Characterization of a modular broad beam ion source

We characterize the performance of a modular broad beam ion source by energy resolved mass spectrometry and beam profile measurements. Using the same source housing and grid system, we performed our experiments powering the source with a hot filament or an ECR excitation, respectively. In the ion energy distribution we detect various peak structures reflecting the potential profile across the ion source as well as different charge exchange and dissociation processes occurring in the beam. The position of these peaks on the energy scale allows conclusions to be formed about the original charge state of the ion, when produced by charge exchange, or on the original molecule, when produced by dissociation. The most important contribution to the ion beam is caused by process gas ions with important impurities from source materials appearing with reactive process gases. Using electronegative gases, a large amount of negative ions generated in the beam is observed. The performance of both excitation types due to the beam parameters is very similar. A comparable power supplied to the source plasma delivers similar beam currents and profiles. The resulting beam current is determined by the density and the potential structure in the source plasma. The axial potential gradient in the plasma, and with it the resulting beam current, is strongly affected by the total voltage applied between the grids and can be derived from the shape of the main peak in the ion energy distribution. The detected radial beam profile is determined simultaneously by the ion optical parameters of the grid system and the space charge potential in the beam. For a fixed grid geometry the shape of the plasma sheath at the screen grid is an important factor controlling the source performance. The shape of the plasma sheath is simultaneously controlled by the voltage difference between the grids and the Debye length. Changes in the shape can be derived from the ion energy distribution with the experimental results being in good agreement with numerical beam simulations. The only important difference between both excitation types is that using the Kaufman type excitation the potential of the source plasma is exactly fixed by the beam potential, whereas with the ECR source it deviates by several tens of volts depending on microwave power and process gas.