Method for optimizing reliability of aluminum nitride copper-clad substrate

Abstract

The invention discloses a method for optimizing the reliability of an aluminum nitride copper-clad substrate, and relates to the technical field of semiconductor devices. According to the method, the aluminum nitride copper-clad substrate is baked twice and the residual copper rate is compensated, so that the reliability is improved; the method specifically comprises the following steps of: baking a sintered and twisted aluminum nitride copper-clad substrate for the first time, transferring a pattern, applying an etching hole at a sharp corner of a pattern surface, applying an etching hole / groove on a non-pattern surface according to a residual copper rate difference value between the non-pattern surface and the pattern surface of the copper-clad substrate for residual copper rate compensation, transferring the pattern, and baking for the second time. By controlling the residual copper rate on the two sides of the substrate and baking twice, the stress on the two sides is balanced, and the overall reliability is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor devices, in particular to a method for optimizing the reliability of an aluminum nitride copper clad substrate. Background Art

[0002] Aluminum nitride (AlN) copper-clad substrates are widely used in high-power electronic devices due to their high thermal conductivity (≥170W / m·K). However, they have low mechanical strength (flexural strength <400MPa) and are prone to ceramic cracking and failure under thermal stress. Existing technologies improve reliability through the following methods:

[0003] 1. Single baking process: thermal stress is released after copper-ceramic sintering, but it cannot solve the secondary stress accumulation after pattern transfer;

[0004] 2. Local etching (such as CN202111163775.7): Etching holes and grooves in the weak areas of the copper layer, but without dynamic control of the residual copper rate, the stress balance effect is limited;

[0005] 3. Nitrogen environment annealing: Oxidation is suppressed but the temperature-time parameters are not optimized. Residual stress still causes warpage > 0.5mm.

[0006] Technical pain points: The existing etching process parameters are single (fixed hole depth / spacing), which cannot match the requirements of different substrate sizes and power density. In addition, baking and etching do not form a synergistic mechanism, and the number of porcelain crack failure cycles is less than 300 times (IEC60749-25 standard). Summary of the Invention

[0007] The object of the present invention is to provide a method for optimizing the reliability of an aluminum nitride copper clad substrate to solve the problems raised in the prior art.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A method for optimizing the reliability of an aluminum nitride copper clad substrate is disclosed. The method improves reliability by performing two bakes on the aluminum nitride copper clad substrate and compensating for residual copper. Specifically, the method comprises the following steps: a sintered, twisted aluminum nitride copper clad substrate is baked for the first time, followed by pattern transfer; etching holes are formed at the sharp corners of the patterned surface; and then, based on the difference in residual copper rates between the non-patterned and patterned surfaces of the copper clad substrate, etching holes / grooves are formed on the non-patterned surface to dynamically compensate for the residual copper rate; and after pattern transfer, a second bake is performed.

[0010] Due to the different thermal expansion coefficients of copper and ceramic in the aluminum nitride copper clad substrate, there is tensile stress in the substrate after sintering, resulting in poor reliability. The most obvious manifestation is the distortion of the aluminum nitride copper clad substrate after sintering, so the first baking (referred to as the first baking) is used to release the stress;

[0011] Preferably, the process parameters of the first baking include: temperature of 75° C. to 85° C., time of 15 min to 25 min, nitrogen concentration ≥ 99.999%, and oxygen content ≤ 10 ppm.

[0012] The stress distribution is uneven at the edge of the copper-ceramic bond, making it the weakest point in reliability. The copper surfaces on both sides of the aluminum nitride copper-clad substrate have different functions during use, including the graphic surface and the non-graphic surface. The large difference in residual copper rates on the two sides leads to uneven stress distribution, which is also the reason for poor reliability. Therefore, during pattern transfer, etched holes are added at the edge of the copper-ceramic bond to release stress. Based on the residual copper rates of the graphic and non-graphic surfaces, etched grooves / holes are added on the non-graphic surface to balance the stress on both sides of the aluminum nitride ceramic sheet.

[0013] Preferably, the calculation formula for the aperture of the etching hole is: aperture d = 0.02 × L, where L is the side length of the aluminum nitride copper clad substrate;

[0014] Preferably, the etched hole has a diameter of 0.5 mm to 1.2 mm;

[0015] Preferably, the etched holes have a spacing of 1.5-2.0 times the hole diameter, forming a uniform stress buffer zone;

[0016] Preferably, the groove width of the etching groove is the thickness of the corresponding surface copper layer + (0.2-1.0) mm, and the groove width is ≤1.0 mm;

[0017] Preferably, the length of the etching groove is 20%-100% of the side length of the aluminum nitride copper clad substrate;

[0018] Preferably, the depth of the etching groove is 40% to 100% of the thickness of the corresponding surface copper layer;

[0019] Preferably, the residual copper rate is dynamically compensated to a residual copper rate difference ΔR between the non-graphic surface and the graphic surface of ≤ 3%;

[0020] Due to the different functionalities of the two sides of the aluminum copper clad substrate, the etching holes / grooves added when compensating the residual copper rate on the non-graphic side are generally different from those on the graphic side:

[0021] The etched holes on the non-graphic surface can not only release stress but also adjust the residual copper rate. They can be placed at any position. The etched grooves are also used to adjust the residual copper rate. The groove depth does not necessarily need to be 100% deep.

[0022] Etched holes on the graphic surface are generally used to relieve stress and are placed at the R corners. Etched grooves are used to separate the copper pads and act as an insulator. The groove depth must be 100% porcelain leakage. Unless otherwise specified in the drawing, grooves for adjusting the residual copper rate are generally not added to the graphic surface to avoid affecting the space utilization of the copper pads on the graphic surface.

[0023] Therefore, the stress distribution on the patterned surface and the non-patterned surface is different. Therefore, after the pattern is transferred, the distorted substrate is baked for the second time (referred to as post-bake) to release the stress.

[0024] Preferably, the process parameters of the second baking include: temperature of 100° C. to 120° C., time of 15 min to 25 min, and heating rate of 3° C. / min.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. By adding copper etching holes to the weak corners, the reliability of the weak spots is enhanced;

[0027] 2. By controlling the residual copper rate on both sides of the ceramic and baking twice, the stress on both sides is balanced, optimizing the overall reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a schematic diagram of the first baking process of the present invention;

[0029] Figure 2 Schematic diagram of residual copper rate compensation in the present invention;

[0030] Figure 3 Schematic diagram of adding copper etching holes to weak corners;

[0031] Figure 4 The non-graphic surface after dynamic compensation of the residual copper rate in Example 1;

[0032] Figure 5 This is the graphic surface after dynamic compensation of residual copper rate in Example 1;

[0033] Figure 6 This is an ultrasonic scanning image of Example 1 after 1000 cycles of TC hot and cold cycle testing;

[0034] Figure 7 It is the graphic surface in Comparative Example 1;

[0035] Figure 8 is the non-graphic surface in Comparative Example 1;

[0036] Figure 9 This is the ultrasonic scanning image of Comparative Example 1 after 300 cycles of TC hot and cold cycle testing;

[0037] Figure 10 This is the graphic surface after dynamic compensation of residual copper rate in comparative example 2;

[0038] Figure 11 This is the non-graphic surface after dynamic compensation of residual copper rate in Comparative Example 2;

[0039] Figure 12This is the graphic surface after dynamic compensation of residual copper rate in comparative example 3;

[0040] Figure 13 This is the non-graphic surface after dynamic compensation of residual copper rate in Comparative Example 3;

[0041] Figure 14 This is the ultrasonic scanning image of comparative example 3 after 600 TC hot and cold cycle tests. DETAILED DESCRIPTION

[0042] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of the present invention.

[0043] In the experiment, the specifications of the twisted aluminum nitride copper-clad substrate after sintering were 138×190mm, with a copper thickness of 0.3mm + a porcelain thickness of 0.635mm + a copper thickness of 0.3mm;

[0044] Example 1: A method for optimizing the reliability of an aluminum copper clad nitride substrate, comprising the following steps:

[0045] 1. Pre-baking: The twisted aluminum nitride copper clad substrate is baked for the first time in a nitrogen oven at a temperature of 75°C for 15 minutes.

[0046] 2. During the graphic transfer, dynamic compensation of residual copper rate is performed, such as Figure 4 、 5 As shown:

[0047] 1) According to the pattern, etch holes are applied at the sharp corners of the pattern surface. The diameter of the etched holes is 0.5mm and the hole spacing is 1.0mm;

[0048] 2) Based on the residual copper rate of the graphic surface, a strip-shaped etching groove is applied on the non-graphic surface. The groove width is 0.8 mm, the length accounts for 100% of the side length, and the groove depth is 100% of the thickness of the copper layer on the non-graphic surface. Etched holes are applied on both sides of the etching groove with a diameter of 1.0 mm and a hole spacing of 1.5 mm.

[0049] 3. Post-baking: After the pattern transfer is completed, the aluminum nitride copper clad substrate is baked for the second time in a nitrogen oven at a temperature of 100°C for 15 minutes at a heating rate of 3°C / min.

[0050] Example 2: A method for optimizing the reliability of an aluminum copper clad nitride substrate, comprising the following steps:

[0051] 1. Pre-baking: The twisted aluminum nitride copper clad substrate is baked for the first time in a nitrogen oven at a temperature of 80°C for 20 minutes.

[0052] 2. During graphic transfer, dynamic compensation of residual copper rate is performed:

[0053] 1) According to the pattern, etch holes are applied on both sides of the etched groove on the pattern surface. The diameter of the etched holes is 0.5mm and the hole spacing is 1.0mm;

[0054] 2) Based on the residual copper rate of the graphic surface, a strip-shaped etching groove is applied on the non-graphic surface. The groove width is 1mm, the length accounts for 100% of the side length, and the groove depth is 100% of the thickness of the copper layer on the non-graphic surface. Etched holes are applied on both sides of the etching groove with a diameter of 1.0mm and a hole spacing of 1.5mm.

[0055] 3. Post-baking: After the pattern transfer is completed, the aluminum nitride copper clad substrate is baked for the second time in a nitrogen oven at a temperature of 110°C for 20 minutes at a heating rate of 3°C / min.

[0056] Example 3: A method for optimizing the reliability of an aluminum copper clad nitride substrate, comprising the following steps:

[0057] 1. Pre-baking: The twisted aluminum nitride copper clad substrate is baked for the first time in a nitrogen oven at a temperature of 80°C for 20 minutes.

[0058] 2. During graphic transfer, dynamic compensation of residual copper rate is performed:

[0059] 1) According to the pattern, etch holes are applied at the sharp corners of the pattern surface. The diameter of the etched holes is 0.5mm and the hole spacing is 1.0mm;

[0060] 2) According to the residual copper rate of the graphic surface, a strip-shaped etching groove is applied on the non-graphic surface. The groove width is 0.8mm, the length accounts for 100% of the side length, and the groove depth is 100% of the thickness of the copper layer on the non-graphic surface. Etched holes are applied on both sides of the etching groove with a diameter of 1.2mm and a hole spacing of 1.5mm.

[0061] 3. Post-baking: After the pattern transfer is completed, the aluminum nitride copper clad substrate is baked for the second time in a nitrogen oven at a temperature of 110°C for 20 minutes at a heating rate of 3°C / min.

[0062] Example 4: A method for optimizing the reliability of an aluminum copper clad nitride substrate, comprising the following steps:

[0063] 1. Pre-baking: The twisted aluminum nitride copper clad substrate is baked for the first time in a nitrogen oven at a temperature of 80°C for 20 minutes.

[0064] 2. During graphic transfer, dynamic compensation of residual copper rate is performed:

[0065] 1) According to the pattern, etch holes are applied at the sharp corners of the pattern surface. The diameter of the etched holes is 0.5mm and the hole spacing is 1.0mm;

[0066] 2) Based on the residual copper rate of the graphic surface, a strip-shaped etching groove is applied on the non-graphic surface. The groove width is 1.0 mm, the length accounts for 100% of the side length, and the groove depth is 100% of the thickness of the copper layer on the non-graphic surface. Etched holes are applied on both sides of the etching groove with a diameter of 1.2 mm and a hole spacing of 1.8 mm.

[0067] 3. Post-baking: After the pattern transfer is completed, the aluminum nitride copper clad substrate is baked for the second time in a nitrogen oven at a temperature of 110°C for 20 minutes at a heating rate of 3°C / min.

[0068] Example 5: A method for optimizing the reliability of an aluminum copper clad nitride substrate, comprising the following steps:

[0069] 1. Pre-baking: The twisted aluminum nitride copper clad substrate is baked for the first time in a nitrogen oven at a temperature of 80°C for 20 minutes.

[0070] 2. During graphic transfer, dynamic compensation of residual copper rate is performed:

[0071] 1) According to the pattern, etch holes are applied at the sharp corners of the pattern surface. The diameter of the etched holes is 0.8mm and the hole spacing is 1.2mm;

[0072] 2) Based on the residual copper rate of the graphic surface, a strip-shaped etching groove is applied on the non-graphic surface. The groove width is 0.5 mm, the length accounts for 100% of the side length, and the groove depth is 60% of the thickness of the copper layer on the non-graphic surface. Etched holes are applied on both sides of the etching groove with a diameter of 1.0 mm and a hole spacing of 2.0 mm.

[0073] 3. Post-baking: After the pattern transfer is completed, the aluminum nitride copper clad substrate is baked for the second time in a nitrogen oven at a temperature of 110°C for 20 minutes at a heating rate of 3°C / min.

[0074] Comparative Example 1: As a control experiment of Example 1, the double baking and dynamic compensation of residual copper rate are not performed. Figure 7 、 8 shown.

[0075] Comparative Example 2: As a control experiment of Example 1, no etching grooves and etching holes are applied to the non-patterned surface, and no post-baking is performed. The specific steps are as follows:

[0076] 1. Pre-baking: The twisted aluminum nitride copper clad substrate is baked for the first time in a nitrogen oven at a temperature of 80°C for 20 minutes.

[0077] 2. During the pattern transfer, dynamic compensation of residual copper rate is performed: according to the pattern, copper etching holes are applied at the sharp corners of the pattern surface. The diameter of the etching holes is 0.5mm and the hole spacing is 1.0mm. Figure 10 、 11 shown.

[0078] Comparative Example 3: As a control experiment of Example 1, etching holes are applied to both the pattern surface and the non-pattern surface, but etching grooves are not applied to the non-pattern surface. Figure 12 、 13 As shown, without pre-baking; the specific steps are as follows:

[0079] 1. During graphic transfer, dynamic compensation of residual copper rate is performed:

[0080] 1) According to the pattern, copper etching holes are applied at the sharp corners of the pattern surface. The diameter of the etching holes is 0.5mm and the hole spacing is 1.0mm;

[0081] 2) According to the residual copper rate of the pattern surface, etch holes are added on both sides of the etching groove with a diameter of 1.0mm and a hole spacing of 1.5mm;

[0082] 2. Post-baking: After the pattern transfer is completed, the aluminum nitride copper clad substrate is baked for the second time in a nitrogen oven at a temperature of 110°C for 20 minutes at a heating rate of 3°C / min.

[0083] In summary, the partially adjusted data of Examples 1-5 and Comparative Examples 1-3 are tabulated and organized as shown in Table 1:

[0084] Table 1

[0085]

[0086] Detection experiment:

[0087] Reliability tests (hot and cold cycle TC) were performed on the substrates treated in Examples 1-5 and Comparative Examples 1-3. The test conditions were: temperature cycling from -55°C to 150°C, with each cycle holding at high and low temperatures for 15 minutes, and a switching time of ≤15 seconds.

[0088] The test results are shown in Table 2

[0089] Table 2

[0090]

[0091] Conclusion: The ultrasonic scanning image of the aluminum nitride copper clad substrate after TC hot and cold cycles after the treatment of Example 1 is as follows Figure 6 As shown in FIG, the aluminum nitride copper clad substrate after 1000 cycles of testing did not fail; the ultrasonic scanning image of the aluminum nitride copper clad substrate after TC hot and cold cycles after treatment in Example 1 is shown in FIG. Figure 9As shown in FIG, the ceramic cracks and fails after 300 cycles of testing; comparing the experimental results of Example 1 and Comparative Example 1, it can be seen that after two bakings and dynamic compensation of residual copper rate, the reliability of the aluminum nitride copper clad substrate is effectively optimized;

[0092] The aluminum nitride copper-clad substrate treated in Comparative Example 2 suffered ceramic cracking and copper layer delamination failure after 500 TC hot and cold cycles. This shows that compensating for the residual copper rate on the non-graphic surface and post-baking play an indispensable role in optimizing substrate reliability.

[0093] The ultrasonic scanning image of the aluminum nitride copper clad substrate after TC hot and cold cycles after treatment in comparative example 3 is as follows: Figure 14 As shown in the figure, the copper-ceramic delamination failed after 600 cycles of testing; it can be seen that the etching groove compensation and pre-baking of the non-graphic surface play an indispensable role in optimizing the reliability of the substrate.

[0094] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the reliability of aluminum nitride copper clad substrate, characterized in that: Specifically, the twisted aluminum nitride copper-clad substrate after sintering is baked for the first time, and then the pattern is transferred. After etching holes are applied at the sharp corners of the pattern surface, etching holes / or etching grooves are applied on the non-pattern surface to dynamically compensate for the residual copper rate based on the difference in residual copper rate between the non-pattern surface and the pattern surface of the copper-clad substrate. After the pattern is transferred, a second baking is performed.

2. The method for optimizing the reliability of an aluminum copper clad substrate according to claim 1, wherein: The process parameters of the first baking include: temperature of 75° C. to 85° C., time of 15 min to 25 min, nitrogen concentration ≥ 99.999%, and oxygen content ≤ 10 ppm.

3. The method for optimizing the reliability of an aluminum copper clad substrate according to claim 1, wherein: The calculation formula of the aperture of the etching hole is: aperture d = 0.02 × L, where L is the side length of the aluminum nitride copper clad substrate.

4. The method for optimizing the reliability of an aluminum copper clad substrate according to claim 1, wherein: The diameter of the etched holes is 0.5 mm to 1.2 mm.

5. The method for optimizing the reliability of an aluminum copper clad substrate according to claim 1, wherein: The hole spacing of the etching holes is 1.5 to 2.0 times the hole diameter of the etching holes.

6. The method for optimizing the reliability of an aluminum copper clad substrate according to claim 1, wherein: The width of the etching groove is the thickness of the corresponding copper layer + (0.2-1.0) mm, and the groove width is ≤1.0 mm, the length is 20%-100% of the side length of the aluminum nitride copper-clad substrate, and the groove depth is 40%-100% of the thickness of the corresponding copper layer.

7. The method for optimizing the reliability of an aluminum copper clad substrate according to claim 1, wherein: The residual copper rate is dynamically compensated to ΔR≤3%, where ΔR is the difference between the residual copper rate of the non-graphic surface and the residual copper rate of the graphic surface.

8. The method for optimizing the reliability of an aluminum copper clad substrate according to claim 1, wherein: The process parameters of the second baking include: temperature of 100° C. to 120° C., time of 15 min to 25 min, and heating rate of 3° C. / min.

Images