High-efficiency protection method for silicon wafer edge-removed back oxide layer

Abstract

The invention discloses an efficient protection method for a silicon wafer edge-removed back oxide layer, and particularly relates to the technical field of semiconductor processing. A chemical vapor deposition reaction chamber sets the edge pressure to be higher than the center through a multi-inlet pressure valve to form a concentration gradient, a shape memory film containing magnetic particles is arranged on the edge of a silicon wafer, and the side edge is wrapped through magnetic field heating; introducing the front surface of the silicon wafer into a microfluidic channel, injecting acid, applying an electric field, soaking in supercritical carbon dioxide, regulating temperature and pressure to form a homogeneous system, and reducing pressure until the humidity is lower than a threshold value. According to the efficient protection method for the silicon wafer edge-removed back oxide layer, the gas concentration gradient is constructed through chemical vapor deposition, the thickness of the oxide layer is monitored, the silicon wafer is fixed and sealed through a magnetic-thermal composite film, corrosion is restrained through a micro-fluidic electric field, accurate edge removal is conducted, and supercritical fluid replacement and reduced-pressure gasification drying are conducted, so that the silicon wafer edge-removed back oxide layer is obtained. And the edge processing precision, reliability and efficiency of the silicon wafer are improved through multi-step cooperation.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor processing, and in particular to a method for efficiently protecting an oxide layer on the back side of a silicon wafer after edge removal. Background Art

[0002] The field of semiconductor processing technology includes several important links such as silicon wafer manufacturing, photolithography, etching, thin film deposition, doping, packaging and testing. Among them, silicon wafer manufacturing, as the cornerstone of the semiconductor industry, is the first key step in the production of semiconductor devices. Its quality directly determines the performance and reliability of subsequent chips. In the silicon wafer manufacturing process, high-purity silicon raw materials are converted into single-crystal silicon ingots through crystal growth; then slicing is performed to cut the silicon ingots into thin slices with precise thickness; then grinding and polishing processes are carried out to obtain silicon wafers with extremely high surface flatness and extremely low roughness. Each step has strict requirements on process parameters, equipment accuracy and environmental conditions. Any slight deviation may affect the final quality of the silicon wafer, and thus affect the overall performance of the semiconductor device.

[0003] The highly efficient method for protecting the oxide layer on the back side of silicon wafers during edge removal is designed to address the current inability of existing technologies to effectively prevent damage to the oxide layer on the chamfered back side of the silicon wafer when removing excess silicon wafer edges during the edge removal process in semiconductor silicon wafer manufacturing. This method involves designing a specific process sequence, pre-treating the silicon wafer, and covering the chamfered back side of the silicon wafer with a special protective material during the edge removal process. Furthermore, the method adjusts the tool parameters, motion trajectory, and processing speed of the edge removal equipment, and controls environmental factors such as temperature and pressure during the process to effectively protect the oxide layer on the chamfered back side of the silicon wafer.

[0004] Existing technologies struggle to effectively protect the oxide layer on the back chamfered surface during the silicon wafer edge removal process. The protective material's lamination accuracy and sealing properties are insufficient, making it unable to fully withstand the mechanical forces or chemical corrosion of the edge removal process. For example, the tool's trajectory and processing speed have a narrow adjustment range, making it easy for the oxide layer to break due to contact or vibration. Inaccurate control of environmental factors such as temperature and pressure can cause the oxide layer to crack or fall off due to sudden changes. The lack of a real-time, precise monitoring and dynamic control mechanism for the oxide layer's thickness distribution prevents differentiated control during the pre-processing phase, resulting in insufficient precision and uniformity in subsequent edge removal. This impacts silicon wafer quality and device performance stability, leading to performance variations and reduced yield. Summary of the Invention

[0005] The main purpose of the present invention is to provide a method for efficiently protecting the oxide layer on the back side of a silicon wafer after edge removal, which can effectively solve the problems involved in the above-mentioned background technology.

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

[0007] A method for efficiently protecting the back oxide layer of a silicon wafer by edge removal comprises the following steps:

[0008] S1: In a chemical vapor deposition reaction chamber, multiple independent gas inlets are equipped with pressure valves, and the edge inlet pressure is set higher than the center inlet pressure, driving the reaction gas to diffuse from the edge to the center to form a concentration gradient;

[0009] S2: A shape memory polymer sealing film containing magnetic particles is placed on the edge of the silicon wafer. A magnetic field is applied so that the magnetic particles are forced to drive the sealing film to initially conform to the edge contour. Heat is then applied to allow the sealing film to return to the preset shape and wrap around the side of the silicon wafer.

[0010] S3: Place the silicon wafer face up in the microfluidic channel, inject weak hydrofluoric acid solution, set electrodes on both sides of the channel and apply an electric field to make the ions in the solution move along the channel constraint direction;

[0011] S4: Immerse the silicon wafer in supercritical carbon dioxide fluid, and adjust the pressure and temperature to allow the fluid to mix with the surface and interstitial liquid of the silicon wafer to form a homogeneous system;

[0012] S5: The pressure of the processing chamber is reduced to allow the supercritical carbon dioxide to gasify and escape. The humidity of the gas in the chamber is monitored by a humidity sensor. When the humidity is lower than a set threshold, the decompression is stopped.

[0013] Preferably, the difference between the edge inlet pressure and the center inlet pressure in S1 is 0.5-2.0 MPa, and the oxide layer thickness of at least 9 evenly distributed areas on the silicon wafer surface is monitored in real time by a laser thickness gauge.

[0014] Preferably, the magnetic particle content of the shape memory polymer sealing film in S2 is 15-35 wt %, the heating temperature is 60-120° C., the heating rate is 5-15° C. / min, and the sealing film laminating pressure is 0.05-0.2 MPa.

[0015] Preferably, the concentration of the weak hydrofluoric acid solution in S3 is 0.5-2.0 mol / L, the width of the microfluidic channel is 0.5-2 mm, the electrode spacing is 5-20 mm, and the oxide layer thickness in the 0.5-3 mm area at the edge of the silicon wafer is monitored by a spectroscopic reflectometer.

[0016] Preferably, the pressure adjustment range of the supercritical carbon dioxide fluid in S4 is 7.3-15 MPa, the temperature adjustment range is 31-80°C, the fluid flow rate is 0.5-3 L / min, the replacement time is 5-20 min, and the deviation of the fluid replacement uniformity parameter value does not exceed ±8%.

[0017] Preferably, the pressure of the processing chamber in S5 is reduced to 0.1-0.5 MPa, the decompression rate is 0.05-0.2 MPa / min, and the residual liquid value on the silicon wafer surface does not exceed 0.1 μg / cm2 .

[0018] Preferably, the shape memory polymer sealing film in S2 has a glass transition temperature of 40 to 60° C. and a thermal expansion coefficient of (50 to 150)×10-6 / ° C.

[0019] Preferably, in said S3, the electrode voltage is adjusted so that the ion migration rate control parameter value is within the range of 10 to 50 μm / min, and the rate regulation within the range of ±15% can be achieved.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] 1. The present invention sets multiple independent gas inlets and configures pressure valves in a chemical vapor deposition reaction chamber, sets the edge inlet pressure higher than the center inlet pressure, drives the reaction gas to diffuse and form a concentration gradient, and simultaneously uses a laser thickness gauge to monitor the oxide layer thickness in at least 9 evenly distributed areas on the silicon wafer surface in real time. The pressure valve is fine-tuned based on the data to achieve precise control of the oxide layer thickness distribution on the silicon wafer surface, so that the oxide layer thickness meets the requirements of subsequent processes, avoids performance differences of silicon wafers due to uneven thickness, and improves the quality and consistency of silicon wafers.

[0022] 2. The present invention uses a shape memory polymer sealing film containing magnetic particles. With the help of a magnetic field, the magnetic particles are subjected to force to drive the sealing film to initially fit the edge contour. By heating, the sealing film is restored to a preset shape and wraps the side of the silicon wafer. At the same time, parameters such as the magnetic particle content, heating temperature, heating rate and sealing film fitting pressure are controlled to achieve rapid and accurate fixation and sealing of the silicon wafer, enhance the sealing and fixing stability, and effectively prevent the contamination of non-processed areas of the silicon wafer by liquids, gases, etc. during subsequent processing.

[0023] 3. The present invention places the silicon wafer face up in a microfluidic channel, injects a weak hydrofluoric acid solution, and sets electrodes on both sides of the channel to apply an electric field, so that ions in the solution move along the channel constraint direction. The electrode voltage is adjusted by combining the microfluidic channel width, weak hydrofluoric acid solution concentration, electrode spacing and other parameter settings and monitoring the thickness of the oxide layer at the edge of the silicon wafer with a spectroscopic reflectometer to achieve high-precision removal of the oxide layer at the edge of the silicon wafer, accurately control the removal thickness and area, and avoid corrosion of non-edge areas.

[0024] 4. The present invention immerses the silicon wafer in a supercritical carbon dioxide fluid, adjusts the pressure and temperature to mix the fluid with the liquid on the surface and in the gaps of the silicon wafer to form a homogeneous system, controls parameters such as pressure, temperature, fluid flow rate and replacement time, and efficiently and thoroughly removes residual liquid on the surface and in the gaps of the silicon wafer. The supercritical carbon dioxide is then vaporized and released by reducing the pressure in the processing chamber. A humidity sensor is used to monitor the humidity of the gas in the chamber to control the decompression operation, quickly and effectively remove residual fluid and moisture, and achieve complete drying of the silicon wafer. Each step works synergistically through specific technical means and parameter control, significantly improving the accuracy, reliability and efficiency of silicon wafer edge processing, and meeting the high quality requirements of silicon wafers in fields such as semiconductor manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is the overall process flow chart of the present invention. DETAILED DESCRIPTION

[0026] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0027] like Figure 1 As shown, the present invention mainly discloses a silicon wafer edge treatment method. The method pre-treats the silicon wafer by establishing a gas concentration gradient in a chemical vapor deposition (CVD) reaction chamber. The silicon wafer is then fixed and sealed by a magnetic thermal composite bonding method. Microfluidic electric field confined etching technology is then used for precise edge removal. The silicon wafer is then removed using a supercritical fluid displacement method. Finally, the silicon wafer is dried by decompression gasification. These steps work together to achieve efficient treatment of the silicon wafer edge.

[0028] Through the synergistic effect of multiple steps, this method can effectively improve the accuracy, reliability and efficiency of silicon wafer edge processing. Compared with traditional methods, it can better meet the high-quality requirements of silicon wafers in fields such as semiconductor manufacturing.

[0029] The specific implementation of the present invention will be described in detail below with reference to specific embodiments;

[0030] Example 1: This example pre-treats the silicon wafer based on S1. In the specific implementation process, this step is performed in a CVD reaction chamber. The CVD reaction chamber is equipped with multiple independent gas inlets, each of which is equipped with a pressure valve. These pressure valves are used to accurately control the inflow pressure of the gas.

[0031] In a confined space, pressure differences drive fluid flow. When the edge inlet pressure of the CVD reaction chamber is set higher than the center inlet pressure, the reactant gas will spontaneously diffuse from the high-pressure edge area to the low-pressure center area under the action of the pressure difference.

[0032] At the same time, during the diffusion process, the concentration of the gas changes with distance and diffusion time, forming a concentration gradient within the chamber. This concentration gradient can cause different areas of the silicon wafer surface to be exposed to different amounts of reactive gas, thereby affecting the growth rate and thickness of the oxide layer on the silicon wafer surface, thereby achieving differentiated regulation of the oxide layer on the silicon wafer surface.

[0033] In actual production, the pressure difference between the edge inlet and the center inlet should be within the range of 0.5 to 2.0 MPa. When the pressure difference is small, the gas diffusion rate is relatively slow and the concentration gradient changes more gently, which is suitable for process scenarios with high requirements for oxide layer thickness uniformity and slow growth rate.

[0034] When the pressure difference is large, the gas diffusion speed is accelerated and the concentration gradient is more significant, which can quickly form differentiated oxide layer growth conditions. It is suitable for situations where the oxide layer thickness distribution needs to be quickly adjusted.

[0035] In this embodiment, the difference between the edge inlet pressure and the center inlet pressure is set to 1.2 MPa. This value is in the middle of the value range, which can not only ensure that the gas forms a stable concentration gradient, but also make the oxide layer thickness distribution vary within a reasonable range, and has good process applicability and representativeness.

[0036] Furthermore, to ensure that the concentration gradient meets the process requirements, a gas concentration sensor is installed in the chamber. The sensor monitors the reaction gas concentration in different areas in real time and feeds the data back to the control system. The control system fine-tunes the pressure valve based on the feedback data to maintain a stable concentration gradient.

[0037] At the same time, a laser thickness gauge is used to monitor the thickness of the oxide layer on the surface of the silicon wafer in real time. During the specific implementation process, the laser thickness gauge measures the oxide layer thickness in nine evenly distributed areas on the surface of the silicon wafer in real time;

[0038] Compare the measured thickness data of each area with the pre-set target thickness, and calculate the deviation between the thickness of each area and the target thickness;

[0039] If the target thickness of a certain area is 500nm and the actual measured thickness is 510nm, the deviation value of this area is 10nm. The control system accurately adjusts the inlet pressure of the corresponding area based on these deviation values, continuously optimizes the growth of the oxide layer, and finally obtains the oxide layer thickness deviation distribution value to meet the requirements of silicon wafer pretreatment.

[0040] Pre-treatment of the silicon wafer can precisely control the thickness distribution of the oxide layer on the surface of the silicon wafer, making the oxide layer thickness more consistent with subsequent process requirements, avoiding differences in silicon wafer performance caused by uneven oxide layer thickness, and improving the quality and consistency of the silicon wafer.

[0041] In this embodiment, the pressure difference setting of 1.2 MPa enables the thickness deviation of the oxide layer on the surface of the silicon wafer to be effectively controlled within the process allowable range, ensuring the stability of subsequent semiconductor device manufacturing, realizing effective regulation of the oxide layer thickness distribution, and meeting the requirements for silicon wafer pretreatment in different process scenarios.

[0042] Example 2: Based on Example 1, this example further performs fixing and sealing operations on the silicon wafer;

[0043] Specifically, a shape memory polymer sealing film containing magnetic particles is selected, and the content of the magnetic particles can be selected within the range of 15 to 35 wt%;

[0044] When the content of magnetic particles is low, the magnetic force on the sealing film under the action of the magnetic field is relatively small, and the initial bonding effect may be weak, but it is helpful to reduce the impact of magnetic particles on other properties of the sealing film;

[0045] Higher magnetic particle contents increase the magnetic force, allowing for a faster and tighter initial fit, but may affect the flexibility and shape memory properties of the sealing film.

[0046] In this embodiment, the magnetic particle content of the sealing film is set to 25wt%. At this content, it can not only ensure that the sealing film achieves good initial bonding under the action of the magnetic field, but also maintain its good shape memory performance and flexibility, and has better comprehensive performance in practical applications.

[0047] According to the size of the silicon wafer, the sealing film is cut into appropriate size and pre-processed to put it in an initial state that is easy to operate.

[0048] Shape memory polymer is a polymer material with special properties. Under certain conditions, it can "remember" its initial shape. When an external temperature change is applied to it, it will return to the preset shape. The magnetic particles contained in the sealing film will be affected by magnetic force under the action of a magnetic field. The shape memory polymer sealing film selected by the present invention has a glass transition temperature of 40-60°C and a thermal expansion coefficient of (50-150)×10~6 / °C.

[0049] Based on this, the force of the magnetic field on the magnetic particles is first used to make the sealing film initially fit the edge contour of the silicon wafer to achieve initial fixation. Then, by heating, the shape memory polymer reaches its phase transition temperature, triggering the conformational change of the polymer molecular chain, prompting the sealing film to restore the preset shape and tightly wrap the side of the silicon wafer, thereby achieving sealing of the silicon wafer.

[0050] In actual operation, lower magnetic field strengths are suitable for sealing film materials with low magnetic field sensitivity or scenarios with strict restrictions on the magnetic impact on silicon wafers. Magnetic field strengths below 0.3T are no longer able to fix the edge contour of the silicon wafer. Higher magnetic field strengths can provide stronger magnetic force and speed up the initial bonding of the sealing film, but may have a certain magnetic impact on the silicon wafer. When the magnetic field strength exceeds 1.2T, the magnetic impact will affect the performance of the silicon wafer and the quality of the final product cannot be guaranteed.

[0051] In this embodiment, a magnetic field of 0.8 T is applied to ensure effective initial bonding of the sealing film while controlling the magnetic effect on the silicon wafer within an acceptable range.

[0052] When the temperature is low, the shape memory polymer recovers its shape more slowly, and it may take a long time to complete the sealing. The minimum sealing temperature must not be lower than 60°C. When the temperature is high, although the recovery speed is faster, it may have an adverse effect on the performance of the sealing film and even damage the silicon wafer. In the actual process of this method, the maximum sealing temperature does not exceed 120°C.

[0053] In this embodiment, the heating temperature is set to 90°C. At this temperature, the sealing film can quickly and stably restore the preset shape within a reasonable time, achieving a good sealing effect;

[0054] The heating rate is 10°C / min, which helps to evenly heat the sealing film and avoid affecting the sealing effect due to local excessively high or low temperatures;

[0055] During the sealing process, a pressure sensor is used to monitor the sealing film bonding pressure in real time. The bonding pressure is usually maintained at 0.05-0.2 MPa. In this embodiment, the sealing film bonding pressure is maintained at 0.1 MPa. If the pressure is abnormal, it can be corrected by adjusting the heating temperature and time to finally obtain the sealing film bonding pressure fluctuation value. The bonding pressure is dynamically adjusted according to the actual airtightness test results. Changes in the workshop temperature and humidity will affect the sealing effect.

[0056] This magnetic thermal composite bonding method can achieve rapid and precise fixation and sealing of silicon wafers. Compared with traditional sealing methods, it has better sealing performance and more stable fixation of silicon wafers. Under the parameter settings of this embodiment, the sealing film can fit tightly to the side of the silicon wafer, effectively preventing liquids, gases and other substances from contaminating the non-processing areas of the silicon wafer during subsequent processing.

[0057] Example 3: Based on the fixation and sealing of the silicon wafer completed in Example 2, this example further performs precise edge removal on the silicon wafer. The silicon wafer is placed face up in a microfluidic channel, a weak hydrofluoric acid solution is injected, electrodes are set on both sides of the channel, and an electric field is applied to cause ions in the solution to move along the channel constraint direction.

[0058] Specifically, the silicon wafer is placed face up and steadily in the microfluidic channel. The width of the microfluidic channel is selected according to the processing accuracy requirements. A narrower channel width can make the flow of the solution in the channel more concentrated and the corrosion effect more precise; a wider channel width has a larger solution flow rate and a more uniform distribution, but the corrosion accuracy may decrease. During the processing of this method, when the width of the microfluidic channel is less than 0.5mm, the flow rate of the solution is too small and the distribution is uneven, affecting the corrosion effect. When the width of the microfluidic channel is greater than 2mm, although the flow rate increases, its distribution range and corrosion efficiency are relatively high, difficult to control, and the accuracy is relatively low.

[0059] A weak hydrofluoric acid solution with a concentration of 1.2 mol / L is slowly injected into the microfluidic channel through a microfluidic pump to ensure that the solution evenly covers the area to be treated at the edge of the silicon wafer. The concentration of the weak hydrofluoric acid solution can be selected within the range of 0.5 to 2.0 mol / L.

[0060] When the concentration is less than 0.5 mol / L, the corrosion rate is slow and the relative processing time will be longer; when the concentration is higher, the corrosion rate is accelerated, but when the concentration is greater than 2.0 mol / L, the risk of corrosion of non-target areas of the silicon wafer will increase. The concentration of 1.2 mol / L selected in this embodiment can better control the scope and degree of corrosion while ensuring a certain corrosion efficiency.

[0061] Furthermore, microfluidics is a technology that precisely controls and manipulates fluids within micron-scale channels. In this step, the precise size and structure of the microfluidic channels enable precise delivery and distribution control of the weak hydrofluoric acid solution.

[0062] At the same time, according to the principle of the effect of electric field on charged particles, after setting electrodes on both sides of the channel and applying an electric field, the ions in the solution will move along the direction of the electric field under the action of the electric field force, that is, move along the channel constraint direction;

[0063] This directional movement of ions can make the corrosion effect of the solution more concentrated and controllable, thereby achieving precise corrosion of the edge of the silicon wafer. By monitoring the thickness of the front oxide layer of the silicon wafer in real time through a spectroscopic reflectometer and adjusting the electrode voltage of the corresponding area according to the thickness data, the ion migration rate can be further accurately controlled, thereby achieving precise regulation of the degree of corrosion.

[0064] In this embodiment, the electrode spacing is 12 mm, and a 30V electric field is applied. Under these parameters, a stable ion migration environment can be formed, achieving efficient and precise etching of the oxide layer at the edge of the silicon wafer;

[0065] A spectroscopic reflectometer is used to monitor the thickness of the oxide layer on the front side of the silicon wafer in real time. During the specific implementation process, the spectroscopic reflectometer monitors the oxide layer thickness in the 2mm area around the edge of the silicon wafer. Based on the monitored thickness data, the corrosion situation in different areas of the silicon wafer is analyzed. The control system adjusts the electrode voltage in the corresponding area to accurately control the ion migration rate.

[0066] For example, if the oxide layer thickness removal rate in a certain area is slow, the electrode voltage in that area can be appropriately increased to speed up the ion migration rate, thereby accelerating corrosion. Ultimately, the ion migration rate control parameter value is obtained to achieve precise edge removal.

[0067] Since the thickness of the oxide layer of the silicon wafer is affected by the production environment and storage environment, the thickness is not uniform. For the thicker part, the voltage can be increased and the electrode spacing can be increased. For the thinner part, the voltage can be reduced and the electrode spacing can be reduced. This can adapt to the processing requirements of different areas and avoid the incomplete processing or over-processing that may damage the silicon wafer due to the different processing efficiency of constant voltage and constant spacing at different thickness positions.

[0068] Specifically, when the electrode spacing is reduced and the voltage is lowered, the ion migration distance becomes shorter and the speed becomes slower, and the corrosion effect is relatively mild. When the electrode spacing is increased and the voltage is increased, the ion migration distance becomes longer and the speed becomes faster, and the corrosion efficiency is improved, but the control difficulty increases.

[0069] To ensure its preparation accuracy, its ion migration rate should not vary by more than ±15% within a cycle. Furthermore, during the processing, to ensure corrosion efficiency, its ion migration rate needs to be maintained in the range of 10 to 50 μm / min.

[0070] This treatment method can achieve high-precision removal of the oxide layer on the edge of the silicon wafer, accurately control the removal thickness and removal area, and avoid unnecessary corrosion of the non-edge area on the front of the silicon wafer.

[0071] Example 4: After completing the edge removal process in Example 3, this example further removes the silicon wafer and performs a supercritical fluid replacement operation. Specifically, the silicon wafer is immersed in a supercritical carbon dioxide fluid, and the pressure and temperature are adjusted to allow the fluid to mix with the silicon wafer surface and the gap liquid to form a homogeneous system.

[0072] Specifically, supercritical carbon dioxide with a purity and storage conditions that meet the requirements is selected as the replacement fluid. During the specific implementation process, the silicon wafer is slowly immersed in a processing container filled with supercritical carbon dioxide fluid to ensure that the silicon wafer is completely immersed.

[0073] Supercritical fluid refers to a special state of matter above the critical temperature and critical pressure. At this time, the fluid has the dual characteristics of gas and liquid, with low viscosity, high diffusion coefficient and good solubility.

[0074] Supercritical carbon dioxide is a commonly used supercritical fluid. In this step, when the pressure and temperature are adjusted to a supercritical state, it can be fully mixed with the liquid on the surface and in the gaps of the silicon wafer to form a homogeneous system. This is because supercritical carbon dioxide has good solubility in many liquids. By adjusting the pressure and temperature, its solubility and physical properties can be changed, allowing it to penetrate into the tiny gaps in the silicon wafer and displace the liquid in the gaps, thereby achieving efficient replacement of the liquid on the surface and in the gaps of the silicon wafer.

[0075] In actual operation, it is necessary to adjust the pressure of the tank carrying the tidal boundary. The solubility of supercritical carbon dioxide is relatively weak at lower pressure and temperature; at higher pressure and temperature, the solubility is enhanced, and the liquid on the surface and gaps of the silicon wafer can be quickly replaced. However, in actual use, when the pressure is lower than 7.3Mpa, the solubility of the liquid on the silicon wafer surface is very low, and it is difficult to replace the liquid on the silicon wafer surface. When the pressure is higher than 15MPa, the excessive pressure will cause the relative temperature to rise. Although the replacement capacity is enhanced, it will have a thermal impact on the silicon wafer and affect the performance of the silicon wafer.

[0076] In this embodiment, the pressure is adjusted to 10 MPa and the temperature is adjusted to 50°C. Under these conditions, supercritical carbon dioxide can not only fully dissolve the liquid on the surface and in the gaps of the silicon wafer, but also control the thermal impact on the silicon wafer within a smaller range, thereby achieving efficient and stable fluid replacement.

[0077] During the replacement process, the flow rate of the supercritical fluid needs to be regulated and the replacement time needs to be controlled. The lower the flow rate, the longer the replacement time will be. The fluid will not mix sufficiently with the silicon wafer surface and the liquid in the gaps, and the replacement effect may not be ideal. According to conventional silicon wafer specifications, the replacement time is a maximum of 20 minutes, and the flow rate should not be less than 0.5L / min.

[0078] The shorter the replacement time, the higher the required flow rate, but this will increase processing costs and time. According to conventional silicon wafer specifications, the maximum flow rate is no more than 3L / min. At the maximum flow rate, the replacement time is at most five minutes. If it exceeds five minutes, the relative cost will increase significantly.

[0079] Real-time regulation is performed based on the displacement rate and uniformity during the actual displacement process. During the specific operation of this embodiment, the initial supercritical carbon dioxide fluid flow rate is set to 1.5 L / min and the displacement time is 12 minutes. While ensuring a good displacement effect, it also takes into account the processing efficiency and cost.

[0080] A concentration distribution detector is used to monitor the uniformity of fluid replacement in real time. During the specific implementation process, the deviation of the fluid replacement uniformity parameter value is controlled within ±5%. If the uniformity does not meet the standard, the pressure and temperature parameters are fine-tuned until a satisfactory replacement effect is achieved, and finally the fluid replacement uniformity parameter value is obtained.

[0081] This method uses a supercritical fluid replacement method, which can efficiently and thoroughly remove residual liquid on the surface and in the gaps of the silicon wafer, preventing the residual liquid from contaminating the silicon wafer or affecting subsequent drying operations.

[0082] Example 5: Based on Example 4, this example further implements drying and decompression gasification processing of the silicon wafer. The pressure of the processing chamber is reduced to allow the supercritical carbon dioxide to gasify and escape. The humidity of the gas in the chamber is monitored by a humidity sensor. When the humidity falls below a set threshold, the decompression is stopped. Specifically, after the silicon wafer is removed, the silicon wafer is dried. First, the sealing of the processing chamber is checked to ensure the normal operation of the decompression equipment. The humidity sensor is calibrated to ensure that the measurement data is accurate and reliable.

[0083] According to the principle of phase change of a substance, when the pressure of the environment in which a substance is located decreases, its boiling point will also decrease accordingly. In this embodiment, supercritical carbon dioxide is in a supercritical state under a certain pressure and temperature. When the pressure of the processing chamber is reduced, the pressure of the supercritical carbon dioxide decreases accordingly. When the pressure is reduced to a certain level, the supercritical carbon dioxide undergoes a phase change, from a supercritical state to a gaseous state, thereby achieving gasification and escape.

[0084] The humidity of the gas in the chamber is monitored in real time through a humidity sensor. When the humidity is lower than the set threshold, it indicates that the moisture on the silicon wafer surface and in the chamber has been basically removed. At this time, the decompression operation is stopped and the silicon wafer drying is completed.

[0085] In actual processing, the pressure of the processing chamber needs to be reduced to 0.1-0.5 MPa within the decompression rate range of 0.05-0.2 MPa / min;

[0086] For silicon wafers with less supercritical carbon dioxide and less moisture remaining on the surface, a high rate can be used to reduce the pressure to a lower final high pressure. During this process, the supercritical carbon dioxide and moisture evaporate quickly, enabling rapid drying. However, for silicon wafers with more supercritical carbon dioxide and moisture remaining on the surface, simple rapid volatilization will result in some residues being unable to be effectively removed, so a lower decompression rate is required to meet the high pressure for a longer time. Although the volatilization rate is relatively low at this time, the longer the time, the larger the range of action on carbon dioxide and moisture, and the more supercritical carbon dioxide and moisture remaining on the silicon wafer surface can be fully removed.

[0087] In this embodiment, the pressure of the processing chamber is reduced to 0.3 MPa, and the decompression rate is 0.1 MPa / min. Under these parameters, the silicon wafer can be completely dried within a reasonable time. The humidity of the gas in the chamber is monitored in real time by a humidity sensor. When the humidity sensor detects that the humidity of the gas in the chamber is lower than 3% RH, the decompression operation is stopped. Finally, the amount of liquid residue on the surface of the silicon wafer is measured by weighing to complete the silicon wafer drying process. During the detection process, the value of the liquid residue on the silicon wafer surface shall not exceed 0.1 μg / cm 2 .

[0088] This reduced-pressure gasification method can quickly and effectively remove residual supercritical carbon dioxide and other moisture from the surface of the silicon wafer, achieving thorough drying of the silicon wafer. Under the parameter settings of this embodiment, the silicon wafer drying effect is good and meets the requirements of subsequent storage, transportation and use.

[0089] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for efficiently protecting the back oxide layer of a silicon wafer by edge removal, characterized in that: The following steps are involved: S1. In a chemical vapor deposition reaction chamber, multiple independent gas inlets are equipped with pressure valves, and the edge inlet pressure is set higher than the center inlet pressure, driving the reaction gas to diffuse from the edge to the center to form a concentration gradient; S2. Place a shape memory polymer sealing film containing magnetic particles on the edge of the silicon wafer, apply a magnetic field so that the magnetic particles are forced to drive the sealing film to initially conform to the edge contour, and heat the sealing film to restore the preset shape and wrap around the side of the silicon wafer; S3, placing the silicon wafer face up in the microfluidic channel, injecting a weak hydrofluoric acid solution, setting electrodes on both sides of the channel and applying an electric field to move ions in the solution along the channel constraint direction; S4, immersing the silicon wafer in a supercritical carbon dioxide fluid, and adjusting the pressure and temperature to allow the fluid to mix with the surface of the silicon wafer and the liquid in the gaps to form a homogeneous system; S5. Reduce the pressure in the processing chamber to allow the supercritical carbon dioxide to gasify and escape. Monitor the humidity of the gas in the chamber through a humidity sensor. Stop reducing the pressure when the humidity is lower than a set threshold.

2. The method for efficiently protecting the back oxide layer of a silicon wafer by edge removal according to claim 1, characterized in that: The difference between the edge inlet pressure and the center inlet pressure in S1 is 0.5-2.0 MPa, and the oxide layer thickness of at least 9 evenly distributed areas on the silicon wafer surface is monitored in real time by a laser thickness gauge.

3. The method for efficiently protecting the back oxide layer of a silicon wafer by edge removal according to claim 1, characterized in that: The magnetic particle content of the shape memory polymer sealing film in S2 is 15-35 wt %, the heating temperature is 60-120° C., the heating rate is 5-15° C. / min, and the sealing film laminating pressure is 0.05-0.2 MPa.

4. The method for efficiently protecting the back oxide layer of a silicon wafer by edge removal according to claim 1, characterized in that: The concentration of the weak hydrofluoric acid solution in S3 is 0.5-2.0 mol / L, the width of the microfluidic channel is 0.5-2 mm, the electrode spacing is 5-20 mm, and the oxide layer thickness in the 0.5-3 mm area around the edge of the silicon wafer is monitored by a spectroscopic reflectometer.

5. The method for efficiently protecting the back oxide layer of a silicon wafer by edge removal according to claim 1, characterized in that: The pressure adjustment range of the supercritical carbon dioxide fluid in S4 is 7.3-15 MPa, the temperature adjustment range is 31-80° C., the fluid flow rate is 0.5-3 L / min, the replacement time is 5-20 min, and the deviation of the fluid replacement uniformity parameter value does not exceed ±8%.

6. The method for efficiently protecting the back oxide layer of a silicon wafer by edge removal according to claim 1, characterized in that: In step S5, the pressure in the processing chamber is reduced to 0.1-0.5 MPa, the decompression rate is 0.05-0.2 MPa / min, and the residual liquid value on the surface of the silicon wafer does not exceed 0.1 μg / cm 2 .

7. The method for efficiently protecting the back oxide layer of a silicon wafer by edge removal according to claim 1, characterized in that: The glass transition temperature of the shape memory polymer sealing film in S2 is 40-60°C, and the thermal expansion coefficient is (50-150)×10 ～6 / ℃.

8. The method for efficiently protecting the back oxide layer of a silicon wafer by edge removal according to claim 1, characterized in that: In the above S3, the electrode voltage is adjusted so that the ion migration rate control parameter value is within the range of 10 to 50 μm / min, and the rate regulation within the range of ±15% can be achieved.

Images