I. INTRODUCTION
As one of the seven Phase I beamlines of Shanghai Synchrotron Radiation Facility (SSRF), BL15U1 is fully dedicated to hard X-ray micro/nano-spectrochemical non-destructive analysis, including X-ray fluorescence (XRF), X-ray absorption spectroscopy (XAS) and X-ray diffraction (XRD) in hard X-ray regime. The beamline permits studies in research fields of medicine, biology, earth and planetary sciences, environmental science, archaeometry and materials science, etc. The beamline provides a non-destructive investigation of the spatial distribution, concentration and speciation of trace elements that correlated to the morphology and crystallographic orientations, with spatial resolutions from a few micrometers to sub-micrometers.
II. X-RAY SOURCE AND BEAMLINE OPTICS
To minimize the gap and obtain continuous tuning curves in desired energy region, the light source of BL15U1 is a 2 m in-vacuum hybrid undulator with 80 periods of 25-mm. It can generate 1–11 harmonics, and provide X-rays of 5–20 keV by tuning its gap from 6 mm to 20 mm. Its upper and lower poles can be adjusted independently, with the maximum gap taper adjustment range of 0–0.2 mm. Figure 1 shows a schematic optical layout of BL15U1.
The source light generated by the undulator is sized at 379 μm (horizontal) and 22 μm (vertical), while the simulation results are 380 μm and 24 μm. Because of the large horizontal source size, two-step focusing scheme (prefocusing-microfocusing scheme) is adopted to obtain micro-sized focused beam. The key optical components of the beamline include white beam slits, prefocusing mirror, double crystal monochromator and the secondary source slits. The white beam slits at 20 m from the source defines a beam acceptance of 80 μrad×50 μrad for the downstream beamline. A toroidal mirror at 22.6 m from the source is used to demagnify the horizontal source by 2:1 and parallelize the beam in vertical direction. This toroidal mirror forms a secondary horizontal source, which is defined by the slits at 33.8 m from the source, enabling rapid change of focal spot size at the sample position. Since the vertical source size is just 22 μm, the toroidal mirror is designed as horizontal reflecting. Thus, the effect of figure error to the beamline performance is greatly reduced. A double crystal monochromator is used to select the desired photon energy. A K-B mirror microfocusing system at 39.2 m from the source is used to focus the beam on the sample, with a pair of focusing mirrors in focal lengths of 517 mm (vertical) and 270 mm (horizontal).
In order to calculate characteristic parameters of the beamline, the tracing software Shadow VUI was used to simulate the beamline performance, such as the photon flux, focal spot size and energy resolution at the sample, the transmission characteristics and thermal load distribution of the main components, etc. The simulation was done under the following parameters [1]: white beam slits opening, 1.7 mm×0.8 mm; length of prefocusing mirror, 800 mm; the K-B system: length of horizontal mirror, 16 cm, length of vertical mirror, 12 cm; grazing incident angle, 3.5 mrad; the secondary source slits and the entrance slit to K-B mirrors were opened completely. Reflector figure error and the first crystal deformation acted in monochromator were taken into account in the simulation.
Here we give just the main calculation results of the focused beam at the sample. The incident divergence angle of the monochromator was about 1 μrad (FWHM). Taking into account the hot deformation of the crystal, the output divergence angle was about 7 μrad (FWHM). As the divergence angle is much smaller than the crystal’s Darwin width, energy resolution of the monochromator shall be close to its intrinsic resolution. The calculated energy resolution is shown in Table 1. It can be seen that the X-ray energy resolutions at the sample are very close to the intrinsic bandwidth of monochromator Si (111). This means that the energy resolution meets the requirements of μ-XAFS experiments.
| X-ray energy (eV) | 6500 | 9600 | 12600 | 15200 | 18000 |
|---|---|---|---|---|---|
| Bandwidth of mono-chromator Si(111) (eV) | 0.842 | 1.26 | 1.66 | 1.97 | 2.39 |
| Bandwidth at sample (eV) | 0.853 | 1.26 | 1.68 | 2.00 | 2.39 |
The K-B mirrors are in helium environment and the atmospheric transmission distance to the sample is about 10 cm. An X-ray beam, without refinement by the secondary source slits, can be focused to 7.7 μm×2.2 μm (H × V) at the sample (Fig. 2(a)). Refining the X-ray beam size to 30 μm×300 μm using horizontal and vertical slits of the secondary source, the focused beam size at the sample is calculated at 1.1 μm×1.6 μm (H × V) (Fig. 2(b)).
As the horizontal source size is larger than the vertical, the refine condition is not consistent with each other. Figure 3 shows the calculated energy spectra at the sample (Si(111), with the grazing incident angle of the micro focus mirror being 3.5 mrad). The black line is without refinement of the secondary source slits, while the red line is with the refined the X-ray beam (30 μm×300 μm). For μ-EXAFS analysis, because of the high flux density, total flux requirement is relatively low, and 1010–1011 photons/s flux can meet the most requirement; while for μ-XANES analysis, meaningful experiments can be carried out with 108 photons/s flux. Therefore, performance of BL15U1 can meet the high spatial resolution and high sensitivity requirements for μ-XRF, μ-XAFS and μ-XRD experiments. Table 2 lists the main specifications of BL15U1. Here we give just the test condition of energy range as a typical example to show the measured results of the beamline performance. Test conditions for Cu and Mo: beam current, 158 mA; undulator gap, 8.74 mm; taper, 0.12 mm; and the secondary light source slits, 50 μm×300 μm. Test conditions for Ti: beam current, 133.9 mA; undulator gap, 7.738 mm; taper, 0.12 mm; and the secondary light source slits, 500 μm×500 μm. Each energy integral is one second. The results are shown in Fig. 4. The absorption edge of Ti is the cumulative result of five tests. The test results show that the photon energy range covers 5–20 keV and achieves the design target. Other test processes were not shown detailedly here. From Table 2, desired energy resolutions are achieved. The minimum focal spot size is better than 2 μm by K-B mirrors, and 150 nm resolution can be achieved by using zone plate focusing.
| Source | Design specifications | Measurement results |
|---|---|---|
| Energy | 5–20 keV | 4.96–20.1 keV |
| Energy resolution (ΔE/E) | < 2×10-4 Si (111) | < 1.37×10-4 Si (111) |
| Minimum beam size (FWHM) | < 2 μm×2 μm (H×V) by K-B mirror | <1.60(19) μm× 1.81(1) μm)(H×V) by K-B mirror |
| < 0.2 μm×0.2 μm (H×V) by zone plate | ∼150 nm×150 nm (H×V) by zone plate | |
| Photon flux (@10 keV) | 1010-1011 photons/s/μm2(0.1% BW)-1 | 1.81×1011 photons/s/μm2(0.1% BW)-1 |
| Beam position stability | < 1 μm/8 h | < 1 μm/8 h |
| Experimental techniques | μ-XRF, μ-XAFS, μ-XRD | μ-XRF, μ-XAFS, μ-XRD |
III. ENDSTATION
A. Microprobe setup
In the experiment hutch, the microprobe setup is installed on a granite table of 3 m(l)×1.5 m (w)×0.4 m (h), which offers high rigidity and stability for equipment installation. A schematic layout of the microprobe setups is shown in Fig. 5. The K-B mirror system produces high-flux beam of micron spot size in the photon energy range of 5–20 keV.
B. K-B mirrors
The K-B mirror system (IDT Inc, UK) is located at 39.2 m from the X-ray source. The horizontal and vertical mirrors are 24 cm in length. A picture of the K-B mirrors system is shown in Fig. 6. The required elliptical mirror surface is produced by using two asymmetric bending moments on both ends of the mirror. By remote-control, the K-B mirror optics can be moved into the beam by translation motion and tilted a specific glancing angle to reflect the beam and focus a desired image point by adjusting mirror curvature.
C. Sample stages
A 7-axis sample stage (KOHZU, Japan) is used for sample positioning and 2-D or 3-D mapping experiments (Fig. 7). The XYZ stages have a scanning precision accuracy of 100 nm, with the encoder resolution of 50 nm. The manipulator has a load capacity of 2.3 kg.
D. Detectors
Detectors of the beamline include: a mini-ionization chamber (2 cm thickness) from ADC, USA; three ionization chambers with the electrode lengths of 5 cm, 14 cm and 28 cm, from OKEN, Japan; a 7-element Si(Li) detector (e2v, USA) with high speed XIA DXP electronics, for fluorescence detection; two single Vortex-90EX Si drift detectors (SDD) from SII, USA, which provide excellent energy resolution and high output counting rate (<136 eV @5.9 keV, at 100 kcps counting rate); and a Mar165 CCD, for X-ray diffraction experiments.
E. Zone plate nanoprobe setup
An overview of the experimental arrangements of hard X-ray nanoprobe is depicted in Fig. 8. The zone plate nanoprobe is composed mainly of three parts: a) focusing optics module including zone plates and an order-sorting aperture (OSA), to focus hard X-ray beams to nanoscale; b) sample module including a piezo stage, to raster the sample on the beam; and c) detector module including a visible light microscope (VLM) to visualize the regions of interest of the sample and align the optical elements such as zone-plates and OSA, and the SDD to detect trace elements. The VLM not only plays an "eye" role for sample searching, but also aligns all main optical elements (zone plates, OSA) with YAG-crystal in optical path. To avoid collision, the sample stage and VLM share the same heavy motor along the beam direction. The zone plates provide practical diffraction efficiency of ∼5% and a monochromatic beam flux of about 4×108 photons@10 keV. The resulting spot size at the focal plane of 135 nm×145 nm (V× H) is shown in Fig. 9.
F. Other instruments
An online microscope for sample alignment is installed above the stage at 45 of the beam direction. Another microscope is installed along the beam direction downstream to the sample. Two offline microscopes are available for users to prepare experiments. A LINKAM temperature controlled stage (TMS-600) is available for users. The temperature is adjustable from -196 ℃ to 600 ℃ with 0.1 ℃ precision.
IV. METHODOLOGY AND DEVELOPMENT
On BL15U1, 2D-mapping experiments of μ-XRF, μ-XAS and μ-XRD can be performed.
A. Micro X-ray fluorescence
XRF spectroscopy is a powerful non-destructive method for analysis of trace elements for researches in geology, archaeology, biomedical science and material science, etc. BL15U1 enables XRF multi-element mapping and quantitative analysis (down to sub-ppm level) with a micro- or submicro-sized beam. Geometry of BL15U1 XRF analysis is shown in Fig. 10. The sample is placed at 45 to the beam incidence, while the X-ray detector, either Si(Li) or SDD, is perpendicular to the beam incidence. An ion chamber is used for measuring the incident photon intensity. An optical microscopy is faced to the sample holder for sample observation. Two codes, GeoPIXE [2] and PyMCA [3], are available for quantitative analysis. The scanning μ-XRF is performed in step-by-step mode. The XRF spectrum of each pixel is saved by sequence for quantitative analysis. Regions of interest (ROIs) in a spectrum can be selected. Once ROIs are created, they are applied to all spectra. ROIs can be used to generate multi-element 2D images.
On BL15U1, a user group developed a method for quantitative imaging of trace elements in sections of bio-tissues using μ-XRF analysis [4]. Another user group used μ-XRF to study the varved sediment of lakes and the modern process of ancient climate index [5-7], finding that μ-XRF provides a potential tool for high resolution in-situ elemental analysis of sedimentary sequences [8]. The μ-XRF has also been applied in nanobiology to study location and biodistribution of nanoparticles in tissues [9] and even in a single cell [10-13], assisting to clarify their biological effect with other techniques. Besides, the fast scanning μ-XRF imaging based on the EPICS platform has been realized on BL15U1. This technique integrated multi-functions such as the precise motion control, light intensity and fluorescence probing, etc. By using this on-the-fly XRF imaging technique, elemental distributions of Cu, Zn, K and Fe in murine spleen slice were imaged in 140 min (at 10 keV, beam size 5 μm×5 μm, scanning time 0.02 s/point, 530× 647 pixels) [14], rather than dozens of hours by conventional XRF technique dozens of hours were needed. Figure 11(1) shows that the on-the-fly XRF imaging technique made a qualitative leap in the acquisition speed under the premise of high quality for fluorescence data.
B. Micro X-ray absorption spectroscopy
X-ray absorption near edge structure (XANES), also known as near edge X-ray absorption fine structure (NEXAFS), is the region of X-ray absorption spectrum within ∼ 50 eV of the absorption edge. The XANES region is sensitive to oxidation state and geometry of the absorbing atom. Much chemical information can be extracted from the XANES region: formal valence (very difficult to experimentally determine in a nondestructive way); coordination environment (e.g., octahedral or tetrahedral coordination) and subtle geometrical distortions of it. Compared with EXAFS, XANES has several advantages. For instance, an XANES spectrum is measured simply and quickly as it concentrates in a small energy region; it can be recorded at reaction temperature due to its weak temperature dependence; it is sensitive to chemical information such as valence and charge transfer; and it can probe unoccupied electronic states, which is important in chemistry. Besides, XANES is often used as simple "fingerprint" to identify presence of a particular chemical species [15].
Generally, XANES spectroscopy is performed using mm-sized X-ray beams, measuring absorption coefficients integrated over a relatively large volume. With high brightness of synchrotron radiation sources and the new developments in the X-rays focusing devices, the intense micro-focusing X-ray beams enable XANES experiments in sub-micro spatial resolution. Currently, BL15U1 has a beam size down to 2 μm by using the K-B mirror for the μ-XANES measurement, which paves the way for a new level of experimental analysis of the heterogeneous samples or individual small structures in biological, environmental and materials science for the domestic users.
The photon energy range of 5–20 keV covers K- or L- absorption edges of most elements. Micro-XANES measurement can be performed in transmission or fluorescence mode. In transmission mode, ionization chambers of an appropriate gas are used for monitoring the incoming and transmitting beam intensity, while in fluorescence mode Si(Li) or SDD is used to detect the X-ray signals. Currently, 80% of the users use μ -XRF measurements while doing their μ-XANES experiments. Generally, 2D elemental mapping by μ-XRF is performed first, followed by μ-XANES measurements to determine the local atomic environment and chemical state of selected elements in microdomains. Especially, users of life science and environmental science prefer combining μ-XANES with μ-XRF [16-19]. For example, to systematically understand the health risks induced by As ingestion, Cui et al. adopted μ-XRF and μ-XANES to explore the As distribution and speciation from ingestion exposures to noninvasive human biomarkers (such as nails and hair) in Shanxi, China [20]. In-situ microdistribution and speciation analysis indicate that As was mainly associated with sulfur in nails and hair. Nails, rather than hair and urine, could be used as a proper biomarker for arsenicosis. Figure 11 shows the spatial distributions of As and S in a nail section and As speciation in nails and hair [20]. This indicates that combined μ-XRF and μ-XANES measurements improve the comprehensive analytical capability of the selected element in a sample.
C. Micro X-ray diffraction
Micro-XRD was developed on BL15U1 since 2010. It was gradually optimized to meet the users’ need, especially for the high pressure μ-XRD. Now μ-XRD is a routine technique of BL15U1. About 30% of beamtime was allocated to μ-XRD users per year.
The μ-XRD setup of BL15U1 was optimized for high pressure μ-XRD experiment. From up-stream to down-stream, 2 μm (FWHM) beam was provided by K-B focusing system. A clean beam profile is obtained by using a pin-hole, which can be of Φ30 μm, Φ50 μm and Φ100 μm, chosen on the users’ requests (Fig. 12, left). The sample stage has six motorized stages. Before an XRD experiment, the first thing is to place the axis center of the rotation stage at the center of beam focusing point by adjusting X1 and Y1 stages (Fig. 12, right). During the experiment, the users just need to adjust X2, Y2 and Z stages to put on the rotation center. A sample, such as a diamond anvil cell (DAC), can be fixed on the sample stage by using a kinematic base (BKL-4, Newport, USA), which can also be used in the off-line ruby system. By using a copper cross hair (Φ20 μm), we can easily transfer coordinate between the on-line XRD and the off-line ruby system. For finding a sample easily, we placed a PIN diode as photon detector between the sample and beam-stop. It can quickly move in and out of the beam path by an actuator. A Mar-165 CCD was used for μ-XRD experiment. And if necessary, we can replace it by a Mar-345 imaging plate.
The control and data collection software of the CCD was written based on EPICS. A code for 2D XRD scan was developed, too. To serve the users requesting high pressure measurement, an off-line ruby system was installed in preparation room of the beamline. A 473 nm laser and a spectrometer of 1200 g/mm and 300 g/mm gratings (Princeton Instruments Acton Series) are used. A by-pass CCD camera is used for sample observation. Coordinate unification of the ruby system and the on-line XRD is mentioned before. So, the sample we can be easily found after pressure measurement. By adopting the high pressure XRD technique on BL15U1, Sun et al. investigated how the iron selenide behaved under pressure, which was a new addition to the high-temperature superconducting family [21]. Both in-situ high-pressure resistance and susceptibility measurements were conducted to detect the superconductivity of the new iron chalcogenides of Tl0.6Rb0.4Fe1.67Se2 and K0.8Fe1.7Se2 at high pressures and low temperatures. Their findings were published in Nature and Physical Review Letters [21, 22]. Besides the high pressure XRD, the submicro-XRD will be open to users needing smaller beam to study their samples.
V. SUMMARY AND OUTLOOK
During the past six years, BL15U1 has rapidly grown up and contributed a lot to much important users’ work, some of them were published in Nature, PNAS, PRL, Advanced Materials, etc. But this does not mean the development of BL15U1 is finished. We plan to make progress in two major fields in the near future: nano probe and micro-scattering.
In 2013, collaborating with Dr. H. K. Mao, we developed X-ray Raman (inelastic X-ray scattering at fixed angle) system for high pressure application. Metallic state of hydrogen is a well-known problem predicted by theorists decades ago [23]. But until now, it lacks convinced experimental evidence. A major challenge is how to detect electronic state of hydrogen at extreme high pressure. X-ray Raman is a promising technique for this problem as it can directly probe energy gap of hydrogen in extreme conditions. We plan to construct an inelastic X-ray scattering (IXS) system based on the success of X-ray Raman experiment at BL15U1. The angle range of IXS will be from 0° to 90°. To have an energy resolution of 0.2 eV for IXS, we will build a high resolution monochromator with two pairs of channel-cut crystal.
A nano-focusing (100 nm) fluorescence system has been built on BL15U1. But it has not been used so often, as the flux is just about 108 photons s-1 μm-2 (0.1% BW)-1@10 keV, with which elements in ppm concentrations can be hardly detected. So, improving the flux and efficiency of detection shall be our main task, with a new detector system composed of backscattering detector and 4-element Si(Li) detector working at transmission mode. Another problem for nano-focusing system is the stability. Nano scale vibration control is a challenge. We plan to build a quick feedback system similar to the Advanced Photon Source [24] and the Hard X-ray Nanoprobe (HXN) beamline at the NSLS-II [25] to improve the stability of our nano-focusing system.
Quantitative analysis of PIXE spectra in geoscience applications
. Nucl Instr Meth B, 1990, 49: 271-276. DOI: 10.1016/0168-583X(90)90259-WA multiplatform code for the analysis of energy-dispersive X-ray fluorescence spectra
. Spectrochim Acta B, 2007, 62: 63-68. DOI: 10.1016/j.sab.2006.12.002Quantitative imaging of element spatial distribution in the brain section of a mouse model of Alzheimer’s disease using synchrotron radiation X-ray fluorescence analysis
. J Anal At Spectrom, 2010, 25: 328-333. DOI: 10.1039/b921201aEvidence for decreasing South Asian summer monsoon in the past 160 years from varved sediment in Lake Xinluhai, Tibetan Plateau
. J Geophys Res, 2011, 116: DO2116. 328-333. DOI: 10.1029/2010JD014454Minor element variations during the past 1300 years in the varved sediments of Lake Xiaolongwan, northeastern China
. GFF, 2013, 135: 265-272. DOI: 10.1080/11035897.2013.788550High resolution elements geochemical record during the past 1200 years in Hugangyan Maar Lake
. Quaternary Sci, 2015, 35: 152-159. (in Chinese) DOI: 10.11928/j.issn.1001-7410.2015.01.14Application of synchrotron radiation technology in high-resolution paleoclomatic and paleoenviromental research
. Quaternary Sci, 2010, 30: 1218-1224. (in Chinese) DOI: 10.3969/j.issn.1001-7410.2010.06.20Synchrotron radiation X-ray fluorescence analysis of biodistribution and pulmonary toxicity of nanoscale titanium dioxide in mice
. Analyst, 2013, 138: 6511-6516. DOI: 10.1039/c3an01267kFates of Fe3O4 and Fe3O4@SiO2 nanoparticles in human mesenchymal stem cells assessed by synchrotron radiation-based techniques
. Biomaterials, 2014, 35: 6412-6421. DOI: 10.1016/j.biomaterials.2014.04.052Double-walled Au nanocage/SiO2 nanorattles: integrating SERS imaging, drug delivery and photothermal therapy
. Biomaterials, 2014, 35: 6412-6421. DOI: 10.1002/smll.201401360Multifunctional photosensitizer-conjugated core-shell Fe3O4@NaYF4:Yb/Er nanocomplexes and their applications in T2-weighted magnetic resonance/upconversion luminescence imaging and photodynamic therapy of cancer cells3
. RSC Adv, 2013, 3: 13915-13925. DOI: 10.1039/c3ra41916aEnhanced doxorubicin transport to multidrug resistant breast cancer cells via TiO2 nanocarriers
. RSC Adv, 2013, 3: 20855-20861. DOI: 10.1039/C3RA42863JFast sanning X-ray microprobe fluorescence imaging based on synchrotron radiation
. Acta Phys Sin, 2011, 60: 090702. (in Chinese) DOI: 10.7498/aps.60.090702EXAFS Data Collection and Analysis Course
.Full assessment of fate and physiological behavior of quantum dots utilizing Caenorhabditis elegans as a model organism
. Nano Lett, 2011, 11: 3174-3183. DOI: 10.1021/nl201391eOral magnetite nanoparticles disturb the development of Drosophila melanogaster from oogenesis to adult emergence
. Nanotoxicology, 2015, 9: 302-312. DOI: 10.3109/17435390.2014.929189Strategy for characterization of distribution and associations of organobromine compounds in soil using synchrotron radiation based spectromicroscopies
. Anal Chem, 2014, 86: 11002-11005. DOI: 10.1021/ac503280vIn-situ determination of metallic variation and multi-association insingle particles by combining synchrotron microprobe, sequentialchemical extraction and multivariate statistical analysis
. J Hazard Mater, 2014, 276: 241-252. DOI: 10.1016/j.jhazmat.2014.05.039Arsenic levels and speciation from ingestion exposures to biomarkers in Shanxi, China: implications for human health
. Environ Sci Technol, 2013, 47: 5419-5424. DOI: 10.1021/es400129sRe-emerging superconductivity at 48 kelvin in iron chalcogenides
. Nature, 2012, 483: 67-69. DOI: 10.1038/nature10813Pressure-driven quantum criticality in iron-selenide superconductors
. Phys Rev Lett, 2012, 108: 197001. DOI: 10.1103/PhysRevLett.108.197001Metallic hydrogen: a high-temperature superconductor?
Phys Rev Lett, 1968, 21: 1748-1749. DOI: 10.1103/PhysRevLett.21.1748Linear actuator system with 1-Å closed-loop control resolution and 50-mm travel range
. Proc SPIE, Optomechanical Design and Engineering, 2002, 4771: 78. DOI: 10.1117/12.482148Piezo control for 1 nm spatial resolution synchrotron X-ray microscopy
. J Phys Conf Ser, 2014, 493: 012026. DOI: 10.1088/1742-6596/493/1/012026