Narrow-band gap thermophotovoltaic cell virtual substrate structure and preparation method thereof
By growing a six-layer compositional gradient buffer layer and a compositional overshoot layer on an InP substrate, the dislocation problem caused by large lattice mismatch in InP-based InGaAs thermophotovoltaic cells was solved, achieving the growth of high-quality epitaxial materials and improving the conversion efficiency and reliability of the cells.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI INST OF SPACE POWER SOURCES
- Filing Date
- 2025-12-29
- Publication Date
- 2026-04-21
AI Technical Summary
Existing InP-based InGaAs thermophotovoltaic cells cannot achieve high-quality material growth due to large lattice mismatch at the 2.6 μm wavelength extension, resulting in dislocations and penetration defects that affect cell lifespan and reliability.
By growing six compositional gradient buffer layers and compositional overshoot layers on an InP substrate, and through stepwise compositional jumps and stress decomposition, combined with a lattice hardening layer, complete stress release and dislocation blocking are achieved, forming a high-quality virtual substrate.
Significantly reducing material defect density, improving device conversion efficiency and long-term reliability, and obtaining high-quality InGaAs thermophotovoltaic cells suitable for 2.6 μm extended wavelength.
Smart Images

Figure CN121908678A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of thermophotovoltaic cell technology, specifically relating to a virtual substrate structure and preparation method for a narrow bandgap thermophotovoltaic cell. Background Technology
[0002] Thermophotovoltaic (TPV) systems are high-efficiency power supply systems that use thermal radiation, such as waste heat from industrial exhaust gases, as an energy source. 20%-50% of industrial energy loss is waste heat emission, of which 18%-30% can be utilized. Thermophotovoltaic cells, as the core energy conversion device, convert the generated thermal radiation energy into electrical energy and are a key component affecting system efficiency. To improve the spectral utilization efficiency of the cells, the absorption wavelength of the cell material needs to be extended to 2.0-2.6 μm, corresponding to a bandgap of 0.5-0.6 eV. InGaAs material is considered an ideal cell material due to its flexible bandgap adjustment capability, strong radiation resistance, and good high-temperature resistance. Traditional InP-based InGaAs cells are widely used in the sub-1.7 μm band, such as in fiber optic communication or infrared detection, with a bandgap typically around 0.74 eV, corresponding to the standard In0.53Ga0.47As material. However, when the absorption wavelength extends to 2.6 μm, corresponding to a 0.5 eV material composition of In0.81Ga0.19As, the nearly 2% lattice mismatch with the InP substrate makes high-quality material growth extremely challenging. Incomplete relaxation of mismatch stress leads to severe dislocations, penetration defects, and surface roughness during epitaxial growth. These dislocations not only disrupt the internal crystal integrity of the material but also become non-radiative recombination centers, significantly reducing the minority carrier lifetime and radiative recombination efficiency of the epitaxial layer. This, in turn, causes a series of problems such as increased dark current, shortened minority carrier lifetime, and decreased device reliability. Numerous studies have shown that if the mismatch exceeds 1%, traditional single-layer buffering or simple linear gradient buffering methods are insufficient to effectively release strain, and the material quality often fails to meet the requirements of high-efficiency thermophotovoltaic devices. Existing virtual substrate techniques, such as thick buffer layers, linear composition gradient layers, and superlattice dislocation blocking layers, can alleviate the stress caused by mismatch to some extent, but they are still insufficient to cope with mismatches as high as 2%. Thick buffer layers typically require thicknesses of several micrometers or even greater, resulting in long epitaxial growth times, high costs, and difficulty in fully releasing residual stress. Linear gradient methods still easily lead to crack formation at stress concentration points. Single superlattice dislocation blocking layers have limited dislocation pinning capabilities under such large mismatches. Therefore, to meet the requirements of thermophotovoltaic cells for 2.6 μm long-wavelength In... 0.81 Ga 0.19 The high-quality requirements of As materials necessitate the development of a high-performance virtual substrate structure and fabrication method designed to address the large mismatch problem in InP-based systems. Summary of the Invention
[0003] The purpose of this invention is to overcome the aforementioned defects and provide a virtual substrate structure and fabrication method for narrow bandgap thermophotovoltaic cells. This solves the technical problem that large lattice mismatch between photovoltaic cell materials and substrates prevents high-quality material growth, ultimately affecting cell lifespan and reliability. This invention can obtain a high-quality virtual substrate suitable for the growth of 2.6 μm extended wavelength InGaAs thermophotovoltaic cells, significantly reducing material defect density and improving device conversion efficiency and long-term reliability.
[0004] To achieve the above-mentioned objectives, the present invention provides the following technical solution: A method for fabricating a virtual substrate structure for a narrow bandgap thermophotovoltaic cell includes: An n-InP buffer layer and an m-layer of n-InAs are sequentially epitaxially grown on an n-InP substrate. x P y Buffer layer, component overshoot layer and lattice hardening layer; m-layer n-InAs x P y In the buffer layer, m = 4~8, x = 0.1~0.6, y = 0.4~0.9; Following the order from the inside out, each layer of n-InAs x P y The component ratio of As / P in the buffer layer gradually increases; Component overpass layer and n-InAs x P y The buffer layers contain the same elements, and the As content in the overshoot layer is higher than that in the m-th layer (n-InAs). x P y The As content in the buffer layer and the P content in the overshoot layer are lower than those in the m-th layer n-InAs. x P y The P content of the buffer layer; the thickness of the overshoot layer is greater than that of each n-InAs layer. x P y Buffer layer; The lattice constant of the lattice-hardened layer is consistent with that of the battery base region and other functional layer materials.
[0005] Furthermore, the As content in the lattice-hardened layer is higher than that in the m-th layer n-InAs. x P y The As content in the buffer layer is lower than that in the composition overshoot layer, and the P content in the lattice hardening layer is lower than that in the m-th n-InAs layer. x P y The P content of the buffer layer is higher than that of the component overshoot layer.
[0006] Furthermore, the lattice constant of the over-stretched crystal is 0.1 to 0.2% higher than that of the lattice-hardened layer.
[0007] Furthermore, the thickness of the n-InP substrate is 300~350 μm; The thickness of the n-InP buffer layer is 0.2~0.4μm; m-layer n-InAs x P y The thickness of the buffer layers is equal, ranging from 0.2 to 0.3 μm. The thickness of the over-penetration layer is 0.5~1.5μm; The thickness of the lattice hardening layer is 0.3~0.7μm.
[0008] Furthermore, m layers of n-InAs x P y In the buffer layer, m=6, 6 layers of n-InAs x P y The buffer layers are as follows: n-InAs 0.18 P 0.82 Buffer layer, n-InAs 0.26 P 0.74 Buffer layer, n-InAs 0.33 P 0.67 Buffer layer, n-InAs 0.42 P 0.58 Buffer layer, n-InAs 0.49 P 0.51 Buffer layer, n-InAs 0.56 P 0.44 Buffer layer.
[0009] Furthermore, the n-InP substrate is doped with S atoms at a concentration of 0.8-8 × 10⁻⁶. 18 cm -3 ; The n-InP buffer layer is doped with Si atoms, and the doping concentration is ≥1×10⁻⁶. 18 cm -3 ; m-layer n-InAs x P y The buffer layer is doped with Si atoms, with a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 ; The overshoot layer is composed of Si atoms with a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 ; The doped atoms of the lattice hardening layer are Si, with a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0010] Furthermore, using metal-organic chemical vapor deposition, an n-InP buffer layer and an m-layer n-InAs layer are sequentially epitaxially grown on an n-InP substrate. x P y The layers consist of a buffer layer, a composition overshoot layer, and a lattice hardening layer, each deposited at a temperature of 630-670°C. o C. Furthermore, the doping source for each layer during deposition is SiH4, the special gases used are AsH3 and PH3, and the MO source used is TMGa and TMIn.
[0011] A narrow bandgap thermophotovoltaic cell virtual substrate structure, obtained by the above-mentioned method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure, includes an n-InP substrate, an n-InP buffer layer, and an m-layer n-InAs layer sequentially arranged. x P y Buffer layer, component overshoot layer and lattice hardening layer.
[0012] Furthermore, the relaxation degree of each layer of the virtual substrate structure of the narrow bandgap thermophotovoltaic cell is >95%, the root mean square surface roughness is <5nm, and there are no penetrating dislocations that extend to the surface.
[0013] This invention provides a virtual substrate structure for a narrow bandgap thermophotovoltaic cell, comprising a composition gradient layer, a composition overshoot layer, and a lattice hardening layer sequentially formed on the substrate. For each sub-gradient layer in the composition gradient layer, from bottom to top, the corresponding As / P composition ratio gradually increases. The lattice constant of the composition overshoot layer is slightly larger than that of the lattice hardening layer, used to balance the tensile and compressive strains during the material epitaxy process. By combining the composition gradient layer and the composition overshoot layer, this invention can distribute large lattice mismatch stresses to the steps of each composition gradient layer and the composition overshoot layer. This not only reduces the stress of large mismatch 2.6μm extended wavelength InGaAs thermophotovoltaic cells... 0.81 Ga 0.19 As the surface roughness of the material can be reduced, the dislocation density can be decreased, thereby obtaining a thermophotovoltaic cell device with better performance.
[0014] Compared with the prior art, the present invention has at least one of the following advantages: (1) The present invention grows six-layer composition-gradient buffer layer steps on an InP substrate. Through the step-by-step change of composition, the large mismatch stress is decomposed and completely released at the interface of each step layer. (2) By adding a thicker component overshoot layer, the tensile strain accumulated in the step layer is offset by the compressive strain of the component overshoot layer, thereby improving the flatness of the epitaxial material and blocking the upward extension of penetrating dislocations. (3) The present invention combines the composition gradient layer and the composition overshoot layer, which can distribute the large lattice mismatch stress to each composition gradient layer step and the composition overshoot layer. This can not only reduce the large mismatch In 0.81 Ga 0.19 As the surface roughness of the material can be reduced, the dislocation density can be decreased, thereby obtaining a thermophotovoltaic cell device with better performance; (4) The present invention provides the optimal composition ratio of each layer. Under this ratio, a substrate with optimal lattice constant matching and low surface roughness can be obtained, thereby maximizing the conversion efficiency and long-term reliability of the battery device. (5) The virtual substrate of the present invention should have comprehensive characteristics such as multi-level stress decomposition, strain compensation, effective dislocation blocking and surface flatness, which can significantly improve the conversion efficiency of the final device while ensuring the controllability of epitaxial growth. Attached Figure Description
[0015] Figure 1 This is a schematic diagram of the virtual substrate structure for InP-based narrow bandgap thermophotovoltaic cells. Figure 2 RSM test results for InAsP and InGaAs epitaxial materials (a) 004 plane (b) 115 plane; Figure 3 (a) TEM image of the battery structure; (b) TEM image of the cross-section of the large mismatch InGaAs epitaxial structure. Figure 4 The images show the AFM diagrams of narrow bandgap InGaAs material epitaxially grown on a six-layer InAsP buffer layer, where (a) is the morphology diagram and (b) is the curve diagram. Detailed Implementation
[0016] The features and advantages of the present invention will become clearer and more apparent from the following detailed description.
[0017] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments. Although various aspects of embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated otherwise.
[0018] This invention provides a virtual substrate structure and preparation method for InP-based narrow bandgap thermophotovoltaic cells, reducing the impact of lattice distortion and penetrating dislocations caused by large lattice mismatch between the thermophotovoltaic cell material and the substrate on the surface roughness of the thermophotovoltaic cell.
[0019] This invention discloses a virtual substrate structure and fabrication method for InP-based narrow-bandgap thermophotovoltaic cells. The method involves growing six compositionally stepped buffer layers on an InP substrate using metal-organic chemical vapor deposition (MOCVD). The gradual compositional transitions decompose and release large mismatch stresses at the interfaces of each step layer. Adding a thicker composition overshoot layer cancels out the tensile strain accumulated within the step layers with the compressive strain of the composition overshoot layer, thereby improving the flatness of the epitaxial material and preventing the upward extension of penetrating dislocations. Finally, a lattice hardening layer with a lattice constant consistent with the cell material is deposited, resulting in a high-quality virtual substrate structure for the direct growth of extended-wavelength thermophotovoltaic cell materials with large lattice mismatches and narrow bandgap.
[0020] like Figure 1 The specific manufacturing steps are as follows: An n-InP substrate is provided, and an n-InP buffer layer and an n-InAs layer are epitaxially grown sequentially on the n-InP substrate. 0.18 P 0.82 Buffer layer 1, n-InAs 0.26 P 0.74 Buffer layer 2, n-InAs 0.33 P 0.67 Buffer layer 3, n-InAs 0.42 P 0.58 Buffer layer 4, n-InAs 0.49 P 0.51 Buffer layer 5, n-InAs 0.56 P 0.44 Buffer layer 6,n-InAs 0.64 P 0.36 Component overpass layer, n-InAs 0.60 P 0.40 Lattice hardening layer.
[0021] The substrate is an n-type doped InP substrate with S as the dopant atom and a doping concentration of 0.8-8 × 10⁻⁶. 18 cm -3 Thickness 300~350 μm.
[0022] The n-InP buffer layer has a thickness of 0.3 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0023] n-InAs 0.18 P 0.82 Buffer layer 1 has a thickness of 0.25 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0024] n-InAs 0.26P 0.74 Buffer layer 2 has a thickness of 0.25 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0025] n-InAs 0.33 P 0.67 The buffer layer 3 has a thickness of 0.25 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0026] n-InAs 0.42 P 0.58 The buffer layer 4 has a thickness of 0.25 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0027] n-InAs 0.49 P 0.51 The buffer layer 5 has a thickness of 0.25 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0028] n-InAs 0.56 P 0.44 The buffer layer 6 has a thickness of 0.25 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0029] n-InAs 0.64 P 0.36 The overpass layer thickness is 1 μm, the doped atoms are Si, and the doping concentration is ≥1×10⁻⁶. 18 cm -3 .
[0030] n-InAs 0.60 P 0.40 The lattice-hardened layer has a thickness of 0.5 μm, is doped with Si atoms, and has a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
[0031] This invention provides a virtual substrate structure for InP-based narrow bandgap thermophotovoltaic cells and its fabrication method. By sequentially growing six compositionally stepped gradient buffer layers on an InP substrate and introducing a composition overshoot layer and a lattice hardening layer, the large mismatch stress is released stepwise and strain is compensated, effectively suppressing through dislocations and improving the flatness of the epitaxial surface. This results in a high-quality virtual substrate suitable for the growth of InGaAs thermophotovoltaic cells with an extended wavelength of 2.6 μm, significantly reducing the material defect density and improving the conversion efficiency and long-term reliability of the device.
[0032] Example: Example 1: A method for fabricating a virtual substrate for InP-based narrow-bandgap thermophotovoltaic cells is disclosed, using a 2600G3 MOCVD (Metal-Organic Chemical Vapor Deposition) system manufactured by AXITRON GmbH, Germany. The substrate is an n-InP substrate with a thickness of 300–350 μm, doped with an S source at a concentration of 0.8E18–8E18 cm⁻¹. -3 ,0 o Chamfering. The MO sources used were TMGa and TMIn, the doping source was SiH4, and the specialty gases used were AsH3 and PH3.
[0033] The specific steps are as follows: A. PH3 is introduced into the MOCVD reaction chamber to raise the temperature of the reaction chamber to 650°C. o C. Deposit an n-InP buffer layer using TMI and TMGa as the source gases, AsH3 and PH3 as the special gases, with a deposition thickness of 0.3 μm. The doping source is SiH4, and the doping concentration is ≥1×10⁻⁶. 18 cm -3 .
[0034] B. On the InP buffer layer at 650 o C-temperature deposition of n-InAs 0.18 P 0.82 Buffer layer 1, n-InAs 0.18 P 0.82 The thickness of buffer layer 1 is 0.25 μm, the doping source is SiH4, and the doping concentration is 1 to 2 × 10⁻⁶. 18 cm -3 .
[0035] C. In n-InAs 0.18 P 0.82 Buffer layer 1 at 650 o C-temperature deposition of n-InAs 0.26 P 0.74 Buffer layer 2, n-InAs 0.26 P 0.74 The thickness of buffer layer 2 is 0.25 μm, the doping source is SiH4, and the doping concentration is 1 to 2 × 10⁻⁶. 18 cm -3 .
[0036] D. In n-InAs 0.26 P 0.74 Buffer layer 2 at 650 o C-temperature deposition of n-InAs 0.33 P 0.67Buffer layer 3, n-InAs 0.33 P 0.67 The thickness of buffer layer 3 is 0.25 μm, the doping source is SiH4, and the doping concentration is 1 to 2 × 10⁻⁶. 18 cm -3 .
[0037] E. In n-InAs 0.33 P 0.67 Buffer layer 3 at 650 o C-temperature deposition of n-InAs 0.42 P 0.58 Buffer layer 4, n-InAs 0.42 P 0.58 The thickness of buffer layer 4 is 0.25 μm, the doping source is SiH4, and the doping concentration is 1 to 2 × 10⁻⁶. 18 cm -3 .
[0038] F, in n-InAs 0.42 P 0.58 Buffer layer 4 at 650 o C-temperature deposition of n-InAs 0.49 P 0.51 Buffer layer 5, n-InAs 0.49 P 0.51 The thickness of buffer layer 5 is 0.25 μm, the doping source is SiH4, and the doping concentration is 1 to 2 × 10⁻⁶. 18 cm -3 .
[0039] G, in n-InAs 0.49 P 0.51 Buffer layer 5 at 650 o C-temperature deposition of n-InAs 0.56 P 0.44 Buffer layer 6, n-InAs 0.56 P 0.44 The thickness of buffer layer 6 is 0.25 μm, the doping source is SiH4, and the doping concentration is 1 to 2 × 10⁻⁶. 18 cm -3 .
[0040] H, in n-InAs 0.56 P 0.44 Buffer layer 6 at 650 o C-temperature deposition of n-InAs 0.64 P 0.36 Component overpass layer, n-InAs 0.64 P 0.36 The thickness of the overpass layer is 1 μm, the doping source is SiH4, and the doping concentration is 1 × 10⁻⁶. 18 cm -3 .
[0041] I. In n-InAs 0.64 P 0.36 The component overpass layer is at 650 o C-temperature deposition of n-InAs 0.60 P 0.40 Lattice hardened layer, n-InAs 0.60 P 0.40 The thickness of the lattice-hardened layer is 0.5 μm, the doping source is SiH4, and the doping concentration is ≥1×10⁻⁶. 18 cm -3 .
[0042] To observe the relaxation of epitaxial materials, their reciprocal space diffraction pattern (RSM) is as follows: Figure 2 As shown. Figure 2 (a) is the RSM plot of the InAsP epitaxial material in the symmetric (004) plane. The results show that the peak intensity of the reciprocal lattice point in the symmetric (004) plane is almost located at the center of the reciprocal lattice point of the substrate (along the vertical line Q=0). This result indicates that the tilt of the epitaxial layer to the substrate is very small. Figure 2 (b) shows the RSM plot of the InGaAs epitaxial material in the asymmetric (115) plane. The results show that the InGaAs epitaxial material in the asymmetric (115) plane... 0.81 Ga 0.19 The reciprocal lattice points of the As layer are precisely aligned on the line extending to the reciprocal lattice points of the substrate and Q = 0. This result indicates that the InGaAs low-temperature thermophotovoltaic cell material has achieved an equilibrium cubic structure with almost no distortion caused by lattice mismatch strain, and the stress caused by lattice mismatch within the buffer layer is completely released. The intensity contour lines corresponding to the compositionally stepped buffer layer and the active layer form an angle of approximately 18° with the intensity contour lines of the substrate's reciprocal lattice, indicating that the epitaxial material is almost completely relaxed.
[0043] from Figure 2 The RSM test results show that by adjusting the component ratio between As and P elements, InAs 0.176 P 0.824 InAs 0.260 P 0.740 InAs 0.334 P 0.666 InAs 0.418 P 0.582 InAs 0.488 P 0.512 InAs 0.556 P 0.444 InAs 0.638 P 0.362 InAs 0.615 P 0.385 and In 0.81 Ga0.19 The relaxation degree of the As layer also changes accordingly. Furthermore, to more clearly compare the effects of different component ratios on different parameters, the calculated values for strain relaxation, parallel strain, and layer tilt for each layer are listed in Table 1. The results show that the 2.07% total mismatch strain was effectively relaxed. The components in Example 1 of this invention are the rounded results from the components in Table 1.
[0044] Table 1. Material properties of InGaAs TPV cells
[0045] To further demonstrate the uniformity of the stepped gradient buffer layer and that the stepped gradient structure is consistent with the expected design, InP / InAs 0.18 P 0.82 / InAs 0.26 P 0.74 / InAs 0.33 P 0.67 / InAs 0.42 P 0.58 / InAs 0.49 P 0.51 / InAs 0.56 P 0.44 Cross-sectional transmission electron microscope image of the component step-gradient buffer layer structure as shown below Figure 3 As shown. Figure 3 (a) is a TEM image of the entire battery structure. Figure 3 (b) is a TEM image of the buffer layer cross-section of the large mismatch InGaAs epitaxial structure. According to... Figure 3 (b) It can be seen that a six-layer stepped buffer layer is deposited on the surface of the InP substrate. At the stepped buffer layer interface, mismatch dislocation lines are generated due to dislocation accumulation. Similarly, it can be observed that dislocations in the sample buffer layer are mainly concentrated near the stepped layer on the substrate side (lower side), forming closed dislocation loops or disappearing at the interface. Notably, the dislocations do not extend to the upper step, and no through dislocations are generated within the composition overshoot layer. This result demonstrates that the stepped interface of the buffer layer can effectively suppress the formation of through dislocations, thereby further improving the quality of epitaxial material growth.
[0046] To further analyze the quality of the epitaxial material in the samples, the surface morphology was characterized using AFM (Artificial Motion Surface Analysis). The results are as follows: Figure 4 As shown, the root mean square roughness of the material surface is 3.85 nm, which proves that the virtual substrate and epitaxial material obtained by this substrate have low roughness and good quality of epitaxial material layer. It is expected that its further application in thermophotovoltaic cell devices will result in higher efficiency.
[0047] The present invention has been described in detail above with reference to specific embodiments and exemplary examples; however, these descriptions should not be construed as limiting the present invention. Those skilled in the art will understand that various equivalent substitutions, modifications, or improvements can be made to the technical solutions and embodiments of the present invention without departing from the spirit and scope of the invention, and all such modifications and improvements fall within the scope of the present invention. The scope of protection of the present invention is defined by the appended claims.
[0048] The contents not described in detail in this specification are common knowledge to those skilled in the art.
Claims
1. A method for fabricating a virtual substrate structure for a narrow bandgap thermophotovoltaic cell, characterized in that, include: An n-InP buffer layer and an m-layer of n-InAs are sequentially epitaxially grown on an n-InP substrate. x P y Buffer layer, component overshoot layer and lattice hardening layer; m-layer n-InAs x P y In the buffer layer, m = 4~8, x = 0.1~0.6, y = 0.4~0.9; Following the order from the inside out, each layer of n-InAs x P y The component ratio of As / P in the buffer layer gradually increases; Component overpass layer and n-InAs x P y The buffer layers contain the same elements, and the As content in the overshoot layer is higher than that in the m-th layer (n-InAs). x P y The As content in the buffer layer and the P content in the overshoot layer are lower than those in the m-th layer n-InAs. x P y The P content of the buffer layer; the thickness of the overshoot layer is greater than that of each n-InAs layer. x P y Buffer layer; The lattice constant of the lattice-hardened layer is consistent with that of the battery base region and other functional layer materials.
2. The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to claim 1, characterized in that, The As content in the lattice-hardened layer is higher than that in the m-th layer n-InAs. x P y The As content in the buffer layer is lower than that in the composition overshoot layer, and the P content in the lattice hardening layer is lower than that in the m-th n-InAs layer. x P y The P content of the buffer layer is higher than that of the component overshoot layer.
3. The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to claim 1, characterized in that, The lattice constant of the over-stretched crystal is 0.1 to 0.2% higher than that of the lattice-hardened crystal.
4. The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to claim 1, characterized in that, The thickness of the n-InP substrate is 300~350 μm; The thickness of the n-InP buffer layer is 0.2~0.4μm; m-layer n-InAs x P y The thickness of the buffer layers is equal, ranging from 0.2 to 0.3 μm. The thickness of the over-penetration layer is 0.5~1.5μm; The thickness of the lattice hardening layer is 0.3~0.7μm.
5. The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to claim 1, characterized in that, m-layer n-InAs x P y In the buffer layer, m=6, 6 layers of n-InAs x P y The buffer layers are as follows: n-InAs 0.18 P 0.82 Buffer layer, n-InAs 0.26 P 0.74 Buffer layer, n-InAs 0.33 P 0.67 Buffer layer, n-InAs 0.42 P 0.58 Buffer layer, n-InAs 0.49 P 0.51 Buffer layer, n-InAs 0.56 P 0.44 Buffer layer.
6. The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to claim 1, characterized in that, The n-InP substrate is doped with S atoms at a concentration of 0.8-8 × 10⁻⁶. 18 cm -3 ; The n-InP buffer layer is doped with Si atoms, and the doping concentration is ≥1×10⁻⁶. 18 cm -3 ; m-layer n-InAs x P y The buffer layer is doped with Si atoms, with a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 ; The overshoot layer is composed of Si atoms with a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 ; The doped atoms of the lattice hardening layer are Si, with a doping concentration ≥ 1 × 10⁻⁶. 18 cm -3 .
7. The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to claim 1, characterized in that, An n-InP buffer layer and an m-layer n-InAs layer were sequentially epitaxially grown on an n-InP substrate using a metal-organic chemical vapor deposition method. x P y The layers consist of a buffer layer, a composition overshoot layer, and a lattice hardening layer, each deposited at a temperature of 630-670°C. o C.
8. The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to claim 7, characterized in that, The doping source for each layer during deposition was SiH4, the special gases used were AsH3 and PH3, and the MO source used was TMGa and TMIn.
9. A virtual substrate structure for a narrow bandgap thermophotovoltaic cell, characterized in that, The method for fabricating a narrow bandgap thermophotovoltaic cell virtual substrate structure according to any one of claims 1-8 includes an n-InP substrate, an n-InP buffer layer, and an m-layer n-InAs layer arranged sequentially. x P y Buffer layer, component overshoot layer and lattice hardening layer.
10. A virtual substrate structure for a narrow bandgap thermophotovoltaic cell according to claim 9, characterized in that, The virtual substrate structure for narrow bandgap thermophotovoltaic cells has a relaxation degree of >95% in each layer, a root mean square surface roughness of <5nm, and no penetrating dislocations extending to the surface.