Thick detectors

This is a long journey in the depth of the physics of the X-ray detector to understand the effect of the thickness of the sensor layer on our data. This study has been sponsored by ESRF’s beamlines ID31 and ID22 with the contribution from the SNBL and from the detector group. Most images were taken at ESRF-ID28 on a Pilatus 1M, mounted on a goniometer stage. Thus I would like to thank Veijo Honkimaki for the subject and Thanh Tra Nguyen for the images.

Disclaimer: The Pilatus detector is a photon counting detector with a threshold in energy. In this demonstration, we consider only the photo-electric effect and consider the full absorption of a photon if it interacts with the sensor material.

This document first presents an experiment on ID28 where the detector position is calibrated on its goniometer stage and highlights some odd behavior at high incidence angle where rings looks shifted and broadened and could be attributed to “thickness effect”. In a second time, the detector is modeled as a 2D array of voxels, hence with some thickness, and the signal is then deconvolved to invert this effect. Finally the corrected images are used to validate the correction.

Another approach is presented where the parallax is modeled as a convolution of the beam profile with the absorption profile, leading to some numerical correction to apply to the position of absorption (WIP).

• Pilatus on a goniometer at ID28
• Modeling of the thickness of the sensor
• Least squares refinement of the pseudo-inverse
• Pseudo inverse with positivitiy constrain and poissonian noise (MLEM)
• Deconvolution of the Thickness effect
• Parallax effect