A Diffraction Simulation and Comparison C.T. Koch, J. Müller, S. Fairman, B. Haas Humboldt University of Berlin Plugin for Nionswift Department of Physics
Page 3 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Research within the Structure Research / EM group • • • • • • • • • -2. • • • δ • • δ • • • •
Page 7 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Large-Angle Rocking-Beam Electron Diffraction Individual pattern LARBED pattern (log scale) Sample: SrTiO3, not energy filtered, microscope: Zeis EM912, 120 kV LARBED data acquisition by QED Plugin to DigitalMicrograph (HREM Research)
Page 8 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 From PED => LARBED Sample: Si (110) zero-loss-filtered LARBED preserves the rocking curve information !
Page 9 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Extracting LARBED Discs from data stacks = 8 ° SrTiO3 (001) not energy-filtered 9 Each pixel within the discs is the integrated background-subtracted Bragg intensity
Page 10 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Correcting Ewald-Sphere Curvature & Separating Laue Zones Extracted Zero-Order Laue Zone 3D View Sum of all frames
Page 11 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Ab-initio Structure Factors Fitting to SrTiO 3 LARBED data Difference ab-initio Fit Experiment Feng Wang IAM potential (121 beams) Fit to exp. (121 beams) Fit to exp. (incl. unmeas. data 456 beams) unit cell F. Wang, et al. Phys. Rev. Lett. 117 (2016) 015501
Page 12 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 LARBED on a 60 nm MgO cube C.T. Koch et al. Microsc. Anal. (2012)
Page 13 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 3D Reconstruction of k-Space for Complex Structure Sample: K 2 O ∙ 7Nb 5 O 2 Reconstructed 3D k-space Volume (a=b=2.9 nm, c = 0.39 nm) from experimental LARBED data z-axis expanded x 5 Rocking curves for each spot in 3D k-space
Page 14 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 LARBED at low kV and variable camera length 3D reconstruction from LARBED on graphene @ 30 kV Graphene oxide @ 30kV MoS 2 @ 30kV MoS 2 @ 30 kV L = 36.5 – 8.5mm L = 36.5 – 11.5mm tilt = 0 ° to 9 ° in 1 °
Page 15 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Nion Swift GUI => Manual Control & Python Scripting Scripts will work both on the SEM and (hopefully soon) on Nion scope
Page 16 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 4D-STEM in the SEM Layer counting in WS 2 4D-STEM of graphene oxide @ 30kV Grain orientation in graphene Two flexible scanning options : • Beam positioning commands to SEM API (slow) • External scan generator (fast)
Page 17 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 A Diffraction Simulation Plugin Uses modules pymatgen and vtk (both installed using pip install)
Page 18 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Managing Large Data & Data Science Decentralized Independent Archives of Data for Electron Microscopy (DIADEM) 100GbE Metadata + DIADEM-Server in-band Metadata GPU-Servers Data base for new Metadata + Data Metadata new in-band Metadata Data archiving and Data Data processing evaluation software acquisition at by user microscope Metadata Data
Page 19 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Fast Bloch-wave simulations with GPUs - Update 97.25x 100 90 90.06x 80 vs. conventional method 66.19x Speedup multiplier 70 63.08x 60 50 Combination of a matrix-exponential 40 approximant & GPU hardware 30 = quick forward simulations without 20 sacrificing numerical accuracy [1] 10 0 0 500 1000 1500 2000 2500 3000 3500 4000 Number of reflections MAGMA cusolver [1] R. S. Pennington, F. Wang, and C. T. Koch, Ultramicroscopy 141 , 32 (2014).
Page 20 C.T. Koch | Humboldt University of Berlin | STEM with Advanced Detectors 4.-7.9.2018 Summary In 4DSTEM, it makes sense to vary 𝑙 𝑗𝑜,⊥ and also 𝑠 • => possibility for solving local structure => local 3 reciprocal space lattice => shape transform of nanocrystals => effect of aberrations can be compensated • 4DSTEM in the SEM ideal for mapping local - orientation & phase - strain and E-field - ideal for 2D materials • 6-axis goniometer + variable camera length transmission diffraction setup => very flexible => many aplications => mapping of beam-sensitive materials • Being able to simulate electron diffraction within Swift allows a quick comparison of exp. & theory
Recommend
More recommend
