A method and system for sputter-electroplating synergistic bonding of DPC ceramic substrates
By employing surface condition detection, vacuum desorption, in-situ plasma cleaning, and segmented electroplating processes, the problem of decreased interfacial adhesion of DPC ceramic substrates under thermal shock was solved, enabling the fabrication of ceramic substrates with high adhesion and excellent thermal reliability.
Patent Information
- Application Number
- CN202610937251.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-26
- Publication Date
- 2026-11-17
AI Technical Summary
After undergoing thermal shock or prolonged high-temperature operation, the bonding force between the metal layer and the ceramic substrate of existing DPC ceramic substrates decreases, leading to blistering, peeling, or microcracks in the circuit layer, which affects the long-term operational reliability of power devices.
By employing a differentiated cleaning and activation strategy driven by surface state detection, combined with vacuum desorption and in-situ plasma cleaning, an activated surface suitable for sputtering metal nucleation is formed. A reactive gradient transition layer and a copper seed layer are then formed by magnetron sputtering in a vacuum environment without disrupting the vacuum. Subsequently, a segmented electroplating thickening process is performed, including low-current flash plating, pulse plating, and DC plating, with heat treatment in between to release internal stress.
It improves the bonding force between the ceramic substrate and the metal layer, enhances thermal cycling reliability, solves the problem of decreased interfacial bonding force, and ensures stability and reliability in harsh environments.
Smart Images

Figure CN123108495A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the fields of printed circuit manufacturing and semiconductor packaging technology, and in particular to a method and system for synergistic enhancement of DPC ceramic substrate sputtering-electroplating. Background Technology
[0002] Direct Plated Copper (DPC) ceramic substrates are a key electronic packaging material widely used in high-power LEDs, IGBT power modules, RF devices, and new energy vehicle electronic control systems. This type of substrate typically uses high thermal conductivity ceramics such as alumina, aluminum nitride, or silicon nitride as the insulating substrate. A titanium / copper seed layer is deposited on the ceramic surface using magnetron sputtering, followed by electroplating to thicken the circuit layer to the designed thickness. This achieves a balance between the excellent thermal conductivity and insulation properties of ceramics and the high current-carrying capacity of metal circuits. Existing fabrication processes mainly involve chemical cleaning of the ceramic substrate, seed layer sputtering deposition in a vacuum environment, and subsequent electroplating thickening treatment. The metallization pattern is constructed through a combination of physical vapor deposition and electrochemical deposition.
[0003] However, in the existing technology, due to the large difference in the coefficient of thermal expansion between ceramic materials and metal layers, and the complex chemical bonding mechanism at the interface, after repeated thermal shocks or long-term high-temperature operation, the bonding force between the metal layer and the ceramic substrate is prone to decrease, which leads to blistering, peeling or microcrack propagation of the circuit layer, affecting the long-term operational reliability of power devices. Summary of the Invention
[0004] This application provides a method and system for synergistic bonding of DPC ceramic substrates by sputtering and electroplating to solve the above-mentioned problems.
[0005] In a first aspect, this application provides a method for synergistic bonding of DPC ceramic substrates by sputtering and electroplating, the method comprising: S1. Perform surface condition detection on the DPC ceramic substrate to be metallized, obtain the ceramic material type, surface roughness, surface contact angle and surface particle residue state of the DPC ceramic substrate, and determine the subsequent cleaning and activation intensity based on the surface condition detection results. S2. The DPC ceramic substrate is subjected to the following steps in sequence according to the cleaning and activation intensity: surface cleaning, ultrasonic water washing, alkaline cleaning, water washing, acid activation, water washing, secondary ultrasonic water washing, drying, and substrate heat treatment, so that the surface of the DPC ceramic substrate forms an activated surface suitable for sputtering metal nucleation. S3. The DPC ceramic substrate processed in step S2 is sent into the magnetron sputtering vacuum chamber for desorption treatment in a vacuum environment, and argon-containing working gas is introduced for in-situ plasma cleaning to remove adsorbed water, oxides and organic residues remaining on the activated surface. S4. Without disrupting the vacuum environment, a first metal adhesion layer, a reactive gradient transition layer, and a copper seed layer are sequentially formed on the surface of the DPC ceramic substrate by magnetron sputtering. The first metal adhesion layer is formed of titanium (Ti) or chromium (Cr). The reactive gradient transition layer is a nitride, oxynitride, or metal-rich gradient film formed by adjusting the gas ratio of nitrogen, argon, or hydrogen during the sputtering of the target material corresponding to the first metal adhesion layer. The ceramic affinity of the reactive gradient transition layer near the DPC ceramic substrate is higher than that near the copper seed layer, and the conductivity near the copper seed layer is higher than that near the DPC ceramic substrate. S5. Perform segmented deposition control on the copper seed layer obtained in step S4. First, form a continuous nucleation copper layer by low-power sputtering, and then form a dense conductive copper layer by sputtering power higher than the low power, so that the copper seed layer simultaneously covers the planar area, edge area and hole wall area of the DPC ceramic substrate. S6. The DPC ceramic substrate after the copper seed layer is formed is cooled under inert gas protection, and the copper seed layer is micro-etched and activated before entering the electroplating thickening process to remove the oxide film on the surface of the copper seed layer and maintain the continuous conductivity of the copper seed layer. S7. Perform segmented electroplating thickening treatment on the DPC ceramic substrate after micro-etching activation. The segmented electroplating thickening treatment includes low-current flash plating, pulse plating and DC plating thickening. The low-current flash plating is used to form an initial electroplated copper layer on the surface of the copper seed layer. The pulse plating is used to refine the electroplated copper grains and reduce the interface stress. The DC plating thickening is used to thicken the copper layer to the thickness that meets the current carrying requirements of the DPC ceramic substrate circuit.
[0006] Through the above technical solution, by setting a differentiated cleaning and activation strategy driven by surface state detection, the cleaning intensity can be precisely matched according to the actual degree of contamination and material characteristics of the ceramic substrate. This solves the problem of over-corrosion or incomplete cleaning caused by uniform cleaning parameters in the prior art, and eliminates the weak bonding layer at the interface from the source. Furthermore, through the synergistic effect of vacuum desorption and in-situ plasma cleaning, the last layer of adsorbed water and organic residues are removed without exposing air, which improves the nucleation density of the first metal adhesion layer. In particular, the introduction of a reactive gradient transition layer, by using dynamic control of gas ratio, achieves a continuous gradual change of composition along the thickness direction, so that a strong chemical bonding region is formed near the ceramic side and a highly conductive metal enrichment region is formed near the copper side. This gradient structure not only acts as a thermal expansion buffer layer to alleviate the interface stress concentration under thermal shock, but also effectively avoids stress cracking caused by abrupt changes in composition at a single interface, thereby improving the bonding force and thermal cycling reliability between the ceramic substrate and the metal layer.
[0007] Optionally, the ceramic material type of the DPC ceramic substrate is alumina, aluminum nitride, silicon nitride, silicon carbide, zirconium oxide, or zirconium oxide toughened alumina ceramic (ZTA). In step S1, the surface condition detection includes detecting the surface roughness, surface contact angle, and surface particle residue state of the DPC ceramic substrate; When the surface roughness is 0.05 to 0.60 μm and the surface contact angle is not greater than 45°, the DPC ceramic substrate is determined to be a substrate suitable for entering the graded cleaning and activation process.
[0008] By defining the type of ceramic material and establishing a quantitative admission standard based on surface roughness and contact angle, the subsequent cleaning and activation parameters can be adapted to the chemical sensitivity of different ceramic materials. This solves the problem of insufficient bonding or excessive surface corrosion caused by using the same process parameters for different ceramic materials. It improves the universality and robustness of the process window for various ceramic substrates. For substrates with small surface contact angles, the activation intensity can be reduced to avoid damaging the crystal lattice, while for substrates with large roughness, cleaning is enhanced to remove resident contaminants, thus improving the yield consistency of the final product.
[0009] Optionally, in step S2, the alkaline cleaning is carried out using sodium hydroxide, potassium hydroxide or organic alkali, the mass concentration of the alkaline cleaning solution is 1-10%, the cleaning temperature is 40-70℃, and the cleaning time is 2-10 min. The acid activation is carried out using oxalic acid, phosphoric acid, hydrochloric acid or nitric acid, with a volume concentration of 1-15%, an activation temperature of 20-50℃, and an activation time of 1-8 min. Furthermore, after alkaline cleaning and acid activation, pure water rinsing is performed to reduce the impact of cleaning residue on subsequent sputtering adhesion.
[0010] By setting the concentration, temperature, and time windows for alkaline cleaning and acid activation, and utilizing two pure water rinses to form a closed loop of chemical dissolution and physical displacement, organic oil stains, metal ions, and inorganic abrasive residues can be specifically removed. This solves the problem of interface bubbling caused by the inability of conventional cleaning to completely remove specific residues. Alkaline conditions are used to efficiently remove grease through saponification, while acidic conditions are used to remove grinding residues through complexation. The low conductivity of pure water rinsing reduces the risk of interface corrosion induced by halogens such as chloride ions during subsequent high-temperature sputtering, providing a reliable foundation for building a clean and activated surface.
[0011] Optionally, in step S3, the desorption process includes vacuum baking the DPC ceramic substrate at 100–300°C, and the vacuum degree of the cavity during vacuum baking is not higher than 5.0 × 10⁻⁶. - ³Pa; The in-situ plasma cleaning uses argon gas, or a working gas formed by a mixture of argon gas and nitrogen or hydrogen gas. The plasma cleaning time is 30 to 300 seconds to remove adsorbed water and residual contaminants from the surface of the DPC ceramic substrate before sputtering.
[0012] The above technical solution effectively removes physically adsorbed water and weakly bonded hydroxyl groups from the ceramic surface by baking and desorption under specific vacuum and temperature conditions, combined with active gas plasma bombardment. It also removes stubborn organic residues by using high-energy ions, thus solving the problem of reduced adhesion caused by the formation of a weak interface layer after re-exposure to air after cleaning. The vacuum environment prevents the re-adsorption of desorbed products. Furthermore, the Ar / N2 or Ar / H2 mixed gas not only physically cleans the surface but also promotes the formation of strong chemical bonds such as Ti-N-Al or reduces surface oxides by introducing active N or H free radicals, thereby improving the cleanliness and reactivity of the initial sputtering interface.
[0013] Optionally, in step S4, the thickness of the first metal adhesion layer is 10–100 nm, the thickness of the reactive gradient transition layer is 20–200 nm, and the thickness of the copper seed layer is 0.3–2.0 μm. The first metal adhesion layer is formed of Ti or Cr, and the reactive gradient transition layer is formed of TiN. x TiO x N CrN x or CrO x N The copper seed layer is formed from copper (Cu).
[0014] By precisely controlling the thickness range and chemical composition of each functional layer, the above technical solution can ensure sufficient interfacial bonding atoms while avoiding thermal stress cracking caused by excessive thickness. The compositional gradient characteristics of the gradient transition layer achieve spatial decoupling between high affinity on the ceramic side and high conductivity on the copper side, thus solving the contradiction between bonding force and conductivity in a single Ti / Cu structure. The 10-100 nm adhesion layer provides sufficient bonding sites, the 20-200 nm gradient layer acts as a stress buffer to suppress element interdiffusion, and the 0.3-2.0 μm copper seed layer maintains a low sheet resistance while satisfying the continuity of electroplating nucleation. The synergistic effect of the three significantly improves the thermal stability and electrical performance of the film system.
[0015] Optionally, the reactive gradient transition layer is formed during the sputtering process of a Ti target or a Cr target; When forming the reactive gradient transition layer, argon gas is first introduced to maintain the sputtering of the metal target, and then nitrogen or hydrogen gas is gradually introduced to adjust the degree of film reaction, so that the reactive gradient transition layer forms a ceramic affinity region on the side close to the DPC ceramic substrate and a metal enriched conductive region on the side close to the copper seed layer. The ceramic affinity region is used to improve the adhesion between the first metal adhesion layer and the DPC ceramic substrate, and the metal-rich conductive region is used to improve the conductivity continuity between the reactive gradient transition layer and the copper seed layer.
[0016] By using the above technical solution, the composition of the film layer gradually changes linearly or exponentially along the thickness direction by dynamically adjusting the proportion of reactive gases during sputtering instead of static co-sputtering, thus avoiding compositional abrupt changes at the interface. This solves the problems of stress concentration caused by abrupt changes in the composition of the gradient layer and interface embrittlement caused by excessive oxidation of the Ti layer. The ceramic affinity region forms a dense chemical bonding network through high N / O content, which greatly enhances the fracture energy. Meanwhile, the metal-rich conductive region maintains good lattice matching and electron mobility through low N / O content. This three-in-one structural design of stress buffer, bonding enhancer and conductive bridge improves the interfacial bonding force and reduces the peeling rate after thermal cycling.
[0017] Optionally, in step S5, the sputtering power of the continuous nucleation copper layer is lower than the sputtering power of the dense conductive copper layer; When forming the continuous nucleated copper layer, the sputtering pressure is 0.2–0.8 Pa and the thickness is 50–300 nm; When forming the dense conductive copper layer, the sputtering pressure is 0.1 to 0.5 Pa and the thickness is 300 nm to 1.7 μm; Through the segmented deposition of the continuous nucleating copper layer and the dense conductive copper layer, the copper seed layer continuously covers the planar region, edge region and hole wall region of the DPC ceramic substrate.
[0018] The above technical solution employs a segmented deposition strategy of low power high pressure and high power low pressure, utilizing the gas scattering effect to improve the angular distribution uniformity of copper atoms in complex three-dimensional structures. At the same time, high-energy particles are used to enhance the film density, thus solving the problem that single-power sputtering cannot simultaneously achieve both hole wall coverage and film density, leading to plating interruptions or excessively high resistance. The low-power stage ensures continuous nucleation without dead angles in hole walls, edges, and micro-recessed areas, avoiding damage to the underlying gradient transition layer from high-energy bombardment. The high-power stage promotes grain growth and defect healing, achieving a good balance between full coverage and low sheet resistance.
[0019] Optionally, in step S6, the inert gas is nitrogen or argon; After the copper seed layer is formed, the DPC ceramic substrate is cooled to 60-100°C in a vacuum environment or an inert gas protected environment before being removed. The micro-etching activation has a micro-etching depth of 0.02–0.30 μm, which is used to remove the oxide film on the surface of the copper seed layer while preserving the continuous conductive path of the copper seed layer.
[0020] By controlling the cooling temperature and implementing precise micro-etching activation under an inert atmosphere, the oxidation kinetics of high-temperature copper surfaces can be effectively suppressed and the nanoscale oxide layer can be completely removed. This solves the key bottleneck of delamination at the electroplating interface caused by oxidation of sputtered copper layers exposed to air. The cooling window of 60–100°C balances efficiency and anti-oxidation requirements, while the micro-etching amount of 0.02–0.30 μm precisely corresponds to the total thickness of Cu2O and CuO. This achieves an atomically clean surface to provide highly active electroplating nucleation sites while avoiding excessive etching that cuts off the conductive network, ensuring zero-delamination bonding between the sputtered layer and the electroplated layer.
[0021] Optionally, in step S7, the current density of the low-current flash plating is 0.2 to 0.8 A / dm², forming an initial electroplated copper layer with a thickness of 0.2 to 2 μm; The pulse electroplating has a forward current density of 1.0 to 3.0 A / dm² and a duty cycle of 40 to 90%, and is used to form a fine-grained electroplated copper layer with a thickness of 3 to 20 μm. The current density for the DC electroplating thickening is 1.5 to 4.0 A / dm², used to thicken the electroplated copper layer to 10 to 100 μm; Furthermore, when the thickness of the electroplated copper layer reaches 20–40 μm, the DPC ceramic substrate is subjected to an intermediate heat treatment at 120–200 °C for 10–60 min to release the internal stress of the electroplated copper layer and reduce the risk of peeling between the sputtered layer and the electroplated layer.
[0022] By constructing a three-stage process chain combining low-current flash plating, pulse electroplating, and DC thickening, and supplementing it with intermediate stress-relieving heat treatment, the growth morphology and internal stress accumulation of copper grains can be controlled in stages. This solves the systemic defects of high current in the early stage of electroplating leading to coarse grains, scorching of orifices, and thermal cycling peeling caused by internal stress accumulation in the thick copper layer. The flash plating layer fills the microscopic defects of the sputtered layer, the pulse electroplating refines the grains and reduces internal stress through periodic current interruption, the DC thickening achieves efficient thickening, and the intermediate heat treatment relaxes the dislocation density below the copper recrystallization temperature, so that the final copper layer has a gradient grain size and extremely low internal stress, which significantly improves the reliability of the product under harsh thermal shock environments.
[0023] Secondly, this application provides a DPC ceramic substrate sputtering-electroplating synergistic bonding system, the system comprising: The information detection module is used to detect the surface condition of the DPC ceramic substrate to be metallized, obtain the ceramic material type, surface roughness, surface contact angle and surface particle residue state of the DPC ceramic substrate, and determine the subsequent cleaning and activation intensity based on the surface condition detection results. The surface activation module is used to sequentially perform the following steps on the DPC ceramic substrate according to the cleaning activation intensity: surface cleaning, ultrasonic water washing, alkaline cleaning, water washing, acid activation, water washing, secondary ultrasonic water washing, drying, and substrate heat treatment, so that the surface of the DPC ceramic substrate forms an activated surface suitable for sputtering metal nucleation. The residue removal module is used to send the DPC ceramic substrate processed in step S2 into the magnetron sputtering vacuum chamber for desorption treatment in a vacuum environment, and to introduce working gas containing argon for in-situ plasma cleaning to remove adsorbed water, oxides and organic residues remaining on the activated surface. The sputtering module is used to sequentially form a first metal adhesion layer, a reactive gradient transition layer, and a copper seed layer on the surface of the DPC ceramic substrate by magnetron sputtering without disrupting the vacuum environment. The first metal adhesion layer is formed of titanium (Ti) or chromium (Cr). The reactive gradient transition layer is a nitride, oxynitride, or metal-rich gradient film formed by adjusting the gas ratio of nitrogen, argon, or hydrogen during the sputtering of the target material corresponding to the first metal adhesion layer. The ceramic affinity of the reactive gradient transition layer near the DPC ceramic substrate is higher than that near the copper seed layer, and the conductivity near the copper seed layer is higher than that near the DPC ceramic substrate. The deposition control module is used to perform segmented deposition control on the copper seed layer obtained in step S4. First, a continuous nucleation copper layer is formed by low-power sputtering, and then a dense conductive copper layer is formed by sputtering power higher than the low power, so that the copper seed layer simultaneously covers the planar area, edge area and hole wall area of the DPC ceramic substrate. The micro-etching activation module is used to cool the DPC ceramic substrate after the copper seed layer is formed under inert gas protection, and to perform micro-etching activation on the copper seed layer before entering the electroplating thickening process, so as to remove the oxide film on the surface of the copper seed layer and maintain the continuous conductivity of the copper seed layer. The electroplating thickening module is used to perform segmented electroplating thickening treatment on the DPC ceramic substrate after micro-etching activation. The segmented electroplating thickening treatment includes low-current flash plating, pulse plating and DC electroplating thickening. The low-current flash plating is used to form an initial electroplated copper layer on the surface of the copper seed layer. The pulse plating is used to refine the electroplated copper grains and reduce the interface stress. The DC electroplating thickening is used to thicken the copper layer to a thickness that meets the current carrying requirements of the DPC ceramic substrate circuit. Attached Figure Description
[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0025] Figure 1 A flowchart of a method for synergistic bonding of DPC ceramic substrate by sputtering and electroplating is provided in one embodiment of this application; Figure 2 This is a comparative SEM cross-section diagram of the DPC ceramic substrate / metal interface provided in an embodiment of this application; Figure 3 A statistical distribution diagram of copper layer grain size provided in an embodiment of this application; Figure 4 This is a schematic diagram of a DPC ceramic substrate sputtering-electroplating synergistic bonding system provided in an embodiment of this application. Detailed Implementation
[0026] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.
[0027] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.
[0028] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.
[0029] Example 1 This embodiment provides a high-adhesion metallization preparation method suitable for alumina DPC ceramic substrates, and verifies the implementation effect of the technical solution of the present invention under optimal process parameters.
[0030] A 0.38 mm thick 96% alumina ceramic substrate was taken and, after polishing, its surface roughness Ra was measured to be 0.20 μm, water contact angle to be 35°, and no obvious large particle residue was found on the surface, indicating suitability for graded cleaning and activation treatment. First, the substrate was immersed in a 5% (w / w) NaOH solution and cleaned at 60°C for 5 min to remove organic oil stains; then it was rinsed with pure water and immersed in a 5% (v / v) oxalic acid-phosphoric acid composite solution for activation at 35°C for 4 min to remove inorganic abrasive residues; next, it underwent pure water rinsing, ultrasonic pure water cleaning (frequency 40 kHz, 5 min), and a second pure water rinsing; finally, it was dried with hot air at 120°C for 30 min.
[0031] The dried substrate is placed into the magnetron sputtering vacuum chamber and evacuated to a base vacuum level of 8×10⁻⁶. -4 The plasma was desorbed by vacuum baking at 180°C for 60 min. Then, a mixture of Ar and H2 gas (Ar flow rate 50 sccm, H2 flow rate 5 sccm) was introduced, and a bias voltage of 200 V was applied for in-situ plasma cleaning for 120 s.
[0032] Without breaking the vacuum, the Ti target was turned on, Ar gas was introduced, and a first metal adhesion layer with a thickness of 40 nm was sputtered. While keeping the Ti target on, N2 gas was gradually introduced, with the N2 flow rate linearly increasing from 0 sccm to 8 sccm, maintained for 60 seconds, and then linearly decreasing back to 0 sccm, forming a TiN layer with a thickness of 80 nm. x A reactive gradient transition layer is formed, with a high N content near the ceramic side and a low N content near the outer side. The N2 gas is turned off, and sputtering continues to form a Cu transition layer with a thickness of 150 nm.
[0033] Next, the copper seed layer was deposited in stages: In the first stage, the sputtering pressure was adjusted to 0.5 Pa, and low power (power density 1.0 W / cm²) was used for sputtering for 10 min to form a continuous nucleated copper layer with a thickness of 200 nm; in the second stage, the sputtering pressure was adjusted to 0.3 Pa, the power was increased to 3.0 W / cm², and sputtering was performed for 25 min to form a dense conductive copper layer with a thickness of 0.8 μm. At this time, the total thickness of the copper seed layer was about 1.15 μm.
[0034] After sputtering, high-purity N2 is introduced into the chamber. The substrate is removed after cooling to 80°C and transferred to the electroplating process within 30 minutes. First, micro-etching activation is performed using a 1% H3PO4 solution, controlling the etching depth to 0.10 μm. Then, the substrate enters the electroplating bath. Initially, low-current flash plating is performed for 5 minutes at a current density of 0.5 A / dm² to form an initial copper layer of approximately 0.5 μm. Next, the process switches to pulse plating mode with a forward current density of 2.0 A / dm², a duty cycle of 60%, and a frequency of 500 Hz to deposit a 10 μm thick fine-grained copper layer. Finally, the process switches to DC plating mode with a current density of 2.5 A / dm² to further thicken the copper layer to a total thickness of 35 μm. When the copper layer thickness reaches 30 μm, the substrate is removed and subjected to an intermediate heat treatment at 180°C in a nitrogen atmosphere for 30 minutes. After this, electroplating continues until the target thickness is reached. Finally, the substrate undergoes low-temperature annealing at 180°C for 30 minutes and surface leveling to obtain the finished product.
[0035] Testing revealed that the copper layer surface of the obtained DPC ceramic substrate was bright and smooth, without pinholes, bubbles, or edge lifting. The cross-cut test result was 5B (no peeling), and the 90° peel strength reached 1.2 N / mm. Cross-sectional SEM observation showed that the ceramic / Ti / TiN... x The Cu interface is tightly bonded with no obvious delamination, and the Ti and N elements exhibit a clear gradient distribution. After 500 thermal shocks (-40℃ to 150℃), the copper layer remains intact without blistering or peeling.
[0036] This embodiment successfully prepared a DPC ceramic substrate with high adhesion and excellent thermal reliability through the synergistic effect of graded cleaning, vacuum plasma activation, gradient transition layer design, and segmented electroplating process, verifying the feasibility and superiority of the technical solution of the present invention.
[0037] Example 2 This embodiment is intended to verify the feasibility of the ceramic material type and surface condition determination threshold specified in this application under the condition of the range edge value.
[0038] Under the same preparation conditions as in Example 1, a zirconia-toughened alumina (ZTA) ceramic substrate with a surface roughness Ra of 0.58 μm (close to the upper limit of 0.60 μm) and a surface contact angle of 44° (close to the upper limit of 45°) was selected. Due to its high roughness, the ultrasonic cleaning time was appropriately extended to 8 min, and the remaining cleaning, sputtering, and electroplating parameters were performed as in Example 1. The first metal adhesion layer was a Cr target, and the reactive gradient transition layer was CrN. x Gradient layer.
[0039] The obtained ZTA ceramic substrate metallized sample exhibited uniform copper layer coverage, with no plating breaks even in areas with high roughness. Adhesion test results showed a cross-cut strength grade of 5B and a peel strength of 1.1 N / mm, slightly lower than Example 1 but fully meeting the usage requirements. No interface failure was observed after 500 cycles of thermal shock.
[0040] The results show that even under critical conditions of high roughness and high contact angle, the graded cleaning and activation strategy of the present invention can still effectively remove resident contaminants and form good bonds, proving that the technical solution has good stability and adaptability within this parameter range.
[0041] Example 3 This embodiment is intended to verify the feasibility of the alkaline cleaning and acidic activation parameters specified in this application under the lower limit conditions.
[0042] The method described in Example 1 is the same, except that the cleaning parameters are adjusted: alkaline cleaning uses a 1% (w / w) NaOH solution at a temperature of 40°C for 2 minutes; acidic activation uses a 1% (v / v) oxalic acid solution at a temperature of 20°C for 1 minute. The remaining steps and parameters are the same as in Example 1.
[0043] The copper layer adhesion test results of the obtained samples were good, with a cross-cut grade of 5B and a peel strength of 1.05 N / mm. Although the cleaning intensity was low, it still achieved an ideal activation effect for substrates with light surface contamination, and no interface blistering caused by insufficient cleaning was observed.
[0044] The results demonstrate that, within the defined lower limit of parameters, the method of the present invention can still achieve effective surface cleaning and activation, reflecting the breadth of the process window.
[0045] Example 4 This embodiment is intended to verify the feasibility of the vacuum desorption temperature and plasma cleaning time specified in this application within the upper limit of the range.
[0046] Based on the process of Example 1, the vacuum baking temperature was adjusted to 300℃ and the baking time to 120 min; the in-situ plasma cleaning time was adjusted to 300 s, and pure Ar gas was used as the working gas. The remaining conditions were the same as in Example 1.
[0047] The resulting product exhibits excellent interfacial adhesion, with a peel strength of 1.25 N / mm. High-temperature desorption more thoroughly removed deeply adsorbed water, and prolonged plasma cleaning further purified the surface. No performance degradation or interfacial embrittlement of the ceramic substrate due to high temperature or prolonged treatment was observed.
[0048] This indicates that under the upper limit of parameters, the technical solution of the present invention remains stable and feasible, and can achieve better interface cleanliness in some cases.
[0049] Example 5 This embodiment is intended to verify the feasibility of the film thicknesses specified in this application under the lower limit conditions.
[0050] The method is the same as in Example 1, except that the film thickness is adjusted: the thickness of the first metal adhesion layer (Ti) is set to 10 nm, and the reactive gradient transition layer (TiN) is... x The thickness is set to 20nm, and the total thickness of the copper seed layer is set to 0.3μm (of which the continuous nucleation layer is 50nm and the dense conductive layer is 250nm).
[0051] The sheet resistance of the obtained sample increased slightly, but still met the conductivity requirements. Adhesion testing showed a cross-cut strength of 5B and a peel strength of 0.95 N / mm. Despite the thin film, the presence of the gradient transition layer effectively buffered the stress, preventing interfacial delamination.
[0052] This demonstrates that within the lower limit of film thickness, the multilayer film structure of the present invention can still exert a synergistic bonding effect, ensuring basic bonding strength and conductivity.
[0053] Example 6 This embodiment is intended to verify the feasibility of the gas regulation method in the reactive gradient transition layer formation mechanism defined in this application.
[0054] Under the same preparation conditions as in Example 1, the formation method of the gradient transition layer was changed: during the sputtering of the Ti target, instead of linear adjustment, N2 gas was introduced in a stepwise manner. That is, 10 nm was sputtered under pure Ar first, then 10 nm was sputtered under Ar+N2 (N2 content 10%), and finally the remaining thickness was sputtered under Ar+N2 (N2 content 20%). Then, the N2 ratio was reduced in reverse stepwise manner to form a film layer with an approximate gradient.
[0055] The interfacial adhesion test of the prepared sample was satisfactory, with a peel strength of 1.15 N / mm. Although the compositional transition was not as smooth as linear adjustment, a basic structure of ceramic affinity region and metal enrichment region was still formed, effectively improving the adhesion.
[0056] This demonstrates that as long as the compositional gradient from the ceramic side to the copper side is achieved, even with a stepwise adjustment, the technical effect of this invention can be achieved, verifying the broad applicability of the technical features.
[0057] Example 7 This embodiment is intended to verify the feasibility of the segmented deposition gas pressure parameters for the copper seed layer as defined in this application under the upper limit condition.
[0058] Keeping the other reaction conditions of Example 1 unchanged, only the sputtering pressure when forming a continuous nucleation copper layer is adjusted to 0.8 Pa (upper limit value), and the pressure when forming a dense conductive copper layer is adjusted to 0.5 Pa (upper limit value).
[0059] The resulting copper seed layer exhibited excellent coverage in the vias and edge areas, and SEM observation showed no dead corners on the via walls. Although the high pressure may have resulted in slightly larger grains in the dense layer, the overall conductivity and adhesion remained excellent, with a peel strength of 1.1 N / mm.
[0060] This demonstrates that under high-pressure boundary conditions, the segmented deposition strategy can still effectively solve the problem of pore wall coverage, and the technical solution is robust.
[0061] Example 8 This embodiment is intended to verify the feasibility of the micro-etching activation amount specified in this application under the lower limit condition.
[0062] The method is the same as in Example 1, except that the amount of micro-etching activated is strictly controlled to 0.02 μm.
[0063] The resulting sample exhibited good adhesion of the electroplated layer without delamination. The minimal etching amount was sufficient to remove the extremely thin natural oxide film while preserving the thickness and integrity of the copper seed layer to the maximum extent, demonstrating the feasibility of this lower limit.
[0064] This indicates that within the lower limit of micro-etching amount, the method of the present invention can still achieve the dual objectives of removing oxide film and maintaining conductive pathway.
[0065] Example 9 This embodiment is intended to verify the feasibility of the pulse electroplating duty cycle and intermediate heat treatment parameters specified in this application under edge value conditions.
[0066] Based on the process of Example 1, the duty cycle of pulse electroplating was adjusted to 40% (lower limit), and intermediate heat treatment was performed when the thickness of the electroplated copper layer reached 20μm (lower limit). The heat treatment temperature was 120℃ (lower limit) and the time was 10min (lower limit).
[0067] The resulting copper layer exhibited significant grain refinement, and internal stress testing revealed a low stress level. No interface cracking was observed after thermal shock testing. Although the heat treatment intensity was low, intervention at this thickness point was sufficient to release some of the accumulated stress.
[0068] This demonstrates that even under parameter edge conditions, the segmented electroplating and intermediate stress release strategy remains effective and can ensure the reliability of thick copper layers.
[0069] Example 10 This embodiment aims to verify, through comparative experiments, the significant improvement effect of the key technical features of the present invention (graded cleaning, gradient transition layer, segmented electroplating, etc.) on the adhesion and reliability of DPC ceramic substrates.
[0070] The test samples include: Sample prepared in Example 1 (full process); Comparative Example 1 (conventional cleaning + Ti / Cu sputtering + direct DC electroplating, without graded cleaning, gradient layer, or segmented electroplating). Comparative Example 2 (same as Example 1, but without low-current flash plating, directly high-current electroplating). Comparative Example 3 (same as Example 1, but without the reactive gradient transition layer, only Ti / Cu bilayer structure).
[0071] Adhesion test (90° peel strength), thermal shock test (-40℃~150℃, 500 times), and high temperature baking test (300℃, 30 min) were performed on each group of samples. The test results are summarized in Table 1.
[0072] Table 1 Performance test results of DPC ceramic substrates under different process conditions
[0073] As shown in Table 1, the peel strength of Example 1 is significantly higher than that of the comparative examples, and it still maintains interface integrity after severe thermal shock and high-temperature baking. In contrast, Comparative Example 1, due to the lack of graded cleaning and gradient layer design, has weak interface bonding and is prone to failure under thermal stress; Comparative Example 2, due to the lack of flash plating nucleation guidance, has more defects in the initial electroplated layer; Comparative Example 3, due to the lack of stress buffering and chemical bonding enhancement effect of gradient transition layer, is extremely prone to catastrophic peeling after thermal cycling.
[0074] like Figure 2 As shown, Figure 2 The images show a comparison of the interface cross-sections of Example 1 and Comparative Example 3 using SEM. It can be clearly seen from the images that the ceramic / metal interface of Example 1 has a smooth and natural transition without obvious boundaries, while the interface of Comparative Example 3 has a clear abrupt change layer and microcrack initiation points. This confirms the key role of the reactive gradient transition layer in alleviating thermal mismatch stress.
[0075] like Figure 3 As shown, Figure 3 The statistical distribution of copper layer grain size at different electroplating stages is shown in the figure. The data shows that after low-current flash plating and pulse electroplating, the copper layer grain size is significantly refined and uniformly distributed, while the grain size of Comparative Example 2 is larger and uneven. This explains why the solution of the present invention can effectively reduce internal stress and suppress the generation of microcracks.
[0076] Based on the above data and graph analysis, this invention, through the coordinated operation of each step, solves the problems of unstable interface bonding, sudden stress changes, and poor thermal reliability in the prior art, and achieves unexpected technical results.
[0077] Experimental results show that the DPC ceramic substrate prepared by this invention exhibits excellent interface stability and thermal conductivity in the thermal cycling test model of high-power LED packaging and IGBT power modules. The metal layer is free from delamination and cracking. Therefore, it can be used as a high-bonding ceramic metallization substrate material to improve the thermal cycling reliability of high-power LED packaging, IGBT power modules, new energy vehicle electronic control modules and radio frequency devices. It can be widely used in the fields of new energy vehicle electronic control, radio frequency devices and high thermal conductivity ceramic packaging.
[0078] Figure 4 This is a schematic diagram of a DPC ceramic substrate sputtering-electroplating synergistic bonding system provided in an embodiment of this application, as shown below. Figure 4 As shown, a DPC ceramic substrate sputtering-electroplating synergistic bonding system 300 of this embodiment includes: an information detection module 301, a surface activation module 302, a residue removal module 303, a sputtering module 304, a deposition control module 305, a micro-etching activation module 306, and an electroplating thickening module 307.
[0079] The information detection module 301 is used to detect the surface condition of the DPC ceramic substrate to be metallized, obtain the ceramic material type, surface roughness, surface contact angle and surface particle residue state of the DPC ceramic substrate, and determine the subsequent cleaning and activation intensity based on the surface condition detection results. The surface activation module 302 is used to sequentially perform the following steps on the DPC ceramic substrate according to the cleaning activation intensity: surface cleaning, ultrasonic water washing, alkaline cleaning, water washing, acid activation, water washing, secondary ultrasonic water washing, drying, and substrate heat treatment, so that the surface of the DPC ceramic substrate forms an activated surface suitable for sputtering metal nucleation. The residue removal module 303 is used to send the DPC ceramic substrate processed in step S2 into the magnetron sputtering vacuum chamber, perform desorption treatment in a vacuum environment, and introduce working gas containing argon for in-situ plasma cleaning to remove adsorbed water, oxides and organic residues remaining on the activated surface. The sputtering module 304 is used to sequentially form a first metal adhesion layer, a reactive gradient transition layer and a copper seed layer on the surface of the DPC ceramic substrate by magnetron sputtering without disrupting the vacuum environment. The first metal adhesion layer is formed of titanium (Ti) or chromium (Cr). The reactive gradient transition layer is a nitride, oxynitride, or metal-rich gradient film formed by adjusting the gas ratio of nitrogen, argon, or hydrogen during the sputtering of the target material corresponding to the first metal adhesion layer. The ceramic affinity of the reactive gradient transition layer near the DPC ceramic substrate is higher than that near the copper seed layer, and the conductivity near the copper seed layer is higher than that near the DPC ceramic substrate. The deposition control module 305 is used to perform segmented deposition control on the copper seed layer obtained in step S4. First, a continuous nucleation copper layer is formed by low-power sputtering, and then a dense conductive copper layer is formed by sputtering power higher than the low power, so that the copper seed layer simultaneously covers the planar area, edge area and hole wall area of the DPC ceramic substrate. The micro-etching activation module 306 is used to cool the DPC ceramic substrate after the copper seed layer is formed under inert gas protection, and to perform micro-etching activation on the copper seed layer before entering the electroplating thickening process, so as to remove the oxide film on the surface of the copper seed layer and maintain the continuous conductivity of the copper seed layer. The electroplating thickening module 307 is used to perform segmented electroplating thickening treatment on the DPC ceramic substrate after micro-etching activation. The segmented electroplating thickening treatment includes low-current flash plating, pulse plating and DC electroplating thickening. The low-current flash plating is used to form an initial electroplated copper layer on the surface of the copper seed layer. The pulse plating is used to refine the electroplated copper grains and reduce the interface stress. The DC electroplating thickening is used to thicken the copper layer to a thickness that meets the current carrying requirements of the DPC ceramic substrate circuit.
[0080] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
Claims
1. A method for synergistic bonding of sputtering and electroplating on DPC ceramic substrates, characterized in that, include: S1. Perform surface condition detection on the DPC ceramic substrate to be metallized, obtain the ceramic material type, surface roughness, surface contact angle and surface particle residue state of the DPC ceramic substrate, and determine the subsequent cleaning and activation intensity based on the surface condition detection results. S2. The DPC ceramic substrate is subjected to the following steps in sequence according to the cleaning and activation intensity: surface cleaning, ultrasonic water washing, alkaline cleaning, water washing, acid activation, water washing, secondary ultrasonic water washing, drying, and substrate heat treatment, so that the surface of the DPC ceramic substrate forms an activated surface suitable for sputtering metal nucleation. S3. The DPC ceramic substrate processed in step S2 is sent into the magnetron sputtering vacuum chamber for desorption treatment in a vacuum environment, and argon-containing working gas is introduced for in-situ plasma cleaning to remove adsorbed water, oxides and organic residues remaining on the activated surface. S4. Without disrupting the vacuum environment, a first metal adhesion layer, a reactive gradient transition layer, and a copper seed layer are sequentially formed on the surface of the DPC ceramic substrate by magnetron sputtering. The first metal adhesion layer is formed of titanium (Ti) or chromium (Cr). The reactive gradient transition layer is a nitride, oxynitride, or metal-rich gradient film formed by adjusting the gas ratio of nitrogen, argon, or hydrogen during the sputtering of the target material corresponding to the first metal adhesion layer. The ceramic affinity of the reactive gradient transition layer near the DPC ceramic substrate is higher than that near the copper seed layer, and the conductivity near the copper seed layer is higher than that near the DPC ceramic substrate. S5. Perform segmented deposition control on the copper seed layer obtained in step S4. First, form a continuous nucleation copper layer by low-power sputtering, and then form a dense conductive copper layer by sputtering power higher than the low power, so that the copper seed layer simultaneously covers the planar area, edge area and hole wall area of the DPC ceramic substrate. S6. The DPC ceramic substrate after the copper seed layer is formed is cooled under inert gas protection, and the copper seed layer is micro-etched and activated before entering the electroplating thickening process to remove the oxide film on the surface of the copper seed layer and maintain the continuous conductivity of the copper seed layer. S7. Perform segmented electroplating thickening treatment on the DPC ceramic substrate after micro-etching activation. The segmented electroplating thickening treatment includes low-current flash plating, pulse plating and DC plating thickening. The low-current flash plating is used to form an initial electroplated copper layer on the surface of the copper seed layer. The pulse plating is used to refine the electroplated copper grains and reduce the interface stress. The DC plating thickening is used to thicken the copper layer to the thickness that meets the current carrying requirements of the DPC ceramic substrate circuit.
2. The method according to claim 1, characterized in that, The ceramic material type of the DPC ceramic substrate is alumina, aluminum nitride, silicon nitride, silicon carbide, zirconium oxide, or zirconium oxide toughened alumina ceramic (ZTA). In step S1, the surface condition detection includes detecting the surface roughness, surface contact angle, and surface particle residue state of the DPC ceramic substrate; When the surface roughness is 0.05 to 0.60 μm and the surface contact angle is not greater than 45°, the DPC ceramic substrate is determined to be a substrate suitable for entering the graded cleaning and activation process.
3. The method according to claim 1, characterized in that, In step S2, the alkaline cleaning is carried out using sodium hydroxide, potassium hydroxide or organic alkali, the mass concentration of the alkaline cleaning solution is 1-10%, the cleaning temperature is 40-70℃, and the cleaning time is 2-10 min. The acid activation is carried out using oxalic acid, phosphoric acid, hydrochloric acid or nitric acid, with a volume concentration of 1-15%, an activation temperature of 20-50℃, and an activation time of 1-8 min. Furthermore, after alkaline cleaning and acid activation, pure water rinsing is performed to reduce the impact of cleaning residue on subsequent sputtering adhesion.
4. The method according to claim 1, characterized in that, In step S3, the desorption process includes vacuum baking the DPC ceramic substrate at 100–300°C, wherein the vacuum degree of the cavity during vacuum baking is not higher than 5.0 × 10⁻⁶. - ³Pa; The in-situ plasma cleaning uses argon gas, or a working gas formed by a mixture of argon gas and nitrogen or hydrogen gas. The plasma cleaning time is 30 to 300 seconds to remove adsorbed water and residual contaminants from the surface of the DPC ceramic substrate before sputtering.
5. The method according to claim 3, characterized in that, In step S4, the thickness of the first metal adhesion layer is 10-100 nm, the thickness of the reactive gradient transition layer is 20-200 nm, and the thickness of the copper seed layer is 0.3-2.0 μm. The first metal adhesion layer is formed of Ti or Cr, and the reactive gradient transition layer is formed of TiN. x TiO x N CrN x or CrO x N The copper seed layer is formed from copper (Cu).
6. The method according to claim 5, characterized in that, The reactive gradient transition layer is formed during the sputtering process of a Ti target or a Cr target; When forming the reactive gradient transition layer, argon gas is first introduced to maintain the sputtering of the metal target, and then nitrogen or hydrogen gas is gradually introduced to adjust the degree of film reaction, so that the reactive gradient transition layer forms a ceramic affinity region on the side close to the DPC ceramic substrate and a metal enriched conductive region on the side close to the copper seed layer. The ceramic affinity region is used to improve the adhesion between the first metal adhesion layer and the DPC ceramic substrate, and the metal-rich conductive region is used to improve the conductivity continuity between the reactive gradient transition layer and the copper seed layer.
7. The method according to claim 1, characterized in that, In step S5, the sputtering power of the continuous nucleation copper layer is lower than that of the dense conductive copper layer; When forming the continuous nucleated copper layer, the sputtering pressure is 0.2–0.8 Pa and the thickness is 50–300 nm; When forming the dense conductive copper layer, the sputtering pressure is 0.1 to 0.5 Pa and the thickness is 300 nm to 1.7 μm; Through the segmented deposition of the continuous nucleating copper layer and the dense conductive copper layer, the copper seed layer continuously covers the planar region, edge region and hole wall region of the DPC ceramic substrate.
8. The method according to claim 1, characterized in that, In step S6, the inert gas is nitrogen or argon; After the copper seed layer is formed, the DPC ceramic substrate is cooled to 60-100°C in a vacuum environment or an inert gas protected environment before being removed. The micro-etching activation has a micro-etching depth of 0.02–0.30 μm, which is used to remove the oxide film on the surface of the copper seed layer while preserving the continuous conductive path of the copper seed layer.
9. The method according to claim 1, characterized in that, In step S7, the current density of the low-current flash plating is 0.2 to 0.8 A / dm², forming an initial electroplated copper layer with a thickness of 0.2 to 2 μm; The pulse electroplating has a forward current density of 1.0 to 3.0 A / dm² and a duty cycle of 40 to 90%, and is used to form a fine-grained electroplated copper layer with a thickness of 3 to 20 μm. The current density for the DC electroplating thickening is 1.5 to 4.0 A / dm², used to thicken the electroplated copper layer to 10 to 100 μm; Furthermore, when the thickness of the electroplated copper layer reaches 20–40 μm, the DPC ceramic substrate is subjected to an intermediate heat treatment at 120–200 °C for 10–60 min to release the internal stress of the electroplated copper layer and reduce the risk of peeling between the sputtered layer and the electroplated layer.
10. A DPC ceramic substrate sputtering-electroplating synergistic bonding system, characterized in that, The method applied to any one of claims 1-9 includes: The information detection module is used to detect the surface condition of the DPC ceramic substrate to be metallized, obtain the ceramic material type, surface roughness, surface contact angle and surface particle residue state of the DPC ceramic substrate, and determine the subsequent cleaning and activation intensity based on the surface condition detection results. The surface activation module is used to sequentially perform the following steps on the DPC ceramic substrate according to the cleaning activation intensity: surface cleaning, ultrasonic water washing, alkaline cleaning, water washing, acid activation, water washing, secondary ultrasonic water washing, drying, and substrate heat treatment, so that the surface of the DPC ceramic substrate forms an activated surface suitable for sputtering metal nucleation. The residue removal module is used to send the DPC ceramic substrate processed in step S2 into the magnetron sputtering vacuum chamber for desorption treatment in a vacuum environment, and to introduce working gas containing argon for in-situ plasma cleaning to remove adsorbed water, oxides and organic residues remaining on the activated surface. The sputtering module is used to sequentially form a first metal adhesion layer, a reactive gradient transition layer, and a copper seed layer on the surface of the DPC ceramic substrate by magnetron sputtering without disrupting the vacuum environment. The first metal adhesion layer is formed of titanium (Ti) or chromium (Cr). The reactive gradient transition layer is a nitride, oxynitride, or metal-rich gradient film formed by adjusting the gas ratio of nitrogen, argon, or hydrogen during the sputtering of the target material corresponding to the first metal adhesion layer. The ceramic affinity of the reactive gradient transition layer near the DPC ceramic substrate is higher than that near the copper seed layer, and the conductivity near the copper seed layer is higher than that near the DPC ceramic substrate. The deposition control module is used to perform segmented deposition control on the copper seed layer obtained in step S4. First, a continuous nucleation copper layer is formed by low-power sputtering, and then a dense conductive copper layer is formed by sputtering power higher than the low power, so that the copper seed layer simultaneously covers the planar area, edge area and hole wall area of the DPC ceramic substrate. The micro-etching activation module is used to cool the DPC ceramic substrate after the copper seed layer is formed under inert gas protection, and to perform micro-etching activation on the copper seed layer before entering the electroplating thickening process, so as to remove the oxide film on the surface of the copper seed layer and maintain the continuous conductivity of the copper seed layer. The electroplating thickening module is used to perform segmented electroplating thickening treatment on the DPC ceramic substrate after micro-etching activation. The segmented electroplating thickening treatment includes low-current flash plating, pulse plating and DC electroplating thickening. The low-current flash plating is used to form an initial electroplated copper layer on the surface of the copper seed layer. The pulse plating is used to refine the electroplated copper grains and reduce the interface stress. The DC electroplating thickening is used to thicken the copper layer to a thickness that meets the current carrying requirements of the DPC ceramic substrate circuit.
