QFN multi-core sealing process

Abstract

The application relates to a QFN multi-core sealing process, and belongs to the technical field of QFN packaging.The process comprises the following steps: S100, a ceramic base is subjected to groove processing to form an annular protrusion, and a preset height gap is formed between the highest point of a lead wire and the top surface of the annular protrusion after lead wire bonding, and the height gap meets the requirement that a colloid can cover the highest point of the lead wire under the action of surface tension; S200, a metal layer is formed in the groove of the ceramic base; S300, the metal layer forms at least two mutually independent circuit unit connecting regions; S400, a chip is installed to the corresponding circuit unit connecting region; S500, the base plate is connected with the lead wire frame; S600, lead wire bonding is performed, and the projection of the highest point of all the lead wires is located in the region defined by the annular protrusion; S700, glue injection is performed on the arch starting position of the lead wire to fix the arch starting position; and S800, injection packaging is performed.The application has the effects of fixing the arch starting position of the lead wire, reducing the risk of colloid overflow, and further reducing the delamination risk of the base plate and the lead wire frame.

Description

Technical Field

[0001] This application relates to the field of QFN packaging technology, and in particular to a QFN multi-core encapsulation process. Background Technology

[0002] QFN packaging is a common leadless packaging form, which has advantages such as small package size, low lead inductance, and good heat dissipation performance, and is widely used in the field of integrated circuit packaging. With the continuous improvement of the functional integration of integrated circuits, multi-chip co-packaging technology has gradually been applied, that is, integrating multiple bare dies in the same package to achieve system-level functional integration and miniaturization.

[0003] In existing QFN multi-core co-package structures, the die is typically mounted on a die pad or metal base island area. Electrical connections between the die and the lead frame, as well as between multiple dies, are achieved through wire bonding. To meet the requirements of wire bonding process and reliability, the leads are typically formed into an arched structure of a certain height during the bonding process.

[0004] However, the arched section of the lead is the most susceptible to deformation, displacement, and parameter drift in the lead structure. Especially in multi-core encapsulation structures, as the number of bare dies increases, the number and density of leads significantly increase, further compressing the spacing between leads. During subsequent packaging processes (such as injection molding), the flow, erosion, and stress of the encapsulating material can easily cause lead displacement or collapse, resulting in the actual distance between leads deviating from the design value. This can lead to enhanced electromagnetic coupling, contact between leads, or short circuits, and in severe cases, cause the packaged product to malfunction, resulting in reduced product yield.

[0005] Currently, no solution exists in the technology to effectively fix the arched part of the lead without affecting the reliability of the package. On the one hand, the arch height of the lead usually needs to meet specific process and reliability requirements and cannot be arbitrarily reduced; therefore, it is often difficult to directly cover the entire substrate surface with a fixing adhesive layer that can cover the height of the lead. This is because the contact angle and flow characteristics of conventional fixing adhesive materials are difficult to ensure that the adhesive stably covers the arched part of the lead, and even if a specific fixing adhesive material is used to achieve coverage of the arched part, adhesive overflow is likely to occur. In existing QFN packaging structures, the base island is usually connected to the lead frame through an intermediate layer. Once the fixative overflows, it can easily flow into the bottom of the base island or the bonding area between the base island and the lead frame, and may interact adversely with the intermediate layer, thereby introducing risks such as delamination and interface failure. Summary of the Invention

[0006] In order to fix the arched part of the lead wire, reduce the risk of colloid overflow, and thus reduce the risk of delamination between the substrate and the lead frame, this application provides a QFN multi-core encapsulation process.

[0007] This application provides a QFN multi-core encapsulation process using the following technical solution: A QFN multi-core co-packing process includes the following steps: S100. Groove processing is performed on the ceramic substrate to form an annular protrusion along its periphery. The depth of the groove is controlled to be greater than the sum of the thickness of the chip and the metal layer. After wire bonding, a preset height gap is formed between the highest point of the lead and the top surface of the annular protrusion. The height gap is such that the colloid can cover the highest point of the lead under the action of surface tension. S200, A metal layer is formed in the groove of the ceramic substrate; S300. Perform photolithography and etching on the metal layer to form at least two independent circuit unit connection regions in the metal layer. S400, Install the chip to the corresponding circuit unit connection area; S500, Connect the substrate to the lead frame; S600, Perform wire bonding so that the projection of the highest point of all leads is located within the area defined by the annular protrusion; S700, Apply adhesive to fix the arched part of the lead wire; S800, perform injection molding encapsulation.

[0008] By adopting the above technical solution, an annular protrusion is formed on the periphery of the substrate and the groove depth is controlled, so that a preset height gap is formed between the highest point of the lead and the top surface of the annular protrusion after the lead bonding. Thus, in the subsequent glue injection process, the glue can stably cover the highest point of the lead under the action of surface tension, and the risk of glue overflow is reduced by the restriction of the annular protrusion.

[0009] Therefore, on the one hand, the arched part of the lead wire can be effectively fixed, reducing the risk of displacement, deformation or mutual contact of the lead wire due to the flow of encapsulant during injection molding; on the other hand, it can reduce the penetration of the colloid into the bottom of the substrate or the bonding area between the substrate and the lead frame, thereby reducing the risk of delamination between the substrate and the lead frame and improving the reliability and yield of the multi-core QFN package structure.

[0010] Optionally, in step S100, the specific method for processing the groove of the ceramic substrate is as follows: the ceramic substrate is removed by a staged laser processing method, a processing allowance is reserved at the groove contour boundary, and the contour boundary is trimmed by low-energy fine finishing processing. The processed groove is cleaned: the groove is pre-rinsed to remove large particles; Subsequently, ultrasonic or mega-sonic cleaning is performed to remove fine particles, and the groove is circulated and replaced with liquid during the cleaning process.

[0011] By adopting the above technical solution, this groove processing method can effectively reduce the heat-affected zone generated in the ceramic substrate during the groove processing, reduce the generation of microcracks, and reduce the risk of edge chipping of the annular protrusion, thereby improving the structural integrity and processing consistency of the annular protrusion. This cleaning method can effectively remove residual processing particles from the groove, especially at the corners of the groove, reducing the accumulation of particles at the corners of the groove. This reduces the risk of discontinuity, poor adhesion, or local defects in the metal layer caused by particle residue during the subsequent metal layer formation process, which is beneficial to improving the film quality and process stability of the metal layer.

[0012] Optionally, step S100 further includes the following steps: after the groove is processed, the top surface of the annular protrusion is modified, and the top surface of the annular protrusion is subjected to surface modification treatment so that the top surface of the annular protrusion is hydrophobic to the colloid.

[0013] By adopting the above technical solution, the top surface of the annular protrusion becomes hydrophobic to the colloid, thereby increasing the contact angle of the colloid at the top surface of the annular protrusion. This helps the colloid maintain a stable shape at the top surface of the annular protrusion, further reducing the risk of the colloid spreading along the outer side of the annular protrusion or overflowing, and enhancing the control over the flow of the colloid.

[0014] Optionally, in step S300, a preset spacing is maintained between adjacent circuit unit connection areas; The circuit unit connection area where the chip is installed is processed into a chip mounting area and a lead connection area, and the chip mounting area and the lead connection area are also controlled to have a preset distance. The preset spacing can reduce electromagnetic coupling between adjacent circuit unit connection areas and between chip mounting area and lead connection area.

[0015] By adopting the above technical solution, the setting of the preset spacing can reduce the electromagnetic coupling between the circuit unit connection areas and between the chip mounting area and the lead connection area. Furthermore, due to the setting of the preset spacing, the coverage area of ​​the metal layer is greatly reduced. In the case of multi-chip packaging and a large number of leads, the electromagnetic coupling between the leads and the base metal layer can be further reduced, which is conducive to ensuring the high-frequency performance and signal integrity of the chip after multi-core packaging. Meanwhile, during wire bonding in the wire connection area, due to the presence of a preset spacing, the heat generated during the bonding process is mainly transferred to the chip mounting part along the extension direction of the wire connection area, reducing the path of heat diffusion in the metal layer through multiple directions, thereby improving the local heating efficiency of the bonding point, which is conducive to shortening the bonding time and thus improving the overall packaging production efficiency.

[0016] Optionally, in step S500, the substrate is connected to the lead frame via nanosilver.

[0017] By adopting the above technical solution, the high thermal conductivity of nano-silver allows heat from the chip on the base island to be rapidly conducted to the lead frame, which helps to reduce the chip's operating temperature and improve heat dissipation efficiency. Simultaneously, the reliable electrical connection formed by nano-silver also helps maintain a low-impedance electrical path between the chip and the lead frame.

[0018] Optionally, in step S600, the lead wire is controlled to form a flat-topped arc shape.

[0019] By adopting the above technical solution, the wire can be controlled to form a flat-top arc shape during the wire bonding process, which can make the height of the highest point of the wire more stable and the distribution more consistent. This reduces the impact of wire height fluctuation on subsequent glue injection and injection molding processes, and is more conducive to covering the highest point of the wire while ensuring that the glue amount is the same each time. Meanwhile, the flat-top arc shape helps the lead maintain a small and controllable arch height when crossing the annular protrusion, reducing the risk of the lead contacting or colliding with the annular protrusion due to the large tilt angle when the lead tilts across it. It also helps the colloid to stably cover the arched part of the lead under the action of surface tension, thereby further improving the reliability of lead fixation and reducing the risk of lead displacement or deformation during injection molding.

[0020] Optionally, in step S700, the specific glue injection method is as follows: control the glue nozzle to align with the lead wire portion located above the annular protrusion, and control the glue injection position to be set close to the inner side of the annular protrusion; After determining the dispensing location, drip the adhesive onto the top surface of the annular protrusion; Repeat the above steps to apply adhesive to each external lead in turn.

[0021] By adopting the above technical solution and applying adhesive to each lead wire individually, the amount of adhesive used and the application position at each lead wire arch can be precisely controlled. Compared with overall potting in the groove, this method helps to reduce the amount of adhesive used and improve the consistency of adhesive application, with a lower risk of overflow, thus achieving more stable coverage and fixation of the lead wire arch. Meanwhile, by setting the dispensing position on the top surface of the annular protrusion and close to its inner side, the adhesive is allowed to expand slightly into the groove before curing, while its spread to the outer side of the annular protrusion is restricted, thereby reducing the risk of the adhesive overflowing to the outer side of the annular protrusion. Under the action of surface tension and wetting, the adhesive can climb along the direction of the lead wire and cover the arched part of the lead wire, which further reduces the possibility of adhesive overflow and delamination risk while ensuring the lead wire fixing effect.

[0022] Optionally, in step S700, the specific method of glue injection is as follows: glue is injected into the groove using an overall glue injection method, so that the glue covers the chip and lead wire in the groove, and the surface of the glue is controlled to be higher than the top surface of the annular protrusion, so that the glue covers the highest point of the lead wire under the action of surface tension. After the adhesive is applied, the adhesive is heated and cured at the application station.

[0023] By adopting the above technical solution, the adhesive is injected into the groove using an integral encapsulation method. Under the action of surface tension, the adhesive covers the highest point of the lead. This not only fixes the highest point of the external connection lead, but also fixes the lead between different circuit unit connection areas and the connection lead between chips. This not only improves the encapsulation efficiency, but also effectively reduces the fluctuation of electrical parameters caused by position changes of the lead in the multi-core encapsulation structure during high-frequency signal transmission, which is beneficial to improving the high-frequency performance and signal integrity of the packaged device.

[0024] In addition, the thermal conductivity of colloids is usually higher than that of injection molding packaging materials. After overall injection molding, the thickness of the injection molding packaging layer above the chip is reduced accordingly. This allows the heat generated by the chip to be dissipated not only through the bottom of the substrate, but also through the colloid layer and the thinner injection molding packaging layer to the top of the package. This helps to improve the overall heat dissipation performance of the packaging structure. Especially in multi-core structures where heat generation increases, it is more conducive to ensuring the stability of chip operation.

[0025] After the adhesive is applied, the adhesive is heated and cured at the adhesive application station. The adhesive needs to be shaped without moving the substrate to avoid overflow due to adhesive flow during the transfer process.

[0026] Optionally, in step S400, an additional heat sink is used for heat dissipation. A heat dissipation groove is processed on the heat dissipation surface of the heat sink, and a sealing layer is set in the heat dissipation groove so that the sealing layer extends out of the heat dissipation groove. A heat conductor is connected to the bottom surface of the heat sink, and the heat conductor is connected to the bottom surface of the groove. In step S800, the upper injection mold is brought into contact with the sealing layer, and the sealing layer is removed after injection molding is completed.

[0027] By adopting the above technical solution, the upper injection mold abuts against the sealing layer during the injection molding process, which can effectively prevent the injection molding material from entering the heat sink, thereby avoiding the heat sink from being blocked during the injection molding and packaging process. After the injection molding and packaging is completed and the sealing layer is removed, the heat sink is exposed to the outside of the package again. The inner sidewall and inner bottom wall of the heat sink form an additional heat dissipation surface, which is beneficial to increase the effective heat dissipation area of ​​the package structure, thereby improving the heat dissipation capacity from the top surface of the chip. Meanwhile, by setting a heat conductor to connect the bottom surface of the heat sink to the bottom surface of the groove, the heat generated by the chip can be transferred to the heat sink through the heat conductor in addition to being conducted downward through the ceramic substrate, and further dissipated to the outside of the package through the heat sink groove. This creates a multi-path heat dissipation channel under multi-core encapsulation conditions, thereby improving the overall heat dissipation performance of the package structure.

[0028] Optionally, in step S400, an expansion layer is first set inside the heat dissipation groove, and then the sealing layer is set on the expansion layer; In step S800, the expansion layer and the sealing layer are removed after injection molding is completed.

[0029] By adopting the above technical solution, the expansion layer expands during the heat preservation stage and pushes the sealing layer to fit the inner wall of the mold, reducing the flow of injection molding material between the sealing layer and the inner wall of the mold during the injection molding process. This compensates for the small gaps between the sealing layer, heat dissipation groove and mold caused by processing or assembly tolerances, so that the sealing layer can be exposed on the surface of the molded body without grinding after injection molding.

[0030] In summary, this application includes at least one of the following beneficial technical effects: 1. It can effectively fix the arched part of the lead wires, reducing the risk of displacement, deformation, or mutual contact of the lead wires due to the flow of encapsulant during injection molding. At the same time, it can reduce the penetration of the colloid into the bottom of the substrate or the bonding area between the substrate and the lead frame, thereby reducing the risk of delamination between the substrate and the lead frame and improving the reliability and yield of the multi-core QFN package structure; 2. Setting the preset spacing can reduce the electromagnetic coupling between circuit unit connection areas and between chip mounting area and lead connection area. Furthermore, due to the preset spacing, the coverage area of ​​the metal layer is significantly reduced. In the case of multi-chip packaging and a large number of leads, it can further reduce the electromagnetic coupling between the leads and the base metal layer, thereby helping to ensure the high-frequency performance and signal integrity of the chip after multi-chip packaging. 3. When wire bonding is performed on the wire connection area, due to the existence of the preset spacing, the heat generated during the bonding process is mainly transferred to the chip mounting part along the extension direction of the wire connection area, which reduces the path of heat diffusion in the metal layer in multiple directions, thereby improving the local heating efficiency of the bonding point, which is conducive to shortening the bonding time and thus improving the overall packaging production efficiency. 4. The adhesive is injected into the groove using an integral encapsulation method. Under the action of surface tension, the adhesive covers the highest point of the lead. This not only fixes the highest point of the external connection lead, but also fixes the lead between different circuit unit connection areas and the connection lead between chips. This not only improves the encapsulation efficiency, but also effectively reduces the fluctuation of electrical parameters caused by the position change of the lead in the multi-core encapsulation structure during high-frequency signal transmission. This is beneficial to improving the high-frequency performance and signal integrity of the packaged device. 5. The heat conductor connects the bottom surface of the heat sink to the bottom surface of the groove, so that the heat generated by the chip can be transferred to the heat sink through the heat conductor in addition to being conducted downward through the ceramic substrate, and further dissipated to the outside of the package through the heat sink groove. This creates a multi-path heat dissipation channel under multi-core encapsulation conditions, improving the overall heat dissipation performance of the package structure. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the overall process of an embodiment of this application.

[0032] Figure 2 This is a structural schematic diagram illustrating the formation of the annular protrusion and the multi-core sealing process in an embodiment of this application.

[0033] Figure 3 This is a schematic diagram of the lead wire structure used to illustrate the flat-top arc structure in an embodiment of this application.

[0034] Figure 4 This is a schematic diagram illustrating the specific process of processing the annular protrusion in an embodiment of this application.

[0035] Figure 5 This is a schematic diagram illustrating the specific dispensing process in an embodiment of this application.

[0036] Figure 6 This is a schematic diagram illustrating the structure of the annular protruding adhesive in an embodiment of this application.

[0037] Figure 7 This is a schematic diagram illustrating the specific process of full-area potting in the embodiments of this application.

[0038] Figure 8 This is a schematic diagram illustrating the structure of the groove for full-area glue filling in the embodiment of this application.

[0039] Figure 9This is a schematic diagram illustrating the specific process of heat sink installation in an embodiment of this application.

[0040] Figure 10 This is a schematic diagram illustrating the installation of the heat sink in an embodiment of this application.

[0041] Figure 11 This is a schematic diagram of Embodiment 1 of this application, illustrating the specific arrangement of the seven chips and the reserved spacing.

[0042] Figure 12 This is a schematic diagram of the structure of a multi-core packaged chip, as shown in Embodiment 1 of this application.

[0043] Figure 13 This is a schematic diagram of the structure of a multi-core packaged chip, as shown in Embodiment 2 of this application.

[0044] Explanation of reference numerals in the attached figures: 1. Ceramic substrate; 11. Groove; 12. Annular protrusion; 13. Height gap; 2. Metal layer; 21. Circuit unit connection area; 211. Chip mounting area; 212. Lead connection area; 22. Preset spacing; 3. Lead; 4. Colloid; 5. Heat sink; 51. Heat sink groove; 6. Expansion layer; 7. Sealing layer; 8. Thermal conductor; 9. Chip. Detailed Implementation

[0045] The following is in conjunction with the appendix Figure 1-13 This application will be described in further detail.

[0046] This application discloses a QFN multi-core encapsulation process.

[0047] like Figure 1 and Figure 2 The QFN multi-core packaging process includes the following steps: S100. Groove 11 is processed on the ceramic substrate 1 to form an annular protrusion 12 along its periphery. The depth of the groove 11 is controlled to be greater than the sum of the thicknesses of the chip 9 and the metal layer 2. After the lead wire 3 is bonded, a preset height gap 13 is formed between the highest point of the lead wire 3 and the top surface of the annular protrusion 12. The height gap 13 satisfies the requirement that the colloid 4 can cover the highest point of the lead wire 3 under the action of surface tension.

[0048] Specifically, the substrate includes a ceramic substrate 1 and a metal layer 2 formed on the surface of the ceramic substrate 1. The groove 11 is processed in the ceramic substrate 1, forming a recessed area in the middle of the ceramic substrate 1, while the peripheral part of the ceramic substrate 1 retains an annular protrusion 12 structure. The groove 11 can be processed by laser etching, mechanical processing, or a combination thereof, with laser processing being preferred to obtain higher processing accuracy.

[0049] The depth of the groove 11 is determined by a combination of the thickness of the chip 9, the thickness of the metal layer 2, and the arching height of the lead 3, so that the depth of the groove 11 is greater than the sum of the thickness of the chip 9 and the thickness of the metal layer 2.

[0050] The preset height gap 13 is the vertical distance between the highest point of the lead 3 and the top surface of the annular protrusion 12 after the lead 3 is bonded. The preset height gap 13 is used to ensure that the adhesive 4 can cover the highest point of the lead 3 under the action of surface tension during subsequent glue injection or dispensing, while the adhesive 4 is restricted by the annular protrusion 12 to prevent it from overflowing.

[0051] S200, A metal layer 2 is formed in the groove 11 of the ceramic substrate 1.

[0052] The metal layer 2 serves as the basic structure for the subsequent circuit unit connection area 21 and chip 9 mounting. Specifically, the metal layer 2 can be formed by physical vapor deposition (PVD), chemical vapor deposition (CVD), sputtering, evaporation, electroplating, or a combination of the above processes. For example, a seed layer can be formed in the groove 11 by sputtering or evaporation, and then the metal layer 2 of the required thickness can be formed by electroplating to improve the density and adhesion of the metal layer 2. The total thickness of the metal layer 2 can be set to approximately 5μm to 30μm, for example, 10μm to 20μm, to meet the process requirements of chip 9 mounting and lead bonding 3, while avoiding excessive thickness of the metal layer 2 from affecting subsequent photolithography and etching processes.

[0053] S300, perform photolithography and etching on the metal layer 2 to form at least two independent circuit unit connection areas 21. At the same time, control the adjacent circuit unit connection areas 21 to have a preset distance 22. The circuit unit connection area 21 on which the chip 9 is mounted is processed into a chip mounting area 211 and a lead connection area 212, and the chip mounting area 211 and the lead connection area are also controlled to have a preset distance 22. The preset spacing 22 can reduce electromagnetic coupling between adjacent circuit unit connection areas 21 and between chip mounting area 211 and lead connection area 212.

[0054] Specifically, photoresist is first coated on the surface of metal layer 2, and corresponding patterned areas are formed on metal layer 2 through mask exposure and development process. Then, wet etching, dry etching or a combination of both are used to remove the metal areas not protected by photoresist, so that metal layer 2 forms multiple electrically isolated circuit unit connection areas 21. The preset spacing 22 can be set according to the device operating frequency and wiring density, for example, about 50μm to 500μm, preferably 100μm to 200μm, so as to ensure the electrical isolation effect without significantly increasing the substrate area.

[0055] S400, Install chip 9 to the corresponding circuit unit connection area 21.

[0056] In this step, chip 9 specifically refers to a bare die with a gold backing, which is fixed to the chip mounting area 211 using conductive or non-conductive adhesive materials. For example, conductive adhesive, silver paste, or solder can be used to fix chip 9 to the metal layer 2 to achieve mechanical fixation while providing electrical connection and heat conduction path; or non-conductive adhesive can be used to fix chip 9 to the chip mounting area 211, and electrical connection can be achieved through subsequent wire bonding 3.

[0057] S500, Connect the substrate to the lead frame 3; The substrate and the lead frame 3 can be connected by welding, bonding or a combination thereof. For example, the substrate can be fixed to the lead frame 3 using solder or adhesive materials. Specifically, nano-silver materials can be used to improve thermal conductivity in order to achieve mechanical fixation and electrical connection.

[0058] S600, perform lead bonding 3 so that the projection of the highest point of all leads 3 is located within the area defined by the annular protrusion 12.

[0059] Specifically, the chip 9 pads are connected to the corresponding lead connection area 212 or lead 3 frame pins using a wire bonding process. During the wire bonding process, the direction and arc shape of the leads 3 are controlled so that the highest point of all leads 3 is located within the area defined by the annular protrusion 12 in the projection direction perpendicular to the substrate.

[0060] Furthermore, it is required that the highest point of the lead 3 is located at least within the area enclosed by the outer perimeter of the annular protrusion 12. Typically, the highest point of the connecting lead 3 inside the base island (e.g., the lead 3 between chips 9) is located within the area enclosed by the inner perimeter of the annular protrusion 12, while the highest point of the lead 3 connecting the base island to the peripheral pins is located above the top surface of the annular protrusion 12. Ideally, the orthographic projection of the highest point of the lead 3 should be located at the annular center line of the annular protrusion 12, or between the annular center line and the inner perimeter of the annular protrusion 12, to prevent the highest point of the lead 3 from approaching or exceeding the edge of the annular protrusion 12, thus providing stable space for subsequent glue injection and fixation of the arched part of the lead 3.

[0061] During the bonding process of lead 3, lead 3 can form different shapes such as arched arcs, flat-topped arcs, or composite arcs. Among these, the highest points of arched arcs are concentrated and their height fluctuates significantly, increasing the risk of contact or collision with the annular protrusion 12 when crossing it due to height instability. Composite arcs are sensitive to process parameters during lead 3 formation, easily leading to poor height consistency in lead 3. For example... Figure 3In contrast, in a preferred embodiment, the lead 3 is controlled to form a flat-topped arc shape, so that the highest point of the lead 3 is distributed more gently and has a higher height consistency along the length of the lead 3, thereby reducing the impact of the height fluctuation of the lead 3 on the subsequent encapsulation process and reducing the risk of the lead 3 contacting or colliding when crossing the annular protrusion 12. However, the flat-topped arc shape is more suitable for situations where there are few leads 3 and few bare wafers inside the base island.

[0062] S700, apply adhesive to fix the arched part of lead wire 3.

[0063] Specifically, after the bonding of the lead 3 is completed, colloid 4 is applied to the arched portion of the lead 3 within the area defined by the annular protrusion 12, so that the colloid 4 fixes the arched portion of the lead 3 after curing. The colloid 4 can be a potting compound, a fixing compound, or other colloid material suitable for fixing the lead 3.

[0064] Adhesive can be applied to the arched area of ​​the lead wire 3 using a dispensing method. This involves applying adhesive 4 to the top surface of the annular protrusion 12, allowing the adhesive 4 to cover the arched area of ​​the lead wire 3 under surface tension and wetting action, thereby fixing the lead wire 3. By controlling the dispensing position, the adhesive can be placed close to the inner circumference of the annular protrusion 12, specifically in the area between the annular center line and the inner circumference line of the annular protrusion 12. Simultaneously, controlling the amount of adhesive 4 allows it to preferentially spread inwards towards the groove 11, and is constrained by the annular protrusion 12 to prevent overflow outwards.

[0065] Alternatively, the adhesive 4 can be injected into the groove 11 using an overall injection method, so that the adhesive 4 covers the chip 9 and lead wire 3 in the groove 11, and the surface of the adhesive 4 is controlled to be higher than the top surface of the annular protrusion 12, so that the adhesive 4 covers the highest point of the lead wire 3 under the action of surface tension, thereby achieving overall fixation of the arched part of the lead wire 3.

[0066] After the glue is injected, the glue 4 is cured to set its shape, thereby suppressing the displacement or deformation of the lead 3 during the subsequent injection molding and encapsulation process.

[0067] S800, perform injection molding encapsulation.

[0068] The substrate and lead frame 3 are placed in an injection mold. Encapsulation material is injected into the mold using an injection molding process, covering the chip 9, lead 3, and the connection structure between the substrate and lead frame within the groove 11, thus forming a molded enclosure. The encapsulation material can be epoxy resin or other injection molding materials suitable for semiconductor packaging. After injection molding, the molded enclosure is cooled and demolded to obtain the encapsulated device.

[0069] Because the annular protrusion 12 is formed around the substrate and the depth of the groove 11 is controlled, a preset height gap 13 is formed between the highest point of the lead 3 and the top surface of the annular protrusion 12 after the lead 3 is bonded. As a result, during the subsequent encapsulation process, the adhesive 4 can stably cover the highest point of the lead 3 under the action of surface tension. Moreover, the adhesive 4 is restricted by the annular protrusion 12, which reduces the risk of overflow. This effectively fixes the arched part of the lead 3 and reduces the risk of displacement, deformation or mutual contact of the lead 3 due to the flow of encapsulation material during the injection molding process.

[0070] like Figure 4 Step S100 specifically includes the following steps: S110, Substrate pretreatment and positioning; The ceramic substrate 1 is loaded and fixed on the machining fixture using vacuum adsorption or clamping. The machining surface is then treated with dust removal and degreasing (e.g., dust-free air blowing + IPA wiping / spraying) to reduce the secondary sintering or re-deposition of particles during subsequent laser processing. Following this, a reference alignment (visual positioning or fixture positioning) is performed to establish the machining coordinate system for the groove 11 pattern. S120, laser rough machining to form grooves, with allowance reserved at the contour boundary; A short-pulse laser is used to remove the ceramic substrate 1 in stages to form the groove 11. Ultraviolet light or a short-pulse mode is preferred to reduce the heat-affected zone. The laser type can be nanosecond UV (e.g., 355 nm) for high-efficiency removal, preferably a picosecond / femtosecond laser (wavelengths of 355 nm, 532 nm, or 1064 nm) to reduce the risk of microcracks and edge chipping.

[0071] The processing strategy employs a "multiple thin-layer removal" method, cutting layer by layer. The removal depth of a single layer can be set to approximately 1–10 μm / layer (preferably 2–5 μm / layer) to avoid heat accumulation and crack propagation caused by a single deep cut. The scanning path can use staggered filling (hatch) and multi-directional cross scanning (e.g., 0° / 90° or 0° / 60° / 120°) to reduce anisotropic stress concentration. The process employs a partitioned machining approach, first machining the main area inside groove 11, and then gradually moving towards the contour boundary. Repetition frequency: 50kHz–1 MHz (higher for picoseconds / femtoseconds; mid-to-low frequencies are used for nanoseconds to reduce heat accumulation), scan speed: 50–2000 mm / s (linked to energy / frequency), line spacing: 5–30μm, trajectory overlap: 30%–80%, optional gas assistance: N2 / air knife purging to remove dust and reduce redeposition.

[0072] A machining allowance is reserved at the contour boundary of the groove 11, which can be about 10–80 μm, preferably 20–50 μm. That is, the shape of the groove 11 is "slightly smaller / slightly recessed" during the rough machining stage.

[0073] S130, low-energy fine-tuning to remove excess material; Low-energy fine-tuning is used to perform secondary processing on the contour boundary of groove 11 to remove the aforementioned reserved allowance, so that the contour of groove 11 reaches the design size and improves the edge quality. The fine-tuning strategy is as follows: the laser energy is reduced, the scanning speed is increased, and the single removal is thinner (e.g., 0.5–3 μm / layer); a combination of "contour tracing + light filling" is used: first perform 1–3 rounds of contour tracing, and then perform light filling trimming; smaller spot size or shorter wavelength can be used for fine-tuning to reduce edge chipping.

[0074] S140, Depth and Shape Detection and Correction; The depth of the groove 11, the flatness of the top surface of the annular protrusion 12, and the contour dimensions are inspected (e.g., by white light interferometry, profilometer, microscopy, or coordinate measuring machine).

[0075] If the depth of groove 11 does not reach the target, return to execute S120 and S130 for compensation processing; if the contour is out of tolerance, return to execute S130 for fine-tuning compensation.

[0076] S150, pre-rinse to remove large particles; Rinse with DI water spray / immersion (optionally adding a low concentration of surfactant to improve wettability), preferably using directional spraying to allow the liquid to enter the groove 11. Rinse for 30 s–3 min. Remove large particles and loose dust first.

[0077] S160, Cleaning removes fine particles; The groove 11 is cleaned by ultrasonic or mega-sound to remove the fine powder adhering to the bottom and corners of the groove 11.

[0078] Specific parameters are as follows: Ultrasonic: typically 20–80 kHz, time 1–10 min; Megasonic: typically 0.8–2 MHz, time 1–10 min. The bath solution is mainly DI water, and a weakly alkaline / neutral cleaning system can be selected. The temperature should be controlled (e.g., 20–50°C) to reduce residues.

[0079] Furthermore, during the ultrasonic / megasonic cleaning process, the cleaning fluid in the groove 11 is circulated or replaced by negative pressure suction to continuously refresh the liquid at the corners of the groove 11, removing the detached particles and preventing secondary deposition. Replacement methods can be selected as: circulation pump circulation, pulse suction, or vacuum-assisted extraction, with a circulation time of approximately 1–10 minutes. Simultaneously, online filtration (e.g., 0.2–1 μm filtration accuracy) is used to reduce backflow particles.

[0080] S170. Rinse and dry the ceramic substrate 1. After cleaning, rinse with DI water 1–3 times, and then dry. The drying method can be N2 blowing, rotary spin drying, vacuum drying or low temperature drying.

[0081] S180, Cleanliness Confirmation; Microscopic inspection or surface residue detection confirms that the particle residue at the bottom and corners of the groove 11 meets the requirements for subsequent metal layer 2 film formation, thereby reducing the risk of poor adhesion, local defects or discontinuities in the subsequent metal layer 2.

[0082] S190, the top surface of the annular protrusion 12 is treated with a non-repellent coating; The modification method can be chemical modification, physical coating, or deposition to form a surface energy reduction layer. To improve the consistency of modification, the surface can be activated before modification (e.g., plasma activation, ultraviolet ozone, etc.) to enhance the adhesion between the modified layer and the substrate surface. After modification, a gentle rinse (e.g., rapid rinsing with DI water / solvent) can be performed, followed by drying or baking for shaping. The baking temperature can be 80–120°C, and the baking time can be 10–30 min to remove unbonded residues and improve the stability of the modified layer.

[0083] By employing the aforementioned staged laser removal, contour pre-reservation, and low-energy fine-tuning processing methods, the risks of heat-affected zones, microcracks, and edge chipping can be reduced. Furthermore, the combination of pre-rinsing, ultrasonic / megasonic cleaning, and cyclic displacement enhances particle removal at the corners of the groove 11, providing a clean and morphologically stable substrate for the subsequent formation of the metal layer 2 within the groove 11. Additionally, surface modification of the annular protrusion 12 increases the contact angle of the colloid 4, thereby reducing the risk of colloid 4 overflow.

[0084] like Figure 5 and Figure 6 When using the dispensing method, step S700 specifically includes the following steps: S711. Determining and aligning the dispensing position; The dispensing position is determined using a vision positioning system of the dispensing equipment. The dispensing equipment can be a precision dispensing machine with a vision system or an integrated die bond and dispensing device, which includes a camera (upward-viewing camera / coaxial optical camera) and a motion platform.

[0085] Positioning reference: The outline of the annular protrusion 12, the boundary of the groove 11, and the solder joint or direction of the lead wire 3 are used as visual identification features to establish the dispensing coordinate system; Dispensing point: The dispensing point is set at a predetermined offset position on the top surface of the annular protrusion 12, close to the inner side of the annular protrusion 12 (i.e. offset inward relative to the center line of the annular protrusion 12), so that the adhesive 4 is preferentially spread inward to the inner side of the groove 11 and the risk of overflowing outward is reduced. Alignment method: The arched area of ​​each external lead 3 is located by visual recognition and program path (teach or CAD import) and the corresponding glue dispensing point is automatically generated.

[0086] S712. Dispensing application and dispensing volume control; Adhesive 4 is applied at the dispensing location using a precision dispensing valve. The dispensing valve can be a jet valve or a contact dispensing valve; when smaller adhesive dots are desired and needle contact with lead 3 is avoided, jet dispensing is preferred.

[0087] Type 4 colloids can be low-volatility or low-ionic potting / fixing adhesives, preferably with a certain degree of thixotropy to reduce flowability; Dispensing volume control method: Volume control is achieved through parameters such as dispensing valve opening time, pressure (or screw displacement), and spraying frequency, and can be calibrated by online weighing or visual measurement of dispensing dot diameter / height.

[0088] The specific range of dispensing volume can be approximately 0.01–0.20 μL for a single dispensing of adhesive 4, preferably 0.02–0.10 μL (the specific amount can be adjusted according to the number of leads 3, the target coverage length, and the viscosity of adhesive 4). The size of the adhesive dots can be: the spreading diameter of the adhesive dots on the top surface of the annular protrusion 12 can be 50–300 μm, preferably 100–200 μm.

[0089] S713. Repeat steps S711 and S712 until all external leads are fixed with 3-point glue. S714. Pre-cure colloid 4 in situ; A specific implementation method involves integrating a heating platform into the dispensing equipment. After dispensing, the workpiece is kept on the same platform for heat preservation, allowing the viscosity of the adhesive to increase and initially set. The pre-curing temperature can be selected from 60–120°C, and the pre-curing time can range from 1–10 minutes. Alternatively, infrared or hot air methods can be used to locally heat the dispensing area to achieve rapid setting.

[0090] S715, Perform final curing of colloid 4; After in-situ pre-curing, final curing can be performed in subsequent processes (e.g., combined with other heat treatment processes) to obtain the final colloid strength that meets the encapsulation reliability requirements. Final curing can also be performed in a separate oven or in an inline curing unit.

[0091] By applying adhesive to each lead 3 individually, the amount and location of the adhesive 4 at the arched point of each lead 3 can be precisely controlled. Compared to overall potting within the groove 11, this method reduces the amount of adhesive 4 used, improves application consistency, and lowers the risk of overflow, thus achieving more stable coverage and fixation of the arched points of the lead 3. This dispensing method is more suitable for chips 9 with fewer internal connecting leads 3 on the base island.

[0092] like Figure 7 and Figure 8 When using the full-area potting method, step S700 specifically includes the following steps: S721. Determination of the glue application area and height reference; The top surface of the annular protrusion 12 serves as the zero reference surface for height. The height of the top surface of the annular protrusion 12 is determined by the height sensing module or vision measurement module of the glue dispensing equipment. The height sensing module can be a laser displacement sensor, a confocal displacement sensor, a white light height measurement module, or other height measurement devices suitable for measuring minute heights. The target dispensing height is determined comprehensively based on the height of the highest point of lead 3, the preset height gap 13, and the rheological and wetting properties of the adhesive 4 material. The height of the highest point of lead 3 can be obtained through the arc detection function of the lead 3 bonding equipment or offline three-dimensional morphology measurement; the preset height gap 13 is the vertical distance between the highest point of lead 3 and the top surface of the annular protrusion 12. The material parameters of adhesive 4 may include viscosity, thixotropy, surface tension, and contact angle on the top surface of the annular protrusion 12, used to characterize the self-leveling, spreading, and wetting / crawling ability of adhesive 4 on lead 3 after dispensing is stopped. Based on these parameters, the adhesive surface height of adhesive 4 after dispensing is set higher than the top surface of the annular protrusion 12, so that adhesive 4 can cover the highest point of lead 3 under surface tension and wetting effects, and minimize the tendency to spread outwards from the annular protrusion 12. Based on the height of the highest point of the lead wire 3 and the preset height gap 13, the target glue filling height is set so that the surface height of the glue 4 is higher than the top surface of the annular protrusion 12 after glue filling, so that it can cover the highest point of the lead wire 3 under the action of surface tension.

[0093] S722, Fill the groove 11 with glue; Adhesive 4 is injected into the groove 11 using a dispensing device, which can be a pressure dispensing machine, a screw-type metering dispensing machine, or other precision dispensing equipment suitable for semiconductor packaging. During the dispensing process, the dispensing nozzle is positioned above the groove 11 and supplies adhesive along the center of the groove 11 area, allowing the adhesive 4 to gradually fill the groove 11 and cover the chip 9 and lead 3 within the groove 11. The type of adhesive 4 can be potting compound, fixing compound, or other low-stress encapsulating adhesive 4; the dispensing method can be continuous dispensing or segmented dispensing; the dispensing pressure, screw speed, or valve opening time are controlled to ensure a smooth dispensing process and avoid the generation of air bubbles or scouring of the lead 3. The volume of adhesive 4 is controlled so that the surface height of adhesive 4 reaches and is slightly higher than the top surface of the annular protrusion 12. After the dispensing stops, the adhesive 4 forms a stable adhesive surface within the area defined by the annular protrusion 12 under the action of surface tension and spreads towards the lead 3, thereby covering the arched part and the highest point of the lead 3.

[0094] S723, In-situ static setting and pre-curing; After the glue is poured, keep the substrate at the glue pouring station and allow the glue 4 to stand in place to allow it to self-level and release internal air bubbles. Then, without transferring the substrate, pre-cure or heat-insulate the glue 4 at the glue pouring station to increase its viscosity and pre-set its shape, reducing the risk of flow or overflow during subsequent transfer or injection molding. The specific in-situ pre-curing method is the same as in step S714.

[0095] S724. Perform final curing on colloid 4; The specific curing method is the same as step S715.

[0096] The adhesive 4 is injected into the groove 11 by the whole-body injection method, so that the adhesive 4 covers the highest point of the lead 3 under the action of surface tension. This not only fixes the highest point of the external connection lead 3, but also fixes the lead 3 located between different circuit unit connection areas 21 and the connection lead 3 between chips 9, which helps to improve the injection efficiency.

[0097] In addition, the thermal conductivity of colloid 4 is usually higher than that of injection molding packaging material. After overall injection molding, the thickness of the injection molding packaging layer above chip 9 is reduced accordingly. This allows the heat generated by chip 9 to be dissipated not only through the bottom of the substrate, but also through colloid 4 and the reduced thickness of the injection molding packaging layer to the top of the package, thereby improving the overall heat dissipation performance of the packaging structure.

[0098] like Figure 9 and Figure 10 In step S400, the following steps can be performed simultaneously while installing chip 9: S421, Process heat sink 5; A copper plate was selected as the heat sink 5. The heat sink 51 was processed by a combination of laser etching and chemical polishing. Subsequently, it was soaked in 5% dilute sulfuric acid solution at room temperature for 10s-60s to reduce the surface roughness of the heat sink 51 to Ra≤1μm. After etching, it was ultrasonically cleaned with deionized water and then dried to remove residual impurities.

[0099] S422, An expansion layer 6 is provided at the bottom of the heat dissipation slot 51; The expansion layer 6 is made of low-modulus silicone rubber modified with nano-calcium carbonate. The low-modulus silicone rubber with thermal expansion filler can expand more than twice its thickness in the thickness direction between 170℃ and 175℃. The low-modulus silicone rubber is coated on the bottom of the heat dissipation groove 51 by screen printing and scraping with a squeegee. Then it is placed in an oven for staged curing: first 30 minutes of pre-curing at 40℃, and then 45 minutes of final curing at 50℃. After curing, the Shore hardness of the expansion layer 6 is controlled at 15-20A.

[0100] S423. A sealing layer 7 is provided on the expansion layer 6; The thickness of the sealing layer 7 is greater than the depth of the heat dissipation groove 11. The bottom of the sealing layer 7 is filled inside the heat dissipation groove 11, while the remaining part protrudes outward from the heat dissipation groove 11. The expansion layer 6 is UV-curable fluorosilicone resin. The UV-curable fluorosilicone resin is applied to the surface of the heat sink 5 using a combination of drop-coating and vacuum spin-coating. Part of the UV-curable fluorosilicone resin flows into the heat dissipation groove 11 and is cured by UV irradiation. The irradiation parameters are: wavelength 365nm, power 800mW / cm², irradiation time 15s. After curing, the Shore hardness is 40-45D, and the adhesion is adjusted to 25-30mN / m. It can maintain a stable coefficient of thermal expansion at continuous operating temperatures of 200°C and 180°C, with 100% shape retention and no creep. Then, the excess UV-curable fluorosilicone resin of the heat sink 5 is removed by photolithography, leaving only the expansion layer 6.

[0101] S424, Use heat conductor 8 to connect to the back of heat sink 5; Flip the heat sink 5 so that its back side faces upward, and then bond the heat conductor 8 to the heat sink 5 using a thermally conductive adhesive. The heat conductor 8 is made of copper-tungsten alloy with a thermal conductivity of 200 W / (m·K) and a coefficient of thermal expansion that matches the silicon chip 9. In this embodiment, the thermally conductive adhesive is a high thermal conductivity silicone gel (thermal conductivity 5 W / (m·K)). The heat conductor 8 is fixed to the back side of the heat sink 5 using a pressure curing method, which ensures that the heat conductor 8 is fixed to the heat sink 5 while maintaining excellent thermal conductivity between the heat conductor 8 and the heat sink 5. Each heat sink 5 is connected to at least two heat conductors 8.

[0102] S425. Connect the heat conductor 8 to the ceramic substrate 1; The end of the heat conductor 8 away from the heat sink 5 is coated with a high thermal conductivity silicone gel. Then, the end of the heat conductor 8 is attached to the corresponding position of the ceramic substrate 1 so that the heat conductor 8 forms a tight contact with the ceramic substrate 1 in the vertical direction.

[0103] During the bonding process, the relative positions of the heat conductor 8 and the ceramic substrate 1 can be aligned and controlled to ensure an effective contact area between them and to prevent the heat conductor 8 from tilting or being partially suspended. After bonding, the heat conductor 8 and the ceramic substrate 1 are kept stationary, and the thermally conductive adhesive is subjected to heat preservation and curing treatment to stably fix the heat conductor 8 on the ceramic substrate 1. The heat preservation and curing can be carried out in a temperature range of approximately 60–150°C for several minutes to tens of minutes, such as 10–60 minutes, to allow the thermally conductive adhesive to cure completely and obtain stable bonding strength.

[0104] In step S800, during the heat preservation process of injection molding, the temperature is gradually increased from low to high to the required injection temperature of 170℃~175℃. Specifically, the mold is heated from room temperature to 80℃ at a rate of 5℃ / min and held for 10min, and then heated to 170-175℃ at a rate of 3℃ / min and held for 2.5min. This avoids uneven expansion caused by sudden local temperature rise and ensures that the expansion layer 6 fully expands to twice its own thickness at this temperature. The expansion layer 6 moves the sealing layer 7 toward the mold and abuts against and adheres to the inner wall of the mold.

[0105] After the heat preservation is completed, epoxy molding compound is injected into the mold cavity at an injection temperature of 175℃, an injection rate of 5mm / s, a holding pressure of 80MPa, and a holding time of 30s. Low speed and high pressure are used to avoid impacting the sealing layer 7. After injection molding, the mold is cooled and then demolded. Since the sealing layer 7 is adhered to the inner wall of the mold, it is not encased inside the molded body but is exposed on the surface of the molded body. Therefore, there is no need to polish the molded body, and the sealing layer 7 is exposed on the surface of the molded body.

[0106] Finally, remove the sealing layer 7 and the expansion layer 6, and invert the molded body into a container filled with professional silicone solvent. The professional silicone solvent is a 1:1 volume ratio of GN6333 and ELAPLUS215. Soak at room temperature for 1–2 hours to dissolve the sealing layer 7 and the expansion layer 6. Then clean the surface with isopropanol or anhydrous ethanol to ensure no solvent residue. After cleaning, the sealing layer 7 dissolves and leaves a heat dissipation channel on the molded body. After the expansion layer 6 is removed, the heat dissipation groove 11 is exposed on the top surface of the molded body.

[0107] By setting a heat conductor 8 to connect the bottom surface of the heat sink 5 with the bottom surface of the groove 11, the heat generated by the chip 9 can be transferred to the heat sink 5 via the heat conductor 8 in addition to being conducted downward through the ceramic substrate 1. The heat can then be further dissipated to the outside of the package through the heat sink 51, thereby constructing a multi-path heat dissipation channel under multi-core encapsulation conditions and improving the overall heat dissipation performance of the package structure. Example

[0108] like Figure 11 and Figure 12 This embodiment takes a 4mm x 4mm package with 7 built-in chips 9 as an example. The ceramic substrate has dimensions of 2500μm x 2500μm x 500μm, a groove depth of 245μm, a metal layer 2 thickness of 15μm, a chip 9 thickness of 200μm, a preset height of 20μm, a reserved spacing of 100μm, and a distance of 50μm between the highest point of the lead 3 and the chip 9. A metal layer 2 is provided on the back of the ceramic substrate to facilitate connection with the lead frame. The specific steps include: S110, Substrate pretreatment and positioning; S120, laser rough machining to form grooves, with allowance reserved at the contour boundary; The process employs a picosecond ultraviolet laser with a wavelength of 355 nm. During laser processing, the groove 11 is fabricated layer by layer using a multi-layer thin-layer removal method, with the single-layer removal depth controlled to approximately 3 μm / layer. The groove 11 is gradually formed to the target depth through repeated scanning, thus avoiding heat accumulation and crack propagation caused by a single deep cut. In this embodiment, the target depth of the groove 11 is 245 μm, corresponding to approximately 80–85 laser layer removals. The laser scanning uses a combination of staggered filling and multi-directional cross-scanning, with the scanning direction alternately set to 0° and 90° to reduce stress concentration in the ceramic substrate 1 in a single direction. Specific process parameters are set as follows: laser repetition frequency: 300 kHz; scanning speed: 600 mm / s; scanning line spacing: 15 μm; overlap rate of adjacent scanning trajectories: approximately 50%. During processing, an N2 air knife is used to assist in purging, removing dust generated during processing from the processing area in real time to reduce dust redeposition within the groove 11.

[0109] At the contour boundary of the groove 11, a partitioned processing strategy is adopted, that is, the main body area inside the groove 11 is processed first, and the laser energy density is reduced near the contour boundary to reserve processing allowance at the contour boundary. In this embodiment, the reserved allowance at the contour boundary is set to about 30μm, so that the contour of the groove 11 formed in the roughing stage is recessed inward relative to the design size, providing processing allowance for subsequent low-energy finishing processing.

[0110] S130, low-energy fine-tuning to remove excess material; The low-energy finishing process is still completed using a short-pulse laser, preferably the same 355nm picosecond laser as in step S120, to ensure processing consistency. In the finishing stage, by lowering the single-pulse energy and increasing the scanning speed, the energy density per unit area is significantly lower than that in the roughing stage, thereby reducing the thermal impact on the edge of the ceramic substrate 1.

[0111] The fine-tuning process employs a combination strategy of "outline stroking + light fill trimming," specifically including: Contour tracing: Multiple tracing scans are performed along the contour boundary of groove 11 to gradually remove the reserved contour allowance. In this embodiment, the number of contour tracing scans is set to 2, and the single-layer removal depth of each scan is controlled to be approximately 1 μm / layer, so as to achieve fine trimming of the contour boundary.

[0112] Light fill trimming: After completing the outline tracing, a light fill scan is performed on a local area near the outline boundary to eliminate any steps or discontinuities that may exist after tracing, making the transition between the outline edge and the bottom surface of the groove 11 smoother. The scan width of the light fill area is preferably 10–30 μm.

[0113] During the finishing process, the specific process parameters can be set as follows: laser repetition frequency: 300kHz; Scanning speed: 1200 mm / s; Scan line spacing: 8 μm; Trajectory overlap: approximately 60%; Single-layer removal depth: approximately 1 μm / layer. Simultaneously, during the finishing stage, a smaller spot diameter focusing condition can be selected to further improve the contour processing resolution and reduce the risk of edge chipping. N2 air knife-assisted purging is also used during processing to reduce the re-deposition of fine powder generated during finishing at the corners of groove 11.

[0114] S140, Depth and Shape Detection and Correction; S150, pre-rinse to remove large particles; Pre-rinsing is performed by a combination of DI water spraying and short-term immersion. The spray direction is preferably towards the opening of the groove 11 and along the depth of the groove 11 to promote the entry of the rinsing solution into the groove 11. To improve the wettability of the rinsing solution on the ceramic surface, a low concentration of surfactant can be added to the DI water.

[0115] The pre-rinsing time is controlled at about 1 minute to prioritize the removal of larger particles and loosely attached dust generated during laser processing, so as to avoid secondary impact or embedding of large particles on the corners of groove 11 during subsequent fine cleaning.

[0116] S160, Cleaning removes fine particles; The cleaning process is carried out using mega-sound cleaning, with the mega-sound frequency set to approximately 1 MHz and the cleaning time to approximately 5 minutes. The cleaning solution is mainly DI water, and a weakly alkaline or neutral cleaning system can be selected. The temperature of the cleaning solution is controlled at approximately 30–40°C to reduce particle residue and inhibit re-adhesion.

[0117] During the cleaning process, the cleaning fluid in the groove 11 is circulated and replaced. The fluid in the groove 11 is continuously renewed by the circulation pump, so that the stripped fine particles are promptly removed from the groove 11 area, and the particles are prevented from redepositing at the corners of the groove 11.

[0118] The circulation replacement time is synchronized with the megasonic cleaning process, approximately 5 minutes, and an online filtration unit is installed in the circulation loop with a filtration accuracy of approximately 0.5 μm to reduce particle backflow with the liquid.

[0119] S170, rinsing and drying; The ceramic substrate 1 is rinsed twice with DI water to remove residual cleaning solution or surfactant; then it is dried by a combination of N2 blowing and low-temperature drying, with the drying temperature controlled at about 100°C for about 15 minutes to avoid water residue and ensure that the inside of the groove 11 is fully dried.

[0120] S180, Cleanliness Confirmation; The particle residue at the bottom and corners of the groove 11 is confirmed by optical microscopy or surface residue detection to ensure that it meets the surface cleanliness requirements of the subsequent metal layer 2 formation process, thereby reducing the risk of poor adhesion, local defects or discontinuities in the metal layer 2.

[0121] S190, the top surface of the annular protrusion 12 is treated with a non-repellent coating; Before performing the surface gelling treatment, the ceramic substrate 1 is first subjected to surface activation treatment to improve the bonding stability between the modified layer and the substrate surface. Specifically, the ceramic substrate 1 is placed in a plasma treatment device, and oxygen plasma or air plasma is used to activate the top surface of the annular protrusion 12. The treatment power is set to about 75W and the treatment time is about 90s to remove surface organic residues and introduce active groups on the surface.

[0122] After surface activation, the top surface of the annular protrusion 12 undergoes a hydrophobic modification treatment. In this embodiment, the hydrophobic modification is achieved through chemical modification, by forming a low surface energy modified layer on the top surface of the annular protrusion 12 to reduce the wettability of the colloid 4 in this area. The modification treatment can be carried out in a gas phase, allowing the modified molecules to form stable adsorption or bonding on the top surface of the annular protrusion 12.

[0123] After the modification treatment is completed, the ceramic substrate 1 is gently rinsed to remove unbonded or excess modification residue. The preferred rinsing method is rapid rinsing with DI water to avoid damaging the modified layer with strong rinsing. Subsequently, the ceramic substrate 1 is dried or baked for shaping. In this embodiment, the baking temperature is set to approximately 100°C, and the baking time is approximately 20 minutes to further improve the stability and durability of the modified layer.

[0124] S200, A metal layer 2 is formed in the groove 11 of the ceramic substrate 1; In this embodiment, the metal layer 2 is prepared by a combination of seed layer formation and electroplating thickening process to improve the density, thickness uniformity and adhesion between the metal layer 2 and the ceramic substrate 1.

[0125] First, a metal seed layer is formed within the groove 11 of the ceramic substrate 1. In this embodiment, a continuous seed layer is formed on the bottom surface of the groove 11 using a physical vapor deposition (PVD) sputtering process. The seed layer can be a titanium / copper (Ti / Cu) composite layer, wherein the titanium layer is used to enhance the adhesion to the ceramic substrate 1, and the copper layer is used to provide a conductive path.

[0126] During the sputtering process, the uniformity of the metal seed layer coverage on the bottom surface of the groove 11 can be improved by adjusting the sputtering angle or by using a substrate rotation method.

[0127] After forming a metal seed layer, an electroplating process is used to deposit metal on the seed layer to form a metal layer 2 of the required thickness. In this embodiment, a copper electroplating process is used to thicken the deposition in the groove 11. During the electroplating process, the current density and electroplating time are controlled to ensure that the metal layer 2 forms a dense and uniformly thick structure on the bottom surface of the groove 11. The total thickness of the formed metal layer 2 is controlled to be approximately 15 μm.

[0128] S300, perform photolithography and etching on metal layer 2; The metal layer 2 is configured to form 8 independent circuit unit connection regions 21, the specific distribution of which is described in reference to [reference needed]. Figure 11 Meanwhile, a preset spacing 22 is maintained between adjacent circuit unit connection areas 21, and the circuit unit connection area 21 on which the chip 9 is mounted is processed into a chip 9 mounting area 211 and a lead wire connection area 212. The chip mounting area 211 and the lead wire connection area are also maintained with a preset spacing 22, which is 100μm.

[0129] S400, Install chip 9 to the corresponding circuit unit connection area 21; Chip 9 is a bare die with a gold backing. Before chip 9 is installed, the chip mounting area 211 in the circuit unit connection area 21 formed in step S300 is cleaned to remove any organic residues or particles. Subsequently, the bare die is mounted to the corresponding chip mounting area 211 through a die bonding process.

[0130] In this embodiment, chip 9 is fixed to metal layer 2 by conductive adhesive, specifically using conductive glue or silver paste as the conductive adhesive material. After the conductive adhesive material is dotted or printed on the chip mounting area 211, the bare die is mounted on it, and the conductive adhesive material is cured by heat curing, thereby achieving mechanical fixation while forming a stable electrical connection and heat conduction path.

[0131] During the chip 9 mounting process, the die position is precisely aligned by the alignment equipment so that the die is located in the predetermined position of the corresponding chip mounting area 211, thereby ensuring that there is a suitable spacing and spatial layout between the die pad and the corresponding lead connection area 212 for lead 3 bonding.

[0132] S500, Connect the substrate to the lead frame 3; In this embodiment, the substrate is a ceramic substrate with a metal layer 2 on its back side. A high thermal conductivity adhesive material is used to fix the substrate to a predetermined position on the lead frame. The high thermal conductivity adhesive material is preferably nano-silver. Through the sintering or curing of the nano-silver material, a stable mechanical fixing relationship is formed between the substrate and the lead frame, while establishing a low thermal resistance thermal conduction path to improve the overall heat dissipation performance of the packaging structure.

[0133] S600, perform wire bonding; During the bonding process of wires 3, the orientation and arc shape of wires 3 are controlled. Wires 3 are bonded using an arched arc shape, ensuring that the highest point of all wires 3 is located 5 μm from the center line of the annular protrusion 12 in the projection direction perpendicular to the substrate, and is positioned close to the inner periphery of the annular protrusion 12. Specifically, for connection wires 3 located inside the base island (e.g., interconnection wires 3 between chips 9), the orthographic projection of the highest point of its wire 3 is controlled within the area enclosed by the inner periphery of the annular protrusion 12.

[0134] S721. Determination of the glue application area and height reference; In this embodiment, the groove 11 serves as the glue-filling area, and the top surface of the annular protrusion 12 serves as the zero reference surface for the glue-filling height. The height of the top surface of the annular protrusion 12 is determined by the height measurement module of the glue-filling equipment.

[0135] The distance between the highest point of lead 3 and the surface of chip 9 is approximately 50 μm, and the preset height gap 13 between the highest point of lead 3 and the top surface of the annular protrusion 12 is 20 μm. The height of the highest point of lead 3 can be obtained by the arc detection function of the lead 3 bonding device, or confirmed by offline three-dimensional morphology measurement.

[0136] The target potting height is determined based on the following factors: the height of the highest point of the lead wire 3; the preset height gap 13; the viscosity, thixotropy, surface tension of the adhesive material 4, and its contact angle on the top surface of the annular protrusion 12 after the adhesive has been treated.

[0137] In this embodiment, after the glue is poured, the surface height of the glue 4 is set to be 18μm higher than the top surface of the annular protrusion 12, so that the glue 4 covers the arched part of the lead wire 3 upward under the action of surface tension and wetting and creeping glue, and extends to the highest point of the lead wire 3.

[0138] S722, Fill the groove 11 with glue; After the height reference is set, glue 4 is injected into the groove 11 using a glue dispensing device. The glue dispensing device can be a screw-type metering glue dispensing machine or a pressure glue dispensing machine, with the dispensing nozzle positioned above the groove 11 area and located in the center of the groove 11.

[0139] The colloid 4 is preferably a low-stress, high-thixotropic silicone potting compound with typical performance parameters as follows: room temperature viscosity: approximately 7000 mPa·s; thermal conductivity: approximately 2.0 W / (m·K); and a low elastic modulus after curing to reduce the stress impact on the lead wire 3 and the chip 9.

[0140] During the potting process, a continuous potting method is adopted. By controlling the screw speed or the potting pressure, the adhesive 4 gradually rises from the bottom of the groove 11 and fills the groove 11, thereby covering the chip 9 and lead wire 3 inside the groove 11.

[0141] By precisely controlling the volume of the colloid 4, the surface height of the colloid 4 is made to reach and slightly exceed the top surface of the annular protrusion 12 by 18 μm.

[0142] S723, In-situ static setting and pre-curing; After the potting process is completed, keep the substrate at the potting station and allow the adhesive 4 to stand in situ to allow it to self-level and release internal air bubbles. The standing time is 3 minutes. The in-situ pre-curing method is the same as in step S714, using a heating table integrated into the potting station for heat preservation at approximately 80°C for 10 minutes.

[0143] S724. Perform final curing on colloid 4; The subsequent processes are completed using an oven or online curing equipment.

[0144] S800, perform injection molding encapsulation. Example

[0145] Reference Figure 13 The difference between this embodiment and Embodiment 1 is that, in step S400, while chip 9 is being installed, a 150μm x 150μm x 300μm copper plate is used for processing to form a 150μm heat dissipation groove 51. The expansion layer 6 is 80μm thick, and the sealing layer 7 is 150μm thick; the heat conductor 8 is a copper-tungsten alloy cylinder with a radius of 25μm and a height of 200μm, and there are two heat conductors 8. After all the bare dies are completed, the heat conductors 8 are installed in the upper right corner area of ​​the ceramic substrate, and finally fixed by full potting. In step S800, after injection molding, the sealing layer 7 and the expansion layer 6 are removed, exposing the heat dissipation groove 11 on the top surface of the molding compound.

[0146] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A QFN multi-core encapsulation process, characterized in that: Includes the following steps: S100. A groove (11) is processed on the ceramic substrate (1) to form an annular protrusion (12) along its periphery. The depth of the groove (11) is controlled to be greater than the sum of the thicknesses of the chip (9) and the metal layer (2). After the lead wire (3) is bonded, a preset height gap (13) is formed between the highest point of the lead wire (3) and the top surface of the annular protrusion (12). The height gap (13) satisfies that the colloid (4) can cover the highest point of the lead wire (3) under the action of surface tension. S200, A metal layer (2) is formed in the groove (11) of the ceramic substrate (1); S300, Photolithography and etching are performed on the metal layer (2) to form at least two independent circuit unit connection regions (21); S400, Install the chip (9) to the corresponding circuit unit connection area (21); S500, Connect the substrate to the lead wire (3) frame; S600, perform lead bonding (3) so that the projection of the highest point of all leads (3) is located within the area defined by the annular protrusion (12); S700, Fix the arched part of the lead wire (3) with glue; S800, perform injection molding encapsulation.

2. The QFN multi-core encapsulation process according to claim 1, characterized in that: In step S100, the specific processing method of the groove (11) of the ceramic substrate (1) is as follows: the ceramic substrate (1) is removed by a staged laser processing method, a processing allowance is reserved at the contour boundary of the groove (11), and the contour boundary is trimmed by low-energy fine finishing processing. Clean the processed groove (11): pre-rinse the groove (11) to remove large particles; Subsequently, ultrasonic or mega-sound cleaning is performed to remove fine particles, and liquid circulation is implemented in the groove (11) during the cleaning process.

3. The QFN multi-core encapsulation process according to claim 1, characterized in that: Step S100 further includes the following steps: after the groove (11) is processed, the top surface of the annular protrusion (12) is modified, and the top surface of the annular protrusion (12) is subjected to surface modification treatment so that the top surface of the annular protrusion (12) is colloid-repellent to the colloid (4).

4. The QFN multi-core encapsulation process according to claim 1, characterized in that: In step S300, a preset spacing (22) is maintained between adjacent circuit unit connection areas (21); The circuit unit connection area (21) on which the chip (9) is installed is processed into a chip mounting area (211) and a lead wire connection area (212), and the chip mounting area (211) and the lead wire connection area are also controlled to have a preset distance (22). The preset spacing (22) can reduce electromagnetic coupling between adjacent circuit unit connection areas (21) and between chip mounting area (211) and lead connection area (212).

5. The QFN multi-core encapsulation process according to claim 1, characterized in that: In step S500, the substrate is connected to the lead wire (3) frame via nanosilver.

6. The QFN multi-core encapsulation process according to claim 5, characterized in that: In step S600, the control lead (3) forms a flat-top arc shape.

7. The QFN multi-core encapsulation process according to claim 1, characterized in that: In step S700, the specific glue injection method is as follows: control the glue nozzle to align with the lead wire (3) portion located above the annular protrusion (12), and control the glue injection position to be set close to the inner side of the annular protrusion (12); After determining the dispensing location, drip the adhesive (4) onto the top surface of the annular protrusion (12); Repeat the above operation to apply glue to each external lead (3) in turn.

8. The QFN multi-core encapsulation process according to claim 1, characterized in that: In step S700, the specific method of glue injection is as follows: glue (4) is injected into the groove (11) by overall glue injection, so that the glue (4) covers the chip (9) and lead wire (3) in the groove (11), and the surface of the glue (4) is controlled to be higher than the top surface of the annular protrusion (12), so that the glue (4) covers the highest point of the lead wire (3) under the action of surface tension. After the glue is applied, the glue (4) is heated and cured at the glue application station.

9. The QFN multi-core encapsulation process according to claim 1, characterized in that: In step S400, an additional heat sink (5) is used for heat dissipation. A heat dissipation groove (51) is processed on the heat dissipation surface of the heat sink (5). A sealing layer (7) is provided in the heat dissipation groove (51) so that the sealing layer (7) extends out of the heat dissipation groove (51). A heat conductor (8) is connected to the bottom surface of the heat sink (5) and the heat conductor (8) is connected to the bottom surface of the groove (11). In step S800, the upper injection mold is brought into contact with the sealing layer (7), and the sealing layer (7) is removed after injection molding is completed.

10. The QFN multi-core encapsulation process according to claim 9, characterized in that: In step S400, an expansion layer (6) is first set inside the heat dissipation groove (51), and then the sealing layer (7) is set on the expansion layer (6); In step S800, the expansion layer (6) and the sealing layer (7) are removed after injection molding is completed.

Images