The influence of post-growth annealing on the optical properties of InAs quantum dot chains grown on pre-patterned GaAs(100)

ArticleinNanotechnology 23(11):115702 · March 2012with14 Reads
Impact Factor: 3.82 · DOI: 10.1088/0957-4484/23/11/115702 · Source: PubMed
Abstract

We report on the effect of post-growth thermal annealing of [011]- ,[011(-)]-, and [010]-oriented quantum dot chains grown by molecular beam epitaxy on GaAs(100) substrates patterned by UV-nanoimprint lithography. We show that the quantum dot chains experience a blueshift of the photoluminescence energy, spectral narrowing, and a reduction of the intersubband energy separation during annealing. The photoluminescence blueshift is more rapid for the quantum dot chains than for self-assembled quantum dots that were used as a reference. Furthermore, we studied polarization resolved photoluminescence and observed that annealing reduces the intrinsic optical anisotropy of the quantum dot chains and the self-assembled quantum dots.

Full-text

Available from: Teemu Hakkarainen, Dec 17, 2015
IOP PUBLISHING NANOTECHNOLOGY
Nanotechnology 23 (2012) 115702 (6pp) doi:10.1088/0957-4484/23/11/115702
The influence of post-growth annealing on
the optical properties of InAs quantum
dot chains grown on pre-patterned
GaAs(100)
T V Hakkarainen, V Poloj
¨
arvi, A Schramm, J Tommila and M Guina
Optoelectronics Research Centre, Tampere University of Technology, PO Box 692, FIN-33101 Tampere,
Finland
E-mail: teemu.hakkarainen@tut.fi
Received 8 December 2011
Published 28 February 2012
Online at stacks.iop.org/Nano/23/115702
Abstract
We report on the effect of post-growth thermal annealing of [011]-, [01
¯
1]-, and [010]-oriented
quantum dot chains grown by molecular beam epitaxy on GaAs(100) substrates patterned by
UV-nanoimprint lithography. We show that the quantum dot chains experience a blueshift of
the photoluminescence energy, spectral narrowing, and a reduction of the intersubband energy
separation during annealing. The photoluminescence blueshift is more rapid for the quantum
dot chains than for self-assembled quantum dots that were used as a reference. Furthermore,
we studied polarization resolved photoluminescence and observed that annealing reduces the
intrinsic optical anisotropy of the quantum dot chains and the self-assembled quantum dots.
(Some figures may appear in colour only in the online journal)
1. Introduction
Thermal annealing has been widely used for improving
material quality and for modifying the properties of semi-
conductor nanostructures, such as quantum wells (QW) [1]
and quantum dots (QDs) [2]. Several studies have been
reported on thermal annealing of self-assembled InAs/GaAs
quantum dots (SAQDs) obtained by the Stranski–Krastanov
(SK) growth mode [39]. These experiments have revealed
that group III intermixing in the QD–matrix interface during
thermal annealing leads to blueshift and narrowing of the
optical emission as well as a reduction of the intersubband
energy separation. These properties have allowed us to exploit
thermal annealing, for example, for tuning the response of
QD photodetectors that utilize ensembles of self-assembled
QDs [10]. Furthermore, it has been demonstrated that
post-growth RTA can be used to reduce the fine structure
splitting in InAs SAQDs [1113].
More recent QD related applications, such as entangled-
photon emitters [14] or nanophotonic waveguides [15]
require a deterministic positioning of QDs with precisely
adjusted optical properties. Recently, we have shown that
formation of site-controlled InAs single dots and quantum
dot chains (QDCs) can be obtained by a combination
of UV-nanoimprint lithography (UV-NIL) and molecular
beam epitaxy (MBE) [1618]. A precise optimization of
growth conditions is necessary in the site-controlled epitaxy
of QDs in order to achieve QD nucleation only in the
pre-determined locations. This might lead to less freedom
in precise adjustment of other QD properties, such as the
emission wavelength, which is often required for specific
applications. Different growth procedures such as high
temperature capping [19] or indium flushing [20] have been
used for blueshifting the emission of site-controlled QDs.
This paper focuses on post-growth rapid thermal annealing
(RTA) which is a potential method not only for modifying
the spectral properties of site-controlled QDs but also for
reducing the defects caused by the lithography process and
subsequent growth on a non-planar surface. So far, only in
situ annealing experiments, showing a reduction of exciton
linewidths, have been reported on site-controlled QDs [21].
In particular, we study the influence of post-growth RTA
on the optical properties of [011]-, [01
¯
1]-, and [010]-oriented
QDCs grown by MBE on UV-NIL patterned GaAs(100)
10957-4484/12/115702+06$33.00
c
2012 IOP Publishing Ltd Printed in the UK & the USA
Page 1
Nanotechnology 23 (2012) 115702 T V Hakkarainen et al
[01-1]
[011]
(a)
(b)
(c)
(d)
Figure 1. AFM pictures of [011]-, [01
¯
1]-, and [010]-oriented QDCs and SAQDs in (a)–(d), respectively. The size of the pictures is
500 nm × 500 nm. The height scale is 20 nm in (a)–(c) and 12 nm in (d).
surfaces. We use room temperature, low temperature, and
polarization resolved photoluminescence (PL) experiments
for analyzing the effect of annealing on the optical properties
of the QDCs and the surrounding GaAs matrix.
2. Experiment
The investigated samples were prepared by a three-stage
procedure combining MBE and UV-NIL. First, a 100 nm
GaAs buffer, a 100 nm AlGaAs cladding layer, and a 100 nm
GaAs were deposited at 590
C on quarters of 2
00
n-GaAs(100)
substrates by MBE. Then, the samples were ex situ patterned
by UV-NIL. Four 10 mm × 10 mm groove patterns oriented
along the [011], [01
¯
1], [010], and [001] directions were
processed on each sample. The grooves were 90 nm wide,
30 nm deep, and had a period of 180 nm. In the final stage, the
patterned surface was covered with a 60 nm GaAs regrowth
buffer at 490
C before the deposition of 2.2 ML of InAs
for the QDC formation at 515
C. The QDCs were covered
with a 20 nm GaAs layer grown at 515
C. Subsequently, the
sample temperature was increased to 590
C for the growth
of a 50 nm GaAs, a 50 nm AlGaAs layer, and a 20 nm GaAs
capping layer. As a reference for the QDCs, we prepared also a
SAQD sample on an unprocessed n-GaAs(100) substrate. The
layer structure and growth conditions of the QDC and SAQD
samples were identical. Furthermore, a sample with uncapped
QDs was prepared for structural analysis of QDCs and SAQDs
by atomic force microscopy (AFM). The UV-NIL process and
chemical cleaning prior to regrowth are discussed in detail
in [17]. The samples with [011]-, [01
¯
1]-, and [010]-oriented
QDCs are from now on referred to as QDC[011], QDC[01
¯
1],
and QDC[010], respectively.
The RTA process was performed for 2 × 2 mm
2
pieces of QDC[011], QDC[01
¯
1], QDC[010], and SAQD
reference sample. The samples were proximity-capped [22]
and annealed on a silicon wafer, face up in N
2
atmosphere
at a temperature of 720
C. The temperature was controlled
by an optical pyrometer. The annealing was performed in
a sequence of 100 and 200 s steps. Room temperature PL
(RT-PL) was measured from as-grown samples and after each
annealing step using excitation at 532 nm and detection with
an InGaAs photodiode. Furthermore, PL emission from the
GaAs matrix and wetting layer was measured with a CCD
detector before and after the annealing procedure. In order
to further analyze the effect of RTA on the optical properties
of the QDCs and SAQDs, we also measured low temperature
PL (LT-PL) and polarization resolved PL (PR-PL) from each
sample piece before and after the annealing sequence. The
LT-PL and PR-PL measurement was performed at 30 K in a
closed-cycle cryostat using excitation at 488 nm and detection
with an InGaAs photodiode.
3. Results and discussion
The AFM pictures in figure 1 show the surface morphology
of the uncapped QDCs and SAQDs. The QDCs grown in the
2
Page 2
Nanotechnology 23 (2012) 115702 T V Hakkarainen et al
Figure 2. The evolution of RT-PL during post-growth RTA at
720
C. (a) and (b) represent [01
¯
1]-oriented QDCs and SAQDs
samples, respectively.
[011]-, [01
¯
1]-, and [010]-oriented grooves are illustrated in
figures 1(a)–(c) and the reference SAQDs in (d). The average
QD height and density in the QDC and SAQD samples are
12 nm and about 1.3 × 10
10
cm
2
. The structural and optical
properties of the QDCs are reported in [16] and [18].
Figure 2 shows RT-PL spectra measured from the
QDC [1] and SAQD reference samples before annealing and
after each annealing step. The RT-PL intensity of the as-grown
QDCs is 11–24% compared with the reference SAQDs [16].
The evolution of RT-PL during annealing is not shown for
QDC[011] and QDC[010] because the results are similar to
those of QDC[01
¯
1]. Both samples (figure 2) show blueshift
of the RT-PL emission with increasing annealing time. The
ground state (GS) peak intensity of QDC[01
¯
1] (figure 2(a))
increases during the first annealing steps. Further increase of
annealing time causes a constant decrease of the GS peak
intensity. As shown in figure 2(b), the SAQDs experience
a constant decrease of the peak intensity during the whole
annealing sequence. The evolution of integrated intensity
and GS peak energy E
GS
during the annealing sequence for
all samples are illustrated in figure 3. The QDC samples
and the SAQDs experience, as a general trend, a decay
of the integrated intensity with increasing annealing time
(figure 3(a)), but the decay is more rapid for the QDCs than
for the SAQDs. All QDC samples show more or less similar
sublinear blueshift of E
GS
with increasing annealing time (as
shown in figure 3(b)); E
GS
increases rapidly in the beginning
and starts to saturate towards the end of the annealing
sequence. The E
GS
of the QDCs after annealing at 720
C
Figure 3. The evolution of PL properties of [011]-, [01
¯
1]-, and
[010]-oriented QDCs and SAQDs during post-growth RTA at
720
C. (a) shows integrated PL intensity and (b) GS peak energy
(E
GS
) as a function of annealing time.
for 1400 s is approximately 1.19 eV, which corresponds to
a blueshift of 130 meV compared with the as-grown samples.
Contrary to the QDCs, the SAQDs experience a superlinear
increase of E
GS
with increasing annealing time, resulting in
a total blueshift of only 90 meV within 1400 s annealing.
After 1400 s annealing the blueshift of the SAQDs is far from
saturation, so we decided to perform one more 200 s annealing
step for them to see if the E
GS
of the SAQDs approaches that
of the QDCs. As shown in figure 3(c), the total blueshift for
the SAQDs reaches 120 meV after the additional annealing
step.
The different annealing behavior of the QDCs and
SAQDs can be understood in terms of defect mediated
diffusion: due to the lithography process and subsequent
MBE growth on a non-planar surface the QDC samples are
expected to possess a higher density of point defects than the
SAQD sample which causes a more a rapid intermixing at the
QD–matrix interface. As a result, the blueshift of E
GS
is more
rapid for the QDCs than for the SAQDs in the beginning of
the annealing sequence. Furthermore, the confinement of the
holes and electrons in the QDs is reduced with increasing E
GS
.
This in turn causes the reduction of the RT-PL intensity for
both QDCs and SAQDs.
In order to study the effect of annealing on the properties
of GaAs matrices in the QDC and SAQD samples, we
measured the RT-PL emission around the GaAs bandgap.
Figures 4(a) and (b) show the wetting layer (WL) and GaAs
RT-PL emission for the QDC[01
¯
1] and SAQD samples before
and after the annealing sequence. Corresponding spectra
for QDC[011] and QDC[010] are not shown because the
3
Page 3
Nanotechnology 23 (2012) 115702 T V Hakkarainen et al
Figure 4. RT-PL spectra showing WL and GaAs emission from the
QDC[01
¯
1] and SAQD samples in (a) and (b), respectively. The
spectra in (a) and (b) are normalized to the GaAs peak of the
as-grown SAQDs. The light and dark gray peaks illustrate PL
emission from n-GaAs substrate and MBE grown undoped GaAs,
respectively.
results are similar to those of QDC[01
¯
1]. Both figures 4(a)
and (b) show the same main features: the PL signal of the
highest excited QD states on the low energy side of the
spectra, the WL peak at 1.34 eV, and the GaAs peak at
1.43 eV. The intensities of all three contributions increase
during annealing. The increased RT-PL from the QDs above
1.25 eV can be attributed to the blueshift of the QD
spectrum. The increase in the WL intensity can be attributed
either to reduced carrier confinement of the QDs due to
QD–matrix intermixing and/or improvement of the matrix
quality, i.e. annealing of the defects in the bulk and at the
interfaces reducing non-radiative recombination. Therefore,
the influence of annealing on the GaAs matrix PL requires
careful consideration. As shown in figure 4(a), the GaAs
peak measured from the as-grown QDC[01
¯
1] sample is broad.
After annealing it is narrower, more intense, and slightly
redshifted. In order to explain these changes, we consider the
contributions from both the undoped GaAs matrix and the
underlying n-GaAs substrate. The light gray and dark gray
peaks in figure 4(a) show RT-PL emission measured from
a bare n-GaAs substrate and from an undoped MBE grown
GaAs layer. A comparison of the shapes of these peaks with
the GaAs peaks of the as-grown and annealed QDC[01
¯
1]
shows that most of the GaAs peak intensity measured from the
Figure 5. LT-PL spectra measured at 30 K before and after
annealing. (a)–(d) represent QDC[011], QDC[01
¯
1], QDC[010], and
SAQD samples, respectively.
as-grown QDC[0-11] originates from the n-doped substrate.
The annealed sample shows a narrowed and slightly redshifted
GaAs peak which suggests increased contribution from the
undoped GaAs matrix. A quantitative analysis was performed
by fitting a linear combination of the light gray and dark
gray curves to the spectra measured from the as-grown and
annealed QDC[01
¯
1] samples. Approximately 23% of the
GaAs peak intensity measured from the as-grown QDC[01
¯
1]
originates from the undoped GaAs matrix and 77% from the
n-GaAs substrate. The corresponding figures for the annealed
QDC[01
¯
1] are 50% and 50%. This yields a 3.5-fold increase
of the GaAs matrix contribution upon annealing, indicating a
reduction of non-radiative defects in the matrix. The GaAs
peaks of both as-grown and annealed SAQD samples are
dominated by the emission from the undoped GaAs matrix,
and therefore annealing has no clear influence on the peak
shape or position. A comparison of peak intensities shows the
SAQDs experience a 2.5-fold increase of the GaAs matrix PL
in annealing. Compared to the SAQD reference sample, the
4
Page 4
Nanotechnology 23 (2012) 115702 T V Hakkarainen et al
QDC samples experience a larger increase of GaAs matrix PL
upon annealing, indicating a more pronounced reduction of
non-radiative defects in the matrix.
The well-known effects of post-growth RTA on the
optical properties of InAs SAQDs include blueshift, spectral
narrowing, reduction of intersubband energy separation, and
increase of low temperature PL intensity [5, 23]. The
room temperature PL spectra (figure 2) showed a clear
blueshift for both QDC samples and for the SAQD reference
sample. In order to assess the other effects of annealing
on the optical properties of QDCs we also measured low
temperature PL spectra from as-grown and annealed QDC
and SAQD samples. The results are shown in figure 5.
The GS peak intensity of the as-grown QDC samples is
approximately 60% compared with the SAQD reference.
The annealed QDC[011], shown in figure 5(a), exhibits
slightly reduced LT-PL intensity, while the LT-PL intensity of
QDC[01
¯
1], QDC[010], and SAQDs increases upon annealing.
Furthermore, the LT-PL spectra in figures 5(a)–(d) show that
annealing causes blueshift, spectral narrowing, and reduction
of intersubband energy separation in all samples investigated.
The low LT-PL intensity of the annealed QDC[011] could
be due to a reduced number of the optically active QDs
after the annealing process. This is supported by the fact that
the spectrum of the annealed QDC[011] is more saturated
than the spectra of annealed QDC[01
¯
1], QDC[010], or
SAQD, although all investigated samples have similar QD
densities [13]. Excluding the reduction of LT-PL intensity in
the QDC[011], the effect of post-growth RTA on the LT-PL
emission from the QDC samples is similar to what has been
reported for SAQDs [5, 23].
In order to investigate the effect of annealing on the
optical anisotropy we performed polarization resolved PL
measurement for the as-grown and annealed QDCs and
SAQDs. The in-plane polarization anisotropy of GS PL
emission can be described by an interplay of intrinsic
and extrinsic components [16]. The intrinsic polarization
anisotropy is caused by the typical slight QD shape elongation
in the [01
¯
1] direction [24] and piezoelectric asymmetry [25]
between the [011] and [01
¯
1] directions. The extrinsic
anisotropy, which tends to align along the QDC axis, can be
caused by lateral interdot coupling [26] and/or an anisotropic
potential environment, for example, due to the existence of
a one-dimensional wetting layer [27]. Figure 6 shows the
GS peak optical anisotropy for the as-grown and annealed
QDCs and SAQDs. The as-grown QDC[011] (figure 6(a))
exhibits optical anisotropy that is oriented along the QDC
axis. The optical anisotropy of QDC[011] increases during
annealing. Compared with the as-grown QDC[011], the
optical anisotropy of the as-grown QDC[01
¯
1] (figure 6(b))
is larger because both intrinsic and extrinsic components
are oriented along the [01
¯
1] direction. However, the optical
anisotropy of QDC[01
¯
1] decreases upon annealing. As shown
in figure 6(c), the as-grown QDC[010] shows an optical
anisotropy aligned in between the [010] and [01
¯
1] directions.
This is caused by the interplay of intrinsic and extrinsic
components that are oriented along the [01
¯
1] and [010]
directions, respectively. The annealed QDC[010] shows a
slightly increased anisotropy that has rotated towards the
[010] directions. The SAQDs (figure 6(d)) show a slight
anisotropy along the [01
¯
1] direction which is reduced upon
annealing.
The effect of annealing on the optical anisotropy of the
QDCs and SAQDs can be understood in terms of a reduction
of the intrinsic anisotropy. In the case of QDC[011] the
reduction of the intrinsic anisotropy causes an increase in the
overall anisotropy because it is oriented perpendicular to the
QDC axis. For the QDC[01
¯
1] we observe an opposite effect
because the intrinsic anisotropy is oriented along the QDC
axis. In QDC[010], the contribution of the extrinsic anisotropy
along the [010] direction is emphasized as the intrinsic
anisotropy along the [01
¯
1] direction decreases, and therefore
we observe the rotation of the optical anisotropy towards the
[010] direction. However, the increase of the overall optical
anisotropy, in addition to the rotation of the polarization,
observed for QDC[010] indicates that the extrinsic component
has also increased during annealing. It should be noted that
annealing may have increased the extrinsic anisotropy in
QDC[011] and QDC[01
¯
1] as well, but it cannot be assessed
from figure 6 because the extrinsic and intrinsic components
are aligned parallel/perpendicular to each other. The reduction
of the intrinsic anisotropy in post-growth RTA can be
attributed to two effects: (i) a reduction of the elastic strain due
to the QD–matrix intermixing during post-growth RTA [28],
and (ii) a decrease of the relative QD size anisotropy due to
a uniform increase of the QD size in all directions [4], the
latter of which may also increase interdot coupling in QDCs
enhancing the extrinsic anisotropy.
4. Conclusions
We have shown that [011]-, [01
¯
1]-, and [010]-oriented QDCs
grown by MBE on UV-NIL patterned GaAs(100) experience
a blueshift of PL emission, spectral narrowing, a reduction
of intersubband energy separation, and an increase in LT-PL
intensity in post-growth RTA. These features were shown
to be similar in QDCs and SAQDs grown on unprocessed
substrates. However, the blueshift is more rapid for the QDCs
than for the SAQDs, indicating more pronounced QD–matrix
intermixing during the annealing process. This is attributed
to a higher density of defects in the GaAs matrix due to the
patterning process and subsequent growth on a non-planar
surface. Furthermore, we have shown that the RT-PL intensity
of the GaAs matrix in the QDC samples experiences a 3.5-fold
increase, compared to a 2.5-fold increase for the SAQD
reference sample during annealing which suggests a more
pronounced reduction of non-radiative defects in the QD
surroundings. PR-PL measurements revealed that the intrinsic
optical anisotropy of both QDCs and SAQDs decreases upon
annealing, which is attributed to a reduction of elastic strain
and QD size anisotropy.
Acknowledgments
The authors thank Mr Risto Ahorinta for the AFM
pictures and the financial support from the Academy of
5
Page 5
Nanotechnology 23 (2012) 115702 T V Hakkarainen et al
Figure 6. Ground state PL polarization measured at 30 K before and after RTA for QDC[011] (a), QDC[01
¯
1] (b), QDC[010] (c), and
SAQDs (d).
Finland projects DAUNTLESS (decision number 123951)
and DROPLET (decision number 138940). T V Hakkarainen
acknowledges financial support from the Finnish National
Graduate School in Materials Physics, J Tommila ac-
knowledges the National Graduate School in Nanoscience
(NGS-NANO) and the Pirkanmaa Regional Fund of the
Finnish Cultural Foundation for the financial support.
References
[1] Zory P S (ed) 1993 Quantum Well Lasers (New York:
Academic)
[2] Grundmann M (ed) 2002 Nano-Optoelectronics (Berlin:
Springer)
[3] Leon R, Kim Y, Jagadish C, Gal M, Zou J and
Cockayne D J H 1996 Appl. Phys. Lett. 69 1888
[4] Kosogov A O et al 1996 Appl. Phys. Lett. 69 3072
[5] Malik S, Roberts C, Murray C and Pate M 1997 Appl. Phys.
Lett. 71 1987
[6] Lobo C, Leon R, Fafard S and Piva P G 1998 Appl. Phys. Lett.
72 2850
[7] Xu S J, Wang X C, Chua S J, Wang C H, Fan W J, Jiang J and
Xie X G 1998 Appl. Phys. Lett. 72 3335
[8] Leon R, Fafard S, Piva P G, Ruvimov S and
Liliental-Weber Z 1998 Phys. Rev. B 58 R4262
[9] Fafard S, Wasilewski Z R, Allen C Ni, Picard D, Spanner M,
McCaffrey J P and Piva P G 1999 Phys. Rev. B 59 15368
[10] Aivaliotis P, Zibik E A, Wilson L R, Cockburn J W,
Hopkinson M and Airey R J 2007 Appl. Phys. Lett.
91 143502
[11] Ellis D J P, Stevenson R M, Young R J, Shields A J,
Atkinson P and Ritchie D A 2007 Appl. Phys. Lett.
90 011907
[12] Seguin R et al 2006 Appl. Phys. Lett. 89 263109
[13] Langbein W, Borri P, Woggon U, Stavarache V, Reuter D and
Wieck A D 2004 Phys. Rev. B 69 161301(R)
[14] Mohan A, Felici M, Gallo P, Dwir B, Rudra A, Faist J and
Kapon E 2010 Nature Photon. 4 302
[15] Wang C-J, Lin L Y and Parviz B A 2005 IEEE J. Sel. Top.
Quantum Electron. 11 500
[16] Hakkarainen T V, Tommila J, Schramm A, Tukiainen A,
Ahorinta R, Dumitrescu M and Guina M 2010 Appl. Phys.
Lett. 97 173107
[17] Tommila J, Tukiainen A, Viheri
¨
al
¨
a J, Schramm A,
Hakkarainen T, Aho A, Stenberg P, Dumitrescu M and
Guina M 2011 J. Cryst. Growth 323 183
[18] Hakkarainen T V, Tommila J, Schramm A, Tukiainen A,
Ahorinta R, Dumitrescu M and Guina M 2011
Nanotechnology 22 295604
[19] Yamaguchi K, Saito Y and Ohtsubo R 2002 Appl. Surf. Sci.
190 212
[20] Sasakura H, Kayamori S, Adachi S and Muto S 2007 J. Appl.
Phys. 102 013515
[21] Helfrich M et al 2011 J. Cryst. Growth 323 187
[22] Babinski A, Jasinski J, Bozek R, Szepielow A and
Baranowski J 2001 Appl. Phys. Lett. 79 2576
[23] Hsu T M, Lan Y S, Chang W H, Yeh N T and Chyi J I 2000
Appl. Phys. Lett. 76 691
[24] Songmuang R, Kiravittaya S and Schmidt O G 2003 J. Cryst.
Growth 249 416
[25] Bester G, Zunger A, Wu X and Vanderbilt D 2006 Phys. Rev.
B 74 081305
[26] Moon P, Yoon E, Sheng W and Leburton J-P 2009 Phys. Rev.
B 79 125325
[27] Wang X, Wang Zh M, Liang B, Salamo G J and
Shih C-K 2006 Nano Lett. 6 1847
[28] Santoprete R, Kratzer P, Scheffler M, Capaz R B and
Koiller B 2007 J. Appl. Phys. 102 023711
6
Page 6
  • [Show abstract] [Hide abstract] ABSTRACT: We present an experimental and theoretical analysis of the influence of a surface nanopattern on the properties of embedded InAs/GaAs quantum dots (QD). In particular, we analyze QDs grown on nanoimprint lithography (NIL) patterned grooves and investigate the influence of the non-planar surface morphology on the size, shape, strain distribution, and electronic structure of the embedded QDs. We show that the height reduction of InAs QDs during GaAs capping is significantly less pronounced for the QDs grown on the pattern than for the self-assembled QDs. Furthermore, the pattern has a strong impact on the strain and composition profile within the QD. The experimentally observed strain distribution was successfully reproduced with a three-dimensional model assuming an inverse-cone type composition gradient. Moreover, we show that the specific morphology of the QDs grown in the grooves gives rise to an increase of the vertically polarized photoluminescence emission which was explained by employing 8-band k.p calculations. Our findings emphasize that the surface curvature of the pattern not only determines the nucleation sites of the QDs but also has a strong impact on their morphological properties including shape, size, composition profile, and strain distribution. These properties are strongly cross-correlated and determine the electronic and optical characteristics of the QDs.
    Full-text · Article · Nov 2013 · Journal of Applied Physics
    0Comments 4Citations
  • [Show abstract] [Hide abstract] ABSTRACT: This paper reports on the evolution of InAs/GaAs quantum dots’ (QDs) intermixing process depending on the QD position with respect to the InGaAs strain reducing layer. The postgrowth intermixing has been ensured by rapid thermal annealing (RTA) at different temperatures and the processed samples were investigated by photoluminescence (PL) spectroscopy. The reference sample, containing InAs QDs in pure GaAs matrix, demonstrates the highest intermixing degree with an emission energy blueshift up to 300 meV. The reduction of the strain around the QDs for dot-in-well (DWELL) structure has been found to slightly reduces the intermixing degree leading to less emission energy blue shift. However, for the QDs with InGaAs underlying layer, a relative prohibition of the intermixing has been found to occur testifying the existing of thermally stable extended defects.
    Full-text · Article · Dec 2014 · Journal of Alloys and Compounds
    0Comments 4Citations