Accuracy comparison of fitting function applied to air-borne alpha detection spectrum

Highlights: • Accuracy of alpha peak shapes from three fitting functions was analyzed. • Spectra was derived with PIPS detector-based airborne detection and MC simulation. • Created fitting curves for low-energy tailing from alpha particles react in air. • Significant differences were found in fitting accuracies of three fitting functions. • A more accurate airborne detection system could be developed based on the results. The ability to detect airborne alpha particles in real time could greatly enhance the safety of workers in radioactive environments. To assist in the development of an air-borne detection system, this study sought to overcome the difficulties in spectral analysis caused by the strong interaction of alpha particles with matter. Three modified Gaussian fitting functions were applied to a radon progeny alpha spectrum and cross-referenced to counteract the overlap phenomenon produced by low-tailing of the alpha peak. The radon progeny alpha spectrum was derived using MCNP6 and a PIPS (passivated implanted planar silicon) detector. The radionuclides were determined pertaining to two cases, where case 1 included ²²⁰Rn, ²¹⁶Po, and ²¹²Bi, and case 2 included ²²²Rn and ²¹⁸Po. The RMS (root mean square) and reduced chi-square values were compared for accuracy. The convolution of an exponential low-energy tail with a Gaussian distribution equation provided 6.68% lower relative error than when two Gaussian distribution functions were combined according to centroid conditions. The relative error was 0.84% lower than with a Gauss function, indicating a spectrum-peak Gaussian distribution. By constructing the algorithm using this equation, it was possible to obtain high-accuracy analysis data even with low-resolution spectrums in the air.