Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
www.elsevier.com/locate/infrared
Institute of Applied Physics, Military University of Technology, 2 Kaliskiego St., 00-908 Warsaw 49, Poland
Abstract
The paper presents progress in infrared (IR) detector technologies during 200 history of their development. Clas-
sication of two types of IR detectors (photon detectors and thermal detectors) is done on the basis of their principle of
operation. The overview of IR systems and detectors is presented. Also recent progress in dierent IR technologies is
described. Discussion is focused mainly on current and the most rapidly developing detectors: HgCdTe heterostructure
photodiodes, quantum well AlGaAs/GaAs photoresistors, and thermal detectors. The outlook for near-future trends in
IR technologies is also presented. 2002 Elsevier Science B.V. All rights reserved.
Keywords: Photon detectors; Thermal detectors; Focal plane arrays; Two-colour detectors
1350-4495/02/$ - see front matter 2002 Elsevier Science B.V. All rights reserved.
PII: S 1 3 5 0 - 4 4 9 5 ( 0 2 ) 0 0 1 4 0 - 8
188 A. Rogalski / Infrared Physics & Technology 43 (2002) 187210
University of Berlin, discovered that lead sulphide class of detectors the radiation is absorbed within
(from natural galena found in Sardinia) was the material by interaction with electrons. The
photoconductive and had response to about 3 lm observed electrical output signal results from the
[8]. changed electronic energy distribution. The pho-
Many materials have been investigated in the ton detectors show a selective wavelength depen-
IR eld. Observing a history of the development of dence of the response per unit incident radiation
the IR detector technology, a simple theorem, after power. They exhibit both perfect signal-to-noise
Norton [9], can be stated: All physical phenomena performance and a very fast response. But to
in the range of about 0.11 eV can be proposed for achieve this, the photon detectors require cryo-
IR detectors. Among these eects are: thermo- genic cooling. Cooling requirements are the main
electric power (thermocouples), change in elec- obstacle to the more widespread use of IR systems
trical conductivity (bolometers), gas expansion based on semiconductor photodetectors making
(Golay cell), pyroelectricity (pyroelectric detec- them bulky, heavy, expensive and inconvenient to
tors), photon drag, Josephson eect (Joseph- use. Depending on the nature of interaction, the
son junctions, SQUIDs), internal emission (PtSi class of photon detectors is further sub-divided
Schottky barriers), fundamental absorption (in- into dierent types. The most important are: in-
trinsic photodetectors), impurity absorption (ex- trinsic detectors, extrinsic detectors, photoemissive
trinsic photodetectors), low-dimensional solids (metal silicide Schottky barriers) detectors, and
(superlattice (SL) and quantum well (QW) detec- quantum well detectors. Table 1 shows the com-
tors), dierent type of phase transitions, etc. parison of various IR detectors.
Fig. 1 gives approximate dates of signicant The second class of IR detectors is composed of
development eorts for the materials mentioned. thermal detectors. In a thermal detector the inci-
The years during World War II saw the origins of dent radiation is absorbed to change temperature
modern IR detector technology. Interest has cen- of the material, and the resultant change in some
tred mainly on the wavelengths of the two atmo- physical properties is used to generate an electrical
spheric windows 35 and 814 lm, though in output. The detector element is suspended on lags,
recent years there has been increasing interest in which are connected to the heat sink. Thermal
longer wavelengths stimulated by space applica- eects are generally wavelength independent; the
tions [10]. signal depends upon the radiant power (or its rate
of change) but not upon its spectral content. In
pyroelectric detectors a change in the internal
2. Classication of IR detectors spontaneous polarisation is measured, whereas in
the case of bolometers a change in the electrical
Progress in IR detector technology is connected resistance is measured. In contrast to photon de-
mainly to semiconductor IR detectors, which are tectors, the thermal detectors typically operate at
included in the class of photon detectors. In this room temperature. They are usually characterized
A. Rogalski / Infrared Physics & Technology 43 (2002) 187210 189
Table 1
Comparison of IR detectors (after Ref. [11])
Detector type Advantages Disadvantages
Thermal Light, rugged, reliable, and low cost Low detectivity at high frequency
Room temperature operation Slow response (ms order)
Photon
Intrinsic
IVVI Available low-gap materials Poor mechanical
Well studied Large permittivity
IIVI Easy band-gap tailoring Non-uniformity over large area
Well-developed theory and exp. High cost in growth and processing
Multicolour detectors
IIIV Good material and dopants Heteroepitaxy with large lattice mismatch
Advanced technology
Possible monolithic integration
Extrinsic Very long wavelength operation Extremely low temperature operation
Relatively simple technology
Free carriers Low-cost, high yields Low quantum eciency
Large and close packed 2D arrays Low temperature operation
Quantum wells
Type I Matured material growth Low quantum eciency
Good uniformity over large area Complicated design and growth
Multicolour detectors
Type II Low Auger recombination rate Complicated design and growth
Easy wavelength control Sensitive to the interfaces
by modest sensitivity and slow response but they best values of NEDT, below 0.1 K, could be
are cheap and easy to use. The greatest utility in reached because eective noise bandwidths < 100
IR technology has found bolometers, pyroelectric Hz can be achieved.
detectors and thermopiles. Uncooled, monolithic focal plane arrays (FPAs)
Up till the nineties, thermal detectors have been fabricated from thermal detectors may revolutio-
considerably less exploited in commercial and nise development of thermal imagers. Recently,
military systems in comparison with photon de- very encouraging results have been obtained with
tectors. The reason for this disparity is that ther- micromachined silicon bolometer [10,12] and py-
mal detectors are popularly believed to be rather roelectric detector arrays [10,13].
slow and insensitive in comparison with photon
detectors. As a result, the worldwide eort to de-
velop thermal detectors was extremely small rela-
tive to that of photon detector. In the last decade 3. Focal plane arrays
however, it has been shown that extremely good
imagery can be obtained from large thermal de- There are a number of architectures used in the
tector arrays operating uncooled at TV frame development of IR FPAs. In general, they may be
rates. The speed of thermal detectors is quite ad- classied as hybrid and monolithic. The central
equate for non-scanned imagers with two-dimen- design questions involve performance advantages
sional (2D) detectors. The moderate sensitivity of versus ultimate producibility. Each application
thermal detectors can be compensated by a large may favour a dierent approach depending on
number of elements in 2D electronically scanned the technical requirements, projected costs, and
arrays. With large arrays of thermal detectors the schedule.
190 A. Rogalski / Infrared Physics & Technology 43 (2002) 187210
In the monolithic approach, some of the mul- ogy with Schottky-barrier detectors is the only
tiplexing is done in the detector material itself than technology, which has matured to a level of
in an external readout circuit. The basic element of practical use.
a monolithic array is a metalinsulatorsemicon- Hybrid FPAs detectors and multiplexers are
ductor (MIS) structure as shown in Fig. 2(c). A fabricated on dierent substrates and mated with
MIS capacitor detects and integrates the IR-gen- each other by the ip-chip bonding (Fig. 3) or
erated photocurrent. Although eorts have been loophole interconnection. In this case we can op-
made to develop monolithic FPAs using narrow- timise the detector material and multiplexer inde-
gap semiconductors, silicon-based FPA technol- pendently. Other advantages of the hybrid FPAs
Fig. 2. Monolithic IR FPAs: (a) all-silicon; (b) heteroepitaxy-on-silicon; (c) non-silicon (e.g., HgCdTe CCD) (after Ref. [14]).
Fig. 3. Hybrid IR FPA with independently optimised signal detection and readout: (a) indium bump technique, (b) loophole tech-
nique.
A. Rogalski / Infrared Physics & Technology 43 (2002) 187210 191
are near 100% ll factor and increased signal- minimum of three transistors per pixel compared
processing area on the multiplexer chip. In the to one per memory cell. Consequently, whereas
ip-chip bonding, the detector array is typically various 64 64 FPAs were available in the early
connected by pressure contacts via indium bumps 1980s, several vendors are now producing mono-
to the silicon multiplex pads. The detector array lithic FPAs in the TV-compatible 1040 1040
can be illuminated from either the frontside or formats. Fig. 4 illustrates the trend of array size
backside (with photons passing through the trans- over the past 25 years and some projections of
parent detector array substrate). In general, the what will evolve in the coming decade. Rockwell
latter approach is most advantageous. In HgCdTe has developed the worlds largest HgCdTe short
hybrid FPAs, photovoltaic detectors are formed wavelength IR (SWIR) FPA for astronomy and
on thin HgCdTe epitaxial layer on transparent low-background applications. The format of the
CdTe or CdZnTe substrates. For HgCdTe ip- device is a hybrid 2048 2048 with a unit cell size
chip hybrid technology, the maximum chip size is of 18 lm 18 lm. Table 2 contains description
of the order of 10 mm square. In order to over- of representative IR FPAs that are commercially
come this problem, the technology is being devel- available as standard products and/or catalogue
oped with sapphire or silicon as the substrate of items from the major manufactures. Ten years
HgCdTe detectors. When using opaque materials, ago, high quality single element detectors often
substrates must be thinned to 1020 lm in order to were priced over $2000, but now, some current IR
obtain sucient quantum eciencies and reduce FPA production costs are less than $1 per detector
the crosstalk. and even greater reductions are expected in the
There is a large research activity directed to- near future.
wards 2D staring arrays detectors consisting of Two types of silicon addressing circuits have been
more than 106 elements. IR FPAs have nominally developed: CCDs and complementary metal-oxide-
the same growth rate as dynamic random access semiconductor (CMOS) switches. The photogener-
memory (RAM) integrated circuits (ICs) (it is ated carriers are rst integrated in the well formed
consequence of Moores Law, which predicts the by a photogate and subsequently transferred to
ability to double transistor integration on each IC slow (vertical) and fast (horizontal) CCD shift
about every 18 months) but lag behind in size by registers.
about 510 years. ROICs are somewhat analo- An attractive alternative to the CCD readout is
gous to dynamic RAMonly readouts require a coordinative addressing with CMOS switches. The
Fig. 4. Increase in array format size over the past 25 years and projections for the coming decade. PtSi, InSb, and HgCdTe have been
following the pace of dynamic RAM, oset by about a decade. QWIP detectors have been recently reported in sizes as large as
640 480 pixels (after Ref. [15]).
192 A. Rogalski / Infrared Physics & Technology 43 (2002) 187210
Table 2
Representative IR FPAs oered by some major manufacturers
Manufacturer/web site Size/architec- Pixel size Detector material Spectral Operating tem- D kp
ture (lm) range (lm) perature (K) (cm Hz1=2 W1 )/
NETD (mK)
Raytheon/ 256 256/H 30 30 InSb 15.5 1077
www.raytheon.com 1024 1024/H 27 27 InSb 0.65.0 35
320 240/H 50 50 Si:As BIB 228 410
128 128/H 40 40 HgCdTe 911 80
256 256/H 30 30 HgCdTe 8.511 77100
320 240/M 48 48 VOx (bolometer) 814 300 50
328 245/H 35 35 Pyro (BST) 814 300 >50
Rockwell/Boeing/ 256 256/H 40 40 HgCdTe >15 77
www.boeing.com 640 480/H 27 27 HgCdTe >10 77
www.rsc.rockwell.com 225 256/H 40 40 HgCdTe 14.6 120 >10
640 480/H 27 27 HgCdTe 14.6 120 >25
2048 2048/H 18 18 HgCdTe 12.5 95120 > 1 1014
320 240/M 48 48 VOx (bolometer) 814 300 50
Mitsubishi/www.mits- 256 256/M 52 40 PtSi 35 77 0.036
ubishi-imaging.com 512 512/M 26 20 PtSi 35 77 0.033
1024 1024/M 17 17 PtSi 35 77 0.10
advantages of CMOS are that existing foundries. gap semiconductor alloys, HgCdTe is the only
Design rules of 0.25 lm are in production with one material covering the whole IR spectral
pre-production runs of the 0.18 lm design rules. range having nearly the same lattice parameter.
At present, CMOS with minimum feature 6 0.5 The dierence of lattice parameter between
lm is also enabling monolithic visible CMOS CdTe (Eg 1:5 eV) and Hg0:8 Cd0:2 Te (Eg 0:1
imagers. eV) is 0.2%. Replacing small fraction of Cd
with Zn or Te with Se can compensate the re-
sidual lattice mismatch. The independence of
4. Viewpoint on IR detector technologies lattice parameter on composition is a major ad-
vantage of HgCdTe over any other materials.
During the past four decades mercury cadmium
telluride (HgCdTe) has became the most im- Heterojunctions are helpful in achieving high
portant semiconductor for the middle and long performance in practice. For example, the narrow-
wavelength (k 330 lm) IR photodetectors. The gap HgCdTe that absorbs IR radiation can be
short wavelength region has been dominated by buried encapsulated in wider gap HgCdTe pre-
IIIV compounds (InGaAs, InAsSb, InGaSb). venting instabilities due to the weak HgTe bonds.
There have been numerous attempts to replace When background-photon noise is the domi-
HgCdTe with alternative materials. At present, nant noise mechanism, the detector is operating in
several other variable gap alloy systems are known an ideal mode, and is said to exhibit background
including closely related mercury alloys HgZnTe, limited performance (BLIP). BLIP temperature is
HgMnTe, lead tin tellurides and selenides, InAsSb, dened that the device is operating at a tempera-
IIIVI compounds with thallium and bismuth, ture at which the dark current equals the back-
free-carrier detectors and low-dimensional solids. ground photocurrent, given a eld of view (FOV),
The main motivations, behind the numerous and a background temperature. In Fig. 5, plots of
attempts to replace HgCdTe, are technological the calculated temperature required for BLIP op-
problems of this material. One of them is weak eration in 30 FOV, are shown as a function of
HgTe bond, which results in bulk and surface and cuto wavelength. We can see that the operat-
interface instabilities. Uniformity and yield are still ing temperature of bulk intrinsic IR detectors
issues. Nevertheless, HgCdTe remains the leading (HgCdTe and PbSnTe) is higher than for other
semiconductor for IR detectors. The most impor- types of photon detectors. HgCdTe detectors with
tant reasons for this are:
Fig. 6. Double layer planar heterostructure cross-section schematics for SWIR InGaAs (a) and HgCdTe (b) photodiodes.
BLIP operate with thermoelectric coolers in the low growth temperature techniques: metalorganic
MWIR range, but the LWIR detectors (8 6 kc 6 12 chemical vapour deposition (MOCVD) and mo-
lm) operate at 100 K. HgCdTe photodiodes lecular beam epitaxy (MBE).
exhibit higher operating temperature compared to The baseline detector architecture for SWIR
extrinsic detectors, silicide Schottky barriers and InGaAs and HgCdTe is P -on-n device (symbol P
QWIPs. However, the cooling requirements for denotes wider gap). Fig. 6 shows the double layer
QWIPs with cuto wavelengths below 10 lm are heterostructure (DLPH) device cross-sections for
less stringent in comparison with extrinsic detec- both types of photodiodes. Apart from the above
tors and Schottky-barrier devices. described benets, incorporation of a buried nar-
Recently, more interest has been focused on pn row-band-gap active layer in the DLPH reduces
junction heterostructure photodiodes. In compar- tunnelling currents and increases the total dose
ison with heterostructures, homojunctions have radiation hardness, both of which are essential
a lower breakdown voltage and a large reverse detector attributes for remote sensing applications.
leakage current. Another aspect of photodiode The thickness of the base region should be opti-
design is the required bandwidth for high-speed mised for near unity quantum eciency and a low
operation. dark current. This is achieved with a base thick-
Photodiodes with their very low power dissi- ness (typically about 57 lm) slightly higher than
pation, easy multiplexing on focal plane silicon the inverse absorption coecient for single pass
chip and less stringent noise requirements for the devices. Low doping is benecial for a low thermal
readout devices and circuits, can be assembled in generation and high quantum eciency. Since the
2D arrays containing a very large number of ele- diusion length in absorbing region is typically
ments, limited only by existing technologies. longer than its thickness, any carriers generated in
Epitaxy is the preferable technique to obtain the base region can be collected giving rise to the
device-quality materials. Among the various epit- photocurrent.
axial techniques, liquid phase epitaxy (LPE) is the
most mature method. LPE growth must be carried 4.1. InGaAs photodiodes
out at relatively high growth temperature with
adherent interdiusion and resulting graded in- The need for high-speed, low-noise Inx Ga1x As
terfaces. Recent eorts are aimed mostly at (InGaAs) photodetectors for use in lightwave
A. Rogalski / Infrared Physics & Technology 43 (2002) 187210 195
detection in the 35 lm spectral range [30,31]. Extrinsic photoresistors are used in a wide
Radiation is transmitted through the p-type silicon range of the IR spectrum extending from a few lm
and is absorbed in the metal PtSi (not in the to 300 lm. They are the principal detectors op-
semiconductor), producing hot holes which are erating in the range k > 20 lm [33]. Detectors
then emitted over the potential barrier into the based on silicon and germanium have found the
silicon, leaving the silicide charged negatively. widest application as compared with extrinsic
Negative charge of silicide is transferred to a CCD photodetectors on other materials. Si has several
by the direct charge injection method. advantages over Ge; for example, three orders of
The eective quantum eciency in the 35 lm magnitude higher impurity solubilities are attain-
atmospheric window is very low, of the order of able, hence thinner detectors with better spatial
1%, but useful sensitivity is obtained by means of resolution can be fabricated from silicon. Si has
near full frame integration in area arrays. The lower dielectric constant than Ge, and the related
quantum eciency has been improved by thin- device technology of Si has now been more thor-
ning PtSi lm and implementation of an opti- oughly developed, including contacting methods,
cal cavity. Due to very low quantum eciency, surface passivation and mature MOS and CCD
the operating temperature of Schottky-barrier technologies. Moreover, Si detectors are charac-
photoemissive detectors is lower than another terized by superior hardness in nuclear radiation
types of IR photon detectors (see Fig. 5). environments. Fig. 12 illustrates the spectral re-
Schottky photoemission is independent of such sponse for several extrinsic detectors.
factors as semiconductor doping, minority carrier The availability of a highly developed silicon
lifetime, and alloy composition, and, as a result of MOS technology facilities the integration of large
this, has spatial uniformity characteristics that are detector arrays with charge-transfer devices for
far superior to those of other detector technolo- readout and signal processing. The well-estab-
gies. Uniformity is only limited by the geometric lished technology also helps in the manufacturing
denition of the detectors. of uniform detector arrays and the formation of
A. Rogalski / Infrared Physics & Technology 43 (2002) 187210 199
low-noise contacts. Although the potential of large Among the dierent types of quantum well IR
extrinsic silicon FPAs for terrestrial applications photodetectors (QWIPs), technology of the GaAs/
has been examined, interest has declined in favour AlGaAs multiple quantum well detectors is the
of HgCdTe and InSb with their more convenient most mature [39,40].
operating temperatures. Strong interest in doped QWIP technology is based on the well-devel-
silicon continues for space applications, particu- oped A3 B5 material system, which has a large
larly in low-background ux and for wavelengths industrial base with a number of military and
from 13 to 20 lm, where compositional control is commercial applications. QWIP cannot compete
dicult for HgCdTe. The shallower impurity en- with HgCdTe photodiode as the single device es-
ergies in germanium allow detectors with spectral pecially at higher temperature operation (>70 K)
response up to beyond 100 lm wavelength and due to fundamental limitations associated with
major interest still exists in extrinsic germanium intersubband transitions [17]. However, the ad-
for wavelengths beyond about 20 lm. vantage of HgCdTe is less distinct in tempera-
To maximize the quantum eciency and de- ture range below 50 K due to problems involved
tectivity of extrinsic photoconductors, the doping in a HgCdTe material (p-type doping, Shockley
level should be as high as possible. This idea is Read recombination, trap-assisted tunnelling,
realized in blocked impurity band (BIB) devices. surface and interface instabilities). Even though
The longer spectral response of the BIB Si:As de- that QWIP is a photoconductor, several its prop-
vice compared with the bulk Si:As device (see Fig. erties such as high impedance, fast response
12) is due to the higher doping level in the former time, long integration time, and low power con-
that reduces the binding energy of an electron. For sumption, well comply requirements of large FPAs
a detailed analysis of the BIB detector see Szmu- fabrication. Due to the high material quality at
lowicz and Madarsz [35]. low temperature, QWIP has potential advantages
BIB devices made from either doped silicon or over HgCdTe for VLWIR FPA applications in
doped germanium are sensitive in the IR wave- terms of the array size, uniformity, yield and cost
length range of 2 and 220 lm. BIB devices in large of the systems.
staring array formats are now becoming com- Fig. 13 shows two detector congurations used
mercially available. The best results have been in fabrication of QWIP FPAs. In the bound-to-
achieved to date for Si:As BIB hybrid FPAs pro- continuum QWIP the photoelectron can escape
duced by Hughes Technology Center in Carlsbad from the quantum well to the continuum transport
[36,37] and Rockwell International Science Center states without being required to tunnel through the
200 A. Rogalski / Infrared Physics & Technology 43 (2002) 187210
or use a lter wheel to spectrally discriminate the consist of back-to-back photovoltaic pn junc-
image focused on single FPA. Consequently, these tions. This device architecture is realised by plac-
approaches are expensive in terms of size, com- ing a longer wavelength HgCdTe photodiode
plexity, and cooling requirements. Both HgCdTe simply behind shorter wavelength photodiode.
photodiodes and QWIPs oer the multicolour ca- Also quaternary device structure is used [60].
pability in the MWIR and LWIR range. Both sequential-mode and simultaneous mode
Considerable progress has been recently dem- detectors are fabricated from the multilayer ma-
onstrated in multispectral HgCdTe detectors em- terials. The mode of detection is determined by the
ploying MBE and MOCVD for the growth of fabrication process. Figs. 17 and 18 show the ele-
variety devices [5461]. Also QWIPs technology ments of arrays of two-colour photovoltaic unit
demonstrates considerably progress in fabrication cells in both modes. The sequential-mode detector
of multicolour FPAs [40,44,6266]. Devices for the has a single indium bump per unit cell that permits
sequential and simultaneous detection of two clo- sequential bias selectivity of the spectral bands
sely spaced sub-bands in the MWIR and LWIR associated with operating tandem photodiodes.
radiation have been demonstrated. The problems with the bias-selectable device are
the following: its construction does not allow in-
5.1. Dual-band HgCdTe FPAs dependent selection of the optimum bias voltage
for each photodiode, and there can be substantial
The two-colour detector arrays are usually medium wavelength crosstalk in the long wave-
based upon an nPN HgCdTe triple layer het- length detector. To overcome the problems of the
erojunction (TLHJ) design. The TLHJ detectors bias-selectable device, the independently accessed
Fig. 17. Cross-section of integrated two-colour detectors in an nPN layer structure for sequential operating mode.
Fig. 18. Cross-section of integrated two-colour detectors in an nPN layer structure for simultaneous operating mode.
204 A. Rogalski / Infrared Physics & Technology 43 (2002) 187210
Fig. 19. Spectral response curves for two-colour HgCdTe detectors in various dual-band combinations of spectral bands (after Ref.
[9]).
A. Rogalski / Infrared Physics & Technology 43 (2002) 187210 205
Fig. 24. Structure cross-section of the interlace dual-band FPA (after Ref. [64]).
A. Rogalski / Infrared Physics & Technology 43 (2002) 187210 207
6. Anticipated evolution of IR technology in the near arrays will capture the low-cost markets. Current
future uncooled bolometer FPAs have achieved NEDT
less than 10 mK with f =1 optics, what open the
The future applications of IR detector systems door to the use of less expensive slower optical
require: systems.
It is supposed that sales of IR thermal imaging
higher pixel sensitivity, equipment to the automobile market will begin to
further increase in pixel density to above 106 rapidly change the relative ratio between military/
pixels, government and commercial IR markets. Today
cost reduction in IR imaging array systems only about 10% of the market is commercial.
through the use of less cooling sensor technol- After a decade the commercial market can grow
ogy combined with integration of detectors and to over 70% in volume and 40% in value, largely
signal-processing functions (with much more connected with volume production of uncooled
on-chip signal processing), imagers for automobile driving [9]. In large volume
improvement in the functionality of IR imaging production for automobiles drivers the cost of
arrays through development of multispectral uncooled imaging systems will decrease to below
sensors. $1000. Of course, these systems will cover other
segments of the transportation industry: trucks,
To reduce the real cost of the IR image systems, trains, ships, barges, buses, and airplanes.
one must take action on all the elements, which For same applications requiring uncooled
make up the cost to the user. The cost can be detectors, the slow response speed is unaccept-
broken down into three parts: the chip (detector able. Recently, a number of concepts (e.g., non-
ROIC), the dewar, integration and tests. The user equilibrium device [68], multijunction HgCdTe
must add the cryogenic machine cost that is not photodiodes [69], optical immersion) and new
negligible compared to the component ones. This materials (InAsSb, InAs/GaSb-based type II su-
explains why the cost of PtSi or QWIPs is not perlattices) [11] have been proposed to improve
markedly less than that of photon detectors of the performance of photon detectors operating at near
same complexity, even though the raw materials room temperature. The measurements show the
(silicon and GaAs) is much less than for HgCdTe. possibility to achieve detectivity of 1 109
A possible reduction in the purchase price is cm Hz1=2 W1 at the 89 lm range and potentially,
counterbalanced by a signicant increase in oper- the devices can be assembled in large FPAs.
ating costs. Despite serious competition from alternative
Detector maturity is a function of the accumu- technologies and slower progress than expected,
lated experience and development eort, the HgCdTe is unlikely to be seriously challenged for
complexity of the device required, and the inherent high-performance applications, applications re-
diculty presented by the material technology. At quiring multispectral capability and fast response.
present, HgCdTe photodiodes and BIB extrinsic The recent successes of competing cryogenically
silicon detectors are not fully mature. PtSi tech- cooled detectors are due to technological, not
nology is mature and has received a plateau. Other fundamental issues. The steady progress in epit-
two detector technologies such as InSb and silicon axial technology would make HgCdTe devices
bolometers are still evolving signicantly as ap- much more aordable in the near future. The
plications for larger array congurations and much higher operation temperature of HgCdTe,
smaller pixel sizes continue to push the technology. compared to Schottky-barrier devices and low-
Thermal detector arrays will increase in size and dimensional solid devices, may become a decisive
improve in thermal sensitivity to a level satisfying argument in this case.
high performance applications at ambient tem- The fundamental performance limits of
perature. It is supposed that the silicon microbo- HgCdTe photodiodes have not been reached yet.
lometers arrays and the monolithic pyroelectric Continued development of the in situ vapour
208 A. Rogalski / Infrared Physics & Technology 43 (2002) 187210
phase epitaxy methods (MBE and MOCVD) will [5] E.S. Barr, The infrared pioneersII. Macedonio Melloni,
allow band-gap engineering heterojunction devices Infrared Phys. 2 (1962) 6773.
[6] E.S. Barr, The infrared pioneersIII. Samuel Pierpont
of increasing quality and complexity. Also con- Langley, Infrared Phys. 3 (1963) 195206.
tinued development of epitaxial growth on alter- [7] T.W. Case, Notes on the change of resistance of certain
native substrates such as silicon will reduce the substrates in light, Phys. Rev. 9 (1917) 305310.
cost of 2D arrays. Development of dual-band ar- [8] R.J. Cushman, Film-type infrared photoconductors, Proc.
ray will continue and three-band detectors will IRE 47 (1959) 14711475.
[9] P.R. Norton, Infrared detectors in the next millennium,
soon be demonstrated. To provide high resolution Proc. SPIE 3698 (1999) 652665.
spectroscopic imaging larger HgCdTe FPAs will [10] A. Rogalski, Infrared Detectors, Gordon and Breach
be used in Fourier-transform (FT) interferometers. Science Publishers, Amsterdam, 2000.
Photodiodes will replace photoresistors for detec- [11] M. Razeghi, Current status and future trends of infrared
tion out to 15 lm since they characterised by more detectors, Opto-Electron. Rev. 6 (1998) 155194.
[12] R.A. Wood, Uncooled microbolometer infrared sensor
linear response. arrays, in: P. Capper, C.T. Elliott (Eds.), Infrared Detec-
The situation concerning quantum well struc- tors and Emitters: Materials and Devices, Kluwer Aca-
tures and superlattices is not clear; however, un- demic Publishers, Boston, 2000, pp. 149174.
ique detection capabilities may arise from the [13] R.W. Whatmore, R. Watton, Pyroelectric materials and
low-dimensional solids. Situation of LWIR QWIPs devices, in: P. Capper, C.T. Elliott (Eds.), Infrared Detec-
tors and Emitters: Materials and Devices, Kluwer Aca-
is clear. The initial results show promise for the demic Publishers, Boston, 2000, pp. 99147.
growth of QWIPs on silicon wafers and appli- [14] D.A. Scribner, M.R. Kruer, J.M. Killiany, Infrared focal
cations for integration with silicon-based elec- plane array technology, Proc. IEEE 79 (1991) 6685.
tronics. It is expected that the QWIP hand-held, [15] P.R. Norton, Status of infrared detectors, Proc. SPIE 3379
cost-eective camera will nd imaging and spec- (1998) 102114.
[16] J. Piotrowski, W. Gawron, Ultimate performance of
troscopy applications in LWIR spectral band. infrared photodetectors and gure of merit of detector
Powerful possibilities of QWIP technology are material, Infrared Phys. Technol. 38 (1997) 6368.
connected with VLWIR FPA applications and [17] A. Rogalski, Assessment of HgCdTe photodiodes and
with multicolour detection. Three-band and four- quantum well infrared photoconductors for long wave-
band FPAs will soon be demonstrated in near the length focal plane arrays, Infrared Phys. Technol. 40 (1999)
279294.
future. [18] A. Rogalski, K. Adamiec, J. Rutkowski, Narrow-Gap
A new IR detector concept is microelectrome- Semiconductor Photodiodes, SPIE Press, Bellingham,
chanical structures (MEMSs). This technology 2000.
is a marriage of photolithography and mechanics. [19] L.J. Kozlowski, K. Vural, J.M. Arias, W.E. Tennant, R.E.
FPAs based on MEMS technology and a visible DeWames, Performance of HgCdTe, InGaAs and quan-
tum well GaAs/AlGaAs staring infrared focal plane arrays,
optical readout system may oer lower-cost LWIR Proc. SPIE 3182 (1997) 213.
imaging systems [70]. [20] W.F.M. Micklethwaite, A.J. Johnson, InSb: materials and
Finally, considerable development of signal- devices, in: P. Capper, C.T. Elliott (Eds.), Infrared Detec-
processing function into FPAs can be anticipated. tors and Emitters: Materials and Devices, Kluwer Aca-
demic Publishers, Boston, 2000, pp. 177204.
[21] J.T. Wimmers, D.S. Smith, Characteristics of InSb photo-
References voltaic detectors at 77 K and below, Proc. SPIE 364 (1983)
123131.
[1] W. Herschel, Experiments on the refrangibility of the [22] J.T. Wimmers, R.M. Davis, C.A. Niblack, D.S. Smith,
invisible rays of the Sun, Philos. Trans. Roy. Soc. London Indium antimonide detector technology at Cincinnati
90 (1800) 284. Electronics Corporation, Proc. SPIE 930 (1988) 125138.
[2] R.A. Smith, F.E. Jones, R.P. Chasmar, The Detection and [23] A.M. Fowler, I. Gatley, P. McIntyre, F.J. Vrba, A.
Measurement of Infrared Radiation, Clarendon, Oxford, Homan, ALADDIN, the 1024 1024 InSb array: design,
1958. description, and results,, Proc. SPIE 2816 (1996) 150160.
[3] P.W. Kruse, L.D. McGlauchlin, R.B. McQuistan, Ele- [24] J.B. Varesi, R.E. Bornfreund, A.C. Childs, W.A. Radford,
ments of Infrared Technology, Wiley, New York, 1962. K.D. Maranowski, J.M. Peterson, S.M. Johnson, L.M.
[4] E.S. Barr, Historical survey of the early development of the Giegerich, T.J. de Lyon, J.E. Jensen, Fabrication of high-
infrared spectral region, Am. J. Phys. 28 (1960) 4254. performance large-format MWIR focal plane arrays from
A. Rogalski / Infrared Physics & Technology 43 (2002) 187210 209
MBE-grown HgCdTe on 4 silicon substrates, J. Electron. [41] A. Rogalski, Quantum well infrared photodetectors among
Mater. 30 (2001) 566573. the other types of semiconductor infrared detectors,
[25] M.B. Reine, Photovoltaic detectors in MCT, in: P. Capper, Infrared Phys. Technol. 38 (1997) 295310.
C.T. Elliott (Eds.), Infrared Detectors and Emitters: [42] J.Y. Andersson, Dark current mechanisms and conditions
Materials and Devices, Kluwer Academic Publishers, of background radiation limitation of n-doped AlGaAs/
Boston, 2000, pp. 279312. GaAs quantum-well infrared detectors, J. Appl. Phys. 78
[26] Y. Nemirovsky, G. Bahir, Passivation of mercury cadmium (1995) 62986304.
telluride surfaces, J. Vac. Sci. Technol. A 7 (1989) 450459. [43] S. Gunapala, S. Bandara, J. Liu, M. Sundaram, Sharp
[27] W.E. Tennant, C.A. Cockrum, J.B. Gilpin, M.A. Kinch, infrared eyes: the journey of QWIPs from concept to large
M.B. Reine, R.P. Ruth, Key issues in HgCdTe-based focal inexpesive sensitive arrays in hand-held infrared cameras,
plane arrays: a industry perspective, J. Vac. Sci. Technol. B Opto-Electron. Rev. 7 (1999) 271282.
10 (1992) 13591369. [44] S.D. Gunapala, S.V. Bandara, J.K. Liu, E.M. Luong, S.B.
[28] A. Rogalski, R. Ciupa, Long wavelength HgCdTe photo- Rafol, J.M. Mumolo, D.Z. Ting, J.J. Bock, M.E. Ressler,
diodes: n -on-p versus p-on-n structures, J. Appl. Phys. 77 M.W. Werner, P.D. LeVan, R. Chehayeb, C.A. Kukkonen,
(1995) 35053512. M. Levy, N. LeVan, M.A. Fauci, Recent developments and
[29] T.J. de Lyon, R.D. Rajavel, J.A. Vigil, J.E. Jensen, O.K. applications of quantum well infrared photodetector focal
Wu, C.A. Cockrum, S.M. Johnson, G.M. Venzor, S.L. plane arrays, Opto-Electron. Rev. 9 (2001) 150163.
Bailey, I. Kasai, W.L. Ahlgren, M.S. Smith, Molecular- [45] M. Sundaram, S.C. Wang, 2-color QWIP FPAs, Proc.
beam epitaxial growth of HgCdTe infrared focal-plane SPIE 4028 (2000) 311317.
arrays on silicon substrates for midwave infrared applica- [46] R.A. Wood, N.A. Foss, Micromachined bolometer arrays
tions, J. Electron. Mater. 27 (1998) 550555. achieve low-cost imaging, Laser Focus World June (1993)
[30] M. Kimata, N. Tsubouchi, Schottky barrier photoemissive 101106.
detectors, in: A. Rogalski (Ed.), Infrared Photon Detec- [47] R.A. Wood, C.J. Han, P.W. Kruse, Integrated uncooled
tors, SPIE Optical Engineering Press, Bellingham, 1995, IR detector imaging arrays, Proc. IEEE Solid State Sensor
pp. 299349. and Actuator Workshop, Hilton Head Island, S.C., June,
[31] M. Kimata, Metal silicide Schottky infrared detector 1992, pp. 132135.
arrays, in: P. Capper, C.T. Elliott (Eds.), Infrared Detec- [48] T. Ishikawa, M. Ueno, K. Endo, Y. Nakaki, H. Hata, T.
tors and Emitters: Materials and Devices, Kluwer Aca- Sone, M. Kimata, Low-cost 320 240 uncooled IRFPA
demic Publishers, Boston, 2000, pp. 7798. using conventional silicon IC process, Opto-Electron. Rev.
[32] A.G.H. Perera, Semiconductor photoemissive structures 7 (1999) 297303.
for far infrared detection, in: M.H. Francombe (Ed.), [49] W. Radford, D. Murphy, A. Finch, K. Hay, A. Kennedy,
Handbook of Thin Devices, vol. 2, Academic Press, San M. Ray, A. Sayed, J. Wyles, R. Wyles, J. Varesi, E.
Diego, 2000, pp. 135170. Moody, F. Cheung, Sensitivity improvements in un-
[33] N. Sclar, Properties of doped silicon and germanium cooled microbolometer FPAs, Proc. SPIE 3698 (1999)
infrared detectors, Prog. Quant. Electr. 9 (1984) 149257. 119130.
[34] P.R. Norton, Infrared image sensors, Opt. Eng. 30 (1991) [50] M. Altman, B. Backer, M. Kohin, R. Blackwell, N. Butler,
16491663. J. Cullen, Lockheed Martin 640 480 uncooled microbo-
[35] F. Szmulowicz, F.L. Madarsz, Blocked impurity band lometer camera, Proc. SPIE 3698 (1999) 137143.
detectorsAn analytical model: Figures of merit, J. Appl. [51] C. Vedel, J.L. Martin, J.L. Ouvrier Buet, J.L. Tissot, M.
Phys. 62 (1987) 25332540. Vilan, J.J. Yon, Amorphous silicon based uncooled mic-
[36] J. Venzon, N. Lum, S. Freeman, G. Domingo, High- robolometer IRFPA, Proc. SPIE 3698 (1999) 276283.
background, longwave IR focal plane array development [52] C.M. Hanson, Uncooled thermal imaging at Texas Instru-
using Si:As IBC detectors, Proc. SPIE 2475 (1995) 3440. ments, Proc. SPIE 2020 (1993) 330339.
[37] S. Solomon, A. Tribble, N. Lum, J. Venzon, G. Domingo, [53] R.K. McEwen, P.A. Manning, European uncooled thermal
A. Hofman, M. Smith, High-background, longwave Si:As imaging sensors, Proc. SPIE 3698 (1999) 322337.
IBC 320 240 IR focal plane array, Proc. SPIE 2816 [54] R.D. Rajavel, D.M. Jamba, O.K. Wu, J.E. Jensen, J.A.
(1996) 161168. Wilson, E.A. Patten, K. Kosai, P. Goetz, G.R. Chapman,
[38] M.G. Stapelbroeck, D.H. Seib, J.E. Human, R.A. Flor- W.A. Radford, High performance HgCdTe two-color
ence, Large-format blocked-impurity-band focal plane infrared detectors grown by molecular beam epitaxy,
arrays for long wavelength infrared astronomy, Proc. SPIE J. Crystal Growth 175 (1997) 653658.
2476 (1995) 4148. [55] P. Mitra, S.L. Barnes, F.C. Case, M.B. Reine, P. ODette,
[39] B.F. Levine, Quantum-well infrared photodetectors, J. R. Starr, A. Hairston, K. Kuhler, M.H. Weiler, B.L.
Appl. Phys. 74 (1993) R1R81. Musicant, MOCVD of bandgap-engineered HgCdTe
[40] S.D. Gunapala, S.V. Bandara, Quantum well infrared pnNP dual-band infrared detector arrays, J. Electron.
photodetectors, in: M.H. Francombe (Ed.), Handbook of Mater. 26 (1997) 482487.
Thin Devices, vol. 2, Academic Press, San Diego, 2000, pp. [56] R.D. Rajavel, D.M. Jamba, J.E. Jensen, O.K. Wu, P.D.
6399. Brewer, J.A. Wilson, J.L. Johnson, E.A. Patten, K. Kasai,
210 A. Rogalski / Infrared Physics & Technology 43 (2002) 187210
J.T. Cauleld, P.M. Goetz, Molecular beam epitaxial [62] W.A. Beck, T.S. Faska, Current status of quantum well
growth and performance of integrated multispectral focal plane arrays, Proc. SPIE 2744 (1996) 193206.
HgCdTe photodiodes for the detection of mid-wave [63] T. Whitaker, Sanders QWIPs detect two color at once,
infrared radiation, J. Crystal Growth 184 (1998) 1272 Compd. Semicond. 5 (7) (1999) 4851.
1278. [64] S.D. Gunapala, S.V. Bandara, A. Sigh, J.K. Liu, S.B.
[57] R.D. Rajavel, D.M. Jamba, J.E. Jensen, O.K. Wu, Rafol, E.M. Luong, J.M. Mumolo, N.Q. Tran, J.D.
J.A. Wilson, J.L. Johnson, E.A. Patten, K. Kasai, P.M. Vincent, C.A. Shott, J. Long, P.D. Le Van, 89 and 14
Goetz, S.M. Johnson, Molecular beam epitaxial growth 15 lm two-color 640 486 quantum well infrared photo-
and performance of HgCdTe-based simultaneous-mode detector (QWIP) focal plane array camera, Proc. SPIE
two-color detectors, J. Electron. Mater. 27 (1998) 747 3698 (1999) 687697.
751. [65] P. Bois, E. Costard, X. Marcadet, E. Herniou, Develop-
[58] M.B. Reine, A. Hairston, P. ODette, S.P. Tobin, F.T.J. ment of quantum well infrared photodetectors in France,
Smith, B.L. Musicant, P. Mitra, F.C. Case, Simultaneous Infrared Phys. Technol. 42 (2001) 291300.
MW/LW dual-band MOCVD HgCdTe 64 64 FPAs, [66] M. Sundaram, S.C. Wang, M.F. Taylor, A. Reisinger, G.L.
Proc. SPIE 3379 (1998) 200212. Milne, K.B. Rei, R.E. Rose, R.R. Martin, Infrared Phys.
[59] S.M. Johnson, J.L. Johnson, W.J. Hamilton, D.B. Leon- Technol. 42 (2001) 301308.
ard, T.A. Strand, A.E. Patten, J.M. Peterson, J.H. [67] M.Z. Tidrow, J.C. Chiang, S.S. Li, K. Bacher, A high
Durham, V.K. Randall, T.J. de Lyon, J.E. Jensen, M.D. strain two-stack two-color quantum well infrared photo-
Gorwitz, HgCdZnTe quaternary materials for lattice- detector, Appl. Phys. Lett. 70 (1997) 859861.
matched two-color detectors, J. Electron. Mater. 29 [68] C.T. Elliott, Photoconductive and non-equilibrium devices
(2000) 680686. in HgCdTe and related alloys, in: P. Capper, C.T. Elliott
[60] P. Ferret, J.P. Zanatta, R. Hamelin, S. Cremer, A. Million, (Eds.), Infrared Detectors and Emitters: Materials and
M. Wolny, G. Destefanis, Status of the MBE technology at Devices, Kluwer Academic Publishers, Boston, 2000,
Leti LIR for the manufacturing of HgCdTe focal plane pp. 279312.
arrays, J. Electron. Mater. 29 (2000) 641647. [69] J. Piotrowski, M. Grudzie n, Z. Nowak, Z. Orman, J.
[61] W.E. Tennant, M. Thomas, L.J. Kozlowski, W.V. McLev- Pawluczyk, M. Romanis, W. Gawron, Uncooled photo-
ige, D.D. Edwall, M. Zandian, K. Spariosu, G. Hilde- voltaic Hg1x Cdx Te LWIR detectors, Proc. SPIE 4130
brandt, V. Gil, P. Ely, M. Muzilla, A. Stoltz, J.H. Dinan, A (2000) 175184.
novel simultaneous unipolar multispectral integrated tech- [70] T. Perazzo, M. Mao, O. Kwon, A. Majumdar, J.B. Varesi,
nology approach for HgCdTe IR detectors and focal plane P. Norton, Infrared vision using uncooled micro-optome-
arrays, J. Electron. Mater. 30 (2001) 590594. chanical camera, Appl. Phys. Lett. 74 (2001) 35673569.