wafer bonding alignment method
By forming diced paths of gradually varying sizes on a semiconductor substrate, the problem of poor bonding caused by differences in thermal expansion coefficients is solved, improving bonding strength and structural integrity, and enhancing electrical connection quality.
Patent Information
- Application Number
- CN202110505952.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2021-03-11
- Filing Date
- 2021-05-10
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2041-05-10
AI Technical Summary
In the semiconductor industry, existing technologies struggle to effectively address issues such as poor bonding and structural integrity caused by differences in the thermal expansion coefficients of different substrates.
By forming cleavage paths of gradually varying sizes on the substrate, the differences in thermal expansion rates of different substrates are compensated, ensuring the alignment of semiconductor devices during bonding operations.
It improves bonding strength, enhances the structural integrity of 3D integrated circuits, reduces device failures, and improves electrical connection quality.
Smart Images

Figure CN114709140B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to a method for wafer bonding alignment. Background Technology
[0002] In the semiconductor industry, bonding is a technique used to form stacked semiconductor devices and three-dimensional integrated circuits. Some examples of bonding include wafer-to-wafer bonding, die-to-wafer bonding, and die-to-die bonding. Summary of the Invention
[0003] According to some embodiments of this disclosure, a method for wafer bonding alignment includes: forming a first plurality of semiconductor devices on a first substrate; forming a first plurality of cleavages between the first plurality of semiconductor devices in the first substrate along a first direction; forming a second plurality of cleavages between the first plurality of semiconductor devices in the first substrate along a second direction, wherein a first width of a first subset of the second plurality of cleavages is greater than a second width of a second subset of the second plurality of cleavages; and performing a bonding operation to bond the first plurality of semiconductor devices to a second plurality of semiconductor devices on a second substrate, wherein the first width is greater than the second width, and the first width and the second width are greater than a third width of a third plurality of cleavages between the second plurality of semiconductor devices, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation.
[0004] According to some embodiments of this disclosure, a method for wafer bonding alignment includes: forming a first plurality of semiconductor devices on a first substrate; forming a first plurality of cleavages between the first plurality of semiconductor devices along a first axis of the first substrate; forming a second plurality of cleavages between the first plurality of semiconductor devices along a second axis of the first substrate, wherein a first width of a first subset of the second plurality of cleavages is smaller than a second width of a second subset of the second plurality of cleavages; and performing a bonding operation to bond the first plurality of semiconductor devices to a second plurality of semiconductor devices on a second substrate, wherein the first width is smaller than the second width and the first width and the second width are smaller than a third width of a third plurality of cleavages between the second plurality of semiconductor devices, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation.
[0005] According to some embodiments of this disclosure, a method for wafer bonding alignment is characterized by comprising: receiving, by a processing device, information associated with a plurality of bonding operations for bonding a first plurality of semiconductor devices on a first plurality of substrates and a second plurality of semiconductor devices on a second plurality of substrates; generating, by the processing device, thermal expansion data of the first plurality of substrates and thermal expansion data of the second plurality of substrates based on the information associated with the plurality of bonding operations; determining, by the processing device, an adjusted width of a plurality of cleavages on at least one of a third substrate or a fourth substrate based on the thermal expansion data of the first plurality of substrates and the thermal expansion data of the second plurality of substrates; and providing an indication of the adjusted width to one or more semiconductor processing stations to form the plurality of cleavages based on the adjusted width. Attached Figure Description
[0006] The best understanding of all aspects of this disclosure will be achieved by reading the following detailed description in conjunction with the accompanying drawings. It should be noted that, in accordance with industry standard practice, the various features are not drawn to scale. In fact, the dimensions of the various features may be arbitrarily increased or decreased for clarity of explanation.
[0007] Figure 1 This is a diagram of an exemplary environment in which the systems and / or methods described herein can be implemented.
[0008] Figures 2A to 2C , Figures 3 to 9 and Figures 10A to 10D This is a diagram of an exemplary implementation described in this document.
[0009] Figure 11 yes Figure 1 A diagram showing exemplary components of one or more devices.
[0010] Figures 12 to 14 This is a flowchart of an exemplary process related to wafer bonding alignment. Detailed Implementation
[0011] The following disclosure provides numerous different embodiments or instances for implementing various features of the provided subject matter. Specific examples of components and arrangements are set forth below to simplify this disclosure. Of course, these are merely examples and are not intended to be limiting. For example, forming a first feature on or on a second feature in the following description may include embodiments in which the first and second features are formed in direct contact, and may also include embodiments in which an additional feature may be formed between the first and second features such that the first and second features are not in direct contact. Furthermore, reference numerals and / or letters may be repeated throughout this disclosure. This repetition is for purposes of brevity and clarity and is not, in itself, indicative of a relationship between the various embodiments and / or configurations discussed.
[0012] Furthermore, for ease of explanation, spatially relative terms such as “beneath,” “below,” “lower,” “above,” and “upper” are used herein to describe the relationship between one element or feature illustrated in the figures and another element or feature. In addition to the orientations shown in the figures, these spatially relative terms are also intended to encompass different orientations of the device in use or operation. The device may be oriented in other ways (rotated 90 degrees or in other orientations), and the spatially relative descriptive terms used herein may be interpreted accordingly.
[0013] Some semiconductor bonding techniques involve heating a substrate (e.g., a semiconductor wafer) to a high temperature. Examples include fusion bonding, eutectic bonding, and hybrid bonding. During the bonding process, the substrate may expand dimensionally due to thermal expansion that occurs as the substrate is heated as part of the bonding process. Some substrates may expand more than others. If substrates with different rates of expansion are bonded together, the difference in expansion rates may cause misalignment of devices and / or structures formed on the substrates. This can lead to poor device connectivity between substrates, reduced bonding strength between substrates, weakened structural integrity of the resulting three-dimensional integrated circuit, poor-quality electrical connections, and / or device failure.
[0014] Some embodiments described herein provide techniques and apparatus for wafer bonding alignment. As described herein, bonding alignment of devices formed on substrates to be bonded can be achieved by using scribe lines between the devices, wherein the size of the scribe lines gradually increases or decreases from the center to the edge of one or more of the substrates to compensate for differences in the thermal expansion rates of the substrates. Due to the gradually increasing or decreasing size of the scribe lines, the devices on the substrates are aligned as the substrates are heated during the bonding operation. The scribe lines can be arranged in a single direction in the substrate to compensate for thermal expansion of the substrate along a single axis, or they can be arranged in multiple directions to compensate for radiation-symmetric thermal expansion. In this way, the techniques and apparatus described herein can reduce device pattern misalignment, improve bonding strength between bonded substrates, improve device structural integrity and electrical connectivity, and / or reduce device failures on the substrates. Figure 1 This is a diagram of an exemplary environment 100 in which the systems and / or methods described herein may be implemented. (See diagram for example.) Figure 1As shown, environment 100 may include multiple semiconductor processing units 102 to 110 and wafer / die transport unit 112. The multiple semiconductor processing units 102 to 110 may include deposition unit 102, exposure unit 104, developing unit 106, etching unit 108, bonding unit 110, and / or another type of semiconductor processing unit. The units included in the exemplary environment 100 may be located in a semiconductor cleanroom, a semiconductor foundry, a semiconductor processing and / or manufacturing facility, and / or another type of semiconductor processing environment.
[0015] Deposition station 102 is a semiconductor processing station that includes a semiconductor processing chamber and one or more means for depositing various types of materials onto a substrate. In some embodiments, deposition station 102 includes a spin coater capable of depositing a photoresist layer on a substrate such as a wafer. In some embodiments, deposition station 102 includes a chemical vapor deposition (CVD) station, such as a plasma-enhanced CVD (PECVD) station, a high-density plasma CVD (HDP-CVD) station, a subatmospheric pressure CVD (SACVD) station, an atomic layer deposition (ALD) station, a plasma-enhanced atomic layer deposition (PEALD) station, or another type of CVD station. In some embodiments, deposition station 102 includes a physical vapor deposition (PVD) station, such as a sputtering station, or another type of PVD station. In some embodiments, exemplary environment 100 includes multiple types of deposition stations 102.
[0016] Exposure equipment 104 is a semiconductor processing equipment capable of exposing a photoresist layer to a radiation source (e.g., an ultraviolet (UV) source (e.g., a deep UV source, an extreme UV (EUV) source, and / or another type of UV source) and / or an X-ray source, an electron beam source, etc.). Exposure equipment 104 exposes the photoresist layer to the radiation source to transfer a pattern from a photomask to the photoresist layer. The pattern may include one or more semiconductor device layer patterns for forming one or more semiconductor devices, patterns for forming one or more structures of a semiconductor device, patterns for etching various portions of a semiconductor device, etc. In some embodiments, exposure equipment 104 includes a scanner, a stepper, or a similar type of exposure equipment.
[0017] The developing station 106 is a semiconductor processing unit capable of developing a photoresist layer exposed to a radiation source to reveal a pattern transferred from the exposure station 104 to the photoresist layer. In some embodiments, the developing station 106 develops the pattern by removing the unexposed portions of the photoresist layer. In some embodiments, the developing station 106 develops the pattern by removing the exposed portions of the photoresist layer. In some embodiments, the developing station 106 develops the pattern by using a chemical developer to dissolve either the exposed or unexposed portions of the photoresist layer.
[0018] Etching equipment 108 is a semiconductor processing equipment capable of etching various types of materials, including substrates, wafers, or semiconductor devices. For example, etching equipment 108 may include wet etching equipment, dry etching equipment, etc. In some embodiments, etching equipment 108 includes a chamber filled with an etchant, and a substrate is placed in the chamber for a specific time period to remove a specific amount of one or more portions of the substrate. In some embodiments, etching equipment 108 may use plasma etching or plasma-assisted etching to etch one or more portions of the substrate, which may involve using an ionized gas to isotropically or directionally etch the one or more portions.
[0019] The bonding station 110 is a semiconductor processing station capable of bonding two or more substrates (e.g., two or more semiconductor substrates, two or more semiconductor wafers, two or more semiconductor devices) together. For example, the bonding station 110 may include a eutectic bonding station capable of forming a eutectic bond between two or more substrates by heating the two or more substrates to form a eutectic system between the materials of the two or more substrates. As another example, the bonding station 110 may include a fusion bonding station capable of forming a direct bond between the substrates (e.g., without using an additional intermediate layer between the two or more substrates) by pre-bonding two or more substrates (e.g., at room temperature or near room temperature) and annealing the substrates to complete the direct bond. In some embodiments, the bonding station 110 uses a combination of bonding techniques (this may be referred to as hybrid bonding) to bond two or more substrates.
[0020] The wafer / die transport platform 112 includes mobile robots, robotic arms, trams or railcars, overhead hoist transport (OHT) systems, automated material handling systems (AMHS), and / or other types of devices for transporting wafers and / or dies between semiconductor processing platforms 102 to 110 and / or to other locations such as wafer racks, storage rooms, etc. In some embodiments, the wafer / die transport platform 112 may be a device programmed to travel a specific path and / or may operate semi-autonomously or autonomously.
[0021] Figure 1 The number and arrangement of the devices shown are provided as one or more examples. In fact, with... Figure 1Compared to the illustrated device, there may be additional devices, fewer devices, different devices, or devices arranged in a different manner. Furthermore, Figure 1 The two or more devices shown can be implemented within a single device, or Figure 1 The single device shown can be implemented as multiple distributed devices. Alternatively, or as another option, a group of devices in environment 100 (e.g., one or more devices) can perform one or more functions described as being performed by another group of devices in environment 100.
[0022] Figures 2A to 2C This is a figure illustrating an exemplary embodiment 200 described herein. Exemplary embodiment 200 may include an example of a dimensionally varying diced channel included on a single substrate of two substrates to be joined. Furthermore, the dimensionally varying diced channel may be included on the substrate along a single direction to compensate for thermal expansion of the substrate in that single direction. In this way, the dimensionally varying diced channel can be used to achieve approximate alignment of semiconductor devices on the two substrates in said single direction.
[0023] Figure 2A A top view (or plan view) of substrates 210 and 220 to be joined is shown. Substrate 210 and substrate 220 may be joined using one or more of the joining techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another semiconductor substrate joining technique). Substrate 210 and substrate 220 may each comprise a semiconductor wafer, a semiconductor substrate, or another type of substrate on which semiconductor devices may be formed. Substrate 210 and substrate 220 may each be formed of silicon (Si), a silicon-containing material, a III-V compound semiconductor material (e.g., gallium arsenide (GaAs), silicon-on-insulator (SOI)), or another type of semiconductor material. Substrate 210 and substrate 220 may each be approximately circular and may each comprise a 200 mm wafer (e.g., a wafer with a diameter of approximately 200 mm), a 300 mm wafer (e.g., a wafer with a diameter of approximately 300 mm), or a wafer of another size.
[0024] like Figure 2AAs shown, substrate 210 may include a plurality of semiconductor devices 212, and substrate 220 may include a plurality of semiconductor devices 222. Semiconductor devices 212 and 222 may be formed using one or more semiconductor processing techniques described herein. By bonding substrates 210 and 220, the corresponding pairs of semiconductor devices 212 and 222 may be formed as three-dimensional semiconductor devices, including transistors (e.g., planar transistors, fin field-effect transistors (finFETs), gate all-around (GAA) transistors), memory devices, microelectromechanical systems (MEMS), application-specific integrated circuits (ASICs), system-on-chips (SoCs), and / or other types of semiconductor devices.
[0025] Semiconductor devices 212 may be arranged in a grid on substrate 210, such that a plurality of cleavages 214 separate rows (or columns) of semiconductor devices 212, and a plurality of cleavages 216 separate columns (or rows) of semiconductor devices 212. The cleavages 214 may be located in substrate 210 along a first direction (e.g., along...). Figure 2A The x-axis shown is formed, and the cut path 216 may be formed in the substrate 210 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 2A The y-axis (as shown) is formed such that cleavage 214 and cleavage 216 intersect at various locations across the substrate 210. Similarly, semiconductor devices 222 may be arranged in a grid on the substrate 220 such that a plurality of cleavage 224 separate rows (or columns) of semiconductor devices 222, and a plurality of cleavage 226 separate columns (or rows) of semiconductor devices 222. Cleavage 224 may be arranged in the substrate 220 along a first direction (e.g., along...). Figure 2A The x-axis shown is formed, and the cut path 226 may be formed in the substrate 220 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 2A The y-axis shown is formed such that cut channels 224 and 226 intersect at various locations across the substrate 220.
[0026] The dicing path may include trenches, recesses, layers, and / or other structures that serve as guides for dicing the substrate into multiple individual dies. For example, dicing path 214 in substrate 210 and dicing path 224 in substrate 220 may be used in a first direction of substrate 210 and substrate 220 (e.g., along...). Figure 2AThe x-axis shown represents a guide that cuts (or slits) the paired semiconductor devices 212 and 222 into individual dies. As another example, cleavage 216 in substrate 210 and cleavage 226 in substrate 220 can be used as guides in a second direction (e.g., along) of substrates 210 and 220. Figure 2A The y-axis (as shown) serves as a guide for cutting (or slicing) the paired semiconductor devices 212 and 222 into individual dies. Furthermore, cleaving tracks 214, 216, 224, and 226 can provide a buffer area between semiconductor devices 212 and 222 to accommodate the kerf (or blade width) of the die-cutting apparatus, allowing the die-cutting apparatus to cut semiconductor devices 212 and 222 into individual dies without cutting into areas of semiconductor devices 212 and 222. Cleaving tracks 214, 216, 224, and / or 226 may include other features such as lithographic alignment marks, overlay marks, thickness measurement structures, electrical test points, and / or other features used to assist in the manufacturing of semiconductor devices 212 and 222.
[0027] The dices 214, 216, 224, and / or 226 can be formed using various semiconductor fabrication processes. In some embodiments, a deposition system 102 forms a photoresist layer on a substrate 210, an exposure system 104 exposes the photoresist layer to a radiation source to pattern the photoresist layer, a developing system 106 develops and removes portions of the photoresist layer to expose the pattern, and an etching system 108 etches portions of the substrate 210 to form openings or recesses in the substrate 210. In some embodiments, the deposition system 102 uses various PVD, CVD, and / or ALD technologies (e.g., sputtering, PECVD, HDP-CVD, SACVD, and / or PEALD, etc.) to deposit one or more layers in the openings or recesses. In some embodiments, techniques such as polysilicon etching, metal etching, and / or another type of etching technology can be used to etch (e.g., using etching system 108) one or more layers formed in the dices 214, 216, 224, and / or 226.
[0028] like Figure 2A As further shown, each of the cutting channels 214, 216, 224, and 226 may have a specific width along the shorter dimension of the cutting channels 214, 216, 224, and 226. The cutting channels 214, 224, and 226 may all have a similar width (w1), which may be in the range of approximately 60 micrometers to approximately 80 micrometers or in the range of approximately 30 micrometers to approximately 100 micrometers, etc.
[0029] During the bonding operation, the diameter of substrate 210 and / or substrate 220 may increase due to thermal expansion. As an example, for a 200 mm substrate, the substrate may expand from approximately 10 micrometers to approximately 100 micrometers. The rate of expansion of the substrate may be based on various factors and / or properties of the substrate, such as the size of the substrate, the material composition of the substrate, and / or the structural composition of the substrate (e.g., films with higher compressive stress may induce a lower rate of expansion), etc.
[0030] Substrate 210 may be referred to as a low-expansion wafer, and substrate 220 may be referred to as a high-expansion wafer, because substrate 220 may have a greater rate of expansion relative to substrate 210. Therefore, dicing 216 may have a greater width (w1) relative to the width of dicing 226 (and dicings 214 and 224) to compensate for the greater thermal expansion of substrate 220 relative to substrate 210. The increased width of dicing 216 relative to dicing 226 may cause semiconductor devices 212 and 222 to be misaligned along the x-axis (e.g., in a first direction) at the start of the bonding operation of bonding substrates 210 and 220. However, as substrates 210 and 220 are heated during the bonding operation (causing them to expand), the combination of the increased width of dicing 216 and the greater rate of expansion of substrate 220 causes semiconductor devices 212 and 222 to move to approximately alignment along the x-axis.
[0031] To further facilitate alignment between semiconductor devices 212 and 222, the width of the cleavage 216 may gradually increase from near the center of substrate 210 toward the edge (or periphery) of substrate 210. This gradual increase in width compensates for the increased rate of expansion from near the center of substrate 220 to the edge of substrate 220, which allows for greater displacement of semiconductor devices 222 near the edge of substrate 220 relative to semiconductor devices 222 near the center of substrate 220 during bonding. Therefore, the width of the cleavage 216 may increase toward the edge of substrate 210, causing the spacing between semiconductor devices 212 to increase toward the edge of substrate 210 to compensate for the greater displacement of semiconductor devices 222 near the edge of substrate 220 during bonding. In this way, substrate 210 may include multiple subsets of cleavage 216, each subset having a different width relative to other subsets of cleavage 216.
[0032] As an example, dicing 216 may include a first subset 216a formed by one or more dicings and a second subset 216b formed by one or more dicings. The first subset 216a may be located closer to the center of the substrate 210 relative to the second subset 216b. The width (w2) of the second subset 216b of dicing 216 may be larger than the width (w3) of the first subset 216a of dicing 216. In some embodiments, the width of dicing 216 gradually increases from near the center of the substrate 210 toward the edge of the substrate 210, from dicing to dicing, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 214, 224, and 226 is approximately 80 micrometers, then the width (w3) of the first subset 216a may range from approximately 81 micrometers to approximately 84 micrometers, the width (w2) of the second subset 216b may range from approximately 82 micrometers to approximately 88 micrometers, and so on. In some embodiments, the cleavage 216 comprises a larger subset of cleavages that gradually increase in width from near the center of substrate 210 to near the edge of substrate 210.
[0033] Figure 2B and Figure 2C A cross-sectional elevation view showing the bonding operation of the bonding substrates 210 and 220 (along...) Figure 2A (See x-axis). It can be used in conjunction with the joining machine 110 to perform joining operations. For example... Figure 2B As shown, substrates 210 and 220 can be arranged such that semiconductor device 212 faces semiconductor device 222. Substrates 210 and 220 can be pre-bonded by placing semiconductor device 212 on and / or abutting against semiconductor device 222 (or vice versa). Figure 2B As further shown, at least one subset of semiconductor device 212 and at least one subset of semiconductor device 222 are misaligned during the pre-bonding stage. This is because the width of dicing 216 (e.g., subsets 216a and 216b) is larger than the width of dicing 226. Furthermore, as the widths of subsets 216a and 216b gradually increase, the amount or magnitude of misalignment can increase from near the center of substrates 210 and 220 to near the edges of substrates 210 and 220.
[0034] like Figure 2C As shown, after the bonding operation, semiconductor device 212 and semiconductor device 222 can be approximately aligned. As explained above, the different expansion rates of substrates 210 and 220 can cause a lateral displacement of semiconductor device 222 (e.g., along...) Figure 2AThe x-axis (in the diagram) is larger relative to semiconductor device 212. The gradual increase in the width of the dicing tracks 216 (e.g., subsets 216a and 216b) compensates for the larger lateral displacement of semiconductor device 222, because the lateral displacement of semiconductor device 222 allows for alignment of semiconductor device 222 with semiconductor device 212, such as... Figure 2C As shown.
[0035] The approximate alignment of semiconductor device 212 and semiconductor device 222 after bonding can be determined based on alignment in one or more directions and / or along one or more axes. For example, the alignment of semiconductor device 212 and semiconductor device 222 can be analyzed for alignment along the x-axis (also known as left / right alignment) and / or alignment along the y-axis (also known as top / bottom alignment). In some embodiments, the x-axis is defined as a vector extending from an alignment mark (e.g., a wafer notch) on the substrate across the widest point of the substrate from a first edge to a second edge of the substrate. In some embodiments, the y-axis is approximately perpendicular to the x-axis and is defined as a vector extending from the first edge of the substrate across the widest point of the substrate to a second edge of the substrate.
[0036] In some implementations, the apparatus (e.g., the joining machine 110, the joining...) Figures 10A to 10D The described processing apparatus 1002 and / or another apparatus generates alignment data and determines, based on the alignment data, whether semiconductor device 212 and semiconductor device 222 are approximately aligned. The alignment data may include a misalignment index (e.g., for the x-axis and y-axis) based on the stacking test pattern of substrates 210 and 220. The alignment index may represent a percentage of the non-overlapping area of semiconductor device 212 and semiconductor device 222 (e.g., the surface area of semiconductor device 212 that does not overlap with the surface area of semiconductor device 222, and vice versa), or a ratio between the non-overlapping area of semiconductor device 212 and semiconductor device 222 and the total surface area.
[0037] In some embodiments, the device may determine approximate alignment of semiconductor device 212 and semiconductor device 222 based on a percentage of non-overlapping area equal to or less than approximately 0.005%. In some embodiments, the device may determine approximate alignment of semiconductor device 212 and semiconductor device 222 based on a percentage of non-overlapping area equal to or less than approximately 0.002%. The techniques described herein can reduce misalignment between joined semiconductor devices by approximately 60% or more.
[0038] As indicated above, Figures 2A to 2C It is provided as an instance. Other instances may be related to... Figures 2A to 2C The content they describe is different.
[0039] Figure 3 This is a figure illustrating the exemplary embodiment 300 described herein. Exemplary embodiment 300 is similar to exemplary embodiment 200 in that it includes diced channels of gradually varying sizes on a single substrate of the two substrates to be joined. Additionally, exemplary embodiment 300 includes multiple sets of diced channels of gradually varying sizes along multiple directions on a single substrate to compensate for thermal expansion of the substrate in multiple directions. In this way, the multiple sets of diced channels of gradually varying sizes can be used to compensate for radiation-symmetric thermal expansion to achieve approximate alignment of semiconductor devices on the two substrates in multiple directions.
[0040] like Figure 3 As shown, exemplary embodiment 300 may include substrates 310 and 320 to be bonded. Substrate 310 and substrate 320 may be bonded using one or more of the bonding techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another type of semiconductor substrate bonding technique). Substrate 310 may include a plurality of semiconductor devices 312, and substrate 320 may include a plurality of semiconductor devices 322. Semiconductor devices 312 and 322 may be formed using one or more semiconductor processing techniques described herein. By bonding substrates 310 and substrate 320, the corresponding pair of semiconductor devices 312 and 322 may be formed as three-dimensional semiconductor devices, including transistors, memory devices, MEMs, ASICs, SoCs, and / or other types of semiconductor devices.
[0041] Semiconductor devices 312 may be arranged in a grid on substrate 310, such that a plurality of cleavage tracks 314 separate rows (or columns) of semiconductor devices 312, and a plurality of cleavage tracks 316 separate columns (or rows) of semiconductor devices 312. The cleavage tracks 314 may be located in substrate 310 along a first direction (e.g., along...). Figure 3 The x-axis shown is formed, and the cut path 316 may be formed in the substrate 310 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 3 The y-axis (as shown) is formed such that cleavage 314 and cleavage 316 intersect at various locations across the substrate 310. Similarly, semiconductor devices 322 may be arranged in a grid on the substrate 320 such that a plurality of cleavage 324 separate rows (or columns) of semiconductor devices 322, and a plurality of cleavage 326 separate columns (or rows) of semiconductor devices 322. Cleavage 324 may be arranged in the substrate 320 along a first direction (e.g., along...). Figure 3 The x-axis shown is formed, and the cut path 326 may be formed in the substrate 320 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 3 The y-axis shown is formed such that cut channels 324 and 326 intersect at various locations across the substrate 320.
[0042] like Figure 3 As further shown, each of the cutting channels 314, 316, 324, and 326 may have a specific width along the shorter dimension of the cutting channels 314, 316, 324, and 326. Cutting channels 324 and 326 may have similar widths (w1), which may be in the range of approximately 60 micrometers to approximately 80 micrometers or in the range of approximately 30 micrometers to approximately 100 micrometers, etc.
[0043] During the bonding operation, due to the thermal expansion of substrates 310 and 320, substrates 310 and 320 may undergo radially symmetrical thermal expansion. Therefore, the dimensions of each of substrates 310 and 320 can be increased in a radially symmetrical manner, such that the diameters of substrates 310 and 320 can be increased along... Figure 3 The x-axis and y-axis are shown as increased. Furthermore, substrate 310 can be referred to as a low-expansion wafer, and substrate 320 can be referred to as a high-expansion wafer, because substrate 320 can have a larger expansion rate relative to substrate 310. To compensate for the greater radiation-symmetric thermal expansion of substrate 320, cleaving channels 314 and 316 can have a wider width (w1) relative to the widths of cleaving channels 324 and 326.
[0044] The increased width of dicing 316 relative to dicing 326 may cause semiconductor devices 312 and 322 to be misaligned along the x-axis (e.g., in a first direction) at the start of the bonding operation between bonding substrates 310 and 320. However, as substrates 310 and 320 are heated during the bonding operation (causing them to expand), the combination of the increased width of dicing 316 and the greater expansion rate of substrate 320 causes semiconductor devices 312 and 322 to move to approximately alignment along the x-axis.
[0045] Similarly, the increased width of dicing 314 relative to dicing 324 can cause semiconductor devices 312 and 322 to be misaligned along the y-axis (e.g., in a second direction) at the start of the bonding operation between bonding substrates 310 and 320. However, as substrates 310 and 320 are heated during the bonding operation (causing them to expand), the combination of the increased width of dicing 314 and the greater expansion rate of substrate 320 causes semiconductor devices 312 and 322 to move into near alignment along the y-axis.
[0046] To further facilitate alignment between semiconductor devices 312 and 322 along the x-axis (e.g., in the first direction), the width of the cleavage 316 may gradually increase from near the center of substrate 310 toward the edge (or periphery) of substrate 310. This gradual increase in width compensates for the increased rate of expansion along the x-axis (e.g., in the first direction) from near the center of substrate 320 to the edge of substrate 320, which can cause a greater displacement of semiconductor devices 322 near the edge of substrate 320 relative to semiconductor devices 322 near the center of substrate 320 during bonding. Therefore, the width of the cleavage 316 may increase toward the edge of substrate 310, causing the spacing between semiconductor devices 312 to increase toward the edge of substrate 310 to compensate for the greater displacement of semiconductor devices 322 near the edge of substrate 320 during bonding. In this way, substrate 310 may include multiple subsets of cleavage 316, each subset having a different width relative to other subsets of cleavage 316.
[0047] As an example, dicing 316 may include a first subset 316a formed by one or more dicings and a second subset 316b formed by one or more dicings. The first subset 316a may be located closer to the center of the substrate 310 relative to the second subset 316b. The width (w2) of the second subset 316b of dicing 316 may be larger than the width (w3) of the first subset 316a of dicing 316. In some embodiments, the width of dicing 316 may gradually increase from near the center of the substrate 310 toward the edge of the substrate 310, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 324 and 326 is approximately 60 micrometers, then the width (w3) of the first subset 316a may be in the range of approximately 61 micrometers to approximately 64 micrometers, the width (w2) of the second subset 316b may be in the range of approximately 62 micrometers to approximately 68 micrometers, and so on. In some embodiments, the dicing track 316 includes a larger subset of tracks whose width gradually increases from near the center of the substrate 310 to near the edge of the substrate 310.
[0048] To further facilitate alignment between semiconductor devices 312 and 322 along the y-axis (e.g., in a second direction), the width of the cleavage 314 may gradually increase from near the center of substrate 310 toward the edge (or periphery) of substrate 310. This gradual increase in width compensates for the increased rate of expansion along the y-axis (e.g., in a second direction) from near the center of substrate 320 to the edge of substrate 320, which can cause a greater displacement of semiconductor devices 322 near the edge of substrate 320 relative to semiconductor devices 322 near the center of substrate 320 during bonding. Therefore, the width of the cleavage 314 may increase toward the edge of substrate 310, causing the spacing between semiconductor devices 312 to increase toward the edge of substrate 310 to compensate for the greater displacement of semiconductor devices 322 near the edge of substrate 320 during bonding. In this way, substrate 310 may include multiple subsets of cleavage 314, each subset having a different width relative to other subsets of cleavage 314.
[0049] As an example, dicing 314 may include a first subset 314a formed by one or more dicings and a second subset 314b formed by one or more dicings. The first subset 314a may be located closer to the center of the substrate 310 relative to the second subset 314b. The width (w4) of the second subset 314b of dicing 314 may be larger than the width (w5) of the first subset 314a of dicing 314. In some embodiments, the width of dicing 314 may gradually increase from near the center of the substrate 310 toward near the edge of the substrate 310, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 324 and 326 is approximately 70 micrometers, then the width (w5) of the first subset 314a may be in the range of approximately 71 micrometers to approximately 74 micrometers, the width (w4) of the second subset 314b may be in the range of approximately 72 micrometers to approximately 78 micrometers, and so on. In some embodiments, the cleavage 314 comprises a larger subset of cleavages that gradually increase in width from near the center of the substrate 310 toward the edge of the substrate 310.
[0050] As indicated above, the widths of the cleavages 314 and 316 may gradually increase from near the center of substrate 310 to near the edge of substrate 310 to compensate for the radially symmetric thermal expansion of substrate 320. Therefore, the increase in the widths of cleavages 314 and 316 may be approximately equal to compensate for the radially symmetric thermal expansion of substrate 320. In some embodiments, the increase in the widths of cleavages 314 and 316 may differ to compensate for the asymmetric thermal expansion of substrate 320, for example, when different types of devices are formed along the x-axis and y-axis of substrate 320 (which can result in asymmetric thermal expansion rates along the x-axis and y-axis of substrate 320).
[0051] As indicated above, Figure 3 It is provided as an instance. Other instances may be related to... Figure 3 The content they describe is different.
[0052] Figure 4 This is a figure illustrating an exemplary embodiment 400 described herein. Exemplary embodiment 400 may include an example of a dimensionally varying diced channel included on a single substrate of two substrates to be joined. Furthermore, the dimensionally varying diced channel may be included in the substrate along a single direction to compensate for thermal expansion of the substrate in that single direction. Exemplary embodiment 400 differs from exemplary embodiment 200 in that the width of the dimensionally varying diced channel decreases from the center of the high-expansion-rate substrate to be joined with the low-expansion-rate substrate towards the edge of the high-expansion-rate substrate. In this way, the decreasing width of the dimensionally varying diced channel facilitates alignment of semiconductor devices on the substrates to be joined, while the reduced diced channel size allows for an increase in semiconductor device density on the substrates.
[0053] Figure 4 A top view (or plan view) of substrates 410 and 420 to be joined is shown. Substrate 410 and substrate 420 may be joined using one or more of the joining techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another type of semiconductor substrate bonding technique). Substrate 410 and substrate 420 may each comprise a semiconductor wafer, a semiconductor substrate, or another type of substrate on which semiconductor devices may be formed and / or thereon.
[0054] like Figure 4 As shown, substrate 410 may include a plurality of semiconductor devices 412, and substrate 420 may include a plurality of semiconductor devices 422. The semiconductor devices 412 may be arranged in a grid on substrate 410 such that a plurality of cleavages 414 separate rows (or columns) of semiconductor devices 412, and a plurality of cleavages 416 separate columns (or rows) of semiconductor devices 412. The cleavages 414 may be located in substrate 410 along a first direction (e.g., along...). Figure 4 The x-axis shown is formed, and the cut path 416 may be formed in the substrate 410 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 4 The y-axis (as shown) is formed such that cleavage 414 and cleavage 416 intersect at various locations across the substrate 410. Similarly, semiconductor devices 422 may be arranged in a grid on the substrate 420 such that a plurality of cleavage 424 separate rows (or columns) of semiconductor devices 422, and a plurality of cleavage 426 separate columns (or rows) of semiconductor devices 422. Cleavage 424 may be formed in the substrate 420 along a first direction (e.g., along...). Figure 4The x-axis shown is formed, and the cut path 426 may be formed in the substrate 420 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 4 The y-axis shown is formed such that cut channels 424 and 426 intersect at various locations across the substrate 420.
[0055] like Figure 4 As further shown, each of the cutting channels 414, 416, 424, and 426 may have a specific width along the shorter dimension of the cutting channels 414, 416, 424, and 426. The cutting channels 414, 416, and 424 may all have a similar width (w1), which may be in the range of approximately 60 micrometers to approximately 80 micrometers or in the range of approximately 30 micrometers to approximately 100 micrometers, etc.
[0056] Substrate 410 may be referred to as a low-expansion wafer, and substrate 420 may be referred to as a high-expansion wafer because substrate 420 may have a greater rate of expansion relative to substrate 410. Therefore, the dimensions (e.g., diameter) of substrate 420 may expand by a greater amount relative to substrate 410. In exemplary embodiment 400, cleavage 426 may have a smaller width (w1) relative to the widths (w1) of cleavages 414, 416, and 424 to compensate for the greater thermal expansion of substrate 420 relative to the thermal expansion of substrate 410. The reduced width of cleavage 426 relative to cleavage 416 may cause semiconductor devices 412 and 422 to be misaligned along the x-axis (e.g., in a first direction) at the start of the bonding operation between substrate 410 and substrate 420. However, as substrates 410 and 420 are heated during the bonding operation (which causes substrates 410 and 420 to expand), the combination of the reduced width of the dicing 426 and the greater expansion rate of substrate 420 causes semiconductor devices 412 and 422 to move into near alignment along the x-axis.
[0057] To further facilitate alignment between semiconductor devices 412 and 422, the width of the cleavage 426 may gradually decrease from near the center of substrate 420 toward the edge (or periphery) of substrate 420. This gradual decrease in width compensates for the increased rate of expansion from near the center of substrate 420 toward the edge of substrate 420. This increased expansion allows for greater displacement of semiconductor devices 422 near the edge of substrate 420 relative to semiconductor devices 422 near the center of substrate 420 during bonding. Therefore, the width of the cleavage 426 may decrease toward the edge of substrate 420, thereby reducing the spacing between semiconductor devices 422 toward the edge of substrate 420 to compensate for the greater displacement of semiconductor devices 422 near the edge of substrate 420 during bonding. In this way, substrate 420 may include multiple subsets of cleavage 426, each subset having a different width relative to other subsets of cleavage 426.
[0058] As an example, dicing 426 may include a first subset 426a formed by one or more dicings and a second subset 426b formed by one or more dicings. The first subset 426a may be located closer to the center of the substrate 420 relative to the second subset 426b. The width (w2) of the second subset 426b of dicing 426 may be smaller than the width (w3) of the first subset 426a of dicing 426. In some embodiments, the width of dicing 426 gradually decreases from near the center of the substrate 420 toward the edge of the substrate 420, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 414, 416, and 424 is approximately 90 micrometers, then the width (w3) of the first subset 426a may be in the range of approximately 86 micrometers to approximately 89 micrometers, the width (w2) of the second subset 426b may be in the range of approximately 82 micrometers to approximately 88 micrometers, and so on. In some embodiments, the cleavage 426 comprises a larger subset of cleavages that gradually decrease in width from near the center of the substrate 420 to near the edge of the substrate 420.
[0059] As indicated above, Figure 4 It is provided as an instance. Other instances may be related to... Figure 4 The content they describe is different.
[0060] Figure 5 This is a figure illustrating the exemplary embodiment 500 described herein. Exemplary embodiment 500 is similar to exemplary embodiment 400 in that it includes dicing channels with progressively decreasing widths on a single substrate of the two substrates to be joined. Additionally, exemplary embodiment 500 includes multiple sets of dicing channels with progressively decreasing widths along multiple directions on a single substrate to compensate for thermal expansion in multiple directions. In this way, the multiple sets of dicing channels with progressively decreasing widths can be used to compensate for radiation-symmetric thermal expansion to achieve approximate alignment of semiconductor devices on the two substrates in multiple directions.
[0061] like Figure 5As shown, an exemplary embodiment 500 may include substrates 510 and 520 to be bonded. Substrate 510 and substrate 520 may be bonded using one or more of the bonding techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another type of semiconductor substrate bonding technique). Substrate 510 may include a plurality of semiconductor devices 512, and substrate 520 may include a plurality of semiconductor devices 522. Semiconductor devices 512 and 522 may be formed using one or more semiconductor processing techniques described herein. By bonding substrates 510 and substrate 520, the corresponding pair of semiconductor devices 512 and 522 may be formed as three-dimensional semiconductor devices, including transistors, memory devices, MEMs, ASICs, SoCs, and / or other types of semiconductor devices.
[0062] Semiconductor devices 512 may be arranged in a grid on substrate 510, such that a plurality of cleavage tracks 514 separate rows (or columns) of semiconductor devices 512, and a plurality of cleavage tracks 516 separate columns (or rows) of semiconductor devices 512. The cleavage tracks 514 may be located in substrate 510 along a first direction (e.g., along...). Figure 5 The x-axis shown is formed, and the cut path 516 may be formed in the substrate 510 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 5 The y-axis (as shown) is formed such that cleavage 514 and cleavage 516 intersect at various locations across the substrate 510. Similarly, semiconductor devices 522 may be arranged in a grid on the substrate 520 such that a plurality of cleavage 524 separate rows (or columns) of semiconductor devices 522, and a plurality of cleavage 526 separate columns (or rows) of semiconductor devices 522. Cleavage 524 may be formed in the substrate 520 along a first direction (e.g., along...). Figure 5 The x-axis shown is formed, and the cut path 526 may be formed in the substrate 520 along a second (e.g., vertical) direction (e.g., along...). Figure 5 The y-axis shown is formed such that cut channels 524 and 526 intersect at various locations across the substrate 520.
[0063] like Figure 5 As further shown, each of the cutting tracks 514, 516, 524, and 526 may have a specific width along the shorter dimension of the cutting tracks 514, 516, 524, and 526. The cutting tracks 514 and 516 may have similar widths (w1), which may be in the range of approximately 60 micrometers to approximately 80 micrometers or in the range of approximately 30 micrometers to approximately 100 micrometers, etc.
[0064] During the bonding operation, due to the thermal expansion of substrates 510 and 520, substrates 510 and 520 may undergo radially symmetrical thermal expansion. Therefore, the dimensions of each of substrates 510 and 520 can be increased in a radially symmetrical manner, such that the diameters of substrates 510 and 520 can be increased along... Figure 5 The x-axis and y-axis are shown as increased. Furthermore, substrate 510 can be referred to as a low-expansion wafer, and substrate 520 can be referred to as a high-expansion wafer, because substrate 520 can have a larger expansion rate relative to substrate 510. To compensate for the greater radiation-symmetric thermal expansion of substrate 520, dices 524 and 526 can have a smaller width relative to the width (w1) of dices 514 and 516.
[0065] The reduced width of dicing 526 relative to dicing 516 allows semiconductor devices 512 and 522 to be misaligned along the x-axis (e.g., in a first direction) at the start of the bonding operation between bonding substrates 510 and 520. However, as substrates 510 and 520 are heated during the bonding operation (causing them to expand), the combination of the reduced width of dicing 526 and the greater expansion rate of substrate 520 causes semiconductor devices 512 and 522 to move into near-alignment along the x-axis.
[0066] Similarly, the reduced width of dicing 524 relative to dicing 514 can cause semiconductor devices 512 and 522 to be misaligned along the y-axis (e.g., in a second direction) at the start of the bonding operation between bonding substrates 510 and 520. However, as substrates 510 and 520 are heated during the bonding operation (causing them to expand), the combination of the reduced width of dicing 524 and the greater expansion rate of substrate 520 causes semiconductor devices 512 and 522 to move into near alignment along the y-axis.
[0067] To further facilitate alignment between semiconductor devices 512 and 522 along the x-axis (e.g., in the first direction), the width of the cleavage 526 may gradually decrease from near the center of substrate 520 toward the edge (or periphery) of substrate 520. This gradual decrease in width compensates for the increased rate of expansion along the x-axis (e.g., in the first direction) from near the center of substrate 520 toward the edge of substrate 520, which can cause a greater displacement of semiconductor devices 522 near the edge of substrate 520 relative to semiconductor devices 522 near the center of substrate 520 during bonding. Therefore, the width of the cleavage 526 may decrease toward the edge of substrate 520, causing the spacing between semiconductor devices 522 to decrease toward the edge of substrate 520 to compensate for the greater displacement of semiconductor devices 522 near the edge of substrate 520 during bonding. In this way, substrate 520 may include multiple subsets of cleavage 526, each subset having a different width relative to other subsets of cleavage 526.
[0068] As an example, dicing 526 may include a first subset 526a formed by one or more dicings and a second subset 526b formed by one or more dicings. The first subset 526a may be located closer to the center of the substrate 520 relative to the second subset 526b. The width (w2) of the second subset 526b of dicing 526 may be smaller than the width (w3) of the first subset 526a of dicing 526. In some embodiments, the width of dicing 526 gradually decreases from near the center of the substrate 520 toward the edge of the substrate 520, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 514 and 516 is approximately 100 micrometers, then the width (w3) of the first subset 526a may be in the range of approximately 96 micrometers to approximately 99 micrometers, the width (w2) of the second subset 526b may be in the range of approximately 92 micrometers to approximately 98 micrometers, and so on. In some embodiments, the dicing 526 comprises a larger subset of dicings that gradually decrease in width from near the center of the substrate 520 toward the edge of the substrate 520.
[0069] To further facilitate alignment between semiconductor devices 512 and 522 along the y-axis (e.g., in a second direction), the width of the cleavage 524 may gradually decrease from near the center of substrate 520 toward the edge (or periphery) of substrate 520. This gradual decrease in width compensates for the increased rate of expansion along the y-axis (e.g., in a second direction) from near the center of substrate 520 toward the edge of substrate 520, which can cause a greater displacement of semiconductor devices 522 near the edge of substrate 520 relative to semiconductor devices 522 near the center of substrate 520 during bonding. Therefore, the width of the cleavage 524 may decrease toward the edge of substrate 520, causing the spacing between semiconductor devices 522 to decrease along the y-axis (e.g., in a second direction) toward the edge of substrate 520 to compensate for the greater displacement of semiconductor devices 522 near the edge of substrate 520 during bonding. In this way, substrate 520 may include multiple subsets of cleavage 524, each subset having a different width relative to other subsets of cleavage 524.
[0070] As an example, dicing 524 may include a first subset 524a formed by one or more dicings and a second subset 524b formed by one or more dicings. The first subset 524a may be located closer to the center of the substrate 520 relative to the second subset 524b. The width (w4) of the second subset 524b of dicing 524 may be smaller than the width (w5) of the first subset 524a of dicing 524. In some embodiments, the width of dicing 524 gradually decreases from near the center of the substrate 520 toward the edge of the substrate 520, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 514 and 516 is approximately 100 micrometers, then the width (w5) of the first subset 524a may be in the range of approximately 96 micrometers to approximately 99 micrometers, the width (w4) of the second subset 524b may be in the range of approximately 92 micrometers to approximately 98 micrometers, and so on. In some embodiments, the dicing track 524 comprises a larger subset of tracks whose width gradually decreases from near the center of the substrate 520 to near the edge of the substrate 520.
[0071] As indicated above, the widths of the cleavages 524 and 526 can gradually decrease from near the center of the substrate 520 to near the edge of the substrate 520 to compensate for the radially symmetrical thermal expansion of the substrate 520. Therefore, the reduction in the widths of the cleavages 524 and 526 can be approximately equal to compensate for the radially symmetrical thermal expansion of the substrate 520. In some embodiments, the reduction in the widths of the cleavages 524 and 526 can be different to compensate for the asymmetric thermal expansion of the substrate 520, for example, when different types of devices are formed along the x-axis and y-axis of the substrate 520 (which can result in asymmetric thermal expansion rates along the x-axis and y-axis of the substrate 520).
[0072] As indicated above, Figure 5 It is provided as an instance. Other instances may be related to... Figure 5 The content they describe is different.
[0073] Figure 6 This is a figure illustrating an exemplary embodiment 600 described herein. Exemplary embodiment 600 includes, for example, a first set of dicing channels with gradually increasing width on a low-thermal-rate substrate to compensate for thermal expansion in a first direction, and a second set of dicing channels with gradually decreasing width on a high-thermal-rate substrate to compensate for thermal expansion in a second direction. In this way, multiple sets of dicing channels can be used to compensate for radiation-symmetric thermal expansion to achieve approximate alignment of semiconductor devices on two substrates in multiple directions.
[0074] like Figure 6 As shown, an exemplary embodiment 600 may include substrates 610 and 620 to be bonded. Substrate 610 and substrate 620 may be bonded using one or more of the bonding techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another type of semiconductor substrate bonding technique). Substrate 610 may include a plurality of semiconductor devices 612, and substrate 620 may include a plurality of semiconductor devices 622. Semiconductor devices 612 and 622 may be formed using one or more semiconductor processing techniques described herein. By bonding substrates 610 and substrate 620, the corresponding pair of semiconductor devices 612 and 622 may be formed as three-dimensional semiconductor devices, including transistors, memory devices, MEMs, ASICs, SoCs, and / or other types of semiconductor devices.
[0075] Semiconductor devices 612 may be arranged in a grid on substrate 610, such that a plurality of cleavages 614 separate rows (or columns) of semiconductor devices 612, and a plurality of cleavages 616 separate columns (or rows) of semiconductor devices 612. The cleavages 614 may be located in substrate 610 along a first direction (e.g., along...). Figure 6 The x-axis shown is formed, and the cut path 616 may be formed in the substrate 610 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 6The y-axis (as shown) is formed such that cleavage 614 and cleavage 616 intersect at various locations across the substrate 610. Similarly, semiconductor devices 622 may be arranged in a grid on the substrate 620 such that a plurality of cleavage 624 separate rows (or columns) of semiconductor devices 622, and a plurality of cleavage 626 separate columns (or rows) of semiconductor devices 622. Cleavage 624 may be formed in the substrate 620 along a first direction (e.g., along...). Figure 6 The x-axis shown is formed, and the cut path 626 may be formed in the substrate 620 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 6 The y-axis shown is formed such that cut channels 624 and 626 intersect at various locations across the substrate 620.
[0076] like Figure 6 As further shown, each of the cutting channels 614, 616, 624, and 626 may have a specific width along the shorter dimension of the cutting channels 614, 616, 624, and 626. The cutting channels 614 and 626 may have similar widths (w1), which may be in the range of approximately 60 micrometers to approximately 80 micrometers or in the range of approximately 30 micrometers to approximately 100 micrometers, etc.
[0077] During the bonding operation, due to the thermal expansion of substrates 610 and 620, substrates 610 and 620 may undergo radially symmetrical thermal expansion. Therefore, the dimensions of each of substrates 610 and 620 can be increased in a radially symmetrical manner, such that the diameters of substrates 610 and 620 can be increased along... Figure 6 The x-axis and y-axis are shown as increased. Furthermore, substrate 610 can be referred to as a low-expansion wafer, and substrate 620 can be referred to as a high-expansion wafer, because substrate 620 can have a larger expansion rate relative to substrate 610. To compensate for the greater radiation-symmetric thermal expansion of substrate 620, cleaving track 616 can have a wider width (w1) relative to the widths (w1) of cleaving tracks 614 and 626, and cleaving track 624 can have a smaller width (w1) relative to the widths (w1) of cleaving tracks 614 and 626.
[0078] The larger width of cleavage 616 and the smaller width of cleavage 624 may cause semiconductor devices 612 and 622 to be misaligned along the x-axis (e.g., in a first direction) and along the y-axis (e.g., in a second direction) at the start of the bonding operation between bonding substrates 610 and 620. However, as substrates 610 and 620 are heated during the bonding operation (causing them to expand radially), the combination of the increased width of cleavage 616 and the greater expansion rate of substrate 620 causes semiconductor devices 612 and 622 to move into near-alignment along the x-axis. Furthermore, the reduced width of cleavage 624 and the greater expansion rate of substrate 620 cause semiconductor devices 612 and 622 to move into near-alignment along the y-axis.
[0079] To further facilitate alignment between semiconductor devices 612 and 622 along the x-axis (e.g., in the first direction), the width of the cleavage 616 may gradually increase from near the center of substrate 610 toward the edge (or periphery) of substrate 610. This gradual increase in width compensates for the increased rate of expansion along the x-axis (e.g., in the first direction) from near the center of substrate 620 toward the edge of substrate 620, which can cause a greater displacement of semiconductor devices 622 near the edge of substrate 620 relative to semiconductor devices 622 near the center of substrate 620 during bonding. Therefore, the width of the cleavage 616 may increase toward the edge of substrate 610, causing the spacing between semiconductor devices 612 to increase toward the edge of substrate 610 to compensate for the greater displacement of semiconductor devices 622 near the edge of substrate 620 during bonding. In this way, substrate 610 may include multiple subsets of cleavage 616, each subset having a different width relative to other subsets of cleavage 616.
[0080] As an example, dicing 616 may include a first subset 616a formed by one or more dicings and a second subset 616b formed by one or more dicings. The first subset 616a may be located closer to the center of the substrate 610 relative to the second subset 616b. The width (w2) of the second subset 616b of dicing 616 may be larger than the width (w3) of the first subset 616a of dicing 616. In some embodiments, the width of dicing 616 gradually increases from near the center of the substrate 610 toward the edge of the substrate 610, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 614 and 626 is approximately 80 micrometers, then the width (w3) of the first subset 616a may be in the range of approximately 81 micrometers to approximately 84 micrometers, the width (w2) of the second subset 616b may be in the range of approximately 82 micrometers to approximately 88 micrometers, and so on. In some embodiments, the cleavage 616 includes a larger subset of cleavages that gradually increase in width from near the center of the substrate 610 to near the edge of the substrate 610.
[0081] To further facilitate alignment between semiconductor devices 612 and 622 along the y-axis (e.g., in a second direction), the width of the cleavage 624 may gradually decrease from near the center of substrate 620 toward the edge (or periphery) of substrate 620. This gradual decrease in width compensates for the increased rate of expansion along the y-axis (e.g., in a second direction) from near the center of substrate 620 to the edge of substrate 620, which can cause a greater displacement of semiconductor devices 622 near the edge of substrate 620 relative to semiconductor devices 622 near the center of substrate 620 during bonding. Therefore, the width of the cleavage 624 may decrease toward the edge of substrate 620, causing the spacing between semiconductor devices 622 to decrease along the y-axis (e.g., in a second direction) toward the edge of substrate 620 to compensate for the greater displacement of semiconductor devices 622 near the edge of substrate 620 during bonding. In this way, substrate 620 may include multiple subsets of cleavage 624, each subset having a different width relative to other subsets of cleavage 624.
[0082] As an example, dicing 624 may include a first subset 624a formed by one or more dicings and a second subset 624b formed by one or more dicings. The first subset 624a may be located closer to the center of the substrate 620 relative to the second subset 624b. The width (w4) of the second subset 624b of dicing 624 may be smaller than the width (w5) of the first subset 624a of dicing 624. In some embodiments, the width of dicing 624 gradually decreases from near the center of the substrate 620 toward the edge of the substrate 620, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 614 and 626 is approximately 80 micrometers, then the width (w5) of the first subset 624a may be in the range of approximately 76 micrometers to approximately 79 micrometers, the width (w4) of the second subset 624b may be in the range of approximately 72 micrometers to approximately 78 micrometers, and so on. In some embodiments, the cleavage 624 includes a larger subset of cleavages that gradually decrease in width from near the center of the substrate 620 to near the edge of the substrate 620.
[0083] As indicated above, the width of cleavage 616 can be gradually increased and the width of cleavage 624 can be gradually decreased to compensate for the difference in radially symmetrical thermal expansion between substrates 610 and 620. Therefore, an increase in the width of cleavage 616 can approximately equal a decrease in the width of cleavage 624 to compensate for the radially symmetrical thermal expansion of substrates 610 and 620. In some embodiments, the increase in the width of cleavage 616 can differ from the decrease in the width of cleavage 624 to compensate for asymmetric thermal expansion of substrates 610 and / or 620, for example, in cases where different types of devices are formed along the x-axis and y-axis of substrates 610 and / or 620 (which can result in asymmetric thermal expansion rates along the x-axis and y-axis of substrates 610 and / or 620).
[0084] As indicated above, Figure 6 It is provided as an instance. Other instances may be related to... Figure 6 The content they describe is different.
[0085] Figure 7This is a figure illustrating an exemplary embodiment 700 described herein. Exemplary embodiment 700 may include, for example, dicing channels with gradually increasing widths on a low-expansion-rate substrate and dicing channels with gradually decreasing widths on a high-expansion-rate substrate. The dicing channels with gradually increasing widths and the dicing channels with gradually decreasing widths are arranged in the same direction (e.g., along the same axis) to compensate for the difference in thermal expansion between the low-expansion-rate substrate and the high-expansion-rate substrate. In this way, the dicing channels with gradually increasing widths in the low-expansion-rate substrate and the dicing channels with gradually decreasing widths in the high-expansion-rate substrate can be used to compensate for a large difference in expansion rates in a single direction between the low-expansion-rate substrate and the high-expansion-rate substrate (e.g., to compensate for a larger difference in expansion rates compared to including only dicing channels with gradually increasing widths or only dicing channels with gradually decreasing widths).
[0086] like Figure 7 As shown, exemplary embodiment 700 may include substrates 710 and 720 to be bonded. Substrate 710 and substrate 720 may be bonded using one or more of the bonding techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another type of semiconductor substrate bonding technique). Substrate 710 may include a plurality of semiconductor devices 712, and substrate 720 may include a plurality of semiconductor devices 722. Semiconductor devices 712 and 722 may be formed using one or more semiconductor processing techniques described herein. By bonding substrates 710 and substrate 720, the corresponding pair of semiconductor devices 712 and 722 may be formed as three-dimensional semiconductor devices, including transistors, memory devices, MEMs, ASICs, SoCs, and / or other types of semiconductor devices.
[0087] Semiconductor devices 712 may be arranged in a grid on substrate 710, such that a plurality of dices 714 separate rows (or columns) of semiconductor devices 712, and a plurality of dices 716 separate columns (or rows) of semiconductor devices 712. The dices 714 may be located in substrate 710 along a first direction (e.g., along...). Figure 7 The x-axis shown is formed, and the cut path 716 may be formed in the substrate 710 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 7 The y-axis (as shown) is formed such that dices 714 and 716 intersect at various locations across the substrate 710. Similarly, semiconductor devices 722 may be arranged in a grid on the substrate 720 such that a plurality of dices 724 separate rows (or columns) of semiconductor devices 722, and a plurality of dices 726 separate columns (or rows) of semiconductor devices 722. The dices 724 may be in the substrate 720 along a first direction (e.g., along...). Figure 7 The x-axis shown is formed, and the cut path 726 may be formed in the substrate 720 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 7The y-axis shown is formed such that cut channels 724 and 726 intersect at various locations across the substrate 720.
[0088] like Figure 7 As further shown, each of the cutting channels 714, 716, 724, and 726 may have a specific width along the shorter dimension of the cutting channels 714, 716, 724, and 726. The cutting channels 714 and 724 may have similar widths (w1), which may be in the range of approximately 60 micrometers to approximately 80 micrometers or in the range of approximately 30 micrometers to approximately 100 micrometers, etc.
[0089] Substrate 710 can be referred to as a low-expansion wafer, and substrate 720 can be referred to as a high-expansion wafer because substrate 720 can have a larger expansion rate relative to substrate 710. To compensate for the expansion rate of substrate 720 along... Figure 7 Due to the greater thermal expansion along the x-axis, cleavage 716 and cleavage 726 can have widths (w1) different from those of cleavage 714 and 724. Specifically, since substrate 710 is a low thermal expansion substrate, the width of cleavage 716 can be larger than the width (w1) of cleavage 714 and 724. Furthermore, since substrate 720 is a high thermal expansion substrate, the width of cleavage 726 can be smaller than the width (w1) of cleavage 714 and 724 (and the width of cleavage 716). As described above, the combination of the larger width of cleavage 716 and the smaller width of cleavage 726 can enable alignment of semiconductor device 712 and semiconductor device 722 by compensating for the larger difference in thermal expansion rates between substrates 710 and 720.
[0090] The larger width of dicing 716 and the smaller width of dicing 726 may cause semiconductor devices 712 and 722 to be misaligned along the x-axis (e.g., in a first direction) at the start of the bonding operation between bonding substrates 710 and 720. However, as substrates 710 and 720 are heated during the bonding operation (causing them to expand), the combination of the increased width of dicing 716, the decreased width of dicing 726, and the greater expansion rate of substrate 720 causes semiconductor devices 712 and 722 to move into near-alignment along the x-axis.
[0091] To further facilitate alignment between semiconductor devices 712 and 722 along the x-axis (e.g., in the first direction), the width of the cleavage 716 may gradually increase from near the center of substrate 710 toward the edge (or periphery) of substrate 710. This gradual increase in width compensates for the increased rate of expansion along the x-axis (e.g., in the first direction) from near the center of substrate 720 toward the edge of substrate 720, which can cause a greater displacement of semiconductor devices 722 near the edge of substrate 720 relative to semiconductor devices 722 near the center of substrate 720 during bonding. Therefore, the width of the cleavage 716 may increase toward the edge of substrate 710, causing the spacing between semiconductor devices 712 to increase toward the edge of substrate 710 to compensate for the greater displacement of semiconductor devices 722 near the edge of substrate 720 during bonding. In this way, substrate 710 may include multiple subsets of the cleavage 716, each subset having a different width relative to other subsets of the cleavage 716.
[0092] As an example, dicing 716 may include a first subset 716a formed by one or more dicings and a second subset 716b formed by one or more dicings. The first subset 716a may be located closer to the center of the substrate 710 relative to the second subset 716b. The width (w2) of the second subset 716b of dicing 716 may be larger than the width (w3) of the first subset 716a of dicing 716. In some embodiments, the width of dicing 716 gradually increases from near the center of the substrate 710 toward the edge of the substrate 710, from dicing to dicing, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 714 and 724 is approximately 40 micrometers, then the width (w3) of the first subset 716a may range from approximately 41 micrometers to approximately 44 micrometers, the width (w2) of the second subset 716b may range from approximately 42 micrometers to approximately 48 micrometers, and so on. In some embodiments, the dicing track 716 includes a larger subset of tracks whose width gradually increases from near the center of the substrate 710 to near the edge of the substrate 710.
[0093] Furthermore, the width of the dicing 726 may gradually decrease from near the center of the substrate 720 toward the edge (or periphery) of the substrate 720. This gradual decrease in width compensates for the increased rate of expansion along the x-axis (e.g., in a first direction) from near the center of the substrate 720 toward the edge, which can cause a greater displacement of the semiconductor devices 722 near the edge of the substrate 720 relative to the semiconductor devices 722 near the center of the substrate 720 during bonding. Therefore, the width of the dicing 726 may decrease toward the edge of the substrate 720, causing the spacing between the semiconductor devices 722 to decrease toward the edge of the substrate 720 to compensate for the greater displacement of the semiconductor devices 722 near the edge of the substrate 720 during bonding. In this way, the substrate 720 may include multiple subsets of the dicing 726, each subset having a different width relative to the other subsets of the dicing 726.
[0094] As an example, dicing 726 may include a first subset 726a formed by one or more dicings and a second subset 726b formed by one or more dicings. The first subset 726a may be located closer to the center of the substrate 720 relative to the second subset 726b. The width (w4) of the second subset 726b of dicing 726 may be smaller than the width (w5) of the first subset 726a of dicing 726. In some embodiments, the width decreases from near the center of the substrate 720 toward the edge of the substrate 720, and from dicing to dicing, the width of dicing 726 may gradually decrease in the range of approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 714 and 724 is approximately 40 micrometers, then the width (w5) of the first subset 726a may be in the range of approximately 36 micrometers to approximately 39 micrometers, the width (w4) of the second subset 726b may be in the range of approximately 32 micrometers to approximately 38 micrometers, and so on. In some embodiments, the dicing track 726 includes a larger subset of tracks whose width gradually increases from near the center of the substrate 720 to near the edge of the substrate 720.
[0095] In some embodiments, the gradual increase in the width of dicing 716 is quantitatively equal to the gradual decrease in the width of dicing 726. For example, the width of dicing 716 may increase gradually in increments of 3 micrometers, and the width of dicing 726 may decrease gradually in increments of 3 micrometers. In some embodiments, the gradual increase in the width of dicing 716 differs quantitatively from the gradual decrease in the width of dicing 726, which allows for finer adjustments to the alignment between semiconductor devices 712 and 722. For example, the width of dicing 716 may increase gradually in increments of 2 micrometers, and the width of dicing 726 may decrease gradually in increments of 4 micrometers.
[0096] As indicated above, Figure 7It is provided as an instance. Other instances may be related to... Figure 7 The content they describe is different.
[0097] Figure 8 This is a figure illustrating an exemplary embodiment 800 described herein. Exemplary embodiment 800 is similar to exemplary embodiment 700 in that it includes, for example, a first set of cleavages with gradually increasing width along a first direction on a low-expansion-rate substrate, and a second set of cleavages with gradually decreasing width along the first direction on a high-expansion-rate substrate. Additionally, in exemplary embodiment 800, a third set of cleavages with gradually increasing width along a second (e.g., approximately perpendicular) direction is included on the low-expansion-rate substrate. The first, second, and third sets of cleavages can compensate for asymmetric thermal expansion of the high-expansion-rate substrate and / or the low-expansion-rate substrate, because the combination of the first and second sets of cleavages can compensate for larger thermal expansion differences along the first axis, and the third set of cleavages can compensate for smaller thermal expansion differences along the second axis.
[0098] like Figure 8 As shown, an exemplary embodiment 800 may include substrates 810 and 820 to be bonded. Substrate 810 and substrate 820 may be bonded using one or more of the bonding techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another type of semiconductor substrate bonding technique). Substrate 810 may include a plurality of semiconductor devices 812, and substrate 820 may include a plurality of semiconductor devices 822. Semiconductor devices 812 and 822 may be formed using one or more semiconductor processing techniques described herein. By bonding substrates 810 and substrate 820, the corresponding pair of semiconductor devices 812 and 822 may be formed as three-dimensional semiconductor devices, including transistors, memory devices, MEMs, ASICs, SoCs, and / or other types of semiconductor devices.
[0099] Semiconductor devices 812 may be arranged in a grid on substrate 810, such that a plurality of cleavages 814 separate rows (or columns) of semiconductor devices 812, and a plurality of cleavages 816 separate columns (or rows) of semiconductor devices 812. The cleavages 814 may be located in substrate 810 along a first direction (e.g., along...). Figure 8 The x-axis shown is formed, and the cut path 816 may be formed in the substrate 810 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 8The y-axis (as shown) is formed such that cleavage 814 and cleavage 816 intersect at various locations across the substrate 810. Similarly, semiconductor devices 822 may be arranged in a grid on the substrate 820 such that a plurality of cleavage 824 separate rows (or columns) of semiconductor devices 822, and a plurality of cleavage 826 separate columns (or rows) of semiconductor devices 822. Cleavage 824 may be formed in the substrate 820 along a first direction (e.g., along...). Figure 8 The x-axis shown is formed, and the cut path 826 may be formed in the substrate 820 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 8 The y-axis shown is formed such that cut channels 824 and 826 intersect at various locations across the substrate 820.
[0100] Substrate 810 can be referred to as a low-expansion wafer, and substrate 820 can be referred to as a high-expansion wafer because substrate 820 can have a larger expansion rate relative to substrate 810. To compensate for the expansion rate of substrate 820 along... Figure 8 Due to the greater thermal expansion along the x-axis, cleavage 816 and cleavage 826 can have widths different from the width (w1) of cleavage 824. Specifically, since substrate 810 is a low thermal expansion substrate, the width of cleavage 816 can be larger than the width (w1) of cleavage 824. Furthermore, since substrate 820 is a high thermal expansion substrate, the width of cleavage 826 can be smaller than the width (w1) of cleavage 824 (and the width of cleavage 816). As described above, the combination of the larger width of cleavage 816 and the smaller width of cleavage 826 can enable alignment of semiconductor device 812 and semiconductor device 822 by compensating for the larger difference in thermal expansion rates between substrates 810 and 820 along the x-axis (e.g., in a direction approximately perpendicular to the lengths of cleavage 816 and cleavage 826).
[0101] To further facilitate alignment between semiconductor devices 812 and 822 along the x-axis (e.g., in the first direction), the width of the cleavage 816 may gradually increase from near the center of substrate 810 toward the edge (or periphery) of substrate 810. This gradual increase in width compensates for the increased rate of expansion along the x-axis (e.g., in the first direction) from near the center of substrate 820 toward the edge of substrate 820, which can cause a greater displacement of semiconductor devices 822 near the edge of substrate 820 relative to semiconductor devices 822 near the center of substrate 820 during bonding. Therefore, the width of the cleavage 816 may increase toward the edge of substrate 810, causing the spacing between semiconductor devices 812 to increase toward the edge of substrate 810 to compensate for the greater displacement of semiconductor devices 822 near the edge of substrate 820 during bonding. In this way, substrate 810 may include multiple subsets of cleavage 816, each subset having a different width relative to other subsets of cleavage 816.
[0102] As an example, dicing 816 may include a first subset 816a formed by one or more dicings and a second subset 816b formed by one or more dicings. The first subset 816a may be located closer to the center of the substrate 810 relative to the second subset 816b. The width (w2) of the second subset 816b of dicing 816 may be larger than the width (w3) of the first subset 816a of dicing 816. In some embodiments, the width of dicing 816 gradually increases from near the center of the substrate 810 toward near the edge of the substrate 810, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicings 814 and 824 is approximately 55 micrometers, then the width (w3) of the first subset 816a may range from approximately 56 micrometers to approximately 59 micrometers, the width (w2) of the second subset 816b may range from approximately 57 micrometers to approximately 63 micrometers, and so on. In some embodiments, the cleavage 816 comprises a larger subset of cleavages that gradually increase in width from near the center of substrate 810 to near the edge of substrate 810.
[0103] Furthermore, the width of the cleavage 826 may gradually decrease from near the center of the substrate 820 toward the edge (or periphery) of the substrate 820. This gradual decrease in width compensates for the increased rate of expansion along the x-axis (e.g., in a first direction) from near the center of the substrate 820 toward the edge of the substrate 820, which can cause a greater displacement of the semiconductor devices 822 near the edge of the substrate 820 relative to the semiconductor devices 822 near the center of the substrate 820 during bonding. Therefore, the width of the cleavage 826 may decrease toward the edge of the substrate 820, causing the spacing between the semiconductor devices 822 to decrease toward the edge of the substrate 820 to compensate for the greater displacement of the semiconductor devices 822 near the edge of the substrate 820 during bonding. In this way, the substrate 820 may include multiple subsets of the cleavage 826, each subset having a different width relative to the other subsets of the cleavage 826.
[0104] As an example, dicing 826 may include a first subset 826a formed by one or more dicings and a second subset 826b formed by one or more dicings. The first subset 826a may be located closer to the center of the substrate 820 relative to the second subset 826b. The width (w4) of the second subset 826b of dicing 826 may be smaller than the width (w5) of the first subset 826a of dicing 826. In some embodiments, the width of dicing 826 gradually decreases from near the center of the substrate 820 toward the edge of the substrate 820, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of dicing 824 is approximately 55 micrometers, then the width (w5) of the first subset 826a may be in the range of approximately 51 micrometers to approximately 54 micrometers, the width (w4) of the second subset 826b may be in the range of approximately 47 micrometers to approximately 53 micrometers, and so on. In some embodiments, the dicing track 826 comprises a larger subset of tracks whose width gradually decreases from near the center of the substrate 820 to near the edge of the substrate 820.
[0105] In some embodiments, the gradual increase in the width of dicing 816 is quantitatively equal to the gradual decrease in the width of dicing 826. In some embodiments, the gradual increase in the width of dicing 816 differs quantitatively from the gradual decrease in the width of dicing 826, which allows for finer adjustments to the alignment between semiconductor devices 812 and 822.
[0106] To further facilitate alignment between semiconductor devices 812 and 822 along the y-axis (e.g., in a second direction), the width of the cleavage 814 may gradually increase from near the center of substrate 810 toward the edge (or periphery) of substrate 810. This gradual increase in width compensates for the increased rate of expansion along the y-axis (e.g., in a second direction) from near the center of substrate 820 toward the edge of substrate 820, which can cause a greater displacement of semiconductor devices 822 near the edge of substrate 820 relative to semiconductor devices 822 near the center of substrate 820 during bonding. Therefore, the width of the cleavage 814 may increase toward the edge of substrate 810, causing the spacing between semiconductor devices 812 to increase toward the edge of substrate 810 to compensate for the greater displacement of semiconductor devices 822 near the edge of substrate 820 during bonding. In this way, substrate 810 may include multiple subsets of cleavage 814, each subset having a different width relative to other subsets of cleavage 814.
[0107] As an example, the dicing 814 may include a first subset 814a formed by one or more dicings and a second subset 814b formed by one or more dicings. The first subset 814a may be located closer to the center of the substrate 810 relative to the second subset 814b. The width (w6) of the second subset 814b of the dicing 814 may be larger than the width (w7) of the first subset 814a of the dicing 814. In some embodiments, the width of the dicing 814 gradually increases from near the center of the substrate 810 toward the edge of the substrate 810, ranging from approximately 1 micrometer to approximately 4 micrometers. As an example, if the width (w1) of the dicing 824 is approximately 55 micrometers, then the width (w7) of the first subset 814a may be in the range of approximately 56 micrometers to approximately 59 micrometers, the width (w6) of the second subset 814b may be in the range of approximately 57 micrometers to approximately 63 micrometers, and so on. In some embodiments, the dicing track 814 comprises a larger number of subsets with gradually increasing widths from near the center of the substrate 810 to near the edge of the substrate 810.
[0108] As indicated above, the widths of the cleavages 814 and 816 can gradually increase from near the center of substrate 810 toward the edge of substrate 810 to compensate for the radially symmetrical thermal expansion of substrate 820. Therefore, the increase in the widths of cleavages 814 and 816 can be approximately equal to compensate for the radially symmetrical thermal expansion of substrate 820. In some embodiments, the increase in the widths of cleavages 814 and 816 can be different to compensate for the asymmetric thermal expansion of substrate 810, for example, in cases where different types of devices are formed along the x-axis and y-axis of substrate 820 (which can result in asymmetric thermal expansion rates along the x-axis and y-axis of substrate 820).
[0109] As indicated above, Figure 8 It is provided as an instance. Other instances may be related to... Figure 8 The content they describe is different.
[0110] Figure 9 This is a figure illustrating an exemplary embodiment 900 described herein. Exemplary embodiment 900 is an example of including multiple sets of cleavage paths along multiple directions on each substrate to be bonded. This allows for compensation of large differences in thermal expansion along multiple axes, and / or allows for finer adjustments to the alignment of the semiconductor device.
[0111] like Figure 9 As shown, an exemplary embodiment 900 may include substrates 910 and 920 to be bonded. Substrate 910 and substrate 920 may be bonded using one or more of the bonding techniques described herein (e.g., eutectic bonding, fusion (or direct) bonding, hybrid bonding, or another type of semiconductor substrate bonding technique). Substrate 910 may include a plurality of semiconductor devices 912, and substrate 920 may include a plurality of semiconductor devices 922. Semiconductor devices 912 and 922 may be formed using one or more semiconductor processing techniques described herein. By bonding substrates 910 and substrate 920, the corresponding pair of semiconductor devices 912 and 922 may be formed as three-dimensional semiconductor devices, including transistors, memory devices, MEMs, ASICs, SoCs, and / or other types of semiconductor devices.
[0112] Semiconductor devices 912 may be arranged in a grid on substrate 910, such that a plurality of cleavages 914 separate rows (or columns) of semiconductor devices 912, and a plurality of cleavages 916 separate columns (or rows) of semiconductor devices 912. The cleavages 914 may be located in substrate 910 along a first direction (e.g., along...). Figure 9 The x-axis shown is formed, and the cut path 916 may be formed in the substrate 910 along a second (e.g., vertical) direction (e.g., along the x-axis). Figure 9 The y-axis (as shown) is formed such that cleavage 914 and cleavage 916 intersect at various locations across the substrate 910. Similarly, semiconductor devices 922 may be arranged in a grid on the substrate 920 such that a plurality of cleavage 924 separate rows (or columns) of semiconductor devices 922, and a plurality of cleavage 926 separate columns (or rows) of semiconductor devices 922. Cleavage 924 may be arranged in the substrate 920 along a first direction (e.g., along...). Figure 9 The x-axis shown is formed, and the cut path 926 may be formed in the substrate 920 along a second (e.g., vertical) direction (e.g., along...). Figure 9 The y-axis shown is formed such that cut channels 924 and 926 intersect at various locations across the substrate 920.
[0113] Substrate 910 can be referred to as a low-expansion wafer, and substrate 920 can be referred to as a high-expansion wafer because substrate 920 can have a larger expansion rate relative to substrate 910. To compensate for the expansion rate of substrate 920 along... Figure 9 To account for the greater thermal expansion along the x-axis, the cleavage 916 may have a width that gradually increases from near the center of the substrate 910 toward the edge of the substrate 910, and the cleavage 926 may have a width that gradually decreases from near the center of the substrate 920 toward the edge of the substrate 920. This is to compensate for the increased thermal expansion along the x-axis of the substrate 920. Figure 9 Due to the greater thermal expansion along the y-axis, the cut 914 may have a width that gradually increases from near the center of the substrate 910 toward the edge of the substrate 910, and the cut 924 may have a width that gradually decreases from near the center of the substrate 920 toward the edge of the substrate 920.
[0114] The dicing 916 may include a first subset 916a formed by one or more dicings and a second subset 916b formed by one or more dicings. The first subset 916a may be located closer to the center of the substrate 910 relative to the second subset 916b. The width (w1) of the second subset 916b of the dicing 916 may be larger than the width (w2) of the first subset 916a of the dicing 916. In some embodiments, the width of the dicing 916 may gradually increase from near the center of the substrate 910 toward the edge of the substrate 910, ranging from approximately 1 micrometer to approximately 4 micrometers.
[0115] The dicing 926 may include a first subset 926a formed by one or more dicings and a second subset 926b formed by one or more dicings. The first subset 926a may be located closer to the center of the substrate 920 relative to the second subset 926b. The width (w3) of the second subset 926b of the dicing 926 may be smaller than the width (w4) of the first subset 926a of the dicing 926. In some embodiments, the width of the dicing 926 gradually decreases from near the center of the substrate 920 toward the edge of the substrate 920, ranging from approximately 1 micrometer to approximately 4 micrometers.
[0116] The dicing 914 may include a first subset 914a formed by one or more dicings and a second subset 914b formed by one or more dicings. The first subset 914a may be located closer to the center of the substrate 910 relative to the second subset 914b. The width (w5) of the second subset 914b of the dicing 914 may be larger than the width (w6) of the first subset 914a of the dicing 914. In some embodiments, the width of the dicing 914 may gradually increase from near the center of the substrate 910 toward the edge of the substrate 910, ranging from approximately 1 micrometer to approximately 4 micrometers.
[0117] The dicing 924 may include a first subset 924a formed by one or more dicings and a second subset 924b formed by one or more dicings. The first subset 924a may be located closer to the center of the substrate 920 relative to the second subset 924b. The width (w7) of the second subset 924b of the dicing 924 may be smaller than the width (w8) of the first subset 924a of the dicing 924. In some embodiments, the width of the dicing 924 may gradually decrease from near the center of the substrate 920 toward the edge of the substrate 920, ranging from approximately 1 micrometer to approximately 4 micrometers.
[0118] In some embodiments, the gradual increase in the width of cleavage 916 is quantitatively equal to the gradual decrease in the width of cleavage 926. In some embodiments, the gradual increase in the width of cleavage 916 differs quantitatively from the gradual decrease in the width of cleavage 926, which allows for finer adjustments to the alignment between semiconductor devices 912 and 922.
[0119] In some embodiments, the gradual increase in the width of cleavage 914 is quantitatively equal to the gradual decrease in the width of cleavage 924. In some embodiments, the gradual increase in the width of cleavage 914 differs quantitatively from the gradual decrease in the width of cleavage 924, which allows for finer adjustments to the alignment between semiconductor devices 912 and 922.
[0120] As indicated above, Figure 9 It is provided as an instance. Other instances may be related to... Figure 9 The content they describe is different.
[0121] Figures 10A to 10D This is a figure illustrating an exemplary embodiment 1000 described herein. Exemplary embodiment 1000 may be an exemplary embodiment of bonding a low-expansion-rate substrate (e.g., one or more of substrates 210 to 910) with a high-expansion-rate substrate (e.g., one or more of substrates 220 to 920). Figure 10AAs shown, exemplary embodiment 1000 may include processing device 1002. In some embodiments, processing device 1002 includes a processor, computing device, and / or controller, etc. In some embodiments, processing device 1002 is included in bonding station 110. In some embodiments, processing device 1002 is separate from bonding station 110 and communicatively coupled to bonding station 110 (e.g., via direct connection or network). In some embodiments, processing device 1002 is included in another station (e.g., wafer analysis station, scanning electron microscopy (SEM) device, transmission electron microscopy (TEM) device, and / or another station).
[0122] like Figure 10A As further shown by reference numeral 1004, the bonding machine 110 can perform multiple bonding operations to bond a first plurality of semiconductor devices on a first plurality of substrates (e.g., a plurality of low thermal expansion substrates) to a second plurality of semiconductor devices on a second plurality of substrates (e.g., a plurality of high thermal expansion substrates). The bonding operations may include fusion bonding operations, eutectic bonding operations, and / or one or more other types of bonding operations.
[0123] The processing device 1002 may receive information associated with the plurality of bonding operations. Additionally and / or alternatively, the processing device 1002 may generate information associated with the plurality of bonding operations. This information may include SEM images of the first plurality of substrates and the second plurality of substrates after the first plurality of substrates and the second plurality of substrates have been bonded, TEM images of the first plurality of substrates and the second plurality of substrates after the first plurality of substrates and the second plurality of substrates have been bonded, and / or other types of information.
[0124] like Figure 10A As further illustrated by reference numeral 1006, the processing apparatus 1002 can generate thermal expansion data for a first plurality of substrates and thermal expansion data for a second plurality of substrates based on information associated with the plurality of bonding operations. The first plurality of substrates may include substrates of the same type, such that thermal expansion data can be generated in the plurality of bonding operations on the same type of substrates. Similarly, the second plurality of substrates may include substrates of the same type, such that thermal expansion data can be generated in the plurality of bonding operations on the same type of substrates.
[0125] Thermal expansion data for multiple substrates may include individual semiconductor expansion measurements for each of the multiple substrates. Semiconductor expansion measurements may include, for example, measurements of the total dimensional expansion of the substrate (e.g., measurements of the increase in the substrate's diameter or the increase in the substrate's dimension, for example, along a certain axis), measurements of the displacement of semiconductor devices on the substrate (or measurements of the displacement at different locations on the substrate), and / or one or more other types of expansion measurements. In these examples, processing device 1002 may perform semiconductor expansion measurements based on SEM and / or TEM images of the first and second multiple substrates after the first multiple substrates and the second multiple substrates have been bonded.
[0126] In some embodiments, semiconductor expansion measurement includes semiconductor device misalignment measurement across the substrate (e.g., from near the center of the substrate to near the edge of the substrate) in one or more directions or along one or more axes. Figure 10A An exemplary semiconductor misalignment map generated by processing device 1002 is shown. The exemplary semiconductor misalignment map can visually indicate the amount and / or direction of misalignment of various portions of a substrate joined by bonding station 110. The amount of misalignment can be indicated as a misalignment percentage or another parameter, and can be visually indicated in the misalignment map by different color coding, by highlighting, by numbering, by shading, and / or by another visual technique. As an example, Figure 10A The misalignment map shown includes a plurality of misalignment regions 1008a to 1008h, each of which includes a region of a different color in the misalignment map to visually indicate the amount and / or direction of misalignment in the plurality of misalignment regions 1008a to 1008h. In some embodiments, the misalignment map may visually indicate the amount and / or direction of misalignment for each individual pair of semiconductor devices that are bonded.
[0127] like Figure 10BAs shown in reference numeral 1010, processing apparatus 1002 can determine an adjusted width for a plurality of cleavages used to bond a third substrate and a fourth substrate. The third substrate may be of the same type as a first plurality of substrates, and the fourth substrate may be of the same type as a second plurality of substrates, such that the adjusted width can be used to reduce misalignment between semiconductor devices on the third and fourth substrates. The adjusted width may be relative to the width of the cleavages on the first plurality of substrates and / or the second plurality of substrates. For example, processing apparatus 1002 can determine an adjusted width of approximately 84 micrometers for a cleavage on the third substrate that corresponds to a cleavage on the first plurality of substrates with a width of approximately 80 micrometers. Processing apparatus 1002 can determine the adjusted width for one or more of the cleavages on the first plurality of substrates and / or the second plurality of substrates. Processing apparatus 1002 can determine the adjusted width based on thermal expansion data of the first plurality of substrates and thermal expansion data of the second plurality of substrates. The adjusted width can be overlaid on a misalignment map, such as... Figure 10B As shown in the example. As an example, the adjusted width can be visually indicated as multiple adjustment areas 1012a to 1012f on the misalignment diagram, where each adjustment area corresponds to the amount and direction of the width adjustment (e.g., an increase or decrease in width).
[0128] In some embodiments, the processing device 1002 determines the adjusted width by determining adjustment parameters based on semiconductor expansion measurements of each of the first plurality of substrates and semiconductor expansion measurements of each of the second plurality of substrates, and by determining the adjusted width for the plurality of dicing channels based on the adjustment parameters.
[0129] As an example, thermal expansion data of the first plurality of substrates can indicate substrate expansion of a micrometer for the first plurality of substrates, and thermal expansion data of the second plurality of substrates can indicate substrate expansion of b micrometers for the second plurality of substrates. The kerfs across the third substrate (e.g., a low-expansion substrate) can be numbered 1 to x from near the center of the third substrate toward the edge of the third substrate. The processing apparatus 1002 can determine the width of each kerf from near the center of the third substrate to near the edge of the third substrate as y+1z, y+2z, ..., y+xz, where y corresponds to the width of the unadjusted kerfs on the first plurality of substrates, and z corresponds to the adjustment parameter of the third substrate. The processing apparatus 1002 can determine the total increase in kerfs on the third substrate based on the following:
[0130] (1+x)xz / 2
[0131] The processing device 1002 can determine the adjustment parameter (z) such that:
[0132]
[0133] The processing apparatus 1002 can determine the width of each cut path from near the center of the fourth substrate (e.g., a high-expansion substrate) to near the edge of the fourth substrate as y–1z, y–2z, ... y–xz based on adjustment parameters.
[0134] In some implementations, when determining adjustment parameters and / or one or more adjusted widths, processing device 1002 uses machine learning and / or neural network processing to make decisions. Machine learning involves a computer learning from data on performing tasks and / or making decisions. Machine learning algorithms are used to train a machine learning model based on sample data (referred to as "training data"). Once trained, the machine learning model can be used to make predictions, decisions, or classifications related to new observations. Here, processing device 1002 may train the machine learning model based on inter-cycle thermal expansion data of a first plurality of substrates and a second plurality of substrates, adjusted widths from one or more historical bonding operations, and / or other types of data. Alternatively, a separate device (e.g., a server device) may train the machine learning model and provide the trained machine learning model for use by processing device 1002.
[0135] Supervised and / or unsupervised learning techniques can be used to train machine learning models based on data from thousands (or more) of concatenation operations. In some implementations, processing device 1002 trains regression models, decision tree models, Naive Bayes classifiers, etc., by specifying a target or outcome variable to be predicted from an independent set of variables. Bayes classifier, k-nearest neighbor (KNN) model, or another type of supervised machine learning model. In some implementations, processing device 1002 uses a neural network (e.g., an artificial neural network comprising thousands or millions (or more) artificial neurons or processing nodes) to train the machine learning model using deep learning techniques. Nodes in the neural network can be hierarchical and weighted. A node in one layer of the neural network can be connected to multiple nodes in the lower layers of the neural network and can receive weights from each of the nodes in the lower layers. If the weights meet a threshold (this may be referred to as the node's "firing"), the node can update the weights and feed the weights forward to one or more nodes in higher layers of the neural network. As the neural network continues to process the training data provided to the machine learning model, the weights in the neural network can be adjusted and / or updated until similar training data parameters result in similar weights through the neural network.
[0136] The machine learning model can generate a function that maps the inputs of the machine learning model (e.g., inter-cycle thermal expansion data of the first and second plurality of substrates, adjusted widths from one or more historical bonding operations, and / or other types of data) to a specified output from the machine learning model (e.g., alignment results of semiconductor devices included on the third and fourth substrates). In this way, the processing device 1002 uses said function (e.g., determined by the machine learning model) to determine adjustment parameters and / or one or more adjusted widths.
[0137] like Figure 10C As shown, processing apparatus 1002 can provide instructions for an adjusted width for forming a plurality of cleavages for joining a third substrate and a fourth substrate to one or more semiconductor processing stations (e.g., one or more of semiconductor processing stations 102 to 108). As indicated by reference numeral 1014, one or more semiconductor processing stations (e.g., one or more of semiconductor processing stations 102 to 108) can form the plurality of cleavages in the third substrate and / or the fourth substrate based on the adjusted width. In some embodiments, the one or more semiconductor processing stations form the plurality of cleavages as shown in one or more of the exemplary embodiments 200 to 900 above. For example, the one or more semiconductor processing stations form the plurality of cleavages in the third substrate and / or the fourth substrate along one or more directions such that the plurality of cleavages gradually increase or decrease to compensate for the thermal expansion of the third substrate and the fourth substrate.
[0138] like Figure 10D As shown by reference numeral 1016, the bonding machine 110 can bond a third substrate and a fourth substrate. The bonding machine 110 can bond the third substrate and the fourth substrate after forming the plurality of cleavages in the third substrate and / or the fourth substrate based on an adjusted width.
[0139] As indicated above, Figures 10A to 10D It is provided as an instance. Other instances may be related to... Figures 10A to 10D The content they describe is different.
[0140] Figure 11 This is a diagram of exemplary components of apparatus 1100. In some embodiments, semiconductor processing stations 102 to 110, wafer / die transport station 112, and / or processing apparatus 1002 may include (or may be included in) one or more apparatuses 1100 and / or one or more components of apparatus 1100. Figure 11 As shown, the device 1100 may include a bus 1110, a processor 1120, a memory 1130, a storage component 1140, an input component 1150, an output component 1160, and a communication component 1170.
[0141] Bus 1110 includes components enabling wired and / or wireless communication between components of device 1100. Processor 1120 includes a central processing unit (CPU), graphics processing unit (GPU), microprocessor, controller, microcontroller, digital signal processor, field-programmable gate array (FPGA), application-specific integrated circuit (ASIC), and / or another type of processing component. Processor 1120 is implemented in hardware, firmware, or a combination of hardware and software. In some embodiments, processor 1120 includes one or more processors that can be programmed to perform functions. Memory 1130 includes random access memory (RAM), read-only memory (ROM), and / or another type of memory (e.g., flash memory, magnetic memory, and / or optical memory).
[0142] Storage component 1140 stores information and / or software related to the operation of device 1100. For example, storage component 1140 may include a hard disk drive, disk drive, optical disk drive, solid-state drive, compact disc, digital versatile disc, and / or another type of non-transitory computer-readable medium. Input component 1150 enables device 1100 to receive input, such as user input and / or sensed input. For example, input component 1150 may include a touch screen, keyboard, keypad, mouse, button, microphone, switch, sensor, GPS component, accelerometer, gyroscope, and / or actuator. Output component 1160 enables device 1100 to provide output, for example, via a display, speaker, and / or one or more light-emitting diodes. Communication component 1170 enables device 1100 to communicate with other devices, for example, via a wired connection and / or wireless connection. For example, communication component 1170 may include a receiver, transmitter, transceiver, modem, network interface card, and / or antenna.
[0143] Device 1100 may perform one or more processes set forth herein. For example, a non-transitory computer-readable medium (e.g., memory 1130 and / or storage component 1140) may store a set of instructions (e.g., one or more instructions, code, software code, and / or program code) for execution by processor 1120. Processor 1120 may execute the set of instructions to perform one or more processes set forth herein. In some embodiments, execution of the set of instructions by one or more processors 1120 causes the one or more processors 1120 and / or device 1100 to perform one or more processes set forth herein. In some embodiments, hardwired circuitry may be used in place of or in combination with the instructions to perform one or more processes set forth herein. Therefore, the embodiments set forth herein are not limited to any particular combination of hardwired circuitry and software.
[0144] Figure 11 The number and arrangement of components shown are provided as examples. Figure 11 Compared to the components shown, device 1100 may include additional components, fewer components, different components, or components arranged in a different manner. Alternatively, one set of components of device 1100 (e.g., one or more components) may perform one or more functions described as being performed by another set of components of device 1100.
[0145] Figure 12 This is a flowchart of an exemplary process 1200 associated with wafer bonding alignment. In some embodiments, Figure 12 The one or more process blocks shown can be performed by one or more semiconductor processing stations (e.g., one or more semiconductor processing stations 102 to 110). In some embodiments, Figure 12 The one or more process blocks shown may be performed by another device or group of devices (e.g., processing device 1002) that is separate from or includes the one or more semiconductor processing units. Alternatively, Figure 12 One or more process blocks shown may be executed by one or more components of device 1100 (e.g., processor 1120, memory 1130, storage component 1140, input component 1150, output component 1160 and / or communication component 1170).
[0146] like Figure 12As shown, process 1200 may include a first plurality of semiconductor devices (block 1210) forming a first substrate. For example, the one or more semiconductor processing stations 102 to 110 may form a first plurality of semiconductor devices (e.g., semiconductor devices 212, 312, 612 and / or 712) on the first substrate (e.g., substrates 210, 310, 610 and / or 710), as described above.
[0147] like Figure 12 As further shown, process 1200 may include forming a first plurality of cleavages (block 1220) in a first substrate between a first plurality of semiconductor devices along a first direction. For example, the one or more semiconductor processing stations 102 to 110 may form the first plurality of cleavages (e.g., cleavages 214, 314, 614 and / or 714) in a first substrate between a first plurality of semiconductor devices along a first direction, as described above.
[0148] like Figure 12 As further shown, process 1200 may include forming a second plurality of cleavages in a first substrate along a second direction between a first plurality of semiconductor devices, wherein a first width of a first subset of the second plurality of cleavages is greater than a second width of a second subset of the second plurality of cleavages (block 1230). For example, the one or more semiconductor processing stations 102 to 110 may form a second plurality of cleavages (e.g., cleavages 216, 316, 616, 716) in a first substrate along a second direction between the first plurality of semiconductor devices, as described above. In some embodiments, the first width of a first subset of the second plurality of cleavages (e.g., subsets 216b, 316b, 616b, and / or 716b) is greater than the second width of a second subset of the second plurality of cleavages (e.g., subsets 216a, 316a, 616a, and / or 716a).
[0149] like Figure 12As further shown, process 1200 may include performing a bonding operation to bond a first plurality of semiconductor devices to a second plurality of semiconductor devices on a second substrate, wherein a first width is greater than a second width and the first width and the second width are greater than a third width of a third plurality of cleavages between the second plurality of semiconductor devices, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation (block 1240). For example, the one or more semiconductor processing stations 102 to 110 may perform a bonding operation to bond a first plurality of semiconductor devices to a second plurality of semiconductor devices (e.g., semiconductor devices 222, 322, 622 and / or 722) on a second substrate (e.g., substrates 220, 320, 620 and / or 720), as described above. In some embodiments, a first width is greater than a second width and a first width and the second width are greater than a third width of a third plurality of cleavages (e.g., cleavages 226, 326, 626 and / or 726), such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation.
[0150] Process 1200 may include additional embodiments, such as any single embodiment or any combination of embodiments set forth below and / or in conjunction with one or more other processes set forth elsewhere herein.
[0151] In a first embodiment, forming the second plurality of dicing channels includes forming a second subset of the second plurality of dicing channels closer to the center of the first substrate than a first subset of the second plurality of dicing channels. In a second embodiment, forming the first plurality of dicing channels, alone or in combination with the first embodiment, includes forming the first plurality of dicing channels with the same width as a third plurality of dicing channels. In a third embodiment, forming the first plurality of dicing channels, alone or in combination with one or more of the first and second embodiments, includes forming a first subset of the first plurality of dicing channels with a fourth width and forming a second subset of the first plurality of dicing channels with a fifth width, wherein the fifth width is larger than the fourth width.
[0152] In a fourth embodiment, forming the first plurality of cleavages, either alone or in combination with one or more of the first to third embodiments, includes forming a first subset and a second subset of the first plurality of cleavages, such that the fourth and fifth widths are larger than the sixth width of the fourth plurality of cleavages between the second plurality of semiconductor devices. In a fifth embodiment, either alone or in combination with one or more of the first to fourth embodiments, the third plurality of cleavages are formed in the second substrate along a second direction, and process 1200 includes forming the fourth plurality of cleavages (e.g., cleavage 624) in the second substrate along a first direction between the second plurality of semiconductor devices, wherein the fourth width of the first subset of the fourth plurality of cleavages (e.g., subset 624b) is smaller than the fifth width of the second subset of the fourth plurality of cleavages (e.g., subset 624a), and wherein the second subset of the fourth plurality of cleavages is closer to the center of the second substrate than the first subset of the fourth plurality of cleavages.
[0153] In a sixth embodiment, either alone or in combination with one or more of the first to fifth embodiments, a third plurality of cleavages are formed in the second substrate along a second direction, and process 1200 includes forming a fourth plurality of cleavages (e.g., cleavage 726) in the second substrate between the second plurality of semiconductor devices along a first direction, wherein a fourth width of a first subset (e.g., subset 726b) of the fourth plurality of cleavages is smaller than a fifth width of a second subset (e.g., subset 726a) of the fourth plurality of cleavages. In a seventh embodiment, either alone or in combination with one or more of the first to sixth embodiments, a second subset of the fourth plurality of cleavages is closer to the center of the second substrate than a first subset of the fourth plurality of cleavages.
[0154] although Figure 12 An exemplary block of process 1200 is shown, but in some embodiments, it is different from... Figure 12 Compared to the boxes shown in the diagram, process 1200 may include additional boxes, fewer boxes, different boxes, or boxes arranged in a different manner. Alternatively, two or more boxes in process 1200 may be executed in parallel.
[0155] Figure 13 This is a flowchart of an exemplary process 1300 associated with wafer bonding alignment. In some embodiments, Figure 13 The one or more process blocks shown can be performed by one or more semiconductor processing units (e.g., the one or more semiconductor processing units 102 to 110). In some embodiments, Figure 13The one or more process blocks shown may be performed by another device or group of devices (e.g., processing device 1002) that is separate from or includes the one or more semiconductor processing units. Alternatively, Figure 13 One or more process blocks shown may be executed by one or more components of device 1100 (e.g., processor 1120, memory 1130, storage component 1140, input component 1150, output component 1160 and / or communication component 1170).
[0156] like Figure 13 As shown, process 1300 may include a first plurality of semiconductor devices (block 1310) forming a first substrate. For example, the one or more semiconductor processing stations 102 to 110 may form a first plurality of semiconductor devices (e.g., semiconductor devices 422, 522, 822 and / or 922) on the first substrate (e.g., substrates 420, 520, 820, 920), as described above.
[0157] like Figure 13 As further shown, process 1300 may include forming a first plurality of cleavages (block 1320) between a first plurality of semiconductor devices along a first axis of the first substrate. For example, the one or more semiconductor processing stations 102 to 110 may form the first plurality of cleavages (e.g., cleavages 424, 524, 824 and / or 924) between the first plurality of semiconductor devices along a first axis of the first substrate, as described above.
[0158] like Figure 13 As further shown, process 1300 may include forming a second plurality of cleavages between a first plurality of semiconductor devices along a second axis of a first substrate, wherein a first width of a first subset of the second plurality of cleavages is smaller than a second width of a second subset of the second plurality of cleavages (block 1330). For example, the one or more semiconductor processing stations 102 to 110 may form a second plurality of cleavages (e.g., cleavages 426, 526, 826, and / or 926) between the first plurality of semiconductor devices along a second axis of a first substrate, as described above. In some embodiments, the first width of the first subset of the second plurality of cleavages (e.g., subsets 426b, 526b, 826b, and / or 926b) is smaller than the second width of the second subset of the second plurality of cleavages (e.g., subsets 426a, 526a, 826a, and / or 926a).
[0159] like Figure 13As further shown, process 1300 may include performing a bonding operation to bond a first plurality of semiconductor devices to a second plurality of semiconductor devices on a second substrate, wherein a first width is smaller than a second width and the first width and the second width are smaller than a third width of a third plurality of cleavages between the second plurality of semiconductor devices, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation (block 1340). For example, the one or more semiconductor processing stations 102 to 110 may perform a bonding operation to bond a first plurality of semiconductor devices to a second plurality of semiconductor devices (e.g., semiconductor devices 412, 512, 812, and 912) on a second substrate (e.g., substrates 410, 510, 810, and / or 910), as described above. In some embodiments, a first width is smaller than a second width and a first width and the second width are smaller than a third width of a third plurality of cleavages (e.g., cleavages 416, 516, 816, and / or 916), such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation.
[0160] Process 1300 may include additional embodiments, such as any single embodiment or any combination of embodiments set forth below and / or in conjunction with one or more other processes set forth elsewhere herein.
[0161] In a first embodiment, forming the first plurality of cleavages includes forming a first subset (e.g., subset 524a) of the first plurality of cleavages with a fourth width and forming a second subset (e.g., subset 524b) of the first plurality of cleavages with a fifth width, wherein the fifth width is smaller than the fourth width. In a second embodiment, either alone or in combination with the first embodiment, forming the first plurality of cleavages includes forming the first subset of the first plurality of cleavages closer to the center of the first substrate than the second subset of the first plurality of cleavages. In a third embodiment, either alone or in combination with one or more of the first and second embodiments, the first width is smaller than the second width, and both the first and second widths are smaller than the third width, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned along a second axis, and wherein the fifth width is smaller than the fourth width, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned along a first axis.
[0162] In a fourth embodiment, either alone or in combination with one or more of the first to third embodiments, a third plurality of cleavages are formed along a second axis in the second substrate. A fourth width of a first subset of the third plurality of cleavages (e.g., subsets 816b and / or 916b) is larger than a fifth width of a second subset of the third plurality of cleavages (e.g., subsets 816a and / or 916a). Process 1300 includes forming a fourth plurality of cleavages (e.g., cleavages 814 and / or 914) between the second plurality of semiconductor devices along a first axis in the second substrate, wherein a sixth width of a first subset of the fourth plurality of cleavages (e.g., subsets 814b and / or 914b) is larger than a seventh width of a second subset of the fourth plurality of cleavages (e.g., subsets 814a and / or 914a). In a fifth embodiment, either alone or in combination with one or more of the first to fourth embodiments, the seventh width of the third plurality of cleavages increases from the center of the second substrate to the edge of the second substrate.
[0163] although Figure 13 An exemplary block of process 1300 is shown, but in some embodiments, with Figure 13 Compared to the boxes shown in the diagram, process 1300 may include additional boxes, fewer boxes, different boxes, or boxes arranged in a different manner. Alternatively, two or more boxes in process 1300 may be executed in parallel.
[0164] Figure 14 This is a flowchart of an exemplary process 1400 associated with wafer bonding alignment. In some embodiments, Figure 14 The one or more process blocks shown can be performed by one or more semiconductor processing units (e.g., the one or more semiconductor processing units 102 to 110). In some embodiments, Figure 14 The one or more process blocks shown may be performed by another device or group of devices (e.g., processing device 1002) that is separate from or includes the one or more semiconductor processing units. Alternatively, Figure 14 One or more process blocks shown may be executed by one or more components of device 1100 (e.g., processor 1120, memory 1130, storage component 1140, input component 1150, output component 1160 and / or communication component 1170).
[0165] like Figure 14As shown, process 1400 may include receiving information associated with multiple bonding operations for bonding a first plurality of semiconductor devices on a first plurality of substrates and a second plurality of semiconductor devices on a second plurality of substrates (block 1410). For example, processing apparatus 1002 may receive information associated with multiple bonding operations for bonding a first plurality of semiconductor devices (e.g., one or more of semiconductor devices 212 to 912) on a first plurality of substrates (e.g., one or more of substrates 210 to 910) and a second plurality of semiconductor devices (e.g., one or more of semiconductor devices 222 to 922) on a second plurality of substrates (e.g., one or more of substrates 220 to 920), as described above. Bonding station 110 may perform the multiple bonding operations.
[0166] like Figure 14 As further shown, process 1400 may include generating thermal expansion data for a first plurality of substrates and thermal expansion data for a second plurality of substrates based on information associated with the plurality of bonding operations (block 1420). For example, processing apparatus 1002 may generate thermal expansion data for the first plurality of substrates and thermal expansion data for the second plurality of substrates based on information associated with the plurality of bonding operations, as described above.
[0167] like Figure 14 As further shown, process 1400 may include determining an adjusted width for a plurality of cleavages of at least one of a third or fourth substrate based on thermal expansion data of a first plurality of substrates and thermal expansion data of a second plurality of substrates (block 1430). For example, processing apparatus 1002 may determine an adjusted width for a plurality of cleavages (e.g., one or more of cleavages 214 to 914, cleavages 216 to 916, cleavages 224 to 924 and / or cleavages 226 to 926) of at least one of a third or fourth substrate based on thermal expansion data of the first plurality of substrates and thermal expansion data of the second plurality of substrates, as described above.
[0168] like Figure 14 As further shown, process 1400 may include providing an indication of the adjusted width to one or more semiconductor processing units to form the plurality of cleavages based on the adjusted width (block 1440). For example, processing unit 1002 may provide an indication of the adjusted width to one or more semiconductor processing units to form the plurality of cleavages based on the adjusted width, as described above. The one or more semiconductor processing units 102 to 110 may form the plurality of cleavages based on the adjusted width.
[0169] Process 1400 may include additional embodiments, such as any single embodiment or any combination of embodiments set forth below and / or in conjunction with one or more other processes set forth elsewhere herein.
[0170] In a first embodiment, process 1400 includes, after forming the plurality of cleavages, bonding (e.g., using bonding machine 110) a third plurality of semiconductor devices on a third substrate to a fourth plurality of semiconductor devices on a fourth substrate, wherein the plurality of cleavages are located between at least one of the third plurality of semiconductor devices or the fourth plurality of semiconductor devices. In a second embodiment, individually or in combination with the first embodiment, thermal expansion data for the first plurality of substrates includes semiconductor expansion measurements for each of the first plurality of substrates, and thermal expansion data for the second plurality of substrates includes semiconductor expansion measurements for each of the second plurality of substrates. In a third embodiment, individually or in combination with one or more of the first and second embodiments, determining an adjusted width for the plurality of cleavages includes determining adjustment parameters based on semiconductor expansion measurements for each of the first plurality of substrates and semiconductor expansion measurements for each of the second plurality of substrates, and determining the adjusted width for the plurality of cleavages based on the adjustment parameters.
[0171] In a fourth embodiment, forming the plurality of cleavage paths, either alone or in combination with one or more of the first to third embodiments, includes forming a first subset of the plurality of cleavage paths in a third substrate along a first axis, and forming a second subset of the plurality of cleavage paths in the third substrate along a second axis approximately perpendicular to the first axis. In a fifth embodiment, forming the plurality of cleavage paths, either alone or in combination with one or more of the first to fourth embodiments, includes forming a first subset of the plurality of cleavage paths in a third substrate, and forming a second subset of the plurality of cleavage paths in a fourth substrate.
[0172] although Figure 14 An exemplary block of process 1400 is shown, but in some embodiments, it is different from... Figure 14 Compared to the boxes shown in the diagram, process 1400 may include additional boxes, fewer boxes, different boxes, or boxes arranged in a different manner. Alternatively, two or more boxes in process 1400 may be executed in parallel.
[0173] In this way, the joining alignment of devices formed on substrates to be joined can be achieved by using cleavage between the devices, wherein the size of the cleavage gradually increases or decreases from the center to the edge of one or more of the substrates to compensate for differences in the thermal expansion rates of the substrates. Due to the gradually increasing or decreasing size of the cleavage, the devices on the substrates are aligned as the substrates are heated during the joining operation. The cleavage can be arranged in a single direction in the substrate to compensate for thermal expansion of the substrate along a single axis, or it can be arranged in multiple directions to compensate for radiation-symmetric thermal expansion. Thus, the techniques and apparatus described herein can reduce device pattern misalignment, improve the bonding strength between joined substrates, improve the structural integrity and electrical connectivity of the devices, and / or reduce device failures on the substrates.
[0174] As described in more detail above, some embodiments described herein provide a method. The method includes forming a first plurality of semiconductor devices on a first substrate. The method includes forming a first plurality of cleavages between the first plurality of semiconductor devices in the first substrate along a first direction. The method includes forming a second plurality of cleavages between the first plurality of semiconductor devices in the first substrate along a second direction, wherein a first width of a first subset of the second plurality of cleavages is greater than a second width of a second subset of the second plurality of cleavages. The method includes performing a bonding operation to bond the first plurality of semiconductor devices to a second plurality of semiconductor devices on a second substrate, wherein the first width is greater than the second width, and the first width and the second width are greater than a third width of a third plurality of cleavages between the second plurality of semiconductor devices, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation.
[0175] In some embodiments, forming the second plurality of cleavages includes: forming a second subset of the second plurality of cleavages closer to the center of the first substrate than a first subset of the second plurality of cleavages. In some embodiments, forming the first plurality of cleavages includes: forming the first plurality of cleavages to have the same width as the third width of the third plurality of cleavages. In some embodiments, forming the first plurality of cleavages includes: forming a first subset of the first plurality of cleavages to have a fourth width; and forming a second subset of the first plurality of cleavages to have a fifth width, wherein the fifth width is greater than the fourth width. In some embodiments, forming the first plurality of cleavages includes: forming the first subset of the first plurality of cleavages and the second subset of the first plurality of cleavages such that the fourth width and the fifth width are greater than the sixth width of the fourth plurality of cleavages between the second plurality of semiconductor devices. In some embodiments, the third plurality of cleavages are along the second direction in the second substrate; and the method further includes: forming a fourth plurality of cleavages between the second plurality of semiconductor devices in the second substrate along the first direction, wherein a fourth width of a first subset of the fourth plurality of cleavages is smaller than a fifth width of a second subset of the fourth plurality of cleavages, and wherein the second subset of the fourth plurality of cleavages is closer to the center of the second substrate than the first subset of the fourth plurality of cleavages. In some embodiments, the third plurality of cleavages are along the second direction in the second substrate; and the method further includes: forming a fourth plurality of cleavages between the second plurality of semiconductor devices in the second substrate along the first direction, wherein a fourth width of a first subset of the fourth plurality of cleavages is smaller than a fifth width of a second subset of the fourth plurality of cleavages. In some embodiments, the second subset of the fourth plurality of cleavages is closer to the center of the second substrate than the first subset of the fourth plurality of cleavages.
[0176] As described in more detail above, some embodiments described herein provide a method. The method includes forming a first plurality of semiconductor devices on a first substrate. The method includes forming a first plurality of cleavages between the first plurality of semiconductor devices along a first axis of the first substrate. The method includes forming a second plurality of cleavages between the first plurality of semiconductor devices along a second axis of the first substrate, wherein a first width of a first subset of the second plurality of cleavages is smaller than a second width of a second subset of the second plurality of cleavages. The method includes performing a bonding operation to bond the first plurality of semiconductor devices to a second plurality of semiconductor devices on a second substrate, wherein the first width is smaller than the second width, and the first width and the second width are smaller than a third width of a third plurality of cleavages between the second plurality of semiconductor devices such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned during the bonding operation.
[0177] In some embodiments, forming the first plurality of cleavages includes: forming a first subset of the first plurality of cleavages with a fourth width; and forming a second subset of the first plurality of cleavages with a fifth width, wherein the fifth width is smaller than the fourth width. In some embodiments, forming the first plurality of cleavages includes: forming the first subset of the first plurality of cleavages closer to the center of the first substrate relative to the second subset of the first plurality of cleavages. In some embodiments, the first width is smaller than the second width and the first width and the second width are smaller than the third width such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned along the second axis; and wherein the fifth width is smaller than the fourth width such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned along the first axis. In some embodiments, the third plurality of cleavages are along the second axis in the second substrate; wherein the fourth width of the first subset of the third plurality of cleavages is greater than the fifth width of the second subset of the third plurality of cleavages; and wherein the method further includes: forming a fourth plurality of cleavages between the second plurality of semiconductor devices along the first axis in the second substrate, wherein a sixth width of the first subset of the fourth plurality of cleavages is greater than a seventh width of the second subset of the fourth plurality of cleavages. In some embodiments, the seventh width of the third plurality of cleavages increases from the center of the second substrate to the edge of the second substrate.
[0178] As described in more detail above, some implementation schemes described herein provide a method. The method includes...
[0179] A processing device receives information associated with multiple bonding operations of bonding first multiple semiconductor devices on first multiple substrates and second multiple semiconductor devices on second multiple substrates. The method includes the processing device generating thermal expansion data for the first multiple substrates and thermal expansion data for the second multiple substrates based on the information associated with the multiple bonding operations. The method includes the processing device determining adjusted widths of multiple cleavages on at least one of a third or fourth substrate based on the thermal expansion data of the first multiple substrates and the thermal expansion data of the second multiple substrates. The method includes the processing device providing an indication of the adjusted widths to one or more semiconductor processing stations to form the multiple cleavages based on the adjusted widths.
[0180] In some embodiments, the method further includes: after forming the plurality of cleavages, joining a third plurality of semiconductor devices on the third substrate to a fourth plurality of semiconductor devices on the fourth substrate, wherein the plurality of cleavages are located between at least one of the third plurality of semiconductor devices or the fourth plurality of semiconductor devices. In some embodiments, the thermal expansion data of the first plurality of substrates includes a semiconductor expansion measurement value for each of the first plurality of substrates; and wherein the thermal expansion data of the second plurality of substrates includes a semiconductor expansion measurement value for each of the second plurality of substrates. In some embodiments, determining the adjusted width of the plurality of cleavages of at least one of the third substrate or the fourth substrate includes: determining an adjustment parameter based on the semiconductor expansion measurement values for each of the first plurality of substrates and the semiconductor expansion measurement values for each of the second plurality of substrates; and determining the adjusted width of the plurality of cleavages based on the adjustment parameter. In some embodiments, the method further includes: forming a first subset of the plurality of cleavages along a first axis in the third substrate; and forming a second subset of the plurality of cleavages along a second axis approximately perpendicular to the first axis in the third substrate. In some embodiments, the method further includes: forming a first subset of the plurality of cleavages in the third substrate; and forming a second subset of the plurality of cleavages in the fourth substrate.
[0181] The foregoing has outlined features of several embodiments to enable those skilled in the art to better understand various aspects of this disclosure. Those skilled in the art will understand that they can readily use this disclosure as a basis for designing or modifying other processes and structures to achieve the same purposes and / or realize the same advantages as the embodiments described herein. Those skilled in the art should also recognize that such equivalent constructions do not depart from the spirit and scope of this disclosure, and that various changes, substitutions, and modifications can be made herein without departing from the spirit and scope of this disclosure.
[0182] [Explanation of Symbols]
[0183] 100: Environment
[0184] 102: Semiconductor processing station / deposition station
[0185] 104: Semiconductor processing equipment / exposure equipment
[0186] 106: Semiconductor processing machine / developing machine
[0187] 108: Semiconductor Processing Equipment / Etching Equipment
[0188] 110: Semiconductor processing station / bonding station
[0189] 112: Wafer / Die Transport Machine
[0190] 200, 300, 400, 500, 600, 700, 800, 900, 1000: Implementation Plan
[0191] 210, 220, 310, 320, 410, 420, 510, 520, 610, 620, 710, 720, 810, 820, 910, 920: Substrate
[0192] 212, 222, 312, 322, 412, 422, 512, 522, 612, 622, 712, 722, 812, 822, 912, 922: Semiconductor devices
[0193] 214, 216, 224, 226, 314, 316, 324, 326, 414, 416, 424, 426, 514, 516, 524, 526, 614, 616, 624, 626, 714, 716, 724, 726, 814, 816, 824, 826, 914, 916, 924, 926: Cutting Track
[0194] 216a, 316a, 426a, 524a, 526a, 616a, 624a, 716a, 726a, 814a, 816a, 826a, 914a, 916a, 924a, 926a: First subset / subset of the cutting path
[0195] 216b, 316b, 426b, 524b, 526b, 616b, 624b, 716b, 726b, 814b, 816b, 826b, 914b, 916b, 924b, 926b: Second subset / subset of the cutting path
[0196] 314a: The first subset of the cutting path
[0197] 314b: The second subset of the cutting path
[0198] 1002: Processing device
[0199] 1004, 1006, 1010, 1014, 1016: Operations
[0200] 1008a, 1008b, 1008c, 1008d, 1008e, 1008f, 1008g, 1008h: Misaligned areas
[0201] 1012a, 1012b, 1012c, 1012d, 1012e, 1012f: Adjustment area
[0202] 1100: Device
[0203] 1110: Bus
[0204] 1120: Processor
[0205] 1130: Memory
[0206] 1140: Storage Component
[0207] 1150: Input component
[0208] 1160: Output Component
[0209] 1170: Communication Components
[0210] 1200, 1300, 1400: Process
[0211] 1210, 1220, 1230, 1240, 1310, 1320, 1330, 1340, 1410, 1420, 1430, 1440: Steps
[0212] w1, w2, w3, w4, w5, w6, w7, w8: Width
[0213] x, y: axes
Claims
1. A method for wafer bonding and alignment, characterized in that, include: A first plurality of semiconductor devices forming a first substrate; A first plurality of cleavage channels are formed in the first substrate between the first plurality of semiconductor devices along a first direction; A second plurality of cleavage channels are formed in the first substrate along a second direction between the first plurality of semiconductor devices. Wherein the first width of the first subset of the second plurality of cutting channels is greater than the second width of the second subset of the second plurality of cutting channels; as well as Perform a bonding operation to bond the first plurality of semiconductor devices to the second plurality of semiconductor devices on the second substrate. The first width is greater than the second width, and the first width and the second width are greater than the third width of the third plurality of cleavages between the second plurality of semiconductor devices, such that the first plurality of semiconductor devices are approximately aligned with the second plurality of semiconductor devices during the bonding operation.
2. The wafer bonding alignment method according to claim 1, wherein forming the second plurality of dicing paths comprises: The second subset of the second plurality of dicing channels is formed closer to the center of the first substrate than the first subset of the second plurality of dicing channels.
3. The wafer bonding alignment method according to claim 1, wherein forming the first plurality of dicing channels comprises: The first plurality of cutting channels are formed to have the same width as the third plurality of cutting channels.
4. The wafer bonding alignment method according to claim 1, wherein forming the first plurality of dicing channels comprises: The first subset of the first plurality of cutting channels is formed as the fourth width; as well as The second subset of the first plurality of cutting channels is formed into the fifth width. The fifth width is greater than the fourth width.
5. The wafer bonding alignment method of claim 4, wherein the fourth width and the fifth width are greater than the sixth width of the fourth plurality of dices between the second plurality of semiconductor devices.
6. The wafer bonding alignment method of claim 1, wherein the third plurality of dicing paths are along the second direction in the second substrate; and The method further includes: A fourth plurality of cleavage channels are formed in the second substrate along the first direction between the second plurality of semiconductor devices. The fourth width of the first subset of the fourth plurality of cutting channels is smaller than the fifth width of the second subset of the fourth plurality of cutting channels, and The second subset of the fourth plurality of dicings is closer to the center of the second substrate than the first subset of the fourth plurality of dicings.
7. The wafer bonding alignment method of claim 1, wherein the third plurality of dicing paths are along the second direction in the second substrate; and wherein the method further comprises: A fourth plurality of cleavages are formed in the second substrate along the first direction between the second plurality of semiconductor devices, wherein a fourth width of a first subset of the fourth plurality of cleavages is smaller than a fifth width of a second subset of the fourth plurality of cleavages.
8. The wafer bonding alignment method of claim 7, wherein the second subset of the fourth plurality of dicing channels is closer to the center of the second substrate than the first subset of the fourth plurality of dicing channels.
9. A method for wafer bonding and alignment, characterized in that, include: A first plurality of semiconductor devices forming a first substrate; A first plurality of cleavages are formed between the first plurality of semiconductor devices along a first axis of the first substrate; A second plurality of cleavage channels are formed between the first plurality of semiconductor devices along a second axis of the first substrate. Wherein the first width of the first subset of the second plurality of cutting channels is less than the second width of the second subset of the second plurality of cutting channels; as well as Perform a bonding operation to bond the first plurality of semiconductor devices to the second plurality of semiconductor devices on the second substrate. Wherein the first width is less than the second width and the first width and the second width are less than the third width of the third plurality of cleavages between the second plurality of semiconductor devices, such that the first plurality of semiconductor devices are approximately aligned with the second plurality of semiconductor devices during the bonding operation.
10. The wafer bonding alignment method of claim 9, wherein forming the first plurality of dicing paths comprises: The first subset of the first plurality of cutting channels is formed as the fourth width; as well as The second subset of the first plurality of cutting channels is formed into the fifth width. The fifth width is smaller than the fourth width.
11. The wafer bonding alignment method of claim 10, wherein forming the first plurality of dicing channels comprises: The first subset of the first plurality of dicings is formed closer to the center of the first substrate than the second subset of the first plurality of dicings.
12. The wafer bonding alignment method of claim 10, wherein the first width is smaller than the second width and the first width and the second width are smaller than the third width, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned along the second axis; and The fifth width is smaller than the fourth width, such that the first plurality of semiconductor devices and the second plurality of semiconductor devices are approximately aligned along the first axis.
13. The wafer bonding alignment method of claim 9, wherein the third plurality of dicing paths are along the second axis in the second substrate; The fourth width of the first subset of the third plurality of cutting channels is greater than the fifth width of the second subset of the third plurality of cutting channels; and The method further includes: A fourth plurality of cleavage channels are formed in the second substrate along the first axis between the second plurality of semiconductor devices. The sixth width of the first subset of the fourth plurality of cutting channels is greater than the seventh width of the second subset of the fourth plurality of cutting channels.
14. The wafer bonding alignment method of claim 13, wherein the seventh width of the third plurality of dicing channels increases from the center of the second substrate to the edge of the second substrate.
Citation Information
Patent Citations
Low warpage wafer bonding through use of slotted substrates
CN103907175A
Semiconductor device
JP1986265842A
