Wafer bonding control method and wafer bonding control system

Abstract

The invention discloses a wafer bonding control method and a wafer bonding control system, and relates to the technical field of semiconductors, and the method comprises the steps: obtaining first detection data of a first wafer included in a first batch, and obtaining second detection data of a second wafer included in at least one second batch; performing wafer matching analysis on the first detection data and the second detection data to obtain a wafer bonding matching relationship of the first wafer, the wafer bonding matching relationship describing a second wafer matched with the first wafer in the at least one second batch; and controlling wafer bonding equipment to carry out wafer taking and bonding according to the wafer bonding matching relationship. According to the invention, the wafer bonding yield can be effectively improved.

Description

Technical Field

[0001] This application relates to the field of semiconductor technology, specifically to a wafer bonding control method and a wafer bonding control system. Background Technology

[0002] In 3D packaging technology, wafer-to-wafer bonding is a core process, which involves stacking (i.e., bonding) multiple layers of wafers, such as two, four, or even eight. Currently, the conventional wafer bonding method usually involves bonding wafers from two batches (LOTs) in a fixed wafer order. Due to the random or systematic defects present on each wafer, bonding in a fixed wafer order can easily result in more defects after bonding the upper and lower wafer layers, leading to poor wafer bonding yield. Summary of the Invention

[0003] This application provides a wafer bonding control scheme that can effectively improve wafer bonding yield.

[0004] The embodiments of this application provide the following technical solutions: According to one embodiment of this application, a wafer bonding control method includes: acquiring first detection data of a first wafer included in a first batch, and acquiring second detection data of a second wafer included in at least one second batch; performing wafer matching analysis on the first detection data and the second detection data to obtain a wafer bonding matching relationship of the first wafer, wherein the wafer bonding matching relationship describes a second wafer in the at least one second batch that matches the first wafer; and controlling a wafer bonding device to perform wafer picking and bonding according to the wafer bonding matching relationship.

[0005] In some embodiments of this application, the step of performing wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer includes: performing wafer matching analysis on the first detection data and the second detection data through a yield management system to obtain the wafer bonding matching relationship of the first wafer, and transmitting the wafer bonding matching relationship to a manufacturing execution system; the step of controlling the wafer bonding equipment to perform wafer picking and bonding according to the wafer bonding matching relationship includes: controlling the wafer bonding equipment to perform wafer picking and bonding according to the wafer bonding matching relationship through the manufacturing execution system and the equipment automation system.

[0006] In some embodiments of this application, the step of performing wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer includes: performing intra-batch wafer matching analysis on the first detection data of the first wafer in the first batch and the second detection data of the second wafer in each of the second batches to obtain the wafer bonding matching relationship of the first wafer.

[0007] In some embodiments of this application, the step of performing wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer includes: performing cross-batch wafer matching analysis on the first detection data of each first wafer in the first batch and the second detection data of each second wafer in the at least one second batch to obtain the wafer bonding matching relationship of the first wafer.

[0008] In some embodiments of this application, the step of performing intra-batch wafer matching analysis on the first detection data of the first wafer in the first batch and the second detection data of the second wafer in each of the second batches to obtain the wafer bonding matching relationship of the first wafer includes: H1 （B #i） With H2 （Tk #j） Stacking, obtaining stacking detection data H3 for the ikj-th stacked wafer within the k-th stacking batch. （B #i）+（Tk #j） H1 （B #i） H2 refers to the first detection data of the i-th first wafer B#i within the first batch. （Tk #j） This refers to the second detection data of the j-th second wafer Tk #j within the k-th second batch, where k belongs to 1 to n, and n is the total number of wafers in the second batch; according to H3 （B #i）+（Tk #j） Determine the number of qualified stacked dies in the j-th stacked wafer, where the j-th stacked wafer is a stacked wafer obtained by stacking B #i and Tk #j; determine the k-th optimal stacking batch based on the number of dies; determine the target optimal stacking batch from the n optimal stacking batches, and obtain the wafer bonding matching relationship of the i-th first wafer B #i based on the target optimal stacking batch.

[0009] In some embodiments of this application, the step of performing cross-batch wafer matching analysis on the first detection data of each first wafer in the first batch and the second detection data of each second wafer in at least one second batch to obtain the wafer bonding matching relationship of the first wafer includes: H1 （B #i） With H2 （Tk #j） Stacking, obtaining stacking detection data H3 for the ikj-th stacked wafer within the k-th stacking batch. （B #i）+（Tk #j） H1 （B #i）H2 refers to the first detection data of the i-th first wafer B#i within the first batch. （Tk #j） This refers to the second detection data of the j-th second wafer Tk #j within the k-th second batch, where k belongs to 1 to n, and n is the total number of wafers in the second batch; according to H3 （B #i）+（Tk #j） Determine the number of qualified stacked dies in the j-th stacked wafer, where the j-th stacked wafer is a stacked wafer obtained by stacking B#i and Tk#j; determine the optimal stacked wafer corresponding to the i-th first wafer B#i based on the number of dies; obtain the wafer bonding matching relationship of the i-th first wafer B#i based on the optimal stacked wafer corresponding to the i-th first wafer B#i.

[0010] In some embodiments of this application, controlling the wafer bonding equipment to perform wafer pick-up and bonding according to the wafer bonding matching relationship includes: dividing the second wafers matched by the first wafer from the second batch of sources into sub-batches according to the wafer bonding matching relationship of the first wafer; generating a new bonding matching relationship for the first wafer, the new bonding matching relationship describing a second wafer matched with the first wafer in the sub-batch; and controlling the wafer bonding equipment to perform wafer pick-up and bonding according to the new bonding matching relationship.

[0011] According to one embodiment of this application, a wafer bonding control system includes a matching module and a control module. The matching module is configured to: acquire first detection data of a first wafer included in a first batch, and acquire second detection data of a second wafer included in at least one second batch; and perform wafer matching analysis on the first detection data and the second detection data to obtain a wafer bonding matching relationship of the first wafer, wherein the wafer bonding matching relationship describes a second wafer that matches the first wafer in the at least one second batch. The control module is configured to: control a wafer bonding device to perform wafer pick-up and bonding according to the wafer bonding matching relationship.

[0012] In some embodiments of this application, the wafer bonding control system includes a yield management system, a manufacturing execution system, and an equipment automation system; the yield management system includes the matching module; the control module belongs to the manufacturing execution system and the equipment automation system.

[0013] In some embodiments of this application, the control module integrates a unit and a control unit; the manufacturing execution system includes the integration unit, and the equipment automation system includes the control unit; the integration unit is configured to: divide the second wafers matched by the first wafer into sub-batches from the second batch of sources according to the wafer bonding matching relationship of the first wafer; and generate a new bonding matching relationship for the first wafer, the new bonding matching relationship describing a second wafer matched with the first wafer in the sub-batch; the control unit is configured to: control the wafer bonding equipment to perform wafer pick-up bonding according to the new bonding matching relationship.

[0014] In this embodiment, first detection data of a first wafer included in the first batch and second detection data of a second wafer included in at least one second batch are obtained; wafer matching analysis is performed on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer, wherein the wafer bonding matching relationship describes a second wafer that matches the first wafer in the at least one second batch; and wafer bonding equipment is controlled to perform wafer picking and bonding according to the wafer bonding matching relationship.

[0015] In this embodiment of the application, the "first detection data of the first wafer included in the first batch" and the "second detection data of the second wafer included in at least one second batch" are fused and analyzed to obtain the wafer bonding matching relationship corresponding to each first wafer in the first batch. For each first wafer in the first batch, a matching second wafer can be found flexibly and reliably from at least one second batch. Wafer picking and bonding are performed according to the wafer bonding matching relationship, which can effectively avoid the drawbacks of bonding according to a fixed wafer order and effectively improve the wafer bonding yield. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A schematic diagram of a wafer bonding method in the prior art is shown.

[0018] Figure 2 A flowchart of a wafer bonding control method according to an embodiment of this application is shown.

[0019] Figure 3 A wafer defect diagram according to an embodiment of this application is shown.

[0020] Figure 4 A grain probe test pattern according to an embodiment of this application is shown.

[0021] Figure 5 A schematic diagram of the stacking of wafer inspection data according to an embodiment of this application is shown.

[0022] Figure 6 A schematic diagram of stacked wafer inspection data according to another embodiment of this application is shown.

[0023] Figure 7 A schematic diagram of stacked wafer inspection data according to another embodiment of this application is shown.

[0024] Figure 8 A block diagram of a wafer bonding control system according to an embodiment of this application is shown.

[0025] Figure 9 A block diagram of a wafer bonding control system according to an embodiment of this application is shown. Detailed Implementation

[0026] The present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the embodiments provided herein are merely illustrative of the present disclosure and are not intended to limit the present disclosure. Furthermore, the embodiments provided below are some embodiments for implementing the present disclosure, and not all embodiments for implementing the present disclosure. Unless otherwise specified, the technical solutions described in the embodiments of the present disclosure can be implemented in any combination. It should be noted that, in the embodiments of this disclosure, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a method or apparatus that includes a list of elements includes not only the elements expressly described, but also other elements not expressly listed, or elements inherent to implementing the method or apparatus. Without further limitations, an element defined by the phrase "comprising a..." does not exclude the presence of other related elements (e.g., steps in the method or units in the apparatus, such as portions of circuitry, processors, programs, or software, etc.) in the method or apparatus that includes that element. For example, the air filtration control method provided in this disclosure includes a series of steps, but the air filtration control method provided in this disclosure is not limited to the steps described. Similarly, the air filtration control system provided in this disclosure includes a series of units, but the device provided in this disclosure is not limited to the units explicitly described, but may also include units that need to be set up for obtaining relevant information or processing based on the information. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of this disclosure.

[0027] In 3D packaging technology, wafer-to-wafer bonding is the core process, which involves stacking (i.e., bonding) multiple layers of wafers, such as two, four, or even eight layers. Currently, the conventional wafer bonding method usually involves bonding wafers from two batches (LOTs) in a fixed wafer order.

[0028] For example, see Figure 1 Taking the bonding of wafers from batches A and B, LOT A includes wafers #01 to #04, and LOT B includes wafers #01 to #04. In the conventional method of bonding wafers in a fixed order, the stacking relationship (sequential) 110 shown in the figure is "#04", "#03", "#02", "#01", where "#04" means bonding wafer #04 from LOT A with wafer #04 from LOT B, and so on. The stacking relationship (reverse) 120 shown in the reverse order is "#04", "#03", "#02", "#01", where "#04" means bonding wafer #04 from LOT A with wafer #01 from LOT B, and so on.

[0029] Due to random or systematic defects present on each wafer, bonding in a fixed wafer order can easily result in a higher defect rate when bonding two layers of wafers (e.g., blindly bonding #04 in LOT A to #04 in LOT B, #03 in LOT A to #03 in LOT B, #02 in LOT A to #02 in LOT B, and #01 in LOT A to #01 in LOT B in a fixed order, ignoring the quality of the two wafers being bonded, the resulting stacked wafers are prone to more quality problems), leading to a poor overall wafer bonding yield.

[0030] In response, this application provides a wafer bonding control scheme that can effectively improve wafer bonding yield.

[0031] The following is a detailed description of the relevant embodiments of the wafer bonding control scheme provided in this application.

[0032] Figure 2 A flowchart illustrating a wafer bonding control method according to an embodiment of this application is shown schematically.

[0033] like Figure 2As shown, the wafer bonding control method may include steps S210 to S230.

[0034] Step S210: Obtain first detection data of the first wafer included in the first batch, and obtain second detection data of at least one second wafer included in the second batch; Step S220: Perform wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer. The wafer bonding matching relationship describes a second wafer that matches the first wafer in at least one second batch. Step S230: Control the wafer bonding equipment to perform wafer picking and bonding according to the wafer bonding matching relationship.

[0035] A wafer production line may include a first batch and at least one second batch. The at least one second batch can be one or more second batches, and the specific number of second batches is not specifically limited in this application. The first batch includes at least one first wafer, and each first wafer can obtain first inspection data (such as defect data and electrical data of the first wafer) detected on the production line. Each second batch includes at least one second wafer, and each second wafer can obtain second inspection data (such as defect data and electrical data of the second wafer) detected on the production line.

[0036] Matching analysis is performed on "first detection data of the first wafer included in the first batch" and "second detection data of the second wafer included in at least one second batch" to obtain a wafer bonding matching scheme. The wafer bonding matching scheme includes the wafer bonding matching relationship of each first wafer in the first batch. The wafer bonding matching relationship of the first wafer describes a second wafer that matches the first wafer in at least one second batch. Each first wafer in the first batch can obtain a corresponding wafer bonding matching relationship.

[0037] For example, the first wafers included in batch B are: B #01, B #02, B #03, B #04, and B #05; the second wafers included in a second batch T01 are: T01 #01, T01 #02, T01 #03, T01 #04, and T01 #05. The second wafers included in another second batch T02 are: T02 #01, T02 #02, T02 #03, T02 #04, and T02 #05. If B #05 matches T01 #01, the corresponding wafer bonding match can be represented as "B #05 & T01 #01"; if B #04 matches T02 #02, the corresponding wafer bonding match can be represented as "B #04 & T02 #02"; and so on, with each first wafer in batch B having its corresponding wafer bonding match.

[0038] Therefore, for each first wafer in the first batch, a matching second wafer can be flexibly and reliably found from at least one second batch. The matching second wafer for each first wafer may no longer be located in a fixed position within the second batch, or may not be located in a fixed second batch. Each first wafer will have better quality after being bonded to its matching second wafer (i.e., a first wafer and a second wafer with a wafer bonding match are bonded).

[0039] Furthermore, by controlling the wafer bonding equipment to perform wafer picking and bonding based on the wafer bonding matching relationship, the wafer bonding equipment can pick up and bond (i.e., stack) the first and second wafers that have a wafer bonding matching relationship. For example, if a wafer bonding matching relationship is "B #04&T02 #02", the wafer bonding equipment can instruct a robotic arm, etc., to pick up the first wafer B #04 from the carrier where the first batch B is located, and to pick up the second wafer T02 #02 from the carrier where the second batch T02 is located, and then bond the two wafers with a wafer bonding matching relationship, the first wafer B #04 and the second wafer T02 #02.

[0040] In summary, by fusing and analyzing the "first detection data of the first wafer included in the first batch" and the "second detection data of the second wafer included in at least one second batch" in this embodiment of the application, the wafer bonding matching relationship corresponding to each first wafer in the first batch can be obtained. For each first wafer in the first batch, a matching second wafer can be found flexibly and reliably from at least one second batch. Wafer picking and bonding can be performed according to the wafer bonding matching relationship, which can effectively avoid the drawbacks of bonding according to a fixed wafer order and effectively improve the wafer bonding yield.

[0041] The following describes further optional embodiments of the steps performed during wafer bonding control according to the embodiments of this application.

[0042] In one embodiment, step S220 involves performing wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer. Specifically, this may include: performing wafer matching analysis on the first detection data and the second detection data through a yield management system to obtain the wafer bonding matching relationship of the first wafer, and transmitting the wafer bonding matching relationship to the manufacturing execution system. Further, step S230 involves controlling the wafer bonding equipment to perform wafer picking and bonding according to the wafer bonding matching relationship. Specifically, this may include: controlling the wafer bonding equipment to perform wafer picking and bonding according to the wafer bonding matching relationship through the manufacturing execution system and the equipment automation system.

[0043] Computer Integrated Manufacturing (CIM) systems consist of dozens of software systems, including Manufacturing Execution System (MES), Equipment Automation System (EAP), Yield Management System (YMS), Statistical Process Control System (SPC), and Recipe Management System (RMS). The Yield Management System (YMS) is a system integrating data management, data analysis, and professional tools, focusing on analyzing and optimizing manufacturing yield. This system can analyze the root causes affecting yield by integrating all data generated during the production process (such as equipment parameters, process flows, and performance test data (such as first and second inspection data)). The Manufacturing Execution System (MES) can collect, analyze, and monitor production data in real time through data exchange with various equipment and systems on the production line, helping production managers make timely decisions, optimize production processes, and improve production efficiency and product quality. The Equipment Automation Program (EAP) is responsible for real-time monitoring and automated control of equipment on the production line (such as wafer bonding equipment). Through communication with the equipment and data acquisition, the EAP monitors the operating status and performance of the equipment in real time, and automatically adjusts the working mode and parameter settings of the equipment (such as wafer bonding equipment) according to preset process parameters and production plans.

[0044] In this embodiment, leveraging the highly integrated MES+YMS+EAP within the CIM system, the integrated interaction between MES, YMS, and EAP enables low-cost and reliable automation of wafer bonding control and personalized matching of wafers requiring bonding (i.e., the first and second wafers with existing wafer bonding matching relationships). Specifically, YMS is no longer limited to data and analysis; beyond its core yield management function, its wafer matching analysis capabilities are expanded. With the help of YMS, personalized wafer matching analysis can be performed at low cost and with high reliability, and YMS can participate in the decision-making process for wafer bonding schemes.

[0045] Specifically, the YMS can collect first inspection data of the first wafer included in the first batch, and second inspection data of at least one second wafer included in the second batch. Then, it performs wafer matching analysis on the first and second inspection data to obtain the wafer bonding matching relationship of the first wafer, and transmits the wafer bonding matching relationship to the MES in the form of a list, etc. Furthermore, the MES and EAP control the wafer bonding equipment to perform wafer picking and bonding according to the wafer bonding matching relationship.

[0046] Furthermore, in one embodiment, the basic configuration in the MES may include "wafer bonding process site configuration," "wafer bonding equipment configuration," and "wafer bonding bill of materials." The wafer bonding process site configuration can describe "which process sites are wafer bonding process sites," the wafer bonding equipment configuration can describe "which devices are wafer bonding devices used for wafer bonding," and the wafer bonding bill of materials may include a first identifier (Base ProductID) for the base batch, a second identifier (Top Product ID) for the candidate batch, a process site identifier, and a matching strategy.

[0047] If the MES determines, based on the basic configuration, that the first batch B and at least one second batch T will enter the wafer bonding process station, the MES can send a bonding matching scheme acquisition request to the YMS. The bonding matching scheme acquisition request can carry a matching strategy. The YMS can respond to the bonding matching scheme acquisition request and perform wafer matching analysis on the first detection data and the second detection data according to the matching strategy to obtain the wafer bonding matching relationship of the first wafer.

[0048] Here, the first batch B can be a batch (LOT) in the base batch, and at least one second batch T can be at least one batch (LOT) in the candidate batch. The MES can determine the first identifier of the base batch from which the first batch B originates, determine the second identifier of the candidate batch from which at least one second batch T originates, and determine the process site identifier of the first batch B and at least one second batch T entering the wafer bonding process site. Then, it determines the matching strategy corresponding to the first identifier, the second identifier, and the process site identifier from the wafer bonding bill of materials.

[0049] Furthermore, in one embodiment, the YMS can collect wafer inspection data, and the basic configuration in the YMS can include the definition of the stacking result. The YMS can determine whether the stacked die obtained by stacking two dies is a qualified stacked die based on the "wafer inspection data" and the "definition of the stacking result".

[0050] Wafer inspection data can include at least two of the following: a defect map, a die probe map (CP map), and a wafer acceptance test map (WAT map). The defect map indicates the location and type of defects within the wafer (such as particle contamination, scratches, etc.). The CP map indicates whether the function or electrical properties of each die within the wafer are good. The WAT map indicates whether the electrical properties of the predetermined test structure (or predetermined test area) containing each die within the wafer are good.

[0051] In one approach, the definition of the first superposition result may include: Defect + CP Bin + WAT Result = OK Die / NG Die. That is, the superposition result of the three detection data—Defect, CP Bin, and WAT Result—is used to determine whether the stacked die obtained by stacking two dies is a qualified stacked die (OK Die). Defect can indicate whether there are defects on the die, CP Bin can indicate whether the function or electrical properties of the die are good, and WAT Result can indicate whether the electrical properties of the predetermined test structure (or predetermined test area) where the die is located are good.

[0052] In another approach, the definition of the second stacking result can include: P + Q = OK Die / NG Die, where P can be any of the Defect, CP Bin, and WAT Result, and Q can also be any of the Defect, CP Bin, and WAT Result. P and Q can be the same or different. That is, the stacked die obtained by stacking the two dies is determined by combining the stacking results from the Defect, CP Bin, and WAT Result.

[0053] YMS can determine whether a stacked die formed by stacking two dies is a qualified stacked die based on "wafer inspection data" and "definition of stacking results". Specifically, this can include: only stacked dies formed by stacking two qualified dies are considered qualified stacked dies (OK Dies). The CP Bin of the die die is... Diei =2, or WAT Result Diei =2, or Defect Diei =When a particle or die is located at the edge of a wafer, the die can be considered a defective grain. CPBin Diei =2 indicates that Diei's function or electrical properties are not good, CP Bin Diei =1 indicates that Diei's function or electrical properties are good; WATResult Diei =2 indicates that the electrical properties of the predetermined test structure (or predetermined test area) where Diei is located are not good. WATResult Diei =1 indicates that the electrical properties of the predetermined test structure (or predetermined test area) where Diei is located are good; Defect Diei =Particle indicates a defect in Diei. Diei =0 indicates that there is no defect on Diei.

[0054] For example, taking the definition based on the first stacking result as an example, when die Die1 in the first wafer and die Die2 in the second wafer are stacked, the stacked die formed is a qualified stacked die (OK Die) only if both Die1 and Die2 are qualified. That is, a "Defective" die is a qualified stacked die (OK Die) if and only if both Die1 and Die2 are qualified. Die1 =0、Defect Die2 =0、CP Bin Die1 =1、CP Bin Die2 =1、WAT Result Die1 =1、WAT Result Die2 =1, and neither Die1 nor Die2 is located at the edge of the wafer. The stacked die formed by Die1 and Die2 is a qualified stacked die (OK Die). Otherwise, the stacked die formed by Die1 and Die2 is an unqualified stacked die (NGDie).

[0055] For example, taking the definition based on the second superposition result as an example, when P=Defect and Q=CP Bin, when Die1 and Die2 are superimposed, the stacked die formed only when both Die1 and Die2 are qualified is it a qualified stacked die (OK Die). Specifically, it can be if and only if "Defect" is not qualified. Die1 =0、CP Bin Die2 =1, and neither Die1 nor Die2 is located at the edge of the wafer. The stacked die formed by Die1 and Die2 is an OK die. Otherwise, the stacked die formed by Die1 and Die2 is an NG die. In this example, the inspection data of the first wafer where Die1 is located may only include the wafer defect map, and the inspection data of the second wafer where Die2 is located may only include the die probe map.

[0056] Furthermore, in one embodiment, performing wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer may include: performing intra-batch wafer matching analysis on the first detection data of the first wafer in the first batch and the second detection data of the second wafer in each second batch to obtain the wafer bonding matching relationship of the first wafer.

[0057] If the matching strategy is "batch yield priority (LOT yield priority)," YMS can perform intra-batch wafer matching analysis on the first inspection data of the first wafer in the first batch and the second inspection data of the second wafer in each second batch to obtain the wafer bonding matching relationship of the first wafer. For example, the first inspection data of the first wafer in the first batch is matched with the second inspection data of the second wafer in the first second batch; the first inspection data of the first wafer in the first batch is matched with the second inspection data of the second wafer in the second second batch; finally, the wafer bonding matching relationship of each first wafer in the first batch is obtained by combining the analysis results of intra-batch wafer matching analysis of multiple batches.

[0058] In other words, the first batch is matched with each of the second batches, with the highest overall batch yield as the benchmark. After matching, the second wafers matched with each first wafer in the first batch all come from the same second batch. The total number of second wafers in the second batch is greater than or equal to the total number of first wafers in the first batch. This "batch yield priority (LOT yield priority)" strategy considers both wafer stacking yield and processing efficiency (since the second wafers matched with each first wafer in the first batch all come from the same second batch, they are loaded into the same carrier (FOUP). During wafer pick-up and bonding, only one loading of this carrier is required, resulting in high processing efficiency).

[0059] Specifically, in one embodiment, the first detection data of the first wafer in the first batch is compared with the second detection data of the second wafer in each second batch to perform intra-batch wafer matching analysis to obtain the wafer bonding matching relationship of the first wafer, which may include: H1 （B #i） With H2 （Tk #j） Stacking, obtaining stacking detection data H3 for the ikj-th stacked wafer within the k-th stacking batch. （B #i）+（Tk #j） H1 （B #i） H2 refers to the first detection data of the i-th first wafer B#i within the first batch. （Tk #j） This refers to the second detection data of the j-th second wafer Tk #j within the k-th second batch, where k belongs to 1 to n, and n is the total number of wafers in the second batch; according to H3 （B #i）+（Tk #j） Determine the number of qualified stacked dies in the ikj-th stacked wafer, where the ikj-th stacked wafer is obtained by stacking B #i and Tk #j; determine the k-th optimal stacking batch based on the number of dies; determine the target optimal stacking batch from the n optimal stacking batches, and obtain the wafer bonding matching relationship of the i-th first wafer B #i based on the target optimal stacking batch.

[0060] The first inspection data for the first wafer may include at least two of the following: a first wafer defect map, a first die probe map (CP map), and a first wafer acceptance test map (WAT map); the second inspection data for the second wafer may include at least two of the following: a second wafer defect map, a second die probe map (CP map), and a second wafer acceptance test map (WAT map). The wafer defect map indicates the location and type of defects within the wafer (such as particle contamination, scratches, etc.). The die probe map (CP map) indicates whether the function or electrical properties of each die within the wafer are good. The wafer acceptance test map (WAT map) indicates the electrical data of the predetermined test structure (or predetermined test area) where each die within the wafer is located.

[0061] For example, in one embodiment, the first detection data may include a first wafer defect map and a first grain probe map (CP map), and the second detection data may include a second wafer defect map and a second grain probe map (CP map). See also... Figure 3 The first batch, B, includes five first wafers: B #01, B #02, B #03, B #04, and B #05. The corresponding first wafer defect maps for these five first wafers are 311, 312, 313, 314, and 315, respectively. The first second batch, T01, includes five second wafers: T01 #01, T01 #02, T01 #03, T01 #04, and T01 #05. The corresponding second wafer defect maps for these five second wafers are 321, 322, 323, 324, and 325, respectively.

[0062] Furthermore, in one example, see [link to example]. Figure 4 The first grain probe test patterns for the five first wafers B #01, B #02, B #03, B #04, and B #05 are 411, 412, 413, 414, and 415, respectively. The second grain probe test patterns for the five second wafers T01 #01, T01 #02, T01 #03, T01 #04, and T01 #05 are 421, 422, 423, 424, and 425, respectively. If a second batch T02 contains T02 #01, T02 #02, T02 #03, T02 #04, and T02 #05, each second wafer in the second batch T02 can also obtain a corresponding second grain probe test pattern.

[0063] In the first wafer defect map, the small black triangles mark the locations of defects within the wafer (e.g., when a small black triangle is marked on a die, it indicates a defect). Die =Particle, when there is no small black triangle marked on the die, it's a defect. Diei =0). In the second grain probe test diagram, the grain (Die) is marked with 1 (CP Bin). Die =1) means that the die has good function or electrical properties, and the die is marked as 2 (CP Bin). Die =2) This means that the die is not functionally or electrically good. In addition, NG is marked on the die at the edge of the wafer (where NG means that the die at the edge of the wafer is a bad die by default).

[0064] When the matching strategy is "batch yield priority (LOT yield priority)", the first detection data H1 of the first wafer B #i in the first batch is first selected. （B #i） "The second detection data H2 of the j-th second wafer Tk #j within the k-th second batch" （Tk #j） "Stacking, obtaining the stacking detection data H3 of the ikj-th stacked wafer within the k-th stacking batch." （B #i）+（Tk #j） Stacking the first wafer B #i and the second wafer Tk #j will result in the ikjth stacked wafer.

[0065] Specifically, the first detection data H1 of the first wafer B#i can be obtained first. （B #i） The first wafer defect map and the first grain probe test map (CP map) are stacked to obtain the first stacked inspection data of the first wafer B#i. For example, see [reference]. Figure 5 When i=01, stack 311 and 411 of B #01 to obtain the first stacked detection data 511, stack 312 and 412 of B #02 to obtain the first stacked detection data 512, and so on. Also, the second detection data H2 of the second wafer Tk #j can be... （Tk #j） The second wafer defect map and the second grain probe test map are stacked to obtain the second stack inspection data of the second wafer Tk #j. For example, see [reference]. Figure 5 When k=01, stacking 321 and 421 of T01 #01 yields the second stacked detection data 521, stacking 322 and 422 of T01 #02 yields the second stacked detection data 522, and so on. Further, stacking the first stacked detection data of the first wafer B #i and the second stacked detection data of the second wafer Tk #j yields the stacked detection data H3 of the ikj-th stacked wafer. （B #i）+（Tk #j）For example, see Figure 6 When i=01 and k=01, the first stacking detection data 511 of B #01 and the second stacking detection data 523 of T01 #03 are stacked to obtain the stacking detection data (H3) of the 01.01.03rd stacked wafer. （B #01）+（T01 #03） )613, stack the first stack detection data 515 of B #05 and the second stack detection data 521 of T01 #01 to obtain the stack detection data (H3) of the 05.01.01st stacked wafer. （B #05）+（T01 #01） )611 and so on.

[0066] Following the above superposition method, the first detection data H1 of the first wafer B #i within the first batch can be performed under different combinations of i, k, and j. （B #i） "The second detection data H2 of the j-th second wafer Tk #j within the k-th second batch" （Tk #j） "Stacking, obtaining the stacking detection data H3 of the ikj-th stacked wafer within the k-th stacking batch." （B #i）+（Tk #j） .

[0067] Furthermore, according to H3 of the ikj-th stacked wafer （B #i）+（Tk #j） The number of qualified stacked dies in the ikj-th stacked wafer can be determined. The ikj-th stacked wafer is obtained by stacking B #i and Tk #j. Dies at the same position in B #i and Tk #j are stacked vertically to form the stacked dies in the stacked wafer. Only stacked dies with two qualified dies are considered qualified stacked dies. In the stacking inspection data of the stacked wafer, the position of a qualified stacked die is marked "OK," and the position of a defective stacked die is marked "NG." If either of the two dies has a defect (as shown by the black triangle), the stacked dies form a defective stacked die. Two dies located at the edge of the wafer also form a defective stacked die. If either of the two dies is marked with a "2," the stacked dies form a defective stacked die. Figure 6 As shown, based on the stacking detection data (H3) of the 05.01.01st stacked wafer... （B #05）+（T01 #01） )611, we can count the number of qualified stacked dies marked OK to be 14, and so on, to get the number of qualified stacked dies in all stacked wafers of the ikjth digit.

[0068] Furthermore, the k-th optimal stacking batch is determined based on the number of wafers. For example, when the first wafer in the first batch B is stacked with the second wafer in the 01st second batch T01, there are multiple permutations and combinations of the first and second wafers. Stacking according to each permutation and combination forms a stacking batch. That is, there are multiple candidate stacking batches when the first wafer in the first batch B is stacked with the second wafer in the 01st second batch T01. For example, a candidate stacking batch D1 may include "B #01&T01 #01, B #02&T01 #02, B #03&T01 #03, B #04&T01 #04, B #05&T01 #05"; see reference. Figure 7 When i=01 and k=01, the first stacking detection data 511 of B #01 and the second stacking detection data 521 of T01 #01 are stacked to obtain the stacking detection data (H3) of the 01.01.01th stacked wafer. （B #01）+（T01 #01） )711, stack the first stack detection data 512 of B #02 and the second stack detection data 522 of T01 #02 to obtain the stack detection data (H3) of the 02.01.02nd stacked wafer. （B #02）+（T01 #02） )712 and so on. Another candidate stacking batch D2 can include "B #05&T01 #01, B #04&T01 #02, B #01&T01 #03, B #02&T01 #04, B #03&T01 #05 stacked wafers", and so on, to obtain different candidate stacking batches.

[0069] At this point, the 01st optimal stacking batch is "the candidate stacking batch with the highest total number of qualified stacked dies among all candidate stacking batches when the first wafer in the first batch B and the second wafer in the 01st second batch T01 are stacked". For example, see [link to relevant documentation]. Figure 6 In candidate stacking batch D2, the number of qualified stacked dies (OK Die count) in the stacked wafer obtained after bonding (stacking) B #05 & T01 #01 is 14; the number of OK Dies after bonding B #04 & T01 #02 is 10, after bonding B #01 & T01 #03 is 11, after bonding B #02 & T01 #04 is 14, and after bonding B #03 & T01 #05 is 12. Therefore, the total number of qualified stacked dies in candidate stacking batch D2 is 61. (See also...) Figure 7In candidate stacking batch D1, the number of qualified stacked dies (OK Die number) in the stacked wafer obtained after bonding (stacking) B #01 & T01 #01 is 10; the number of OK Dies after bonding B #02 & T01 #02 is 9, the number of OK Dies after bonding B #03 & T01 #03 is 9, the number of OK Dies after bonding B #04 & T01 #04 is 10, and the number of OK Dies after bonding B #05 & T01 #05 is 11. The total number of qualified stacked dies in candidate stacking batch D1 is 49. If the total number of qualified stacked dies in all candidate stacking batches is at most 61, then the candidate stacking batch is the 01st optimal stacking batch. Figure 7 The stacking method shown is the result of the conventional method of bonding according to a fixed wafer order. It can be seen that the total number of qualified stacked dies after stacking according to the fixed wafer order is usually less, that is, the yield of stacked wafers is lower.

[0070] Similarly, when the first wafer in the first batch B is stacked with the second wafer in the second batch T02, the second optimal stacking batch can be determined; when the first wafer in the first batch B is stacked with the second wafer in the kth second batch Tk, the kth optimal stacking batch can be determined; when the first wafer in the first batch B is stacked with the second wafer in the nth second batch Tn, the nth optimal stacking batch can be determined.

[0071] Furthermore, one of the n optimal stacking batches can be selected as the target optimal stacking batch. The target optimal stacking batch is the "optimal stacking batch with the highest total number of qualified stacked dies". For example, if the total number of qualified stacked dies in the 01st optimal stacking batch is 61, and the total number of qualified stacked dies in the 02nd to nth optimal stacking batches is less than 61, then the 01st optimal stacking batch is the target optimal stacking batch.

[0072] Furthermore, based on the target optimal stacking batch, the wafer bonding matching relationship of the i-th first wafer B#i within the first batch B can be obtained. For example, if the target optimal stacking batch is candidate stacking batch D2, which includes "B #05&T01 #01, B #04&T01 #02, B #01&T01 #03, B #02&T01 #04, B #03&T01 #05 stacked wafers", then the wafer bonding matching relationship of the first first wafer B #01 is "B #01&T01 #03", the wafer bonding matching relationship of the second first wafer B #02 is "B #02&T01 #04", the wafer bonding matching relationship of the third first wafer B #03 is "B #03&T01 #05", the wafer bonding matching relationship of the fourth first wafer B #04 is "B #04&T01 #02", and the wafer bonding matching relationship of the fifth first wafer B #05 is "B #05&T01 #01". It can be seen that the second wafers matched by the i-th first wafer B#i in the first batch B all come from the same second batch T01.

[0073] As mentioned above, the strategy of "lot yield priority" takes into account both wafer stacking yield and processing efficiency. Since the second wafers matched by each first wafer in the first batch all come from the same second batch, the second wafers matched by each first wafer B #i in the first batch are loaded in the same carrier (FOUP). When performing wafer pick-up bonding, only one loading is needed on the carrier to obtain the second wafers matched by each first wafer B #i in the first batch from the carrier, resulting in high processing efficiency.

[0074] Furthermore, in one embodiment, performing wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer may include: performing cross-batch wafer matching analysis on the first detection data of each first wafer in the first batch and the second detection data of each second wafer in at least one second batch to obtain the wafer bonding matching relationship of the first wafer.

[0075] If the matching strategy is "single wafer yield priority (Wafer yield priority)", YMS can perform cross-batch wafer matching analysis on the first detection data of each first wafer in the first batch and the second detection data of each second wafer in at least one second batch to obtain the wafer bonding matching relationship of the first wafer.

[0076] In other words, the first wafer is individually matched with each of the second wafers, with the highest single-wafer yield as the benchmark. After matching, the second wafers matched with each first wafer in the first batch can come from different second batches. This "wafer yield priority" strategy ignores the impact of processing efficiency (because when the second wafers matched with each first wafer in the first batch come from different second batches, the second wafers matched with each first wafer in the first batch are mounted in different FOUPs, requiring multiple loadings of these different FOUPs during wafer pick-up and bonding, resulting in lower processing efficiency). However, the yield of each first wafer bonded to its matched second wafer is higher.

[0077] Specifically, in one embodiment, the first detection data of each first wafer in the first batch is compared with the second detection data of each second wafer in at least one second batch to perform cross-batch wafer matching analysis to obtain the wafer bonding matching relationship of the first wafer. This may include: setting H1 （B #i） With H2 （Tk #j） Stacking, obtaining stacking detection data H3 for the ikj-th stacked wafer within the k-th stacking batch. （B #i）+（Tk #j） H1 （B #i） H2 refers to the first detection data of the i-th first wafer B#i within the first batch. （Tk #j） This refers to the second detection data of the j-th second wafer Tk #j within the k-th second batch, where k belongs to 1 to n, and n is the total number of wafers in the second batch; according to H3 （B #i）+（Tk #j） Determine the number of qualified stacked dies in the ikj-th stacked wafer, where the ikj-th stacked wafer is the stacked wafer obtained by stacking B #i and Tk #j; determine the optimal stacked wafer corresponding to the i-th first wafer B #i based on the number of dies; obtain the wafer bonding matching relationship of the i-th first wafer B #i based on the optimal stacked wafer corresponding to the i-th first wafer B #i.

[0078] When the matching strategy is "single-wafer yield priority (Wafer yield priority)", the first detection data H1 of the first wafer B #i in the first batch is first selected. （B #i） "The second detection data H2 of the j-th second wafer Tk #j within the k-th second batch" （Tk #j） "Stacking, obtaining the stacking detection data H3 of the ikj-th stacked wafer within the k-th stacking batch." （B #i）+（Tk #j） Stacking the first wafer B #i and the second wafer Tk #j will result in the ikjth stacked wafer.

[0079] Specifically, the first detection data H1 of the first wafer B#i can be obtained first. （B #i）The first wafer defect map and the first grain probe test map (CP map) are stacked to obtain the first stacked inspection data of the first wafer B#i. For example, see [reference]. Figure 5 When i=01, stack 311 and 411 of B #01 to obtain the first stacked detection data 511, stack 312 and 412 of B #02 to obtain the first stacked detection data 512, and so on. Also, the second detection data H2 of the second wafer Tk #j can be... （Tk #j） The second wafer defect map and the second grain probe test map are stacked to obtain the second stack inspection data of the second wafer Tk #j. For example, see [reference]. Figure 5 When k=01 and j=01, stack 321 and 421 of T01 #01 to obtain the second stacked detection data 521, stack 322 and 422 of T01 #02 to obtain the second stacked detection data 522, and so on. Further, stack the first stacked detection data of the first wafer B #i and the second stacked detection data of the second wafer Tk #j to obtain the stacked detection data H3 of the ikj-th stacked wafer. （B #i）+（Tk #j） For example, see Figure 6 When i=01, k=01 and j=01, the first stacking detection data 511 of B #01 and the second stacking detection data 523 of T01 #03 are stacked to obtain the stacking detection data (H3) of the 01.01.03rd stacked wafer. （B #01）+（T01 #03） )613, stack the first stack detection data 515 of B #05 and the second stack detection data 521 of T01 #01 to obtain the stack detection data (H3) of the 05.01.01st stacked wafer. （B #05）+（T01 #01） )611 and so on.

[0080] Following the above superposition method, the first detection data H1 of the first wafer B #i within the first batch can be performed under different combinations of i, k, and j. （B #i） "The second detection data H2 of the j-th second wafer Tk #j within the k-th second batch" （Tk #j） "Stacking, obtaining the stacking detection data H3 of the ikj-th stacked wafer within the k-th stacking batch." （B #i）+（Tk #j） .

[0081] Furthermore, according to H3 of the ikj-th stacked wafer （B #i）+（Tk #j）The number of qualified stacked dies in the ikj-th stacked wafer can be determined. The ikj-th stacked wafer is obtained by stacking B #i and Tk #j. Dies at the same position in B #i and Tk #j are stacked vertically to form the stacked dies in the stacked wafer. Only stacked dies with two good dies are considered qualified stacked dies. The position of a qualified stacked die is marked "OK" in the stacking inspection data of the stacked wafer. The position of a defective stacked die is marked "NG". A defective stacked die is formed when either of the two dies has a defect (as shown by the black triangle). Two dies located at the edge of the wafer are also considered defective stacked dies. A defective stacked die is formed when either of the two dies is marked "2". For example... Figure 6 As shown, based on the stacking detection data (H3) of the 05.01.01st stacked wafer... （B #05）+（T01 #01） )611, we can count the number of qualified stacked dies marked OK to be 14, and so on, to get the number of qualified stacked dies in all stacked wafers of the ikjth digit.

[0082] Furthermore, based on the number of qualified stacked dies within the ikj-th stacked wafer, the optimal stacked wafer corresponding to the i-th first wafer B#i is determined. For example, when i=01, stacking the first wafer B#01 can yield several stacked wafers (01.kj-th stacked wafers (different combinations of k and j correspond to different stacked wafers)). The stacked wafer with the highest number of qualified stacked dies among these several stacked wafers is the optimal stacked wafer corresponding to the first wafer B#01; when i=02, stacking the first wafer B#02 can yield several stacked wafers (02.kj-th stacked wafers (different combinations of k and j correspond to different stacked wafers)). The stacked wafer with the highest number of qualified stacked dies among these several stacked wafers is the optimal stacked wafer corresponding to the first wafer B#02; and so on, the optimal stacked wafer corresponding to each i-th first wafer B#i can be obtained.

[0083] Furthermore, based on the optimal stacked wafers corresponding to the i-th first wafer B#i, the wafer bonding matching relationship of the i-th first wafer B#i can be obtained. For example, when the optimal stacked wafers corresponding to the first wafer B #01 are stacked as "B #01 and T01 #05", the wafer bonding matching relationship of the 01st first wafer B #01 is "B #01 & T01 #05"; when the optimal stacked wafers corresponding to the first wafer B #02 are stacked as "B #02 and T05 #14", the wafer bonding matching relationship of the 02nd first wafer B #02 is "B #02 & T05 #14"; when the optimal stacked wafers corresponding to the first wafer B #03 are stacked as "B #03 and T03 #25", the wafer bonding matching relationship of the 03rd first wafer B #03 is "B #03 & T03 #25"; when the optimal stacked wafers corresponding to the first wafer B #04 are stacked as "B #04 and T02 #02", the wafer bonding matching relationship of the 04th first wafer B #04 is "B #01 & T01 #05"; When the optimal stacked wafer corresponding to the first wafer B #05 is obtained by stacking "B #05 and T04 #13", the wafer bonding matching relationship of the 5th first wafer B #05 is "B #05&T04 #13". It can be seen that the second wafer matched by the i-th first wafer B #i in the first batch B comes from different second batches T01, T02, T03, T04, T05.

[0084] As mentioned above, the strategy of "wafer yield priority" ignores the impact of processing efficiency (because when the second wafers matched by each first wafer in the first batch come from different second batches, the second wafers matched by each first wafer in the first batch are loaded in different carriers (FOUP), and multiple loading of these different carriers is required when performing wafer pick-up and bonding, resulting in low processing efficiency). However, each first wafer B #i usually has a higher yield after being bonded to its matched second wafer.

[0085] In one embodiment, controlling the wafer bonding equipment to perform wafer pick-up and bonding based on the wafer bonding matching relationship may include: dividing the second wafers matched by the first wafer into sub-batches from the second batch of sources according to the wafer bonding matching relationship of the first wafer; generating a new bonding matching relationship for the first wafer, the new bonding matching relationship describing a second wafer matched with the first wafer in the sub-batch; and controlling the wafer bonding equipment to perform wafer pick-up and bonding based on the new bonding matching relationship.

[0086] The manufacturing execution system (MES) can: based on the wafer bonding matching relationship of the first wafer, divide the second wafers that match the first wafer into sub-batches from the second batch of sources; generate a new bonding matching relationship for the first wafer, which describes a second wafer that matches the first wafer in the sub-batch. Then, the MES can convert the new bonding matching relationship into executable production instructions and issue these instructions to the equipment automation system. The equipment automation system can then load the carriers containing the first and second wafers onto the wafer bonding equipment according to these production instructions, and direct robotic arms or similar devices to remove the first and second wafers with the new bonding matching relationship from the carriers and feed them into the wafer bonding equipment for bonding.

[0087] Specifically, when adopting a "batch yield priority (LOT yield priority)" strategy, all second wafers matched by the i-th first wafer B#i within the first batch B come from the same second batch (called the parent batch, such as T01). If the parent batch includes C1 second wafers, and C2 (C2 < C1) of these second wafers are matched by first wafers within the first batch B, then, based on the wafer bonding matching relationship of the first wafers, when dividing the second wafers matched by the first wafers from the source second batch (parent batch) into sub-batches, it can be done by "separating the second wafers matched by the first wafers within the first batch B from the parent batch to form a new sub-batch, and the second wafers not separated from the parent batch to form another sub-batch." Then, a new bonding matching relationship for the first wafer is generated, describing a second wafer in the sub-batch that matches the first wafer.

[0088] For example, if a "batch yield priority (LOT yield priority)" strategy is used, the resulting wafer bonding matching relationships include: “B #01&T01 #03”; “B #02&T01 #04”; “B #03&T01 #05”; “B #04&T01 #02”; "B #05&T01 #01".

[0089] If T01 (mother batch) includes 15 second wafers from T01 #01 to T01 #15, then the second wafers matched with the first wafer in batch B include T01 #01 to T01 #05. T01 #01 to T01 #05 can be divided into a single sub-batch, which can be: T01 (#01-#15, 15 pieces in total) = T01.00 (#06-#15, 10 pieces in total) + T01.01 (#01-#05, 5 pieces in total); at this time, sub-batch T01.00 and T01.01 are both in the same vehicle (FOUP 01 where T01 is located).

[0090] Furthermore, the new bonding matching relationships for generating the first wafer may include: “B #01&T01.01 #03”; “B #02&T01.01 #04”; “B #03&T01.01 #05”; “B #04&T01.01 #02”; "B #05&T01.01 #01".

[0091] Subsequently, based on this new bonding matching relationship, the second wafer belonging to sub-batch T01.01 (#01-#05, a total of 5 wafers) can be accurately taken out from the carrier (FOUP 01) and bonded to the first wafer matched in the first batch. By dividing the batch into sub-batch according to this logic, the consistency of the materials on the production line can be further ensured during wafer bonding, and the reliability of wafer bonding can be further guaranteed.

[0092] Furthermore, when adopting a "wafer yield priority" strategy, the second wafers matched by the first wafer B#i within the first batch B will come from different second batches (called parent batches, such as T01, T02, T03, T04, T05). Similarly, if the parent batch includes C1 second wafers, and C2 (C2 less than C1) of these second wafers are matched by the first wafer within the first batch B, then, based on the wafer bonding matching relationship of the first wafer, when dividing the second wafers matched by the first wafer into sub-batches from the source second batch (parent batch), it can be done by "separating the second wafers matched by the first wafer within the first batch B from the parent batch to form a new sub-batch, and the second wafers not separated from the parent batch to form another sub-batch." Then, a new bonding matching relationship for the first wafer is generated, describing a second wafer matched with the first wafer in the sub-batch.

[0093] For example, wafer bonding matching relationships obtained using a "wafer yield priority" strategy include: “B #01&T01 #05”; “B #02&T05 #14”; “B #03&T03 #25”; “B #04&T02 #02”; "B #05&T04 #13".

[0094] Furthermore, if T01 (#01-#05, 5 wafers total), T02 (#01-#10, 10 wafers total), T03 (#01-#25, 25 wafers total), T04 (#01-#25, 25 wafers total), and T05 (#01-#20, 20 wafers total), then the second wafers matched by the first wafer in B include T01 #05, T05 #14, T03 #25, T02 #02, and T04 #13. These second wafers are respectively divided from their respective parent batches into a sub-batch, which can be: T01 (#01-#05, 5 pieces in total) = T01.00 (#01-#04, 4 pieces in total) + T01.01 (#05, 1 piece in total); At this time, sub-batch T01.00 and sub-batch T01.01 are both in the same vehicle (FOUP 01 where T01 is located); T02 (#01-#010, 10 pieces in total) = T02.00 (#01, #03-#10, 9 pieces in total) + T02.01 (#02, 1 piece in total); at this time, sub-batch T02.00 and sub-batch T02.01 are both in the same vehicle (FOUP 02 where T02 is located); T03 (#01-#25, 25 pieces in total) = T03.00 (#01-#24, 24 pieces in total) + T03.01 (#25, 1 piece in total); at this time, sub-batch T03.00 and sub-batch T03.01 are both in the same vehicle (FOUP 03 where T03 is located); T04 (#01-#25, 25 pieces in total) = T04.00 (#01-#12, #14-#25, 24 pieces in total) + T04.01 (#13, 1 piece in total); at this time, sub-batch T04.00 and sub-batch T04.01 are both in the same vehicle (FOUP 04 where T04 is located); T05 (#01-#20, 20 pieces in total) = T05.00 (#01-#13, #15-#20, 19 pieces in total) + T05.01 (#14, 1 piece in total); at this time, sub-batch T05.00 and sub-batch T05.01 are both in the same vehicle (FOUP 05 where T05 is located).

[0095] Furthermore, the new bonding matching relationships for generating the first wafer may include: “B #01&T01.01 #05”; "B #02&T05.01 #14"; "B #03&T03.01 #25"; "B #04&T02.01 #02"; "B #05&T04.01 #13".

[0096] In addition, T01.01 #05, T05.01 #14, T03.01 #25, T02.01 #02, and T04.01 #13 can be transferred to the same new vehicle (FOUP 20).

[0097] Subsequently, based on this new bonding matching relationship, a second wafer matching the first wafer in the first batch can be accurately taken out from the new carrier (FOUP 20) and bonded. By dividing the batches into sub-batch according to this logic, the consistency of the materials on the production line can be ensured during wafer bonding, further ensuring the reliability of wafer bonding.

[0098] To facilitate better implementation of the wafer bonding control method provided in the embodiments of this application, the embodiments of this application also provide a wafer bonding control system applicable to the above-described wafer bonding control method. The meanings of the terms used are the same as in the above-described wafer bonding control method, and specific implementation details can be found in the descriptions in the method embodiments. Figure 8 A block diagram of a wafer bonding control system according to an embodiment of this application is shown.

[0099] like Figure 8 As shown, the wafer bonding control system 800 may include a matching module 810 and a control module 820; The matching module 810 can be used to: acquire first detection data of a first wafer included in the first batch, and acquire second detection data of a second wafer included in at least one second batch; and perform wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer, wherein the wafer bonding matching relationship describes a second wafer that matches the first wafer in the at least one second batch; the control module 820 can be used to: control the wafer bonding equipment to perform wafer picking and bonding according to the wafer bonding matching relationship.

[0100] Furthermore, in one embodiment, see [reference] Figure 9 The wafer bonding control system includes a yield management system 910, a manufacturing execution system 920, and an equipment automation system 930; the yield management system 910 includes the matching module 810; the control module 820 belongs to the manufacturing execution system 920 and the equipment automation system 930.

[0101] Furthermore, in one embodiment, the control module 820 integrates the unit 821 and the control unit 822; the manufacturing execution system 920 includes the integration unit 821; and the equipment automation system 930 includes the control unit 822. The integration unit is used to: divide the second wafers matched with the first wafer into sub-batches from the second batch of sources according to the wafer bonding matching relationship of the first wafer; and generate a new bonding matching relationship for the first wafer, the new bonding matching relationship describing a second wafer matched with the first wafer in the sub-batch. The control unit is used to: control the wafer bonding equipment to perform wafer picking and bonding according to the new bonding matching relationship.

[0102] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0103] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0104] It should be understood that this application is not limited to the embodiments described above and shown in the accompanying drawings, but various modifications and changes can be made without departing from its scope.

Claims

1. A wafer bonding control method characterized by, The method comprises: obtaining first detection data of a first wafer included in a first batch, and obtaining second detection data of a second wafer included in at least one second batch; performing wafer matching analysis on the first detection data and the second detection data to obtain a wafer bonding matching relationship of the first wafer, the wafer bonding matching relationship describing a second wafer matched with the first wafer in the at least one second batch; controlling a wafer bonding device to perform wafer picking bonding according to the wafer bonding matching relationship.

2. The method of claim 1, wherein, The wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer comprises: performing wafer matching analysis on the first detection data and the second detection data by a yield management system to obtain the wafer bonding matching relationship of the first wafer, and transmitting the wafer bonding matching relationship to a manufacturing execution system; The controlling of the wafer bonding device to perform wafer picking bonding according to the wafer bonding matching relationship comprises: controlling the wafer bonding device to perform wafer picking bonding according to the wafer bonding matching relationship by the manufacturing execution system and a device automation system.

3. The method according to claim 1 or 2, characterized in that, The wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer comprises: performing intra-batch wafer matching analysis on the first detection data of the first wafer in the first batch and the second detection data of the second wafer in each of the second batches to obtain the wafer bonding matching relationship of the first wafer.

4. The method according to claim 1 or 2, characterized in that, The wafer matching analysis on the first detection data and the second detection data to obtain the wafer bonding matching relationship of the first wafer comprises: performing cross-batch wafer matching analysis on the first detection data of each of the first wafers in the first batch and the second detection data of each of the second wafers in the at least one second batch to obtain the wafer bonding matching relationship of the first wafer.

5. The method of claim 3, wherein, The intra-batch wafer matching analysis on the first detection data of the first wafer in the first batch and the second detection data of the second wafer in each of the second batches to obtain the wafer bonding matching relationship of the first wafer comprises: H1 （B #i） and H2 （Tk #j） stacking, to obtain the stacking detection data H3 of the i.k.jth wafer in the kth stacking batch （B #i）+（Tk #j） , H1 （B #i） denotes the first detection data of the ith first wafer B #i in the first batch, H2 （Tk #j） denotes the second detection data of the jth second wafer Tk #j in the kth second batch, k belongs to 1 to n, and n is the total number of the second batches; According to H3 （B #i）+（Tk #j） determining a die number of qualified stacked dies in an i.k.jth stacked wafer, the i.k.jth stacked wafer being a stacked wafer stacked by Bi and Tk #j; determining a kth optimal stacking batch according to the number of dies; determining a target optimal stacking batch from the n optimal stacking batches, and obtaining a wafer bonding matching relationship of an ith first wafer B #i according to the target optimal stacking batch.

6. The method of claim 4, wherein, The cross-batch wafer matching analysis on the first detection data of each of the first wafers in the first batch and the second detection data of each of the second wafers in the at least one second batch to obtain the wafer bonding matching relationship of the first wafer comprises: H1 （B #i） and H2 （Tk #j） stacking, to obtain the stacking detection data H3 of the i.k.jth wafer in the kth stacking batch （B #i）+（Tk #j） , H1 （B #i） denotes the first detection data of the ith first wafer B #i in the first batch, H2 （Tk #j） denotes the second detection data of the jth second wafer Tk #j in the kth second batch, k belongs to 1 to n, and n is the total number of the second batches; According to H3 （B #i）+（Tk #j） determining a die number of qualified stacked dies in an i.k.jth stacked wafer, the i.k.jth stacked wafer being a stacked wafer stacked by Bi and Tk #j; determining an optimal stacking wafer corresponding to an ith first wafer B #i according to the number of dies; obtaining a wafer bonding matching relationship of the ith first wafer B #i according to the optimal stacking wafer corresponding to the ith first wafer B #i.

7. The method according to claim 1 or 2, characterized in that, The controlling of the wafer bonding device to perform wafer picking bonding according to the wafer bonding matching relationship comprises: According to the wafer bonding matching relationship of the first wafer, a second wafer matched with the first wafer is batched into a sub-batch from a source second batch; A new bonding matching relationship of the first wafer is generated, the new bonding matching relationship describing one second wafer matched with the first wafer in the sub-batch; According to the new bonding matching relationship, a wafer bonding device is controlled to perform wafer pick-and-place bonding.

8. A wafer bonding control system, comprising: The wafer bonding control system comprises a matching module and a control module; The matching module is configured to: acquire first detection data of a first wafer included in a first batch, and acquire second detection data of a second wafer included in at least one second batch; and perform wafer matching analysis on the first detection data and the second detection data to obtain a wafer bonding matching relationship of the first wafer, the wafer bonding matching relationship describing one second wafer matched with the first wafer in the at least one second batch; The control module is configured to: control a wafer bonding device to perform wafer pick-and-place bonding according to the wafer bonding matching relationship.

9. The wafer bonding control system of claim 8, wherein, The wafer bonding control system comprises a yield management system, a manufacturing execution system and a device automation system; the matching module is included in the yield management system; the control module belongs to the manufacturing execution system and the device automation system.

10. The wafer bonding control system of claim 9, wherein, The control module comprises an integration unit and a control unit, the integration unit being included in the manufacturing execution system, and the control unit being included in the device automation system; The integration unit is configured to: according to the wafer bonding matching relationship of the first wafer, batch a second wafer matched with the first wafer into a sub-batch from a source second batch; and generate a new bonding matching relationship of the first wafer, the new bonding matching relationship describing one second wafer matched with the first wafer in the sub-batch; The control unit is configured to: control a wafer bonding device to perform wafer pick-and-place bonding according to the new bonding matching relationship.

Images