\/h2>\n
The reactor behind all of it is sitting at zero<\/h2>\n