Deep-level transient spectroscopy in InGaAsN lattice-matched to GaAs
ABSTRACT Deep-level transient spectroscopy (DLTS) measurements have been performed on the quaternary semiconductor InGaAsN. A series of as-grown, metalorganic chemical vapor deposited samples having varying composition were grown and measured. A GaAs sample was used as a baseline for comparison. After adding only In to GaAs, we did not detect significant additional defects; however, adding N and both N and In led to larger hole-trap peaks and additional electron-trap peaks in the DLTS data. The samples containing about 2% N, with and without about 6% In, had electron traps with activation energies of about 0.2 and 0.3 eV. A sample with 0.4% N had an electron trap with an activation energy of 0.37 eV.
Spectroscopy in InGaAsN
Lattice-Matched to GaAs
May 2002 • NREL/CP-520-31401
S.W. Johnston, R.K. Ahrenkiel, D.J. Friedman,
and Sarah R. Kurtz
To be presented at the 29th IEEE PV Specialists Conference
New Orleans, Louisiana
May 20-24, 2002
National Renewable Energy Laboratory
1617 Cole Boulevard
Golden, Colorado 80401-3393
NREL is a U.S. Department of Energy Laboratory
Operated by Midwest Research Institute • • • • Battelle • • • • Bechtel
Contract No. DE-AC36-99-GO10337
The submitted manuscript has been offered by an employee of the Midwest Research Institute (MRI), a
contractor of the US Government under Contract No. DE-AC36-99GO10337. Accordingly, the US
Government and MRI retain a nonexclusive royalty-free license to publish or reproduce the published
form of this contribution, or allow others to do so, for US Government purposes.
This report was prepared as an account of work sponsored by an agency of the United States
government. Neither the United States government nor any agency thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents
that its use would not infringe privately owned rights. Reference herein to any specific commercial
product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily
constitute or imply its endorsement, recommendation, or favoring by the United States government or any
agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect
those of the United States government or any agency thereof.
Available electronically at http://www.osti.gov/bridge
Available for a processing fee to U.S. Department of Energy
and its contractors, in paper, from:
U.S. Department of Energy
Office of Scientific and Technical Information
P.O. Box 62
Oak Ridge, TN 37831-0062
Available for sale to the public, in paper, from:
U.S. Department of Commerce
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
online ordering: http://www.ntis.gov/ordering.htm
Printed on paper containing at least 50% wastepaper, including 20% postconsumer waste
DEEP-LEVEL TRANSIENT SPECTROSCOPY IN InGaAsN LATTICE-MATCHED TO
S.W. Johnston, R.K. Ahrenkiel, D.J. Friedman, and Sarah R. Kurtz
National Renewable Energy Laboratory, Golden, CO 80401
Deep-level transient spectroscopy (DLTS) measure-
ments have been performed on the quaternary semicon-
ductor InGaAsN. A series of as-grown, metal-organic
chemical vapor deposited samples having varying com-
position were grown and measured. A GaAs sample was
used as a baseline for comparison. After adding only In
to GaAs, we did not detect significant additional defects;
however, adding N and both N and In led to larger hole-
trap peaks and additional electron-trap peaks in the DLTS
data. The samples containing about 2% N, with and with-
out about 6% In, had electron traps with activation ener-
gies of about 0.2 and 0.3 eV. A sample with 0.4% N had
an electron trap with an activation energy of 0.37 eV.
The quaternary semiconductor InGaAsN can be
grown epitaxially on GaAs, and a bandgap near 1 eV can
be attained [1-5]. Both of these properties are advanta-
geous to developing a four-junction high-efficiency solar
cell, consisting of GaInP, GaAs, InGaAsN, and Ge. Such
a structure has an ideal AM0 efficiency of 41% , but to
date, InGaAsN minority-carrier properties have not been
good enough to be useful in multijunction cells [5,7,8].
Deep-level transient spectroscopy (DLTS) is a powerful
technique for characterizing material defects and impuri-
ties and can provide information to identify lifetime-killing
defects that degrade device performance.
A series of metal-organic chemical vapor deposited
(MOCVD) samples having varying composition were
grown on conductive GaAs substrates. An ohmic contact
was deposited on the back surface, and Schottky contacts
0.75 mm in diameter were deposited on the as-grown
layers. Estimates of N content were obtained by x-ray
diffraction . The samples were then characterized by
measuring room-temperature carrier concentration using
the capacitance-voltage (C-V) technique . Table 1
shows each sample's composition and room-temperature
DLTS data were collected using up to 12 rate windows of
different time constants. Temperature was scanned at a
rate of 10 K per minute, and data for each rate window
were collected during both upward and downward tem-
perature sweeps. These data were then averaged to ac-
count for any temperature lag between the thermocouple
and the sample.
Table 1. Carrier Concentration of Samples Measured by
p-type [Na] (cm-3)
DLTS data for the 0.1-ms rate window during the heating
cycle are shown in Fig. 1.
NT [x 10
400 350300250 200
Fig. 1. Trap concentration from DLTS measurement us-
ing 0.1-ms rate window, 1-V reverse bias, 1-V pulse
height, and 10-ms pulse width.
Data were collected with an applied reverse bias of 1 V, a
pulse amplitude of 1 V (to 0 V), and a pulse width of 10
ms. The ∆C values are the capacitance changes during
the rate windows and are plotted as trap density by using
the relation of Eq. 1 .
NT is trap concentration, Na is p-type hole concentration
determined from room-temperature C-V curves, and C is
the capacitance under the reverse-bias condition of 1 V.
The maximum temperature points are then plotted on an
Arrhenius plot with the corresponding rate-window time
constants. These data for the five samples are shown in
Fig. 2. Open symbols correspond to majority-carrier hole
traps (negative peaks), whereas filled symbols correspond
to electron traps (positive peaks). The data points are fit
by linear equations, giving slope and intercept values.
Fig. 2. Arrhenius plot of DLTS data collected on samples
using 1-V reverse bias, 1-V pulse amplitude, and 10-ms
pulse width. Open symbols show hole traps.
Optical DLTS data were also collected by using a pulsed
xenon flashlamp, in substitution of electrical pulses, to fill
trap levels. The samples were similarly reverse-biased at
1 V. These data are summarized in the Arrhenius plot
shown in Fig. 3. Only positive peaks were detected; thus,
all defect levels are electron traps.
The DLTS data from the Arrhenius plots are summarized
in Table 2. The activation energies, Ea, and capture cross
sections, σ∞, are determined from the Arrhenius plots'
slope and intercept values .
Electrical-pulse DLTS data show that each sample
contains a hole trap with the peak occurring near 350 K,
as seen in Fig. 1. These activation energies range from
~0.62 eV for no N content to ~0.77 eV with N added.
Hole trap concentrations increase from about 1014 cm-3
with little to no N, to ~1015 cm-3 with about 2% N. The
capture cross section for this hole trap increases from
~10-14 cm2 with no N, to ~10-12 cm2 with added N. This
trap may be a native defect or a complex including a na-
tive defect that has previously been reported at 0.72 eV in
GaAs grown under Ga-rich conditions .
Fig. 3. Arrhenius plot of optical DLTS data collected on
samples using a xenon flashlamp and applied reverse
bias of 1 V.
The band-to-impurity lifetime can be estimated from Eq. 2
τ = σ∞vthNT
where vth is thermal velocity and is defined in Eq. 3 .
Using Boltzmann's constant, kB, a temperature of 300 K,
and a GaAs hole conductivity effective mass of 0.34 mo,
vth is 2.0 x 107 cm/s. The lifetime of Eq. 2 assumes a
temperature-independent capture cross section, low in-
jection of excess carriers, and a deep trap level. The life-
time values decrease from an average value of 35 ns with
no N, to an average value of 0.023 ns with N added.
These lifetimes correspond to previously measured values
using time-resolved photoluminescence . Here, life-
times for n-type GaAsN ranged from ~10 ns with no N, to
values between 0.1 and 1 ns with N added.
Table 2. Summary of Electrical-Pulse and Optical DLTS on the Series of InGaAsN Alloy Samples
Electrical pulseOptical pulse (All electron traps)
GaAshole 9.2 x1013
electron 0.82 140
InGaAselectron 0.13.0341.1 x1013
710.26 4.4 x1014
electron0.37 1.5 x1015
0.120.28 6.6 x1014
electron 0.310.46 0.17
electron 0.194.20.151.4 x1015
Electron traps are identified as positive peaks in the DLTS
spectra. Normally, when measuring DLTS using a
Schottky barrier contact, minority carriers are not injected
into the space-charge region (SCR). However, minority
carrier traps can be observed by generating carriers opti-
cally . Others have reported minority-carrier-trap de-
tection using Schottky barriers without the use of optical
generation [14-17]. In these examples, minority-carrier
traps were detected by either using forward-biased pulses
to inject minority carriers [14,15] or by using a high
Schottky barrier [16,17].
We have detected electron traps using a pulse height of 1
V and reverse biases of 1, 2, 3, and 4 V (see Fig. 4).
Thus, the case of forward bias does not apply. However,
a large Schottky barrier can explain electron trapping. We
have measured the built-in potential of the Schottky bar-
rier by C-V, and this value is about 0.8 V for the N-
containing samples. In the neutral bulk, the difference
between the Fermi level and the valence band edge is 0.1
eV. This leads to a Schottky barrier of 0.9 V. The
GaAs0.98N0.02 alloy has a bandgap of 1.2 eV as measured
by photoluminescence . Thus, the Fermi level is
pinned within 0.3 eV of the conduction band at the
When reverse bias is applied, the band bending in-
creases, and the position of the quasi-Fermi level, EFn,
favors filling electron traps near the interface. When a
smaller reverse bias is applied (the filling pulse), the band
bending relaxes, allowing EFn to remain closer to the trap
energy level for more distance into the semiconductor.
This favors electron traps to fill as generation creates
electron-hole pairs in the SCR. Likewise, free-carrier
holes fill hole traps at the retreated edge of the SCR re-
gion near the quasi-neutral p-type bulk. After the filling
pulse, the SCR widens to the corresponding larger re-
verse bias value, and electrons and holes trapped in the
SCR favor emission according to the positions of their
respective quasi-Fermi levels.
Fig. 4. ∆C from DLTS measurement of the GaAs0.98N0.02
sample using 0.1-ms rate window, varying reverse biases,
and 1-V pulse height with 10 ms width.