Plasma surface treatment device for integrated circuit board chip processing
The plasma surface treatment device, which uses multi-module collaborative sensing and decision-making, solves the consistency problem of existing devices in processing different types of circuit boards and process environment fluctuations, realizes fine control of radio frequency power, and improves the reliability and processing quality of semiconductor devices.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGXI JINYUFENG ELECTRONIC TECHNOLOGY CO LTD
- Filing Date
- 2026-03-10
- Publication Date
- 2026-05-29
AI Technical Summary
Existing plasma surface treatment devices lack multi-parameter collaborative sensing and intelligent decision-making capabilities, making it unable to effectively cope with the characteristic differences of different integrated circuit boards, fluctuations in process environment parameters, and abnormal conditions. This results in poor consistency of plasma treatment effects, affecting the reliability and lifespan of semiconductor devices.
A multi-module collaborative sensing and decision-making system is adopted. Through the basic feedback module, process environment assessment module, arc-temperature adaptation module and feedforward prediction module, the plasma state, process environment stability and the characteristics of the processed object are quantified in real time, and the radio frequency power is dynamically adjusted to adapt to complex working conditions.
It achieves a high degree of consistency in plasma treatment effects, improves the yield rate of integrated circuit board surface mount processing and the reliability of semiconductor devices, avoids under-processing or over-processing, and ensures equipment safety.
Smart Images
Figure CN122121068A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of semiconductor manufacturing technology, and in particular relates to a plasma surface treatment device for integrated circuit board mounting. Background Technology
[0002] In the back-end packaging and system-level assembly stages of semiconductor device manufacturing, semiconductor chips need to be precisely soldered onto integrated circuit boards using surface mount technology to form complete electronic functional modules. The reliability of this process directly determines the performance and lifespan of the final semiconductor device. As semiconductor technology develops towards high density and miniaturization, chip packaging forms such as BGA, CSP, and WLP have increasingly stringent requirements for mounting accuracy, and the requirements for the cleanliness and chemical activity of the integrated circuit board surface have reached unprecedented levels.
[0003] During the initial processing, transportation, and storage of integrated circuit boards, an oxide layer inevitably forms on the surface of the solder pads, and organic contaminants adhere to them. These microscopic contaminants severely weaken the wettability between the solder and the pads, leading to defects such as poor soldering and cold solder joints after the semiconductor chips are mounted, directly reducing the reliability and lifespan of the final semiconductor devices. To address this issue, plasma surface treatment technology has been introduced into this field. It uses high-energy active particles in plasma to physically bombard and chemically react with the PCB surface, effectively removing contaminants and activating the surface.
[0004] Currently, existing plasma surface treatment devices typically include basic components such as a vacuum chamber, a radio frequency (RF) generator, an electrode system, a vacuum pump assembly, and a gas path system. Among these, the RF generator serves as the core energy source, generating a radio frequency electric field that ionizes the process gas to form plasma. However, the inventors have discovered that such devices in the prior art have significant limitations: they mostly employ open-loop or simple single-parameter closed-loop control strategies, and the output power of the RF generator is usually fixed or subject to limited adjustment based on only a single parameter.
[0005] This crude control method is ill-suited to the complex conditions in semiconductor device manufacturing. First, different types of integrated circuit boards have different characteristics, such as the copper foil area ratio, board thickness, and initial contamination level, which significantly alter the plasma loading characteristics. Second, fluctuations in process environment parameters, such as chamber pressure and gas flow rate, directly affect plasma stability. Third, abnormal conditions such as arc discharge and PCB surface temperature rise that may occur during processing threaten processing quality and equipment safety. Existing devices lack the ability to collaboratively sense and intelligently decide on these multiple parameters, resulting in inconsistent plasma processing effects and failing to meet the stringent requirements of high-reliability semiconductor device manufacturing.
[0006] Therefore, there is an urgent need for a plasma surface treatment device that can sense multiple process parameters in real time, intelligently evaluate the system status, and accurately adjust the radio frequency power accordingly, so as to ensure the surface quality of the integrated circuit board before it carries the semiconductor chip, thereby improving the yield and reliability of the entire semiconductor device manufacturing process. Summary of the Invention
[0007] The purpose of this invention is to provide a plasma surface treatment apparatus for integrated circuit board mounting, thereby solving the aforementioned problems.
[0008] This invention is implemented as follows: a plasma surface treatment apparatus for integrated circuit board surface mount processing includes one or more processors and a memory for storing one or more computer programs. The one or more computer programs are configured to be executed by the one or more processors, and the processors form the following functional modules: a basic feedback module, which outputs a plasma state coefficient based on the plasma's reflected power, density, and optical emission spectral intensity through a basic feedback model; a process environment assessment module, which outputs a process environment coefficient based on the absolute pressure of the process chamber and the process gas flow rate through a process environment assessment model; an arc-temperature adaptation module, which outputs an arc-temperature adaptation degree based on the plasma state coefficient, the process environment coefficient, the arc discharge count, and the PCB surface temperature through an arc-temperature adaptation model; a feedforward prediction module, which outputs a characteristic coefficient of the processed object based on the copper foil area ratio, overall thickness, and initial surface energy of the PCB board through a feedforward prediction model; and a power control module, which outputs a target power and adjusts the current power to the target power based on the characteristic coefficient of the processed object, the arc-temperature adaptation degree, and the RF generator efficiency through a power control model.
[0009] In a further technical solution, the power control model is configured such that the target power is positively correlated with the characteristic coefficient of the object being processed and the arc-temperature adaptability, and negatively correlated with the efficiency of the radio frequency generator.
[0010] A further technical solution, the specific steps of outputting the characteristic coefficient of the processed object through a feedforward prediction model based on the copper foil area ratio, overall thickness, and initial surface energy of the PCB board are as follows: normalize the copper foil area ratio, overall thickness, and initial surface energy of the PCB board to obtain the copper foil area ratio index, overall thickness index, and initial surface energy index; weight and combine the copper foil area ratio index, overall thickness index, and initial surface energy index to obtain the characteristic coefficient of the processed object, wherein the characteristic coefficient of the processed object is proportional to the copper foil area ratio index, overall thickness index, and initial surface energy index.
[0011] A further technical solution is that the copper foil area ratio, overall thickness, and initial surface energy of the PCB board are normalized as follows: the copper foil area ratio and overall thickness of the current PCB board are compared with their upper limit of specifications, and the upper limit is limited to 1 by a min function to obtain the copper foil area ratio index and the overall thickness index; the water contact angle, which represents the initial surface energy, is normalized by a maximum-minimum process to obtain the initial surface energy index.
[0012] A further technical solution, the specific steps for outputting the arc-temperature fit degree based on the plasma state coefficient, process environment coefficient, arc discharge count, and PCB surface temperature through an arc-temperature fit model, are as follows: The arc discharge count and PCB surface temperature are normalized to obtain the arc discharge count index and PCB surface temperature index; based on the plasma state coefficient, process environment coefficient, arc discharge count index, and PCB surface temperature index, the arc-temperature fit degree is calculated using a decay function, wherein the arc-temperature fit degree is configured to decay with the combined deterioration of the product of the plasma state coefficient and process environment coefficient, the arc discharge count index, and the PCB surface temperature index.
[0013] A further technical solution involves normalizing the arc discharge count and PCB surface temperature by: comparing the arc discharge count with the set maximum allowable arc rate, and using a min function to limit the upper limit to 1 to obtain the arc discharge count index; and performing maximum-minimum normalization on the PCB surface temperature to obtain the PCB surface temperature index.
[0014] A further technical solution involves the following steps for outputting the plasma state coefficient based on the plasma's reflection power, density, and optical emission spectral intensity using a fundamental feedback model: Normalizing the plasma's reflection power, density, and optical emission spectral intensity to obtain the reflection power index, density index, and optical emission spectral intensity index; and comprehensively calculating the reflection power index, density index, and optical emission spectral intensity index to obtain the plasma state coefficient. The plasma state coefficient is configured to be negatively correlated with the reflection power index, and its value is limited to the minimum value among the reflection power index, density index, and optical emission spectral intensity index.
[0015] A further technical solution involves normalizing the plasma's reflection power, density, and optical emission spectral intensity by performing maximum-minimum normalization on the current plasma's reflection power, density, and optical emission spectral intensity to obtain the reflection power index, density index, and optical emission spectral intensity index.
[0016] A further technical solution involves the following steps for outputting a process environment coefficient based on the absolute pressure of the process chamber and the process gas flow rate through a process environment assessment model: Normalizing the absolute pressure of the process chamber and the process gas flow rate to obtain the absolute pressure index and the process gas flow rate index; multiplying the absolute pressure index and the process gas flow rate index to obtain a process environment coefficient with an output range of 0-1. , The closer the value is to 1, the more stable the process environment.
[0017] A further technical solution involves normalizing the absolute pressure of the process chamber and the process gas flow rate by: calculating the absolute pressure index of the process chamber based on the deviation of the absolute pressure of the process chamber from its target set value; and calculating the process gas flow rate index based on the deviation of the process gas flow rate from its target set value.
[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By constructing a multi-module collaborative sensing and decision-making system that integrates basic feedback, process environment assessment, arc-temperature adaptation, and feedforward prediction, a comprehensive quantitative assessment of plasma state, process environment stability, system safety margin, and characteristics of the processed material is achieved, overcoming the sensing limitations of traditional open-loop or single-parameter control strategies.
[0019] 2. By employing unique model algorithms (such as state coefficient calculation based on geometric mean and minimum constraint, environment coefficient evaluation based on Gaussian membership function, and fitness calculation based on exponential decay function), the effective fusion of multi-source heterogeneous parameters is achieved, avoiding the evaluation distortion that may be caused by simple weighted averaging and significantly improving the accuracy of state evaluation.
[0020] 3. The feedforward prediction module enables predictive power regulation by pre-quantifying key PCB characteristics (copper foil area ratio, thickness, and initial surface energy). It can adjust the processing intensity in advance for circuit boards with different load characteristics, effectively preventing under-processing or over-processing.
[0021] 4. The power control model integrates processing requirements (characteristic coefficients of the processed object), safety constraints (arc-temperature compatibility), and equipment performance (RF generator efficiency) to achieve dynamic and precise control of RF power. While ensuring processing effect, it minimizes arc risk and thermal damage.
[0022] 5. The entire control strategy significantly enhances the device's adaptability to complex and variable operating conditions (such as different PCB models, process parameter fluctuations, and abnormal events), thereby ensuring a high degree of consistency in plasma surface treatment effects and improving the yield rate of integrated circuit board chip assembly and the reliability of the final semiconductor devices. Attached Figure Description
[0023] Figure 1 A flowchart of a plasma surface treatment apparatus for integrated circuit board mounting processing provided by the present invention. Detailed Implementation
[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0025] In the back-end packaging and system-level assembly stages of semiconductor device manufacturing, the cleanliness and chemical activity of the integrated circuit board surface have a decisive impact on mounting reliability. Existing plasma surface treatment equipment employs open-loop or simple single-parameter closed-loop control strategies, with the output power of the RF generator set at a fixed value or adjusted only according to a single parameter. This fails to effectively address the dynamic response requirements of different integrated circuit board models, fluctuations in process environment parameters, and abnormal conditions. Specifically, variations in the copper foil area ratio, board thickness, and initial contamination level significantly alter the plasma loading characteristics. Fluctuations in the absolute pressure and gas flow rate of the process chamber directly affect plasma uniformity and stability. Abnormal changes in arc discharge count and PCB surface temperature are difficult to assess and suppress in real time. Consequently, the plasma treatment effect is inconsistent, the wettability between solder and pads is weakened, and the occurrence of cold solder joints and poor solder joints increases, ultimately impacting the reliability and lifespan of semiconductor devices.
[0026] For example, when processing integrated circuit boards using high-density BGA packages, these boards have a high copper foil area ratio, a thin overall thickness, and a low initial surface energy. During plasma processing, fluctuations in the absolute pressure of the process chamber were monitored, the process gas flow rate exhibited instability, multiple arc discharges were recorded, and the PCB surface temperature rose abnormally. Existing control strategies cannot coordinately adjust to these multiple parameter changes, resulting in uneven plasma conditions, insufficient processing in some areas, poor pad wettability, and ultimately, solder joint defects after semiconductor chip mounting.
[0027] If the above problems are not solved, the quality consistency of the plasma surface treatment process cannot be guaranteed, the frequency of defects such as poor soldering and cold soldering will increase, the overall performance of semiconductor devices will be affected, the system stability will decrease, which may increase the risk of production interruption, reduce the safety margin of equipment, and ultimately weaken the yield and product reliability.
[0028] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.
[0029] like Figure 1 As shown, a plasma surface treatment apparatus for integrated circuit board surface mounting is provided in one embodiment of the present invention. It includes one or more processors and a memory for storing one or more computer programs. The one or more computer programs are configured to be executed by the one or more processors, and the following functional modules are formed on the processors: The basic feedback module outputs plasma state coefficients based on plasma reflection power, density, and optical emission spectrum intensity through a basic feedback model. The plasma state coefficient is a quantitative index that characterizes the stability of the plasma's working state. It can be calculated by weighted average of reflected power, density, and optical emission spectrum intensity, for example, by normalizing each parameter and summing it according to a preset weight; or by using a threshold-based rule judgment method, for example, outputting a high state coefficient value when the reflected power is lower than the first threshold and the density is higher than the second threshold.
[0030] The process environment assessment module outputs process environment coefficients based on the absolute pressure of the process chamber and the flow rate of the process gas through the process environment assessment model. The process environment factor is a quantitative indicator that reflects the stability of the process chamber environment. It can be achieved by linearly combining the absolute value of the deviation between the absolute pressure of the process chamber and the absolute value of the deviation between the absolute pressure and the target value of the process gas flow rate and the target value. For example, the weighted average of the pressure deviation and the flow rate deviation can be calculated. Alternatively, a lookup table method can be used to look up a preset environment factor table based on the actual values of pressure and flow rate.
[0031] The arc-temperature adaptation module, based on the plasma state coefficient and process environment coefficient, as well as the arc discharge count and PCB surface temperature, outputs the arc-temperature adaptation degree through the arc-temperature adaptation model. Arc-temperature fit is a quantitative indicator for evaluating the safety margin of a system. It can be achieved by using the minimum value of the ratio of arc discharge count to the allowable threshold and the ratio of PCB surface temperature to the safety threshold. For example, the smaller value of the two can be taken as the fit. Alternatively, a safety margin calculation method can be used, where the maximum fit value is output when the arc count is zero and the temperature is below the threshold, otherwise it is reduced proportionally.
[0032] The feedforward prediction module outputs the characteristic coefficients of the processed object based on the copper foil area ratio, overall thickness and initial surface energy of the PCB board through the feedforward prediction model. The characteristic coefficient of the processed object refers to a quantitative indicator that characterizes the intensity of the processing demand of the PCB board. It can be achieved by simply multiplying the copper foil area ratio, the overall thickness and the initial surface energy, for example, by multiplying the parameters after normalization; or by directly mapping the PCB model code to the preset coefficient value, for example, by querying the characteristic coefficients corresponding to different PCB models through a database.
[0033] The power control module, based on the characteristic coefficients of the processed object, the arc-temperature adaptability, and the efficiency of the radio frequency generator, outputs the target power through the power control model and adjusts the current power to the target power.
[0034] The power regulation module dynamically adjusts the RF power by comprehensively considering the characteristics of the processed object, the arc-temperature adaptability, and the efficiency of the RF generator. This avoids the limitations of open-loop or single-parameter control strategies in existing technologies and effectively solves the problem of poor consistency in processing results caused by different PCB characteristics, process environment fluctuations, and abnormal events.
[0035] The device comprises several key components: a basic feedback module, which can be implemented by an embedded processing unit integrating a spectrum analyzer and an RF power sensor, for high-precision acquisition of reflected power, plasma density, and optical emission spectral intensity; a process environment assessment module, which uses a high-precision pressure transmitter and a thermal gas mass flow meter in conjunction with a microprocessor, to analyze process chamber pressure and gas flow data in real time; an arc-temperature adaptation module, which uses a high-frequency current detection circuit to monitor arc discharge events and obtains the PCB surface temperature distribution using a non-contact infrared thermometer; a feedforward prediction module, which uses a machine vision system to analyze PCB images to determine the copper foil area ratio and combines a laser thickness gauge and a contact angle measuring instrument to obtain the overall thickness and initial surface energy parameters; and a power control module, which uses a programmable logic controller to execute the core control algorithm to precisely adjust the output power of the RF generator. Thus, through the deep integration of feedforward prediction and real-time feedback, the device enables the RF power to adapt to different PCB characteristics, process environment fluctuations, and potential abnormal operating conditions, avoiding the limitations of traditional open-loop control strategies. Furthermore, the predictive function of the characteristic coefficients of the processed object ensures that the power setting matches the actual needs of the PCB, effectively preventing under- or over-processing caused by differences in copper foil area ratio or contamination level. The dynamic evaluation mechanism of arc-temperature adaptability suppresses the risk of arc discharge and heat accumulation effect in real time, ensuring the continuous stability of the processing process. Specifically, the multi-parameter collaborative sensing system significantly improves the adaptability to complex working conditions, effectively controlling the consistency of plasma processing results, thereby solving the problem of processing quality fluctuations caused by coarse control strategies in existing technologies, and ultimately achieving reliable assurance of the surface treatment quality of integrated circuit boards.
[0036] like Figure 1 As shown, in a preferred embodiment of the present invention, the power control model is configured such that the target power is positively correlated with the characteristic coefficient of the object being processed and the arc-temperature adaptability, and negatively correlated with the efficiency of the radio frequency generator. The power regulation model can specifically be: ,in, For the target power, Based on the base power value, The characteristic coefficient of the processed object, For arc-temperature compatibility, For the efficiency of the radio frequency generator.
[0037] A power regulation model is a mathematical decision-making mechanism used to dynamically calculate target power. It can be implemented using an embedded microprocessor combined with a pre-stored algorithm program or a dedicated integrated circuit. Its purpose is to establish an accurate mapping relationship between multi-dimensional parameters and target power.
[0038] In one specific implementation, the power control module can be implemented using a microcontroller. This microcontroller periodically receives the processed object characteristic coefficients output from the feedforward estimation module, the arc-temperature adaptation degree output from the arc-temperature adaptation module, and the efficiency parameters fed back from the RF generator monitoring unit. It then calculates the target power value according to a preset mathematical model. During the calculation, the microcontroller uses the base power value as an initial reference, multiplies it sequentially by the processed object characteristic coefficients to enhance the processing intensity response, multiplies it by the arc-temperature adaptation degree to implement safety constraints, and divides it by the RF generator efficiency to calibrate energy conversion deviations. The final target power value is output to the RF generator's power control interface via a digital-to-analog converter. When the system detects an abnormally high arc discharge count, the arc-temperature adaptation degree rapidly decays, causing the target power to automatically decrease, effectively suppressing arc propagation. When the PCB copper foil area reaches the upper limit of the specification, the processed object characteristic coefficients approach 1, and the target power is correspondingly increased to the maximum safe value allowed by the process, ensuring sufficient surface activation.
[0039] Through the above technical solution, precise dynamic control of radio frequency power is achieved. It can adaptively optimize according to the actual characteristics of the PCB board and the real-time safety status of the system, effectively quantifying and balancing the contradictory relationship between processing intensity and safety margin. It avoids the problem of arc aggravation caused by over-adjustment of power or insufficient surface activation caused by insufficient power, thereby ensuring the cleanliness and chemical activity of PCB pads during plasma surface treatment and improving the reliability of semiconductor chip mounting and soldering.
[0040] like Figure 1 As shown, in a preferred embodiment of the present invention, the specific steps for outputting the characteristic coefficients of the processed object through a feedforward prediction model based on the copper foil area ratio, overall thickness, and initial surface energy of the PCB board are as follows: The copper foil area ratio and overall thickness of the current PCB board are compared with their upper limit, and the upper limit is limited to 1 by the min function to obtain the copper foil area ratio index and the overall thickness index. The copper foil area ratio index is a dimensionless parameter obtained by comparing the copper foil area ratio of the current PCB board with its upper limit of specifications and then using a min function with a limit of 1. It can be achieved by direct proportional calculation combined with threshold limitation. The purpose is to eliminate the absolute size difference between different PCB designs, convert the parameter into a relative load characteristic index, and avoid evaluation distortion caused by exceeding the specification limit.
[0041] The water contact angle, which characterizes the initial surface energy, is subjected to maximum-minimum normalization to obtain the initial surface energy index. The initial surface energy index is a standardized parameter obtained by performing maximum-minimum normalization on the water contact angle that characterizes the initial surface energy. It can be achieved by mapping the original water contact angle data to the [0,1] interval. The purpose is to unify surface energy parameters of different dimensions to a standardized range and accurately quantify the degree of PCB surface contamination and activation requirements.
[0042] The copper foil area ratio index, overall thickness index, and initial surface energy index are weighted and combined to obtain the characteristic coefficient of the processed object, wherein the characteristic coefficient of the processed object is directly proportional to the copper foil area ratio index, overall thickness index, and initial surface energy index. The weighted combination specifically involves importing the copper foil area ratio index, overall thickness index, and initial surface energy index into the formula. Obtain the characteristic coefficients of the processed object. , Output range 0-1, The larger the value, the higher the processing intensity required for the PCB. This is the copper foil area ratio index. This refers to the overall thickness index. The initial surface energy index, , and All are weighting coefficients with values ranging from 0 to 1, and The weighting coefficient is a parameter used to adjust the importance of different factors in decision-making. It can be assigned values through preset strategies or dynamic algorithms.
[0043] Specifically, the weighted combination formula is a mathematical expression that linearly combines the copper foil area ratio index, the overall thickness index, and the initial surface energy index through weighting coefficients. It can be implemented by microprocessor performing floating-point operations. The purpose is to allow the contribution of each parameter to be adjusted according to process experience through flexible configuration of weighting coefficients, while ensuring that the output range is strictly limited to 0-1 to avoid dimensional conflicts and numerical overflow.
[0044] As a specific implementation method, the solution of this application is implemented as follows: The feedforward estimation module uses an industrial-grade microcontroller as the processing unit. This processing unit receives copper foil area ratio data, board thickness measurement value, and water contact angle sensor output signal from the PCB inspection system. The processing unit first divides the copper foil area ratio and board thickness by their respective upper limit and applies a min function to limit the amplitude, generating a copper foil area ratio index and an overall thickness index. At the same time, the water contact angle data is subjected to maximum-minimum normalization processing to obtain an initial surface energy index. Subsequently, the processing unit calculates the characteristic coefficient of the processed object by linear weighting formula based on the preset weighting coefficients of the three indices. This coefficient is transmitted to the power control module in real time for dynamically adjusting the output power of the RF generator.
[0045] Through the above solution, this application can accurately quantify the differences in processing requirements of different PCBs, so that the RF power regulation can be matched with the actual load characteristics of the PCB, effectively improving the consistency and reliability of plasma surface treatment, and avoiding the problem of insufficient or excessive processing due to power mismatch.
[0046] like Figure 1 As shown, in a preferred embodiment of the present invention, the specific steps for outputting the arc-temperature fit degree based on the plasma state coefficient, process environment coefficient, arc discharge count, and PCB surface temperature using an arc-temperature fit model are as follows: The arc discharge count is compared with the set maximum allowable arc rate, and the upper limit is limited to 1 using the min function to obtain the arc discharge count index. The arc discharge count index is a technical feature that transforms raw arc count data into a standardized index through ratio processing. It can be achieved by real-time monitoring of arc pulse signals and dividing by a process safety threshold. The purpose is to quantify the relative severity of arc activity and avoid evaluation bias caused by differences in magnitude of the raw count data.
[0047] The PCB surface temperature is normalized to its maximum and minimum values to obtain the PCB surface temperature index. The PCB surface temperature index can be understood as a dimensionless parameter generated based on the temperature measurement boundary value. It can be achieved by linearly transforming the temperature data obtained by an infrared thermometer. The purpose is to eliminate the influence of the temperature dimension and enable the temperature rise effect to be directly integrated with other dimensionless parameters.
[0048] Based on the plasma state coefficient, process environment coefficient, arc discharge count index, and PCB surface temperature index, the arc-temperature fit is calculated using a decay function, wherein the arc-temperature fit is configured to decay with the combined deterioration of the product of the plasma state coefficient and the process environment coefficient, the arc discharge count index, and the PCB surface temperature index. The attenuation function is calculated by importing the arc discharge count index and the PCB surface temperature index into the formula. Obtain arc-temperature compatibility , Output range 0-1, The closer the value is to 1, the more secure and stable the system is. The plasma state coefficient, For process environment coefficient, The arc discharge count index. This refers to the PCB surface temperature index. The fitness attenuation width factor (a constant greater than 0) controls the fitness. The rate at which the system deteriorates. The smaller the value, the lower the system's tolerance to any anomalies, the more drastic the decrease in fitness, and the greater the fitness decay width factor. It can be understood as an adjustable parameter for controlling the risk sensitivity of the system. It can be implemented by using a preset constant or by dynamically configuring it according to the process type. The purpose is to enable the system to flexibly adjust its response characteristics to abnormal states according to actual process requirements.
[0049] As a specific implementation method, the solution of this application is implemented as follows: Arc discharge counting can be performed by real-time acquisition of arc pulse signals by a high-frequency current sensor and counting processing. The count data is then compared with the maximum allowable arc rate stored in the process database, and the result is limited to the range of 0-1 using a min function to form an arc discharge count index. The PCB surface temperature can be obtained by a non-contact infrared temperature measurement device. This temperature data is linearly normalized based on the upper and lower limits of the device's preset temperature measurement range to generate a PCB surface temperature index. The arc-temperature compatibility calculation unit can employ an embedded microprocessor, inputting the plasma state coefficient, process environment coefficient, arc discharge count index, and PCB surface temperature index into a preset attenuation function, wherein… The parameters are retrieved from the parameter library according to the current process type, and the final output is an arc-temperature compatibility value in the range of 0-1. This value is transmitted to the power control module in real time for target power calculation.
[0050] Through the above technical solution, this application realizes a refined and quantitative assessment of the safety status of the plasma processing system, enabling the arc-temperature compatibility to accurately reflect the combined impact of arc risk and temperature anomaly. This ensures that the power control module can implement gradual power adjustment in the early stages of system deterioration, effectively avoiding processing quality fluctuations or equipment safety hazards caused by response lag, and significantly improving the stability and reliability of the plasma surface treatment process.
[0051] like Figure 1 As shown, in a preferred embodiment of the present invention, the specific steps for outputting the plasma state coefficient based on the plasma's reflection power, density, and optical emission spectral intensity through a fundamental feedback model are as follows: The reflection power, density, and optical emission spectral intensity of the current plasma are all subjected to maximum-minimum normalization to obtain the reflection power index, density index, and optical emission spectral intensity index. Max-min normalization is a standardization method that linearly transforms the original physical quantities to a unified dimensionless range. It can be implemented using a linear scaling algorithm based on historical data extrema. Its purpose is to eliminate the incomparability caused by dimensional differences in reflected power, density, and optical emission spectral intensity, laying the foundation for multi-parameter fusion. The reflected power index is a normalized measure of reflected power, which can be understood as an indicator of the degree of radio frequency energy reflection. Specifically, it can be obtained by real-time acquisition of the reflected power signal and dividing it by a preset safety threshold. Its purpose is to directly reflect the influence of impedance matching state on plasma stability. The density index and optical emission spectral intensity index refer to the normalized plasma density and optical emission spectral intensity, respectively. Their introduction aims to provide a quantitative basis for plasma activity levels and chemical reaction intensity, and can be achieved by converting sensor output signals through calibration curves.
[0052] The plasma state coefficient is obtained by comprehensively calculating the reflection power index, density index, and optical emission spectral intensity index. The plasma state coefficient is configured to be negatively correlated with the reflection power index, and its value is limited to the minimum value among the reflection power index, density index, and optical emission spectral intensity index. The comprehensive calculation involves importing the reflection power index, density index, and optical emission spectral intensity index into the formula. Obtaining plasma state coefficients , Output range 0-1, The closer the value is to 1, the better the plasma state. The reflection power index, Density index It is the intensity index of the optical emission spectrum.
[0053] Specifically, the scheme in this application converts the reflection power, density, and optical emission spectral intensity into comparable indices through normalization processing, wherein... The term achieves a negative correlation mapping with reflected power to quantify the degree of impedance mismatch; geometric mean part The synergistic effect of the density index and the optical emission spectral intensity index smooths short-term fluctuations and avoids interference from transient noise on the state coefficients; the min term forces the state coefficients to be constrained by... The minimum value among the density index and optical emission spectrum intensity index is used to ensure that the assessment results truly reflect the weakest link. Finally, the formula outputs a state coefficient in the range of 0-1. This coefficient comprehensively characterizes the plasma health level and provides a high-fidelity feedback signal for the power control module, thereby avoiding the risk of arcing or inadequate handling caused by distortion in the state assessment.
[0054] As a specific implementation method, the solution of this application is implemented as follows: the signal acquisition unit of the basic feedback module may include a wideband reflected power monitor, a Langmuir probe density sensor and a fiber-coupled spectrometer, and the processing unit may use an embedded microcontroller to execute normalization and synthesis calculation algorithms; for example, the microcontroller is equipped with a non-volatile memory to store historical data extreme value parameters, and calculates the state coefficients in real time through a floating-point arithmetic unit, wherein the normalization processing is based on the pre-stored minimum-maximum threshold for dynamic adjustment, and the synthesis calculation module uses hardware acceleration circuits to realize the parallel operation of geometric mean and minimum value functions.
[0055] Through the above scheme, this application can accurately quantify the comprehensive health status of plasma, effectively solve the problem of state coefficient distortion caused by insufficient integration of multi-dimensional information and neglect of bottleneck effects in traditional methods, thereby improving the accuracy of power control and system stability, and ensuring the consistency and reliability of plasma surface treatment process in integrated circuit board chip processing.
[0056] like Figure 1 As shown, in a preferred embodiment of the present invention, the specific steps for outputting the process environment coefficient based on the absolute pressure of the process chamber and the process gas flow rate through the process environment assessment model are as follows: The absolute pressure index of the process chamber is calculated based on the degree of deviation between the absolute pressure of the process chamber and its target set value. Specifically, this involves importing the absolute pressure of the process chamber into the formula. Obtain the absolute pressure index of the process chamber. ,in, The absolute pressure of the process chamber. The target pressure value set for the process. The standard deviation allowed for pressure control.
[0057] The absolute pressure index of the process chamber is a quantitative indicator based on the degree of deviation between the absolute pressure of the process chamber and its target set value. It can be implemented using a Gaussian distribution model. Specifically, it can be achieved by calculating the square of the difference between the actual pressure value and the target value and applying an exponential decay function. Its purpose is to smoothly reflect the impact of pressure fluctuations on the stability of the process environment.
[0058] The process gas flow rate index is calculated based on the degree of deviation between the process gas flow rate and its target set value; Specifically, this involves: introducing the process gas flow rate into the public... Obtain the process gas flow index ,in, For process gas flow rate, The target flow rate value set for the process, The standard deviation allowed for flow control.
[0059] The process gas flow rate index is a quantitative indicator based on the degree of deviation between the process gas flow rate and its target setpoint. It can be implemented using the same Gaussian distribution model as the pressure index. Specifically, it can be achieved by the squared decay of the difference between the actual flow rate and the target value. Its purpose is to accurately quantify the impact of flow rate fluctuations on plasma formation. The Gaussian function can be understood as a continuously decaying function based on the degree of deviation. Its specific implementation can be in the form of an exponential function. Its purpose is to provide a smooth quantitative output and avoid the abrupt changes of traditional threshold judgment.
[0060] The absolute pressure index of the process chamber and the process gas flow rate index are multiplied to obtain the process environment coefficient, which has an output range of 0-1. , The closer the value is to 1, the more stable the process environment. Product processing refers to the operation of multiplying the absolute pressure index of the process chamber and the process gas flow rate index. This can be achieved using simple numerical multiplication. Its purpose is to comprehensively consider the interaction between pressure and flow parameters, as the stability of the process environment depends on the combined state of both.
[0061] As a specific implementation method, the solution of this application is implemented as follows: In the actual system, the absolute pressure of the process chamber is monitored in real time by a high-precision pressure sensor, and the process gas flow rate is measured by a mass flow controller; the processing unit uses a microcontroller to perform Gaussian function calculations, wherein the allowable standard deviation of pressure control and the allowable standard deviation of flow control are preset according to process requirements; the process environment coefficient is calculated by a software algorithm and used to adjust the radio frequency power in real time.
[0062] Through the above-mentioned scheme, this application can accurately quantify the deviation of process environment parameters, accurately reflect the stability of the process environment, thereby improving the accuracy of arc-temperature adaptability and power control, and ensuring the stability and reliability of the plasma surface treatment process.
[0063] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A plasma surface treatment apparatus for integrated circuit board mounting, comprising one or more processors and a memory for storing one or more computer programs, characterized in that, The one or more computer programs are configured to be executed by the one or more processors, and to form the following functional modules on the processors: The basic feedback module outputs plasma state coefficients based on plasma reflection power, density, and optical emission spectrum intensity through a basic feedback model. The process environment assessment module outputs process environment coefficients based on the absolute pressure of the process chamber and the flow rate of the process gas through the process environment assessment model. The arc-temperature adaptation module, based on the plasma state coefficient and process environment coefficient, as well as the arc discharge count and PCB surface temperature, outputs the arc-temperature adaptation degree through the arc-temperature adaptation model. The feedforward prediction module outputs the characteristic coefficients of the processed object based on the copper foil area ratio, overall thickness and initial surface energy of the PCB board through the feedforward prediction model. The power control module, based on the characteristic coefficients of the processed object, the arc-temperature adaptability, and the efficiency of the radio frequency generator, outputs the target power through the power control model and adjusts the current power to the target power.
2. The plasma surface treatment apparatus for integrated circuit board surface mounting according to claim 1, characterized in that, The power control model is configured such that the target power is positively correlated with the characteristic coefficient of the object being processed and the arc-temperature adaptability, and negatively correlated with the efficiency of the radio frequency generator.
3. The plasma surface treatment apparatus for integrated circuit board mounting according to claim 2, characterized in that, The specific steps for outputting the characteristic coefficients of the processed object through a feedforward prediction model based on the copper foil area ratio, overall thickness, and initial surface energy of the PCB board are as follows: The copper foil area ratio, overall thickness, and initial surface energy of the PCB board are normalized to obtain the copper foil area ratio index, overall thickness index, and initial surface energy index. The copper foil area ratio index, overall thickness index, and initial surface energy index are weighted and combined to obtain the characteristic coefficient of the processed object, wherein the characteristic coefficient of the processed object is directly proportional to the copper foil area ratio index, overall thickness index, and initial surface energy index.
4. The plasma surface treatment apparatus for integrated circuit board mounting according to claim 3, characterized in that, The copper foil area ratio, overall thickness, and initial surface energy normalization method of the PCB board are as follows: The copper foil area ratio and overall thickness of the current PCB board are compared with their upper limit, and the upper limit is limited to 1 by the min function to obtain the copper foil area ratio index and the overall thickness index. The water contact angle, which characterizes the initial surface energy, is subjected to maximum-minimum normalization to obtain the initial surface energy index.
5. The plasma surface treatment apparatus for integrated circuit board mounting according to claim 2, characterized in that, The specific steps for outputting the arc-temperature fit degree based on the plasma state coefficient, process environment coefficient, arc discharge count, and PCB surface temperature using the arc-temperature fit model are as follows: The arc discharge count and PCB surface temperature are normalized to obtain the arc discharge count index and PCB surface temperature index. Based on the plasma state coefficient, process environment coefficient, arc discharge count index, and PCB surface temperature index, the arc-temperature fit is calculated using a decay function, wherein the arc-temperature fit is configured to decay with the combined deterioration of the product of the plasma state coefficient and the process environment coefficient, the arc discharge count index, and the PCB surface temperature index.
6. The plasma surface treatment apparatus for integrated circuit board mounting according to claim 5, characterized in that, The normalization method for the arc discharge count and PCB surface temperature is as follows: The arc discharge count is compared with the set maximum allowable arc rate, and the upper limit is limited to 1 using the min function to obtain the arc discharge count index. The PCB surface temperature is normalized to its maximum and minimum values to obtain the PCB surface temperature index.
7. The plasma surface treatment apparatus for integrated circuit board mounting according to claim 5, characterized in that, The specific steps for outputting the plasma state coefficient based on the plasma's reflection power, density, and optical emission spectral intensity through the fundamental feedback model are as follows: The reflection power, density, and optical emission spectral intensity of the plasma are normalized to obtain the reflection power index, density index, and optical emission spectral intensity index. The plasma state coefficient is obtained by comprehensively calculating the reflection power index, density index, and optical emission spectral intensity index. The plasma state coefficient is configured to be negatively correlated with the reflection power index, and its value is limited to the minimum value among the reflection power index, density index, and optical emission spectral intensity index.
8. The plasma surface treatment apparatus for integrated circuit board mounting according to claim 7, characterized in that, The plasma's reflection power, density, and optical emission spectral intensity are normalized as follows: The reflection power, density, and optical emission spectral intensity of the current plasma are all subjected to maximum-minimum normalization to obtain the reflection power index, density index, and optical emission spectral intensity index.
9. The plasma surface treatment apparatus for integrated circuit board surface mounting according to claim 5, characterized in that, The specific steps for outputting the process environment coefficient based on the absolute pressure of the process chamber and the process gas flow rate through the process environment assessment model are as follows: The absolute pressure of the process chamber and the flow rate of the process gas are normalized to obtain the absolute pressure index and the flow rate index of the process chamber. The absolute pressure index of the process chamber and the process gas flow rate index are multiplied to obtain the process environment coefficient, which has an output range of 0-1. , The closer the value is to 1, the more stable the process environment.
10. The plasma surface treatment apparatus for integrated circuit board mounting according to claim 9, characterized in that, The method for normalizing the absolute pressure of the process chamber and the process gas flow rate is as follows: The absolute pressure index of the process chamber is calculated based on the degree of deviation between the absolute pressure of the process chamber and its target set value. The process gas flow rate index is calculated based on the degree of deviation between the process gas flow rate and its target set value.