GEM Detector Calibration Philosophy for Disordered Materials

There are web pages which give the GEM detector calibration files for gudrun and for ATLAS.

A time-of-flight neutron diffractometer depends upon having correct Lsinq values for each detector element. (L and 2q are the total flight path and scattering angle for the detector element.) These calibration files contain suitable L and 2q values for each detector element. The initial values used were obtained from the engineering design of GEM. These values were then refined using neutron diffraction data. The aim of the refinement is to make the peaks in the correlation function, derived by the standard method, to occur at the correct distances. There are two stages to this process, firstly a relative calibration, and secondly an absolute calibration.

Relative calibration

Firstly, the calibrations are produced so that the first moment of a Bragg peak is the same for different detectors at different scattering angles. The reason is that the resolution function broadens the ideal diffraction pattern and we need a calibration which does not shift area under features in the diffraction pattern – this is important because we are going to Fourier transform the diffraction pattern. This procedure produces a consistent relative calibration for the set of detectors. Note that this calibration is not optimal for Rietveld refinement, or any other analysis method which explicitly includes the asymmetric resolution function which applies for t-o-f neutron diffraction.

Absolute calibration

Secondly, an absolute calibration is obtained by analysis of data for a crystalline standard. The correlation function is measured and the positions of its peaks are used to determine the calibration. Only correlation function peaks whose position depends only on the lattice parameters are used. We do not use peaks where the distance depends upon any (x,y,z) atomic coordinates, because these are less reliable than the lattice parameters.

Last updated on 05 May 2006 by Alex Hannon (a.c.hannon@rl.ac.uk)