Single photon avalanche detector based on cerium-doped InP for space communication and preparation method thereof

Abstract

This invention discloses a cerium-doped InP-based single-photon avalanche detector for space communication and its fabrication method. The detector's epitaxial structure, from bottom to top, includes a substrate, an N-type buffer layer, an absorption layer, a band transition layer, an N-type charge layer, an intrinsic multiplication layer, an intrinsic InP layer, and a cerium-doped P-type contact layer. The core innovation lies in the dual radiation-resistant design of the Ce-doped P-type contact layer and the Ce-doped modified stacked passivation film. The fabrication method involves MOCVD epitaxial growth, spin-coating source diffusion to prepare the P-type contact layer, ICP etching and wet repair, magnetron sputtering doping, and electrode fabrication. This invention utilizes Ce's unique electronic structure to suppress radiation-induced defects and combines this with an SACM structure to optimize the electric field distribution, achieving low dark count rate, high detection sensitivity, and excellent neutron radiation resistance. The process is compatible with existing semiconductor production lines and is suitable for wafer-level mass production, meeting the high-reliability single-photon detection requirements of aerospace missions such as space laser communication and quantum key distribution.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of semiconductor optoelectronic devices and space radiation effects, specifically involving a cerium-doped InP-based single-photon avalanche detector for space communication and its fabrication method. This single-photon avalanche detector is a near-infrared single-photon avalanche diode (SPAD) for applications such as space laser communication and quantum key distribution. Background Technology

[0002] SPAD (Single-Photon Avalanche Diode) is a semiconductor detector operating in Geiger mode, achieving high-sensitivity detection at the single-photon level through the avalanche multiplication effect. Its core materials include semiconductors such as InGaAs. SPADs have irreplaceable application value in critical space missions such as deep-space laser communication and inter-satellite quantum communication, and are key components for ensuring the reliability of space communication and detection systems, promoting the upgrading of space technology, and developing related industries. Although SPADs based on III-V compounds (such as InP / InGaAs) have fundamental advantages such as high detection sensitivity and adaptability to the extreme space environment, the harshness of the space radiation environment and the limitations of traditional doping processes mean that traditional SPAD technologies still have many shortcomings, severely restricting their long-term stable application and mass production in the space field. These shortcomings are mainly reflected in the following aspects: 1. In the space radiation environment, a large number of high-energy particles (protons, electrons, neutrons, etc.) generated by galactic cosmic rays and solar activity can cause severe displacement and ionization damage to SPAD semiconductor devices. For SPADs based on III-V compounds (such as InP / InGaAs), displacement damage introduces deep-level defects into the material's bandgap. These defects act as carrier generation-recombination centers (GR centers), which exponentially increase the device's dark count rate (DCR). As a core noise indicator for SPADs, an increase in the dark count rate can severely affect detection accuracy and device stability. It can also alter the device's avalanche breakdown characteristics, significantly shortening the space application lifetime of SPADs and becoming a major bottleneck limiting their space deployment.

[0003] 2. Traditional SPADs typically use group II elements such as zinc (Zn) or cadmium (Cd) to dope the P-type region. While these group II elements can form relatively shallow acceptor levels in InP, meeting the basic electrical performance requirements of the device, they have significant process and performance defects: On the one hand, the high atomic mobility of Zn and Cd leads to poor process thermal stability, affecting the consistency and reliability of device fabrication; more importantly, when exposed to neutron irradiation, Zn atoms readily form complex, electrically highly reactive defect complexes (such as Zn-V-P) with irradiation-induced point defects (such as phosphorus vacancies V_P and indium interstitials In_i). These complexes are usually deep energy levels, becoming highly efficient GR centers, significantly exacerbating dark count noise, further deteriorating the device's radiation resistance, and failing to meet the long-term use requirements of the space radiation environment.

[0004] 3. Due to their unique 4f electronic structure, rare earth elements exhibit superior properties in semiconductors that differ from traditional dopants. Among them, cerium (Ce) can act as an acceptor in InP, and its electrical activity originates from Ce. 3+ / Ce 4+ The valence change process. Theoretical calculations and experimental studies (Phys.Rev.B40,9812(1989)) show that Ce and InP anions (P) have a strong bonding tendency. This strong localization of bonding, coupled with the 4f electron being absorbed by the outer 5s electron, leads to the formation of the valence change process. 2 5p 6 The effective shielding of electrons ensures that the electronic configuration of Ce-doped centers remains relatively stable when bombarded by high-energy particles. This results in a high energy barrier for interaction with point defects, making it difficult to form new, highly electrically active recombination centers. Furthermore, the large mass of Ce atoms allows their substitutional behavior in the crystal lattice to exert a "pinning" effect on the surrounding lattice, inhibiting defect migration and aggregation. These characteristics suggest that Ce-doped InP has the potential to become a neutron-resistant P-type material. However, this property has never been applied to the design and radiation hardening of SPAD devices, failing to meet the high radiation resistance requirements of space-use SPADs, thus creating a contradiction between technological needs and applications.

[0005] Therefore, how to solve the technical bottlenecks of existing single-photon avalanche diodes (SPADs), such as performance degradation due to space radiation, poor radiation resistance of traditional P-type doping processes, and the lack of application of new radiation-resistant doping technologies, and how to overcome the stringent limitations of the space radiation environment, develop new P-type doping technologies by utilizing the excellent radiation resistance properties of Ce-doped InP, improve the radiation resistance and space application lifetime of SPADs, and promote their long-term stable application in space missions such as deep space laser communication and inter-satellite quantum communication, has become an important issue that urgently needs to be addressed by those skilled in the art. Summary of the Invention

[0006] This invention addresses the core problems of existing InP / InGaAs-based SPADs for space applications, namely poor neutron radiation resistance, high dark count rate, and short space operating lifetime. It provides a cerium-doped InP-based single-photon avalanche detector for space communication with a reasonable structure and controllable fabrication process, along with its fabrication method. This invention achieves dual radiation protection for both the bulk and surface phases of the detector by incorporating an intrinsic InP layer / diffusion layer in the detector's epitaxial structure and fabricating a Ce-doped InP P-type contact layer using a spin-coating source diffusion process. Combined with Ce-doping modification of a stacked passivation dielectric film, this results in low dark noise and high detection sensitivity. The fabrication process is compatible with mature III-V compound semiconductor processes, requiring no additional specialized equipment and possessing promising prospects for industrial-scale mass production. To achieve the above objectives, this invention employs the following technical solutions: The first aspect of this invention provides a single-photon avalanche detector for space communication based on cerium-doped InP. The epitaxial structure of the detector comprises, from bottom to top: a substrate (1), an N-type buffer layer (2) for reducing series resistance, an absorption layer (3) for absorbing photons in the communication band, a band transition layer (4) for smoothing band shift and reducing carrier barrier, an N-type charge layer (5) for adjusting the electric field distribution between the absorption region and the multiplication region, an intrinsic multiplication layer (6) for serving as the region where avalanche multiplication occurs, and an intrinsic InP layer (7) serving as a diffusion layer; a cerium (Ce)-doped P-type contact layer (71) is grown on the surface of the intrinsic InP layer (7).

[0007] Furthermore, the cerium (Ce) doping concentration of the P-type contact layer (71) ranges from 1 × 10⁻⁶. 17 cm -3 ~5×10 18 cm -3 This concentration range can achieve stable P-type conductivity while maximizing the radiation resistance properties of Ce.

[0008] Furthermore, the substrate (1) is N + The N-type InP substrate provides a stable growth substrate for the entire epitaxial structure; the N-type buffer layer (2) has a thickness of 0.3~1.0μm and a Si doping concentration of (0.2~8)×10⁻⁶. 18 cm -3 N + A type InP buffer layer is used to reduce the series resistance of devices and optimize the efficiency of electrical signal transmission.

[0009] Furthermore, the absorption layer (3) is intrinsically In 0.53 Ga 0.47The As absorption layer, with a thickness of 1.5~2.5μm, is specifically used to absorb photons in the 1550nm communication band and generate photogenerated electron-hole pairs; the band transition layer (4) is an unintentionally doped InGaAsP layer with a band gap between InGaAs and InP and a thickness of 0.1~0.3μm, used to smooth the band shift of the two materials, reduce the energy barrier for carriers to be transported from the absorption layer to the multiplication region, and reduce carrier interface recombination.

[0010] Furthermore, the N-type charge layer (5) has a thickness of 0.1~0.6μm and a Si doping concentration of (0.4~2)×10⁻⁶. 17 cm -3 The N-type InP charge layer is used to precisely adjust the electric field distribution in the absorption region and the multiplication region to ensure the stable occurrence of carrier directional transport and avalanche multiplication; the intrinsic multiplication layer (6) is a high-purity intrinsic InP layer with a thickness of 0.8~1.2μm, which is the core region where the avalanche multiplication effect occurs. Under reverse bias, a high electric field region is formed to cause photogenerated carriers to undergo avalanche multiplication; the intrinsic InP layer (7) is an intrinsic InP diffusion layer with a thickness of 0.8~3μm, which provides a uniform lattice substrate for Ce atom diffusion and ensures the doping uniformity of the P-type contact layer (71).

[0011] The superior radiation resistance of the Ce-doped InP-based P-type contact layer in this invention stems from the unique interaction mechanism between Ce and the InP lattice and irradiation defects. Its microscopic physical basis is as follows: neutron irradiation in InP semiconductors primarily generates primary knockout atoms (PKA) through elastic collisions, triggering cascade collisions and ultimately producing a large number of Frenkel defect pairs (vacancy-interstitial atom pairs). Among these, phosphorus vacancies (V_P) are typical irradiation-induced defects with strong donor characteristics. In traditional Zn doping, Zn, as a substitutional acceptor (Zn_In), readily reacts with adjacent V_P to form Zn_In-V_P pairs (A-center type defects). These defects are located deep within the principal energy level and are highly efficient electron-hole generation-recombination centers, significantly exacerbating the dark count rate. In contrast, Ce in InP primarily substitutes for In sites to form Ce... 3+ Form exists, through Ce 3+ / Ce 4+ The valence change process releases holes to achieve P-type conduction, where the 4f electron shell is completely filled by the outer 5s electron shell. 2 5p 6 Electron-high shielding, with a highly localized electron cloud and weak coupling to the lattice, results in strong bonding with P atoms. When irradiated point defects such as V_P are close to the Ce doping center, the probability of Ce forming an electrically active defect complex with V_P is much lower than that of Zn. Simultaneously, Ce's... The variable valence property provides a charge compensation mechanism, effectively neutralizing or passivating the donor states introduced by V_P, preventing them from evolving into efficient generation-recombination centers (GR centers). Furthermore, Ce doping can increase the formation energy of point defects such as V_P and reduce their mobility, suppressing the formation rate of harmful defect complexes at their source. Therefore, after neutron irradiation, the InP:Ce layer exhibits a significantly lower density of electrically active deep-level defects than the traditional InP:Zn layer, providing core theoretical support for improving the radiation resistance of detectors.

[0012] The first aspect of this case provides a method for fabricating a single-photon avalanche detector for space communication based on cerium-doped InP, which is used to fabricate the single-photon avalanche detector described in the first aspect. The method includes the following steps: Step S1, Epitaxial Structure Growth: Prepare a substrate (1) and use chemical deposition technology to perform epitaxial growth from bottom to top on the front surface of a single-sided polished InP substrate to form a complete detector epitaxial structure; the detector epitaxial structure includes a substrate (1), an N-type buffer layer (2), an absorption layer (3), a band transition layer (4), an N-type charge layer (5), an intrinsic multiplication layer (6), and an intrinsic InP layer (7). Step S2, P-type contact layer preparation: A dielectric film is grown on the surface of the intrinsic InP layer (7) using PECVD. The dielectric film is opened to expose the intrinsic InP layer (7), and diffusion holes are formed. The cerium (Ce) doped P-type contact layer (71) is prepared by spin coating source diffusion. Step S3, Device Forming Process: The detector epitaxial structure with P-type contact layer (71) is etched sequentially to form mesa and repair sidewall defects, grow stacked passivation dielectric film and perform Ce doping modification, open metal contact area and deposit contact metal to complete the detector fabrication.

[0013] Further, step S2 includes the following sub-steps: S21. Dielectric film growth and pore opening: A dielectric film is grown on the surface of the intrinsic InP layer (7) using PECVD. The dielectric film is then opened to expose the diffusion layer and a diffusion hole is formed to define a precise region for Ce atom diffusion. S22. Spin coating and drying: Prepare a cerium-based doped source solution, place the detector epitaxial structure on a spin coater to spin coat the doped source solution, and after spin coating, perform drying treatment to remove the solution solvent and form a uniform doped source film. S23. Diffusion annealing and residue removal: The dried detector epitaxial structure is subjected to diffusion annealing to allow cerium atoms to diffuse into the lattice of the intrinsic InP layer (7) to form a P-type contact layer (71). Then, the surface residual dopant source and dielectric film are removed by cleaning to obtain a pure P-type contact layer (71).

[0014] Furthermore, the cerium-based dopant source solution in step S22 is a 0.1 mol / L tris(cyclopentadienyl)cerium toluene solution, and a surfactant is added to the solution to improve the wettability of the solution on the semiconductor surface and avoid droplet aggregation or missed coating during spin coating; the spin coating speed is 3000 rpm, the spin coating time is 30~90 seconds, the drying temperature is 120℃, and the drying time is 5 minutes. These parameters can ensure that the dopant source solution uniformly covers the diffusion hole area; the diffusion annealing in step S23 is carried out under N2 atmosphere, the annealing temperature is 650℃, and the annealing time is 45 minutes, which provides suitable thermodynamic conditions for Ce atom diffusion and ensures that Ce atoms uniformly enter the InP lattice; acetone and isopropanol are used for sequential ultrasonic cleaning to remove residues and thoroughly remove surface organic impurities and unreacted dopant sources.

[0015] Further, step S3 includes the following sub-steps: S31. Mesa preparation and sidewall repair: The detector epitaxial structure is patterned by dry etching process to form a mesa structure, with the etching depth reaching the N-type buffer layer (2); then the mesa sidewall is repaired by wet process to eliminate the sidewall lattice damage introduced by dry etching and reduce sidewall leakage. S32. Stacked passivation and Ce doping modification: On the repaired mesa surface, a SiO2 / SiN stacked passivation dielectric film is grown using PECVD process to neutralize dangling bonds on the semiconductor surface and reduce the surface state density; then, using CeO2 as the target material, the stacked passivation dielectric film is modified by Ce doping using magnetron sputtering process. After sputtering, rapid thermal annealing is performed to integrate Ce into the dielectric film to improve its radiation resistance. S33. Electrode fabrication: P-type and N-type metal contact areas are formed on the surface of the P-type contact layer (71) and the N-type buffer layer (2) respectively by photolithography and etching technology; contact metal is deposited in the contact area by EB evaporation or magnetron sputtering process to form ohmic contact electrodes and complete the detector fabrication.

[0016] Furthermore, the dry etching process in step S31 is an ICP dry etching process, which has high etching precision, good anisotropy, and can form a regular mesa structure; the wet repair uses a citric acid-hydrogen peroxide mixed solution, and the repair time is 30-60 seconds, which can effectively eliminate the sidewall lattice damage introduced by dry etching, while avoiding corrosion of the functional layer of the device; the magnetron sputtering process in step S32 has a sputtering power of 60-90W, a sputtering atmosphere of argon, and a deposition temperature of 280-320℃. The injection pressure is 0.5-0.8 Pa, which ensures that Ce atoms are uniformly doped into the stacked passivation dielectric film; the rapid thermal annealing is RTP annealing, with an annealing temperature of 450℃ and an annealing time of 1 min, which solidifies the Ce doping effect, allowing Ce to be incorporated into the dielectric film in the form of "substitutional doping" or "mixed oxide network"; the contact metal in step S33 is a Ti / Pt / Au metal layer, which forms an ohmic contact electrode after peeling and cleaning, reducing contact resistance and ensuring efficient and stable transmission of electrical signals.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. Superior Neutron Radiation Resistance: This invention is the first to apply the radiation-resistant properties of Ce-doped materials to SPAD devices. A Ce-doped P-type contact layer is formed on the surface of the intrinsic InP / diffusion layer using a spin-coating source diffusion process. The 4f electron shell of Ce atoms is highly shielded by outer electrons, forming strong bonds with P atoms in InP. This makes it difficult for Ce to form electrically active complexes with radiation-induced point defects. Simultaneously, Ce's variable valence properties neutralize donor states introduced by point defects, suppressing the formation of deep-level defects (especially A-center type defects) that cause a surge in dark counts at the source. Therefore, after exposure to equivalent space neutron fluence irradiation, the increase in dark count rate is significantly lower than that of traditional Zn-doped devices. Furthermore, combined with Ce-doped modification of the stacked passivation dielectric film, dual radiation protection is achieved at both the bulk and surface levels, effectively reducing the increase in dark count rate under radiation conditions and significantly improving the detector's space operating lifetime.

[0018] 2. High Stability and Low Noise: The detector adopts the classic SACM structure, with precise optimization of the thickness and doping concentration of each functional layer. The band transition layer effectively solves the band mismatch problem between InGaAs and InP, improving carrier transport efficiency. The intrinsic InP diffusion layer provides a uniform substrate for Ce atom diffusion, ensuring the doping uniformity of the P-type contact layer, precisely controlling the detector's electric field distribution, reducing breakdown voltage fluctuations and band-edge tunneling current, and achieving low background dark noise and high detection sensitivity. Ce has an extremely low diffusion coefficient in InP, forming an extremely steep and thermally stable doping profile, which is beneficial for precisely controlling the electric field distribution in the SACM structure, reducing device breakdown voltage fluctuations and band-edge tunneling current, thereby obtaining even lower background dark noise.

[0019] 3. Strong process feasibility and compatibility: The fabrication process of this invention is fully compatible with mature III-V compound semiconductor MOCVD epitaxy, PECVD film deposition, ICP etching, magnetron sputtering, and other processes, without the need to develop extremely special equipment or process parameters. The spin-coating source diffusion process is simple to operate and the doping concentration is controllable, making it suitable for large-scale and batch production. The process parameters of each step are precisely defined, ensuring the repeatability and consistency of detector fabrication and high product yield. In addition, the fabrication process follows the logic of "epitaxy growth → P-type contact layer preparation → device forming". First, a complete functional layer stack is formed, then the core radiation-resistant P-type contact layer is prepared through diffusion process, and finally the device is formed through etching, passivation, and electrode preparation. Each step is progressive, and processes such as wet repair, stacked passivation, and Ce doping modification specifically solve problems such as leakage and defects in the device fabrication process, further improving the performance stability of the detector. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the preparation method steps in Example 2; Figure 2 This is a schematic diagram of the epitaxial wafer structure of the detector of the present invention; Figure 3 This is a schematic diagram of the structure after the diffusion process is completed during the fabrication of the detector of the present invention; Figure 4 This is a schematic diagram of the structure of the detector product of the present invention; Figure 5 This is a schematic diagram of the fabrication process of the P-type contact layer of the single-photon avalanche detector of the present invention.

[0021] In the figure: 1-substrate, 2-N-type buffer layer, 3-absorber layer, 4-band transition layer, 5-N-type charge layer, 6-intrinsic multiplication layer, 7-intrinsic InP layer, 71-P-type contact layer. Detailed Implementation

[0022] The present invention will be further described in detail below with reference to specific embodiments. These embodiments are only used to explain the present invention and are not intended to limit the scope of protection of the present invention. All technical solutions obtained by equivalent substitution or equivalent transformation fall within the scope of protection of the present invention. Example

[0023] like Figures 2-5 As shown, the epitaxial structure of the cerium-doped InP-based single-photon avalanche detector for space communication, from bottom to top, includes: Substrate (1), N-type buffer layer (2), absorption layer (3), band transition layer (4), N-type charge layer (5), intrinsic multiplication layer (6), and intrinsic InP layer (7) as diffusion layer; The diffusion layer (7) has a cerium (Ce) doped P-type contact layer (71) grown on its surface.

[0024] As described above, the cerium-doped InP-based single-photon avalanche detector for space communication provided in this embodiment forms a complete integrated separate absorption, charge multiplication (SACM) system through clear functional division and layer-by-layer adaptation of each layer. Furthermore, the orderly stacking of each layer achieves a full-function coverage structure encompassing "support-signal transmission-photon absorption-bandgap transition-electric field modulation-avalanche multiplication-doped substrate," providing robust structural support for the detector's high detection sensitivity and low background noise. Furthermore, the single-photon avalanche detector in this case innovatively grows a Ce-doped P-type contact layer on the surface of the intrinsic InP layer (7), which serves as the diffusion layer. Ce atoms have a unique 4f electronic structure, and their outer electrons can effectively shield radiation damage and form strong bonds with the InP lattice. They are not prone to forming electrically active defect complexes with point defects generated by space neutron irradiation. At the same time, the variable valence characteristics of Ce can compensate for the charge imbalance introduced by radiation defects, suppress the formation of carrier generation-recombination centers (GR centers), reduce the surge in dark count rate after irradiation from the source, and significantly improve the long-term working stability and service life of the detector in the high-energy particle radiation environment of space, solving the core pain point of weak radiation resistance of traditional Zn-doped P-type layers. Moreover, in this structure, the intrinsic InP layer (7) also serves as a diffusion layer. Its lattice integrity is high and its thickness is suitable, providing a uniform and stable substrate for Ce atom diffusion, ensuring that the Ce doping concentration of the P-type contact layer (71) is uniform, achieving stable P-type conductivity and maximizing the radiation resistance characteristics of Ce. Meanwhile, the interface between the intrinsic InP layer (7) and the upper P-type contact layer (71) and the lower intrinsic multiplication layer (6) is well connected, which does not affect the stable occurrence of the avalanche multiplication effect, nor does it hinder the directional transport of charge carriers, thus achieving dual optimization of radiation resistance and photoelectric detection performance. In addition, the design of the absorption layer (3) and the band transition layer (4) is precisely adapted to the 1550nm and other communication bands to ensure efficient photon absorption; the N-type charge layer (5) precisely controls the electric field distribution to ensure the stability of avalanche multiplication and reduce signal interference; the P-type contact layer (71) is well adapted to the subsequent electrode process and can achieve efficient extraction of electrical signals. The overall structure fits the core requirements of space communication for detectors of "high sensitivity, low noise, radiation resistance and long life" and has strong adaptability. Example

[0025] like Figures 1-5 As shown, a method for fabricating a cerium-doped InP-based single-photon avalanche detector for space communication includes: Step S1: Epitaxial Structure Growth Select N with (100) crystal orientation +Using a single-sided polished InP substrate as the substrate (1), functional layers are epitaxially grown sequentially from bottom to top on the positive surface of the substrate using MOCVD chemical deposition technology: The growth thickness is 0.3~1.0μm, such as 0.5μm, and the Si doping concentration is 4×10⁻⁶. 18 cm -3 N + Type InP buffer layer, as N-type buffer layer (2); Intrinsic In atoms with a growth thickness of 1.5~2.5μm (e.g., 2.0μm) 0.53 Ga 0.47 As the absorption layer, serving as the absorption layer (3); An unintentionally doped InGaAsP bandgap transition layer (band gap ≈ 1.0 eV) with a thickness of 0.1~0.3 μm (e.g., 0.2 μm) is grown as a bandgap transition layer (4). The growth thickness is 0.1~0.6μm (e.g., 0.3μm), and the Si doping concentration is (0.4~2)×10⁻⁶. 17 cm -3 (e.g., 1×10) 17 cm -3 The N-type InP charge layer is used as the N-type charge layer (5); A high-purity intrinsic InP multiplication layer with a thickness of 0.8~1.2μm (e.g. 1.0μm) was grown as an intrinsic multiplication layer (6). An intrinsic InP layer (7) with a thickness of 0.8~3μm (e.g. 1.5μm) is grown as a diffusion layer; After epitaxial growth is completed, a complete detector epitaxial structure is obtained.

[0026] Thus, in step S1, multiple functional layers are grown in an orderly manner from bottom to top using the MOCVD process. The thickness and doping concentration of each layer are strictly matched to ensure high lattice matching degree between layers and excellent interface quality. This effectively reduces the accumulation of interlayer stress and defect transmission, and avoids problems such as epitaxial layer delamination and cracking in subsequent processes. Among them, the N-type buffer layer (2) significantly reduces the series resistance of the device through high Si doping design, providing a guarantee for efficient transmission of electrical signals; the absorption layer (3) has a thickness precisely controlled at 1.5~2.5μm, which can maximize the absorption of photons in the 1550nm communication band and improve the photoresponse efficiency; the band transition layer (4) smooths the band shift between the absorption layer and the charge layer, reducing carrier interface recombination loss; the doping concentration of the N-type charge layer (5) is precisely controlled, realizing the optimization of the electric field distribution in the absorption region and the multiplication region, providing a prerequisite for the stable occurrence of avalanche multiplication; the high-purity intrinsic InP multiplication layer (6) ensures the low noise characteristics of the avalanche multiplication process; the intrinsic InP diffusion layer (7) serves as a dedicated substrate for Ce atom diffusion, with high lattice integrity, providing a good foundation for the subsequent formation of a uniformly doped P-type contact layer (71). The entire epitaxial growth process adopts mature MOCVD technology, which has strong controllability and good repeatability, and is suitable for wafer-level mass production, laying a solid foundation for the efficient development of subsequent processes.

[0027] Step S2, P-type contact layer preparation: A dielectric film is grown on the surface of the intrinsic InP layer (7) using PECVD. The dielectric film is then opened to expose the intrinsic InP layer (7), and diffusion holes are formed. The cerium-doped P-type contact layer (71) is prepared by diffusion through spin-coating source. The specific steps include the following: S21. Dielectric film growth and opening: A dielectric film is grown on the surface of the diffusion layer (7) using PECVD process. The dielectric film is then opened by photolithography and etching technology to expose a designated area of ​​the diffusion layer (7) and form a diffusion hole. S22. Spin coating and drying: Prepare a 0.1 mol / L tri(cyclopentadienyl)cerium toluene solution with the chemical formula Cp3Ce. Add a small amount of surfactant to the solution to improve wettability and stir evenly. Place the above detector epitaxial structure on a spin coater and spin coat the doped source solution at a speed of 3000 rpm for 30-90 seconds. After spin coating, place the epitaxial structure on a hot plate at 120°C and dry for 5 minutes. S23. Diffusion annealing and residue removal: The dried epitaxial structure is placed in a tube annealing furnace and diffused under N2 atmosphere protection. The annealing temperature is 650℃ and the annealing time is 45 minutes, so that Ce atoms diffuse into the lattice of the diffusion layer (7). After annealing, the epitaxial structure is placed in acetone and isopropanol for ultrasonic cleaning to remove the residual organic dopant source and dielectric film on the surface, and a Ce-doped P-type contact layer (71) is formed on the surface of the diffusion layer (7).

[0028] As described above, step S2 precisely prepares a Ce-doped P-type contact layer (71) through a combination of "dielectric film opening + spin-coating dopant source + diffusion annealing". Compared with traditional MOCVD in-situ doping, this process does not require complex in-situ doping equipment debugging, making it more flexible and cost-effective. Among them, the dielectric film growth and opening steps precisely define the diffusion area of ​​Ce atoms, avoiding uncontrolled doping range; the 0.1 mol / L Cp3Ce toluene solution combined with a surfactant significantly improves the wettability of the solution on the semiconductor surface; the spin-coating speed of 3000 rpm and the drying process at 120°C ensure the formation of a uniform dopant source film, providing a guarantee for the uniform diffusion of Ce atoms; the diffusion annealing at 650°C for 450 minutes in an N2 atmosphere provides suitable thermodynamic conditions for Ce atoms to enter the InP lattice, and the Ce doping concentration of the final P-type contact layer (71) is stable at 1×10⁻⁶. 17 cm -3 ~5×10 18 cm -3 This achieves stable P-type conductivity while fully utilizing Ce's radiation resistance properties. Sequential ultrasonic cleaning with acetone and isopropanol thoroughly removes residual organic impurities and unreacted dopant sources from the surface, avoiding interference from impurities on device performance and ensuring the purity of the P-type contact layer (71).

[0029] Step S3, Device Fabrication Process: The detector epitaxial structure with the P-type contact layer (71) is sequentially etched to form mesa and repair sidewall defects, grown in a stacked passivation dielectric film and modified with Ce doping, and the metal contact region is opened and the contact metal is deposited to complete the detector fabrication. Specifically, this includes the following sub-steps: S31. Mesa preparation and sidewall repair: The epitaxial structure with a P-type contact layer (71) was patterned by ICP dry etching process, and the etching depth was up to the N-type buffer layer (2) to form a regular detector mesa structure. Then the epitaxial structure was placed in a citric acid-hydrogen peroxide mixed solution for wet repair for 45 seconds to eliminate the sidewall lattice damage introduced by dry etching. After removal, it was rinsed with deionized water and dried. S32. Stacked Passivation and Ce Doping Modification: A SiO2 / SiN stacked passivation dielectric film was grown on the repaired mesa surface using PECVD. CeO2 was used as the target material, and Ce doping modification of the stacked passivation dielectric film was performed using magnetron sputtering. The sputtering power was 60-90W, the atmosphere was argon, the deposition temperature was 280-320℃, and the sputtering pressure was 0.5-0.8Pa. After sputtering doping, RTP annealing was performed at 450℃ for 1 minute to allow Ce atoms to integrate into the dielectric film. S33. Electrode fabrication: Using photolithography and etching techniques, a P-type metal contact region (101) is formed on the surface of the P-type contact layer (71), and an N-type metal contact region (102) is formed on the surface of the N-type buffer layer (2). Ti / Pt / Au metal layers are deposited in the two contact regions using EB evaporation process. After peeling and cleaning, ohmic contact electrodes are formed, and the fabrication of a cerium-doped InP-based single-photon avalanche detector for space communication is completed.

[0030] As described above, step S31 of step S3 uses ICP dry etching to form a mesa structure, with the etching depth precisely controlled to the N-type buffer layer (2). The etching anisotropy is good and the edges are regular, effectively avoiding electrical crosstalk between adjacent detector units. The subsequent wet repair with a citric acid-hydrogen peroxide mixed solution specifically eliminates the sidewall lattice damage introduced by dry etching, reduces sidewall dangling bonds and defect states, reduces the risk of sidewall leakage from the source, and further optimizes the dark counting performance of the device. This combined process retains the high precision advantage of dry etching and makes up for its lattice damage shortcomings through wet repair, achieving the dual effects of precise forming and defect repair, ensuring the structural integrity and electrical stability of the detector. The growth of the SiO2 / SiN stacked passivation dielectric film in step S32 effectively neutralizes dangling bonds on the semiconductor surface, reduces surface state density, and decreases surface leakage current and carrier recombination. Based on this, the passivation film is modified by magnetron sputtering with a CeO2 target. Precise control of sputtering power (60-90W), deposition temperature (280-320℃), and sputtering pressure (0.5-0.8Pa) ensures uniform integration of Ce atoms into the dielectric film. Further RTP annealing at 450℃ for 1 min ensures that Ce exists stably in the passivation film as a "substitutional dopant" or "mixed oxide network," significantly improving the radiation resistance of the dielectric film. This step, synergistically with the P-type contact layer formed in step S2, constructs a dual system of bulk and surface radiation resistance, effectively suppressing the formation of deep-level defects induced by space neutron irradiation and significantly extending the detector's space operating lifetime. Furthermore, step S33 precisely creates the P / N electrode metal contact area using photolithography and etching techniques, achieving high positioning accuracy and avoiding electrical performance abnormalities caused by contact area misalignment. The Ti / Pt / Au metal layer deposited by EB vapor deposition forms the ohmic contact electrode. The Ti layer ensures good adhesion between the electrode and the semiconductor surface, the Pt layer blocks Au atom diffusion, and the Au layer reduces contact resistance. These three elements work together to ensure the stability and conductivity of the electrode. After stripping and cleaning, the electrode pattern is regular and the size is controllable, enabling efficient electrical connection between the detector and external circuits, reducing losses and interference during electrical signal transmission, and further optimizing the detector's detection sensitivity and response speed.

[0031] In a preferred embodiment, the Ce doping concentration in the P-type contact layer (71) / InP:Ce layer is in the range of 1×10⁻⁶.17 cm -3 ~5×10 18 cm -3 .

[0032] The selection criteria and analysis for the Ce doping concentration range are as follows: I. Lower limit (1×10) 17 cm -3 The drawback of radiation protection is its limited effectiveness. 1. In a SACM structure, the surface charge density of the charge layer determines the electric field distribution between the absorption and multiplication regions. If the doping concentration is too low, the charge layer cannot provide sufficient surface charge density, causing the electric field to extend excessively into the absorption region, increasing the tunneling dark current. When the Ce doping concentration is 5 × 10⁻⁶... 16 cm -3 (less than 1×10) 17 cm -3 When the charge layer thickness is increased to >1μm, the target electric field needs to be maintained. However, the increased defect density introduced by the thick layer causes the device breakdown voltage (Vbr) to drift by more than ±2V, and the dark count rate (DCR) increases to >1kHz at 220K (target value <300Hz).

[0033] 2. Low-concentration Ce cannot effectively "capture" or "neutralize" irradiation-induced point defects (such as phosphorus vacancies Vp). The large interatomic spacing of Ce results in a low probability of encountering Ce during defect migration, making it difficult to form stable Ce-defect complexes, leading to a still high deep-level defect density after irradiation. At 1×10⁻⁶... 16 cm -3 (less than 1×10) 17 cm -3 In Ce-doped samples, after neutron irradiation with 1×10¹¹ n / cm², deep-level transient spectroscopy (DLTS) detected a significant defect peak (energy level Ec-0.52 eV) near the band gap center, corresponding to a defect density of ~6×10¹¹ n / cm². 14 cm -3 After irradiation, the DCR increased by more than 100 times, and the radiation resistance performance was not significantly different from that of the undoped sample.

[0034] Therefore, if the Ce doping concentration is below 1×10⁻⁶, 17 cm -3 This approach cannot guarantee that the P-type layer has sufficient conductivity and electric field control capability, leading to increased dark count and deterioration of device electrical performance. It also cannot form an effective radiation defect suppression effect, resulting in significantly higher defect density and dark count growth rate after irradiation, which cannot meet the requirements of space applications.

[0035] II. Upper limit (5×10) 18 cm -3Degradation of materials and device performance: deterioration of material quality and device performance 1. Ce has a larger atomic radius (~1.03 Å) than In (~0.94 Å), and high-concentration doping leads to significant lattice distortion. X-ray diffraction (XRD) shows that when the Ce concentration is >5 × 10⁻⁶, significant lattice distortion occurs. 18 cm -3 When the (004) diffraction peak's full width at half maximum (FWHM) broadens from ~25 arcsec to >60 arcsec, it indicates a decrease in crystal quality. Excess Ce may form Ce-P precipitates (such as CeP or Ce3P4), which act as nonradiative recombination centers and leakage channels, degrading device performance. When the Ce doping concentration is 1×10⁻⁶... 19 cm -3 (greater than 5×10) 18 cm -3 When the Ce concentration distribution was observed, a cluster of ~5 nm in size was observed near the interface using transmission electron microscopy (TEM), and secondary ion mass spectrometry (SIMS) showed a "tailing" and accumulation peaks in the Ce concentration distribution.

[0036] 2. High-concentration doping introduces enhanced scattering of ionized impurities, resulting in a significant decrease in hole mobility. Hall effect measurements show that when the Ce concentration increases from 5 × 10⁻⁶, the scattering of ionized impurities decreases. 18 cm -3 Increased to 1×10 19 cm -3 (greater than 5×10) 18 cm -3 When the mobility decreases from ~60 cm² / Vs to ~25 cm² / V*s, the mobility drops from ~60 cm² / V*s. This decrease in mobility leads to an increase in the series resistance of the P-type contact layer, affecting the transient response speed and high-frequency performance of the device.

[0037] Therefore, it can be concluded that if the Ce doping concentration is higher than 5 × 10⁻⁶... 18 cm -3 On the one hand, the difference in atomic radius will cause significant lattice distortion, which will easily lead to the precipitation of second phase and interface clusters, introduce non-radiative recombination centers and leakage channels, resulting in significant deterioration of crystal quality and device performance. On the other hand, it will exacerbate the scattering of ionized impurities, reduce hole mobility and increase series resistance, and degrade the transient response and high-frequency operating performance of the device, which will also fail to meet the reliability and performance requirements of space applications.

[0038] III. Validation of Optimal Concentration Range Within this range, Ce concentration exhibits a good linear relationship with charge density at the charge level, allowing for precise control of Vbr within 65 ± 1 V (target value). Hall effect measurements show that hole concentration increases linearly with Ce concentration, with ionization approaching 100%, indicating that Ce acts as an effective acceptor. Ce concentrations within 1 × 10⁻⁶... 17 cm -3 ~5×1018 cm -3 At that time, the deep-level defects generated after irradiation were <8×10 13 cm -3 At this concentration, the interatomic spacing of Ce matches the mean free path of the irradiated defects, enabling the most efficient defect trapping and passivation.

[0039] Note: Neutron irradiation conditions are 1×10⁻⁶. 11 n / cm2 (1MeVSi equivalent), the test temperature was 220K.

[0040] As shown in the table above, when the Ce doping concentration is 1×10⁻⁶... 17 cm -3 ~5×10 18 cm -3 Within the specified range, the device breakdown voltage Vbr is stable and controllable, the initial dark count (DCR) is low, and the DCR growth factor after neutron irradiation is significantly smaller than that of the low-concentration or excessively high-concentration group. At the same time, the detection efficiency (PDE) remains at a high level, resulting in the best overall photoelectric performance and radiation resistance.

[0041] In summary, the preparation method in Example 1 of this case features smooth transitions between steps and controllable process parameters. Each step specifically addresses issues in traditional processes such as poor radiation resistance, high dark count, and insufficient process compatibility. The resulting cerium-doped InP-based SPAD not only possesses excellent neutron radiation resistance but also boasts advantages such as low dark noise and high detection sensitivity. Furthermore, the preparation process is compatible with existing semiconductor equipment, making it suitable for large-scale mass production. This provides strong support for its promotion in high-end applications such as space laser communication and quantum key distribution. Example

[0042] Example 3 is a comparative example of Example 2. Except for replacing the P-type contact layer fabrication process with a Zn-doped InP (InP:Zn) process of the same concentration, the epitaxial structure and fabrication steps in Example 3 are completely identical to those in Example 2; including the following process steps: 1. Diffusion process: After growing a dielectric film and opening holes on the surface of the diffusion layer using PECVD, Zn is diffused using MOCVD to form a Zn-doped InP P-type contact layer. 2. Device fabrication: The diffusion holes are micro-processed using an acidic solution, followed by electrochemical sulfidation to form a passivation layer. Then, a SiO2 passivation layer is grown using PECVD, and finally, the contact holes and P / N electrodes are fabricated.

[0043] The performance tests and results analysis of Example 3 and Example 2 are as follows: The detectors prepared in Examples 2 and 3 were subjected to equivalent space neutron fluence irradiation performance tests. The core test indicator was the dark count rate (DCR). The test results are as follows: Before irradiation, the detector prepared in Example 2 had a background dark count rate of 80 Hz, while the background dark count rate in Comparative Example 2 / Example 3 was 95 Hz. The two had comparable basic photoelectric performance, proving that the Ce doping process of the present invention did not sacrifice the basic detection performance of the detector. After irradiation with an equivalent space neutron fluence, the dark count rate of the detector prepared in Example 2 increased to 1.5 kHz, while the dark count rate of the detector in the comparative example increased sharply to over 12 kHz. Thus, it can be seen that the dark count rate of the detector in Example 2 after irradiation increases by only 1 / 10 of that in the comparative example, which fully demonstrates that the cerium-doped InP-based SPAD of the present invention has excellent resistance to neutron irradiation, can effectively suppress radiation-induced dark count rate spikes, and significantly improve the reliability and lifespan of the detector in the space radiation environment.

[0044] In particular, this invention introduces cerium (Ce), a rare earth metal with excellent resistance to neutron irradiation, as the acceptor doping source to prepare a highly stable p-type indium phosphide (InP) charge layer and contact layer, thereby fundamentally improving the long-term operational reliability of the detector in space particle radiation environments (especially neutron irradiation).

[0045] In summary, this invention discloses a single-photon avalanche detector for space communication based on cerium-doped InP and its fabrication method. The detector's epitaxial structure, from bottom to top, includes a substrate, an N-type buffer layer, an absorption layer, a band transition layer, an N-type charge layer, an intrinsic multiplication layer, an intrinsic InP layer, and a cerium-doped P-type contact layer. The core innovation is the dual radiation-resistant design of the Ce-doped P-type contact layer and the Ce-doped modified stacked passivation film. The fabrication method involves MOCVD epitaxial growth, spin-coating source diffusion to prepare the P-type contact layer, ICP etching and wet repair, magnetron sputtering doping, and electrode fabrication. This invention utilizes Ce's unique electronic structure to suppress radiation-induced defects and combines it with an SACM structure to optimize the electric field distribution, achieving low dark count rate, high detection sensitivity, and excellent neutron radiation resistance. The process is compatible with existing semiconductor production lines and is suitable for wafer-level mass production, meeting the high-reliability single-photon detection requirements of aerospace missions such as space laser communication and quantum key distribution.

[0046] The cerium-doped InP-based single-photon avalanche detector for space communication of this invention has a reasonable structure, excellent resistance to neutron radiation, low dark noise, and high detection sensitivity, which fully meets the requirements of aerospace missions such as space laser communication and quantum key distribution. Its preparation method is compatible with mature semiconductor processes, the process parameters are controllable, and the steps are smoothly connected, which can realize large-scale and batch production, and has good industrial applicability and industrialization prospects.

[0047] The above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A single-photon avalanche detector for space communication based on cerium-doped InP, characterized in that, Its epitaxial wafer structure includes, from bottom to top: substrate (1), N-type buffer layer (2), absorption layer (3), band transition layer (4), N-type charge layer (5), intrinsic multiplication layer (6), and intrinsic InP layer (7) as diffusion layer; a cerium-doped P-type contact layer (71) is grown on the surface of the intrinsic InP layer (7).

2. The single-photon avalanche detector according to claim 1, characterized in that, The cerium doping concentration of the P-type contact layer (71) ranges from 1×10⁻⁶. 17 cm -3 ~5×10 18 cm -3 .

3. The single-photon avalanche detector according to claim 1, characterized in that, The substrate (1) is N + Type InP substrate; The N-type buffer layer (2) has a thickness of 0.3~1.0 μm and a Si doping concentration of (0.2~8)×10⁻⁶. 18 cm -3 N + Type InP buffer layer.

4. The single-photon avalanche detector according to claim 1, characterized in that, The absorption layer (3) is intrinsic In 0.53 Ga 0.47 As absorption layer, with a thickness of 1.5~2.5μm; The band transition layer (4) is an unintentionally doped InGaAsP layer with a band gap between InGaAs and InP and a thickness of 0.1~0.3μm.

5. The single-photon avalanche detector according to claim 1, characterized in that, The N-type charge layer (5) has a thickness of 0.1~0.6 μm and a Si doping concentration of (0.4~2)×10⁻⁶. 17 cm -3 N-type InP charge layer; The intrinsic multiplication layer (6) has a thickness of 0.8~1.2μm; the intrinsic InP layer (7) is intrinsically InP diffused and has a thickness of 0.8~3μm.

6. A method for fabricating a single-photon avalanche detector for space communication based on cerium-doped InP, characterized in that, Includes the following steps: Step S1, epitaxial structure growth: Prepare a substrate (1) and use chemical deposition technology to perform epitaxial growth from bottom to top on the positive surface of the substrate (1) to form a complete detector epitaxial structure; the detector epitaxial structure includes a substrate (1), an N-type buffer layer (2), an absorption layer (3), a band transition layer (4), an N-type charge layer (5), an intrinsic multiplication layer (6), and an intrinsic InP layer (7). Step S2, P-type contact layer preparation: A dielectric film is grown on the surface of the intrinsic InP layer (7) using PECVD. The dielectric film is opened to expose the intrinsic InP layer (7), and diffusion holes are formed. The cerium-doped P-type contact layer (71) is prepared by spin coating source diffusion. Step S3, Device Forming Process: The detector epitaxial structure with P-type contact layer (71) is etched sequentially to form mesa and repair sidewall defects, grow stacked passivation dielectric film and perform Ce doping modification, open metal contact area and deposit contact metal to complete the detector fabrication.

7. The fabrication method of the cerium-doped InP-based single-photon avalanche detector for space communication according to claim 6, characterized in that, Step S2 includes the following sub-steps: S21. Dielectric film growth and opening: A dielectric film is grown on the surface of the intrinsic InP layer (7) using PECVD. The dielectric film is then opened to expose the diffusion layer and a diffusion hole is formed. S22. Spin coating and drying: Prepare a cerium-based doped source solution, place the detector epitaxial structure on a spin coater to spin coat the doped source solution, and then perform a drying process after spin coating; S23. Diffusion annealing and residual removal: The dried detector epitaxial structure is subjected to diffusion annealing to allow cerium atoms to diffuse into the intrinsic InP layer (7) to form a P-type contact layer (71). Then, the surface residual dopant source and dielectric film are cleaned to remove them.

8. The fabrication method of the cerium-doped InP-based single-photon avalanche detector for space communication according to claim 7, characterized in that, The cerium-based doping source solution mentioned in step S22 is a 0.1 mol / L tris(cyclopentadienyl)cerium toluene solution, and a surfactant is added to the solution; the spin coating speed is 3000 rpm, the spin coating time is 30~90 seconds; the drying temperature is 120℃, and the drying time is 5 minutes; The diffusion annealing described in step S23 is carried out under a N2 atmosphere at a temperature of 650°C for 45 minutes; the residue is removed by ultrasonic cleaning with acetone and isopropanol in sequence.

9. The fabrication method of the cerium-doped InP-based single-photon avalanche detector for space communication according to claim 7, characterized in that, Step S3 includes the following sub-steps: S31. Mesa preparation and sidewall repair: The detector epitaxial structure is patterned by dry etching process to form a mesa structure, with the etching depth reaching the N-type buffer layer (2); then the mesa sidewall is repaired by wet process to eliminate the sidewall damage introduced by dry etching. S32. Stacked passivation and Ce doping modification: On the repaired mesa surface, a SiO2 / SiN stacked passivation dielectric film is grown using PECVD process; then, using CeO2 as the target material, the stacked passivation dielectric film is modified by Ce doping using magnetron sputtering process. After sputtering, rapid thermal annealing is performed to integrate Ce into the dielectric film to improve its radiation resistance. S33. Electrode fabrication: Using photolithography and etching techniques, a P-type metal contact area is formed on the surface of the P-type contact layer (71), and an N-type metal contact area is formed on the surface of the N-type buffer layer (2). The contact metal is deposited in the contact area using EB evaporation or magnetron sputtering to form an ohmic contact electrode and complete the detector fabrication.

10. The fabrication method of the cerium-doped InP-based single-photon avalanche detector for space communication according to claim 9, characterized in that, The dry etching process in step S31 is the ICP dry etching process, and the wet repair is carried out using a citric acid-hydrogen peroxide mixed solution, with a repair time of 30-60 seconds. In step S32, the magnetron sputtering process has a sputtering power of 60-90W, a sputtering atmosphere of argon, a deposition temperature of 280-320℃, and a sputtering pressure of 0.5-0.8Pa; the rapid thermal annealing is RTP annealing, with an annealing temperature of 450℃ and an annealing time of 1min. The contact metal mentioned in step S33 is a Ti / Pt / Au metal layer, which is peeled off and cleaned to form an ohmic contact electrode.

Images