Université du Québec Institut national de la recherche scientifique INRS

Transcription

1 McGill University, CSACS Course, March 5th 2010 Scanning Probe Microscopy Techniques Federico Rosei Canada Research Chair in Nanostructured Organic and Inorganic Materials Énergie, Matériaux et Télécommunications élécommunications, Université du Québec, Varennes (Québec)

2 Contents Intro on nano-tools Scanning Tunneling Microscopy (STM) STM imaging i of semiconductor surfaces STM imaging of metal surfaces Atomic Force Microscopy (AFM)

3 Nano tools By increasing by a factor of 10 the resolving power of the Human eye, Galileo was able to discover Jupiter s satellites The principles of physics, as far as I can see, do not speak against the possibility of maneuvering things atom by atom The problems of chemistry and biology can be greatly helped if our ability to see what we are doing, and to do things on an atomic level, is ultimately developed a development which I think cannot be avoided From: R. Feynman: There's Plenty Science in ACTION of for a World Room in EVOLUTION at the Bottom (1959)

4

5 Length scales Nano tools Université du Québec Human hair Cells 100 μm Optical Microscopy 10 μm Lithography Integrated circuits Biological Macromolecules 1 μm 100 nm 10 nm Electron Microscopy Atoms and molecules 1 nm Scanning Probe Microscopy

6 Scanning Probe Microscopy Principle of a scanning probe microscope. Surface is scanned line by line with a probe using a fine positioning system (scanner). Vibration isolation shields the microscope from external vibrations. With a coarse positioning device, the distance between the sample and the probe is reduced until the interaction regime is reached.

7 The STM principle G. Binnig and H. Rohrer, (Nobel Prize in Physics, 1986) A sharp metal tip (W, Pt Ir) is brought into close proximity of a conducting sample, and a bias is applied: electrons tunnel from tip to sample (or viceversa)

8 STM working model Animation on:

9 The STM principle Principle of a scanning tunnelling microscope. Once the gap between tip and sample is about as small as the diameter of an atom, a tunnelling current flows between a conductive cti tip and sample.

10 Operation of an STM 1,2 [1] C. Julian Chen, Introduction to Scanning Tunnelling Microscopy, Oxford (1993) [2] G.A.D. Briggs and A. J. Fisher, Surf. Sci. Rep. 33, 1 (1999)

11 Scanning Tunneling Microscopy Once the tip is in tunneling contact, it is scanned above the surface using three separate piezoelectric transducers for precise movements in x,y,z Pt(110) (1x2) with atomic resolution

12 Principle of a local probe gentle touch of a nanofinger: If the interaction between tip and sample decays sufficiently rapidly on the atomic scale, only the two atoms that are closest to each other are able to feel each other. G. Binnig, H. Röhrer, Rev. Mod. Phys. 71, S324 (1999)

13 Tunneling Current I t V t exp(-a Θ z) Θ. Workfunction, typically 3-5 ev z.. Tip-sample separation, typically 4-10 A Δ z = 1 Å --> Δ I one order of magnitude!

14 Imaging Si(111) 7x7 Si(111) clean surface (reconstructed 7x7) imaged with atomic resolution First STM work: Binnig et al., Phys. Rev. Lett. (1983) Si(111) 7x7 in real space

15 The Bottom Up approach Semiconductor quantum dot: the ultimate quantum confined structure Narrow band gap material nanostructure embedded in wide band material Unique electronic properties: δ-function like energy dependence of the density of states: energy and charge quantization quantum confinement of carriers in all 3 dimensions can be treated like an atom Needs: lateral dimensions < λ de Broglie (50 nm in GaAs) Uniformity in shape and dimensions Reliable ordered distribution direct synthesis of devices by epitaxial growth Self assembled epitaxial growth: Damage free (coherent) structures Coherent crystals Possibility of integration with microelectronic fabrication

16 Crystal Growth: Université du Québec a non equilibrium phenomenon Epitaxial thin film growth is a non equilibrium kinetic phenomenon. At thermodynamic equilibrium, all atomic processes proceed din opposite directions at equal rates (principle of detailed balance ) Adsorption/desorption from the gasphase and cluster nucleation/decay occur at equal rates => no net growth at equilibrium => average macroscopic quantities (e.g. surface coverage, roughness) stay constant For a net growth rate, one has to be away from equilibrium Competing effects: thermodynamics (surface and interface energies) versus kinetics (temperature, deposition rate)

17 Epitaxy Epitaxy comes from the greek words: επί (on top) ταξισ (to order) In this growth mode, the atoms attach to an existing crystalline surface by forming layers with the same order as the original matrix.

18 Growth Techniques Chemical Vapor Deposition Molecular Beam Epitaxy

19 Three types of Epitaxy The forces that act between the substrate and the deposited layer give rise to three different growth modes: layer-by-layer (Frank Van der Merwe), island formation on the bare substrate (Volmer Weber), layer by layer up to a critical thickness, followed by island formation (Stranski Krastanov), which is a mix of the first two.

20

21

22 Homoepitaxy Examples: Growth of Si on Si Growth of Pt on Pt (previous STM movies) There is no strain. The growth dynamics is governed by kinetic effects (e.g. substrate temperature, deposition rate)

23 Heteroepitaxy heteroepitaxy : epitaxial growth realized by depositing one or more atomic species on a substrate with different composition (e.g. Ge on Si, InAs on GaAs). Typically the two (or more) species have a different lattice parameter, and this often induces a strain in the grown film

24 Lattice parameter: x Ge = 5.65 Å x Si =543Å 5.43 Lattice mismatch Pseudomorphic Growth ε = (x Ge x Si )/x Ge = 4.2% Up to a critical thickness, it is possible to grow layer by layer (Wetting Layer: The grown material B wets the surface of A) Rule of thumb: If B wets A, then A does not wet B (the type of strain is inverted)

25 Heteroepitaxy: Ge/Si growth Université du Québec 1 ML = 0.34 nm: WL ~ 3 ML Si Wetting Layer 4.2% Ge/Si lattice mismatch: wetting layer -> 3D structures (SK growth mode) Up to 3 5 ML flat WL More than 3 5 ML strain driven roughening transition Ge/Si(111) is a model quantum dot system Islands are very large due to alloying 1 μm LEEM movie: T = 550 C, 3 to 10 ML

26 Ge/Si: Reconstruction change Si(111) reconstructs 7x7 The Ge/Si(111) Wetting Layer reconstructs 5x5 7x7 5x5

27 Ge/Si(111): island nucleation Nucleation of 3 D islands: 20 Å Ge / Si(111), T = 550 C Abrupt transition of the WL? Islands begin to form as truncated tetrahedra Gradient image 8.5 nm nm 40 Θ 0

28 Ge/Si(111): island evolution 25 Å Ge / Si(111), T = 450 C Gradient image Island height: 38 nm 43 <117> Insertion facets of Initial formation trenches new of 230 nm <111> 55 <100> 60 Θ 0

29 Ge/Si(111): different reconstructions 2000 nm 9MLGedeposition 0.1 nm/min T = 500 ºC 3 D Islands: 7x7-5x5 tall triangular (strained) (180 nm wide x 10 nm high) low rounded (ripened) (350 nm wide x 2.5 nm high) 8 nm

30 Ge/Si(111) island evolution After nucleation: the islands grow vertically up to a critical value the strain energy introduces dislocations morphological transition lateral growth material flow from the top - central hole formation nm 0.6 nm n 40 m

31 Ge/Si(111) island evolution 45 Å Ge/Si(111) T=450 C Gradient image Island height: 63 nm Insertion of dislocations nm 260 Dislocations

32 Ge/Si(111): island ripening 30 Å Ge / Si(111), T = 500 C 20 Å Ge / Si(111), T = 550 C Gradient image - Main features: Gradient image nm Ripening effect: island is rounded Substrate erosion: formation of a trench around the island. 500 nm - Full Ripening: Atoll like shape: formation of a central hole Substrate erosion

33 Growth Movies -1 STM Movies: B. Voigtlander observation of step-flow growth Homoepitaxy of Si on Si(001): Layer by Science inlayer ACTION for a World EVOLUTION growth

34 Growth Movies - 2 Homoepitaxy of Si on Si(111): Layer by Layer growth 2 D island nucleation and coalescence

35 Growth Movies - 3 heteroepitaxy of Ge on Si(001): Stranski Krastanow growth (layer by layer, then 3 D island nucleation)

36 Growth Movies - 4 heteroepitaxy of Ge on Si(111): Stranski Krastanow growth

37

38 The Herringbone Reconstruction on the Au(111) Surface 1211 x 1227 Å 2 87 x 90 Å 2

39 Au(110)-(1x2) (1x2) Missing row reconstruction

40 Cu(110) with atomic resolution Cu(110) 1x1 [1 1 0] [0 0 1] 100x100 Å 2

41 Oxygen nanopattern on Cu(110) Cu(110) O 2 chemisorption: Cu O rows (2x1) Cu O rows (2x1) 4 6 Langmuir O 2 at 625 K [001] [1-10] 10] partial 2x1 reconstruction ( patches ) [001] [1-10] 10] 65x65 Å 2 Bare Cu 65x65 Å 2 added row structure : Cu atoms : O atom

42 Coseupo Close on Cu O O 2 chemisorption on Cu(110): partial 2x1 reconstruction ti ( patches ) ) 35x35 Å 2 20x20 Å 2

43 Surface diffusion: in general, it is a 2 D random walk E D ν = ν 0 exp(-e D /kt) Hopping rate: counting the proportion of molecules that have not moved between two consecutive images: P 0 = M / N = F(ht)

44 Surface diffusion i E D hopping rate: h = h 0 exp(-e D /kt) random walk: ( x) 2 = λ 2 ht λ - RMS jump length P x T 1 T 2 tracer diffusion coefficient: D = ( x) 2 /2t x D = D 0 exp(-e D /kt) ln D D 0 E D with D 0 = h 0 λ 2 /2 1/kT

45 STM Université du Québec movies: Diffusion of Pt adatoms Diffusion of Pt adatoms on Pt(110) (1x2) (1x2) T.R. Linderoth et al., Phys. Rev. Lett. 78, 4978 (1997) More STM Movies: Science in ACTION for a World in EVOLUTION

46 STM movie: Dynamics of Pt dimers Slightly higher Pt coverage Diffusion of vacancies along the rows Formation and diffusion of Pt dimers T. Linderoth et al., Phys. Rev. B 61, R2448 (2000) Å 2 STM Movies:

47 Surface Université du Québec Diffusion of Large Molecules DC HtBDC 1 D Diffusion along the close packed direction [1-10] of Cu(110) [1,-1,0] 1,0] [001] 50 Jump length ht HtBDC/Cu(110): λ = 6.8 ± x500 Å 2 T = 235 K 500x500 Å 2 T = 194 K DC/Cu(110): λ = 3.9 ± 0.2 Pt/Pt(110): λ = 1.11 ± 0.01 Comparative diffusion of DC and HtBDC on Cu(110): M. Schunack, T. Linderoth, F. Rosei et al., Phys. Rev. Lett. 88, (2002)

48 Displacement Institut national de recherche scientifique distribution: HtBDC Temperature: 185 K Time per image: 13.5 sec Temperature: 194 K Time per image: 13.7 sec OBSER RVATION NS OBSER RVATION NS Δx [N.N. DISTANCES] Δx [N.N. DISTANCES]

49 Institut national de la Arrhenius recherche scientifique analysis: D DC K HtBDC K E D =0.71±0.05 ev E D =0.62±0.04eV D =10-1.0± cm s D =10 0.9± cm s => Arrhenius parameters are related to Science molecular in ACTION for a World in EVOLUTION structure

50 Institut national de la recherche Arrhenius scientifique analysis: h DC K HtBDC K E D=0.71±0.05 ev h 0 = ±0.7 s -1 E D=0.62±0.04eV h 0 = ±0.4 s -1 In general, E D is a fraction of the adsorption energy

51 Diffusion coupled to rotation Diffusion of DC on Cu(110), high resolution movie 200 x 200 Å 2 with high resolution we observed the rotation of single molecules, coupled to diffusion: the molecules behave like nano-disks

52 Diffusion of C 60 on Pd(110): rolling motion? J. Weckesser, J.V. Barth, K. Kern, Phys. Rev. B 64, (2001) E D = 1.4 ± 0.2 ev, h 0 = ±0.4 s -1 Sequence of STM images monitoring the thermal motions of C 60 molecules on Science Pd(110) in ACTION for a World in EVOLUTION at T = 437 K.

53 Diffusion i conclusions DC HtBDC E D (ev) 0.71 ± ± 0.04 λ (nn dist) ) 3.9 ± ± long jumps may be predominant in surface diffusion of large organic molecules 2. tailoring diffusion properties (using specially designed molecules)

54 Further Reading F. Besenbacher, STM studies of metal surfaces Rep. Prog. Phys. 59, 1737 (1996) G. Binnig et al., Atomic Force Microscope, Phys. Rev. Lett. 56, 930 (1986) G. Binnig et al., 7x7 reconstruction on Si(111) resolved in real space, Phys. Rev. Lett. 50, 120 (1983) G. Binnig, H. Rohrer, In touch with atoms, Rev. Mod. Phys. 71, S324 (1999) K. Bobrov, A.J. Mayne, G. Dujardin, Atomic - scale imaging of insulating diamond through resonant electron injection Nature 413, 616 (2001) C. Barth, M. Reichling, Imaging the atomic arrangements on the high-temperature reconstructed α-al 2 O 3 (0001) surface, Nature 414, 54 (2001)

55 Review Institut national Article de la recherche in: scientifique Progress in Surface Science 71,, (2003).

56 Atomic Force Microscope (AFM) It also works on insulators! Binnig, Quate, Gerber (1986)

57 Atomic Force Microscope

58 The first AFM G. Binning, Ch. Gerber, C.F. Quate, Phys. Rev. Lett. 56, 930 (1986)

59 Atomic Force Microscopy

60 Principle of an atomic force microscope. A sharp tip is brought close to the sample. AFM principle The forces acting between tip and sample lead to a deflection of the spring. The deflection is then measured e.g. optically

61 AFM on insulators Previous studies had indicated that AFM can attain atomic resolution on Si(111) 7x7 Structure of the α-al 2 O 3 (0001) surface in its 31x 31 R ±9 high-temperature reconstruction, as measured by dynamic scanning force Science microscopy in ACTION for a World in EVOLUTION

62 Measuring Forces

63 Forces between atoms Back of the envelope : Atomic energy scale: E bond ~ 1-4 ev ~ J bond Typical bonding length: a~0.2 nm Typical forces: F = E/a ~ 1-3 nn Bonding energies: Quantum mechanical (covalent, metallic bonds): 1 3 nn Coulomb (dipole, ionic): nn Polarization (induced dipoles): nN J. Israelachvili, Intermolecular and Surface Forces, Academic Press

64 Comparing forces Comparison of the distance dependence of a short range force with the tunneling current.

65 F vdw = AR/6z 2 vdw Van der Waals forces A: Hamaker s constant R: tip radius z: Tip sample separation A depends d on the type of material (polarizability). For most materials and vacuum, A~1 ev (Krupp, Advances Colloidal Interface Sci. 1, 113 (1967) R~100 nm is a typical effective radius => F vdw ~10 nn at z~0.5 nm

66

67 Deflection sensors

68 Laser beam deflection Interferometric cantilever detection systems as detection system, a often used in ambient popular choice for low- AFMs. temperature t AFMs.

69 Feedback modes

70 Piezoelectric scanners

71 Creating images from the feedback signal

72 Electrostatic forces

73 Imaging artifacts t

74 Thermal drift Touching the microscope (e.g. sample, cantilever) will change its temperature T. Shining light on it too! Cantilever has a mass of ~ 1 ng, and thus a VERY small heat capacity. So what!?! L/L = const T; const ~ 10-5

75 AFM on biological systems (a) The cytoplasmic surface of the hexagonally packed intermediate layer is an essential part of the cell envelope of Deinococcus radiodurans. It has a protective function and acts as a molecular sieve. The pores in protruding cores are the channels of this sieve, and exhibit two conformations that change dynamically. The unit cell size is 18 nm, and the brightness range corresponds to 3 nm. (b) Two-dimensional crystals of bacteriophage F 29 head-tail connectors recorded with AFM in buffer solution. The connectors are packed in up-and-down originations, exposing their narrow ends that connect to the tail and their wide ends that connect to the head. The unit cell size is 16.5 nm, whereas the brightness range corresponds to 4 nm.

76 Examples of writing by AFM

77 Dip-pen pen nanolithography Chad Mirkin (1999)

78 Université Dynamic du Québec (AFM) Nanostencil Lüthi et al., Appl. Phys. Lett. 75 (1999) 1314 Scheme of the experimental apparatus: - The material is deposited from source A through series of collimating apertures - Optical beam from source B reflects off a cantilever (D) - Position-sensitive sensiti e detector (C) used to regulate proximity of the cantilever tip with respect to substrate surface (E) - Series of small apertures: the cantilever defines pattern of deposited material at E

79 Université Dynamic du Québec (AFM) Nanostencil SEM images of a Si 3 N 4 cantilever featuring several holes in and near pyramidal tip - milled by focused ion beam (Ga ions, 70 pa, 30 kv). Lüthi et al., Appl. Phys. Lett. 75 (1999) 1314 AFM images of atomically clean Cu lines grown by dynamic stenciling: (a) line, (b) circle, (c) interdigitated line, and (d) lines connecting two pads (height of the lines: nm).

80 Applications: nanosensors Ways to functionalize cantilevers: - Evaporation of metal layers through shadow masks - Self assembly of monolayers - Spray coating with polymers - Coating using a microfluidic network IBM and the Nanomechanical ln Nose Ch. Gerber Science and in ACTION for a World in EVOLUTION co workers, Science 288, 216 (2000)