P e r f o r m a n c e o f B L 3 5 X U f o r H i g h R e s o l u t i o n I n e l a s t i c X - r a y S c a t t e r i n g Fig. 1. Energy resolution for different power loads on th e backscattering monochromator at 21.75 keV. A flat crysta l (at fixed temperature) was used to analyze the beam reflected during a temperature scan of the monochromator. Each mm of aluminum reduces the incident power by a factor of 2. The grazing incidence geometry removes all of the broadening caused by the ~100 mW of beam (no Al). It als o reduces the shift that occurs because the temperature sensor is not exactly at the beam s pot. (Note: the r elevant scal e factor is about 18 mK /m eV at this X-ray energy ). Temperature Difference (m K) bservin Investigation of sample dynamics at meV energy scales and Å l ength scales is now possible at beamline BL35X U [1] using inelastic X-ray scattering (I X S). I XS o ffers some unique advantages as c ompared to neutron scattering techniques, but comes at the expense of an extremely technically demanding setup. T he primary advantages of IX S appear in m easurements of disordered material s and s mall s amples. More precisely, kinematic crystals having 2- meV resolution in t he s pac e of five m onths, (including t he two and a hal f m onth summer s hutdown), very much faster t han any comparable facility. Since the first user experiments in October 2001, steady progres s has improved t he resolution to between 1.5 and 1. 8 meV (depending on the analyzer crystal ) with as much as 5 × 10 9 photons/s in a 0. 8- me V bandwidth and a φ 75 - μ m s pot at the sample. restrictions in neutron scattering ca n m a k e i t d i f f i c u l t t o w o r k a t s m a l l momentum transfers, which is a crucia l region for o g co llective modes in non-crystalline materials [2]. In addition, typical neutron beam sizes are ~ cm 2 , so t hat c omparably l arge s amples are needed. T his may be compared to an X -ray beam from a 3rd generation synchrotron s ource t hat c an easily be focused to a spot of ~ 100 μ m or less in diameter. This is a great advantage fo r t he investigation of materials t hat ar e difficult to prepar e, or for t he use of extreme sample environments (hig h pressure and/or high temperature). T he technical difficulty of IXS stems f r o m t h e e x t r e m e l y h i g h e n e r g y resolution r equired. W hereas me V resolved measurements using neutrons with ~ 20 -m eV energy require resolutions of ∆ E/E ~ %, with X-rays of energy a fe w t ens of keV, t he r equired resolution jumps some five orders of magnitude to ∆ E/E ~ 10 -7 , l eading to a commensurat e increase in difficulty. Despite this , B L35XU has gone from t he very firs t tests of the spectrometer using a singl e analyzer in May and June of 2001 to a working setup with four analyzer crystals 104 N o A l 4 5 m K F W H M @ 7 4 1 m m A l 2 8 m K F W H M @ 2 5 2 m m A l 2 4 m K F W H M @ 0 N o r m a l I n c i d e n c e G r a z i n g I n c i d e n c e I n t e n s i t y ( a . u . ) I n t e n s i t y ( a . u . ) – 5 0 0 5 0 1 0 0 1 5 0 2 0 0 0 . 5 0 1 0 . 5 0 1 N o A l 2 4 m K F W H M @ 1 8 1 m m A l 2 2 m K F W H M @ 5 2 m m A l 2 3 m K F W H M @ – 2 Fig. 1 Figure 2 Table I Here , we briefly discuss so me of the instrumentation t hat has made this possible. T wo e ssential components in t he operatio n o f t he s pectrometer are t he high resolutio n backscattering m onochromator and t he analyze r crystals. The monochromator, having to accommodat e only the divergence of the undulator beam (som e 15 × 40 μ rad 2 , or less) is the easier one, but, even so, it provided one notable surprise ( beyond th e expected difficulty of dealing with mK temperature control). In particular, it was found that the beam from the high heat load (Si (111)) monochromator, ~ 100 mW of power, c aused local heating and distortion of the backscattering crystal over the beam spot. T hus, it was necessary to reduce the power density onto t he backscattering m onochromator in order to achieve resolution better than 2 meV. This was done by replacing the normal incidenc e backscattering monochromator with a grazin g incidence version: working at a grazing angle of some 2.5 degrees reduc ed t he power density by a factor of 20, and removed nearly all the e ffects of the heat load, as is evident in . T he analyzer crystals are more difficult, as they must a ccommodate a solid angle of ~ 10 × 10 mrad 2 , some five orders of m agnitude more t han t he monochromator . T hus t here has been an ongoing R&D program with NEC F undamental Research Laboratory over t he last f our y ears to achieve good quality crystals. We have been following a prescription similar to that of [3]: first a flat wafer is cross-cut to produce m any free- standing crystallites with a thin common back-wall ; t hen these crystallites are bonded to a substrate of appropriate curvature; and finally the common bac k wall is removed, leading to many independent and, importantly, unstrained crystallites with the correc t orientation. This process remains a bit of an art, but slow and mostly steady progress has led to a set of four crystals giving resolutions of 1.5, 1.6, 1. 6 of 1.8 meV (operating at a 10 -m radius, without any limiting apertures) using the Si (11 11 11) reflection at 21.75 keV. P resently , a 60 - or 70 - μ m thick sa w blade is used to cut 2.9 mm into a flat wafer on a 750 μ m pitch. T he crystallites are then bonded to a s pherically curved s ubstrate using a high temperatur e gold diffusion bond, and the back-wall is removed, leaving about 15000 independent crystallites wit h the appropriate geometry (rms deviation from th e 9. 8- m radius of the substrate is ~ 15 μ rad, or better). Notably, a fter all etching, t he active area of th e crystallites is 60 to 65% of t he s ubstrate area, which is rather good. shows one of th e analyzer crystal s, while the gives detailed Fig. 2. 10-cm diameter analyzer crystal. Resolutions are presented in the Table I. Table I. Resolution at various silicon orders with the best analyzer crystal. FWHM = Full Width at Half Maximu m FWM/X = Full Width at Maximum over X E n e r g y S i ( n n n ) R e l . F l u x 1 5 . 8 1 ( 8 8 8 ) 1 5 . 3 8 . 3 4 . 2 4 7 2 9 1 6 1 7 . 7 9 2 1 . 7 5 ( 9 9 9 ) ( 1 1 1 1 1 1 ) 6 . 1 3 . 1 1 . 5 8 2 1 F W H M ( m e V ) F W M / 1 0 ( m e V ) F W M / 1 0 0 ( m e V ) 105 Vacuum Flight Pat h Chamber at 10 m 4 Analyzer Crystal s 4 Channel Detector Incident Beam Sampl e Granite Base with Airpads IXS Spectromete r 10 m Horizontal Arm to 55 º in 2 Θ Fig. 3 Fig. 3. The horizontal arm of the IXS spectrometer. Focused, highl y monochromatic beam is incident from the right and is scattered by th e sample into a set of four analyzer crystals that focus it into four detectors, allowing simultaneous measurement of four momentum transfers. References [1] A.Q.R. Baron, Y. Tanaka, S. Goto, K. Takeshita, T. Matsushita and T. Ishikawa. J. Phys. and C hem. Solid s 61 ( 2000) 461; A.Q.R. Baron, Y. T anaka, D. Miwa, D. Ishikawa, T. Mochizuki, K. Takeshita, S. Goto, T. Matsushita and T. Ishikawa, Nucl . Instrum. and Meth. in Phys. Res. A 467-468 (2001) 627. [2] F. Sette et al . , Scienc e 280 (1998) 1550; S. Hosokawa et al . , in this issue of SPring-8 Researc h Frontiers . [3] C. Masciovecchio et al . , Nucl . Instrum. and Meth . B 111 (1996) 181. parameters. We are presently working to improv e all crystals to the same 1. 5- meV level of resolution. The IXS spectrometer is shown in the final figure ( ). In the 19 months since first beam from a single analyzer crystal, and t he 14 m onths sinc e opening f or user operation, t he s pectrometer has been used by many user groups to investigate a variety of liquid and solid materials, including several types of molten metals, mercury near th e liquid-gas critical point (so both at high temperature and high pressure), and many crystalline materials, including several types of superconductors and one quasi-crystal. Alfred Q. R. Baron, Yoshikaz u Tanaka and Satosh i Tsutsu i SPring-8 / JASR I E-mail: baron@spring8.or.jp 106
home
- JP
- EN