A stress regulation based tgv manufacturing process

Abstract

The application discloses a kind of TGV manufacturing processes based on stress regulation, belong to TGV etching technical field, using picosecond laser Bessel beam to carry out multilayer coaxial modification to glass, and process auxiliary etching path inside modified profile to form composite modification structure;Again, through hot alkali solution combined with ultrasonic wave assisted wet etching, modified area is selectively removed to form through hole;Finally, through hole is aligned with high thermal expansion coefficient mold, and is finished by thermal expansion, so that glass microplasticity flows to repair hole wall and copy mold shape, then gradient cooling, and utilize thermal expansion difference natural demolding.The application uses the above-mentioned TGV manufacturing process based on stress regulation, significantly improves etching efficiency and inhibits micro load effect;At the same time, auxiliary etching path reduces etching defects from source, combined with the microplasticity flow of thermal expansion finishing heals microcrack, reduces roughness and copies mold shape, realizes high-quality hole wall of near-zero taper.

Description

Technical Field

[0001] This invention relates to the field of TGV etching technology, and in particular to a stress-controlled TGV manufacturing process. Background Technology

[0002] Through-Glass Via (TGV) technology, as a key interconnect technology in the field of 3D integrated packaging, has received widespread attention in recent years. TGV technology achieves high-density vertical interconnects between chips by creating through-holes in a glass substrate and filling them with conductive metal. It possesses excellent high-frequency electrical performance, an adjustable coefficient of thermal expansion, and the potential for large-size panel-level processing, and is considered one of the important technological routes for next-generation advanced packaging. With the continuous improvement of packaging integration, the demand for TGV manufacturing with high aspect ratios (depth-to-width ratio greater than 10:1), high hole wall quality, and high interface reliability is becoming increasingly urgent.

[0003] Currently, the manufacturing of TGV has the following defects: 1) High aspect ratio TGV etching efficiency is low, and micro-loading effect is severe: Traditional Bessel beam coaxial modification only processes cylindrical profiles, leaving an unmodified solid core. The etching solution can only etch from the top and bottom of the substrate towards the center. The core acts as a physical barrier to material exchange, preventing fresh etching solution from entering the depths of the hole and hindering the removal of reaction products. Deep hole etching inherently exhibits a micro-loading effect. Once the blind hole depth reaches a certain value, material exchange within the hole relies entirely on diffusion, creating an axial gradient in etching solution concentration. The concentration at the bottom is extremely low, leading to a sharp decrease in the axial etching rate. In the later stages of etching, the core material dissolves slowly, consuming a large amount of the limited bottom etching solution and further exacerbating the loading effect.

[0004] 2) The TGV via walls prepared by wet etching have inherent quality defects: During laser modification, the interaction between the picosecond laser and the glass generates a thermal effect, forming a heat-affected zone around the modified area, leading to microcracks and the formation of a recast layer. Wet etching is essentially isotropic etching; the hole's opening has the longest contact time with fresh etching solution, resulting in the most severe lateral etching. Conversely, the etching solution concentration is lower in the middle of the hole, leading to a slower lateral etching rate and creating a tapered shape that is wider at the top and narrower at the bottom. Microscopic loading effects cause uneven etching rates at different axial positions, further exacerbating the distortion of the hole wall profile. The longer the etching time, the more severe the lateral etching, and the greater the hole wall tapering and roughness.

[0005] 3) TGV metallization interfaces have low reliability and are prone to interface delamination and Cu coating peeling: The coefficients of thermal expansion of glass and copper differ by approximately 34 times. Under the thermal cycling effects of post-plating annealing, chip bonding, and device operation, the interface generates significant tensile stress. Microcracks, protrusions, and depressions in the hole walls cause stress concentration at these locations. When the stress exceeds the interfacial bonding strength, cracks form and propagate, ultimately leading to interfacial delamination. After wet etching, the glass surface retains a large number of Si-OH chemical functional groups and dangling bonds, making it prone to moisture absorption and the formation of a weak boundary layer. During subsequent electroplating, these locations easily form voids, significantly reducing the interfacial fracture toughness. At high temperatures, the axial expansion of copper causes uneven plastic slippage along the rough hole walls, leading to increased interfacial shear stress and further accelerating interfacial failure.

[0006] 4) Existing methods for relieving interfacial stress have inherent flaws: Polymer buffer layers require multiple processing steps, including coating, photolithography, and curing, which increases process complexity and manufacturing costs. Furthermore, they introduce two new interfaces: polymer-glass and Cu-polymer. However, the adhesion between polymers and glass / copper is limited, and polymers themselves have poor thermal stability, making them prone to aging at high temperatures, which actually increases the likelihood of interfacial failure. Polymer buffer layers can only alleviate stress; they cannot eliminate stress concentration sources such as microcracks and roughness in the pore walls, nor can they change the significant difference in thermal expansion coefficients between glass and copper. Summary of the Invention

[0007] The purpose of this invention is to provide a stress-controlled TGV manufacturing process that significantly improves etching efficiency and suppresses micro-loading effects through the synergistic effect of auxiliary etching paths and ultrasonic assistance. At the same time, the auxiliary etching path reduces etching defects at the source, and the microplastic flow of thermal expansion finishing heals microcracks, reduces roughness, and replicates the mold shape, achieving high-quality hole walls with near-zero taper. Thermal expansion finishing also eliminates stress concentration sources and chemically weak boundary layers, and introduces residual compressive stress during cooling to offset thermal mismatch tensile stress, thereby significantly improving the bonding strength and thermal cycling reliability of the metallization interface.

[0008] To achieve the above objectives, the present invention provides a stress-controlled TGV manufacturing process, comprising the following steps: Step 1: Bessel laser multi-layer coaxial modification: Picosecond laser Bessel beams are used to perform multi-layer coaxial modification inside the glass substrate. Auxiliary etching paths are processed inside the cylindrical modification contour to form a composite modification structure that penetrates the thickness of the substrate. Step 2, Ultrasonic-assisted hot alkaline solution wet etching: The modified glass substrate is placed in a hot alkaline solution, and wet etching is performed with ultrasonic assistance. By utilizing the etching selectivity ratio between modified and unmodified glass, the laser-modified area is selectively removed to form glass vias. Step 3, Vacuum thermal expansion finishing and cooling demolding: Align the etched glass through-hole with the cylindrical mold, heat it in a vacuum environment for thermal expansion finishing, causing the glass to produce micro-plastic flow, and then cool it. Utilize the difference in thermal expansion between the mold and the glass to achieve demolding and separation, and introduce residual compressive stress on the surface of the glass hole wall.

[0009] Preferably, in step one, the auxiliary etching path is a cross-shaped structure that penetrates the entire thickness of the glass.

[0010] Preferably, in step one, the glass substrate is one of fused silica glass, borosilicate glass, or soda-lime glass.

[0011] Preferably, in step one, the process parameters for multi-layer coaxial modification of the picosecond laser Bessel beam are: center wavelength 1030nm or 1060nm, pulse width 1-15ps, pulse repetition frequency 100kHz-1MHz, single pulse energy 1-20μJ, equivalent focal depth of the Bessel beam 0.5-10mm, number of coaxial modification layers 1-5 with inter-layer coaxiality deviation ≤1μm, and scanning rate 10-500mm / s.

[0012] Preferably, in step two, the hot alkaline solution is an aqueous solution of one or a mixture of potassium hydroxide and sodium hydroxide, with a mass concentration of 5-30 wt% and a temperature of 80-200℃; the ultrasonic-assisted frequency is 20-100 kHz, and the power density is 0.5-5 W / cm². 2 The etching time is 30-600 min.

[0013] Preferably, in step three, the coefficient of thermal expansion of the cylindrical mold material is 12 × 10⁻⁶. -6 / ℃-25×10 -6 At room temperature, the initial single-sided gap between the mold and the through hole is 1-8 μm, and the surface roughness Ra of the mold working surface is ≤5 nm.

[0014] Preferably, in step three, the vacuum level of the process environment is 1×10⁻⁶. -2 Pa-5×10 -4 Pa, heating rate of 2-15℃ / min, finishing and heat preservation temperature of 30-120℃ higher than glass transition temperature, heat preservation and shape preservation time of 5-60min.

[0015] Preferably, in step three, the cooling adopts a gradient cooling method: the cooling rate is 0.5-3℃ / min in the range above the glass transition temperature; and the cooling rate is 3-10℃ / min in the range below the glass transition temperature.

[0016] Preferably, in step three, the material of the column mold is selected from iron-nickel-chromium alloy, nickel-based high-temperature alloy, or iron-nickel alloy.

[0017] Therefore, the present invention employs the above-mentioned stress-controlled TGV manufacturing process, which has the following beneficial effects: (1) Significantly improves the etching efficiency of high aspect ratio TGV and effectively suppresses micro-loading effect: In the laser modification stage, the present invention introduces an auxiliary etching path, which allows the etching solution to quickly penetrate into the deep part of the substrate through the internal channel, realizes multi-point synchronous etching of the upper and lower end faces and the internal channel, shortens the etching time by more than 50%, and effectively suppresses the micro-loading effect.

[0018] (2) Eliminate inherent hole wall quality defects in wet etching TGV: This invention eliminates hole wall defects from two dimensions: reducing at the source and repairing later: the auxiliary etching path shortens the etching time to reduce lateral etching and taper, vacuum thermal expansion finishing heals microcracks and reduces roughness through glass microplastic flow, and replicates the precise shape of the mold to form a uniform through hole with near-zero taper.

[0019] (3) Improve the reliability of TGV metallization interface and eliminate interface delamination and Cu coating peeling: This invention eliminates stress concentration sources on the hole wall through thermal expansion smoothing, and high temperature hot pressing causes the glass surface to recrystallize to eliminate Si-OH functional groups and dangling bonds, thus solving the weak boundary layer problem; during cooling, residual compressive stress is introduced by the difference in thermal expansion between the mold and the glass to offset the thermal mismatch tensile stress, while smoothing the hole wall to avoid uneven plastic slip of copper, thus significantly improving the interface bonding strength and thermal cycle life.

[0020] (4) Avoiding the inherent defects of existing polymer buffer layer methods: This invention replaces the traditional polymer buffer layer with molding finishing and residual stress control, eliminating the need for complex processes such as coating, photolithography, and curing, avoiding the introduction of two new interfaces, polymer-glass and Cu-polymer, eliminating the problems of high-temperature aging and insufficient adhesion of polymer, fundamentally eliminating stress concentration sources and actively offsetting thermal mismatch stress.

[0021] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of Bessel laser multilayer coaxial modification based on stress-controlled TGV manufacturing process according to the present invention, wherein (a) is a Bessel laser multilayer coaxial modification operation diagram, and (b) is a schematic diagram of auxiliary etching path for coaxial modification processing. Figure 2 This is a schematic diagram of ultrasonic-assisted hot alkaline solution wet etching of a stress-controlled TGV manufacturing process according to the present invention, wherein (a) is an operation diagram of ultrasonic-assisted hot alkaline solution wet etching, and (b) is a result diagram of ultrasonic-assisted hot alkaline solution wet etching. Figure 3This is a schematic diagram of vacuum thermal expansion finishing and cooling demolding of a stress-controlled TGV manufacturing process according to the present invention, wherein (a) is a schematic diagram of vacuum thermal expansion finishing and (b) is a schematic diagram of the TGV obtained by cooling demolding. Detailed Implementation

[0023] This invention provides a stress-controlled TGV manufacturing process, comprising the following steps: Step 1: Bessel laser multilayer coaxial modification: such as Figure 1 As shown, a picosecond laser Bessel beam is used to perform multi-layer coaxial modification inside the glass substrate, and an auxiliary etching path is processed inside the cylindrical modification contour to form a composite modification structure that penetrates the thickness of the substrate.

[0024] The auxiliary etching path has a cross-shaped structure that runs through the entire thickness of the glass.

[0025] The glass substrate is one of fused silica glass, borosilicate glass, or soda-lime glass.

[0026] The process parameters for multi-layer coaxial modification of picosecond laser Bessel beams are as follows: center wavelength 1030nm or 1060nm, pulse width 1-15ps, pulse repetition frequency 100kHz-1MHz, single pulse energy 1-20μJ, equivalent focal depth of Bessel beam 0.5-10mm, number of coaxial modification layers 1-5 with interlayer coaxiality deviation ≤1μm, and scanning rate 10-500mm / s.

[0027] Step 2: Ultrasonic-assisted hot alkaline solution wet etching: (e.g.) Figure 2 As shown, the modified glass substrate is placed in a hot alkaline solution and wet etching is performed with ultrasonic assistance. By utilizing the etching selectivity ratio between the modified and unmodified glass, the laser-modified areas are selectively removed to form glass vias.

[0028] The hot alkaline solution is an aqueous solution of one or a mixture of potassium hydroxide and sodium hydroxide, with a mass concentration of 5-30 wt% and a temperature of 80-200℃; the ultrasonic-assisted frequency is 20-100 kHz, and the power density is 0.5-5 W / cm³. 2 The etching time is 30-600 min.

[0029] Step 3: Vacuum thermal expansion finishing and cooling demolding: (e.g.) Figure 3 As shown, the etched glass through-hole is aligned with the cylindrical mold, and thermal expansion is performed under vacuum to induce microplastic flow in the glass. Then, it is cooled, and the difference in thermal expansion between the mold and the glass is used to achieve demolding and separation, and residual compressive stress is introduced on the surface of the glass hole wall.

[0030] The coefficient of thermal expansion of the cylindrical mold material is 12×10⁻⁶. -6 / ℃-25×10-6 At room temperature, the initial single-sided gap between the mold and the through hole is 1-8 μm, and the surface roughness Ra of the mold working surface is ≤5 nm.

[0031] The process environment vacuum level is 1×10 -2 Pa-5×10 -4 Pa, heating rate of 2-15℃ / min, finishing and heat preservation temperature of 30-120℃ higher than glass transition temperature, heat preservation and shape preservation time of 5-60min.

[0032] The cooling adopts a gradient cooling method: the cooling rate is 0.5-3℃ / min in the range above the glass transition temperature; and the cooling rate is 3-10℃ / min in the range below the glass transition temperature.

[0033] The material for the column mold is selected from iron-nickel-chromium alloy, nickel-based high-temperature alloy or iron-nickel alloy.

[0034] First, this invention introduces an auxiliary etching path during the laser modification stage, upgrading the traditional solid cylindrical modification structure into a composite modification structure with internal open channels. This structure breaks down the material transport barrier formed by the solid core, allowing the etching solution to quickly penetrate deep into the substrate through the internal channels, achieving multi-point simultaneous etching of the upper and lower end faces and the internal channels. The auxiliary etching path increases the reaction contact area, accelerates the discharge of reaction products, fundamentally suppresses the micro-loading effect of deep hole etching, and reduces etching time by more than 50%.

[0035] Secondly, this invention addresses the hole wall quality issue from two dimensions: reducing defects at the source and repairing defects laterally. On the one hand, the auxiliary etching path significantly shortens the etching time, substantially reducing lateral etching and hole wall taper caused by isotropic etching. On the other hand, the subsequent vacuum thermal expansion finishing step, through the microplastic flow of glass under high temperature and pressure, further heals laser-induced microcracks and recast layers, reducing hole wall roughness from the micrometer level to the nanometer level. Simultaneously, the forced constraint of the high-precision mold precisely replicates the ideal cylindrical shape of the mold onto the glass hole wall, forming a near-zero taper uniform through-hole, completely eliminating the barrel distortion and flared-mouth defects inherent in wet etching.

[0036] Furthermore, this invention eliminates stress concentration sources such as microcracks, protrusions, and depressions in the hole walls through thermal expansion finishing, preventing stress concentration at these locations during thermal cycling. The high-temperature hot pressing process causes recrystallization on the glass surface, eliminating Si-OH functional groups and dangling bonds remaining from wet etching, fundamentally solving the weak boundary layer problem and significantly improving the bonding strength of the Cu-glass interface. During cooling, the difference in thermal expansion between the mold and the glass introduces controllable residual compressive stress into the hole wall surface, which can offset the tensile stress generated by thermal mismatch between the glass and copper during thermal cycling, keeping the interface under compressive stress at all times. In addition, the smooth hole walls prevent uneven plastic slippage of copper at high temperatures, further reducing interfacial shear stress.

[0037] Finally, this invention completely replaces the traditional polymer stress buffer layer through molding finishing and residual stress control technology. It eliminates the need for complex processes such as coating, photolithography, and curing, significantly simplifying the process and reducing manufacturing costs. Simultaneously, it avoids introducing two new failure interfaces—polymer-glass and Cu-polymer—eliminating problems such as high-temperature polymer aging and insufficient adhesion. Unlike polymer buffer layers that only alleviate stress, this invention fundamentally eliminates stress concentration sources and actively counteracts thermal mismatch stress through residual compressive stress, achieving a qualitative leap in interface reliability.

[0038] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention should be considered equivalent substitutions and are included within the protection scope of the present invention. Furthermore, it should be understood that after reading the contents of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims and are all within the protection scope of the present invention.

[0039] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0040] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.

[0041] In this invention, unless otherwise specified, all other test materials and instruments are conventional test materials in the field and can be purchased through commercial channels.

[0042] Example 1 This invention provides a stress-controlled TGV manufacturing process, using 6mm thick fused silica glass as the substrate, and includes the following steps: Step 1: Using a picosecond laser Bessel beam with a center wavelength of 1030nm, a pulse width of 5ps, a repetition frequency of 500kHz, and a single pulse energy of 12μJ, and an equivalent focal depth of 3.5mm, two layers of coaxial modification are performed, with the interlayer coaxiality deviation controlled within 1μm. The diameter of the modified cylinder is 80μm. At the same time, a cross-shaped auxiliary etching path is processed inside the cylinder, penetrating the entire thickness of the glass, with a scanning rate of 100mm / s.

[0043] Step 2: Place the modified glass substrate in a mixed aqueous solution of 30wt% potassium hydroxide and 5wt% sodium hydroxide at a solution temperature of 150℃, and simultaneously apply an electrical flux at a frequency of 28kHz and a power density of 2.5W / cm². 2 With ultrasonic assistance and an etching time of 150 minutes, the high etching selectivity between modified and unmodified glass is utilized to selectively remove the laser-modified areas and form glass vias.

[0044] Step 3: Select a nickel-based high-temperature alloy as the material for the cylindrical mold, with a coefficient of thermal expansion of 15 × 10⁻⁶. -6 At room temperature, the initial single-sided gap between the mold and the through hole is 5 μm, and the surface roughness Ra of the mold working surface is ≤5 nm. After precisely aligning the mold and the through hole, place it under a vacuum of 1×10⁻⁶. -3 In an environment of Pa, the glass is heated to 1150°C at a heating rate of 3°C / min (this temperature is about 50°C higher than the glass transition temperature of fused silica), and held at this temperature for 30 minutes to induce microplastic flow. Then, a gradient cooling method is used, cooling at a rate of 1°C / min in the range above the glass transition temperature and at a rate of 4°C / min in the range below the glass transition temperature to room temperature. Natural demolding is achieved by utilizing the difference in thermal expansion between the mold and the glass, and residual compressive stress is introduced into the surface layer of the glass hole wall.

[0045] Performance results: The etching time is reduced by 52% compared with the traditional process; the finished through hole has a depth-to-diameter ratio of 37.5:1, a taper of 0.04°, a hole wall roughness Ra of 5.8 nm, and no subsurface crack residue; the residual compressive stress of the hole wall is -95MPa; the interface bonding strength after copper plating is 58MPa, and there is no plating failure after 400 thermal cycles.

[0046] Comparative Example 1 The only difference between this comparative example and Example 1 is that the laser modification in this comparative example uses a cylindrical profile, without a cross-shaped auxiliary etching path, and does not include the operation in step three; all other conditions are the same.

[0047] Performance results: The etching time to complete penetration was 142 minutes, 137% longer than the previous example; the finished through-hole taper was 0.87°, the hole wall roughness Ra was 1.2 μm, and obvious subsurface microcracks were present. The interfacial bonding strength after copper plating was only 22 MPa, and the plating peeled off after 80 thermal cycles.

[0048] Comparative Example 2 The only difference between this comparative example and Example 1 is that step three is not included; all other conditions are the same.

[0049] Performance results: The etching efficiency was similar to that of the previous example, but the hole wall taper was 0.52°, the roughness Ra was 0.8μm, and the subsurface microcracks were not eliminated. The interfacial bonding strength after copper plating was 28MPa, and interfacial delamination occurred after 20 thermal cycles.

[0050] Comparative Example 3 The only difference between this comparative example and Example 1 is that the thermal expansion finishing temperature in step three is 1050°C, while all other conditions are the same.

[0051] Performance results: The glass did not undergo sufficient microplastic flow; the hole wall roughness Ra was 0.3 μm; and microcracks only partially healed. The through-hole taper was 0.31°, and there was no effective residual compressive stress on the hole wall. The interfacial bonding strength after copper plating was 35 MPa, and plating peeled off after 50 thermal cycles.

[0052] Comparative Example 4 The comparative example uses the same substrate as Example 1. The TGV manufacturing process uses traditional solid modification and wet etching. After coating the hole walls with a polyimide buffer layer, copper electroplating is completed.

[0053] Performance results: The interfacial bonding strength is 32MPa. After 100 thermal cycles, the polymer layer shows signs of aging and cracking, and the copper plating peels off along with the buffer layer. The process steps are increased by 5, and the manufacturing cost is increased by more than 80% compared to this solution.

[0054] Comparing Example 1 and Comparative Example 1, it can be seen that the difference between Comparative Example 1 and Example 1 is that the laser modification does not have a cross-assisted etching path and omits the thermal expansion finishing operation in step three. Its etching penetration time is 142 minutes, which is 137% longer than that of Example 1. The taper is as high as 0.87° and the roughness Ra is 1.2μm. There are obvious subsurface microcracks. The interface bonding strength is only 22MPa, and the coating peels off after 80 thermal cycles. In contrast, the etching time of Example 1 is significantly shortened. The taper is only 0.04°, the roughness Ra is 5.8nm, there are no microcracks, the interface bonding strength reaches 58MPa, and there is no failure after 400 thermal cycles. The above comparison proves that the auxiliary etching path is the key to significantly improving etching efficiency by breaking the material transport barrier of the solid core and achieving multi-point synchronous penetration of the etching solution; while thermal expansion finishing heals microcracks and replicates the precise shape of the mold through the microplastic flow of glass at high temperature, and introduces residual compressive stress by utilizing the difference in thermal expansion. It is the core step that determines the final hole wall quality and interface reliability. Both are indispensable.

[0055] Comparing Example 1 and Comparative Example 2, the only difference between Comparative Example 2 and Example 1 is that the thermal expansion finishing operation in step three is omitted. Although the etching efficiency of Comparative Example 2 is close to that of Example 1, the taper is still 0.52°, the roughness Ra is 0.8μm, the subsurface microcracks are not eliminated, the interface bonding strength is only 28MPa, and delamination occurs after 20 thermal cycles. In contrast, after thermal expansion finishing, the taper of Example 1 is reduced to 0.04°, the roughness Ra is reduced to 5.8nm, the microcracks are completely healed, the interface bonding strength is increased to 58MPa, and the thermal cycle life is extended to 400 cycles. This comparison proves that although the auxiliary etching path can partially improve the hole wall quality by shortening the etching time, the improvement is limited. The microcracks and chemical residues inherent in wet etching cannot be eliminated by shortening the etching time alone. Thermal expansion finishing is irreplaceable in eliminating hole wall defects, repairing weak chemical boundary layers, and introducing residual compressive stress.

[0056] Secondly, the thermal cycling life of Comparative Example 2 (20 cycles) was worse than that of Comparative Example 1 (80 cycles). The reason is that Comparative Example 2 has a large number of residual Si-OH functional groups and dangling bonds on the hole wall, resulting in a prominent weak boundary layer problem. In addition, the micron-level rough surface causes non-uniform plastic slip of copper at high temperature, which significantly increases the interfacial shear stress. The absence of step three means that the above-mentioned chemical and physical defects cannot be repaired, and there is no residual compressive stress protection. Although Comparative Example 1 also lacks step three, its excessive etching time leads to excessive corrosion of the hole wall, which to some extent washes away some of the surface active functional groups, resulting in a slightly longer thermal cycling life. This abnormal result proves that simply pursuing etching efficiency without interface modification cannot obtain a reliable metallization interface.

[0057] Comparing Example 1 and Comparative Example 3, the only difference between Comparative Example 3 and Example 1 is that the thermal expansion finishing temperature in step 3 is set to 1050°C, which is lower than the fused silica glass transition temperature (approximately 1100°C). As a result, the roughness only drops to Ra 0.3 μm, the microcracks only partially close physically and fail to heal completely, the taper is 0.31°, and there is no effective residual compressive stress on the hole wall. The interface bonding strength is only 35 MPa, and the coating peels off after 50 thermal cycles. In contrast, Example 1 sets the finishing temperature to 1150°C, which achieves complete healing of microcracks, a roughness reduction to Ra 5.8 nm, a taper close to zero, and successfully introduces a residual compressive stress of -95 MPa. This comparison demonstrates that the temperature conditions in the thermal expansion finishing process are critical: when the temperature is below the glass transition temperature, the glass remains brittle and cannot produce sufficient viscoplastic flow, thus failing to achieve complete healing of microcracks and high-precision contour replication, and also failing to establish an effective residual compressive stress field after cooling; only within an appropriate range above the glass transition temperature, the glass softens sufficiently and fits tightly against the mold, and a uniform and stable residual compressive stress can be formed due to the difference in thermal expansion during the cooling process, thereby achieving a qualitative improvement in interface reliability.

[0058] Comparing Example 1 and Comparative Example 4, it can be seen that Comparative Example 4, using a traditional polyimide buffer layer, has an interfacial bonding strength of only 32 MPa. After 100 thermal cycles, the polymer layer shows signs of aging and cracking, and the copper plating peels off along with the buffer layer. Furthermore, it adds five process steps, increasing manufacturing costs by over 80%. In contrast, Example 1 achieves an interfacial bonding strength of 58 MPa, exhibits no plating failure after 400 thermal cycles, and requires no additional coating, photolithography, or curing processes. This comparison demonstrates that traditional polymer solutions can only passively alleviate stress and cannot eliminate stress concentration sources such as microcracks and roughness in the hole walls. Moreover, it introduces two new interfaces—polymer-glass and Cu-polymer—which have inherent defects such as limited polymer adhesion and susceptibility to high-temperature aging and degradation. Example 1, through thermal expansion finishing, physically eliminates stress concentration sources and actively introduces residual compressive stress to offset thermal mismatch tensile stress, fundamentally solving the interfacial failure problem. It also significantly simplifies the process and reduces manufacturing costs, demonstrating outstanding overall advantages.

[0059] Therefore, this invention employs the aforementioned stress-controlled TGV manufacturing process, which significantly improves etching efficiency and suppresses micro-loading effects through the synergistic effect of auxiliary etching paths and ultrasonic assistance. Simultaneously, the auxiliary etching path reduces etching defects at the source, and combined with the microplastic flow of thermal expansion finishing to heal microcracks, reduce roughness, and replicate mold shape, it achieves high-quality hole walls with near-zero taper. Thermal expansion finishing also eliminates stress concentration sources and chemically weak boundary layers, and introduces residual compressive stress during cooling to offset thermal mismatch tensile stress, thereby significantly improving the bonding strength and thermal cycling reliability of the metallization interface.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A stress-controlled TGV manufacturing process, characterized in that: Includes the following steps: Step 1: Bessel laser multi-layer coaxial modification: Picosecond laser Bessel beams are used to perform multi-layer coaxial modification inside the glass substrate. Auxiliary etching paths are processed inside the cylindrical modification contour to form a composite modification structure that penetrates the thickness of the substrate. Step 2, Ultrasonic-assisted hot alkaline solution wet etching: The modified glass substrate is placed in a hot alkaline solution, and wet etching is performed with ultrasonic assistance. By utilizing the etching selectivity ratio between modified and unmodified glass, the laser-modified area is selectively removed to form glass vias. Step 3, Vacuum thermal expansion finishing and cooling demolding: Align the etched glass through-hole with the cylindrical mold, heat it in a vacuum environment for thermal expansion finishing, causing the glass to produce micro-plastic flow, and then cool it. Utilize the difference in thermal expansion between the mold and the glass to achieve demolding and separation, and introduce residual compressive stress on the surface of the glass hole wall.

2. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step one, the auxiliary etching path is a cross-shaped structure that runs through the entire thickness of the glass.

3. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step one, the glass substrate is one of fused silica glass, borosilicate glass, or soda-lime glass.

4. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step one, the process parameters for multi-layer coaxial modification of the picosecond laser Bessel beam are as follows: center wavelength 1030nm or 1060nm, pulse width 1-15ps, pulse repetition frequency 100kHz-1MHz, single pulse energy 1-20μJ, equivalent focal depth of Bessel beam 0.5-10mm, number of coaxial modification layers 1-5 with inter-layer coaxiality deviation ≤1μm, and scanning rate 10-500mm / s.

5. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step two, the hot alkaline solution is an aqueous solution of one or a mixture of potassium hydroxide and sodium hydroxide, with a mass concentration of 5-30 wt% and a temperature of 80-200℃; the ultrasonic-assisted frequency is 20-100 kHz, and the power density is 0.5-5 W / cm³. 2 The etching time is 30-600 min.

6. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step three, the coefficient of thermal expansion of the cylindrical mold material is 12 × 10⁻⁶. -6 / ℃-25×10 -6 At room temperature, the initial single-sided gap between the mold and the through hole is 1-8 μm, and the surface roughness Ra of the mold working surface is ≤5 nm.

7. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step three, the vacuum level of the process environment is 1×10⁻⁶. -2 Pa-5×10 -4 Pa, heating rate of 2-15℃ / min, finishing and heat preservation temperature of 30-120℃ higher than glass transition temperature, heat preservation and shape preservation time of 5-60min.

8. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step three, a gradient cooling method is adopted: the cooling rate is 0.5-3℃ / min in the range above the glass transition temperature; and the cooling rate is 3-10℃ / min in the range below the glass transition temperature.

9. The TGV manufacturing process based on stress regulation according to claim 1, characterized in that: In step three, the material for the column mold is selected from iron-nickel-chromium alloy, nickel-based high-temperature alloy, or iron-nickel alloy.

Images