Microchip laser mass transfer method, device and equipment and storage medium

Abstract

The invention relates to the technical field of laser mass transfer, in particular to a microchip laser mass transfer method, device and equipment and a storage medium, and the method comprises the steps: firstly obtaining a morphology data set of an acceptor substrate, constructing a morphology digital map based on the morphology data set, and constructing a process parameter query table based on the morphology digital map; then acquiring a plurality of chip sub-region images of the donor substrate and a plurality of welding plate sub-region images of the acceptor substrate, and performing visual fine alignment processing on each chip sub-region image and each welding plate sub-region image to obtain a plurality of plane position offsets; the method comprises the following steps of: firstly, performing instruction generation processing on the basis of a process parameter lookup table and a plurality of plane position offsets to obtain an execution instruction set corresponding to each chip sub-region image, and finally, finishing laser mass transfer of all chip sub-region images on the basis of a plurality of execution instruction sets, so as to solve the transfer defect caused by non-uniform morphology and improve the transfer efficiency. The mass transfer yield of microchips is improved, the manufacturing cost is reduced, and the mass production process is accelerated.

Description

Technical Field

[0001] This invention relates to the field of laser mass transfer technology, and in particular to a microchip laser mass transfer method, apparatus, device and storage medium. Background Technology

[0002] In the field of laser mass transfer for Micro LEDs (micro-light-emitting diodes), current process control relies on two potential and critical technical assumptions: the ideal acceptor assumption and the universal parameter assumption. The ideal acceptor assumption presupposes that the surface of the acceptor substrate electrode pads is an ideally smooth and highly uniform plane. The universal parameter assumption assumes that reliable bonding of all pads in the same batch of chips can be achieved using a fixed set of laser energy, pulse width, and other parameters. Based on these assumptions, the core process of existing technologies involves precise alignment, fixed-parameter transfer, offline inspection of results, and defect repair. The technological focus is concentrated on improving alignment accuracy and transfer speed, while the control of interface contact quality remains in an open-loop, blind state.

[0003] Existing technologies suffer from numerous core defects and shortcomings. Before transfer, the equipment only identifies the planar position of the pads using two-dimensional vision, failing to acquire three-dimensional microscopic morphological information. It lacks prior knowledge of pad convexity, tilt, and other characteristics, resulting in inherent blind spots in the process. Furthermore, the use of a one-size-fits-all, fixed-parameter process cannot customize laser parameters based on the unique microscopic geometric characteristics of each pad. Raised pads are prone to chip damage or stress concentration due to excessive energy, while recessed pads may experience weak bonding due to insufficient energy. Yield is limited by the area with the worst substrate morphology. Quality control relies entirely on offline inspection after transfer, a post-inspection model. Defects can only be identified and addressed after they occur, leading to high sunk costs, and the repair process can easily introduce secondary defects, further extending the production cycle. In addition, the process data chain is broken; it does not collect data related to the acceptor morphology and transfer results, making it difficult to quickly pinpoint the root cause of batch anomalies. Summary of the Invention

[0004] In order to overcome the shortcomings of the prior art, the present invention aims to provide a method, apparatus, device and storage medium for mass transfer of microchip lasers, which aims to solve the transfer defects caused by uneven morphology, improve the yield of mass transfer of microchips, reduce manufacturing costs and accelerate the mass production process.

[0005] The first aspect of this invention provides a microchip laser mass transfer method, comprising: acquiring a topography dataset of a recipient substrate; constructing a topography digital map based on the topography dataset; constructing a process parameter lookup table based on the topography digital map; acquiring multiple chip sub-region images of a donor substrate and multiple pad region images of a recipient substrate; performing visual fine alignment processing on each chip sub-region image and each pad region image to obtain multiple planar position offsets; performing instruction generation processing based on the process parameter lookup table and the multiple planar position offsets to obtain an execution instruction set corresponding to each chip sub-region image; and completing the laser mass transfer of all chip sub-region images based on the multiple execution instruction sets.

[0006] Optionally, in a first implementation of the first aspect of the present invention, the acceptor substrate includes a plurality of pads; the step of obtaining a topography dataset of the acceptor substrate and constructing a topography digital map based on the topography dataset includes: performing three-dimensional measurement processing on the acceptor substrate using a topography scanning tool to obtain the topography dataset, the topography dataset including topography data corresponding one-to-one with each of the pads; obtaining the pad coordinates of each pad; and matching each pad coordinate with each topography data one-to-one to construct the topography digital map.

[0007] Optionally, in a second implementation of the first aspect of the present invention, the digital topography map includes multiple topography data; the step of constructing a process parameter query table based on the digital topography map includes: calling a pre-trained data-driven model; calculating laser process compensation parameters for each topography data based on the data-driven model to obtain a laser parameter compensation dataset; and constructing a process parameter query table based on the laser parameter compensation dataset and the digital topography map.

[0008] Optionally, in a third implementation of the first aspect of the present invention, the laser parameter compensation dataset includes multiple laser parameter compensation data, and the step of constructing a process parameter query table based on the laser parameter compensation dataset and the topography digital map includes: obtaining the pad coordinates corresponding to each laser parameter compensation data based on the topography digital map; establishing a mapping relationship between each laser parameter compensation data and the corresponding pad coordinates to generate multiple related items; and integrating all the related items to construct the process parameter query table.

[0009] Optionally, in a fourth implementation of the first aspect of the present invention, acquiring multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate includes: acquiring a preset partitioning rule, a chip array image of the donor substrate, and a pad array image of the recipient substrate; performing initial position matching processing on the donor substrate and the recipient substrate to establish a positional correspondence between the donor substrate and the recipient substrate; and partitioning the chip array image and the pad array image into regions based on the partitioning rule to obtain multiple chip sub-region images and a pad region image corresponding to each chip sub-region image.

[0010] Optionally, in a fifth implementation of the first aspect of the present invention, the step of performing visual fine alignment processing on each chip sub-region image and each solder pad region image to obtain multiple planar position offsets includes: obtaining multiple preset chip design coordinates; transferring each chip sub-region image and its corresponding solder pad region image to a laser working position respectively; performing visual fine alignment processing on the chip sub-region image and its corresponding solder pad region image at the laser working position to obtain actual transfer coordinates corresponding to each chip sub-region image; and calculating the position offset between each chip sub-region image and its corresponding solder pad region image based on the multiple chip design coordinates and the multiple actual transfer coordinates to obtain multiple planar position offsets.

[0011] Optionally, in a sixth implementation of the first aspect of the present invention, the step of generating execution instructions based on the process parameter lookup table and multiple planar position offsets to obtain an execution instruction set corresponding to each chip sub-region image includes: using multiple chip design coordinates as indexes, obtaining laser parameter compensation data corresponding one-to-one with each chip design coordinate from the process parameter lookup table; performing visual offset correction processing on the corresponding chip design coordinates based on each planar position offset to obtain chip execution coordinates corresponding one-to-one with each chip design coordinate; and performing data fusion processing based on each laser parameter compensation data and the corresponding chip execution coordinates to obtain an execution instruction set corresponding to each chip sub-region image.

[0012] A second aspect of the present invention provides a microchip laser mass transfer apparatus, comprising: a topography digital map construction module for acquiring a topography dataset of a recipient substrate and constructing a topography digital map based on the topography dataset; a process parameter lookup table construction module for constructing a process parameter lookup table based on the topography digital map; an offset calculation module for acquiring multiple chip sub-region images of a donor substrate and multiple pad region images of a recipient substrate, performing visual fine alignment processing on each chip sub-region image and each pad region image to obtain multiple planar position offsets; an instruction generation module for performing instruction generation processing based on the process parameter lookup table and the multiple planar position offsets to obtain an execution instruction set corresponding to each chip sub-region image; and an instruction execution module for completing the laser mass transfer of all chip sub-region images based on the multiple execution instruction sets.

[0013] A third aspect of the present invention provides a microchip laser mass transfer device, the microchip laser mass transfer device comprising: a memory and at least one processor, the memory storing instructions; the at least one processor calling the instructions in the memory to cause the microchip laser mass transfer device to perform the various steps of the microchip laser mass transfer method described in any of the preceding claims.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the microchip laser mass transfer method described in any of the preceding claims.

[0015] In the technical solution of this invention, firstly, a morphology dataset of the recipient substrate is acquired, a morphology digital map is constructed based on the morphology dataset, and a process parameter lookup table is constructed based on the morphology digital map. Then, multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate are acquired. Visual fine alignment processing is performed on each chip sub-region image and each pad region image to obtain multiple planar position offsets. Next, instruction generation processing is performed based on the process parameter lookup table and multiple planar position offsets to obtain an execution instruction set corresponding to each chip sub-region image. Finally, laser mass transfer of all chip sub-region images is completed based on multiple execution instruction sets. This aims to solve the transfer defects caused by morphological inhomogeneity, improve the yield of microchip mass transfer, reduce manufacturing costs, and accelerate the mass production process. Attached Figure Description

[0016] Figure 1 A logic flowchart of the microchip laser mass transfer method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the microchip laser mass transfer device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of the microchip laser mass transfer device provided in an embodiment of the present invention. Detailed Implementation

[0017] This invention provides a microchip laser mass transfer method, apparatus, device, and storage medium. In this invention, the terms "first," "second," "third," "fourth," etc. (if present)," in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatuses.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the microchip laser mass transfer method of the present invention includes: 101. Obtain the morphology dataset of the receptor substrate, and construct a morphology digital map based on the morphology dataset; In this embodiment, a high-precision three-dimensional optical sensor is integrated into a topography pre-scanning module independent of the laser transfer head to perform non-contact three-dimensional measurement of the loaded acceptor substrate, covering the entire area. The system collects key geometric feature information such as height, tilt, and roughness of each pad area, forming a complete acceptor substrate topography dataset. The collected topography data is calibrated and correlated, and each set of pad topography feature data is precisely matched with the corresponding pad design coordinates to construct a digital map of the entire board pad topography, ensuring that the map strictly corresponds to the design coordinate system. This map records the geometric features such as height, tilt, and roughness of each pad area. This method can complete the full-area topography perception before the laser transfer process starts. Non-contact measurement avoids physical damage to the acceptor substrate, and the construction of the topography digital map provides reliable data support for the calculation of subsequent laser process parameters, breaking through the limitation of traditional two-dimensional vision that can only identify planar positions, and eliminating the process blindness caused by the lack of topography information from the source.

[0019] 102. Construct a process parameter lookup table based on the aforementioned topographic digital map; In this embodiment, the built-in physical model or data-driven model of the intelligent compensation controller is invoked to calculate the laser parameter compensation values ​​required for transferring the corresponding chip. This calculation is performed on the topography data corresponding to each pad coordinate in the topography digital map, including key parameters such as height deviation and slope. This covers core parameters such as the magnitude of laser energy increase or decrease. Subsequently, using the pad coordinates as an index, each compensation value is associated and integrated with its corresponding coordinate to construct a process parameter lookup table. This process achieves personalized adaptation design of laser process parameters, breaking through the limitations of traditional fixed-parameter processes. Model-based calculations ensure the accuracy of compensation values, and the coordinate-indexed lookup table structure enables rapid retrieval of subsequent parameters, providing a reliable basis for precise control of the laser process on a pad-by-pad basis. Simultaneously, it effectively avoids the problem of excessive or insufficient laser energy caused by topography differences, improving chip bonding reliability from the process source and laying a core technological foundation for improving transfer yield.

[0020] 103. Acquire multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate, perform visual fine alignment processing on each chip sub-region image and each pad region image to obtain multiple planar position offsets; In this embodiment, after the donor substrate is loaded, it is first macroscopically aligned with the recipient substrate. The equipment adopts a step-scan working mode to divide the chip array image of the donor substrate and the pad array image of the recipient substrate into corresponding regions, forming multiple matching chip sub-region images and pad region images. The motion platform moves each set of matching chip sub-region images and pad region images to the laser working position area by area. At the working position, static visual fine alignment is performed on the two, capturing and quantifying the positional deviation information between them, and finally obtaining the real-time planar position offset corresponding to each chip sub-region image. The area-by-area static visual fine alignment achieves high-precision positional deviation detection. The quantified planar position offset provides accurate real-time positional basis for subsequent coordinate correction and process parameter data fusion, effectively making up for the accuracy limitations of macroscopic alignment, avoiding chip bonding misalignment problems caused by positional deviation, and balancing alignment accuracy and operational efficiency.

[0021] 104. Based on the process parameter lookup table and multiple planar position offsets, execute instruction generation processing to obtain an execution instruction set corresponding to each chip sub-region image; In this embodiment, execution instruction generation is performed based on the process parameter lookup table and the planar position offset corresponding to each chip sub-region image. Precise adaptation of position and parameters is achieved through data fusion. First, the chip design coordinates are visually offset corrected based on the planar position offset, mapping the design coordinates to actual execution coordinates that fit the actual working conditions. Simultaneously, the corresponding pre-calculated compensation parameters are retrieved from the process parameter lookup table based on the chip design coordinates. These compensation parameters are precisely associated with the corrected actual execution coordinates, completing the fusion of position information and process compensation parameters. Finally, an execution instruction set corresponding one-to-one with each chip sub-region image is generated, ensuring that each chip in the region has its own dedicated execution instruction. This achieves coordinated control of position accuracy and process parameters, breaking through the limitations of separate control of position and parameters in traditional processes. The execution instructions not only take into account the correction of microscopic position deviations but also adapt to the compensation requirements corresponding to the morphological characteristics of the target pads, significantly improving the process accuracy and adaptability of laser transfer. By using dedicated instructions to ensure that the transfer parameters and position status of each chip are highly matched, bonding defects caused by position offset or parameter mismatch are effectively avoided. This provides a stable and reliable process execution basis for the mass transfer of microchip lasers, taking into account both mass production efficiency and transfer yield, and promoting the upgrade of process control to a refined adaptive mode.

[0022] 105. Based on multiple execution instruction sets, complete the laser mass transfer of images of all chip sub-regions.

[0023] In this embodiment, the device retrieves the execution instruction set corresponding to the image of each chip sub-region area area area by area. The laser and beam control system synchronously load the actual execution coordinates corrected for visual offset and the laser process parameters compensated for fused topography from the instruction set, precisely controlling the laser energy output, pulse characteristics, and action position, triggering the laser to complete the peeling, transfer, and bonding operations of all chips in the current sub-region. After the transfer of a single sub-region is completed, the motion platform moves the substrate to the next target sub-region, repeating the above instruction retrieval, parameter loading, and laser triggering process, sequentially completing the transfer operation of all chip sub-region images, and finally realizing the mass laser transfer of the entire board of microchips. By implementing a dedicated instruction set for each chip, the transfer parameters are precisely matched with the shape and actual position of the target pad. This completely eliminates the limitations of the traditional fixed-parameter transfer mode, effectively avoiding defects such as chip damage, poor soldering, and bonding misalignment. It significantly improves transfer yield and bonding reliability, balancing the mass production efficiency of mass transfer with the process precision of a single chip. At the same time, relying on feedforward control logic reduces the dependence on subsequent offline detection and repair processes, shortens the production cycle, and reduces overall manufacturing costs, providing stable technical support for the large-scale, high-precision mass production of microchip laser mass transfer.

[0024] In this embodiment of the invention, the acceptor substrate includes a plurality of pads; the step of obtaining the morphology dataset of the acceptor substrate and constructing a morphology digital map based on the morphology dataset includes: performing three-dimensional measurement processing on the acceptor substrate using a morphology scanning tool to obtain the morphology dataset, the morphology dataset including morphology data corresponding one-to-one with each of the pads; obtaining the pad coordinates of each pad; and matching each pad coordinate with each morphology data one-to-one to construct the morphology digital map.

[0025] In this embodiment, multiple pads are distributed on the surface of the acceptor substrate. When acquiring the topography dataset and constructing a digital topography map, a topography scanning tool integrating a high-precision three-dimensional optical sensor and independent of the laser transfer head is used. Before the laser transfer process begins, a full-coverage non-contact three-dimensional measurement is performed on the loaded acceptor substrate. This measurement process accurately collects the topography information corresponding to each pad, forming a topography dataset containing key geometric features such as the height, tilt, and roughness of each pad, while simultaneously acquiring the pad coordinates corresponding to each pad. After calibrating the collected topography data, each set of pad topography data is precisely matched one-to-one with the corresponding pad coordinates. Through data integration, a digital topography map of the entire board's pads is constructed, completely and accurately representing the microscopic geometric state of all pads on the substrate surface. The calibration of the collected topography data involves standardization and normalization of various systematic errors and random deviations generated during the three-dimensional optical measurement process. Non-contact measurement can effectively avoid physical damage to the substrate and pads. The binding of morphology data with pad coordinates provides reliable data support for the calculation of subsequent laser process parameters. It breaks through the technical limitation of traditional two-dimensional vision that can only identify the planar position of pads, and eliminates the blindness of process control caused by the lack of pad morphology information from the source.

[0026] In this embodiment of the invention, the digital topography map includes multiple topography data; the step of constructing a process parameter query table based on the digital topography map includes: calling a pre-trained data-driven model; calculating laser process compensation parameters for each topography data based on the data-driven model to obtain a laser parameter compensation dataset; and constructing a process parameter query table based on the laser parameter compensation dataset and the digital topography map.

[0027] In this embodiment, the digital topography map contains topography data corresponding to each pad. When constructing a process parameter lookup table based on this map, a pre-trained data-driven model is first invoked. This model employs a deep learning architecture combining a multilayer perceptron and an attention mechanism. The input layer connects to the pad topography feature dimension, performs non-linear feature mapping through multiple fully connected hidden layers, and embeds a channel attention module in between. Features are extracted through pooling, and weights are calculated using a small multilayer perceptron. Core features such as height deviation are weighted and strengthened before being passed to the output layer via a fully connected layer, outputting the corresponding laser process compensation parameters. The training data comes from massive pad topography feature data, corresponding laser process parameter data, and corresponding process feedback data such as chip laser transfer bonding intensity and transfer yield. After annotation, normalization, noise reduction, and data augmentation preprocessing, the above data completes the training and iterative optimization of the model, ensuring the model's parameter prediction accuracy and scenario adaptability for different topography features. Based on this pre-trained data-driven model, targeted calculations of laser process compensation parameters are performed on the morphology data of each pad in the morphology digital map, combined with micro-geometric features such as pad height deviation and slope. Compensation values ​​for core process parameters such as laser energy and pulse width are output, forming a laser parameter compensation dataset. Subsequently, a process parameter lookup table is constructed based on the laser parameter compensation dataset and the morphology digital map. This process uses a data-driven model trained on a large amount of process data, which can capture the inherent correlation between pad morphology features and laser process parameters, significantly improving the accuracy and adaptability of compensation parameter calculations, allowing compensation parameters to highly fit the micro-morphological states of different pads.

[0028] In this embodiment of the invention, the laser parameter compensation dataset includes multiple laser parameter compensation data. The step of constructing a process parameter query table based on the laser parameter compensation dataset and the topography digital map includes: obtaining the pad coordinates corresponding to each laser parameter compensation data based on the topography digital map; establishing a mapping relationship between each laser parameter compensation data and the corresponding pad coordinates to generate multiple related items; and integrating all the related items to construct the process parameter query table.

[0029] In this embodiment, after the laser parameter compensation data is calculated by the data-driven model, the pad coordinates corresponding to each set of laser parameter compensation data are extracted from the digital topography map. This ensures that the laser parameter compensation data and the target pads correspond precisely. These coordinates are associated with both the spatial location of the pads and their core topographic features such as height deviation and slope, providing a foundation for the precise binding of compensation data and pads. Based on this, a unique mapping relationship is established between each set of laser parameter compensation data and the corresponding pad coordinates, generating several associated items. Each associated item carries the spatial location information of a specific pad and the appropriate laser process compensation parameters, achieving a deep binding between the compensation requirements derived from topographic features and the spatial location of the pads. Subsequently, all associated items are structurally integrated, and a process parameter lookup table is constructed using the pad coordinates as an index. This enables the lookup table to quickly locate and retrieve the corresponding compensation parameters by coordinates, ensuring that subsequent process steps can efficiently match the compensation requirements of the target pads. Through coordinate association, a one-to-one precise matching between compensation parameters and pads is achieved, abandoning the coarseness of the traditional fixed parameter mode and allowing each set of compensation parameters to be specifically matched to the topographic features of the corresponding pads. The indexed lookup table architecture significantly improves the efficiency of retrieving compensation parameters in subsequent processes, adapts to the mass production rhythm of laser mass transfer, and builds a complete data association link between morphological features, pad coordinates and laser parameters, providing reliable support for personalized process control on a pad-by-pad basis. It effectively avoids the problem of laser energy mismatch caused by morphological differences, reduces defects such as chip damage and cold solder joints, and improves transfer yield and bonding reliability.

[0030] In this embodiment of the invention, acquiring multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate includes: acquiring a preset partitioning rule, a chip array image of the donor substrate, and a pad array image of the recipient substrate; performing initial position matching processing on the donor substrate and the recipient substrate to establish a positional correspondence between the donor substrate and the recipient substrate; and partitioning the chip array image and the pad array image into regions based on the partitioning rule to obtain multiple chip sub-region images and a pad region image corresponding to each chip sub-region image.

[0031] In this embodiment, a preset partitioning rule, a chip array image of the donor substrate, and a pad array image of the recipient substrate are first acquired. The partitioning rule is preset based on chip density, transfer efficiency, and equipment movement stroke, clearly defining the partition size, quantity, and boundary judgment criteria. The chip array image and the pad array image are respectively the orderly arranged chip set on the donor substrate and the corresponding pad set on the recipient substrate, serving as the basic carrier for region partitioning. Subsequently, initial position matching processing is performed on the donor substrate and the recipient substrate. Through equipment positioning reference calibration and visual coarse positioning, macroscopic positional deviations caused by substrate loading and initial placement are eliminated, establishing an overall positional correspondence between the two, ensuring that the spatial orientation of the chip array image and the pad array image is initially aligned. Based on the preset partitioning rule, the chip array image and the pad array image are synchronously partitioned. According to the boundaries set by the rule, the chip array image is divided into multiple independent chip sub-region images, and the pad array image is correspondingly partitioned to obtain several pad region images, ensuring that each chip sub-region image can accurately correspond to a unique pad region image, guaranteeing the regionalized collaborative execution of subsequent transfer processes. The initial matching establishes a positional correspondence that ensures zonal synchronization, preventing subsequent transfer deviations caused by sub-region misalignment. Zonal processing adapts to the step-scanning working mode of laser mass transfer, enabling precise alignment and transfer operations zone by zone, balancing mass production efficiency and positioning accuracy. Simultaneously, the clear sub-regional correspondence provides a clear regional benchmark for subsequent accurate retrieval of process parameters and visual precision alignment, reducing cross-regional interference and improving the overall process consistency and controllability.

[0032] In this embodiment of the invention, the step of performing visual fine alignment processing on each chip sub-region image and each solder pad region image to obtain multiple planar position offsets includes: obtaining multiple preset chip design coordinates; transferring each chip sub-region image and its corresponding solder pad region image to a laser working position; performing visual fine alignment processing on the chip sub-region image and its corresponding solder pad region image at the laser working position to obtain the actual transfer coordinates corresponding to each chip sub-region image; and calculating the position offset between each chip sub-region image and its corresponding solder pad region image based on the multiple chip design coordinates and the multiple actual transfer coordinates to obtain multiple planar position offsets.

[0033] In this embodiment, the preset chip design coordinates are the theoretical layout coordinates of the chips on the donor substrate, derived from the design drawings of the chip array image. These coordinates define the ideal position of each chip and its preset alignment with the corresponding pads on the recipient substrate, serving as the core reference for subsequent position calibration, deviation calculation, and process parameter retrieval. After the donor substrate is loaded, it is first macroscopically aligned with the recipient substrate to eliminate macroscopic positional deviations caused by initial placement and loading, establishing a basic attitude reference for subsequent precise processing. Simultaneously, the preset chip design coordinates are retrieved. The equipment adopts a step-scan working mode. The motion platform moves the image of each chip sub-region and its corresponding pad region to the laser workstation area area by area. At the workstation area, a high-precision vision inspection system performs static visual precision alignment processing on the two. This system is based on an industrial high-resolution area scan camera, equipped with a telecentric lens and a coaxial light source to form an imaging unit, and also integrates a vision processing unit to achieve high-precision imaging and real-time coordinate calculation. During the precision alignment process, a complete coordinate transformation is completed. First, the imaging unit acquires clear images of the chip and pads, extracts the image pixel coordinates of the feature points of the two, and then uses the internal and external participation hand-eye calibration matrix obtained from the camera's pre-calibration to convert the pixel coordinates into visual coordinates. Finally, combined with the coordinate system mapping relationship of the equipment's motion platform, it is converted into the equipment motion coordinates of the laser workstation, that is, the actual transfer coordinates corresponding one-to-one with each chip sub-region image. These coordinates truly reflect the actual alignment state of the chip and pads at the laser workstation area and are the positional basis for the chip to actually perform the transfer operation. Using the preset chip design coordinates as a reference, the coordinates are compared and quantized point by point with the measured actual transfer coordinates. The planar position offset between each chip sub-region image and the corresponding pad region image is calculated, which fully characterizes the microscopic position deviation between the actual alignment state and the theoretical design state.

[0034] In this embodiment, by clearly defining the baseline and measured attributes of theoretical design coordinates and actual transfer coordinates, a precise comparison system between theory and practice is established. The differential analysis between the two enables precise capture and quantification of microscopic positional deviations, effectively compensating for the limitations of macroscopic alignment accuracy. This provides a clear theoretical reference and a realistic working condition basis for positional deviation calculations. The step-scanning zone-by-zone processing mode adapts to the mass production rhythm of laser mass transfer, while the actual transfer coordinates obtained through static vision precision alignment ensure that each position calibration matches the actual working conditions of the current workstation, avoiding accuracy errors caused by uniform calibration across the entire domain. The quantified planar position offset provides accurate real-time position data for subsequent coordinate correction and data fusion, ensuring that the coordinates and process parameters of subsequent transfer operations accurately match the actual alignment state. This effectively avoids chip bonding misalignment problems caused by microscopic positional deviations. Simultaneously, the correlation and comparison between theoretical and actual coordinates constructs a traceable data link for positional deviations, providing precise positional data support for subsequent process optimization, balancing the mass production efficiency and microscopic alignment accuracy of laser mass transfer.

[0035] In this embodiment of the invention, the step of generating execution instructions based on the process parameter lookup table and multiple planar position offsets to obtain an execution instruction set corresponding to each chip sub-region image includes: using multiple chip design coordinates as indexes, obtaining laser parameter compensation data corresponding one-to-one with each chip design coordinate from the process parameter lookup table; performing visual offset correction processing on the corresponding chip design coordinates based on each planar position offset to obtain chip execution coordinates corresponding one-to-one with each chip design coordinate; and performing data fusion processing based on each laser parameter compensation data and the corresponding chip execution coordinates to obtain an execution instruction set corresponding to each chip sub-region image.

[0036] In this embodiment, chip design coordinates are used as an index to accurately extract laser parameter compensation data corresponding to each set of design coordinates from the process parameter lookup table. Based on the preset association between design coordinates and pad morphology, the compensation data is ensured to accurately match the microscopic morphology requirements of the target chip and corresponding pads, providing a foundation for subsequent process parameter adaptation. For each set of chip design coordinates, visual offset correction processing is performed using the planar position offset obtained from visual alignment. Using the chip design coordinates as a reference, the X-axis translation deviation and Y-axis translation deviation decoupled from the corresponding planar position offset are used to perform coordinate compensation transformation on each set of chip design coordinates. This calibrates the theoretical design coordinates to chip execution coordinates adapted to the actual working conditions of the laser working position, eliminating microscopic position deviations caused by factors such as substrate loading deformation and motion platform positioning fluctuations, ensuring that the coordinate data adapts to the current transfer conditions. Based on this, data fusion processing is carried out to precisely associate and bind each set of laser parameter compensation data with the corresponding chip execution coordinates. The position guidance information and personalized process compensation requirements are integrated into one, and finally, an execution instruction set corresponding to each chip sub-region image is generated. This ensures that the transfer operation of each chip in the region can obtain exclusive position and parameter guidance. Data fusion realizes the deep binding of position information and process parameters, so that laser transfer has both a precise position reference and personalized adaptation of process parameters according to the pad morphology. It breaks through the limitations of traditional fixed parameters and single coordinate control. The integrated instruction set takes into account the mass production efficiency of mass transfer and the transfer accuracy of single chip, promotes the upgrade of process control from open-loop extensive mode to closed-loop precise mode, and enhances the traceability of process data, significantly improving the yield and stability of chip transfer.

[0037] The microchip laser mass transfer method in the embodiments of the present invention has been described above. The microchip laser mass transfer device in the embodiments of the present invention will be described below. Please refer to [link / reference]. Figure 2 One embodiment of the microchip laser mass transfer device of the present invention includes: Topographic digital map construction module 201: used to acquire the topographic dataset of the receptor substrate and construct a topographic digital map based on the topographic dataset; Process parameter query table construction module 202: used to construct a process parameter query table based on the topographic digital map; Offset calculation module 203: used to acquire multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate, perform visual fine alignment processing on each chip sub-region image and each pad region image to obtain multiple planar position offsets; Instruction generation module 204: is used to perform instruction generation processing based on the process parameter lookup table and multiple planar position offsets to obtain an execution instruction set corresponding to each chip sub-region image; Instruction execution module 205: used to complete the laser mass transfer of all chip sub-region images based on multiple execution instruction sets.

[0038] Based on the same ideas as the methods in the above embodiments, the apparatus provided in this application can implement the methods in the above embodiments.

[0039] above Figure 2 The microchip laser mass transfer device in the embodiments of the present invention will be described in detail from the perspective of modular functional entities. The microchip laser mass transfer device in the embodiments of the present invention will be described in detail from the perspective of hardware processing.

[0040] Figure 3 This is a schematic diagram of a microchip laser mass transfer device 300 provided in an embodiment of the present invention. The microchip laser mass transfer device 300 can vary significantly due to different configurations or performance characteristics. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors) and a memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) storing application programs 333 or data 332. The memory 320 and storage media 330 can be temporary or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the microchip laser mass transfer device 300. Furthermore, the processor 310 may be configured to communicate with the storage media 330 and execute the series of instruction operations in the storage media 330 on the microchip laser mass transfer device 300 to implement the steps of the microchip laser mass transfer method provided in the above-described method embodiments.

[0041] The microchip laser mass transfer device 300 may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 3 The structure of the microchip laser mass transfer device shown does not constitute a limitation on the microchip laser mass transfer device. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0042] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the microchip laser mass transfer method.

[0043] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0044] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0045] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for mass laser transfer in a microchip, characterized in that, include: Obtain a morphology dataset of the receptor substrate, and construct a morphology digital map based on the morphology dataset; A process parameter lookup table is constructed based on the aforementioned digital topography map; Multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate are acquired. Visual fine alignment processing is performed on each chip sub-region image and each pad region image to obtain multiple planar position offsets. Based on the process parameter lookup table and multiple planar position offsets, an instruction generation process is performed to obtain an execution instruction set corresponding to each chip sub-region image; Laser mass transfer of images from all chip sub-regions is completed based on multiple execution instruction sets.

2. The microchip laser mass transfer method according to claim 1, characterized in that, The acceptor substrate includes multiple pads; the step of acquiring the morphology dataset of the acceptor substrate and constructing a morphology digital map based on the morphology dataset includes: The acceptor substrate is subjected to three-dimensional measurement processing using a topology scanning tool to obtain the topology dataset, which includes topology data corresponding to each of the pads. Obtain the pad coordinates for each of the pads; Each pad coordinate is matched one-to-one with each topography data to construct the topography digital map.

3. The microchip laser mass transfer method according to claim 1, characterized in that, The digital topography map includes multiple topography data; the construction of a process parameter lookup table based on the digital topography map includes: Call the pre-trained data-driven model; Based on the data-driven model, laser process compensation parameters are calculated for each of the morphology data to obtain a laser parameter compensation dataset. A process parameter lookup table is constructed based on the laser parameter compensation dataset and the topographic digital map.

4. The microchip laser mass transfer method according to claim 3, characterized in that, The laser parameter compensation dataset includes multiple laser parameter compensation data sets. The step of constructing a process parameter lookup table based on the laser parameter compensation dataset and the topographic digital map includes: Based on the digital topography map, obtain the pad coordinates corresponding to each of the laser parameter compensation data; Establish a mapping relationship between each laser parameter compensation data and the corresponding pad coordinates, and generate multiple related items; Integrate all the aforementioned related items to construct the process parameter lookup table.

5. The microchip laser mass transfer method according to claim 1, characterized in that, The acquisition of multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate includes: Obtain the preset partitioning rules, the chip array image of the donor substrate, and the pad array image of the acceptor substrate; The donor substrate and the recipient substrate are subjected to initial position matching processing to establish the positional correspondence between the donor substrate and the recipient substrate; Based on the division rules, the chip array image and the pad array image are divided into regions to obtain multiple chip sub-region images and a pad region image corresponding to each chip sub-region image.

6. The microchip laser mass transfer method according to claim 1, characterized in that, The visual alignment processing of each chip sub-region image and each solder pad region image yields multiple planar position offsets, including: Obtain multiple preset chip design coordinates; Each chip sub-region image and its corresponding solder pad region image are respectively transferred to the laser working position; Visual alignment processing is performed on the chip sub-region image and the corresponding solder pad region image at the laser working position to obtain the actual transfer coordinates corresponding to each chip sub-region image; Based on multiple chip design coordinates and multiple actual transfer coordinates, the positional offset between each chip sub-region image and the corresponding pad region image is calculated to obtain multiple planar positional offsets.

7. The microchip laser mass transfer method according to claim 6, characterized in that, The instruction generation process based on the process parameter lookup table and multiple planar position offsets yields an execution instruction set corresponding to each chip sub-region image, including: Using multiple chip design coordinates as indexes, laser parameter compensation data corresponding to each chip design coordinate is obtained from the process parameter lookup table; Based on each of the plane position offsets, the corresponding chip design coordinates are visually offset corrected to obtain chip execution coordinates that correspond one-to-one with each of the chip design coordinates. Data fusion processing is performed based on the laser parameter compensation data and the corresponding chip execution coordinates to obtain the execution instruction set corresponding to each chip sub-region image.

8. A microchip laser mass transfer device, characterized in that, include: Topographic digital map construction module: used to acquire the topographic dataset of the receptor substrate and construct a topographic digital map based on the topographic dataset; Process parameter query table construction module: used to construct a process parameter query table based on the topographic digital map; Offset calculation module: used to acquire multiple chip sub-region images of the donor substrate and multiple pad region images of the recipient substrate, perform visual fine alignment processing on each chip sub-region image and each pad region image to obtain multiple planar position offsets; Instruction generation module: used to perform instruction generation processing based on the process parameter lookup table and multiple planar position offsets to obtain an execution instruction set corresponding to each chip sub-region image; Instruction execution module: used to complete the laser mass transfer of images of all chip sub-regions based on multiple execution instruction sets.

9. A microchip laser mass transfer device, characterized in that, The microchip laser mass transfer device includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the microchip laser mass transfer device to perform the steps of the microchip laser mass transfer method as claimed in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the various steps of the microchip laser mass transfer method as described in any one of claims 1-7.

Images