Adaptive electron beam shaping using a photoemission gun and spatial light modulator

The need for precisely defined beam shapes in photoelectron sources has been well established. In this paper, we use a spatial light modulator and simple shaping algorithm to create arbitrary, detailed transverse laser shapes with high fidelity. We transmit this shaped laser to the photocathode of a high voltage dc gun. Using beam currents where space charge is negligible, and using an imaging solenoid and fluorescent viewscreen, we show that the resultant beam shape preserves these detailed features with similar fidelity. Next, instead of transmitting a shaped laser profile, we use an active feedback on the unshaped electron beam image to create equally accurate and detailed shapes. We demonstrate that this electron beam feedback has the added advantage of correcting for electron optical aberrations, yielding shapes without skew. The method may serve to provide precisely defined electron beams for low current target experiments, space-charge dominated beam commissioning, as well as for online adaptive correction of photocathode quantum efficiency degradation.

Figure 1

532 nm diode laser light enters from the bottom along $z$ and is polarized along $x$. A quarter-wave plate and SLM act as a polarization rotator with spatial dependence, which shapes the light when used with a polarizing beam splitter (PBS). The surface of the SLM is then 4-f imaged onto an intermediate plane to preserve the beam divergence, and then this intermediate plane is imaged with a single long focal length lens onto either the photocathode or a CCD. An ultrahigh vacuum (UHV) mirror reflects light to the center of the photocathode. A dc bias of 300 kV accelerates the photoemitted electrons.

Figure 2

Coordinate systems of both the SLM and target planes, showing a possible skew transformation between the two.

Figure 3

(a) The initial near-Gaussian unshaped laser profile, $U$. (b) Shaped profile, a 2 mm diameter flattop, prior to feedback, $I^0$. (c) After seven feedback iterations, the rms error is reduced to 10%. Each image is scaled to its respective maximum.

Figure 4

Layout of the electron beam line, including the photogun (purple equipotential lines), imaging solenoid, and BeO fluorescent viewscreen and fluorescence profiling camera.

Figure 5

(a) Resultant beam profile when the flattop shown in Fig. 3 is transported to the cathode and the beam imaged on the viewscreen. (b) Beam result when a pinhole is placed in the image plane of the second lens downstream of the SLM. (c) 1D profiles along the lines shown in (a) and (b). (d) Background of (a), taken from yellow boxed region. (e) Background of (b), taken from the yellow boxed region.

Figure 6

(a) Laser target (see text for details). (b) Final measured laser distribution. (c) Resultant beam distribution, rotated by 47 degrees in after acquisition, with no median filter.

Figure 7

(a) Initial unshaped electron beam imaged on the fluorescent screen. (b) Final output of electron beam-based feedback, with no profiling of the drive laser. (c) Progress of feedback versus iteration. The distribution shown in (b) is circled.

Figure 8

(a) Electron beam distribution created with beam feedback: A truncated Gaussian. (b) A beam shape for a hypothetical photocathode with a local QE dip, thus requiring a local maximum on the laser power (electron beam shown).

Figure 9

(a) Phase applied to the SLM to create the shape in (b). The phase ranges from $[0,\pi ]$. (b) Electron beam distribution created with electron beam feedback, which does not directly profile the laser. The target distribution is that of Fig. 6, scaled by the electron magnification.

Figure 10

(a) Simulated laser profile output for setting the MMDM actuator voltages to alternating zero and maximum on each radial hexagonal ring group, using a Gaussian input laser. (b) Experimental output for the conditions simulated in (a).

Figure 11

Simulated output of a 3 mm diameter flattop assuming an ideal MMDM with analytic solution, or assuming the MMDM is composed of varied numbers of perfect actuator rings. The MMDM used in Fig. 10 is four-ring, with each ring being composed of multiple hexagonal actuators.