Highly Linearly Polarized Emission from Quantum Dash Excitons
Modelling and Experiment at the 3
rd
Telecom Window
Paweł Mrowiński
1
, Sven Höfling
2,3
, Johannes P. Reithmaier
4
and Grzegorz Sęk
1
1
LOSN, Faculty of Fundamental Problems of Technology, Departement of Experimental Physics,
Wrocław University of Science and Technology, Wybrzeze Wyspianskiego 27, Wrocław, Poland
2
Technische Physik & W. C. Röntgen-Center for Complex Material Systems, Universität Würzburg, Germany
3
SUPA, School of Physics and Astronomy, University of St. Andrews, North Haugh, KY16 9SS, St. Andrews, U.K.
4
Institute of Nanostructure Technologies and Analytics (INA), CINSaT, University of Kassel,
Heinrich-Plett-Str. 40, 34132 Kassel, Germany
Keywords: Photoluminescence, Spectroscopy, Exciton, InAs, Quantum Dash, Polarization Anisotropy, FDTD,
Telecommunication.
Abstract: This work is focused on controlling the polarization anisotropy of emission from single self-assembled
InAs/InGaAlAs quantum dashes grown by molecular beam epitaxy on InP substrate. We studied the degree
of linear polarization of excitonic emission for submicrometer mesa photonic structures of asymmetric in-
plane geometry. We present both experimental and numerical analysis performed at 1550 nm wavelength (3
rd
telecommunication window for optical fibers), and we discussed the impact of anisotropy of the dielectric
confinement, which paves the way towards a single photon source characterized by a degree of linear
polarization exceeding 0.9.
1 INTRODUCTION
Single nanostructures like quantum dots or dashes
embedded in semiconductor matrix have been found
as promising candidates for quantum emitters such as
single photon sources (Michler et al., 2000; Fiore et
al., 2007; Dusanowski et al., 2014) or entangled
photon pairs (Akopian et al., 2006; Stevenson et al.,
2006). By modifying their external environment, one
can obtain enhancement or inhibition of spontaneous
emission rate due to Purcell effect (Purcell, 1946).
Advanced microstructures allow also for controlling
the polarization state of the emitted light when
transition dipole moment of excitons couples with the
optical far field modes (Munsch et al., 2012; Foster et
al., 2015). By measuring the polarization anisotropy
of emission, it is possible to analyse the coupling
selectivity in terms of polarization, and by using a
simulation tool based on Finite-Difference Time-
Domain (FDTD) one can design a unique
microstructure that enables to outcouple photons with
a well-defined polarization state (Konishi et al., 2011;
Mrowiński et al., 2016).
Figure 1: a) SEM image of uncapped InAs quantum dash
sample b) chemically etched mesa of submicrometer in-
plane size and 300 nm height c) photoluminescence spectra
of both planar and processed samples showing isolated
spectral lines for mesas due to limited number of optically
active nanostructures.
238
Mrowi
´
nski, P., Höfling, S., Reithmaier, J. and S˛ek, G.
Highly Linearly Polarized Emission from Quantum Dash Excitons - Modelling and Experiment at the 3rd Telecom Window.
DOI: 10.5220/0006635602380241
In Proceedings of the 6th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2018), pages 238-241
ISBN: 978-989-758-286-8
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
In this work, we focus on quantum dashes
(QDashes) of the InAs/InGaAlAs material system
grown on InP substrate by molecular beam epitaxy.
The system consists of 200 nm barrier
In
0.53
Ga
0.23
Al
0.24
As layer (lattice matched to InP),
InAs 2-5 monolayers leading to nucleation of
QDashes, 100 nm the same barrier and 10 nm InP.
The QDashes are rather non-uniform nanostructures
of high surface density of ~10
10
cm
-1
and elongated
along [1-10] crystallographic direction reaching
length of more than 50 nm (Sauerwald et al., 2005;
Reithmaier, Eisenstein and Forchel, 2007) (Fig. 1a).
A spectral range of emission can be tuned from 1200
to 2000 nm by controlling the amount of nominally
deposited InAs material, thus enabling their use in the
active region of nanophotonic or optoelectronic
devices devoted to operate at telecommunication
windows of O-band (1310 50 nm) and C-band
(1550 15 nm).
By using chemically etched submicrometer
structures like asymmetric mesas shown in Fig 1b),
the number of investigated nanostructures is limited
to a few tens of optically active QDashes on 200 x
400 nm
2
area and 300 nm height. In Fig. 1 c) we
demonstrated results of photoluminescence
experiment performed on both ensemble and etched
mesa structure showing well isolated spectral features
for the mesa, which can be assigned to optical
transitions from excitonic complexes confined in the
QDash.
2 POLARIZATION ANISOTROPY
OF QUANTUM DASH IN
ASYMMETRIC MESA
In this section, we study the influence of mesa
structure geometry on the polarization anisotropy of
Figure 2: a) Polarization resolved emission of quantum dash
exciton for rectangular mesa of 400 x 200 nm
2
oriented
paralallel and b) perpendicular to the QDash axis showing
DOLP of 0.5 and -0.2, respectively. c) A size dependence
of DOLP measured for several excitons using fixed lateral
aspect ratio.
excitonic emission around 1550 nm spectral range.
All results can be quantified by a degree of linear
polarization (DOLP) defined as 
 

 
,
where
is emission intensity from QDash
polarized along [1-10] ([110]) in-plane directions.
First, we examined the reference sample, which is
200 x 200 nm
2
and we obtained a non-zero
polarization anisotropy exciton line, which is
typically in a range from 0.2 to 0.3 (Musiał,
Podemski, et al., 2012; Mrowiński et al., 2016). Such
noticeable anisotropy seen for QDashes results from
the anisotropic quantum confinement due to the in-
plane elongated geometry, which leads to a mixing in
the valence-band and thus induces a difference in the
oscillator strengths for the linearly polarized emission
from mixed exciton bright states (Musiał,
Kaczmarkiewicz, et al., 2012; Tonin et al., 2012;
Singh, Kumar and Singh, 2017). Polarization
anisotropy originated from the electronic structure is
then an intrinsic contribution. Next, we performed
similar measurements for asymmetric mesa
structures, namely rectangular one of 400 x 200 nm
2
size, oriented both along the QDash elongation axis
[1-10] and perpendicular to it. Such configurations
results in DOLP of about 0.55 and -0.20, respectively
(Fig. 2a,b). This effect has also been examined in
function of the mesa size. In Fig. 2c) we present the
measured DOLP of exciton emission for several
mesas of fixed lateral aspect ratio of 2:1 oriented
along QDash main axis and for increasing its base
dimension up to 900 x 450 nm
2
. It shows that the mesa
influence on the DOLP decreases with the increasing
size approaching 0.26 for the largest one, which is
close to the value obtained for reference and in-plane
symmetric mesa.
3 FDTD SIMULATIONS
We perform simple modelling using commercial-
grade simulator (‘Lumerical Solutions, Inc.) based on
FDTD to explain the experimental results concerning
the polarization anisotropy. A spontaneous emission
(SE) rate of exciton can be described in a dipole-
approximation as




 



,
where 

is local density of optical states,
is
transition dipole moment and


is electric field
amplitude for the far-field mode at the position of the
emitter. In the weak-coupling regime, the main
contribution to the anisotropy is expected for the
electric field, which is localized in the asymmetric
mesa structures depending on its polarization. The
Highly Linearly Polarized Emission from Quantum Dash Excitons - Modelling and Experiment at the 3rd Telecom Window
239
Figure 3: a) Simple model used for FDTD simulations of
the optical field confinement inside the mesa structure for
the far field modes. b) Simulation results of the polarized
electric field intensity distributions using two representative
projections for asymmetric mesa structure of 400 x 200 nm
2
and 300 nm height.
polarization degree of freedom is taken into account
in the scalar product where the linearly polarized
H(V) dipoles couples to the parallel component of the
localized electric field.
We calculated the electric field intensity
distribution inside a mesa structure using linearly
polarized pulsed excitation normal to the sample
surface, which is schematically shown in Fig. 3a). A
spectral range is centred at 1550 nm. By using
simplified geometry of InAs rectangular mesa on the
InP substrate, we performed calculations for 400 x
200 nm
2
in-plane size and 300 nm height. In Fig. 3b)
we demonstrate Fourier transformed results of
polarized electric field distributions in a linear scale
for both the in-plane projections and cross-sections
along the longer mesa edge. We find that parallel
configuration exhibits a noticeable localization of the
field while cross-polarized case shows excluded field
from the inside. By comparing the squared absolute
amplitude values in the middle point of the mesa. We
obtain about three times higher value for co-polarized
component, which gives an optical DOLP defined as

of 0.32. Such mesa
induced polarization anisotropy influences the
Figure 4: Optical DOLP calculated for mesa structures of
high in-plane asymmetry and height of a) 300 nm and b)
500 nm showing qualitatively different distributions and
maximum DOLP as high as 0.85.
emitter polarization and in the case of single QDash
with intrinsic DOLP of 0.25 it results in anisotropy of
SE rate on the level of 0.53 when the mesa is oriented
parallel to QDash axis, or -0.08 when it is
perpendicular to it. These results are in good
agreement with the experimentally obtained DOLP
for exciton emission for asymmetric mesa structures.
We can now use this modelling to find the limits
of tailoring the polarization anisotropy of emission at
1550 nm wavelength by using asymmetric mesa
structures of this kind. In Fig. 4 a) we demonstrate the
calculated cross-sectional distribution of optical
DOLP (mesa induced, i. e. for a symmetric quantum
dot with no intrinsic DOLP) for mesa structure of 700
x 100 nm
2
in-plane size and 300 nm height. The
distribution is rather uniform and optical DOLP is
more than 0.85 inside the mesa. By exploiting a
QDash as a quantum emitter embedded parallel in
such a mesa structure, it would be possible to obtain
anisotropy of SE rate above 0.9 due to additional
intrinsic anisotropy of the emitter. In Fig. 4 b) we
examined also mesa of 700 x 200 nm
2
and 500 nm
height and we find a non-trivial distribution of optical
DOLP showing the highest value of 0.9 at 50 nm
below the top edge and lower than 0.6 in a lower
region. Although high DOLP can be obtained in this
configuration, it is crucial to control the position of
QDash precisely along z-direction making fabrication
procedure more problematic.
4 CONCLUSIONS
We demonstrated both experimental and numerical
studies on developing highly linearly polarized
emission from a single InAs quantum dashes emitting
in the spectral range of the 3
rd
telecommunication
window, i. e. at 1550 nm. We investigated anisotropic
submicrometer mesa structures and we found their
influence on a degree of linear polarization of exciton
emission of about 0.3 by using parallel or
perpendicular alignment with respect to the
elongation axis of a quantum dash. Next, we
presented simulations based on FDTD numerical
modelling of polarized electric field intensity
distributions inside a mesa structure, and thus we
found a quantitative agreement with the experiment
by calculating the polarization anisotropy of the
spontaneous emission rates. Such methodology
allowed proposing mesa structures of higher in-plane
asymmetries to find conditions for the degree of
linear polarization of more than 0.85 for a symmetric
quantum dot or above 0.9 for quantum dashes.
PHOTOPTICS 2018 - 6th International Conference on Photonics, Optics and Laser Technology
240
ACKNOWLEDGEMENTS
P. Mrowiński acknowledges financial support from
National Science Centre in Poland within Grant No.
2015/17/N/ST7/03858.
REFERENCES
Akopian, N. et al. (2006) ‘Entangled Photon Pairs from
Semiconductor Quantum Dots’, Physical Review
Letters, 96(13), p. 130501.
Dusanowski, Ł. et al. (2014) ‘Single photon emission at
1.55 μ m from charged and neutral exciton confined in
a single quantum dash’, Applied Physics Letters.
American Institute of Physics Inc., 105(2), p. 21909.
Fiore, A. et al. (2007) ‘Telecom-wavelength single-photon
sources for quantum communications’, Journal of
Physics: Condensed Matter, 19(22), pp. 225005
225014.
Foster, A. P. et al. (2015) ‘Linearly Polarized Emission
from an Embedded Quantum Dot Using Nanowire
Morphology Control’, Nano Letters, p.
150216083859008.
Konishi, K. et al. (2011) ‘Circularly polarized light
emission from semiconductor planar chiral
nanostructures’, Physical Review Letters, 106(5), pp.
14.
‘Lumerical Solutions, Inc.’ (no date). Available at:
http://www.lumerical.com/tcad-
products/fdtd/%0A%0A.
Michler, P. et al. (2000) ‘A Quantum Dot Single-Photon
Turnstile Device’, Science, 290(5500), pp. 22822285.
Mrowiński, P. et al. (2016) ‘Tailoring the
photoluminescence polarization anisotropy of a single
InAs quantum dash by a post-growth modification of its
dielectric environment’, Journal of Applied Physics,
120(7), p. 74303.
Munsch, M. et al. (2012) ‘Linearly polarized, single-mode
spontaneous emission in a photonic nanowire’,
Physical Review Letters, 108(7), pp. 15.
Musiał, A., Kaczmarkiewicz, P., et al. (2012) ‘Carrier
trapping and luminescence polarization in quantum
dashes’, Physical Review B - Condensed Matter and
Materials Physics, 85(3), p. 35314.
Musiał, A., Podemski, P., et al. (2012) ‘Height-driven
linear polarization of the surface emission from
quantum dashes’, Semiconductor Science and
Technology, 27, p. 105022.
Purcell, E. M. (1946) ‘Spontaneous Emission Probabilities
at Radio Frequencies’, Phys. Rev., 69(1112), p. 681.
Reithmaier, J. P., Eisenstein, G. and Forchel, A. (2007)
‘InAs/InP quantum-dash lasers and amplifiers’,
Proceedings of the IEEE, 95(9), pp. 17791790.
Sauerwald, A. et al. (2005) ‘Size control of InAs quantum
dashes’, Applied Physics Letters, 86(25), p. 253112.
Singh, R., Kumar, R. and Singh, V. (2017) ‘Optical
anisotropy and the direction of polarization of exciton
emissions in a semiconductor quantum dot : Effect of
heavy- and light-hole mixing’, Chinese Physics B,
26(8), p. 87303.
Stevenson, R. M. et al. (2006) ‘A semiconductor source of
triggered entangled photon pairs.’, Nature, 439(7073),
pp. 17982.
Tonin, C. et al. (2012) ‘Polarization properties of excitonic
qubits in single self-assembled quantum dots’, Physical
Review B - Condensed Matter and Materials Physics,
85(15), pp. 110.
Highly Linearly Polarized Emission from Quantum Dash Excitons - Modelling and Experiment at the 3rd Telecom Window
241