You are on page 1of 7

Materials Today  Volume 16, Number 6  June 2013 RESEARCH

Bandgap engineering in a nanowire:

RESEARCH: Review
self-assembled 0, 1 and 2D quantum
structures
Jordi Arbiol1,2,*, Maria de la Mata1, Martin Eickhoff3 and Anna Fontcuberta i Morral4,*
1
Institut de Ciencia de Materials de Barcelona, ICMAB-CSIC, Campus de la UAB, E-08193 Bellaterra, CAT, Spain
2
Institucio Catalana de Recerca i Estudis Avancats (ICREA), E-08010 Barcelona, CAT, Spain
3
I. Physikalisches Institut, Justus-Liebig-Universitat Gieen, Heinrich-Buff-Ring 16, DE-35392 Gieen, Germany
4
Laboratoire des Materiaux Semiconducteurs, Ecole Polytechnique Federale de Lausanne, 1015 Lausanne, Switzerland

Inherent to the nanowire morphology is the exciting possibility of fabricating materials organized at the
nanoscale in three dimensions. Composition and structure can be varied along and across the nanowire,
as well as within coaxial shells. This opens up a manifold of possibilities in nanoscale materials science
and engineering which is only possible with a nanowire as a starting structure. As the variation in
composition and structure is accompanied by a change in the band structure, it is possible to conne
carriers within the nanowire. Interestingly, this results in the formation of local two, one and zero-
dimensional structures from an electronic point of view within the nanowire. This novel palette of nano-
structures paves the way toward novel applications in many engineering domains such as lasers, high-
mobility transistors, quantum information and energy harvesting. In the present review we summarize
and give an overview on recent achievements in the design and growth of advanced quantum structures
starting from nanowire templates. The quantum structures presented have been grown by molecular
beam epitaxy and correspond to different connement approaches: quantum wells (2D), quantum wires
(1D) and quantum dots (0D).

Introduction absorption [1820]. In the case of energy harvesting, for example,


The latest developments in nanoscale self-assembly and in parti- tandem solar cells based on semiconductor heterostructures in a
cular in the area of nanowire growth have opened up many nanowire [21], allow a better utilization of the solar spectrum and
possibilities in materials science and engineering. In particular, thereby result in an increase in the efciency of the cells by
the control that scientists have gained in achieving the required absorbing a wider range of light wavelengths [22]. Furthermore,
composition, shape and structure of nanowires [15] and also on improved thermoelectric efciencies can be also obtained by
the variation in the three dimensions of space at the nanoscale using core-shell nanomaterials with a previous band energy selec-
level is notable. Change in composition and structure is always tion [2325]. High mobility eld-effect transistors can also be
related to variations in the electronic structure. In this way, a designed by selecting the appropriate heterostructures allowing
3D variation in composition and crystal phase at the nanoscale higher gain and carrier mobilities [15,26]. Among all the nanos-
leads necessarily to the formation of low dimensional conne- tructures used as templates for bandgap engineering, semicon-
ment potentials within the nanowire. This results in the possibi- ductor nanowires are one of the most promising. Because of their
lity of designing new advanced applications in the area of one-dimensional (1D) morphology, with smaller cross-section
electronics [6,7] optoelectronic devices [810], energy harvesting compared to their length, they can be easily integrated in circuits
[1114,102], energy transport [15], light emission [16,17] or light and devices, allowing higher performance of operation [27,28]. In
addition, when positioned vertically on a substrate, they show a
*Corresponding authors:. Arbiol, J. (arbiol@icrea.cat), more efcient coupling with light when compared to planar
Morral, A.F.i. (anna.fontcuberta-morral@ep.ch) structures, meaning that they can become better emitters or

1369-7021/06/$ - see front matter 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mattod.2013.06.006 213
RESEARCH Materials Today  Volume 16, Number 6  June 2013

absorbers, when used as light emitting diodes (LEDs) or solar cells, the dimensionality of the structure, we may be able to conne the
respectively [14,2931]. density of states (DOS) in one to three dimensions, resulting,
In this manuscript, we review different approaches used respectively, in the formation of two to zero dimensional struc-
recently to modulate the composition at the nanoscale in a tures. In Fig. 1ac we show schematically how these low dimen-
three-dimensional manner within the nanowire, resulting in sional structures can be arranged within a nanowire. 2D structures
the integration of self-assembled quantum structures in a self- can be simply formed by varying the composition within the shell
assembled nanowire template. To give a complete overview of the of the nanowire; while 1D and 0D structures might be formed by
different quantum structures attending to their dimensionality we further composition variations within the shell. One should also
will start with two-dimensional (2D) structures such as quantum note that 0D structures can also be formed by varying the com-
wells; we will summarize the latest efforts to integrate 1D quantum position and structure along the nanowire [3234] if the diameter
RESEARCH: Review

structures such as quantum wires in a nanowire and nally we will is small enough, which we will not consider here for reasons of
show a couple of efcient ways to self-assemble quantum dots (0D space. High angle annular dark eld (HAADF) scanning transmis-
structures) in a nanowire template. In all the cases, we will focus on sion electron microscopy (STEM) and high resolution transmission
arsenide and nitride-based heterostructures synthesized by means electron microscopy (HRTEM) images of the 2D, 1D and 0D
of molecular beam epitaxy, which allows a high control of the structures reviewed here are shown in Fig. 1df. They correspond,
material deposition even at the monolayer scale. For a good respectively, to GaAs quantum wells embedded in an AlAs barrier,
understanding of the synthesized heterostructures, atomic resolu- GaN quantum wires on an AlN/GaN nanowire template and InAs
tion imaging is necessary to obtain their 3D atomic structure, quantum dots on a GaAs nanowire template. They will be
which can be then simulated and directly correlated to the optical described in detail below. For completeness, we show in Fig. 1g
properties. In particular we will show how advanced electron i how the density of states varies when we conne carriers in low
microscopy tools are the key for achieving an accurate under- dimensional structures. In the 2D case, the carriers are free to move
standing of such advanced structures. in two directions and are conned in the third one. As a conse-
quence, the energy of free carriers splits into quantized electronic
Strategies to achieve quantum connement in sub-bands, the quantized levels directly related to the connement
nanowires in one of the directions. This leads to the transformation of the
As shown in Fig. 1, several strategies can be applied to obtain density of states (DOS) from a parabolic (3D case) to a ladder shape
[(Figure_1)TD$IG]
quantum connement within a nanowire template. Depending on dependence, each step of the ladder coinciding with the conned

FIGURE 1
Sketches of different quantum structures classied by their dimensionality: (a) quantum wells or 2D structures (QWs); (b) quantum wires or 1D structures
(QWRs); (c) quantum dots or 0D structures (QDs). (df) Experimental STEM and HRTEM images of the quantum structures, QWs, QWRs and QDs, respectively.
(d) Reproduced by permission of The Royal Society of Chemistry (RSC) on behalf of the Centre National de la Recherche Scientique (CNRS) and the RSC
[64]. (f) Adapted with permission from Ref. [85]. Copyright (2010) American Chemical Society. (gi) Electronic density of states and corresponding energy
levels for the QWs, QWRs and QDs, respectively.

214
Materials Today  Volume 16, Number 6  June 2013 RESEARCH

states. When we proceed to the further connement in an addi- increase the energy of the emitted light. This accurate control is an
tional dimension, carriers can only move in one of the three essential property for obtaining precisely engineered light emitters,
directions of space. This leads to the formation of additional being able to modulate the light emission energy up to 3.98 eV [42].
p
parabolic sub-bands. The DOS follows then a 1= E law, the peak Another additional improvement is that due to the quantum con-
coinciding with the energy of the quantized states. Finally, the nement and the condensation of the DOS in the conned levels,
complete connement of carriers in three directions leads to the we obtain a better monochromaticity in the emitted light, with
formation of energy levels and a DOS similar to that of atoms. In lower energy dispersion and narrower energy peaks [42]. Interest-
conclusion, as a consequence of the signicant modication of ingly, thanks to the lamentary nature of the nanowire, the band-
energy levels and DOS, low dimensional structures within nano- gap modulation can be also easily modied by embedding it in the
wires offer many perspectives in terms of engineering the electro- appropriate strained environment [43]. It has been recently shown

RESEARCH: Review
nic properties of materials. that it is possible to change the strain conditions in the surroundings
of the QW by modifying the Al content in the AlxGa1xN barriers. In
Quantum connement for the formation of 2D this way, we can induce strain to the QW structure, thus obtaining a
structures slight modication of its cell parameters and thus obtaining a
When the diameter of the nanowire is so large that no quantum change in the emitted wavelengths [42].
connement effects arise in the radial direction, a relatively straight- Piling up axial homostructures consisting of the very same
forward manner to create a 2D quantum structure consists in the chemical composition but with different crystal structures
creation of heterostructures by changing the chemical composi- [34,33,44] has also been proved to be an approach to engineer
tion/structure during the growth in the axial direction [3541]. the nal system bandgap. Semiconductor binary systems in the
Fig. 2 shows an example. Here a GaN nanowire is rst grown as a form of NWs allow their crystallization in several polytypic crystal
base/stem. By varying the ux of growth precursors it is possible to phases, the most common being the hexagonal wurtzite (WZ) and
modulate the chemical composition while the growth continues. In the cubic zinc-blende (ZB) [4547]. Existence of both polytypes
this way, wide bandgap AlxGa1xN barriers could be successfully basically depends on the nal atomic plane stacking along the ZB
grown on top of GaN NW templates and be used as axial conne- [1 1 1] or the WZ [0 0 0 1] axes [48]. Switching from one structure
ment of narrower GaN thin layers acting as quantum wells because to the other can be achieved by creating ordered twin boundaries
of the 1D quantum connement of the carriers allowed by their 2D which will imply a change in the nal structural stacking of the
aspect ratio [42]. These axial QW systems allow a wide range of axial planes detailed before. In this way, many methods have been
possibilities for electronic-property engineering, as the dimensions developed to create ordered twinning arrays or superlattices [49
of the sections and thereby the quantum connement can be 51] to achieve the desired stacking. The quantum connement in
precisely controlled by the growth conditions. It is possible to this latter case is possible because of the small size (axial height) of
inuence the nal conned states by selecting the appropriate the insertions and because of the dependence of the band structure
thickness and composition of the QWs or barriers, for example, with respect to the crystalline phase [5256].
when decreasing the axial dimension (thickness) we automatically Until now, we have seen that 2D quantum structures can be
increase the energy of corresponding quantized states, and thus obtained by axial variation of the composition/structure; however,
[(Figure_2)TD$IG]

FIGURE 2
Axial GaN QWs in a NW. (a) ADF image of axial GaN QWs in a GaN/AlN NW template, showing the GaN regions (core and wells) with brighter contrast.
(b) Sketch of the NW with axial QWs: red arrows are pointing the QWs and green arrows the barriers between them. (c) Simulation demonstrating that the QWs
are under a strain gradient, being more compressive for the wells closer to the base as a result of the thickness variation of the AlN covering shell (being thicker
close to the base). (d) HRTEM image showing the GaN axial QWs in darker contrast. The red squared detail is displayed in (e) with atomic resolution. (f) Strain
variation dependence of the GaN QW versus the Al content in the AlxGa1xN barrier. Copyright 2011 by the American Physical Society [42].

215
RESEARCH Materials Today  Volume 16, Number 6  June 2013

the possibility to form radial heterostructures also exists [57]. This bottom facets of the inclined NWs (Fig. 3). Growth of the radial
constitutes again a rich playground for materials engineering. QWs in a molecular beam epitaxy system allowed obtaining
First, homogeneous or inhomogeneous shell thicknesses around perfectly sharp and defect-free interfaces, controlling the material
the different six NW facets can be achieved. In the case of most of deposition to the monolayer. With such promising quantum
the semiconductor binary systems (e.g. GaAs, GaSb, GaN, ZnTe) all systems, there are plenty of possibilities to perform accurate
the NWs tend to grow in only one polarity. In particular the bandgap engineering [6367]. Depending on the growth condi-
(1 1 1)B is the most commonly observed growth direction (e.g. tions, it is possible to modify either the QW thickness or the barrier
in GaAs that corresponding to an As-terminated top facet) [5861]. compositions, similarly to the case of axial QWs, thus obtaining
The nanowires present {1 1 0} or {1 1 2} lateral facets and a hex- the desired quantum connement conditions, obtaining, for
agonal prismatic cross section or they even present sections with example, 1.55 eV PL emission at 4.2 K with just 5 meV dispersion
RESEARCH: Review

various families of facets coexisting (typically in antimonides) (full width at half maximum, FWHM). Controlling the QW thick-
[62]. Taking into account this polarity-driven growth mechanism, ness on the same heterostructures, wavelength-controlled lasers
it was demonstrated that depending on the GaAs substrate orien- [68] and other electroluminescent devices [69] such as photode-
tation, the NWs growing on top were obtained at different angles tectors [70] could be designed. Moreover, coupled plasmon-pho-
(with respect to the substrate), always keeping a perfect epitaxial non modes have been observed in radial quantum well
relationship. Because of this, it was shown that perfect vertically heterostructures, opening the possibility for such heterostructures
aligned NWs could be obtained on top of (1 1 1)B substrates, while to be used for the connement of high-density of charge in
358 inclined NWs could be obtained on top of (0 0 1) oriented nanowire systems [71].
substrates (Fig. 3) [63,64]. Taking advantage of this latter property,
the growth conditions could be selected to obtain radial hetero- One-dimensional connement: 1D quantum wires
structures with a homogeneous thickness on vertically aligned Moving one step further from the quantum connement point of
wires (those grown on top of (1 1 1)B substrates), while otherwise, view, new quantum structures have been obtained by decreasing
the thickness of the shells on top of the different six facets could be their dimensionality in one more dimension thus creating the so-
inhomogeneous because of the shadowing effects occurring on the called quantum wires (QWRs) [7274]. Most of the published
[(Figure_3)TD$IG]

FIGURE 3
Radial QWs in-a-NW. (a,b) Homogeneous multi quantum well heterostructure (GaAs/AlAs) promoted by the growth on a (1 1 1)B GaAs substrate that allows
the vertical growth of the NWs. (c,d) (0 0 1) GaAs substrates force the tilted growth of the wires, preventing their homogeneous coverage. The resulting
QWs did not homogeneously cover the different facets of the wire.

216
Materials Today  Volume 16, Number 6  June 2013 RESEARCH

results dealing on QWRs present them embedded in bulk materials structures could be successfully self-assembled to the six edges of
or layers [7582], and only few have integrated them in NW GaN/AlN core-shell NW template systems.
systems, although only as core elements in core/shell NWs [83]. In Fig. 1b,e,h, we show the formation of self-assembled QWRs
Just recently, GaN QWRs could be grown by second order self- on a core-shell NW template. As it is well known, GaN NWs tend to
assembly on the edges of self-assembled GaN/AlN NW templates present a hexagonal cross-section presenting six {1 1 0 0} facets
[84]. In this latter case the quantum structures extend along the terminated into sharp edges placed along the [1120] direction
NW length with a high aspect ratio (Fig. 1b,e), being 1D quantum [42]. However, when growing an AlN shell on top of the GaN core,
structures and thus presenting quantum connement in 2D the AlN {1 1 2 0} edges are not sharp anymore presenting trun-
(Fig. 1h). As it was presented recently, this kind of 1D quantum cated surfaces (equivalent to {1 1 2 0} planes)[42]. In a third step,
[(Figure_4)TD$IG]

RESEARCH: Review

FIGURE 4
Top section: InAs Stranski-Krastanov QDs on the external facets of GaAs/AlAs NW templates. (a) Sketch of the followed synthesis process. (b) Atomic model
of the QD. (c) HAADF image of one of the dots (the scale bar represent 100 nm). The inset shows the EELS prole crossing the NW diameter including the
embedded QD. Adapted with permission from Ref. [85]. Copyright (2010) American Chemical Society. Bottom section: GaAs QDs developed in a multishell
GaAs/AlGaAs NW. (d) Z-contrast image of the cross-section of one wire. The squared corner is shown in (e) false-colored to enhance the contrast
differences. (f) Atomic resolution ADF image, also false-colored, of the QD and surrounding area. Reprinted with permission from Ref. [97].
Copyright 2013.

217
RESEARCH Materials Today  Volume 16, Number 6  June 2013

GaN can be deposited on top of the AlN shell. Nevertheless, GaN The StranskiKrastanov method of QD formation in NWs is one
under N-rich growth conditions does not nucleate on top of the of the most used for the design of QD structures in planar samples
AlN {1 1 0 0}-facets, but nucleation is observed on top of the and more rarely used in NW systems. A completely new method
{1 1 2 0} facets, which in fact are the ones located at the truncated consisting of self-induced diffusion processes has been recently
edges [84]. With this third step, the edges are lled with GaN until developed for the creation of QDs with superior optical perfor-
they become sharp again; leaving a nucleated GaN QWR structure mance (Fig. 4df) [97]. The growth of AlxGa1 xAs radial shell layers
extended along the six NW edges, with a characteristic triangular covering GaAs NW cores presents an inhomogeneous composition
cross-section. The rst synthesized QWRs presented bright cath- because of the different diffusion lengths of the Ga and Al adatoms
oluminescence (CL) spectra even at room temperature. Analyses on the nanowire facets.
performed by means of PL and CL on individual NWs showed that As a result we observe the formation of Al-rich stripes along the
RESEARCH: Review

the emission obtained from the QWRs presented a clear blue-shift, six edges of the AlxGa1 xAs radial shell (Fig. 4d). Similar observa-
with energies up to 4.4 eV, far away from the emission attributed tions had been reported in similar systems grown by MOCVD [98
to the GaN NW base (3.4 eV). In addition the width of a single 100,67,103]. However, even more interesting is that these Al-rich
QWR emission is below 10 meV, displaying an excellent quantum stripes can suffer inhomogeneities which develop by creating Y-
connement. In the rst studies performed, the diameter of the shape bifurcations that result in self-assembled conned Ga-rich
QWRs differed from each other, even in the same NW template, as QDs (Fig. 4ef)[97]. These QDs can be created at will by controlling
there was just a rough control on the nal AlN edge truncation the growth conditions of the shell layer, and appear regularly
during the AlN shell growth process [84]. Further work has to be along the entire length of the NW. The optical properties of this
carried out to improve the template design and control; however, new type of QDs are outstanding, presenting a modulated bandgap
this newly developed process of QWR self-assembling on NW (clear blue-shift with respect to the GaAs core), a narrow line-
templates opens a new range of possibilities for bandgap engineer- width, and even extremely bright single photon emission. A
ing and the formation of one-dimensional quantum structures. discontinuous emission line along the NW at higher energy (in
the range between 630 and 700 nm) was found, corresponding to
Toward the limits of quantum connement: quantum the presence of QDs, acting as extremely localized emitters. These
dots QDs emit within the red region, with a bright signal, and have
When lowering the dimensionality of an object in three dimen- been presented as potential emitters for ber optics communica-
sions at the nanoscale, we obtain a zero-dimensional structure or a tions, including their use in quantum computing. In addition,
dot, thus obtaining optical properties similar to those of the atoms these new QD-in-a-NW systems could be easily integrated to the
but with the emission lines engineered by the composition and CMOS technology as they were grown on top of Si substrates [101].
size of the QDs (Fig. 1j). The QDs are in fact tiny structures that can
easily be coupled to a NW [8587]. For many years, much work has Conclusions
been carried out to embed QD arrays inside NWs by piling different The development of quantum structures in a nanowire system has
heterostructures in nanowires showing quantum connement in been shown to be a promising methodology to achieve accurate
the radial direction, and obtaining localized QDs. These QDs have and precise bandgap modulations and thus extend the concept of
been sometimes ordered in arrays in the inner core of the NW. bandgap engineering to more complex and advanced devices. The
Some other growth methodologies allow the spontaneous growth NW systems can be easily used as building blocks for such systems
of such 0D quantum structures on the facets or even on the edges and furthermore can easily be integrated into the leading CMOS
of NWs. These later QDs have been obtained in some cases by technology. Beneting from all the possibilities of current growth
taking advantage of the different growth peculiarities of the technologies, these structures provide a promising platform for the
material heterostructures used. Recently, it was demonstrated that future of nanotechnology in optics and electronics.
using a GaAs NW acting as the core, and covering it with an AlAs
shell, it was possible to attach InAs QDs on top of the shell that Acknowledgements
developed on the NW facets and edges following the Stranski JA acknowledges the funding from the Spanish MICINN project
Krastanov growth mode. The intermediate AlAs layer was neces- MAT2010-15138 (COPEON) and Generalitat de Catalunya (2009
sary to allow the QDs growth, as the InAs may wet the GaAs layer SGR 770 and NanoAraCat). MdlM thanks CSIC JAE Pre-Doc
but just forming planar shells and not developing dots. This special program. ME acknowledges nancial support from the European
method for self-assembly was possible as the InAs can be easily Commission (Contract No. 224818 DOTSENSE) and the Deutsche
grown forming dots on top of the {1 1 0} AlAs facets [8891]. Here, Forschungsgemeinschaft DFG (NAWACS, EI 518/2). AFiM
as previously, atomic resolution transmission electron microscopy acknowledges funding from ERC StG (UpCon) and SNF through
was necessary to determine the morphology of the QDs. To under- NCCR-QSIT and grant nr. 134506. JA, MdlM and AFiM also
stand the optical properties of the resulting QD-in-a-NW system, acknowledge funding from PRI-PIMERU-2011-1422 (InCoSiN).
accurate 3D atomic models were obtained from the experimental
microscopy images [9296]. The optical characterization was per- References
formed by means of PL measurements, resulting in emission peaks [1] S.R. Plissard, et al. Nano Lett. 12 (2012) 1794.
in the range of 857930 nm. In this case, the energy of the light [2] S. Assali, et al. Nano Lett. 13 (2013) 1559.
[3] I.R. Musin, M.A. Filler, Nano Lett. 12 (2012) 3363.
emitted by the QDs was higher than the expected because of the [4] N. Shin, M.A. Filler, Nano Lett. 12 (2012) 2865.
interdiffusion of InAs in the AlAs layer as demonstrated by EELS [5] R.E. Algra, et al. Nano Lett. 11 (2011) 1690.
and HRTEM experiments [85]. [6] Y. Cui, C.M. Lieber, Science 29 (1) (2001) 851.

218
Materials Today  Volume 16, Number 6  June 2013 RESEARCH

[7] X.F. Duan, et al. Nature 409 (2001) 66. [56] G. Jacopin, et al. J. Appl. Phys. 110 (2011) 064313.
[8] F. Qian, et al. Appl. Phys. Lett. 87 (2005) 173111. [57] F. Qian, et al. Nano Lett. 4 (2004) 1975.
[9] F. Qian, et al. Nano Lett. 5 (2005) 2287. [58] A.F.I. Morral, et al. Appl. Phys. Lett. 92 (2008) 063112.
[10] H.P.T. Nguyen, et al. Nano Lett. 11 (2011) 1919. [59] C. Colombo, et al. Phys. Rev. B 77 (2008) 155326.
[11] B.M. Kayes, et al. Appl. Phys. 97 (2005) 114302. [60] E. Uccelli, et al. Nano Lett. 11 (2011) 3827.
[12] B.Z. Tian, et al. Nature 449 (2007) 885. [61] M. de la Mata, et al. Nano Lett. 12 (2012) 2579.
[13] C. Colombo, et al. Appl. Phys. Lett. 94 (2009) 173108. [62] T. Xu, et al. Nanotechnology 23 (2012) 095702.
[14] P. Krogstrup, et al. Nat. Photon. 7 (2013) 306. [63] A.F.I. Morral, et al. Small 4 (2008) 899.
[15] Y. Cui, et al. Nano Lett. 3 (2003) 149. [64] M. Heigoldt, et al. J. Mater. Chem. 19 (2009) 840.
[16] R. Songmuang, et al. Nano Lett. 10 (2010) 3545. [65] G. Ferrari, et al. Nano Lett. 9 (2009) 1631.
[17] R. Armitage, K. Tsubaki, Nanotechnology 21 (2010) 195202. [66] A. Bertoni, et al. Phys. Rev. B 84 (2011) 205323.
[18] M.T. Borgstrom, et al. Nano Lett. 5 (2005) 1439. [67] M. Fickenscher, et al. Nano Lett. 13 (2013) 1016.

RESEARCH: Review
[19] G. Bulgarini, et al. Appl. Phys. Lett. 100 (2012) 121106. [68] F. Qian, et al. Nat. Mater. 7 (2008) 701.
[20] M.E. Reimer, et al. Nat. Commun. 3 (737) (2012) 1746. [69] R. Koester, et al. Nano Lett. 11 (2011) 4839.
[21] M. Heurlin, et al. Nano Lett. 11 (2011) 2028. [70] A. de luna Bugallo, et al. Appl. Phys. Lett. 98 (2011) 233107.
[22] A. Kandala, et al. Phys. Status Solidi A 206 (2009) 173. [71] B. Ketterer, et al. Phys. Rev. B 83 (2011) 245327.
[23] M. Ibanez, et al. ACS Nano 7 (2013) 2573. [72] B. Gudden, R. Pohl, Z. Phys. 17 (1923) 331.
[24] A.I. Hochbaum, et al. Nature 451 (2008) 163. [73] E.V. Kovarskii Fizika, Tverdogo Tela 19 (1977) 905.
[25] A.I. Boukai, et al. Nature 451 (2008) 168. [74] V.K. Arora, Phys. Rev. B 23 (1981) 5611.
[26] K. Tomioka, et al. Nature 488 (2012) 189. [75] S. Koshiba, et al. Appl. Phys. Lett. 64 (1994) 363.
[27] J.C. Johnson, et al. Phys. Chem. B 107 (2003) 8816. [76] H. Fujikura, H. Hasegawa, J. Electron. Mater. 25 (1996) 619.
[28] M.A. Zimmler, et al. Nano Lett. 8 (2008) 1695. [77] H. Weman, et al. J. Mater. Sci. Mater. Electron. 20 (2009) 94.
[29] T. M.Bjork, et al. Nano Lett. 2 (2002) 87. [78] P. Kroger, et al. Appl. Phys. Lett. 100 (2012) 263114.
[30] A. Pfund, et al. Appl. Phys. Lett. 89 (2006) 252106. [79] P.M. Petroff, et al. Appl. Phys. Lett. 41 (1982) 635.
[31] E.D. Minot, et al. Nano Lett. 7 (2007) 367. [80] W. Wegscheider, et al. Phys. Rev. Lett. 71 (1993) 4071.
[32] L. Samuelson, et al. Physica E 35 (2004) 313. [81] B.J. van Wees, et al. Phys. Rev. Lett. 60 (1988) 848.
[33] N. Akopian, et al. Nano Lett. 10 (2010) 1198. [82] L.T. Canham, Appl. Phys. Lett. 57 (1990) 1046.
[34] D. Spirkoska, et al. Phys. Rev. B 80 (2009) 245325. [83] H.-J. Choi, et al. Phys. Chem. B 107 (2003) 8721.
[35] M. Wolz, et al. Appl. Phys. Lett. 98 (2011) 261907. [84] J. Arbiol, et al. Nanoscale 4 (2012) 7517.
[36] J. Lahnemann, et al. Phys. Rev. B 84 (2011) 155303. [85] E. Uccelli, et al. ACS Nano 4 (2010) 5985.
[37] J. Renard, et al. Phys. Rev. B 80 (2009) 121305(R). [86] X. Yan, et al. Nano Lett. 12 (2012) 1851.
[38] L. Rigutti, et al. Phys. Rev. B 82 (2010) 235308. [87] M. Heiss, et al. Nanotechnology 22 (2011) 195601.
[39] L.F. Zagonel, et al. Nano Lett. 11 (2011) 568. [88] D. Wasserman, et al. Appl. Phys. Lett. 83 (2003) 5050.
[40] M. Hei, et al. Nanotechnology 20 (2009) 075603. [89] E. Uccelli, et al. Nanotechnology 19 (2008) 045303.
[41] A.M. Graham, et al. Phys. Rev. B 87 (2013) 125304. [90] E. Uccelli, et al. Superlattices Microstruct. 44 (2008) 425.
[42] F. Furtmayr, et al. Phys. Rev. B 84 (2011) 205303. [91] X.B. Niu, et al. Appl. Phys. Lett. 95 (2009) 023119.
[43] B. Bouwes, et al. Nano Lett. 12 (2012) 6206. [92] S. Bernal, et al. Ultramicroscopy 72 (1998) 135.
[44] K. Pemasiri, et al. Nano Lett. 9 (2009) 648. [93] M. Ibanez, et al. Cryst. Growth Des. 12 (2012) 1085.
[45] A.F.I. Morral, et al. Adv. Mater. 19 (2007) 1347. [94] See other 3D atomic models of complex nanostructures at: http://www.icmab.
[46] J. Arbiol, et al. Nanotechnology 18 (2007) 305606. cat/gaen/research/
[47] A. Mishra, et al. Appl. Phys. Lett. 91 (2007) 263104. [95] J. Arbiol, et al. Appl. Phys. Lett. 80 (2002) 329.
[48] J. Arbiol, et al. J. Appl. Phys. 104 (2008) 064312. [96] M.I.B. Utama, et al. Adv. Funct. Mater. 23 (2013) 1636.
[49] J. Bao, et al. Nano Lett. 8 (2008) 836. [97] M. Heiss, et al. Nat. Mater. 12 (2013) 439.
[50] P. Caroff, et al. Nat. Nanotechnol. 4 (2009) 50. [98] M. Keplinger, et al. Nano Lett. 9 (2009) 1877.
[51] R.E. Algra, et al. Nature 456 (2008) 369. [99] G. Biasol, et al. Phys. Rev. Lett. 81 (1998) 2962.
[52] I. Zardo, et al. Phys. Rev. B 80 (2009) 245324. [100] N. Skold, et al. Nano Lett. 6 (2006) 2743.
[53] M. Heiss, et al. Phys. Rev. B 83 (2011) 045303. [101] E. Russo-Averchi, et al. Nanoscale 4 (2012) 1486.
[54] B. Ketterer, et al. ACS Nano 5 (2011) 7585. [102] J. Wallentin, et al. Science 339 (2013) 1057.
[55] J. Arbiol, et al. Nanotechnology 20 (2009) 145704. [103] C. Zeng, et al. Nano Lett. (2013), http://dx.doi.org/10.1021/nl401680k.

219

You might also like