Ultraviolet laser transverse profile shaping for improving x-ray free electron laser performance

The photocathode rf gun is one of the most critical components in x-ray free electron lasers. The drive laser strikes the photocathode surface, which emits electrons with properties that depend on the shape of the drive laser. Most free electron lasers use photocathodes with work function in the ultraviolet, a wavelength where direct laser manipulation becomes challenging. In this paper, we present a novel application of a digital micromirror device (DMD) for the 253 nm drive laser at the Linear Coherent Light Source. Laser profile shaping is accomplished through an iterative algorithm that takes into account shaping error and efficiency. Next, we use laser shaping to control the X-ray laser output via an online optimizer, which shows improvement in FEL pulse energy. Lastly, as a preparation for electron beam shaping, we use the DMD to measure the photocathode quantum efficiency across cathode surface with an averaged laser rms spot size of $59\mu \mathrm{m}$. Our experiments demonstrate promising outlook of using DMD to shape ultraviolet lasers for photocathode rf guns with various applications.

Figure 1

Simplified layout of the LCLS injector laser imaging system with the DMD and grating installed.

Figure 2

Original beam profile (left) and Stanford tree profile at VCC (right).

Figure 3

Flow chart of the shaping algorithm, with example VCC images of the laser beam in the original shape, after the first iteration, and after the fourth iteration.

Figure 4

Example VCC images of laser shaping. Top left is the original laser profile. Top right is cut-Gaussian shaping (error: 18%, efficiency: 77%). Bottom left is flat-top shaping (error: 23%, efficiency: 71%). Bottom right is parabolic shaping (error: 19%, efficiency: 67%). Note that the color scale for the shaped profiles (top right, lower left, and lower right) is 60% of the color scale of the original beam image (top left), in order to show stronger contrast. Since the DMD can only reduce intensity, the fine structures in the shaped profiles, for example the lower right corner of the flat-top image, come from the corresponding low intensity part of the original beam. This indicates that the shaping performance also depends on the uniformity of the initial beam.

Figure 5

Left: initial beam profile corresponding to $175\mu \mathrm{J}$. Middle: optimal beam profile corresponding to $372\mu \mathrm{J}$. Right: gas detector reading evolution through 39 iterations.

Figure 6

The evolution of the FEL gas detector reading as a function of each of the variables.

Figure 7

Drive laser profiles for preliminary emittance measurements.

Figure 8

Illustration of raster scan to obtain a QE map across the laser illumination area on cathode.

Figure 9

QE map from combination of two partially overlapping raster scans. Blue region is not scanned by the laser.