monochromator, which is kept closed during the experiment, since re-equilibration
can take hours.
Other issues for monochromator performance include the harmonic composition
of the transmitted beam, backlash and encoders during scans, and the notorious
monochromator artefacts cleverly called “glitches.” These will be discussed in the
application chapters where relevant.
4.6.3 Crystal Analyzers
For photon-in photon-out experiments, an analyzer is needed to characterize the
energies of the emitted photons. This is a harder task than for the first monochromator, because the synchrotron source beam is nearly parallel, while the emitted
photons come out in all directions. This forces analyzers to resolve different energies
with a single reflection, so they all work relatively close to backscattering. To
capture a reasonable solid angle, most analyzers arrange the crystal surface in a
spherical (Rowland circle) geometry. This is accomplished either by bending a
single crystal to the desired surface (Fig. 4.22), by arranging small flat crystals
close to such a surface, or possibly by grouping such analyzers into bigger arrays
(Figs. 8.23 and 8.26).
For moderate (~100 meV) resolution, the radius of the Rowland circle is usually
on the order of 1–2 m. However, in some cases, say to see vibrational features,
~1 meV resolution is needed. In these cases, rather heroic 10–20 m Rowland circles
are used. To cover a reasonable solid angle at that distance, some resort to multiple
spherical analyzers. Even with a large bending radius, residual strain in the analyzer
crystal would diminish the energy resolution. To avoid that strain, ultra-high-resolution analyzers are typically “diced,” with a single crystal diced into thousands of
pillars (Fig. 4.22). With typical dimensions of 0.9 Â 0.9 Â 5 mm, these pillars
mitigate the strain that would ordinarily result from bending a flat surface into a
spherical one.
Fig. 4.22 (a) The Rowland circle geometry; (b) resolution is improved by combining a moderately
diced analyzer with a position sensitive detector [131]; (c) diced optics with sub-mm lateral
dimensions for high-resolution experiments; (d) a large crystal diced into thousands of individual
pillars [132]
94
4 X-ray Optics and Synchrotron Beamlines
can take hours.
Other issues for monochromator performance include the harmonic composition
of the transmitted beam, backlash and encoders during scans, and the notorious
monochromator artefacts cleverly called “glitches.” These will be discussed in the
application chapters where relevant.
4.6.3 Crystal Analyzers
For photon-in photon-out experiments, an analyzer is needed to characterize the
energies of the emitted photons. This is a harder task than for the first monochromator, because the synchrotron source beam is nearly parallel, while the emitted
photons come out in all directions. This forces analyzers to resolve different energies
with a single reflection, so they all work relatively close to backscattering. To
capture a reasonable solid angle, most analyzers arrange the crystal surface in a
spherical (Rowland circle) geometry. This is accomplished either by bending a
single crystal to the desired surface (Fig. 4.22), by arranging small flat crystals
close to such a surface, or possibly by grouping such analyzers into bigger arrays
(Figs. 8.23 and 8.26).
For moderate (~100 meV) resolution, the radius of the Rowland circle is usually
on the order of 1–2 m. However, in some cases, say to see vibrational features,
~1 meV resolution is needed. In these cases, rather heroic 10–20 m Rowland circles
are used. To cover a reasonable solid angle at that distance, some resort to multiple
spherical analyzers. Even with a large bending radius, residual strain in the analyzer
crystal would diminish the energy resolution. To avoid that strain, ultra-high-resolution analyzers are typically “diced,” with a single crystal diced into thousands of
pillars (Fig. 4.22). With typical dimensions of 0.9 Â 0.9 Â 5 mm, these pillars
mitigate the strain that would ordinarily result from bending a flat surface into a
spherical one.
Fig. 4.22 (a) The Rowland circle geometry; (b) resolution is improved by combining a moderately
diced analyzer with a position sensitive detector [131]; (c) diced optics with sub-mm lateral
dimensions for high-resolution experiments; (d) a large crystal diced into thousands of individual
pillars [132]
94
4 X-ray Optics and Synchrotron Beamlines