Source

BL02U1 and BL02U2 share a straight section in sector 02 of the SSRF, which is a third-generation light source with a 3.5 GeV storage ring 432 m in circumference. The upstream undulator (02U1), which delivers photons to BL02U1, is canted inboard by 3 mrad compared with the storage ring. The downstream undulator (02U2) is canted outboard by 3 mrad compared with the storage ring. The total canting angle between the two undulators is 6 mrad. Based on the 4.03 nm-rad emittance of the storage ring, the X-ray beam source has a size of 386 µm (FWHM) and divergence of 80 µrad (FWHM) horizontally, and a size of 26 µm (FWHM) and divergence of 23 µrad (FWHM) vertically.

The beamline source is a 1.52 m long in-vacuum undulator (IVU) comprising 69 periods of NdFeB magnets, each with a length of 22 mm. The IVU has a minimum gap of 6 mm according to the current operating conditions, a peak magnetic field of 0.902 T, and a maximum K value of 1.81. The energy of the fundamental harmonic ranges from 2.0 to 5.1 keV when the IVU gap is tuned in the range of 6–30 mm. The undulator phase error is approximately 1.7° (rms) over the entire gap range of 6–30 mm. The IVU produces a maximum of 2.74 kW of power in total, of which only 65 W is confined in the central cone accepted by the beamline (100 µrad horizontally and 50 µrad vertically). It delivers X-ray photons at 12.6 keV with a flux of 1.94×10¹⁴ phs/s/0.1%BW in the central cone when the IVU gap is set to 7.375 mm.

Beamline optics

The optical design was optimized in order to be compatible with the canted layout, deliver as high a photon flux as possible in a focused beam smaller than 20 µm, cover a photon energy range of 6–16 keV, monochromize the photons with an energy resolution of 2×10^-4, and obtain as large a working distance as possible for the experimental station. Along the beamline, the main optical components are the white-beam slit, white-beam scanning wire, DCM, first diagnostic screen, first HRM, second HRM, second diagnostic screen, first beam position monitor (BPM), monochromatic-beam slit, second BPM, and Kirkpatrick–Baez (K-B) focusing mirror pair, as shown in Fig. 1.

Fig. 1Full-screen View

(Color online) Schematic optical layout (not to scale) of BL02U1. All distances are given with respect to the undulator source position

The white-beam slit is located 20.5 m from the undulator, which defines the acceptance angle of the beamline as 0.1 mrad × 0.05 mrad horizontally and vertically. To withstand a heat load of up to 1500 W, the absorber is made of tantalum and has a sloped incident surface to accept the beam at a grazing angle of 2° to lower the power density. Water cooling is used to protect the absorber from the high heat loads.

The monochromator has a vertical geometry. To cover the photon energy range of 6–16 keV, a pair of Si(111) crystals in a (+n, -n) arrangement is used to monochromize the beam without dispersion. The Bragg rotation axis carries both crystals, with the first crystal at the rotation center. The exit beam position is fixed by translating the gaps between the crystals. The length of the second crystal covers the beam motion upon the variation of the Bragg angle to avoid the need for a second translation. Two piezo-positioners are mounted on the second crystal for fine pitch and roll motion adjustments.

The reflecting mirrors for the two HRMs, which are located at 28 and 30 m, were manufactured with the same geometric size for easy maintenance and cost saving. The grazing angle of each HRM is 3.5 mrad and they are coated with rhodium, which provides a high reflectivity of greater than 89% in the energy range of 6–16 keV.

The monochromatic beam slit is located at 35.8 m, which defines the acceptance angle of the downstream focusing system. The blades of the slits are two L-shaped tantalum blocks facing each other. Each block has two motors arranged perpendicularly to achieve horizontal and vertical motion. The blocks can be moved to overlap and completely cut off the beam.

Two elliptical surface mirrors with fixed curvatures, located at 40.575 and 41.845 m, are arranged as a K-B mirror pair to focus the beam at the sample position at 43.245 m, with demagnifications of 15.2 and 29.89 in the vertical and horizontal directions, respectively. As a compromise between the mirror acceptance angle and working distance in the experimental station, the grazing angle is 4 mrad for each mirror, which restricts the reflectivity of the rhodium coating for photon energies above 16 keV, but can improve the photon flux at 12.6 keV.

A white-beam scanning wire located upstream of the DCM and two diagnostic screens located downstream of the DCM are used for commissioning. The white-beam scanning wire comprises two crossing wires, and the scanning plane is perpendicular to the incoming beam. The scanning direction is 45°, with each wire in the scanning plane, to obtain photoelectric information in the horizontal and vertical directions simultaneously. Two diagnostic screens are placed downstream of the DCM and second HFM. Diagnostic screens equipped with YAG:Ce scintillators are used for monochromatic-beam analysis.

Two BPMs are placed at 32.9 and 39.9 m and used for monochromatic-beam position and direction monitoring. The BPMs are backscattered photodiode QBPMs from FMB Oxford. Each BPM has four diodes for detecting the scattered radiation from two switchable metal foils (one 0.5 µm nickel foil and one 0.5 µm titanium foil) used to detect photons in different energy ranges.

Beamline specification measurement

Through precise alignment, the beamline can efficiently deliver photons from the undulator to a sample. When the undulator gap is set to 7.375 mm, the photon flux spectrum has a range of 11–13 keV, as shown in Fig. 2 (a), which is the 5^th order harmonic of the undulator.

Fig. 2Full-screen View

(Color online) Characteristics of the photon beam: (a) spectrum of 5^th-order harmonic of undulator with a gap of 7.375 mm, (b) normalized absorption spectrum at Cu K-edge, (c) beam profile at sample position, and (d) photon flux from 6 to 16 keV at sample position

To evaluate the energy resolution, an absorption spectrum was obtained at the Cu K-edge, as shown in Fig. 2 (b). The absorption edge characteristics indicated that the beamline's energy resolution was better than 2×10^-4.

Knife-edge scans were performed to measure the beam profiles horizontally and vertically. At the sample position, the beam had horizontal and vertical profiles with 11.6 and 4.8 µm FWHM values, respectively, as shown in Fig. 2 (c).

Considering the tuning curve of the undulator, absorption in the Be window, and remaining air path to the sample, the photon flux at the sample position should be greater than 10¹² phs/s over an energy range of 6–16 keV. The flux was measured when the storage ring current was 200 mA and then normalized to 300 mA, as shown in Fig. 2 (d).

With its single-stage focusing geometry, the instability of the focused beam is caused by vibrations from the source, DCM crystal, and mirrors, which are present as beam spot vibrations and flux fluctuations. A fluorescence-based beam position online monitoring device [25] was used to measure the beam spot vibration with a sampling rate of 100 Hz and total sampling time of 10 s. An ionization chamber with high-frequency readout electronics was set downstream of the sample position to measure the fluctuation, which had a readout frequency of 100 Hz and total sampling time of 10 s. The results shown in Fig. 3 indicate that the highest vibrations occurred at 8–9 Hz. The liquid-nitrogen supply pump of the DCM, which was set to 28 Hz, made a small contribution to the vibration in the vertical direction. The stability (rms) of the beam position had values of 0.6 and 0.5 µm in the horizontal and vertical directions, respectively. The stability (rms) of the beam flux was 0.34%, according to the measurements.

Fig. 3Full-screen View

(Color online) Beam stability: (a) position stability measured at 100 Hz over period of 10 s, where the red spot in the window at the top right is the center position of the beam fitted by a Gaussian function for each frame; (b) Fourier transform of beam position stability; (c) photon flux stability measured at 100 Hz over period of 10 s, and (d) Fourier transform of photon flux stability

All the specifications of the beamline are listed in Table 1. At the sample position, the photon flux at 12.6 keV was 4.5×10¹² phs/s for a spot size of 11.6 µm × 4.8 µm, and the beam divergence was approximately 1.3 mrad × 0.25 mrad.

Detector

The detector used was a Dectris Eiger2 S 9M with a frame rate of up to 40 Hz. It has 3108 × 3262 pixels with a 75 μm × 75 μm size, for a total detection area of 233 mm × 245 mm. The layout of the experimental station includes a detector-to-sample distance range of 150–750 mm, which provides the maximum detectable 2-theta angle of 39.2° when the detector center superposes with the beam.

Diffractometer

The rebuilt experimental station employs a homemade diffractometer called Sealion, as shown in Fig. 4, which is composed of a one-axis air-bearing rotation motor, an on-axis microscope, a collimator, a beamstop, and docks for the cryostream and fluorescence detectors.

The on-axis microscope is composed of a 10× objective lens, 90° reflecting mirror, 12.5:1 zoom lens, and 1× camera tube with a 1/2″ CCD camera. Two holes with a diameter of 2 mm were drilled through the objective lens and reflecting mirror, which were precisely aligned to allow the passage of the beam. The microscope has a resolution of up to 900 lp/mm within a field of view (FOV) of 0.14 mm × 0.19 mm at the highest magnification mode, and a resolution of 270 lp/mm within an FOV of 1.7 mm × 2.3 mm at the lowest magnification mode. The working distance is 34 mm, which creates space to mount a sample.

The collimator is composed of two tantalum metal blocks with holes of different sizes (50 and 300 µm) for the purpose of reducing the stray beam.

Automation

There is a homemade automatic sample changer called Swordfish employed for automatic sample mounting. It is composed of a 6-axis robot arm, gripper, dewar that can contain 21 uni-pucks, and gas–liquid phase separator for a liquid nitrogen supply.

The data analysis pipeline Aquarium [26] is used for automatic data processing and can process raw data images automatically after data collection is completed.

User interface

A homemade user interface based on webpage access called Finback is shown in Fig. 5. It lists all the key information about the experiment at the top, including the electron beam current, photon energy, photon flux, sample ambient temperature, experimental hutch humidity and temperature, sample orientation angle, and detector distance. The control command module for the sample-changer robot is placed on the left side, which also shows the sample information. The signal from the sensor on the goniometer indicates whether the sample is mounted. The window in the middle is an interactive interface for sample centering, and the goniometer moves the sample to the cross-hairs when the user clicks it using the microscope view. The control command module for data collection is placed on the right side and lists all the key parameters for the experimental setting, which can be a test of either a single image or the whole dataset.

Fig. 5Full-screen View

(Color online) User interface of BL02U1 experimental station, which includes the key experimental information, robot control module, sample alignment module, and data collection module

Data collection

The beamline can deliver a brilliant beam with a flux density greater than 10¹⁰ phs/s/µm² to a sample, which could cause significant radiation damage and destroy a crystal in seconds, even at a temperature of 100 K. Thus, it is necessary to determine a suitable strategy for data collection using either beam attenuation or a short exposure time. The latter requires higher timing synchronization precision for the goniometer, shutter, and detector. The experimental station offers a sampling rate of up to 40 Hz, which means that the shortest exposure time for each image is 0.025 s, and 40 images are generated in 1 s. To verify the reliability, a set of comparative experiments was designed, as listed in Table 2. Six thaumatin crystals were used for data collection, and each crystal was exposed three times at three locations with different exposure times and transmittances. The transmittance was inversely proportional to the exposure time to maintain the same dose rate for each dataset. The sequence ordered by the exposure time for each crystal was reversed to prevent radiation damage from affecting the consistency of the experimental data. The data analysis results showed that the R_merge value for the inner shell of each dataset was approximately 0.04. There was no noticeable difference in the data quality obtained with long and short exposure times. However, a short exposure time with a high-transmittance beam obviously increased the experimental efficiency.

Complex structure: RBD-hACE2

The omicron and delta Covid-19 mutants are currently the two most important of the five "variants of concern (VOC)" defined by the World Health Organization (WHO). Omicron has appeared in 128 countries and regions and has attracted worldwide attention. The receptor-binding domain (RBD) of this mutant strain, which is the most critical RBD of the spike S protein, carries up to 15 amino acid mutations that cover the features of the alpha, beta, and gamma mutant strains. The delta variant is the most contagious variant of Covid-19 so far. Understanding the mechanisms by which the omicron and delta mutants recognize receptors and invade cells is critical for vaccine and drug development. Gao et al. resolved the crystal structures of hACE2 in complex with the RBDs of the omicron and delta variants at a resolution of 3.0 Å, showing the recombination of the interaction region and changes in the combined capability caused by the mutant [27].

Anti-infective therapy

The clpP gene is highly conserved. Abnormal activation of bacterial ClpP is another way to discover new antibiotics. However, it is still difficult to develop a selective Staphylococcus aureus ClpP activator that does not interfere with the function of ClpP in Homo sapiens. By resolving the crystal structures, the research group of Caiguang Yang identified (R) - and (S) - ZG197 as highly selective Staphylococcus aureus ClpP activators [28]. The key structural elements in Homo sapiens ClpP, especially W146, and their combined action with a C-terminal motif, are very helpful in the recognition of activators. These findings indicate that a species-specific activator of Staphylococcus aureus ClpP is a promising therapeutic agent for the treatment of Staphylococcal infections.

Methyltransferase ribozyme

Ribozymes can be utilized in a limited range of chemical catalyses, such as site-specific methyl transfer reactions. The research group of Lin Huang resolved the crystal structure of a ribozyme at a resolution of 2.3 Å, which showed a three-channel structure and RNA folds that generated a specific binding site for the methyl donor substrate. The space group of the ribozyme crystal was P222₁, with unit cell dimensions of a = 50.38 Å, b = 52.5 Å, and c = 99.9 Å (PDB entry:7V9E) [29]. The diffraction data were collected at a wavelength of 0.97918 Å and phased using a single-wavelength anomalous dispersion method with eleven barium atoms.