A simulation method for retaining ring structure in chemical mechanical polishing process

Abstract

The present invention provides a simulation method for a retaining ring structure in a chemical mechanical polishing process, comprising constructing a CFD model of the chemical mechanical polishing process, dividing a fluid region into discrete grids, wherein the grids are non-repetitive control volumes, and within each control volume, using a momentum equation, a continuity equation, and a VOF model to control the motion of the polishing liquid fluid, and a DPM model to control the motion of the abrasive, integrating the momentum equation, the continuity equation, and the VOF model for each control volume to obtain a discrete equation, solving the discrete equation to obtain simulation data, the simulation data including abrasive motion data, plotting the abrasive distribution and the abrasive aggregation state on the polishing pad and / or the wafer edge based on the abrasive motion data, and having the beneficial effects of visualizing the polishing liquid / abrasive flow through a finite element method, helping to understand the influence of the retaining ring on the polishing liquid flow, providing a retaining ring structure improvement to solve wafer edge over-polishing, conforming to actual production processes, and having high simulation accuracy.

Description

Technical Field

[0001] The present invention relates to a simulation method for a retaining ring structure in a chemical mechanical polishing process, and in particular to a simulation method for studying how the shape of a groove on a retaining ring affects the passage of abrasives through the groove and the overall distribution of the abrasives in a polishing liquid. Background Art

[0002] In the 1970s, multilayer metallization technology was introduced into integrated circuit manufacturing, effectively utilizing the vertical space within the chip and significantly increasing device integration. However, multilayer metallization exacerbated wafer surface unevenness during IC manufacturing, leading to a series of problems, such as uneven photoresist thickness and consequently photolithography limitations, which severely hampered the development of large-scale integrated circuits (LSIs). In the late 1980s, IBM developed chemical-mechanical planarization (CMP) technology, which significantly improved wafer flatness compared to traditional techniques. CMP has become the most critical planarization technology in LSI manufacturing.

[0003] CMP requires the use of dedicated polishing equipment, such as Figure 1 As shown, the wafer 13 is clamped by the retaining ring 12, and the retaining ring 12 is fixed on the back film 11 and the load shaft (not shown). The load shaft rotates to drive the wafer 12 and press the wafer 12 on the surface of the polishing pad 31 for grinding. The polishing liquid 2 is supplied to the surface of the polishing pad 31 at any time to corrode and soften the wafer 13. While softening the part near the surface of the wafer 13, the polishing pad 31 continuously grinds to remove the softened part. During this process, the polishing table 32 drives the polishing pad 31 to rotate, wherein the direction of rotation of the polishing pad 31 is generally the same as the direction of rotation of the wafer 12. The polishing liquid 2 contains abrasives and other chemical components, and the abrasives are squeezed, rolled, scraped and ground by the polishing pad 31 to grind the surface of the wafer 13.

[0004] From the perspective of material removal, during the CMP process, the abrasive in the polishing liquid reacts with the chemical components in the polishing liquid, and the softened surface of the wafer has come into contact with the abrasive. The softened layer is removed by mechanical force, reducing the height of the higher protrusions. By continuously repeating this process, a highly flat wafer surface can be obtained. Therefore, the abrasive in the polishing liquid is the main factor affecting the quality of wafer surface flatness.

[0005] However, during the CMP polishing process, the polishing liquid containing abrasives flows from the infusion tube above the polishing pad to the surface of the polishing pad. Under the rotation of the polishing pad and the sweeping of the adjustment disk, it is spread over the entire surface of the polishing pad. In order to ensure that the wafer is horizontally fixed and remains in a non-vibrating state during the polishing process, a retaining ring is usually used to clamp the wafer. The retaining ring is fixed to the periphery of the wafer, and the load pressure is generally higher than the wafer. The retaining ring has the following characteristics: Figure 2 The groove 121 shown is for the polishing liquid and abrasive to enter the wafer-polishing pad contact gap. If the groove setting of the retaining ring is unreasonable and blocks the polishing liquid from entering the wafer-polishing pad contact gap, it is easy to cause a series of problems, such as the groove width is too narrow, resulting in the abrasive being unable to enter the wafer area, resulting in a low removal rate; or causing the abrasive to gather at the edge of the wafer, thereby causing "edge over-polishing" 1-2 cm at the edge of the wafer, resulting in an uneven wafer surface.

[0006] At the same time, since the CMP equipment is in a closed state during operation to prevent external dust contamination, it is difficult to observe the flow of polishing liquid on the polishing pad and in the wafer-polishing pad contact gap in real time. Therefore, it is impossible to study the distribution and flow of polishing liquid when using different retaining ring structures in real time, and it is impossible to determine the distribution of abrasive between the wafer and the polishing pad.

[0007] Currently, there are methods to study the effect of wafer pressure on CMP, and there are also methods to study the effect of the distance between the retaining ring and the wafer on CMP. Both are divorced from the main factor of actual wafer grinding, which is the abrasive in the polishing liquid. Summary of the Invention

[0008] To solve the above technical problems, the present invention provides a simulation method for the retaining ring structure in the chemical mechanical polishing process. The finite element method is used to visualize the flow of polishing liquid / abrasive, which helps to understand the influence of the retaining ring on the flow of polishing liquid. The retaining ring structure is improved to solve the problem of over-polishing of wafer edges. It fits the actual production process and has high simulation accuracy.

[0009] The present invention provides a simulation method for a retaining ring structure in a chemical mechanical polishing process, comprising:

[0010] Step 100: constructing a polishing pad compression model, including a wafer and a polishing pad, wherein the wafer applies a load to the polishing pad, and the deformation of the polishing pad after compression is d;

[0011] Step 200: Constructing a CFD model of the chemical mechanical polishing process, the CFD model including:

[0012] at least one of the size, rotation speed, porosity, porosity, elastic modulus, Poisson's ratio, and density of the polishing pad;

[0013] at least one of the size, rotation speed, elastic modulus, Poisson's ratio, and density of the wafer;

[0014] at least one of the flow rate, viscosity, and density of the polishing liquid, the fluid area of ​​the polishing liquid, and the size of the abrasive in the polishing liquid;

[0015] The shape of the retaining ring, including the distribution and shape of the grooves on the surface of the retaining ring;

[0016] It also includes steps 210 to 250,

[0017] Step 210: Divide the fluid region into discrete grids, where the grids are non-repeating control volumes.

[0018] Step 220: In each of the control volumes, the momentum equation, continuity equation and VOF model are used to control the movement of the polishing liquid, and the DPM model is used to control the movement of the abrasive.

[0019] The momentum equation, the continuity equation and the VOF model are integrated for each control volume to obtain a discrete equation, and the discrete equation is solved to obtain simulation data, wherein the simulation data includes abrasive motion data.

[0020] Step 230: plotting abrasive distribution based on the abrasive motion data,

[0021] Step 240: Observe the abrasive aggregation state on the polishing pad and / or wafer edge. If the aggregation state improves, execute step 300. If the aggregation state does not improve, execute step 250 and then execute steps 210 to 240 in sequence.

[0022] Step 250: improving the retaining ring;

[0023] Step 300: Output the result.

[0024] The continuity equation uses formula (4), and the momentum equation uses formula (5) to control the movement of the polishing liquid in the flow area.

[0025]

[0026] Wherein, ρ is the density of the abrasive in the control volume, t is the flow time of the polishing liquid, is the flow velocity of the polishing liquid, S m is the mass of the abrasive removed from the dispersed phase and added to the continuous phase;

[0027]

[0028] Wherein, p is the static pressure of the polishing liquid, is the stress tensor of the polishing solution, and It is the attraction and external force generated by the interaction of the dispersed phase of the abrasive.

[0029] The VOF model equation (6) solves the momentum equation and the volume fraction of the fluid in the fluid region to simulate the mass transfer of at least two immiscible fluids, including at least the polishing liquid phase and the air gas phase, to calculate the free surface height of the polishing liquid.

[0030]

[0031] in, is the mass transfer from one q phase to another p phase, is the mass transfer from one phase p to another phase q, and α is the volume fraction of each phase;

[0032] The constraint equation adopts formula (7) to solve the volume fraction of multiple immiscible fluids.

[0033]

[0034] The DPM equation uses formula (8) to simulate the relationship between the abrasive and the polishing liquid to obtain the abrasive motion data.

[0035]

[0036] Among them, m p is the mass of the abrasive, is the polishing liquid fluid phase velocity, is the velocity of the abrasive, ρ is the polishing liquid density, is the particle density of the abrasive, is the additional force on the abrasive, is the drag force on the abrasive, and τr is the relaxation time between the polishing liquid and the abrasive.

[0037] Preferably, in step 100, the compression model includes the shape of the wafer and the shape of the polishing pad.

[0038] Preferably, the minimum potential energy equation uses formula (1) to calculate the deformation variable d,

[0039] П=Up-Vp (1)

[0040] Where π is the total potential energy, U p is the total strain energy in the system, V p is the work done by the load on the polishing pad,

[0041] U p Using formula (2),

[0042]

[0043] Where π is the total potential energy, Up is the total strain energy in the system, V p is the work done by the load on the polishing pad,

[0044] U p Using formula (2),

[0045]

[0046] Where {ε} is the row vector of the polishing pad strain, {σ} is the row vector of the stress on the polishing pad, V is the volume of the polishing pad,

[0047] V p Using formula (3):

[0048] Vp=∫∫∫ V {d} T {Fb}dV+∫∫ s {d} T {Td}dS (3)

[0050] Wherein, {Fb} is the body force on the polishing pad, {Td} is the surface load on the contact surface of the polishing pad, and S is the area on the polishing pad to which the load is applied.

[0051] Preferably, in step 210, the fluid region includes a fluid region thickness and a polishing gap, the polishing gap being the distance between the polishing pad and the wafer in the chemical mechanical polishing process, and the discrete grid is the fluid region thickness divided into at least 10 layers from bottom to top, and the thickness of each layer is no more than one quarter of the polishing gap.

[0052] Preferably, in step 230, the abrasive distribution is rendered in color according to the modulus concentration, and the abrasive track is a track in which the abrasive is continuously marked at positions within equal steps.

[0053] Preferably, the simulation data further includes:

[0054] Steady-state / transient pressure distribution at various locations of the fluid region, the wafer surface, and the polishing pad surface,

[0055] Velocity distribution and shear force distribution of the polishing liquid and the abrasive;

[0056] The abrasive passes through the movement trajectory of the groove.

[0057] Preferably, improvement means that the abrasive is more evenly distributed than before at least on the portion of the wafer projected onto the polishing pad.

[0058] The present invention provides a simulation method for a retaining ring structure in a chemical mechanical polishing process, which has the beneficial effect of visualizing the polishing liquid / abrasive flow through a finite element method, helping to understand the influence of the retaining ring on the polishing liquid flow, and providing an improvement to the retaining ring structure to solve the problem of wafer edge over-polishing. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Attachment Figure 1 It is a schematic diagram of chemical mechanical polishing process equipment;

[0060] Attachment Figure 2 (a) to (d) are respectively a side view, a perspective view, a bottom view and an enlarged view of part A in the perspective view of the prior art retaining ring;

[0061] Attachment Figure 3 This is a flow chart of a process simulation method for a retaining ring in chemical mechanical polishing according to the present invention;

[0062] Attachment Figure 4 is a schematic diagram of polishing pad deformation;

[0063] Attachment Figure 5 This is a schematic diagram of the discrete grid of the polishing liquid fluid area;

[0064] Attachment Figure 6 This is a simulation diagram of the abrasive passing through the retaining ring groove;

[0065] Attachment Figure 7 It is a rendering diagram of abrasive distribution;

[0066] Attachment Figure 8 It is a retaining ring structure based on simulation of the present invention;

[0067] Attachment Figure 9 (a) to (d) are Figure 8 Enlarged views of the first to fourth designs in Part B;

[0068] Attachment Figure 10 yes Figure 8 A magnified view of the fifth design in Part B;

[0069] Attachment Figure 11 (a) to (b) are Figure 8 Enlarged views of the sixth and seventh designs in Part B;

[0070] Attachment Figure 12 It is a two-dimensional finite element model of the polishing pad, wafer, and retaining ring structure;

[0071] Attachment Figure 13 (a) to (e) are simulation diagrams of abrasive distribution in the wafer area of ​​the retaining ring with different groove-ring area ratios at the end of 5s under given conditions;

[0072] Attachment Figure 14This is the distribution diagram of polishing removal amount of retaining ring wafer with different groove-ring area ratios. DETAILED DESCRIPTION

[0073] The following describes in detail a specific embodiment of a simulation method for a retaining ring structure in a chemical mechanical polishing process of the present invention with reference to the accompanying drawings.

[0074] In the accompanying drawings, for the sake of convenience of description, the size ratios of layers and regions are not actual ratios. When a layer (or film) is referred to as being "on" another layer or substrate, it can be directly on the other layer or substrate, or there can also be an intermediate layer. In addition, when a layer is referred to as being "under" another layer, it can be directly below, and there can also be one or more intermediate layers. In addition, when a layer is referred to as being between two layers, it can be the only layer between the two layers, or there can also be one or more intermediate layers. The same reference numerals always represent the same elements. In addition, when two components are referred to as being "connected", this includes a physical connection, unless expressly defined in the specification, and such physical connection includes but is not limited to electrical connection, contact connection, and wireless signal connection.

[0075] The present invention provides a simulation method for retaining ring structure in chemical mechanical polishing process, such as Figure 3 Shown, including:

[0076] Step 100: Figure 4 As shown, a polishing pad compression model is constructed, including a wafer 13 and a polishing pad 31, wherein the wafer 13 applies a load, i.e., a pressure, to the polishing pad 31; and the shapes of the wafer 13 and the polishing pad 31 are established. The dimensions and physical properties of the wafer 13 and the polishing pad 31 are input. The pressure applied to the wafer 13 is input, and the compressed thickness of the polishing pad 31 is simulated. Through the stress-strain relationship, the deformation 311 of the polishing pad after being compressed is obtained as △h, i.e., the polishing pad deformation d below. The height of the rough peak on the surface of the physical polishing pad 31 is h0, and h0-△h can be used to obtain the film thickness h of the polishing liquid 2 in the gap between the wafer 13 and the polishing pad 31. It should be noted that the same is applicable to the compression of the retaining ring 12-polishing pad 31, which is used to simulate the gap between the retaining ring 12-polishing pad 31.

[0077] In this embodiment, specifically, the minimum potential energy equation (1) is used to calculate the deformation amount d of the polishing pad,

[0078] ∏=Up-Vp (1)

[0079] Where ∏ is the total potential energy, U p is the total strain energy in the system, V p is the work done by the load on the polishing pad,

[0080] U p Formula is (2),

[0081]

[0082] Where {ε} is the strain row vector, {σ} is the stress row vector, V is the polishing pad volume,

[0083] V p Formula (3):

[0084] Vp=∫∫∫ V {d} T {Fb}dV+∫∫ s {d} T {Td}dS (3)

[0086] Wherein, d is the displacement of the load on the polishing pad, i.e., the deformation Δh of the polishing pad after being compressed as described above, {Fb} is the body force, {Td} is the surface load, and S is the area to which the load is applied.

[0087] Step 200: Constructing a CFD (Computational Fluid Dynamics) model of the chemical mechanical polishing process, the CFD model includes the size, rotation speed, porosity, porosity, elastic modulus, Poisson's ratio, and density of the polishing pad 31; the size, rotation speed, elastic modulus, Poisson's ratio, and density of the wafer 13; the flow rate, viscosity, and density of the polishing liquid 2, the fluid area of ​​the polishing liquid, and the size of the abrasive in the polishing liquid 2; the shape of the retaining ring 12, including the distribution and shape of the grooves 121 on the surface of the retaining ring 12; and further including steps 210 to 250.

[0088] Step 210: Figure 5 As shown, the polishing liquid 2 area is divided into discrete grids, and the grids are non-repeating control volumes. The fluid area includes the fluid area thickness and the polishing gap. The polishing gap is the distance between the polishing pad and the wafer in the chemical mechanical polishing process. The discrete grid is the fluid area thickness divided into at least 10 layers from bottom to top, and the thickness of each layer is not greater than one-quarter of the polishing gap. In the process of solving the CFD model, the momentum equation, continuity equation and VOF model can all be regarded as consisting of an infinite number of continuous function values ​​in the discrete grids within the model area, and these continuous function values ​​are replaced by the function values ​​of the discrete grids. In the solution process, the momentum equation, continuity equation and VOF model in the CFD model can be converted into discrete equations of the relationship between the function values ​​to be solved in the discrete grids. The function values ​​of the discrete grids can be obtained by solving the discrete equations.

[0089] In this embodiment, in step 210, the distance between the polishing pad 31 and the wafer 13 is generally 10-100 microns. To ensure that dramatic changes in parameters near the polishing pad 31 and the wafer 13 surface are reflected, the number of grid layers is guaranteed to be 4 or more, and the grid size is less than 1 / 4 of the gap between the polishing pad 31 and the wafer 13. The thickness of the fluid region is larger than the polishing gap, ensuring that the number of grid layers is 10 or more.

[0090] Step 220: In each of the control volumes, the momentum equation, continuity equation and DPM model are used to control the fluid motion of the polishing liquid 2, and the finite volume method is used to calculate the simulation data, including the abrasive motion data. The basic idea of ​​the finite volume method is: divide the calculation domain into a series of non-repeating control volumes, and make a control volume around each grid point in the discrete grid obtained in step 210; integrate the differential equation to be solved for each control volume to obtain a set of discrete equations. The physical meaning of the discrete equation is the conservation principle of the dependent variable in a control volume of finite size, just like the differential equation represents the conservation principle of the dependent variable in an infinitesimal control volume. The discrete equation obtained by the finite volume method requires that the integral conservation of the dependent variable is satisfied for any set of control volumes and for the entire calculation area.

[0091] In this embodiment, the continuity equation (4) and momentum equation (5) are used to control the movement of the polishing liquid 2 in the polishing area.

[0092]

[0093] Wherein, ρ is the density of the abrasive in the control volume, t is the flow time of the polishing liquid, is the flow velocity of the polishing liquid, S m is the mass of the abrasive removed from the dispersed phase and added to the continuous phase;

[0094]

[0095] Wherein, p is the static pressure of the polishing liquid, is the stress tensor of the polishing solution, and It is the attraction and external force generated by the interaction of the dispersed phase of the abrasive.

[0096] The momentum equation and the volume fraction of the fluid in the fluid region are solved using the VOF model equation (6) to simulate at least two immiscible fluids, generally liquid and gas phases, to solve the free surface height of the polishing liquid.

[0097]

[0098] in, is the mass transfer from one q phase to another p phase, is the mass transfer from one phase p to another phase q, α is the volume fraction;

[0099] According to the constraint equation (7), the volume fractions of various immiscible fluids are solved.

[0100]

[0101] According to the DPM equation (8), the relationship between the abrasive and the polishing liquid 2 is simulated to obtain the simulation data.

[0102]

[0103] Among them, m p is the mass of the abrasive, is the polishing liquid fluid phase velocity, is the velocity of the abrasive, ρ is the polishing liquid density, is the particle density of the abrasive, is the additional force on the abrasive, is the drag force on the abrasive, and τr is the relaxation time between the polishing liquid and the abrasive.

[0104] Step 230: Draw the abrasive distribution according to the abrasive motion data. The abrasive distribution is rendered in color according to the modulus concentration. The abrasive trajectory is a trajectory generated by the abrasive that is continuously marked at positions within equal steps.

[0105] In this embodiment, according to the obtained simulation results, Figure 7 As shown, the abrasive distribution is rendered in different colors / shades, and the position continuity of the abrasive within the same step length is derived as a particle motion trajectory (not shown). For step 230, the obtained data also includes: steady-state / transient pressure distribution of the fluid region, the surface of the wafer 13 and the polishing pad 31 at various positions, the velocity distribution and shear force distribution of the polishing liquid 2 and the abrasive; and also includes Figure 6 As shown, the abrasive passes through the movement trajectory of the groove 121, wherein the retaining ring 12 rotates counterclockwise, as shown in FIG. Figure 6 As shown, the abrasive in the polishing liquid 2 entering the groove 12 from the outside will slide out of the groove 12 due to hitting the rounded part, while the abrasive entering from the middle part will smoothly enter the polishing gap. This simulation result will help to improve the groove 121 of the retaining ring 12 in step 250.

[0106] Step 240: Figure 7As shown, observe the aggregation state of the abrasive on the polishing pad and / or the edge of the wafer. If the aggregation state improves, execute step 300. Improvement means that the abrasive is more evenly distributed than before at least on the portion of the wafer projected on the polishing pad. If the aggregation state does not improve, execute step 250 and then execute steps 210 to 240 in sequence.

[0107] Step 250: Improve the retaining ring. Based on the simulation results from step 240, optimize the structure and shape of the retaining ring 12. Repeat this process to find the optimal retaining ring structure 12 and process parameters. For the first iteration, parameters such as the shape of the retaining ring 12 and the distribution and shape of the grooves 121 on the surface of the retaining ring 12 may not be included in step 200 to obtain zero-baseline control group simulation data. Subsequent improvements to the retaining ring 12 can be made by adjusting the number, shape, size, inclination angle, presence of steps, and groove angle shape of the grooves.

[0108] Step 300: Output the results; output the abrasive distribution, abrasive motion trajectory, fluid area pressure distribution, and fluid flow rate distribution data corresponding to various retaining ring 12 shapes.

[0109] The present invention provides a simulation method for a retaining ring structure in a chemical mechanical polishing process, which has beneficial effects: the finite element method is used to visualize the flow of polishing liquid / abrasive, which helps to understand the influence of the retaining ring on the flow of polishing liquid, provides an improvement to the retaining ring structure to solve the problem of over-polishing of wafer edges, fits the actual production process, and has high simulation accuracy.

[0110] According to the above simulation method of the retaining ring structure in the chemical mechanical polishing process, the present invention proposes a retaining ring 12 used in chemical mechanical polishing, such as Figure 8 As shown, it comprises an annular body (not shown), the annular body comprising a top surface (not shown) and a bottom surface (not shown), the top surface being connected to a load device (not shown). Figures 9-11 As shown, the lower surface is provided with a combination groove (not shown), which includes through grooves 1211 / 1212 pointing toward the interior of the annular body and extending through the lower surface, and a groove cavity 1210 located on the inner side of the annular body and communicating with the through groove 1211 but not extending through / or partially extending through the lower surface. It should be noted that both the through grooves 1211 / 1212 and the groove cavity 1210 are open on the lower surface. The combination groove is used for the inlet and outlet of the polishing liquid 2.

[0111] like Figures 9-11As shown, the ratio of the opening area of ​​the groove cavity and the through groove on the inner side of the annular body is in the range of 5 to 30, that is, the groove cavity 1210 has a larger accommodating space than the through groove 1211 / 1212. On the one hand, it avoids the retaining ring 12 from clamping the wafer 13 in this part, hindering the movement of the abrasive and causing it to gather at the boundary between the retaining ring 12 and the wafer 13, thereby causing the edge of the wafer 13 to be over-polished. At the same time, after the shape of the groove cavity 1210 is improved, a single combined groove can be realized to simultaneously allow new polishing liquid 2 to enter the polishing gap inside the annular body and allow the polishing liquid 2 containing wafer waste to be discharged from the polishing gap.

[0112] It should be noted that the ratio of the opening area of ​​the groove cavity to the through groove on the inner side of the annular body is in the range of 5-30. Alternatively, the ratio of the volume of the groove cavity to the through groove in the inner cavity of the annular body is in the range of 5-30.

[0113] It should be further noted that the ratio of the opening area of ​​the groove cavity to the through groove on the inner side of the annular body is 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30. Alternatively, the ratio of the volume of the groove cavity to the through groove is 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30.

[0114] like Figure 9 As shown in (a) to (d), the through grooves have two 1211 and 1212, one of which is arranged on one side of the groove cavity 1210. This combination of grooves is conducive to discharging the polishing liquid 2 containing wafer waste.

[0115] like Figure 9 As shown in (a), the groove cavity 1210 and the through groove 1211 have different depths on the lower surface, that is, a step is formed between the through groove 1211 and the groove cavity 1210. This combination groove is conducive to discharging the polishing liquid 2 containing wafer waste.

[0116] As shown in Figure 9(c), a combined groove includes both a step formed between the through groove 1212 and the groove cavity 1210, and a through groove 1211 with the same depth as the groove cavity 1210. The pressure in the groove cavity 1210 helps the single combined groove to simultaneously allow new polishing liquid 2 to enter the polishing gap inside the annular body through the through groove 1211, and allow the polishing liquid 2 containing wafer waste to be discharged from the polishing gap through the through groove 1212.

[0117] As shown in Figure 9(b), the through grooves 1211 and 1212 have the same depth as the groove cavity 1210. The opening of the through groove 1212 on the outside of the annular body is trumpet-shaped, and the inside is a uniformly sized channel, while the opening of the through groove 1211 on the outside of the annular body is a uniformly sized channel and the inside is trumpet-shaped. That is, the portion of the groove cavity 1210 connecting to the through groove 1211 near the outside of the annular body is an arc-shaped portion that bends toward the inside of the annular body. The opening area of ​​the through groove 1211 on the outside of the annular body is smaller than the area of ​​the arc-shaped portion formed by the through groove 1211 and the groove cavity 1210, so that the polishing liquid 2 containing wafer waste is discharged from the through groove 1211. The opening of the through groove 1212 is configured such that one groove edge gradually opens away from the other groove edge from the outer quarter of the annular body, so that the polishing liquid 2 is injected into the polishing gap from the through groove 1212. Thus, a single combined tank can not only update the polishing liquid 2 in the polishing gap, but also discharge the waste liquid containing wafer waste chips.

[0118] like Figure 9 (b) and Figure 9 As shown in (d), the bottom of the groove cavity 1210 can be configured as a uniform, smooth, or unevenly curved surface along the annular body. For example, the edge of the groove cavity 1210 corresponding to the through groove 1211 is a smooth curved surface that evenly transitions to the vertical edge of the groove cavity 1210 corresponding to the through groove 1212, thereby enabling the polishing liquid 2 to be injected into the polishing gap from the through groove 1211 and the waste liquid containing wafer scraps to be discharged from the through groove 1212.

[0119] Through simulation, it is found that as long as the groove cavity 1210 is a closed space relative to the polishing pad 31, there is a certain probability that abrasives will accumulate and cause the edge of the wafer 13 to be over-polished. Figure 9 (b) Further improvements based on the combined slot design, such as Figure 10 As shown, the depth of the groove cavity 1210 opening on the outer side of the annular body is less than half the depth of the through groove 1211 to avoid over-polishing of the edge portion of the wafer 13 corresponding to the groove cavity 1210.

[0120] like Figure 9 As shown in (a) to (d), the combined grooves all include two through grooves 1211 and 1212, both located on both sides of the groove cavity 1210 and connected to the latter. The applicant found through simulation that for the through grooves that need to allow the polishing liquid 2 to enter the polishing gap, Figure 9 The through groove 1212 in (b) extends to the edge of the polishing gap, which will hinder the polishing liquid 2 from being quickly and evenly distributed in the polishing gap. Figure 9 (b) Further improvements based on the combined slot design, such as Figure 11As shown in (b), there are two through grooves 1211 and 1212, one of which is configured to open into the groove cavity 1210, thereby helping the polishing liquid 2 to drive the abrasive to be quickly and evenly distributed in the polishing gap.

[0121] Similarly, if Figure 11 As shown in (a), the applicant has designed a new combination groove, wherein the groove cavity 1210 and the through groove 1211 have different depths on the lower surface, i.e., a step is formed between the through groove 1211 and the groove cavity 1210. This combination groove is conducive to discharging the polishing liquid 2 containing wafer waste, and the through groove 1212 is configured to open to the groove cavity 1210, thereby helping the polishing liquid 2 to drive the abrasive to be quickly and evenly distributed in the polishing gap.

[0122] Through simulation, the applicant found that the number of the above-mentioned combination slots provided on the retaining ring 12 ranges from 8 to 40. Preferably, the applicant sets the number of combination slots to 8, 13, 18, 23, 28, 36, or 40. Of course, the number of combination slots can also be set to any integer in the range of 8 to 40. It should be noted that, under the premise of satisfying the same functional design of the combination slots on the retaining ring, the number of combination slots can also be 3, 4, 5, 6, or 7.

[0123] Based on the simulation method of the retaining ring structure in the chemical mechanical polishing process, the present invention designs a new retaining ring structure according to the simulation data, which has the technical effects of directionally introducing abrasives contained in the polishing liquid into the polishing gap and directionally removing the polishing liquid containing wafer waste from the polishing gap, thereby avoiding over-polishing of the wafer edge.

[0124] According to the simulation method of the retaining ring structure in the above chemical mechanical polishing process, from the perspective of retaining ring manufacturing, only uniform through grooves 1211 (such as Figure 13 (a) to (e)) instead of using a combined groove consisting of through grooves 1211 / 1212 and groove cavity 1210 (as shown in Figures 8 to 11 The applicant conducted a series of simulations with the cutting edge of the retaining ring 12 improved by further adjusting the groove-ring area ratio.

[0125] The groove-ring area ratio is the ratio of the total area of ​​the through grooves 1211 to the surface area of ​​the annular body of the retaining ring 12 (i.e., the ring area); when the through grooves 1211 have the same shape, the groove-ring area ratio can also be expressed as the area of ​​a single through groove 1211 (e.g., Figure 13 (a) to (e)) area and number thereof, a design is adopted in which the through grooves 1211 with uniform width are distributed on the retaining ring 12 in the same regular pattern. The groove-ring area ratio can also be expressed as the number of through grooves*through groove width.

[0126] like Figures 12 to 14As shown, the applicant established a two-dimensional finite element model of the polishing pad 31, wafer 13, and retaining ring 12 structure, wherein d1 is the outer diameter of the annular body of the retaining ring 12, which is 348 mm, d2 is the inner diameter of the annular body of the retaining ring 12, which is 301 mm, and D is the size of the polishing pad 31, which is 850 mm.

[0127] The motion of the polishing liquid 2 is controlled by the continuity equation and the momentum equation, where

[0128] The continuity equation (9) is:

[0129]

[0130] The momentum equations (10) and (11) are:

[0131]

[0132]

[0133] Where u is the fluid velocity of the polishing liquid 2, p is the fluid pressure of the polishing liquid 2, ρ is the fluid density of the polishing liquid 2, μ is the fluid dynamic viscosity of the polishing liquid 2, and μ T is the eddy viscosity coefficient, and F is the additional force acting on the fluid of the polishing liquid 2.

[0134] The flow rate of the polishing liquid 2 is calculated based on the k-epsilon equation. The abrasive is completely immersed in the polishing liquid 2 and is affected by the drag force between the abrasive and the polishing liquid 2. The movement of the abrasive in the flow field of the polishing liquid 2 can be expressed as equations (12)(13):

[0135]

[0136]

[0137] Among them F D is the drag force on the abrasive, u, v, m p Represent the polishing liquid 2 speed, abrasive speed and abrasive mass ρ respectively p is the abrasive density, d p is the abrasive diameter and μ is the fluid dynamic viscosity of the polishing liquid 2.

[0138] The simulation shows that the abrasive in the polishing liquid 2 reaches a stable state at the end of 5 seconds when the rotation speed ratio of the polishing pad 31 to the wafer 13 is 120 rpm·min-1 / 120 rpm·min-1.

[0139] As shown in Table 1 and Figure 14As shown, the width of the through grooves 1211 is uniform. Taking the retaining ring 12 with 18 3mm wide through grooves 1211 as a control, the wafer polishing material removal amount of the retaining ring 12 is analyzed. Within a certain range of groove-ring area ratio, the removal amount is proportional to the groove-ring area ratio. Among the 36 13mm wide through grooves 1211, the groove-ring area ratio is 71.82%, and the material removal amount reaches the highest.

[0140] Table 1 Ratio of average removal of retaining rings with different structures to 18*3mm retaining ring

[0141]

[0142]

[0143] At the same time, the applicant found that within a certain range of the groove-ring area ratio, the increase in the amount of removal is not proportional to the increase in the groove-ring area ratio, and there is a trend that the increase in the amount of removal gradually decreases, such as Figure 13 (a) to (e) show the simulated distribution of abrasive in the wafer 13 region in the retaining ring 12 with different groove-ring area ratios at the end of 5s, where the groove-ring area ratios of the retaining ring 12 are (a) 8.69% and (b) 8.69% respectively.

[0144] (b) 19.99% (c) 35.91% (d) 41.22% (e) 71.82%. When the groove-to-ring area ratio is 71.82%, the abrasive distribution in the wafer 13 region is essentially the same as the abrasive distribution outside the wafer 13. The applicant also simulated the abrasive distribution (not shown) and the abrasive removal amount distribution for retaining rings 12 with number*through grooves 1211 widths of 39*13mm, 42*13mm, and 45*13mm, respectively. The simulations were similar to the retaining ring 12 with number*through grooves 1211 widths of 36*13mm, i.e., a groove-to-ring area ratio of 71.82%.

[0145] Therefore, a groove-to-ring area ratio of 60% to 75% for retaining ring 12 is optimal. Within this range, the amount of removal changes minimally. A groove-to-ring area ratio of 71.82% is the optimal design. It should be noted that the boundary values ​​of 60% and 75% within this range are integer values ​​obtained by the applicant by adjusting the orientation of through-groove 1211. As the orientation of through-groove 1211 deviates from the center of retaining ring 12, the groove-to-ring area ratio gradually increases, and vice versa. The applicant will not elaborate further here.

[0146] When the groove-ring area ratio is the same, Figure 13 The single through groove 1211 on the retaining ring 12 in (a) to (e) can also be formed as follows Figures 8 to 11 The applicant will not elaborate on the shape design of the through grooves 1211 / 1212 in (b) here.

[0147] The present invention provides a simulation method for a retaining ring structure in a chemical mechanical polishing process, and designs a retaining ring structure with a specific groove-ring area ratio according to simulation data, which has the technical effect of high grinding removal.

[0148] The above is only a preferred embodiment of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and modifications without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for simulating a retaining ring structure in a chemical mechanical polishing process, comprising: Step 100: constructing a polishing pad compression model, including a wafer and a polishing pad, wherein the wafer applies a load to the polishing pad, and the deformation of the polishing pad after compression is d; Step 200: Constructing a CFD model of the chemical mechanical polishing process, the CFD model including: at least one of the size, rotation speed, porosity, porosity, elastic modulus, Poisson's ratio, and density of the polishing pad; at least one of the size, rotation speed, elastic modulus, Poisson's ratio, and density of the wafer; at least one of the flow rate, viscosity, and density of the polishing liquid, the fluid area of ​​the polishing liquid, and the size of the abrasive in the polishing liquid; The shape of the retaining ring, including the distribution and shape of the grooves on the surface of the retaining ring; It also includes steps 210 to 250, Step 210: Divide the fluid region into discrete grids, where the grids are non-repeating control volumes. Step 220: In each of the control volumes, the momentum equation, continuity equation and VOF model are used to control the movement of the polishing liquid, and the DPM model is used to control the movement of the abrasive. The momentum equation, the continuity equation and the VOF model are integrated for each control volume to obtain a discrete equation, and the discrete equation is solved to obtain simulation data, wherein the simulation data includes abrasive motion data. Step 230: plotting abrasive distribution based on the abrasive motion data, Step 240: Observe the abrasive aggregation state on the polishing pad and / or wafer edge. If the aggregation state improves, execute step 300. If the aggregation state does not improve, execute step 250 and then execute steps 210 to 240 in sequence. Step 250: improving the retaining ring; Step 300: Output the result; The continuity equation uses formula (4), and the momentum equation uses formula (5) to control the movement of the polishing liquid in the fluid region. Wherein, ρ is the density of the abrasive in the control volume, t is the flow time of the polishing liquid, u is the flow velocity of the polishing liquid, S m is the mass of the abrasive removed from the dispersed phase and added to the continuous phase; Wherein, p is the static pressure of the polishing liquid, is the stress tensor of the polishing solution, and It is the attraction and external force generated by the interaction of the dispersed phase of the abrasive.

2. The method according to claim 1, characterized in that The VOF model equation (6) solves the momentum equation and the volume fraction of the fluid in the fluid region to simulate the mass transfer of at least two immiscible fluids, including at least the polishing liquid phase and the air gas phase, to calculate the free surface height of the polishing liquid. in, is the mass transfer from one q phase to another p phase, is the mass transfer from one phase p to another phase q, and α is the volume fraction of each phase; The constraint equation adopts formula (7) to solve the volume fraction of multiple immiscible fluids.

3. The method according to claim 1, characterized in that The DPM equation uses formula (8) to simulate the relationship between the abrasive and the polishing liquid to obtain the abrasive motion data. Among them, m p is the mass of the abrasive, is the polishing liquid fluid phase velocity, is the velocity of the abrasive, ρ is the polishing liquid density, is the particle density of the abrasive, is the additional force on the abrasive, is the drag force on the abrasive, and τr is the relaxation time between the polishing liquid and the abrasive.

4. The method according to claim 1, characterized in that In step 100, a compression model is formed including the shape of the wafer and the shape of the polishing pad.

5. The method according to claim 4, characterized in that: The minimum potential energy equation uses formula (1) to calculate the deformation variable d, Π=Up-Vp (1) Where π is the total potential energy, U p is the total strain energy in the system, V p is the work done by the load on the polishing pad, U p Using formula (2), Where {ε} is the row vector of the polishing pad strain, {σ} is the row vector of the stress on the polishing pad, V is the volume of the polishing pad, and V p Using formula (3): Vp=∫∫∫ V {d} T {Fb}dV+∫∫ s {d} T {Td}dS (3) Wherein, {Fb} is the body force on the polishing pad, {Td} is the surface load on the contact surface of the polishing pad, and S is the area on the polishing pad to which the load is applied.

6. The method according to claim 1, characterized in that In step 210, the fluid region includes a fluid region thickness and a polishing gap, wherein the polishing gap is the distance between the polishing pad and the wafer in the chemical mechanical polishing process, and the discrete grid is the fluid region thickness divided into at least 10 layers from bottom to top, and the thickness of each layer is no more than one quarter of the polishing gap.

7. The method according to claim 1, characterized in that: In step 230 , the abrasive distribution is rendered using colors according to the modulus concentration, and the abrasive trajectory is a trajectory in which the abrasive is continuously marked at positions within a constant step length.

8. The method according to claim 1, characterized in that: The simulation data also includes: Steady-state / transient pressure distributions of the fluid region, the wafer surface, and various locations on the polishing pad surface, and velocity and shear force distributions of the polishing liquid and the abrasive; The abrasive passes through the movement trajectory of the groove.

9. The method according to claim 1, characterized in that: In step 240 , improvement means that the abrasive is more evenly distributed than before at least on the portion of the wafer projected onto the polishing pad.

Images