P13, the EMBL macromolecular crystallography beamline at the low-emittance PETRA III ring for high- and low-energy phasing with variable beam focusing

Abstract

The macromolecular crystallography P13 beamline is part of the European Molecular Biology Laboratory Integrated Facility for Structural Biology at PETRA III (DESY, Hamburg, Germany) and has been in user operation since mid-2013. P13 is tunable across the energy range from 4 to 17.5 keV to support crystallographic data acquisition exploiting a wide range of elemental absorption edges for experimental phase determination. An adaptive Kirkpatrick-Baez focusing system provides an X-ray beam with a high photon flux and tunable focus size to adapt to diverse experimental situations. Data collections at energies as low as 4 keV (λ = 3.1 Å) are possible due to a beamline design minimizing background and maximizing photon flux particularly at low energy (up to 10¹¹ photons s^-1 at 4 keV), a custom calibration of the PILATUS 6M-F detector for use at low energies, and the availability of a helium path. At high energies, the high photon flux (5.4 × 10¹¹ photons s^-1 at 17.5 keV) combined with a large area detector mounted on a 2θ arm allows data collection to sub-atomic resolution (0.55 Å). A peak flux of about 8.0 × 10¹² photons s^-1 is reached at 11 keV. Automated sample mounting is available by means of the robotic sample changer `MARVIN' with a dewar capacity of 160 samples. In close proximity to the beamline, laboratories have been set up for sample preparation and characterization; a laboratory specifically equipped for on-site heavy atom derivatization with a library of more than 150 compounds is available to beamline users.

Keywords: 2θ-detector arm; low-energy data collection; phasing; softer X-rays; sulfur SAD; tunability; variable beam.

Figure 1

P13 beamline. (a) Optical layout; assembly drawing of (b) the optics hutch, (c) the experimental hutch, (d) predicted flux of the 2 m U29 undulator through a 1 mm × 1 mm aperture at 40 m distance from the source at the PETRA III storage ring (DESY, Hamburg), (e) reflectivity of photon beam versus energy through the double horizontal deflecting mirror for Rh coating (black line) and uncoated silica (red line).

Figure 2

Beam size and photon flux. (a) Total photon flux as a function of energy at the sample position for fully focused beam; (b) time course of beam profile during the vertical defocusing procedure from 24 µm to 70 µm over a period of 180 s; (c) beam profile of the fully focused beam 30 µm (H) × 24 µm (V) at the sample position, determined with the built-in scintillator of MD2; (d) defocused beam of 150 µm (H) × 70 µm (V) size; (e) MD2 penta-aperture with aperture sizes indicated.

Figure 3

P13 sample environment around the MD2 goniometer with the MARVIN sample changer on the left and the BCU at the back.

Figure 4

P13 ancillary facilities. (a) Drawing of the detector table with vertical 2Θ-arm. (b) Diffraction resolution versus energy at the edge of the PILATUS 6M-F at 2Θ = 0° (full line) and at 2Θ = 25° (dashed line). (c) PILATUS 6M-F equipped with the helium cone assembly. (d) Back panel of the DECTRIS PILATUS 6M-F with the standard N₂ gas inlet highlighted, used also to vent the complete detector with helium gas. (e) Detail of the sample position with helium cone assembly on; the post-sample air gap is reduced to 30 mm at the closest sample-to-detector distance of 135 mm. (f) EMBL derivative laboratory (DLab) located about 15 m away from the beamline.

Figure 5

Diffraction images acquired at λ = 2.75 Å (4.5 keV). 1.0° rotation images with an exposure time of 1 s were collected with 3.8 × 10¹⁰ photons s⁻¹ in a beam defocused to 50 µm in diameter at 135 mm crystal-to-detector distance on a crystal of lysozyme (120 µm × 120 µm × 80 µm in size) mounted in a litho-loop (MiTeGEN, Ithaca, USA). Images were displayed with ADXV (Arvai, 2015 ▸) using identical contrast levels (−1/+10) with inverted colours. Diffraction images were taken in face-on (top) and edge-on (bottom) orientation of the mounting loop.

Figure 6

Data quality statistics for crystals of Zn-free insulin as calculated by SHELXC. (a) Anomalous signal-to-noise d′′/sig as a function of resolution. Statistics for data collected at 13 keV are shown in red, for data collected at 4 keV without and with 2Θ-offset in blue and green, respectively, for the combination of the two 4 keV data sets in black. (b) Multiplicity as a function of resolution. Colour codes are the same as in panel (a).