​There are two types of cosmic rays, one produced by neutrons and the other muons. These neutrons or muons make a semiconductor device more vulnerable by causing a failure via soft errors. The team wished to understand how and why this happens, as well as understanding the mechanism that makes these soft errors.

Around six years ago, a team at ISIS conducted experiments using a SRAM memory device​, where they discovered that a positively charged muon can cause effects. Later a team in Japan found at J-PARC that negatively charged muon creates more soft errors and the team wanted to investigate why there is a difference between the positive and the negative.

The team found that the negative muon makes a nuclear reaction inside the semiconductor device and this nuclear reaction significantly affects the soft errors. The muons that remain in a semiconductor device can produce various charged particles such as proton, deutron, triton, alpha and beta, as well as neutrons, that may cause the soft errors and the aim was to understand the yield and energy profile of this emitted radiation. 

Dr Niikura, RIKEN, explained, “ISIS is the only facility where we can do this kind of experiment. It is technically exceedingly difficult to measure the charged particles and gamma-rays together in one measurement, so we must separate them in order to measure the energy and the probability independently. The RIKEN beamline here at ISIS allows us to do this, and, in both experiments, we must make the sample close to the target and we cannot put detectors for charged particles and gamma rays in at the same time."

Dr Kawase, Kyushu University, explained the thickness of the target in this experiment is important as it is only twenty-five micrometres thick. The spread in muon energy makes it difficult to control their stopping depth. Nevertheless, the muons need to be close to the surface as the emitted charged particles can be self-absorbed or lose energy. This is the first time in a study that  low-energy alpha particles have been measured and characterised to understand the effects of soft errors in semiconducting devices.

Once the team understand the probability, they can simulate semiconductor devices and estimate the critical level for soft errors which can be caused by cosmic radiation. They intend to measure the cosmic ray intensity in multiple locations for future experiments. Understanding this will enable better device design to reduce the number of errors that are critically important for most computing and memory devices.