Nuclear reactors produce neutrons from fission of \({}^{235}\rm {U}\). Fission produces neutrons with very high energies and these neutrons must be moderated at least partially to allow for the nuclear chain reaction needed to sustain the fission process. In a reactor neutron source, special attention is given to the moderator design to ensure that no higher energy (epithermal) neutrons remain in the spectrum that the reactor is designed to emit. Some neutron facilities also have a secondary moderator that is cooled to cryogenic temperatures to produce very low energy (long wavelength) “cold” neutrons. Other than for magnetic scattering, cold neutrons are not very useful for powder diffraction as the maximum Q value that can be obtained is too low. In fact, the ideal neutron source for reactor-based powder diffraction would have a moderator operating at a somewhat elevated temperature, should someone with very deep pockets ever wish to design an optimal instrument.
The first component in a neutron diffractometer is typically a Soller collimator. These are long thin blades arranged vertically that limit the horizontal divergence of the neutron beam, but do not limit the vertical divergence. The second component is a monochromator, but there are several differences between the monochromator used for neutrons vs. synchrotrons though commonly the same materials, such as silicon and germanium are commonly used. The first difference is that while synchrotrons typically use perfect crystals which offer very high angular and energy resolution, neutron monochromators use highly imperfect crystals that have degraded angular and energy resolution. The intensity loss from higher resolution cannot be afforded. Creating these highly imperfect crystals is something of an art. One needs to take perfect crystals and create highly mosiac domains through some form of mechanical process. Secondly, usually the monochromator is constructed from multiple crystals and these are arranged to focus the horizontally diverging beam back onto the sample.
There is likely to be a second Soller collimator between the monochomator and the sample. Samples may be as large as finger-sized (1 cm by 8 cm) and are commonly placed in vanadium cans, though aluminum is also common. Vanadium at natural abundance has almost no coherent scattering, but does add some background to the diffraction pattern. Aluminum has significant Bragg scattering, but depending on the wavelength this will be in a fairly small number of peaks, but Al does not add any significant amount of background.
Most instruments place a large number of detectors around the sample (more than 20 typically) with a third Soller collimator between the sample and the detector. Detection is typically done with discrete \({}^3\rm {He}\) tubes, which produce an electronic pulse similar to a scintillator detector for x-rays, but other types of neutron detectors exist, including linear position sensitive detectors and area detectors.
The diffractometer will have a resolution curve that has optimum performance when the \(2\theta \) angle for the diffractometer is about the same as the \(2\theta \) angle for the monochromator and it is common to design the instrument to have this angle in the range of 90 degrees.
CW diffraction patterns appear not very different from x-ray diffraction patterns, though closer examination will note that the intensities fall off with \(2\theta \) at a much slower rate than with x-rays due to the flat scattering factors for neutrons. Peaks are usually fairly broad with Gaussian peaks that often do not show any sample-broadening effects. One may note that the instrumental resolution will improve with Q. CW neutrons can offer quite good resolution, but only for a limited Q range.