Quantum Well Infrared Detectors

Wan-Ching Hung,Jie Zhang

Department of Electrical and Computer Engineering, University of Rochester, Rochester, New York, 14627-0231

Abstract:Quantum well infrared photodetectors (QWIP) have been developed very quickly over the past twenty years and large format focal plane arrays (FPA) with low noise equivalent temperature differences (NETD), high uniformity and operability have been achieved. In this paper, we make brief comparison of QWIPs with HgCdTe (MCT) detectors. Basic device physics and structures, characteristics, performance and benefits of QWIPs were demonstrated. The state-of-the-art of the QWIP FPA technology and its application were presented.

©Department of Electrical and Computer Engineering

I. INTRODUCTION

Infrared (IR) radiation is electromagnetic radiation of a wavelength longer than that of visible light, shorter than that of radio wave. It spans three orders of magnitude and has wavelengths between approximately 750 nm and 1 mm.IR detector technologies are very important nowadays both in military and civilian applications and have been widely investigated over the past century. Commercial applications of IR FPAs could cover astronomy, art history and archaeology, biological and medical systems, spectroscopy, fire control, surveillance and driver’s vision enhancement. The military applications could include night vision, rifle sight, surveillance, missile guidance, tracking, and interceptors. Fig 1 shows the atmosphere transmittance in the IR region which indicates the prospect applications in the astronomy, communication and military. [1-5]

Fig. 1 Plot of atmospheric transmittance in the infrared region of the electromagnetic spectrum [6]

In this paper, we focus on devices which involve IR excitation of carriers in quantum wells. A distinguishing feature of QW infrared detectors is that they can be implemented in chemically stable wide-band-gap materials as a result of the use of intersubband (intraband) processes. Till now, different types of quantum well infrared detectors have been achieved, among which, the technology of GaAs/AlGaAs multiple quantum well (MQW) detectors is the most mature. Rapid progress has been made recently in the performance of these detectors. Infrared focal plane arrays with high sensitivity, high uniformity, large format, and flexible wavelength are fabricated. Based on the high requirement in surveillance sensors and interceptor seekers, quantum well infrared photodetectors (QWIP) focal plane arrays (FPA) with lattice matched GaAs/AlGaAs material system which can provide high uniform, multicolor and long-wavelength operation is currently a hot topic in the worldwide. Here, the characteristics, performance and benefits of QWIP are demonstrated and discussed. The basic device physics and detector design of QWIP structures are given. The stability, reproducibility, yield, cost, maintenance, and manufacturability are also very important issues.

II. HISTORY: Overview of current IR detectors

The first IR photoconductor was developed by Case in 1917. Since then, many materials have been investigated in the IR field. Observing the IR detector technology development history, a simple theorem can be stated: All physics phenomena in the range of about 0.1-1 eV can be proposed for IR detectors. Among these effects are: change in electrical conductivity (bolometers), gas expansion (Golay cell), pyroelectricity (pyroelectric detectors), photon drag, Josephson effect (Josephson junctions, SQUIDs), internal emission (PtSi Schottky barriers), fundamental absorption (intrinsic photodetectors), impurity absorption (extrinsic photodetectors), low dimensional solids [superlattice (SL) and quantum well (QW) detectors], different type of phase transitions, etc. Since the initial proposal by Esaki and Tsu and the advent of molecular beam epitaxy (MBE), interest in semiconductor superlattices (SL’s) and quantum well (QW) structures has increased continuously over the years. As a result, a new class of materials and heterojunctions with unique electronics and optical properties has been developed.

Fig. 2 History of the development of IR detectors [3]

Fig. 2 gives approximate dates of significant development efforts for the materials mentioned. The modern IR detector technology started from the World War II. Till recently, photon IR detector technology combined with semiconductor material science, photolithography technology developed for integrated circuits and research progress in optical engineering have propelled extremely advances in IR capabilities in just a fraction of the last century.

Current cooled IR detector systems use material systems, such as HgCdTe (MCT), InSb, PtSi, and doped Si. The quantum well infrared photodetector is a relatively new technology for IR sensor applications. Among these cooled IR detector systems, PtSi FPAs are highly uniform and manufacturable, but have very low quantum efficiency and can only operate in the MWIR range. The InSb FPA technology is mature with very high sensitivity, but it also can only operate in the MWIR range. Neither PtSi nor InSb IR detectors have wavelength tunabiligty or multicolor capabilities. Doped silicon has a wide spectral bandwidth from 0.8 to 30 um, with no wavelength tenability or multicolor capability, and it can only operate at very low temperature (around 12 K). Both MCT and QWIPs offer high sensitivity with wavelength flexibility in the middle-wavelength infrared (MWIR), long-wavelength infrared (LWIR) and very-long-wavelenth infrared (VLWIR) regions, as well as multicolor capabilities.

Table 1. Comparison of infrared detectors [2]

Detector type / Advantages / Disadvantages
Thermal (thermopile, bolometers, pyroelectric) / Light, rugged, reliable, and low cost
Room temperature operation / Low detectivity at high frequncey
Slow response (ms order)
IV-VI
(PbS, PbSe, PbSnTe) / Easier to prepare
More stable materials / Very high thermal expansion coefficient, Large permittivity
Intrinsic II-VI (HgCdTe)
III-V (InGaAs, InAs, InSb, InAsSb)
Photon
Extrinsic
(Si:Ca, Si:As, Ge:Cu, Ge:Hg)
Free carriers
(PtSi, Pt2Si, IrSi)
Type I
(GaAs/AlGaAs, InGaAs/AlGaAs)
Quantum
Wells Type II
(InAs/InGaSb, InAs/InAsSb)
Quantum dots InAs/GaAs, InGaAs/InGaP, Ge/Si / Easy band gap tailoring
Well developed theory & expiment
Multicolor detectors
Good material and dopants
Advanced technology
Possible monolithic integration
Very-long-wavelength operation
Relatively simple technology
Low-cost, high yields
Large and close-packed 2D arrays
Matured material growth
Good uniformity over large area
Multicolor detectors
Low Auger recombination rate
Easy wavelength control
Normal incidence of light
Low thermal generation / Nonuniformity over large area
High cost in growth and processing
Surface instability
Heteroepitaxy with large
Lattice mismatch
Long wavelength cutoff limited
To 7 um (at 77K)
High thermal generation
Extremely low-temperature operation
Low quantum efficiency
Low-temperature operation
High thermal generation
Complicated design and growth
Complicated design and growth
Sensitive to the interfaces
Complicated design and growth

HCdTe (MCT) is a variable-gap semiconductor most often used in the production of IR photodetectors. It is nearly the most perfect IR detector material in terms of fundamental properties. But, since the conventional interband optical absorption involves photoexciting acrriers across the band gap Eg, i.e., promoting an electron to jump from the valence band to the conduction band. These photocarriers are then collected, in order to produce a photocurrent. This process is comparatively simpler in the visible light or ultra violet light spectrum ranges. However, for an infrared radiation whose wavelength ranges from 0.1 um to about 100 um, it requires an extremely small band gap which is in the order of 100 meV. Such small-band-gap materials are well known to be more difficult to grow, process, and fabricate into devices than are larger-band-gap semiconductors. In addition, the week Hg-Te band resulting in bulk, surface, and interface instabilities as well as the uniformity and yield are still unresolved issues. These difficulties thus motivate the study of novel artificial low effective band-gap materials which use quantum wells in large-band-gap (Eg>1 eV) semiconductors. [7]

III. DEVICE PHYSICS & CHARACTERISTICS

The QWIP is a semiconductor device using intersubband transitions within either the conduction band (n-type) or the valence band (p-type). The quantum well is formed by using an ultra thin layer of narrow band gap semiconductor (e.g. GaAs) sandwiched between two thin wider band gap semiconductors (e.g. AlGaAs) barrier layers. The motion of the charge carriers perpendicular to the layers becomes quantized so that localized two-dimensional (2-D) subbands of quantized states are formed inside the quantum well. When an optical beam is incident with an angle to the QWIP surface, an electron in the ground state of the quantum well absorbs an infrared photon and excites to a higher state during an intersubband optical transition.

A typical GaAs/AlGaAs QWIP consists of 30-50 quantum well periods. Using GaAs as the well region and AlGaAs as the barrier region, confined quantum well structures can be formed when the well width is small (less than an electron’s de Broglie wavelength). The thickness of the GaAs layer determines the well width, and the x value in AlxGa1-xAs determines the barrier height. The well region has one bound ground state and one or more excited states, depending on the barrier structure.

A. Classification


Fig.3 Energy Band Diagram for the type-I, type-II staggered, type-II misaligned, and type-III. [8]

A majority of the studies on quantum well infrared photodetectors (QWIPs) have been focused on GaAs/AlGaAs. However, other material such as n-type InGaAs/InAlAs, GaAs/GaInP, InGaAsP/InP, type II AlAs/Al0.5Ga0.5As, and SiGe/Si have also been investigated for QWIPs applications. In general, the hetero-interface quantum well structures may be classified into four categories: type-I, type-II staggered, type-II misaligned, and type-III, as shown in Fig. 3

Fig. 3 Type-I is the most used structure for QWIPs, which maybe fabricated from n-type GaAs/AlGaAs, InGaAs/InAlAs, GaSb/AlSb, GaAs/GaInP material. In a type-II quantum well structure, electrons and holes are confined in different semiconductor layers at their heterojunctions and super lattices. The type-III QWIPs involve the use of a zero band gap material such as HgCdTe.

B. Working principle

(1) Intersubband absorption

Fig. 4 is a typical figure for single well of type I. For instance, the low-band-gap material in the quantum well is GaAs and the high-band-gap material is AlGaAs. The intersubband transition energy between the lowest and first excited state is (1)

where Lw is the width of the quantum well, m* is the effective mass in the well.


Fig. 4 Band structure of quantum-well (depths Ec, and Ev. Intersubband absorption between electron levels E, and E2 or hole levels HI to H2 is schematically shown. [1]

Furthermore this transition has a large dipole matrix element (z) = 16L/92~0.18 L, and an integrated absorption strength of

Where c=NDLW is the two-dimensional density of carriers in the well, ND is the three-dimensional carrier density, LW is the number of doped well, nr is the index of refraction, θ’ is the angle between the direction of the optical beam and the surface normal, and f is the oscillator strength. This oscillator strength is very large and is given for this quantum well with infinitely high barriers as

(where z is the direction normal to the quantum well). By changing the quantum-well width L, this Intersubband transition energy can be varied over a wide range from the short wave infrared SWIR (A-2 pm), the medium wave infrared MWIR (n-4 pm>, through long-wave LWIR (n-10 pm) and into the very long-wave VWIR spectral regions (A > 14 pm). Since the oscillator strength only has a component along z, the optical electric field must also have a component parallel to z in order to induce an intersubband absorption; thus, normal incidence radiation will not be absorbed. Based on the above limitation, several optical coupling methods are adopted, such as gratings, random scattering and microlenses.

(2) Sequential resonant tunneling

Consider the application of electric fields to a multiquantum-well structure, and the tunneling escape and subsequent transport of these photoexcited electrons. The infrared absorption due to the intersubband transition from the bound ground state to the bound excited state is followed by the photoexcited electrons tunneling out of the well (as shown in Fig. 5). These photocarriers, which escape from the well, are transported by the electric field in the continuum above the barriers for an excited-state lifetime τL during which they travel a distance L (which is the mean free path for recapture back into the quantum wells) and thereby produce a photocurrent. The total current I can be written as I=Ist+Ith+Ipt, where Ist is the sequential resonant tunneling contribution, Ith is due to thermionic emission, and Ipt is phonon assisted tunneling.

In principle, due to the two-dimensional (2D) nature of the electron gas in the well, resonant tunneling is possible only when the energy levels in each well coincide. The presence of acoustic phonons and impurity scattering within each well, conservation of energy and momentum is possible provided that , where Vp is the potential difference per period between the adjacent wells and τ1 is the ground-state scattering time. Therefore, at small bias the electrons are able to conduct by ground-state resonant tunneling through the ground states of each well.

At high bias, , ground-state resonant tunneling is not possible and as a result negative differential resistance occurs. As each period breaks off from the resonant condition, the resistance across this period becomes much larger and a high-field domain forms. Any subsequent increase in the bias will appear across this domain until the ground level rises to within of the first excited level E2 in the next well whereupon the resonant tunneling condition is restored.

(3)Bound-to-bound State

All QWIPs are based on ‘‘band gap engineering’’ of layered structures of wide-band-gap (relative to thermal IR energies) materials. These structures are designed in such a way that the energy separation between two selected states in the structure matches the energy of the infrared photons to be detected. By using different well widths and barrier heights, several QWIP configurations have been reported based on transitions from bound-to-bound, bound-to-continuum states, bound-to-quasibound states, and bound-to-miniband states.


Fig. 5 Schematic energy band diagram [9]

For Fig. 5 (a) after absorption of the infrared photon, the photoexcited carrier can either be transported along the quantum well direction (with an applied parallel bias voltage), orperpendicular to the wells (with an applied field along the growth direction). However, as far as detection is concerned, perpendicular transport is superior to parallel transporting since the difference between the excited-state and ground-state mobilities is much larger in the latter case, and thus the photo response is substantially greater. The photocurrent from this detector arises solely from the high-field domain, since only from this region can the photoexcited carriers tunnel out of the well and escape. But the escape probability by tunneling is not high due to the confinement of the excited level. We can thus express the photocurrent Ip as

(2)

where np is the number of photogenerated carriers/cm3and v is the transport velocity along the super lattice. Furthermore, the dark current is much lower since the heterobarriers effectively block the transport of the carriers in the doped quantum-well ground state. For this reason QWIPs based on the escape and perpendicular transport of photoexcited carriers are to be preferred.

(4)Bound-to-Continuum State

By decreasing the size of the quantum well, the strong oscillator strength of the excited bound state can be pushed up into the continuum resulting in a strong bound-to-continuum state absorption. This extended state structure has the major advantage that the photoexcited electron can escape from the quantum well without tunneling through the energy barrier. Thus, the bias voltage required for the photoelectron to efficiently escape from the well can be dramatically reduced, strongly lowering the dark current. In addition, the barrier thickness can now be substantially increased thereby further reducing the ground-state sequential tunneling by many orders of magnitude.

C. Characteristics and performance

(1) Quantum efficiency η

Fig. 6. Quantum efficiency versus wavelength for a HgCdTe photodiode and GaAs/AlGaAs QWIP detector with similar cutoffs. [2]

The quantum efficiency value describes how well the detector is coupled to the radiation to be detected. It is usually defined as the number of electron-hole pairs generated per incident photon. Due to the intersubband transition in the conduction band, the n-typed QWIP detection mechanism requires photons with a non-normal angle of incidence to provide proper polarization for photon absorption. The absorption quantum efficiency of QWIP is relatively small with a 2D grating. Fig. 6 compares the spectral η of a HgCdTe photodiode to that of a QWIP. A higher bias voltage is used to boost η. However, an increase in the reverse bias voltage also causes an increase of the leakage current. This limits any potential improvement in the system performance. New grating designs are under study to improveη, such as an enhanced QWIP, antenna gratings, and corrugated gratings. [10, 11] It is well known that using a smaller number of quantum wells and bound-to-continuum structures, increased optical gain and improved detector performance at low temperatures are possible. Tidrow presented [12] a highperformance QWIP consisting of only three quantum wells with the conversion efficiencies up to 29% at a bias voltage of 20.8 V and a peak wavelength of 8.5 mm.