Ptychographic Characterization of the Wavefield in the Focus of Reflective

ARTICLE IN PRESS

Ultramicroscopy 110 (2010) 325–329

Contents lists available at ScienceDirect

Ultramicroscopy

journal homepage: www.elsevier.com/locate/ultramic

Ptychographic characterization of the waveﬁeld in the focus of reﬂective hard X-ray optics

Cameron M. Kewish a,*, Pierre Thibault a, Martin Dierolf b,1, Oliver Bunk a, Andreas Menzel a, Joan Vila-Comamala a, Konstantins Jeﬁmovs c, Franz Pfeiffer b,1 a Paul Scherrer Institut, CH-5232 Villigen PSI, Switzerland b Technische Universitat¨ Munchen,¨ D-85748 Garching, Germany c EMPA, CH-8600 Dubendorf,¨ Switzerland article info abstract

Article history: A technique for quantitatively characterizing the complex-valued focal waveﬁeld of arbitrary optics is Received 1 September 2009 described and applied to reconstructing the coherent focused beam produced by a reﬂective/diffractive Received in revised form hard X-ray mirror. This phase-retrieval method, based on ptychography, represents an important 23 December 2009 advance in X-ray optics characterization because the information obtained and potential resolution far Accepted 12 January 2010 exceeds that accessible to methods of directly probing the focus. Ptychography will therefore be well- suited for characterizing and aligning future nanofocusing X-ray optics. Keywords: & 2010 Elsevier B.V. All rights reserved. X-ray optics Wavefront characterization Ptychography Phase retrieval Diffractive imaging

1. Introduction been employed to measure probe intensity profiles of the order of 25–100 nm-wide, e.g., using fluorescence from nanofabricated Nanofocused X-ray beams promise to provide high spatial structures [4–6], or modulation transfer function measurements resolution for X-ray spectroscopy, scanning microscopy and from periodic samples [7]. However, these methods are limited to diffraction. X-ray optics, such as zone plates [1], refractive lenses providing linear profiles of the beam intensity, and are therefore [2], total reflection mirrors [3], and multilayer optics [4,5], have insensitive to wavefront aberrations and are restricted to working been used to focus hard X-ray beams to o100 nm, and in some directly in the focal plane. A procedure similar to autocollimation cases below 40 nm. These focal spot sizes represent significant was used to infer wavefront aberrations introduced by a hard achievements in the spatial resolution of hard X-ray microscopy, X-ray mirror in one-dimension (1-D), measuring beam displace- but the technological developments to achieve the ultimate ments during slit-scanning [4]. This can be extended to two wavelength-limited resolution with X-rays are still far out of dimensions (2-D) by scanning a pinhole rather than a slit, but the reach. resolution is limited by the pinhole size and diffraction between When trying to increase the resolution it is crucial to the aperture and the optics. characterize the aberrations in the focus, to give insight into the Phase retrieval algorithms are often used to reconstruct relationship between fabrication tolerances and optical perfor- complex-valued wavefields from far-field coherent diffraction mance. Precise measurement of the intensity profile in the patterns. So-called ‘lensless’ imaging replaces the X-ray image- vicinity of the focus, or ‘probe’, becomes more difficult as the forming system with computations and produces images with beam size decreases. Conventional ‘knife-edge’ scans are expected resolution dictated by the range of diffraction angles measured. to fail at higher photon energies and numerical apertures, because These techniques are usually based on iterative algorithms [8,9], the depth of focus can be shorter than the X-ray attenuation although analytical methods are being developed (see e.g., [10]). length in the knife-edge material. Alternative procedures have Once the phase problem is solved, the far field exit wave can be Fourier transformed to yield the projection of the complex refractive index distribution of a sample. Phase retrieval calcula- tions incorporating optical metrology data were used to refine the E-mail address: [email protected] (C.M. Kewish). phases of focal plane intensity measurements at high resolution in 1 Previous address: Swiss Light Source, Paul Scherrer Institut, CH-5232 Villigen PSI, Switzerland and E´ cole Polytechnique Fe´de´rale de Lausanne, CH- 1-D [11,12], but the nanostructure used to probe the focus 1015 Lausanne, Switzerland. restricted the technique to obtaining linear wavefront profiles.

326 C.M. Kewish et al. / Ultramicroscopy 110 (2010) 325–329

An iterative method based on curved wavefronts was proposed the chosen wavelength of 1.54 A˚ provides coherent radiation over for reconstructing the 2-D probe wavefield from the far-field an area of 250ðvÞ25ðhÞ mm2 transverse to the beam in the intensity distribution [13,14]. It relies on an accurate knowledge experiment hutch 34.5 m downstream [20]. of the wavefront curvature and the dimensions of the focusing The X-ray focusing optic used for this experiment is a optic, or the pupil function, to provide the real-space support commercial 2-D focusing device [21], whose paraboloidal reflec- constraint for the reconstruction algorithm. However, for reflective surface is coated with a laterally graded multilayer optimized tive optics, which necessarily operate at grazing incidence angles, for X-rays of this wavelength. This mirror was designed to or volume diffractive optics, the exit pupil plane is not as easily collimate the divergent beam from a laboratory X-ray source into defined. a parallel beam. Instead, the mirror was used in a reversed Recently, an iterative phase-retrieval method related to geometry, to focus the undulator beam to a size of the order of a ptychography was reported, which makes use of multiple micron at a distance approximately 120 mm from the center of exposures to solve the phase problem [15,16]. Combining the mirror. redundant information from overlapping illuminations on an A coherent portion of the beam was selected by a 20 mm- extended sample additionally allows one to recover the 2-D probe diameter pinhole located approximately 12 mm upstream of the wavefield in an essentially arbitrary plane perpendicular to the entrance window of the mirror chamber. The apertured beam was focused beam between the optic and the detector [17–19]. incident upon the mirror at a grazing angle of y ¼ 1:33, By numerically propagating the reconstructed wavefield along corresponding to the Bragg angle of the multilayer coating at this the optical axis, it is possible to diagnose aberrations in the optical wavelength. In the focal plane, a nanofabricated sample was system resulting from astigmatism, or misalignment. The infor- scanned with a high-precision piezoelectric positioning stage mation obtained in this process can allow optimization of a (nominal 0.3 nm resolution with o2 nm repeatability). The beamline optics, alignment of a focusing system for microscopy or sample stage was mounted on an optical bench that was rotated scanning probe applications. Propagating the wavefield back to to an angle of 2y with respect to the incident beam. the focusing optic can further allow an assessment of the The radiation transmitted and diffracted by the sample fabrication errors that are present, by analysis of the wavefront traveled through a He-filled ‘flight tube’, to minimize absorption aberrations produced by the optics. We have applied ptychogra- and scattering along the 7.26 m path to a PILATUS 2M detector phy to assess the focusing performance of a curved multilayer [22]. This temperature-stabilized detector samples the far-field X-ray focusing optic, as a demonstration of principle, to assess intensity with 1475 1679 pixels of size 172 172 mm2, having whether this method will be useful to characterize future X-ray 20-bit dynamic range, no readout noise or dark current, and a optics providing nanometer-sized hard X-ray probes. point-spread function of one pixel [23]. Despite the large dynamic range of the detector, and in order to avoid the use of a beam-stop, we had to attenuate the incident beam with a 200 mm- thick Si 2. Experimental procedure single-crystal such that no pixel was subjected to more than a few million counts per second per pixel, corresponding to the dead- The experiments were carried out at the cSAXS beamline at the time of the detector [24]. Swiss Light Source, an undulator beamline that is optimized for For hard X-rays traversing thin samples, the exit wave is to a coherent X-ray scattering techniques and small angle-scattering good approximation equal to the product of the probe wavefield, ˚ in the range of wavelengths between 0.65 and 3.0 A. A schematic P(r) and the complex transmission function of the sample, O(r), of the experiment geometry is shown in Fig. 1. e.g. A double-crystal Si(1 1 1) monochromator selected 1.54 A˚ c ðrÞ¼Pðr-r ÞOðrÞ; ð1Þ wavelength radiation from the undulator spectrum, which was j j incident at a grazing angle of 0:233 on a Rh-coated mirror to reject where cjðrÞ represents the exit wave resulting from the j th 497:5% of the higher-harmonics, prior to the beam entering the sample position rj. The intensity in the far-field, Ij, can be written experiment hutch. The radiation source is approximately 20 mm as (vertical) by 200 mm (horizontal) in size at the undulator, which at ~ 2 IjðqÞ¼jcj ðqÞj ; ð2Þ where the tilde denotes the two-dimensional Fourier transform of the exit wave, and the diffraction vector q is the reciprocal-space coordinate dual to the sample plane coordinate r. Eqs. (1) and (2) comprise the constraints used to guide the iterative search to recover the phases that are lost when measuring the far-field diffraction pattern. When multiple intensity measurements from partially overlapping probe positions are given as input to the ptychography algorithm, a unique solution can be obtained. Such a dataset also contains sufficient information to refine the probe function from an initial assumption. This procedure, described in detail by Thibault et al. [18], consists of numerically solving the simultaneous equations X ^ P ðr-rjÞc ðrÞ O^ ðrÞ¼ P j ð3Þ j ^ 2 jjPðr-rjÞj ;

X ^ O ðrþrjÞcjðrþrjÞ Fig. 1. Schematic diagram of the experiment. A pinhole (P) selects a coherent P^ ðrÞ¼ P ð4Þ j ^ 2 portion of the synchrotron beam (X), to be focused by a multilayer mirror (M) onto jjOðrþrjÞj ; the sample (S). The sample can be scanned in the XY plane, perpendicular to the optical axis Z. The radiation transmitted and diffracted by the sample is incident where the circumﬂex represents the discrete, reconstructed on a detector (D) located in the far-ﬁeld downstream. (Figure not to scale.) images of the exit wave and the probe, respectively, and the ARTICLE IN PRESS

C.M. Kewish et al. / Ultramicroscopy 110 (2010) 325–329 327 superscript asterisk denotes complex conjugation. The ptycho- in Fig. 2(b), where the light colored structures are Au and the graphy algorithm used here assumes that the illumination darker background is the polyimide layer on the membrane. function is coherent, without allowing for the possibility of The central 512 512 pixels were extracted from the mea- partial coherence resulting from the finite source size. sured full-frame diffraction patterns for reconstruction. For the A ptychography scan consisted of collecting a series of experiment geometry outlined above, this results in reconstructed diffraction patterns from a sample illuminated at various over- pixels of side-length 12.72 nm. The complex probe wavefield is lapping positions. The locations of the scan points were chosen to reconstructed with pixels of the same size, shown in Fig. 2(c), avoid the ‘raster grid pathology’ which can arise when ptycho- where the image width corresponds to 6:51 mm. This resolution graphy data are collected from sample positions which are allows us to see the detailed structure in the beam, resulting periodic, as on a rectangular grid [18]. Specifically, the scan primarily from aberrations and roughness present in the surface points were located on concentric circles, whose radii varied from of the focusing mirror. 0.5 to 5 mm in steps of 0:5 mm. The scan started in the center, and Intensity profiles were calculated by integrating all pixels in each circle contained five positions more than the previous, the detector at each scan point, as a thick (opaque) knife-edge was requiring a total of 330 positions to scan a 10 mm- diameter translated by 72:5 mm in 20 nm increments across the beam, in area and ensure at least 50% overlap of neighboring points. Two both the X and Y directions. Taking the derivative of these profiles, diffraction patterns were collected at each of the scan positions, we find the full-width at half-maximum (FWHM) to be integrating for 0.1 and 30 s, respectively. The values in the center wx ¼ 1:32 mm and wy ¼ 1:26 mm in the X and Y directions, of the shorter exposure were then used to correct overexposed respectively. To compare the reconstructed probe with the pixels in the center of the longer exposure. knife-edge scans of the focus, we find the FWHM of the probe intensity integrated in the perpendicular direction, i.e. Z ^ ^ 2 3. Results IðxÞ¼ jPðx; yÞj dy; ð5Þ

For probe retrieval it is important to employ a strongly which reveals that the reconstructed beam is slightly smaller than scattering sample, which includes fine details that will diffract the directly measured beam, as shown in Fig. 2(d). The size of the to sufficiently high angles to achieve the desired resolution. As a reconstructed focus in the X and Y directions is w^ x ¼ 0:77 mm and sample we used a nanofabricated Snellen chart, consisting of Au w^ y ¼ 0:78 mm, respectively. The difference is due to the convolu- structures produced by electron beam lithography and electro- tion of the beam size with the knife-edge shape function. plating [1] into a polyimide mold resist layer on a silicon nitride Measurements were also taken with the sample out of focus, to supporting membrane. The sample plane was positioned near the demonstrate that this beam characterization method is robust focal plane of the mirror, determined by scanning the opaque with respect to positioning of the sample and different wavefront edge of the sample-holder through the beam. Fig. 2(a) shows the curvatures. Since we reconstruct a complex-valued wavefield, we reconstructed complex exit wave in the sample plane. The phase can numerically propagate it along the optical axis. Fig. 3(a) difference of approximately p=3 rad introduced by the Au shows a sagittal view of the probe wavefield shown in Fig. 2(c), structures compared to the surrounding polyimide layer where the horizontal axis is the direction of propagation. indicates a Au thickness of around 0:4 mm725%, where larger Comparing this directly with a coronal view of the wavefield structures toward the top of the chart are on the upper end of this (not shown) reveals that there is no astigmatism in the focus, range. Although the Au thickness could not be directly measured, as the horizontal and vertical foci coincide. The beam waist this range of thicknesses is reasonable because the polyimide can clearly be seen and the location of the original scanning layer was 1 mm thick, and the Au coating process was halted well plane, shown with a dashed line in the center, indicates that below this thickness to avoid overplating. As a result, some of the the sample was well inside the focus. A different sample, a structures were unfilled or partially-filled, and the average 175 mm- thick silicon chip with integrated circuits, was scanned at thickness was expected to be well below a micron. This is also distances of Dz ¼þ8:64 mm (positive indicating downstream) evidenced by a scanning electron microscope (SEM) image shown and 16.28 mm from the previous scanning plane.

Fig. 2. (a) Reconstructed complex-valued image of a nanofabricated test object, showing Au structures deposited on a silicon nitride membrane. (b) Scanning electron micrograph of the sample. (c) Reconstructed illumination function obtained simultaneously to the exit wave. (d) Comparison of line-proﬁles (solid lines: X-axis shown in blue, Y in red) through the normalized intensity of the reconstructed probe, and the intensity calculated from knife-edge scans (circles). The color wheel in (c) shows the mapping for complex values in (a) and (c) displayed with phase mapped to hue and amplitude to value. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.) ARTICLE IN PRESS

328 C.M. Kewish et al. / Ultramicroscopy 110 (2010) 325–329

Fig. 3. (a) Numerically propagating the reconstructed probe from Fig. 2(c) along the optical axis reveals the beam waist. (b) Complex exit wave reconstructed for a silicon chip sample, scanned 8.62 mm downstream of the original ‘in-focus’ position (indicated in (a) by dashed lines). Comparison of the propagated probe (c) and the ‘out-of- focus’ reconstructions reveals good agreement. The intensity proﬁles along the X-axis are compared in (e), where the red points are from the propagated probe, and the blue from the out-of-focus reconstruction. The scale bars in images (b–d) represent 2 mm, and the color wheel in (b) shows the mapping of the phase and amplitude of the complex images, to hue and value, respectively. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

Fig. 3(b) shows the image of the complex exit wave ptychography can be seen clearly in the details of the recon- reconstructed for the chip sample in the downstream position, structed wavefields. The fine details in the reconstructed sample indicated by the dashed line in Fig. 3(a). This reconstruction used images agree well with SEM images: despite the focal spot size of 128 128 pixels from each of the measured diffraction patterns, the mirror being rather large, we can resolve the ‘holes’ in the resulting in a pixel size of 50.86 nm. The top interconnect layer numerals 6 and 8 that can be seen in Fig. 2(b) are o70 nm in and via structures, of 0:25 mm dimensions, are the dominant diameter. The spatial resolution is limited in principle only by the features in the sample reconstruction. The reconstructed probe highest detection angle which receives statistically significant compares qualitatively very well with the probe numerically coherent diffraction intensity. In practise, however, the resolution propagated by Dz from the focal plane [Figs. 3(c) and (d)], and a of the probe (and object) reconstruction is ultimately limited by line profile through the two wavefield intensities shows the the accuracy to which one knows the sample positions, or can quantitative agreement wherein the RMS difference is of the order refine them from the data. of 2% [Fig. 3(e)]. The RMS difference between the phases of these Comparing the intensity of the reconstructed probe wavefield wavefields, in regions where there is non-negligible amplitude, is with the knife-edge scans shows that the directly measured approximately 0.5 rad corresponding to a wavefront reconstruc- values overestimate the beam size. This results primarily tion accuracy of about l=12:7. Similar agreement was obtained for from the convolution of the knife-edge shape with the beam the upstream position. profile. In the present case, the effect is quite noticeable, because the knife-edge intersecting the beam was not optimized for the measurement. In fact, an edge of the 3 mm-thick sample holder 4. Discussion was used, and therefore the measurement represents a beam size measurement at up to 3 mm downstream from the sample The potential of ptychography for beam characterization in position. Simply propagating the focused probe by this distance X-ray optics is demonstrated by the agreement between the increases its FWHM by approximately 40%, which improves the wavefields reconstructed in different planes, and for different agreement between the two measurements, but indicates that samples. The full complex wavefields before and after the sample there is a further contribution, e.g., from the roughness of the can be recovered at high resolution without a priori knowledge of edge. This highlights a difficulty of physically probing a focused the focal length, or the sample position relative to the focus along beam, even using specialized knife-edges and scanning techni- the optical axis. Only the wavelength, sample-to-detector dis- ques, since the beam size and depth-of-focus will decrease tance and the sample positions perpendicular to the optical axis appreciably with diffraction-limited hard X-ray nanofocusing need to be known accurately. The resolution enhancement of optics. ARTICLE IN PRESS

C.M. Kewish et al. / Ultramicroscopy 110 (2010) 325–329 329

5. Summary and outlook nanoprobe based on refractive X-ray lenses, Appl. Phys. Lett. 87 (12) (2005) 124103. [3] H. Mimura, H. Yumoto, S. Matsuyama, Y. Sano, K. Yamamura, Y. Mori, M. The objective of this experiment was to characterize the focus Yabashi, Y. Nishino, K. Tamasaku, T. Ishikawa, K. Yamauchi, Efficient focusing of an X-ray mirror using ptychography. The reconstruction of hard X rays to 25 nm by a total reflection mirror, Appl. Phys. Lett. 90 (5) algorithm allows the unique recovery of the complex wavefield (2007) 051903. [4] O. Hignette, P. Cloetens, G. Rostaing, P. Bernard, C. Morawe, Efficient sub in the plane of a specimen which is scanned through a coherent 100 nm focusing of hard X rays, Rev. Sci. Instrum. 76 (6) (2005) 063709. beam. The reconstructed wavefield can then be numerically [5] H.-C. Kang, J. Maser, G.B. Stephenson, C. Liu, R. Conley, A.T. Macrander, S. Vogt, propagated along the focused beam path, making available Nanometer linear focusing of hard X rays by a multilayer Laue lens, Phys. Rev. Lett. 96 (12) (2006) 127401. detailed information about the optical system, including astig- [6] W. Liu, G.E. Ice, J.Z. Tischler, A. Khounsary, C. Liu, L. Assoufid, A.T. Macrander, matism and other aberrations which are not obtainable from Short focal length Kirkpatrick–Baez mirrors for a hard X-ray nanoprobe, Rev. knife-edge scans. We reconstructed exit waves from samples Sci. Instrum. 76 (11) (2005) 113701. [7] W. Chao, B.D. Harteneck, J.A. Liddle, E.H. Anderson, D.T. Attwood, Soft X-ray located at several positions relative to the focus of an X-ray microscopy at a spatial resolution better than 15 nm, Nature 435 (2005) mirror, along with a self-consistent set of illumination functions. 1210–1213. The beam size obtained is smaller than that measured with knife- [8] R.W. Gerchberg, W.O. Saxton, A practical algorithm for the determination of edge scans of the intensity profile, a result which can be phase from image and diffraction plane pictures, Optik 35 (1972) 237–246. [9] J.R. Fienup, Phase retrieval algorithms: a comparison, Appl. Opt. 21 (15) attributed to instrumental broadening. Ptychography does not (1982) 2758–2769. require accurate knowledge of the dimensions of the focusing [10] S.G. Podorov, K.M. Pavlov, D.M. Paganin, A non-iterative reconstruction mirror or the form of the pupil function, instead relying on very method for direct and unambiguous coherent diffractive imaging, Opt. Express 15 (16) (2007) 9954–9962. accurate sample positioning and far-field intensity measurements [11] H. Yumoto, H. Mimura, S. Matsuyama, S. Handa, Y. Sano, M. Yabashi, Y. collected with a high dynamic range, low noise detector. We Nishino, K. Tamasaku, T. Ishikawa, K. Yamauchi, At-wavelength figure conclude that this technique represents an important advance in metrology of hard X-ray focusing mirrors, Rev. Sci. Instrum. 77 (2006) 063712. [12] H. Mimura, H. Yumoto, S. Matsuyama, S. Handa, T. Kimura, Y. Sano, M. so-called ‘at-wavelength’ X-ray optics characterization which Yabashi, Y. Nishino, K. Tamasaku, T. Ishikawa, K. Yamauchi, Direct does away with the need to directly probe the focus. Ptycho- determination of the wave field of an X-ray nanobeam, Phys. Rev. A 77 (1) graphy will therefore be useful in the testing and alignment of (2008) 15812. [13] G.J. Williams, H.M. Quiney, B.B. Dhal, C.Q. Tran, K.A. Nugent, A.G. Peele, D. beamlines employing nanofocusing optics, and assessing the Paterson, M.D. de Jonge, Fresnel coherent diffractive imaging, Phys. Rev. Lett. effect of optics on the X-ray wavefronts by propagating the 97 (2) (2006) 025506. reconstructed probe wavefield back to the focusing optic is [14] H.M. Quiney, A.G. Peele, Z. Cai, D. Paterson, K.A. Nugent, Diffractive imaging of highly focused X-ray fields, Nat. Phys. 2 (2006) 101–104. expected to provide useful information for both metrology and [15] H.L. Faulkner, J. Rodenburg, Movable aperture lensless transmission micro- fabrication. scopy: a novel phase retrieval algorithm, Phys. Rev. Lett. 93 (2) (2004) 023903. [16] J.M. Rodenburg, A.C. Hurst, A.G. Cullis, B.R. Dobson, F. Pfeiffer, O. Bunk, C. David, K. Jefimovs, I. Johnson, Hard-X-ray lensless imaging of extended Acknowledgments objects, Phys. Rev. Lett. 98 (3) (2007) 034801. [17] P. Thibault, M. Dierolf, A. Menzel, O. Bunk, C. David, F. Pfeiffer, This work was carried out at the cSAXS beamline at the High-resolution scanning X-ray diffraction microscopy, Science 321 (2008) 379–383. Swiss Light Source, Paul Scherrer Institut, CH-5232 Villigen PSI, [18] P. Thibault, M. Dierolf, O. Bunk, A. Menzel, F. Pfeiffer, Probe retrieval in Switzerland. We are grateful for the continuous technical support ptychographic coherent diffractive imaging, Ultramicroscopy 109 (2009) from Xavier Donath at cSAXS, and we thank Peter Høghøj at 338–343. [19] A.M. Maiden, J.M. Rodenburg, An improved ptychographical phase retrieval Xenocs for the loan of an earlier 2-D focusing mirror during algorithm for diffractive imaging, Ultramicroscopy 109 (10) (2009) preparation for this experiment. P.T. acknowledges financial 1256–1262. support from the Fonds Que´bećois de la recherche sur la nature [20] F. Pfeiffer, O. Bunk, C. Schulze-Briese, A. Diaz, T. Weitkamp, C. David, J.F. van der Veen, I. Vartanyants, I.K. Robinson, Shearing interferometer for quantify- et les technologies (FQRNT). M.D. and F.P. acknowledge support ing the coherence of hard X-ray beams, Phys. Rev. Lett. 94 (16) (2005) 164801. through the DFG Cluster of Excellence Munich—Center for [21] Manufactured by Xenocs: 19, rue Franc-ois Blumet, F-38360 Sassenage, Advanced Photonics. France, Model: FOX-2D-CU-12_INF. [22] Manufactured by Dectris Ltd.: Neuenhoferstrasse 107, CH-5400 Baden, Switzerland, Model: PILATUS 2M. References [23] C. Broennimann, E.F. Eikenberry, B. Henrich, R. Horisberger, G. Huelsen, E. Pohl, B. Schmitt, C. Schulze-Briese, M. Suzuki, T. Tomizaki, H. Toyokawa, A. Wagner, The PILATUS 1M detector, J. Synchrotron Radiat. 13 (2) (2006) [1] K. Jefimovs, O. Bunk, F. Pfeiffer, D. Grolimund, J.F. van der Veen, C. David, 120–130. Fabrication of Fresnel zone plates for hard X-rays, Microelectron. Eng. [24] P. Kraft, A. Bergamaschi, C. Broennimann, R. Dinapoli, E.F. Eikenberry, B. 84 (5–8) (2007) 1467–1470. Henrich, I. Johnson, A. Mozzanica, C.M. Schleputz,¨ P.R. Willmott, B. Schmitt, [2] C.G. Schroer, O. Kurapova, J. Patommel, P. Boye, J. Feldkamp, B. Lengeler, M. Performance of single-photon-counting PILATUS detector modules, J. Syn- Burghammer, C. Riekel, L. Vincze, A. van der Hart, M. Kuchler,¨ Hard X-ray chrotron Radiat. 16 (2009) 368–375.