Data Acquisition

Course Material Index  Section Index  Previous Page  Next Page 

Data Acquisition

Wavelength Calibration

For a laboratory source, the wavelength is exactly known from the constant and precise values of Kα₁, which are given in the International Tables for Crystallography. As already mentioned, the neutron wavelength of powder diffractometer on reactor source is determined by the choice of monochromator crystal and its associated planes, (hkl), plus the monochromator angle, 2θ. Whereas the selected d spacing from the crystal monochromator planes is usually known precisely, the precise value of the instrument take-off angle is not known until the instrument is calibrated.

Each wavelength available from the monochromator of a powder neutron diffractometer should therefore be calibrated. In addition, it is good laboratory practice to recalibrate the wavelength each time the monochromator is moved (or rotated to a new position). The wavelength is best calibrated using a standard powdered sample of, say, silicon or nickel metal. From the position of the diffraction peaks, both the wavelength and any diffractometer zero errors should become apparent. It is not uncommon given their large size (and weight) for powder neutron diffractometers to be run with large 2θ zero errors. The latter can usually be corrected for after the data acquisition. Measurement of a standard sample also allows the user to check that the instrument is working correctly, and that there are no spurious background peaks due to sample environment or impurity wavelength(s).

Detector Calibration

With both high-flux and high-resolution powder neutron diffractometers, a second type of calibration is required. For the PSD type detector used on the high-flux instruments, it is important to measure the efficiency of each wire (as with a PSD X-ray detector). For neutrons, the ideal sample for producing a flat background at all scattering angles is a rod of solid vanadium metal since it has a very low coherent cross-section, but a relatively high incoherent one. If the detector is scanned during the calibration procedure, even the very weak Bragg peaks from the vanadium rod can be avoided. The efficiency of each wire is simply proportional to the inverse of the counts from the vanadium sample.

With a bank of step-scanning detectors, both the efficiency and precise relative angular position of each detector has to be determined. (The efficiency measurement takes into account the transmission of the Soller collimator in front of each detector as well as the efficiency of the detector itself.) The ideal procedure for calibration is to step scan each detector through an identical low-angle reflection, for example, the Si 111. A calibration table can then be constructed which makes the relative position and height of the peak identical for each detector. With very large detector banks, this procedure may not be possible without the detector bank either passing through the incident beam or hitting the concrete shielding. Under these circumstance, several peaks from a standard sample may be used and the results combined.

Data Collection

So far we have discussed the neutron radiation source, its monochromatization, the powder sample and the detection of its diffraction pattern. As in the X-ray case, data collection times may vary from seconds to days depending on the type of experiment. Clearly for the high-flux diffractometers acquisition times are going to be relatively short compared to using a step-scanning high-resolution instrument. The major difference in data collection with the X-ray case is that powder neutron diffraction patterns do not suffer intensity diminution with increasing scattering angle. Therefore most of the points in the powder diffraction pattern should be measured for a similar time interval. In practice, with the large step-scanning detector banks the very high angle data is measured for slightly less time, but at the very highest angles, the resolution of the instrument is rapidly deteriorating. The data from each of the detectors of a step-scanning diffractometer is normally merged into a single diffraction pattern (assuming that the individual detector efficiency factors and electronic background counts are not too dissimilar)

For most neutron powder diffraction experiments the value of the absorption coefficient, μ, multiplied by the sample diameter, 2R, is very small. For most samples, an absorption measurement is therefore not required. In a few instances, it may be useful to measure the absorption of the sample so that this factor can be taken into account in any data-fitting procedure. This is readily achieved by measuring the transmission of the sample with the detector using a small incident slit (with an attenuated beam if so needed).

Given the ease with which the wavelength can be changed on many powder neutron diffractometers, it is often useful to measure the pattern with more than one wavelength. For example, a data set collected at long wavelength may be useful for indexing, while one collected at a shorter wavelength may be required for structure refinement. These particular aspects of data collection will be discussed further in later sections.

Course Material Index  Section Index  Previous Page  Next Page