A multi-chip package offset compensation method and system
By acquiring material expansion and contraction characterization data and actual alignment offset data, a quantitative correspondence is established to determine the target packaging material and its offset compensation parameters. This solves the position offset problem caused by material mismatch during multi-chip packaging, thereby improving packaging accuracy and yield.
Patent Information
- Application Number
- CN202610966555.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-01
- Publication Date
- 2026-07-31
AI Technical Summary
In the process of multi-chip packaging, factors such as the mismatch of thermal expansion coefficients between the packaging material and the chip and dielectric layer, differences in chip arrangement, and stress release after debonding and bonding can cause the reconstructed wafer to expand, shrink, warp, or deform locally, affecting chip position offset and photolithography pattern alignment accuracy, and thus affecting the product's electrical performance and manufacturing yield.
By acquiring material expansion and contraction characterization data and actual alignment offset data of candidate packaging materials, a quantitative correspondence is established to determine the target packaging material and its offset compensation parameters. The target packaging material is used for offset compensation, including adjusting the resin matrix, curing components and additive components, to form a multi-chip packaging structure. Offset compensation is then performed before photolithography.
This enables the selection of packaging materials to shift from relying on experience to predictive selection based on measurable parameters, reducing the cost of offset measurement throughout the process and improving the accuracy of alignment and manufacturing yield.
Smart Images
Figure CN122497407A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of semiconductor packaging technology, and in particular to a method and system for offset compensation in multi-chip packaging. Background Technology
[0002] With the development of high-density integrated packaging technology, Fanout multi-chip co-packaging technology is widely used in multi-chip integrated products. In this type of process, multiple chips usually need to go through process steps such as reconstruction, EMC encapsulation and curing, debonding, dielectric layer coating and curing, and photolithography.
[0003] In actual production, due to factors such as EMC curing shrinkage, mismatch of thermal expansion coefficients between packaging materials and chips and dielectric layers, chip arrangement differences, stress release after debonding, and dielectric layer curing stress, reconstructed wafers are prone to expansion, warping, or local deformation. These deformations can further cause chip positions or photolithography pattern positions to shift, making it difficult for patterns to accurately fall into the target position during subsequent alignment processes, thereby affecting the electrical performance of the product and manufacturing yield.
[0004] The information disclosed in this background section is intended only to enhance the understanding of the general background of this disclosure and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention
[0005] This invention provides a method and system for offset compensation in multi-chip packaging, which can effectively solve the problems in the background art.
[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: Multi-chip package offset compensation method, the method includes: Acquire material expansion and contraction characterization data for multiple candidate packaging materials. The material expansion and contraction characterization data is used to characterize the dimensional change characteristics of the candidate packaging materials during the packaging process. Each candidate packaging material is used to form a corresponding multi-chip package structure, and the actual alignment offset data of each multi-chip package structure is obtained. Based on the material expansion and contraction characterization data and actual alignment offset data of each candidate packaging material, the target packaging material and the offset compensation parameters corresponding to the target packaging material are determined. The target packaging material is used to form a multi-chip package structure to be compensated, and offset compensation is performed on the alignment process of the multi-chip package structure to be compensated based on the offset compensation parameters.
[0007] Further, the target packaging material and the corresponding offset compensation parameters are determined, including: Establish the correspondence between the material expansion and contraction characterization data of each candidate packaging material and the corresponding actual alignment offset data, and determine the target packaging material and offset compensation parameters based on the correspondence.
[0008] Furthermore, the candidate encapsulation material is formed by modifying at least one of the resin matrix, curing component, filler component, or additive component.
[0009] Furthermore, the material expansion and contraction characterization data include the curing dimensional change rate and the coefficient of thermal expansion. The curing dimensional change rate is determined based on the ratio of the dimensional difference of the candidate encapsulation material before and after curing to the dimensional difference before curing. The coefficient of thermal expansion is determined based on the ratio of the dimensional change of the candidate encapsulation material per unit initial dimension within a preset temperature range to the temperature change.
[0010] Furthermore, the actual alignment offset data is determined based on the graphic displacement data of the offset magnification measurement graphic.
[0011] Furthermore, the offset magnification measurement pattern includes a reference pattern and a pattern to be measured; The reference pattern has a first pattern period, and the pattern to be tested has a second pattern period, and the first pattern period and the second pattern period are different; The graphic displacement data includes the relative displacement between the reference graphic and the graphic to be measured. The actual alignment offset data is determined based on the relative displacement and the period difference between the first graphic period and the second graphic period.
[0012] Furthermore, the correspondence between the material expansion and contraction characterization data of each candidate packaging material and the corresponding actual alignment offset data is established, including: Before the polyimide layer formed on each multi-chip package structure is cured, the first thermal expansion data and the first alignment offset data of each multi-chip package structure are obtained. After the polyimide layer is cured, the second thermal expansion data and the second alignment offset data of each multi-chip package structure are obtained. The polyimide curing-induced thermal expansion change data are determined based on the second thermal expansion data and the first thermal expansion data, and the polyimide curing-induced offset change data are determined based on the second paraposition offset data and the first paraposition offset data. Based on the data of polyimide curing-induced thermal expansion and polyimide curing-induced offset, a polyimide curing offset contribution parameter is generated, and the polyimide curing offset contribution parameter is used as a component of the corresponding relationship.
[0013] Furthermore, the correspondence between the material expansion and contraction characterization data of each candidate packaging material and the corresponding actual alignment offset data is established, including: Collect data samples from multiple batches and groups of candidate packaging materials or multiple groups of packaging process conditions, and store the data samples in the offset database; The data samples include material identification, material expansion and contraction characterization data, and actual alignment offset data; The mapping relationship between material expansion and contraction characterization data and actual alignment offset data is established based on the offset database.
[0014] Furthermore, the target packaging material and offset compensation parameters are determined based on the correspondence, including: Based on the mapping relationship in the offset database, candidate packaging materials and their predicted alignment offset data that match the packaging process conditions of the multi-chip packaging structure to be compensated are determined. The target packaging material and offset compensation parameters are determined based on the predicted alignment offset data.
[0015] A multi-chip package offset compensation system includes: The material data acquisition module is used to acquire material expansion and contraction characterization data of multiple candidate packaging materials; The offset data acquisition module is used to acquire the actual alignment offset data of the multi-chip package structure formed using each candidate packaging material; The parameter determination module is used to determine the target packaging material and the offset compensation parameters corresponding to the target packaging material based on the material expansion and contraction characterization data and actual alignment offset data of each candidate packaging material. The compensation module is used to perform offset compensation on the alignment process of the multi-chip package structure to be compensated, which is formed using the target packaging material, based on the offset compensation parameters.
[0016] The technical solution of this invention can achieve the following technical effects: By quantitatively correlating the intrinsic expansion and contraction characteristics of the packaging material with the final alignment offset data generated by the packaging structure, the selection of packaging materials is transformed from relying on trial and error based on experience to predictive selection based on measurable parameters. By separating the polyimide curing-induced offset as an independent parameter from the total offset and establishing a correspondence including this parameter, the offset compensation parameter can reflect the contributions of the two different sources: the intrinsic expansion and contraction of the packaging material and the curing of the dielectric layer. After forming a mapping relationship between the material expansion and contraction characterization data and the actual alignment offset data based on the offset database, it is only necessary to query the predicted offset data that matches the current packaging process conditions and take its negative value as the compensation parameter before alignment processing. It is not necessary to remeasure the offset for each batch of products, thus reducing the implementation cost of offset measurement throughout the entire process.
[0017] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description
[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 A flowchart illustrating a multi-chip package offset compensation method; Figure 2 This is a schematic diagram of a multi-chip package offset compensation system. Detailed Implementation
[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.
[0022] like Figure 1 As shown, this application provides a multi-chip package offset compensation method, including: Step S1: Obtain material expansion and contraction characterization data for multiple candidate packaging materials. The material expansion and contraction characterization data is used to characterize the dimensional change characteristics of the candidate packaging materials during the packaging process. Step S2: Use each candidate packaging material to form the corresponding multi-chip packaging structure, and obtain the actual alignment offset data of each multi-chip packaging structure; Step S3: Determine the target packaging material and the offset compensation parameter corresponding to the target packaging material based on the material expansion and contraction characterization data and actual alignment offset data corresponding to each candidate packaging material. Step S4: Use the target packaging material to form the multi-chip package structure to be compensated, and perform offset compensation during the alignment process of the multi-chip package structure to be compensated based on the offset compensation parameters.
[0023] In this embodiment, a quantitative correspondence is established between the material expansion and contraction characterization data obtained in step S1 and the actual alignment offset data obtained in step S2 through step S3. This correspondence directly links the intrinsic curing shrinkage and thermal expansion behavior of the encapsulation material with the measured position offset in the final encapsulation structure, so that the offset compensation parameter determined in step S3 simultaneously carries the comprehensive contribution of multiple factors such as the material's own expansion and contraction, stress transmission of the encapsulation structure, and offset induced by the curing of the dielectric layer. Step S4 applies the compensation parameter in reverse to the alignment processing process, realizing a closed-loop chain from material selection to compensation execution.
[0024] In a specific implementation, as one example, given that the final alignment offset in a real multi-chip packaging structure is affected by a combination of process factors, the dimensional changes of the packaging material during the curing reaction and subsequent heating process are independently measurable variables that can be intervened through material formulation adjustments. However, there is no directly usable quantitative correspondence between this variable and the final offset. Therefore, this variable needs to be characterized separately to provide a basis for material selection and compensation parameter determination. This embodiment directly measures the curing dimensional change rate and thermal expansion coefficient of candidate packaging materials under simulated process conditions to form material expansion and contraction characterization data, as detailed below: Step S11: Select at least one set of candidate encapsulation materials. Each set of encapsulation materials is formed by adjusting at least one of the resin matrix, curing component, filler component, or additive component. Mix each set of candidate encapsulation materials uniformly according to its recommended ratio, and after vacuum degassing, prepare sheet-like samples of specified sizes using molding or injection molding processes. The minimum size of the sample is not less than 100 times the resolution of the measuring equipment, and the maximum size is not greater than the effective working space of the thermomechanical analyzer furnace. The thickness-to-diameter ratio of the sample is not greater than 1:50, so that the measurement results mainly reflect the expansion and contraction behavior in the in-plane direction. Prepare at least 5 parallel samples for each candidate encapsulation material. Step S12: Measure the initial dimensions of the prepared sheet-like sample in a specific direction. Measurements are performed using an optical non-contact measuring device, such as a three-dimensional measuring system equipped with a telecentric lens and a micrometer displacement stage. The measurement temperature is controlled at 25 degrees Celsius. Mark at least three measurement positions on each sample surface, and repeat the measurement three times at each position, taking the arithmetic mean as the sample's pre-curing dimension. Place the sample in a curing process identical to the target packaging process, ensuring the curing temperature profile, heating rate, holding time, and cooling rate are consistent with the curing conditions used in multi-chip packaging fabrication. After curing, measure the dimensions at the marked positions again at the same temperature of 25 degrees Celsius. The curing dimension change rate is determined by the ratio of the dimension difference before and after curing to the dimension before curing, expressed as a percentage. For each sample, take the average of the change rates at each measurement position; for each candidate packaging material, take the average and standard deviation of all parallel samples. Step S13: For the same batch of sheet samples prepared, after complete curing, a thermomechanical analyzer is used for measurement; the test mode is selected as expansion mode, and a constant load is applied; the sample is placed in the furnace of the thermomechanical analyzer, and the temperature is increased from room temperature to a preset temperature range at a set constant heating rate. This heating rate should ensure uniform temperature distribution inside the sample, for example, a rate within the range of 1 degree Celsius to 10 degrees Celsius per minute can be selected; the preset temperature range should cover the entire temperature range that occurs in the target packaging process, for example, from 25 degrees Celsius to 200 degrees Celsius; the length change of the sample during the heating process is recorded; based on the recorded length-temperature data, a curve of length versus temperature is plotted; the inflection point where the slope changes significantly is identified on the curve, along the glassy state segment and the rubbery state segment respectively. Plot tangents and determine the temperature corresponding to the intersection of the two tangents as the glass transition temperature of the material. Calculate the coefficient of thermal expansion for two temperature ranges based on the length change data of the cured sample during the heating process. The first temperature range is below the glass transition temperature, with its upper limit at least 20 degrees Celsius lower than the glass transition temperature of the sample. The second temperature range is above the glass transition temperature, with its lower limit at least 20 degrees Celsius higher than the glass transition temperature of the sample. For each temperature range, the coefficient of thermal expansion is determined by the ratio of the change in size per unit initial dimension to the change in temperature within that range. Measure at least three samples for each candidate encapsulation material, measuring each sample once, and take the arithmetic mean of the calculated values for each sample within each temperature range as the coefficient of thermal expansion for that material in that range. Step S14: Use the obtained curing dimensional change rate and the obtained thermal expansion coefficient of each temperature range as material expansion and contraction characterization data of a set of candidate encapsulation materials.
[0025] In this embodiment, the curing dimensional change rate directly reflects the degree of volume shrinkage of the encapsulation material due to chemical cross-linking during the curing reaction. This shrinkage is converted into in-plane displacement of the chip surface in a multi-chip encapsulation structure. The coefficients of thermal expansion in the two temperature ranges reflect the thermal response behavior of the material in the glass and rubber states, respectively. The glass state corresponds to the dimensional stability of the encapsulation structure in working environments at room temperature and below, while the rubber state corresponds to the dimensional changes in high-temperature processes such as dielectric layer curing and solder reflow. By combining the above three parameters into a set of material expansion and contraction characterization data, the dimensional change driving capability of the material at different process stages can be described, thereby providing a physical basis for establishing a segmented correspondence between material properties and final alignment offset.
[0026] In a specific implementation, as one example, after obtaining the material expansion and contraction characterization data, it is necessary to obtain the actual alignment offset data generated by each candidate packaging material in the actual packaging structure in order to establish a quantitative correspondence between the intrinsic properties of the material and the offset consequences. Existing alignment offset measurement methods mostly use single-cycle alignment marks, such as cross marks or square marks, to obtain offset data by directly measuring the offset of the mark center coordinates. When the reconstructed wafer undergoes overall expansion and contraction or local deformation, the measurement results of this method simultaneously include two components: pattern displacement and pattern scaling, which are difficult to separate. When the expansion and contraction of the packaging material causes the overall size of the wafer to change, the alignment mark itself also scales proportionally with the wafer. At this time, the measured mark center offset cannot accurately reflect the true positional deviation of the chip or the pattern to be processed relative to the exposure grid. Based on the above problems, this embodiment decouples the two effects of displacement and scaling through a dual-cycle structure, and uses each candidate packaging material to fabricate a corresponding multi-chip packaging structure, and obtains the actual alignment offset data on this structure. The specific implementation is as follows: Step S21: Design an offset magnification measurement pattern, which includes two sub-patterns: a reference pattern and a pattern to be measured. The reference pattern uses a periodically arranged array of lines or dots, and its period is denoted as the first pattern period. The pattern to be measured uses a periodic arrangement structure of the same type as the reference pattern but with a different period, and its period is denoted as the second pattern period. Under ideal conditions with no offset and no deformation, the relative positional relationship between the two sub-patterns is a known design value. Step S22: Create patterns of the reference pattern and the pattern to be tested on the mask respectively, and set the relative positional relationship of the two patterns on the mask according to the ideal state. Step S23: Select each group of candidate packaging materials for which material expansion and contraction characterization data has been obtained. For each group of candidate packaging materials, form a multi-chip packaging structure according to the same reconstructed wafer fabrication process. The process includes: mounting multiple chips on a temporary bonding substrate according to a predetermined layout; encapsulating the chips with the candidate packaging materials to form a reconstructed wafer; fabricating at least three parallel reconstructed wafers for each group of candidate packaging materials; curing the packaging materials according to their preset curing program; removing the temporary bonding substrate; coating a dielectric layer on the reconstructed wafer and fabricating photolithographic alignment marks on the surface of the dielectric layer; the dielectric layer can be polyimide. Step S24: Using the fabricated mask, an offset magnified test pattern is formed on the dielectric layer surface of the reconstructed wafer through photolithography. The exposure conditions of the photolithography process are consistent with the photolithography conditions in the actual packaging process. After development, a magnified test pattern consisting of a reference pattern and a test pattern is formed on the surface of the reconstructed wafer. The pattern is fabricated in at least three different regions on each reconstructed wafer. The three regions are located at the wafer center, the wafer edge, and the middle position between the center and the edge, respectively. Step S25: Using high-resolution optical measurement equipment, such as a measurement microscope equipped with autofocus and sub-pixel resolution image processing algorithms, image the magnified measurement pattern; for each measurement area, acquire images of the reference pattern and the pattern to be measured respectively; use image processing algorithms to extract the feature positions of the two patterns; for a periodically arranged line array, extract the center line position of each line; for a dot matrix, extract the centroid position of each point; calculate the relative offset between the feature positions of the reference pattern and the feature positions of the pattern to be measured within the same measurement area; this offset is represented in the form of a two-dimensional vector, including a component along a first direction and a component along a second direction, the two directions being perpendicular to each other and parallel to the wafer plane; Step S26: Let the first graphic period be P1, the second graphic period be P2, and the relative displacement between the measured reference graphic and the graphic to be measured be ΔL. Then, the actual alignment offset data D is determined by the following formula: ; This formula uses the first graphic period P1 as the reference; if the second graphic period P2 is used as the reference, the formula is: ; The two formulas differ in result by a sign and a numerical factor. In practical use, one formula should be selected and kept consistent based on the coordinate system definition and graphic design. In the above two formulas... >0, and ; Step S27: Measure at least three regions for each wafer and take the arithmetic mean of all measurements as the final actual alignment offset data for that group of materials.
[0027] In this embodiment, by designing two periodic patterns with unequal periods, an analytical linear relationship is formed between the measured relative displacement ΔL and the overall wafer scaling and in-plane displacement: when the overall scaling acts alone, ΔL is proportional to |P1-P2|, and when the in-plane displacement acts alone, ΔL is proportional to either P1 or P2. Thus, by multiplying ΔL by the reciprocal of the difference between P1 and P2 and then by one of P1 or P2, the contribution of the overall scaling in the actual alignment offset data D is canceled out, and only the contribution of the in-plane displacement is retained, thereby achieving decoupling between the alignment offset measurement and the overall wafer scaling effect.
[0028] In one embodiment, candidate encapsulation material A was selected as the EMC material formulation for testing. Candidate encapsulation material A includes epoxy resin, curing agent, inorganic filler, and additives, wherein the epoxy resin accounts for 18.0% by mass, the curing agent accounts for 7.0% by mass, the spherical silica filler accounts for 72.0% by mass, and the coupling agent and stress-modifying additive account for 3.0% by mass. The curing conditions for candidate encapsulation material A are 175°C for 120 minutes.
[0029] Material expansion and contraction characterization tests were performed on candidate packaging material A, and the following data were obtained:
[0030] The curing dimensional change rate is calculated as follows: Cured dimensional change rate = (Test length after curing L1 - Test length before curing L0) / Dimension before curing L0 × 100% Right now: Cured dimensional change rate = (99.936 - 100.000) / 100.000 × 100% = -0.064% Furthermore, a Fanout multi-chip package structure was fabricated using candidate packaging material A, and PI coating and curing were completed after debonding. The PI curing conditions were 250℃ for 60 min, and actual alignment offset data were acquired by offset amplification measurement after PI curing.
[0031] In this embodiment, a reference pattern and a pattern to be tested are included. The period of the first pattern P1 is 10.00 μm, the period of the second pattern P2 is 10.20 μm, the period difference ΔP is 0.20 μm, and the corresponding theoretical magnification is approximately 50 times.
[0032] After PI curing, the actual alignment offset data for each region are as follows:
[0033] Based on the above test results, the statistical results of the actual alignment offset after PI for candidate packaging material A are as follows:
[0034] The results above show that the curing dimensional change rate of candidate encapsulation material A is -0.064%, the thermal expansion coefficient in the low-temperature region is 13.8 ppm / ℃, and the average actual alignment offset after PI curing is +0.39 μm in the X direction and -0.29 μm in the Y direction. Based on the material expansion and contraction characterization data and actual alignment offset data of candidate encapsulation material A, it can be used as a sample data in the offset database, and offset compensation parameters corresponding to the material formulation can be further generated.
[0035] In the subsequent photolithography alignment process, the X-direction compensation parameter can be set to -0.39μm and the Y-direction compensation parameter can be set to +0.29μm based on the offset data of the candidate packaging material A. For cases where the offset in the edge region is large, a partitioned compensation method can be further adopted, setting independent compensation parameters for regions such as East, North-East, and South-East to reduce the impact of local nonlinear offset on the photolithography alignment accuracy.
[0036] In a specific implementation, as one example, the target packaging material and its corresponding offset compensation parameters are determined so that the compensation parameters can be directly called for alignment correction in mass production, avoiding the need to remeasure the offset for each batch of products. Existing offset compensation methods mostly use fixed compensation coefficients or linear correction based on a single variable, without distinguishing the contributions of different material sources to the final offset. When the packaging structure contains multiple materials and undergoes multiple thermal processes, a single compensation coefficient cannot adapt to the differences in offset behavior under different material combinations. Based on this, this embodiment establishes a correspondence between material expansion and contraction characterization data and actual alignment offset data, and incorporates the offset contribution during the dielectric layer curing process as an independent parameter into this correspondence, thereby achieving the decomposition and quantification of the offset source. The specific implementation is as follows: Step S31: Using the candidate packaging materials for which material expansion and contraction characterization data has been obtained in Step S1, form corresponding multi-chip packaging structures according to the method in Step S2, and obtain the actual alignment offset data of each structure before the dielectric layer is coated, which is recorded as the first alignment offset data; at the same time, before the dielectric layer is coated on each multi-chip packaging structure, measure its thermal expansion behavior and obtain the first thermal expansion data; the first thermal expansion data is obtained by placing the packaging structure on a temperature-controlled platform and measuring its in-plane dimension change within the range of room temperature to a preset temperature, and is used to characterize the thermal response characteristics of the packaging structure without a dielectric layer; the preset temperature range should be consistent with the temperature range experienced during the curing process of polyimide, or at least cover the curing temperature range; Step S32: After completing each multi-chip package structure, a polyimide layer is coated; the coating thickness and curing procedure of the polyimide layer are consistent with the dielectric layer process in the target packaging process; after the polyimide layer is completely cured, the actual alignment offset data of each package structure is measured again and recorded as the second alignment offset data; at the same time, the thermal expansion behavior is repeatedly measured in the same temperature range to obtain the second thermal expansion data. Step S33: For each group of candidate encapsulation materials, subtract the first thermal expansion data from the second thermal expansion data to obtain polyimide curing-induced thermal expansion change data; this difference reflects the amount of change in thermal expansion characteristics of the polyimide layer due to chemical shrinkage and thermal mismatch with adjacent layer materials during the curing process; subtract the first alignment offset data from the second alignment offset data to obtain polyimide curing-induced offset change data; this difference reflects the alignment offset increment contributed solely by the polyimide layer. Step S34: Perform a correlation analysis between the obtained polyimide curing-induced thermal expansion change data and the polyimide curing-induced offset change data. The correlation method is as follows: using the curing-induced thermal expansion change data as the independent variable and the curing-induced offset change data as the dependent variable, the proportionality coefficient between the two is calculated using a linear regression method. This proportionality coefficient is denoted as the polyimide curing offset contribution parameter, and its physical meaning is the offset change corresponding to a unit thermal expansion change. This parameter is independent of the intrinsic expansion and contraction characteristics of the encapsulation material and is only related to the dielectric layer material itself and its curing process. Step S35: Pair the curing dimensional change rate and thermal expansion coefficient of each candidate encapsulation material obtained in step S1 with the actual alignment offset data corresponding to each candidate encapsulation material obtained in step S2 to form a set of correspondence pairs; in this correspondence, the generated polyimide curing offset contribution parameter is introduced as an offset correction term; the correction method is: for a certain candidate encapsulation material, its actual alignment offset data is equal to the sum of the intrinsic expansion and contraction contribution of the material and the polyimide curing offset contribution parameter multiplied by the measured thermal expansion change of the encapsulation structure; Step S36: For each group of candidate encapsulation materials, repeat steps S31 to S35 for at least three independent batches. Each batch involves independent sample preparation, measurement, and calculation. Store the data samples of each batch in an offset database. Each data sample includes material identification, curing dimensional change rate, glassy thermal expansion coefficient, rubbery thermal expansion coefficient, polyimide curing offset contribution parameter, and actual alignment offset data. After accumulating data samples from at least five different candidate encapsulation materials, establish a mapping relationship based on this database. The mapping relationship can be in the form of a multidimensional lookup table indexed by material identification, or a multiple linear regression model can be used with material expansion and contraction characterization data as input and actual alignment offset data as output. Step S37: Obtain the packaging process conditions of the multi-chip package structure to be compensated. These conditions include the expected expansion and contraction characteristics range of the packaging material used, the type of dielectric layer and the curing process, and the chip layout density. Based on these packaging process conditions, search the offset database for the candidate packaging material that best matches them. The matching criterion is that the deviation between the material expansion and contraction characterization data of the candidate packaging material and the expected expansion and contraction characteristics of the structure to be compensated does not exceed a preset threshold, for example, the deviation is less than 10%. After a successful search, the candidate packaging material is determined as the target packaging material, and the actual alignment offset data corresponding to the candidate packaging material in the database is used as the predicted alignment offset data. The negative value of the predicted alignment offset data is determined as the offset compensation parameter.
[0037] In this embodiment, by separating the dielectric layer curing-induced offset as an independent parameter from the total offset and establishing a correspondence including the polyimide curing offset contribution parameter, the final determined offset compensation parameter can reflect the contributions from two different sources: the intrinsic expansion and contraction of the encapsulation material and the dielectric layer curing. When the dielectric layer process of the encapsulation structure to be compensated is consistent with the sample dielectric layer process in the database, the offset compensation parameter can be directly called without re-measuring the dielectric layer contribution. The construction of the offset database enables the reuse of data from multiple batches and multiple material systems, reducing the dependence on performing full-process offset measurements for each batch of products.
[0038] In a specific implementation, as one example, this embodiment achieves offset compensation by applying the determined target packaging material and offset compensation parameters to the fabrication and alignment process of the multi-chip packaging structure to be compensated. Specifically, the following operations are performed: Step S41: Based on the determined target packaging material, obtain actual batch samples of the material; Step S42: Form a multi-chip package structure to be compensated using the target packaging material; mount multiple chips on a temporary bonding substrate according to a predetermined layout; encapsulate the chips using the selected target packaging material through molding or injection molding processes to form a reconstructed wafer; process the wafer in the curing process; after curing is completed, remove the temporary bonding substrate and coat the reconstructed wafer with a dielectric layer. Step S43: Obtain the offset compensation parameters of the multi-chip package structure to be compensated, and input the offset compensation parameters into the lithography equipment used to fabricate the redistribution layer; the input method includes: manual input into the equipment control terminal, automatic download through the manufacturing execution system, or import through the storage medium; after receiving the offset compensation parameters, the lithography equipment superimposes them onto the exposure grid coordinates; the superposition method is: before exposure, the lithography equipment adds the original exposure position coordinates to the negative value of the offset compensation parameters to obtain the corrected exposure position coordinates; the lithography equipment performs exposure according to the corrected coordinates; Step S44: On the lithography equipment with the compensation parameters imported, the redistribution layer of the multi-chip package structure to be compensated is processed by lithography. During the processing, the alignment system of the lithography equipment first identifies the alignment marks on the wafer, calculates the measured offset, and then compares the measured offset with the offset compensation parameters. If the absolute value of the difference between the measured offset and the offset compensation parameters exceeds the preset threshold, the lithography equipment issues an alarm and suspends processing. The operator then determines whether it is necessary to remeasure or re-determine the compensation parameters. If the difference is within the preset threshold range, the lithography equipment completes the exposure according to the corrected exposure position coordinates. Step S45: After completing the photolithography process, select at least three wafers, and select at least five different positions on each wafer, such as the wafer center, edge, and the middle position between the center and the edge, to measure the residual offset between the actual pattern position and the designed position; calculate the average value after taking the absolute value of the residual offset of each measurement point; if the average value exceeds the preset qualified threshold, it is determined that the compensation is insufficient, and the measured offset in this process is stored as a new data sample in the offset database for the purpose of updating the offset compensation parameters in subsequent batches.
[0039] In this embodiment, the offset compensation parameter acts directly on the exposure coordinates of the lithography equipment in the form of a two-dimensional vector. Its physical meaning is to predict the negative value of the alignment offset data, that is, to move the exposure pattern in the opposite direction of the offset by the same distance before the alignment process. This compensation method does not change the physical state of the wafer itself, but only offsets the offset by adjusting the exposure position. Therefore, it is applicable to any packaging structure that generates an offset and does not introduce additional process steps. The compensation parameter comes from the correspondence established in step S3. This correspondence is constructed based on multiple batches of data samples. Its predicted offset already includes the combined contribution of the intrinsic expansion and contraction of the packaging material and the offset induced by the curing of the dielectric layer. When the structure to be compensated matches the packaging process conditions of the database sample, the existing compensation parameter is directly called without re-measuring the offset, so as to realize the reuse of the compensation parameter between different batches. The residual offset threshold set in step S46 is used to determine whether the compensation is effective and to feed back cases that exceed the threshold to the database, so that the mapping relationship has self-correction capability.
[0040] Based on the same inventive concept as the multi-chip package offset compensation method in the foregoing embodiments, the present invention also provides a multi-chip package offset compensation system, such as... Figure 2 As shown, the system includes: The material data acquisition module is used to acquire material expansion and contraction characterization data of multiple candidate packaging materials; The offset data acquisition module is used to acquire the actual alignment offset data of the multi-chip package structure formed using each candidate packaging material; The parameter determination module is used to determine the target packaging material and the offset compensation parameters corresponding to the target packaging material based on the material expansion and contraction characterization data and actual alignment offset data of each candidate packaging material. The compensation module is used to perform offset compensation on the alignment process of the multi-chip package structure to be compensated, which is formed using the target packaging material, based on the offset compensation parameters.
[0041] The adjustment system described above in this invention can effectively realize the multi-chip package offset compensation method, and the technical effects it can achieve are as described in the above embodiments, which will not be repeated here.
[0042] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.
Claims
1. A method for offset compensation in a multi-chip package, the method comprising: include: Acquire material expansion and contraction characterization data of multiple candidate packaging materials, wherein the material expansion and contraction characterization data is used to characterize the dimensional change characteristics of the candidate packaging materials during the packaging process; The candidate packaging materials are used to form corresponding multi-chip packaging structures, and the actual alignment offset data of each multi-chip packaging structure is obtained. Based on the material expansion and contraction characterization data and the actual alignment offset data corresponding to each of the candidate packaging materials, the target packaging material and the offset compensation parameter corresponding to the target packaging material are determined. The target packaging material is used to form a multi-chip package structure to be compensated, and the alignment process of the multi-chip package structure to be compensated is offset compensated based on the offset compensation parameters.
2. The multi-chip package offset compensation method according to claim 1, characterized in that, Determining the target packaging material and the offset compensation parameters corresponding to the target packaging material includes: Establish a correspondence between the material expansion and contraction characterization data of each candidate packaging material and the corresponding actual alignment offset data, and determine the target packaging material and the offset compensation parameters based on the correspondence.
3. The multi-chip package offset compensation method according to claim 1, characterized in that, The candidate encapsulation material is formed by adjusting at least one of the resin matrix, curing component, filler component, or auxiliary component.
4. The multi-chip package offset compensation method according to claim 1, characterized in that, The material expansion and contraction characterization data includes the curing dimensional change rate and the coefficient of thermal expansion. The curing dimensional change rate is determined based on the ratio of the dimensional difference of the candidate encapsulation material before and after curing to the dimensional difference before curing. The coefficient of thermal expansion is determined based on the ratio of the dimensional change of the candidate encapsulation material per unit initial size to the temperature change within a preset temperature range.
5. The multi-chip package offset compensation method according to claim 1, characterized in that, The actual alignment offset data is determined based on the graphic displacement data of the offset magnification measurement graphic.
6. The multi-chip package offset compensation method according to claim 5, characterized in that, The offset magnification measurement pattern includes a reference pattern and a pattern to be measured; The reference pattern has a first pattern period, the pattern to be tested has a second pattern period, and the first pattern period and the second pattern period are different; The graphic displacement data includes the relative displacement between the reference graphic and the graphic to be measured, and the actual alignment offset data is determined based on the relative displacement and the period difference between the first graphic period and the second graphic period.
7. The multi-chip package offset compensation method according to claim 2, characterized in that, The step of establishing the correspondence between the material expansion and contraction characterization data of each candidate packaging material and the corresponding actual alignment offset data includes: Before the polyimide layer formed on each of the multi-chip package structures is cured, first thermal expansion data and first alignment offset data of each of the multi-chip package structures are obtained. After the polyimide layer is cured, the second thermal expansion data and the second alignment offset data of each of the multi-chip package structures are obtained; Based on the second thermal expansion data and the first thermal expansion data, determine the polyimide curing-induced thermal expansion change data, and based on the second alignment offset data and the first alignment offset data, determine the polyimide curing-induced offset change data. Based on the polyimide curing-induced thermal expansion change data and the polyimide curing-induced offset change data, a polyimide curing offset contribution parameter is generated, and the polyimide curing offset contribution parameter is used as a component of the correspondence.
8. The multi-chip package offset compensation method according to claim 2, characterized in that, The step of establishing the correspondence between the material expansion and contraction characterization data of each candidate packaging material and the corresponding actual alignment offset data includes: Collect data samples from multiple batches and groups of candidate packaging materials or multiple groups of packaging process conditions, and store the data samples in an offset database; The data sample includes material identification, material expansion and contraction characterization data, and actual alignment offset data. Based on the offset database, a mapping relationship is formed between the material expansion and contraction characterization data and the actual alignment offset data.
9. The multi-chip package offset compensation method according to claim 8, characterized in that, Determining the target packaging material and the offset compensation parameters based on the correspondence includes: Based on the mapping relationship in the offset database, candidate packaging materials and their predicted alignment offset data that match the packaging process conditions of the multi-chip packaging structure to be compensated are determined. The target packaging material and the offset compensation parameters are determined based on the predicted alignment offset data.
10. A multi-chip package offset compensation system, characterized in that, include: The material data acquisition module is used to acquire material expansion and contraction characterization data of multiple candidate packaging materials; The offset data acquisition module is used to acquire the actual alignment offset data of the multi-chip package structure formed using each of the candidate packaging materials; The parameter determination module is used to determine the target packaging material and the offset compensation parameter corresponding to the target packaging material based on the material expansion and contraction characterization data and the actual alignment offset data corresponding to each candidate packaging material. The compensation module is used to perform offset compensation on the alignment process of the multi-chip package structure to be compensated formed using the target packaging material, based on the offset compensation parameters.