A high thermal conductivity tungsten-copper composite heat dissipation substrate and its preparation method

By introducing a tungsten carbide transition layer and bimodal tungsten granules into the tungsten-copper composite heat dissipation substrate, and combining it with copper-based infiltrating material containing zirconium and boron, a continuous tungsten skeleton and copper network are formed. This solves the problem of organizational coordination of the tungsten-copper composite heat dissipation substrate in high-power packaging scenarios, achieves a balance between heat conduction, interface bonding and thermal matching, and improves the stability and manufacturability of the packaging system.

CN122033252BActive Publication Date: 2026-06-30YIXING CITY JITAI ELECTRONICS CO LTD

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
YIXING CITY JITAI ELECTRONICS CO LTD
Filing Date
2026-04-17
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Existing tungsten-copper composite heat dissipation substrates are difficult to simultaneously achieve the synergistic structure of the tungsten phase, copper phase, and interface layer in high-power packaging scenarios, resulting in uneven local heat transfer and stress concentration, which limits the stability and manufacturability of the packaging system.

Method used

A continuous tungsten framework is formed by using tungsten powder with a tungsten carbide transition layer on the surface and bimodal tungsten granules. Combined with copper-based infiltrating materials containing zirconium and boron, a continuous copper network is formed through pressing, sintering and infiltration, thereby optimizing the interfacial bonding stability and thermal matching.

Benefits of technology

It achieves a balance between thermal conductivity stability, interface bonding stability, and thermal matching capability of the high thermal conductivity tungsten-copper composite heat dissipation substrate in thin-plate scenarios, making it suitable for the assembly adaptability and long-term reliability requirements of high-power electronic devices.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122033252B_ABST
    Figure CN122033252B_ABST
Patent Text Reader

Abstract

This invention belongs to the field of electronic packaging heat dissipation materials, and provides a high thermal conductivity tungsten-copper composite heat dissipation substrate and its preparation method. By constructing a continuous tungsten skeleton, a continuous copper network and a tungsten carbide transition layer, and by using the synergistic design of tungsten powder with a tungsten carbide transition layer on the surface, bimodal tungsten granules, tungsten preforms and copper-based infiltrating materials containing zirconium and boron, the interfacial bonding stability and infiltration uniformity are improved. This makes the heat dissipation substrate have high thermal conductivity, suitable coefficient of linear expansion, high relative density and low residual open porosity. It solves the problems of balancing the strength of tungsten preforms with sufficient infiltration, interfacial bonding with low thermal resistance, and thin plate processing with thermal matching. It is suitable for heat dissipation packaging of high-power electronic devices.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of heat dissipation materials for electronic packaging, specifically to a high thermal conductivity tungsten-copper composite heat dissipation substrate and its preparation method. Background Technology

[0002] As high-power electronic devices, microwave modules, power semiconductor packaging units, and high heat flux density components evolve towards miniaturization, high integration, and long-term reliable operation, the role of heat dissipation substrates in the entire packaging system has gradually shifted from simple thermal support to a crucial foundational material that balances heat conduction, thermal matching, structural support, and processing adaptability. The tungsten-copper system possesses complementary characteristics of both tungsten and copper phases. The tungsten phase helps maintain framework stability and control thermal expansion, while the copper phase facilitates the formation of continuous heat transfer paths and improves subsequent processing adaptability. Therefore, this system has clear application value in high-power packaging scenarios. If the framework support is insufficient, the copper interconnects are inadequate, or the interface transition is unstable, a chain reaction of problems such as uneven local heat transfer and stress concentration may occur during the material preparation, leveling, grinding, and service stages, thereby limiting the stability and manufacturability of the packaging system. Therefore, constructing a tungsten-copper composite heat dissipation substrate that is suitable for both efficient heat transfer and thin-plate processing, based on the synergistic organization of the tungsten phase, copper phase, and interface layer, has clear engineering significance and application requirements.

[0003] Existing tungsten-copper composite heat dissipation technologies mainly revolve around tungsten framework construction, copper phase introduction, and improved interface bonding. However, different technical approaches often emphasize only one aspect, making it difficult to simultaneously achieve framework strength, sufficient infiltration, and interface stability in heat dissipation substrate scenarios. For example, Chinese patent CN111041318A discloses a tungsten-copper alloy and its preparation method. This solution mainly relies on hierarchical structure control to achieve performance improvement, with relatively limited involvement in the overall synergy between the continuous tungsten framework and the continuous copper network inside the substrate. Another example is Chinese patent CN120082783A, which discloses a tungsten framework copper-infiltrated billet, its preparation method, and its application. This solution helps improve densification and copper phase distribution, but there is still room for further optimization in achieving a comprehensive balance between interface thermal resistance, thermal expansion matching, and subsequent precision processing adaptability under thin-plate heat dissipation substrate conditions. Therefore, how to establish a more stable, continuous, and suitable microstructure for thin-plate applications among the tungsten phase, copper phase, and interface transition layer remains a technical problem to be solved in this field. Summary of the Invention

[0004] The purpose of this invention is to provide a high thermal conductivity tungsten-copper composite heat dissipation substrate and its preparation method, which solves the current pain point problem of difficulty in balancing the high strength of tungsten preforms with sufficient copper phase infiltration, high bending strength with low interfacial thermal resistance, and the precision machining window of thin plates with low thermal expansion thermal matching.

[0005] This invention improves the skeleton stability, pore structure connectivity and copper phase infiltration path of tungsten preforms by synergistic design of tungsten powder with a tungsten carbide transition layer on the surface, bimodal tungsten granules, continuous tungsten skeleton and copper-based infiltrator containing zirconium and boron. On the other hand, it enhances the interfacial bonding stability between the tungsten phase and the copper phase, so that thermal conductivity, thermal matching and thin plate processing adaptability are balanced and improved within the same system.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A high thermal conductivity tungsten-copper composite heat dissipation substrate includes 60-85 wt% tungsten phase and the balance copper phase based on the total mass of the heat dissipation substrate, wherein the copper phase is 15-40 wt% based on the total mass of the heat dissipation substrate; the tungsten phase forms a continuous tungsten skeleton, and the copper phase fills the spaces between the tungsten skeleton to form a continuous copper network; a tungsten carbide transition layer with an average thickness of 2-10 nm is provided between the tungsten phase and the copper phase.

[0008] The tungsten framework comprises a first tungsten region with a characteristic size of 2-12 μm and a second tungsten region with a characteristic size of 20-80 μm; the copper phase contains 0.05-0.30 wt% zirconium and 0.02-0.10 wt% boron by weight; the relative density of the heat dissipation substrate is 98.5-99.8%.

[0009] The continuous tungsten framework is derived from a tungsten preform formed by pressing and sintering a bimodal tungsten granulation material made from tungsten powder with a tungsten carbide transition layer on its surface and tungsten powder with a particle size D50 of 15-45 μm. The continuous copper network is formed by impregnating the tungsten preform with a copper-based infiltrate containing zirconium and boron.

[0010] Furthermore, the tungsten powder with a tungsten carbide transition layer on its surface is prepared through the following steps:

[0011] A1. Add 100 parts by weight of tungsten powder with a particle size D50 of 1-8 μm to a solution containing 0.5-3.0 parts by weight of sucrose and 20-60 parts by weight of deionized water, and stir for 0.5-2.0 h;

[0012] A2. Dry at 60-90℃ for 4-10 hours to obtain the coated precursor;

[0013] A3. In a mixed atmosphere of hydrogen and argon, heat to 780-950℃ at a rate of 3-10℃ / min and hold at that temperature for 0.5-2.0h;

[0014] A4. Cooling under argon protection yields tungsten powder with a tungsten carbide transition layer on the surface;

[0015] The tungsten powder with a tungsten carbide transition layer on its surface has a total carbon content of 0.03-0.25 wt% and an oxygen content of 200-800 ppm.

[0016] Furthermore, the bimodal tungsten granulation material is prepared through the following steps:

[0017] B1. Mix 15-40 parts by weight of tungsten powder with a tungsten carbide transition layer on the surface with the remainder of tungsten powder with a particle size D50 of 15-45 μm;

[0018] B2. After adding 1-3 parts by weight of polyvinyl alcohol and 20-40 parts by weight of deionized water to prepare the adhesive solution, knead for 0.5-2.0 hours.

[0019] B3. After granulation through a 20-60 mesh sieve, dry at 60-80℃ for 4-8 hours to obtain bimodal tungsten granules with a particle size of 250-850μm and a moisture content of no more than 0.5wt%.

[0020] Furthermore, the tungsten preform is prepared through the following steps:

[0021] C1. Press the bimodal tungsten granules at 150-300MPa to obtain a green body;

[0022] C2. Heat the green billet to 1150-1350℃ in a hydrogen atmosphere at a rate of 5-15℃ / min and hold for 30-90min.

[0023] C3. After cooling, a tungsten preform with a bimodal pore structure of 2-10 μm and 20-80 μm is obtained.

[0024] Furthermore, the zirconium- and boron-containing copper-based infiltration material is prepared through the following steps:

[0025] D1. Based on a total of 100 parts by weight of copper powder, zirconium powder, and boron powder, 0.05-0.30 parts by weight of zirconium powder, 0.02-0.10 parts by weight of boron powder, and the balance of copper powder are loaded into a vacuum induction melting device.

[0026] D2. Under vacuum or argon protection, heat to 1250-1350℃ and hold for 10-40 minutes to melt the zirconium and boron into the copper liquid.

[0027] D3. Cast the melt into copper-based infiltrate ingots or atomize it into copper-based infiltrate powder;

[0028] D4. The oxygen content in the copper-based infiltration material is not higher than 500 ppm, and the total content of zirconium and boron is 0.07-0.40 wt%.

[0029] Furthermore, the volume ratio of the first tungsten region to the second tungsten region is 15:85 to 40:60, the thickness of the heat dissipation substrate is 0.3-3.0 mm, and the flatness is 0.02-0.08 mm / 100 mm.

[0030] Furthermore, the thermal conductivity of the heat dissipation substrate at room temperature is 232-275 W / m·K, and the coefficient of linear expansion in the range of 25-300℃ is 5.8 × 10⁻⁶. -6 / ℃ to 7.2×10 -6 / ℃, the total oxygen content of the heat dissipation substrate is 300-800ppm, and the residual open porosity is not higher than 1.5vol.

[0031] As a concept of this invention, the microstructure design employing a continuous tungsten framework, a continuous copper network, and a tungsten carbide transition layer is primarily used to enhance the thermal conductivity stability, interfacial bonding stability, and thermal matching capability of the high thermal conductivity tungsten-copper composite heat dissipation substrate. The combination of the first and second tungsten regions helps to maintain a suitable spatial structure for copper phase entry and communication while ensuring the strength of the tungsten preform, resulting in a more uniform infiltration process. The tungsten carbide transition layer, positioned between the tungsten and copper phases, helps improve interfacial continuity and mitigates the adverse effects of interfacial mismatch. The zirconium- and boron-containing copper phase, in conjunction with the tungsten framework and interfacial layer, further promotes copper phase filling, interfacial bonding, and microstructure stability. Therefore, the heat dissipation substrate, even in thin-plate applications, can balance thermal conductivity, low thermal resistance, low thermal expansion, and adaptability to subsequent processing.

[0032] This invention also discloses a method for preparing a high thermal conductivity tungsten-copper composite heat dissipation substrate, comprising the following steps:

[0033] S1. Provide a tungsten preform and a copper-based infiltrate containing zirconium and boron, wherein the tungsten preform is formed by pressing and sintering a bimodal tungsten granule obtained by pressing tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 15-45μm.

[0034] S2. Place the zirconium- and boron-containing copper-based infiltration material above or around the tungsten preform;

[0035] S3. Heat to 1180-1280℃ in a vacuum or argon atmosphere and hold for 5-20 minutes to allow the zirconium- and boron-containing copper-based infiltrate into the tungsten preform.

[0036] S4. Cool under vacuum or inert atmosphere protection to obtain tungsten-copper composite heat dissipation substrate.

[0037] Furthermore, the tungsten preform provided in step S1 is obtained by pressing and sintering at 150-300 MPa, with a dimensional shrinkage rate of 4-12%; the zirconium- and boron-containing copper-based infiltration material provided in step S1 is an infiltration ingot or infiltration powder, with a particle size D50 of 20-80 μm; the vacuum degree before heating in step S3 is 1×10⁻⁶. -8 MPa to 1×10 -5 MPa, or the oxygen content in the argon atmosphere is not higher than 100 ppm.

[0038] Furthermore, after step S4, at least one of the following post-processing steps is included: annealing at 650-800℃ for 1-3 hours; light rolling and leveling with a reduction rate of 5-20%; and double-sided grinding to make the thickness tolerance of the obtained heat dissipation substrate ±0.05mm and the surface roughness Ra 0.2-1.0μm.

[0039] Furthermore, step A1 employs mechanical stirring or magnetic stirring.

[0040] Furthermore, in step A3, the mixed atmosphere of hydrogen and argon is kept constant during the heating and holding processes.

[0041] Furthermore, in step A4, the furnace is cooled to room temperature using either in-furnace cooling or programmed cooling.

[0042] Furthermore, the tungsten carbide transition layer includes a W2C phase and / or a WC phase, the phase composition of which is characterized by X-ray diffraction and / or selected area electron diffraction, and the layer thickness is characterized by transmission electron microscopy.

[0043] Furthermore, in step B2, the mass fraction of polyvinyl alcohol in the mixture is determined according to a statistical standard consistent with the amounts of polyvinyl alcohol, tungsten powder, and deionized water.

[0044] Furthermore, in step B2, the degree of hydrolysis and / or molecular weight of polyvinyl alcohol are specified, and step B2 is performed by mixing using a mixer.

[0045] Furthermore, step C2 is carried out under a hydrogen atmosphere.

[0046] Furthermore, after sintering in step C2, the mixture is cooled under the protection of hydrogen or argon.

[0047] Furthermore, the open porosity of the tungsten preform is set according to the target copper phase content and the amount of copper-based infiltrate.

[0048] Furthermore, the vacuum level or oxygen content in the protective argon gas in step D2 is limited during the heating and holding processes.

[0049] Furthermore, step D3 can cast the melt into copper-based ingots or atomize it into copper-based ingot powder.

[0050] Furthermore, in step S2, the amount of copper-based infiltrate relative to the tungsten preform is determined based on the theoretical filling amount of the pore volume of the tungsten preform.

[0051] Furthermore, step S3 is performed under a vacuum or argon atmosphere.

[0052] Furthermore, step S4 involves cooling under vacuum or an inert atmosphere.

[0053] Furthermore, the post-processing annealing is carried out under vacuum or inert atmosphere protection.

[0054] As another concept of this invention, the preparation design combining the pre-construction of a tungsten preform with the directional infiltration of a copper-based infiltrate containing zirconium and boron is mainly used to enhance the microstructure controllability, infiltration sufficiency, and dimensional stability of high thermal conductivity tungsten-copper composite heat dissipation substrates. First, a tungsten preform is obtained by forming a bimodal tungsten granule with tungsten powder having a tungsten carbide transition layer on its surface and tungsten powder of a specific particle size, followed by sintering. This facilitates the pre-establishment of a suitable pore structure and framework support for copper phase entry. Then, by limiting the composition, infiltration temperature, holding time, and atmosphere of the copper-based infiltrate containing zirconium and boron, the copper phase is made to enter the interior of the tungsten preform more uniformly and cooperate with the existing interface structure. If necessary, annealing, light rolling leveling, and double-sided grinding are performed to improve the thickness control, surface condition, and subsequent packaging adaptability of the thin-plate heat dissipation substrate.

[0055] In zirconium- and boron-containing copper-based infiltrates, zirconium primarily improves the interfacial bonding stability between the copper and tungsten phases and helps regulate the microstructure at the interface, ensuring good connectivity and stability of the infiltrated copper phase within the continuous copper network. Boron, on the other hand, focuses on facilitating melt flow and interfacial reaction processes, promoting more thorough filling of the pore structure of the tungsten preform by the copper-based infiltrate, and reducing microstructure inhomogeneity caused by insufficient local infiltration. When both work together, it benefits the formation and maintenance of the continuous copper network and the stability of the interfacial state near the tungsten carbide transition layer, achieving a synergistic balance between heat conduction paths, interfacial bonding, and thermal matching within the same system.

[0056] Beneficial technical effects

[0057] 1. Through the combination of continuous tungsten skeleton, continuous copper network and tungsten carbide transition layer, a relatively continuous heat conduction path is formed inside the material, and the interface continuity between tungsten phase and copper phase is improved, which helps to reduce the interface heat transfer barrier and enable the heat dissipation substrate to maintain a good balance between heat conduction and thermal matching.

[0058] 2. By forming bimodal tungsten granules with tungsten powder having a tungsten carbide transition layer on the surface and tungsten powder of a specific particle size, and further obtaining tungsten preforms, it is helpful to balance the structural strength and pore connectivity of the tungsten preforms, providing a stable skeleton and accessible channels for subsequent copper-based infiltration.

[0059] 3. By introducing zirconium and boron into the copper phase and cooperating with the tungsten preform and tungsten carbide transition layer, it is helpful to improve the uniformity of copper phase infiltration and the stability of interfacial bonding, and reduce the adverse effects of local microstructure inhomogeneity on thermal conductivity, flatness and processing quality.

[0060] 4. By limiting the thickness, flatness, residual opening porosity, and post-processing conditions of the heat dissipation substrate, the present invention is more suitable for forming a thin, flattenable, and grindable heat dissipation substrate, which is beneficial to meeting the assembly adaptability and long-term reliability requirements in the packaging of high-power electronic devices. Attached Figure Description

[0061] Figure 1 The X-ray diffraction patterns are those of Example 1, Comparative Example 1, and Comparative Example 7.

[0062] Figure 2 The images show the particle size difference distribution diagrams obtained by scanning electron microscopy for Example 1 and Comparative Example 2.

[0063] Figure 3 The image shows the macroscopic morphology of the high thermal conductivity tungsten-copper composite heat dissipation substrate prepared in Example 1.

[0064] Figure 4 The image shows the EDS elemental distribution of the high thermal conductivity tungsten-copper composite heat dissipation substrate prepared in Example 1. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0066] Example 1

[0067] This embodiment provides a high thermal conductivity tungsten-copper composite heat dissipation substrate, comprising 72 wt% tungsten phase and 28 wt% copper phase based on the total mass of the heat dissipation substrate in this embodiment; the tungsten phase forms a continuous tungsten skeleton, and the copper phase fills the spaces between the tungsten skeleton to form a continuous copper network; a tungsten carbide transition layer with a thickness of 5 nm is provided between the tungsten phase and the copper phase in this embodiment.

[0068] In this embodiment, the tungsten framework includes a first tungsten region with a characteristic size of 6 μm and a second tungsten region with a characteristic size of 45 μm; the copper phase in this embodiment contains 0.15 wt% zirconium and 0.05 wt% boron by weight of copper phase; the relative density of the heat dissipation substrate in this embodiment is 99.2%.

[0069] In this embodiment, the continuous tungsten framework is derived from a tungsten preform formed by pressing and sintering a bimodal tungsten granulation material made from tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 28 μm. In this embodiment, the continuous copper network is formed by impregnating the tungsten preform of this embodiment with a copper-based infiltrate containing zirconium and boron.

[0070] In this embodiment, tungsten powder with a tungsten carbide transition layer on its surface is prepared through the following steps:

[0071] A1. Add 100 parts by weight of tungsten powder with a particle size D50 of 4μm to a solution containing 1.5 parts by weight of sucrose and 35 parts by weight of deionized water, and mechanically stir for 1.0 h at a stirring speed of 300 rpm and a stirring temperature of 25℃.

[0072] A2. Dry at 75℃ for 6 hours to obtain the coated precursor;

[0073] A3. In a mixed atmosphere of hydrogen and argon (hydrogen to argon volume ratio of 3:1, total flow rate of 2L / min), the temperature is increased to 850℃ at 6℃ / min and held for 1.0h.

[0074] A4. Under argon protection, the temperature is reduced to room temperature at a programmed rate of 5℃ / min to obtain tungsten powder with a tungsten carbide transition layer on the surface.

[0075] In this embodiment, the tungsten powder with a tungsten carbide transition layer on its surface has a total carbon content of 0.12 wt% and an oxygen content of 450 ppm. The tungsten carbide transition layer in this embodiment includes a W2C phase and a WC phase, the phase composition of which is characterized by X-ray diffraction and selected area electron diffraction, and the layer thickness is characterized by transmission electron microscopy as 5 nm.

[0076] The bimodal particle size tungsten granules in this embodiment are prepared through the following steps:

[0077] B1. Mix 25 parts by mass of tungsten powder with a tungsten carbide transition layer on the surface with 75 parts by mass of tungsten powder with a particle size D50 of 28 μm;

[0078] B2. After adding 2 parts by mass of polyvinyl alcohol with a degree of alcoholysis of 88% and a degree of polymerization of 1750 and 28 parts by mass of deionized water to prepare a binder, mix it in a planetary mixer at 40°C for 1.0 h.

[0079] B3. After granulation through a 40-mesh sieve, the material is dried at 70℃ for 6 hours to obtain bimodal tungsten granules with a particle size of 450μm and a moisture content of 0.3wt%.

[0080] The tungsten preform in this embodiment is prepared through the following steps:

[0081] C1. Press the bimodal tungsten granules at 220 MPa to obtain a green body;

[0082] C2. The green compact is heated to 1250°C at a rate of 10°C / min and held at that temperature for 60 min in a hydrogen atmosphere with a purity of 99.99%.

[0083] C3. After cooling to 200°C under hydrogen protection, the tungsten preform with a bimodal pore structure of 6μm and 45μm is obtained.

[0084] The copper-based infiltration material containing zirconium and boron in this embodiment is prepared through the following steps:

[0085] D1. Based on a total of 100 parts by weight of copper powder, zirconium powder, and boron powder, 0.15 parts by weight of zirconium powder, 0.05 parts by weight of boron powder, and 99.80 parts by weight of copper powder are loaded into a vacuum induction melting device.

[0086] D2, under a vacuum degree of 5×10 -3 The temperature was raised to 1300℃ at MPa and held for 25 minutes to melt zirconium and boron into the copper liquid in this embodiment.

[0087] D3. The molten material is poured into a graphite mold to obtain a copper-based ingot with a casting temperature of 1250℃.

[0088] D4. In this embodiment, the oxygen content in the copper-based infiltration material is 300 ppm, and the total content of zirconium and boron is 0.20 wt%.

[0089] The preparation method of the high thermal conductivity tungsten-copper composite heat dissipation substrate in this embodiment includes the following steps:

[0090] S1. A tungsten preform and a copper-based infiltration material containing zirconium and boron are provided. In this embodiment, the tungsten preform is formed by pressing and sintering a bimodal tungsten granule material obtained by pressing tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 28 μm. In this embodiment, the tungsten preform is obtained by pressing and sintering at 220 MPa, and its dimensional shrinkage rate is 7.5%. In this embodiment, the copper-based infiltration material containing zirconium and boron is an infiltration ingot.

[0091] S2. Place the copper-based infiltration material containing zirconium and boron in this embodiment above the tungsten preform in this embodiment. The amount of copper-based infiltration material is determined according to the theoretical filling amount of 105% of the pore volume of the tungsten preform.

[0092] S3. Under a vacuum atmosphere, heat to 1230℃ at a rate of 8℃ / min and hold for 12min. The vacuum degree during the heating process is 5×10⁻⁶. -6 MPa, so that the copper-based infiltrate containing zirconium and boron in this embodiment is infiltrated into the tungsten preform of this embodiment;

[0093] S4. Cool to room temperature at 6℃ / min under vacuum protection to obtain tungsten-copper composite heat dissipation substrate.

[0094] Following S4, the following post-processing is also included: at a vacuum level of 1×10⁻⁶ -4Annealing at 720℃ for 2 hours under MPa conditions; light rolling and leveling with a reduction rate of 10%; double-sided grinding to obtain a heat dissipation substrate with a thickness of 1.5 mm, a thickness tolerance of ±0.05 mm, and a surface roughness Ra of 0.5 μm.

[0095] In this embodiment, the volume ratio of the first tungsten region to the second tungsten region is 25:75. The thickness of the heat dissipation substrate in this embodiment is 1.5 mm, and the flatness is 0.05 mm / 100 mm.

[0096] In this embodiment, the thermal conductivity of the heat dissipation substrate at room temperature is 260 W / m·K, and the coefficient of linear expansion in the range of 25-300℃ is 6.8 × 10⁻⁶. -6 / ℃, the total oxygen content of the heat dissipation substrate in this embodiment is 400ppm, and the residual open porosity is 0.8vol.

[0097] This embodiment uses a moderately stable parameter combination, which is suitable for heat dissipation applications of conventional power devices with high requirements for performance stability, such as medium-power IGBT modules and LED lighting heat dissipation substrates.

[0098] Example 2

[0099] This embodiment provides a high thermal conductivity tungsten-copper composite heat dissipation substrate, comprising 78 wt% tungsten phase and 22 wt% copper phase based on the total mass of the heat dissipation substrate in this embodiment; the tungsten phase forms a continuous tungsten skeleton, and the copper phase fills the spaces between the tungsten skeleton to form a continuous copper network; a tungsten carbide transition layer with a thickness of 7 nm is provided between the tungsten phase and the copper phase in this embodiment.

[0100] In this embodiment, the tungsten framework includes a first tungsten region with a characteristic size of 8 μm and a second tungsten region with a characteristic size of 58 μm; the copper phase in this embodiment contains 0.22 wt% zirconium and 0.07 wt% boron by weight of copper phase; the relative density of the heat dissipation substrate in this embodiment is 99.4%.

[0101] In this embodiment, the continuous tungsten framework is derived from a tungsten preform formed by pressing and sintering a bimodal tungsten granulation material made from tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 35 μm. In this embodiment, the continuous copper network is formed by impregnating the tungsten preform of this embodiment with a copper-based infiltrate containing zirconium and boron.

[0102] In this embodiment, tungsten powder with a tungsten carbide transition layer on its surface is prepared through the following steps:

[0103] A1. Add 100 parts by weight of tungsten powder with a particle size D50 of 6μm to a solution containing 2.2 parts by weight of sucrose and 45 parts by weight of deionized water, and mechanically stir for 1.5 hours at a stirring speed of 350 rpm and a stirring temperature of 30℃.

[0104] A2. Dry at 80℃ for 7 hours to obtain the coated precursor;

[0105] A3. In a mixed atmosphere of hydrogen and argon (hydrogen to argon volume ratio of 4:1, total flow rate of 2.5 L / min), the temperature is increased to 900℃ at 8℃ / min and held for 1.5 h.

[0106] A4. Under argon protection, the temperature is reduced to room temperature at a programmed rate of 4℃ / min to obtain tungsten powder with a tungsten carbide transition layer on the surface.

[0107] In this embodiment, the tungsten powder with a tungsten carbide transition layer on its surface has a total carbon content of 0.18 wt% and an oxygen content of 600 ppm. The tungsten carbide transition layer in this embodiment is mainly WC phase, and its thickness, characterized by X-ray diffraction and selected area electron diffraction, is 7 nm, as characterized by transmission electron microscopy.

[0108] The bimodal particle size tungsten granules in this embodiment are prepared through the following steps:

[0109] B1. Mix 32 parts by weight of tungsten powder with a tungsten carbide transition layer on the surface with 68 parts by weight of tungsten powder with a particle size D50 of 35 μm;

[0110] B2. After adding 2.5 parts by weight of polyvinyl alcohol with a degree of hydrolysis of 88% and a degree of polymerization of 2000 and 32 parts by weight of deionized water to prepare a binder liquid, mix it in a planetary mixer at 45°C for 1.3 hours.

[0111] B3. After granulation through a 35-mesh sieve, the material is dried at 72℃ for 6.5 hours to obtain bimodal tungsten granules with a particle size of 550μm and a moisture content of 0.35wt%.

[0112] The tungsten preform in this embodiment is prepared through the following steps:

[0113] C1. Press the bimodal tungsten granules at 240 MPa to obtain a green body;

[0114] C2. The green blank is heated to 1300℃ at a rate of 12℃ / min and held for 70min in a hydrogen atmosphere with a purity of 99.995% and a dew point of -70℃.

[0115] C3. After cooling to 180°C under hydrogen protection, the tungsten preform with a bimodal pore structure of 8μm and 58μm is obtained.

[0116] The copper-based infiltration material containing zirconium and boron in this embodiment is prepared through the following steps:

[0117] D1. Based on a total of 100 parts by weight of copper powder, zirconium powder, and boron powder, 0.22 parts by weight of zirconium powder, 0.07 parts by weight of boron powder, and 99.71 parts by weight of copper powder are loaded into a vacuum induction melting device.

[0118] D2. Heat to 1320℃ under a vacuum of 3×10-3 MPa and hold for 30 min to melt zirconium and boron into the copper liquid in this embodiment.

[0119] D3. The melt is atomized under argon protection to obtain copper-based infiltration powder. The atomizing gas is high-purity argon and the atomization pressure is 3MPa. The resulting infiltration powder has a particle size D50 of 50μm.

[0120] D4. In this embodiment, the oxygen content in the copper-based infiltration material is 350 ppm, and the total content of zirconium and boron is 0.29 wt%.

[0121] The preparation method of the high thermal conductivity tungsten-copper composite heat dissipation substrate in this embodiment includes the following steps:

[0122] S1. A tungsten preform and a copper-based infiltration material containing zirconium and boron are provided. In this embodiment, the tungsten preform is formed by pressing and sintering a bimodal tungsten granule material obtained by pressing tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 35 μm. In this embodiment, the tungsten preform is obtained by pressing and sintering at 240 MPa, and its dimensional shrinkage rate is 8.5%. In this embodiment, the copper-based infiltration material containing zirconium and boron is an infiltration powder with a particle size D50 of 50 μm.

[0123] S2. The copper-based infiltration powder containing zirconium and boron in this embodiment is evenly spread around the tungsten preform in this embodiment. The amount of copper-based infiltration powder is determined according to the theoretical filling amount of 108% of the pore volume of the tungsten preform.

[0124] S3. Under a vacuum atmosphere, heat to 1250℃ at a rate of 10℃ / min and hold for 15min. The vacuum degree during the heating process is 3×10⁻⁶. -6 MPa, so that the copper-based infiltrate containing zirconium and boron in this embodiment is infiltrated into the tungsten preform of this embodiment;

[0125] S4. Cool to room temperature at 5℃ / min under vacuum protection to obtain tungsten-copper composite heat dissipation substrate.

[0126] Following S4, the following post-processing is also included: at a vacuum level of 5×10 -5 Annealed at 750℃ for 2.5h under MPa conditions; lightly rolled and leveled with a reduction rate of 12%; double-sided grinding was performed to obtain a heat dissipation substrate with a thickness of 1.2mm, a thickness tolerance of ±0.05mm, and a surface roughness Ra of 0.4μm.

[0127] In this embodiment, the volume ratio of the first tungsten region to the second tungsten region is 32:68. The thickness of the heat dissipation substrate in this embodiment is 1.2 mm, and the flatness is 0.04 mm / 100 mm.

[0128] In this embodiment, the thermal conductivity of the heat dissipation substrate at room temperature is 245 W / m·K, and the coefficient of linear expansion in the range of 25-300℃ is 6.2 × 10⁻⁶. -6 / ℃, the total oxygen content of the heat dissipation substrate in this embodiment is 500ppm, and the residual open porosity is 0.6vol.

[0129] This embodiment is particularly suitable for high-reliability applications with strict requirements for matching thermal expansion coefficients, such as heat sinks for high-power lasers, packaging substrates for high-frequency microwave devices, and heat dissipation components for aerospace electronic equipment.

[0130] Example 3

[0131] This embodiment provides a high thermal conductivity tungsten-copper composite heat dissipation substrate, comprising 65 wt% tungsten phase and 35 wt% copper phase based on the total mass of the heat dissipation substrate in this embodiment; the tungsten phase forms a continuous tungsten skeleton, and the copper phase fills the spaces between the tungsten skeleton to form a continuous copper network; a tungsten carbide transition layer with a thickness of 3 nm is provided between the tungsten phase and the copper phase in this embodiment.

[0132] The tungsten framework in this embodiment includes a first tungsten region with a characteristic size of 4 μm and a second tungsten region with a characteristic size of 32 μm; the copper phase in this embodiment contains 0.10 wt% zirconium and 0.03 wt% boron by weight of copper phase; the relative density of the heat dissipation substrate in this embodiment is 99.0%.

[0133] In this embodiment, the continuous tungsten framework is derived from a tungsten preform formed by pressing and sintering a bimodal tungsten granulation material made from tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 22 μm. In this embodiment, the continuous copper network is formed by impregnating the tungsten preform of this embodiment with a copper-based infiltrate containing zirconium and boron.

[0134] In this embodiment, tungsten powder with a tungsten carbide transition layer on its surface is prepared through the following steps:

[0135] A1. Add 100 parts by weight of tungsten powder with a particle size D50 of 2.5 μm to a solution containing 0.8 parts by weight of sucrose and 28 parts by weight of deionized water, and stir magnetically for 0.8 h at a stirring speed of 250 rpm and a stirring temperature of 20 °C.

[0136] A2. Dry at 68℃ for 5 hours to obtain the coated precursor;

[0137] A3. In a mixed atmosphere of hydrogen and argon (hydrogen to argon volume ratio of 2:1, total flow rate of 1.5 L / min), the temperature is increased to 820℃ at a rate of 5℃ / min and held for 0.8 h.

[0138] A4. Under argon protection, the furnace is cooled to room temperature to obtain tungsten powder with a tungsten carbide transition layer on the surface;

[0139] In this embodiment, the tungsten powder with a tungsten carbide transition layer on its surface has a total carbon content of 0.08 wt% and an oxygen content of 350 ppm. The tungsten carbide transition layer in this embodiment is mainly W2C phase, and its thickness, characterized by X-ray diffraction and selected area electron diffraction, is 3 nm, as characterized by transmission electron microscopy.

[0140] The bimodal particle size tungsten granules in this embodiment are prepared through the following steps:

[0141] B1. Mix 18 parts by mass of tungsten powder with a tungsten carbide transition layer on the surface with 82 parts by mass of tungsten powder with a particle size D50 of 22 μm;

[0142] B2. After adding 1.5 parts by weight of polyvinyl alcohol with a degree of hydrolysis of 87% and a degree of polymerization of 1500 and 25 parts by weight of deionized water to prepare the adhesive liquid, it is mixed in a double-arm mixer at 35°C for 0.8 hours.

[0143] B3. After granulation through a 45-mesh sieve, the material is dried at 68℃ for 5.5 hours to obtain bimodal tungsten granules with a particle size of 380μm and a moisture content of 0.4wt%.

[0144] The tungsten preform in this embodiment is prepared through the following steps:

[0145] C1. Press the bimodal tungsten granules at 195 MPa to obtain a green body;

[0146] C2. The green compact is heated to 1200℃ at a rate of 8℃ / min and held for 45min in a hydrogen atmosphere with a purity of 99.99%.

[0147] C3. After cooling to 220°C under hydrogen protection, the tungsten preform with a bimodal pore structure of 4μm and 32μm is obtained.

[0148] The copper-based infiltration material containing zirconium and boron in this embodiment is prepared through the following steps:

[0149] D1. Based on a total of 100 parts by weight of copper powder, zirconium powder, and boron powder, 0.10 parts by weight of zirconium powder, 0.03 parts by weight of boron powder, and 99.87 parts by weight of copper powder are loaded into a vacuum induction melting device.

[0150] D2, under a vacuum degree of 8×10 -3 The temperature was raised to 1270℃ at MPa and held for 18 minutes to melt zirconium and boron into the copper liquid in this embodiment.

[0151] D3. The molten material is poured into a graphite mold to obtain a copper-based ingot with a casting temperature of 1220℃.

[0152] D4. In this embodiment, the oxygen content in the copper-based infiltration material is 280 ppm, and the total content of zirconium and boron is 0.13 wt%.

[0153] The preparation method of the high thermal conductivity tungsten-copper composite heat dissipation substrate in this embodiment includes the following steps:

[0154] S1. A tungsten preform and a copper-based infiltration material containing zirconium and boron are provided. In this embodiment, the tungsten preform is formed by pressing and sintering a bimodal tungsten granule material obtained by pressing tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 22 μm. In this embodiment, the tungsten preform is obtained by pressing and sintering at 195 MPa, and its dimensional shrinkage rate is 6.0%. In this embodiment, the copper-based infiltration material containing zirconium and boron is an infiltration ingot.

[0155] S2. Place the copper-based infiltration material containing zirconium and boron in this embodiment above the tungsten preform in this embodiment. The amount of copper-based infiltration material is determined according to the theoretical filling amount of 110% of the pore volume of the tungsten preform.

[0156] S3. Under an argon atmosphere, the temperature is increased to 1200℃ at 6℃ / min and held for 8min. During the heating process, the oxygen content in the argon is 60ppm, so that the copper-based infiltrate containing zirconium and boron in this embodiment is infiltrated into the tungsten preform of this embodiment.

[0157] S4. Cool to room temperature at 7℃ / min under argon protection to obtain tungsten-copper composite heat dissipation substrate.

[0158] Following S4, the following post-processing steps are performed: annealing at 680°C for 1.5 hours under an argon atmosphere; light rolling with a reduction rate of 8% for leveling; and double-sided grinding to achieve a heat dissipation substrate with a thickness of 2.0 mm, a thickness tolerance of ±0.05 mm, and a surface roughness Ra of 0.6 μm.

[0159] In this embodiment, the volume ratio of the first tungsten region to the second tungsten region is 18:82. The thickness of the heat dissipation substrate in this embodiment is 2.0 mm, and the flatness is 0.06 mm / 100 mm.

[0160] In this embodiment, the thermal conductivity of the heat dissipation substrate at room temperature is 275 W / m·K, and the coefficient of linear expansion in the range of 25-300℃ is 7.2 × 10⁻⁶. -6 / ℃, the total oxygen content of the heat dissipation substrate in this embodiment is 300ppm, and the residual open porosity is 1.0vol.

[0161] This embodiment is designed with a high copper content, making it suitable for applications with extremely high heat dissipation requirements, such as heat dissipation of high power density power electronic devices, heat dissipation of power amplifiers in 5G communication base stations, and heat dissipation substrates for high-power motor controllers in electric vehicles.

[0162] Example 4

[0163] This embodiment provides a high thermal conductivity tungsten-copper composite heat dissipation substrate, comprising 82 wt% tungsten phase and 18 wt% copper phase based on the total mass of the heat dissipation substrate in this embodiment; the tungsten phase forms a continuous tungsten skeleton, and the copper phase fills the spaces between the tungsten skeleton to form a continuous copper network; a tungsten carbide transition layer with a thickness of 9 nm is provided between the tungsten phase and the copper phase in this embodiment.

[0164] In this embodiment, the tungsten framework includes a first tungsten region with a characteristic size of 10 μm and a second tungsten region with a characteristic size of 70 μm; the copper phase in this embodiment contains 0.27 wt% zirconium and 0.09 wt% boron by weight of copper phase; the relative density of the heat dissipation substrate in this embodiment is 99.6%.

[0165] In this embodiment, the continuous tungsten framework is derived from a tungsten preform formed by pressing and sintering a bimodal tungsten granulation material made from tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 40 μm. In this embodiment, the continuous copper network is formed by impregnating the tungsten preform of this embodiment with a copper-based infiltrate containing zirconium and boron.

[0166] In this embodiment, tungsten powder with a tungsten carbide transition layer on its surface is prepared through the following steps:

[0167] A1. Add 100 parts by weight of tungsten powder with a particle size D50 of 7 μm to a solution containing 2.7 parts by weight of sucrose and 52 parts by weight of deionized water, and mechanically stir for 1.7 h at a stirring speed of 380 rpm and a stirring temperature of 35℃.

[0168] A2. Dry at 85℃ for 8 hours to obtain the coated precursor;

[0169] A3. In a mixed atmosphere of hydrogen and argon (hydrogen to argon volume ratio of 4.5:1, total flow rate of 2.8 L / min), the temperature is increased to 920℃ at a rate of 9℃ / min and held for 1.7 h.

[0170] A4. Under argon protection, the temperature is reduced to room temperature at a programmed rate of 3℃ / min to obtain tungsten powder with a tungsten carbide transition layer on the surface.

[0171] In this embodiment, the tungsten powder with a tungsten carbide transition layer on its surface has a total carbon content of 0.22 wt% and an oxygen content of 720 ppm. The tungsten carbide transition layer in this embodiment includes W₂C and WC phases, and its thickness, characterized by X-ray diffraction and selected area electron diffraction, is 9 nm, as characterized by transmission electron microscopy.

[0172] The bimodal particle size tungsten granules in this embodiment are prepared through the following steps:

[0173] B1. Mix 37 parts by mass of tungsten powder with a tungsten carbide transition layer on the surface with 63 parts by mass of tungsten powder with a particle size D50 of 40 μm;

[0174] B2. After adding 2.7 parts by weight of polyvinyl alcohol with a degree of hydrolysis of 88% and a degree of polymerization of 2200 and 36 parts by weight of deionized water to prepare a binder liquid, the mixture is kneaded in a planetary mixer at 48°C for 1.7 hours.

[0175] B3. After granulation through a 25-mesh sieve, the material is dried at 76℃ for 7 hours to obtain bimodal tungsten granules with a particle size of 750μm and a moisture content of 0.4wt%.

[0176] The tungsten preform in this embodiment is prepared through the following steps:

[0177] C1. Press the bimodal tungsten granules at 275 MPa to obtain a green body;

[0178] C2. The green blank is heated to 1320°C at a rate of 13°C / min and held for 80 min in a hydrogen atmosphere with a purity of 99.998% and a dew point of -75°C.

[0179] C3. After cooling to 150°C under hydrogen protection, the tungsten preform with a bimodal pore structure of 10μm and 70μm is obtained.

[0180] The copper-based infiltration material containing zirconium and boron in this embodiment is prepared through the following steps:

[0181] D1. Based on a total of 100 parts by weight of copper powder, zirconium powder, and boron powder, 0.27 parts by weight of zirconium powder, 0.09 parts by weight of boron powder, and 99.64 parts by weight of copper powder are loaded into a vacuum induction melting device.

[0182] D2, under a vacuum degree of 2×10 -3 The temperature was raised to 1335℃ at MPa and held for 35 minutes to melt zirconium and boron into the copper liquid in this embodiment.

[0183] D3. The melt is atomized under argon protection to obtain copper-based infiltration powder. The atomizing gas is high-purity argon and the atomization pressure is 3.5 MPa, resulting in infiltration powder with a particle size D50 of 65 μm.

[0184] D4. In this embodiment, the oxygen content in the copper-based infiltration material is 450 ppm, and the total content of zirconium and boron is 0.36 wt%.

[0185] The preparation method of the high thermal conductivity tungsten-copper composite heat dissipation substrate in this embodiment includes the following steps:

[0186] S1. A tungsten preform and a copper-based infiltration material containing zirconium and boron are provided. In this embodiment, the tungsten preform is formed by pressing and sintering a bimodal tungsten granule material obtained by combining tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 40 μm. In this embodiment, the tungsten preform is obtained by pressing and sintering at 275 MPa, and its dimensional shrinkage rate is 10.5%. In this embodiment, the copper-based infiltration material containing zirconium and boron is an infiltration powder with a particle size D50 of 65 μm.

[0187] S2. The copper-based infiltration powder containing zirconium and boron in this embodiment is evenly spread around the tungsten preform in this embodiment. The amount of copper-based infiltration powder is determined according to the theoretical filling amount of 103% of the pore volume of the tungsten preform.

[0188] S3. Under a vacuum atmosphere, heat to 1265℃ at a rate of 11℃ / min and hold for 18min. The vacuum degree during the heating process is 2×10⁻⁶. -6 MPa, so that the copper-based infiltrate containing zirconium and boron in this embodiment is infiltrated into the tungsten preform of this embodiment;

[0189] S4. Cool to room temperature at 4℃ / min under vacuum protection to obtain tungsten-copper composite heat dissipation substrate.

[0190] Following S4, the following post-processing is also included: at a vacuum level of 3×10 -5 Annealed at 780℃ for 2.7h under MPa conditions; lightly rolled and leveled with a reduction rate of 18%; double-sided grinding was performed to obtain a heat dissipation substrate with a thickness of 0.5mm, a thickness tolerance of ±0.05mm, and a surface roughness Ra of 0.9μm.

[0191] In this embodiment, the volume ratio of the first tungsten region to the second tungsten region is 37:63. The thickness of the heat dissipation substrate in this embodiment is 0.5 mm, and the flatness is 0.03 mm / 100 mm.

[0192] The thermal conductivity of the heat dissipation substrate in this embodiment is 232 W / m·K at room temperature, and the coefficient of linear expansion in the range of 25-300℃ is 5.8 × 10⁻⁶. -6 / ℃, the total oxygen content of the heat dissipation substrate in this embodiment is 800ppm, and the residual open porosity is 1.2vol.

[0193] This embodiment uses parameter combinations close to the end of the range, which is suitable for special application scenarios with extremely strict matching of thermal expansion coefficients and requirements for ultra-thinness, such as precision optoelectronic packaging substrates, heat dissipation of satellite payload electronic devices, thermal control components of high-precision laser collimation systems, and replacement applications of thin high-power LED ceramic substrates.

[0194] Comparative Example 1: It is basically the same as Example 1, except that step A3 does not involve carbonization treatment, and the tungsten powder dried in step A2 is used directly as raw material. That is, the surface of the tungsten powder does not have a tungsten carbide transition layer. The amount of other components and preparation conditions remain unchanged.

[0195] Comparative Example 2: It is basically the same as Example 1, except that bimodal tungsten granules are not used. Instead, 100 parts by mass of tungsten powder with a particle size D50 of 28 μm are used to prepare unimodal tungsten granules. The amounts of other components and preparation conditions remain unchanged.

[0196] Comparative Example 3: It is basically the same as Example 1, except that zirconium powder and boron powder are not added to the copper-based infiltration material, and only 100 parts by weight of pure copper powder is used as the copper-based infiltration material. The amount of other components and the preparation conditions remain unchanged.

[0197] Comparative Example 4: It is basically the same as Example 1, except that the tungsten phase content is 55wt% and the copper phase content is 45wt%. The amount of tungsten powder and copper-based infiltrator is adjusted accordingly, while other preparation conditions remain unchanged.

[0198] Comparative Example 5: It is basically the same as Example 1, except that the tungsten phase content is 88wt% and the copper phase content is 12wt%. The amount of tungsten powder and copper-based infiltrator is adjusted accordingly, while other preparation conditions remain unchanged.

[0199] Comparative Example 6: It is basically the same as Example 1, except that the impregnation temperature in step S3 is 1350℃ and the temperature is maintained for 12 minutes. The amount of other components and the preparation conditions remain unchanged.

[0200] Comparative Example 7: It is basically the same as Example 1, except that the carbonization temperature in step A3 is reduced to 750°C and held for 0.5 h, so that the thickness of the tungsten carbide transition layer is 1 nm. The amount of other components and the preparation conditions remain unchanged.

[0201] Comparative Example 8: It is basically the same as Example 1, except that the particle size D50 of the tungsten powder with the tungsten carbide transition layer on the surface in step B1 is 15 μm, resulting in a first tungsten region feature size of 15 μm. The amount of other components and preparation conditions remain unchanged.

[0202] Performance testing:

[0203] Thermal conductivity was evaluated using the laser flare method to assess the thermal conductivity of the tungsten-copper composite heat sink substrate. The substrate was fabricated into circular samples with a diameter of 12.7 mm and a thickness of 1.5 mm, and a graphite coating was applied to the surface to improve absorptivity and emissivity. The thermal diffusivity was measured at room temperature (25℃±2℃) using a laser flare thermophysical analyzer. The front side of the sample was heated by a laser pulse, and the temperature response on the back side was detected by an infrared detector. Specific heat capacity was measured using a differential scanning calorimeter, and density was determined using the Archimedes method. Thermal conductivity was calculated using the formula λ=α·ρ·Cp. The laser pulse energy density was 200-400 J / m², the pulse width was 0.3-1.0 ms, the ambient humidity was <30%RH, and the sample surface parallelism was ≤0.01 mm. Each sample was tested five times, and the average value was taken. The standard deviation was ≤5%. The Cowan+Pulse Correction model was used to correct for the effects of finite pulse width and heat loss. The final result was reported as mean ± standard deviation (n≥5).

[0204] The coefficient of linear expansion was tested using a push-rod dilatometer to evaluate the thermal expansion characteristics of the tungsten-copper composite heat sink substrate and its thermal compatibility with the silicon chip. The heat sink substrate was processed into a sample with a length of 25 mm and a cross-section of 3 mm × the actual thickness of the heat sink substrate, and the surface was polished to a roughness Ra ≤ 0.4 μm. Using a push-rod dilatometer, the temperature was increased from 25 °C to 300 °C at a rate of 5 °C / min under an argon protective atmosphere. The sample underwent thermal expansion under programmed temperature control, and the length change was detected by a high-sensitivity sensor. The temperature-displacement curve was continuously recorded, and the average coefficient of linear expansion within the range of 25-300 °C was calculated using the formula α = (ΔL / L0) / ΔT. The argon flow rate was 100 mL / min, the argon purity was ≥ 99.99%, the sensor resolution was ≤ 0.01 μm, and the sample parallelism was ≤ 0.02 mm. Each sample was tested three times, and the average value was taken, with a standard deviation ≤ 0.2 × 10⁻⁻⁻⁶. 6 / ℃, the system error is calibrated using quartz standard samples, and the final report is in the form of mean ± standard deviation (n≥3).

[0205] The bending strength test evaluated the overall bending resistance of the tungsten-copper composite heat sink substrate using the three-point bending method. The heat sink substrate was processed into strip-shaped specimens with a length of 40 mm, a width of 5 mm, and a thickness of 1.5 mm, and the oxide layer was removed by surface grinding. A universal testing machine was used for the three-point bending test, with a support span of 30 mm and a loading rate of 0.5 mm / min. A transverse load was applied to induce bending until fracture. The load-displacement curve was recorded, and the bending strength was calculated using the formula σ = 3FL / (2bh²). After testing, the fracture morphology was observed using a scanning electron microscope to analyze the interfacial bonding state and failure mode. The support radius was 2 mm, the test temperature was 25℃ ± 2℃, and the relative humidity was 50% ± 10%. At least 10 samples were tested in each group. After removing outliers, the average value was taken, with a standard deviation ≤ 45 MPa. Weibull statistical analysis was used to assess reliability, and the final report was presented as mean ± standard deviation (n ≥ 10).

[0206] The relative density test was conducted using Archimedes' displacement method to evaluate the densification degree and residual porosity of the tungsten-copper composite heat sink substrate. The heat sink substrate was processed into rectangular samples of 10mm × 10mm × 1.5mm × the actual thickness of the heat sink substrate, and the surface was polished to remove the oxide layer and oil. Deionized water was used as the impregnation medium. First, a vacuum of <10Pa was applied for 30 minutes to remove adsorbed gases from the surface and gases in the open pores. Then, the dry weight (ma) of the sample in air, the wet weight (mw) after immersion in deionized water, and the buoyant weight (ms) suspended in water were measured. The bulk density was calculated using the formula ρ = ma / (ma-ms) × ρw, and the theoretical density was calculated using ρth = 1 / (w1 / ρ1 + w2 / ρ2). The relative density was calculated as ρ / ρth × 100%, and the residual open porosity was calculated as (mw-ma) / (mw-ms) × 100%. The water temperature was controlled at 25℃ ± 1℃, and the balance accuracy was ±0.1mg. Each sample was tested 3 times and the average value was taken. The standard deviation was ≤0.3%. The final report was in the form of mean ± standard deviation (n≥3).

[0207] XRD phase analysis was used to identify the tungsten carbide phase type (W2C and / or WC) and crystal structure parameters in the tungsten carbide transition layer and tungsten-copper interface region of the tungsten-copper composite heat sink substrate. The surface of the heat sink substrate sample was ground to expose a fresh interface, and phase analysis was performed using an X-ray diffractometer. Cu Kα radiation (λ = 0.15406 nm) was used; when X-rays were incident on the crystal, diffraction peaks were generated on crystal planes satisfying the condition 2dsinθ = nλ. The scanning range was 2θ = 10–90°, the scanning step size was 0.02°, the scanning rate was 2° / min, the tube voltage was 40 kV, the tube current was 40 mA, and both the divergence slit and the anti-scattering slit were 1°. The obtained XRD patterns were compared with the PDF-2 database using Jade software to identify the phases, and the cell parameters and the content of each phase were refined using the Rietveld full-spectrum fitting method. Peak position fitting and quantitative phase analysis were performed using Jade or HighScore Plus software. The main diffraction peaks of W2C were located at approximately 2θ=39.5° and 2θ=37.8°, while the main diffraction peaks of WC were located at approximately 2θ=35.6° and 2θ=48.3°. The relative contents of each phase were calculated using the peak area integration method.

[0208] TEM interface characterization is used to directly observe the thickness, morphology, and crystal structure of the tungsten-copper phase interface and the tungsten carbide transition layer in tungsten-copper composite heat sink substrates, and to evaluate the interface bonding state. Ultrathin samples with a thickness of approximately 80-100 nm were prepared in the tungsten-copper interface region using focused ion beam (FIB) technology. Bright-field imaging, high-resolution imaging, and selected area electron diffraction (SEED) analysis were performed using transmission electron microscopy at an accelerating voltage of 200 kV. Nanoscale resolution microstructural information was obtained through diffraction and imaging after the electron beam passed through the ultrathin sample. The thickness of the tungsten carbide transition layer was measured at at least 10 interface locations, and the average value was taken. The lattice fringes and crystal orientation of the tungsten carbide layer were analyzed using high-resolution TEM images. The tungsten carbide phase type (W2C or WC) was identified by combined SEED pattern analysis, and the elemental distribution in the interface region was analyzed using EDS energy dispersive spectroscopy. The point resolution was ≤0.19 nm, the information resolution was ≤0.14 nm, and the FIB sample preparation ion beam energy was 30 kV. The thickness of the tungsten carbide layer was measured using Digital Micrograph or Gatan software. The lattice fringe spacing was analyzed by Fast Fourier Transform (FFT), and the interplanar spacing and crystal structure were determined based on the selected area electron diffraction rings or spots.

[0209] Figure 1The X-ray diffraction patterns of Example 1, Comparative Example 1, and Comparative Example 7 are shown. The basic parameters are a scanning range of 10° to 90°, a step of 0.02°, and intensity as the ordinate. The variable parameter is the variation in the intensity of tungsten carbide-related diffraction peaks caused by the difference in the degree of carbon source and interface reaction in different samples. The results show that Example 1 has more obvious tungsten carbide characteristic peaks near 35.6°, 37.8°, 39.5°, and 48.3°, while Comparative Example 1 basically does not show corresponding peaks, and Comparative Example 7 only retains a weak peak signal. This indicates that the sample of Example 1 has relatively more tungsten carbide-related peaks, suggesting that the composition design and sintering control are beneficial to promoting the formation of the carbide phase.

[0210] Figure 2 The scanning electron microscope (SEM) particle size distribution maps for Example 1 and Comparative Example 2 are shown. The basic parameters are: 170 statistical particles, particle size range of 0 to 80 μm, and probability density on the ordinate. The variable parameters are the particle size composition of different sample powders. The results show that Example 1 exhibits a bimodal distribution at approximately 6 μm and 45 μm, while Comparative Example 2 mainly forms a unimodal distribution around approximately 28 μm. This indicates that the particle size distribution used in Example 1 can simultaneously provide fine particle filling and coarse particle skeleton support, which is beneficial for improving packing efficiency and tissue uniformity. Therefore, this particle size design is clearly reasonable.

[0211] Figure 3 The image shows the macroscopic morphology of the high thermal conductivity tungsten-copper composite heat dissipation substrate prepared in Example 1. Its warm steel gray appearance and uniform semi-gloss ground surface indicate that the sample has good surface processing quality, providing a sample basis for subsequent characterization.

[0212] Figure 4 The EDS elemental distribution diagram of the thermal tungsten-copper composite heat dissipation substrate of Example 1 shows that W and Cu elements are complementary in the observed area, and the surface distribution of C, Zr and B elements is characterized. Combined with the relative density and residual open porosity test results, it can be shown that the tungsten preform of Example 1 obtained a relatively dense structure after copper-based infiltration.

[0213] Table 1 Performance Comparison Summary Table

[0214]

[0215] As can be seen from the performance of the examples and comparative examples in Table 1, the tungsten-copper composite heat dissipation substrate of the present invention has a better overall balance in key performance indicators such as thermal conductivity, bending strength, and relative density. Comparative Example 1, due to the absence of a tungsten carbide transition layer on the surface of the tungsten powder, has a bending strength of only 425 MPa, significantly lower than the 685 MPa of Example 1. Simultaneously, its relative density decreases to 97.5%, and its thermal conductivity decreases to 218 W / m·K, indicating that the tungsten carbide transition layer plays a crucial role in improving the wettability of the tungsten-copper interface, promoting sufficient copper phase infiltration, and enhancing bending strength. Comparative Example 2 uses a single-peaked tungsten skeleton, achieving a thermal conductivity of 238 W / m·K, but its bending strength is only 520 MPa, lower than the 685 MPa of Example 1, indicating that the bi-peaked tungsten skeleton design can enhance structural strength while ensuring the heat conduction path. Comparative Example 3, with a copper phase lacking zirconium and boron microalloying elements, has a bending strength of 580 MPa, lower than Example 1, indicating that zirconium-boron microalloying helps improve the microstructure and enhance the overall bending performance of the material. Comparative Examples 4 and 5 represent cases where the tungsten phase content deviates from the optimized range. In Comparative Example 4, the excessively low tungsten phase content caused the linear expansion coefficient to rise to 8.2 × 10⁻⁻⁻⁶. 6 At a temperature of / ℃, the thermal compatibility with the silicon chip deteriorated. In Comparative Example 5, the excessive tungsten phase content led to a decrease in thermal conductivity to 215 W / m·K and a relative density of only 96.8%, indicating insufficient copper phase infiltration. In Comparative Example 6, the excessively high infiltration temperature resulted in excessive interfacial reaction and copper phase volatilization, causing the relative density to drop to 97.2% and the bending strength to decrease to 520 MPa. In Comparative Examples 7 and 8, the excessively thin tungsten carbide transition layer and the excessively large size of the first tungsten region, respectively, resulted in poor interfacial bonding, with bending strengths of 485 MPa and 495 MPa, respectively, both lower than the examples. Comprehensive comparative analysis shows that the present invention, through the synergistic design of the tungsten carbide transition layer, the bimodal tungsten skeleton, and the copper phase zirconium boron microalloying, successfully resolved the coupling contradictions between the high strength of the tungsten preform and sufficient copper phase infiltration, the high bending strength and low interfacial thermal resistance, and the precision machining window of the thin plate and low thermal expansion thermal compatibility, achieving a thermal conductivity ≥232 W / m·K and a linear expansion coefficient of 5.8-7.2 × 10⁻⁻⁻⁶. 6 Excellent comprehensive properties including a temperature of ℃, relative density ≥99.0%, and flexural strength ≥650MPa.

[0216] 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 it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that any equivalent structural transformations made under the concept of the present invention and using the contents of the specification and drawings of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A high thermal conductivity tungsten-copper composite heat dissipation substrate, characterized in that, It includes 60-85 wt% tungsten phase and the balance copper phase based on the total mass of the heat dissipation substrate, wherein the copper phase is 15-40 wt% based on the total mass of the heat dissipation substrate; the tungsten phase forms a continuous tungsten skeleton, and the copper phase fills the spaces between the tungsten skeleton to form a continuous copper network; a tungsten carbide transition layer with an average thickness of 2-10 nm is provided between the tungsten phase and the copper phase. The tungsten framework comprises a first tungsten region with a characteristic size of 2-12 μm and a second tungsten region with a characteristic size of 20-80 μm; the copper phase contains 0.05-0.30 wt% zirconium and 0.02-0.10 wt% boron by weight; the relative density of the heat dissipation substrate is 98.5-99.8%. The continuous tungsten framework is derived from a tungsten preform formed by pressing and sintering a bimodal tungsten granulation material made from tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 15-45μm. The continuous copper network is formed by impregnating the tungsten preform with a copper-based infiltrate containing zirconium and boron. The tungsten powder with a tungsten carbide transition layer on its surface is prepared by the following steps: A1. Add 100 parts by weight of tungsten powder with a particle size D50 of 1-8 μm to a solution containing 0.5-3.0 parts by weight of sucrose and 20-60 parts by weight of deionized water, and stir for 0.5-2.0 h; A2. Dry at 60-90℃ for 4-10 hours to obtain the coated precursor; A3. In a mixed atmosphere of hydrogen and argon, heat to 780-950℃ at a rate of 3-10℃ / min and hold at that temperature for 0.5-2.0h; A4. Cooling under argon protection yields tungsten powder with a tungsten carbide transition layer on the surface; The tungsten powder with a tungsten carbide transition layer on its surface has a total carbon content of 0.03-0.25 wt% and an oxygen content of 200-800 ppm. The bimodal particle size tungsten granules are prepared through the following steps: B1. Mix 15-40 parts by weight of tungsten powder with a tungsten carbide transition layer on the surface with the remainder of tungsten powder with a particle size D50 of 15-45 μm; B2. After adding 1-3 parts by weight of polyvinyl alcohol and 20-40 parts by weight of deionized water to prepare the adhesive solution, knead for 0.5-2.0 hours. B3. After granulation through a 20-60 mesh sieve, dry at 60-80℃ for 4-8 hours to obtain bimodal tungsten granules with a particle size of 250-850μm and a moisture content of no more than 0.5wt%.

2. The high thermal conductivity tungsten-copper composite heat dissipation substrate as described in claim 1, characterized in that, The tungsten preform is prepared by the following steps: C1. Press the bimodal tungsten granules at 150-300MPa to obtain a green body; C2. Heat the green billet to 1150-1350℃ in a hydrogen atmosphere at a rate of 5-15℃ / min and hold for 30-90min. C3. After cooling, a tungsten preform with a bimodal pore structure of 2-10 μm and 20-80 μm is obtained.

3. The high thermal conductivity tungsten-copper composite heat dissipation substrate as described in claim 1, characterized in that, The zirconium and boron-containing copper-based infiltration material is prepared by the following steps: D1. Based on a total of 100 parts by weight of copper powder, zirconium powder, and boron powder, 0.05-0.30 parts by weight of zirconium powder, 0.02-0.10 parts by weight of boron powder, and the balance of copper powder are loaded into a vacuum induction melting device. D2. Under vacuum or argon protection, heat to 1250-1350℃ and hold for 10-40 minutes to melt the zirconium and boron into the copper liquid. D3. Cast the melt into copper-based infiltrate ingots or atomize it into copper-based infiltrate powder; D4. The oxygen content in the copper-based infiltration material is not higher than 500 ppm, and the total content of zirconium and boron is 0.07-0.40 wt%.

4. The high thermal conductivity tungsten-copper composite heat dissipation substrate as described in claim 1, characterized in that, The volume ratio of the first tungsten region to the second tungsten region is 15:85 to 40:60, the thickness of the heat dissipation substrate is 0.3-3.0 mm, and the flatness is 0.02-0.08 mm / 100 mm.

5. The high thermal conductivity tungsten-copper composite heat dissipation substrate as described in claim 1, characterized in that, The thermal conductivity of the heat dissipation substrate at room temperature is 232-275 W / m·K, and the coefficient of linear expansion in the range of 25-300℃ is 5.8 × 10⁻⁶. -6 / ℃ to 7.2×10 -6 / ℃, the total oxygen content of the heat dissipation substrate is 300-800ppm, and the residual open porosity is not higher than 1.5vol.

6. A method for preparing a high thermal conductivity tungsten-copper composite heat dissipation substrate as described in any one of claims 1-5, characterized in that, Includes the following steps: S1. Provide a tungsten preform and a copper-based infiltrate containing zirconium and boron, wherein the tungsten preform is formed by pressing and sintering a bimodal tungsten granule obtained by pressing tungsten powder with a tungsten carbide transition layer on the surface and tungsten powder with a particle size D50 of 15-45μm. S2. Place the zirconium- and boron-containing copper-based infiltration material above or around the tungsten preform; S3. Heat to 1180-1280℃ in a vacuum or argon atmosphere and hold for 5-20 minutes to allow the zirconium- and boron-containing copper-based infiltrate into the tungsten preform. S4. Cool under vacuum or inert atmosphere protection to obtain tungsten-copper composite heat dissipation substrate.

7. The method as described in claim 6, characterized in that, The tungsten preform provided in step S1 is obtained by pressing and sintering at 150-300 MPa, with a dimensional shrinkage rate of 4-12%; the zirconium and boron-containing copper-based infiltration material provided in step S1 is an infiltration ingot or infiltration powder, with a particle size D50 of 20-80 μm; the vacuum degree before heating in step S3 is 1×10⁻⁶. -8 MPa to 1×10 - 5 MPa, or the oxygen content in the argon atmosphere is not higher than 100 ppm.

8. The method as described in claim 6, characterized in that, After step S4, at least one of the following post-processing steps is included: annealing at 650-800℃ for 1-3 hours; light rolling and leveling with a reduction rate of 5-20%; and double-sided grinding to make the thickness tolerance of the obtained heat dissipation substrate ±0.05mm and the surface roughness Ra 0.2-1.0μm.