Coating system based on inert gas protection

Abstract

The invention relates to the technical field of physical vapor deposition coating, in particular to a coating system based on inert gas protection, and aims at solving the technical problems that existing coating equipment is insufficient in inert gas protection reliability, uneven in film layer quality, low in intelligent level, prone to sample oxidation, low in gas utilization rate and the like. The system comprises a vacuum coating cavity module, an inert gas protection module, a coating execution module, an exhaust and tail gas treatment module, a temperature control module, a PLC intelligent control module and an auxiliary support module. The core of the system is that the inert gas protection module constructs a full-flow closed-loop protection system of inert gas supply, purification, proportioning, uniform distribution, circulation, monitoring and emergency. The inert gas protection reliability and the film layer quality are remarkably improved, the inert gas utilization rate is increased, and the device has the advantages of being high in intelligent degree, convenient and fast to operate and high in universality and can be widely adapted to the high-end film coating requirements in the fields of semiconductors, lithium batteries, precise optical devices and the like.

Description

Technical Field

[0001] This invention relates to the field of physical vapor deposition coating technology, specifically a coating system based on inert gas protection. Background Technology

[0002] As is well known, in the field of physical vapor deposition coating technology, the core requirement for thin film preparation is to ensure the purity and quality of the film layer and avoid oxidation, hydrolysis and other reactions between the substrate and the film layer materials during the coating process. Such reactions can lead to various defects in the film layer, affecting the performance and service life of the product. This problem is particularly prominent for various sensitive substrates, which seriously limits the application of coating technology in high-end fields.

[0003] Existing inert gas protected coating equipment has many technical limitations, making it difficult to meet the stringent requirements of high-end coating scenarios. In terms of inert gas protection, the gas supply and control methods of existing equipment are relatively simple, failing to achieve uniform gas distribution. This easily leads to an uneven gas atmosphere in the coating area, thus affecting the uniformity of the film layer. Furthermore, gas protection is mostly passive, only able to remedy the situation after parameters become abnormal, unable to predict abnormal situations in advance, resulting in insufficient protection reliability.

[0004] In terms of sample handling, existing equipment struggles to provide complete inert gas protection for samples from pretreatment to loading and unloading. Samples are easily exposed to the atmosphere during transfer, leading to oxidative contamination of sensitive substrates and affecting film quality. Regarding thermal management, critical equipment components are prone to thermal expansion and deformation under high-power operation, which can damage equipment parts, compromise the stability of the coating process, and further compromise the effectiveness of inert gas protection.

[0005] Furthermore, existing equipment suffers from low inert gas utilization rates, with most employing a single-use emission mode, increasing consumable costs during production. Simultaneously, insufficient precision in coating parameter control and a lack of multi-algorithm collaborative regulation result in poor film quality repeatability, making it difficult to stably produce coated products meeting high-end demands. These overlapping technical challenges mean that the practicality, economy, and intelligence of existing coating equipment need improvement, failing to fully adapt to the diverse coating needs of high-end fields. Therefore, developing a novel inert gas protective coating system capable of solving these technical problems is of great significance. Summary of the Invention

[0006] (a) Technical problems to be solved To address the technical pain points of existing coating equipment, such as insufficient reliability of inert gas protection, uneven film quality, low level of intelligence, easy oxidation of samples, and low gas utilization, this invention provides a coating system based on inert gas protection.

[0007] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a coating system based on inert gas protection, comprising a vacuum coating chamber module, an inert gas protection module, a coating execution module, an exhaust and tail gas treatment module, a temperature control module, a PLC intelligent control module, and an auxiliary support module; the modules cooperate to achieve precise control of the entire inert gas protection process through multi-algorithm fusion, thereby completing automated coating under active inert gas protection throughout the process; wherein, the inert gas protection module is the core innovative module, and the PLC intelligent control module is the carrier of multi-algorithm fusion, used to integrate multiple intelligent algorithms and drive the linkage control of each module.

[0008] Furthermore, the vacuum coating chamber module is a sealed stainless steel chamber with its inner wall polished and passivated. The chamber is equipped with front and rear double doors, a high-transmittance quartz glass observation window, and a heat-insulating partition, which divides the interior of the chamber into a substrate mounting area and a coating execution area. The ultimate vacuum degree of the chamber is ≤3×10⁻⁻⁻⁶. 5 Pa, the pressure after 12 hours of pressure holding after shutdown is ≤5Pa; the cavity is equipped with a real-time sealing monitoring sensor, which is linked with the algorithm module of the PLC intelligent control module to trigger leakage warning and emergency gas replenishment mechanism.

[0009] Furthermore, the inert gas protection module includes an inert gas supply unit, a gas purification unit, a dynamic adaptive dual-channel uniform distribution unit, a multi-source inert gas intelligent proportioning unit, a gas circulation and recovery unit, an atmosphere detection unit, and a leak prediction and emergency gas replenishment unit; the inert gas protection module can realize the purification, precise proportioning, dynamic uniform distribution, recycling, real-time monitoring, anomaly prediction, and emergency remediation of inert gas.

[0010] Furthermore, the dynamic adaptive dual-channel uniform distribution unit adopts an integrated ring structure and an anode integrated design, integrating an electromagnetically controllable injection orifice, a gas concentration detection array, and an active cooling circuit; the orifice diameter of the electromagnetically controllable injection orifice can be dynamically adjusted within the range of 0.5~1.5mm, and the gas concentration detection array has 16 detection points evenly distributed; the dynamic adaptive dual-channel uniform distribution unit integrates machine vision algorithm and particle swarm optimization algorithm, which can dynamically adjust the injection parameters to ensure that the inert gas concentration deviation in the coating area is ≤0.05Pa.

[0011] Furthermore, the multi-source inert gas intelligent mixing unit includes a gas mixing chamber, a mixing ratio regulating valve, and a mixed gas detection sensor, and incorporates a genetic algorithm to automatically optimize the inert gas mixing ratio based on the substrate type, membrane material, and coating power. The mixing ratio can be adjusted within a range of 70:30 to 95:5, with a mixing accuracy deviation of ≤0.5%. The leak prediction and emergency gas replenishment unit integrates wavelet analysis and Bayesian inference algorithms, achieving a leak prediction accuracy of ≥98%. It is also equipped with a spare inert gas cylinder, which can automatically calculate the leakage amount and quickly replenish the inert gas to maintain the inert gas protection effect.

[0012] Furthermore, the coating execution module includes an arbitrarily combinable evaporation source assembly, a magnetron sputtering target assembly, and a substrate clamping assembly; the evaporation source assembly includes an electron beam evaporation source and a tungsten filament heated evaporation boat, with an evaporation temperature control range of 500~3000℃; the magnetron sputtering target assembly adopts DC / RF dual-mode control and is equipped with an independent cooling circuit; the rotating sample stage of the substrate clamping assembly can achieve stepless speed regulation from 0 to 60 r / min and is linked with the particle swarm optimization algorithm to ensure the uniformity of coating on the substrate surface.

[0013] Furthermore, the exhaust and tail gas treatment module includes a two-stage vacuum pumping unit consisting of a molecular pump and a mechanical pump connected in series, and a tail gas purification unit consisting of an activated carbon filter and a molecular sieve dryer arranged in sequence. The vacuum pumping unit uses a PID-fuzzy adaptive algorithm to control the pumping speed, and the tail gas purification unit uses a machine learning algorithm to optimize the processing parameters, and can further recover and reuse valuable inert gases.

[0014] Furthermore, the temperature control module includes a closed-loop water circulation cooling system and an auxiliary heating system; the water temperature of the closed-loop water circulation cooling system is stabilized at 20~25℃ and is regulated by a PID-fuzzy adaptive algorithm in conjunction with inert gas protection parameters; the auxiliary heating system can control the cavity ambient temperature within the range of 25~35℃ with a control accuracy of ±2℃, and is linked with an LSTM neural network algorithm to predict the risk of water vapor condensation and adjust it in advance.

[0015] Furthermore, the PLC intelligent control module integrates PID-fuzzy adaptive algorithm, LSTM neural network algorithm, genetic algorithm, particle swarm optimization algorithm, machine learning algorithm, wavelet analysis algorithm and Bayesian inference algorithm to form a multi-algorithm collaborative control system; the PLC intelligent control module has parameter setting and adjustment, real-time monitoring and active early warning, automated operation, data storage and traceability and safety protection functions, with parameter repeatability ≥98% and data storage time ≥1 year.

[0016] Furthermore, the auxiliary support module includes a stainless steel frame with shock-absorbing pads and adjustable feet, a fully enclosed inert atmosphere glove box integrated assembly, and an in-situ film thickness detection assembly; a transition chamber gas purification module is provided between the glove box and the vacuum coating chamber, and this transition chamber gas purification module is linked with the algorithm for control; the detection accuracy of the in-situ film thickness detection assembly is ≤0.1nm, which can ensure film thickness uniformity ≥98%, and it is linked with the particle swarm optimization algorithm to adjust coating parameters and gas protection parameters in real time.

[0017] (III) Beneficial Effects Compared with the prior art, the present invention provides a coating system based on inert gas protection, which has the following advantages: This inert gas-protected coating system achieves full-process control of inert gas from supply, purification, proportioning, and distribution to circulation, monitoring, and emergency response by constructing a complete closed-loop inert gas protection system. Combined with multi-algorithm collaborative regulation, it overcomes the limitations of passive regulation in existing equipment, enabling early prediction of abnormal gas atmosphere and leakage risks, and timely initiation of emergency remedial measures. This ensures the continuity and stability of inert gas protection, effectively preventing oxidation and hydrolysis reactions between the substrate and the film layer, and guaranteeing the purity of the film layer.

[0018] This invention optimizes the inert gas distribution method and combines it with intelligent control of the sample stage to achieve uniform gas atmosphere distribution within the coating area, ensuring uniform film thickness. At the same time, through precise temperature control and process parameter optimization, it avoids the negative impacts of thermal expansion and deformation of equipment components and fluctuations in coating parameters, reduces the generation of various defects in the film, improves the density and adhesion of the film, ensures the consistency and repeatability of film quality, and enables the stable production of coated products that meet high-end requirements.

[0019] This invention adds an inert gas recycling system and, combined with algorithm optimization, significantly improves the utilization rate of inert gas, reduces ineffective gas consumption, and lowers material costs in the production process. At the same time, through the purification treatment of exhaust gas, it achieves emission standards, avoids environmental pollution, and balances economic efficiency and environmental protection, meeting the development needs of green production.

[0020] This invention integrates a PLC intelligent control module with a multi-algorithm collaborative control system, realizing automated operation of the entire coating process. It can automatically optimize process parameters and inert gas protection parameters, reduce manual intervention, and lower the operating threshold. At the same time, it is equipped with complete data storage and traceability functions, which facilitates process optimization and product quality control, and also improves the ease of equipment operation and maintenance. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall workflow of the system of the present invention; Figure 2 This is a schematic diagram of the entire process of the inert gas protection module of the present invention; Figure 3 This is a schematic diagram of the vacuum extraction, inert gas filling, and circulation process of the present invention; Figure 4 This is a schematic diagram of the core process of coating execution in this invention; Figure 5 This is a schematic diagram of the PLC multi-algorithm collaborative control process of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Please see Figures 1 to 5 This invention relates to a coating system based on inert gas protection, comprising a vacuum coating chamber module, an inert gas protection module, a coating execution module, an exhaust and tail gas treatment module, a temperature control module, a PLC intelligent control module, and an auxiliary support module. These modules work in concert, achieving precise control of the entire inert gas protection process through multi-algorithm fusion, thereby completing automated coating under active inert gas protection throughout the process. The inert gas protection module is the core innovative module, and the PLC intelligent control module serves as the carrier for multi-algorithm fusion, integrating various intelligent algorithms and driving the coordinated control of each module.

[0024] In this design, the vacuum coating chamber module is a sealed stainless steel chamber with its inner wall polished and passivated. The chamber is equipped with front and rear double doors, a high-transmittance quartz glass observation window, and a heat-insulating partition. The heat-insulating partition divides the interior of the chamber into a substrate mounting area and a coating execution area. The ultimate vacuum degree of the chamber is ≤3×10⁻⁻⁻⁶. 5The pressure after 12 hours of pressure holding at shutdown is ≤5Pa. A real-time sealing monitoring sensor is installed inside the cavity. This sensor, linked to the algorithm module of the PLC intelligent control module, can trigger leakage warnings and emergency gas replenishment mechanisms. The design features front and rear double doors and a quartz glass observation window, balancing ease of equipment maintenance with visualization of the coating process. A heat-insulating partition separates the substrate installation area from the coating execution area, blocking heat transfer during the coating process and preventing the substrate performance from being affected by high temperatures. Precise control of the cavity's ultimate vacuum and pressure holding performance creates a vacuum environment that meets the requirements of high-end coating. The addition of a real-time sealing monitoring sensor, linked to the PLC algorithm module, detects abnormalities in the cavity's sealing in real time, triggering timely warnings and emergency gas replenishment, thus preventing inert gas loss and air infiltration caused by sealing leaks at the source. The cavity is ensured to have good sealing performance, cleanliness, and thermal stability, with an ultimate vacuum ≤3×10⁻⁻⁻⁶. 5 Pa, 12-hour holding pressure ≤5Pa, meeting the vacuum environment requirements of high-end coating; effectively avoids contamination of the film layer by impurities in the cavity, and prevents heat transfer damage to the substrate; sealing monitoring and emergency linkage mechanism, avoids the risk of sealing leakage in advance, ensures the continuity of inert gas protection, reduces film oxidation and scrapping caused by sealing problems, and improves the convenience and safety of equipment operation and maintenance.

[0025] In this solution, the inert gas protection module includes an inert gas supply unit, a gas purification unit, a dynamic adaptive dual-channel uniform distribution unit, a multi-source inert gas intelligent proportioning unit, a gas circulation and recovery unit, an atmosphere detection unit, and a leak prediction and emergency gas replenishment unit. The inert gas protection module can achieve inert gas purification, precise proportioning, dynamic uniform distribution, recycling, real-time monitoring, anomaly prediction, and emergency remediation. The inert gas supply unit provides a stable gas source, the gas purification unit removes impurities, the dynamic adaptive uniform distribution unit achieves uniform distribution, the multi-source proportioning unit adapts to different scenarios, the circulation and recovery unit improves utilization, the atmosphere detection unit monitors parameters in real time, and the leak prediction and emergency gas replenishment unit avoids abnormal risks. All units work together to ensure that inert gas protection covers the entire coating process, achieving precise and comprehensive protection. The core structure and functional boundaries of the inert gas protection module were clearly defined, enabling full-process control of inert gas from supply to emergency response. This solves the problems of missing inert gas protection links and incomplete control in existing technologies. It effectively removes impurities from inert gas, achieves uniform distribution and recycling, accurately matches the needs of different coating scenarios, monitors and avoids protection anomalies in real time, and comprehensively improves the reliability, pertinence and economy of inert gas protection, providing core guarantee for the preparation of high-purity film layers.

[0026] In this scheme, the dynamic adaptive dual-channel uniform distribution unit adopts an integrated ring structure and an integrated anode design, integrating an electromagnetically controllable injection orifice, a gas concentration detection array, and an active cooling circuit. The orifice diameter of the electromagnetically controllable injection orifice can be dynamically adjusted within the range of 0.5~1.5mm, and the gas concentration detection array has 16 detection points evenly distributed. The dynamic adaptive dual-channel uniform distribution unit integrates machine vision algorithms and particle swarm optimization algorithms to dynamically control the injection parameters, ensuring that the inert gas concentration deviation in the coating area is ≤0.05Pa. The orifice diameter is dynamically adjusted (0.5~1.5mm) through the electromagnetically controllable injection orifice, and combined with the gas concentration detection array with 16 detection points, the gas concentration distribution in the coating area is captured in real time. The integration of machine vision algorithms and particle swarm optimization algorithms dynamically adjusts the injection orifice diameter and angle based on the concentration detection data and plasma state, forming a closed-loop control to ensure uniform inert gas concentration in the coating area. At the same time, the active cooling circuit controls the control unit temperature to avoid thermal expansion and deformation. Achieving uniform distribution of inert gas in the coating area with a concentration deviation ≤0.05Pa completely solves the problems of uneven film thickness and parasitic plasma caused by uneven gas distribution in existing technologies; the dynamic adjustment design of the electromagnetically controllable injection orifice adapts to different coating power and gas source flow scenarios, improving the flexibility of uniform distribution; the active cooling circuit avoids thermal expansion deformation of the unit, reduces particulate contamination, further improves film thickness uniformity and film purity, and at the same time improves the structural stability and service life of the unit.

[0027] In this solution, the multi-source inert gas intelligent mixing unit includes a gas mixing chamber, a mixing ratio regulating valve, and a mixed gas detection sensor. It also incorporates a genetic algorithm, which can automatically optimize the inert gas mixing ratio based on the substrate type, membrane material, and coating power. The mixing ratio can be adjusted within a range of 70:30 to 95:5, with a mixing accuracy deviation of ≤0.5%. The leak prediction and emergency gas replenishment unit integrates wavelet analysis and Bayesian inference algorithms, achieving a leak prediction accuracy of ≥98%. It is also equipped with a spare inert gas cylinder, which can automatically calculate the leakage amount and quickly replenish the inert gas to maintain the inert gas protection effect. The multi-source mixing unit achieves mixing of multiple inert gases through a gas mixing chamber and a mixing ratio regulating valve. It integrates a genetic algorithm to automatically optimize the mixing ratio (70:30~95:5) based on parameters such as substrate type, membrane material, and coating power, and ensures mixing accuracy through a mixed gas detection sensor. The leak prediction and emergency gas replenishment unit integrates wavelet analysis and Bayesian inference algorithms to analyze multi-sensor data, accurately predict leak risks, and, in conjunction with a backup gas cylinder, automatically calculates the leakage amount and quickly replenishes gas to maintain the protective effect. The multi-source mixing unit achieves intelligent optimization and precise control of the inert gas mixing ratio, with a mixing deviation ≤0.5%, adapting to the protection needs of different substrates and membranes, especially improving the protection effect of active substrates, balancing protection performance and coating efficiency. The leak prediction accuracy is ≥98%, and the emergency gas replenishment mechanism responds quickly, rapidly replenishing inert gas to avoid protection interruption. The backup gas cylinder ensures uninterrupted gas supply, significantly improving the targeting, reliability, and continuity of inert gas protection, and reducing membrane oxidation and scrapping caused by mixing deviations and leaks.

[0028] In this solution, the coating execution module includes an arbitrarily combinable evaporation source assembly, a magnetron sputtering target assembly, and a substrate clamping assembly. The evaporation source assembly includes an electron beam evaporation source and a tungsten filament heated evaporation boat, with an evaporation temperature control range of 500~3000℃. The magnetron sputtering target assembly adopts DC / RF dual-mode control and is equipped with an independent cooling circuit. The rotating sample stage of the substrate clamping assembly can achieve stepless speed regulation from 0 to 60 r / min and is linked with the particle swarm optimization algorithm to ensure the uniformity of the coating on the substrate surface. This allows for flexible switching between multiple coating processes (thermal evaporation, magnetron sputtering), adapting to different melting points and types of coating materials and substrates, improving the system's versatility. Precise control of evaporation temperature and sputtering mode avoids problems such as unstable evaporation of coating materials and target overheating damage. The rotating sample stage and algorithm linkage ensure film thickness uniformity, and with inert gas protection, further improves film density and adhesion, reduces defects such as pinholes and color differences, and improves film quality and process repeatability.

[0029] In this solution, the exhaust and tail gas treatment module includes a two-stage vacuum pumping unit consisting of a molecular pump and a mechanical pump connected in series, and a tail gas purification unit consisting of an activated carbon filter and a molecular sieve dryer arranged sequentially. The vacuum pumping unit uses a PID-fuzzy adaptive algorithm to regulate the pumping speed, while the tail gas purification unit uses a machine learning algorithm to optimize processing parameters and enables further recovery and utilization of valuable inert gases. The two-stage pumping structure, with the mechanical pump performing pre-vacuuming and the molecular pump performing deep pumping, gradually increases the vacuum level of the cavity to meet coating requirements. The PID-fuzzy adaptive algorithm regulates the pumping speed to prevent excessively rapid pumping, which could lead to inert gas loss and substrate damage. The tail gas purification unit uses a two-stage treatment of activated carbon filtration and molecular sieve drying to remove impurities and moisture from the tail gas. Machine learning algorithms optimize processing parameters, and the valuable inert gases are recovered, achieving environmentally friendly emissions and resource utilization. It achieves rapid and precise improvement of cavity vacuum to meet ultimate vacuum requirements. The algorithm controls the pumping speed to avoid damaging the inert gas atmosphere and substrate. The exhaust gas is purified in two stages before being discharged in compliance with standards, avoiding environmental pollution. The valuable inert gas is recycled and reused. Combined with algorithm optimization of processing parameters, the exhaust gas treatment efficiency is improved, while the cost of inert gas consumption is reduced, further enhancing the system's economy and environmental friendliness.

[0030] In this solution, the temperature control module includes a closed-loop water circulation cooling system and an auxiliary heating system. The water temperature of the closed-loop water circulation cooling system is stabilized at 20~25℃ and is regulated by a PID-fuzzy adaptive algorithm in conjunction with inert gas protection parameters. The auxiliary heating system can control the ambient temperature of the cavity within the range of 25-35℃ with a control accuracy of ±2℃, and is linked with an LSTM neural network algorithm to predict the risk of water vapor condensation and adjust it in advance. The closed-loop water circulation cooling system cools key components such as the uniformly distributed unit and the target material. The PID-fuzzy adaptive algorithm, in conjunction with inert gas protection parameters, regulates the cooling water temperature (20~25℃) to ensure stable component temperatures. The auxiliary heating system heats the cavity environment, controlling the ambient temperature (25~35℃) with a control accuracy of ±2℃. It is also linked with an LSTM neural network algorithm to predict the risk of water vapor condensation based on inert gas water vapor content data, and adjusts the temperature in advance to prevent water vapor contamination of the inert gas atmosphere. Ensure that the temperature of key components of the equipment (uniform distribution unit, target material, etc.) is stable at 20-25℃ to avoid problems such as parasitic plasma and particulate contamination caused by thermal expansion and deformation; ensure that the temperature of the cavity environment is stable at 25-35℃, and precisely control the temperature deviation (±2℃) to avoid water vapor condensation contaminating the inert gas and ensure the purity of the inert gas protection; link the temperature control and inert gas protection parameters to achieve synergistic optimization of temperature and gas atmosphere, and further improve process stability and film quality.

[0031] In this solution, the PLC intelligent control module integrates a PID-fuzzy adaptive algorithm, an LSTM neural network algorithm, a genetic algorithm, a particle swarm optimization algorithm, a machine learning algorithm, a wavelet analysis algorithm, and a Bayesian inference algorithm, forming a multi-algorithm collaborative control system. The PLC intelligent control module possesses parameter setting and adjustment, real-time monitoring and proactive early warning, automated operation, data storage and traceability, and safety protection functions. Its parameter repeatability is ≥98%, and data storage time is ≥1 year. The PID-fuzzy adaptive algorithm controls real-time parameters, the LSTM neural network algorithm achieves time-series prediction, the genetic algorithm optimizes the proportions, the particle swarm optimization algorithm controls the uniformity of distribution and rotation speed, the machine learning algorithm optimizes efficiency, and the wavelet analysis + Bayesian inference algorithm predicts leakage. Through algorithm collaboration, the various modules are driven to work together, achieving full-function control of parameter setting, real-time early warning, automated operation, data storage, and safety protection, ensuring stable system operation. The system achieves intelligent and precise control over the entire process, with parameter repeatability ≥98%, ensuring process stability and consistent membrane quality. An active early warning mechanism proactively mitigates abnormal risks, reducing membrane scrap and equipment damage. Automated operation reduces manual intervention and improves production efficiency. Data storage (≥1 year) and traceability facilitate process optimization and product quality control. Multiple safety protection mechanisms ensure personnel and equipment safety, while algorithm self-optimization continuously improves system operational stability and inert gas protection reliability.

[0032] In this solution, the auxiliary support module includes a stainless steel frame with shock-absorbing pads and adjustable feet, a fully enclosed inert atmosphere glove box integrated assembly, and an in-situ film thickness detection assembly. A transition chamber gas purification module is installed between the glove box and the vacuum coating chamber, and this transition chamber gas purification module is linked to the algorithm for control. The in-situ film thickness detection assembly has a detection accuracy of ≤0.1nm, ensuring film thickness uniformity of ≥98%, and is linked to the particle swarm optimization algorithm to adjust coating parameters and gas protection parameters in real time. This reduces the impact of equipment vibration on the coating process, avoiding problems such as uneven gas distribution and film thickness deviation caused by vibration. The sample is kept in an inert atmosphere throughout the process, with no air exposure, completely solving the oxidation problem during sample transfer, especially suitable for sensitive substrates. The film thickness detection accuracy of ≤0.1nm ensures film thickness uniformity of ≥98%, and the parameter adjustment in linkage with the algorithm further improves film quality. Overall, this improves the stability, safety, and practicality of the system, expanding its adaptability.

[0033] The PID-fuzzy adaptive algorithm is the core of the system's real-time parameter control. It is mainly used for closed-loop control of parameters such as inert gas flow rate, cooling water temperature, and cavity ambient temperature. It combines the accuracy of the PID algorithm with the adaptive capability of the fuzzy algorithm to solve the problem of control lag when the single PID algorithm fluctuates in the system parameters, and ensures the stability of parameters such as inert gas concentration and temperature.

[0034] Core algorithm formula (taking inert gas flow control as an example, adapted to actual system application scenarios): u(k)=Kp(k)⋅e(k)+Ki(k)⋅i=0ke(i)⋅Ts+Kd(k)⋅e(k)−e(k−1)Ts u(k): The output control quantity of inert gas flow rate (unit: sccm) during the kth control cycle, which directly acts on the high-precision mass flow controller (MFC) to adjust the output flow rate of inert gas. It is the final execution signal of the algorithm control.

[0035] Kp(k): The adaptive proportional coefficient (unitless) in the k-th control cycle, is the core adaptive parameter of the PID-fuzzy algorithm, which is adjusted in real time by the fuzzy controller based on the deviation e(k) and its rate of change. When the inert gas flow deviation is large (such as during system startup or emergency leakage), Kp(k) increases to accelerate the control response speed; when the deviation is small, Kp(k) decreases to avoid parameter fluctuations and ensure that the flow rate is stable at the set value.

[0036] Ki(k): The adaptive integral coefficient (unit: 1 / s) in the kth control cycle, which is also optimized in real time by the fuzzy controller to eliminate steady-state deviations in inert gas flow (such as small flow drifts during long-term operation). The integral term accumulates historical deviations to avoid the problem of "flow rate close to the set value but unable to reach it", ensuring that the inert gas concentration deviation is ≤0.05Pa and ensuring stable protection effect.

[0037] Kd(k): The adaptive differential coefficient (unit: s) in the k-th control cycle, controlled by the fuzzy controller, is used to predict the changing trend of the inert gas flow rate and make early control compensation. For example, when a large fluctuation in flow rate is detected (such as changes in cylinder pressure or signs of minor pipeline leakage), $$K_d(k)$$ will adjust the output in advance to avoid drastic fluctuations in flow rate and reduce the impact on the coating atmosphere.

[0038] e(k): The deviation between the actual flow rate and the set flow rate of the inert gas during the k-th control cycle (unit: sccm), i.e., e(k) = set flow rate - actual detected flow rate. The deviation data is collected in real time by the atmosphere detection unit (multi-parameter detection array) of the inert gas protection module, and is the core input basis for algorithm control, directly determining the adjustment direction of Kp(k), Ki(k), and Kd(k).

[0039] i=0ke(i)⋅Ts: Integral term (unit: sccm·s), where e(i) is the flow deviation in the i-th control cycle, and Ts is the control cycle (unit: s, set to 0.1s in this system), used to accumulate historical deviations, eliminate steady-state errors, and ensure that the inert gas flow rate remains stable at the set value for a long time.

[0040] e(k)−e(k−1)Ts: Differential term (unit: sccm / s), used to calculate the rate of change of flow deviation, reflecting the fluctuation trend of flow, helping the algorithm to predict deviation changes in advance, achieving "early control and active stabilization", and avoiding the impact of flow fluctuations on the uniform distribution effect of inert gas.

[0041] Ts: Algorithm control cycle (unit: s, fixed at 0.1s in this system). Combined with the response speed setting of the PLC intelligent control module, it ensures the real-time performance of flow control and matches the detection frequency of the atmosphere detection unit (synchronous acquisition of 16 detection points) to achieve closed-loop parameter control.

[0042] Example 1: Magnetron sputtering coating for semiconductor chip electrodes.

[0043] Equipment configuration: A 700mm (L) × 700mm (W) × 800mm (H) vacuum coating chamber is selected; the dynamic adaptive dual-channel uniform distribution unit is made of stainless steel and aluminum anode; the inert gas is an argon-krypton mixture (genetic algorithm optimized ratio 90:10), with a purity of 99.999%; the magnetron sputtering target assembly uses a 3-inch titanium target, DC sputtering mode; it is equipped with an in-situ film thickness tester (detection accuracy 0.05nm) and a fully enclosed inert atmosphere glove box (water and oxygen content ≤0.05ppm); the PLC intelligent control module uses all seven algorithms for coordinated control.

[0044] Process parameters: Vacuum degree ≤3×10⁻ 5 Pa, total inert gas flow rate 200 sccm, chamber pressure 0.5 Pa; substrate heating temperature 200℃, chamber ambient temperature 30℃, coating time 15 min, film thickness 100 nm; sample stage rotation speed 30 r / min (controlled by particle swarm optimization algorithm).

[0045] Operating procedures: Strictly follow the illustrated procedures. Figure 1-5 In the pretreatment stage, the chip substrate is cleaned by plasma cleaning in a glove box, and the gas parameters in the transition chamber are controlled by an algorithm. Vacuum extraction is carried out by mechanical pump pre-extraction and molecular pump deep extraction. After purification and proportioning, the inert gas is uniformly injected by a dynamic adaptive uniform distribution unit and the circulation recovery is initiated. During the coating process, the LSTM algorithm predicts the water and oxygen content, the PID-fuzzy algorithm controls the flow rate and temperature, and the wavelet analysis + Bayesian inference algorithm predicts leakage in real time. After the film thickness reaches the standard, it is kept at a constant temperature for 10 minutes for curing, and the part is removed after the inert gas is replenished and the pressure is released.

[0046] Performance testing: The titanium film surface is smooth with no oxidation spots, film thickness uniformity is 98.5%, adhesion is Grade 1 (no peeling in cross-cut test), purity is 99.995% (meets semiconductor chip electrode requirements); inert gas utilization rate is 93%, leakage prediction accuracy is 100%, and maximum fluctuation of water and oxygen content is ≤0.02ppm (verifying the reliability of inert gas protection); after 10 consecutive batches of operation, the film quality repeatability is 98.8% (verifying algorithm synergy and process stability).

[0047] Example 2: Thermal evaporation coating for lithium battery electrodes.

[0048] Equipment configuration: A small vacuum chamber of 360mm×360mm×500mm (H) is selected; the inert gas is a mixture of argon and xenon (genetic algorithm optimized ratio 85:15); the evaporation source component consists of 2 sets of tungsten wire heated evaporation boats; the sample stage can simultaneously load 4 lithium battery substrates; the glove box is equipped with a high-temperature furnace; the PLC intelligent control module focuses on activating the leakage prediction and emergency gas replenishment functions.

[0049] Process parameters: Vacuum degree ≤ 5 × 10⁻ 5 Pa, total inert gas flow rate 150 sccm, chamber pressure 0.3 Pa; substrate heating temperature 100℃, coating time 8 min, film thickness 50 nm; PID-fuzzy algorithm controls evaporation power and gas flow rate, LSTM algorithm predicts water oxygen content.

[0050] Operating procedure: Following the core steps of the flowchart, the substrate is loaded into the glove box and evacuated to the target vacuum level. The prepared inert gas is introduced and the circulation recovery is started. The tungsten filament evaporation boat is started, and the evaporation power is controlled by the algorithm. The sample stage is rotated for coating. A small leak is artificially set to simulate the actual working condition and verify the leak prediction and emergency gas replenishment functions. After the film thickness reaches the standard, it is kept at a constant temperature for 5 minutes for curing, and then the pressure is released and the sample is taken.

[0051] Performance testing: After a leak occurs, the algorithm completes the prediction and initiates emergency gas replenishment within 6 seconds. The cavity pressure and gas purity remain stable, and no air seeps in. The lithium film is free from oxidation and pinholes, with a film thickness uniformity of 98% and a tight bond with the substrate (improving the cycle life of lithium batteries). The product qualification rate is over 99.5%, and the inert gas utilization rate is 92.5% (verifying the reliability of multi-source ratio and emergency protection).

[0052] Example 3: Ion-assisted magnetron sputtering coating for precision optical devices.

[0053] Equipment configuration: A 600mm×600mm×700mm vacuum chamber is selected, with a dynamic adaptive dual-channel uniform distribution unit and an ion-assisted jet interface added; pure argon (99.999% purity) is used as the inert gas, and the gas purification unit is upgraded to three-stage purification; a 3-inch silicon target is used for the magnetron sputtering target assembly, with RF sputtering mode (automatic impedance matching); equipped with a high-precision in-situ ellipsometer (film thickness detection accuracy ≤0.05nm) and a glove box integrated assembly; the PLC intelligent control module optimizes algorithm parameters to adapt to the coating requirements of optical devices.

[0054] Process parameters: Vacuum degree ≤2×10⁻ 5 Pa, inert gas flow rate 180 sccm, chamber pressure 0.4 Pa; substrate heating temperature 150℃, chamber ambient temperature 28℃, coating time 20 min, film thickness 150 nm; ion-assisted power 300 W, sample stage rotation speed 40 r / min (controlled by particle swarm optimization algorithm); machine learning algorithm optimizes gas purification and circulation parameters.

[0055] Operational Procedure: After pretreatment in the glove box and loading in the transition chamber, the precision optical substrate is evacuated to the target vacuum level and purified argon gas is introduced. The uniform distribution unit is linked to the ion-assisted interface to achieve gas uniformity and ion-assisted synergy. The RF magnetron sputtering target is started, and the sputtering power and ion-assisted power are controlled by the algorithm. The ellipsometer detects the film thickness and refractive index in real time. Throughout the process, the water and oxygen content is predicted by the LSTM algorithm, and the leakage is predicted by the wavelet analysis algorithm to ensure a stable gas atmosphere. After the film thickness reaches the target, it is kept at a constant temperature for 15 minutes for curing, and then the pressure is released and samples are taken.

[0056] Performance testing: The optical film surface is smooth, without scratches or oxidation defects, with a film thickness uniformity of 99% and a refractive index deviation of ≤0.002 (meeting the requirements of precision optical devices); the film density is increased by 20%, and the light transmittance is ≥98.5% (verifying the synergistic effect of ion assistance and gas uniform distribution); after 8 consecutive batches of operation, the film quality repeatability is 99.2%, and the inert gas utilization rate is 94% (verifying the system's versatility, algorithm adaptability, and economy).

[0057] Summary of the effects of Example 1: Film thickness uniformity: 98.5%; Inert gas utilization rate: 93%; Leakage prediction accuracy: 100%; Maximum fluctuation of water oxygen content ≤0.02ppm; Core effects of membrane layer: no oxidation, first-class adhesion, purity 99.995%; Batch repeatability: 98.8% (10 batches).

[0058] Summary of the effects of Example 2: Film thickness uniformity: 98%; Inert gas utilization rate: 92.5%; Leakage prediction accuracy: not separately marked (effective in emergency response); Maximum fluctuation of water and oxygen content: not separately marked (no oxidation); Core effect of membrane layer: no oxidation, no pinholes, and tight bonding with the substrate; Batch repeatability: not separately marked (pass rate 99.5%+).

[0059] Summary of the effects of Example 3: Film thickness uniformity: 99%; Inert gas utilization rate: 94%; Leakage prediction accuracy: not separately marked (stable atmosphere with no leakage); Maximum fluctuation of water and oxygen content: not separately marked (no oxidation defects); Core effect of film layer: no defects, light transmittance ≥98.5%, refractive index deviation ≤0.002; Batch repeatability: 99.2% (8 batches).

[0060] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A coating system based on inert gas protection, characterized in that, It includes a vacuum coating chamber module, an inert gas protection module, a coating execution module, an exhaust and tail gas treatment module, a temperature control module, a PLC intelligent control module, and an auxiliary support module. These modules work together to achieve precise control of the entire inert gas protection process through multi-algorithm fusion, thereby completing automated coating under active inert gas protection throughout the process. The inert gas protection module is the core innovative module, and the PLC intelligent control module serves as the carrier for multi-algorithm fusion, integrating various intelligent algorithms and driving the coordinated control of each module.

2. The coating system based on inert gas protection according to claim 1, characterized in that, The vacuum coating chamber module is a sealed stainless steel chamber with its inner wall polished and passivated. The chamber is equipped with front and rear double doors, a high-transmittance quartz glass observation window, and a heat-insulating partition. The heat-insulating partition divides the interior of the chamber into a substrate mounting area and a coating execution area. The ultimate vacuum degree of the chamber... The pressure after 12 hours of pressure holding after shutdown is ≤5Pa; the cavity is equipped with a real-time sealing monitoring sensor, which is linked with the algorithm module of the PLC intelligent control module to trigger a leakage warning and emergency gas replenishment mechanism.

3. The coating system based on inert gas protection according to claim 1, characterized in that, The inert gas protection module includes an inert gas supply unit, a gas purification unit, a dynamic adaptive dual-channel uniform distribution unit, a multi-source inert gas intelligent proportioning unit, a gas circulation and recovery unit, an atmosphere detection unit, and a leak prediction and emergency gas replenishment unit. The inert gas protection module can realize the purification, precise proportioning, dynamic uniform distribution, recycling, real-time monitoring, anomaly prediction, and emergency remediation of inert gas.

4. A coating system based on inert gas protection according to claim 3, characterized in that, The dynamic adaptive dual-channel uniform distribution unit adopts an integrated ring structure and an anode integrated design, integrating an electromagnetically controllable injection orifice, a gas concentration detection array, and an active cooling circuit. The orifice diameter of the electromagnetically controllable injection orifice can be dynamically adjusted within the range of 0.5~1.5mm, and the gas concentration detection array has 16 detection points evenly distributed. The dynamic adaptive dual-channel uniform distribution unit integrates machine vision algorithms and particle swarm optimization algorithms, which can dynamically control the injection parameters to ensure that the inert gas concentration deviation in the coating area is ≤0.05Pa.

5. A coating system based on inert gas protection according to claim 3, characterized in that, The multi-source inert gas intelligent mixing unit includes a gas mixing chamber, a mixing ratio regulating valve, and a mixed gas detection sensor. It also incorporates a genetic algorithm to automatically optimize the inert gas mixing ratio based on the substrate type, membrane material, and coating power. The mixing ratio can be adjusted within a range of 70:30 to 95:5, with a mixing accuracy deviation of ≤0.5%. The leak prediction and emergency gas replenishment unit integrates wavelet analysis and Bayesian inference algorithms, achieving a leak prediction accuracy of ≥98%. It is equipped with a spare inert gas cylinder and can automatically calculate the leakage amount and quickly replenish the inert gas to maintain the inert gas protection effect.

6. The coating system based on inert gas protection according to claim 1, characterized in that, The coating execution module includes an arbitrarily combinable evaporation source assembly, a magnetron sputtering target assembly, and a substrate clamping assembly. The evaporation source assembly includes an electron beam evaporation source and a tungsten filament heated evaporation boat, with an evaporation temperature control range of 500~3000℃. The magnetron sputtering target assembly adopts DC / RF dual-mode control and is equipped with an independent cooling circuit. The rotating sample stage of the substrate clamping assembly can achieve stepless speed regulation from 0 to 60 r / min and is linked with the particle swarm optimization algorithm to ensure the uniformity of coating on the substrate surface.

7. A coating system based on inert gas protection according to claim 1, characterized in that, The exhaust and tail gas treatment module includes a two-stage vacuum pumping unit consisting of a molecular pump and a mechanical pump connected in series, and a tail gas purification unit with an activated carbon filter and a molecular sieve dryer arranged in sequence. The vacuum pumping unit uses a PID-fuzzy adaptive algorithm to control the pumping speed, and the tail gas purification unit uses a machine learning algorithm to optimize the processing parameters, and can further recover and reuse valuable inert gases.

8. A coating system based on inert gas protection according to claim 1, characterized in that, The temperature control module includes a closed-loop water circulation cooling system and an auxiliary heating system. The water temperature of the closed-loop water circulation cooling system is stabilized at 20~25℃ and is regulated by a PID-fuzzy adaptive algorithm in conjunction with inert gas protection parameters. The auxiliary heating system can control the ambient temperature of the cavity within the range of 25~35℃ with a control accuracy of ±2℃. It is also linked with an LSTM neural network algorithm to predict the risk of water vapor condensation and adjust it in advance.

9. A coating system based on inert gas protection according to claim 1, characterized in that, The PLC intelligent control module integrates PID-fuzzy adaptive algorithm, LSTM neural network algorithm, genetic algorithm, particle swarm optimization algorithm, machine learning algorithm, wavelet analysis algorithm and Bayesian inference algorithm, forming a multi-algorithm collaborative control system; The PLC intelligent control module has functions such as parameter setting and adjustment, real-time monitoring and proactive early warning, automated operation, data storage and traceability, and safety protection. Its parameter repeatability is ≥98%, and the data storage time is ≥1 year.

10. A coating system based on inert gas protection according to claim 1, characterized in that, The auxiliary support module includes a stainless steel frame with shock-absorbing pads and adjustable feet, a fully enclosed inert atmosphere glove box integrated assembly, and an in-situ film thickness detection assembly; a transition chamber gas purification module is provided between the glove box and the vacuum coating chamber, and this transition chamber gas purification module is linked with the algorithm for control; the detection accuracy of the in-situ film thickness detection assembly is ≤0.1nm, which can ensure film thickness uniformity ≥98%, and it is linked with the particle swarm optimization algorithm to adjust coating parameters and gas protection parameters in real time.

Images