Optically addressed near and long-wave infrared multiband photodetectors
O. O. Cellek, J. L. Reno, and Y.-H. Zhang
Citation: Applied Physics Letters 100, 241103 (2012); doi: 10.1063/1.4729004
View online: http://dx.doi.org/10.1063/1.4729004
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/100/24?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Optically-aligned visible/near-infrared dual-band photodetector materials and devices on GaAs using
metamorphic epitaxy
J. Appl. Phys. 110, 063109 (2011); 10.1063/1.3642981
Binary superlattice quantum-well infrared photodetectors for long-wavelength broadband detection
Appl. Phys. Lett. 84, 5127 (2004); 10.1063/1.1764932
Quantum-well infrared photodetectors with digital graded superlattice barrier for long-wavelength and broadband
detection
Appl. Phys. Lett. 75, 3207 (1999); 10.1063/1.125279
Optical coupling in quantum well infrared photodetector by diffraction grating
J. Appl. Phys. 84, 5750 (1998); 10.1063/1.368840
Resonant-cavity InGaAs/InAlGaAs/InP photodetector arrays for wavelength demultiplexing applications
Appl. Phys. Lett. 70, 2347 (1997); 10.1063/1.118870

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 17:35:35

APPLIED PHYSICS LETTERS 100, 241103 (2012)

Optically addressed near and long-wave infrared multiband photodetectors
O. O. Cellek,1 J. L. Reno,2 and Y.-H. Zhang1,a)
1

Center for Photonics Innovation and School of Electrical, Computer, and Energy Engineering,
Arizona State University, Tempe, Arizona 85287, USA
2
Center for Integrated Nanotechnologies, Sandia National Laboratories, Albuquerque, New Mexico 87185,
USA

(Received 11 April 2012; accepted 24 May 2012; published online 11 June 2012)
Optically addressed dual-band photodetector incorporating of a 0.82 lm cut-off wavelength near-infrared
(NIR) AlGaAs/GaAs p-i-n photodetector and a 8.2 lm peak wavelength long-wave infrared (LWIR)
AlGaAs/GaAs quantum well infrared photodetector on GaAs substrate is fabricated and
characterized. Switching between NIR and LWIR bands is demonstrated by using external light bias.
The dual-band photodetector gives 65% quantum efficiency in NIR band and specific detectivity of
2  109 cm Hz1/2/W in LWIR band at 68 K. Spectral crosstalk is better than 25 dB. These devices
enable the use of only a single indium-bump per pixel for multiband image sensor arrays to have
C 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4729004]
maximum fill factor. V
Mid-wave infrared band (MWIR, 3–5 lm) and long-wave
infrared band (LWIR, 8–12 lm) photodetectors and their focal
plane arrays (FPA) are important for chemical sensing, environmental monitoring, natural resources survey, and defense
applications. Adding near infrared (NIR, 0.75–1.1 lm) and
short-wave infrared (SWIR, 1.1–2.5 lm) band detection capability to them allows new applications.
Band switching and spectral tuning by electrical bias polarity switching and voltage tuning have been demonstrated with
HgCdTe and quantum well infrared photodetectors (QWIP),
respectively.1,2 Multiband infrared FPA using these principles or
using more than two terminals per pixel have been reported.2–9
Dual-band type-II superlattice infrared photodetectors,10,11 fourcolor QWIPs,12 voltage-tunable quantum-dot infrared photodetectors (QDIP),13 and voltage-switchable NIR/LWIR imaging
with QWIPs were also reported.14 Without band switching,
InSb, HgCdTe, QWIP,15–17 and InGaAs (Ref. 18) detectors can
detect light with broad band from visible/NIR to MWIR/LWIR
ranges. Theory of voltage-tuning in multi-color QWIP stacks
has been discussed in the literature.19,20 However, none of the
demonstrated technologies were able to switch bands with externally applied light in the infrared.
Optically addressed multiband detectors21,22 can switch
detection bands by external optical bias and can be fabricated by using two terminals (single indium-bump per
pixel), which maximizes the pixel fill factor in large infrared (IR) FPAs using small pixel pitch. It also simplifies the
read-out integrated circuit (ROIC) for multiband FPA. Furthermore, when utilized with existing multiband detector
technologies, it enables adding at least one more extra
detection band without the need to change the ROIC design.
The proposed architecture also allows low spectral crosstalk, high band rejection ratio, and, in principle, can have
an infinite number of bands.
The preliminary proof of concept has been recently demonstrated using a triple-junction photodetector covering visible, NIR, and SWIR bands.21 The photodetector consisted of a
set of p-n junction sub-photodetectors with different cutoff
a)

Electronic mail: yhzhang@asu.edu.

0003-6951/2012/100(24)/241103/4/$30.00

wavelengths and series connected through tunnel junctions.
Incident photons pass through all the sub-photodetectors
whose bandgaps have higher energies than the photons’ and
are absorbed by the corresponding sub-photodetector. All subphotodetectors except one are illuminated by a series of light
emitting diodes (LED) with their corresponding detection
wavelengths, and the non-illuminated sub-photodetector
remains the only active photodetector since it limits the total
current. In order to extend this idea into technologically important NIR and LWIR bands, we use unipolar and bipolar
devices to eliminate the need for the tunnel junctions. Equivalent circuit modeling of this design showed that NIR/LWIR
band switching is indeed feasible using the optical addressing
technique.23
In this paper, we experimentally demonstrate the
NIR/LWIR dual-band optically addressed photodetector and
characterize its performance. The photodetector consisted
of two sub-photodetectors. For the LWIR band detection, a
AlGaAs/GaAs QWIP is used as first sub-photodetector, and
for the NIR band detection, a GaAs p-i-n photodetector is
grown on the QWIP as the second sub-photodetector (Sample Reference: VA0411). The optical bias source for optical
addressing is placed outside of the signal beam path (Fig. 1).
The photodetector is defined by mesa etching and connected
with conventional ohmic contacts. The entire device is kept
under constant electrical bias such that mesa top is negative
and p-i-n sub-photodetector is reverse biased. The total
current is limited by the higher resistance NIR p-i-n subphotodetector without any optical bias, because the two
sub-photodetectors are series connected. For front-side illumination, when the NIR p-i-n sub-photodetector is grown on
top of the LWIR sub-photodetector, the optical bias source
illuminates the p-i-n layers first and decreases its resistance.
Since both sub-photodetectors are series connected, when
the NIR optical bias is on, the photodetector current is limited by the smaller current of QWIP sub-photodetector. As
the NIR optical bias photon flux is increased, the p-i-n subphotodetector impedance can be reduced to be orders of
magnitude lower than that of the QWIP. The output photocurrent can be written as23

100, 241103-1

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 17:35:35

241103-2

Cellek, Reno, and Zhang

Appl. Phys. Lett. 100, 241103 (2012)

FIG. 1. Schematic of NIR/LWIR optically addressed photodetector using
single optical bias source.

iout ¼ iQWIP

rQWIP
rPIN
þ iPIN
;
rQWIP þ rPIN
rQWIP þ rPIN

(1)

where iQWIP and iPIN are the AC-currents and rQWIP and rPIN
are the AC-resistances of the QWIP and the p-i-n subphotodetector, respectively. Without optical bias, rQWIP rPIN
holds and the output current is dominated by p-i-n subphotodetector photocurrent, and the device is in the NIR mode
of operation. When optical bias is applied onto the p-i-n subphotodetector, this changes to rQWIP  rPIN and the output
becomes dominated by QWIP sub-photodetector photocurrent
and the device is in the LWIR mode of operation.
The device structure used for this study is grown with
molecular beam epitaxy on (100) GaAs substrate. The QWIP
sub-photodetector structure consists of a 1000 nm thick
1018 cm3 Si doped n-type GaAs layer for a bottom contact
and an absorbing layer consisting of 30 periods of 4.6 nm
GaAs quantum wells with 50 nm Al0.27Ga0.73As barriers. The
center 4.0 nm regions in the quantum wells are n-doped with
Si at 4  1017 cm3. On top of the QWIP structure, the layers
of the p-i-n sub-photodetector are grown in following
sequence: 500 nm n-GaAs doped with Si at 1018 cm3,
2500 nm n-GaAs doped with Si at 2  1016 cm3, and 400 nm
p-Al0.27Ga0.73As window layer doped with C at 1018 cm3
concentration. The cap layer is a 10 nm p-GaAs doped with C
at 1.4  1018 cm3. The device mesa size 150 lm  150 lm is
defined by wet chemical etching. The packaged photodetector
is mounted in a cryostat with a ZnSe front window and a
quartz side window. The ZnSe front window is used for normal incidence optical response characterization, while the
quartz window is used for the optical biasing light.
The I-V curves plotted in Fig. 2 are obtained at different
detector temperatures without any optical bias and with
0.12 W/cm2 optical bias at 780 nm. With an applied optical
bias the device is in LWIR mode of operation where the
QWIP is the current limiting sub-photodetector. A typical
AlGaAs/GaAs QWIP I-V curve24 is observed and showed
current saturation under voltage bias from 1 to 3 V. Open
circuit voltage of the p-i-n sub-photodetector under optical
bias shifts the zero crossing of the I-V curve by 1.4 V to positive bias region. As the temperature reaches 90 K, the dark
current of the QWIP sub-photodetector increases so that the

FIG. 2. Absolute value of the current density vs. electrical bias measured in a
dewar with 180 field of view at temperatures 10, 30, 50, 68, 77, and 90 K.
While the detector is in NIR mode, the p-i-n photodetector determines the current, while the detector is in LWIR mode the QWIP determines the current.

p-i-n sub-photodetector under optical bias becomes the current limiting device. Without optical bias, the photodetector
is in NIR mode of operation, and typical NIR p-i-n photodiode characteristics is observed in reverse bias with leakage
current on the order of 106 A/cm2. In forward bias, the p-i-n
sub-photodetector turns on at 1.4 V, and then the QWIP subphotodetector becomes the current limiting device. Under
higher bias voltages, the current density with optical bias follows the current density without optical bias within a band of
less than 10% increase. This increase in QWIP current is due
to heating of the pixel with the bias laser light.
NIR responsivity is measured with a 780 nm LED. LWIR
responsivity is measured under optical bias from a 780 nm
laser diode and a chopped blackbody source with an IR filter.
The output photocurrent signal is amplified with a current preamplifier connected to a lock-in amplifier. While the photodetector is kept at constant bias of 1 V (mesa top negative),
NIR, and LWIR responsivity is measured under different optical bias levels (Fig. 3). The NIR/LWIR band switching threshold occurs at optical bias power interval of 1–10 mW/cm2 at
68 K. As the operating temperature increases, the increased
QWIP dark current requires a higher optical bias threshold to
switch bands. Figure 3 shows that LWIR peak responsivity is
75 mA/W and remains constant for optical bias down to
6 mW/cm2. The LWIR signal was too weak to be measured at
an optical bias below 6 mW/cm2. The result proves that the
LWIR band can be switched off effectively by decreasing the
optical bias below the optical bias threshold. The measured
peak responsivity corresponds to a LWIR quantum efficiency
of 1%, in good agreement with the typical value for this frontilluminated normal incidence QWIPs without grating coupling.24 Figure 3 shows that NIR peak responsivity at the cut-off
wavelength is constant at 0.43 A/W and the NIR photodetector
is operational with QE of 65% under optical biases below
3 mW/cm2. At optical biases higher than 6 mW/cm2, the device
operates in the LWIR mode with a measured NIR responsivity
of 2.7 mA/W. The NIR band rejection ratio (BRR) at the cutoff
wavelength can be defined as

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 17:35:35

241103-3

Cellek, Reno, and Zhang

Appl. Phys. Lett. 100, 241103 (2012)

FIG. 3. Peak responsivity vs. applied optical bias at 68 K. Band switching
between NIR and LWIR occurs at 3–6 mW/cm2 optical bias at 780 nm
wavelength.

BRRNIR ¼ 10 log

RNIR
0:43 A=W
NIR
¼ 22dB; (2)
¼ 10 log
2:7
mA=W
RLWIR
NIR

LWIR
where RNIR
NIR and RNIR are peak NIR responsivities (A/W
units) in NIR and LWIR mode of operation, respectively.
The LWIR band rejection is higher than the measurement
limit which is determined by the 54 dB dynamic range of the
lock-in amplifier (Stanford Research Systems SR830) and
under the measurement conditions BRRLWIR > 30 dB. The
spectral crosstalk between the bands can be written using the
same notation in the literature.25 Assuming a scenario of
constant irradiance in all wavelengths, crosstalk (CT) of NIR
on LWIR can be written as

ð1
RLWIR ðkÞ
NIR
CTLWIR

¼ 10 log ð

0
1:1 lm

 25 dB;

(3)

RLWIR ðkÞ

FIG. 4. Spectral response in NIR and LWIR modes of operation at 77 K
measured with different beam splitters in FTIR.

Peak specific detectivity (D*) of the detector in LWIR
mode is measured to be 2  109 cm Hz1/2/W under 290 K background and 180 field of view, at an optical bias of 0.12 W/cm2
and an electrical bias of 1 V. The detectivity will improve
with grating coupling, the same QWIP design with back-side
illumination and top-side etched grating coupling is suitable for
FPA imaging with long integration times.6,24,26 Figure 5 shows
the peak responsivity and D* vs. bias voltage for the LWIR
mode of operation. Measurements of the LWIR D* under different electrical biases show that the peak D* decreases as the
electrical bias increases due to an increase in QWIP dark current. LWIR responsivity has a negative differential slope on 1
to 2 V electrical bias range, consistent with early reports and
related with intervalley transfer of electrons.24,27,28 The result
shows that the p-i-n sub-photodetector does not affect the transport characteristics of the QWIP under the reported operating
conditions, and the LWIR QWIP signal is as high as its single
band counterparts.24 In NIR mode, photodetector noise is limited by the system noise which is above the shot noise limit due
to low current density, on the order of 106 A/cm2 at 68 K.

0

where RLWIR(k) is spectral responsivity of the device in LWIR
mode. The numerator is the integrated photocurrent in both
bands, while the denominator is the integrated photocurrent in
the NIR band only. Crosstalk of LWIR on NIR is also limited
LWIR
> 30 dB.
by the measurement set-up and CTNIR
Spectral response of the photodetector is measured with
an FTIR system equipped with KBr and quartz beam splitters
which are installed separately for the measurements in
LWIR and NIR bands, respectively. Calibrated Si and MCT
detectors are used with a halogen lamp as visible and a globar as infrared light sources. Under 1 W/cm2 780 nm optical
bias and at 77 K, the LWIR spectrum is measured, and the
peak wavelength of the QWIP is 8.2 lm with a full width at
half maximum of 1.4 lm (Fig. 4). NIR spectrum is measured
without optical bias, and the cut-off wavelength is at 820 nm.
Visible response of the photodetector is strongly attenuated
due to the thick AlGaAs window layer.

FIG. 5. Detectivity and responsivity vs. applied bias in LWIR band at
780 nm optical bias of 0.12 W/cm2.

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 17:35:35

241103-4

Cellek, Reno, and Zhang

In summary, NIR/LWIR optically addressed dual band
infrared photodetector has been experimentally demonstrated.
The architecture allows single-indium bump per pixel infrared
FPAs, and using a bipolar NIR sub-photodetector with a unipolar LWIR sub-photodetector eliminated the need for tunnel
junctions. Due to large impedance difference between NIR
and LWIR sub-photodetectors, only a single optical bias
source is needed for dual-band switching. Further addition of
more detection bands is possible by using different active
layers grown on the same substrate and by adding more corresponding optical bias sources. Measured LWIR D* value
shows that the device is suitable for passive imaging in LWIR
band.
This work was performed, in part, at the Center for Integrated Nanotechnologies, a U.S. Department of Energy,
Office of Basic Energy Sciences user facility. Sandia National
Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of
Lockheed Martin Corporation, for the U.S. Department of
Energy’s National Nuclear Security Administration under
contract DE-AC04-94AL85000. O.O.C. and Y.H.Z. also
acknowledge funding by the AFOSR under Grant No.
FA9550-10-1-0129.
1

I. Gravé, A. Shakouri, N. Kuze, and A. Yariv, Appl. Phys. Lett. 60, 2362
(1992).
2
E. R. Blazejewski, J. M. Arias, G. M. Williams, W. McLevige, M. Zandian,
and J. Pasko, J. Vac. Sci. Tech. B 10, 1626 (1992).
3
E. P. G. Smith, L. T. Pham, G. M. Venzor, E. M. Norton, M. D. Newton,
P. M. Goetz, V. K. Randall, A. M. Gallagher, G. K. Pierce, E. A. Patten,
R. A. Coussa, K. Kosai, W. A. Radford, L. M. Giegerıch, J. M. Edwards,
S. M. Johnson, S. T. Baur, J. A. Roth, B. Nosho, T. J. De Lyon, J. E. Jensen, and R. E. Longshore, J. Electron. Mater. 33, 509 (2004).
4
P. Abbott, L. Pillans, P. Knowles, and R. K. M. Ewen, Proc. SPIE 7660,
766035 (2010).
5
M. Sundaram, S. C. Wang, M. F. Taylor, A. Reisinger, G. L. Milne, K. B.
Reiff, R. E. Rose, and R. R. Martin, Infrared Phys. Technol. 42, 301
(2001).

Appl. Phys. Lett. 100, 241103 (2012)
6

S. D. Gunapala, S. V. Bandara, J. K. Liu, J. M. Mumolo, D. Z. Ting, C. J.
Hill, J. Nguyen, B. Simolon, J. Woolaway, S. C. Wang, W. Li, P. D.
LeVan, and M. Z. Tidrow, IEEE J. Quantum Electron. 46, 285 (2010).
7
A. Majumdar, K. K. Choi, J. L. Reno, and D. C. Tsui, Appl. Phys. Lett. 86,
261110 (2005).
8
S. U. Eker, M. Kaldirim, Y. Arslan, and C. Besikci, IEEE Electron Device
Lett. 29, 1121 (2008).
9
A. Walther, R. Rehm, J. Fleissner, J. Schmitz, J. Ziegler, W. Cabanski,
and R. Breiter, Proc. SPIE 6542, 654206 (2007).
10
A. Khoshakhlagh, J. B. Rodriguez, E. Plis, G. D. Bishop, Y. D. Sharma, H. S.
Kim, L. R. Dawson, and S. Krishna, Appl. Phys. Lett. 91, 263504 (2007).
11
E. K. Huang, A. Haddadi, G. Chen, B.-M. Nguyen, M.-A. Hoang, R.
McClintock, M. Stegall, and M. Razeghi, Opt. Lett. 36, 2560 (2011).
12
M. Z. Tidrow, X. Jiang, S. S. Li, and K. Bacher, Appl. Phys. Lett. 74,
1335 (1999).
13
S. Krishna, D. Forman, S. Annamalai, P. Dowd, P. Varangis, T. Tumolillo,
A. Gray, J. Zilko, K. Sun, M. Liu, J. Campbell, and D. Carothers, Appl.
Phys. Lett. 86, 193501 (2005).
14
N. Cohen, G. Mizrahi, G. Sarusi, and A. Sa’ar, Infrared Phys. Technol. 47,
43 (2005).
15
A. Bezinger, H. C. Liu, B. Aslan, M. Buchanan, A. J. Springthorpe, P. J.
Poole, D. A. Cardimona, and G. J. Brown, Electron. Lett. 43, 685 (2007).
16
E. Cho, B. McQuinston, W. Lim, D. Rafol, C. Hanson, R. Nguyen, and A.
Hutchinson, Proc. SPIE 5074, 735 (2003).
17
S. V. Bandara, S. D. Gunapala, D. Z. Ting, J. K. Liu, C. J. Hill, J. M.
Mumolo, and S. Keo, Infrared Phys. Technol. 50, 211 (2007).
18
T. Martin, R. Brubaker, P. Dixon, M.-A. Gagliardi, and T. Sudol, Proc.
SPIE 5783, 13 (2005).
19
H. C. Liu, J. Li, J. R. Thompson, Z. R. Wasilewski, M. Buchanan, and
J. G. Simmons, IEEE Electron Dev. Lett. 14, 566 (1993).
20
L. C. Lenchyshyn, H. C. Liu, M. Buchanan, and Z. R. Wasilewski, J.
Appl. Phys. 79, 8091 (1996).
21
E. H. Steenbergen, M. J. DiNezza, W. H. G. Dettlaff, S. H. Lim, and Y.-H.
Zhang, Appl. Phys. Lett. 97, 161111 (2010).
22
E. H. Steenbergen, M. J. DiNezza, W. H. G. Dettlaff, S. H. Lim, and Y.-H.
Zhang, Infrared Phys. Technol. 54, 292 (2011).
23
O. O. Cellek and Y.-H. Zhang, Proc. SPIE 8268, 8268N (2012).
24
B. F. Levine, J. Appl. Phys. 74, R1 (1993).
25
A. Nedelcu, X. Marcadet, O. Huet, and P. Bois, Appl. Phys. Lett. 88,
191113 (2006).
26
S. D. Gunapala, B. F. Levine, L. Pfeiffer, and K. West, J. Appl. Phys. 69,
6517 (1991).
27
R. Rehm, H. Schneider, C. Schönbein, and M. Walther, Physica E 7, 124
(2000).
28
O. O. Cellek and C. Besikci, Semicond. Sci. Tech. 19, 183 (2004).

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 17:35:35