Id01 diffraction imaging
Transcription
Id01 diffraction imaging
ESRF IMAGING SEMINAR SERIES Microscopy with and without microscopes: X-ray imaging under diffraction conditions TOBIAS SCHÜLLI ESRF seminar April 28th 2015 Page 1 INTENTION OF THIS SEMINAR To be interesting Show the motivation and the wider context & origin of our methods Be comprehensive for everybody Abbé limit of resolution in microscopy 1.22*λ ∆x = 2*n*sinα Avoid equations sinα 0.66 ∆x ≥ λ n Heisenberg uncertainty Page 2 ∆x∆p ≥ h OUTLINE Imaging with light and microscopes Diffraction (and) imaging techniques, review of the potential of x-rays Short examples on what can be done with diffraction imaging What exactly is “coherence” and why do we need it ? 3 LIGHT MICROSCOPY (SINCE 16TH CENTURY) Refraction used to produce lenses defocusing lens: Focusing lens: Makes an image “short sighted person” Page 4 At equal lens diameter a short sighted person can look with a larger collection angle at objects. MAGINFICATION (THE VERY INTEREST OF MICROSCOPY) Magnification depends on : distance of object from lens and the lens curvature, The magnification depends NOT on the diameter of the lens. Be careful with “1000-times magnification” Microscopes Magnification is not in itself a figure of merit Page 5 MAGNIFICATION AND RESOLUTION Be careful with “1000-times magnification” Microscopes Bigger lens Very big lens Magnification: geometrical optics (no reasonable limits, everything is allowed) Resolution (=∆ ∆x): real information: limited (at least) by quantum mechanics Page 6 MICROSCOPY, RESOLUTION AND WAVELENGTH Abbé limit for resolution (1870): ∆x is the smallest object that can be resolved ∆x= 1.22 * λ λ ≈ 0.66 2 * n* sin α sin α at best ∆x ~ 0.6..1.0*λ n: refractive index (1 in air) Ernst Abbé (1840 (Eisenach) -1905 (Jena) ), Physicist and Mathematician, 1863: Lecturer in Jena 1866: Hired as Research Director by Carl Zeiss Optics (Jena) 1870: Formulation of Abbé’s law 1871: Marriage with Else Snell (Karl Snell’s daughter) 1874: Co-owner of Zeiss Optics 1884: Co-founder (with C.&R. Zeiss and Otto Schott) of Schott company 1889 (+ C. Zeiss): Sole owner of Zeiss Optics 1890-1900: Zeiss foundation: company becomes independent of personal ownerships’ interests. 2α High resolution means small wavelengths and large apertures (large collection angles) Page 7 RESOLUTION, X-RAY WAVELENGTH AND LENSES 2α Abbé limit for resolution: ∆x ≈ 0.66 Light wavelength λ~400…800 nm X-ray wavelength (typical) λ=0.01…0.1 nm X-rays have a 10 000 times smaller wavelength than light To produce lenses for x-rays we need refraction (as for light), refraction depends on the polarizability of a material X-rays have a way too high frequency for electron clouds to follow , -> they polarize materials only very little (almost no refraction) Page 8 λ sin α USING X-RAYS WITHOUT OPTICS 1. Shadow play: Resolution depends (roughly) on source size 2. X-ray (light) scattering (no images but averaged information) Moon Halo: refraction by ice particles. Average shape (angles between facets) can be extracted. Page 9 (BRAGG) DIFFRACTION OF X-RAYS Constructive interference of waves reflected by atomic planes spaced by d depends on angle of incidence θ and wavelength λ: Bragg’s law Page 10 sinθ = n λ 2d , n =1,2,3,... DIFFRACTION (AND) IMAGING TECHNIQUES Radiography vs. Diffraction Imaging: full field technique with spatial resolution ~sub mm (traditional sources) Page 11 Diffraction : spatial resolution limited in any case and traded in for angular resolution DIFFRACTION AND SCATTERING: ADVANTAGES Objects can be far away (leaves a lot of space around the sample) Angular resolution obtained by diffraction leads to spatial information below λ -> “interferometric” technique(~0.0001 nm for Bragg diffraction in crystals) Limits: requires spatially homogeneous samples Position 1: interatomic distance a Position 2: interatomic distance b In many interesting systems, heterogeneity happens to be on the “mesoscale” (not atomic scale). Combining small x-ray beams with diffraction Page 12 DEVELOPMENT OF X-RAY OPTICS Mirrors: P. Kirkpatrick (1894-1992) A. Baez (1912-2007) Refractive lenses: small radii of curvature+ use many of them, Excellent imaging devices Diffractive optics: Fresnel zone plates Excellent focusing devices Some geometric drawbacks in imaging Today, focusing down to 100 nm (or less) is possible Page 13 COMBINATION OF LIMITING OPTICS AND DIFFRACTION Abbé limit for resolution: ∆x ≈ 0.66 λ sin α 2α Collection angle 2α Wavelength λ Light ~90° 500 nm X-rays ~0.1° 0.05 nm X-rays 1000 times worse X-rays 10 000 better Disappointing.. It looks like we can only win a factor of 10. BUT Combining the “good” spatial resolution of ~100 nm with diffraction (small fractions of atomic distances resolved) offers a great potential 14 DIFFRACTION IMAGING: SCANNING PROBE Use of focused beam/ scanning technique. Resolution limited by beam spot Sub 100 nm are possible 15 Detector Diffracted signal DIFFRACTION IMAGING: X-RAYS SEE MORE THAN LIGHT Light interference microscopy Scanning x-ray diffraction microscopy After growth After polishing Zöllner, Richard, Chahine: Appl. Mat.&Interf. In press Page 16 Surface roughness Lattice undulations SCANNING DIFFRACTION TECHNIQUES 0.1mmx0.1 mm ESRF logo written in a Si crystal: imaging of lattice strain and tilt: Relative strain levels of ∆a/a = few 10-6 can trace a landscape: We can “see” a ∆T of a few °C potentially in buried systems (working devices), G. A. Chahine, M.-I. Richard, R. A. Homs-Regojo et al., J. Appl. Cryst. (2014) Page 17 17 DIFFRACTION IMAGING: FULL FIELD Detector Full field technique Resolution ultimately limited by numerical aperture of the Imaging optics Sub 100 nm within reach Diffracted signal 18 T. U. Schulli DIFFRACTION IMAGING: FULL FIELD (DARK FIELD IMAGING) Similar information as before but “in one shot”(no scanning across the sample) Ideal full-field application: J. Hilhorst et al. J. Appl. Cryst. (2014). Processes (growth, annealing, …. Page 19 FROM FULL FIELD TO COHERENT DIFFRACTION IMAGING (CDI) Detector Full field technique Resolution ultimately limited by numerical aperture of the imaging optics Sub 100 nm within reach 20 T. U. Schulli COHERENT DIFFRACTION IMAGING (CDI) Coherent Diffraction Imaging: Detector Theory of optics is very well known. -> Measure all emitted rays from the sample and replace the lens by a computer to calculate the image What we can do The physics of resolution remains the same ∆x= 1.22*λ 2*n*sinα 2α Instead of the lens we need a detector with a large opening angle. We need perfect detectors: A noisy detector is like a sandblasted lens. And we need single photon detection at near 100% efficiency 21 T. U. Schulli COHERENT DIFFRACTION IMAGING (CDI) Another Problem : Waves have amplitudes and phases: they interfere to form an image; Refraction (as in a real lens) preserves the phase information, crucial for the image Detection measures only the intensity (number of photons) and not their phase From a quantum mechanical point of view, refraction (preserves ∆p) by a lens cannot be replaced by detection (destroys ∆p): Equivalence between Abbé and Heisenberg. Detector The loss of phase information cannot be recovered by a computer. We have thus to know the phase beforehand. The sample has to be illuminated with photons that are all in phase with each other This is the definition of a coherent beam 22 PHASE OF PHOTONS AND COHERENT BEAMS Marathon: photons=runners Phase depends on the exact departure time of the runners We have to select one single phase. The rest of the runners cannot be used for the experiment. We select the “coherent fraction” (Runners that all have roughly the same departure time). ESRF coherent fraction: <1% Flux available for “normal” light imaging and coherent diffraction x-ray imaging: 5 Watt LED: 1019 photons/second (incoherent but with optics we can use them all) ESRF coherent flux: 1011 photons/second (@ 8keV ) -> 1019 photons in 1 year X-ray tube coh. flux: few photons/second,1019 photons in 1010 years (the age of this world) Page 23 COHERENT DIFFRACTION IMAGING: ANTIPHASE DOMAINS IN GAN WIRES GaN: semiconductor for optoelectronic and high power applications. Problem: crystal quality N-polarity Ga-polarity 2 µm SEM image from J. Eymery (CEA-Grenoble) GaN nanowires: M. I. Richard Page 24 SUMMARY AND OUTLOOK -Diffraction imaging allows combinations of x-rays diffraction with the limiting performances of optics to obtain information in two length scale regimes. -Lensless imaging using coherent beams can push the resolution below 10 nm. This technique can currently only use a small fraction of the beam, the coherent fraction. -After phase II this fraction will increase by a factor of 40 and be of the order of 10 % for low x-ray energies (<10 keV). -Diffraction imaging techniques are nowadays exploited at ID01, ID16, ID13, ID11, ID10, ID19, ID03,… and likely by even more BLs after phase II. Thank you Page 25