A method, system, and apparatus for thinning

Abstract

The application relates to the field of semiconductor silicon wafer processing, in particular to a thinning method applied to a thinning device, the thinning device comprising: a grinding wheel, the edge of the grinding wheel is provided with a plurality of grinding teeth, and the grinding teeth are used for grinding a silicon wafer; a first damage function f' is established for a grinding tooth quantity n, a grinding wheel rotating speed w s , a silicon wafer rotating speed w w , and a pressure P of the grinding wheel on the silicon wafer; the first damage factor refers to a damage depth of the grinding wheel on the silicon wafer; and the first damage factor is determined based on a set damage layer depth and the first damage function f'. The optimal parameters are determined by establishing the silicon wafer damage function, so that the technical effect of improving the thinning quality is achieved.

Description

Technical Field

[0001] This application relates to the field of semiconductor silicon wafer processing, and in particular to a thinning method, system, and apparatus. Background Technology

[0002] Silicon wafer thinning equipment is a key piece of equipment used in semiconductor manufacturing to reduce the thickness of silicon wafers. Its core objective is to meet the requirements of chip packaging, heat dissipation, and electrical performance optimization by precisely removing the surface material of the silicon wafer.

[0003] In existing technologies, silicon wafer thinning is achieved through the interaction between a grinding wheel and the silicon wafer surface. The contact area between the cup-shaped grinding wheel and the silicon wafer is usually a local line contact or point contact. The grinding wheel uses the sharp edges of diamond abrasive grains to impact and cut the silicon wafer surface through high-speed rotation and axial feed. Local contact can concentrate the pressure in a very small area, significantly improving the material removal efficiency. However, during the grinding process of the grinding wheel on the silicon wafer, there are many factors that affect the processing of the silicon wafer, resulting in a thicker silicon wafer damage layer and a decrease in the quality of silicon wafer thinning.

[0004] Therefore, the technical problem with existing technologies is that the thinning quality is relatively low. Summary of the Invention

[0005] This application provides a thinning method, system, and apparatus that determines optimal parameters by establishing a silicon wafer damage function, thereby achieving the technical effect of improving thinning quality.

[0006] Firstly, this application provides a thinning method, which adopts the following technical solution:

[0007] A thinning method is applied to a thinning device, the thinning device comprising:

[0008] A grinding wheel, the edge of which has a plurality of grinding teeth for grinding silicon wafers;

[0009] Establish the number of grinding teeth n and the grinding wheel speed w s Silicon wafer rotation speed w w And the first damage function f' of the pressure P of the grinding wheel on the silicon wafer with respect to the first damage factor; the first damage factor refers to the depth of damage to the silicon wafer by the grinding wheel;

[0010] The first damage factor is determined based on the set damage layer depth and the first damage function f'.

[0011] Preferably, the first damage function f' is:

[0012]

[0013] Where n is the number of grinding teeth; w s The rotational speed of the grinding wheel; w wdenoted as silicon wafer rotation speed; P is the pressure of the grinding wheel on the silicon wafer; e is the influence coefficient of the grinding wheel pressure on the silicon wafer on damage; f is the influence coefficient of silicon wafer rotation speed on silicon wafer damage; g is the influence coefficient of the number of grinding teeth on silicon wafer damage; h is the influence coefficient of grinding wheel rotation speed on silicon wafer damage.

[0014] As a preferred option, it also includes:

[0015] Establish a second damage function g' for sludge removal efficiency E and silica sludge concentration C with respect to the second damage factor; the second damage factor refers to the depth of damage to silicon wafers by silica sludge.

[0016] Based on the set damage layer depth, the first damage function f', and the second damage function g', the first damage factor and the second damage factor are determined.

[0017] As a preferred option, it also includes:

[0018] Establish a sludge removal function with respect to the sludge removal efficiency E, which includes the number of grinding teeth n, the grinding fluid flow rate Q, the silica mud concentration C, and the grinding fluid pressure difference ΔP.

[0019] Based on the collected or acquired data, including the number of grinding teeth n, grinding fluid flow rate Q, silica mud concentration C, grinding fluid pressure difference ΔP, and mud removal function, the mud removal efficiency E is determined.

[0020] Preferably, the sludge removal function is:

[0021]

[0022] Where k1 is the equipment constant; n is the number of grinding teeth; Q is the grinding fluid flow rate; C is the silica mud concentration; ΔP is the grinding fluid pressure difference; a is the influence coefficient of grinding fluid flow rate on mud removal efficiency; b is the influence coefficient of grinding fluid pressure difference on mud removal efficiency; c is the influence coefficient of number of grinding teeth on mud removal efficiency; and d is the influence coefficient of silica mud concentration on mud removal efficiency.

[0023] As a preferred option, it also includes:

[0024] Establish concentration functions of silicon wafer density ρ, silicon wafer area A, silicon wafer thickness removal amount ΔH, grinding fluid flow rate Q, and processing time t with respect to silicon sludge concentration C;

[0025] Based on the collected or acquired data such as silicon wafer density ρ, silicon wafer area A, silicon wafer thickness removal amount ΔH, grinding fluid flow rate Q, processing time t, and concentration function, the silicon sludge concentration C is determined.

[0026] Preferably, the concentration function is:

[0027]

[0028] Where ρ is the silicon wafer density; A is the silicon wafer area; ΔH is the silicon wafer thickness removal amount; Q is the grinding fluid flow rate; and t is the processing time.

[0029] Preferably, the second damage function g' is:

[0030] g'=(1-E) i ·C j

[0031] Where E is the sludge removal efficiency; C is the silica sludge concentration; i is the influence coefficient of the proportion of unremoved silica sludge on the damage; j is the influence coefficient of the silica sludge concentration on the damage.

[0032] A third damage function F is established based on the first damage function f' and the second damage function g'. The third damage function F is:

[0033]

[0034] Wherein, k2 is the grinding wheel damage weighting coefficient; k3 is the silica mud damage weighting coefficient;

[0035] Based on the set damage layer depth and the third damage function F, the number of grinding teeth n and the grinding wheel speed w are determined. s Silicon wafer rotation speed w w The pressure P of the grinding wheel on the silicon wafer and the flow rate Q of the grinding fluid.

[0036] Secondly, the thinning system provided in this application adopts the following technical solution:

[0037] A thinning system, comprising:

[0038] The function creation module is used to establish the number of grinding teeth n and the grinding wheel speed w. s Silicon wafer rotation speed w w The pressure P of the grinding wheel on the silicon wafer and the sludge removal efficiency E are related to the first damage factor and the first damage function f'.

[0039] The setting module is used to set the depth of the damage layer; and

[0040] The determination module is used to determine the first damage factor based on the set damage layer depth and the first damage function f'.

[0041] Thirdly, the thinning device provided in this application adopts the following technical solution:

[0042] A thinning device, the thinning device comprising:

[0043] A stage for supporting and rotating silicon wafers;

[0044] A grinding wheel is located above the stage, and the edge of the grinding wheel has a plurality of grinding teeth for grinding silicon wafers.

[0045] In summary, this application includes at least one of the following beneficial technical effects:

[0046] This application establishes a function for the depth of the damaged layer of the silicon wafer, and the optimal processing parameters can be determined based on the preset depth of the damaged layer, which is beneficial to improving the quality of silicon wafer thinning. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the thinning device described in this application;

[0048] Figure 2 This is a schematic diagram of the grinding wheel of the thinning device described in this application;

[0049] Figure 3 This is a first schematic diagram of the thinning method described in this application;

[0050] Figure 4 This is a second schematic diagram of the thinning method described in this application;

[0051] Figure 5 This is a third schematic diagram of the thinning method described in this application.

[0052] Explanation of reference numerals in the attached figures: 100, stage; 200, grinding wheel; 210, grinding teeth; W, silicon wafer. Detailed Implementation

[0053] The serial numbers assigned to components in this document, such as "first" and "second," are used solely to distinguish the described objects and have no sequential or technical meaning. The terms "connection" and "linkage" used in this application, unless otherwise specified, include both direct and indirect connections (linkages). It should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are used solely for the convenience of describing this application and simplifying the description. They do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0054] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0055] This application provides a thinning method, system, and apparatus that determines optimal parameters by establishing a silicon wafer damage function, thereby achieving the technical effect of improving thinning quality.

[0056] To better understand the above technical solutions, a detailed description of the technical solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit the scope of this application.

[0057] In semiconductor device manufacturing, silicon wafer thinning is a key process for improving device performance and integration. With the rapid development of 5G communication, artificial intelligence, and the Internet of Things, chips are evolving towards higher density, higher frequency, and thinner profiles, placing increasingly stringent requirements on the thickness accuracy and surface quality of silicon wafers. Silicon wafer thinning, by precisely removing surface material, not only reduces chip package size but also lowers parasitic capacitance and resistance, improves signal transmission speed, and enhances heat dissipation performance. For example, when the silicon wafer thickness of power devices is reduced from the traditional 500μm to below 100μm, heat dissipation efficiency can be improved by more than 40%. Among these methods, planar thinning, as the mainstream silicon wafer processing method, directly determines the feasibility of subsequent photolithography, bonding, and other processes, thus affecting device yield and reliability.

[0058] Quality control in planar thinning processes is influenced by the coupling of multiple process parameters. Key parameters of general industry concern include: axial feed rate (directly determining the grinding pressure of the grinding wheel 200 on the silicon wafer W), grinding wheel 200 size (affecting the contact area distribution with the silicon wafer W), grinding time, and grinding wheel speed (affecting the cutting linear velocity). In existing technologies, engineers often optimize the combination of these parameters through orthogonal experimental methods. For example, when the grinding wheel speed is fixed, the material removal rate is controlled by adjusting the ratio of feed rate to time, aiming to reduce the depth of the damaged layer.

[0059] However, actual production data shows that relying solely on the aforementioned parameter adjustments is insufficient to effectively control the damage layer. Experimental records indicate that even after optimizing with traditional parameters, the damage layer depth of the 6-inch silicon wafer W still fluctuates between 8-15 μm, with a surface scratch rate as high as 3%-5%, far exceeding the advanced process requirements of a damage layer ≤5 μm and a scratch rate <0.5%. After multiple process iterations, it was found that even reducing the feed rate by 30% and extending the grinding time to 1.5 times the original plan could not prevent problems such as edge chipping and localized deep scratches on the silicon wafer W, severely restricting the mass production capability of high-precision chips.

[0060] After 18 months of systematic research, the applicant team discovered that the core of the problem lies in the insufficient understanding of the grinding mechanism of the cup-shaped grinding wheel 200. As a special tool for planar thinning, the structural characteristics of the cup-shaped grinding wheel 200 have a decisive influence on the thinning effect. The edge of this type of grinding wheel 200 is machined with an annular array of grinding teeth 210, and the cutting is completed by forming line contact with the silicon wafer W through the edge grinding teeth 210. This line contact mode is different from the traditional surface contact grinding wheel 200. Its local contact stress is much higher than the fracture strength of silicon material, which inevitably leads to a unique damage mechanism.

[0061] Further high-speed imaging and cross-sectional electron microscopy analysis revealed the formation process of dual damage: the first grinding was initiated by direct cutting of the 200 teeth of the grinding wheel, where sharp diamond abrasive grains formed a plastic deformation layer on the surface of silicon wafer W, simultaneously generating radial cracks; the second grinding originated from the secondary effect of silica sludge, where silicon debris generated during grinding mixed with coolant to form silica sludge. If not removed in time, this sludge would form a "grinding medium layer" in the inter-tooth gaps, continuously rubbing against the surface of silicon wafer W as the grinding wheel 200 rotated, causing the initial cracks to propagate and introducing irregular scratches. The superposition of these two types of damage ultimately aggravated the depth of the damage layer, with the edge areas being more severely affected by silica sludge accumulation, exhibiting a damage level 30%-50% higher than the central area.

[0062] The limitation of existing technologies lies in their focus on parameter optimization for a single grinding process. They fail to consider the impact of structural parameters such as the number and shape of the grinding teeth on the first grinding cycle, nor do they establish a quantitative correlation between silicon sludge removal efficiency and damage during the second grinding cycle. Therefore, a systematic control scheme integrating the structural characteristics of the grinding wheel 200, process parameters, and the dual damage mechanism is urgently needed to overcome the current bottleneck in silicon wafer W thinning quality.

[0063] The thinning method of the present invention is mainly applied to thinning devices, such as... Figure 1 , 2As shown, the thinning device includes a grinding wheel 200; the grinding wheel 200 refers to a cup-shaped grinding wheel 200, which is used to thin the silicon wafer W in a planar manner. The edge of the grinding wheel 200 is provided with several grinding teeth 210, which can grind the surface of the silicon wafer W. During the thinning process, by precisely controlling the interaction parameters between the grinding wheel 200 and the silicon wafer W (such as the grinding wheel speed, the silicon wafer speed, the pressure of the grinding wheel on the silicon wafer, etc.), as well as the influence of the silicon sludge generated during the process on the silicon wafer W (such as the silicon sludge concentration, the sludge removal efficiency, etc.), efficient and low-damage thinning of the silicon wafer W can be achieved.

[0064] like Figure 3 As shown, the thinning method includes: S1.1: establishing the number of grinding teeth n and the grinding wheel speed w. s Silicon wafer rotation speed w w And the first damage function f' of the pressure P of the grinding wheel on the silicon wafer with respect to the first damage factor;

[0065] The first damage factor refers to the depth of damage caused by the grinding wheel 200 to the silicon wafer W.

[0066] S1.2: Determine the first damage factor based on the set damage layer depth and the first damage function f'.

[0067] This application first addresses the damage effect of the grinding wheel 200 on the silicon wafer W by establishing a first damage function f' to describe the damage depth (i.e., the first damage factor) of the grinding wheel 200 on the silicon wafer W in relation to the number of grinding teeth n and the grinding wheel rotation speed w. s Silicon wafer rotation speed w w The relationship between the pressure P of the grinding wheel on the silicon wafer and the relationship between the grinding wheel and the silicon wafer; specifically, the first damage function f' is:

[0068]

[0069] Where n is the number of grinding teeth; w s The rotational speed of the grinding wheel; w w denoted as silicon wafer rotation speed; P is the pressure of the grinding wheel on the silicon wafer; e is the influence coefficient of the grinding wheel pressure on the silicon wafer on damage; f is the influence coefficient of silicon wafer rotation speed on silicon wafer W damage; g is the influence coefficient of the number of grinding teeth on silicon wafer W damage; h is the influence coefficient of grinding wheel rotation speed on silicon wafer W damage.

[0070] During the thinning process of silicon wafer W, the mechanical impact damage to silicon wafer W caused by the grinding teeth 210 is mainly determined by contact stress and cutting energy. According to cutting mechanics theory, the cutting force F c The relationship with the grinding wheel 200 parameter is simplified to:

[0071] When the grinding teeth 210 come into contact with the silicon wafer W, the local stress exceeds the fracture strength of silicon, leading to crack propagation; the thickness of the damaged layer D1 (the first damage factor determined by the first damage function f') is affected by the following factors:

[0072] Positive effect: The pressure P of the grinding wheel on the silicon wafer W increases, the load on a single grinding tooth 210 increases, and the crack depth ∝ P. e (e is the coefficient of influence of the pressure of the grinding wheel on the silicon wafer on the damage, i.e., the load sensitivity coefficient);

[0073] silicon wafer rotation speed w w Increased size, accumulated damage ∝ (f is the coefficient of influence of silicon wafer rotation speed on silicon wafer W damage, i.e., the cumulative coefficient);

[0074] Negative effect: As the number of grinding teeth (n) increases, the load on each individual grinding tooth (210) is dispersed, and the damage layer thickness (D1) increases by ∝n. -g (g is the influence coefficient of the number of grinding teeth on the damage of silicon wafer W, i.e., the load dispersion coefficient);

[0075] Grinding wheel speed w s Improved cutting uniformity, and increased damage layer thickness D1∝ (h is the influence coefficient of grinding wheel speed on silicon wafer W damage, i.e., dynamic stability coefficient).

[0076] Assume the relationship between the damage layer depth D1 and the parameters is a power function (due to the nonlinear relationship between the cutting force and the parameters):

[0077]

[0078] Where k0 is a material constant, and since the properties of silicon wafer W are fixed, it is normalized to k0 = 1, resulting in the first damage function f':

[0079]

[0080] Experimental design: Fix three parameters and change only the target parameter to measure the damage depth (e.g., depth measurement under a microscope): determine e, f, g, and h through calibration. For example, calibrating e: fix n = 100, w s =3000rpm,w w =100rpm, change P (0.3-0.8MPa), and fit the slope of lnD1 and lnP as e (typical value e = 1.2);

[0081] For example, calibration g: fixed P = 0.5 MPa, w s =3000rpm,w w =100rpm, change n(80-150), and fit the absolute value of the slope of lnD1 and lnn to g (typical value g=0.9).

[0082] Furthermore, this application considers the wear and tear on the silicon wafer W caused by the friction of silica sludge against the silicon wafer W, thereby damaging the silicon wafer W. In other words, unremoved silica sludge particles will remain between the grinding wheel 200 and the silicon wafer W, causing damage to the silicon wafer W. The higher the silica sludge concentration, the more severe the interaction between particles, resulting in an increased coefficient of friction; the lower the sludge removal efficiency, the higher the proportion of unremoved silica sludge, and the longer the contact time between the particles and the silicon wafer W.

[0083] Based on this, such as Figure 4 As shown, the thinning method provided in this application further includes: S2.1: establishing a second damage function g' of the sludge removal efficiency E and the silica sludge concentration C with respect to the second damage factor; the second damage factor refers to the depth of damage to the silicon wafer W by the silica sludge;

[0084] S2.2: Based on the set damage layer thickness, the first damage function f' and the second damage function g', determine the first damage factor and the second damage factor.

[0085] The process of determining the sludge removal efficiency E is as follows: establish the sludge removal function of the number of grinding teeth n, the flow rate of grinding fluid Q, the concentration of silica sludge C, and the pressure difference of grinding fluid ΔP with respect to the sludge removal efficiency E.

[0086] Based on the collected or acquired data, including the number of grinding teeth n, grinding fluid flow rate Q, silica mud concentration C, grinding fluid pressure difference ΔP, and mud removal function, the mud removal efficiency E is determined.

[0087] Specifically, the sludge removal function is:

[0088]

[0089] Where k1 is the equipment constant; n is the number of grinding teeth; Q is the grinding fluid flow rate; C is the silica sludge concentration; ΔP is the grinding fluid pressure difference; a is the influence coefficient of grinding fluid flow rate on sludge removal efficiency; b is the influence coefficient of grinding fluid pressure difference on sludge removal efficiency; c is the influence coefficient of the number of grinding teeth on sludge removal efficiency; and d is the influence coefficient of silica sludge concentration on sludge removal efficiency. High-precision pressure sensors can be installed inside and outside the grinding wheel 200 to obtain pressure values ​​and calculate the grinding fluid pressure difference ΔP.

[0090] It needs to be explained that the sludge discharge efficiency E is defined as:

[0091]

[0092] Its scouring ability is affected by the hydrodynamics of the grinding fluid: scouring capacity ∝ Q a ·△P b The kinetic energy of the grinding fluid is determined by both the flow rate Q and the pressure difference ΔP.

[0093] Blocking resistance ∝n c ·C dIncreasing the number of grinding teeth (n) will reduce the inter-tooth gap, and increasing the concentration of silica sludge (C) will increase the annual flow rate of fluid.

[0094] Introduce the device constant k1 and establish the relationship:

[0095]

[0096] Furthermore, the above coefficients were calibrated through fluid dynamics experiments: fluorescent silica mud was injected into a grinding wheel 200 simulation device, and the discharge process was tracked by high-speed photography. a, b, c, and d can be determined through calibration experiments. For example, calibrating a: fixing ΔP = 0.2 MPa, n = 100, C = 10%, changing Q (5-20 L / min), and fitting the slopes of lnE and lnQ (typical value a = 0.7); calibrating c: fixing Q = 10 L / min, ΔP = 0.2 MPa, C = 10%, changing n (80-140), and fitting the absolute values ​​of the slopes of lnE and lnn (typical value c = 1.1).

[0097] Furthermore, it also includes: establishing the concentration functions of silicon wafer density ρ, silicon wafer area A, silicon wafer thickness removal amount ΔH, grinding fluid flow rate Q, and processing time t with respect to silicon sludge concentration C;

[0098] Based on the collected or acquired data such as silicon wafer density ρ, silicon wafer area A, silicon wafer thickness removal amount ΔH, grinding fluid flow rate Q, processing time t, and concentration function, the silicon sludge concentration C is determined.

[0099] The concentration function is:

[0100]

[0101] Where ρ is the silicon wafer density; A is the silicon wafer area; ΔH is the silicon wafer thickness removal amount; Q is the grinding fluid flow rate; and t is the processing time.

[0102] It needs to be explained that the amount of silica sludge generated, M... gen =ρ·A·△H(density * area * thickness removed);

[0103] Silica mud discharge M out = C·Q·t(concentration * flow rate * time);

[0104] In steady state, M gen =M out ,Right now:

[0105] ρ·A·△H=C·Q·t

[0106] Furthermore, the second damage function g' is:

[0107] g = (1 - E) i ·C j

[0108] Where E is the sludge removal efficiency; C is the silica sludge concentration; i is the influence coefficient of the proportion of unremoved silica sludge on the damage; and j is the influence coefficient of the silica sludge concentration on the damage.

[0109] The second damage function g' determines the secondary damage depth D2, that is:

[0110] D² = g = (1 - E) i ·C j ;

[0111] The exponents i and j represent the nonlinear sensitivity coefficients of the proportion of unremoved silica sludge (1-E) and the silica sludge concentration C on the secondary damage depth of silicon wafer W, respectively. It can be understood that the higher the proportion of unremoved silica sludge (1-E), the greater the probability of contact between particles and silicon wafer W; the higher the silica sludge concentration C, the more intense the agglomeration between particles, and the greater the friction coefficient.

[0112] Therefore, assume D²∝(1-E) i ·C j Among them, i reflects the sensitivity of insufficient sludge removal efficiency to damage (the larger i is, the more significant the effect of reduced E on D2); j reflects the sensitivity of increased silica sludge concentration to damage (the larger j is, the more significant the effect of increased C on D2). Therefore, i and j can be measured independently. For example, i is measured while the silica sludge concentration C is fixed; j is measured while the sludge removal efficiency E is fixed.

[0113] like Figure 5 As shown in S3.1: A third damage function F is established based on the first damage function f' and the second damage function g'. The third damage function F is:

[0114]

[0115] Wherein, k2 is the damage weighting coefficient of grinding wheel 200; k3 is the damage weighting coefficient of silica mud.

[0116] S3.2: Based on the set damage layer thickness and the third damage function F, determine the number of grinding teeth n and the grinding wheel speed w. s Silicon wafer rotation speed w w The pressure P of the grinding wheel on the silicon wafer and the flow rate Q of the grinding fluid.

[0117] In other words, weighting coefficients k2 and k3 are introduced into the calculation of D = D1 + D2 to balance the contributions of the two types of damage:

[0118]

[0119] The weighting calibration method is implemented by designing a damage-dominant scenario for a 200mm grinding wheel: shutting down the sludge removal system, making E = 0, measuring D, and solving the equation:

[0120] D = k2·f' + k3·Cj

[0121] After experimental fitting of the k2 / k3 ratio, typical values ​​can be taken as k2 = 0.6 and k3 = 0.4.

[0122] Thus, the parameters can be solved in reverse based on the third damage function F:

[0123] Silicon wafer constants: silicon wafer density ρ, silicon wafer area A, silicon wafer thickness removal ΔH;

[0124] Function coefficients: a~j, k1~k3 (calibrated through prior experiments);

[0125] Target damage depth D target (e.g. D) target =2μm);

[0126] based on:

[0127]

[0128] Set initial values, for example, make n = 100, w s =3000rpm; iteratively adjust P,w w ,Q,t, until F≤D target Among these, the constraints on process parameters can be adjusted within the allowable range of the equipment, for example, P min ≤P≤P max .

[0129] More specifically, regarding the target damage depth D target Let n = 100, w s =3000rpm, and given the silicon wafer density ρ, silicon wafer area A, and silicon wafer thickness removal amount ΔH are constant, adjust the grinding wheel pressure P and grinding wheel speed w on the silicon wafer iteratively. w The grinding fluid flow rate Q is used to achieve inverse parameter calculation:

[0130] For example, based on the third damage function F, substituting n=100, w s =3000rpm, current P, w w And the calculated E and C:

[0131]

[0132] Calculate D and D target Deviation (e.g., |DD) target (Continue iteration when |>0.1μm), adjust parameters according to the deviation direction:

[0133] If D > D target If the damage is too great, proceed as follows: Reduce the pressure P of the grinding wheel on the silicon wafer or the grinding wheel speed w.w ; and / or increase the grinding fluid flow rate Q.

[0134] If D > D target If the damage is less, the parameters can be relaxed to improve efficiency: Execute: Increase the pressure P of the grinding wheel on the silicon wafer or the grinding wheel speed w. w ; and / or reduce the grinding fluid flow rate Q.

[0135] The optimal solution is automatically found by adjusting the step size (pressure P of the grinding wheel on the silicon wafer, grinding wheel speed w). w Grinding fluid flow rate (Q).

[0136] In this application, adjusting the grinding fluid flow rate Q will simultaneously affect the silica sludge concentration C and the sludge removal efficiency E, requiring a balance between the weights of the damage layer depth D1 caused by the grinding wheel and the secondary damage depth D2 caused by the silica sludge.

[0137] This application also provides a thinning system, including a function establishment module, a setting module, and a determination module. The function establishment module is used to establish the number of grinding teeth n and the grinding wheel speed w. s Silicon wafer rotation speed w w The first damage function f' is used to determine the first damage factor based on the pressure P of the grinding wheel on the silicon wafer and the mud removal efficiency E. The setting module is used to set the damage layer depth. The determining module is used to determine the first damage factor based on the set damage layer depth and the first damage function f'.

[0138] This application also provides a thinning device, including a stage 100 and a grinding wheel 200. The stage 100 is used to support and rotate a silicon wafer W. The grinding wheel 200 is located above the stage 100, and the edge of the grinding wheel 200 has a plurality of grinding teeth 210 for grinding the silicon wafer W.

[0139] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0140] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A thinning method, characterized in that, It is used in a thinning device, which includes: A grinding wheel (200) has a plurality of grinding teeth (210) on its edge, the grinding teeth (210) being used for grinding silicon wafers (W); Establish grinding tooth quantity Grinding wheel speed Silicon wafer rotation speed and the pressure of the grinding wheel on the silicon wafer. The first damage function regarding the first damage factor The first damage factor refers to the depth of damage to the silicon wafer by the grinding wheel (200); Based on the set damage layer depth and the first damage function Determine the first damage factor; The first damage function for: in, This refers to the number of teeth being ground. This refers to the grinding wheel speed; The rotational speed of the silicon wafer; This refers to the pressure exerted by the grinding wheel on the silicon wafer. This is the coefficient representing the influence of the pressure exerted by the grinding wheel on the silicon wafer on the damage. This is the coefficient representing the influence of silicon wafer rotation speed on silicon wafer damage. The coefficient representing the influence of the number of grinding teeth on silicon wafer damage; This is the coefficient representing the influence of grinding wheel rotation speed on silicon wafer damage.

2. The thinning method according to claim 1, characterized in that, Also includes: Establish sludge removal efficiency Silica mud concentration The second damage function regarding the second damage factor ; The second damage factor refers to the depth of damage to silicon wafers caused by silica mud. Based on the set damage layer depth and the first damage function and the second damage function The first damage factor and the second damage factor are determined.

3. The thinning method according to claim 2, characterized in that, Also includes: Establish grinding tooth quantity Grinding fluid flow rate Silica mud concentration and grinding fluid pressure difference Regarding sludge removal efficiency The sludge discharge function; Based on the number of grinding teeth collected or acquired Grinding fluid flow rate Silica mud concentration Grinding fluid pressure difference And the sludge removal function to determine the sludge removal efficiency. .

4. The thinning method according to claim 3, characterized in that, The sludge removal function is: in, For equipment constants; This refers to the number of teeth being ground. This refers to the grinding fluid flow rate; The concentration of silica mud; For grinding fluid pressure difference; The coefficient representing the influence of grinding fluid flow rate on sludge removal efficiency; The coefficient representing the influence of grinding fluid pressure difference on mud removal efficiency; The coefficient representing the influence of the number of grinding teeth on the mud removal efficiency; The coefficient representing the influence of silica mud concentration on sludge discharge efficiency.

5. The thinning method according to claim 4, characterized in that, Also includes: Establish silicon wafer density silicon wafer area Silicon wafer thickness removal amount Grinding fluid flow rate and processing time Regarding silica mud concentration The concentration function; Based on the collected or acquired silicon wafer density silicon wafer area Silicon wafer thickness removal amount Grinding fluid flow rate Processing time And concentration functions to determine silica mud concentration .

6. The thinning method according to claim 5, characterized in that, The concentration function is: in, Silicon wafer density; The area of ​​the silicon wafer; This refers to the amount of silicon wafer thickness removed. This refers to the grinding fluid flow rate; This refers to the processing time.

7. The thinning method according to claim 6, characterized in that, The second damage function for: in, To improve sludge removal efficiency; The concentration of silica mud; The coefficient representing the influence of the silica mud ratio on the damage was not excluded. The coefficient representing the effect of silica mud concentration on damage; Based on the first damage function Second damage function Establish a third damage function The third damage function for: in, The damage weighting coefficient for the grinding wheel (200); This represents the weighting coefficient for silica mud damage. Based on the set damage layer depth and the third damage function Determine the number of grinding teeth Grinding wheel speed Silicon wafer rotation speed Pressure of the grinding wheel on the silicon wafer Grinding fluid flow rate .

8. A thinning system, characterized in that, The thinning system comprising the thinning method as described in any one of claims 1-7, wherein the thinning method includes: The function creation module is used to establish the number of grinding teeth. Grinding wheel speed Silicon wafer rotation speed Pressure of the grinding wheel on the silicon wafer and sludge removal efficiency The first damage function regarding the first damage factor ; The setting module is used to set the depth of the damage layer; and The determination module is used to determine the damage layer depth and the first damage function. The first damage factor is determined.

9. A thinning device, characterized in that, The thinning device, comprising the thinning method as described in any one of claims 1-7, includes: A stage (100) for carrying and rotating a silicon wafer (W); A grinding wheel (200) is located above the stage, and the edge of the grinding wheel (200) has a plurality of grinding teeth (210) for grinding silicon wafers (W).

Images