Apparatus and method for measuring bonding strength of multi-layer d2w hybrid bonded chip

Abstract

This invention relates to an apparatus and method for measuring the bonding strength of multilayer D2W hybrid bonded chips, belonging to the field of semiconductor packaging technology. It addresses one of the problems of inaccurate and large errors in measuring the bonding strength of multilayer, small, and thin D2W hybrid bonded chips. The method of this invention uses the centripetal force generated during the chip's high-speed rotation to replace the tensile force applied during the pulling process. Since the centripetal force always acts perpendicularly towards the center of rotation, it is always uniformly applied to the bonding interface, avoiding stress concentration. The test is conducted under vacuum conditions, avoiding the influence of air resistance on the test structure's side surface on the uneven distribution of centripetal force under high-speed rotation. Extensive testing revealed that a correction factor α needs to be added to the calculation of bonding strength. This correction factor was obtained through comparison with currently widely accepted pull-out tests and is highly representative.

Description

Technical Field

[0001] This invention relates to the field of semiconductor packaging technology, and in particular to an apparatus and method for measuring the bonding strength of multilayer D2W hybrid bonding chips. Background Technology

[0002] The demand for high-performance computing is driving the development of 3D chip packaging technology, among which hybrid bonding is one of the most promising technologies for achieving high-density interconnects in the future. Chip-to-wafer (D2W) hybrid bonding, by bonding known high-performance chips to a substrate wafer, offers advantages such as high yield and multifunctionality, and is the mainstream development direction in the field of hybrid bonding. However, due to the small size and thinness of the chips in D2W hybrid bonding, they are not easy to handle and subject to large external forces. Furthermore, D2W in commercial applications requires multi-layer stacked structures, and strength testing of multi-layer D2W hybrid bonded chips requires repeated testing on the same structure multiple times.

[0003] Currently, the Czochralski method, shearing method, crack propagation method, and four-point bending method are all commonly used methods in the process of bonding strength measurement. However, these methods are only applicable to the testing of wafer-to-wafer (W2W) bonding strength. This is because the upper chip of the D2W structure is small and thin, the area that can be used for testing is small and it is difficult to withstand large external forces. In particular, for multilayer D2W structures, after the first fracture, the test cannot continue due to problems such as inability to clamp or uneven fracture surface. The Czochralski method and the shear method are two of the most commonly used bond strength testing methods. When measuring multilayer D2W structures, the Czochralski method requires high-precision fixtures; if the applied tensile force is not perfectly perpendicular to the structure under test, stress concentration can occur, and the load loading speed is slow, resulting in a long testing time. The shear method is difficult to apply loads to the edges of multilayer thin chips, and the application of loads often leads to breakage of the thin chip itself, resulting in a large error in the measured bond strength. When the crack propagation method and the four-point bending method are applied to multilayer D2W structures, the small and thin chips are prone to bond failure during notch fabrication, and the lack of suitable clamping tools for the small chips during testing often causes fracture to occur within the chip itself rather than at the bonding interface. Therefore, there is currently no suitable method for measuring the bond strength of chips in multilayer stacked D2W structures. Summary of the Invention

[0004] Based on the above analysis, the present invention aims to provide an apparatus and method for measuring the bonding strength of multilayer D2W hybrid bonding chips, in order to solve one of the problems of inaccurate measurement and large error in the bonding strength of multilayer, small and thin D2W hybrid bonding chips.

[0005] In a first aspect, the present invention provides an apparatus for measuring the bonding strength of multilayer D2W hybrid bonding chips. The apparatus includes a rotating base and a test rod, the test rod being arranged perpendicularly to the rotating base. The rotating base and the test rod are disposed inside a vacuum chamber, and a buffer layer is disposed on the inner wall of the vacuum chamber. The chip is attached to the test rod for measuring the bonding strength.

[0006] Furthermore, the test rod has a cuboid structure, and the chip is bonded to two opposite sides of the cuboid structure.

[0007] Furthermore, the cross-section of the chip under test is a cuboid, with a length or width of 1–25 mm and a cross-sectional area of ​​1–625 mm². 2 .

[0008] In a second aspect, the present invention provides a method for measuring the bonding strength of a multilayer D2W hybrid bonding chip using the above-mentioned apparatus, comprising the following steps:

[0009] (1) Cut the multilayer D2W hybrid bonding chip to form a D2D structure;

[0010] (2) Take two of the D2D structures and bond the surfaces of the two D2D structures to the two opposite sides of the test rod;

[0011] (3) Under vacuum conditions, turn on the motor and rotate the test rod on the rotating base. Record the motor speed n when the D2D structure falls off the test rod.

[0012] (4) Calculate the bond strength of the fallen layer = α*(π 2 n 2 mr) / 900s, where n is the rotational speed in r / min, m is the total mass of the D2D structure and the counterweight in kg, r is the distance from the center of the test rod to the center of gravity of the D2D structure and the counterweight combination in m, s is the cross-sectional area of ​​the upper chip, and α is the correction coefficient, ranging from 0.8 to 0.9.

[0013] Furthermore, in step (1), the D2D structure includes an upper chip and a lower chip, both of which are cuboid structures. The connecting line between the center of the upper chip and the center of the lower chip is perpendicular to the horizontal line. The length of the upper chip is less than the length of the lower chip, and the width of the upper chip is less than the width of the lower chip.

[0014] Furthermore, step (2) also includes bonding a counterweight to the upper surface of the D2D structure, wherein the mass of the counterweight is 1-500 times the mass of the D2D structure.

[0015] Furthermore, in step (3), the ambient vacuum level of the testing device is 1000-5000 Pa.

[0016] Furthermore, in step (3), after the motor is turned on, the speed is gradually increased from 0 to accelerate, and the angular acceleration a is satisfied. t Less than centripetal acceleration a n .

[0017] Furthermore, a t <0.1%*a n .

[0018] Furthermore, in step (3), if the rotating base rotates for more than 30 minutes during the test and no D2D structure has fallen off the test rod, then step (3) also includes a pre-test. Specifically, after the motor is turned on, the speed of the test rod starts to increase from 0 and the speed of increase is 1-60 revolutions per minute until the D2D structure breaks and falls off the test rod. This is recorded as the motor speed at this time being n0.

[0019] Remove the remaining D2D structure on the test bar. Following steps (1) and (2), select the same D2D structure and paste it onto the test bar. Turn on the motor and set the initial speed of the test bar to 0.9n0. Then increase the speed at an incremental rate of 1-60 revolutions per minute until the D2D structure breaks and falls off the test bar. Record the motor speed at this time as n.

[0020] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0021] (1) The device of the present invention enables the multilayer D2W hybrid bonding chip to act uniformly on the entire bonding area by centripetal force. The centripetal force generated by the rotating base replaces the tensile force applied during the pulling process. Since the direction of the centripetal force is always perpendicular to the center of rotation, the centripetal force always acts uniformly on the bonding interface, avoiding stress concentration. In addition, the bonding strength is measured under vacuum conditions, avoiding the air resistance on the side surface of the test structure under high-speed rotation, and avoiding the uneven distribution of centripetal force. The buffer layer is to prevent the chip from being damaged when it falls. The sample can be effectively protected after being thrown out, so that the remaining part of the multilayer bonded chip can continue to be tested.

[0022] (2) The method of the present invention uses the centripetal force generated during the high-speed rotation of the chip to replace the tensile force applied during the pulling process. Since the centripetal force always acts perpendicularly to the center of rotation, the centripetal force always acts uniformly on the bonding interface, avoiding stress concentration. The test of the present invention is carried out under vacuum conditions, avoiding the influence of air resistance on the side surface of the test structure on the uneven distribution of centripetal force under high-speed rotation. Since small-sized chips are greatly affected by the uniformity of adhesive coating, they are prone to stress concentration and other problems. The present invention has found through a large number of experiments that the calculation of bonding strength needs to add a correction factor α. This correction factor is obtained by comparing with the currently widely accepted pull test and has strong representativeness.

[0023] (3) The test method described in this invention can repeatedly apply glue and test the structure, thereby realizing the test of the bonding strength of the D2W structure with multilayer bonding.

[0024] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings. Attached Figure Description

[0025] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0026] Figure 1 This is a schematic diagram of the device for measuring the bonding strength of a multilayer D2W hybrid bonded chip according to the present invention;

[0027] Figure 2 This is a schematic diagram illustrating the connection relationship between a test probe and a chip under test according to the present invention.

[0028] Figure 3 This is a schematic diagram of another device for measuring the bonding strength of multilayer D2W hybrid bonded chips according to the present invention;

[0029] Figure 4 A schematic diagram showing the cutting of a multilayer D2W hybrid bonding chip according to the present invention;

[0030] Figure 5 This is a schematic diagram of the adhesive coating process for a chip mounting according to the present invention;

[0031] Figure 6 This is a schematic diagram of the adhesive coating process for a chip according to the present invention.

[0032] Figure label:

[0033] 1-Rotating base, 2-Test rod, 3-Chip under test, 31-Lower chip, 32-Upper chip, 4-Counterweight, 5-Adhesive layer, 6-Vacuum chamber, 7-Buffer layer, 8-Wafer. Detailed Implementation

[0034] The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which constitute a part of the present invention and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0035] A specific embodiment of the present invention, such as Figure 1-6 As shown, an apparatus for measuring the bonding strength of multilayer D2W hybrid bonding chips is disclosed. The apparatus includes a rotating base 1 and a test rod 2, the test rod 2 being arranged perpendicularly to the rotating base 1. The rotating base 1 and the test rod 2 are disposed inside a vacuum chamber 6, and a buffer layer 7 is disposed on the inner wall of the vacuum chamber 6. The chip is attached to the test rod 2 for measuring the bonding strength.

[0036] Specifically, the test rod 1 has a cuboid structure, and the chip is bonded to two opposite sides of the cuboid structure.

[0037] Specifically, the rotating base 1 is connected to a motor. The rotation speed of the rotating base 1 is controlled by the motor.

[0038] Specifically, the chip under test 3 has 2 to 21 layers, and the thickness of a single chip layer is 250 to 800 μm.

[0039] Preferably, the cross-section of the chip under test 3 is a cuboid, the length or width of which is 1-25 mm, and the cross-sectional area is 1-625 mm². 2 .

[0040] The device of this invention enables the centripetal force to act uniformly on the entire bonding area of ​​the multilayer D2W hybrid bonded chip. The centripetal force generated by the rotating base 1 replaces the tensile force applied during the pulling process. Since the direction of the centripetal force is always perpendicular to the center of rotation, the centripetal force always acts uniformly on the bonding interface, avoiding stress concentration. In addition, the bonding strength is measured under vacuum conditions, avoiding the air resistance on the side surface of the test structure under high-speed rotation, and avoiding the uneven distribution of centripetal force. The buffer layer 7 is to prevent damage to the chip when it falls. The sample is effectively protected after being thrown out, so that the remaining part of the multilayer bonded chip can continue to be tested.

[0041] Another specific embodiment of the present invention discloses a method for measuring the bonding strength of multilayer D2W hybrid bonding chips using the above-described apparatus, comprising the following steps:

[0042] (1) Cut the multilayer D2W hybrid bonding chip to form a D2D (chip to chip) structure;

[0043] (2) Take two of the D2D structures and bond the surfaces of the two D2D structures to the two opposite sides of the test rod 2;

[0044] (3) Under vacuum conditions, turn on the motor and rotate the test rod 2 on the rotating base 1. Record the motor speed n when the D2D structure falls off the test rod 2.

[0045] (4) Calculate the bond strength of the fallen layer = α*(π 2 n 2 mr) / 900s, where n is the rotational speed in r / min, m is the total mass of the D2D structure and the counterweight in kg, r is the distance from the center of the test rod 2 to the center of gravity of the combination of the D2D structure and the counterweight in m, s is the cross-sectional area of ​​the upper chip 32, and α is the correction coefficient, with a value range of 0.8-0.9.

[0046] The method of this invention uses the centripetal force generated during the high-speed rotation of the chip to replace the tensile force applied during the pulling process. Since the direction of the centripetal force is always perpendicular to the center of rotation, the centripetal force always acts uniformly on the bonding interface, avoiding stress concentration. The test of this invention is carried out under vacuum conditions, avoiding the influence of air resistance on the side surface of the test structure on the uneven distribution of centripetal force under high-speed rotation. Since small-sized chips are greatly affected by the uniformity of adhesive coating and are prone to stress concentration and other problems, this invention has found through a large number of experiments that the calculation of bonding strength needs to add a correction factor α. This correction factor is obtained by comparing with the currently widely accepted pull test and has strong representativeness.

[0047] The method of this invention allows the centripetal force to act uniformly over the entire bonding region. Therefore, unlike the shearing method, fracture does not occur in a partial area but throughout the entire bonding region. This means that the test method is suitable for side lengths of 1–25 mm and cross-sectional areas of 1–625 mm². 2 It is applicable to chips with a single-layer chip thickness of 250-800μm.

[0048] The method of the present invention does not require complex pretreatment of the structure to be tested. It only requires that the upper surface of the small and thin chip be bonded to the metal rod with a high-strength adhesive. This process requires that the strength of the adhesive be higher than that of the D2D bonding region.

[0049] Specifically, in step (1), the D2D structure (i.e., the chip under test 3) includes an upper chip 32 and a lower chip 31. Both the upper chip 32 and the lower chip 31 are cuboid structures. The connecting line between the center of the upper chip 32 and the center of the lower chip 31 is perpendicular to the horizontal line. The length of the upper chip 32 is less than the length of the lower chip 31, and the width of the upper chip 32 is less than the width of the lower chip 31.

[0050] Preferably, the side length of the upper chip 32 or the lower chip 31 is 1–25 mm, and the cross-sectional area is 1–625 mm². 2 .

[0051] Specifically, step (2) also includes bonding a counterweight 4 to the upper surface of the D2D structure, wherein the mass of the counterweight 4 is 1-500 times the mass of the D2D structure.

[0052] It should be noted that in this invention, the lower surface of the D2D structure is bonded to the test rod 2 with an adhesive, and a counterweight is bonded to the upper surface of the D2D structure with an adhesive. The weight is adjusted by the counterweight. The counterweight 4 is a cuboid metal counterweight, and the length and width of the metal counterweight are the same as the length and width of the upper surface of the D2D structure. The thickness of the metal counterweight is 1-10 mm. The bonding strength between the upper and lower surfaces of the D2D structure is greater than the bonding strength between any two layers in the multilayer D2W hybrid bonding chip.

[0053] Small-sized chips are more susceptible to external interference during bonding strength testing. The single-sided bonding structure vibrates during high-speed rotation, reducing the accuracy of the bonding strength test. This invention adds an additional structure identical to the test structure on the opposite side for balancing, ensuring the stability of the substrate during high-speed rotation.

[0054] Specifically, in step (3), the ambient vacuum level of the test device is 1000-5000Pa, for example, 1000Pa, 1500Pa, 2000Pa, 2500Pa, 3000Pa, 3500Pa, 4000Pa, 4500Pa, 5000Pa.

[0055] It should be noted that selecting the above-mentioned vacuum range can effectively avoid the influence of air resistance.

[0056] Specifically, in step (3), after the motor is turned on, the speed is gradually increased from 0 to accelerate, and the angular acceleration a is satisfied. t Much smaller than the centripetal acceleration a n Preferred, a t <0.1%*a n This is to reduce the impact of angular acceleration.

[0057] It should be noted that the method of the present invention can not only test the bonding strength of a certain layer, but also re-bond the fallen structure to the test rod or re-coat the chip of the remaining test rod with glue and add a counterweight to continue testing.

[0058] Specifically, if the rotation time of the rotating base 1 is greater than 30 minutes during the test and no D2D structure has fallen off the test rod, then step (3) also includes a pre-test. Specifically, after the motor is turned on, the rotation speed of the test rod starts from 0 and increases at a rate of 1-60 revolutions per minute until the D2D structure breaks and falls off the test rod. This is recorded as the motor speed at this time being n0.

[0059] Remove the remaining D2D structure on the test bar. Following steps (1) and (2), select the same D2D structure and paste it onto the test bar. Turn on the motor and set the initial speed of the test bar to 0.9n0. Then increase the speed at an incremental rate of 1-60 revolutions per minute until the D2D structure breaks and falls off the test bar. Record the motor speed at this time as n.

[0060] The method of this invention offers high testing speed and accuracy: Pre-testing steps can be added based on the chip breakage and drop time to quickly determine the approximate rotational speed at which the bonding interface fractures; during formal testing, a steplessly adjustable motor with a speed range of 0–6000 rpm is used, enabling rapid increase of the metal rod's rotational speed in the initial stage, reducing the loading time. When the rotational speed reaches the point where the bonding interface is about to fracture, the speed can be gradually increased in increments of at least 1 rpm, while maintaining a constant angular acceleration α during this stage. t Much smaller than the centripetal acceleration a n (a t <0.1%*a n This is to reduce the impact of angular acceleration.

[0061] The technical solution of the present invention will be further explained below with reference to specific embodiments.

[0062] Example 1

[0063] This embodiment discloses an apparatus for measuring the bonding strength of multilayer D2W hybrid bonding chips. The apparatus includes a rotating base 1 and a test rod 2. The test rod 2 is arranged perpendicularly to the rotating base 1. The rotating base 1 and the test rod 2 are disposed inside a vacuum chamber 6. A buffer layer 7 is disposed on the inner wall of the vacuum chamber 6. The chip is attached to the test rod for measuring the bonding strength.

[0064] In a further embodiment, the test rod 2 has a cuboid structure, and the chip is bonded to two opposite sides of the cuboid structure. The rotating base 1 is connected to a motor, and the rotation speed of the rotating base 1 is controlled by the motor.

[0065] Example 2

[0066] This embodiment provides a method for measuring the bonding strength of a multilayer D2W hybrid bonded chip using the apparatus described in Embodiment 1, comprising the following steps:

[0067] (1) As Figure 4 As shown, the multilayer D2W hybrid bonding chip is formed by bonding multiple upper chips 32 onto the wafer 8. The multilayer D2W hybrid bonding chip is cut to form a D2D (chip-to-chip) structure, that is, a cuboid structure with upper chips 32 and lower chips 32 connected. Both upper chips 32 and lower chips 31 are made of silicon material. The upper chip is a thin sheet with a size of 3mm*5mm and a thickness of 50μm. The upper chip 32 has a multilayer stacked structure. The lower chip 31 is a thick sheet with a size of 6mm*11mm and a thickness of 775μm. The connecting line between the center of the upper chip 32 and the center of the lower chip 31 is perpendicular to the horizontal line.

[0068] (2) Take two of the aforementioned D2D structures respectively, and each according to... Figure 1-3 As shown in Figure 5, an adhesive layer 5 is coated on the upper surface of the upper chip 32 of the D2D structure. The upper chip 32 is bonded to the test rod 2 through the adhesive layer 5, as shown in Figure 5. Figure 6 As shown, an adhesive layer 5 is coated on the lower surface of the lower chip 31, and the counterweight 4 is bonded to the adhesive layer 5. The length and width of the counterweight 4 are the same as those of the upper chip 32. The counterweight 4 is made of metal and weighs 20g. The lower surfaces of the two D2D structures are bonded to the two opposite sides of the test rod according to the above method. The distance from the center of the test rod to the center of gravity of any D2D structure is r = 20cm.

[0069] (3) Under vacuum conditions, the vacuum degree is 1000Pa. The motor is turned on and the test rod rotates on the rotating base 1. During the test, when the rotating base 1 rotates for more than 30 minutes, no D2D structure has fallen off the test rod.

[0070] Specifically, the pre-test included: the test bar was rotated at a rate of 1 revolution per minute, and the D2D structure broke and detached from the test bar. The rotation speed at which the breakage occurred was recorded as 1850 revolutions per minute.

[0071] Remove the remaining D2D structure from the test bar. Following steps (1) and (2), select the same D2D structure and attach it to the test bar. Turn on the motor. Set the initial rotation speed of the test bar to 0.9 × 1850 = 1665 rpm. Gradually increase the rotation speed at a rate of 10 rpm. At this time, the tangential acceleration... Until a certain bonded region breaks, record the rotational speed of the test rod at this point as 1800 rpm (angular velocity ω = 60π rad / s), and calculate the angular acceleration a. t =r*α=0.02094m / s 2 centripetal acceleration a n =ω 2 r = 7106.4 m / s 2 , satisfying a t <0.1%*a n . ,

[0072] (4) Calculate the bond strength of the fallen layer = α*(π 2 n 2 mr) / 900s, where n is the rotational speed in r / min, m is the total mass of the D2D structure and the counterweight in kg, r is the distance from the center of the test rod to the center of gravity of the D2D structure and the counterweight combination in m, s is the cross-sectional area of ​​the upper chip, and α is 0.8.

[0073] Where m = 0.0201 kg, r = 0.2 m, s = 1.5 * 10 -5 m 2 If n = 1800 r / min, then the bond strength of the fallen layer is 7.618 MPa.

[0074] (5) Repeat steps (2)-(4) on the fallen part to calculate the bonding strength of each layer.

[0075] Example 3

[0076] The method for bonding strength of multilayer D2W hybrid bonding chip in this embodiment is the same as that in embodiment 1, except that the vacuum degree is 2500 Pa in step (3) and α is 0.85 in step (4).

[0077] The bond strength of the first fractured layer tested in this embodiment was 7.773 MPa.

[0078] Example 4

[0079] The method for bonding strength of multilayer D2W hybrid bonding chip in this embodiment is the same as that in embodiment 1, except that the vacuum degree is 5000Pa in step (3) and α is 0.9 in step (4).

[0080] The bond strength of the first fractured layer tested in this embodiment was 7.648 MPa.

[0081] Experimental Example 1

[0082] Four chips of the same size and structure as in Example 1 were selected, and the bonding strength of the fractured layer was tested using the Czochralski method. The average bonding strength was calculated to be 7.695 MPa.

[0083] The bond strength of the fractured layer tested in Examples 2-4 of this invention is between 7.618 and 7.773 MPa, which is not significantly different. Therefore, it can be seen that the method of this invention is not significantly different from the Czochralski method, and within an error range of 0.95-1.05 (i.e., 7.310-8.080 MPa), the testing method of this invention is relatively reliable.

[0084] Comparative Example 1

[0085] The measurement method in this comparative example is the same as in Example 2, except that no counterweight is added.

[0086] Comparative Example 1 could not be tested because the chip itself is too small. During high-speed rotation, the centripetal force generated by the chip's own mass is too small compared to the chip's bonding force, thus requiring an excessively high motor speed, making it difficult for the chip to break.

[0087] Comparative Example 2

[0088] The measurement method of this comparative example is the same as that of Example 2, except that the test is not conducted in a vacuum environment in step (3).

[0089] Comparative Example 2 was still able to be tested, and the bond strength value of the first fracture was 6.856 MPa. When tested outside a vacuum environment, the bond strength error was relatively large, and compared with the Czochralski method, the error was larger, falling outside the error range of 0.95-1.05.

[0090] Comparative Example 3

[0091] The test method for this comparative example is the same as that for Example 1, except that the vacuum degree in step (3) is 6000 Pa.

[0092] The bond strength of the first fracture tested using the comparative method was 6.427 MPa.

[0093] The bond strength error is relatively large when tested outside the vacuum level specified in this invention. Compared with the Czochralski method, the error is larger and does not fall within the error range of 0.95-1.05.

[0094] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. An apparatus for measuring the bonding strength of multilayer D2W hybrid bonded chips, characterized in that, The device includes a rotating base and a test rod, the test rod being arranged perpendicularly to the rotating base. The rotating base and the test rod are disposed inside a vacuum chamber, and a buffer layer is provided on the inner wall of the vacuum chamber. The chip is attached to the test rod for measuring the bonding strength.

2. The apparatus for measuring the bonding strength of multilayer D2W hybrid bonded chips according to claim 1, characterized in that, The test rod is a cuboid structure, and the chip is bonded to two opposite sides of the cuboid structure.

3. The apparatus for measuring the bonding strength of multilayer D2W hybrid bonded chips according to claim 1, characterized in that, The chip under test has a cuboid cross-section, with a length or width of 1–25 mm and a cross-sectional area of ​​1–625 mm². 2 .

4. A method for measuring the bonding strength of a multilayer D2W hybrid bonded chip using the apparatus described in any one of claims 1-3, characterized in that, Includes the following steps: (1) Cut the multilayer D2W hybrid bonding chip to form a D2D structure; (2) Take two of the D2D structures and bond the surfaces of the two D2D structures to the two opposite sides of the test rod; (3) Under vacuum conditions, turn on the motor and rotate the test rod on the rotating base. Record the motor speed n when the D2D structure falls off the test rod. (4) Calculate the bond strength of the fallen layer = α*(π 2 n 2 mr) / 900s, where n is the rotational speed in r / min, m is the total mass of the D2D structure and the counterweight in kg, r is the distance from the center of the test rod to the center of gravity of the D2D structure and the counterweight combination in m, s is the cross-sectional area of ​​the upper chip, and α is the correction coefficient, ranging from 0.8 to 0.

9.

5. The method according to claim 4, characterized in that, In step (1), the D2D structure includes an upper chip and a lower chip. Both the upper chip and the lower chip are cuboid structures. The connecting line between the center of the upper chip and the center of the lower chip is perpendicular to the horizontal line. The length of the upper chip is less than the length of the lower chip, and the width of the upper chip is less than the width of the lower chip.

6. The method according to claim 4, characterized in that, Step (2) also includes attaching a counterweight to the upper surface of the D2D structure, wherein the mass of the counterweight is 1-500 times the mass of the D2D structure.

7. The method according to claim 4, characterized in that, In step (3), the ambient vacuum level of the testing device is 1000-5000 Pa.

8. The method according to claim 4, characterized in that, In step (3), after the motor is turned on, the speed is gradually increased from 0 to accelerate, and the angular acceleration a is satisfied. t Less than centripetal acceleration a n .

9. The method according to claim 8, characterized in that, a t <0.1%*a n 。 10. The method according to claim 4, characterized in that, In step (3), if the rotating base rotates for more than 30 minutes during the test and no D2D structure falls off the test rod, then step (3) also includes a pre-test. Specifically, after the motor is turned on, the speed of the test rod increases from 0 to 60 revolutions per minute until the D2D structure breaks and falls off the test rod. This is recorded as the motor speed n0 at this time. Remove the remaining D2D structure on the test bar. Following steps (1) and (2), select the same D2D structure and paste it onto the test bar. Turn on the motor and set the initial speed of the test bar to 0.9n0. Then increase the speed at an incremental rate of 1-60 revolutions per minute until the D2D structure breaks and falls off the test bar. Record the motor speed at this time as n.

Images