Photoluminescence studies of type-II CdSe/CdTe superlattices
Jing-Jing Li, Leijun Yin, Shane R. Johnson, B. J. Skromme, Shumin Wang, Xinyu Liu, Ding Ding, Cun-Zheng
Ning, Jacek K. Furdyna, and Yong-Hang Zhang
Citation: Applied Physics Letters 101, 061915 (2012); doi: 10.1063/1.4745199
View online: http://dx.doi.org/10.1063/1.4745199
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/101/6?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Time-resolved and excitation-dependent photoluminescence study of CdTe/MgCdTe double heterostructures
grown by molecular beam epitaxy
J. Vac. Sci. Technol. B 32, 040601 (2014); 10.1116/1.4878317
Optimization of growth conditions of type-II Zn(Cd)Te/ZnCdSe submonolayer quantum dot superlattices for
intermediate band solar cells
J. Vac. Sci. Technol. B 31, 03C119 (2013); 10.1116/1.4797486
Radiative transitions in stacked type-II ZnMgTe quantum dots embedded in ZnSe
J. Appl. Phys. 112, 063521 (2012); 10.1063/1.4754451
CdSe/CdTe type-II superlattices grown on GaSb (001) substrates by molecular beam epitaxy
Appl. Phys. Lett. 100, 121908 (2012); 10.1063/1.3697676
Coexistence of type-I and type-II band lineups in Cd ( Te , Se ) ∕ Zn Se quantum-dot structures
Appl. Phys. Lett. 89, 123110 (2006); 10.1063/1.2355439

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
209.147.144.22 On: Wed, 04 Feb 2015 17:23:29

APPLIED PHYSICS LETTERS 101, 061915 (2012)

Photoluminescence studies of type-II CdSe/CdTe superlattices
Jing-Jing Li,1,2 Leijun Yin,1,3 Shane R. Johnson,1,2 B. J. Skromme,2 Shumin Wang,1,2
Xinyu Liu,4 Ding Ding,1,2 Cun-Zheng Ning,1,2 Jacek K. Furdyna,4
and Yong-Hang Zhang1,2,a)
1

Center for Photonics Innovation, Arizona State University, Tempe, Arizona 85287-5706, USA
School of Electrical, Computer, and Energy Engineering, Arizona State University, Tempe,
Arizona 85287-5706, USA
3
Department of Physics, Arizona State University, Tempe, Arizona 85287-1504, USA
4
Department of Physics, University of Notre Dame, Notre Dame, Indiana 46556, USA
2

(Received 14 May 2012; accepted 27 July 2012; published online 9 August 2012)
CdSe/CdTe type-II superlattices grown on GaSb substrates by molecular beam epitaxy are studied
using time-resolved and steady-state photoluminescence (PL) spectroscopy at 10 K. The relatively
long carrier lifetime of 188 ns observed in time-resolved PL measurements shows good material
quality. The steady-state PL peak position exhibits a blue shift with increasing excess carrier
concentration. Self-consistent solutions of the Schr€odinger and Poisson equations show that
this effect can be explained by band bending as a result of the spatial separation of electrons and
holes, which is critical confirmation of a strong type-II band edge alignment between CdSe and
C 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4745199]
CdTe. V

The monolithic integration of 6.1 Å II-VI (MgZnCd)
(SeTe) and III-V (AlGaIn)(AsSb) semiconductors on commercially available GaSb or InAs substrates offers the freedom to integrate many photonic and electronic devices onto
a single substrate.1–3 Because these materials can be grown
lattice matched on GaSb or InAs substrates, high quality
materials can be obtained with minimal misfit dislocations to
ensure the best possible device performance. In particular,
CdSe/CdTe superlattices have recently been grown on GaSb
substrates using molecular beam epitaxy.4 The excellent
structural and optical properties of these superlattices make
them suitable for applications such as multi-junction solar
cells, light emitting diodes, and photodetectors. The CdSe/
CdTe heterostructure has a type-II band edge alignment, and
therefore offers some unique optical properties.4 This paper
reports a detailed study of the optical properties of the typeII superlattices using time-resolved and steady-state photoluminescence (PL) spectroscopy. The superlattice sample used
as an example consists of 50 periods of alternating CdSe and
CdTe layers with thicknesses of 6.0 6 0.1 nm and
0.8 6 0.1 nm, respectively, on a ZnTe buffer layer grown
using molecular beam epitaxy on the GaSb substrate.
Detailed growth conditions and structural characterization
results are reported in Ref. 4.
The time-dependent PL decay of the superlattice was
measured at 10 K using time-correlated single photon counting and is shown in Fig. 1 for two optical excitation intensities. The excitation source was a 405 nm pulsed laser with
a 50 ps pulse width and a 2.5 MHz repetition rate, and the detector was a Hamamatsu H10330-75 near infrared photomultiplier tube. Assuming there is no carrier recombination
during the laser pulse, the initial excess carrier concentrations are estimated to be 5  1018 6 1  1018 cm3 (red
curve) and 4  1017 6 1  1017 cm3 (blue curve) using the

measured laser peak powers of 554 6 20 mW and
44 6 2 mW, laser spot diameter of 15 6 2 lm, and the
absorption coefficient of 1.5  105 cm1 measured using
UV-visible variable-angle spectroscopic ellipsometry. The
laser spot size is estimated by imaging the focused laser spot
on a reference sample using an Indigo Alpha NIR camera on
a micro-PL setup, with an accuracy of about 1 lm as limited
by the camera pixel size.
The PL decay evolves from a steep initial descent with a
rapidly increasing instantaneous lifetime in the first 75 ns to
a gradual decay with a single exponential tail. The strong dependence of carrier lifetime on the excess carrier concentration in the initial decay can be partly attributed to
concentration-dependent recombination mechanisms such as
radiative and Auger recombination.5 However, the similarity
in the initial slopes of the PL decay curves, in spite of the
very different initial excess carrier concentrations, suggests
that other mechanisms are also involved in the initial steep
PL decay, e.g., carrier diffusion and the direct band-to-band

a)

FIG. 1. Time-resolved photoluminescence decays measured at 10 K with initial carrier concentrations of 5  1018 cm3 and 4  1017 cm3, respectively.

Author to whom correspondence should be addressed. Electronic mail:
yhzhang@asu.edu.

0003-6951/2012/101(6)/061915/4/$30.00

101, 061915-1

C 2012 American Institute of Physics
V

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
209.147.144.22 On: Wed, 04 Feb 2015 17:23:29

061915-2

Li et al.

transition in CdSe. The latter was directly observed in
steady-state PL on a similar sample.
The excess carrier concentrations at the onset of the exponential decay tails are on the order of 1018 cm3 (red
curve) and 1017 cm3 (blue curve), respectively. The net
background doping concentration at room temperature is
estimated to be below 1017 cm3 from the free carrier
absorption measured using an infrared variable angle spectroscopic ellipsometer. Therefore, the excess carrier concentrations at the onset of the exponential decay tails are
assumed to be much larger than the background carrier concentration considering carrier freeze-out at 10 K. In this case,
the constant lifetimes in the exponential decay tails can be
attributed to Shockley-Read-Hall (SRH) recombination, and
the PL intensity has a quadratic dependence on the excess
carrier concentration. The PL decay lifetimes in the exponential decay tails are fit as 174 6 6 ns and 202 6 7 ns for the
initial excess carrier concentrations of 5  1018 cm3 and
4  1017 cm3, respectively. The slight difference between
the observed lifetimes may be due to the limited time range
taken in the measurements, and the SRH lifetime is taken as
188 6 5 ns from the average of those fit values.
The steady-state PL of the superlattice was measured at
10 K with excess carrier concentrations ranging from approximately 9  1015 cm3 to 6  1018 cm3. The sample was
excited by a 488 nm Ar-ion laser with a focused spot diameter
of 134 6 2 lm measured using the knife-edge method, and the
PL signals were collected using a grating monochromator and
measured using a Ge detector cooled to liquid nitrogen temperature. As shown in Fig. 2(a), the PL peak energy of the
superlattice is smaller than the bandgap energy of either CdTe
(1.61 eV)9,12,13 or cubic CdSe (1.76 eV),9,10 suggesting a typeII band edge alignment. Furthermore, the PL peak shifts to
higher energy as the excess carrier concentration increases.
This blue shift is proposed to be caused by the spatial separation of electrons and holes generated by the optical excitation
in the type-II superlattice. Figure 2(b) shows the calculated
band edge alignments, the superlattice minibands, and the corresponding ground state probability densities of the conduc-

Appl. Phys. Lett. 101, 061915 (2012)

tion band and heavy-hole band at an excess carrier
concentration of 6  1018 cm3. (Details of the calculations
are described below.) The electrons and holes are spatially
separated as shown by their probability densities, and their
Coulomb attraction pulls them toward the interfaces of the
two materials. Consequently, the bands are bent down towards
the interfaces, pushing the minibands to higher energies as the
excess carrier concentration increases with increasing excitation power density. A blue shift in the PL peak should, therefore, result from the enhanced band bending caused by the
increased excess carrier concentration.
Pronounced band bending of this type has been observed
primarily in type-II quantum wells and quantum dots.6,7 In
superlattices, however, the coupling between the quantum
wells tends to reduce the charge localization and separation
as shown by the probability densities in Fig. 2(b). To confirm
the band bending effect in the type-II superlattices, selfconsistent solutions of the Schr€odinger and Poisson equations8 are performed to model the increase of the ground state
transition energy with the accumulation of excess carriers.
The modeling starts with the calculation of strain effects
on the band edges. In CdSe/CdTe superlattices, the CdSe
layer is tensilely strained, while the CdTe layer is compressively strained to match the GaSb lattice constant. The strain
effects on the band edges are caused by a superposition of
the hydrostatic strain and shear strain. The hydrostatic strain
shifts the conduction band and valence band edges, while the
shear strain splits the heavy-hole and light-hole bands. The
superlattice minibands are formed from a hybridization of
bound states of the quantum wells coupled through the barriers, which can be obtained by solving the Schr€odinger-like
envelope function equation


h2 d
1 d
WðzÞ þ ½VSL ðzÞ  VE ðzÞWðzÞ ¼ EWðzÞ;

2 dz mðzÞ dz
(1)
where m(z) is the effective mass in the growth direction z,
W(z) is the wave function of the energy state E, and VSL(z)

FIG. 2. (a) Steady-state photoluminescence spectra measured at 10 K with estimated excess carrier concentrations ranging from 9  1015 cm3 to
6  1018 cm3. (b) Calculated band edge alignments (black), minibands (green), and the corresponding ground state probability densities of the conduction
band and heavy-hole band (red) at a carrier concentration of 6  1018 cm3.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
209.147.144.22 On: Wed, 04 Feb 2015 17:23:29

061915-3

Li et al.

Appl. Phys. Lett. 101, 061915 (2012)

and VE(z) are, respectively, the band edge alignment of
the superlattice and the electrostatic potential induced by the
charges. For a superlattice of N periods of a period L, the
wave vector q has N values in the first Brillouin zone with a
spacing of 2p/NL according to the envelope function approximation and the cyclic boundary condition. The energy states
corresponding to þq and –q are degenerate. Equation (1) is
solved, respectively, for the conduction band, the heavy-hole
band, and the light-hole band using the propagation matrix
method, in order to obtain the energy states and wave functions corresponding to the wave vectors.
The charges accumulated under optical excitation generate an electrostatic potential VE(z) as described by the Poisson equation (neglecting ionized dopants)


d
d
eðzÞ
VE ðzÞ ¼ jej½pðzÞ  nðzÞ;
(2)
dz
dz
where e(z) is the dielectric function and n(z) and p(z) are the
electron and hole concentrations. The electron concentration
n(z) is
XX
2
nðzÞ ¼
jWnm ðzÞj Nnm ;
(3)
m

n

where jWnm ðzÞj2 and Nnm are the electron probability density
and the sheet concentration in the nth energy state Enm of the
mth miniband in the conduction band. The electron sheet
concentration for the energy level Enm is
Nnm ¼

kB Tme
lnð1 þ e½FN Enm =kB T Þ:
ph2

(4)

The electron quasi-Fermi level FN can thus be found using
Eqs. (3) and (4) assuming the incident light is evenly
absorbed in every period of the superlattices. The hole quasiFermi level FP is obtained similarly considering both the
heavy-hole and light-hole bands. Consequently, the electron,
heavy-hole, and light-hole concentrations are obtained, and
Eq. (2) is solved for the electrostatic potential VE(z). This
procedure of solving the coupled Schr€odinger and Poisson
equations is iterated until it converges.
The material parameters used in the calculations are
shown in Table I. The absorption coefficient and the surface
reflectivity at the 488 nm laser wavelength are 105 cm1 and
28%, respectively, as estimated from the ellipsometry measurements. The electron and hole concentrations are assumed
for simplicity to be the same in each period of the superlat-

tice that contributes to the PL. The surface recombination is
not expected to be significant because of the strong hole confinement in the structure. The carrier lifetime used in the calculations is 188 ns as obtained from the time-resolved PL
measurements, because most of the carrier concentrations
that cause the blue shift in the PL peak position are in the
range of carrier concentrations that exist during the PL decay
tails.
The calculated ground state transition energy is compared with the blue shift in the PL peak position as a function
of the excess carrier concentration in Fig. 3(a). The calculated ground state transition energy increases by 54 meV,
which is in reasonable agreement with the 40 meV blue shift
observed in the PL peak position. The larger blue shift of the
measured PL peak position at low carrier concentrations
could be caused by filling of the tail states below the band
edges, which is not considered in the calculations. The
smaller blue shift of the measured PL peak position at large
carrier concentrations could be attributed to a decrease of
carrier lifetime due to the increase of wave function overlap
in the type-II superlattice, the direct band-to-band transition
in CdSe or a heating effect at the high excitation levels.
Fig. 3(b) shows the maximum PL peak intensity as a
function of excess carrier concentration. For excess carrier
concentrations of less than 3  1017 cm3, the SRH recombination dominates and, therefore, the PL signal increases
superlinearly with the excess carrier concentration. As the
excess carrier concentration increases, the PL internal quantum efficiency increases until the peak intensity is directly
proportional to excess carrier concentration as shown by the
dashed line, because the competing SRH process is saturated. The PL intensity increases sublinearly when the excess
carrier concentration becomes larger than 9  1017 cm3.
This behavior could be caused by spatially direct band-toband recombination in CdSe, which is out of the detection
range of the Ge detector, or by the heating effect at the high
excitation levels. These results indicate that the carrier lifetime may start to decrease at the excess carrier concentration
of 3  1017 cm3 because the radiative recombination
becomes dominant. Note that the decrease of carrier lifetime
happens at a slightly higher carrier concentration in the PL
decay with an initial excess carrier concentration of
5  1018 cm3. This discrepancy could be due to errors in the
estimation of carrier concentrations for these two
measurements.
In summary, the optical properties of a type-II CdTe/
CdSe superlattice have been studied using time-resolved

TABLE I. Material parameters for zinc-blende CdTe and CdSe [a: lattice constant, Eg: bandgap energy, VBO: valence band offset, me: electron effective
mass, mzhh (mzlh ) and mthh (mtlh ): heavy-hole (light-hole) effective mass in the growth direction and in plane, ac, av, and b: deformation potentials, e: dielectric
constant, and C11 and C12: elastic stiffness constants].

CdSe
CdTe

a (Å)

Eg (eV)

VBO (eV) (Ref. 4)

me

6.052 (Ref. 9)
6.482 (Refs.
9, 11, and 12)

1.76 (Refs. 9 and 10)
1.61 (Refs.
9, 12, and 13)

0.63

0.12 (Ref. 14)
0.088 (Ref. 15)

av (eV) (Ref. 18)
1.81
2.14

b (eV) (Ref. 19)
0.8
1

e
9.6 (Ref. 19)
10.4 (Refs. 12 and 19)

ac (eV) (Ref. 18)
CdSe
3.77
CdTe
5.09

mthh

mzhh

mtlh

mzlh

0.17 (Ref. 17) 0.83 (Ref. 17) 0.36 (Ref. 17) 0.13 (Ref. 17)
0.14 (Refs.
0.53 (Refs.
0.28 (Refs.
0.11 (Refs.
15 and 16)
15 and 16)
15 and 16)
15 and 16)
C11 (1011 dyne/cm2)
8.8 (Ref. 11)
5.35 (Refs. 11, 12, and 20)

C12 (1011 dyne/cm2)
5.3 (Ref. 11)
3.68 (Refs. 11, 12, and 20)

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
209.147.144.22 On: Wed, 04 Feb 2015 17:23:29

061915-4

Li et al.

Appl. Phys. Lett. 101, 061915 (2012)

FIG. 3. (a) The blue shift DE in the measured PL peak position and the calculated ground state transition energy and (b) the maximum PL peak intensity, both
as a function of the excess carrier concentration. The dashed line in part (b) shows the slope expected when the PL intensity is directly proportional to the
excess carrier concentration.

photoluminescence and steady-state photoluminescence. The
photoluminescence decay lifetime due to Shockley-Read-Hall
recombination is obtained as 188 ns. As a result of the type-II
band alignment, the photoluminescence peak position shifts to
shorter wavelengths with an increase of excess carrier concentration. A comparison of the measured blue shift with the
increase in the ground state transition energy obtained from
self-consistent solutions of the Schr€odinger and Poisson equations suggests that this blue shift is primarily due to band
bending associated with the type II band offsets.
This work was partially supported by the Science Foundation Arizona, Contract No. SRG 0339-08, the Air Force
Research Laboratory/Space Vehicles Directorate, Contract
No. FA9453-08-2-0228, and by NSF Grant ECCS-1002072.
The authors would like to acknowledge Dr. Vijay D’Costa
for the ellipsometry measurements.
1

Y.-H. Zhang, S.-N. Wu, D. Ding, S.-Q. Yu, and S. R. Johnson, in Proc.
33rd IEEE Photovoltaic Specialist Conference (2008), p. 20.
2
S. Wang, D. Ding, X. Liu, X.-B. Zhang, D. J. Smith, J. K. Furdyna, and
Y.-H. Zhang, J. Cryst. Growth 311, 2116 (2009).

3

S.-N. Wu, D. Ding, S. R. Johnson, S.-Q. Yu, and Y.-H. Zhang, Prog.
Photovolt. Res. Appl. 18, 328 (2010).
4
J.-J. Li, X. Liu, S. Liu, S. Wang, D. J. Smith, D. Ding, S. R. Johnson, J. K.
Furdyna, and Y.-H. Zhang, Appl. Phys. Lett. 100, 121908 (2012).
5
B. C. Connelly, G. D. Metcalfe, H. Shen, and M. Wraback, Appl. Phys.
Lett. 97, 251117 (2010).
6
N. N. Ledentsov, J. B€
ohrer, M. Beer, F. Heinrichsdorff, M. Grundmann,
and D. Bimberg, Phys. Rev. B 52, 14058 (1995).
7
C. H. Wang, T. T. Chen, K. W. Tan, Y. F. Chen, C. T. Cheng, and P. T.
Chou, J. Appl. Phys. 99, 123521 (2006).
8
K. T. Kim, S. S. Lee, and S. L. Chuang, J. Appl. Phys. 69, 6617 (1991).
9
S.-H. Wei, S. B. Zhang, and A. Zunger, J. Appl. Phys. 87, 1304 (2000).
10
W. Shan, J. Song, H. Luo, and J. Furdyna, Phys. Rev. B 50, 8012 (1994).
11
E. Deligoz, K. Colakoglu, and Y. Ciftci, Physica B 373, 124 (2006).
12
A. J. Strauss, Rev. Phys. Appl. (Paris) 12, 167 (1977).
13
J. Camassel and D. Auvergne, Phys. Rev. B 12, 3258 (1975).
14
Y. D. Kim, M. V. Klein, S. F. Ren, Y. C. Chang, H. Luo, N. Samarth, and
J. K. Furdyna, Phys. Rev. B 49, 7262 (1994).
15
L. S. Dang, G. Neu, and R. Romestain, Solid State Commun. 44, 1187
(1982).
16
P. Lawaetz, Phys. Rev. B 4, 3460 (1971).
17
M. Willatzen, M. Cardona, and N. Christensen, Phys. Rev. B 51, 17992
(1995).
18
S.-H. Wei and A. Zunger, Phys. Rev. B 60, 5404 (1999).
19
S. Adachi, Properties of Group-IV, III-V and II-VI Semiconductors (Wiley,
Chichester, 2005).
20
H. J. McSkimin and D. G. Thomas, J. Appl. Phys. 33, 56 (1962).

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
209.147.144.22 On: Wed, 04 Feb 2015 17:23:29