Gallium nitride device and preparation method thereof

Abstract

The invention discloses a gallium nitride device and a preparation method thereof, and relates to the technical field of semiconductor devices. The gallium nitride device comprises a sapphire substrate, an AlN nucleating layer, an N-type AlGaN buffer layer, an intrinsic AlGaN active layer, an electron blocking layer and a P-type AlGaN layer which are sequentially stacked from bottom to top, a gradient doping edge stop layer is arranged on the edge of the P-type AlGaN layer, and meanwhile a bottom electrode in ohmic contact with the N-type AlGaN buffer layer and a top electrode in ohmic contact with the P-type AlGaN layer are formed respectively. According to the invention, by optimizing the structural design of the substrate, the nucleating layer, the buffer layer, the active layer, the electron blocking layer, the edge termination layer and the electrode, and combining with the seagull optimization algorithm to accurately regulate and control the preparation process parameters, the comprehensive improvement of the light extraction efficiency, the carrier recombination efficiency, the breakdown voltage and the long-term working stability is realized; and the photoelectric property and the reliability of the gallium nitride device are obviously enhanced.

Description

Technical Field

[0001] This invention relates to the field of semiconductor device technology, specifically to a gallium nitride device and its fabrication method. Background Technology

[0002] Gallium nitride (GaN)-based semiconductor devices, as typical representatives of wide-bandgap semiconductor materials, possess characteristics such as large bandgap, high electron mobility, high temperature resistance, radiation resistance, and excellent chemical stability, demonstrating enormous application potential in fields such as ultraviolet light detection, high-frequency power devices, and optoelectronic devices. Among them, gallium nitride-based photodiodes (such as PIN or Schottky structures) have become core devices for ultraviolet light detection by achieving efficient electron-hole pair generation and collection through epitaxial heterostructure design, and are widely used in key areas such as environmental monitoring, flame detection, and biomedicine.

[0003] In the performance metrics of gallium nitride-based photodiodes, light extraction efficiency is one of the key indicators limiting the improvement of external quantum efficiency (EQE). In existing technologies, the device structure typically employs a simple stacking of a planar substrate and an epitaxial layer. This results in light propagation within the device being easily limited by total internal reflection caused by high refractive index differences (such as between a sapphire substrate and an epitaxial layer). A large number of endogenous photons are trapped inside the device and cannot be effectively emitted, causing severe photon loss.

[0004] Therefore, overcoming the limitations of total internal reflection, enhancing light scattering efficiency, and optimizing carrier transport and stress release have become key technical challenges in improving the light extraction efficiency and overall performance of gallium nitride (GaN) devices. This invention aims to provide a GaN device with a highly efficient light scattering structure and optimized epitaxial layer design, as well as its fabrication method, to significantly improve the device's light extraction efficiency and operational stability. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art, provide a gallium nitride device and its fabrication method, solve the problem of total internal reflection of light at the device-air interface, and improve the device breakdown voltage and long-term reliability, thereby making up for the defects of the prior art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A gallium nitride device includes, from bottom to top, a sapphire substrate, an AlN nucleation layer, an N-type AlGaN buffer layer, an intrinsic AlGaN active layer, an electron blocking layer, and a P-type AlGaN layer. A gradient-doped edge termination layer is provided at the edge of the P-type AlGaN layer. A bottom electrode that is in ohmic contact with the N-type AlGaN buffer layer and a top electrode that is in ohmic contact with the P-type AlGaN layer are also formed.

[0008] Furthermore, the surface of the sapphire substrate is patterned to form periodic micro / nano structures to enhance ultraviolet light scattering and light absorption efficiency; the AlN nucleation layer has multiple pores, forming a light scattering structure with a pore density of 10. 9 Up to 10 12 pcs / cm 2 The pore size is 50 to 500 nm.

[0009] Furthermore, the Al composition in the N-type AlGaN buffer layer gradually increases along the growth direction, and its gradient distribution function of Al composition is:

[0010]

[0011] Where Al(z) represents the Al composition; z represents the z-direction coordinate value on the N-type AlGaN buffer layer; x0 represents the initial Al composition value, which can be taken as 0.1 in practical applications; represents the change in composition, which can be taken as 0.3 in practical applications; L represents the thickness of the N-type AlGaN buffer layer; e represents the natural constant; k represents the gradient adjustment factor.

[0012] Furthermore, the intrinsic AlGaN active layer comprises multiple pairs of indium-component gradient subwells formed by alternating stacking of InGaN and GaN. The band gradient thickness of the indium-component gradient subwells near the N-type AlGaN buffer layer is 0.5 nm, the band maintenance thickness in the middle portion of the indium-component gradient subwells is 1.1 nm, and the band gradient thickness of the indium-component gradient subwells near the P-type AlGaN layer is 1.3 nm.

[0013] Furthermore, a current flow control window is formed in the electron blocking layer; the formula for calculating the average ion implantation depth in the electron blocking layer is as follows:

[0014]

[0015] Among them, R p The average depth of implanted ions is represented by E, and the ion energy is represented by S. n Indicates nuclear deterrence capability, S e represents the electron blocking ability, and ln represents the natural logarithm function;

[0016] The ion implantation dose was determined through Monte Carlo simulation optimization, with the optimization objective being to satisfy the following vertical electric field distribution:

[0017]

[0018] Where Y represents the electron blocking layer thickness, E(y) represents the electric field distribution function, y represents the z-axis coordinate value on the electron blocking layer, d represents the differential sign, and Vbr η represents the breakdown voltage, and η represents the electron utilization factor.

[0019] Furthermore, the gradient-doped edge termination layer uses N + →N - Ion gradient doping distribution forms space charge compensation to suppress edge electric field peaks;

[0020] The gradient doping distribution function is as follows:

[0021]

[0022] Where N(x) represents the net doping concentration, x represents the coordinate value along the z-direction on the gradient doped edge termination layer, and N D0 N represents + Peak ion doping concentration, N A0 N represents - The peak value of ion doping concentration, exp represents an exponential function with base e as the natural constant, and x0 represents N. + The location of the peak ion doping concentration, x1 represents N - The location of the peak ion doping concentration, where σ represents the Gaussian width.

[0023] Furthermore, the top electrode is an asymmetric metal electrode, and the top electrode is etched into a circular or rectangular finger structure with a spacing of 3 μm between the finger structures.

[0024] On the other hand, the present invention also provides a method for fabricating gallium nitride devices as described above, comprising:

[0025] S1: Construct a simulation model of gallium nitride devices;

[0026] S2: Determine the breakdown voltage of gallium nitride devices under various structural parameters through simulation models;

[0027] S3: Determine the optimal structural parameters of the gallium nitride device based on the breakdown voltage of the gallium nitride device under various structural parameters;

[0028] S4: According to the optimal structural parameters, an AlN nucleation layer is deposited on the sapphire substrate by radio frequency sputtering;

[0029] S5: By selectively depositing argon ions in certain areas, multiple pores are generated on the AlN nucleation layer, and these pores form a light scattering structure.

[0030] S6: An N-type AlGaN buffer layer is deposited on the AlN nucleation layer using chemical vapor deposition technology;

[0031] S7: Using chemical vapor deposition technology, InGaN and GaN are alternately stacked and deposited on an N-type AlGaN buffer layer to form an intrinsic AlGaN active layer with an indium composition gradient sub-well;

[0032] S8: An electron blocking layer is deposited on the intrinsic AlGaN active layer using chemical vapor deposition technology;

[0033] S9: A P-type AlGaN layer is deposited on the electron blocking layer using chemical vapor deposition technology;

[0034] S10: N ions are doped into the edge of the P-type AlGaN layer to form a gradient-doped edge termination layer with a gradually changing charge distribution.

[0035] S11: Make an ohmic contact between the N-type AlGaN buffer layer and the bottom electrode, and an ohmic contact between the P-type AlGaN layer and the top electrode.

[0036] Specifically, the structural parameters mentioned in step S2 include the thickness of the N-type AlGaN buffer layer, the Al composition distribution in the N-type AlGaN buffer layer, the thickness of the intrinsic AlGaN active layer, the thickness of the electron blocking layer, the ion implantation dose and depth in the electron blocking layer, the thickness of the P-type AlGaN layer, the thickness of the gradient-doped edge termination layer, and the N ion doping concentration in the gradient-doped edge termination layer.

[0037] Specifically, in step S3, the optimal structural parameters of the gallium nitride device are determined using the Seagull optimization algorithm.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] (1) In this invention, the periodic micro-nano structure on the surface of the sapphire substrate is introduced into the light scattering interface through patterning, which makes it easier for incident or endogenous light to break the total internal reflection limitation, expand the emission angle coverage range, and significantly improve the light extraction efficiency.

[0040] (2) The present invention introduces multiple pores on the AlN nucleation layer to form a light scattering structure, which helps to enhance the scattering effect of ultraviolet light inside the device, effectively breaks the total reflection limitation caused by high refractive index difference, and significantly improves the light emission efficiency. At the same time, this porous structure can also release stress and reduce lattice defect density during epitaxial growth, thereby improving the crystal quality of the subsequent AlGaN layer and the overall optoelectronic performance of the device.

[0041] (3) This invention achieves an optimized bandgap structure by employing a gradually increasing Al composition along the growth direction in the N-type AlGaN buffer layer. As the Al composition gradually increases, the band gap of the material widens along the growth direction, which helps electrons migrate to the active region more efficiently, while avoiding nonradiative recombination of charge carriers in the buffer layer. By adjusting the initial Al composition value x0, the composition change Δx, and the gradient adjustment factor k, the steepness of the bandgap can be precisely controlled, further optimizing the electron injection efficiency and thus enhancing the internal quantum efficiency and overall performance of the optoelectronic device.

[0042] (4) The intrinsic AlGaN active layer employs multiple pairs of InGaN and GaN stacked alternately to form indium-component gradient quantum wells. By precisely designing the band thickness in different regions of the quantum well (0.5 nm near the N-type buffer layer, 1.1 nm in the middle maintenance section, and 1.3 nm near the P-type layer), the gradient optimization of the band structure is achieved. This asymmetric trapezoidal band design ensures efficient injection of electrons from the N-type layer into the quantum well while effectively limiting the diffusion of holes in the quantum well, thus increasing the probability of electron-hole recombination. At the same time, the gradient distribution of indium composition reduces stress concentration inside the quantum well, lowers the defect density caused by lattice mismatch, and further improves the luminescence efficiency and stability of the active layer.

[0043] (5) The electron blocking layer forms a current flow control window through ion implantation, and the average ion implantation depth is optimized based on a formula. The implantation dose is determined by Monte Carlo simulation, thus achieving precise control of the vertical electric field distribution. This design can effectively prevent electrons from leaking from the active layer to the P-region, ensuring that electrons concentrate in the recombination region and combine with holes to emit light. At the same time, it avoids the local concentration of the electric field in the blocking layer, which significantly improves the breakdown voltage (breakdown voltage reaches 1.2kV) and electron utilization factor (reaches 0.95), thereby improving the energy conversion efficiency and long-term operational reliability of the device.

[0044] (6) The gradient doped edge termination layer uses N + →N - The ion gradient doping distribution achieves a smooth transition in doping concentration through a Gaussian function. This design creates a space charge compensation region at the edge of the p-type AlGaN layer, effectively suppressing edge electric field peaks and avoiding breakdown risks or local current leakage caused by excessive electric fields. Compared to traditional planar structures, this edge termination layer allows the electric field to gradually transition at the device edge, reducing lattice damage caused by abrupt electric field changes and significantly improving the device's breakdown voltage and long-term operational stability.

[0045] (7) The top electrode adopts an asymmetric metal electrode and is etched into a circular or rectangular finger structure. The spacing between the finger structures is precisely controlled at 3 μm, which can effectively optimize the current distribution in the P-type AlGaN layer, reduce the concentration of current density, and avoid local overheating or electrical breakdown. At the same time, the asymmetric structure can adjust the electrode contact area according to the internal carrier transport characteristics of the device, further improving the uniformity of current injection, thereby improving the luminescence consistency and reliability of the device.

[0046] (8) In the process of fabricating gallium nitride devices, this invention uses the Seagull optimization algorithm to determine the optimal structural parameters. Using breakdown voltage as the fitness function, and combining global and local search strategies, it achieves coordinated optimization of multi-dimensional structural parameters such as the thickness of the N-type AlGaN buffer layer, Al composition distribution, and ion implantation parameters of the electron blocking layer. This algorithm can efficiently explore the solution space, avoid getting trapped in local optima, and ensure that the device achieves optimal comprehensive performance such as light extraction efficiency and carrier recombination efficiency while maintaining a high breakdown voltage, thereby improving the consistency of mass production of the device. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the structure of the gallium nitride device provided by the present invention;

[0048] Figure 2 A schematic diagram of the band structure variation of the indium composition gradient sub-well provided by the present invention;

[0049] Figure 3 This is a schematic diagram of the process for fabricating gallium nitride devices in this invention.

[0050] The corresponding names of the reference numerals in the figure are: 1-sapphire substrate, 2-AlN nucleation layer, 3-N-type AlGaN buffer layer, 4-intrinsic AlGaN active layer, 5-electron blocking layer, 6-P-type AlGaN layer, 7-gradient doped edge termination layer, 8-top electrode, 9-bottom electrode. Detailed Implementation

[0051] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following examples. In the following description, descriptions of well-known structures and techniques are omitted to avoid unnecessarily obscuring the concepts disclosed in the present invention.

[0052] The present invention provides a structure for a gallium nitride device, such as... Figure 1As shown, from bottom to top, it includes: a sapphire substrate 1, an AlN nucleation layer 2, an N-type AlGaN buffer layer 3, an intrinsic AlGaN active layer 4, an electron blocking layer 5, and a P-type AlGaN layer 6. A gradient-doped edge termination layer 7 is provided at the edge of the P-type AlGaN layer 6. At the same time, a bottom electrode 9 is formed in ohmic contact with the N-type AlGaN buffer layer 3, and a top electrode 8 is formed in ohmic contact with the P-type AlGaN layer 6. The specific arrangement of each layer structure is as follows:

[0053] The sapphire substrate 1 serves as the basic carrier for the entire device epitaxial structure. Its surface is patterned to form a periodic micro-nano structure. This micro-nano structure can construct a light scattering interface. When ultraviolet light from inside the device is incident on this interface, it can change the direction of light propagation, break the total internal reflection limitation between the semiconductor and air interface, expand the range of light emission angle, and enhance the absorption efficiency of ultraviolet light. In addition, the periodic micro-nano structure can also increase the contact surface area between the substrate and the upper AlN nucleation layer 2, and improve the bonding strength and long-term stability of the heterostructure.

[0054] AlN nucleation layer 2 is disposed on sapphire substrate 1. AlN nucleation layer 2 has multiple pores, forming a light scattering structure. The pore density ranges from 10⁻⁶. 9 Up to 10 12 pcs / cm 2 The density of pores directly affects the light scattering effect and the stress release between layers. Excessively high pore density may lead to structural instability or lattice defects, while too low a density may fail to effectively improve the scattering effect. The density range in this invention ensures effective light scattering by the pores and avoids strong impacts on the substrate, especially in controlling stress and improving scattering efficiency. The pore size ranges from 50 to 500 nm. The pore size distribution has a significant impact on the light scattering effect; generally, a larger pore size results in a wider scattering angle and a stronger effect, but it may also lead to stress concentration. Furthermore, a log-normal distribution can be used to control the pore size range, making the scattering structure more diverse and covering different wavelengths of light. The AlN nucleation layer 2, as an important transition layer to alleviate lattice mismatch between sapphire and AlGaN materials, introduces multiple pore structures on its surface, forming a multi-scale light scattering interface. These pore structures act as scattering centers at the bottom of the device, effectively increasing the complexity of the internal photon emission path, further suppressing total internal reflection loss, and enhancing the probability of photons escaping from the device, thus improving the external quantum efficiency (EQE). Simultaneously, the pore structures also help release stress during the epitaxial process, reduce dislocation density, and improve the crystal quality of subsequent epitaxial layers.

[0055] The N-type AlGaN buffer layer 3 is disposed on the AlN nucleation layer 2, mainly used to regulate carrier transport and guide electron injection. By optimizing its doping and composition distribution, it can effectively balance carrier concentration and lattice stability, providing a stable electron supply channel for the active region above.

[0056] The intrinsic AlGaN active layer 4 is disposed on the N-type AlGaN buffer layer 3. It is the main region for electron-hole recombination and plays a decisive role in the emission wavelength, efficiency, and response sensitivity of the device. By precisely controlling its thickness and composition, a quantum well structure or an enhanced bandgap modulation structure can be formed, thereby increasing the radiative recombination probability and enhancing the internal quantum efficiency (IQE).

[0057] An electron blocking layer 5 is disposed on the intrinsic AlGaN active layer 4. Its function is to prevent excessive electrons from leaking from the active layer to the P-region, ensuring that electrons are concentrated in the recombination region and combine with holes to emit light, thereby improving energy utilization and device stability.

[0058] The p-type AlGaN layer 6 is disposed on the electron blocking layer 5. Its main function is to provide a source of holes to participate in the recombination process, while forming an ohmic contact with the top electrode to optimize the carrier injection efficiency. Since p-type GaN or AlGaN materials have high electrical resistance, high conductivity needs to be achieved through Mg doping or controlled growth process.

[0059] The gradient-doped edge termination layer 7 is disposed at the edge of the P-type AlGaN layer 6. It can form a space charge compensation region, significantly reduce the electric field concentration effect at the edge, effectively suppress the electrical breakdown phenomenon, and improve the breakdown voltage and long-term operating reliability of the device.

[0060] In this invention, the periodic micro / nano structures formed by patterning the sapphire substrate surface create a light scattering interface, making it easier for incident light or light generated inside the device to overcome the total internal reflection limitation, expanding the emission angle coverage range, and significantly improving the light extraction efficiency. Multiple pores introduced on the AlN nucleation layer constitute a light scattering structure, which helps enhance the scattering effect of ultraviolet light inside the device, effectively breaking the total internal reflection limitation caused by the high refractive index difference and significantly improving the light extraction efficiency. Simultaneously, this porous structure can also release stress and reduce the lattice defect density during epitaxial growth, thereby improving the crystal quality of the subsequent AlGaN layer and the overall optoelectronic performance of the device.

[0061] In one embodiment, the Al composition in the N-type AlGaN buffer layer 3 gradually increases along the growth direction to achieve band gradient modulation and reduce carrier recombination.

[0062] Optionally, the gradient distribution function of the Al component is specifically:

[0063]

[0064] Where Al(z) represents the Al composition; z represents the z-direction coordinate value on the N-type AlGaN buffer layer; x0 represents the initial Al composition value, which can be taken as 0.1 in practical applications; represents the change in composition, which can be taken as 0.3 in practical applications; L represents the thickness of the N-type AlGaN buffer layer; e represents the natural constant; k represents the gradient adjustment factor.

[0065] Those skilled in the art can set the size of the gradient adjustment factor according to the actual situation; this embodiment does not impose any limitations.

[0066] Specifically, the gradient adjustment factor determines the rate of change of Al composition and the steepness of the gradient in the AlGaN buffer layer, controlling the rate of change of Al composition along the z-direction, thereby regulating the bandgap change rate of the material. By reasonably adjusting the value of the gradient adjustment factor, carrier migration paths can be optimized, recombination losses can be reduced, and the overall efficiency and stability of the device can be improved.

[0067] In this embodiment, the Al composition is gradually distributed along the growth direction in the N-type AlGaN buffer layer, enabling band gradient modulation and effectively reducing carrier recombination. Specifically, as the Al composition gradually increases, the band gap of the material also widens, forming a band gradient structure. This structure facilitates the migration of electrons and holes to the active region while preventing nonradiative recombination of carriers in the buffer layer. By adjusting the amount of change in Al composition and the gradient factor, the electron injection efficiency can be optimized, enhancing the performance of optoelectronic devices, especially in improving internal quantum efficiency and reducing carrier loss.

[0068] In one embodiment, the intrinsic AlGaN active layer 4 includes multiple pairs of indium composition gradient quantum wells formed by alternating stacking of InGaN and GaN, preferably 5 pairs. Figure 2 The diagram illustrates the band structure of an indium-component gradient quantum well according to this embodiment, which exhibits an asymmetric trapezoidal shape. Specifically, the band transition thickness near the N-type AlGaN buffer layer is 0.5 nm, the band maintenance thickness in the middle portion of the quantum well is 1.1 nm, and the band transition thickness near the P-type AlGaN layer is 1.3 nm.

[0069] The indium-based gradient quantum well features different bandgap thicknesses in different regions, a design that effectively modulates carrier distribution and recombination. A 0.5 nm thin bandgap region near the N-type AlGaN buffer layer helps improve electron injection efficiency, ensuring easier electron entry into the quantum well. A thicker 1.1 nm bandgap in the middle region maintains a high electron-hole recombination probability, enhancing light emission efficiency. A thicker 1.3 nm bandgap region near the P-type AlGaN layer helps mitigate hole loss, ensuring sufficient hole recombination within the quantum well. Overall, this design, by optimizing the bandgap structure in different regions, improves quantum efficiency, enhances luminescence intensity, and improves device performance and stability.

[0070] In one embodiment, a current flow control window is formed in the electron blocking layer 5.

[0071] Forming a current flow control window in the electron blocking layer 5 has the main advantage of precisely controlling the electron flow path, ensuring that electrons flow only within a predetermined area and avoiding unnecessary leakage or diffusion. This design effectively prevents electrons from penetrating into areas where they should not emit light, thereby improving the recombination efficiency of electrons and holes and enhancing light generation. Furthermore, the current flow control window can optimize the electric field distribution, reduce the risk of adverse current concentration or breakdown, and thus improve the stability and reliability of the device.

[0072] The formula for calculating the average ion implantation depth in electron blocking layer 5 is as follows:

[0073]

[0074] Among them, R p The average depth of implanted ions is represented by E, and the ion energy is represented by S. n Indicates nuclear deterrence capability, S e represents the electron blocking ability, and ln represents the natural logarithm function.

[0075] Nuclear stopping power describes the interaction between ions and atomic nuclei in a target material, primarily through elastic collisions that transfer ion energy to atoms within the material. Electron stopping power, on the other hand, describes the interaction between ions and electrons in the material, mainly through inelastic collisions, i.e., energy loss due to electron excitation. Specifically, Bethe's stopping theory can be used to quantify nuclear and electron stopping power.

[0076] This implementation method is based on ion energy E and nuclear stopping power S. n and electron blocking capability S eBy adjusting the ion implantation depth, precise current blocking and electric field modulation can be achieved. This design ensures uniform current distribution within the blocking layer, preventing excessive current concentration or uneven flow, reducing the risk of current leakage and electrical breakdown, and improving the device's operational stability and long-term reliability.

[0077] The ion implantation dose was determined through Monte Carlo simulation optimization, with the optimization objective being to satisfy the following condition for the vertical electric field distribution:

[0078]

[0079] Where Y represents the electron blocking layer thickness, E(y) represents the electric field distribution function, y represents the z-axis coordinate value on the electron blocking layer, d represents the differential sign, and V br η represents the breakdown voltage, and η represents the electron utilization factor.

[0080] By optimizing the ion implantation dose using Monte Carlo simulations and setting specific conditions for the vertical electric field distribution (ensuring a breakdown voltage of 1.2 kV and an electron utilization factor of 0.95) as the optimization objective, the electric field distribution in the electron blocking layer can be effectively controlled, resulting in a more uniform electric field distribution and preventing breakdown or localized overheating caused by electric field concentration. This optimization process ensures both a high breakdown voltage and maximizes electron utilization in the effective region, thereby improving the device's stability, reliability, and long-term performance.

[0081] In one embodiment, the gradient-doped edge termination layer 7 uses N + →N - Ion gradient doping distribution forms space charge compensation to suppress edge electric field peaks.

[0082] The gradient doping distribution function is as follows:

[0083]

[0084] Where N(x) represents the net doping concentration, x represents the coordinate value along the z-direction on the gradient doped edge termination layer, and N D0 N represents + Peak ion doping concentration, N A0 N represents - The peak value of ion doping concentration, exp represents an exponential function with base e as the natural constant, and x0 represents N. + The location of the peak ion doping concentration, x1 represents N - The location of the peak ion doping concentration, where σ represents the Gaussian width.

[0085] It should be noted that Gaussian doping structures can form smooth charge compensation regions at the edge of the device, effectively reducing abrupt changes in the electric field.

[0086] This embodiment employs N in the gradient doping edge termination layer. + →N - Ion gradient doping can create space charge compensation at the device edges, effectively suppressing the concentration of the edge electric field and thus avoiding the risk of breakdown or local current leakage caused by excessively strong electric fields. Specifically, this gradient doping distribution uses a Gaussian function to make N + and N - The smooth transition in ion concentration distribution allows the electric field to gradually transition at the edges, reducing the peak electric field and improving the device's breakdown voltage and operational stability. Furthermore, this design effectively enhances the device's long-term reliability, enabling stable operation under higher voltages and more demanding operating conditions.

[0087] In one embodiment, the top electrode 8 is an asymmetric metal electrode, and the top electrode is etched into a circular or rectangular finger structure with a spacing of 3 μm between the finger structures.

[0088] In this embodiment, using asymmetric metal electrodes and etching the top electrode into a circular or rectangular finger structure effectively optimizes current distribution and electric field uniformity. The 3μm spacing between the finger structures helps reduce the concentration of current density, avoiding localized overheating or electrical breakdown.

[0089] The gallium nitride devices (taking LEDs as an example) actually manufactured using the structure of this invention were tested. The results showed that the GaN-based LED with an epitaxially grown AlN nucleation layer had a light output power of 754 mW at an injection current of 600 mA, which was about 4.0% higher than that of the conventional sample. Under the same conditions, the reverse breakdown voltage of the LED with an epitaxially grown AlN nucleation layer was about 10V higher than that of the conventional sample. The peak wavelength of the spectrum of the LED with an epitaxially grown AlN nucleation layer red-shifted at low injection current, which indicates that the device has a large compressive strain.

[0090] The present invention also provides a method for fabricating the above-mentioned gallium nitride device, the process of which is shown in the figure below. Figure 3 As shown, it includes:

[0091] S1: Constructing a simulation model of a gallium nitride device:

[0092] The simulation model of gallium nitride devices is described using the Poisson equation, the continuity equations for electrons and holes, and the drift-diffusion current equations for electrons and holes.

[0093] The Poisson equation is used to calculate the potential distribution within a semiconductor device, and its expression is:

[0094]

[0095] Where ε represents the dielectric constant of the semiconductor, This represents gradient operation. Let N represent electric potential, q represent the elementary charge of an electron, p represent the hole density, n represent the electron density, and N represent the electron density. D N represents the donor concentration of ionization. A This indicates the concentration of ionized acceptors.

[0096] The continuity equations for electrons and holes describe the conservation of charge, as follows:

[0097]

[0098] Among them, J n J represents the electron current density. p R represents the current density of holes. n R represents the recombination rate of electrons. p G represents the recombination rate of holes. n G represents the rate of electron formation. p This represents the hole generation rate. t represents partial derivative operation, and t represents time.

[0099] The drift-diffusion current equations for electrons and holes are as follows:

[0100]

[0101] Where, μ n μ represents the mobility of electrons. p D represents the hole mobility. n D represents the diffusion coefficient of electrons. p Let E represent the hole diffusion coefficient, and E represent the electric field distribution.

[0102] S2: Determine the breakdown voltage of gallium nitride devices under various structural parameters using simulation models:

[0103] Breakdown voltage refers to the voltage at which, in electronic devices (such as semiconductor diodes, transistors, or other power electronic devices), the internal electric field strength becomes sufficient to cause the insulation properties of the material to fail, resulting in uncontrolled current flow through the device and electrical breakdown. It is one of the key indicators for evaluating device performance, representing the maximum voltage the device can withstand.

[0104] The structural parameters include: the thickness of the N-type AlGaN buffer layer, the Al composition distribution in the N-type AlGaN buffer layer, the thickness of the intrinsic AlGaN active layer, the thickness of the electron blocking layer, the ion implantation dose and depth in the electron blocking layer, the thickness of the P-type AlGaN layer, the thickness of the gradient-doped edge termination layer, and the N ion doping concentration in the gradient-doped edge termination layer.

[0105] S3: Determine the optimal structural parameters of the gallium nitride (GaN) device based on its breakdown voltage under various structural parameters. In practice, the Seagull optimization algorithm can be used to determine the optimal structural parameters of the GaN device.

[0106] Specifically, the breakdown voltage is used as the fitness function of the Seagull optimization algorithm.

[0107] Initialize seagull individuals. Each seagull individual represents a feasible set of structural parameters. Each seagull individual consists of multiple dimensional components, and each component represents a structural parameter.

[0108] During the global search phase, it is necessary to avoid collisions between individual seagulls and to encourage them to move towards the optimal individual:

[0109]

[0110]

[0111]

[0112] in, This represents the position of the i-th seagull after the global search phase in the t-th iteration. Let A represent the position of the i-th seagull after collision avoidance processing in the t-th iteration, and let A represent the control factor. This represents the position of the i-th seagull in the t-th iteration. Let B represent the displacement of the i-th seagull individual towards the optimal individual in the t-th iteration, and let B represent the search balance factor. This represents the optimal individual position at the t-th iteration.

[0113] In this invention, during the global search phase, by introducing anti-collision processing and a strategy of moving towards the optimal individual, the collisions and excessive aggregation of individual seagulls in the solution space can be effectively avoided, thereby maintaining the breadth and diversity of the search.

[0114] Specifically, the control factor is calculated as follows:

[0115]

[0116] Where t represents the current iteration number, T represents the maximum iteration number, and f c denoted by , where lg represents the linear decreasing frequency, and lg represents the logarithmic function to the base 10.

[0117] In this invention, the control factor A gradually decreases with increasing iterations. This helps maintain a large search range in the early stages of the optimization process, encouraging individual seagulls to explore the solution space more extensively, thus avoiding premature entrapment in local optima. As iterations progress, the control factor gradually decreases, making the search process more refined, focusing on the vicinity of the optimal solution, improving convergence, and accelerating the arrival at the optimal solution. This method balances exploration and development during the optimization process, effectively broadening the search while ensuring precise searching for the optimal solution.

[0118] The formula for calculating the search balance factor B is:

[0119]

[0120] Where r1 represents a random number between 0 and 1.

[0121] In this invention, the speed and range at which an individual approaches the optimal solution during the search process can be dynamically adjusted. The introduction of random numbers increases uncertainty, enabling the algorithm to maintain a certain balance between global and local searches, avoiding premature convergence. The balance factor B decreases as the control factor A decreases, thereby gradually reducing randomness in the search process, enhancing fine-grained search capabilities, and helping to accurately find the optimal solution. This method can maintain the breadth of the search while avoiding excessive fluctuations in the search process, improving the stability and convergence efficiency of the algorithm.

[0122] During the local search phase, a random number r2 is generated. Based on the random number r2, a choice is made in parallel between the spiral search strategy and the encirclement strategy, and displacement is performed in a spiral motion manner.

[0123]

[0124]

[0125]

[0126]

[0127]

[0128] in, Let represent the position of the i-th individual seagull after spiraling motion in the t-th iteration, x represent the spiral flight coefficient in the x-direction, y represent the spiral flight coefficient in the y-direction, z represent the spiral flight coefficient in the z-direction, r represent the radius of the spiral flight trajectory, θ represent a random number between 0 and 2π, u and v represent the spiral constants, and e represents the natural constant.

[0129] This invention enhances the diversity and flexibility of the search process by combining a spiral search strategy with an encirclement strategy during the local search phase. Random numbers are generated to determine which strategy to adopt. The spiral movement allows individual seagulls to explore effectively in different directions, forming a purposeful spiral trajectory in the search space, thus avoiding getting trapped in local optima. By randomly selecting between spiral search and encirclement strategies, the algorithm can adaptively adjust its search method based on the current search state. This enhances the local search capability, especially when approaching the optimal solution, allowing for more refined optimization and improving the algorithm's convergence speed and accuracy.

[0130] Perform mutation operations on each individual seagull:

[0131]

[0132] in, P represents the position of the i-th seagull individual after mutation in the t-th iteration. r Let ω represent a random individual and ω represent an adaptive scaling factor.

[0133] This invention introduces randomness and diversity into the search process by performing mutation operations, preventing the algorithm from getting trapped in local optima. It increases the ability of individuals to jump within the solution space, improves the flexibility of the search process, and enhances the algorithm's global search capability and convergence speed.

[0134] The formula for calculating the adaptive scaling factor ω is:

[0135]

[0136] Where, ω max ω represents the maximum scaling factor. min This represents the minimum scaling factor.

[0137] In the early stages of the optimization process, a larger scaling factor ω promotes a broader search, helping individual seagulls explore the solution space and avoid getting trapped in local optima. As the number of iterations increases, the scaling factor ω gradually decreases, reducing the mutation amplitude and making the search more refined, focusing on the neighborhood of the optimal solution. This dynamic adjustment helps maintain a balance between global and local search, improving the algorithm's search efficiency and convergence speed, while also enhancing its stability and adaptability.

[0138] Determine if the fitness value of the mutated position is greater than the fitness value of the original position. If it is greater, replace the original position with the mutated position. Otherwise, leave the original position unchanged.

[0139] Update the fitness values ​​of each individual seagull and the global best individual.

[0140] Determine if the current iteration count has reached the maximum iteration count. If it has, output the set of structural parameters represented by the seagull with the highest fitness; otherwise, return to continue iterating.

[0141] This invention utilizes the Seagull optimization algorithm to optimize the breakdown voltage under different structural parameters, efficiently finding the optimal combination of structural parameters for gallium nitride (GaN) devices. This method combines global and local search strategies to explore a broad solution space and precisely adjusts each structural parameter to maximize the breakdown voltage. The optimization process avoids getting trapped in local optima, ensuring the device achieves the highest breakdown voltage while maintaining stability. This improves the device's electrical performance, reliability, and long-term stability, providing a scientific and efficient optimization approach for GaN device design.

[0142] S4: According to the optimal structural parameters, an AlN nucleation layer is deposited on a sapphire substrate by radio frequency sputtering.

[0143] S5: By selectively depositing argon ions into certain areas, multiple pores are generated on the AlN nucleation layer, which form a light scattering structure.

[0144] S6: An N-type AlGaN buffer layer is deposited on the AlN nucleation layer using chemical vapor deposition technology.

[0145] S7: Using chemical vapor deposition technology, InGaN and GaN are alternately stacked and deposited on an N-type AlGaN buffer layer to form an intrinsic AlGaN active layer with an indium composition gradient quantum well.

[0146] S8: An electron blocking layer is deposited on the intrinsic AlGaN active layer using chemical vapor deposition technology.

[0147] S9: A P-type AlGaN layer is deposited on the electron blocking layer using chemical vapor deposition technology.

[0148] S10: N ions are doped into the edge of the P-type AlGaN layer to form a gradient-doped edge termination layer with a gradually changing charge distribution.

[0149] S11: Make an ohmic contact between the N-type AlGaN buffer layer and the bottom electrode, and an ohmic contact between the P-type AlGaN layer and the top electrode.

[0150] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the preferred embodiments, while those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0151] The above embodiments are merely one of the preferred embodiments of the present invention and should not be used to limit the scope of protection of the present invention. Any modifications or refinements made to the main design concept and spirit of the present invention that are not of substantial significance, but solve the same technical problem as the present invention, should be included within the scope of protection of the present invention.

Claims

1. A gallium nitride device, characterized in that, The structure consists of a sapphire substrate (1), an AlN nucleation layer (2), an N-type AlGaN buffer layer (3), an intrinsic AlGaN active layer (4), an electron blocking layer (5), and a P-type AlGaN layer (6) stacked sequentially from bottom to top. A gradient-doped edge termination layer (7) is provided at the edge of the P-type AlGaN layer (6). At the same time, a bottom electrode (9) is formed that is in ohmic contact with the N-type AlGaN buffer layer (3), and a top electrode (8) is formed that is in ohmic contact with the P-type AlGaN layer (6).

2. A gallium nitride device according to claim 1, characterized in that, The surface of the sapphire substrate (1) is patterned to form a periodic micro / nano structure to enhance ultraviolet light scattering and light absorption efficiency; the AlN nucleation layer (2) has multiple pores to form a light scattering structure with a pore density of 10. 9 Up to 10 12 pcs / cm 2 The pore size is 50 to 500 nm.

3. A gallium nitride device according to claim 1, characterized in that, The Al composition in the N-type AlGaN buffer layer (3) gradually increases along the growth direction, and its gradual distribution function of Al composition is: Where Al(z) represents the Al composition; z represents the z-direction coordinate value on the N-type AlGaN buffer layer; x0 represents the initial Al composition value, which can be taken as 0.1 in practical applications; represents the change in composition, which can be taken as 0.3 in practical applications; L represents the thickness of the N-type AlGaN buffer layer; e represents the natural constant; k represents the gradient adjustment factor.

4. A gallium nitride device according to claim 1, 2, or 3, characterized in that, The intrinsic AlGaN active layer (4) includes multiple pairs of InGaN and GaN stacked alternately to form indium composition gradient sub-wells. The band gradient thickness of the indium composition gradient sub-wells near the N-type AlGaN buffer layer is 0.5 nm, the band maintenance thickness of the middle part of the indium composition gradient sub-wells is 1.1 nm, and the band gradient thickness of the indium composition gradient sub-wells near the P-type AlGaN layer is 1.3 nm.

5. A gallium nitride device according to claim 4, characterized in that, A current flow control window is formed in the electron blocking layer (5); the calculation formula for the average ion implantation depth in the electron blocking layer (5) is as follows: Among them, R p The average depth of implanted ions is represented by E, and the ion energy is represented by S. n Indicates nuclear deterrence capability, S e represents the electron blocking ability, and ln represents the natural logarithm function; The ion implantation dose was determined through Monte Carlo simulation optimization, with the optimization objective being to satisfy the following vertical electric field distribution: Where Y represents the electron blocking layer thickness, E(y) represents the electric field distribution function, y represents the z-axis coordinate value on the electron blocking layer, d represents the differential sign, and V br η represents the breakdown voltage, and η represents the electron utilization factor.

6. A gallium nitride device according to claim 5, characterized in that, The gradient doped edge termination layer (7) uses N + →N - Ion gradient doping distribution forms space charge compensation to suppress edge electric field peaks; The gradient doping distribution function is as follows: Where N(x) represents the net doping concentration, x represents the coordinate value along the z-direction on the gradient doped edge termination layer, and N D0 N represents + Peak ion doping concentration, N A0 N represents - The peak value of ion doping concentration, exp represents an exponential function with base e as the natural constant, and x0 represents N. + The location of the peak ion doping concentration, x1 represents N - The location of the peak ion doping concentration, where σ represents the Gaussian width.

7. A gallium nitride device according to claim 6, characterized in that, The top electrode (8) is an asymmetric metal electrode, and the top electrode (8) is etched into a circular or rectangular finger structure with a spacing of 3 μm between the finger structures.

8. A method for fabricating a gallium nitride device, used to fabricate the gallium nitride device according to any one of claims 1 to 7, characterized in that, include: S1: Construct a simulation model of gallium nitride devices; S2: Determine the breakdown voltage of gallium nitride devices under various structural parameters through simulation models; S3: Determine the optimal structural parameters of the gallium nitride device based on the breakdown voltage of the gallium nitride device under various structural parameters; S4: According to the optimal structural parameters, an AlN nucleation layer is deposited on the sapphire substrate by radio frequency sputtering; S5: By selectively depositing argon ions in certain areas, multiple pores are generated on the AlN nucleation layer, and these pores form a light scattering structure. S6: An N-type AlGaN buffer layer is deposited on the AlN nucleation layer using chemical vapor deposition technology; S7: Using chemical vapor deposition technology, InGaN and GaN are alternately stacked and deposited on an N-type AlGaN buffer layer to form an intrinsic AlGaN active layer with an indium composition gradient sub-well; S8: An electron blocking layer is deposited on the intrinsic AlGaN active layer using chemical vapor deposition technology; S9: A P-type AlGaN layer is deposited on the electron blocking layer using chemical vapor deposition technology; S10: N ions are doped into the edge of the P-type AlGaN layer to form a gradient-doped edge termination layer with a gradually changing charge distribution. S11: Make an ohmic contact between the N-type AlGaN buffer layer and the bottom electrode, and an ohmic contact between the P-type AlGaN layer and the top electrode.

9. The method for preparing gallium nitride devices according to claim 8, characterized in that, The structural parameters mentioned in step S2 include the thickness of the N-type AlGaN buffer layer, the Al composition distribution in the N-type AlGaN buffer layer, the thickness of the intrinsic AlGaN active layer, the thickness of the electron blocking layer, the ion implantation dose and depth in the electron blocking layer, the thickness of the P-type AlGaN layer, the thickness of the gradient-doped edge termination layer, and the N ion doping concentration in the gradient-doped edge termination layer.

10. The method for fabricating a gallium nitride device according to claim 9, characterized in that, In step S3, the optimal structural parameters of the gallium nitride device are determined using the Seagull optimization algorithm.

Images