High-precision MEMS piezoresistive pressure sensor and preparation method thereof
By introducing an active electrostatic force compensation mechanism into the piezoresistive pressure sensor and utilizing a combination of a vacuum chamber and a Wheatstone bridge, the measurement error caused by membrane structure deformation hysteresis was solved, achieving high-precision and stable air pressure detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-30
- Publication Date
- 2026-04-03
AI Technical Summary
Traditional piezoresistive pressure sensors suffer from measurement errors due to membrane deformation hysteresis under dynamic pressure changes or long-term operating conditions, a problem that existing technologies struggle to effectively address.
An active electrostatic compensation mechanism is adopted. A vacuum cavity is prepared inside the support layer of a silicon substrate, and a Wheatstone bridge is prepared on the functional layer. External air pressure is applied to the sensitive film through the vent hole. Electrostatic force is used to counteract the deformation of the film layer and keep the sensitive film in its initial state. The correspondence between air pressure and applied voltage is established for measurement.
It effectively suppressed the hysteresis error of the membrane structure, improved the accuracy and stability of absolute pressure detection, simplified the process, and reduced the packaging cost.
Smart Images

Figure CN121783422A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a sensor and its fabrication method, and more particularly to a MEMS piezoresistive pressure sensor and its fabrication method. Background Technology
[0002] Pressure measurement is a core technology in fields such as industrial automation control, aerospace navigation, medical device monitoring, and intelligent terminal sensing. Its measurement accuracy and stability directly affect the reliability of system operation. To meet the pressure detection needs in different scenarios, pressure sensors based on various principles have been widely developed and applied. According to the core detection mechanism, they can be divided into: (1) piezoresistive pressure sensors, which realize pressure-to-electrical signal conversion based on the semiconductor piezoresistive effect; (2) capacitive pressure sensors, which sense pressure through the change in the distance between the plates caused by the deformation of the membrane structure; (3) piezoelectric pressure sensors, which convert pressure into charge signals by utilizing the positive piezoelectric effect of piezoelectric materials; and (4) resonant pressure sensors, which detect pressure based on the change law of resonant frequency with pressure. Among them, capacitive pressure sensors have high sensitivity and low power consumption, but have problems such as parasitic capacitance interference; piezoelectric pressure sensors have fast response speed and are suitable for dynamic pressure detection, but cannot realize continuous measurement of static pressure; resonant pressure sensors have high measurement accuracy and good stability, but have complex structural design, difficult manufacturing process and high cost. In contrast, piezoresistive pressure sensors, with their outstanding features such as simple structure, strong compatibility with MEMS processes, fast response speed, and excellent linearity, have become one of the most widely used types of pressure sensors.
[0003] However, traditional piezoresistive pressure sensors employ a passive detection mode, whose core mechanism relies on the bending deformation of a membrane under external air pressure, converting the deformation into an electrical signal output via a Wheatstone bridge. Due to the inherent elastic hysteresis characteristics of the membrane material and the tendency for stress accumulation during deformation, the sensor may exhibit measurement errors under dynamic pressure changes or long-term operating conditions. Summary of the Invention
[0004] Objective of this invention: To address the aforementioned limitations of existing technologies, this invention proposes a high-precision MEMS piezoresistive pressure sensor to solve the measurement accuracy problem caused by membrane structure deformation hysteresis. Another objective of this invention is to propose a method for fabricating a high-precision MEMS piezoresistive pressure sensor, simplifying the process and improving structural compatibility.
[0005] Technical solution: A high-precision MEMS piezoresistive pressure sensor, comprising: a silicon substrate support layer, a functional layer and a silicon-based top electrode layer bonded together sequentially from bottom to top; The silicon substrate support layer is pre-etched to form a groove on the side facing the functional layer, and after bonding with the functional layer, it forms a complete vacuum cavity. The functional layer includes a device layer, the area of which facing the vacuum cavity is a sensitive film, and a Wheatstone bridge composed of a varistor and leads is provided on the sensitive film; the device layer also functions as a bottom electrode, and an N-type heavily doped connection region for device layer conductivity is provided on the device layer, and the bottom electrode is led out through a bottom electrode bonding block; The silicon-based top electrode layer includes a vent hole, a top electrode bonding block, and a top N-type heavily doped connection region disposed on a silicon substrate. The vent hole penetrates the silicon-based top electrode layer vertically and faces the vacuum cavity. The top N-type heavily doped connection region is used to make the silicon-based top electrode layer conductive. The top electrode bonding block is used to lead out the top electrode, so that the silicon-based top electrode layer and the device layer form an electrode pair.
[0006] Furthermore, the electrode pair is used to apply a voltage between the device layer and the silicon-based top electrode layer to generate electrostatic force, so that the sensitive film can maintain a flat initial state under different air pressures. Low-hysteresis pressure measurement is completed by establishing the correspondence between air pressure and the applied voltage.
[0007] Furthermore, the plurality of vent holes are arranged in an array on the silicon-based top electrode layer.
[0008] Furthermore, the functional layer also includes a buried oxide layer, which is bonded to the silicon substrate support layer, and the device layer is located on the buried oxide layer.
[0009] Furthermore, the device layer also includes a Wheatstone bridge bonding pad; the Wheatstone bridge is formed by doping on the device layer and is located at the edge of the sensitive film. The Wheatstone bridge bonding pad uses heavily doped connecting lines to lead the voltage application terminal and differential output terminal of the Wheatstone bridge to a region outside the sensitive film.
[0010] Furthermore, the functional layer also includes a protective layer, which is made of silicon dioxide and covers the surface of the device layer for the protection and insulation of the Wheatstone bridge.
[0011] The method for fabricating the high-precision MEMS piezoresistive pressure sensor includes the following steps: Step 1): Fabricate the silicon substrate support layer and functional layer; Step 2): Prepare the silicon-based top electrode layer; Step 3): Bond the silicon-based top electrode layer prepared in Step 2) to the functional layer prepared in Step 1).
[0012] Furthermore, step 1) specifically includes: 1.1): SOI substrate is prepared by wet etching to form grooves on an N-bulk silicon substrate, initially forming a cavity and forming the silicon substrate support layer; the buried oxide layer is formed by thermal oxidation on another N-silicon substrate, and the unoxidized part is the device layer; the buried oxide layer and the silicon substrate support layer are bonded to silicon dioxide under vacuum to form a vacuum cavity; 1.2): To fabricate a Wheatstone bridge, a thin silicon dioxide layer is generated on the surface of the device layer using a thermal oxidation method as a mask. The P- implantation window of the varistor region is formed by photolithography and etching. Then, light doping implantation is performed to form a P-type varistor. The silicon dioxide mask is removed and a new silicon dioxide mask is generated. The P+ implantation window of the heavily doped region is formed by photolithography and etching. The heavily doped region is then used as a heavily doped connection line and a P-type heavily doped connection region. The P-type heavily doped connection region is used as the ohmic contact region of the varistor. 1.3): Form ohmic contacts in the device layer, wash away the mask, regenerate a thin silicon dioxide layer, then use photolithography to form the window for N+ implantation of the heavily doped region, and then perform heavy doping implantation as the N-type heavily doped connection region of the device layer. 1.4): Fabricate leads and bonding pads. Remove the silicon dioxide mask layer from the previous step. Use PECVD to grow a layer of silicon dioxide on the wafer surface as an intermediary layer between Si and Al metals. Photolithography and etching are used to form contact vias between the metal and the heavily doped region. Use magnetron sputtering to generate a layer of Al metal on the front side. Photolithography and etching are used to form metal leads, Wheatstone bridge bonding pads and bottom electrode bonding pads.
[0013] Furthermore, step 2) specifically includes: 2.1): Forming electrode gaps: Photolithography is performed on an N-silicon substrate, and a groove is formed as an electrode gap using photoresist as a mask and dry etching is employed. 2.2): The top electrode is brought out, the mask is washed away, and a thin silicon dioxide layer is generated. Then, the window for N+ implantation of the heavily doped region is formed by photolithography. Then, heavy doping implantation is performed to form the top N-type heavily doped connection region. The silicon dioxide mask layer is removed, and a silicon dioxide layer is grown on the wafer surface as an intermediary layer between Si and Al metal using PECVD. The contact vias between the metal and the heavily doped region are formed by photolithography and etching. An Al metal layer is generated on the front side using magnetron sputtering to form the top electrode bonding block. 2.3): Prepare vent holes by photolithography and using photoresist as a mask, and dry etching to form vent holes.
[0014] Beneficial Effects: The MEMS piezoresistive pressure sensor of this invention fabricates a vacuum cavity inside a silicon substrate support layer using a cavity-sealed insulator silicon-on-insulator process. A Wheatstone bridge detection unit is fabricated on the surface of the sensitive thin film of the functional layer. This functional layer also serves as the bottom electrode, forming a silicon-based electrode pair together with the silicon-based top electrode layer with vent holes. During operation, external air pressure acts on the upper surface of the sensitive thin film through the vent holes of the silicon-based top electrode layer, creating a pressure difference with the vacuum cavity. This causes the sensitive thin film to bend and deform. At this time, the Wheatstone bridge output deviates from the initial zero bias voltage due to film stress. The control system then applies a voltage between the silicon-based electrode pairs, generating electrostatic force to counteract the bending tendency of the film layer, allowing the bridge output to recover and stabilize at the initial zero bias voltage. Since the sensitive thin film always remains in its initial state, the material elastic hysteresis caused by repeated deformation of the film structure during air pressure loading and unloading is avoided, thereby effectively suppressing hysteresis error and improving the accuracy, stability, and repeatability of absolute air pressure detection.
[0015] Compared with existing technologies: (1) The pressure sensor adopts an electrostatic active compensation mechanism, which suppresses film deformation through electrostatic force, so that the silicon sensitive film is always kept in the initial state, effectively reducing hysteresis error.
[0016] (2) The pressure sensor uses the device layer as the bottom electrode directly. Its conductivity and lead-out process can reuse the Wheatstone bridge fabrication process, without the need to fabricate heterogeneous electrodes and adapter structures, saving space and simplifying the process.
[0017] (3) The pressure sensor adopts an integrated packaging design. Each functional layer is directly bonded to form a closed sensing structure, which reduces packaging costs and reduces reliability risks caused by assembly gaps. Attached Figure Description
[0018] Figure 1 This is a schematic diagram of the overall three-dimensional structure of the MEMS piezoresistive pressure sensor of the present invention; Figure 2 This is a partial cross-sectional three-dimensional structural schematic diagram of the MEMS piezoresistive pressure sensor of the present invention; Figure 3 This is a cross-sectional view along line A-A' of the MEMS piezoresistive pressure sensor of the present invention. Figure 4 This is a circuit diagram of the Wheatstone bridge in the MEMS piezoresistive pressure sensor of the present invention. Figure 5 This is a process flow diagram of the MEMS piezoresistive pressure sensor of the present invention.
[0019] The attached figures are labeled as follows: Silicon substrate support layer 1, vacuum cavity 2, functional layer 3, buried oxide layer 301, device layer 302, protective layer 303, Wheatstone bridge 304, Wheatstone bridge bonding pad 305, bottom electrode bonding pad 306, P-type heavily doped connection region 307, device layer N-type heavily doped connection region 308, sensitive film 309, silicon-based top electrode layer 4, vent hole 401, top electrode bonding pad 402, top N-type heavily doped connection region 403. Detailed Implementation Plan
[0020] A high-precision MEMS piezoresistive pressure sensor, such as Figure 1 As shown, from bottom to top, there are silicon substrate support layer 1, functional layer 3 and silicon-based top electrode layer 4, with adjacent layers connected by bonding.
[0021] The silicon substrate support layer 1 corresponds to the bottom silicon structure of the SOI substrate. Its side facing the functional layer is pre-etched to form a groove, which, after bonding with the functional layer, forms a complete vacuum cavity 2. Figure 2 As shown.
[0022] like Figure 3 As shown, functional layer 3 includes a buried oxide layer 301, a device layer 302, a protective layer 303, a Wheatstone bridge 304, a Wheatstone bridge bonding pad 305, a bottom electrode bonding pad 306, a P-type heavily doped connection region 307, and a device layer N-type heavily doped connection region 308. The buried oxide layer 301 and device layer 302 are inherent layers of a standard SOI substrate, corresponding to the silicon dioxide isolation layer and top silicon layer of the SOI substrate, respectively. The silicon substrate support layer 1 is connected to the buried oxide layer 301 through bonding. The device layer 302 region located directly above the vacuum cavity 2 is the sensitive thin film 309. The Wheatstone bridge 304 is doped on the device layer 302 and located at the edge of the sensitive thin film 309. The P-type heavily doped connection region 307 serves as the ohmic contact region for the varistor. The varistors constituting the Wheatstone bridge are connected through heavily doped connection lines. Figure 4 As shown, the Wheatstone bridge bonding pad 305 is used for voltage application and differential output measurement of the Wheatstone bridge 304. The Wheatstone bridge 304 is led out to a region outside the sensitive thin film 309 using heavily doped connecting lines. Device layer 302 also functions as a bottom electrode. The N-type heavily doped connection region 308 of the device layer is used for conductivity of the device layer, and the bottom electrode bonding pad 306 above the N-type heavily doped connection region 308 is used to lead out the bottom electrode. The protective layer 303 is made of silicon dioxide and covers the surface of device layer 302 to achieve a protective function, while also serving as an insulating layer to prevent direct conduction between device layer 302 and the silicon-based top electrode layer 4.
[0023] The silicon-based top electrode layer 4 includes a vent 401, a top electrode bonding pad 402, and a top N-type heavily doped connection region 403 disposed on a silicon substrate. The vent 401 penetrates vertically through the silicon-based top electrode layer 4 and faces the vacuum cavity 2 within the silicon substrate support layer 1, allowing external air pressure to act vertically on the upper surface of the sensitive thin film 309 through the vent 401. The top N-type heavily doped connection region 403 is used to conduct electricity into the silicon-based top electrode layer. The top electrode bonding pad 402 above the top N-type heavily doped connection region 403 is used to lead out the top electrode, forming an electrode pair between the silicon-based top electrode layer 4 and the device layer 302, which serves as the bottom electrode. The silicon-based top electrode layer 4 is made of silicon and directly serves as the top electrode, forming an electrode pair with the device layer 302, which serves as the bottom electrode, without the need for additional independent electrode fabrication.
[0024] The MEMS piezoresistive pressure sensor of the present invention has an integrated closed sensing structure formed by bonding the silicon substrate support layer 1, the functional layer 3 and the silicon-based top electrode layer 4, without the need for an additional external packaging shell.
[0025] During operation, external air pressure first acts vertically on the upper surface of the sensitive film 309 of the functional layer 3 through the vent 401 of the silicon substrate top electrode layer 4. A pressure difference is formed between the lower surface of the sensitive film 309 and the vacuum cavity 2 in the silicon substrate support layer 1. This pressure difference drives the sensitive film 309 to bend and deform. Since the sensitive film 309 is a specific area of the device layer 302, the Wheatstone bridge 304 in the stress concentration area at its surface edge will experience a change in resistance due to the stress generated by the deformation trend, causing the bridge output to deviate from the preset initial reference state. After the control system detects the output deviation in real time, it immediately applies an adaptation voltage to the electrode pair formed by the silicon substrate top electrode layer 4 and the device layer 302 through the bottom electrode bonding block 306 and the top electrode bonding block 402, respectively, so that an electrostatic force is generated between the two plates in the opposite direction to the deformation trend of the film. This electrostatic force precisely cancels the deformation driving force brought about by the pressure difference, allowing the sensitive film 309 to maintain an initial state with almost no substantial deformation, thereby enabling the output of the Wheatstone bridge 304 to quickly recover and stabilize at the preset reference state. By establishing a correlation between air pressure and applied voltage for measurement, this active deformation cancellation mechanism avoids the hysteresis error caused by repeated thin-film deformation in traditional piezoresistive sensors.
[0026] like Figure 5 As shown, a method for fabricating a high-precision MEMS piezoresistive pressure sensor includes the following steps: Step 1): Prepare silicon substrate support layer 1 and functional layer 3.
[0027] Step 1.1): SOI substrate fabrication. A groove is formed on an N-bulk silicon substrate using wet etching, initially creating a cavity and forming the silicon substrate support layer 1. A buried oxide layer 301 is formed on another N-silicon substrate by thermal oxidation; the unoxidized portion forms the device layer 302. The buried oxide layer 301 and the silicon substrate support layer 1 are bonded to silicon dioxide under vacuum, forming a vacuum cavity 2.
[0028] Step 1.2): Fabrication of the Wheatstone bridge 304. A thin silicon dioxide layer is formed on the surface of device layer 302 using a thermal oxidation method as a mask. A window for P-implantation of the varistor region is formed by photolithography and etching, followed by light doping implantation to form a P-type varistor. The silicon dioxide mask is removed and a new one is formed. A window for P+ implantation of the heavily doped region is formed by photolithography and etching, and then heavily doped to form the heavily doped interconnect and the P-type heavily doped interconnect region 307.
[0029] Step 1.3): Forming ohmic contacts in the device layer. The mask is washed away, and a thin silicon dioxide layer is formed again. Then, the heavily doped N+ implantation window is formed by photolithography, and then heavy doping implantation is performed as the N-type heavily doped connection region 308 in the device layer.
[0030] Step 1.4): Fabrication of leads and bonding pads. The silicon dioxide mask layer from the previous step is removed. A silicon dioxide layer is grown on the wafer surface using PECVD as an interlayer between Si and Al metals. Contact vias between the metal and heavily doped regions are formed by photolithography and etching. An Al metal layer is formed on the front side using magnetron sputtering. Metal leads, Wheatstone bridge bonding pad 305, and bottom electrode bonding pad 306 are formed by photolithography and etching.
[0031] Step 2): Prepare silicon-based top electrode layer 4.
[0032] Step 2.1): Forming the electrode gap. Photolithography is performed on the N-silicon substrate, and using photoresist as a mask, a groove is formed as the electrode gap by dry etching.
[0033] Step 2.2): Lead out the top electrode. Remove the mask to form a thin silicon dioxide layer. Then, perform photolithography to form the window for N+ implantation in the heavily doped region. Next, perform heavy doping implantation to form the top-layer N-type heavily doped connection region 403. Remove the silicon dioxide mask layer and use PECVD to grow a silicon dioxide layer on the wafer surface as an intermediary layer between Si and Al metal. Photolithography and etching are then used to form contact vias between the metal and the heavily doped region. Use magnetron sputtering to form an Al metal layer on the front side, forming the top electrode bonding pad 402.
[0034] Step 2.3): Fabrication of vent hole 401. The vent hole 401 is formed by photolithography and dry etching using photoresist as a mask.
[0035] Step 3): Bond the silicon-based top electrode layer 4 prepared in step 2) to the functional layer 3 prepared in step 1).
[0036] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A high-precision MEMS piezoresistive pressure sensor, characterized in that, include: The silicon substrate support layer (1), the functional layer (3) and the silicon-based top electrode layer (4) are bonded together sequentially from bottom to top. The silicon substrate support layer (1) is pre-etched to form a groove on the side facing the functional layer (3), and after being bonded to the functional layer (3), it forms a complete vacuum cavity (2). The functional layer (3) includes a device layer (302), and the area of the device layer (302) facing the vacuum cavity (2) is a sensitive film (309). The sensitive film (309) is provided with a Wheatstone bridge (304) composed of a varistor and leads. The device layer (302) also functions as a bottom electrode. The device layer (302) is provided with a device layer N-type heavily doped connection region (308) for device layer conductivity, and the bottom electrode is led out through the bottom electrode bonding pad (306). The silicon-based top electrode layer (4) includes a vent hole (401), a top electrode bonding block (402), and a top N-type heavily doped connection region (403) disposed on a silicon substrate. The vent hole (401) penetrates the silicon-based top electrode layer (4) vertically and faces the vacuum cavity (2). The top N-type heavily doped connection region (403) is used to make the silicon-based top electrode layer conductive. The top electrode bonding block (402) is used to lead out the top electrode, so that the silicon-based top electrode layer (4) and the device layer (302) form an electrode pair.
2. The high-precision MEMS piezoresistive pressure sensor according to claim 1, characterized in that, The electrode pair is used to apply a voltage between the device layer (302) and the silicon-based top electrode layer (4) to generate electrostatic force, so that the sensitive film (309) can maintain a flat initial state under different air pressures. Low hysteresis pressure measurement is completed by establishing the correspondence between air pressure and the applied voltage.
3. The high-precision MEMS piezoresistive pressure sensor according to claim 1, characterized in that, Multiple vent holes (401) are arranged in an array on the silicon-based top electrode layer (4).
4. The high-precision MEMS piezoresistive pressure sensor according to claim 1, characterized in that, The functional layer (3) further includes a buried oxide layer (301), which is bonded to the silicon substrate support layer (1), and the device layer (302) is located on the buried oxide layer (301).
5. The high-precision MEMS piezoresistive pressure sensor according to claim 1, characterized in that, The device layer (302) also includes a Wheatstone bridge bonding pad (305); the Wheatstone bridge (304) is formed by doping on the device layer (302) and is located at the edge of the sensitive film (309). The Wheatstone bridge bonding pad (305) uses heavily doped connecting lines to lead the voltage application terminal and differential output terminal of the Wheatstone bridge (304) to a region outside the sensitive film (309).
6. The high-precision MEMS piezoresistive pressure sensor according to claim 1, characterized in that, The functional layer (3) also includes a protective layer (303), which is a silicon dioxide material and covers the surface of the device layer (302) for the protection and insulation of the Wheatstone bridge (304).
7. The method for fabricating a high-precision MEMS piezoresistive pressure sensor according to any one of claims 1-6, characterized in that, Includes the following steps: Step 1): Prepare the silicon substrate support layer (1) and functional layer (3); Step 2): Prepare the silicon-based top electrode layer (4); Step 3): Bond the silicon-based top electrode layer (4) prepared in step 2) to the functional layer (3) prepared in step 1).
8. The preparation method according to claim 7, characterized in that, Step 1) specifically includes: 1.1): SOI substrate is prepared by wet etching to form a groove on an N-bulk silicon substrate, initially forming a cavity and forming the silicon substrate support layer (1); the buried oxide layer (301) is formed by thermal oxidation on another N-silicon substrate, and the unoxidized part is the device layer (302); the buried oxide layer (301) and the silicon substrate support layer (1) are bonded to silicon dioxide under vacuum to form a vacuum cavity (2). 1.2): To prepare a Wheatstone bridge (304), a thin silicon dioxide layer is generated on the surface of the device layer (302) using a thermal oxidation method as a mask. The window for P- implantation of the varistor region is formed by photolithography and etching. Then, light doping implantation is performed to form a P-type varistor. The silicon dioxide mask is removed and a new silicon dioxide mask is generated. The window for P+ implantation of the heavily doped region is formed by photolithography and etching. The heavily doped region is then used as a heavily doped connection line and a P-type heavily doped connection region (307). The P-type heavily doped connection region (307) is used as the ohmic contact region of the varistor. 1.3): Form ohmic contacts in the device layer, wash off the mask, regenerate a thin silicon dioxide layer, then form the heavily doped N+ implantation window by photolithography, and then perform heavy doping implantation as the N-type heavily doped connection region (308) of the device layer. 1.4): Prepare leads and bonding pads. Remove the silicon dioxide mask layer from the previous step. Use PECVD to grow a layer of silicon dioxide on the wafer surface as an intermediary layer between Si and Al metals. Photolithography and etching are used to form contact vias between the metal and the heavily doped region. Use magnetron sputtering to generate a layer of Al metal on the front side. Photolithography and etching are used to form metal leads, Wheatstone bridge bonding pads (305) and bottom electrode bonding pads (306).
9. The preparation method according to claim 7, characterized in that, Step 2) specifically includes: 2.1): Forming electrode gaps: Photolithography is performed on an N-silicon substrate, and a groove is formed as an electrode gap using photoresist as a mask and dry etching is employed. 2.2): The top electrode is brought out, the mask is washed away, and a thin silicon dioxide layer is generated. Then, the window for N+ implantation of the heavily doped region is formed by photolithography. Then, heavy doping implantation is performed to form the top N-type heavily doped connection region (403). The silicon dioxide mask layer is removed, and a layer of silicon dioxide is grown on the wafer surface by PECVD as an intermediary layer between Si and Al metal. The contact vias between the metal and the heavily doped region are formed by photolithography and etching. A layer of Al metal is generated on the front side by magnetron sputtering to form the top electrode bonding pad (402). 2.3): Prepare the ventilation hole (401), perform photolithography and use photoresist as a mask, and dry etch to form the ventilation hole (401).