A waste gas purification system in a semiconductor production process

Abstract

The present application relates to the technical field of semiconductor waste gas treatment, and particularly relates to a waste gas purification system in a semiconductor production process. The system comprises a flow field reconstruction pretreatment module, a multiphase reaction decomposition module, a temperature control anti-blocking washing module, a dynamic load matching module and a real-time pressure difference compensation module. The control method receives an operation load message sent by a power-driven execution unit through a purification scheduling control center, determines a load characteristic sequence or a state symbol sequence corresponding to a purification unit, and then determines an operation mode type and issues a control instruction. The present application synchronizes the process stage number with the process machine step sequence, combines a spiral guide and a thermal balance jacket structure, can effectively solve the short flow dead zone problem caused by uneven flow field distribution, prevent system fouling and blocking caused by by-product deposition, and significantly improve the dynamic matching precision of purification energy efficiency and the long-term operation stability of the system.

Description

Technical Field

[0001] This invention relates to the field of semiconductor associated waste gas treatment technology, and more specifically, to a waste gas purification system for semiconductor manufacturing processes. Background Technology

[0002] In precision manufacturing scenarios such as chemical vapor deposition, plasma etching, and epitaxial growth, the exhaust gases are rich in silanes, fluorine-containing gases, acidic and alkaline components, and various volatile organic compounds, which increases the emission treatment load and makes the purification process more complex.

[0003] Currently, conventional semiconductor exhaust gas purification solutions mainly rely on spray towers or internal reaction chamber structures for treatment. These solutions face challenges such as limited flow field uniformity, scaling caused by byproduct deposition, and insufficient energy efficiency matching accuracy. These factors affect the equipment's operation under high airflow or high concentration conditions, placing higher demands on the continuous and stable compliance required in actual production and the equipment's long-term operational capabilities. Specifically, existing purification devices typically employ spray components, filter grids, or heat exchangers for physical and chemical treatment. However, these methods have room for improvement. Specifically, the relatively simple internal chamber structure can easily lead to short-flow phenomena or local dead zones during operation, resulting in uneven distribution of the gas-liquid or gas-solid two-phase contact interface and causing real-time fluctuations in purification efficiency. Limited temperature control and self-cleaning capabilities cause exhaust dust and solid salt crystals generated during reactions to easily accumulate in key parts of the flow channel, resulting in physical accumulation or precipitation of insoluble substances, increasing system pressure drop and maintenance frequency. Furthermore, the constant operating mode has limited dynamic response capabilities to semiconductor batch process loads, leading to energy consumption and reagent depletion during process idle periods.

[0004] In summary, conventional purification systems have room for improvement in terms of processing efficiency stability and pipeline anti-clogging, which has a certain impact on the vacuum level assurance and resource conservation of upstream precision production equipment. Therefore, there is an urgent need for a semiconductor exhaust gas purification system that can optimize flow field contact efficiency while taking into account by-product suppression and dynamic matching of energy efficiency, in order to improve the continuity of system operation and enhance the economy of the purification process. Summary of the Invention

[0005] In view of this, the present invention proposes a waste gas purification system and its control method in semiconductor manufacturing process, aiming to solve the problems of short-flow dead zone caused by uneven flow field distribution in the current semiconductor waste gas purification process, as well as the problems of system scaling and blockage caused by by-product deposition and insufficient dynamic energy efficiency matching accuracy.

[0006] In a first aspect, embodiments of this application provide a control method for a waste gas purification system in a semiconductor manufacturing process. This method can be applied to a purification scheduling and control center, which may be a central processing unit, a module (e.g., a circuit, chip, or chip system) in a purification management network element, or a functional node, logic module, or software capable of implementing all or part of the purification management functions. For example, the purification management network element may be a load coordination control network element or other network elements with logical operation functions. The method includes: determining a load characteristic sequence corresponding to the first purification unit based on multiple operating load messages from the power-driven execution unit for the first purification unit; wherein each operating load message includes an identifier of the first purification unit, a processing timestamp, and a process stage number, the process stage number indicating the response order when the first purification unit first generates a processing fluctuation, the processing timestamp and the process stage number indicating the order in which the first purification unit switches between multiple purification load intervals, the load characteristic sequence including multiple load point information of the first purification unit, each load point information corresponding to one of the multiple operating load messages, each load point information including the processing timestamp and process stage number from one of the multiple operating load messages; and determining the type of operating mode corresponding to the first purification unit based on the load characteristic sequence, the type of operating mode indicating the purification efficiency output mode of the first purification unit.

[0007] Based on the above scheme, the cleanroom scheduling and control center receives the operating load message from the power-driven execution unit, which includes the identifier of the first cleanroom unit, the processing timestamp, and the process stage number. Then, based on the multiple received operating load messages, it determines the load characteristic sequence of the first cleanroom unit and, based on the load characteristic sequence, determines the type of operating mode corresponding to the first cleanroom unit, thus accurately determining the type of operating mode. In this method, the operating load message received by the cleanroom scheduling and control center does not directly use the real-time concentration value from physical sensor feedback, but instead includes a process stage number generated based on a control algorithm. This process stage number is synchronized with the step sequence of the front-end semiconductor process equipment and is not affected by physical layer sensor detection lag or environmental noise, thus avoiding the risk of system response delays. Furthermore, since the generated load characteristic sequence is not forcibly bound to specific sensor hardware but is a sequence generated based on logical step sequence, this method of generating load characteristic sequences has good generalization ability, enabling rapid deployment in cleanroom equipment with different process technologies (such as dry etching or thin film deposition), improving system compatibility.

[0008] As one possible implementation method, the type of operation mode corresponding to the first purification unit is determined based on the load characteristic sequence, including: inputting the load characteristic sequence into the purification evaluation model to obtain the type of operation mode corresponding to the first purification unit.

[0009] Based on the above scheme, by introducing a trained purification assessment model, it is possible to extract deep operating patterns from the nonlinear load fluctuation characteristics. Using the model to determine the type of operating mode corresponding to the first purification unit helps to accurately determine the purification intensity matching strategy in the complex and ever-changing semiconductor process environment and avoid energy waste.

[0010] As one possible implementation method, determining the load feature sequence corresponding to the first purification unit based on multiple operating load messages for the first purification unit from the power drive execution unit includes: when the number of multiple operating load messages is greater than a preset sample threshold, determining the load feature sequence corresponding to the first purification unit based on multiple operating load messages for the first purification unit from the power drive execution unit.

[0011] Based on the above scheme, the sequence determination operation is performed only after the number of received operating load messages accumulates to a preset sample threshold, instead of updating the complete sequence every time a momentary message is received. This can effectively filter out electromagnetic pulses or airflow disturbance noise generated during process switching, reduce the computational overhead of the purification management algorithm, and improve the efficiency and stability of determining the load characteristic sequence.

[0012] As one possible implementation method, the type of operation mode corresponding to the first purification unit is determined according to the load characteristic sequence, including: when the number of load point information in the load characteristic sequence is greater than a preset judgment threshold, the type of operation mode corresponding to the first purification unit is determined according to the load characteristic sequence.

[0013] Based on the above scheme, when the amount of load point information in the load characteristic sequence is sufficiently rich, it can cover the entire process cycle. The operation mode determined based on this can more realistically reflect the long-term operation trend of the purification system, help to accurately determine the trend of pollutant fluctuations, and prevent the decline in treatment efficiency caused by misjudgment.

[0014] As one possible implementation method, a control method for an exhaust gas purification system in a semiconductor manufacturing process further includes: sending a control command to a power drive execution unit, the control command including the identifier of a first purification unit and the type of its operating mode, and the control command being used to generate a purification intensity compensation strategy for the first purification unit.

[0015] Based on the above scheme, the power-driven execution unit can generate precise energy consumption and reagent dosage strategies for the first purification unit according to the type of operation mode. For example, it can increase the fluid drive power when the high load mode is identified, and reduce the power operation when the standby mode is identified, thereby significantly improving the economic efficiency of semiconductor waste gas treatment.

[0016] As one possible implementation method, a control method for a waste gas purification system in a semiconductor manufacturing process further includes: determining a deposition risk type based on a load characteristic sequence, wherein the deposition risk type indicates the load status of by-product accumulation in the purification channel.

[0017] Based on the above scheme, by analyzing the degree of deviation of the load sequence, it is possible to predict in advance whether there is a trend of salt crystal deposition on the inner wall of the washing chamber, thereby achieving accurate monitoring of similar load changes during the reduction of the roll diameter and ensuring unobstructed flow.

[0018] Secondly, embodiments of this application provide a control method for a waste gas purification system in a semiconductor manufacturing process. This method can be applied to a power drive execution unit, which may be a flow field drive module, a module (e.g., a circuit, chip, or chip system) in a frequency converter control unit, or a functional node, logic module, or software capable of realizing all or part of the purification power control function. For example, the flow field drive module may be a frequency converter network element or other network element with power electronic control functions. The method includes: receiving a first load adjustment completion message, the first load adjustment completion message including an identifier of a first purification unit, a first processing timestamp, and an identifier of a first processing interval accessed by the first purification unit after adjustment; determining a first process stage number corresponding to the identifier of the first processing interval according to a first correspondence relationship corresponding to the first purification unit; wherein, the first correspondence relationship includes the correspondence between the identifiers of processing intervals accessed by the first purification unit after historical adjustment, the historical processing timestamps of the first purification unit, and the process stage numbers corresponding to the identifiers of processing intervals accessed by the first purification unit after historical adjustment; the process stage number is used to indicate the response order when the first purification unit first generates a processing fluctuation, and the processing timestamp and the process stage number are used to indicate the order in which the first purification unit adjusts among multiple processing intervals; sending an operating load message to a purification scheduling control center, the operating load message including the identifier of the first purification unit, the first processing timestamp, and the first process stage number; wherein, the operating load message is used to determine the type of operating mode corresponding to the first purification unit, and the type of operating mode is used to indicate the purification efficiency output mode of the first purification unit.

[0019] Based on the above scheme, the operating load message reported by the power drive execution unit adopts the process stage number based on the logic algorithm mapping, rather than the sensitive original signal fed back by physical sensors. This makes the control system more resistant to interference during data transmission, and will not cause a decrease in control accuracy due to plasma noise generated during the chemical reaction process, thus ensuring the stable operation of the system.

[0020] As one possible implementation method, determining the first process stage number corresponding to the identifier of the first processing interval according to the first correspondence relationship includes: if the first correspondence relationship includes the first process stage number corresponding to the identifier of the first processing interval, then obtaining the first process stage number from the first correspondence relationship; or, if the first correspondence relationship does not include the first process stage number corresponding to the identifier of the first processing interval, then generating the first process stage number according to the maximum number in the first correspondence relationship.

[0021] Based on the above scheme, by generating new process stage numbers through incremental logic, it is possible to quickly lock the current purification stage when the process flow changes. The method is simple and the logic is rigorous.

[0022] As one possible implementation method, a control method for an exhaust gas purification system in a semiconductor manufacturing process further includes: adding a correspondence between a first processing timestamp, a first processing interval identifier, and a first process stage number to the first correspondence relationship.

[0023] Based on the above scheme, dynamic maintenance of the purification trajectory is achieved, ensuring that the system can accurately record the evolution of the purification unit as the production cycle changes.

[0024] As one possible implementation method, a control method for an exhaust gas purification system in a semiconductor manufacturing process further includes: receiving a second load adjustment completion message, the second load adjustment completion message including an identifier of a first purification unit, a second processing timestamp, and an identifier of a second processing interval accessed after the first purification unit is adjusted; if the difference between the second processing timestamp and the timestamp corresponding to the most recent historical adjustment of the first purification unit indicated by the first correspondence is greater than a preset time threshold, then the first correspondence is deleted.

[0025] Based on the above scheme, by introducing an aging and cleaning mechanism based on the time axis, it can be ensured that the data stored in the first correspondence always reflects the current process status, and avoid old data from misleading the purification logic of new production batches.

[0026] Based on any of the implementation methods of the first and second aspects mentioned above, as one possible implementation method, the type of operation mode is full-load processing mode, low-energy cruise mode, or process switching transition mode.

[0027] Thirdly, embodiments of this application provide a control method for a waste gas purification system in a semiconductor manufacturing process. This method can be applied to a purification scheduling and control center. The method includes: receiving an operating load message from a power-driven execution unit; wherein the operating load message includes an identifier of a first purification unit and a purification status symbol sequence corresponding to the first purification unit, each purification status symbol in the purification status symbol sequence indicating the flow field fluctuation characteristics of the first purification unit; and determining the type of operating mode corresponding to the first purification unit based on the purification status symbol sequence, the type of operating mode indicating the purification efficiency output mode of the first purification unit.

[0028] Based on the above scheme, by symbolizing complex fluid dynamics and chemical reaction parameters, logic similar to tension control is transformed into symbol sequence recognition. This approach does not rely on specific analog feedback, but rather includes a state symbol sequence generated based on the flow field equilibrium state, effectively avoiding the risk of misjudgment caused by the drift of analog signals in the high-frequency electromagnetic environment of a semiconductor workshop.

[0029] As one possible implementation, each purified state is denoted as S or mW, where S represents the steady state of the flow field, mW represents the disturbed state of the flow field, m indicates the disturbance intensity and response slope, and m is a non-zero constant.

[0030] Based on the above scheme, using symbolic methods to quantitatively describe the level of turbulence activity during the purification process helps the control system to more intuitively grasp the intensity of the internal reaction, thereby making accurate compensation decisions.

[0031] As one possible implementation method, a control method for an exhaust gas purification system in a semiconductor manufacturing process further includes: determining the current process sub-environment type based on a purification state symbol sequence, wherein the current process sub-environment type indicates the concentration gradient of chemical components in the exhaust gas.

[0032] Based on the above scheme, the change gradient of pollutant concentration at the intake end can be deduced by the change frequency of the symbol sequence, thereby achieving accurate process identification in the absence of concentration sensors.

[0033] Fourthly, embodiments of this application provide a control method for a waste gas purification system in a semiconductor manufacturing process. This method can be applied to a power-driven execution unit. The method includes: determining a purification state symbol sequence corresponding to a first purification unit based on a first correspondence relationship. Each purification state symbol in the purification state symbol sequence is used to indicate the flow field fluctuation characteristics of the first purification unit. The first correspondence relationship includes at least one record. Each record in the at least one record indicates the correspondence between the identifier of a response interval historically experienced by the first purification unit, the weighted rank corresponding to the identifier of the response interval historically experienced by the first purification unit, and the purification state symbol corresponding to the identifier of the response interval historically experienced by the first purification unit. The weighted rank represents the deduplicated sorting of a response interval within the response intervals historically experienced by the first purification unit.

[0034] Based on the above scheme, the power-driven actuator achieves quantification of the nonlinear changes in flow channel pressure resistance by rank-sorting the historical response intervals. This logical sorting effectively avoids the shortcomings of physical sensor measurement lag. Furthermore, since it does not involve the direct exposure of actual pressure or flow parameters, it enhances the safety protection at the system control level.

[0035] Fifthly, embodiments of this application provide a control method for a waste gas purification system in a semiconductor manufacturing process, which is applied to a purification scheduling and control center. The method includes: determining a differential pressure time sequence corresponding to a first purification unit based on multiple operating load messages from a power-driven execution unit for a first purification unit; wherein each operating load message includes an identifier of the first purification unit, a processing timestamp, and a performance deviation value between flow channel coupling; the differential pressure time sequence includes multiple deviation time combinations of the first purification unit, each deviation time combination corresponding one-to-one with a single operating load message, and each deviation time combination including a performance deviation value between a processing timestamp and flow channel coupling from one of the multiple operating load messages; and determining the operating mode type corresponding to the first purification unit based on the differential pressure time sequence.

[0036] Based on the above scheme, by establishing a deviation time series, accurate modeling of the coupling relationship between multi-stage purification structures was achieved. The system no longer directly relies on the lagging end-emission feedback, but instead determines the most suitable operating mode by the evolution trend of the internal flow resistance deviation sequence, significantly improving the system's robustness.

[0037] As one possible implementation method, the control command also includes the confidence level of the control result and the generation time. When the confidence level is greater than a preset confidence level, the power drive execution unit adjusts the speed of the induced draft fan according to the control command; otherwise, it maintains the current parameters, thereby ensuring the self-protection capability of the purification system in the event of a momentary sensor failure.

[0038] Sixthly, embodiments of this application provide a control method for a waste gas purification system in a semiconductor manufacturing process. This method can be applied to a power-driven execution unit. The method includes: receiving a load adjustment completion message, the load adjustment completion message including an identifier of a first purification unit, a first processing timestamp, and an identifier of a first response interval accessed after adjustment by the first purification unit; obtaining a first deviation between the first response interval and a second response interval accessed before adjustment based on a first correspondence relationship corresponding to the first purification unit; wherein the first correspondence relationship includes the identifier of the first purification unit, the identifiers of response intervals accessed by the first purification unit before and after historical adjustment, the deviation between response intervals accessed by the first purification unit before and after historical adjustment, and the correspondence relationship between the historical processing timestamps of the first purification unit; and sending an operating load message to a purification scheduling control center.

[0039] Based on the above scheme, by reporting relative deviation rather than absolute physical quantities, the influence of sensor zero-point drift caused by equipment aging can be eliminated, thereby improving the real-time performance and accuracy of control decisions.

[0040] Seventhly, this application provides a waste gas purification system for semiconductor manufacturing processes according to embodiments of this application, comprising: a flow field reconstruction pretreatment module, the inlet of which is connected to the exhaust port of the semiconductor process equipment, the flow field reconstruction pretreatment module having an anisotropic flow guiding structure inside for converting the inlet flow field into a spirally rising turbulent flow field; a multiphase reaction decomposition module connected to the output end of the flow field reconstruction pretreatment module, the multiphase reaction decomposition module having a built-in high-temperature reaction generator; a temperature control anti-clogging washing module located downstream of the multiphase reaction decomposition module, the temperature control anti-clogging washing module including a multi-stage atomizing spray structure and a heat balance jacket structure covering the outer periphery of the washing chamber; a dynamic load matching module connected to the control system of the semiconductor process equipment and the power components of the purification system respectively; and a real-time differential pressure compensation module deployed in the purification flow channel, the real-time differential pressure compensation module including a sensor for monitoring changes in flow resistance and a self-cleaning actuator for executing self-cleaning commands.

[0041] Based on the above scheme, the flow field reconstruction pretreatment module eliminates dead zones in the flow field through spiral flow guidance and increases the gas-liquid contact area; the multiphase reaction decomposition module fully decomposes gases such as silane through a high-temperature environment; the temperature control anti-clogging washing module prevents reaction byproducts from precipitating and crystallizing on the wall surface through a thermal balance jacket; and the dynamic load matching module and the real-time differential pressure compensation module jointly ensure the stability and energy efficiency of the system during long-term operation.

[0042] As one possible implementation method, the flow field reconstruction pretreatment module includes a variable-diameter venturi tube and helical guide vanes disposed on the inner wall of the variable-diameter venturi tube; wherein, the pitch of the helical guide vanes gradually decreases from bottom to top. Through the gradient reduction of the pitch, the airflow is continuously accelerated and generates stronger centrifugal force during its upward movement, thereby causing the heavy particulate components in the exhaust gas to move closer to the wall surface, enhancing the initial separation effect.

[0043] As one possible implementation method, the multiphase reaction decomposition module is equipped with a honeycomb catalytic carrier and a microwave-assisted heating unit. The honeycomb catalytic carrier exhibits a gradient pore size distribution, with the pore size near the center being larger than that at the edges. The large pore size at the center helps reduce the resistance of the core airflow and prevents overheating due to excessive central load, while the small pore size at the edges increases the catalytic specific surface area, ensuring the complete conversion of the waste gas.

[0044] As one possible implementation method, the thermal balance jacket of the temperature-controlled anti-clogging washing module maintains the inner wall temperature of the washing chamber within a preset temperature difference range above the crystallization point of the salts formed by the reaction through circulating cooling medium. Precise temperature boundary control prevents corrosive crystals from forming on the wall surface due to the condensation of acidic gases, thus physically suppressing the risk of channel blockage.

[0045] Eighthly, this application provides a purification scheduling and control device, which has the function of implementing any of the implementation methods in the first, third and / or fifth aspects described above.

[0046] Ninthly, this application provides a power drive device that has the function of implementing any of the methods described in the second, fourth and / or sixth aspects above.

[0047] In a tenth aspect, this application provides a control device, including an interface circuit and one or more processors, the processors executing a computer program to cause the device to implement the aforementioned method.

[0048] In the eleventh aspect, this application provides a computer-readable storage medium storing instructions that, when executed, implement the above-described control method.

[0049] In a twelfth aspect, this application provides a computer program product that implements the above-described purification control method when the instructions are executed.

[0050] Regarding the beneficial effects of the various implementation methods in aspects eight through twelfth above, please refer to the beneficial effects of the corresponding parts in the aforementioned methods and system implementation schemes, which will not be repeated here. Attached Figure Description

[0051] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This application provides an overall structural block diagram of a waste gas purification system for semiconductor manufacturing processes. Figure 2 A flowchart of a control method applied to a purification scheduling and control center, provided in an embodiment of this application; Figure 3 This is a flowchart of a control method applied to a power-driven actuator provided in an embodiment of this application; Figure 4 This is a schematic diagram of the flow field reconstruction preprocessing module provided in an embodiment of this application; Figure 5 This is a schematic diagram of the multiphase reaction decomposition module and the temperature control anti-clogging washing module provided in the embodiments of this application.

[0052] Among them, 100 is a semiconductor process equipment; 200 is an exhaust gas purification system; 210 is a flow field reconstruction pretreatment module; 211 is a variable diameter venturi tube; 212 is a spiral guide vane; 220 is a multiphase reaction decomposition module; 221 is a honeycomb catalytic carrier; 222 is a microwave-assisted heating unit; 230 is a temperature-controlled anti-clogging washing module; 231 is an atomizing spray structure; 232 is a thermal balance jacket structure; 240 is a dynamic load matching module; 250 is a real-time differential pressure compensation module; 300 is a purification scheduling and control center; and 400 is a power drive execution unit. Detailed Implementation

[0053] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey its scope to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0054] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0055] In semiconductor manufacturing processes, semiconductor process equipment 100 generates various complex waste gases during operation. These waste gases are typically highly toxic, highly corrosive, flammable, and explosive, and also contain a large amount of particulate matter. To address the purification of these waste gases, this embodiment provides a waste gas purification system 200. Please refer to... Figure 1 The exhaust outlet of the semiconductor process equipment 100 is connected to the inlet of the exhaust gas purification system 200 via a corrosion-resistant sealed pipe. Inside the exhaust gas purification system 200, arranged sequentially according to the direction of exhaust gas flow, are a flow field reconstruction pretreatment module 210, a multiphase reaction decomposition module 220, and a temperature control anti-clogging scrubbing module 230. Furthermore, the entire system integrates a dynamic load matching module 240 and a real-time differential pressure compensation module 250 to ensure stable internal pressure and matching processing efficiency. The entire exhaust gas purification system 200 is under the unified command of the purification scheduling control center 300, and is provided with specific physical driving force by the power drive execution unit 400.

[0056] Specifically, the structure of the flow field reconstruction preprocessing module 210 is as follows: Figure 4 As shown. The air inlet of this module is directly flange-connected to the exhaust pipe of the semiconductor process equipment 100. The core component inside the flow field reconstruction pretreatment module 210 is a variable-diameter venturi tube 211. The variable-diameter venturi tube 211 includes a contraction section, a throat section, and an expansion section. On the inner wall of the contraction section, multiple spiral guide vanes 212 are arranged circumferentially at equal intervals. The helical angle of these spiral guide vanes 212 is set according to the rated flow velocity of the exhaust gas, usually between 30 and 60 degrees. When the exhaust gas is discharged from the semiconductor process equipment 100 and enters the contraction section, the airflow changes from linear motion to spiral rotation under the action of the spiral guide vanes 212. As the cross-sectional area of ​​the pipe decreases, the axial velocity and tangential velocity of the airflow increase simultaneously. This high-speed rotating flow field can cause large dust particles in the exhaust gas to move towards the pipe wall under the action of centrifugal force. Because the throat section of the variable-diameter Venturi tube 211 has the smallest diameter, the airflow velocity reaches its maximum value here, and the huge kinetic energy gradient helps to break up the agglomerated microparticles in the exhaust gas. In the expansion section of the variable-diameter Venturi tube 211, the airflow velocity gradually decreases, the pressure begins to rise, and the rotating flow field is further reconstructed here, which increases the contact area between the exhaust gas molecules and the subsequent reaction medium.

[0057] After being treated by the flow field reconstruction pretreatment module 210, the exhaust gas enters the multiphase reaction decomposition module 220. For example... Figure 5 As shown, the multiphase reaction decomposition module 220 is located downstream of the flow field reconstruction pretreatment module 210, and the two are fastened together by a flange with a sealing gasket. The multiphase reaction decomposition module 220 is filled with a honeycomb catalytic carrier 221. The honeycomb catalytic carrier 221 is made of high-temperature resistant ceramic material, and its surface is loaded with catalytically active components specific to semiconductor waste gases. To improve reaction efficiency, multiple microwave-assisted heating units 222 are arranged circumferentially on the outer side of the shell of the multiphase reaction decomposition module 220. The emitting ends of the microwave-assisted heating units 222 penetrate the shell and point towards the honeycomb catalytic carrier 221. During operation, the microwaves emitted by the microwave-assisted heating units 222 penetrate the catalytic carrier and directly act on the waste gas molecules and catalytically active sites, generating dipole reversal and ion conduction, thereby achieving rapid heating and inducing a multiphase catalytic reaction. This heating method avoids the hysteresis of traditional heat conduction, enabling the efficient decomposition of recalcitrant components such as silanes and perfluorinated compounds in the waste gas at a relatively low overall temperature. The honeycomb structure ensures a low drag drop as exhaust gas passes through, while also providing physical space for the uniform distribution of microwaves.

[0058] Following the multiphase reaction decomposition module 220 is the temperature-controlled anti-clogging washing module 230. The air inlet of this module connects to the exhaust outlet of the multiphase reaction decomposition module 220. An atomizing spray structure 231 is installed on the upper part of the temperature-controlled anti-clogging washing module 230. The atomizing spray structure 231 consists of a spray main pipe, branch pipes, and high-pressure atomizing nozzles installed on the branch pipes. The high-pressure atomizing nozzles atomize the washing liquid into micron-sized droplets, which fully contact the high-temperature reaction products discharged from the multiphase reaction decomposition module 220. To prevent salt crystallization and deposition on the washing chamber wall due to excessive temperature difference, a heat balance jacket structure 232 is wrapped around the outer side of the temperature-controlled anti-clogging washing module 230. The heat balance jacket structure 232 is circulated with cooling water or warm water, maintaining the temperature of the inner wall of the washing chamber above the preset dew point temperature through heat exchange, effectively preventing the adhesion of solid by-products to the wall surface. After the atomized droplets capture dust and acidic gases in the exhaust gas, they are collected in the collection tank at the bottom of the washing module, while the purified gas is discharged from the outlet at the top of the module.

[0059] In terms of physical connection, the dynamic load matching module 240 is connected to the main air inlet of the exhaust gas purification system 200 via a sensor interface to monitor the exhaust gas flow and pressure fluctuation signals from the semiconductor process equipment 100 in real time. These signals are transmitted to the purification scheduling and control center 300. The real-time differential pressure compensation module 250 is connected between the inlet of the flow field reconstruction pretreatment module 210 and the outlet of the temperature control anti-clogging scrubbing module 230, and obtains the resistance of the entire purification link through a pressure transmitter.

[0060] Please see Figure 2 This describes the control method and process applied to the purification scheduling and control center 300. The purification scheduling and control center 300 first obtains the operating status parameters of the semiconductor process equipment 100 through the dynamic load matching module 240, including the current process sequence, exhaust flow limit, and waste gas composition. Based on this real-time data, the control center calculates the required processing load. Subsequently, the control center sends control commands to the power drive execution unit 400 to adjust the speed of the system's main fan and the flow rate of the circulating pump. Simultaneously, based on the differential pressure value fed back by the real-time differential pressure compensation module 250, the control center determines whether there is local blockage or excessive flow resistance within the system and dynamically adjusts the opening of the make-up air valve at the inlet of the variable-diameter venturi tube 211 to maintain a constant negative pressure at the system front end relative to the process equipment.

[0061] Figure 3 The specific execution logic applied to the power-driven execution unit 400 is demonstrated. After receiving instructions from the purification scheduling and control center 300, the power-driven execution unit 400 first activates the microwave-assisted heating unit 222 for preheating. After the temperature sensor indicates that the honeycomb catalyst carrier 221 has reached the preset reaction temperature, the atomizing spray structure 231 is activated. Based on the real-time flow rate of the exhaust gas, the power-driven execution unit 400 precisely controls the power of the pump group through a frequency converter to ensure that the liquid-to-gas ratio of the washing liquid is always maintained within the optimal range. When the temperature difference between the inlet and outlet water of the heat balance jacket structure 232 is detected to exceed the safety threshold, the power-driven execution unit 400 automatically adjusts the heat exchange efficiency of the heat exchanger to prevent scaling on the inner wall of the temperature-controlled anti-clogging washing module 230.

[0062] In the specific description of the operating principle, this invention achieves efficient purification of waste gas through the physical synergy of multiple modules. First, the flow field reconstruction pretreatment module 210 uses a mechanical structure to forcibly change the airflow dynamics characteristics. When the waste gas enters the variable-diameter venturi tube 211 at a speed of 10 to 20 meters per second, the spiral guide vanes 212 endow it with strong angular momentum. This rotating flow field, upon entering the multiphase reaction decomposition module 220, can eliminate the stagnant zone at the front end of the honeycomb catalyst carrier 221, allowing the waste gas to pass evenly through each honeycomb channel. Within the multiphase reaction decomposition module 220, microwave energy is efficiently converted into chemical energy and localized thermal energy, and complex molecules in the waste gas undergo bond breaking and recombination on the catalyst surface. For example, fluorine-containing gas undergoes a hydrolysis reaction with water vapor on the carrier surface under microwave excitation, generating hydrogen fluoride and silica particles.

[0063] Next, the airflow carrying particles and acidic gases enters the temperature-controlled anti-clogging washing module 230. At this point, the fine droplets sprayed by the atomizing spray structure 231 undergo a violent exchange of momentum and mass with the airflow. Due to the large surface area of ​​the droplets, acidic gases such as hydrogen fluoride are rapidly absorbed. Simultaneously, the thermal balance jacket structure 232 maintains a constant wall temperature, reducing the migration of fine particles to the wall surface using the principle of thermophoresis. The real-time differential pressure compensation module 250 compensates for the resistance generated by the spray droplets and filter cake layer through feedback adjustment of the power drive execution unit 400, ensuring that the exhaust pressure fluctuation on the semiconductor process equipment 100 side is less than ±10 Pascals, which is crucial for the stability of precision semiconductor processes.

[0064] In practical applications, such as silicon epitaxy processes, the exhaust gas contains large amounts of dichlorosilane and hydrogen chloride. The exhaust gas first undergoes preliminary gas-solid centrifugal separation in the flow field reconstruction pretreatment module 210 to reduce the risk of subsequent catalyst clogging. In the multiphase reaction decomposition module 220, microwave heating rapidly raises the catalyst support to a reaction temperature above 300 degrees Celsius, causing chlorosilane catalytic hydrolysis. Subsequently, the temperature-controlled anti-clogging washing module 230 neutralizes the hydrogen chloride using an alkaline washing solution, and the thermal balance jacket structure 232 maintains the temperature at 60 degrees Celsius through circulating hot water, effectively preventing the byproduct ammonium chloride from crystallizing and clogging the nozzles and pipes.

[0065] The system's dynamic load matching module 240 can automatically switch operating modes according to different working conditions of the process equipment, such as etching, thin film deposition, or cleaning. During the standby phase of the equipment, the system enters a low-energy maintenance mode, maintaining only a small amount of spraying and basic negative pressure; during peak process phases, the power-driven execution unit 400 operates at full speed to ensure that all waste gases are fully decomposed and absorbed before being emitted.

[0066] The combination of the variable-diameter venturi tube 211 and the helical guide vane 212 in the flow field reconstruction pretreatment module 210 not only accelerates the flow but, more importantly, eliminates dead zones in the pipeline through flow field reconstruction. The root of the helical guide vane 212 is fixed to the inner wall of the contraction section, while its tip extends towards the central axis but does not contact the axis, forming a hollow rotating flow core. This design allows the high-speed fluid in the center of the exhaust gas to generate strong turbulent pulsations without physical obstruction, greatly enhancing the collision frequency between the airflow and the catalyst carrier inlet.

[0067] The microwave-assisted heating unit 222 in the multiphase reaction decomposition module 220 employs a multi-source feeding method to form a uniform electromagnetic field distribution within the honeycomb catalyst support 221. The pore size distribution of the honeycomb catalyst support 221 is optimized, with the porosity controlled between 60% and 80%, ensuring sufficient surface area for the reaction to occur while preventing rapid pore blockage by particulate matter. Microwave radiation enables the active centers on the catalyst surface to instantly generate high-temperature hotspots. The temperature of these micro-regions is much higher than the average temperature of the gas flow, thereby significantly reducing the macroscopic activation energy of the reaction.

[0068] The temperature-controlled anti-clogging washing module 230 employs a counter-current contact arrangement in its atomizing spray structure 231, where the nozzles spray downwards while the exhaust gas flows upwards. This counter-current motion increases the relative velocity and contact time between the two phases. The heat balance jacket structure 232 utilizes a multi-layer spiral flow channel design, where the heat transfer medium circulates within the channels and transfers heat to the washing chamber through the aluminum or stainless steel inner wall. This structure precisely controls the heat flux to the inner wall surface, ensuring no cold spots are generated throughout the washing path.

[0069] The connection between the real-time differential pressure compensation module 250 and the purification dispatch control center 300 adopts a high-speed fieldbus, with signal transmission delay controlled at the millisecond level. When the differential pressure transmitter detects an increase in system resistance, the control center calculates the resistance increment and instructs the variable frequency fan in the power drive execution unit 400 to increase the corresponding output frequency. At the same time, it optimizes the spray pressure by adjusting the flow control valve, thereby maintaining the dynamic balance of the system at the physical level.

[0070] In terms of structural assembly, all modules are connected via standard flanges, and the internal seals are made of perfluoroelastomer rubber, which is resistant to high temperatures and strong acids and alkalis. Each module is externally covered with an insulation layer to reduce heat loss and prevent burns. The power drive actuators 400 are centrally located in an independent cabinet at the bottom of the system and are connected to each actuator via cables and conduits. The purification dispatch control center 300 is installed behind the system's operation panel and is equipped with a real-time human-machine interface to display the operating parameters of each module and the system status.

[0071] The entire working cycle of the exhaust gas purification system 200 is a closed-loop physical control process. Starting from the exhaust gas entering the flow field reconstruction pretreatment module 210, to molecular pyrolysis through the multiphase reaction decomposition module 220, to physical capture and chemical neutralization through the temperature control and anti-clogging scrubbing module 230, and finally to global feedback control through the real-time differential pressure compensation module 250 and the dynamic load matching module 240, each link is coordinated with each other through precise physical structure connections and rigorous logical instructions.

[0072] In the specific physical implementation of flow field reconstruction, the contraction ratio (the ratio of inlet diameter to throat diameter) of the variable-diameter Venturi tube 211 is designed to be between 2:1 and 4:1. The length of the helical guide vane 212 accounts for two-thirds of the length of the contraction section. When the exhaust gas passes through the helical guide vane 212, part of its kinetic energy is converted into rotational kinetic energy. After entering the throat, the centrifugal force of the rotating flow field causes a gradient in the radial distribution of airflow density, with higher pressure near the tube wall and lower pressure at the central axis. This pressure gradient triggers strong backflow and vortices after entering the expansion section, forcing gas molecules in the exhaust gas to be pushed towards the inner surface of the honeycomb catalyst carrier 221.

[0073] The honeycomb catalyst support 221 is installed in the middle of the multiphase reaction decomposition module 220 and is fixed by a high-temperature resistant bracket. The waveguide opening of the microwave-assisted heating unit 222 is covered with microwave-transparent quartz glass to prevent dust generated during the reaction from contaminating the microwave generator. The operating frequency of the microwave-assisted heating unit 222 is typically set at 2.45 GHz, which can effectively couple the polar groups on the catalyst to generate a thermal effect.

[0074] The inner wall of the temperature-controlled anti-clogging washing module 230 is treated with a Teflon coating, resulting in extremely low surface energy. Combined with the temperature regulation of the heat balance jacket structure 232, even if a small number of particles adhere, they can be quickly washed away by the liquid film flowing down from the atomized spray structure 231. The nozzles in the atomized spray structure 231 are arranged in a staggered pattern to ensure that the spray area covers the entire cross-section of the washing chamber, leaving no blind spots.

[0075] The power drive actuator 400 includes not only a motor and pump, but also a series of proportional control valves and solenoid valves. For example, a proportional control valve is installed on the water supply line of the temperature control and anti-clogging washing module 230. The purification scheduling and control center 300 dynamically adjusts the flow rate of the washing water based on the temperature sensor reading at the outlet of the multiphase reaction decomposition module 220 to achieve precise cooling and purification effects.

[0076] In actual physical operation, the real-time differential pressure compensation module 250 achieves its function by adjusting a bypass damper driven by a servo motor. When the main flow resistance increases due to the accumulation of reaction byproducts, the bypass damper finely adjusts its opening, changing the total airflow into the fan and thus altering the fan's operating point. This allows the fan's suction force to offset the increased system resistance, maintaining a constant pressure at the process equipment outlet. This compensation process is continuous and smooth, avoiding the impact of sudden pressure changes on the semiconductor film deposition quality.

[0077] Through the close physical connection and cooperation of the aforementioned components, the exhaust gas purification system 200 can achieve continuous, efficient, and stable purification of complex semiconductor exhaust gases. The positional relationship between the modules ensures the logical consistency and tight physical connection of the processing, providing reliable end-of-pipe treatment for semiconductor manufacturing.

[0078] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0079] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0080] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0081] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A waste gas purification system for semiconductor manufacturing processes, characterized in that, include: The flow field reconstruction pretreatment module has its air inlet connected to the exhaust port of the semiconductor process equipment. The flow field reconstruction pretreatment module is equipped with an anisotropic flow guiding structure to convert the air inlet flow field into a spiral upward turbulent flow field. The multiphase reaction decomposition module is connected to the output of the flow field reconstruction pretreatment module. The multiphase reaction decomposition module has a built-in high-temperature reaction generator. The temperature-controlled anti-clogging washing module is located downstream of the multiphase reaction decomposition module. The temperature-controlled anti-clogging washing module includes a multi-stage atomizing spray structure and a heat balance jacket structure covering the outer periphery of the washing chamber. The dynamic load matching module is connected to the control system of the semiconductor process equipment and the power components of the purification system, respectively. The real-time differential pressure compensation module is deployed in the purification flow channel. The real-time differential pressure compensation module includes sensors for monitoring changes in flow resistance and self-cleaning actuators for executing self-cleaning commands.

2. The waste gas purification system for semiconductor manufacturing process according to claim 1, characterized in that, The flow field reconstruction preprocessing module includes a variable-diameter venturi tube and helical guide vanes disposed on the inner wall of the variable-diameter venturi tube; wherein, The pitch of the spiral guide vanes gradually decreases from bottom to top.

3. The waste gas purification system for semiconductor manufacturing process according to claim 1, characterized in that, The multiphase reaction decomposition module is internally equipped with a honeycomb catalytic support and a microwave-assisted heating unit; among which, The pore size distribution of the honeycomb catalyst support exhibits a gradient change, with the pore size near the center region being larger than that at the edge region.

4. The waste gas purification system for semiconductor manufacturing process according to claim 1, characterized in that, The temperature-controlled anti-clogging washing module also includes: Circulating washing liquid tank and multi-stage centrifugal atomizer; Multi-stage centrifugal atomizers produce liquid droplets with a median particle size of 30 to 50 micrometers; The heat balance jacket maintains the inner wall temperature of the washing chamber at 10 to 20 degrees Celsius above the crystallization point of the salts formed by the reaction through a circulating cooling medium.

5. The waste gas purification system for semiconductor manufacturing process according to claim 1, characterized in that, The dynamic load matching module includes a process signal acquisition unit, a frequency converter, and a logic processing unit. The process signal acquisition unit is used to obtain the operating status of the semiconductor process equipment, including process execution status, standby status, and maintenance status.

6. The waste gas purification system for semiconductor manufacturing process according to claim 5, characterized in that, When the logic processing unit detects that the operating state is standby, it reduces the frequency of the induced draft fan in the purification system to 20% to 40% of the rated frequency through the frequency converter controller, and reduces the liquid supply of the multi-stage centrifugal atomizer.

7. The waste gas purification system for semiconductor manufacturing process according to claim 1, characterized in that, The real-time differential pressure compensation module includes a high-precision differential pressure sensor, an ultrasonic descaling device, and a high-pressure pulse jet cleaning device; the high-precision differential pressure sensor is distributed at both the inlet and outlet ends of the temperature-controlled anti-clogging washing module, and is used to calculate the actual pressure drop value under the current flow rate in real time.

8. The waste gas purification system for semiconductor manufacturing process according to claim 1, characterized in that, The multiphase reaction decomposition module and the temperature control and anti-clogging washing module are connected by a flexible expansion joint with heat insulation properties. The inner lining material of the flexible expansion joint is made of polytetrafluoroethylene reinforced composite material.

9. The waste gas purification system for semiconductor manufacturing process according to claim 5, characterized in that, The dynamic load matching module also runs a predictive control algorithm; among which, The predictive control algorithm is based on establishing a correlation model between exhaust gas concentration and process steps, obtaining the expected pollutant load within a preset range, and adjusting the reaction intensity of the multiphase reaction decomposition module.

10. The waste gas purification system for semiconductor manufacturing process according to claim 7, characterized in that, During the self-cleaning process of the real-time differential pressure compensation module, the dynamic load matching module will increase the speed compensation of the induced draft fan by 10% to 15%. The circulating washing liquid tank integrates a solid-liquid separation unit, which includes an inclined plate sedimentation structure and an automatic slag discharge mechanism.

Images