The electronic properties of graphene

A. H. Castro Neto
Department of Physics,
Boston University,
590 Commonwealth Avenue,
Boston, MA 02215,
USA

F. Guinea
Instituto de Ciencia de Materiales de Madrid. CSIC. Cantoblanco. E-28049 Madrid,
Spain

N. M. R. Peres
Center of Physics and Department of Physics,
Universidade do Minho,
P-4710-057, Braga,
Portugal

K. S. Novoselov and A. K. Geim

Department of Physics and Astronomy,
University of Manchester,
Manchester, M13 9PL,
UK

(Dated: December 26, 2012)

This article reviews the basic theoretical aspects of graphene, a one atom thick allotrope
of carbon, with unusual two-dimensional Dirac-like electronic excitations. The Dirac
electrons can be controlled by application of external electric and magnetic fields, or by
altering sample geometry and/or topology. We show that the Dirac electrons behave in
unusual ways in tunneling, confinement, and integer quantum Hall effect. We discuss
the electronic properties of graphene stacks and show that they vary with stacking order
and number of layers. Edge (surface) states in graphene are strongly dependent on the
edge termination (zigzag or armchair) and affect the physical properties of nanoribbons.
We also discuss how different types of disorder modify the Dirac equation leading to
unusual spectroscopic and transport properties. The effects of electron-electron and
electron-phonon interactions in single layer and multilayer graphene are also presented.

CONTENTS N. Spin orbit coupling 21

I. Introduction 2 III. Flexural phonons, elasticity, and crumpling 22

II. Elementary electronic properties of graphene 4 IV. Disorder in graphene 24

A. Single layer: tight-binding approach 4 A. Ripples 25
1. Cyclotron mass 5 B. Topological lattice defects 26
2. Density of states 5 C. Impurity states 27
B. Dirac fermions 6 D. Localized states near edges, cracks, and voids 27
1. Chiral Tunneling and Klein paradox 7 E. Self-doping 28
2. Confinement and zitterbewegung 8 F. Vector potential and gauge field disorder 29
C. Bilayer graphene: tight-binding approach 9 1. Gauge field induced by curvature 29
D. Epitaxial graphene 10 2. Elastic strain 30
E. Graphene stacks 11 3. Random gauge fields 31
1. Electronic structure of bulk graphite 12 G. Coupling to magnetic impurities 31
F. Surface states in graphene 13 H. Weak and strong localization 31
G. Surface states in graphene stacks 14 I. Transport near the Dirac point 33
H. The spectrum of graphene nanoribbons 15 J. Boltzmann Equation description of DC transport in
1. Zigzag nanoribbons 15 doped graphene 33
2. Armchair nanoribbons 16 K. Magnetotransport and universal conductivity 35
I. Dirac fermions in a magnetic field 17 1. The full self-consistent Born approximation
J. The anomalous integer quantum Hall effect 18 (FSBA) 35
K. Tight-binding model in a magnetic field 19
L. Landau levels in graphene stacks 20 V. Many-body effects 37
M. Diamagnetism 20 A. Electron-phonon interactions 37
 2

B. Electron-electron interactions 38
1. Screening in graphene stacks 41 FIG. 1 (Color online) Graphene (top left) is a honeycomb
C. Short range interactions 41 lattice of carbon atoms. Graphite (top right) can be viewed
1. Bilayer graphene: exchange 42 a stack of graphene layers. Carbon nanotubes are rolled-
2. Bilayer graphene: short range interactions 42 up cylinders of graphene (bottom left). Fullerenes (C60 ) are
D. Interactions in high magnetic fields 43 molecules consisting of wrapped graphene by the introduction
of pentagons on the hexagonal lattice.
VI. Conclusions 43

VII. Acknowledgments 44
Graphene was eventually spotted due to the subtle op-
References 44
tical effect it creates on top of a cleverly chosen SiO2
substrate (Novoselov et al., 2004) that allows its observa-
I. INTRODUCTION tion with an ordinary optical microscope (Abergel et al.,
2007; Blake et al., 2007; Casiraghi et al., 2007). Hence,
Carbon is the materia prima for life on the planet and graphene is relatively straightforward to make, but not
the basis of all organic chemistry. Because of the flexi- so easy to find.
bility of its bonding, carbon-based systems show an un- The structural flexibility of graphene is reflected in its
limited number of different structures with an equally electronic properties. The sp2 hybridization between one
large variety of physical properties. These physical prop- s-orbital and two p-orbitals leads to a trigonal planar
erties are, in great part, the result of the dimensional- structure with a formation of a σ-bond between carbon
ity of these structures. Among systems with only car- atoms which are separated by 1.42 Å. The σ-band is re-
bon atoms, graphene - a two-dimensional (2D) allotrope sponsible for the robustness of the lattice structure in all
of carbon - plays an important role since it is the ba- allotropes. Due to the Pauli principle these bands have a
sis for the understanding of the electronic properties filled shell and hence, form a deep valence band. The un-
in other allotropes. Graphene is made out of carbon affected p-orbital, which is perpendicular to the planar
atoms arranged on a honeycomb structure made out of structure, can bind covalently with neighboring carbon
hexagons (see Fig. 1), and can be thought as composed atoms leading to the formation of a π-band. Since each
of benzene rings stripped out from their hydrogen atoms p-orbital has one extra electron, the π-band is half-filled.
(Partoens and Peeters, 2006). Fullerenes (Andreoni, Half-filled bands in transition elements have played an
2000) are molecules where carbon atoms are arranged important role in the physics of strongly correlated sys-
spherically, and hence, from the physical point of view, tems since, due to its strong tight binding character, the
are zero-dimensional objects with discrete energy states. Coulomb energies are very large, leading to strong col-
Fullerenes can be obtained from graphene with the in- lective effects, magnetism, and insulating behavior due
troduction of pentagons (that create positive curvature to correlation gaps or Mottness (Phillips, 2006). In fact,
defects), and hence, fullerenes can be thought as wrapped Linus Pauling proposed in the 1950’s that, on the basis
up graphene. Carbon nanotubes (Charlier et al., 2007) of the electronic properties of benzene, graphene should
are obtained by rolling graphene along a given direc- be a resonant valence bond structure (RVB) (Partoens
tion and reconnecting the carbon bonds. Hence, carbon and Peeters, 2006). RVB states have become very pop-
nanotubes have only hexagons and can be thought as ular in the literature of transition metal oxides, and
one-dimensional (1D) objects. Graphite, a three dimen- particularly in studies of cuprate oxides superconduc-
sional (3D) allotrope of carbon, became widely known to tors (Maple, 1998). This point of view, should be con-
mankind after the invention of the pencil in 1564 and its trasted with contemporaneous band structure studies of
usefulness as an instrument for writing comes from the graphene (Wallace, 1947) that found it to be a semimetal
fact that graphite is made out of stacks of graphene layers with unusual linearly dispersing electronic excitations
that are weakly coupled by van der Waals forces. Hence, called Dirac electrons. While most of the current exper-
when one presses a pencil against a sheet of paper one imental data in graphene supports the band structure
is actually producing graphene stacks and, somewhere point of view, the role of the electron-electron interac-
among them, there could be individual graphene layers. tions in graphene is a subject of intense research.
Although graphene is the mother for all these different al- It was P. R. Wallace who in 1946 wrote the first pa-
lotropes and has been presumably produced every time pers on the band structure of graphene and showed the
someone writes with a pencil, it was only isolated 440 unusual semimetallic behavior in this material (Wallace,
years after its invention (Novoselov et al., 2004). The 1947). At that point in time, the thought of a purely
reason is that, first, no one actually expected graphene 2D structure was a mere fantasy and Wallace’s stud-
to exist in the free state and, second, even with the ben- ies of graphene served him as a starting point to study
efit of hindsight, no experimental tools existed to search graphite, a very important material for nuclear reactors
for one-atom-thick-flakes among the pencil debris cov- in the post-World War II era. During the following years,
ering macroscopic areas (Geim and MacDonald, 2007). the study of graphite culminated with the Slonczewski-
 3

Weiss-McClure (SWM) band structure of graphite which graphene these electrostatic potentials can be easily gen-
provided a detailed description of the electronic prop- erated by disorder. Since disorder is unavoidable in any
erties in this material (McClure, 1957; Slonczewski and material, there has been great interest in trying to un-
Weiss, 1958) and was very successful in describing the ex- derstand how disorder affects the physics of electrons in
perimental data (Boyle and Nozières, 1958; Dillon et al., graphene and its transport properties. In fact, under
1977; McClure, 1958; Soule et al., 1964; Spry and Scherer, certain conditions, Dirac fermions are immune to local-
1960; Williamson et al., 1965). Interestingly enough, ization effects observed in ordinary electrons (Lee and
from 1957 to 1968, the assignment of the electron and Ramakrishnan, 1985) and it has been established exper-
hole states within the SWM model were the opposite imentally that electrons can propagate without scatter-
to what is accepted today. In 1968, Schroeder et al. ing over large distances of the order of micrometers in
(Schroeder et al., 1968) established the currently ac- graphene (Novoselov et al., 2004). The sources of disor-
cepted location of electron and hole pockets. The SWM der in graphene are many and can vary from ordinary
model has been revisited in recent years because of its effects commonly found in semiconductors, such as ion-
inability to describe the van der Waals-like interactions ized impurities in the Si substrate, to adatoms and vari-
between graphene planes, a problem that requires the ous molecules adsorbed in the graphene surface, to more
understanding of many-body effects that go beyond the unusual defects such as ripples associated with the soft
band structure description (Rydberg et al., 2003). These structure of graphene. In fact, graphene is unique in the
issues, however, do not arise in the context of a single sense that it shares properties of soft membranes and at
graphene crystal but they show up with great importance the same time it behaves in a metallic way, so that the
when graphene layers are stacked on top of each other, as Dirac fermions propagate on a locally curved space. Here,
in the case, for instance, of the bilayer graphene. Stack- analogies with problems of quantum gravity become ap-
ing can change the electronic properties considerably and parent. The softness of graphene is related with the fact
the layering structure can be used in order to control the that it has out-of-plane vibrational modes (phonons) that
electronic properties. cannot be found in 3D solids. These flexural modes, re-
One of the most interesting aspects of the graphene sponsible for the bending properties of graphene, also
problem is that its low energy excitations are massless, account for the lack of long range structural order in soft
chiral, Dirac fermions. In neutral graphene the chemical membranes leading the phenomenon of crumpling. Nev-
potential crosses exactly the Dirac point. This particular ertheless, the presence of a substrate or scaffolds that
dispersion, that is only valid at low energies, mimics the hold graphene in place, can stabilize a certain degree of
physics of quantum electrodynamics (QED) for massless order in graphene but leaves behind the so-called ripples
fermions except by the fact that in graphene the Dirac (which can be viewed as frozen flexural modes).
fermions move with a speed vF which is 300 times smaller It was realized very early on that graphene should also
than the speed of light, c. Hence, many of the unusual present unusual mesoscopic effects (Katsnelson, 2007a;
properties of QED can show up in graphene but at much Peres et al., 2006a). These effects have their origin in the
smaller speeds. Dirac fermions behave in very unusual boundary conditions required for the wavefunctions in
ways when compared to ordinary electrons if subjected mesoscopic samples with various types of edges graphene
to magnetic fields, leading to new physical phenomena can have (Akhmerov and Beenakker, 2008; Nakada et al.,
(Gusynin and Sharapov, 2005; Peres et al., 2006c) such 1996; Peres et al., 2006c; Wakabayashi et al., 1999). The
as the anomalous integer quantum Hall effect (IQHE) most studied edges, zigzag and armchair, have drastically
measured experimentally (Zhang et al., 2005). Besides different electronic properties. Zigzag edges can sustain
being qualitatively different from the IQHE observed in edge (surface) states and resonances that are not present
Si and GaAlAs (heterostructures) devices, the IQHE in in the armchair case. Moreover, when coupled to con-
graphene can be observed at room temperature because ducting leads, the boundary conditions for a graphene
of the large cyclotron energies for “relativistic” electrons ribbon strongly affects its conductance and the chiral
(Novoselov et al., 2007). In fact, the anomalous IQHE is Dirac nature of the fermions in graphene can be ex-
the trademark of Dirac fermion behavior. ploited for applications where one can control the valley
Another particularly interesting feature of Dirac flavor of the electrons besides its charge, the so-called val-
fermions is their insensitivity to external electrostatic leytronics. Furthermore, when superconducting contacts
potentials due to the so-called Klein paradox, that is, are attached to graphene, they lead to the development
the fact that Dirac fermions can be transmitted with of supercurrent flow and Andreev processes character-
probability one through a classically forbidden region istic of superconducting proximity effect. The fact that
(Calogeracos and Dombey, 1999; Itzykson and Zuber, Cooper pairs can propagate so well in graphene attests for
2006). In fact, Dirac fermions behave in a very un- the robust electronic coherence in this material. In fact,
usual way in the presence of confining potentials lead- quantum interference phenomena such as weak localiza-
ing to the phenomenon of zitterbewegung, or jittery mo- tion, universal conductance fluctuations (Morozov et al.,
tion of the wavefunction (Itzykson and Zuber, 2006). In 2006), and the Aharonov-Bohm effect in graphene rings
 4

have already been observed experimentally (Recher et al., vectors can be written as:
2007; Russo et al., 2008). The ballistic electronic prop- a √ a √
agation in graphene can be used for field effect devices a1 = (3, 3) , a2 = (3, − 3) , (1)
2 2
such as p-n (Cheianov et al., 2007a; Cheianov and Fal’ko,
2006; Fogler et al., 2008; Huard et al., 2007; Tworzydlo where a ≈ 1.42 Å is the carbon-carbon distance. The
et al., 2007; Williams et al., 2007; Zhang and Fogler, reciprocal lattice vectors are given by:
2008) and p-n-p (Ossipov et al., 2007) junctions, and as 2π √ 2π √
“neutrino” billiards (Berry and Mondragon, 1987; Miao b1 = (1, 3) , b2 = (1, − 3) . (2)
3a 3a
et al., 2007). It has also been suggested that Coulomb
interactions are considerably enhanced in smaller geome- Of particular importance for the physics of graphene are
tries, such as graphene quantum dots, leading to unusual the two points K and K 0 at the corners of the graphene
Coulomb blockade effects (Geim and Novoselov, 2007) Brillouin zone (BZ). These are named Dirac points for
and perhaps to magnetic phenomena such as the Kondo reasons that will become clear later. Their positions in
effect. The amazing transport properties of graphene al- momentum space are given by:
low for their use in a plethora of applications ranging 
2π 2π
 
2π 2π

from single molecule detection (Wehling et al., 2008) to K= , √ , K0 = ,− √ . (3)
3a 3 3a 3a 3 3a
spin injection (Cho et al., 2007; Ohta, 1971b).
Because of its unusual structural and electronic flexi- The three nearest neighbors vectors in real space are
bility, graphene can be tailored chemically and/or struc- given by:
turally in many different ways: deposition of metal atoms a √ a √
(Calandra and Mauri, 2007) or molecules (Leenaerts δ1 = (1, 3) δ2 = (1, − 3) δ3 = −a(1, 0)
2 2
et al., 2008; Wehling et al., 2008) on top; intercalation (4)
(as it is done in graphite intercalated compounds (Dres- while the six second-nearest neighbors are located at:
selhaus and Dresselhaus, 2002; Dresselhaus et al., 1983)); δ10 = ±a1 , δ20 = ±a2 , δ30 = ±(a2 − a1 ).
incorporation of nitrogen and/and boron in its structure The tight-binding Hamiltonian for electrons in
(Martins et al., 2007; Peres et al., 2007a) (in analogy graphene considering that electrons can hop both to near-
with what has been done in nanotubes (Stephan et al., est and next nearest neighbor atoms has the form (we use
1994)); using different substrates that modify the elec- units such that ~ = 1):
tronic structure (Calizo et al., 2007; Faugeras et al., 2008; X  † 
Giovannetti et al., 2007; Varchon et al., 2007). The con- H =−t aσ,i bσ,j + h.c.
trol of graphene properties can be extended in new di- hi,ji,σ
X  
rections allowing for creation of graphene-based systems −t 0
a†σ,i aσ,j + b†σ,i bσ,j + h.c. , (5)
with magnetic and superconducting properties (Uchoa hhi,jii,σ
and Castro Neto, 2007) that are unique in their 2D prop-
erties. Although the graphene field is still in its infancy, where ai,σ (a†i,σ ) annihilates (creates) an electron with
the scientific and technological possibilities of this new spin σ (σ =↑, ↓) on site Ri on sublattice A (an equiva-
material seem to be unlimited. The understanding and lent definition is used for sublattice B), t (≈ 2.8 eV) is the
control of the properties of this material can open doors nearest neighbor hopping energy (hopping between dif-
for a new frontier in electronics. As the current status of ferent sublattices), t0 1 is the next nearest neighbor hop-
the experiment and potential applications have recently ping energy (hopping in the same sublattice). The en-
been reviewed (Geim and Novoselov, 2007), in this article ergy bands derived from this Hamiltonian have the form
we mostly concentrate on the theory and more technical (Wallace, 1947):
aspects of electronic properties of this exciting new ma- p
terial. E± (k) = ±t 3 + f (k) − t0 f (k) ,

√ √ !  
 3 3
f (k) = 2 cos 3ky a + 4 cos ky a cos kx a ,
2 2
II. ELEMENTARY ELECTRONIC PROPERTIES OF
(6)
GRAPHENE

A. Single layer: tight-binding approach

1 The value of t0 is not well known but ab initio calculations (Reich
et al., 2002) find 0.02t . t0 . 0.2t depending on the tight-binding
Graphene is made out of carbon atoms arranged in parameterization. These calculations also include the effect of a
hexagonal structure as shown in Fig. 2. The structure third nearest neighbors hopping, which has a value of around
is not a Bravais lattice but can be seen as a triangular 0.07 eV. A tight binding fit to cyclotron resonance experiments
lattice with a basis of two atoms per unit cell. The lattice (Deacon et al., 2007) finds t0 ≈ 0.1 eV.
 5

FIG. 2 (Color online) Left: Lattice structure of graphene, FIG. 4 (Color online) Cyclotron mass of charge carriers in
made out of two interpenetrating triangular lattices (a1 and graphene as a function of their concentration n. Positive and
a2 are the lattice unit vectors, and δi , i = 1, 2, 3 are the near- negative n correspond to electrons and holes, respectively.
est neighbor vectors); Right: corresponding Brillouin zone. Symbols are the experimental data extracted from tempera-
The Dirac cones are located at the K and K’ points. ture dependence of the SdH oscillations; solid curves is the
best fit to Eq. (13). m0 is the free electron mass.

FIG. 3 (Color online) Left: Energy spectrum (in units of t)

for finite values of t and t0 , with t =2.7 eV and t0 = 0.2t. 1. Cyclotron mass
Right: zoom-in of the energy bands close to one of the Dirac
points. The energy dispersion (7) resembles the energy of
ultra-relativistic particles; these particles are quantum
mechanically described by the massless Dirac equation
where the plus sign applies to the upper (π) and the (see section II.B for more on this analogy). An immedi-
minus sign the lower (π ∗ ) band. It is clear from (6) that ate consequence of this massless Dirac-like dispersion is
the spectrum is symmetric around zero energy if t0 = a cyclotron mass that depends on the electronic density
0. For finite values of t0 the electron-hole symmetry is as its square root (Zhang et al., 2005). The cyclotron
broken and the π and π ∗ bands become asymmetric. In mass is defined, within the semiclassical approximation
Fig. 3 we show the full band structure of graphene with (Ashcroft and Mermin, 1976), as
both t and t0 . In the same figure we also show a zoom  
in of the band structure close to one of the Dirac points 1 ∂A(E)
m∗ = , (10)
(at the K or K’ point in the BZ). This dispersion can 2π ∂E E=EF
be obtained by expanding the full band structure, eq.(6),
close to the K (or K0 ) vector, eq.(3), as: k = K + q, with with A(E) the area in k−space enclosed by the orbit and
|q|  |K| (Wallace, 1947): given by:

E2
E± (q) ≈ ±vF |q| + O((q/K)2 ) , (7) A(E) = πq(E)2 = π . (11)
vF2
where q is the momentum measured relatively to the Using (11) in (10) one obtains:
Dirac points and vF represents the Fermi velocity, given
by vF = 3ta/2, with a value vF ' 1 × 106 m/s. This EF kF
result was first obtained by Wallace (Wallace, 1947). m∗ = = . (12)
vF2 vF
The most striking difference between this result and
the usual case, (q) = q 2 /(2m) where m is the electron The electronic density, n, is related to the Fermi momen-
mass, is that the Fermi velocity in (7) does not depend tum, kF , as kF2 /π = n (with contributions from the two
on the energy
p or momentum: in the usual case we have Dirac points K and K 0 and spin included) which leads
v = k/m = 2E/m and hence the velocity changes sub- to:
stantially with energy. The expansion of the spectrum √
π√
around the Dirac point including t0 up to second order in ∗
m = n. (13)
q/K is given by: vF

Fitting (13) to the experimental data (see Fig.4) provides

9t0 a2 3ta2
 
0
E± (q) ' 3t ±vF |q|− ± sin(3θq ) |q|2 , (8) an estimation for the Fermi velocity and the hopping pa-
4 8 rameter as vF ≈ 106 ms−1 and t ≈ √ 3 eV, respectively.
The experimental observation of the n dependence of
where the cyclotron mass provides evidence for the existence of
  massless Dirac quasiparticles in graphene (Deacon et al.,
qx 2007; Jiang et al., 2007a; Zhang et al., 2005) - the usual
θq = arctan , (9)
qy parabolic (Schrödinger) dispersion implies a constant cy-
clotron mass.
is the angle in momentum space. Hence, the presence of t0
shifts in energy the position of the Dirac point and breaks
electron-hole symmetry. Notice that up to order (q/K)2 2. Density of states
the dispersion depends on the direction in momentum
space and has a three fold symmetry. This is the so- The density of states per unit cell, derived from (6), is
called trigonal warping of the electronic spectrum (Ando given in Fig. 5 for both t0 = 0 and t0 6= 0, showing in
et al., 1998; Dresselhaus and Dresselhaus, 2002). both cases a semimetallic behavior (Bena and Kivelson,
 6
√
where Ac is the unit cell area given by Ac = 3 3a2 /2. It
FIG. 5 (Color online) Density of states per unit cell as a
is worth noting that the density of states for graphene is
function of energy (in units of t) computed from the energy
dispersion (6), t0 = 0.2t (top) and for t0 = 0 (bottom). Also very different from the density of states of carbon nan-
√
shown is a zoom in of the density of states close to the neu- otubes (Saito et al., 1992a,b). The latter show 1/ E
trality point of one electron per site. For the case t0 = 0 the singularities due to the 1D nature of their electronic
electron-hole nature of the spectrum is apparent and the den- spectrum, which comes about due to the quantization
sity of states close to the neutrality point can be approximated of the momentum in the direction perpendicular to the
by ρ() ∝ |—. tube axis. From this perspective, graphene nanoribbons,
which also have momentum quantization perpendicular
to the ribbon length, have properties very similar to car-
2005; Wallace, 1947). For t0 = 0 it is possible to derive bon nanotubes.
an analytical expression for the density of states per unit
cell, which has the form (Hobson and Nierenberg, 1953):
B. Dirac fermions

4 |E| 1 π
r !
Z1 Let us consider Hamiltonian (5) with t0 = 0 and con-
ρ(E) = 2 2 √ F , sider the Fourier transform of the electron operators:
π t Z0 2 Z0
 
2  2
E 1 X −ik·Rn


 E
2
t −1 an = √ e a(k), , (16)
 1+

 − ; −t ≤ E ≤ t Nc k
Z0 = t 4
where Nc is the number of unit cells. Using this transfor-
mation, let us write the field an as a sum of two terms,


4 E


; −3t ≤ E ≤ −t ∨ t ≤ E ≤ 3t coming from expanding the Fourier sum around K 0 and

t
K. This produces an approximation for the representa-
tion of the field an as a sum of two new fields, written

E

 4 ; −t ≤ E ≤ t as
 t



0
Z1 = 
2  2 an ' e−iK·Rn a1,n + e−iK ·Rn
a2,n ,
E

 E
2
t −1 −iK·Rn 0
−iK ·Rn
bn ' e

 1+

 − ; −3t ≤ E ≤ −t ∨ t ≤ E ≤ 3t b1,n + e b2,n , (17)
t 4
(14) where the index i = 1 (i = 2) refers to the K (K’)
where F(π/2, x) is the complete elliptic integral of the point. These new fields, ai,n and bi,n are assumed to
first kind. Close to the Dirac point the dispersion is ap- vary slowly over the unit cell. The procedure for de-
proximated by (7) and the expression for the density of riving a theory that is valid close to the Dirac point
states per unit cell is given by (with a degeneracy of four consists in using this representation in the tight-binding
included): Hamiltonian and expanding the operators up to a lin-
ear
P order in δ.PIn the derivation one uses the fact that
±iK·δ ±iK 0 ·δ
e = δe = 0. After some straightfor-
2Ac |E| δ
ρ(E) = (15) ward algebra we arrive at (Semenoff, 1984):
π vF2

Z  √   √  
0 √ 3a(1 − i 3)/4 0√ 3a(−i − 3)/4
H ' −t dxdy Ψ̂†1 (r) ∂x + ∂y Ψ̂1 (r)
−3a(1 + i 3)/4 0 −3a(i − 3)/4 0
 √   √  
† 0 √ 3a(1 + i 3)/4 0 √ 3a(i − 3)/4
+ Ψ̂2 (r) ∂x + ∂y Ψ̂2 (r)
−3a(1 − i 3)/4 0 −3a(−i − 3)/4 0
Z  
= −ivF dxdy Ψ̂†1 (r)σ · ∇Ψ̂1 (r) + Ψ̂†2 (r)σ ∗ · ∇Ψ̂2 (r) , (18)

with Pauli matrices σ = (σx , σy ), σ ∗ = (σx , −σy ), and guage, the two-component electron wavefunction, ψ(r),
Ψ̂†i = (a†i , b†i ) (i = 1, 2). It is clear that the effective close to the K point, obeys the 2D Dirac equation:
Hamiltonian (18) is made of two copies of the massless
Dirac-like Hamiltonian, one holding for p around K and
other for p around K 0 . Notice that, in first quantized lan-
−ivF σ · ∇ψ(r) = Eψ(r) . (19)
 7

The wavefunction, in momentum space, for the mo-

FIG. 6 (Color online) Top: Schematic picture of the scatter-
mentum around K has the form:
ing of Dirac electrons by a square potential. Bottom: defini-
 −iθ /2  tion of the angles φ and θ used in the scattering formalism in
1 e k
ψ±,K (k) = √ iθk /2 , (20) the three regions I, II, and III.
2 ±e

for HK = vF σ · k, where the ± signs correspond to the 1. Chiral Tunneling and Klein paradox
eigenenergies E = ±vF k, that is, for the π and π ∗ band,
respectively, and θk is given by (9). The wavefunction In this section we want to address the scattering of chi-
for the momentum around K 0 has the form: ral electrons in two dimensions by a square barrier (Kat-
 iθ /2  snelson, 2007c). The one dimensional scattering of chiral
1 e k electrons was discussed earlier in the context of carbon
ψ±,K0 (k) = √ −iθ /2 , (21)
2 ±e k nanotubes (Ando et al., 1998; McEuen et al., 1999)
We start by noticing that by a gauge transformation
for HK 0 = vF σ ∗ · k. Notice that the wavefunctions the wavefunction (20) can be written as:
at K and K0 are related by time reversal symmetry:  
if we set the origin of coordinates in momentum space 1 1
ψK (k) = √ iθ . (24)
in the M-point of the BZ (see Fig.2), time reversal be- 2 ±e k
comes equivalent to a reflection along the kx axis, that We further assume that the scattering does not mix the
is, (kx , ky ) → (kx , −ky ). Also note that if the phase θ is momenta around K and K 0 points. In Fig. 6 we depict
rotated by 2π the wavefunction changes sign indicating the scattering process due to the square barrier of width
a phase of π (in the literature this is commonly called a D.
Berry’s phase). This change of phase by π under rotation The wavefunction in the different regions can be writ-
is characteristic of spinors. In fact, the wavefunction is a ten in terms of incident and reflected waves. In region I
two component spinor. we have:
A relevant quantity used to characterize the eigenfunc- 1

1

tions is their helicity defined as the projection of the ψI (r) = √ iφ ei(kx x+ky y)
2 se
momentum operator along the (pseudo)spin direction.  
The quantum mechanical operator for the helicity has r 1
+√ i(π−φ) ei(−kx x+ky y) , (25)
the form: 2 se
with φ = arctan(ky /kx ), kx = kF cos φ, ky = kF sin φ
1 p
ĥ = σ· . (22) and kF the Fermi momentum. In region II we have:
2 |p|  
a 1
ψII (r) = √ 0 iθ ei(qx x+ky y)
It is clear from the definition of ĥ that the states ψK (r) 2 se
 
and ψK 0 (r) are also eigenstates of ĥ: b 1
+√ 0 i(π−θ) ei(−qx x+ky y) , (26)
2 se
1
ĥψK (r) = ± ψK (r), (23) with θ = arctan(ky /qx ) and
2 q
and an equivalent equation for ψK 0 (r) with inverted sign. qx = (V0 − E)2 /(vF2 ) − ky2 , (27)
Therefore electrons (holes) have a positive (negative) he- and finally in region III we have a transmitted wave only:
licity. Equation (23) implies that σ has its two eigen-  
values either in the direction of (⇑) or against (⇓) the t 1
ψIII (r) = √ iφ ei(kx x+ky y) , (28)
momentum p. This property says that the states of the 2 se
system close to the Dirac point have well defined chi- with s = sgn(E) and s0 = sgn(E − V0 ). The coefficients
rality or helicity. Notice that chirality is not defined in r, a, b and t are determined from the continuity of the
regards to the real spin of the electron (that has not yet wavefunction, which implies that the wavefunction has
appeared in the problem) but to a pseudo-spin variable to obey the conditions ψI (x = 0, y) = ψII (x = 0, y) and
associated with the two components of the wavefunction. ψII (x = D, y) = ψIII (x = D, y). Unlike the Schödinger
The helicity values are good quantum numbers as long equation we only need to match the wavefunction but
as the Hamiltonian (18) is valid. Therefore the existence not its derivative. The transmission through the barrier
of helicity quantum numbers holds only as an asymptotic is obtained from T (φ) = tt∗ and has the form:
property, which is well defined close to the Dirac points
K and K 0 . Either at larger energies or due to the pres- cos2 θ cos2 φ
T (φ) = .
ence of a finite t0 the helicity stops being a good quantum [cos(Dqx ) cos φ cos θ]2 +sin2 (Dqx)(1−ss0 sin φ sin θ)2
number. (29)
 8

an image of the source. This behavior is the same as

FIG. 7 (Color online) Angular behavior of T (φ) for two dif-
that of photons moving in a medium with negative re-
ferent values of V0 : V0 = 200 meV dashed line, V0 = 285 meV
solid line. The remaining parameters are D = 110 nm (top), flection index (Cheianov et al., 2007a). Similar effects
D = 50 nm (bottom) E = 80 meV, kF = 2π/λ, λ = 50 nm. can occur in graphene quantum dots, where the inner
and outer regions contain electrons and holes, respec-
tively (Cserti et al., 2007b). Note that the fact that bar-
This expression does not take into account a contribu- riers do not impede the transmission of normally incident
tion from evanescent waves in region II, which is usually electrons does not preclude the existence of sharp reso-
negligible, unless the chemical potential in region II is at nances, due to the confinement of electrons with a finite
the Dirac energy (see section IV.I). parallel momentum. This leads to the possibility of fab-
Notice that T (φ) = T (−φ) and for values of Dqx sat- ricating quantum dots with potential barriers (Silvestrov
isfying the relation Dqx = nπ, with n an integer, the and Efetov, 2007). Finally, at half-filling, due to disorder
barrier becomes completely transparent since T (φ) = 1, graphene can be divided in electron and hole charge pud-
independently of the value of φ. Also, for normal in- dles. Transport is determined by the transmission across
cidence (φ → 0 and θ → 0) and for any value of Dqx the p-n junctions between these puddles (Cheianov et al.,
one obtains T (0) = 1, and the barrier is again totally 2007b; Shklovskii, 2007). There is a rapid progress in the
transparent. This result is a manifestation of the Klein measurement of transport properties of graphene ribbons
paradox (Calogeracos and Dombey, 1999; Itzykson and with additional top gates that play the role of tunable
Zuber, 2006) and does not occur for non-relativistic elec- potential barriers (Han et al., 2007; Huard et al., 2007;
trons. In this latter case and for normal incidence, the Özyilmaz et al., 2007; Williams et al., 2007).
transmission is always smaller than one. In the limit A magnetic field and potential fluctuations break both
|V0 |  |E|, eq. (29) has the following asymptotic form inversion symmetry of the lattice and time reversal sym-
metry. The combination of these effects break also the
cos2 φ symmetry between the two valleys. The transmission
T (φ) ' . (30)
1 − cos2 (Dqx ) sin2 φ coefficient becomes valley dependent, and, in general,
electrons from different valleys propagate along different
In Fig. 7 we show the angular dependence of T (φ) for paths. This opens the possibility of manipulating the val-
two different values of the potential V0 ; it is clear that ley index (Tworzydlo et al., 2007) (valleytronics) in a way
there are several directions for which the transmission is similar to the control of the spin in mesoscopic devices
one. Similar calculations were done for a graphene bilayer (spintronics). For large magnetic fields, a p-n junction
with its most distinctive behavior being the absence of separates regions with different quantized Hall conduc-
tunneling in the forward (ky = 0) direction. tivities. At the junction, chiral currents can flow at both
The simplest example of a potential barrier is a square edges (Abanin and Levitov, 2007), inducing backscatter-
potential discussed previously. When intervalley scat- ing between the Hall currents at the edges of the sample.
tering and the lack of symmetry between sublattices are The scattering of electrons near the Dirac point
neglected, a potential barrier shows no reflection for elec- by graphene-superconductor junctions differs from An-
trons incident in the normal direction. Even when the dreev scattering process in normal metals (Titov and
barrier separates regions where the Fermi surface is elec- Beenakker, 2006). When the distance between the Fermi
tron like on one side and hole like on the other, a normally energy and the Dirac energy is smaller than the supercon-
incident electron continues propagating as a hole with ducting gap, the superconducting interaction hybridizes
100% efficiency. This phenomenon is another manifesta- quasiparticles from one band with quasiholes in the other.
tion of the chirality of the Dirac electrons within each As in the case of scattering at a p-n junction, the trajec-
valley, which prevents backscattering in general. The tories of the incoming electron and reflected hole (note
transmission and reflection probabilities of electrons at that hole here is meant as in the BCS theory of supercon-
different angles depend on the potential profile along the ductivity) are different from those in similar processes in
barrier. A slowly varying barrier is more efficient in re- metals with only one type of carrier (Bhattacharjee and
flecting electrons at non-zero incident angles (Cheianov Sengupta, 2006).
and Fal’ko, 2006).
Electrons moving through a barrier separating p- and
n-doped graphene, a p-n junction, are transmitted as 2. Confinement and zitterbewegung
holes. The relation between the velocity and the mo-
mentum for a hole is the inverse of that for an elec- Zitterbewegung, or jittery motion of the wavefunction
tron. This implies that, if the momentum parallel to of the Dirac problem, occurs when one tries to confine the
the barrier is conserved, the velocity of the quasipar- Dirac electrons (Itzykson and Zuber, 2006). Localization
ticle is inverted. When the incident electrons emerge of a wavepacket leads, due to the Heisenberg principle, to
from a source, the transmitting holes are focused into uncertainty in the momentum. For a Dirac particle with
 9

FIG. 8 (Color on line) Energy spectrum (in units of t) for a FIG. 9 (Color online)(a) Lattice structure of the bilayer with
graphene ribbon 600a wide, as a function
√ of the momentum the various hopping parameters according to the SWM model.
k along the ribbon (in units of 1/( 3a)), in the presence of The A-sublattices are indicated by the darker spheres. (b)
confining potential with V0 = 1 eV, λ = 180a. Brillouin zone. Adapted from Malard et al., 2007.

zero rest mass, uncertainty in the momentum translates spectrum for a graphene ribbon of width L = 600a, in
into uncertainty in the energy of the particle as well (this the presence of a confining potential,
should be contrasted with the non-relativistic case where h i
the position-momentum uncertainty relation is indepen- V (x) = V0 e−(x−L/2)/λ + e−(L/2−x)/λ , (35)
dent of the energy-time uncertainty relation). Thus, for
where x is the direction of confinement and V0 the
a ultra-relativistic particle, a particle-like state can have
strength of the potential. One can clearly see that in
hole-like states in its time evolution. Consider, for in-
the presence of the confining potential the electron-hole
stance, if one tries to construct a wave packet at some
symmetry is broken and, for V0 > 0, the hole part of
time t = 0, and let us assume, for simplicity, that this
the spectrum is strongly distorted. In particular, for k
packet has a Gaussian shape of width w with momentum
close to the Dirac point, we see that the hole disper-
close to K:
sion is given by: En,σ=−1 (k) ≈ −γn k 2 − ζn k 4 where n
2
e−r /(2w
2
) is a positive integer, and γn < 0 (γn > 0) for n < N ∗
ψ0 (r) = √ eiK·r φ , (31) (n > N ∗ ). Hence, at n = N ∗ the hole effective mass di-
πw
verges (γN ∗ = 0) and, by tuning the chemical potential,
where φ is spinor composed of positive energy states (as- µ, via a back gate, to the hole region of the spectrum
sociated with ψ+,K of (20)). The eigenfunction of the (µ < 0) one should be able to observe an anomaly in the
Dirac equation can be written in terms of the solution Shubnikov-de Haas (SdH) oscillations. This is how zit-
(20) as: terbewegung could manifest itself in magnetotransport.

d2 k X
Z
ψ(r, t) = αa,k ψa,K (k)e−ia(k·r+vF kt) (32)
, C. Bilayer graphene: tight-binding approach
(2π)2 a=±1
The tight-binding model developed for graphite can be
where α±,k are Fourier coefficients. We can rewrite (31)
easily extended to stacks with a finite number of graphene
in terms of (32) by inverse Fourier transform and find
layers. The simplest generalization is a bilayer (McCann
that:
and Fal’ko, 2006). A bilayer is interesting because the
√ 2 2 † IQHE shows anomalies, although different from those ob-
α±,k = πwe−k w /2 ψ±,K (k)φ . (33)
served in a single layer, and also a gap can open between
Notice that the relative weight of positive energy states the conduction and valence band (McCann and Fal’ko,
with respect to negative energy states, |α+ /α− |, given 2006). The bilayer structure, with the AB stacking of
by (20) is one, that is, there are as many positive energy 3D graphite, is shown in Fig.9.
states as negative energy states in a wavepacket. Hence, The tight-binding Hamiltonian for this problem can be
these will cause the wavefunction to be delocalized at written as:
any time t 6= 0. Thus, a wave packet of electron-like X †
Ht.b. = −γ0 (am,i,σ bm,j,σ + h.c.)
states has hole-like components, a result that puzzled
<i,j>
many researchers in the early days of QED (Itzykson and m,σ
Zuber, 2006).
X †
− γ1 (a1,j,σ a2,j,σ + h.c.),
Consider the tight-binding description (Chen et al., j,σ
2007a; Peres et al., 2006b) of Sec. II.A when a poten- X †
tial Vi on site Ri is added to the problem: − γ3 (a1,j,σ b2,j,σ + a†2,j,σ b1,j,σ + h.c.)
j,σ
X X †
He = Vi ni , (34) − γ4 (b1,j,σ b2,j,σ + h.c.), (36)
i j,σ

where ni is the local electronic density. For simplicity, we where am,i,σ (bm,iσ ) annihilates an electron with spin σ,
assume that the confining potential is 1D, that is, that on sublattice A (B), in plane m = 1, 2, at site Ri . Here
Vi vanishes in the bulk but becomes large at the edge we use the graphite nomenclature for the hopping pa-
of the sample. Let us assume a potential that decays rameters: γ0 = t is the in-plane hopping energy and γ1
exponentially away from the edges into the bulk with a (γ1 = t⊥ ≈ 0.4 eV in graphite (Dresselhaus and Dressel-
penetration depth, λ. In Fig. 8 we show the electronic haus, 2002)) is the hopping energy between atom A1 and
 10

FIG. 10 (Color online) Band structure for bilayer graphene FIG. 11 (Color online) Band structure for bilayer graphene
for V = 0 and γ3 = 0. for V 6= 0 and γ3 = 0.

atom A2 (see Fig. 9), and γ3 (γ3 ≈ 0.3 eV in graphite arguments (Mañes et al., 2007). The winding number
(Dresselhaus and Dresselhaus, 2002)) is the hopping en- of a closed curve in the plane around a given point is
ergy between atom A1 (A2 ) and atom B2 (B1 ), and γ4 an integer representing the total number of times that
(γ4 ≈ −0.04 eV in graphite (Dresselhaus and Dressel- the curve travels counterclockwise around the point so
haus, 2002)) that connects B1 and B2 . that the wavefunction remains unaltered. The winding
In the continuum limit, by expanding the momentum number of the point where the two parabolic bands come
close to the K point in the BZ, the Hamiltonian reads, together for γ3 = 0 has winding number +2. The trigo-
X † nal warping term, γ3 , splits it into a Dirac point at k = 0
H= Ψk · H K · Ψk (37) and winding number −1, and three Dirac points at k 6= 0
k and winding numbers +1. An in-plane magnetic field, or
where (ignoring γ4 for the time being): a small rotation of one layer with respect to the other
splits the γ3 = 0 degeneracy into two Dirac points with
−V vF k 0 3γ3 ak ∗
 
winding number +1.
 vF k ∗ −V γ1 0  The term V in (38) breaks the equivalence of the two
HK ≡ , (38)
 0 γ1 V vF k  layers, or, alternatively, inversion symmetry. In this case,
3γ3 ak 0 vF k ∗ V the dispersion relation becomes:

where k = kx + iky is a complex number, and we have 2±,k = V 2 + vF2 k 2 + t2⊥ /2

added V which is here half the shift in electro-chemical q
potential between the two layers (this term will appear ± 4V 2 vF2 k 2 + t2 vF2 k 2 + t4⊥ /4 , (41)
if a potential bias is applied between the layers), and
given rise to the dispersion shown in Fig. 11, and to the
opening of a gap close, but not directly at, the K point.
 
Ψ†k = a†1 (k), a†2 (k), b†1 (k), b†2 (k) (39)
For small momenta, and V  t, the energy of the con-
is a four component spinor. duction band can be expanded:
If V = 0 and γ3 , vF k  γ1 , one can eliminate the
k ≈ V − (2V vF2 k 2 )/t⊥ + (vF4 k 4 )/(2t2⊥ V ) . (42)
high energy states perturbatively and write an effective
Hamiltonian: The dispersion for the valence band can be obtained
2 2
vF k ∗
! by replacing k by −k . The bilayer has a gap at
0 γ1 + 3γ3 ak k 2 ≈ (2V 2 )/vF2 . Notice, therefore, that the gap in the
HK ≡ v2 (k∗ )2 . (40)
F
γ1 + 3γ3 ak 0 biased bilayer system depends on the applied bias and
hence can be measured experimentally (McCann, 2006;
The hopping γ4 leads to a k dependent coupling between McCann and Fal’ko, 2006). The ability to open a gap
the sublattices or a small renormalization of γ1 . The makes bilayer graphene most interesting for technologi-
same role is played by the inequivalence between sublat- cal applications.
tices within a layer.
For γ3 = 0, (40) gives two parabolic bands, k,± ≈
±vF2 k 2 /t⊥ which touch at  = 0 (as shown in Fig.10). D. Epitaxial graphene
The spectrum is electron-hole symmetric. There are two
additional bands which start at ±t⊥ . Within this approx- It has been known for a long time that monolayers
imation, the bilayer is metallic, with a constant density of of graphene could be grown epitaxially on metal sur-
states. The term γ3 changes qualitatively the spectrum faces by using catalytic decomposition of hydrocarbons or
at low energies since it introduces a trigonal distortion, carbon oxide (Campagnoli and Tosatti, 1989; Eizenberg
or warping, of the bands (notice that this trigonal dis- and Blakely, 1979; Oshima and Nagashima, 1997; Shelton
tortion, unlike the one introduced by large momentum et al., 1974; Sinitsyna and Yaminsky, 2006). When such
in (8), occurs at low energies). The electron-hole sym- surfaces are heated, oxygen or hydrogen desorbs, and the
metry is preserved but, instead of two bands touching at carbon atoms form a graphene monolayer. The resulting
k = 0, we obtain three sets of Dirac-like linear bands. graphene structures could reach sizes up to a microm-
One Dirac point is at  = 0 and k = 0, while the three eter, with few defects and were characterized by differ-
other Dirac points, also at  = 0, lie at three equivalent ent surface-science techniques and local scanning probes
points with a finite momentum. The stability of points (Himpsel et al., 1982). For example, graphene grown on
where bands touch can be understood using topological ruthenium has zigzag edges and also ripples associated
 11

with a (10 × 10) reconstruction (Vázquez de Parga et al.,

FIG. 12 (Color online) Sketch of the three inequivalent ori-
2008).
entations of graphene layers with respect to each other.
Graphene can also be formed on the surface of SiC.
Upon heating, the silicon from the top layers desorbs, and
a few layers of graphene are left on the surface (Berger fermions. Local probes reveal a rich structure of terraces
et al., 2004; Bommel et al., 1975; Forbeaux et al., 1998; (Mallet et al., 2007) and interference patterns due to de-
Hass et al., 2007; de Heer et al., 2007; Rollings et al., fects at or below the graphene layers.
2006). The number of layers can be controlled by limit-
ing time or temperature of the heating treatment. The
quality and the number of layers in the samples depends E. Graphene stacks
on the SiC face used for their growth (de Heer et al.,
2007) (the carbon terminated surface produces few layers In stacks with more than one graphene layer, two con-
but with a low mobility, whereas the silicon terminated secutive layers are normally oriented in such a way that
surface produces several layers but with higher mobil- the atoms in one of the two sublattices, An , of the hon-
ity). Epitaxially grown multilayers exhibit SdH oscilla- eycomb structure of one layer are directly above one
tions with a Berry phase shift of π (Berger et al., 2006), half of the atoms in the neighboring layer, sublattice
which is the same as the phase shift for Dirac fermions An±1 . The second set of atoms in one layer sits on
observed in a single layer as well as for some subbands top of the (empty) center of an hexagon in the other
present in multilayer graphene (see further) and graphite layer. The shortest distance between carbon atoms in
(Luk’yanchuk and Kopelevich, 2004). The carbon layer different layers is dAn An±1 = c = 3.4Å. The next dis-
√
directly on top of the substrate is expected to be strongly tance is dAn Bn±1 = c2 + a2 . This is the most common
bonded to it, and it shows no π √ bands (Varchon
√ et al., arrangement of nearest neighbor layers observed in na-
2007). The next layer shows a (6 3 × 6 3) reconstruc- ture, although a stacking order in which all atoms in one
tion due to the substrate, and has graphene properties. layer occupy positions directly above the atoms in the
An alternate route to produce few layers graphene is neighboring layers (hexagonal stacking) has been consid-
based on synthesis from nanodiamonds (Affoune et al., ered theoretically (Charlier et al., 1991) and appears in
2001). graphite intercalated compounds (Dresselhaus and Dres-
Angle resolved photo-emission experiments (ARPES) selhaus, 2002).
show that epitaxial graphene grown on SiC has linearly The relative position of two neighboring layers allows
dispersing quasiparticles (Dirac fermions) (Ohta et al., for two different orientations of the third layer. If we la-
2006), in agreement with the theoretical expectation. bel the positions of the two first atoms as 1 and 2, the
Nevertheless, these experiments show that the electronic third layer can be of type 1, leading to the sequence 121,
properties can change locally in space indicating a certain or it can fill a third position different from 1 and 2 (see
degree of inhomogeneity due to the growth method. Sim- Fig. 12), labeled 3. There are no more inequivalent po-
ilar inhomogeneities due to disorder in the c-axis orien- sitions where a new layer can be placed, so that thicker
tation of graphene planes is observed in graphite. More- stacks can be described in terms of these three orien-
over, graphene grown this way is heavily doped due to the tations. In the most common version of bulk graphite
charge transfer from the substrate to the graphene layer the stacking order is 1212 · · · (Bernal stacking). Re-
(with the chemical potential well above the Dirac point) gions with the stacking 123123 · · · (rhombohedral stack-
and therefore all samples have strong metallic charac- ing) have also been observed in different types of graphite
ter with large electronic mobilities (Berger et al., 2006; (Gasparoux, 1967). Finally, samples with no discernible
de Heer et al., 2007). There is also evidence for strong stacking order (turbostratic graphite) are also commonly
interaction between a substrate and the graphene layer reported.
leading to the appearance of gaps at the Dirac point. Beyond two layers, the stack ordering can be arbitrar-
Indeed, gaps can be generated by the breaking of the ily complex. Simple analytical expressions for the elec-
sublattice symmetry and, as in the case of other carbon tronic bands can be obtained for perfect Bernal ( 1212 · · ·
based systems such as polyacethylene (Su et al., 1979, ) and rhombohedral ( 123123 · · · ) stacking (Guinea
1980), it can lead to soliton-like excitations (Hou et al., et al., 2006). Even if we consider one interlayer hop-
2007; Jackiw and Rebbi, 1976). Multilayer graphene ping, t⊥ = γ1 , the two stacking orders show rather dif-
grown on SiC have also been studied with ARPES (Booth ferent band structures near  = 0. A Bernal stack with
et al., 2008; Ohta et al., 2006) and the results seem to N layers, N even, has N/2 electron like and N/2 hole
agree quite well with band structure calculations (Mat- like parabolic subbands touching at  = 0. When N
tausch and Pankratov, 2007). Spectroscopy measure- is odd, an additional subband with linear (Dirac) dis-
ments also show the transitions associated with Lan- persion emerges. Rhombohedral systems have only two
dau levels (Sadowski et al., 2006), and weak localization subbands that touch at  = 0. These subbands disperse
effects at low magnetic fields, also expected for Dirac as k N , and become surface states localized at the top
 12

layers with an even number of layers and Bernal stacking

FIG. 13 (Color online) Electronic bands of graphene multi-
have inversion symmetry, leading to degeneracies with
layers: top left: biased bilayer; top right: trilayer with Bernal
stacking; bottom left: trilayer with orthorhombic stacking; winding number +2, as in the case of a bilayer. The
bottom right: stack with four layers where the top and bot- trigonal lattice symmetry implies that these points can
tom layers are shifted in energy with respect to the two middle lead, at most, to four Dirac points. In stacks with an odd
layers by +0.1 eV. number of layers, these degeneracies can be completely
removed. The winding number of the degeneracies found
in stacks with N layers and orthorhombic ordering is ±N .
and bottom layer when N → ∞. In this limit, the re- The inclusion of trigonal warping terms will lead to the
maining 2N − 2 subbands of a rhombohedral stack be- existence of many weaker degeneracies near  = 0.
come Dirac like, with the same Fermi velocity as a sin- Furthermore, it is well known that in graphite the
gle graphene layer. The subband structure of a tri-layer planes can be rotated relative each other giving rise to
with the Bernal stacking includes two touching parabolic Moiré patterns that are observed in STM of graphite sur-
bands, and one with Dirac dispersion, combining the fea- faces (Rong and Kuiper, 1993). The graphene layers can
tures of bilayer and monolayer graphene. be rotated relative to each other due to the weak coupling
The low energy bands have different weights on the two between planes that allows for the presence of many dif-
sublattices of each graphene layer. The states at a site ferent orientational states that are quasidegenerate in en-
directly coupled to the neighboring planes are pushed to ergy. For certain angles the graphene layers become com-
energies  ≈ ±t⊥ . The bands near  = 0 are localized mensurate with each other leading to a lowering of the
mostly at the sites without neighbors in the next layers. electronic energy. Such phenomenon is quite similar to
For the Bernal stacking, this feature implies that the den- the commensurate-incommensurate transitions observed
sity of states at  = 0 at sites without nearest neighbors in in certain charge density wave systems or adsorption of
the contiguous layers is finite, while it vanishes linearly gases on graphite (Bak, 1982). This kind of dependence
at the other sites. In stacks with rhombohedral stack- of the electronic structure on the relative rotation angle
ing, all sites have one neighbor in another plane, and between graphene layers leads to what is called super-
the density of states vanishes at  = 0 (Guinea et al., lubricity in graphite (Dienwiebel et al., 2004), namely,
2006). This result is consistent with the well known fact the vanishing of the friction between layers as a function
that only one of the two sublattices at a graphite surface of the angle of rotation. In the case of bilayer graphene,
can be resolved by scanning tunneling microscopy (STM) a rotation by a small commensurate angle leads to the
(Tománek et al., 1987). effective decoupling between layers and the recovery of
As in the case of a bilayer, an inhomogeneous charge the linear Dirac spectrum of the single layer albeit with a
distribution can change the electrostatic potential in the modification on the value of the Fermi velocity (Lopes dos
different layers. For more than two layers, this breaking Santos et al., 2007).
of the equivalence between layers can take place even in
the absence of an applied electric field. It is interest-
ing to note that a gap can open in a stack with Bernal 1. Electronic structure of bulk graphite
ordering and four layers, if the electronic charge at the
two surface layers is different from that at the two in- The tight-binding description of graphene described
ner ones. Systems with a higher number of layers do not earlier can be extended to systems with an infinite num-
show a gap, even in the presence of charge inhomogene- ber of layers. The coupling between layers leads to hop-
ity. Four representative examples are shown in Fig. 13. ping terms between π orbitals in different layers, leading
The band structure analyzed here will be modified by to the so called Slonczewski-Weiss-McClure model (Slon-
the inclusion of the trigonal warping term, γ3 . Experi- czewski and Weiss, 1958). This model describes the band
mental studies of graphene stacks have showed that, with structure of bulk graphite with the Bernal stacking or-
increasing number of layers, the system becomes increas- der in terms of seven parameters, γ0 , γ1 , γ2 , γ3 , γ4 , γ5 and
ingly metallic (concentration of charge carriers at zero en- ∆. The parameter γ0 describes the hopping within each
ergy gradually increases), and there appear several types layer, and it has been considered previously. The cou-
of electron-and-hole-like carries (Morozov et al., 2005; pling between orbitals in atoms that are nearest neigh-
Novoselov et al., 2004). An inhomogeneous charge dis- bors in successive layers is γ1 , which we called t⊥ earlier.
tribution between layers becomes very important in this The parameters γ3 and γ4 describe the hopping between
case, leading to 2D electron and hole systems that occupy orbitals at next nearest neighbors in successive layers and
only a few graphene layers near the surface and can com- were discussed in the case of the bilayer. The coupling be-
pletely dominate transport properties of graphene stacks tween orbitals at next nearest neighbor layers are γ2 and
(Morozov et al., 2005). γ5 . Finally, ∆ is an on site energy which reflects the in-
The degeneracies of the bands at  = 0 can be studied equivalence between the two sublattices in each graphene
using topological arguments (Mañes et al., 2007). Multi- layer once the presence of neighboring layers is taken into
 13

γ0 3.16 eV
FIG. 14 (Color online) Ribbon geometry with zigzag edges.
γ1 0.39 eV
γ2 -0.020 eV
γ3 0.315 eV where c is the lattice constant in the out of plane di-
γ4 -0.044 eV rection, equal to twice the interlayer spacing. The ma-
0
γ5 0.038 eV trix elements of H K can be obtained by replacing kx
∆ -0.008 eV by −kx (other conventions for the unit cell and the ori-
entation of the lattice lead to different phases). Re-
TABLE I Band structure parameters of graphite (Dressel- cent ARPES experiments performed in epitaxially grown
haus and Dresselhaus, 2002). graphene stacks (Berger et al., 2004) confirm the main
features of this model, formulated mainly on the basis of
Fermi surface measurements (McClure, 1957; Soule et al.,
account. The values of these parameters, and their de- 1964). The electronic spectrum of the model can also be
pendence with pressure, or, equivalently, the interatomic calculated in a magnetic field (Ghosal et al., 2007), and
distances, have been extensively studied (Dillon et al., the results are also consistent with STM on graphite sur-
1977; Dresselhaus and Mavroides, 1964; McClure, 1957; faces (Kobayashi et al., 2005; Matsui et al., 2005; Niimi
Nozières, 1958; Soule et al., 1964). A representative set et al., 2006), epitaxially grown graphene stacks (Mallet
of values is shown in Table[I]. It is unknown, however, et al., 2007), and with optical measurements in the in-
how these parameters may vary in graphene stacks with frared range (Li et al., 2006).
a small number of layers.
The unit cell of graphite with Bernal stacking includes
two layers, and two atoms within each layer. The tight- F. Surface states in graphene
binding Hamiltonian described previously can be repre-
sented as a 4 × 4 matrix. In the continuum limit, the So far, we have discussed the basic bulk properties of
two inequivalent corners of the BZ can be treated sepa- graphene. Nevertheless, graphene has very interesting
rately, and the in plane terms can be described by the surface (edge) states that do not occur in other systems.
Dirac equation. The next terms in importance for the A semi-infinite graphene sheet with a zigzag edge has a
low energy electronic spectrum are the nearest neighbor band of zero energy states localized at the surface (Fu-
couplings γ1 and γ3 . The influence of the parameter γ4 jita et al., 1996; Nakada et al., 1996; Wakabayashi et al.,
on the low energy bands is much smaller, as discussed 1999). In section II.H we will discuss the existence of
below. Finally, the fine details of the spectrum of bulk edge states using the Dirac equation. Here will discuss
graphite are determined by ∆, which breaks the electron- the same problem using the tight-binding Hamiltonian.
hole symmetry of the bands preserved by γ0 , γ1 and γ3 . To see why these edge states exist we consider the ribbon
It is usually assumed to be much smaller than the other geometry with zigzag edges shown in Fig. 14. The ribbon
terms. width is such that it has N unit cells in the transverse
We label the two atoms from the unit cell in one layer cross section (y direction). We will assume that the rib-
as 1 and 2, and 3 and 4 correspond to the second layer. bon has infinite length in the longitudinal direction (x
Atoms 2 and 3 are directly on top of each other. Then, direction).
the matrix elements of the Hamiltonian can be written Let us rewrite (5), with t0 = 0, in terms of the integer
as: indices m and n, introduced in Fig. 14, and labeling the
K
unit cells:
H11 = 2γ2 cos(2πkz /c) X
K
H12 = vF (kx + iky ) H = −t [a†σ (m, n)bσ (m, n) + a†σ (m, n)bσ (m − 1, n)
3γ4 a m,n,σ
K
1 + eikz c (kx + iky )


H13 =
2 +a†σ (m, n)bσ (m, n − 1) + h.c.]. (44)
K 3γ3 a
1 + eikz c (kx − iky )


H14 = Given that the ribbon is infinite in the a1 direction one
2
K
H22 = ∆ + 2γ5 cos(2πkz /c) can introduce a Fourier decomposition of the operators
leading to
H23 = γ1 1 + eikz c
K

3γ4 a
Z
K dk X †
1 + eikz c (kx + iky ) [a (k, n)bσ (k, n) + eika a†σ (k, n)bσ (k, n)


H24 = H = −t
2 2π n,σ σ
K
H33 = ∆ + 2γ5 cos(2πkz /c)
K +a†σ (k, n)bσ (k, n − 1) + h.c.] , (45)
H34 = vF (kx + iky )
K
H44 = 2γ2 cos(2πkz /c) (43) where c†σ (k, n) |0i = |c, σ, k, ni, and c = a, b. The one-
 14

particle Hamiltonian can be written as:

Z FIG. 15 (Color online) Sketch of a zigzag termination of a
1p
X graphene bilayer. As discussed in (Castro et al., 2008), there
H = −t dk [(1 + eika ) |a, k, n, σi hb, k, n, σ| is a band of surface states completely localized in the bottom
n,σ layer, and another surface band which alternates between the
+ |a, k, n, σi hb, k, n − 1, σ| + h.c.]. (46) two.

The solution of the Schrödinger equation, H 1p |µ, k, σi =

Eµ,k |µ, k, σi, can be generally expressed as: Although the boundary conditions defined by Eqs. (54)
X and (55) are satisfied for solutions (56) and (57) in the
|µ, k, σi = [α(k, n) |a, k, n, σi + β(k, n) |b, k, n, σi], semi-infinite system, they are not in the ribbon geome-
n
(47) try. In fact, Eqs. (58) and (59) are eigenstates only in
where the coefficients α and β satisfy the following equa- the semi-infinite system. In the graphene ribbon the two
tions: edge states, which come from both sides of the edge, will
overlap with each other. The bonding and anti-bonding
Eµ,k α(k, n) = −t[(1 + eika )β(k, n) + β(k, n − 1)], (48) states formed by the two edge states will then be the rib-
Eµ,k β(k, n) = −t[(1 + e−ika )α(k, n) + α(k, n + 1)].
(49) bon eigenstates (Wakabayashi et al., 1999) (note that at
zero energy there are no other states with which the edge
As the ribbon has a finite width we have to be careful states could hybridize). As bonding and anti-bonding
with the boundary conditions. Since the ribbon only ex- states result in a gap in energy the zero energy flat bands
ists between n = 0 and n = N − 1 at the boundary of edge states will become slightly dispersive, depending
Eqs. (48) and (49) read: on the ribbon width N . The overlap between the two
Eµ,k α(k, 0) = −t(1 + eika )β(k, 0) , (50) edge states is larger as ka approaches 2π/3 and 4π/3.
−ika This means that deviations from zero energy flatness will
Eµ,k β(k, N − 1) = −t(1 + e )α(k, N − 1). (51) be stronger near these points.
The surface (edge) states are solutions of Eqs. (48-51) Edge states in graphene nanoribbons, just as the case
with Eµ,k = 0: of carbon nanotubes, are predicted to be Luttinger liq-
uids, that is, interacting one-dimensional electron sys-
0 = (1 + eika )β(k, n) + β(k, n − 1) , (52) tems (Castro Neto et al., 2006). Hence, clean nanorib-
0 = (1 + e−ika )α(k, n) + α(k, n + 1) , (53) bons must have 1D square root singularities in their den-
0 = β(k, 0) , (54) sity of states (Nakada et al., 1996) that can be probed by
0 = α(k, N − 1) . (55) Raman spectroscopy. Disorder may smooth out these sin-
gularities, however. In the presence of a magnetic field,
Equations (52) and (55) are easily solved giving: when the bulk states are gapped, the edge states are re-
ka sponsible for the transport of spin and charge (Abanin
α(k, n) = [−2 cos(ka/2)]n ei 2 n α(k, 0), (56) and Levitov, 2007; Abanin et al., 2006, 2007a,b).
N−1−n −i ka
2 (N−1−n)
β(k, n) = [−2 cos(ka/2)] e β(k, N −1).
(57)
Let us consider, for simplicity, a semi-infinite system
G. Surface states in graphene stacks
with a single edge. We must require the convergence con-
dition |−2 cos(ka/2)| < 1, in (57) because otherwise the
Single layer graphene can be considered a zero gap
wavefunction would diverge in the semi-infinite graphene
semiconductor, which leads to the extensively studied
sheet. Therefore, the semi-infinite system has edge states
possibility of gap states, at  = 0, as discussed in the
for ka in the region 2π/3 < ka < 4π/3, which corre-
previous section. The most studied such states are those
sponds to 1/3 of the possible momenta. Note that the
localized near a graphene zigzag edge (Fujita et al., 1996;
amplitudes of the edge states are given by,
s Wakabayashi and Sigrist, 2000). It can be shown analyt-
2 −n/λ(k) ically (Castro et al., 2008) that a bilayer zigzag edge,
|α(k, n)| = e , (58) like that shown in Fig. 15, analyzed within the nearest
λ(k)
s neighbor tight-binding approximation described before,
2 −(N −1−n)/λ(k) has two bands of localized states, one completely local-
|β(k, n)| = e , (59) ized in the top layer and indistinguishable from similar
λ(k)
states in single layer graphene, and another band which
where the penetration length is given by: alternates between the two layers. These states, as they
lie at  = 0, have finite amplitudes on one half of the sites
λ(k) = −1/ ln |2 cos(ka/2)|. (60)
only.
It is easily seen that the penetration length diverges when These bands, as in single layer graphene, occupy one
ka approaches the limits of the region ]2π/3, 4π/3[. third of the BZ of a stripe bounded by zigzag edges. They
 15

FIG. 16 (Color online) A piece of a honeycomb lattice dis- FIG. 17 (Color online) Left: Energy spectrum, as calcu-
playing both zigzag and armchair edges. lated from the tight-binding equations, for a nanoribbon with
armchair(top) and zigzag(bottom) edges. The width of the
nanoribbon is N = 200 unit cells. Only fourteen eigenstates
become dispersive in a biased bilayer. As graphite can are depicted. Right: Zoom of the low energy states shown on
be described in terms of effective bilayer systems, one the right.
for each value of the perpendicular momentum, kz , bulk
graphite with a zigzag termination should show one sur-
and armchair edges directly from the Dirac equation.
face band per layer.
This was originally done both with and without a mag-
netic field (Brey and Fertig, 2006a,b; Nakada et al., 1996).
H. The spectrum of graphene nanoribbons
1. Zigzag nanoribbons
The spectrum of graphene nanoribbons depend very
much on the nature of their edges – zigzag or armchair In the geometry of Fig. 16
(Brey and Fertig, 2006a,b; Nakada et al., 1996). In Fig. √ the
unit cell vectors are a1 =
a0 (1, 0) and a2 = a0 1/2, 3/2 , which generate the unit
16 we show a honeycomb lattice having zigzag edges √ √

vectors of the BZ given by b1 = 4π/(a0 3) 3/2, −1/2
along the x direction and armchair edges along the y √
direction. If we choose the ribbon to be infinite in the x and b2 = 4π/(a0 3)(0, 1). From these two vectors
direction we produce a graphene nanoribbon with zigzag we find two inequivalent Dirac points given by K =
edges; conversely choosing the ribbon to be macroscop- (4π/3a0 , 0)√= (K, 0) and K 0 = (−4π/3a0 , 0) = (−K, 0),
ically large along the y but finite in the x direction we with a0 = 3a. The Dirac Hamiltonian around the Dirac
produce a graphene nanoribbon with armchair edges. point K reads in momentum space:
In Fig. 17 we show the fourteen energy levels, calcu- !
lated in the tight-binding approximation, closest to zero 0 px − ipy
HK = vF , (61)
energy for a nanoribbon with zigzag and armchair edges px + ipy 0
and of width N = 200 unit cells. We can see that they
are both metallic, and that the zigzag ribbon presents a and around the K 0 as:
band of zero energy modes that is absent in the armchair !
case. This band at zero energy is the surface states liv- 0 px + ipy
HK 0 = v F . (62)
ing near the edge of the graphene ribbon. More detailed px − ipy 0
ab initio calculations of the spectra of graphene nanorib-
bons show that interaction effects can lead to electronic The wavefunction, in real space, for the sublattice A is
gaps and magnetic states close to the graphene edges, in- given by:
dependent of their nature (Son et al., 2006; Yang et al., 0
·r 0
ΨA (r) = eiK·r ψA (r) + eiK ψA (r) , (63)
2007a,b).
From the experimental point of view, however, and for sublattice B is given by
graphene nanoribbons currently have a high degree of
0
roughness at the edges. Such edge disorder can change ·r 0
ΨB (r) = eiK·r ψB (r) + eiK ψB (r) , (64)
significantly the properties of edge states (Areshkin et al.,
2007; Zhu and Cai, 2007), leading to Anderson local- where ψA and ψB are the components of the spinor wave-
ization, and anomalies in the quantum Hall effect (Cas- 0 0
function of Hamiltonian (61) and ψA and ψB have iden-
tro Neto et al., 2006; Martin and Blanter, 2009) as well as tical meaning but relatively to (62). Let us assume that
Coulomb blockade effects (Sols et al., 2007). Such effects the edges of the nanoribbons are parallel to the x−axis.
have already been observed in lithographically engineered In this case, the translational symmetry guarantees that
graphene nanoribbons (Han et al., 2007; Özyilmaz et al., the spinor wavefunction can be written as:
2007). Furthermore, the problem of edge passivation by !
hydrogen or other elements is not clearly understood ex- φ A (y)
perimentally at this time. Passivation can be modeled in ψ(r) = eikx x , (65)
φB (y)
the tight-binding approach by modifications of the hop-
ping energies or via additional phases in the boundary and a similar equation for the spinor of Hamiltonian (62).
conditions (Kane and Mele, 1997). Theoretical modeling For zigzag edges the boundary conditions at the edge of
of edge passivation indicate that those have a strong ef- the ribbon (located at y = 0 and y = L, where L is the
fect on the electronic properties at the edge of graphene ribbon width) are:
nanoribbons (Barone et al., 2006; Hod et al., 2007).
In what follows we derive the spectrum for both zigzag ΨA (y = L) = 0, ΨB (y = 0) = 0 , (66)
 16

leading to: The solutions of (77) correspond to confined modes in

the graphene ribbon.
0 = eiKx eikx x φA (L) + e−iKx eikx x φ0A (L) , (67) If we apply the same procedure to the Dirac equation
0=e iKx ikx x
e φB (0) + e−iKx eikx x φ0B (0) . (68) around the Dirac point K 0 we obtain a different eigen-
value equation given by:
The boundary conditions (67) and (68) are satisfied for
any x by the choice: kx + z
e−2zL = . (78)
kx − z
φA (L) = φ0A (L) = φB (0) = φ0B (0) = 0. (69)
This equation supports real solutions for z if kx is neg-
We need now to find out the form of the envelope func- ative. Therefore we have edge states for negative values
tions. The eigenfunction around the point K has the kx , with momentum around K 0 . As in the case of K,
form: the system also supports confined modes, given by:
! ! !
0 kx − ∂y φA (y) φA (y) kn
= ˜ , (70) kx = − . (79)
kx + ∂y 0 φB (y) φB (y) tan(kn L)

with ˜ = /vF and  the energy eigenvalue. The eigen- One should note that the eigenvalue equations for K 0 are
problem can be written as two linear differential equa- obtained from those for K by inversion, kx → −kx .
tions of the form: We finally notice that the edge states for zigzag
( nanoribbons are dispersionless (localized in real space)
(kx − ∂y )φB = ˜φA , when t0 = 0. When electron-hole symmetry is broken
(71)
(kx + ∂y )φA = ˜φB . (t0 6= 0) these states become dispersive with a Fermi ve-
locity ve ≈ t0 a (Castro Neto et al., 2006).
Applying the operator (kx + ∂y ) to the first of Eqs. (71)
leads to:
2. Armchair nanoribbons
(−∂y2 + kx2 )φB = ˜2 φB , (72)
Let us now consider an armchair nanoribbon with arm-
with φA given by: chair edges along the y direction. The boundary condi-
1 tions at the edges of the ribbon (located at x = 0 and
φA = (kx − ∂y )φB . (73) x = L, where L is the width of the ribbon):
˜
The solution of (72) has the form: ΨA (x = 0) = ΨB (x = 0) = ΨA (x = L) = ΨB (x = L) = 0 .
(80)
φB = Aezy + Be−zy , (74) Translational symmetry guarantees that the spinor wave-
function of Hamiltonian (61) can be written as:
leading to an eigenenergy ˜2 = kx2 − z 2 . The boundary !
conditions for a zigzag edge require that φA (y = L) = 0 φ A (x)
and φB (y = 0) = 0, leading to: ψ(r) = eiky y , (81)
φB (x)
(
φB (y = 0) = 0 ⇔ A + B = 0 , and a similar equation for the spinor of the Hamiltonian
,
φA (y = L) = 0 ⇔ (kx − z)AezL + (kx + z)Be−zL = 0 (62). The boundary conditions have the form:
(75)
0 = eiky y φA (0) + eiky y φ0A (0) , (82)
which leads to an eigenvalue equation of the form:
0= e φB (0) + eiky y φ0B (0) ,
iky y
(83)
kx − z
e −2zL
= . (76) 0= eiKL eiky y φA (L) + e−iKL eiky y φ0A (L) , (84)
kx + z
0= eiKL eiky y φB (L) + e−iKL eiky y φ0B (L) , (85)
Equation (76) has real solutions for z, whenever kx is pos-
and are satisfied for any y if:
itive; these solutions correspond to surface waves (edge
states) existing near the edge of the graphene ribbon. In φµ (0) + φ0µ (0) = 0 , (86)
section II.F we discussed these states from the point of
view of the tight-binding model. In addition to real solu- and
tions for z, (76) also supports complex ones, of the form
z = ikn , leading to: eiKL φµ (L) + e−iKL φ0µ (L) = 0 , (87)

kn with µ = A, B. It is clear that these boundary conditions

kx = . (77)
tan(kn L) mix states from the two Dirac points. Now we must find
 17

the form of the envelope functions obeying the boundary

FIG. 18 (Color online) SdH oscillations observed in longitu-
conditions (86) and (87). As before, the functions φB
dinal resistivity ρxx of graphene as a function of the charge
and φ0B obey the second order differential equation (72) carrier concentration n. Each peak corresponds to the pop-
(with y replaced by x) and the function φA and φ0A are ulation of one Landau level. Note that the sequence is not
determined from (73). The solutions of (72) have the interrupted when passing through the Dirac point, between
form: electrons and holes. The period of oscillations ∆n = 4B/Φ0 ,
where B is the applied field and Φ0 is the flux quantum.
φB = Aeikn x + Be−ikn x , (88)
φ0B = Ceikn x + De−ikn x . (89)
non-relativistic problem where the cyclotron energy is
Applying the boundary conditions: (86) and (87), one linear in B. This implies that the energy scale associated
obtains: with the Dirac fermions is rather different from the one
0 = A+B +C +D, (90) find in the ordinary 2D electron gas. For instance, for
fields of the order B ≈ 10 T the cyclotron energy in the
0 = Aei(kn +K)L + De−i(kn +K)L
2D electron gas is of the order of 10 K. In contrast, for the
+ Be−i(kn −K)L + Cei(kn −K)L . (91) Dirac fermion problem, for the same fields, the cyclotron
The boundary conditions are satisfied with the choice: energy is of the order of 1, 000 K, that is, two orders of
magnitude bigger. This has strong implications for the
A = −D , B = C = 0, (92) observation of the quantum Hall effect at room temper-
ature (Novoselov et al., 2007). Furthermore, for B = 10
which leads to sin[(kn + K)L] = 0. Therefore the allowed T the Zeeman energy is relatively small, gµB B ≈ 5 K,
values of kn are given by and can be disregarded.
nπ 4π Let us now consider the Dirac equation in more detail.
kn = − , (93) Using the minimal coupling in (19) (i.e., replacing −i∇
L 3a0
by −i∇ + eA/c) we find:
and the eigenenergies are given by:
vF [~σ · (−i∇ + eA/c)] ψ(r) = E ψ(r) , (97)
˜2 = ky2 + kn2 . (94)
in the Landau gauge the generic solution for the wave-
No surface states exist in this case. function has the form ψ(x, y) = eikx φ(y), and the Dirac
equation reads:
I. Dirac fermions in a magnetic field
" #
0 ∂y −k+Bey/c
vF φ(y)=Eφ(y) (98)
,
Let us now consider the problem of a uniform magnetic −∂y −k+Bey/c 0
field B applied perpendicular to the graphene plane 2 .
that can be rewritten as:
We use the Landau gauge: A = B(−y, 0). Notice that " #
the magnetic field introduces a new length scale in the 0 O
problem: ωc φ(ξ) = E φ(ξ) , (99)
O† 0
r
c
`B = , (95) or equivalently:
eB
which is the magnetic length. The only other scale in (Oσ + + O† σ − )φ = (2E/ωc )φ , (100)
the problem is the Fermi-Dirac velocity. Dimensional
analysis shows that the only quantity with dimensions of where σ ± = σx ± iσy , and we have defined the dimen-
energy we can make is vF /`B . In fact, this determines sionless length scale:
the cyclotron frequency of the Dirac fermions: y
√ vF ξ= − `B k , (101)
`B
ωc = 2 (96)
`B
√ and 1D harmonic oscillator operators:
(the 2 factor comes from the quantization of the prob-
1
lem, see below). Eqs. (96)√ and (95) show that the cy- O = √ (∂ξ + ξ) ,
clotron energy scales like B, in clear contrast with the 2
1
O† = √ (−∂ξ + ξ) , (102)
2
2 The problem of transverse magnetic and electric fields can also that obey canonical commutation relations: [O, O† ] = 1.
be solved exactly. See: (Lukose et al., 2007). The number operator is simply: N = O† O.
 18

Firstly, we notice that (100) allows for a solution with

FIG. 19 (Color online) Geometry of Laughlin’s thought ex-
zero energy:
periment with a graphene ribbon: a magnetic field B is ap-
plied normal to the surface of the ribbon, a current I circles
(Oσ + + O† σ − )φ0 = 0 , (103) the loop, generating a Hall voltage VH , and a magnetic flux
Φ.
and since the Hilbert space generated by ~σ is of dimension
2, and the spectrum generated by O† is bounded from
below, we just need to ensure that: FIG. 20 (Color online) Quantum Hall effect in graphene as a
function of charge carrier concentration. The peak at n = 0
Oφ0 = 0 , shows that in high magnetic fields there appears a Landau
level at zero energy where no states exist in zero field. The
σ − φ0 = 0 , (104) field draws electronic states for this level from both conduc-
tion and valence bands. The dashed line indicate plateaus in
in order for (103) to be fulfilled. The obvious zero mode σxy described by Eq. (111).
solution is:

φ0 (ξ) = ψ0 (ξ) ⊗ | ⇓i , (105) states with hole character (µ < 0) from states with elec-
tron character (µ > 0). The components of the resistivity
where | ⇓i indicates the state localized on sublattice A
and conductivity tensors are related by:
and | ⇑i indicates the state localized on sublattice B.
Furthermore, σxx
ρxx = 2 + σ2
,
σxx xy
Oψ0 (ξ) = 0 , (106) σxy
ρxy = 2 2
, (109)
σxx + σxy
is the ground states of the 1D harmonic oscillator. All
the solutions can now be constructed from the zero mode: where σxx (ρxx ) is the longitudinal component and σxy
(ρxy ) is the Hall component of the conductivity (resistiv-
φN,± (ξ) = ψN −1 (ξ) ⊗ | ⇑i ± ψN (ξ) ⊗ | ⇓i ity). When the chemical potential is inside of a region
!
ψN −1 (ξ) of localized states the longitudinal conductivity vanishes,
= , (107) σxx = 0, and hence: ρxx = 0, ρxy = 1/σxy . On the other
±ψN (ξ)
hand, when the chemical potential is a region of delo-
and their energy is given by (McClure, 1956): calized states, when the chemical potential is crossing a
Landau level, we have σxx 6= 0 and σxy varies continu-
√
E± (N ) = ±ωc N , (108) ously (Sheng et al., 2006).
The value of σxy in the region of localized states can
where N = 0, 1, 2, ... is a positive integer, ψN (ξ) is be obtained from Laughlin’s gauge invariance argument
the solution of the 1D Harmonic oscillator (explicitly: (Laughlin, 1981): one imagines making a graphene rib-
ψN (ξ) = 2−N/2 (N !)−1/2 exp{−ξ 2 /2}HN (ξ) where HN (ξ) bon such as the one in Fig. 19 with a magnetic field B
is a Hermite polynomial). The Landau levels at the op- normal through its surface and a current I circling its
posite Dirac point, K’, have exactly the same spectrum loop. Due to the Lorentz force the magnetic field pro-
and hence each Landau level is doubly degenerate. Of duces a Hall voltage VH perpendicular to the field and
particular importance for the Dirac problem discussed current. The circulating current generates a magnetic
here is the existence of a zero energy state N = 0 which flux Φ that threads the loop. The current is given by:
is responsible, as we are going to show, to the anoma-
δE
lies observed in the quantum Hall effect. This particular I=c , (110)
Landau level structure has been observed by many differ- δΦ
ent experimental probes, from Shubnikov-de Haas oscilla- where E is the total energy of the system. The localized
tions in single layer graphene (see Fig. 18) (Zhang et al., states do not respond to changes in Φ, only the delocal-
2005), to infrared spectroscopy (Jiang et al., 2007a), and ized ones. When the flux is changed by a flux quantum
to scanning tunneling spectroscopy (STS) on a graphite δΦ = Φ0 = hc/e the extended states remain the same
surface. by gauge invariance. If the chemical potential is in the
region of localized states, all the extended states below
the chemical potential will be filled both before and after
J. The anomalous integer quantum Hall effect the change of flux by Φ0 . However, during the change of
flux an integer number of states enter the cylinder at one
In the presence of disorder Landau levels get broadened edge and leave at the opposite edge.
and mobility edges appear (Laughlin, 1981). Notice that The question is how many occupied states are trans-
there will be a Landau level at zero energy that separates ferred between edges. Let us consider a naive and as
 19

shown further incorrect calculation in order to show the ductivity is (Gusynin and Sharapov, 2005; Herbut, 2007;
importance of the zero mode in this problem. Each Lan- Peres et al., 2006c,d; Schakel, 1991):
dau level contributes with one state times its degeneracy
g. In the case of graphene we have g = 4 since there I c δE e2
σxy = = = ±2(2N + 1) , (111)
are 2 spin states and 2 Dirac cones. Hence, we would VH VH δΦ h
expect that when the flux changes by one flux quantum without any Hall plateau at N = 0. This amazing result
the change in energy would be δEinc. = ±4N eVH , where has been observed experimentally (Zhang et al., 2005) as
N is an integer. The plus sign applies to electron states shown in Fig.20.
(charge +e) and the minus sign to hole states (charge
−e). Hence, we would conclude that Iinc. = ±4(e2 /h)VH
and hence σxy,inc. = I/VH = ±4N e2 /h, which is the K. Tight-binding model in a magnetic field
naive expectation. The problem with this result is that
when the chemical potential is exactly at half-filling, that In the tight-binding approximation the hopping inte-
is, at the Dirac point, it would predict a Hall plateau at grals are replaced by a Peierls substitution:
N = 0 with σxy,inc. = 0 which is not possible since there R R0 R R0
2π
is a N = 0 Landau level, with extended states at this en- eie R
A·dr
tR,R0 = ei Φ0 R
A·dr
tR,R0 , (112)
ergy. The solution for this paradox is rather simple: be-
cause of the presence of the zero mode which is shared by where tR,R0 represents the hopping integral between the
the two Dirac points, there are exactly 2 × (2N + 1) occu- sites R and R0 , with no field present. The tight-binding
pied states that are transferred from one edge to another. Hamiltonian for a single graphene layer, in a constant
Hence, the change in energy is δE = ±2(2N + 1)eVH for magnetic field perpendicular to the plane, is conveniently
a change of flux of δΦ = hc/e. Therefore, the Hall con- written as,

Φ 1+z Φ Φ z−1
a†σ (m, n)bσ (m, n)+e−iπ Φ0 n a†σ (m, n)bσ (m−1, n−(1−z)/2)+eiπ Φ0 n
X
H = −t [eiπ Φ0 n 2 2
a†σ (m, n)bσ (m, n−z)+h.c.],
m,n,σ
(113)

holding for a graphene stripe with a zigzag (z = 1) and

armchair (z = −1) edges oriented along the x−direction.
Fourier transforming along the x direction gives,

Φ 1+z Φ Φ z−1
a†σ (k, n)bσ (k, n) + e−iπ Φ0 n eika a†σ (k, n)bσ (k, n − (1 − z)/2) + eiπ Φ0 n
X
H = −t [eiπ Φ0 n 2 2
a†σ (k, n)bσ (k, n − z) + h.c.].
k,n,σ

Let us now consider the case of zigzag edges. The lattice-position-state states as:
eigenproblem can be rewritten in terms of Harper’s equa- X
tions (Harper, 1955), and for zigzag edges we obtain |ψ(k)i = (α(k, n)|a; k, n, σi + β(k, n)|b; k, n, σi) .
n,σ
(Rammal, R., 1985):
(116)
Eqs. (114) and (115) hold in the bulk. Considering that
Φ ka the zigzag ribbon has N unit cells along its width, from
Eµ,k α(k, n) = −t[eika/2 2 cos(π n− )β(k, n) n = 0 to n = N −1, the boundary conditions at the edges
Φ0 2
are obtained from Eqs. (114) and (115), and read
+ β(k, n − 1)], (114)  
Φ ka ka
Eµ,k β(k, n) = −t[e−ika/2 2 cos(π n − )α(k, n) Eµ,k α(k, 0) = −teika/2 2 cos β(k, 0) , (117)
Φ0 2 2
+ α(k, n + 1)], (115)  
Φ ka
Eµ,k β(k, N−1)=−2te−ika/2 cos π (N−1)− α(k, N−1).
(118)
Φ0 2
where the coefficients α(k, n) and β(k, n) show up in Similar equations hold for a graphene ribbon with arm-
Hamiltonian’s eigenfunction, |ψ(k)i, written in terms of chair edges.
 20

FIG. 21 (Color online) Fourteen Energy levels of tight- FIG. 23 (Color online) Relative spectral strength of the low
binding electrons in graphene in the presence of a magnetic energy optical transitions between Landau levels in graphene
flux Φ = Φ0 /701, for a finite stripe with N = 200 unit cells. stacks with Bernal ordering and an odd number of layers. The
The bottom panels are zoom in images of the top ones. The applied magnetic field is 1 T. Top left: 3 layers. Top right:
dashed line represents the chemical potential µ. 11 layers. Bottom: 51 layers. The large red circles are the
transitions in a single layer.

FIG. 22 (Color online) Landau levels of the graphene stacks

shown in Fig.13. The applied magnetic field is 1 T. graphene single layer, and only transitions between sub-
bands with Landau level indices M and N such that
|N | = |M ±1| are allowed. The resulting transitions, with
In Fig. 21 we show fourteen energy levels, around zero their respective spectral strengths, are shown in Fig. 23.
energy, for both the zigzag and the armchair cases. The The transitions are grouped into subbands, which give
formation of the Landau levels is signaled by the presence rise to a continuum when the number of layers tends to
of flat energy bands, following the bulk energy spectrum. infinity. In Bernal stacks with an odd number of lay-
From Fig. 21 it is straightforward to obtain the value of ers, the transitions associated to Dirac subbands with
the Hall conductivity in the quantum Hall effect regime. linear dispersion have the largest spectral strength, and
Let us assume that the chemical potential is in between they give a significant contribution to the total absorp-
two Landau levels at positive energies, as represented by tion even if the number of layers is large, NL . 30 − 40
the dashed line in Fig. 21. The Landau level structure (Sadowski et al., 2006).
shows two zero energy modes, one of them is electron-like
(hole-like), since close to the edge of the sample its energy
is shifted upwards (downwards). The other Landau levels
M. Diamagnetism
are doubly degenerate. The determination of the values
for the Hall conductivity is done by counting how many
energy levels (of electron-like nature) are below chemical Back in 1952 Mrozowski (Mrozowski, 1952) stud-
potential. This counting produces the value (2N + 1), ied diamagnetism of polycrystalline graphite and other
with N = 0, 1, 2, . . . (for the case of Fig. 21 one has condensed-matter molecular-ring systems. It was con-
N = 2). From this counting the Hall conductivity is cluded that in such ring systems diamagnetism has two
given, including an extra factor of two accounting for the contributions: (1) diamagnetism from the filled bands of
spin degree of freedom, by electrons; (2) Landau diamagnetism of free electrons and
holes. For graphite the second source of diamagnetism is
e2 e2

1
 by far the largest of the two.
σxy = ±2 (2N + 1) = ±4 N+ . (119) McClure (McClure, 1956) computed diamagnetism of
h h 2
a 2D honeycomb lattice using the theory introduced by
Eq. (119) leads to the anomalous integer quantum Hall Wallace (Wallace, 1947), a calculation he later general-
effect discussed previously, which is the hallmark of single ized to three dimensional graphite (McClure, 1960). For
layer graphene. the honeycomb plane the magnetic susceptibility in units
of emu/g is
 
L. Landau levels in graphene stacks 0.0014 2 µ
χ=− γ0 sech2 , (120)
T 2kB T
The dependence of the Landau level energies on the
Landau index N roughly follows the dispersion relation where µ is the Fermi energy, T the temperature, and
of the bands, as shown in Fig. 22. Note that, in√ a trilayer kB the Boltzmann constant. For graphite the magnetic
with Bernal stacking, two sets of levels have a N depen- susceptibility is anisotropic and the difference between
dence, while the energies of other two depend linearly on the susceptibility parallel to the principal axis and that
N . In an infinite rhombohedral stack, the Landau lev- perpendicular to the principal axis is -21.5×10−6 emu/g.
els remain discrete and quasi-2D (Guinea et al., 2006). The susceptibility perpendicular to the principal axis is
Note that, even in an infinite stack with the Bernal struc- about the free-atom susceptibility of -0.5×10−6 emu/g.
ture, the Landau level closest to E = 0 forms a band In the 2D model the susceptibility turns out to be large
which does not overlap with the other Landau levels, due to the presence of fast moving electrons, with a ve-
leading to the possibility of a 3D integer quantum Hall ef- locity of the order of vF ' 106 m/s, which in turn is a
fect (Bernevig et al., 2007; Kopelevich et al., 2006, 2003; consequence of the large value of the hopping parameter
Luk’yanchuk and Kopelevich, 2004). γ0 . In fact the susceptibility turns out to be proportional
The optical transitions between Landau levels can also to the square of γ0 . Later Sharma et al. extended the cal-
be calculated. The selection rules are the same as for a culation of McClure for graphite by including the effect
 21

of trigonal warping and showed that the low temperature 1979):

diamagnetism increases (Sharma et al., 1974). gv gs 2 2
Safran and DiSalvo (Safran and DiSalvo, 1979), in- χ(F ) = − e γ0 δ(F ) . (123)
6π
terested in the susceptibility of graphite intercalation
compounds, recalculated, in a tight-binding model, the
susceptibility perpendicular to a graphite plane using N. Spin orbit coupling
Fukuyama’s theory (Fukuyama, 1971), which includes
interband transitions. The results agree with those pub- Spin-orbit coupling describes a process in which an
lished by McClure (McClure, 1956). Later, Safran com- electron changes simultaneously its spin and angular mo-
puted the susceptibility of a graphene bilayer showing mentum or, in general, moves from one orbital wavefunc-
that this system is diamagnetic at small values of the tion to another. The mixing of the spin and the orbital
Fermi energy, but there appears a paramagnetic peak motion is a relativistic effect, which can be derived from
when the Fermi energy is of the order of the interlayer Dirac’s model of the electron. It is large in heavy ions,
coupling (Safran, 1984). where the average velocity of the electrons is higher. Car-
The magnetic susceptibility of other carbon based ma- bon is a light atom, and the spin orbit interaction is ex-
terials, as carbon nanotubes and C60 molecular solids pected to be weak.
was measured (Heremans et al., 1994) showing a dia- The symmetries of the spin orbit interaction, however,
magnetic response at finite magnetic fields different from allow the formation of a gap at the Dirac points in clean
that of graphite. The study of the magnetic response graphene. The spin orbit interaction leads to a spin de-
of nanographite ribbons with both zig-zag and arm-chair pendent shift of the orbitals, which is of a different sign
edges was done by Wakabayashi et al. using a numeri- for the two sublattices, acting as an effective mass within
cal differentiation of the free energy (Wakabayashi et al., each Dirac point (Dresselhaus and Dresselhaus, 1965;
1999). From these two systems, the zig-zag edge state Kane and Mele, 2005; Wang and Chakraborty, 2007a).
is of particular interest since the system supports edge The appearance of this gap leads to a non trivial spin
states even in the presence of a magnetic field, leading to Hall conductance, in similar way to other models which
very high density of states near the edge of the ribbon. study the parity anomaly in relativistic field theory in
The high temperature response of these nanoribbons was (2+1) dimensions (Haldane, 1988). When the inversion
found to be diamagnetic whereas the low temperature symmetry of the honeycomb lattice is broken, either be-
susceptibility is paramagnetic. cause the graphene layer is curved or because an external
The Dirac-like nature of the electronic quasiparti- electric field is applied (Rashba interaction) additional
cles in graphene led (Giesbers et al., 2007) to consider terms, which modulate the nearest neighbor hopping,
in general the problem of the diamagnetism of nodal are induced (Ando, 2000). The intrinsic and extrinsic
fermions and Nakamura to study the orbital magnetism spin orbit interactions can be written as (Dresselhaus and
of Dirac fermions in weak magnetic fields(Nakamura, Dresselhaus, 1965; Kane and Mele, 2005):
2007). Koshino and Ando considered the diamagnetism Z
of disordered graphene in the self consistent Born approx- HSO;int≡∆so d2 rΨ̂† (r)ŝz σ̂z τ̂z Ψ̂(r) ,
imation, with a disorder potential of the form U (r) = Z
1ui δ(r − R) (Koshino and Ando, 2007). At the neu- HSO;ext≡λR d2 rΨ̂† (r)(−ŝx σ̂y+ŝy σ̂x τ̂z )Ψ̂(r) , (124)
trality point and zero temperature the susceptibility of
disordered graphene is given by where σ̂ and τ̂ are Pauli matrices which describe the sub-
lattice and valley degrees of freedom, and ŝ are Pauli ma-
gv gs 2 2 2W trices acting on actual spin space. ∆so is the spin-orbit
χ(0) = − e γ0 , (121)
3π 2 Γ0 coupling and λR is the Rashba coupling.
where gs = gv = 2 is the spin and valley degeneracies, W The magnitude of the spin orbit coupling in graphene
is a dimensionless parameter for the disorder strength, can be inferred from the known spin orbit coupling in
defined as W = ni u2i /4πγ02 , ni the impurity concentra- the carbon atom. This coupling allows for transitions
tion, and Γ0 is given by Γ0 = c exp[−1/(2W )] with c between the pz and px and py orbitals. An electric field
a parameter defining a cut-off function used in the the- induces also transitions between the pz and s orbitals.
ory. At finite Fermi energy F and zero temperature the These intra-atomic processes mix the π and σ bands in
magnetic susceptibility is given by graphene. Using second order perturbation theory, one
obtains a coupling between the low energy states in the
gv gs 2 2 W π band. Tight-binding (Huertas-Hernando et al., 2006;
χ(F ) = − e γ0 . (122) Zarea and Sandler, 2007) and band structure calculations
3π |F |
(Min et al., 2006; Yao et al., 2007) give estimates for
In the clean limit the susceptibility is given by (Koshino the intrinsic and extrinsic spin-orbit interactions in the
and Ando, 2007; McClure, 1956; Safran and DiSalvo, range 0.01 − 0.2 K, and hence much smaller than the
 22

at low energies (low temperatures) and long wavelengths.

FIG. 24 (Color online) Suspended graphene sheet. (a)
It is instructive to understand how these modes appear
Bright-field transmission-electron-microscope image of a
graphene membrane. Its central part (homogeneous and fea- from the point of view of elasticity theory. Consider, for
tureless region) is monolayer graphene. (b) Despite being instance, a graphene sheet in 3D and let us parameterize
only one atom thick, graphene remains a perfect crystal as the position of the sheet relative of a fixed coordinate
this atomic resolution image shows. The image is obtained frame in terms of the in-plane vector r and the height
in a scanning transmission electron microscope. The visible variable h(r) so that a position in the graphene is given
periodicity is given by the lattice of benzene rings. by the vector R = (r, h(r)). The unit vector normal to
the surface is given by:
other energy scales in the problem (kinetic, interaction, z − ∇h
N= p , (125)
and disorder). 1 + (∇h)2
where ∇ = (∂x , ∂y ) is the 2D gradient operator, and
III. FLEXURAL PHONONS, ELASTICITY, AND z is the unit vector in the third direction. In a flat
CRUMPLING graphene configuration all the normal vectors are aligned
and therefore ∇ · N = 0. Deviations from the flat config-
Graphite, in the Bernal stacking configuration, is a lay- uration requires misalignment of the normal vectors and
ered crystalline solid with 4 atoms per unit cell. Its basic costs elastic energy. This elastic energy can be written
structure is essentially a repetition of the bilayer struc- as:
ture discussed earlier. The coupling between the lay- Z Z
κ 2 κ
2
ers, as we discussed, is weak and, therefore, graphene E0 = d2 r (∇ · N) ≈ d 2 r ∇2 h (126)
2 2
has been always the starting point for the discussion of
phonons in graphite (Wirtz and Rubio, 2004). Graphene where κ is the bending stiffness of graphene, and the
has two atoms per unit cell and if we consider graphene expression in terms of h(r) is valid for smooth distortions
as strictly 2D it should have 2 acoustic modes (with dis- of the graphene sheet ((∇h)2  1). The energy (126) is
persion ωac (k) ∝ k as k → 0) and 2 optical modes (with valid in the absence of a surface tension or a substrate
dispersion ωop (k) ∝ constant, as k → 0) solely due to which break the rotational and translational symmetry
the in-plane translation and stretching of the graphene of the problem, respectively. In the presence of tension
lattice. Nevertheless, graphene exists in the 3D space there is an energy cost for solid rotations of the graphene
and hence the atoms can oscillate out-of-plane leading to sheet (∇h 6= 0) and hence a new term has to be added
2 out-of-plane phonons (one acoustic and another opti- to the elastic energy:
cal) called flexural modes. The acoustic flexural mode Z
γ 2
has dispersion ωflex (k) ∝ k 2 as k → 0 which represents ET = d2 r (∇h) , (127)
the translation of the whole graphene plane (essentially 2
a one atom thick membrane) in the perpendicular di- where γ is the interfacial stiffness. A substrate, described
rection (free particle motion). The optical flexural mode by a height variable s(r), pins the graphene sheet through
represents the out-of-phase out-of-plane oscillation of the van der Waals and other electrostatic potentials so that
neighboring atoms. In first approximation, if we neglect the graphene configuration tries to follow the substrate
the coupling between graphene planes, graphite has es- h(r) ∼ s(r). Deviations from this configuration cost ex-
sentially the same phonon modes, albeit they are degen- tra elastic energy that can be approximated by a har-
erate. The coupling between planes has two main effects: monic potential (Swain and Andelman, 1999):
(1) it lifts the degeneracy of the phonon modes, and (2) Z
v 2
leads to a strong suppression of the energy of the flexural ES = d2 r (s(r) − h(r)) , (128)
modes. Theoretically, flexural modes should become or- 2
dinary acoustic and optical modes in a fully covalent 3D where v characterizes the strength of the interaction po-
solid, but in practice, the flexural modes survive due to tential between substrate and graphene.
the fact the planes are coupled by weak van der Waals- Firstly, let us consider the free floating graphene prob-
like forces. These modes have been measured experimen- lem (126) that we can rewrite in momentum space as:
tally in graphite (Wirtz and Rubio, 2004). Graphene
κX 4
can also be obtained as a suspended membrane, that is, E0 = k h−k hk . (129)
2
without a substrate, being supported only by a scaffold k
or bridging micron-scale gaps (Bunch et al., 2007; Meyer We now canonically quantize the problem by introducing
et al., 2007). Figure 24 shows a suspended graphene sheet a momentum operator Pk that has the following commu-
and an atomic resolution image of its crystal lattice. tator with hk :
Because the flexural modes disperse like k 2 they dom-
inate the behavior of structural fluctuations in graphene [hk , Pk0 ] = iδk,k0 , (130)
 23

and write the Hamiltonian as: is of the order of kmin ∼ 2π/L. Using kmin as a lower cut-
off in the integral (136) we find:
X  P−k Pk κk 4

H= + h−k hk , (131)  
2σ 2 π 1
k Nph = 2 ln , (137)
2 2
LT 1 − e−LT /L
where σ is graphene’s 2D mass density. From the Heisen-
berg equations of motion for the operators it is trivial where
to find that hk oscillates harmonically with a frequency
2π  κ 1/4
given by: LT = √ , (138)
kB T σ
 κ 1/2
ωflex (k) = k2 , (132) is the thermal wavelength of the flexural modes. Notice
σ
that that when L  LT the number of flexural phonons
which is the long wavelength dispersion of the flexural in (137) diverges logarithmically with the size of the sys-
modes. In the presence of tension it is easy to see that tem:
the dispersion is modified to:  
2π L
r Nph ≈ 2 ln , (139)
κ 2 γ LT LT
ω(k) = k k + , (133)
σ σ
indicating that the system cannot be structurally ordered
indicating that the dispersion of the flexural modes be- at any finite temperature. This is nothing but the crum-
comes linear in k, as k → 0, under tension. That is what pling instability of soft membranes. For L  LT one
happens in graphite where the interaction between layers finds that Nph goes to zero exponentially with the size
breaks the rotational symmetry of the graphene layers. of the system indicating that systems with finite size can
Eq. (132) also allows us to relate the bending energy be flat at sufficiently low temperatures. Notice that for
of graphene with the Young modulus, Y , of graphite. κ ≈ 1 eV, ρ ≈ 2200 kg/m3 , t = 3.4 Å (σ ≈ 7.5 × 10−7
The fundamental resonance frequency of a macroscopic kg/m2 ), and T ≈ 300 K, we find LT ≈ 1 Å indicat-
graphite sample of thickness t is given by (Bunch et al., ing that free floating graphene should always crumple at
2007): room temperature due to thermal fluctuations associated
with flexural phonons. Nevertheless, the previous discus-
 1/2
Y sion only involves the harmonic (quadratic part) of the
ν(k) = t k2 , (134) problem. Non-linear effects such as large bending defor-
ρ
mations (Peliti and Leibler, 1985), the coupling between
where ρ = σ/t is the 3D mass density. Assuming that flexural and in-plane modes (or phonon-phonon interac-
(134) works down to the single plane level, that is, when tions (Bonini et al., 2007)) and the presence of topologi-
t is the distance between planes, we find: cal defects (Nelson, D.R. and Peliti, L., 1987) can lead to
strong renormalizations of the bending rigidity, driving
κ = Y t3 , (135) the system toward a flat phase at low temperatures. This
which provides a simple relationship between the bending situation has been confirmed in numerical simulations of
stiffness and the Young modulus. Given that Y ≈ 1012 free graphene sheets (Abedpour et al., 2007).
N/m and t ≈ 3.4 Å we find, κ ≈ 1 eV. This result is in The situation is rather different if the system is un-
good agreement with ab initio calculations of the bend- der tension or in the presence of a substrate or scaffold
ing rigidity (Tu and Ou-Yang, 2002) and experiments in that can hold the graphene sheet. In fact, static rippling
graphene resonators (Bunch et al., 2007). of graphene flakes suspended on scaffolds have been ob-
The problem of structural order of a “free floating” served for single layer as well as bilayers (Meyer et al.,
graphene sheet can be fully understood from the exis- 2007). In this case the dispersion, in accordance with
tence of the flexural modes. Consider, for instance, the (133), is at least linear in k, and the integral in (136)
number of flexural modes per unit of area at a certain converges in the infrared (k → 0) indicating that the
temperature T : number of flexural phonons is finite and graphene does
not crumple. We should notice that these thermal fluctu-
Z ∞
d2 k ations are dynamic and hence average to zero over time,
Z
1 k
Nph = nk = dk √ (136) therefore, the graphene sheet is expected to be flat under
(2π)2 2π 0 e β κ/σk2
−1
these circumstances. Obviously, in the presence of a sub-
where nk is the Bose-Einstein occupation number (β = strate or scaffold described by (128) static deformations
1/(kB T )). For T 6= 0 the above integral is logarithmically of the graphene sheet are allowed. Also, hydrocarbon
divergent in the infrared (k → 0) indicating a divergent molecules that are often present on top of free hanging
number of phonons in the thermodynamic limit. For a graphene membranes could quench flexural fluctuations
system with finite size L the smallest possible wave vector making them static.
 24

Finally, one should notice that in the presence of a 2007; Shytov et al., 2007). Its solution has many of
metallic gate the electron-electron interactions lead to the features of the 3D relativistic hydrogen atom prob-
the coupling of the phonon modes to the electronic ex- lem (Baym, 1969). Just as in the case of the 3D prob-
citations in the gate. This coupling could be partially lem the nature of the eigenfunctions depends strongly on
responsible to the damping of the phonon modes due to graphene’s dimensionless coupling constant:
dissipative effects (Seoánez et al., 2007) as observed in
Ze2
graphene resonators (Bunch et al., 2007). g= . (143)
 0 vF
Notice, therefore, that the coupling constant can be var-
IV. DISORDER IN GRAPHENE ied by either changing the charge of the impurity, Z,
or modifying the dielectric environment and changing
Graphene is a remarkable material from the electronic 0 . For g < gc = 1/2 the solutions of this problem are
point of view. Because of the robustness and specificity given in terms of Coulomb wavefunctions with logarith-
of the sigma bonding, it is very hard for alien atoms to mic phase shifts. The local density of states (LDOS) is
replace the carbon atoms in the honeycomb lattice. This affected close to the impurity due the electron-hole asym-
is one of the reasons why the electron mean free path metry generated by the Coulomb potential. The local
in graphene can be so long, reaching up to one microm- charge density decays like 1/r3 plus fast oscillations of
eter in the existing samples. Nevertheless, graphene is the order of the lattice spacing (in the continuum limit
not immune to disorder and its electronic properties are this would give rise to a Dirac delta function for the den-
controlled by extrinsic as well as intrinsic effects that are sity (Kolezhuk et al., 2006)). Just like in 3D QED, the
unique to this system. Among the intrinsic sources of 2D problem becomes unstable for g > gc = 1/2 leading to
disorder we can highlight: surface ripples and topologi- super-critical behavior and the so-called fall of electron
cal defects. Extrinsic disorder can come about in many to the center. In this case the LDOS is strongly affected
different forms: adatoms, vacancies, charges on top of by the presence of the Coulomb impurity with the ap-
graphene or in the substrate, and extended defects such pearance of bound states outside the band and scattering
as cracks and edges. resonances within the band (Pereira et al., 2007) and the
It is easy to see that from the point of view of single local electronic density decays monotonically like 1/r2 at
electron physics (that is, terms that can be added to (5)), large distances.
there are two main terms that disorder couples to. The It has been argued that without high vacuum environ-
first one is a local change in the single site energy, ment these Coulomb effects can be strongly suppressed
X  †  by large effective dielectric constants due to the presence
Hdd = Vi ai ai + b†i bi , (140) of a nanometer thin layer of absorbed water. In fact, ex-
i
periments in ultra-high vacuum conditions (Chen et al.,
where Vi is the strength of the disorder potential on site 2008) display strong scattering features in the transport
Ri , which is diagonal in the sublattice indices and hence, that can be associated to charge impurities. Screening
from the point of view of the Dirac Hamiltonian (18), can effects that affect the strength and range of the Coulomb
be written as: interaction, are rather non-trivial in graphene (Fogler
Z X et al., 2007; Shklovskii, 2007) and, therefore, important
Hdd = d2 r Va (r)Ψ̂†a (r)Ψ̂a (r) , (141) for the interpretation of transport data (Bardarson et al.,
a=1,2 2007; Lewenkopf et al., 2008; Nomura and MacDonald,
which acts as a chemical potential shift for the Dirac 2006; San-Jose et al., 2007).
fermions, that is, shifts locally the Dirac point. Another type of disorder is the one that changes the
Because of the vanishing of the density of states in sin- distance or angles between the pz orbitals. In this case,
gle layer graphene, and by consequence the lack of elec- the hopping energies between different sites are modified
trostatic screening, charge potentials may be rather im- leading to a new term to the original Hamiltonian (5):
portant in determining the spectroscopic and transport X n (ab)  † 
Hod = δtij ai bj + h.c.
properties (Adam et al., 2007; Ando, 2006b; Novikov,
i,j
2007a). Of particular importance is the Coulomb impu-  o
(aa)
rity problem where, + δtij a†i aj + b†i bj , (144)
2
e 1 or in Fourier space:
Va (r) = , (142)
0 r X † X (ab) 0 ~ 0
Hod = ak bk0 δti ei(k−k )·Ri −iδaa ·k + h.c.
where 0 is the dielectric constant of the medium. The k,k0 i,~
δab
solution of the Dirac equation for the Coulomb potential  X 0 0
~
+ a†k ak0+b†k bk0
(aa)
in 2D can be studied analytically (Biswas et al., 2007; Di- δti ei(k−k )·Ri −iδab ·k ,(145)
Vincenzo and Mele, 1984; Novikov, 2007b; Pereira et al., i,~
δaa
 25

(ab) (aa)
where δtij (δtij ) is the change of the hopping en-
FIG. 25 Bending of the surface of graphene by a radius R
ergy between orbitals on lattice sites Ri and Rj on the and its effect on the pz orbitals.
same (different) sublattices (we have written Rj = Ri + ~δ
where ~δab is the nearest neighbor vector, and ~δaa is the
next nearest neighbor vector). Following the procedure A. Ripples
of Sec. II.B we project out the Fourier components of the
operators close to the K and K’ points of the BZ using Graphene is a one atom thick system, the extreme
(17). If we assume that δtij is smooth over the lattice case of a soft membrane. Hence, just like soft mem-
spacing scale, that is, it does not have an Fourier com- branes, it is subject to distortions of its structure either
ponent with momentum K − K0 (so the two Dirac cones due to thermal fluctuations (as we discussed in Sec. III)
are not coupled by disorder), we can rewrite (145) in real or interaction with a substrate, scaffold, and absorbands
space as: (Swain and Andelman, 1999). In the first case the fluc-
Z tuations are time dependent (although with time scales
much longer than the electronic ones), while in the sec-
h
Hod = d2 r A(r)a†1 (r)b1 (r) + h.c.
ond case the distortions act as quenched disorder. In
both cases, the disorder comes about because of the mod-
 i
+ φ(r) a†1 (r)a1 (r) + b†1 (r)b1 (r) , (146)
ification of the distance and relative angle between the
carbon atoms due to the bending of the graphene sheet.
with a similar expression for the cone 2 but with A re- This type of off-diagonal disorder does not exist in ordi-
placed by A∗ , where, nary 3D solids, or even in quasi-1D or quasi-2D systems,
X ~ where atomic chains and atomic planes, respectively, are
A(r) = δt(ab) (r)e−iδab ·K , (147) embedded in a 3D crystalline structure. In fact, graphene
~
δab is also very different from other soft membranes because
~ it is (semi) metallic, while previously studied membranes
X
φ(r) = δt(aa) (r)e−iδaa ·K . (148)
~
were insulators.
δaa
The problem of the bending of graphitic systems and
Notice that whereas φ(r) = φ∗ (r), because of the in- its effect on the hybridization of the π orbitals has been
version symmetry of the two triangular sublattices that studied a great deal in the context of classical minimal
make up the honeycomb lattice, A is complex because of surfaces and applied to fullerenes and carbon nanotubes
lack of inversion symmetry for nearest neighbor hopping. (Kane and Mele, 1997; Tersoff, 1992; Tu and Ou-Yang,
Hence, 2002; Xin et al., 2000; Zhong-can et al., 1997). In order
to understand the effect of bending on graphene, con-
A(r) = Ax (r) + iAy (r) . (149) sider the situation shown in Fig.25. The bending of the
graphene sheet has three main effects: the decrease of the
In terms of the Dirac Hamiltonian (18) we can rewrite distance between carbon atoms, a rotation of the pZ or-
(146) as: bitals (compression or dilation of the lattice are energet-
ically costly due to the large spring constant of graphene
≈ 57 eV/Å2 (Xin et al., 2000)), and a re-hybridization
Z h
Hod = d2 r Ψ̂†1 (r)σ · A(r) ~ Ψ̂1 (r)
between π and σ orbitals (Kim and Neto, 2008). Bend-
i ing by a radius R decreases the distance between the
+ φ(r)Ψ̂†1 (r)Ψ̂1 (r) , (150) orbitals from ` to d = 2R sin[`/(2R)] ≈ ` − `3 /(24R2 )
for R  `. The decrease in the distance between the or-
where A ~ = (Ax , Ay ). This result shows that changes in bitals increases the overlap between the two lobes of the
0
the hopping amplitude lead to the appearance of vector, pZ orbital (Harrison, 1980): Vppa ≈ Vppa [1 + `2 /(12R2 )],
~ and scalar, Φ, potentials in the Dirac Hamiltonian.
A, 0
where a = π, σ, and Vppa is the overlap for a flat graphene
The presence of a vector potential in the problem indi- sheet. The rotation of the pZ orbitals can be understood
cates that an effective magnetic field B~ = c/(evF )∇ × A ~ within the Slater-Koster formalism, namely, the rotation
should also be present, naively implying a broken time re- can be decomposed into a pz − pz (π bond) plus a px − px
versal symmetry, although the original problem was time (σ bond) hybridization with energies Vppπ and Vppσ ,
reversal invariant. This broken time reversal symmetry respectively (Harrison, 1980): V (θ) = Vppπ cos2 (θ) −
is not real since (150) is the Hamiltonian around only Vppσ sin2 (θ) ≈ Vppπ −(Vppπ +Vppσ )(`/(2R))2 , leading to a
one of the Dirac cones. The second Dirac cone is re- decrease in the overlap. Furthermore, the rotation leads
lated to the first by time reversal symmetry indicating to re-hybridization between π and σ orbitals leading to
that the effective magnetic field is reversed in the second a further shift in energy of the order of (Kim and Neto,
2 2
cone. Therefore, there is no global broken symmetry but 2008): δπ ≈ (Vspσ + Vppσ )/(π − a ).
a compensation between the two cones. In the presence of a substrate, as we discussed in
 26

[Link], elasticity theory predicts that graphene can be

FIG. 26 (Color online) Sketch of the boundary conditions
expected to follow the substrate in a smooth way. In-
associated to a disclination (pentagon) in the honeycomb lat-
deed, by minimizing the elastic energy (126), (127), and tice.
(128) with respect to the height h we get (Swain and
Andelman, 1999):
induce long range deformations, which modify the elec-
κ∇4 h(r) − γ∇2 h(r) + vh(r) = vs(r) , (151) tron trajectories.
that can be solved by Fourier transform: Let us consider first a disclination. This defect is equiv-
alent to the deletion or inclusion of a wedge in the lattice.
s(k) The simplest one in the honeycomb lattice is the absence
h(k) = , (152)
1 + (`t k)2 + (`c k)4 of a 60◦ wedge. The resulting edges can be glued in such
where a way that all sites remain three-fold coordinated. The
 γ 1/2 honeycomb lattice is recovered everywhere, except at the
`t = , apex of the wedge, where a fivefold ring, a pentagon, is
v formed. One can imagine a situation where the nearest
 κ 1/4
`c = . (153) neighbor hoppings are unchanged. Nevertheless, the ex-
v istence of a pentagon implies that the two sublattices of
Eq. (152) gives the height configuration in terms of the the honeycomb structure can no longer be defined. A
substrate profile, and `t and `c provide the length scales trajectory around the pentagon after a closed circuit has
for elastic distortion of graphene on a substrate. Hence, to change the sublattice index.
disorder in the substrate translates into disorder in the The boundary conditions imposed at the edges of a
graphene sheet (albeit restricted by elastic constraints). disclination are sketched in Fig. 26, showing the identi-
This picture has been confirmed by STM measurements fication of sites from different sublattices. In addition,
on graphene (Ishigami et al., 2007; Stolyarova et al., the wavefunctions at the K and K 0 points are exchanged
2007) in which strong correlations were found between when moving around the pentagon.
the roughness of the substrate and the graphene topog- Far away from the defect, a slow rotation of the com-
raphy. Ab initio band structure calculations also give ponents of the spinorial wavefunction can be described
support to this scenario (Dharma-wardana, 2007). by a gauge field which acts on the valley and sublattice
The connection between the ripples and the electronic indices. This gauge field is technically non-abelian, al-
problem comes from the relation between the height field though a transformation can be defined which makes the
h(r) and the local curvature of the graphene sheet, R: resulting Dirac equation equivalent to one with an effec-
tive abelian gauge field. The final continuum equation
2 gives a reasonable description of the electronic spectrum
≈ ∇2 h(r) , (154)
R(r) of fullerenes of different sizes, and other structures which
and, hence we see that due to bending the electrons are contain pentagons (Kolesnikov and Osipov, 2004, 2006;
subject to a potential which depends on the structure of Lammert and Crespi, 2004; LeClair, 2000; Osipov et al.,
a graphene sheet (Kim and Neto, 2008): 2003). It is easy to see that an heptagon leads to the
opposite effective field.
V (r) ≈ V 0 − α (∇2 h(r))2 , (155) An in-plane dislocation, that is, the inclusion of a semi-
infinite row of sites, can be considered as induced by
where α (α ≈ 10 eV Å2 ) is the constant that depends on
a pentagon and a heptagon together. The non-abelian
microscopic details. The conclusion from (155) is that the
field described above is canceled away from the core. A
Dirac fermions are scattered by ripples of the graphene
closed path implies a shift by one (or more) lattice spac-
sheet through a potential which is proportional to the
ings. The wavefunctions at the K and K 0 points acquire
square of the local curvature. The coupling between ge-
phases, e±2πi/3 , under a translation by one lattice unit.
ometry and electron propagation is unique to graphene,
Hence, the description of a dislocation in the continuum
and results in additional scattering and resistivity due to
limit requires an (abelian) vortex of charge ±2π/3 at its
ripples (Katsnelson and Geim, 2008).
core. Dislocations separated over distances of the order
of d lead to an effective flux through an area of perimeter
B. Topological lattice defects l of the order of (Morpurgo and Guinea, 2006):
d
Structural defects of the honeycomb lattice like pen- Φ∼ (156)
kF l2
tagons, heptagons and their combinations such as Stone-
Wales defect (a combination of two pentagon-heptagon where kF is the Fermi vector of the electrons.
pairs) are also possible in graphene and can lead to scat- In general, a local rotation of the axes of the honey-
tering (Cortijo and Vozmediano, 2007a,b). These defects comb lattice induces changes in the hopping which lead
 27

to mixing of the K and K 0 wavefunctions, leading to a

FIG. 27 (Color online) Sketch of a rough graphene surface.
gauge field like the one induced by a pentagon (González
The full line gives the boundary beyond which the lattice can
et al., 2001). Graphene samples with disclinations and be considered undistorted. The number of mid-gap states
dislocations are feasible in particular substrates (Coraux changes depending on a difference in the number of removed
et al., 2008), and gauge fields related to the local cur- sites for two sublattices.
vature are then expected to play a crucial role in such
structures. The resulting electronic structure can be an-
alyzed using the theory of quantum mechanics in curved
neighbor square lattice with half flux per plaquette, or
space (Birrell and Davies, 1982; Cortijo and Vozmediano,
the nearest neighbor square lattice with two flavors per
2007a,b).
site. The latter has been extensively studied in relation
to the effects of impurities on the electronic structure
C. Impurity states of d-wave superconductors (Balatsky et al., 2006), and
numerical results are in agreement with the existence of
Point defects, similar to impurities and vacancies, can this solution. As the state is localized on one sublattice
nucleate electronic states in their vicinity. Hence, a con- only, the solution can be generalized for the case of two
centration of ni impurities per carbon atom leads to a vacancies.
change in the electronic density of the order of ni . The
√
corresponding shift in the Fermi energy is F ' vF ni .
In addition, impurities lead to a finite elastic mean free
−1/2
path, lelas ' ani , and to an elastic scattering time D. Localized states near edges, cracks, and voids
−1
τelas ' (vF ni ) . Hence, the regions with few impurities
can be considered low-density metals in the dirty limit,
−1
as τelas ' F . Localized states can be defined at edges where the
The Dirac equation allows for localized solutions that number of atoms in the two sublattices is not compen-
satisfy many possible boundary conditions. It is known sated. The number of them depend on details of the edge.
that small circular defects result in localized and semi- The graphene edges can be strongly deformed, due to the
localized states (Dong et al., 1998), that is, states whose bonding of other atoms to carbon atoms at the edges.
wavefunction decays as 1/r as a function of the distance These atoms should not induce states in the graphene π
from the center of the defect. A discrete version of these band. In general, a boundary inside the graphene ma-
states can be realized in a nearest neighbor tight-binding terial will exist, as sketched in Fig. 27, beyond which
model with unitary scatterers such as vacancies (Pereira the sp2 hybridization is well defined. If this is the case,
et al., 2006). In the continuum, the Dirac equation (19) the number of mid-gap states near the edge is roughly
for the wavefunction, ψ(r) = (φ1 (r), φ2 (r)), can be writ- proportional to the difference in sites between the two
ten as: sublattices near this boundary.
Along a zigzag edge there is one localized state per
∂w φ1 (w, w∗ ) = 0 ,
three lattice units. This implies that a precursor struc-
∂w∗ φ2 (w, w∗ ) = 0 , (157) ture for localized states at the Dirac energy can be found
where w = x + iy is a complex number. These equa- in ribbons or constrictions of small lengths (Muñoz Rojas
tions indicate that at zero energy the components of the et al., 2006), which modifies the electronic structure and
wavefunction can only be either holomorphic or anti- transport properties.
holomorphic with respect to the variable w (that is, Localized solutions can also be found near other defects
φ1 (w, w∗ ) = φ1 (w∗ ) and φ2 (w, w∗ ) = φ2 (w)). Since the that contain broken bonds or vacancies. These states do
boundary conditions require that the wavefunction van- not allow an analytical solution, although, as discussed
ishes at infinity the only possible solutions have the form: above, the continuum Dirac equation is compatible with
ΨK (r̃) ∝ (1/(x + iy)n , 0) or ΨK 0 (r̃) ∝ (0, 1/(x − iy)n ). many boundary conditions, and it should describe well lo-
The wavefunctions in the discrete lattice must be real, calized states that vary slowly over distances comparable
and at large distances the actual solution found near to the lattice spacing. The existence of these states can
a vacancy tends to a superposition of two solutions be investigated by analyzing the scaling of the spectrum
formed from wavefunctions from the two valleys with near a defect as a function of the size of the system, L
equal weight, in a way similar to the mixing at armchair (Vozmediano et al., 2005). A number of small voids and
edges (Brey and Fertig, 2006a). elongated cracks show states whose energy scales as L−2 ,
The construction of the semi-localized state around a while the energy of extended states scales as L−1 . A state
vacancy in the honeycomb lattice can be extended to with energy scaling L−2 is compatible with continuum
other discrete models which leads to the Dirac equation states for which the modulus of the wavefunction decays
in the continuum limit. A particular case is the nearest as r−2 as a function of the distance from the defect.
 28

E. Self-doping
FIG. 28 (Color online) Top: Self-consistent analysis of the
displaced charge density (in units of number of electrons per
The band structure calculations discussed in previous carbon) is shown as a continuous line, and the corresponding
sections show that the electronic structure of a single electrostatic potential (in units of t) is shown as a dashed
graphene plane is not strictly symmetrical in energy (Re- line, for a graphene ribbon with periodic boundary conditions
ich et al., 2002). The absence of electron-hole symmetry along the zig-zag edge (with a√length of L = 960a) and with a
shifts the energy of the states localized near impurities circumference of size W = 80 3a. The inset shows the charge
above or below the Fermi level, leading to a transfer of density near the edge. Due to the presence of the edge, there
is a displaced charge in the bulk (bottom panel) that is shown
charge from/to the clean regions. Hence, the combina-
as a function of width W . Notice that the displaced charge
tion of localized defects and the lack of perfect electron- vanishes in the bulk limit (W → ∞), in agreement with (161).
hole symmetry around the Dirac points leads to the pos- Adapted from Peres et al., 2006c.
sibility of self-doping, in addition to the usual scattering
processes.
Extended lattice defects, like edges, grain boundaries, surface states is mostly localized near the edges of the
or micro-cracks, are likely to induce a number of elec- system.
tronic states proportional to their length, L/a, where a The charge transfer is suppressed by electrostatic ef-
is of the order of the lattice constant. Hence, a distribu- fects, as large deviations from charge neutrality have an
tion of extended defects of length L at a distance equal associated energy cost (Peres et al., 2006c). In order to
to L itself gives rise to a concentration of L/a carriers study these charging effects we add to the free-electron
per carbon in regions of size of the order of (L/a)2 . The Hamiltonian (5) the Coulomb energy of interaction be-
resulting system can be considered a metal with a low tween electrons:
density of carriers, ncarrier ∝ a/L per unit cell, and an
X
HI = Ui,j ni nj , (159)
elastic mean free path lelas ' L. Then, we obtain: i,j
vF
where ni = σ (a†i,σ ai,σ +b†i,σ bi,σ ) is the number operator
P
F ' √
aL at site Ri , and
1 vF
' (158) e2
τelas L Ui,j = , (160)
0 |Ri − Rj |
−1
and, therefore, (τelas )  F when a/L  1. Hence, is the Coulomb interaction between electrons. We expect,
the existence of extended defects leads to the possibil- on physics grounds, that an electrostatic potential builds
ity of self-doping but maintaining most of the sample in up at the edges, shifting the position of the surface states,
the clean limit. In this regime, coherent oscillations of and reducing the charge transferred to/from them. The
transport properties are expected, although the observed potential at the edge induced by a constant doping δ
electronic properties may correspond to a shifted Fermi per carbon atom is roughly, ∼ (δe2 /a)(W/a) (δe2 /a is
energy with respect to the nominally neutral defect–free the Coulomb energy per carbon), and W the width of
system. the ribbon (W/a is the number of atoms involved). The
One can describe the effects that break electron-hole charge transfer is stopped when the potential shifts the
symmetry near the Dirac points in terms of a finite next- localized states to the Fermi energy, that is, when t0 ≈
nearest neighbor hopping between π orbitals, t0 , in (148). (e2 /a)(W/a)δ. The resulting self-doping is therefore
Consider, for instance, electronic structure of a ribbon of
width L terminated by zigzag edges, which, as discussed, t0 a2
δ∼ , (161)
lead to surface states for t0 = 0. The translational sym- e2 W
metry along the axis of the ribbon allows us to define that vanishes when W → ∞.
bands in terms of a wavevector parallel to this axis. On We treat Hamiltonian (159) within the Hartree P approx-
the other hand, the localized surface bands, extending imation (thatP is, we replace HI by HM.F. = i Vi ni
from kk = (2π)/3 to kk = −(2π)/3 acquire a dispersion where Vi = j U i,j hn j i, and solve the problem self-
of order t0 . Hence, if the Fermi energy remains unchanged consistently for hni i). Numerical √ results for graphene
at the position of the Dirac points (Dirac = −3t0 ), this ribbons of length L = 80 3a and different widths are
band will be filled, and the ribbon will no longer be charge shown in Fig. 28 (t0 /t = 0.2 and e2 /a = 0.5t). The
neutral. In order to restore charge neutrality, the Fermi largest width studied is ∼ 0.1µm, and the total number
level needs to be shifted by an amount of the order of of carbon atoms in the ribbon is ≈ 105 . Notice that as
t0 . As a consequence, some of the extended states near W increases, the self-doping decreases indicating that,
the Dirac points are filled, leading to the phenomenon of for a perfect graphene plane (W → ∞), the self-doping
self-doping. The local charge is a function of distance to effect disappears. For realistic parameters, we find that
the edges, setting the Fermi energy so that the ribbon is the amount of self-doping is 10−4 − 10−5 electrons per
globally neutral. Note that the charge transferred to the unit cell for sizes 0.1 − 1µm.
 29

B
FIG. 29 (Color online) Gauge field induced by a simple elas-
tic strain. Top: The hopping along the horizontal bonds is lB h=1 nm, l=10nm, a=0.1nm
assumed to be changed on the right hand side of the graphene l


Intrinsic curvature l h
0.06T
lattice, defining a straight boundary between the unperturbed q
t l3
and perturbed regions (green dashed line). Bottom: The Extrinsic curvature l 2 0.006T
qE ah
modified hopping acts like a constant gauge field, which dis- Elastic strains l 1 al
6T
β h2
places the Dirac cones in opposite directions at the K and K 0
points of the Brillouin zone. The conservation of energy and
momentum parallel to the boundary leads to a deflection of TABLE II Estimates of the effective magnetic length, and ef-
electrons by the boundary. fective magnetic fields generated by the deformations consid-
ered in this section. The intrinsic curvature entry also refers
to the contribution from topological defects.

F. Vector potential and gauge field disorder

A comparison of the relative strengths of the gauge
As discussed in Sec. IV, lattice distortions modify the fields induced by intrinsic curvature, π − σ hybridization
Dirac equation that describes the low energy band struc- (extrinsic curvature), and elastic strains, arising from a
ture of graphene. We consider here deformations that ripple of typical height and size is given in Table II.
change slowly on the lattice scale, so that they do not mix
the two inequivalent valleys. As shown earlier, perturba-
tions that hybridize the two sublattices lead to terms 1. Gauge field induced by curvature
that change the Dirac Hamiltonian from vF σ · k into
vF σ · k + σ · A. Hence, the vector A can be thought of as As we discussed in Sec. IV.A, when the π orbitals are
if induced by an effective gauge field, A. In order to pre- not parallel, the hybridization between them depends on
serve time reversal symmetry, this gauge field must have their relative orientation. The angle θi determines the
opposite signs at the two Dirac cones, AK = −AK 0 . relative orientation of neighboring orbitals at some posi-
tion ri in the graphene sheet. The value of θi depends
A simple example is a distortion that changes the hop-
on the local curvature of the layer. The relative angle of
ping between all bonds along a given axis of the lattice.
rotation of two pz orbitals at position ri and rj can be
Let us assume that the sites at the ends of those bonds de-
written as: cos(θi − θj ) = Ni · Nj , where Ni is the unit
fine the unit cell, as sketched in Fig. 29. If the distortion
vector perpendicular to the surface, defined in (125). If
is constant, its only effect is to displace the Dirac points
rj = ri + uij we can write:
away from the BZ corners. The two inequivalent points
are displaced in opposite directions. This uniform dis- 1
Ni ·Nj ≈ 1+Ni ·[(uij ·∇)Ni ]+ Ni ·[(uij ·∇)2 Ni ](162)
,
tortion is the equivalent of a constant gauge field, which 2
does not change the electronic spectrum. The situation
where we assume smoothly varying N(r). We use (125)
changes if one considers a boundary that separates two
in terms of the height field h(r) (N(r) ≈ z − ∇h(r) −
domains where the magnitude of the distortion is differ-
(∇h)2 z/2) to rewrite (162) as:
ent. The shift of the Dirac points leads to a deflection
of the electronic trajectories that cross the boundary, as 1
Ni · Nj ≈ 1 − [(uij · ∇)∇h(ri )]2 . (163)
also sketched in Fig. 29. The modulation of the gauge 2
field leads to an effective magnetic field, which is of op-
Hence, bending of the graphene sheet leads to a modifi-
posite sign for the two valleys.
cation of the hopping amplitude between different sites
We have shown in Section IV.B how topological lat- in the form:
tice defects, such as disclinations and dislocations, can be
described by an effective gauge field. Those defects can t0ij
δtij ≈ − [(uij · ∇)∇h(ri )]2 , (164)
only exist in graphene sheets that are intrinsically curved, 2
and the gauge field only depends on topology of the lat- where t0ij is the bare hopping energy. A similar ef-
tice. Changes in the nearest neighbor hopping also lead fect leads to changes the electronic states in a carbon
to effective gauge fields. We consider next two physical nanotubes (Kane and Mele, 1997). Using the results
processes that induce effective gauge fields: (i) changes of Sec. IV, namely (147), we can now see that a vec-
in the hopping induced by hybridization between π and tor potential is generated for nearest neighbor hopping
σ bands, which arise in curved sheets, and (ii) changes in (u = ~δab ) (Kim and Neto, 2008):
the hopping due to modulation in the bond length, which
is associated with elastic strain. The strength of these 3Eab a2  2 2
A(h) (∂x h) − (∂y2 h)2

fields depends on parameters that describe the value of x =−
8
the π-σ hybridization, and the dependence of hopping on 3Eab a2 2
A(h) ∂x h + ∂y2 h ∂x h∂y h


the bond length. y = (165)
4
 30

where the coupling constant Eab depends on microscopic where κ is a dimensionless quantity that depends on mi-
details (Kim and Neto, 2008). The flux of effective mag- croscopic details. Changes in the bond length lead to
netic field through a ripple of lateral dimension l and changes in the hopping amplitude:
height h is given approximately by:
∂tij
tij ≈ t0ij + δui , (172)
Eab a2 h2 ∂a
Φ≈ (166)
vF l 3 and we can write:
−1 −2
where the radius of curvature is R ≈ hl . For a ripple δu(r)
δt(ab) (r) ≈ β , (173)
with l ≈ 20nm, h ≈ 1nm, taking Eab /vF ≈ 10 Å−1 , we a
find Φ ≈ 10−3 Φ0 .
where
∂t(ab)
2. Elastic strain β= . (174)
∂ ln(a)
The elastic free energy for graphene can be written in Substituting (170) into (173) and the final result into
terms of the in-plane displacement u(r) = (ux , uy ) as: (147), one finds (Ando, 2006a):
 
1
Z 3
F [u] = d2 r (B − G)(
X
uii )2 + 2G
X
u2ij (167)
, A(s)
x = β κ (uxx − uyy ) ,
2 4
i=1,2 i,j=1,2 3
A(s)
y = β κ uxy . (175)
2
where B is the bulk modulus, G is the shear modulus,
and We assume that the strains induced by a ripple of di-
  mension l and height h scale as uij ∼ (h/l)2 . Then,
1 ∂ui ∂uj
uij = + , (168) using β/vF ≈ a−1 ∼ 1Å−1 , we find that the total flux
2 ∂xj ∂xj through a ripple is:
is the strain tensor (x1 = x and x2 = y).
h2
There are many types of static deformation of the hon- Φ≈ . (176)
eycomb lattice which can affect the propagation of Dirac al
fermions. The simplest one is due to changes in the For ripples such that h ∼ 1nm and l ∼ 20nm, this es-
area of the unit cell either due to dilation or contrac- timate gives Φ ∼ 10−1 Φ0 in reasonable agreement with
tion. Changes in the unit cell area lead to local changes the estimates in ref. (Morozov et al., 2006).
in the density of electrons and, therefore, local changes in The strain tensor must satisfy some additional con-
the chemical potential in the system. In this case, their straints, as it is derived from a displacement vector field.
effect is similar to the one found in (148), and we must These constraints are called Saint Venant compatibility
have: conditions:
φdp (r) = g(uxx + uyy ) , (169) ∂uij ∂ukl ∂uil ∂ujk
Wijkl = + − − = 0.
∂xk ∂xl ∂xi ∂xj ∂xj ∂xk ∂xi ∂xl
and their effect is diagonal in the sublattice index. (177)
The nearest neighbor hopping depends on the length of An elastic deformation changes the distances in the crys-
the carbon bond. Hence, elastic strains that modify the tal lattice and can be considered as a change in the met-
relative orientation of the atoms also lead to an effective ric:
gauge field, which acts on each K point separately, as
first discussed in relation to carbon nanotubes (Maiti and gij = δij + uij (178)
Sengupta, 2007; Suzuura and Ando, 2002b). Consider
two carbon atoms located in two different sublattices in The compatibility equations (177) are equivalent to the
the same unit cell at Ri . The change in the local bond condition that the curvature tensor derived from (178) is
length can be written as: zero. Hence, a purely elastic deformation cannot induce
~δab intrinsic curvature in the sheet, which only arises from
δui = · [uA (Ri ) − uB (Ri + ~δab )] . (170) topological defects. The effective fields associated with
a elastic strains can be large (Morozov et al., 2006), lead-
The local displacements of the atoms in the unit cell can ing to significant changes in the electronic wavefunctions.
be related to u(r) by (Ando, 2006a): An analysis of the resulting state, and the possible insta-
bilities that may occur can be found in (Guinea et al.,
(~δab · ∇)u = κ−1 (uA − uB ) , (171) 2008).
 31

3. Random gauge fields that can be comparable to the bandwidth allows the for-
mation of mid-gap states and changes the phase-shift as-
The preceding discussion suggests that the effective sociated to spin scattering (Hentschel and Guinea, 2007).
fields associated with lattice defects can modify signif- These phase-shifts have a weak logarithmic dependence
icantly the electronic properties. This is the case when on the chemical potential, and a Kondo resonance can
the fields do not change appreciably on scales comparable exist, even close to the Dirac energy.
to the (effective) magnetic length. The general problem The RKKY interaction between magnetic impurities
of random gauge fields for Dirac fermions has been exten- is also modified in graphene. At finite fillings, the ab-
sively analyzed before the current interest in graphene, sence of intra-valley backscattering leads to a reduction
as the topic is also relevant for the IQHE (Ludwig et al., of the Friedel oscillations, which decay as sin(2kF r)/|r|3
1994) and d-wave superconductivity. The one electron (Ando, 2006b; Cheianov and Fal’ko, 2006; Wunsch et al.,
nature of this two dimensional problem makes it possible, 2006). This effect leads to an RKKY interaction, at fi-
at the Dirac energy, to map it onto models of interact- nite fillings, which oscillate and decay as |r|−3 . When
ing electrons in one dimension, where many exact results intervalley scattering is included, the interaction re-
can be obtained (Castillo et al., 1997). The low energy verts to the usual dependence on distance in two di-
density of states, ρ(ω), acquires an anomalous exponent, mensions, |r|−2 (Cheianov and Fal’ko, 2006). At half-
ρ(ω) ∝ |ω|1−∆ , where ∆ > 0. The density of states is filling extended defects lead to an RKKY interaction with
enhanced near the Dirac energy, reflecting the tendency an |r|−3 dependence (Dugaev et al., 2006; Vozmediano
of disorder to close gaps. For sufficiently large values of et al., 2005). This behavior is changed when the impu-
the random gauge field, a phase transition is also possible rity potential is localized on atomic scales (Brey et al.,
(Chamon et al., 1996; Horovitz and Doussal, 2002). 2007; Saremi, 2007), or for highly symmetrical couplings
Perturbation theory shows that random gauge fields (Saremi, 2007).
are a marginal perturbation at the Dirac point, leading
to logarithmic divergences. These divergences tend to
have the opposite sign with respect to those induced by H. Weak and strong localization
the Coulomb interaction (see Sec. V.B). As a result, a
renormalization group (RG) analysis of interacting elec- In sufficiently clean systems, where the Fermi wave-
trons in a random gauge field suggests the possibility length is much shorter than the mean free path, kF l  1,
of non-trivial phases (Aleiner and Efetov, 2006; Altland, electronic transport can be described in classical terms,
2006; Dell’Anna, 2006; Foster and Ludwig, 2006a,b; No- assuming that electrons follow well defined trajectories.
mura and MacDonald, 2006; Stauber et al., 2005), where At low temperatures, when electrons remain coherent
interactions and disorder cancel each other. over long distances, quantum effects lead to interfer-
ence corrections to the classical expressions for the con-
ductivity, the weak localization correction (Bergmann,
G. Coupling to magnetic impurities 1984; Chakravarty and Schmid, 1986). These corrections
are usually due to the positive interference between two
Magnetic impurities in graphene can be introduced paths along closed loops, traversed in opposite directions.
chemically by deposition and intercalation (Calandra and As a result, the probability that the electron goes back
Mauri, 2007), or self-generated by the introduction of de- to the origin is enhanced, so that quantum corrections
fects (Kumazaki and Hirashima, 2006, 2007). The energy decrease the conductivity. These interferences are sup-
dependence of the density of states in graphene leads to pressed for paths longer than the dephasing length, lφ ,
changes in the formation of a Kondo resonance between determined by interactions between the electron and en-
a magnetic impurity and the graphene electrons. The vironment. Interference effects can also be suppressed
vanishing of the density of states at the Dirac energy by magnetic fields that break down time reversal sym-
implies that a Kondo singlet in the ground state is not metry and adds a random relative phase to the process
formed unless the exchange coupling exceeds a critical discussed above. Hence, in most metals, the conductivity
value, of the order of the electron bandwidth, a problem increases when a small magnetic field is applied (negative
already studied in connection with magnetic impurities in magnetoresistance).
d-wave superconductors (Cassanello and Fradkin, 1996, Graphene is special in this respect, due to the chirality
1997; Fritz et al., 2006; Polkovnikov, 2002; Polkovnikov of its electrons. The motion along a closed path induces a
et al., 2001). For weak exchange couplings, the magnetic change in the relative weight of the two components of the
impurity remains unscreened. An external gate changes wavefunction, leading to a new phase, which contributes
the chemical potential, allowing for a tuning of the Kondo to the interference processes. If the electron traverses
resonance (Sengupta and Baskaran, 2008). The situation a path without being scattered from one valley to the
changes significantly if the scalar potential induced by the other, this (Berry) phase changes the sign of the ampli-
magnetic impurity is taken into account. This potential tude of one path with respect to the time-reversed path.
 32

As a consequence, the two paths interfere destructively, namely a conductance of the order of e2 /h, is obtained
leading to a suppression of backscattering (Suzuura and for disordered graphene bilayers where a self-consistent
Ando, 2002a). Similar processes take place in materi- calculation leads to universal conductivity at the neutral-
als with strong spin orbit coupling, as the spin direction ity point (Katsnelson, 2007b; Nilsson et al., 2007, 2006b).
changes along the path of the electron (Bergmann, 1984; In a biased graphene bilayer, the presence of impurities
Chakravarty and Schmid, 1986). Hence, if scattering be- leads to the appearance of impurity tails in the density
tween valleys in graphene can be neglected, one expects a of states due to the creation of mid-gap states which are
positive magnetoresistance, i. e., weak anti-localization. sensitive to the applied electric field that opens the gap
In general, intra- and intervalley elastic scattering can between the conduction and valence bands (Nilsson and
be described in terms of two different scattering times, Castro Neto, 2007).
τintra and τinter , so that if τintra  τinter one expects
weak anti-localization processes, while if τinter  τintra One should point out that most of the calculations
ordinary weak localization will take place. Experimen- of transport properties assume self-averaging, that is,
tally, localization effects are always strongly suppressed that one can exchange a problem with lack of transla-
close to the Dirac point but can be partially or, in rare tional invariance by an effective medium system with
cases, completely recovered at high carrier concentra- damping. Obviously this procedure only works when
tions, depending on a particular single-layer sample (Mo- the disorder is weak and the system is in the metallic
rozov et al., 2006; Tikhonenko et al., 2008). Multilayer phase. Close to the localized phase this procedure breaks
samples exhibit an additional positive magnetoresistance down, the system divides itself into regions of different
in higher magnetic fields, which can be attribued to clas- chemical potential and one has to think about transport
sical changes in the current distribution due to a vertical in real space, usually described in terms of percolation
gradient of concentration (Morozov et al., 2006) and anti- (Cheianov et al., 2007b; Shklovskii, 2007). Single elec-
localization effects. tron transistor (SET) measurements of graphene show
The propagation of an electron in the absence of inter- that this seems to be the situation in graphene at half-
valley scattering can be affected by the effective gauge filling.
fields induced by lattice defects and curvature. These
fields can suppress the interference corrections to the con- Finally, we should point out that graphene stacks suf-
ductivity (Morozov et al., 2006; Morpurgo and Guinea, fer from another source of disorder, namely, c-axis dis-
2006). In addition, the description in terms of free Dirac order that is either due to impurities between layers
electrons is only valid near the neutrality point. The or rotation of graphene planes relative to each other.
Fermi energy acquires a trigonal distortion away from the In either case the in-plane and out-of-plane transport
Dirac point, and backward scattering within each valley is directly affected. This kind of disorder has been
is no longer completely suppressed (McCann et al., 2006), observed experimentally by different techniques (Hass
leading to a further suppression of anti-localization ef- et al., 2008). In the case of the bilayer, the rotation
fects at high dopings. Finally, the gradient of external of planes changes substantially the spectrum restoring
potentials induce a small asymmetry between the two the Dirac fermion description (Lopes dos Santos et al.,
sublattices (Morpurgo and Guinea, 2006). This effect 2007). The transport properties in the out of plane di-
will also contribute to reduce anti-localization, without rection are determined by the interlayer current operator,
ĵn,n+1 = it (c†A,n,s cA,n+1,s − c†A,n+1,s cA,n,s ), where n is
P
giving rise to localization effects.
The above analysis has to be modified for a graphene a layer index, and A is a generic index that defines the
bilayer. Although the description of the electronic states sites coupled by the interlayer hopping t. If we only con-
requires a two component spinor, the total phase around sider hopping between nearest neighbor sites in consecu-
a closed loop is 2π, and backscattering is not suppressed tive layers, these sites belong to one of the two sublattices
(Kechedzhi et al., 2007). This result is consistent with in each layer.
experimental observations, which show the existence of
weak localization effects in a bilayer (Gorbachev et al., In a multilayer with Bernal stacking, these connected
2007). sites are the ones where the density of states vanishes at
When the Fermi energy is at the Dirac point, a replica zero energy, as discussed above. Hence, even in a clean
analysis shows that the conductivity approaches a univer- system, the number of conducting channels in the direc-
sal value of the order of e2 /h (Fradkin, 1986a,b). This re- tion perpendicular to the layers vanishes at zero energy
sult is valid when intervalley scattering is neglected (Os- (Nilsson et al., 2007, 2006b). This situation is reminis-
trovsky et al., 2006, 2007; Ryu et al., 2007). Localization cent of the in plane transport properties of a single layer
is induced when these terms are included (Aleiner and graphene. Similar to the latter case, a self-consistent
Efetov, 2006; Altland, 2006), as also confirmed by numer- Born approximation for a small concentration of impu-
ical calculations (Louis et al., 2007). Interaction effects rities leads to a finite conductivity, which becomes inde-
tend to suppress the effects of disorder. The same result, pendent of the number of impurities.
 33

I. Transport near the Dirac point twice bigger than the one in single layer graphene.
The calculation of the conductance of clean graphene
In clean graphene, the number of channels available in terms of transmission coefficients, using the Landauer
for electron transport decreases as the chemical poten- method leads to an effective conductivity which is equal
tial approaches the Dirac energy. As a result, the con- to the value obtained for bulk graphene using diagram-
ductance through a clean graphene ribbon is, at most, matic methods, the Kubo formula (Peres et al., 2006d),
4e2 /h, where the factor of 4 stands for the spin and val- in the limit of zero impurity concentration and zero dop-
ley degeneracy. In addition, only one out of every three ing. Moreover, this correspondence remains valid for the
possible clean graphene ribbons have a conduction chan- case of a bilayer without and with trigonal warping effects
nel at the Dirac energy. The other two thirds are semi- (Cserti et al., 2007a; Koshino and Ando, 2006).
conducting, with a gap of the order of vF /W , where W Disorder at the Dirac energy changes the conduc-
is the width. This result is a consequence of the addi- tance of graphene ribbons in two opposite directions
tional periodicity introduced by the wavefunctions at the (Louis et al., 2007): i) a sufficiently strong disorder, with
K and K 0 points of the Brillouin Zone, irrespective of short range (intervalley) contributions, lead to a local-
the boundary conditions. ized regime, where the conductance depends exponen-
A wide graphene ribbon allows for many channels, tially on the ribbon length, and ii) at the Dirac energy,
which can be approximately classified by the momentum disorder allows mid-gap states that can enhance the con-
perpendicular to the axis of the ribbon, ky . At the Dirac ductance mediated by evanescent waves discussed above.
energy, transport through these channels is inhibited by A fluctuating electrostatic potential also reduces the ef-
the existence of a gap, ∆ky = vF ky . Transport through fective gap for the transverse channels, enhancing fur-
these channels is suppressed by a factor of the order of ther the conductance. The resonant tunneling regime
e−ky L , where L is the length of the ribbon. The number mediated by mid-gap state was suggested by analytical
of transverse channels increases as W/a, where W is the calculations (Titov, 2007). The enhancement of the con-
width of the ribbon and a is a length of the order of the ductance by potential fluctuations can also be studied
lattice spacing. The allowed values of ky are ∝ ny /W , semi-analytically. In the absence of intervalley scatter-
where ny is an integer. Hence, for a ribbon such that ing, it leads to an effective conductivity which grows
W  L, there are many channels which satisfy ky L  1. with ribbon length (San-Jose et al., 2007). In fact, an-
Transport through these channels is not strongly inhib- alytical and numerical studies (Bardarson et al., 2007;
ited, and their contribution dominates when the Fermi Lewenkopf et al., 2008; Nomura and MacDonald, 2006;
energy lies near the Dirac point. The conductance aris- San-Jose et al., 2007) show that the conductivity obeys
ing from these channels is given approximately by (Kat- a universal scaling with the lattice size L:
snelson, 2006b; Tworzydlo et al., 2006):
2e2
σ(L) = (A ln(L/ξ) + B) , (180)
e2 W e2 W h
Z
G∼ dky e−ky L ∼ . (179)
h 2π h L where ξ is a length scale associated with range of inter-
actions and A and B are numbers of the order of unit
The transmission at normal incidence, ky = 0, is one,
(A ≈ 0.17 and B ≈ 0.23 for a graphene lattice in the
in agreement with the absence of backscattering in
shape of a square of size L(Lewenkopf et al., 2008)). No-
graphene, for any barrier that does not induce intervalley
tice, therefore, that the conductivity is always of the or-
scattering. The transmission of a given channel scales as
der of e2 /h and has a weak dependence on size.
T (ky ) = 1/ cosh2 (ky L/2).
Eq.(179) shows that the contribution from all trans-
verse channels lead to a conductance which scales, simi- J. Boltzmann Equation description of DC transport in
lar to a function of the length and width of the system, doped graphene
as the conductivity of a diffusive metal. Moreover, the
value of the effective conductivity is of the order of e2 /h. It was shown experimentally that the DC conductiv-
It can also be shown that the shot noise depends on cur- ity of graphene depends linearly on the gate potential
rent in the same way as in a diffusive metal. A detailed (Novoselov et al., 2004, 2005), except very close to the
analysis of possible boundary conditions at the contacts neutrality point (see Fig.30). Since the gate potential
and their influence on evanescent waves can be found depends linearly on the electronic density, n, one has
in (Robinson and Schomerus, 2007; Schomerus, 2007). a conductivity σ ∝ n. As shown by Shon and Ando
The calculations leading to eq.(179) can be extended to (Shon and Ando, 1998) if the scatterers are short range
a graphene bilayer. The conductance is, again, a summa- the DC conductivity should be independent of the elec-
tion of terms arising from evanescent waves between the tronic density, at odds with the experimental result. It
two contacts, and it has the dependence on sample di- has been shown (Ando, 2006b; Nomura and MacDon-
mensions of a 2D conductivity of the order of e2 /h (Sny- ald, 2007; Novikov, 2007a) that by considering a scat-
man and Beenakker, 2007), although there is a prefactor tering mechanism based on screened charged impurities
 34

The relaxation time τk is defined as:

FIG. 30 (Color online) An example of changes in conductivity
Z 0 0
σ of graphene with varying gate voltage, Vg , and, therefore,
Z
1 k dk
carrier concentration n. Here σ is proportional to n as dis- = ni d θ S(k, k0 )(1 − cos θ) , (184)
τk (2π)2
cussed in the text. Note that samples with higher mobility
(> 1 m2 /Vs) normally show a sublinear dependence, presum- where ni is impurity concentration per unit of area, and
ably indicating the presence of different types of scatterers. the transition rate S(k, k0 ) is given, in the Born approx-
Inset: scanning-electron micrograph of one of experimental
devices (in false colors matching those seen in visible optics.
imation, by:
The scale of the micrograph is given by the width of the Hall 1
bar, which is one micrometer. S(k, k0 ) = 2π|Hk0 ,k |2 δ(k 0 − k) , (185)
vF
where the vF k is the dispersion of Dirac fermions in
it is possible to obtain from a Boltzmann equation ap- graphene and Hk0 ,k is defined as
proach a conductivity varying linearly with the density, Z
in agreement with the experimental result (Ando, 2006b; Hk0 ,k = drψk∗ 0 (r)US (r)ψk (r) , (186)
Katsnelson and Geim, 2008; Novikov, 2007c; Peres et al.,
2007b; Trushin and Schliemann, 2007). with US (r) the scattering potential and ψk (r) is the elec-
The Boltzmann equation has the form tronic spinor wavefunction of a clean graphene sheet. If
the potential is short range,(Shon and Ando, 1998) of the
∂fk
−vk ·∇r f (k )−e(E +vk ×H)·∇k f (k ) = − . form US = v0 δ(r), the Boltzmann relaxation time turns
∂t scatt. out to be
(181)
The solution of the Boltzmann equation in its general 4vF 1
τk = . (187)
form is difficult and one needs therefore to rely upon ni v02 k
some approximation. The first step in the usual approx-
On the other hand, if the potential is the Coulomb po-
imation scheme is to write the distribution as f (k ) =
tential, given by US (r) = eQ/(4π0 r) for charged impu-
f 0 (k ) + g(k ) where f 0 (k ) is the steady state distri-
rities of charge Q, the relaxation time is given by
bution function and g(k ) is assumed to be small. In-
serting this ansatz in (181) and keeping only terms that vF
τk = k. (188)
are linear in the external fields one obtains the linearized u20
Boltzmann equation which reads
where u20 = ni Q2 e2 /(1620 2 ). As we argue below, the
∂f 0 (k ) phenomenology of Dirac fermions implies that the scat-
   
k − ζ 1
− vk · − ∇r T + e E − ∇r ζ = tering in graphene must be of the form (188).
∂k T e
Within the relaxation time approximation the solution
∂fk
− + vk · ∇r gk + e(vk × H) · ∇k gk . (182) of the linearized Boltzmann equation when an electric
∂t scatt.
field is applied to the sample is
The second approximation has to do with the form of the ∂f 0 (k )
scattering term. The simplest approach is to introduce a gk = − eτk vk · E , (189)
∂k
relaxation time approximation:
and the electric current reads (spin and valley indexes
∂fk gk included)
− → , (183)
∂t scatt. τk
4 X
J= evk gk . (190)
where τk is the relaxation time, assumed to be momen- A
k
tum dependent. This momentum dependence is deter-
mined phenomenologically in such way that the depen- Since at low temperatures the following relation
dence of the conductivity upon the electronic density −∂f 0 (k )/∂k → δ(µ − vF k) holds, one can easily see
agrees with experimental data. The Boltzmann equation that assuming (188) where k is measured relatively to
is certainly not valid at the Dirac point, but since many the Dirac point, the electronic conductivity turns out to
experiments are performed at finite carrier density, con- be
trolled by an external gate voltage, we expect the Boltz- e2 µ2 e2 πvF2
mann equation to give reliable results if an appropriate σxx = 2 2 =2 n, (191)
h u0 h u20
form for τk is used (Adam et al., 2007).
Let us compute the Boltzmann relaxation time, τk , where u0 is the strength of the scattering potential (with
for two different scattering potentials:(i) a Dirac delta dimensions of energy). The electronic conductivity de-
function potential; (ii) a unscreened Coulomb potential. pends linearly on the electron density, in agreement with
 35

the experimental data. We stress that the Coulomb po-

FIG. 31 (Color online) Top: Electronic density of states
tential is one possible mechanism of producing a scatter-
(DOS), ρ(ω), as a function of ω/ωc (ωc = 0.14 eV) in a mag-
ing rate of the form (188) but we do not exclude that netic field B = 12 T for different impurity concentrations ni .
other mechanisms may exist (see, for instance, (Katsnel- Bottom: ρ(ω), as a function of ω/ωc (ωc = 0.1 eV is the cy-
son and Geim, 2008)). clotron frequency) in a magnetic field B = 6 T. The solid line
shows the DOS in the absence of disorder. The position of
the Landau levels in the absence of disorder are shown as ver-
K. Magnetotransport and universal conductivity tical lines. The two arrows in the top panel show the position
of the renormalized Landau levels (see Fig.32) given by the
The description of the magnetotransport properties of solution of Eq. (202). Adapted from Peres et al., 2006c.
electrons in a disordered honeycomb lattice is complex
because of the interference effects associated with the
Hofstadter problem (Gumbs and Fekete, 1997). We shall where
simplify our problem by describing electrons in the hon-
eycomb lattice as Dirac fermions in the continuum ap- √
E(α, n) = αωc n (195)
proximation, introduced in Sec. II.B. Furthermore, we
will only focus on the problem of short range scattering in
the unitary limit since in this regime many analytical re- are the Landau levels for this problem (α = ±1 labels the
sults are obtained (Kumazaki and Hirashima, 2006; Mar- two bands). Notice that G0 (k, n, α; iω) is effectively k-
iani et al., 2007; Pereira et al., 2006, 2008; Peres et al., independent, and E(α, −1) = 0 is the zero energy Landau
2006c; Skrypnyk and Loktev, 2006, 2007). The problem level. When expressed in the Landau basis, the scatter-
of magnetotransport in the presence of Coulomb impuri- ing Hamiltonian (192) connects Landau levels of negative
ties, as discussed in the previous section is still an open and positive energy.
research problem. A similar approach was considered by
Abrikosov in the quantum magnetoresistance study of
non-stoichiometric chalcogenides (Abrikosov, 1998). In
the case of graphene, the effective Hamiltonian describ- 1. The full self-consistent Born approximation (FSBA)
ing Dirac fermions in a magnetic field (including disorder)
can be written as: H = H0 + Hi where H0 is given by In order to describe the effect of impurity scattering
(5) and Hi is the impurity potential reading (Peres et al., on the magnetoresistance of graphene, the Green’s func-
2006c): tion for Landau levels in the presence of disorder needs
to be computed. In the context of the 2D electron gas,
Ni
X an equivalent study was performed by Ohta and Ando,
Hi = V δ(r − rj )I (192) (Ando, 1974a,b, 1975; Ando and Uemura, 1974; Ohta,
j=1
1971a; Ohta et al., 2007) using the averaging procedure
The formulation of the problem in second quantization over impurity positions of Duke (Duke, 1968). Below
requires the solution of H0 , which was done in Section the averaging procedure over impurity positions is per-
II.I. The field operators, close to the K point, are defined formed in the standard way, namely, having determined
as (the spin index is omitted for simplicity): the Green’s function for a given impurity configuration
! (r1 , . . . rNi ), the position averaged Green’s function is
X eikx 0 determined from:
Ψ(r) = √ ck,−1
k
L φ0 (y)
X eikx 2
! hG(p, n, α; iω; r1 , . . . rNi )i ≡ G(p, n, α; iω)
φn (y − klB )
+ √ ck,n,α , (193)
 
2 Ni Z
2L φn+1 (y − klB ) Y
n,k,α = L−2Ni  drj  G(p, n, α; iω; r1 , . . . rNi ) .(196)
j=1
where ck,n,α destroys an electron in band α = ±1, with
2
energy level n and guiding center klB ; ck,−1 destroys an
electron in the zero Landau level; the cyclotron frequency In the presence of Landau levels the average over im-
is given by (96). The sum over n = 0, 1, 2, . . . , is cut off at purity positions involves the wavefunctions of the one-
n0 given by E(1, n0 ) = W , where W is of the order of the dimensional harmonic oscillator. After a lengthy algebra,
electronic bandwidth. In this representation H0 becomes the Green’s function in the presence of vacancies, in the
diagonal, leading to Green’s functions of the form (in FSBA, can be written as:
Matsubara representation):
1 G(p, n, α; ω + 0+ ) = [ω − E(n, α) − Σ1 (ω)]−1 ,(197)
G0 (k, n, α; iω) = , (194)
iω − E(α, n) G(p, −1; ω + 0+ ) = [ω − Σ2 (ω)]−1 , (198)
 36

FIG. 32 (Color online) Imaginary (right) and real (left) parts FIG. 33 (Color online) Conductivity kernel, K(ω) (in units
of Σ1 (ω) (top) and Σ2 (ω)(bottom), in units of ωc , as a func- of e2 /(πh)), as a function of energy ω for different magnetic
tion of ω/ωc . The right panels also show the intercept of fields and for ni = 10−3 . The horizontal lines mark the uni-
ω − E(α, n) with ReΣ(ω) as required by Eq. (202). Adapted versal limit of the conductivity per cone, σ0 = 2e2 /(πh). The
from Peres et al., 2006c. vertical lines show the position of the Landau levels in the
absence of disorder. Adapted from Peres et al., 2006c.

where
The study of the magnetoresistance properties of the
Σ1 (ω) = −ni [Z(ω)]−1 , (199) system requires the calculation of the conductivity ten-
Σ2 (ω) = −ni [gc G(p, −1; ω + 0+ )/2 + Z(ω)]−1 ,(200) sor. We compute the current-current correlation function
Z(ω) = gc G(p, −1; ω + 0+ )/2 and from it the conductivity tensor is derived. The de-
X tails of the calculations are presented in (Peres et al.,
+ gc G(p, n, α; ω + 0+ )/2 , (201) 2006c). If we however neglect the real part of the self-
n,α energy, assume for ImΣi (ω) = Γ (i = 1, 2) a constant
2 value, and consider that E(1, 1)  Γ, these results re-
and gc = Ac /(2πlB ) is the degeneracy of a Landau level
duce to those of (Gorbar et al., 2002).
per unit cell. One should notice that the Green’s func-
It is instructive to consider first the case ω, T → 0,
tions do not depend upon p explicitly. The self-consistent
which leads to (σxx (0, 0) = σ0 ):
solution of Eqs. (197), (198), (199), (200) and (201) gives
the density of states, the electron self-energy, and the e2 4 ImΣ1 (0)/ImΣ2 (0) − 1

change of Landau level energy position due to disorder. σ0 =
h π 1 + (ImΣ1 (0)/ωc )2
The effect of disorder in the density of states of Dirac 
n0 + 1
fermions in a magnetic field is shown in Fig. 31. For ref- + , (203)
erence we note that E(1, 1) = 0.14 eV, for B = 14 T, and n0 + 1 + (ImΣ1 (0)/ωc )2
E(1, 1) = 0.1 eV, for B = 6 T. From Fig. 31 we see that, where we include a factor 2 due to the valley degeneracy.
for a given ni , the effect of broadening due to impurities In the absence of a magnetic field (ωc → 0) the above
is less effective as the magnetic field increases. It is also expression reduces to:
clear that the position of Landau levels is renormalized
e2 4 [ImΣ1 (0)]2
 
relatively to the non-disordered case. The renormaliza-
σ0 = 1− , (204)
tion of the Landau level position can be determined from h π (vF Λ)2 + [ImΣ1 (0)]2
poles of (197) and (198):
where we have introduced the energy cut-off, vF Λ. Either
ω − E(α, n) − ReΣ(ω) = 0 . (202) when ImΣ1 (0) ' ImΣ2 (0) and √ ωc 2 ImΣ1 (0) (or n0 
ImΣ1 (0)/ωc , ωc = E(0, 1) = 2vF /lB ), or when ΛvF 
Of course, due to the importance of scattering at low ImΣ1 (0), in the absence of an applied field, Eqs. (203)
energies, the solution to Eq. (202) does not represent and (204) reduce to:
exact eigenstates of system since the imaginary part of
4 e2
the self-energy is non-vanishing. However, these energies σ0 = , (205)
do determine the form of the density of states, as we π h
discuss below. which is the so-called universal conductivity of graphene
In Fig. 32, the graphical solution to Eq. (202) is given (Fradkin, 1986a,b; Katsnelson, 2006b; Lee, 1993; Ludwig
for two different energies (E(−1, n), with n = 1, 2), its is et al., 1994; Peres et al., 2006c; Tworzydlo et al., 2006;
clear that the renormalization is important for the first Yang and Nayak, 2002; Ziegler, 1998). This result was
Landau level. This result is due to the increase in scat- obtained previously by Ando and collaborators using the
tering at low energies, which is present already in the second order self-consistent Born approximation (Ando
case of zero magnetic field. The values of ω satisfying et al., 2002; Shon and Ando, 1998).
Eq. (202) show up in the density of states as the en- Because the DC magnetotransport properties of
ergy values where the oscillations due to the Landau level graphene are normally measured with the possibility
quantization have a maximum. In Fig. 31, the position of tuning its electronic density by a gate potential
of the renormalized Landau levels is shown in the upper (Novoselov et al., 2004), it is important to compute the
panel (marked by two arrows), corresponding to the bare conductivity kernel, since this has direct experimental
energies E(−1, n), with n = 1, 2. The importance of this relevance. In the the case ω → 0 we write the conductiv-
renormalization decreases with the reduction of the num- ity σxx (0, T ) as:
ber of impurities. This is clear in Fig. 31 where a visible
∞
e2
Z
shift toward low energies is evident when ni has a small ∂f ()
σxx (0, T ) = d KB () , (206)
10% change, from ni = 10−3 to ni = 9 × 10−4 . πh −∞ ∂
 37

FIG. 34 (Color online) Frequency dependent conductivity per FIG. 35 (Color online) Top: The continuous line is the rel-
cone, σ(ω) (in units of e2 /(πh)) at T = 10 K and ni = 10−3 , ative phonon frequency shift as a function of µ/ω0 , and the
as a function of the energy ω (in units of ωc ) for different val- dashed line is the damping of the phonon due to electron-
ues of the magnetic field B. The vertical arrows in the upper hole pair creation; Bottom: (a) Electron-hole process that
left panel, labeled a, b, and c, show the positions of the tran- leads to phonon softening (ω0 > 2µ), and (b) electron-hole
sitions between different Landau levels: E(1, 1) − E(−1, 0), process that leads to phonon hardening (ω0 < 2µ).
E(2, 1) − E(−1, 0), and E(1, 1) − E(1, −1), respectively. The
horizontal continuous lines show the value of the universal
conductivity. The lower right panel shows the conductivity √ Dirac equation (150).
for different values of magnetic field as a function of ω/ B.
For optical modes the situation is slightly different
Adapted from Peres et al., 2006c.
since the optical mode displacement is (Ando, 2006a,
2007a):
where the conductivity kernel KB () is given in the Ap- 1
pendix of Ref. (Peres et al., 2006c). The magnetic field uop = √ (uA − uB ) , (207)
2
dependence of kernel KB () is shown in Fig. 33. Ob-
serve that the effect of disorder is the creation of a re- that is, the bond length deformation vector. To calculate
gion where KB () remains constant before it starts to the coupling to the electrons we can proceed as previously
increase in energy with superimposed oscillations com- and compute the change in the nearest neighbor hopping
ing from the Landau levels. The same effect, but with energy due to the lattice distortion through (172), (173),
the absence of the oscillations, was identified at the level (170), and (207). Once again the electron-phonon inter-
of the self-consistent density of states plotted in Fig. 31. action becomes a problem of the coupling of the electrons
Together with σxx (0, T ), the Hall conductivity σxy (0, T ) with a vector potential as in (150) where the components
allows the calculation of the resistivity tensor (109). of the vector potential are:
Let us now focus on the optical conductivity, σxx (ω) r
(Gusynin et al., 2007; Peres et al., 2006c). This quantity (op) 3 β op
Ax = − u ,
can be probed by reflectivity experiments in the subter- 2 a2 y
r
ahertz to mid-infrared frequency range (Bliokh, 2005). 3 β op
A(op)
y =− u , (208)
This quantity is depicted in Fig. 34 for different mag- 2 a2 x
netic fields. It is clear that the first peak is controlled
by the E(1, 1) − E(1, −1), and we have checked that it where β = ∂t/∂ ln(a) was p defined 2in (174). Notice that
~ op
we can write: A = − 3/2(β/a ) ~σ × uop . A similar
does not obey any particular scaling form as a function
of ω/B. On the other hand, as the effect of scattering expression is valid close to the K’ point with A ~ replaced
becomes less important the high energy by −A. ~
√ conductivity os-
cillations start obeying the scaling ω/ B, as we show in Optical phonons are particularly important in
the lower right panel of Fig. 34. graphene research because of Raman spectroscopy. The
latter has played a particularly important role in the
study of carbon nanotubes because of the 1D character
V. MANY-BODY EFFECTS of these systems, namely, the presence of van Hove singu-
larities in the 1D spectrum lead to colossal enhancements
A. Electron-phonon interactions of the Raman signal that can be easily detected, even for
a single isolated carbon nanotube. In graphene the situ-
In Sec. IV.F.1 and Sec. IV.F.2 we discussed how static ation is rather different since its 2D character leads to a
deformations of the graphene sheet due to bending and much smoother density of states (except for the van Hove
strain couple to the Dirac fermions via vector potentials. singularity at high energies of the order of the hopping
Just as bending has to do with the flexural modes of the energy t ≈ 2.8 eV). Nevertheless, graphene is an open
graphene sheet (as discussed in Sec. III), strain fields are surface and hence is readily accessible by Raman spec-
related to optical and acoustic modes (Wirtz and Rubio, troscopy. In fact, it has played a very important role be-
2004). Given the local displacements of the atoms in each cause it allows the identification of the number of planes
sublattice, uA and uB , the electron-phonon coupling has (Ferrari et al., 2006; Graf et al., 2007; Gupta et al., 2006;
essentially the form discussed previously for static fields. Yan et al., 2007), and the study of the optical phonon
The coupling to acoustic modes is the most straight- modes in graphene, particularly the ones in the center of
forward one, since it already appears in the elastic the- the BZ with momentum q ≈ 0. Similar studies have been
ory. If uac is the acoustic phonon displacement, then performed in graphite ribbons (Cançado et al., 2004).
the relation between this displacement and the atom dis- Let us consider the effect of the Dirac fermions on
placement is given by equation (171), and its coupling the optical modes. If one treats the vector potential,
to electrons is given by the vector potential (175) in the electron-phonon coupling, (150) and (208) up to second
 38

order perturbation theory, its main effect is the polar- is the dimensionless electron-phonon coupling. Notice
ization of the electron system by creating electron-hole that (213) has a real and imaginary part. The real part
pairs. In the QED language, the creation electron-hole represents the actual shift in frequency, while the imagi-
pairs is called pair (electron/anti-electron) production. nary part gives the damping of the phonon mode due to
Pair production is equivalent to a renormalization of the pair production (see Fig. 35). There is a clear change in
phonon propagator by a self-energy that is proportional behavior depending whether µ is larger or smaller than
to the polarization function of the Dirac fermions. ω0 /2. For µ < ω0 /2 there is a decrease in the phonon
The renormalized phonon frequency, Ω0 (q), is given frequency implying that the lattice is softening, while for
by (Ando, 2006a, 2007a; Lazzeri and Mauri, 2006; Saha µ > ω0 /2 the lattice hardens. The interpretation for this
et al., 2007): effect is also given in Fig. 35. On the one hand, if the
frequency of the phonon is larger than twice the chemical
2β 2
Ω0 (q) ≈ ω0 − χ(q, ω0 ) , (209) potential, real electron-hole pairs are produced, leading
a2 ω0 to stronger screening of the electron-ion interaction and
where ω0 is the bare phonon frequency, and the electron- hence, to a softer phonon mode. At the same time the
phonon polarization function is given by: phonons become damped and decay. On the other hand,
X Z d2 k f [Es (k + q)]−f [Es0 (k)] if the frequency of the phonon is smaller than the twice
χ(q, ω) = (210)
, the chemical potential, the production of electron-hole
0
(2π)2 ω0 −Es (k + q)+Es0 (k)+iη
s,s =±1 pairs is halted by the Pauli principle and only virtual
where Es (q) is the Dirac fermion dispersion (s = +1 for excitations can be generated leading to polarization and
the upper band, and s = −1 for the lower band), and lattice hardening. In this case, there is no damping and
f [E] is the Fermi-Dirac distribution function. For Ra- the phonon is long lived. This amazing result has been
man spectroscopy, the response of interest is at q = 0 observed experimentally by Raman spectroscopy (Yan
where clearly only the interband processes such that et al., 2007). Electron-phonon coupling has also been
ss0 = −1 (that is, processes between the lower and up- investigated theoretically in the case of a finite magnetic
per cones) contribute. The electron-phonon polariza- field (Ando, 2007b; Goerbig et al., 2007). In this case,
tion function can be easily calculated using the linearized resonant coupling occurs due to the large degeneracy of
Dirac fermion dispersion (7) and the low energy density the Landau levels and different Raman transitions are
of states (15): expected as compared with the zero-field case. The cou-
√ Z pling of electrons to flexural modes on a free standing
6 3 vF Λ

1 graphene sheet was discussed in ref. (Mariani and von
χ(0, ω0 ) = 2 dEE (f [−E]−f [E])
πvF 0 ω0 +2E +iη Oppen, 2008).

1
− , (211)
ω−2E +iη
B. Electron-electron interactions
where we have introduced the cut-off momentum Λ (≈
1/a) so that the integral converges in the ultraviolet. At Of all disciplines of condensed matter physics, the
zero temperature, T = 0, we have f [E] = θ(µ − E) and study of electron-electron interactions is probably one
we assume electron doping, µ > 0, so that f [−E] = 1 (for of the most complex since it involves the understand-
the case of hole doping, µ < 0, is obtained by electron- ing of the behavior of a macroscopic number of variables.
hole symmetry). The integration in (211) gives: Hence, the problem of interacting systems is a field in
√ 
constant motion and we shall not try to give here a com-

6 3 ω0 ω0 /2 + µ
χ(0, ω0 ) = 2 vF Λ − µ + ln prehensive survey of the problem for graphene. Instead,
πvF 4 ω0 /2 − µ
we will focus on a small number of topics that are of
+ iπθ (ω0 /2 − µ))] , (212)
current discussion in the literature.
where the cut-off dependent term is a contribution com- Since graphene is a truly 2D system, it is informative
ing from the occupied states in the lower π band and to compare it with the more standard 2DEG that has
hence is independent of the value of the chemical poten- been studied extensively in the last 25 years since the
tial. This contribution can be fully incorporated into the development of heterostructures and the discovery of the
bare value of ω0 in (209). Hence the relative shift in the quantum Hall effect. At the simplest level, metallic sys-
phonon frequency can be written as: tems have two main kind of excitations: electron-hole
δω0 λ

µ ω0 /2 + µ
 pairs and collective modes such as plasmons.
≈− − + ln + iπθ (ω0 /2 − µ)(213)
, Electron-hole pairs are incoherent excitations of the
ω0 4 ω0 ω0 /2 − µ
Fermi sea and a direct result of Pauli’s exclusion princi-
where ple: an electron inside the Fermi sea with momentum k is
√
36 3 β2 excited outside the Fermi sea to a new state with momen-
λ= , (214) tum k + q, leaving a hole behind. The energy associated
π 8M a2 ω0
 39

Undoped, clean graphene is a semimetal, with a van-

FIG. 36 (Color online) Electron-hole continuum and collec-
ishing density of states at the Fermi level. As a result the
tive modes of: (a) a 2DEG; (b) undoped graphene; (c) doped
graphene. linear Fermi Thomas screening length diverges, and the
long range Coulomb interaction is not screened. At finite
electron density n, the Thomas-Fermi screening length
with such an excitation is simply: ω = k+q − k and for reads :
states close to the Fermi surface (k ≈ kF ) their energy 1 1 1 1
scales linearly with the excitation momentum, ωq ≈ vF q. λT F ≈ = √ , (216)
4α kF 4α πn
In a system with non-relativistic dispersion such as nor-
mal metals and semiconductors, the electron-hole contin- where
uum is made out of intra-band transitions only and exists
even at zero energy since it is always possible to produce e2
α= , (217)
electron-hole pairs with arbitrarily low energy close to 0 v F
the Fermi surface, as shown in Fig. 36(a). Besides that, is the dimensionless coupling constant in the problem
the 2DEG can also sustain collective excitations such as (the analogue of (143) in the Coulomb impurity prob-
√
plasmons that have dispersion: ωplasmon (q) ∝ q, and lem). Going beyond the linear Thomas-Fermi regime, it
exist outside the electron-hole continuum at sufficiently has been shown that the Coulomb law is modified (Fogler
long wavelengths (Shung, 1986a). et al., 2007; Katsnelson, 2006a; Zhang and Fogler, 2008).
In systems with relativistic-like dispersion, such as The Dirac Hamiltonian in the presence of interactions
graphene, these excitations change considerably, espe- can be written as:
cially when the Fermi energy is at the Dirac point. In this Z
case the Fermi surface shrinks to a point and hence intra- H ≡ −ivF d2 r Ψ̂† (r)σ · ∇Ψ̂(r)
band excitations disappear and only interband transi-
e2
Z
tions between the lower and upper cones can exist (see 1
+ d2 rd2 r0 ρ̂(r)ρ̂(r0 ) , (218)
Fig.36(b)). Therefore, neutral graphene has no electron- 20 |r − r0 |
hole excitations at low energy, instead each electron-hole
pair costs energy and hence the electron-hole occupies where
the upper part of the energy versus momentum diagram.
ρ̂(r) = Ψ̂† (r)Ψ̂(r) , (219)
In this case, plasmons are suppressed and no coherent
collective excitations can exist. If the chemical poten- is the electronic density. Observe that Coulomb in-
tial is moved away from the Dirac point then intra-band teraction, unlike in QED, is assumed to be instanta-
excitations are restored and the electron-hole continuum neous since vF /c ≈ 1/300 and hence retardation ef-
of graphene shares features of the 2DEG and undoped fects are very small. Moreover, the photons propagate
graphene. The full electron-hole continuum of doped in 3D space whereas the electrons are confined to the
graphene is shown in Fig. 36(c), and in this case plasmon 2D graphene sheet. Hence, the Coulomb interaction
modes are allowed. As the chemical potential is raised breaks the Lorentz invariance of the problem and makes
away from the Dirac point, graphene resembles more and the many-body situation rather different from the one
more the 2DEG. in QED. Furthermore, the problem depends on two pa-
These features in the elementary excitations of rameters, vF and e2 /0 . Under a dimensional scaling,
graphene reflect its screening properties as well. In r → λr, t → λt, Ψ → λ−1 Ψ, both parameters remain in-
fact, the polarization and dielectric functions of undoped variant. In RG language, the Coulomb interaction is a
graphene are rather different from the ones of the 2DEG marginal variable, whose strength relative to the kinetic
(Lindhard function). In the random phase approxima- energy does not change upon a change in scale. If the
tion (RPA), the polarization function can be calculated units are chosen in such a way that vF is dimensionless,
analytically (Shung, 1986a): the value of e2 /0 will also be rendered dimensionless.
q2 This is the case in theories considered renormalizable in
Π(q, ω) = p 2 , (215) quantum field theory.
4 vF q 2 − ω 2
The Fermi velocity in graphene is comparable to that
and hence, for ω > vF q the polarization function is imag- in half-filled metals. In solids with lattice constant a, the
inary indicating the damping of electron-hole pairs. No- total kinetic energy per site, 1/(ma2 ), where m is the
tice that the static polarization function (ω = 0) vanishes bare mass of the electron, is of the same order of mag-
linearly with q, indicating the lack of screening in the nitude as the electrostatic energy, e2 /(0 a). The Fermi
system. This polarization function has been also calcu- velocity for fillings of the order of unity is vF ∼ 1/(ma).
lated in the presence of a finite chemical potential (Ando, Hence, e2 /(0 vF ) ∼ 1. This estimate is also valid in
2006b; Hwang and Das Sarma, 2007; Shung, 1986a,b; graphene. Hence, unlike in QED, where αQED = 1/137,
Wunsch et al., 2006). the coupling constant in graphene is α ∼ 1.
 40

This analysis confirms that the Coulomb interaction is

FIG. 37 Hartree-Fock self-energy diagram which leads to a
logarithmic renormalization of the Fermi velocity.
renormalized towards lower values.
The enhancement in the Fermi velocities leads to a
widening of the electronic spectrum. This is consistent
Despite the fact that the coupling constant is of the with measurements of the gaps in narrow single wall nan-
order of unity, a perturbative RG analysis can be applied. otubes, which show deviations from the scaling with R−1 ,
RG techniques allow us to identify stable fixed points where R is the radius, expected from the Dirac equa-
of the model, which may be attractive over a broader tion (Kane and Mele, 2004). The linear dependence of
range than the one where a perturbative treatment can the inverse quasiparticle lifetime with energy is consistent
be rigorously justified. Alternatively, an RG scheme can with photo-emission experiments in graphite, for energies
be reformulated as the process of piecewise integration of larger with respect to the interlayer interactions (Sug-
high energy excitations (Shankar, 1994). This procedure awara et al., 2007; Xu et al., 1996). Note that in graphite,
leads to changes in the effective low energy couplings. band structure effects modify the lifetimes at low energies
The scheme is valid if the energy of the renormalized (Spataru et al., 2001). The vanishing of the quasiparti-
modes is much larger than the scales of interest. cle peak at low energies can lead to an energy dependent
The Hartree-Fock correction due to Coulomb interac- renormalization of the interlayer hopping (Vozmediano
tions between electrons (given by the diagram in Fig. 37) et al., 2002). Other thermodynamic properties of un-
gives a logarithmic correction to the electron self-energy: doped and doped graphene can also be calculated (Barlas
  et al., 2007; Vafek, 2007).
α Λ Non-perturbative calculations of the effects of the long
ΣHF (k) = k ln (220)
4 k range interactions in undoped graphene show that a tran-
sition to a gapped phase is also possible, when the num-
where Λ is a momentum cutoff which sets the range of ber of electronic flavors is large (Khveshchenko, 2001;
validity of the Dirac equation. This result remains true Luk’yanchuk and Kopelevich, 2004). The broken sym-
even to higher order in perturbation theory (Mishchenko, metry phase is similar to the excitonic transition found in
2007) and is also obtained in large N expansions (Rosen- materials where it becomes favorable to create electron-
stein et al., 1989, 1991; Son, 2007) (N is the number of hole pairs that then form bound excitons (excitonic tran-
flavors of Dirac fermions), with the only modification be- sition).
ing the prefactor in (220). This result implies that the Undoped graphene cannot have well defined plasmons,
Fermi velocity is renormalized towards higher values. As as their energies fall within the electron-hole continuum,
a consequence, the density of states near the Dirac en- and therefore have a significant Landau damping. At fi-
ergy is reduced, in agreement with the general trend of nite temperatures, however, thermally excited quasipar-
repulsive interactions to induce or increase gaps. ticles screen the Coulomb interaction, and an acoustic
This result can be understood from the RG point of collective charge excitation can exist (Vafek, 2006).
view by studying the effect of reducing the cut-off from Doped graphene shows a finite density of states at the
Λ to Λ − dΛ and its effect on the effective coupling. It Fermi level, and the long range Coulomb interaction is
can be shown that α obeys the equation : screened. Accordingly, there are collective plasma p in-
∂α α teractions near q → 0, which disperse as ωp ∼ |q|,
Λ =− . (221) since the system is 2D (Campagnoli and Tosatti, 1989;
∂Λ 4
Shung, 1986a,b). The fact that the electronic states are
Therefore, the Coulomb interaction becomes marginally described by the massless Dirac equation implies that
irrelevant. These features are confirmed by a full rela- ωP ∝ n1/4 , where n is the carrier density. The static
tivistic calculation, although the Fermi velocity cannot, dielectric constant has a continuous derivative at 2kF ,
obviously, surpass the velocity of light. This result indi- unlike in the case of the 2D electron gas (Ando, 2006b;
cates that strongly correlated electronic phases, such as Das Sarma et al., 2007; Wunsch et al., 2006). This
ferromagnetism (Peres et al., 2005) and Wigner crystals fact is associated with the suppressed backward scat-
(Dahal et al., 2006) are suppressed in clean graphene. tering in graphene. The simplicity of the band struc-
A calculation of higher order self-energy terms leads ture of graphene allows analytical calculation of the en-
to a wavefunction renormalization, and to a finite quasi- ergy and momentum dependence of the dielectric func-
particle lifetime, which grows linearly with quasiparticle tion (Das Sarma et al., 2007; Wunsch et al., 2006). The
energy (González et al., 1996). The wavefunction renor- screening of the long-range Coulomb interaction implies
malization implies that the quasiparticle weight tends to that the low energy quasiparticles show a quadratic de-
zero as its energy is reduced. A strong coupling expan- pendence on energy with respect to the Fermi energy
sion is also possible, assuming that the number of elec- (Hwang et al., 2007).
tronic flavors justifies an RPA expansion, keeping only One way to probe the strength of the electron-electron
electron-hole bubble diagrams (González et al., 1999). interactions is via the electronic compressibility. Mea-
 41

surements of the compressibility using a single electron to twice the distance between layers. This effect greatly
transistor (SET) show very little sign of interactions in enhances Friedel like oscillations in the charge distribu-
the system, being well fitted by the non-interacting re- tion in the out of plane direction, which can lead to the
sult that, contrary to the two-dimensional electron gas changes in the sign of the charge in neighboring layers
(2DEG) (Eisenstein et al., 1994; Giovani and Vignale, (Guinea, 2007). Away from half-filling a graphene bi-
2005), is positively divergent (Polini et al., 2007). Bi- layer behaves, from the point of view of screening, in a
layer graphene, on the other hand, shares characteristics way very similar to the 2DEG (Wang and Chakraborty,
of the single layer and the 2DEG with a non-monotonic 2007b).
dependence of the compressibility on the carrier density
(Kusminskiy et al., 2008). In fact, bilayer graphene very
close to half-filling has been predicted to be unstable to- C. Short range interactions
wards Wigner crystallization (Dahal et al., 2007), just
like the 2DEG. Furthermore, according to Hartree-Fock In this section we discuss the effect of short range
calculations, clean bilayer graphene is unstable towards Coulomb interactions on the physics of graphene. The
ferromagnetism (Nilsson et al., 2006a). simplest carbon system with a hexagonal shape is the
benzene molecule. The value of the Hubbard interaction
among π-electrons was, for this system, computed long
1. Screening in graphene stacks ago by Parr et al. (Parr et al., 1950), yielding a value of
U = 16.93 eV. For comparison purposes, in polyacetylene
The electron-electron interaction leads to the screen- the value for the Hubbard interaction is U '10 eV and
ing of external potentials. In a doped stack, the charge the hopping energy is t ≈ 2.5 eV (Baeriswyl et al., 1986).
tends to accumulate near the surfaces, and its distribu- These two examples just show that the value of the on-
tion is determined by the dielectric function of the stack site Coulomb interaction is fairly large for π−electrons.
in the out-of-plane direction. The same polarizability As a first guess for graphene, one can take U to be of the
describes the screening of an external field perpendicular same order as for polyacethylene, with the hopping inte-
to the layers, like the one induced by a gate in electri- gral t ' 2.8 eV. Of course in pure graphene the electron-
cally doped systems (Novoselov et al., 2004). The self- electron interaction is not screened, since the density of
consistent distribution of charge in a biased graphene bi- states is zero at the Dirac point, and one should work
layer has been studied in ref. (McCann, 2006). From the out the effect of Coulomb interactions by considering the
observed charge distribution and self-consistent calcula- bare Coulomb potential. On the other hand, as we have
tions, an estimate of the band structure parameters and seen before, defects induce a finite density of states at the
their relation with the induced gap can be obtained. Dirac point, which could lead to an effective screening of
In the absence of interlayer hopping, the polarizability the long-range Coulomb interaction. Let us assume that
of a set of stacks of 2D electron gases can be written as the bare Coulomb interaction is screened in graphene and
a sum of the screening by the individual layers. Using that Coulomb interactions are represented by the Hub-
the accepted values for the effective masses and carrier bard interaction. This means that we must add to the
densities of graphene, this scheme gives a first approx- Hamiltonian (5) a term of the form:
imation to screening in graphite (Visscher and Falicov, Xh †
HU = U a↑ (Ri )a↑ (Ri )a†↓ (Ri )a↓ (Ri )
1971). The screening length in the out of plane direction
Ri
is of about 2 graphene layers (Morozov et al., 2005). This i
model is easily generalizable to a stack of semimetals de- + b†↑ (Ri )b↑ (Ri )b†↓ (Ri )b↓ (Ri ) (222)
scribed by the 2D Dirac equation (González et al., 2001).
At half filling, the screening length in all directions di- The simplest question one can ask is whether this system
verges, and the screening effects are weak. shows a tendency toward some kind of magnetic order
Interlayer hopping modifies this picture significantly. driven by the interaction U . Within the simplest Hartree-
The hopping leads to coherence (Guinea, 2007). The out Fock approximation (Peres et al., 2004), the instability
of plane electronic dispersion is similar to that of a one line toward ferromagnetism is given by:
dimensional conductor. The out of plane polarizability 2
of a multilayer contains intra- and interband contribu- UF (µ) = , (223)
ρ(µ)
tions. The subbands in a system with the Bernal stack-
ing have a parabolic dispersion, when only the nearest which is nothing but the Stoner criterion. Similar results
neighbor hopping terms are included. This band struc- are obtained in more sophisticated calculations (Herbut,
ture leads to an interband susceptibility described by a 2006). Clearly, at half-filling the value for the density of
sum of terms like those in (228), which diverges at half- states is ρ(0) = 0 and the critical value for UF is arbi-
filling. In an infinite system, this divergence is more pro- trarily large. Therefore we do not expect a ferromagnetic
nounced at k⊥ = π/c, that is, for a wave vector equal ground state at the neutrality point of one electron per
 42

carbon atom. For other electronic densities, ρ(µ) be- Writing n↑ = (n + s)/2, n↓ = (n − s)/2, where s is the
comes finite producing a finite value for UF . We note magnetization, (227) predicts a second order transition to
that the inclusion of t0 does not change these findings, a ferromagnetic state for n = (4e4 t2 )/(81π 3 vF4 0 ). Higher
since the density of states remains zero at the neutrality order corrections to (227) lead to a first order transition
point. at slightly higher densities (Nomura et al., 2007). For a
The line toward an antiferromagnetic ground state is ratio γ1 /γ0 ≈ 0.1, this analysis implies that a graphene
given by (Peres et al., 2004) bilayer should be ferromagnetic for carrier densities such
that |n| . 4 × 1010 cm−2 .
2
UAF (µ) = 1
P 1 , (224) A bilayer is also the unit cell of Bernal graphite, and
N k,µ>0 |E+ (k)|
the exchange instability can also be studied in an infi-
where E+ (k) is given in (6). This result gives a finite nite system. Taking into account nearest neighbor in-
UAF at the neutrality point (Martelo et al., 1997; Sorella terlayer hopping only, bulk graphite should also show an
and Tosatti, 1992): exchange instability at low doping. In fact, there is some
experimental evidence for a ferromagnetic instability in
UAF (0) = 2.23t . (225) strongly disordered graphite (Esquinazi et al., 2002, 2003;
Kopelevich and Esquinazi, 2007).
Quantum Monte Carlo calculations (Paiva et al., 2005;
The analysis described above can be extended to the
Sorella and Tosatti, 1992), raise however its value to:
biased bilayer, where a gap separates the conduction and
UAF (0) ' 5t . (226) valence bands (Stauber et al., 2007). The analysis of this
case is somewhat different, as the Fermi surface at low
Taking for graphene the same value for U as in poly- doping is a ring, and the exchange interaction can change
acetylene and t = 2.8 eV, one obtains U/t ' 3.6, which its bounds. The presence of a gap reduces further the
put the system far from the transition toward an antifer- mixing of the valence and conduction band, leading to an
romagnet ground state. Yet another possibility is that enhancement of the exchange instability. At all doping
the system may be in a sort of a quantum spin liquid levels, where the Fermi surface is ring shaped, the biased
(Lee and Lee, 2005) (as originally proposed by Pauling bilayer is unstable towards ferromagnetism.
(Partoens and Peeters, 2006) in 1956) since mean field
calculations give a critical value for U to be of the or-
der of U/t ' 1.7. Whether this type of ground state
really exists and whether quantum fluctuations pushes
2. Bilayer graphene: short range interactions
this value of U toward larger values is not known.

The band structure of a graphene bilayer, at half fill-

1. Bilayer graphene: exchange ing, leads to logarithmic divergences in different response
functions at q = 0. The two parabolic bands that are
The exchange interaction can be large in an unbiased tangent at k = 0 lead to a susceptibility which is propor-
graphene bilayer with a small concentration of carriers. tional to:
It was shown previously that the exchange contribution
to the electronic energy of a single graphene layer does Z !
2 1 Λ
not lead to a ferromagnetic instability (Peres et al., 2005). χ(~q, ω) ∝ d k ∝ log
ω − (vF2 /t)|k|2
p
The reason for this is a significant contribution from the |~
q |<Λ (ωt)/vF2
interband exchange, which is a term usually neglected in p (228)
doped semiconductors. This contribution depends on the where Λ ∼ t2 /vF2 is a high momentum cutoff. These
overlap of the conduction and valence wavefunctions, and logarithmic divergences are similar to the ones which
it is modified in a bilayer. The interband exchange en- show up when the Fermi surface of a 2D metal is near a
ergy is reduced in a bilayer (Nomura et al., 2007), and a saddle point in the dispersion relation (González et al.,
positive contribution that depends logarithmically on the 1996). A full treatment of these divergences requires a
bandwidth in graphene is absent in its bilayer. As a re- RG approach (Shankar, 1994). Within a simpler mean
sult, the exchange energy becomes negative, and scales as field treatment, however, it is easy to notice that the di-
n3/2 , where n is the carrier density, similar to the 2DEG. vergence of the bilayer susceptibility gives rise to an in-
The quadratic dispersion at low energies implies that the stability towards an antiferromagnetic phase, where the
kinetic energy scales as n2 , again as in the 2DEG. This carbon atoms which are not connected to the neighbor-
expansion leads to: ing layers acquire a finite magnetization, while the mag-
netization of the atoms with neighbors in the contiguous
πvF2 n2 e2 n3/2 layers remain zero. A scheme of the expected ordered
E = Ekin + Eexc ≈ − √ (227)
8t⊥ 27 π0 state is shown in Fig. 38.
 43

VI. CONCLUSIONS
FIG. 38 (Color online) Sketch of the expected magnetization
of a graphene bilayer at half-filling.
Graphene is a unique system in many ways. It is truly
2D, has unusual electronic excitations described in terms
of Dirac fermions that move in a curved space, it is an in-
teresting mix of a semiconductor (zero density of states)
D. Interactions in high magnetic fields
and a metal (gaplessness), and has properties of soft mat-
ter. The electrons in graphene seem to be almost insen-
The formation of Landau levels enhances the effect of sitive to disorder and electron-electron interactions and
interactions due to the quenching of the kinetic energy. have very long mean free paths. Hence, graphene’s prop-
This effect is most pronounced at low fillings, when only erties are rather different from what is found in usual
the lowest levels are occupied. New phases may appear metals and semiconductors. Graphene has also a very
at low temperatures. We consider here phases different robust but flexible structure with unusual phonon modes
from the fractional quantum Hall effect, which has not that do not exist in ordinary 3D solids. In some sense,
been observed in graphene so far. The existence of new graphene brings together issues in quantum gravity and
phases can be inferred from the splitting of the valley or particle physics, and also from soft and hard condensed
spin degeneracy of the Landau levels, which can be ob- matter. Interestingly enough, these properties can be
served in spectroscopy measurements (Jiang et al., 2007a; easily modified with the application of electric and mag-
Sadowski et al., 2006), or in the appearance of new quan- netic fields, addition of layers, by control of its geometry,
tum Hall plateaus (Abanin et al., 2007b; Goswami et al., and chemical doping. Moreover, graphene can be directly
2007; Jiang et al., 2007b; Zhang et al., 2006). and relatively easily probed by various scanning probe
techniques from mesoscopic down to atomic scales, be-
Interactions can lead to new phases when their effect cause it is not buried inside a 3D structure. This makes
overcomes that of disorder. An analysis of the com- graphene one of the most versatile systems in condensed
petition between disorder and interactions is found in matter research.
ref. (Novikov, 2007a). The energy splitting of the differ- Besides the unusual basic properties, graphene has the
ent broken symmetry phases, in a clean system, is deter- potential for a large number of applications (Geim and
mined by lattice effects, so that it is reduced by factors Novoselov, 2007), from chemical sensors (Chen et al.,
of order a/lB , where a is a length of the order of the lat- 2007b) to transistors (Nilsson et al., 2008). Graphene
tice spacing, and lB is the magnetic length (Alicea and can be chemically and/or structurally modified in order
Fisher, 2006, 2007; González et al., 1992; Wang et al., to change its functionality and henceforth its potential
2008). The combination of disorder and a magnetic field applications. Moreover, graphene can be easily obtained
may also lift the degeneracy between the two valleys, fa- from graphite, a material that is abundant on earth’s
voring valley polarized phases (Abanin et al., 2007a). surface. This particular characteristic makes graphene
one of the most readily available materials for basic re-
In addition to phases with enhanced ferromagnetism or
search since it frees economically challenged research in-
with broken valley symmetry, interactions at high mag-
stitutions in developing countries from the dependence of
netic fields can lead to excitonic instabilities (Gusynin
expensive sample growing techniques.
et al., 2006) and Wigner crystal phases (Zhang and
Joglekar, 2007). When only the n = 0 state is occu- Many of graphene’s properties are currently subject of
pied, the Landau levels have all their weight in a given intense research and debate. The understanding of the
sublattice. Then, the breaking of valley degeneracy can nature of the disorder and how it affects the transport
be associated with a charge density wave, which opens a properties (a problem of fundamental importance for ap-
gap (Fuchs and Lederer, 2007). It is interesting to note plications), the effect of phonons on electronic transport,
that in these phases new collective excitations are possi- the nature of electron-electron interactions and how they
ble (Doretto and Smith, 2007). modify its physical properties are research areas that are
still in their infancy. In this review, we have touched
Interactions modify the edge states in the quantum only the surface of a very deep sea that still has to be
Hall regime. A novel phase can appear when the n = 0 explored.
is the last filled level. The Zeeman splitting shifts the Whereas hundreds of papers have been written on
electron and hole like chiral states, which disperse in op- monolayer graphene in the last few years, only a small
posite directions near the boundary of the sample. The fraction actually deals with multilayers. The majority of
resulting level crossing between an electron like level with the theoretical and experimental efforts have been con-
spin anti-parallel to the field, and a hole like level with centrated on the single layer, perhaps because of its sim-
spin parallel to the field, may lead to Luttinger liquid plicity and the natural attraction that a one atom thick
features in the edge states (Abanin et al., 2007b; Fertig material, which can be produced by simple methods in
and Brey, 2006). almost any laboratory in the world, creates for human
 44

imagination. Nevertheless, few layer graphene is equally 317 (5838), 641–643.

interesting and unusual with a technological potential Abanin, Dmitry A, Patrick A. Lee, and Leonid S. Levitov
perhaps bigger than the single layer. Indeed, the the- (2006), “Spin-filtered edge states and quantum hall effect
oretical understanding and experimental exploration of in graphene,” Phys. Rev. Lett. 96, 176803.
Abanin, Dmitry A, Patrick A. Lee, and Leonid S. Levitov
multilayers is far behind the single layer. This is a fertile (2007a), “Charge and spin transport at the quantum hall
and open field of research for the future. edge of graphene,” Solid State Communications 143, 77 –
Finally, we have focused entirely on pure carbon 85.
graphene where the band structure is dominated by the Abanin, Dmitry A, Kostya S. Novoselov, Uli Zeitler,
Dirac description. Nevertheless, chemical modification Patrick A. Lee, A. K. Geim, and L. S. Levitov (2007b),
of graphene can lead to entirely new physics. Depend- “Dissipative quantum hall effect in graphene near the dirac
ing on the nature of chemical dopants and how they are point,” Phys. Rev. Lett. 98, 196806.
Abedpour, N, M. Neek-Amal, Reza Asgari, F. Shahbazi,
introduced into the graphene lattice (adsorption, substi- N. Nafari, and M. Reza Rahimi Tabar (2007), “Roughness
tution, or intercalation) the results can be many. Small of undoped graphene and its short-range induced gauge
concentrations of adsorbed alkali metal can be used to field,” Phys. Rev. B 76, 195407.
change the chemical potential while adsorbed transition Abergel, D S L, A. Russell, and Vladimir I. Fal’ko (2007),
elements can lead to strong hybridization effects that af- “Visibility of graphene flakes on a dielectric substrate,” Ap-
fect the electronic structure. In fact, the introducion of d- plied Physics Letters 91 (6), 063125.
and f-electron atoms in the graphene lattice may produce Abrikosov, A A (1998), “Quantum magnetoresistance,” Phys.
Rev. B 58, 2788–2794.
a significant enhancement of the electron-electron inter- Adam, Shaffique, E. H. Hwang, V. M. Galitski, and
actions. Hence, it is easy to envision a plethora of many- S. Das Sarma (2007), “A self-consistent theory for graphene
body effects that can be induced by doping and have transport,” Proceedings of the National Academy of Sci-
to be studied in the context of Dirac electrons: Kondo ences 104 (47), 18392–18397.
effect, ferromagnetism, antiferromagnetism, charge and Affoune, AM, B.L.V Prasad, Hirohiko Sato, Toshiaki Enoki,
spin density waves. The study of chemically induced Yutaka Kaburagi, and Yoshihiro Hishiyama (2001), “Ex-
many-body effects in graphene would add a new chap- perimental evidence of a single nano-graphene,” Chemical
Physics Letters 348 (1–2), 17–20.
ter to the short but fascinating history of this material. Akhmerov, A R, and C. W. J. Beenakker (2008), “Boundary
Only future will tell but the potential for more amaze- conditions for dirac fermions on a terminated honeycomb
ment is lurking on the horizon. lattice,” Phys. Rev. B 77, 085423.
Aleiner, I L, and K. B. Efetov (2006), “Effect of disorder on
transport in graphene,” Phys. Rev. Lett. 97, 236801.
VII. ACKNOWLEDGMENTS Alicea, Jason, and Matthew P. A. Fisher (2006), “Graphene
integer quantum hall effect in the ferromagnetic and para-
We have benefited immensely from discussions with magnetic regimes,” Phys. Rev. B 74, 075422.
many colleagues and friends in the last few years but Alicea, Jason, and Matthew P.A. Fisher (2007), “Interplay
between lattice-scale physics and the quantum hall effect
we would like to thank especially, Boris Altshuler, Eva
in graphene,” Solid State Communications 143 (11–12),
Andrei, Alexander Balatsky, Carlo Beenakker, Sankar 504–509.
Das Sarma, Walt de Heer, Millie Dresselhaus, Vladimir Altland, Alexander (2006), “Low-energy theory of disordered
Falko, Andrea Ferrari, Herb Fertig, Eduardo Fradkin, graphene,” Phys. Rev. Lett. 97, 236802.
Ernie Hill, Mihail Katsnelson, Eun-Ah Kim, Philip Kim, Ando, Tsuneya (1974a), “Theory of quantum transport in a
Valery Kotov, Alessandra Lanzara, Leonid Levitov, Al- two-dimensional electron system under magnetic fields. iii.
lan MacDonald, Serguey Morozov, Johan Nilsson, Vitor many-site approximation,” Journal of the Physical Society
of Japan 37 (3), 622–630.
Pereira, Philip Phillips, Ramamurti Shankar, João Lopes Ando, Tsuneya (1974b), “Theory of quantum transport in a
dos Santos, Shan-Wen Tsai, Bruno Uchoa, and Maria two-dimensional electron system under magnetic fields. iv.
Vozmediano. oscillatory conductivity,” Journal of the Physical Society of
N.M.R.P. acknowledges financial support from POCI Japan 37 (5), 1233–1237.
2010 via project PTDC/FIS/64404/2006. F.G. was sup- Ando, Tsuneya (1975), “Theory of cyclotron resonance line-
ported by MEC (Spain) grant No. FIS2005-05478-C02-01 shape in a two-dimensional electron system,” Journal of the
and EU contract 12881 (NEST). A. H. C. N was sup- Physical Society of Japan 38 (4), 989–997.
Ando, Tsuneya (2000), “Spin-orbit interaction in carbon nan-
ported through NSF grant DMR-0343790. K.S.N. and otubes,” Journal of the Physical Society of Japan 69 (6),
A. K. G. were supported by EPSRC (UK) and the Royal 1757–1763.
Society. Ando, Tsuneya (2006a), “Anomaly of optical phonon in
monolayer graphene,” Journal of the Physical Society of
Japan 75 (12), 124701.
REFERENCES Ando, Tsuneya (2006b), “Screening effect and impurity scat-
tering in monolayer graphene,” Journal of the Physical So-
Abanin, D A, and L. S. Levitov (2007), “Quantized trans- ciety of Japan 75 (7), 074716.
port in graphene p-n junctions in a magnetic field,” Science Ando, Tsuneya (2007a), “Anomaly of optical phonons in bi-
 45

layer graphene,” Journal of the Physical Society of Japan a time-of-flight experiment with conduction electrons,”
76 (10), 104711. Physics Reports 107 (1), 1–58.
Ando, Tsuneya (2007b), “Magnetic oscillation of optical Bernevig, B Andrei, Taylor L. Hughes, Srinivas Raghu, and
phonon in graphene,” Journal of the Physical Society of Daniel P. Arovas (2007), “Theory of the three-dimensional
Japan 76 (2), 024712. quantum hall effect in graphite,” Phys. Rev. Lett. 99,
Ando, Tsuneya, Takeshi Nakanishi, and Riichiro Saito 146804.
(1998), “Berry’s phase and absence of back scattering Berry, M V, and R. J. Mondragon (1987), “Neutrino bil-
in carbon nanotubes,” Journal of the Physical Society of liards: Time-reversal symmetry-breaking without magnetic
Japan 67 (8), 2857–2862. fields,” Proceedings of the Royal Society of London. A.
Ando, Tsuneya, and Yasutada Uemura (1974), “Theory of Mathematical and Physical Sciences 412 (1842), 53–74.
quantum transport in a two-dimensional electron system Bhattacharjee, Subhro, and K. Sengupta (2006), “Tunneling
under magnetic fields. i. characteristics of level broadening conductance of graphene nis junctions,” Phys. Rev. Lett.
and transport under strong fields,” Journal of the Physical 97, 217001.
Society of Japan 36 (4), 959–967. Birrell, ND, and P.C.W. Davies (1982), Quantum Fields in
Ando, Tsuneya, Yisong Zheng, and Hidekatsu Suzuura Curved Space (Cambridge Univ. Press, Cambridge).
(2002), “Dynamical conductivity and zero-mode anomaly Biswas, Rudro R, Subir Sachdev, and Dam T. Son (2007),
in honeycomb lattices,” Journal of the Physical Society of “Coulomb impurity in graphene,” Phys. Rev. B 76, 205122.
Japan 71 (5), 1318–1324. Blake, P, E. W. Hill, A. H. Castro Neto, K. S. Novoselov,
Andreoni, W (2000), The Physics of Fullerene-Based and D. Jiang, R. Yang, T. J. Booth, and A. K. Geim (2007),
Fullerene-related materials (Springer). “Making graphene visible,” Applied Physics Letters 91 (6),
Areshkin, Denis A, Daniel Gunlycke, and Carter T. White 063124.
(2007), “Ballistic transport in graphene nanostrips in the Bliokh, KYu (2005), “Weak antilocalization of ultrarelativis-
presence of disorder: Importance of edge effects,” Nano tic fermions,” Physics Letters A 344, 127 – 130.
Letters 7 (1), 204–210, pMID: 17212465. Bommel, AJ Van, J.E. Crombeen, and A. Van Tooren (1975),
Ashcroft, NW, and N.D. Mermin (1976), Solid State Physics “Leed and auger electron observations of the sic(0001) sur-
(Saunders College, Philadelphia, PA). face,” Surface Science 48 (2), 463 – 472.
Baeriswyl, Dionys, David K. Campbell, and Sumit Mazum- Bonini, Nicola, Michele Lazzeri, Nicola Marzari, and
dar (1986), “Correlations and defect energies,” Phys. Rev. Francesco Mauri (2007), “Phonon anharmonicities in
Lett. 56, 1509–1509. graphite and graphene,” Phys. Rev. Lett. 99, 176802.
Bak, P (1982), “Commensurate phases, incommensurate Booth, Tim J, Peter Blake, Rahul R. Nair, Da Jiang,
phases and the devil’s staircase,” Reports on Progress in Ernie W. Hill, Ursel Bangert, Andrew Bleloch, Mhairi
Physics 45 (6), 587. Gass, Kostya S. Novoselov, M. I. Katsnelson, and A. K.
Balatsky, A V, I. Vekhter, and Jian-Xin Zhu (2006), Geim (2008), “Macroscopic graphene membranes and their
“Impurity-induced states in conventional and unconven- extraordinary stiffness,” Nano Letters 8 (8), 2442–2446,
tional superconductors,” Rev. Mod. Phys. 78, 373–433. pMID: 18593201.
Bardarson, J H, J. Tworzydlo, P. W. Brouwer, and C. W. J. Boyle, W S, and P. Nozières (1958), “Band structure and
Beenakker (2007), “One-parameter scaling at the dirac infrared absorption of graphite,” Phys. Rev. 111, 782–785.
point in graphene,” Phys. Rev. Lett. 99, 106801. Brey, L, and H. A. Fertig (2006a), “Electronic states of
Barlas, Yafis, T. Pereg-Barnea, Marco Polini, Reza Asgari, graphene nanoribbons studied with the dirac equation,”
and A. H. MacDonald (2007), “Chirality and correlations Phys. Rev. B 73, 235411.
in graphene,” Phys. Rev. Lett. 98, 236601. Brey, L, H. A. Fertig, and S. Das Sarma (2007), “Diluted
Barone, Veronica, Oded Hod, and Gustavo E. Scuseria graphene antiferromagnet,” Phys. Rev. Lett. 99, 116802.
(2006), “Electronic structure and stability of semiconduct- Brey, Luis, and H. A. Fertig (2006b), “Edge states and the
ing graphene nanoribbons,” Nano Letters 6 (12), 2748– quantized hall effect in graphene,” Phys. Rev. B 73, 195408.
2754, pMID: 17163699. Bunch, J Scott, Arend M. van der Zande, Scott S. Verbridge,
Baym, G (1969), Lectures on Quantum Mechanics (Benjamin, Ian W. Frank, David M. Tanenbaum, Jeevak M. Parpia,
New York). Harold G. Craighead, and Paul L. McEuen (2007), “Elec-
Bena, Cristina, and Steven A. Kivelson (2005), “Quasiparti- tromechanical resonators from graphene sheets,” Science
cle scattering and local density of states in graphite,” Phys. 315 (5811), 490–493.
Rev. B 72, 125432. Calandra, Matteo, and Francesco Mauri (2007), “Publisher’s
Berger, Claire, Zhimin Song, Tianbo Li, Xuebin Li, As- note: Electronic structure of heavily doped graphene: The
merom Y. Ogbazghi, Rui Feng, Zhenting Dai, Alexei N. role of foreign atom states [phys. rev. b 76, 161406 (2007)],”
Marchenkov, Edward H. Conrad, Phillip N. First, and Phys. Rev. B 76, 199901.
Walt A. de Heer (2004), “Ultrathin epitaxial graphite: 2d Calizo, Irene, Wenzhong Bao, Feng Miao, Chun Ning Lau,
electron gas properties and a route toward graphene-based and Alexander A. Balandin (2007), “The effect of sub-
nanoelectronics,” The Journal of Physical Chemistry B strates on the raman spectrum of graphene: Graphene-
108 (52), 19912–19916. on-sapphire and graphene-on-glass,” Applied Physics Let-
Berger, Claire, Zhimin Song, Xuebin Li, Xiaosong Wu, Nate ters 91 (20), 201904.
Brown, Cécile Naud, Didier Mayou, Tianbo Li, Joanna Calogeracos, A, and N. Dombey (1999), “History and physics
Hass, Alexei N. Marchenkov, Edward H. Conrad, Phillip N. of the klein paradox,” Contemporary Physics 40 (5), 313–
First, and Walt A. de Heer (2006), “Electronic confinement 321.
and coherence in patterned epitaxial graphene,” Science Campagnoli, G, and E. Tosatti (1989), Progress on Electron
312 (5777), 1191–1196. Properties of Metals, edited by R. Girlinda (Kluwer Aca-
Bergmann, Gerd (1984), “Weak localization in thin films: demic Publishing).
 46

Casiraghi, C, A. Hartschuh, E. Lidorikis, H. Qian, H. Haru- Coraux, Johann, Alpha T. N‘Diaye, Carsten Busse, and
tyunyan, T. Gokus, K. S. Novoselov, and A. C. Ferrari Thomas Michely (2008), “Structural coherency of graphene
(2007), “Rayleigh imaging of graphene and graphene lay- on ir(111),” Nano Letters 8 (2), 565–570, pMID: 18189442.
ers,” Nano Letters 7 (9), 2711–2717. Cortijo, A, and M. A. H. Vozmediano (2007a), “Electronic
Cassanello, Carlos R, and Eduardo Fradkin (1996), “Kondo properties of curved graphene sheets,” EPL (Europhysics
effect in flux phases,” Phys. Rev. B 53, 15079–15094. Letters) 77 (4), 47002.
Cassanello, Carlos R, and Eduardo Fradkin (1997), “Over- Cortijo, Alberto, and Marı̀a A.H. Vozmediano (2007b), “Ef-
screening of magnetic impurities in dx2 −y2 -wave supercon- fects of topological defects and local curvature on the elec-
ductors,” Phys. Rev. B 56, 11246–11261. tronic properties of planar graphene,” Nuclear Physics B
Castillo, Horacio E, Claudio de C. Chamon, Eduardo Frad- 763 (3), 293–308.
kin, Paul M. Goldbart, and Christopher Mudry (1997), Cserti, József, András Csordás, and Gyula Dávid (2007a),
“Exact calculation of multifractal exponents of the criti- “Role of the trigonal warping on the minimal conductivity
cal wave function of dirac fermions in a random magnetic of bilayer graphene,” Phys. Rev. Lett. 99, 066802.
field,” Phys. Rev. B 56, 10668–10677. Cserti, József, András Pályi, and Csaba Péterfalvi (2007b),
Castro, Eduardo V, N. M. R. Peres, J. M. B. Lopes dos San- “Caustics due to a negative refractive index in circular
tos, A. H. Castro Neto, and F. Guinea (2008), “Localized graphene p-n junctions,” Phys. Rev. Lett. 99, 246801.
states at zigzag edges of bilayer graphene,” Phys. Rev. Lett. Dahal, H P, T. O. Wehling, K. S. Bedell, J.-X. Zhu, and
100, 026802. A. V. Balatsky (2007), “Wigner crystallization in a single
Castro Neto, A H, F. Guinea, and N. M. R. Peres (2006), and bilayer graphene,” arXiv:0706.1689 [[Link]-el].
“Edge and surface states in the quantum hall effect in Dahal, Hari P, Yogesh N. Joglekar, Kevin S. Bedell, and
graphene,” Phys. Rev. B 73, 205408. Alexander V. Balatsky (2006), “Absence of wigner crystal-
Cançado, L G, M. A. Pimenta, B. R. A. Neves, G. Medeiros- lization in graphene,” Phys. Rev. B 74, 233405.
Ribeiro, Toshiaki Enoki, Yousuke Kobayashi, Kazuyuki Das Sarma, S, E. H. Hwang, and Wang-Kong Tse (2007),
Takai, Ken-ichi Fukui, M. S. Dresselhaus, R. Saito, and “Many-body interaction effects in doped and undoped
A. Jorio (2004), “Anisotropy of the raman spectra of graphene: Fermi liquid versus non-fermi liquid,” Phys. Rev.
nanographite ribbons,” Phys. Rev. Lett. 93, 047403. B 75, 121406.
Chakravarty, Sudip, and Albert Schmid (1986), “Weak local- Deacon, R S, K.-C. Chuang, R. J. Nicholas, K. S. Novoselov,
ization: The quasiclassical theory of electrons in a random and A. K. Geim (2007), “Cyclotron resonance study of the
potential,” Physics Reports 140 (4), 193 – 236. electron and hole velocity in graphene monolayers,” Phys.
Chamon, Claudio de C, Christopher Mudry, and Xiao-Gang Rev. B 76, 081406.
Wen (1996), “Instability of the disordered critical points of Dell’Anna, Luca (2006), “Disordered d-wave superconductors
dirac fermions,” Phys. Rev. B 53, R7638–R7641. with interactions,” Nuclear Physics B 758 (3), 255 – 303.
Charlier, J-C, J.-P. Michenaud, X. Gonze, and J.-P. Vigneron Dharma-wardana, M W C (2007), “Electronic structure and
(1991), “Tight-binding model for the electronic properties the minimum conductance of a graphene layer on sio 2
of simple hexagonal graphite,” Phys. Rev. B 44, 13237– from density functional methods,” Journal of Physics: Con-
13249. densed Matter 19 (38), 386228.
Charlier, Jean-Christophe, Xavier Blase, and Stephan Roche Dienwiebel, Martin, Gertjan S. Verhoeven, Namboodiri
(2007), “Electronic and transport properties of nanotubes,” Pradeep, Joost W. M. Frenken, Jennifer A. Heimberg,
Rev. Mod. Phys. 79, 677–732. and Henny W. Zandbergen (2004), “Superlubricity of
Cheianov, Vadim V, Vladimir Fal’ko, and B. L. Altshuler graphite,” Phys. Rev. Lett. 92, 126101.
(2007a), “The focusing of electron flow and a veselago lens Dillon, RO, I.L. Spain, and J.W. McClure (1977), “Elec-
in graphene p-n junctions,” Science 315 (5816), 1252–1255. tronic energy band parameters of graphite and their de-
Cheianov, Vadim V, and Vladimir I. Fal’ko (2006), “Selec- pendence on pressure, temperature and acceptor concen-
tive transmission of dirac electrons and ballistic magne- tration,” Journal of Physics and Chemistry of Solids 38 (6),
toresistance of n-p junctions in graphene,” Phys. Rev. B 635 – 645.
74, 041403. DiVincenzo, D P, and E. J. Mele (1984), “Self-consistent
Cheianov, Vadim V, Vladimir I. Fal’ko, Boris L. Altshuler, effective-mass theory for intralayer screening in graphite
and Igor L. Aleiner (2007b), “Random resistor network intercalation compounds,” Phys. Rev. B 29, 1685–1694.
model of minimal conductivity in graphene,” Phys. Rev. Dong, Shi-Hai, Xi-Wen Hou, and Zhong-Qi Ma (1998), “Rel-
Lett. 99, 176801. ativistic levinson theorem in two dimensions,” Phys. Rev.
Chen, Hong-Yi, Vadim Apalkov, and Tapash Chakraborty A 58, 2160–2167.
(2007a), “Fock-darwin states of dirac electrons in graphene- Doretto, R L, and C. Morais Smith (2007), “Quantum
based artificial atoms,” Phys. Rev. Lett. 98, 186803. hall ferromagnetism in graphene: Su(4) bosonization ap-
Chen, J-H, C. Jang, S. Adam, M. S. Fuhrer, E. D. Williams, proach,” Phys. Rev. B 76, 195431.
and M. Ishigami (2008), “Charged-impurity scattering in Dresselhaus, G, and M. S. Dresselhaus (1965), “Spin-orbit
graphene,” Nature Physics 4, 377–381, arXiv:0708.2408 interaction in graphite,” Phys. Rev. 140, A401–A412.
[[Link]]. Dresselhaus, M S, and G. Dresselhaus (2002), “Intercalation
Chen, Zhihong, Yu-Ming Lin, Michael J. Rooks, and Phae- compounds of graphite,” Advances in Physics 51 (1), 1–
don Avouris (2007b), “Graphene nano-ribbon electronics,” 186.
Physica E: Low-dimensional Systems and Nanostructures Dresselhaus, MS, G. Dresselhaus, J.E. Fischer, and M.J.
40 (2), 228 – 232. Moran (1983), Intercalated graphite (North-Holland, New
Cho, Sungjae, Yung-Fu Chen, and Michael S. Fuhrer (2007), York).
“Gate-tunable graphene spin valve,” Applied Physics Let- Dresselhaus, MS, and J.G. Mavroides (1964), IBM J. Res.
ters 91 (12), 123105. Dev , 262.
 47

Dugaev, V K, V. I. Litvinov, and J. Barnas (2006), “Ex- Japan 65 (7), 1920–1923.

change interaction of magnetic impurities in graphene,” Fukuyama, Hidetoshi (1971), “Theory of orbital magnetism
Phys. Rev. B 74, 224438. of bloch electrons: Coulomb interactions,” Progress of The-
Duke, C B (1968), “Quantum theory of transport in narrow oretical Physics 45 (3), 704–729.
channels,” Phys. Rev. 168, 816–831. Gasparoux, H (1967), “Modification des proprits magntiques
Eisenstein, J P, L. N. Pfeiffer, and K. W. West (1994), du graphite par cration de sequences rhombodriques,” Car-
“Compressibility of the two-dimensional electron gas: Mea- bon 5 (5), 441 – 451.
surements of the zero-field exchange energy and fractional Geim, A K, and K. S. Novoselov (2007), “The rise of
quantum hall gap,” Phys. Rev. B 50, 1760–1778. graphene,” Nature Materials 6, 183.
Eizenberg, M, and J.M. Blakely (1979), “Carbon monolayer Geim, Andrey K, and Allan H. MacDonald (2007),
phase condensation on ni(111),” Surface Science 82 (1), 228 “Graphene: Exploring carbon flatland,” Physics Today
– 236. 60 (8), 35–41.
Esquinazi, P, A. Setzer, R. Höhne, C. Semmelhack, Y. Kopele- Ghosal, Amit, Pallab Goswami, and Sudip Chakravarty
vich, D. Spemann, T. Butz, B. Kohlstrunk, and M. Lösche (2007), “Diamagnetism of nodal fermions,” Phys. Rev. B
(2002), “Ferromagnetism in oriented graphite samples,” 75, 115123.
Phys. Rev. B 66, 024429. Giesbers, A J M, U. Zeitler, M. I. Katsnelson, L. A. Pono-
Esquinazi, P, D. Spemann, R. Höhne, A. Setzer, K.-H. Han, marenko, T. M. Mohiuddin, and J. C. Maan (2007),
and T. Butz (2003), “Induced magnetic ordering by proton “Quantum-hall activation gaps in graphene,” Phys. Rev.
irradiation in graphite,” Phys. Rev. Lett. 91, 227201. Lett. 99, 206803.
Faugeras, C, A. Nerriere, M. Potemski, A. Mahmood, E. Du- Giovani, GF, and G. Vignale (2005), Quantum theory of the
jardin, C. Berger, and W. A. de Heer (2008), “Few-layer electron liquid (Cambridge Press).
graphene on sic, pyrolitic graphite, and graphene: A raman Giovannetti, Gianluca, Petr A. Khomyakov, Geert Brocks,
scattering study,” Applied Physics Letters 92 (1), 011914. Paul J. Kelly, and Jeroen van den Brink (2007),
Ferrari, A C, J. C. Meyer, V. Scardaci, C. Casiraghi, “Substrate-induced band gap in graphene on hexagonal
M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. boron nitride: Ab initio density functional calculations,”
Novoselov, S. Roth, and A. K. Geim (2006), “Raman spec- Phys. Rev. B 76, 073103.
trum of graphene and graphene layers,” Phys. Rev. Lett. Goerbig, M O, J.-N. Fuchs, K. Kechedzhi, and Vladimir I.
97, 187401. Fal’ko (2007), “Filling-factor-dependent magnetophonon
Fertig, H A, and L. Brey (2006), “Luttinger liquid at the resonance in graphene,” Phys. Rev. Lett. 99, 087402.
edge of undoped graphene in a strong magnetic field,” Phys. González, J, F. Guinea, and M. A. H. Vozmediano (1996),
Rev. Lett. 97, 116805. “Unconventional quasiparticle lifetime in graphite,” Phys.
Fogler, M M, D. S. Novikov, L. I. Glazman, and B. I. Rev. Lett. 77, 3589–3592.
Shklovskii (2008), “Effect of disorder on a graphene p-n González, J, F. Guinea, and M. A. H. Vozmedi-
junction,” Phys. Rev. B 77, 075420. ano (1999), “Marginal-fermi-liquid behavior from two-
Fogler, M M, D. S. Novikov, and B. I. Shklovskii (2007), dimensional coulomb interaction,” Phys. Rev. B 59,
“Screening of a hypercritical charge in graphene,” Phys. R2474–R2477.
Rev. B 76, 233402. González, J, F. Guinea, and M. A. H. Vozmediano (2001),
Forbeaux, I, J.-M. Themlin, and J.-M. Debever (1998), “Het- “Electron-electron interactions in graphene sheets,” Phys.
eroepitaxial graphite on 6h − SiC(0001) : interface forma- Rev. B 63, 134421.
tion through conduction-band electronic structure,” Phys. González, José, Francisco Guinea, and M. Angeles H. Voz-
Rev. B 58, 16396–16406. mediano (1992), “Continuum approximation to fullerene
Foster, Matthew S, and Andreas W. W. Ludwig (2006a), “In- molecules,” Phys. Rev. Lett. 69, 172–175.
teraction effects on two-dimensional fermions with random Gorbachev, R V, F. V. Tikhonenko, A. S. Mayorov, D. W.
hopping,” Phys. Rev. B 73, 155104. Horsell, and A. K. Savchenko (2007), “Weak localization
Foster, Matthew S, and Andreas W. W. Ludwig (2006b), in bilayer graphene,” Phys. Rev. Lett. 98, 176805.
“Metal-insulator transition in hubbard-like models with Gorbar, E V, V. P. Gusynin, V. A. Miransky, and I. A.
random hopping,” Phys. Rev. B 74, 241102. Shovkovy (2002), “Magnetic field driven metal-insulator
Fradkin, Eduardo (1986a), “Critical behavior of disordered phase transition in planar systems,” Phys. Rev. B 66,
degenerate semiconductors. i. models, symmetries, and for- 045108.
malism,” Phys. Rev. B 33, 3257–3262. Goswami, Pallab, Xun Jia, and Sudip Chakravarty (2007),
Fradkin, Eduardo (1986b), “Critical behavior of disordered “Quantum hall plateau transition in the lowest landau level
degenerate semiconductors. ii. spectrum and transport of disordered graphene,” Phys. Rev. B 76, 205408.
properties in mean-field theory,” Phys. Rev. B 33, 3263– Graf, D, F. Molitor, K. Ensslin, C. Stampfer, A. Jungen,
3268. C. Hierold, and L. Wirtz (2007), “Spatially resolved ra-
Fritz, Lars, Serge Florens, and Matthias Vojta (2006), “Uni- man spectroscopy of single- and few-layer graphene,” Nano
versal crossovers and critical dynamics of quantum phase Letters 7 (2), 238–242.
transitions: A renormalization group study of the pseudo- Guinea, F (2007), “Charge distribution and screening in lay-
gap kondo problem,” Phys. Rev. B 74, 144410. ered graphene systems,” Phys. Rev. B 75, 235433.
Fuchs, Jean-Noël, and Pascal Lederer (2007), “Spontaneous Guinea, F, A. H. Castro Neto, and N. M. R. Peres (2006),
parity breaking of graphene in the quantum hall regime,” “Electronic states and landau levels in graphene stacks,”
Phys. Rev. Lett. 98, 016803. Phys. Rev. B 73, 245426.
Fujita, Mitsutaka, Katsunori Wakabayashi, Kyoko Nakada, Guinea, F, M. I. Katsnelson, and M. A. H. Vozmediano
and Koichi Kusakabe (1996), “Peculiar localized state at (2008), “Midgap states and charge inhomogeneities in cor-
zigzag graphite edge,” Journal of the Physical Society of rugated graphene,” Phys. Rev. B 77, 075422.
 48

Gumbs, Godfrey, and Paula Fekete (1997), “Hofstadter but- transitions and the density of states of two-dimensional ran-
terfly for the hexagonal lattice,” Phys. Rev. B 56, 3787– dom dirac hamiltonians,” Phys. Rev. B 65, 125323.
3791. Hou, Chang-Yu, Claudio Chamon, and Christopher Mudry
Gupta, A, G. Chen, P. Joshi, S. Tadigadapa, and Eklund (2007), “Electron fractionalization in two-dimensional
(2006), “Raman scattering from high-frequency phonons graphenelike structures,” Phys. Rev. Lett. 98, 186809.
in supported n-graphene layer films,” Nano Letters 6 (12), Huard, B, J. A. Sulpizio, N. Stander, K. Todd, B. Yang, and
2667–2673, pMID: 17163685. D. Goldhaber-Gordon (2007), “Transport measurements
Gusynin, V P, V. A. Miransky, S. G. Sharapov, and I. A. across a tunable potential barrier in graphene,” Phys. Rev.
Shovkovy (2006), “Excitonic gap, phase transition, and Lett. 98, 236803.
quantum hall effect in graphene,” Phys. Rev. B 74, 195429. Huertas-Hernando, Daniel, F. Guinea, and Arne Brataas
Gusynin, V P, and S. G. Sharapov (2005), “Unconventional (2006), “Spin-orbit coupling in curved graphene, fullerenes,
integer quantum hall effect in graphene,” Phys. Rev. Lett. nanotubes, and nanotube caps,” Phys. Rev. B 74, 155426.
95, 146801. Hwang, E H, and S. Das Sarma (2007), “Dielectric function,
Gusynin, V P, S G Sharapov, and J P Carbotte (2007), screening, and plasmons in two-dimensional graphene,”
“Magneto-optical conductivity in graphene,” Journal of Phys. Rev. B 75, 205418.
Physics: Condensed Matter 19 (2), 026222. Hwang, E H, Ben Yu-Kuang Hu, and S. Das Sarma (2007),
Haldane, F D M (1988), “Model for a quantum hall effect “Inelastic carrier lifetime in graphene,” Phys. Rev. B 76,
without landau levels: Condensed-matter realization of the 115434.
”parity anomaly”,” Phys. Rev. Lett. 61, 2015–2018. Ishigami, Masa, J. H. Chen, W. G. Cullen, M. S. Fuhrer, and
Han, Melinda Y, Barbaros Özyilmaz, Yuanbo Zhang, and E. D. Williams (2007), “Atomic structure of graphene on
Philip Kim (2007), “Energy band-gap engineering of sio2,” Nano Letters 7 (6), 1643–1648.
graphene nanoribbons,” Phys. Rev. Lett. 98, 206805. Itzykson, C, and J.-B. Zuber (2006), Quantum Field Theory
Harper, P G (1955), “Single band motion of conduction elec- (Dover).
trons in a uniform magnetic field,” Proceedings of the Phys- Jackiw, R, and C. Rebbi (1976), “Solitons with fermion num-
ical Society. Section A 68 (10), 874. ber 12 ,” Phys. Rev. D 13, 3398–3409.
Harrison, WA (1980), Solid State Theory (Dover). Jiang, Z, E. A. Henriksen, L. C. Tung, Y.-J. Wang, M. E.
Hass, J, R. Feng, J. E. Millán-Otoya, X. Li, M. Sprin- Schwartz, M. Y. Han, P. Kim, and H. L. Stormer (2007a),
kle, P. N. First, W. A. de Heer, E. H. Conrad, and “Infrared spectroscopy of landau levels of graphene,” Phys.
C. Berger (2007), “Structural properties of the multilayer Rev. Lett. 98, 197403.
graphene/4h-SiC(0001) system as determined by surface x- Jiang, Z, Y. Zhang, H. L. Stormer, and P. Kim (2007b),
ray diffraction,” Phys. Rev. B 75, 214109. “Quantum hall states near the charge-neutral dirac point
Hass, J, F. Varchon, J. E. Millán-Otoya, M. Sprinkle, in graphene,” Phys. Rev. Lett. 99, 106802.
N. Sharma, W. A. de Heer, C. Berger, P. N. First, L. Mag- Kane, C L, and E. J. Mele (1997), “Size, shape, and low en-
aud, and E. H. Conrad (2008), “Why multilayer graphene ergy electronic structure of carbon nanotubes,” Phys. Rev.
on 4h-SiC(0001) behaves like a single sheet of graphene,” Lett. 78, 1932–1935.
Phys. Rev. Lett. 100, 125504. Kane, C L, and E. J. Mele (2004), “Electron interactions
de Heer, Walt A, Claire Berger, Xiaosong Wu, Phillip N. First, and scaling relations for optical excitations in carbon nan-
Edward H. Conrad, Xuebin Li, Tianbo Li, Michael Sprin- otubes,” Phys. Rev. Lett. 93, 197402.
kle, Joanna Hass, Marcin L. Sadowski, Marek Potemski, Kane, C L, and E. J. Mele (2005), “Quantum spin hall effect
and Grard Martinez (2007), “Epitaxial graphene,” Solid in graphene,” Phys. Rev. Lett. 95, 226801.
State Communications 143 (1–2), 92–100. Katsnelson, M I (2006a), “Nonlinear screening of charge im-
Hentschel, Martina, and Francisco Guinea (2007), “Orthog- purities in graphene,” Phys. Rev. B 74, 201401.
onality catastrophe and kondo effect in graphene,” Phys. Katsnelson, M I (2006b), “Zitterbewegung, chirality, and min-
Rev. B 76, 115407. imal conductivity in graphene,” The European Physical
Herbut, Igor F (2006), “Interactions and phase transitions Journal B - Condensed Matter and Complex Systems 51,
on graphene’s honeycomb lattice,” Phys. Rev. Lett. 97, 157–160, 10.1140/epjb/e2006-00203-1.
146401. Katsnelson, M I (2007a), “Conductance quantization in
Herbut, Igor F (2007), “Theory of integer quantum hall effect graphene nanoribbons: adiabatic approximation,” The Eu-
in graphene,” Phys. Rev. B 75, 165411. ropean Physical Journal B - Condensed Matter and Com-
Heremans, J, C. H. Olk, and D. T. Morelli (1994), “Mag- plex Systems 57, 225–228, 10.1140/epjb/e2007-00168-5.
netic susceptibility of carbon structures,” Phys. Rev. B 49, Katsnelson, M I (2007b), “Scattering of charge carriers by
15122–15125. point defects in bilayer graphene,” Phys. Rev. B 76,
Himpsel, FJ, K. Christmann, P. Heimann, D.E. Eastman, 073411.
and Peter J. Feibelman (1982), “Adsorbate band disper- Katsnelson, MI, and A.K Geim (2008), “Electron scatter-
sions for c on ru(0001),” Surface Science 115 (3), L159 – ing on microscopic corrugations in graphene,” Philosoph-
L164. ical Transactions of the Royal Society A: Mathematical,
Hobson, J P, and W. A. Nierenberg (1953), “The statistics Physical and Engineering Sciences 366 (1863), 195–204.
of a two-dimensional, hexagonal net,” Phys. Rev. 89, 662– Katsnelson, Mikhail I (2007c), “Graphene: carbon in two di-
662. mensions,” Materials Today 10, 20 – 27.
Hod, Oded, Verónica Barone, Juan E. Peralta, and Gus- Kechedzhi, K, Vladimir I. Fal’ko, E. McCann, and B. L. Alt-
tavo E. Scuseria (2007), “Enhanced half-metallicity in edge- shuler (2007), “Influence of trigonal warping on interference
oxidized zigzag graphene nanoribbons,” Nano Letters 7 (8), effects in bilayer graphene,” Phys. Rev. Lett. 98, 176806.
2295–2299. Khveshchenko, D V (2001), “Ghost excitonic insulator tran-
Horovitz, Baruch, and Pierre Le Doussal (2002), “Freezing sition in layered graphite,” Phys. Rev. Lett. 87, 246802.
 49

Kim, Eun-Ah, and A. H. Castro Neto (2008), “Graphene sorption of h2 O, Nh3 , co, No2 , and no on graphene: A
as an electronic membrane,” EPL (Europhysics Letters) first-principles study,” Phys. Rev. B 77, 125416.
84 (5), 57007. Lewenkopf, C H, E. R. Mucciolo, and A. H. Castro Neto
Kobayashi, Yousuke, Ken-ichi Fukui, Toshiaki Enoki, Koichi (2008), “Numerical studies of conductivity and fano factor
Kusakabe, and Yutaka Kaburagi (2005), “Observation of in disordered graphene,” Phys. Rev. B 77, 081410.
zigzag and armchair edges of graphite using scanning tun- Li, Z Q, S.-W. Tsai, W. J. Padilla, S. V. Dordevic, K. S.
neling microscopy and spectroscopy,” Phys. Rev. B 71, Burch, Y. J. Wang, and D. N. Basov (2006), “Infrared
193406. probe of the anomalous magnetotransport of highly ori-
Kolesnikov, D, and V. Osipov (2004), “Electronic structure ented pyrolytic graphite in the extreme quantum limit,”
of carbon nanohorns near the fermi level,” JETP Letters Phys. Rev. B 74, 195404.
79, 532–536, 10.1134/1.1787100. Louis, E, J. A. Vergés, F. Guinea, and G. Chiappe (2007),
Kolesnikov, D V, and V. A. Osipov (2006), “The continuum “Transport regimes in surface disordered graphene sheets,”
gauge field-theory model for low-energy electronic states of Phys. Rev. B 75, 085440.
icosahedral fullerenes,” The European Physical Journal B Ludwig, Andreas W W, Matthew P. A. Fisher, R. Shankar,
- Condensed Matter and Complex Systems 49, 465–470, and G. Grinstein (1994), “Integer quantum hall transition:
10.1140/epjb/e2006-00087-y. An alternative approach and exact results,” Phys. Rev. B
Kolezhuk, Alexei, Subir Sachdev, Rudro R. Biswas, and 50, 7526–7552.
Peiqiu Chen (2006), “Theory of quantum impurities in spin Lukose, Vinu, R. Shankar, and G. Baskaran (2007), “Novel
liquids,” Phys. Rev. B 74, 165114. electric field effects on landau levels in graphene,” Phys.
Kopelevich, Y, and P. Esquinazi (2007), “Ferromag- Rev. Lett. 98, 116802.
netism and superconductivity in carbon-based systems,” Luk’yanchuk, Igor A, and Yakov Kopelevich (2004), “Phase
Journal of Low Temperature Physics 146, 629–639, analysis of quantum oscillations in graphite,” Phys. Rev.
10.1007/s10909-006-9286-5. Lett. 93, 166402.
Kopelevich, Y, J.C. Medina Pantoja, R.R. da Silva, Mañes, J L, F. Guinea, and Marı́a A. H. Vozmediano (2007),
F. Mrowka, and P. Esquinazi (2006), “Anomalous hall ef- “Existence and topological stability of fermi points in mul-
fect in graphite,” Physics Letters A 355 (3), 233 – 236. tilayered graphene,” Phys. Rev. B 75, 155424.
Kopelevich, Y, J. H. S. Torres, R. R. da Silva, F. Mrowka, Maiti, Moitri, and K. Sengupta (2007), “Josephson effect
H. Kempa, and P. Esquinazi (2003), “Reentrant metallic in graphene superconductor/barrier/superconductor junc-
behavior of graphite in the quantum limit,” Phys. Rev. tions: Oscillatory behavior of the josephson current,” Phys.
Lett. 90, 156402. Rev. B 76, 054513.
Koshino, Mikito, and Tsuneya Ando (2006), “Transport in Malard, L M, J. Nilsson, D. C. Elias, J. C. Brant, F. Plentz,
bilayer graphene: Calculations within a self-consistent born E. S. Alves, A. H. Castro Neto, and M. A. Pimenta (2007),
approximation,” Phys. Rev. B 73, 245403. “Probing the electronic structure of bilayer graphene by
Koshino, Mikito, and Tsuneya Ando (2007), “Diamagnetism raman scattering,” Phys. Rev. B 76, 201401.
in disordered graphene,” Phys. Rev. B 75, 235333. Mallet, P, F. Varchon, C. Naud, L. Magaud, C. Berger, and
Kumazaki, Hideki, and Dai S. Hirashima (2006), “Effects J.-Y. Veuillen (2007), “Electron states of mono- and bilayer
of impurities in two-dimensional graphite,” Journal of the graphene on sic probed by scanning-tunneling microscopy,”
Physical Society of Japan 75 (5), 053707. Phys. Rev. B 76, 041403.
Kumazaki, Hideki, and Dai S. Hirashima (2007), Maple, MB (1998), “High-temperature superconductivity,”
“Nonmagnetic-defect-induced magnetism in graphene,” Journal of Magnetism and Magnetic Materials 177181,
Journal of the Physical Society of Japan 76 (6), 064713. Part 1 (0), 18 – 30.
Kusminskiy, S Viola, Johan Nilsson, D. K. Campbell, and Mariani, Eros, Leonid I. Glazman, Alex Kamenev, and Felix
A. H. Castro Neto (2008), “Electronic compressibility of a von Oppen (2007), “Zero-bias anomaly in the tunneling
graphene bilayer,” Phys. Rev. Lett. 100, 106805. density of states of graphene,” Phys. Rev. B 76, 165402.
Lammert, Paul E, and Vincent H. Crespi (2004), “Graphene Mariani, Eros, and Felix von Oppen (2008), “Erratum: Flex-
cones: classification by fictitious flux and electronic prop- ural phonons in free-standing graphene [phys. rev. lett. 100,
erties,” Phys. Rev. B 69, 035406. 076801 (2008)],” Phys. Rev. Lett. 100, 249901.
Laughlin, R B (1981), “Quantized hall conductivity in two Martelo, LM, M. Dzierzawa, L. Siffert, and D. Baeriswyl
dimensions,” Phys. Rev. B 23, 5632–5633. (1997), “Mott-hubbard transition and antiferromagnetism
Lazzeri, Michele, and Francesco Mauri (2006), “Nonadiabatic on the honeycomb lattice,” Zeitschrift fr Physik B Con-
kohn anomaly in a doped graphene monolayer,” Phys. Rev. densed Matter 103, 335–338, 10.1007/s002570050384.
Lett. 97, 266407. Martin, I, and Y. M. Blanter (2009), “Transport in disordered
LeClair, André (2000), “Relevance of disorder for dirac graphene nanoribbons,” Phys. Rev. B 79 (23), 235132,
fermions with imaginary vector potentials,” Phys. Rev. arXiv:0705.0532 [[Link]-hall].
Lett. 84, 1292–1295. Martins, T B, R. H. Miwa, Antônio J. R. da Silva, and
Lee, Patrick A (1993), “Localized states in a d -wave super- A. Fazzio (2007), “Electronic and transport properties of
conductor,” Phys. Rev. Lett. 71, 1887–1890. boron-doped graphene nanoribbons,” Phys. Rev. Lett. 98,
Lee, Patrick A, and T. V. Ramakrishnan (1985), “Disordered 196803.
electronic systems,” Rev. Mod. Phys. 57, 287–337. Matsui, T, H. Kambara, Y. Niimi, K. Tagami, M. Tsukada,
Lee, Sung-Sik, and Patrick A. Lee (2005), “U(1) gauge theory and Hiroshi Fukuyama (2005), “Sts observations of landau
of the hubbard model: Spin liquid states and possible ap- levels at graphite surfaces,” Phys. Rev. Lett. 94, 226403.
plication to κ-(BEDT-TTF)2 cu2 (CN)3 ,” Phys. Rev. Lett. Mattausch, Alexander, and Oleg Pankratov (2007), “Ab Ini-
95, 036403. tio study of graphene on sic,” Phys. Rev. Lett. 99, 076802.
Leenaerts, O, B. Partoens, and F. M. Peeters (2008), “Ad- McCann, E, K. Kechedzhi, Vladimir I. Fal’ko, H. Suzuura,
 50

T. Ando, and B. L. Altshuler (2006), “Weak-localization Nilsson, Johan, and A. H. Castro Neto (2007), “Impurities
magnetoresistance and valley symmetry in graphene,” in a biased graphene bilayer,” Phys. Rev. Lett. 98, 126801.
Phys. Rev. Lett. 97, 146805. Nilsson, Johan, A. H. Castro Neto, F. Guinea, and N. M. R.
McCann, Edward (2006), “Asymmetry gap in the electronic Peres (2007), “Transmission through a biased graphene bi-
band structure of bilayer graphene,” Phys. Rev. B 74, layer barrier,” Phys. Rev. B 76, 165416.
161403. Nilsson, Johan, A. H. Castro Neto, F. Guinea, and N. M. R.
McCann, Edward, and Vladimir I. Fal’ko (2006), “Landau- Peres (2008), “Electronic properties of bilayer and multi-
level degeneracy and quantum hall effect in a graphite bi- layer graphene,” Phys. Rev. B 78, 045405.
layer,” Phys. Rev. Lett. 96, 086805. Nilsson, Johan, A. H. Castro Neto, N. M. R. Peres, and
McClure, J W (1956), “Diamagnetism of graphite,” Phys. F. Guinea (2006a), “Electron-electron interactions and the
Rev. 104, 666–671. phase diagram of a graphene bilayer,” Phys. Rev. B 73,
McClure, J W (1957), “Band structure of graphite and de 214418.
haas-van alphen effect,” Phys. Rev. 108, 612–618. Nilsson, Johan, A. H. Castro Neto, F. Guinea, and N. M. R.
McClure, J W (1958), “Analysis of multicarrier galvanomag- Peres (2006b), “Electronic properties of graphene multilay-
netic data for graphite,” Phys. Rev. 112, 715–721. ers,” Phys. Rev. Lett. 97, 266801.
McClure, J W (1960), “Theory of diamagnetism of graphite,” Nomura, Kentaro, Mikito Koshino, and Shinsei Ryu (2007),
Phys. Rev. 119, 606–613. “Topological delocalization of two-dimensional massless
McEuen, Paul L, Marc Bockrath, David H. Cobden, Young- dirac fermions,” Phys. Rev. Lett. 99, 146806.
Gui Yoon, and Steven G. Louie (1999), “Disorder, pseu- Nomura, Kentaro, and A. H. MacDonald (2007), “Quantum
dospins, and backscattering in carbon nanotubes,” Phys. transport of massless dirac fermions,” Phys. Rev. Lett. 98,
Rev. Lett. 83, 5098–5101. 076602.
Meyer, JC, A.K. Geim, M.I. Katsnelson, K.S. Novoselov, Nomura, Kentaro, and Allan H. MacDonald (2006), “Quan-
D. Obergfell, S. Roth, C. Girit, and A. Zettl (2007), “On tum hall ferromagnetism in graphene,” Phys. Rev. Lett.
the roughness of single- and bi-layer graphene membranes,” 96, 256602.
Solid State Communications 143, 101 – 109. Novikov, D S (2007a), “Elastic scattering theory and trans-
Miao, F, S. Wijeratne, Y. Zhang, U. C. Coskun, W. Bao, and port in graphene,” Phys. Rev. B 76, 245435.
C. N. Lau (2007), “Phase-coherent transport in graphene Novikov, D S (2007b), “Numbers of donors and acceptors from
quantum billiards,” Science 317 (5844), 1530–1533. transport measurements in graphene,” Applied Physics
Min, Hongki, J. E. Hill, N. A. Sinitsyn, B. R. Sahu, Leonard Letters 91 (10), 102102.
Kleinman, and A. H. MacDonald (2006), “Intrinsic and Novikov, D S (2007c), “Transverse field effect in graphene
rashba spin-orbit interactions in graphene sheets,” Phys. ribbons,” Phys. Rev. Lett. 99, 056802.
Rev. B 74, 165310. Novoselov, K S, A. K. Geim, S. V. Morozov, D. Jiang,
Mishchenko, E G (2007), “Effect of electron-electron interac- Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A.
tions on the conductivity of clean graphene,” Phys. Rev. Firsov (2004), “Electric field effect in atomically thin car-
Lett. 98, 216801. bon films,” Science 306 (5696), 666–669.
Morozov, S V, K. S. Novoselov, M. I. Katsnelson, F. Schedin, Novoselov, K S, D. Jiang, F. Schedin, T. J. Booth, V. V.
L. A. Ponomarenko, D. Jiang, and A. K. Geim (2006), Khotkevich, S. V. Morozov, and A. K. Geim (2005),
“Strong suppression of weak localization in graphene,” “Two-dimensional atomic crystals,” Proceedings of the Na-
Phys. Rev. Lett. 97, 016801. tional Academy of Sciences of the United States of America
Morozov, S V, K. S. Novoselov, F. Schedin, D. Jiang, A. A. 102 (30), 10451–10453.
Firsov, and A. K. Geim (2005), “Two-dimensional electron Novoselov, K S, Z. Jiang, Y. Zhang, S. V. Morozov, H. L.
and hole gases at the surface of graphite,” Phys. Rev. B 72, Stormer, U. Zeitler, J. C. Maan, G. S. Boebinger, P. Kim,
201401. and A. K. Geim (2007), “Room-temperature quantum hall
Morpurgo, A F, and F. Guinea (2006), “Intervalley scatter- effect in graphene,” Science 315 (5817), 1379.
ing, long-range disorder, and effective time-reversal sym- Nozières, Philippe (1958), “Cyclotron resonance in graphite,”
metry breaking in graphene,” Phys. Rev. Lett. 97, 196804. Phys. Rev. 109, 1510–1521.
Mrozowski, S (1952), “Semiconductivity and diamagnetism Ohta, Kuniichi (1971a), “Broadening of landau levels in two-
of polycrystalline graphite and condensed ring systems,” dimensional electron gas. ii. transverse magnetoconduc-
Phys. Rev. 85, 609–620. tance,” Journal of the Physical Society of Japan 31 (6),
Nakada, Kyoko, Mitsutaka Fujita, Gene Dresselhaus, and 1627–1638.
Mildred S. Dresselhaus (1996), “Edge state in graphene rib- Ohta, Kuniichi (1971b), “Broadening of landau levels in two-
bons: Nanometer size effect and edge shape dependence,” dimensional electron gas: Its effect on surface capacitance,”
Phys. Rev. B 54, 17954–17961. Japanese Journal of Applied Physics 10 (7), 850–863.
Nakamura, Masaaki (2007), “Orbital magnetism and trans- Ohta, Taisuke, Aaron Bostwick, J. L. McChesney, Thomas
port phenomena in two-dimensional dirac fermions in a Seyller, Karsten Horn, and Eli Rotenberg (2007), “In-
weak magnetic field,” Phys. Rev. B 76, 113301. terlayer interaction and electronic screening in multilayer
Nelson, D.R.,, and Peliti, L. (1987), “Fluctuations in mem- graphene investigated with angle-resolved photoemission
branes with crystalline and hexatic order,” J. Phys. France spectroscopy,” Phys. Rev. Lett. 98, 206802.
48 (7), 1085–1092. Ohta, Taisuke, Aaron Bostwick, Thomas Seyller, Karsten
Niimi, Y, H. Kambara, T. Matsui, D. Yoshioka, and Hiroshi Horn, and Eli Rotenberg (2006), “Controlling the elec-
Fukuyama (2006), “Real-space imaging of alternate local- tronic structure of bilayer graphene,” Science 313 (5789),
ization and extension of quasi-two-dimensional electronic 951–954.
states at graphite surfaces in magnetic fields,” Phys. Rev. Oshima, Chuhei, and Ayato Nagashima (1997), “Ultra-thin
Lett. 97, 236804. epitaxial films of graphite and hexagonal boron nitride
 51

on solid surfaces,” Journal of Physics: Condensed Matter Peres, N M R, F. Guinea, and A. H. Castro Neto (2006c),
9 (1), 1. “Electronic properties of disordered two-dimensional car-
Osipov, V, E. Kochetov, and M. Pudlak (2003), “Elec- bon,” Phys. Rev. B 73, 125411.
tronic structure of carbon nanoparticles,” Journal of Peres, N M R, F. D. Klironomos, S.-W. Tsai, J. R. Santos,
Experimental and Theoretical Physics 96, 140–149, J. M. B. Lopes dos Santos, and A. H. Castro Neto (2007a),
10.1134/1.1545393. “Electron waves in chemically substituted graphene,” EPL
Ossipov, A, M. Titov, and C. W. J. Beenakker (2007), “Reen- (Europhysics Letters) 80 (6), 67007.
trance effect in a graphene n-p-n junction coupled to a su- Peres, N M R, J. M. B. Lopes dos Santos, and T. Stauber
perconductor,” Phys. Rev. B 75, 241401. (2007b), “Phenomenological study of the electronic trans-
Ostrovsky, P M, I. V. Gornyi, and A. D. Mirlin (2006), “Elec- port coefficients of graphene,” Phys. Rev. B 76, 073412.
tron transport in disordered graphene,” Phys. Rev. B 74, Peres, NMR, F. Guinea, and A.H. Castro Neto (2006d),
235443. “Electronic properties of two-dimensional carbon,” Annals
Ostrovsky, P M, I. V. Gornyi, and A. D. Mirlin (2007), of Physics 321 (7), 1559 – 1567.
“Quantum criticality and minimal conductivity in graphene Phillips, Philip (2006), “Mottness,” Annals of Physics
with long-range disorder,” Phys. Rev. Lett. 98, 256801. 321 (7), 1634 – 1650.
Özyilmaz, Barbaros, Pablo Jarillo-Herrero, Dmitri Efetov, Polini, Marco, Reza Asgari, Yafis Barlas, T. Pereg-Barnea,
Dmitry A. Abanin, Leonid S. Levitov, and Philip Kim and A.H. MacDonald (2007), “Graphene: A pseudochiral
(2007), “Electronic transport and quantum hall effect in fermi liquid,” Solid State Communications 143 (1–2), 58–
bipolar graphene p-n-p junctions,” Phys. Rev. Lett. 99, 62.
166804. Polkovnikov, Anatoli (2002), “Kondo effect in d-wave super-
Paiva, Thereza, R. T. Scalettar, W. Zheng, R. R. P. conductors,” Phys. Rev. B 65, 064503.
Singh, and J. Oitmaa (2005), “Ground-state and finite- Polkovnikov, Anatoli, Subir Sachdev, and Matthias Vojta
temperature signatures of quantum phase transitions in the (2001), “Impurity in a d -wave superconductor: Kondo ef-
half-filled hubbard model on a honeycomb lattice,” Phys. fect and stm spectra,” Phys. Rev. Lett. 86, 296–299.
Rev. B 72, 085123. Rammal, R., (1985), “Landau level spectrum of bloch elec-
Vázquez de Parga, A L, F. Calleja, B. Borca, M. C. G. trons in a honeycomb lattice,” J. Phys. France 46 (8), 1345–
Passeggi, J. J. Hinarejos, F. Guinea, and R. Miranda 1354.
(2008), “Periodically rippled graphene: Growth and spa- Recher, P, B. Trauzettel, A. Rycerz, Ya. M. Blanter, C. W. J.
tially resolved electronic structure,” Phys. Rev. Lett. 100, Beenakker, and A. F. Morpurgo (2007), “Aharonov-bohm
056807. effect and broken valley degeneracy in graphene rings,”
Parr, Robert G, David P. Craig, and Ian G. Ross (1950), Phys. Rev. B 76, 235404.
“Molecular orbital calculations of the lower excited elec- Reich, S, J. Maultzsch, C. Thomsen, and P. Ordejón (2002),
tronic levels of benzene, configuration interaction in- “Tight-binding description of graphene,” Phys. Rev. B 66,
cluded,” The Journal of Chemical Physics 18 (12), 1561– 035412.
1563. Robinson, John P, and Henning Schomerus (2007), “Elec-
Partoens, B, and F. M. Peeters (2006), “From graphene to tronic transport in normal-conductor/graphene/normal-
graphite: Electronic structure around the k point,” Phys. conductor junctions and conditions for insulating behavior
Rev. B 74, 075404. at a finite charge-carrier density,” Phys. Rev. B 76, 115430.
Peliti, L, and S. Leibler (1985), “Effects of thermal fluctu- Muñoz Rojas, F, D. Jacob, J. Fernández-Rossier, and J. J.
ations on systems with small surface tension,” Phys. Rev. Palacios (2006), “Coherent transport in graphene nanocon-
Lett. 54, 1690–1693. strictions,” Phys. Rev. B 74, 195417.
Pereira, Vitor M, F. Guinea, J. M. B. Lopes dos Santos, Rollings, E, G.-H. Gweon, S.Y. Zhou, B.S. Mun, J.L. Mc-
N. M. R. Peres, and A. H. Castro Neto (2006), “Disor- Chesney, B.S. Hussain, A.V. Fedorov, P.N. First, W.A.
der induced localized states in graphene,” Phys. Rev. Lett. de Heer, and A. Lanzara (2006), “Synthesis and charac-
96, 036801. terization of atomically thin graphite films on a silicon car-
Pereira, Vitor M, Johan Nilsson, and A. H. Castro Neto bide substrate,” Journal of Physics and Chemistry of Solids
(2007), “Coulomb impurity problem in graphene,” Phys. 67 (9–10), 2172–2177.
Rev. Lett. 99, 166802. Rong, Zhao Y, and Pieter Kuiper (1993), “Electronic ef-
Pereira, Vitor M, J. M. B. Lopes dos Santos, and A. H. Cas- fects in scanning tunneling microscopy: Moiré pattern on
tro Neto (2008), “Modeling disorder in graphene,” Phys. a graphite surface,” Phys. Rev. B 48, 17427–17431.
Rev. B 77, 115109. Rosenstein, Baruch, Brian J. Warr, and Seon H. Park
Peres, N M R, M. A. N. Araújo, and Daniel Bozi (2004), (1989), “Four-fermion theory is renormalizable in 2+1 di-
“Phase diagram and magnetic collective excitations of the mensions,” Phys. Rev. Lett. 62, 1433–1436.
hubbard model for graphene sheets and layers,” Phys. Rev. Rosenstein, Baruch, Brian J. Warr, and Seon H. Park (1991),
B 70, 195122. “Dynamical symmetry breaking in four-fermion interaction
Peres, N M R, A. H. Castro Neto, and F. Guinea (2006a), models,” Physics Reports 205 (2), 59 – 108.
“Conductance quantization in mesoscopic graphene,” Phys. Russo, S, J. B. Oostinga, D. Wehenkel, H. B. Heersche, S. S.
Rev. B 73, 195411. Sobhani, L. M. K. Vandersypen, and A. F. Morpurgo
Peres, N M R, A. H. Castro Neto, and F. Guinea (2006b), (2008), “Observation of Aharonov-Bohm conductance os-
“Dirac fermion confinement in graphene,” Phys. Rev. B 73, cillations in a graphene ring,” Phys. Rev. B 77 (8), 085413,
241403. arXiv:0711.1508 [[Link]-hall].
Peres, N M R, F. Guinea, and A. H. Castro Neto (2005), Rydberg, H, M. Dion, N. Jacobson, E. Schröder,
“Coulomb interactions and ferromagnetism in pure and P. Hyldgaard, S. I. Simak, D. C. Langreth, and B. I.
doped graphene,” Phys. Rev. B 72, 174406. Lundqvist (2003), “Van der waals density functional for
 52

layered structures,” Phys. Rev. Lett. 91, 126402. B 34, 979–993.

Ryu, Shinsei, Christopher Mudry, Hideaki Obuse, and Akira Shung, Kenneth W K (1986b), “Lifetime effects in low-stage
Furusaki (2007), “z2 topological term, the global anomaly, intercalated graphite systems,” Phys. Rev. B 34, 1264–
and the two-dimensional symplectic symmetry class of an- 1272.
derson localization,” Phys. Rev. Lett. 99, 116601. Shytov, A V, M. I. Katsnelson, and L. S. Levitov (2007),
Sadowski, M L, G. Martinez, M. Potemski, C. Berger, and “Vacuum polarization and screening of supercritical impu-
W. A. de Heer (2006), “Landau level spectroscopy of ultra- rities in graphene,” Phys. Rev. Lett. 99, 236801.
thin graphite layers,” Phys. Rev. Lett. 97, 266405. Silvestrov, P G, and K. B. Efetov (2007), “Quantum dots in
Safran, S A (1984), “Stage dependence of magnetic suscepti- graphene,” Phys. Rev. Lett. 98, 016802.
bility of intercalated graphite,” Phys. Rev. B 30, 421–423. Sinitsyna, O V, and Igor V Yaminsky (2006), “Atomic reso-
Safran, S A, and F. J. DiSalvo (1979), “Theory of magnetic lution probe microscopy of the graphite surface,” Russian
susceptibility of graphite intercalation compounds,” Phys. Chemical Reviews 75 (1), 23.
Rev. B 20, 4889–4895. Skrypnyk, Yu V, and V. M. Loktev (2006), “Impurity ef-
Saha, S K, U. V. Waghmare, H. R. Krishnamurthy, and A. K. fects in a two-dimensional system with the dirac spectrum,”
Sood (2007), “Frequency shift of optical phonons in doped Phys. Rev. B 73, 241402.
graphene layers,” arXiv:cond-mat/0702627. Skrypnyk, Yuriy V, and Vadim M. Loktev (2007), “Local
Saito, R, M. Fujita, G. Dresselhaus, and M. S Dresselhaus spectrum rearrangement in impure graphene,” Phys. Rev.
(1992a), “Electronic structure of chiral graphene tubules,” B 75, 245401.
Applied Physics Letters 60 (18), 2204–2206. Slonczewski, J C, and P. R. Weiss (1958), “Band structure
Saito, Riichiro, Mitsutaka Fujita, G. Dresselhaus, and M. S. of graphite,” Phys. Rev. 109, 272–279.
Dresselhaus (1992b), “Electronic structure of graphene Snyman, I, and C. W. J. Beenakker (2007), “Ballistic trans-
tubules based on c60 ,” Phys. Rev. B 46, 1804–1811. mission through a graphene bilayer,” Phys. Rev. B 75,
San-Jose, Pablo, Elsa Prada, and Dmitry S. Golubev (2007), 045322.
“Universal scaling of current fluctuations in disordered Sols, F, F. Guinea, and A. H. Castro Neto (2007), “Coulomb
graphene,” Phys. Rev. B 76, 195445. blockade in graphene nanoribbons,” Phys. Rev. Lett. 99,
Lopes dos Santos, J M B, N. M. R. Peres, and A. H. Cas- 166803.
tro Neto (2007), “Graphene bilayer with a twist: Electronic Son, D T (2007), “Quantum critical point in graphene ap-
structure,” Phys. Rev. Lett. 99, 256802. proached in the limit of infinitely strong coulomb interac-
Saremi, Saeed (2007), “Rkky in half-filled bipartite lattices: tion,” Phys. Rev. B 75, 235423.
Graphene as an example,” Phys. Rev. B 76, 184430. Son, Young-Woo, Marvin L. Cohen, and Steven G. Louie
Schakel, Adriaan M J (1991), “Relativistic quantum hall ef- (2006), “Energy gaps in graphene nanoribbons,” Phys. Rev.
fect,” Phys. Rev. D 43, 1428–1431. Lett. 97, 216803.
Schomerus, Henning (2007), “Effective contact model for Sorella, S, and E. Tosatti (1992), “Semi-metal-insulator tran-
transport through weakly-doped graphene,” Phys. Rev. B sition of the hubbard model in the honeycomb lattice,” EPL
76, 045433. (Europhysics Letters) 19 (8), 699.
Schroeder, P R, M. S. Dresselhaus, and A. Javan (1968), Soule, D E, J. W. McClure, and L. B. Smith (1964), “Study
“Location of electron and hole carriers in graphite from of the shubnikov-de haas effect. determination of the fermi
laser magnetoreflection data,” Phys. Rev. Lett. 20, 1292– surfaces in graphite,” Phys. Rev. 134, A453–A470.
1295. Spataru, Catalin D, Miguel A. Cazalilla, Angel Ru-
Semenoff, Gordon W (1984), “Condensed-matter simulation bio, Lorin X. Benedict, Pedro M. Echenique, and
of a three-dimensional anomaly,” Phys. Rev. Lett. 53, Steven G. Louie (2001), “Anomalous quasiparticle lifetime
2449–2452. in graphite: Band structure effects,” Phys. Rev. Lett. 87,
Sengupta, K, and G. Baskaran (2008), “Tuning kondo physics 246405.
in graphene with gate voltage,” Phys. Rev. B 77, 045417. Spry, W J, and P. M. Scherer (1960), “de haas-van alphen
Seoánez, C, F. Guinea, and A. H. Castro Neto (2007), “Dis- effect in graphite between 3 and 85 kilogauss,” Phys. Rev.
sipation in graphene and nanotube resonators,” Phys. Rev. 120, 826–829.
B 76, 125427. Stauber, T, F. Guinea, and M. A. H. Vozmediano
Shankar, R (1994), “Renormalization-group approach to in- (2005), “Disorder and interaction effects in two-dimensional
teracting fermions,” Rev. Mod. Phys. 66, 129–192. graphene sheets,” Phys. Rev. B 71, 041406.
Sharma, M P, L. G. Johnson, and J. W. McClure (1974), Stauber, T, N. M. R. Peres, F. Guinea, and A. H. Castro Neto
“Diamagnetism of graphite,” Phys. Rev. B 9, 2467–2475. (2007), “Fermi liquid theory of a fermi ring,” Phys. Rev. B
Shelton, JC, H.R. Patil, and J.M. Blakely (1974), “Equi- 75, 115425.
librium segregation of carbon to a nickel (111) surface: A Stephan, O, P. M. Ajayan, C. Colliex, Ph. Redlich, J. M. Lam-
surface phase transition,” Surface Science 43 (2), 493 – 520. bert, P. Bernier, and P. Lefin (1994), “Doping graphitic
Sheng, D N, L. Sheng, and Z. Y. Weng (2006), “Quantum and carbon nanotube structures with boron and nitrogen,”
hall effect in graphene: Disorder effect and phase diagram,” Science 266 (5191), 1683–1685.
Phys. Rev. B 73, 233406. Stolyarova, Elena, Kwang Taeg Rim, Sunmin Ryu, Janina
Shklovskii, B I (2007), “Simple model of coulomb disorder Maultzsch, Philip Kim, Louis E. Brus, Tony F. Heinz,
and screening in graphene,” Phys. Rev. B 76, 233411. Mark S. Hybertsen, and George W. Flynn (2007), “High-
Shon, Nguyen Hong, and Tsuneya Ando (1998), “Quantum resolution scanning tunneling microscopy imaging of meso-
transport in two-dimensional graphite system,” Journal of scopic graphene sheets on an insulating surface,” Proceed-
the Physical Society of Japan 67 (7), 2421–2429. ings of the National Academy of Sciences 104 (22), 9209–
Shung, Kenneth W K (1986a), “Dielectric function and plas- 9212.
mon structure of stage-1 intercalated graphite,” Phys. Rev. Su, W P, J. R. Schrieffer, and A. J. Heeger (1979), “Solitons
 53

in polyacetylene,” Phys. Rev. Lett. 42, 1698–1701. Vozmediano, M A H, M. P. López-Sancho, T. Stauber, and
Su, W P, J. R. Schrieffer, and A. J. Heeger (1980), “Soliton F. Guinea (2005), “Local defects and ferromagnetism in
excitations in polyacetylene,” Phys. Rev. B 22, 2099–2111. graphene layers,” Phys. Rev. B 72, 155121.
Sugawara, K, T. Sato, S. Souma, T. Takahashi, and Wakabayashi, Katsunori, Mitsutaka Fujita, Hiroshi Ajiki,
H. Suematsu (2007), “Anomalous quasiparticle lifetime and and Manfred Sigrist (1999), “Electronic and magnetic prop-
strong electron-phonon coupling in graphite,” Phys. Rev. erties of nanographite ribbons,” Phys. Rev. B 59, 8271–
Lett. 98, 036801. 8282.
Suzuura, Hidekatsu, and Tsuneya Ando (2002a), “Crossover Wakabayashi, Katsunori, and Manfred Sigrist (2000), “Zero-
from symplectic to orthogonal class in a two-dimensional conductance resonances due to flux states in nanographite
honeycomb lattice,” Phys. Rev. Lett. 89, 266603. ribbon junctions,” Phys. Rev. Lett. 84, 3390–3393.
Suzuura, Hidekatsu, and Tsuneya Ando (2002b), “Phonons Wallace, P R (1947), “The band theory of graphite,” Phys.
and electron-phonon scattering in carbon nanotubes,” Rev. 71, 622–634.
Phys. Rev. B 65, 235412. Wang, Hao, D. N. Sheng, L. Sheng, and F. D. M. Hal-
Swain, Peter S, and David Andelman (1999), “The influence dane (2008), “Broken-symmetry states of dirac fermions
of substrate structure on membrane adhesion,” Langmuir in graphene with a partially filled high landau level,” Phys.
15 (26), 8902–8914. Rev. Lett. 100, 116802.
Tersoff, J (1992), “Energies of fullerenes,” Phys. Rev. B 46, Wang, Xue-Feng, and Tapash Chakraborty (2007a), “Collec-
15546–15549. tive excitations of dirac electrons in a graphene layer with
Tikhonenko, F V, D. W. Horsell, R. V. Gorbachev, and A. K. spin-orbit interactions,” Phys. Rev. B 75, 033408.
Savchenko (2008), “Weak localization in graphene flakes,” Wang, Xue-Feng, and Tapash Chakraborty (2007b),
Phys. Rev. Lett. 100, 056802. “Coulomb screening and collective excitations in a
Titov, M (2007), “Impurity-assisted tunneling in graphene,” graphene bilayer,” Phys. Rev. B 75, 041404.
EPL (Europhysics Letters) 79 (1), 17004. Wehling, T O, K. S. Novoselov, S. V. Morozov, E. E. Vdovin,
Titov, M, and C. W. J. Beenakker (2006), “Josephson effect M. I. Katsnelson, A. K. Geim, and A. I. Lichtenstein
in ballistic graphene,” Phys. Rev. B 74, 041401. (2008), “Molecular doping of graphene,” Nano Letters
Tománek, David, Steven G. Louie, H. Jonathon Mamin, 8 (1), 173–177, pMID: 18085811.
David W. Abraham, Ruth Ellen Thomson, Eric Ganz, Williams, J R, L. DiCarlo, and C. M. Marcus (2007),
and John Clarke (1987), “Theory and observation of “Quantum hall effect in a gate-controlled p-n junction of
highly asymmetric atomic structure in scanning-tunneling- graphene,” Science 317 (5838), 638–641.
microscopy images of graphite,” Phys. Rev. B 35, 7790– Williamson, S J, S. Foner, and M. S. Dresselhaus (1965),
7793. “de haas-van alphen effect in pyrolytic and single-crystal
Trushin, Maxim, and John Schliemann (2007), “Minimum graphite,” Phys. Rev. 140, A1429–A1447.
electrical and thermal conductivity of graphene: A quasi- Wirtz, Ludger, and Angel Rubio (2004), “The phonon dis-
classical approach,” Phys. Rev. Lett. 99, 216602. persion of graphite revisited,” Solid State Communications
Tu, Zhan-chun, and Zhong-can Ou-Yang (2002), “Single- 131, 141 – 152.
walled and multiwalled carbon nanotubes viewed as elastic Wunsch, B, T Stauber, F Sols, and F Guinea (2006), “Dy-
tubes with the effective young’s moduli dependent on layer namical polarization of graphene at finite doping,” New
number,” Phys. Rev. B 65, 233407. Journal of Physics 8 (12), 318.
Tworzydlo, J, I. Snyman, A. R. Akhmerov, and C. W. J. Xin, Zhou, Zhou Jianjun, and Ou-Yang Zhong-can (2000),
Beenakker (2007), “Valley-isospin dependence of the quan- “Strain energy and young’s modulus of single-wall carbon
tum hall effect in a graphene p-n junction,” Phys. Rev. B nanotubes calculated from electronic energy-band theory,”
76, 035411. Phys. Rev. B 62, 13692–13696.
Tworzydlo, J, B. Trauzettel, M. Titov, A. Rycerz, and Xu, S, J. Cao, C. C. Miller, D. A. Mantell, R. J. D. Miller,
C. W. J. Beenakker (2006), “Sub-poissonian shot noise in and Y. Gao (1996), “Energy dependence of electron life-
graphene,” Phys. Rev. Lett. 96, 246802. time in graphite observed with femtosecond photoemission
Uchoa, Bruno, and A. H. Castro Neto (2007), “Supercon- spectroscopy,” Phys. Rev. Lett. 76, 483–486.
ducting states of pure and doped graphene,” Phys. Rev. Yan, Jun, Yuanbo Zhang, Philip Kim, and Aron Pinczuk
Lett. 98, 146801. (2007), “Electric field effect tuning of electron-phonon cou-
Vafek, Oskar (2006), “Thermoplasma polariton within scal- pling in graphene,” Phys. Rev. Lett. 98, 166802.
ing theory of single-layer graphene,” Phys. Rev. Lett. 97, Yang, Li, Marvin L. Cohen, and Steven G. Louie (2007a),
266406. “Excitonic effects in the optical spectra of graphene
Vafek, Oskar (2007), “Anomalous thermodynamics of nanoribbons,” Nano Letters 7 (10), 3112–3115.
coulomb-interacting massless dirac fermions in two spatial Yang, Li, Cheol-Hwan Park, Young-Woo Son, Marvin L. Co-
dimensions,” Phys. Rev. Lett. 98, 216401. hen, and Steven G. Louie (2007b), “Quasiparticle energies
Varchon, F, R. Feng, J. Hass, X. Li, B. Ngoc Nguyen, and band gaps in graphene nanoribbons,” Phys. Rev. Lett.
C. Naud, P. Mallet, J.-Y. Veuillen, C. Berger, E. H. Conrad, 99, 186801.
and L. Magaud (2007), “Electronic structure of epitaxial Yang, Xiao, and Chetan Nayak (2002), “Electrical and ther-
graphene layers on sic: Effect of the substrate,” Phys. Rev. mal transport by nodal quasiparticles in the d -density-
Lett. 99, 126805. wave state,” Phys. Rev. B 65, 064523.
Visscher, P B, and L. M. Falicov (1971), “Dielectric screening Yao, Yugui, Fei Ye, Xiao-Liang Qi, Shou-Cheng Zhang, and
in a layered electron gas,” Phys. Rev. B 3, 2541–2547. Zhong Fang (2007), “Spin-orbit gap of graphene: First-
Vozmediano, M A H, M. P. López-Sancho, and F. Guinea principles calculations,” Phys. Rev. B 75, 041401.
(2002), “Confinement of electrons in layered metals,” Phys. Zarea, Mahdi, and Nancy Sandler (2007), “Electron-electron
Rev. Lett. 89, 166401. and spin-orbit interactions in armchair graphene ribbons,”
 54

Phys. Rev. Lett. 99, 256804. Kim (2005), “Experimental observation of quantum hall
Zhang, C-H, and Yogesh N. Joglekar (2007), “Wigner crys- effect and berry’s phase in graphene,” NATURE 438, 201.
tal and bubble phases in graphene in the quantum hall Zhong-can, Ou-Yang, Zhao-Bin Su, and Chui-Lin Wang
regime,” Phys. Rev. B 75, 245414. (1997), “Coil formation in multishell carbon nanotubes:
Zhang, L M, and M. M. Fogler (2008), “Nonlinear screening Competition between curvature elasticity and interlayer ad-
and ballistic transport in a graphene p-n junction,” Phys. hesion,” Phys. Rev. Lett. 78, 4055–4058.
Rev. Lett. 100, 116804. Zhu, Yun, and J. W. Cai (2007), “Ultrahigh sensitivity hall
Zhang, Y, Z. Jiang, J. P. Small, M. S. Purewal, Y.-W. Tan, effect in magnetic multilayers,” Applied Physics Letters
M. Fazlollahi, J. D. Chudow, J. A. Jaszczak, H. L. Stormer, 90 (1), 012104.
and P. Kim (2006), “Landau-level splitting in graphene in Ziegler, K (1998), “Delocalization of 2d dirac fermions: The
high magnetic fields,” Phys. Rev. Lett. 96, 136806. role of a broken supersymmetry,” Phys. Rev. Lett. 80,
Zhang, Yuanbo, Yan-Wen Tan, Horst L. Stormer, and Philip 3113–3116.