Process optimization method and system applied to liquid resin wafer plastic package

Abstract

The invention relates to the technical field of wafer plastic packaging, in particular to a process optimization method and system applied to liquid resin wafer plastic packaging, and the method comprises the steps: obtaining the initial characteristics of liquid resin, the initial state of a release film, and the size and structure characteristics of a wafer; on the basis of the initial characteristics of the liquid resin and the initial state of the release film, release film preheating control parameters are determined, and a preheating procedure is carried out; after the preheating process is finished, performing feature recognition on the target area of the release film to obtain surface features of the release film and size features of the target area; and carrying out fluidity simulation on the surface characteristics of the release film, the size characteristics of the target area and the initial characteristics of the liquid resin, determining glue extrusion parameters according to a fluidity simulation result, and carrying out a glue extrusion process. The filling behaviors of resin under different working conditions are pre-judged in advance through fluidity simulation and a mold closing control model, and the problem of non-uniform filling caused by fixed parameters is solved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of wafer plastic packaging, and in particular to a process optimization method and system applied to liquid resin wafer plastic packaging. BACKGROUND

[0002] As a core material for semiconductor chip packaging, epoxy molding compound (EMC) accounts for more than 97% of the total amount of packaging materials, and plays an irreplaceable role in protecting chips from physical damage, isolating moisture and contaminants, and improving heat dissipation performance. With the development of semiconductor packaging towards high density and thinness, liquid resin is increasingly widely used in wafer-level plastic packaging due to its good flowability and adaptability to complex structures.

[0003] In the existing liquid resin wafer plastic packaging process, each process parameter is fixedly set in advance, and the differences in the initial characteristics of the liquid resin, the state of the release film and the wafer structure are not considered, which makes it difficult to adapt to the flow characteristics of the liquid resin, resulting in uneven filling of the liquid resin during the plastic packaging process, forming uneven surfaces and color difference defects, and there is a height difference of 2-3 microns in the color difference area. Such uneven surfaces not only affect the chip step difference accuracy, but also cause a large number of abnormal images during optical detection, causing equipment jamming or missed detection risk, which seriously reduces the production yield and efficiency. SUMMARY

[0004] The application provides a process optimization method and system applied to liquid resin wafer plastic packaging by flowability simulation and mold closing control model, which can effectively solve the problems in the background art.

[0005] To achieve the above-mentioned purpose, in a first aspect, the application provides a process optimization method applied to liquid resin wafer plastic packaging, comprising: obtaining the initial characteristics of the liquid resin, the initial state of the release film and the size and structure characteristics of the wafer; determining the release film preheating control parameters based on the initial characteristics of the liquid resin and the initial state of the release film, and performing the preheating process according to the parameters; after the preheating process is completed, identifying the features of the target area of the release film to obtain the surface features of the release film and the size features of the target area; performing flowability simulation on the surface features of the release film, the size features of the target area and the initial characteristics of the liquid resin, and determining the extrusion parameters according to the flowability simulation results, and performing the extrusion process according to the parameters; after the extrusion process is completed, obtaining the real-time characteristics of the liquid resin, and inputting the wafer size and structure characteristics into the plastic packaging mold closing control model to obtain the mold closing process parameters, and performing the mold closing process according to the parameters.

[0006] With reference to the first aspect, in a possible design of the first aspect, the preheating control parameter of the release film comprises a preheating temperature and a preheating time. The preheating temperature is determined based on a viscosity-temperature curve of the initial property of the liquid resin and a heat conduction performance of the initial state of the release film. The preheating time is determined based on a target area size feature of the release film and an expected flowability improvement time of the liquid resin at the preheating temperature, so that the target area of the release film can reach a temperature state suitable for the initial property of the liquid resin after preheating.

[0007] With reference to the first aspect, in a possible design of the first aspect, a calculation formula of the preheating temperature is as follows: ; wherein, represents the preheating temperature; represents a resin viscous flow activation energy; represents a gas constant; represents a target viscosity; represents a pre-exponential factor, which is used to quantify a reference value of a material property change with temperature; represents an ambient temperature.

[0008] With reference to the first aspect, in a possible design of the first aspect, a calculation formula of the preheating time is as follows: ; wherein, represents the preheating time; represents a release film density, which is determined based on the initial state of the release film; represents a specific heat capacity of the release film, which is determined based on the initial state of the release film; represents a target area feature size, which is an equivalent diameter of the target area and is calculated based on the target area size feature; represents a heat conduction coefficient of the release film, which is determined based on the initial state of the release film; represents an initial temperature difference; represents a final temperature difference.

[0009] With reference to the first aspect, in a possible design of the first aspect, the surface feature of the release film comprises a surface roughness, a surface flatness and a surface temperature.

[0010] With reference to the first aspect, in a possible design of the first aspect, the extrusion parameter comprises a spiral path, an extrusion speed and an extrusion pressure.

[0011] With reference to the first aspect, in a possible design of the first aspect, the mold closing process parameter comprises a mold closing speed and a mold closing pressure.

[0012] With reference to the first aspect, in a possible design, the formula of the flowability simulation is: ; wherein U represents the radial diffusion uniformity of the resin in the ring extrusion process; k r represents a path correction coefficient; P represents the extrusion pressure; represents the extrusion speed; d represents the maximum radius of the spiral trajectory; represents the dynamic viscosity of the resin; R a represents the surface roughness of the release film; represents a roughness influence coefficient; s represents the spiral spacing; and r represents the current radial position.

[0013] With reference to the first aspect, in a possible design, the mold clamping control model is constructed based on finite element analysis, and includes a geometric modeling module, a physical parameter module, and a numerical solution module. The geometric modeling module is configured to generate a three-dimensional geometric model consistent with an actual mold cavity according to wafer size structure characteristics. The physical parameter module is configured to store physical properties of the resin and mechanical parameters of the mold. The numerical solution module is configured to simulate the flow and curing behavior of the resin in the mold clamping process by using a fluid dynamics and curing dynamics coupling algorithm.

[0014] In a second aspect, the present application further provides a process optimization system applied to liquid resin wafer plastic packaging, comprising: a characteristic acquisition module configured to acquire initial characteristics of the liquid resin, an initial state of a release film, and wafer size structure characteristics; a preheating control module configured to determine release film preheating control parameters based on the initial characteristics of the liquid resin and the initial state of the release film, and control a preheating process based on the release film preheating control parameters; a target area identification module configured to identify characteristics of a target area of the release film after the preheating process, and obtain surface characteristics of the release film and size characteristics of the target area; an extrusion parameter determination module configured to simulate flowability of the surface characteristics of the release film, the size characteristics of the target area, and the initial characteristics of the liquid resin, and determine extrusion parameters based on a flowability simulation result; an extrusion execution module configured to execute an extrusion process based on the extrusion parameters; a mold clamping parameter determination module configured to acquire real-time characteristics of the liquid resin after the extrusion process, and input the real-time characteristics of the liquid resin and the wafer size structure characteristics into a plastic packaging mold clamping control model to obtain mold clamping process parameters; a mold clamping execution module configured to execute a mold clamping process based on the mold clamping process parameters.

[0015] The technical scheme of the present application can realize the following technical effects: the present application can obtain the initial characteristics of liquid resin, the release film state and the wafer structure features in real time, so that the preheating, extrusion and mold closing parameters can be accurately matched with the actual working conditions, and the uneven filling problem caused by the parameter rigidity can be reduced from the root; the preheating, extrusion and mold closing processes are linked to form a closed loop through feature acquisition and process parameter linkage, the preheating parameters are used as the liquidity influencing factors for the subsequent extrusion, the extrusion parameters are dynamically adjusted according to the release film state after preheating, and the mold closing parameters are optimized based on the real-time state of the resin after extrusion, so that the cumulative errors caused by independent control of each process can be avoided; through the liquidity simulation and mold closing control model, the filling behavior of the resin under different working conditions can be predicted in advance, the uneven filling problem caused by fixed parameters can be solved, and the surface concave-convex and color difference defects can be reduced. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The logic flowchart of the process optimization method for the wafer plastic packaging of liquid resin in the present application; Figure 2 The structural block diagram of the process optimization system for the wafer plastic packaging of liquid resin in the present application. DETAILED DESCRIPTION

[0017] The present application will be described below in combination with the drawings in the present application.

[0018] As shown in the drawings, Figure 1 The process optimization method for the wafer plastic packaging of liquid resin in the present application specifically includes the following steps: Step S100, obtaining the initial characteristics of liquid resin, the initial state of release film and the size structure features of wafer; Step S200, determining the release film preheating control parameters based on the initial characteristics of liquid resin and the initial state of release film, and performing the preheating process according to the parameters; Step S300, after the preheating process is completed, performing feature recognition on the target area of release film to obtain the surface features of release film and the size features of target area; the surface features of release film include surface roughness, surface flatness and surface temperature; the target area is the extrusion area of liquid resin; Step S400, performing liquidity simulation on the surface features of release film, the size features of target area and the initial characteristics of liquid resin, and determining the extrusion parameters according to the liquidity simulation results, and performing the extrusion process according to the parameters; the extrusion parameters include spiral path, extrusion speed and extrusion pressure; Step S500, after the extrusion process is completed, obtaining the real-time characteristics of liquid resin, and inputting the characteristics and the size structure features of wafer into the plastic packaging mold closing control model to obtain the mold closing process parameters, and performing the mold closing process according to the parameters; the mold closing process parameters include mold closing speed and mold closing pressure.

[0019] In the present embodiment, the present application matches the preheating, extrusion and mold closing parameters with the actual working conditions by acquiring the initial characteristics of the liquid resin, the release film state and the wafer structure features in real time, so as to reduce the uneven filling problem caused by rigid parameters from the root; the preheating, extrusion and mold closing processes form a closed loop through feature acquisition and process parameter linkage, the preheating parameters are used as the flowability influencing factors for the subsequent extrusion, the extrusion parameters are dynamically adjusted according to the release film state after preheating, and the mold closing parameters are optimized based on the real-time state of the resin after extrusion, so as to avoid the cumulative error caused by independent control of each process; through the flowability simulation and mold closing control model, the filling behavior of the resin under different working conditions is predicted in advance, the uneven filling problem caused by fixed parameters is solved, and the surface concave-convex and color difference defects are reduced; The parameter deviation of a single process in the existing process is amplified in the subsequent process, while the method realizes cross-process error compensation through the logic of providing an adaptive basis for the subsequent process by the previous parameters and adjusting the subsequent parameters based on the real-time state of the previous process, for example, when the surface temperature of the release film after preheating is slightly lower than expected, the extrusion speed can be adjusted through flowability simulation in the extrusion stage, and the pressure can be optimized through real-time characteristics in the mold closing stage, so that the small deviation of a single process is offset through multi-linkage cooperation, thereby improving the control precision of surface flatness; When facing different batches of liquid resin, release films with different aging degrees or special-shaped wafer structures, the existing fixed parameters will cause a sharp increase in defect rate due to the problem of fixed setting; while the method can automatically adapt to extreme differences through full-process feature acquisition and model calculation, for example, for the combination of high-viscosity resin and rough release film, the preheating time can be lengthened, the extrusion speed can be reduced, and the mold closing speed can be reduced, etc., and the three are linked to stabilize the product yield under complex working conditions; The isolated optimization of each process can only reduce a single defect, while the overall logic of the method through the prediction, adjustment and feedback cycle not only directly reduces the rework rate caused by surface defects, but also indirectly eliminates the chain costs caused by defects, such as the manpower cost of rejudging abnormal images in AOI detection and the adaptation cost of subsequent packaging processes caused by step difference, forming a positive cycle of process optimization, cost reduction and efficiency improvement, and the comprehensive benefits are better than the simple superposition of individual optimization of each process.

[0020] In some embodiments of the present application, for the acquisition of the initial characteristics of the liquid resin, first, an appropriate amount of sample is extracted from the batch of liquid resin to be used; for example, a certain volume of liquid resin is collected using a clean sampling tool to ensure that the sample is representative and can reflect the characteristics of the entire batch of resin; then, using a rotational viscometer, the viscosity, viscous flow activation energy, power law index and other non-Newtonian fluid characteristic parameters of the liquid resin are detected at a specific temperature and shear rate according to relevant industry standards; for example, the rotational speed of the viscometer and the test time are set, the viscosity values corresponding to different shear rates are recorded, and the key parameters such as viscous flow activation energy are calculated through data fitting and other methods; then, Fourier transform infrared spectroscopy, differential scanning calorimetry and other analysis methods are used; Fourier transform infrared spectroscopy can detect the types and contents of functional groups of the resin and determine whether its chemical structure meets the process requirements; differential scanning calorimetry can measure the glass transition temperature and reaction heat of the resin to determine the phase change and reaction characteristics of the resin during the plastic packaging process; at the same time, the density and surface tension of the resin can also be measured; the density can be measured by the specific gravity bottle method, the resin sample with a known volume is placed in the specific gravity bottle, and the weight is calculated using a balance; the surface tension is tested by the pendant drop method, capillary rise method, etc., and the surface tension will affect the spreading and filling behavior of the resin on the release film surface, serving as an input parameter for flow simulation.

[0021] For the acquisition of the initial state of the release film, first, the release film to be used is determined, and if the release film has a protective film, etc., it needs to be properly pretreated in a clean environment to remove the protective film, etc., and expose the effective surface in contact with the liquid resin; then, the thermal conductivity of the release film is measured using a thermal conductivity meter, and the specific heat capacity and other parameters are tested using differential scanning calorimetry; for example, the release film sample is shaped into a specified shape and placed in the thermal conductivity meter test chamber, and the thermal conductivity data is obtained according to the instrument operation process; the specific heat capacity change of the release film at different temperature intervals is obtained through differential scanning calorimetry testing, which is used for subsequent calculation of preheating control parameters; then, a high-precision temperature sensor is used to measure the initial temperature of the release film to ensure uniform distribution of the measurement points, and the average value is obtained through multiple measurements to ensure the accuracy of the temperature data; the surface cleanliness of the release film is detected by a visual detection system or a contact angle measuring instrument; visual detection can observe whether there are dust, oil stains and other pollutants on the surface, and the contact angle measuring instrument can indirectly judge the surface cleanliness and hydrophilic / hydrophobic property by measuring the contact angle of a liquid water droplet on the surface of the release film; if the contact angle is abnormal, it indicates that there may be pollutants or improper chemical modification on the surface, and the release film needs to be cleaned or replaced.

[0022] For the acquisition of the initial state of the release film, first, the wafer to be plastic encapsulated is placed on a high-precision bearing platform to ensure that the wafer is horizontal and stable, and can be fixed by vacuum suction, mechanical clamping, etc. to avoid the wafer moving in the subsequent measurement process affecting the accuracy of the data; the basic dimensions of the wafer such as diameter and thickness are measured using optical profilometers, laser range finders and other equipment; the optical profilometer can quickly scan the wafer surface to obtain wafer thickness distribution data to determine whether there is a problem of uneven thickness; the laser range finder can accurately measure the distance from the edge of the wafer to the center to determine whether the wafer diameter meets the process requirements; the wafer detection equipment, such as a wafer defect detection system, includes optical imaging, electron beam detection and other modules to identify the chip array layout on the wafer, the chip size, the spacing between chips, i.e. the target area size, and the step structure on the chip surface; for example, through high-resolution optical imaging, the geometric shape and arrangement of the chips are clearly presented, and the gap size between the chips is measured. The above chip structure features will affect the filling path of the liquid resin and the pressure distribution during mold closing, etc.

[0023] In the present embodiment, by standardizing and accurately acquiring the initial characteristics, the starting conditions of plastic encapsulation for each batch and each wafer are controllable; whether it is the replacement of liquid resin of different batches or the replacement of release film and wafer, the subsequent process parameters can be adjusted based on accurate initial data to ensure the consistency of the plastic encapsulation process under different production cycles and different production environments, and to reduce product quality differences caused by fluctuations in initial conditions; by mastering the initial state of the liquid resin, release film and wafer, risk factors that may affect the quality of plastic encapsulation can be identified in advance, such as abnormal resin viscosity, release film surface contamination, wafer structure defects, etc.; for adjustable factors such as resin viscosity which can be improved by pre-treatment, release film surface contamination which can be cleaned, etc., intervention can be made before the process starts; for non-adjustable factors such as the inherent structure of the wafer, subsequent process parameter adaptation should fully consider it to prevent defects such as bubbles, underfilling and uneven surface from the source.

[0024] In some embodiments of the present application, the preheating process of the release film has an important influence on the subsequent extrusion and mold closing. The fluidity and filling behavior of the liquid resin are significantly affected by temperature changes, and the initial state of the release film also changes the flow characteristics of the resin on its surface. Therefore, in order to ensure that the liquid resin can uniformly and smoothly fill the mold cavity during the extrusion process, it is necessary to accurately determine the preheating temperature and preheating time according to the initial characteristics of the liquid resin and the initial state of the release film, which helps the release film to reach a temperature state that is suitable for the liquid resin, thereby improving the fluidity and filling performance of the resin, reducing filling defects, uneven surface and color difference caused by insufficient or excessive preheating, and improving the overall performance of the wafer surface quality and plastic encapsulated products.

[0025] Specifically, the determination method of the preheating temperature is as follows: Step S211, determine the viscosity variation law of the liquid resin at different temperatures according to the viscosity-temperature curve of the initial properties of the liquid resin; the viscosity-temperature curve reflects the trend of the viscosity of the liquid resin changing with temperature; Step S212, calculate the resistance and efficiency of the release film to heat transfer in combination with the heat conduction performance of the initial state of the release film; the release film with poor heat conduction performance needs a higher preheating temperature or a longer preheating time to reach the required temperature state; Step S213, determine the target viscosity range that the liquid resin should reach after preheating according to the target flowability requirement; lower viscosity helps to improve the flowability of the resin, but it cannot be too low to affect the subsequent mold closing and curing effect; Step S214, based on the above factors, determine a suitable preheating temperature range through experimental data and simulation calculation, so that the release film and the liquid resin can reach a good matching state in the subsequent glue extrusion process.

[0026] More specifically, the calculation formula of the preheating temperature is: ; Among them, represents the preheating temperature; represents the resin viscous flow activation energy, which is determined by the initial properties of the liquid resin, by measuring the viscosity of the liquid resin at different temperatures, and is obtained by fitting the Arrhenius equation, reflecting the internal characteristics of the viscosity of the liquid resin changing with temperature; represents the gas constant, R=8.314 J / (mol·K); represents the target viscosity, which is adjusted according to the flowability requirement, and is selected by referring to process experience and experimental data, combined with factors such as wafer structure complexity, to ensure that the liquid resin has good flowability and filling performance in the glue extrusion process; represents the pre-exponential factor, which is used to quantify the benchmark value of the change of material properties with temperature, and is obtained by measuring the viscosity of the liquid resin and data fitting; represents the ambient temperature, which is measured by using a high-precision temperature sensor near the preheating equipment or in the production workshop, and records the real-time ambient temperature value.

[0027] In this embodiment, the calculation formula of the preheating temperature directly relates the viscous flow activation energy and the pre-exponential factor of the resin, making the preheating temperature strongly related to the characteristics of the resin material. For example, when the resin of different batches or formulations is replaced and the formula automatically calculates the new preheating temperature, the flowability problem caused by the traditional fixed temperature is avoided. The ambient temperature is introduced as a reference, so that the preheating temperature automatically adjusts with the fluctuation of the workshop environment temperature. For example, in the high-temperature environment in summer and the low-temperature environment in winter, the preheating temperature calculated by the formula will change accordingly. Through the formula calculation method, the complex material characteristics and process requirements are converted into specific numerical parameters, making the preheating process more controllable and repeatable. In different production batches, production lines, or when different types of liquid resin and release film are replaced, only the relevant parameters need to be measured and substituted into the formula to accurately determine the preheating temperature, enhancing the stability and consistency of the process.

[0028] The determination method of the preheating time is as follows: Step S221, considering the target area size characteristics of the release film, including the length, width, shape, etc. of the target area; a larger target area needs a longer preheating time to ensure that heat can be uniformly transmitted to the entire area; Step S222, analyzing the expected flowability improvement time of the liquid resin at the preheating temperature; different types of liquid resin have different response speeds to temperature, and the time required to reach the expected flowability also differs; Step S223, according to the thermal inertia of the release film and the heating efficiency of the heating equipment, calculating the time required for heat transfer to the target area of the release film during the preheating process; Step S224, combining the above factors, determining a reasonable preheating time range to ensure that the target area of the release film can reach a temperature state suitable for the initial characteristics of the liquid resin after preheating, while avoiding excessive preheating leading to energy waste or material property changes.

[0029] More specifically, the calculation formula of the preheating time is: ; Wherein, represents the preheating time; represents the density of the release film, which is measured from the initial state of the release film, measured by professional material testing instruments such as densitometers and differential scanning calorimeters, reflecting the thermal physical properties of the release film material; represents the specific heat capacity of the release film, which is measured from the initial state of the release film; represents the target area characteristic size, usually taking the equivalent diameter of the target area, measured and calculated by geometric calculation or image analysis software according to the shape and size of the target area of the release film, representing the size of the target area and the distance of heat transfer; Heat transfer coefficient of release film, measured from the initial state of the release film, tested according to the relevant standards using a thermal conductivity meter, reflecting the ability of the release film material to transfer heat; Initial temperature difference, i.e. ; Final temperature difference, i.e. , Temperature of the heating source.

[0030] In the present embodiment, the calculation formula of the preheating time comprehensively considers the thermal physical characteristic parameters of the release film, such as density, specific heat capacity, and heat transfer coefficient, as well as the size characteristics and temperature difference change of the target area, and can accurately calculate the time that meets the preheating requirement, ensuring that the release film reaches the expected temperature state after preheating, avoiding process problems caused by insufficient or excessive preheating time, and improving the quality and efficiency of the preheating process; by accurately calculating the preheating time, energy waste caused by excessive preheating is avoided, and energy consumption in the production process is reduced, which helps enterprises to save energy and reduce emissions, and reduce production costs; reasonable determination of the preheating time can reduce unnecessary preheating waiting time, improve the utilization rate of equipment and production efficiency; under the premise of ensuring the preheating quality, the preheating time is shortened as much as possible, making the entire plastic packaging process more compact and efficient, which is conducive to meeting large-scale production demand and shortening production cycle.

[0031] In some embodiments of the present application, for step S300, the roughness and flatness of the surface of the release film directly affect the flow and spreading behavior of the liquid resin during the glue extrusion process, and the surface temperature affects the viscosity and fluidity of the resin. If the preset glue extrusion parameters are directly used, the actual state of the release film and the characteristics of the resin may not match, causing uneven filling, bubbles and other defects. In addition, the size characteristics of the target area determine the size and shape of the area to be filled, which directly affects the flow path of the resin.

[0032] Specifically, the release film surface feature acquisition method is as follows: a. Surface roughness: use a laser confocal microscope or an atomic force microscope to scan the target area of the release film, the scanning range covers the entire glue extrusion area, and the step is not more than 0.5 microns; the key detection parameters are the arithmetic average roughness and the maximum profile height, and the distribution density of the surface micro convex or concave is recorded, such as the number of convexes greater than 1 micron per square millimeter; the purpose is that the roughness directly affects the flow resistance of the resin, such as high roughness which can cause a sudden drop in local flow rate, forming a filling dead angle, and the micro resistance parameter needs to be provided for the flowability simulation; b. Surface flatness: Use a white light interferometer to perform a global scan of the target area with a resolution not exceeding 0.1 microns, generating a three-dimensional topography map; calculate parameters including flatness error, i.e. peak-to-valley value, local tilt angle, such as the height difference between the edge and center of the target area; the purpose is that flatness deviation will cause the resin to climb or accumulate during the flow process, which needs to be corrected by detecting data in the flow simulation of the gravity field distribution; c. Surface temperature: Use high-precision temperature sensors such as thermocouples or infrared thermometers to measure the surface temperature of the target area of the release film; thermocouples can directly contact the surface of the release film for measurement, while infrared thermometers measure temperature non-contact by detecting infrared radiation emitted by the surface of the release film; place the temperature sensor at multiple locations on the target area of the release film to obtain comprehensive temperature distribution data; record the measurement data, calculate the average surface temperature, and analyze the uniformity of the temperature.

[0033] Target area size feature acquisition method, as follows: According to the size structure characteristics of the wafer and the extrusion process requirements, the target area on the release film is determined, i.e. the area where the liquid resin will be extruded; mark the boundary of the target area on the release film to facilitate subsequent measurement and identification; Use optical measuring equipment such as optical microscopes or image analyzers, or mechanical measuring tools such as vernier calipers or coordinate measuring machines to measure the size features of the target area, including length, width, area, shape, etc.; for complex-shaped areas, use image analysis software to process and analyze the measurement images to extract the required size information; record the measurement data and calculate the characteristic size of the target area, such as equivalent diameter, etc.

[0034] In this embodiment, by detecting the surface temperature distribution after preheating in real time, local areas that do not reach the temperature can be found, such as the edge where the temperature is lower due to fast heat dissipation, and the resin flow parameters can be adjusted accordingly in the flow simulation, such as increasing the extrusion speed in this area, to avoid filling defects caused by uneven preheating; surface roughness and flatness data can be converted into wall shear force coefficients in the flow simulation, making the simulation results more realistic.

[0035] As a preferred embodiment of the above, in the process of liquid resin wafer plastic packaging, the annular extrusion method moves the glue outlet in a spiral trajectory to uniformly extrude the liquid resin into the mold cavity, achieving efficient packaging of the wafer; the annular extrusion method moves the glue outlet in a spiral trajectory by a mechanical device; the glue outlet moves along the set spiral path to continuously extrude the liquid resin into the mold cavity; as the glue outlet moves and the glue continuously extrudes, the liquid resin gradually fills the mold cavity.

[0036] Specifically, to enable the liquid resin to uniformly fill the mold cavity and reduce defects such as uneven filling and bubbles, flow simulation analysis needs to be performed on the surface features of the release film, the size features of the target area, and the initial properties of the liquid resin, and then the extrusion parameters are obtained, including: Spiral path: The parameters such as the pitch, radius, and starting point position of the spiral will affect the filling effect and filling time of the liquid resin. Spiral path planning can ensure that the liquid resin uniformly fills the mold cavity and reduces defects such as uneven filling and bubbles. According to the size features of the target area, the spiral motion path of the glue port is planned, including the starting point, end point, and parameters such as the pitch and radius of the spiral. Extrusion speed: The extrusion speed determines the filling rate of the liquid resin. Too fast extrusion speed may lead to uneven filling or bubbles, and too slow will reduce production efficiency. By controlling the motor speed or hydraulic flow of the extrusion equipment, precise control of the extrusion speed is realized to ensure that the liquid resin can fill the mold cavity at an appropriate rate. Extrusion pressure: Extrusion pressure is the driving force for the flow of liquid resin. Suitable extrusion pressure can ensure that the liquid resin fills the mold cavity smoothly, but excessive pressure may cause overflow or damage to the mold. According to the viscosity of the liquid resin and the resistance of the mold cavity, adjust the extrusion pressure to make the liquid resin flow out smoothly and fill the mold cavity.

[0037] Further, the flow characteristics of the liquid resin are complex and are affected by multiple factors. The flow behavior of the liquid resin is not only affected by its initial properties, but also closely related to the surface features of the release film and the size features of the target area. In existing processes, the extrusion parameters are mostly set based on experience, which cannot accurately adapt to the dynamic changes of the above-mentioned multiple factors, and may easily lead to defects such as uneven resin filling, bubble residue, or incomplete filling. Through flow simulation, the flow state of the resin under complex boundary conditions can be predicted before actual extrusion, and then the extrusion parameters can be optimized based on the simulation results to ensure that the resin can uniformly fill the target area.

[0038] Specifically, the formula for liquid resin flow simulation is as follows: ; Where U represents the radial diffusion uniformity of the resin during the annular extrusion process, which characterizes the radial diffusion uniformity of the resin in the mold cavity. The closer the value of U is to 1, the more uniform the filling is. k r is the path correction coefficient, which is determined by the spiral starting point position. When the starting point is aligned with the center of the target area, k r =1, and when it is offset, k r <1. The larger the offset is, the smaller k r is. It is used to correct the influence of the starting position deviation of the spiral path on the diffusion uniformity to ensure that the glue starts to diffuse uniformly from the center area. P represents the extrusion pressure, which is the basic driving force for resin extrusion and directly determines the flow potential energy of the resin. It needs to match the resin viscosity and the resistance of the mold cavity. High viscosity resin needs higher P to ensure flow; V represents the extrusion speed, which is the moving speed of the glue outlet along the spiral trajectory. It is used to determine the rate and trajectory density of the glue extrusion. Too fast speed can easily cause local turbulent flow to generate bubbles, and too slow speed can reduce efficiency and may cause defects due to premature curing of the resin; d represents the maximum radius of the spiral trajectory, which corresponds to the size characteristics of the target area. It needs to cover the wafer to be packaged. d needs to match the radius of the target area. Too large may cause overflow of the mold cavity, and too small may not completely cover the wafer.

[0039] η represents the dynamic viscosity of the resin, which is the internal friction of the resin flow at temperature T. It is affected by temperature T and related to the preheating time of the release film. The longer the preheating time, the higher the T, and the lower the dynamic viscosity of the resin. The dynamic viscosity of the resin under different preheating times is determined by experiment; R a R represents the surface roughness of the release film, which is the arithmetic mean of the microscopic concave-convex of the release film surface. It affects the flow resistance of the resin. Too high roughness will increase the friction resistance of the resin flow. The influence needs to be offset by adjusting the extrusion pressure or speed.

[0040] K represents the roughness influence coefficient, which quantifies the amplification effect of roughness on resistance. It is obtained by experiment fitting; s represents the spiral pitch, which is the distance between adjacent spiral trajectories. Too small pitch can easily cause glue accumulation, and too large pitch can easily form filling dead angles. It needs to match the resin viscosity. Low viscosity resin can appropriately increase s, and high viscosity resin needs to reduce s; r represents the current radial position, which is the distance from the center of the mold cavity to the current detection point, reflecting the diffusion state of different areas. The edge area needs higher diffusion power. In the formula, U is reduced by increasing r, prompting the adjustment of the extrusion parameters of the edge area, such as increasing the local pressure or speed; For the above formula, the numerator represents the positive factor driving the uniform diffusion of the resin, which is used to comprehensively reflect the dynamic strength and coverage efficiency of the spiral extrusion. The larger the value, the easier the resin diffuses uniformly. The denominator represents the negative factor hindering the uniform diffusion of the resin, which is used to comprehensively reflect the total resistance of the resin flow. The larger the value, the more difficult it is to ensure uniform diffusion. By calculating the U value at different radial positions, it can be judged whether the ring extrusion parameters are reasonable: if the U value of a certain area is too large, the extrusion speed of that area needs to be increased or decreased; if the U value of a certain area is too small, P needs to be increased or s needs to be reduced to ensure sufficient diffusion of the resin; finally, through optimization of the parameters, the U value of the whole area is consistent, realizing the uniform filling of the resin.

[0041] More specifically, the process of determining the extrusion parameters in combination with the simulation results needs to aim at making the resin diffuse uniformly in the target area; according to the process target, the U value deviation of each position in the target area needs to be controlled within the preset deviation threshold, the smaller the deviation, the more uniform the filling, if the simulation shows that the local U value is abnormal, the extrusion parameters need to be adjusted accordingly, and the specific adjustment is as follows: Determination of spiral path parameters: including pitch, radius and starting point; if the simulation shows that the U value of a certain radial area is too small, it indicates that the resin diffusion in this area is insufficient, which may lead to filling dead angle or color difference, the spiral pitch needs to be checked whether it is too large, the spiral pitch can be reduced to make the adjacent glue tracks more dense and enhance the resin coverage in this area; if the U value of the edge area is continuously small, it may be that the spiral radius does not cover the edge of the target area, the spiral radius needs to be appropriately increased, within the range of not exceeding the cavity range; if the simulation shows that the local U value is too large, it indicates that the resin accumulates in this area, which may lead to height difference, the spiral pitch needs to be checked whether it is too small, causing excessive superposition of glue, the spiral pitch can be increased to reduce the local glue amount; if the U value of the center area is too large, it may be that the spiral starting point deviates from the center of the target area, the starting point needs to be adjusted to the center position to make the diffusion uniformly advance from the center to the edge; Determination of extrusion speed: if the simulation shows that the overall U value is too small, it indicates that the resin diffusion is slow and the filling efficiency is low, which means that the extrusion speed is too slow, the spiral movement speed of the glue outlet needs to be increased to speed up the glue extrusion frequency and promote the resin diffusion; but the speed needs to be combined with the resin characteristics to avoid excessive speed causing local turbulence and causing bubbles or surface ripples; if the simulation shows that the local U value rises sharply, such as at the turning point of the spiral track, it indicates that the extrusion speed and the path curvature at this position do not match, that is, the speed is not adapted to cause glue accumulation when turning, the local speed of the turning area needs to be reduced, the spiral movement speed can be dynamically adjusted by mechanical device to restore the U value to be stable; Determination of extrusion pressure: if the simulation shows that the overall U value is low and uniformly distributed, it indicates that the resin flowability is insufficient and the overall filling is slow, the extrusion pressure needs to be increased to enhance the flow dynamics of the resin and reduce the resistance caused by viscosity; but the pressure needs to be controlled within the range that the mold can withstand; if the simulation shows that the U value of the edge area is significantly lower than the center, it may be that the resistance of the edge area cavity is large, the local extrusion pressure of the edge area needs to be increased accordingly to ensure that the resin can diffuse to the edge and reduce the height difference.

[0042] In the embodiment, the defects caused by the dependence of the existing process on the experience setting parameters are effectively solved by the liquidity simulation and accurate extrusion parameter optimization. On the one hand, the dynamic adaptation of the multi-factor coupling influence is realized, the dynamic changes of the initial characteristics of the liquid resin, the release film surface features and the target area size features are comprehensively considered, and the problems such as uneven filling, bubble residue, surface height difference and color difference are reduced from the root; on the other hand, the filling effect is quantified by the radial diffusion uniformity, the influence of each parameter on the diffusion is clear by means of the formula, the spiral path, the extrusion speed and the extrusion pressure can be adjusted, the uniform filling of the resin in the target area is ensured, and the wafer surface flatness is improved; the process stability and yield are improved, the rework and AOI detection complex judgment pressure caused by defects are reduced, and the production efficiency is improved.

[0043] In some embodiments of the application, the mold closing process is a key link for compacting and shaping the resin after extrusion; if the mold closing speed is too fast during the mold closing process, it will cause a violent impact on the liquid resin in the mold cavity, resulting in turbulent flow or local accumulation of the resin under the sudden pressure drop, which will destroy the uniform distribution formed in the extrusion process; if the mold closing pressure is improper, too high will cause the resin to be excessively extruded and overflow the mold cavity, and too low will not be able to eliminate bubbles or ensure that the resin is fully attached to the wafer surface, eventually exacerbating the unevenness of the wafer surface and color difference; therefore, based on the real-time characteristics of the resin and the wafer structure, the mold closing parameters are accurately determined by the plastic encapsulation mold closing control model to ensure uniform solidification of the resin and realize controllable wafer surface flatness, which is implemented as follows: Step S501, after the extrusion process is completed, the real-time viscosity, temperature and flow speed of the liquid resin are collected by online viscometer, temperature sensor and other devices; the above data can reflect the actual state changes of the liquid resin during the extrusion process; Step S502, the size and structure feature data of the wafer are retrieved from the wafer detection system or the production database, including the diameter, thickness, chip layout, wiring level and other information of the wafer; the above data has been obtained and stored by optical measurement equipment or other detection means during the wafer manufacturing process; Step S503, the mold closing control model is initialized and set according to the size and structure features of the wafer and the initial characteristics of the liquid resin, the geometry, boundary conditions and initial state of the mold cavity are defined; for example, the length, width and height of the mold cavity are set in the mold closing control model, the positions of the mold cavity inlet and outlet are defined, and the viscosity, temperature and flow speed of the liquid resin at the initial time of mold closing are set as parameters; Step S504, the mold closing control model simulates the flow front position, pressure change curve and filling time of the liquid resin under different mold closing speeds and pressures according to the input real-time characteristics of the liquid resin and wafer size structure characteristics, and selects the optimal parameter combination, i.e., the optimal mold closing speed and mold closing pressure, that can meet the filling quality requirements, to ensure that the liquid resin can be uniformly distributed and fully filled without defects during the mold closing process. Step S505, according to the mold closing speed and mold closing pressure output by the mold closing control model, the operating parameters of the mold closing equipment are set, such as the hydraulic pressure and mechanical transmission speed of the mold closing machine. Step S506, the mold closing equipment is started, and the mold is closed according to the set mold closing speed and pressure, so that the liquid resin is uniformly distributed in the mold cavity and solidified and formed; during the mold closing process, the changes of the mold closing pressure and speed are monitored in real time to ensure that the mold closing process is carried out according to the predetermined process parameters; for example, the pressure and speed during the mold closing process are monitored in real time by a pressure sensor and a displacement sensor, and feedback control and adjustment are performed in time if there is deviation.

[0044] In this embodiment, by accurately calculating the mold closing speed and pressure, the liquid resin can be uniformly distributed in the mold cavity, and defects such as internal stress concentration, uneven surface, bubble residue and the like caused by improper mold closing parameters can be reduced; based on the accurate control of the real-time characteristics of the liquid resin and the wafer size structure characteristics, process fluctuations caused by material characteristics and size differences can be reduced, and the stability and consistency of the mold closing process can be improved; for example, when switching between different batches of wafers or different types of liquid resin, only the relevant data need to be collected and input into the model, and the appropriate mold closing process parameters can be quickly and accurately determined to ensure the stable operation of the process.

[0045] In some embodiments of the present application, the mold closing control model is a numerical simulation tool for accurately calculating the mold closing speed and mold closing pressure; based on the real-time characteristics of the liquid resin and the wafer size structure characteristics, the flow state of the resin under different mold closing parameters is simulated, and the optimal mold closing speed and pressure that can ensure uniform solidification and no defects of the resin are output, to realize closed-loop control from parameter calculation, simulation verification to process execution. The mold closing control model is constructed based on finite element analysis, and includes a geometric modeling module, a physical parameter module and a numerical solution module; the geometric modeling module generates a three-dimensional geometric model consistent with the actual mold cavity according to the wafer size structure characteristics, and accurately restores the length, width, height and internal steps, wiring and the like of the mold cavity; the physical parameter module stores the physical characteristics of the resin and the mechanical parameters of the mold; the numerical solution module has a fluid dynamics and solidification dynamics coupling algorithm built in, which is used to simulate the flow and solidification behavior of the resin during the mold closing process.

[0046] According to the cavity geometry parameters, data is called from the wafer detection system, the boundary and internal structure of the cavity are defined in the model, the initial state of the resin is set, and the real-time viscosity, temperature, flow speed and other data collected after the glue squeezing are input into the model as the resin state parameters at the initial moment of the mold closing; the constraint conditions are defined, and the rigid boundary of the mold and the curing threshold of the resin are set.

[0047] The mold closing control model simulates the resin flow behavior through the following logic: the flow simulation is based on the non-Newtonian fluid constitutive equation, the flow front position and pressure distribution of the resin under different mold closing speeds are calculated, and it is judged whether there is turbulent flow or accumulation; the curing simulation combines the curing kinetics equation of the resin, calculates the curing rate of the resin under different mold closing pressures, and avoids premature curing caused by excessive pressure or incomplete curing caused by too low pressure.

[0048] The mold closing control model simulates through multiple parameter combinations, and selects mold closing process parameters that meet the following conditions: the resin flow front uniformly advances without local stagnation; the pressure distribution in the cavity is stable without sudden rise and fall; the degree of curing reaches the preset value at the end of mold closing.

[0049] In this embodiment, a multi-module collaborative system is constructed through finite element analysis, the real-time characteristics of the resin and the structure characteristics of the wafer are combined, the resin flow and curing state under different mold closing parameters are accurately simulated, the optimal mold closing speed and pressure are selected, the closed-loop control from parameter calculation to process execution is realized, the problems such as turbulent flow, accumulation, glue overflow and bubbles caused by excessive mold closing speed or improper pressure are avoided, the uniform curing of the resin is ensured, the unevenness and color difference of the wafer surface are effectively improved, and the product quality stability is improved; at the same time, the model can be adapted to different resin batches and wafer types, reduce process fluctuations, facilitate repeated implementation, and help improve production efficiency and yield.

[0050] As shown in Figure 2 , the present application also provides a process optimization system applied to liquid resin wafer plastic packaging, which specifically comprises the following modules: A feature acquisition module is used to acquire the initial characteristics of the liquid resin, the initial state of the release film, and the size and structure characteristics of the wafer. A preheating control module is used to determine the release film preheating control parameters based on the initial characteristics of the liquid resin and the initial state of the release film, and control the execution of the preheating process. A target area identification module is used to identify the characteristics of the target area of the release film after the preheating process, and obtain the surface characteristics of the release film and the size characteristics of the target area. An extrusion parameter determination module is used to simulate the flowability of the release film surface characteristics, the target area size characteristics, and the initial characteristics of the liquid resin, and determine the extrusion parameters according to the flowability simulation results. An extrusion execution module is used to execute the extrusion process according to the extrusion parameters. A mold closing parameter determination module is configured to acquire real-time characteristics of the liquid resin after the glue extrusion process ends, and input the wafer size structure characteristics into a plastic package mold closing control model to obtain mold closing process parameters. A mold closing execution module is configured to execute the mold closing process according to the mold closing process parameters.

[0051] In the embodiment, the feature acquisition module, the preheating control module, the target area identification module, the glue extrusion parameter determination module, and the mold closing parameter determination module are connected in series to realize the full-process linkage from initial feature acquisition to dynamic adjustment of process parameters. Unlike the traditional fixed parameter system, the system can capture the differences in real-time characteristics of the liquid resin, the release film state, and the wafer structure, and convert these differences into adaptive process parameters, thereby avoiding the problem of uneven filling caused by one-size-fits-all parameters from the root cause. The target area identification module is specially designed to identify the surface characteristics of the preheated release film and the size of the glue extrusion area, thereby reducing the glue extrusion deviation caused by unknown surface state. The glue extrusion parameter determination module generates adaptive parameters through flow simulation to solve the problems of bubbles and insufficient glue caused by the sensitivity of the flow characteristics of the liquid resin. The mold closing parameter determination module adjusts the mold closing speed and pressure based on the real-time characteristics of the resin after glue extrusion to improve the surface concave-convex and color difference defects, and indirectly improve the AOI detection yield.

[0052] The basic principles, main features and advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above embodiments, and the above embodiments and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A process optimization method for liquid resin wafer encapsulation, characterized in that: include: Obtain the initial properties of the liquid resin, the initial state of the release film, and the wafer-scale structural characteristics; Determining release film preheating control parameters based on the initial properties of the liquid resin and the initial state of the release film, and performing a preheating process accordingly; After the preheating process is completed, the target area of ​​the release film is characterized to obtain the surface characteristics of the release film and the size characteristics of the target area; Performing fluidity simulation on the surface characteristics of the release film, the size characteristics of the target area, and the initial characteristics of the liquid resin, and determining extrusion parameters based on the fluidity simulation results, and performing the extrusion process accordingly; After the extrusion process is completed, the real-time characteristics of the liquid resin are obtained and input into the plastic encapsulation mold control model together with the wafer size and structure characteristics to obtain the mold closing process parameters, and the mold closing process is carried out accordingly.

2. The process optimization method for liquid resin wafer molding according to claim 1, characterized in that: The release film preheating control parameters include preheating temperature and preheating time; The preheating temperature is determined based on the viscosity-temperature curve of the initial characteristics of the liquid resin and the thermal conductivity of the release film in its initial state; The preheating time is determined based on the size characteristics of the target area of ​​the release film and the expected fluidity improvement time of the liquid resin at the preheating temperature, so that the target area of ​​the release film can reach a temperature state that is compatible with the initial characteristics of the liquid resin after preheating.

3. The process optimization method for liquid resin wafer molding according to claim 2, characterized in that: The calculation formula of the preheating temperature is: ； in, Indicates preheating temperature; It represents the resin viscosity flow activation energy; represents the gas constant; Indicates target viscosity; It represents the pre-exponential factor, which is used to quantify the base value of the change of material properties with temperature; Indicates the ambient temperature.

4. The process optimization method for liquid resin wafer encapsulation according to claim 3, characterized in that: The calculation formula for the preheating time is: ； in, Indicates the warm-up time; Indicates the density of the release film, which is measured from the initial state of the release film; represents the specific heat capacity of the release film, which is measured from the initial state of the release film; Indicates the characteristic size of the target area, taking the equivalent diameter of the target area, which is calculated from the size characteristics of the target area; Indicates the thermal conductivity of the release film, measured from the initial state of the release film; represents the initial temperature difference; Indicates the final temperature difference.

5. The process optimization method for liquid resin wafer encapsulation according to claim 4, characterized in that: The surface characteristics of the release film include surface roughness, surface flatness and surface temperature.

6. The process optimization method for liquid resin wafer molding according to claim 5, characterized in that: The extrusion parameters include a spiral path, an extrusion speed and an extrusion pressure.

7. The process optimization method for liquid resin wafer encapsulation according to claim 6, characterized in that: The mold closing process parameters include mold closing speed and mold closing pressure.

8. The process optimization method for liquid resin wafer encapsulation according to claim 6, characterized in that: The formula for the liquidity simulation is: ； Where U represents the radial diffusion uniformity of the resin during the annular extrusion process; k r represents the path correction coefficient; P represents the extrusion pressure; Indicates the extrusion speed; d indicates the maximum radius of the spiral trajectory; Indicates the dynamic viscosity of the resin; R a Indicates the surface roughness of the release film; represents the roughness influence coefficient; s represents the spiral pitch; and r represents the current radial position.

9. The process optimization method for liquid resin wafer encapsulation according to claim 7, characterized in that: The mold clamping control model is constructed based on finite element analysis and includes a geometric modeling module, a physical parameter module and a numerical solution module; The geometric modeling module is used to generate a three-dimensional geometric model consistent with the actual mold cavity based on the wafer size and structural characteristics; The physical parameter module is used to store the physical properties of the resin and the mechanical parameters of the mold; The numerical solution module has a built-in fluid dynamics and curing dynamics coupling algorithm to simulate the flow and curing behavior of the resin during the mold closing process.

10. A process optimization system for liquid resin wafer encapsulation, characterized in that: include: Feature acquisition module, used to obtain the initial characteristics of liquid resin, the initial state of release film and wafer size structure characteristics; a preheating control module, configured to determine release film preheating control parameters based on the initial characteristics of the liquid resin and the initial state of the release film, and to control the execution of a preheating process accordingly; The target area recognition module is used to perform feature recognition on the target area of ​​the release film after the preheating process is completed, and obtain the surface features of the release film and the size features of the target area; a glue extrusion parameter determination module, configured to simulate the fluidity of the release film surface characteristics, the target area size characteristics, and the initial characteristics of the liquid resin, and determine the glue extrusion parameters based on the fluidity simulation results; A glue extrusion execution module, used for executing the glue extrusion process according to the glue extrusion parameters; A mold closing parameter determination module is used to obtain the real-time characteristics of the liquid resin after the extrusion process is completed, and input the real-time characteristics and the wafer size and structure characteristics into the plastic encapsulation mold closing control model to obtain the mold closing process parameters; The mold closing execution module is used to execute the mold closing process according to the mold closing process parameters.

Images