A transfocator for X-ray focusing

Since their development a little over 12 years ago [1], the use of X-ray refractive lenses has rapidly expanded [2,3] and they are now in common use on synchrotron beamlines. Unlike visible light, the real part of the index of refraction of X-rays is close to and slightly below one, meaning that a large number of concave lenses is needed to produce a marked focusing effect. Compared to other focusing elements, refractive lenses present several attractive features, being simple to align and relatively insensitive to misorientations. As in-line optics, they can be conveniently inserted and removed from the beam to allow fast switching of beam size. As the index of refraction is small and energy dependent, a substantial, well-defined number of lenses is necessary to focus X-rays of a given energy at a given distance. For this reason, systems with a tunable number of lenses have been proposed [4,5] to provide permanent energy tunability.

Figure 1. Implantation of the in-vacuum transfocator. Other beamline optics are not shown in the figure.

Based on the success of an in-air version, an in-vacuum transfocator (Figures 1 and 2) has now been implemented at beamline ID11. Installed in vacuum, the transfocator benefits from being closer to the source (31.5 m), thus capturing a larger proportion of the diverging X-rays. It is also water cooled for optional use in the white beam. The in-vacuum transfocator consists of nine pneumatically actuated, water cooled, cylindrical cartridges containing 1, 2, 4, 8, 16, and 32 beryllium lenses, and 32 and 64 aluminium lenses. The combination of these cartridges allows complete tunability between 25 and 125 keV at 94 m, and 25 to 75 keV at 42 m (Figure 3). At 94 m from the source, gains on the order of 5×10⁴ are achieved with respect to an unfocussed beam. The gain in the microfocused beam at 42 m from the source is even more substantial, over 10⁵.

Figure 2. Design and mounting of the in-vacuum transfocator.

The in-vacuum transfocator is very flexible, and has been used in several different configurations, either as a stand alone focusing device in the monochromatic beam, giving vertical spot sizes ranging from 7 to 42 μm depending on the focal distance, or as a pre-focusing device used in conjunction with a downstream micro- or nano-focusing element, leading to enormous flux gains. Finally, without any other optics, the transfocator acts as a longitudinally dispersive monochromator, producing beams with about 1% band pass and several micrometres vertical size.

Figure 3. Theoretical (lines) and measured number (points) of Be lenses needed to optimise focus at 94 m. From left to right, the different curves correspond to coupling the Be lenses with 0, 32, 64 or 96 Al lenses.

For micro- and nano-focusing applications at long (ca. 100 m) focal length, the small beam size from the transfocator corresponds well with the acceptance of downstream focusing optics. The pre-focused beam from the transfocator has been used with subsequent focusing from either other refractive lenses, a silicon nano-lens chip [6,7], or a multilayer system. The flux gain is substantial, with only slight degradation in the spot size. For example, focusing the transfocator behind the focal point of a series of short focal length lenses produces a factor of 30 gain in the focal spot, with only a 75% increase in spot size. An even larger flux gain of 40 is found when using the nanofocusing chip, with the spot being broadened from about 200 to 450 nm vertically at 50 keV.

As a single optical device in the white beam, the transfocator can act as a fundamentally new kind of monochromator, delivering impressive flux in a ~1% band pass beam (a factor of 75 with respect to the ~10-3 band pass beam when used in conjunction with the double Laue monochromator). The energy selectivity of the transfocator comes from the chromatic nature of the focusing; with a relatively short focal length (10 m in this case), the depth of focus is sufficiently short that only a narrow band pass is focused at a given longitudinal point. At the focal point, this leads to a 2 order of magnitude increase in the flux at this selected energy with respect to the rest of the spectrum. By further tuning the undulator to a peak at the same energy the effect can be enhanced.

This concept has been demonstrated by translating a pinhole perpendicular to the beam at the focal point for 35.5 keV, and scanning an analyser crystal in the transmitted beam (Figure 4). The band pass (FWHM) at focus was found to be about 337 eV, and the vertical FWHM at the central energy about 10 μm (broadened by the pinhole in this measurement). When used with monochromatic beams, this chromatic nature of the focusing means that it suppresses harmonics at the focal point.

Figure 4. a) Measured spectrum through a 10 ×10 µm2 pinhole of the beam optimized for 35.5 keV. b) spatial distribution of the flux at 35.500 ± 0.005 keV perpendicular to the beam at the focus.

By varying the number of lenses, the transfocator can also be used to collimate the beam for high resolution experiments such as peak shape analysis, SAXS, or spectroscopy switching between phase and absorption contrast modes.

Finally, refractive optics are insensitive to vibrations which cause angular changes, the predominant vibrational modes of the mounting. This leads to a very high stability. In particular, using the transfocator and a double Laue monochromator, we were able to image a vertical source of 21 μm, close to the true source size and among the smallest observed in a monochromatic beam. This high stability also means that the transfocator can be used to characterise the source. At ID11’s endstation 94 m from the source the focusing ratio is about 1:2, which means the source is focused and can be measured with a normal high resolution detector, allowing rapid source characterisation (up to 1 kHz with a fast camera) and the measurement of sub-micro-radian angular source size.

References
[1] A. Snigirev, V. Kohn, I. Snigireva, B. Lengeler, Nature 384, 49 (1996).
[2] B. Lengeler, C. Schroer, J. Tummler, B. Benner, M. Richwin, A. Snigirev, I. Snigireva, Drakopoulos M, J. Synchrotron Rad. 6, 1153 (1999).
[3] B. Lengeler, C. Schroer, M. Richwin, J. Tummler, M. Drakopoulos, A. Snigirev, I. Snigireva, Appl. Phys. Lett. 74, 3924 (1999).
[4] I. Snigireva, A. Snigirev, V. Yunkin, M. Drakopoulos, M. Grigoriev, S. Kuznetsov, M. Chukalina, M. Hoffmann, D. Nuesse, E. Voges, AIP conference proceedings, 705, 708 (2004).
[5] A. Snigirev, I. Snigireva, G.B.M. Vaughan, J. Wright, M. Rossat, A. Bytchkov, C. Curfs, Journal of Physics, Conference Series, in press.
[6] A. Snigirev, I. Snigireva, M. Grigoriev, V. Yunkin, M. Di Michiel, S. Kuznetsov, G.B.M. Vaughan, Proc. of SPIE 6705, 670506-1 (2007).
[7] P. Van Vaerenbergh, A. Snigirev, M.A. Nicola, I. Snigireva, M. Grigoriev, V. Yunkin, G.B.M. Vaughan, L. Claustre, H-P. Van Der Kleij, and J-Y. Massonnat, Proc. of SPIE 6705, 670508 (2007).

Principal publication and authors
G.B.M. Vaughan, A. Snigirev, M. Rossat, J.P. Wright, A. Bytchkov, H. Gleyzolle, in preparation
ESRF