Polarisation optimisation for ferromagnetic diffraction: A case study of iron

The magnetic effect from a ferromagnet can be measured in a non-resonant X-ray diffraction experiment by controlling the polarisation of the incident beam with a phase retarder. However, the results strongly depend on the parasitic polarisation caused by fluctuations of the orbit. We found that by inserting a Si (3 3 3) channel-cut polariser with a Bragg angle close to 45° in front of the diamond phase-plate, most of the unwanted polarisation was rejected and the magnetic effect was enhanced. We demonstrate the performance of this experimental configuration for non-resonant ferromagnetic diffraction with the example of iron. The ratio of magnetic to charge scattering from ferromagnetic iron was increased up to three times for the (2 2 0) reflection of Fe by combining the channel-cut crystal and the phase-plate. Also, by setting the ψ angle of the phase retarder to 11° rather than 45°, the measurements of the magnetic effect were less affected by the presence of multiple reflections occurring in the diamond phase-plate.