Circular polarization control for the LCLS baseline in the soft X-ray regime

TL;DR

This paper proposes a simple, cost-effective method to generate highly circularly polarized soft X-ray radiation at the LCLS baseline, achieving 10 GW power and 99% polarization, with potential applications to other XFELs.

Contribution

A novel approach using a short APPLE II radiator after the baseline undulator to produce high-power, circularly polarized X-rays with minimal modifications.

Findings

Achieved 10 GW power at the fundamental harmonic.

Attained 99% degree of circular polarization.

Method is low cost, simple, and easily implementable.

Abstract

The LCLS baseline includes a planar undulator system, producing linearly polarized light in the range 0.15-1.5 nm. Polarization control in the soft X-ray region from linear to circular is highly desirable. Several schemes using helical undulators have been discussed for the LCLS. One consists in replacing three of the last planar undulator segments by APPLE III. A second proposal, the 2nd harmonic helical afterburner, uses short, crossed undulators tuned to the second harmonic. This last scheme is expected to be the better one. Its advantages are a high and stable degree of circular polarization and a low cost. Its disadvantage is a small output power and a narrow wavelength range. We propose a novel method to generate 10 GW level power at the fundamental harmonic with 99% degree of circular polarization from the LCLS baseline. Its merits are low cost, simplicity and easy implementation. After the baseline undulator, the electron beam is sent through a 40 m long straight section, and subsequently passes through a short APPLE II radiator. The microbunch structure is easily preserved, and intense coherent radiation is emitted in the helical radiator. The background radiation from the baseline undulator is suppressed by letting radiation and electrons through horizontal and vertical slits upstream the helical radiator, where the radiation spot size is about ten times larger than the electron bunch transverse size. Using thin Beryllium foils for the slits the divergence of the electron beam halo increases by Coulomb scattering, but the beam will propagate through the setup without electron losses. The applicability of our method is not restricted to the LCLS baseline setup. LCLS II or the European XFEL may benefit from this work as well, due to availability of sufficiently long free space at the end of undulator tunnel.