NIST's cryogenic sensors resolve the X-ray lines that blur plutonium and uranium assay

An assay reads a spectrum and asks which lines belong to which isotope. Some of the elements in that spectrum emit X-rays at the same energies as the gamma rays being counted, and that masks the signal ^[2]. Knowing precisely where the X-ray lines sit is what lets an analyst subtract them, and the NIST-led group measured them in the overlap band for uranium, plutonium and neptunium, a decay product of uranium ^[1].

A transition edge sensor is a superconducting film held a fraction of a degree above absolute zero, right at the boundary between zero resistance and ordinary metallic behaviour. A single X-ray photon landing on it deposits a minuscule amount of heat, enough to push the resistance up sharply ^[3]. The resistance is directly proportional to the photon's energy, which is what gives the device its energy resolution ^[4].

The reported gain is on the uncertainty of those X-ray energies, which NIST's announcement puts at a reduction of "one-third to one-eighth" compared with previous measurements ^[5]. Read as a factor, that means uncertainties fall to between a third and an eighth of the old values, or 67% to 87.5% less ^[18]. The announcement gives the range without saying which elements or emission lines sit at the eight-fold end ^[20].

What the assay has to resolve is a ratio, because isotope ratios indicate whether material is meant for a power plant or a weapon ^[10]. Uranium-235 is 0.7% of natural uranium, a few percent in reactor fuel, and 90% in weapons-grade material ^[11]. That is a span of about 129 in relative abundance ^[17], and the X-ray background is what a gamma detector has to remove before it can place a sample anywhere on it ^[9].

Speed is the second claim. Fuel composition has to be reassessed at each step of a multistep process before the next one starts, so a faster measurement should shorten the hold-up time between steps, with the potential to raise efficiency and reduce costs ^[13]. Jonathan Dean, a physicist at NIST and the University of Colorado Boulder ^[8], said rapid assessments should come in handy at nuclear power plants ^[12].

The refrigerator is the limit on where this runs. Cooling equipment is too bulky to be handheld, though a TES array will operate at any site with enough electricity to power it ^[14]. Where no TES is available, samples go to a laboratory that has one. "Our instruments are compatible with both approaches," Dean said ^[15]. Working with Los Alamos National Laboratory, NIST has installed TES detectors at three Department of Energy laboratories ^[16].

The thing this does not tell you is how much an inspector's accounting error bar actually shrinks. The improvement is reported as a reduction in X-ray energy uncertainty, and no resulting figure for isotope-ratio or material-accounting uncertainty is given ^[19]. The benefit is stated by the researcher: "Our measurements support international nuclear safeguards by enabling more precise accounting of material in nuclear facilities," Dean said ^[7]. The work appeared in Physical Review Letters, with co-authors at NIST, the University of Colorado Boulder, Los Alamos, Houghton University in New York and the Kastler Brossel Laboratory at Sorbonne University in Paris ^[6].