Single electron tunneling in large scale nanojunction arrays with
Shilpi Karmakar,*aSusmit Kumar,aPasquale Marzo,aElisabetta Primiceri,aRiccardo Di Corato,a
Ross Rinaldi,abPier Giorgio Cozzi,cAlessandro Paolo Bramantidand Giuseppe Maruccio*ae
Received 27th August 2011, Accepted 15th January 2012
We report on the fabrication and single electron tunneling behaviour of large scale arrays of nanogap
electrodes bridged by bisferrocene–gold nanoparticle hybrids (BFc–AuNP). Coulomb staircase was
observed in the low temperature current–voltage curves measured on the junctions with asymmetric
tunnel barriers. On the other hand, junctions with symmetric tunneling barrier exhibited mere
nonlinear current voltage characteristics without discrete staircase. The experimental results agreed
well with simulations based on the orthodox theory. The junction resistance showed thermally
activated conduction behaviour at higher temperature. The overall voltage and temperature dependent
results show that the transport behaviour of the large arrays of single particle devices obtained by
a facile optical lithography and chemical etching process corresponds with the behaviour of single
particle devices fabricated by other techniques like e-beam lithography and mechanical breaking
Single electron devices exploit controlled electron tunneling
between electrodes and coulomb islands for many applications
from nanoelectronics to metrology and frontier research.1–10The
precision and architecture of the nanoscale separation between
the device elements play a decisive role in controlling electron
tunneling and device properties/response.11To date various
techniques have been implemented, viz. electron-beam lithog-
raphy,12shadow evaporation,13mechanically controllable break
approaches are valid for fabrication of only a small number of
units and not for large-scale processes due to the cost and time
Several novel strategies aspiring for parallel processing have
been researched, although their applicability to large-scale
fabrication of single-electron devices has not yet been fully
established. Krahne et al.18have reported a procedure to obtain
nanometric gaps based on optical lithography and wet etching of
a quantum well heterostructure. As a drawback, however, its
usability was restricted to cryogenic temperatures because of the
high current leakage. Recently G. Maruccio et al.19,20have
further developed this procedure to permit the usage at room
temperature of these nanogap electrodes. The importance of this
approach lies in its low cost, simple fabrication process and its
wafer scale applicability.
A smart approach to probe molecular conduction is to use
metal nanoparticles (NPs) as a bridge between organic mono-
layers formed on metallic electrodes.21In this method, a metal/
molecule/NP/molecule/metal bridge is formed, and the current
through the molecular/nanoparticle layers is measured in series
and characterized. Previous studies using the nanoparticle bridge
approach focused primarily on conductance through conjugated
molecules with thiol terminal groups linked to the gold nano-
particle. Amlani et al.22presented the particle bridge concept by
measuring conductance through a monolayer of (1-nitro-2,5-
diphenylethynyl-4*-thioacetyl)benzene. Long and co-workers23
have studied magnetic nanoparticle assembly by comparing
conductance through undecanethiol, oligo(phenylene ethynyl
ene)-dithiol, and oligo(phenylene vinylene)dithiol. Dadosh and
coworkers21extended this approach to include analysis of three
different conjugated dithiol molecules assembled between
nanoparticles, where the particle bridge enables the analysis of
single molecules. Potentially, nanoparticles with functional
organic shells are particularly interesting because of size and
shape-dependent properties arising from quantum confinement
along with an enhanced stability and easier interconnection
strategy compared to solo-molecular or solo-nanoparticle
aNational Nanotechnology Laboratory, Istituto Nanoscienze-CNR, Via
Arnesano, I-73100 Lecce, Italy. E-mail: firstname.lastname@example.org;
bDepartment of Innovation Engineering, University of Salento, Via
Arnesano, I-73100 Lecce, Italy
cDipartimento di Chimica ‘‘G. Ciamician’’, Universit? a di Bologna, Via
F. Selmi 2, I-40126 Bologna, Italy
dSTMicroelectronics Srl, Distretto Tecnologico, Via per Arnesano, I-73100
eDepartment of Physics, University of Salento, Via Arnesano, I-73100
† Electronic supplementary information (ESI) available. See DOI:
This journal is ª The Royal Society of Chemistry 2012Nanoscale, 2012, 4, 2311–2316 | 2311
Dynamic Article LinksC
Cite this: Nanoscale, 2012, 4, 2311
devices. In this letter we report on the fabrication and transport
studies of devices based on bisferrocene molecule–gold nano-
particle, BFc–AuNP, hybrid systems incorporated in mesa
nanojunctions19(Fig. 1(a) and (b)). The present study is an
attempt to exploit the peculiar properties of metallocene mole-
cules and to add further functionality to the nanodevices due to
the presence of multivalent redox states, which in principle could
be utilized for ON–OFF switching,24,25bistable operations,26
molecular spintronics,27etc. Depending on the tunneling rates
and regime, symmetric and asymmetric double barrier tunneling
junctions (DBTJ) were obtained with the latter exhibiting a clear
Coulomb staircase. No magnetoresistance effects were observed
as a result of an external magnetic field, pointing out that the BFc
molecule remains diamagnetic
even after immobilization
Preparation and characterization of the hybrids
To prepare the hybrids (Fig. 1(b)), the oleyl amine stabilized gold
nanoparticles (?5 nm) were first mixed with bisferrocene
solution (1 : 50) in chloroform for 48 hours at room temperature.
The detailed synthesis procedure for these thiol ended BFc
molecules has been described elsewhere.31The hybrids were then
precipitated from chloroform by adding equal volume of acetone
and ethanol, washed four times after centrifugation and finally
redispersed in chloroform to obtain the desired concentration.
To confirm the formation of hybrids, electrochemical studies
were performed on the solutions and the results were compared
with those obtained from solutions containing the separate
components (i.e. gold nanoparticles and bisferrocene molecules).
Cyclic voltammograms are shown in Fig. 1(c) (tetra butyl
ammonium hexafluorophosphate, TBAPF6, was used as the
supporting electrolyte and dichloromethane as the solvent). For
both hybrids (blue triangles) and bisferrocenes (green triangles),
clear forward anodic and return cathodic peaks could be
observed with an average electrode potential value, E0av ¼
0.429 V and 0.459 V, respectively, which in both cases could be
ascribed to a 2-electron process associated with the reversible
oxidation of the two equivalent ferrocenyl moieties32with only
a small shift in the position and reduction/oxidation peaks
becoming closer after immobilization on gold nanoparticles. This
undoubtedly confirms the formation of hybrids33since this
signature is not present in the case of gold nanoparticles alone
Nanodevices realization and characterization
Fig. 1(d) shows the empty device configuration used for this
study. The final layer sequence consists of a 200 nm thick GaAs
buffer layer, a 300 nm thick AlxGa1?xAs lower barrier-layer,
a 20 nm GaAs quantum well, and a 100 nm thick AlxGa1?xAs
upper barrier-layer. The quantum well was selectively wet etched
to achieve 10 nm gaps, according to a procedure described in
detail elsewhere.18–20One typical mesa structure consisted of 18
nanojunctions (Fig. S1, ESI†), and we have analysed 11 sets of
such mesa arrays, in total 198 nanojunctions. In order to
immobilize the hybrids on the Au electrodes, we initially devel-
oped a self-assembled monolayer (SAM) of hexane-1,6-dithiols
on the electrode surface by incubating empty nanojunctions for
24 hoursinto a solution of 10?3Mhexane–1,6-dithiols in ethanol.
Then the nanojunctions were kept in the hybrid solution of BFc–
AuNP’s for 24 hours and successively washed with ethanol and
dried using nitrogen flow. The concentrated BFc–AuNP hybrid
solution was deliberately diluted to obtain preferably single
hybrid particles between the electrodes, although in some cases
we have also observed more than one hybrid particle attached
between the mesa electrodes. Out of 198 prepared nanojunctions,
around 35% showed hybrid attachment and hence conduction
after immobilization. The resistance of any typical empty
nanojunction without hybrids was in the order of teraohms at
room temperature. After immobilization of the hybrids, the
junction resistances decreased by several orders and were in the
range from a few hundreds of kohms to tens of mega ohms.
Fig. 1(e) shows the SEM image of one of the typical measured
nanojunctions with self-assembled BisFc–AuNP hybrids. The
final layer sequence in the device junction can be visualized as
a double barrier tunneling junction with hybrids coupled to
source/drain electrodes and NP expected to exchange electrons
Structure of the bisferrocene molecules used for this study. (b) Cartoon
representation of how the bisferrocene molecules are attached to the Au
nanoparticles. (c) Cyclic voltammograms demonstrating the assembly of
(BFc–AuNP) hybrids. (d) Schematic view of the mesa nanogapelectrodes
fabricated on GaAs/AlxGa1?xAs structures with a self-assembled hybrid
particleattachedin the gap.(e)SEMimageof oneofthe typicalmeasured
nanojunctions. The particular image shows 2–3 nanoparticles attached in
Realization of the nanodevices with (BFc–AuNP) hybrids. (a)
2312 | Nanoscale, 2012, 4, 2311–2316This journal is ª The Royal Society of Chemistry 2012
via bisferrocene molecules (?1.25 nm) through tunnel barriers.
Transport measurements were carried out in a cryogen-free
superconducting magnet in the temperature range from 1.5–
Results and discussion
In the weak coupling limit, transport through the nanojunctions
occurs through single electron charging of the conducting island
and is determined by overall dynamics of this process and ulti-
mately by the specific tunneling rates into and out of the hybrid
which define the tunneling regime (shell-tunneling or shell
filling)34–37and are related to the resistances and capacitances of
the first and the second electrodes (R1, C1and R2, C2respec-
tively). A typical property of single-electron transport is the
Coulomb staircase, the stepwise increase of electric current as
a function of source–drain voltage, where each step corresponds
to an addition/subtraction of one electron to/from the Coulomb
island. According to the orthodox theory,10,11,37,38the Coulomb
staircase could be observed when the tunneling junctions are
electrically asymmetric i.e. R1C1/R2C2[ 1 or ? 1 (shell filling
regime). This configuration is mostly satisfied in our nano-
junctions due to the asymmetric positioning of the nanocrystals
on the nonplanar gap. The hybrid particles are typically more
strongly attached to the lower electrode (1) than to the upper one
(2). As a result, we have observed a Coulomb staircase like effect
at low temperature in many devices using ?12.5 nm hybrids
(10 nm NP coated with a ?1.25 nm molecular layer). If, however,
symmetric junctions (R1C1z R2C2) were formed (which could
occur when a hybrid is positioned at the same tunneling distances
a pronounced staircase was generally observed.12
Fig. 2(a) shows a typical I–V curve from an asymmetric
junction (JAS) demonstrating a Coulomb staircase. The experi-
mental data (open circles) agreed well with the simulated results39
(red solid curve) using the orthodox theory and DBTJ model
with the following parameters: C1¼ 1.8 aF, C2¼ 1.1 aF, R1¼
4 MU, R2¼ 46 MU and Q0¼ 0.001e. This means that the hybrid
particle is more strongly attached to side 1 than side 2 (Fig. 1(d)).
On the other hand, in Fig. 2(b) an I–V curve from a symmetric
junction (JS) is reported with the Coulomb blockade but no
observable staircase. The corresponding simulated I–V charac-
teristics (red curve) were calculated with the following parame-
ters: C1¼ 1.1 aF, C2¼ 1.2 aF, R1¼ 65 MU, R2¼ 50 MU and
Q0¼ ?0.03e, showing that R1C1z R2C2in this case. The non-
conducting voltage interval DVDSis ?108 mV for JSand ?80 mV
for JAS. According to the orthodox theory, it is directly related to
the total capacitance CSand charging energy Ecrequired for
adding one electron to the Coulomb island:
DVDS¼ 2e/CS¼ 4Ec/e
This gives CS(and Ec) as ?2.96 aF (27 meV) and 4 aF (20 meV)
for JSand JASrespectively. The characteristic charging energy
can also be calculated from the self-capacitance value CHz
2p303d ¼ 2.1 aF for one isolated hybrid spherical particle of
represent the theoretically calculated behavior from the orthodox model of single electron transport. The parameters used for generating the simulated
curvesare indicatedin the figure.Insetsshowthe variationof junctioncurrentat higherbiasvoltages.Solid linesare fits toV3/2and quadraticdependence
for JAS(inset (a)) and JS(inset (b)). (c) Mean deviation of current–voltage behavior in 50 nanojunctions at 2 K. Histograms showing the distribution of
junction resistance (d) and bias gap (e) in 50 separate nanojunctions with AuNp–BisFe hybrids, measured at 2 K.
Current (I)–voltage (V) characteristics of the typical (a) asymmetric junction, JAS, and (b) symmetric junction, JS, at 2 K. The solid curves
This journal is ª The Royal Society of Chemistry 2012Nanoscale, 2012, 4, 2311–2316 | 2313
diameter d ¼ 12.5 nm using the expression Ec¼ e2/2CHgiving
Ecz 38 meV, which is not too far from the experimental value,
taking into account that other capacitances from electrodes were
not considered for the calculation.
Notably, a markedly difference in the high voltage–current
behaviour between the symmetric and asymmetric junctions was
also observed (insets of Fig. 2). The asymmetric junction shows
V3/2voltage dependence at higher biases whereas the symmetric
one exhibits a quadratic dependence13on the bias voltage, I f
V2. The red solid lines in the insets of Fig. 2 are the best fits to V3/2
and V2dependence of junction current on the bias voltage for
JASand JS, respectively. This deviation from linearity in I–V
curves, outside the Coulomb blockade region, indicates the
presence of small multiple barriers in the surrounding BFc
molecule capping the AuNP, which is suppressed at higher bias
voltages.14Fig. 2(c) shows the statistical deviation of junction
current from the mean current value. Apart from 4–5 junctions
showing very large deviation (not included in the distribution
analysis), we did not observe a significant deviation in the
symmetry and current value. This could be further understood
from the histograms showing junction resistance (Fig. 2(d)) and
bias gap voltage (Fig. 2(e)) distribution. The typical bias voltage
gap and junction resistance were about 76 mV and 10 ? 108U
respectively at 2 K.
Fig. 3(a) and (b) display a two dimensional colour plot of the
I–V characteristics at different temperatures for two typical JAS
and JSrespectively. The Coulomb blockade can be clearly seen in
both the figures at low temperature (black region). In the case of
JAS the thermal smearing out is much less prominent and
nonlinearity exists in the I–V curves till 200 K, which could be
ascribed to the presence of a larger tunneling barrier in the case
of the asymmetric junction. In the case of JSthere is a sharp
increase in current as temperature is raised. The I–V behaviour
becomes linear above 30 K. This can also be further understood
from the temperature dependence of resistances of the
Fig. 4 shows the junction resistances (RASand RS) plotted as
a function of inverse of temperature (1/T). According to the
orthodox theory, for this type of system, the thermal depen-
dencies are usually explained by Arrhenius activated behaviour
R ¼ R0exp (Eg/kBT)(2)
where Egis the activation energy of the charge carrier. Fig. 4
shows that for the symmetric junction the thermal dependence
above 30 K is mainly governed by Arrhenius dynamics. In the
case of the asymmetric junction, in the higher temperature region
(T > 50 K) the Arrhenius law is obeyed as well. The value of Eg
calculated from the best fits of eqn (2) with experimental data
was found to be 7.12 meV (JAS) and 8.24 meV (JS), which
correspond nicely with previous reports on monolayer alka-
nethiol coated AuNP’s.40The observed difference between Ec
and Eg(Eg? Ec) might be due to cotunneling, barrier suppres-
sion phenomenon, etc. However since the junction resistances
were much larger than the quantum resistance (?104U), we
discarded cotunneling to be responsible for this discrepancy.41
Also, since small barrier suppression does not affect the
Coulomb blockade threshold voltage,42we can neglect it too.
This difference in energy scales might instead be ascribed to
a voltage divider effect in the junction which takes into account
the voltage drop in the BisFc layer capping the AuNP.19,43In this
scenario, only a fraction of total bias voltage actually gets
current with temperature T and bias voltage V for (a) an asymmetric and
(b) a symmetric junction. The color axis represents the absolute value of
the junction current. The black region near zero bias at low temperature
indicates the Coulomb blockade region.
Three dimensional plots showing the variation of the junction
and RSare the resistances of the asymmetric and symmetric junctions
respectively. The solid lines represent the best fits to the Arrhenius law
(eqn (2)), showing thermally activated behavior at higher temperature.
(a) Plots of ln RASand ln RSversus inverse of temperature, RAS
2314 | Nanoscale, 2012, 4, 2311–2316This journal is ª The Royal Society of Chemistry 2012
applied to the AuNP and consequently, larger bias voltages are
required to charge the AuNP. From simple lever arm consider-
ations,43a rescaling factor around 1.5–1.7 can be estimated if
a constant resistivity is assumed, but this value should reasonably
increase considering that the metal NP is much more conductive
than the molecular layer. So, a factor around 3 seems realistic
and this would allow a good agreement among the different data.
Hence, the values of Eg derived from the T dependence are
considered to be more consistent, in this case. The conduction
below 10 K was found to be almost temperature independent,
which can be explained by the presence of multiple small barriers
in the BFc molecule as discussed above, leading to a multibarrier
conduction at low T. Conversely, as T is increased, the electrons
overcome these small barriers thermally, but still show tunneling
behaviour due to remaining larger barriers.
For the sake of completing our analysis, we also measured the
I–V curves in the presence of magnetic field, up to 7 T at 2 K.
Fig. 5 shows the variation of junction current and its numerically
calculated first derivative with applied bias at several external
magnetic fields from 0 to 7 T, for a typical nanojunction. Within
experimental limits, there was no observable shift or change in
the Coulomb blockade region or its threshold (Fig. 5(b)) with
applied magnetic field. It could be seen that the curves are stable
and almost identical for different externally applied magnetic
fields. This suggests that the bisferrocene molecule stays in its
inherent (neutral) diamagnetic state28–30even after immobiliza-
tion, which was not obvious since other somewhat similar
molecules were observed to change their magnetism once grafted
on a substrate as a result of a change in the oxidation state of the
In summary we present here single electron tunneling charac-
teristics of asymmetric and symmetric nanodevices based on
metal/molecule/nanoparticle/molecule/metal bridges. It is also
worth highlighting that this study demonstrates chip-level
fabrication of single-electron devices, but does not use any of the
sophisticated nanoscale pattern definition techniques that
generally have intrinsic limitations for large-scale processing.
The temperature-dependent conductance and I–V characteristics
were measured in the T range of 2–300 K. The conductance
decreased with a decrease in T and the I–V characteristics were
strongly nonlinear at low temperatures. Except for electrical
measurements, at no time was any device treated individually.
Fabrication of other nanoscale devices and sensors,20e.g.,
nanopores, magnetic molecules, and spin based nanodevices,
may also benefit from the approach described in this study.
This work was financially supported by the SpiDME European
Project (Grant: EU-FP6-029002) and the MIUR-FIRB Project
1 K. K. Likharev, Proc. IEEE, 1999, 87, 606–632.
2 D. L. Feldheim and C. D. Keating, Chem. Soc. Rev., 1998, 27, 1–12.
3 J. J. Welser, S. Tiwari, S. Rishton, K. Y. Lee and Y. Lee, IEEE
Electron Device Lett., 1997, 18, 278–280.
4 D. L. Klein, R. Roth, A. K. L. Lim, A. P. Alivisatos and
P. L. McEuen, Nature, 1997, 389, 699–701.
5 D. Loss and D. P. DiVincenzo, Phys. Rev. A, 1998, 57, 120–126.
6 G. Schon and A. D. Zaikin, Phys. Rep., 1990, 198, 237–412.
7 W. G. van der Wiel, S. De Franceschi, J. M. Elzerman, T. Fujisawa,
S. Tarucha and L. P. Kouwenhoven, Rev. Mod. Phys., 2003, 75, 1–22.
8 C. Kergueris, J. P. Bourgoin, S. Palacin, D. Esteve, C. Urbina,
M. Magoga and C. Joachim, Phys. Rev. B: Condens. Matter, 1999,
9 M. Rontani, E. Molinari, G. Maruccio, M. Janson, A. Schramm,
C. Meyer, T. Matsui, C. Heyn, W. Hansen and R. Wiesendanger,
J. Appl. Phys., 2007, 101, 081714.
10 D. Goldhaber-Gordon, M. S. Montemerlo, J. C. Love, G. J. Opiteck
and J. C. Ellenbogen, Proc. IEEE, 1997, 85, 521–540.
11 A. Schmid and Y. Leblebici, Robust Circuit and System Design
Methodologies for Nanometer-Scale Devices and Single-Electron
Transistors, in Third IEEE Conference on Nanotechnology, IEEE-
12 C. Vieu, F. Carcenac, A. P? epin, Y. Chen, M. Mejias, A. Lebib,
L. Manin-Ferlazzo, L. Couraud and H. Launois, Appl. Surf. Sci.,
2000, 164, 111–117.
13 G. Philipp, T. Weimann, P. Hinze, M. Burghard and J. Weis,
Microelectron. Eng., 1999, 46, 157–160.
14 M. A. Reed, C. Zhou, C. J. Muller, T. P. Burgin and J. M. Tour,
Science, 1997, 278, 252–254.
15 H. Park, A. K. L. Lim, A. P. Alivisatos, J. Park and P. L. McEuen,
Appl. Phys. Lett., 1999, 75, 301–303.
16 T. Deng, M. Prentiss and G. M. Whitesides, Appl. Phys. Lett., 2002,
17 S. Karmakar, S. Kumar, R. Rinaldi and G. Maruccio, J. Phys.: Conf.
Ser., 2011, 292, 012002.
18 R. Krahne, A. Yacoby, H. Shtrikman, I. Bar-Joseph, T. Dadosh and
J. Sperling, Appl. Phys. Lett., 2002, 81, 730–732.
calculated first derivative, dI/dV versus voltage (V) for a typical mesa
nanojunction with BFc–AuNP hybrids, at 2 K, in several magnetic fields
(0 to 7 T).
Plots of (a) current (I) versus voltage (V) and its (b) numerically
This journal is ª The Royal Society of Chemistry 2012Nanoscale, 2012, 4, 2311–2316 | 2315
19 G. Maruccio, P. Marzo, R. Krahne, A. Passaseo, R. Cingolani and Download full-text
R. Rinaldi, Small, 2007, 3, 1184–1188.
20 G. Maruccio, E. Primiceri, P. Marzo, V. Arima, A. Della Torre,
R. Rinaldi, T. Pellegrino, R. Krahne and R. Cingolani, Analyst,
2009, 134, 2458–2461.
21 T. Dadosh, Y. Gordin, R. Krahne, I. Khivirich, D. Mahalu,
V. Frydman, J. Sperling, A. Yacoby and I. Bar-Joseph, Nature,
2005, 436, 677.
22 I.Amlani,A. M.Rawlett, L. A.NagaharaandR. K. Tsui,Appl.Phys.
Lett., 2002, 80, 2761–2763.
23 D. P. Long,C. H. Patterson, M. H. Moore, D. S. Seferos,G. C.Bazan
and J. G. Kushmerick, Appl. Phys. Lett., 2005, 86, 153105.
24 H. X. He, X. L. Li, N. J. Tao, L. A. Nagahara, I. Amlani and R. Tsui,
Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 68, 045302.
25 M. Galperin, M. A. Ratner and A. Nitzan, Nano Lett., 2004, 5, 125–
26 J. Chen, W. Wang, J. Klemic, M. A. Reed, B. W. Axelrod,
D. M. Kaschak, A. M. Rawlett, D. W. Price, S. M. Dirk,
J. M. Tour, D. S. Grubisha and D. W. Bennett, Ann. N. Y. Acad.
Sci., 2002, 960, 69–99.
27 N. Baadji, M. Piacenza, T. Tugsuz, F. D. Sala, G. Maruccio and
S. Sanvito, Nat. Mater., 2009, 8, 813–817.
28 D. O. Cowan, J. Park, M. Barber and P. Swift, J. Chem. Soc. D, 1971,
29 W. H. Morrison and D. N. Hendrickson, Inorg. Chem., 1975, 14,
30 W. H. Morrison, S. Krogsrud and D. N. Hendrickson, Inorg. Chem.,
1973, 12, 1998–2004.
31 P. G. Cozzi, et al., Adv. Funct. Mater., 2011, submitted.
32 V. Arima, M. Iurlo, L. Zoli, S. Kumar, M. Piacenza, F. Della Sala,
F. Matino, G. Maruccio, R. Rinaldi, F. Paolucci, M. Marcaccio,
P. G. Cozzi and A. P. Bramanti, Nanoscale, 2012, 4, 813–823.
33 M. Yamada and H. Nishihara, Langmuir, 2003, 19, 8050–8056.
34 D. V. Averin and K. K. Likharev, Mesoscopic Phenomena in Solids,
North-Holland, Amsterdam, 1991.
35 H. van Houten, C. W. J. Beenakker and A. A. M. Staring, Single-
Charge Tunneling, Plenum, New York, 1992.
36 L. P. Kouwenhoven and P. L. McEuen, Single Electron Transport
Through A Quantum Dot, AIP, New York, 1998.
37 G. Maruccio and R. Wiesendanger, in Quantum Materials,
Nanocrystals, ed. D. Heitmann, Springer, Berlin Heidelberg, 2010,
38 A. E. Hanna and M. Tinkham, Phys. Rev. B: Condens. Matter, 1991,
39 P. Hadley, Single-Electron Device Java Simulation Programs, Institute
of Solid State Physics, Austria, http://lamp.tu-graz.ac.at/?hadley/set/
40 D. Greshnykh, A. Fr€ omsdorf, H. Weller and C. Klinke, Nano Lett.,
2008, 9, 473–478.
41 D. V. Averin and Y. V. Nazarov, Phys. Rev. Lett., 1990, 65, 2446.
42 A. N. Korotkov and Y. V. Nazarov, Phys. B, 1991, 173, 217–222.
43 G. Maruccio, M. Janson, A. Schramm, C. Meyer, T. Matsui,
E. Molinari, Nano Lett., 2007, 7, 2701–2706.
44 J. Brede, N. Atodiresei, S. Kuck, P. Lazi? c, V. Caciuc, Y. Morikawa,
G. Hoffmann, S. Bl€ ugel and R. Wiesendanger, Phys. Rev. Lett., 2010,
2316 | Nanoscale, 2012, 4, 2311–2316 This journal is ª The Royal Society of Chemistry 2012