Testing an electrostatic deflection innovation: Initial experimental results

When deflecting a charged particle beam into small angles, electrostatic deflection is commonly used. For larger angles, magnetic is the usual choice mainly because the deflection aberrations are two- or threefold less. Methods to reduce electrostatic deflection aberrations have been proposed over the years but a major solution has not been identified. This subject has recently been revisited and, based on computations, it has been proposed that electrostatic deflection aberrations can be reduced 10-fold if the beam is injected asymmetrically into the gap between two conventional shaped oppositely charged conducting plates. According to this theory, with the proper injection offset, the only surviving aberration is astigmatism that is totally correctable with a quadrupole prior to deflection. The predicted optimal offset was either 33% or 42% (depending on the method of calculation) from the center toward the attracting plate. We have built a demountable test bed to experimentally determine if the innovation is correct. The beam is deflected by shaped plates according to a previously reported design. This plate design was predicted to provide ±38.1° of deflection for a beam 0.050 in. in diameter with undetectable deflection aberrations at optimal offsets. Compared to the calculated design, the voltages used in the demountable test bed are smaller (6 compared to 20 kV), the throw distance from the plate entrance to the screen is smaller (4.5 compared to 12 in.), and the deflection angle is ±28.2° but otherwise the setups are very similar. Initial data from the test bed are deflection of a 0.010 in. diam probe after injection into the deflection plate gap at various offset positions. It is found that 1/3 of the way from center toward the attracting plate, there is a 0.2 in. wide section of the deflection versus offset curve that is flat indicating a segment in which the aberrations are not detectable. Based on these initial results, the concept is valid and the injection offset was where it was predicted to be.