advantage of this configuration is that the following focusing optics and diffractometer do not have to be moved when the energy is changed.
The monochromators at crystallography beamlines in the future may follow some
of the same principles of the classic DCM, or they might not even exist. The flux
limiting optical element on most crystallography beamlines is the monochromator –
three orders of magnitude can be lost in a silicon DCM. This is due to the extremely
narrow rocking curve, which has been advantageous for spectral purity; however,
this narrow level of bandpass is not necessary for most crystallographic experiments.
The humble sealed tube has a much lower energy resolution, usually quantified as
E/ΔE, equalling 150, whereas a typical Si DCM is closer to 7,000. Reducing this
resolution to a more moderate level will not harm the crystallographic experiment,
but it will allow access to unprecedented flux levels. There are various methods for
increasing the bandpass of a traditional DCM – replacing the silicon crystals with
multilayer mirrors can give E/ΔE values similar to those of a sealed tube. Bending
the silicon crystals will also increase the bandpass. The most radical method for
decreasing the flux loss at the monochromator is to eliminate it completely. The
undulator harmonics on fourth-generation rings are expected to become more
Lorentzian/Gaussian in shape, and it is likely that the harmonic peak will be suitably
narrow not to need further monochromation. The harmonics can be dispersed
spatially using a prism made from aluminium, or some other suitable material, and
the required harmonic selected with a set of slits (the harmonic separator concept
[35]).
Downstream of the monochromator there are usually mirrors to focus the beam in
both the horizontal and the vertical planes to best match the sample size. Again, the
design and configuration of the focusing optics can differ considerably between
beamlines with the type and number of focusing elements varying to provide the
desired beam size. This is commonly performed with either Kirkpatrick-Baez
mirrors [36], a so-called KB pair, or a toroidal mirror. The KB pair consists of two
mirrors that are initially flat, but then bent along their length and that are arranged
horizontally and vertically to focus the beam in those planes. The toroidal mirror is a
cylindrical mirror that is bent along the beam path to focus both horizontally and
vertically. All mirrors, however, work at a grazing incidence angle (usually with an
incidence angle of only a few milliradians), and they have coatings to supress the
reflectivity of the higher energy harmonics from the monochromator (for Si (1 1 1)
the strongest harmonic arises from the λ/3 component, associated with the (3 3 3)
reflection, but it is also possible to observe the ‘forbidden’ (2 2 2) reflections, which
manifests itself as a very weak λ/2 harmonic, for very strongly scattering samples).
This would be detrimental in that resulting data will be contaminated by these
harmonics and could produce effects such as unit cell doubling and tripling. The
mirror coating needs to be selected to suppress both of these harmonics. As the
mirrors are required to suppress harmonics over a large energy range (typically
5 keV to 35 keV), they are often prepared with parallel lanes of different coating
materials, with each lane being used in a specific energy window. The coating
materials have different atomic weights, e.g. Si, Cr or Pt, and by choosing
80
S. J. Coles et al.
The monochromators at crystallography beamlines in the future may follow some
of the same principles of the classic DCM, or they might not even exist. The flux
limiting optical element on most crystallography beamlines is the monochromator –
three orders of magnitude can be lost in a silicon DCM. This is due to the extremely
narrow rocking curve, which has been advantageous for spectral purity; however,
this narrow level of bandpass is not necessary for most crystallographic experiments.
The humble sealed tube has a much lower energy resolution, usually quantified as
E/ΔE, equalling 150, whereas a typical Si DCM is closer to 7,000. Reducing this
resolution to a more moderate level will not harm the crystallographic experiment,
but it will allow access to unprecedented flux levels. There are various methods for
increasing the bandpass of a traditional DCM – replacing the silicon crystals with
multilayer mirrors can give E/ΔE values similar to those of a sealed tube. Bending
the silicon crystals will also increase the bandpass. The most radical method for
decreasing the flux loss at the monochromator is to eliminate it completely. The
undulator harmonics on fourth-generation rings are expected to become more
Lorentzian/Gaussian in shape, and it is likely that the harmonic peak will be suitably
narrow not to need further monochromation. The harmonics can be dispersed
spatially using a prism made from aluminium, or some other suitable material, and
the required harmonic selected with a set of slits (the harmonic separator concept
[35]).
Downstream of the monochromator there are usually mirrors to focus the beam in
both the horizontal and the vertical planes to best match the sample size. Again, the
design and configuration of the focusing optics can differ considerably between
beamlines with the type and number of focusing elements varying to provide the
desired beam size. This is commonly performed with either Kirkpatrick-Baez
mirrors [36], a so-called KB pair, or a toroidal mirror. The KB pair consists of two
mirrors that are initially flat, but then bent along their length and that are arranged
horizontally and vertically to focus the beam in those planes. The toroidal mirror is a
cylindrical mirror that is bent along the beam path to focus both horizontally and
vertically. All mirrors, however, work at a grazing incidence angle (usually with an
incidence angle of only a few milliradians), and they have coatings to supress the
reflectivity of the higher energy harmonics from the monochromator (for Si (1 1 1)
the strongest harmonic arises from the λ/3 component, associated with the (3 3 3)
reflection, but it is also possible to observe the ‘forbidden’ (2 2 2) reflections, which
manifests itself as a very weak λ/2 harmonic, for very strongly scattering samples).
This would be detrimental in that resulting data will be contaminated by these
harmonics and could produce effects such as unit cell doubling and tripling. The
mirror coating needs to be selected to suppress both of these harmonics. As the
mirrors are required to suppress harmonics over a large energy range (typically
5 keV to 35 keV), they are often prepared with parallel lanes of different coating
materials, with each lane being used in a specific energy window. The coating
materials have different atomic weights, e.g. Si, Cr or Pt, and by choosing
80
S. J. Coles et al.