LED chip high-heat-dissipation packaging method based on heat diffusion path reconstruction

Abstract

This application relates to the field of semiconductor packaging technology, and in particular to a high-heat-dissipation packaging method for LED chips based on thermal diffusion path reconstruction. The method includes: performing void detection on an LED chip package in a pre-cured or semi-cured state to obtain spatial distribution information of micro-scale voids and form a void topology map; determining the main thermal diffusion region within the LED chip package; identifying key void clusters based on the void topology map and the main thermal diffusion region; classifying the key void clusters according to their thermal resistance mechanisms and performing matching void topology reshaping processing; constructing a continuous thermally conductive structure along the heat flow path within the main thermal diffusion region; and performing partitioned curing of the continuous thermally conductive structure, ensuring that non-main thermal regions retain controlled micro-hole buffer zones. This application constructs a packaging structure combining a continuous thermally conductive channel in the main thermal region with a controlled micro-hole buffer zone in the non-main thermal region, effectively reducing interface thermal resistance and improving packaging reliability.

Description

Technical Field

[0001] This application relates to the field of semiconductor packaging technology, and in particular to a high heat dissipation packaging method for LED chips based on thermal diffusion path reconstruction. Background Technology

[0002] As a core component of modern lighting and display technologies, LED chips are experiencing continuously increasing power density, placing higher demands on their heat dissipation performance. In the semiconductor packaging field, thermal management is crucial for ensuring the long-term stable operation of LED chips. Current technologies typically employ multi-layered LED chip packaging designs, including the chip body, die-bonding layer, support layer, encapsulant layer, and substrate layer. Heat is conducted outwards layer by layer through these media. To improve heat dissipation performance, the industry commonly employs methods to increase the thermal conductivity of materials, such as selecting die-bonding adhesives and encapsulation materials with high thermal conductivity. Simultaneously, structures such as metal heat sinks, ceramic substrates, or external heat sinks are used to enhance the heat conduction path. Furthermore, optimizing the package shape, increasing the heat dissipation interface area, and using processes such as vacuum degassing and agitation degassing to reduce internal air bubbles are also employed to improve overall heat conduction efficiency.

[0003] However, in existing technologies, the relationship between the distribution of microscale voids within the package and the heat diffusion path has not received sufficient attention, leading to increased local thermal resistance during heat transfer. (Summary of the Invention)

[0004] This application provides a high heat dissipation packaging method for LED chips based on thermal diffusion path reconstruction to solve the above problems. The method includes:

[0005] S1. Perform void detection on the LED chip package in the pre-cured or semi-cured state, and divide it into multiple continuous detection units along the preset heat transfer direction from the main heat-generating area of ​​the LED chip to the external heat dissipation interface. Obtain the spatial distribution information of micro-scale voids in each detection unit, extract at least one void topology parameter from the void position, equivalent aperture, aspect ratio, spacing between adjacent voids and void main axis direction, and form a void topology distribution map.

[0006] S2. Based on the heat transfer relationship between the main heat-generating area of ​​the LED chip and the external heat dissipation interface, determine the main heat diffusion area within the LED chip package. The main heat diffusion area includes at least two of the following: the main heat diffusion core area, the axial heat conduction area, and the lateral transition heat conduction area.

[0007] S3. Based on the void topology distribution map and the main heat diffusion zone, cluster and identify multiple voids located in the same detection unit and arranged sequentially along the main heat flow direction. When the multiple voids meet at least two of the following conditions, they are determined to be key void clusters: the total projected length in the main heat flow direction is greater than a preset multiple of the average aperture of a single void; the spacing between adjacent voids is less than a preset connection threshold; the angle between the void main axis direction and the main heat flow direction is less than a preset angle threshold; and they are located in the main heat diffusion core area or the lateral transition heat conduction area.

[0008] S4. The key void clusters are classified according to their heat resistance mechanisms, including volume-dominant void clusters, orientation-dominant void clusters, and connectivity-dominant void clusters.

[0009] S5. Perform matching void topology reshaping processing for different types of critical void clusters. Specifically, for volume-dominant void clusters, perform local filling and / or local compression; for orientation-dominant void clusters, perform directional extrusion, local hot pressing, and / or void orientation deflection processing; and for connectivity-dominant void clusters, perform thermal bridging and / or connectivity chain disconnection processing.

[0010] S6. After completing the void topology reshaping process, a continuous thermal conduction structure is constructed sequentially along the heat flow path in the main heat diffusion region. The continuous thermal conduction structure includes at least two of the following: a low void thermal conduction region directly below the chip, a continuous thermal conduction section extending along the package thickness direction, and a heat flow turning bridging section.

[0011] S7. The continuous thermally conductive structure is partitioned and cured, wherein the main heat diffusion area is cured in stages first, and the non-main heat area is cured as a whole, so that the non-main heat area retains a controlled micropore buffer zone, so as to obtain a high heat dissipation packaging structure for LED chips.

[0012] Through the above technical solutions, by constructing a void topology distribution map, defining the three-dimensional main heat diffusion zone, identifying key void clusters through multi-condition clustering, classifying by heat resistance mechanism type, matching topology reshaping, constructing a continuous thermally conductive structure in segments, and solidifying the main heat zone / non-main heat zone in stages, we have achieved a higher-dimensional understanding and targeted intervention of microscale voids inside the package from random defects to modelable thermal objects. By using topological parameters such as void location, equivalent aperture, aspect ratio, adjacent spacing, and main axis direction, and spatial-directional-connectivity coupling with the main heat diffusion path, we can accurately identify the key void clusters that truly affect the continuity of heat flow. Then, based on their dominant heat resistance mechanisms (volume blocking, directional blocking, connectivity cutting), we can perform differentiated physical reshaping to form a continuous thermally conductive chain with low voids, low blocking, and low connectivity heat resistance in the main heat zone. At the same time, we retain a controlled micro-hole buffer zone in the non-main heat zone to effectively release thermal expansion mismatch stress, and ultimately significantly improve the heat dissipation performance of the LED chip without sacrificing the reliability of the package.

[0013] Optionally, the gap detection in step S1 may employ at least one of the following: X-ray detection, industrial CT detection, scanning acoustic microscopy detection, infrared thermography-assisted inversion detection, and section microscopy detection.

[0014] The formation of the void topology distribution map includes: statistically analyzing the degree of overlap of the projections of voids in each detection unit along the preset heat transfer direction, and combining the spacing between adjacent voids to form a distribution map characterizing the void aggregation state and connectivity trend.

[0015] By integrating multiple void detection methods, the above technical solutions cover the blind spots in the identification of spherical, cracked, interface debonding, and dynamic thermal response voids, thereby improving the completeness and confidence of void detection in S1. Furthermore, through dual-dimensional modeling using projection overlap statistics and adjacent void spacing analysis, the void topology map is upgraded from a simple location map to a functional map with heat flow influence prediction capabilities. This provides reliable input for the clustering and identification of key void clusters in step S3 of this application, ensuring that subsequent topology reshaping actions accurately target the void set that truly influences the main heat diffusion path, rather than random voids. This supports the effective implementation of the zonal governance strategy of dense heat conduction in the main heat zone and buffering and retention in non-main heat zones.

[0016] Optionally, in step S2, the main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the package thickness direction;

[0017] The axial heat conduction zone is a heat conduction area formed by extending the main heat diffusion core area toward the external heat dissipation interface.

[0018] The lateral transition thermal conduction zone is a heat flow transition area formed by the extension of the LED chip edge toward the bracket or substrate.

[0019] Through the above technical solution, the main heat diffusion core area anchors the heat source output end, the axial heat conduction zone establishes the longitudinal path of the main heat flow, and the lateral transition heat conduction zone covers the heat flow turning interface. The three elements work together to construct a three-dimensional main heat diffusion domain model that highly matches the actual physical path of heat flow. On this basis, the determination of key void clusters in S3 no longer relies on subjective experience or global statistics, but has a technical foundation that allows for spatial location, path tracing, and regional verification. This ensures the accuracy and stability of cluster identification based on void topology distribution map and main heat diffusion zone in this application.

[0020] Optionally, in step S3, when clustering and identifying multiple gaps located within the same detection unit and arranged sequentially along the main heat flow direction, multiple gaps that meet at least two of the following conditions are identified as key gap clusters:

[0021] The total projected length in the main heat flow direction is greater than 2 to 10 times the average pore size of a single pore;

[0022] The spacing between adjacent gaps is less than a preset connection threshold;

[0023] The angle between the main axis direction of the void and the main heat flow direction is less than a preset angle threshold.

[0024] The multiple gaps are located in the main heat diffusion core area or the lateral transition heat conduction area.

[0025] By using the above technical solution, the total projected length is limited to 2 to 10 times the average aperture of a single gap, combined with the spacing between adjacent gaps being less than a preset connection threshold, the angle between the main axis of the gap and the main heat flow direction being less than a preset angle threshold, and the positional constraints of multiple gaps located in the main heat diffusion core area or the lateral transition heat conduction area, multidimensional coupling identification of key gap clusters is achieved. On this basis, the identified key gap clusters can accurately reflect the spatial distribution and structural characteristics of the real thermal resistance bottleneck inside the package.

[0026] Optionally, in step S4, the key void clusters are classified according to their heat-insulating mechanisms, specifically including:

[0027] Key void clusters dominated by aperture-enlarged voids are classified as volume-dominant void clusters.

[0028] The key void clusters, which are mainly composed of slender, fissure-type voids and whose angle between the void principal axis and the main heat flow direction is less than the preset angle threshold, are classified as direction-dominant void clusters.

[0029] Key void clusters formed by connecting two or more adjacent voids in series along the main heat flow direction are classified as connected dominant void clusters.

[0030] By classifying key void clusters into three categories—volume-dominant, direction-dominant, and connectivity-dominant—the physical essence of the void thermal insulation mechanism is mapped: volume-dominant corresponds to heat flow cross-section obstruction and thermal capacity disturbance, direction-dominant corresponds to heat flow path bypass induction, and connectivity-dominant corresponds to structural cutting of heat flow channels. Based on this, the three classification results serve as the triggering basis for the corresponding reshaping actions, ensuring that the void topology reshaping strategy is strictly matched with the actual thermal insulation physical mechanism, and avoiding the limitation of heat dissipation performance improvement or the decrease in reliability caused by the mismatch between processing methods and thermal insulation morphology.

[0031] Optionally, for the volume-dominant void cluster, the void topology reshaping process in step S5 includes:

[0032] A thermally conductive filler material is introduced into the region where the volume-dominant void cluster is located, allowing the thermally conductive filler material to penetrate along the gap between the void and the surrounding encapsulation medium, and applying local pressure to the region after the thermally conductive filler material has penetrated, so as to reduce the cavity volume of the volume-dominant void cluster.

[0033] The direction of the applied local pressure is intersected with the direction of the main heat flow;

[0034] The thermally conductive filler material includes at least one of silver filler, alumina filler, boron nitride filler, aluminum nitride filler, graphene filler, and carbon nanotube filler.

[0035] Through the above technical solution, by infiltrating the thermally conductive filling material along the gap between the voids and the surrounding encapsulation medium, combined with the local pressure set at the intersection with the main heat flow direction, and the synergistic application of at least one filler among silver, alumina, boron nitride, aluminum nitride, graphene, and carbon nanotubes, a triple effect of targeted filling and coverage, structural compression, and interface anchoring of volume-dominant void clusters is achieved. On this basis, the risk of material intrusion into non-target areas is avoided by using the gap infiltration path, the shear component introduced by the intersecting pressure direction is improved by improving the interface bonding stability, and the selection of multiple types of fillers covers the differentiated requirements of different encapsulation levels for thermal conductivity, insulation, thermal matching, and anisotropy. Finally, without destroying the overall encapsulation structure, the local thermal resistance in the main heat diffusion region is significantly reduced, providing a repeatable, controllable, and verifiable process implementation path for the LED chip high heat dissipation encapsulation method based on microscale void topology reshaping in this application.

[0036] Optionally, for the direction-dominant void cluster, the void topology reshaping process in step S5 includes:

[0037] When the LED chip package is in a pre-cured or semi-cured state, directional extrusion or local hot pressing is applied to the area where the direction-dominant void cluster is located, so that the long axis direction of the direction-dominant void cluster deviates from the main heat flow direction, or the direction-dominant void cluster is split into multiple sub-voids that are dispersed along the non-main heat flow direction. After directional extrusion or local hot pressing, the area is pre-cured to fix the void orientation after reshaping.

[0038] Through the above technical solution, by performing directional extrusion or local hot pressing on the directional dominant void clusters in the pre-cured state, the long axis of the voids is deflected or split by the viscoelastic response of the encapsulation medium. Then, by rapidly freezing the new orientation through pre-curing, the slender cracks that originally formed low-resistance bypass channels along the main heat flow direction are transformed into discontinuous barrier structures with multi-directional dispersion, tortuous paths, and enhanced heat flow scattering. On this basis, the stable curing of the void orientation avoids the performance degradation caused by stress relaxation during thermal cycling, thereby significantly improving the continuity of the heat flow path and the uniformity of thermal resistance in the main heat diffusion zone without increasing the overall densification cost.

[0039] Optionally, for the connectivity-dominant gap cluster, the gap topology reshaping process in step S5 includes:

[0040] A thermally conductive bridging band is constructed between adjacent gaps, or the encapsulation medium between adjacent gaps is reclosed by localized thermal pressing to break the series connection relationship of the dominant gap cluster along the main heat flow direction;

[0041] The thermally conductive bridging strip extends in a direction that intersects with the main heat flow direction to form a cross-gap thermally conductive connection section between adjacent gaps.

[0042] Through the above technical solution, by constructing thermally conductive bridging bands extending in cross directions between adjacent gaps and supplementing this with localized hot pressing to reclose the encapsulation medium, a dual-severing mechanism for the series connection relationship of the dominant gap cluster is achieved: On the one hand, the thermally conductive bridging bands, acting as high thermal conductivity bridges, force heat flow to traverse the gaps along the shortest effective path, and their cross-orientation design significantly improves the thermally conductive cross-sectional area and thermal contact length per unit heat flow direction; on the other hand, localized hot pressing, by regulating the rheological behavior of the material, repairs the weak medium layer between gaps at the microscale, eliminating the connection path from a physical perspective. The synergistic effect of these two methods avoids the material redundancy and interface weakening risks caused by simple filling, and overcomes the high dependence of single hot pressing on gap spacing and medium state. This allows for stable continuous reconstruction of the main heat flow path under different process windows, supporting the implementation of S5 thermally conductive bridging and / or connection chain disconnection techniques, and providing a microscopic basis for the macroscopic structural construction of the heat flow redirection bridging segment.

[0043] Optionally, in step S6, a continuous heat conduction structure is constructed sequentially along the heat flow path within the main heat diffusion zone. Specifically, this includes: first, constructing a low-gap heat conduction zone directly below the chip in the main heat diffusion core zone; then, constructing a continuous heat conduction segment extending along the package thickness direction in the axial heat conduction zone; and finally, constructing a heat flow turning bridge segment in the lateral transition heat conduction zone, so that a continuous heat conduction chain is formed between the main heat-generating area of ​​the LED chip and the external heat dissipation interface.

[0044] The above technical solution involves constructing a low-gap thermal conductivity zone directly beneath the chip in the main heat diffusion core area to receive the chip's heat source output and suppress initial thermal resistance. Next, a continuous thermal conductivity section extending along the package thickness direction is constructed in the axial thermal conductivity zone to open up the longitudinal main heat flow channel. Finally, a heat flow turning bridge section is constructed in the lateral transition thermal conductivity zone to bridge the thermal conductivity gap at the heat flow transition interface. These three components are constructed in strict order according to the physical path of heat flow and are connected step by step, together forming a continuous, uninterrupted heat conduction chain from the main heat-generating area of ​​the LED chip to the external heat dissipation interface. This significantly improves the continuity of the heat flow path and the efficiency of heat conduction without changing the overall packaging material system, solving the problem of increased local thermal resistance and hot spot accumulation caused by mismatched gap space distribution in existing technologies.

[0045] Optionally, the partition curing in step S7 includes:

[0046] After completing the void topology reshaping process of the main heat diffusion core area and the axial heat conduction area, the main heat diffusion core area and the axial heat conduction area are first solidified in the first stage; after the heat flow turning bridge section is constructed, the lateral transition heat conduction area is solidified in the second stage; and after the main heat diffusion area is solidified, the non-main heat area is solidified as a whole so that the non-main heat area retains the controlled micropore buffer zone.

[0047] The controlled micropore buffer zone has a micropore volume fraction of 0.1% to 8%.

[0048] Through the above technical solution, the low-void thermal conductivity structure of the main heat diffusion core area and the axial heat conduction area is solidified and locked in the first stage. The heat flow turning bridging section in the lateral transition heat conduction area is stabilized by solidifying in the second stage. Finally, the non-main heat area is solidified as a whole in the third stage and its micropore volume fraction is precisely controlled to be 0.1% to 8%. This achieves the synergistic unity of the stability of the main heat path structure and the stress buffering function of the non-main heat area. On this basis, the thermal conductivity structure of each section of the main heat diffusion area will not deteriorate due to subsequent process disturbances, and the micropore buffer zone of the non-main heat area effectively absorbs interface stress in the thermal cycle. Thus, while ensuring the high heat dissipation performance of the LED chip, the long-term reliability of the package is significantly improved. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a flowchart illustrating a high heat dissipation packaging method for LED chips based on thermal diffusion path reconstruction, as provided in one embodiment of this application. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0052] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0053] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0054] See Figure 1 As shown in the figure, an embodiment of this application provides a high heat dissipation packaging method for LED chips based on microscale gap topology reshaping. The method includes the following steps:

[0055] Step 1: Perform void detection on the LED chip package in the pre-cured or semi-cured state, and divide it into multiple continuous detection units along the preset heat transfer direction from the main heat-generating area of ​​the LED chip to the external heat dissipation interface. Obtain the spatial distribution information of micro-scale voids in each detection unit, extract at least one void topology parameter from the void position, equivalent aperture, aspect ratio, spacing between adjacent voids, and void principal axis direction, and form a void topology distribution map.

[0056] Among them, void detection (one or more of X-ray detection, industrial CT detection, scanning acoustic microscopy detection, infrared thermography-assisted inversion detection, and section microscopy detection) is used to obtain the three-dimensional spatial coordinates, morphological characteristics, and neighborhood relationships of microscale voids inside the package during the pre-curing or semi-curing stage when the packaging medium still has a certain degree of fluidity; the preset heat transfer direction is the vector direction from the center of the main heat-generating area of ​​the LED chip to the external heat dissipation interface (such as the lower surface of the ceramic substrate, the heat-conducting part of the metal bracket, or the heat-conducting area of ​​the lead frame); the continuous detection unit is a layered region divided along the preset heat transfer direction with a fixed step size or adaptive thickness, used to support subsequent projection analysis and connectivity statistics of voids on the heat flow path; the void position is the coordinate value of the geometric center of the microscale void in the three-dimensional coordinate system; the equivalent aperture is the void volume... The diameter obtained when the product is equivalent to a sphere is used to characterize the thermal shielding capability of the gap; the aspect ratio is the ratio of the maximum projected length to the minimum projected length of the gap, used to distinguish between approximately spherical gaps (aspect ratio ≈ 1~2) and slender, slit-like gaps (aspect ratio > 3); the distance between adjacent gaps is the Euclidean distance between the geometric centers of the gaps, used to determine whether the gaps may form a connected thermal barrier chain; the direction of the gap principal axis is the unit vector corresponding to the direction of the maximum moment of inertia of the gap, used to quantify its spatial orientation relationship with the main heat flow direction; the gap topology distribution map is a visualization map that integrates at least one of the above gap topology parameters and maps it to the spatial coordinates of each detection unit. Its formation includes: statistically analyzing the degree of projection overlap of the gaps in each detection unit in the preset heat transfer direction, and combining the distance between adjacent gaps to form a distribution map characterizing the gap aggregation state and connectivity trend.

[0057] This application can, for example, identify gap boundaries based on regions of abrupt grayscale gradient changes in X-ray transmission images and obtain gap spatial distribution information through a 3D reconstruction algorithm; it can also, for example, invert the equivalent aperture and aspect ratio of gaps based on the time delay and amplitude attenuation characteristics of reflected signals from scanning acoustic microscopy (SAM) combined with a gap size-acoustic impedance model; further, it can also invert the gap location and principal axis direction based on anomalous regions of heat wave propagation during transient heating in infrared thermography combined with the heat diffusion equation. This application obtains a gap topology distribution map based on any of the above methods to support subsequent delineation of the main heat diffusion region and identification of key gap clusters.

[0058] Step 2: Based on the heat transfer relationship between the main heat-generating area of ​​the LED chip and the external heat dissipation interface, determine the main heat diffusion area inside the LED chip package. The main heat diffusion area includes at least two of the following: the main heat diffusion core area, the axial heat conduction area, and the lateral transition heat conduction area.

[0059] The main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the package thickness direction, which is used to characterize the initial heat flow convergence area radiated from the chip heat source into the package interior; the axial heat conduction area is the heat conduction area formed by the main heat diffusion core area extending towards the external heat dissipation interface, and its extension path is along the preset heat transfer direction, which is used to carry the chip heat to the substrate or bracket axially; the lateral transition heat conduction area is the heat flow turning area formed by the edge of the LED chip extending towards the bracket or substrate, which is used to realize the turning conduction of heat flow from the chip plane direction to the vertical direction or oblique heat dissipation interface; the three together constitute a three-dimensional spatial governance domain around the main heat diffusion path, the boundary of which is jointly determined by the chip geometric center, pn junction heat source distribution, high heat area near the electrode, driving current concentration area, heat dissipation outlet position and package structure level.

[0060] This application, for example, can define the main heat diffusion core region as a columnar region within ±0.15mm of the chip center based on the peak position and size of the infrared thermal image of the PN junction region of the chip, combined with the finite element thermal simulation results of the packaging structure. Alternatively, it can define a fan-shaped region extending 0.2mm to 0.4mm outward from the outer edge of the chip as a lateral transition heat conduction region based on the geometric path and heat flux density gradient field from the chip edge to the support connection point. Furthermore, it can define the heat flow path envelope region, where the cumulative thermal resistance contribution is over 70%, as the axial heat conduction region based on the thermal resistance contribution ratio of each path in the thermal resistance network model. This application obtains the main heat diffusion region based on any of the above methods, which is used to define the spatial identification range and reshaping operation area of ​​subsequent key void clusters.

[0061] Step 3: Based on the void topology distribution map and the main heat diffusion zone, cluster identification is performed on multiple voids located in the same detection unit and arranged sequentially along the main heat flow direction. When multiple voids meet at least two of the following conditions, they are identified as key void clusters: the total projected length in the main heat flow direction is greater than a preset multiple of the average aperture of a single void; the spacing between adjacent voids is less than a preset connection threshold; the angle between the void's main axis direction and the main heat flow direction is less than a preset angle threshold; and the voids are located in the main heat diffusion core area or the lateral transition heat conduction zone.

[0062] The total projected length in the main heat flow direction can refer to the total coverage length of the projected line segments of multiple gaps in the preset heat transfer direction, which is used to quantify their cumulative shading effect on the main heat flow cross section. The preset multiple of the average aperture of a single gap is 2 to 10 times, which is used to exclude interference from isolated small gaps. The distance between adjacent gaps refers to the projected distance between the geometric centers of the two gaps in the main heat flow direction. When it is less than the preset connection threshold (15μm to 50μm), it indicates that the two gaps have a physical connection tendency in the heat flow path. The angle between the main axis direction of the gap and the main heat flow direction reflects the orientation blocking strength of the gap on the heat flow. The smaller the angle, the more significant the heat flow interception along the main heat flow direction. The preset angle threshold is 5° to 30°. Being located in the main heat diffusion core area or the lateral transition heat conduction area is a condition for determining the space of the gap, ensuring that the identified object is in the actual dominant thermal resistance contribution area.

[0063] This application can, for example, filter out spatially overlapping void sets by performing Boolean intersection operations between the coordinates of each void in the void topology distribution map and the three-dimensional mesh of the main heat diffusion zone; it can also sort the filtered voids according to their projected coordinates in the main heat flow direction, identify continuous arrangement sequences using the sliding window method, and calculate the ratio of their total projected length to the average aperture; furthermore, this application can also perform connectivity weighted determination on void sequences that meet the angle condition based on the comparison results of the cosine value of the void principal axis direction and the included angle threshold. This application obtains key void clusters based on any of the above methods, which are used to drive the generation of differentiated topology reshaping strategies.

[0064] Step 4: Classify the key void clusters according to their heat-insulating mechanisms, including volume-dominant void clusters, orientation-dominant void clusters, and connectivity-dominant void clusters;

[0065] Among them, the volume-dominant type void cluster is a key void cluster mainly composed of voids with enlarged apertures. Its heat resistance mechanism mainly stems from the heat capacity shielding and local heat flow cross-sectional area loss caused by the large cavity volume. The direction-dominant type void cluster is a key void cluster mainly composed of slender slit-type voids, and the angle between the void's main axis and the main heat flow direction is less than a preset angle threshold. Its heat resistance mechanism mainly stems from the fact that the long axis of the void is highly consistent with the heat flow direction, forming a low-resistance heat flow channel interruption. The connectivity-dominant type void cluster is a key void cluster formed by two or more adjacent voids connected in series along the main heat flow direction. Its heat resistance mechanism mainly stems from the fact that multiple voids form a continuous blocking zone on the heat flow path, inducing the heat flow to detour over long distances.

[0066] This application can, for example, classify voids as volume-dominant based on the histogram of aspect ratio distribution of each void in the key void cluster. If the peak value is in the range of 1 to 2 and the volume fraction is >60%, then the voids are classified as volume-dominant. Alternatively, it can be classified as direction-dominant based on the consistency index of the void principal axis direction (such as the von Mises distribution concentration). If the concentration is >0.8 and the average angle is <10°, then the voids are classified as direction-dominant. Furthermore, this application can be classified as connectivity-dominant based on the projection connectivity of the void sequence in the main heat flow direction (i.e., the ratio of projection line segment overlap to void size). If the connectivity is >0.7 and the number of voids is ≥3, then the voids are classified as connectivity-dominant. This application completes the type classification based on any of the above methods, providing a physical basis for the aforementioned matching reshaping.

[0067] Step 5: Perform matching void topology reshaping processing for different types of critical void clusters. Specifically, for volume-dominant void clusters, perform local filling and / or local compression; for orientation-dominant void clusters, perform directional extrusion, local hot pressing, and / or void orientation deflection processing; and for connectivity-dominant void clusters, perform thermal bridging and / or connectivity chain disconnection processing.

[0068] Local filling involves introducing thermally conductive filler material into the region where the volume-dominant void cluster is located, allowing it to penetrate along the capillary gaps between the void and the surrounding encapsulation medium to reduce the cavity volume and improve local thermal conductivity. Local compression involves applying local pressure intersecting the main heat flow direction to cause the incompletely cured encapsulation medium to flow in a controlled manner, compressing the void volume and changing its aspect ratio. Directional extrusion involves applying pressure perpendicular to the main heat flow direction to the directional-dominant void cluster, forcing the slender voids to deform or split, causing their main axis direction to deviate from the main heat flow direction. Local thermo-pressing involves simultaneously applying temperature and pressure to the directional-dominant void cluster region in a semi-cured state, utilizing the material's viscoelastic response to achieve void orientation deflection. Thermal bridging involves constructing thermally conductive bridges extending in a direction intersecting the main heat flow direction between adjacent voids connecting the dominant void cluster, allowing heat flow to cross the voids for conduction. Disconnection of the connecting chain involves using local thermo-pressing to reclose the encapsulation medium between adjacent voids, or injecting a highly fluid thermally conductive medium to fill the gaps, thereby severing the series connection relationship along the main heat flow direction.

[0069] This application can, for example, involve injecting a low-viscosity thermally conductive adhesive containing alumina filler into the volume-dominant void cluster region, and then applying a vacuum negative pressure of 0.05 MPa to promote penetration; another application can involve applying a lateral pressure of 0.2 MPa to the direction-dominant void cluster region and maintaining it for 60 seconds, causing its main axis direction to deflect to an angle >45° with the main heat flow; further, this application can involve depositing a conductive and thermally conductive slurry containing graphene filler in the regions connecting the two sides of the dominant void cluster, and forming a 50 μm wide and 8 μm thick trans-void thermally conductive bridging band through local hot pressing. This application achieves void topology reshaping based on any of the above methods, thereby specifically weakening the dominant thermal resistance mechanism of the key void clusters.

[0070] Step 6: After completing the void topology reshaping process, a continuous thermal conduction structure is constructed sequentially along the heat flow path in the main heat diffusion region. The continuous thermal conduction structure includes at least two of the following: a low void thermal conduction region directly below the chip, a continuous thermal conduction section extending along the package thickness direction, and a heat flow turning bridging section.

[0071] The low-gap thermal conductivity zone directly below the chip is located directly below the main heat diffusion core area. Through the reshaping process of the volume-dominant or direction-dominant gap clusters, the gap volume fraction is reduced to less than 0.3%, forming the first-level low thermal resistance thermal conductivity interface between the chip and the packaging medium. The continuous thermal conductivity section extending along the packaging thickness direction is located within the axial thermal conductivity zone. Through the disconnection process of the interconnected dominant gap clusters and thermal bridging, it forms a through path without significant gap obstruction in the main heat flow direction, with a heat flow cross-section continuity >95%. The heat flow deflection bridging section is located within the lateral transition thermal conductivity zone. By constructing a thermally enriched zone or a locally densified structure in the chip edge and bracket connection area, the heat flow is smoothly deflected and diverted from the chip plane direction to the vertical / oblique heat dissipation interface.

[0072] This application can, for example, form a low-void thermally conductive region with a thermal conductivity >12 W / (m·K) within a 0.1 mm range directly below the chip by combining localized adhesive application with thermoforming. Alternatively, it can construct a continuous thermally conductive segment extending along the main heat flow direction by in-situ growing a boron nitride nanosheet array around the void clusters in the axial thermally conductive region. Furthermore, this application can deposit a thermally conductive bridging paste containing silver nanowires within a 0.15 mm range at the chip edge of the lateral transition thermally conductive region using inkjet printing, forming a heat flow redirection bridging segment at an angle of 30°–60° to the main heat flow direction. This application constructs a continuous thermally conductive structure based on any of the above methods, establishing a main heat conduction chain from the chip to the external heat dissipation interface.

[0073] Step 7: Perform partitioned curing on the continuous thermally conductive structure. First, the main heat diffusion area is cured in stages, and then the non-main heat area is cured as a whole. The non-main heat area retains a controlled micropore buffer zone to obtain a high heat dissipation packaging structure for LED chips.

[0074] The main heat diffusion zone is cured in stages, including: after completing the void topology reshaping and continuous thermal conduction structure construction, the first stage of curing (temperature 60℃~100℃, time 5min~30min) is performed on the main heat diffusion core area and the axial thermal conduction area to initially lock the thermal conduction structure morphology; then, after the heat flow turning bridge section is constructed, the second stage of curing (temperature 100℃~160℃, time 20min~120min) is performed on the lateral transition thermal conduction area to enhance the structural stability of the turning area; the overall curing of the non-main heat diffusion area can refer to the final curing of the packaging area outside the main heat diffusion area (such as the chip peripheral encapsulation glue, the non-main heat connection area at the edge of the bracket) under a low heating rate and constant pressure, so that it retains a controlled micropore buffer with a volume fraction of 0.1%~8% to absorb the interfacial stress caused by the difference in the thermal expansion coefficient of the material during the thermal cycle.

[0075] This application can, for example, employ a stepped heating curve: a first stage at 60℃ for 15 minutes, a second stage at 120℃ for 45 minutes, and a third stage at 160℃ for 20 minutes, wherein the first and second stages only affect the main heat diffusion zone, while the third stage covers the entire encapsulation. Alternatively, this application can utilize coordinated control of a local heating module and a global heating platform to ensure that the temperature of the main heat diffusion zone is consistently 20℃–40℃ higher than that of the non-main heat diffusion zone. Furthermore, this application can introduce a controllable nitrogen atmosphere during the curing stage of the non-main heat diffusion zone to suppress the crosslinking reaction rate, thereby stabilizing the microporous structure. This application achieves zoned curing based on any of the above methods, balancing the stability of the thermally conductive structure in the main heat diffusion zone with the stress buffer reliability of the non-main heat diffusion zone.

[0076] Example 2, in an optional embodiment, this application further provides that the void detection in step S1 employs at least one of X-ray detection, industrial CT detection, scanning acoustic microscopy detection, infrared thermographic-assisted inversion detection, and section microscopy detection; the formation of the void topology distribution map includes:

[0077] Step 1: The void detection in step S1 uses at least one of the following: X-ray detection, industrial CT detection, scanning acoustic microscopy detection, infrared thermography-assisted inversion detection, and section microscopy detection.

[0078] Among them, X-ray detection can refer to the use of X-rays to penetrate the package and form a grayscale image on the imaging plate or detector, and to identify the spatial location and equivalent size of the internal pores through density differences. It is suitable for detecting closed spherical pores with a depth in the range of 50μm to 500μm.

[0079] Industrial CT inspection refers to the technology of reconstructing three-dimensional voxel models based on multi-angle X-ray projection. It can output the three-dimensional coordinates, volume, surface curvature and spatial connectivity of voids, and is suitable for full-scale reconstruction of micro-scale voids at the interface between die-bonding layers, encapsulation layers and substrates.

[0080] Scanning acoustic microscopy (SAMS) is a technique that uses high-frequency ultrasonic waves to image the reflection / scattering characteristics of material interfaces. It has high sensitivity to debonding cracks, micropore clusters, and directional voids at the layered interfaces such as chip / die bond layer and die bond layer / scaffold. It is especially suitable for detecting slender thermally insulating cracks with an aspect ratio greater than 3 and a thickness of less than 20 μm.

[0081] Infrared thermal imaging-assisted inversion detection refers to capturing the temperature response curve of the package surface using a high-speed infrared thermal imager after applying a transient pulse current to an LED chip, and then using a heat conduction inversion model to estimate the location of internal voids and the area contributing to thermal resistance. It is suitable for identifying key void clusters that dynamically block the main heat flow path.

[0082] Slice microscopy inspection refers to the process of precisely grinding or cutting the package along a preset cross section, and then observing the morphology, size and adjacency relationship of the voids under an optical microscope. This provides submicron-level resolution verification data, which is used to calibrate the aforementioned non-destructive testing results and support the calibration of void topology parameters.

[0083] This application can, for example, obtain a two-dimensional projection image of the voids in the area directly beneath the chip using X-ray detection, and supplement the identification of crack-type voids in the edge heat flow transition zone by combining the SAM detection results; this application can also, for example, reconstruct the three-dimensional void distribution of the package using industrial CT detection, and locate and label the abnormal thermal response voids in the main heat diffusion core area using infrared thermography inversion results; further, this application can also use slice microscopy to perform cross-sectional verification on the suspected connected thermal resistance chains identified by CT, confirming whether they truly constitute a continuous void arrangement extending along the main heat flow direction. This application obtains the void spatial distribution information required in step S1 based on any of the above methods, ensuring that the detection covers all key thermal resistance levels from the main heat-generating area of ​​the LED chip to the external heat dissipation interface, avoiding missed detection of bulb-type voids or misjudgment of crack-type voids due to a single detection method.

[0084] Step 2: The formation of the void topology distribution map includes: statistically analyzing the degree of overlap of the projections of voids in each detection unit in the preset heat transfer direction, and combining the spacing between adjacent voids to form a distribution map characterizing the void aggregation state and connectivity trend;

[0085] Among them, the degree of projection overlap can refer to the area overlap rate or length overlap integral value of the projection contour of all gaps in each detection unit along the preset heat transfer direction (i.e., from the main heat-generating area of ​​the LED chip to the external heat dissipation interface) after orthogonal projection. This degree of overlap reflects the cumulative effect of the gaps blocking the heat flow cross section. The higher the overlap rate, the more gaps are arranged in layers in the main heat flow direction, forming a superimposed blockage of the heat flow.

[0086] The spacing between adjacent gaps can refer to the shortest Euclidean distance between the geometric centers of any two gaps within the same detection unit. After normalization, it is compared with a preset connection threshold to determine whether there is physical connection or heat flow bypass induction potential between gaps. When the spacing is less than the preset connection threshold, it indicates that the two gaps are within the range of heat flow disturbance interaction and may jointly constitute a local thermal resistance abrupt change point.

[0087] The void clustering state can refer to the spatial clustering density of voids in the same detection unit along the projection direction, characterized by the number of voids per unit projection length or the gradient change rate of the projection overlap integral value, and is used to identify dense void zones on the main heat flow channel.

[0088] Connectivity trend can refer to the spatial continuity characteristic of multiple gaps whose adjacent spacing is consistently lower than a preset connection threshold on the basis of projection overlap. It is used to predict whether a connected heat-resistant chain extending along the main heat flow direction is likely to be formed.

[0089] This application, for example, can use statistical results of projection overlap to indicate the occlusion intensity level of different detection units in the gap topology distribution map with varying color shades, where darker areas correspond to high-occlusion units with a projection overlap rate ≥15%. Alternatively, it can mark gap pairs with a spacing of less than 40 μm in the gap topology distribution map by connecting them with lines based on the distribution of adjacent gap spacing, and extend them with dashed lines to indicate the direction of potential connectivity chains. Furthermore, this application can jointly model the projection overlap degree and adjacent spacing to generate a two-dimensional heat flow influence index matrix, where each element is a weighted product of the projection overlap rate and the reciprocal of the spacing, used to quantify the comprehensive interference degree of each detection unit on the main heat flow path. Based on any of the above methods, this application obtains a distribution map characterizing the gap aggregation state and connectivity trend, enabling the gap topology distribution map to not only present static spatial information but also to predict and support the determination of heat flow bypass paths, local hotspot locations, and S3 key gap clusters.

[0090] In Example 3, another optional embodiment, this application further provides that in step S2, the main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the package thickness direction; the axial heat conduction area is the heat conduction area formed by extending from the main heat diffusion core area toward the external heat dissipation interface; and the lateral transition heat conduction area is the heat flow turning area formed by extending from the edge of the LED chip toward the bracket or substrate, specifically:

[0091] Step 1: The main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the direction of package thickness;

[0092] Among them, the main heat diffusion core area can refer to the two-dimensional area covered by the orthogonal projection made in the direction of the package thickness perpendicular to the chip surface, with the main heat-generating area of ​​the LED chip as the reference.

[0093] This region can be the geometric projection of the LED chip pn junction light-emitting region, the heat source projection of the high current density region adjacent to the chip electrode, the equivalent heat source projection of the driving current concentration region, or the union projection of multiple heat source regions mentioned above.

[0094] In this embodiment, the main heat diffusion core area is used to anchor the starting output surface of the main heat flow. Its boundary determination does not rely on empirical estimation, but is based on spatial mapping of the actual heat source distribution of the chip and the hierarchical relationship of the packaging structure, thereby ensuring the consistency of the spatial reference for subsequent gap topology identification and key gap cluster determination.

[0095] Step 2: The axial heat conduction zone is the heat conduction area formed by extending from the main heat diffusion core area toward the external heat dissipation interface;

[0096] Among them, the axial heat conduction zone can refer to the three-dimensional heat conduction channel area formed by extending unidirectionally from the boundary of the main heat diffusion core area along the preset heat transfer direction (i.e., the encapsulation thickness direction is the main axis direction) to the external heat dissipation interface.

[0097] This region can be columnar, conical, or truncated conical in shape, and its extension length depends on the total thickness of the package, the thickness of the die-attach layer, the thickness of the support / substrate, and the location of the external heat dissipation interface.

[0098] In this embodiment, the axial heat-guiding zone is used to characterize the main path of the main heat flow in the thickness direction of the package. Its existence gives the clustering identification of multiple gaps arranged sequentially along the main heat flow direction in S3 a clear spatial constraint, avoiding misjudging non-axially distributed gaps as key gap clusters.

[0099] Step 3: The lateral transition heat conduction zone is the heat flow transition area formed by the extension of the LED chip edge toward the bracket or substrate;

[0100] The lateral transition heat conduction area can refer to the heat flow turning and redistribution area that extends radially or fan-shaped from the edge segment of the LED chip to the adjacent bracket pins, the heat conduction part of the metal bracket, the edge of the ceramic substrate, or the heat conduction area of ​​the lead frame.

[0101] This area may include the interface between the chip edge and the support, the junction between the chip side and the encapsulant, the contact zone between the encapsulant and the substrate sidewall, and the interface convergence area during the process of heat flow changing from the vertical direction to the horizontal direction.

[0102] In this embodiment, the lateral transition thermal conduction zone is used to cover the key thermal resistance link where the direction of heat flow changes at the edge of the chip. Its explicit definition supports the location determination condition of the key gap cluster in the main heat diffusion core area or the lateral transition thermal conduction zone in Embodiment 4, and also directly constitutes the basic area for constructing the heat flow turning bridge section in Embodiment 9.

[0103] This application may, for example, determine the chip heat source output reference plane based on the projection mapping relationship of the main heat diffusion core region; this application may also, for example, determine the longitudinal conduction backbone of the main heat flow based on the extension path of the axial heat conduction zone; further, this application may also determine the distribution of the heat flow transition interface based on the geometric radiation range of the lateral transition heat conduction zone. This application obtains the three-dimensional spatial definition of the main heat diffusion region based on any of the above methods, so that the delineation of the main heat diffusion region in step S2 has structural reproducibility, heat flow physics interpretability, and cross-package adaptability.

[0104] In Example 4, another embodiment, this application further provides that in step S3, when clustering and identifying multiple gaps located in the same detection unit and arranged sequentially along the main heat flow direction, multiple gaps that meet at least two of the following conditions are identified as key gap clusters, including:

[0105] Step 1: The total projected length in the main heat flow direction is greater than 2 to 10 times the average aperture of a single pore;

[0106] The total projection length can refer to the maximum continuous projection length of the projection contours of multiple gaps in the preset heat transfer direction (i.e., the main heat flow direction) after being superimposed along that direction; the average aperture of a single gap can refer to the arithmetic mean of the equivalent apertures of all identified gaps in the same detection unit.

[0107] The total projected length is used to characterize the cumulative shading effect of multiple gaps on the heat flow cross section in the main heat flow direction. When the total projected length is less than twice the average aperture of a single gap, it indicates that the physical shading of the gap set on the main heat flow path is weak and insufficient to constitute a significant increase in thermal resistance. When the total projected length is greater than 10 times the average aperture of a single gap, it may contain gaps that are discontinuously distributed or non-collinearly arranged, and their connectivity and thermal resistance synergy are reduced, which can easily lead to misjudgment. Therefore, the numerical range of 2 to 10 times is limited to ensure the effective capture of the real connected thermal resistance chain and avoid including the gaps that are scattered, isolated or on non-main heat paths in the key cluster identification category.

[0108] This application can, for example, calculate the continuous projection length along the main heat flow direction of the void using the maximum envelope method based on the two-dimensional / three-dimensional projection coordinates of the void in the main heat flow direction; it can also construct the void projection interval based on the coordinates of the void center point and the principal axis direction vector, and merge the intervals to obtain the total projection length; furthermore, this application can also obtain the total projection length by using pixel-level projection statistics based on the preset heat transfer direction in the void topology distribution map, and then using grayscale accumulation and threshold segmentation. This application obtains a quantitative evaluation result of the heat flow obstruction degree of the void cluster based on any of the above methods, so as to support the reliable identification of key void clusters.

[0109] Step 2: The spacing between adjacent gaps is less than the preset connection threshold;

[0110] The spacing between adjacent gaps can be the sum of the projected distance between the geometric centers of the two gaps in the main heat flow direction and the projected half length of each gap in that direction; the preset connection threshold is 10μm~80μm, and can be selected as 15μm~50μm;

[0111] This spacing is used to determine whether multiple gaps have the spatial prerequisite to form a connected thermal resistance chain. When the spacing is too large, there is no thermal flow coupling effect between the gaps, and the heat flow can bypass through the gap area. When the spacing is too small, the encapsulation medium still has the ability to flow and rearrange in the pre-cured state, and the gaps are prone to merging or bridging, which weakens their independent thermal resistance characteristics. Therefore, the preset connection threshold is set to an engineering-adjustable range to ensure that the identification results are consistent with the actual heat conduction behavior.

[0112] This application can, for example, calculate the minimum spacing between two gaps in the main heat flow direction based on the gap center coordinates and the principal axis direction; it can also extract the Euclidean distance of the nearest neighbor gap projection boundary in the main heat flow direction based on the edge distance function of the gap projection profile; further, it can extract the skeleton distance as a spacing characterization based on the connected component analysis results of adjacent gap pixel clusters in the gap topology distribution map. This application obtains the criteria for determining the spatial proximity between adjacent gaps based on any of the above methods to support the structural identification of connected heat-resistant chains.

[0113] Step 3: The angle between the main axis direction of the void and the main heat flow direction is less than the preset angle threshold;

[0114] Among them, the main axis direction of the gap can refer to the major axis direction of the gap geometry, which is obtained by fitting an ellipse to the gap profile or by principal component analysis; the main heat flow direction is the main heat transfer vector from the main heat-generating area of ​​the LED chip to the external heat dissipation interface; the preset angle threshold is 10°~45°, and can be selected as 15°~30°;

[0115] The included angle is used to measure the directional blocking strength of the gap on the main heat flow path. When the included angle is close to 0°, the gap cuts off the heat flow in a waist-like manner, and the blocking efficiency is the highest. When the included angle increases to above the critical value, the gap is more of a lateral deflection rather than a direct blockage, and its direct thermal resistance contribution to the main heat flow decreases significantly. Therefore, setting an angle threshold can effectively eliminate a large number of gaps that are only directionally related but have no substantial thermal resistance effect.

[0116] This application can, for example, extract the first principal component direction as the principal axis direction based on the eigenvector of the covariance matrix of the gap contour point cloud, and calculate the angle between it and the unit vector of the main heat flow direction; this application can also, for example, detect the linear principal axis based on the Hough transform of the gap projection image, and then calculate its angle with the main heat flow direction; further, this application can also determine the principal axis direction by using the peak interval of the gap direction histogram in the gap topology distribution map, combined with the weighted average of direction consistency. This application obtains the matching degree between the gap orientation and the heat flow path based on any of the above methods to support the identification basis of the direction-dominated heat resistance effect.

[0117] Step 4: Multiple gaps are located in the main heat diffusion core region or the lateral transition heat conduction region;

[0118] Among them, the main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the package thickness direction; the lateral transition heat conduction area is the heat flow turning area formed by the extension of the LED chip edge toward the bracket or substrate;

[0119] This location condition is used to constrain the spatial assignment of key gap clusters, ensuring that identification focuses on two highly sensitive areas: the heat source output end (core area) and the key node for heat flow path reconstruction (lateral area). The core area corresponds to the low thermal resistance main channel directly below the chip, and the gaps in this area directly cut off the main heat flow. The lateral area corresponds to the stress concentration and path abrupt change area where the heat flow changes from vertical to horizontal. The gaps in this area are prone to inducing heat flow detours and local hot spots. Together, they constitute the most prominent physical location of the heat diffusion bottleneck. Limiting it to this location can significantly improve identification accuracy and process response efficiency.

[0120] Example 5, in one embodiment, this application also provides the classification of key void clusters according to their heat-insulating mechanisms in step S4, specifically including:

[0121] Step 1: Classify the key void clusters dominated by aperture enlargement voids into volume-dominant void clusters;

[0122] Among them, the enlarged aperture void can refer to a microscale void in which the equivalent aperture D in the detection unit is 1.5 to 3 times larger than the average equivalent aperture of all voids in the unit, and the aspect ratio AR is ≤2.0.

[0123] The role of this technical feature in this embodiment is to reflect the direct shielding ability of the gap on the main heat flow cross section and the local thermal capacity disturbance effect. When the gap volume increases, its projected area in the main heat flow direction increases significantly, resulting in a decrease in the effective heat conduction cross section area per unit heat flow path. At the same time, the thermal capacity space occupied by the low thermal conductivity medium (air or residual gas) inside the gap increases, weakening the local thermal response speed.

[0124] This application identifies pore enlargement-type pores, for example, based on the ratio of the equivalent pore diameter D to the average pore diameter of the detection unit, combined with the morphological constraint of aspect ratio AR ≤ 2.0. It also determines whether a pore belongs to the pore enlargement-type pore category by checking whether the pore's projected shading rate η in the main heat flow direction exceeds a preset shading rate threshold (10%) and simultaneously verifying whether its aspect ratio meets AR ≤ 2.0. Furthermore, this application identifies key pore clusters composed mainly of large-volume pores based on the ratio of the pore cluster volume fraction to the average volume of a single pore. This application achieves accurate identification of volume-dominant pore clusters based on any of the above methods, supporting the targeting and necessity of subsequent local filling and compression processing.

[0125] Step 2: Classify the key void clusters that are mainly composed of slender, crack-type voids and whose angle between the void principal axis direction and the main heat flow direction is less than a preset angle threshold as direction-dominant void clusters;

[0126] Among them, slender crack-type voids can refer to microscale voids with an aspect ratio AR of 3 to 15, a clear principal axis direction, and an equivalent pore size D of 5 μm to 60 μm;

[0127] The angle between the void principal axis direction and the main heat flow direction is calculated by extracting the void contour ellipse using an image processing algorithm, fitting the principal axis, and then calculating the minimum angle θ between the principal axis vector and the main heat diffusion direction vector.

[0128] The role of this technical feature in this embodiment is to characterize the directional induction and bypass amplification effect of the gap on the heat flow path. When the main axis direction of the slender crack is highly consistent with the main heat flow direction (θ < preset angle threshold, 15°), the heat flow is forced to bypass significantly along its long axis direction, resulting in an extension of the equivalent heat conduction path and an increase in local thermal resistance.

[0129] This application identifies slender, fracture-type voids based on the dual criteria of an aspect ratio (AR) greater than 3 and an included angle (θ) less than 15°. It also determines directional dominance based on the ratio of the void's projected length in the main heat flow direction to its projected length in the perpendicular direction being greater than a preset ratio (5:1), combined with an included angle (θ) check. Furthermore, this application identifies directionally dominant void clusters based on the matching degree between the peak interval width in the void's principal axis distribution histogram and the deviation range of the main heat flow direction. This application achieves accurate identification of directionally dominant void clusters using any of the above methods, ensuring that subsequent directional extrusion, hot pressing, or orientation deflection processes have a clear physical basis and direction of action.

[0130] Step 3: Divide the critical void clusters formed by two or more adjacent voids connected in series along the main heat flow direction into connected dominant void clusters;

[0131] Among them, two or more adjacent gaps are connected in series along the main heat flow direction. In the main heat flow direction, the center coordinate projections of multiple gaps are approximately collinear, and the nearest neighbor distance L between any two adjacent gaps in this direction is less than the preset connection threshold (30μm). At the same time, the gaps are not completely separated by a continuous heat-conducting medium.

[0132] The role of this technical feature in this embodiment is to characterize the structural cutting effect of the void set on the main heat flow path. When multiple voids are arranged in a tightly spaced linear pattern along the main heat flow direction, they together form a heat-resistant chain that the heat flow cannot penetrate, forcing the heat flow to detour to non-main heat areas, causing local hot spots and a jump in thermal resistance.

[0133] This application, for example, identifies subsequences with consecutive projection point spacing less than 30 μm based on the projected coordinate sequence of the void center point in the main heat flow direction using the sliding window method, and counts whether the number of voids contained in each subsequence is ≥2. This application, for example, determines whether a void constitutes a series structure along the main heat flow direction based on the combination of void connectivity C > 0.6 and projection overlap rate > 40%. Furthermore, this application, for example, uses the ratio of the total projected length of the void cluster in the main heat flow direction to the average single void diameter to assist in verifying its connectivity dominance based on whether it is ≥5. This application obtains accurate identification of connectivity-dominant void clusters based on any of the above methods, ensuring the timing and spatial positioning accuracy of subsequent thermal bridging or connectivity chain disconnection processing.

[0134] Example 6, in an optional embodiment, this application also provides a void topology reshaping process in step S5 for volume-dominant void clusters, including:

[0135] Step 1: Introduce thermally conductive filler material into the area where the volume-dominant void cluster is located, allowing the thermally conductive filler material to penetrate along the gap between the void and the surrounding encapsulation medium, and apply local pressure to the area after the thermally conductive filler material has penetrated, so as to reduce the cavity volume of the volume-dominant void cluster.

[0136] Among them, volume-dominant void clusters can refer to key void clusters dominated by pore size-enlarged voids, whose equivalent pore size is significantly larger than that of adjacent voids, and whose projected shading rate in the main heat flow direction is higher than a preset threshold (e.g., 10%–25%), forming the main thermal resistance area on the heat flow cross section; the thermally conductive filler material can be a low-viscosity thermally conductive adhesive, a capillary-penetrating thermally conductive slurry, or a secondary wetting thermally conductive medium, with an apparent viscosity of 0.1 Pa·s to 5 Pa·s, a contact angle of less than 45°, and the ability to spontaneously penetrate along the micron-level gap at the interface of the incompletely cured encapsulation medium; penetration along the gap between the void and the surrounding encapsulation medium can guide the thermally conductive filler material not to be directly injected into the void, but to enter the surrounding area of ​​the void through the capillary channels formed by the void edge and the adjacent encapsulation medium, in the void The cavity wall forms a continuous coating layer or interface anchoring structure, thereby synergistically compressing the cavity and enhancing the interface bonding strength during subsequent pressurization. The local pressure can be mechanical pressure film, micro-area pneumatic pressure head, vacuum-assisted back pressure, or thermo-pressure composite pressure. Its effective area covers the entire range of the detection unit where the volume-dominant cavity cluster is located, with a pressure amplitude of 0.05MPa to 0.4MPa. The direction of application of the local pressure intersects with the main heat flow direction and can be set vertically, obliquely (with an angle of 30° to 75°), or segmented with different directions. The technical purpose is to apply a shear component to the cavity wall while compressing the cavity volume, so as to promote the directional spreading and micro-intercalation of the thermally conductive filling material at the cavity-medium interface, and avoid colloidal extrusion, interface peeling, or stress concentration caused by pure axial pressure.

[0137] This application can achieve the infiltration of thermally conductive filler materials along gaps through, for example, the synergistic effect of capillary penetration and interfacial wetting; it can also achieve the infiltration of thermally conductive filler materials along gaps through vacuum-assisted negative pressure driving and surface energy gradient guidance; furthermore, it can also achieve the infiltration of thermally conductive filler materials along gaps through temperature difference-induced viscosity gradient and interfacial tension regulation. Based on any of the above methods, this application achieves targeted filling and coverage of volume-dominant void clusters and interfacial anchoring effects, providing a structural basis for subsequent local pressure application.

[0138] Step 2: The direction of the applied local pressure is set to intersect with the direction of the main heat flow;

[0139] The main heat flow direction refers to the direction of the path with the highest heat flow density between the main heat-generating area of ​​the LED chip and the external heat dissipation interface. It is usually vertical in the package thickness direction, and inclined or curved in the lateral transition heat conduction area. The intersecting arrangement can be orthogonal (angle 90°), acute angle (angle 30° to 85°), or obtuse angle (angle 95° to 150°), but does not include parallel arrangement (angle 0° or 180°). The technical function of this intersecting relationship is to reduce the cavity volume, introduce shear stress components, suppress the axial overflow of the thermally conductive filling material during the pressurization process, and enhance its adhesion stability on the cavity wall, so as to prevent the thermal resistance from rising due to interface debonding during subsequent thermal cycling.

[0140] This application can achieve synergistic cavity volume compression and interface anchoring by applying local pressure perpendicular to the main heat flow direction; it can also achieve synergistic cavity volume compression and interface anchoring by applying oblique local pressure at a 45° angle to the main heat flow direction; further, it can achieve synergistic cavity volume compression and interface anchoring by segmented pressure direction switching (first applying pressure vertically and then holding pressure obliquely). Based on any of the above methods, this application achieves a structurally stable compression effect on volume-dominant void clusters, ensuring their long-term thermal conductivity under thermal load.

[0141] Step 3: The thermally conductive filling material includes at least one of the following: silver filler, alumina filler, boron nitride filler, aluminum nitride filler, graphene filler, and carbon nanotube filler;

[0142] Among them, silver filler can refer to spherical or plate-shaped silver particles with a particle size of 0.1μm to 5μm, which have high thermal conductivity (≥400W / (m·K)) and good metal interface wettability, and are suitable for die-bonding layer scenarios that require both electrical conductivity and thermal conductivity; alumina filler can refer to α-Al2O3 phase ceramic particles, which have a thermal conductivity of 20W / (m·K) to 35W / (m·K), good insulation, low cost, and high chemical stability, and are suitable for general-purpose thermally conductive adhesive systems; boron nitride filler can refer to hexagonal boron nitride (h-BN) plate-shaped filler, which has an in-plane thermal conductivity of up to 300W / (m·K) and a vertical thermal conductivity of about 30W / (m·K), and has both high insulation and good thermal conductivity. Anisotropic thermal conductivity makes it suitable for scenarios requiring suppression of lateral thermal crosstalk; aluminum nitride filler can refer to AlN ceramic particles with a thermal conductivity of 150 W / (m·K) to 200 W / (m·K), and its coefficient of thermal expansion matches well with silicon chips, making it suitable for high-reliability packaging; graphene filler can refer to single-layer or few-layer graphene sheets with a theoretical in-plane thermal conductivity exceeding 5000 W / (m·K), which can form a two-dimensional thermally conductive network and is suitable for constructing cross-pore bridging structures; carbon nanotube filler can refer to multi-walled carbon nanotubes with a length of 1 μm to 20 μm, which have a high aspect ratio and axial thermal conductivity (>3000 W / (m·K)) and are suitable for constructing directional thermally conductive pathways in the pores.

[0143] This application may employ a combination of silver and alumina fillers to balance high thermal conductivity and insulation requirements; alternatively, it may use a combination of boron nitride and graphene sheet fillers to construct a heterogeneous laminated structure with high in-plane thermal conductivity and controllable vertical thermal conductivity; furthermore, it may employ a blend of carbon nanotubes and aluminum nitride particles to form a point-to-line composite anchoring structure on the void walls. Based on any of the above filler combinations, this application achieves differentiated thermal responses and interface adaptability to volume-dominant void clusters, supporting the feasibility of localized filling and / or localized compression processes in this application.

[0144] Example 7, in yet another embodiment, this application also provides a void topology reshaping process for direction-dominant void clusters, including:

[0145] Step 1: When the LED chip package is in a pre-cured or semi-cured state, apply directional extrusion or local hot pressing to the area where the directional dominant void cluster is located, so that the long axis of the directional dominant void cluster deviates from the main heat flow direction, or the directional dominant void cluster splits into multiple sub-voids that are dispersed along the non-main heat flow direction.

[0146] Among them, the direction-dominant void cluster can refer to a key void cluster that is dominated by slender crack-type voids and whose angle between the void principal axis direction and the main heat flow direction is less than a preset angle threshold.

[0147] Direction-dominated void clusters are collections of micro-scale voids with an aspect ratio greater than 3, and the angle between their main axis direction and the main heat flow direction is less than 15° when undisturbed.

[0148] Direction-dominated void clusters form low thermal resistance bypass channels in the direction of main heat flow. Their heat flow blocking rate is lower than that of volume-dominated void clusters, but the heat flow path deflection effect caused by unit projected area is stronger.

[0149] Directional extrusion involves applying mechanical pressure at an angle of 20° to 70° to the direction of the main heat flow, causing the slender fissures to undergo elastic buckling or viscous rotation, thereby changing their principal axis direction.

[0150] Local hot pressing involves applying localized heating and pressure simultaneously to the region where the directional dominant void cluster is located. By utilizing the viscoelastic response of the encapsulation medium in the pre-cured state, it promotes directional flow of the material around the void, induces deflection of the long axis of the void, or induces the void to split along the shear direction.

[0151] By deviating the long axis direction of the direction-dominant void cluster from the main heat flow direction, the original included angle is increased from less than 15° to greater than 30°, thereby significantly increasing the actual path length required for heat flow to pass through the void cluster;

[0152] The direction-dominant void cluster is split into multiple sub-voids that are dispersed along the non-main heat flow direction. This is achieved by inducing stress concentration points in the middle of the void through asymmetric thermal pressure or gradient pressure field, causing it to break into 2 to 4 independent sub-voids along the direction perpendicular to the original main axis. The main axis directions of each sub-void are distributed at an angle of 20° to 90° to each other.

[0153] This application, for example, involves deflecting the long axis of a directionally dominant void cluster using directional extrusion; it also involves causing buckling deformation of the directionally dominant void cluster and altering its principal axis orientation using localized hot pressing; further, it involves splitting the directionally dominant void cluster into multiple dispersed sub-voids along the shear direction using a gradient temperature field coupled with unilateral pressure. This application achieves spatial orientation control of the directionally dominant void cluster based on any of the above methods, thereby weakening its low-resistance conductivity in the main heat flow direction.

[0154] Step 2: After directional extrusion or local hot pressing, the area is pre-cured to fix the orientation of the reshaped voids;

[0155] Among them, pre-curing is to perform rapid thermal initiation cross-linking on the region where the directional dominant void cluster is located after the directional external force intervention is completed, so that the encapsulation medium completes the initial network structure locking under the new void configuration;

[0156] Pre-curing is one of photo-initiated curing, thermal-initiated curing, or dual-mode initiated curing, and its curing depth is controlled within the range of 50μm to 300μm to ensure that only the surrounding medium structure of the voids is stabilized without affecting the overall fluidity of the package.

[0157] Pre-curing is carried out at 60℃~100℃ for 5min~30min, or under ultraviolet light irradiation (365nm wavelength, energy density of 100mJ / cm²~500mJ / cm²) for 10s~60s;

[0158] The orientation of the voids after reshaping is fixed in order to suppress the rebound, closure or reorientation of the voids due to residual stress relaxation or thermal expansion differences during subsequent processes.

[0159] This application, for example, pre-cures the reshaped orientation-dominant void cluster region using a low-temperature, short-time thermosetting method; this application, for example, pre-cures the reshaped orientation-dominant void cluster region using localized ultraviolet irradiation; further, this application uses a photothermal synergistic triggering method to perform layered pre-curing of the reshaped orientation-dominant void cluster region. This application achieves structural stability of the void orientation based on any of the above methods, preventing morphological regression during subsequent axial thermal section construction or overall curing.

[0160] Example 8, in an optional embodiment, this application also provides a gap topology reshaping process for connectivity-dominant gap clusters, including:

[0161] Step 1: Construct a thermally conductive bridging strip between adjacent gaps, or reclose the encapsulation medium between adjacent gaps through localized thermal pressing, so as to sever the series connection relationship of the dominant gap cluster along the main heat flow direction;

[0162] Among them, the thermally conductive bridging strip can refer to a continuous or quasi-continuous thermally conductive structure that is artificially introduced between two or more gaps arranged sequentially along the main heat flow direction and has a thermal conductivity significantly higher than that of the surrounding encapsulation medium; local thermal pressure can refer to the physical action of applying a controllable combination of temperature and pressure to the area where the connected dominant gap cluster is located when the LED chip package is in a pre-cured or semi-cured state, so as to cause the incompletely cross-linked encapsulation medium in this area to undergo plastic flow and interface refusion; the series connection relationship can refer to multiple gaps arranged end to end in the main heat flow direction, and the shortest distance between adjacent gaps is less than a preset connection threshold, thereby forming a continuous shielding strip or heat flow bypass induction channel on the heat flow path.

[0163] Thermally conductive bridging bands can be high thermal conductivity particle bands formed by the enrichment of thermally conductive filling materials around the voids; they can be thermally conductive thin layers formed by the directional stacking of sheet-like fillers; or they can be cross-void thermally conductive chains formed by fibrous fillers bridging both sides of the voids. Localized hot pressing can be applying a pressure of 0.1MPa to 0.5MPa to the target area within a temperature range of 100℃ to 160℃ and maintaining it for 10s to 120s; it can also be using a temperature gradient field combined with unidirectional pressure to drive the encapsulation medium to flow into the void gaps and achieve physical closure; or it can be hot pressing under vibration-assisted conditions to enhance the fluidity of the medium and the wettability of the interface. The breaking of the series connection relationship can refer to eliminating the direct thermal connectivity between voids in the direction of the main heat flow, so that the heat flow can no longer continuously bypass or be blocked along the original void chain path, but instead achieves cross-void conduction through thermally conductive bridging bands or densified media.

[0164] This application can form a cross-gap thermally conductive connection segment by constructing a thermally conductive bridging strip; it can also sever the physical connection between gaps by using localized thermal pressing to promote the refusion of the encapsulation medium; further, it can also be a composite method of simultaneously performing the construction of a thermally conductive bridging strip and localized thermal pressing closure, enhancing the interfacial bonding strength between the bridging structure and the gap sidewall while constructing the bridging structure. Based on any of the above methods, this application effectively severs the series connection relationship of the dominant gap cluster along the main heat flow direction, thereby weakening its cutting effect on the main heat flow path and improving the continuity of heat flow through the gap region and the equivalent thermally conductive cross-sectional area.

[0165] Step 2: The thermal bridging strip extends in a direction that intersects with the main heat flow direction to form a cross-gap thermal connection section between adjacent gaps;

[0166] Among them, the thermal bridging strip can refer to a functional thermally conductive structure used to establish a transverse heat conduction path between adjacent gaps; the cross direction can refer to the main extension direction of the thermal bridging strip being non-parallel to the main heat flow direction, with an included angle ranging from 30° to 90°, and optionally from 45° to 75°; the cross-gap thermally conductive connection section can refer to a structural unit that guides the thermal bridging strip to simultaneously connect at least two gap sidewalls in space and form a continuous heat conduction path inside it.

[0167] The thermal bridging strip can be a strip-shaped structure extending at a 45° angle to the main heat flow direction; it can be a zigzag or sawtooth-shaped tortuous strip structure; or it can be a multi-point connection structure that spreads out in an arc or fan shape between adjacent gaps. The intersecting direction extension can refer to the thermal bridging strip forming a non-zero angle with the main heat flow direction on the horizontal projection plane; it can also refer to its extension along an inclined plane in three-dimensional space, making the entire bridging strip orthogonal or intersecting with the heat flow vector at a large angle; or it can refer to the bridging strip having the maximum thermally conductive cross-sectional area on the section perpendicular to the main heat flow direction. The cross-gap thermally conductive connection section can refer to the bridging strip being anchored at both ends to the sidewall surfaces of the two gaps; it can refer to the bridging strip spanning the gap in the middle and forming extended contact areas on both sides; or it can refer to the bridging strip forming micro-protrusions or embedded interlocking structures at the gap sidewalls to enhance thermal contact reliability.

[0168] This application can increase the effective thermally conductive projection length in the main heat flow direction by extending the thermally conductive bridging strip at a 45° angle; it can also extend the actual heat conduction path and improve the uniformity of heat diffusion by arranging the thermally conductive bridging strip in a zigzag pattern; further, it can enhance the interfacial bonding strength and thermal contact stability between the bridging strip and the encapsulation medium by forming a micro-protrusion anchoring structure on the sidewall of the gap. Based on any of the above methods, this application obtains a cross-gap thermally conductive connection segment with high thermal bridging efficiency and strong structural stability between adjacent gaps.

[0169] Example 9, in an optional implementation, this application further provides that step S6 involves sequentially constructing a continuous heat-conducting structure along the heat flow path within the main heat diffusion region, specifically including:

[0170] Step 1: First, construct a low-void thermal conductivity zone directly below the chip in the main heat diffusion core area;

[0171] Among them, the main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the packaging thickness direction (see Example 3). This area corresponds to the vertical projection range of the high heat-generating area near the chip pn junction or electrode at the interface of the die bond layer, the encapsulating adhesive layer and the substrate. The low void thermal conduction area directly below the chip can refer to the dense thermal conduction sub-region formed by void topology reshaping treatment in this projection area, which has a void volume fraction that is significantly lower than that of the surrounding area.

[0172] The function of this heat-conducting zone is to reduce the initial interface thermal resistance at the heat source output end and avoid local heat flow interruption and hot spot accumulation caused by the accumulation of voids directly below the chip. Its formation depends on the reshaping actions such as local filling, local compression or thermal bridging performed by S5 in Example 1 on the key void cluster located in the main heat diffusion core area, and is based on the infiltration of thermally conductive filling material and the synergistic effect of pressure in Example 6.

[0173] This application can, for example, use a local filling method to inject a low-viscosity thermally conductive supplementary adhesive into the main heat diffusion core region, and use vacuum assistance and local pressure to promote the full penetration of the thermally conductive material along the gap between the voids and the surrounding encapsulation medium, thereby forming a physical region with low porosity and high thermal conductivity continuity directly below the chip. Alternatively, this application can use a local compression method to apply local pressure intersecting the main heat flow direction in the main heat diffusion core region, causing the incompletely cured encapsulation medium to flow in a controlled manner and compress the void volume, thereby improving the thermal conductivity continuity of this region. Furthermore, this application can use a thermal bridging method to enrich lamellar boron nitride or spherical alumina fillers around the critical void clusters directly below the chip, forming a cross-void thermally conductive compensation band to maintain the integrity of the longitudinal heat flow path in this region. Based on any of the above methods, this application obtains a low-void thermally conductive region directly below the chip, with a void obstruction rate of less than 10% in the main heat flow direction, optionally less than 5%, thereby ensuring the initial continuity of heat flow at the heat source output end.

[0174] Step 2: Then, construct a continuous thermally conductive section extending along the thickness direction of the package in the axial thermally conductive zone;

[0175] Among them, the axial heat conduction zone is a heat conduction area formed by extending from the main heat diffusion core area toward the external heat dissipation interface (see Example 3). Its spatial orientation is basically consistent with the main heat flow direction, covering the longitudinal path between the die-bonding layer, the encapsulating adhesive layer and the substrate heat conduction interface; the continuous heat conduction segment can refer to a columnar heat conduction structure formed in this area that runs through the encapsulation thickness direction and has no significant gaps blocking the heat conduction path or connecting the heat resistance chain.

[0176] The function of this heat-conducting section is to open up the longitudinal main channel between the chip heat source and the external heat dissipation interface, to receive the heat flow output from the low-gap heat-conducting zone, and to suppress the heat flow dispersion and lateral circulation caused by the discrete distribution of axial gaps. Its construction starts from the low-gap heat-conducting zone formed above, continues the reshaping action performed by S5 in embodiment 1 on the key gap clusters in the axial heat-conducting zone, and strictly corresponds to the timing logic constructed in embodiment 9.

[0177] This application can be implemented, for example, by using a thermally conductive bridging method, to set thermally conductive particle enrichment bands at intervals along the main heat flow direction in the axially heat-conducting zone, with each band having a width of 5μm to 20μm and a band spacing less than 3 times the average pore size of a single void, so that heat flow can form a cross-void jump conduction between adjacent enrichment bands; this application can also be implemented, for example, by using a local hot-pressing method, to apply gradient pressure along the thickness direction in the axially heat-conducting zone, causing the high thermal conductivity filler in the encapsulation medium to migrate and accumulate directionally along the main heat flow direction, forming a continuous thermally conductive channel with an increasing filler concentration gradient; furthermore, this application can also be implemented, for example, by using a method of treating interconnected dominant void clusters (see Example 8), after identifying three or more voids arranged in series along the main heat flow direction in the axially heat-conducting zone, a thermally conductive bridging band is constructed between adjacent voids that intersects with the main heat flow direction, so as to cut off their longitudinal interconnection thermal resistance relationship and rebuild the transverse thermal coupling path. This application obtains a continuous thermally conductive section extending along the thickness direction of the package based on any of the above methods, such that the continuity of the heat flow section in the axial direction is not less than 90%, and optionally not less than 95%.

[0178] Step 3: Finally, construct a heat flow redirection bridging section in the lateral transition heat conduction zone;

[0179] The lateral transition thermal conduction zone is a heat flow turning area formed by the extension of the LED chip edge toward the bracket or substrate (see Example 3). Its geometry is fan-shaped or cone-shaped, corresponding to the interface transition zone where the heat flow at the chip edge turns toward the heat conduction part of the bracket or the heat conduction area of ​​the lead frame. The heat flow turning bridge section can refer to the cross-interface thermal conduction structure constructed in this area to bridge the interface thermal resistance increase and heat flow dispersion effect caused by the sudden change in heat flow direction.

[0180] The function of this bridging section is to compensate for the heat flow divergence, local thermal resistance jump and hot spot shift caused by geometric discontinuity and material interface differences at the heat flow turning point; its construction uses the above-mentioned continuous heat-conducting section as the heat flow input end, receives the longitudinal heat flow from the axial heat-conducting zone, and guides the heat flow to smoothly transition to the lateral heat dissipation path through structural design; its technical implementation is directly related to the limitation of the heat-conducting bridging strip in Embodiment 8 extending in the direction intersecting with the main heat flow direction, and does not introduce the parameter content of the controlled micropore buffer zone in Embodiment 10.

[0181] This application can, for example, employ a thermal bridging method by arranging sheet-like graphene filler enrichment strips along the chip edge and the interface connecting the chip and the support in the lateral transition thermally conductive zone. The strips extend in the same direction as the main heat flow turning direction and have a width of 8μm to 15μm, enabling the heat flow to achieve direction reconstruction and flux redistribution at the interface. Alternatively, this application can employ a localized hot-pressing method by applying oblique pressure to the heat flow turning zone at the chip edge, causing the thermally conductive filler in the packaging medium to accumulate at the interface corner and form a thermally conductive bridging protrusion structure with curvature matching characteristics. Furthermore, this application can employ a void orientation deflection method (see Example 7) by applying directional extrusion to the elongated, slit-type voids in the lateral transition thermally conductive zone, causing their main axis direction to change from parallel to the chip edge to an angle of 30° to 60° with the heat flow turning direction, thereby reducing their positive shielding rate on the turning heat flow and enhancing lateral thermal coupling capability. This application obtains a heat flow turning bridge section based on any of the above methods, such that the equivalent heat conduction area at the heat flow turning interface is increased by no less than 40%, and optionally no less than 60%, compared with the untreated area.

[0182] Example 10, another optional embodiment, this application also provides partitioning and curing in step S7, including:

[0183] Step 1: After completing the void topology reshaping treatment of the main heat diffusion core area and the axial heat conduction zone, first perform the first stage of curing on the main heat diffusion core area and the axial heat conduction zone;

[0184] The main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the package thickness direction (see Example 3), and the axial heat conduction area is the heat conduction area formed by the main heat diffusion core area extending towards the external heat dissipation interface (see Example 3).

[0185] The first stage of curing is used to lock the low-gap thermal conductive area directly below the chip and the continuous thermal conductive section extending along the package thickness direction, to prevent subsequent process disturbances from causing deformation of the main heat flow channel structure or gap rebound.

[0186] The process parameters for the first stage of curing can be: temperature of 60℃~100℃, time of 5min~30min, pressure of 0MPa~0.2MPa, and ultraviolet irradiation dose of 100mJ / cm²~500mJ / cm² (when using a photocuring system).

[0187] This application, for example, can employ a low-temperature, short-time pre-curing method to initially cross-link and fix the main heat diffusion core region and the axial heat conduction zone, based on the requirement of prioritizing stability of the main heat flow path. Alternatively, based on the rheological properties of the encapsulating medium, this application can initiate pre-curing by applying localized thermo-pressure to the main heat diffusion core region while simultaneously raising the temperature, allowing the incompletely flowed thermally conductive filler to complete initial network formation under pressure constraint. Furthermore, this application can also achieve space-constrained thermally initiated curing by selectively heating with infrared radiation only to the corresponding areas of the main heat diffusion core region and the axial heat conduction zone. Based on any of the above methods, this application obtains early stability of the structure of the main heat diffusion core region and the axial heat conduction zone, ensuring that they do not collapse, shift, or re-cavitation in subsequent processes.

[0188] Step 2: After the heat flow diversion bridging section is constructed, the lateral transition heat conduction zone is cured in the second stage.

[0189] The heat flow turning bridge section is a heat flow turning heat conduction structure built in the lateral transition heat conduction zone to connect the chip edge with the bracket or substrate (see Example 9). The lateral transition heat conduction zone is a heat flow turning area formed by the LED chip edge extending toward the bracket or substrate (see Example 3).

[0190] The second stage of curing is used to fix the spatial configuration and thermal conductivity of the heat flow diversion bridging section, ensuring that it does not debond or break under thermal stress.

[0191] The process parameters for the second stage of curing can be: temperature of 100℃~160℃, time of 20min~120min, pressure of 0MPa~0.1MPa, or irradiation with 365nm wavelength ultraviolet light in the photocuring system with a dose of 800mJ / cm²~2000mJ / cm².

[0192] This application, for example, leverages the characteristic that the thermal flow transition zone's function depends on the integrity of the bridging structure by employing a medium-temperature delayed curing method to promote the full directional arrangement of fillers within the thermally conductive bridging strip and enhance interfacial bonding. Alternatively, it can utilize the irregular geometry of the lateral transition thermally conductive zone by simultaneously applying gradient pressure during curing, enabling progressive bonding and curing between the two ends of the bridging segment and the chip edge and support interface. Furthermore, this application can control the surface temperature distribution of the lateral transition thermally conductive zone through thermal imaging feedback, ensuring the temperature in the center of the bridging segment is slightly higher than at the ends to compensate for the hysteresis effect of heat conduction and improve the overall thermal uniformity of the bridging segment. Based on any of the above methods, this application achieves structural robustness and thermal continuity of the thermal flow transition bridging segment under thermal cycling loads.

[0193] Step 3: After curing in the main heat diffusion zone, cure the non-main heat zone as a whole to preserve the controlled micropore buffer zone in the non-main heat zone;

[0194] The main heat diffusion zone includes the main heat diffusion core zone, the axial heat conduction zone, and the lateral transition heat conduction zone (see Examples 2, 3, and 9), while the non-main heat diffusion zone is the remaining encapsulation area other than the main heat diffusion zone.

[0195] Controlled micropore buffer zones can refer to a collection of microscale voids with a defined volume fraction range that are retained in non-main hot zones. Their purpose is to mitigate the interfacial stress generated by the mismatch of the thermal expansion coefficients of the material during thermal cycling of the encapsulation.

[0196] The volume fraction of micropores in the controlled micropore buffer zone is 0.1% to 8%, which has been determined by thermal cycling reliability verification: when the volume fraction is below 0.1%, the stress buffering capacity is insufficient and it is easy to induce interfacial microcracks; when the volume fraction is above 8%, the mechanical strength decreases significantly, leading to structural failure of the package in drop or vibration tests.

[0197] This application can, for example, utilize a slow heating and pressure-free overall curing method based on the functional positioning of non-primary heat zones, which do not require high heat flux density. This allows the encapsulating medium in these areas to naturally shrink while preserving the original microporous structure. Alternatively, this application can control the ambient humidity and atmosphere composition in the non-primary heat zone area, introducing controllable water vapor or inert gas in the later stages of curing to suppress excessively high local crosslinking density, thereby stabilizing the micropore volume fraction. Furthermore, this application can dynamically adjust the curing endpoint temperature and holding time of the non-primary heat zone based on thermal imaging feedback results. When the thermal expansion strain rate of this area is detected to be below a threshold, curing is terminated early to maintain the preset microporosity. Based on any of the above methods, this application achieves reproducible and controllable stress buffering capabilities in the non-primary heat zone while maintaining structural integrity.

[0198] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A high-heat-dissipation packaging method for LED chips based on microscale gap topology reshaping, characterized in that, include: S1. Perform void detection on the LED chip package in the pre-cured or semi-cured state, and divide it into multiple continuous detection units along the preset heat transfer direction from the main heat-generating area of ​​the LED chip to the external heat dissipation interface. Obtain the spatial distribution information of micro-scale voids in each detection unit, extract at least one void topology parameter from the void position, equivalent aperture, aspect ratio, spacing between adjacent voids and void main axis direction, and form a void topology distribution map. S2. Based on the heat transfer relationship between the main heat-generating area of ​​the LED chip and the external heat dissipation interface, determine the main heat diffusion area within the LED chip package. The main heat diffusion area includes at least two of the following: the main heat diffusion core area, the axial heat conduction area, and the lateral transition heat conduction area. S3. Based on the void topology distribution map and the main heat diffusion zone, cluster and identify multiple voids located in the same detection unit and arranged sequentially along the main heat flow direction. When the multiple voids meet at least two of the following conditions, they are determined to be key void clusters: the total projected length in the main heat flow direction is greater than a preset multiple of the average aperture of a single void; the spacing between adjacent voids is less than a preset connection threshold; the angle between the void main axis direction and the main heat flow direction is less than a preset angle threshold; and they are located in the main heat diffusion core area or the lateral transition heat conduction area. S4. The key void clusters are classified according to their heat resistance mechanisms, including volume-dominant void clusters, orientation-dominant void clusters, and connectivity-dominant void clusters. S5. Perform matching void topology reshaping processing for different types of critical void clusters. Specifically, for volume-dominant void clusters, perform local filling and / or local compression; for orientation-dominant void clusters, perform directional extrusion, local hot pressing, and / or void orientation deflection processing; and for connectivity-dominant void clusters, perform thermal bridging and / or connectivity chain disconnection processing. S6. After completing the void topology reshaping process, a continuous thermal conduction structure is constructed sequentially along the heat flow path in the main heat diffusion region. The continuous thermal conduction structure includes at least two of the following: a low void thermal conduction region directly below the chip, a continuous thermal conduction section extending along the package thickness direction, and a heat flow turning bridging section. S7. The continuous thermally conductive structure is partitioned and cured, wherein the main heat diffusion area is cured in stages first, and the non-main heat area is cured as a whole, so that the non-main heat area retains a controlled micropore buffer zone, so as to obtain a high heat dissipation packaging structure for LED chips.

2. The method according to claim 1, characterized in that, The void detection in step S1 uses at least one of the following: X-ray detection, industrial CT detection, scanning acoustic microscopy detection, infrared thermography-assisted inversion detection, and section microscopy detection. The formation of the void topology distribution map includes: statistically analyzing the degree of overlap of the projections of voids in each detection unit along the preset heat transfer direction, and combining the spacing between adjacent voids to form a distribution map characterizing the void aggregation state and connectivity trend.

3. The method according to claim 2, characterized in that, In step S2, the main heat diffusion core area is the projection area of ​​the main heat-generating area of ​​the LED chip in the packaging thickness direction; The axial heat conduction zone is a heat conduction area formed by extending the main heat diffusion core area toward the external heat dissipation interface. The lateral transition thermal conduction zone is a heat flow transition area formed by the extension of the LED chip edge toward the bracket or substrate.

4. The method according to claim 1, characterized in that, In step S3, when clustering and identifying multiple gaps located within the same detection unit and arranged sequentially along the main heat flow direction, multiple gaps that meet at least two of the following conditions are identified as key gap clusters: The total projected length in the main heat flow direction is greater than 2 to 10 times the average pore size of a single pore; The spacing between adjacent gaps is less than a preset connection threshold; The angle between the main axis direction of the void and the main heat flow direction is less than a preset angle threshold. The multiple gaps are located in the main heat diffusion core area or the lateral transition heat conduction area.

5. The method according to claim 4, characterized in that, In step S4, the key void clusters are classified according to their heat-insulating mechanisms, specifically including: Key void clusters dominated by aperture-enlarged voids are classified as volume-dominant void clusters. The key void clusters, which are mainly composed of slender, fissure-type voids and whose angle between the void principal axis and the main heat flow direction is less than the preset angle threshold, are classified as direction-dominant void clusters. Key void clusters formed by connecting two or more adjacent voids in series along the main heat flow direction are classified as connected dominant void clusters.

6. The method according to claim 5, characterized in that, For the volume-dominant void cluster, the void topology reshaping process in step S5 includes: A thermally conductive filler material is introduced into the region where the volume-dominant void cluster is located, allowing the thermally conductive filler material to penetrate along the gap between the void and the surrounding encapsulation medium, and applying local pressure to the region after the thermally conductive filler material has penetrated, so as to reduce the cavity volume of the volume-dominant void cluster. The direction of the applied local pressure is intersected with the direction of the main heat flow; The thermally conductive filler material includes at least one of silver filler, alumina filler, boron nitride filler, aluminum nitride filler, graphene filler, and carbon nanotube filler.

7. The method according to claim 5, characterized in that, For the direction-dominant void cluster, the void topology reshaping process in step S5 includes: When the LED chip package is in a pre-cured or semi-cured state, directional extrusion or local hot pressing is applied to the area where the direction-dominant void cluster is located, so that the long axis direction of the direction-dominant void cluster deviates from the main heat flow direction, or the direction-dominant void cluster is split into multiple sub-voids that are dispersed along the non-main heat flow direction. After directional extrusion or local hot pressing, the area is pre-cured to fix the void orientation after reshaping.

8. The method according to claim 5, characterized in that, For the aforementioned connectivity-dominated gap cluster, the gap topology reshaping process in step S5 includes: A thermally conductive bridging band is constructed between adjacent gaps, or the encapsulation medium between adjacent gaps is reclosed by localized thermal pressing to break the series connection relationship of the dominant gap cluster along the main heat flow direction; The thermally conductive bridging strip extends in a direction that intersects with the main heat flow direction to form a cross-gap thermally conductive connection section between adjacent gaps.

9. The method according to claim 1, characterized in that, In step S6, a continuous heat conduction structure is constructed sequentially along the heat flow path in the main heat diffusion zone. Specifically, this includes: first, constructing a low-gap heat conduction zone directly below the chip in the main heat diffusion core zone; then, constructing a continuous heat conduction segment extending along the package thickness direction in the axial heat conduction zone; and finally, constructing a heat flow turning bridge segment in the lateral transition heat conduction zone, so that a continuous heat conduction chain is formed between the main heat-generating area of ​​the LED chip and the external heat dissipation interface.

10. The method according to claim 9, characterized in that, Step S7, partition curing, includes: After completing the void topology reshaping process of the main heat diffusion core area and the axial heat conduction area, the main heat diffusion core area and the axial heat conduction area are first solidified in the first stage; after the heat flow turning bridge section is constructed, the lateral transition heat conduction area is solidified in the second stage; and after the main heat diffusion area is solidified, the non-main heat area is solidified as a whole so that the non-main heat area retains the controlled micropore buffer zone. The controlled micropore buffer zone has a micropore volume fraction of 0.1% to 8%.

Images