Transport in Graphene: Klaus Richter

Transport in Graphene
Klaus Richter
Regensburg Universitat
Strasbourg, July 2012
Graphene layers
Graphite Bilayer graphene
Monolayer graphene
3d layered material with hexagonal 2d layers

gures: Courtesy of E McCann
2-dimensional material; zero-gap semiconductor; Dirac spectrum of electrons
Outline:
monolayer graphene:
electronic properties and tight-binding model physics near K -points: Dirac equation
quantum transport (Landauer formalism) transport in bulk graphene: Chiral tunneling and Klein paradox transport through graphene nanoribbons: edge magnetism and spin currents
Tight binding model of monolayer graphene
+ lecture notes by Ed McCann

Transport in Graphene Strasbourg, July 2012
Bond formation of graphene

Carbon: 6 electrons
2 core electrons 4 valence electrons: one 2s and three 2p orbitals
sp2 hybridization:
2s and two 2p orbitals form three -bonds in plane
stability of graphene as real 2D-system
Bond formation of graphene

Carbon: 6 electrons
2 core electrons 4 valence electrons: one 2s and three 2p orbitals
sp2 hybridization:
remaining 2pz orbital: orbital perpendicular to plane
at energies of interest for transport measurements:

only -orbital relevant keep only this orbital per site in a corresponding tight-binding model
Graphene lattice
honeycomb lattice
two different ways of bond orientation two different types of atomic sites (chemical identical)
Graphene lattice
two different atomic sites two interpenetrating triangular sub-lattices triangular lattice with a basis of two atoms (A and B) per unit cell 2 sublattice degrees of freedom:
Reciprocal lattice
triangular reciprocal lattice hexagonal Brillouin zone with two unique corner points: K-points or K-valleys, 2 valley degrees of freedom:
Tight binding model

Schrodinger equation: H (k, r) = p2 + V (r) (k, r) = E (k)(k, r) 2m
one -orbital per site two orbitals per unit cell tight binding approximation: expansion in Bloch functions (label j = 1 (A sites) and j = 2 (B sites)): 1 j (k, r) = N with
P
Rj :
eikRj j (r Rj )
Rj
sum over all type j atomic sites
j : atomic wavefunction
Tight binding model

Eigenfunction j (k, r) (for j = 1, 2) written as a linear combination of Bloch functions:
2
j (k, r) =
j =1
Cjj (k) j (k, r) = CjA (k) A (k, r) + CjB (k) B (k, r)
Eigenvalues Ej (for j = 1, 2) written as: Ej (k) = in terms of Bloch functions: Ej (k) =

i,l Cji Cjl i |H |l C Cji jl i |l
j |H |j j |j
i,l i,l
Cji Cjl Hil C S Cji jl il
i,l
with transfer matrix elements Hil = i |H |l and overlap matrix elements Sil = i |l
Tight binding model

energy: Ej (k) =
i,l i,l Cji Cjl Hil C S Cji jl il
variational principle: if the Hil and Sil are known, determine energy by minimizing with respect to variations of Cjm : Ej =0 Cjm
2 2
l=1
Hml Cjl = Ej
l=1
Sml Cjl
written as a matrix equation Cj = (CjA , CjB ) : H Cj = Ej S Cj the secular equation gives the eigenvalues Ej : det (H ES ) = 0
Tight binding model

calculation of transfer and overlap integrals: 1 j (k, r) = eikRj j (r Rj ) N Rj Hij = i |H |j , Sij = i |j
diagonal matrix elements: HAA = HBB =

0
SAA = SBB = 1
A and B sites are chemically identical; every A site has three neighbours off-diagonal matrix elements: (nearest-neighbour transfer integrals)
3
HAB = A |H |B = tf (k)
SAB = sf (k)
f (k) =
j =1
eikj
Tight binding model: band structure

Schrodinger equation:
0
tf (k)
0 3
tf (k)
E (k)
1 sf (k) 1 sf (k)
CA CB
(k) = 0
calculation of f (k) = j =1 eikj and energy from secular equation yeilds (Wallace (1947)): E (k) = t with nearest neighbour coupling t and w(k) = 2 cos( 3ky a) + 4 cos ab initio calculation: =0 ; t 3 eV 3 ky a cos 2 ; 3 kx a 2 3 + w(k)
s 0.13 eV
Tight binding model: band structure
six corners of the Brillouin zone (K -points), but only two are non-equivalent
Tight binding: band structure
2 spin-degenerate bands touching each other at the corners of the 1. BZ.: zero gap semiconductor Fermi energy exactly at touching points linear dispersion for low energy excitations expand E (k) near K -points
Graphene effective hamiltonian

linear expansion near K -points: with K = K+1 , K = K1 and valley index = 1: k E (k) = = K + q vF |q| + O((q/K )2 )
Fermi velocity vF = 3ta/(2 ) 106 m/s does not depend on energy ! usually: E (p) = p2 /(2m) ; vF = p/m = 2E/m.
effective wave equation: vF with Cj = CAj CBj = A B 0 px + ipy px ipy 0 = E = vF |q|
Bloch functions amplitudes on the A,B sites mimic pseudospin effective wave equation resembles Dirac equation for massless fermions (with pseudo spin 1/2) in 2 dimensions. : not the full wave function; slowly oscillating envelop wave function
take into account both, K and K -points: (AK , BK , AK , BK )t effective 4 4-Hamiltonian: 0 px + ipy H = vF 0 0 px ipy 0 0 0 0 0 0 px + ipy 0 0 px ipy 0
inter-valley coupling often of relevance

Hamiltonian for one K -point, e.g. = 1: H = vF with x = helical electrons:
chiral operator n projects pseudospin onto quantization axis n eigenstates of H are also eigenstates of n with eigenvalues 1 pseudospin direction is linked to momentum direction
0 px + ipy
px ipy 0 0 1 1 0
= vF (x px + y py ) = vF p = vF p n 0 i i 0
y =
conduction band (electrons): n=1 valence band (holes): n = -1 absence of backscattering in ideal graphene !
Quantum transport: Landauer formalism
From macroscopic to microscopic conductors
classical (Ohms law)

quantum
Conductance quantization at a point contact

Topinka et al., Westervelt lab.
Conductance quantization at a point contact
conductance:
G=2
e2 h
Landauer formula for mesoscopic conductors

conductance: G=2
e2 h
(Landauer formula) quantum transmission:

N
T =
m,n
t mn tmn
applicable to phase-coherent diffusive or ballistic conductors quantum effects:

conductance steps coherent backsctattering: weak localization universal conductance uctuations quantum noise how to compute the conductance numerically ?
Green function formalism for transport

Conductance: G = (e2 /h) T
N N
with
T
m n retarded Green function: self-energies: coupling to lead l:
|tnm |2 = Tr(l Gr l Ga )
n=1 m=1
Gr = (E Hscat r )1 r =
leads
r l
a l = i(r l l )
use recursive Green function techniques (matrix reordering strategies; graph-theoretical approaches)
M. Wimmer and KR, J. Comp. Phys (2009), arXiv:0806.2739
Tight-binding model for graphene
tight-binding model H=
ij nn nearest-neighbor hopping
t |i j |
ij nnn
t |i j |
next-nearest neighbor hopping
+
i
Ui |i i|
on-site energy
Ui =
Vi + Mi Vi Mi
for i in sublattice A for i in sublattice B

V : potential, M : staggered potential

Quantum transport through graphene

Landauer approach + tight-binding approximation
Transport in bulk graphene: chiral tunneling and Klein paradox
Chiral tunneling and Klein paradox in graphene

Potential barrier for ...
... monolayer graphene cos2 1 cos2 [qx (E V0 )D] sin2
T =
... bilayer graphene
M.I. Katsnelson, K.S. Novoselov, A.K. Geim, Nat. Phys. 2, 620 (2006)
Chiral tunneling and Klein paradox in graphene

MLG: R|=0 = 0 BLG: T |=0 = 0
(Klein tunneling at = 0)
(anti-Klein tunneling at = 0)
V.V. Cheianov, V.I. Falko, Phys. Rev. B 74, 041403 (2006) M.H. Liu, J. Bundesmann, KR, Phys. Rev. B 85, 085406 (2012)
Klein backscattering phase shift

consider a phase-coherent npn junction prediction of a half-period phase shift in the associated Fabry-Perot oscillation: = 2WKB + 1 + 2
A.V. Shytov, M.S. Rudner, L.S. Levitov, Phys. Rev. Lett. 101, 156804 (2008).
A key experiment on Klein backscattering
A.F. Young and P. Kim, Nature Phys. 5, 222 (2009)

Observation of Klein backscattering phase shift
A.F. Young and P. Kim, Nature Phys. 5, 222 (2009)

Transport in graphene nanoribbons: edge magnetism and spin currents
Spins in graphene
small intrinsic spin-orbit coupling long spin lifetimes expected graphene as prospective material for spin electronics successful spin injection from ferromagnetic contacts into graphene
Hill et al., Trans. Magn. (2006); Oishi et al., Jpn. J. Appl. Phys. (2007); Tombros et al., Nature (2007):
how to generate spin currents in graphene without ferromagnets? zigzag nanoribbons

Edge states in zigzag nanoribbons

Zigzag graphene nanoribbons: existence of a localized state at the edges
(see eg. Fujita, Wakabayashi, Nakada, Kusakabe (1996)
experimental observation:
(eg. Kobayashi et al. Phys. Rev B (2006); Niimi et al. Phys. Rev. B (2006); see also: Ritter and Lyding, Nature Mat. (2009) )
Edge states in zigzag nanoribbons
scanning tunneling experiment DFT calculation of local DoS
X. Zhang et al., Crommie lab. (2012)

Edge magnetism in zigzag nanoribbons

Zigzag graphene nanoribbons: at-band high density of states magnetism of the edge state
Fujita et al., JPSJ (1996) (mean eld Hubbard model) Y.-W. Son et al., Nature (2007) (DFT) H. Feldner et al. (2011)
edge magnetization within each sublattice A and B ground state: opposite magnetization of A and B edge states staggered magnetization


3 2
E-EF [eV]
1 0 -1
-2 -3
X
Modelling edge state magnetism

staggered magnetization of edge state: Hmag = Ms M s on sublattice A on sublattice B
t magnetization to DFT calculations:
Spin transport at the edges

magnetic edge states are strongly localized at the edges transport mediated through edge states nnn-coupling in tight-binding Hamiltonian essential ! two-wire mechanism of spin transport in graphene ribbon:
break up-down symmetry to achieve spin conductance !
0000000000000 1111111111111 0000000000000 1111111111111 1111111111111 0000000000000 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 1111111111111 0000000000000 0000000000000 1111111111111 0000000000000 1111111111111
Spin injection in graphene
spin Hall type effect

M. Wimmer, I. Adagideli, S. Berber, D. Tomanek, KR, Phys. Rev. Lett. (2009)
Spin injection in graphene
net spin conductance !

M. Wimmer, I. Adagideli, S. Berber, D. Tomanek, KR, Phys. Rev. Lett. (2009)
Ribbons with rough edges

more realistic in todays experiments: rough edges:
0000000000000 1111111111111 1111111111111 0000000000000 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111
R1
1111111111111 0000000000000 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111 0000000000000 1111111111111
R2
Spin conductance: Gs = G G If R1 = R2 , Gs = 0. However: mesoscopic conductance uctuations! Spin conductance uctuations: Var Gs = Var G + Var G = Var Gtot rms Gs = Var Gs = rms Gtot
Spin conductance uctuations

typical spin density in a disordered graphene nanoribbon:
corresponding spin current density
Spin conductance uctuations

Disordered graphene nanoribbon: single ribbon
Further transport phenomena in graphene
spintronics in graphene graphene as a topological insulator quantum Hall effect weak (anti-)localization in graphene transport through graphene quantum dots / billiards
Hints for further reading about graphene...

tight-binding hamiltonian and electronic structure:
Saito, Dresselhaus, Dresselhaus: Physical Properties of Carbon Nanostructures E. McCann: Electronic properties of monolayer and bilayer graphene, arXive cond-mat 1205.4849 A.H. Castro Neto et al.: The Electronic properties of graphene, Rev. Mod. Phys. 81, 109 (2009)
transport:
N.M.R. Peres: The transport properties of graphene: an introduction, Rev. Mod. Phys. 82, 2673 (2010)
chiral tunneling:
M.I. Katsnelson, K.S. Novoselov, A.K. Geim, Nat. Phys. 2, 620 (2006)
nanoribbons and edge magnetism:

O. Yazyev, Emergence of magnetism in graphene materials and nanostructures, Rep. Prog. Phys., 73, 056501 (2010)

Transport in Graphene: Klaus Richter

Uploaded by

Copyright:

Available Formats

Transport in Graphene: Klaus Richter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Transport in Graphene: Klaus Richter

Uploaded by

Copyright:

Available Formats

Transport in Graphene

Strasbourg, July 2012

3d layered material with hexagonal 2d layers

2-dimensional material; zero-gap semiconductor; Dirac spectrum of electrons

Strasbourg, July 2012

Strasbourg, July 2012

Tight binding model of monolayer graphene

+ lecture notes by Ed McCann

Bond formation of graphene

stability of graphene as real 2D-system

Strasbourg, July 2012

Bond formation of graphene

at energies of interest for transport measurements:

Strasbourg, July 2012

Strasbourg, July 2012

Strasbourg, July 2012

Strasbourg, July 2012

Tight binding model

sum over all type j atomic sites

Strasbourg, July 2012

Tight binding model

Cjj (k) j (k, r) = CjA (k) A (k, r) + CjB (k) B (k, r)

Eigenvalues Ej (for j = 1, 2) written as: Ej (k) = in terms of Bloch functions: Ej (k) =

Cji Cjl Hil C S Cji jl il

Tight binding model

Tight binding model

diagonal matrix elements: HAA = HBB =

Strasbourg, July 2012

Tight binding model: band structure

Strasbourg, July 2012

Tight binding model: band structure

Strasbourg, July 2012

Tight binding: band structure

Strasbourg, July 2012

Graphene effective hamiltonian

Strasbourg, July 2012

Graphene effective hamiltonian

effective wave equation: vF with Cj = CAj CBj = A B 0 px + ipy px ipy 0 = E = vF |q|

Strasbourg, July 2012

Graphene effective hamiltonian

inter-valley coupling often of relevance

Strasbourg, July 2012

Graphene effective hamiltonian

Quantum transport: Landauer formalism

Strasbourg, July 2012

From macroscopic to microscopic conductors

classical (Ohms law)

Conductance quantization at a point contact

Strasbourg, July 2012

Conductance quantization at a point contact

Strasbourg, July 2012

Landauer formula for mesoscopic conductors

(Landauer formula) quantum transmission:

applicable to phase-coherent diffusive or ballistic conductors quantum effects:

Green function formalism for transport

Tight-binding model for graphene

for i in sublattice A for i in sublattice B

V : potential, M : staggered potential

Quantum transport through graphene