A New Oata Processing Method For Measuring Atomic Inner-Shell Ionization Cross Sections Impacted By KeV Electron Using Thin Films With Thin Substrate Target

The experimental research on atomic inner-shell ionization cross sections(ISICSs)impacted by electron near the threshold energy plays an important role in both engineering applications and theoretical studies.On the one hand,it can test the correctness of the theoretical model of the interaction between electron near the threshold energy and atomic,and promote the development of the theoretical research.On the other hand,accurate ISICSs can supplement the fundamental data required in fields such as radiation medicine,radiation physics,and materials science.The thintarget method can effectively improve the experimental accuracy.When the incident electron energy is greater than 3-5 times the ionization threshold energy,the x-ray production cross sections(XRPCSs)measured by different researchers using the thintarget method are consistent with the predicted value of the distorted wave Born approximation(DWBA).However,when the incident electron energy is close to the ionization threshold energy,the experimental data published by different researchers compared with the DWBA theoretical XRPCSs will also occasionally exhibit significant discrepancies,which are greater than the quoted experimental uncertainty.This is because the target in this case can no longer be considered a thin target,the experimental characteristic x-rays yield and XRPCSs no longer simply follow the linear mapping relationship in the calculation formula of the thin-target method.Therefore,it is necessary to develop a new data processing method to calculate the experimental XRPCSs.This paper proposes that the mapping relationship between characteristic x-rays yield and XRPCSs can be obtained based on the neural network method,and then the experimental yield can be input to solve the experimental XRPCSs.Firstly,the neural network was built.Then,the simulated yield data set and cross-section data set were obtained by Monte Carlo simulation software,and the constructed neural network was used to train the data sets and fit the mapping relationship between yield and XRPCSs.Finally,the experimental yield of the target element under each incident electron energy are input into the trained neural network model to obtain the corresponding XRPCSs.In this paper,the neural network method is applied to obtain the K-shell ISICSs of Al by 4-20 keV electron impacted,the total L-shell XRPCSs of Ag and Sn by 4-30 keV electron impacted,and the Mαβ XRPCSs of W and Bi by 3-30 keV electron impacted.The DWBA theoretical XRPCSs are also presented in the paper as a contrast.When the incident electron energy is close to the ionization threshold energy,the experimental XRPCSs obtained by the neural network method is generally higher than that obtained by the thin-target formula method,and are in good agreement with the predicted value of the DWBA.When the incident electron energy is greater than 3-5 times to the ionization threshold energy,the experimental XRPCSs obtained by the neural network method is generally close than that obtained by the thin-target formula method.The DWBA theoretical model can well predict XRPCSs of these elements.The neural network method can be used as a new data processing method for measuring atomic inner-shell ionization cross sections impacted by keV electron using thin films with thin substrate target.