3GW8218 08/26/2014 20:43:33 Page 1

EPITAXIAL QUANTUMDOT INFRAREDPHOTODETECTORS

INTRODUCTION

Infrared (IR) radiation, which was accidently discovered

by the musician and astronomer Sir Frederick William

Herschel (1738–1822), has played an ever increasing role

in improving the quality of life in living beings. The

thermometer played a key role in this discovery and can

be categorized as the first-ever infrared detector. However,

more than a century later, World War II paved the way for

the rapid development of the infrared detector technology.

IR detectors are used for various applications from astron-

omy to zoology, for example in the fields of medical diag-

nostics, environmental monitoring, thermal imaging,

military defense and offense, and space research. Early

detectors used naturally existing materials and gradually

developed to include various semiconductor compounds

such as PbS, InSb, and HgCdTe. The epitaxial growth

techniques (1) developed in the 1970s allowed the growth

of various other semiconductor compounds such as GaAs,

InGaAs, AlGaAs, InAs, and InP, promoting the develop-

ment of infrared detectors using the electronic transitions

between quantized states. The strain induced by the lattice

mismatch of the substrate and the epilayermaterials is the

key to the formation of the self-assembled quantum dots

(QDs) (2).

Theoretical predictions (3) of higher carrier lifetimes,

temperature-independent density of states, and reduced

scattering efficiencies leading to reduced dark current in

ideal zero-dimension devices due to the carrier confine-

ment increased the interest of quantum dot structures to

be used as detectors. Quantum dot infrared photodetectors

(QDIPs) employ self-assembled InAs/GaAs (QDs) in the

active region and represent an emerging detector technol-

ogy. QDIPs exploit three-dimensional confinement of car-

riers and atom-like discrete energy states in quantum dots

to achieve low dark-current levels, high operating temper-

ature, 10–100 times longer carrier lifetimes, and normal

incidence photoresponse. There are several reports on the

enhancement in QDIP performance using novel hetero-

structure designs, most of which chiefly rely on barrier

engineering. Some of these include AlGaAs (4, 5) and

InGaAs (6, 7) barriers, development of an AlAs/InAs/

GaAs QD superlattice structure, (8) submonolayer (ML)

QDs, (9) tunneling QDIP structures (10, 11), the dot-in-a-

well (DWELL) recipe (12), dot-in-double-well heterostruc-

ture (13), confinement-enhancing (CE) barriers (14), and

quaternary InAlGaAs capping with uncoupled InGaAs

QDs (15). Several researchers have attempted to realize

a multispectral response in a quantum well (QW) hetero-

structure (16), a dot-in-well photodetector (17), an InGaAs/

GaAs-based QD intersubband transition (18, 19), and

other devices. However, QDIPs that simultaneously oper-

ate over a broad spectral range and provide superior

performance are not always obtained.

In comparison with quantum well infrared photodetec-

tors (QWIPs), additional degrees of confinement in QDs of

QDIPs lead to threemajor advantages (20): 1) sensitivity to

normal-incidence radiation, which is forbidden in n-type

QWIPs due to polarization selection rules; 2) compara-

tively long (hundreds of picoseconds) effective carrier life-

times, which has been predicted by theory (21) and

confirmed by experiment (6); and 3) low dark current.

Hence, QDIPs are expected to show improved performance

characteristics such as high responsivity, high detectivity,

and high operating temperatures. The dark current of the

QDIPs has been further reduced using the resonant tun-

neling concept. In addition to the aforementioned advan-

tages, QDIPs show improved radiation hardness (22, 23)

and polarization-sensitive spectral responses (24, 25).

Although the potentials and benefits of QD-based struc-

tures as photodetectors have been identified, several areas

still need to be understood and developed. One of themajor

problems associated with QD-based devices is related to

the QD size and shape, which play a major role in QDIPs.

The growth of QDs is a self-assemble process that results in

an unintentional size fluctuation. In general, the size

fluctuation of QDs negatively affects the electrical and

optical properties of the detectors (decreasing absorption

coefficient and increasing dark current), limiting the over-

all performance (26).

QDIPs ranging from single-element detectors (17,

27–31) to focal plane arrays (FPAs) (32–34) have been

demonstrated, whereas operation at temperatures above

77K (7, 28, 35–39) indicate the possibility of developing

uncooled IR imaging systems. In a recent publication,

Matthews et al. (40) reported long carrier lifetime of

3–600 ns for a DWELL detector, which also exhibits a

photoconductive gain of 104 to 105 in the 20–100K tem-

perature range. QDIPs are being developed, in addition

to InAs/GaAs (or InGaAs/GaAs) material systems, using

SiGe/Si (41–43) and GaN/AlN material systems (44). The

behavior of QDs under an applied magnetic field (45, 46)

has recently become a point of interest to understand

physical mechanisms of QDs as well as future spintronic

devices. The spin of an electron in a QD can be used as a

qubit (47–50) for quantum information processing.

In general, an IR detector or a focal plane array camera

captures the intensity profile of the scene. However, if the

information of the scene can be captured using two or more

spectral bands, it would be useful to reconstruct the com-

plete thermal profile of the scene and reduce false posi-

tives. Hence, the development of detectors with multiband

characteristics and the ability to select spectral bands will

immensely aid various applications including land mine

detection, missile-warning sensors, identification of muz-

zle flashes from firearms, and space situational awareness.

QDIPs have become a potential choice for multiband detec-

tion applications as multiple electronic transitions in QDs

can facilitate such a detection capability.

THEORETICAL BACKGROUND

Calculation of Energy States in Quantum Dots

A substantial amount of theoretical work (51–56) has been

carried out to solve a three-dimensional (3-D) confined

J. Webster (ed.),Wiley Encyclopedia of Electrical and Electronics Engineering. Copyright# 2014 John Wiley & Sons, Inc.

3GW8218 08/26/2014 20:43:33 Page 2

quantum mechanical system, such as QDs. Finding solu-

tions for a hypothetical 3-D system is complicated but

achievable with high accuracy. However, the QDs formed

by a self-assembled process, having odd shapes, introduce

many complications for modeling. These complications

become more challenging when the stress, primarily

resulting from lattice mismatch, is involved, which in

general plays a role in the formation of QDs in the self-

assembled process. Hence, the existing theoretical models

involving complicated calculations will always present a

high uncertainty in the results. The most commonly

reported model for solving the energy spectrum of QDs

is the eight-band k.pmodel (51). This model uses the strain

in the QD, calculated using the valence force field (VFF)

model, which has been successful in calculating the strain

tensor in self-assembled QDs.

Although the calculation of the band structure and

energy states in QD system is useful to determine potential

device designs, the investigation of optical properties of a

QD system is also invaluable. The intersubband absorption

coefficient of a photon with energy �hv in a QD layer can be

expressed as (28) follows:

að�hvÞ ¼ pe2�h

e0n0cm20Vav

1

�hv

X

f i

a � pf i

�

�

�

�

2Nð�hvÞ (1)

where Van is the average QD volume, a is the polarization

of the incident light, pfi is the momentum matrix element

between energy states, andN(�hv) is the electron density of

states. Considering a Gaussian inhomogeneous broaden-

ing caused by the large fluctuation in QD size, N(�hv) is

given by N(�hv) ¼

Nð�hvÞ ¼ 1ffiffiffiffiffiffi

2pp

sexp � Ef i � �hv

� �2

2sð Þ2

!

(2)

where Efi is the energy separation between states and s is

the line width of the transition. The momentum matrix

element is calculated from the QD wavefunctions, which

can be obtained from the eight-band k.p model. The spec-

tral response of a QD-based detector is characterized by

peak wavelength (lp), peak responsivity (Rp), and the peak

quantum efficiency (QE, hp). Responsivity is given by R¼qhl/hc, where q is the electron charge, l is the wavelength,

h is the Planck constant, and c is the speed of light.

Quantum efficiency can be calculated from the absorption

coefficient and the thickness of the absorption region.

Apart from this method, several other approaches were

tested for solving a QD-based system. For example, an

energy-level calculation model for a DWELL system was

proposed by Amtout et al. (52). DWELL structures with

different QDs have been tested experimentally, and elec-

tronic spectra obtained by themodel are in good agreement

with the experimental results (52). The pyramidal shape

QDs have been confirmed by transmission electron micros-

copy. The Hamiltonian of a system with a quasi-zero-

dimensional QDplaced in a two-dimensional QW is defined

with a potential energy consisting of four terms: the poten-

tial in the QD region, the potential in the QW region, the

potential in the barrier region, and the potential from the

applied electric field. The eigenfunctions of the Hamilto-

nian are derived using a Bessel function expansion. The

DWELL detectors tested by Amtout have QDs with base

dimensions of 110 and 140A�and heights of 65 and 50A

�,

respectively. The energy spacing between the first and

second energy levels obtained from the model for the

two samples are 132 and 150meV, whereas experimental

analysis showed energy spacing of 123 and 140meV,

respectively. Although the energy states are shifted by

the electrostatic potential from the bias field, the energy

spacing between the first two energy states was not

changed significantly by the applied electric field. This

can be observed in the spectral response of many DWELL

detectors (17, 57).

Using these theoretical models, QDIPs can be designed

using the transitions of carriers between QD states. Sim-

ply, the energy spacing between QD states is adjusted by

varying the QD size, shape, and thematerials of the QD, as

well as the barrier. However, for the design to be realistic,

growth limitations, particularly the QD size, will have to

be taken into account. The most common size possible has

base dimensions of �20nm and a height of �6nm. Hence,

tailoring the operating wavelength by changing the QD

size is practically limited. Nonetheless, there is an

increased interest in growing different size QDs, and

any successful achievement would be beneficial for future

QDIP development. Keeping the growth possibilities

restrained at the current level, there will also be alterna-

tive approaches to tune the transition energy states to

achieve the desired operating wavelength. One of these

approaches is the use of intersubband electron transitions

in QDs (58), and the other approach uses resonant tunnel-

ing (10) to block the dark current selectively without

blocking the photocurrent.

Dark Current and Detectivity

Aiming for high-temperature operation, a theoretical

assessment of QDIPs, comparing QDIPs with other avail-

able detectors, has been recently carried out by Martyniuk

et al. (26). Assuming that the detection is the result of

thermal generation of carriers within the active region, the

dark current and detectivity of QDIPs are calculated using

the model proposed by Ryzhii et al. (59, 60). The 3-D

quantum confinement of electrons in QDs is the key to

increasing the lifetime and reducing the dark current,

which consequently makes QDIPs a potential choice for

high-temperature operation. A typical QD structure

consists of a number of QD layers, which are indexed as

k¼ 1, 2, . . . , K. Then, the average value of dark current

density across the entire structure, as a result of carrier

trapping into the QDs and thermionic emission from QDs,

is expressed as follows (26):

hJdarki ¼ qX

QD

Gk

pk

(3)

where SQD is the dot density, Gk is the rate of the electron

thermoexcitation from QDs in the kth layer, pk is the

capture cross section, and q is the electron charge. As

shown in Figure 1(a), the calculated dark current of QDIPs

2 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:33 Page 3

is more than an order of magnitude lower than that of

HgCdTe (MCT) detectors in the 200–300K temperature

range. In this figure, the 300K background photocurrent

for a field of view (FOV) of 2p is shown for reference.

Martyniuk et al. (26) have also calculated the thermal

detectivity of QDIPs, which is expressed as follows:

D� ¼ h

2hnffiffiffiffiffiffiffiffi

Gth

p (4)

where h is the quantum efficiency, h is the Planck con-

stant, and n is the frequency of light. A similar trend for

detectivity can be observed as shown in Figure 1(b). The

calculated quantum efficiency of QDIPs is significantly

higher than that of MCT detectors within the temperature

range from 200K to 300K. However, the experimentally

demonstratedQDIPs show lowerD� values, represented by

solid symbols, indicating the predicted level of perform-

ance has yet to be reached.

Size of Self-Assembled Quantum Dots

InQDIPs, the size of QDs plays amajor role, particularly in

determining the response wavelength (frequency) range.

However, with the current growth techniques, it is not

possible to obtain QDs with any size as desired. The typical

size of the near-pyramidal InAs/GaAs self-organized

QDs (17, 36) is �60–70A�(height)/�200–250A

�(base)

with QD density varying between 5� 1010 cm�2 and

10� 1010 cm�2. For In0.4Ga0.6As/GaAs QDs (10), the typi-

cal height and base are 60 and 200A�, respectively. The

electron intersublevel energy separation in these QDs

ranges between 40 and 80meV. As reported (24), the

height of the disk-shaped QDs can also be as narrow as

25A�in the growth direction (height), whereas the base

remains around 180A�. Hence, the confinement along

the growth direction is strong, whereas the in-plane

confinement is weak. As reported by Su et al. (61), the

growth of smaller QDs (40A�of height and 130A

�of width),

leading to a large energy spacing (�124meV) between the

QD ground and first excited states, is also possible with

InAlAs/GaAs material combination. In such QDs, the

intersubband transitions between lower QD states lead

to shorter wavelength detection, whereas the transitions

from the upper states (particularly in T-QDIPs) lead to

longer wavelength detection (61). Smaller QDs also pro-

vide a large QD density for the same amount of adatom

change, which increases the absorption coefficient.

MULTIBAND DOTS-IN-A-WELL (DWELL) INFRAREDDETECTORS

In a typical DWELL structure (7, 17, 31, 52, 62–65), InAs

QDs are placed in a thin InGaAs QW, which in turn is

positioned in a GaAs matrix. The DWELL heterostructure

provides strong confinement for carriers trapped in QDs by

lowering the ground state of the QD with respect to the

GaAs band edge, resulting in low thermionic emission.

There can be one or more confined energy states in the

QD, with the position and separation of energy states

dependent on the size of the QD as well as the confinement

potential. The detection mechanism of a DWELL detector

involves the transitions of electrons from the QD ground

state to an excited state in either the QD or QW. Energy

states associated with the QW can be bound, quasi-bound,

or part of the continuum. These different possible transi-

tions lead to multicolor characteristics. A schematic dia-

gram of the conduction band (CB) profile of a DWELL

structure is shown in Figure 2, along with different tran-

sitions between energy states as indicated by the arrows.

The photocurrent, a result of the photoexcitation of carri-

ers, is proportional to the product of the oscillator strength

and the carrier escape probability. A response peak

Figure 1. (a) Comparison between the dark current density of HgCdTe photodiodes and QDIPs as a function of wavelength. Thebackground-generated photocurrent is also shown. (b) Calculated detectivity (D�) of QDIPs andHgCdTe photodetectors at 200K, 250K, and300K temperatures. Background-limited detectivity (FOV¼ 2p, TBLIP¼ 300K, and h¼1) is also shown for comparison. The symbolsrepresent experimental data gathered from literature and data from commercially available detectors (see Reference 26 for details). After

Reference 26.

Epitaxial Quantum Dot Infrared Photodetectors 3

3GW8218 08/26/2014 20:43:34 Page 4

resulting from a bound-to-bound transition has stronger

oscillator strength and a smaller escape probability than a

response peak resulting from a bound-to-continuum tran-

sition. However, the escape probability can be increased by

applying an external electric field. Hence, a bound-to-

continuum peak can be observed even at low biases,

whereas a bound-to-bound peak dominates at high applied

fields because of the enhanced escape probability by field-

assisted tunneling. The energy states in the QD and the

QWcan be adjusted independently by changing the param-

eters associated with each. As a result, DWELL structures

open up a variety of possible designs, leading to multiband

IR detectors. In this section, three-color InAs/InGaAs

DWELL detectors with different well widths are discussed

as previously reported by Ariyawansa et al. (62) Three

DWELL detectors (labeled as DWELL1, DWELL2, and

DWELL3) with different well widths (90A�, 110A

�, and

120A�, respectively) were studied and reported. These

detectors showed response peaks at three distinct wave-

lengths in the mid-wavelength infrared (MWIR), long-

wavelength infrared (LWIR), and very-long-wavelength

infrared (VLWIR) regions.

DWELL Device Structure

The DWELL detector structures reported by Ariyawansa

et al. (62) (DWELL1, DWELL2, and DWELL3) were grown

(66) in a VG-80 solid-sourcemolecular beam epitaxy (MBE)

system with a cracked As2 source at the University of New

Mexico. The GaAs layers were grown at a substrate tem-

perature Tsub¼ 580�C, whereas the In0.15Ga0.85As QW and

the InAs QDs were grown at Tsub¼ 480�C. The tempera-

ture was measured using an optical pyrometer. Using

standard lithography, metal evaporation and wet etching

techniques, n� i�n detector mesas were fabricated for

top-side illumination. Mesas with various circular opti-

cally active areas (diameters ranging from 25–300mm)

were fabricated to test for leakage current and uniformity

of structures. A more detailed discussion on the growth

process can be found in Reference 66. The structure of the

DWELL3 detector is shown in Figure 3. The QDs were

doped n-type with silicon to a sheet density of

5� 1011 cm�2, which corresponds to about 1 electron per

QD. The QW was not intentionally doped. It has been

found (65) that the optimal doping concentration for

DWELL detectors corresponds to about one electron per

QD. The size of the QDs is a critical parameter in the

detector design and is controlled by growth conditions,

especially the temperature and growth rate. Any

inhomogeneous QD size fluctuation will lead to a broaden-

ing of spectral response peaks. The width of the QW, i.e.,

the combined thickness of In0.15Ga0.85As layers, is denoted

byw. The other two detectors (DWELL1 andDWELL2) are

identical to theDWELL3 sample except for thewidth of the

QW. In DWELL2 and DWELL1 detectors structures, the

thickness of the bottom In0.15Ga0.85As layer is 50A�and

30A�, respectively, whereas the top In0.15Ga0.85As layer

thickness is the same (60A�) for both structures, thus

providing a 110 and 90A�well width, respectively.

Figure 2. Conduction band profile of a DWELL structure (a) under zero bias and (b) under negative bias. The energy states correspondingto possible transitions leading to spectral response peaks are indicated by arrows.

Figure 3. Structure of a DWELL detector. The width of the QW,i.e., the combined thickness of In0.15Ga0.85As layers (indicated aswin the figure), is different for each detector. Structures DWELL1,DWELL2, and DWELL3 have well widths of 90, 110, and 120A

�

respectively, whereas all other parameters are the same. After

Reference 62.

4 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:34 Page 5

Effects of the Well Width on Response Peaks

To explain the transition between energy states leading to

response peaks, the experimental results of the three

DWELL detectors reported by Ariyawansa et al. (62) are

discussed. A comparison of dark current voltage (I-V)

characteristics for all three structures is shown inFigure 4.

The sample DWELL3 showed the lowest dark current, and

it increased when the width of the QW was decreased. The

spectral responsivity for the DWELL3 detector under dif-

ferent bias voltages is shown in Figure 5. The band dia-

gram, with corresponding transitions indicated by arrows,

is shown in the inset to Figure 5. The origin of each peak

will be explained in detail in the following sections. The

spectral response of the three detectors in the MWIR/

LWIR regions for�0.5V and�1.4V bias voltages is shown

in Figure 6. All three detectors showed two distinct peaks

in this wavelength range. The DWELL1 detector exhibited

the first peak at �4.2mm and the second peak at �8.1mm.

A semiempirical estimate based on the photoluminescence

spectra with a 60:40 split of the bandgap difference gives a

225–250meV (�4.9–5.5mm) energy separation between

the ground-state of the QD and the conduction band

edge of GaAs. Hence, the 4.2-mm peak is probably a result

of transitions from the ground state of the QD to the

continuum state of the QW, and the second peak is a result

of transitions from the ground state of the QD to a bound

state in the QW, as shown in Figure 2. Moreover, it has

been shown (67) that the line width (Dl/l) of a peak

resulting from transitions from bound-to-bound states is

narrower than that of transitions from bound-to-contin-

uum states. The line width of the 4.2-mm peak is approxi-

mately 42%, whereas the line width of the 8.1-mm peak is

approximately 28%; this observation is consistent with the

aforementioned description. The escape probability of car-

riers excited to a bound state increases with increasing

bias because of field-assisted tunneling. Thus, the bound-

to-bound peak (at 8.1mm) shows a stronger dependency on

the applied bias than the bound-to-continuum peak (at

4.2mm), as evident from Figure 6.

When the width of a QW is increased, the energy

spacing between the states in QW decreases; as a result,

the energy spacing between QD ground state and states in

QW also decreases. Thus, a red shift of the first and second

peaks is observed. The results for the DWELL1 and

DWELL3 detectors confirm this notion. In addition, the

DWELL3 detector exhibits a quasi-bound state close to the

band edge of the GaAs barrier, and hence, the first peak of

the DWELL3 detector is a result of transitions from the

ground state of the QD to the quasi-bound state in the QW.

This can be confirmed by the red shift and the reduced line

width for the first peak of the DWELL3 detector compared

with the first peak of the DWELL1 detector. Based on

width of the QW in detector DWELL2, its peaks are

expected to be located in between the peaks of DWELL1

and DWELL3 detectors. However, the experiment showed

a longer red shift than expected in both the first and the

second peaks of the DWELL2 detector with respect to the

DWELL1 detector. This discrepancy in the result for the

DWELL2 detector could be explained by an unintentional

–2 –1 0 1 2 310

–10

10–8

10–6

10–4

DWELL1

DWELL2

DWELL3

Dark

Curr

ent (A

)

Bias Voltage (V)

T = 4.6 K

Figure 4. A comparison of dark I-V characteristics of threeDWELL structures (DWELL1, DWELL2, and DWELL3) at4.6K. The mesas tested have the same electrical area. The sampleDWELL3showed the lowest dark current and a decrease of darkcurrent is observed as the width of the QW increases. Modified

after Reference 62.

Figure 5. Three color response of DWELL3 detector under dif-ferent bias voltages at 4.6K. Band diagrams with the correspond-ing transitions indicated by arrows are shown in the inset. After

Reference 62.

4 6 8 10 12 140

10

0

5

10

0

1

2

3

248 meVX 5

4.2 µm

8.1 µm

DWELL1

Wavelength ( µm)

248 meV

X 4

9.67 µm6.25 µm

DWELL2

R

esponsiv

ity (

mA

/W)

4.6 K

10.5 µm

6.25 µm DWELL3

–1.4V

–0.5V

Figure 6. The first two peaks of the three DWELL detectorsbiased with �1.4V and �0.5V at 4.6K. Arrows indicate thepeak positions and the “�” sign implies that the curve has beenmultiplied by the specified number. Modified after Reference 62.

Epitaxial Quantum Dot Infrared Photodetectors 5

3GW8218 08/26/2014 20:43:34 Page 6

change in the QD size during the growth process. Further

discussion on this will be given in following sections.

Moreover, several unexplained features in the responsivity

spectra, such as line splitting, were also observed. Based

on doping concentration and sheet density of QDs, it has

been found (31) that a single QD consists of one electron.

Multiple electrons within a QD could lead to a splitting of

photoresponse peaks because of intralevel and interlevel

Coulomb interactions (68). Therefore, the secondary peaks

superimposed on the primary peaks may result from

either different QD sizes in the same DWELL structure

or Coulomb interactions between multiple electrons in

the QD. The expected red shift caused by the Coulomb

interaction with an applied electric field could be compen-

sated by the blue shift caused by the Stark effect (68).

Splitting of absorption peaks is also possible through

interdot coupling (69), which depends on the random

distribution of QDs.

The spectral responsivity of the third peak of detectors

in the VLWIR region is shown in Figure 7. For QDs with a

base diameter of 20nm and a height of 7–8nm, the energy

separation between the ground state and first excited state

calculated from an eight-band k.p model (51) is about

50–60meV. Thus, it is believed that the VLWIR peak is

a result of transitions between two bound states within the

QD. Moreover, the VLWIR peak for DWELL1 and

DWELL3 detectors occurs at approximately the same

wavelength (�23.3mm). Changing the width of the QW

does not affect the energy states in theQD, thus confirming

that the VLWIR peak is a result of transitions between QD

states. However, the VLWIR peak of the DWELL2 detector

appeared at 25.5mm and is red shifted with respect to the

VLWIR peaks of DWELL1 and DWELL3 detectors. This

observation was attributed to the unintentional increase of

QD size in the DWELL2 detector during the growth pro-

cess, resulting in decreased energy spacing between the

ground and the first excited states in the QD. In addition,

this would also decrease the energy spacing between the

ground state in the QD and the bound state in the QW. As a

result, the first two peaks will appear at longer wave-

lengths than expected. This shift was observed in the

spectral response curves of the DWELL2 detector (see

Figure 6).

It is also reported that the VLWIR peak of the DWELL2

detector could be obtained up to 60K, whereas the VLWIR

peaks of DWELL1 and DWELL3 detectors were observed

up to 80K. The highest observed detectivity for the

DWELL3 detector at 23.3mm under a �2.2V bias at

4.6K was reported as �7.9� 1010 cmHz1/2/W. At 80K, a

detectivity of 3.2� 108 cmHz1/2/W was reported under a

�1.4V bias for the DWELL3 detector, whereas DWELL3

detector exhibited lower responsivity and a lower noise

current than those of the DWELL1 detector, resulting in a

higher detectivity for the DWELL3 detector than for the

DWELL1 detector. The improvement in operating temper-

ature of the VLWIR response (up to 80K), compared

with a typical VLWIR QWIP (70) operating at �10–20K,

demonstrates the benefit of a quasi-zero-dimensional

confinement.

Voltage-Tunable Dual-Band DWELL Detector

Some of the other multiband quantum dot Infrared

detectors reported in the literature include a voltage

tunable 3–5mm and 8–12mm dual-band detector with

InAs/Al0.3Ga0.7As/In0.15Ga0.85As confinement-enhanced

DWELL device with response peaks at 5.0mm and

8.6mm operating up to a temperature of 140K. (71). At

77K, the response ratio of the 8.6mm peak over the

5.0mm peak changes from 0.29 at �3V to 5.8 at

þ4.8V. Combined quaternary In0.21Al0.21Ga0.58As and

GaAs capping that relieves strain and maintains strong

carrier confinement was used to demonstrate a four-color

infrared response with peaks in the MWIR (5.7mm),

LWIR (9.0mm and 14.5mm), and far-infrared region

(17mm) (72).

MULTIBAND TUNNELING QUANTUM DOT STRUCTURES

Currently, most of the commercially available IR detectors

do not have low enough dark current to work at higher

temperatures than cryogenic (liquid nitrogen) tempera-

ture. In general, a detector’s dark current should be lower

than its background-limited current at the operating tem-

perature. For some imaging applications, the background

current, which also depends on optics and operating tem-

perature, will determine the highest operating tempera-

ture given the dark current is lower than the background

current. However, there are some other applications that

can benefit from uncooled detectors mainly because of the

reduced complexity and cost. In that case, the detectors

should exhibit low dark current to be eligible for uncooled

operations, which is a challenge as the rate of thermal

excitations leading to the dark current increases exponen-

tially with temperature. Although a QD-based structure is

a potential choice, conventional QDIP structures have not

yet shown adequate performance above 150K. At temper-

atures above 150K, electron occupation is dominated by

the excited states in the QDs; as a result, the reduction in

the dark current is not significant. As a solution, Bhatta-

charya et al. (10) have explored a new resonant tunnel-

based QD device architecture, demonstrating room-tem-

perature IR detection at 6 and 17mm.

20 25 30 350

20

40

60 w = 90

120

110

Re

sp

on

siv

ity (

mA

/W)

Wavelength (µm)

DWELL1

DWELL2

DWELL3

Figure 7. Comparison of the VLWIR response peak of DWELLdetectors. After Reference 62.

6 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:34 Page 7

Aslan et al. (73) has observed resonant tunneling

through a QD layer. In general, any device structure

designed to reduce the dark current will also reduce the

photocurrent. Recently reported tunneling QDIPs (T-

QDIPs) (10, 61, 74, 75) use resonant tunneling to collect

the photocurrent generated within the QDs selectively,

whereas the tunneling barriers (double barriers) block the

majority of carriers contributing to the dark current. The

characteristics of the room-temperature T-QDIP reported

by Bhattacharya et al. (10), showing two-color response at

wavelengths of �6 and �17mm, are discussed in this

section.

A T-QDIP structure can be considered as an extended

DWELL structure. That is, a DWELL structure coupled

with a double-barrier transforms into a T-QDIP structure,

which has several advantages over DWELL structures.

Conventional QDIPs have inherently low dark current,

which can be further reduced using a DWELL structure.

Compared with DWELL detectors, T-QDIPs exhibit lower

dark current as a result of the dark current blocking by the

double barrier. As a result, T-QDIPs have the potential to

achieve the highest possible operating temperature. Addi-

tionally, photocurrent filtering by means of resonant tun-

neling in T-QDIPs offers a solution for low quantum

efficiency, which has been observed in other QD-based

devices. Quantum efficiency can be increased even more

through resonant cavity enhancement (76–78), which

increases the absorption in the active region without

increasing the dark current. In addition, several other

important properties of T-QDIPs include the tunability

of the operating wavelength and the multicolor (band)

nature of the photoresponse based on different transitions

in the structure. The operating wavelength can be tailored

by changing the device parameters of QW, QD, and double

barrier. Using transitions between the energy levels of the

QD and the energy levels of QW, it is possible to obtain

detectors with multiple distinct response peaks.

Modeling T-QDIP Structures

A typical T-QDIP consists of InGaAs QDs embedded

in a AlGaAs/GaAs QW, which is then coupled to a

AlGaAs/InGaAs/AlGaAs double barrier. The T-QDIP

structure [reported by Bhattacharya et al. (10)] and its

conduction band profile under an applied reverse bias are

schematically shown in Figure 8. Pulizzi et al. (79) have

reported resonant tunneling phenomena for a similar QD-

based structure coupled with a double barrier. The photo-

current generated by a transition from a state in the QD

(E0, E1 or E2) to a state in the QW, which is coupled with a

state in the double barrier, can be collected by resonant

tunneling. In this discussion, the energy state in the QW is

denoted as the resonant state Er because it is associated

with resonance tunneling. The double barrier blocks the

majority of carriers contributing to the dark current (car-

riers excited to any state other than the resonant state in

the QW). It can be shown that the tunneling probability is

near unity for carriers excited by radiation with energy

equal to the energy difference between the QD ground

state and the resonant state.

The first step in designing a T-QDIP is the calculation of

the QD energy levels as explained before. The size of QDs

and their barrier material (the confinement potential)

determine the energy spacing between QD bound states.

The width and the confinement potential of QW are

adjusted to obtain the resonant state at the right location

with respect to the QD ground state. In addition, the

doping concentration in QDs should be sufficiently high

to populate all the states that take part in the detection

mechanism. The energy states in the QW with the pres-

ence of the wetting layer and the double-barrier system

are calculated by solving the one-dimensional Schr€odinger

equation. The transmission probability for the double-

barrier structure is calculated using the transfer matrix

method (80). The double barrier (AlGaAs/InGaAs/AlGaAs)

is integrated with each QD layer of the QDIP and is

designed such that the resonant state coincides with a

bound state in the double barrier under certain bias

conditions. In this way, a higher potential barrier for

thermal excitations can be introduced, whereas the

photoexcitation energy remains very low. Because of

the energy-dependent tunneling rate of the double

barrier, the dark current resulting from carriers with a

broad energy distribution is suppressed. Thus, the dark

Figure 8. (a) Structure of a T-QDIP grown by MBE. InGaAs QDs are placed in a GaAs/AlGaAs QW. The AlGaAs/InGaAs/AlGaAs layersserve as a double-barrier system to decouple the dark and photocurrents. The letter “i” stands for intrinsic. (b) Schematic diagram of theconduction band profile of a T-QDIP structure under bias.E0,E1, andE2 are the energy level positions in theQDwith respect to the resonantstate Er. Only the carriers excited to the resonant state contribute to the photocurrent. Modified after Reference 10.

Epitaxial Quantum Dot Infrared Photodetectors 7

3GW8218 08/26/2014 20:43:35 Page 8

current can be significantly reduced, particularly at high

temperatures.

To achieve background limited infrared performance

(BLIP) conditions at high temperatures, the detector

should exhibit an extremely low dark current density. A

TQDIP detector designed to have strong resonant tunnel-

ing can achieve high BLIP temperatures. The dark current

Id of a T-QDIP structure at a bias V is given (74) by

IdðVÞ ¼ evðVÞnemðVÞA (5)

where A is the device area, and v and nem (given by Eqs. 6

and 7 are the average electron drift velocity in the barrier

material and the concentration of electrons excited out of

the QD, respectively. The electron drift velocity is given by

vðVÞ ¼ mFðVÞffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

1þ mFðVÞvs

� �2r (6)

where m is the electron mobility, F is the electric field, and

vs is the electron saturation velocity. The excited electron

density from the QD is given by

nemðVÞ ¼Z 1

�1NðEÞf ðEÞTðE;VÞdE (7)

where f(E) is the Fermi-Dirac distribution function,T(E, V)

is the tunneling probability calculated by the transfer

matrix method (80), and N(E) is the density of states,

which is given by the following equation.

NðEÞ ¼X

i

2Nd

Lp

1ffiffiffiffiffiffiffiffiffi

2psp exp

�ðEf i � EÞ22s2

( )

þ 4pm

Lph2HðE� EWÞ

þ 8pffiffiffi

2p

h3ðm�Þ32

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

E� ECHp

ðE� ECÞ ð8Þ

where the first term is the density of states of the QD state

and Nd is the surface density of QDs. The second term is

the density of the wetting layer states, where EW is the

wetting layer state andH(x) is a step function withH(x)¼ 1

for x� 0 andH(x)¼ 0 for x< 0. The third term describes the

density of states in the barrier material, where EC is the

conduction band edge of the barrier material.

For efficient dark current blocking, the broadening of

the resonant state has to be at a minimum. That is, the

resonant state should be strongly bound. The basic param-

eters should be adjusted so that the tunneling probability

remains close to unity and the carrier escape lifetime is

smaller than the carrier recombination lifetime. The Fermi

level in the QD (hence QD ground state) should be below

the band edge of the QW; however, adjusting the ground

state will change the energy difference between the QD

ground state and the resonant state, which will affect the

peak response wavelength. Thus, all these factors need to

be taken into account to design an optimized detector

exhibiting low dark current.

Two-Color Room Temperature T-QDIP Detector

As reported by Bhattacharya et al. (10) a T-QDIP structure

(TQDIP1) was grown by MBE and then characterized

using I-V, spectral response, and noise measurements.

The results demonstrated the detector’s ability to operate

at room temperature because of the resonant tunneling

phenomenon present in the structure. The detector showed

a two-color response at wavelengths of �6 and �17mm up

to room temperature. A detailed explanation of the device

structure, spectral response and device performance are

given in the following sections.

Device Structure and Dark Current Characteristics. The

T-QDIP structure, TQDIP1, is schematically shown in

Figure 8. Self-assembled In0.4Ga0.6As QDs were grown

on a GaAs substrate. A stack of Al0.3Ga0.7As/

In0.1Ga0.9As/Al0.3Ga0.7As layers serve as the double bar-

rier. The GaAs and AlGaAs layers were grown at 610�Cand the InGaAs or InAlAs QD layers were grown at 500�C.Dark I-V characteristics of TQDIP1, at different tempera-

tures ranging from 80–300K, are shown in Figure 9.

Positive (or negative) bias denotes positive (or negative)

polarity on the top contact. For comparison, the dark

current density between DWELL (DWELL3) at 80K is

also shown in Figure 9. Dark current densities at a bias

of �2V are 3� 10�1 and 1.8� 10�5A/cm2 for DWELL and

T-QDIP, respectively. The reduction in dark current of the

T-QDIP is associated with dark current blocking by the

double barrier in the structure.Moreover, the dark current

densities for TQDIP1 at a bias of 1V were 0.21, 0.96, and

1.55A/cm2 at 240K, 280K, and 300K, respectively. These

dark current density values are lower than the dark cur-

rent values of other IR detectors operating in comparable

wavelength regions at the same temperature. Based on the

dark current variation as a function of bias, negative

conductance peaks were not visible even though resonant

tunneling takes place in the structure. This observation

can be expected for T-QDIPs because sequential resonant

tunneling through ground state is not possible. In T-QDIP

structures, there is no coupling between the QD ground

state and states in the double barrier [unlike in super-

lattice structures (81)]. Also, each active region of the T-

QDIP is separated by a thick spacer layer (400A�GaAs),

which does not allow any significant coupling between two

–4 –2 0 2 4

10–6

10–4

10–2

100

DWELL3

(T = 80 K) Dark

Curr

ent D

ensity (

A/c

m2)

300 K

280

240

200

160

120

80

Bias (V)

Figure 9. Dark current density of the T-QDIP (TQDIP1) as afunction of bias in the temperature range of 80–300K. The darkcurrent density of the DWELL detector (DWELL3) at 80K is alsoshown for comparison. The reduction in dark current observed forT-QDIP is attributed to the blocking barrier in the architecture.

Modified after Reference 10.

8 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:35 Page 9

active regions (two periods). Thus, I-V curves are not

expected to display negative conductance regimes. Fur-

thermore, it is important to underline the thin AlGaAs

barriers (30 or 40A�) on both sides of the QW. Even though

the double barrier is placed only on one side of the QW,

tunneling through the single barrier on the opposite side is

also possible. However, the transmission through this

barrier is lower compared with that through the double

barrier. Therefore, an asymmetric I-V characteristic was

observed.

Spectral Responsivity. The spectral response of TQDIP1

at 80K, 200K, 240K, 280K, and 300K under different bias

values is shown in Figure 10(a), and the variation of the

peak responsivity at 6.2mm is shown in Figure 10(b).

Based on calculations, the allowed confined energy

states in the QD E0, E1, and E2 are located at �161,

�103, and �73meV with respect to the resonant state

(see Figure 8(b)). Thus, the peak at �6mm is a result of

transitions from the ground state of the QD to the resonant

state in the structure, which is consistent with the calcu-

lated energy spacing between corresponding states (DE¼ 161meV). The peak responsivity and the external quan-

tum efficiency (the product of quantum efficiency and the

photoconductive gain) of the 6mm peak at 80K and �4.5V

are �0.75A/W and 16%, respectively. Under reverse bias

(top contact is negative), the photoexcited electrons tunnel

through the double barrier by resonant tunneling. Simi-

larly, under forward bias, photoexcited electrons tunnel

through the single barrier (on the opposite side of the

double barrier). Because of the variations in transmission

through the single and double barriers, the response under

reverse bias is significantly higher than the response

under forward bias, as evident from Figure 10(b). How-

ever, the responsivity shows a strong dependence on the

applied bias in both positive and negative directions. This

behavior is attributed to resonant tunneling similar to that

of double-barrier superlattice structures (73, 82). Applying

a bias across the structure can fine tune the alignment of

the bound state in the QW (resonant state) and the bound

state in the double barrier, allowing for resonant tunneling

conditions. The observed bias dependence of the respon-

sivity indicates that resonant conditions are satisfied over

a considerable range of applied bias voltages. This behav-

ior could be associated with thin barriers and the broad-

ening of the energy states (dE) in the system.

At high temperatures, two distinct peaks centered

around �6 and �17mm were observed, and a weak

response around 11mm was also present. The peak at

17mm results from transitions between the second excited

state of the QD and the resonant state (DE¼ 73meV). The

line width is �26meV, which corresponds to the

inhomogeneous broadening of QD states at 300K. Because

of the symmetry of QD geometry, excited-states have a

higher degeneracy (8) than the ground state (2). The

carrier density in excited-states increases with increasing

temperature, as compared to that in the ground-state. As a

result, the 17mm peak was dominant above 200K. The

weak response at �11mm corresponds to the energy sepa-

ration between the first excited QD state and the resonant

state (DE¼ 102meV).

Noise Characteristics and Detectivity. In general, the

total noise current is a contribution of Johnson noise, 1/f

noise, and generation-recombination noise (g-r) compo-

nents. However, QWIPs (83–85) and QDIPs (4) are

expected to show very low 1/f noise characteristics. The

frequency dependent noise spectrum, which is due to 1/f

and generation-recombination noise components, has the

form (86, 87)

Sðf Þ ¼ Cþ B

fþ A

1þ ff c

� �2(9)

where A, B, and C are constants. The cut-off frequency fc is

given by fc¼ 1/2ptwhere t is the electron lifetime, which is

given by

t / T�2 expEA

kT

�

(10)

where T is the temperature, k is the Boltzmann constant

and EA is the activation energy of the thermally activated

trap level. Noise current density spectra are used to deter-

mine the variation of t with temperature. Based on

equation 8, a plot of log(t/T2) against 1/T would result in

5 10 15 20

0.00

0.05

0.10

0.15

0.20

(a)

V = –2 V

Responsiv

ity (

A/W

)

Wavelength (µm)

300 K

280 K

240 K

200 K

80 K

System Noise Level @ 300 K

–6 –4 –2 0 2 4 610–4

10–3

10–2

10–1

100

Responsiv

ity (

A/W

)

Bias (V)

(b) T = 80 K

Figure 10. (a) Spectral responsivity of T-QDIP (TQDIP1) in thetemperature range of 80–300K under a �2V bias. Two distinctpeaks centered around �6 and �17mm can be observed at hightemperatures, and a weak response around 11mm is also visible.(b) Variation of the peak responsivity at 6mmwith applied bias at80K. Modified after Reference 10.

Epitaxial Quantum Dot Infrared Photodetectors 9

3GW8218 08/26/2014 20:43:35 Page 10

a straight line with a slope of EA/k, which can be used to

calculate the activation energy of the trap level. Further-

more, the electronic states, carrier capture, and carrier

transport properties in a QD-based structure have

been studied (88) using capacitance-voltage spectroscopy

(89, 90) and deep-level transient spectroscopy (91, 92). A

strong negative capacitance phenomenon, which is origi-

nated from carrier capture and emission at interface

states, has been observed in homojunction (93) and

QWIP (94) detector structures. The specific detectivity

(D�) of TQDIP1 detector at different temperatures and

applied biases was obtained from the measured peak

responsivity and noise density spectra. At 80K and under

a bias of �2V, the maximum D� was found to be 1.2� 1010

cmHz1/2/W. The variation in D� at 6.2mm with changing

bias at 80K is shown in Figure 11. The rate of increasing

noise current with increasing bias was much higher than

the rate of increasing responsivity with increasing bias,

resulting in lowerD� at higher bias voltages (beyond2V).

This variation in D� as a function of bias is expected for a

typical photodetector. The value of D� at 17mm and 300K

was of the order of 107 cmHz1/2/W, and with some redesign-

ing of the device heterostructure, a higher D� could be

obtained for the same conditions.

Bias-Selectable Tricolor T-QDIP Detector

Following the same idea of the T-QDIP described previ-

ously, Ariyawansa et al. (95) reported a multicolor T-QDIP

that not only responds at three wavelengths but also

provides a response wavelength selectivity based on the

applied bias voltage. The structure uses two double barri-

ers coupled with QDs through the top and bottom. The

structure of this detector grown bymolecular beam epitaxy

(MBE) is shown in Figure 12. The active region consists of

pyramidal-shape In0.4Ga0.6As QDs sandwiched between

two double barriers that consist of an In0.1Ga0.9As quan-

tum well in 30-A�thick Al0.2Ga0.8As barriers. The widths of

the In0.1Ga0.9As wells in bottom-double barrier and top-

double barrier systems are 60 and 40 A�, respectively. There

are 10 periods of these QDs coupled with double barriers,

and each period is separated with an undoped 400-A�thick

GaAs layer. The GaAs and AlGaAs layers were grown at

610�C. The In0.4Ga0.6AsQDswere grown at 500�C on top of

a 3-mL wetting layer. QDs with height and base dimen-

sions of the �6nm and �20nm, respectively, are n-doped

to 1018 cm�3 using Si as the dopant, whereas all other

layers are undoped except the GaAs bottom and top contact

layers (n-doped to 2� 1018 cm�3).

The band structure for the tricolor T-QDIP detector

with calculated energy levels in the QDs and double-bar-

rier systems is shown in Figure 13(a), (b), and (c) for zero,

forward (top contact is positive), and reverse (bottom con-

tact is positive) bias conditions, respectively. There are

three bound states located at �0.156, �0.065, and

�0.026 eV (ground, first excited, and second excited state,

respectively) with respect to the GaAs conduction band

edge (¼0 eV). The energy states in the wetting layer (pur-

ple dashed line [Color version is available online]) and the

double barriers (blue short dashed lines) are also shown in

Figure 13. Although these states are localized in the

corresponding regions, they also can extend across the

whole structure, especially the wetting layer state. As

shown in Figure 13(a), photoabsorption takes place in

the QDs and electrons are excited from the QD ground

state to the first QD excited state (transition 3 with DE�91meV), to the secondQD excited state (transition 2with

DE �130meV), and to the wetting layer state (transition

1). The electric field-dependent tunneling of excited carri-

ers leads to a selectivity for photoresponse peaks. Under a

certain forward bias condition [see Figure 13(b)], the sec-

ond QD excited state will overlap with the state in the top

double barrier. Hence, the carriers excited to the second

QD state (DE¼ 130meV) will have the maximum tunnel-

ing probability and will be collected as the photocurrent,

leading to a photoresponse at 9.5mm. Similarly, under a

–4 –2 0 2 4 610

7

108

109

1010

D*(

cm

Hz

0.5/W

)

Bias (V)

= 6.2 µm

T = 80 K

Figure 11. Variation of detectivity of the T-QDIP detector(TQDIP1) at 6.2mm as a function of bias at 80K. The rate ofincreasing noise current with bias is much higher than the rate ofincreasing responsivity with bias, resulting in lower D� at higherbias voltages (above 2V).

Figure 12. A 3-D view of a processed tricolor T-QDIP structuregrown by MBE. In0.4Ga0.6As QDs are placed in between two DBs(top and bottom double-barrier systems are indicated by TDB andBDB, respectively). The letter “i” indicates that the layer isintrinsic. After Reference 95.

10 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:36 Page 11

certain reverse bias condition [see Figure 13(c)], the first

QD excited state will overlap with a state in the bottom

double barrier leading to a response peak at 13.6mm(DE¼ 91meV). Under both forward and reverse bias con-

ditions, the carriers excited to the wetting layer also can

tunnel through the barriers. Hence, a short wavelength

peak in the 4.5–5mm range is also expected.

The measured and calibrated spectral response for for-

ward and reverse bias at 50K is shown in Figure 14(a).

Under forward bias (2V), two peaks were observed at

4.5 and 9.5mm because of the transitions from QD ground

state to thewetting layer state and to the secondQDexcited

state, respectively. Under reverse bias (�3.25V), two

peaks were observed at 4.9 and 16.9mm because of the

transitions from QD ground state to the wetting layer state

and to the firstQDexcited state,whereas the peak at 9.5mm(observed for forward bias) is not apparent. The observed

peak selectivity is consistent and in good agreement with

the theoretical calculations, as explained previously. The

shift of the peak associated with the wetting layer is also

expected and in good agreementwith the calculated results.

The calculated transition energies (DE for transition 1) are

264 and 249meV for forward and reverse bias, respectively.

This change is caused by the energy shift of the wetting

layer state by the applied electric field relative to the QD

states. Also, the carriers excited to the wetting layer can

escape even under nonresonant conditions because this

state is close to the top band edge of the barriers. Compared

with the calculated peak positions, the observed peak posi-

tions are in reasonably good agreement and any deviation

can be associated with the uncertainty in structure param-

eters (such asmaterial composition, layer thickness in DBs,

and size of QDs) and the uncertainty in the calculation. It is

also shownthatall thepeakswith thepeakselectivity canbe

observed up to 80K [Figure 14(b)], whereas the short peak

(4.5–5mm) can be observed up to 100K. The dark current

densities at 80K and 300K are �4� 108A/cm2 at 4V and

�8� 104A/cm2 at 2V, respectively. This comparatively

(96) low dark current is attributed to dark current blocking

by the double barriers. The detectivity values at 50K for the

peaks at 4.5 (2V), 9.5 (2V), and 16.9mm (�3.25V) are 3.0,

1.6, and 6.0� 1012 Jones, respectively.

SUPERLATTICE QUANTUM DOT DETECTORS

Anew quantum dot structure-superlattice quantum dot

infrared photodetector (SL-QDIP), which can detect

(c) Reverse Bias

∆E1= 264 meV

∆E2= 130 meV

(b) Forward Bias(a) Zero BiasBD B

–0.3

–0.2

–0.1

0.0

0.1

0.2

E (m

eV

) 1

23

T DB

QD

2

1

∆E1= 249 meV

∆E3= 91 meVQD

3

1

QD

QDDBWL

Figure 13. Schematic diagram of the conduction band profile of the T-QDIP structure under (a) zero, (b) forward, and (c) reverse biasconditions. The calculated bound state energies in the dots (red solid lines), wetting layer (pink dashed lines), and double barriers (blue shortdashed lines) are also indicated. The photoexcited carriers are collected by resonant tunneling through the double barriers. AfterReference

95. DB¼double barrier.

5 10 15

1

2

Responsiv

ity (

mA

/W)

Wavelength (µm)

9.5 µm

16.9 µm

–3.25 V

2 V

T = 50 K

5 10 150.00

0.05

0.10

6420–2–4–610

–9

10–7

10–5

10–3

50 K

Dark

Cu

rre

nt D

ensity (

A/c

m2)

Bias Voltage (V)

300 K

2 VResponsiv

ity (

mA

/W)

Wavelength (µm)

T = 80 K

–3.5 V

)b()a(

Figure 14. (a) Spectral responsivity of the T-QDIP detector at 50K under 2 and�3.25V bias and (b) variation of the peak responsivity withapplied bias. By alternating the bias polarity, the detector can be operated at 9.5 or 16.9mm. The peak response becomes maximumwhen resonant tunneling condition is met for a certain bias value (2 and �3.25V for 9.5 and16.9mm peaks, respectively). Spectralresponsivity at 80K under 2 and �3.5V bias is shown in the inset. After Reference 95.

Epitaxial Quantum Dot Infrared Photodetectors 11

3GW8218 08/26/2014 20:43:36 Page 12

radiation in two spectral bands with improved wavelength

selection capability, is reported by Ariyawansa et al. (97).

As an additional advantage compared with the bias-select-

able multiband T-QDIP (95), the SL-QDIP provides

response wavelength tunability at the detector design

stage without changing the QD size. Compared with the

previously reported superlattice QWIP structure (98),

which showed a response at 10K with either 45-incidence

configuration or corrugated geometry, the SL-QDIP shows

a similar responsivity at 80Kwith normal incidence geom-

etry. Hence, SL-QDIP demonstrates a significant improve-

ment in the operating temperature with normal incidence

detection.

A schematic diagram of the SL-QDIP structure grown

by MBE is shown in Figure 15. The structure consists of

two QD-SLs (labeled as top and bottom QD-SL) separated

by a graded AlxGa1�xAs barrier (x¼ 0.09–0.3), which are

sandwiched in two highly doped (n¼ 2� 1018 cm�3) GaAs

contact layers. The two QD-SLs are identical and consist of

self-assembled In0.4Ga0.6As QDs placed in a superlattice

made of five periods of 90A�GaAs/30A

�Al0.4Ga0.6As quan-

tumwells. The pyramidal shape QDs have height and base

dimensions of �6 and �20nm, respectively, and they were

n-doped to 1.5� 1018 cm�3 using Si. The GaAs and AlGaAs

layers were grown at 610�C, whereas the In0.4Ga0.6As QDs

were grown at 500�C on top of a wetting layer with a

thickness of �3 monolayers. Although this structure con-

sists of one active period (top-QD-SLs/graded barrier/bot-

tom-QD-SL), it is also possible to use multiple periods,

which can be expected to show high performance because

of the increase in light absorption. However, the number of

periods has to be determined to optimize the performance,

taking growth issues into account.

In Figure 16(a), a band diagram with the bound states

in QDs calculated by an eight-band k.pmodel (51) (E0, E1,

andE2) andminibands (M1 andM2) in both SLs are shown.

The approach to calculate the minibands in the two SLs is

explained elsewhere (80). These bound states in the QDs

(E0, E1, and E2) are located at �0.156, �0.065, and

�0.026 eV with respect to the GaAs conduction band

edge. The SLs exhibit two minibands located at 0.093

and 0.269 eV with respect to the GaAs conduction band

edge. In both SLs, the effect of the wetting layer has been

taken into account. As the QDs are doped and the highly

doped GaAs contact layers are separated only by a thin

AlGaAs layer, all the QD energy states are filled with

carriers. In this structure, optical absorption takes place

in the SLs, exciting carriers from all QD states (E0, E1, and

E2) to the minibands. Two sets of closely spaced peaks are

expected because of the excitations from QD states to M1

and M2 minibands. The excited carriers escape over the

graded barrier with the support of the applied electric field

and are collected at the contacts as the photocurrent. The

most important fact is that only one SL becomes active for

photocurrent generation under a given bias direction (for-

ward or reverse). As shown in Figure 16(b), under forward

bias (top-positive), a response resulted in three peaks at

5.1, 7.8, and 10.5mm is expected because of the electronic

transitions from QD states to the lower miniband state

(M1). Similarly, under reverse bias (top-negative), a

response resulted in three peaks at 2.9, 3.7, and 4.2mmis expected because of the electronic transitions from QD

states to the upper miniband (M2). Under this condition,

the carriers excited to the lower miniband (M1) cannot

escape the AlGaAs graded barrier and do not contribute to

the photocurrent.

The experimentally observed response peaks of the SL-

QDIP at 80 and 120K closely follow the theoretical predic-

tions, as shown in Figure 17. A response with three peaks

at 2.9, 3.2, and 4.9mm with the 4.9 peak being the domi-

nant one was observed under reverse bias. The longer

wavelength threshold was observed at�6mm. These peaks

are in good agreement with the theoretically predicted

peaks at 2.9, 3.7, and 4.2mm for reverse bias. Similarly,

a response due to three peaks at 4.3, 7.4, and 11mm with

the 7.4mm peak being the dominant one was observed

Figure 15. Structure of a SL-QDIP detector. Two QD-SLs (top and bottom QD-SLs) are separated by a thick, linearly graded AlxGa1�xAs(x¼ 0.09–0.3) barrier, which in turn are sandwiched between twohighly doped (n¼1.5�1018 cm�3) GaAs contact layers. Each SL consists offive n-doped (1.5� 1018 cm�3) In0.4Ga0.6As QD layers placed in GaAs/Al0.4Ga0.6As wells, as shown in the expanded view. AfterReference97.

12 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:36 Page 13

under forward bias. The long wavelength threshold for this

response bandwas observed at�13mm. This set of peaks is

also in good agreement with the predicted peak locations

(5.1, 7.8, and 10.5mm). As in the band diagram shown in

Figure 16, a response from the transition from the M1

miniband to theM2miniband sensitive to normal incidence

radiation could also be expected. However, in this struc-

ture, electron transition from M1 is not observed because

the doping is such that the Fermi level is kept below M1,

leaving theM1miniband empty. If theM1-to-M2 transition

occurs, then based on the calculated miniband locations

(mentioned before), it should correspond to a peak around

�7mm, and this peak should be dominant for forward bias.

However, this was not observed for forward bias, confirm-

ing that M1-to-M2 transition does not take place in this

structure. As an overall comment, the dominant response

peaks were based on the electronic transitions from the

upper QD states, implying that the transitions from upper-

most states are more efficient. However, for the response

under forward bias, the transition from the E3 state is

weaker than that from theE2 state. This observation could

be due to incomplete carrier occupancy in the E3 state.

Moreover, the variation of the responsivity with bias volt-

age at 7.4mm for forward and at 4.9mm for reverse bias is

shown in the inset in Figure 17(a). Also, the response at

120K for reverse and forward bias is shown in Figure 17(b)

and (c), respectively.

The experimental dark current density up to 200K is

shown in Figure 18(a). The asymmetry in the dark current

can be attributed to the difference in the effective barrier

for the electrons in the two QD-SLs. The wavelength

threshold, which corresponds to the activation energy

obtained from Arrhenius model, is shown in Figure 18

(b). For low bias (�0.5V), the wavelength threshold agrees

with the theoretically predicted response, and it rapidly

increases with bias. This could be caused by tunneling

M2

M1

n+ G

aA

s T

C

Top QD-SLAlGaAs Graded

Barrier

(x= 0.09–0.3)

n+ G

aA

s B

C

Bottom QD-SL

E0

E1

E2

E0

E1

E2

n-In0.4

Ga0.6

As

QDs

M2

M1

(b) Forward Bias (c) Reverse Bias

(a)

Bottom QD-SL

Top QD-SL Bottom QD-SL

Top QD-SL

Figure 16. The conduction band profile of the SL-QDIP under (a) zero, (b) forward (top positive), and (c) reverse bias conditions. The boundstates in QDs (E0, E1, and E2) and minibands (M1 andM2) in both SLs are also shown. In (b) and (c), possible transitions from QD states tominibands leading to spectral response peaks are indicated by vertical arrows, whereas escape of carriers is indicated by horizontal arrows.

After Reference 97.

5 10 150

1

2

Wavelength (µm)

T = 120 K

V = –2 V

(b)

5 10 150

2

4

T = 120 K

V = 1.5 V(c)

(a)

5 10 150.0

0.5

1.0

1.5

2.0

0

50

100

2.9

7.4

–2 V

Re

sp

on

siv

ity (

mA

/W)

Responsiv

ity (

mA

/W)

T = 80 K

4.9

11 4.4

3.2

2 V

20–2

0.1

1

10

100

Re

sp

on

siv

ity (

mA

/W)

Bias Voltage (V)

T = 80 K

Figure 17. Calibrated spectral response of the SL-QDIP underforward and reverse bias conditions at 80K. Peaks at 4.9 and7.4mm are observed because of the transitions of electrons fromQD states to the upper (for reverse bias) and lower minibands (forforward bias), respectively. Variation in the responsivity with biasvoltage at 7.4mm for forward bias and at 4.9mm for reverse bias isshown in the inset. The response at 120K is shown in (b) forreverse and (c) for forward bias. After Reference 97.

Epitaxial Quantum Dot Infrared Photodetectors 13

3GW8218 08/26/2014 20:43:36 Page 14

dominant dark current as opposed to thermal, at high bias.

However, the threshold for forward bias is longer than that

of reverse bias, as expected. Using the measured noise

current spectra, the detectivity values were obtained as

3.2� 109 and 2.6� 109 Jones at 80K for the peaks at

4.9mm (under �2V bias) and 7.4mm (under 2V bias),

respectively. Assuming that the photoconductive gain is

similar12 to the noise gain, the quantum efficiency of the

SL-QDIP was obtained to be �0.4 and 5% at 4.9 and

7.4mm, respectively.

Sensitivity of QDIPs to polarized light is an area that

has not been extensively studied. As reported by Aslan

et al. (24), polarization-dependent electron transitions

exist in QDIPs and only the response peaks sensitive to

in-plane polarized light provide the benefit of normal

incidence detection. To carry out a direct comparison

between 90�-angle (normal) and 45�-angle incidence

responses, a separately processed SL-QDIP with a 45�-angle polished facet was used. The 90�-angle incidence

response of this detector, shown in Figure 19(a), is consist-

ent with the results discussed earlier in this section,

whereas it exhibits a factor of 3 increase in response for

45�-angle incidence radiation over 90�-angle incidence

radiation. This observation can be either because of the

quantum well nature of QDs, although not expected, or

because of the nature of particular electronic transitions

leading to the response peaks [transitions from QD states

(3-D) to superlattice minibands (1-D)]. To investigate the

polarization sensitivity, the response of the SL-QDIP was

measured by varying the angle of polarization of the inci-

dent light. As shown in Figure 19(b), both peaks (4.9 and

7.4mm) have shown a polarization extinction ratio of

about 100:40, which is comparable with typical QWIPs.

The measurement geometry is also shown in the inset to

Figure 19(b). A detailed analysis to this effect would be

beneficial to the advancement of QDIP detectors.

P-TYPE QUANTUM DOT DETECTORS

Despite its promising characteristics, a major challenge

associated with QDIPs is the lowQE (99). For example, the

typical QE was obtained to be about 2%. (100) Optimiza-

tion to 12% can be achieved by using bound-to-bound

transitions in a GaAs-based n-type DWELL structure

(101) that, however, is still less than the QE of HgCdTe

and type II superlattice detectors (99). The relatively low

QE of QDIPs results in part from the large fluctuation of

the dot size in the Stranski-Krastanov growth mode. This,

along with the low QD density compared with the density

of dopants in QWIPs (102), gives rise to the lower absorp-

tion efficiency than expected. Designing optimized struc-

tures to improve QE may include a change of operating

carrier type frommajority electrons to holes, which offers a

few unique characteristics, such as optical transitions

associated with three valence bands and higher effective

mass of the holes. The former leads to a broad response

allowing for convenient tailoring of the spectral response.

The latter features increased the density of states and thus

enhanced absorption, as a great number of holes are

allowed in QDs. Also, a higher effective mass of holes

means the lower dark current compared with conduction

through electrons (103). For conduction through holes

versus conduction through electrons, the higher effective

mass can reduce the photocurrent as well. However, for the

tunneling current, the higher effective mass is an advan-

tage. In addition, although the photocurrent is reduced

because of the higher effective mass, a recent experiment

210–1–2

10

20

Wa

ve

len

gth

Thre

sh

old

(µm

)

Bias Voltage (V)

(a)

(b)

420–2–410

–8

10–6

10–4

10–2

100

Da

rk C

urr

en

t

Density (

A/c

m2)

200 K

160

120

80

20–20.0

0.2

0.4

0.6

No

ise

Ga

in

(c)

Figure 18. Dark current density of the SL-QDIP at temperaturesbetween 80 and 200K. (b) Variation in the wavelength thresholdwith bias calculated based on the Arrheniusmodel. (c) Variation inthe noise gain based on the measured noise current and darkcurrent. After Reference 97.

9060300–30–60–900.2

0.4

0.6

0.8

1.0 s-Polarized

No

rma

lized

Re

spo

nse

(a

.u.) 4.9 µm

7.3 µm

Plarization Angle

p-Polarized

Detector

Substrate

s

p

14121086420.0

0.2

0.4

0

10

20

30

Normal

45-Angle

7.4 µm4.9 µm

0.5 V

Re

sp

on

siv

ity (

mA

/W)

Wavelength (µm)

T = 80 K

–1 V

Re

sp

on

siv

ity (

mA

/W)

(a)(b)

Figure 19. (a) The response of the SL-QDIP at 80K for 90� angle (normal) and 45� angle incidence radiation. (b) The sensitivity of the SL-QDIP to polarized light when the polarization angle is varied. The measurement geometry is shown in the inset. The point at u¼ 0�

corresponding to s-polarized light and u¼90� corresponds to p-polarized light, as indicated by vertical arrows.

14 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:36 Page 15

has demonstrated higher efficiencies for p-type QDIPs

(104). A schematic structure of the p-type QDIP grown

by molecular beam epitaxy is shown in Figure 20(a). The

absorbing region consists of 10 periods of InAs QDs,

between which is an 80-nm-thick undoped GaAs barrier.

The pyramidal shapeQDs have the height and base dimen-

sions of 5 and �20–25nm, respectively. The dot density

is about 5� 1010 cm�2. Free holes are introduced by a

d-doping technique. A sheet density of 5� 1011 cm�2

p-type dopants is placed above the QDs, with a 15-nm-

thick spacer (GaAs) between them, which gives about 10

holes per dot.

The hole states in QDs were calculated as shown in

Figure 20(b) by using an 8� 8k�p model described in

Reference 51. In contrast to only one electron state in

the CB, many hole states are allowed in the dots. From

numerical computation point of view, thismeans amassive

number of Eigenvalues to be solved simultaneously from

the eight-band Hamiltonian, which becomes even difficult

in the higher hole energy range where dense states are

included. To facilitate the computation, the spin-orbit

split-off (SO) states were obtained by treating the QD as

aQWand using an effective-massmethod. Themuchwider

in-plane dimension of the dots than the height partially

validates such a treatment. The comparison of QD and

QW states for the heavy-hole (HH) level, as shown in

Figure 20(b), indicates that the obtained ground states

from two approaches are close to each other. We use the

QWSO state to represent for the QD SO hole and interpret

the spectral response, which should be acceptable for

analysis on distinguishing respective contributions of

valence-band (VB) hole transitions to the response.

The computed electronic structure of QDs is used to

interpret the spectral response of p-QDIP, as shown in

Figure 21(a), displaying two primary response bands at

1.5–3 and 3–10mm. The overall spectral profile is analogy

with that of the p-type GaAs heterojunction detector (105)

[Figure 21(b)]. However, the responsivity of p-QDIP is

about 10–20 times higher than that of the heterojunction

detector, as shown in Figure 21(b) and (c), even though

QDIP contains a much thinner absorbing region than the

heterojunction. This characteristic indicates that the ori-

gin of response should be dominant because of the QDs but

not the p-type GaAs contact layers [Figure 20(a)], benefit-

ing from the longer hole lifetime of QDs. The small spacing

between hole states (<30meV) leads to varying responses

with the photon energy in accordance with the band struc-

ture of InAs/GaAs. For example, the two response bands lie

above and below the SO splitting energy of InAs (0.39 eV or

3.2mm in wavelength). The experimental short-wave-

length response peak at 0.552 eV corresponds to the hole

transition from the HH ground state to the SO state

[0.609 eV by calculation as indicated by transition I of

Figure 20(b)], whereas the long-wavelength peak at

0.247 eV [also see inset of Figure 21(a)] corresponds to

the hole transition from the HH ground state to the state

Figure 20. (a) Schematic of the p-typeQDIP structure. Free holesare introduced into QDs by d-doping above the QD layer.(b) Computed valence band structure of the QDs, where solidlines represent for hole states obtained by using an 8�8k�pmodel.The thick lines are the band edges. The dashed lines are the holeground states of a QW, which has the same thickness as the QDheight. The QW states were obtained using an effective-massmethod, with the potential confinement adopted from that ofthe QD in the growth direction through the dot center. The statesbeyond the LH level become denser because of the larger holeeffectivemass and because of the LH confinement potential, whichleads to a quasi-continuum of tenuously bound states. Threetransitions (I, II, and III) indicated contribute to the responseas observed experimentally, where the transition III only observedat higher biases has the much weaker contribution compared withthe other two. After Reference 104.

Figure 21. (a) Spectral response of the p-type QDIP at 78K. Tworesponse bands at 1.5–3 and 3–10mm originate from hole transi-tions between the SO-HH and HH-HH QD levels, respectively.Inset: Gaussian fits to the spectral response show lower-energyresponse peak at 0.148 eV because of the LH-HH transition.(b) Comparison between the response of the p-type QDIP andheterojunction detector (105). The bias voltages are selected suchthat they lead to nearly the same electric field. The heterojunctiondetector consists of 30 periods of p-type GaAs (3� 1018 cm�3,18.8nm) and Al0.28Ga0.72As (60nm). Its spectral response wasreported elsewhere (105). (c) The variation of peak responsivitywith the electric field. The responsivity of p-QDIP is about 10–20times higher than that of the heterojunction detector. After Ref-

erence 104.

Epitaxial Quantum Dot Infrared Photodetectors 15

3GW8218 08/26/2014 20:43:37 Page 16

near the GaAs barrier [transition II, calculated as

0.260 eV]. In view of the assumed ideal pyramidal shape

in the calculations, the experiment and computed elec-

tronic structure results agree reasonably well.

The hole transition which contributes to the primary

response peak at 0.247 eV may end up to quasi-bound

states. It can be seen from Figure 20(b) that hole states

become denser (106) at the higher energy portion of theHH

confinement potential, partly because of the larger hole

effective mass and the light-hole (LH) confinement poten-

tial, which leads to a continuum of tenuously bound states

(107). This characteristic is consistent with the broad

nature of the response peak with Dl/l¼ 0.42, where land Dl are 5.2mm and 2.2mm, respectively. However,

the HH-bound-to-HH quasi-bound transition may domi-

nate over the HH bound-to-LH continuum transition, as

the bound-to-quasi-bound transition has the higher

absorption than the bound-to-continuum transition

(101). Compared to the HH to HH response, the short-

wavelength response contributed by the HH to SO transi-

tion is not as strong as in the heterojunction case, as shown

in Figure 21(b). A possible cause is the impact of strain on

the local band edges (108), leading to a much shallower SO

confinement potential than the HH band and giving rise to

continuum SO states. Additionally, scattering events are

required to transfer holes in the SO states to the HH states

of the barrier to facilitate transport, which somewhat

reduces the escape efficiency.

Typical QDIPs display bias-dependent multiple-wave-

length responses as a result of the change between bound-

to-bound and bound-to-continuum transitions. Such an

effect is less pronounced in the present p-type detector

compared with n-type QDIP (18). It was observed that a

long-wavelength response tail rises at higher bias as

shown in Figure 21(a), the Gaussian fit of which [inset

of Figure 21(a)] gives its peak at 0.148 eV. In view of its

width (Dl/l¼ 0.23, where l and Dl are 8.3mm and 1.9mm,

respectively), this response originates from the HH bound-

to-LH bound transition as indicated by transition III of

Figure 20(b) (0.160 eV by calculation).

The dark current-voltage characteristic of the p-QDIP is

shown in Figure 22 and is used, along with the experi-

mentally measured noise current (in), to obtain the noise

gain (g) through the expression: g ¼ i2n=4eId, where Id is the

dark current. Figure 22(b) plots the noise gain as a function

of bias. Assuming that the photoconductive gain equals the

noise gain, the value of QE can be calculated from the

relationship between the responsivity (R) and gain: QE¼Rhn/eg, where hn is the photon energy. As shown in

Figure 22(c), QE reaches the maximum of 17% at

�0.6V. The specific detectivity is given by D�¼R(A�Df)1/2/in, where A is the device area and Df is the band-

width. The detectivity at 78K for the response peak at

5mm as a function of bias is shown in Figure 23, with a

maximum value of 1.8� 109 cmHz1/2/W at �0.4V.

It was reported that the bias at which the maximum

detectivity is obtained progressively varies with the tran-

sition type. The optimum operating bias of the bound-to-

continuum transition is close to 0V (101). A recently

demonstrated p-type Ge/Si(001) QDIP (109) based on the

bound-to-continuum transition also confirmed this opera-

tion. For the InAs/GaAs p-QDIP reported here, the bias-

dependent detectivity as shown in Figure 23 indicates the

bound-to-quasi-bound transition as the primary transition

contributing to the response, in agreement with the

prior analysis based on computed electronic structure

(Figure 20).

It should be noted that the demonstrated p-QDIP

does not employ a dark current blocking layer [see

Figure 20(a)]. This causes the dark current to be more

than three order of magnitudes higher than that of typical

n-QDIPs using a blocking layer (14). Although the current-

blocking layer also reduces the responsivity, the noise is

suppressedmore effectively, leading to increased signal-to-

noise ratio and detectivity. Based on the current, struc-

ture, further improvement on the QE is possible through

optimizing the transition type and enhancing the absorp-

tion. Barve et al. (101) reported the QE values for the

Figure 22. (a) Dark current density of the p-type QDIP at differ-ent temperatures. (b) and (c) (shown as insets) are the noise gainand QE, respectively. The noise gain is calculated by using exper-imentally measured noise current and dark current. QE isobtained by assuming that the photoconductive gain equals thenoise gain. After Reference 104.

Figure 23. The specific detectivity of the p-type QDIP for theresponse peak at 5mm, plotted as a function of the bias. After

Reference 104.

16 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:38 Page 17

bound-to-continuum, bound-to-quasi-bound, and bound-

to-bound transitions, the maximums of which are 2%,

6%, and 12%, respectively. It is therefore expecting an

increase in the QE of the p-QDIP by operating based on

the bound-to-bound hole transitions. The enhancement of

absorption could be attained by optimizing the QDs, such

as the number, density, and uniformity of the dot layers

and by employing a resonant cavity (105).

TERAHERTZ QUANTUM DOT ANDQUANTUM RINGDETECTORS

With an increasing interest in the terahertz region of the

spectrum (0.1–3.0THz), there is a need for terahertz detec-

tors exhibiting low dark current and high operating tem-

peratures for applications in imaging, communication,

security, and defense. The primary challenge in developing

terahertz detectors is the reduction of dark current

(because of thermal excitations) as the energy correspond-

ing to the terahertz detection mechanism is few meV.

Currently, terahertz detectors such as Ge BIB detectors

(110), photoconductors triggered by femtosecond laser

pulses (111), QWIP detectors (112), homojunction detec-

tors (113), heterojunction detectors (114), and thermal

detectors, such as bolometers and pyroelectric detectors,

are being studied. However, all of these detectors operate

at low temperatures. A typical detector structure, in which

the transitions leading to terahertz detection occur

between two electronic states with an energy difference

of DE (� 4.1meV for 1THz), would not be suitable for high-

temperature terahertz detection because the dark currents

from thermal excitations become dominant even at 77 K

because of the small DE. A T-QDIP structure (61), in which

the photocurrent is selectively collected while the dark

current is blocked, can be adjusted to obtain terahertz

response, thus offering a suitable platform for high-oper-

ating-temperature terahertz detectors. Huang et al. (58)

have reported a QDIP structure using intersubband detec-

tion, which has reduced complexity in the device structure.

Advancing quantum dot growth process further, a process

in which quantum dots are transformed into quantum

rings by postepitaxy thermal annealing has been reported

(115), also demonstrating the feasibility of terahertz detec-

tion using the electronics transitions in quantum rings

(116–119). The following discussion has more details about

those three approaches.

Tunneling Quantum Dot Structures for Terahertz Detection

Su et al. (61) have explored a resonant tunneling QDIP

device architecture demonstrating a detector sensitive to

6THz. The basic idea is same as other T-QDIPs discussed

previously, but in the terahertz T-QDIP structure, smaller

QDs are used to accommodate electronic transitions lead-

ing to terahertz detection. The structure and conduction

band profile of the structure reported by Su et al. (61) is

illustrated in Figure 24. In this approach, the transition

energy between the E1 state and the Er state corresponds

to 5THz radiation. The energy spacing for this transition

can be adjusted by changing the GaAs QW parameters

without changing the QDparameters. Hence, this provides

an advantage when the QD size and other parameters are

restrained by the growth technique. Flexibility for reduc-

ing the QD size would raise the upper energy states closer

to the Er state. However, to obtain a strong transition

between these two states, the carrier occupancy in the

E1 state should be increased. One possibility is to adjust

the doping concentration in the QDs so that the Fermi level

rises above the E1 state. Also, in this case the photores-

ponse can be enhanced by increasing the temperature,

similar to the T-QDIP detectors discussed before.

The structure of the MBE grown terahertz T-QDIP

detector contains self-organized In0.6Al0.4As QDs grown

on a GaAs layer and doped with Si. The stack of

Al0.3Ga0.7As/In0.1Ga0.9As/Al0.3Ga0.73As layers served as

the double-barrier system. The GaAs and AlGaAs layers

were grown at 610�C and the rest of the structure was

grown at 500�C. The top and bottom GaAs contact layers

were n-doped with Si to a level of 2� 1018 cm�3. To obtain a

transition leading to a response in terahertz region, the

excited states in the QD was pushed toward the resonant

state by forming smaller QDs. The QDswere doped to raise

the Fermi level so that photoexcitations take place from an

upper state in the QD to the resonant state. To reduce QD

size, In0.6Al0.4As was used instead of InGaAs because the

Figure 24. (a) Structure of a T-QDIPTHz detector. In0.6Al0.4AsQDs aren-dopedwith Si. The growth of smallerQDs comparedwith InAs orInGaAs QDs was achieved using InAlAs material. The QD size has been considerably reduced to 40A

�(height) and 130A

�(width).

(b) Conduction band profile of the THz T-QDIP under reverse applied bias along with the calculated bound state energies in the dots andwells. After Reference 61.

Epitaxial Quantum Dot Infrared Photodetectors 17

3GW8218 08/26/2014 20:43:39 Page 18

Al-containing islands (QDs) are smaller in size compared

with InAs (or InGaAs) islands because of the smaller

migration rate of Al adatoms on the growing surface

during epitaxy. The QD size was reported (61) to be 40A�

(height) and 130A�(width).

The dark current density of terahertz T-QDIP at differ-

ent temperatures is shown in Figure 25. At �2V bias, the

dark current density values are 6.5� 10�8, 7.9� 10�2, and

8.3A/cm2 at 4.2, 80, and 150K, respectively. At tempera-

tures above 80K, this terahertz T-QDIP detector showed a

lower dark current density compared with other terahertz

detectors (120, 121) operating in the 15-5THz (�20–60mm)

range irrespective of the bias voltage. However, compared

with the terahertz QD-based detector, which will be dis-

cussed in the next section, this T-QDIP has a lower dark

current density at low temperatures irrespective of the

bias. At high temperatures (above 80K), it showed a lower

dark current density only for bias voltages above 1V. This

variation of the dark current for the T-QDIP with bias

voltage is attributed to dark current blocking from the

tunneling barriers.

Referring to the conduction band profile of the structure

shown in Figure 24(b), this structure exhibits only two

bound states in the QD because of the reduced dot size. The

electrons excited from the upper bound state (E1) to the

resonant state (Er) lead to the photocurrent. The location of

the upper bound state is �24.6meV (5.5 THz, 54.6mm)

with respect to the resonant state and a response peak

centered around 5.5THz is theoretically expected. At the

frequency of 6 THz, the measured spectral response of the

detector at 80 and 150K showing responsivities of 6 and

0.6mA/W is shown in Figure 26. The sharp dip around

8.1THz (37mm) is due to the Reststrahlen band of GaAs,

which is present in all GaAs-based photon detectors (114,

121). The observed full width at half maximum (FWHM) of

the spectral response is�35meV. This spectral broadening

arises from the inhomogeneous size distribution of self-

organized QDs. Whereas this broadening is useful to

obtain a broad spectral response, it will weaken the reso-

nant conditions. For efficient resonant tunneling leading to

low dark current, tight resonant conditions are required.

Hence, the QD size and shape fluctuations need to be

minimum, which could be achieved using improved growth

techniques in the future. Furthermore, the measured

detectivity of the terahertz T-QDIP at 6 THz was

�5� 107 Jones at 80K under a bias of 1V. Although

this value of D� is not high enough for applications, higher

D� could be achieved through design and growth optimi-

zation. To achieve terahertz detection in 1–3THz region at

high temperatures, several issues need to be resolved such

as the growth of small QDs with reduced size fluctuation,

optimization of structure parameters, and tight resonant

conditions to maintain low dark current without reducing

the photocurrent.

Intersub-Band Terahertz QDIP

The theoretical calculations carried out in Reference 58

show that the In0.4Ga0.6As QDs in GaAs barriers exhibit

multiple confined electronic energy states. The transitions

of electrons from these energy states to the continuum

would lead to multiple response peaks, with at least one

falling in the terahertz region. Using those transitions,

Huang et al. (58) reported a In0.4Ga0.6As/GaAs QD-based

terahertz detector exhibiting two color characteristics with

response peaks in the 3–13 and 20–55mm (15–5.4THz)

ranges and operating up to 150K. With this detector

structure, two major disadvantages have been overcome:

One is the use of a simple heterostructure compared with

complicated terahertz detector structures (61, 122) and the

other is the higher temperature operation for the terahertz

region compared with terahertz QWIPs (112, 121, 122).

The schematics of the In0.4Ga0.6As/GaAs QDIP hetero-

structure grown by MBE on a (001)-oriented semi-insulat-

ing GaAs substrate is shown in Figure 27(a). The near-

pyramidal (61) shape QDs had an average base width of

21nm and a height of 5nm as estimated from atomic force

microscopy measurements (61). The active region con-

sisted of 20 periods of In0.4Ga0.6As/GaAs QD layers

3210–1–2–310

–8

10–6

10–4

10–2

100

150 K

80 KD

ark

Cu

rre

nt

De

nsity (

A/c

m2)

Bias Voltage (V)

4.2 K

Figure 25. The dark current density of THz T-QDIP as afunction of bias in the temperature range 4.2–150K. In thereported response range, the T-QDIP detector shows a lowerdark current density compared to other THz detectors operatingin the �20–60mm range. After Reference 61.

0

4

8

12

80 K, 2 V

80 K, 1 V

150 K, 1 V

20 40 5030

1014 6

Responsiv

ity (

mA

/W)

Frequency (THz)

Wavelength (µm)

Figure 26. Spectral responsivity of THz T-QDIP in the tempera-ture range 80–150K. The dip at 37mm is the Reststrahlen regionof GaAs. The calculated energy difference between the two energylevels leading to the response is 24.6meV (50.4mm). The peakresponsivity at 80 and 150K are 6 and 0.6mA/W, respectively.

After Reference 61.

18 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:39 Page 19

separated by a 500-A�-thick GaAs spacer layer. The QDs

and the spacer layers were grown continuously at 500�C.One of the critical parameters in this design is the doping

concentration because the electron occupancy in the bound

states is important for various transitions at different

temperatures. QDs were doped with silicon as the dopant

to a level of 1 electron per QD. The active layers were then

sandwiched between 0.2- and 0.8-mm-thick highly doped

(2� 1018 cm�3) GaAs layers, which act as top and bottom

contact layers, respectively.

Theoretical calculations based on the eight-band k.p

model (51) exhibited four bound states (E0, E1, E2, and E3)

for this structure as shown in Figure 27(b). The locations of

the energy states with respect to the GaAs band edge and

two sets of possible transitions in the QDs are also shown

in the figure. One set includes the transitions of electrons

from any of the states to the continuum, leading to

response peaks at 37.5, 17.6, 15.8, and 5.2THz (8, 17,

19, and 58mm), whereas the other set consists of the

transitions to the third excited state (E3) leading to

response peaks at 32.6, 12.6, and 10.7THz (9.2, 23.8,

and 28mm). The photocurrent observed from transitions

to the E3 state is weak, as these excited carriers are still

bound. Also, a few of the transitions in the two sets overlap

in the energy scale (wavelength) resulting in only a few

dominant response peaks.

The dark IV characteristics of this detector, as shown in

Figure 28, show a typical variation of current density with

the temperature. At high temperatures (80K and above),

the dark current density is lower compared with any other

detector (excluding T-QDIPs) responding in comparable

wavelength regions. At�2V bias, the dark current density

values are 6.5� 1�6, 5.5� 10�2, and 2.3A/cm2 at 4.2K,

80K, and 150K, respectively. This observed dark current

reduction is attributed to the high carrier confinement in

QDs. The calibrated responsivity at 5.4K, 80K, and 120K

is shown in Figure 29. The agreement of calculated and

experimental data confirms the validity of the k.pmodel for

pyramidal QDs. The peak centered around 11.5THz

(26mm) could be matched with the transition from the

E2 state to the E3 state [DE¼ 44meV (28mm)], whereas

the shoulder around 13.6THz (22mm) was assigned to

transition from the E1 state to the E3 state [DE¼ 52meV

(23.8mm)]. The peak centered around 6THz (50mm) was

assigned to the transitions from E3 state to the continuum

[DE¼ 21.3meV (58mm)]. In addition, there is a clear peak

at 37.5THz (8mm) as shown in Figure 29(b) (this has not

been measured at 5.4K, although expected). This peak

matched with the transition from the E0 state to the

continuum [DE¼ 155meV (8mm)]. The sharp dip at

8.1THz (�37mm) is a result of the longitudinal optical

phonon absorption in GaAs.

To confirm the terahertz mechanism in this structure,

the temperature dependence of the peaks at 11.5 and 6THz

can be considered. At low temperatures, carrier occupancy

in upper states, for example E3, is relatively low. Hence,

the transitions from upper states are weak, and as a result,

the peak at 6THz is relatively weak at 5.4K. At high

temperatures, because the carrier density in upper states

becomes high, an enhanced responsivity was expected for

the response peak originating from upper states (6THz

peak). This effect of the temperature is evident from

Figure 29. However, at 80K and 120K, the separation

of the two peaks (11.5 and 6THz) is not clear. The thermal

broadening of the energy states and the enhanced transi-

tions from upper states resulted in a combined broad peak

in the 15–5.26THz (20–57mm) range instead of two sepa-

rate peaks. Moreover, it is clear that the responsivity at

6THz has increased, whereas the responsivity at 11.5THz

has decreased drastically when the temperature was

increased from 5.4K to 120K. The specific detectivity

(D�), which is a measure of the signal-to-noise ratio of

the device, for the 37.5THz peak were reported to be

2� 109 Jones at 80K under �3V bias and 7.2� 107 Jones

at 120K under �2V bias. For the 7.5THz peak, the values

of D� were 1.3� 108 Jones at 80K under �3V bias,

Figure 27. (a) Structure of the terahertz QDIP detector. (b) Conduction band profile along with the energy states and possible electronictransitions between states. After Reference 58.

3210–1–2–310

–8

10–6

10–4

10–2

100

150 K

80 K

Dark

Curr

ent D

ensity (

A/c

m2)

Bias Voltage (V)

4.2 K

Figure 28. Dark current density of the terahertz QDIP at 4.2, 80,and 150K. After Reference 58.

Epitaxial Quantum Dot Infrared Photodetectors 19

3GW8218 08/26/2014 20:43:41 Page 20

2.8� 107 Jones at 120K under�2V bias, and 2� 107 Jones

at 150K under�2V bias. TheseD� values for the terahertztransition are larger compared with terahertz QWIPs. It

can be concluded that the 5THz detection up to 150K in

the QD-based detector have been achieved based on a

thermal-assisted mechanism in QDs.

Terahertz Quantum Ring Infrared Detectors

An observation of the intermixing and shape changes

during the growth of InAs QDs has been reported (123),

and growth of self-organized InGaAs QRs on GaAs sub-

strate was proposed by Lorke et al. (115). This process

starts with growth of QDs capped by a thin GaAs layer.

Then, a short annealing process is performed, which is the

critical step for transforming QDs to QRs. It is also

reported that the ring morphology is preserved even after

the growth of cladding layers on top of InGaAs islands.

Utilizing this process, a number of quantum ring (QR)

detectors, mainly sensitive to terahertz, has been recently

demonstrated (116–119). The QRs can be incorporated into

device structures, just like QDs in QDIP structures, in

which the electronic transitions from confined energy

states in QRs lead to the detection of infrared radiation.

Dai et al. (118) reported a terahertz QRIP sensitive in the

3–100THz range with a peak responsivity of 127mA/W at

23mm (detectivity of 2.3� 1011 cmHz1/2/W) at 10K. This

work was followed by the demonstration of another ter-

ahertz QRIP by the same authors (119). This detector

exhibits three peaks at 7.1, 3.5, and 2.3THz (42, 85, and

130mm, respectively) with the lowest frequency response

peak showing a cut-off of 1.7THz (175mm). Electronic

transitions, leading to the three peaks, are from the ground

state in QRs to the continuum (first peak) and to two

subbands (second and third peaks). Through an analysis,

it is shown that the observed spectral characteristics are

from QRs, but not from conventional QDs, which proves

the existence of QRs. This detector also showed a BLIP

temperature of 50K. In a recent article, Zavvari et al. (124)

presented a theoretical work on modeling QRIP detectors,

calculating their absorption, responsivity, and dark cur-

rent characteristics. Electronic energy states and wave-

functions are numerically calculated using effective mass

approximation and an intersubband absorption coefficient

of QRs is obtained. A clear dependence of spectral charac-

teristic on QR size, described by the height, inner diame-

ter, and outer diameter, has been observed, providing

various design options of QRs for covering the terahertz

spectrum.

In the work reported by Huang et al. (116), a tunnel

quantum ring detector sensitive to terahertz radiation in

the 3–8THz range up to 120K is demonstrated. The

height, inner diameter, and outer diameter of QRs are

approximately 2–4, 50, and 80nm, respectively. The detec-

tion mechanism is exactly same as that in a TQDIP struc-

ture (illustrated in Figure 24), and it involves transition of

electrons from the top-most quantum ring state to the

resonant state. As reported (116), the detector showed

low dark current density values (�5� 10�5, 4.7� 10�2

and 3.5� 10�1A/cm2 under a �1V bias, at 4.2, 80 and

300K, respectively). Three prominent response peaks were

observed at �6.5, 10, and 12.5THz up to 120K. The

detectivity values for the 10 THz response peak at 5 and

80K were 4.9� 109 and 9.5� 107 Jones, respectively, at a

bias of �2V.

Followed by the work by Huang et al., Bhowmick et al.

(117) have reported a terahertz QRIP based on intersub-

band electronic transitions in small QRs. With small QRs,

the top-most energy state was pushed close enough to the

edge of the GaAs barrier to achieve terahertz detection at

1.82THz. This is the lowest frequency (longest wave-

length) among all the terahertz detectors discussed in

this article. This detector containing multiple QR layers

was grown by MBE on semi-insulating (001) GaAs sub-

strates. The growth parameters for the initial InAs quan-

tum dots and the anneal conditions to convert them to

quantum rings were carefully tuned to produce the desired

smaller size of these nanostructures. First, a 0.8-mmGaAs

buffer layer was grown at 600�C. The substrate tempera-

ture was then lowered to 530�C and 2.1monolayers of InAs

20 30 40 50 60

14 12 10 8 6 50.0

0.5

1.0

1.5

2.0

20 30 40 50

15 100.00

0.05

0.10

Frequency (THz)

Wavelength (µm)

Re

sp

on

siv

ity (

A/W

)

7.5

Frequency (THz)

Wavelength (µm)

T = 5.4 K

V = –3 V

Responsiv

ity (

A/W

)

0.0

0.2

0.4

0.6

0.8

20 30 40 50 605 10 15

Frequency (THz)

Wavelength (µm)

Responsiv

ity (

A/W

)

Responsiv

ity (

A/W

)

GaAs-Like

Phonon

15 10 50.00

0.05

0.10

2550150 7.5

80 K, –3 V

120 K, –2 V

)b()a(

Figure 29. (a) Calibrated spectral response of the terahertz QD-based detector at 5.4K. The observed peak positions are indicated byvertical arrows. The response around 6THz (50mm) has been expanded in the inset. (b) Calibrated spectral response of the terahertz QD-based detector at 80 and 120K. The response below 20THz (15mm) is scaled with left vertical axis, while the response above 15THz (20mm)is scaled with the right vertical axis. After Reference 58.

20 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:41 Page 21

was deposited at a rate of 0.08 monolayer/s. Self-organized

quantum dots were formed after the growth of 1.8 mono-

layers of InAs wetting layer. A 10A�AlAs cap layer was

grown on the InAs islands at 530�C. Growth was inter-

rupted and the capped islands were annealed at 580�C for

30 s under an As2 flux to form quantum rings. The function

of the AlAs layer was to reduce the surface mobility of

group III atoms on the AlAs surface because AlAs has a

higher bonding strength than GaAs. Consequently, the

ring shape is better preserved during the annealing pro-

cess. The complete detector heterostructure, as shown in

Figure 30(a), consists of 10 quantum ring layers grown and

formed under the conditions described previously. The 50-

nm GaAs barrier layers shown in this figure are grown

immediately after the formation of the quantum rings. A

single Al0.2Ga0.8As barrier is inserted at the top to reduce

the dark current without substantially affecting the pho-

tocurrent. In Figure 30(b), an AFM image of small quan-

tum rings is shown. It was estimated that the density of the

rings is 1010 cm�2 and the rings have a height in the range

of 1.2–1.5 nm and inner and outer radii of 25 and 40nm,

respectively.

Carrier confinement in the rings primarily results from

their width and height. With the dimensions considered,

there is only one bound state in the potential well, which is

also the ground state. Then, the transition energy from the

ground state to the continuum in the ring corresponds to

the frequency range of 1–3THz. It is important to note that

a single confined state in QRs can be obtained only if they

are made small. The measured dark current character-

istics for this detector at 4.2K and 80K are shown in

Figure 31(a), whereas the dark current density values

observed here can be considered extremely low (�10�8A/

cm2 and 10�6A/cm2 at 4.2K for bias of �2V and þ2V,

respectively). The calibrated spectral response of this ter-

ahertz QRIP at 5.3K and 10K is also shown in Figure 31(b).

The peak at 1.82THz (165mm) corresponds to an energy of

7.52meV, which agrees well with the calculated ground

state-to-continuum transition energy for a ring with height

of 1.25nm. It is reported that the peak external quantum

efficiencies (internal quantum efficiency� gain) are �15

and 19% at �1 and 1V bias values, respectively. Moreover,

the values of specific detectivity at a bias of 1V are 1� 1016

and 3� 1015 Jones at 5.2K and 10K, respectively. These

Figure 30. (a) Heterostructure of the terahertz QRIP grown by MBE. It has 10 layers of InAs quantum rings in the active region and asingle Al0.2Ga0.8As barrier at the top. (b) AFM image of InAs quantum rings formed by postepitaxy thermal annealing of quantum dots.Inset shows a magnified image and dimensions. Modified after Reference 117.

43210–1–2–3–4–5

10–6

10–4

10–2

T = 80 K

Dark

Curr

ent D

ensity (

A/c

m2)

Bias Voltage (V)

T = 4.2 K

300250200150100500

5

10

15

20

25

100 200 3000

1

2

314 4 2

Frequency (THz)1

T = 10 K

V = 1 V

Re

sp

on

siv

ity (

A/W

)

Wavelength (µm)

14 4 2Frequency (THz)

1

T = 5.2 K

V = 1 V

Responsiv

ity (

A/W

)

Wavelength (µm)

)b()a(

Figure 31. (a) dark current characteristics and (b) measured spectral responsivity of the terahertz QRIP at 5.2K under 1V bias (insetshows the responsivity at 10K under 1V bias). Modified After Reference 117.

Epitaxial Quantum Dot Infrared Photodetectors 21

3GW8218 08/26/2014 20:43:41 Page 22

valuesofdetectivity,whicharehigher thanthedetectivityof

bolometers (�1014 Jones), prove that quantum ring detec-

tors are another alternative for terahertz detectors,

especially where detection at a specific terahertz frequency

is required.

IMPROVEMENT OF QDIP PERFORMANCE

Potential approaches to improve QDIP performance

include using multiple periods of QD active layers (75),

resonant cavity (RC) enhancement (78), and photonic crys-

tal cavity (PhC) (125) enhancement. To increase IR absorp-

tion, Chakrabarti et al. (75) have reported an MBE grown

QDIP consisting of 70 periods of QD active layers. A 400A�

Al0.3Ga0.7As current-blocking barrier was incorporated

underneath the GaAs top-contact layer. A low dark current

density (1.83� 10�2Acm�2 at 175K under a �2.0V bias)

was observed at high operating temperatures. The

response in the 2–6mm range, peaking at 3.9mm, was

caused by transitions of electrons from the QD ground

state to the continuum state. At 150K, a peak responsivity

of 0.12A/W was observed. A maximum detectivity of

1011 cmHz1/2/W at 100K for �2V, which is reasonably

high at this temperature, was also reported. The improve-

ment in detector performance was a result of the increased

number of QD active layers.

Using photonic crystals as amicrocavity resonator (125)

has demonstrated a significant improvement in the

responsivity, conversion efficiency, and detectivity for a

DWELL detector. A two-dimensional array of holes (PhC)

with a 2–3mmdiameter and a lattice spacing of 2.4mmwas

fabricated on the DWELL structure using e-beam lithog-

raphy. A schematic diagram and an image of a hexagonal

PhC are shown in Figure 32(a) and (b). PhCs provide a

photonic bandgap for normal incidence radiation. By intro-

ducing a defect into the crystal, a cavity can be built, which

operates in the wavelength range corresponding to the

photonic bandgap. The above parameters were chosen

(125) to obtain resonant cavity effects at 8.1mm. Two

structures, a DWELL detector and a PhC-DWELL detector

(125), were tested and performance was compared to deter-

mine the effects of PhC. The DWELL detector exhibited

two-color response characteristics with peaks at 6 and

10mm. The 6-mmpeak (dominant at a low bias) was caused

by transitions from the QD ground state to the continuum,

whereas the 10-mm peak (dominant at high bias) origi-

nated from transitions from the QD ground state to a

bound state in the QW. These electronic transitions and

corresponding response peaks are similar to those of the

DWELL detectors reported previously in this chapter. An

increase in photocurrent by an order of magnitude was

reported for PhC-DWELL, whereas both DWELL and

PhC-DWELL had the same dark current. A comparison

of the responsivity of a PhC-DWELL and a standard

DWELL is shown in Figure 32(c). It should be noted

that the spectra have been scaled for comparison of spec-

tral shape and the y-axis does not represent the true

relative response between the two spectra. At 9mm and

78K under a �2.6V bias, the PhC-DWELL detector had a

conversion efficiency of 95%, as opposed to the conversion

efficiency of 7.5% for the DWELL detector. Using conven-

tional optical lithography, PhCs with 2–3mm hole diame-

ter can be easily incorporated into FPA fabrication.

Similar to PhC-DWELL, an RC-DWELL, in which the

active QD region was placed in a Fabry–P�erot resonant

cavity, was reported by Attaluri et al. (78) This allows

radiation to pass multiple times through the active region

within the cavity. The RC was designed to increase the

optical field strength at the peak wavelength of a standard

DWELL detector. The RC-DWELL consists of a regular

DWELL (including top-and bottom-contact layers) grown

on a distributed Bragg reflector (bottom reflector), as

shown in Figure 33(a). The top reflector was the semi-

conductor-air interface (top-contact layer), which has

about 30% reflectivity. The DWELL detector showed two

response peaks at 6 and 10mm. The resonant optical cavity

was active at 9.5mm, consequently, the RC-DWELL

showed an enhance response at 9.5mm (see Figure 33(b)).

A peak responsivity of 0.76A/W, detectivity of

Figure 32. (a) Photonic crystal resonant cavity comprising of a hexagonal pattern of air holes. (b) Image of a PhC defined in a DWELLdetector. (c) A comparison of the spectral responsivity of a PhC-DWELLand a standardDWELLdetector at 50Kunder a�3V bias. It shouldbe noted that the spectra have been scaled for comparison of spectral shape and the y-axis does not represent the true relative responsebetween the two spectra. After Reference 125.

22 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:42 Page 23

1.4� 1010 cmHz1/2/W, and 10%conversion efficiencyat 1.4V

bias and 77K were reported for RC-DWELL, whereas

DWELL showed only 1.25% conversion efficiency at

10mm under the same conditions. The detectivity of the

RC-DWELL was 3� 109 cmHz1/2/W at 77K under a 1.2V

bias, showing a factor of 3 improvements comparedwith the

conventional DWELL detector. Another important feature

of incorporating an RC into a DWELL structure is the

increase in photocurrent without increasing the dark

current.

By controlling the excited state energy between the CE

barriers and the continuum level, thin AlGaAs barrier

layers in the quantum dots in a well heterostructure

enhanced the quantum confinement of carriers in the

excited energy level, while maintaining high escape prob-

ability (14). A reported postgrowth technique for improv-

ing the peak detectivity is to implant the structure with a

low energy light ions(H�) (126). A proposed simulation of a

resonant cavity enhanced quantum dot infrared photo-

detector consisting of a conventional n-i-n QDIP sand-

wiched by a bottom GaAs/Al2O3 distributed Bragg

reflector and a top mirror of Ge/SiO2 subwavelength gra-

ting predicts an enhanced external quantum efficiency

more than 50% (127).

PRESENT PERFORMANCE AND CAPABILITIES OF QDIPs

Pal et al. (128) reported a superlattice QDIPwith 50-period

InAs/GaAs QD active layers and an AlGaAs current-block-

ing layer to suppress the dark current. The spectral

responsivity and detectivity of this detector is shown in

Figure 34(a) and (b). Gunapala et al. (33) have demon-

strated up to 1A/W response at 77K for a DWELL detector

coupled with a grating. A two-color DWELL detector with

response peaks at 3.2 and 4.1mm operating up to room

Figure 33. (a) Schematic structure of a RC-DWELL detector with eight-period QD active layers (InAs/In0.15Ga0.85As). (b) A comparison ofthe spectral responsivity of a RCDWELL detector and a standard DWELL (without the photon reflector) detector at 30K under a �1.8Vbias. After Reference 78.

Figure 34. (a) Spectral responsivity of a QDIP with a current-blocking layer composed of an Al0.3Ga0.7As layer. This enabled operation upto 220K. (b) Variation of the peak detectivity with bias. The peak responsivity as a function of bias at 100K is shown in the inset. After

Reference 128.

Epitaxial Quantum Dot Infrared Photodetectors 23

3GW8218 08/26/2014 20:43:42 Page 24

temperature was demonstrated by Lim et al. (39). Also,

there are few other reports demonstrating QDIPs (6, 75)

operating at high temperatures. A historical analysis of

the QDIP development shows rapid progression toward

high-performance detectors capable of room-temperature

operation.

In a recent investigation by Jayaweera et al. (129), a

new concept for developing photon detectors, based on the

displacement currents in a QD-embedded dielectric media,

operating at room temperature has been reported. A pre-

liminary detector structure along with its photoresponse

results was also reported. The detector structure consists

of PbS QDs embedded in a dielectric thin film, which is

sandwiched between two conducting glass contacts. The

dielectric thin film was made of paraffin wax, which has a

high dielectric constant. The incident photons are absorbed

in QDs, generating an electron-hole pair. When an electric

field applied across the contacts, the electrons drift toward

the positive contact, while holes drift toward the negative

contact. Because carrier transport through the dielectric

medium is not possible, there will be a charge separation in

the medium.

This polarizability with photon absorption, which

changes the capacitance of the device, is the key detection

mechanism in the structure proposed by Jayaweera et al.

QDs in a dielectric medium has shown several orders of

magnitude higher polarizability than atoms and mole-

cules. The incident radiation is modulated and the photo-

generated displacement current is measured under an

applied bias across the contacts. The displacement cur-

rent is a direct measurement of the incident radiation

intensity and does not contain any contribution from the

dark current. Hence, this approach is suitable for devel-

oping room-temperature detectors. At 300K and 40V

bias, the responsitivity and the detectivity at the peak

absorption (lP¼ 540 nm) were found to be 195V/W and

3� 108 cmHz1/2/W, respectively. As an advantage, sophis-

ticated growth and fabrication techniques are not

required for this type of detector development. However,

in this approach the incident light has to be modulated in

order to obtain a photoresponse. A closely related device

structure, which operates in photocurrent mode, has

been reported by Konstantatos et al. (130). The device

structure consists of spin-coating colloidal PbS QDs fab-

ricated onto interdigitated gold electrodes. Detectivity in

the order of 1013 cmHz1/2/W has been reported at 1.3mmat room temperature. A high photo-conductive gain

(�10,000) and consequently a high responsivity of

103A/W were observed at room temperature.

QUANTUM DOT FOCAL PLANE ARRAYS

The development of FPAs using successful single-element

detector structures has remarkably increased potential

applications in IR imaging. Multi-color FPAs would lead

to additional detection capabilities such as the construc-

tion of a true thermal map of a target. FPAs operating in

atmospheric windows (3–5 and 8–14mm) are required for a

number of applications including night vision cameras and

missile tracking, whereas FPAs operating in the FIR

region are particularly suited for applications in astron-

omy and space situational awareness (131). Most high-

performance FPAs developed so far have been based on

InSb, MCT, and QWIPs; however, QD-based FPAs have

also attracted the attention of infrared community because

of the development of successful single-element QDIPs

(6, 39, 128). Krishna et al. (132) have reported a

320� 256 pixel two-color FPA using an InAs/InGaAs

DWELL detector. According to Krishna et al., one of the

major drawbacks in developing QD-based FPAs is the

growth of thick active layers, providing sufficient absorp-

tion of radiation without causing misfit dislocations. The

growth of self-assembled QDs needs sufficient strain in the

QD regions, reducing the possibility of growing thick QD

layers. The two-color DWELL detector, which was used for

the FPA, showed wavelength selectivity with applied bias,

where a 5.5-mm peak was enabled at low bias and 8- to 10-

mm peak was activated by high biases. This has a peak

responsivity of 1A/W and D� of 2–7� 1010 cmHz1/2/W.

Thermal imaging using the FPA was demonstrated at

80K with different optical filters (3–5 and 8–12mm). More-

over, the operability of the FPA was reported to be greater

than 99%, and the noise equivalent temperature difference

(NEDT) was estimated to be less than 100mK for f/1 (for 3–

5mm) and f/2 (5–9mm) optics.

Gunapala et al. (33) have demonstrated a 640� 512

pixel QD-based IR imaging FPA operating in the LWIR

region. The detector was modeled on a DWELL structure,

where In-GaAs QDs were placed in a GaAs/AlGaAs QW.

The 30-stack DWELL detector has shown a peak absorp-

tion quantum efficiency of 2.7%. Under normal incidence

configuration, the detector had a much stronger (almost

one order of magnitude) responsivity compared to the

responsivity of a typical QWIP under 45� incidence config-

uration. In addition, it was found that the 45� incidence

responsivity of the detector (�1A/W) was four to five times

stronger than the normal incidence responsivity, as shown

in Figure 35(a). DWELL detectors can be fabricated with a

reflection grating to couple normal incidence light to the z-

polarization sensitive absorption mechanism in the

DWELL structure. A DWELL detector (reported by Guna-

pala et al. (33)) fabricated with a reflecting grating showed

a factor of 3–4 enhancement in the responsivity, as shown

in Figure 35(b). Peak detectivity of 1� 1010 cmHz1/2/W at

8.1mm and 77K was reported. Using this detector struc-

ture, a 640� 512 pixel FPA has been fabricated. Addition-

ally, a LWIR imagery system with a noise-equivalent

temperature difference of 40mK at 60K showing a

high uniformity was constructed. A 3-inch GaAs wafer

with twelve 640� 512 pixels QDIP FPAs is shown in

Figure 36(a), while an image of a human target

taken with the 640� 512 pixels FPA camera is shown in

Figure 36(b).

In a recent publication, Razeghi’s group demonstrated

(34) a 320� 256 pixel MIR focal plane array operating at

temperatures up to 200K. The FPA was based on a

DWELL structure (39), which consists of InAs QDs

imbedded in a InGaAs/InAlAs QW. It was reported that

the detector has two color characteristics up to room

24 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:42 Page 25

temperature with response peaks at 3.2 and 4.1mm, as

shown in Figure 37. The photoresponse peak at 4.1mm

originated from bound-to-bound intersubband transitions

between delocalizedmixed states in the QD andQWunlike

the bound-to-bound transitions between pure QD states

and QW states for DWELL structures. The detector had a

peak responsivity of 34mA/W and maximum D� of

2.8� 1011 cmHz1/2/W at 120K, whereas a D� of 6� 107

cmHz1/2/W was obtained at 300K. The DWELL FPA

showed imaging capabilities at 120K and 200K. A conver-

sion efficiency of 1.1% and a noise equivalent temperature

difference of 344mK at 120K were observed. The FPA

camera is capable of imaging human targets up to 150K.

The images of a human target and a hot soldering iron

taken with the FPA camera at 130K and 200K, respec-

tively, are shown in Figure 38. Based on these results, it

Figure 35. (a) A comparison between the spectral responsivity of a DWELL detector measured under normal incidence and 45� incidenceconfigurations. (b) Normal incidence spectral responsivity with and without reflection gratings. After Reference 33.

Figure 36. (a) A 3-in. GaAs wafer with twelve 640� 512 pixels QDIP FPAs. (b) An image of a human target takenwith the 640�512 pixelsQDIP FPA camera. After Reference 33.

Figure 37. (a) Spectral responsivity of a DWELL detector at different temperatures under a (a) �1V bias and (b) �5V bias (response atroom temperature was obtained under a �2V bias). The variation of the response at room temperature is shown in the inset. After

Reference 39.

Epitaxial Quantum Dot Infrared Photodetectors 25

3GW8218 08/26/2014 20:43:48 Page 26

can be concluded that the development of QD-based FPAs

is feasible, even though the present QDIP FPAs show

lower overall performance as compared to MCT and

QWIP FPAs. Major improvements in QDIP FPAs include

high responsivity and low dark current at high tempera-

tures, improved growth quality, and uniformity in detector

material for large wafer sizes and device processing.

CONCLUSION

The electronic transitions between energy states in the QD

and QW give rise to multicolor response in DWELL detec-

tor structures. Response peaks experimentally obtained

from each detector correspond to the energy spacing cal-

culated by theoretical models. By changing the well width

and the size of the QD, detectors can be designed to operate

at different wavelength regions depending on the applica-

tion. The operating wavelength can be selected by varying

the applied bias. Advantages of DWELL detectors over

typical n-type QWIPs include their ability to operate under

normal incidence and at high temperatures particularly in

the VLWIR/FIR region. In a different approach, T-QDIP

detectors designed for room temperature operation and

terahertz detection were reported. As evident from the

results, T-QDIPs exhibit lower dark current, and higher

operating temperatures compared to typical QDIPs, which

were made possible by the incorporation of double-barriers

into the structure. A 17-mm T-QDIP detector that can

operate at room temperature and a 6 terahertz detector

operating at 150K were demonstrated. T-QDIPs are pho-

ton detectors, which are inherently fast, and do not use

slower thermalization processes, which is the primary

detection mechanism in thermal detectors such as bolom-

eter or TGS. In comparison with the Si BIB detectors, the

main advantage will be in the increase in operating tem-

perature to 77K or higher, which will greatly reduce the

cooling requirements. This would be an important feature

for space-based applications where moving from 20K to

higher temperatures is a major design advantage. Fur-

thermore, T-QDIP structures open up a wide range of

possible modifications and designing options to aid in

wavelength tuning and performance optimization.

Superlattice QDIP is another successful device architec-

ture for the bias-selectable dual-band detectors discussed

in this article. A SL-QDIP detector, showing the proof of

concept, exhibits two response peaks at 4.9 and 7.4mm,

which can be selected by changing the applied bias voltage.

In order to increase the quantum efficiency, a strategy

discussed here includes the use of p-type doped quantum

dots. Another region of the spectrum covered in this article

is the terahertz region and three different types of tera-

hertz structures (T-QDIP, intersubband QDIP, and QRIP)

were discussed. Among the three terahertz structures,

QRIP which utilized quantum rings (instead of quantum

dots) as light-absorbing domains showed the shortest

detection frequency (1.8THz). Promising results have

been observed for QD-based FPAs, even though the overall

performance is low compared with MCT and QWIP FPAs.

ACKNOWLEDGMENT

This work is supported in part by the National Science

Foundation under Grant ECCS 1232184. The authors

acknowledge the contributions of Prof. P. Bhattacharya

and his group at the University of Michigan and Prof. S.

Krishna and his group at University of New Mexico, and

the group members at GSU, especially Dr. Y. Lao and

Mr. S. Wolde.

BIBLIOGRAPHY

1. A. Y. Cho. Film Deposition by Molecular-Beam Techniques.

J. Vac. Sci. Tech. 1971, 8, S31–S38.

2. B. J. Riel. An Introduction to Self-Assembled QuantumDots.

Am. J. Phys. 2008, 76, 750–757.

3. Y. Arakawa andH. Sakaki.Multidimensional QuantumWell

Laser and Temperature Dependence of its Threshold Cur-

rent. Appl. Phys. Lett. 1982, 40, 939–941.

4. A. D. Stiff, S. Krishna, Pallab Bhattacharya, and S. W.

Kennerly. Normal-Incidence, High-Temperature, Mid-Infra-

red, InAs-GaAs Vertical Quantum-Dot Infrared Photo-

detector. IEEE J. Quantum Electron. 2001, 37, 1412–1419.

5. Z. Ye, J. C. Campbell, Z. Chen, O. Baklenov, E. T. Kim, I.

Mukhametzhanov, J. Tie, and A. Madhukar. High-

Figure 38. Images taken with 320�256 pixel QDIP FPA camera at (a) 130K and (b) 200K. The maximum operating temperature of thearray is 200K. After Reference 34.

26 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:48 Page 27

Performance InAs/GaAs Quantum Dots Infrared Photo-

detector With/Without Al0.2Ga0.8As Blocking Layers, in

2001 MRS Fall Meeting, vol. 692; 2001 H9.17.1.

6. E.-Tae Kim, A. Madhukar, Z. Ye, and J. C. Campbell. High

Detectivity InAs Quantum Dot Infrared Photodetectors.

Appl. Phys. Lett. 2004, 84, 3277–3279.

7. Z. Ye, J. C. Campbell, Z. Chen, E.-Tae Kim, and A.

Madhukar. InAs Quantum Dot Infrared Photodetectors

with In0.15Ga0.85As Strain-Relief Cap Layers. J. Appl.

Phys. 2002, 92, 7462–7468.

8. S. Chakrabarti, A. D. Stiff-Roberts, P. Bhattacharya, and S.

W. Kennerly. High Responsivity Alas/InAs/GaAs Superlat-

tice Quantum Dot Infrared Photodetector. Electron. Lett.

2004, 40, 197–198.

9. D. Z.-Y. Ting, S. V. Bandara, S. D. Gunapala, J. M. Mumolo,

S. A. Keo, C. J. Hill, J. K. Liu, E. R. Blazejewski, B. Rafol, and

Y.-ChungChang. SubmonolayerQuantumDot Infrared Pho-

todetector. Appl. Phys. Lett. 2009, 94, 111107–111107-3.

10. P. Bhattacharya, X. H. Su, S. Chakrabarti, G. Ariyawansa,

and A. G. U. Perera. Characteristics of a Tunneling Quan-

tum-Dot Infrared Photodetector Operating at Room Temper-

ature. Appl. Phys. Lett. 2005, 86, 191106–191103.

11. X. H. Su, S. Chakrabarti, A. D. Stiff-Roberts, J. Singh, and P.

Bhattacharya. Quantum Dot Infrared Photodetector Design

Based on Double-Barrier Resonant Tunnelling. Electron.

Lett. 2004, 40, 1082–1083.

12. K. Sanjay. Quantum Dots-In-A-Well Infrared Photodetec-

tors. J. Phys. D: Appl. Phys. 2005, 38, 2142–2150.

13. Y. D. Sharma, M. N. Kutty, R. V. Shenoi, Ajit V. Barve, S.

Myers, J. Shao, E. Plis, S. Lee, S. Noh, and S. Krishna.

Investigation of multistack InAs/InGaAs/GaAs self-

assembled quantum dots-in-double-well structures

for infrared detectors. J. Vac. Sci. Tech. B 2010, 28,

C3G1–C3G7.

14. A. V. Barve, S. Sengupta, J. O. Kim, Y. D. Sharma, S.

Adhikary, T. J. Rotter, S. J. Lee, Y. H. Kim, and S. Krishna.

Confinement Enhancing Barriers for High Performance

Quantum Dots-In-A-Well Infrared Detectors. Appl. Phys.

Lett. 2011, 99, 191110–191113.

15. S. Chakrabarti, S. Adhikary, N. Halder, Y. Aytac, and A. G.

U. Perera. High-performance, Long-Wave (�10.2 Mm)

InGaAs/GaAs Quantum Dot Infrared Photodetector with

Quaternary In0.21Al0.21Ga0.58As Capping. Appl. Phys.

Lett. 2011, 99, 181102–181102–3.

16. S. S. Li. Multi-Color, Broadband Quantum Well Infrared

Photodetectors for Mid-, Long-, and Very Long-Wavelength

Infrared Applications. Int. J. High Speed Electron. Syst.

2002, 12, 761–801.

17. S. Krishna, S. Raghavan, G. von Winckel, A. Stintz, G.

Ariyawansa, S. G. Matsik, and A. G. U. Perera. Three-Color

(Lp1�3.8 mm,Lp2�8.5 mm, andLp3�23.2 mm) InAs/InGaAs

Quantum-Dots-In-A-Well Detector. Appl. Phys. Lett. 2003,

83, 2745–2747.

18. S. Chakrabarti, X. H. Su, P. Bhattacharya, G. Ariyawansa,

and A. G. U. Perera. Characteristics of a Multicolor InGaAs-

GaAs Quantum-Dot Infrared Photodetector. IEEE Photon.

Tech. Lett. 2005, 17, 178–180.

19. G. Ariyawansa, A. G. U. Perera, X. H. Su, S. Chakrabarti,

and P. Bhattacharya. Multi-Color Tunneling Quantum Dot

Infrared Photodetectors Operating at Room Temperature.

Infrared Phys. Tech. 2007, 50, 156–161.

20. A. Rogalski. Infrared Detectors: Status and Trends. Prog.

Quantum Electron. 2003, 27, 59–210.

21. J. Urayama, T. B. Norris, J. Singh, and P. Bhattacharya.

Observation of Phonon Bottleneck in Quantum Dot

Electronic Relaxation. Phys. Rev. Lett. 2001, 86, 4930–4933.

22. V. Ryzhii. The Theory Of Quantum-Dot Infrared Phototran-

sistors. Semiconduct. Sci. Tech. 1996, 11, 759–765.

23. R. Leon, G.M. Swift, B.Magness,W. A. Taylor, Y. S. Tang, K.

L. Wang, P. Dowd, and Y. H. Zhang. Changes In Lumines-

cence Emission Induced by Proton Irradiation: InGaAs/GaAs

Quantum Wells and Quantum Dots. Appl. Phys. Lett. 2000,

76, 2074–2076.

24. B. Aslan, H. C. Liu, M. Korkusinski, S.-J. Cheng, and P.

Hawrylak. Response Spectra From Mid- to Far-Infrared,

Polarization Behaviors, and Effects of Electron Numbers

in Quantum-Dot Photodetectors. Appl. Phys. Lett. 2003,

82, 630–632.

25. D. Pal, L. Chen, and E. Towe. Intersublevel Photoresponse of

(In,Ga)As/GaAs Quantum-Dot Photodetectors: Polarization

and Temperature Dependence. Appl. Phys. Lett. 2003, 83,

4634–4636.

26. P. Martyniuk, S. Krishna, and A. Rogalski. Assessment of

QuantumDot Infrared Photodetectors forHighTemperature

Operation. J. Appl. Phys. 2008, 104, 034314–034314-6.

27. S. S. Kim, Y. S. Choi, Kibum Kim, J. H. Kim, and S. Im.

Fabrication Of P-Pentacene/N-Si Organic Photodiodes and

Characterization of Their Photoelectric Properties. Appl.

Phys. Lett. 2003, 82, 639–641.

28. B. Kochman, A. D. Stiff-Roberts, S. Chakrabarti, J. D. Phil-

lips, S. Krishna, J. Singh, and P. Bhattacharya. Absorption,

Carrier Lifetime, and Gain in InAs-GaAs Quantum-Dot

Infrared Photodetectors. IEEE J. Quantum Electron. 2003,

39, 459–467.

29. H. C. Liu, M. Gao, J. McCaffrey, Z. R. Wasilewski, and S.

Fafard. Quantum Dot Infrared Photodetectors. Appl. Phys.

Lett. 2001, 78, 79–81.

30. L. Jiang, S. S. Li, N. -Tze Y., J.-Inn Chyi, C. E. Ross, andK. S.

Jones. In0.6Ga0.4As/GaAs Quantum-Dot Infrared Photo-

detector with Operating Temperature up to 260 K. Appl.

Phys. Lett. 2003, 82, 1986–1988.

31. S. Raghavan, P. Rotella, A. Stintz, B. Fuchs, S. Krishna, C.

Morath, D. A. Cardimona, and S.W.Kennerly. High-Respon-

sivity, Normal-Incidence Long-Wave Infrared (L�7.2 Mm)

InAs/In0.15Ga0.85As Dots-In-A-Well Detector. Appl. Phys.

Lett. 2002, 81, 1369–1371.

32. 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. Demonstration of a 320 � 256 Two-Color

Focal Plane Array Using InAs/InGaAs Quantum Dots in

Well Detectors. Appl. Phys. Lett. 2005, 193501–193501-3.

33. S. D. Gunapala, S. V. Bandara, C. J. Hill, D. Z. Ting, J. K. Liu,

S. B. Rafol, E. R. Blazejewski, J. M. Mumolo, S. A. Keo, S.

Krishna, Y. C. Chang, and C. A. Shott. Demonstration of

640 � 512 Pixels Long-Wavelength Infrared (LWIR) Quan-

tumDot Infrared Photodetector (QDIP) Imaging Focal Plane

Array. Infrared Phys. Tech. 2007, 50, 149–155.

34. S. Tsao, H. Lim, W. Zhang, and M. Razeghi. High Operating

Temperature 320 � 256 Middle-Wavelength Infrared Focal

Plane Array Imaging Based on an InAs/InGaAs/Inalas/Inp

QuantumDot InfraredPhotodetector.Appl. Phys. Lett. 2007,

90, 201109–201109-3.

35. A. Sergeev, V. Mitin, and M. Stroscio. Quantum-Dot Photo-

detector Operating at Room Temperatures: Diffusion-

Limited Capture. Phys. B: Condens. Matter. 2002, 316–317,

369–372.

Epitaxial Quantum Dot Infrared Photodetectors 27

3GW8218 08/26/2014 20:43:49 Page 28

36. J. Phillips, K. Kamath, and P. Bhattacharya. Far-Infrared

Photoconductivity in Self-Organized InAs Quantum Dots.

Appl. Phys. Lett. 1998, 72, 2020–2022.

37. S. Maimon, E. Finkman, G. Bahir, S. E. Schacham, J. M.

Garcia, and P. M. Petroff. Intersublevel Transitions in InAs/

GaAs Quantum Dots Infrared Photodetectors. Appl. Phys.

Lett. 1998, 73, 2003–2005.

38. D. Pan, E. Towe, and S. Kennerly. Normal-Incidence Inter-

subband (In, Ga)As/GaAs Quantum Dot Infrared Photode-

tectors. Appl. Phys. Lett. 1998, 73, 1937–1939.

39. H. Lim, S. Tsao, W. Zhang, and M. Razeghi. High-Perform-

ance InAs Quantum-Dot Infrared Photodetectors Grown on

Inp Substrate Operating at Room Temperature. Appl. Phys.

Lett. 2007, 90, 131112–131113.

40. M. R. Matthews, R. J. Steed, M. D. Frogley, C. C. Phillips, R.

S. Attaluri, and S. Krishna. Transient Photoconductivity

Measurements of Carrier Lifetimes in an InAs/

In0.15Ga0.85As Dots-In-A-Well Detector. Appl. Phys. Lett.

2007, 90, 103519–103519-3.

41. C.-H. Lin, C.-Y. Yu, C.-Y. Peng, W. S. Ho, and C. W. Liu.

Broadband SiGe/Si Quantum Dot Infrared Photodetectors.

J. Appl. Phys. 2007, 101, 033117–033117-3.

42. S. Tong, H.-Jun Kim, and K. L. Wang. Normal Incidence

Intersubband Photoresponse from Phosphorus D-Doped Ge

Dots. Appl. Phys. Lett. 2005, 87, 081104–081104-3.

43. D. Cha, M. Ogawa, C. Chen, S. Kim, J. Lee, K. L. Wang, J.

Wang, and T. P. Russell. Intersubband Absorption In P-Type

Si1–XGex Quantum Dots on Pre-Patterned Si Substrates

Made by a Diblock Copolymer Process. J. Cryst. Growth

2007, 301–302, 833–836.

44. A. Vardi, N. Akopian, G. Bahir, L. Doyennette, M. Tcher-

nycheva, L. Nevou, F. H. Julien, F. Guillot, and E. Monroy.

Room Temperature Demonstration of Gan/Aln Quantum

Dot Intraband Infrared Photodetector at Fiber-Optics Com-

munication Wavelength. Appl. Phys. Lett. 2006, 88,

143101–143101-3.

45. A. Babinski, G. Ortner, S. Raymond, M. Potemski, M. Bayer,

W. Sheng, P. Hawrylak, Z. Wasilewski, S. Fafard, and A.

Forchel. Ground-State Emission from a Single InAs/GaAs

Self-Assembled Quantum Dot Structure in Ultrahigh Mag-

netic Fields. Phys. Rev. B 2006, 74, 075310–075310-6.

46. R. J. A. Hill, A. Patan�e, P. C. Main, L. Eaves, B. Gustafson,

M.Henini, S. Tarucha, andD. G. Austing.Magnetotunneling

Spectroscopy of an Individual Quantum Dot in a Gated

Tunnel Diode. Appl. Phys. Lett. 2001, 79, 3275–3277.

47. D. Loss and D. P. DiVincenzo. Quantum Computation with

Quantum Dots. Phys. Rev. A 1998, 57, 120–126.

48. A. Imamoglu, D. D. Awschalom, G. Burkard, D. P. DiVin-

cenzo, D. Loss, M. Sherwin, and A. Small. Quantum Infor-

mation Processing Using Quantum Dot Spins and Cavity

Qed. Phys. Rev. Lett. 1999, 83, 4204–4207.

49. M. N. Leuenberger, M. E. Flatt�e, and D. D. Awschalom.

Teleportation Of Electronic Many-Qubit States Encoded in

theElectron Spin ofQuantumDots Via Single Photons.Phys.

Rev. Lett. 2005, 94, 107401–107401-4.

50. V. V. Dobrovitski, J. M. Taylor, andM. D. Lukin. Long-Lived

Memory for Electronic Spin in a Quantum Dot: Numerical

Analysis. Phys. Rev. B 2006, 73, 245318–245318-6.

51. H. Jiang and J. Singh. Strain Distribution and Electronic

Spectra of InAs/GaAs Self-Assembled Dots: An Eight-Band

Study. Phys. Rev. B 1997, 56, 4696–4701.

52. A. Amtout, S. Raghavan, P. Rotella, G. von Winckel,

A. Stintz, and S. Krishna. Theoretical Modeling and

Experimental Characterization of InAs/InGaAs Quantum

Dots in a Well Detector. J. Appl. Phys. 2004, 96, 3782–3786.

53. M Califano and P Harrison. Presentation and Experimental

Validation of a Single-Band, Constant-Potential Model for

Self-Assembled InAs/GaAs Quantum Dots. Phys. Rev. B

2000, 61, 10959–10965.

54. K. H. Schmidt, G. Medeiros-Ribeiro, M. Oestreich, P. M.

Petroff, and G. H. D€ohler. Carrier Relaxation and Electronic

Structure in InAs Self-Assembled QuantumDots. Phys. Rev.

B 1996, 54, 11346–11353.

55. E. P. Pokatilov, V. A. Fonoberov, V. M. Fomin, and J. T.

Devreese. Electron and Hole States in Quantum Dot Quan-

tum Wells Within a Spherical Eight-Band Model. Phys. Rev.

B 2001, 64, 245329–245329-7.

56. Y. Yin Lin and J. Singh. Self-Assembled Quantum Dots: A

Study of Strain Energy and Intersubband Transitions.

J. Appl. Phys. 2002, 92, 6205–6210.

57. G. Ariyawansa, A. G. U. Perera, G. S. Raghavan, G. Von

Winckel, A. Stintz, and S. Krishna. Effect of Well Width on

Three-Color Quantum Dots-In-A-Well Infrared Detectors.

IEEE Photon. Tech. Lett. 2005, 17, 1064–1066.

58. G. Huang, J. Yang, P. Bhattacharya, G. Ariyawansa, and A.

G. U. Perera. A Multicolor Quantum Dot Intersublevel

Detector with Photoresponse in the Terahertz Range.

Appl. Phys. Lett. 2008, 92, 011117–011113.

59. V. Ryzhii, I. Khmyrova, V. Pipa, V. Mitin, and M. Willander.

Device Model for Quantum Dot Infrared Photodetectors and

Their Dark-Current Characteristics.Semiconduct. Sci. Tech.

2001, 16, 331–338.

60. V. Ryzhii, I. Khmyrova, V. Mitin, M. Stroscio, and M. Will-

ander. On the Detectivity of Quantum-Dot Infrared Photo-

detectors. Appl. Phys. Lett. 2001, 78, 3523–3525.

61. X. H. Su, J. Yang, P. Bhattacharya, G. Ariyawansa, and A. G.

U. Perera. Terahertz Detection with Tunneling Quantum

Dot Intersublevel Photodetector. Appl. Phys. Lett. 2006, 89,

031117–031117-3 .

62. G. Ariyawansa, A. G. U. Perera, G. S. Raghavan, G. von

Winckel, A. Stintz, and S. Krishna. Effect of Well Width on

Three-Color Quantum Dots-In-A-Well Infrared Detectors.

IEEE Photon. Tech. Lett. 2005, 17, 1064–1064.

63. S. Raghavan, D. Forman, P. Hill, N. R. Weisse-Bernstein, G.

von Winckel, P. Rotella, S. Krishna, S. W. Kennerly, and J.

W. Little. Normal-Incidence InAs/In0.15Ga0.85As Quantum

Dots-In-A-Well Detector Operating in the Long-Wave Infra-

red AtmosphericWindow (8–12Mm). J. Appl. Phys. 2004, 96,

1036–1039.

64. X. Han, J. Li, J. Wu, G. Cong, X. Liu, Q. Zhu, and Z. Wang.

Intersubband Optical Absorption in Quantum Dots-In-A-

Well Heterostructures. J. Appl. Phys. 2005, 98, 053703–

053703–5.

65. R. S. Attaluri, S. Annamalai, K. T. Posani, A. Stintz, and S.

Krishna. Effects of Si Doping on Normal Incidence InAs/

In0.15Ga0.85As Dots-In-Well Quantum Dot Infrared Photo-

detectors. J. Appl. Phys. 2006, 99, 083105–083105-083103.

66. S.Krishna, S. Raghavan,G. vonWinckel, P.Rotella, A. Stintz,

C. P. Morath, D. Le, and S. W. Kennerly. Two Color InAs/

InGaAs Dots-In-A-Well Detector with Background-Limited

Performance At 91 K. Appl. Phys. Lett. 2003, 82, 2574–2576.

67. B. F. Levine. Quantum-Well Infrared Photodetectors.

J. Appl. Phys. 1993, 74, R1–R81.

68. K. David and Y.-C. Chang. Effects of Coulomb Blockade on

the Photocurrent in Quantum Dot Infrared Photodetectors.

Phys. Rev. B 2003, 67, 035313–7.

28 Epitaxial Quantum Dot Infrared Photodetectors

3GW8218 08/26/2014 20:43:49 Page 29

69. V. Apalkov. Quantum Dot Infrared Photodetectors: Interdot

Coupling. J. Appl. Phys. 2006, 100, 076101–076101–3.

70. A. G. U. Perera, W. Z. Shen, S. G. Matsik, H. C. Liu, M.

Buchanan, and W. J. Schaff. GaAs/Algaas Quantum Well

Photodetectors with a Cutoff Wavelength at 28 Mm. Appl.

Phys. Lett. 1998, 72, 1596–1598.

71. H.-Shi Ling, S.-Yu Wang, W.-Cheng Hsu, and C.-Ping Lee.

Voltage-Tunable Dual-Band Quantum Dot Infrared Photo-

detectors for Temperature Sensing. Opt. Express 2012, 20,

10484–10489.

72. S. Adhikary, Y. Aytac, S. Meesala, S. Wolde, A. G. U. Perera,

and S. Chakrabarti. A Multicolor, Broadband (5–20 Mm),

Quaternary-Capped InAs/GaAs Quantum Dot Infrared Pho-

todetector. Appl. Phys. Lett. 2012, 101, 261114–261114–3.

73. B. Aslan, H. C. Liu, J. A. Gupta, Z. R. Wasilewski, G. C. Aers,

S. Raymond, and M. Buchanan. Observation of Resonant

Tunneling Through a Self-Assembled InAs Quantum Dot

Layer. Appl. Phys. Lett. 2006, 88, 043103–043103.

74. X. Su, S. Chakrabarti, P. Bhattacharya, G. Ariyawansa, and

A. G. U. Perera. A Resonant Tunneling Quantum-Dot Infra-

red Photodetector. IEEE J. Quantum Electron. 2005, 41,

974–979.

75. S. Chakrabarti, A. D. Stiff-Roberts, X. H. Su, P. Bhatta-

charya, G. Ariyawansa, and A. G. U. Perera. High-Perform-

ance Mid-Infrared Quantum Dot Infrared Photodetectors.

J. Phys. D 2005, 38, 2135–2141.

76. D. G. Esaev, S. G. Matsik, M. B. M. Rinzan, A. G. U. Perera,

H. C. Liu, andM. Buchanan. Resonant Cavity Enhancement

inHeterojunction GaAs/Algaas Terahertz Detectors. J. Appl.

Phys. 2003, 93, 1879–1883.

77. M. S. €Unl€u and S. Strite. Resonant Cavity Enhanced Pho-

tonic Devices. J. Appl. Phys. 1995, 78, 607–639.

78. R. S. Attaluri, J. Shao, K. T. Posani, S. J. Lee, J. S. Brown, A.

Stintz, and S. Krishna. Resonant Cavity Enhanced InAs/

In0.15Ga0.85As Dots-In-A-Well Quantum Dot Infrared Pho-

todetector. J. Vac. Sci. Tech. B 2007, 25, 1186–1190.

79. F. Pulizzi, D. Walker, A. Patan�e, L. Eaves, M. Henini, D.

Granados, J. M. Garcia, V. V. Rudenkov, P. C. M. Christi-

anen, J. C. Maan, P. Offermans, P. M. Koenraad, and G. Hill.

Excited States of Ring-Shaped (InGa)As Quantum Dots in a

GaAs/(Alga)As Quantum Well. Phys. Rev. B 2005, 72,

085309–085309-6.

80. E. Anemogiannis, E. N. Glytsis, and T. K. Gaylord. Quasi-

Bound States Determination Using a Perturbed Wavenum-

bers Method in a Large Quantum Box. IEEE J. Quantum

Electron. 1997, 33, 742–752.

81. K. K. Choi, B. F. Levine, R. J. Malik, J. Walker, and C. G.

Bethea. Periodic Negative Conductance by Sequential Reso-

nant Tunneling Through an Expanding High-Field Super-

lattice Domain. Phys. Rev. B 1987, 4172–4175.

82. K. K. Choi, B. F. Levine, C. G. Bethea, J. Walker, and R. J.

Malik. Photoexcited Coherent Tunneling in aDouble-Barrier

Superlattice. Phys. Rev. Lett. 1987, 59, 2459–2462.

83. A. Carbone and P. Mazzetti. Current Noise Spectroscopy of

Deep Energy Levels in Photoconductors. J. Appl. Phys. 1996,

80, 1559–1566.

84. H. Schneider R. Rehm, C. Sch€onbein, and M. Walther. Noise

Current Investigations of G–RNoise Limited and Shot Noise

Limited Qwips Physica E-low-dimensional Systems & Nano-

structures. Phys. E 2000, 7, 124–129.

85. Y. Paltiel, N. Snapi, A. Zussman, andG. Jung. Non-Gaussian

Dark Current Noise in P-Type Quantum-Well Infrared Pho-

todetectors. Appl. Phys. Lett. 2005, 87, 231103–231103-3.

86. Y. Haddab, A. P. Friedrich, and R. S. Popovic. Observation of

Dopant–Vacancy Defects by Low-Frequency Noise Measure-

ments in Heavily Doped Nþ/Pþ Silicon Diodes. Solid-State

Electron. 1999, 43, 413–416.

87. R. A. Rupani, S. Ghosh, X. Su, and P. Bhattacharya. Low

Frequency Noise Spectroscopy in InAs/GaAs Resonant Tun-

neling QuantumDot Infrared Photodetectors.Microelectron.

J. 2008, 39, 307–313.

88. S. Schulz, A. Schramm, C. Heyn, and W. Hansen, Tunneling

Emission from Self-Assembled InAs Quantum Dots Probed

with Capacitance Transients.Phys. Rev. B 2006, 74, 033311–

033311-4.

89. R. Wetzler, A. Wacker, E. Sch€oll, C. M. A. Kapteyn, R. Heitz,

and D. Bimberg. Capacitance–Voltage Characteristics of

InAs/GaAs Quantum Dots Embedded in a Pn Structure.

Appl. Phys. Lett. 2000, 77, 1671–1673.

90. C. Bock, K. H. Schmidt, U. Kunze, S. Malzer, and G. H.

D€ohler. Valence-Band Structure of Self-Assembled InAs

Quantum Dots Studied by Capacitance Spectroscopy.

Appl. Phys. Lett. 2003, 82, 2071–2073.

91. S. Anand, N. Carlsson, M-E Pistol, L. Samuelson, and W.

Seifert. Deep Level Transient Spectroscopy of Inp Quantum

Dots. Appl. Phys. Lett. 1995, 67, 3016–3018.

92. C. M. A. Kapteyn, F. Heinrichsdorff, O. Stier, R. Heitz, M.

Grundmann, N. D. Zakharov, D. Bimberg, and P. Werner.

Electron Escape from InAs Quantum Dots. Phys. Rev. B

1999, 60, 14265–14268.

93. A. G. U. Perera, W. Z. Shen, M. Ershov, H. C. Liu, M.

Buchanan, and W. J. Schaff. Negative Capacitance of

GaAs Homojunction Far-Infrared Detectors. Appl. Phys.

Lett. 1999, 74, 3167–3169.

94. M. Ershov, H. C. Liu, L. Li, M. Buchanan, Z. R. Wasilewski,

and A. K. Jonscher. Negative Capactiance Effect in Semi-

conductor Devices. IEEE Trans. Electron Dev. 1998, 45,

2196–2206.

95. G. Ariyawansa, V. Apalkov, A. G. U. Perera, S. G. Matsik, G.

Huang, and P. Bhattacharya. Bias-Selectable Tricolor Tun-

neling QuantumDot Infrared Photodetector for Atmospheric

Windows. Appl. Phys. Lett. 2008, 92, 111104–111103.

96. D. G. Esaev, M. B. M. Rinzan, S. G. Matsik, A. G. U. Perera,

H. C. Liu, B. N. Zvonkov, V. I. Gavrilenko, and A. A. Belya-

nin. High Performance Single Emitter Homojunction Inter-

facial Work Function Far Infrared Detectors. J. Appl. Phys.

2004, 95, 512–519.

97. G. Ariyawansa, A. G. U. Perera, G. Huang, and P. Bhat-

tacharya. Wavelength Agile Superlattice Quantum Dot

Infrared Photodetector. Appl. Phys. Lett. 2009, 94,

131109–131109-3.

98. A. Majumdar, K. K. Choi, J. L. Reno, and D. C. Tsui.

Voltage Tunable Superlattice Infrared Detector For Mid-

and Long-Wavelength Detection. Appl. Phys. Lett. 2005,

86, 261110–261113.

99. A. Rogalski, J. Antoszewski, and L. Faraone. Third-Genera-

tion Infrared Photodetector Arrays. J. Appl. Phys. 2009, 105,

091101–091101-6.

100. H. S. Ling, S. Y. Wang, C. P. Lee, and M. C. Lo. High

Quantum Efficiency Dots-In-A-Well Quantum Dot Infrared

Photodetectors with Algaas Confinement Enhancing Layer.

Appl. Phys. Lett. 2008, 92, 193506–193506-3.

101. A. V. Barve, T. Rotter, Y. Sharma, S. J. Lee, S. K. Noh, and S.

Krishna. Systematic Study of Different Transitions in High

Operating Temperature Quantum Dots in a Well Photode-

tectors. Appl. Phys. Lett. 2010, 97, 061105–061105-3.

Epitaxial Quantum Dot Infrared Photodetectors 29

3GW8218 08/26/2014 20:43:49 Page 30

102. S. D. Gunapala, S. V. Bandara, C. J. Hill, D. Z. Ting, J. K. Liu,

S. B. Rafol, E. R. Blazejewski, J. M. Mumolo, S. A. Keo, S.

Krishna, Y. C. Chang, and C. A. Shott. 2006 (unpublished).

103. J. R. Hoff, M. Razeghi, and G. J. Brown. Effect of the Spin

Split-Off Band on Optical Absorption in P-Type Ga1-Xinxa-

syp1-Y Quantum-Well Infrared Detectors. Phys. Rev. B

1996, 54, 10773–10783.

104. Y.-Feng Lao, S. Wolde, A. G. U. Perera, Y. H. Zhang, T. M.

Wang, H. C. Liu, J. O. Kim, T. Schuler-Sandy, Z.-Bing Tian,

and S. S. Krishna. InAs/GaAs P-Type QuantumDot Infrared

Photodetector withHigher Efficiency.Appl. Phys. Lett. 2013,

103, 241115–241115-3.

105. Y. F. Lao, P. K. D. D. P. Pitigala, A. G. U. Perera, H. C. Liu,

M. Buchanan, Z. R. Wasilewski, K. K. Choi, and P. Wije-

warnasuriya. Light-Hole and Heavy-Hole Transitions for

High-Temperature Long-Wavelength Infrared Detection.

Appl. Phys. Lett. 2010, 97, 091104–091104-3.

106. W. Sheng and J.-Pierre Leburton. Interband Transition Dis-

tributions in theOptical Spectra of InAs/GaAsSelf-Assembled

Quantum Dots. Appl. Phys. Lett. 2002, 80, 2755–2757.

107. M. A. Cusack, P. R. Briddon, and M. Jaros. Electronic

Structure of InAs/GaAs Self-Assembled Quantum Dots.

Phys. Rev. B 1996, 54, R2300–R2303.

108. A. Schliwa, M. Winkelnkemper, and D. Bimberg. Impact Of

Size, Shape, and Composition on Piezoelectric Effects and

Electronic Properties of In(Ga)As/GaAs Quantum Dots.

Phys. Rev. B 2007, 76, 205324–205324-6.

109. A. I. Yakimov, A. A. Bloshkin, V. A. Timofeev, A. I. Nikiforov,

and A. V. Dvurechenskii. Effect of Overgrowth Temperature

on the Mid-Infrared Response of Ge/Si(001) Quantum Dots.

Appl. Phys. Lett. 2012, 100, 053507–053507-3.

110. E. E. Haller. Advanced Far-Infrared Detectors. Infrared

Phys. Tech. 1994, 35, 127–146.

111. M. Suzuki and M. Tonouchi. Fe-Implanted InGaAs Photo-

conductive Terahertz Detectors Triggered By 1.56mmFemto-

second Optical Pulses. Appl. Phys. Lett. 2005, 86, 163504–

163504-3.

112. H. C. Liu, C. Y. Song, A. J. SpringThorpe, and J. C. Cao.

Terahertz Quantum-Well Photodetector. Appl. Phys. Lett.

2004, 84, 4068–4070.

113. A. G. U. Perera, W. Z. Shen, H. C. Liu, M. Buchanan, and W.

J. Schaff. GaAs Homojunction Interfacial Workfunction

Internal Photoemission (HIWIP) Far-Infrared Detectors.

Mater. Sci. Eng.: B 2000, 74, 56–60.

114. M. B. M. Rinzan, A. G. U. Perera, S. G. Matsik, H. C. Liu, Z.

R. Wasilewski, and M. Buchanan. Algaas Emitter/GaAs

Barrier Terahertz Detector with A 2.3 Thz Threshold.

Appl. Phys. Lett. 2005, 86, 071112–071112-3.

115. A. Lorke, R. J. Luyken, J. M. Garcia, and P. M. Petroff.

Growth and Electronic Properties of Self-Organized Quan-

tum Rings. Jpn. J. Appl. Phys. 2001, 40, 1857–1859.

116. G.Huang,W.Guo,P.Bhattacharya,G.Ariyawansa, andA.G.

U.Perera.AQuantumRingTerahertzDetectorwithResonant

TunnelBarriers.Appl.Phys.Lett.2009,94, 101115–101115-3.

117. S. Bhowmick, G. Huang, W. Guo, C. S. Lee, P. Bhattacharya,

G. Ariyawansa, and A. G. U. Perera. High-Performance

Quantum Ring Detector for the 1–3 Terahertz Range.

Appl. Phys. Lett. 2010, 96, 231103–231103-3.

118. J.-Horng Dai, J.-Han Lee, Y-Lung Lin, and S.-Chen Lee. In

(Ga)As Quantum Rings for Terahertz Detectors. Jpn. J.

Appl. Phys. 2008, 47, 2924–2926.

119. L. Jheng-Han, D. Jong-Horng, C. Chi-Feng, and S-Chen Lee.

In(Ga)As Quantum Ring Terahertz Photodetector with Cut-

offWavelength at 175mm. IEEEPhoton. Tech. Lett. 2009, 21,

721–723.

120. D. G. Esaev, M. B. M. Rinzan, S. G. Matsik, and A. G. U.

Perera. Design and Optimization of GaAs/Algaas Heterojunc-

tion Infrared Detectors. J. Appl. Phys. 2004, 96, 4588–4597.

121. H. Luo, H. C. Liu, C. Y. Song, and Z. R. Wasilewski. Back-

ground-Limited Terahertz Quantum-Well Photodetector.

Appl. Phys. Lett. 2005, 86, 231103–231103-3.

122. M. Graf, G. Scalari, D. Hofstetter, J. Faist, H. Beere, E.

Linfield, D. Ritchie, and G. Davies. Terahertz Range Quan-

tum Well Infrared Photodetector. Appl. Phys. Lett. 2004, 84,

475–477.

123. J. M. Garcıa, G. Medeiros-Ribeiro, K. Schmidt, T. Ngo, J. L.

Feng, A. Lorke, J. Kotthaus, and P. M. Petroff. Intermixing

and Shape Changes During the Formation of InAs Self-

Assembled Quantum Dots. Appl. Phys. Lett. 1997, 71,

2014–2016.

124. M. Zavvari, K. Abedi, and M. Karimi. Numerical Analysis of

Quantum Ring Intersubband Photodetector for Far Infrared

Detection. Opt. Quant. Electron. 2013, 1, 1–10.

125. K. T. Posani, V. Tripathi, S. Annamalai, N. R. Weisse-Bern-

stein, S. Krishna, R. Perahia, O. Crisafulli, and O. J. Painter.

Nanoscale Quantum Dot Infrared Sensors with Photonic

Crystal Cavity. Appl. Phys. Lett. 2006, 88, 151104–151104-

151103.

126. A.Mandal, A. Agarwal,H.Ghadi, G.KumariK. C. A. Basu, N.

B. V. Subrahmanyam, P. Singh, and S. Chakrabarti. More

than one Order Enhancement in Peak Detectivity (D�) for

Quantum Dot Infrared Photodetectors Implanted with Low

Energy Light Ions (H–). Appl. Phys. Lett. 2013, 102, 051105–

051105-3.

127. C.-Cheng Wang and S.-Di Lin. Resonant Cavity-Enhanced

Quantum-Dot Infrared Photodetectors with Sub-Wave-

length Grating Mirror. J. Appl. Phys. 2013, 113, 213108–

213108-6.

128. D. Pal and E. Towe. Characteristics of High-Operating-

Temperature InAs/GaAs Quantum-Dot Infrared Detectors.

Appl. Phys. Lett. 2006, 88, 153109–153109-3.

129. P. V. V. Jayaweera, A. G. U. Perera, and K. Tennakone.

Displacement Currents in Semiconductor Quantum Dots

Embedded Dielectric Media: A Method for Room Tempera-

ture Photon Detection. Appl. Phys. Lett. 2007, 91, 063114–

063114-3.

130. G. Konstantatos, I. Howard, A. Fischer, S. Hoogland, J.

Clifford, E. Klem, L. Levina, and E. H. Sargent. Ultrasensi-

tive Solution-Cast Quantum Dot Photodetectors. Nature

2006, 442, 180–183.

131. P. M. Alsing, D. A. Cardimona, D. H. Huang, T. Apostolova,

W. R. Glass, and C. D. Castillo. Advanced Space-Based

Detector Research at the Air Force Research Laboratory.

Infrared Phys. Tech. 2007, 50, 89–94.

132. 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. Two-Color Focal Plane Arrays Based on

Self Assembled Quantum Dots in A Well Heterostructure.

Phys. Status Solid C 2006, 3, 439–443.

A. G. U. PERERA

Georgia State University, Atlanta

G. ARIYAWANSA

UES Inc., Dayton, Ohio, USA

30 Epitaxial Quantum Dot Infrared Photodetectors