Electron transport in quantum wire superlattices

Thomas Grange

Walter Schottky Institut, Technische Universitat M unchen,

Am Coulombwall 3, D-85748, Garching, Germany
(Dated: April 25, 2014)
Electronic transport is theoretically investigated in laterally conned semiconductor superlattices
using the formalism of non-equilibrium Greens functions. Velocity-eld characteristics are calcu-
lated for nanowire superlattices of varying diameters, from the quantum dot superlattice regime
to the quantum well superlattice regime. Scattering processes due to electron-phonon couplings,
phonon anharmonicity, charged impurities, surface and interface roughness and alloy disorder are
included on a microscopic basis. Elastic scattering mechanisms are treated in a partial coherent
way beyond the self-consistent Born approximation. The nature of transport along the superlattice
is shown to depend dramatically on the lateral dimensionality. In the quantum wire regime, the
electron velocity-eld characteristics are predicted to deviate strongly from the standard Esaki-Tsu
form. The standard peak of negative dierential velocity is shifted to lower electric elds, while
additional current peaks appear due to integer and fractional resonances with optical phonons.
I. INTRODUCTION
Electron transport in superlattices (SLs) has been
widely studied in the case of 1D periodic arrangement of
2D semiconductor layers
13
(Fig. 1a). In contrast, the na-
ture of charge transport in superlattices of lower dimen-
sionality remains an open question. The nature and e-
ciency of dissipative processes are well known to strongly
depend on the dimensionality, with longest lifetimes and
coherence times in 3D quantum conned structures
47
.
Quantum dot (QD) SLs are hence expected to dis-
play radically dierent non-equilibrium transport prop-
erties than quantum well (QW) SLs. Potential appli-
cations for high performance thermoelectric converters
8
,
photovoltaics
9
, and quantum cascade lasers
10,11
further
motivates the fundamental understanding of the non-
equilibrium transport properties of SLs with 3D quantum
connement.
Though ballistic transport in a QW SL is ideally a
pure 1D problem, scattering processes couples all spatial
degrees of freedom (Fig. 1a). In QW SLs, the in-plane
free electron motion possesses a continuous dispersion.
During the scattering processes, energy transfer occurs
between the axial and lateral motions. Hence the eec-
tive energy conservation laws in the motion along the SL
axis can greatly dier from the 3D one of the involved
scattering mechanisms. For example, in QW heterostruc-
tures, 3D elastic scattering processes due to static defects
act eectively as 1D inelastic scattering mechanisms with
respect to the motion along the SL. In contrast, in the
limit of purely 1D SL with 0D lateral motion (Fig. 1d),
the lateral state is frozen so that the 3D energy conser-
vation laws of the scattering processes apply directly to
the 1D motion along the SL.
The eect of dissipation on 1D quantum electron trans-
port has been investigated in nite quantum region such
as atomic wires
12
and stacked QDs
13
. Quantum dissi-
pative transport in QD SLs has been studied in Ref. 14
for high electric elds. However full velocity-eld char-
acteristics of 1D SLs have not been calculated so far. In
(a) (b) (c) (d)
z
y
x

FIG. 1: (Color online). Schematic of superlattices with dier-
ent lateral geometries. (a) shows a standard quantum well su-
perlattice consisting of 2D (x,y) innite layers of various ma-
terials stacked periodically along the z direction. (b),(c) and
(d) represent nanowire superlattices of dierent lateral sizes.
Dotted arrows are sketched of classical trajectories of charge
carriers through the various superlattices under an electric
eld applied along the z direction. Scattering processes occur
due to various eects (interaction with phonons, structural
disorder...). It illustrates the transition from transport in-
volving 3D scattering processes towards a pure 1D motion
with increasing quantum lateral connement. Cylindrical co-
ordinates (,,z) used in the calculations are indicated for the
(b) nanowire.
particular it is still unkown how the Esaki-Tsu negative
dierential velocity (NDV)
1
evolves when the lateral di-
mensionality is reduced towards QD SLs.
In this article, we present a detailed theory of elec-
tron transport in semiconductor SLs laterally conned
in nanowires using the framework of non-equilibrium
Greens functions (NEGF). The formalism presented here
allows to describe the transport in SLs for any lat-
eral dimensionality from the regime of coupled 0D QDs
(Fig. 1d) to the one of coupled 2D QWs (Fig. 1a). For
this purpose, we include in our modeling the scatter-
ing mechanisms that are known to be important in pla-
nar QW heterostructures as well as the ones specic to
a
r
X
i
v
:
1
3
1
2
.
5
2
1
8
v
2


[
c
o
n
d
-
m
a
t
.
m
e
s
-
h
a
l
l
]


2
4

A
p
r

2
0
1
4
2
quasi-0D systems. Specically, we consider the mecha-
nism of anharmonic polaron decay which has been shown
to be dominant in InGaAs self-assembled QDs
6,15
. In
addition, as the nature of elastic scattering mechanisms
evolves from incoherent to coherent as the dimensional-
ity decreases, we include a coherent treatment of elas-
tic scattering processes for lateral state-conserving pro-
cesses. The electronic transport properties are shown
to depend dramatically on the SL lateral dimensionality.
With increasing lateral connement, the drift velocity-
voltage characteristics evolve from the Esaki-Tsu form for
usual QW SL, dominated by a single NDV eect, towards
richer characteristics where (i) the linear conductivity in-
creases, (ii) the standard Esaki-Tsu NDV occurs at much
lower voltages, and (iii) large peaks due to integer and
fractional resonances with optical phonons appear.
The paper is organized as follows. The theory
of electron transport in laterally conned SLs is de-
scribed in section II. The application to the transport
in GaAs/AlGaAs nanowire superlattices of varying di-
ameters is presented and discussed in section III.
II. THEORY
Our aim is to model the non-equilibrium electron
transport in a nanowire SL heterostructure. We rst in-
troduce the general form of the Hamiltonian. We then
introduce an electronic basis set, and present the NEGF
formalism. We then treat the relevant scattering terms
arising from electron-phonon interactions and static dis-
order eects. Finally we present the method used for
solving the self-consistent NEGF equations.
A. Hamiltonian of the nanowire superlattice
We consider charge carriers in a nanowire SL in inter-
action with phonon modes. The following Hamiltonian
is used to model the system:

H =

H
0
+

V
e
+

V
e-ph
+

H
vib
+

V
m-f
e-e
, (1)
where

H
0
is the electronic Hamiltonian of a single
charge carrier in the idealized heterostructure under a
homogeneous electric eld (the length gauge will be used
throughout this paper);

V
e
is the static electronic poten-
tial due to disorder eects such as interface and surface
roughness, charged impurities and alloy disorder;

V
e-ph
represents the interaction terms between electrons and
phonons;

H
vib
is the vibrational Hamiltonian of the crys-
tal, including its anharmonic part;

V
m-f
e-e
represents the
mean-eld potential arising from the Coulomb interac-
tion with other charge carriers. The standard caret no-
tation is used to refer to operators.
B. Electronic structure and basis states
We consider cylindrical nanowires of diameter D = 2R
having SL heterostructures along their longitudinal axis.
We use cylindrical coordinates (z, , ) where z is the
nanowire axis (see Fig. 1b). The material composition
and modeling parameters are assumed to depend only
on the z coordinate. The electronic structure is modeled
considering a single band within the envelope function
approximation. We consider a basis set of the form:

,n
(z, , ) =

(z)
n
(, ), (2)
where the
i
(z) are dened below and
n
(, ) are eigen-
states of the Schrodinger equation on a homogeneous disk
(i.e. the assumed nanowire cross-section):

n
(, ) =
1

RJ
(mn+1))
(
ln
mn
)
J
mn
_

ln
mn

R
_
e
imn
(3)
where J
m
is the Bessel function of order m and
l
m
is its
lth zero. The indexes m
n
and l
n
are chosen such that
they are associated with the nth eigenvalue of the 2D
lateral motion at the z coordinate:
E
n
(z) =

2
_

ln
mn
_
2
2m

(z)R
2
(4)
where m

(z) is the material-dependent eective mass.

For each lateral mode we are then left with a z-dependent
electronic Hamiltonian:
n|

H
0
|p =
n,p
_

h
0
eF z +E
n
(z)
_
, (5)
where F is the electric eld applied along the z axis
and

h
0
=

2
2

z
1
m

(z)

z
+V
b
(z), (6)
with V
b
(z) being the material-dependent band-oset po-
tential. The eigenstates of the periodic

h
0
Hamilto-
nian satisfy the Bloch theorem and can be classied by
wavevector
z
[/l
p
, /l
p
] and miniband index

h
0

z
(z) =

z
(z) (7)
In order to discretize the basis set in the z direc-
tion, various choices are possible. A full real-space
discretization
16,17
by e.g. nite dierences requires a
large basis size in order to quantitatively describe the
miniband structure, rendering the NEGF numerical im-
plementation challenging. On the other hand, mode-
space approaches are also possible, and allow the de-
scription of the transport properties with a much smaller
basis
14,18,19
. Here, the Hilbert space is reduced by keep-
ing only a nite number of minibands, i.e. the low lying
states of the unbiased structure. An energy cut-o E
c
is
3
xed and only the minibands separated from the ground
state by less than E
c
are kept. Formally we dene the
following projection operator

P
z
=

0
0
<Ec

,z
|

z
| (8)
In the following, the non-equilibrium dynamics will be
computed within the subspace P generated by

P
z
. The
various operators

O dened on the full Hilbert space need
to be transformed in operators

O
r
acting within the P
subspace. The restrictions of the

h
0
unperturbed Hamil-
tonian and z position operator are simply dened by:

h
0r
=

h
0

P
z
(9)
z
r
=

P
z
z

P
z
(10)
Within the P subspace, we consider the localized ba-
sis which is constructed by diagonalizing the z
r
position
operator
14,19
:
z
r
|

(11)
where the |

( = 1, 2, ...) are sorted by increasing

eigenvalues. In addition, their phase is chosen in order

to fullll periodicity:

+pM
(z) =

(z pL) (12)
where p is an integer, M is the number of minibands and
L is the period length.
All the terms

V
(j)
in the Hamiltonian other than

h
0
are functions of the sole z operator. In order to con-
serve the commutation relation [

V
(j)
, z] = [

V
(j)r
, z
r
] = 0,
i.e. to conserve their local nature, we use the following
transformation:

V
(j)r
=

V
(j)
|

| (13)
This transformation allows us to keep exactly the same
form for the current operator
34
:

j
zr
=
i

[

H
r
, z
r
] =
i

h
0r
, z
r
] (14)
We also keep only a nite number of lateral states sat-
isfying E
i
E
0
< E
c
. In the following NEGF simulations,
it will be checked that the E
c
cut-o value is chosen high
enough so that adding higher minibands and/or higher
lateral states does not change signicantly the calculated
current. In the following, we will use exclusively oper-
ators acting in the P subspace and all the r subscripts
indicating restricted operators will be omitted for sim-
plicity.
C. Non-equilibrium Greens functions
1. Electron Greens functions
The NEGF formalism is generally introduced in terms
of four Greens functions (GFs)
2023
. Here the lesser,
upper, retarded and advanced electron GFs are dened
respectively in terms of the following quantities:
G
<
,,n
(t
1
, t
2
) = i c
+
,n
(t
2
) c
,n
(t
1
) (15a)
G
>
,,n
(t
1
, t
2
) = i c
,n
(t
1
) c
+
,n
(t
2
) (15b)
G
R
,,n
(t
1
, t
2
) = i(t
1
t
2
){ c
,n
(t
1
), c
+
,n
(t
2
)} (15c)
G
A
,,n
(t
1
, t
2
) = i(t
2
t
1
){ c
,n
(t
1
), c
+
,n
(t
2
)} (15d)
where c
+
,n
and c
,n
are respectively the creation and an-
nihilation operators in the electronic state |
,n
. It is
convenient to introduce the GF operators

G(t
1
, t
2
) de-
ned here by:
, n|

G(t
1
, t
2
)|, n

=
n,n
G
,,n
(t
1
, t
2
), (16)
where G refers to any of the four GFs. Since we do not
consider spin-dependent eects, the spin index is not in-
dicated here for simplicity and all quantities are implic-
itly spin diagonal. The spectral function reads in opera-
tor form:

A = i(

G
R

G
A
) = i(

G
>

G
<
) (17)
The retarded GF can be expressed from the spectral
function:

G
R
(t
1
, t
2
) = i(t
1
t
2
)

A(t
1
, t
2
) (18)
The lesser GF (15a) is an extension of the concept of
density matrix for two dierent times. Its value at equal
times corresponds to the density matrix:
(t) = i

G
<
(t, t). (19)
Here, o-diagonal terms of the GFs are considered in
the z direction but not in the lateral directions where
only the diagonal terms of the |
n
eigenmodes are con-
sidered. The calculations are thus independent of the
basis choice along the z-axis, allowing full coherent de-
scription of transport along the nanowire. In contrast,
the use of only diagonal GF terms in the lateral direc-
tions does not allow us to describe coherent transport
eects in the lateral directions. This seems a reason-
able approximation for thin nanowires where the lateral
scattering potentials are weak compared to the lateral
4
quantization energies. On the other hand, in the limit
of large diameter nanowire and planar heterostructures,
this is equivalent to considering only GF terms which are
diagonal in the in-plane momentum, similarly to previ-
ous implementations of NEGF in planar quantum well
heterostructures
1618
. In this later case, lateral localiza-
tion eects might become relevant at low temperature
and/or in presence of strong disorder eects, which re-
mains beyond the scope of this work.
In steady state, the GFs depend only on the time dier-
ence (t
2
t
1
). In this case we will use one-time-dependent
GFs

G(t
1
t
2
) =

G(t
1
, t
2
). Energy-dependent GFs are
then dened by Fourier-transforming these quantities

G(E) =
1

_
dt

G(t)e
iEt/
(20)
Moreover, in steady state, the advanced and retarded
GFs are linked by

G
A
(E) =

G
R
(E)
+
, (21)
so that the GFs are forming only two independent quan-
tities.
2. Phonon Greens functions
The Hamiltonian of the vibration of the crystal reads

H
vib
=

H
ph
+

V
a
(22)

H
ph
=

q,p
a
+
q,p
a
q,p
(23)
where

H
ph
and

V
a
are respectively the harmonic and an-
harmonic parts, and a
+
q,p
and a
q,p
are the creation and
annihilation operators for a phonon mode of branch p
with wavevector q. The phonon GFs are dened by
24
:
D
<
q,p
(t
1
, t
2
) = i

A
+
q,p
(t
2
)

A
q,p
(t
1
) (24)
D
>
q,p
(t
1
, t
2
) = i

A
q,p
(t
1
)

A
+
q,p
(t
2
) (25)
D
R
q,p
(t
1
, t
2
) = i(t
1
t
2
)[

A
q
(t
1
),

A
+
q,p
(t
2
)], (26)
with

A
q,p
= a
q,p
+ a
+
q,p
. (27)
In steady-state, we will use the same one-time notation
D(t) and energy-dependent GFs D(E) as for electronic
GFs (Eq. 20). When only the harmonic part of the
Hamiltonian is taken into account, the lesser and re-
tarded equilibrium phonon GFs for a mode of frequency
read respectively:
D
<(0)

(t) = i[(N

+ 1)e
it
+N

e
it
] (28)
D
R(0)

(t) = i(t)[e
it
e
it
], (29)
where N

is the Bose factor at the energy . In the

energy domain, it reads:
D
<(0)

(E) = 2i[(N

+ 1)(E + ) +N

(E )]
(30)
D
R(0)

(E) =
1
E +i0
+

1
E +i0
+
+
. (31)
Harmonic phonon GFs are usually directly used in elec-
tron transport calculations. However, in this work, we
will use instead anharmonic GFs for optical modes which
will be derived below (Sec. II F). This will allow us to
account for the mechanism of anharmonic polaron decay.
D. Equations of motion
The equations of motion in the NEGF formalism can
be expressed in terms of the so-called Dyson and Keldysh
relations; in steady-state, they read respectively
2023
:

G
R
(E) =
_
E

I

H
0


V
c

R
(E)
_
1
(32)

G
<
(E) =

G
R
(E)

<
(E)

G
A
(E) (33)
where

V
c
is the part of the interaction which is treated
coherently. Here it reads

V
c
=

V
m-f
e-e
+

V
rand
e
, where

V
m-f
e-e
is the mean-eld Coulomb potential and

V
rand
e
is the part
of the electronic disordered potential which is generated
randomly and treated coherently (see below). Similarly
to the spectral GF, the spectral self-energy is dened as

= i(

A
) = i(

<

>
), (34)
and the retarded self-energy can be expressed as:

R
(t) = i(t)

(t) = (t)(

<
(t)

>
(t)). (35)
In the energy domain, this relation reads:

R
(E) =
i
2

(E) +
1
2
H[

](E) (36)
where H denotes the Hilbert tranform H[

](E) =
P
_
dE

(E

)/(E E

).
5
E. Self-energies due to electron-phonon interaction
We consider the interactions of electrons with longi-
tudinal optical (LO), surface optical (SO) and longitudi-
nal acoustic (LA) phonon modes laterally conned in the
nanowire. Each interaction is of the form:

V
e-ph
=

q,p

f
q,p

A
q,p
. (37)
The form factors

f
q,p
are given in appendix A for the dif-
ferent modes. In the self-consistent Born approximation
(SCBA), the self-energies due to electron-phonon inter-
action are of the form:

<
(t) = i

q,p

f
q,p

G
<
(t)D
<
q,p
(t)

f
q,p
. (38)
Within the {|

} basis considered here, the nonvan-

ishing coupling terms being diagonal with respect to the
axial wavefunctions, the self-energy reads

<
n
(t) = i

q,p

f
(n,n

)
q,p
()f
(n

,n)
q,p
()G
<
n
(t)D
<
q,p
(t),
(39)
where
f
(n,n

)
q,p
() =
,n
|

f
q,p
|
,n
. (40)
1. Optical phonons
For longitudinal optical (LO) and surface optical
(SO) phonons, we assume wavevector-independent op-
tical phonon GFs. Indeed only long-wavelength optical
phonons with negligible dispersion are eectively coupled
to the relevant electronic states. We can then write

<(iO)
n
(t) = iD
<
LO
(t)

W
(n,n

)
iO
(, )G
<
n
(t), (41)
where
W
(n,n

)
iO
(, ) =

q
f
(n,n

)
q,iO
()f
(n

,n)
q,iO
(). (42)
In the energy domain, it reads

<(iO)
n
(E) = i

W
(n,n

)
iO
(, )

_
dE

2
D
<
LO
(E

)G
<
n
(E E

).
(43)
If the phonon GF D
<
LO
is taken harmonic, only quan-
tized energy exchanges are allowed. In QDs, it has
been shown that energy exchanges strongly diering from
the optical phonon energy take place due to simulta-
neous electron-phonon interaction and anharmonic cou-
plings among phonons
6,15
. In these previous works,
the electron-phonon interaction was diagonalized exactly,
and the Fermi golden rule was used to calculate scatter-
ing rates among polaronic states induced by anharmonic
couplings. Here, instead, the SCBA is used in order to
treat the electronoptical-phonon interaction, and anhar-
monicity is included in the optical-phonon GFs. As al-
ready checked in Ref. 14, the SCBA very well reproduces
the exact polaron formation, especially for temperatures
T verifying k
b
T < E
LO
where E
LO
is the optical phonon
energy. The calculation of anharmonic GFs is presented
below.
2. Acoustic phonons
In contrast to optical phonons, the acoustic phonons
are assumed to be harmonic and to have a linear dis-
persion. The corresponding self-energy can be expressed
as:

<
n
(t) = i

K
(n,n

)
,
(t)G
<
n
(t), (44)
where
K
(n,n

)
,
(t) =

q,LA
D
<(0)
q
(t)f
(n,n

)
q,LA
()f
(n

,n)
q,LA
() (45)
F. Greens functions of anharmonic phonons
We now calculate the anharmonic phonon GFs in-
volved in the electronoptical-phonon self-energy. We
retain only the cubic couplings in the anharmonic terms,
and thermal equilibrium population of phonon modes is
assumed. As we consider only diagonal terms in the
phonons GFs, the Dyson equation reads
D
R
q,p
(E) =
1
_
D
R(0)
q,p
(E)
_
1

R
q,p
(E)
, (46)
where
R
is the phonon retarded self-energy. The
lesser GF is then calculated using the Keldysh relation:
D
<
q,p
(E) = D
R
q,p
(E)
<
q,p
(E)D
A
q,p
(E). (47)
The phonon self-energy is calculated in the Born approx-
imation, taking into account cubic anharmonic terms.
The phonon lesser self-energy reads

<
q,p
(t) = i

q1,p1,q2,p2
|V
a
(q, q
1
, q
2
)|
2
D
<(0)
q1,p1
(t)D
<(0)
q2,p2
(t),
(48)
in which we have taken the harmonic phonon GFs in the
right hand side. The cubic anharmonic coupling V
a
is ex-
pressed in terms of the Gr uneisen constant as reported in
Ref. 15. The phonon spectral and retarded self-energies
are then given respectively by

j
(E) = i(
<
j
(E)
<
j
(E)) (49)
6

R
j
(E) =
1
2
H(
j
)(E)
i
j
(E)
2
(50)
where H denotes the Hilbert transform. Applying the
Dyson equation, we obtain the phonon lesser GF:
D
<
j
(E) =
4E
2
j

<
(E)
_
E
2
E
2
j
+E
j
H(
j
)(E)

2
+E
2
j

2
j
(E)
. (51)
G. Elastic scattering
Disordered potentials due to randomly distributed
charged impurities, alloy disorder, rough interfaces and
rough surfaces break the ideal symmetry of the struc-
ture, and couple the dynamics in the various directions.
In particular, they couple the 1D motion along the z su-
perlattice axis to the the lateral motion. In planar QW
heterostructures, as the lateral dispersion forms a con-
tinuum, the static disorder produces an incoherent and
irreversible evolution within the z-electron basis. In con-
trast, in the limit of purely 1D heterostructures, where
only one lateral state is involved, the evolution is com-
pletely reversible, as elastic couplings induce a unitary
evolution within the z-electron basis. A treatment of
elastic scattering beyond the SCBA is thus required in
these low-dimensional structures.
In previous studies, the surface roughness in nanowires
has been treated coherently by random generation of
surfaces
25,26
. Here, we consider in addition interface
roughness, impurity scattering and alloy scattering. In-
stead of generating microscopic random realizations of all
surfaces, interfaces, charge positions, and alloy atom po-
sitions, we develop a simpler approach based on their cor-
relations. We believe that it is sucient for low or mod-
erate disorder eects, which are anyway often not known
precisely from the microscopic point of view. The disor-
dered potentials are split into diagonal and o-diagonal
parts with respect to the lateral eigenstates:

V
diag
e
=

n
|

V
e
|
n
, (52)

V
o-diag
e
=

n=m

n
|

V
e
|
m
. (53)
The diagonal component is treated coherently, while
the o-diagonal part can only be treated incoherently
since we consider only diagonal elements of the GFs with
respect to the lateral states.
1. Elastic scattering self-energies
The o-diagonal elastic coupling component

V
o-diag
e
,
which does not conserve the lateral state, is treated
within the SCBA. The corresponding self-energy reads:

<
n
(t) = i

=n
V
nn
()V
n

n
()G
<
n
(t), (54)
where
V
nn
() =
,n
|

V
e
|
,n
. (55)
2. Elastic scattering coherent terms
In order to treat coherently the diagonal component

V
diag
e
, we rst calculate the following covariance matrix
M
n
(, ) = V
nn
()V
nn
(). (56)
We then generate random potentials with correlations
that obey this calculated covariance. To this purpose,
we need to rotate the covariance matrix in its principal
component basis. More precisely, the M
n
symmetric ma-
trix can be diagonalized in
M
n
= R
n
D
n
R
+
n
, (57)
where D
n
is a diagonal matrix and R
n
is a unitary ma-
trix. We then generate random diagonal matrices d
rand
n
in the following way: for each diagonal element of index
i, we generate a d
n
(i, i) random number obeying a Gaus-
sian distribution with a variance D
n
(i, i). The random
potential is then obtained by a back transformation into
the original basis:

V
rand
e
=

n
|
n

V
rand
e,n

n
|, (58)

V
rand
e,n
= R
n
d
rand
n
R
+
n
. (59)
The

V
rand
potential is then included in V
c
in the Dyson
equation (32) in order to be treated coherently. The all
NEGF calculation is then made for dierent generated
random potentials until the distribution of the calculated
observables reached the desired accuracy.
H. Field-periodic boundary conditions
We assume the nanowire SL to contain an innite num-
ber of periods of length L. The possible formation of
electrical eld domains is not considered in this work
3,27
.
Instead, the electric eld is assumed to have the same
periodicity as the SL. The eld-periodic boundary con-
dition for the mean-eld component of the electrostatic
potential V
m-f
reads
V
m-f
(z +pFL) = V
m-f
(z), (60)
7
where p is an arbitrary integer. Note that the externally
applied electric eld F is already included in H
0
. The
Poisson equation reads

2
V
m-f
z
2
=
e

s
(
e
(z) +
d
(z)) , (61)
where
e
(z) and
d
(z) are the charge densities of carriers
and dopants respectively. From Eq. 61, this eld-periodic
boundary assumption implies that the electron probabil-
ity distribution
e
(z) is periodic. In the length gauge
used here, the periodicity condition for the GFs reads
G
+pM,+pM,n
(E) = G
,,n
(E +pFL) (62)
where M is the number of minibands. Note that in
nanowires, the SL periodicity can be broken by the in-
clusion of the coherent disordered potentials (II G2). In
order to investigate such eects, the simulated period can
be increased to several SL periods. In this case, the pe-
riod length L in Eqs. 60 and 62 is replaced by L
p
= m
p
L
where m
p
is the number of SL periods included in one
computational period. In order to know the needed num-
ber of periods, the integer m
p
is increased until the cal-
culated observables tend towards constant values.
I. Electron density and current
The expectation values of the various observables can
be derived using the relation between the density matrix
and the lesser GF at equal times (Eq. 19). The electron
probability distribution is given by:

e
(z) = 2ie

,n
_
dE
2
G
<
n
(E)|

(z)|
2
, (63)
where the factor 2 arises from the spin index. Using
the expression of the current operator in terms of the
restricted operators (Eq. 13), the current reads
J
z
=
2e
V

,
|[

h
0
, z]|

n
_
dE
2
G
<
n
(E). (64)
J. Numerical details
The self-consistent problem formed by the Keldysh
equation, the Dyson relation, the expression of the var-
ious self-energies and the electrostatic Poisson equation
is solved iteratively. Starting from an initial guess of the
GFs, we calculate iteratively (i) the lesser

<
and upper

>
self-energies; (ii) the retarded self-energy (Eq. 36);
(iii) the coherent part of the interaction, comprising the
mean-eld electrostatic potential and the random poten-
tial; (iv) the retarded GF from the Dyson equation; (v)
the lesser GFs from the Keldysh relation. This procedure
is repeated until a self-consistent solution is reached, with
two convergence criteria based on both the lesser GF and
the calculated current. Before each step (i), the GFs are
renormalized in order to fulll the charge neutrality for
each period. It is checked that this renormalization factor
converges accurately towards unity as the convergence is
achieved.
The iterative steps (i) and (ii) are performed numer-
ically in the time-domain. It is advantageous since the
self-energies within the SCBA involves the product in
the time domain. In contrast, energy domain calcula-
tions would involve numerically expensive convolutions,
as a broad continuum of anharmonic optical phonons is
considered. On the contrary, the iterative steps (iv) and
(v) are more easily computed in the energy domain, so
that Fourier transforms of the GFs and the self-energies
are used respectively in between steps (v)-(i) and (ii)-
(iv). Though adaptive energy grids have been shown to
be useful in order to resolve GF peaks with a reduced
number of energy points
17
, a homogeneous energy grid is
used here in order to make use of a fast Fourier transform
algorithm.
A maximum coherence length l
c
is set in the calcula-
tions, so that we consider only the GF terms G
,,n
for
|

| l
c
. Due to the eld-periodic boundary con-
ditions, the GFs G
,,n
are calculated for belonging
to a given period, while varies away from within the
set coherence length l
c
. In the simulation of the nanowire
SL, we check that l
c
is taken large enough by increasing it
until convergence of the calculated current. In the regime
of QW SL, we have veried that there is no dierence in
the calculated observables considering one or several pe-
riods, provided the doping densities remain low enough
to prevent the formation of eld domain instabilities.
III. RESULTS AND DISCUSSIONS
A. Transport in nanowire SLs: transition from QW
to QD SLs
As an application of the theory presented above, we
present calculations of the electronic transport in a
GaAs/Al
0.3
Ga
0.7
As nanowire SL. One period consists in
5 nm GaAs (well) and 5 nm Al
0.3
Ga
0.7
As (barrier). The
SL is assumed to be homogeneously n-doped with a con-
centration of 10
16
cm
3
. Assumed values for surface and
interface roughnesses are given in appendix B. The fun-
damental subband has a width of 9.7 meV. It is separated
from the rst excited subband by a minigap of 122 meV.
In the following calculations only the fundamental sub-
band is considered (the cut-o value E
c
is taken around
100 meV). Fig. 2 shows the drift velocity as a function
of the nanowire diameter for two dierent electric elds
along the SL. For large nanowire diameters, the current
density tends toward a constant value. This is interpreted
as the 2D regime for the lateral motion. In contrast, for
smaller diameters, the quantum wire regime is reached,
and the current density varies with the lateral conne-
8
10 50 100 200
0
1
2
3
Electric field
10 kV/cm
20 kV/cm

D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
Nanowire diameter (nm)
100 10 1
E
1
-E
0
(meV)
FIG. 2: (Color online). Drift velocity in a nanowire superlat-
tice as a function of the nanowire diameter for various electric
elds. The superlattice period consists of 5 nm GaAs followed
by 5 nm AlGaAs. The temperature is set to 300 K.
0
10
20
50
100
150
20 40 60 80 100 120
0.0
0.1
0.2
0.3
0.5
1.0


Nanowire diameter
15 nm
30 nm
50 nm
150 nm
S
p
e
c
t
r
a
l

f
u
n
c
t
i
o
n

(
a
.
u
.
)


Nanowire diameter
15 nm
30 nm
50 nm
150 nm
E
l
e
c
t
r
o
n

d
e
n
s
i
t
y

(
a
.
u
.
)
Energy (meV)
FIG. 3: (Color online). Spectral function A
lat
(E) (upper
panel) and electron density n
lat
(E) (lower panel) for various
nanowire diameters (see text for denitions). The electric
eld is 20 kV/cm and the temperature is 300 K.
ment. In order to conrm this interpretation, we plot in
Fig. 3 the quantities
A
lat
(E) =
1
s

n
A
,,n
(E) (65)
n
lat
(E) =
1
s
Im
_

n
G
<
,,n
(E)
_
, (66)
which represent respectively the spectral function and the
electron density summed over the lateral states; s = R
2
is the nanowire cross-section and is the single basis
state per period considered. For a nanowire diameter of
0 10 20 30 40 50 60 70
0
1
2
3
4
5
6
7
0 10 20 30 40 50 60 70
Electric field (kV/cm)
D
r
i
f
t

v
e
l
o
c
i
t
y

(
m
.
s
-1
)
Voltage drop per period (mV)
Nanowire diameter
150 nm
50 nm
30 nm
15 nm


D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
Voltage drop per period (mV)
0.0 0.5 1.0 1.5
0
100
200

FIG. 4: (Color online). Drift velocityvoltage characteristics
at 300K for various diameters of the nanowire superlattice.
150 nm, the smooth shape of the spectral function indi-
cates that the lateral quantization eects are overcome
by broadening eects. Its at shape is characteristic of a
2D subband. On the other hand, for nanowire diameters
below 17 nm, the current displays relatively small varia-
tions in Fig. 2. This corresponds to a quasi-0D regime for
the lateral motion, where only the ground lateral state is
occupied, as shown in Fig. 3 for a nanowire diameter of
15 nm. In between these two limit regimes, large varia-
tions of the current densities are calculated in Fig. 2, cor-
responding to an intermediate dimensional regime where
several but still distinct lateral states are involved, as de-
picted in Fig. 3 for nanowire diameters of 30 and 50 nm.
Fig. 4 shows the calculated velocity-voltage character-
istics for various nanowire diameters. For a nanowire di-
ameter of 150 nm, where the 2D lateral regime is reached
in good approximation, the standard Esaki-Tsu behavior
is observed: negative dierential velocity (NDV) is ex-
pected to occur when the carriers are excited beyond an
inection point in the miniband dispersion
1
. For smaller
diameters, the characteristics deviate from this simple
standard behavior. The most remarkable dierence is the
appearance of a large peak when the voltage drop per pe-
riod matches the LO-phonon energy (
LO
=37 meV in
GaAs). In addition, smaller resonances become visible
at
LO
/2 and
LO
/3. The interpretation is the fol-
lowing: in the 2D regime, the lateral motion acts as a
continuous energy reservoir, which renders the 3D en-
ergy conservation laws of the scattering mechanisms not
directly visible on the SL transport properties. In par-
ticular, though LO phonons provide the most ecient
inelastic scattering processes, their quasi-monochromatic
nature is not evidenced on the velocityvoltage charac-
teristics. In contrast, in the limit of 1D SLs with 0D
lateral motion, the energy is exchanged directly between
the quantized SL levels (i.e. the Wannier-Stark states)
9
and the phonons. Hence resonances in the transport oc-
cur when two consecutive Wannier-Stark levels are sepa-
rated by the LO-phonon energy
LO
. The smaller res-
onances around
LO
/2 and
LO
/3 are attributed to
LO-phonon assisted transport between states distant of
2 and 3 periods respectively.
In the inset of Fig. 4, we observe an increase in the
electron mobility with decreasing NW thickness. This is
attributed to the decrease of the eciency of the scat-
tering processes with decreasing dimensionality. In addi-
tion, the position of the rst NDV peak shifts to smaller
voltage values with decreasing NW thickness. This can
be understood within the Esaki and Tsu semi-classical
model
1
, where the NDV peak is predicted to occur for
edF/ = 1 ( being the scattering time). Beyond this
critical point, the Esaki-Tsu model predicts that elec-
trons are accelerated beyond an inection point in the
energymomentum dispersion relation and hence expe-
rience a negative eective mass. This edF = / crit-
ical point can also be interpreted as the transition be-
tween the domain of validity of miniband transport and
Wannier-Stark hopping models
3
. In both pictures, the
overall reduction of scattering processes with decreasing
dimensionality explains qualitatively the observed shift
of the NDV peak towards lower voltages.
B. Role and nature of elastic scattering processes
In order to analyze the role and the nature of elas-
tic scattering processes, simulations with dierent treat-
ments of elastic scattering are shown in Fig. 5. First, we
discuss the comparison between full calculations with the
one neglecting elastic scattering. These two calculations
notably dier for thick NWs. In contrast, for thin NWs,
the relative dierence in the calculated transport is very
small, indicating that elastic scattering only plays a mi-
nor role. This strong reduction of elastic scattering pro-
cesses with decreasing dimensionality is consistent with
the intuitive expectation that elastic mechanisms are sup-
pressed by the discretization of energy levels.
Now we compare the calculations using the standard
SCBA, the selective SCBA dened above and the full
calculation. In the QW limit the standard SCBA, the
selective SCBA and the full calculation are found to
give the same results. In contrast, in the quantum wire
regime, the standard SCBA totally diers from the two
other models. The reason is that the standard SCBA
fails when the uctuations of the energy levels become
larger than their linewidths. In contrast, in the present
full calculation, the main coupling elements, which are di-
agonal with respect to the lateral eigenstates, are treated
coherently. If we now compare the full calculation to the
selective SCBA calculation, we can see that the selec-
tive SCBA provides an excellent approximation to the
full calculation, down to 15 nm thin nanowires. Our in-
terpretation is that the typical energy uctuations of the
levels stay smaller than the miniband width and are too
0 10 20 30 40 50 60 70
0
1
2
3
4
5
6
D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
D =15 nm
Full calculation
No elastic scattering
Standard SCBA
Selective SCBA

Voltage drop per period (mV)
0 10 20 30 40 50 60 70
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
D =30 nm
Full calculation
No elastic scattering
Standard SCBA
Selective SCBA


D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
Voltage drop per period (mV)
0 10 20 30 40 50 60 70
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
D =50 nm
Full calculation
No elastic scattering
Standard SCBA
Selective SCBA


D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
-
2
)
Voltage drop per period (mV)
0 10 20 30 40 50 60 70
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
D =150 nm
Full calculation
No elastic scattering
Standard SCBA
Selective SCBA


D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
Voltage drop per period (mV)
FIG. 5: (Color online). Drift velocityvoltage characteristics
at 300K for dierent treatments of the scattering mechanisms.
10
0 10 20 30 40 50 60 70
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
D =30 nm
Anharmonic phonons
Harmonic phonons


D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
Voltage drop per period (mV)
0 10 20 30 40 50 60 70
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
D =50 nm
Anharmonic phonons
Harmonic phonons


D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
Voltage drop per period (mV)
0 10 20 30 40 50
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
D =150 nm
Anharmonic phonons
Harmonic phonons


D
r
i
f
t

v
e
l
o
c
i
t
y

(
1
0
3

m
.
s
-
1
)
Voltage drop per period (mV)
FIG. 6: (Color online). Drift velocityvoltage characteristics
at 300K with and without the inclusion of phonon anhar-
monicity.
weak to induce localization eects.
C. Role of phonon anharmonicity
In Fig. 6 we show a comparison of the calculations
with and without the inclusion of phonon anharmonicity
for various NW diameters. While the inclusion of anhar-
monicity is found to be almost irrelevant for thick NWs
in the QW SL limit, striking dierences are observed for
thin NWs. Indeed the direct LO-phonon emission is grad-
ually suppressed with decreasing dimensionality towards
0D systems, while the mechanism of anharmonic polaron
decay becomes dominant. For very thin NW diameters
corresponding to the pure 1D SL limit, note that the sim-
ulation without phonon anharmonicity does not converge
due to the strong reduction of the linewidths below our
numerical possibilities in energy resolution. Hence our
calculations show a clear transition in the dominant en-
ergy dissipation pathway as the lateral dimensionality is
reduced, from direct LO-phonon emission to polaron an-
harmonicity.
IV. CONCLUSION
Within the NEGF formalism, we have developed a the-
ory of electronic transport in nanowire SLs. This model
allows the calculation of transport in a wide range of
NW SL thickness, for lateral dimensionality ranging from
0D to 2D, i.e. from QD SL to QW SL. We have stud-
ied how the transport evolves with this change of lat-
eral dimensionality. While velocity-voltage characteris-
tics are dominated by the standard Esaki-Tsu NDV in
QW SLs, electron-phonon resonances appear when the
lateral quantum regime is reached. In addition, the elec-
tron mobility increases and the NDV peaks are strongly
shifted to lower electric elds. This is accompanied by
important changes in the dominant scattering mecha-
nisms. In QW SLs, the transport is mainly controlled
by LO-phonon scattering and elastic scattering processes.
In contrast, when the lateral dimensionality is reduced,
these two mechanisms are progressively suppressed, and
anharmonic polaron decay becomes dominant.
V. ACKNOWLEDGMENTS
This work has been supported by the Alexander von
Humboldt foundation and the Austrian Science Fund
FWF through Project No. F25-P14 (SFB IR-ON).
S. Birner, M. Branstetter, C. Deutsch, P. Greck, G.
Koblm uller, M. Krall, T. Kubis, S. Rotter, K. Unter-
rainer, and P. Vogl are gratefully acknowledged for fruit-
ful discussions.
Appendix A: Electron-phonon interactions in
cylindrical nanowires
In this appendix we give the form factor of the
couplings of electrons with longitudinal-optical (LO),
surface-optical (SO), and longitudinal-acoustical (LA)
phonon modes conned in a cylindrical nanowire. We
consider the full connement of phonons by the nanowire
surfaces but neglect the eect of the SL interfaces.
11
1. Interaction with longitudinal optical modes
The polar interaction between electrons and optical
phonons conned in a cylindrical nanowire is treated
within the dielectric continuum model
28
. The zone-
center LO phonons involved in the couplings with elec-
trons are assumed to be monochromatic with frequency

LO
. The interaction with the longitudinal optical (LO)
modes of the nanowire reads

H
e-LO
=

m,l,qz
f
LO
m,l,qz
(, , z)

A
LO
m,l,qz
, (A1)
where

A
LO
m,l,qz
= a
+
m,l,qz
+ a
m,l,qz
(A2)
is the sum of phonon creation and annihilation operators
with opposite momentum. The form factor reads
f
LO
m,l,qz
(, , z) = C
LO
m,l,qz
J
m
_

l
m

R
_
e
im
e
iqzz
, (A3)
where
|C
LO
m,l,qz
|
2
=
e
2

LO
2
0
LJ
2
m+1
(
l
m
)(
l2
m
+R
2
q
2
z
)
_
1

s
_
.
(A4)

0
is the vacuum permittivity;
s
and

are respectively
the static and high frequency relative permittivities.
2. Interaction with surface optical modes
The interaction with the surface optical modes (SO)
reads
28

H
e-SO
=

m,l,qz
C
SO
m,l,qz
I
m
(q
z
)e
im
e
iqzz

A
SO
m,l,qz
, (A5)
where
|C
SO
m,l,qz
|
2
=
e
2

m,l,qz
2
0
LI
m
(q
z
R)(I
m1
(q
z
R) +I
m+1
(q
z
R))

1
q
z
R
_
1

s

_
(A6)
and
=
I
m
(q
z
R)((K
m1
(q
z
R) +K
m+1
(q
z
R))
K
m
(q
z
R)((I
m1
(q
z
R) +I
m+1
(q
z
R))

ext
(A7)
where
ext
is the relative permittivity outside the
nanowire.

2
m,l,qz
=
_
1 +

0

2
TO
(A8)
3. Interaction with acoustic phonons
We consider the deformation potential interaction with
acoustic phonons conned laterally in the nanowire.
In principle, the boundary condition couple longitudi-
nal acoustic (LA) and transverse acoustic (TA) modes.
Coupled LA-TA modes interacting with electrons have
been considered by Yu et al
29
. However, there, only
phonon modes with axial symmetry have been consid-
ered. We expect this approximation to be valid in
the thin nanowire limit but to completely fail for large
nanowires where LA phonons do not necessarily prop-
agate parallel to the nanowire axis. Here, instead, we
neglect the coupling between LA and TA modes (i.e. be-
tween compressive and shear modes) at the surface of the
nanowire, and consider the interaction of electrons with
all the LA modes. This treatment is exact in the bulk
limit and is expected to provide a reasonable approxima-
tion for nanowires of moderate size, as far as the NW
diameter is large compared to the lattice constant. We
believe such assumption for the boundary conditions does
not change signicantly the results for the electron trans-
port properties, as (i) the nanowire thicknesses studied
here are relatively large compared to the atomic scale
and (ii) the eciency of the acoustic phonon scattering
mechanism is always much weaker than other scattering
mechanisms. The displacement elds for purely compres-
sive phonon modes can be written as
u = (A9)
We consider the free boundary condition, so that u = 0
at the surface and can be written as
=

m,l,qz
C
m,l,qz
J
m
_

l
m

R
_
e
im
e
iqzz
(A10)
The phonon quantization reads
u =

m,l,qz
u
m,l,qz

A
LA
m,l,qz
, (A11)
with the normalization condition:
_
d
3
r|u
m,l,qz
(r)|
2
=

2
m,l,qz
, (A12)
is the material density and
m,l,qz
the phonon fre-
quency.
The integration using the form of Eq. A10 gives:
|C
m,l,qz
|
2
=

2L(
l2
m
+R
2
q
2
z
)J
2
m+1
(
l
m
)
(A13)
The Hamiltonian describing the electronLA-phonon
interaction is taken of the form

H
e-LA
= D . u, (A14)
12
where D is the deformation potential of the band of in-
terest. Here it gives:

H
e-LA
= D(
l2
m
/R
2
+q
2
z
)

, (A15)
and nally:

H
e-LA
=

m,l,qz
f
LA
m,l,qz
(, , z)

A
LA
m,l,qz
, (A16)
where the form factor reads:
f
LA
m,l,qz
(, , z) = C
LA
m,l,qz
J
m
_

l
m

R
_
e
im
e
iqzz
, (A17)
|C
LA
m,l,qz
|
2
=
D
2

l2
m
+R
2
q
2
z
2RLc
s
J
2
m+1
(
l
m
)
, (A18)
where c
s
the sound celerity. The dispersion relation reads

m,l,qz
= c
s
_

l2
m
R
2
+q
2
z
. (A19)
Appendix B: Elastic scattering mechanisms
Imperfections and intrinsic atomic disorder break the
structure symmetry and the crystal periodicity, thus re-
sponsible for elastic scattering processes. We consider
below the eects of surface roughness of the nanowires,
interface roughness of the superlattice heterostructures,
randomly located charged impurities and alloy disorder.
1. Interface roughness
We note z
0
i
the mean z-coordinate of the ith interface.
The deviation from this average position at the in-plane
coordinate r is denoted by z
i
(r). The interface rough-
ness can be statistically characterized by (i) the distri-
bution of the interface position along the heterostructure
growth axis, and (ii) its in-plane autocorrelation. All
interfaces are assumed to obey the same statistics. We
note P(z) the distribution of the z
i
interface deviation
and its root mean square. We note C(r) the in-plane
autocorrelation form factor so that:
z
i
(r
1
)z
i
(r
2
) =
2
C(|r
2
r
1
|) (B1)
We note V the heterostructure band oset. The cou-
pling correlations involved in the calculation of elastic
scattering processes read:
V
(ir)
nn
()V
(ir)
n

n
() = V
2
Y
nn

i
W
(i)

, (B2)
where the lateral form factor reads
Y
nn
=
_
dr
1
dr
2

n
(r
1
)
n
(r
2
)C(|r
2
r
1
|), (B3)
and the axial form factor for the ith interface reads
W
(i)

=
_
z
0
i+1
z
0
i1
dz
i
P(z
i
z
0
i
)w
(i)

(z
i
)w
(i)

(z
i
), (B4)
with
w
(i)

(z
i
) =
_
[z
0
i
,zi]
dz|

(z)|
2
. (B5)
In the numerical study, P(z) is taken as a Gaussian dis-
tribution with a standard deviation = 0.15 nm and the
autocorrelation C(r) decreases as an exponential e
r/
with a correlation length = 8 nm.
2. Surface roughness
We consider small uctuations of the surface around its
ideal cylindrical position. The surface of the nanowires
are considered as innite barriers for the electrons. Mo-
tivated by the fact that a homogeneous change R in the
nanowire diameter would induce a change
En
R
R in the
n-th lateral state energy, we make the following assump-
tion for the surface roughness couplings
30
:
n|V
(sr)
(, z)|n

=
_
E
n
R
E
n

R
R
,z
(B6)
We assume an exponential autocorrelation of the form
R(
1
, z
1
)R(
2
, z
2
) = R
2
e
|z2z1|/L
s
c
e
R| arg[e
i(
2

1
)
]|/L
s
c
(B7)
in which we have factorized the axial and azimuthal cor-
relation terms in order to simplify the calculations. Fi-
nally we nd that the correlations of the coupling terms
are given by
V
(sr)
nn
()V
(sr)
n

n
() =
4R
2
R
2
E
n
E
n
Y
(sr)
nn
W
(sr)

, (B8)
where the lateral form factor reads
Y
(sr)
nn
=
_

d
2
e
i(mnm
n
)
e
R||/L
s
c
=
_
_
_
L
s
c
R
_
1 e
R/L
s
c

if m
n
= m
n

R[1e
R/L
s
c
(1)
mnm
n

]
L
s
c
[(mnm
n
)
2
+(R/L
s
c
)
2
]
if m
n
= m
n

,
(B9)
and the axial form factor reads
W
(sr)

=
_
dz
_
dz

(z)|
2
|

(z

)|
2
e
|zz

|/L
s
c
. (B10)
In the numerical study, the standard deviation is taken
as R = 0.15 nm and the surface correlation length is
taken as L
s
c
= 8 nm.
13
3. Charged impurities
The random location of charged impurities is also a
source of elastic scattering. For simplicity of the calcu-
lation and for clear comparison with the 2D limit, we
assume that the lling factor of the nanowires is unity.
The Coulomb potential created by ionized impurities at
the position r
d
reads
V
(c)
r
d
(r) =
e
2
e
(|rr
d
|)/Ls
4
0

s
|r r
d
|
(B11)
where L
s
is the screening length. Following Nelander
and Wacker
31
, L
s
is calculated within the Debye screen-
ing model for a 3D homogeneous electron gas with the
same average electron density and at the lattice temper-
ature. After performing an in-plane Fourier transform
this potential reads
32
:
V
(c)

d
,z
d
(, z) =
_
d
2
k

4
2
e
ik

.
e
2
2
0

s
k
s
e
ks|zz
d
|
(B12)
with k
s
=
_
k

+ 1/L
2
s
. The coupling correlation terms
are obtained by averaging over the dierent possible po-
sition of the charged impurities:
V
(c)
nn
()V
(c)
n

n
()
z
d
,
d
=
_
dz
d
_
d
2

d
N
d
(z
d
)

,n
|V
(c)
z
d
,
d
|
,n

,n
|V
(c)
z
d
,
d
|
,n

. (B13)
After some algebra we nd:
V
(c)
nn
()V
(c)
n

n
() =
e
4
2
2
0

2
s
_
+
0
k

dk

k
2
s
F
2
k

(n, n

)
k

(, ),
(B14)
where F
k
are Hankel transform coecients:
F
k
(n, n

) =
_
R
0
d
n
()
n
()J
m
n,n

(k) , (B15)
with m
n,n
= |m
n
m

n
|, and the axial form factor reads:

k
(, ) =
_
dz
1
_
dz
2
|

(z
1
)|
2
|

(z
2
)|
2

k
(z
1
, z
2
),
(B16)
with

k
(z
1
, z
2
) =
_
dzN
d
(z)e
k(|z2z|+|z1z|)
. (B17)
4. Alloy disorder
We consider a ternary alloy of the form A
x
B
1x
C. We
assume a random and uncorrelated distribution of the
atoms A and B. We assume that, at the atomic scale,
the carriers experience local band oset potentials V
AC
or V
BC
with probability x and 1 x, respectively
16,33
.
Here these potentials are taken as the one of the binary
compounds. The mean potential reads
V (z) = xV
AC
+ (1 x)V
BC
, (B18)
while the alloy scattering potential is dened by:
V = V
AC
V
BC
(B19)
The variance of the local band oset is then:
(V V )
2
= x(V
AC
V )
2
+ (1 x)(V
BC
V )
2
= x(1 x)V
2
(B20)
We denote V the volume occupied by a binary pair of
atoms (V = a
3
/4 for zinc-blende structures, a being the
lattice constant). The alloy disorder coupling terms read
V
(a)
nn
()V
(c)
n

n
() = VV
2
A

R
nn
(B21)
where x(z) is the z-dependent alloy composition, and
where the axial and lateral form factors read respectively:
A

=
_
Alloy
dz|

(z)|
2
|

(z)|
2
x(z)[1 x(z)], (B22)
R
nn
=
_
d
2
|
n
()|
2
|
n
()|
2
. (B23)

Electronic address: [email protected]; Present

address: Institut NeelCNRS, 25 av. des Martyrs, 38042
Grenoble, France.
1
L. Esaki and R. Tsu, IBM Journal of Research and Devel-
opment 14, 61 (1970).
2
A. Sibille, J. Palmier, H. Wang, and F. Mollot, Phys. Rev.
Lett. 64, 52 (1990).
3
A. Wacker, Physics Reports 357, 1 (2002).
4
U. Bockelmann and G. Bastard, Phys. Rev. B 42, 8947
(1990).
5
H. Benisty, C. M. Sotomayor-Torr`es, and C. Weisbuch,
Phys. Rev. B 44, 10945 (1991).
6
E. Zibik, T. Grange, B. Carpenter, N. Porter, R. Ferreira,
G. Bastard, D. Stehr, S. Winnerl, M. Helm, H. Liu, et al.,
Nature materials 8, 803 (2009).
7
T. Grange, Phys. Rev. B 80, 245310 (2009).
8
Y.-M. Lin and M. Dresselhaus, Phys. Rev. B 68, 075304
(2003).
14
9
A. Nozik, Physica E 14, 115 (2002).
10
N. Wingreen and C. Staord, Quantum Electronics, IEEE
Journal of 33, 1170 (1997), ISSN 0018-9197.
11
T. Grange, arXiv preprint arXiv:1301.1258 (2013).
12
T. Frederiksen, M. Paulsson, M. Brandbyge, and A.-P.
Jauho, Physical Review B 75, 205413 (2007).
13
C. Gnodtke, G. Kielich, E. Sch oll, and A. Wacker, Phys-
ical Review B 73, 115338 (2006).
14
N. Vukmirovic, Z. Ikonic, D. Indjin, and P. Harrison, Phys.
Rev. B 76, 245313 (2007).
15
T. Grange, R. Ferreira, and G. Bastard, Phys. Rev. B 76,
R241304 (2007).
16
R. Lake, G. Klimeck, R. C. Bowen, and D. Jovanovic, Jour-
nal of Applied Physics 81, 7845 (1997).
17
T. Kubis, C. Yeh, P. Vogl, A. Benz, G. Fasching, and
C. Deutsch, Phys. Rev. B 79, 195323 (2009).
18
S. Lee and A. Wacker, Phys. Rev. B 66, 245314 (2002).
19
S. Lee, F. Banit, M. Woerner, and A. Wacker, Physical
Review B 73, 245320 (2006).
20
S. Datta, Electronic transport in mesoscopic systems
(Cambridge university press, 1997).
21
H. Haug and A.-P. Jauho, Quantum kinetics in transport
and optics of semiconductors, vol. 123 (Springer, 2008).
22
C. Caroli, R. Combescot, P. Nozieres, and D. Saint-James,
Journal of Physics C: Solid State Physics 4, 916 (1971).
23
G. D. Mahan, Physics Reports 145, 251 (1987).
24
E. N. Economou, Greens functions in quantum physics,
vol. 3 (Springer, 1984).
25
J. Wang, E. Polizzi, A. Ghosh, S. Datta, and M. Lund-
strom, Applied Physics Letters 87, 043101 (2005).
26
A. Lherbier, M. P. Persson, Y.-M. Niquet, F. Triozon, and
S. Roche, Physical Review B 77, 085301 (2008).
27
L. Esaki and L. Chang, Phys. Rev. Lett. 33, 495 (1974).
28
H. Xie, C. Chen, and B. Ma, Physical Review B 61, 4827
(2000).
29
S. Yu, K. Kim, M. A. Stroscio, and G. Iafrate, Physical
Review B 51, 4695 (1995).
30
H. Sakaki, T. Noda, K. Hirakawa, M. Tanaka, and T. Mat-
susue, Applied physics letters 51, 1934 (1987).
31
R. Nelander and A. Wacker, Journal of Applied Physics
106, 063115 (2009).
32
H. Haug and S. S. W. Koch, Quantum theory of the optical
and electronic properties of semiconductors (World scien-
tic, 2009).
33
G. Bastard, Wave mechanics applied to semiconductor het-
erostructures (New York, NY (USA); John Wiley and Sons
Inc., 1990).
34
In the work of Lee and coworkers
18,19
, an implicit trun-
cation of the Hamiltonian leads instead to an additional
current term Jscatt which is found to be numerically negli-
gible.