Semiconductor manufacturing temperature regulation and control system

Abstract

The invention discloses a semiconductor manufacturing temperature regulation and control system which comprises a modular heat pump unit, a multi-parameter distributed sensor network, an intelligent distributed control platform and an energy efficiency optimization module. The modular heat pump unit adopts a pluggable design, the refrigeration, heating and humidity adjusting functions are supported through a cascade loop, the interface standard is unified, and interchange is supported. The multi-parameter sensor network is arranged in a manufacturing process area and collects environment temperature and humidity, pressure difference and equipment vibration data in real time, and accurate monitoring is achieved. The intelligent control platform optimizes heat pump operation parameters based on a self-adaptive learning algorithm in combination with real-time and historical data, dynamically adjusts regional temperature and humidity and realizes resource priority allocation. The energy efficiency optimization module achieves recovery, storage and distribution of heat energy and cold energy through a waste heat recovery and energy storage device, and the energy efficiency of the system is improved. The system meets the requirement of semiconductor manufacturing for high-precision temperature and humidity control, and the energy utilization efficiency and the system reliability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of temperature control, and in particular to a semiconductor manufacturing temperature control system. Background Art

[0002] In the field of semiconductor manufacturing, the production process of chips and other products is extremely precise and complex, and the requirements for the manufacturing environment are almost harsh. Among them, precise control of temperature and humidity is one of the key factors to ensure high-quality product output.

[0003] The temperature and humidity control of traditional semiconductor manufacturing workshops mostly relies on a combination of decentralized air conditioning systems and independent heating devices. In terms of temperature control, the air conditioning unit compresses the refrigerant through the compressor according to the set temperature to realize the refrigeration cycle, and changes the air temperature through the heat exchange of the evaporator and condenser. When heating, it uses electric heating elements or heat pump technology (mostly conventional single-stage air source heat pumps) to increase the heat energy of the air; humidity control mainly relies on the dehumidification and humidification modules of the air conditioner. Dehumidification uses condensation tubes to condense water vapor into liquid water for discharge, and humidification uses ultrasonic atomization, electrode humidification and other means to increase the moisture content of the air. For different process areas, based on experience and preset temperature and humidity standards, multiple air conditioning equipment and heating devices are set up in different areas. They operate independently and rely on manual regular inspections and manual adjustment of parameters to maintain environmental stability. At the same time, some high-end production lines will be equipped with simple sensors to monitor the temperature and humidity values ​​of each area and feed them back to the central control room. Operators remotely control the start and stop of the equipment and adjust the temperature and humidity settings based on the data, but the coordination of the equipment in each area is poor, and basically they are regulated independently.

[0004] Each device in the traditional distributed system operates independently, making it difficult to achieve high-precision coordinated control across regions and the entire process. The temperature and humidity fluctuations in different regions often exceed the ideal process requirements. For example, the temperature fluctuation in the lithography area exceeds ±0.5°C and the humidity deviation reaches ±3%, which cannot meet the needs of improving lithography accuracy. It is easy to cause defects such as uneven width of lithography lines and incomplete patterns, which seriously affect chip process accuracy and product yield. The air conditioning and heating devices in each area are frequently started and stopped on demand and operate independently. There is a lack of waste heat recovery and comprehensive energy allocation mechanisms. Especially under extreme conditions in winter and summer, the cooling and heating equipment are operated at high load at the same time, resulting in serious energy waste. Summary of the invention

[0005] Purpose of the invention: In order to overcome the shortcomings of the prior art, the present invention provides a semiconductor manufacturing temperature control system, which aims to solve the problems existing in the prior art: first, poor accuracy. The traditional distributed temperature control equipment has poor coordination; second, low energy efficiency. Conventional equipment is isolated, without waste heat utilization and dynamic allocation; and third, poor flexibility.

[0006] A semiconductor manufacturing temperature control system constructed by the present invention includes a modular heat pump unit. The heat pump unit adopts a pluggable design and has a multifunctional heat exchange module, which can support refrigeration, heating and humidity control at the same time. The realization of each function relies on different circuits and operates in coordination with the help of a cascade design. Different circuits achieve functional coordination through heat exchange and flow regulation. Specifically, the heat exchange of each circuit promotes directional heat transfer based on the temperature difference, and the flow regulation dynamically adjusts the flow of refrigerant or medium based on the real-time load feedback of each process area. The interface standards of the heat pump unit are unified and support the interchange of heat pump units of different models. The interchange process does not require tools to operate, and unit replacement and maintenance can be easily realized; The multi-parameter distributed sensor network is deployed in multiple semiconductor manufacturing process areas, including the photolithography annealing and wafer film thickness detection areas, to collect environmental temperature and humidity, pressure difference and equipment vibration data. The layout of the sensor network is optimized according to process requirements and air flow. With the help of simulation software, the location of each sensor is determined according to the sensitivity of different areas to parameters, air flow direction and speed. For example, in the photolithography area, temperature, humidity and pressure difference sensors are deployed at key heat dissipation and easily fluctuating locations according to the heat dissipation of the light source and the direction of airflow; An intelligent distributed control platform is connected to the sensor network and the heat pump unit in communication, optimizes the working parameters of the heat pump unit based on the real-time collected data in combination with an adaptive learning algorithm, dynamically adjusts the temperature and humidity in different areas, and controls by prioritizing resource allocation; The energy efficiency optimization module is coupled to the heat pump unit and is used to recover and store waste heat and distribute heat and cold energy according to load conditions to ensure system operation. The energy efficiency optimization module includes a waste heat recovery unit and an energy storage device. The waste heat recovery unit includes a plate heat exchanger and a heat energy converter, which can realize energy recovery under low and high temperature conditions; the energy storage device is used to store the recovered heat energy and release it during peak demand, supporting efficient charging and discharging.

[0007] Furthermore, the modular heat pump unit described in the present application includes: a variable flow refrigerant control module, which is used to dynamically adjust the refrigerant flow according to the load requirements of different process areas. The refrigerant flow adjustment range is 0 to its rated maximum flow, and it is adjusted through an electronic expansion valve. Its adjustment logic is based on the temperature, pressure and other parameters of each process area collected by the sensor network and compared with the preset standard values. When the difference reaches the set threshold, the electronic expansion valve opening is determined according to the built-in established algorithm to accurately control the refrigerant flow. For example, when the temperature of a certain photolithography area is higher than the preset upper limit of 3°C, the system increases the opening of the electronic expansion valve through calculation, increases the refrigerant flow and enhances the cooling effect; The efficient heat exchange component has a multi-channel design, supports the parallel operation of cooling, heating and dehumidification functions, and realizes functional coordination through heat exchange and transfer of fluids in different channels. Fluids in different functional channels flow and exchange heat in an orderly manner according to process requirements, such as the low-temperature refrigerant in the cooling channel absorbing heat and the heat medium in the heating channel releasing heat to achieve the corresponding temperature adjustment function; The quick-plug interface is used to realize the rapid disassembly and maintenance of the heat pump unit. The interface design supports the interchange of heat pump units of different models. Interchangeable heat pump units must comply with the unified modular interface standard and have the convenience of tool-free operation, which facilitates on-site unit replacement.

[0008] Furthermore, the multifunctional heat exchange module described in the present application adopts a cascade design, which includes a carbon dioxide cascade heat pump circuit for providing low-temperature refrigeration capacity, and the refrigeration capacity working fluid temperature range is -40°C to -10°C, which meets the cooling requirements of the annealing equipment and provides cooling on demand according to the annealing process temperature change curve; an air source heat pump circuit is used for thermal energy regulation in the normal temperature range, and the ambient temperature adjustment range is 15°C to 35°C, which is suitable for wafer film thickness detection, and can flexibly adjust the temperature as the temperature and humidity fluctuate in the detection area; a water cooling circuit is used for efficient heat conduction and supports waste heat recovery, which conducts heat through medium circulation and intercepts waste heat for storage and standby.

[0009] Furthermore, the multi-parameter distributed sensor network described in this application includes high-precision temperature and humidity sensors, which are distributed at key locations in each manufacturing area and are used to collect environmental parameters in real time. Under ambient temperature conditions ranging from -20°C to 40°C, the temperature measurement accuracy is ±0.1°C, and the humidity measurement accuracy is ±1%, laying a solid data foundation for system temperature and humidity control; pressure difference sensors are used to monitor changes in airflow between areas to ensure the stability of the clean environment. The pressure difference measurement range is 0-500Pa to ensure stable and compliant airflow in the process environment; vibration sensors are used to detect the operating status of manufacturing equipment and detect mechanical vibrations in the range of 1Hz to 500Hz to assist in monitoring equipment health and warn of faults.

[0010] Furthermore, the sensor network described in the present application is connected to the intelligent distributed control platform through a wireless communication module, supports real-time data transmission, adopts the Zigbee 3.0 protocol for low-power, high-stability data transmission, and is backward compatible with the old version of the protocol to ensure unimpeded data transmission.

[0011] Furthermore, the intelligent distributed control platform described in the present application includes a real-time data processing unit for receiving data collected by the sensor network and performing real-time analysis, quickly extracting effective information to assist in subsequent control decisions, an adaptive learning algorithm module, which optimizes the operating parameters of the heat pump unit by combining historical data and real-time data, and adopts convolutional neural network (CNN) and long short-term memory network (LSTM) algorithms to predict the temperature and humidity requirements during the process, optimize the startup and adjustment cycle of the heat pump unit, and plan the temperature and humidity control strategy in advance. The multi-region coordinated control module prioritizes resource allocation according to the needs of different manufacturing areas to ensure the accuracy of temperature and humidity control. The coordinated control range temperature is -80°C to +150°C, the control accuracy is ±0.1°C, and resources are finely allocated to meet the needs of multiple processes.

[0012] Furthermore, the energy efficiency optimization module described in the present application includes a waste heat recovery unit for recovering heat energy from the exhaust heat of the heat pump unit, including a plate heat exchanger and a heat energy converter. Under low temperature conditions, the plate heat exchanger allows the low-temperature fluid (such as low-temperature refrigerant) and the high-temperature fluid (such as preheating medium) to flow in opposite directions on both sides of the plate, and uses the temperature difference to transfer heat from the high-temperature side to the low-temperature side to recover energy; under high temperature conditions, the heat energy converter converts heat energy into electrical energy and other forms for recycling with the help of mechanisms such as heat-to-work conversion. The energy storage device stores the recovered heat energy and releases it during peak demand. The energy storage device supports efficient charging and discharging, balancing energy supply and demand by shaving peaks and filling valleys. The dynamic allocation module adjusts the distribution ratio of heat energy and cold energy according to real-time load demand, and flexibly allocates energy to adapt to changes in working conditions.

[0013] Furthermore, in the present application, the waste heat recovery unit includes a plate heat exchanger and a heat energy converter, which can realize energy recovery under low and high temperature conditions, and realize energy interception and reuse by temperature difference heat transfer at low temperature and heat-to-work conversion at high temperature.

[0014] It can be seen from the above technical solution that the present invention has the following beneficial effects: In terms of precise temperature and humidity control, the intelligent distributed control platform optimizes the working parameters of the modular heat pump unit based on the multi-parameter sensor network data, using CNN and LSTM algorithms. Its variable flow refrigerant control module and multi-functional heat exchange module work together to achieve a temperature accuracy of ±0.1°C in each area and a humidity deviation of ±1%, reducing process defects and improving product yield. In terms of energy utilization, the waste heat recovery unit (plate heat exchanger and heat energy converter) of the energy efficiency optimization module recovers and converts waste heat, the energy storage device stores it, and the dynamic allocation module is deployed on demand. When the system is flexibly adapted, the quick-plug interface of the modular heat pump unit is interchangeable and can be quickly reorganized according to changes in the workshop and process; the multi-parameter sensor network optimizes the layout according to the process and airflow, and uses the Zigbee 3.0 protocol to stably transmit data to adapt to industrial upgrades. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is an architecture diagram of a semiconductor manufacturing temperature control system according to the present invention. DETAILED DESCRIPTION

[0016] like Figure 1 As shown, this embodiment is deployed in a medium-sized semiconductor manufacturing plant with two main production lines covering key process links such as lithography, annealing and wafer film thickness detection. The production workshop area is 2,000 square meters, and the daily fluctuation range of ambient temperature and humidity is 18°C-28°C and 40%-60%. Different process areas have more stringent requirements on temperature and humidity. For example, the lithography area requires 22°C±0.2°C and 48%±1% humidity.

[0017] Before installing the system, the power and ventilation systems of the workshop were renovated. The power system was upgraded to three-phase 380V, 50Hz AC, with 50kW of capacity reserved for future expansion. The ventilation ducts are made of SUS304 stainless steel, with a diameter of 200mm-500mm designed according to regional requirements, and cleanliness and durability verification was carried out.

[0018] 1. Modular heat pump unit Selection and configuration Based on the detailed calculation of the heat load of each process area, the peak heat load of the lithography area is about 30kW (considering factors such as the long-term operation heat generation of the lithography equipment and the heat dissipation of the lighting), the annealing area has special cooling requirements, and the heat load frequently switches between -25kW (cooling conditions) and -10kW (heating conditions, equipment preheating and maintaining ambient temperature), and the stable heat load of the wafer film thickness detection area is about 15kW. Based on this, a modular heat pump unit with corresponding cooling and heating capabilities is selected, and its carbon dioxide cascade heat pump loop cooling capacity can reach 35kW (-40°C to -10°C working conditions), which meets the deep cooling of the annealing equipment; the air source heat pump loop heating capacity of 20kW (15°C-35°C working conditions) is suitable for daily temperature adjustment of wafer film thickness detection, and the water cooling loop flow is designed to be 5m³ / h, which meets the requirements of efficient heat conduction and waste heat recovery.

[0019] Each heat pump unit is equipped with a quick-plug interface. The interface insertion and extraction force has been tested and controlled between 50N-80N. The contact resistance of the electrical connection contacts is less than 0.1mΩ, ensuring stable and efficient signal and power transmission. The mechanical locking structure can withstand 500N of lateral tension without loosening, ensuring long-term stable operation.

[0020] Installation Layout The heat pump units are centrally located in an independent equipment room covering an area of ​​80 square meters. Each heat pump unit is 1.5 meters apart, and the refrigerant pipes are covered with a 50mm polyurethane insulation layer. The heat loss of the pipes is less than 0.5°C under the ambient temperature difference.

[0021] 2. Multi-parameter distributed sensor networks Sensor Selection and Calibration In the photolithography area, 20 high-precision temperature and humidity sensors and 8 pressure difference sensors are installed around the key parts of 5 photolithography equipment (light source, objective lens, etc.) and key nodes of airflow. The temperature and humidity sensor model is SHT31. After calibration, the temperature measurement error is controlled within ±0.1°C and humidity within ±1% in an environment of 18°C-28°C. The pressure difference sensor has a range of 0-500Pa and an accuracy of ±2Pa, which meets the environmental stability monitoring needs of this area.

[0022] The annealing area is equipped with 15 temperature and humidity sensors and 6 pressure difference sensors, which are mainly arranged at the air inlet and outlet of the annealing furnace and around the furnace body. The high-temperature resistant temperature and humidity sensor HTM-220 is selected, which can withstand an ambient temperature of up to 80°C and meet the accuracy standards, ensuring accurate feedback of environmental parameters under high-temperature processes.

[0023] Because the wafer film thickness detection area is sensitive to micro-vibrations, 12 VS-100 vibration sensors are evenly arranged with a detection range of 1Hz-500Hz and a resolution of 0.1Hz. 18 temperature and humidity sensors are calibrated to ensure that the temperature and humidity are stably controlled at 22°C±0.1°C and 45%±2%, providing an ideal environment for high-precision detection.

[0024] Network layout and communication construction According to the laminar airflow direction (wind speed 0.3m / s-0.5m / s) designed for the workshop purification air conditioning system, the sensor layout is optimized, and the distance between adjacent sensors is 3m-5m to form a grid monitoring. Each sensor is networked through a wireless communication module. The communication module uses a ZigBee-Pro 3.0 chip with a transmission power of 10dBm. The communication packet loss rate is less than 0.1% in the complex electromagnetic environment of the workshop, and the communication distance can reach 50 meters indoors, ensuring stable data transmission to the intelligent distributed control platform. When testing compatibility with the old version of the protocol, the data transmission success rate exceeds 99%.

[0025] 3. Intelligent distributed control platform Hardware Setup The control platform is built based on the Dell PowerEdge R740 server, equipped with 2 Intel Xeon Gold 6240 processors (2.6GHz, 24 cores and 48 threads), 384GB DDR4 memory and 2 1.92TB solid-state drives to form a RAID1 array. The server is connected to the workshop Gigabit Industrial Ethernet with a network delay of less than 1ms and is equipped with dual redundant 1600W power modules to ensure 99.99% power availability.

[0026] Software deployment and algorithm debugging The self-developed "Semicon-TempCtrl" system management software was installed on the Windows Server 2019 operating system, integrating the real-time data processing unit, adaptive learning algorithm module, and multi-region coordination control module functions. The convolutional neural network (CNN) and long short-term memory network (LSTM) algorithm codes were embedded in the adaptive learning algorithm module, and the prediction deviation was controlled within ±0.1°C (temperature) and ±1% (humidity) by pre-training 500GB of production data.

[0027] (IV) Energy efficiency optimization module Waste heat recovery unit installation The plate heat exchanger (BRP-500, heat exchange area 50 square meters, material 316L stainless steel) is connected in series to the heat pump unit exhaust pipe. In the range of 30°C - 80°C exhaust temperature, the measured heat recovery efficiency is 60%-70%. The thermal energy converter (ETC-200, power 20kW, conversion efficiency of 15%-20% under 80°C-120°C working conditions) is installed adjacent to the heat pump unit to convert the recovered waste heat into electrical energy storage as needed.

[0028] Energy storage devices and dynamic allocation configuration The energy storage device uses a composite phase change material with a phase change temperature of 50°C-60°C to fill the heat storage tank, with a volume of 20 cubic meters. The charging and discharging efficiency has been tested to be over 85%. The dynamic allocation module accurately allocates heat and cold energy based on the real-time load of each process area (through sensor data updated and analyzed every second, combined with the production scheduling task weight), with the help of electric control valves (flow adjustment accuracy ±2%) and variable frequency pumps (frequency adjustment range 20Hz-50Hz, speed accuracy ±1%) to meet the instantaneous and continuous temperature and humidity control needs of each area.

[0029] 3. System operation test and verification 1. Initial operation and debugging When the system is powered on for the first time, each component goes through a 5-minute self-test process, the sensor network is activated first, and the first data collection is completed within 1 minute and transmitted to the control platform. The control platform instructs the heat pump unit to start operating initially according to the preset startup strategy (formulated based on the historical average temperature and humidity values ​​of the workshop during the same period, the weather forecast of the day, and other information). For example, at 6 a.m., when the workshop temperature is 18°C ​​and the humidity is 50%, the air source heat pump circuit preheats the wafer film thickness detection area, starting with a power of 5kW, observes the temperature and humidity change curves of each area, fine-tunes the sensor sensitivity (adjusts the gain parameter between 0.9 and 1.1), and the heat pump flow adjustment parameters (the initial adjustment step of the electronic expansion valve opening is set to 5%). After 2 hours of fine debugging, it is ensured that the system smoothly transitions to normal working state.

[0030] (II) Load condition test Conventional production load: Simulate the daily single production line with 80% load production, the lithography equipment is running continuously, the annealing furnace performs an annealing process every 2 hours, and the wafer film thickness detection is working uninterruptedly. Monitoring shows that the temperature in the lithography area is stable at 22°C±0.2°C and the humidity is 48%±1%. The temperature in the annealing area drops rapidly to -10°C (within 2 minutes) during cooling and maintains -10°C±0.3°C throughout the annealing cycle (about 30 minutes). The temperature reaches 15°C within 10 minutes during the heating and temperature recovery stage. The wafer film thickness detection area maintains 22°C±0.1°C and the humidity is 45%±1%. The pressure difference in each area is stable at 100Pa-200Pa, meeting the process requirements.

[0031] Peak production load: When the two production lines are in full load and running in parallel, the heat load of the workshop surges by 80%. The heat load of the lithography area light source is fully turned on, reaching 40kW. The heat load fluctuates violently due to frequent annealing operations. At this time, the system responds quickly, and the heat pump unit runs at full load. The cooling capacity of the carbon dioxide cascade heat pump loop is increased to 32kW, and the heating capacity of the air source heat pump loop is 18kW. The temperature and humidity in each area can still be controlled within the allowable deviation of the process, such as the temperature in the lithography area is 23°C±0.3°C, and the annealing cooling is -10°C±0.5°C, verifying the system's control ability and stability in dealing with complex working conditions.

[0032] 3. Energy efficiency assessment and optimization Energy consumption data statistics: In the first month of operation, the total power consumption of the system was 30,000 kWh, which is 28% energy-saving compared with the original traditional temperature control method (distributed air conditioning and heating equipment). Breaking down the energy consumption of each component, the heat pump unit accounts for 60% (about 18,000 kWh), the sensor network and communication module accounts for 10% (3,000 kWh), and the control platform and auxiliary equipment account for 30% (9,000 kWh).

[0033] Optimization of waste heat recovery and distribution: Analyzing the data of the waste heat recovery unit, the plate heat exchanger recovers an average of 5,000 kWh of heat per day during the low temperature period in winter and 8,000 kWh during the high temperature period in summer. After optimizing the conversion efficiency threshold setting of the thermal energy converter (increased from 10% to 12%) and the charging and discharging strategy of the energy storage device (adjusting the charging and discharging time window according to the peak and valley electricity prices and production load), the overall energy efficiency ratio of the system was increased to 3.2 (originally 2.8), achieving efficient operation and reducing operating costs.

[0034] 4. Long-term stability monitoring The system was continuously operated for 6 months, and the daily operation status and fault alarm information of each component were recorded. During regular inspections, 2 minor refrigerant leaks were found (due to aging of the pipe joint seals, which were repaired by tightening and replacing the seals), and 3 brief interruptions of the communication module signal (due to electromagnetic interference from large equipment in the workshop, which was solved by adjusting the communication channel). Monthly sensor calibration drift detection showed that the drift of the temperature and humidity sensor was less than 0.1°C (temperature), ±0.5% (humidity), the pressure difference sensor was ±3Pa, and the vibration sensor resolution was stable. The system operation and maintenance manual was improved based on the operation and maintenance records to ensure that the system can serve the temperature and humidity control needs of the semiconductor manufacturing process in a long-term and stable manner.

[0035] In this embodiment, the modules work closely together to achieve precise and efficient temperature and humidity control and energy optimization.

[0036] The modular heat pump unit interacts with the multi-parameter distributed sensor network: the sensor network collects temperature, humidity, pressure difference and vibration data of process areas such as lithography, annealing, and wafer film thickness detection in real time. For example, the high-precision temperature and humidity sensor (SHT31) and pressure difference sensor in the lithography area accurately monitor environmental fluctuations and transmit the data to the control platform using the Zigbee 3.0 protocol. The control platform analyzes and calculates based on the data and adaptive learning algorithms (CNN, LSTM). If the temperature in the lithography area exceeds 22°C±0.2°C, the algorithm determines the adjustment strategy and issues instructions to the modular heat pump unit. The variable flow refrigerant control module of the heat pump unit (through an electronic expansion valve) accurately adjusts the refrigerant flow, and the high-efficiency heat exchange component (based on multi-channel design) coordinates the switching of cooling, heating or dehumidification functions. Its carbon dioxide cascade heat pump circuit, air source heat pump circuit, and water cooling circuit efficiently provide cooling, heating or heat conduction to recover waste heat according to the working conditions, maintain the stability of regional temperature and humidity, and realize a closed loop of precise temperature and humidity control.

[0037] Intelligent distributed control platform and energy efficiency optimization module collaboration: The control platform predicts system load and temperature and humidity trends based on real-time data from the sensor network and algorithm models, providing decision-making basis for the energy efficiency optimization module. During low-load periods, such as when some production lines are shut down at night, the control platform instructs the energy efficiency optimization module energy storage device (composite phase change material heat storage tank) to store the heat pump unit to recover waste heat, and the plate heat exchanger (heat transfer based on a temperature difference of 30°C-80°C) and the heat energy converter (conversion efficiency of 15%-20% under specific working conditions) to recover and convert electrical energy storage on demand. When the heat load surges during production peaks, the platform instructs the dynamic allocation module (electric control valves and variable frequency pumps are adjusted according to accuracy standards) to accurately allocate heat and cold energy, and to coordinate energy supply from the energy storage device and the heat pump unit to ensure that the temperature control of each area improves energy utilization efficiency while achieving dynamic energy optimization management.

[0038] Modular heat pump unit and energy efficiency optimization module linkage: The waste heat recovery unit of the energy efficiency optimization module is closely connected to the heat pump unit exhaust pipe, recovering exhaust heat to preheat fresh air or converting electricity. The water cooling circuit and the plate heat exchanger work together to efficiently recover heat at 30°C-80°C, and convert and store it through the thermal energy converter. The charging and discharging of the energy storage device is flexibly adjusted according to the system load and electricity price, charging at low electricity prices and discharging at peak times. It is coupled with the heat pump unit through the dynamic allocation module to accurately supply energy according to the needs of the process area, improve the overall energy efficiency of the system, and ensure efficient energy utilization and cost reduction under stable temperature and humidity environments in each area. The interactive integration of various modules supports the efficient and stable temperature control needs of semiconductor manufacturing.

[0039] Through the above detailed implementation process, the advantages of the coordinated working of various components of the semiconductor manufacturing temperature control system are fully demonstrated, which effectively meets the complex and stringent process requirements of semiconductor manufacturing in terms of precise temperature and humidity control, high-efficiency operation and long-term stability, and helps factories improve product yield and production efficiency.

[0040] The above embodiments are exemplary, and their purpose is to illustrate the technical concept and features of the present invention so that people familiar with the technology in this field can understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any changes or modifications made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A semiconductor manufacturing temperature control system, characterized in that: include: Modular heat pump unit, the heat pump unit adopts a pluggable design, has a multifunctional heat exchange module, can support cooling, heating and humidity adjustment at the same time, and each function is realized by different circuits and works together through a cascade design, wherein different circuits realize the coordination between functions through heat exchange and flow regulation, the interface standard of the heat pump unit is unified and supports the interchange of heat pump units of different models, and the interchange process does not require tool operation; A multi-parameter distributed sensor network is deployed in multiple semiconductor manufacturing process areas, including photolithography annealing and wafer film thickness detection areas, to collect ambient temperature and humidity, pressure difference, and equipment vibration data. The layout of the sensor network is optimized according to process requirements and air flow to achieve the precise monitoring required in each area; An intelligent distributed control platform is connected to the sensor network and the heat pump unit in communication, optimizes the working parameters of the heat pump unit based on the real-time collected data in combination with an adaptive learning algorithm, dynamically adjusts the temperature and humidity in different areas, and controls by prioritizing resource allocation; The energy efficiency optimization module is coupled to the heat pump unit and is used to recover and store waste heat and distribute heat and cold energy according to load conditions to ensure system operation. The energy efficiency optimization module includes a waste heat recovery unit and an energy storage device. The waste heat recovery unit includes a plate heat exchanger and a heat energy converter, which can realize energy recovery under low and high temperature conditions; the energy storage device is used to store the recovered heat energy and release it during peak demand, supporting efficient charging and discharging.

2. A semiconductor manufacturing temperature control system according to claim 1, characterized in that: The modular heat pump unit comprises: Variable flow refrigerant control module, used to dynamically adjust the refrigerant flow according to the load requirements of different process areas. The refrigerant flow adjustment range is 0 to its rated maximum flow, and is adjusted through an electronic expansion valve; High-efficiency heat exchange components with multi-channel design to support the parallel operation of cooling, heating and dehumidification functions, and achieve the coordination of different functions through heat exchange and transfer of fluids in different channels; The quick-plug interface is used to realize the rapid disassembly and maintenance of the heat pump unit. The interface design supports the interchange of heat pump units of different models. The interchangeable heat pump units must comply with the unified modular interface standards and have the convenience of tool-free operation.

3. A semiconductor manufacturing temperature control system according to claim 2, characterized in that: The multifunctional heat exchange module adopts a cascade design, which includes: The CO2 cascade heat pump circuit is used to provide low-temperature refrigeration capacity with a working fluid temperature range of -40°C to -10°C to meet the cooling needs of annealing equipment; Air source heat pump circuit, used for thermal energy regulation in the normal temperature range, with an ambient temperature adjustment range of 15°C to 35°C, suitable for wafer film thickness detection; Water cooling circuit for efficient heat transfer and support of waste heat recovery.

4. A semiconductor manufacturing temperature control system according to claim 1, characterized in that: The multi-parameter distributed sensor network comprises: High-precision temperature and humidity sensors are distributed in key locations in each manufacturing area to collect environmental parameters in real time. In the ambient temperature range of -20°C to 40°C, the temperature measurement accuracy is ±0.1°C and the humidity measurement accuracy is ±1%; Pressure difference sensor, used to monitor airflow changes between areas to ensure the stability of the clean environment. The pressure difference measurement range is 0-500Pa; Vibration sensors are used to detect the operating status of manufacturing equipment and detect mechanical vibrations in the range of 1Hz to 500Hz.

5. A semiconductor manufacturing temperature control system according to claim 4, characterized in that: The sensor network is connected to the intelligent distributed control platform through a wireless communication module, supports real-time data transmission, adopts the Zigbee 3.0 protocol for data transmission, and is backward compatible with the old version protocol.

6. A semiconductor manufacturing temperature control system according to claim 1, characterized in that: The intelligent distributed control platform includes: A real-time data processing unit, used to receive data collected by the sensor network and perform real-time analysis; Adaptive learning algorithm module, which optimizes the operating parameters of the heat pump unit by combining historical data and real-time data, uses convolutional neural network (CNN) and long short-term memory network (LSTM) algorithms to predict the temperature and humidity requirements during the process and optimize the startup and adjustment cycle of the heat pump unit; The multi-zone coordinated control module prioritizes resource allocation according to the needs of different manufacturing areas to ensure the accuracy of temperature and humidity control. The coordinated control range temperature is -80°C to +150°C, with a control accuracy of ±0.1°C.

7. A semiconductor manufacturing temperature control system according to claim 1, characterized in that: The energy efficiency optimization module comprises: A waste heat recovery unit for recovering heat energy from the exhaust heat of the heat pump unit; Energy storage devices, which store the recovered thermal energy and release it when demand peaks. Energy storage devices support efficient charging and discharging; Dynamic allocation module adjusts the distribution ratio of heat and cold energy according to real-time load demand.

8. A semiconductor manufacturing temperature control system according to claim 7, characterized in that: The waste heat recovery unit includes a plate heat exchanger and a heat energy converter, and can realize energy recovery under low temperature and high temperature conditions.

Images