A method for real-time monitoring of a silicon carbide ohmic contact annealing process

Abstract

The application provides a real-time monitoring method for a silicon carbide ohmic contact annealing process, and belongs to the technical field of semiconductor manufacturing; a technical problem to be solved is that the control precision of an ohmic contact metal layer is low and the quality stability is poor in the existing rapid thermal annealing process; a technical solution for solving the technical problem is that: step 1: a reference area and a comparison area are prepared on a wafer; step 2: according to the reference area and the comparison area, a test sample is prepared, micro-morphology images of the metal layer are collected at different temperatures, and feature parameters are extracted through image processing software; the feature parameters and the corresponding metal layer micro-morphology images are stored in a standard atlas database; step 3: the wafer is placed in an RTP thermal annealing device, and a synchronous microscopic imaging system, an atmosphere control system and a multi-stage annealing execution system are started; and step 4: the collected images are compared with the standard atlas database, the temperature of the wafer surface is evaluated, and the process parameters are adjusted.

Description

Technical Field

[0001] This application relates to the field of semiconductor manufacturing technology, and in particular to a real-time monitoring method for the silicon carbide ohmic contact annealing process. Background Technology

[0002] Traditional silicon carbide ohmic contact annealing processes employ rapid thermal annealing equipment, heat-treating the entire wafer using a pre-set single temperature-time profile. Operators set the heating rate, target temperature, and holding time based on prior experimental experience, monitoring only the ambient temperature within the equipment chamber during annealing. After annealing, the wafer is removed, and the ohmic contact quality is evaluated offline using methods such as four-point probe testing or transmission electron microscopy (TEM) cross-sectional observation. This method is commonly used on mass production lines with a fixed process formulation, adjusting the temperature profile only when batch sampling reveals excessive resistivity.

[0003] In the prior art, Chinese invention patent CN121038452A, entitled "A Method for Annealing Temperature Control Based on Differentiated Ohmic Contact Materials," proposes a solution to the problems of inaccurate monitoring of the annealing temperature of the Au / Ge ohmic contact layer on the N-side of AlGaInP-based red Micro LEDs, difficulties in detecting uneven temperature fields, and delayed response to anomalies. This method uses a seven-step process (two patterning steps, two metal coating steps, two resist removal steps, annealing, and observation) to prepare a gold / germanium first ohmic contact layer (for functional use) and a gold / germanium / nickel second ohmic contact layer (for detection) on the N-side of the wafer. Utilizing the property of nickel to inhibit gold / germanium agglomeration, the temperature is determined by observing the difference in agglomeration state between the two layers using an optical microscope (normally, the former agglomerates while the latter does not). This method is convenient for detection, supports batch testing, can detect temperature anomalies in real time, improves process stability, and is suitable for industrial mass production.

[0004] While the above scheme introduces a dual-region design in the fabrication of N-sided ohmic contacts for red Micro-LEDs, its core flaw lies in the lag of the control logic and the singularity of the judgment dimension. The scheme discloses in step S7 a technique of observing the agglomeration state of the two regions after annealing using optical observation equipment, and then determining whether the annealing temperature is normal. This offline detection mode is essentially still a post-event inspection. When an abnormal agglomeration state is observed, the annealing process for the entire batch of wafers has already ended, and no remedial adjustments can be made to the current batch. If both the first and second ohmic contact layers have agglomerated or neither has agglomerated, the operator can only determine that the temperature is too high or too low. However, at this point, the wafer has completed the entire annealing process, and the over-agglomerated or incompletely alloyed ohmic contact layers cannot be recovered, directly leading to a decrease in the yield of that batch of wafers. Furthermore, the judgment criteria of this scheme are based solely on the presence or absence of agglomeration, lacking the collection and analysis of quantitative indicators such as agglomerate particle diameter, growth rate, and distribution density. When the annealing temperature is at the edge of the process window but has not yet triggered extreme agglomeration, operators cannot identify this critical deviation, thus masking potential reliability risks.

[0005] In summary, in traditional annealing processes, operators can only control the process based on preset temperature profiles, making it impossible to observe the microscopic morphological evolution of the metal layer on the wafer surface in real time at high temperatures. In particular, the agglomeration and diffusion behavior of contact materials such as gold and germanium at high temperatures is highly temperature sensitive. Excessive temperature can cause excessive metal agglomeration, resulting in discontinuous contact layers, while insufficient temperature can prevent the formation of effective alloy contacts. Furthermore, different metal combinations, such as the introduction of nickel to suppress agglomeration in the contrast layer, exhibit different temperature responses, making it difficult for a single temperature measurement method to accurately reflect the effective process state of the wafer surface. At the same time, due to the lack of real-time quantitative feedback on key parameters such as the degree of metal agglomeration, including average particle diameter and growth rate, it is impossible to dynamically adjust the heating rate, holding time, or atmosphere ratio to correct process deviations when equipment temperature field fluctuations or batch-to-batch differences occur. Ultimately, this results in large fluctuations in the specific contact resistivity of ohmic contacts and low device yield, severely restricting the consistency and reliability of silicon carbide device performance. Summary of the Invention

[0006] To address the aforementioned technical problems, this application proposes a real-time monitoring method for the silicon carbide ohmic contact annealing process.

[0007] The technical solution adopted in this application is: a real-time monitoring method for the silicon carbide ohmic contact annealing process, comprising the following steps: Step 1: Prepare a test area on the surface of the wafer to be processed, the test area including a reference area and at least one contrast area; Step 2: Based on the test area, prepare multiple sets of test samples, collect microscopic morphology images of the metal layer of the test samples at different annealing temperatures and different holding times, and extract the characteristic parameters of the metal layer microscopic morphology images at different annealing temperatures using image processing software. Store the characteristic parameters and the corresponding metal layer microscopic morphology images together in a standard spectral database. Step 3: Place the wafer in the RTP thermal annealing equipment, and start the synchronous microscopic imaging system, atmosphere control system, and multi-stage annealing execution system. The synchronous microscopic imaging system is used to monitor the microscopic morphology changes of the test area in real time. The atmosphere control system is used to ensure that the atmosphere in the RTP thermal annealing equipment matches the high-temperature reaction and interface evolution characteristics of the metal layer in the test area. The multi-stage annealing execution system is used to perform multi-stage annealing treatment on the wafer. Step 4: Compare the images acquired by the synchronous microscopic imaging system with the data in the standard image database to evaluate the annealing temperature of the wafer surface in real time and adjust the process parameters during the annealing process. The multi-stage annealing process includes a low-temperature stage, a medium-temperature stage, a high-temperature stage, and a rapid cooling stage.

[0008] Furthermore, in step 1, the test area is prepared by photolithography and electron beam evaporation technology.

[0009] Furthermore, the reference region includes a first ohmic contact layer composed of a gold layer and a germanium layer.

[0010] Furthermore, the comparison region includes a second ohmic contact layer composed of a gold layer, a germanium layer, and any one of the metals selected from nickel, aluminum, and molybdenum.

[0011] Furthermore, the comparison region is a second ohmic contact layer composed of a gold layer, a germanium layer, and a nickel layer, wherein the thickness of the gold layer is 30-50 nm, the thickness of the germanium layer is 10-20 nm, and the thickness of the nickel layer is 5-10 nm.

[0012] Furthermore, in step 2, the annealing temperature ranges from 300°C to 850°C.

[0013] Furthermore, the characteristic parameters include the average diameter of the metal particles, growth rate, distribution density, and time of agglomeration.

[0014] Furthermore, the multi-stage annealing process specifically includes the following steps: Step 3.1: Enter the low-temperature stage, heat the wafer to 300℃~350℃, and continue for 5min~10min; Step 3.2: Enter the intermediate temperature stage, heat the wafer to 500℃~600℃, and continue for 3min~5min; Step 3.3: Enter the high-temperature stage and heat the wafer to 800℃~850℃ for 2min~3min; Step 3.4: Enter the rapid cooling stage, cool the wafer to 20℃~30℃, and continue cooling for 1min~2min.

[0015] Furthermore, the atmosphere includes nitrogen, hydrogen, and argon.

[0016] Furthermore, the concentration of nitrogen is 90% to 98%, the concentration of hydrogen is 1% to 5%, and the concentration of argon is 1% to 2%.

[0017] The advantages of this application compared to the prior art are as follows: 1. This application prepares a nickel-free gold-germanium reference layer and a nickel-containing gold-germanium-nickel contrast layer on the same wafer. Utilizing the inhibitory properties of nickel on metal agglomeration, a difference in temperature sensitivity is created. The difference in agglomeration behavior between the two regions under the same annealing environment is used as the basis for real-time judgment of the effective temperature of the wafer surface: agglomeration in the reference layer while the contrast layer remains uniform indicates a suitable temperature; rapid agglomeration in the contrast layer indicates an excessively high temperature. This built-in reference directly reflects the reaction temperature of the metal / semiconductor interface, overcoming the limitations of thermocouple temperature measurement due to interference from cavity airflow and radiation. 2. This application utilizes multi-stage annealing processes at low temperatures (300-350℃), medium temperatures (500-600℃), and high temperatures (800-850℃) with multiple metal combinations of gold-germanium and gold-germanium-nickel. This allows for the collection of data on the average diameter, growth rate, and distribution density of metal particles at different temperatures, establishing a temperature-agglomeration state correlation. This quantitative relationship provides data support for the narrow process window of silicon carbide ohmic contacts, enabling precise setting of target temperatures for each stage, replacing the traditional trial-and-error method. 3. This application integrates a synchronous microscopic imaging system into the RTP thermal annealing equipment to directly observe the time of agglomeration points, average particle diameter, and growth rate of metal layers such as gold and germanium on the wafer surface at high temperatures. By comparing real-time images with preset standard spectra in a standard spectra database, the system can identify the degree of deviation between the actual agglomeration state and the process window. When the monitored data exceeds a threshold, such as when the particle diameter in the gold, germanium, and nickel region exceeds 60 nm, the control system immediately adjusts the heating rate or holding time to avoid contact failures caused by insufficient annealing or excessive agglomeration in the entire batch of wafers, thereby improving the consistency of the single-batch process. Attached Figure Description

[0018] The following description, in conjunction with the accompanying drawings, further illustrates this application: Figure 1 This is a schematic diagram of the method flow of this application; Figure 2 This is a flowchart illustrating the multi-stage annealing process in this application; Figure 3 This is a flowchart illustrating step 4 of this application; Figure 4 This is a temperature curve of the multi-stage annealing process in this application; Figure 5 This is a schematic diagram showing the change of metal agglomeration state with temperature in this application. Detailed Implementation

[0019] like Figures 1 to 5 As shown, this application provides a real-time monitoring method for the silicon carbide ohmic contact annealing process, including the following steps: Step 1: Using photolithography and electron beam evaporation, a test area is prepared on the surface of the silicon carbide wafer to be processed. The test area includes a reference area and at least one contrast area. The reference area includes a first ohmic contact layer composed of a gold layer and a germanium layer. The contrast area includes a second ohmic contact layer composed of a gold layer, a germanium layer, and any one of the metal layers selected from nickel, aluminum, and molybdenum. When the contrast area is a second ohmic contact layer composed of a gold layer, a germanium layer, and a nickel layer, the thickness of the gold layer is 30-50 nm, the thickness of the germanium layer is 10-20 nm, and the thickness of the nickel layer is 5-10 nm. Step 2: Based on the test area in Step 1, and using the same metal layer combination and film thickness as in Step 1, prepare multiple independent test samples. At different annealing temperatures (annealing temperature range covering 300℃~850℃, with a collection node set every 20℃~50℃), collect microscopic morphology images of the metal layer of the test samples at different holding times. Extract characteristic parameters (including the average diameter, growth rate, distribution density, and agglomeration point appearance time of metal particles) from the metal layer microscopic morphology images at different annealing temperatures using image processing software. Store the characteristic parameters and the corresponding metal layer microscopic morphology images together in a standard spectral database. The standard spectral database is used to represent the correspondence between annealing temperature and metal agglomeration state. Step 3: Place the wafer in the RTP thermal annealing equipment and activate the synchronous microscopic imaging system, atmosphere control system, and multi-stage annealing execution system. The synchronous microscopic imaging system is used to monitor the microscopic morphology changes of the test area in real time, focusing on: the appearance time, growth rate, and final morphology of metal agglomerates, and comparing the dynamic changes of the reference and control areas at different annealing stages. The atmosphere control system is used to ensure that the atmosphere in the RTP thermal annealing equipment (a mixture of nitrogen, hydrogen, and argon, with nitrogen concentration of 90%~98%, hydrogen concentration of 1%~5%, and argon concentration of 1%~2%) matches the high-temperature reaction and interface evolution characteristics of the metal layer in the test area. The multi-stage annealing execution system is used to perform multi-stage annealing on the wafer, specifically including the following steps: Step 3.1: Enter the low temperature stage, heat the wafer to 300℃~350℃, and hold for 5min~10min; Step 3.2: Enter the medium temperature stage, heat the wafer to 500℃~600℃, and continue for 3min~5min; Step 3.3: Enter the high-temperature stage, heat the wafer to 800℃~850℃, and continue for 2min~3min; Step 3.4: Enter the rapid cooling stage, cool the wafer to 20℃~30℃, and continue cooling for 1min~2min; Step 4: Compare the images acquired by the synchronous microscopic imaging system with the data in the standard image database to evaluate the annealing temperature of the wafer surface in real time and adjust the process parameters during the annealing process. In the embodiments of this application, nitrogen is used as the main atmosphere, hydrogen is used to regulate the diffusion rate of the metal and inhibit the agglomeration rate, and argon is used as an inert gas to maintain the stability of the system and control the concentration ratio of the atmosphere. By adjusting the atmosphere, the diffusion rate and agglomeration behavior of the metal layer are controlled, reducing the oxidation reaction that may occur during the multi-stage annealing process.

[0020] The addition of a nickel layer to the comparative region aims to regulate the diffusion behavior of the metal layer by leveraging its unique ability to inhibit metal agglomeration, thereby studying its impact on annealing temperature sensitivity. The addition of aluminum or molybdenum is commonly used to improve the stability of the metal layer, especially at higher temperatures, where they inhibit diffusion behavior. Other possible alloy combinations can also be used to prepare the comparative region, as long as different metal combinations can be added to simulate various annealing process conditions; this application does not impose specific limitations on these combinations. In practical work, the amount of metal vapor deposition should be strictly controlled to ensure that the effects of different metal combinations can be compared under the same experimental conditions.

[0021] In step 3, the temperature and holding time were adjusted at different annealing stages to test the diffusion and aggregation characteristics of the metal layer under different temperature conditions. For example, at low temperatures (300-350℃), the diffusion and aggregation of the metal layer were relatively slow, mainly aimed at removing the oxide layer and wafer surface stress; at medium temperatures (500-600℃), the metal layer began to undergo initial diffusion and aggregation, and excessive aggregation was avoided through temperature control; while at high temperatures (800-850℃), the aggregation behavior of the metal layer was most significant, especially the formation of gold, germanium, and nickel alloy layers. At high temperatures, the inhibitory effect of different alloys on metal aggregation and their temperature sensitivity were further studied by adjusting the combination of different metal layers. Under different metal combinations and annealing conditions, the diffusion and aggregation behavior of the metal layer was monitored in real time, and the changes in the microstructure of the metal layer at each stage were captured using a synchronous microscopic imaging system.

[0022] Example 1: Using gold and germanium as reference areas and gold, germanium, and nickel as comparison areas, the specific process of the multi-stage annealing treatment in step 3 is explained.

[0023] Phase 1 (Low Temperature Phase): The wafer is heated from room temperature to 300℃-350℃ for 5-10 minutes. This phase is used to remove the oxide layer and stress on the wafer surface, while preventing premature diffusion or agglomeration of the metal layer. With a target temperature of 350℃ and a holding time of 5 minutes, the diffusion rate of gold, germanium, and gold, germanium, and nickel combinations is low, and the structure of the metal layer has not yet undergone significant changes. Adding nickel to the contrast region in Step 1 will make the metal layer in the contrast region more stable in the initial stage, preventing premature agglomeration of the metal layer in the contrast region.

[0024] The second stage (intermediate temperature stage): The wafer is heated to 500-600℃ and maintained for 3-5 minutes. At this stage, the metal layer begins to diffuse and aggregate initially. The target temperature is set to 550℃ and held for 4 minutes. The metal diffusion rate accelerates at this point. The combination of gold and germanium undergoes slight diffusion at lower temperatures, but the addition of nickel inhibits excessive diffusion and aggregation, ensuring the metal layer in the contrast region remains relatively uniform. Simultaneous microscopic imaging reveals initial diffusion in the metal layer, particularly noticeable in the gold-germanium reference region, while the diffusion in the nickel-added contrast region is slower.

[0025] The third stage (high-temperature stage): The wafer is rapidly heated to 800-850℃ and held for 2-3 minutes. The aggregation behavior of the metal layer is significantly enhanced at this stage. The target temperature is set to 820℃ and held for 2 minutes. During this stage, significant diffusion and aggregation of the metal layer occur in both the reference and control regions. Significant aggregation begins in the reference region; however, nickel effectively suppresses the rate and extent of metal aggregation, resulting in slower and more uniform diffusion of the metal layer in the control region.

[0026] Phase Four (Rapid Cooling): After the high-temperature phase, the wafer will be rapidly cooled to room temperature, with the cooling time controlled within 1-2 minutes. Rapid cooling solidifies the final morphology of the metal layer and prevents excessive agglomeration or morphological changes. At this point, the agglomeration state of the metal layer should be stable and will not undergo significant changes. Controlling the cooling process is crucial for ensuring the stability and reliability of the metal contact layer, especially since the agglomeration behavior of the metal layer in the contrast region is more stable than that in the reference region.

[0027] Example 2: Taking a comparative region composed of gold, germanium, and nickel layers as an example, this example details how to quantify the relationship between metal agglomeration state and temperature based on data acquired by a synchronous microscopic imaging system.

[0028] The basic data for the comparison area are as follows: the thickness of the gold layer is 30nm~50nm, the thickness of the germanium layer is 10nm~20nm, and the thickness of the nickel layer is 5nm~10nm; the target temperature for the low-temperature stage is set at 350℃, the target temperature for the medium-temperature stage is 550℃, the critical temperature for the high-temperature stage is 750℃, and the peak process temperature for the high-temperature stage is 820℃.

[0029] Timing begins during the multi-stage annealing process. A synchronous microscopic imaging system is activated to continuously photograph the test area and observe it frame by frame. The earliest time from the uniform state of the metal film to the appearance of the first obvious agglomeration particle can be observed. This time point is the time when the agglomeration point appears. After monitoring and recording, it can be found that when the target temperature of the low-temperature stage is 350℃, the final average diameter is 20nm, the initial diameter is 10nm, the net growth is 10nm, and the actual growth time is 5min.

[0030] Particle growth rate = (final average diameter - initial diameter) / growth time; The calculated growth rate is 2 nm / min.

[0031] Similarly, we can conclude that: when the target temperature in the medium-temperature stage is 550℃, the growth rate is 5nm / min; when the critical temperature in the high-temperature stage is 750℃, the growth rate is 10nm / min; and when the peak process temperature in the high-temperature stage is 820℃, the growth rate is 15nm / min.

[0032] Based on the above theoretical derivation, aggregation data at different time points were obtained and added to the standard spectral database. Some parameters are shown in the table below.

[0033] Table 1. Reunion data at different time points.

[0034]

[0035] Here, particle growth rate refers to the growth rate of metal particles per unit time; particle distribution refers to the distribution of metal particles during the annealing process (uniform or aggregated). The agglomeration state during the actual multi-stage annealing process is then compared in real-time with data from the standard spectral database. When the temperature is set to 750℃, if the average diameter and growth rate of the metal particles exceed the expected range—for example, if the average diameter exceeds 60nm and the growth rate exceeds 10nm / min—the temperature or holding time during the annealing process can be adjusted immediately.

[0036] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method of real-time monitoring of a silicon carbide ohmic contact annealing process, comprising: Includes the following steps: Step 1: Prepare a test area on the surface of the wafer to be processed, the test area including a reference area and at least one contrast area; Step 2: Based on the test area, prepare multiple sets of test samples, collect microscopic morphology images of the metal layer of the test samples at different annealing temperatures and different holding times, and extract the characteristic parameters of the metal layer microscopic morphology images at different annealing temperatures using image processing software. Store the characteristic parameters and the corresponding metal layer microscopic morphology images together in a standard spectral database. Step 3: Place the wafer in the RTP thermal annealing equipment, and start the synchronous microscopic imaging system, atmosphere control system, and multi-stage annealing execution system. The synchronous microscopic imaging system is used to monitor the microscopic morphology changes of the test area in real time. The atmosphere control system is used to ensure that the atmosphere in the RTP thermal annealing equipment matches the high-temperature reaction and interface evolution characteristics of the metal layer in the test area. The multi-stage annealing execution system is used to perform multi-stage annealing treatment on the wafer. Step 4: Compare the images acquired by the synchronous microscopic imaging system with the data in the standard image database to evaluate the annealing temperature of the wafer surface in real time and adjust the process parameters during the annealing process. The multi-stage annealing process includes a low-temperature stage, a medium-temperature stage, a high-temperature stage, and a rapid cooling stage.

2. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 1, characterized in that: In step 1, the test area is prepared by photolithography and electron beam evaporation.

3. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 1, characterized in that: The reference region includes a first ohmic contact layer consisting of a gold layer and a germanium layer.

4. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 3, characterized in that: The contrast region includes a second ohmic contact layer composed of a gold layer, a germanium layer, and any one of the following metals: nickel, aluminum, and molybdenum.

5. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 4, characterized in that: The comparison region is a second ohmic contact layer composed of a gold layer, a germanium layer, and a nickel layer, wherein the thickness of the gold layer is 30-50 nm, the thickness of the germanium layer is 10-20 nm, and the thickness of the nickel layer is 5-10 nm.

6. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 1, characterized in that: In step 2, the annealing temperature ranges from 300°C to 850°C.

7. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 1, characterized in that: The characteristic parameters include the average diameter of the metal particles, growth rate, distribution density, and time of agglomeration.

8. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 1, characterized in that: The multi-stage annealing process specifically includes the following steps: Step 3.1: Enter the low-temperature stage, heat the wafer to 300℃~350℃, and continue for 5min~10min; Step 3.2: Enter the intermediate temperature stage, heat the wafer to 500℃~600℃, and continue for 3min~5min; Step 3.3: Enter the high-temperature stage and heat the wafer to 800℃~850℃ for 2min~3min; Step 3.4: Enter the rapid cooling stage, cool the wafer to 20℃~30℃, and continue cooling for 1min~2min.

9. The real-time monitoring method for the silicon carbide ohmic contact annealing process according to claim 1, characterized in that: The atmosphere includes nitrogen, hydrogen, and argon.

10. A real-time monitoring method for a silicon carbide ohmic contact annealing process according to claim 9, characterized in that: The concentration of nitrogen is 90% to 98%, the concentration of hydrogen is 1% to 5%, and the concentration of argon is 1% to 2%.

Images