This paper reports experiments and computer modeling that together suggest energetic alpha particles can start short collision cascades in compressed deuterium gas that produce measurable D–D (deuterium–deuterium) fusion neutrons. The authors placed a small radioactive alpha source inside a steel chamber filled with pressurized D2 and watched for neutrons with two helium‑3 detectors. They also ran detailed Monte Carlo simulations of the same setup.

In the experiment a 5‑milliCurie polonium‑210 alpha source sat inside a cylindrical steel chamber (12 cm long, 3 cm diameter) filled with D2 gas at several bar of pressure. Neutron counts were recorded for 18 hours with two Mirion SN‑S helium‑3 neutron detectors. With only D2 gas the data showed an excess of about 74 neutrons above background, which the authors translate to roughly 2.24 neutrons per second. When lithium deuteride (LiD) was added to the same cell the excess rose to about 268 neutrons, or 8.1 neutrons per second. Because D–D fusion has two main branches with about equal probability, those neutron rates correspond to roughly 4.5 and 16.2 fusion events per second, respectively.

To support the measured signal the team ran Monte Carlo n‑Particle (MCNP) transport simulations that included the chamber geometry, source activity, gas pressure (4–10 bar), and detector placement. The simulations used ~5.5 MeV alpha particles, a stopping range of about 1.7 cm in D2 at 10 bar, and recorded how many deuterons and other ions were given kinetic energy by collisions. For very large simulated source samples (10^11 alphas) the code produced tens of millions of energized deuterons; the authors emphasize that “created” in the simulation means particles were given kinetic energy during collisions, not that new nuclei were produced. The simulated neutron yields matched the measured counts to within about 5.4% and reproduced the detector response within quoted uncertainties.