Non-existence of positive glitches in spectra using the YB₆₆ double-crystal monochromator of BL15XU at SPring-8

YB₆₆ is suitable for dispersing synchrotron radiation in the 1–2 keV energy range with a 2d lattice spacing of 1.17 nm. When used with an undulator there are no positive glitches at 1385.6 and 1438 eV in spectra dispersed by a YB₆₆ 400 double-crystal monochromator as observed using bending-magnet or wiggler beamlines. The transmission function of a YB₆₆ double-crystal monochromator has been measured by means of a Si PIN photodetector, and X-ray absorption near-edge structure (XANES) of Mg, Al and Si were measured at high resolution. From all of these experiments it has been clarified that a YB₆₆ double-crystal monochromator is well suited for soft X-ray beamlines on third-generation light sources.

1. Introduction

We have adopted a YB₆₆ double-crystal monochromator (Fukushima et al., 1999) for use in the 1–2 keV energy range. YB₆₆ has a f.c.c. packing structure of B₁₅₆ units (Slack et al., 1971, 1977), which has a lattice plane (400) whose large d-lattice spacing of 0.586 nm is suitable for dispersing soft X-rays (Tanaka et al., 1985). Moreover, since YB₆₆ is more resistant than other crystals (beryl, quartz, InSb) to synchrotron radiation damage (Wong et al., 1982, 1999; Funabashi et al., 1989) and has a higher resolution than other crystals, it is best suited for dispersing synchrotron radiation at a third-generation light source in the 1–2 keV energy range. On the other hand, YB₆₆ has the disadvantages that its resolution is likely to be affected by the heat load because of its very low thermal conductivity (Slack et al., 1971) and it exhibits non-uniform crystalline quality (Rek et al., 1992). In addition, there are positive glitches at 1385.6 and 1438 eV in the spectrum dispersed by YB₆₆ when using a bending magnet (Tanaka et al., 1997; Wong et al., 1999) or a wiggler (Kinoshita et al., 1998).

The mechanism of positive-glitch generation was discussed by Tanaka et al. (1997), and it was confirmed that positive glitches are caused by sharp reflectivity increases associated with anomalous scattering for the 006 reflection at the Y L₃- and L₂-edges by means of reflectivity measurements and structure-factor calculations for the 006 reflections.

2. Experimental

YB₆₆ crystals were installed in the double-crystal monochromator on the helical-undulator beamline B15XU at SPring-8 (Kitamura et al., 2002, 2003). The experimental configuration is shown in Fig. 1. A helical undulator was adopted because the angular distribution of flux density is mainly distributed off the optical axis and power on the crystal is lower than that from a planar undulator (Hara et al., 2001). Although a helical undulator decreases the thermal load compared with a planar undulator, additional thermal-load reduction is necessary with YB₆₆ crystals (Avery, 1984; Bilderback, 1986 Bilderback et al., 1986). The thermal load from a helical undulator was reduced by adjusting the aperture size of the front-end slit and adding a Be filter. Prior to using the YB₆₆ crystal, the effects of power density and total power on the first YB₆₆ crystal were evaluated using SPECTRA (Tanaka & Kitamura, 2001). The transmission function of the YB₆₆ double-crystal monochromator was evaluated by measuring the photocurrent from a Si PIN photodetector at an experimental position while simultaneously scanning the monochromator and the helical undulator gap. XANES of Mg, Al and Si were also measured. The current before the sample was monitored with a Au mesh.

Figure 1 Outline of the experimental configuration to the first experimental hutch of beamline BL15XU at SPring-8. A tandem-type monochromator consists of two monochromators for which one can select an analyzer crystal from Si, YB66 and a multilayer. In the 1–2 keV energy range we adopted the YB66 crystal for dispersing synchrotron radiation.

3. Results and discussion

The transmission function of the YB₆₆ double-crystal monochromator is shown in Fig. 2. There are no positive glitches at 1385.6 and 1438 eV in the transmission function of the YB₆₆ double-crystal monochromator of our undulator beamline. It is considered that this result supports the explanation of the glitch-generation mechanism in the previous study; positive glitches are generated due to the 006 reflection with an energy 1.5 times higher than that of the YB₆₆ 400 reflection (Tanaka et al., 1997). The spectrum from a helical undulator in the case of the first harmonics at 1385 eV calculated using SPECTRA is shown in Fig. 3(a). In the spectrum, the second harmonics are 2760 eV away from the photon energy of the Y L₃- and L₂-edges; the flux at 2080 eV corresponding to the Y L₃-edge was estimated to be 1.5 × 10⁹ photons s⁻¹ (0.1% bandwidth)⁻¹ (100 mA)⁻¹ with a Be filter (thickness 25 µm) and the flux at 2156 eV corresponding to the Y L₂-edge was estimated to be 1.8 × 10⁹ photons s⁻¹ (0.1% bandwidth)⁻¹ (100 mA)⁻¹ with a Be filter (thickness 25 µm). These flux values are three orders of magnitude smaller than that of the first harmonics with a Be filter (thickness 25 µm) and not sufficient to generate positive glitches.

Figure 2 The transmission function of the YB66 double-crystal monochromator measured using the photocurrent of a Si PIN photodetector at an experimental position on the beamline. There are no positive glitches in the transmission function. Ring current = 100 mA, 25 µm Be filter, 0.04 mm × 0.6 mm front-end slit.

Figure 3 Spectrum of the first harmonics at (a) 1385 eV and (b) 1438 eV from a helical undulator of BL15XU calculated using SPECTRA. Ring current = 100 mA, 0.04 mm × 0.6 mm front-end slit. The flux is measured in units of photons s^−1 (0.1% bandwidth)^−1 (100 mA)^−1.

Similarly, the spectrum from a helical undulator in the case of the first harmonics at 1438 eV calculated using SPECTRA is shown in Fig. 3(b). The flux at 2080 eV corresponding to the Y L₃-edge was estimated to be 1.2 × 10⁹ photons s⁻¹ (0.1% bandwidth)⁻¹ (100 mA)⁻¹ with a Be filter (thickness 25 µm) and the flux at 2156 eV corresponding to the Y L₂-edge was estimated to be 1.3 × 10⁹ photons s⁻¹ (0.1% bandwidth)⁻¹ (100 mA)⁻¹ with a Be filter (thickness 25 µm). These flux values are estimated to be four orders of magnitude smaller than that of the first harmonics with a Be filter (thickness 25 µm) and not sufficient to generate positive glitches. In the experiment using the YB₆₆ crystals, a scan of the photon energy was performed by automatic tuning of a helical-undulator gap. Thus it is considered that the spectra of the first harmonics at 1385 and 1438 eV are the same as those estimated using SPECTRA.

From these discussions it is confirmed that positive glitches are not generated in the undulator beamline by automatic tuning of a helical-undulator gap. These results indicate that our undulator beamline effectively utilizes the YB₆₆ 400 monochromator without the cut-off mirror which minimizes the intensity of anomalous high harmonics at glitches compared with using bending-magnet or wiggler beamlines.

To confirm that there were no positive glitches we measured XANES of Mg, Al and Si (Wong et al., 1994, 1995). The results of the XANES measurements shown in Figs. 4, 5 and 6 clearly show the good normalization of the sample signal, I, by the input signal, I₀, with no glitches at 1385.6 or 1438 eV. In Fig. 4 the pre-edge at 1303 eV of Mg metal is observed clearly and the intermediate region between 1305 and 1309 eV is also shown clearly in the XANES. In Fig. 5 it is shown that the near-edge structure of Al metal has a hollow at approximately 1567 eV. Fig. 6 shows the near-edge structure of Si metal measured at a high resolution (Sham et al., 1992). Kossel theoretically explained that the fine structure of the near edge is due to the excitation of inner-shell electrons to an unoccupied level (Kawai & Mayers, 2000). The Kossel theory is called the short-range-order theory because the electronic structures of unoccupied levels are mostly determined by hybridization between the centre atom and the nearest-neighbour atoms (Kawai & Mayers, 2000). Thus, we expect our undulator beamline BL15XU to provide users with substantial experimental data of electronic structures of unoccupied levels using the YB₆₆ double-crystal monochromator at the third-generation light source SPring-8.

Figure 4 XANES of Mg metal. The pre-edge at 1303 eV is clearly observed and the intermediate region between 1305 and 1309 eV is also shown clearly.

Figure 5 XANES of Al metal. It is shown that the near-edge structure has a hollow at approximately 1567 eV.

Figure 6 XANES of Si metal. It is shown that the near-edge structure of Si metal is measured at high resolution.

4. Conclusions

A double-crystal YB₆₆ monochromator was constructed on a helical-undulator beamline at the third-generation light source SPring-8. It was confirmed that positive glitches do not appear in the transmission function of a YB₆₆ 004 double-crystal monochromator. Furthermore, XANES of Mg, Al and Si were measured without a low-pass mirror filter. It is therefore possible to investigate the electronic structures of unoccupied levels of advanced materials with this undulator beamline. From these experiments it has been clarified that a YB₆₆ double-crystal monochromator is well suited for soft X-ray beamlines on third-generation light sources.