Wafer thinning method, apparatus, storage medium, and program product

Abstract

The application provides a wafer thinning method, device, storage medium and program product, and belongs to the technical field of semiconductor manufacturing. The method comprises the following steps: in a first grinding stage, a wafer thickness measuring device is used to detect the wafer thickness in real time to control grinding until the detected wafer thickness reaches a lower limit of detection; in response to the wafer thickness reaching the lower limit of detection, a second grinding stage is triggered: determining an actual position of a Z-axis in the wafer thinning device as a first position; determining a remaining removal amount of the wafer based on a target thickness and the lower limit of detection; determining a first loss amount of a grinding wheel in a process of grinding the wafer thickness from the lower limit of detection to the target thickness based on a loss change rate and the remaining removal amount; determining a target feed position of the Z-axis based on the first position, the remaining removal amount and the first loss amount; and controlling the Z-axis to feed to the target feed position to complete wafer grinding. The application can meet the wafer thinning demand of ultra-thin wafers and realize stable grinding with high precision.

Description

Technical Field

[0001] This application belongs to the field of semiconductor manufacturing technology, and more specifically, relates to a wafer thinning method, apparatus, storage medium, and process product. Background Technology

[0002] Wafer thinning is a key process in semiconductor manufacturing. By thinning wafers, the packaging height can be reduced, the chip package size can be minimized, and the chip's heat dissipation efficiency can be effectively improved, electrical and mechanical performance can be optimized, and the workload of subsequent dicing can be reduced. As device performance requirements continue to increase, the demand for ultra-thin wafer thinning is becoming increasingly urgent, which places higher demands on the thickness monitoring capabilities during the grinding process.

[0003] Currently, wafer thinning primarily employs grinding processes, using grinding wheels to remove material from the back side of the wafer. During grinding, wafer thickness measurement devices (such as non-contact optical measuring instruments) are typically used to monitor the wafer thickness in real time to precisely control the grinding endpoint. However, when the real-time wafer thickness decreases to near or beyond the detection limit of the wafer thickness measurement device, effective monitoring of the wafer thickness becomes impossible. This problem is particularly pronounced in ultra-thin wafers. Summary of the Invention

[0004] In view of the above problems, this application provides a wafer thinning method, apparatus, storage medium and program product, thereby solving or at least alleviating one or more of the above-mentioned problems and other problems existing in the prior art.

[0005] A first aspect of this application provides a wafer thinning method, comprising:

[0006] The target thickness for wafer grinding and the detection limit of the wafer thickness measuring device are obtained; wherein the target thickness is less than the detection limit.

[0007] In the first grinding stage, the wafer thickness is detected in real time using a wafer thickness measuring device to control the grinding process until the detected wafer thickness reaches the detection lower limit.

[0008] The second grinding stage is triggered when the wafer thickness reaches the detection limit.

[0009] The actual position of the Z-axis in the wafer thinning equipment is obtained by using a grating ruler and is taken as the first position;

[0010] The remaining amount of material removed from the wafer is determined based on the target thickness and the detection limit.

[0011] The wear rate of the grinding wheel is obtained. The wear rate is the rate of change of the grinding wheel wear during the process of grinding the wafer from the initial thickness to the detection limit.

[0012] Based on the rate of change of wear and the amount of remaining removal, the first wear of the grinding wheel is determined during the process of the wafer thinning equipment grinding the wafer thickness from the detection lower limit to the target thickness.

[0013] Based on the first position, the remaining removal amount, and the first loss amount, determine the target feed position of the Z-axis;

[0014] Control the Z-axis feed to the target feed position to complete the wafer grinding.

[0015] In one embodiment, before obtaining the wear rate of the grinding wheel, the wafer thinning method further includes:

[0016] Obtain the initial reference position of the Z-axis. The initial reference position is the reference position of the Z-axis before the wafer grinding begins.

[0017] Based on the first position, the initial reference position, and the detection lower limit, the second loss amount is determined. The second loss amount is the wear amount of the grinding wheel during the process of grinding the wafer from the initial thickness to the detection lower limit.

[0018] Calculate the change in wafer thickness from the initial thickness to the detection limit during the grinding process;

[0019] The ratio between the second loss and the thickness change is determined as the wear rate of the grinding wheel.

[0020] In one embodiment, determining the second loss amount based on the first position, the initial reference position, and the detection lower limit includes:

[0021] Based on the initial reference position and the detection lower limit, the second position is determined. The second position is the theoretical position of the Z-axis when the wafer is ground from the initial thickness to the detection lower limit.

[0022] The difference between the first position and the second position is determined as the second loss.

[0023] In one embodiment, before obtaining the wear rate of the grinding wheel, the wafer thinning method further includes:

[0024] The amount of thickness removed and the corresponding grinding wheel wear of the first historical grinding wafer during the grinding process are obtained. The first historical grinding wafer is the first preset number of wafers that have been recently ground.

[0025] Based on the amount of thickness removed during the grinding process of the first historical grinding wafer and the corresponding grinding wheel wear, the linear wear rate of the grinding wheel was obtained by fitting.

[0026] The wear rate of the grinding wheel is determined based on the linear wear rate.

[0027] In one embodiment, determining the wear rate of the grinding wheel based on the linear wear rate includes:

[0028] Obtain the actual wear rate corresponding to the second historical grinding wafer; the second historical grinding wafer is the second preset number of wafers that have been recently ground, and the actual wear rate is determined based on the historical wear amount and historical removal amount obtained by real-time measurement; the historical wear amount is the wear amount of the grinding wheel during the grinding process of the second historical grinding wafer; the historical removal amount is the amount of thickness removed from the second historical grinding wafer.

[0029] The error between the actual rate of change of loss and the linear rate of change of loss is determined as the prediction error.

[0030] If the prediction error is not greater than the preset error threshold, then the linear loss change rate is determined as the wear change rate of the grinding wheel.

[0031] If the prediction error is greater than the preset error threshold, the nonlinear wear rate of the grinding wheel is obtained, and the wear rate of the grinding wheel is determined based on the linear wear rate and the nonlinear wear rate.

[0032] The rate of change of nonlinear loss is determined in the following way:

[0033] Obtain the cumulative total grinding thickness corresponding to the grinding wheel. The cumulative total grinding thickness is the cumulative wafer thickness ground by the grinding wheel since its installation.

[0034] Obtain the motor power and coolant flow rate per unit time of the spindle in the wafer thinning equipment;

[0035] The nonlinear wear rate of the grinding wheel is determined based on the total cumulative grinding thickness, motor power, and coolant flow rate per unit time.

[0036] In one embodiment, the nonlinear wear rate of the grinding wheel is determined based on the cumulative total grinding thickness, motor power, and coolant flow rate per unit time, including:

[0037] The additional wear rate of the grinding wheel caused by the cumulative total grinding thickness, the fluctuation of the motor power, and the coolant flow rate per unit time are determined respectively, and are respectively used as the first additional wear rate, the second additional wear rate, and the third additional wear rate.

[0038] The nonlinear loss rate is determined based on the first additional loss rate of change, the second additional loss rate of change, the third additional loss rate of change, and their respective weight parameters.

[0039] In one embodiment, the weight parameters are obtained based on a recursive least squares algorithm and iteratively updated in the following manner:

[0040] Obtain the weight parameters corresponding to the previous iteration and the historical additional loss change rate. The historical additional loss change rate includes: the first additional loss change rate, the second additional loss change rate, and the third additional loss change rate corresponding to the second historical grinding wafer.

[0041] Based on the historical additional loss change rate and the weight parameters corresponding to the previous iteration, the predicted loss change rate of the second historical grinding wafer is determined.

[0042] Determine the difference between the actual loss change rate and the predicted loss change rate, and update the weight parameters corresponding to the previous iteration based on the difference to obtain the weight parameters corresponding to the current iteration, which are then used as the weight parameters.

[0043] In one embodiment, the wafer thinning method further includes:

[0044] The initial thickness of the wafer is measured using a wafer thickness measuring device;

[0045] Based on the initial thickness, determine whether to grind the wafer;

[0046] If so, the initial grinding position of the Z-axis is determined based on the initial thickness;

[0047] Determine whether there is a risk of collision between the spindle and the wafer in the wafer thinning equipment at the initial grinding position;

[0048] If not, control the Z-axis feed to the initial grinding position;

[0049] If present, an alarm signal will be issued and a shutdown operation will be performed.

[0050] A second aspect of this application provides a wafer thinning method, comprising:

[0051] The target thickness for wafer grinding and the detection limit of the wafer thickness measuring device are obtained; wherein the target thickness is less than the detection limit.

[0052] In the first grinding stage, the wafer thickness is detected in real time using a wafer thickness measuring device to control the grinding process until the detected wafer thickness reaches the detection lower limit.

[0053] In response to the wafer thickness reaching the detection lower limit, the second grinding stage is triggered: based on the target thickness and the detection lower limit, the target feed position of the Z-axis in the wafer thinning equipment is determined, and the wafer grinding is continued until the Z-axis feed reaches the target feed position to complete the wafer grinding.

[0054] In one embodiment, the wafer thickness measuring device is a non-contact measuring device. After obtaining the detection lower limit of the wafer thickness measuring device, the wafer thinning method further includes:

[0055] The detection limit is dynamically calibrated at predetermined time intervals:

[0056] At the current moment, obtain the vibration amplitude and frequency of the wafer thinning equipment, the single sampling time and standard measurement temperature of the wafer thickness measurement device, and the wafer temperature;

[0057] The first error caused by the vibration of the wafer thinning equipment is determined based on the phase matching relationship between the vibration frequency and the single sampling time, as well as the vibration amplitude.

[0058] The second error caused by temperature is determined based on the error between the wafer temperature and the standard measurement temperature;

[0059] The detection limit is calibrated based on the first error and the second error.

[0060] A third aspect of this application provides a wafer thinning apparatus, comprising:

[0061] An adsorption platform is used to support the wafer and drive its rotation.

[0062] The grinding device is lifted and positioned above the adsorption platform. The grinding device includes a spindle, a Z-axis, and grinding wheels for grinding wafers.

[0063] The controller is used to perform the operations corresponding to the wafer thinning method in the above embodiments.

[0064] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the wafer thinning method described above.

[0065] A fifth aspect of this application provides a computer program product, including a computer program or computer-executable instructions, which, when executed by a processor, implement the steps of the wafer thinning method described above.

[0066] The beneficial effects of the embodiments of this application are as follows:

[0067] This application embodiment considers the impact of the measurement range of the wafer thickness measuring device on wafer thickness detection and grinding accuracy during the grinding process of ultra-thin wafers. Therefore, this application embodiment first obtains the detection lower limit of the wafer thickness measuring device to determine the detection effectiveness of the wafer thickness measuring device based on the relative magnitude of the real-time wafer thickness and the detection lower limit. Furthermore, by dynamically calibrating the detection lower limit, compensation is made for measurement errors caused by equipment vibration and wafer temperature changes, making the determination of the detection lower limit more accurate, thereby improving the system's adaptability and robustness under different operating conditions. Based on this, this application embodiment can take corresponding measures to achieve effective monitoring of wafer thickness when the wafer thickness measuring device is effectively detecting and when it is not, thereby ensuring grinding accuracy.

[0068] Specifically, in this embodiment, the wafer grinding process is divided into two grinding stages: a first grinding stage when the wafer thickness has not been ground to the detection lower limit of the wafer thickness measuring device, and a second grinding stage when the wafer thickness is ground from the detection lower limit to the target thickness. In the first grinding stage, this embodiment uses a wafer thickness measuring device to detect the wafer thickness, thereby more accurately controlling the progress of wafer grinding. In the second grinding stage, considering that the wafer thickness measuring device is no longer effective for detecting the wafer thickness, it is no longer used to determine the grinding endpoint. Instead, the target feed position of the Z-axis in the wafer thinning equipment is determined by parameters such as the target thickness of the wafer and the detection lower limit of the wafer thickness measuring device, thus determining the grinding endpoint. This allows for control of the grinding thickness outside the detection range of the wafer thickness measuring device, avoiding detection fluctuations that occur when the wafer thickness exceeds the detection range, thereby reducing the risk of over-grinding and improving the accuracy and yield of wafer grinding.

[0069] Furthermore, in this embodiment, considering the wear and tear of the grinding wheel itself during wafer grinding, this embodiment not only combines the target thickness of the wafer grinding and the detection lower limit of the wafer thickness measuring device when determining the target feed position, but also introduces the wear change rate of the grinding wheel to quantitatively compensate for the amount of wear during the grinding process from the detection lower limit to the target thickness. This effectively eliminates the cumulative feed error caused by grinding wheel wear and ensures the positioning accuracy of the grinding endpoint. This advantage is particularly significant in mass continuous production. Furthermore, this embodiment combines linear and nonlinear wear change rates to effectively compensate for additional losses caused by nonlinear factors such as abrasive passivation and coolant flow fluctuations, further improving the accuracy and stability of wafer grinding and fully meeting the high-precision and stable grinding requirements of ultra-thin wafers. Attached Figure Description

[0070] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0071] Figure 1 This is a schematic diagram of grinding control provided in an embodiment of this application;

[0072] Figure 2 This is a schematic diagram of the structure of a wafer thinning apparatus provided in one embodiment of this application;

[0073] Figure 3 This is a schematic flowchart of a wafer thinning method provided in an embodiment of this application;

[0074] Figure 4 A schematic flowchart of a wafer thinning method provided in another embodiment of this application;

[0075] Figure 5 A schematic flowchart of a wafer thinning method provided in another embodiment of this application;

[0076] Figure 6 A schematic flowchart of a wafer thinning method provided in another embodiment of this application;

[0077] Figure 7 A schematic flowchart for calculating the wear rate of a grinding wheel according to an embodiment of this application;

[0078] Figure 8 A schematic flowchart of a wafer thinning method provided in another embodiment of this application;

[0079] Figure 9 This is a schematic flowchart of a dynamic calibration method for the lower limit of detection provided in an embodiment of this application;

[0080] Figure 10 This is a schematic diagram of the controller provided in one embodiment of this application.

[0081] Figure label:

[0082] 21. Adsorption platform; 22. Grinding device; 23. Rotary worktable; 24. Wafer thickness measuring device; 221. Spindle; 222. Z-axis; 223. Grinding wheel; 241. Swing arm. Detailed Implementation

[0083] To enable those skilled in the art to better understand the technical solutions in the embodiments of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments in the specific implementation of this application should fall within the protection scope of the embodiments of this application.

[0084] To keep the drawings concise, each drawing only schematically shows the parts relevant to the disclosure; these do not represent the actual structure of the product. Furthermore, for ease of understanding, in some drawings, only one of components with the same structure or function is schematically shown, or only one is labeled. In this document, "one" not only means "only one," but can also mean "more than one," and "several" includes "two" and "more than two."

[0085] In this document, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0086] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0087] In the description of this embodiment, terms such as "upper," "lower," "left," and "right" are based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of description and simplification of operation, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0088] Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0089] It should be understood that, unless the context clearly states otherwise, the terms "comprising," "including," or "having" as used herein refer to the presence of an element, but do not exclude the presence or addition of one or more other elements. Furthermore, "comprising" and / or "including" as used herein specify the presence of shapes, numbers, steps, operations, members, elements, and / or combinations thereof, and do not exclude the presence or addition of one or more other shapes, numbers, operations, elements, and / or combinations thereof. Some embodiments of this application are described in detail below with reference to the accompanying drawings. Where there is no conflict between the embodiments, the following embodiments and features can be combined with each other. The steps in the following method embodiments are for illustrative purposes only and are not intended to limit this application.

[0090] To make the objectives, technical solutions, and advantages of this application clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0091] During wafer thinning, wafer thickness is typically measured using a wafer thickness measuring device. In this embodiment, the wafer thickness measuring device can be either a non-contact or contact device; this embodiment does not limit the type of device. See also Figure 1 The following problems exist in the practical application of wafer thickness measurement devices to detect wafer thickness:

[0092] On the one hand, wafer thickness measurement devices have an inherent detection lower limit. When the wafer is ground to a thickness below this detection lower limit, the wafer thickness measurement device will be unable to output valid thickness data, resulting in thickness monitoring failure. For example, refer to... Figure 1 If you want to Figure 1The wafer shown is ground to the target thickness, requiring a reduction of wafer thickness H1 + H2. However, in some cases, when the removed wafer thickness reaches H1, it reaches the detection limit of the wafer thickness measurement device. At this point, if the removed wafer thickness exceeds H1, the device will no longer be able to effectively detect thickness changes, leading to thickness monitoring failure. Here, H1 represents the wafer thickness range that the device can accurately measure; H2 represents the wafer thickness range that the device cannot accurately measure when the real-time wafer thickness is below the detection limit. This problem is particularly prominent for ultra-thin wafers (where the target thickness is often below the detection limit). On the other hand, even if the wafer thickness is above the detection limit, when it approaches the limit, the measurement data fluctuates significantly due to factors such as wafer material and surface condition, easily leading to misjudgments and increasing the risk of over-grinding. Furthermore, considering the continuous wear of the grinding wheel itself during the grinding process... If the grinding endpoint is controlled solely based on wafer thickness measurements or Z-axis feed stroke, ignoring the deviation between the actual removal amount and the theoretical feed amount caused by grinding wheel wear, grinding accuracy will decrease, making it difficult to meet the high-precision processing requirements of ultra-thin wafers. Therefore, this embodiment provides a wafer thinning method to achieve high-precision and high-stability grinding control when the wafer thickness exceeds the detection range of the measuring device.

[0093] First, let's explain the wafer thinning scenarios involved in the wafer thinning method of this application embodiment: Please refer to... Figure 2 , Figure 2 The wafer thinning equipment provided in this embodiment may include: an adsorption platform 21 for supporting and rotating the wafer; a grinding device 22, which is vertically mounted above the adsorption platform 21 and includes a spindle 221, a Z-axis 222, and a grinding wheel 223 for grinding the wafer; and a controller (not shown in the figure). In this embodiment, as... Figure 2 As shown, the wafer thinning equipment has three adsorption platforms 21, which are identical in structure and function. Each adsorption platform 21 carries a wafer and is used to support and rotate the wafer. The three adsorption platforms 21 can rotate around the axis of the rotary table 23, allowing the wafer to rotate between different stations to achieve wafer loading, rough grinding, and fine grinding.

[0094] In this embodiment, the wafer is placed on the adsorption platform 21, which is used to adsorb and fix the wafer and drive it to rotate continuously at a set speed. On this basis, the spindle 221 of the grinding device 22 drives the grinding wheel 223 to rotate, while the Z-axis 222 drives the grinding device 22 to descend, so that the grinding wheel 223 makes contact with the back of the wafer. According to the preset process program, the Z-axis 222 drives the grinding wheel 223 to feed towards the wafer to realize wafer grinding.

[0095] In this embodiment, the wafer thinning equipment may further include a wafer thickness measuring device 24. The swing arm 241 of the wafer thickness measuring device 24 can drive the wafer thickness measuring device 24 to measure the wafer thickness on different adsorption platforms 21. In this embodiment, during the continuous feeding of the Z-axis 222, the wafer thickness data fed back by the wafer thickness measuring device 24 can be acquired in real time. When the wafer thickness approaches or reaches the preset detection lower limit of the wafer thickness measuring device 24, the wafer thickness measuring device 24 stops detecting the wafer thickness. Based on the target thickness and the detection lower limit, the target feed position of the Z-axis 222 is determined, and the feed of the Z-axis 222 is controlled based on the determined target feed position.

[0096] In this embodiment, the controller is used to control the grinding device 22, specifically to control the Z-axis feed process in the grinding device 22. In one embodiment, the controller executes the following wafer thinning method during the thinning process, such as... Figure 3 As shown, the method may include:

[0097] S1011: Obtain the target thickness for wafer grinding and the detection limit of the wafer thickness measuring device.

[0098] In this embodiment, the target thickness is less than the detection limit of the wafer thickness measuring device.

[0099] In this embodiment, the target thickness refers to the final thickness the wafer needs to achieve after grinding, and the detection lower limit refers to the lower limit of the measuring range of the wafer thickness measuring device. When the wafer thickness is less than the detection lower limit, the wafer thickness measuring device cannot complete an effective thickness measurement. The target thickness can be preset; for example, the same target thickness is usually set for wafers in the same batch. In this application, the detection lower limit of the wafer thickness measuring device refers to the minimum critical thickness value at which the device can output a valid thickness measurement value that meets preset accuracy requirements. When the wafer thickness is below this critical value, the output data of the measuring device is invalid or the error exceeds the allowable range. The detection lower limit of the wafer thickness measuring device can be determined with reference to the instruction manual of the wafer thickness measuring device, or it can be set based on experience; this embodiment does not limit this.

[0100] S1021: In the first grinding stage, the wafer thickness is detected in real time using a wafer thickness measuring device to control the grinding process until the detected wafer thickness reaches the detection lower limit.

[0101] In this embodiment, the first grinding stage refers to the grinding stage in which the wafer thickness is ground from its initial thickness to the detection lower limit of the wafer thickness measuring device. During the first grinding stage, since the wafer thickness measuring device can measure the wafer thickness within its effective range (i.e., when the wafer thickness has not reached the detection lower limit), the measurement accuracy of the wafer thickness measuring device is high and the data is real-time. Therefore, in this embodiment, the wafer thickness measuring device is used to detect the wafer thickness during the first grinding stage, and thickness monitoring is performed based on the detected wafer thickness during the wafer grinding process.

[0102] S1031: In response to the wafer thickness reaching the detection lower limit, the second grinding stage is triggered: wafer grinding is achieved by controlling the Z-axis position feed in the wafer thinning equipment.

[0103] In this embodiment, when the wafer thickness reaches the detection lower limit, the wafer thickness measuring device will no longer be able to effectively detect the wafer thickness, thus causing the wafer thickness monitoring to fail. At this time, the Z-axis position can be obtained, and the change in the Z-axis feed position can be used to characterize the change in wafer thickness. Then, wafer grinding can be achieved by controlling the Z-axis position feed amount.

[0104] In this embodiment, wafer grinding by controlling the Z-axis position feed can specifically include: determining the target feed position of the Z-axis in the wafer thinning equipment based on the target thickness and the detection lower limit, and continuing to control the grinding of the wafer until the Z-axis feeds to the target feed position to complete the wafer grinding. Therefore, this embodiment can also refer to... Figure 4 The controller can also perform actions such as thinning during the process. Figure 4 The wafer thinning method shown may include:

[0105] S1012: Obtain the target thickness for wafer grinding and the detection limit of the wafer thickness measuring device, wherein the target thickness is less than the detection limit of the wafer thickness measuring device.

[0106] S1022: In the first grinding stage, the wafer thickness is detected in real time using a wafer thickness measuring device to control the grinding process until the detected wafer thickness reaches the detection lower limit.

[0107] S1032: In response to the wafer thickness reaching the detection lower limit, the second grinding stage is triggered: Based on the target thickness and the detection lower limit, the target feed position of the Z-axis in the wafer thinning equipment is determined, and the wafer is continued to be ground until the Z-axis feed reaches the target feed position to complete the wafer grinding.

[0108] In this embodiment, reference may also be made to Figure 5 , Figure 5 The specific implementation method of step S1032 is given, such as Figure 5As shown, step S1032 can be described in detail as follows:

[0109] S501: Obtain the actual position of the Z-axis in the wafer thinning equipment using a grating ruler, and use it as the first position.

[0110] S502: Determine the remaining amount of material removed from the wafer based on the target thickness and the detection limit.

[0111] S503: Obtain the wear rate of the grinding wheel. The wear rate is the rate of change of the grinding wheel's wear during the process of grinding the wafer from its initial thickness to the detection limit.

[0112] S504: Based on the rate of change of wear and the amount of remaining removal, determine the first wear of the grinding wheel during the process of the wafer thinning equipment grinding the wafer thickness from the detection lower limit to the target thickness.

[0113] S505: Determine the target feed position of the Z-axis based on the first position, the remaining removal amount, and the first loss amount.

[0114] S506: Controls the Z-axis feed to the target feed position to complete the wafer grinding.

[0115] In this embodiment, the remaining removal amount of the wafer is the amount of removal required to grind the wafer from the detection lower limit to the target thickness.

[0116] In this embodiment, the second grinding stage refers to the grinding stage where the wafer thickness is ground from the detection lower limit of the wafer thickness measuring device to the target thickness. In this stage, the actual position of the Z-axis in the wafer thinning equipment can be obtained by means of an optical scale, magnetic scale, or motor encoder, wherein the motor encoder is installed on the motor that controls the lifting and lowering of the Z-axis.

[0117] In this embodiment, when the wafer thickness measuring device detects that the wafer thickness has reached the detection lower limit, the wafer thickness measuring device can no longer effectively measure the wafer thickness. Therefore, in the second grinding stage, this embodiment stops the process of grinding the wafer based on the wafer thickness detected by the wafer thickness measuring device. Instead, it determines the target feed position and controls the Z-axis based on the target feed position to determine the grinding endpoint and complete the wafer grinding.

[0118] In this embodiment, the target feed position refers to the final axial position of the Z-axis after feeding vertically downwards along the Z-axis from the first position, with the feed distance being the sum of the remaining removal amount and the first loss amount. The target feed position can be determined by the target thickness and the detection lower limit.

[0119] In this embodiment, under ideal conditions, the Z-axis axial feed stroke equals the wafer material removal thickness. However, in actual grinding, the grinding wheel continuously rubs against the wafer, resulting in continuous axial wear. This weakens the cutting height of the grinding wheel, meaning the grinding wheel also experiences thickness loss. Consequently, for the same Z-axis feed stroke, the actual wafer material removal thickness is less than the theoretically preset removal thickness. If this wear variable is ignored during stroke calculation, the actual material removal in the later stages will be insufficient, and the final wafer thickness will fail to reach the target thickness, thus failing to meet the wafer thinning requirements.

[0120] Therefore, in this embodiment, the difference between the detection lower limit and the target thickness can be used as the remaining removal amount of the wafer. Combined with the remaining removal amount, the axial wear reduction caused by abrasive grain shedding and substrate wear on the grinding wheel cutting surface during the process cycle of removing the remaining material is calculated. This wear reduction amount is defined as the first loss amount. In this embodiment, the wear change rate refers to the axial wear of the grinding wheel corresponding to the thickness removed per unit wafer during the process of grinding the wafer from its initial thickness to the detection lower limit; it is a dimensionless parameter. The first loss amount can be the product of the wear change rate and the remaining removal amount.

[0121] Based on this, the target feed position can be calculated more accurately by combining the first loss amount, the first position, and the remaining removal amount, thereby ensuring precise grinding.

[0122] In one embodiment of this application, reference may be made to Figure 6 , Figure 6 A detailed process for a wafer thinning method is provided:

[0123] like Figure 6 As shown, after grinding is started, the adsorption platform rotates at the set speed, and the swing arm of the wafer thickness measuring device swings to the measurement position to complete the measurement of the initial thickness of the wafer. Then the Z-axis descends to start grinding. During the grinding process, the wafer thickness measuring device continuously monitors the change in wafer thickness. When the wafer thickness does not reach the detection lower limit of the wafer thickness measuring device, grinding and thickness monitoring are maintained until the wafer thickness reaches the detection lower limit. At this time, the wafer thickness measuring device stops monitoring the thickness, and the wafer thinning equipment continues to control the Z-axis feed until the Z-axis feed is completed (reaching the target feed position), and finally the entire grinding process ends.

[0124] As can be seen from the above, the embodiments of this application take into account the influence of the measurement range of the wafer thickness measuring device on the wafer thickness detection and grinding accuracy during the grinding process of ultra-thin wafers. Therefore, the embodiments of this application first obtain the detection lower limit of the wafer thickness measuring device.

[0125] In this embodiment, the wafer grinding process is divided into two grinding stages: a first grinding stage when the wafer thickness has not been ground to the detection lower limit of the wafer thickness measuring device, and a second grinding stage when the wafer thickness has been ground from the detection lower limit to the target thickness. In the first grinding stage, the wafer thickness is measured using a wafer thickness measuring device to more accurately control the grinding progress. In the second grinding stage, considering that the wafer thickness measuring device is no longer effective for measuring the wafer thickness, it is no longer used to determine the grinding endpoint. Instead, the target feed position of the Z-axis in the wafer thinning equipment is determined using parameters such as the target grinding thickness and the detection lower limit of the wafer thickness measuring device to determine the grinding endpoint. This allows for control of the grinding thickness outside the detection range of the wafer thickness measuring device, avoiding detection fluctuations that occur when the measurement exceeds the device's range. This reduces the risk of over-grinding and improves the accuracy and yield of the wafer grinding process.

[0126] Furthermore, in this embodiment, considering that the grinding wheel itself will also have certain wear during the grinding process of the wafer, this embodiment not only combines the target thickness of the wafer grinding and the detection lower limit of the wafer thickness measuring device when determining the target feed position, but also combines the wear of the grinding wheel during the grinding process from the detection lower limit to the target thickness, thereby improving the accuracy of wafer grinding and meeting the high-precision and stable grinding requirements of the wafer.

[0127] In one embodiment of this application, the wear rate of the grinding wheel can be determined in the following manner:

[0128] Obtain the initial reference position of the Z-axis. The initial reference position is the reference position of the Z-axis before the wafer grinding begins.

[0129] Based on the first position, the initial reference position, and the detection lower limit, the second loss amount is determined. The second loss amount is the wear amount of the grinding wheel during the process of grinding the wafer from the initial thickness to the detection lower limit.

[0130] Calculate the change in wafer thickness from the initial thickness to the detection limit during the grinding process;

[0131] The ratio between the second loss and the thickness change is determined as the wear rate of the grinding wheel.

[0132] In this embodiment, the theoretical position that the Z-axis should reach under ideal wear-free conditions can be determined using the initial reference position and the detection lower limit. The deviation between this theoretical position and the first position is calculated. This deviation characterizes the cumulative axial wear of the grinding wheel during the grinding process, starting from the initial reference position of the Z-axis at the beginning of grinding and ending at the actual Z-axis position (first position) acquired when the wafer thickness reaches the detection lower limit. This wear is the second wear amount described in this embodiment. Based on this, the ratio between the second wear amount and the thickness change amount can be determined as the wear change rate of the grinding wheel.

[0133] In one embodiment of this application, the second loss amount is determined based on the first position, the initial reference position, and the detection lower limit, which may specifically include:

[0134] Based on the initial reference position and the detection lower limit, the second position is determined. The second position is the theoretical position of the Z-axis when the wafer is ground from the initial thickness to the detection lower limit. The difference between the first position and the second position is determined as the second loss amount.

[0135] The determination of the second position based on the initial reference position and the detection lower limit can specifically include:

[0136] The axial position after feeding the Z-axis is determined as the second position:

[0137] Starting from the initial reference position, with the Z-axis vertically downward as the feed direction, and the difference between the initial thickness of the wafer and the detection lower limit as the feed distance, the axial position of the Z-axis after feeding is determined.

[0138] In another embodiment of this application, the wear rate of the grinding wheel can also be determined in the following manner:

[0139] The amount of thickness removed and the corresponding grinding wheel wear of the first historical grinding wafer during the grinding process are obtained. The first historical grinding wafer is the first preset number of wafers that have been recently ground.

[0140] Based on the amount of thickness removed during the grinding process of the first historical grinding wafer and the corresponding grinding wheel wear, the linear wear rate of the grinding wheel was obtained by fitting.

[0141] The wear rate of the grinding wheel is determined based on the linear wear rate.

[0142] In this embodiment, considering that the loss change rate calculated solely based on the grinding data of the current single wafer is easily affected by factors such as instantaneous feed fluctuations during a single processing cycle and uneven cooling flow at a single point, the grinding data of the first historical grinding wafer that was recently completed is selected to determine the loss change rate in this embodiment.

[0143] In one embodiment, the ratio of the grinding wheel wear to the corresponding removal amount during the grinding process of the first historical grinding wafer can be determined as the linear wear rate of the grinding wheel. Alternatively, the slope of the fitted straight line can be solved by data fitting and the slope can be determined as the linear wear rate of the grinding wheel.

[0144] In one embodiment, the linear wear rate of the grinding wheel can be fitted using the least squares method or a sliding window averaging method. Taking the least squares method as an example, the linear wear rate of the grinding wheel can be calculated using the following formula based on the least squares method:

[0145] ,

[0146] in, This represents the linear wear rate of the grinding wheel. Indicates the first The amount of thickness removed during the grinding process of the first historically ground wafer. Indicates the first The amount of thickness removed during the grinding process of a first-ever ground wafer corresponds to the amount of grinding wheel wear. This indicates the first preset quantity, which can be set according to actual needs.

[0147] In one embodiment of this application, determining the wear rate of the grinding wheel based on the linear wear rate can specifically include:

[0148] The linear loss rate is directly used as the wear rate of the grinding wheel.

[0149] In one embodiment of this application, considering that during the grinding process, the wear rate of the grinding wheel is affected not only by linear factors such as thickness removal and grinding wheel wear, but also by various nonlinear factors, this embodiment can further determine the calculation accuracy corresponding to the linear wear rate to improve the accuracy of the target feed position determination. If the calculation accuracy of the linear wear rate meets the requirements, it is directly used as the wear rate of the grinding wheel. If the calculation accuracy of the linear wear rate does not meet the requirements, the nonlinear wear rate of the grinding wheel is also determined, and the final wear rate of the grinding wheel is determined by combining the aforementioned linear and nonlinear wear rates. Based on this, in this embodiment, determining the wear rate of the grinding wheel based on the linear wear rate can specifically include:

[0150] Obtain the actual wear rate corresponding to the second historical grinding wafer; the second historical grinding wafer is the second preset number of wafers that have been recently ground; the actual wear rate is obtained based on the historical wear amount and the historical removal amount obtained by real-time measurement; the historical wear amount is the wear amount of the grinding wheel during the grinding process of the second historical grinding wafer; the historical removal amount is the amount of thickness removed from the second historical grinding wafer.

[0151] The error between the actual rate of change of loss and the linear rate of change of loss obtained by fitting is determined as the prediction error;

[0152] If the prediction error is not greater than the preset error threshold, then the linear loss change rate is determined as the wear change rate of the grinding wheel.

[0153] If the prediction error is greater than the preset error threshold, the nonlinear wear rate of the grinding wheel is obtained, and the wear rate of the grinding wheel is determined based on the linear wear rate and the nonlinear wear rate.

[0154] In this embodiment, the second preset quantity can be less than the first preset quantity or equal to the first preset quantity.

[0155] In this embodiment, the grinding data of the first historical grinding wafer (the thickness removal amount of the wafer and the wear amount of the grinding wheel) is used for data fitting to obtain the linear wear rate; the grinding data of the second historical grinding wafer is used to verify the linear wear rate and determine whether the fitted linear wear rate can effectively characterize the wear rate of the grinding wheel during the grinding process of the second historical grinding wafer. The second preset quantity can be set to 1.

[0156] In this embodiment, if the second preset quantity is greater than 1, the historical wear amount of the second historical grinding wafer can be the sum of the historical wear amounts of each second historical grinding wafer, and the historical removal amount of the second historical grinding wafer can be the sum of the historical removal amounts of each second historical grinding wafer.

[0157] In this embodiment, the actual wear change rate corresponding to the second historical grinding wafer is the ratio of the historical wear amount to the historical removal amount of the second historical grinding wafer.

[0158] In this embodiment, the prediction error is used to characterize the deviation between the fitted linear loss rate of change and the actual loss rate of change obtained from on-site measurement. If the prediction error is not greater than a preset error threshold, it indicates that the linear loss rate of change can effectively characterize the wear rate of the grinding wheel during the grinding process of the second historical grinding wafer. If the prediction error is greater than the preset error threshold, it indicates that the linear loss rate of change cannot effectively characterize the wear rate of the grinding wheel during the grinding process of the second historical grinding wafer. In this case, referring to the aforementioned reasons, the wear rate of the grinding wheel can be determined by combining the linear loss rate of change with the nonlinear loss rate of change. In this embodiment, the sum of the linear loss rate of change and the nonlinear loss rate of change can be determined as the wear rate of the grinding wheel.

[0159] In this embodiment, considering that with the use of the grinding wheel and the increase in cumulative grinding amount, the abrasive grains on the grinding wheel surface will become passivated, which will affect the wear change of the grinding wheel. Therefore, the cumulative grinding amount of the grinding wheel will affect the wear change rate of the grinding wheel. Furthermore, the power fluctuation of the spindle will change the grinding force between the grinding wheel and the wafer, thus also affecting the wear change rate of the grinding wheel. Secondly, in this embodiment, it is also considered that if the coolant flow rate for cooling the grinding wheel is insufficient per unit time during the grinding process, it will lead to severe heating of the grinding wheel, which will accelerate the wear of the abrasive grains in the grinding wheel. Therefore, the coolant flow rate will also affect the wear change rate of the grinding wheel. Therefore, in this embodiment, the calculation of the nonlinear wear change rate mainly considers three influencing factors: the cumulative grinding thickness corresponding to the grinding wheel, the motor power, and the coolant flow rate per unit time. Specifically, the nonlinear wear change rate is determined in the following way:

[0160] Obtain the cumulative total grinding thickness corresponding to the grinding wheel, which is the cumulative wafer thickness ground by the grinding wheel since its installation; obtain the motor power of the spindle and the coolant flow rate per unit time in the wafer thinning equipment; based on the cumulative total grinding thickness, motor power and coolant flow rate per unit time, determine the nonlinear loss change rate of the grinding wheel.

[0161] In this embodiment, the cumulative grinding thickness corresponding to the grinding wheel can be obtained by retrieving the historical grinding work records of the wafer thinning equipment where the grinding wheel is located, and the motor power of the spindle and the coolant flow rate per unit time can be obtained by the corresponding sensors or monitoring devices.

[0162] In this embodiment, the spindle motor power and coolant flow rate per unit time are data from the current wafer grinding process, that is, the spindle motor power and coolant flow rate per unit time during the process of grinding the wafer from its initial thickness to the detection lower limit of the wafer thickness measuring device.

[0163] In one embodiment of this application, the nonlinear wear rate of the grinding wheel is determined based on the cumulative total grinding thickness, motor power, and coolant flow rate per unit time, including:

[0164] The cumulative grinding thickness is determined as the rate of change of additional wear caused by the grinding wheel, which is the first rate of change of additional wear; the fluctuation of motor power is determined as the rate of change of additional wear caused by the grinding wheel, which is the second rate of change of additional wear; and the coolant flow rate per unit time is determined as the rate of change of additional wear caused by the grinding wheel, which is the third rate of change of additional wear.

[0165] The nonlinear loss rate is determined based on the first additional loss rate of change, the second additional loss rate of change, the third additional loss rate of change, and their respective weight parameters.

[0166] In this embodiment, a mapping relationship between each influencing factor and the rate of change of additional wear can be pre-defined. Based on this, the rate of change of additional wear brought to the grinding wheel by each influencing factor is quantitatively characterized using the pre-defined mapping relationship corresponding to each influencing factor, thereby obtaining the first rate of change of additional wear, the second rate of change of additional wear, and the third rate of change of additional wear. Based on this, a weighted calculation is performed on each rate of change of additional wear to obtain the nonlinear rate of change of wear. Based on this, this embodiment provides a method for quantifying the nonlinear rate of change of wear of the grinding wheel, thereby ensuring more accurate calculation of the nonlinear wear of the grinding wheel, and thus more accurately locating the target feed position to achieve precision grinding.

[0167] In this embodiment, the aforementioned preset mapping relationship can be constructed by pre-calibrating a mapping table, or the aforementioned preset mapping relationship can be constructed by establishing a mapping relationship between each influencing factor and the rate of change of additional loss based on the different influence characteristics of each influencing factor. This embodiment does not limit this approach.

[0168] In one possible implementation of this embodiment, the nonlinear wear rate of the grinding wheel can be determined by the following formula:

[0169] ,

[0170] in, The nonlinear loss change rate, This represents the cumulative thickness of the wafer removed since the grinding wheel was installed, which is also the total cumulative grinding thickness. The root mean square ripple value of the spindle power. The coolant flow rate per unit time. This is an exponential growth rate function induced by the cumulative total grinding thickness. The first additional loss change rate is used to characterize the contribution of the cumulative total grinding thickness to the wear of the grinding wheel; The main spindle power fluctuation function. This is the second additional loss variation rate, used to characterize the contribution of spindle power fluctuations to grinding wheel losses; The coolant flow rate penalty function. The third additional loss change rate is used to characterize the contribution of coolant flow rate to the wear of the grinding wheel; These are pre-defined weight parameters.

[0171] in, , express The spindle power signal at any given time, Indicates within the time window The average value of the internal power signal. , The time window represents the duration of the sliding integral window, and its value can range from 0.5s to 2s. In this embodiment, the time window... This can refer to the time window corresponding to the grinding of the wafer from its initial thickness to the detection limit.

[0172] In this embodiment, ,in, For the preset nonlinear loss threshold, when At that time, the rate of change of wear of the grinding wheel is approximately constant, therefore It can take the value 0; when At this point, the wear rate of the grinding wheel begins to increase, requiring the introduction of nonlinear compensation. The characteristic constant of exponential growth, with units of and . same. Used to characterize how quickly the rate of change of wear increases with the increase of the total cumulative grinding thickness. The smaller the value, the faster the rate of change in losses increases.

[0173] In this embodiment, It can be preset. The thickness can be predetermined or obtained by fitting experimental data. For example, continuous grinding can be performed using grinding wheels from the same batch. The thickness of the grinding wheel is measured after grinding a certain number of wafers, and the wear rate is calculated. When the wear rate exceeds 1.2-1.3 times the average value of the initial linear segment, the corresponding cumulative grinding thickness can be recorded as follows: .

[0174] In this embodiment, , The preset standard power.

[0175] In this embodiment, , For the preset standard coolant flow rate, The coolant influence intensity factor ranges from 0.2 to 0.5. In this embodiment, Specifically, it can be the coolant flow rate per unit time during the process of grinding the wafer from its initial thickness to the detection limit of the wafer thickness measuring device.

[0176] As can be seen from the above, this embodiment fits the linear loss rate of change using grinding data from historically ground wafers, avoiding errors caused by random fluctuations in the wafer thinning equipment during the processing of a single wafer. Secondly, this embodiment also uses grinding data from a recently processed second historically ground wafer to verify the fitted linear loss rate of change, determining whether the linear loss rate of change can effectively characterize the wear rate of the grinding wheel during the grinding process of the second historically ground wafer, and thus determining whether a nonlinear loss rate of change needs to be added to the linear loss rate. These steps make the calculation of the grinding wheel wear rate more accurate, improving the wafer grinding precision.

[0177] In one embodiment of this application, reference is made to Figure 7 , Figure 7 A process for calculating the rate of change of wear on grinding wheels is provided:

[0178] Figure 7 In the diagram, StartPos represents the initial grinding position, and DetectLowerLimit represents the detection lower limit. The initial grinding position refers to the Z-axis position when the grinding wheel contacts the upper surface of the wafer before grinding. Figure 7 As shown, after grinding starts, the theoretical Z-axis position when the wafer is thinned to the detection lower limit is first calculated. Then, the initial wafer thickness is measured to determine whether grinding can proceed. If grinding is deemed impossible, the equipment alarms and terminates grinding. If grinding is possible, the initial Z-axis grinding position is determined. Under the premise of determining grinding can proceed, the risk of collision between the spindle and the wafer is assessed. If a collision risk exists, an alarm is triggered. If no collision risk exists, the Z-axis is controlled to rapidly feed to the initial grinding position. After confirmation of the position, the Z-axis is controlled to drive the grinding wheel to slowly feed and begin grinding. During grinding, the wafer thickness change is continuously monitored. Grinding continues as long as the wafer thickness does not reach the detection lower limit of the wafer thickness measuring device. When the wafer thickness reaches the detection lower limit, the current Z-axis position is recorded, and the grinding wheel wear rate is calculated accordingly, thus completing the grinding wheel wear rate calculation process.

[0179] In one embodiment of this application, reference may be made to Figure 8 , Figure 8 A process flow for wafer thinning when the wafer thickness reaches the detection lower limit is provided:

[0180] Figure 8SparkOut indicates idling, Escape indicates the process of the grinding wheel being lifted a small distance to leave the wafer, and Home indicates the initial reference position. When the wafer thickness is reduced to the detection lower limit, the following process is initiated:

[0181] First, calculate the wafer removal amount when the thickness reaches the detection lower limit. Then, calculate the grinding wheel wear change rate. Sequentially calculate the remaining removal amount and the grinding wheel wear amount corresponding to the reduction of the remaining wafer thickness (i.e., the first wear amount). Finally, determine the target feed position of the Z-axis. Then, control the Z-axis to grind at the target feed position. After confirming that it is in place, stop the feed and perform an idle operation. After completion, control the grinding wheel to lift it a small distance away from the wafer. Then, quickly lift the Z-axis to the initial reference position to complete the entire grinding process.

[0182] In one embodiment of this application, the weight parameters are obtained based on a recursive least squares algorithm and iteratively updated in the following manner:

[0183] Obtain the weight parameters corresponding to the previous iteration and the historical additional loss change rate, which includes the first additional loss change rate, the second additional loss change rate, and the third additional loss change rate corresponding to the second historical grinding wafer. Based on the historical additional loss change rate and the weight parameters corresponding to the previous iteration, determine the predicted loss change rate of the second historical grinding wafer. Determine the difference between the actual loss change rate and the predicted loss change rate, and update the weight parameters corresponding to the previous iteration based on the difference to obtain the weight parameters corresponding to the current iteration, which serve as the weight parameters.

[0184] In this embodiment, the weighting parameter can be iteratively updated after each grinding of the wafer and the process is completed. The second historical grinding wafer can refer to the wafer that was previously ground using the method described in the foregoing embodiment.

[0185] In this embodiment, the second historical grinding wafer can be any historical grinding wafer, preferably a wafer from the same batch as the current grinding wafer.

[0186] In this embodiment, the wafer currently being ground is taken as the first... The first wafer, the second historical grinding wafer is the first Taking a single wafer as an example, the weighting parameters can be updated using the following formula:

[0187] ,

[0188] in, Indicates the first The weight coefficient vector after wafer grinding is completed and updated. Indicates the first The weight coefficient vector after the wafer grinding is completed and updated, i.e., for the th wafer... The weighted coefficient vector of each wafer before grinding. This represents the gain vector, used to control the current iteration (the...). Update step size and orientation after each wafer grinding is completed. The larger the value, the greater the correction. This represents the measured rate of change of loss. Represents the eigenvectors of the function transpose, . Indicates the predicted rate of change of loss. This represents the difference between the actual rate of change of loss and the predicted rate of change of loss.

[0189] In this embodiment, This is used to determine the adjustment amount of each weight parameter based on the difference between the actual wear rate and the predicted wear rate, thereby achieving adaptive updates of the weight parameters and more accurately calculating the wear rate of the grinding wheel.

[0190] In this embodiment, since the linear loss change rate varies relatively little, the difference between the actual loss change rate and the predicted loss change rate can be directly used to reduce the computational complexity. calculate.

[0191] In another embodiment of this application, to further improve the calculation accuracy, the weight parameters can also be updated using the following formula:

[0192] ,

[0193] in, Indicates the first The linear wear rate of the grinding wheel during the grinding of a wafer.

[0194] In one embodiment of this application, considering that the grinding wheel may be replaced during the wafer grinding process, it is not possible to use a uniform fixed coordinate as the standard initial grinding position for the Z-axis. If the fixed initial grinding position is directly used, it is easy for the grinding wheel to collide with the wafer surface during the high-speed descent. Therefore, in this embodiment, before grinding the wafer, the following steps are also included:

[0195] The initial thickness of the wafer is measured by a wafer thickness measuring device; based on the initial thickness, it is determined whether the wafer should be ground; if so, the initial grinding position of the Z-axis is determined according to the initial thickness, and it is judged whether there is a risk of collision between the spindle and the wafer in the wafer thinning equipment at the initial grinding position; if not, the Z-axis feed is controlled to the initial grinding position.

[0196] In this embodiment, if the initial thickness of the wafer is greater than the target thickness, it is determined that the wafer will be ground; otherwise, it is determined that the wafer will not be ground.

[0197] In one embodiment of this application, it is determined whether there is a risk of collision between the spindle and the wafer in the wafer thinning equipment at the initial grinding position, and may further include: if there is a risk, issuing an alarm signal and performing a shutdown operation.

[0198] In one embodiment of this application, the wafer thickness measuring device is a non-contact measuring device. After obtaining the detection lower limit of the wafer thickness measuring device, the wafer thinning method may further include:

[0199] The detection limit is dynamically calibrated at predetermined time intervals. Specific steps for the dynamic calibration method of the detection limit can be found in [reference needed]. Figure 9 ,like Figure 9 As shown, the dynamic calibration method for the detection lower limit can specifically include:

[0200] S901: At the current moment, acquire the vibration amplitude and frequency of the wafer thinning equipment, the single sampling time and standard measurement temperature of the wafer thickness measurement device, and the wafer temperature;

[0201] S902: The first error caused by the vibration of the wafer thinning equipment is determined based on the phase matching relationship between the vibration frequency and the single sampling time, and the vibration amplitude.

[0202] S903: Determine the second error caused by temperature based on the error between the wafer temperature and the standard measurement temperature;

[0203] S904: Calibrate the detection limit based on the first error and the second error.

[0204] In this embodiment, considering that the detection lower limit of the wafer thickness measuring device may change due to the influence of its measurement accuracy, this embodiment can also dynamically calibrate the detection lower limit in order to achieve wafer grinding more accurately.

[0205] In this embodiment, the predetermined time interval can be determined according to actual process requirements, and this embodiment does not limit it.

[0206] In this embodiment, the non-contact measurement device measures wafer thickness based on the principle of infrared interferometry. On the one hand, considering that the vibration of the wafer thinning equipment affects the optical path length during wafer thickness measurement, thus affecting the measurement accuracy and consequently the detection limit of the wafer thickness measurement device, the vibration information of the wafer thinning equipment can be considered as one of the factors influencing changes in the detection limit. On the other hand, considering that the refractive index of a wafer is significantly affected by temperature, the principle of using a non-contact measurement device to measure wafer thickness is as follows: ,in, For optical path difference, Let be the refractive index of the wafer. Changes in wafer thickness, i.e., the wafer's refractive index, can lead to variations in the measurement accuracy of the wafer thickness measuring device, thus affecting the detection limit. Therefore, wafer temperature can also be considered as one of the factors influencing changes in the detection limit. Thus, this embodiment calibrates the detection limit from both vibration and temperature perspectives.

[0207] In this embodiment, considering that aliasing occurs when the vibration frequency is close to an integer multiple of the single sampling time, resulting in the largest error, the first error caused by the vibration of the wafer thinning equipment can be determined by the phase matching relationship between the vibration frequency and the single sampling time, as well as the vibration amplitude. In one possible implementation, the first error can be determined by the following formula:

[0208] ,

[0209] in, The first error, The amplitude of vibration. The vibration frequency, This is the time for a single sampling. When... This indicates that the vibration of the wafer thinning equipment has completed at least half a cycle within the sampling time of the wafer thickness measurement device. At this time, the displacement measured in a single sampling may fall between the positive and negative peaks of the vibration curve, and can be partially offset by averaging. Therefore, the first error can be taken as... ;then This indicates that the wafer thickness measurement device only captures a small segment of the vibration cycle of the wafer thinning equipment in a single sampling. In this case, the displacement measured by the wafer thickness measurement device will monotonically increase or decrease, and cannot be evenly offset. The first error can be taken as the largest displacement change, i.e. .in, This is to ensure that the error does not exceed the vibration amplitude itself.

[0210] In this embodiment, the standard measurement temperature refers to a pre-set standard temperature that ensures the measurement accuracy of the wafer thickness measurement device. Based on this, if the wafer temperature differs significantly from the standard measurement temperature, the error will also increase. Therefore, the second error caused by temperature can be determined by the error between the wafer temperature and the standard measurement temperature. In one possible implementation, the second error can be determined by the following formula:

[0211] ,

[0212] in, This is the second error. For wafer thickness, The temperature coefficient of the refractive index of the wafer material. This is the error between the wafer temperature and the standard measurement temperature. is the coefficient of thermal expansion of the wafer material.

[0213] In this embodiment, calibrating the detection lower limit based on the first error and the second error may specifically include:

[0214] The first error and the second error are both added to the lower detection limit to achieve detection limit calibration. That is, the calibrated lower detection limit is equal to the original lower detection limit, the first error, and the second error.

[0215] Analysis of this embodiment shows that it provides a dynamic detection lower limit to adapt to actual wafer thinning scenarios, thereby more accurately identifying the effectiveness of the wafer thickness measurement device. When the wafer thickness measurement device cannot effectively measure the thickness, the wafer can be ground in a timely manner by controlling the feed position of the Z-axis, thus improving the wafer grinding accuracy.

[0216] See Figure 10 , Figure 10 This is a schematic block diagram of a controller provided in one embodiment of this application. Figure 10 The controller 1000 in this embodiment may include: a processor 1001, a memory 1002, a communication interface 1003, and a communication bus 1004. The processor 1001, the memory 1002, and the communication interface 1003 communicate with each other through the communication bus 1004. The memory 1002 is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the wafer thinning method in the aforementioned embodiment.

[0217] It should be understood that, in the embodiments of this application, the processor 1001 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0218] The memory 1002 may include read-only memory and random access memory, and provides instructions and data to the processor 1001. A portion of the memory 1002 may also include non-volatile random access memory. For example, the memory 1002 may also store device type information.

[0219] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0220] The computer-readable storage medium can be an internal storage unit of the controller in any of the foregoing embodiments, such as the controller's hard disk or memory. The computer-readable storage medium can also be an external storage device of the controller, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., mounted on the controller. Furthermore, the computer-readable storage medium can include both internal storage units and external storage devices of the controller. The computer-readable storage medium is used to store computer programs and other programs and data required by the controller. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0221] This application provides a computer program product, which includes computer-executable instructions or a computer program. The computer-executable instructions or computer program are stored in a computer-readable storage medium. The processor of the controller reads the computer-executable instructions from the computer-readable storage medium and executes the computer-executable instructions, causing the controller to perform the wafer thinning method described in this application embodiment.

[0222] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0223] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the controller and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0224] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A wafer thinning method, characterized in that, include: The target thickness for wafer grinding and the detection limit of the wafer thickness measuring device are obtained; wherein the target thickness is less than the detection limit; In the first grinding stage, the wafer thickness is detected in real time using the wafer thickness measuring device to control the grinding process until the detected wafer thickness reaches the detection lower limit. In response to the wafer thickness reaching the detection lower limit, the second grinding stage is triggered: The actual position of the Z-axis in the wafer thinning equipment is obtained by using a grating ruler and is taken as the first position; Based on the target thickness and the detection lower limit, the remaining amount of material removed from the wafer is determined; The wear rate of the grinding wheel is obtained, wherein the wear rate of the grinding wheel is the rate of change of wear of the grinding wheel during the process of grinding the wafer from the initial thickness to the detection lower limit; Based on the loss change rate and the remaining removal amount, determine the first wear amount of the grinding wheel during the process of the wafer thinning equipment grinding the wafer thickness from the detection lower limit to the target thickness; Based on the first position, the remaining removal amount, and the first loss amount, determine the target feed position of the Z-axis; Control the Z-axis feed to the target feed position to complete the wafer grinding.

2. The wafer thinning method as described in claim 1, characterized in that, Before obtaining the wear rate of the grinding wheel, the process also includes: Obtain the initial reference position of the Z-axis, which is the reference position of the Z-axis before the grinding of the wafer begins; Based on the first position, the initial reference position, and the detection lower limit, a second loss amount is determined, wherein the second loss amount is the wear amount of the grinding wheel during the process of grinding the wafer from the initial thickness to the detection lower limit; Calculate the change in wafer thickness during the process of grinding the wafer from its initial thickness to the detection lower limit; The ratio between the second amount of wear and the thickness change is determined as the wear rate of the grinding wheel.

3. The wafer thinning method as described in claim 2, characterized in that, Determining the second loss amount based on the first position, the initial reference position, and the detection lower limit includes: Based on the initial reference position and the detection lower limit, a second position is determined, which is the theoretical position of the Z-axis when the wafer is ground from the initial thickness to the detection lower limit; The difference between the first position and the second position is determined as the second loss amount.

4. The wafer thinning method as described in claim 1, characterized in that, Before obtaining the wear rate of the grinding wheel, the process also includes: The amount of thickness removed and the corresponding grinding wheel wear of the first historical grinding wafer during the grinding process are obtained. The first historical grinding wafer is the first preset number of wafers that have been recently ground. Based on the amount of thickness removed during the grinding process of the first historical grinding wafer and the corresponding grinding wheel wear, the linear wear rate of the grinding wheel is obtained by fitting. The wear rate of the grinding wheel is determined based on the linear wear rate.

5. The wafer thinning method as described in claim 4, characterized in that, Determining the wear rate of the grinding wheel based on the linear wear rate includes: Obtain the actual wear rate corresponding to the second historical grinding wafer; the second historical grinding wafer is the second preset number of wafers that have been recently ground, and the actual wear rate is determined based on the historical wear amount and historical removal amount obtained by real-time measurement; the historical wear amount is the wear amount of the grinding wheel during the grinding process of the grinding wheel grinding the second historical grinding wafer; the historical removal amount is the amount of thickness removed from the second historical grinding wafer. The error between the actual loss change rate and the linear loss change rate is determined as the prediction error; If the prediction error is not greater than a preset error threshold, then the linear loss change rate is determined as the wear change rate of the grinding wheel; If the prediction error is greater than the preset error threshold, the nonlinear wear rate of the grinding wheel is obtained, and the wear rate of the grinding wheel is determined based on the linear wear rate and the nonlinear wear rate. The nonlinear loss change rate is determined in the following way: Obtain the cumulative total grinding thickness corresponding to the grinding wheel, where the cumulative total grinding thickness is the wafer thickness that the grinding wheel has ground since its installation. The motor power of the spindle and the coolant flow rate per unit time in the wafer thinning equipment are obtained. The nonlinear wear rate of the grinding wheel is determined based on the cumulative total grinding thickness, the motor power, and the coolant flow rate per unit time.

6. The wafer thinning method as described in claim 5, characterized in that, The determination of the nonlinear wear rate of the grinding wheel based on the cumulative total grinding thickness, the motor power, and the coolant flow rate per unit time includes: The additional wear rate of the grinding wheel caused by the cumulative total grinding thickness, the fluctuation of the motor power, and the coolant flow rate per unit time are determined respectively, and are respectively used as the first additional wear rate, the second additional wear rate, and the third additional wear rate; The nonlinear loss change rate is determined based on the first additional loss change rate, the second additional loss change rate, the third additional loss change rate, and their respective weight parameters.

7. The wafer thinning method as described in claim 1, characterized in that, Before grinding the wafer, the process also includes: The initial thickness of the wafer is measured using the wafer thickness measuring device; Based on the initial thickness, determine whether to grind the wafer; If so, the initial grinding position of the Z-axis is determined based on the initial thickness; Determine whether there is a risk of collision between the spindle and the wafer in the wafer thinning equipment at the initial grinding position; If it does not exist, then control the Z-axis feed to the initial grinding position; If present, an alarm signal will be issued and a shutdown operation will be performed.

8. A wafer thinning method, characterized in that, include: The target thickness for wafer grinding and the detection limit of the wafer thickness measuring device are obtained; wherein the target thickness is less than the detection limit; In the first grinding stage, the wafer thickness is detected in real time using the wafer thickness measuring device to control the grinding process until the detected wafer thickness reaches the detection lower limit. In response to the wafer thickness reaching the detection lower limit, a second grinding stage is triggered: based on the target thickness and the detection lower limit, the target feed position of the Z-axis in the wafer thinning equipment is determined, and the wafer is continued to be ground until the Z-axis feeds to the target feed position to complete the wafer grinding.

9. The wafer thinning method as described in claim 8, characterized in that, The wafer thickness measuring device is a non-contact measuring device. After obtaining the detection lower limit of the wafer thickness measuring device, it further includes: The detection limit is dynamically calibrated at predetermined time intervals. At the current moment, acquire the vibration amplitude and frequency of the wafer thinning equipment, the single sampling time and standard measurement temperature of the wafer thickness measuring device, and the wafer temperature; The first error caused by the vibration of the wafer thinning equipment is determined based on the phase matching relationship between the vibration frequency and the single sampling time, and the vibration amplitude. A second error caused by temperature is determined based on the error between the wafer temperature and the standard measurement temperature; The detection limit is calibrated based on the first error and the second error.

10. A wafer thinning apparatus, characterized in that, include: An adsorption platform is used to support the wafer and drive its rotation. A grinding device is raised and lowered above the adsorption platform. The grinding device includes a spindle, a Z-axis, and a grinding wheel for grinding wafers. A controller for performing operations corresponding to the wafer thinning method as described in any one of claims 1-7 or 8-9.

11. A computer storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the wafer thinning method as described in any one of claims 1-7 or 8-9.

12. A computer program product, characterized in that, Includes computer instructions that instruct a computing device to perform operations corresponding to the wafer thinning method as described in any one of claims 1-7 or 8-9.

Images