Introductory lectures on

Angle-resolved photoemission spectroscopy (ARPES)

and its application to the experimental study of the electronic structure of solids
Andrs Felipe Santander-Syro Universit Paris-Sud and Ecole Suprieure de Physique et Chimie Industrielles - Paris

Resources
BOOKS S. Hfner. Photoelectron Spectroscopy Principles and Applications, third edition, Springer (Berlin), 2003. REVIEW ARTICLES F. Reinert and S. Hfner, New Journal of Physics 7, 97 (2005). A. Damascelli, Z.-X. Shen, S. Hussain, Rev. Mod. Phys. 75, 473 (2003). J. C. Campuzano, M. R. Norman, M. Randeria, cond-mat/0209476. J. Braun. The theory of angle-resolved ultraviolet photoemission and its application to ordered materials. Rep. Prog. Phys. 59, 1267-1338 (1996). INTERNET www-bl7.lbl.gov/BL7/who/eli/SRSchoolER.pdf (by Eli Rotenberg, Advanced Light Source, Berkeley) www.physics.ubc.ca/~quantmat/ARPES/PRESENTATIONS/ Lectures/Exciting2003.pdf (by Andrea Damascelli, UBC) ftp://ftp.espci.fr/shadow/bontemps/cargese2005.zip (ARPES lectures by Ralph Claessen, Augsburg)

Condensed (crystalline) matter in a nutshell

Allowed electronic states Repeated-zone scheme
-kF kF Many properties of solids are determined by electrons within a narrow energy slice (~kBT) around EF (dc conductivity, magnetism, superconductivity) Fermi Surface

3D

2D 1D
Adapted from A. Damascellis Exciting-2003 lecture and E. Rotenbergs lecture

(at room temperature, kBT = 25 meV)

ARPES: energy and momentum conservation

Kinetic energy analyzer

v h , A

z
eDetector

v h , A
Ekin, W, ,

Fixed during experiment Measured

Conservation laws

Sample

Ekin = h W E B
k vacuum = k solid || ||
hk vacuum = 2mEkin sin ||

In the solid

EB k ||

To make it work
Pull the electron out of its bound state
W ~ 2 5 eV E B ~ 0 1 eV (Valence band) E B 1500 eV (interesting core levels)
0 < Ekin = h W E B
Fi xe d

h ~ 10 2000 eV

The electron has to make its way up to the samples surface

esc = electron escape depth (Ekin~ 10 2000 eV) Experimentally: [10 2000 eV] 10 50

PES is a surface technique: one needs clean surfaces and work under ultrahigh vacuum ARPES needs, furthermore, atomically-flat surfaces (for ideal conservation of surface-parallel momentum): prepare surfaces in-situ, cleave,

Furthermore, one has to make sure that the photoemission process itself does not modify the electronic structure of the material
SUDDEN APPROXIMATION The ejected electron should be fast enough to neglect its interaction with the hole left behind At h 25 eV 2

e- - e- interaction

ee

Ekin ~ 20 eV

e- time of escape

2 Ekin esc / m
1 esc P >> m 2 2
2

esc ~ 10
For cuprates

<< ee Ekin

/ ee 0.1

h P ~ 1 eV

When all of this works

ARPES gives direct access to the single-particle electronic structure of a crystal: Band structure Spectral line-shapes and widths: electron scattering rate Interactions

a = CuO = 3.82
EB [eV]
M

X
0.6

kx

3 /a

ky
h = 22 eV

Bi2Sr2CaCu2O8+

Instrumentation and implementation

Kinetic energy analyzer Light source

v h , A

z
eDetector

Sample

Samples surface preparation Sample moving, rotating, cooling

Radiation sources
Laboratory sources Gas discharge lamps (h ~ 20-50 eV) X-ray tubes (h ~ 1500 eV) Synchrotron radiation Tunable (h ~ 10 eV 10 keV) Brilliant Polarized (linear and circular) Temporal structure (time-resolved experiments) Laser IR laser + 2*(frequency doubling): h ~ 6 7 eV (Ekin ~ 1-2 eV) esc ~ 50 100 Sudden approximation ?! Probes reciprocal space only near -point Under development: UV and soft X-ray laser (IR laser + highharmonic generation inside rare-earth gas)

Electron energy analyzer

Adapted from A. Damascellis Exciting-2003 lecture

Interaction effects on ARPES spectra

I (k , ) = I 0 (k , , A ) f ( )A(k , )

A(k,) = Probability of adding or removing one electron at (k,)

(k, ) A(k, ) = [ k (k, )]2 + [(k, )]2 1

E B h , h = 1
Many-body physics
k()
I(k,)

Energy renormalization Lifetime of dressed e-

k
EF

(k,) (renormalization) (k,) (life-time)

Binding energy EB

EF

Lifetime of the photo-electron and measured line-widths

Assuming Lorentzian line-shapes, the total (measured) width is given by:
tot

vh tot h + e ve
~ meV ~ eV

Spectrum dominated by finalstate (photo-electron) line-widths, unless h e

<< v

2D and 1D systems !
Adapted from: T.-C. Chiang, Chemical Physics 251, 133-140 (2000)
Adapted from R. Claessens Cargese-2005 lectures

Spectra analysis: EDCs

Line-shapes and widths many-body physics
(k , ) [ k (k , )]2 + [(k , )]2 1

A(k , ) =

EDC: Lorentzian if and only if

and independent of

Spectra analysis: MDCs

Line-shapes and widths many-body physics
1 A(k , ) =

(k , ) [ k (k , )]2 + [(k , )]2

MDC: Lorentzian if and only if

0 k HWHM = ( ) / vF 0 kc = k F + [ ( )] / vF

Many-body physics Effects of the interactions on the band structure: Example of surface states of Mo(110)

Mo(110) band along N T = 70 K e-imp = const

ee ~ 2 e-ph Eliashberg

T. Valla et al., PRL 83, 2085 (1999)

Strongly-correlated electron systems (brief recall)

Transition-metal oxides: solid-state SCES

MH ~ 1-2 eV
Cu O Cu

Transition-metal oxides Strongly-correlated electrons systems New physics displaying exotic electronic states in a solid sample

Cuprates: antiferromagnetic insulators that become high-Tc superconductors upon doping!

Crystal unit-cell

Antiferromagnetic unit cell

Cuprates: (rough) phase diagram

Coexistence ?

Electron-doped cuprates: Generalities

R2-xCexCuO4 : crystal structure

CuO2 plane (R,Ce)2O2 block CuO2 plane (R,Ce)2O2 block CuO2 plane

R/Ce

Electron-doped cuprates: effects of Ce-doping and annealing

H. J. Kang et al., Nature Materials 6, 224 (Feb. 2007). L. Shan et al., cond-mat/0703256 (March 2007).

Fermi surfaces, Brillouin zones and AF-zones: hole-doped vs electron-doped cuprates

Nd2-xCexCuO4 (x = 0.15)

Tl2Ba2CuO6+d

N. P. Armitage et al., PRL 88, 257001 (2002)

M. Plat et al., PRL 95, 077001 (2005)

Antiferromagnetic-induced band-folding in underdoped Sm2-xCexCuO4 (x = 0.14)

S. R. Park et al., cond-mat/0612419 (Dec. 2006)

Effects of annealing on band structure at optimal doping: the case of Pr1.85Ce0.15CuO4

Annealed

As-grown

P. Richard et al., cond-mat/0704.0453 (Apr. 2007)

Electronic structure and signatures of interactions in Sm1.84Ce0.16CuO4

Coworkers - collaborators Takeshi Kondo, Adam Kaminski (Ames Lab - Iowa) Stphane Pailhs (PSI and LLB) Johan Chang, Ming Shi, Luc Patthey (PSI) Alexandre Zimmers (CSR Maryland and INP P6) Bing Liang, Pengcheng Li, Rick Greene (CSR - Maryland)

Sm2-xCexCuO4 : Fermi surface and doping

0.0 Binding energy [eV]

(,)

-0.1 -0.2 -0.3

Max

(,0)

Min
0.6 1.2
-

0.2

0.4

0.8

1.0

(,)

Momentum along AFZB [2 /a]

Doping from FS volume: x = 0.16 0.01 Single-band FS (no band-folding) Suppressed spectral weight at hot-spots
Tight-binding fit

(,0)

1.4

-0.4

Sm2-xCexCuO4 (x = 0.16) : Nodal vs anti-nodal ARPES spectra

N AN

0.0 Binding energy [eV] -0.2

AN

0.0 Binding energy [eV] -0.2

Max

Max

-0.4 -0.6
0.1
-1

-0.4 -0.6
0.1
-1

Min

Min

-0.8

kAFZB

-0.8

kAFZB

Sm2-xCexCuO4 (x = 0.16) : Nodal and anti-nodal MDCs

AN
= -200 meV

N AN

= -200 meV

= -100 meV

= -100 meV

0.1

-1

= 0 meV

0.1

-1

= 0 meV

Relative momentum

Relative momentum

Anti-nodal and nodal MDCs are Lorentzians MDC peak maximum gives the quasi-particle dispersion MDC width is proportional to quasi-particle scattering rate (self-energy)

Sm2-xCexCuO4 (x = 0.16) : Nodal vs anti-nodal EDCs

Hump N AN AN

AN

AN

-0.8 -0.4 0.0 Binding energy [eV]

Anti-node Peak-hump structure close to kF Lorentzian EDCs at energies > 300 meV

-0.8

-0.6

-0.4

-0.2

0.0

Binding energy [eV]

Node No clear peak-hump structure EDCs are not Lorentzian

Sm2-xCexCuO4 (x = 0.16) : Nodal vs anti-nodal dispersions

0.0 -0.1 Binding energy [eV] -0.2 -0.3 -0.4

AN

N
-0.5 -0.6 Relative momentum 0.1
-1

AN

Sm2-xCexCuO4 (x = 0.16) : Nodal vs anti-nodal line-widths

0.4 Im() [eV] 0.3 0.2
AN
0 ( ) = vband k HWHM

0.1

From EDC-width N

0.0 -0.5

-0.4

-0.3

-0.2

-0.1

0.0

Binding energy [eV]

Conclusions
When it works ARPES is a powerful technique for the study of the electronic structure of complex systems Detailed band structures and Fermi surfaces k-dependent Fermi velocity and effective mass Gaps Many-body effects in the QP dispersion Kinks Fermi-surface nesting

Outlook (and dreams)

Spin-resolved ARPES Time-resolved ARPES Micro-ARPES