short communications
Twin helical undulator beamline for soft
at SPring-8aJAERI-RIKEN SPring-8 Project Team, Kamigori, Ako-gun, Hyogo 678-12, Japan, bJapan Synchrotron Radiation Reseach Institute, Kamigori, Ako-gun, Hyogo 678-12, Japan, cDepartment of Physics, Tokyo Metropolitan University, Tokyo 192-03, Japan, dSRL, ISSP, The University of Tokyo, Tanashi, Tokyo 188, Japan, eSchool of Engineering, Nagoya University, Nagoya 464-01, Japan, fDepartment of Material Physics, Osaka University, Osaka 560, Japan, and gGraduate School of Material Science, Nara Institute of Science and Technology, Nara 630-01, Japan
*Correspondence e-mail: ysaitoh@spring8.or.jp
A very high resolution soft X-ray beamline, BL25SU, has been designed and is under construction at SPring-8. Completely right or left circularly polarized light is supplied on a common axis of a newly designed twin helical undulator. A helicity modulation up to 10 Hz can be performed using five kicker magnets. The fundamental radiation covers the region 0.5–3 keV. Higher-order radiation is rather weak on the axis. A monochromator with varied-line-spacing plane gratings is installed to cover the region below 1.5 keV. A very high resolution beyond 104 is expected for the whole energy region.
1. Introduction
Integrated studies of electronic, magnetic and geometrical structures are now becoming more and more necessary. Soft
such as photoemission, photoelectron diffraction and using circularly polarized undulator radiation is the means to satisfy such requirements. Higher-order light rejection, helicity modulation, as well as high-resolution measurements, will open up wider applications.2. Light source and monochromator
Our light source is the `twin helical undulator' consisting of two left- and right-handed helical undulators installed in tandem in a 4.5 m straight section. This device was designed by Kitamura's group (Hara et al., 1998) and covers the energy range 0.5–3 keV with fundamental radiation. Previously developed helical undulators which produce both helicities emit light on slightly different optical axes (Goulon et al., 1996). In our undulator, each helicity light comes out on the same optical axis (0 direction) and the helicity can be alternatively reversed by the five kicker magnets as shown in Fig. 1. The circularly polarized light with unwanted helicity comes out in the off-axis direction (1 direction) and can be absorbed by the water-cooled components of a mask and an xy slit set downstream. Quick helicity modulation up to 10 Hz or more with a switching time within 50 ms is the target.
The exclusive use of the circularly polarized light from the helical undulator of a high-energy storage ring has some characteristic advantages. First, the on-axis power density from the helical undulator is very low, especially for high K values, compared with that of the planar undulator, because most of the higher harmonics radiation is emitted off the axis of the helical undulator. Fig. 2 shows results calculated from the storage ring parameters for the power density and density (dashed and solid curves, respectively; the horizontal arrows indicate the corresponding ordinates). The dashed curves on the left-hand side of Fig. 2 have two maxima at θx = ±θm and a minimum at θx = 0 corresponding to the upper geometry of Fig. 1 (Kitamura, 1996). The two different dashed curves shifted by 300 µrad correspond to the radiation from the RCP (right circularly polarized) and LCP (left circularly polarized) undulators in the upper configuration of Fig. 1. For example, the on-axis power density of each undulator at K = 3.0 (corresponding to the fundamental energy ∊1 = 500 eV) for the stored current of 100 mA is about 0.2 kW mrad−2 for the total radiation power of 0.8 kW and θm is about 190 µrad. The density for a 0.1% bandwidth at ∊1 = 500 eV is still as high as 2.6 × 1016 photons s−1 mrad−2 on the axis, as shown by the solid curve. Under this condition, the opposite helicity radiation along the 1 direction has the distribution shown on the right-hand side of Fig. 2.
When the helicity is changed from the upper geometry to the lower geometry in Fig. 1, undesirable radiation passes through the optical axis (0 direction) depending on the angle, θk, between the radiation directions of 0 and 1. θm decreases with decreasing K value as approximated by K/γ ≃ 64K mrad.
The value of θk between the two directions (0 and 1) is so determined that the power from the unused undulator is sufficiently suppressed in using a particular helicity. Restriction of the power supply for the kickers is also taken into account. As a result, θk is set to 300 µrad, where the power density in the 0 direction is at most 1.5 kW mrad−2 in the ∊1 range 0.5–1.5 keV covered by the monochromator described below.
In the initial stage, the energy range 0.5–1.5 keV is covered by a grating monochromator. The L and M edges of most transition-metal and rare-earth magnetic materials are covered by this energy range.
Fig. 3 schematically shows the major optical elements of BL25SU. Detailed parameters of the components are listed in Table 1. In order to preserve the source circular polarization, a grazing-incidence monochromator is used. The beamline consists of three main parts: (i) the pre-focusing mirrors, Mh and Mv, in a hutch, (ii) the monochromator, from the entrance slit S1 to the exit slit S2, and (iii) the post-focusing mirrors, M3 and M4. A variably bendable cylindrical mirror, Mh, horizontally deflects the undulator radiation by 3° and horizontally converges the light onto the sample positions (ST1–ST3). The spherical mirror, Mv, vertically focuses the beam onto the entrance slit. Water-cooled Si substrates are used for these optical components to minimize possible effects of thermal instability. The water-cooled entrance slit is continuously variable from 2 µm to 1 mm.
|
A varied-line-spacing plane grating is mounted in this monochromator. A constant-deviation type is employed in contrast to the monochromators designed by Itou et al. (1989) or by Fujisawa et al. (1996). The type of monochromator employed here was first designed by Hettric (1988) using a spherical mirror and a varied-line-spacing plane grating. The present design has two angles of deviation, 176 and 174°, to cover the wide photon energy range from 200 to 1500 eV, by in situ selecting of one of the spherical mirrors, M1 and M2, and one of the three gratings, G1, G2 and G3. For the 174° deviation angle, mirror M1 is retracted from the beam path. One of the gratings is selected by a linear translation mechanism. The gratings are mechanically ruled in the Au thin film deposited on an SiC substrate to avoid the heat load problem. Such mechanically ruled varied-line-spacing plane gratings are used in the BL-19B planar undulator beamline at the Photon Factory and have shown high-resolution performance. However, simultaneous rotation of a large plane mirror and the grating to scan the wavelength induced some problems with the wavelength reproducibility and mechanical stability in the design by Fujisawa et al. (1996). In the present monochromator, the wavelength scanning only requires a rotation of the grating around an axis parallel to the grooves. Consequently, high precision of the wavelength scanning is expected. The spherical mirror, M1 or M2, converges the vertically divergent beam from S1 to a virtual source at a distance 10 m from the grating centre. This type of monochromator has the excellent feature that the focal point of the grating is almost unchanged near the exit slit throughout the scanning range, when the space-variation parameters of the grating are properly determined.
The groove-spacing profile is determined from a light-path function theory. In the case of a mechanically ruled plane grating, the groove spacing, σ, is given by
where σ0 is the central groove spacing and w is the distance along the grating length. The space-variation parameters b2, b3 and b4 can be determined by cancelling the aberration terms of defocus (F20), coma (F30) and spherical aberrations (F40), respectively, at a certain photon energy. The parameters are optimized for hν0 in Table 1, so that high energy resolution is realized in the scanning range. Other designed parameters of the gratings G1 and G2 are also summarized in Table 1. The blaze angles are determined to give the highest output considering the diffraction efficiency and the reflectivity.
The designed performance of the monochromator is evaluated by ray tracing. The hν/Δhν, of the monochromator estimated by this ray-tracing method is shown in Fig. 4. The beam characteristics were calculated from storage ring parameters (Kitamura, 1996). Δhν was given by FWHM × (reciprocal linear dispersion). The width of the entrance slit is assumed to be 10 µm. The full circles are the results obtained by neglecting the slope error on the surface of the optical components and the triangles represent the results with possible slope errors (1 µrad for M1 and 0.5 µrad for G1). The inset of Fig. 4 shows an example of the ray tracing on the exit plane for a photon energy of 1.5 keV for the combination of M1 and G1, where no slope errors are assumed. Comparison of the image size with the value evaluated from magnification has shown that the contribution of aberrations is 1 µm or so. That is, it provides an almost source-size-limited resolution. The entrance slit-width-limited resolution curves are also shown for comparison by the solid curves in Fig. 4. Although the resolution strongly depends on the slope error of the optical components, we can conservatively expect a resolution of about 104 or more for the present system.
At ST3 of the third experimental end-station, we expect an intensity of the order of 1011 photons s−1 at a of 104 with a beam size of less than 1 mm2, considering the transmittance obtained by the ray tracing, reflectivity and grating efficiency.
3. End-stations
Three experimental stations are on this beamline. (i) High-resolution photoemission apparatus composed of an SES200 spherical mirror analyser combined with a closed-cycle liquid-He cryostat is located at ST1 in Fig. 3. (ii) Apparatus for spectroscopy of of core absorption, in which the sample temperature is also controlled by a liquid-He cryostat. At present, two permanent-magnet dipoles are used for changing the direction of the magnetization. In future, a superconducting magnet will be used in combination with the rapid helicity modulation of the undulator light. (iii) A two-dimensional display-type photoelectron analyser to be used for angle-resolved photoemission and photoelectron diffraction experiments is located at ST3 in Fig. 3. A two-dimensional photoelectron angular-distribution pattern at one particular is recorded simultaneously by a CCD camera. In order to realize a resolution of 1000 for the electron pass energy of the analyser, the radius of the outer hemisphere of this analyser is set to 300 mm.
The whole beamline and experimental apparatus will be commissioned in early 1998.
Acknowledgements
The authors would like to thank Professor H. Kitamura, Dr T. Hara and Dr T. Tanaka of SPring-8 for frequent and helpful discussions about the twin helical undulator. We also acknowledge many staff of the SPring-8 with whom we have discussed this beamline.
References
Fujisawa, M., Harasawa, A., Agui, A., Watanabe, M., Kakizaki, A., Shin, S., Ishii, T., Kita, T., Harada, T., Saitoh, Y. & Suga, S. (1996). Rev. Sci. Instrum. 67, 345–349. CrossRef CAS Web of Science
Goulon, J., Malgrange, C., Giles, C., Neumann, C., Rogalev, A., Moguiline, E., De Bergrevin, F. & Vettier, C. (1996). J. Synchrotron Rad. 3, 272–281. CrossRef CAS Web of Science IUCr Journals
Hara, T., Tanaka, T., Tanabe, T., Marechal, X. M., Kumagai, K. & Kitamura, H. (1998). J. Synchrotron Rad. 5, 426–427. Web of Science CrossRef CAS IUCr Journals
Hettric, M. C. (1988). Nucl. Instrum. Methods, A266, 404–413.
Itou, M., Harada, T. & Kita, T. (1989). Appl. Opt. 28, 146–153. CrossRef CAS PubMed
Kitamura, H. (1996). Insertion Device Handbook '96, pp. 83–124. Japan Synchrotron Radiation Reseach Institute, Japan. (In Japanese.)
© International Union of Crystallography. Prior permission is not required to reproduce short quotations, tables and figures from this article, provided the original authors and source are cited. For more information, click here.