Electron spectroscopy with a commercial 4H-SiC photodiode

Abstract

A Commercial-Off-The-Shelf (COTS) 4H-SiC p-n photodiode (sold as a UV detector) was investigated as detector of electrons (βparticles) over the temperature range 100 °C to 20 °C. The photodiode had an active area of 0.06 mm 2. The currents of the photodiode were measured in dark condition and under the illumination of a 63 Ni radioisotope βparticle source (endpoint energy = 66 keV). The photodiode was then coupled to a custom-made low-noise charge-sensitive preamplifier to make a direct detection particle counting electron spectrometer. 63 Ni βparticle spectra were accumulated with the spectrometer operating at temperatures up to 100 °C. The quantum efficiency of the photodiode as well as the spectrum expected to be detected were calculated via Monte Carlo simulations produced using the CASINO computer program. Comparisons between the simulated and detected 63 Ni βparticle spectra are presented. The work was motivated by efforts to apply COTS technologies to develop low-cost space science instrumentation; a low-cost electron spectrometer of this type could be included on a university-led CubeSat mission for space plasma physics and magnetosphere experiments.

Hard X-ray and γ-ray spectroscopy at high temperatures using a COTS SiC photodiode

“…Given the manufacturer stated donor concentration (8 × 10 15 cm -3 ) [29] and the abrupt junction approximation [30], a depletion width of 3.7 µm at 100 V would have been expected, although no uncertainty in the concentration was stated by the manufacturer. However, despite the depletion region seemingly being only a portion of the epitaxial layer thickness, results previously reported for a detector of the same type (but in that case used as an electron spectrometer [31]) suggested that the entire epitaxial width of the detector was active for radiation detection. It is possible that a significant portion of the epitaxial layer was not depleted and that despite this the whole of the epitaxial layer contributed to the detected signal since the electron and hole recombination lengths in ND-NA = 8 × 10 15 cm -3 n-type and NA-ND = 6 × 10 18 cm -3 p-type doped SiC are ≈ 1.1 µm [32] and ≈ 20 µm [33] respectively.…”

“…Measurements of capacitance as a function of reverse bias suggested that the detector could be depleted to 2.41 µm ± 0.03 µm at 20 °C, short of what was expected from the abrupt junction calculation [30] (3.7 µm at 100 V). Notwithstanding this, a previous electron spectroscopy study suggested that most of the epitaxial width (5.15 µm) was depleted and active [31]. It was also noted that carrier recombination lengths in the material were sufficiently large to address the difference in depletion width calculated from the capacitance -voltage measurements and the previous study that suggested the entire epitaxial thickness was active.…”

“…The reason for the discrepancy in depletion width between the measurement of capacitance as a function of reverse bias, the doping density calculation, and the epitaxial layer thickness are unknown. However, given that the electron spectroscopy measurements of Zhao et al [31] indicated that the whole of the epitaxial thickness was active, the full thickness of the epitaxial layer (5.15 µm) was assumed to be active and depleted for the quantum detection efficiency calculation below. Nevertheless, the simplicity of this assumption is recognized.…”

AlGaAs two by two pixel detector for electron spectroscopy in space environments

“…Secondary electron generation was not included in the simulations. The same CASINO settings and presently reported methodology were used in refs [35], [36], and [37], to simulate similar situations in GaAs, SiC, and InGaP electron detectors. Fig.…”

“…A bank of 14 computers, each with an Intel i7-6700 (4 cores, 3.40 GHz) processor and 32 GB of random access memory, was used to perform the simulations. The presently reported methodology was the same as that used in refs [35], [36], and [37].…”