Historical research guide · 0950 · 1999
Measurements of low-energy (d,n) reactions for BNCT
N. Colonna · L. Beaulieu · L. Phair · G. J. Wozniak · L. G. Moretto · W. T. Chu · B. A. Ludewigt
In one minute
Boron neutron capture therapy needs a very particular kind of neutron: slow enough to be captured in a tumour, not so fast that it burns the skin on the way in. Reactors make them, but a small accelerator would be cheaper, safer and installable in a city hospital — so the question is which nuclear reaction to fire the beam at. Colonna and his colleagues at the Lawrence Berkeley National Laboratory, with Bari’s nuclear-physics institute, went and measured it. Using 1.5 megaelectronvolt deuterons from the laboratory’s 88-inch cyclotron, they counted the neutrons coming off beryllium, carbon-12 and carbon-13 targets, timing each one against the cyclotron’s own radio-frequency clock to get its energy, at five angles at once. The answer is specific: beryllium is spoiled by a hard 3.6 megaelectronvolt component, carbon-12 is clean but too weak, and carbon-13 is the one worth pursuing — plentiful, low in energy, and reachable with a very small machine.
Connection
Chapter 12 is about getting useful nuclear output at the lowest beam energy you can arrange, and this is that question asked with a detector in front of it: what does a deuteron at one and a half million volts actually produce, reaction by reaction, angle by angle. It is also a clean model of the discipline the chapter depends on — the yield, the spectrum and the contamination are measured and reported together, so the next person can build on the number instead of the hope.
From the source
The target list is chosen by physics before anything is measured. To keep the accelerator small the deuteron energy must stay low, which restricts the search to reactions with a positive or only slightly negative Q-value; the Coulomb barrier then restricts the targets to light elements; and the requirement of a stable, mechanically and thermally convenient material leaves essentially lithium-6, beryllium-9, boron-10, carbon-12 and carbon-13.LBNL-42219, section 1, Introduction
The measurement is a time-of-flight measurement against the accelerator’s own clock. Deuteron beams of 1.5 megaelectronvolts and proton beams of 2.5 megaelectronvolts were taken from the 88-inch cyclotron at Berkeley; neutrons were counted in five liquid-scintillator cells 12.7 centimetres across and 5 centimetres thick, set at 0, 30, 60, 90 and 140 degrees at half a metre from the target, and each neutron’s energy was read from its flight time relative to the cyclotron radio-frequency signal. A detection threshold near 10 kiloelectronvolts electron-equivalent was held throughout so that neutrons as slow as 100 kiloelectronvolts would still be counted.LBNL-42219, section 2, Experimental Method
Beryllium-9 looks promising and is not. The reaction does give the strong 400 kiloelectronvolt neutron peak that had been proposed as a therapy source, but a second peak sits at about 3.6 megaelectronvolts, from the 2.15 megaelectronvolt level of the boron-10 residue. After efficiency correction the high-energy neutrons are about 38 per cent of the total yield for a thin target and about 50 per cent for a thick one, and the authors judge it unlikely that raising or lowering the bombarding energy will bring that fraction low enough for therapy.LBNL-42219, section 3A, the beryllium-9 reaction
Carbon-12 is the clean one and the weak one. Its slightly negative Q-value of minus 0.28 megaelectronvolts gives a clean low-energy emission peaked at 500 kiloelectronvolts with no high-energy contamination at all, and carbon’s very high melting point makes it an easy target to cool — but the yield at every angle is too small for practical use, with preliminary estimates calling for a deuteron beam of several hundred milliamperes at this energy.LBNL-42219, section 3, the carbon-12 subsection, and Table 1
Carbon-13 is the result. At 1.5 megaelectronvolts the total neutron yield of the carbon-13 reaction is about 1.9 times ten to the eighth neutrons per microcoulomb of beam, some 30 per cent below the 2.4 times ten to the eighth that published cross-sections imply — a gap comparable to the uncertainty in their own absolute normalisation. About 30 per cent of the neutrons come out above 1 megaelectronvolt, but most of those sit just above 1 megaelectronvolt and only 6 per cent exceed 2 megaelectronvolts. Preliminary moderator calculations indicate that a beam current below 100 milliamperes would keep treatment time under an hour, at a beam quality superior to currently available reactor beams.LBNL-42219, section 3, the carbon-13 subsection, and Table 1
What to watch is stated by the authors as an invitation rather than a conclusion. Because of the limited resolution for high-energy neutrons and the uncertainty on the absolute yield, they write that the present results are by no means conclusive and are meant to stimulate further discussion and measurement — while recording their own verdict that carbon-13 is the only potentially interesting alternative among the deuteron-induced reactions studied so far. The measurements that would settle it are named: a better absolute yield, and a full simulation of reflection and moderation.LBNL-42219, section 4, Summary and Conclusions
The way in
https://doi.org/10.1118/1.598599How to cite it
N. Colonna, L. Beaulieu, L. Phair, G. J. Wozniak, L. G. Moretto, W. T. Chu, B. A. Ludewigt (1999) Measurements of low-energy (d,n) reactions for BNCT doi:10.1118/1.598599