3D and Atomic-resolution Imaging with Coherent Electron Nanobeams - Opportunities and Challenges for X-rays David A. Muller Lena Fitting Kourkoutis, Megan Holtz, Robert Hovden, Qingyun Mao, Julia Mundy, Yingchao Yu, Huolin L. Xin, Ye Zhu BaTiO 3 /SrTiO 3 Pt-Co Fuel Cell Catalysts coalescence in 3D Integrated Circuits
Scanning Transmission Electron Microscopy 100 kev Electron beam lenses 1 atom wide (1 Å) beam is scanned across the sample to form a 2-D image Elastic Scattering ~ "Z contrast" EELS: Chemical Imaging La Ti sample Mn ADF Detector EELS: Photonic LDOS Electron Spectrometer Energy (ev) Scattering angle ( rad) Muller, Kourkoutis, Murfitt, Song, Hwang, Silcox, Dellby, Krivanek, Science 319, 1073 (2008).
2D imaging of electronic structure EuTiO 3 on DyScO 3 (Lee, Schlom, Nature 466, 954 (2010)) 2 10 0 Intensity (a.u.) 1 10 0 Reference spectra Ref1 Ref2 Eu 2+ Eu 3+ 2-component least squares fit to the background subtracted spectrum 0 1090 1100 1110 1120 1130 Energy Loss (ev) Increased Eu valence in one atomic layer at the interface
Single Atom Imaging: Catalysis and Dopants 10 nm Th on a-c Crewe et al., Science 168, 1338 (1970) U on a-c Treacy & Rice, J. Microscopy 156, 211 (1989) Pt on a-al 2 O 3 Blom et al, Microsc. Microanal. 12, 483, 2006 2 nm Sb in Si Voyles et al., Nature 416 826 (2002) K-doped C 60 peapods Guan et al., PRL 94, 045502 (2005) Graphene Grain Boundaries Huang et al, Nature 469, 389 (2011)
Hardware Advances in Microscopy Resolving Power (Å -1 ) 1 10-1 10-2 10-3 10-4 light microscope electron microscope Amici Abbe aberration-corrected electron microscope Stuttgart TEAM/ (1.2 MV) CREST Dietrich NION (200 kv) Haider (200 kv) Zach, Haider (1kV SEM) Borries and Ruska Marton, Ardenne (100 kv) Ruska (75 kv) 1 10 10 2 10 3 10 4 Feature Size (Å) Corrected optics have enabled practical Sub-Angstrom resolution Ross 1800 1850 1900 1950 2000 2050 Year Muller, Nature Materials, (2009), Adapted from Rose (2009)
Relativistic Charge as a Virtual Light Source q v/c ~ 1 b E 1 E 2 Broad-band source of photons E 1 (t) STEM offers sub-nanometer spatial resolution γ v / b b /γ v E 1 (ω) t log(ω)
Scanning Transmission Electron Microscopy 100 kev Electron beam sample lenses 1 atom wide (1 Å) beam is scanned across the sample to form a 2-D image Intensity (arb. units) 10 8 10 7 10 6 10 5 Incident Beam ADF Signal Valence Excitations 3 Å Si L edge O-K edge Er M 4 Edge ADF Detector 10 4 Elastic Scattering ~ "Z contrast" Annular Dark Field (ADF) detector Science 319, 1073 (2008) 0 0.5 1 1. Distance (nm) 0 100 200 300 400 500 600 700 Energy Loss (ev) Electron Spectrometer Single atom Sensitivity: ADF: P. Voyles, D. Muller, J. Grazul, P. Citrin, H. Gossmann, Nature 416 826 (2002) EELS: U. Kaiser, D. Muller, J. Grazul, M. Kawasaki, Nature Materials, 1 102 (2002)
How Bad is Radiation Damage? R. Henderson, Quarterly Reviews of Biophysics 28 (1995) 171-193. It s not the cross-section, but How many damaging events per useful imaging event? Least Damage: Elastic imaging - Electrons wins Inelastic imaging - Soft X-rays win
Radiation Damage as a Fundamental limit More Elastic information per damaging event Data from Breedlove and Trammell, Science 170 (1970) 1310-1313 For electrons σ i / σ e ~ln (E)
Dose Required for 2D-Imaging Dose (e - /nm 2 ) 10 5 10 4 1000 100 10 C r =0.1 C r =0.3 n 0.45 nm 1.3 nm 0 k fc 2 2 r r 2 Bio-sections Resolution 1/(Dose) 1/2 Resolution 1/(Contrast) 1 0.1 1 10 Resolution (nm) It s almost impossible to do atomic-resolution phase contrast imaging with biological samples (except by averaging over many similar molecules)!
Dose Required for 3-D Reconstructions is high B. F. McEwen et al, Journal of Structural Biology 138 47 57 (2002) Saxberg & Saxton, W.O., Ultramicroscopy 6, 85 90 (1981) M.R. Howells et al, J. Electron Spec. Rel. Phenom. 170 4 (2009) 1% 40% 80% SiO 2 r 1/(Dose) 1/4 r 1/(Contrast) 1/2 10 kev x-rays Resolution (nm) Dose-limited resolution for oxides is ~0.2-2 nm and ~3-10 nm for polymers
High Resolution= Thin Sections Small features have low contrast (and for a fixed dose we trade 2D resolution for contrast) Resolution Sample Thickness Need to make thin samples (true for x-rays as well as electrons) (unless we have a fluorescence detection method & there is only 1 object)
Site-Specific Focused Ion Beam high-k : Xeon X3220, 2.5 GHz, Quad Core Thin-section Mounted for TEM Liftout
Ultramicroscopy, 107 (2007) p8
Electron Tomography Experiment Walter Hoppe, Angew. Chem. Int. Ed. Engl. 22 (1983) 456-485 Field of View : 127 nm 3D resolution function along X, dx ~ 0.2-1 nm along Y, dy ~ >1 nm along Z, dz ~ >1 nm (due to limited tilt range and finite number of projection images)
Three-Dimensional Imaging at the Nanoscale: How Metal Contacts Form on a Carbon Nanotube 2 nm Nano Letters 7 3770-3773 (2007)
[101] Discrete Tomography at Atomic Resolution? Discretize a small set of atomic-resolution zone-axis projections [100] Van Aert et al, Nature 470 (2011) 374
Discrete Tomography at Atomic Resolution? Discretize a small set of atomic-resolution zone-axis projections -60 [1123] -60-30 0 30 60 [2110] [1010] [1120] [0110] [1210] 60 [1123] Higher resolution More projections See also Jinschek et al, Ultramicroscopy 108 (2008) 589
Discrete Tomography at Atomic Resolution? Digitize a small set of atomic-resolution zone-axis projections Er in Si Uniqueness?
Atom-Probe Tomography: 3D Composition at Sub-nanometer Resolution (~0.5 nm resolution, 50-60% of atoms detected) CoFe/Cu/CoFe Magnetic Multilayer Fe Co Cu Co Ni X.W. Zhou et al, Acta Mater. 49, 4005 (2001)
TEM in Liquids Peckys et al, "Fully Hydrated Yeast Cells Imaged with Electron Microscopy". Biophysical Journal 100, 2522 (2011). STEM in liquid Optical Fluorescence 2 μm
Electrons go a Long Way E 0 =200 kev E 0 =20 kev Electron Range (in m): R 0.064 E 1.5 0.3 0 (density in g/cm 3, E 0 in kev) 1 m of Carbon R~ 150 m at 200 kev in silicon
Beam Spreading (in Water) Beam Spreading b Z E t3/2 Electron s Home court advantage 10 μm 0.1 μm 1 μm 1 MeV 10 MeV 200 kev By changing the collection angle, can trade off a linear improvement in resolution (~ x 2) for an exponential drop in signal Appl. Phys. Lett. 88 243116 (2006)
Why not build an electron beamline? (you have all the pieces already ) Resolution Limits Electron optics: Higher brightness, Ultrafast, 1 nm spot @ 1 μm Radiation Damage limits the imaging of UV-sensitive materials (~10x better for electrons) Tomography Quantitative measurements of 3D objects -1-3 nm resolution at 0.1-1 m thick - 10 nm for 10-20 microns @ MeV -Atomic resolution may be possible in inorganics EELS Imaging (not time-resolved) 2D Chemical information at the atomic scale 3D may be possible at the nanometer scale Plasmon 24Map of Si in SiO 2
Outlook Nanoscale Photonics V Liquid cell Graphene AtomicallyEngineered Oxides