Abstract
This paper reviews and summarizes the development history and technical characteristics of the infrared detectors for the Fengyun meteorological satellites, and looks forward to the development trend of infrared detectors. Since the successful launch of the Fengyun-1 satellite, China's Fengyun satellite series has become an important part of the global meteorological observation system. As one of the core components of meteorological satellite payloads, the performance improvement of infrared detectors directly affects the accuracy and timeliness of meteorological forecasts. This paper first introduces the basic principles and key technologies of infrared detectors, then sorts out the evolution process of Fengyun meteorological satellites’ infrared detectors from the early stage to the latest version, including technological innovation, performance improvements, and scale expansion. At the same time, this paper also analyzes the current technical challenges and future development directions of infrared detectors, and discusses relevant technical pathways.
0 Introduction
Objects with temperatures above absolute zero radiate electromagnetic waves. Objects at different temperatures generally have different peak wavelengths of radiation. The global average surface temperature is around 15℃, and the peak value of the electromagnetic spectrum radiated by the Earth into space is approximately 10 μm, falling within the long-wave infrared band. Simultaneously, because certain gases in the Earth's atmosphere (such as water vapor, carbon dioxide, and ozone) strongly absorb infrared radiation of specific wavelengths, there exists a so-called atmospheric infraredwindow. By measuring this Earth and atmospheric radiation using infrared detectors, information such as Earth's surface temperature and atmospheric composition can be obtained. Furthermore, infrared remote sensing instruments in space are not limited by sunlight conditions, thus enabling all-weather, day-and-night operation.
Therefore, one of the key capabilities of photoelectric remote sensing instruments, which serve as payloads for meteorological satellites, is infrared imaging and detection. With increasingly severe global climate change and frequent meteorological disasters, higher demands are being placed on the accuracy and timeliness of weather forecasts. Meteorological satellites are a crucial means of atmospheric monitoring, and their importance is becoming increasingly prominent. Infrared remote sensing instruments, as the core payload of meteorological satellites, acquire infrared radiation information from the Earth's atmosphere and surface, providing vital data support for fields such as weather forecasting, disaster monitoring, agricultural production, and environmental protection. Infrared detectors are the key components in these instruments that enable photoelectric conversion.
The launch of the US polar-orbiting meteorological satellite on April1, 1960, marked the beginning of the era of satellite remote sensing for meteorological observation. The Fengyun series of satellites are meteorological satellites independently developed by China. In the1960s, China began preparations for developing polar-orbiting meteorological satellites. In 1970, Premier Zhou Enlai proposed "developing our own meteorological satellites" and personally assigned related tasks, thus beginning the research and development of China's first generation of polar-orbiting meteorological satellites − FY-1. The first batch of FY-1A satellites was launched on September 7, 1988, and FY-1B on September 3, 1990. The second batch of FY-1C satellites was launched on May 10, 1999, and FY-1D on May 15, 2002. Currently, all Fengyun-1 satellites have ceased operation. The Fengyun-2 (FY-2) satellite is China's first generation of geostationary meteorological satellites. Eight FY-2 satellites have been successfully launched, with six − FY-2A, FY-2B, FY-2C, FY-2D, FY-2E, and FY-2F − having ceased operations. Currently, FY-2G and FY-2H remain in orbit and continueto provide operational services. FY-2, along with the polar-orbiting meteorological satellites (FY-1/3) , complement each other, forming China's meteorological satellite application system, playing a crucial role in monitoring major severe weather events and significant meteorological natural disasters. As China's second-generation polar-orbiting meteorological satellite, the Fengyun-3 (FY-3) satellite is a development and improvement upon the technology of the FY-1 meteorological satellite, capable of acquiring global, all-weather, three-dimensional, quantitative, and multispectral atmospheric, surface, and sea surface characteristic parameters. The Fengyun-4 (FY-4) satellite is China's new-generation geostationary meteorological satellite, achieving for the first time in the world integrated observation of geostationary imaging observation and infrared hyperspectral atmospheric vertical sounding.
Currently, China has launched 21 meteorological satellites, of which 9 are in orbit and have achieved operational status for both polar-orbiting and geostationary satellites (see Figure1) . China is the third country after the United States and Russia to possess both polar-orbiting and geostationary meteorological satellites. The Fengyun series of meteorological satellites has become a representative of China's strength and has a wide international reputation for Earth observation. It has been included in the global operational Earth observation satellite sequence by the United Nations World Meteorological Organization, which has improved the forecast timeliness and accuracy of China's and the world's most advanced medium-and long-term numerical weather prediction models. The Fengyun series of meteorological satellites also undertakes the Chinese duty satellite mission of the international charter mechanism for disaster reduction, and plays an increasingly important role in international meteorological disaster prevention and mitigation work [1]. All of these achievements are inseparable from the infrared remote sensing instruments of meteorological satellites.
From the multi-channel scanning radiometer of Fengyun-1 to the scanning radiometer of Fengyun-2, and the ten-channel scanning radiometer, medium-resolution spectral imager, infrared spectrometer, and Earth radiation budget instrument of Fengyun-3, and then to the advanced multi-channel scanning radiometer and atmospheric vertical sounder of Fengyun-4, all payloads were developed by the Shanghai Institute of Technical Physics, Chinese Academy of Sciences. All the infrared detectors used in these instruments were independently developed by the institute. Due to differences in orbital altitude and satellite platform stabilization methods, Earth observation remote sensing instruments differ significantly, thus different payloads place different requirements on infrared detectors. This paper reviews the basic situation of infrared detectors based on the characteristics of Fengyun satellite instruments, then analyzes and discusses some problems in detector development, and looks forward to the future technological direction of the detectors. This paper aims to sort out and summarize the development history, technical characteristics, performance improvement, and future development trends of the Fengyun meteorological satellite infrared detectors, contributing to the further development of meteorological satellite technology in China.
1 Basic Principles and Technological Basis of Infrared Detectors
1.1 Working principle of infrared detectors
An infrared detector is a device that converts infrared radiation into electrical signals. Its working principle is based on various effects produced by the interaction between infrared radiation and matter, such as the infrared heating effect and the infrared photoelectric effect. In meteorological satellites, the commonly used infrared detectors are mainly of two types: thermoelectric detectors and photoelectric detectors.
Thermoelectric detectors, sometimes called thermal detectors, such as thermistor infrared detectors, utilize the heating effect of infrared light on the detector material. Under infrared light irradiation, the thermistor's temperature rises, causing a change in the device's resistance. Since the infrared light energy flow to be detected is generally very weak, the heat capacity of the light absorber in the thermal detector must be minimized while simultaneously reducing heat conduction losses. This results in a relatively long dynamic response time for thermal detectors, typically ranging from several milliseconds to tens of milliseconds. In the early development stages, when single-unit devices were the primary focus, thermistor infrared detectors, operating at room temperature and with relatively simple principles and manufacturing processes, were initially well-suited for use in aerospace instruments such as horizon sensors. However, as the scanning speed requirements of infrared imaging units have increased, thermal detectors, due to their long response times, have become increasingly difficult to meet system requirements.
Infrared photodetectors are detectors built using the photoelectric effect within narrow-bandgap semiconductors. Common types include photoconductive and photovoltaic detectors. Infrared photoconductive detectors are sensitive elements whose resistance decreases due to infrared light illumination. High-performance photoconductive detectors can be fabricated using intrinsic semiconductors. Since the energy of theincident photon must be greater than the semiconductor's bandgap to excite intrinsic photoelectric transitions of valence band electrons into the conduction band, causing an increase in conductivity, semiconductor photodetectors all have a cutoff wavelength in their response spectrum. All photodetectors face the challenge of matching this cutoff wavelength with the operating wavelength.
From the perspective of detector response spectra, thermal detectors with blackened surfaces are generally considered to have poor spectral selectivity, with a spectrum approximately a flat line. Photodetectors, on the other hand, involve the conversion of photons into electrons, and their ideal response spectrum is a triangle with a clear long-wavelength limit (i.e., long-wavelength cutoff wavelength) . The longer the cutoff wavelength, the narrower the corresponding bandgap, and the more severe the thermal excitation in the semiconductor. Therefore, infrared photodetectors generally need to be cooled to low temperatures to achieve high performance. Due to direct detection after photoelectric conversion and the excellent conditions of environmental cooling, the sensitivity of photodetectors is generally two orders of magnitude higher than that of thermal detectors. The sensitivity of infrared detectors is generally characterized by specific detectivity (usually denoted by D*) . It refers to the signal-to-noise ratio obtained by a unit area of detector per unit incident energy per unit electronic bandwidth. The D* of thermal detectors is generally 107–108 cm-1·Hz1/2·W-1, while the D* of long-wavelength photodetectors is generally 1010–1011 cm-1·Hz1/2·W-1.
Photovoltaic detectors utilize the photovoltaic effect of a semiconductor pn junction. Photo-excited non-equilibrium carriers are collected on both sides of the pn junction under the influence of the built-in electric field and output as a photocurrent proportional to the incident light from two short-circuited electrodes. Obviously, fabricating a pn junction requires a doping process. Relatively lower operating temperatures are also needed to reduce dark current and maintain junction impedance. However, photovoltaic detectors can operate with zero bias, making array integration easier. In particular, they can be easily interconnected with silicon-based complementary metal-oxide-semiconductor (CMOS) readout circuits via flip-flopbonding to form a two-dimensional array with a back-incidence structure, i.e., an infrared focal plane array detector.
1.2 Narrow bandgap semiconductor materials are the foundation for the development of infrared detectors
Detector materials are a core component of infrared detectors, and their performance directly affects the detector's sensitivity and resolution. Currently, commonly used infrared photodetector materials include mercury cadmium telluride, indium antimonide, and indium gallium arsenide, while commonly used infrared thermal detector materials include vanadium oxide, amorphous silicon, and manganese cobalt nickel thermistors. These materials possess different spectral response characteristics and operating temperature ranges, allowing for selection based on specific application requirements. In recent years, with the continuous development of nanotechnology and thin-film technology, significant progress has been made in the research and development of novel artificial structure infrared detector materials.
Mercury cadmium telluride (Hg1-xCdxTe) remains the best material for fabricating high-sensitivity infrared detectors. Due to its direct bandgap with selectable bandgap, as well as its advantages such as low effective electron mass, high light absorption coefficient, long minority carrier lifetime, high electron mobility, low thermal carrier generation rate, and relatively low dielectric constant, it has dominated the field of infrared materials and devices since its invention. To date, despite the continuous development and improvement of various materials and devices with the advancement of semiconductor bandgap engineering theory and practice, mercury cadmium telluride infrared detectors have maintained their leading position.
As early as 1968, the Shanghai Institute of Technical Physics of the Chinese Academy of Sciences laid out the research on mercury cadmium telluride materials. Researchers represented by Shen Jie and Tang Dingyuan started the research and development of mercury cadmium telluride bulk materials [2]. Mercury cadmium telluride is an alloy ofsemiconductor material cadmium telluride and semi-metallic material mercury telluride. By changing the composition x, the band gap of the material can be adjusted, so infrared detectors with cutoff wavelength matching detection working bands can be made. For example, a room temperature mercury cadmium telluride detector with a short wave of 1.64 μm can be made with a material with x of 0.62, a detector with a medium wave of 3.93 μm (working at 105 K) can be made with mercury cadmium telluride material with x of 0.34, a detector with a long wave of 12.5 μm can be made with mercury cadmium telluride material with x of 0.20, and a very long wave detector requires a smaller composition. Chu Junhao, Tang Dingyuan and others have given an empirical formula for the relationship between the band gap of mercury cadmium telluride and composition and temperature [3].
As the operating wavelength of HgCdTe materials extends further into the long wavelength range, the component values become smaller and smaller; however, the change in cutoff wavelength caused by a small change in component values becomes increasingly larger. Therefore, whether from the perspective of material preparation or material selection, the requirement for the uniformity of HgCdTe composition increases with the increase in operating wavelength. It is easy to see that the development of very long-wave infrared detectors is far more difficult than that of mid-wave infrared detectors.
However, mercury cadmium telluride materials also have drawbacks. Due to the high vapor pressure and easy segregation of mercury, it is difficult to prepare large-area bulk single crystals without macroscopic defects and with uniform optical and electrical properties. The Hg-Te bonds in the crystal are weak and will dissociate at relatively low temperatures, forming mercury vacancies and interstitial atoms, which are easily mobile. Once they move to the surface, they may volatilize out of the body, impairing the stability of the crystal. The operating temperature in the device manufacturing process must generally be kept below 80°C to ensure chip characteristics.
Furthermore, due to limitations in the bulk material crystal fabrication characteristics, it is virtually impossible to fabricate large-scale focal plane detectors using mercury cadmium telluride bulk materials. Therefore, it is necessary to develop HgCdTe epitaxial growth technology to fabricate large-scale area array detectors. Commonly used epitaxial thin film growth techniques for HgCdTe include liquid phase epitaxy, molecular beam epitaxy, and vapor phase epitaxy. Among them, liquid phase epitaxy grows near the liquid-solid point, which is a near-equilibrium growth process, and therefore it is relatively easy to obtain materials with better lattice quality. However, techniques such as molecular beam epitaxy can precisely monitor the material growth process, so it is easier to obtain epitaxial materials with specific functional structures. From the perspective of developing hybrid interconnect large-area arrays, obtaining HgCdTe thin films through heteroepitaxial growth on silicon-based materials has a thermal expansion coefficient that matches that of silicon CMOS circuits, making it the preferred route for ultra-large-scale HgCdTe focal plane detectors. However, this requires overcoming many problems in the epitaxial growth of large-area HgCdTe materials [4-5].
The ultimate foundation of epitaxial technology lies in the lattice matching degree of the substrate and the defect density of the epitaxial substrate surface. Currently, lattice-matched materials with mercury cadmium telluride crystals are mostly cadmium zinc telluride crystals. Therefore, the level of bulk crystal growth technology for lattice-matched substrate materials is playing an increasingly important role.
The same mechanism is also demonstrated in the indium gallium arsenide epitaxial layer and the indium phosphide substrate. We can observe that the performance of the lattice-matched 1.7 μm indium gallium arsenide detector significantly outperforms wavelength-extended devices [6].
1.3 Multi-band integration technology is the only way to develop infrared detectors for aerospace infrared remote sensing instruments
Since infrared atmospheric windows exist in the wavelength ranges of 1–3 μm, 3–5 μm, and 8–14 μm, meteorological satellite infrared remote sensing instruments require theselection of appropriate materials corresponding to these bands. Different wavelength bands correspond to different photon energies. Materials capable of detecting different photon energies ideally possess corresponding band gaps, achieving photoelectric conversion quantum efficiency exceeding80%. Since the band gap of mercury cadmium telluride is adjustable, we utilized HgCdTe covering the entire1–12 μm infrared band on the Fengyun-1 satellite. Different HgCdTe compositions require different growth process conditions; therefore, fabricating HgCdTe materials with suitable compositions is the primary task in the development process. Secondly, selecting the appropriate composition of the material wafer is another crucial step in ensuring suitable wavelengths. Furthermore, different material compositions exhibit varying characteristics in detector chip fabrication processes, requiring careful differentiation.
After the detector chips of different bands are manufactured and screened, they are also integrated with micro-filters. Cold filters can limit background radiation. In the near-photosensitive surface packaging of micro-filters, attention needs to be paid to the low-temperature characteristics, low-temperature reliability and stray light crosstalk of the filters. The low-stress integrated packaging technology of micro-filters is also one of the considerations to ensure that the component response spectrum meets the inner and outer frame limit requirements of the system working band. This issue was recognized and strengthened during the development of the Fengyun-2 03 batch of components. The different temperature stress caused by different filter metal support materials has significantly affected the shape of the component response spectrum. In addition, Dewar packaging is also one of the contents of infrared detector technology. It consists of infrared detector chips, Dewar, command and electrical control components, etc. The main technologies of Dewar component packaging include Dewar structure design, thermodynamic optimization design, Dewar lead technology, high-precision installation and optical registration technology of infrared chips, welding technology, leak detection technology, surface treatment technology, component reliability technology, etc. [7]
Obviously, integrated packaging not only provides the optimal solution for detector performance, but also plays a significant role in reducing the size and weight of remote sensing instrument systems. Therefore, multi-band integrated detector component technology is essentially an inevitable choice for aerospace instruments to pursue ultimate performance across all bands [8].
2 Development History of Fengyun Meteorological Satellite Infrared Detectors
2.1 Fengyun-1 Satellite Infrared Detector
Fengyun-1 satellite is China's first generation of polar-orbiting meteorological satellites, employing an orbital altitude of 860 km and a three-axis stabilized satellite platform. The onboard scanning radiometer utilizes the satellite's motion and the rotation of the scanning mirror to complete two-dimensional scanning imaging. The first satellite, FY-1A, was successfully launched in 1988. Its infrared detector adopted then-advanced mercury cadmium telluride optical guide detectors and radiation cooler technology. The scanning rate of the remote sensing instruments was improved from 36 rpm in the pre-research stage to 120 rpm and then to 360 rpm, achieving a continuous improvement in the ground resolution of Earth observation from 8 km to 4 km and then to 1.1 km, reaching the international advanced level at the time. FY-1A/B satellites have five channels with wavelength ranges of 0.48–0.53 μm, 0.53–0.58 μm, 0.58–0.68 μm, 0.725–1.1 μm, 10.5–12.5 μm; the FY-1C/D satellite has 10 channels with wavelengths ranging from 0.43 to 0.43–0.46 μm, 0.48–0.53 μm, 0.53–0.58 μm, 0.58–0.68 μm, 0.84–0.89 μm, 0.900–0.965 μm, 1.58–1.64 μm, 3.55–3.93 μm, 10.3–11.3 μm, 11.5–12.5 μm. The infrared detector of Fengyun-1A satellite is a single-unit mercury cadmium telluride optical guide long-wave infrared detector. Fengyun-1C satellite of batch 02 had been extended to a three-channel medium-and long-wave detector assembly. This assembly operates on a second-stage radiation cooler at 105 K. The short-wave band uses a room-temperature mercury cadmium telluride photovoltaic detector (see Figure2) .
During the development of the infrared detector of the Fengyun-1 meteorological satellite, Fang Jiaxiong et al. proposed the concept of band detection rate based on the actual needs of the development work, so as to more appropriately characterize the performance of the infrared detector working in a specific band; in addition, they pioneered the electrical aging and life test to determine the average life of the device; at the same time, they confirmed the test standard of the meteorological satellite infrared detector aged for 1500 h, which is still in use today [9].
Fig.2Photographs of the FY-1C three-channel HgCdTe infrared detector and short-wave infrared detector.
2.2 Fengyun-2 Satellite Infrared Detector
Fengyun-2 is China's first-generation geostationary meteorological satellite, independently developed by China. Its orbital altitude is 36, 000 km, and the satellite platform is spin-stabilized. The onboard scanning radiometer utilizes the satellite's spin and the rotation of the scanning mirror to complete two-dimensional scanning imaging of the Earth's disk, acquiring daytime visible light cloud images, daytime and nighttime infrared cloud images, and water vapor distribution maps. Simultaneously, it collects meteorological monitoring data from meteorological, hydrological, and oceanographic data collection platforms, monitors solar activity and the space environment of the satellite's orbit, and provides monitoring data for satellite engineering and spaceenvironment science research. It can achieve directional coverage and continuous remote sensing monitoring of the Earth's surface and atmospheric distribution, with advantages such as strong real-time performance, high temporal resolution, objectivity, and vividness. China's geostationary meteorological satellite has five predetermined fixed-point locations: 79°E, 86.5°E, 105°E, 112.5°E, and 123°E. Among them, 105°E is the fixed-point location for operational satellites, and 79°E is the fixed-point location for the "Belt and Road Initiative" satellites.
China's first generation of geostationary meteorological satellites were divided into three batches: (1) Batch 01 consisted of experimental geostationary meteorological satellites, including two satellites: FY-2A was successfully launched from the Xichang Satellite Launch Center on June10, 1997, aboard a Long March 3 rocket, and FY-2B was successfully launched from the Xichang Satellite Launch Center on June25, 2000, aboard a Long March 3 rocket. (2) Batch 02 consisted of operational geostationary meteorological satellites. Compared with the Batch 01 satellites, the technical performance of the Batch 02 satellites was greatly improved (mainly the onboard scanning radiometer increased from 3 channels in the Batch 01 to 5 channels) . Batch 02 consisted of three satellites: FY-2C was launched on October 19, 2004, FY-2D was launched on December 8, 2006, and FY-2E was launched on December 23, 2008. (3) The main purpose of the 03 batch of satellites was to ensure a continuous and stable transition from the first generation of geostationary meteorological satellites to the second generation, with appropriate improvements to the performance of the satellites compared to the 02 batch. The 03 batch included three satellites: FY-2F, launched on January 13, 2012; FY-2G, launched on December 31, 2014; and FY-2H, launched on June5, 2018. Currently, FY-2G and FY-2H satellites are in orbit and providingapplication services.
The infrared detector of the Fengyun-2 01 batch is a dual-band component using both water vapor and thermal infrared wavelengths. Xu G S et al. reported on the performance and related aspects of the quadruple mercury cadmium telluride detector with dual water vapor/thermal infrared bands (6.3–7.6 μm and 10.5–12.5 μm) used inthe first Fengyun-2 satellite [10]. At an operating temperature of 100 K, the D* value for the thermal infrared band detector is 3.4×1010 cm·Hz1/2·W-1; the D* value for the water vapor band detector is 1.1×1011 cm·Hz1/2·W-1.
The infrared detectors on Fengyun-2 satellites of batches 02 and 03 have been expanded to four-band components consisting of mid-wave, water vapor, thermal infrared 1, and thermal infrared 2 (these two bands are also called thermal infrared split windows) (see Figure3) . Table1 lists the performance data of the Fengyun-2 satellite infrared detector (component number 2010/11) at 93 K temperature. In the development of the Fengyun-2 detector, the mechanism of equal thickness interference between the reflected light from the surface of the mercury cadmium telluride photosensitive element and the reflected light from the sapphire substrate was also discovered. Li X Y, Xu J T, Zhu L Y and others designed and verified the high-performance detector spectral modulation technology route with anti-reflection film and wall structure, which improved the problem of the response spectral shape of the optical guide detector to a certain extent (see Figure4) [11].
Fig.3Photo of the die of the Fengyun-2 four-band infrared detector before packaging.
Table1Performance data of the Fengyun-2 infrared detector labeled as 2010/11 at 93 K
Fig.4Response spectra of the Fengyun-2 four-channel infrared detector assembly (the inner and outer boxes resembling regular rectangles are the quantitative assessment requirement limits proposed by the satellite application department) .
2.3 Fengyun-3 satellite infrared detector
FY-3A is China's second-generation polar-orbiting meteorological satellite. Launched on May 27, 2008 (Note: China uses odd-numbered naming conventions for polar-orbiting satellites and even-numbered naming conventions for geostationary satellites; during ground development, satellites are numbered sequentially, i.e., 01, 02, etc.; after successful launch, they are renamed A, B, etc.) . The satellite employs a three-axis stabilization system and carries multiple instruments. Among these, the infrared-channel payloads include a ten-channel scanning radiometer (VIRR) , an infrared spectrometer (IRAS) , a medium-resolution spectrophotometer (MERSI) , and an Earth radiation budget instrument (ERM) . The ERM uses a room-temperature operating, spectrally flat, broad-spectrum-coverage thermistor infrared detector to detect total infrared radiation from Earth. The other three instruments all use high-sensitivity cooled mercury cadmium telluride (MCT) detectors. The scanning radiometer and the imager both utilize a band-matched MCT detector chip integrated with a low-temperature micro-filter in a single package. The infrared spectrometer employs a large-area MCT detector with a light-cone focusing design, operating by switching channels via a filter wheel. Because the infrared spectrometer requires detection in the very longwave band above15 μm, a key focus was providing high-performance large-area detectors based on small-scale MCT materials in the development of the Fengyun-3 01 satellite. Furthermore, the long-wave channel of the medium-resolution spectral imager employed a40-element detector, which was the largest-scale space-grade infrared MCT detector chip at the time. The satellite instruments not only boasted diverse functions but also achieved a significant leap forward in ground resolution, detection sensitivity, and spectral range.
Starting with Fengyun-3 04 satellite, the infrared spectrometer was replaced by a Fourier-Transform Infrared (FTIR) hyperspectral instrument. Unlike filter-based methods, this instrument employs a Michelson interferometer configuration. It generates interference patterns by moving a mirror, then applies a Fourier transform to derive spectral information. The first successful space application of this FTIR remote sensing instrument was achieved on Fengyun-4A satellite. Due to the requirement for a wide response spectrum and high detection performance, the development of FTIR instruments requires extra attention to the width and shape of the spectrum compared to detectors operating on specific wavelengths. Therefore, the detector development specifications typically specify three wavelength performance evaluation points: the shortest required wavelength, the longest required wavelength, and the detection rate at the peak response position. Furthermore, because the moving mirror's speed is within acertain range, the detector also needs to pay special attention to its detection rate performance at different frequencies, particularly in the range of tens to hundreds of kilohertz.
The Fengyun-3 satellite carries numerous infrared instruments with complex channel arrangements and diverse types. Taking the medium-wave component (M210304) of the medium-resolution spectral imager on satellite 06 as an example, the detector's performance data is shown in Table2.
2.4 Fengyun-4 Satellite Infrared Detector
FY-4A is China's second-generation geostationary meteorological satellite. Launched on December 11, 2016, the satellite employs a three-axis stabilization system, with a scanning radiometer and a vertical sounding instrument as its two main payloads. On February 27, 2017, FY-4A, China's new-generation geostationary meteorological satellite, obtained its first batch of images and data. The world's first geostationary orbit hyperspectral image of the Earth's atmosphere was officially released. The scanning radiometer has 10 channels: the short-wave channel uses an indium gallium arsenide
The detector performance of the radiation-cooled prototype components (serial numbers LW2015/02, WV2014/07, SM2014/18) for the Fengyun-4 01 satellite radiometer is shown in Table3. The detectivity range refers to the range from the minimum to the maximum value of the four or eight photosensitive detectors. Figure4 shows the response spectrum of the long-wavelength component LW2015/02 of the Fengyun-4 01 satellite.
Table3Performance data of the infrared detector of Fengyun-4 (01) satellite
Fig.4Response spectra of Fengyun-4 (01) long-wave assembly labelled as LW2015/02.
In the development of the very long-wavelength optical guide array detector of the Fengyun-4 vertical sounder, a multi-metal three-dimensional relay interconnection technology for the optical guide array was achieved by laser-etching a sapphire micro-via array. This alleviated the problem of array lead congestion, improved the detector's duty cycle, and effectively enhanced the overall photoelectric conversion efficiency. During the development process, elemental purification techniques were advanced to synthesize mercury cadmium telluride crystal materials featuring high mobility, long minority carrier lifetime, and a bandgap width below 75 meV. Low-noise chip fabrication processes were enhanced to produce high-quantum-efficiency detector chips with a cutoff wavelength of 16.5 μm. The fabricated mercury cadmium telluride very long-wave infrared optical guide array detector (see Figure5) achieved an average peak detectivity of 7.26×1010 cm·Hz1/2·W-1 and a response cutoff wavelength greater than 16.5 μm.
Fig.5Photo of the FY-4B VLWIR photoconductive detector array (16×8) .
3 Technical characteristics of the Fengyun meteorological satellite infrared detector
3.1 Improved detection sensitivity and resolution
Sensitivity is one of the most critical performance parameters of remote sensing instruments, and the detectivity of the infrared detector can reflect this to a certainextent. Most of the infrared remote sensing instruments on the Fengyun satellites use high-detectivity MCT detectors. The detection rate of currently developed infrared detectors has approached the background limit of 300 K. The total noise of the detector is determined when the fluctuation in the number of photons from the infrared background radiation incident on the detector becomes the main source of noise, and at this point, the detector performance reaches the background limit. In the low-temperature filter assembly of multispectral remote sensing instruments, since the filter is integrated with the detector, the filter plays a role in limiting the background photon flux. In our experiments, we can clearly observe that the D* value of the encapsulated detector is significantly higher than the screening test value during chip fabrication.
When the capabilities of a single detector approach their limits, we can leverage the scale effect of the detector to achieve better imaging and detection. Figure6 shows the trend of the maximum size of the infrared detectors used by the Fengyun satellites over time. It is clear from the development of the Fengyun satellite detectors that as requirements for detection accuracy, ground resolution, and detection channels increase, the size of the detectors also continuously increases. Statistical data from over thirty years shows that the total number of photosensitive elements in the entire satellite roughly doubles every three years.
Fig.6The trends of total elements of infrared detectors in Fengyun satellites.
3.2 Evolution from narrow-band detection to broad-spectral coverage
From the perspective of the spectrum of remote sensing instruments, scanning radiometers have narrowband characteristics: one spectral channel–one narrowband filter–one detector photoelement (later developed to have multiple detector photoelement per spectral channel) . This type of detector can approximately focus only on the performance at the peak wavelength position. Hyperspectral spectrometers and FTIR hyperspectral instruments, on the other hand, utilize changes in the instrument's operating wavelength. In this case, the detector not only needs to ensure peak detectivity but also needs to monitor the performance across the entire wavelength range.
The fabrication of very long-wavelength infrared FTIR devices using photoconductive MCT detectors offers a feasible solution to the limitations of wide wavelength range, high performance, and operating temperature. The basic principle is roughly as follows: First, very long wavelengths represent significant thermal excitation in semiconductors. Photoconductive devices based on the intrinsic narrow bandgap semiconductor photoelectric effect have a clear advantage in operating temperature, while photovoltaic devices using doped semiconductors to form pnjunctions generally require even lower operating temperatures. Second, high-performance detectors require semiconductor materials with high absorption coefficients. However, the use of such high-absorption direct bandgap semiconductor materials means that the absorption of short-wavelength photons occurs near the material surface. But the surface and interfaces of the material often contain deep energy levels or surface dangling bonds, forming non-equilibrium carrier recombination centers. In other words, the magnitude of the surface recombination velocity profoundly affects the shape of the short-wavelength portion of the detector's response spectrum; that is, the surface and interfaces can significantly influence the short-wavelength spectrum of the detector. The working principle of photovoltaic devices utilizes the built-in electric field of pn junctions constructed with different dopants. Since the electric field is located within the material and is always some distance from the surface, the short-wavelength spectrum of photovoltaic detectors cannot compare with that of photoconductive devices in the same wavelength range.
Photovoltaic devices typically require surfaces with near-flat energy bands to enhance R0A values and ensure detection efficiency. In contrast, photoconductive devices can incorporate heavily-accumulating light-incident surfaces to suppress surface recombination rates and broaden the response spectrum. Consequently, mercury cadmium telluride photoconductive detectors remain the primary choice for very-long-wave FTIR instruments in current laboratory equipment. However, there are serious difficulties in fabricating large-scale arrays using photoconductive detectors. Photovoltaic array devices show advantages in scalability.
3.3 Continuous innovation and the pursuit of excellence are the driving forces behind the development of the Fengyun meteorological satellite infrared detector
Looking back, the infrared detectors of the Fengyun meteorological satellites have evolved from nothing to something, with their performance continuously improving. The performance of the photoelectric detector units is increasingly approaching the background limit, the array size is constantly increasing, and the spectral shape isbecoming more and more perfect. From usable to easy-to-use and then to aesthetically pleasing, they have been refined to continuously meet application requirements.
Meanwhile, in order to improve the overall system performance, the Shanghai Institute of Technical Physics of the Chinese Academy of Sciences has leveraged its professional vertical integration characteristics and, from the perspective of meeting national application needs, has continuously iterated and optimized between technology-driven and demand-driven approaches, developing some technologies unique to the Fengyun series detectors. For example, mercury cadmium telluride is a relatively brittle material, so leading out gold wire electrodes on the detector is a technology that tests design and process. We initially had indium ball bonding technology, later developed ultrasonic bonding, and currently mostly use gold wire ball bonding, sapphire via interconnect technology, and indium pillar flip-bonding interconnect technology. Other technologies include: (1) low-temperature bonding technology, high aspect ratio etching technology, and photosensitive area confinement protection ring technology; (2) the unidirectional stacking technology of the Fengyun-2 medium-wave detector to suppress the sweep-out effect; (3) the Fengyun-3 40-element long-wave detector adopts the tilted photosensitive surface design technology matched with scanning imaging, and completes the registration of pixel imaging timing in conjunction with the overall timing sequence; (4) the anti-reflection structure with lead electrodes placed below the FTIR instrument, designed according to the characteristics of the instrument, and the low-temperature preamplifier integration technology in conjunction with the radiated photovoltaic detector. These technologies are the result of close collaboration between system engineers and detector engineers, contributing to the excellent performance of the entire system.
4 Technological challenges and future development trends of infrared detectors for fengyun meteorological satellites
4.1 Technical challenge analysis
The next generation of satellites will have higher ground resolution, higher temporal resolution, finer spectral resolution, higher temperature resolution, and higher accuracy in radiometric quantification. These requirements place demands on the detectors: detector size needs to be further increased, the photoelectric performance and response spectrum of the detectors need further optimization, and the stability and reliability of the detectors need further improvement. Therefore, the wafer size of the substrate material needs to be increased, the wafer size of the epitaxial material needs to be increased and its performance needs to be improved, the control level of the chip fabrication process needs to be improved, and the size and precision of the packaging need to be increased. Furthermore, with the continuous increase in detector size, the massive data transmission and processing on the detector chip will also face challenges.
4.2 Future development trend forecast
As the requirements for future remote sensing instruments continue to increase, the development of detectors will move towards increased scale, optimized performance, and greater intelligence.
In scenarios with high photon flux density and long integration times, uncooled focal plane array detectors, leveraging their advantages of large scale, long integration times, and low cost, may find some applications in imaging. Cooled photon-detector-based focal plane array devices will remain dominant in hyperspectral, high-resolution, and rapid imaging. Multi-band, multi-color integrated focal plane array detectors may see some new applications. Detector chips will incorporate dedicated functions such as analog-to-digital conversion and information processing to preprocess the massive amounts of acquired data for subsequent tasks. Future photon detector chips will see significant improvements in dynamic range, and the equivalent temperature difference (NETD) of focal plane array detectors will reach the 1 mK level.
In terms of detector materials, substrate materials are evolving towards 6-8 inches, andepitaxial technology is advancing to 6−8 inches in wafer size. Multilayer epitaxial technology with bandgap engineering features is developing detector chips with more functions. In detector manufacturing, wafer processing capabilities are expanding from the current 4–6 inches to 6–8 inches, and the size of key lithography lines is reaching the sub-micron level. Regarding process control, automation technology is widely adopted to improve manufacturing yield, and artificial intelligence is being combined to enhance process control. It is expected that within 5 to 10 years, detector sizes will reach 12 k×12k, enabling meteorological satellites in geostationary orbit to obtain real-time staring observation videos (10 frames per second) of the Earth's disk with 1 km resolution and 1 mK sensitivity.
4.3 Discussion of Technical Paths
The history of my country's independent development of Fengyun meteorological satellites tells us that building a path of independent development across the entire process, from materials to devices to instruments and finally to the entire satellite, is both correct and necessary. The development of infrared detectors must first emphasize the fundamental role of materials; secondly, while targeting application needs, the joint development of materials and devices is an effective way to achieve rapid technological breakthroughs. Developing process technologies and researching and developing specialized process equipment are among the current directions for the self-reliant development of detectors.
5 Conclusions and Outlook
This paper reviews the development history and related technical characteristics of the Fengyun meteorological satellite and its detectors, and looks forward to the development trend of infrared detectors. As one of the core components of meteorological satellite payloads, the performance of infrared detectors directly affects the performance of meteorological remote sensing instruments, thus influencing theaccuracy and timeliness of weather forecasts. The development of Fengyun satellite infrared detectors is closely related to the development of Fengyun satellite remote sensing instruments. The entire process reflects the spirit of Chinese researchers' self-reliance, continuous self-improvement, and constant innovation and transcendence in order to successfully complete national missions. This paper also analyzes the current technical challenges and future development directions of infrared detectors, and proposes possible solutions and technical paths.