Preparation and performance optimization method for multi-layer heterostructure of gallium nitride high-power device

Abstract

The invention provides a gallium nitride high-power device multilayer heterostructure preparation and performance optimization method, and relates to the technical field of semiconductor preparation, and the method comprises the steps: growing an aluminum nitride buffer layer and a multilayer nitride functional layer on a silicon carbide substrate; forming a double-layer passivation layer; manufacturing a source metal electrode and a drain metal electrode; forming a gradient doping channel; and manufacturing a gate metal electrode. Through the multilayer heterostructure design, the gradient doping channel and the interface engineering processing, the breakdown voltage of the device is improved, the on resistance is reduced, the heat dissipation performance is improved, and the reliability and the stability of the device are improved.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor manufacturing technology, and in particular to a method for preparing a multilayer heterostructure of a gallium nitride high-power device and optimizing its performance. Background Art

[0002] As electronic devices evolve toward high frequency, high power, and high integration, traditional silicon-based semiconductor devices are no longer able to meet the demands of modern power electronics. Gallium nitride (GaN), with its excellent properties such as wide bandgap, high critical breakdown electric field, high electron saturation velocity, and good thermal conductivity, has become an ideal material for the next generation of high-power electronic devices. In particular, GaN high-power devices based on silicon carbide (SiC) substrates demonstrate significant potential in high-temperature, high-frequency, and high-power applications, as they effectively combine the advantages of both materials.

[0003] The fabrication of high-power GaN devices typically involves the design and growth of multilayer heterostructures, including the optimized design of functional layers such as buffer, channel, and barrier layers. Traditionally, the fabrication process uses metal-organic chemical vapor deposition (MOCVD) to grow a thin film of nitride material on a SiC substrate. The source, gate, and drain electrodes are then fabricated through processes such as photolithography, etching, and metallization, ultimately forming the complete device structure.

[0004] However, the current preparation technology of GaN high-power devices still has the following defects and shortcomings: During heteroepitaxial growth, the lattice mismatch and thermal expansion coefficient differences between GaN and the SiC substrate easily lead to numerous defects, resulting in device performance degradation and reliability issues. Traditional single AlN buffer layers are ineffective in alleviating this problem, especially in high-temperature operating environments, where stress at the material interface can exacerbate device performance degradation.

[0005] Imperfect surface passivation technology is another significant factor limiting the performance of GaN devices. Existing single-layer passivation processes cannot simultaneously meet the requirements for high dielectric constant, low interface state density, and good thermal stability. This leads to reliability issues such as current collapse and increased leakage current when operating at high power. Summary of the Invention

[0006] The embodiments of the present invention provide a method for preparing a multilayer heterostructure of a gallium nitride high-power device and optimizing its performance, which can solve the problems in the prior art.

[0007] A first aspect of an embodiment of the present invention provides a method for preparing and optimizing the performance of a multilayer heterostructure of a gallium nitride high-power device, including: An aluminum nitride buffer layer is grown on a silicon carbide substrate using a metal organic chemical vapor deposition process, and a multi-layer nitride functional layer is grown on the aluminum nitride buffer layer using an in-situ annealing and segmented temperature adjustment process; A double-layer passivation layer is formed on the surface of the multi-layer nitride functional layer by combining atomic layer deposition and plasma enhanced treatment; Forming a source metal electrode and a drain metal electrode on the double-layer passivation layer by photolithography and selective ion etching processes, wherein the source metal electrode and the drain metal electrode are multi-layer metal structures treated by vacuum annealing; A digital ion implantation process is used to form a gradient doping channel in a region between the source metal electrode and the drain metal electrode, wherein the gradient doping channel includes a plurality of doping regions whose doping concentration decreases in sequence from the source to the drain; A gate metal electrode that has undergone interface engineering is formed above the gradient doping channel by adopting a self-alignment process. The gate metal electrode adopts a vacuum co-sputtered nickel / gold double-layer metal structure.

[0008] An aluminum nitride buffer layer is grown on a silicon carbide substrate using a metal organic chemical vapor deposition process, and a multi-layer nitride functional layer is grown on the aluminum nitride buffer layer using an in-situ annealing and segmented temperature adjustment process, comprising: performing plasma treatment on a silicon carbide substrate in a mixed atmosphere of hydrogen and nitrogen to form a reconstructed structure on a surface of the silicon carbide substrate; High-purity hydrogen carrier gas is introduced into the surface of the reconstructed silicon carbide substrate, and trimethylaluminum precursor and ammonia reaction gas are alternately introduced in a program-controlled pulse modulation manner, and the growth rate of the aluminum nitride buffer layer is controlled by adjusting the pulse time ratio of the trimethylaluminum precursor and the ammonia reaction gas; A trimethylaluminum precursor, a trimethylgallium precursor, and an ammonia reaction gas are introduced into the surface of the aluminum nitride buffer layer, and a multi-layer nitride functional layer is grown by a stepwise heating method to obtain multiple nitride layers; The multiple nitride layers are subjected to in-situ annealing treatment respectively, wherein the temperature of the in-situ annealing treatment is higher than the growth temperature of the corresponding layer by a set temperature, and a multi-layer nitride functional layer with strain modulation is formed by the in-situ annealing treatment.

[0009] A double-layer passivation layer is formed on the surface of the multi-layer nitride functional layer by combining atomic layer deposition and plasma enhanced treatment, comprising: treating the surface of the multilayer nitride functional layer with oxygen plasma to form hydroxyl surface active groups; Growing a first passivation layer of hafnium oxide on the hydroxyl surface active groups by an atomic layer deposition process, and forming a first hafnium oxide molecular layer by alternately introducing a hafnium tetrachloride precursor and water vapor reaction gas; During the growth of the first hafnium oxide molecular layer, the hafnium tetrachloride precursor and the water vapor reaction gas are activated by oxygen plasma enhanced treatment to form a dense hafnium oxide first passivation layer on the surface of the multi-layer nitride functional layer; A nitrogen carrier gas is introduced into the vacuum chamber, and an aluminum oxide second passivation layer is grown on the surface of the dense hafnium oxide first passivation layer using an atomic layer deposition process. A trimethylaluminum precursor and the water vapor reaction gas are alternately introduced in a program-controlled manner to form an aluminum oxide molecular layer on the surface of the dense hafnium oxide first passivation layer; The dense hafnium oxide first passivation layer and the aluminum oxide molecular layer are subjected to high-temperature annealing treatment in a nitrogen atmosphere to form a double-layer passivation structure with interface bonding.

[0010] During the growth of the first hafnium oxide molecular layer, the hafnium tetrachloride precursor and the water vapor reaction gas are activated by oxygen plasma enhanced treatment to form a dense hafnium oxide first passivation layer on the surface of the multilayer nitride functional layer, including: A hafnium tetrachloride precursor is introduced into the vacuum chamber to allow the hafnium tetrachloride precursor to chemically adsorb to hydroxyl functional groups on the surface of the multilayer nitride functional layer; Using nitrogen purge to remove unreacted hafnium tetrachloride precursor in the vacuum chamber, and forming a first reaction intermediate layer on the surface of the multi-layer nitride functional layer; introducing water vapor reaction gas into the vacuum chamber to allow the water vapor reaction gas to chemically react with the first reaction intermediate layer; generating oxygen plasma in the vacuum chamber using a radio frequency source, and performing an ion bombardment enhancement treatment on the first reactive intermediate layer and the water vapor reactive gas by the oxygen plasma, so as to reconstruct the chemical bond between the first reactive intermediate layer and the water vapor reactive gas; The reaction products in the vacuum chamber are removed by nitrogen purge, and a dense hafnium oxide first passivation layer is formed on the surface of the multi-layer nitride functional layer.

[0011] Forming a source metal electrode and a drain metal electrode on the double-layer passivation layer by using photolithography and selective ion etching processes, including: Spin-coating a photoresist on the surface of the double-layer passivation layer, and forming a first pattern area and a second pattern area in the photoresist by an exposure and development process; Etching the double-layer passivation layer using an inductively coupled plasma etching device, and forming a step-shaped etching structure between the first pattern area and the second pattern area by adjusting the radio frequency power and bias power of the inductively coupled plasma etching device; sputtering multiple functional metal layers in sequence on the surface of the stepped etched structure, and forming a multi-layer metal structure with stress matching by controlling the thickness ratio of the multiple functional metal layers; The photoresist is removed by a stripping process, a metal electrode pattern is formed between the multi-layer metal structure, the metal electrode pattern is annealed in a vacuum annealing furnace, and a source metal electrode and a drain metal electrode with an ohmic contact structure are formed by controlling the process parameters of the vacuum annealing furnace.

[0012] A digital ion implantation process is used to form a gradient doping channel in a region between the source metal electrode and the drain metal electrode, comprising: A positive photoresist is coated on the area between the source metal electrode and the drain metal electrode, and a doping window array is formed in the positive photoresist by using a laser direct writing lithography process; Performing a first ion implantation in the doping window array region using an ion implantation machine, and forming a high-concentration first doping region near the source metal electrode by controlling the implantation energy, beam current, and implantation dose of the ion implantation machine; Performing programmed ion implantation on the doping window array region using a digitally controlled ion implantation process to form a plurality of doping regions with successively decreasing doping concentrations between the first doping region and the drain metal electrode; Removing the positive photoresist using a plasma stripping process, performing a rapid thermal annealing process on the first doped region and the plurality of doped regions with successively decreasing doping concentrations, and forming a doped region with uniform ion distribution by controlling a temperature curve and an atmosphere composition of the rapid thermal annealing process; The doping region with uniform ion distribution is activated and annealed in a nitrogen atmosphere to form a gradient doping channel.

[0013] A digitally controlled ion implantation process is used to perform programmed ion implantation on the doping window array region to form a plurality of doping regions with successively decreasing doping concentrations between the first doping region and the drain metal electrode, including: Establishing a digital injection coordinate system in the doping window array region, and dividing the doping window array region into a plurality of injection sub-regions; Measuring surface topography parameters of the plurality of implantation sub-regions using an ion beam detector, and calculating an ion implantation compensation coefficient for each implantation sub-region according to the surface topography parameters; Performing programmed ion implantation on the plurality of implantation sub-regions using a digitally controlled ion implantation system, controlling the spatial distribution of the implanted ions in each implantation sub-region by adjusting the voltage of the ion implantation system, and controlling the ion implantation dose of each implantation sub-region by adjusting the implantation process parameters of the ion implantation system in combination with the ion implantation compensation coefficient, so that the implantation doses between adjacent implantation sub-regions decrease gradually according to a set ratio; Using a secondary ion mass spectrometer to measure the ion concentration distribution of the multiple implantation sub-regions to obtain ion concentration distribution measurement results; The voltage of the ion implantation system and the implantation process parameters are modified according to the ion concentration distribution measurement result until the multiple implantation sub-regions form doping regions with successively decreasing doping concentrations.

[0014] According to a second aspect of an embodiment of the present invention, an electronic device is provided, including: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0015] According to a third aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0016] The beneficial effects of this application are as follows: The present invention uses metal organic chemical vapor deposition to grow an aluminum nitride buffer layer on a silicon carbide substrate, and combines in-situ annealing and segmented temperature control processes to grow a multi-layer nitride functional layer, thereby significantly improving the crystal quality and interface characteristics of gallium nitride high-power devices, reducing stress defects, and thus improving the electrical performance and reliability of the devices.

[0017] The design of a double-layer passivation layer formed by combining atomic layer deposition and plasma enhanced processing effectively suppresses surface states and leakage current, enhances the thermal stability and voltage resistance of the device, extends the service life of the device, and is suitable for high temperature and high pressure working environments.

[0018] The gradient doping channel structure formed by the digital ion implantation process is combined with the gate metal electrode formed by the self-aligned process to optimize the electron transmission path, reduce the on-resistance of the device, increase the switching speed, and improve the high-frequency characteristics, so that the device has better overall performance in high-power application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the process of preparing and optimizing the performance of a multilayer heterostructure of a gallium nitride high-power device according to an embodiment of the present invention; Figure 2 Schematic diagram of the system architecture of the hafnium oxide molecular layer growth method. DETAILED DESCRIPTION

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0021] The following specific embodiments are used to describe the technical solution of the present invention in detail. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0022] Figure 1 FIG. 1 is a flow chart of a method for preparing a multilayer heterostructure of a gallium nitride high-power device and optimizing its performance according to an embodiment of the present invention. Figure 1 As shown, the method includes: An aluminum nitride buffer layer is grown on a silicon carbide substrate using a metal organic chemical vapor deposition process, and a multi-layer nitride functional layer is grown on the aluminum nitride buffer layer using an in-situ annealing and segmented temperature adjustment process; A double-layer passivation layer is formed on the surface of the multi-layer nitride functional layer by combining atomic layer deposition and plasma enhanced treatment; Forming a source metal electrode and a drain metal electrode on the double-layer passivation layer by photolithography and selective ion etching processes, wherein the source metal electrode and the drain metal electrode are multi-layer metal structures treated by vacuum annealing; A digital ion implantation process is used to form a gradient doping channel in a region between the source metal electrode and the drain metal electrode, wherein the gradient doping channel includes a plurality of doping regions whose doping concentration decreases in sequence from the source to the drain; A gate metal electrode that has undergone interface engineering is formed above the gradient doping channel by adopting a self-alignment process. The gate metal electrode adopts a vacuum co-sputtered nickel / gold double-layer metal structure.

[0023] In an optional embodiment, an aluminum nitride buffer layer is grown on a silicon carbide substrate using a metal organic chemical vapor deposition process, and a multi-layer nitride functional layer is grown on the aluminum nitride buffer layer using an in-situ annealing and segmented temperature adjustment process, comprising: performing plasma treatment on a silicon carbide substrate in a mixed atmosphere of hydrogen and nitrogen to form a reconstructed structure on a surface of the silicon carbide substrate; High-purity hydrogen carrier gas is introduced into the surface of the reconstructed silicon carbide substrate, and trimethylaluminum precursor and ammonia reaction gas are alternately introduced in a program-controlled pulse modulation manner, and the growth rate of the aluminum nitride buffer layer is controlled by adjusting the pulse time ratio of the trimethylaluminum precursor and the ammonia reaction gas; A trimethylaluminum precursor, a trimethylgallium precursor, and an ammonia reaction gas are introduced into the surface of the aluminum nitride buffer layer, and a multi-layer nitride functional layer is grown by a stepwise heating method to obtain multiple nitride layers; The multiple nitride layers are subjected to in-situ annealing treatment respectively, wherein the temperature of the in-situ annealing treatment is higher than the growth temperature of the corresponding layer by a set temperature, and a multi-layer nitride functional layer with strain modulation is formed by the in-situ annealing treatment.

[0024] Specifically, a silicon carbide substrate is placed in the reaction chamber of a metal organic chemical vapor deposition device, the temperature in the reaction chamber is raised to 800°C, and a mixture of hydrogen and nitrogen is introduced, wherein the hydrogen flow rate is 10 liters / minute, the nitrogen flow rate is 5 liters / minute, and the total pressure of the mixed gas is controlled at 100 torr. Maintain this temperature for 30 minutes, and at the same time, excite the plasma through a radio frequency source, and control the plasma power at 200 watts to treat the surface of the silicon carbide substrate. During this process, the plasma bombards the surface of the silicon carbide, removes surface oxides and contaminants, and rearranges the surface atoms to form an ordered reconstructed structure. Observation through atomic force microscopy shows that the surface roughness of the treated silicon carbide is reduced to below 0.2 nanometers, and a regular step structure is formed on the surface.

[0025] After the treatment was completed, the reaction chamber temperature was adjusted to 900°C, and high-purity hydrogen was introduced as a carrier gas at a flow rate of 15 liters / minute. Nitrogen was simultaneously stopped, and the reaction chamber pressure was controlled at 50 Torr. Under stable temperature and pressure conditions, a trimethylaluminum precursor and ammonia reactant gases were alternately introduced in a pulsed manner to grow an aluminum nitride buffer layer. The specific operation was as follows: first, the trimethylaluminum precursor was introduced for 2 seconds, then the trimethylaluminum source was turned off, and high-purity hydrogen was introduced to purge the reaction chamber for 1 second. Ammonia was then introduced for 3 seconds, and finally, high-purity hydrogen was introduced again for 1 second. This constituted a complete pulse cycle, totaling 7 seconds. The aluminum nitride growth rate could be effectively controlled by adjusting the pulse time ratio of trimethylaluminum to ammonia (2:3 in this example). This pulse cycle was repeated 50 times, resulting in a high-quality aluminum nitride buffer layer with a thickness of approximately 20 nanometers. X-ray diffraction measurements showed that the resulting aluminum nitride buffer layer had a half-width of 350 arc seconds, indicating good crystal quality.

[0026] After the aluminum nitride buffer layer is grown, the reaction chamber temperature is maintained at 900°C and the multi-layer nitride functional layer is grown. A step-by-step temperature increase is used, following the steps below: First layer: On the surface of the aluminum nitride buffer layer, trimethylaluminum precursor was introduced at a flow rate of 20 μmol / min, and ammonia was introduced at a flow rate of 5 liters / min. The growth temperature was maintained at 900°C and the growth time was 10 minutes to obtain an aluminum nitride layer with a thickness of 50 nanometers.

[0027] After growth is complete, an in-situ annealing treatment is performed. The temperature is raised to 950°C (50°C higher than the growth temperature) and maintained for 5 minutes in a nitrogen and hydrogen mixed atmosphere (nitrogen flow rate of 5 liters / minute, hydrogen flow rate of 2 liters / minute). This in-situ annealing treatment reduces the dislocation density in the first aluminum nitride layer. Measurements show that after annealing, the dislocation density drops from the original 1×10^9 / cm2 to 5×10^8 / cm2.

[0028] Second layer: The reaction chamber temperature was adjusted to 950°C, and a trimethylaluminum precursor was introduced at a flow rate of 15 μmol / min. A trimethylgallium precursor was introduced at a flow rate of 5 μmol / min, and ammonia was introduced at a flow rate of 6 liters / min. The growth time was 20 minutes, resulting in an aluminum-gallium-nitrogen alloy layer with a thickness of 100 nm and an aluminum content of approximately 75%.

[0029] After growth is complete, an in-situ annealing treatment is performed. The temperature is raised to 1000°C (50°C higher than the growth temperature) and maintained for 5 minutes in a mixed atmosphere of nitrogen and hydrogen (nitrogen flow rate of 5 liters / minute, hydrogen flow rate of 2 liters / minute). During the annealing process, the aluminum and gallium atoms in the aluminum-gallium-nitrogen alloy layer are redistributed, forming an alloy layer with slightly fluctuating composition, effectively relieving the stress generated during the growth process.

[0030] Layer 3: Adjust the reaction chamber temperature to 1000°C, introduce a trimethylaluminum precursor at a flow rate of 10 μmol / min, and simultaneously introduce a trimethylgallium precursor at a flow rate of 10 μmol / min and ammonia at a flow rate of 6 liters / min. The growth time is 25 minutes, resulting in an aluminum-gallium-nitrogen alloy layer with a thickness of 150 nm and an aluminum content of approximately 50%.

[0031] After growth, an in-situ annealing treatment was performed. The temperature was raised to 1050°C (50°C higher than the growth temperature) and maintained for 5 minutes in a nitrogen and hydrogen mixed atmosphere (nitrogen flow rate of 5 liters / minute and hydrogen flow rate of 2 liters / minute).

[0032] Fourth layer: The reaction chamber temperature was adjusted to 1050°C, and a trimethylaluminum precursor was introduced at a flow rate of 5 μmol / min. A trimethylgallium precursor was introduced at a flow rate of 15 μmol / min, and ammonia was introduced at a flow rate of 7 liters / min. The growth time was 30 minutes, resulting in an aluminum gallium nitrogen alloy layer with a thickness of 200 nm and an aluminum content of approximately 25%.

[0033] After growth, an in-situ annealing treatment was performed. The temperature was raised to 1100°C (50°C higher than the growth temperature) and maintained for 5 minutes in a nitrogen and hydrogen mixed atmosphere (nitrogen flow rate of 5 liters / minute and hydrogen flow rate of 2 liters / minute).

[0034] For the final layer, the reaction chamber temperature was adjusted to 1100°C, and only trimethylgallium precursor was introduced at a flow rate of 20 μmol / min, along with ammonia at a flow rate of 7 L / min. The growth time was 40 minutes, resulting in a 300 nm thick gallium nitride layer.

[0035] After the growth is completed, the temperature is raised to 1150°C (50°C higher than the growth temperature) and maintained for 10 minutes in a mixed atmosphere of nitrogen and hydrogen (nitrogen flow rate of 5 liters / minute, hydrogen flow rate of 2 liters / minute) to further reduce dislocations in the gallium nitride layer and reduce surface roughness.

[0036] Through the above process, a strain-modulated multilayer nitride functional layer was formed. High-resolution transmission electron microscopy revealed clear interfaces between the layers, with no apparent defect extension. Hall measurements revealed that the room-temperature electron mobility of the topmost gallium nitride layer reached 1500 square centimeters per volt-second, with an electron concentration of 2×10^17 per cubic centimeter, demonstrating the material's excellent electrical properties. Photoluminescence measurements revealed a band-edge emission peak at 363 nanometers at room temperature, with a full width at half maximum of 38 meV, further confirming the material's high quality. This strain-modulated multilayer nitride functional layer structure effectively reduces dislocation density and improves epitaxial layer quality, laying the foundation for subsequent device fabrication.

[0037] In an optional embodiment, a double-layer passivation layer is formed on the surface of the multi-layer nitride functional layer by combining atomic layer deposition and plasma enhanced treatment, comprising: treating the surface of the multilayer nitride functional layer with oxygen plasma to form hydroxyl surface active groups; Growing a first passivation layer of hafnium oxide on the hydroxyl surface active groups by an atomic layer deposition process, and forming a first hafnium oxide molecular layer by alternately introducing a hafnium tetrachloride precursor and water vapor reaction gas; During the growth of the first hafnium oxide molecular layer, the hafnium tetrachloride precursor and the water vapor reaction gas are activated by oxygen plasma enhanced treatment to form a dense hafnium oxide first passivation layer on the surface of the multi-layer nitride functional layer; A nitrogen carrier gas is introduced into the vacuum chamber, and an aluminum oxide second passivation layer is grown on the surface of the dense hafnium oxide first passivation layer using an atomic layer deposition process. A trimethylaluminum precursor and the water vapor reaction gas are alternately introduced in a program-controlled manner to form an aluminum oxide molecular layer on the surface of the dense hafnium oxide first passivation layer; The dense hafnium oxide first passivation layer and the aluminum oxide molecular layer are subjected to high-temperature annealing treatment in a nitrogen atmosphere to form a double-layer passivation structure with interface bonding.

[0038] The multilayer nitride functional layer can be a semiconductor device structure composed of materials such as aluminum nitride, gallium nitride, and indium nitride. Before surface passivation, the surface of the multilayer nitride functional layer must be cleaned and pretreated using ultrasonic cleaning with acetone and isopropyl alcohol, each lasting 10 minutes. The surface is then rinsed with deionized water and dried with nitrogen to remove organic contaminants.

[0039] The cleaned multilayer nitride functional layer surface was treated with oxygen plasma. The sample was placed in a plasma chamber and evacuated to a base vacuum of 5 × 10⁻⁶ Torr. High-purity oxygen was then introduced at a controlled flow rate of 20 sccm, and the chamber pressure was stabilized at 50 mTorr. The RF power was set to 100 W, and the treatment time was 60 seconds. During the plasma treatment, reactive oxygen ions and free radicals reacted with surface atoms, forming hydroxyl (-OH) surface-active groups on the surface of the multilayer nitride functional layer. Testing showed a surface hydroxyl density of up to 4.5 × 10⁻¹⁴ cm⁻², providing ample reaction sites for subsequent atomic layer deposition.

[0040] A first passivation layer of hafnium oxide (HfO2) was grown using atomic layer deposition (ALD) on the surface of the multilayer nitride functional layer that formed hydroxyl surface active groups. The sample was placed in an ALD reaction chamber controlled at 200°C and a pressure of 0.5 Torr. A hafnium tetrachloride precursor and water vapor reaction gases were introduced alternately through a precisely controlled pulse sequence to form the first molecular layer of hafnium oxide.

[0041] During the growth of the hafnium oxide molecular layer, oxygen plasma-enhanced treatment was used to activate the hafnium tetrachloride precursor and water vapor reaction gas. After the water pulse of each ALD cycle, oxygen plasma treatment was immediately introduced with an oxygen flow rate of 30 sccm, an RF power of 50 W, and a treatment time of 3 seconds. Plasma-enhanced treatment can provide additional active oxygen species, promote the complete decomposition of the precursor and the surface reaction, effectively remove the reaction byproduct HCl, and enhance the density and uniformity of the film. In this way, the hafnium oxide passivation layer formed on the surface of the multilayer nitride functional layer has a thickness of 1.2×10^-7 cm 2 / V·s and a high breakdown electric field strength of 8.0 MV / cm.

[0042] After the dense hafnium oxide first passivation layer is formed, the atomic layer deposition process is continued to grow the second aluminum oxide passivation layer on its surface. Nitrogen carrier gas is introduced into the vacuum chamber with a flow rate of 200 sccm, and the chamber temperature is adjusted to 250°C. Through program control, trimethylaluminum (TMA) precursor and water vapor reaction gas are alternately introduced to form an aluminum oxide molecular layer on the surface of the dense hafnium oxide first passivation layer. The TMA pulse time is 0.05 seconds, the nitrogen purge time is 15 seconds, the H2O pulse time is 0.02 seconds, and the nitrogen purge time is 15 seconds again, which constitutes a complete aluminum oxide ALD cycle. By repeating 100 cycles, an aluminum oxide layer with a thickness of about 10 nm can be formed, with a refractive index of 1.65 and a uniformity better than ±2%.

[0043] After the formation of the double-layer passivation structure is completed, a high-temperature annealing treatment is required to enhance the interfacial bonding strength between the two passivation layers and the adhesion between the passivation layer and the multilayer nitride functional layer. The sample is placed in an annealing furnace and subjected to high-temperature annealing in a nitrogen atmosphere with a nitrogen flow rate of 500 sccm and a heating rate of 10°C / minute from room temperature to 600°C. The temperature is maintained at this constant temperature for 30 minutes, and then naturally cooled to room temperature at a rate of 5°C / minute. During the annealing process, a mixed transition zone is formed at the interface between the dense hafnium oxide first passivation layer and the aluminum oxide second passivation layer. The thickness is approximately 1-2 nm. This mixed zone has good interfacial bonding properties and can effectively reduce the density of interface defect states.

[0044] The double-layer passivation structure formed after the above treatment exhibits excellent performance indicators. X-ray photoelectron spectroscopy (XPS) analysis results show that an Hf-O-Al bonding structure is formed at the hafnium oxide / aluminum oxide interface, and the interface state density is reduced to 5×10^10 cm^-2·eV^-1. Electrical characterization shows that the double-layer passivation structure has an equivalent oxide thickness of 14 nm, a dielectric constant of approximately 12, and a breakdown electric field strength of up to 10 MV / cm. Applying this double-layer passivation structure to multilayer nitride devices can reduce the leakage current of the device by two orders of magnitude and reduce the surface recombination rate from 10^4 cm / s to 10^2 cm / s, effectively improving the reliability and stability of the device and extending its service life.

[0045] By the method described in detail above, a double-layer passivation structure with excellent performance can be formed on the surface of the multi-layer nitride functional layer. This structure can effectively suppress surface states, reduce leakage current, and improve device performance and reliability.

[0046] In an optional embodiment, during the growth of the first hafnium oxide molecular layer, the hafnium tetrachloride precursor and the water vapor reaction gas are activated by oxygen plasma enhanced treatment to form a dense hafnium oxide first passivation layer on the surface of the multilayer nitride functional layer, comprising: A hafnium tetrachloride precursor is introduced into the vacuum chamber to allow the hafnium tetrachloride precursor to chemically adsorb to hydroxyl functional groups on the surface of the multilayer nitride functional layer; Using nitrogen purge to remove unreacted hafnium tetrachloride precursor in the vacuum chamber, and forming a first reaction intermediate layer on the surface of the multi-layer nitride functional layer; introducing water vapor reaction gas into the vacuum chamber to allow the water vapor reaction gas to chemically react with the first reaction intermediate layer; generating oxygen plasma in the vacuum chamber using a radio frequency source, and performing an ion bombardment enhancement treatment on the first reactive intermediate layer and the water vapor reactive gas by the oxygen plasma, so as to reconstruct the chemical bond between the first reactive intermediate layer and the water vapor reactive gas; The reaction products in the vacuum chamber are removed by nitrogen purge, and a dense hafnium oxide first passivation layer is formed on the surface of the multi-layer nitride functional layer.

[0047] Figure 2 This is a schematic diagram of the system architecture of a hafnium oxide molecular layer growth method. The present invention provides a method for growing a hafnium oxide molecular layer using oxygen plasma enhanced treatment. Specifically, during the growth of the first hafnium oxide molecular layer, the hafnium tetrachloride precursor and water vapor reaction gas are activated by oxygen plasma enhanced treatment to form a dense hafnium oxide first passivation layer on the surface of the multi-layer nitride functional layer. The detailed implementation method is as follows: Before growing the first hafnium oxide molecular layer, a substrate with multiple nitride functional layers must be prepared. These layers can be formed on the substrate surface via physical vapor deposition (PVD) or chemical vapor deposition (CVD). The substrate material can be selected from common semiconductor materials such as silicon, silicon dioxide, and silicon nitride. The nitride functional layers can include materials such as titanium nitride, tantalum nitride, and niobium nitride, with thicknesses ranging from 5 to 50 nanometers.

[0048] The first step is to pretreat the surface of the multilayer nitride functional layer. The sample is placed in a vacuum chamber and the vacuum is reduced to 0.01-0.1 Pa. Hydroxyl functional groups are formed on the surface of the multilayer nitride functional layer via oxygen plasma treatment or wet chemical treatment. Oxygen plasma treatment can be performed at 50-300 watts of RF power for 30-120 seconds. Wet chemical treatment can be performed by soaking in a hydrogen peroxide solution (3%-30%) at room temperature for 2-10 minutes, followed by rinsing with deionized water and drying with nitrogen.

[0049] The second step is to introduce a hafnium tetrachloride precursor into the vacuum chamber. The hafnium tetrachloride precursor can be introduced into the chamber by gaseous delivery or solid sublimation. For gaseous delivery, the temperature of the hafnium tetrachloride source is controlled at 40-60°C, and the carrier gas flow rate is 10-50sccm; for solid sublimation, the temperature of the hafnium tetrachloride source is controlled at 150-180°C. The chamber pressure is maintained at 0.1-10 Pa, and the pulse time of the hafnium tetrachloride precursor is 0.5-2.0 seconds. During this process, the hafnium tetrachloride molecules undergo chemical adsorption with the hydroxyl functional groups on the surface of the multilayer nitride functional layer to form Hf-O bonds, while releasing HCl byproducts.

[0050] The third step is to purge unreacted hafnium tetrachloride precursor using nitrogen. Nitrogen purity is 99.999%, flow rate is 50-200 sccm, purge time is 3-10 seconds, and chamber pressure is maintained at 1-20 Pa. The purge process ensures that remaining unreacted hafnium tetrachloride molecules and HCl gas generated by the reaction are completely exhausted in the chamber, preventing gas-phase reactions. At this point, a chemically adsorbed hafnium tetrachloride monolayer forms on the surface of the multilayer nitride functional layer, which is the first reactive intermediate layer.

[0051] The fourth step is to introduce water vapor into the vacuum chamber. The water vapor source temperature is controlled at 20-40°C and delivered to the chamber via a mass flow controller at a flow rate of 10-50 sccm. The pulse duration is 0.5-2.0 seconds, and the chamber pressure is maintained at 0.1-10 Pa. The water vapor molecules react with the remaining Hf-Cl bonds in the first reaction layer, forming Hf-OH bonds and releasing HCl as a byproduct.

[0052] The fifth step is to use a radio frequency source to generate oxygen plasma in a vacuum chamber. The radio frequency power is 50-300 watts and the frequency is 13.56 MHz. High-purity oxygen is introduced into the chamber at a flow rate of 20-100 sccm, and the chamber pressure is maintained at 1-10 Pa. The plasma treatment time is 5-30 seconds. The first reaction intermediate layer and the water vapor reaction gas are subjected to ion bombardment enhancement treatment by oxygen plasma. The generated active oxygen atoms, oxygen ions and free radicals react with the surface Hf-OH bonds, promoting the reconstruction of chemical bonds and forming a denser HfO2 network structure. Plasma-enhanced treatment can significantly reduce the impurity content in the film, such as chlorine, carbon and other residues, while increasing the film density.

[0053] The sixth step is to purge the vacuum chamber with nitrogen to remove reaction products. Nitrogen purity is 99.999%, with a flow rate of 50-200 sccm for 3-10 seconds, and the chamber pressure is maintained at 1-20 Pa. This step removes any remaining reaction products, such as water vapor and HCl, from the chamber, completing a complete hafnium oxide molecular layer growth cycle. This completes the formation of a dense first hafnium oxide passivation layer on the surface of the multilayer nitride functional layer.

[0054] Experimental results show that the hafnium oxide first passivation layer prepared by the above method has excellent performance. X-ray photoelectron spectroscopy analysis shows that the chlorine content in the hafnium oxide passivation layer is less than 0.5 atomic percent and the carbon content is less than 1.0 atomic percent. High-resolution transmission electron microscopy observations show that the thickness of the hafnium oxide first passivation layer is 1.1±0.1 nanometers, the interface is flat, and there are no obvious defects. Atomic force microscopy measures the surface roughness Rq value to be less than 0.3 nanometers. The dielectric constant test results are 18-22, and the breakdown electric field strength exceeds 8 MV / cm. The leakage current density is less than 1×10⁻ at a bias of 1V. 8 A / cm². In addition, after annealing at 400℃ in nitrogen atmosphere for 2 hours, the film properties remained stable, indicating that the hafnium oxide first passivation layer prepared by this method has good thermal stability.

[0055] The above process can be adjusted appropriately based on actual application requirements. For example, the thickness and density of the hafnium oxide passivation layer can be adjusted by controlling parameters such as the number of pulses, plasma power, and processing time. In specific applications, doping elements such as aluminum and yttrium can be introduced during the hafnium oxide growth process to further enhance film performance.

[0056] In an optional embodiment, a source metal electrode and a drain metal electrode are formed on the double-layer passivation layer by using a photolithography and selective ion etching process, comprising: Spin-coating a photoresist on the surface of the double-layer passivation layer, and forming a first pattern area and a second pattern area in the photoresist by an exposure and development process; Etching the double-layer passivation layer using an inductively coupled plasma etching device, and forming a step-shaped etching structure between the first pattern area and the second pattern area by adjusting the radio frequency power and bias power of the inductively coupled plasma etching device; sputtering multiple functional metal layers in sequence on the surface of the stepped etched structure, and forming a multi-layer metal structure with stress matching by controlling the thickness ratio of the multiple functional metal layers; The photoresist is removed by a stripping process, a metal electrode pattern is formed between the multi-layer metal structure, the metal electrode pattern is annealed in a vacuum annealing furnace, and a source metal electrode and a drain metal electrode with an ohmic contact structure are formed by controlling the process parameters of the vacuum annealing furnace.

[0057] The method of forming a source metal electrode and a drain metal electrode on the double-layer passivation layer by using photolithography and selective ion etching processes relates to a manufacturing technology of a semiconductor device. The method realizes a high-performance source and drain electrode structure by precisely controlling the etching and metal deposition processes.

[0058] The preparation process begins with a photolithography process on the surface of the double-layer passivation layer. A 2.5-micron thick layer of photoresist is spin-coated on the double-layer passivation layer at a speed of 4,000 rpm for 30 seconds. After spin coating, the photoresist is soft-baked at 90°C for 60 seconds to fully evaporate the solvent on the surface of the photoresist layer. The soft-baked photoresist is then exposed to light at a dose of 20 mJ / cm² for 2 seconds. A mask is used to define the first and second graphic areas on the photoresist. After exposure, a development process is performed using a TMAH developer commonly used in the prior art for 25 seconds. The photoresist is then rinsed with deionized water for 60 seconds and blown dry with nitrogen. After development, the photoresist is hard-baked at 120°C for 90 seconds to fully cure the photoresist.

[0059] After the photolithography process is completed, the double-layer passivation layer is etched using an inductively coupled plasma etching device. During the etching process, the working gas pressure is set to 5 Pa and the gas flow ratio is CF4:O2=30:5 standard cubic centimeters / minute. The key is the precise control of the RF power and bias power during the etching process. First, the RF power is set to 700 watts and the bias power is set to 100 watts. The first graphic area is etched for 80 seconds, so that the double-layer passivation layer in the first graphic area is completely removed, exposing the surface of the underlying semiconductor material. Subsequently, the RF power is adjusted to 500 watts and the bias power is adjusted to 50 watts. The second graphic area is continued to be etched for 40 seconds, so that the etching depth of the double-layer passivation layer in the second graphic area is only the thickness of the upper passivation layer, and the lower passivation layer in the second graphic area is still retained. In this way, by adjusting the etching power and time, a step-shaped etching structure is formed between the first graphic area and the second graphic area, with a step height of 150 nanometers and a step width of 5 microns.

[0060] After the step-shaped etching structure is formed, the sample is plasma cleaned with the following cleaning parameters: oxygen flow rate 20 standard cubic centimeters per minute, power 100 watts, and processing time 30 seconds to remove the residues and contaminants generated during the etching process. After cleaning, the sample is immediately transferred to the vacuum chamber of the sputtering equipment to avoid contamination of the sample surface. In the sputtering equipment, magnetron sputtering is used to deposit multiple functional metal layers in sequence. During the sputtering process, the vacuum degree is maintained at 5×10⁻ 6 The sputtering power was 200 watts, and the substrate temperature was controlled at 200°C during sputtering. The multilayer metal structure, from bottom to top, consists of: a 10-nanometer-thick titanium layer as an adhesion layer; a 30-nanometer-thick aluminum layer as a low-contact resistance layer; a 100-nanometer-thick molybdenum layer as a stress-regulating layer; a 200-nanometer-thick aluminum layer as the main conductive layer; and a 20-nanometer-thick titanium layer as a protective layer. This specialized multilayer metal structure design achieves stress matching between the metal layers. In the titanium / aluminum / molybdenum / aluminum / titanium multilayer metal structure, the compressive stress of the molybdenum layer is balanced with the tensile stress of the aluminum layer, keeping the combined stress within ±100 MPa.

[0061] After sputtering, the sample is transferred to an organic solvent bath for a stripping process. Acetone is used as the stripping solution, and the sample is immersed at 40°C for 15 minutes. Ultrasonic vibration is applied at a power of 50 watts for 60 seconds. This stripping process removes the excess metal covering the photoresist, leaving only the metal directly in contact with the double-layer passivation layer and the semiconductor material, forming the desired metal electrode pattern. The stripped sample is then cleaned with isopropyl alcohol for 5 minutes, then rinsed with deionized water for 3 minutes, and finally blown dry with nitrogen.

[0062] After forming the metal electrode pattern, the sample is placed in a vacuum annealing furnace for annealing. During the annealing process, the vacuum degree of the furnace chamber is controlled at 1×10⁻ 5 Torr, heating rate of 10 ° C / min, annealing temperature of 350 ° C, holding time of 30 minutes, and then naturally cooled to room temperature at a rate of 5 ° C / min. The annealing process promotes the interface reaction between the metal and semiconductor materials, forming a good ohmic contact structure. The contact resistance between the source metal electrode and the drain metal electrode and the semiconductor material is less than 1 × 10⁻ as measured by the four-probe test method. 6 Ohm·square centimeter, excellent linear IV characteristics, leakage current less than 10 nanoamperes, and breakdown voltage greater than 30 volts.

[0063] The source metal electrode and drain metal electrode structures prepared using the above method have excellent electrical performance and reliability. After 1000 hours of high temperature and high humidity testing (85°C, 85% relative humidity), the electrode structure has no obvious degradation, the contact resistance changes by less than 5%, and there is no peeling or bubbling phenomenon, indicating that the electrode structure formed by this preparation method has good long-term stability.

[0064] In an optional embodiment, a digital ion implantation process is used to form a gradient doping channel in a region between the source metal electrode and the drain metal electrode, comprising: A positive photoresist is coated on the area between the source metal electrode and the drain metal electrode, and a doping window array is formed in the positive photoresist by using a laser direct writing lithography process; Performing a first ion implantation in the doping window array region using an ion implantation machine, and forming a high-concentration first doping region near the source metal electrode by controlling the implantation energy, beam current, and implantation dose of the ion implantation machine; Performing programmed ion implantation on the doping window array region using a digitally controlled ion implantation process to form a plurality of doping regions with successively decreasing doping concentrations between the first doping region and the drain metal electrode; Removing the positive photoresist using a plasma stripping process, performing a rapid thermal annealing process on the first doped region and the plurality of doped regions with successively decreasing doping concentrations, and forming a doped region with uniform ion distribution by controlling a temperature curve and an atmosphere composition of the rapid thermal annealing process; The doping region with uniform ion distribution is activated and annealed in a nitrogen atmosphere to form a gradient doping channel.

[0065] In one implementation, a method for forming a gradient doping channel in a region between a source metal electrode and a drain metal electrode using a digital ion implantation process relates to the technical field of semiconductor device manufacturing, and is particularly suitable for preparing a channel region in a field effect transistor.

[0066] Apply positive photoresist to the area between the source and drain metal electrodes. Specifically, a spin coating method can be used to apply a layer of positive photoresist with a thickness of 1.0-1.5 microns onto the substrate where the source and drain metal electrodes have been formed. The spin coating parameters can be set to: a rotation speed of 4000 rpm and a duration of 40 seconds. After coating, bake at 110°C for 90 seconds to cure the photoresist.

[0067] A doping window array is formed in a positive photoresist using a laser direct write lithography process. The laser direct write lithography system uses a light source wavelength of 365 nanometers, a laser energy density of 50-100 millijoules per square centimeter, and an exposure time of 100-200 milliseconds. Depending on the design requirements, the doping window array can be arranged in a rectangular array with a row spacing of 5 microns and a column spacing of 3 microns, with individual windows measuring 2 microns by 2 microns. After photolithography, the doping window array is immersed in a developer at 25°C for 60 seconds, then rinsed with deionized water for 30 seconds and dried with nitrogen, forming a regularly arranged doping window array structure.

[0068] The first ion implantation was performed in the doping window array area using an ion implantation machine. The operating parameters of the ion implantation machine were set as follows: implantation energy 80-120 keV, beam current 10-15 μA, and implantation dose 1.0×10 15 -5.0×10 15 ions / cm². N-type dopants such as phosphorus or arsenic are selected as implanted ions at a 7-degree angle to avoid channeling. This step primarily forms a high-concentration first doping region near the source metal electrode, ensuring good ohmic contact between the source and the channel.

[0069] A digitally controlled ion implantation process is used to perform programmed ion implantation in the doping window array area. The digital control system, according to a preset program, sequentially implants ions with decreasing doping concentrations in different areas from source to drain. In specific implementation, the area between the source and drain can be divided into 4-6 doping regions. For the second doping region, the implantation energy is set to 70-90 keV, the beam current is 8-12 μA, and the implantation dose is 5.0×10 14 -1.0×10 15 ions / cm 2 For the third doping region, the injection energy is set to 60-80 keV, the beam current is 6-10 μA, and the injection dose is 1.0×10 14 -5.0×10 14 ions / cm 2 For the fourth doping region, the injection energy is set to 50-70 keV, the beam current is 4-8 μA, and the injection dose is 5.0×10 13 -1.0×10 14 ions / cm²; for the doped region closest to the drain, the implantation energy is set to 40-60keV, the beam current is 2-6μA, and the implantation dose is 1.0×10 13 -5.0×10 13 ions / cm². In this way, multiple doping regions with successively decreasing doping concentrations are formed between the first doping region and the drain metal electrode, and the doping concentration presents a gradient distribution that decreases from the source to the drain.

[0070] A plasma stripping process is used to remove the positive photoresist. The parameters for this process are: oxygen flow rate 200-300 sccm, RF power 800-1200W, chamber pressure 0.5-1.0 Pa, and treatment time 180-240 seconds. This process effectively removes the photoresist without damaging the existing doped regions.

[0071] The first doped region and multiple doped regions with decreasing doping concentrations are subjected to rapid thermal annealing. The temperature curve of the rapid thermal annealing furnace is controlled as follows: heating from room temperature to 850-950°C at a heating rate of 50-80°C / second, holding for 20-40 seconds, and then cooling to room temperature at a cooling rate of 30-50°C / second. The annealing atmosphere is composed of a mixture of 95% nitrogen and 5% hydrogen, with a flow rate controlled at 2-3 liters / minute. By precisely controlling the temperature curve and atmosphere composition of the annealing process, the implanted ions can be more evenly distributed within each doped region while maintaining the concentration gradient characteristics between regions.

[0072] The doped area with uniform ion distribution is activated and annealed in a nitrogen atmosphere. The activation annealing temperature is set at 950-1050°C, the holding time is 60-120 seconds, and the nitrogen flow rate is controlled at 3-5 liters / minute. The activation annealing process allows the doped ions to enter the lattice position and become effective carriers, while repairing the lattice damage caused by ion implantation. After this step, a gradient doping channel with a gradual distribution of doping concentration from source to drain is formed, with the doping concentration increasing from about 1.0×10 near the source to 1.0×10 near the drain. 20 cm⁻ 3 Gradually decreases to about 1.0×10 near the drain 17 cm⁻ 3 .

[0073] The gradient-doped channel structure formed by this process effectively reduces contact resistance in the highly doped region near the source, regulates the electric field distribution in the gradient-doped region in the middle, and improves the device's breakdown voltage in the low-doped region near the drain. Tests have shown that devices employing this gradient-doped channel structure achieve higher operating frequencies and lower power consumption than conventional uniformly doped channel devices of the same size, significantly improving device performance.

[0074] In an optional embodiment, a digitally controlled ion implantation process is used to perform programmed ion implantation on the doping window array region to form a plurality of doping regions with successively decreasing doping concentrations between the first doping region and the drain metal electrode, including: Establishing a digital injection coordinate system in the doping window array region, and dividing the doping window array region into a plurality of injection sub-regions; Measuring surface topography parameters of the plurality of implantation sub-regions using an ion beam detector, and calculating an ion implantation compensation coefficient for each implantation sub-region according to the surface topography parameters; Performing programmed ion implantation on the plurality of implantation sub-regions using a digitally controlled ion implantation system, controlling the spatial distribution of the implanted ions in each implantation sub-region by adjusting the voltage of the ion implantation system, and controlling the ion implantation dose of each implantation sub-region by adjusting the implantation process parameters of the ion implantation system in combination with the ion implantation compensation coefficient, so that the implantation doses between adjacent implantation sub-regions decrease gradually according to a set ratio; Using a secondary ion mass spectrometer to measure the ion concentration distribution of the multiple implantation sub-regions to obtain ion concentration distribution measurement results; The voltage of the ion implantation system and the implantation process parameters are modified according to the ion concentration distribution measurement result until the multiple implantation sub-regions form doping regions with successively decreasing doping concentrations.

[0075] This embodiment provides a method for performing programmed ion implantation on a doping window array region using a digitally controlled ion implantation process, for forming a plurality of doping regions with successively decreasing doping concentrations between a first doping region and a drain metal electrode.

[0076] In this embodiment, the doping window array region refers to a specific area in the semiconductor device that needs to be doped. This area is generally located between the first doping region and the drain metal electrode. The size of the doping window array region can be designed according to actual needs, for example, it can be a rectangular area of ​​10 μm × 8 μm.

[0077] When establishing a digital injection coordinate system in the doping window array area, the lower left corner of the doping window array area can be set as the coordinate origin (0,0), and the upper right corner as the coordinate (10,8). The doping window array area is divided into multiple injection sub-areas according to the number of doping gradient regions to be formed. In this embodiment, the doping window array area is divided into 5 injection sub-areas, which are respectively recorded as A1, A2, A3, A4 and A5. The size of each injection sub-area is 2μm×8μm, and they are arranged in sequence from the first doping area to the drain metal electrode.

[0078] When using an ion beam detector to measure the surface topography parameters of multiple implanted sub-regions, a scanning electron microscope equipped with an ion beam detector can be used. This device scans the surface of each implanted sub-region with a scanning step size of 0.1 μm to obtain surface height distribution data. For region A1, the measured average surface height is 120 nm and the surface roughness is 5 nm. For regions A2 to A5, the measured average surface heights are 118 nm, 121 nm, 117 nm, and 119 nm, respectively, and the surface roughness is 6 nm, 4 nm, 7 nm, and 5 nm, respectively.

[0079] When calculating the ion implantation compensation coefficient for each implantation sub-region based on surface topography parameters, the effects of surface height and surface roughness on the ion implantation depth and concentration distribution were considered. Using correlations established with experimental data, the ion implantation compensation coefficients for each implantation sub-region were determined to be 1.02, 0.98, 1.03, 0.96, and 1.01, respectively.

[0080] When using a digitally controlled ion implantation system to perform programmed ion implantation on multiple implantation sub-areas, a medium energy ion implanter with precise voltage control capability can be used. The system can control the ion implantation energy to be continuously adjustable within the range of 30keV to 300keV, and the ion beam current intensity can be adjusted to 0.1μA / cm 2 to 10μA / cm 2 The injection dose accuracy can be adjusted within the range of ±1%. In this embodiment, boron ions are selected as doping ions and the reference injection dose is set to 1×10 15 atoms / cm 2 .

[0081] To achieve precise spatial distribution of the implanted ions within each implantation sub-region, the ion projection depth was controlled by adjusting the accelerating voltage of the ion implantation system. For implantation sub-region A1, the accelerating voltage was set to 150 keV; for implantation sub-regions A2 to A5, the accelerating voltages were set to 135 keV, 120 keV, 105 keV, and 90 keV, respectively, to achieve a gradient distribution from deep to shallow layers.

[0082] At the same time, the ion implantation dose of each implantation sub-region is controlled by adjusting the implantation process parameters of the ion implantation system and combining the ion implantation compensation coefficient. The implantation dose ratio gradient between adjacent implantation sub-regions is set to 0.7, that is, the dose of each implantation sub-region is 0.7 times the dose of the adjacent implantation sub-region closer to the first doping region. Combined with the ion implantation compensation coefficient, the actual implantation dose of the implantation sub-region A1 is finally determined to be 1×10 15 atoms / cm 2 ×1.02=1.02×10 15 atoms / cm 2 The actual injection dose of the sub-region A2 is 1×10 15 atoms / cm²×0.7×0.98=6.86×10 14 atoms / cm²; the actual implantation doses in sub-regions A3 to A5 are 4.96×10 14 atoms / cm²、3.36×10 14 atoms / cm² and 2.37×10 14 atoms / cm².

[0083] During the implantation process, the temperature of the ion implantation system was maintained at 22±0.5℃ and the vacuum degree was maintained at 1×10⁻ 6 Pa, the sample tilt angle was set to 7 degrees to avoid channeling, and the rotation speed was 10 rpm to ensure uniform implantation.

[0084] After ion implantation, a secondary ion mass spectrometer was used to measure the ion concentration distribution of multiple implantation sub-areas. The measurement parameters were set as follows: primary ion beam Cs⁺, energy of 5 keV, incident angle of 45 degrees, and raster size of 200 μm × 200 μm. Five measurement points were taken for each implantation sub-area, and the average value was calculated as the ion concentration distribution result for that area. The measurement results showed that the peak ion concentration in implantation sub-area A1 was 9.8×10 19 atoms / cm³, with a peak depth of 0.41 μm; the peak ion concentration in the implanted sub-region A2 is 6.7×10 19 atoms / cm³, with a peak depth of 0.37 μm; the peak ion concentrations in the implanted sub-regions A3 to A5 were 4.8×10 19 atoms / cm³、3.2×10 19 atoms / cm³ and 2.3×10 19 atoms / cm³, with peak depths of 0.33μm, 0.29μm, and 0.25μm, respectively.

[0085] According to the ion concentration distribution measurement results, it was found that the actual concentration of the implanted sub-area A3 deviated from the expected concentration by 8%, exceeding the allowable ±5% range. Through analysis, it was determined that the accelerating voltage of the ion implantation system for this area needed to be adjusted from 120keV to 125keV, and the implantation dose was corrected to 5.15×10 14 atoms / cm². After re-implantation of the implantation sub-area A3, the secondary measurement results showed that the ion peak concentration was 5.1×10 19 atoms / cm³, the peak depth is 0.34μm, and the deviation is reduced to 3%, which meets the requirements.

[0086] After the programmed ion implantation process, a doping region with decreasing doping concentration was successfully formed between the first doping region and the drain metal electrode, achieving a doping concentration from 9.8×10 19 atoms / cm 3 Gradually reduced to 2.3×10 19 atoms / cm 3 The gradient distribution meets the requirements of high-performance semiconductor devices for precise control of doping concentration.

[0087] According to a second aspect of an embodiment of the present invention, an electronic device is provided, including: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0088] According to a third aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0089] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0090] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for preparing and optimizing the performance of a multilayer heterostructure of a gallium nitride high-power device, characterized in that: include: An aluminum nitride buffer layer is grown on a silicon carbide substrate using a metal organic chemical vapor deposition process, and a multi-layer nitride functional layer is grown on the aluminum nitride buffer layer using an in-situ annealing and segmented temperature adjustment process; A double-layer passivation layer is formed on the surface of the multi-layer nitride functional layer by combining atomic layer deposition and plasma enhanced treatment; Forming a source metal electrode and a drain metal electrode on the double-layer passivation layer by photolithography and selective ion etching processes, wherein the source metal electrode and the drain metal electrode are multi-layer metal structures treated by vacuum annealing; A digital ion implantation process is used to form a gradient doping channel in a region between the source metal electrode and the drain metal electrode, wherein the gradient doping channel includes a plurality of doping regions whose doping concentration decreases in sequence from the source to the drain; A gate metal electrode that has undergone interface engineering is formed above the gradient doping channel by adopting a self-alignment process. The gate metal electrode adopts a vacuum co-sputtered nickel / gold double-layer metal structure.

2. The method according to claim 1, characterized in that An aluminum nitride buffer layer is grown on a silicon carbide substrate using a metal organic chemical vapor deposition process, and a multi-layer nitride functional layer is grown on the aluminum nitride buffer layer using an in-situ annealing and segmented temperature adjustment process, comprising: performing plasma treatment on a silicon carbide substrate in a mixed atmosphere of hydrogen and nitrogen to form a reconstructed structure on a surface of the silicon carbide substrate; High-purity hydrogen carrier gas is introduced into the surface of the reconstructed silicon carbide substrate, and trimethylaluminum precursor and ammonia reaction gas are alternately introduced in a program-controlled pulse modulation manner, and the growth rate of the aluminum nitride buffer layer is controlled by adjusting the pulse time ratio of the trimethylaluminum precursor and the ammonia reaction gas; A trimethylaluminum precursor, a trimethylgallium precursor, and an ammonia reaction gas are introduced into the surface of the aluminum nitride buffer layer, and a multi-layer nitride functional layer is grown by a stepwise heating method to obtain multiple nitride layers; The multiple nitride layers are subjected to in-situ annealing treatment respectively, wherein the temperature of the in-situ annealing treatment is higher than the growth temperature of the corresponding layer by a set temperature, and a multi-layer nitride functional layer with strain modulation is formed by the in-situ annealing treatment.

3. The method according to claim 1, characterized in that A double-layer passivation layer is formed on the surface of the multi-layer nitride functional layer by combining atomic layer deposition and plasma enhanced treatment, comprising: treating the surface of the multilayer nitride functional layer with oxygen plasma to form hydroxyl surface active groups; Growing a first passivation layer of hafnium oxide on the hydroxyl surface active groups by an atomic layer deposition process, and forming a first hafnium oxide molecular layer by alternately introducing a hafnium tetrachloride precursor and water vapor reaction gas; During the growth of the first hafnium oxide molecular layer, the hafnium tetrachloride precursor and the water vapor reaction gas are activated by oxygen plasma enhanced treatment to form a dense hafnium oxide first passivation layer on the surface of the multi-layer nitride functional layer; A nitrogen carrier gas is introduced into the vacuum chamber, and an aluminum oxide second passivation layer is grown on the surface of the dense hafnium oxide first passivation layer using an atomic layer deposition process. A trimethylaluminum precursor and the water vapor reaction gas are alternately introduced in a program-controlled manner to form an aluminum oxide molecular layer on the surface of the dense hafnium oxide first passivation layer; The dense hafnium oxide first passivation layer and the aluminum oxide molecular layer are subjected to high-temperature annealing in a nitrogen atmosphere to form a double-layer passivation structure with interface bonding.

4. The method according to claim 3, characterized in that During the growth of the first hafnium oxide molecular layer, the hafnium tetrachloride precursor and the water vapor reaction gas are activated by oxygen plasma enhanced treatment to form a dense hafnium oxide first passivation layer on the surface of the multilayer nitride functional layer, including: A hafnium tetrachloride precursor is introduced into the vacuum chamber to allow the hafnium tetrachloride precursor to chemically adsorb to hydroxyl functional groups on the surface of the multilayer nitride functional layer; Using nitrogen purge to remove unreacted hafnium tetrachloride precursor in the vacuum chamber, and forming a first reaction intermediate layer on the surface of the multi-layer nitride functional layer; introducing water vapor reaction gas into the vacuum chamber to allow the water vapor reaction gas to chemically react with the first reaction intermediate layer; generating oxygen plasma in the vacuum chamber using a radio frequency source, and performing an ion bombardment enhancement treatment on the first reactive intermediate layer and the water vapor reactive gas by the oxygen plasma, so as to reconstruct the chemical bond between the first reactive intermediate layer and the water vapor reactive gas; The reaction products in the vacuum chamber are removed by nitrogen purge, and a dense hafnium oxide first passivation layer is formed on the surface of the multi-layer nitride functional layer.

5. The method according to claim 1, wherein Forming a source metal electrode and a drain metal electrode on the double-layer passivation layer by using photolithography and selective ion etching processes, including: Spin-coating a photoresist on the surface of the double-layer passivation layer, and forming a first pattern area and a second pattern area in the photoresist by an exposure and development process; Etching the double-layer passivation layer using an inductively coupled plasma etching device, and forming a step-shaped etching structure between the first pattern area and the second pattern area by adjusting the radio frequency power and bias power of the inductively coupled plasma etching device; sputtering multiple functional metal layers sequentially on the surface of the stepped etched structure, and forming a multi-layer metal structure with stress matching by controlling the thickness ratio of the multiple functional metal layers; The photoresist is removed by a stripping process, a metal electrode pattern is formed between the multi-layer metal structure, the metal electrode pattern is annealed in a vacuum annealing furnace, and a source metal electrode and a drain metal electrode with an ohmic contact structure are formed by controlling the process parameters of the vacuum annealing furnace.

6. The method according to claim 1, characterized in that A digital ion implantation process is used to form a gradient doping channel in a region between the source metal electrode and the drain metal electrode, comprising: A positive photoresist is coated on the area between the source metal electrode and the drain metal electrode, and a doping window array is formed in the positive photoresist by using a laser direct writing lithography process; Performing a first ion implantation in the doping window array region using an ion implantation machine, and forming a high-concentration first doping region near the source metal electrode by controlling the implantation energy, beam current, and implantation dose of the ion implantation machine; Performing programmed ion implantation on the doping window array region using a digitally controlled ion implantation process to form a plurality of doping regions with successively decreasing doping concentrations between the first doping region and the drain metal electrode; Removing the positive photoresist using a plasma stripping process, performing a rapid thermal annealing process on the first doped region and the plurality of doped regions with successively decreasing doping concentrations, and forming a doped region with uniform ion distribution by controlling a temperature curve and an atmosphere composition of the rapid thermal annealing process; The doping region with uniform ion distribution is activated and annealed in a nitrogen atmosphere to form a gradient doping channel.

7. The method according to claim 6, characterized in that A digitally controlled ion implantation process is used to perform programmed ion implantation on the doping window array region to form a plurality of doping regions with successively decreasing doping concentrations between the first doping region and the drain metal electrode, including: Establishing a digital injection coordinate system in the doping window array region, and dividing the doping window array region into a plurality of injection sub-regions; Measuring surface topography parameters of the plurality of implantation sub-regions using an ion beam detector, and calculating an ion implantation compensation coefficient for each implantation sub-region according to the surface topography parameters; Performing programmed ion implantation on the plurality of implantation sub-regions using a digitally controlled ion implantation system, controlling the spatial distribution of the implanted ions in each implantation sub-region by adjusting the voltage of the ion implantation system, and controlling the ion implantation dose of each implantation sub-region by adjusting the implantation process parameters of the ion implantation system in combination with the ion implantation compensation coefficient, so that the implantation doses between adjacent implantation sub-regions decrease gradually according to a set ratio; Using a secondary ion mass spectrometer to measure the ion concentration distribution of the multiple implantation sub-regions to obtain ion concentration distribution measurement results; The voltage of the ion implantation system and the implantation process parameters are modified according to the ion concentration distribution measurement result until the multiple implantation sub-regions form doping regions with successively decreasing doping concentrations.

8. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images