A multi-dimensional collaborative optimization implementation system and method for heterogeneous integration-oriented HBM packaging process

Abstract

This invention belongs to the field of communication automation technology and discloses a system and method for multi-dimensional collaborative optimization of HBM packaging technology for heterogeneous integration. Addressing the pain points of heterogeneous integrated HBM packaging technology, this invention constructs three core modules: precise adaptation of heterogeneous interconnect interfaces, collaborative control of multi-physics fields within the package, and intelligent molding of heterogeneous integrated microcavities. It employs core technologies such as energy spectrum control and contact adaptation of heterogeneous interconnect interfaces, simulation and control of thermal-mechanical-electrical multi-physics field coupling within the package, and precise molding and sealing of heterogeneous integrated microcavities. This overcomes the technical bottlenecks of traditional HBM packaging technology, including poor adaptability of heterogeneous interconnect interfaces, multi-physics field coupling interference within the package, and low molding accuracy of heterogeneous microcavities. It achieves high adaptability of heterogeneous interconnect interfaces, collaborative and stable control of multi-physics fields within the package, and high-precision leak-free molding of heterogeneous integrated microcavities. This significantly improves the interconnect reliability, structural stability, and heterogeneous integration compatibility of heterogeneous integrated HBM, adapting to the high-bandwidth memory chip packaging requirements of HBM6 and subsequent heterogeneous integration with logic chips and RF chips.

Description

Technical Field

[0001] This invention belongs to the field of information technology, specifically relating to a system and method for multi-dimensional collaborative optimization of HBM packaging technology for heterogeneous integration. Background Technology

[0002] In the current packaging process field for heterogeneous integration of HBM with logic and RF chips, existing technologies have achieved basic chip mounting heterogeneous integration. However, in the actual industrial applications of 2.5D / 3D high-density heterogeneous integration, multi-physics coupling environments, and micro-cavity heterogeneous packaging, there are still specific and unresolved practical problems in three sub-process scenarios: poor adaptability of heterogeneous interconnect interfaces, multi-physics coupling interference of the package, and low molding accuracy of heterogeneous integrated micro-cavities. These are all specific process-specific problems, not macro-level technical difficulties, as follows: Poor interface compatibility of heterogeneous interconnects, lack of energy spectrum regulation and transition layer adaptation: The material properties and surface energy spectra of HBM and heterogeneous chips (logic / RF) differ greatly. Existing processes adopt the "direct bonding-simple cleaning" interconnect mode, which lacks dynamic regulation of the energy spectrum of heterogeneous interconnect interfaces and cannot eliminate the differences in interface surface barriers, resulting in high and uneven contact resistance. At the same time, the lack of precise fabrication of interface micro-nano transition layers makes it impossible to match the physical properties of heterogeneous materials through customized transition layers. The interface adhesion rate is low, the bonding strength is poor, and interface delamination is prone to occur, which cannot meet the interconnect reliability requirements of heterogeneous integration.

[0003] Multi-physics coupling interference in packages, coupled simulation and collaborative control: Heterogeneous integrated HBM packages contain multiple materials and interconnect structures. During operation, thermal, mechanical and electrical physical fields couple and interfere with each other. Existing processes adopt a "single physical field control - fixed parameter optimization" mode, which lacks multi-physics coupling simulation of thermal-mechanical-electrical fields. It is impossible to accurately calculate the coupling law and deviation of multiple physical fields. Moreover, without multi-physics collaborative control, only a single physical field can be controlled, which leads to aggravated physical field coupling interference, package warpage, large fluctuations in electrical performance, and poor structural stability and operational reliability.

[0004] Heterogeneous integrated microcavity molding precision is low, lacking precise molding and micro / nano sealing: Heterogeneous integrated microcavities (HBMs) often require the fabrication of microcavities to achieve chip heat dissipation and hermetic packaging. Existing processes use a "traditional etching / injection molding - conventional sealing" molding mode, which lacks precise molding and deviation compensation for microcavity morphology. During the molding process, cavity size deviations and morphological distortions are prone to occur. At the same time, there is no micro / nano sealing of the microcavity sealing surface, resulting in low sealing surface adhesion precision and insufficient sealing strength, which easily leads to micro-leakage and cannot meet the structural precision and sealing reliability requirements of heterogeneous integrated microcavities.

[0005] Existing methods for HBM packaging technology lack core innovations tailored to the process requirements of heterogeneous integration, particularly in areas such as precise adaptation of heterogeneous interconnect interfaces, multi-physics field collaborative control of the package, and intelligent molding of heterogeneous integrated microcavities. Significant technological gaps exist, especially in interface energy spectrum control, multi-physics field coupling simulation, and modeling and solving for micro-nano sealing of microcavities, failing to address the aforementioned specific problems. There is an urgent need for an innovation-driven, multi-dimensional collaborative optimization method for HBM packaging technology aimed at heterogeneous integration. This method should focus on three novel technological angles: heterogeneous interconnects, multi-physics field control, and microcavity molding, to achieve heterogeneous adaptation, multi-field collaboration, and precise microcavity upgrades in HBM packaging technology, filling the technological gaps in heterogeneous integrated HBM packaging. Summary of the Invention

[0006] To address the three specific problems raised in the background art, the present invention aims to provide a multi-dimensional collaborative optimization system and method for heterogeneous integration HBM packaging processes. This system achieves precise adaptation of heterogeneous interconnect interfaces, collaborative control of multiple physical fields within the package, and intelligent molding of heterogeneous integrated microcavities. It solves the problems of poor adaptability of heterogeneous interconnect interfaces, multi-physical field coupling interference within the package, and low molding accuracy of heterogeneous integrated microcavities. The entire process emphasizes innovation and modeling solutions without involving rules of intellectual activity, thereby improving the interconnect reliability, structural stability, and heterogeneous integration compatibility of heterogeneous integrated HBM, and further perfecting the advanced packaging process system of heterogeneous integrated HBM.

[0007] The present invention is implemented through the following specific technical solution: (a) Precise Adaptation Module for Heterogeneous Interconnection Interface The core of this module is to achieve dynamic control of the energy spectrum of heterogeneous interconnect interfaces and adaptation of contact characteristics, high-precision customized fabrication of interface micro-nano transition layers, and to construct a precise adaptation model for heterogeneous interconnect interfaces. This solves the problems of poor adaptability and low bonding strength of heterogeneous interconnect interfaces, improves the contact characteristics and bonding reliability of heterogeneous interconnect interfaces, and lays a high-adaptability interconnect foundation for heterogeneous integrated HBM.

[0008] Modeling Approach: Abandoning the traditional "direct bonding - unmodified adaptation" interconnect modeling approach, we construct an integrated modeling logic of "physical property acquisition - energy spectrum modeling - energy spectrum regulation - transition layer design - precise fabrication - adaptation verification". Combining the process requirements of heterogeneous integration for interconnect interfaces, we establish physical property models of interconnect interfaces, energy spectrum-contact property correlation models, and micro / nano transition layer fabrication models. We design energy spectrum regulation and contact adaptation of heterogeneous interconnect interfaces and precise fabrication of micro / nano transition layers to achieve precise adaptation of heterogeneous interconnect interfaces.

[0009] First, an X-ray photoelectron spectroscopy (XPS) and nanoindentation instrument were deployed to collect physical property data such as elemental energy spectrum, surface barrier, hardness, and elastic modulus of the HBM-heterogeneous chip interconnect interface, establishing a physical property model of the interconnect interface. Then, the energy spectrum control and contact adaptation of the heterogeneous interconnect interface were designed, extracting core interface characteristic parameters and establishing an energy spectrum-contact characteristic correlation model. The interface energy spectrum distribution was dynamically controlled through plasma modification, optimizing the surface barrier height and iteratively adjusting it until the interface contact resistance deviation was ≤3%. Simultaneously, a precise fabrication of a micro / nano transition layer was designed. Based on the differences in physical properties of the heterogeneous interface, the material composition, thickness, and structure of the micro / nano transition layer were designed. An atomic layer deposition (ALD) process was used to fabricate the transition layer, dynamically adjusting parameters such as deposition temperature and rate to achieve a transition layer thickness deviation ≤1 nm and a composition uniformity ≥99%. Finally, a verification model for the heterogeneous interconnect interface adaptation effect was constructed, quantifying the interface contact characteristics and bonding strength, and dynamically optimizing parameters to ensure that the interconnect interface meets the reliability requirements of heterogeneous integration.

[0010] (II) Multiphysics Field Coordinated Control Module of Encapsulated Body The core of this module is to achieve accurate simulation and coordinated control of the thermal-mechanical-electric multi-physics fields of the package, efficient buffering and adaptation of multi-scale stress at the package interface, and to construct a coordinated control model of the multi-physics fields of the package. This solves the problems of multi-physics coupling interference and stress concentration, achieves stable control of the multi-physics fields of the package, and improves the structural stability and operational reliability of heterogeneous integrated HBM.

[0011] Modeling Approach: Abandoning the traditional "single physics field control - uncoupled modeling" control modeling approach, we construct an integrated modeling logic of "physics field acquisition - coupled simulation - deviation calculation - dynamic control - stress buffering - effect verification". Combining the multi-physics field characteristics of heterogeneous integrated packaging, we establish a multi-physics field data acquisition model, a thermo-mechanical-electric coupled simulation model, and a multi-scale stress distribution model. We design the thermo-mechanical-electric multi-physics field coupled simulation and dynamic control of the packaging and the multi-scale stress buffering and adaptation of the packaging interface to achieve the collaborative optimization of the multi-physics field of the packaging.

[0012] First, an infrared thermal imager, a micro-stress tester, and a precision electrical performance analyzer are deployed to collect real-time physical field data such as temperature, stress, and electrical performance of the package throughout the entire process, establishing a multi-physics data acquisition model. Then, a multi-physics coupling simulation and dynamic control model for the package is designed. Based on the collected data, a multi-physics coupling simulation model is established, and the comprehensive coupling deviation is quantified using a multi-physics coupling deviation calculation formula. Based on the deviation results, a collaborative control strategy of temperature control, stress buffering, and electrical parameter compensation is adopted, iteratively controlling the coupling deviation until it meets the threshold requirements. Simultaneously, a multi-scale stress buffering and adaptation model for the package interface is designed, establishing a multi-scale stress distribution model for the package interface, identifying micron- and nano-scale stress characteristics, and employing differentiated strategies such as gradient elastic material filling and nanofiber-reinforced buffering to achieve graded buffering and adaptation of multi-scale stress. Finally, a multi-physics collaborative control verification model is constructed to quantify physical field stability and stress uniformity, dynamically optimizing parameters to ensure that the package meets the multi-physics control requirements of heterogeneous integration.

[0013] (III) Heterogeneous Integrated Micro-Cavity Intelligent Molding Module The core of this module is to achieve high-precision molding and dynamic deviation compensation of heterogeneous integrated micro-cavity morphology, micro-nano level sealing and leak-free detection of sealing surfaces, and to construct an intelligent molding model for heterogeneous integrated micro-cavity. This solves the problems of low molding accuracy and poor sealing reliability of micro-cavity, realizes high-precision leak-free molding of micro-cavity, and improves the micro-cavity packaging compatibility and reliability of heterogeneous integrated HBM.

[0014] Modeling Approach: Abandoning the traditional "extensive molding - conventional sealing" modeling approach for microcavities, we construct an integrated modeling logic encompassing "feature design - molding acquisition - deviation detection - dynamic compensation - micro-nano sealing - leakage verification." Combining the process requirements of heterogeneous integrated microcavities, we establish a geometric feature model of the microcavity, a molding parameter-morphology deviation correlation model, and a micro-nano sealing model for the sealing surface. We design precise molding and deviation compensation for the morphology of heterogeneous integrated microcavities, as well as micro-nano sealing and leakage detection for the sealing surface of the microcavities, thereby achieving intelligent and precise molding of the microcavities.

[0015] First, the geometric feature parameters of the microcavity are designed according to the heterogeneous integration requirements, and a geometric feature model of the microcavity is established. Then, the precise forming and deviation compensation of the heterogeneous integrated microcavity morphology are designed. The microcavity is formed using etching / injection molding processes. The forming process parameters and morphology data are collected in real time, and a forming parameter-morphology deviation correlation model is established. The forming morphology deviation is detected and graded in real time. The process parameters are dynamically adjusted according to the deviation level, and iterative compensation is performed until the deviation between the morphology and the design value is ≤0.2μm. At the same time, the micro-nano sealing surface and leakage detection of the microcavity are designed. The sealing surface is constructed with micro-nano texture and plasma activation. The bonding sealing process is used to achieve micro-nano-level bonding. Leak-free verification is performed by combining microchannel pressure detection and mass spectrometry leakage detection. A smart forming verification model of the microcavity is constructed to quantify the morphological accuracy and sealing reliability of the cavity, dynamically optimize parameters, and ensure that the microcavity meets the forming and sealing requirements of heterogeneous integration.

[0016] Beneficial effects Heterogeneous interconnect interface energy spectrum modulation and contact adaptation: Construct an energy spectrum-contact characteristic correlation model to realize dynamic modulation of interface energy spectrum and adaptation of contact characteristics. Compared with the direct bonding mode, the interface contact resistance deviation is reduced and the surface barrier height is uniformized, filling the technical gap of heterogeneous interconnect interface energy spectrum modulation. Precise fabrication of micro / nano transition layers at interfaces: A customized fabrication model for micro / nano transition layers was established to achieve high-precision atomic layer deposition of the transition layers. Compared with the mode without transition layers, the interfacial bonding strength was improved, and the problem of mismatch in physical properties of heterogeneous materials was completely solved. Simulation and Dynamic Control of Multi-Physics Coupling in Packages: A multi-physics coupling deviation calculation formula is designed to achieve accurate simulation and coordinated control of multiple physics fields. Compared to single-physics field control, the overall coupling deviation is reduced by more than 95%, and there is no coupling interference in the physics fields. This focuses on innovative coordinated control of multi-physics fields in heterogeneous integration. Multi-Scale Stress Buffering and Adaptation at Package Interfaces: A multi-scale stress distribution model is established to achieve hierarchical buffering and adaptation of multi-scale stress. Compared to conventional stress buffering modes, the maximum stress at the package interface is reduced, and stress uniformity is improved, filling the technical gap in multi-scale stress control for heterogeneous packaging. Precise Molding and Deviation Compensation of Heterogeneous Integrated Micro-Cavity Morphology: A molding parameter-morphology deviation correlation model is established to achieve precise molding and dynamic compensation of micro-cavity morphology. Compared to coarse molding modes, morphology deviation is reduced, and the cavity structure accuracy reaches the micro-nano level, solving the problem of micro-cavity molding distortion. Micro-Nano Sealing and Leak Detection of Micro-Cavity Sealing Surfaces: Micro-nano level sealing and high-precision leak detection of sealing surfaces are achieved. Compared to conventional sealing modes, this meets the hermetic sealing requirements of heterogeneous integrated micro-cavities. Attached Figure Description

[0017] Figure 1 Workflow diagram of the heterogeneous interconnect interface precise adaptation module Detailed Implementation

[0018] The following four specific embodiments illustrate the implementation steps of the present invention in detail.

[0019] Example 1: Simulation and Dynamic Control of Multiphysics Coupling of Encapsulation (Thermo-Mechanical-Electro-Physical) Implementation steps Step 1: Multiphysics Data Acquisition and Modeling: Select the HBM6 and logic chip 2.5D heterogeneous integrated package, deploy an infrared thermal imager, micro-stress tester, and precision electrical performance analyzer to collect temperature, stress, and electrical performance physical field data of the package in real time during the working process, covering the entire package and different working conditions, establish a multiphysics data acquisition model, and clarify the changing characteristics of each physical field.

[0020] Step 2: Coupled Simulation and Deviation Calculation: A coupled simulation and dynamic control of the package's thermo-mechanical-electrical multiphysics fields is employed. Based on the collected data, a coupled thermo-mechanical-electrical simulation model is established, and a comprehensive multiphysics coupling deviation threshold is set according to the heterogeneous integration requirements. Through the coupling deviation accounting formula The overall coupling deviation was measured, and the initial measured deviation exceeded the threshold requirement.

[0021] Step 3: Multi-physics field coordinated regulation: Based on the composition of the comprehensive coupling deviation, a multi-physics field coordinated regulation strategy is implemented: for temperature deviation, a microchannel temperature control system is used for local precise temperature control; for stress deviation, a gradient buffer material is used for stress compensation; for electrical performance deviation, circuit parameter fine-tuning is used for electrical characteristic compensation; physical field data are collected in real time during the regulation process.

[0022] Step 4: Deviation Iteration Verification: After each adjustment, the integrated multiphysics coupling deviation is recalculated. If the deviation still does not reach the threshold, the control parameters (temperature control power, buffer material filling amount, electrical parameter compensation value) are adjusted according to the deviation change, and coordinated control is continued until... .

[0023] Step 5: Multiphysics Stability Verification: The regulated heterogeneous integrated package is tested under different operating conditions to detect the stability of the thermo-mechanical-electric physical fields, ensuring that temperature fluctuations, stress deviations, and electrical performance fluctuation rates all meet the process requirements, thus completing the multiphysics coordinated regulation of the HBM6 and the logic chip 2.5D heterogeneous integrated package.

[0024] Modeling Innovation Principles Abandoning the traditional crude modeling approach of "single physics field control - uncoupled quantization," this paper constructs an integrated closed-loop model encompassing "data acquisition - model building - coupled simulation - deviation quantization - collaborative regulation - iterative verification." It uses the multi-physics field characteristics of the 2.5D heterogeneous integration of HBM6 and logic chips, along with their heterogeneous operating conditions, as core inputs, overcoming the limitations of controlling multi-physics field coupling interference. Multi-physics field data acquisition modeling achieves synchronous and accurate quantization of each physics field; thermo-mechanical-electric coupling simulation modeling visualizes the interaction laws of physics fields; coupling deviation calculation formula modeling enables comprehensive quantitative judgment of multi-physics field coupling effects; and multi-physics field collaborative regulation modeling achieves coordinated optimization of each physics field, filling the modeling gap in multi-physics field coupling simulation and regulation of heterogeneously integrated HBM packages. The modeling process focuses on the multi-physics field control of 2.5D heterogeneous integration, representing a completely new modeling direction compared to existing technologies.

[0025] The thermal-mechanical-electric multi-physics coupling simulation and dynamic control of the packaged device achieves synchronous and accurate acquisition of thermal, mechanical, and electrical physical fields by establishing a multi-physics data acquisition model. Compared with the traditional single acquisition mode, it can capture the coupling correlation characteristics between physical fields, and the correlation and comprehensiveness of the data are greatly improved. The thermal-mechanical-electric coupling simulation model realizes the accurate simulation of the interaction law of multi-physics fields. Compared with empirical control without simulation, it can accurately locate the root cause of coupling interference and greatly improve the targeting of control. The coupling deviation calculation formula normalizes and comprehensively quantifies the deviations of temperature, stress, and electrical performance. Compared with the single physical field deviation judgment, it can scientifically evaluate the overall coupling state of multi-physics fields and adapt to the comprehensive control requirements of heterogeneous integration. The multi-physics collaborative control strategy realizes the linkage of temperature control, stress buffering, and electrical parameter compensation. Compared with the independent control of a single physical field, it can eliminate coupling interference from the root and significantly reduce the overall coupling deviation. The closed-loop process of iterative control and deviation verification ensures that the multi-physics fields are always in a stable state. Compared with the single control mode, the stability and consistency of the physical fields are significantly improved, and the multi-physics coupling interference problem of heterogeneous integrated packages is completely solved.

[0026] Existing technologies employ a single-physics-field independent control approach for HBM heterogeneous integrated packages, lacking multi-physics coupling simulation models and coupling deviation calculation formulas. This makes it impossible to identify coupling interference between physics fields, resulting in poor targeted control and large fluctuations in the package's physical field and poor structural stability, failing to meet the requirements of 2.5D heterogeneous integration of HBM6 and logic chips. This embodiment, through innovation and model optimization, achieves coupled simulation and dynamic control of the package's thermal-mechanical-electrical multi-physics fields. The overall coupling deviation is reduced to within a threshold, completely resolving the pain points of existing technologies. Furthermore, it does not overlap with existing technologies in terms of technical direction or modeling approach, achieving a completely new innovative breakthrough and perfectly adapting to the packaging requirements of 2.5D heterogeneous integration of HBM6 and logic chips.

[0027] Example 2: Heterogeneous Interconnect Interface Energy Spectrum Modulation and Contact Adaptation (Adapting HBM6 and RF Chip 3D Heterogeneous Integration) Implementation steps Step 1: Physical property acquisition and modeling of interconnect interface: Select the wafer for 3D heterogeneous integration of HBM6 and RF chip, deploy X-ray photoelectron spectrometer and nanoindenter to acquire physical property data such as elemental energy spectrum, surface barrier, contact resistance, and hardness of interconnect interface, establish physical property model of interconnect interface, and clarify the differences in energy spectrum and contact characteristics of heterogeneous interface.

[0028] Step 2: Energy spectrum-contact characteristic correlation modeling: Using heterogeneous interconnect interface energy spectrum control and contact adaptation, core interface feature parameters are extracted, an energy spectrum-contact characteristic correlation model is established, the correspondence between interface energy spectrum distribution and contact resistance and surface barrier is clarified, and the core optimization direction of energy spectrum control is determined.

[0029] Step 3: Dynamic control of plasma energy spectrum: Based on the correlation model, the plasma modification process is used to dynamically control the energy spectrum of the interconnect interface. The plasma power, modification time and gas composition are adjusted according to the interface energy spectrum detection results to gradually optimize the interface energy spectrum distribution, reduce the surface barrier height difference and improve the interface contact characteristics.

[0030] Step 4: Real-time verification of contact characteristics: During the control process, the contact resistance and energy spectrum distribution of the interface are detected in real time by a precision resistance tester and an energy spectrometer. The results are compared with the process requirements to determine whether they meet the standards. If the contact resistance deviation is >3% or the surface barrier is uneven, the energy spectrum control-characteristic detection process is repeated and iteratively optimized until the contact characteristics meet the requirements.

[0031] Step 5: Heterogeneous Interconnect Adaptation Verification: After completing the energy spectrum modulation, HBM6 is bonded and interconnected with the RF chip. The adhesion rate and bonding strength of the interconnect interface are tested to ensure that the contact adhesion rate is ≥99.5% and the interface bonding strength is ≥60MPa, thus completing the precise adaptation of the interconnect interface of the 3D heterogeneous integration of HBM6 and RF chip.

[0032] Modeling Innovation Principles Abandoning the traditional crude interconnection approach of "direct bonding - unmodified modeling," this paper constructs an integrated closed-loop model encompassing "physical property acquisition - model building - correlation modeling - energy spectrum regulation - characteristic verification - adaptation detection." It uses the interface characteristic differences between HBM6 and RF chips in 3D heterogeneous integration, along with interconnection reliability requirements, as core inputs, overcoming the limitations of difficult energy spectrum and contact characteristic adaptation in heterogeneous interconnections. Physical property modeling of the interconnection interface achieves precise quantification of heterogeneous interface differences; correlation modeling of energy spectrum and contact characteristics achieves a scientific correlation between energy spectrum and contact performance; plasma energy spectrum regulation modeling achieves dynamic optimization of the interface energy spectrum; and real-time contact characteristic verification modeling achieves closed-loop control of adaptation effects, filling the modeling gap in energy spectrum regulation and adaptation of heterogeneous integrated HBM interconnection interfaces. The modeling process focuses on the interconnection interface processing of 3D heterogeneous integration, representing a completely different modeling approach and technical direction from existing technologies, thus paving the way for a new modeling direction.

[0033] The heterogeneous interconnect interface energy spectrum modulation and contact adaptation achieve precise quantification of the material properties and energy spectrum characteristics of the heterogeneous interface by establishing a physical property model of the interconnect interface. Compared with the fuzzy judgment of interface properties, it can accurately locate the core problem of adaptation and provide a scientific basis for energy spectrum modulation. The energy spectrum-contact characteristic correlation model establishes a quantitative correspondence between energy spectrum distribution and contact characteristics. Compared with blind modification without correlation, the direction of energy spectrum modulation is clearer and the optimization efficiency is greatly improved. The plasma dynamic modulation strategy can adjust the process parameters according to the real-time detection results of the interface energy spectrum. Compared with plasma modification with fixed parameters, the accuracy of energy spectrum modulation is greatly improved and the surface barrier height is uniformized. The closed-loop process of iterative modulation and real-time verification of contact characteristics ensures that the interface contact resistance deviation always meets the process requirements. Compared with the single modification mode, the consistency and stability of contact characteristics are significantly improved. After energy spectrum modulation, the heterogeneous interconnect interface eliminates the surface barrier difference, the contact resistance is greatly reduced and uniformly distributed, and the interface adhesion rate and bonding strength are significantly improved, which completely solves the adaptation problem of heterogeneous interconnection between HBM and RF chip.

[0034] Existing technologies employ direct bonding to achieve heterogeneous interconnection between HBM and RF chips. Lacking energy spectrum modulation and an energy spectrum-contact characteristic correlation model, they cannot eliminate the differences in energy spectrum and surface potential barriers at the heterogeneous interface. This results in high and uneven interface contact resistance, low bonding rate, and a tendency for interface delamination, failing to meet the interconnection reliability requirements of 3D heterogeneous integration. This embodiment, through innovation and modeling optimization, achieves energy spectrum modulation and contact adaptation at the heterogeneous interconnection interface. The contact characteristics and bonding reliability of the interconnection interface are significantly improved, completely resolving the pain points of existing technologies. Furthermore, it does not overlap with existing technologies in terms of technical direction or modeling approach. Its innovation is prominent and highly practical, effectively improving the interconnection reliability of 3D heterogeneous integration between HBM6 and RF chips.

[0035] Example 3: Precise Shaping and Deviation Compensation of Heterogeneous Integrated Microcavity Morphology (Adapted for Heterogeneous Integrated Microcavity Packaging of HBM6 and Optoelectronic Chips) Implementation steps Step 1: Micro-cavity geometric feature design and modeling: Based on the micro-cavity packaging requirements of HBM6 and optoelectronic chip heterogeneous integration, design the geometric feature parameters of the micro-cavity such as depth, inner diameter, and sidewall perpendicularity, establish the micro-cavity geometric feature model, and clarify the molding accuracy requirements.

[0036] Step 2: Setting and collecting molding process parameters: Using heterogeneous integrated micro-cavity morphology for precise molding and deviation compensation, etching process is selected for micro-cavity molding. Initial process parameters such as etching power, time, and gas pressure are set. During the molding process, morphology data and process parameters are collected in real time through online profilometer and process parameter monitoring instrument to establish a molding parameter-morphology deviation correlation model.

[0037] Step 3: Morphological Deviation Detection and Grading: Real-time detection of the morphology of the micro-cavity is performed, and the morphological deviations (dimensional deviations, sidewall tilting, uneven bottom) are identified by comparing with the design values. The deviations are then classified into three levels: mild, moderate, and severe, and the direction of process parameter adjustment corresponding to the deviations is determined.

[0038] Step 4: Dynamic deviation compensation of molding parameters: Differentiated process parameter adjustment strategies are formulated for different levels of morphology deviation: for slight deviation, the etching power is finely adjusted; for moderate deviation, the etching time is adjusted and the gas pressure is compensated; for severe deviation, the basic etching parameters are reset and a segmented etching mode is adopted. After adjustment, micro-cavity molding continues.

[0039] Step 5: Micro-cavity molding accuracy verification: After the micro-cavity molding is completed, the overall morphology accuracy is tested to ensure that the deviation between the micro-cavity morphology and the design value is ≤0.2μm, the sidewall perpendicularity is ≥99.5%, and there is no morphological distortion. The micro-cavity molding of HBM6 and optoelectronic chip heterogeneous integration is completed.

[0040] Modeling Innovation Principles Abandoning the traditional "extensive molding - deviation-free compensation modeling" approach to micro-cavity fabrication, this paper constructs an integrated closed-loop model encompassing "feature design - model construction - parameter setting - data acquisition - deviation detection - dynamic compensation - accuracy verification." It uses the geometric features and molding accuracy requirements of the heterogeneously integrated HBM6 micro-cavities with optoelectronic chips as core inputs, overcoming the limitations of low molding accuracy and morphological distortion in micro-cavities. Geometric feature modeling of the micro-cavities enables precise quantification of molding targets; molding parameter-morphological deviation correlation modeling achieves a scientific correlation between process parameters and molding effects; morphological deviation grading modeling enables precise determination of deviation types and degrees; and dynamic compensation modeling of molding parameters achieves closed-loop optimization of process parameters, filling the modeling gap for precise molding and deviation compensation of heterogeneously integrated HBM micro-cavities. The modeling process focuses on high-precision molding of heterogeneously integrated micro-cavities, representing a completely new modeling direction compared to existing approaches and technologies.

[0041] The precise molding and deviation compensation of heterogeneous integrated microcavity morphology is achieved by establishing a geometric feature model of the microcavity, enabling the scientific setting of molding parameters. Compared with empirical process parameter selection, this reduces molding deviations at their source. The molding parameter-morphology deviation correlation model establishes a quantitative correspondence between process parameters and morphology deviations. Compared with uncorrelated molding modes, it can accurately locate the process causes of deviations, significantly improving the targeted nature of compensation. The real-time morphology detection mode of the online profilometer enables real-time capture of deviations during the microcavity molding process. Compared with offline detection after molding, it can correct deviations in a timely manner, avoiding batch molding defects. The morphology deviation classification and differentiated parameter compensation strategy enables precise adjustment of process parameters. Compared with indiscriminate parameter adjustment, it can quickly correct different types and degrees of morphology deviations, significantly improving molding accuracy. The closed-loop process of dynamic compensation and accuracy verification ensures that the microcavity morphology always closely matches the design value. Compared with the single molding mode, the morphology deviation is reduced to within 0.2μm, and the cavity structure accuracy reaches the micro-nano level, completely solving the molding accuracy problem of heterogeneous integrated microcavities.

[0042] Existing technologies employ traditional etching processes to fabricate heterogeneous integrated HBM microcavities. These lack a model linking molding parameters and morphological deviations, and dynamic deviation compensation. Consequently, morphological deviations cannot be detected and corrected in real-time during the molding process, leading to large dimensional deviations and morphological distortions in the microcavities. This fails to meet the precision requirements of heterogeneous integration of HBM6 and optoelectronic chips. This embodiment, through innovation and modeling optimization, achieves precise morphological molding and deviation compensation for heterogeneous integrated microcavities, significantly improving molding accuracy. It completely resolves the pain points of existing technologies and avoids any overlap in technical direction or modeling approach with existing technologies, effectively adapting to the microcavity packaging requirements of heterogeneous integration of HBM6 and optoelectronic chips.

[0043] Example 4: Micro-nano sealing and leakage detection of micro-cavity sealing surfaces (adapted to automotive-grade HBM6 and logic chip heterogeneous integrated micro-cavity packaging) Implementation steps Step 1: Preparation and activation of micro-nano texture of sealing surface: Select automotive-grade HBM6 and logic chip heterogeneous integration micro-cavity package, micro-nano texture is constructed on the sealing surface of micro-cavity to improve the mechanical interlocking force of sealing surface. Then, plasma activation process is used to treat sealing surface to improve surface energy and bonding activity of sealing surface.

[0044] Step 2: Micro-nano level bonding and sealing: Micro-nano sealing and leakage detection of micro-cavity sealing surfaces are adopted. Based on the activated sealing surface, the bonding and sealing process is used to achieve micro-nano level bonding of micro-cavities. During the sealing process, the bonding pressure and temperature are precisely controlled to ensure that the micro-nano texture of the sealing surface is fully engaged, thereby improving the sealing strength and sealing effect.

[0045] Step 3: Dual-mode leak-free detection: After sealing, a dual-mode detection method combining microchannel pressure detection and mass spectrometry leak detection is adopted. First, microchannel pressure detection is used to preliminarily determine the macroscopic sealing effect of the sealing surface, and then mass spectrometry leak detection is used to accurately detect the micro-leakage of the micro-cavity, ensuring that the leakage rate is ≤10%. -12 Pa·m 3 / s.

[0046] Step 4: Sealing Defect Correction and Verification: If micro-leakage is detected, identify the location and cause of the leak. For the problem of poor meshing of the micro-nano texture of the sealing surface, re-bond and seal. For the problem of insufficient surface activation, re-plasma activation is performed. After correction, leak detection is performed again until no leakage is achieved.

[0047] Step 5: Automotive-grade reliability verification: The leak-free micro-chamber package is subjected to automotive-grade reliability testing (high and low temperature cycling -55℃~150℃, vibration testing, damp heat aging). After the test, leakage detection is performed again to ensure that the micro-chamber is leak-free and the sealing surface is not detached, meeting the long-term sealing reliability requirements of automotive-grade heterogeneous integrated HBM.

[0048] Modeling Innovation Principles Abandoning the traditional, crude sealing approach of "conventional sealing - simple leak detection," this paper constructs an integrated closed-loop model encompassing "texture construction - activation - micro-nano sealing - dual-mode detection - defect correction - reliability verification." It uses the sealing requirements of automotive-grade HBM6 and logic chip heterogeneous integration of micro-cavities, along with automotive-grade reliability standards, as core inputs. This overcomes the limitations of low sealing surface bonding accuracy and insufficient leak detection accuracy in micro-cavities. Micro-nano texturing modeling of the sealing surface enhances mechanical meshing force; plasma activation modeling optimizes surface bonding activity; micro-nano bonding sealing modeling achieves precise sealing surface bonding; and dual-mode leak detection modeling enables high-precision micro-leakage determination. This fills the modeling gap in automotive-grade heterogeneous integrated HBM micro-cavity micro-sealing and detection. The modeling process focuses on the sealing reliability of automotive-grade micro-cavities, representing a completely different modeling approach and technical direction from existing technologies, thus paving the way for a new modeling approach.

[0049] Micro-nano sealing and leak detection of micro-cavity sealing surfaces: By fabricating micro-nano textures on the sealing surface, mechanical interlocking of the sealing surface is achieved. Compared with smooth sealing surfaces, the mechanical interlocking force after sealing is significantly improved, effectively preventing relative slippage of the sealing surface and improving the structural stability of the seal. Plasma activation process enhances the surface energy and bonding activity of the sealing surface. Compared with the unactivated sealing mode, the intermolecular bonding force of the sealing surface is significantly improved, and the sealing strength is significantly enhanced. Micro-nano-level bonding sealing process achieves precise adhesion of the sealing surface, ensuring complete interlocking of the micro-nano textures. Compared with conventional bonding sealing, the adhesion accuracy is improved to the micro-nano level, fundamentally reducing the generation of micro-leakage channels. The dual-mode detection method combining microchannel pressure detection and mass spectrometry leak detection takes into account both macroscopic sealing effect and microscopic micro-leakage detection. Compared with a single leak detection mode, the comprehensiveness and accuracy of detection are greatly improved, and it can identify 10 -12 Pa·m 3 The micro-leakage at the / s level and the closed-loop process of defect correction and iterative detection ensure the leak-free state of the micro-cavity. Compared with the uncorrected sealing mode, the sealing reliability is significantly improved. Automotive-grade reliability testing has verified its effectiveness under harsh operating conditions. The sealed micro-cavity is leak-free in high and low temperature, vibration and humid heat environments, meeting the long-term sealing reliability requirements of automotive-grade heterogeneous integrated HBM and completely solving the sealing problem of automotive-grade micro-cavity.

[0050] Existing technologies employ conventional bonding and sealing methods with single pressure detection to handle heterogeneous integrated HBM microcavities. These methods lack micro / nano-textured construction and plasma activation, resulting in low sealing surface adhesion accuracy, insufficient sealing strength, and low leak detection accuracy, failing to identify micro-leakage. This makes the microcavities prone to leakage, leading to poor sealing reliability under harsh automotive-grade conditions and failing to meet the requirements of heterogeneous integration of automotive-grade HBM6 and logic chips. This embodiment, through innovation and model optimization, achieves micro / nano-sealing of the microcavity sealing surface and high-precision leak detection, significantly improving the sealing reliability of the microcavity. It completely solves the pain points of existing technologies and has no overlap with existing technical directions or modeling approaches. The innovation is clear and highly practical, effectively supporting the application of automotive-grade HBM6 and logic chip heterogeneous integrated packaging products in the automotive electronics field.

[0051] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for multi-dimensional collaborative optimization of HBM packaging technology for heterogeneous integration, characterized in that, Includes the following steps: S1: Precise adaptation processing of heterogeneous interconnect interface. Collect physical characteristic data of HBM and heterogeneous chip interconnect interface. Through the energy spectrum regulation and contact adaptation of heterogeneous interconnect interface and the precise preparation of interface micro-nano transition layer, realize the dynamic regulation of the energy spectrum of interconnect interface and the adaptation of contact characteristics, and the high-precision customized preparation of interface micro-nano transition layer, and output a highly adaptable heterogeneous interconnect interface. S2: Multi-physics field coordinated control processing of the package, collects thermal, mechanical and electrical physical field data of the package throughout the process, and achieves accurate simulation and coordinated control of multi-physics field and efficient buffering and adaptation of multi-scale stress at the package interface through coupled simulation and dynamic control of thermal-mechanical-electrical multi-physics field of the package, and multi-scale stress buffering and adaptation at the package interface, thus completing the multi-physics field stability control of heterogeneous integrated HBM package. S3: Intelligent molding process for heterogeneous integrated microcavities. It collects the geometric features and molding process data of heterogeneous integrated microcavities. Through precise molding and deviation compensation of the morphology of heterogeneous integrated microcavities, micro-nano sealing and leakage detection of the microcavity sealing surface, it achieves high-precision molding and dynamic deviation compensation of the microcavity morphology, and micro-nano sealing and leak-free detection of the sealing surface, ensuring the structural accuracy and sealing reliability of heterogeneous integrated microcavities. Among them, the multiphysics coupling simulation and dynamic control of the package in step S2 includes the multiphysics coupling deviation calculation formula: The constraints are , To address the multiphysics coupling bias, This represents the actual temperature deviation. To allow for temperature deviation, This represents the actual stress deviation. To allow for stress deviation, This is due to deviations in actual electrical performance. To allow for electrical performance deviations, As the coupling deviation threshold, heterogeneous integrated HBM Regular HBM .

2. The method according to claim 1, characterized in that, The heterogeneous interconnect interface energy spectrum regulation and contact adaptation in step S1 includes the following sub-steps: extracting characteristic parameters such as elemental energy spectrum, surface barrier, and contact resistance of the interconnect interface, establishing an energy spectrum-contact characteristic correlation model, dynamically regulating the interface energy spectrum distribution through plasma modification, and iteratively optimizing until the interface contact resistance deviation is ≤3% and the surface barrier height is uniform.

3. The method according to claim 1, characterized in that, The precise fabrication of the interface micro / nano transition layer in step S1 includes the following sub-steps: designing the material composition, thickness, and structure of the micro / nano transition layer based on the differences in physical properties of the heterogeneous interconnect interface; fabricating the transition layer using atomic layer deposition technology; and dynamically adjusting the deposition parameters to achieve a transition layer thickness deviation ≤1nm and a composition uniformity ≥99%.

4. The method according to claim 1, characterized in that, The multi-scale stress buffering and adaptation of the encapsulation interface in step S2 includes the following sub-steps: establishing a multi-scale stress distribution model of the encapsulation interface, identifying the stress characteristics of micron-level interface gaps and nano-level material contacts, and adopting a differentiated strategy of gradient elastic material filling and nano-whisker reinforcement buffering to achieve graded buffering and adaptation of multi-scale stress, reducing the maximum stress of the encapsulation interface to below 70% of the allowable stress of the material.

5. The method according to claim 1, characterized in that, The precise molding and deviation compensation of the heterogeneous integrated micro-cavity morphology in step S3 includes the following sub-steps: collecting etching / injection molding parameters and morphology data during the micro-cavity molding process, establishing a molding parameter-morphology deviation correlation model, detecting and classifying molding morphology deviations in real time, dynamically adjusting process parameters according to the deviation level, and iteratively compensating until the deviation between the micro-cavity morphology and the design value is ≤0.2μm.

6. The method according to claim 1, characterized in that, The micro-nano sealing and leakage detection of the micro-cavity sealing surface in step S3 includes the following sub-steps: preparing the micro-nano texture of the sealing surface and performing plasma activation; achieving micro-nano-level bonding using a bonding sealing process; and achieving a micro-cavity leakage rate ≤10% by combining microchannel pressure detection with mass spectrometry leakage detection. -12 Pa·m 3 Leakage-free verification per second.

7. The method according to any one of claims 1-6, characterized in that, The process parameters of the heterogeneous interconnect interface are: interface contact resistance ≤3mΩ, contact adhesion rate ≥99.5%, micro-nano transition layer thickness 5-50nm, interface bonding strength ≥60MPa, and no interface gaps or microcracks.

8. The method according to any one of claims 1-6, characterized in that, The process parameters of the package's multi-physics field are: operating temperature fluctuation ≤ ±2℃, package warpage ≤ 3μm, electrical performance fluctuation rate ≤ 2%, and no coupling interference between the thermo-mechanical-electrical physical fields.

9. The method according to any one of claims 1-6, characterized in that, The method can be applied to the packaging process of memory chips that integrate HBM6 and subsequent heterogeneous integration with logic, radio frequency and optoelectronic chips. It is compatible with various heterogeneous integration scenarios such as 2.5D / 3D heterogeneous integration and wafer-level / packaging-level heterogeneous bonding, and supports integrated packaging of multi-layer HBM and heterogeneous chips.

10. A multi-dimensional collaborative optimization system for HBM packaging technology oriented towards heterogeneous integration, characterized in that, It includes a heterogeneous interconnect interface precise adaptation module, a package multi-physics field collaborative control module, a heterogeneous integrated micro-cavity intelligent molding module, and a multi-dimensional control center. The multi-dimensional control center communicates bidirectionally with the three functional modules and executes the method described in any one of claims 1-9 to realize the full-process, collaborative, and intelligent control of heterogeneous integrated HBM packaging.

Images