rus
Issue #1/2022
N. A. Kulchitsky, A. V. Naumov, V. V. Startsev
«Quantum Wells» Cooled Infrared Photodetectors – the State and Prospects of Development
«Quantum Wells» Cooled Infrared Photodetectors – the State and Prospects of Development
DOI: 10.22184/1993-7296.FRos.2022.16.1.22.36
In this paper the quantum wells infrared detectors (QWIP) of thermal imaging equipment are considered. These devices are in demand in the security surveillance systems and complexes, as well as in personal night vision and security systems. The QWs cooled photodetectors (PDs) for the spectral range 3–5 µm and 8–12 µm are of the special interest. A comparison of thermal detectors of various types from different world manufacturers is presented. An expert forecast of changes in the dynamics of market growth and its development trends is given.
In this paper the quantum wells infrared detectors (QWIP) of thermal imaging equipment are considered. These devices are in demand in the security surveillance systems and complexes, as well as in personal night vision and security systems. The QWs cooled photodetectors (PDs) for the spectral range 3–5 µm and 8–12 µm are of the special interest. A comparison of thermal detectors of various types from different world manufacturers is presented. An expert forecast of changes in the dynamics of market growth and its development trends is given.
Теги: photodetectors quantum wells thermal imagers квантовые ямы тепловизоры фотоприемные устройства
“Quantum Wells” Cooled Infrared Photodetectors – the State and Prospects of Development
N. A. Kulchitsky , A. V. Naumov, V. V. Startsev
Moscow Technological University (MIREA),
Lomonosov Moscow State University,
State Scientific Center of the Russian Federation, Moscow, Russia
Orion RPA JSC, Moscow, Russia
RDB Astrohn JSC, Lytkarino, Moscow region, Russia
In this paper the quantum wells infrared detectors (QWIP) of thermal imaging equipment are considered. These devices are in demand in the security surveillance systems and complexes, as well as in personal night vision and security systems. The QWs cooled photodetectors (PDs) for the spectral range 3–5 µm and 8–12 µm are of the special interest. A comparison of thermal detectors of various types from different world manufacturers is presented. An expert forecast of changes in the dynamics of market growth and its development trends is given.
Keywords: thermal imagers, quantum wells, photodetectors
Received on: 10.01.2022
Accepted on: 28.01.2022
In recent years, the pace of development of new generation thermal imaging equipment has accelerated noticeably. Devices using thermal imagers allow good performance in conditions of poor visibility, provide a search for and tracking of targets, etc. The spheres of application are also infrared astronomy, medicine, metrology, military equipment [1,2]. According to Yole Development, the market for thermal imaging cameras for IR systems (civilian and military) amounted to USD6.2 billion in 2021, will reach USD 8.7 billion in 2026 (Fig. 1)
Basic semiconductor photosensitive materials and their features.
Back in the late 90s of the last century various reviews of the development of cooled IR photodetectors included references to a rather extensive list of devices based on various photosensitive materials, such as impurity silicon, platinum silicides, lead-tin chalcogenides, narrow-gap solid solutions InAsSb, HgCdTe, photodetector structures based on quantum wells (QW), etc. By the present day the list has been reduced to four main technologies which are the following: 1) technology of HgCdTe solid solutions; 2) technologies of InGaAs compounds; 3) technologies of quantum wells quantum-dimensional structures (or QWIP – quantum well infrared photodetector) and its later derivatives, technology of quantum dot structures (QDIP) and 4) technology of type 2 superlattices (T2SL) in the InAs / GaSb system [2, 3]. The share of QWIP in the total production volumes today is quite small (Fig. 2) [1], but the dynamics of development of QWIP can be assessed, as it seems to us in Fig. 3
Comparison of properties of various photodetector materials
The solid solutions of HgCdTe (cadmium- mercury-tellurium, CMT) are known as one of the most promising photosensitive materials. However, these expectations have so far been only partly justified. While very high performance has been achieved in the creation of an PDs for the range of 3–5 µm based on CMT, for the range of 8–12 µm and longer wavelengths with acceptable characteristics, it is still possible to produce only linear and matrix photodetectors of a relatively small format. The reason for this was the high inhomogeneity of the CMT characteristics for the long-wavelength range over the crystal area, as well as the high sensitivity with respect to various external influencing factors (temperature, radiation, etc.). As a result, the technical feasibility of using long-wavelength CMT for 3rd generation photodetectors is limited: such photodetectors turn out to be extremely expensive even for special applications. The possibility of this problem was recognized by the developers of IR systems more than 20 years ago. At the same time, the search began for alternative photosensitive materials that would make it possible to overcome the noted shortcomings of CMT and preserve its advantages to the maximum. Numerous materials were tested during the search. The conducted searches have led to the emergence of a new class of photosensitive materials – quantum-dimensional heterostructures or structures with quantum wells, QWs [3,6]. The size quantization effect underlies the operation not only of QWIP, but also of quantum dots devices, QDIP, and type 2 superlattices, T2SL, which are more complex in terms of design and technology.
However, the devices of the last two types are still at the stage of creating experimental samples, and the technological problems that stand in the way of their improvement today do not yet make it possible to unequivocally determine their prospects. Thus, in devices based on QDIP, for which theoretical estimates predict significant advantages over QWIP according to the main photoelectric characteristics, today it is possible to realize, mainly, only one of them, which consists in the absence of selectivity with respect to the signal radiation polarization. [5, 6] The main reason for this situation is the lack of technological capability to reliably control the size and uniformity of the distribution of quantum dots in the structure. Instruments based on T2SL show good results: they have a higher quantum efficiency, a longer lifetime and, as a result, a higher operating temperature than QWIP, they can operate in the photovoltaic mode. However, T2SL The structures still have several times higher residual impurity concentration in the active region compared to CMT, which causes excess tunneling currents. The problem of reliable passivation of the lateral surface of mesa diodes remains unresolved, which causes surface leakage and, in combination with tunnel leakage, leads to a deterioration in the threshold characteristics of the PDs. As a result, today the T2SL PDs, in terms of their characteristics, they are close to their CMT counterparts, but are still far from instrumental implementation on an industrial scale [7].
This article considers some properties and characteristics of quantum wells (QWs or QWIP structures) cooled thermal detectors sensitive in the infrared (IR) range with a long-wave boundary from 3 to 20 microns, as well as possible prospects for the development of the market.
QWIP principle of operation
The operation of most photonic PDs is based on one common phenomenon: a current carrier (electron, hole, ion), which is in a bound (neutral) state and does not participate in the creation of an electric current, passes into a free state when a photon is absorbed, i. e., it can move under the influence of an electric field and create an electric current. Such PDs can be considered as a medium containing many potential wells, in which there is at least one bound energy state occupied by charge carriers. The principle of formation of a photosensitive medium from a set of quantum wells has been known for a long time (classical quantum wells were theoretically studied at the dawn of quantum mechanics) [4, 5]. PDs based on impurity photoconductivity became one of the first analogues of such a medium. However, impurity atoms in silicon (germanium, gallium arsenide, etc.) set strictly fixed energy characteristics of the quantum well, which limits the possibility of adjusting the PDs material to the requirements of a particular problem. Therefore, PDs developers have always been interested in materials in which it is possible, by controlling the “geometric” parameters of the well: the thickness, the height of the barriers, to change the energy position of the levels in it. And thanks to precision methods of epitaxy – metal-organic gas phase epitaxy (MOGPE) and molecular beam epitaxy (MBE), such materials have appeared. The most successful implementation of this idea turned out to be in the system of GaAs / AlGaAs materials. Fig. 4 and 5 explain the principle of operation of the QWIP. Structurally, the QW structure is a multilayer epitaxial structure in which GaAs wells layers alternate with AlGaAs barriers (Fig. 4). These materials have different values of the band gap. QWIP are the devices based on majority charge carriers, i. e., those in which the carriers, which determine the type of conductivity of the material, also determine its photosensitivity.
Thus, in electronic structures, quantum wells are doped with a donor impurity, in hole structures, with an acceptor impurity, and photosensitivity arises due to the optical excitation of electrons or holes from a bound state to the conduction band or valence band, respectively. Typical thicknesses of GaAs layers, at which 1–2 levels are located in the well, have a value of 40–50 A. The thicknesses of AlGaAs barriers is 400–500 A. To fill the wells with electrons (holes), GaAs layers doped with silicon (beryllium) to concentrations of about 1018 cm–3. With sufficient cooling, the electrons mainly fill the lower energy level in the wells. There are relatively few mobile carriers in the conduction band above the barriers, and their concentration is determined by thermal generation. Therefore, mainly these carriers determine the ‘dark’ conductivity of a QW structure. When a QW structure is irradiated with photons whose energy exceeds the energy distance between the lower level in the well and the upper level or the top of the barrier, an electron transition can occur between these levels or between the lower level and the bottom of the barrier conduction band [5]. Further, the electron becomes free and can move under the action of an electric field, thus creating a photocurrent (Fig. 5).
According to theoretical calculations, the sensitivity of a QWIP depends on the polarization of the absorbed photons, i. e. for efficient absorption, the electric polarization vector of a photon should be perpendicular to the layers of the structure with QWs [5]. To control the orientation of the polarization vector relative to the words with QWs, radiation input devices in the form of diffraction gratings (usually phase gratings) began to be formed on the surface of photosensitive elements, which make it possible to provide the required direction of radiation propagation. The use of diffraction gratings turns out to be useful for increasing the quantum efficiency of the PDs. The principle of operation of a diffraction grating is shown in Fig. 6
When creating multi-element PDs, it becomes necessary to read signals from sensitive elements. Typically, in cooled multi-element PDs, including QWs, signals are read using silicon microcircuits connected to the photodetector section (Fig. 7) [6,9]. Thus, the PDs includes a matrix of photosensitive elements coupled to the readout circuit. Cooling is provided by a micro-cryogenic cooling system (MCS) operating on the Stirling cycle.
The emergence of QWIPs and their applications.
The first reports on imaging with photodetectors based on QW structures appeared in 1991 [5]. A group of researchers from Bell Laboratories (USA) (B. F. Levine et al.) demonstrated a thermal image obtained with a QW-based linear photodetector (range 8–12 µm) using optical-mechanical scanning. By that time, only 4 years had passed since the first experimental observation of the photoconductivity effect in QWs. In 1997, a thermal imaging camera with a PD was demonstrated, in which a matrix photodetector based on a QW with a format of 256 × 256 elements was used [6]. To date, many different IR systems have already been created on the basis of QWIP, which are used in France, Germany, the USA, etc. [6, 7]. Today, reports about commercially available thermal imaging devices with a photodetector matrix of 640 × 480 elements based on QWs for the spectral ranges of both 3–5 μm and 8–12 μm are already commonplace. There are also reports on the creation of photodetector matrices with a format of more than 1 000 × 1 000 elements, multispectral photodetectors [8,9]. In this case, the advantages of structures with QWs are most convincingly manifested in the spectral range of 8–12 μm and longer wavelengths. Thus, today the QW technology is one of the main technologies due to which IR systems are equipped with the cooled PDs for various purposes. At the same time, the greatest demand for QWIPs is observed in IR systems, which require PDs in the long-wavelength range of a large format, as well as multispectral devices [1, 6].
Comparative analysis of cooled PDs of the long-wavelength spectrum
longwave thermal imaging cooled photodetectors are designed and manufactured for the spectrum from 8 to 14 µm based on several technologies. Cooled long-wavelength PDs are produced on the basis of heterostructures of ternary compounds of cadmium-mercury-tellurium (CMT) and on the basis of heterostructures with QWs. The main manufacturers are presented in table. one.
The characteristics of long-wavelength CMT and QW are different while each has its pros and cons. In Russia, over the past few years, several research and development works have been carried out to develop and master the PDs of the long-wave range. The main developers were Orion RPA, RDB Astrohn and ISP SB RAS. Matrix PDs with a resolution of 640 × 480 px for the spectrum range of 8–14 µm are currently being produced at RDB Astrohn and Orion RPA. RDB Astrohn produces uncooled PDs based on Astrohn‑64017 type microbolometer arrays and cooled PDs based on Astrohn‑640KYa20 type QW heterostructures based on QW heterostructures industrially produced at Svetlana-Rost JSC (St. Petersburg). RDB Astrohn also produces cooled PDs based on heterostructures with CMT of the Astrohn‑640KRT15A810 type based on heterostructures with CMT produced at the ISP SB RAS. The PDs produced by Orion RPA for the range of 8–12 µm on CMT have a resolution of 320 × 240 px.
The main advantages of cooled PDs are their speed (up to 200 frames per second) and high sensitivity. It should be noted that the sensitivity of long-wave cooled PDs (about 35 mK) is inferior to short-wave receivers (about 15–20 mK) by about 2 times and is comparable to the sensitivity of microbolometric arrays.
The most famous receivers in Russia are CMT Scorpio LW and QW Sirius LW (France). Let us compare the receivers LYNRED (France), used in the Russian Federation, based on the Scorpio-LW type CMT connection, on Sirius LW quantum wells and RDB Astrohn quantum wells receiver, Table 2
French receivers show a difference in the number of valid pixels between CMT and QWIP receivers in favor of the latter (99.9%). This is primarily due to the physics of the quantum well heterostructures themselves and the mature growth technology. Numerous publications note the great stability and repeatability of technological processes for creating structures based on quantum wells. This is also confirmed by the indicators of domestic PDs based on quantum wells grown at the Svetlana-Rost JSC for the receivers by RDB Astrohn.
The CMT receiver has a slightly larger spectral range (7.7–9.3 µm) with a peak in the 8.5 µm region and 0.7 µm in both directions. For QW receivers with a peak of 8.5 µm, the range is narrower, for Astrohn it is ±0.3 µm in both directions, for Sirius it is ±0.5 µm in both directions.
Also noteworthy is the high uniformity of the pixel sensitivity on the Astrohn‑640KYa20 receiver matrix, which is clearly seen in the histogram in Fig. 8
The CMT receivers have better performance in sensitivity both in the range of 3–5 microns and in the range of 7–14 microns. However, the level of usable pixels on long-wavelength CMT matrices is 98%, which is unacceptable for many tasks. The QW receivers have industrial quality repeatability, good uniformity of pixel sensitivity on the plate and a high (99.9%) level of usable pixels. Comparison of QW receivers by LYNRED and RDB Astrohn shows the identity of the main indicators both in the pixel size (20 μm) and in the average pixel sensitivity on the matrix. (Fig.9)
Situation in Russia
In Russia, the problem of providing development and production in the field of IR photoelectronics with domestic photosensitive materials is quite acute. The development of photodetectors for various purposes in Russia is carried out by a number of enterprises concentrated in Shvabe JSC and Ruselectronics JSC, in the Russian Academy of Sciences, as well as private enterprises. The main suppliers of thermal imaging PDs and MPDs are Orion RPA JSC and MZ Sapfir JSC, which are part of Shvabe JSC, as well as the private enterprise RDB Astrohn JSC. Orion RPA JSC develops and manufactures cooled and uncooled photodetectors, the material science base of the enterprise is focused on molecular beam epitaxy. MZ Sapfir JSC produces cooled and uncooled PDs based on Si, Ge, InSb, CdHgTe. The Institute of Semiconductor Physics of the Siberian Branch of the Russian Academy of Sciences is developing PDs based on CdHgTe, InAs, microbolometers and quantum wells. RDB Astrohn JSC (Lytkarino) develops and manufactures civil thermal imaging devices based on uncooled PDs of its own production, as well as cooled QWIPs together with Svetlana-Rost and the Astrohn-MKS500 microrefrigeration system. [9]
Thus, at present, Russia has developed PDs of the second generation, as well as large-format and IR viewing PDs. The results achieved are close in terms of their indicators to the world level [9, 10].
Trends in the development of QWIPs
Despite the impressive progress in the field of development of QWIPs, today there are a number of problems that limit the range of applications of this technology. The development of a cooled PDs of the 3rd generation aims at obtaining a large-format device with a threshold sensitivity (in one or several spectral ranges) close to the limit at the highest possible operating temperature. It is precisely in terms of threshold sensitivity and operating temperature that QWIP are inferior today to analogues based on CMT. In order to improve the indicated characteristics of QWIPs, it is necessary to carry out fundamental studies in order to clarify the physical mechanisms underlying the operation of QWIPs, as well as to search for constructive and technological solutions that take into account new research results. The most important factor determining the threshold sensitivity of any PD is the quantum efficiency. According to this indicator, QWIPs are still inferior to their narrow-gap counterparts. However, as shown in several publications, such a lag can be minimized by improving the designs of PD sensitive elements [5,6].
The operating temperature, in addition to the quantum efficiency, also depends on the carrier lifetime in the active region and on the design of the QW. As already noted, structures on QWs are similar in nature to impurity photoresistors. This similarity makes it possible to use the same methods to increase the QW lifetime as in impurity photoresistors, in which the probability of carrier capture by an impurity center can be reduced by several orders of magnitude by converting this center from neutral to charged [5]. A similar technique can also be applied in a structure with a QW by doping not a quantum well, but a barrier, as a result of which the well will be negatively charged and free carriers will be repelled from it, i. e. the probability of being trapped in the pit will decrease, and the lifetime will increase. To implement this method of increasing the lifetime, both constructive and technological improvements are required in the creation of QWIPs.
Also, the sharpness of the barrier – well boundaries can significantly affect the lifetime, the dark current, and, as a result, the threshold sensitivity [5, 8]. In viewing PDs, the homogeneity of sensitivity across the matrix and the number of defective elements have acquired particular relevance.
One of the advantages of QWIPs is their potentially high resistance to various kinds of external influencing factors. First, this is resistance to ionizing radiation and laser exposure to high energy. However, there are practically no reports on research in these areas in foreign periodicals. In view of the particular importance of the indicated properties of the PDs for various special applications, it is important to study the resistance of the QWIPs.
The physical and technological problems of creating QWIPs, noted above, characterize this area as a whole, however, they are of particular importance for Russia, where the backlog of domestic applied science in this area remains [11].
This requires both basic researches in the field of physics of low-dimensional semiconductors and IR-photonics [12–14], and perspectives of applications in different fields [15–17].
Conclusion
The last decade can be characterized as a period of rapid development of QW technologies and devices based on it. Such PDs are developed by almost all leading companies. The increase in the information content and the probability of detection and recognition achieved with their use, as well as the compactness of complex optoelectronic systems, are the main driving forces for the development of this direction. In the next decade, two-spectrum PDs, along with megapixels, will become the main commercially available products in IR photoelectronics. To implement two-spectral PDs sensitive in the ranges of 3–5 and 8–12 µm, QW-based technologies are used.
It seems to us that Russia continues to lag behind the leading foreign companies in the development of PDs, and simply financial, even if significant, “injections” into existing organizational structures are not enough to overcome it. The task of achieving parity with the world level, as well as creating scientific, technical, and technological groundwork for the development of PDs, can be solved by a program-target method that provides for a set of technical, financial, and organizational measures within the framework of a public-private partnership mechanism.
ABOUT THE AUTHORS
Kulchitsky Nikolai Alexandrovich, Dr. of Sciences (Eng.), e-mail: n.kulchitsky@gmail.com, Professor, Moscow Technological University (MIREA), Chief Specialist, State Scientific Center of the Russian Federation, Orion Research and Production Association JSC, Moscow, Russia.
ORCID: 0000-0003-4664-4891
Naumov Arkady Valerievich Analyst Engineer, Astrohn Optical-Mechanical Design Bureau JSC, https://astrohn.ru, Lytkarino, Moscow region, Russia.
ORCID: 0000–0001–6081–8304
Startsev Vadim Valerievich, Cand. of Sciences (Eng.), Chief Designer, Astrohn Optical -Mechanical Design Bureau, https://astrohn.ru, Lytkarino, Moscow. region, Russia.
ORCID: 0000-0002-2800-544
N. A. Kulchitsky , A. V. Naumov, V. V. Startsev
Moscow Technological University (MIREA),
Lomonosov Moscow State University,
State Scientific Center of the Russian Federation, Moscow, Russia
Orion RPA JSC, Moscow, Russia
RDB Astrohn JSC, Lytkarino, Moscow region, Russia
In this paper the quantum wells infrared detectors (QWIP) of thermal imaging equipment are considered. These devices are in demand in the security surveillance systems and complexes, as well as in personal night vision and security systems. The QWs cooled photodetectors (PDs) for the spectral range 3–5 µm and 8–12 µm are of the special interest. A comparison of thermal detectors of various types from different world manufacturers is presented. An expert forecast of changes in the dynamics of market growth and its development trends is given.
Keywords: thermal imagers, quantum wells, photodetectors
Received on: 10.01.2022
Accepted on: 28.01.2022
In recent years, the pace of development of new generation thermal imaging equipment has accelerated noticeably. Devices using thermal imagers allow good performance in conditions of poor visibility, provide a search for and tracking of targets, etc. The spheres of application are also infrared astronomy, medicine, metrology, military equipment [1,2]. According to Yole Development, the market for thermal imaging cameras for IR systems (civilian and military) amounted to USD6.2 billion in 2021, will reach USD 8.7 billion in 2026 (Fig. 1)
Basic semiconductor photosensitive materials and their features.
Back in the late 90s of the last century various reviews of the development of cooled IR photodetectors included references to a rather extensive list of devices based on various photosensitive materials, such as impurity silicon, platinum silicides, lead-tin chalcogenides, narrow-gap solid solutions InAsSb, HgCdTe, photodetector structures based on quantum wells (QW), etc. By the present day the list has been reduced to four main technologies which are the following: 1) technology of HgCdTe solid solutions; 2) technologies of InGaAs compounds; 3) technologies of quantum wells quantum-dimensional structures (or QWIP – quantum well infrared photodetector) and its later derivatives, technology of quantum dot structures (QDIP) and 4) technology of type 2 superlattices (T2SL) in the InAs / GaSb system [2, 3]. The share of QWIP in the total production volumes today is quite small (Fig. 2) [1], but the dynamics of development of QWIP can be assessed, as it seems to us in Fig. 3
Comparison of properties of various photodetector materials
The solid solutions of HgCdTe (cadmium- mercury-tellurium, CMT) are known as one of the most promising photosensitive materials. However, these expectations have so far been only partly justified. While very high performance has been achieved in the creation of an PDs for the range of 3–5 µm based on CMT, for the range of 8–12 µm and longer wavelengths with acceptable characteristics, it is still possible to produce only linear and matrix photodetectors of a relatively small format. The reason for this was the high inhomogeneity of the CMT characteristics for the long-wavelength range over the crystal area, as well as the high sensitivity with respect to various external influencing factors (temperature, radiation, etc.). As a result, the technical feasibility of using long-wavelength CMT for 3rd generation photodetectors is limited: such photodetectors turn out to be extremely expensive even for special applications. The possibility of this problem was recognized by the developers of IR systems more than 20 years ago. At the same time, the search began for alternative photosensitive materials that would make it possible to overcome the noted shortcomings of CMT and preserve its advantages to the maximum. Numerous materials were tested during the search. The conducted searches have led to the emergence of a new class of photosensitive materials – quantum-dimensional heterostructures or structures with quantum wells, QWs [3,6]. The size quantization effect underlies the operation not only of QWIP, but also of quantum dots devices, QDIP, and type 2 superlattices, T2SL, which are more complex in terms of design and technology.
However, the devices of the last two types are still at the stage of creating experimental samples, and the technological problems that stand in the way of their improvement today do not yet make it possible to unequivocally determine their prospects. Thus, in devices based on QDIP, for which theoretical estimates predict significant advantages over QWIP according to the main photoelectric characteristics, today it is possible to realize, mainly, only one of them, which consists in the absence of selectivity with respect to the signal radiation polarization. [5, 6] The main reason for this situation is the lack of technological capability to reliably control the size and uniformity of the distribution of quantum dots in the structure. Instruments based on T2SL show good results: they have a higher quantum efficiency, a longer lifetime and, as a result, a higher operating temperature than QWIP, they can operate in the photovoltaic mode. However, T2SL The structures still have several times higher residual impurity concentration in the active region compared to CMT, which causes excess tunneling currents. The problem of reliable passivation of the lateral surface of mesa diodes remains unresolved, which causes surface leakage and, in combination with tunnel leakage, leads to a deterioration in the threshold characteristics of the PDs. As a result, today the T2SL PDs, in terms of their characteristics, they are close to their CMT counterparts, but are still far from instrumental implementation on an industrial scale [7].
This article considers some properties and characteristics of quantum wells (QWs or QWIP structures) cooled thermal detectors sensitive in the infrared (IR) range with a long-wave boundary from 3 to 20 microns, as well as possible prospects for the development of the market.
QWIP principle of operation
The operation of most photonic PDs is based on one common phenomenon: a current carrier (electron, hole, ion), which is in a bound (neutral) state and does not participate in the creation of an electric current, passes into a free state when a photon is absorbed, i. e., it can move under the influence of an electric field and create an electric current. Such PDs can be considered as a medium containing many potential wells, in which there is at least one bound energy state occupied by charge carriers. The principle of formation of a photosensitive medium from a set of quantum wells has been known for a long time (classical quantum wells were theoretically studied at the dawn of quantum mechanics) [4, 5]. PDs based on impurity photoconductivity became one of the first analogues of such a medium. However, impurity atoms in silicon (germanium, gallium arsenide, etc.) set strictly fixed energy characteristics of the quantum well, which limits the possibility of adjusting the PDs material to the requirements of a particular problem. Therefore, PDs developers have always been interested in materials in which it is possible, by controlling the “geometric” parameters of the well: the thickness, the height of the barriers, to change the energy position of the levels in it. And thanks to precision methods of epitaxy – metal-organic gas phase epitaxy (MOGPE) and molecular beam epitaxy (MBE), such materials have appeared. The most successful implementation of this idea turned out to be in the system of GaAs / AlGaAs materials. Fig. 4 and 5 explain the principle of operation of the QWIP. Structurally, the QW structure is a multilayer epitaxial structure in which GaAs wells layers alternate with AlGaAs barriers (Fig. 4). These materials have different values of the band gap. QWIP are the devices based on majority charge carriers, i. e., those in which the carriers, which determine the type of conductivity of the material, also determine its photosensitivity.
Thus, in electronic structures, quantum wells are doped with a donor impurity, in hole structures, with an acceptor impurity, and photosensitivity arises due to the optical excitation of electrons or holes from a bound state to the conduction band or valence band, respectively. Typical thicknesses of GaAs layers, at which 1–2 levels are located in the well, have a value of 40–50 A. The thicknesses of AlGaAs barriers is 400–500 A. To fill the wells with electrons (holes), GaAs layers doped with silicon (beryllium) to concentrations of about 1018 cm–3. With sufficient cooling, the electrons mainly fill the lower energy level in the wells. There are relatively few mobile carriers in the conduction band above the barriers, and their concentration is determined by thermal generation. Therefore, mainly these carriers determine the ‘dark’ conductivity of a QW structure. When a QW structure is irradiated with photons whose energy exceeds the energy distance between the lower level in the well and the upper level or the top of the barrier, an electron transition can occur between these levels or between the lower level and the bottom of the barrier conduction band [5]. Further, the electron becomes free and can move under the action of an electric field, thus creating a photocurrent (Fig. 5).
According to theoretical calculations, the sensitivity of a QWIP depends on the polarization of the absorbed photons, i. e. for efficient absorption, the electric polarization vector of a photon should be perpendicular to the layers of the structure with QWs [5]. To control the orientation of the polarization vector relative to the words with QWs, radiation input devices in the form of diffraction gratings (usually phase gratings) began to be formed on the surface of photosensitive elements, which make it possible to provide the required direction of radiation propagation. The use of diffraction gratings turns out to be useful for increasing the quantum efficiency of the PDs. The principle of operation of a diffraction grating is shown in Fig. 6
When creating multi-element PDs, it becomes necessary to read signals from sensitive elements. Typically, in cooled multi-element PDs, including QWs, signals are read using silicon microcircuits connected to the photodetector section (Fig. 7) [6,9]. Thus, the PDs includes a matrix of photosensitive elements coupled to the readout circuit. Cooling is provided by a micro-cryogenic cooling system (MCS) operating on the Stirling cycle.
The emergence of QWIPs and their applications.
The first reports on imaging with photodetectors based on QW structures appeared in 1991 [5]. A group of researchers from Bell Laboratories (USA) (B. F. Levine et al.) demonstrated a thermal image obtained with a QW-based linear photodetector (range 8–12 µm) using optical-mechanical scanning. By that time, only 4 years had passed since the first experimental observation of the photoconductivity effect in QWs. In 1997, a thermal imaging camera with a PD was demonstrated, in which a matrix photodetector based on a QW with a format of 256 × 256 elements was used [6]. To date, many different IR systems have already been created on the basis of QWIP, which are used in France, Germany, the USA, etc. [6, 7]. Today, reports about commercially available thermal imaging devices with a photodetector matrix of 640 × 480 elements based on QWs for the spectral ranges of both 3–5 μm and 8–12 μm are already commonplace. There are also reports on the creation of photodetector matrices with a format of more than 1 000 × 1 000 elements, multispectral photodetectors [8,9]. In this case, the advantages of structures with QWs are most convincingly manifested in the spectral range of 8–12 μm and longer wavelengths. Thus, today the QW technology is one of the main technologies due to which IR systems are equipped with the cooled PDs for various purposes. At the same time, the greatest demand for QWIPs is observed in IR systems, which require PDs in the long-wavelength range of a large format, as well as multispectral devices [1, 6].
Comparative analysis of cooled PDs of the long-wavelength spectrum
longwave thermal imaging cooled photodetectors are designed and manufactured for the spectrum from 8 to 14 µm based on several technologies. Cooled long-wavelength PDs are produced on the basis of heterostructures of ternary compounds of cadmium-mercury-tellurium (CMT) and on the basis of heterostructures with QWs. The main manufacturers are presented in table. one.
The characteristics of long-wavelength CMT and QW are different while each has its pros and cons. In Russia, over the past few years, several research and development works have been carried out to develop and master the PDs of the long-wave range. The main developers were Orion RPA, RDB Astrohn and ISP SB RAS. Matrix PDs with a resolution of 640 × 480 px for the spectrum range of 8–14 µm are currently being produced at RDB Astrohn and Orion RPA. RDB Astrohn produces uncooled PDs based on Astrohn‑64017 type microbolometer arrays and cooled PDs based on Astrohn‑640KYa20 type QW heterostructures based on QW heterostructures industrially produced at Svetlana-Rost JSC (St. Petersburg). RDB Astrohn also produces cooled PDs based on heterostructures with CMT of the Astrohn‑640KRT15A810 type based on heterostructures with CMT produced at the ISP SB RAS. The PDs produced by Orion RPA for the range of 8–12 µm on CMT have a resolution of 320 × 240 px.
The main advantages of cooled PDs are their speed (up to 200 frames per second) and high sensitivity. It should be noted that the sensitivity of long-wave cooled PDs (about 35 mK) is inferior to short-wave receivers (about 15–20 mK) by about 2 times and is comparable to the sensitivity of microbolometric arrays.
The most famous receivers in Russia are CMT Scorpio LW and QW Sirius LW (France). Let us compare the receivers LYNRED (France), used in the Russian Federation, based on the Scorpio-LW type CMT connection, on Sirius LW quantum wells and RDB Astrohn quantum wells receiver, Table 2
French receivers show a difference in the number of valid pixels between CMT and QWIP receivers in favor of the latter (99.9%). This is primarily due to the physics of the quantum well heterostructures themselves and the mature growth technology. Numerous publications note the great stability and repeatability of technological processes for creating structures based on quantum wells. This is also confirmed by the indicators of domestic PDs based on quantum wells grown at the Svetlana-Rost JSC for the receivers by RDB Astrohn.
The CMT receiver has a slightly larger spectral range (7.7–9.3 µm) with a peak in the 8.5 µm region and 0.7 µm in both directions. For QW receivers with a peak of 8.5 µm, the range is narrower, for Astrohn it is ±0.3 µm in both directions, for Sirius it is ±0.5 µm in both directions.
Also noteworthy is the high uniformity of the pixel sensitivity on the Astrohn‑640KYa20 receiver matrix, which is clearly seen in the histogram in Fig. 8
The CMT receivers have better performance in sensitivity both in the range of 3–5 microns and in the range of 7–14 microns. However, the level of usable pixels on long-wavelength CMT matrices is 98%, which is unacceptable for many tasks. The QW receivers have industrial quality repeatability, good uniformity of pixel sensitivity on the plate and a high (99.9%) level of usable pixels. Comparison of QW receivers by LYNRED and RDB Astrohn shows the identity of the main indicators both in the pixel size (20 μm) and in the average pixel sensitivity on the matrix. (Fig.9)
Situation in Russia
In Russia, the problem of providing development and production in the field of IR photoelectronics with domestic photosensitive materials is quite acute. The development of photodetectors for various purposes in Russia is carried out by a number of enterprises concentrated in Shvabe JSC and Ruselectronics JSC, in the Russian Academy of Sciences, as well as private enterprises. The main suppliers of thermal imaging PDs and MPDs are Orion RPA JSC and MZ Sapfir JSC, which are part of Shvabe JSC, as well as the private enterprise RDB Astrohn JSC. Orion RPA JSC develops and manufactures cooled and uncooled photodetectors, the material science base of the enterprise is focused on molecular beam epitaxy. MZ Sapfir JSC produces cooled and uncooled PDs based on Si, Ge, InSb, CdHgTe. The Institute of Semiconductor Physics of the Siberian Branch of the Russian Academy of Sciences is developing PDs based on CdHgTe, InAs, microbolometers and quantum wells. RDB Astrohn JSC (Lytkarino) develops and manufactures civil thermal imaging devices based on uncooled PDs of its own production, as well as cooled QWIPs together with Svetlana-Rost and the Astrohn-MKS500 microrefrigeration system. [9]
Thus, at present, Russia has developed PDs of the second generation, as well as large-format and IR viewing PDs. The results achieved are close in terms of their indicators to the world level [9, 10].
Trends in the development of QWIPs
Despite the impressive progress in the field of development of QWIPs, today there are a number of problems that limit the range of applications of this technology. The development of a cooled PDs of the 3rd generation aims at obtaining a large-format device with a threshold sensitivity (in one or several spectral ranges) close to the limit at the highest possible operating temperature. It is precisely in terms of threshold sensitivity and operating temperature that QWIP are inferior today to analogues based on CMT. In order to improve the indicated characteristics of QWIPs, it is necessary to carry out fundamental studies in order to clarify the physical mechanisms underlying the operation of QWIPs, as well as to search for constructive and technological solutions that take into account new research results. The most important factor determining the threshold sensitivity of any PD is the quantum efficiency. According to this indicator, QWIPs are still inferior to their narrow-gap counterparts. However, as shown in several publications, such a lag can be minimized by improving the designs of PD sensitive elements [5,6].
The operating temperature, in addition to the quantum efficiency, also depends on the carrier lifetime in the active region and on the design of the QW. As already noted, structures on QWs are similar in nature to impurity photoresistors. This similarity makes it possible to use the same methods to increase the QW lifetime as in impurity photoresistors, in which the probability of carrier capture by an impurity center can be reduced by several orders of magnitude by converting this center from neutral to charged [5]. A similar technique can also be applied in a structure with a QW by doping not a quantum well, but a barrier, as a result of which the well will be negatively charged and free carriers will be repelled from it, i. e. the probability of being trapped in the pit will decrease, and the lifetime will increase. To implement this method of increasing the lifetime, both constructive and technological improvements are required in the creation of QWIPs.
Also, the sharpness of the barrier – well boundaries can significantly affect the lifetime, the dark current, and, as a result, the threshold sensitivity [5, 8]. In viewing PDs, the homogeneity of sensitivity across the matrix and the number of defective elements have acquired particular relevance.
One of the advantages of QWIPs is their potentially high resistance to various kinds of external influencing factors. First, this is resistance to ionizing radiation and laser exposure to high energy. However, there are practically no reports on research in these areas in foreign periodicals. In view of the particular importance of the indicated properties of the PDs for various special applications, it is important to study the resistance of the QWIPs.
The physical and technological problems of creating QWIPs, noted above, characterize this area as a whole, however, they are of particular importance for Russia, where the backlog of domestic applied science in this area remains [11].
This requires both basic researches in the field of physics of low-dimensional semiconductors and IR-photonics [12–14], and perspectives of applications in different fields [15–17].
Conclusion
The last decade can be characterized as a period of rapid development of QW technologies and devices based on it. Such PDs are developed by almost all leading companies. The increase in the information content and the probability of detection and recognition achieved with their use, as well as the compactness of complex optoelectronic systems, are the main driving forces for the development of this direction. In the next decade, two-spectrum PDs, along with megapixels, will become the main commercially available products in IR photoelectronics. To implement two-spectral PDs sensitive in the ranges of 3–5 and 8–12 µm, QW-based technologies are used.
It seems to us that Russia continues to lag behind the leading foreign companies in the development of PDs, and simply financial, even if significant, “injections” into existing organizational structures are not enough to overcome it. The task of achieving parity with the world level, as well as creating scientific, technical, and technological groundwork for the development of PDs, can be solved by a program-target method that provides for a set of technical, financial, and organizational measures within the framework of a public-private partnership mechanism.
ABOUT THE AUTHORS
Kulchitsky Nikolai Alexandrovich, Dr. of Sciences (Eng.), e-mail: n.kulchitsky@gmail.com, Professor, Moscow Technological University (MIREA), Chief Specialist, State Scientific Center of the Russian Federation, Orion Research and Production Association JSC, Moscow, Russia.
ORCID: 0000-0003-4664-4891
Naumov Arkady Valerievich Analyst Engineer, Astrohn Optical-Mechanical Design Bureau JSC, https://astrohn.ru, Lytkarino, Moscow region, Russia.
ORCID: 0000–0001–6081–8304
Startsev Vadim Valerievich, Cand. of Sciences (Eng.), Chief Designer, Astrohn Optical -Mechanical Design Bureau, https://astrohn.ru, Lytkarino, Moscow. region, Russia.
ORCID: 0000-0002-2800-544
Readers feedback
