the Scanning Tunnelling Microscope Rubn Prez Departamento de Fsica - - PowerPoint PPT Presentation

the scanning tunnelling
SMART_READER_LITE
LIVE PREVIEW

the Scanning Tunnelling Microscope Rubn Prez Departamento de Fsica - - PowerPoint PPT Presentation

Theoretical Modelling and the Scanning Tunnelling Microscope Rubn Prez Departamento de Fsica Terica de la Materia Condensada Universidad Autnoma de Madrid Curso Introduccin a la Nanotecnologa Mster de fsica de la


slide-1
SLIDE 1

Theoretical Modelling and the Scanning Tunnelling Microscope

Rubén Pérez Departamento de Física Teórica de la Materia Condensada Universidad Autónoma de Madrid Curso “Introducción a la Nanotecnología” Máster de física de la materia condensada y nanotecnología

slide-2
SLIDE 2

Theoretical modelling of SPM

  • 1. STM: contrast mechanisms

(lecture, 20/01/13)

  • 2. AFM: Resolution?

(lecture, 30/01/13)

  • 3. New developments

(discussion based on 4 recent papers, 03/02/13)

  • 4. STM simulations

(hands-on, 06/02/13)

  • 5. AFM simulations

(hands-on, 10/02/13)

slide-3
SLIDE 3

Understanding the STM contrast : GaAs (110)

As Ga

  • Only one atomic specie

imaged for each voltage?

  • Shift between the

position of the maxima? STM Experiments at different polarities

slide-4
SLIDE 4

References

  • JM. Blanco, F. Flores and R. Perez. Progress in Surface

Science 81, 403-443 (2006).

  • W. A. Hofer. Progress in Surface Science 71, 147-183

(2003).

  • C. J. Chen. “Introduction to Scanning Tunneling

Microscopy”. 2nd Edition. (Oxford University Press, Oxford, 2008).

  • R. Wiesendanger. “Scanning Probe Microscopy &

Spectroscopy”. (Cambridge University Press, Cambridge, 1994).

  • D. Bonell, Editor. “Scanning Probe Microscopy &

Spectroscopy”. 2nd Edition. (Wiley-VCH, New York, 2001).

slide-5
SLIDE 5

Principle of operation

“It soon become apparent that it was one thing to obtain an image and quite another to understand the structure that was seen” G.A.D. Briggs and A.J. Fisher, Surf. Sci. Rep. 33 (1999) 1-81

  • Atomic protrusions on the tip are usually

random, and with luck one atom may protrude sufficiently to dominate the tunneling geometry.

  • Atomic resolution: Tunnelling probability

changes an order of magnitude for every angstrom change.

  • Contrast: combined effects of

topography and electronic structure.

slide-6
SLIDE 6

The problem we are facing…

TIP SURFACE d: Tip-surface distance

STM implies describing tip, sample + tunnelling process. Applying V  system out of equilibrium Most theoretical tools for systems in equilibrium,so... Tip-surface distance ~ 5-10 Å  exchange- correlation and image potential effects are important ( are well described by DFT??) Conventional approaches: Sample description is usually good, while transport and tip are treated with very rough approximations... (perturbative, s-wave for the tip, no image effects)  qualitative description, but can we make it quantitative...? Non-perturbative approaches for tunneling + first-principles description of the electronic properties of tip and sample

slide-7
SLIDE 7

1) Different STM approaches:

Perturbative method: Bardeen, Tersoff-Haman, Chen

Tersoff – Haman approximation (T-H) Chen’s improvement to T-H Bardeen: Transfer Hamiltonian + Bardeen tunnelling current

Non-perturbative approaches to transport:

2) Combining STM and theoretical modelling: Examples

OUTLINE

3) Recent developments & Challenges: (tip-sample interaction, electric field, spin-polarized STM)

Scattering matrix Keldysh-Green function formalism Landauer formalism (only elastic contributions)

slide-8
SLIDE 8

Perturbative Methods: Bardeen , Tersoff-Haman and Chen’s approach

slide-9
SLIDE 9

Transfer Hamiltonian + TUNNELLING CURRENT

k k k S U k         2 2 / 1

' ' ' ' 2 2 / 1 k k k U k

T

       

Uncoupled system Coupled system:

S T

U U H     

2

2 / 1 ˆ

(J. Bardeen, PRL 6 (1961) 57)

slide-10
SLIDE 10

Transfer Hamiltonian + TUNNELLING CURRENT

Bardeen showed that under certain assumptions,

) ( 2

' ' ' k k k k S kk

S d m T        

   

(Tkk’tunnelling matrix element between k and  k’)

Current (1st order perturbation theory)

 

' 2 . , ' . , ' ) ' (k

/ 2

k k empt k

  • cc

k kk k

T e I      

Empty states Occupied states

Energy V

k k k S U k         2 2 / 1

' ' ' ' 2 2 / 1 k k k U k

T

       

Uncoupled system Coupled system:

S T

U U H     

2

2 / 1 ˆ

(J. Bardeen, PRL 6 (1961) 57)

slide-11
SLIDE 11

TERSOFF-HAMAN APPROXIMATION: S k

WF

Ideal tip, with an s-like orbital in the apex

) (

' tip k kk

r T   

) , ( ) (

2 2 ' k tip sample tip k kk

r r T       

) , (

Fermi tip sample tunnel

r I    

V0

) , (

k tip sample tunnel

r dV dI    

  

    d f f r I

S T tip sample tunnel

  

  ) ( ) ( ) , (

slide-12
SLIDE 12

TERSOFF-HAMAN APPROXIMATION: Tip Sample D.O.S. Sample

height ~ 5 – 7 Å

D.O.S. near the Fermi level controls the current STM images are not topographic.

slide-13
SLIDE 13

Atomic resolution on the Si(111)-7x7

Calculated charge distribution on the states (“dangling bonds”) localized on adatom and rest atom

faulted half unfaulted half 12 adatoms 6 rest atoms corner hole dimers

6-7Å 100 Å

slide-14
SLIDE 14

Understanding the bias dependence: GaAs (110)

As Ga

  • Only one atomic specie

imaged for each voltage?

  • Shift between the

position of the maxima? STM Experiments

slide-15
SLIDE 15
  • 2

2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

E

(E)

As Ga

tip sample VT -VS >0 VS -VT >0 As Ga

GaAs (110): Understanding the bias dependece

slide-16
SLIDE 16
  • 4
  • 2

2 4

  • 4
  • 3
  • 2
  • 1

1 2 3 4

14.95 14.97 14.99 15.01 15.03 15.05 15.07 15.09 15.11 15.13 15.15 15.17 15.19 15.21 15.23 15.25 15.27 15.29 15.31 15.33 15.35 15.37 15.39 15.41 15.43 15.45 15.47 15.49 15.51 15.53 15.55

X Y

  • 4
  • 2

2 4

  • 4
  • 3
  • 2
  • 1

1 2 3 4

1.44E1 1.442E1 1.445E1 1.447E1 1.449E1 1.452E1 1.454E1 1.456E1 1.459E1 1.461E1 1.463E1 1.466E1 1.468E1 1.47E1 1.473E1 1.475E1 1.477E1 1.48E1 1.482E1 1.484E1 1.487E1 1.489E1 1.491E1 1.494E1 1.496E1 1.498E1 1.501E1 1.503E1 1.505E1 1.508E1 1.51E1

X

Y

VT-VS > 0 VS-VT > 0 GaAs (110): Theoretical STM images

 

slide-17
SLIDE 17

A typical application of Tersoff-Hamann Approach

(S-H. Lee et al, PRL 85 (2000) 3890) Novel surface geometry for GaAs(100) under low As pressure First principles calculations of sample + T-H approach for tunneling

slide-18
SLIDE 18

CHEN’s IMPROVEMENT TO TERSOFF-HAMAN

Directional p or d-like orbitals at the tip apex needed

i tip k kk

dx r d T ) (

'

  

For a p-like orbital For a d-like orbital

dxj dx r d T

i tip k kk

) (

2 '

   ) ( 3 2 ) (

2 2 ' tip k tip k kk

r W dz r d T      

T-H reproduces qualitatively large period surface reconstructions + adsorbates on metals

But CANNOT reproduce:

Lateral atomic resolution in closed-packed metal surfaces Inverted contrast images Large atomic corrugations (C.J. Chen, PRL 65 (1990) 448 ; PRB 42 (1990) 8841; PRL 69 (1992) 1656)

slide-19
SLIDE 19

PROBLEMS WITH B — T-H — CHEN APPROACH:

5) T-H: Neglects the dependence on the tip structure 4) T-H: At typical tip-sample distances, sample(rtip , ) can’t be used. 1) Gives just the 1st order perturbation term

Atomic oxygen on Pd(111) imaged with two diffent tips (M. Salmeron group)

3) Long T-S distances : Tkk’ smaller than actual values due to long range atomic potentials. 2) Small T-S distances: Tkk’ don’t include the effect of tip-sample chemical interaction. 6) Chen: Not easy to combine different tip-orbital symmetries to get real image.

slide-20
SLIDE 20

Approaches based on Bardeen’s tunneling currents and First Principles calculations

  • O. Paz et al PRL 94, 056103 (2005)

Propagating the sample wfn’s with the vaccuum Green’s function G

  • W. Hofer & J. Redinger Surf. Sci.

447, 51 (2000)

  • W. Hofer et al RMP 75, 1287 (2003)

FLAPW calculations for isolated tip & sample + Numerical evaluation of the Bardeen integral over a plane located at the medium distance

slide-21
SLIDE 21

Non-perturbative approaches to electronic transport: Calculating the STM current

slide-22
SLIDE 22

MULTIPLE SCATTERING formalism

1) Electron tunnelling viewed as a scattering process. 2) Tunnel gap treated as a 2-dimensional defect. 3) Scattering matrix contains the probability amplitudes for conduction electrons. 2D defect Semi-infinite solid Semi-infinite solid

Incident Transmitted Reflected (P. Sautet, Chem. Rev. 97 (1997) 1097; SS 374 (1997) 406; J. Cerdá et al, PRB 56 (1997) 15885)

slide-23
SLIDE 23

MULTIPLE SCATTERING formalism

Calculating Smm’(E)?

  • ESQC: transfer matrix tech. (both sides

have to be identical, only zero bias).

  • Surface Green’s function matching (finite

bias, more robust computationally)

slide-24
SLIDE 24

MULTIPLE SCATTERING formalism: Applications

A theoretical approach to adsorbate identification B C N O

(P. Sautet, SS 374 (1997) 406)

slide-25
SLIDE 25

KELDYSH-GREEN’S FUNCTION METHOD

  • Non-equilibrium diagramatic technique.
  • Formalism equivalent to the other non-perturbative approaches in the elastic

tunneling and for the limit of zero bias

  • BUT also appropriate for finite bias, inelastic and correlation effects!!!.
  • Naturally formulated in a local orbital basis (using atomic-like orbitals).
  • It can be efficiently linked with first-principles local basis codes (Fireball,

Siesta) to calculate the effective hamiltonians (H = Hsample + Htip + Hcoupling).

 

 

     

 c c T n HT ˆ ˆ ˆ ˆ

 

 

 

     i i i i i P

c c T c c T H ˆ ˆ ˆ ˆ ˆ

 

 

i j i ij i i S

c c T n H ˆ ˆ ˆ ˆ 

slide-26
SLIDE 26

1) Lowest order in Ti ss(), tt() = D.O.S. matrices for the sample and the tip. Tst, Tts = hopping matrices between the tip and the sample. fS(),fT() = Fermi distribution functions for the sample and the tip.

 

    

     d f f T T Trace e I

S T TT ST SS TS tunnel

 

  

) ( ˆ ˆ ˆ ˆ / 4 

KELDYSH-GREEN’S FUNCTION METHOD

(N. Mingo et al, PRB 54 (1996) 2225; L. Jurczyszyn et al, SS 482 (2001)1350)

slide-27
SLIDE 27

2) Exact solution to all orders in the tip-sample hoppings

 

    

     d f f T T Trace e I

S T TT eff ST SS eff TS tunnel

 

  

) ( ˆ ˆ ˆ ˆ / 4 

TTS

) (

R TT

G

TTS TST TTS TST TTS

) (

R SS

G ) (

R SS

G ) (

R TT

G

TTS TST TTS

) (

R SS

G ) (

R TT

G

+ + + ...

...... ˆ ˆ ˆ ˆ ˆ ˆ   

TS R TT ST R SS TS TS

T G T G T T

 

TS TS R TT ST R SS eff TS

T T G T G T ˆ ˆ ˆ ˆ ˆ 1 ˆ

1 

 

KELDYSH-GREEN’S FUNCTION METHOD

(N. Mingo et al, PRB 54 (1996) 2225; L. Jurczyszyn et al, SS 482 (2001)1350)

slide-28
SLIDE 28

2) Combining STM and theoretical modelling:

 STM currents: from tunneling to the contact regime  Conflicting images: Role of tip and imaging conditions

slide-29
SLIDE 29

From tunneling to the contact regime…

J.M. Blanco et al, PRB 70, 085405 (2004) Al tip on Al(111)

slide-30
SLIDE 30

From tunneling to the contact regime…

J.M. Blanco et al, PRB 70, 085405 (2004) Al tip on Al(111)

Corrugation: combined effects of atomic relaxation & current saturation !!! Corrugation Conductance Keldysh- Green´s function formalism

slide-31
SLIDE 31

Comparison of different transport formalisms

)] ( ˆ ) ( ˆ ˆ ) ( ˆ ) ( ˆ ˆ [ Im 4

2 F a TT F TT ST F r SS F SS TS

E D E T E D E T Tr e dV dI G       )] ( ˆ ) ( ˆ [ Im 2

2 F F

E t E t Tr h e dV dI G

  Landauer formalism Keldish-Green’s function formalism )] ( ˆ ) ( ˆ [ Im 4 2

2 2 F F

E t E t Tr h e dV dI G

   ) ( ˆ ˆ ) ( ˆ ) ( ˆ ˆ

2 / 1 2 / 1 F TT ST F r SS F SS

E T E D E t   

slide-32
SLIDE 32

Directly related with catalysis of gases over surfaces. Discrepancies between different experimental images of O/Pd(111). Motivation

Oxigen with circular shape. Oxigen with triangular shape Inverse contrast V = -1.4

  • J. Méndez et al., Berlin

V ~ 0

  • M. Salmeron et al., Berkeley

V = 0.29 V

  • J. Méndez et al., Berlin

...that cannot be explained with approximations like Tersoff-Hamann.

J.M.Blanco et al., PRB 71, 113402 (2005)

Conflicting experiments: O(2x2) / Pd(111)

slide-33
SLIDE 33

Theoretical simulations

W tip d=6.46 (top) V=-1.4 V Pt tip: d=6.46 (top) V=-0.030 V 6.30 6.35 6.40 6.45 top O hcp top

Å

O hcp 6.20 6.28 6.36 6.44 top

Å

top

slide-34
SLIDE 34

¿D ¿Dif iffer erent compos ent composit ition ion or dif

  • r differ

erent ent volta

  • ltage?

ge?

Pt tip: W tip:

  • Corrug. for different tips:
  • Corrug. for different geometries:

Analysis in tip orbitals

  • 4.E-07
  • 2.E-07

0.E+00 2.E-07 4.E-07 6.E-07 8.E-07 1.E-06 1.E-06 1.E-06

  • 5
  • 4
  • 3
  • 2
  • 1

1 2 3 4 5

Posición (Å) Conductancia (A/V)

Total Sólo ápex Capa 3Pt

  • Cruz. ápex-3Pt

pz+s-pz

6.30 6.35 6.40 6.45 6.50

  • 5
  • 4
  • 3
  • 2
  • 1

1 2 3 4 5

Posición (Å) Distancia punta-muestra (Å)

  • Conduct. (3010 nA / V)

V=-1.0 V ( 254 nA) V=+1.0 V (432 nA)

6.30 6.35 6.40 6.45 6.50

  • 5
  • 4
  • 3
  • 2
  • 1

1 2 3 4 5

Posición (Å) Distancia punta-muestra (Å)

  • Conduct. (1600 nA / V )

V= -1.0 V ( 325 nA) V= +1.0 V ( 472 nA)

slide-35
SLIDE 35

Con

  • ntr

trast ast in inver ersi sion

  • n:

Sometimes, images have inverted contrast: Oxigen contaminated tip :

slide-36
SLIDE 36

CONCLUSIONS

3) Keldysh-Green´s function formalism, exact to all orders of the coupling, provides a clear picture of the physics involved and offers great flexibility in its application. (+ it can handle inelastic and correlation effects) 1) A proper theoretical treatment of the STM has to describe not only the properties of the tip and the sample, but also should include a good description of the tunnelling process. 2) Perturbative methods of calculating STM images (in particular Tersoff- Hamman approx.) can in some cases give an approximate qualitative behavior, but don’t take into account effects that can be crucial to understand the experiments.

 

    

     d f f T T Trace e I

S T TT eff ST SS eff TS tunnel

 

  

) ( ˆ ˆ ˆ ˆ / 4 

4) More effort is needed in the characterization (both experimental and theoretically) of the tips and their properties.