Growth, steady-state, and time-resolved photoluminescence study of CdTe/MgCdTe
double heterostructures on InSb substrates using molecular beam epitaxy
Michael J. DiNezza, Xin-Hao Zhao, Shi Liu, Alexander P. Kirk, and Yong-Hang Zhang
Citation: Applied Physics Letters 103, 193901 (2013); doi: 10.1063/1.4828984
View online: http://dx.doi.org/10.1063/1.4828984
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/103/19?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Minority carrier lifetime of lattice-matched CdZnTe alloy grown on InSb substrates using molecular beam epitaxy
J. Vac. Sci. Technol. B 33, 011207 (2015); 10.1116/1.4905289
Determination of CdTe bulk carrier lifetime and interface recombination velocity of CdTe/MgCdTe double
heterostructures grown by molecular beam epitaxy
Appl. Phys. Lett. 105, 252101 (2014); 10.1063/1.4904993
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
CdSe/CdTe type-II superlattices grown on GaSb (001) substrates by molecular beam epitaxy
Appl. Phys. Lett. 100, 121908 (2012); 10.1063/1.3697676
Steady-state and time-resolved photoluminescence from relaxed and strained GaN nanowires grown by catalystfree molecular-beam epitaxy
J. Appl. Phys. 103, 124309 (2008); 10.1063/1.2940732

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:
209.147.144.24 On: Tue, 27 Jan 2015 21:16:03

APPLIED PHYSICS LETTERS 103, 193901 (2013)

Growth, steady-state, and time-resolved photoluminescence study
of CdTe/MgCdTe double heterostructures on InSb substrates using
molecular beam epitaxy
Michael J. DiNezza,1,2 Xin-Hao Zhao,1,3 Shi Liu,1,2 Alexander P. Kirk,1,2
and Yong-Hang Zhang1,2,a)
1

Center for Photonics Innovation, Arizona State University, Tempe, Arizona 85287, USA
School of Electrical, Computer and Energy Engineering, Arizona State University, Tempe,
Arizona 85287, USA
3
School for Engineering of Matter, Transport and Energy, Arizona State University, Tempe,
Arizona 85287, USA
2

(Received 4 July 2013; accepted 21 October 2013; published online 5 November 2013)
CdTe/MgCdTe double heterostructures (DHs) are grown on InSb substrates using molecular beam
epitaxy and reveal strong photoluminescence with over double the intensity of a GaAs/AlGaAs
DH with an identical layer structure design grown on GaAs. Time-resolved photoluminescence of
the CdTe/MgCdTe DH gives a Shockley-Read-Hall recombination lifetime of 86 ns, which is more
than one order of magnitude longer than that of typical polycrystalline CdTe films. These findings
indicate that monocrystalline CdTe/MgCdTe DHs effectively reduce surface recombination, have
limited nonradiative interface recombination, and are promising for solar cells that could reach
C
2013 AIP Publishing LLC.
power conversion efficiencies similar to that of GaAs. V
[http://dx.doi.org/10.1063/1.4828984]
Recently, thin-film polycrystalline CdTe solar cells
have shown impressive improvements in device performance and market acceptance, only second to that of Si cells.
Polycrystalline CdTe solar cells have demonstrated efficiencies over 19% under AM1.5G illumination at 298 K,1
while monocrystalline GaAs solar cells have reached a single junction record efficiency of 28.8%,1 which is close to
the detailed balance limit of 33.2%. Our recent modeling
results have shown that monocrystalline CdTe solar cells
can potentially achieve efficiencies greater than 25% based
on an 86 ns bulk Shockley-Read-Hall (SRH) carrier lifetime.2 In order to understand the fundamental physics that
limits the record efficiencies of polycrystalline and monocrystalline CdTe solar cells, it is highly desirable to have a
model system in which the effects of various kinds of
defects can be carefully studied in a controlled fashion. A
double heterostructure (DH) with type-I band edge alignment provides a model system with reduced interface and
surface recombination so that the middle CdTe region can
be isolated and studied. However, few efforts have been
undertaken to explore the use of monocrystalline CdTe and
related DHs grown on lattice-matched substrates for such a
purpose. In this letter, we report the molecular beam epitaxial growth and characterization of a CdTe/MgCdTe DH
grown on an InSb substrate using X-ray diffraction (XRD),
steady-state photoluminescence (PL) and time-resolved PL
(TRPL) measurements. Also, the CdTe/MgCdTe DH is
compared with a GaAs/AlGaAs DH grown on GaAs and a
plain CdTe layer grown on InSb using steady-state PL
measurements.
The layer structures of all three samples used in this
study are shown in Fig. 1. The two DH samples A and B consist of a 1 lm thick undoped (middle) layer of either CdTe or
a)

Electronic mail: yhzhang@asu.edu

0003-6951/2013/103(19)/193901/4/$30.00

GaAs sandwiched between two 30 nm thick Mg0.18Cd0.82Te
or Al0.30Ga0.70As barrier layers, respectively. Both
CdTe/MgCdTe and GaAs/AlGaAs have a type-I band edge
alignment, and the CdTe/MgCdTe heterojunction valence
band offset is 30% of the difference in the bandgaps of the
two materials.3 The composition of each barrier is chosen so
that the conduction and valence band offsets are greater than
several kT, so as to reduce the number of photogenerated carriers from reaching to the top surface and the bottom buffer
layer. For sample A, the CdTe/Mg0.18Cd0.82Te heterojunction
has conduction and valence band offsets of approximately
188 meV and 80 meV, respectively, assuming a MgTe
bandgap of 3.0 eV.3 For sample B, the GaAs/Al0.30Ga0.70As
barriers have conduction and valance band offsets of
237 meV and 138 meV, respectively. A 10 nm CdTe and
GaAs layer is used to cap the MgCdTe and the AlGaAs top
barrier layers, respectively, in order to prevent them from exposure to air.4 The cap layers have a relatively negligible
impact on the PL intensity from the middle CdTe and GaAs
layers from a comparison point of view. Sample C consists of
an undoped 1.57 lm thick CdTe single layer grown on InSb.
This sample is used to determine the effectiveness of the DH
for photogenerated carrier confinement.
The growth is carried out using a dual-chamber VG
V80H molecular beam epitaxy (MBE) system equipped with
two separate III-V and II-VI growth chambers and an ultrahigh vacuum (UHV) transfer chamber. The purity of all
source materials is at least 7N, with the exception of Mg
which is 6N. Effusion cells are used for the group-II materials, and a valved effusion cell is used for Te. The surface
reconstruction during growth is monitored by reflection high
energy electron diffraction (RHEED), and the substrate temperature is measured using 1550 nm and 950 nm infrared
pyrometers for the InSb and GaAs wafers, respectively.
RHEED oscillations observed during the growth of CdTe are

103, 193901-1

C 2013 AIP Publishing LLC
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:
209.147.144.24 On: Tue, 27 Jan 2015 21:16:03

193901-2

DiNezza et al.

Appl. Phys. Lett. 103, 193901 (2013)

FIG. 2. RHEED pattern along the [011] direction recorded 10 min after the
start of CdTe growth on InSb. Half order reconstructions are also seen along
the [001] direction, giving a (2  1) þ c(2  2) pattern, which indicates a Cdrich surface.

FIG. 1. Layer structures for the (a) CdTe/MgCdTe DH, (b) GaAs/AlGaAs
DH, and (c) plain CdTe samples.

used to calibrate the 1:1 Cd/Te flux ratio based on the saturation of the growth rate with increasing Cd flux.5
The growth procedure starts with thermal deoxidation of
III-V substrates under a group-V overpressure, followed by
the growth of a 500 nm thick III-V buffer layer. Typical
growth conditions are used for all III-V layers,6 and streaky
RHEED reconstructions are observed. Samples A and C are
transferred directly to the II-VI growth chamber under UHV
after the InSb buffer layer growth. Immediately prior to
CdTe growth, InSb surfaces are exposed to a Cd flux for several minutes to prevent the formation of a group III-VI alloy
at the interface. A CdTe buffer layer is then grown using an
initial Cd/Te flux ratio of 3.5 in order to further prevent the
formation of In3Te2 at the interface.7 After 2 min of growth,
the flux ratio is then reduced to 1.5. Upon initiation of CdTe
growth on InSb, the RHEED pattern becomes slightly hazy
as the surface reconstruction transitions from InSb to CdTe.
After 10 min of growth the pattern becomes streaky, as
shown in Fig. 2. Both (2  1) and c(2  2) RHEED reconstructions are observed which indicate a Cd-rich growth condition.8 The substrate temperature for samples A and C are
265  C and 280  C, respectively, and the growth rate for both
samples is 9.6 nm/min as determined by RHEED oscillations. The Cd and Te fluxes are kept constant during the
MgCdTe layer growth, since Mg has a larger sticking coefficient and will displace the excess Cd.9
The structural properties of sample A are investigated
using high-resolution XRD measurements. An x-2h scan is
taken in the vicinity of the (004) diffraction peak, and reciprocal space maps (RSMs) are taken in the vicinity of the
(004) and (115) diffraction peaks. A narrow diffraction peak
with a FWHM of 19 arc sec is observed for CdTe (004)

diffraction, as shown in Fig. 3. Pendell€osung fringes of the
CdTe layer are clearly observed, indicating smooth DH interfaces. A weak and broad MgCdTe peak is also observed,
which is expected for the thin barrier layers. Interference
fringes for MgCdTe give a thickness of approximately
30 nm, which is in excellent agreement with the designed
thickness. The (115) RSM in Fig. 4 shows diffraction peaks
that are aligned parallel to the vertical axis, and a similar pattern is observed for the (004) RSM. The (004) RSM indicates
that there is no tilt present in the crystal structure, while the
(115) RSM indicates that the in-plane lattice constants of
CdTe and MgCdTe are matched to InSb. Therefore, the
CdTe and MgCdTe layers are coherently strained. The composition of the MgCdTe barrier layer is determined using the
biaxial strain relationship
az  ao
2C12 ax  ao
¼

;
ao
C11
ao

(1)

where az is the measured lattice constant of MgCdTe in the
growth direction determined using the Bragg equation and
the (004) x-2h scan, ao is the lattice constant of unstrained
MgCdTe, C11 and C12 are elastic stiffness coefficients of

FIG. 3. XRD x-2h scan of the CdTe/MgCdTe double-heterostructure sample taken in the vicinity of the (004) diffraction peak.

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:
209.147.144.24 On: Tue, 27 Jan 2015 21:16:03

193901-3

DiNezza et al.

FIG. 4. RSM of the CdTe/MgCdTe double heterostructure in the vicinity of
the (115) peak with contours plotted on a logarithmic scale. A similar RSM
is observed in the vicinity of the (004) peak. The alignment of the InSb,
CdTe, and MgCdTe diffraction peaks parallel to the vertical axis indicates
that all epilayers are coherently strained to the InSb substrate.

MgCdTe, and ax is the in-plane lattice constant of MgCdTe
which is equal to that of bulk InSb. The values of C11
and C12 for Mg0.18Cd0.82Te are 5.34  1011 dyn/cm2 and
3.68  1011 dyn/cm2, respectively, which are obtained using
linear interpolation of the elastic stiffness coefficients for
CdTe and MgTe.10 Also, it is assumed that MgTe has a lattice constant of 6.42 Å.11 Based on the XRD scans and parameters above, the Mg composition is determined to be
18% 6 1%.
A comparison of the room-temperature steady-state PL
of all three samples is shown in Fig. 5. The pump laser wavelength is chosen to be 532 nm, since the difference between
the absorption coefficients of CdTe and GaAs at this wavelength is less than 6%.12 The laser excitation power density
for samples A and B is 0.10 W/cm2, while sample C needed a
stronger excitation of 2.1 W/cm2 to increase the signal to
noise ratio. As shown in the figure, sample A has over double
the PL intensity as sample B. The strong PL intensity from
sample A is attributed to both improved carrier confinement
provided by the CdTe/MgCdTe DH and the use of a lattice
matched substrate to obtain high quality monocrystalline material. The effectiveness of the carrier confinement is evident
from the relatively low PL intensity of sample C, which gives
over three orders of magnitude lower PL intensity compared
with sample A when measured with the same excitation density. This comparison shows that the MgCdTe barriers have a
low interface state density and effectively reduce both surface
recombination and transport of photogenerated carriers to the
narrow gap (0.17 eV) InSb buffer layer. Further improvements
in CdTe material quality with respect to GaAs are expected
by optimizing the growth temperature and flux ratios.
The photoluminescence decay of sample A shown in
Fig. 6 is obtained using time-correlated single photon counting TRPL measurements. A 530 nm pulsed laser operating at

Appl. Phys. Lett. 103, 193901 (2013)

FIG. 5. Room-temperature PL spectra of the (a) CdTe/MgCdTe and (b)
GaAs/AlGaAs DHs grown on InSb and GaAs, respectively, as well as a (c)
plain CdTe layer grown on InSb. A higher excitation density is used for the
plain CdTe layer to improve the signal-to-noise ratio, and the curve is magnified by a factor of 15. The PL intensity of sample A is over three orders of
magnitude stronger compared with sample C when measured under the
same excitation power density.

an average power density of 0.14 W/cm2 is used. The pulse
width is 6 ps, and the repetition rate is 2 MHz. These conditions give an initial carrier concentration of less than
1  1015 cm3, which is determined by assuming a normal
incidence reflection loss of 27% and a cap layer absorption
loss of 7% at 530 nm.12 A spectrometer with 16 nm resolution is set to a detection wavelength of 820 nm. Fig. 6 shows
a decay time of 86 ns when fitted using a single exponential
decay model in the range of 90 ns–400 ns. The absence of a
second decay time constant in this region suggests that the
86 ns decay time is the SRH carrier lifetime of the middle

FIG. 6. Time-resolved photoluminescence decay of the CdTe/MgCdTe double heterostructure grown on InSb at room temperature. The initial carrier
concentration is less than 1  1015 cm3. A Shockley-Read-Hall lifetime of
86 ns is extracted using a single exponential decay model.

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:
209.147.144.24 On: Tue, 27 Jan 2015 21:16:03

193901-4

DiNezza et al.

CdTe layer. Furthermore, photon recycling is negligible
because the InSb substrate absorbs luminescent photons and
SRH recombination likely dominates in this sample. This
carrier lifetime is much longer than that of polycrystalline
CdTe solar cells, which is typically less than 6 ns.13 It should
be noted that as-deposited polycrystalline CdTe film has a
carrier lifetime of only 0.087 ns, and several processing techniques are needed to improve the material quality.13 The carrier lifetime of sample A is also longer than the current
record of 66 ns reported for bulk single-crystal CdTe measured using a sub-bandgap two-photon method at 850 nm.14
This method selectively excites the bulk material several
hundred microns below the surface, therefore surface recombination is negligible and a fair comparison can be made to
our sample.
In summary, CdTe/MgCdTe DHs grown on InSb substrates show a SRH lifetime of 86 ns and over twice the PL
intensity as an identical GaAs/AlGaAs structure grown on
GaAs. Such a fair comparison is enabled by the use of identical DH layer designs grown on lattice-matched substrates to
provide strong photogenerated carrier confinement and to
reduce the misfit dislocation density, respectively. Coherently
strained high quality MgCdTe barrier layers with relatively
low interface recombination are evident from the over three
order of magnitude improvement in PL intensity for
CdTe/MgCdTe DHs compared with a plain CdTe 1.57 lm
thick layer grown on InSb. The strong PL intensity and long
PL decay time show the promise of monocrystalline CdTe
DHs for high efficiency solar cell applications. Furthermore,
this CdTe/MgCdTe DH sample design provides a model system in which the effects of various defects in CdTe layers can
be investigated in a controlled fashion.
The authors would like to thank Bas A. Korevaar and
Aharon Yakimov at General Electric Global Research Center
and Su Lin at ASU for assistance with TRPL measurements.
This work was partially supported by AFOSR (Grant No.
FA9550-12-1-0444), Science Foundation Arizona (Grant No.

Appl. Phys. Lett. 103, 193901 (2013)

SRG 0339-08), and NSF (Grant No. 1002114). M.J.D. was
supported by the National Science Foundation Graduate
Research Fellowship (Grant No. DGE-0802261) and A.P.K.
is supported by the Bisgrove Postdoctoral Scholars program.
The authors gratefully acknowledge the use of facilities in the
LeRoy Eyring Center for Solid State Science at Arizona State
University.
1

M. A. Green, K. Emery, Y. Hishikawa, W. Warta, and E. D. Dunlop, Prog.
Photovoltaics 21, 827 (2013).
2
A. P. Kirk, M. J. DiNezza, S. Liu, X.-H. Zhao, and Y.-H. Zhang, in
Proceedings of the 39th IEEE Photovoltaic Specialists Conference,
Tampa, Florida, USA, 16–21 June, 2013.
3
A. Waag, F. Fischer, Th. Litz, B. Kuhn-Heinrich, U. Zehnder, W. Ossau,
W. Spahn, H. Heinke, and G. Landwehr, J. Cryst. Growth 138, 155 (1994).
4
A. Waag, H. Heinke, S. Scholl, C. R. Becker, and G. Landwehr, J. Cryst.
Growth 131, 607 (1993).
5
Th. Litz, Th. Behr, D. Hommel, A. Waag, and G. Landwehr, J. Appl.
Phys. 72, 3492 (1992).
6
The growth conditions for the GaAs/AlGaAs sample are very similar to
the ones reported in the following paper without the use of Sb as a surfactant during growth. J.-B. Wang, D. Ding, S. R. Johnson, S.-Q. Yu, and
Y.-H. Zhang, Phys. Status Solidi B 244, 2740 (2007).
7
T. D. Golding, M. Martinka, and J. H. Dinan, J. Appl. Phys. 64, 1873
(1988).
8
S. Tatarenko, F. Bassani, J. C. Klein, K. Saminadayar, J. Cibert, and V. H.
Etgens, J. Vac. Sci. Technol. A 12, 140 (1994).
9
G. Mackh, M. Hilpert, D. R. Yakovlev, W. Ossau, H. Heinke, T. Litz, F.
Fischer, A. Waag, G. Landwehr, R. Hellmann, and E. O. G€
obel, Phys.
Rev. B 50, 14069 (1994).
10
S. Adachi, Handbook on Physical Properties of Semiconductors: II-VI
Compound Semiconductors (Kluwer Academic Publishers, Boston, MA,
2004), Vol. III, pp. 54 and 367.
11
J. M. Hartmann, J. Cibert, F. Kany, H. Mariette, M. Charleux, P.
Alleysson, R. Langer, and G. Feuillet, J. Appl. Phys. 80, 6257 (1996).
12
S. Adachi, Optical Constants of Crystalline and Amorphous
Semiconductors-Numerical Data and Graphical Information (Kluwer
Academic Publishers, Boston, MA, 1999), pp. 222 and 540.
13
D. Kuciauskas, J. N. Duenow, A. Kanevce, J. V. Li, M. R. Young, P.
Dippo, and D. H. Levi, in Proceedings of the 38th IEEE Photovoltaic
Specialists Conference, Austin, Texas, USA, 3–8 June, (Institute of
Electrical and Electronics Engineers, 2012), pp. 001721–001726.
14
D. Kuciauskas, A. Kanevce, J. M. Burst, J. N. Duenow, R. Dhere, D. S.
Albin, D. H. Levi, and R. K. Ahrenkiel, IEEE J. Photovoltaics 3, 1319
(2013).

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:
209.147.144.24 On: Tue, 27 Jan 2015 21:16:03