Stripping units take a key role in the neutral particle analyzer (NPA). A renovated gas-stripping unit was constructed for the newly designed E//B NPA. Using \({\textrm{H}}_2\) as the working gas, we measured the gas inlet pressure ( \(P_0\) ) and vacuum chamber pressure ( \(P_3\) ). The pressure distribution inside the gas-stripping room was calculated with Ansys Fluent, using the measured \(P_0\) and \(P_3\) as boundary conditions. The stripping efficiency of the stripping unit was then simulated utilizing the Geant4 Monte Carlo code for the H and D particles. The pressure \(P_0={40\,\textrm{Pa}}\) , which is one-sixth of what found in the previous design and corresponds to a thickness of \({1.27}\times 10^{17}{\mathrm{atoms/cm}}^2\) , was obtained as the optimum working pressure for the upgraded stripping unit. An 50 kV electron cyclotron resonance (ECR) ion source platform was designed and constructed for E//B NPA calibration, and its performance has been measured. Using the ECR ion source platform, we measured the efficiency of the stripping unit through an inverse experiment with proton beams. We compared the current ratios of measurements with and without \({\textrm{H}}_2\) gas to Geant4 simulation results. We found adequate agreement between the overall trends of the experiment and the simulation. The significant deviation for incident energies below 20 keV may result from the scattering effects of low-energy protons, leading to reduced accuracy in single-scattering physics in Geant4 simulations. Applying the scattering corrections observed in the reverse experiments obtains more accurate stripping efficiencies for H and D atoms in the energy range of 20–200 keV and the global efficiency with the maximum values of 95.0% for H atoms and 78.9% for D atoms at 200 keV.
