|
Antimony (Sb) based materials are suitable for the fabrication of optoelectronic devices in the mid-infrared wavelength range. The availability of substrates allows the growth of multilayer structures, where lattice-matched ternary and quaternary layers could be tailored to detect wavelengths in the range of 0.8 to .1 Such detectors are useful for several applications, including atmospheric remote sensing. In such applications, the simultaneous detection of optical signals at different wavelengths allows monitoring several atmospheric species with a single excitation source and receiver. This addresses the necessity for two-color and multi-color detectors, which will reduce system complexity, weight, and cost.2 In this paper, the fabrication and characterization of a dual-band detector are presented. The first band consists of an pn junction for 600 to wavelength detection, while the second band consists of a pn junction for 600 to detection. Compared to and quantum-well (QWIP) two- and multi-color detector technologies, which usually require liquid nitrogen operation, the presented Sb-based detector is capable of room temperature operation.3, 4 The growth of the dual-band detector films, as shown in Fig. 1a, was carried out using in-house built horizontal metal-organic vapor phase epitaxy (MOVPE) equipment. The carrier gas was ultra-pure hydrogen at a flow rate of , and the growth pressure was maintained at with temperature stabilization at . Organometallic sources include trimethylindium, trimethylgallium, tertiarybutylarsine, and trimethylantimony, and the n- and p-dopant sources were diethyltelluride and silane (0.01% in ), respectively. Dimethylzinc (0.1% in ) was also used as a p-type dopant for certain cases. The n-type doping density is about for the layer and for the layer. The growth rate was approximately 3.8 and for the and layers, respectively. Multilayer structures for the dual-band devices were grown on nominally undoped p-type substrates. The structure schematic is shown in Fig. 1a for p/n/n/p-on-p photodetectors. The as-grown multilayer films were characterized and analyzed with optical microscopes, atomic force microscopes (AFM), x-rays, and electron microprobes. The typical alloy composition for the quaternary material is as determined by x-ray diffraction and electron microprobe analyses. Device fabrication was carried out by applying standard processing using three masks. The first mask was used to delineate the upper diode mesa and the second mask was used to define the lower mesa diode structures [see Fig. 1b]. The third mask was used to pattern the contact metals using liftoff technique. Mesa diodes were patterned using wet chemical etching. The contact metals were evaporated using e-beam evaporation. The metal contacts consisted of Pd ( , first layer)/Ge ( , last layer) for the front contact. The back contact was achieved by evaporating on the backside of the substrate. No passivation or antireflective coating was used for this device. Several wafer-based devices were fabricated and characterized. No attempt at dicing or packaging the detectors was performed.5 The detector characterization included dark current and spectral response measurements obtained at . The dark current curves for the upper and lower photodetectors, shown in Fig. 1b, indicate typical exponential dependence on the bias voltage, conforming to the diode theory. The spectral response of the dual-band photodetectors was measured by applying the substitution method, using two reference detectors.6 The first is a 1-cm-diameter Si detector for the 500 to wavelength range, and the second is a detector for up to . Figure 2 shows the spectral response of the upper and lower detectors under different bias voltages. To operate both devices separately, the top and middle contacts were used to access the upper p-n junction detector and the middle and bottom contacts were used to access the lower p-n junction detector. The cutoff wavelength of the upper device is about , which corresponds to the band gap, while the cutoff of the lower device corresponds to the and was tuned to about . As indicted in Fig. 2, the response of the lower detector totally overlaps that of the upper one. This could be tailored using different composition, and yet preserving the lattice-match condition. Controlling the composition will also shift the cutoff wavelength of the quaternary material to a longer wavelength up to . Biasing the pn junction changes the responsivity of the detector. For the reverse bias case the depletion width increases leading to enhanced drift of the photogenerated charge carriers, as opposed to the forward bias case in which it results in narrowing the same region. Figure 3 shows the spectral response of the same sample, normalized to the device area. In this case both devices (upper and lower are serially connected) were accessed only by using the top and bottom contacts. While the bottom contact grounded, applying a voltage to the top contact leads to forward bias of one of the p-n junctions and reverse bias of the other. This results in selectable detection for the devices by the polarity of the applied bias voltage as indicated in Fig. 3. This behavior is attributed to the simultaneous change in the depletion regions with the bias voltage as discussed before, but for the two devices at the same time. In this letter, fabrication and characterization of an Sb-based dual-band two-color detector are presented. and pn junctions were grown lattice-matched to substrate using MOVPE. Dark current measurements illustrated the diode behavior of both lattice-matched detectors. Spectral response measurements indicated either independent operation of both detectors simultaneously, or selective operation of one detector, by the polarity of the bias voltage, while serially accessing both devices. AcknowledgementsThe authors would like to acknowledge Stephen Sandford and Chris Moore for their continuous support. ReferencesA. Srivastava,
J. DeWinter,
C. Caneau,
M. Pollack, and
J. L. Zyskind,
“High performance GaInAsSb/GaSb p-n photodiodes for the 1.8–2.3 mm wavelength range,”
Appl. Phys. Lett., 48
(14), 903
–904
(1986). https://doi.org/10.1063/1.96653 0003-6951 Google Scholar
N. Abedin,
T. Refaat,
I. Bhat,
Y. Xiao,
S. Bandra, and
S. Gunapala,
“Progress of multicolor single detector to detector array development for remote sensing,”
Proc. SPIE, 5543 239
–247
(2004). 0277-786X Google Scholar
X. Jiang,
S. Li, and
M. Tidrow,
“Investigation of a multistack voltage-tunable four-color quantum-well infrared photodetector for mid- and long-wavelength infrared detection,”
IEEE J. Quantum Electron., 35
(11), 1685
–1692
(1999). https://doi.org/10.1109/3.798092 0018-9197 Google Scholar
R. Rajavel,
D. Jamba,
O. Wu,
J. Jensen,
J. Wilson,
E. Patten,
K. Kosai,
P. Goetz,
G. Chapman, and
W. Radford,
“High performance HgCdTe two-color infrared detectors grown by molecular beam epitaxy,”
J. Cryst. Growth, 175/176 653
–658
(1997). 0022-0248 Google Scholar
Y. Xiao,
I. Bhat,
T. Refaat,
N. Abedin, and
Q. Shao,
“MOVPE of multilayers and fabrication of dual band photodetectors,”
Proc. SPIE, 5881 588101
(2005). 0277-786X Google Scholar
L. Thomas,
S. Bruce, and
A. Parr,
“Spectroradiometric detector measurements: part I—ultraviolet detectors and part II—visible to near-infrared detectors,”
NIST Spec. Publ., 250–41 1
–84
(1998). 1048-776X Google Scholar
|