WIR SCHAFFEN WISSEN —HEUTE FÜR MORGEN Dr. Tim Grüne :: Paul Scherrer Institut :: [email protected] Electron Crystallography for Structure Solution — Latest Res- ults from the PSI Electron Diffraction Group Research Complex at Harwell Seminars 13th April 2017 Tim Grüne Electron Crystallography at PSI 1 - Crystallography under the Microscope Instrumentation available at C–CINA, University Basel 1. FEI Polara, 200keV, 512x512 Timepix detector, single axis ±30◦ 2. FEI Talos, 200keV, 1024x1024 Timepix detector, single axis ±40◦ RCa Harwell Seminars 13th April 2017 1/34 Tim Grüne Electron Crystallography at PSI The Lens System Gun / Source C1 • Lenses C1–C3 shape beam C2 • Crystallography: Parallel beam • Objective lens: sets effective detector distance to back- C3 focal plane = diffraction mode • C3 not present in all microscopes Sample Lenses cause distortions. Objective lens Diffraction Plane Imaging Plane RCa Harwell Seminars 13th April 2017 2/34 Tim Grüne Electron Crystallography at PSI Electron Microscope: Imaging Mode Plane Wave Object Lense Image Plane (Detector) Rays of equal origin focus on detector RCa Harwell Seminars 13th April 2017 3/34 Tim Grüne Electron Crystallography at PSI Electron Microscope: Imaging Mode Detector noise and radi- ation senstivity require low contrast images Martinez-Rucobo et al. Mo- lecular Cell (2015) 58, 1079– 1089 Plane Wave Object Lense Image Plane (Detector) RCa Harwell Seminars 13th April 2017 4/34 Tim Grüne Electron Crystallography at PSI Electron Microscope: Diffraction Mode Plane Wave Object Lense RCa Harwell Seminars 13th April 2017 5/34 Tim Grüne Electron Crystallography at PSI Electron Microscope: Diffraction Mode Image at Backfocal Plane = kFouriertransform of objectk If object = crystal: diffraction spots according to Laue conditions Backfocal Plane Rays of equal direction focus on detector Plane Wave Object Lense RCa Harwell Seminars 13th April 2017 6/34 Tim Grüne Electron Crystallography at PSI 2 - Experimental Considerations 1. Ewald sphere: refinement stability 2. Dynamic scattering 3. Detectors RCa Harwell Seminars 13th April 2017 7/34 Tim Grüne Electron Crystallography at PSI X–rays: The Ewald Sphere λ = 1Å, “normal” resolution: 2θmax = 40◦ • Assume: wrong detector distance • Diffraction spot calculated wrongly (red circle) • Reciprocal lattice becomes distorted 1 /λ correct distance incorrect distance Curvature of the Ewald sphere gauges the diffraction geometry RCa Harwell Seminars 13th April 2017 8/34 Tim Grüne Electron Crystallography at PSI Electrons: The Ewald “Plane” 1/λe = 1/0.025Å 1 • Typical X–ray wavelength λ = Å X − Typical e wavelength λe 0 025Å • = . • Radius of Ewald sphere 40x greater • Ewald sphere nearly flat 1 1 1 /λX = / Å RCa Harwell Seminars 13th April 2017 9/34 Tim Grüne Electron Crystallography at PSI Electrons: The Ewald “Plane” 1 1 0 025 /λe = / . Å “normal” resolution: • typical wavelength with X–rays: 1Å 2 1◦ θmax = typical wavelength with electrons: 0.025Å • 1 4◦ • opening angle ≤ − • Ewald sphere virtually flat Without curvature: impossible to refine both de- • tector distance and cell RCa Harwell Seminars 13th April 2017 10/34 Tim Grüne Electron Crystallography at PSI Summary: The Ewald “Plane” • Detector distance and unit cell parameters are strongly related • Wrongly set distance can lead to incorrect bond lengths • Distance refinement with X–ray data routine • Distance refinement with electron data = unstable • good: Distance calibration from powder sample • better: Distance calibration from chemical bond lengths RCa Harwell Seminars 13th April 2017 11/34 Tim Grüne Electron Crystallography at PSI Dynamic Scattering • Kinematic Theory of Diffraction: Every photon / electron / neutron scatters once in the crystal • |Fideal(hkl)| ∝ pIexp(hkl) • Dynamic Scattering: Multiple Scattering events occur • Electron Diffraction: Multiple Scattering occurs even with nanocrystals RCa Harwell Seminars 13th April 2017 12/34 Tim Grüne Electron Crystallography at PSI Dynamic Scattering Data from SAPO–34: I(−2,−1,1) > Idirect beam (Eiger chip, 256x256 px) RCa Harwell Seminars 13th April 2017 13/34 Tim Grüne Electron Crystallography at PSI Multiple (Dual) Scattering 1 • Outgoing ray ~So acts as incoming ray for re- 2 flection ~So. • Probability of re–reflection thickness de- ~Si pendent 1 ~So 2 ~So RCa Harwell Seminars 13th April 2017 14/34 Tim Grüne Electron Crystallography at PSI Multiple (Dual) Scattering Laue Conditions (accordingly~b and ~c): 1 ~S ~S a h1 ~Si ( o − i) ·~ = ~S2 ~S1 a h′ 1 ( o − o) ·~ = ~So 2 ′ (~So −~Si) ·~a = h1 + h = h2 Requirement for detrimental effect on I(h2,k2,l2) • I(h1k1l1) must be strong • I(h′,k′,l′) must be strong 2 2 • I(h2k2l2) must be weak ~So ~So • I(h1k1l1) and I(h2k2l2) on same frame RCa Harwell Seminars 13th April 2017 15/34 Tim Grüne Electron Crystallography at PSI Multiple (Dual) Scattering I(hkl) 10 5 • Re-reflection more likely for thicker crystal(path) • Percentage similar for all reflections on frame (2θ ≈ 0) 3 • 10% of strong reflection affects weak reflection Measured intensities “shifted” from strong to weak 2 ⇒ ⇒ Low resolution reflection under–, high resolution reflec- 1 tions overestimated ⇒ Covered during refinement by reduced B–factor: elec- I(h 1k1l1) I(h 2k2l2) tron diffraction includes map–sharpening RCa Harwell Seminars 13th April 2017 16/34 Tim Grüne Electron Crystallography at PSI e− Diffraction Studies with Timepix and Eiger TVIPS CMOS 1024x1024 Timepix 256x256 Eiger (PSI) 20 ≤ I ≤ 21 cts 0 ≤ I ≤ 1 ct Lysozyme (invd) 0 ≤ I ≤ 1 ct SAPO–34 crystal ≈ 1kHz,50µm × 50µm ≤ 23kHz,75µm × 75µm cut–off: 11809 cut–off: 16, 64, or 4096 (4, 8, 12 bit) dead time ≈ 0.01s dead time 3µs RCa Harwell Seminars 13th April 2017 17/34 Tim Grüne Electron Crystallography at PSI 3 - Structures RCa Harwell Seminars 13th April 2017 18/34 Tim Grüne Electron Crystallography at PSI Pharmaceutical I: Visualisation of Hydrogen Atoms H–atom positions can be refined against electron diffraction data CCDC: IRELOH, Dai et al., Eur. J. Org. Chem (2010), 6928-6937 Sample courtesy Novartis 3 m • Field of view: µ • dmin < 0.8Å • Hydrogen atoms in difference map even • Crystal: 1.6µm × 400nm • P212121: 85% completeness with poor model with 3 crystals • 1334 reflections, 195 parameters, 156 re- • a=8.06Å b=10.00Å c=17.73Å straints (RIGU) • R1 = 15.5%,Rcomplete = 18.5% RCa Harwell Seminars 13th April 2017 19/34 Tim Grüne Electron Crystallography at PSI Pharmaceutical II: Differentiation of Atom Types Data quality: recognition of atom types, C vs. O vs. N etc. (CCDC: EPICZA) d [Å] 5 2.5 1.25 1.0 100 12 11 80 10 9 60 8 I σ 7 I/ 40 6 Completeness 5 20 4 I/ Completeness [%] σI 3 0 2 0 0.1 0.2 0.3 0.4 0.5 0.6 d* [1/Å] 3 m • Field of view: µ • dmin = 0.87Å • 2545 refl., 258 param., 267 restraints • Crystal: 400nm diameter • a=11.35Å, b=12.7Å, c=13.0Å (RIGU) 1 15 9% 19 1% • P212121: completeness with 4 • all data: R = . ,Rcomplete = . crystals: 86% • R1 = 14.7%,Rcomplete = 18.0% RCa Harwell Seminars 13th April 2017 20/34 Tim Grüne Electron Crystallography at PSI Pharmaceutical II (EPICZA): Structure Solution Process Direct methods reveal H atoms HFIX: all except 1 H Final Structure =data quality =model quality RCa Harwell Seminars 13th April 2017 21/34 Tim Grüne Electron Crystallography at PSI Summary: Electron Diffraction of Organic Compounds • Structures can be solved with X–ray knowledge and methods. • Radiation damage present, but not (always) limiting • Kinematic approximation sufficient for high quality structures IRELOH EPICZA 1.6 2.2 x x 1.55 2.1 line fit f(x) = 0.92 * x + 1.06e-01 line fit f(x) = 0.96 * x + 3.52e-02 2 1.5 1.9 1.45 1.8 1.4 1.7 1.35 1.6 1.5 1.3 1.4 Bond lengths X model [Å] 1.25 Bond lengths X-ray model [Å] Bond lengths X-ray 1.3 1.2 1.2 1.15 1.1 1.2 1.25 1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 Bond lengths e- model [Å] Bond lengths e- model [Å] RCa Harwell Seminars 13th April 2017 22/34 Tim Grüne Electron Crystallography at PSI Thermolysin (sample courtesy Ilme Schlichting) • Spacegroup P6122 • Unit Cell 94.3 94.3 130.4 90◦ 90◦ 120◦ • dmin = 3.5Å • 72.4% completeness • MR with 3DNZ poly Alanine: TFZ=26.4, LLG=433 • Buccaneer: side chain extension 315/316 • Refmac5: R1/“Rfree” = 28.0% / 29.9% (4N5P w/o water) RCa Harwell Seminars 13th April 2017 23/34 Tim Grüne Electron Crystallography at PSI Summary: Electron Diffraction of Proteins • Structures can be solved with X–ray knowledge and methods. • Kinematic approximation sufficient for high quality structures • Radiation damage major limit • Possibility: “serial” electron crystallography RCa Harwell Seminars 13th April 2017 24/34 Tim Grüne Electron Crystallography at PSI 4 - Crystallographic Lens Corrections RCa Harwell Seminars 13th April 2017 25/34 Tim Grüne Electron Crystallography at PSI Garnet Andradite 3 • The garnet Andradite, Ca3Fe2 +(SiO4)3, radiation hard • 2 grids courtesy Xiaodong Zou (Stockholm) • Space group Ia3¯d, a = 12.06314(1)Å (ICSD No. 187908) (Wikipedia) • Summed images from Garnet (200keV) • 66.8◦ rotation • good coverage of detector surface RCa Harwell Seminars 13th April 2017 26/34 Tim Grüne Electron Crystallography at PSI Spatial Correction for the Detector Surface XDS Correction Table X–coordinate • Spot positions determined through Laue Conditions ~S.~a = h ~S.~b = k ~S.~c = l +10 • Deviations between calculated and observed positions • per–pixel look-up tables for X– and Y–coordinates • Independent of Source of Error -5 RCa Harwell Seminars 13th April 2017 27/34 Tim Grüne Electron Crystallography at PSI Directly Visible Improvements Garnet Data set processed before spatial correction: BEAM DIVERGENCE: 0.16◦ REFLECTING RANGE: 0.47◦ Garnet Data set processed after spatial correction: BEAM DIVERGENCE: 0.15◦ REFLECTING RANGE: 0.28◦ RCa Harwell Seminars 13th April 2017 28/34 Tim Grüne Electron Crystallography at PSI Improved Cell Accuracy with Look–up Tables 1.